杏吧原创

GLASS skyscrapers, Gothic cathedrals, yurts, Georgian terraces, Shinto
shrines, wattle and daub, Victorian railway stations, Bauhaus, igloos,
mock-Tudor. Imagine that all the architectural styles that human ingenuity could
ever devise appeared during one 35-year period, sometime in the middle of the
15th century. Imagine how today鈥檚 historians would be trampling over each other
in their eagerness to learn what made that window of profound creativity
possible. That鈥檚 roughly how palaeontologists feel about the Cambrian
explosion.

In just 35 million years, the blinking of an eye for evolution, animal life
erupted in an explosion of inventiveness that far outshines anything the planet
has seen before or since. During the preceding three billion years of life on
Earth, the most sophisticated creatures evolution could come up with were the
organismal equivalent of mud huts鈥攁lgae, the mysterious Ediacarans
(see 鈥淭he Garden of Ediacara鈥�), and perhaps flatworms. Then, with little
warning, at least little that left any trace in the fossil record, all hell
broke loose. Early in the Cambrian period about 535 million years ago, large,
diverse and lavishly elaborate animals suddenly appeared.

They had heads, middles, rear ends, segments and guts. Some had four legs,
some a dozen. Some a shell, some antennae, some gills. In short, almost all the
body forms familiar in modern animals, and several more that have long since
become extinct, sprung up in this one great orgy of creativity. Then it all
stopped, as abruptly as it had started. And in over 500 million years since the
end of the Cambrian, evolution has rested on its laurels, content to spin off
variations on old themes.

The palaeontologists struggling to explain the Cambrian explosion face a
tough task鈥�500 million years separate them from their subject. Until
recently their only option was to study the animal fossil record. But now, more
and more researchers are taking a different path to enlightenment: scrutinising
the genetic record that has been handed down through the ages to today鈥檚
creatures. Comparing the genes of living animals has enabled biologists to crawl
back down the evolutionary tree to deduce which genes were present back then and
what roles they might have played.

This exploration has led some researchers to make the heretical claim that
the Cambrian explosion never happened, and others to say it certainly did, and
that they have found the likely mechanism: the genes that help the cells of a
developing embryo know front from back, top from bottom, and near from far. They
believe that these genes were a necessary prerequisite for the
explosion鈥攖hough they may not have been what set it off.

The most famous of these genes belong to a set known as the hox
cluster. Hox genes are the mapmakers that tell the embryo鈥檚 cells where
they are on the body鈥檚 front-to-back axis and thus what they should become. In
fruit flies, in which they were first discovered, eight hox genes line
up on their chromosome like a train of boxcars. The gene at the front of the
train tells cells there to make a head and other paraphernalia characteristic of
that part of the body, number two takes over a bit further back, and so on back
to the guard鈥檚 van (Americans call it the caboose), which holds sway over the
hindmost end of the animal. By mucking up this orderly sequence, researchers
create startling freaks such as flies sprouting legs from their heads. Other
sets of genes lay out the body鈥檚 up-down axis or distinguish the base of a leg
or a wing from its tip.

The discovery of these genes solved one of developmental biology鈥檚 central
puzzles. How does the embryo of a multicellular animal know what structures to
build where? And about two years ago, three palaeontologists began to suspect
that these genes could be the answer to another great enigma, the Cambrian
explosion. The modular way in which the genes map body regions would have
provided evolution with a mechanism to modify one part of the body without
changing the rest, to duplicate segments, or to add new appendages where none
existed before. This is the line taken by James Valentine, of the University of
California at Berkeley, David Jablonski of the University of Chicago and Douglas
Erwin of the National Museum of Natural History in Washington DC. Just as spoken
language (also modular) is flexible enough to support the whole of human
culture, says Valentine, this genetic language is so basic, so powerful, and so
adaptable that it could underlie the whole amazing diversity of animal life.

Sounds great, but making bold claims is the easy bit鈥攖o convince your
colleagues, you usually need some evidence. Valentine, Jablonski and Erwin
needed to show that these mapmaking genes actually existed in the Cambrian. That
posed a problem鈥擩urassic Park notwithstanding, genes don鈥檛
fossilise, least of all for half a billion years and more.

Fortunately, living organisms hold much of the secret. 鈥淭he present is the
key to the past,鈥� says Rudolf Raff, an evolutionary developmental biologist at
Indiana University in Bloomington. Raff鈥檚 reasoning is simple. Living organisms
are the tips of branches in the evolutionary tree. If the same gene occurs in
two of these animals, that gene must also have been present in their common
ancestor, back where their lineages first split apart. By comparing ever more
distantly related organisms, biologists can move down toward the earliest
branches near the tree鈥檚 root.

Fruit flies and frogs鈥攖wo veterans of the biology labs鈥攈ave very
different body designs. And they occupy twigs on two of the most fundamental
branches of the animal tree, the groups known to biologists as protostomes and
deuterostomes respectively (named after the way in which the embryonic mouth
forms). Palaeontologists know from the fossil record that these two lineages
diverged no later than the early Cambrian, 535 million years ago. They could
scarcely be more distant cousins.

Yet over the past few years, developmental biologists have accumulated more
and more evidence of astounding similarities between the DNA sequences in the
mapmaking genes of these two groups. Flies and frogs, and by inference their
common ancestor in the Cambrian or before, share six hox genes. Another
gene called decapentaplegic, which marks the upper half of the fly
embryo, is kin to Bmp-4, which marks the lower half of the frog embryo.
Somewhere in the course of evolution, the developmental processes triggered by
the genes did a half somersault in one of the two lineages. Apart from that, the
two have diverged so little in more than half a billion years that scientists
can snip the gene out of a fly, plug it into a frog, and it will work perfectly,
triggering the development of the bottom half of a frog wherever you insert it
in the embryo. You can do the same thing with a gene calledpax6, which
controls eye development. Even genes that dictate the development of such
modern-seeming accoutrements as hearts, nervous systems,and body segments appear
to have been present in the frog-fruit fly common ancestor, way back in the
Cambrian or earlier.

This was just the evidence the three palaeontologists needed鈥攁nd its
implications are still sinking in. For a start, it questions the old assumption
that the common ancestor of horses and horseflies, lobsters and
Londoners鈥攖he giga-great-grandmother of the early Cambrian or
before鈥攚as a 鈥渞oundish flatworm鈥�, little more than an oozing blob of
cells. After all, if the genetic evidence is to be believed, that ancestor could
equally well have been a sophisticated, segmented worm with eyes, a heart, a
nervous system, possibly even antennae or legs, says Eddy De Robertis, an
embryologist at the University of California at Los Angeles.

Sensory bump

Another explanation, and the one that fits in best with Valentine & Co鈥檚
hypothesis, is that those key developmental genes existed in that roundish
flatworm, but didn鈥檛 trigger the growth of eyes, limbs, hearts, and segments as
they do in post-Cambrian animals. 鈥淭he problem is that we see what these genes
do now and we assume they were doing the same thing way back then,鈥� explains
Nipam Patel, a developmental biologist at the University of Chicago. 鈥淏ut you
have to be very careful.鈥� The Precambrian 鈥渆ye鈥� that pax6 helped form
may have been nothing more than a crude photosensor with a pigment cell to back
it up, for instance. Indeed, the fossil record shows no trace of anything as
sophisticated as the insect and vertebrate eyes in the earliest protostomes and
deuterostomes. Likewise, the 鈥渁ppendage鈥� for which the ancestor carried
mapmaking genes may have been just a little sensory bump.

There is other evidence that Jablonksi, Valentine and Erwin鈥檚 hypothesis
could be correct. In a paper published in August in the Biological
Bulletin(vol 193, p 62), John Finnerty and Mark Martindale of the
University of Chicago report that sea anemones have a rich set of hox
genes, even though these most primitive of animals have no head or
tail鈥攖hey face the world equally well in all directions. No one knows why
sea anemones should have such genes, but Finnerty suspects they may map out the
anemones鈥� far simpler up-and-down axis instead. The best guess is that the
Precambrian flatworm then commandeered these crude mapmaking genes for use in
the crucial front-to-back axis of its more complex body plan鈥攑ossibly
their first big role as the 鈥渓anguage of evolution鈥�.

鈥淐reationists have often said that the one thing we can鈥檛 explain is the
extremely rapid appearance of body plans in the Cambrian,鈥� says Valentine. 鈥淎nd
this is the answer. We haven鈥檛 understood it until these development guys.鈥�

But not all the new discoveries in evolutionary molecular biology fit as well
with the whole idea of a Cambrian boom. There have always been rumblings from
certain quarters that perhaps the Cambrian explosion is nothing more than an
illusion鈥攁 lovely, imaginative, but completely wrong reconstruction of
evolutionary history. The main evidence for the explosion has been the total
absence鈥攄espite a lot of diligent searching鈥攐f fossils showing
modern-style body plans until the start or just before the start of the
Cambrian. 鈥淚f you go to the same kind of rocks and the animals are not there,
and you look all over the world for tracks and trails and they鈥檙e not there, the
most parsimonious explanation is that complex large [animals] just were not
there. It takes a lot of special pleading to argue they were there and just were
not preserved. It would require a truly perverse fossil record,鈥� says
Jablonski.

In fact, that鈥檚 just what the anti-Cambrians believe鈥攖hat somehow,
through not knowing where to look, or simply through sheer bad luck, fossil
hunters have missed the bodies, burrowings and footprints of complex animals鈥�
Precambrian history. And now they have some molecular ammunition to bolster
their claim. In a study published one year ago in Science (vol 274, p
568), Gregory Wray and his colleagues at the State University of New York at
Stony Brook measured the changes in the sequences of seven genes in dozens of
living species, and concluded that the protostome-deuterostome split happened
roughly a billion years ago, long before the alleged Cambrian explosion. More
recent branches on the animal tree, such as the split between echinoderms
(starfish and sea urchins) and chordates (the group to which vertebrates belong)
also happened well before the beginning of the Cambrian. 鈥淚f this guy is right,鈥�
says Finnerty. 鈥渢here鈥檚 no explosion.鈥�

But the 鈥渕olecular clock鈥� technique Wray and his colleagues used is
controversial. First, they assumed that the longer ago two lineages split, the
more different their gene sequences should be. Then, they counted how many
mutations had accumulated since several branchings that can be clearly dated
from fossil records. For example, bony fishes split from sharks 415 million
years ago and crocodiles from birds 235 million years ago. They used this to
estimate the mutation rate, and so to calculate the age of unknown splits, such
as protostomes from deuterostomes

The big fizzle

The problem is that mutations may occur in spurts, throwing off estimates of
branch dates, says Erwin. Moreover, all the known branch dates Wray used to
calibrate his clock are relatively young, so he had to work backwards to date
the branches of interest. Like the wavering tip of a long pole gripped at one
end, such extrapolations are erratic. To Erwin and other sceptics, this
uncertainty means that the 鈥渂illion-year-old鈥� branches could turn out to be just
600 million years old or less, compatible with the notion of a Cambrian
explosion.

The proponents of molecular clocks admit that they are not very precise, but
deny that they are way off mark. What鈥檚 more, other analyses that attempt to
take into account variations in mutation rates also put the key lineage splits
much earlier than the Cambrian. Taken together, that evidence convinces at least
some that the Cambrian explosion never happened. Instead, evolution burned slow
and steady for at least 100 million years to create all the different animal
forms. 鈥淭he Cambrian is a dull period in evolutionary history. It鈥檚 not where
the interesting things really happened,鈥� says Jeffrey Levinton, one of Wray鈥檚
co-authors.

So what should we believe: the hard copy embedded in rocks around the world,
or the high-tech virtual reality of the molecular clocks? Perhaps both, says
Simon Conway Morris, a palaeobiologist at Cambridge University. Even if the
molecular data tell us that the lineages鈥攖he echinoderms and the
chordates, the protostomes and the deuterostomes鈥攕plit early on, that
doesn鈥檛 mean they acquired their spiffy new body plans prior to the Cambrian.
鈥淵ou can have your animals ticking away for as long as you like, but if they鈥檙e
not doing anything morphologically, who cares?鈥� he asks.

Which brings us back to the Cambrian. If the lineages had gone their separate
ways long before, and if we know mapmaking genes were present before the split,
what happened 535 million years ago to trigger the explosion in body plans?
Before the evidence for the very early existence of the mapmaking genes,
Valentine and his colleagues had toyed with the idea that the genes themselves
might have been the trigger rather than part of the explosive mix. That theory
has been laid to rest, but as always there are still plenty of others to choose
from. Perhaps the first hard-bodied predator chanced to evolve, sparking off an
evolutionary arms race in which everybody had to get larger, better armoured, or
cleverer to stay alive? Or the oxygen content of the atmosphere increased,
suddenly making larger, more energetic animals feasible? Or a burst of speedy
continental drift scattered Earth鈥檚 nearshore habitats, forcing animals to come
up with different solutions to new adaptive challenges? 鈥淲e have several
plausible and a lot of nutty suggestions,鈥� says Valentine. Until some fresh
evidence pops up鈥攎ore genes, or perhaps, some major fossil find鈥攖he
Cambrian trigger is destined to remain mere speculation.

Carefree worms

But whatever happened to kick-start the Precambrian worm into an evolutionary
frenzy, it probably found it much easier to spin off new body plans than a
creature would today. 鈥淵ou can do a lot to that animal in terms of its basic
body arrangement for two reasons,鈥� says Raff. 鈥淥ne is that the world is empty
ecologically, so there can be a lot of experimentation. Whatever you鈥檙e making
doesn鈥檛 have to be very good, because there isn鈥檛 much competition. The second
thing is that because the body鈥檚 relatively simple, changes that would now be
horrendous鈥攍ike changing dorsal and ventral鈥攚ould have been nothing
at all.

鈥淚n the post-Cambrian world, competition is severe, so if you鈥檙e not good at
making a living, you鈥檙e dead meat. And body plans have become more elaborate, so
changes that you make have more consequences. You鈥檝e connected the genetic
machinery together in a certain way, and it may be hard to unconnect it. That鈥檚
left us with a world in which evolution is largely within body plans.鈥�

In short, animal life on Earth has grown up. The heady, experimental days of
its carefree youth are past and now, saddled with a job, a mortgage,
responsibilities, it has settled down into a steady, plodding respectability.
Looking back now, we can shed a tear for those days so long ago when life was
young and we were worms.

* * *

The Garden of Ediacara

WHEN something goes kaboom in a good action film, there have usually been
some mysterious characters hanging about beforehand. So, too, for the Cambrian
explosion, that orgy of evolutionary activity that created the menagerie of
animal body shapes around today. In this case, the skulkers are the Ediacarans,
named after the Ediacara Hills in South Australia, where palaeontologists
stumbled on the first big deposit of their fossils in the 1940s. These creatures
dominated the Earth for tens of millions of years leading right up to the
beginning of the Cambrian, according to some major fossil finds in Namibia over
the past two years. That means that the Ediacarans were at the very least eye
witnesses to the Cambrian explosion. Or they might have been at the explosion鈥檚
ground zero鈥攖he simple forebears of the elaborate animals that it
created.

Palaeontologists who are trying to reconstruct the course of events that led
up to evolution鈥檚 big bang disagree wildly over where these strange creatures
fit in. 鈥淭he problem is that the same fossils are interpreted in completely
different ways by different people,鈥� says Simon Conway Morris of Cambridge
University.

To some people the long, graceful plumes, strange ribbed discs and amorphous,
quilt-like creatures up to a metre long look like nothing else on Earth. One,
Ernietta, looks like a round-bottomed mug made out of drinking straws. These
people argue that some Ediacarans鈥攑erhaps all鈥攄eserve their own
kingdom, albeit an extinct one, on a par with today鈥檚 plants, animals and
fungi.

But other Ediacaran fossils show affinities to segmented worms and relatives
of corals and jellyfish, says James Gehling of the University of South Australia
in Adelaide. And one, Kimberella, has many hallmarks of a primitive mollusc,
Russian and American researchers announced in August in Nature (vol 388, p 868).
Evidence like that suggests that at least some Ediacarans are the forerunners of
modern animals.

Absolutely not, says Mark McMenamin of Mount Holyoke College in South Hadley,
Massachusetts. 鈥淚鈥檝e figured out the Ediacaran problem. I鈥檓 convinced now that
these things are not animals.鈥� His new analysis shows that during their
embryonic development, Ediacarans never pass through a stage consisting of a
hollow ball of cells known as a blastula, as all animals do. He plans to present
his evidence at next week鈥檚 meeting of the Geological Society of America in Salt
Lake City, Utah.

McMenamin鈥檚 claim is sure to prove contentious, but even those who don鈥檛 buy
it think that he might be right on a second point: his assessment of Ediacarans鈥�
fate in the Cambrian. McMenamin imagines the late Precambrian as a 鈥淕arden of
Ediacara鈥�, an idyllic time where large, soft-bodied creatures lolled about on
the seabed in a peaceful coexistence, drawing nourishment from photosynthetic
algae or microorganisms living within their tissues, much as many corals do
today.

Their idyll came to a sticky end when the animal forebear (be it an Ediacaran
or some other organism) evolved into the first large predator. For a creature
with the right equipment鈥攕ense organs to find prey and mouthparts to bite
it with鈥攖his peaceable kingdom was a well-stocked supermarket full of
tasty treats.

The evolution of such predators around the beginning of the Cambrian would
spark an evolutionary arms race in which the organisms that could get faster,
fiercer, or more furtive survived to give rise to the animals we see in the
Cambrian, says Bruce Runnegar of the University of California at Los Angeles.
Those that couldn鈥檛 adapt鈥攊ncluding the biggest, blobbiest
Ediacarans鈥攎ade a quick exit.

This theory, appealing as it is, remains untested so far. But
palaeontologists are already scouring rocks at the Cambrian boundary for signs
of the sudden appearance of predation. 鈥淚鈥檓 sure we鈥檙e going to have many
discoveries in the next 12 months,鈥� says Gehling.