THE planet is in crisis. The stench of death is everywhere as whole branches of the tree of life are pruned almost to oblivion – and all because of the waste gas pumped into the atmosphere by one incredibly successful species. Welcome to Earth, 2.4 billion years ago.
This was arguably the most tumultuous episode in life’s history. It had been thriving for well over a billion years when a new kind of cell appeared on the scene, one that harvested the sun’s energy using a process that generates a highly toxic by-product – oxygen. These cells were soon growing in such unimaginable numbers in the primordial oceans that they transformed Earth’s atmosphere.
At the time, this was a catastrophe. The rise of oxygen may have wiped out a greater proportion of life than in any other mass extinction. But the very property that makes oxygen so dangerous – its high reactivity – also makes it a rich source of energy. Life soon started to exploit this, including, of course, our animal ancestors.
Advertisement
In the past decade, our view of this crucial episode has been turned upside down. The textbooks will tell you that oxygen levels began climbing soon after photosynthesis evolved, but we now know that some cells started photosynthesising as long as 3.4 billion years ago, long before oxygen levels began to rise. The question is, why did it take so long for them to start pumping out oxygen?
At its heart, photosynthesis is about harvesting the sun’s energy. Plants use this energy to make food, by building chains of carbon from carbon dioxide. The process produces sugars that can be used as an energy source or to make more complex molecules, from proteins to DNA. But contrary to what you might expect, it does not necessarily produce oxygen. In fact many bacteria turn light and CO2 into food without producing oxygen. What’s more, recent discoveries suggest they have been doing so for nearly as long as there has been life on Earth.
“Bacteria have been photosynthesising for nearly as long as there has been life on Earth. So why did it take a billion years for them start making oxygen?”
In 2004, Michael Tice and Donald Lowe, both then at Stanford University in California, were studying rocks in South Africa that formed in shallow water 3.41 billion years ago. They found fossil structures rather like the microbial mats formed by photosynthetic bacteria today, but no sign that any oxygen was produced (). The most likely explanation, they think, is that these cells were carrying out anoxygenic photosynthesis.
Since that discovery we have actually come face-to-face with some of these early photosynthetic microbes. In 2011, Martin Brasier at the University of Oxford and colleagues discovered fossils of individual bacterial cells in rocks that formed 3.43 billion years ago, in what is now western Australia (). “They occurred in a well-lit intertidal or supratidal setting,” says Brasier. The chemical make-up of the rocks, along with the plentiful light, strongly suggests that some of the cells photosynthesised without producing oxygen.
It may seem surprising that anoxygenic photosynthesis evolved so soon after life itself – the earliest fossils we know of are only slightly more ancient, at 3.49 billion years old. But Nick Lane of University College London, who studies life’s origins, thinks that once cells capable of living on chemical energy had evolved, it was not a huge step for them to start exploiting light energy instead. “Really, light just gets electrons flowing through the same equipment,” he says.
For researchers like Lane, the mystery is instead why it took so long for the oxygen-producing form of photosynthesis to evolve. It may not have emerged until around 2.4 billion years ago, perhaps a billion years after anoxygenic photosynthesis appeared. Given the advantages of oxygen-producing photosynthesis, why the delay?
Photosynthesis has two main steps. In the second, electrons are added to CO2 to help convert the molecule into sugars. The first step is getting the electrons. They are stripped from a source molecule and used to generate an electrochemical gradient that powers the second step.
The billion-year delay
In oxygenic photosynthesis, the source molecule is water. Removing electrons splits water molecules into hydrogen ions and oxygen gas. The hydrogen ions and electrons play a key role in turning CO2 into sugars. The oxygen, though, is an unneeded by-product.
In anoxygenic photosynthesis, different molecules provide the electrons. One of the most common donors is hydrogen sulphide. Splitting it generates sulphur as a waste product instead of oxygen. The advantage of hydrogen sulphide is that it is very easy to remove electrons from, or oxidise. It was also common in the early ocean, but probably got used up quickly in surface waters where anoxygenic photosynthesis took place.
The great advantage of using water as the electron donor instead is that there is an endless supply of it in the oceans. But there is a big drawback, too. “Water is incredibly difficult to oxidise,” says Robert Blankenship at Washington University in St Louis, Missouri. We’re still struggling to do it: researchers have been trying for decades to develop cheap, energy-efficient ways of splitting water to produce hydrogen gas for fuel.
So it makes sense that photosynthesising bacteria first exploited easy-to-oxidise molecules before switching to water. The conventional view, supported by Blankenship and many other researchers, is that oxygenic photosynthesis gradually evolved from the anoxygenic version through a series of intermediate steps. But over the past decade, John Allen at Queen Mary, University of London, has devised an alternative scenario that is almost deliberately implausible. “This process has to have happened by accident,” he says. Only that can explain the billion-year delay, he argues.
Any scenario for how oxygenic photosynthesis got started has to deal with four significant facts. Fact one: there are two related but distinct types of anoxygenic photosynthesis. Some bacteria have what is called a type-I reaction centre, which takes electrons from sources like hydrogen sulphide and sends them down a one-way street: each electron is used just once. Other bacteria carry a type-II reaction centre that recycles electrons internally, making them less dependent on an external electron source (see illustration).
Fact two: oxygenic photosynthesis involves a type-I and a type-II reaction centre working in tandem. Fact three: even though cyanobacteria have both reaction centres, it is only the type-II centre that splits water and generates oxygen, at a site that contains four manganese atoms arranged around a calcium atom. Finally, fact four: anoxygenic photosynthetic bacteria that have a type-II reaction centre lack this cluster of manganese and calcium.
Blankenship thinks it is the final two facts that are most important and point towards a simple scenario. The type-I centre evolved first, he thinks. Then the genes encoding its machinery were acquired by another group of bacteria – gene-swapping was and is rife among bacteria. In this group, the machinery gradually became modified, forming the first type-II reaction centre. Later, the descendants of these bacteria began to incorporate metal atoms into it. Eventually they arrived at a configuration that included four atoms of manganese and one of calcium. They could now oxidise water and perform oxygen-generating photosynthesis using just a type-II reaction centre.
Only later, claims Blankenship, did this group’s descendants acquire the type-I machinery via gene transfer, giving rise to cyanobacteria. So Blankenship thinks it is just a coincidence that cyanobacteria have two different reaction centres.
This scenario makes one clear prediction – there were once bacteria that generated oxygen through photosynthesis, but were distinct from cyanobacteria. They would have been the missing link between the anoxygenic bacteria with a type-II reaction centre – including what are called purple bacteria, alive today – and the oxygen-generating cyanobacteria, so let’s call them “indigo” bacteria. No indigo bacteria have ever been found, though. Instead, Blankenship and others have tried to show that they could have existed.
Perhaps most significantly, a team at Arizona State University in Tempe has tried to turn a purple bacterium into something like an indigo bacterium. The researchers modified the purple one so it could bind a manganese ion to its reaction centre and use it to react with molecules containing oxygen ( It’s not oxygenic photosynthesis, but it’s a step towards it.
Marine disaster
Even if biologists do one day engineer an indigo bacterium in the lab, though, this wouldn’t prove they could evolve naturally. And to Allen, the gradual evolution scenario cannot explain all the facts. Why would such an apparently simple sequence of events have taken up to a billion years to occur? Why did oxygenic photosynthesis evolve only once, in cyanobacteria, as far as we know? (Plants acquired the ability to photosynthesise by allowing cyanobacteria to live inside them – their chloroplasts are descended from cyanobacteria.) And why do all cyanobacteria have both kinds of reaction centres?
Allen also thinks the type-I centre evolved first. But from there, his scenario is very different. Allen thinks that early in their history, these bacteria experienced some kind of genetic glitch which duplicated the entire set of genes for making a type-I reaction centre. The spare copy was free to take on a different role, and it evolved the ability to recycle electrons – the first type II reaction centre. Having two distinct reaction centres allowed these “proto-cyanobacteria” to thrive in a wide range of environments, Allen proposes. When there was plenty of hydrogen sulphide, they used their type-I reaction centre. When hydrogen sulphide ran low, the bacteria switched to using their type-II reaction centre, recycling the electrons they had gathered.
Then one day, disaster struck: some proto-cyanobacteria drifted into a shallow marine environment rich in manganese but poor in hydrogen sulphide. The bacteria duly switched to a type-II reaction. But when ultraviolet light hits manganese it strips off electrons, so there were actually plenty available – and these electrons quickly clogged the cyclic type-II reaction centre. The resulting manganese ions would have reacted with water to form manganese oxide, but there was plenty more manganese around, producing more than enough electrons to kill the microbes.
Well, almost all of them. One lucky proto-cyanobacterium survived, Allen suggests, because a mutation wrecked the switch that turned on only one kind of reaction centre at any time. With both kinds in action together, electrons from the manganese could flow through the type-II centre before being siphoned off by the type-I centre, preventing a blockage. In other words, the two reaction centres would have been working together, just as they do in cyanobacteria today ().
But how did the descendants of this bacterium go from getting electrons from manganese to getting them from water? Well, in a way they didn’t. To this day manganese provides the electrons needed for photosynthesis in all plants. However, the electrons now come from a cluster of manganese atoms within the type-II reaction centre, and this cluster has a remarkable ability – after it has given up electrons, it steals others from water molecules, splitting them apart and liberating oxygen.
Once early cyanobacteria had evolved this kind of type-II centre, they needed only trace amounts of manganese. They could then spread from manganese-rich environments and start exploiting the abundant CO2 available at the time, with the help of an unlimited supply of water and sunshine. Soon immense numbers of cyanobacteria were spewing out enough oxygen to transform the atmosphere.
If Allen’s hypothesis is correct, proto-cyanobacteria had to stumble into a highly unusual manganese-rich environment and lose control of a key genetic switch at the same time. Allen agrees this is improbable, but this could be why oxygenic photosynthesis took a billion years to appear. “The way I look at it, it was only a matter of time until one of these bacteria had two accidents at once,” he says.
Remarkably, there is now hard evidence to back Allen’s idea: we’ve found one of those rare manganese-rich environments. Woodward Fischer at the California Institute of Technology in Pasadena and his colleagues have been studying rocks laid down in what is now South Africa just before levels of oxygen began to rise. In one spot they found a superabundance of manganese oxide in rocks that formed, significantly, in the absence of oxygen. Not even ultraviolet light could have generated manganese oxide on the scale found in the rocks. This leaves photosynthesis as it existed in Allen’s proto-cyanobacteria as the only plausible scenario, the team told a meeting in December.
“It is big news, hugely exciting – and spot on for John’s hypothesis,” says William Martin at Heinrich Heine University in Düsseldorf, Germany, who studies early evolution. Martin is a supporter of Allen’s scenario, and has been working with him to gather supporting evidence. But Blankenship is sticking to his guns. He describes his many discussions with Allen and Martin on the origin of oxygenic photosynthesis as “very spirited, yet friendly”.
What would settle the debate once and for all is the discovery of living representatives of one of the proposed intermediate forms – either indigo bacteria or proto-cyanobacteria. Surprisingly, Blankenship and Allen are both confident that their respective organisms still exist somewhere in the world. “You find specialist environments today that correspond to typical conditions 2.4 billion years ago,” says Allen. “It’s not absurd to think that these microorganisms are still out there.”
Whatever the ancestor of cyanobacteria turns out to be like, we have reason to be very grateful to it. “This organism – maybe by accident – was hugely important,” says Allen. “Quite simply, it changed the world forever.”
