by Charles Baum
Biochemical textbooks describe dozens of other examples of elabo-
rate synthetic pathways: photosynthesis to make the sugar glucose, gly-
colysis (splitting glucose) to make the energy-rich molecule ATP
(adenosine triphosphate), metabolism via the citric acid cycle, the pro-
duction of urea, and countless other vital chemical processes. Over and
over, we find that cells zero in on a few key molecules. DNA and RNA,
which carry the genetic code, rely on ribose and deoxyribose alone,
eschewing the dozens of other 5-carbon sugars. Proteins are con-
structed from only 20 of the hundreds of known amino acids. What’s
more, sugars and amino acids often come in mirror-image pairs—so-
called “right-handed” and “left-handed” variants—but life uses right-
handed sugars and left-handed amino acids almost exclusively.
The take-home lesson is that life is exceedingly choosy about its
chemistry. Of the millions of known organic molecules with up to a
dozen carbon atoms, cells typically employ just a few hundred. This
selectivity is perhaps the single most diagnostic characteristic of living
versus nonliving systems. If an ancient rock is found to hold a diverse
and nondescript suite of organic molecules, then there’s little we can
conclude, yea or nay, about its biological origins. It may once have held
life, or it may simply represent an abiotic accumulation of organic junk.
If, on the other hand, an old rock holds a highly selective suite of car-
bon-based molecules—predominantly even-numbered hydrocarbon
chains or left-handed amino acids, for example—then that’s strong
evidence that life was involved.
A crucial requirement, if this logic is to be implemented in the
search for life here or on other worlds, is that biomolecules must be
stable over time spans of billions of years. Large protein molecules
won’t last that long, and neither will the 20 amino acids that comprise
the building blocks of proteins. Nor will most carbohydrates or hydro-
carbon chains. Over time, water attacks the bonds of these
biomolecules, breaking them into smaller fragments of no diagnostic
use. But polycyclic compounds, like sterols, degrade more slowly and
might survive over geological spans of time. Therein lies a possible
top-down path to the discovery of life that is distant in space or time.
IDIOSYNCRASIES
65
THE HOPANE STORY
Once in a very great while, extremely old rocks are found to hold mi-
croscopic droplets of a petroleum-like black residue—hydrocarbons
that represent the remains of ancient marine algae. When such drop-
lets were first discovered, decades ago, most scientists discounted the
possibility that these organic remains were very old; no oil could sur-
vive billions of years of geological processing, they said. But subse-
quent discoveries and improved analytical techniques have convinced
the geological community that a hardy breed of organic hydrocarbons
can survive in ancient rock provided that temperatures never got too
high.
In their quest for life signs, a group of Australian scientists has
focused upon what are perhaps the ideal biomarkers—distinctive ste-
rol-derived polycyclic hydrocarbon molecules called hopanes. This
group of elegant 5-ring molecules is known in nature only from the
biochemical processes of cellular life, where it concentrates in protec-
tive cell membranes. Furthermore, different variants of hopanes point
to specific groups of microbes with distinctive biochemical lifestyles. If
an ancient rock happens to encase and preserve hopane-related mo-
lecular fragments with the diagnostic structures of once-living
biomolecules, then we have convincing evidence of ancient life.
In 1999, a team of scientists led by Roger Summons (then at the
Australian Geological Survey Organisation) presented compelling evi-
dence for the survival of hopanes in a sequence of 2.7-billion-year-old
This 5-ring structure is characteristic of hopane, a distinctive biomolecule whose backbone may be preserved for billions of years in ancient sediments.
66
GENESIS
sedimentary rocks called the Pilbara Craton, in Western Australia. The
black, carbon-rich shale layers in question came from a section of drill
core extracted from a depth of about 700 meters. The mineralogy of
the shale revealed that it had never experienced a temperature higher
than about 300°C—an unusually benign history for such an ancient
deposit.
Hopanes have been common biomolecules for a long time, so the
Australian team’s principal challenge was ruling out contamination
from more recent life. The rocks might have been contaminated hun-
dreds of millions of years ago by subsurface microbes, or by ground-
water carrying biomolecules from the surface, or perhaps even by oil,
seeping from some other sedimentary horizon. Summons and his col-
leagues discounted the last of these possibilities because they found no
trace of petroleum in adjacent sediment layers. Modern contamina-
tion from living cells, which abound in the lubricants that scientists
use to drill their deep holes in the host rock, was also a concern. The
team ruled out such contamination, too, because the suite of molecules
preserved in the shale was “mature,” containing none of the fragile or-
ganic species that would point to recent lubricants and accompanying
microbial activity.
Summons and his co-workers had to develop meticulous proce-
dures to expose and clean unadulterated fresh rock surfaces: Break the
rock, wash the surface, and measure the wash for contamination. They
resorted to smaller and smaller rock fragments to avoid the inevitable
impurities that had seeped in along cracks. Summons found that prop-
erly prepared powdered shale contained a distinct suite of ancient hy-
drocarbon molecules at hundreds of times higher concentrations than
in adjacent chert and basalt layers from the same drill core. Neverthe-
less, the amount of hopanes was minuscule: of all the carbon-rich ma-
terial extracted from the rock, no more than a precious few hundred
parts per million were hopanes and related polycyclic molecules. Still,
the very presence of hopanes provided evidence for ancient microbial
life.
Having overcome daunting hurdles, Summons and his colleagues
announced their finding in August 1999, in two remarkable papers,
one in Science and the other in Nature. The Science article detailed extraction of hopanes from 2.7-billion-year-old shale—results that broke
the previous record for the oldest molecular biomarker by about a bil-
lion years. The Nature article described the discovery of hopanes from
IDIOSYNCRASIES
67
2.5-billion-year-old Australian black shale—a younger nearby forma-
tion, but with a twist. That formation included 2-methylhopanoid, a
hopane variant known to occur in cyanobacteria, which are the primi-
tive photosynthetic microbes responsible for generating Earth’s oxy-
gen-rich atmosphere. The Australian team had found suggestive
; evidence that cyanobacteria were thriving long before 2 billion years
ago, when Earth’s atmosphere is thought to have achieved modern lev-
els of oxygen.
By extracting and identifying unambiguous biomarkers in ancient
rocks, Summons and colleagues had made a major advance in detect-
ing and characterizing ancient life. They also helped close the gap in
our ignorance of life’s emergence by embellishing the top-down story
and pushing it just a little bit further back in time.
BIOSIGNATURES AND ABIOSIGNATURES
The quest for unambiguous “biosignatures,” including hopanes and
other distinctive molecules, represents an effective strategy in the
search for ancient life on Earth and other worlds. However, the identi-
fication of “abiosignatures”—chemical evidence that life was never
present in a particular environment—might also prove important in
constraining models of life’s emergence.
Abiosignatures hold special significance to astrobiologists, who
search for life in Martian meteorites and other exotic specimens. Are
there physical or chemical tests that might preclude the presence of
past life in those specimens? “NO LIFE ON MARS!” would be a bum-
mer of a headline, but would nevertheless carry great scientific, not to
mention philosophical, implications about the frequency of life’s
emergence.
Hopanes not only represent biosignatures for ancient life on Earth,
but they also point to a search strategy for other worlds, especially our
nearest neighbor, Mars. Based on what we now know about life and its
fossil preservation, we are unlikely to find unambiguous Martian fos-
sils of single cells, much less animals or plants, at least not any time
soon. A Mars sample return mission won’t happen for at least a dozen
years, while human exploration of the red planet is many decades away.
And even with such hands-on exploration, we’d be incredibly lucky to
find a convincing fossil. We’re much more likely to find local concen-
trations of carbon-based molecules, from which we can determine the
68
GENESIS
carbon-isotopic composition. However, as the Greenland incident re-
veals, a simple isotopic ratio may not be sufficient to distinguish
nonbiological chemical systems from those that were once living.
Suites of carbon-based molecules, if we can find them, hold much
greater promise. An array of molecular fragments derived from a
colony of cells, if not too degraded, will differ fundamentally from a
geochemical suite synthesized in the absence of life. For now, molecules
represent our best hope of finding proof of life both here and else-
where in our solar system.
The ideal molecular biosignatures—and abiosignatures as well—
must display three key characteristics. First, biosignatures should con-
sist of distinctive molecules or their diagnostic fragments that are
essential to cellular processes. Similarly, abiosignatures should consist
of molecules that clearly point to nonbiological processes.
The second criterion is stability: biosignatures—and abiosigna-
tures as well—must be molecules able to survive through geological
time. Even the least altered ancient sediments have been subjected to
billions of years of temperatures greater than the boiling point of wa-
ter—conditions that significantly alter the chemical characteristics of
any suite of organic molecules, whether biological or not. This crite-
rion of stability, consequently, focuses our attention on unusually
stable molecules.
Finally, the molecules must occur commonly and in reasonable
abundance. A molecular biosignature or abiosignature is of no use un-
less it can be detected by mass spectrometry or other standard analyti-
cal techniques.
Hopanes, and the related sterols, are unquestionably excellent di-
agnostic biosignatures from the standpoint of stability, and they’re rea-
sonably easy to analyze. Many ancient deposits yield traces of these
molecules, and they will continue to be a tempting target for analysis,
as well as a model for finding other biosignatures. But hopanes are
probably not the ultimate answer in the search for signs of life: They
seldom occur in abundance, and their absence cannot be taken as a
reliable abiosignature.
An alternative to the search for reliable biosignatures and
abiosignatures might be to identify diagnostic ratios of molecular frag-
ments, akin to the carbon-12/carbon-13 isotopic ratio. However, we’re
confronted with a vast multitude of possible molecule pairs. Which
pair of molecules should we study?
IDIOSYNCRASIES
69
My first foray into the search for biomarkers occurred in the sum-
mer of 2004. Preliminary studies by George Cody on organic com-
pounds in meteorites prompted us to look at the ratio of two of the
commonest PAHs: anthracene and phenanthrene. These 3-ring poly-
cyclic molecules, both made up of 14 carbon atoms and 10 hydrogen
atoms (C H ), differ only in the arrangement of the rings: In an-
14
10
thracene the rings form a line, in phenanthrene a dogleg. We realized
that the ratio of these two molecules might fulfill the essential
biomarker requirements: Both are distinctive, relatively stable, com-
mon in the geological record, and easy to detect in trace amounts.
Phenanthrene and anthracene form in abundance through a vari-
ety of nonbiological processes, including any burning process that pro-
duces soot. These cyclic compounds are also synthesized in deep space,
where they contribute to the molecular inventory of the carbon-rich
meteorites called carbonaceous chondrites. The most celebrated of
these is the Murchison meteorite, which fell to Earth in a cow field
outside the small town of Murchison, about 100 miles north of
Melbourne, Australia, on September 28, 1969. Meteorites hit Earth all
the time, but the Murchison fall was special. For one thing, it was big—
several kilograms of rock. For another, it was fresh and relatively un-
Phenathrene (top) and anthracene are 3-ring polycyclic molecules that differ only
in their shape. The ratio of these two molecules differs in abiotic and biological systems.
70
GENESIS
contaminated—a number of pieces were collected while they were still
warm. But, most important, the Murchison was a carbonaceous chon-
drite, containing more than 3 percent by weight of organic molecules.
That black, resinous matter, formed billions of years ago in dense mo-
lecular clouds and protoplanetary disks, held a treasure trove of the
molecules that could have accumulated on the prebiotic Earth.
George Cody had found that such meteorites often display about
a 1:1 ratio of phenanthrene to anthracene. But biochemical processes
seem to produce a different ratio. Many polycyclic biomolecules—in-
cluding sterols and the varied hopanes—incorporate a 3-ring dogleg,
so phenanthrene is a common and expected biomolecular fragment,
and it should persist in rocks and soils, even
when larger molecules
break down. But for some reason, life almost never uses anthracene’s
linear arrangement of three rings. Anthracene would thus seem to be
correspondingly rare as a biomolecular fragment. Cody had found that
biogenic coals typically hold 10 times more phenanthrene than
anthracene.
Is the ratio of phenanthrene to anthracene a useful biomarker?
Testing this idea required measuring the ratios of cyclic compounds in
lots of samples, so that was the task I gave to Rachel Dunham, a bright
and energetic undergraduate summer intern from Amherst College.
Over the course of her 10-week stay in Washington, Rachel assembled
dozens of natural and synthetic PAH-containing samples from around
the world, analyzed them with our gas chromatograph/mass spectrom-
eter, and managed to track down many more analyses from the vast
coal and petroleum literature, since it turns out that PAHs are espe-
cially abundant in some fossil fuels.
The first few data points seemed to support the hypothesis. The
Murchison and Allan Hills meteorites showed phenanthrene-to-an-
thracene ratios of 1.7:1 and 2:1, respectively. The biogenic Burgess
Shale and a mature coal, on the other hand, yielded much higher ra-
tios, close to 15:1. But then results began to scatter. Some low-grade
coals had ratios less than 5:1, while the black, fossil-rich Enspel Shale
was only about 2:1. The promising hypothesis began to crumble.
After several weeks of effort, and a thorough review of the pub-
lished literature, Rachel discovered that we were simply reinventing
the wheel. Coal experts have long known that anthracene is slightly
less stable than phenanthrene. Consequently, “high-grade” coals that
have experienced prolonged high temperatures and pressures have a
IDIOSYNCRASIES
71
higher ratio of phenanthrene to anthracene, topping 20:1 in some
specimens. Such a high ratio in any sample may have resulted from
prolonged heating and have nothing at all to do with a biogenic past.
Our only useful conclusion was that unambiguously biological
specimens never seem to display ratios less than about 2. So a lower
ratio of phenanthrene to anthracene, as found in meteorites, synthetic-
run products, and soot from burning carbon, may provide a valid
abiomarker.
The next step? Perhaps we’ll try another PAH ratio, such as that of
