In 1952, Stanley Miller and Harold Urey derived a number of racemic amino acids from a handful of small molecules. These were electrifying results because they suggested that the methods of synthetic chemistry might finally explain the origins of life. The excitement was justified, but premature.1 Origins of life (OOL) research has, to be sure, become progressively more sophisticated, but its goal—to explain the origins of life—remains as distant today as it was in 1952. This is not surprising. The protocols in use have remained unchanged: buy highly purified chemicals; mix them together in high concentrations and in a specific order under carefully devised laboratory conditions; derive a mixture of compounds; and publish a paper making bold claims about OOL. These protocols are as unrealistic as they are unimproved.
This essay comprises an argument, but it also contains an appeal to the OOL community. The history of science suggests that on occasion what is required for research to flourish is not further research—at least to the extent that further research involves doing the same thing. This is one of those times.
Needed for Life
Four molecules are needed for life: nucleotides, carbohydrates, proteins, and lipids. Nucleotides are composed of a trimeric nucleobase-carbohydrate-phosphate combination, and once polymerized, constitute DNA and RNA. Five different nucleobases comprise the entire alphabet for DNA and RNA. The nucleotides and their subsequent DNA and RNA structures are homochiral, yielding one of two possible enantiomers. Amino acids are most often homochiral. When amino acids are polymerized, they form proteins and enzymes. Proteins and enzymes also display a tertiary homochirality. Lipids are dipolar molecules with a polar water-soluble head and a nonpolar water-insoluble tail. They, too, are most often homochiral. Cells use carbohydrates for energy, and carbohydrates, along with proteins, are identification receptors. Carbohydrates are also homochiral, and their polymeric forms take on tertiary homochiral shapes. OOL researchers have spent a great deal of time trying to make these four classes of molecules, but with scant success.
Constructing the molecules necessary for life from their prebiotic precursors represents one goal of OOL research; putting them together, another. Some of synthetic chemistry is pedestrian, and some ingenious. Fundamental questions remain unaddressed. Claims that these structures could be prepared under prebiotic conditions in high enantiomeric purity using inorganic templates, or any presumed templates, have never been realized. The carbohydrates, amino acids, lipids, and other compounds within each of these classes require specific methods in order to control their regiochemistry and stereochemistry. The differences in reaction rates often require chiral systems acting upon chiral molecules. If this were possible under prebiotic conditions, it is odd that it cannot be replicated by synthetic chemists.
They have, after all, had 67 years to try.
Synthetic Hyperbole
Consider the class of experiments that deals with the assembly of chemicals into what are referred to as protocells—“a self-organized, endogenously ordered, spherical collection of lipids proposed as a stepping-stone to the origin of life.”2 In 2017, a team from the Origins of Life Initiative at Harvard University performed a type of polymerization reaction in water known as the reversible addition–fragmentation chain transfer.3 This reaction type is not seen in nature, and neither are the monomers that figure in the experiment. Still, this is standard chemistry. Polymers are made by a controlled radical polymerization reaction, where two different monomer types are added sequentially to a chain bearing both a hydrophobic and a hydrophilic block. Researchers observed polymeric vesicles forming during polymerization—interesting, but not extraordinary. The vesicles grew to bursting as researchers kept the radical chain growing through ultraviolet light activation. There is, in this, nothing surprising: the forces between the growing vesicle and the surrounding water dictate a critical growth volume before the vesicle ruptures.
The claims should have ended there.
Here is how the work was portrayed in the published article:
The observed net oscillatory vesicle population grows in a manner that reminds one of some elementary modes of sustainable (while there is available “food”!) population growth seen among living systems. The data supports an interpretation in terms of a micron scale self-assembled molecular system capable of embodying and mimicking some aspects of “simple” extant life, including self-assembly from a homogenous but active chemical medium, membrane formation, metabolism, a primitive form of self-replication, and hints of elementary system selection due to a spontaneous light triggered Marangoni instability [provoked by surface tension gradients].4
These claims were then rephrased and presented to the public by the Harvard Gazette:
A Harvard researcher seeking a model for the earliest cells has created a system that self-assembles from a chemical soup into cell-like structures that grow, move in response to light, replicate, and exhibit signs of rudimentary evolutionary selection [emphasis added].5
This degree of hyperbole is excessive.6 Nothing in this experiment had growing cell-like structures with replication, or that exhibited aspects of evolutionary selection.
Teams from the University of California and the University of New South Wales recently conducted lipid bilayer assembly experiments, publishing a summary of their work in 2017.7 They combined nucleotides and lipids in water to form lamellae, with the nucleotides sandwiched between the layers. Nucleotides are trimers of nucleobase-carbohydrate-phosphate, and, in this case, both nucleotides and lipids were purchased in pure homochiral form. Both teams then demonstrated that a condensation polymerization of the nucleotides can take place within the lamella upon dehydration. Polymerization takes place by means of a reaction between preloaded phosphate and the purchased stereo-defined alcohol moiety found on a neighboring nucleotide. Similar reactions, they conjectured, may have occurred at the edge of hydrothermal fields, volcanic landmasses providing the necessary heat for reactions.
The chemistry that figures in these experiments is unremarkable. Bear in mind that derivatives were all preloaded. To provide the essential concentrations for the reactions, researchers removed the water, thus driving the intermolecular reactions to form oligomers that resembled nucleic acids. The problem with condensation polymerization is obvious: any alcohol can compete for the reactive electrophilic site. In the case under consideration, researchers added no other alcohols. They were scrupulous, but the system was stacked. Condensation polymerization reactions need to be very pure, free from competing nucleophilic and electrophilic components. Witness the Carothers equation, which defines degrees of polymerization based upon monomer purity.8 If there happened to be amino acids or carbohydrates mixed with the nucleotides, they would terminate or interrupt the growth of the oligonucleotides. What is more, the researchers did not confirm the integrity of the structures they claimed to have derived. If carefully analyzed, these structures would likely have shown attacks from unintended hydroxyl sites. Since their sequences are essentially random, short oligonucleotides are not realistic precursors to RNA. An alphabet soup is not a precursor to a poem. The authors go on to suggest that the lamella sandwiching oligonucleotides eventually break off to form lipid bilayer vesicles. These contain the oligonucleotide-within-vesicle constructs, which they call protocells. The conversion of planar lamella into multilamellar vesicles as they hydrate is well established, but shearing forces are generally required to form the requisite lipid bilayer vesicle. For this reason, yields were likely to be low.9 It is hard to imagine finding highly purified homochiral nucleotides trapped in a pure lipid lamella on the prebiotic earth.
But set all that aside. These vesicles bear almost no resemblance to cellular lipid bilayers. Lipid bilayer balls are not cellular lipid bilayers. One would never know this from reading the authors’ account. “Then, in the gel phase,” they write, “protocells pack together in a system called a progenote and exchange sets of polymers, selecting those that enhance survival during many cycles.”10 Chemicals, of course, are indifferent to their survival. No mechanism is described to demonstrate how protocells would bear different sets of polymers or exchange polymers among them. Terms from biology have generally been misappropriated in a way that makes no chemical sense. This is not an isolated or incidental defect. It reappears when the authors write that “[t]he best-adapted protocells spread to other pools or streams, moving by wind and water.”11 Best-adapted? Microbial communities apparently “evolve into a primitive metabolism required by the earliest forms of life.” Molecules do not evolve, and nothing is being metabolized. Condensation polymerization is a simple chemical reaction based upon the addition of nucleophiles to electrophiles with loss of water. Such a reaction is never referred to as a form of metabolism within synthetic chemistry.
Terminology is one thing, non sequiturs quite another. “After much trial and error,” the authors write, “one protocell assembles the complicated molecular machinery that enables it to divide into daughter cells. This paves the way for the first living microbial community.”12 How is the molecular machinery made? They do not say. The mechanisms needed for cellular division are complex, requiring cascades of precisely functioning enzymes. There is nothing between what the authors demonstrate and what they claim to have established, and nothing they propose “paves the way for the first living microbial community.”
The Emerging Cell
A functioning cell contains a complex noncovalent interactive system. Nobody knows how a cell emerges from its molecular components. An interactome is the set of molecular interactions in a given cell.13 Interactions may be between proteins, genes, or molecules. Information is transferred within the cell through these molecular interactions. Electrostatic potentials permit information to flow through noncovalent molecular arrays, but these arrays require specific orientation.14 The interactome defines these intermolecular orientations, alignments that are unattainable through random mixing.
Peter Tompa and George Rose have calculated that if one considered only protein combinations in a single yeast cell, the result would be an estimated 1079,000,000,000 combinations.15 The authors understand that this is a very large number, one that precludes “formation of a functional interactome by trial and error complex formation within any meaningful span of time.” What Tompa and Rose call “a complicated cellular sorting/trafficking and assembly system” is required. Sophisticated scaffolding notwithstanding, “in the absence of energy even this well developed infrastructure would be insufficient to account for the generation of the interactome, which requires a continuous expenditure of energy to maintain steady state.”16 In their concluding paragraph, Tompa and Rose remark that
[t]he inability of the interactome to self-assemble de novo imposes limits on efforts to create artificial cells and organisms, that is, synthetic biology. In particular, the stunning experiment of “creating” a viable bacterial cell by transplanting a synthetic chromosome into a host stripped of its own genetic material has been heralded as the generation of a synthetic cell (although not by the paper’s authors). Such an interpretation is a misnomer, rather like stuffing a foreign engine into a Ford and declaring it to be a novel design. The success of the synthetic biology experiment relies on having a recipient interactome … that has high compatibility with donor genetic material. The ability to synthesize an actual artificial cell using designed components that can self-assemble spontaneously still remains a distant challenge.17
The fact is that interactomes add a massive layer of complexity to all cellular structures. It is one that underscores the difference between a real cell and the protocells or extant cells made by OOL researchers.
In 2010, a team led by Craig Venter made a copy of a known bacterial genome and transplanted it into another cell.18 In 2016, they did something better, removing all but 473 genes from a natural genome and transplanting it into another cell.19 Venter and his team were circumspect; the press was enthusiastic. More recently, Henrike Niederholtmeyer, Cynthia Chaggan, and Neal Devaraj have made what they term, “mimics of eukaryotic cells.”20 Science declared them “the most lifelike artificial cells yet.”21 Microcapsules made of plastic and containing clay were prepared using microfluidic techniques. Clay has a high affinity for binding DNA. Thus, when DNA was added to the solution, it diffused through the semiporous plastic microcapsules and bound to the clay. The requisite RNA polymerases, together with the ribosomes, tRNA, amino acids, enzymatic cofactors, and energy sources were either purchased or extracted from living systems. The expected chemical reactions did result in protein synthesis. Newly formed proteins diffused from their microcapsules of origin to other microcapsules. The nearer the neighboring microcapsule, the greater the exchange of reagents between them. Diffusion between microcapsules the authors dubbed quorum sensing. The chemistry would work no matter the container, whether a test tube or a large-scale industrial production tank. If the experimental design is clever, the synthesis is unremarkable. Phys.org reported these modest results in markedly flamboyant terms, referring to “gene expression and communication rivaling that of living cells.”22 There is no rivalry here. All of the active chemical components were extracted from living systems. If these are “the most lifelike artificial cells yet,” this serves only to underscore the point that no one has ever come close to the real thing.23
Life as a Lucky Fluke
In an article entitled “How Did Life Begin?” Jack Szostak asks whether the appearance of life on earth is “a lucky fluke or an inevitable consequence of the laws of nature.”24 It is a good but premature question, a point obvious from his own appreciation of current research. Having vetted the usual suspects of asteroids, dust clouds, volcanoes, lightning, and time, Szostak appeals to “a concentrated stew of reactive chemicals”:
Life as we know it requires RNA. Some scientists believe that RNA emerged directly from these reactive chemicals, nudged along by dynamic forces in the environment. Nucleotides, the building blocks of RNA, eventually formed, then joined together to make strands of RNA. Some stages in this process are still not well understood. … Once RNA was made, some strands of it became enclosed within tiny vesicles formed by the spontaneous assembly of fatty acids (lipids) into membranes, creating the first protocells. … As the membranes incorporated more fatty acids, they grew and divided; at the same time, internal chemical reactions drove replication of the encapsulated RNA.25
The thesis that “RNA emerged directly [emphasis added] from these reactive chemicals, nudged along by dynamic forces” is painful to a synthetic chemist. A complex pathway of reactions would have been needed, incorporating purification, assembly, polymerization, and sequencing. Nothing emerged directly in Szostak’s scenario, let alone something as complex as RNA. Phrases such as “nudged along by dynamic forces” have no meaning in terms of synthetic chemistry. Nucleotides never form and join together to make strands of RNA without complex protecting and deprotecting steps. It is perfectly true that “[s]ome stages in this process are still not well understood,” if only because we are clueless about the chemistry needed on a prebiotic earth.
In the diagram to which Szostak appeals, the compounds listed as simple sugars are, in fact, glycerol and ethylene glycol. There are known routes to convert them to simple sugars, but only in gross relative and absolute stereochemically mixed states, and as a mixture of several different polyols.26 Carbohydrate synthesis is a difficult prebiotic problem.27 Szostak’s carbohydrates would be useless in their mixed states, and separations are hard. The diagram’s cyanide derivatives are unrecognizable as cyanide derivatives. In an act of grace, let us attribute these chemical structural errors to the faulty renderings of a staff artist. The chemical errors are Szostak’s own. There is simply no way that heat and light can directly make a nucleotide from simple sugars and cyanide derivatives. Such glossy presentations have become the standard of the OOL community when it tries to build upon the careful work of exacting synthetic chemistry.
I have discussed these issues with OOL researchers, and I am amazed that they fail to appreciate the magnitude of the problem in building molecules. They see little difficulty in accepting a chemical synthesis where a desired product is mixed with a large array of closely related yet undesired compounds. They seem unaware that separations would be enormously complex, and subsequent reactions unavailing. In a 2018 article for Progress in Biophysics and Molecular Biology, Edward Steele et al. concede the following.
The transformation of an ensemble of appropriately chosen biological monomers (e.g. amino acids, nucleotides) into a primitive living cell capable of further evolution appears to require overcoming an information hurdle [emphasis added] of superastronomical proportions, an event that could not have happened within the time frame of the Earth except, we believe, as a miracle. All laboratory experiments attempting to simulate such an event have so far led to dismal failure.28
“At this stage of our scientific understanding,” they write, “we need to place on hold the issue of life’s actual biochemical origins [emphasis added]—where, when and how may be too difficult to solve on the current evidence.”29 All is not lost. If life on earth did not arise on earth, “[i]t would thus seem reasonable,” Steele et al. remark, “to go to the biggest available ‘venue’ in relation to space and time. A cosmological origin of life thus appears plausible and overwhelmingly likely.”30 Why chemical reactions that are unlikely on the earth should prove likely somewhere else, Steele et al. do not say.
Facing Facts
John Sutherland, one of OOL’s giants and the most skilled synthetic chemist to engage in OOL research, has recently proposed that “chemical determinism can no longer be relied on as a source of innovation, and further improvements have to be chanced upon instead.”31 Chanced upon? It appears that Sutherland has come to appreciate the depths of the problems facing OOL researchers. In 2017, Ramanarayanan Krishnamurthy et al. showed that diamidophosphate can phosphorylate nucleosides, nucleotides, and stereo-scrambled lipid precursors. These can further result in the formation of random oligonucleotides and oligopeptides. The fundamental challenges with respect to synthesis and assembly remain unaddressed. Krishnamurthy was rightly measured in writing about “the pitfalls of extrapolating extant biochemical pathways backwards all the way to prebiotic chemistry and vice versa.”32 In 2018, Clemens Richert argued that “the ideal experiment does not involve any human intervention.”33 This is a step in the right direction. So, too, is the fact that he scrupled at the pure chemicals used by the OOL community.
It is time for a temporary time out. Why not admit what we cannot yet explain: the mass transfer of starting materials to the molecules needed for life; the origin of life’s code; the combinatorial complexities present in any living system; and the precise nonregular assembly of cellular components?
It would be helpful if leading researchers, among them very sophisticated synthetic chemists, were to step back, pause, and join forces. If the origins of life remain a mystery, two goals are within reach: an agreement about the rational standards by which OOL research should be judged, and a candid acknowledgment of the problems that remain to be overcome. A statement of this sort would be reassuring in its candor.34
