New Scientist has some intriguing recent stories on particular fronts of research, including a discussion of the search for so-called ‘shadow life‘ — meaning, essentially, life forms here on earth that don’t share the last common ancestor of currently known life forms. Exobiologists, who hope to study life beyond earth, tend to focus on nearby candidate habitats (such as Mars and some volcanic moons elsewhere) that are thought to be energetically and chemically amenable to life as we know it. But it’s intriguing to consider the possibility that critters with ancestry outside the currently understood ancestral tree of life are living under (or perhaps even in…) our noses. In mulling the potential importance of understanding such life, and how it might differ from our own, the New Scientist article quotes Carol Cleland with an instructive analogy:
“If you were an alien biologist who’s interested in understanding what a mammal was, and all you had was zebras, it’s very unlikely that you would focus on their mammary glands, because only half the specimens have them. You’d probably focus on the stripes, which are ubiquitous.”
This analogy, though pithy and thought-provoking, strikes me as incomplete. If alien biologists thought like human biologists, for example, they’d likely first focus on traits that they shared with zebras (such a bias toward expecting and recognizing the familiar is reflected, comically, in the vast array of subtle twists on human morphology that seem to characterize most sentient life forms — cosmos-wide — in the fictional world of Star Trek). But Cleland’s point — that life, perhaps like any phenomenon, is most effectively understood through systematic comparison of many examples — is well taken.
So how might a ‘shadow’ life form differ from us other life forms? Well, for starters, it might not share the micron-scaled compartmental (cell- and other membrane sac-defined) architecture that we know so well. Or, for that matter, the familiar centralized (‘genomic’) storage of replicative ‘instructions’. Or, more fundamentally, the carbon ‘backbone’-based chemistry that we recognize to be so versatile for building molecules that interact with each other in richly complex ways. Yet even if a newfound life form shared all of the foregoing traits with ‘non-shadow’ life, a key way in which it might differ from us might be in not sharing one of the most salient ‘stripes’ that characterize life as we know it — our genetic code.
The latter term is so frequently (and dismayingly) misused in popular science journalism, that a definition is in order here. By genetic code, biologists literally mean a key: that is, a specific translation scheme for making proteins (the molecules that catalyze or otherwise centrally direct most of the chemical interactions that we recognize as crucial to cellular function) from instructions written in the form of nucleic acids (which make up genomes). The ‘texts’ to be translated into proteins, via this code, are found in the specialized segments of a genome that we call genes. Such texts are simply strings of 3-letter DNA ‘words’ called codons, each of which is translated into a particular amino acid ‘meaning’ as the protein (a polymer of amino acids strung together in the same order as the codons that specify them) is built.
Because the ‘alphabet’ of DNA has exactly four ‘letters’ (A,C,G,T), there are exactly sixty-four (i.e., 4 x 4 x 4) distinct codons. But there are, with minor exceptions, just twenty distinct amino acids used to make proteins in life as we know it. Other than these twenty amino acids, the only other ‘meaning’ to be conveyed by codons is the ‘stop sign’ used to signal that a newly minted protein is complete. The genetic code, having more distinct words (64) than distinct meanings (21), is thus said to be degenerate — that is, some of its words have the same meaning.
While such degeneracy can help buffer a code against ‘transmission error’, degenerate codes are inherently inefficient. One could, for example, easily devise a genetic code using just sixteen ‘2-letter’ codons (one of them with a ‘shift key’ meaning), rather than the sixty-four ‘3-letter’ ones, to encode exactly the same ‘meaning’ information as the real genetic code does, in roughly 30% less ‘text’. And when one considers other optimality criteria, too, the standard genetic code likewise falls short of ideal: Stephen Freeland and Laurence Hurst, for example, measured codes’ abilities to buffer the effects of genomic mutation — a major source of ‘transmission error’ — at the protein level, and estimated that the standard code, while remarkably good, was worse than trillions of possible alternative codes (and they even limited the field of contenders to codes that had exactly the same synonymous codon ‘families’ found in the standard code, with the same such family assigned to the ‘stop sign’ meaning).
Yet when one surveys earthly life — from blue whales to bacteria (and including humans, of course) — one finds that nearly all known organisms use this standard code. Known exceptions (including, incidentally, the code used within mitochondria) are very rare, likewise suboptimal, and involve reassignment of no more than a handful of codons relative to the standard code. The striking near-ubiquity of the standard code, with its readily apparent functional shortcomings, likely won’t sit well with proponents of ‘intelligent design’, and continues to prompt debate among evolutionary geneticists themselves.
The leading explanation, championed by Francis Crick (who, in addition to discovering the structure of DNA, gleaned the first details of the standard genetic code for us), posits that the code is simply a ‘frozen accident’. In this view, the standard code, like the QWERTY keyboard, has been more lucky than good: out of oodles of possible variants, it worked well enough (potentially quite a bit better than early rivals), early in the evolution of a system of information flow, to become overwhelmingly common and well entrenched. As the system (earthly life, or earthly typing) evolved, important infrastructure (in the QWERTY analogy, not just keyboards themselves, but the programming details of word processing software, the synapses of typing instructors, etc.) became well adapted to the variant in question but poorly suited to any potential rival, even if the latter were much more efficient.
The ‘weight’ of such infrastructure now serves, in this view, to anchor the standard genetic code against intrinsic improvement. In evolutionary terms, consider the implications of even a tiny tweak of the genetic code in an organism that, like a human or a banana plant, relies on the code in making thousands of distinct proteins, each with finely tailored functionality. Swapping one amino acid for another, in thousands of proteins at once, would be overwhelmingly likely to wreak utter havoc with organismal function. Simply put, tinker with the code and the organism will, with near perfect certainty, fail to develop.
‘Shadow life’, however, could readily have a genetic code that differs from the one we know so well. If such differences were slight, we might infer that the newfound critter in question was a fairly close relative, despite its novelty in our systema naturae. Such a discovery might push back our understanding of the origin of life on earth, setting it in a temporally deeper and/or spatially broader context. If the differences between the two codes (assuming the ‘shadow’ life form even had a recognizable genetic code) were vast, on the other hand, we might infer that the two domains of life had distinct evolutionary origins.
All of which raises the question — crucial to consider in any quest to recognize ‘shadow life’ — of how we are to define life in the first place. Historian Samuel Moyn, a keen thinker and a longtime friend, recently told me that, having informally polled some biologists on how they define life, he’d been surprised by the lack of consensus — or even individual clarity — in their answers. He wondered how a whole field could focus, with any lasting fruitfulness, on something that its researchers could not rigorously define. In noting the definitional slipperiness of life, Moyn is on to something; I’m not sanguine that any rigorous candidate definition of living matter will fully accord our ‘intuitive’ answers to the question of whether something is alive. Textbook cases of ambiguity — such as ‘are viruses alive?’ — often hinge on some concept of ‘autonomous replication’; but, of course, nothing that we think of as unambiguously living (such as ourselves) can truly replicate ‘autonomously’, i.e., independently of necessary cofactors that we think of as external to the organism.
That said, I’m not sure that Moyn’s point applies only to biologists; mathematicians, for example, may ultimately have a hard time rigorously defining, say, number (at least in terms that withstand criticism from within their own ranks). While admitting that we biologists sweep a huge question under the rug in failing to rigorously define life, I’m nonetheless excited to be living and working in an era during which I anticipate that our understanding of the scope of life (however defined) in the cosmos may expand significantly. Here’s hoping we find something surprising, and not too hungry.