Power, Sex, Suicide: Mitochondria and the Meaning of Life (10 page)

This scenario is only one of several possibilities, and perhaps the timing holds the key. Even if energy was not the basis of the relationship, the rise in oxygen levels might still explain the initial benefits. Oxygen is toxic to anaerobic (oxygen-hating) organisms—it ‘corrodes’ unprotected cells in the same way that it rusts iron nails. If the guest was an aerobic bacterium, using oxygen to generate its energy, while the host was an anaerobic cell (generating energy by fermentation), then the aerobic bacterium may have protected its host against toxic oxygen—it could have worked as an internally fitted ‘catalytic converter’, guzzling up oxygen from the surroundings and converting it into harmless water. Siv Andersson calls this the ‘Ox-Tox’ hypothesis.

Let’s recapitulate the argument. A bacterium loses its cell wall but survives because it has an internal cytoskeleton, which it had made use of before to keep in shape. It now resembles a modern archaeon. With a few modifications to its cytoskeleton, the wall-less archaeon learns to eat food by phagocytosis. As it grows larger it wraps its genes in a membrane and develops a nucleus. It has now turned into an archezoon, perhaps resembling cells like
Giardia
. One such hungry archezoon happens to engulf a smaller aerobic bacterium but fails to digest it, let’s say because the bacterium is a parasite like the modern
Rickettsia
,
and has learned to evade the defences of its host. The two get along together in a benign parasitic relationship, but as atmospheric oxygen levels rise, the relationship begins to pay dividends to both the host and parasite: the parasite still gets its free lunch, but the host is now getting a better deal—it’s protected from toxic oxygen from within by its catalytic converter. Then, finally, in an act of breathtaking ingratitude, the host plugs a ‘tap’ into the membrane of its guest and drains off its energy. The modern eukaryotic cell is born, and never looks back.

This long chain of reasoning is a good example of how science can piece together a plausible story and back it up with evidence at almost every point. To me there is a feeling of inevitability about the whole process: it could happen here and it could happen anywhere else in the universe—no single step is particularly improbable. There is simply a bottleneck, as postulated by Christian de Duve, in which the evolution of the eukaryotes is unlikely when there is not much oxygen around, but almost inevitable as soon as the oxygen levels rise. While everybody agrees this story is broadly speculative, it was widely believed to be plausible, and made use of most of the known facts. Nothing prepared the field for the reversal that was to follow in the late 1990s. As sometimes happens to the ‘good’ stories in science, virtually the entire edifice collapsed in the space of just five years. Nearly every point has now been contradicted. But perhaps the writing was on the wall. If the eukaryotes only evolved once, then a plausible story may be exactly the wrong kind of story.

The first stone to crumble was the ‘primitively amitochondriate’ status of the archezoa. This term, if you recall, means that the archezoa never did have any mitochondria. But when more genes from different archezoa were sequenced, it began to look as if postulated progenitors of eukaryotic cells, such as
Entamoeba histolytica
(the cause of amoebic dysentery), were not the earliest representatives of their group after all. Other types of cell in the same group appeared to be even older—but did have mitochondria. Unfortunately, the genetic dating techniques were approximate and liable to error, and so the results were controversial. But if the estimated dates were correct, then the results could only mean that
Entamoeba histolytica
did have ancestors that had once possessed mitochondria, and so must have lost its own, rather than never having had any at all. If the archezoa are defined as a group of primitive eukaryotes that never had mitochondria, then
E. histolytica
could not be an archezoon.

In 1995, Graham Clark at the National Institutes of Health in the United States, and Andrew Roger at Dalhousie University in Canada, went back to look
more closely at
E. histolytica
to see if there were any traces that it had formerly possessed mitochondria. There were. Hidden away in the nuclear genome were two genes that, from their DNA sequences, almost certainly derived from the original mitochondrial merger. They were presumably transferred from the early mitochondria to the host cell nucleus, and the cell later lost all physical traces that it had ever had any mitochondria. We should note that the transfer of genes from the mitochondria to the host is quite normal, for reasons that we’ll consider in
Part 3
. Modern mitochondria have retained only a handful of genes, and the rest were either lost altogether or transferred across to the nucleus. The proteins encoded by these nuclear genes are often targeted back to the mitochondria. Interestingly
E. histolytica
does actually possess some oval organelles that might be the corrupt remains of mitochondria; they resemble mitochondria in their size and shape, and several of the proteins that have been isolated from them have also been found in the mitochondria in other organisms.

Not surprisingly, the burning question transferred to the other supposedly primitively amitochondriate groups. Had they, too, once possessed mitochondria? Similar studies were carried out, and so far all the ‘archezoa’ that have been tested turn out to have once possessed mitochondria, and lost them later on. For example, not only did
Giardia
apparently once have mitochondria, but it, too, may still preserve relics, in the form of tiny organelles called mitosomes, which continue to carry out some of the functions of mitochondria (if not the best known, aerobic respiration). Perhaps the most surprising results concerned the microsporidia. This supposedly ancient group not only
did
possess mitochondria in the past, but now turns out
not
to be an ancient group at all—they are most closely related to the higher fungi, a relatively recent group of eukaryotes. The apparent antiquity of the microsporidia is merely an artefact of their parasitic lifestyle inside other cells. And the fact that they infect so many different groups is but a testament to their success.

While it remains possible that the real archezoa are still out there, just waiting to be found, the consensus view today is that the entire group is a mirage—every single eukaryote that has ever been examined either has, or once had, mitochondria. If we believe the evidence, then there never were any primitive archezoa. And if this is true, then the mitochondrial merger took place at the very beginning of the eukaryotic line, and was perhaps inseparable from it: the merger
was
the unique event that gave rise to the eukaryotes.

If the prototype eukaryote was not an archezoon—in other words, not a simple cell that made its living by engulfing its food by phagocytosis—then what
did
it look like? The answer might possibly lie in the detailed DNA sequences of eukaryotes living today. We have seen it is possible to identify ex-mitochondrial genes by comparing their gene sequences; perhaps we can
do the same with those genes inherited from the original host. The idea is simple. Because we know that mitochondria are related to a particular group of bacteria, the α-proteobacteria, we can exclude any genes that seem to derive from this source, and look to see where the rest come from. Of the rest, we can assume that some are unique to eukaryotes—they evolved in the last two thousand million years since the merger—while some might have been transferred from elsewhere. Even so, at least a few ought to line up with the original host. These genes should have been inherited by all the descendents of the original merger, and gradually accumulated modifications ever since; but they should still bear
some
resemblance to the original host cell.

This was the approach employed by Maria Rivera and her colleagues at the University of California, Los Angeles, published in 1998 and in more detail in
Nature
in 2004. This team compared complete genome sequences from representatives of each of the three domains of life, and found that eukaryotes possess two distinct classes of genes, which they referred to as
informational
and
operational
genes. The
informational
genes encoded all the fundamental inheritance machinery of the cell, enabling it to copy and transcribe DNA, to replicate itself, and to build proteins. The
operational
genes encoded the workaday proteins involved in cellular metabolism—in other words, the proteins responsible for generating energy and manufacturing the basic building blocks of life, such as lipids and amino acids. Interestingly, almost all the operational genes came from the α-proteobacteria, presumably by way of the mitochondria, and the only real surprise was how many more of these genes there were than expected—it seems the genetic contribution of the ancestor of the mitochondria was greater than anticipated. But the biggest surprise was the allegiance of the
informational
genes. These genes lined up with the archaea, as anticipated, but they bore a strong resemblance to the genes in a completely unexpected group of archaea: they were most similar to
methanogens
, those swamp lovers that shun oxygen and produce the marsh gas methane.

This is not the only piece of evidence to point a suspicious finger at the methanogens. John Reeve and his colleagues at Ohio State University, Columbus, have shown that the structure of eukaryotic histones (the proteins that wrap DNA) is closely related to methanogen histones. This similarity is surely no coincidence. Not only are the structures of the histones themselves closely related, but also the three-dimensional conformation of the whole DNA-protein package is amazingly similar. The chances of finding exactly the same structure in two organisms that are supposedly unrelated, like the methanogens and the eukaryotes, is equivalent to finding the same jet engine in two aeroplanes produced independently by two competing companies. Of course, we might well find the same engine, but we’d be incredulous if told that
it had been ‘invented’ twice, without any knowledge of the rival company’s version, or of the prototype: we would assume that the engine had been bought or stolen from another company. In the same way, the packaging of DNA with histones is so similar in the methanogens and the eukaryotes that the most likely explanation is that they derived the full package from a common ancestor—both were developed from the same prototype.

All this adds up to quite a package. Two tell-tale wisps of smoke curl out of the same smoking gun. If these wisps are believable, it seems we inherited both our informational genes and our histone proteins from the methanogens. Suddenly our most venerable ancestor is no longer the vile parasite that we suspected, but an even more alien entity, which survives today in stagnant swamps and the intestinal tract of animals. The original host in the eukaryotic merger was a methanogen.

We are now in a position to see what kind of a hopeful monster the first eukaryotic cell might have been—the product of a merger between a methanogen (which gained its energy by generating methane gas) and an α-proteo-bacterium, for example a parasite like
Rickettsia
. This is a startling paradox. Few organisms hate oxygen more than the methanogens do—they can only be found living in the stagnant, oxygen-free pits of the world. Conversely, few organisms depend more on oxygen than
Rickettsia
—they are tiny parasites living inside other cells, and have streamlined themselves to their specialist niche by throwing away redundant genes, leaving them with only the genes needed to reproduce themselves—and the genes needed for oxygen respiration. Everything else has gone. So the paradox is this: if the eukaryotic cell was supposedly born of a symbiosis between an oxygen-hating methanogen and an oxygen-loving bacterium, how could the methanogen possibly benefit from having α-proteobacteria inside it? For that matter, how did the α-proteobacteria benefit from being inside? Indeed, if the host was incapable of phagocytosis—and methanogens are certainly not able to change shape and eat other cells—how on earth did it get inside?

It is possible that Siv Andersson’s Ox-Tox hypothesis still applies—in other words, the oxygen-guzzling bacterium protected its host from toxic oxygen, enabling the methanogen to venture into pastures new. But there is a big difficulty with this scenario now. Such a relationship makes sense for a primitive archezoon that lives by fermenting organic remains. This will prosper if it is able to migrate to any environment where such remains can be found. Such scavenging cells are the single-celled equivalent of jackals prowling Africa, covering vast distances in the search for a fresh carcass. But this roving existence would kill a methanogen. A methanogen is as tied to a low-oxygen environment as a hippo is to waterholes. The methanogens can
tolerate
the presence of oxygen, but they can’t generate any energy in its presence, because
they depend on hydrogen for fuel, and this is very rarely found in the same environment as oxygen. So if a methanogen does leave its watering hole, it must starve until it gets back: festering organic remains mean nothing to a methanogen—it would do better never to leave. Thus there is a deep tension between the interests of the methanogen, which gains nothing from venturing to pastures new, and those of an oxygen-guzzling parasite, which can’t generate any energy at all in the anoxic environment favoured by methanogens.

This paradox is heightened because, as we have seen, their relationship could not have depended on energy in the form of exchangeable ATP—bacteria do not have ATP exporters, and never benevolently ‘feed’ each other. The tryst could still have been a parasitic relationship, in which the bacteria consumed the organic products of the methanogen from within—but again, there are problems with this, as an oxygen-dependent bacterium could not generate any energy from the innards of a methanogen unless it could persuade the methanogen to leave its waterhole, those comfortable oxygen-free surroundings. One might picture the α-proteobacteria herding the methanogens and driving them like cattle to an oxygen-rich slaughter field, but for bacteria this is nonsense. In short, the methanogens would starve if they left their waterhole; the oxygen-dependent bacteria would starve if they lived in the waterhole, and the middle ground, a little oxygen, must have been equally disadvantageous to both parties. Such a relationship seems to be mutually insufferable—is this really how the stable symbiotic relationship of the eukaryotic cell began? It is not just improbable, but downright preposterous. Luckily there is another possibility, which until recently seemed fanciful, but is now looking far more persuasive.