How did life survive Earth's great...

How did life survive Earth’s great oxygenation?Woodward W Fischer1, James Hemp1 andJoan Selverstone Valentine1,2

Available online at www.sciencedirect.com

ScienceDirect

Life on Earth originated and evolved in anoxic environments.

Around 2.4 billion-years-ago, ancestors of Cyanobacteria

invented oxygenic photosynthesis, producing substantial

amounts of O2 as a byproduct of phototrophic water oxidation.

The sudden appearance of O2 would have led to significant

oxidative stress due to incompatibilities with core cellular

biochemical processes. Here we examine this problem through

the lens of Cyanobacteria — the first taxa to observe significant

fluxes of intracellular dioxygen. These early oxygenic organisms

likely adapted to the oxidative stress by co-opting preexisting

systems (exaptation) with fortuitous antioxidant properties. Over

time more advanced antioxidant systems evolved, allowing

Cyanobacteria to adapt to an aerobic lifestyle and become the

most important environmental engineers in Earth history.

Addresses1 Division of Geological & Planetary Sciences, California Institute of

Technology, Pasadena, CA 91125, United States2 Department of Chemistry and Biochemistry, UCLA, Los Angeles, CA

90095, United States

Corresponding authors: Fischer, Woodward W ([email protected])

and Valentine, Joan Selverstone ([email protected])

Current Opinion in Chemical Biology 2016, 31:166–178

This review comes from a themed issue on Bioinorganic chemistry

Edited by R David Britt and Emma Raven

http://dx.doi.org/10.1016/j.cbpa.2016.03.013

1367-5931/# 2016 Elsevier Ltd. All rights reserved.

IntroductionData from the geological record indicate that life was

present on Earth very early in its history. Intriguing obser-

vations of graphitic carbon in some of the oldest rocks and

minerals have been proposed to be traces of Earth’s early

biosphere [1,2]. By �3.4–3.2 billion years ago (giga annum

or Ga), a range of observations indicate the presence of

microbial cells [3,4] with diverse anaerobic metabolisms

[5–8]. Even using the most conservative estimate of 3.2 Ga

for the origin of life, it was clearly present long before O2

appeared in significant amounts in Earth’s atmosphere.

The first organisms to encounter significant and sustained

oxidative stress due to O2 were ancestral Oxyphotobac-teria — members of the bacterial phylum Cyanobacteria,


capable of oxygenic photosynthesis. These organisms like-

ly supported themselves largely by means of anoxygenic

photosynthesis at first, perhaps with only intermittent

production of O2 in relatively small amounts. At this stage

in Earth history, life did not have the defenses necessary to

deal with an oxidant as powerful as O2 [9��]. How did early

Oxyphotobacteria survive the intracellular production of O2

and make the transition from a strictly anaerobic to aerobic

metabolism? Here we integrate data from bioinorganic

chemistry and comparative biology to infer the evolution

of oxygen tolerance in Oxyphotobacteria. Notably, ancestral

Oxyphotobacteria could have coped with small amounts of

O2 during the transition to oxygenic phototrophy by co-

opting preexisting systems with fortuitous antioxidant

properties. This would have allowed time for more modern

antioxidant systems to evolve. We discuss non-enzymatic,

small-molecule solutions to mitigate O2 stress that were

present in early cells, followed by some specific adaptations

that the Oxyphotobacteria developed to enable the transition

to an oxygenated world.

Geological record of O2

Today O2 comprises nearly 21% of the atmosphere,

however a wide array of observations made over the past

sixty years from the geological record illustrates that it was

extremely scarce prior to the evolution of oxygenic pho-

tosynthesis [10�]. How scarce? An exact paleobarometer

for O2 remains out of reach, but there are several types of

geological and geochemical data that can be converted

into O2 concentrations that are thought to be accurate

within an order of magnitude or two (Figure 1). Sedimen-

tary rocks older than 2.4 Ga, composed of pebbles and

sand eroded from the crust and deposited by rivers, are

distinct from those seen in younger strata because they

contain abundant physically rounded redox-sensitive

minerals, such as pyrite (FeS2), uraninite (UO2), and

siderite (FeCO3) [11]. These minerals are quickly oxi-

dized and destroyed in the presence of even trace O2, so

their presence constrains O2 levels to less than �10�5 atm

before 2.4 Ga [11]. Additional independent geochemical

proxies support this view. For example, multiple sulfur

isotopes in marine sedimentary rocks older than 2.4 Ga

display a widespread and unusual type of mass indepen-

dent fractionation (MIF) caused by the photochemistry of

SO2 in an atmosphere largely devoid of O2 (�10�5 atm

and likely closer to 10�10 atm). This corresponds to

astonishingly low environmental levels of O2 — suffi-

ciently low that the anaerobic microorganisms that

existed at the time may not have ever encountered

biologically meaningful amounts of oxygen related stress.

www.sciencedirect.com

http://crossmark.crossref.org/dialog/?doi=10.1016/j.cbpa.2016.03.013&domain=pdf

mailto:[email protected]

mailto:[email protected]

http://dx.doi.org/10.1016/j.cbpa.2016.03.013

http://www.sciencedirect.com/science/journal/13675931

How did life survive Earth’s great oxygenation? Fischer, Hemp and Valentine 167

Figure 1

Age (Ma)

Log[

Mn]

(pp

m)

101

102

104

103

100

10–1

Today35004000 3000 2500 2000 500100015004500

Orig

in o

f the

Ear

th ~

4.5

5 G

a

Hadean Archean Proterozoic Phan.

Old

est s

edim

enta

ry r

ocks

GO

E

Manganese deposits

MIF of S isotopes

Fe in Paleosols & Red bedsredox-sensitivedetrital grains

Sulfate deposits

Charcoal

animals

plants

Current Opinion in Chemical Biology

Geological and geochemical data reveal a large first-order, irreversible change in the redox state of Earth surface environments marked by the rise of O2

2.4–2.35 billion years ago. Upper panel: The widespread presence of redox-sensitive detrital grains, like pyrite and uraninite, in Archean and early

Paleoproterozoic sandstones and conglomerates illustrates that O2 levels were lower than 10�5 atm in the atmosphere and Earth surface waters to

explain their survival through the rock formation processes of weathering, transport, deposition and lithification. A similar pattern is provided by the

mass-independent fractionation (MIF) of multiple S isotopes in Archean and early Proterozoic strata [89,90]. The MIF signal results from photochemistry

involving SO2 in the early atmosphere and models constructed to evaluate the O2 concentrations consistent with these observations suggest that O2

levels were exceedingly low, much less than 10�5 atm, and perhaps closer to 10�10 atm [10�]. These observations are juxtaposed by a number of

observations that show a rise of O2 between 2.4 and 2.35 Ga. MIF and redox-sensitive detrital grains disappear from sedimentary rocks [11,12]. Iron

becomes oxidized and retained in preserved soils (paleosols) during weathering of bedrock [13], and it forms abundant hematite (Fe2O3) grains and

cements in sedimentary rocks (red beds). And sulfate salts become conspicuous sulfur cycle-sinks of sulfate derived from weathering of sulfide-bearing

minerals [15]. Though young by comparison, the charcoal record [91] provides an important constraint on atmospheric oxygen because it illustrates that

as long as there was plant biomass around on the land surface that could burn, it did. Mn deposits (Mn-rich sedimentary rocks >1 wt.% Mn) do not

occur until just prior to the rise of O2, an observation that motivates the hypothesis that Mn(II) was a substrate for phototrophy prior to the evolution of

the water-oxidizing complex of PSII, and photosynthetic O2 fluxes [29��]. The rise of O2 marks the oldest certain age for the evolution of biological water

splitting by Oxyphotobacteria [10�]. The origins of animals and plants are shown for context. Lower panel: Mn(II) content of calcite-bearing (CaCO3) and

dolomite-bearing (CaMg(CO3)2) sedimentary rocks provides a coarse measure of the amount of Mn2+ present in seawater (data from Shields and Veizer

[18]). Before the GOE, carbonates precipitated from seawater are highly enriched in Mn(II), implying high concentrations in seawater. After the rise of

oxygen, Mn-oxide minerals form an important sink of Mn, and the Mn content of seawater subsequently decreased [14]. These measurements exhibit

substantial variation due to post-depositional recrystallization and interaction with later fluids, which tend to enrich carbonates in Mn(II).

The geological record indicates that a rapid and irrevers-

ible rise of O2 occurred between 2.4 and 2.35 Ga — a

transition known as the Great Oxygenation Event (GOE)

(Figure 1) [10�,12]. Many of these observations record


changes in the biogeochemical cycles of major redox-

active elements, such as Fe, Mn, and S. At this time MIF

[10�] and redox sensitive detrital grains disappeared [11].

Ferrous iron in igneous minerals became oxidized and


168 Bioinorganic chemistry

was retained in ancient soils [13]. Iron oxide cements

become widespread in fluvial (river) and near-shore ma-

rine sandstones, called red beds. Thick deposits of Mn-

oxide minerals like those in the Kalahari Manganese

Field in South Africa [14] postdate the GOE, with an

important exception (more on this below). In addition

sulfate salts begin to accumulate in sedimentary basins,

recording the oxidative weathering of sulfide-bearing

minerals [15]. Because these redox proxies are sensitive

to small amounts of O2 (e.g., ppm levels), we do not have

good estimates for how high O2 rose during the GOE

(some hypotheses suggest perhaps only 1% of modern).

However it is clear from the geological record that, after

the GOE, O2 became widespread across vast swaths of the

Earth’s surface and was thereafter an important compo-

nent of the atmosphere and surface waters.

Changes in metal ion availability: Metal use in biology is

primarily determined by evolutionary history, which is

the result of dynamic interactions between environmen-

tal availability, biochemical demands, and ecological

flexibility. The metal concentrations in seawater have

changed dramatically over Earth history as a function of

the differential solubility, complexation, and redox be-

havior of distinct metals — modulated by changes in the

redox state of Earth’s surface environments and geo-

chemical budgets set by different rock-forming minerals

within the crust [16,17]. Iron and manganese are the first

and third most abundant transition elements in the crust

and substitute for one another in a wide range of rock-

forming silicate minerals, dominantly as Fe(II) and ex-

clusively as Mn(II). Consequently, prior to the GOE

chemical weathering of silicate minerals would have

provided substantial fluxes of Fe2+ and Mn2+ into the

oceans, resulting in high concentrations in seawater [18].

In particular, Archean-age carbonate platforms are loaded

with remarkably high levels of Mn(II) — much higher

than observed in younger carbonate platforms (Figure 1)

[14,18]. It is challenging to convert these data accurately

into seawater concentrations because the partitioning

depends on the kinetics and mechanisms of carbonate

mineral precipitation, but experimental relationships and

solubility constraints suggest that seawater Mn2+ concen-

trations prior to the GOE ranged from �5 mM to

�120 mM [19]. After the GOE, iron availability dramati-

cally decreased in surface seawater due to removal of

insoluble iron oxides, marked by red beds and ferric iron

in paleosols (Figure 1). However iron was so thoroughly

integrated into cellular biology by that point that creative

strategies were required so that organisms would be able

to obtain it in oxygenated environments. Mn concentra-

tions appear not to have fallen as precipitously, likely

because the kinetics of Mn2+ oxidation are substantially

slower than those of Fe2+ [20]. In contrast to iron, for

example, the bioavailability of Zn2+ does not appear to

have changed dramatically over time [21,22]. O2-driven

oxidative weathering of sulfide-bearing minerals in the


crust — evidenced by the loss of detrital pyrite from the

record (Figure 1) — sourced and solubilized chalcophilic

elements like Cu, which was uniquely suited for the

evolution of high potential metabolisms and aerobic

biology [16].

Biological record of O2

The relative timing of the appearance of O2 from com-

parative biology and biochemistry is consistent with

observations from the geological record. These data sug-

gest that O2 was not used by early life in either biosyn-

thetic reactions or respiration. Network analyses of

metabolic pathways from diverse microbes identified a

strictly anaerobic core, with O2 requiring reactions

appearing only in the terminal steps of a small fraction

of molecules [23]. This implies that O2 was not available

to early organisms as a substrate for biochemical reactions

and was only incorporated into metabolism after the

evolution of oxygenic phototrophy.

A similar view is provided by phylogenomic analyses of

the distribution of aerobic respiration. The majority of

Bacteria and Archaea phyla have anaerobic basal mem-

bers with aerobic members found only in derived (i.e. non

ancestral) positions (Figure 2a). For example, in the

Actinobacteria phylum members of the basal classes Rubro-bacteria and Coriobacteria are anaerobic. Classes with

aerobic members (Acidimicrobiia, Nitriliruptoria, and Acti-nobacteria) are derived. Comparisons of the different O2

reductases used by Actinobacteria for aerobic respiration

shows two major independent acquisitions (once in the

Thermoleophilia and once in the Acidimicrobiia + Nitrilir-uptoria + Actinobacteria clade), with a subsequent loss in

the Bifidobacteria genus after they adapted to anoxic gut

ecosystems (Figure 2a). Many other phyla exhibit similar

phylogenetic patterns, suggesting that the major radiation

of Bacteria occurred at a time when the Earth was anoxic

and that aerobic respiration was acquired after they had

diverged from one another, likely after the GOE.

The biological record over the past two billion years

shows that O2 has driven major revolutions in biology,

forming a biochemical veil that makes it challenging to

infer some aspects of early life prior to O2 with a high

degree of confidence. Deductions from comparative bi-

ology and biochemistry have tremendous value for study-

ing this problem, but it is important to remember that

they present a view biased by extinction. The ‘winners’

wrote the biological record. Mechanisms to cope with O2

and reactive oxygen species (ROS) toxicity have been

thoroughly integrated into the function of modern

cells — even for what are typically viewed as strictly

anaerobic organisms [24]. All we easily study are the

biological solutions selected after several billion years

of evolution. The organisms and biochemistries that

did not survive can only be inferred from geological

observations and imagined via evolutionary hypotheses



Figure 2

0.1

AerobicAnaerobic

RubrobacteriaThermoleophiliaCoriobacteriiaAcidimicrobiiaNitriliruptoriaActinobacteria

(a) (b)

AerobicAnaerobic

12

3


Observations from comparative biology suggest that O2 and aerobic metabolisms evolved late in Earth history — a pattern that matches

expectations based on the extremely low O2 levels on the early Earth-derived geological observations shown in Figure 1. (a) Phylogenetic tree of

the Actinobacteria phylum based on the conserved marker gene RpoB — a useful phylogenetic marker protein from the b subunit of the bacterial

RNA polymerase. Aerobic respiration was acquired two times independently within this phylum: once in the last common ancestor of the clade

comprised of Acidimicrobiia, Nitriliruptoria, and Actinobacteria (marked 1) classes, and once in the ancestor of the Thermoleophilia (marked 2).

Members of the Bifidobacteria genus lost the capacity for aerobic respiration as they became specialized for anaerobic gut environments (marked

3). The phylogenetic distribution seen here is characteristic of many known bacterial phyla, suggesting that O2 was not present prior to the main

divergences of bacterial phyla from one another. This supports the late evolution of aerobic respiration, with lateral gene transfer playing a

significant role in its modern distribution. (b) A network view of the metabolic networks in the KEGG database containing diverse metabolisms

from all three domains of life from the data reduction by Raymond and Segre [23]. Nodes mark metabolites and edges reflect reactions. Red

denotes anaerobic metabolism, blue marks areas of the network that are aerobic, and green highlights networks where aerobic reactions are

known to have replaced ones that did not involve O2 (e.g. Raymond and Blankenship [92]). Note that aerobic parts of the network decorate the

outer edges of the network map and are not central to these metabolic processes. These data imply that O2 was not a founding molecule for

cellular metabolism.

from the vestiges we see. While it is often envisioned that

systems for dealing with the toxicity of O2 must have

been in place for cells to survive the GOE, it is, however,

just as reasonable that such systems co-evolved in time

with dioxygen because they would have had little selec-

tive advantage prior to the GOE. Instead cells may have

initially coped with O2 stresses using molecules that

performed other functions. And those that were the most

helpful early on were those that had fortuitous antioxidant

chemistry.

The transition to oxygenic phototrophyNew insights into the evolution of Cyanobacteria: We can gain

considerable insights about how life adapted to O2 by

examining Oxyphotobacteria, the first organisms to witness

substantial fluxes of intracellular O2. Genomic studies

over the last couple years have dramatically changed the

way we think about the diversity and evolution of the


Cyanobacteria phylum. New members have been found in

many aphotic environments and analysis of genomes

assembled from environmental metagenomic datasets

(and one cultured isolate) demonstrate that these organ-

isms are missing genes for phototrophy [10�,25�,26��,27].

Current evolutionary relationships between members of

the Cyanobacteria phylum are shown in Figure 3; it is

important to note that oxygenic photosynthesis exhibits a

derived position [10�]. A new classification scheme has

been proposed [26��] in which the Cyanobacteria are now

comprised of three classes: the Oxyphotobacteria (oxygenic

Cyanobacteria), the Melainabacteria, and ML635J-21.

After their divergence from the Melainabacteria, it was

stem group members of the Oxyphotobacteria (between

markings 1 and 2 in Figure 3) — perhaps long ex-

tinct — that developed oxygenic photosynthesis. Be-

tween molecular clock estimates [28] and geological



Figure 3

other phyla

0.1

Melainabacteria

Oxyphotobacteria

ML635J-21


Phylogeny of the Cyanobacteria phylum (shown here from 16S ribosomal DNA) has grown substantially in recent years with advances in genomic

and metagenomic sequencing (e.g., [26��]). Current relationships show that all known Cyanobacteria that produce O2 via oxygenic

photosynthesis — a class now termed Oxyphotobacteria — sit in a derived position within the phylum. The Oxyphotobacteria have a close sister

clade: the Melainabacteria [25�]. No known Melainabacteria are capable of photosynthesis; most are anaerobic. Basal members of the phylum

form a paraphyletic group only known from environmental samples, currently termed ML635J-21. This topology suggests that oxygenic

photosynthesis in the Oxyphotobacteria is a relatively recent innovation in the context of Earth history [10�]. Molecular clock estimates suggest

that the divergence between the Oxyphotobacteria and Melainabacteria (marked 1) occurred between 2.5 and 2.6 Ga, whereas the radiation of

crown group Oxyphotobacteria (marked 2) occurred after �2.0 Ga [28].

data [10�,12,29��], this appears to have occurred around

2.4–2.35 billion years ago.

Manganese: a gateway to the high potential world: The transi-

tion from anoxygenic phototrophy to oxygenic photosyn-

thesis in Oxyphotobacteria required the evolution of two

important features: a high potential reaction center, and a

CaMn4O5 bioinorganic cluster called the water-oxidizing

complex (WOC) capable of oxidizing water [30]. Modern

anoxygenic reaction centers are unable to generate high

redox potentials (<+500 mV). The highest potential elec-

tron donor for anoxygenic phototrophy is nitrite

(+430 mV) [31]. One hypothesis emerging in recent years

from both geological and biological data is that Mn-

oxidizing phototrophy was likely the direct precursor to

oxygenic photosynthesis [10�]. Mn2+ was in abundant in

the early oceans (Figure 1) and has redox potentials near

to that of water. Biochemical and structural analyses of

reaction centers strongly suggest that the ability to oxi-

dize Mn compounds drove the evolution of high potential

reaction centers in early Oxyphotobacteria [30]. In fact,

modern Oxyphotobacteria still oxidize Mn using PSII dur-

ing the photoassembly of their WOCs. The WOC may


therefore have originated by the accumulation of oxidized

Mn in the high potential reaction center, with catalytic

Mn oxidation eventually being replaced by water oxida-

tion. Importantly, evidence for this transition is found in

the geological record where O2-independent oxidation of

Mn was observed to have occurred prior to the GOE

(Figure 1) [29��].

Initial threats from O2

Early O2 production: fluxes, concentrations, and timescales:Once oxygenic photosynthesis evolved, the timescale for

O2 to titrate the existing pools of geochemically derived

reductants in Earth’s surface environments, which there-

after became oxygenated, is relatively short in the context

of geological time — on the order of tens of thousands to a

hundred thousand years [32]. A number of anoxic envir-

onments would endure through the GOE (e.g., pore fluids

in marine sediments), and provide refuge for anaerobic

physiologies. But for oxygenic phototrophs, O2 was an

undeniable problem molecule. What are the concentra-

tions and fluxes of O2 that typify photosynthetic cells?

Recent calculations from theory suggest that the O2

concentrations inside active photosynthetic cells are



much lower than expected — tens of nanomolar greater

than the external environment for solitary planktonic

cells, though much higher for cells living in biofilms

and mats, which can achieve highly supersaturated O2

levels [33��]. This work highlights an important asym-

metry between photosynthetic O2 production and con-

sumption by aerobic respiration: the rates of O2

generation are slow compared to diffusive fluxes. This

lessens the degree of oxidative stress for the earliest

Oxyphotobacteria [33��]. Even functioning at or near mod-

ern photosynthetic rates (�10�18 mol cell�1 s�1), intra-

cellular concentrations were maximally 250 nM [33��];presumably early Oxyphotobacteria did not produce O2 at

anywhere near these high rates. It is important to note

that these fluxes are still large compared to the concen-

trations required for O2 to disrupt the intracellular pools

of Fe [20], sulfide [34], thiols, cysteine [35��], flavins [36],

and metal centers [37].

1O2 as the first oxidative threat: In addition to ROS formed by

O2 reduction at various points along electron transport

chains (e.g. Complex I, Complex III, PSI), photoexcited

singlet O2 presented a unique problem for Oxyphotobacteriaas a substantial and inchoate oxidative threat. Photosystems

are studded with chlorophylls as the major light harvesting

molecules. Energy transfer from excited long-lived triplet

states of chlorophyll to ground state triplet O2 can create1O2 with extremely high quantum efficiency [38]. Further-

more 1O2 is highly reactive with a wide range of molecules

including proteins and lipids, has a short half-life within

cells [39], and is one of the most critical sources of oxidative

damage in photosynthetic cells [40]. For phototrophic

organisms that generate O2, there was no easy way around

this problem. Light-harvesting can be carefully tuned by

mechanisms of non-photochemical quenching (NPQ), and

carotenoids can be used to dissipate energy from photo-

sensitized chlorophyll; however a good solution is simply to

keep intracellular O2 levels low, particularly near the

photosystems. This logic appears to have been important

for early Oxyphotobacteria, which appear to have employed

flavodiiron proteins to remove O2 and alleviate the produc-

tion of this problematic molecule [41�].

Early fortuitous antioxidant systemsManganese(II) and reactive oxygen species: In modern aerobic

organisms, antioxidant enzymes such as superoxide dis-

mutases (SOD), superoxide reductases (SOR), catalases

and peroxidases, play a crucial role in defending cells

from superoxide, hydrogen peroxide, and other perox-

ides. Based on their widespread distribution and phylo-

genetic relationships, the iron-containing SOD, FeSOD,

appears to be the earliest known form of the SODs, and

FeSOR likewise appears to be quite ancient, but there is

no evidence that it was in place when the earliest Oxy-photobacteria first experienced O2 [42]. The earliest cata-

lase appears to be the manganese catalase [43], and the

earliest heme-containing peroxidase known is the short


peroxicin found in crown group Oxyphotobacteria including

the basal genus Gloeobacter [44�]. Whether or not the

precursors of these enzymes were present in the earliest

Oxyphotobacteria, it is likely that these cells contained

relatively high levels of Mn2+, which is itself competent

to catalyze both the SOD and the catalase reactions [45].

The early oceans contained relatively high concentrations

of both Fe2+ and Mn2+, and it is possible that significant

concentrations of both of these cations were also present

intracellularly in early Oxyphotobacteria. The reactivity of

Fe2+ and Mn2+ with superoxide and hydrogen peroxide

are entirely different, with the result that iron functions as

a prooxidant and manganese as an antioxidant [46]. Fe2+

reacts with hydrogen peroxide via the Fenton reaction,

forming highly toxic hydroxyl radicals and Fe3+ as pro-

ducts. Superoxide can act to reduce Fe3+ back to Fe2+,

and both reactions combined constitute the so-called iron

catalyzed Haber-Weiss reaction (Reaction 1). By contrast,

Mn2+ does not do Fenton-type chemistry but instead

catalyzes disproportionation (dismutation) of either su-

peroxide (Reaction 2) or hydrogen peroxide (Reaction 3).

In addition, Fe2+, but not Mn2+, reacts readily with O2 to

produce Fe3+.

O�2 þ H2O2 �!Fe2þ=Fe3þ

OH� þ OH� (1)

2O�2 þ 2Hþ�!Mn2þO2 þ H2O2 (2)

2H2O2�!Mn2þ

O2 þ 2H2O (3)

It has been shown repeatedly that the O2-sensitivity of

cells entirely missing superoxide dismutase enzymes can

be completely rescued when manganese levels are high,

levels well within that observed in early seawater

(Figure 1) [45,47��,48].

Manganese also offered another mode of protection.

Iron(II)-containing non-redox enzymes are frequently

inactivated by superoxide or hydrogen peroxide which

oxidize the metal center to iron(III) which then dissoci-

ates as Fe3+, leaving behind the inactive apoprotein. If the

iron(II) is replaced by manganese(II), such enzymes

frequently retain a large fraction of their enzymatic activ-

ities and they are relatively resistant to oxidative damage

[46,49,50].

Iron and sulfur: Modern organisms, both aerobic and

anaerobic, rely upon numerous iron–sulfur cluster-con-

taining proteins [37], and it has recently become apparent

their numbers may have been seriously underestimated

in the past [51�]. Iron–sulfur clusters are generally quite

sensitive to oxidants, which oxidize the clusters and cause

them to fall out of the proteins. The fact that this ancient

type of metalloproteins survived the rise of O2 in the



atmosphere is a testament to their importance to all forms

of life.

There are several different complex pathways known for

the in vivo assembly of iron–sulfur clusters and their

incorporation into proteins, but the earliest of these

appears to be the Suf system [52��]. When it occurs in

modern aerobic organisms, the Suf system is highly

complex, consisting of multiple components, and it relies

upon cysteine rather than inorganic sulfide, S2�, as its

source of sulfur. But it is possible to deduce that in some

anaerobic organisms, the minimum functional unit is

SufB plus SufC — the first is an iron–sulfur scaffold

protein and the second is an ATPase. This minimal

system is expected to be sufficient for iron–sulfur cluster

assembly on the scaffold and insertion into the apoen-

zyme, but only in the case where abundant Fe(II) and

inorganic sulfide were readily available, as would have

been the case in early organisms [52��].

Inorganic sulfide (S2�, HS�, or H2S) was relatively abun-

dant in early anaerobic cells, but the concentrations are

very low in modern aerobic cells, and instead cysteine acts

as a source of inorganic sulfur when it is needed. But other

sulfur-containing compounds, in particular relatively con-

centrated pool of thiols, maintain their importance, acting

as redox buffers in modern intracellular biochemistry, and

it is likely that the anaerobic ancestors of the Oxyphoto-bacteria and other microbial groups used thiols in a similar

fashion prior to the GOE [35��].

In addition to inorganic sulfide, another big difference to be

expected between pre-GOE cells and their modern des-

cendants lies in the concentration of freely available Fe(II),

since early cells would have had little need to sequester

available iron sources (see discussion of ferritins below).

Photosynthetic cells require a large amount of iron, and we

can assume that this was true also of the first phototrophic

Oxyphotobacteria; iron would have been abundant and pres-

ent as Fe(II), just as both sulfide and thiols are likewise

expected to have been abundant. Fe(II) would have readi-

ly formed Fe(II)–thiolate complexes in addition to Fe(II)–sulfide complexes, using as ligands free cysteine, cysteine

side chains on proteins, and whatever other thiolates were

present. When O2 first appeared, the Fe(II)–sulfide and

Fe(II)–thiolate complexes would have been rapidly oxi-

dized to Fe(III). Sulfide ligands are likely to have been

oxidized to elemental sulfur as well as polysulfides, sulfite,

thiosulfate, and sulfate [34], and thiolates would have been

oxidized to disulfides, cysteine to cystine, for example, or to

sulfinic acids (RSO2H) or sulfonic acids (RSO3H) [35��].Any reactive oxygen species formed, such as superoxide or

hydrogen peroxide would likewise be rapidly scavenged by

this Fe(II)–sulfide system.

Thus in early anaerobic Oxyphotobacteria, very small

amounts of O2 were not likely to be catastrophic so long


as Fe(II), thiols, and sulfide remained abundant and did

not become limiting; rapid scavenging of small amounts

of O2 by the Fe(II)–thiolate plus the Fe(II)–sulfide

system could have removed O2 and any resulting H2O2

temporarily from the intracellular compartment. With

higher O2 fluxes, however, excessive consumption of

sulfide and thiolate ligands and the accumulation of

Fe(III) would have certainly been toxic to these anaero-

bic cells.

Greater amounts of O2 could have been tolerated if these

new Fe(II)–sulfide and Fe(II)–thiolate antioxidant sys-

tems were catalytic. Early members of the Oxyphotobac-teria may have been capable of recycling sulfite and

sulfate using reductase enzymes [53], and the Fe(III)

formed could have been reduced by the excess sulfide,

making the Fe(II)–sulfide antioxidant system catalytic. In

addition, the enzyme thioredoxin may have acted in a

similar fashion in the Fe(II)–thiolate antioxidant system

by catalyzing reduction of disulfide-containing species.

Thioredoxin is a class of small proteins present in most or

all cells that facilitates thiol-disulfide exchange reactions

on other proteins [54,55]. In combination with thiore-

doxin reductase, a flavoprotein that uses NADPH to keep

thioredoxin in its reduced functional state [93], this

system is used to return cysteine side chains of many

proteins to their reduced states. The recent discovery of a

functional thioredoxin system in the strictly anaerobic

methanogens, presumably to facilitate thiol-disulfide ex-

change reactions, suggests that thioredoxins may have

been present in cells long before the rise of O2. This is

important because it implies that thioredoxins may have

already been present in early Oxyphotobacteria and other

coeval microbial groups, where they could have contrib-

uted to the rapid development of a thiol-disulfide antiox-

idant system by catalyzing the reduction of the disulfides

thus returning them to their reduced states [56�].

The modern antioxidant most related to this primitive

Fe(II)-sulfide-thiol antioxidant systems is glutathione, an

antioxidant molecule that appears to have originated in

Oxyphotobacteria [57]. Prior to the advent of glutathione,

cysteine and then g-Glu-Cys probably were used as

antioxidants [35��]. One enormous advantage of g-Glu-

Cys and glutathione over cysteine is that their metal

complexes are much more slowly oxidized by O2 than

most other metal-thiol complexes [35��]. These other

thiol complexes would have been rapidly depleted by

higher O2 levels via redox metal-catalyzed reactions. The

evolution of glutathione was enormously important be-

cause it allowed for the sequestration of Fe(II), and later

Cu(I), as glutathione–metal complexes. These complexes

resisted rapid oxidation by O2 and thus protected the

other thiols present from redox metal-catalyzed destruc-

tion by O2. It is interesting in this regard to note that the

‘labile iron pool’ present in modern eukaryotic cells

consists primarily of Fe(II)–glutathione [58]; cysteine



could never have functioned in this manner in aerobic

cells because Fe(II)–cysteine complexes would be rapid-

ly oxidized by O2 [35��].

Sequestration of iron(III) — ferritin and polyphosphate: Dur-

ing early encounters with O2, microbes may have relied

upon sulfide and thiolates as antioxidants, but only at

great cost to the budget of sulfides, thiols, and other

reducing agents present in cells. Moreover, the cellular

redox potentials would have become more oxidizing as

these cells evolved an aerobic lifestyle, until ultimately

the soluble Fe(II) would have been converted to insolu-

ble Fe(III)oxyhydroxides. The modern solution to this

problem is the ferritin family of proteins: ferritin, bacter-

ioferritin, and the Dps proteins [59–61]. These proteins

are soluble hollow polypeptide spheres that are reversibly

filled with nanospheres of hydrated ferric oxides (ferrihy-

drite). Ferritin proteins function as iron storage, but they

also act as antioxidants. The antioxidant action of ferritins

is probably best illustrated by the Dps proteins. These act

as enclosed nanoreactors in which Fe(II) is oxidized in

two steps, first by O2 producing hydrogen peroxide, but

then even faster by hydrogen peroxide without releasing

any reactive oxygen species in the process [62–64].

An additional strategy that early aerobic cells may have

used to sequester Fe(III) might have been to bind it to

polyphosphate, the synthesis of which is up-regulated in

bacteria in response to oxidative stress [65�]. Coordination

of Fe(III) to polyphosphate stabilizes it in that oxidation

state, and thus inhibits its ability to catalyze Fenton

chemistry [66].

Quinols: Lipid peroxidation also posed a new kind of

threat to the membranes of the ancestral microbes, par-

ticularly if those membranes contained unsaturated

lipids. The membranes of modern aerobic cells would

be susceptible to oxidative damage in the presence of O2

were in not for the presence of quinols such as tocopherols

and ubiquinol, which act as chain-breakers of the free

radical autoxidation process that peroxidizes lipids. Qui-

nols such as menaquinol were already present in early

anoxygenic photosynthetic cells, where they played im-

portant functions as redox-active molecules in electron

transport chains [67–70]. Such quinols would have been

poised to act as inhibitors of free radical autoxidation of

the membrane components.

The modern widespread quinol antioxidant system found in

aerobic cells is alpha-tocopherol (vitamin E), which is only

synthesized by oxygenic photosynthetic organisms and

must be consumed in the diets of humans and other animals.

It is interesting to note that, just like glutathione, this system

appears to have first arisen in Oxyphotobacteria [71,72].

Carotenoids and singlet O2: Early Oxyphotobacteria encoun-

tered for the first time not only ground state triplet O2 but


also highly reactive photoexcited singlet O2 [73]. Fortu-

nately carotenoid pigments, which are excellent singlet

O2 quenchers, were already part of the photosynthetic

apparatus in anoxygenic phototrophs [74]. In modern

Oxyphotobacteria, this role is played by the orange carot-

enoid protein (OCP) [75].

Evolution of advanced antioxidant systems inOxyphotobacteriaFlavodiiron proteins: Flavodiiron proteins (FDPs) are oxi-

doreductases that are predominantly found in Oxyphoto-bacteria and anaerobes, where they reduce NO and O2 to

N2O and H2O respectively, with reducing equivalents

from NAD(P)H, or, in the case of methanogens, F420H2

[78,80]. All FDPs contain two conserved structural

domains: an N-terminus metallo-b-lactamase-like do-

main that contains a non-heme Fe–Fe center, and a C-

terminus flavodoxin-like domain. Crystal structures re-

veal that FDPs form a homodimeric head-to-tail arrange-

ment that places the FMN moiety of one subunit near the

diiron site of the metallo-b-lactamase-like domain of the

other for efficient electron transfer during substrate re-

duction [76–79]. While all FDPs have this conserved

flavodiiron compound domain structure, some have gene

fusions of additional redox domains at their C-termini

(Figure 4) [80].

In non-phototrophic organisms, FDPs appear to function

in NO and O2 detoxification, commonly showing a pref-

erence for one molecule or the other [80,81]. FDPs

provide a biochemically efficient means of removing

O2 and alleviating oxidative stress. For example, the

anaerobic protist Giardia lamblia lacks respiratory O2

reductases and does not contain the typical suite of

enzymes for dealing with ROS, but it does contain an

FDP with a high affinity for O2 [82]. FDPs found in

methanogens also display a strong preference for O2 [78].

Interestingly, FDPs are present in many paralogous

copies (typically between two and six) in the genomes

of all extant Oxyphotobacteria, except for losses in non-

phototrophic symbionts [83]. These form both a mono-

phyletic clade of FDPs and an independent class marked

by the fusion of a flavodoxin-like (RutF) domain at the C-

terminus (Figure 4). This phylogenetic distribution

implies a rich evolutionary history of this family of

proteins in the Oxyphotobacteria and suggests that FDPs

were important during the evolution of oxygenic photo-

synthesis. The exact function and physiological utility of

the different flavors of FDPs found in Oxyphotobacteriaare still not well known [41�]. In recombinant studies,

FDPs from Oxyphotobacteria are capable of direct reduc-

tion of O2 to water without the production of ROS [80].

Two genes encoding FDPs, Flv1 and Flv3, promote a

Mehler-like reaction in Oxyphotobacteria, modulating the

photoreduction of O2 with electrons downstream of PSI

[84]. Unlike the Mehler reaction in plants and algae,



Figure 4

non-redox Fe(II) diiron (or Zn) hydrolase

Fe(III) diironoxidoreductase diiron

flavoproteinO2 reductase

complex

flavodiironO2 reductasehomodimer

C-terminaldiversification(other taxa)

flavodiironNAD(P)H-O2 oxidoreductase

homodimer

fusion offlavodoxin

gene duplication &paralogous evolution

environmental O2 (~2.4 Ga) crown group Oxyphotobacteria (< 2.0 Ga)

fusion

flavodiironflavodoxin-O2

oxidoreductasecomplex

crown group FDPs

0.1

root?

ABCD

(b)

Fe Fe

FMN

Rd

Flv

O2 , NO H2O , N 2O

NADH NAD+

Class A

Fe Fe

FMN

Rd

Flv

O2 H2O

NADH NAD+

Class B

Fe Fe

FMN

Rd

O2 H2O

NAD(P)H NADP+

Flv

Class D

Fe Fe

FMN

Flv

O2 H2O

NAD(P)H NADP+

Class COxyphotobacteria

(a)

(c)


(a) Ferric diiron proteins (O2 and NO reductases) comprise a large, and somewhat modular, protein family commonly classified by their domain

structure and diversity of fusions of different redox domains that occur at their C terminus. Cartoons highlight domain structure of the difference

classes. (b) Phylogenetic relationships of ferric diiron proteins (FDPs) constructed using sequence alignments of their common flavodiiron core.

The classes with additional C terminal fusions are well captured by clades in the phylogentic relationships, with positions derived within the

structurally simple Class A FDPs. Note that the fusion of rubredoxin domains appear to have occurred twice independently each in the Class B

and Class D FDPs. These proteins are rather sparsely and variably distributed among Bacteria, Archaea, and a number of eukaryotic protists —

mainly in anaerobes. Importantly all Oxyphotobacteria contain FDPs, often many paralogous copies, where these proteins play an important role in

cell redox balance, removing O2, and protecting photosystems against singlet oxygen. These proteins in Oxyphotobacteria form a diverse clade

highlighting their importance in evolutionary history within the group. They have a unique domain structure with a fusion of a flavodoxin domain,

and they directly oxidize NAD(P)H. These proteins are responsible for a tremendous O2 reduction flux — up to 40% of photosynthetically produced

O2 — in modern Oxyphotobacteria [85]. (c) Evolutionary hypothesis for the origin and early evolution of FDPs in Oxyphotobacteria. Currently there

is no solved crystal structure of a Class C FDP from the Oxyphotobacteria, and the orientation and placement of the C-terminal flavodoxin domain

remains uncertain; it could feed electrons into either active site in the dimer. We view FDPs as important proteins in stem group Oxyphotobacteria

because they would have allowed for the maintenance of low cellular O2 levels — in a sense buying time to explore and test a range of possible

antioxidant systems to cope with aerobic stress.

however, this FDP-dependent reaction in Oxyphotobacteriadoes not produce ROS. From observations of the differen-

tial mass law fractionations of multiple oxygen isotopes

(16O, 17O, 18O), Helman et al. [85] estimated that as much as

40% of the O2 leaving PSII was re-reduced by FDPs in

photosynthetically active cells, compared with only 6%

going to respiration at Complex IV. The amount of O2 that

flows to FDPs appears to depend on the availability of CO2

and light and is particularly important during fluctuating

light conditions, providing a harmless mechanism to

maintain cellular redox balance by dissipating excess elec-

trons downstream of PSI [86]. Other FDPs in Oxyphoto-bacteria appear to play roles in photoprotection of PSII

against 1O2 and may function as heterodimers [84,87]; yet

others appear to provide heterocysts with anaerobic envir-

onments conducive to nitrogen fixation [41�]. Altogether,


the diversity and omnipresence of FDPs in Oxyphotobac-teria supports the view that the FDPs were integral to the

development of oxygenic photosynthesis [41�].

Several aspects of the structural biology and biochemistry

of FDPs permit one to develop a hypothesis for the

evolution of this protein family as an important comple-

ment to oxygenic photosynthesis (Figure 4). Because all

FDPs have in common a conserved two-domain struc-

ture, one might infer that they result from the fusion of

proteins that once had little to do with one another.

Metallo-b-lactamase-like domains appear in proteins

with diverse functions, and on the basis of their distribu-

tion in metabolism and the tree of life appear to be

exceptionally old [77]. We envision this domain originally

contained a non-redox metal center comprised of either



Fe(II) or Zn(II), and probably functioned as a hydrolytic

enzyme. With early photosynthetic O2 fluxes, the metal

center became a ferric Fe-Fe site with a high affinity for

O2, and formed an early O2 reductase complex in part-

nership with the flavoprotein. Due to the distances be-

tween the FMN moiety and the diiron site (40 A), this

complex could only function as an O2 reductase in the

configuration of a multi-subunit homodimer. The FDPs

family, thus, was born from the fusion of these two

interacting domains that were probably organized togeth-

er as genes within an operon. The radiation of the FDPs

occurred, providing a useful mechanism for many anaer-

obic microorganisms to detoxify O2. This co-occured with

the functional diversification of the C-terminus in a

number of different groups, with lateral gene transfer

playing an important role in the extant distribution of

FDPs [80]. In stem group Oxyphotobacteria, the fusion of

another flavodoxin domain occurred before the adaptive

radiation of the many different FDPs found within them,

because this fusion is observed in all members of crown

group Oxyphotobacteria. Thus based on the phylogenetic

distribution of FDPs in Oxyphotobacteria, their general

absence or evolutionary incongruence in the Melainabac-teria, and their diverse roles in protecting cells against O2

stresses, we hypothesize that FDPs played an key role in

allowing stem group Oxyphotobacteria to complete early

experiments in phototrophic O2 production.

Ascorbate was not an ancient antioxidant: The ascorbate–ascorbate peroxidase antioxidant system, which is a major

antioxidant system in plants and most animals, is absent

in Oxyphotobacteria. In fact, it seems likely that prokar-

yotes in general do not synthesize ascorbate [88]. This is

in sharp contrast to two other very important modern

antioxidant systems, glutathione and alpha-tocopherol,

both of which appear to have originated in the Cyano-bacteria phylum and were then widely spread to other

species.

ConclusionsThe evolution of oxygenic photosynthesis in Oxyphoto-bacteria led to the GOE and, along with it, to the greatest

interval of environmental change in Earth history. In turn,

O2 fed back on biology with both substantial risk and

substantial reward. While the availability of O2 provided

life with new bioenergetic opportunities, it also produced

significant oxidative stress. Virtually all modern cells —

including strict anaerobes — rely upon diverse mecha-

nisms for coping with oxidative stress, but most of these

systems would not have been in place when O2 first

appeared. Nevertheless, there were likely a number of

preexisting biochemical systems and inorganic reactions

that had fortuitous antioxidant chemistry, which with

time enabled the evolution of more complex enzymatic

antioxidant systems. Oxyphotobacteria provide an interest-

ing test case for understanding how life survived the

GOE, because these cells were the first to deal with large


fluxes of intracellular dioxygen. Importantly, manganese

appears to have played major roles in both the production

and early detoxification of O2. Current data also suggest

that the Oxyphotobacteria were important incubators for a

wide variety of solutions to deal with oxidative stress (like

glutathione, alpha-tocopherol, and FDPs).

Going forward to better understand how life responded to

the GOE, we advocate greater focus on non-enzymatic

solutions that were likely present in early cells, because it

is possible that many of the enzymatic systems important

today were not available for biology at this time. We think

that efforts using comparative biology to infer the relative

timing of different antioxidant systems will continue to

provide useful insight into this problem. This is a good

time to be asking these questions because the landscape

for understanding microbial and metabolic evolution is

rapidly changing, enabled by new single-cell and meta-

genomic sequencing technologies, which probe microbial

diversity deeply and make it possible to populate evolu-

tionary analyses with substantial amounts of sequence

data. We think this approach is promising and will enable

integrative hypothesis generation and testing as new

groups of microbes are discovered and characterized.

Conflict of interestThe authors declare no conflict of interest.

AcknowledgementsWe thank Usha Lingappa, Hope Johnson, Jena Johnson, Dianne Newman,and two anonymous reviewers for helpful feedback on ideas synthesized inthis paper. We acknowledge support from a David and Lucile PackardFoundation Fellowship in Science and Engineering (WWF), and theAgouron Institute (JH and WWF).

References and recommended readingPapers of particular interest, published within the period of review,have been highlighted as:

� of special interest�� of outstanding interest

1. Bell EA, Boehnke P, Harrison TM, Mao WL: Potentially biogeniccarbon preserved in a 4.1 billion-year-old zircon. Proc NatlAcad Sci U S A 2015, 112:14518-14521.

2. Rosing MT: 13C-Depleted carbon microparticles in >3700-Masea-floor sedimentary rocks from west Greenland. Science1999, 283:674-676.

3. Javaux EJ, Marshall CP, Bekker A: Organic-walled microfossilsin 3.2-billion-year-old shallow-marine siliciclastic deposits.Nature 2010, 463:934-938.

4. Sugitani K, Mimura K, Nagaoka T, Lepot K, Takeuchi M:Microfossil assemblage from the 3400Ma Strelley PoolFormation in the Pilbara Craton, Western Australia: resultsform a new locality. Precambrian Res 2013, 226:59-74.

5. Bontognali TRR, Sessions AL, Allwood AC, Fischer WW,Grotzinger JP, Summons RE, Eiler JM: Sulfur isotopes of organicmatter preserved in 3.45-billion-year-old stromatolites revealmicrobial metabolism. Proc Natl Acad Sci U S A 2012,109:15146-15151.

6. Slotznick SP, Fischer WW: Examining Archean methanotrophy.Earth Planet Sci Lett 2016, 441:52-59.


http://refhub.elsevier.com/S1367-5931(16)30020-5/sbref0005





















7. Tice MM, Lowe DR: Photosynthetic microbial mats in the 3416-Myr-old ocean. Nature 2004, 431:549-552.

8. Ueno Y, Yamada K, Yoshida N, Maruyama S, Isozaki Y: Evidencefrom fluid inclusions for microbial methanogenesis in the earlyArchaean era. Nature 2006, 440:516-519.

9.��

Imlay JA: The molecular mechanisms and physiologicalconsequences of oxidative stress: lessons from a modelbacterium. Nat Rev Microbiol 2013, 11:443-454.

Comprehensive review of the challenges and solutions for cells dealingwith O2, from the biology of the facultative anaerobe Escherichia coli asperhaps the best understood model organism.

10.�

Fischer WW, Hemp J, Johnson JE: Evolution of oxygenicphotosynthesis. Annu Rev Earth Planet Sci 2016, 44 in press.

A comprehensive review of the evolution of phototrophy and oxygenicphotosynthesis, from both geological and comparative biological obser-vations. Importantly, oxygenic photosynthesis is a relatively late inventionin the context of the history of life.

11. Johnson JE, Gerpheide A, Lamb MP, Fischer WW: O2 constraintsfrom Paleoproterozoic detrital pyrite and uraninite. Geol SocAm Bull 2014, 126:813-830.

12. Rasmussen B, Bekker A, Fletcher IR: Correlation ofPaleoproterozoic glaciations based on U–Pb zircon ages fortuff beds in the Transvaal and Huronian Supergroups. EarthPlanet Sci Lett 2013, 382:173-180.

13. Rye R, Holland HD: Paleosols and the evolution of atmosphericoxygen: a critical review. Am J Sci 1998, 298:621-672.

14. Johnson JE, Webb SM, Ma C, Fischer WW: Manganesemineralogy and diagenesis in the sedimentary rock record.Geochim Cosmochim Acta 2016, 173:210-231.

15. Cameron EM: Sulphate and sulphate reduction in earlyPrecambrian oceans. Nature 1982, 296:145-148.

16. Crichton RR, Pierre JL: Old iron, young copper: from Mars toVenus. Biometals Int J Role Met Ions Biol Biochem Med 2001,14:99-112.

17. Rickaby REM: Goldilocks and the three inorganic equilibria:how Earth’s chemistry and life coevolve to be nearly in tune.Philos Trans R Soc Math Phys Eng Sci 2015, 373 20140188-20140188.

18. Shields G, Veizer J: Precambrian marine carbonate isotopedatabase: version 1.1: CARBONATE ISOTOPE DATABASE.Geochem Geophys Geosystems 2002, 3 1 of 12-12 of 12.

19. Mucci A: Manganese uptake during calcite precipitation fromseawater: conditions leading to the formation of apseudokutnahorite. Geochim Cosmochim Acta 1988, 52:1859-1868.

20. Morgan JJ: Kinetics of reaction between O2 and Mn(II) speciesin aqueous solutions. Geochim Cosmochim Acta 2005, 69:35-48.

21. Liu X-M, Kah LC, Knoll AH, Cui H, Kaufman AJ, Shahar A,Hazen RM: Tracing Earth’s O2 evolution using Zn/Fe ratios inmarine carbonates. Geochem Perspect Lett 2016, 2:24-34.

22. Robbins LJ, Lalonde SV, Saito MA, Planavsky NJ,Mloszewska AM, Pecoits E, Scott C, Dupont CL, Kappler A,Konhauser KO: Authigenic iron oxide proxies for marine zincover geological time and implications for eukaryoticmetallome evolution. Geobiology 2013, 11:295-306.

23. Raymond J, Segre D: The effect of oxygen on biochemicalnetworks and the evolution of complex life. Science 2006,311:1764-1767.

24. Baughn AD, Malamy MH: The strict anaerobe Bacteroidesfragilis grows in and benefits from nanomolar concentrationsof oxygen. Nature 2004, 427:441-444.

25.�

Di Rienzi SC, Sharon I, Wrighton KC, Koren O, Hug LA,Thomas BC, Goodrich JK, Bell JT, Spector TD, Banfield JF et al.:The human gut and groundwater harbor non-photosyntheticbacteria belonging to a new candidate phylum sibling toCyanobacteria. eLife 2013, 2:e01102.

First description of the Melainabacteria as a non-phototrophic, closesister clade of the oxygenic Cyanobacteria from gut and groundwatermetagenomic data.


26.��

Soo RM, Skennerton CT, Sekiguchi Y, Imelfort M, Paech SJ,Dennis PG, Steen JA, Parks DH, Tyson GW, Hugenholtz P: Anexpanded genomic representation of the phylumcyanobacteria. Genome Biol Evol 2014, 6:1031-1045.

This study presents a new classification of the Cyanobacteria phylum toinclude diverse members of the Melainabacteria based a number ofmetagenomic datasets. The members capable of oxygenic photosynth-esis now constitute the class Oxyphotobacteria.

27. Soo RM, Woodcroft BJ, Parks DH, Tyson GW, Hugenholtz P: Backfrom the dead; the curious tale of the predatorycyanobacterium Vampirovibrio chlorellavorus. PeerJ 2015,3:e968.

28. Johnson JE, Webb SM, Thomas K, Ono S, Kirschvink JL,Fischer WW: Correcting mistaken views of sedimentarygeology, Mn-oxidation rates, and molecular clocks. Proc NatlAcad Sci U S A 2013, 110:E4119-E4120.

29.��

Johnson JE, Webb SM, Thomas K, Ono S, Kirschvink JL,Fischer WW: Manganese-oxidizing photosynthesis before therise of cyanobacteria. Proc Natl Acad Sci U S A 2013, 110:11238-11243.

This work presents the first suite of geological observations that Mn2+

oxidation preceded the GOE. This scenario was predicted by severalevolutionary hypotheses for the origin of the WOC of PSII, wherein Mn-oxidizing phototrophy was an evolutionary intermediate during the devel-opment of high-potential photosynthesis and light-driven water-splitting.

30. Fischer WW, Hemp J, Johnson JE: Manganese and the evolutionof photosynthesis. Orig Life Evol Biosphere J Int Soc Study OrigLife 2015, 45:351-357.

31. Hemp J, Lucker S, Schott J, Pace LA, Johnson JE, Schink B,Daims H, Fischer WW: Genomics of a phototrophic nitriteoxidizer: insights into the evolution of photosynthesis andnitrification. ISME 2016. in press.

32. Ward LM, Kirschvink JL, Fischer WW: Timescales ofoxygenation following the evolution of oxygenicphotosynthesis. Orig Life Evol Biospheres 2015. in press.

33.��

Kihara S, Hartzler DA, Savikhin S: Oxygen concentration inside afunctioning photosynthetic cell. Biophys J 2014, 106:1882-1889.

This paper presents a set of theoretical calculations for the dioxygenconcentrations inside photosynthetically active Oxyphotobacteria cells.Interestingly, in solitary planktonic cells the steady-state O2 concentra-tions are only mildly higher than that of the external environment.

34. Li Q, Lancaster JR: Chemical foundations of hydrogen sulfidebiology. Nitric Oxide Biol Chem Off J Nitric Oxide Soc 2013,35:21-34.

35.��

Fahey RC: Glutathione analogs in prokaryotes. Biochim BiophysActa 2013, 1830:3182-3198.

This paper presents a comprehensive synthesis of the different thiol-based protective systems found in Bacteria and Archaea, along with theargument that thiols would have been broadly distributed among micro-organisms prior to the GOE, and valuable as early antioxidants.

36. Chaiyen P, Fraaije MW, Mattevi A: The enigmatic reaction offlavins with oxygen. Trends Biochem Sci 2012, 37:373-380.

37. Imlay JA: Iron–sulphur clusters and the problem with oxygen.Mol Microbiol 2006, 59:1073-1082.

38. Vass I, Styring S, Hundal T, Koivuniemi A, Aro E, Andersson B:Reversible and irreversible intermediates duringphotoinhibition of photosystem II: stable reduced QA speciespromote chlorophyll triplet formation. Proc Natl Acad Sci U S A1992, 89:1408-1412.

39. Gorman AA, Rodgers MA: Current perspectives of singletoxygen detection in biological environments. J PhotochemPhotobiol B 1992, 14:159-176.

40. Latifi A, Ruiz M, Zhang C-C: Oxidative stress in cyanobacteria.FEMS Microbiol Rev 2009, 33:258-278.

41.�

Allahverdiyeva Y, Isojarvi J, Zhang P, Aro E-M: Cyanobacterialoxygenic photosynthesis is protected by flavodiiron proteins.Life Basel Switz 2015, 5:716-743.

This work reviews what is currently known about the distribution anddiverse functions of FDPs in Oxyphotobacteria.




















































































































42. Sheng Y, Abreu IA, Cabelli DE, Maroney MJ, Miller A-F, Teixeira M,Valentine JS: Superoxide dismutases and superoxidereductases. Chem Rev 2014, 114:3854-3918.

43. Zamocky M, Gasselhuber B, Furtmuller PG, Obinger C: Molecularevolution of hydrogen peroxide degrading enzymes. ArchBiochem Biophys 2012, 525:131-144.

44.�

Zamocky M, Hofbauer S, Schaffner I, Gasselhuber B, Nicolussi A,Soudi M, Pirker KF, Furtmuller PG, Obinger C: Independentevolution of four heme peroxidase superfamilies. ArchBiochem Biophys 2015, 574:108-119.

This paper presents molecular evolution analyses and a nice synthesis ofthe biochemistry of the heme peroxidase superfamilies — proteins that areimportant because they catalyze a number of high-potential oxidationreactions using hydrogen peroxide. This analysis highlights Oxyphotobac-teria as important innovators in the heme peroxidases and also shows thatMn-catalase appears to be a relatively late addition to the superfamily.

45. Barnese K, Gralla EB, Valentine JS, Cabelli DE: Biologicallyrelevant mechanism for catalytic superoxide removal bysimple manganese compounds. Proc Natl Acad Sci U S A 2012,109:6892-6897.

46. Latour J-M: Manganese, the stress reliever. Met Integr BiometalSci 2015, 7:25-28.

47.��

Culotta VC, Daly MJ: Manganese complexes: diverse metabolicroutes to oxidative stress resistance in prokaryotes and yeast.Antioxid Redox Signal 2013, 19:933-944.

Review of the current state of knowledge of low molecular weightmanganous species functioning as antioxidants in vivo.

48. Sanchez RJ, Srinivasan C, Munroe WH, Wallace MA, Martins J,Kao TY, Le K, Gralla EB, Valentine JS: Exogenous manganousion at millimolar levels rescues all known dioxygen-sensitivephenotypes of yeast lacking CuZnSOD. J Biol Inorg Chem JBICPubl Soc Biol Inorg Chem 2005, 10:913-923.

49. Imlay JA: The mismetallation of enzymes during oxidativestress. J Biol Chem 2014, 289:28121-28128.

50. Kaushik MS, Srivastava M, Verma E, Mishra AK: Role ofmanganese in protection against oxidative stress under ironstarvation in cyanobacterium Anabaena 7120. J Basic Microbiol2015, 55:729-740.

51.�

Rouault TA: Iron–sulfur proteins hiding in plain sight. Nat ChemBiol 2015, 11:442-445.

This recent commentary highlights how tricky it can be to identify proteinsbearing Fe–S clusters with different approaches, with the implication thatthese proteins are probably more common and important than previouslythought.

52.��

Boyd ES, Thomas KM, Dai Y, Boyd JM, Outten FW: Interplaybetween oxygen and Fe–S cluster biogenesis: insights fromthe Suf pathway. Biochemistry (Mosc.) 2014, 53:5834-5847.

This work contains a nice synthesis of the mechanisms used for thegenesis of iron–sulfur clusters, presented in an evolutionary context ofchanges in Suf system engendered by the rise of dioxygen.

53. Grein F, Ramos AR, Venceslau SS, Pereira IAC: Unifyingconcepts in anaerobic respiration: insights from dissimilatorysulfur metabolism. Biochim Biophys Acta 2013, 1827:145-160.

54. Ingles-Prieto A, Ibarra-Molero B, Delgado-Delgado A, Perez-Jimenez R, Fernandez JM, Gaucher EA, Sanchez-Ruiz JM,Gavira JA: Conservation of protein structure over four billionyears. Struct Lond Engl 1993 2013, 21:1690-1697.

55. Lu J, Holmgren A: The thioredoxin antioxidant system. FreeRadic Biol Med 2014, 66:75-87.

56.�

Susanti D, Wong JH, Vensel WH, Loganathan U, DeSantis R,Schmitz RA, Balsera M, Buchanan BB, Mukhopadhyay B:Thioredoxin targets fundamental processes in a methane-producing archaeon, Methanocaldococcus jannaschii. ProcNatl Acad Sci U S A 2014, 111:2608-2613.

This report describes the presence and function of thioredoxins in strictlyanaerobic methanogens, suggesting that these proteins may have oper-ated in the absence of O2 to regulate metabolism long before the GOE.

57. Musgrave WB, Yi H, Kline D, Cameron JC, Wignes J, Dey S,Pakrasi HB, Jez JM: Probing the origins of glutathionebiosynthesis through biochemical analysis of glutamate-cysteine


ligase and glutathione synthetase from a model photosyntheticprokaryote. Biochem J 2013, 450:63-72.

58. Hider RC, Kong X: Iron speciation in the cytosol: an overview.Dalton Trans Camb Engl 2003 2013, 42:3220-3229.

59. Andrews SC: The Ferritin-like superfamily: evolution of thebiological iron storeman from a rubrerythrin-like ancestor.Biochim Biophys Acta 2010, 1800:691-705.

60. Bradley JM, Moore GR, Le Brun NE: Mechanisms of ironmineralization in ferritins: one size does not fit all. J Biol InorgChem JBIC Publ Soc Biol Inorg Chem 2014, 19:775-785.

61. Theil EC, Behera RK, Tosha T: Ferritins for chemistry and for life.Coord Chem Rev 2013, 257:579-586.

62. Calhoun LN, Kwon YM: Structure, function and regulation of theDNA-binding protein Dps and its role in acid and oxidativestress resistance in Escherichia coli: a review. J Appl Microbiol2011, 110:375-386.

63. Haikarainen T, Papageorgiou AC: Dps-like proteins: structuraland functional insights into a versatile protein family. Cell MolLife Sci 2010, 67:341-351.

64. Santos SP, Mitchell EP, Franquelim HG, Castanho MARB,Abreu IA, Romao CV: Dps from Deinococcus radiodurans:oligomeric forms of Dps1 with distinct cellular functions andDps2 involved in metal storage. FEBS J 2015, 282:4307-4327.

65.�

Gray MJ, Jakob U: Oxidative stress protection bypolyphosphate — new roles for an old player. Curr OpinMicrobiol 2015, 24:1-6.

This paper highlights the (likely ancient) role of polyphosphate in reducingoxidative stress as a chaperone and metal chelator in diverse groups oforganisms.

66. Rachmilovich-Calis S, Masarwa A, Meyerstein N, Meyerstein D:The effect of pyrophosphate, tripolyphosphate and ATP on therate of the Fenton reaction. J Inorg Biochem 2011, 105:669-674.

67. Dibrova DV, Cherepanov DA, Galperin MY, Skulachev VP,Mulkidjanian AY: Evolution of cytochrome bc complexes: frommembrane-anchored dehydrogenases of ancient bacteria totriggers of apoptosis in vertebrates. Biochim Biophys Acta2013, 1827:1407-1427.

68. Kao W-C, Hunte C: The molecular evolution of the Qo motif.Genome Biol Evol 2014, 6:1894-1910.

69. Schoepp-Cothenet B, Lieutaud C, Baymann F, Vermeglio A,Friedrich T, Kramer DM, Nitschke W: Menaquinone as poolquinone in a purple bacterium. Proc Natl Acad Sci U S A 2009,106:8549-8554.

70. Schoepp-Cothenet B, van Lis R, Atteia A, Baymann F, Capowiez L,Ducluzeau A-L, Duval S, ten Brink F, Russell MJ, Nitschke W: Onthe universal core of bioenergetics. Biochim Biophys Acta 2013,1827:79-93.

71. Cheng Z, Sattler S, Maeda H, Sakuragi Y, Bryant DA,DellaPenna D: Highly divergent methyltransferases catalyze aconserved reaction in tocopherol and plastoquinonesynthesis in cyanobacteria and photosynthetic eukaryotes.Plant Cell 2003, 15:2343-2356.

72. Sattler SE, Cahoon EB, Coughlan SJ, DellaPenna D:Characterization of tocopherol cyclases from higher plantsand cyanobacteria. evolutionary implications for tocopherolsynthesis and function. Plant Physiol 2003, 132:2184-2195.

73. Schmitt F-J, Renger G, Friedrich T, Kreslavski VD,Zharmukhamedov SK, Los DA, Kuznetsov VV, Allakhverdiev SI:Reactive oxygen species: re-evaluation of generation,monitoring and role in stress-signaling in phototrophicorganisms. Biochim Biophys Acta 2014, 1837:835-848.

74. Hamilton TL, Bryant DA, Macalady JL: The role of biology inplanetary evolution: cyanobacterial primary production in lowoxygen Proterozoic oceans. Environ Microbiol 2015, 18:325-340.

75. Sedoud A, Lopez-Igual R, Ur Rehman A, Wilson A, Perreau F,Boulay C, Vass I, Krieger-Liszkay A, Kirilovsky D: Thecyanobacterial photoactive orange carotenoid protein is anexcellent singlet oxygen quencher. Plant Cell 2014, 26:1781-1791.



























































































































76. Di Matteo A, Scandurra FM, Testa F, Forte E, Sarti P, Brunori M,Giuffre A: The O2-scavenging flavodiiron protein in the humanparasite Giardia intestinalis. J Biol Chem 2008, 283:4061-4068.

77. Frazao C, Silva G, Gomes CM, Matias P, Coelho R, Sieker L,Macedo S, Liu MY, Oliveira S, Teixeira M et al.: Structure of adioxygen reduction enzyme from Desulfovibrio gigas. NatStruct Biol 2000, 7:1041-1045.

78. Seedorf H, Hagemeier CH, Shima S, Thauer RK, Warkentin E,Ermler U: Structure of coenzyme F420H2 oxidase (FprA), a di-iron flavoprotein from methanogenic Archaea catalyzing thereduction of O2 to H2O. FEBS J 2007, 274:1588-1599.

79. Silaghi-Dumitrescu R, Kurtz DM, Ljungdahl LG, Lanzilotta WN: X-ray crystal structures of Moorella thermoacetica FprA. Noveldiiron site structure and mechanistic insights into ascavenging nitric oxide reductase. Biochemistry (Mosc.) 2005,44:6492-6501.

80. Vicente JB, Carrondo MA, Teixeira M, Frazao C: Flavodiironproteins: nitric oxide and/or oxygen reductases. InEncyclopedia of Inorganic and Bioinorganic Chemistry. Edited byScott RA. John Wiley & Sons, Ltd.; 2011.

81. Romao CV, Vicente JB, Borges PT, Frazao C, Teixeira M: The dualfunction of flavodiiron proteins: oxygen and/or nitric oxidereductases. J Biol Inorg Chem 2016, 21:39-52.

82. Mastronicola D, Falabella M, Forte E, Testa F, Sarti P, Giuffre A:Antioxidant defence systems in the protozoan pathogen Giardiaintestinalis. Mol Biochem Parasitol 2015 http://dx.doi.org/10.1016/j.molbiopara.2015.12.002. Published online Dec. 7, 2015.

83. Bombar D, Heller P, Sanchez-Baracaldo P, Carter BJ, Zehr JP:Comparative genomics reveals surprising divergence of twoclosely related strains of uncultivated UCYN-A cyanobacteria.ISME J 2014, 8:2530-2542.

84. Helman Y, Tchernov D, Reinhold L, Shibata M, Ogawa T,Schwarz R, Ohad I, Kaplan A: Genes encoding A-type


flavoproteins are essential for photoreduction of O2 incyanobacteria. Curr Biol 2003, 13:230-235.

85. Helman Y, Barkan E, Eisenstadt D, Luz B, Kaplan A: Fractionationof the three stable oxygen isotopes by oxygen-producing andoxygen-consuming reactions in photosynthetic organisms.Plant Physiol 2005, 138:2292-2298.

86. Allahverdiyeva Y, Mustila H, Ermakova M, Bersanini L, Richaud P,Ajlani G, Battchikova N, Cournac L, Aro E-M: Flavodiiron proteinsFlv1 and Flv3 enable cyanobacterial growth andphotosynthesis under fluctuating light. Proc Natl Acad Sci U S A2013, 110:4111-4116.

87. Zhang P, Allahverdiyeva Y, Eisenhut M, Aro E-M: Flavodiironproteins in oxygenic photosynthetic organisms:photoprotection of photosystem II by Flv2 and Flv4 inSynechocystis sp. PCC 6803. PLoS ONE 2009, 4:e5331.

88. Wheeler G, Ishikawa T, Pornsaksit V, Smirnoff N: Evolution ofalternative biosynthetic pathways for vitamin C followingplastid acquisition in photosynthetic eukaryotes. eLife 2015, 4.

89. Farquhar J, Bao H, Thiemens M: Atmospheric influence ofEarth’s earliest sulfur cycle. Science 2000, 289:756-759.

90. Paris G, Adkins JF, Sessions AL, Webb SM, Fischer WW:Neoarchean carbonate-associated sulfate records positiveD33S anomalies. Science 2014, 346:739-741.

91. Scott AC, Glasspool IJ: The diversification of Paleozoic firesystems and fluctuations in atmospheric oxygenconcentration. Proc Natl Acad Sci U S A 2006, 103:10861-10865.

92. Raymond J, Blankenship RE: Biosynthetic pathways, genereplacement and the antiquity of life. Geobiology 2004, 2:199-203.

93. Balsera M, Uberegui E, Schurmann P, Buchanan BB: Evolutionarydevelopment of redox regulation in chloroplasts. AntioxidRedox Signal 2014, 21:1327-1355.




























http://dx.doi.org/10.1016/j.molbiopara.2015.12.002

http://dx.doi.org/10.1016/j.molbiopara.2015.12.002







































How did life survive Earth's great...

Documents

Transcript of How did life survive Earth's great...