The first emergent step in life’s chemical origin, wherever it may have occurred, must have been the synthesis of organic molecules. Stanley Miller’s experiments amply demonstrated one potential ancient energy source with which to effect this synthesis—lightning near the ocean–atmosphere interface. Subsequent workers, notably Thomas Gold, postulated deeper origins and chemically derived energy sources. Any plausible source of carbon-based molecules should be fair game for origin researchers. Every plausible source deserves rigorous study.

In 1996, when George Cody, Hat Yoder, and I began our research on hydrothermal chemistry and the origins of life, we were novices in the origins game, too naïve to suspect that we were wasting our time. We thought that our experimental studies of organic synthesis in hot, high-pressure water provided compelling evidence that hydrothermal zones could have played a role in the prebiotic production of life’s molecules. Stanley Miller and his students, most notably chemist Jeffrey Bada at the Scripps Institution of Oceanography, would have been happy to set us straight, but we had yet to cross paths with them. Harold Morowitz, whose ideas instigated the research, had warned us that we might meet with some criticism. He was right.

It must have seemed odd to the scientific community that a group of geologists and geophysicists was studying the origin of life. For al-

most a century, our home base, the Carnegie Institution’s Geophysical Laboratory, had supported research on the physics and chemistry of Earth materials at extreme temperatures and pressures. Carnegie scientists had melted granite, synthesized basalt, and heated and squeezed just about every known rock-forming mineral to understand how our dynamic planet works. They routinely achieved temperatures of thousands of degrees and pressures of millions of atmospheres—extreme environments that would seem to have little to do with life’s delicate chemistry. Nevertheless, the lab’s arsenal of furnaces and pressure vessels stood ready to heat and squeeze almost any sort of sample, and experiments on carbon compounds weren’t intrinsically different from any other chemical we might have chosen to study.

Now, more than 2,000 experiments later, our hydrothermal organic-synthesis program is expanded, well funded by NASA, and going strong. Our simple strategy, much in the spirit of Stanley Miller’s original experiments, has been to subject plausible prebiotic chemical mixes of water, carbon dioxide, and minerals to a controlled environment—in our case, a high-pressure and high-temperature environment typical of deep-ocean hydrothermal systems—still often employing the classic gold-tube protocol described in the Prologue.

Our first experiments (the ones with pyruvate and carbon dioxide) were exhilarating. In that series we synthesized sugars, alcohols, and a host of larger molecules, many incorporating dozens of carbon atoms. Under extreme hydrothermal conditions, the pyruvate molecules polymerize; in some cases they form rings and branching structures reminiscent of molecules found in modern cells.

In a later set of experiments, we sealed water, nitrogen gas, and powdered minerals into the gold tubes—a reasonable proxy for primitive conditions at some hydrothermal vents. In these experiments, we produced ammonia, an essential starting material for synthesizing amino acids and other biomolecules. Our work suggests that hydrothermal vents may have been one of the principal sources of the early Earth’s ammonia. Carrying on with this line of experiments, we put ammonia into a capsule with pyruvate and found that these two compounds react to form the amino acid alanine. By 1999, we had estab-

lished one plausible chemical pathway leading from simple carbon compounds, nitrogen gas, and common minerals to key biomolecules.

In 1997, shortly after our initial exciting results, Cody, Yoder, and I submitted our first papers for peer review. We believed optimistically that our demonstrations of facile hydrothermal organic synthesis would be welcomed as a logical extension of the Miller–Urey approach. After all, we, like Miller a half-century before, had subjected a closed vessel containing water and simple carbon-containing molecules to heat. We had followed Miller’s lead in attempting to simulate a plausible prebiotic environment. And we, too, had found an abundance of synthetic organic molecules. We thought the Miller camp might actually welcome our contribution. Boy, were we wrong!

Miller and his scientific cohort had staked their claim to a surface origin of life, and they seemed determined to systematically head off dissenting opinions. Their reasons were straightforward and clearly argued. While organic synthesis may occur near the 100°C limit of liquid water, they argued, at the much higher temperatures of our experiments, the biomolecules of interest, notably amino acids, decompose at a higher rate than they can be manufactured. Hydrothermal zones on the early Earth would have destroyed many essential biomolecules in the prebiotic ocean much faster than they could have been replenished. Such hostile zones were antithetical to life. We weren’t the first researchers to suffer these criticisms: The Oregon State group who first outlined the hypothesis of a hydrothermal origin of life following deep-sea discoveries in Alvin had found themselves repeatedly thwarted by this argument, their manuscript rejected time and again. However, we were not about to follow their example by publishing in obscure conference proceedings.

Stanley Miller had been relentless in his offensive ever since the hydrothermal-origins idea arose. In a 1988 article in Nature, Miller and Jeffrey Bada wrote, “This proposal is based on a number of misunderstandings concerning the organic chemistry involved…. The high temperatures in the vents would not allow synthesis of organic compounds, but would decompose them.” Miller was more blunt in a Discover feature in 1992, when he called the vent hypothesis “a real loser.”

In The Spark of Life, their popular book on origin theories, Bada and Christopher Wills dubbed those of us working on the hydrothermal hypothesis “ventists,” making us sound a little like the members of

some fanatical religious cult. Their presentation is engaging and generally accurate, but like an argumentative rondo the old refrain appears again and again: “Vents are more likely to destroy organic compounds than create them.” “This high temperature rapidly destroys organic compounds rather than synthesizing them.” “At these high temperatures, amino acids undergo total decomposition on a time scale of only a few minutes.” “Vents … do terrible things to the primordial soup.” Piled onto this catchy litany were other criticisms: Our starting solutions were much too concentrated, the resulting products were much too dilute, and our geological assumptions about vent-water circulation were incorrect.

Our immediate reaction to the assault was defensive and angry. We saw the Miller camp as self-serving, defending old ideas while discounting new possibilities. Such turf-protecting tactics are not unknown in science, but we were more than a little dismayed by such focused efforts to thwart what seemed an interesting and testable idea. Our sense of a conspiracy arrayed against us was heightened when anonymous reviewers recommended rejecting one of our grant proposals on the basis of familiar, party-line criticisms.

The most maddening aspect of Stanley Miller’s criticisms wasn’t his persistence or even the public way he voiced them. The most maddening aspect was that he was basically right. Our first experiments were admittedly flawed. Our starting mixtures of pyruvate were much more concentrated than any plausible prebiotic solution; indeed, unstable pyruvate is unlikely as an important prebiotic reactant in any environment. What’s more, we couldn’t control the acidity inside our gold capsule (acidity can drastically alter reaction pathways), we couldn’t be sure that all our products formed at high temperature (as opposed to during cooling), and we hadn’t proved that the gold was an inert container (might the gold have triggered the observed reactions?). In short, it was a typical start to a complex scientific research program.

The art of science isn’t necessarily to avoid mistakes; rather, progress is often made by making mistakes as fast as possible, while avoiding making the same mistake twice. After several hundred experiments under a wide range of pressure and temperature conditions, we began to realize that water, carbon dioxide, and nitrogen are not sufficient by themselves to synthesize very many useful biomolecules. Under the right circumstances we could trigger some interesting organic reactions, but many key biomolecules were missing from the inventory

of our products; and the ones we did manage to make were indiscriminate, with numerous molecules of no possible use to biology. We needed some other chemical ingredient, or ingredients, and so our research led us back to my specialty—mineralogy. Minerals, it turns out, have the potential to promote key chemical reactions, while protecting the products from the ravages of high temperature.

Before biology, the only raw materials our planet had to play with were the ocean, the atmosphere, and rocks. It now appears that rocks and minerals, in roles as varied as containers, templates, catalysts, and chemical reactants, may provide a key to understanding life’s ancient origins.

The view of minerals’ roles in the origin of life came to the fore with the publication of a seminal paper in 1988: “Before enzymes and templates: Theory of surface metabolism,” by the brilliant, combative Günter Wächtershäuser. Wächtershäuser, a chemist by training but a Munich patent attorney by day, erected a sweeping theory of organic evolution in which minerals—mostly iron and nickel sulfides, which abound at deep-sea volcanic vents—provided catalytic, energy-rich surfaces for the synthesis and assembly of life.

Among the many striking aspects of Wächtershäuser’s grand scheme is the unusual extent to which he roots his theory in the philosophy of science. Most researchers tend to ignore science philosophers, whose work often seems removed from the practical day-to-day details of experimentation. But Wächtershäuser is an avowed disciple (and a personal friend) of the influential philosopher of science Karl Popper, who encouraged the new theory’s development. “During breakfast I mentioned my ideas on an alternative and he urged me to work it out,” Wächtershäuser told a New York Times reporter not long ago.

According to Popper’s dictum, theories, to be theories at all in the scientific sense, must make testable predictions that might be proved false by empirical tests. A theory that makes no testable predictions is pseudoscience according to Popper. By the same token, the more precise (and, therefore, the more falsifiable) its predictions, the more believable a theory becomes when it survives scrutiny.

Most origins researchers before Wächtershäuser had adopted an

ad hoc, trial-and-error approach: They cooked up likely geochemical recipes to see what worked, then patched together a theory around those observations. Stanley Miller’s experiment made amino acids, so his followers became convinced that that’s how it happened in nature. “Ventists,” who found primitive life in hydrothermal zones, often fell into the same trap. Theories flourished around particular sets of observations rather than grand falsifiable frameworks. By contrast, Wächtershäuser set out to formulate his theory from scratch, starting with a few plausible assumptions about the nature of the first living entity.

Assumption 1. Reject the “primordial soup” concept. In Wächtershäuser’s model, random prebiotic synthesis plays no essential role in life’s origin. He argues that the prebiotic emergence of biomolecules à la Miller–Urey was irrelevant because the primordial soup was too dilute and contained too many molecular species that could not contribute to life. It takes a lot of nerve to throw out decades of research, but that’s Wächtershäuser’s style.

Assumption 2. The first life-form made its own molecules. Most other workers assume that the first living entity scavenged amino acids, hydrocarbons, and other useful molecules from its surroundings—a strategy called heterotrophy (from the Greek for “other nourishment”). Wächtershäuser denies that possibility, since the soup was so dilute and unreliable. He counters that the first life must have been autotrophic (“self-nourishing”), manufacturing its own molecular building blocks from scratch.

Assumption 3. The first life-form relied on the chemical energy of minerals, not the Sun. Sunlight and lightning are both too violent, and they are uncharacteristic of the energy sources most cells use today. Moreover, photosynthesis is an immensely complex sequence of chemical reactions, requiring numerous proteins and other specialized molecules. Surface reactions on minerals, by contrast, are simple and similar to the synthesis strategies central to many cellular processes today.

Assumption 4. Metabolism came first. Many researchers claim that life requires encapsulation and membranes, but simple membranes are ill suited to let food in or waste out. Others, seduced by discoveries of modern molecular genetics, are convinced that life began with a self-replicating genetic molecule like RNA (see Chapter 16), but even the simplest genetic molecule is vastly more complex than anything in the

Miller–Urey soup. Wächtershäuser, in sharp contrast to these more conventional ideas, assumes that life began with metabolism, which he defines as a simple cycle of chemical reactions that duplicates itself.

These key ideas are amplified by one additional assumption, common to all origin theories—that of biological continuity. Today’s biochemistry, no matter how intricate and dependent on specialized catalysts, has evolved in an unbroken path from primordial geochemistry. Thus, for example, Wächtershäuser’s scheme builds on the observation that many of the molecules that enable today’s living cells to process energy have, at their core, a mineral-like cluster of iron or nickel and sulfur atoms.

But Wächtershäuser’s theory is much more than a list of assumptions. Over 100 pages of detailed chemical reactions—testable steps that lead from simple geochemical raw materials to biology—amplify his epic proposition. “You don’t mind if I brag a little,” he told a reporter for Earth in 1998, “but something like this has never been done in the entire field.”

We’ll look at the details of Günter Wächtershäuser’s theory in Chapter 15, but the central chemical idea is that iron and nickel sulfides, notably the common iron–sulfur minerals pyrrhotite and pyrite, served as template, catalyst, and energy source for biosynthesis. In Wächtershäuser’s view, simple molecules like carbon monoxide and hydrogen react on sulfide surfaces to produce larger molecules. These molecules tend to have negative charges and so they stick to the positively charged sulfide surfaces, where additional reactions build larger and larger molecules. Various surface-bound molecules begin to feed off each other, eventually forming a chemical cycle that copies itself. Voilà! It’s alive! In Wächtershäuser’s view, this emergent process is both inevitable and fast. “It takes maybe two weeks,” he estimated in answer to a question at a Carnegie seminar.

Wächtershäuser’s theory has proven more than a little controversial, and it has evoked much discussion and comment, both pro and con. “There’s probably nothing there because, otherwise, people would have found it already,” Jeffrey Bada says. In a recent paper in Science, he and his colleague Antonio Lazcano dismiss Wächtershäuser’s originality: “It’s not a new idea,” they write, pointing to a 1955 paper by the Lithuanian-American microbiologst Martynas Ycas—a two-page “Note on the origin of life” in the Proceedings of the National Academy of Sciences that had comparatively little impact on the origins commu-

nity. Whatever one thinks of Wächtershäuser’s full-fledged and widely cited theory, it cannot be dismissed as a mere restatement of earlier ideas.

More than most scientists, Wächtershäuser seems to keep track of who cites his work, who offers praise, who follows up with relevant experiments. Though we had never met, he considered our Carnegie group a friend. His provocative theoretical ideas had influenced our experimental program. We had focused on the possibility that in the presence of minerals—especially iron and nickel sulfides—metabolism can proceed without protein catalysts. Many of our experiments served to test his ideas. So in January 1998, as a member of the Lab’s seminar committee, I invited Wächtershäuser to give a lecture. He responded by phone less than a week later, and we agreed on a visit near the end of March.

Intense and earnest, combative and to the point, Wächtershäuser delivered his lecture like a legal summary to a packed, attentive room. An hour proved not nearly enough to touch on the philosophy and outline of his epic hypothesis, but many of us had read his papers and followed as one does a favorite, oft-told tale. When he did touch on other theories or dissenting views, he dismissed them quickly, efficiently, as if he were brushing away a pesky fly.

I should be clear: None of us was a Wächtershäuser disciple, nor did any of us believe everything he proposed; indeed, the idea that prebiotic chemistry and the primordial soup played no role in life’s origins runs counter to the assumptions of many of our hydrothermal synthesis experiments. But his origin-of-life model is amazingly rich and detailed. Virtually every chemical step is testable by experiments. Nothing is fuzzy or left to speculation. His daring lies in the prediction of specific chemicals that arise by specific reaction pathways. Right or wrong, Wächtershäuser has produced a synthesis that will be studied by scientists for decades to come (and historians of science long after that).

Following the seminar, several of us were eager to help test the model in any way we could. That afternoon we showed him our extensive experimental facilities, described our origins research, and raised the possibility of collaboration. Our offer was for him to propose the experiments and be the lead author on any resulting publications. A week later I wrote, “We would like to explore the possibility of collaborative experimental work, in particular the effects of modest pres-

sure on the rates of reactions you propose.” He received our offers politely, but was clearly reluctant and declined to accept our help. I was left with the strong impression that Günter Wächtershäuser is eager to maintain control over both his theory and its experimental verification.

The hypothesis that minerals might enhance the stability of some biomolecules received a boost from a series of experiments in 1999–2000 conducted by my Carnegie colleague Jay Brandes. Brandes had arrived at the Geophysical Lab fresh from his 1996 University of Washington PhD in oceanography—a study of nitrogen chemistry in the Central Amazonian Basin of Brazil. Jay’s original proposal for work at the Lab was to elaborate on the nitrogen work, but he soon shifted his efforts to our origin-of-life program. It proved a good decision.

His first paper at Carnegie, “Abiotic nitrogen reduction on the early Earth,” appeared in the September 24, 1998, issue of Nature and immediately gained widespread attention. Brandes demonstrated that hydrothermal vents could have provided a reliable source of ammonia—an essential nitrogen-bearing compound for prebiotic synthesis. He followed up by demonstrating the synthesis of amino acids under ammonia-rich vent conditions.

In spite of these successes, studies by Stanley Miller’s group on the rapid breakdown of amino acids had raised a question that would not go away. Miller and his colleagues objected to the hydrothermal-origin hypothesis, in part, because amino acids decompose rapidly at the elevated temperatures of ocean vents. How could we be making appreciable quantities of amino acids in our runs at 200°C, if these chemicals aren’t stable? So Jay’s next round of experiments focused on the breakdown of amino acids under a variety of geochemically relevant conditions.

Brandes’s new experiments found evidence that the amino acid leucine, which breaks down in a few minutes in 200°C pressurized water, may persist for days when pyrrhotite, an iron–sulfur mineral commonly found at submarine volcanic vents, is added to the mix. While the exact protection mechanism is still under study, Brandes’s experiments seemed to demonstrate that minerals may greatly enhance the stability of essential biomolecules.

Two other intriguing lines of research point to the complexity of determining amino acid stability. One recent set of experiments focuses on fossil bones, which often preserve ancient proteins. Fossil bones are composite materials in which a strong but brittle mineral matrix interweaves with flexible fibers of the proteins osteocalcin and collagen. Unprotected, these proteins would break down in a matter of a few centuries, but some fossil bones are known to preserve proteins for many millions of years. Recent work in Andrew Steele’s laboratory has even revealed hints of fossil collagen in dinosaur bones more than 70 million years old. The secret to such exceptional preservation is strong bonding between minerals and the amino acid constituents of the proteins. Minerals can thus protect and preserve amino acids.

Additional evidence for amino acid survivability comes from experiments conducted by Stanford graduate student Kono Lemke and geochemist David Ross of the U.S. Geological Survey laboratories in Menlo Park. Working with financial support from our NASA Astrobiology Institute grant, they placed amino acids in a superhot pressure cooker and watched for them to decompose. But, unlike the protocol of our restrictive gold-tube experiments, Lemke and Ross employed an exotic flexible “gold bag” apparatus—a great improvement on the sealed-tube experiments.

The heart of the gold-bag apparatus is a thin-walled flexible bag about the size of a grapefruit, meticulously crafted from pure gold foil. The bag opens at one end into a titanium-valve system, with which Lemke and Ross filled and emptied their experimental solutions. They immersed the entire gold assembly in a water-filled pressure chamber that was compressed to several hundred atmospheres and heated to several hundred degrees—conditions similar to those found at deep-ocean hydrothermal zones. Once they loaded and sealed the apparatus, hot, pressurized water compressed and heated their sample container uniformly on all sides.

The principle of the gold-bag apparatus is much the same as that of our smaller gold-tube experiments. Both rely on soft, inert, deformable gold to exert uniform pressure and temperature on a fluid sample. The great advantage of the gold-bag setup is that small samples of the reacting fluid can be extracted every few hours or days throughout the duration of a long experiment. But trade-offs are a fact of scientific life. The disadvantage is that the thin-walled gold bag is frustratingly fragile and can be a pain to use. A slight miscalculation and the bag will

rip, ruining an experiment and requiring a tedious and exacting welded repair. But once properly filled with reactants, the device usually works wonderfully well.

Lemke and Ross gingerly placed the bag into a water-filled pressure vessel, slowly filled the gold container with a solution of the amino acid glycine and water, sealed the assembly, and ramped up the pressure and temperature. They ran their samples for weeks, monitoring the solution, watching for the glycine to decompose. Over time, the concentration of glycine steadily declined, but even though they didn’t use minerals in their experiments, they observed a much slower rate of breakdown than had been reported in previous studies at lower pressure.

In addition, they found a surprisingly fast rate of peptide-bond formation—amino acids linking together to form molecular chains. This result was unexpected, because hot water tends to break apart rather than form peptide bonds. (That’s one reason boiled foods are so squishy—the sturdy bonds between amino acids that give food texture break down.) Once formed, these peptides decomposed rather quickly, but their formation pointed to more complex behavior than had been expected.

Lemke and Ross found hints of another potentially important behavior in their gold bag. Single amino acid molecules and small clusters with just two or three molecules linked by peptide bonds readily dissolve in water at room temperature, but longer peptide chains proved much less soluble. Lemke and Ross imagine a scenario in which peptides form rapidly in vents and are then exposed to the cooler seawater. Given a high enough concentration of long amino acid chains, these molecules might separate out as a relatively stable concentrated phase—just the kind of emergent molecular selection and organization that life’s origin required. There’s a lot more work to be done, but it appears that the book is not yet closed on amino acid stability in hydrothermal systems.

The most fundamental biological reaction—and one of our group’s primary goals in prebiotic-synthesis experiments—is carbon fixation, the incorporation of more carbon atoms (starting with carbon dioxide) into organic molecules. After all, the first chemical step in the path

to life must be to make bigger molecules, like amino acids and sugars, out of smaller ones, like carbon dioxide, ammonia, and water. Such reactions occur rapidly in our experiments, but they follow two rather different paths, depending on the mineral employed.

Many common minerals, including most oxides and sulfides of iron, copper, and zinc, promote carbon addition by a routine, industrially important process known as the Fischer–Tropsch (F–T) synthesis. In its idealized form, F–T synthesis builds long chainlike organic molecules from carbon dioxide and hydrogen that are exposed to hot, dry metal surfaces. Our gold-tube experiments and studies in several other labs display similar reactions in the presence of wet mineral surfaces at high pressure, though a lot less efficiently than the industrial process.

Field studies complement these experiments. Recent intriguing analyses of organic molecules emanating from hydrothermal vents reveal similar Fischer–Tropsch-like products, and it now appears that F-T synthesis constantly manufactures larger organic molecules from smaller building blocks in Earth’s hydrothermal zones. Many of these molecules are hydrocarbons of the type that form petroleum. (Who knows, maybe Tommy Gold is correct and at least some petroleum forms abiotically at depth.)

Alternatively, when we use nickel or cobalt sulfides, we observe that carbon addition occurs primarily by the insertion of carbon monoxide, a molecule with one carbon atom and one oxygen atom, which readily attaches itself to nickel or cobalt atoms. By repeating these simple kinds of reactions—add a carbon atom here, an oxygen or hydrogen atom there—over and over again, new and more complex organic molecules emerge.

One conclusion seems certain: Mineral-rich hydrothermal systems contributed to the early Earth’s varied inventory of potential bio-building blocks. With tens of thousands of miles of deep-ocean hydrothermal ridges, billions of cubic kilometers of warm wet crust, and hundreds of millions of years to process the raw materials, organic molecules must have been produced in prodigious amounts. But the Geophysical Lab synthesis experiments have done more than simply add to the catalog of interesting molecules that could have been formed on the early Earth. These experiments are now uncovering something quite new about the possible role of minerals in the origin of life.

Previous origin-of-life studies, such as those of Günter Wächtershäuser, treat minerals essentially as solid and relatively stable

platforms for synthesis and assembly of organic molecules. But our experiments reveal another, more complex behavior that may have important consequences for origin-of-life chemistry. We find that in the presence of high-temperature and high-pressure water, minerals often start to dissolve. In the process, the dissolved atoms and molecules from the minerals themselves become crucial reactants in the prebiotic milieu. Sulfur, dissolved from sulfide minerals, combines with carbon dioxide and water to form thiols and thioesters—reactive molecules that can jump-start new synthetic pathways.

Even more dramatic is the behavior of iron, which can dissolve in water to form brilliantly colorful organic solutions. After one experimental run, George Cody bounced from office to office on the second floor, showing off a particularly striking orange-red solution he had just extracted from a pressure capsule. The deep color was exciting because it pointed to the formation of iron complexes—iron atoms surrounded by a starburst of organic molecules. Chemists have long known that similar iron complexes promote chemical reactions, so Cody speculated that our cheerful solutions might contain a kind of primitive catalyst that promoted the assembly of more complex molecular structures.

Such behavior is not entirely unexpected, for hydrothermal fluids are well known to dissolve and concentrate mineral matter. Many of the world’s richest ore deposits arise from hydrothermal processes. Similarly, spectacular sulfide pillars tens of meters tall grow rapidly at volcanic vents called black smokers, where rising plumes of hot, mineral-rich solutions contact the frigid water of the deep ocean.

Yet there’s so much we don’t know about hydrothermal systems and the chemical processes that might occur in their vicinity. And in spite of their prevalence, the role of these dissolved ingredients has not yet figured significantly in origin scenarios. No one yet knows how this rich mix of organic compounds and dissolved minerals might influence the synthesis and assembly of biomolecules. But we’re poised to find out.

This page intentionally left blank.