Javed Mohammed Khan and Shoba Ranganathan

1

A multi-species comparative structural bioinformatics analysis of

inherited mutations in -D-Mannosidase reveals strong

genotype-phenotype correlation

Javed Mohammed Khan and Shoba RanganathanMacquarie University, Sydney,

[email protected]

2

-D-Mannosidase Lysosomal -mannosidase acts to degrade N-linked

oligosaccharides or mannose-rich compounds.

The deficiency of this enzyme results in a recessively inherited lysosomal storage disease, called -mannosidosis.

Mutations in the gene (MAN2B1) encoding lysosomal -D-mannosidase cause the disease.

Structure of bovine-D-mannosidase shows five chains held togetherby four disulfide bonds.

Missense, nonsense, insertions, deletions and also some splicing mutations.

A

B

C

D

E

3

1. Malm D, Nilssen Ø: Alpha-mannosidosis. Orphanet J Rare Dis. 2008, 3:21.

2. Ockerman PA. A generalised storage disorder resembling Hurler's syndrome. Lancet. 1967;2:239.

Artemisia annua

• Coarse face, prominent forehead, mental retardation, skeletal malformation, hearing loss and cerebral dysfunction leading to paralysis and death1.

• A severe infantile form– Leads to early death

• A mild juvenile form – Survival into adult life

• Domestic cats, cattle and guinea pigs.

• First characterized in humans by Ockerman in 19672.

Phenotypes and species affected

4

Why study -D-mannosidosis? It can be comparatively well studied due to its occurrence

across various species.

Clinicians, geneticists and molecular biologists have not been able to correlate the genotypic mutations with the observed phenotype1, 3.

The high phenotypic variability, has so far prevented adoption of a standardized clinical typing.

-mannosidosis occurs in 1 of 500,000 live births1.

– Errors of lysosomal metabolism occur in approximately 1:5000 live births.

– This approach can be extended to other lysosomal disorders.


3. Lyons MJ, Wood T, Espinoza L, Stensland HM, Holden KR: Dev Med Child Neurol 2007, 49:854-857.

5

Our approach Use protein structure to unravel how the mutations affect the

function of the enzyme, especially with respect to the enzyme active site, to explain observed phenotypes.

Data sources for mutations OMIM: Online Mendelian Inheritance in Man Publications from PubMed. OMIA: Online Mendelian Inheritance in Animals

Sequence information GenBank, Swiss-Prot

Sequence alignment and visualization CLUSTAL, WEBLOGO

3D structural modeling MODELLER

3D structure assessment Biotech Validation Suite

6

Issues related to modelling

• The X-ray crystal structure for bovine lysosomal -D-mannosidase (PDB ID: 1O7D)4 lacks:

– Two of the four vital disulfide bonds (as annotated by Swiss-Prot). – A few structurally and functionally important residues.

• MAN2B1 produces a five chain protein.

• Existence of 11 essential ligands.

• Presence of cleavable signal peptide (50 AA) in database sequences and water molecules in 3D structure.

• Swiss-Prot and WEBLOGO depict a well conserved propeptide region.

4. Heikinheimo P, et al.: J Mol Biol 2003, 327:631-644.

7

• A single comprehensive molecular modeling procedure.

• Addresses all the issues related to -D-mannosidase modeling.

What is the solution?

First Alpha Helix of Chain B

New SS bond New SS bond





Mo

lecu

lar

mo

del

ing

Seq

uen

ce r

etri

eval

Seq

uen

ce

alig

nm

ent

Template sequence

Extract template protein sequence using MODELLER

Get protein sequencesSwiss-Prot

Align sequences with template sequence using ClustalX

Input alignment file to

MODELLER

Template structure file from PDB

MODELLER command file

Template selection

Build homology modelsusing MODELLER

Optimize structure iteratively

Verify model quality withBiotech validation suite

High quality structural models

Refine lowquality

structuresMo

lecu

lar

mo

del

ing

Seq

uen

ce r

etri

eval

Seq

uen

ce

alig

nm

ent

Template sequence

Extract template protein sequence using MODELLER

Get protein sequencesSwiss-Prot

Align sequences with template sequence using ClustalX

Input alignment file to

MODELLER

Template structure file from PDB

MODELLER command file

Template selection

Build homology modelsusing MODELLER

Optimize structure iteratively

Verify model quality withBiotech validation suite

High quality structural models

Refine lowquality

structures

• Rebuilt the two missing disulfide bridges along with all the existing ligands.

8

Assessment of the Wild-Type structural models

Wild typeModels

PROVE PROCHECK WHATIFRMSD (Å)

Z-Scoreaverage

Residues in most favoured regions

Residues inallowed regions

Z-Score

Bovine 0.08 89.9% 8.8% 1.31 0.70

Human -0.04 89.1% 9.7% 1.34 0.64

Guinea pig 0.05 86.7% 11.5% 1.50 0.89

Feline 0.05 89.0% 9.9% 1.40 0.71

• High degree of similarity between target and template sequences and strict modeling protocols adopted.

• Excellent RMSD and WHATIF Z-Scores.

• Above average PROVE Z-Score and PROCHECK validation results.

• Generation of high quality WT models.

9

Mapping mutations to WT structural models • We have constructed a

mutation map– mapping available mutations

in the context of the enzyme active site

– to understand where the observed mutations occur

• Most mutations with lethal phenotypes are located in and around the active site.

• Thereby affecting the functionality of the enzyme.

10

Genotype-Phenotype correlation

Non-sense Mutations Insertions Deletions

Feline Truncation

A B C D E

S. No Species Mutated residues Structurallocation

Effect on Structure Phenotypic effect

1. Human

H72L In the active site

Disrupts the active site

Lethal(Type 3)

H200N, T355P,P356R, E402K, S453Y

Close to the active site

Perturbs the fold Harmful(Type 2)

W714R, R750W,G801D, L809P

Away from the active site

Minor perturbation Mild or Viable(Type 1)

2.Bovine

R220H In the active site

Disrupts the active site

Lethal(Type 3)

F320L Close to the active site


3. Guinea pig R227W Close to the active site


• The three phenotypes correspond to Type 3, Type 2 and Type 1 clinical phenotypes described by Malm and Nilssen1.

• Mapped all truncation mutations to different chains of human WT protein.


11

Prediction of potentially harmful mutations 51 85 185 195 MA2B1_BOVIN AAGYKTCPKVKPDMLNVHLVPHTHDDVGWLKTVDQ FGSDGRPRVAW MA2B1_FELCA AAGYETCPMVHPDMLNVHLVAHTHDDVGWLKTVDQ FGKDGRPRVAW MA2B1_HUMAN AGGYETCPTVQPNMLNVHLLPHTHDDVGWLKTVDQ FGNDGRPRVAW MA2B1_CAVPO AAGYETCPMVQPGMLNVHLVAHTHDDVGWLKTVDQ FGSDGRPRVAW *.**:*** *:*.******:.************** **.******** 201 235 317 327 MA2B1_BOVIN GHSREQASLFAQMGFDGFFFGRLDYQDKKVRKKTL TMGSDFQYENA MA2B1_FELCA GHSREQASLFAQMGFDGLFFGRLDYQDKRVREENL TMGSDFQYENA MA2B1_HUMAN GHSREQASLFAQMGFDGFFFGRLDYQDKWVRMQKL TMGSDFQYENA MA2B1_CAVPO GHSREQASLFAQMGFDGVFFGRIDYQDKLVRKKRR TMGSDFQYENA *****************.****:***** ** : *********** 354 367 401 410 451 457 517 525 MA2B1_BOVIN VLYSTPACYLWELN LKRYERLSYN AVSGTSR VYNPLGRKV MA2B1_FELCA VLYSTPACYLWELN LKRYERLSYN AVSGTSK IYNPLGRKI MA2B1_HUMAN VLYSTPACYLWELN LKRYERLSYN AVSGTSR VYNPLGRKV MA2B1_CAVPO VLYSTPACYLWELN LKRYERLSYN AVSGTSK VYNPLGRKV ************** ********** ******: :*******: 561 571 637 646 759 767 917 926 MA2B1_BOVIN Q----ELLFSA LLPVRQAFYW RRNYRPTWK ETLLLRLEHQ MA2B1_FELCA Q----ELLFPA LLPVRQAFYW RRDYRPTWK KTLLLRLEHQ MA2B1_HUMAN QAHPPELLFSA LLPVRQTFFW RRDYRPTWK EMVLLRLEHQ MA2B1_CAVPO P----ELLFPA TLPVNQAFFW RRDYRPSWK DTLLLRLEHQ ****.* ***.*:*:* **:***:** . :*******

• Highly conserved sequence across all species.

• Almost all mutated residues causing fatal/harmful phenotypes are highly conserved.

• All these positions can be considered potential disease-causing mutations for all mammals represented here.

• Mutations in residues comprising the active site of the enzyme could have serious effects

• This residue set represents a structurally-derived mutation hot-spot.

12

Sequence-based mutational hot-spot regions in the MAN2B1 gene

Missense H72L

H200N

H200L R220H

R227W F320L

T355P

P356R E402K

L518P W714R R750W

G801D L809P

R916S

Nonsense

E53X

R188X

Y359X

E563X

Y461X

W77X

5' 3'

1 3468

157 - 323 562 - 679 961 - 1204 1383 - 1815 2140 - 2746

Insertions 293–294 insA

322–323 insA

1076–1077 insA

1153–1154 insA

1197–1198 insA

Deletions 965delAT

1748del4

1815delA

2548delC

2660delC

MAN2B1

• Mutations are scattered along the length of the gene.

• Mutations seem to cluster into groups over segments of varying sequence length called mutational hot-spot regions.

• Five distinct mutational hot-spot regions with lengths varying from 117 to 606 nucleotides.

• Residues coded for by the nucleotides within the range 961-1204 are most likely to undergo mutations.

• Occurrence of a harmful mutation is most likely to be between 157-323, 562-679 and 961-1204 hot-spot regions due to their close proximity to the active site.

13

Significance of this study• Can be used as a predictive approach for detecting likely α-

mannosidosis across various species.

• Novel prediction protocol for new disease mutations related to α-mannosidosis.

• Approach can be extended to other inherited disorders.

• Provides a way for detecting mutation hotspots in the gene, where novel mutations could be implicated in disease.

• A rational approach for predicting the phenotype of a disease, based on observed genotypic variations.

14

• Establishes a significant correlation between the genotype and the phenotype of the disease.

• Highlights the effect of disease mutations on protein structure and forms the basis for understanding the molecular determinants for phenotypic variations.

• High degree of mutational heterogeneity of α-mannosidosis is comparable to that observed in many other lysosomal disorders.

• Suggests that rather than drug/inhibitor design, this disease could be tackled through gene therapy.

• This study could play a vital role in developing therapies for inherited diseases.

Conclusions

15

Acknowledgements

Prof. Shoba Ranganathan.

Macquarie University for MQRES scholarship.

Colleagues and friends at Macquarie University.

The InCoB 2009 program and organising committees.

16

Questions ?

Javed Mohammed Khan and Shoba Ranganathan

Documents

Transcript of Javed Mohammed Khan and Shoba Ranganathan