Energy, Metabolism, and Photosynthesis

Originally I sat down to write about the large-scale history of Earth, and line up the big developments that our biosphere has undergone in the last 4 billion years.  But after writing about the reason that Earth is unique in our solar system (photosynthesis being an option here), I guess I needed to explore photosynthesis and other forms of metabolism on Earth in a little more detail and before I knew it I’d written more than 3000 words about it.  So, here we are, taking a deep dive into photosynthesis and energy metabolism.

Again, fans of the ‘rare earth hypothesis’ will tend to take a look at everything along the history of Earth and argue that if anything had happened differently, we wouldn’t have the big happy biosphere we have today.  This only holds water, however, if there was no other way to have what happened happen.  It also only really holds water if there is only one instance of the major transitions in Earth’s biosphere – if something happened multiple times on Earth you can argue that it is likely to happen anywhere there is life, while if something only happened once here it is at least possible that we are here because of a very rare event that doesn’t often happen, and necessarily see that unusual event.  Here, I will try to dissect energy metabolism across the set of living things on Earth and argue that at least coupling light to it is probably likely to happen anywhere there is light falling onto a biosphere.

I would like to refer all my readers to a wonderful recent book on photosynthesis, from which I derived more than half of my understanding of it:  “Oxygen, a Four Billion Year History” by Donald Canfield.  It is an amazing walk through the current state of the art of research on both the history/origins of photosynthesis, and the history of molecular oxygen in Earth’s atmospheric and geological chemistry.  I would also like to refer everyone to another recent book that touches on energy metabolism and the origin of life by Dr. Nick Lane, entitled “The Vital Question: Energy, Evolition, and the Origins of Complex Life”.  The first third of the book is a thorough look at a particular strand of origin of life research I find compelling in many of its details, and the latter two thirds explores the origin and attributes of Eukatyotes from a perspective I am rather less of a fan of but think makes some important points, and will get to later.

Redox Metabolism

I am going to make a bold claim:  all metabolism on Earth depends ultimately on redox gradients, and photosynthesis is just a way to generate these internally in a cell from light.

Redox reactions – named for oxidation and its opposite, reduction – are chemical reactions that involve the movement of electrons from one substance to another during the reaction.  A substance that loses electrons in a reaction is said to be oxidized (so named long ago because oxygen is a fantastic oxidizing agent that will suck electrons away from almost anything) and something that gains electrons is said to be reduced (a horrible piece of nomenclature – blame 18th century chemists who didn’t understand electrons and only thought about redox reactions in terms of gaining or losing oxygen).  When hydrogen burns in oxygen, the result is water.  The OH bond in water is asymmetrical – oxygen pulls on the shared electrons in the bond much harder and gains a partial negative charge.  The oxygen is reduced, as it has gained some electron density from the hydrogen, and the hydrogen has been oxidized as it has lost some of its electron density.  This definition can be extended to any substance, even ionic substances like iron which can exist in two redox states – Fe2+ is more reduced than Fe3+ because it has one more electron.

The most obvious way that metabolism depends on redox reactions is the fact that all carbon in biomass ultimately comes from CO2.  This is a highly oxidized and inert form of carbon, and it must be reduced and some oxygens pulled off of it before it can do the interesting chemistry of life.  In all of life, there are six known molecular pathways through which carbon is stripped out of CO2 and built into biomolecules.  In all of them, both cellular energy and highly reduced ‘carrier’ molecules react with CO2 to reduce it.  The energy applied massively increases the reaction rate, and the reduced carrier molecules strip away the oxygens as water and reduce the carbon so it can bond with other carbons and hydrogen.  These reducing carriers either need to be reduced themselves by reducing agents in the environment that the organism eats, in the case of chemolithotrophic bacteria that live far away from sunlight, or they need to be reduced by photosynthesis forcing electrons from another substance out in the word onto it by the energy of photosynthesis.

There is one carbon fixation pathway that is special that I will talk about more in later blog posts.  This is sometimes called the acetyl-coenzyme A pathway, and sometimes the Wood-Ljungdahl pathway.  It is the pathway by which methanogenic archaea and acetogenic bacteria react hydrogen gas from geological or biological sources with CO2 to build biomass.  Of all six known pathways, it is the only one found in both the archaebacteria and the eubacteria – the two deepest branches on the tree of life that split off the last universal common ancestor, ‘LUCA’.  This alone should be a clue that it could be important.  But secondly, it is the only carbon fixation pathway that is on the whole spontaneous and does not require an external input of energy – its early stages require an input of energy, but its later stages actually generate more energy than is consumed.  This is because hydrogen gas is such a strong reducing agent – most substances when reacting with it will pull electrons away from it.  As such, fixing carbon using hydrogen gas has been called, “a free lunch you are paid to eat,” by certain origin of life researchers.  It is actually energetically favorable for CO2 and hydrogen gas to interact to build biomass.  This pathway may have a key role to play in the origin of life under certain assumptions, which I will discuss another time.

Secondly, life gets energy to DO things from redox gradients, not just build things.  Redox gradients are perfect for life to insert itself into as a middleman to get energy out of.  Temperature gradients, for example, are pretty much a nonstarter – a cell is usually at most microns wide, and there are no natural temperature gradients that matter that are that sharp in the natural world.  Redox gradients can get inside a cell – you can have an oxidized molecule and a reduced molecule right next to each other inside an organism, and it can require a catalyst to get them to react.  The energy released by that reaction is, by and large, harnessed by a mechanism called chemiosmosis.  Cells make sure that the reaction that pulls an electron away from a reducing agent and the reaction that adds one to an oxidizing agent happen in different places on the cell, usually on a cell membrane.  They then pass electrons between these places through what’s called an electron transport chain, a series of membrane-bound proteins carrying metal ions and other reactive redox reaction centers that the electron  must pass along in order to complete the energetically favorable redox reaction.

The electron transport chain of animal mitochondria, in which food is oxidized by oxygen to pump protons out of the mitochondrial inner membrane.  These protons leak back across their concentration gradient, driving the production of ATP via the ATP synthase.  There are many other quite diverse electron transport chains across biology.

The electron moving from protein to protein to complete the reaction is coupled to and forces the pumping of protons – hydrogen ions – through the membrane from the inside of a cell to the outside.  As these reactions proceed a proton gradient (and thus a pH gradient) is built up across the membrane.  This gradient is then allowed to dissipate through something called the membrane ATP synthase – a huge rotary complex that couples the dissipation of the proton gradient to the production of ATP, a highly energetic molecule that most processes that use energy in biology get their energy from.  Why this roundabout coupling?  Because ATP is a highly energetic molecule and many redox reactions do not release enough energy to cause a whole molecule of ATP to be produced.  But by pumping protons across a membrane and building a gradient, the effects of multiple redox reactions can be accumulated and saved up until enough energy has been stored to produce one molecule of ATP.

The only energy metabolism on Earth that doesn’t use membrane-bound redox manipulating proteins is fermentation – a whole slew of processes that take large biomolecules and rearrange their atoms into slightly lower energy states.  But even that rearrangement is an intramolecular redox reaction – the bonds are rearranged in such a way to slightly reduce oxygen further, slightly oxidize hydrogen further.  Through some clever chemistry a tiny amount of the energy provided by their rearrangement – say from sugar into CO2, alcohols, and acids – is captured but it’s very small compared to the energy you get from bigger electron transfers between different substances.  But ultimately, all carbon fixation from CO2 into biomass derives from either reducing CO2 using a reducing agent out in the environment and energy from a collapsing a redox disequilibrium, or photosynthesis.  As such, organisms that ferment biomolecules all ultimately depend on other living things for their energy.  They also all show evidence of descending from precursors that used redox reactions pumping protons for energy.  As such, chemiosmotic energy production appears to be ancestral to all life, likely via chemolithotrophy, consuming chemical redox gradients be it between CO2 and H2 or other slight disequlibria within Earth’s geosphere.


As I mentioned in my last post, “The Solar System:  Why Earth?” photosynthesis is the reason that Earth is what it is today.  It drives the vast majority of metabolic activity on the planet and has chemically transformed the atmosphere and geosphere.  The story of Earth since this set of metabolic pathways was invented is, arguably, mostly the story of photosynthesis and its interaction with the geosphere and the evolutionary pressures it created for all life on Earth.  As the source of almost all biomass on Earth and the force that built the planet we have, I think it deserves a really deep look.

Light is a very interesting biological energy source.  There are multiple reasons that all photosynthesis makes use of what we call ‘visible light’.  One, there’s a LOT of it here – at our sun’s temperature, the vast majority of the radiation it emits is in this band of wavelengths, and our atmosphere is really only strongly transparent to two bands of frequencies, the visible and the radio.  Two, the energies that visible light photons have really are special.  Longer wavelengths, like infrared or microwaves or radio, can transfer energy and heat things up but individual photons do not have enough energy to move electrons around inside the organic molecules that living things are composed of.  They do often contain enough energy to vibrate or rotate bonds, but the electrons stay stuck exactly where they are and no chemical reactions are driven since photons and electrons interact for the most part in a binary way, transferring the exact amount of energy embodied in the photon.  Shorter wavelength photons, like ultraviolet light or X-rays, are so energetic that they can completely blow apart organic molecules or break bonds turning biomolecules into highly reactive poisons.

Visible light, on the other hand, has just the range of energies that can kick electrons around within organic molecules without completely breaking them, forcing the molecules into excited states or different shapes.  Strong enough to shuffle electrons, weak enough to not shatter the molecules completely.  This is especially interesting given what I said above about the metabolism of all life ultimately deriving from redox gradients – the movement of electrons from one substance to another during chemical reactions.

Photosynthesis has originated at least twice in Earth’s biosphere.  Or, more technically, phototrophy has – there is a distinction to be made between organisms that use sunlight to produce energy to DO things (phototrophs) or MAKE things (photosynthesizers).  All cells that photosynthesize are phototrophs, as they must be able to extract energy from sunlight, but only a subset of phototrophs are able to take that energy and embody it in long-lived organic biomolecules rather than just run their energy-using processes off it while they get their organic molecules from elsewhere.  Only one of these two origins of phototrophy has been elaborated into photosynthesis, in multiple inventive ways (one of these being oxygen-producing photosynthesis).  It is photosynthesis that has then gone on and remade the world.

Bacteriorhodopsin, via the Protein Databank (and Wikipedia).  Retinal pigment caged inside in purple.

The simplest form of phototrophy on Earth is the plethora of microbes carrying a molecule called bacteriorhodopsin (or a whole bunch of evolved derivatives of it) – a membrane-bound protein clutching a little purple dye molecule called retinal (note the name: it’s the same dye that in our own eyes is clutched by different proteins with a similar shape but probably a separate evolutionary origin).  When a photon of the proper frequency range hits this molecule (mostly green, leaving purple behind to see) it can excite it and flip the molecular structure of one particular bond 180 degrees and change its shape, in the process forcing a proton across the membrane from the inside to the outside of the prokaryotic cell.  This proton can then be brought back through the membrane by the ATP synthase or other channels that couple the breakdown of the proton gradient to other activities.

Light flips a particular bond in retinal from ‘trans’ to ‘cis’ arrangement, causing the pumping of one proton per photon across the cell membrane through the bacteriorhodopsin protein.  Image by Darekk2, from Wikipedia.

Bacteriorhodopsin really just gives a cell that carries it energy, not the ability to fix carbon in and of itself.  It is also rather inefficient, with every photon captured only pumping one proton and thus wasting a lot of the available energy.  But it’s simple and inexpensive.  All you need for it is the metabolic pathway required to build the retinal pigment (not much protein per cell all things considered) and the single bacteriorhodopsin gene that represents the pump itself.  This phototrophy pathway has been horizontally transferred all over the eubacteria and archaea by these microbes swapping genes around (and a few times into the Eukaryotes where it is usually substantially altered to do something completely different) to the point it’s pretty hard to figure out where and when it originated within the eubacteria or archaea.

The second, substantially more complex form of phototrophy looks a bit more like the redox-driven metabolisms I previously spoke about and is ultimately a way of transforming the highly organized energy of visible light into a redox gradient internal to the cell.  Everything with this second form of phototrophy uses some variation on the theme of chlorophyll or bacteriochlorophyll as important pigments, caged up in a massive membrane-bound protein complex called a photosystem.

An incomplete Photosystem II structure from cyanobacteria.  Proteins as ribbons, pigments as green ball-and-sticks wedged between them.  Illustration by Curtis Neveu, from Wikipedia.

You can tell just from looking at this that it’s a much much more complicated system, with dozens of proteins involved (not all pictured and not all part of the photosystem) and multiple types of pigments embedded in it.  What happens in these photosystems is special:  they are where the energy of light is turned into an internal redox gradient.  Light hits one of the pigments in the photosystem, or one of some dispersed ‘antenna’ pigments elsewhere in the membrane of the cell.  These pigments become electrically excited, and are capable of transferring this excitation back and forth between themselves, until the pigment at the active center of the photosystem is excited.  At the active center, this electronically excited pigment transfers an electron to another molecule, losing that electron.  When this happens, suddenly you have one molecule with an overabundance of electrons (a very strong reducing agent) and one with a deficit of electrons (a very strong oxidizing agent).  It is very energetically favorable for those molecules to react and release this energy like any other favorable redox reaction.  But the structure of the photosystem catalyzes the creation of that gradient from the excited pigment, not its dissipation by letting them react.  To dissipate the new redox gradient between these two molecules, the electrons have to go through a Rube Goldberg apparatus that wrings as much energy as possible from them in the process.

Some phototrophic organisms (green nonsulfur bacteria, most purple nonsulfur bacteria, etc) basically just run the electrons from one photosystem through a circular electron transport chain, running them from the photosystem through membrane protein complexes until the end of the chain dumps the now lower-energy electron back into the oxidized pigment, restoring it to its original state but having pumped several protons across the membrane in the process.  This, again, just extracts energy from the light and puts it into the form of a proton gradient that the rotary ATP synthase or other processes can dissipate to let the cell do its business but does it much more efficiently than bacteriorhodopsin can at the expense of complexity.

Other organisms actually take the redox gradient created within a photosystem and use it to fix CO2 into biomolecules, performing photosynthesis.  These organisms (the green sulfur bacteria, the purple sulfur bacteria, some purple and green nonsulfur bacteria, etc) take the electron rich reducing agent generated in the photosystem and through the various carbon fixation pathways use it to reduce the carbon in CO2 into biomass (average formula: CH2O).  This, however, leaves the extremely strong oxidizing agent in the photosystem un-neutralized.  It has to be neutralized or else the cycle cannot repeat – for every carbon that is reduced using light energy, something else has to be oxidized by the photosystem using light energy. A whole slew of substances are possible – hydrogen sulfide can be oxidized to elemental sulfur as in the various sulfur bacteria, ferrous iron (green Fe2+) can be oxidized to ferric iron (red Fe3+, rust), or water can be oxidized into oxygen.  Ultimately, photosynthesis is thus a process for ripping electrons from something in the environment, thus oxidizing it, and using the electrons to reduce CO2 into biomass.

Oxygen-producing photosynthesis appears to have only appeared from photosystem-type photosynthesis once, in the cyanobacteria.  But today it is by far the most common kind.  This has very little to do with the fact that it makes oxygen in and of itself.  Dissolved ferrous iron, hydrogen sulfide, and water are all perfectly good electron donors for photosynthesis.  But oxygenic photosynthesis both sidesteps factors that limit all other forms of photosynthesis and simultaneously forces these other forms to the sidelines.

If you are fixing carbon using sulfur producing photosynthesis, you need hydrogen sulfide – two H2S molecules for every CO2 molecule you turn into biomass.  If you are using ferrous iron, you need four Fe2+ ions for every CO2 you fix.  These molecules are produced geologically, iron leaching out of rocks as they are weathered by water and hydrogen sulfide belching out of volcanoes and coming up from underground in certain types of hydrothermal vents.  But the rate at which carbon can be fixed by a biosphere running on these forms of photosynthesis is constrained by the rate at which these substances are produced geologically.  There’s a cap, and these substances are the limiting factor.  Oxygen producing photosynthesis, on the other hand, uses water itself as the chemical source of electrons, two H2O for every CO2.  It’s hard to think up a more common substance in biology.  In one stroke, the evolution of oxygenic photosynthesis removed a primary chemical limiting factor for biomass production and turned the limiting factor into mineral nutrients, atmospheric CO2 levels, or light energy itself, depending on the ecosystem.

On top of this, once oxygenic photosynthesis became common enough the oxygen byproduct of biomass production started rapidly oxidizing the geosphere and atmosphere with massive side effects.  In the presence of absolutely TINY levels of atmospheric oxygen, ferrous iron immediately turns into ferric iron oxide – rust – and falls out of solution, eliminating one major source of nonoxygenic biomass production from the biosphere.  Huge bands of iron oxide from the ancient ocean floors show this process from what is known as the ‘great oxidation’ 2.3 billion years ago when atmospheric oxygen levels first rose from nothing to just barely something.  Oxygen also oxidizes hydrogen sulfide into water and sulfate salts – and living things accelerate this process, building electron transport chains to get energy for themselves using oxygen as the oxidizing agent and hydrogen sulfide as the reducing agent.  This practically eliminates it from the biosphere except in the places where it is generated (which often have no available light).  In short, oxygen rapidly oxidizes all the reduced substances that can be used by other forms of photosynthesis as electron sources, so not only does oxygenic photosynthesis remove limiting factors on biomass production, but it crowds out the competition.  These other forms of photosynthesis most assuredly exist today, but they are marginal, making tiny fractions of the biomass of the planet in special environments.

Oxygenic photosynthesis is a funny hack of the photosystem-based photosynthesis pathway.   Oxygen is a VERY strong oxidizing agent, much stronger than sulfur or Fe3+, and it takes a lot more energy to rip the electrons out of water to make it than it takes to rip them out of hydrogen sulfide or ferrous iron.  Given the efficiency that photosystems have, there’s not enough energy captured from a single visible light photon to do it even though there is enough energy to do so to these nonoxygenic substrates.  In cyanobacteria, the one place that oxygenic photosynthesis emerged, the two different uses for photosystems were combined.  In essence, an energy-producing photosystem and a sulfur-producing photosystem were tied together by a short electron transport chain in between them in the same cell.  The energy producing photosystem stacks its energy on top of the energy captured by the sulfur producing photosystem, allowing the combined pathway to use the energy of two photons rather than one to rip apart water and allow this ubiquitous but tough substrate to be used.  This also necessitated the evolution of the part of the oxygenic photosystem II that splits water, the oxygen-evolving complex – there’s some weird chemistry that happens there that is unlike what happens when hydrogen sulfide is involved and there is quite a bit of argument over how it came to be.

This is weird.  No other phototrophic microbes have both the energy-producing and the photosynthesizing photosystem in the same cell – two very similar but still divergent protein complexes – but instead only have one or the other.  It’s hard but not impossible to imagine horizontal gene transfer moving around that many dozens of genes in a coherent way.  One popular hypothesis surrounding the origin of oxygenic photosynthesis suggests that an ancestral photosynthesizing microbe actually had both the circular electron transport chain photosystem and the sulfur / iron / etc oxidizing photosystem which had been duplicated within its genome and diverged, and would switch between them depending on the environment and what it needed.  In this situation the lineage leading to the cyanobacteria chained them together, while in all other known lineages of phototrophic/photosynthetic bacteria one or the other photosystem would have been lost (loss being close to irreversible for such a big complicated system and likely to happen to something if it’s ever not needed by a lineage for a long enough while).  This hypothesis is, however, unconfirmed, and the details of the origin of oxygenic photosynthesis are not well understood other than that it appears to have come from a lineage that was capable of both using light for energy and using light to make biomass using hydrogen sulfide.  The question is still open as to how this happened.


No one really knows the time at which oxygenic photosynthesis first appeared.  People like to point to the ‘great oxidation’ about 2.3 billion years ago, but that just represents the time at which the main geological sinks that sucked up biogenic oxygen filled up and oxygen became a measurable force in the atmosphere.  Volcanic or biogenic methane exhalations, as well as the oxidation of crustal minerals and the enormous pool of dissolved iron in the oceans would have kept the air completely anoxic for quite some time before the rate of production of oxygen could overwhelm the rate of it being sequestered away in sediments as oxides or turned back into CO2.  The slowing of Earth’s geology as the interior of the planet cooled down leading to less volcanism or changes in the mineral nutrients available to drive different forms of metabolism as a result of this could have been confounding factors, changing when this tipping point was reached.

There is ample evidence of photosynthesis going on all over Earth long before the great oxidation, but it can be difficult to tell oxygenic and nonoxygenic photosynthesis apart unless you can get preserved chemicals from the cells themselves out of fossil microbial mats.  The older banded iron oxide formations from ancient seafloor may have been produced by non-oxygenic photosynthesis turning ferrous iron to rust, without the help of oxygen, or at the same time that oxygenic photosynthesis was also oxidizing iron indirectly.  Molecular biomarkers from preserved microbial mats have suggested cyanobacteria were definitely around by 2.7 billion years ago, and in 2014 evidence came to light in the form of various metal oxides in the rock record that certain shallow marine settings covered in microbial mats were getting oxidized before the ocean basins as a whole up to 3 billion years ago (see Planavsky et al., 2014:  Fossilized microbial mats consistent with an origin from photosynthesis of some sort are seen far before even then, though.

Stromatolites are bulbous formations of photosynthesizing bacteria, trapped sediment, and crystallized minerals excreted by the bacteria.  They only appear when grazing animals are completely absent, and are exceedingly rare today but show up all through the rock record of early Earth.  The oldest unequivocal evidence for photosynthesis of any type was, until recently, 3.5 gigayear old fossil stromatolites seen in Australia (the oldest pieces of the Earth’s crust that survive to this day can be found in Australia, Northern Canada, and Greenland).  But just this August, previously unseen rock strata revealed in Greenland by retreating ice have been found to contain a new deposit bearing what are believed to be the oldest known stromatolites in the world, at 3.7 gigayears old (see Nutman et al., 2016, ‘Rapid emergence of life shown by discovery of 3,700 million year old microbial structures’ at (popular press:

Oldest known stromatolites, layered microbial fossils created by photosynthetic microbial mats growing towards light and trapping sediment.  Figure 1 reproduced from Nutman et al., 2016.

As photosynthesis is a very complicated process, oxygenic or non, it’s striking that these structures were well established by 3.7 gigayears ago.  The entire solar system is about 4.6 gigayears old, and there’s precious little rock on Earth that has survived the cycles of plate tectonics and geologic recycling back to 3.7 gigayears ago.  This is an absolute MINIMUM age for a well established, photosynthetic, high-biomass biosphere that was transforming a lot of matter and energy, as it would be very difficult to find evidence if this was happening earlier.  Life got going FAST, and it is the preservation of evidence of it that is the factor that keeps limiting our knowledge of its age.  I would bet money on photosynthesis being found to be significantly older than this latest discovery – the earliest known date DID just slip back by 200 megayears after all, and there are equivocal traces of biogenic carbon in the tiny grains of older rocks that have been preserved that not everyone believes going all the way back to 4.3 gigayears ago (!).

In short, photosynthesis in general goes back all the way to the edge of our ability to look for it in some of the oldest coherent, non-ground-up rocks that we can look for microbial structures in and we have probably not yet found its true age.  The origin of oxygenic photosynthesis is fuzzier.  It could be a relative latecomer stitched together by horizontal gene transfer, or it could have come from one of the first microbes to use photosystems, with most other photosynthetic or phototrophic microbes losing the full set of pieces that were stitched together to create that pathway.


So, is photosynthesis and later oxygenic photosynthesis a weird unlikely event, or something that you’d expect in any biosphere?  The fact that there are two separate origins of phototrophy, getting energy from light, argues to me that such a great energy source WILL be tapped anywhere there are living things and sunlight in the same place.  The photosystem-driven form is the only form that gave rise to photosynthesis, however.  Does that mean that this is a rare event on cosmic scales?  Probably not, though I can’t say with certainty.

All photosynthetic organisms use fantastically sophisticated machinery, with little to nothing in the way of intermediate forms that show another independent elaboration of a simpler, worse starting point.  This suggests that once even rudimentary photosynthesis was established, it was souped up to full efficiency and effectiveness very quickly with little opportunity for diversification of ‘half-finished’ lineages, and that the ‘complete’ lineage then radiated rapidly into all the various forms we see today.  This could’ve represented a strong ‘first mover’ advantage, where any newcomers with very bad rudimentary photosynthesis would be outcompeted in environments where photosynthesis were possible by the well-established old guard of efficient photosynthesizers.  I really have no idea how to rigorously tell apart a circumstance in which the first-mover advantage prevents another evolution of sophisticated photosynthesis, and a circumstance in which a fantastically rare evolution of a complicated system happened to occur here and take over the world, though I strongly favor the former due to the lack of half-built intermediate forms and the apparent age of the pathway.

Building phylogenetic trees of bacteria to understand their relationship with each other by analyzing sequence data is a very difficult proposition and not my area of expertise – they trade genes back and forth with each other like mad, passing enzymes around and taking them up from their neighbors.  “Species” is a dicey concept for them, as the complement of all genes present in every member of a given species can be larger than the number of genes in any given individual by a factor of 3.  Most people who do bacterial taxonomy thus focus on a very particular set of genes – those of the protein translation system, which are very resistant to horizontal transfer.  So now, I’m probably going to severely anger any microbial taxonomists in my audience and make a claim based on my analysis of a recent evolutionary tree created from thousands of sequenced microbial genomes (Hug et al., 2016 “A new view of the tree of life” in Nature Microbiology,

Recent reconstruction of the full tree of life using new sequence analysis and thousands of microbial genomes.  Archaea and eukaryotes branch together below, bacteria above.  Red dots indicate bacteria no one has ever cultured and which are only known from genome sequence data in the last decade or so.  LUCA lies somewhere in the long branch between the bacteria and the archaea/eukaryotes. Figure 1 reproduced from Hug et al., 2016.

This tree will get a lot more analysis in later posts.  But for now, within the Bacterial radiation in the upper half of the tree, I note that the Cyanobacteria, Chloroflexi, several branches within the proteobacteria, the Acidobacteria, and Chlorobi – various photosynthetic lineages with different variations on photosynthesis of the photosystem style – are widely separated along the tree.  This suggests that either photosystem-based photosynthesis underwent repeated massive horizontal transfer among the bacteria – possibly difficult for a large complex system and also allowing a later origin for oxygenic photosynthesis, but possible – or it actually originated very very deep near the base of the bacterial tree, very long ago, and most lineages just lost it and specialized into eating organic molecules created by photosynthesizers or going back to chemosynthesis.  Or that the tree’s topology is wrong, of course.

In conclusion, I think it’s clear from its multiple origins that phototrophy, using light for energy, is likely to show up anywhere there is light and life.  I strongly suspect, but cannot rigorously prove, that photosynthesis of biomass was a very early development in life on Earth near the root of the Bacteria and just produced a very strong first-mover advantage crowding out secondary origins of it, and would probably also show up where there is life and light.  As for oxygen-producing photosynthesis, its origin from more mundane forms of photosynthesis is still being studied.  It required a strange chaining together of multiple modes of photosynthesis to make it work, and only ever happened once.  Its time of emergence, early or late, is pretty unconstrained and I don’t think there’s sufficient evidence to say one way or another if it is likely to happen anywhere there is photosynthesis.  It could be subject to the same sort of ‘first mover advantage’ situation that other photosynthesis may have encountered as well.  But whatever the case once it got going, it would naturally take over biomass production and crowd out other forms of photosynthesis due to its inherent chemical advantages on any wet planet and its effects suppressing other forms of photosynthesis.

Oxygen in the atmosphere had some important side effects, one which most people care about being allowing big complicated energy-gobbling organisms like animals – all that energy that organisms can get burning biomass in oxygen lets organisms that do so do a lot of interesting stuff.  Looking for oxygen in the atmospheres of other terrestrial planets would be an extremely informative experiment, as the presence of this substance would suggest that a process very similar to the process that created our huge diverse and active biosphere were underway.


In the week since I posted this, I’ve been pointed to some recent amazing work from the last five years, in journals that I don’t frequent much, supporting the idea that photosynthesis is very ancient.  I will link to them below.  The end result of recent work by people who study photosynthesis specifically suggests that the two different photosystems did in fact originate in the same cell via a duplication, and that cyanobacteria are just the only photosynthetic lineage that kept both rather than combining them horizontally later.  On top of this, looking at the evolution of the photosystems and the enzymes that make the various pigments suggests that most of the photosynthetic lineages may have inherited everything vertically from their ancestors rather than acquiring them horizontally, and in particular that cyanobacteria never acquired any of these things horizontally.

The possibility that most of the photosynthetic lineages inherited photosynthesis vertically along with the topology of most suggested bacterial trees suggests that an actual majority of bacterial groups are descended from photosynthesizers.  This puts photosystem-based photosynthesis very close to the root of the bacterial tree, and very early in the history of life on Earth, bolstering the idea that the origin of photosynthesis was relatively easy but crowded out second origins of this process because the original pathway was so much better than any new origins that would have started out very bad at it.  Oxygen producing photosynthesis is still unclear, since nobody really knows how long ago the two photosystems were chained together to break water.  On the one hand, everything needed to create oxygenic photosynthesis was around quite early in the same cell, on the other hand it only arose once and the case for the crowding out of secondary origins is not as clear.


Cardona T, 2016, Origin of Bacteriochlorophyll a and the Early Diversification of Photosynthesis., PLoS One, Vol: 11

Cardona T, 2015, A fresh look at the evolution and diversification of photochemical reaction centers., Photosynth Res, Vol: 126, Pages: 111-134

Cardona T, 2016, Reconstructing the Origin of Oxygenic Photosynthesis: Do Assembly and Photoactivation Recapitulate Evolution?, Frontiers in Plant Science, Vol: 7

Blankenship RE. 2010. Early evolution of photosynthesis. Plant Physiol. 154:434–38


Boy, nobody can agree on this.  Additional work I’ve been pointed to in the last few weeks has been used to support the possibility of horizontal transfer being the main method of photosystem gene transfer around the bacterial tree.  This looks not at the relative tree topologies of different parts of the photosynthetic apparatus but the position of electron transport chain components and photosystem components in blocks of genes next to each other.  It also makes use of molecular clock estimates which I am usually wary of, but points to clear shifts in the electron transport chains of these organisms as evidence for horizontal transfer.

I find myself hoping that the large amount of horizontal gene transfer that happens in prokaryotes has not completely scrambled the signal from the past that would allow us to figure this out once and for all…

Reference:  Soo et al., 2017.  On the origins of oxygenic photosynthesis and aerobic respiration in Cyanobacteria. Science 355, 1436–1440


Author: Tony

Biology postdoc at Georgia Tech and astronomy enthusiast who thinks about astrobiology way too freaking much.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s