by Nick Lane
If eukaryotic cells have things inside, bacteria are inscrutable. They have little of the eukaryotic riot of internal membrane systems, apart from their single external cell membrane, which is occasionally folded in upon itself to give some texture to the cell. Even so, the flourishing eukaryotic membranes share the same basic composition with the sparse membranes of bacteria. Both are composed of a water-soluble ‘head’ of glycerol phosphate, to which are bound several long fatty chains, which are soluble only in oils. Just as detergents form naturally into tiny droplets, so too the chemical structure of lipids enables them to coalesce naturally into membranes, in which the fatty chains are buried inside the membrane, while the water-soluble heads protrude from either side. This kind of consistency in the bacteria and eukaryotes helps to convince biochemists that both ultimately share a common inheritance.
Before we move on to consider the meaning of all these similarities and differences, let’s just complete our whistlestop tour of the eukaryotic cell. There are two remaining differences with bacteria that I’d like to touch on. First, besides their membrane structures and organelles, eukaryotic cells contain a dense internal scaffolding of protein fibres known as the cytoskeleton. Second, unlike bacteria, eukaryotes do not have a cell wall, or at least not a bacterial-style cell wall (plant cells, and some algae and fungi, do actually possess cell walls, but these are very different from bacterial walls, and evolved much later).
The internal cytoskeleton and the external cell wall are utterly different conceptions, but nonetheless have equivalent functions—both provide structural support, in the same way that the external cuticle of an insect and own internal skeleton both provide structural support. Bacterial cell walls vary in structure and composition, but in general they provide a rigid skeleton that maintains the shape of the bacterium, preventing it from swelling to bursting point, or collapsing, if its environment suddenly changes. In addition, the bacterial cell wall provides a solid surface for anchoring the chromosome (containing its genes), along with various locomotive devices, such as whip-like threads, or flagellae. In contrast, eukaryotic cells usually have a flexible outer membrane, which is stabilized by the internal cytoskeleton. This is not at all a fixed structure but is constantly being remodelled—a highly energetic process—giving the cytoskeleton a dynamism unattainable by a cell wall. This means that eukaryotic cells (or protozoa at least) are not as robust as bacteria, but they have the immeasurable advantage that they can change shape, often quite vigorously. The classic example is the amoeba, which crawls around and engulfs its food by phagocytosis: temporary cellular projections, known as pseudopodia (literally, false feet) flow around the prey and meld together again, forming a food vacuole inside the cell. The pseudopodia are stabilized by dynamic changes in the cytoskeleton. They meld together again so easily because the lipid membranes are as fluid as soap bubbles, and can easily bud off into vesicles, then meld back together. Their ability to change shape and engulf food by phagocytosis enables single-celled eukaryotic organisms to become true predators, setting them apart from bacteria.
The road less travelled—from bacteria to eukaryotes
Eukaryotic cells and bacteria are constructed from essentially the same building materials (nucleic acids, proteins, lipids, and carbohydrates). They have exactly the same genetic codes, and very similar membrane lipids. Clearly they share a common inheritance. On the other hand, the eukaryotes are different from bacteria in virtually every aspect of their structure. Eukaryotic cells are, on average, 10 000 to 100 000 times the volume of bacteria, and contain a nucleus and many membranes and organelles. They generally carry orders of magnitude more genetic material and fragment their genes into short sections, in no particular order. Their chromosomes are straight rather than circular, and are wrapped in histone proteins. Most reproduce by sex, at least occasionally. They are supported internally by a dynamic cytoskeleton and may lack an external cell wall, which enables them to scavenge food and ingest whole bacteria.
The mitochondria are only one element in this catalogue of differences, and might seem to be just another added extra. They are not, as we shall see. But we are left with the question: why did the eukaryotes make such a complicated evolutionary pilgrimage while bacteria barely changed in nearly four billion years?
The origin of the eukaryotic cell is one of the hottest topics in biology, what Richard Dawkins has termed the ‘Great Historic Rendezvous’. It furnishes exactly the right balance of science and speculation to generate violent passions among supposedly dispassionate scientists. Indeed, it sometimes feels as if each new piece of evidence throws up a new hypothesis to explain the evolutionary roots of the eukaryotic cell. Such hypotheses have traditionally fallen into two groups, those which try to explain the eukaryotes on the basis of mergers between a variety of bacterial cells, and those which try to derive most eukaryotic features from within the group, without recourse to so many mergers. As we saw in the Introduction, Lynn Margulis argued that both mitochondria and chloroplasts are derived from free-living bacteria. She also argued that several other features of eukaryotic cells, including the cytoskeleton, along with its organizing centres, the centrioles, are derived from bacterial mergers, but she has been less successful at drawing the field with her. The problem is that resemblances in cellular structures may derive from a direct evolutionary relationship, in which the endosymbiont has degenerated to the point that its ancestry can only just be made out. Alternatively, similarities in structure may be the result of convergent evolution, in which similar selection pressures inevitably generate similar structures, as there are only a few possible engineering solutions to a particular problem, as discussed earlier.
In the case of cellular objects like the cytoskeleton, which, unlike the chloroplasts and the mitochondria, do not have a genome of their own, it’s difficult to establish provenance. If genealogy can’t be traced directly, it is not easy to prove whether an organelle is symbiotic or an invention of the eukaryotes. Most biologists lean towards the simplest view, that most eukaryotic traits, including the nucleus and the organelles, except for the mitochondria and chloroplasts, are purely eukaryotic inventions.
To trace a path through this maze of contradictions we’ll consider just two of the competing theories on the origin of the eukaryotic cell, which seem to me to be the most likely possibilities—the ‘mainstream’ view and the ‘hydrogen hypothesis’. The mainstream view has superseded Lynn Margulis’ original ideas in many details, and in its present form is largely attributable to Oxford biologist Tom Cavalier-Smith. Few researchers have quite as detailed an understanding of the molecular structures of cells and their evolutionary relationships as Cavalier-Smith, and he has put forward numerous important and contentious theories on cellular evolution. The hydrogen hypothesis is an utterly different theory, argued forcefully by Bill Martin, an American biochemist at Heinrich-Heine University in Düsseldorf, Germany. Martin is a geneticist by background, and tends to prefer biochemical, rather than structural, insights into the origins of the eukaryotes. His ideas are counter-intuitive, and have generated a heated, even vitriolic, response in some quarters, but they are underpinned by a crisp ecological logic that cannot be ignored. The pair often clash at conferences, and their views seem to hang over such meetings with an almost Victorian sense of melodrama, reminiscent of Conan Doyle’s Professor Challenger. At a splendid discussion meeting on the origin of eukaryotic cells at the Royal Society of London in 2002, Cavalier-Smith and Martin contested each other’s views throughout the meeting, and I was impressed to find them still embroiled in debate hours afterwards in the local pub.
2
Quest for a Progenitor
How did the eukaryotic cell evolve from bacteria? The mainstream view assumes that it was by way of a sequence of tiny steps, through which a bacterium was gradually transformed into a primitive eukaryotic cell, possessing everything that characterises the modern eukaryotes, except for mitochondria. But what were these steps? And how did they get starte
d down a path that in the end found a way across the deep chasm separating the eukaryotes from bacteria?
Tom Cavalier-Smith has argued that the key step forcing the evolution of the eukaryotes was the catastrophic loss of the cell wall. According to the Oxford English Dictionary, the word ‘catastrophe’ means ‘a calamitous fate’ or ‘an event producing a subversion of the order of things’. For any bacteria that lose their cell wall, either definition may easily come true. Most wall-less bacteria are extremely fragile, and unlikely to survive long outside the cosy laboratory environment. This does not mean that such calamities are rare events, though. In the wild, bacterial cell walls might be lost quite often, either by mutation or active sabotage. For example, some antibiotics (such as penicillin) work by blocking the formation of the cell wall. Bacteria engaged in chemical warfare may well have produced such antibiotics. This is not at all improbable—most new antibiotics are isolated from bacteria and fungi engaged in exactly this kind of struggle. So, the first step, the calamitous loss of the cell wall, might not have posed any problem. What of the second step: survival and subversion of the order of things?
As we noted in the previous chapter, there are potentially big advantages to getting rid of the unwieldy cell wall, not least being able to change shape and engulf food whole by phagocytosis. According to Cavalier-Smith, phagocytosis is the defining feature that set the eukaryotes apart from bacteria. Any bacterium that solved the problem of structural support and movement could certainly subvert the established order of things. Yet, for a long time, it looked as if surviving without a cell wall was a magic trick equivalent to pulling a rabbit out of a hat. Bacteria were believed to lack an internal cytoskeleton, and if that was the case, the eukaryotes must have evolved their complex skeleton in a single generation, or faced extinction. In fact this assumption turns out to be groundless. In two seminal papers, published in the journals Cell and Nature in 2001, Laura Jones and her colleagues at Oxford, and Fusinita van den Ent and her colleagues in Cambridge, showed that some bacteria do indeed have a cyto-skeleton as well as a cell wall—they wear a belt and braces, as Henry Fonda put it in Once Upon a Time in the West (‘never trust a man who can’t even trust his own trousers’). Unlike Fonda’s risk-averse cowboy, however, bacteria do need both to maintain their shape.
Many bacteria are spherical (cocci) while others are rod-shaped (bacilli), filamentous, or helical. Some oddballs have been found that even have triangular or square shapes. Quite what advantages these different shapes might confer is an interesting question, but it seems that the default bacterial shape is spherical, and any other shape requires internal support. Non-spherical bacteria possess protein filaments very similar in microscopic structure to those found in eukaryotes like yeast, as well as in humans and plants. In each case, the cytoskeleton filaments are composed of a protein akin to actin, best known for its role in muscle contraction. In non-spherical bacteria, these filaments form into a helical swirl underneath the cell membrane, which apparently provides structural support. What is clear is that if the genes encoding the filaments are deleted then bacteria that are normally rod-like in shape (bacilli) develop as spherical cocci instead. Impressions resembling bacilli have been found in rocks 3500 million years old, so it is conceivable that the cytoskeleton evolved not long after the appearance of the earliest cells. This reverses the problem. If a cytoskeleton was there all along, then why do so few bacteria survive the loss of the cell wall? We’ll return to this theme in Part 3. For now, let’s satisfy ourselves with the possible consequences.
‘Discovery’ of the archaea—a missing link?
Only two groups of cells have thrived in the absence of a cell wall—the eukaryotes themselves, and the Archaea, a remarkable group of prokaryotes (cells that lack a nucleus, like bacteria). The Archaea were discovered by Carl Woese and George Fox at the University of Illinois in 1977, and named from the Greek for ‘ancient’. Most archaea do, in fact, have a cell wall, but their walls are rather different in chemical composition from those of bacteria, and some groups (such as the boiling-acid loving Thermoplasma) do not have a cell wall at all. Curiously, antibiotics like penicillin don’t affect the synthesis of archaeal cell walls, lending support to the idea that cell walls might have been the target of bacterial chemical warfare. Like bacteria, archaea are tiny, typically measuring a few thousandths of a millimetre (microns) across, and they do not have a nucleus. Like bacteria, they have a single circular chromosome. Again, like bacteria, the archaea take on many shapes and forms, and so presumably have some sort of cytoskeleton. One reason why they were discovered so recently is that archaea are mostly ‘extremophiles’, that is, they thrive in the most extreme and arcane of environments, from boiling acid-baths beloved of Thermo-plasma, to putrid marshes (inhabited by marsh-gas producing methanogens) and even buried oilfields. In the latter case, the archaea responsible have attracted commercial interest, or rather annoyance, as they ‘sour’ the wells—they raise the sulphur content of oil, which corrodes the well-casings and metal pipelines. Greenpeace could hardly conceive a more wily sabotage.
The ‘discovery’ of the archaea is a relative term, as some of them had been known about for decades (particularly the oil-souring archaea and swamp-gas producing methanogens), but their small size and lack of nucleus meant that they were invariably mistaken for bacteria. In other words, they were not so much discovered as reclassified; and even now, some researchers prefer to classify them with the bacteria, as just another diverse group of inventive prokaryotes. But the painstaking genetic studies of Woese and others have convinced most impartial observers that the archaea really do differ in profound ways from bacteria, ways that go well beyond the construction of their cell walls. We now know that about 30 per cent of archaeal genes are unique to the group. These unique genes code for forms of energy metabolism (such as the generation of methane gas) and cell structures (such as membrane lipids) that are not found in any other bacteria. The differences are important enough for most scientists to regard the archaea as a separate ‘domain’ of life. This means that we now classify all living things into three great domains—the bacteria, the archaea, and the eukaryotes (which, as we have seen, includes all multicellular plants, animals, and fungi). The bacteria and the archaea are both prokaryotic (lacking a cell nucleus) while the eukaryotes all do have a nucleus.
Despite their love of extreme environments and unique characteristics, the archaea also share a mosaic of traits with both bacteria and eukaryotes. I say ‘mosaic’ advisedly, as many of these traits are self-contained modules, encoded by groups of genes that work together as a unit (such as the genes for protein synthesis, or for energy metabolism). These individual modules fit together like the pieces of a mosaic, to construct the overall pattern of an organism. In the case of the archaea, some pieces are similar to those used by eukaryotes, while others are more reminiscent of bacteria. It is almost as if they were selected at random from a lucky dip of cell characteristics. So, for example, even though the archaea are prokaryotes, easily mistaken for bacteria when viewed down the microscope, some of them nonetheless wrap their chromosome in histone proteins, in a very similar manner to eukaryotes.
The parallels between archaea and eukaryotes go further. The presence of histones means that archaeal DNA is not easily accessible, so, like the eukaryotes, archaea need complicated transcription factors to copy or to transcribe their DNA (reading off the genetic code to construct a protein). The detailed mechanism of genetic transcription in the archaea parallels that in eukaryotes, albeit in a simpler fashion. There are also similarities in the way that the two groups construct their proteins. As we saw in the Introduction, all cells assemble their proteins using the tiny molecular factories called ribosomes. The ribosomes are broadly similar in all three domains of life, implying that they share a common ancestry, but they differ in many details. Interestingly, there are more differences between the bacterial and archaeal ribosomes than there are between archaeal and eukaryotic ribosomes. For exa
mple, toxins like diphtheria toxin block protein assembly on ribosomes in both the archaea and eukaryotes, but not in bacteria. Antibiotics like chloramphenicol, streptomycin, and kanamycin block protein synthesis in the bacteria, but not in the archaea or eukaryotes. These patterns are explained by differences in the way that protein synthesis is initiated, and in the detailed structure of the ribosome factories themselves. The ribosomes of eukaryotes and archaea have more in common with each other than either do with bacteria.
All this means the archaea are about as close to a missing link between the bacteria and the eukaryotes as we are ever likely to find. The archaea and the eukaryotes probably share a relatively recent common ancestor, and are best seen as ‘sister’ groups. This seems to back up Cavalier-Smith’s view that the loss of the cell wall, possibly in the common ancestor of the archaea and the eukaryotes, was the catastrophic step that later propelled the evolution of eukaryotes. The earliest eukaryotes may have looked a little like modern archaea. Intriguingly, though, no archaea ever learnt to change shape to scavenge a living by engulfing food in the eukaryotic fashion. On the contrary, instead of developing a flexible cytoskeleton as the eukaryotes did, the archaea developed quite a stiff membrane system, and remained nearly as rigid as bacterial cells. So there is more to being ‘eukaryotic’ than just lacking a cell wall; but might it be no more complex than lifestyle? Were the ancestral eukaryotes simply wall-less archaea, which modified their existing cytoskeleton into a more dynamic scaffolding that enabled them to change shape and eat food in lumps, by phagocytosis? Might this alone account for how they came by their mitochondria—they simply ate them? And if so, might there still be a few living fossils from the age before mitochondria lurking in hidden corners, relics of those primitive eukaryotes that shared more traits with the archaea?