doc.anet.be Page | 1 Faculty of Sciences Department of Biology Unveiling the evolutionary history &...

Page | 1

Faculty of Sciences

Department of Biology

Unveiling the evolutionary history & molecular epidemiology of

Mycobacterium ulcerans

De onthulling van de evolutionaire geschiedenis en de moleculaire epidemiologie van

Mycobacterium ulcerans

Proefschrift voorgelegd tot het behalen van de graad van

Doctor in de Wetenschappen: Biologie

aan de Universiteit Antwerpen te verdedigen door

Koen Vandelannoote

Promotoren: Antwerpen, 2016

Prof. dr. Bouke de Jong

Prof. dr. Herwig Leirs

Co-Promotor:

Prof. dr. Dissou Affolabi

Page | 2

Front cover image: Koen Vandelannoote Back cover image: Sophie Gryseels

Copyright © 2016 Universiteit Antwerpen - Faculteit Wetenschappen

Self-published by Koen Vandelannoote, Lode de Boningestraat 12, 2650 Edegem

(België).

All rights reserved. No part of this publication may be reproduced in any form by

print, photoprint, microfilm, electronic or any other means without written

permission from the publisher.

Alle rechten voorbehouden. Niets uit deze uitgave mag worden vermenigvuldigd

en/of openbaar gemaakt worden door middel van druk, fotokopie, microfilm,

elektronisch of op welke andere wijze ook zonder voorafgaande schriftelijke

toestemming van de uitgever.

Page | 3

Para Luquinha

Page | 4

Page | 5

Members of the Jury

Chairperson:

Prof. dr. Han Asard University of Antwerp (Belgium)

Secretary:

Prof. dr. Surbhi Malhotra Kumar University of Antwerp (Belgium)

Promotors:

Prof. dr. Bouke de Jong Institute of Tropical Medicine (Belgium)

Prof. dr. Herwig Leirs University of Antwerp (Belgium)

Co-Promotor:

Prof. dr. Dissou Affolabi University of Abomey-Calavi (Benin)

Members:

Prof. dr. Philippe Lemey University of Leuven (Belgium)

Prof. dr. Philip Supply CNRS & Genoscreen (France)

Em. Prof. dr. Françoise Portaels Institute of Tropical Medicine (Belgium)

Page | 6

Page | 7

Contents

Abbreviations ..................................................................................................... 9

Summary ........................................................................................................... 11

Samenvatting .................................................................................................... 13

Chapter 1 .......................................................................................................... 15

1.1 Buruli ulcer ......................................................................................... 16

1.2 M. ulcerans genomic characteristics and evolutionary origin ........... 19

1.3 Etiology .............................................................................................. 21

1.4 Reservoir(s) and mode(s) of transmission ......................................... 22

1.5 BU transmission in Australia .............................................................. 25

1.6 Molecular phylogenetics and epidemiology ..................................... 27

1.7 Research Goals: breaking the Buruli enigma ..................................... 28

Chapter 2 .......................................................................................................... 31

2.1 Abstract .............................................................................................. 32

2.2 Introduction ....................................................................................... 32

2.3 Material and Methods ....................................................................... 35

2.4 Results ................................................................................................ 44

2.5 Discussion .......................................................................................... 52

2.6 Acknowledgements ........................................................................... 56

Chapter 3 .......................................................................................................... 57

3.1 Abstract .............................................................................................. 58

3.2 Introduction ....................................................................................... 59

3.3 Methods ............................................................................................. 61

3.4 Results ................................................................................................ 65

3.5 Discussion .......................................................................................... 70

3.6 Supporting Information ..................................................................... 72

3.7 Acknowledgments ............................................................................. 72

3.8 Funding Statement ............................................................................ 72

3.9 Data Availability ................................................................................. 73

Chapter 4 .......................................................................................................... 75

4.1 Abstract .............................................................................................. 76

4.2 Introduction ....................................................................................... 76

4.3 Methods ............................................................................................. 78

Page | 8

4.4 Results & Discussion ........................................................................... 84

4.5 Conclusion .......................................................................................... 95

4.6 Supporting Information ...................................................................... 96

4.7 Data access ....................................................................................... 107

4.8 Acknowledgements .......................................................................... 107

Chapter 5......................................................................................................... 109

5.1 Abstract ............................................................................................ 110

5.2 Introduction ..................................................................................... 110

5.3 Materials and Methods .................................................................... 113

5.4 Results .............................................................................................. 118

5.5 Discussion ......................................................................................... 129

5.6 Supporting Information .................................................................... 132

5.7 Acknowledgements .......................................................................... 141

Chapter 6......................................................................................................... 143

6.1 Abstract ............................................................................................ 144

6.2 Author summary .............................................................................. 144

6.3 Introduction ..................................................................................... 145

6.4 Methods ........................................................................................... 146

6.5 Results .............................................................................................. 148

6.6 Discussion ......................................................................................... 153

6.7 Supporting Information .................................................................... 155

Chapter 7......................................................................................................... 157

Chapter 8......................................................................................................... 167

Acknowledgements ........................................................................................ 171

List of Publications .......................................................................................... 173

References ...................................................................................................... 175

Page | 9

Abbreviations

aDNA: ancient DNA

BAPS: Bayesian Analysis of genetic Population Structure

BU: Buruli ulcer

DRC: Democratic Republic of Congo

EBSP: Extended Bayesian Skyline Plot

ESS: Effective Sample Size

FAO: Food and Agriculture Organization

GTR: Generalized Time Reversible

HPD: Highest Posterior Density

IME: Institut Médical Evangélique

Indel: insertion / deletion

IS: Insertion Sequence

ISE-SNP: Insertion Sequence Element - Single Nucleotide Polymorphism

ITM: Institute of Tropical Medicine

LJ: Löwenstein Jensen

MRCA: Most Recent Common Ancestor

ML: Maximum Likelihood

MP: Maximum Parsimony

NGS: Next Generation Sequencing

NJ: Neighbor-Joining

PNG: Papua New Guinea

PS: Path Sampling

qPCR: Quantitative Polymerase Chain Reaction

RC: Republic of the Congo

SNPs: Single Nucleotide Polymorphisms

SRA: Sequence Read Archive

SRTM: Shuttle Radar Topography Mission

WGS: Whole Genome Sequencing

WHO: World Health Organization

Page | 10

Page | 11

Summary

Buruli ulcer is a neglected tropical disease of skin and subcutaneous tissue,

caused by infection with Mycobacterium ulcerans. Cases are reported around

the world, with rural wetlands of West- and Central African countries most

affected. Prevention and control of the disease is complicated by limited

understanding of the mode(s) of transmission of BU, probably involving

inoculation directly by a vector, or from a contaminated skin surface. While

direct human-to-human transmission does not occur, the potential

environmental reservoir(s), and the role of the human reservoir in sustaining

endemic BU, are controversial.

We redesigned a SNP-based (ISE-SNP) fingerprinting assay to gain

fundamental insights into the population structure and evolutionary history of

African M. ulcerans. Even though ISE-SNP genotyping resulted in higher

geographical resolution than previously achieved, the resolution was still too

limited to reconstruct evolutionary events on anything smaller than the

continental scale. Therefore we used state-of-the-art second and third

generation genome sequencing technologies and advanced statistical

approaches to understand the detailed evolutionary and spatio-temporal

history of African M. ulcerans. We also explored the use of the increased

resolution offered by low-cost genomics to distinguish disease relapses from

reinfections in patients with multiple BU episodes. African M. ulcerans was

found to be evolving entirely through clonal expansion and its genetic

diversity proved to be very restricted because of the pathogen’s slow

genome-wide substitution rate coupled with its relative recent origin.

Exploration of the genetic population structure revealed the existence of two

specific lineages within the African continent. We used temporal associations

and studied the past demographic history of M. ulcerans in a BU endemic

region to implicate the role of humans as a major reservoir in BU

transmission. In previous studies, M. ulcerans DNA has been detected in the

environment, possibly originating from BU patients with active, openly

discharging lesions. Transmission can then occur indirectly in the same

community water source, when the superficial skin surface of a naïve

Page | 12

individual is contaminated, and the bacilli present on the contaminated skin

are subsequently inoculated subcutaneously through some form of

penetrating (micro)trauma.

Our observations on the role of humans as maintenance reservoir to sustain

new BU infections suggest that interventions in a region aimed at reducing the

human BU burden will at the same time break the transmission chains within

that region. Active case-finding programs and the early treatment of pre-

ulcerative infections with specific antibiotics will decrease the amounts of

mycobacteria shed into the environment, which may ultimately reduce

disease transmission in Africa.

Page | 13

Samenvatting

Buruli ulcer (BU) is een verwaarloosde tropische ziekte veroorzaakt door

infectie met Mycobacterium ulcerans. De ziekte tast het huidoppervlak en de

onderhuidse weefsels aan. Hoewel er gevallen gemeld worden van over de

hele wereld zijn de rurale wetlands van West- en Centraal-Afrika het ergst

getroffen. Momenteel zijn BU preventie en -bestrijding gehinderd door een

beperkt begrip van de ziektetransmissie. Er wordt vermoed dat infectie kan

gebeuren na inoculatie via een vector of via het gecontamineerde

huidoppervlak. Hoewel overdracht van mens op mens niet voorkomt zijn

zowel potentiele reservoirs uit het milieu als de rol van het menselijke

reservoir in het onderhouden van een BU epidemie controversieel.

We herontwierpen een SNP-gebaseerde (ISE-SNP) fingerprinting assay om

fundamentele inzichten te verwerven in de populatiestructuur en de

evolutionaire geschiedenis van Afrikaanse M. ulcerans. Hoewel ISE-SNP

genotypering resulteerde in een ongezien hoge geografische resolutie, bleek

deze toch te beperkt om evolutionaire gebeurtenissen te reconstrueren op

een niveau kleiner dan de continentale schaal. Daarom hebben we gebruik

gemaakt van zowel state-of-the-art eerste en tweede generatie sequencing

technologieën, als geavanceerde statistische benaderingen om de

gedetailleerde evolutionaire en spatio-temporele geschiedenis van Afrikaanse

M. ulcerans te begrijpen. We verkenden eveneens het gebruik van de

aanzienlijk toegenomen resolutie gebracht door goedkope genomics om

herbesmetting te onderscheiden van herval in patiënten die verschillende BU-

episoden doormaakten. Afrikaanse M. ulcerans bleek volledig te evolueren

door middel van klonale expansie. Bovendien bleek de genetische diversiteit

vrij beperkt vanwege de trage genoomwijde substitutiesnelheid

gecombineerd met de relatief recente oorsprong van het pathogeen.

Onderzoek naar de genetische populatiestructuur onthulde het bestaan van

twee specifieke lineages binnen het Afrikaanse continent. We maakten

gebruik van temporele associaties en we bestudeerden daarnaast de

demografische geschiedenis van M. ulcerans in een BU endemisch gebied om

de mens te impliceren als een belangrijk reservoir in BU transmissie. In

Page | 14

voorgaande studies werd in het milieu M. ulcerans DNA gedetecteerd dat

mogelijk afkomstig was van BU patiënten met actieve open etterende laesies.

Transmissie kan indirect optreden in hetzelfde gemeenschappelijke waterpunt

wanneer het huidoppervlak van een naïef individu wordt gecontamineerd, en

de bacillen aanwezig op de besmette huid vervolgens subcutaan worden

geïnoculeerd via een bepaalde vorm van penetrerend (micro)trauma.

Onze waarneming over de rol die de mens speelt als onderhoudsreservoir in

het tot stand brengen van nieuwe BU infecties suggereert dat interventies in

een gebied gericht op het reduceren van de menselijke BU ziektelast

tegelijkertijd transmissieketens zullen breken in dat gebied. Actieve case-

finding programma’s en vroege behandeling van laesies in het pre-ulceratieve

stadium met specifieke antibiotica zal het aantal mycobacteriën dat in de

omgeving uitgescheden wordt doen afnemen wat uiteindelijk verdere

besmettingen van de ziekte in Afrika zou doen afnemen.

Page | 15

Chapter 1

General Introduction

Page | 16

1.1 Buruli ulcer

Mycobacterium ulcerans is the causative agent of Buruli ulcer (BU), the third

most common mycobacterial disease in humans after tuberculosis and leprosy

[1]. BU is a slowly progressing necrotizing disease of the skin and

subcutaneous tissues and the affliction is recognized as one of 17 “neglected”

tropical diseases [2].

BU was first recognized and documented in 1948 in the Bairnsdale Region of

Victoria, Australia [3], where the disease until today is known as “Bairnsdale

ulcer”. However the most commonly used name of the disease originates

from the Buruli County in Uganda (presently the Nakasongola District) where

a large number of cases was described during the 1960s [4]. BU has gone

unnoticed for a long period of time, probably because it occurs in remote

communities that used to be poorly covered by national health surveillance

systems. Presently, the condition has been reported (but not always

microbiologically confirmed), in more than 30 countries spread over Africa,

the Americas, Asia, and Oceania [5] (Figure 1.1). However, Africa is by far the

worst affected continent with 2151 new cases reported to the WHO in 2014

[6]. The highest documented notification rates are found in the West African

nations of Côte d’Ivoire, Benin, and Ghana where in some specific areas the

number of BU cases may exceed those of tuberculosis and leprosy [7]. The

disease remains very uncommon in non-African countries. Nevertheless,

important disease foci located outside of Africa have persisted in Australia,

French Guiana, and Papua New Guinea [8]. Furthermore, a number of

imported cases have been reported in non-endemic European and American

countries after international travel [9, 10].

Page | 17

Figure 1.1: BU endemic countries. The number of new cases reported to the WHO in 2014.

Image created using global health observatory data (WHO).

The disease manifests itself as painless, necrotizing cutaneous lesions that are

often located on the limbs and can take both ulcerative and non-ulcerative

forms. The non-ulcerative spectrum of the disease includes papules, nodules,

plaques, edema, and osteomyelitis. However, untreated infection with M.

ulcerans often leads to extensive destruction of the skin and soft tissues

through the formation of large ulcers, with characteristically undermined

edges and edema of the surrounding skin (Figure 1.2) [8]. Any age group can

be affected, yet incidence peaks in 5-15 year old children in Africa [7]. No

gender, ethnicity or social group is exempt.

Page | 18

Figure 1.2: The spectrum of clinical BU disease. 1, A nodule (Photo: Kingsley Asiedu) 2, A

papule (Photo: John Hayman) 3, A plaque (Photo: Mark Evans) 4, Small ulcerative lesion; note

the characteristic yellowish-white necrotic base (cotton wool-like appearance) and the

undermined edge of the lesion (Photo: John Hayman) 5, Extensive ulceration on the left arm

(Photo: Kingsley Asiedu) 6, Extensive ulceration on the leg (Photo: Henri Assé) 7, Extensive

ulceration involving the chest and abdominal walls, the pubic region, the penis and the

scrotum; post excision and ready for grafting (Photo: Pius Agbenorku) 8, Non-ulcerative

oedema (Photo: Samuel Etuaful) 9, Amputation of the left leg (Photo: Linda Lehman) 10,

Contracture deformity of knee joint (Photo: Henri Assé) 11, An eye complication as a result of

BU (Photo: Kingsley Asiedu) 12, Bone lesion (osteomyelitis) visualized with an X-ray (Photo:

WHO) 13, Squamous cell carcinoma secondary to BU (Photo: Samuel Etuaful).

In Africa, BU affects mostly poor, remote rural communities [1, 7], where the

disease is known to have a serious impact on public health. While the

mortality is low, the morbidity is considerable, as without proper treatment

Page | 19

the disease leads to scarring, contractures and other physical deformities

associated with stigma and a marked socioeconomic burden (Figure 1.2) [8].

Patients who present with advanced stages of the disease often require

extensive surgical intervention involving excision and skin grafting. This places

a lot of added burden on already stretched local health structures, as patients

often require extended hospitalization, with school dropouts and a reduced

workforce as major consequences for the affected communities [7]. Such

sequelae can be avoided by early case detection, as the early forms of the

disease can be treated relatively easily with a specific combined antibiotic

regimen (rifampicin and streptomycin) administered during 8 weeks, often

without need for surgery [11].

1.2 M. ulcerans genomic characteristics and evolutionary origin

The genome of the Ghanaian M. ulcerans strain Agy99 was sequenced in 2007

at the Pasteur Institute, Paris [12]. The chromosome contains 4160 genes and

771 pseudogenes and has a GC content of 65.7%. The Agy99 reference

genome is comprised of two circular replicons: a bacterial chromosome of

5.63 Mb and on average 1.9 copies of a giant circular virulence plasmid

pMUM001. This 174 kb megaplasmid encodes the genes required for the

biosynthesis of the endotoxin mycolactone [13]. This macrocyclic polyketide

molecule has cytotoxic, analgesic, and immunomodulatory properties that

cause the chronic ulcerative skin lesions with limited inflammation and thus

plays a key role in the pathogenesis of BU [14]. Recent major advances

identified the underlying molecular targets for mycolactone. First,

mycolactone prevents the co-translational translocation (and therefore

production) of many proteins that pass through the endoplasmic reticulum for

secretion or placement in cell membranes [15]. Second, it can target specific

scaffolding proteins, which control actin polymerization dynamics in adherent

cells and therefore lead to detachment and cell death [16]. Finally, the

hypoesthesia results from mycolactone eliciting signaling through type 2

angiotensin II receptors [17].

A recent genomic study of 35 M. marinum - M. ulcerans complex isolates

confirmed earlier indications [12] that all mycolactone-producing

mycobacteria evolved by a process of lateral gene transfer (pMUM001

Page | 20

acquisition from another actinobacterium) and subsequent reductive

evolution from a common M. marinum-like progenitor [18-20]. Differences in

the extent of genome reduction along 3 deep branches of the complex

suggest that three major lineages have adapted to slightly different niche

environments [19]. The three lineages that are responding to different

environmental pressures can therefore be considered M. ulcerans ecovars.

Ecovar 1 encompasses globally distributed fish and frog isolates and isolates

that (rarely) cause BU in humans in the Americas [21]. Ecovar 2 is represented

by Japanese isolates from human BU cases. Finally, M. ulcerans disease

isolates from West and Central Africa belong to the third ecovar which

comprises a highly clonal group that is also responsible for BU in South East

Asia and Australia (Figure 1.3).

Figure 1.3: Neighbour joining phylogenetic tree based on 128,463 variable common

nucleotide positions across 34 sequenced isolates. The tree was rooted using M. tuberculosis

as an outgroup (not shown). The major clustering of isolates are M. marinum isolates (4) -

blue; Fish and frog isolates (5) - green; Japanese isolate (1) - Mu_8765; French Guiana isolate -

Mu_1G897; Australian isolates (10) - orange; African isolates (13) - red. The scale bars show

the median pairwise divergence for the set of isolates they span. The isolates that produce

mycolactone are highlighted in pale yellow. This image has been modified from [19].

Page | 21

The reductive evolution of M. ulcerans has been mediated by the high-copy

number insertion sequence (IS) elements IS2404 and IS2606 [18-20]. As

IS2404 is present in the genome in approximately 200 copies [12], it is the

preferred target for the molecular detection of M. ulcerans by PCR in both

clinical and environmental samples [22]. For some M. ulcerans lineages a

second IS element, IS2606, is also present in a high copy number

(approximately 90 copies) [19]. IS elements code for a transposase that

catalyzes the enzymatic reaction allowing the elements to move along the

genome. Their widespread genome distribution and high copy number

indicate their potential to actively jump or duplicate in the genome and so, to

act as substrates for ongoing genome rearrangements. As such these IS

elements profoundly enhance mycobacterial genome plasticity by modifying

gene expression and inactivating genes [23]. As a direct result the M. ulcerans

Agy99 reference genome has accumulated 771 pseudogenes [12].

1.3 Etiology

The isolation of M. ulcerans in primary culture both from clinical and

environmental origins is particularly challenging as M. ulcerans is a slow

grower with an estimated generation time of 23 hours [24]. As a result,

cultures easily become contaminated by other less fastidious organisms that

outgrow M. ulcerans. To improve sensitivity, harsh decontamination methods

are employed that eliminate non mycobacterial cells prior to culture [25].

Nevertheless, the sensitivity of in vitro culture of M. ulcerans from clinical

specimens remains rather low and cultures can take several months to grow,

even under optimal conditions [26]. Since, furthermore, culture laboratories

are not easily accessible in most remote BU endemic regions of Africa,

samples from lesions often have to be stored for extended periods of time

prior to processing [27], furthermore decreasing culture sensitivity. These

difficulties present limitations for BU studies on molecular epidemiology,

treatment efficacy, and drug susceptibility where primary isolation of M.

ulcerans remains crucial.

Page | 22

1.4 Reservoir(s) and mode(s) of transmission

A wide knowledge gap exists in our understanding of the mode(s) of

transmission and the environmental reservoir(s) of BU, complicating

prevention and control of the disease. Unlike leprosy and tuberculosis, which

are characterized by person-to-person transmission, BU is rarely, if ever,

contagious, and is considered a non-communicable disease [8]. A

characteristic of BU is its focal distribution pattern: even within endemic

regions, endemic and non-endemic communities may be separated by only a

couple of kilometers [28]. Epidemiological studies established early on that

the disease is related to rural wetlands, especially those with stagnant to

slow-flowing water [8]. People living or working close to these water bodies

are more at risk for contracting the disease in endemic areas [29, 30] which

suggests the presence of the mycobacterium in or near the watery

environment. A systematic review of all papers studying risk factors

associated with M. ulcerans infection throughout the world concluded

furthermore that (i) poor wound care, and (ii) failure to wear protective

clothing, were the only other two commonly identified risk factors in all eight

papers [31]. Review of these risk factors suggests that transmission of M.

ulcerans occurs through inoculation of mycobacteria into the skin after

(micro)trauma following contact with a contaminated environment. As such, it

is generally believed that M. ulcerans is an environmental mycobacterium,

which can initiate infection after being inoculated into the subcutaneous

tissues of the skin through some form of microtrauma [32, 33]. These

observations led to numerous studies that made use of polymerase chain

reaction (PCR) screening to detect M. ulcerans DNA in various different types

of environmental samples adjacent to, or within water bodies. Studies carried

out in Australia [34] and West Africa [35] during the late 90s were the first of

their kind to successfully detect M. ulcerans DNA, and they were soon

followed by a plethora of others. IS2404, was detected in several

environmental samples including water [36], fish [37], aquatic insects [35, 38,

39], biofilms [39], detritus [40], mollusks [41], and mosquitos [42]. However,

these findings need to be interpreted with caution as:

• The bacillary concentration (as quantitated by PCR) in IS2404 positive

samples proved nearly always insignificantly low, indicating that it was

Page | 23

unlikely that M. ulcerans was multiplying in the tested specimens [36,

43, 44].

• PCR detects DNA, and not intact organisms. The death of an M.

ulcerans bacillum will lead to the release of its DNA into the

environment where it can stick to other substrates. Actual

confirmation of the presence of living M. ulcerans in IS2404 positive

specimens has proven problematic as numerous studies repeatedly

failed to isolate M. ulcerans from the environment [1, 24, 45]. There is

one notable exception; in 2008, a landmark paper was published by

Portaels et al. [46], which described the first isolation of M. ulcerans in

pure culture from an aquatic environment in Benin. The fully

characterized isolate was obtained from a water strider (Gerris sp.)

and required three rounds of passaging through mice followed by

culture. Although this result presented considerable technical

prowess, it has until this day not been successfully repeated.

• Early studies furthermore relied entirely on the detection of a single

genetic marker to identify M. ulcerans DNA: IS2404. As this target

proved unspecific [22], attempts were made to augment the specificity

of environmental screening by increasing the amount of tested

molecular targets with IS2606, and pMUM001 domains like

ketoreductase B (KR-B) and enoyl reductase (ER) [22, 39]. However,

the discovery [47] of distinct globally distributed M. ulcerans Ecovar 1

isolates that cause disease in fish and ectotherms and contain both

IS2404 and pMUM001 still severely complicate the identification of M.

ulcerans from the environment that is pathogenic to humans.

• A series of external quality assessment programs for PCR detection of

M. ulcerans in environmental specimens indicated furthermore that

false positives were common, especially in the earliest round,

jeopardizing the credibility of a substantial segment of this literature

[48].

• True IS2404 positive environmental samples might reflect only the

presence of M. ulcerans in the local environment and play no role in

human transmission. Very stringent criteria, building on Koch’s

postulates, exist in biomedical research for indicating the roles of living

Page | 24

organisms as biologically significant reservoirs and/or vectors of a

pathogen [49].

In spite of these serious limitations, several organisms have been put forward

as “potential” vectors or reservoirs of M. ulcerans in Africa [38]. Furthermore,

published results on M. ulcerans DNA ubiquitously dispersed in the

environment [50] also spawned the notion that M. ulcerans exists

saprophytically as a free-living organism in the aquatic environment. This is

however unlikely, as M. ulcerans is relatively “fragile” compared to true

opportunistically pathogenic saprophytic mycobacteria (e.g. M. fortuitum and

M. marinum):

• M. ulcerans exhibits a narrow in vitro temperature tolerance: its

optimal growth temperature ranges between 30-32°C and the

mycobacterium is sensitive to temperatures of 37°C or higher [51, 52].

This temperature requirement favors development of BU lesions on

the skin and subcutaneous tissues of the limbs where the human body

temperature is generally lower.

• The bacterium is very sensitive to oxygen concentration in vitro: its a

microaerophile [53]. Anaerobic respiration is furthermore impossible

as M. ulcerans lacks nitrate and fumarate reductase systems present in

M. marinum [12].

• M. ulcerans is sensitive to light as it has lost the ability to produce

light-inducible carotenoids that protect M. marinum from incident

sunlight [12].

• The susceptibility of M. ulcerans to streptomycin and rifampicin also

suggests that it is not free-living, where it would have encountered

these naturally occurring antibiotics in the environment [54].

• A recent study identified an inverse correlation between salt tolerance

and host adaptation [55]. Very much like the obligatory pathogen M.

tuberculosis, M. ulcerans exhibited a low (3%) salt tolerance compared

to M. fortuitum (8%).

There is furthermore additional strong genetic evidence that unlike M.

marinum, M. ulcerans is a commensal, associated with another organism that

protects the mycobacterium against unfavorable physical parameters in the

Page | 25

environment. In depth analysis of the M. ulcerans genome indicates it carries

six genomic signatures that are indicative of organisms that passed through an

evolutionary bottleneck and are adapting to a new ecological niche:

proliferation of insertion sequence elements, accumulation of pseudogenes,

chromosomal rearrangements, genome downsizing, a high degree of

relatedness (“clonality”), and acquisition of foreign genes (through plasmids

or bacteriophages) that confer a fitness advantage in the new niche [12]. M.

ulcerans shares these specific signatures with other bacteria like M. leprae

[56], Bordetella pertussis [57], and Yersinia pestis [58].

1.5 BU transmission in Australia

Australia is the only developed country in the world reporting significant local

transmission of M. ulcerans. Infection with M. ulcerans occurs infrequently in

the wet tropical northern state of Queensland, where the climate is

somewhat similar to the sub-Saharan African tropics [59]. However, more

than 80% of Australian cases of BU in the past 15 years have originated from

the temperate southeastern state of Victoria [60]. M. ulcerans DNA was

detected at low levels by qPCR in soil, sediment, water residue, aquatic plant

biofilm and terrestrial vegetation collected in the Point Lonsdale hotspot [22].

More importantly, again in Point Lonsdale, higher levels of M. ulcerans DNA

were detected in the feces of common ringtail and common brushtail

possums. Systematic testing of possum feces with adequate controls revealed

that M. ulcerans DNA could be detected in 41% of fecal samples collected in

Point Lonsdale compared with less than 1% of fecal samples collected from

non-endemic areas [61]. Until now, Australian researchers have not been able

to culture M. ulcerans from the strongly PCR positive possum feces. However,

capture and clinical examination of live possums in Point Lonsdale revealed

that from both ringtail and brushtail possums M. ulcerans could be cultured

from skin lesions [62]. As such, possum species have been indicated as a likely

reservoir in which M. ulcerans can multiply. Whole genome sequencing

revealed furthermore an extremely close genetic relationship between human

and possum M. ulcerans isolates [61].

In temperate Australia, at this point in time, M. ulcerans infection has also

been identified in other wild, domesticated, and zoo animals [24]. Laboratory-

Page | 26

confirmed cases have been diagnosed in koalas (Phascolarctos cinereus), a

mountain brushtail possum (Trichosurus cunninghami), a long-footed potoroo

(Potorous longipes), alpacas (Vicugna pacos), horses, dogs, and a cat [62]. All

animal cases were identified in locations where human cases of BU have been

reported.

M. ulcerans DNA has furthermore been found in mosquitos captured during

an outbreak of BU in the state of Victoria [42]. In a particularly interesting

case, BU lesions developed on the ear of a child who only briefly visited an

outbreak area; the child’s mother suspected a mosquito bite as the initiating

event [42]. In a case control study performed in southeastern Australia, the

reporting of mosquito bites on the forearms and lower legs was associated

with increased risk, while the use of insect repellent was associated with a

reduced risk [63]. Moreover, over a period of seven years, BU notifications

were found to correlate with those of Ross River virus, another notifiable

“local” disease which is transmitted by mosquitos [64].

These investigations combined seem to suggest that in Victoria, Australia, BU

is a zoonosis transmitted from possums to humans, via a mosquito vector,

although definite proof is to date still lacking. It is very hard to translate these

findings to the endemic regions in the Afrotropic ecozone. In the temperate

state of Victoria people have less direct contact with the environment than in

tropical Africa (e.g. bathing, washing, cooking, and farming). Also, medical

health-seeking behavior differs substantially with Australians predominantly

presenting during the early pre-ulcerative onset of the disease. M. ulcerans

infections in domesticated and wild animals have never been reported in

Africa [36, 43, 65, 66]. Furthermore, marsupial mammals (possums) are not

endemic in the African continent. In addition, the results of studies looking

into mosquito related risk factors in Africa are contradictory: two studies [29,

67] found a decreased risk of infection with mosquito net use, while a third

study [30] found no association between bed net use and infection.

Additionally, BU lesion distribution proves not to be consistent with mosquito

biting patterns [68]. Finally, a recent field study in Benin found no trace of M.

ulcerans DNA in a total of 4322 mosquitoes and 5407 mosquito larvae,

indicating that mosquitoes are probably not involved in the ecology and

dissemination of M. ulcerans in Africa [69].

Page | 27

1.6 Molecular phylogenetics and epidemiology

The study of phylogenetics is essentially concerned with reconstructing the

phylogenetic tree that represents the evolutionary history of related genes,

individuals, populations or species. The phylogenetic tree is estimated from a

multiple sequence alignment of aligned homologous molecular sequences.

When a tree has been estimated from individuals sampled from the same

population, statistical properties of the tree can be used to learn more about

the population from which the sample was drawn. In particular, the size of the

population can be estimated using Kingsman’s so called n-coalescent.

Coalescent theory is today used to infer fundamental parameters driving

molecular evolution and population dynamics [70].

Pathogen molecular sequences sampled through space and time and

particular traits associated with those sequences (like sampling location) can

be analyzed using Bayesian statistical inference approaches [71]. The major

advantage of the Bayesian statistical framework is the fact that it takes into

account the uncertainty of the unknown genealogy when trying to understand

the processes that ultimately gave rise to that genealogy [72].

Heterochronous sampling times of the samples drawn from a pathogen

population represent the most important layer of information to incorporate

into Bayesian phylogenetic analysis as these dates allow calibration of

phylogenetic trees in calendar years [73]. This approach has revolutionized

phylogenetic inference as it allows measuring evolutionary histories in units of

time instead of genetic distance. When the relationship between sequence

divergence and time (better known as substitution rate) is matching in all

branches of a time-tree we can assume a strict molecular clock model [74].

However, many biological systems do not adhere to the strict molecular clock

model [73]. Therefore, when there are variations in the substitution rate of a

time-tree, a relaxed molecular clock model needs to be considered, as that

model will consider such variance [72].

During the phylogenetic reconstruction of a pathogen, a very important early

step is choosing the proper models that drive the pathogen’s evolution like

molecular clock, coalescent, and nucleotide substitution models. When

choosing from a set of potential models, the “best” one is the model that has

Page | 28

the flexibility to explain a specific dataset without being overly parameterized.

Different methods of model selection exist in Bayesian phylogenetics [75].

The theory of statistical phylogenetics and mathematical epidemiology were

integrated into phylodynamics; a relatively new field of study that offers a

framework to investigate the evolutionary dynamics of an epidemic [76]. In

phylodynamic analysis, genomic sequence data sampled through space and

time is investigated for mutations that accumulated in the genomes of a

pathogen during an epidemic. These mutations can represent the molecular

imprint of epidemiological processes that can otherwise not directly be

observed. This can give insights to establish: the origin of an epidemic, how

the epidemic spread historically, when outbreaks of a disease occurred in a

specific region, and how it might have been transmitted [71].

BEAST2 was used in this PhD thesis to date evolutionary events, determine

substitution rates, determine demographic histories, and produce time-trees

of M. ulcerans, as this software allows for inference of phylogenies with a

diverse set of molecular clock and population parameters in the face of

phylogenetic uncertainty [77]. Presently, the BEAST package is the most

widely used software for these inferences in RNA viruses [71] but also in

animal mitochondria [78]. Its use in bacteria has however been very limited to

date [79-82] as the main limitation to the application of such methods is the

low genetic diversity and extensive homologous recombination associated

with many bacterial populations.

1.7 Research Goals: breaking the Buruli enigma

After almost 70 years of study in Africa the mode of transmission and the non-

human reservoir(s) of BU are still largely unknown, even though various

hypotheses have been formulated. This is the reason why BU carries the

epithets: the “mysterious” and “enigmatic” disease. This knowledge gap in

large part is due to the clonal population structure of M. ulcerans and the

resulting lack of effective high resolution typing methods to discriminate

different strains of M. ulcerans to enable transmission pattern tracking.

Therefore, the aim of CHAPTER 2 was to apply a redesigned ISE-SNP

fingerprinting technique to a vast panel of M. ulcerans disease isolates and

Page | 29

clinical samples originating from multiple African disease foci in order to gain

fundamental insights into the population structure, evolutionary history, and

phylogeographic relationships of the pathogen.

Even though ISE-SNP genotyping resulted in the highest geographical

resolution of genotyping achieved at that time, the resolution was still too

limited to reconstruct evolutionary events on anything smaller than the

continental scale. However, the start of this millennium brought about the

development and application of the first genomic sequencing technologies

that permitted population genetics laboratories to study full genomic

variation rather than inferring relationships from a few relatively well-

characterized genes. Unfortunately, for most of the genomics era, population

genomics was not feasible for many research groups due to the expense,

expertise, and resources required to generate even one representative

genome sequence from a species. The development of second (e.g. Illumina)

and third (e.g. Pacific Biosciences) generation sequencing technologies moved

genomics out of the genome center and into the research laboratory, which

made it possible to perform the first sophisticated, comparative genomic

studies on African M. ulcerans. We aimed to use the considerably elevated

resolution afforded by comparative next-generation genomics for the first

time to explore the molecular epidemiology of BU at the continental

(CHAPTER 4), and at the smaller geographical “village scale” in a BU endemic

region of Ghana (CHAPTER 3) and one in the Democratic Republic of Congo

(CHAPTER 5). We also meant to explore the use of the increased resolution

offered by low-cost genomics to for the first time distinguish disease relapses

from reinfections in patients with multiple BU episodes (CHAPTER 6).

We believe our findings will make an important contribution to breaking the

Buruli enigma by offering new insights into the mycobacterial epidemiological

dynamics. It is our hope that this will contribute to paving the road towards

improved BU prevention and control.

Page | 30

Page | 31

Chapter 2

Insertion sequence element single nucleotide

polymorphism typing provides insights into

the population structure and evolution of

Mycobacterium ulcerans across Africa

This chapter is published as:

Koen Vandelannoote, Kurt Jordaens, Pieter Bomans, Herwig Leirs, Lies Durnez,

Dissou Affolabi, Ghislain Sopoh, Julia Aguiar, Delphin Mavinga Phanzu, Kapay Kibadi,

Sara Eyangoh, Louis Bayonne Manou, Richard Odame Phillips, Ohene Adjei, Anthony

Ablordey, Leen Rigouts, Françoise Portaels, Miriam Eddyani, Bouke C. de Jong.

Insertion sequence element single nucleotide polymorphism typing provides insights

into the population structure and evolution of Mycobacterium ulcerans across

Applied and Environmental Microbiology 2014 Feb;80(3):1197-209

Conceived and designed the experiments: KV, LR, FP,ME, BCdJ.

Performed the experiments: KV, PB, .

Analyzed the data: KV, KJ.

Contributed reagents/materials/analysis tools: KV, HL, LD, DA, GS, JA, DMP, KK, SE,

LBM, ROP, OA, AA, LR, FP, ME, BCdJ.

Wrote the paper: KV, KJ

Page | 32

2.1 Abstract

Buruli ulcer is an indolent, slowly progressing necrotizing disease of the skin

caused by infection with Mycobacterium ulcerans. In the present study we

applied a redesigned technique to a vast panel of M. ulcerans disease isolates

and clinical samples originating from multiple African disease foci (i) to gain

fundamental insights into the population structure and evolutionary history of

the pathogen and (ii) to disentangle the phylogeographic relationships within

the genetically conserved cluster of African M. ulcerans. Our analyses

identified 23 different African insertion sequence element single nucleotide

polymorphism (ISE-SNP) types which dominate in different Buruli ulcer

endemic areas. These ISE-SNP types appear to be the initial stages of clonal

diversification from a common, possibly ancestral, ISE-SNP type. ISE-SNP types

were found unevenly distributed over the greater West African hydrological

drainage basins. Our findings suggest that geographical barriers bordering the

basins to some extent prevented bacterial gene flow between basins and that

this resulted in independent focal transmission clusters associated with the

hydrological drainage areas. Different phylogenetic methods yielded two well

supported sister clades within the African ISE-SNP types. The ISE-SNP types

from the “pan-African clade” were found widespread throughout Africa while

the ISE-SNP types of the “Gabonese/Cameroonian clade” were much rarer

and found in a more restricted area, which suggested that the latter clade

evolved more recently. Additionally, the Gabonese/Cameroonian clade was

found to form a strongly supported monophyletic group with Papua New

Guinean ISE-SNP type 8, which was unrelated to other Southeast Asian ISE-

SNP types.

2.2 Introduction

Buruli ulcer (BU), is a slowly progressing necrotizing disease of skin and

subcutaneous tissue caused by infection with Mycobacterium ulcerans [2]. BU

is the third most common mycobacterial disease in humans, after tuberculosis

and leprosy, and the least understood of the three [83]. Even though the

infection affects all age groups, at least half of all cases occur in children under

age 15 [84]. More than 30 countries world-wide have reported (but not

always confirmed) this emerging disease, with the highest incidence in West

Page | 33

and Central Africa, where the disease occurs in foci among people living in

rural marshes, wetlands, and riverine areas [2, 85]. As proximity to these slow

flowing to stagnant water bodies is a known risk factor for M. ulcerans

infection [31] and as M. ulcerans DNA has been detected in a variety of

aquatic specimens [35, 36], it is generally believed that M. ulcerans is an

environmental mycobacterium which can initiate infection after micro-

traumata of the skin [32]. However, the exact mode of transmission and the

environmental reservoir(s) of M. ulcerans remain largely unknown [60] as (i)

culturing the slow growing mycobacterium from an environmental source is

particularly difficult [46] and (ii) the significance of the detection of M.

ulcerans DNA by PCR in environmental samples remains unclear in the disease

ecology of BU [35-37, 39, 86-89].

Multilocus sequence typing analyses [18] and subsequent whole-genome

comparisons [19] proved that M. ulcerans recently evolved from a

Mycobacterium marinum progenitor by acquisition of the virulence plasmid

pMUM001. This plasmid harbors genes required for the synthesis of the

macrocyclic polyketide toxin mycolactone [13], which has cytotoxic and

immunosuppressive properties that cause chronic ulcerative skin lesions with

limited inflammation and thus plays a key role in the pathogenesis of BU [14].

Both the acquisition of the plasmid and a reductive evolution [12, 20] led the

generalist M. marinum to become a highly specialized mycobacterium, more

adapted to a restricted environment such as that of a vertebrate host.

Analysis of the genome sequence suggests that this new niche is likely to be

an obscure, aerated, osmotically stable, extracellular environment where slow

growth, the loss of several immunogenic proteins and production of

mycolactone provided selective advantages [12, 19]. Many of the changes in

this evolutionary process were mediated by two insertion sequence elements

(ISE), IS2404 and IS2606, which are present in the M. ulcerans genome in ±

200 and ± 90 copies respectively [12]. These short, mobile genetic DNA

elements promote genetic rearrangements by modifying gene expression and

sequestering genes, profoundly affecting mycobacterial genome plasticity

[23]. Increased ISE numbers are expected as the aforementioned lifestyle shift

causes many loci to become excessive as they are no longer essential for the

survival in the new environment [90]. Subsequent whole-genome

Page | 34

comparisons [19] furthermore showed that the resulting niche-adapted

genomic signature was established in a M. ulcerans progenitor before its

intercontinental dispersal.

Deciphering the structure of pathogenic bacterial populations is instrumental

for the understanding of the epidemiology, global spread and evolutionary

history of bacterial infectious diseases. Moreover, understanding the

population structure allows for studying meaningful bacterial differences

affecting disease control, including public health interventions, such as

vaccination programs [91]. Differences in the ratio of genetic variation caused

by de novo mutations relative to recombination brings about a spectrum of

different bacterial population structures ranging from “clonal” (no

recombination) to “non-clonal” (where a lot of recombination of alleles

prevents the emergence of stable clones) [92]. The clonal population

structure of M. ulcerans has meant that conventional genetic fingerprinting

methods have largely failed to genetically differentiate clinical disease

isolates, complicating molecular analyses on the elucidation of the disease

ecology and the population structure and evolutionary history of the

pathogen [93]. However, in 2009, Käser et al. [94] identified single nucleotide

polymorphisms (SNPs) within M. ulcerans haplotype-specific IS2404 elements

MUL_2990 and MUL_3871, located respectively in region of difference RD1

and RD12 [95]. The identified SNPs differentiated multiple genotypes among

isolates originating from one region in Ghana, resulting in the highest

geographical resolution of genotyping achieved to date without the use of

whole genome sequencing. Given the apparent rarity of recombination in M.

ulcerans, ISE-SNP types should contain sufficient phylogenetic signal to

reconstruct recent evolutionary events on a continental scale. Hence, in the

present study we applied a redesigned form of the ISE-SNP typing technique

as described in Käser et al. [94] to a vast panel of M. ulcerans isolates

originating from multiple African disease foci to gain deeper insights into the

population structure and evolutionary history of the pathogen and continue

to disentangle the phylogeographic relationships within the genetically

conserved cluster of African M. ulcerans.

Page | 35

2.3 Material and Methods

A panel (n = 171) of 157 M. ulcerans clinical isolates and 14 clinical specimens

with a quantification cycle Cq (IS2404) ≤ 32 (see further) originating from

disease foci in 11 different African countries, was selected to assess the

polymorphisms in the RD1 and RD12-associated haplotype-specific copies of

IS2404 (Table 2.1 and 2.2). Clinical specimens consisted of tissue fragments

and swabs originating from ulcerated and non-ulcerated BU lesions. These

surplus samples had been collected for routine diagnostic purposes and for

re-checking for quality control. All isolates and specimens were selected from

the comprehensive mycobacterial collection of the Institute of Tropical

Medicine (ITM) and were chosen to maximize temporal and spatial diversity

within countries in which more than 20 isolates / specimens were available.

Isolates and specimens were processed and analyzed for bacterial

polymorphisms without use of any patient identifiers, except for country- and

village of origin if this information was available.

Based on conventional phenotypic and genotypic methods bacterial isolates

had previously been assigned to the species M. ulcerans. They had all tested

positive for IS2404 using primers that amplify all copies of IS2404 routinely

used for diagnostic PCR [96]. Mycobacterial isolates were maintained for

prolonged storage at ≤ 70°C in Dubos-broth enriched with growth supplement

and glycerol. They were recultured on solid Löwenstein-Jensen medium. DNA

was obtained by scraping 1-2 loopfuls of colonies into 400 μL of TE followed

by heat inactivation at 100°C for 5 min and subsequent centrifugation to

remove cellular debris. Clinical specimens were maintained (after

decontamination) at ≤ -18°C. The modified Boom DNA extraction procedure

was carried out on all clinical specimens as previously described [65].

As the original ISE-SNP typing method described by Käser et al. [94] resulted in

aspecific bands, short sequence reads and high background signals, we

redesigned and optimized primers and conditions for PCR and sequencing for

application directly on clinical specimens. Primer pair RD1_SENSE

(GGTGCTTAACGAAACGTGCTG) and RD1_ANTI_SENSE

(ACGGGCTATCTGGAGAACGA) was designed to amplify a fragment of 1431 bp

in RD1 that comprises IS2404 (MUL_2990) while the primer pair RD12_SENSE

Page | 36

(CGTTGGCGCGGTACAAGCTTCCCAA) and RD12_ANTI_SENSE

(GATGGTCGCGGTGCTGCTTGCCCT) was used to amplify a 1871 bp PCR product

in RD12 that comprises IS2404 (MUL_3871). Primers were designed with

Primer premier 6 (Premier Biosoft, California, US) and evaluated in silico with

Amplify 3.1.4 (Bill Engels, University of Wisconsin). The PCR design was

challenging as only haplotype specific copies of IS2404 (MUL_2990 and

MUL_3871) were to be amplified and because the relevant regions are of

considerable size (1730bp for RD1 / 1905bp for RD12). Although we reduced

the size of the amplicons in both assays (by 299bp for RD1 and by 48bp for

RD12), they still contained all the variable nucleotide positions described by

Käser et al. [94]. PCR reaction mixtures contained 1.0 U of HotStarTaq

polymerase (QIAGEN, Hilden, Germany), 3.0 μL 10X PCR buffer, 6.0 μL

Qsolution, 1.5 mM MgCl2, 200 μM of each dNTP, and 0.5 μM of each primer in

a total volume of 30 µL. PCR reactions were carried out on a Biometra

TProfessional Thermal cycler under the following conditions: an initial

denaturation step of 15 min at 95°C followed by 40 cycles of denaturation for

1 min at 95°C, annealing for 1 min at 65°C (RD1) or 70°C (RD12) and

elongation for 2 min at 72°C, and ending with a final elongation step of 10 min

at 72°C. PCR products were visualized with ethidium-bromide on 1% agarose

gels by electrophoresis (30 min, 100 V). PCR products were purified by

automated gel excision. Bi-directional sequencing was performed by the

Genetic Service Facility of the Flanders Institute for Biotechnology (GSF-VIB)

on an Applied Biosystems 3730 DNA Analyzer capillary sequencer using the

ABI PRISM BigDye Terminator cycle sequencing v3.1 kit using the PCR primers.

We estimated the bacterial load by a quantitative PCR (qPCR) for IS2404 as

described by Fyfe et al. [22] on a set of 122 clinical specimens to determine

the quantification cycle Cq (IS2404) below which the optimized genotyping

PCRs were always successful.

The sequences of RD1 and RD12 were concatenated to yield a 3278 bp

fragment and aligned in Clustal X v2.1 [97]. Sequences were trimmed to equal

length and all currently known ISE-SNP types (including 5 ISE-SNP types from

Papua New Guinea, Australia and Malaysia) [94] were added to this dataset.

We mapped SNPs according to the Agy99 bacterial reference chromosome

(GenBank accession no. NC_008611). We constructed a Neighbor-Joining (NJ)

Page | 37

tree based on p-distances between ISE-SNP types [98] in MEGA v5 [99].

Maximum Parsimony (MP) and Maximum Likelihood (ML) trees were

estimated in the same program using a heuristic search with the tree-

bisection-reconnection branch-swapping algorithm and random addition of

taxa. Relative branch support was evaluated with 1000 bootstrap replicates

[100] for the NJ and MP tree, and 200 for the ML tree. Phylogenetic trees for

ML analysis were inferred with the nucleotide substitution model selected

using jModelTest v0.1.1 [101]. Phylogenetic relationships were inferred with

ISE-SNP 28 (strain ITM_030524) from Papua New Guinea as outgroup, isolated

from a patient who had never travelled outside of the region around Yarapos

in the East Sepik Province (J Taylor, personal communication). Trees were

drawn using FigTree software [102].

A haplotype network was derived using the Median Joining algorithm after

processing the data with the reduced median method as implemented by

Network v4.6.1.0 with default settings [103].

The open source Geographic Information System “Quantum GIS” (QGIS) [104]

was used to generate the figure on the geographical distribution of African M.

ulcerans. The geographical location of the residence of BU patients at the time

of clinical visit was rendered as points. In the case where residence

information was missing we used the location of the hospital supplying the

sample. A modification of the QGIS Python plugin “Shift Points” was used to

modify this point shape file where point features with the same position

overlapped. Point displacement rendered such features in a circle around the

original “real” position. The river layer is translated from the “River-Surface

Water Body Network” dataset of the “African Water Resource database” of

the Food and Agriculture Organization (FAO) of the United Nations [105]. The

administrative borders of countries are rendered from the Global

Administrative Unit Layers dataset of FAO.

All statistical testing was performed in R v2.15.2 [106]. The correlation

between the number of isolates per country and the number of ISE-SNP types

per country was checked with a Spearman’s rank order correlation coefficient.

To examine the relation between ISE-SNP types and the greater West African

hydrological drainage basins a Fisher's exact test was used.

Page | 38

Table 2.1: Isolates used in this study; CDTUB: Centre de Dépistage et de traitement de l'Ulcère

de Buruli, CPC : Centre Pasteur du Cameroun, DRC: Democratic Republic of Congo, IME:

Institut Médical Evangélique, KCCR: Kumasi Centre for Collaborative Research in Tropical

Medicine, NCTC: National Collection of Type Cultures, PNLUB: Programme National de Lutte

contre l'Ulcère de Buruli, YOI : Year of isolation.

ISE-SNP

type Culture n°

Country of

Origin

First-level

Administrative

Division

Second-level

Administrative

Division

Third-level

Administrative

Division

Source YOI Remarks

1 ITM_940511 Ivory Coast Moyen-Cavally Duékoué Niambli ITM 1994

1 ITM_000483 Ivory Coast Moyen-Cavally Duékoué Niambli ITM 2000

1 ITM_000870 Ivory Coast Dix-Huit

Montagnes Zouan-Hounien Ouyatouo ITM 2000

1 ITM_063519 DRC Bas-Congo Cataractes /

Songololo

Luima / Cité

Songololo IME 2006

1 ITM_071924 Congo Kouilou Madingo-Kayes Loukouala ITM 2007 Originates from same

patient as 071925

1 ITM_071925 Congo Kouilou Madingo-Kayes Loukouala ITM 2007 Originates from same

patient as 071924


Songololo

Bamboma /

Mbanza-Manteke IME 2007


Songololo Palabala / Nkamuna IME 2007




Songololo

Luima-Mayanga /

Ngombe IME 2007


Songololo

Bamboma /

Mbanza-Manteke IME 2007


Songololo Luima / Luvuvamu IME 2007



Originates from same

patient as 072841




patient as 072840



1 ITM_073459 Benin Kouffo Lalo Ahojinako CDTUB Lalo 2007


Songololo Luima / Kisonga IME 2007


Songololo

Luima / Cité

Songololo IME 2007

1 ITM_073478 Angola Malanje Marimba Kafufu / Luremo

(Kwango River) IME 2007


Songololo Luima / Kisonga IME 2007


Songololo Kilueka / Nzundu IME 2008


Songololo Lovo / Tole IME 2010


Songololo Mayanga / Mpelo IME 2010


patient as 100141


Songololo Luima / Luvuvamu IME 2010


patient as 100142




Songololo Mayanga / Mpelo IME 2010


Songololo

Luima / Nkondo-

Kiomba IME 2003

1 ITM_040149 Ghana Ashanti Asante Akim North Agogo Presbyterian

Hospital ITM 2003

1 ITM_991591 Togo Maritime Vo Anagali ITM 1999

1 ITM_050303 Congo Kouilou

ITM 1979 Originates from same

patient as 050304

Page | 39

ISE-SNP

type Culture n°

Country of

Origin

First-level

Administrative

Division

Second-level

Administrative

Division

Third-level

Administrative

Division

Source YOI Remarks

1 ITM_050304 Congo Kouilou

ITM 1979 Originates from same

patient as 050303

1 ITM_960658 Angola Bengo Dande Caxito ITM 1996 Originates from same

patient as 960657

1 ITM_960657 Angola Bengo Dande Caxito ITM 1996 Originates from same

patient as 960658

1 ITM_072662 Ghana Ashanti Asante Akim North Ananekrom KCCR 2007

1 ITM_072646 Ghana Ashanti Atwima Mponua Abofrom KCCR 2007

1 ITM_072651 Ghana Ashanti KMA Kaase KCCR 2007

1 ITM_120140 Cameroon Adamawa

Region Maya-Banyo Bankim / Mbondji II CPC 2011

1 ITM_030950 Benin Kouffo Lalo Adoukandji CDTUB Lalo 2003

1 ITM_030716 Benin Kouffo Lalo Tchito / Village

Aboeti CDTUB Lalo 2003

1 ITM_102686 Nigeria Oyo State Ibadan Ibadan ITM 2010

1 ITM_083232 Angola Lunda Norte Xa-Muteba (Kwango River) ITM 2008

1 ITM_000869 Ivory Coast Moyen-Cavally Duékoué Guezon ITM 2000

1 ITM_990007 Ivory Coast Haut-

Sassandra Issia Guetuzon II ITM 1998

1 ITM_991633 Ivory Coast Moyen-Cavally Duékoué Guezon ITM 1999

2 ITM_030791 Benin Kouffo Lalo Tchito / Gare CDTUB

Zagnanado 2003

2 ITM_970680 Benin Mono Houéyogbé Sahoué CDTUB Lalo 1997

2 ITM_022876 Benin Kouffo Lalo Tohou CDTUB Lalo 2002

2 ITM_021434 Benin Kouffo Klouékanmè Adjassagon CDTUB Lalo 2002

2 ITM_012596 Benin Mono Bopa Lobogo CDTUB Lalo 2001

2 ITM_071938 Benin Kouffo Lalo Tandji CDTUB Lalo 2007

4 ITM_5150 DRC Bandundu Kwilu

ITM 1962

5 ITM_940512 Benin Zou Ouinhi Ouokon CDTUB

Zagnanado 1994

5 ITM_010157 Benin Zou Zogbodomè Domè-

Houandougon

CDTUB

Zagnanado 2001

5 ITM_000951 Benin Zou Zogbodomè Domè-

Houandougon

CDTUB

Zagnanado 2000

5 ITM_970435 Benin Ouémé Bonou Bonou CDTUB

Zagnanado 1997


patient as 970301

5 ITM_970301 Benin Ouémé Bonou Bonou CDTUB

Zagnanado 1997


patient as 970435

5 ITM_000479 Benin Zou Zagnanado Zagnanado / Doga CDTUB

Zagnanado 2000

5 ITM_092100 Benin Zou Zagnanado Doga-Domè CDTUB

Zagnanado 2009

5 ITM_083865 Benin Zou Ouinhi Tohoue /

Hounnoumè

CDTUB

Zagnanado 2008

5 ITM_093013 Benin Zou Ouinhi Ouinhi /

Monzoungoudo

CDTUB

Zagnanado 2009


Monzoungoudo

CDTUB

Zagnanado 2009

5 ITM_101300 Benin Zou Ouinhi Sagon / Adamè CDTUB

Zagnanado 2010

5 ITM_101302 Benin Zou Ouinhi Dasso / Bossa CDTUB

Zagnanado 2010

5 ITM_102554 Benin Zou Ouinhi Dasso / Agonkon CDTUB

Zagnanado 2010

5 ITM_081919 Benin Zou Ouinhi Dasso / Yaago &

Akantomè

CDTUB

Zagnanado 2008

Page | 40

ISE-SNP

type Culture n°

Country of

Origin

First-level

Administrative

Division

Second-level

Administrative

Division

Third-level

Administrative

Division

Source YOI Remarks

5 ITM_092997 Benin Zou Djidja Oungbègame CDTUB

Zagnanado 2009

5 ITM_080066 Benin Zou Ouinhi Sagon / Ayizè CDTUB

Zagnanado 2008

5 ITM_070381 Benin Zou Ouinhi Dasso / Yaago CDTUB

Zagnanado 2007


Monzoungoudo

CDTUB

Zagnanado 2007

5 ITM_070131 Benin Zou Zagnanado Dovi-Dove / Tévedji CDTUB

Zagnanado 2007

5 ITM_092473 Benin Zou Ouinhi Tohoue /

Midjannangon

CDTUB

Zagnanado 2009

5 ITM_082549 Benin Zou Ouinhi Tohoue / Akassa CDTUB

Zagnanado 2008

5 ITM_090149 Benin Zou Zagnanado Dovi-Dove / Tévedji CDTUB

Zagnanado 2009

5 ITM_091800 Benin Zou Ouinhi Tohoue / Gangban CDTUB

Zagnanado 2009

5 ITM_083584 Benin Zou Ouinhi Ouinhi / Ahicon CDTUB

Zagnanado 2008

5 ITM_9146 Benin Zou Zagnanado Kpedekpo / Loko-

Alankpe

CDTUB

Zagnanado 1992

5 ITM_991721 Benin Atlantique Toffo Séhoué CDTUB

Zagnanado 1999

5 ITM_092472 Benin Atlantique Toffo Séhoué / Agaga CDTUB

Zagnanado 2009

5 ITM_070383 Benin Ouémé Dangbo Dékin CDTUB

Zagnanado 2007

5 ITM_070625 Nigeria Ogun State Yewa North Odja Odan CDTUB

Zagnanado 2007

5 ITM_061509 Benin Zou Zagnanado Zagnanado CDTUB

Zagnanado 2006

5 ITM_081676 Benin Plateau Adja-Ouere Tatonnoukon CDTUB

Zagnanado 2008

5 ITM_081681 Benin Plateau Issaba Onigbolo CDTUB

Zagnanado 2008

5 ITM_082696 Benin Ouémé Adjohoun Abato CDTUB

Zagnanado 2008

5 ITM_091801 Benin Zou Zogbodomè Kpokissa /

Hinzounmè

CDTUB

Zagnanado 2009

5 ITM_092101 Benin Ouémé Dangbo Gbéko CDTUB

Zagnanado 2009


Zagnanado 2009

5 ITM_100126 Benin Zou Zogbodomè Kpokissa CDTUB

Zagnanado 2010

5 ITM_951009 Benin Zou Zagnanado

CDTUB

Zagnanado 1995

6 ITM_5151 DRC Maniema Kasongo

ITM 1972

7 ITM_970359 Ghana Ashanti Amansie West Manso-Afraso ITM 1997

7 ITM_970606 Ghana Ashanti Amansie West Yaw Kasakrom ITM 1997

7 ITM_970677 Ghana Ashanti Amansie West Manso Dominase ITM 1997

7 ITM_970678 Ghana Ashanti Asante Akim North Afrisre ITM 1997

7 ITM_970959 Ghana Ashanti Amansie West Manso-Afraso ITM 1997

7 ITM_970964 Ghana Ashanti Amansie West Offinho Asaman ITM 1997

7 ITM_971351 Ghana Ashanti Atwima Mponua Achiase ITM 1997

7 ITM_980063 Ghana Ashanti Atwima Mponua Achiase ITM 1998

7 ITM_940662 Ivory Coast Moyen-Cavally Duékoué Nanandi ITM 1994


Sassandra Issia Guetuzon I ITM 1998

Page | 41

ISE-SNP

type Culture n°

Country of

Origin

First-level

Administrative

Division

Second-level

Administrative

Division

Third-level

Administrative

Division

Source YOI Remarks

7 ITM_990734 Ivory Coast Moyen-Cavally Duékoué Duékoué ITM 1999


Sassandra Issia Bediegbeu ITM 1999

7 ITM_072634 Ghana Ashanti Asante Akim North Adoniem KCCR 2007

7 ITM_072652 Ghana Ashanti Atwima Mponua Achiase KCCR 2007



7 ITM_072658 Ghana Western

Region Wassa West Owusukrom KCCR 2007

7 ITM_072650 Ghana Ashanti Atwima

Nwabiagya Kyereyase KCCR 2007

7 ITM_072630 Ghana Central Upper Denkyira Nkotumso KCCR 2007

7 ITM_072656 Ghana Ashanti Atwima Mponua Abompe KCCR 2007

7 ITM_072655 Ghana Ashanti Atwima Mponua Sireso KCCR 2007

7 ITM_072653 Ghana Ashanti Atwima Mponua Amadaa KCCR 2007



Zagnanado 2007

13 ITM_021433 Benin Kouffo Lalo Gnizoumè /

Hangbanou CDTUB Lalo 2002

13 ITM_022045 Benin Kouffo Lalo Adoukandji /

Yamontouhoué CDTUB Lalo 2002

13 ITM_022287 Benin Zou Agbangnizoun Kpota CDTUB Lalo 2002

13 ITM_022875 Benin Kouffo Lalo Gnizoumè CDTUB Lalo 2002

13 ITM_030717 Benin Kouffo Lalo Ahomadégbé CDTUB Lalo 2003

13 ITM_030718 Benin Kouffo Lalo Lalo CDTUB Lalo 2003

13 ITM_031892 Benin Kouffo Lalo Hlassamè CDTUB Lalo 2003

13 ITM_071804 Benin Kouffo Lalo Zalli CDTUB Lalo 2007

14 ITM_991590 Togo Maritime Vo Tchekpo Deve ITM 1999 Originates from same

patient as 000909

14 ITM_000909 Togo Maritime Vo Tchekpo Deve ITM 2000 Originates from same

patient as 991590

14 ITM_993354 Togo Maritime Vo Tchekpo Deve ITM 1999

14 ITM_042407 Togo Maritime Vo Kodji Kopé ITM 2004


Songololo

Kimpese / Cité-

Kimpese IME 2007


patient as 070123


Songololo

Kimpese / Cité-

Kimpese IME 2007


patient as 070404


Songololo

Kimpese / Cité-

Kimpese IME 2009


Sassandra Issia Zakogbeu ITM 1998

18 ITM_070386 Nigeria Anambra State Ayamelum Ifite Ogwari ITM 2007

19 ITM_001211 Ivory Coast Dix-Huit

Montagnes Zouan-Hounien Zouan-Hounien ITM 2000

20 ITM_020279 Cameroon Centre Region Nyong-et-

Mfoumou Ayos CPC 2002

20 ITM_091067 Gabon Moyen-

Ogooué

Ogooue et des

Lacs Junkville ITM 2009


Ogooué

Ogooue et des

Lacs Gravier ITM 2011

Page | 42

ISE-SNP

type Culture n°

Country of

Origin

First-level

Administrative

Division

Second-level

Administrative

Division

Third-level

Administrative

Division

Source YOI Remarks


Mfoumou Akolo CPC 2002


Mfoumou Akonolinga CPC 2011


Mfoumou Obis CPC 2002

23 ITM_9102 Cameroon Centre Region

ITM 1970

23 ITM_9103 Cameroon Centre Region

ITM 1970


Ogooué

Ogooue et des

Lacs

Lambaréné /

Adaghe ITM 2010


Ogooué

Ogooue et des

Lacs Issac ITM 2011




Mfoumou

Akonolinga /

Wouma CPC 2011


Mfoumou Medjap CPC 2011


Mfoumou

Akonolinga /

Ekolman CPC 2011

23 ITM_120534 Cameroon Centre Region Nyong-Et-Soo Bembé CPC 2011




Mfoumou Akonolinga / Djo'o CPC 2011









24 ITM_051459 Uganda Northern

Region Adjumani Adjumani NCTC 2005

25 ITM_120537 Cameroon Centre Region Nyong-Et-Soo Edjom CPC 2011


Mfoumou Akam-Engali CPC 2011



Page | 43

Table 2.2: Clinical specimens used in this study; DRC: Democratic Republic of Congo, PNLUB:

Programme National de Lutte contre l'Ulcère de Buruli, YOI : Year of isolation.

ISE-

SNP

type

Sample n° Country of

Origin First-level Second-level Third-level Source YOI

1 BK121032 Nigeria Cross river

state Ogoja

TBL hospital

monaiya ITM 2012

1 BK120888 DRC Maniema Kibombo Likeri PNLUB 2012

1 BK120890 DRC Maniema Kasongo Samba / Malela PNLUB 2012

1 BK120891 DRC Maniema Kasongo Kankumba PNLUB 2012

1 BK065361 Nigeria Enugu State Igbo Eze North Nkpo Hamida ITM 2006

5 BK121025 Nigeria Ogun State Abeokuta North Abeokuta / Ijaye

State hospital ITM 2012

5 BK121026 Nigeria Ogun State Abeokuta North Abeokuta / Ijaye

State hospital ITM 2012

5 BK121031 Nigeria Ogun State Yewa South Oke-Odan / PHC

Oke-Odan ITM 2012

17 BK065369 Nigeria Ebonyi State Ohaozora Iburu ITM 2006

20 BK105250 Gabon Nyanga Douigni Moussamou

kougou ITM 2010

21 BK101660 Gabon Moyen-

Ogooué

Ogooue et des Lacs

Department

Lambaréné / Point

V ITM 2010


Ogooué

Ogooue et des Lacs

Department

Lambaréné /

Bellevue ITM 2010


Ogooué

Ogooue et des Lacs

Department Lambaréné / Isaac ITM 2010

Page | 44

2.4 Results

Primers and conditions for PCR and sequencing were redesigned and

optimized from those described by Käser et al. [94] for application directly on

clinical specimens (Figure 2.1). Isolates ITM_5150, ITM_5151 ITM_940511,

ITM_940512, ITM_960658, ITM_940662, ITM_970359, and ITM_970680 were

included in the panel to validate the redesigned assays. These isolates were

also included in the panel of Käser et al. [94] and gave identical genotypes as

with the redesigned assays.

Figure 2.1: The 1431 bp fragment of IS2404 (MUL_2990) in RD1 and the 1871 bp fragment of

IS2404 (MUL_3871) in RD12; L: GeneRuler 100 bp Plus DNA Ladder (Fermentas); 1,

ITM_951009; 2, ITM_072645; 3, ITM_072653; 4, ITM_072658; 5, No Template Control (NTC).

As our collection of African M. ulcerans isolates did not represent all endemic

countries and their different regions to the same extent owing to the low

sensitivity of culture, we fine-tuned the technique for application directly on

clinical specimens by adjusting individual PCR component concentrations and

optimizing the thermal PCR profile. We were thus able to deduce sequence

information of clinical samples with a modest bacterial load corresponding to

a Cq (IS2404) ≤ 32. Failure of PCR amplification for specimens with Cq (IS2404)

> 32 was caused by low mycobacterial DNA concentration.

Amplification and sequencing of IS2404 in MUL_2990 (RD1) and MUL_3871

(RD12) was successful on the entire collection of 157 (100%) clinical isolates

(Table 2.1 and 2.2). The optimized method also proved successful in all 14

clinical specimens analyzed with a Cq (IS2404) ≤ 32. A total of 75 (31 in

Page | 45

MUL_2990 (RD1), 44 in MUL_3871 (RD12)) variable nucleotide positions were

identified, including four insertions / deletions (indels) (Figure 2.2). This

resulted in 28 ISE-SNP types, of which 23 were found on the African continent.

Sixteen of these were newly identified types while the other seven

corresponded to the ISE-SNP types described by Käser et al. [94]. The Papua

New Guinean ISE-SNP type 28, used in the phylogenetic analyses as outgroup,

was also a novel type.

Figure 2.2: Sequence variation in two haplotype-specific concatenated IS2404 elements;

MUL_2990 (RD1) and MUL_3871 (RD12). Only variable nucleotides in the aligned sequences

are shown for all 28 ISE-SNP types. SNP position numbers are given according to Käser et al.

[94], with 1 corresponding to position 3313231 in RD1 and 1498 to position 4326896 in RD12

according to the Agy99 bacterial reference chromosome.

The Spearman's rank correlation coefficient showed a significant relationship

between the number of isolates per country and number of ISE-SNP types

identified per country (r[9] = 0.79, p < 0.01). We were able to identify all

African ISE-SNP types described by Käser et al. [94] except ISE-SNP type 3,

which was found in the Greater Accra Region of Ghana, a region not covered

by our panel. Some ISE-SNP types were common (types 1, 2, 5, 7, 23) while

others were represented by one isolate/clinical specimen only (types 4, 6, 12,

16, 17, 18, 19, 21, 24, 25, 26, and 27). Our panel included a number of linked

isolates originating from the same patient. In all eight occurrences (Table 2.1)

these linked isolates revealed the same ISE-SNP type. The geographical

distribution of all African ISE-SNP types is shown in Figure 2.3A. Most ISE-SNP

types had a distinct restricted geographical localization. For example, all 41

isolates of ISE-SNP type 5 were recovered from an area in West Africa with a

60 km radius. Other African ISE-SNP types were more widely dispersed. ISE-

Page | 46

SNP type 1 for instance, although also emerging in clusters in Central Africa,

was identified throughout Central and West Africa. Furthermore, some

regions harbored a multitude of different ISE-SNP types while others yielded

just one. In southern Benin, for example, the greatest variety of allelic

patterns is found with as much as five ISE-SNP types (viz. types 1, 2, 5, 12, and

13) circulating.

We found a strong relationship (Fisher’s exact test: p < 0.0001) between the

distribution of ISE-SNP types and the greater West African hydrological

drainage basins (Table 2.3) as shown in figure 2.3B. Hydrologically this region

can be divided into separate main drainage areas: the Mono, the Kouffo, the

Oueme, the Yewa, the Ogun, and the Togolese Coastal Rivers Basin. The main

rivers of these border-crossing basins all arise on the central west-African

plateau and form broad fertile richly inundated plains when they reach the

lowlands of the coastal regions, where BU-endemic areas are concentrated.

Here, the basins can be divided into an inland region drained by a network of

freshwater rivers and streams that discharges into a region of extensive

brackish-water swamps interconnected with lakes, narrow lagoons and

streams parallel to the coastline. Haplotype ISE-SNP 5 dominates in the BU-

endemic areas of the Oueme, the Yewa, and the Ogun Basins. The haplotype

is best represented along the Oueme and its last tributary, the Zou River,

where most Beninese BU cases are reported [107]. After the confluence, the

Oueme traverses over 1500 km² of floodplains after which the river discharges

into Lake Nokoue, Porto-Novo Lagoon, and the coastal lagoons of Nigeria

which are all interconnected by the numerous channels of the deltaic fan of

the Oueme River. Two other drainage units, the Ogun and the Yewa discharge

in this same system of lagoons and streams. So, although the basins are

separate drainage systems, they discharge into this collective interconnected

system, which could potentially explain the observed shared distribution of

haplotype ISE-SNP 5. Haplotypes ISE-SNP 2 and 13 dominate in the BU-

endemic areas of the Kouffo Basin [28, 108]. After draining the highly BU

endemic regions of the commune of Lalo, the Kouffo discharges via Lake

Aheme in the “western lagoonal complex“ that is in contact with the Gulf of

Guinea at “Bouche du Roi”. Although this system is part of a semi-continuous

line of narrow lagoons that runs behind the dunes along the entire coastal

Page | 47

strip until the Ghanaian border, it is not in contact with the interconnecting

drainage system of the Oueme Delta. The lower course of the Mono River

forms the border between Togo and Benin and discharges in the same

“western lagoonal system” as the Kouffo River. The Mono River basin

however has no known areas of BU-endemicity, despite similar riverine

habitats. Even further west, in southern Togo, three small coastal rivers (Boko,

Haho, and Zio) form a third small basin. The basin encompasses a couple of

BU endemic regions in which the Togolese haplotype ISE-SNP 14 is

represented.

Table 2.3: Distribution of ISE-SNP types over the hydrological drainage basins of southern

Benin, southern Togo, and southwestern Nigeria.

Coastal Rivers Basin Mono Kouffo Oueme Yewa Ogun

ISE-SNP type 1 1 0 3 0 0 0

ISE-SNP type 2 0 1 5 0 0 0

ISE-SNP type 5 0 0 0 37 2 2

ISE-SNP type 12 0 0 0 1 0 0

ISE-SNP type 13 0 0 8 0 0 0

ISE-SNP type 14 3 0 0 0 0 0

Page | 48

Figure 2.3: (A) The geographical distribution of African M. ulcerans. The location of residence

at the time of clinical visit of the individual BU patients was retrospectively correlated with

the ISE-SNP typing results. ISE-SNP types represented by only one clinical isolate or specimen

are depicted as numbers while more common ISE-SNP types are color coded. (B) The uneven

distribution of ISE-SNP types over the different greater hydrological drainage basins of West

Africa.

The NJ method yielded two well supported sister clades within the African ISE-

SNP types (Figure 2.4A). The first clade comprised ISE-SNP types 20 and 21

which circulate in different BU endemic regions of Cameroon and Gabon; this

clade had also a high bootstrap support for the MP and ML analyses. A second

“pan-African clade” comprised all other African ISE-SNP types (Figure 2.4B).

Support for other nodes within the “pan-African clade” was very low

Page | 49

(bootstrap values < 70%) except in the NJ analysis for (i) a clade of ISE-SNP

types 7 and 19 which circulate in Ghana and Ivory coast (ii) a clade of ISE-SNP

types 22, 23, 25, 26, and 27 which all circulate in Cameroon and neighboring

Gabon (iii) a modest support in the NJ analysis for a clade of ISE-SNP types 1,

2, 3, 4, 5, 12, 13, 14, 15, 16, 17, found throughout the continent.

Mycobacterium ulcerans haplotypes from Australia and South East Asia were

also included in the analysis as these, together with African haplotypes,

belong to the more virulent and distinct “classic” phylogenetic lineage [95,

109], relative to M. ulcerans isolates elsewhere. It is of particular interest that

ISE-SNP type 8 from Papua New Guinea forms a strongly supported

monophyletic group with ISE-SNP types 20 and 21 from Cameroon and Gabon

and is distinctly unrelated to other Southeast Asian clinical isolates, which

belong to ISE-SNP types 9, 10 and 28 (Figure 2.4A). In contrast, other ISE-SNP

types found in Papua New Guinea are related to Malaysian and Australian

clinical isolates.

Page | 50

Figure 2.4: A: Neighbor-joining tree showing the phylogenetic relationships between the 28

currently known ISE-SNP types of M. ulcerans with haplotype ISE-SNP 28 from Papua New

Guinea as outgroup. Bootstrap values (if >70%) for the Neighbor-Joining (NJ), Maximum

Likelihood (ML) and Maximum Parsimony (MP) analysis are given at the nodes as NJ/ML/MP.

ISE-SNP types belonging to “pan-African clade” and the “Gabonese/Cameroonian clade” are

highlighted in grey and red respectively. B: The geographical distribution of the pan-African

clade and the Gabonese/Cameroonian clade. The location of residence at the time of clinical

visit of the individual BU patients was retrospectively correlated with the ISE-SNP typing

results.

Page | 51

The phylogenetic network (Figure 2.5) showed that a number of African ISE-

SNP types are closely related and only differ in a single, or few, mutational

steps. However, other African ISE-SNP types are more distantly related and

even differ in a number of mutational steps that is similar to the number of

mutational steps between African and non-African types. Analogous with the

phylogenetic tree analysis, the “pan-African clade” is divided into two major

clusters. A first major cluster comprised the common Central and West

African ISE-SNP type 1 and several, closely related, yet rarer, ISE-SNP types.

The second cluster comprised ISE-SNP types 22, 23, 25, 26 and 27, which are

circulating in Cameroon and neighboring Gabon. The network also showed

several more distantly related African haplotypes of which most are relatively

rare (types 18, 19, 20, 21, and 24) and one (type 7) is common.

Page | 52

Figure 2.5: Phylogenetic network showing patterns of descent among the 28 currently know

ISE-SNP types of M. ulcerans in relation to their geographic origin. The network was derived

by using the Median Joining algorithm after processing the data with the reduced median

method as implemented by Network v.4.6.1.0. Each circle represents a unique ISE-SNP type,

and the size of the circle is proportional to the number of individuals sharing that type.

Numbers in boxes represent the number of mutational steps (if not given, then a single

mutational step). Position at which mutations occurred are given in Figure 1. Color codes

represent the country of origin as shown in the key.

2.5 Discussion

In this study, we applied an optimized ISE-SNP genotyping technique to a

comprehensive panel of isolates from all African countries that ever yielded

culture confirmed BU cases. This analysis, unparalleled in size and scope,

allowed us to assess the diversity and population structure across BU endemic

regions on a continental scale, and to explore the phylogenetic and

phylogeographic relationships within the genetically conserved cluster of

African M. ulcerans ISE-SNP types.

Analysis of polymorphisms in the RD1 and RD12 genomic regions, which have

been identified among the most variable of the M. ulcerans bacterial

chromosome [109], over our comprehensive sample panel spanning 11

endemic African countries, identified 23 different African ISE-SNP types. The

observed low level of polymorphisms [110], together with the characteristic

geographical restriction of most ISE-SNP types suggests a highly clonal

population structure of African M. ulcerans. This is in agreement with the

findings of Doig et al. [19], who found that clinical isolates from Ghana and

Benin were only separated by an average pairwise distance of 160 SNPs over

the entire 5.6 Mbp sequenced bacterial chromosome. This low sequence

diversity is in strong contrast with other pathogens like Helicobacter pylori,

where microevolution can be observed even within serial bacterial isolates

from individual humans with prolonged infection [111]. The genetic

conservation among African M. ulcerans might reflect a short evolutionary

history since its intra-continental dispersal but might also be explained by a

low mutation rate. Reliable estimates of mutation rates are required to

resolve these issues [110].

Page | 53

Closely related ISE-SNP types dominate in different BU endemic areas. The

identified SNPs describe a phylogenetic path wherein these individual ISE-SNP

types document the sequential accumulation of mutations from a common

root. If we assume that (i) an ancestral ISE-SNP type will be more

geographically dispersed than a more recently derived type and (ii) that the

geographical distribution of the ISE-SNP types is not explained by selective

effects, this common root node is represented here by ISE-SNP 1; the most

common type distributed over the entire continent. The different ISE-SNP

types thereby represent the initial stages of clonal diversification through de

novo mutations from this, possibly ancestral type, after its intra-continental

spread.

The unevenly distributed ISE-SNP types circulating within small regions of

West Africa are furthermore suggestive of the existence of independent

transmission clusters. We found a strong association between the distribution

of ISE-SNP types and the greater West African hydrological drainage basins.

Genetic differences between clinical isolates originating from two neighboring

drainage areas in Benin have previously been reported [19, 46]. It appears

that geographic barriers (eg. elevated regions and salt water) bordering these

hydrological basins, separated an ancestral genotype to a certain extent into

discontinuous parts by the formation of a physical barrier to bacterial gene

flow. Our data suggest that this resulted in differentiation by the slow

accumulation of point mutational changes of the original founder clone (ISE-

SNP 1) into different closely related types distributed over the various basins

(Figure 2.3B and Table 2.3). New ISE-SNP types derived from the founder type

did not easily spread but formed focal transmission clusters associated with

the hydrological drainage areas. Hence, BU infections in these areas probably

resulted from locally confined transmission of a single circulating clone, with

only occasional transfer of clones between basins. Our findings confirm a

study of Röltgen et al. [112], in which a number of M. ulcerans haplotypes

within the Densu hydrological basin of Ghana (with SNP typing based on

whole genome data) were differentiated, revealing similar focal transmission

clusters within the basin itself. Hence, our findings provide additional

evidence that both transmission and fine-grained evolutionary events play at

the local level and we consequently hypothesize that potential reservoirs have

Page | 54

a limited mobility. Such a scenario would correspondingly account for the

presence of endemic and non-endemic villages in close proximity (<10 km)

within the same drainage basin [108].

Our phylogenetic analyses did not result in a fully resolved phylogenetic tree

since most nodes had low bootstrap support. Nevertheless, there was support

for a “pan-African” and a “Gabonese / Cameroonian” sister clade. The ISE-SNP

types from the pan-African clade are found widespread throughout Africa

while the ISE-SNP types of the Gabonese/Cameroonian clade are much rarer,

and are found in a more restricted area (Figure 2.4), which suggests that the

latter clade evolved more recently. Alternatively, this may also be the result of

a sampling artifact; indeed the Spearman’s rank correlation indicated that the

higher the sampling effort per country the more ISE-SNP types are found.

However, the entirety of the Gabonese/Cameroonian region in itself was well

sampled, with five isolates/clinical samples belonging to the

Gabonese/Cameroonian clade and 25 isolates/clinical samples to the Pan

African clade (Table 2.1 and 2.2, Figure 2.3). Furthermore, the fact that we did

not encounter ISE-SNP types of the Gabonese/Cameroonian clade in

neighboring countries like DRC, where sampling was higher, also suggests that

the ISE-SNP types belonging to the Gabonese/Cameroonian clade are not only

rare but also have a limited distribution. Interestingly, the only ISE-SNP 1

isolate from Cameroon (ITM_120140) came from a patient from Bankim, a

district located along the Mapé river (Sanaga basin), while other studied

isolates all came from around the Nyong river basin. Bankim has been recently

identified as an additional BU endemic area in Cameroon. However, whether

BU was emerging in Bankim, or constitutes a newly recognized preexisting

disease focus, remains unclear [68, 113].

The Gabonese/Cameroonian clade was found to form a strongly supported

monophyletic group with Papua New Guinean ISE-SNP type 8, which is

distinctly unrelated to other ISE-SNP types found in Southeast Asia. Using a

different genotyping technique, the relatedness of a Papua New Guinean

clinical isolate (not included in this study) to African rather than to Southeast

Asian clinical isolates has been reported elsewhere [114]. The process

(historical events, restricted bacterial gene flow, etc.) that led to this inter-

continental association of ISE-SNP haplotypes remains elusive.

Page | 55

In this report we have analyzed a large collection of isolates representative of

the African M. ulcerans population to analyze its population structure

accurately and appropriately. The panel used in this study is, to our

knowledge, the most comprehensive one used so far. It covered disease foci

from all 11 well-documented BU-endemic countries ranging from West,

Central to East Africa. Six countries (Burkina Faso, Equatorial Guinea, Guinea,

Kenya, Liberia, and South Sudan) that have reported a limited number of BU

cases in the past [5] were not included in the study as we were unable to

include specimens, nor isolates, from them. Moreover, cases from Central

African Republic, Senegal and Sierra Leone were never confirmed by

laboratory tests [5]. Although we tried to maximize spatial diversity within our

panel some countries are better represented than others again due to the

limited availability of clinical isolates. We might have missed some ISE-SNP

types in these countries because there was a significant relationship between

the sampling effort per country and the amount of different ISE-SNP types

identified per country. Because of all these limitations we successfully

optimized the genotyping PCR technique for application directly on clinical

specimens, which allowed us to include clinical specimens from certain

geographical regions of Gabon and Nigeria of which no M. ulcerans isolates

were available (Table 2.2).

To our knowledge ISE-SNP typing currently yields the greatest resolution

within M. ulcerans, save for whole genome sequencing. The method may be

an easy, low cost, powerful, reliable, and reproducible tool for reference

laboratories to assist in the tracking of M. ulcerans ISE-SNP types of

epidemiological studies on a continental scale [10].

Because African M. ulcerans shows such low genetic variation, further studies

require a whole genome approach to comprehensively evaluate the genetic

diversity, the evolution, and the phylogenetic relatedness of African M.

ulcerans and to delineate the exact origin and spread of the pathogen at the

local and the continental level. It is specifically the paucity of genetic diversity

and the sequential order of the genetic changes that have occurred between

individual isolates that render M. ulcerans as such a promising model to reveal

evolutionary bacterial mechanisms. Furthermore, given the comprehensive

nature of full genome data, sequences could also serve in large scale micro-

Page | 56

epidemiological studies focusing on the elucidation of transmission pathways

and relevant reservoirs of M. ulcerans. Indeed, different studies in

mycobacterial genomics [19, 112, 115] already showed that, at the whole-

genome level, substantial genetic variation exists in African M. ulcerans, which

can be exploited for phylogenetically robust strain classification. In order to

capture as much diversity as possible and to minimize phylogenetic discovery

bias [116] in such impending large sequencing endeavors it will be desirable to

select representatives from all the central and radial ISE-SNP types defined in

this study.

2.6 Acknowledgements

The present study pays tribute to the extensive collection of M. ulcerans

isolates generated over decades by Emerita Prof. dr. Françoise Portaels and

her collaborators in endemic countries, who initiated and fueled research into

the pathogenesis, diagnosis and management of M. ulcerans disease.

Koen Vandelannoote was supported by a PhD-grant of the Flemish

Interuniversity Council - University Development Cooperation (Belgium).

Funding for this work was provided by the Stop Buruli Consortium supported

by the UBS Optimus Foundation, the European Community's Seventh

Framework Programme under grant agreement n° 241500 (BURULIVAC), the

European Commission (project no. INCO-CT-2005-051476-BURULICO), and the

Fund for Scientific Research Flanders (Belgium) (FWO grant n° G.0321.07N).

The funders had no role in study design, data collection and analysis, decision

to publish, or preparation of the manuscript.

We thank Tim Stinear for helpful discussions and a critical comments to the

manuscript. We thank Pim de Rijk, Krista Fissette, Elie Nduwamahoro, and

Anita Van Aerde (ITM) for their excellent technical assistance. We thank three

anonymous reviewers for their constructive and insightful comments, which

helped us to improve the manuscript

Page | 57

Chapter 3

Whole genome comparisons suggest random

distribution of Mycobacterium ulcerans

genotypes in a Buruli ulcer endemic region of

Ghana


Anthony S. Ablordey, Koen Vandelannoote, Isaac A. Frimpong, Evans K. Ahortor,

Nana Ama Amissah, Miriam Eddyani, Lies Durnez, Françoise Portaels, Bouke C. de

Jong, Herwig Leirs, Jessica L. Porter, Kirstie M. Mangas, Margaret M. C. Lam, Andrew

Buultjens, Torsten Seemann, Nicholas J. Tobias, and Timothy P. Stinear

Whole genome comparisons suggest random distribution of Mycobacterium ulcerans

genotypes in a Buruli ulcer endemic region of Ghana

PLoS Neglected Tropical Diseases 2015 Mar 31;9(3):e0003681

Conceived and designed the experiments: ASA KV TPS.

Performed the experiments: ASA KV IAF EKA NAA JLP KMM MMCL NJT TPS. Analyzed

the data: ASA KV IAF NAA ME LD FP BCdJ HL AB NJT TPS.

Contributed reagents/materials/analysis tools: ME LD FP BCdJ HL TS.

Wrote the paper: ASA KV TPS.

Page | 58

3.1 Abstract

Efforts to control the spread of Buruli ulcer - an emerging ulcerative skin

infection caused by Mycobacterium ulcerans - have been hampered by our

poor understanding of reservoirs and transmission. To help address this issue,

we compared whole genomes from 18 clinical M. ulcerans isolates from a

30km² region within the Asante Akim North District, Ashanti region, Ghana,

with 15 other M. ulcerans isolates from elsewhere in Ghana and the

surrounding countries of Ivory Coast, Togo, Benin and Nigeria. Contrary to our

expectations of finding minor DNA sequence variations among isolates

representing a single M. ulcerans circulating genotype, we found instead two

distinct genotypes. One genotype was closely related to isolates from

neighbouring regions of Amansie West and Densu, consistent with the

predicted local endemic clone, but the second genotype (separated by 138

single nucleotide polymorphisms (SNPs) from other Ghanaian strains) most

closely matched M. ulcerans from Nigeria, suggesting another introduction of

M. ulcerans to Ghana, perhaps from that country. Both the exotic genotype

and the local Ghanaian genotype displayed highly restricted intra-strain

genetic variation, with less than 50 SNP differences across a 5.2 Mbp core

genome within each genotype. Interestingly, there was no discernible spatial

clustering of genotypes at the local village scale. Interviews revealed no

obvious epidemiological links among BU patients who had been infected with

identical M. ulcerans genotypes but lived in geographically separate villages.

We conclude that M. ulcerans is spread widely across the region, with

multiple genotypes present in any one area. These data give us new

perspectives on the behavior of possible reservoirs and subsequent

transmission mechanisms of M. ulcerans. These observations also show for

the first time that M. ulcerans can be mobilized, introduced to a new area and

then spread within a population. Potential reservoirs of M. ulcerans thus

might include humans, or perhaps M. ulcerans-infected animals such as

livestock that move regularly between countries.

Page | 59

3.2 Introduction

Buruli ulcer (BU) is a neglected tropical disease caused by infection

with Mycobacterium ulcerans. Each year 5000-6000 cases are reported from

15 of the 33 countries where BU cases have been reported, predominantly

from rural regions across West and Central Africa [117]. The disease involves

subcutaneous tissue and has several manifestations but necrotic skin ulcers

are a common presentation, caused by the proliferation of bacteria beneath

the dermis by virtue of a secreted bioactive lipid called mycolactone [14]. The

role of mycolactone in the natural ecology of M. ulcerans is not understood,

but it has been shown to possess several specific activities against mammalian

cells from activating actin polymerization, blocking secreted protein

translocation, to interacting with neuronal angiotensin type II receptors

causing hypoesthesia [15, 16]. These collective biological activities of

mycolactone, while diverse, might collectively help explain the tissue

destruction, lack of inflammation, and painlessness associated with BU. BU is

rarely fatal and early diagnosis followed by combined antibiotic therapy

(rifampicin and streptomycin) is key to preventing complications that can arise

from severe skin ulceration [11].

Epidemiological studies frequently link BU occurrence with low-lying and

wetland areas and human-to-human transmission seems rare, suggesting an

environmental source of the mycobacterium [29, 68, 113, 118-131].

Frustratingly however, the environmental reservoir(s) and mode(s) of

transmission of M. ulcerans remain unknown. M. ulcerans has the genomic

signature of a niche-adapted mycobacterium, indicating that it is unlikely to

be found free-living in diverse aquatic (or other) environments, but more

likely in close association with a host organism. In south eastern Australia,

native marsupials have been identified as both susceptible hosts and

reservoirs of M. ulcerans, with high numbers of the bacteria shed in the feces

of infected animals. Mosquitoes have also been found to harbor the bacteria

in this region and a zoonotic model of disease transmission has been

proposed involving possums, biting insects and humans [42, 61]. No such

animal reservoir has yet been identified in African BU endemic areas and

studies of BU lesion distribution are thought not consistent with mosquito

biting patterns [68, 89]. On the other hand, case-control studies in Cameroon

Page | 60

have shown that bed nets are protective, supporting a role for insects in

transmission [132].

A feature of M. ulcerans is the close correlation between genotype and the

geographic origin of a strain, but its restricted genetic diversity has limited the

application of traditional molecular epidemiological methods such as VNTR-

typing to discriminate between isolates at the village or even regional scales.

The advent of low cost genomics has opened up new possibilities to explore

and track the movement and spread of this pathogen within communities

[112, 133].

Agogo is the principal town of 30,000 inhabitants in the Asante Akim North

(AAN) district within the Ashanti region of Ghana and BU has been reported in

about half of the sixty-four communities in this district since mid-1975 [121].

The AAN district covers an area of 650 km2 in the forest belt of Ghana and it is

the third most endemic district in Ghana [134]. Five of the communities

(Ananekrom, Serebouso, Nshyieso, Serebuoso and Dukusen) in this district are

among the communities reported with the highest burden of the disease in

Ghana [134]. About 120 laboratory-confirmed new cases are reported

annually in this district [134]. Subsistence farming and petty trading are the

principal occupations of inhabitants of these endemic communities. People

generally live in simple dwellings constructed from local materials. Houses are

often close together with 3-5 households in a compound. Many inhabitants

raise animals such as goats, sheep, and pigs in the immediate vicinity of their

houses. Farming is the main occupation with some people engaged in fishing

and petty trading. Farms may be distant, ranging 5-20 km from a given

domicile. Fishing is usually undertaken close to home. Water sources are of

two types. Water for drinking and cooking is usually fetched from bore holes

fitted with mechanical pumps, within or near a village. Water for bathing and

domestic chores such as washing of clothes is drawn from local natural water

sources (rivers, streams, ponds). These natural sources are usually no more

than 500 metres from a given village.

In this study we sequenced and compared the genomes of 18 M. ulcerans

isolates obtained from 10 BU endemic villages in the AAN district and

uncovered genetic evidence supporting the introduction of a foreign clone

Page | 61

of M. ulcerans to this region. This observation indicates that M. ulcerans can

be mobilized and spread throughout a region, indicating that reservoirs of the

bacterium are themselves potentially highly mobile.

3.3 Methods

Ethics statement

M. ulcerans isolates were obtained from BU diagnostic samples, collected as

part of routine laboratory diagnosis. Ethical approval to interview patients and

use bacterial isolates resulting from diagnostic specimens for research was

obtained from the ethical review board of the Noguchi Memorial Institute for

Medical Research, University of Ghana, Legon, Accra, Ghana (FWA 00001824),

with written informed consent obtained from all adult patients or the

parents/guardians of the participating children.

Study site and case reporting

The study was carried out in ten endemic villages including Ananekrom,

Nshyieso, Serebouso, Dukusen, Afreserie, Afreserie OK, Baama,

Nysonyameye, Kwame Addo and Bebuso, in the Asante Akim North (AAN)

district of Ghana (Table 3.1). These are small villages and hamlets, 5 to 10 km

from each other with populations between 120-1500 inhabitants. Ananekrom

is the largest of these communities and is the closest (15 km) to the district

capital, Agogo. An asphalt road connects Agogo to Ananekrom, Dukusen and

Afriserie, while the other communities are located off this main road and are

connected to each other by unmade roads and foot-tracks. A community

health centre Ananefromh (near Ananekrom) is usually the first point of call

for patients seeking medical treatment. Patients suspected of having BU are

referred to the Agogo Presbyterian Hospital for diagnosis and treatment.

Patient information including name and place of residence were obtained

from hospital records and patients were visited in their homes for more

detailed interviews that included questions about possible travel to other BU

endemic areas outside the AAN district. GPS coordinates in the vicinity of each

patient’s residence were recorded in order to map the spatial distribution of

Page | 62

cases in the villages, based on the assumption that the patient acquired their

infection near their domicile.

Table 3.1: M. ulcerans isolate and DNA sequencing information. Notes: 1: parentheses

indicate alternate strain code, 2: expressed as latitude and longitude, 3: Democratic Republic

of the Congo, 4: reference genome

No. Strain ID1 Isolation date

Patient Age

(years)

Patient

Gender Village/Genotype District/ Country Location

2 Seq Plat. Avg cov. ENA Ref.

1 S15 2/02/2010 5 M Ananekrom/

Agogo-1 Ashanti/Ghana

6.91481, -

1.01658 SE PGM 106

SAMEA321257

8 This study

2 S43 9/05/2010 13 F Ananekrom/

Agogo-2 Ashanti /Ghana

6.91481, -

1.01658 SE PGM 120

SAMEA321257

0 This study

3 S38 23/06/2010 12 M Serebuso/ Agogo-

1 Ashanti /Ghana

6.94838, -

1.02373 SE PGM 104

SAMEA321258

0 This study

4 F64 8/09/2010 3 M Nsonyameye/


6.93744, -

0.96464 SE PGM 99

SAMEA321257

4 This study

5 F70 15/09/2010 32 F Baama/ Agogo-1 Ashanti /Ghana 6.96810, -

0.94013 SE PGM 61

SAMEA321258

2 This study

6 1510 (F79) 22/09/2010 28 F Bebuso/ Agogo-2 Ashanti /Ghana 6.88909, -

0.97209 PE Illumina 81

SAMEA321256

5 This study

7 F75 7/10/2010 18 F Afriserie OK/


7.01992, -

0.92325 SE PGM 68

SAMEA321258

1 This study

8 F85 20/10/2010 28 F Nshyieso/ Agogo-2 Ashanti /Ghana 6.95910, -

1.01860 SE PGM 71

SAMEA321256

6 This study

9 S77 24/11/2010 3 F Serebuso/ Agogo-

2 Ashanti /Ghana

6.94838, -

1.02373 SE PGM 83

SAMEA321257

1 This study

10 2610 (F92) 24/12/2010 14 F Nsonyameye/


6.93744, -


SAMEA321256

7 This study

11 F13 2/02/2011 8 F Bebosu Ado/


6.88909, -

0.97209 SE PGM 92

SAMEA321258

3 This study

12 F65 2/09/2011 28 F Ananekrom/


6.91481, -

1.01658 SE PGM 89

SAMEA321257

6 This study

13 F74 21/09/2011 38 M Dukusen/ Agogo-2 Ashanti /Ghana 6.97683, -

0.98278 SE PGM 46

SAMEA321257

2 This study

14 S72 12/12/2011 11 M Afriserie/ Agogo-1 Ashanti /Ghana 7.01992, -

0.92325 SE PGM 128

SAMEA321257

5 This study

15 612 (F36) 18/07/2012 37 M Ananekrom/


6.91481, -


SAMEA321256

8 This study

16 212 (F3) 8/02/2012 11 F Serebuso/ Agogo-

2 Ashanti /Ghana

6.94838, -


SAMEA321257

7 This study

17 712 (S24) 2/08/2012 40 F Wenamda/ Agogo-

2 Volta/Ghana

7.07738,


SAMEA321256

9 This study

18 412 (F37) 2/08/2012 12 M Ananekrom/


6.91481, -


SAMEA321258

4 This study

19 IC21 jul/00

Bondoukou Cote d’Ivoire 8.03333, -

2.8000 SE PGM 79

SAMEA321257

9 This study

20 IC38 jul/00

Dimbokro Cote d’Ivoire 6.64445, -

4.70540 SE PGM 63

SAMEA321257

3 This study

21 ITM000909 2000

Tchekpo Deve Maritime/Togo 6.48434,

1.369718 SE PGM 45

[135]

22 ITM991591 1999

Anagali Maritime/Togo 6.48434,

1.369718 SE PGM 55

[135]

23 ITM102686 2010

M Ibadan Oyo State/Nigeria 7.50194,

3.982327 SE PGM 51

[135]

24 ITM5151 1971

Maniema/DRC3

-4.18655,

26.43937 SE PGM 51

[135]

25 NM14.01 2001

Densu/Ghana 5.72532, -


[19]

26 NM43.02 2002



[19]

27 NM49.02 2002



[19]

28 NM54.02 2002



[19]

29 NM33.04 2004

Amansie West/Ghana 6.68712, -


[19]

30 Agy994 aug/99 5 F

Amansie West/Ghana

6.68712, -

1.62197 Sanger

[12]

31 ITM980535 1998 12 F Djigbé Atlantique/Benin 6.885076,


[19]

Page | 63

No. Strain ID1 Isolation date

Patient Age

(years)

Patient

Gender Village/Genotype District/ Country Location

2 Seq Plat. Avg cov. ENA Ref.

32 ITM000945 2000 21 F Hwegoudo Atlantique/Benin 6.7234,


[19]

33 ITM001506 2000 6 F Wokon Zou/Benin 7.099851,


[19]

Culture isolation and identification of M. ulcerans from patients

The isolates examined in this study are listed in Table 3.1 and were recovered

from fine needle aspirates (FNA) or swabs, obtained from pre-ulcerative

lesions and ulcers respectively. Specimens were stored in transport medium

and PBS and transported in cool boxes to the Noguchi Memorial Institute for

Medical Research (NMIMR) for diagnosis [136, 137]. Tubes containing swabs

were vortexed in 3 ml of transport medium for 30 sec and the swabs

removed. A volume of 250μl of the transport medium from either specimen

type was transferred into 1.5 ml microfuge tubes and decontaminated using

the oxalic acid method as previously described [138]. The pellets were

resuspended in 100 μl phosphate buffered-saline (PBS) and 100 μl volume of

the decontaminated sample was inoculated onto Löwenstein Jensen (LJ)

slopes and incubated at 33°C. The cultures were observed weekly for growth.

Suspected M. ulcerans colonies were harvested and DNA extracted as

described above [139]. The DNA extract was tested with the IS2404 PCR for

the identification of M. ulcerans [140]. Colonies positive for IS2404 were

suspended in 1 ml of Middlebrook 7H9 broth and stored at -80°C. All 18

bacterial samples analyzed were selected from this stored collection and were

subcultured on LJ medium and DNA for whole genome sequencing was

extracted from resulting growth as described [139]. The isolation date refers

to the date when colonies became visible on LJ medium following primary

cultivation.

Genome sequencing and analysis

DNA sequencing was performed using two methods. The Ion Torrent Personal

Genome Machine was employed, with a 316 chip and 200bp single-end

sequencing chemistry (Life Technologies). Genomic libraries for Ion Torrent

sequencing were prepared using Ion Express, with size selection using the

Pippin Prep (Sage Sciences) and emulsion PCR run using a One-Touch

Page | 64

instrument (Life Technologies). The Illumina MiSeq was also used, with

Nextera XP library preparation and 2x250 bp sequencing chemistry. Read data

for the study isolates have been deposited in the European Nucleotide

Archive (ENA) under accession ERA401876. Prior to further analysis, reads

were filtered to remove those containing ambiguous base calls, any reads less

than 50 nucleotides in length, and containing only homopolymers. All reads

were furthermore trimmed removing residual ligated Nextera adaptors and

low quality bases (less than Q10) at the 3' end. Resulting sequence Fastq

sequence read files from either platform were subjected to read-mapping to

the M. ulcerans Agy99 reference genome (Genbank accession

number CP000325) using Bowtie2 v2.1.0 [141] with default parameters and

consensus calling to identify SNPs (indels excluded) using Nesoni v0.109, a

Python utility that uses the reads from each genome aligned to the core

genome to construct a tally of putative differences at each nucleotide position

(including substitutions, insertions, and deletions)

(www.bioinformatics.net.au). Those positions in the Agy99 reference genome

that were covered by at least 3 reads from every isolate defined a core

genome. Note that the pMUM001 plasmid (required for mycolactone

synthesis) was not included in the reference genome [13]. Testing of the

plasmid sequences revealed less than 10 polymorphic sites among the

genomes under investigation and the highly repetitive sequence structure of

the mycolactone genes impaired unambiguous read-mapping. An unpaired t

test with Welch’s correction was used to assess the differences between

mean nucleotide pairwise identities for different groups of genomes. The null

hypothesis (no difference between means) was rejected for p<0.01. The

inputs for subsequent phylogenomic analyses were the nucleotide sequence

alignments of the concatenated variable nucleotide positions for the core

genome among all isolates. A maximum-likelihood (ML) phylogeny was

inferred using RAxML v 7.2.8, with the GTR model of nucleotide substitution

(plugin within Geneious v 8.0.4). We performed 1000 rapid pseudo-replicate

bootstrap analyses to assess support for the ML phylogeny. We used

Consensus-Tree-Builder (Geneious v8.0.4) to collapse nodes in the tree with

bootstrap values below a set threshold of 70%. The resulting phylogenomic

tree was exported in Newick format and visualized using FigTree v1.4.0

(tree.bio.ed.ac.uk/software/figtree). A haplotype network was derived using

Page | 65

the median-joining algorithm as implemented in SplitsTree v4.13.1 [103, 142].

A correction to the source attribution of the M. ulcerans Agy99 reference

genome was also made in the course of this study, where it was realized that

this isolate was actually obtained from a patient attending St Martin’s

Hospital in Agroyesum (Amansie West) and not the Ga District Hospital as

originally published [12] (K. Asiedu and J., Hayman pers comms), thereby

explaining the inconsistent geographic clustering reported in previous

molecular epidemiological studies [19, 112, 143].

3.4 Results

Genome sequence comparisons of 18 M. ulcerans isolates from Agogo

region

Eighteen M. ulcerans isolates were randomly selected for whole genome

sequencing. The isolates represented 20% (total of 92 isolates from 2010-

2012) of all culture-confirmed BU cases referred to the Agogo Presbyterian

Hospital between 2010 and 2012 (Table 3.1). There were no differences in

colony phenotype or growth characteristics among the isolates. The DNA

sequence reads for each genome were mapped to the M. ulcerans Agy99

reference sequence. Sequencing and read-mapping summary statistics are

given in Table 3.1. In addition to the 18 Agogo isolates sequenced here, 15

other genomes (including some previously described) were included in

comparisons making a total of 33 isolates (Table 3.1). These additional

genomes were from M. ulcerans isolates in other regions of Ghana and from

surrounding countries to provide appropriate genetic context for interpreting

the diversity and evolution of M. ulcerans isolates from around Agogo. Read-

mapping and SNP identification revealed 320 variable nucleotide positions

across a 5.2Mb core genome for the 33 isolates. A phylogeny was inferred

from this alignment, showing the clustering typical of M. ulcerans genotypes

with geographic origin (Figure 3.1). A separate SNP alignment was performed

taking the genome sequences for only the 18 isolates from the Agogo region,

and 10 of them (called Agogo-1) clustered with isolates from the neighboring

district of Amansie West and also the Ivory Coast, the country which borders

this region to the west (Figure 3.1). This close relationship is indicative of a

local clone that has spread and persisted within the greater region for some

Page | 66

time. Unexpectedly however, this analysis also revealed the presence of a

second distinct M. ulcerans genotype co-circulating with Agogo-1. This second

genotype (called Agogo-2) was substantially more diverse from all other

Ghanaian M. ulcerans genotypes (138 SNPs), suggesting the re-introduction

of M. ulcerans to the Agogo region, potentially from a source outside Ghana

(Figure 3.1, Table 3.S1). The intra-genotype variation within either cluster was

low. The mean nucleotide pairwise identity was 94.7% (SEM ± 0.4) for Agogo-

1 versus 97.2% (SEM ± 0.4) for Agogo 2. The mean pairwise nucleotide

identity was significantly lower for Agogo-2 genomes compared with Agogo-1

(p<0.001).

Figure 3.1: Genetic relationship among the 33 M. ulcerans isolates used in this study. A

maximum-likelihood consensus phylogeny was inferred based on whole genome alignments

of each of the isolates against the M. ulcerans Agy99 reference genome. The alignment file

from pairwise comparisons of the resulting 320 variable nucleotide positions was used as

input for RaxML. Nodes with less than 70% bootstrap support (1000 replicates) were

collapsed.

Page | 67

A likely foreign origin for Agogo-2

To investigate the possible origin of the Agogo-2 isolates we compared SNP

profiles among our panel of M. ulcerans genomes from across West and

Central Africa. The closest match obtained was to isolate ITM102686,

obtained from a patient originating from Ibadan, Nigeria, with 29 SNPs

different when only this genome was compared to the Agogo-2 cluster. This

close association may indicate that Nigeria was the source of the Agogo-2

cluster. Some circumspection is needed when interpreting these data, as only

two M. ulcerans genomes were sampled from countries east of Benin. There is

however a compelling patient history behind this isolate to support Nigeria as

the correct origin. The Caucasian patient, a long-term resident in Ibadan and

an employee of a non-government organisation, believes he was infected on

an Ibadan golf course, when he was bitten by black biting flies (his description

suggests they may have been moth flies [Psychodidae]) that began plaguing

the course when ground works started adjacent to a lake on the course. The

patient developed a painful ulcer on the site of the insect bites. A couple of

months later he developed a second ulcer on an adjacent site on the same

limb that was microbiologically diagnosed as a Buruli ulcer. This patient

history, combined with the documented cases of BU in Ibadan, with cases

occurring around the Ibadan University campus and other nearby institutions

[144], support Ibadan as the likely origin of M. ulcerans isolate ITM102686.

M. ulcerans genotype clustering breaks down at a local scale

We next explored the distribution of M. ulcerans genotypes in the Agogo

region at the village scale and observed no obvious pattern or relationship

between genotype, patient, strain and village (Figure 3.2). There is complete

intermixing of Agogo-1 and Agogo-2 clusters amongst the population.

Median-joining-network analysis suggested the independent radiation of the

two clusters throughout the region (Figure 3.2). Furthermore, within either

cluster there was a broad distribution of cluster subtypes across the region.

For example, isolates F70 and S38 (Agogo-1) have identical SNP profiles but

the patients came from Baama and Serebouso, villages separated by 10-15

km. Similarly isolates F74 and 1510 (Agogo-2), came from patients who live in

two different villages (Figure 3.2). Patient interviews did not identify any

Page | 68

travel histories or other epidemiological links that might explain these

distribution patterns. An 11-year old girl from Serebouso was the third child

within her family to have BU (isolate 212, November 2012), eight years after

two of her siblings had the disease. The family of this child lived very close to

that of another BU patient, a 3 year-old infant (isolate S77, February 2010).

Both isolates belonged to the Agogo-2 cluster but their genome sequence

differed in nine nucleotide positions, a significant amount of genetic variation

given that S77 shares a near identical genotype with F74 and 1510. Again, we

could not identify any specific activity or travel history such as attending a

common community event that was shared by Agogo-2 genotype patients.

These data suggest that (i) the disease is acquired locally, (ii) multiple M.

ulcerans genotypes are circulating simultaneously within the local region and

(iii) a single clone can have the propensity to spread through a region. Further

support for local acquisition of infection comes from observations of infants

with no travel history with BU such as a locally-born 2-year old infant from

Ananekrom identified over the time of this study.

Page | 69

Figure 3.2: Micro-molecular epidemiology of BU in the Asante Akim North District revealed by

M. ulcerans whole genome sequence comparisons. (A) Median-joining network graph

showing the genetic relationship between 18 M. ulcerans clinical isolates comprising the

Agogo-1 and Agogo-2 genotypes (shaded), inferred from whole genome sequence

alignments. Node sizes in the graph are proportional to the frequency of genotype occurrence

and have been colour-coded accordingly. Edges are labelled in red with the number of

mutational steps between each node. (B) Map of Asante Akim North District study area,

showing the location of endemic villages and the origin of each of the 18 BU cases, with a

coloured circle corresponding with the genotype displayed in the network graph in (A). The

number “2” within some coloured circles indicates an Agogo-2 genotype.

Page | 70

3.5 Discussion

The clonal population structure of M. ulcerans has made identifying and

comparing genetic variation in isolates at anything less than a continental

scale very difficult. Here we have used the high resolution afforded by

comparative genomics to explore the molecular epidemiology of BU at the

regional and village scale. Like recent studies using a single polymorphic

genetic locus or whole genome sequence comparisons to assess M. ulcerans

genetic diversity across a range of African countries, we found a highly

significant relationship between the genotype of an isolate and its geographic

origin at a national and regional scale [19, 135]. These repeated observations

indicate that M. ulcerans, when introduced to an area, remains localized and

isolated for a sufficiently long period to allow mutations to become fixed in

the bacterial population and a local genotype to evolve. It is reasonable to

infer therefore, that the environmental reservoirs of the bacterium in these

areas are also likely to be somewhat localized and isolated.

However, the current study has shown for the first time how this focal

distribution pattern breaks down at a local scale with the presence of identical

genotypes appearing concurrently in separate areas of the same district.

There was no discernible distribution pattern for either the Agogo-1 or Agogo-

2 genetic clusters, with both M. ulcerans genotypes appearing at the same

times and within the same villages across the region. Interestingly, there were

several examples of isolates with identical genome sequences (e.g. isolates

F74, 1510 or F85, F65) that were obtained from patients living in four

different villages, each separated by distances in excess of 10km (Figure 3.2).

There are several potential explanations for these patterns. The bacteria (or a

vector spreading the bacteria) may be widely distributed across the region

and infections are being acquired locally, or it may be that people are

traveling and becoming infected from a common point source. Patient

interviews and travel histories did not reveal any common activity that might

explain a point-source transmission scenario, although the long incubation

time for this disease (4-months) is likely to make recall of any such events

unreliable [145]. However, on balance the former scenario seems most likely,

and we suggest that each genotype of M. ulcerans has now spread equally

widely across the region. If this assumption is correct, then the lack of genetic

Page | 71

variation among isolates suggests that the spread of M. ulcerans throughout

the region has occurred relatively rapidly, with insufficient time elapsed for

mutations to accumulate. Reliable mutation rates for M. ulcerans have not

been established and some solid data here would allow inferences regarding

the time particular clones have been extant within a population.

To our knowledge, this is the first report to employ whole genome sequencing

to explore the molecular epidemiology of BU at a local scale. A previous study

utilizing high-resolution SNP assays to explore M.ulcerans genetic variation

did uncover some suggestion of local genotype clustering and a recent report

used VNTR to examine the link between human and environmental sources

of M. ulcerans [66, 112]. However such approaches rely on variable

nucleotides that have been defined from a limited reference genome set. If

this reference genome set does not represent the genetic variation of the

isolates under investigation then data analysis can be flawed, with

phenomena such as long-branch attraction and phylogenetic discovery bias

confounding analyses [146]. Whole genome sequencing and comparisons of

all isolates under investigation as in our study here overcomes the potential

weaknesses of targeted SNP-based typing. SNP-typing could however be

employed to classify patient samples as Agogo-1 and Agogo-2 genotypes

without relying on sequencing of cultured isolates, as culture sensitivity is only

around 30%, depending on transport duration. Future studies could thus

search for clinical phenotypes between these two distinct bacterial genotypes,

although no differences were observed in pathology or treatment outcomes

among the patients associated with this study.

There are interesting parallels between M. ulcerans and Mycobacterium

leprae, the causative agent of leprosy, where genomics has shown that the

leprosy bacillus is another example of a niche-adapted, highly clonal, zoonotic

mycobacterial pathogen, with the potential to spread from environment-to-

human [56, 147-149]. Mycobacterium tuberculosis might also be considered in

a similar context, with genomic population analysis also suggesting

interactions among genetically distinct M. tuberculosis lineages [150, 151].

One potential issue arising from this study is the risk of incorrectly attributing

Nigeria as the origin M. ulcerans genome sequence ITM102686, as it

Page | 72

represents only one isolate. While the patient history makes a persuasive

argument for Ibadan as the source of the infection, additional M. ulcerans

isolates are clearly required from patients in different BU endemic regions of

Nigeria and surrounding countries, to further explore the relationship and

disease transmission patterns we propose here.

Regardless of the precise origin of Agogo-2 isolates, the data presented here

suggest that M. ulcerans can be introduced into a region and then be spread

extensively. How might M. ulcerans be imported into a region? We speculate

that movements of people or perhaps animals between countries could be

one likely means, where infected individuals with BU lesions that can contain

very high bacterial burdens might inadvertently contaminate aquatic

environments during bathing or other water contact activities. Now is the

time to undertake more intensive and extensive whole-genome M. ulcerans

sequencing surveys across West Africa, to assess the extent of genotype

admixture such as we’ve revealed here. Enriching our genome data will also

inform other research programs that are identifying reservoirs of M. ulcerans,

leading to the new knowledge required to design interventions and stop the

spread of BU.

3.6 Supporting Information

Table 3.S1: Summary of the 127 variable nucleotides that differentiate the

Agogo-1 and Agogo-2 clusters, with predicted CDS consequences, based on

whole genome sequence comparisons.

� Object too large to print. Available online (http://tinyurl.com/h9xpemc).

3.7 Acknowledgments

We thank Conor Meehan for critical review of the analysis methods and

Laurent Marsollier for the provision of bacterial isolates.

3.8 Funding Statement

This study was supported in part by the Stop Buruli Initiative

(www.stopburuli.org, UBS Optimus Foundation, Zurich, Switzerland) and the

Page | 73

Flemish Interuniversity Council -University Development Cooperation

(www.vliruos.be, VLIR-UOS). TPS was supported by a Fellowship from the

National Health and Medical Research Council of Australia

(www.nhmrc.gov.au). KV was supported by a VLADOC PhD scholarship of

VLIR-UOS (Belgium). The funders had no role in study design, data collection

and analysis, decision to publish or preparations of the manuscript.

3.9 Data Availability

All relevant data are within the paper and its Supporting Information files. All

DNA sequence files are available from the European Nucleotide Archive (ENA)

under accession ERA401876.

Page | 74

Page | 75

Chapter 4

Population genomics aligns the spread of

Mycobacterium ulcerans and Buruli ulcer with

the scramble for Africa.

Koen Vandelannoote, Conor J. Meehan, Miriam Eddyani, Dissou Affolabi, Delphin

Mavinga Phanzu, Sara Eyangoh, Kurt Jordaens, Françoise Portaels, Kirstie Mangas,

Torsten Seemann, Laurent Marsollier, Estelle Marion, Annick Chauty, Jordi Landier,

Arnaud Fontanet, Herwig Leirs, Timothy P. Stinear, Bouke C. de Jong

Designed research: KV, CM, ME, FP, HL, TPS, BJ

Performed research: KV, KM, TPS

Contributed new reagents or analytic tools: KV, CM, DA, DP, SE, TS, LM, EM, AC, JL,

AF, TPS

Analysed data: KV, CM, TPS

Wrote the paper: KV, CM, TPS, KJ, BJ

Page | 76

4.1 Abstract

Buruli ulcer (BU) is an insidious neglected tropical disease. Cases are reported

around the world but the rural regions of West and Central Africa are most

affected. How BU is transmitted and spreads has remained a mystery, even

though the causative agent, Mycobacterium ulcerans, has been known for

more than 70 years. Here, using the tools of population genomics, we

reconstruct the evolutionary history of M. ulcerans by comparing 165 isolates

spanning 48 years and representing 11 endemic countries across Africa. The

genetic diversity of African M. ulcerans was found to be restricted due to the

bacterium’s slow substitution rate coupled with its relatively recent origin. We

identified two specific M. ulcerans lineages within the African continent, and

showed that M. ulcerans lineage Mu_A1 existed in Africa for several hundreds

of years, unlike lineage Mu_A2, which was introduced much more recently

during the early 20th century. Additionally, we observed that specific M.

ulcerans epidemic Mu_A1 clones were introduced during the same time

period in the three hydrological basins that were well covered in our panel.

The estimated time span of the introduction events coincides with the Neo-

imperialism period, during which time the European colonial powers divided

the African continent among themselves. Using this temporal association, and

in the absence of a known BU reservoir or –vector on the continent, we

postulate that the so-called “Scramble for Africa” played a significant role in

the spread of the disease across the continent through the displacement of

BU-infected humans.

4.2 Introduction

Buruli ulcer (BU) is a slowly progressing necrotizing disease of skin and

subcutaneous tissue caused by the pathogen Mycobacterium ulcerans [2]. BU

is considered a neglected tropical disease, and in some highly endemic areas it

is more prevalent than the most notorious mycobacterial diseases,

tuberculosis and leprosy [83]. Even though BU can affect all age groups, the

majority of cases occur in children under age 15 [84]. On the African

continent, the first detailed clinical descriptions of ulcers caused by M.

ulcerans have been attributed to Sir Albert Cook, a missionary physician who

worked in Uganda in 1897 [152]. Since this first description BU was reported

Page | 77

in 16 Sub-Saharan African countries over the course of the 20th

and the early

21th

century. Today, more than 30 countries worldwide have reported the

emerging disease, although the highest incidence by far is still observed in

impoverished, rural communities of West and Central Africa [5].

BU is known to occur primarily in foci around rural marshes, wetlands, and

riverine areas [2, 85]. As proximity (but not contact) to these slow flowing or

stagnant water bodies is a known risk factor for M. ulcerans infection [31], it is

generally believed that M. ulcerans is an environmental mycobacterium that

can initiate infection after a micro-trauma of the skin [32, 33]. Indeed, M.

ulcerans DNA has been detected in a variety of aquatic specimens [35, 36], yet

the significance of the detection of M. ulcerans DNA by PCR in environmental

samples remains unclear in the disease ecology of BU [35-37, 39, 86-89]. This

is largely due to the fact that definite evidence for the presence of viable M.

ulcerans in potential environmental reservoirs is lacking owing to the

challenge of culturing the slow growing mycobacterium from non-clinical,

environmental sources [46]. Consequently, the mode of transmission and

non-human M. ulcerans reservoir(s) remain poorly understood [153]. As until

today no animal reservoir for M. ulcerans has been identified in the Afrotropic

ecozone [89], we hypothesize that humans with active, openly discharging BU

lesions may play a pivotal role in the spread of the bacterium.

Multilocus sequence typing analyses [18] and subsequent whole-genome

comparisons [19] indicated that M. ulcerans evolved from a Mycobacterium

marinum progenitor by acquisition of the virulence plasmid pMUM001. This

plasmid harbors genes required for the synthesis of the macrocyclic

polyketide toxin mycolactone [13], which has cytotoxic and

immunosuppressive properties that can cause chronic ulcerative skin lesions

with limited inflammation and thus plays a key role in the pathogenesis of BU

[14]. Both the acquisition of the plasmid and a reductive evolution [12, 20]

suggested that a generalist proto-M. marinum became a specialized

mycobacterium, more adapted to a restricted environment, perhaps within a

vertebrate host. Analysis of the genome suggests that this new niche is likely

to be protected from sunlight, non-anaerobic, osmotically stable, and an

extracellular environment where slow growth, the loss of several

immunogenic proteins, and production of the immunosuppressive molecule,

Page | 78

mycolactone, provide selective advantages [12, 19]. The evolution of M.

ulcerans has been mediated by the insertion sequence element (ISE) IS2404,

which is present in the M. ulcerans genome in ~200 copies [12]. For some M.

ulcerans lineages a second ISE, IS2606, is also present in a high copy number

(~90 copies). These short, mobile genetic DNA elements promote genetic

rearrangements by modifying gene expression and sequestering genes,

profoundly enhancing mycobacterial genome plasticity [23]. Increased ISE

numbers are a signature for bacteria that have passed through an

evolutionary bottleneck and undergone a lifestyle shift to a new niche,

causing loss of genetic loci that are no longer required for the survival in the

new environment [90]. Subsequent whole-genome comparisons showed that

this “niche-adapted” genomic signature was established in a M. ulcerans

progenitor before its intercontinental dispersal [19].

The clonal population structure of M. ulcerans has meant that conventional

genetic fingerprinting methods have largely failed to differentiate clinical

disease isolates, complicating molecular analyses on the elucidation of the

disease ecology, the population structure, and the evolutionary history of the

pathogen [93]. Whole genome sequencing (WGS) is currently replacing

conventional genotyping methods for M. ulcerans [19, 154-156]. Hence, in the

present study we sequenced and compared the genomes of 165 M. ulcerans

disease isolates originating from multiple African disease foci to gain deeper

insights into the population structure and evolutionary history of the

pathogen, and to untangle the phylogeographic relationships within the

genetically conserved cluster of African M. ulcerans.

4.3 Methods

Bacterial isolates and sequencing

We analyzed a panel of 165 M. ulcerans disease isolates originating from

disease foci in 11 African countries that had been cultured between 1964 and

2012 (Table 4.S1). Isolates were chosen to maximize temporal and spatial

diversity within countries in which more than 20 isolates were available

(Figure 4.S1). Even though most well-documented BU endemic countries were

well represented, we were unable to include isolates from several African

Page | 79

countries (Equatorial Guinea, Kenya, Liberia, Sierra Leone, and South Sudan),

that have reported, if not isolated, (a limited number of) BU cases in the past

[5]. Two isolates from Papua New Guinea (PNG) were included as out-groups

to root the African phylogenetic tree. PNG M. ulcerans was specifically chosen

as it (together with African and other Southeast Asian M. ulcerans) belongs to

the more virulent and distinct “classic” phylogenetic lineage [109], relative to

M. ulcerans isolates elsewhere.

Permission for the study was obtained from the ITM Institutional Review

Board. Isolates were processed and analyzed without use of any patient

identifiers, except for country and village of origin if this information was

available. Based on conventional phenotypic and genotypic methods,

bacterial isolates had previously been assigned to the species M. ulcerans

[157]. Mycobacterial isolates were maintained for prolonged storage at ≤-

70°C in Dubos broth enriched with growth supplement and glycerol. DNA was

obtained by harvesting the growth of three Löwenstein-Jensen (LJ) slants

followed by heat inactivation, mechanical disruption, enzymatic digestion and

DNA purification on a Maxwell 16 automated platform [156].

Index-tagged paired-end sequencing-ready libraries were prepared from

gDNA extracts with the Nextera XT DNA Library Preparation Kit. Genome

sequencing was performed on Illumina HiSeq 2000 and Miseq DNA

sequencers according to the manufacturers’ protocols with 150bp, 250bp or

300bp paired-end sequencing chemistry. Sequencing statistics are provided in

Table 4.S1. The quality of raw Illumina reads was investigated with FastQC

v0.11.3 [158].

Prior to further analysis, reads were cleaned with clip, a tool in the Python

utility toolset Nesoni v0.130 [159]. Reads were filtered to remove those

containing ambiguous base calls, any reads <50 nucleotides in length, and

reads containing only homopolymers. All reads were further trimmed to

remove residual ligated Nextera adaptors and low quality bases (<Q10) at the

3' end. The total amount of read-pairs kept after clipping and their average

read length are summarized for all isolates in Table 4.S1.

Page | 80

Read mapping and SNP / large deletion detection

Read mapping and SNP detection were performed using the Snippy v2.6

pipeline [160]. The Burrows-Wheeler Aligner (BWA) v0.7.12 [161] was used

with default parameters to map clipped read-pairs to two M. ulcerans Agy99

reference genomes: the M. ulcerans Agy99 bacterial chromosome (Genbank:

CP000325) and the M. ulcerans pMUM001 plasmid (Genbank: BX649209).

Due to the unreliability of read mapping in mobile repetitive regions all ISE

elements (IS2404 and IS2606) and all (plasmid-encoded) polyketide synthase

genes were masked in these reference genomes (397 kb / 5,63 Mb - 7% of

Agy99, 118 kb / 174 kb - 67% of pMUM001). After read mapping to M.

ulcerans Agy99 and pMUM001, average read depths were determined with

SAMtools v1.2 [162] and are summarized for all isolates in Table 4.S1. SNPs

were subsequently identified using the variant caller FreeBayes v0.9.21 [163],

with a minimum depth of 10 and a minimum variant allele proportion of 0.9.

Snippy was used to pool all identified SNP positions called in at least one

isolate and interrogate all isolates of the panel at that position. As such a

multiple sequence alignment of “core SNPs” was generated.

The number of reads mapping to unique regions of the plasmid and the

bacterial chromosome were used to roughly infer plasmid copy number. This

was achieved by calculating the ratio of the mode of read depth of all

positions in the plasmid to that of the chromosome: Mo(read depth unique

positions plasmid):Mo(read depth unique positions chromosome) [82].

Large (>1500 bp) chromosomal deletions were detected with Breseq v.0.27.1

[164], a reference-based alignment pipeline that has been specifically

optimized for microbial genomes. Breseq was used with Bowtie2 v.2.2.6 [141]

to map clipped read-pairs to Agy99 and the resulting missing coverage

evidence was used to detect large deleted chromosomal regions.

Population genetic analysis

Bayesian model-based inference of the genetic population structure was

performed using the “Clustering with linked loci” module [165] in BAPS v.6.0

[166]. This particular module takes potential linkage within the employed

Page | 81

molecular information into consideration, which is advisable when performing

genetic mixture analysis on SNP data from a haploid organism. A

concatenation of the core-SNP alignment of both the bacterial chromosome

and the plasmid was loaded as a sequential BAPS formatted file. This entry

was complemented with a “linkage map” file that differentiated two linkage

groups (bacterial chromosome & plasmid). The optimal number of genetically

diverged BAPS-clusters (K) was estimated in our data by running the

estimation algorithm with the prior upper bound of K varying in the range of

3-20. Since the algorithm is stochastic, the analysis was run in 20 replicates for

each value of K as to increase the probability of finding the posterior optimal

clustering with that specific value of K.

QGIS v.2.10 [104] was used to generate the figures on the geographical

distribution of African M. ulcerans. The residence of BU patients at the time of

their clinical visit was represented as points. In the case where residence

information was missing we used the location of the hospital supplying the

sample. The administrative borders of countries were obtained from the

Global Administrative Unit Layers dataset of FAO.

Detection of recombination

Evidence for recombination between different BAPS-clusters was assessed

using several methods as studies show that no single method is optimal,

whereas multiple approaches may maximize the chances of detecting

recombination events [167]. A whole genome alignment was constructed with

Snippy using default parameters. Recombination was assessed using the

pairwise homoplasy index test Φw [168] (with significance set at 0.05), as

implemented in Splitstree v 4.13.1 [169] on the whole genome alignment. We

used BRAT-NextGen [170] to detect recombination events within our isolate

panel using the whole genome alignment. BRAT-NextGen was specifically

developed to detect homologous recombinant segments among a group of

closely related bacteria over the process of their diversification and has been

shown to be highly accurate when applied to mycobacteria [171]. The

estimation of recombination was started with a partitioning of the whole

genome alignment into 5kb segments and running a clustering analysis

separately on each of these segments. The alpha hyper-parameter was

Page | 82

estimated with default settings. The proportion of shared ancestry (PSA) tree

was cut at 0.15 differentiating a total set of 4 clusters for all taxa.

Recombination profiles were calculated with 100 iterations, at which stage

parameter estimations had successfully converged. Significance (p<0.05) of

each putative recombinant segment was determined with 100 pseudo

replicate permutations.

Maximum-likelihood phylogenetic analysis

A maximum-likelihood (ML) phylogeny was estimated ten times from the SNP

alignment using RAxML v8.2.0 [172] under a plain generalized time reversible

(GTR) model (no rate heterogeneity) with likelihood calculation correction for

ascertainment bias using the Stamatakis method [173]. Identical sequences

were removed before the RAxML runs. For each run we performed 10,000

rapid bootstrap analyses to assess support for the ML phylogeny. The tree

with the highest likelihood across the ten runs was selected. We used

TreeCollapseCL v4 [174] to collapse nodes in the tree with bootstrap values

below a set threshold of 70% [175] to polytomies while preserving the length

of the tree. Phylogenetic relationships were inferred with the two PNG strains

as outgroups. Root-to-tip distances were extracted from the ML phylogeny

using TreeStat v1.2 [176]. The relationship between root-to-tip distances and

tip dates [177] was determined using linear regression analysis in R v3.2.0

[178].

Bayesian phylogenetic analysis

We used BEAST2 v2.2.1 [77] to date evolutionary events, determine the

substitution rate and produce a time-tree of African M. ulcerans, as this

approach allows for inference of phylogenies with a diverse set of molecular

clock and population parameters [72]. Path sampling (PS) [75] was used to

determine the best clock and population model priors by computing the

marginal likelihoods of competing models of evolution, as this method has

been shown to outperform other methods of model selection [75, 179]. We

compared three clock models (strict, uncorrelated exponential relaxed, and

uncorrelated log-normal relaxed) in combination with two demographic

coalescent models (constant and exponential). The required number of steps

Page | 83

in PS analysis was determined by running one of the more complex models

(uncorrelated log-normal relaxed clock / constant coalescent tree prior) with a

different amount of steps, starting from 100 steps until 400 using increments

of 50. As no difference in marginal likelihood estimates was observed after

100 steps, each model was run for 100 path steps, each with 200 million

generations, sampling every 20,000 MCMC generations and with a burn-in of

30%. Likelihood log files of all individual steps were inspected with Tracer v1.6

[180] to see whether the chain length produced an effective sample size (ESS)

larger than 400, indicating sufficient sampling. Marginal likelihoods of the

models were then used to calculate natural log Bayes factors (LBF=2*(ln

mL(model1)-mL(model2)), which evaluate the relative merits of competing

models (see also 4.SI1 text).

The best clock/demographic model (UCLD relaxed clock with a constant

coalescent tree prior - see also 4.SI1 text) was then used to infer a genome

scale African M. ulcerans time-tree under the GTR substitution model and

with tip-dates defined as the year of isolation (Table 4.S1). Analysis was

performed in BEAST2 using a total of 10 independent chains of 200 million

generations, with samples taken every 20,000 MCMC generations. Log files

were inspected for convergence and mixing with Tracer v1.6. LogCombiner

v2.2.1 [77] was then used to combine log and tree files of the independent

BEAST2 runs, after having removed a 30% burn-in from each run. Thus,

parameter medians and 95% highest posterior density (HPD) intervals were

estimated from over 1.6 billion visited MCMC generations. To ensure prior

parameters were not over-constraining the calculations, the entire analysis

was furthermore run while sampling only from the prior. Finally, we also

checked for the robustness of our findings under different priors, as states of

low mutation rate and large troot are hard to distinguish from otherwise

identical states of large mutation rate and smaller troot [181]. This additional

analysis was required as the tree prior and the clock prior interact when

adding sequence data, and the strength of this interaction is not visible when

sampling exclusively from the prior.

TreeAnnotator was used to summarize the posterior sample of time-trees so

as to produce a maximum clade credibility tree with the posterior estimates

of node heights visualized on it (posterior probability limit ≥ 0.8).

Page | 84

A permutation test was used to assess the validity of the temporal signal in

the data. This was undertaken by performing 20 additional BEAST2 runs (of

200 million MCMC generations each) with identical substitution (GTR), clock

(uncorrelated log-normal relaxed) and demographic models (constant

coalescent) but with tip dates randomly reassigned to sequences. This random

“null set” of tip-date and sequence correlations was then compared with the

substitution rate estimate of the genuine tip-date and sequence correlations

[82, 177].

Discrete phylogeographic analysis.

To assess the geospatial distribution of African M. ulcerans through time, an

additional BEAST2 analysis was performed. In this analysis the posterior

probability distribution of the location state (geographic region) of each node

in the tree was inferred, in addition to the parameters described above (tree

topology, evolutionary & demographic model). Sampled isolates were

associated with fixed discrete location states and a discrete BSSVS geospatial

model [182] was subsequently used to reconstruct the ancestral location

states of internal nodes in the tree from these isolate regions. To prevent loss

of the signal in the data by considering too many discrete regions compared

to the number of isolates, we limited the amount of discrete regions by

merging neighboring countries (similarly as carried out in [183, 184]. As such

we differentiated 5 regions: Ivory Coast (20 isolates), Ghana-Togo (25

isolates), Benin-Nigeria (65 isolates), Cameroon-Gabon (24 isolates), and

Angola-DRCongo-Congo (30 isolates). Ten independent chains were run for

200 million generations, with subsamples recorded from the posterior every

20,000 MCMC generations. LogCombiner was then used to combine tree files

of the independent BEAST2 runs, after having removed a 30% burn-in from

each run. TreeAnnotator was used as described above to summarize the

posterior sample of time-trees.

4.4 Results & Discussion

In order to understand how and when M. ulcerans has spread across Africa we

sequenced the genomes of 165 African isolates that were obtained between

1964 and 2012 and spanned most of the known endemic areas of BU in 11

Page | 85

different African countries (Figure 4.S1). This collection captured as much

diversity as possible within Africa while minimizing the phylogenetic discovery

bias implicit to SNP typing [112, 146]. Resulting sequence reads were mapped

to the Ghanaian M. ulcerans Agy99 reference genome and, after excluding

mobile repetitive IS elements and small insertion-deletions (indels), we

detected a total of 9,193 SNPs uniformly distributed across the M. ulcerans

chromosome with approximately 1 SNP per 613 bp (0.17% nucleotide

divergence) (Figure 4.S2). Similarly, a total of 81 SNPs were identified in the

non-repetitive regions of pMUM001, which resulted in a very comparable

nucleotide divergence of 0.14%. The maximum chromosomal genetic distance

between two African isolates was 5,157 SNPs.

The population structure of M. ulcerans in Africa

Large DNA deletions are excellent evolutionary markers since they are very

unlikely to occur independently in different lineages but rather are the result

of unique irreversible events in a common progenitor [185]. We explored the

distribution of large chromosomal deletions (relative to Agy99) and identified

differential genome reduction that supports the existence of two specific M.

ulcerans lineages within the African continent, hereafter referred to as

Lineage Africa I (Mu_A1) and Lineage Africa II (Mu_A2).

SNP-based exploration of the genetic population structure agreed with the

above deletion analysis and subdivided the African M. ulcerans population

into four major clusters. Clusters 1 to 3 constitute Mu_A1 while BAPS-cluster

4 corresponds to Mu_A2. The composition of these clusters is detailed in

Table 4.S1. Cluster 1 circulates throughout the African continent and

represents the majority of the isolates, n=136 (82.4%). This cluster is also the

most genetically diverse with an intra-cluster average pairwise SNP difference

(SNPΔ) of 171 SNPs (SD=73). Clusters 2, 3, and 4 were considerably smaller,

encompassing 20 (12.1%), 1 (0.6%) and 8 (4.8%) isolates respectively. Cluster

2 circulates in different regions of Cameroon and neighboring Gabon and

corresponds to an SNPΔ of 64 SNPs (SD=54). Cluster 3 (1 isolate) was only

found in Uganda. Finally, Cluster 4 (which encompasses Mu_A2 entirely) has a

SNPΔ of 81 SNPs (SD=41) and is common in Gabon (40%, n=5), but quite rare

in Cameroon (5%, n=19) and Benin (8%, n=59).

Page | 86

African M. ulcerans evolves through clonal expansion, not recombination

Ignoring recombination when analyzing evolving bacterial pathogens can be

misleading as the process has the potential to both distort phylogenetic

inference and create a false signal of apparent mutational evolution by

(horizontally) introducing additional divergence between heterochronously

sampled disease isolates [186]. The pairwise homoplasy index (Φw) did not

find statistically significant evidence for recombination (p= 0.1545) between

different BAPS-clusters. Correspondingly, BRAT-NextGen did not detect any

recombined segments in any isolate, supporting a strongly clonal population

structure for M. ulcerans that is evolving by vertically inherited mutations.

Phylogenetic analysis reveals strong geographical restrictions on M. ulcerans

dispersal

A phylogeny was reconstructed from the chromosomal SNP alignment using

both maximum-likelihood (RAxML) and Bayesian (BEAST2) approaches. Figure

4.S3 shows a well-supported ML-phylogeny that resolved the two major

African lineages Mu_A1 and Mu_A2 and distinguished between the 4 BAPS-

clusters within the African panel. Both the lineages and the BAPS-clusters had

100% bootstrap support and a Bayesian posterior probability of 1 (BEAST2

tree – Figure 4.1). The genome-based phylogeny was consistent with

previously constructed phylogenies based on discriminating ISE-SNP markers

even though these previous trees suffered from low branch support [135].

The tree also indicated that Mu_A1 is much more widely dispersed within the

African continent than Mu_A2.

We identified an unambiguous relationship between the genotype of an

isolate and its geographical origin. This is illustrated by the explicit regional

clustering of M. ulcerans within the phylogenetic tree, indicating significant

geographical structure in the African mycobacterial population. For instance,

all strains isolated from patients living in the hydrological basin of the Kouffo

River of southern Benin clustered together in a “basin-specific” clade in the

Bayesian phylogeny (Figure 4.1). Our observations confirm and extend

previous data showing geographical subdivisions [19, 112, 135, 154, 155], and

indicate that when M. ulcerans is introduced in a particular area, it remains

Page | 87

isolated and localized for a sufficient amount of time to allow mutations to

become fixed in that population. As such, a local genotype that is associated

with that area is allowed to evolve.

We also identified a strong association between the distribution of particular

genotypes and hydrological drainage areas. It appears that the borders of

hydrological basins (consisting of elevated regions, and salt water) also form a

barrier to bacterial spread. For instance, the isolates of the Kouffo Basin are

distinct from isolates originating from the neighboring Oueme Basin, and are

in fact more related to isolates from Ghana and Ivory Coast (Figure 4.1).

Page | 88

Page | 89

Figure 4.1: Bayesian maximum clade credibility phylogeny for African M. ulcerans. The tree

was visualized and colored in Figtree v1.4.2 [102]. Branches are color coded according to their

branch specific substitution rate (legend at top). Branches defining major lineages are

annotated on the tree. Tip labels are color coded according to their respective BAPS-clusters

(the best visited BAPS partitioning scheme of our sample yielded a natural log marginal

likelihood of -95857). Divergence dates (mean estimates and their respective 95% HDP) are

indicated in green for major nodes. Note 95% HDP intervals grow larger closer to the root of

the tree as increasingly less timing calibration information is available the further one goes

back in time. Geographically localized clonal expansions associated with four particular

hydrological basins (Congo, Kouffo, Oueme, and Nyong) are highlighted with boxes and their

corresponding t(MRCA) & 95% HDP are specified in green.

A central role for the mycolactone producing plasmid

All sequenced isolates carried the pMUM001 plasmid. By comparing read

depths of all plasmid and chromosome positions we roughly estimated an

average pMUM001 copy number of 1.3 copies per cell (range 0.4– 1.7). This

asserted the central role of the mycolactone producing plasmid in the

evolution of African M. ulcerans. The plasmid ML-tree (built with the

discovered 81 SNPs) closely matches the topology and relative branch lengths

of the chromosome ML-tree (Figure 4.S3A), which is consistent with co-

evolution of the plasmid with the host-chromosome, stable maintenance of

the plasmid, and absence of transfer of plasmid variants between host

bacteria.

The substitution rate of African M. ulcerans is remarkably low

The major objective of this study was to estimate the rate of evolution of M.

ulcerans in order to estimate the temporal dynamics of the spread of the

pathogen across Africa. Like Mycobacterium tuberculosis [184], M. ulcerans

does not exhibit a strict molecular clock with substitution mutations occurring

at a fixed regular rate, complicating temporal inferences. To overcome this

limitation, we used a Bayesian approach with a relaxed molecular clock model

to infer the evolutionary dynamics of the African mycobacterial population

(See also 4.SI1 text). As such, a genome scale African M. ulcerans time-tree

was inferred (Figure 4.1) while also providing estimates of nucleotide

substitution rates and divergence times for key M. ulcerans clades.

Page | 90

We estimated a mean genome wide substitution rate of 6.32E-8 per site per

year (95% HPD interval [3.90E-8 - 8.84E-8]), corresponding to the

accumulation of 0.33 SNPs per chromosome per year (95% HPD interval [0.20

- 0.46]) (excluding IS elements). To test the validity of the discovered temporal

signal in the data we performed 20 permutation tests. This produced a null set

of 20 “randomized” substitution rate distributions, which were significantly

different (Wilcoxon test, p<2.2E-16) to the substitution rate estimate of the

genuine tip-date and sequence correlation (Figure 4.S4). This clearly indicated

that the tip dates were informative and could provide sufficient calibrating

information for the analysis [82, 177]. The estimated genome wide

substitution rate is lower than the estimate for Clostridium difficile (1.88E-7)

[183] and Shigella sonnei (6.0E-7) [82] yet slightly higher than that of M.

tuberculosis (2.6E-9), the bacterium with the slowest rate currently described

[184]. The analysis also indicated that the genealogy has undergone very

moderate rate variation, with a 2.8-fold difference between the slowest

(3.18E-8) and the fastest (8.87E-8) evolving branches. Rate accelerations and

decelerations are found interspersed in the time-tree (Figure 4.1). The

observed slight rate variation is probably attributable to fluctuations in the

number of bacterial replication cycles per time unit, changes in selection

pressures through time, or combinations of these factors.

M. ulcerans has existed in Africa for centuries and was recently re-

introduced

African Mu_A2 strains were found to form a very strongly supported

(posterior probability = 1) monophyletic group with two PNG strains that were

included in the analysis as outgroup, indicating a closer relationship with

strains from PNG than African Mu_A1 stains. The Bayesian analysis (Figure

4.1) demonstrated furthermore that lineage Mu_A1 has been endemic in the

African continent for hundreds of years (tMCRA(Mu_A1) = 68 B.C. (95% HPD

1093 B.C. - 719 A.D.)). Conversely, Lineage Mu_A2 was introduced much more

recently in the African continent (tMCRA(Mu_A2) = 1800 A.D. (95% HPD 1689

A.D. - 1879 A.D.)), explaining why the lineage is less common and more

geographically restricted. The estimated time span of the Mu_A2 introduction

event coincides with a historical event of particular interest: the period of

Neo-imperialism (late 19th - early 20th century). During this period the

Page | 91

European powers divided the African continent among themselves through

the invasion, colonization and annexation of territory. In the absence of a

known alternative reservoir, nor vector, this specific temporal association

implies a human mediated Mu_A2 introduction event, whether through the

introduction of M. ulcerans bacteria within diseased humans, or an alternative

reservoir or vector.

Recent introduction of M. ulcerans in the Congo, Kouffo, Oueme and Nyong

basins

The time-tree of African M. ulcerans also reveals evidence of the likely role

that the so-called “Scramble for Africa” played in the spread of endemic

Mu_A1 clones in three hydrological basins (Congo, Oueme & Nyong) that are

particularly well covered by our isolate panel (Figure 4.1) [187]. Since, to our

knowledge, no epidemiological studies were conducted in these hydrological

basins until the late 1900s, whether BU was a newly introduced, versus an old

expanding illness in these regions [5] had remained unclear to date. Close

inspection of the time-tree revealed that, similar to the Mu_A2 introduction

event, the basin-specific introduction events coincide with the start of colonial

rule (Table 4.1).

To situate the historical model we are suggesting here, it is important to note

that inhabitants of the three regions of interest have long exploited the river

and forest ecologies prior to the arrival of the European colonial powers [188].

Many of the basin’s inhabitants relied on natural resources for their survival

and as such, were continuously exposed to the lentic environments. However,

it was only after the start of colonial rule that the basin associated epidemic

Mu_A1 clones were introduced, presumably through the introduction of a M.

ulcerans reservoir or vector. However, given the fact that no vector or

reservoir species is known in the Afrotropic ecozone other than Homo sapiens

[36, 65, 66], we postulate that it was the arrival of displaced BU-infected

humans that played a pivotal role in the observed spread of M. ulcerans.

Colonialism was commonly violent and introduced significant socio-economic

changes in the three basins that often involved population displacement. In all

likelihood, displaced BU-infected humans were not directly infecting other

humans as human-to-human transmission of M. ulcerans is extremely rare

Page | 92

[189]. Humans were nevertheless in all probability an important reservoir as

displaced BU-infected patients with active, openly discharging lesions could

contaminate the environment during water contact activities by shedding

concentrated clumps of mycobacteria. Transmission could occur indirectly in

the same community water source, when the superficial skin surface of a

naïve individual was contaminated, and the bacilli present on the

contaminated skin were subsequently inoculated subcutaneously through

some form of penetrating (micro)trauma [190]. Alternatively, we cannot rule

out the possibility that the displaced BU-infected patients were instead source

reservoirs for mechanical transmission via an (unknown) kind of biting aquatic

insect vector. In spite of this uncertainty, since tMCRA of the three hydrological

basins did not predate colonization, it seems likely that M. ulcerans was

introduced after the instigation of colonial rule through an influx of BU

infected humans.

A fourth noteworthy hydrological basin is that of the Beninese Kouffo River.

The timing of its basin specific introduction event (1977 A.D. (95% HPD 1959

A.D. - 1988 A.D.)) is much more recent than the three previously discussed

basins. Notably, the first BU cases from this region were identified and treated

in 1977 [191], concurrent with the estimated introduction event.

Page | 93

Table 4.1: Timing of introduction events of four selected epidemic lineage Mu_A1 clones in

their respective hydrological basin. a

Founding of the Belgian Congo Free State; b

Kingdom of

Dahomey annexed into the French colonial empire; c German Empire claimed the colony of

Kamerun and began a steady push inland; d Based on interviews and observations of healed

lesions in the villages of the Songololo territory it was believed that M. ulcerans infections

already existed in the area in 1935 [120]. Abbreviations: HDP, highest probability density

interval; t(MRCA), time to most recent common ancestor.

Hydrological

Basin Endemic hotspot

Approx. start of

colonial rule Mean t(MRCA)

95% HPD

t(MRCA)

First reported

cases

Congo Songololo Territory 1885a 1905 1855-1941 1961

d [192]

Nyong Between Ayos and

Akonolinga 1884

c 1901 1848-1937 1969 [193, 194]

Oueme Southeastern Benin 1892b 1890 1835-1932 1988 [195]

Kouffo Southeastern Benin 1892b 1977 1959-1988 1977 [191]

The historic spread of M. ulcerans lineage Mu_A1

The phylogeographic analysis also offered new insights on the geospatial

spread of M. ulcerans lineage Mu_A1 through time (Figure 4.2). Uganda,

Cameroon, and Gabon occupy basal branches in the Mu_A1 time-tree

indicating that the bacterium has been extant in these regions for the longest

time. The lineage subsequently expanded from these regions into West and

Central Africa. Ancestral state reconstruction analysis indicated that this

expansion most likely occurred from the region that encompasses Benin and

Nigeria (posterior probability = 0.91). From there Mu_A1 spread further west

(into Togo, Ghana and Ivory Coast) and back east (into Congo, DRC and

Angola).

In other bacterial pathogens, the discipline of microbial phylogeography has

also proved to be a powerful means of investigating not only the spread of

microbes but also the movement of their hosts. For example, interesting

associations were found between the genotypes of Helicobacter pylori strains,

their places of origin, and the migration and ethnicity of their human hosts

[196]. Additionally, the spread of M. tuberculosis and M. leprae also reflects

the migrations of early humans [56, 184].

Page | 94

Comparable as in these landmark studies, the phylogeographic approach

applied here is limited in the respect that it studies a sample of the

mycobacterial population from which it then infers information about the

entire population through various statistical methods. Even though our isolate

panel originates from all African disease foci that have ever yielded positive

M. ulcerans cultures, the spatial coverage of disease isolates is moderately

restricted to specific geographical areas, which might have confounded our

interpretation of the historic spread of African M. ulcerans.

Page | 95

Figure 4.2: Geospatial distribution of African M. ulcerans through time. A Bayesian maximum

clade credibility phylogeny is drawn for lineage Mu_A1 with branches colour coded according

to their most likely location state (legend at top). Pie charts indicate location state posterior

probability distributions of major nodes. The amount of location states was limited to five by

merging the disease isolates of certain neighboring countries. The genetically distinct

Ugandan singleton node (which represents its own BAPS-cluster) was omitted from the

analysis as multiple isolates are required per cluster. Divergence dates (mean estimates and

their respective 95% HDP) are indicated in green for nodes that fall outside of the time scale.

A number of oversampled localized clonal expansions are collapsed in the tree with the size of

their representing cartoon proportional to the amount of collapsed taxa. The tips of the tree

are connected to the location of residence of patients from whom the isolate was grown.

4.5 Conclusion

Here we reconstructed the population structure and evolutionary history of

African M. ulcerans using the molecular and bioinformatics tools of modern

population genomics. The genetic diversity of M. ulcerans proved restricted

because of its slow substitution rate coupled with its recent origin. Sequence

types appear to be maintained in geographically separated subpopulations

that are associated with hydrological drainage areas. Our data show for the

first time that the spread of M. ulcerans across Africa is a relatively modern

phenomenon and one that has escalated since the late 19th and early 20th

centuries. Using temporal associations, this work implicates human-induced

changes and activities behind the expansion of BU in Africa. We propose that

humans with actively infected, openly discharging BU lesions inadvertently

contaminated aquatic environments during water contact activities and thus

played the pivotal role in the spread of the mycobacterium. Our observations

on the role of humans as potential maintenance reservoir to sustain new BU

infections suggests that interventions in a region aimed at reducing the

human BU burden will at the same time break the transmission chains within

that region. Active case-finding programs, improved disease surveillance, and

the early treatment of pre-ulcerative infections with specific antibiotics will

decrease the amounts of mycobacteria shed into the environment and may as

a result reduce disease transmission. Our findings are supported by the

observed decline of BU incidence recorded in some areas which profited from

both improved BU surveillance and early treatment [197].

Page | 96


Figure 4.S1: Distribution of BU in Africa by country, as of 2016. Relative endemicity is denoted

as high (red), moderate (yellow), and low (green); dots denote countries with suspected BU

cases. Stars represent the residence of BU patients from whom M. ulcerans disease isolates

were grown at the time of clinical visit. In the case where residence information was missing

we used the location of the hospital supplying the sample. Five countries (Equatorial Guinea,

Kenya, Liberia, Sierra Leone, and South Sudan) that have reported a limited number of BU

cases in the past [5] were not included in the study due to a lack of isolates.

Page | 97

Figure 4.S2: Distribution of SNPs identified compared to the Ghanaian M. ulcerans Agy99

reference genome. The Y-axis corresponds to SNP counts per 10,000bp window, the dashed

line indicates the average rate of 16 SNPs per 10,000bp (or 1 SNP per 613 bp window).

Page | 98

Page | 99

Figure 4.S3: Maximum-likelihood phylogeny of African M. ulcerans based on 9,193 SNP

differences detected across the whole core genome of 167 sequenced isolates. Branches

defining major lineages are annotated on the tree. Nodes in the tree with bootstrap support

below a set threshold of 70% were collapsed to polytomies, while preserving the length of the

tree. The tree was rooted with the two PNG strains as outgroup. Branches are color coded

according to their respective BAPS-clusters as indicated in the legend. Branches defining

major lineages are further annotated on the tree. Tip labels are color coded according to

country of isolation. Inset A: Maximum-likelihood phylogeny of the plasmid which was

constructed from 81 SNPs identified in the non-repetitive regions of pMUM001. The topology

and relative branch lengths of the plasmid tree match those of the chromosomal tree. Inset B:

Linear regression analysis of year of isolation vs root-to-tip distances extracted from

chromosomal ML-phylogeny. Linear regression lines (with 95% CI) and Spearman's rank

correlation coefficients are indicated separately for each lineage.

Page | 100

Figure 4.S4: Comparison of Bayesian estimates of nucleotide substitution rates for real and

randomized tip dates. Filled squares & circles represent mean estimates, while bars indicate

values of the 95% highest probability density (HDP) interval. The estimate obtained using the

real tip date associations (circle) is shown to the far right while estimates from random

associations (squares) are shown to the left. All randomized data sets were analyzed in

BEAST2 using identical model settings as used in the analysis of the real tip date data. Note

the y-axis is on the log scale.

Page | 101

Table 4.S1: M. ulcerans isolate and DNA sequencing information. 1: Democratic Republic of

the Congo

Isolate n° Lineage BAPS

cluster YOI Supplier Country of Origin

Administrative

Division First-

level

Administrative

Division Second-

level

Administrative Division

Third-level Latitude Longitude

Coverage

(x)

Average

read

length

(bp)

Sequencing Platform

ITM_072814 Mu_A1 BAPS-1 2007 CDTUB

Zagnanado Benin Ouémé Dangbo Gbéko 6.606738 2.454321 44 136

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Zou Ouinhi Ouokon 7.099851 2.464233 63 136

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Zou Zogbodomey Domè-Houandougon 7.100052 2.300224 39 135

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Zou Zogbodomey Domè-Houandougon 6.906959 2.453384 73 137

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Ouémé Bonou Bonou 6.906959 2.453384 83 138

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Ouémé Bonou Bonou 6.906959 2.453384 58 138

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Zou Zagnanado Doga-Domè 7.202393 2.336118 50 131

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Zou Ouinhi Tohoue / Hounnoumè 6.974698 2.419704 92 115

Illumina MiSeq (PE 2 x

250 bp)


Zagnanado Benin Zou Ouinhi Ouinhi / Monzoungoudo 7.073294 2.527386 83 243


250 bp)




250 bp)


Zagnanado Benin Zou Ouinhi Sagon / Adamè 7.157755 2.425868 111 241


250 bp)


Zagnanado Benin Zou Ouinhi Dasso / Bossa 6.997382 2.454390 116 243


250 bp)


Zagnanado Benin Zou Ouinhi Dasso / Agonkon 6.970885 2.438508 97 240


250 bp)


Zagnanado Benin Zou Ouinhi

Dasso / Yaago &

Akantomè 7.051987 2.415819 58 103

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Zou Djidja Oungbègame 7.278137 2.023043 55 245


250 bp)


Zagnanado Benin Zou Ouinhi Sagon / Ayizè 7.166206 2.500697 94 129


250 bp)


Zagnanado Benin Zou Ouinhi Dasso / Yaago 6.989340 2.461358 77 127


250 bp)




250 bp)


Zagnanado Benin Zou Zagnanado Dovi-Dove / Tévedji 7.097947 2.409703 98 111


250 bp)


Zagnanado Benin Zou Ouinhi Tohoue / Midjannangon 6.967335 2.419675 76 244


250 bp)


Zagnanado Benin Zou Ouinhi Tohoue / Akassa 7.051987 2.415819 51 118


250 bp)


Zagnanado Benin Zou Zagnanado Dovi-Dove / Tévedji 7.097947 2.409703 86 231


250 bp)


Zagnanado Benin Zou Ouinhi Ouinhi / Ahicon 7.074330 2.505618 317 184


250 bp)


Zagnanado Benin Atlantique Toffo Séhoué 6.902940 2.254384 71 133

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Atlantique Toffo Séhoué / Agaga 6.925428 2.270933 39 130

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Ouémé Dangbo Dékin 6.557344 2.451185 44 135

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Nigeria Ogun State Yewa North Odja Odan 6.896438 2.845491 62 137

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Plateau Adja-Ouere Tatonnoukon 6.910516 2.538600 72 137

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Plateau Issaba Onigbolo 7.170837 2.650620 66 139

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Ouémé Adjohoun Abato 6.738905 2.484311 62 137

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Zou Zogbodomey Kpokissa / Hinzounmè 7.000827 2.399765 41 130

Illumina HiSeq 2000

(PE 2 x 150 bp)



Illumina HiSeq 2000

(PE 2 x 150 bp)

Page | 102



Administrative

Division First-

level

Administrative

Division Second-

level



Coverage

(x)

Average

read

length

(bp)

Sequencing Platform



Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Zou Zogbodomey Kpokissa 7.000827 2.399765 76 138

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Zou Zagnanado NA 7.209312 2.341557 60 139

Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_021433 Mu_A1 BAPS-1 2002 CDTUB Lalo Benin Kouffo Lalo Gnizoumè / Hangbanou 6.941835 1.957969 60 136 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_022045 Mu_A1 BAPS-1 2002 CDTUB Lalo Benin Kouffo Lalo Adoukandji /

Yamontouhoué 6.917139 1.959558 75 138

Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_022287 Mu_A1 BAPS-1 2002 CDTUB Lalo Benin Zou Agbangnizoun Kpota 7.023358 1.973188 59 136 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_022875 Mu_A1 BAPS-1 2002 CDTUB Lalo Benin Kouffo Lalo Gnizoumè 6.943016 1.942924 64 135 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_030717 Mu_A1 BAPS-1 2003 CDTUB Lalo Benin Kouffo Lalo Ahomadégbé 6.834510 2.001984 54 137 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_030718 Mu_A1 BAPS-1 2003 CDTUB Lalo Benin Kouffo Lalo Lalo 6.929381 1.888870 57 136 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_031892 Mu_A1 BAPS-1 2003 CDTUB Lalo Benin Kouffo Lalo Hlassamè 6.902215 1.941018 40 128 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_071804 Mu_A1 BAPS-1 2007 CDTUB Lalo Benin Kouffo Lalo Zalli 6.980859 1.926673 61 136 Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Kouffo Lalo Tchito / Gare 6.919671 2.046356 56 130

Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_970680 Mu_A1 BAPS-1 1997 CDTUB Lalo Benin Mono Houéyogbé Sahoué 6.538114 1.807039 68 137 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_021434 Mu_A1 BAPS-1 2002 CDTUB Lalo Benin Kouffo Klouékanmè Adjassagon 7.001751 1.951723 46 135 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_012596 Mu_A1 BAPS-1 2001 CDTUB Lalo Benin Mono Bopa Lobogo 6.624900 1.907600 81 137 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_071938 Mu_A1 BAPS-1 2007 CDTUB Lalo Benin Kouffo Lalo Tandji 6.943208 1.970882 59 136 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_000909 Mu_A1 BAPS-1 1999 ITM Togo Maritime Vo Tchekpo Deve 6.484348 1.369718 49 142 Illumina MiSeq (PE 2 x

300 bp)

ITM_070123 Mu_A1 BAPS-1 2006 IME DRC1 Bas-Congo

Cataractes /

Songololo Kimpese / Cité-Kimpese -5.56652 14.43447 72 136

Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_092479 Mu_A1 BAPS-1 2009 IME DRC Bas-Congo Cataractes /

Songololo Kimpese / Cité-Kimpese -5.563273 14.445951 67 132

Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_000483 Mu_A1 BAPS-1 2000 ITM Ivory_Coast Moyen-Cavally Duékoué Niambli 6.746215 -7.278825 150 136 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_000870 Mu_A1 BAPS-1 2000 ITM Ivory_Coast Dix-Huit

Montagnes Zouan-Hounien Ouyatouo 6.872745 -8.109516 45 133

Illumina HiSeq 2000

(PE 2 x 150 bp)


Songololo Luima / Cité Songololo -5.6969 14.05914 56 135

Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_071924 Mu_A1 BAPS-1 2007 ITM Congo Kouilou Madingo-Kayes Loukouala -4.433392 11.700150 53 136 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_071925 Mu_A1 BAPS-1 2007 ITM Congo Kouilou Madingo-Kayes Loukouala -4.433392 11.700150 54 136 Illumina HiSeq 2000

(PE 2 x 150 bp)


Songololo

Bamboma / Mbanza-

Manteke -5.46421 13.78163 65 137

Illumina HiSeq 2000

(PE 2 x 150 bp)


Songololo Palabala / Nkamuna -5.76785 13.92162 74 138

Illumina HiSeq 2000

(PE 2 x 150 bp)



Illumina HiSeq 2000

(PE 2 x 150 bp)


Songololo Luima / Ngombe -5.39494 14.12631 57 137

Illumina HiSeq 2000

(PE 2 x 150 bp)


Songololo

Palabala / Km 70

Vemadiyo -5.68719 13.89086 40 136

Illumina HiSeq 2000

(PE 2 x 150 bp)


Songololo Luima / Luvuvamu -5.70311 14.1483 40 135

Illumina HiSeq 2000

(PE 2 x 150 bp)



Illumina HiSeq 2000

(PE 2 x 150 bp)



Illumina HiSeq 2000

(PE 2 x 150 bp)


Songololo Luima / Kisonga -5.75556 13.98184 67 137

Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_073459 Mu_A1 BAPS-1 2007 CDTUB Lalo Benin Kouffo Lalo Ahojinako 6.737059 1.978464 84 139 Illumina HiSeq 2000

(PE 2 x 150 bp)


Songololo Songololo / Luvituku -5.67491 14.01439 74 137

Illumina HiSeq 2000

(PE 2 x 150 bp)

Page | 103



Administrative

Division First-

level

Administrative

Division Second-

level



Coverage

(x)

Average

read

length

(bp)

Sequencing Platform


Songololo

Bamboma / Mbanza-

Manteke -5.46071 13.78244 48 136

Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_073478 Mu_A1 BAPS-1 2007 IME Angola Malanje Marimba Kafufu / Luremo (Kwango

River) -8.096696 17.503546 47 135

Illumina HiSeq 2000

(PE 2 x 150 bp)


Songololo Luima / Kisonga -5.75556 13.98184 50 138

Illumina HiSeq 2000

(PE 2 x 150 bp)


Songololo Kimpese / Nzundu -5.41166 14.49669 41 136

Illumina HiSeq 2000

(PE 2 x 150 bp)


Songololo Kimpese / Tole -5.70385 14.39294 51 137

Illumina HiSeq 2000

(PE 2 x 150 bp)


Songololo Luima / 5km -5.73816 14.04989 69 137

Illumina HiSeq 2000

(PE 2 x 150 bp)


Songololo Luima / 5km -5.73816 14.04989 48 137

Illumina HiSeq 2000

(PE 2 x 150 bp)



Illumina HiSeq 2000

(PE 2 x 150 bp)


Songololo Mayanga / Mpelo -5.29764 14.03725 54 137

Illumina HiSeq 2000

(PE 2 x 150 bp)


Songololo Luima / Nkondo -5.60384 14.13927 89 133

Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_040149 Mu_A1 BAPS-1 2003 ITM Ghana Ashanti Asante Akim North Agogo Presbyterian

Hospital 6.799841 -1.083548 60 133

Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_991591 Mu_A1 BAPS-1 1999 ITM Togo Maritime Vo Anagali 6.500000 1.366667 79 144 Illumina MiSeq (PE 2 x

300 bp)

ITM_050303 Mu_A1 BAPS-1 1979 ITM Congo Kouilou NA NA -4.432895 12.299497 59 131 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_960658 Mu_A1 BAPS-1 1996 ITM Angola Bengo Dande Caxito -8.565601 13.669590 55 137 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_960657 Mu_A1 BAPS-1 1996 ITM Angola Bengo Dande Caxito -8.565601 13.669590 79 138 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_072646 Mu_A1 BAPS-1 2007 KCCR Ghana Ashanti Atwima Mponua Abofrom 6.616723 -1.986334 74 137 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_072651 Mu_A1 BAPS-1 2007 KCCR Ghana Ashanti KMA Kaase 6.644303 -1.632192 55 137 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_120140 Mu_A1 BAPS-1 2011 CPC Cameroon Adamawa

Region Maya-Banyo Bankim / Mbondji II 6.084052 11.486537 78 139

Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_030950 Mu_A1 BAPS-1 2003 CDTUB Lalo Benin Kouffo Lalo Adoukandji 6.818579 1.986384 70 132 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_030716 Mu_A1 BAPS-1 2003 CDTUB Lalo Benin Kouffo Lalo Tchito / Village Aboeti 6.917797 2.023378 67 138 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_102686 Mu_A1 BAPS-1 2010 ITM Nigeria Oyo State Ibadan Ibadan 7.387891 3.920866 56 137 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_083232 Mu_A1 BAPS-1 2008 ITM Angola Lunda Norte Xa-Muteba (Kwango River) -8.522231 17.746601 56 137 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_000869 Mu_A1 BAPS-1 2000 ITM Ivory_Coast Moyen-Cavally Duékoué Guezon 6.733249 -7.107404 48 136 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_990007 Mu_A1 BAPS-1 1998 ITM Ivory_Coast Haut-Sassandra Issia Guetuzon II 6.746001 -6.929517 59 138 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_991633 Mu_A1 BAPS-1 1999 ITM Ivory_Coast Moyen-Cavally Duékoué Guezon 6.733249 -7.107404 94 139 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_990008 Mu_A1 BAPS-1 1998 ITM Ivory_Coast Haut-Sassandra Issia Zakogbeu 6.782380 -6.814253 81 139 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_070386 Mu_A1 BAPS-1 2007 ITM Nigeria Anambra State Ayamelum Ifite Ogwari 6.609054 6.951123 68 137 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_5151 Mu_A1 BAPS-1 1972 ITM DRC Maniema Kasongo NA -4.186555 26.439372 59 145 Illumina MiSeq (PE 2 x

250 bp)

ITM_970359 Mu_A1 BAPS-1 1997 ITM Ghana Ashanti Amansie West Manso-Afraso 6.343930 -1.984469 46 137 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_970606 Mu_A1 BAPS-1 1997 ITM Ghana Ashanti Amansie West Yaw Kasakrom 6.450216 -1.791701 64 138 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_970677 Mu_A1 BAPS-1 1997 ITM Ghana Ashanti Amansie West Manso Dominase 6.464442 -1.893037 78 138 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_970678 Mu_A1 BAPS-1 1997 ITM Ghana Ashanti Asante Akim North Afrisre 7.050181 -0.962048 69 138 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_970959 Mu_A1 BAPS-1 1997 ITM Ghana Ashanti Amansie West Manso-Afraso 6.343930 -1.984469 81 139 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_970964 Mu_A1 BAPS-1 1997 ITM Ghana Ashanti Amansie West Offinho Asaman 6.373639 -1.999730 65 137 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_971351 Mu_A1 BAPS-1 1997 ITM Ghana Ashanti Atwima Mponua Achiase 6.599467 -2.029728 76 138 Illumina HiSeq 2000

(PE 2 x 150 bp)

Page | 104



Administrative

Division First-

level

Administrative

Division Second-

level



Coverage

(x)

Average

read

length

(bp)

Sequencing Platform

ITM_980063 Mu_A1 BAPS-1 1998 ITM Ghana Ashanti Atwima Mponua Achiase 6.599467 -2.029728 66 137 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_940662 Mu_A1 BAPS-1 1994 ITM Ivory_Coast Moyen-Cavally Duékoué Nanandi 6.706187 -7.106573 43 136 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_990006 Mu_A1 BAPS-1 1998 ITM Ivory_Coast Haut-Sassandra Issia Guetuzon I 6.741000 -6.940000 73 139 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_990734 Mu_A1 BAPS-1 1999 ITM Ivory_Coast Moyen-Cavally Duékoué Duékoué 6.683885 -7.313537 44 136 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_991632 Mu_A1 BAPS-1 1999 ITM Ivory_Coast Haut-Sassandra Issia Bediegbeu 6.67837 -6.82040 112 139 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_072634 Mu_A1 BAPS-1 2007 KCCR Ghana Ashanti Asante Akim North Adoniem 6.884259 -0.974580 86 138 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_072652 Mu_A1 BAPS-1 2007 KCCR Ghana Ashanti Atwima Mponua Achiase 6.599467 -2.029728 73 137 Illumina HiSeq 2000

(PE 2 x 150 bp)


(PE 2 x 150 bp)


(PE 2 x 150 bp)

ITM_072658 Mu_A1 BAPS-1 2007 KCCR Ghana Western Region Wassa West Owusukrom 5.295855 -1.995595 51 135 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_072650 Mu_A1 BAPS-1 2007 KCCR Ghana Ashanti Atwima Nwabiagya Kyereyase 6.703054 -1.837654 78 136 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_072630 Mu_A1 BAPS-1 2007 KCCR Ghana Central Upper Denkyira Nkotumso 6.001297 -1.918974 45 135 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_072656 Mu_A1 BAPS-1 2007 KCCR Ghana Ashanti Atwima Mponua Abompe 6.532913 -2.045321 47 137 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_072655 Mu_A1 BAPS-1 2007 KCCR Ghana Ashanti Atwima Mponua Sireso 6.619944 -2.302405 65 137 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_072653 Mu_A1 BAPS-1 2007 KCCR Ghana Ashanti Atwima Mponua Amadaa 6.645268 -1.991520 53 136 Illumina HiSeq 2000

(PE 2 x 150 bp)


(PE 2 x 150 bp)

ITM_001211 Mu_A1 BAPS-1 2000 ITM Ivory_Coast Dix-Huit

Montagnes Zouan-Hounien Zouan-Hounien 6.919888 -8.204394 57 136

Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_020279 Mu_A2 BAPS-4 2002 CPC Cameroon Centre Region Nyong-et-

Mfoumou Ayos 3.903651 12.536624 51 136

Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_091067 Mu_A2 BAPS-4 2009 ITM Gabon Moyen-Ogooué Ogooue et des Lacs Junkville -0.777504 10.192464 48 131 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_110450 Mu_A2 BAPS-4 2011 ITM Gabon Moyen-Ogooué Ogooue et des Lacs Gravier -0.629682 10.446592 88 137 Illumina HiSeq 2000

(PE 2 x 150 bp)


Mfoumou Akolo 3.833876 12.182044 39 133

Illumina HiSeq 2000

(PE 2 x 150 bp)


Mfoumou Obis 3.983916 12.467150 52 136

Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_9103 Mu_A1 BAPS-2 1970 ITM Cameroon Centre Region NA NA 3.782691 12.249357 59 131 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_101500 Mu_A1 BAPS-2 2010 ITM Gabon Moyen-Ogooué Ogooue et des Lacs Lambaréné / Adaghe -0.703731 10.240318 70 138 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_110893 Mu_A1 BAPS-2 2011 ITM Gabon Moyen-Ogooué Ogooue et des Lacs Issac -0.487290 9.925157 73 136 Illumina HiSeq 2000

(PE 2 x 150 bp)


Mfoumou Medjap 3.600827 11.983612 90 139

Illumina HiSeq 2000

(PE 2 x 150 bp)


Mfoumou Akonolinga / Ekolman 3.773791 12.243754 51 137

Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_051459 Mu_A1 BAPS-3 1964 NCTC Uganda Northern

Region Adjumani Adjumani 3.381441 31.783100 119 136

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Nigeria Ondo State Ifedore Ilara-Mokin 7.348108 5.110065 187 127

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Nigeria Ogun State Ewekoro NA 6.934564 3.218125 39 120

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Mono Bopa Lobogo / Lobogo 6.624900 1.907600 48 118

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Zou Agbangnizoun Sahe / Gbozoun 7.052104 1.956116 98 121

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Zou Agbangnizoun Sahe / Gbozoun 7.052104 1.956116 87 122

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Kouffo Klouékanmey Ahogbeya / Ahogbeya 7.026145 1.910315 52 120

Illumina HiSeq 2000

(PE 2 x 150 bp)


Zagnanado Benin Ouémé Porto-Novo

Porto-Novo / Kpota

Sandodo 6.472724 2.620146 57 118

Illumina HiSeq 2000

(PE 2 x 150 bp)

Page | 105



Administrative

Division First-

level

Administrative

Division Second-

level



Coverage

(x)

Average

read

length

(bp)

Sequencing Platform


Zagnanado Benin Zou Zogbodomey Kpokissa / Kpokissa 7.015763 2.381552 102 181


300 bp)


Zagnanado Benin Zou Zogbodomey Kpokissa / Dehounta 6.999813 2.400830 53 165


300 bp)

Mu_12-29 Mu_A1 BAPS-2 2012 CPC Cameroon NA NA NA NA NA 117 231 Illumina MiSeq (PE 2 x

250 bp)

Mu_08-51 Mu_A1 BAPS-2 2008 CPC Cameroon Centre Region Nyong-et-

Mfoumou Eboa 3.756858 12.250822 115 232


250 bp)


Mfoumou Ndjong Medjap 3.655066 12.173620 114 227


250 bp)


Mfoumou Djoudjoua 3.677862 12.018625 112 228


250 bp)


Mfoumou Mengou 3.895708 12.149926 99 222


250 bp)


Mfoumou Ngonanga 3.705152 12.233626 99 244


250 bp)



Illumina HiSeq 2000

(PE 2 x 150 bp)




250 bp)


Mfoumou Medoumou 3.624172 12.052166 91 244


250 bp)


Mfoumou Nkoldja'a 3.697345 12.115912 88 242


250 bp)



Illumina HiSeq 2000

(PE 2 x 150 bp)

Mu_11-20 Mu_A1 BAPS-2 2011 CPC Cameroon NA NA NA NA NA 77 137 Illumina HiSeq 2000

(PE 2 x 150 bp)

ITM_121474 Mu_A1 BAPS-1 2012 ITM Nigeria Cross river state Ogoja TBL hospital monaiya 6.661395 8.797635 48 139 Illumina MiSeq (PE 2 x

300 bp)

ITM_111662 Mu_A1 BAPS-2 2011 ITM Gabon Moyen-Ogooué Ogooue et des Lacs Lambaréné / Isaac -0.703731 10.240318 47 142 Illumina MiSeq (PE 2 x

300 bp)

ITM_940659 Mu_A1 BAPS-1 1994 ITM Ivory_Coast NA NA NA NA NA 39 141 Illumina MiSeq (PE 2 x

300 bp)


300 bp)


300 bp)


300 bp)


300 bp)


300 bp)


300 bp)


300 bp)


300 bp)

ITM_993355 Mu_A1 BAPS-1 1999 ITM Togo Maritime NA NA 6.348805 1.405662 49 141 Illumina MiSeq (PE 2 x

300 bp)

Mu_282-5 Mu_A2 BAPS-4 2011 CDTUB Pobé Benin Ouémé Bonou Dame Wogon 6.946733 2.418495 84 135 Illumina HiSeq 2000

(PE 2 x 150 bp)

Mu_283-8 Mu_A2 BAPS-4 2011 CDTUB Pobé Benin Ouémé Bonou Dame Wogon 6.946733 2.418495 75 135 Illumina HiSeq 2000

(PE 2 x 150 bp)

Mu_V71_200

5 Mu_A2 BAPS-4 2005 VIDRL Papua New Guinea NA NA NA NA NA 44 182


300 bp)

ITM_9537 Mu_A2 BAPS-4 1971 ITM Papua New Guinea NA NA NA NA NA 39 156 Illumina MiSeq (PE 2 x

300 bp)

Page | 106

Supplemental Text 4.SI1: The search for a molecular clock to date

phylogenetic events:

Several robust approaches exist to date bacterial phylogenies. Ideally,

molecular clocks are calibrated using temporal information from ancient DNA

(aDNA) sequences [198]. When aDNA is not available, “short-term” mutation

rates estimated from molecular epidemiological data [199] or animal infection

models [200] can be used. Since none of the above approaches are currently

possible for M. ulcerans a short-term molecular clock was estimated here

using correlations between phylogenetic divergence and isolation times of

heterochronous disease isolates.

Maximum likelihood phylogenetic analysis (Figure 4.S3B) revealed a very weak

(Mu_A1 ρ=0.013 and Mu_A2 ρ=0.34) correlation between root-to-tip branch

lengths and dates of isolation of the sequenced isolates. This observation is

indicative of the lack of rapid strict-clock-like evolution in which substitution

mutations occur at a fixed regular rate. The finding is in line with studies of

other bacterial species with notably slower clock rates that also failed to find

significant correlations between isolation time and phylogenetic divergence

[183, 184, 201].

Since M. ulcerans conflicted with strict clock-like molecular evolution we

subsequently used a Bayesian approach that allowed for the use of relaxed

molecular clock models (BEAST2) to infer the evolutionary dynamics of the

African mycobacterial population. PS analysis was used to select the best of

three clock models and the better of two demographic coalescent tree priors.

The model with the highest marginal likelihood was the UCLD relaxed clock

with a constant coalescent tree prior. Since this model was also one of the

more parameter-rich models we calculated LBF’s, which evaluate the relative

merits of competing models. A difference of more than 10 LBF units (which is

considered as very strong evidence against a competing model [202]) was

used as the threshold for accepting a more parameter-rich model. This

confirmed the UCLD relaxed clock model with a constant coalescent tree prior

to be the best model and thus was used for all subsequent BEAST2 runs.

Page | 107

Table: Natural log marginal likelihoods, Natural log Bayes factors and model probabilities of 6

competing models of molecular evolution of African M. ulcerans. Abbreviations: LBF, Natural

log Bayes factor; LmL, Natural log marginal likelihood; UCED, Uncorrelated exponential

distribution; UCLD; Uncorrelated log-normal distribution.

Clock model Coalescent tree prior LmL LBF Model probability

UCLD relaxed Constant -52551,97 0 0,99997

UCLD relaxed Exponential -52557,15 -10,36 3,17E-05

Strict Constant -52594,84 -85,74 5,80E-38

UCED relaxed Constant -52594,91 -85,88 5,04E-38

Strict Exponential -52601,46 -98,98 1,03E-43

UCED relaxed Exponential -52602,69 -101,44 8,81E-45

4.7 Data access

Read data for the study isolates have been deposited in the NCBI Sequence

Read Archive (SRA) under BioProject accession PRJNA313185.


KV was supported by a PhD-grant of the Flemish Interuniversity Council -

University Development Cooperation (Belgium). BdJ & CM were supported by

the European Research Council-INTERRUPTB starting grant (nr.311725). TPS

was supported by a fellowship from the National Health and Medical Research

Council of Australia (1105525).

Funding for this work was provided by the Department of Economy, Science

and Innovation of the Flemish Government, the Stop Buruli Consortium

supported by the UBS Optimus Foundation, and the Fund for Scientific

Research Flanders (Belgium) (FWO grant n° G.0321.07N). The computational

resources used in this work were provided by the HPC core facility CalcUA and

VSC (Flemish Supercomputer Center), funded by the University of Antwerp,

the Hercules Foundation and the Flemish Government - department EWI.

Aspects of the research in Cameroon and Benin were funded by the Raoul

Follereau Fondation France. The funders had no role in study design, data

collection and analysis, decision to publish, or preparation of the manuscript.

Page | 108

We thank Rodenbrandt Rex for helpful discussions and critical comments to

the manuscript. We thank Pim de Rijk, Wim Mulders, Krista Fissette, Elie

Nduwamahoro, and Cécile Uwizeye for their excellent technical assistance.

Page | 109

Chapter 5

The population size of Mycobacterium

ulcerans in the Congo river basin reflects the

intensity of control efforts: is the human

reservoir important?

Koen Vandelannoote, Delphin Mavinga Phanzu, Kapay Kibadi, Miriam Eddyani, Conor

J. Meehan, Kurt Jordaens, Herwig Leirs, Françoise Portaels, Timothy P. Stinear, Simon

R. Harris, Bouke C. de Jong

The population size of Mycobacterium ulcerans in the Congo river basin reflects the

intensity of control efforts: is the human reservoir important?

Designed research: KV, DMP, ME, CM, FP, TPS, SRH, BCJ.

Performed research: KV, DMP, KK.

Contributed new reagents or analytic tools: KV, DMP, KK, ME, CM, KJ, HL, FP, TPS,

SRH, BCJ.

Analysed data: KV, DMP.

Wrote the paper: KV.

Page | 110

5.1 Abstract


subcutaneous tissue caused by infection with the pathogen Mycobacterium

ulcerans. After almost 70 years of study in Africa, the mode of transmission

and the non-human reservoir(s) of BU are still largely unknown. The vastly

greater resolution offered by genomics is opening up new possibilities to

explore the pathogen’s cryptic epidemiology and disease ecology. In this

study, we aimed to use comparative second and third generation genomics to

explore the molecular epidemiology of BU at the continental scale, and at the

smaller geographical “village scale” in a BU endemic region of The Democratic

Republic of Congo. We used both temporal associations and the study of the

mycobacterial demographic history to estimate the contribution of humans as

a reservoir in BU transmission. We identified a relationship between the

observed past population dynamics of M. ulcerans from the Songololo

Territory and the timing of health policy changes managing the BU epidemic in

that region. We propose that humans with actively infected, openly

discharging BU lesions inadvertently contaminate community water sources

during bathing/wading and as such indirectly expose naïve individuals to the

etiological agent.

5.2 Introduction


subcutaneous tissue caused by the pathogen Mycobacterium ulcerans [2]. In

BU patients, early diagnosis followed by 8 weeks of treatment with a

combined antibiotic regimen (rifampicin and streptomycin) is key to

preventing complications that can arise from severe skin ulcerations [11]. BU

is a neglected tropical disease, and in some highly endemic areas it is more

prevalent than the most notorious mycobacterial diseases, tuberculosis and

leprosy [83]. Despite the fact that more than 30 countries worldwide have

reported BU, the highest incidence by far is still observed in impoverished,

rural communities of West and Central Africa [5], with 2151 new cases

reported to the WHO in 2014 [6].

Page | 111

M. ulcerans evolved from a Mycobacterium marinum progenitor by

acquisition of the virulence plasmid pMUM001 [18]. This plasmid harbors

genes required for the biosynthesis of the macrocyclic polyketide toxin

mycolactone [13]. This toxin plays a key role in the pathogenesis of BU as it

has cytotoxic, analgesic, and immunomodulatory properties that can cause

painless chronic ulcerative skin lesions with limited inflammation [14]. The

acquisition of the virulence plasmid also introduced insertion sequence (IS)

elements IS2404 and IS2606 into the bacterial chromosome. These IS

elements subsequently expanded extensively in the genome, promoting

genetic rearrangements by modifying gene expression and sequestering

genes, profoundly enhancing mycobacterial genome plasticity [23].

BU epidemiology is characterized by its patchy focal distribution within

endemic countries [5]. Disease foci are known to primarily occur around low-

lying rural marshes, wetlands, and riverine areas [83]. As living or working

close to these slow flowing or stagnant water bodies is a known risk factor for

M. ulcerans infection [31], and as human-to-human transmission is rare, it is

generally believed that M. ulcerans is an environmental mycobacterium that

can infect humans through introduction via a micro-trauma of the skin [32,

33]. Indeed, M. ulcerans DNA has been detected in a variety of aquatic

specimens [35, 36], yet the significance of the detection of M. ulcerans DNA

by PCR in environmental samples remains unclear in the disease ecology of BU

[35-37, 86, 203]. Indeed, environmental sampling surveys show M. ulcerans

DNA can be present in the environment, but the concentration of bacteria in

positive samples is nearly always insignificantly low, indicating that it is

unlikely that M. ulcerans was multiplying in the tested specimens [36, 44].

Furthermore, the frequency of positive sample occurrence is nearly always

low [43]. Finally, definite evidence for the presence of viable M. ulcerans in

potential environmental reservoirs is lacking owing to the strenuous challenge

of culturing the slow growing mycobacterium from non-clinical,

environmental sources [46]. The combination of these limitations clearly

indicates how difficult it is to detect evidence of M. ulcerans in aquatic

environments. As a direct result the mode of transmission and non-human M.

ulcerans reservoir(s) remain poorly understood in Africa [153].

Page | 112

In Victoria, Australia, BU is a considered a zoonosis transmitted from possums

to humans, possibly via a mosquito vector [42, 61-64], although definite proof

is to date still lacking. It is very hard to translate these findings to the endemic

regions in the Afrotropic ecozone. Marsupial mammals like possums are not

endemic in the African continent. Furthermore, M. ulcerans infections in

domesticated and wild animals have never been reported in Africa [36, 65,

66]. Additionally, the results of case-control studies looking into mosquito

related risk factors in Africa are contradictory [29, 30, 67]. A recent extensive

environmental survey in Benin also found no trace of M. ulcerans DNA in

mosquitos [69].

As M. ulcerans has the genome signature of a “niche-adapted”

mycobacterium it is unlikely to be found free-living in various aquatic or

terrestrial environments and is rather more likely living in close association

with a host organism [12]. A recent study used temporal associations to

implicate human-induced changes and activities in the spread of BU across

Africa during the period of Neo-imperialism (late 19th - early 20th century)

(CHAPTER 4). African lineage Mu_A2 was found to be introduced to the

continent during this specific time period. Additionally, close inspection of the

M. ulcerans time-tree revealed that introduction of BU in three well sampled

disease foci coincided with the instigation of colonial rule. Since these disease

foci were inhabited prior to the arrival of the European powers and since it

was only after the start of colonial rule that the specific epidemic clones were

introduced, the study posited that - in the absence of a known BU reservoir or

vector on the African continent [89] - displaced humans with actively infected,

openly discharging BU lesions inadvertently contaminated aquatic

environments during water contact activities and as such spread the

mycobacterium.

The first BU case in the Democratic Republic of the Congo (DRC) was reported

in 1950, in the Kwilu region of the Bandundu Province [204]. Since this first

description, microbiologically confirmed cases have been identified in the

provinces of Bandundu, Katanga, Kinshasa, Kongo Central, Maniema and

South-Kivu [205]. The main endemic focus of BU in the country is located in

the Songololo Territory of the Kongo Central Province and encompasses the

highly endemic rural health zones of Kimpese and Nsona-Mpangu. The

Page | 113

General Reference Hospital of the Institut Médical Evangélique (IME) in

Kimpese receives most of the hospitalized BU patients of this endemic

hotspot. A 2008 study [206] estimated the overall BU prevalence of the

Songololo Territory at 3.3/1000 population, with considerable variation (0.0 to

27.5/1000 population) among the 40 health areas of the territory. Since no

epidemiological studies were conducted in the Territory until the 1960s and

1970s [120, 207, 208], it remained unclear whether BU was a newly

introduced, versus an old expanding illness in the region. Prior to 2002, BU

control in DRC suffered from decades of neglect and conflict affecting the vast

nation’s health and sanitation infrastructure [5].

A better understanding of the transmission and the disease dynamics of M.

ulcerans infection could have a direct impact on the development of effective

and appropriate control strategies against the disease. However, conventional

genetic fingerprinting methods have largely failed to differentiate clinical

disease isolates in this monomorphic organism [93], leading to their

replacement with whole genome sequencing (WGS) [19, 154-156] (CHAPTER

4). The vastly greater resolution offered by genomics is opening up new

possibilities to explore the pathogen’s cryptic epidemiology and disease

ecology. In this study, we have sequenced and compared the genomes of 179

M. ulcerans strains isolated from patients in the Democratic Republic of the

Congo (DRC), The Republic of the Congo (RC) and Angola over a 52 year period

to investigate the microevolution and population dynamics of this pathogen

during its establishment in the region.

5.3 Materials and Methods

Study sites

The study covered all BU foci in DRC, RC and Angola that have ever yielded

positive M. ulcerans cultures. The endemic foci of DRC are located in the

provinces of Bandundu, Kongo Central (previously known as Bas-Congo), and

Maniema. The vast majority of isolates originated from the Songololo

Territory of the Kongo Central Province. Isolates from the low endemic health

zone of Tshela, northwest in the Kongo Central province are also included.

Isolates from the Maniema province originated from the historical BU focus of

Page | 114

the Kasongo territory [209] which was recently assessed to be still active

[210]. Finally, the province of Bandundu is represented by a recently

discovered endemic focus along the Kwango River, a tributary of the Congo

River that forms the border between Angola and DRC [211].

Bacterial isolates and sequencing

We analyzed a panel of 179 M. ulcerans strains originating from disease foci in

DRC, RC and Angola that had been isolated between 1966 and 2014 (Table

5.S1). In addition to the isolates sequenced here, 144 other African genomes

(described in CHAPTER 4) were included to provide appropriate genetic

context for interpreting the diversity and evolution of Central African M.

ulcerans.

Permission for the study was obtained from the ITM Institutional Review

Board (Belgium) and the Ethics Committee of the Public Health School of the

University of Kinshasa (DRC). All patients provided written informed consent

for the collection of samples and subsequent analyses, except when cultures

were analyzed retrospectively, after they had been collected for diagnostic

purposes, as part of routine clinical care. Isolates were processed and

analyzed without use of any patient identifiers, except for approximate

residence location if this information was available.

Based on conventional phenotypic and genotypic methods, bacterial isolates

had previously been assigned to the species M. ulcerans [157]. Mycobacterial

isolates were maintained for prolonged storage at ≤-70°C in Dubos broth

enriched with OAD growth supplement and glycerol.

Index-tagged paired-end sequencing-ready libraries were prepared from

gDNA extracts with the Nextera XT DNA Library Preparation Kit. Genome

sequencing was performed on an Illumina HiSeq 2000 sequencer according to

the manufacturers’ protocols with 100bp or 150bp paired-end sequencing

chemistry. Sequencing statistics are provided in Table 5.S1. The quality of raw

Illumina reads was investigated with FastQC v0.11.3 [158]. Read data for the

study isolates have been deposited in the NCBI Sequence Read Archive (SRA)

under BioProject accession XXX. Prior to further analysis, reads were cleaned

Page | 115

with clip, a tool in the Python utility toolset Nesoni v0.130 [159]. Reads were

filtered to remove those containing ambiguous base calls, any reads <50

nucleotides in length, and reads containing only homopolymers. All reads

were further trimmed to remove residual ligated Nextera adaptors and low

quality bases (<Q10) at the 3' end. The total amount of read-pairs kept after

clipping and their average read length are summarized for all isolates in Table

5.S1.

For isolate Mu_ITM032481, intact, pure, high-molecular-weight gDNA was

obtained by harvesting the growth of 10 Löwenstein-Jensen (LJ) slants

followed by heat inactivation (80°C – 1h), enzymatic digestion (Proteinase K,

lysozyme and RNAse) and purification with the Genomic DNA Buffer Set

(Qiagen, cat. no. 19060) and 100/G Genomic-tips (Qiagen, cat. no. 10243).

This gDNA sample was submitted to the Duke “Sequencing and Genomic

Technologies Shared Resource” for sequencing on a Pacific Biosciences RSII

instrument. Libraries of 15 to 20 kb were constructed and sequenced on 3

SMRT cells using P5-C3 chemistry. This yielded 895 Mbp from a total of

161,629 subreads. The average subread length was 5,536 bp with a

sequencing depth of 160X. Data were analyzed using SMRT Analysis v2.3.0

(Pacific Biosciences). The continuous long reads (CLR) were assembled de

novo using the PacBio Hierarchical Genome Assembly Process 3 (HGAP.3) and

polished using Quiver as previously described [212]. This resulted in a single

contig that was circularized and subsequently annotated using Prokka v1.11

[213]. The annotated closed genome was then manually curated and

visualized using Artemis v.16 [214] and Geneious v9.0.5 [215]. The Congolese

M. ulcerans Mu_ITM032481 bacterial reference chromosome sequence

received strain name SGL03 and was submitted to GenBank with accession

number XXX.

Read mapping and SNP detection

Read mapping and SNP detection were performed using the Snippy v3.0

pipeline [160]. The Burrows-Wheeler Aligner (BWA) v0.7.12 [161] was used

with default parameters to map clipped read-pairs to the new Congolese

SGL_03 reference genome. Due to the unreliability of read mapping in mobile

repetitive regions all ISE elements (IS2404 and IS2606) were hard masked in

Page | 116

these reference genomes (0.398 Mb / 5.625 Mb, i.e. 7% of SGL03). After read

mapping to M. ulcerans SGL03, average read depths were determined with

SAMtools v1.2 [162] and are summarized for all isolates in Table 5.S1. SNPs

were subsequently identified using the variant caller FreeBayes v0.9.21 [163],

with a minimum depth of 10 and a minimum variant allele proportion of 0.9.

Snippy was used to pool all identified SNP positions called in at least one

isolate and interrogate all isolates of the panel at that position. As such a

multiple sequence alignment of “core SNPs” was generated.

Population genetic analysis

Bayesian model-based inference of the genetic population structure was

performed using the “Clustering with linked loci” module [165] in BAPS v.6.0

[166]. The optimal number of genetically diverged BAPS-clusters (K) was

estimated in our data by running the estimation algorithm with the prior

upper bound of K varying in the range of 1-20. Since the algorithm is

stochastic, the analysis was run in 20 replicates for each value of K as to

increase the probability of finding the posterior optimal clustering with that

specific value of K.

On the assumption that patients were infected near their residences, the

latitude and longitude coordinates of a location in the vicinity of patients’

residences at the time of the first clinical visit were collected, including for

retrospective isolates, by using handheld GPS devices (Garmin eTrex 20).

When exact residence locations were missing we used the latitude and

longitude of the village center. QGIS v.2.14.1 [104] was used to generate the

figures of the geographical distribution of Congolese M. ulcerans. The QGIS

Python plugin “Points displacement” was used to modify point shape files,

where point features with the same position overlapped. Point displacement

rendered such features in a circle around the original “real” position.

Geographical analysis of diversity and the overlaying of a phylogenetic tree

was performed with GenGIS v2.5.0 [216], based on the household GPS

coordinates of patients and whole-genome ML phylogenies of the

corresponding M. ulcerans isolates.

Page | 117

Maximum-likelihood phylogenetic analysis

Maximum-likelihood (ML) phylogenies were estimated ten times from SNP

alignments using RAxML v8.2.4 [172] under a plain generalized time reversible

(GTR) model (no rate heterogeneity) with likelihood calculation correction for

ascertainment bias using the Stamatakis method [173]. Identical sequences

were removed before the RAxML runs. For each run we performed 10,000

rapid bootstrap analyses to assess support for the ML phylogeny. The tree

with the highest likelihood across the ten runs was selected. We used

TreeCollapseCL v4 [174] to collapse nodes in the tree with bootstrap values

below a set threshold of 70% to polytomies while preserving the length of the

tree.

Bayesian phylogenetic analysis

We used BEAST2 v2.4.0 [77] to date evolutionary events, determine the

substitution rate, estimate the demographic history, and produce a time-tree

of DRC M. ulcerans. An uncorrelated log-normal relaxed molecular clock [72]

was used with the extended Bayesian skyline demographic method [217] to

infer a genome scale Congolese M. ulcerans time-tree under the GTR

substitution model and with tip-dates defined as the year of isolation (Table

5.S1) (CHAPTER_4). Analysis was performed in BEAST2 using a total of 10

independent chains of 200 million generations, with samples taken every

20,000 MCMC generations. Log files were inspected in Tracer v1.6 [180] for

convergence, proper mixing, and to see whether the chain length produced an

effective sample size (ESS) for all parameters larger than 400, indicating

sufficient sampling. LogCombiner v2.4.0 [77] was then used to combine log

and tree files of the independent BEAST2 runs, after having removed a 30%

burn-in from each run. Thus, parameter medians and 95% highest posterior

density (HPD) intervals were estimated from 70,000 sampled MCMC

generations. To ensure prior parameters were not over-constraining the

calculations, the entire analysis was also run while sampling only from the

prior, and the resulting parameter distributions were compared in Tracer.

TreeAnnotator v2.4.0 [77] was used to summarize the posterior sample of

time-trees in order to as to produce a maximum clade credibility tree with the

posterior estimates of node heights visualized on it.

Page | 118

The estimated timing of population increases and decreases is dependent on

the estimated substitution rate. A potential source of error when estimating

the substitution rate is that tip dates alone, rather than the link of tip dates

associated with sequence data might be driving the results, especially when

the sequence data lacks temporal phylogenetic information [218]. Therefore,

a permutation test was used to assess the validity of the temporal signal in

the data. This was undertaken by performing 20 additional BEAST2 runs (of

200 million MCMC generations each) with identical substitution (GTR), clock

(uncorrelated log-normal relaxed) and demographic models (extended

Bayesian skyline) but with tip dates randomly reshuffled to sequences [219].

This random “null set” of tip-date and sequence correlations was then

compared with the substitution rate estimate of the genuine tip-date and

sequence correlations (CHAPTER 4) [82, 220].

5.4 Results

Genome sequence comparisons of 178 M. ulcerans isolates from Central

Africa

To understand the dynamics and timing of the spread of M. ulcerans across

Central Africa, we sequenced the genomes of 179 clinical isolates that were

obtained between 1966 and 2014 and spanned most of the known endemic

areas of BU in the DRC, RC and Angola (Table 5.S1). Using PacBio reads, a new

complete, closed DRC M. ulcerans reference genome was assembled that

received the strain name SGL03 (“Songololo Territory 2003”). The strain

originated from a patient (Male / 10 years old) from the hamlet Nkondo-

Kiombia (Minkelo Health area) who presented with a severe disseminated

form of BU in 2003. The patient was BU confirmed with all four diagnostic

tests: Ziehl–Neelsen microscopy, IS2404 qPCR, culture and histopathology.

SGL03 comprises a single 5,625,184 bp (6422 bp smaller than Agy99) circular

bacterial chromosome with a G+C content of 65.5 %. Illumina sequence reads

of the sample panel were mapped to the newly assembled M. ulcerans SGL03

reference genome and, after excluding mobile repetitive IS elements and

small insertion-deletions (indels), we detected a total of 6655 SNPs uniformly

distributed across the M. ulcerans chromosome with approximately 1 SNP per

846 bp (0.12% nucleotide divergence) (Figure 5.S1).

Page | 119

The population structure of M. ulcerans in Central Africa

A Bayesian maximum clade credibility phylogeny was inferred from a whole-

genome alignment of the isolates (Figure 5.1). Both known lineages of African

M. ulcerans were identified within the Central African isolate panel: 177/178

(99.4%) corresponded to lineage Africa I (Mu_A1) and 1/178 (0.6%)

corresponded to the uncommon lineage Africa II (Mu_A2). The single Mu_A2

isolate originated from a patient (F/40) from the hamlet Kilima in the

Songololo Territory (Nkamuna Health area).

The average pairwise SNP difference (SNPΔ) between Mu_A1 Central African

isolates was low indicating that the majority of the discovered diversity

derived from the large genetic distance between Mu_A1 and the single

Mu_A2 isolate from the region.

Page | 120

Figure 5.1: Bayesian maximum clade credibility phylogeny for DRC, RC and Angolan M.

ulcerans. The tree was visualized and colored in Figtree v1.4.2 [102]. Branches are color coded

according to their branch specific substitution rate (legend at top). Branches defining major

lineages are annotated on the tree. Tip labels of Songololo Territory isolates are color coded

according to their respective BAPS-clusters. Divergence dates (mean estimates and their

respective 95% HDP) are indicated in black for major nodes.

Page | 121

Phylogenetic analysis reveals strong geographical restrictions on M. ulcerans

dispersal at high-level geographical scales

Within an M. ulcerans phylogeny of the entire African continent (Figure 5.S2),

the single Songololo Mu_A2 isolate clusters together with a clade of 8 other

Mu_A2 isolates originating from Benin, Gabon, and Cameroon. Furthermore,

a distinct Mu_A1 isolate from RC (ITM_071925) clusters together with a small

clade of Nigerian and Cameroonian M. ulcerans. More importantly however,

all other 176 Mu_A1 isolates of the Central African panel form a strongly

supported (posterior probability = 1) monophyletic group within that

continental African phylogeny. Within that clade we can discern an

unambiguous relationship between the genotype of an isolate and its

geographical origin. This is illustrated by the clear spatial clustering of M.

ulcerans from the different endemic BU foci within the Bayesian phylogeny.

For instance, all 123 isolates of the endemic BU focus of the Songololo

Territory form a strongly supported monophyletic group (posterior probability

= 1) (Figure 5.2). The Songololo Territory isolates are unrelated to the 4

isolates form the neighboring Tshela Territory (northwest in the Kongo Central

province) which form a separate monophyletic group (posterior probability =

1) (Figure 5.2).

Page | 122

Figure 5.2: Phylogeography of DRC, RC and Angolan M. ulcerans. A Maximum-likelihood

phylogeny is drawn for lineage Mu_A1 with branches color coded according to BU disease

focus (legend bottom right). The ML phylogeny is based on 1373 SNP differences detected

across the whole core genome of 135 sequenced isolates with GPS data. Nodes in the tree

with bootstrap support below a set threshold of 70% were collapsed to polytomies, while

preserving the length of the tree. The green clade formed by 123 isolates from the Songololo

Territory disease focus is collapsed in the tree. The tips of the tree are connected to the

location of residence of patients from whom the strain was isolated. The administrative

borders of countries were obtained from the Global Administrative Unit Layers dataset of

FAO. The river layer was translated from the River-Surface Water Body Network data set of

the African Water Resource database of FAO.

Page | 123

The clustering of M. ulcerans genotypes ends at low-level geographical

scales

We then explored the geographical distribution of M. ulcerans genotypes at a

lower geographical scale: that of the Songololo Territory. The territory covers

an area of 8190 km² and is limited to the north by the Congo River, to the east

by the Kwilu-Ngongo health zone, to the south by the northern border of

Angola, and to the west by the Sekebanza health zone (Figure 5.S3). The

Songololo Territory has the highest case notifications, yet the limits of this

endemic focus are somewhat extended into the above mentioned bordering

health zones. The 123 Songololo isolates originating from 123 unique patients

are spread rather evenly over the territory and the majority of Health Areas

with a “modest” to “high” BU burden are well represented (Figure 5.S3).

Bayesian model-based inference of the genetic population structure revealed

the existence of six BAPS-clusters within the territory (Figure 5.3). The six

clusters occur somewhat intermixed, as in some regions of the territory,

multiple clusters are found circulating simultaneously. In the Health Area of

Lovo for instance, up to 5 different clusters are co-circulating (BAPS 1, 2, 3, 4,

5). The clusters are however distributed differently over the study region:

clusters 2, 4 and 5 are found rather widely dispersed, while cluster 1, 3 and 6

are more geographically restricted (Figure 5.3). Cluster 1 (n=20) is found

almost exclusively in the east of the Territory while cluster 3 (n=31) is

localized in the west (Figure 5.S4). Cluster 6, finally, is uncommon (n=4) and

found solely in the south-west. Moreover, within the clusters there are some

distinct cluster subtypes, which occasionally also have a limited distribution

across the region. For example, one specific sub-cluster of BAPS cluster 2

consist of 7 isolates that all originate from a 90 km² zone covering the

neighboring health areas of Mukimbungu and Kasi (Figure 5.S4). However,

other sub-clusters can be far more broadly distributed; with the extreme

example of identical genomes identified in different BU patients separated by

large geographical distances (Figure 5.3; I-X). A total of ten such genomes that

were identified multiple times in the Songololo Territory can be discerned

(Table 5.1). The average geographical distance between the domiciles of

patients identified with an identical genome is 17,3 ± 18,1 km.

Page | 124

Table 5.1: Identical genomes identified in different BU patients of the

Songololo Territory. YOI: Year Of Isolation; Not Applicable: “-“

Isolate 1

(YOI)

Isolate 2

(YOI)

Isolate 3

(YOI)

Geographical

distance (km)

Years between

isolation dates

Identical Genome Set I ITM102560

(2009)

ITM131951

(2008)

ITM141716

(2013)

26,6 5

Identical Genome Set II ITM130328

(2011)

ITM130330

(2011)

- 56,5 0

Identical Genome Set III ITM130336

(2012)

ITM131959

(2013)

- 0,0 1

Identical Genome Set IV ITM081364

(2007)

ITM082600

(2007)

- 0,1 0

Identical Genome Set V ITM141715

(2013)

ITM141729

(2014)

- 37,2 1

Identical Genome Set VI ITM072731

(2007)

ITM141740

(2014)

- 18,7 7

Identical Genome Set VII ITM112015

(2011)

ITM112016

(2011)

- 6,0 0

Identical Genome Set VIII ITM073463

(2007)

ITM141700

(2013)

- 11,2 6

Identical Genome Set IX ITM081935

(2007)

ITM110809

(2009)

- 5,7 2

Identical Genome Set X ITM141709

(2013)

ITM141717

(2013)

- 11,4 0

Page | 125

Figure 5.3: Phylogeography of the Songololo Territory BU disease focus. A Maximum-

likelihood phylogeny is drawn for lineage Mu_A1. The ML phylogeny is based on 684 SNP

differences detected across the whole core genome of 123 sequenced isolates from the

Songololo Territory with GPS data. Branches are color coded according to their respective

BAPS-clusters as indicated in the legend. Nodes in the tree with bootstrap support below a set

threshold of 70% were collapsed to polytomies, while preserving the length of the tree. The

location of residence of patients from whom the isolate was grown is colored according to the

BAPS-cluster the corresponding isolate belonged to. Identical genomes identified in different

patients are interconnected by the green curves, which are annotated with Roman numerals.

The background map was created using elevation data from the Shuttle Radar Topography

Mission (SRTM). The river layer (Congo river and its tributaries) was digitized from declassified

Soviet military topographic maps xb33-13, xb33-14, xb33-15, xb33-16, and xb33-17 (scale

1:200k) and xb33-1, and xb33-1 (scale 1:500k).

Page | 126

The Central African mutation rate of M. ulcerans is similar to the one

inferred on a continental scale

A genome scale Central African M. ulcerans time-tree was inferred (Figure 5.1)

while also providing estimates of nucleotide substitution rates and divergence

times for key M. ulcerans clades. In this process, a molecular clock was

estimated using correlations between phylogenetic divergence and isolation

times of heterochronous disease isolates. We estimate a mean genome wide

substitution rate of 4.38E-8 per site per year (95% HPD interval [2.83E-8 -

6.03E-8]), corresponding to the accumulation of 0.23 SNPs per chromosome

per year (95% HPD interval [0.15 – 0.32]) (excluding IS elements). The analysis

also indicates that the genealogy has undergone very moderate substitution

rate variation, with a 6-fold difference between the slowest (1.22E-08) and

the fastest (8.06E-08) evolving branches. Rate accelerations and decelerations

are found interspersed in the time-tree (Figure 5.1). When we estimated the

substitution rate by informing the prior distribution of the substitution rate in

the Bayesian analysis with that of a previous estimate (CHAPTER 4) we obtain

a similar result (4.56E-8 per site per year (95% HPD interval [2.53E-8 - 6.01E-

8])).

The Bayesian analysis (Figure 5.1) indicates that lineage Mu_A1 has been

endemic in the Central Africa for hundreds of years (tMCRA(Mu_A1) = 1372

A.D.(95% HPD 913 - 1776)). The time-tree of Central African M. ulcerans also

reveals the timing of the BU introduction event in the Songololo Territory:

tMCRA(Songololo)= 1865 A.D. (95% HPD 1803 - 1915). Finally, the time-tree

also indicates that the separated “eastern” (tMCRA (BAPS-1)=1941 A.D. (95%

HPD 1908 - 1969)) and “western” (tMCRA (BAPS-3)=1922 A.D. (95% HPD 1885

- 1954)) Songololo clusters have most likely remained segregated over a

timespan of half a century.

The demographic history of M. ulcerans in the Songololo Territory

The reconstruction of the demographic history of M. ulcerans in the Songololo

Territory involved the estimation of the phylogeny and the inference of the

mycobacterial population size (Ne.τ) at different points along the timescale of

that phylogeny. There were sufficient event times informing the population

Page | 127

dynamics between 1880 and 2014. Inspection of the extended Bayesian

skyline plot (EBSP) (Figure 5.4) suggested that the M. ulcerans population size

remained stable until the early 1980s, after which it increased slightly during

the course of the nighties, until it reached a peak around 2004. This was

followed by a small decline that persisted until 2014.

We identified a relationship between the observed past population dynamics

of M. ulcerans from the Songololo Territory and the timing of health policy

changes managing the BU epidemic in that region. More than 500 cases of BU

had been reported before 1980, after which there was a 20 year long “silent”

period in the scientific literature during which no cases were reported [5].

During this period the hospital lost the majority of its specialized personnel,

which was partially due to the political situation in DRC at that time. This lead

to IME’s lowest recorded (all-cause) admission rate of 4,5 patients/year

between 1989 and 1999 [221]. Later, in 2002, a national BU program was

started (Program National de Lutte contre L’Ulcère de Buruli - PNLUB) and

during 2002-2004 an apparent resurgence of BU was reported in the

Songololo Territory [222]. Since the end of 2004 the General Reference

Hospital of the Institut Médical Evangélique (IME) in Kimpese launched a

specialized BU program (sponsored by American Leprosy Missions), offering

free-of-charge treatment and supplementary aid. Since the start of the BU

control project a strong increase was noted in the number of notified BU

cases, including those admitted to IME hospital [223].

We checked for two factors that could bias the reconstruction of the

mycobacterial population size over time: population structure and lack of

temporal information in the alignment. To test whether our results were

influenced by any of these factors we conducted extensive resampling and

randomization experiments. To test the validity of the discovered temporal

signal in the data we performed 20 date-randomization tests. The estimate of

the substitution rate passed this test as its mean did not fall within the 95%

HPD intervals of rate estimates obtained using replicate data sets in which the

sampling times were randomized (Figure 5.S5). This indicated the spread of

sampling times was sufficient to allow the substitution rate to be estimated

reliably.

Page | 128

Figure 5.4: The demographic history of M. ulcerans in the Songololo Territory. Extended

Bayesian Skyline plot; the central dotted line represents the median mycobacterial population

size (Ne.τ) with its 95% central posterior density (CPD) interval represented by the upper and

lower lines. Note the y-axis is on the log scale. The EBSP method was implemented in a

framework in which the phylogeny, the demographic history and all other parameters of the

model of molecular evolution were co-estimated in a single analysis. As a result the resulting

plot of the population history includes credibility intervals that represent the combined

phylogenetic and coalescent uncertainty. Ne reflects the number of individuals that contribute

offspring to the next generation and will always be smaller than the “real” population size. τ

represents the generation time.

Page | 129

5.5 Discussion

The demographic history of a pathogen population leaves a “signature” in the

genomes of modern representatives of that population [224]. Reconstructing

this history can allow us to gain valuable insights into the processes that drove

past population dynamics [76]. We recognized a relationship between the

mycobacterial population dynamics and the timing of health policy changes

managing the BU epidemic in the Songololo Territory. In the Territory, the

mycobacterial population size increased in the period without BU control

activities, as reflected by lack of notified BU cases, with a subsequent subtle

decline when BU control activities resumed. To test whether our results were

not biased we conducted extensive resampling and randomization

experiments and used a wide range of parameter settings. All additional

analyses confirmed our results. The bacterial population size increased prior

to the start of a national BU program in the country during a period of

decreased attention to BU that resulted in the loss of specialized personnel.

After the start of the program and the implementation of free-of-charge

treatment, a strong increase was noted in the number of admitted BU cases

which concorded with a detected inflection - perhaps a small drop - of the

mycobacterial population size. These observations suggest that control

strategies at the public health level may have decreased the size of the human

M. ulcerans reservoir and that this reservoir is important in sustaining new

infections. These analyses furthermore indicate that even if other

environmental reservoirs exists, the number of M. ulcerans infections will

decrease, even if only human cases are treated.

The M. ulcerans phylogeography revealed one almost exclusively

predominant sublineage of Mu_A1 that arose in Central Africa and

proliferated in the different endemic foci of DRC, Angola, and RC during the

Age of Discovery (15th - 18th century). The principal sublineage of Mu_A1 was

introduced into the Songololo Territory around 1865 (95% HPD 1803-1915)

and, over the subsequent century (1865-1974), it established itself and

evolved in six distinct clusters across the territory. This timing is consistent

with in-depth interviews with former patients and observations of healed

lesions that had suggested that M. ulcerans infections already occurred in the

Songololo Territory in 1935 and probably even earlier [120]. The genome-

Page | 130

based time-tree of Central African M. ulcerans thus revealed that the

Songololo Territory became endemic during the time when the region was

being colonized by Belgium. Early during the Belgian occupation, the

Songololo territory was developed heavily to link the oceanic harbor of

Matadi by rail with Kinshasa, where the Congo River becomes navigable,

opening up the entire interior of DRC for economic exploitation [225]. The

Songololo Territory was already inhabited long before the arrival of the

European colonizers. The Kongo People are believed to have settled at the

mouth of the Congo River before 500 BC, as part of the larger Bantu migration

[226]. However, our data reveal that it was only after the start of colonial rule

that the epidemic Songololo M. ulcerans clone was introduced, possibly

through the arrival of displaced BU-infected humans.

Similar to recent studies that used comparative genomics to investigate the

microevolution of M. ulcerans during its establishment in a continent

(CHAPTER 4) or region [154, 155], the genotypes in Central Africa show strong

spatial segregation (Figure 5.2). This was illustrated by the explicit regional

clustering of M. ulcerans from the different endemic BU foci (Songololo,

Tshela, Kwango) within the phylogenies. This clustering of cases within

endemic foci reflects a common source of infection within the disease focus.

These repeated findings indicate that when M. ulcerans is introduced in a

particular area, it remains isolated, resulting in a localized clonal expansion

associated with that area. Inspection of the time tree shows that a clonal

complex associated with an endemic focus often has been pervasive in that

region for a considerable time; in the case of Songololo around 150 years.

Even within the Songololo Territory we observed that the separated eastern

(tMCRA (BAPS-1)=1941) and western (tMCRA (BAPS-2)=1922) Songololo

clusters have most likely remained segregated over a timespan of half a

century, indicating that M. ulcerans spreads relatively slowly between

neighboring regions. This also indicates that environmental reservoirs of the

mycobacterium in that region had to remain localized and relatively isolated.

Unlike the larger geographic scale data, at smaller geographical scales

genotypes start to intermingle. Firstly, four Songololo BAPS clusters were

found to be co-circulating. Moreover, we observed even completely identical

genomes originating from patients living in villages separated by distances of

Page | 131

on average 17 km, similarly to the findings of a recent studies [154, 156]. We

believe the observed “breakup” of the focal distribution pattern at smaller

geographical scales can be explained by the determined low substitution rate

that corresponds to the accumulation of just 0.23 SNPs per bacterial

chromosome per year. The slow substitution rate severely limits the

accumulation of point mutations and as such lowers the resolving power of

the genomes. This explains why (in the most extreme case), over a period of

five years an identical genome was discovered in three patients who lived in

three different villages separated by 26 km: insufficient time has elapsed for

point mutations to accumulate.

Finally, an old debate relating to the role of Angolan refugees on a resurge of

BU in the Songololo territory [227] can be settled. In the aftermath of the

Angolan civil war (which ended in 2002) BU was frequently diagnosed in

Angolan refugees who lived in refugee camps located in the Songololo

Territory. As cases have been reported in Angola [228] the possibility existed

that these patients were infected in Angola and re-introduced BU in the

region. Most of these Angolans however had already lived in DRC for several

years before their diagnosis, and some young Angolan patients were born in

DRC and had never visited Angola. Analysis of the phylogenies shows that no

typical Angolan genotypes were detected in Songololo, indicating that the

refugees were in all likelihood infected in the DRC.

In conclusion, in the present study we used both temporal associations and

the study of the mycobacterial demographic history in an endemic focus to

implicate human-induced changes and activities over (recent) historical scales

behind the expansion of BU in Central Africa. We propose that humans with

actively infected, openly discharging BU lesions inadvertently contaminate

aquatic environments during water contact activities and thus play the pivotal

role in the spread of the mycobacterium. A total of 74% of BU patients

identified during a cross-sectional study [206] of the Songololo Territory had

ulcerative lesions (49% category I, 31% category II, and 20 % category III)

indicating a high percentage of patients might be shedding bacteria into the

environment potentially infecting others. Based on these conclusions we

suggest that in BU-affected areas, chains of transmission will be broken and

the spread of disease stopped through improved disease surveillance, active

Page | 132

case-finding programs and early treatment of pre-ulcerative infection. This

view is supported by the decline of BU incidence recorded in some areas

which profited from both improved BU surveillance and early treatment [197].

Future longitudinal micro-epidemiological studies involving comparative

genomics and SNP-typing of clinical and environmental samples could possibly

provide even deeper insights into the transmission pathways of M. ulcerans.

We anticipate that the SNPs described here will provide useful genetic

markers for future M. ulcerans transmission pathway tracing in the Songololo

Territory.


Figure 5.S1: Distribution of SNPs identified in the entire sequenced sample panel (Mu_A1 &

Mu_A2) compared to the Congolese M. ulcerans SNG03 reference genome. The Y-axis

corresponds to SNP counts per 10,000bp window, the dashed line indicates the average rate

of 11.8 SNPs per 10,000 bp (or 1 SNP per 846 bp window).

Page | 133

Figure 5.S2: The position of Central African isolates in the continental African M. ulcerans

tree. Bayesian maximum clade credibility phylogeny drawn for lineage Mu_A1 and Mu_A2 for

the 179 isolates sequenced in this study plus 144 African M. ulcerans isolates sequenced

previously (total n=323). DRC, RC, and Angolan isolates are highlighted in pink in the

phylogeny. Branches defining major lineages are annotated on the tree.

Page | 134

Figure 5.S3: Sampling effort and the distribution of the BU disease burden in the Health zones

and Health areas of the Kongo Central province of RDC. The distribution of all BU cases per

Health zone (top) or Health area (bottom) was determined from all disease notifications

reported since the start of the national BU program (PNLUB) in 2002 until 2014. The red

crosses denote the General Reference Hospitals of the Institut Médical Evangélique (IME) in

Kimpese and that of Nsona-Mpangu. Yellow points represent the residence of BU patients

from whom M. ulcerans disease isolates were grown at the time of clinical visit.

Page | 135

Figure 5.S4: Detailed view of the phylogeography of BAPS clusters 1 (“eastern”) and 3

(“western”) and sub cluster Mukimbungu-Kasi of the Songololo Territory BU disease focus.

For clarity, not all connecting lines are plotted. The distribution of isolates and the overlaying

of the phylogenetic tree was performed with GenGIS v2.5.0, based on the household GPS

coordinates of each patient and the whole-genome ML phylogeny (same as in Figure 5.3) of

their corresponding M. ulcerans isolates.

Page | 136

Figure 5.S5: Comparison of Bayesian estimates of nucleotide substitution rates for real and

randomized tip dates. Filled squares & circles represent mean estimates, while bars indicate

values of the 95% highest probability density (HDP) interval. The estimate obtained using the

real tip date associations (circle) is shown to the far right while estimates from random

associations (squares) are shown to the left. All randomized data sets were analyzed in

BEAST2 using identical model settings as used in the analysis of the real tip date data. Note

the y-axis is on the log scale.

Page | 137

Table 5.S1: M. ulcerans isolate and DNA sequencing information. 1: Democratic Republic of

the Congo

Isolate n° Lineage BAPS cluster YOI Country of Origin Latitude Longitude Coverage (x) Average read length (bp)

ITM030221 Mu_A1 BAPS_1 2002 DRC1

-5.38942 14.30403 63 95

ITM030719 Mu_A1 BAPS_1 2002 DRC -5.42357 14.48067 56 95

ITM030952 Mu_A1

2002

66 96

ITM031677 Mu_A1 BAPS_1 2003 DRC -5.49655 14.69358 60 95

ITM032481 Mu_A1 BAPS_2 2003 DRC -5.60384 14.13927 89 138

ITM040358 Mu_A1 BAPS_2 2003 DRC -5.73894 14.04787 61 95

ITM041706 Mu_A1

2003 DRC -4.722099 13.041641 61 95

ITM050303 Mu_A1

1966 Congo -4.432895 12.299497 59 138

ITM050715 Mu_A1 BAPS_1 2004 DRC -5.55598 14.45739 71 91

ITM060976 Mu_A1

2005 Congo -4.761918 11.868028 52 95

ITM060977 Mu_A1 BAPS_2 2005 DRC -5.09303 14.07327 52 95

ITM062479 Mu_A1

2006

52 95

ITM063519 Mu_A1 BAPS_3 2006 DRC -5.6969 14.05914 56 139

ITM070118 Mu_A1 BAPS_4 2006 DRC -5.55598 14.45739 65 95

ITM070121 Mu_A1 BAPS_6 2006 DRC -5.76913 13.92102 66 95

ITM070122 Mu_A1 BAPS_2 2006 DRC -5.50590 14.52313 67 95

ITM070123 Mu_A1 BAPS_1 2006 DRC -5.56652 14.43447 72 139

ITM070124 Mu_A1 BAPS_3 2006 DRC -5.67882 13.96647 69 95

ITM071491 Mu_A1 BAPS_2 2006 DRC -5.68600 13.88951 53 95

ITM071499 Mu_A1

2006 Angola

69 96

ITM071910 Mu_A1 BAPS_5 2006 DRC -5.46310 13.78095 59 95

ITM071912 Mu_A1 BAPS_2 2007 DRC -5.09275 14.02277 62 96

ITM071913 Mu_A1 BAPS_2 2007 DRC -5.12371 14.01776 57 95

ITM071925 Mu_A1

2007 Congo -4.433392 11.700150 54 140

ITM072398 Mu_A1 BAPS_5 2006 DRC -5.52036 13.90937 65 140

ITM072399 Mu_A1 BAPS_2 2006 DRC -5.76827 13.92186 60 95

ITM072400 Mu_A1 BAPS_6 2007 DRC -5.76913 13.92102 60 95

ITM072401 Mu_A1 BAPS_3 2007 DRC -5.76785 13.92162 74 140

ITM072731 Mu_A1 BAPS_2 2007 DRC -5.15632 14.18982 56 95

ITM072732 Mu_A1 BAPS_3 2007 DRC -5.76854 13.91985 44 140

ITM072733 Mu_A1 BAPS_2 2007 DRC -5.39494 14.12631 57 140

ITM072734 Mu_A1 BAPS_2 2007 DRC -5.68719 13.89086 40 140

ITM072735 Mu_A1 BAPS_2 2007 DRC -5.70311 14.1483 40 139

ITM072737 Mu_A1 BAPS_2 2006 DRC -5.68680 13.89025 62 95

ITM072840 Mu_A1 BAPS_2 2007 DRC -5.76805 13.91836 83 140

ITM073461 Mu_A1 BAPS_2 2007 DRC -5.75556 13.98184 66 92

ITM073463 Mu_A1 BAPS_2 2007 DRC -5.67491 14.01439 74 139

ITM073477 Mu_A1 BAPS_2 2007 DRC -5.46071 13.78244 48 139

ITM073676 Mu_A1 BAPS_3 2007 DRC -5.76855 13.92024 63 95

ITM080059 Mu_A1 BAPS_1 2007 DRC -5.70821 14.04022 59 95

ITM080060 Mu_A1 BAPS_2 2007 DRC -5.77422 13.94787 59 95

ITM080062 Mu_A1 BAPS_3 2007 DRC -5.77508 14.16654 63 91

ITM080064 Mu_A1 BAPS_2 2007 DRC -5.56748 14.12015 65 95

ITM080098 Mu_A1

2007 Angola -8.515200 17.828573 65 95

Page | 138


ITM080285 Mu_A1 BAPS_4 2007 DRC -5.75498 14.43446 58 95

ITM080288 Mu_A1 BAPS_4 2007 DRC -5.75519 14.43433 52 95

ITM081018 Mu_A1 BAPS_1 2007 DRC -5.73305 14.45571 57 95

ITM081364 Mu_A1 BAPS_5 2007 DRC -5.41084 14.49715 57 95

ITM081433 Mu_A1

2007 DRC -4.722099 13.041641 62 95

ITM081434 Mu_A1

2007 DRC -4.722099 13.041641 67 95

ITM081436 Mu_A1

2007 DRC -4.722099 13.041641 59 95

ITM081935 Mu_A1 BAPS_2 2007 DRC -5.75556 13.98184 62 95

ITM081937 Mu_A1 BAPS_2 2007 DRC -5.76938 13.93092 72 96

ITM082001 Mu_A1

2008 DRC

68 95

ITM082494 Mu_A1 BAPS_1 2007 DRC -5.56121 14.44862 64 95

ITM082495 Mu_A1 BAPS_4 2007 DRC -5.56121 14.44862 81 96

ITM082498 Mu_A1 BAPS_5 2007 DRC -5.41089 14.49713 68 95

ITM082499 Mu_A1 BAPS_4 2008 DRC -5.81782 14.19031 64 95

ITM082575 Mu_A1

2008 DRC

66 95

ITM082600 Mu_A1 BAPS_5 2007 DRC -5.41166 14.49669 41 139

ITM083232 Mu_A1

2008 Angola -9.418594 18.432287 56 140

ITM083545 Mu_A1

2008 DRC

67 95

ITM083546 Mu_A1

2008 DRC

63 95

ITM083547 Mu_A1

2008 DRC

72 95

ITM083663 Mu_A1

2008 DRC

82 96

ITM083664 Mu_A1

2008 DRC

83 96

ITM083665 Mu_A1

2008 DRC

80 96

ITM090058 Mu_A1

2008 DRC

65 95

ITM090059 Mu_A1 BAPS_3 2007 DRC -5.76785 13.92162 47 93

ITM090244 Mu_A1 BAPS_3 2008 DRC -5.37190 14.37605 58 95

ITM090245 Mu_A1 BAPS_4 2008 DRC -5.72645 14.39808 58 95

ITM090250 Mu_A1

2008 DRC

67 96

ITM090252 Mu_A1

2008

66 96

ITM090253 Mu_A1 BAPS_2 2008 DRC -5.56121 14.44862 64 95

ITM091058 Mu_A1

2008 DRC

72 96

ITM091059 Mu_A1

2008 DRC

78 96

ITM091155 Mu_A1 BAPS_3 2008 DRC -5.72898 14.44607 87 96

ITM091795 Mu_A1 BAPS_2 2008 DRC -5.64428 14.18763 83 96

ITM092479 Mu_A1

2009

UNKNOWN UNKNOWN 67 138

ITM092522 Mu_A1

2008

66 95

ITM100140 Mu_A1 BAPS_2 2009 DRC -5.70385 14.39294 90 139

ITM100831 Mu_A1 BAPS_2 2009 DRC -5.73816 14.04989 70 95

ITM100833 Mu_A1 BAPS_3 2009 DRC -5.29764 14.03725 54 140

ITM101864 Mu_A1

2009 DRC -4.825456 15.179014 68 95

ITM102560 Mu_A1 BAPS_3 2009 DRC -5.73978 14.04599 86 96

ITM102561 Mu_A1 BAPS_3 2009 DRC -5.73978 14.04599 63 95

ITM102562 Mu_A1 BAPS_2 2010 DRC -5.51255 13.90563 57 95

ITM102563 Mu_A1

2010

81 96

ITM102564 Mu_A1 BAPS_3 2010 DRC -5.701165 14.038949 66 96

ITM102565 Mu_A1 BAPS_2 2009 DRC -5.38319 14.51654 57 95

ITM110183 Mu_A1 BAPS_3 2009 DRC -5.09500 14.02389 68 96

Page | 139


ITM110798 Mu_A1 BAPS_5 2008 DRC -5.25116 14.14875 70 96

ITM110799 Mu_A1 BAPS_3 2008 DRC -5.51917 13.90905 82 96

ITM110800 Mu_A1 BAPS_3 2008 DRC -5.67940 13.96586 90 96

ITM110801 Mu_A1 BAPS_2 2009 DRC -5.67933 13.96667 80 96

ITM110804 Mu_A1 BAPS_2 2011 DRC -5.68124 13.91436 82 96

ITM110805 Mu_A1

2008 DRC

59 95

ITM110807 Mu_A1 BAPS_3 2009 DRC -5.40057 14.11641 70 96

ITM110808 Mu_A1 BAPS_2 2009 DRC -5.76894 13.92052 66 96

ITM110809 Mu_A1 BAPS_2 2009 DRC -5.74082 14.03083 72 96

ITM110810 Mu_A1 BAPS_5 2010 DRC -5.47175 14.46689 75 96

ITM110811 Mu_A1 BAPS_1 2010 DRC -5.41990 14.47082 79 96

ITM110812 Mu_A1

2010 DRC

81 96

ITM110813 Mu_A1 BAPS_1 2010 DRC -5.55463 14.46311 89 96

ITM110891 Mu_A1 BAPS_3 2010 DRC -5.6969 14.05914 81 96

ITM112014 Mu_A1 BAPS_1 2011 DRC -5.45031 14.46770 58 95

ITM112015 Mu_A1 BAPS_2 2011 DRC -5.14118 14.03652 63 95

ITM112016 Mu_A1 BAPS_2 2011 DRC -5.09462 14.02354 81 96

ITM120108 Mu_A1

2011

82 96

ITM120109 Mu_A1

2011

73 96

ITM130324 Mu_A1 BAPS_2 2011 DRC -5.73737 14.04630 74 96

ITM130325 Mu_A1

2011 DRC

87 96

ITM130326 Mu_A1 BAPS_2 2011 DRC -5.76861 13.92081 79 96

ITM130327 Mu_A1 BAPS_2 2011 DRC -4.910070 14.166960 88 96

ITM130328 Mu_A1 BAPS_1 2011 DRC -5.54598 13.95093 70 96

ITM130329 Mu_A1

2011

72 96

ITM130330 Mu_A1 BAPS_1 2011 DRC -5.47032 14.45564 70 96

ITM130333 Mu_A1 BAPS_4 2007 DRC -5.60976 14.08427 67 96

ITM130334 Mu_A1 BAPS_1 2012 DRC -5.56121 14.44862 67 96

ITM130335 Mu_A1 BAPS_3 2011 DRC -5.75556 13.98184 70 96

ITM130336 Mu_A1 BAPS_1 2012 DRC -5.50590 14.52313 64 96

ITM130337 Mu_A1 BAPS_2 2012 DRC -5.42084 13.82826 81 96

ITM130338 Mu_A1 BAPS_2 2012 DRC -5.37215 13.78772 75 96

ITM130340 Mu_A2

2013 DRC -5.76894 13.93317 73 96

ITM130341 Mu_A1 BAPS_2 2013 DRC -5.76913 13.92102 70 96

ITM130342 Mu_A1 BAPS_6 2012 DRC -5.75556 13.98184 66 96

ITM130343 Mu_A1 BAPS_2 2012 DRC -5.09502 14.02358 67 96

ITM130347 Mu_A1 BAPS_6 2012 DRC -5.74139 14.03020 70 96

ITM130348 Mu_A1 BAPS_4 2012 DRC -5.47705 13.78943 67 96

ITM131942 Mu_A1

2012

86 96

ITM131943 Mu_A1 BAPS_5 2012 DRC -5.70424 14.39377 79 96

ITM131944 Mu_A1 BAPS_2 2012 DRC -5.74140 14.03005 73 96

ITM131945 Mu_A1 BAPS_2 2012 DRC -5.75556 13.98184 73 96

ITM131947 Mu_A1 BAPS_1 2012 DRC -5.59428 14.44455 74 96

ITM131948 Mu_A1 BAPS_3 2012 DRC -5.68378 13.91524 65 96

ITM131950 Mu_A1 BAPS_1 2012 DRC -5.50590 14.52313 73 96

ITM131951 Mu_A1 BAPS_3 2008 DRC -5.58945 13.94445 77 96

ITM131953 Mu_A1 BAPS_3 2012 DRC -5.68112 13.91754 87 96

Page | 140


ITM131954 Mu_A1 BAPS_2 2012 DRC -5.38895 13.85555 72 96

ITM131955 Mu_A1

2013 DRC -6.479574 16.811128 68 96

ITM131957 Mu_A1 BAPS_1 2013 DRC -5.50590 14.52313 76 96

ITM131959 Mu_A1 BAPS_1 2013 DRC -5.50590 14.52313 75 96

ITM131960 Mu_A1 BAPS_3 2013 DRC -5.75556 13.98184 80 96

ITM131961 Mu_A1 BAPS_2 2013 DRC -5.50388 13.87502 72 96

ITM131963 Mu_A1 BAPS_3 2008 DRC -5.58674 13.95612 71 96

ITM131964 Mu_A1

2013 DRC

81 96

ITM141700 Mu_A1 BAPS_2 2013 DRC -5.76605 13.96978 118 125

ITM141701 Mu_A1 BAPS_3 2013 DRC -5.77753 14.16762 120 125

ITM141706 Mu_A1

2013 DRC

120 125

ITM141707 Mu_A1

2013 DRC

120 125

ITM141709 Mu_A1 BAPS_2 2013 DRC -5.61116 14.08526 113 125

ITM141710 Mu_A1 BAPS_2 2013 DRC -5.60220 14.13882 118 125

ITM141711 Mu_A1

2013 DRC

129 125

ITM141712 Mu_A1

2013 DRC

128 125

ITM141715 Mu_A1 BAPS_3 2013 DRC -5.37203 13.78832 138 125

ITM141716 Mu_A1 BAPS_3 2013 DRC -5.79074 14.05777 124 125

ITM141717 Mu_A1 BAPS_2 2013 DRC -5.62628 14.18642 138 125

ITM141719 Mu_A1

2013 DRC

131 125

ITM141722 Mu_A1 BAPS_3 2013 DRC -5.52901 14.07047 128 125

ITM141723 Mu_A1

2013 DRC

133 125

ITM141724 Mu_A1 BAPS_1 2013 DRC -5.72464 14.47332 118 125

ITM141725 Mu_A1 BAPS_2 2013 DRC -5.74056 14.04931 132 125

ITM141727 Mu_A1

2013 DRC

121 125

ITM141728 Mu_A1 BAPS_3 2013 DRC -5.58575 13.95549 141 125

ITM141729 Mu_A1 BAPS_3 2014 DRC -5.68276 13.91298 142 125

ITM141732 Mu_A1

2013 DRC

131 125

ITM141733 Mu_A1

2013 DRC

141 125

ITM141736 Mu_A1 BAPS_3 2013 DRC -5.76594 13.97003 121 125

ITM141737 Mu_A1 BAPS_2 2013 DRC -5.74126 14.05073 118 125

ITM141738 Mu_A1 BAPS_4 2013 DRC -5.70475 14.39368 118 125

ITM141739 Mu_A1

2013 DRC

125 125

ITM141740 Mu_A1 BAPS_2 2014 DRC -5.14118 14.03652 121 125

ITM141741 Mu_A1

2014 DRC

105 125

ITM141742 Mu_A1

2014 DRC

108 125

ITM5122 Mu_A1

1962

59 96

ITM5151 Mu_A1

1972 DRC -4.186555 26.439372 59 158

ITM5152 Mu_A1

1974

65 96

ITM5153 Mu_A1

1975

63 96

ITM960657 Mu_A1

1996 Angola -8.565601 13.669590 79 140

Page | 141


KV was supported by a PhD-grant of the Flemish Interuniversity Council -

University Development Cooperation (Belgium). BdJ & CM were supported by

the European Research Council-INTERRUPTB starting grant (nr.311725).

Funding for this work was provided by the Department of Economy, Science

and Innovation of the Flemish Government, and the Stop Buruli Consortium

supported by the UBS Optimus Foundation. The computational resources

used in this work were provided by the HPC core facility CalcUA and VSC

(Flemish Supercomputer Center), funded by the University of Antwerp, the

Hercules Foundation and the Flemish Government - department EWI. The

funders had no role in study design, data collection and analysis, decision to

publish, or preparation of the manuscript.

We thank Wim Mulders, Krista Fissette, Elie Nduwamahoro, and Cécile

Uwizeye for their excellent technical assistance.

Page | 142

Page | 143

Chapter 6

A Genomic Approach to Resolving Relapse

versus Reinfection among Four Cases of

Buruli Ulcer


Miriam Eddyani , Koen Vandelannoote, Conor J. Meehan, Sabin Bhuju, Jessica L.

Porter, Julia Aguiar, Torsten Seemann, Michael Jarek, Mahavir Singh, Françoise

Portaels, Timothy P. Stinear, Bouke C. de Jong

A Genomic Approach to Resolving Relapse versus Reinfection among Four Cases of

Buruli Ulcer.

PLoS Neglected Tropical Diseases 2015 Nov 30;9(11):e0004158.

Conceived and designed the experiments: ME, FP, BCdJ.

Performed the experiments: ME, SB, JLP

Analyzed the data: ME, KV, CJM, SB.

Contributed reagents/materials/analysis tools: JA, MJ, MS.

Wrote the paper: ME, KV, CJM, TS, TPS, BCdJ.

Page | 144

6.1 Abstract

Background: Increased availability of Next Generation Sequencing (NGS)

techniques allows, for the first time, to distinguish relapses from reinfections

in patients with multiple Buruli ulcer (BU) episodes.

Methodology: We compared the number and location of single nucleotide

polymorphisms (SNPs) identified by genomic screening between four pairs of

Mycobacterium ulcerans isolates collected at the time of first diagnosis and at

recurrence, derived from a collection of almost 5000 well characterized

clinical samples from one BU treatment center in Benin.

Principal Findings: The findings suggest that after surgical treatment - without

antibiotics - the second episodes were due to relapse rather than reinfection.

Since specific antibiotics were introduced for the treatment of BU, the one

patient with a culture available from both disease episodes had M. ulcerans

isolates with a genomic distance of 20 SNPs, suggesting the patient was most

likely reinfected rather than having a relapse.

Conclusions: To our knowledge, this study is the first to study recurrences in

M. ulcerans using NGS, and to identify exogenous reinfection as causing a

recurrence of BU. The occurrence of reinfection highlights the contribution of

ongoing exposure to M. ulcerans to disease recurrence, and has implications

for vaccine development.

6.2 Author summary

We compared the whole genomes of four pairs of Mycobacterium ulcerans

isolates collected at the time of first diagnosis and at recurrence, derived from

a collection of almost 5000 well characterized clinical samples from one BU

treatment center in Benin. Our findings suggest that after surgical treatment -

without antibiotics - the second episodes were due to relapse rather than

reinfection. Since specific antibiotics were introduced for the treatment of BU,

the one patient with a culture available from both disease episodes had M.

ulcerans isolates with a larger genomic distance, suggesting that the patient

was most likely reinfected rather than having a relapse. To our knowledge,

this study is the first to assess recurrences in M. ulcerans using whole

Page | 145

genomes, and to identify exogenous reinfection as causing a recurrence of BU.

The occurrence of reinfection highlights the contribution of ongoing exposure

to M. ulcerans to disease recurrence, and has implications for vaccine

development.

6.3 Introduction

Buruli ulcer (BU) is a neglected necrotizing skin and bone disease caused by

the enigmatic pathogen Mycobacterium ulcerans that occurs in riverine

regions of West and Central Africa.

The clonal nature of M. ulcerans has complicated molecular analyses of the

epidemiology of the pathogen, as genotyping methods with sufficient

resolution have been lacking [93]. Using insertion sequence element single

nucleotide polymorphism (ISE-SNP) typing, a technique in which two relatively

small regions (1,431 and 1,871 bp) of the M. ulcerans genome are screened

for polymorphisms, strains could be distinguished to the level of the river

basin in West Africa [135]. However, such molecular genotyping techniques

lack sufficient resolution to distinguish relapse from reinfection with an

unrelated exogenous strain of M. ulcerans among BU recurrences. The World

Health Organization (WHO) defines a relapse as a recurrence of BU within one

year after termination of antibiotic treatment [229]. A recurrence that

appears after that period is consequently considered a reinfection. The

contribution of relapses versus reinfections to BU recurrences and their

biological basis are to date unknown.

Until the routine use of antibiotics (rifampicin and streptomycin) was

advocated by the WHO in 2004, surgery was the mainstay of BU therapy. After

surgical treatment only, a recurrence rate of 6% was reported in Benin [107,

230]. Higher recurrence rates were reported in Ghana (16%-35%) [231, 232],

Ivory Coast (17%) [233], Uganda (20%) [234], and Australia (32%) [235]. When

specific antibiotics are used very few, if any, recurrences are observed [11,

236, 237].

The introduction of Next Generation Sequencing (NGS) now allows for the

first time to distinguish relapses from reinfections in patients with multiple BU

episodes. Similar studies have been conducted for Mycobacterium

Page | 146

tuberculosis [238, 239] and other monomorphic bacterial infections, such as

Clostridium difficile [240]. In the present study we compared the number and

location of SNPs identified by genomic screening between four pairs of M.

ulcerans isolates collected at the time of first diagnosis and at recurrence,

derived from a collection of almost 5000 well characterized clinical samples

from one BU treatment center in Zagnanado, Benin, between 1989 and 2010.

6.4 Methods

Study design and participants

We defined an episode as a clinical suspicion of BU and a recurrence as the

presence of two episodes separated by at least six months. Four such patients

were identified for this study.

We compared the number of SNP differences separating the paired isolates

and a random selection of 36 isolates from 36 patients living in the same

geographical area and diagnosed with BU in the same time frame (1998-2008)

as the patients with multiple episodes. We also compared genomic

relatedness between six patients with two M. ulcerans cultures isolated from

the same disease episode. This genetic background helped to avoid

misclassifying any second episodes with similar M. ulcerans strains prevalent

in the patient’s environment as relapses rather than reinfections. As such a

total of 58 M. ulcerans isolates obtained from 46 patients was included in this

study (Figure 6.1).

Figure 6.1: Flow chart outlining the M. ulcerans isolates included in this analysis.

Page | 147

Genome sequencing and analysis

DNA was obtained by harvesting the growth of three Löwenstein-Jensen

slants followed by heat inactivation, mechanical disruption, enzymatic

digestion and DNA purification on a Maxwell 16 automated platform, a

technique modified from Käser et al. [241].

Whole genome sequencing of the isolates was performed using an Illumina

HiSeq 2000 DNA sequencer and an Illumina MiSeq DNA sequencer with

Nextera XT or TruSeq (Illumina Inc., San Diego, CA, USA) library preparation

and 2x36bp 2x100bp 2x150bp 2x250bp sequencing paired-end chemistry.

Sequencing statistics are provided in Table 6.S1.

The sequencing data analysis was done using the Nesoni software [159].

Firstly, reads were filtered to remove those containing ambiguous base calls,

any reads <50 nucleotides in length, and reads containing only

homopolymers. All reads were furthermore trimmed removing residual

ligated Nextera of TruSeq adaptors and low quality bases (<Q10) at the 3' end.

Average read lengths after clipping are summarized for all isolates in Table

6.S1. Bowtie2 v2.1.0 [141] was used with default parameters to map clipped

sequence reads sets to the Ghanaian M. ulcerans Agy99 reference genome

(Genbank accession number: CP000325). Due to the unreliability of read

mapping in repetitive regions, all ISE elements (IS2404 and IS2606) were hard

masked in this reference genome. Average read depths after mapping to M.

ulcerans Agy99 are summarized for all isolates in Table 6.S1.

We compared the number and location of SNPs between isolates collected at

baseline and at recurrence. At each of the loci called as a variant in any read

set, Nesoni was used to generate a multi-way summary of consensus allele

calls at the corresponding locus in all other read sets of the investigated panel.

By concatenating all these loci a multiple SNP sequence alignment was

generated containing all 282 variant loci across the Agy99 reference

chromosome sequence. A maximum likelihood (ML) phylogenetic tree was

constructed from this alignment using RAxML v8.0.19 [242] under a GTR

model of evolution (no rate heterogeneity) and with an ascertainment bias

likelihood correction for SNP data. The resulting tree was visualized in Figtree

v1.4.0 [102] with nodes of interest highlighted. A haplotype network was

Page | 148

derived using the median joining algorithm [103] as implemented within

SplitsTree v.4.13.1 [142] with default settings. This network was subsequently

visualized with Hapstar v.0.7 [243].

The open source geographic information system Quantum GIS (QGIS v1.8.0)

[244] was used to generate the illustration of the geographical distribution of

all included M. ulcerans genomes in Figure 6.2. The geographical locations of

the residences of BU patients at the time of their first consultation are shown.

The river layer (Ouémé river and its tributaries) was digitized from declassified

Soviet military topographic maps b31-03, b31-09, and b31-15 (scale 1:200k).

The administrative borders of African countries were rendered from the

Global Administrative Unit Layers data set of FAO [245].

Ethics

This retrospective study on stored isolates was approved by the Institutional

Review Board of the Institute of Tropical Medicine.

Accession Numbers

Read data for the study isolates have been deposited in the NCBI Sequence

Read Archive (SRA) under accession n° PRJNA296792.

6.5 Results

Among the 4951 clinically BU suspected patients who consulted the BU

treatment centre of Gbemotin in Zagnanado in southern Benin between 1989

and 2010, we identified 100 who presented with multiple BU episodes (Figure

6.3). A majority of 93 patients had two disease episodes while 7 had three

episodes. Twenty recurrence patients received streptomycin and/or

rifampicin during their first BU episode. The distribution of patients that had

received (partially) effective antibiotics is shown in the Figure 6.S1. Only for

seven of the 100 recurrence patients were we able to successfully culture

isolates from each of two or three disease episodes, owing to the limited

sensitivity of culture for isolation of M. ulcerans from skin biopsies [136].

These mycobacterial isolates were stored at ≤70°C in Dubos broth enriched

with growth supplement and glycerol. However, paired cultures were found

to be viable for only four of these seven patients. The first two patients of

Page | 149

these four patients each had three paired isolates while the other two

patients each had two (Table 1).

Figure 6.3: Flow chart outlining the patients contributing isolates to this analysis. SM:

streptomycin; RMP: rifampicin.

Table 6.1: Details on the four patients with recurrent BU. PCN: penicillin; GM: gentamycin;

CXA: cloxacillin, SM: streptomycin, RMP: rifampicin.

ID Patient A Patient B Patient C Patient D

Gender M F M M

Age at first consultation 12y 9y 18y 10y

Time between two episodes 9.5 months 7 months 22,5 months 9 months

Episode 1 Location right lower limb left lower limb left upper limb left back/flank

Date of presentation 7-Apr-1999 10-Jul-2000 27-May-2002 1-Mar-2004

Treatment PCN GM CXA GM CXA GM CXA SM RMP

surgery surgery surgery surgery

Cultures stored 1 1 1 1

ITM992421 ITM001249 ITM021902 ITM041698

Episode 2 Location right lower limb left lower limb left lower limb right buttock

Cultures stored 2 2 1 1

ITM000562 ITM010548 ITM041716 ITM051516

ITM000563 ITM010623

SNP difference 1 0 0 20

Page | 150

We assessed the paired M. ulcerans isolates of four patients (letter-coded

from A to D) with two BU episodes each (Table 6.1). Each of these 4 patients

presented for the first time at the BU treatment centre of Gbemotin between

1999 and 2004. Patients A & B had lesions at the same location during both

episodes, while patients C & D had lesions at another body site. The time

between the diagnoses of both episodes ranged from seven months to almost

two years. All patients underwent surgery, while patient D also received

specific antibiotics, which were introduced for the routine management of BU

in 2004 [246]. Comparative genomic analysis identified two patients (B & C)

with no detectable genetic differences (0 SNPs) between isolates originating

from two disease episodes. Patient A had one SNP difference between his first

episode isolate and both second episode isolates although at different

positions, suggesting that micro-evolution took place. For patient D however

the isolates of both disease episodes were differentiated from each other by

no less than 20 SNPs, which were found distributed throughout the genome.

This is a genetic distance similar to that observed between different isolates

from the same geographic region as illustrated in the haplotype network

(Figure 6.2B) and the phylogenetic tree (red nodes, Figure 6.S2). The

difference in SNPs between control patients varied from 0 to 53 SNPs. To

exclude cross-contamination of the M. ulcerans culture obtained from the

second episode of patient D, the genome obtained from a strongly positive

biopsy that was treated on the same day as the biopsy of patient D was

sequenced as well and found to differ by 21 SNPs.

There were 2 clusters of 5 and 4 patients having identical M. ulcerans

genomes, who lived in an area of respectively 70 km² (4.7-27 km between

villages) and 14 km² (3-18 km between villages). Four patients with multiple

M. ulcerans isolates from the same BU episode had identical paired genomes

while one such patient differed in 1 SNP and another one in 6 SNPs between

the paired genomes with respectively 8 and 7 days between the times of

sampling of the biopsies from which the M. ulcerans cultures were isolated

(Figure 6.2B).

Among the total of 24 substitutions that were identified in patient A and

patient D, four occurred in intergenic regions, while 20 were found in coding

sequences resulting in 3 synonymous changes (i.e. ’silent’ changes) and 17

Page | 151

non-synonymous changes (i.e. resulting in a change in amino acid), in genes

encoding proteins with various functions (Table 6.S2).

Page | 152

Figure 6.2: (A): Geographical origin of the 58 isolates included in the study. Patients of special

interest are colored accordingly. The size of the dots is proportional to the number of patients

originating from a specific village varying from 1 to 3. The Ouémé river and its tributaries, and

the BU treatment center of Zagnanado are shown as well. (B): Median-Joining network

showing patterns of descent among the 58 studied M. ulcerans isolates. Each circle represents

a unique genotype, and the size of the circle is proportional to the number of individuals

sharing that type. Every edge represents a single mutational step (or SNP) separating the

sampled and hypothetical genotypes. Genotypes obtained from patients of special interest

are colored accordingly.

Page | 153

6.6 Discussion

This study is the first to study recurrences in M. ulcerans using NGS, and to

identify exogenous reinfection as causing a recurrence of BU. In relapses,

paired isolates are genetically more related to each other than to isolates

from other patients living in the same region and infected during the same

time-period. In reinfections on the other hand, paired isolates are potentially

not more related to each other than to isolates from other patients living in

the same region and infected during the same time-period.

Our results suggest that the second BU episode of patients A, B and C was

most likely due to relapses. The second BU episode of the 4th patient was

probably due to a reinfection. This patient is also the only one who had

received specific antibiotics during his first BU episode. We can however not

be entirely certain that patients A, B and C were not infected a second time by

an identical M. ulcerans strain, as we identified other genetically identical

clusters among patients living in the same area and time period. As the mode

of transmission of M. ulcerans remains enigmatic, detailed investigation of

these genetic clusters may provide leads to a common point source of

exposure.

To our knowledge, this study is the first using NGS to assess recurrences in M.

ulcerans, which is important to understand BU epidemiology. In BU, when

disease re-occurs within one year after the end of treatment it is assumed to

be a relapse which is considered the primary endpoint in several studies [11,

236, 237] and in patient management. This study for the first time indicates

that exogenous reinfection plays a role in recurrence of BU. However, the

restricted phylogeny presented here based on four patients should be

interpreted with some circumspection because of the limited sample size,

which is due to the overall low recurrence rate at this treatment center,

combined with the low sensitivity of in vitro culture of M. ulcerans.

In other mycobacterial infections, most notably infection with M. tuberculosis,

the number of different SNPs between relapse and reinfection pairs can be

large, with a clear distinction between pairs with a small difference (≤6 SNPs),

classified as probably relapses, and those with a large difference (≥1306

SNPs), deemed to be reinfections [238]. Epidemiologically linked M.

Page | 154

tuberculosis populations have been reported with a mean SNP difference of

3.4 demonstrating high genomic stability [247]. Walker and colleagues [248]

reported that in patients with an epidemiological link the divergence between

their M. tuberculosis genomes generally does not exceed 14 SNPs, with most

patients having fewer than 5 SNP differences during one disease episode.

During recurrences of Clostridium difficile infection, paired isolates ≤2 SNPs

apart were considered relapses while paired isolates >10 SNPs apart were

considered reinfections [240]. In the present study we detected only one

genomic difference between the relapse pairs, reaffirming the high genomic

stability previously reported for M. ulcerans [249].

Apparent reinfections could theoretically result from differential sampling of

an initially mixed infection. Patient D could possibly have had a simultaneous

infection by two M. ulcerans strains although the results from the 6 patients

with multiple isolates from one BU episode suggest that different M. ulcerans

strains causing a single disease episode is, at best, an infrequent occurrence.

The same was observed among four Cameroonian BU patients with multiple

isolates from one BU episode [250]. This suggests mixed infection would have

been an unlikely explanation for the genetic differences between the paired

isolates of patient D.

The occurrence of reinfection highlights the contribution of ongoing exposure

to M. ulcerans to disease recurrence. Since the delay between relapse (7, 9.5,

and 22.5 months) and reinfection (9 months) episodes overlap, the use of NGS

on cultured isolates is required to distinguish these two scenarios. BU

recurrences within a period of one year after antibiotic treatment that are

considered relapses by WHO [246] may therefore also be reinfections.

Since BU transmission is probably not human-to-human and the mode of

transmission from the environment is not yet clarified, epidemiological links in

support of transmission routes are speculative at best. However, we expect to

be able to determine a SNP-threshold which can be interpreted as an

epidemiological link consistent with a common source of infection in an

ongoing study with a greater number of M. ulcerans genomes from the

Ouémé river valley. The definition of transmission clusters could help to

unravel the enigmatic transmission routes of M. ulcerans.

Page | 155

NGS of paired M. ulcerans strains collected from patients with multiple

episodes of BU has sufficient resolution to distinguish relapse from

reinfection. Our results on a small number of patients suggest that after

surgical treatment without antibiotics the second episodes were due to

relapse rather than reinfection. Since specific antibiotics were introduced for

the treatment of BU, the only patient with a culture available from both

disease episodes had M. ulcerans isolates with a greater genetic distance,

suggesting this patient was most likely reinfected.


Table 6.S1: Sequencing statistics of the study isolates.

� Object too large to print. Available online (http://tinyurl.com/zpwrfdh).

Table 6.S2: Description of substitutions in patients A and D.

� Object too large to print. Available online (http://tinyurl.com/zpwrfdh).

Figure 6.S1: Proportion of clinical suspects that had either not received any specific antibiotics

or (partially) effective antibiotics (rifampicin, streptomycin, amikacin or clarithromycin) during

their first disease episode stratified by delay between episodes (6 months: cutoff in this study;

12 months: WHO cutoff).

Page | 156

Figure 6.S2: Unrooted phylogenetic SNP tree: RAxML was used to build a maximum likelihood

tree from the genomic SNP data with 282 variable positions among 58 genomes. Nodes of

interest are colored according to subject. The scale indicates expected number of

substitutions per site.

Page | 157

Chapter 7

General Discussion

Page | 158

Deciphering the structure of pathogenic bacterial populations is instrumental

for understanding the evolutionary history, spread, and epidemiology of

bacterial infectious diseases. Additionally, a better understanding of disease

transmission and the disease dynamics can have a direct impact on the

development of effective control strategies against the spread of the disease.

This is why, in this PhD thesis, we have used Bayesian phylogenetics and

phylogeography as key methodologies to understand M. ulcerans evolution

and spread using genomic sequence data. Major advances in these fields were

made over the last decade through the development of new statistical models

and rapid advances in low-cost next-generation sequencing technologies.

Whereas 15 years ago it would have taken hundreds of individuals months of

work and millions of Euro to obtain a single bacterial genome sequence,

today, the same sequence can be obtained by a single individual in a couple of

weeks for a few hundred Euro. This made it possible in this PhD thesis, to

perform the first large scale comparative genomic studies on African M.

ulcerans.

The genetic diversity of African M. ulcerans proved to be very restricted

because of the pathogen’s slow genome-wide substitution rate coupled with

its relatively recent origin. As no evidence for recombination was identified

using various comprehensive genome-wide analyses, African M. ulcerans was

found to be evolving entirely through clonal expansion. As a result, the clonal

population structure of African M. ulcerans evolves exclusively by the

accumulation of vertically inherited mutations. Furthermore, the virulence

plasmid pMUM001 was found to be entirely co-evolving with the bacterial

chromosome.

Throughout the work presented in this thesis we repeatedly observed how

genotypes of M. ulcerans show strong spatial segregation at higher

geographical scales, confirming and extending previous data showing

geographical subdivisions [19, 112, 135, 154, 155]. These repeated findings

indicate that when M. ulcerans is introduced in an area it remains localized

and isolated for a sufficient amount of time to allow new mutations to

accumulate in the mycobacterial population of that area. As a direct result a

local genotype associated with the region evolves. This clustering of cases

within endemic foci reflects a common source of infection within foci. We

Page | 159

observed that clonal complexes associated with particular disease foci usually

have been pervasive in these foci for a considerable amount of time; in the

case of the Songololo Territory hotspot around 150 years. This indicates that

an M. ulcerans outbreak spreads relatively slowly, which consequentially also

means that M. ulcerans reservoirs and/or (potential) vectors should also

remain relatively localized and isolated for extended amounts of time within a

disease focus.

On the other hand, we also observed that M. ulcerans genotype clustering

breaks down at lower geographical scale as, at the village level, genotypes

started to intermingle both in Ghana and DRC. Furthermore, in both studies,

various BU patients were identified who had been infected by identical M.

ulcerans genotypes but lived in geographically separated villages and

developed BU in different years. In the Ghanaian study, no obvious

epidemiological links (e.g. travel histories) were revealed in interviews with

patients who had been infected with identical genotypes. We believe this

breakup of the focal distribution of genotypes can be explained by M.

ulcerans’ low substitution rate: insufficient time has elapsed to allow for the

accumulation of discriminatory mutations in the genomes. This complicates

the elucidation of chains of transmission of BU in both DRC and Ghana.

Differential genome reduction analysis and SNP-based exploration of the

genetic population structure revealed the existence of two specific lineages

within the African continent. Phylogeographic analysis indicated that

dominant lineage Mu_A1 has been endemic in the African continent for

hundreds of years, during which time it established itself in different endemic

foci in West and Central Africa. Conversely, lineage Mu_A2 was found to be

less widespread and is rarer, and introduced into Africa much more recently.

We used temporal associations and studied the past demographic history of

M. ulcerans in a BU endemic region to implicate the role of humans as a major

reservoir in BU transmission. African lineage Mu_A2 was found to be

introduced during the period of Neo-imperialism. Additionally, detailed

phylogeographic analysis revealed that introduction of BU in three well

sampled disease foci coincided with the instigation of colonial rule. These

regions were already inhabited long before the arrival of the European

Page | 160

colonizers, when inhabitants were constantly exposed to the lentic

environments in their search for natural resources. However, it was only after

the start of colonial rule that the M. ulcerans clones associated with the

present disease foci were introduced, probably through the arrival of

displaced BU infected human hosts. Furthermore, we identified a relationship

between the past mycobacterial population dynamics of M. ulcerans from the

Songololo Territory and the timing of health policy changes in managing the

BU epidemic in that region. The inverse association we identified over time

between BU control activities and the size of the human M. ulcerans reservoir,

while no proof of causality, nevertheless suggests an impact. Conversely, the

human reservoir appears to be important to sustain new infections. These

combined observations suggest that humans with actively infected, open,

discharging BU lesions can indirectly cause new infections.

In Africa, humans are currently the only known reservoir of M. ulcerans as M.

ulcerans infections in domesticated and wild animals have never been

reported in the African tropics [36, 65, 66]. In all likelihood, humans are not

directly infecting other humans as human-to-human transmission of M.

ulcerans is extremely rare [189]. Humans are nevertheless in all probability an

important reservoir as BU patients with active, openly discharging lesions

contaminate the environment during water contact activities by shedding

concentrated clumps of mycobacteria. Transmission can occur indirectly in the

same community water source, when the superficial skin surface of a naïve

individual is contaminated, and the bacilli present on the contaminated skin

are subsequently inoculated subcutaneously through some form of

penetrating (micro)trauma. This trauma may be as slight as a hypodermic

injection or as serious as a gunshot or a landmine wound [32]. Alternatively,

the inoculation of bacteria can occur through deep bites of transiently

contaminated insects [32], or even scorpions [32] and snakes [251].

Experiments with a Guinea pig infection model showed that passive

application of M. ulcerans onto a preexisting superficial abrasion was

insufficient as the mycobacteria needed to be inoculated at least

intradermally to establish infection [190]. Interestingly, M. marinum, the

closest relative of M. ulcerans, occasionally also causes human infection (fish

Page | 161

tank granulomas) by inoculation through small skin lesions which are often

not remembered by the patient because of the long incubation period [252].

Our observations on the role of humans as potential maintenance reservoir to

sustain new BU infections suggests that interventions in a region aimed at

reducing the human BU burden will at the same time break the transmission

chains within that region. Active case-finding programs and the early

treatment of pre-ulcerative infections with specific antibiotics will decrease

the amounts of mycobacteria shed into the environment and may as a result

reduce disease transmission. Our findings are also supported by the observed

decline of BU incidence recorded in some areas which profited from both

improved BU surveillance and early treatment [197].

However, our findings do not allow us to definitely determine whether

humans are a spillover or a maintenance host species for M. ulcerans [253]. In

a maintenance host an infection can persist through intra-species

transmission alone while in a spillover host, infection will not persist

indefinitely unless there is occasional reinfection from another species.

Nevertheless, our analyses of the demographic history of M. ulcerans from

the Songololo Territory imply that even if other environmental reservoirs

exist, the number of BU notifications can decrease when only human cases

are treated.

In South Eastern Australia (Victoria), BU is considered a zoonosis transmitted

from possums to humans, potentially via a mosquito vector, although to date

definite proof is still lacking [61]. Along these lines, Fyfe et al. [61] drew an

interesting parallel between the epidemiology of BU in Victoria and

leptospirosis, the most commonly identified bacterial zoonosis worldwide

[254]. Similar to BU, leptospirosis is not known to spread directly between

humans. The disease is transmitted through contact with water, food, or soil

contaminated with urine from animals (e.g. rodents) infected with bacteria of

the genus Leptospira. These bacteria can enter the human body through skin

or mucous membranes, especially if the skin is cut or scratched.

Unlike temperate Victoria [24], in tropical North Eastern Australia

(Queensland) M. ulcerans infection has never been identified in wild,

domesticated, nor zoo animals. To date, there are no (potential) vectors

Page | 162

and/or non-human reservoir species known in the region. The epidemiology

of BU in Queensland resembles BU in the African continent with a static

endemic area (Daintree River catchment) that has seen little change since the

1950’s [255].

In Australia, the role of humans in indirectly establishing new BU infections is

probably negligible, as Australians predominantly seek treatment in the early

pre-ulcerative onset of the disease. As a result, patients cannot contaminate

aquatic environments. Additionally, Australians are not dependent on

communal water sources for day-to-day activities like bathing, washing, and

cooking, which furthermore decreases indirect transmission. However, the

close genetic relationship between M. ulcerans from Africa and Australia

suggests that findings from studies of BU transmission in Australia may find

corollaries in African BU endemic settings. For example, more mammal

reservoirs might still remain to be identified that, in a scenario where all BU

cases have been treated in a region, could still spill over the infection to the

human population. The current absence of evidence of a non-human reservoir

of M. ulcerans in Africa is not the evidence of its absence.

A number of comprehensive environmental sampling surveys show that M.

ulcerans DNA can be present in African aquatic environments. However, in

these surveys, the frequency of positive sample occurrence is usually very

low, and furthermore, the bacillary concentration in PCR positive

environmental samples is nearly always insignificantly low [36, 43, 44].

Additionally, actual confirmation of the presence of living M. ulcerans in PCR

positive specimens is extremely difficult to achieve [46]. All of these

limitations combined clearly indicate how difficult it is to detect evidence of

viable M. ulcerans in aquatic environments. These difficulties have raised the

suspicion that M. ulcerans might not have a maintenance reservoir in African

aquatic environments [253]. Nevertheless, our findings in this PhD thesis on

the role of humans in indirectly establishing new BU infections indicate that

future detailed, controlled environmental sampling surveys in Africa should

move away from the commonly applied “fishing expedition” approach and

instead, have an increased focus on community water sources where BU

patients are known to have water contact activities like bathing or washing

[256]. By determining the relative concentrations of M. ulcerans DNA among

Page | 163

the different sample types collected around these community water sources,

researches might find a gradient of M. ulcerans DNA that could lead to the

discovery of the first, non-human reservoir of M. ulcerans in Africa. This

approach in environmental surveys has been successfully applied in Australia

to establish possum species as reservoirs of M. ulcerans [257].

In this PhD thesis, we documented a case of a patient who suffered two

different BU episodes, separated by a period of nine months, caused by

reinfection with an unrelated exogenous strain. Paired isolates originating

from the two episodes of this particular patient were not more related to

each other than two isolates from other patients living in the same region and

infected during the same time period. The occurrence of this reinfection case

clearly highlighted the contribution of ongoing exposure to M. ulcerans. On

the other hand, in three other studied recurrence cases, the second BU

episode was determined to be most likely due to a relapse, as paired isolates

from the different episodes of the three patients were more related to each

other than to isolates from other patients living in the same region, and

infected during the same time-period. Interestingly, all three relapse patients

were treated during a time period (<2004) when surgery was the mainstay of

BU therapy and the use of specific antibiotics (rifampicin and streptomycin)

was not yet advocated by the WHO.

Until today, ISE-SNP genotyping offers the highest geographical resolution of

genotyping achieved, save for methods that rely on WGS. In this thesis, this

method allowed us to gain some fundamental insights into the population

structure and evolutionary history of M. ulcerans that were confirmed in

subsequent genomic work. Furthermore, ISE-SNP typing led to the discovery

of “rare” lineage Mu_A2. In the ISE-SNP technique, two relatively small (<1,9

kb) polymorphic regions of the M. ulcerans genome are screened for

polymorphisms. However, WGS queries nearly the entire genome for

polymorphisms, and consequently, its approach offers vastly greater

geographical genotyping resolution while minimizing phylogenetic discovery

bias [146] to an absolute minimum. As a result, in this thesis, WGS offered the

possibility to comprehensibly study the spread and the evolutionary history of

M. ulcerans at the lower geographical “village” level. In the face of the vastly

greater resolution offered by WGS, a future for ISE-SNP typing can still be

Page | 164

envisaged in its use as an easy, reliable, fast, low tech means to allow

epidemiological tracking of M. ulcerans on a continental scale. The technique

could as such still be applied to trace cases reported after international travel

[9, 10]. An additional advantage of the technique is that, in the case an isolate

failed to grow in culture, the method can be applied directly on clinical

specimens (like swabs or a biopsies) with even very modest mycobacterial

loads. ISE-SNP typing relies solely on widely available conventional PCR

methodology and classical Sanger sequencing and hence it does not require

access to state-of-the-art sequencing equipment and know-how to perform

complex computational data analysis. However, in the future, it will become

ever harder to justify the consumable cost of ISE-SNP typing (+/-30€) against

that of WGS (+/- 70€) in the face of the great difference in resolution.

Some important caveats need to be considered for the various performed

phylogeographic analyses, in particular pertaining to sampling and weak

temporal signals:

A first major limitation we encountered in this PhD thesis was that we studied

a sample of the population from which we then tried to infer information

about the entire population through various statistical methods. The isolate

panels used in this thesis are, to our knowledge, the most comprehensive

ones studied so far in the literature. Even though all well-documented BU

endemic countries were well represented, we were unable to acquire isolates

from several African countries (Equatorial Guinea, Kenya, Liberia, Sierra

Leone, and South Sudan) that have reported (albeit a limited number of) BU

cases in the past [5]. As a result, the spatial coverage of disease isolates is

moderately restricted to specific geographical areas, which might have

restricted our interpretation of the historic spread of African M. ulcerans.

A second important caveat is related to the weak temporal signal. M. ulcerans

molecular sequences sampled through space and time were analyzed in this

thesis using statistical inference approaches. We used a simple linear

regression of the root-to-tip distances of a ML-tree as a function of sampling

times of a heterochronous isolate panel to give an exploratory insight into the

temporal signal present in our panels [219]. The difficulties we encountered

while identifying a linear regression indicated the temporal signal in our data

Page | 165

was rather weak. Nonetheless, in such a situation, relaxed molecular clocks

can perform reasonably well, even though additional attention is required to

avoid that the increased impact of the tree prior in the analysis leads to

incorrect temporal estimates. As a result, we conducted extensive resampling

and randomization experiments, which confirmed that the tip dates

associated with isolates were informative.

Several alternative approaches to date bacterial phylogenies exist that might

be considered to confirm the genome-wide substitution rate identified here.

Ideally, molecular clocks are calibrated using temporal information from

ancient DNA sequences. In a particularly noteworthy Bayesian dating analysis,

Bos et al. [198] used radiocarbon dates obtained from 1000 year old Peruvian

human skeletons as tip-date calibrations for the M. tuberculosis genomes that

were reconstructed from the studied archeological remains. Alternatively,

bacterial phylogenies of M. tuberculosis have been calibrated with temporal

information from human mitochondrial genome data which to a certain

extent overlap in tree topology as tuberculosis is spread from human to

human [184]. As to date there are no known archaeological remains of BU

infected humans, and as BU is a non-communicable disease, yet other

approaches could be considered to confirm the identified genome-wide

substitution rate. Alternative “short-term” mutation rates can be estimated

from molecular epidemiological data or animal infection models. Walker et al.

[199] studied within-host longitudinal diversity between paired isolates of 30

tuberculosis patients to estimate the genome wide substitution rate of M.

tuberculosis using a coalescent-based ML approach. We envisaged doing a

similar analysis for the BU epidemic of southern Benin (CHAPTER 6) but we

identified insufficient patients that fit the inclusion criteria (three from a total

of 4951). Since the introduction of specific antibiotics in the management of

BU in Africa (2004), the amount of patient relapses decreased significantly,

likely impeding similar future investigations. As a result, the only remaining

approach that could still be applied is studying longitudinal diversity between

paired isolates obtained from an animal infection model. Ford et al. [200]

used WGS to compare the accumulation of mutations in M. tuberculosis

isolated from macaques with different tuberculosis disease states. As a result

authors were able to determine the amount of mutations that accumulated

Page | 166

per day of infection. We can envisage a similar experiment using pigs (Sus

scrofa) as experimental infection model for BU as the skin of pigs has striking

structural and physiological similarities with human skin and infection leads to

the development of lesions that closely resemble human BU lesions [258].

Page | 167

Chapter 8

Future Perspectives

Page | 168

The collective efforts of a handful of research teams around the world will

need to continue if we ever are to understand the exact means by which BU is

spreading. Various research lines should be further explored to potentially

shed some more light on this “mystery disease”:

In the future, attempts can be made to obtain M. ulcerans fingerprint patterns

directly from DNA extracts, especially from environmental specimens with

modest to high bacterial loads. Sequenced genomes from an endemic focus

can be used to identify a minimal set of specific “cluster-defining” SNPs. Real-

time PCR based SNP-assays can be developed for this minimal set. The SNP

profiles of environmental specimens can be matched with those of clinical

isolates recovered from patients within the same endemic area in order to

make inferences of potential transmission routes. The development of these

real-time PCR assays, when combined with the ability to rapidly type M.

ulcerans strains, could be essential for understanding the ecology of M.

ulcerans disease.

In this thesis we focused on M. ulcerans from the continent that is worst

affected by BU: Africa. In future work it would be interesting to study the

evolutionary history, phylogeography, and inter-continental spread of M.

ulcerans on a global scale by enriching our genome data further with M.

ulcerans strains from the Americas, Asia, and Oceania.

By virtue of their repetitive nature, IS elements are difficult to infer accurately

using the relatively short reads produced by second generation sequencing

technologies like Illumina sequencing. As a result, in our analyses, 7% of the

bacterial chromosome was ignored. Closing the genome of the new Congolese

reference sequence “SGL03” proved that state-of-the art sequence

technologies like Pacific Biosciences Single Molecule, Real-Time (SMRT)

sequencing can easily span these regions with single reads. This could

significantly increase the geographical genotyping resolution of future

comparative genomics studies even further.

Finally, the ability to interrogate full microbial genomes practically in real time

is changing the face of bacterial population analyses. As next-generation

sequencing technology permeates microbial population genetics further, not

only will new life be infused into long-standing questions like the ones

Page | 169

discussed here, but a more elaborate understanding of microbial diversity and

function will inevitably arise, and from this, entirely new questions will

emerge.

Page | 170

Page | 171

Acknowledgements

Acknowledgements are quite commonplace in academic communication and

virtually obligatory in dissertation writing as they lead to a sense of closure at

the end of a lengthy and demanding research period. The acknowledgements

section is as a result the means of publicly recognizing the role of mentors, the

sacrifices of loved ones, and sharing the relief of completing the PhD process.

However, I feel expressing thanks to others is an entirely personal affair that

shouldn’t necessarily be put in writing for all to read. I believe I gave personal

thanks to all who provided me with love, advice, and guidance throughout the

preparation of my dissertation, and to all who have helped to shape the

accompanying text. Thank you for all your assistance and contributions. I will

forever be indebted to you for being there for me.

Page | 172

Page | 173

List of Publications

� Eddyani M, Vandelannoote K, Meehan CJ, Bhuju S, Porter JL, Aguiar J, Seemann T, Jarek

M, Singh M, Portaels F, Stinear TP, de Jong BC. A Genomic Approach to Resolving Relapse

versus Reinfection among Four Cases of Buruli Ulcer. PLoS Negl Trop Dis. 2015 Nov

30;9(11):e0004158.

� Ablordey AS, Vandelannoote K, Frimpong IA, Ahortor EK, Amissah NA, Eddyani M, Durnez

L, Portaels F, de Jong BC, Leirs H, Porter JL, Mangas KM, Lam MM, Buultjens A, Seemann

T, Tobias NJ, Stinear TP. Whole genome comparisons suggest random distribution of

Mycobacterium ulcerans genotypes in a Buruli ulcer endemic region of Ghana. PLoS Negl

Trop Dis. 2015 Mar 31;9(3):e0003681.

� Amissah NA, Gryseels S, Tobias NJ, Ravadgar B, Suzuki M, Vandelannoote K, Durnez L,

Leirs H, Stinear TP, Portaels F, Ablordey A, Eddyani M. Investigating the role of free-living

amoebae as a reservoir for Mycobacterium ulcerans. PLoS Negl Trop Dis. 2014 Sep

4;8(9):e3148.

� Barletta F, Vandelannoote K, Collantes J, Evans CA, Arévalo J, Rigouts L. Standardization

of a TaqMan-based real-time PCR for the detection of Mycobacterium tuberculosis-

complex in human sputum. Am J Trop Med Hyg. 2014 Oct;91(4):709-14.

� Vandelannoote K, Jordaens K, Bomans P, Leirs H, Durnez L, Affolabi D, Sopoh G, Aguiar J,

Phanzu DM, Kibadi K, Eyangoh S, Manou LB, Phillips RO, Adjei O, Ablordey A, Rigouts L,

Portaels F, Eddyani M, de Jong BC. Insertion sequence element single nucleotide

polymorphism typing provides insights into the population structure and evolution of

Mycobacterium ulcerans across Africa. Appl Environ Microbiol. 2014 Feb;80(3):1197-209.

� Bayonne Manou LS, Portaels F, Eddyani M, Book AU, Vandelannoote K, de Jong BC.

[Mycobacterium ulcerans disease (Buruli ulcer) in Gabon: 2005-2011]. Med Sante Trop.

2013 Oct-Dec;23(4):450-7. (Article in French)

� Gryseels S, Amissah D, Durnez L, Vandelannoote K, Leirs H, De Jonckheere J, Silva MT,

Portaels F, Ablordey A, Eddyani M. Amoebae as potential environmental hosts for

Mycobacterium ulcerans and other mycobacteria, but doubtful actors in Buruli ulcer

epidemiology. PLoS Negl Trop Dis. 2012;6(8):e1764.

� Affolabi D, Sanoussi N, Vandelannoote K, Odoun M, Faïhun F, Sopoh G, Anagonou S,

Portaels F, Eddyani M. Effects of decontamination, DNA extraction, and amplification

procedures on the molecular diagnosis of Mycobacterium ulcerans disease (Buruli ulcer).

J Clin Microbiol. 2012 Apr;50(4):1195-8.

� Vandelannoote K, Durnez L, Amissah D, Gryseels S, Dodoo A, Yeboah S, Addo P, Eddyani

M, Leirs H, Ablordey A, Portaels F. Application of real-time PCR in Ghana, a Buruli ulcer-

endemic country, confirms the presence of Mycobacterium ulcerans in the environment.

FEMS Microbiol Lett. 2010 Mar;304(2):191-4

Page | 174

Page | 175

References

1. Portaels F: Epidemiology of mycobacterial diseases. Clinics in dermatology 1995,

13:207-222.

2. Portaels F, Silva MT, Meyers WM: Buruli ulcer. Clinics in dermatology 2009, 27:291-

305.

3. Mac CP, Tolhurst JC, et al.: A new mycobacterial infection in man. The Journal of

pathology and bacteriology 1948, 60:93-122.

4. Clancey JK, Dodge OG, Lunn HF, Oduori ML: Mycobacterial skin ulcers in Uganda.

The Lancet 1961:951-954.

5. Janssens P, Pattyn S, Meyers W, Portaels F: Buruli Ulcer: An historical overview with

updating. Bull Seances Acad R Sci Outre Mer 2005, 51:265-299.

6. WHO: Buruli ulcer. 2015.

7. Debacker M, Aguiar J, Steunou C, Zinsou C, Meyers WM, Guedenon A, Scott JT,

Dramaix M, Portaels F: Mycobacterium ulcerans disease (Buruli ulcer) in rural

hospital, Southern Benin, 1997-2001. Emerging infectious diseases 2004, 10:1391-

1398.

8. Portaels F, Silva MT, Meyers WM: Buruli ulcer. ClinDermatol 2009, 27:291-305.

9. Evans MR, Mawdsley J, Bull R, Lockwood DN, Thangaraj H, Shanahan D,

Rajakulasingam K: Buruli ulcer in a visitor to London. The British journal of

dermatology 2003, 149:907-909.

10. McGann H, Stragier P, Portaels F, Gascoyne Binzi D, Collyns T, Lucas S, Mawer D:

Buruli ulcer in United Kingdom tourist returning from Latin America. Emerging

infectious diseases 2009, 15:1827-1829.

11. Nienhuis WA, Stienstra Y, Thompson WA, Awuah PC, Abass KM, Tuah W, Awua-

Boateng NY, Ampadu EO, Siegmund V, Schouten JP, et al: Antimicrobial treatment

for early, limited Mycobacterium ulcerans infection: a randomised controlled trial.

Lancet 2010, 375:664-672.

12. Stinear TP, Seemann T, Pidot S, Frigui W, Reysset G, Garnier T, Meurice G, Simon D,

Bouchier C, Ma L, et al: Reductive evolution and niche adaptation inferred from the

genome of Mycobacterium ulcerans, the causative agent of Buruli ulcer. Genome

research 2007, 17:192-200.

13. Stinear TP, Mve-Obiang A, Small PL, Frigui W, Pryor MJ, Brosch R, Jenkin GA, Johnson

PD, Davies JK, Lee RE, et al: Giant plasmid-encoded polyketide synthases produce

the macrolide toxin of Mycobacterium ulcerans. Proceedings of the National

Academy of Sciences of the United States of America 2004, 101:1345-1349.

14. George KM, Chatterjee D, Gunawardana G, Welty D, Hayman J, Lee R, Small PL:

Mycolactone: a polyketide toxin from Mycobacterium ulcerans required for

virulence. Science 1999, 283:854-857.

15. Hall BS, Hill K, McKenna M, Ogbechi J, High S, Willis AE, Simmonds RE: The

pathogenic mechanism of the Mycobacterium ulcerans virulence factor,

mycolactone, depends on blockade of protein translocation into the ER. PLoS

pathogens 2014, 10:e1004061.

16. Guenin-Mace L, Veyron-Churlet R, Thoulouze MI, Romet-Lemonne G, Hong H, Leadlay

PF, Danckaert A, Ruf MT, Mostowy S, Zurzolo C, et al: Mycolactone activation of

Page | 176

Wiskott-Aldrich syndrome proteins underpins Buruli ulcer formation. The journal of

clinical investigation 2013, 123:1501-1512.

17. Marion E, Song OR, Christophe T, Babonneau J, Fenistein D, Eyer J, Letournel F,

Henrion D, Clere N, Paille V, et al: Mycobacterial toxin induces analgesia in buruli

ulcer by targeting the angiotensin pathways. Cell 2014, 157:1565-1576.

18. Yip MJ, Porter JL, Fyfe JA, Lavender CJ, Portaels F, Rhodes M, Kator H, Colorni A,

Jenkin GA, Stinear T: Evolution of Mycobacterium ulcerans and other mycolactone-

producing mycobacteria from a common Mycobacterium marinum progenitor.

Journal of bacteriology 2007, 189:2021-2029.

19. Doig KD, Holt KE, Fyfe JA, Lavender CJ, Eddyani M, Portaels F, Yeboah-Manu D,

Pluschke G, Seemann T, Stinear TP: On the origin of Mycobacterium ulcerans, the

causative agent of Buruli ulcer. BMC genomics 2012, 13:258.

20. Demangel C, Stinear TP, Cole ST: Buruli ulcer: reductive evolution enhances

pathogenicity of Mycobacterium ulcerans. Nature reviews Microbiology 2009, 7:50-

60.

21. Tobias NJ, Doig KD, Medema MH, Chen H, Haring V, Moore R, Seemann T, Stinear TP:

Complete genome sequence of the frog pathogen Mycobacterium ulcerans ecovar

Liflandii. Journal of bacteriology 2013, 195:556-564.

22. Fyfe JA, Lavender CJ, Johnson PD, Globan M, Sievers A, Azuolas J, Stinear TP:

Development and application of two multiplex real-time PCR assays for the

detection of Mycobacterium ulcerans in clinical and environmental samples.

Applied and environmental microbiology 2007, 73:4733-4740.

23. Mahillon J, Chandler M: Insertion sequences. Microbiology and molecular biology

reviews : MMBR 1998, 62:725-774.

24. Portaels F, Chemlal K, Elsen P, Johnson PDR, Hayman JA, Hibble J, Kirkwood R, Meyers

WM: Mycobacterium ulcerans in wild animals. Revue Scientifique et Technique de

l'Office Internationale des Epizooties 2001, 20:252-264.

25. Portaels F, Agular J, Fissette K, Fonteyne PA, De Beenhouwer H, de Rijk P, Guedenon

A, Lemans R, Steunou C, Zinsou C, et al: Direct detection and identification of

Mycobacterium ulcerans in clinical specimens by PCR and oligonucleotide-specific

capture plate hybridization. Journal of clinical microbiology 1997, 35:1097-1100.

26. Bratschi MW, Bolz M, Grize L, Kerber S, Minyem JC, Um Boock A, Yeboah-Manu D, Ruf

MT, Pluschke G: Primary cultivation: factors affecting contamination and

Mycobacterium ulcerans growth after long turnover time of clinical specimens.

BMC Infectious Diseases 2014, 14:636.

27. Eddyani M, Debacker M, Martin A, Aguiar J, Johnson CR, Uwizeye C, Fissette K,

Portaels F: Primary culture of Mycobacterium ulcerans from human tissue

specimens after storage in semisolid transport medium. Journal of clinical

microbiology 2008, 46:69-72.

28. Johnson RC, Makoutode M, Sopoh GE, Elsen P, Gbovi J, Pouteau LH, Meyers WM,

Boko M, Portaels F: Buruli ulcer distribution in Benin. Emerging infectious diseases

2005, 11:500-501.

29. Nackers F, Johnson RC, Glynn JR, Zinsou C, Tonglet R, Portaels F: Environmental and

health-related risk factors for Mycobacterium ulcerans disease (Buruli ulcer) in

Benin. The American journal of tropical medicine and hygiene 2007, 77:834-836.

30. Raghunathan PL, Whitney EA, Asamoa K, Stienstra Y, Taylor TH, Jr., Amofah GK,

Ofori-Adjei D, Dobos K, Guarner J, Martin S, et al: Risk factors for Buruli ulcer disease

(Mycobacterium ulcerans Infection): results from a case-control study in Ghana.

Page | 177

Clinical infectious diseases : an official publication of the Infectious Diseases Society

of America 2005, 40:1445-1453.

31. Jacobsen KH, Padgett JJ: Risk factors for Mycobacterium ulcerans infection.

International journal of infectious diseases : IJID : official publication of the

International Society for Infectious Diseases 2010, 14:e677-681.

32. Meyers WM, Shelly WM, Connor DH, Meyers EK: Human Mycobacterium ulcerans

infections developing at sites of trauma to skin. The American journal of tropical

medicine and hygiene 1974, 23:919-923.

33. Williamson HR, Mosi L, Donnell R, Aqqad M, Merritt RW, Small PLC: Mycobacterium

ulcerans Fails to Infect through Skin Abrasions in a Guinea Pig Infection Model:

Implications for Transmission. PLoS neglected tropical diseases 2014, 8.

34. Ross BC, Johnson PD, Oppedisano F, Marino L, Sievers A, Stinear T, Hayman JA, Veitch

MG, Robins-Browne RM: Detection of Mycobacterium ulcerans in environmental

samples during an outbreak of ulcerative disease. Applied and environmental

microbiology 1997, 63:4135-4138.

35. Portaels F, Elsen P, Guimaraes-Peres A, Fonteyne PA, Meyers WM: Insects in the

transmission of Mycobacterium ulcerans infection. Lancet 1999, 353:986.

36. Vandelannoote K, Durnez L, Amissah D, Gryseels S, Dodoo A, Yeboah S, Addo P,

Eddyani M, Leirs H, Ablordey A, Portaels F: Application of real-time PCR in Ghana, a

Buruli ulcer-endemic country, confirms the presence of Mycobacterium ulcerans in

the environment. FEMS microbiology letters 2010, 304:191-194.

37. Eddyani M, Ofori-Adjei D, Teugels G, De Weirdt D, Boakye D, Meyers WM, Portaels F:

Potential role for fish in transmission of Mycobacterium ulcerans disease (Buruli

ulcer): an environmental study. Applied and environmental microbiology 2004,

70:5679-5681.

38. Marsollier L, Robert R, Aubry J, Saint Andre JP, Kouakou H, Legras P, Manceau AL,

Mahaza C, Carbonnelle B: Aquatic insects as a vector for Mycobacterium ulcerans.

Applied and environmental microbiology 2002, 68:4623-4628.

39. Williamson HR, Benbow ME, Nguyen KD, Beachboard DC, Kimbirauskas RK, McIntosh

MD, Quaye C, Ampadu EO, Boakye D, Merritt RW, Small PL: Distribution of

Mycobacterium ulcerans in buruli ulcer endemic and non-endemic aquatic sites in

Ghana. PLoS neglected tropical diseases 2008, 2:e205.

40. Stinear T, Davies JK, Jenkin GA, Hayman JA, Oppedisano F, Johnson PD: Identification

of Mycobacterium ulcerans in the environment from regions in Southeast Australia

in which it is endemic with sequence capture-PCR. Applied and environmental

microbiology 2000, 66:3206-3213.

41. Kotlowski R, Martin A, Ablordey A, Chemlal K, Fonteyne PA, Portaels F: One-tube cell

lysis and DNA extraction procedure for PCR-based detection of Mycobacterium

ulcerans in aquatic insects, molluscs and fish. Journal of Medical Microbiology 2004,

53:927-933.

42. Johnson PDR, Azuolas J, Lavender CJ, Washart E, Stinear TP, Hayman JA, Brown L,

Jenkin GA, Fyfe JAM: Mycobacterium ulcerans in mosquitoes captured during

outbreak of Buruli ulcer, Southeastern Australia. Emerging infectious diseases 2007,

13 1653-1660.

43. Tobias N, Ammisah N, Ahortor E, Wallace J, Ablordey A, Stinear T: Snapshot fecal

survey of domestic animals in rural Ghana for Mycobacterium ulcerans. PeerJ 2016,

4.

Page | 178

44. Tian RB, Niamke S, Tissot-Dupont H, Drancourt M: Detection of Mycobacterium

ulcerans DNA in the Environment, Ivory Coast. Plos One 2016, 11:e0151567.

45. Barker DJP, Clancey JK, Rao S: Mycobacteria on vegetation in Uganda. East African

Medical Journal 1972, 49:667-671.

46. Portaels F, Meyers WM, Ablordey A, Castro AG, Chemlal K, de Rijk P, Elsen P, Fissette

K, Fraga AG, Lee R, et al: First cultivation and characterization of Mycobacterium

ulcerans from the environment. PLoS neglected tropical diseases 2008, 2:e178.

47. Trott KA, Stacy BA, Lifland BD, Diggs HE, Harland RM, Khokha MK, Grammer TC,

Parker JM: Characterization of a Mycobacterium ulcerans-like infection in a colony

of African tropical clawed frogs (Xenopus tropicalis). Comparative medicine 2004,

54:309-317.

48. Eddyani M, Lavender C, de Rijk WB, Bomans P, Fyfe J, de Jong B, Portaels F:

Multicenter external quality assessment program for PCR detection of

Mycobacterium ulcerans in clinical and environmental specimens. Plos One 2014,

9:e89407.

49. Barnett H: The incrimination of arthropods as vectors of disease. In Proceedings of

the 11th International Congress of Entomology (1962). Edited by Strouhal H, Beier M.

1960: 341–345.

50. Hennigan CE, Myers L, Ferris MJ: Environmental distribution and seasonal

prevalence of Mycobacterium ulcerans in Southern Louisiana. Applied and

environmental microbiology 2013, 79:2648-2656.

51. Meyers WM, Shelly WM, Connor DH: Heat treatment of Mycobacterium ulcerans

infections without surgical excision. The American journal of tropical medicine and

hygiene 1974, 23:924-929.

52. Eddyani M, Portaels F: Survival of Mycobacterium ulcerans at 37 degrees C. Clinical

Microbiology and Infection 2007, 13:1033-1035.

53. Palomino JC, Obiang AM, Realini L, Meyers WM, Portaels F: Effect of oxygen on

growth of Mycobacterium ulcerans in the BACTEC system. Journal of clinical

microbiology 1998, 36:3420-3422.

54. van Ingen J, Boeree MJ, van Soolingen D, Iseman MD, Heifets LB, Daley CL: Are

phylogenetic position, virulence, drug susceptibility and in vivo response to

treatment in mycobacteria interrelated? Infection, genetics and evolution : journal

of molecular epidemiology and evolutionary genetics in infectious diseases 2012,

12:832-837.

55. Asmar S, Sassi M, Phelippeau M, Drancourt M: Inverse correlation between salt

tolerance and host-adaptation in mycobacteria. BMC research notes 2016, 9:249.

56. Monot M, Honore N, Garnier T, Zidane N, Sherafi D, Paniz-Mondolfi A, Matsuoka M,

Taylor GM, Donoghue HD, Bouwman A, et al: Comparative genomic and

phylogeographic analysis of Mycobacterium leprae. Nature genetics 2009, 41:1282-

1289.

57. Parkhill J, Sebaihia M, Preston A, Murphy LD, Thomson N, Harris DE, Holden MT,

Churcher CM, Bentley SD, Mungall KL, et al: Comparative analysis of the genome

sequences of Bordetella pertussis, Bordetella parapertussis and Bordetella

bronchiseptica. Nature genetics 2003, 35:32-40.

58. Parkhill J, Wren BW, Thomson NR, Titball RW, Holden MT, Prentice MB, Sebaihia M,

James KD, Churcher C, Mungall KL, et al: Genome sequence of Yersinia pestis, the

causative agent of plague. Nature 2001, 413:523-527.

Page | 179

59. Francis G, Whitby M, Woods M: Mycobacterium ulcerans infection: a rediscovered

focus in the Capricorn Coast region of central Queensland. The Medical journal of

Australia 2006, 185:179-180.

60. Merritt RW, Walker ED, Small PL, Wallace JR, Johnson PD, Benbow ME, Boakye DA:

Ecology and transmission of Buruli ulcer disease: a systematic review. PLoS

neglected tropical diseases 2010, 4:e911.

61. Fyfe JA, Lavender CJ, Handasyde KA, Legione AR, O'Brien CR, Stinear TP, Pidot SJ,

Seemann T, Benbow ME, Wallace JR, et al: A major role for mammals in the ecology

of Mycobacterium ulcerans. PLoS neglected tropical diseases 2010, 4:e791.

62. O'Brien CR, Handasyde KA, Hibble J, Lavender CJ, Legione AR, McCowan C, Globan M,

Mitchell AT, McCracken HE, Johnson PD, Fyfe JA: Clinical, microbiological and

pathological findings of Mycobacterium ulcerans infection in three Australian

Possum species. PLoS neglected tropical diseases 2014, 8:e2666.

63. Quek TY, Athan E, Henry MJ, Pasco JA, Redden-Hoare J, Hughes A, Johnson PD: Risk

factors for Mycobacterium ulcerans infection, southeastern Australia. Emerging

infectious diseases 2007, 13:1661-1666.

64. Johnson PD, Lavender CJ: Correlation between Buruli ulcer and vector-borne

notifiable diseases, Victoria, Australia. Emerging infectious diseases 2009, 15:614-

615.

65. Durnez L, Stragier P, Roebben K, Ablordey A, Leirs H, Portaels F: A comparison of DNA

extraction procedures for the detection of Mycobacterium ulcerans, the causative

agent of Buruli ulcer, in clinical and environmental specimens. Journal of

microbiological methods 2009, 76:152-158.

66. Narh CA, Mosi L, Quaye C, Dassi C, Konan DO, Tay SC, de Souza DK, Boakye DA,

Bonfoh B: Source tracking Mycobacterium ulcerans infections in the Ashanti region,

Ghana. PLoS neglected tropical diseases 2015, 9:e0003437.

67. Pouillot R, Matias G, Wondje CM, Portaels F, Valin N, Ngos F, Njikap A, Marsollier L,

Fontanet A, Eyangoh S: Risk factors for buruli ulcer: a case control study in

Cameroon. PLoS neglected tropical diseases 2007, 1:e101.

68. Bratschi MW, Bolz M, Minyem JC, Grize L, Wantong FG, Kerber S, Njih Tabah E, Ruf

MT, Mou F, Noumen D, et al: Geographic distribution, age pattern and sites of

lesions in a cohort of buruli ulcer patients from the mape basin of cameroon. PLoS

neglected tropical diseases 2013, 7:e2252.

69. Zogo B, Djenontin A, Carolan K, Babonneau J, Guegan JF, Eyangoh S, Marion E: A

Field Study in Benin to Investigate the Role of Mosquitoes and Other Flying Insects

in the Ecology of Mycobacterium ulcerans. PLoS neglected tropical diseases 2015,

9:e0003941.

70. Kingman JFC: Origins of the coalescent: 1974-1982. Genetics 2000, 156:1461-1463.

71. Pybus OG, Rambaut A: Evolutionary analysis of the dynamics of viral infectious

disease. Nature reviews Genetics 2009, 10:540-550.

72. Drummond AJ, Ho SY, Phillips MJ, Rambaut A: Relaxed phylogenetics and dating

with confidence. PLoS biology 2006, 4:e88.

73. Ho SY, Lanfear R, Bromham L, Phillips MJ, Soubrier J, Rodrigo AG, Cooper A: Time-

dependent rates of molecular evolution. Molecular ecology 2011, 20:3087-3101.

74. Zuckerkandl E, Pauling L: Evolutionary divergence and convergence in proteins.

Evolving genes and proteins 1965, 97:97-166.

75. Lartillot N, Philippe H: Computing Bayes factors using thermodynamic integration.

Systematic biology 2006, 55:195-207.

Page | 180

76. Grenfell BT, Pybus OG, Gog JR, Wood JL, Daly JM, Mumford JA, Holmes EC: Unifying

the epidemiological and evolutionary dynamics of pathogens. Science 2004,

303:327-332.

77. Bouckaert R, Heled J, Kuhnert D, Vaughan T, Wu CH, Xie D, Suchard MA, Rambaut A,

Drummond AJ: BEAST 2: a software platform for Bayesian evolutionary analysis.

PLoS computational biology 2014, 10:e1003537.

78. Achilli A, Olivieri A, Soares P, Lancioni H, Hooshiar Kashani B, Perego UA, Nergadze

SG, Carossa V, Santagostino M, Capomaccio S, et al: Mitochondrial genomes from

modern horses reveal the major haplogroups that underwent domestication.

Proceedings of the National Academy of Sciences of the United States of America

2012, 109:2449-2454.

79. Eppinger M, Baar C, Linz B, Raddatz G, Lanz C, Keller H, Morelli G, Gressmann H,

Achtman M, Schuster SC: Who ate whom? Adaptive Helicobacter genomic changes

that accompanied a host jump from early humans to large felines. PLoS genetics

2006, 2:e120.

80. Jaenike J, Dyer KA: No resistance to male-killing Wolbachia after thousands of years

of infection. Journal of evolutionary biology 2008, 21:1570-1577.

81. Roumagnac P, Weill FX, Dolecek C, Baker S, Brisse S, Chinh NT, Le TA, Acosta CJ,

Farrar J, Dougan G, Achtman M: Evolutionary history of Salmonella typhi. Science

2006, 314:1301-1304.

82. Holt KE, Baker S, Weill FX, Holmes EC, Kitchen A, Yu J, Sangal V, Brown DJ, Coia JE,

Kim DW, et al: Shigella sonnei genome sequencing and phylogenetic analysis

indicate recent global dissemination from Europe. Nature genetics 2012, 44:1056-

1059.

83. Walsh DS, Portaels F, Meyers WM: Buruli ulcer: Advances in understanding

Mycobacterium ulcerans infection. Dermatologic clinics 2011, 29:1-8.

84. Debacker M, Aguiar J, Steunou C, Zinsou C, Meyers WM, Scott JT, Dramaix M,

Portaels F: Mycobacterium ulcerans disease: role of age and gender in incidence

and morbidity. Tropical medicine & international health : TM & IH 2004, 9:1297-

1304.

85. Wagner T, Benbow ME, Burns M, Johnson RC, Merritt RW, Qi J, Small PL: A

Landscape-based model for predicting Mycobacterium ulcerans infection (Buruli

Ulcer disease) presence in Benin, West Africa. EcoHealth 2008, 5:69-79.

86. Gryseels S, Amissah D, Durnez L, Vandelannoote K, Leirs H, De Jonckheere J, Silva MT,

Portaels F, Ablordey A, Eddyani M: Amoebae as potential environmental hosts for

Mycobacterium ulcerans and other mycobacteria, but doubtful actors in Buruli

ulcer epidemiology. PLoS neglected tropical diseases 2012, 6:e1764.

87. Merritt RW, Benbow ME, Small PLC: Unraveling an emerging disease associated

with disturbed aquatic environments: the case of Buruli ulcer. Frontiers in Ecology

and the Environment 2005, 3:323-331.

88. Marsollier L, Brodin P, Jackson M, Kordulakova J, Tafelmeyer P, Carbonnelle E, Aubry

J, Milon G, Legras P, Andre JP, et al: Impact of Mycobacterium ulcerans biofilm on

transmissibility to ecological niches and Buruli ulcer pathogenesis. PLoS pathogens

2007, 3:e62.

89. Durnez L, Suykerbuyk P, Nicolas V, Barriere P, Verheyen E, Johnson CR, Leirs H,

Portaels F: Terrestrial Small Mammals as Reservoirs of Mycobacterium ulcerans in

Benin. Applied and environmental microbiology 2010, 76:4574-4577.

Page | 181

90. Moran NA, Plague GR: Genomic changes following host restriction in bacteria.

Current opinion in genetics & development 2004, 14:627-633.

91. Musser JM: Molecular population genetic analysis of emerged bacterial pathogens:

selected insights. Emerging infectious diseases 1996, 2:1-17.

92. Spratt BG, Maiden MC: Bacterial population genetics, evolution and epidemiology.

Philosophical transactions of the Royal Society of London Series B, Biological sciences

1999, 354:701-710.

93. Roltgen K, Stinear TP, Pluschke G: The genome, evolution and diversity of

Mycobacterium ulcerans. Infection, genetics and evolution : journal of molecular

epidemiology and evolutionary genetics in infectious diseases 2012, 12:522-529.

94. Kaser M, Hauser J, Pluschke G: Single nucleotide polymorphisms on the road to

strain differentiation in Mycobacterium ulcerans. Journal of clinical microbiology

2009, 47:3647-3652.

95. Rondini S, Kaser M, Stinear T, Tessier M, Mangold C, Dernick G, Naegeli M, Portaels F,

Certa U, Pluschke G: Ongoing genome reduction in Mycobacterium ulcerans.

Emerging infectious diseases 2007, 13:1008-1015.

96. Guimaraes-Peres A, Portaels F, de Rijk P, Fissette K, Pattyn SR, van Vooren J, Fonteyne

P: Comparison of two PCRs for detection of Mycobacterium ulcerans. Journal of

clinical microbiology 1999, 37:206-208.

97. Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, McWilliam H,

Valentin F, Wallace IM, Wilm A, Lopez R, et al: Clustal W and Clustal X version 2.0.

Bioinformatics 2007, 23:2947-2948.

98. Saitou N, Nei M: The neighbor-joining method: a new method for reconstructing

phylogenetic trees. Molecular biology and evolution 1987, 4:406-425.

99. Tamura K, Dudley J, Nei M, Kumar S: MEGA4: Molecular Evolutionary Genetics

Analysis (MEGA) software version 4.0. Molecular biology and evolution 2007,

24:1596-1599.

100. Felsenstein J: Confidence-Limits on Phylogenies - an Approach Using the Bootstrap.

Evolution 1985, 39:783-791.

101. Posada D: jModelTest: phylogenetic model averaging. Molecular biology and

evolution 2008, 25:1253-1256.

102. Rambaut A: FigTree v.1.4.2, Available from

http://tree.bio.ed.ac.uk/software/figtree/. 2015.

103. Bandelt HJ, Forster P, Rohl A: Median-joining networks for inferring intraspecific

phylogenies. Molecular biology and evolution 1999, 16:37-48.

104. Quantum_GIS: Quantum GIS Geographic Information System Open Source

Geospatial Foundation Project.; 2012.

105. Jenness J, Dooley J, Aguilar-Manjarrez J, Riva C: African Water Resource Database.

GIS-based tools for inland aquatic resource management. Rome: FAO; 2007.

106. R_Core_Team: R: A Language and Environment for Statistical Computing. 2012.

107. Sopoh GE, Johnson RC, Chauty A, Dossou AD, Aguiar J, Salmon O, Portaels F, Asiedu K:

Buruli ulcer surveillance, Benin, 2003-2005. Emerging infectious diseases 2007,

13:1374-1376.

108. Johnson RC, Sopoh GE, Boko M, Zinsou C, Gbovi J, Makoutode M, Portaels F:

[Distribution of Mycobacterium ulcerans (Buruli ulcer) in the district of Lalo in

Benin]. Tropical medicine & international health : TM & IH 2005, 10:863-871.

Page | 182

109. Kaser M, Rondini S, Naegeli M, Stinear T, Portaels F, Certa U, Pluschke G: Evolution of

two distinct phylogenetic lineages of the emerging human pathogen

Mycobacterium ulcerans. BMC evolutionary biology 2007, 7:177.

110. Achtman M: Evolution, population structure, and phylogeography of genetically

monomorphic bacterial pathogens. Annual review of microbiology 2008, 62:53-70.

111. Morelli G, Didelot X, Kusecek B, Schwarz S, Bahlawane C, Falush D, Suerbaum S,

Achtman M: Microevolution of Helicobacter pylori during prolonged infection of

single hosts and within families. PLoS genetics 2010, 6:e1001036.

112. Roltgen K, Qi W, Ruf MT, Mensah-Quainoo E, Pidot SJ, Seemann T, Stinear TP, Kaser

M, Yeboah-Manu D, Pluschke G: Single nucleotide polymorphism typing of

Mycobacterium ulcerans reveals focal transmission of buruli ulcer in a highly

endemic region of Ghana. PLoS neglected tropical diseases 2010, 4:e751.

113. Marion E, Landier J, Boisier P, Marsollier L, Fontanet A, Le Gall P, Aubry J, Djeunga N,

Umboock A, Eyangoh S: Geographic expansion of Buruli ulcer disease, Cameroon.

Emerging infectious diseases 2011, 17:551-553.

114. Stinear T, Davies JK, Jenkin GA, Portaels F, Ross BC, Oppedisano F, Purcell M, Hayman

JA, Johnson PD: A simple PCR method for rapid genotype analysis of Mycobacterium

ulcerans. Journal of clinical microbiology 2000, 38:1482-1487.

115. Qi W, Kaser M, Roltgen K, Yeboah-Manu D, Pluschke G: Genomic diversity and

evolution of Mycobacterium ulcerans revealed by next-generation sequencing.

PLoS pathogens 2009, 5:e1000580.

116. Pearson T, Okinaka RT, Foster JT, Keim P: Phylogenetic understanding of clonal

populations in an era of whole genome sequencing. Infection, genetics and

evolution : journal of molecular epidemiology and evolutionary genetics in infectious

diseases 2009, 9:1010-1019.

117. WHO: Buruli ulcer. 2014.

118. unknown: Epidemiology of Mycobacterium ulcerans infection (Buruli ulcer) at

Kinyara, Uganda. Transactions of the Royal Society of Tropical Medecine and

Hygiene 1971, 65:763-775.

119. Barker DJ: Epidemiology of Mycobacterium ulcerans infection. Transactions of the

Royal Society of Tropical Medicine and Hygiene 1973, 67:43-50.

120. Meyers WM, Connor DH, McCullough B, Bourland J, Moris R, Proos L: Distribution of

Mycobacterium ulcerans infections in Zaire, including the report of new foci.

Annales de la Societe belge de medecine tropicale 1974, 54:147-157.

121. van der Werf TS, van der Graaf WT, Groothuis DG, Knell AJ: Mycobacterium ulcerans

infection in Ashanti region, Ghana. Transactions of the Royal Society of Tropical

Medicine and Hygiene 1989, 83:410-413.

122. Amofah GK, Sagoe-Moses C, Adjei-Acquah C, Frimpong EH: Epidemiology of Buruli

ulcer in Amansie West district, Ghana. Transactions of the Royal Society of Tropical

Medecine and Hygiene 1993, 87:644-645.

123. Marston BJ, Diallo MO, Horsburgh CR, Jr., Diomande I, Saki MZ, Kanga JM, Patrice G,

Lipman HB, Ostroff SM, Good RC: Emergence of Buruli ulcer disease in the Daloa

region of Cote d'Ivoire. The American journal of tropical medicine and hygiene 1995,

52:219-224.

124. Veitch MGK, Johnson PDR, Flood PE, Leslie DE, Street AC, Hayman JA: A large

localized outbreak of Mycobacterium ulcerans infection on a temperate southern

Australian island. Epidemiology and Infection 1997, 119:313-318.

Page | 183

125. Aiga H, Amano T, Cairncross S, Adomako J, Nanas OK, Coleman S: Assessing water-

related risk factors for Buruli ulcer: a case-control study in Ghana. The American

journal of tropical medicine and hygiene 2004, 71:387-392.

126. Johnson RC, Makoutod‚ M, Sopoh GE, Elsen P, Gbovi J, Pouteau LH, Meyers WM, Boko

M, Portaels F: Buruli ulcer distribution in Benin. Emerging infectious diseases 2005,

11:500-501.

127. Debacker M, Portaels F, Aguiar J, Steunou C, Zinsou C, Meyers W, Dramaix M: Risk

factors for Buruli ulcer, Benin. Emerging infectious diseases 2006, 12:1325-1331.

128. Wagner T, Benbow ME, Burns M, Johnson RC, Merritt RW, Qi J, Small PL: A

Landscape-based model for predicting Mycobacterium ulcerans infection (Buruli

ulcer disease) presence in Benin, West Africa. Ecohealth 2008, 5:69-79.

129. Porten K, Sailor K, Comte E, Njikap A, Sobry A, Sihom F, Meva'a A, Eyangoh S, Myatt

M, Nackers F, Grais RF: Prevalence of Buruli ulcer in Akonolinga health district,

Cameroon: results of a cross sectional survey. PLoSNeglTropDis 2009, 3:e466.

130. van Ravensway J, Benbow ME, Tsonis AA, Pierce SJ, Campbell LP, Fyfe JA, Hayman JA,

Johnson PD, Wallace JR, Qi J: Climate and landscape factors associated with Buruli

ulcer incidence in Victoria, Australia. PLoS One 2012, 7:e51074.

131. Huang GK, Johnson PD: Epidemiology and management of Buruli ulcer. Expert

review of anti-infective therapy 2014, 12:855-865.

132. Landier J, Boisier P, Fotso Piam F, Noumen-Djeunga B, Sime J, Wantong FG,

Marsollier L, Fontanet A, Eyangoh S: Adequate wound care and use of bed nets as

protective factors against Buruli Ulcer: results from a case control study in

Cameroon. PLoS neglected tropical diseases 2011, 5:e1392.

133. Roltgen K, Assan-Ampah K, Danso E, Yeboah-Manu D, Pluschke G: Development of a

temperature-switch PCR-based SNP typing method for Mycobacterium ulcerans.

PLoS neglected tropical diseases 2012, 6:e1904.

134. Ampadu E: Annual Report of the National Buruli ulcer Control Programme

Department of Health, Accra, Ghana: Government of Ghana. 2008.

135. Vandelannoote K, Jordaens K, Bomans P, Leirs H, Durnez L, Affolabi D, Sopoh G,

Aguiar J, Phanzu DM, Kibadi K, et al: Insertion Sequence Element Single Nucleotide

Polymorphism Typing Provides Insights into the Population Structure and Evolution

of Mycobacterium ulcerans across Africa. Applied and environmental microbiology

2014, 80:1197-1209.

136. Eddyani M, Debacker M, Martin A, Aguiar J, Johnson RC, Uwizeye C, Fissette K,

Portaels F: Primary culture of Mycobacterium ulcerans from human tissue

specimens after storage in semisolid transport medium. Journal of clinical

microbiology 2008, 46:69-72.

137. Eddyani M, Fraga AG, Schmitt F, Uwizeye C, Fissette K, Johnson C, Aguiar J, Sopoh G,

Barogui Y, Meyers WM, et al: Fine-needle aspiration, an efficient sampling

technique for bacteriological diagnosis of nonulcerative Buruli ulcer. Journal of


138. Portaels F, De Muynck A, Sylla MP: Selective isolation of mycobacteria from soil: a

statistical analysis approach. Journal of General Microbiology 1988, 134:849-855.

139. Boom R, Sol CJA, Salimans MMM, Jansen CL, Wertheim-van Dillen PME, van der

Noordaa J: Rapid and simple method for purification of nucleic acids. Journal of


Page | 184

140. Ross BC, Marino L, Oppedisano F, Edwards R, Robins-Browne RM, Johnson PD:

Development of a PCR assay for rapid diagnosis of Mycobacterium ulcerans

infection. Journal of clinical microbiology 1997, 35:1696-1700.

141. Langmead B, Salzberg SL: Fast gapped-read alignment with Bowtie 2. Nature

methods 2012, 9:357-359.

142. Huson DH: SplitsTree: analyzing and visualizing evolutionary data. Bioinformatics

1998, 14:68-73.

143. Hilty M, Yeboah-Manu D, Boakye D, Mensah-Quainoo E, Rondini S, Schelling E, Ofori-

Adjei D, Portaels F, Zinsstag J, Pluschke G: Genetic diversity in Mycobacterium

ulcerans isolates from Ghana revealed by a newly identified locus containing a

variable number of tandem repeats. Journal of bacteriology 2006, 188:1462-1465.

144. Oluwasanmi JO, Itayemi SO, Alabi GO: Buruli (mycobacterial) ulcers in Caucasians in

Nigeria. British Journal of Plastic Surgery 1975, 28:111-113.

145. Trubiano JA, Lavender CJ, Fyfe JA, Bittmann S, Johnson PD: The incubation period of

Buruli ulcer (Mycobacterium ulcerans infection). PLoS neglected tropical diseases

2013, 7:e2463.

146. Pearson T, Busch JD, Ravel J, Read TD, Rhoton SD, U'Ren JM, Simonson TS, Kachur

SM, Leadem RR, Cardon ML, et al: Phylogenetic discovery bias in Bacillus anthracis

using single-nucleotide polymorphisms from whole-genome sequencing.

Proceedings of the National Academy of Sciences of the United States of America

2004, 101:13536-13541.

147. Monot M, Honor‚ N, Garnier T, Araoz R, Copp‚e JY, Lacroix C, Sow S, Spencer JS,

Truman RW, Williams DL, et al: On the origin of leprosy. Science 2005, 308:1040-

1042.

148. Truman RW, Singh P, Sharma R, Busso P, Rougemont J, Paniz-Mondolfi A,

Kapopoulou A, Brisse S, Scollard DM, Gillis TP, Cole ST: Probable zoonotic leprosy in

the southern United States. The New England Journal of Medicine 2011, 364:1626-

1633.

149. Wheat WH, Casali AL, Thomas V, Spencer JS, Lahiri R, Williams DL, McDonnell GE,

Gonzalez-Juarrero M, Brennan PJ, Jackson M: Long-term survival and virulence of

Mycobacterium leprae in amoebal cysts. PLoS neglected tropical diseases 2014,

8:e3405.

150. Boritsch EC, Supply P, Honore N, Seemann T, Stinear TP, Brosch R: A glimpse into the

past and predictions for the future: the molecular evolution of the tuberculosis

agent. Molecular microbiology 2014, 93:835-852.

151. Supply P, Marceau M, Mangenot S, Roche D, Rouanet C, Khanna V, Majlessi L,

Criscuolo A, Tap J, Pawlik A, et al: Genomic analysis of smooth tubercle bacilli

provides insights into ancestry and pathoadaptation of Mycobacterium

tuberculosis. Nature genetics 2013, 45:172-179.

152. Cook A: Mengo hospital notes, 1897, Makerere Medical School Library. BMJ 1970,

2:378-379.

153. Röltgen K, Pluschke G: Mycobacterium ulcerans Disease (Buruli Ulcer): Potential

Reservoirs and Vectors. Current Clinical Microbiology Reports 2015, 2:35-43.

154. Ablordey AS, Vandelannoote K, Frimpong IA, Ahortor EK, Amissah NA, Eddyani M,

Durnez L, Portaels F, de Jong BC, Leirs H, et al: Whole genome comparisons suggest

random distribution of Mycobacterium ulcerans genotypes in a Buruli ulcer

endemic region of Ghana. PLoS neglected tropical diseases 2015, 9:e0003681.

Page | 185

155. Bolz M, Bratschi MW, Kerber S, Minyem JC, Um Boock A, Vogel M, Bayi PF, Junghanss

T, Brites D, Harris SR, et al: Locally Confined Clonal Complexes of Mycobacterium

ulcerans in Two Buruli Ulcer Endemic Regions of Cameroon. PLoS neglected tropical

diseases 2015, 9:e0003802.

156. Eddyani M, Vandelannoote K, Meehan CJ, Bhuju S, Porter JL, Aguiar J, Seemann T,

Jarek M, Singh M, Portaels F, et al: A Genomic Approach to Resolving Relapse versus

Reinfection among Four Cases of Buruli Ulcer. PLoS neglected tropical diseases 2015,

9:e0004158.

157. WHO: Laboratory diagnosis of Buruli ulcer – A manual for health care providers. 1st

edn. Geneva: WHO; 2014.

158. Andrews S: FastQC: A quality control tool for high throughput sequence data,

Available from http://www.bioinformatics.babraham.ac.uk/projects/fastqc/.

2015.

159. Harrison P, Seeman T: Nesoni, Available from https://github.com/Victorian-

Bioinformatics-Consortium/nesoni. Victorian Bioinformatics Consortium; 2014.

160. Seemann T: Snippy, Available from https://github.com/tseemann/snippy. 2015.

161. Li H, Durbin R: Fast and accurate short read alignment with Burrows-Wheeler

transform. Bioinformatics 2009, 25:1754-1760.

162. Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, Marth G, Abecasis G,

Durbin R: The Sequence Alignment/Map format and SAMtools. Bioinformatics 2009,

25:2078-2079.

163. Garrison E, Marth G: Haplotype-based variant detection from short-read

sequencing. arXiv preprint 2012, arXiv:1207.3907.

164. Barrick JE, Colburn G, Deatherage DE, Traverse CC, Strand MD, Borges JJ, Knoester

DB, Reba A, Meyer AG: Identifying structural variation in haploid microbial

genomes from short-read resequencing data using breseq. BMC genomics 2014,

15:1039.

165. Corander J, Tang J: Bayesian analysis of population structure based on linked

molecular information. Mathematical biosciences 2007, 205:19-31.

166. Corander J, Marttinen P, Siren J, Tang J: Enhanced Bayesian modelling in BAPS

software for learning genetic structures of populations. BMC bioinformatics 2008,

9:539.

167. Posada D, Crandall KA: Evaluation of methods for detecting recombination from

DNA sequences: computer simulations. Proceedings of the National Academy of

Sciences of the United States of America 2001, 98:13757-13762.

168. Bruen TC, Philippe H, Bryant D: A simple and robust statistical test for detecting the

presence of recombination. Genetics 2006, 172:2665-2681.

169. Huson DH, Bryant D: Application of phylogenetic networks in evolutionary studies.

Molecular biology and evolution 2006, 23:254-267.

170. Marttinen P, Hanage WP, Croucher NJ, Connor TR, Harris SR, Bentley SD, Corander J:

Detection of recombination events in bacterial genomes from large population

samples. Nucleic acids research 2012, 40:e6.

171. Mortimer TD, Pepperell CS: Genomic signatures of distributive conjugal transfer

among mycobacteria. Genome biology and evolution 2014, 6:2489-2500.

172. Stamatakis A: RAxML version 8: a tool for phylogenetic analysis and post-analysis

of large phylogenies. Bioinformatics 2014, 30:1312-1313.

173. Stamatakis A: The RAxML v8.1.X Manual, Available from http://sco.h-

its.org/exelixis/web/software/raxml/index.html. 2015.

Page | 186

174. Hodcroft E: TreeCollapserCL 4 http://emmahodcroft.com/TreeCollapseCL.html.

University of Edinburgh; 2013.

175. Hillis DM, Bull JJ: An Empirical-Test of Bootstrapping as a Method for Assessing

Confidence in Phylogenetic Analysis. Systematic biology 1993, 42:182-192.

176. Rambaut A: TreeStat v1.2, Available from

http://tree.bio.ed.ac.uk/software/treestat/. 2008.

177. Rieux A, Balloux F: Inferences from tip-calibrated phylogenies: a review and a

practical guide. Molecular ecology 2016, 25:1911-1924.

178. R Core Team: R v3.2.0: A language and environment for statistical computing,

Available from: http://www.R-project.org/. 2015.

179. Baele G, Lemey P, Bedford T, Rambaut A, Suchard MA, Alekseyenko AV: Improving

the accuracy of demographic and molecular clock model comparison while

accommodating phylogenetic uncertainty. Molecular biology and evolution 2012,

29:2157-2167.

180. Rambaut AS, MA. Xie, D. & Drummond, AJ. : Tracer v1.6, Available from

http://beast.bio.ed.ac.uk/Tracer. 2014.

181. Drummond AJ, Nicholls GK, Rodrigo AG, Solomon W: Estimating mutation

parameters, population history and genealogy simultaneously from temporally

spaced sequence data. Genetics 2002, 161:1307-1320.

182. Lemey P, Rambaut A, Drummond AJ, Suchard MA: Bayesian phylogeography finds its

roots. PLoS computational biology 2009, 5:e1000520.

183. He M, Miyajima F, Roberts P, Ellison L, Pickard DJ, Martin MJ, Connor TR, Harris SR,

Fairley D, Bamford KB, et al: Emergence and global spread of epidemic healthcare-

associated Clostridium difficile. Nature genetics 2013, 45:109-113.

184. Comas I, Coscolla M, Luo T, Borrell S, Holt KE, Kato-Maeda M, Parkhill J, Malla B, Berg

S, Thwaites G, et al: Out-of-Africa migration and Neolithic coexpansion of

Mycobacterium tuberculosis with modern humans. Nature genetics 2013, 45:1176-

1182.

185. Brosch R, Gordon SV, Marmiesse M, Brodin P, Buchrieser C, Eiglmeier K, Garnier T,

Gutierrez C, Hewinson G, Kremer K, et al: A new evolutionary scenario for the

Mycobacterium tuberculosis complex. Proceedings of the National Academy of

Sciences of the United States of America 2002, 99:3684-3689.

186. Croucher NJ, Harris SR, Grad YH, Hanage WP: Bacterial genomes in epidemiology--

present and future. Philosophical transactions of the Royal Society of London Series

B, Biological sciences 2013, 368:20120202.

187. Pakenham T: The Scramble for Africa: White Man's Conquest of the Dark Continent

from 1876 to 1912. New York: Avon Books; 1993.

188. Giles-Vernick T, Owona-Ntsama J, Landier J, Eyangoh S: The puzzle of Buruli ulcer

transmission, ethno-ecological history and the end of "love" in the Akonolinga

district, Cameroon. Social science & medicine 2015, 129:20-27.

189. Debacker M, Zinsou C, Aguiar J, Meyers WM, Portaels F: First case of Mycobacterium

ulcerans disease (Buruli ulcer) following a human bite. Clinical infectious diseases :

an official publication of the Infectious Diseases Society of America 2003, 36:e67-68.

190. Williamson HR, Mosi L, Donnell R, Aqqad M, Merritt RW, Small PL: Mycobacterium

ulcerans fails to infect through skin abrasions in a guinea pig infection model:

implications for transmission. PLoS neglected tropical diseases 2014, 8:e2770.

191. Aguiar J: L'ulcère de Buruli, une maladie mycobactérienne importante et en

recrudescence au Bénin. Bull Seances Acad R Sci Outre Mer 1997, 43:325-356.

Page | 187

192. Andersen FO: Mycobacterial Skin Ulcers. Clinical Experience. The Central African

journal of medicine 1965, 11:131-135.

193. Ravisse P: L'ulcère cutané à Mycobacterium ulcerans au Cameroun. Bulletin De La

Societe De Pathologie Exotique 1977, 70:109-124.

194. Ravisse P, Rocques M, Le Bourthe F, Tchuembou C, Menard J: Une affection

méconnue au Cameroun, l'ulcère à Mycobactérie. Med Trop 1975:471–474.

195. Muelder K, Nourou A: Buruli ulcer in Benin. Lancet 1990, 336:1109-1111.

196. Linz B, Balloux F, Moodley Y, Manica A, Liu H, Roumagnac P, Falush D, Stamer C,

Prugnolle F, van der Merwe SW, et al: An African origin for the intimate association

between humans and Helicobacter pylori. Nature 2007, 445:915-918.

197. WHO: WHO meeting on Buruli ulcer Control and Research: Summary report of the

control group. 2015.

198. Bos KI, Harkins KM, Herbig A, Coscolla M, Weber N, Comas I, Forrest SA, Bryant JM,

Harris SR, Schuenemann VJ, et al: Pre-Columbian mycobacterial genomes reveal

seals as a source of New World human tuberculosis. Nature 2014, 514:494-497.

199. Walker TM, Ip CL, Harrell RH, Evans JT, Kapatai G, Dedicoat MJ, Eyre DW, Wilson DJ,

Hawkey PM, Crook DW, et al: Whole-genome sequencing to delineate

Mycobacterium tuberculosis outbreaks: a retrospective observational study. The

Lancet Infectious diseases 2013, 13:137-146.

200. Ford CB, Lin PL, Chase MR, Shah RR, Iartchouk O, Galagan J, Mohaideen N, Ioerger

TR, Sacchettini JC, Lipsitch M, et al: Use of whole genome sequencing to estimate

the mutation rate of Mycobacterium tuberculosis during latent infection. Nature

genetics 2011, 43:482-486.

201. Wagner DM, Klunk J, Harbeck M, Devault A, Waglechner N, Sahl JW, Enk J, Birdsell

DN, Kuch M, Lumibao C, et al: Yersinia pestis and the plague of Justinian 541-543

AD: a genomic analysis. The Lancet Infectious diseases 2014, 14:319-326.

202. Kass RE, Raftery AE: Bayes Factors. Journal of the American Statistical Association

1995, 90:773-795.

203. Durnez L, Suykerbuyk P, Nicolas V, Barriere P, Verheyen E, Johnson CR, Leirs H,

Portaels F: Terrestrial small mammals as reservoirs of Mycobacterium ulcerans in

benin. Applied and environmental microbiology 2010, 76:4574-4577.

204. van Oye E, Ballion M, Janssens PG: Faudra-t-il tenir compte d'une nouvelle affection

à bacilles acido-résistants en Afrique? Ann Soc Belg Med Trop 1950, 30:619-621.

205. Kibadi K, Tiendrebeogo A, Ekoue Kinvi B, De Jong B, Boelaert M, Portaels F: [Buruli

ulcer in the health districts of the Democratic Republic of Congo from 1950 to 2013:

literature review and new distribution map]. Medecine et sante tropicales 2014,

24:420-429.

206. Phanzu DM, Suykerbuyk P, Saunderson P, Lukanu PN, Minuku JBM, Imposo DBB,

Diengidi BM, Kayinua M, Muyembe JJT, Lutumba PT, et al: Burden of Mycobacterium

ulcerans Disease (Buruli Ulcer) and the Underreporting Ratio in the Territory of

Songololo, Democratic Republic of Congo. PLoS neglected tropical diseases 2013, 7.

207. Andersen FO: Mycobacterial Skin Ulcers--Clinical Experience. The Central African

journal of medicine 1965, 11:131-135.

208. Smith JH: Epidemiologic observations on cases of Buruli ulcer seen in a hospital in

the Lower Congo. The American journal of tropical medicine and hygiene 1970,

19:657-663.

209. Quertinmont MJ: Etude clinique d'une affection nouvelle à Mycobacterium au

Maniéma (Congo belge). Acta Chirurgica Belgica 1959, 9:862-892.

Page | 188

210. Suykerbuyk P, Wambacq J, Phanzu DM, Haruna H, Nakazawa Y, Ooms K, Kamango K,

Stragier P, Singa JN, Ekwanzala F, et al: Persistence of Mycobacterium ulcerans

disease (Buruli Ulcer) in the historical focus of Kasongo Territory, the Democratic

Republic of Congo. The American journal of tropical medicine and hygiene 2009,

81:888-894.

211. Kibadi K, Panda M, Tamfum JJ, Fraga AG, Longatto Filho A, Anyo G, Pedrosa J,

Nakazawa Y, Suykerbuyk P, Meyers WM, Portaels F: New foci of Buruli ulcer, Angola

and Democratic Republic of Congo. Emerging infectious diseases 2008, 14:1790-

1792.

212. Chin CS, Alexander DH, Marks P, Klammer AA, Drake J, Heiner C, Clum A, Copeland A,

Huddleston J, Eichler EE, et al: Nonhybrid, finished microbial genome assemblies

from long-read SMRT sequencing data. Nature methods 2013, 10:563-569.

213. Seemann T: Prokka: rapid prokaryotic genome annotation. Bioinformatics 2014,

30:2068-2069.

214. Rutherford K, Parkhill J, Crook J, Horsnell T, Rice P, Rajandream MA, Barrell B:

Artemis: sequence visualization and annotation. Bioinformatics 2000, 16:944-945.

215. Kearse M, Moir R, Wilson A, Stones-Havas S, Cheung M, Sturrock S, Buxton S, Cooper

A, Markowitz S, Duran C, et al: Geneious Basic: an integrated and extendable

desktop software platform for the organization and analysis of sequence data.

Bioinformatics 2012, 28:1647-1649.

216. Parks DH, Mankowski T, Zangooei S, Porter MS, Armanini DG, Baird DJ, Langille MG,

Beiko RG: GenGIS 2: geospatial analysis of traditional and genetic biodiversity, with

new gradient algorithms and an extensible plugin framework. Plos One 2013,

8:e69885.

217. Heled J, Drummond AJ: Bayesian inference of population size history from multiple

loci. BMC evolutionary biology 2008, 8:289.

218. Miller HC, Moore JA, Allendorf FW, Daugherty CH: The evolutionary rate of tuatara

revisited. Trends in genetics : TIG 2009, 25:13-15; author reply 16-18.

219. Duchene S, Duchene D, Holmes EC, Ho SY: The Performance of the Date-

Randomization Test in Phylogenetic Analyses of Time-Structured Virus Data.

Molecular biology and evolution 2015, 32:1895-1906.

220. Stiller M, Baryshnikov G, Bocherens H, Grandal d'Anglade A, Hilpert B, Munzel SC,

Pinhasi R, Rabeder G, Rosendahl W, Trinkaus E, et al: Withering away--25,000 years

of genetic decline preceded cave bear extinction. Molecular biology and evolution

2010, 27:975-978.

221. Bafende AE, Phanzu MD, Imposo BB: Buruli ulcer in the Democratic Republic of

Congo: epidemiology, presentation and outcome. Tropical doctor 2004, 34:82-84.

222. Phanzu DM, Bafende EA, Dunda BK, Imposo DB, Kibadi AK, Nsiangana SZ, Singa JN,

Meyers WM, Suykerbuyk P, Portaels F: Mycobacterium ulcerans disease (Buruli

ulcer) in a rural hospital in Bas-Congo, Democratic Republic of Congo, 2002-2004.

The American journal of tropical medicine and hygiene 2006, 75:311-314.

223. Phanzu DM, Suykerbuyk P, Imposo DB, Lukanu PN, Minuku JB, Lehman LF,

Saunderson P, de Jong BC, Lutumba PT, Portaels F, Boelaert M: Effect of a control

project on clinical profiles and outcomes in buruli ulcer: a before/after study in Bas-

Congo, Democratic Republic of Congo. PLoS neglected tropical diseases 2011,

5:e1402.

224. Ho SY, Shapiro B: Skyline-plot methods for estimating demographic history from

nucleotide sequences. Molecular ecology resources 2011, 11:423-434.

Page | 189

225. Guy A: World Strategic Highways. Routledge; 2000.

226. Ehret C: Bantu expansions: Re-envisioning a central problem of early African

history. International Journal of African Historical Studies 2001, 34:5-41.

227. Kibadi K, Tsakala M, Mputu-Yamba JB, Muyembe T, Kashongwe M, Imposso B, Nsiala

A: [Buruli ulcer in Angolese refugees in the Kimpese area, Lower Congo, D.R.

Congo]. Sante 2003, 13:39-41.

228. Bar W, Rusch-Gerdes S, Richter E, de Bar GM, Dittmer C, Papsdorf H, Stosiek P, de Rijk

PB, Meyers WM, Portaels F: Mycobacterium ulcerans infection in a child from

Angola: diagnosis by direct detection and culture. Tropical Medicine & International

Health 1998, 3:189-196.

229. WHO: Treatment of Mycobacterium ulcerans disease (Buruli ulcer), Guidance for

health workers. Geneva; 2012.

230. Debacker M, Aguiar J, Steunou C, Zinsou C, Meyers WM, Portaels F: Buruli ulcer

recurrence, Benin. EmergInfectDis 2005, 11:584-589.

231. Amofah G, Asamoah S, Afram-Gyening C: Effectiveness of excision of pre-ulcerative

Buruli lesions in field situations in a rural district in Ghana. Tropical doctor 1998,

28:81-83.

232. Teelken MA, Stienstra Y, Ellen DE, Quarshie E, Klutse E, van der Graaf WTA, van der

Werf TS: Buruli ulcer: differences in treatment outcome between two centres in

Ghana. Acta Trop 2003, 88:51-56.

233. Kanga J, Kacou D, Sangaré A, Dabila Y, Asse N, Djakeaux S: Les rechutes observées

après le traitement chirurgical des cas d'ulcère de Buruli en Côte d'Ivoire. Bull Soc

Pathol Exot 2003, 96:406-409.

234. Lunn HF, Connor DH, Wilks NE, Barnley GR, Kamunvi F, Clancey JK, Bee JDA: Buruli

(Mycobacterial) ulceration in Uganda. (A new focus of Buruli ulcer in Madi district,

Uganda). East Afr Med J 1965, 42:275-288.

235. O'Brien DP, Walton A, Hughes AJ, Friedman ND, McDonald A, Callan P, Rhadon R,

Holten I, Athan E: Risk factors for recurrent Mycobacterium ulcerans disease after

exclusive surgical treatment in an Australian cohort. MedJAust 2013, 198:436-439.

236. Chauty A, Ardant MF, Marsollier L, Pluschke G, Landier J, Adeye A, Goundote A, Cottin

J, Ladikpo T, Ruf T, Ji B: Oral treatment for Mycobacterium ulcerans infection:

results from a pilot study in Benin. Clinical infectious diseases : an official publication

of the Infectious Diseases Society of America 2011, 52:94-96.

237. Kibadi K, Boelaert M, Fraga AG, Kayinua M, Longatto-Filho A, Minuku JB, Mputu-

Yamba JB, Muyembe-Tamfum JJ, Pedrosa J, Roux JJ, et al: Response to treatment in a

prospective cohort of patients with large ulcerated lesions suspected to be Buruli

ulcer ( Mycobacterium ulcerans disease). PLoSNeglTropDis 2010, 4:e736.

238. Bryant JM, Harris SR, Parkhill J, Dawson R, Diacon AH, van HP, Pym A, Mahayiddin

AA, Chuchottaworn C, Sanne IM, et al: Whole-genome sequencing to establish

relapse or re-infection with Mycobacterium tuberculosis: a retrospective

observational study. Lancet RespirMed 2013, 1:786-792.

239. Schurch AC, Kremer K, Kiers A, Daviena O, Boeree MJ, Siezen RJ, Smith NH, van SD:

The tempo and mode of molecular evolution of Mycobacterium tuberculosis at

patient-to-patient scale. Infect GenetEvol 2010, 10:108-114.

240. Eyre DW, Babakhani F, Griffiths D, Seddon J, Del Ojo EC, Gorbach SL, Peto TE, Crook

DW, Walker AS: Whole-genome sequencing demonstrates that fidaxomicin is

superior to vancomycin for preventing reinfection and relapse of infection with

Clostridium difficile. JInfect Dis 2014, 209:1446-1451.

Page | 190

241. Kaser M, Ruf MT, Hauser J, Marsollier L, Pluschke G: Optimized method for

preparation of DNA from pathogenic and environmental mycobacteria.

ApplEnvironMicrobiol 2009, 75:414-418.

242. Stamatakis A: RAxML version 8: a tool for phylogenetic analysis and post-analysis

of large phylogenies. Bioinformatics 2014, 30:1312-1313.

243. Teacher AG, Griffiths DJ: HapStar: automated haplotype network layout and

visualization. MolEcolResour 2011, 11:151-153.

244. Team QD: QGIS Geographic Information System. http://qgis.osgeo.org. Open

Source Geospatial Foundation Project 2012.

245. Global Administative Unit Layers data set

[http://www.fao.org/geonetwork/srv/en/metadata.show?id=12691]

246. World Health O: Treatment of Mycobacterium ulcerans disease (Buruli ulcer),

Guidance for health workers. Geneva; 2012.

247. Bryant JM, Schurch AC, van DH, Harris SR, de Beer JL, de JV, Kremer K, van Hijum SA,

Siezen RJ, Borgdorff M, et al: Inferring patient to patient transmission of

Mycobacterium tuberculosis from whole genome sequencing data. BMCInfect Dis

2013, 13:110.

248. Walker TM, Ip CL, Harrell RH, Evans JT, Kapatai G, Dedicoat MJ, Eyre DW, Wilson DJ,

Hawkey PM, Crook DW, et al: Whole-genome sequencing to delineate

Mycobacterium tuberculosis outbreaks: a retrospective observational study. The

Lancet infectious diseases 2013, 13:137-146.

249. Roltgen K, Stinear TP, Pluschke G: The genome, evolution and diversity of

Mycobacterium ulcerans Infect GenetEvol 2012, 12:522-529.

250. Bolz M, Bratschi MW, Kerber S, Minyem JC, Um BA, Vogel M, Bayi PF, Junghanss T,

Brites D, Harris SR, et al: Locally confined clonal complexes of Mycobacterium

ulcerans in two Buruli ulcer endemic regions of Cameroon. PLoS NeglTropDis 2015,

9:e0003802.

251. Hofer M, Hirschel B, Kirschner P, Beghetti M, Kaelin A, Siegrist CA, Suter S, Teske A,

Bottger EC: Brief report: disseminated osteomyelitis from Mycobacterium ulcerans

after a snakebite. The New England Journal of Medicine 1993, 328:1007-1009.

252. Tebruegge M, Curtis N: Mycobacterium marinum infection. Advances in

experimental medicine and biology 2011, 719:201-210.

253. Nugent G: Maintenance, spillover and spillback transmission of bovine tuberculosis

in multi-host wildlife complexes: a New Zealand case study. Veterinary Microbiology

2011, 151:34-42.

254. Adler B, de la Pena Moctezuma A: Leptospira and leptospirosis. Veterinary

Microbiology 2010, 140:287-296.

255. Steffen CM, Smith M, McBride WJ: Mycobacterium ulcerans infection in North

Queensland: the 'Daintree ulcer'. ANZ journal of surgery 2010, 80:732-736.

256. Bratschi MW, Ruf MT, Andreoli A, Minyem JC, Kerber S, Wantong FG, Pritchard J,

Chakwera V, Beuret C, Wittwer M, et al: Mycobacterium ulcerans persistence at a

village water source of Buruli ulcer patients. PLoS neglected tropical diseases 2014,

8:e2756.

257. Fyfe JA, Lavender CJ, Handasyde KA, Legione AR, O'Brien CR, Stinear TP, Pidot SJ,

Seemann T, Benbow ME, Wallace JR, et al: A major role for mammals in the ecology

of Mycobacterium ulcerans PLoSNeglTropDis 2010, 4:e791.

Page | 191

258. Bolz M, Ruggli N, Ruf MT, Ricklin ME, Zimmer G, Pluschke G: Experimental infection

of the pig with Mycobacterium ulcerans: a novel model for studying the

pathogenesis of Buruli ulcer disease. PLoS neglected tropical diseases 2014, 8:e2968.

Page | 192

doc.anet.be Page | 1 Faculty of Sciences Department of Biology Unveiling the evolutionary history &...

Documents

Transcript of doc.anet.be Page | 1 Faculty of Sciences Department of Biology Unveiling the evolutionary history &...