“THE past isn’t dead,” wrote William Faulkner. “It isn’t even past.” As far
as biology goes, he had a point. Relics of the most distant biological past live
on in every creature on Earth.
Now a team of researchers have pieced together some of these relics to
produce a surprising picture of our primitive ancestors. They say that, in
certain ways, the earliest life on Earth was more like the complex cells of
animals and plants than the bacteria normally regarded as life’s most ancient
ancestors. If they’re correct, evolutionary biologists will have to redraw the
whole tree of life and re-evaluate the origins of modern cells.
The researchers, biologists from Massey University in New Zealand, started
with the assumption that life began in an RNA world. Today, most RNA molecules
act as cellular intermediaries, converting the genetic information stored in DNA
into the proteins that provide cells with structure and carry out their
metabolic reactions. But according to one school of thought, which is popular
but far from universally accepted (see “Look, No Genome”), RNA used to do
much more. At one time, it may have stored genetic information and catalysed
metabolic reactions, performing the roles that DNA and proteins play today.
Thus, RNA on its own might once have handled all the tasks necessary for a cell
to survive (“Let there be life”, Â鶹´«Ă˝, 6 July 1996, p 22).
Advertisement
This RNA world, the theory goes, arose nearly four billion years ago, before
the earliest fossils were formed and before the most ancient threads of DNA
locked themselves into their first twisty-turny embrace. It came to an end when
a new division of labour evolved, with more stable DNA taking over the
information-storing duties and more versatile proteins taking over the chemical
engineering tasks.
Although the RNA world is gone, its ghosts persist within the molecular
machinery of the modern cell. And according to Anthony Poole, Daniel Jeffares
and David Penny of Massey University, some RNA molecules around today pre-date
DNA and protein, though their functions may have changed through the aeons.
Studying these RNA relics, the researchers developed a picture of a hypothetical
RNA-based organism that they call Riborgis eigensis. This fanciful
beast—the RNA organism that ruled before proteins came into
vogue—made its debut last year in the Journal of Molecular
Evolution (vol 46, p 1 and p 18). Poole and his colleagues make a good case
for the idea that Riborgis would have kept its genes strung out in
pieces along a set of linear chromosomes, the organisation preferred by most
complex cells today.
Why should an RNA organism divide its genome—its entire collection of
genetic instructions—into chromosomes? RNA is less stable than DNA, so RNA
copying mechanisms tend to have high error rates. And the larger the genome, the
greater the chances of catastrophic copying mistakes turning gene sequences into
garbage. Such propensity for error puts a limit on how large a genome can be.
This cutoff size is called the Eigen limit, after the German biologist Manfred
Eigen.
One way an organism can get around the Eigen limit is to cut its genome into
chunks and keep several copies of each bit. If copying mistakes damage one copy
of an important gene, the cell still has others. There seems no reason why
Riborgis should not have availed itself of this molecular “loophole”, and
there’s some evidence that it did, argue the New Zealand researchers.
If it did, then in one important respect Riborgis was more like the
complex cells of plants and animals than like bacteria. Eukaryotic cells, such
as those of plants, animals and some single-celled organisms, split their
genomes into chromosomes. Prokaryotic cells such as bacteria, on the other hand,
are much simpler and generally store only one copy of each gene on a single,
circular chromosome.
One of the problems with linear chromosomes is that their ends tend to fray.
Eukaryotes get round this with telomerase, an enzyme that maintains
telomeres—the structures that cap the ends of chromosomes. Telomerase, it
turns out, contains an RNA molecule that helps guide its work. Riborgis
might have used such a functional RNA to care for its chromosomes, say Poole and
his colleagues.
The logic? Proteins are better catalysts than RNA, so it seems reasonable
that once proteins appeared on the scene they would have taken on all the new
catalytic jobs that evolution threw up. Jobs still being performed by RNA, then,
are likely to be ones that existed before proteins did. So if looking after
chromosome ends requires active RNAs, then chromosome ends must have been around
before proteins appeared, in RNA-based beasts like Riborgis.
Prokaryotes, the researchers note, seem to lack telomerase. They don’t need it
because their circular chromosomes have no ends. Prokaryotes may have lost these
catalytic RNAs after they evolved from more complex cells.
This sort of reasoning, plausible but not conclusive, underlies most of the
team’s arguments. They apply a similar logic to their analysis of the most
complex RNA-based machine left in existence, the ribosome.
Modern ribosomes are molecular factories that assemble amino acids into
proteins. They’re made of large pieces of RNA studded with dozens of ancillary
proteins and smaller RNAs. To make a protein, a ribosome reads the genetic
information encoded in messenger RNA (mRNA) and adds the appropriate amino acids
to the growing peptide chain (see Diagram).
Ribosomal RNAs act as enzymes, and thus might be considered relics of the RNA
world. But this world, by definition, had no need of the proteins that ribosomes
now make. So what did the earliest ribosomes do?
According to the New Zealand researchers and others, ribosomes were
originally used to replicate RNA. In an organism like Riborgis, they
say, protoribosomes would pluck small strands of RNA from their surroundings. If
one of these fragments matched the sequence on the RNA template, it would be cut
and pasted into the growing RNA strand.
How did ribosomes get into the protein-making business? Perhaps these little
strands of RNA started associating with free-floating amino acids—to boost
their stability, for example. And joining a few of these amino acids together
might have offered additional advantages. The resulting peptide chains might
have acted as a scaffolding to support larger pieces of RNA, such as the RNA
that makes up the ribosome itself.
In fact, some of the proteins that appear to be the oldest—those that
crop up in similar forms all over the tree of life— tend not to be
catalysts, but structural proteins, such as the proteins that adorn the
ribosome, says Penny. This is just the sort of protein, he argues, that an
evolving protoribosome would be able to put together. Such proteins might not
need to be assembled with every amino acid just so, so they could be made before
a precise genetic code had evolved. Selection pressure would then have made the
code more robust so that cells could reliably produce the most effective forms
of these proteins.
In an organism such as Riborgis, these early protein-coding bits of
RNA might have arisen in “spacers” that separated the “genes” that coded for
catalytic RNAs, the New Zealand researchers speculate. Such spacers might have
restrained the catalytic RNA sequences, preventing them from cutting themselves
up in a self-catalytic frenzy at times when they were meant to be passively
storing information. Because the exact sequence of the bases in the spacers
probably did not matter, these regions might have provided a fertile ground in
which a protein-coding system could start to grow.
This kind of organisation—sequences that code for catalytic RNA
molecules interspersed between sequences that encode primitive
proteins—mimics the gene structure found in almost all higher eukaryotes,
including animals, plants and some complex single-celled organisms such as
yeast. Eukaryotic genes are typically fragmented into sequences called exons and
sequences called introns. Only the exons encode proteins. When a gene is
transcribed from DNA to RNA, the introns that separate the protein-coding exons
are snipped out—normally by other RNA molecules—and the exons are
sewn together, forming an mRNA that is ready for the ribosomes.
That’s not to say that introns are trash. Quite a few code for RNA fragments
that have important functions of their own. Some are involved in splicing other
introns out of RNA transcripts. Others belong to a specific family of RNAs,
called small nucleolar RNAs, whose members are needed to trim the RNA molecules
that eventually become part of the ribosome. According to the Massey
researchers, these functional RNAs are commonly found in genes in which the
exons code for nonspecific scaffolding-type proteins, the sort they think must
have evolved earliest. Poole points out that two genes have now been discovered
in which the exons don’t code for any protein at all: the noncoding RNA introns
are all that matters. In both cases, the small RNAs from the introns are
involved in processing the much larger RNAs that make up the ribosome. All these
functional introns are prime candidates to be considered as relics of the RNA
world.
Bacteria and other prokaryotes lack the intron/exon arrangement. So if Penny,
Poole and Jeffares are correct—and introns did arise early in
evolution—then RNA-based organisms must have resembled today’s complex
eukaryotes in their genome organisation and in their chromosome structure. This
means that organisms without introns, including bacteria, must have lost them,
as devotees of the early-intron argument claim (see “Bacteria rule, OK?”,
Â鶹´«Ă˝, 3 June 1995, p 34).
If this is true, then the simplicity of the bacterial genome—a single
loop of DNA in most cases—is evidence not of primitiveness, but of a
certain streamlined sophistication. By shedding the complexities of the
eukaryotic genome, bacteria may have become capable of exploring or surviving in
environments that no eukaryote could tolerate. A simpler genome might have
allowed bacteria to reproduce more quickly and efficiently, or to withstand high
temperatures at which RNA is much less stable. In such circumstances, there
might be no time for the complicated splicing and processing needed, for
example, to remove introns from mRNAs. This evolutionary downsizing might well
have set the stage for prokaryotes’ successful colonisation of virtually every
ecosystem on the planet.
Fundamentally flawed?
Today, when molecular genealogists draw life’s family trees, they still place
prokaryotes at the root, with eukaryotes evolving much later. But if the New
Zealand team is right, this picture is fundamentally flawed. According to their
theory, some sort of creature with a eukaryotic way of organising its genome
gave rise to both prokaryotes and eukaryotes.
But many researchers who study the origins of life think the whole concept of
Riborgis is dubious. There’s a lot about the imaginary RNA organism
that the New Zealanders have not yet begun to speculate about, such as where it
got its energy. And much of what they have published is at best an argument for
the plausibility of such RNA-based organisms, rather than proof.
The main problem, say critics, lies in the assumption that modern RNAs
provide a direct window to the evolutionary past. “The idea that RNAs have been
frozen in evolution for 3 billion years is hopelessly naive,” says Gary Olsen of
the University of Illinois. “If the rest of the cellular components have
evolved, developed and improved for all of these years, how could it possibly be
that the RNAs have been protected from similar improvements?”
It’s now up to Poole and his colleagues to uncover compelling evidence for
the idea that the first genomes were eukaryotic. Studying the entire genome
sequences of primitive eukaryotes, such as the protozoan parasite
Giardia, might provide such evidence. And supporting arguments may come
from the test tube. Poole is now in Sweden seeing if he can coax random
sequences of RNA into producing nonspecific protein.
If he succeeds, Poole will have shown that the first exons could have evolved
in the way that he and his colleagues suggest. In the end, the best way to show
that a beast like Riborgis existed may be to re-create it.
While some researchers debate what type of genome Earth’s early life forms
may have had, others are delving even further back into the origins of life.
“Unless you believe in Immaculate Conception of the genome,” says Andrew
Pohorille, a theoretical biophysicist at NASA’s Ames Research Center in
California, “there must have been something before the genome.”
Pohorille’s money is on peptides. According to this hypothesis, genome-less
life began when short peptides emerged that could catalyse the production of
other peptides. Amino acids, the precursors of peptides, pop up in just about
any experiment designed to mimic the chemical conditions on the prebiotic Earth,
so it is possible to envisage how those first catalytic peptides were formed.
Most of them, however, would have been weak, nonspecific catalysts. The key to
life lies with the few more efficient ones. They would churn out more peptides
than the others and, by so doing, up the chances that some of these new peptides
would be even more efficient enzymes, as well as capable of performing new
functions.
Eventually, this process of increasing efficiency and functionality would
throw up enzymes that could catalyse the production of RNA and DNA molecules,
and the coupling of DNA, RNA and peptides, generating genomes that work just
like those we know today. Before that stage is reached, however, the peptides
would have evolved in the absence of a genome.
Now, Pohorille is leading a team that is about to test the feasibility of the
non-genomic evolution hypothesis. The first step will be to use computer
simulations to work out what conditions might make such a process possible.
Next, Pohorille and Jack Szostak of Massachusetts General Hospital in Boston
will attempt to re-create non-genomic evolution in a test tube using a new
technique that Szostak helped develop (Proceedings of the National Academy
of Sciences, vol 94, p 12 297).
Pohorille and Szostak plan to generate peptides that catalyse the production
of other peptides. If they succeed, says Pohorille, they will have shown that
non-genomic evolution is a “plausible” hypothesis.
Rachel Nowak