by Pross, Addy
In the light of the above experiments and arguments, the reader has hopefully been convinced that the processes of abiogenesis and evolution are actually one single physicochemical process governed by one single mechanism, rather than two discrete processes governed by two different mechanisms. That insight will turn out to be of utmost value as it leads to a whole range of both chemical and biological insights. If our conclusion is correct it means we can apply chemical insights from the chemical phase to better understand the biological phase, and we can also apply biological insights derived from 150 years of studying Darwinian evolution to provide greater insights into the chemical phase. Win-win for sure! But beyond that, the unification tells us that chemistry and biology are one, that there is a complexity continuum that connects them, that biology is just an elaborate extension of replicative chemistry. Interestingly, as noted in the prologue, Darwin, in his genius, foresaw the existence of some underlying principle governing abiogenesis and biological evolution. However, thanks to the inspiring work of gifted systems chemists these past decades, we don’t have to speculate about the nature of a general life principle—the life principle can now be formulated based on hard facts.
So what new insights does this merging of chemistry and biology provide us with? Before answering this question and in order to fully benefit from this conceptual merging, we now need to rephrase Fig. 6. Traditionally one would describe the first phase, the chemical one, in chemical terms, and the second biological phase in biological terms—each process in its own language. But, as we all know from foreign travel, a dialogue in two languages, when the two parties do not speak the other’s tongue, may be frustratingly less than useful. Misunderstandings galore can arise. In order to avail ourselves of the deeper insight of one continuous process, the two phases need to be described in one language. So which is it to be—the language of chemistry, or that of biology? The answer is clear-cut: the entire process—chemical and biological—needs to be described in chemical terms. Let me explain why.
In an earlier chapter (chapter 3), I described how understanding in science is achieved at different hierarchical levels. Phenomena at a higher hierarchical level of complexity are normally explained in terms of scientific principles associated with a lower hierarchical level. Thus we conventionally explain biological phenomena in chemical terms and chemical phenomena in physical terms, not the other way around. Recall, Steven Weinberg’s comment: ‘Explanatory arrows always point downward.’24 To bring this point home and to illustrate how fundamental this hierarchical aspect of explanation is, consider the two sciences, chemistry and psychology, and how they might interrelate. Let us say you find some psychological phenomenon of interest and you tried to explain it in molecular terms. Scientifically speaking that is quite acceptable. For example, if you came up with a molecular explanation for schizophrenia that would certainly be of interest—drug companies would likely be knocking on your door! However, if one went the other way and attempted to explain some molecular phenomenon in psychological terms, that would only attract derision! Schizophrenic molecules? Neurotic molecules? No way! The message is clear: the temptation to interpret phenomena that are inherently chemical in nature in biological terms—fitness, natural selection, adaptation, survival of the fittest, cooperation, information, etc., should be firmly resisted. Open any chemical text that deals with chemical reactivity and those biological expressions will not be found there. Chemical phenomena are explained in chemical (and physical) terms, as chemistry is the more fundamental science. On this basis a reinterpretation of Fig. 6 in terms of just one scientific discipline makes clear that the discipline of choice must be the lower-level one, chemistry, not the higher-level one, biology. So let us proceed to do just that. Let us reinterpret the entire process of Fig. 6—part chemical, part biological—solely in chemical terms.
Natural selection is kinetic selection
When several replicating molecules are mixed with their component molecular building blocks, as described in chapter 4, they compete with one another, in much the same way as biological entities compete for a limited supply of food. But as explained above we shouldn’t discuss that competitive process as natural selection at the molecular level. Such reactions are dealt with by a specific branch of chemistry that deals with the rates of chemical reactions called chemical kinetics. That sub-discipline of chemistry, going back some 100 years to the pioneering work of Alfred Lotka, has no difficulty in dealing with the situation in which two replicating molecules compete for the same building blocks. It comes up with a clear-cut prediction that is applicable in most cases—the faster replicating molecule will out-replicate the slower replicating molecule and drive it to extinction. That result comes out directly by solving the relevant rate equations. In other words when two replicating molecules compete for the same chemical building blocks, the outcome is readily explained by a process that chemists call kinetic selection. Kinetic selection in everyday language just means ‘the faster one wins’. Since the faster replicator is capable of assembling building blocks into new replicating molecules more effectively (for a variety of chemical reasons), the number of those faster replicators grows quickly while the number of slower replicators drops until those slower replicators disappear entirely.
But that strictly chemical result does ring a biological bell. It sounds very much like the way in which natural selection operates in biology. When two biological species compete for the same resource, the one that can utilize that resource more effectively drives the other to extinction. That result is the basis for the competitive exclusion principle that we discussed earlier. But then, natural selection and kinetic selection are really the same concept, so let us state that explicitly:
natural selection = kinetic selection
Biological natural selection merely emulates chemical kinetic selection. Natural selection is the biological term, kinetic selection is the chemical term.
At this point the reader may ask why the chemical description is to be preferred over the biological one. Despite the earlier comment that explanatory arrows always point downward, aren’t the chemical and biological explanations really saying the same thing, that faster, and therefore more effective replicators, whether chemical or biological, will out-replicate less effective ones? Not quite. The reason is that the chemical explanation is more fundamental and probes the issue of selection more deeply. The chemical term is more quantifiable than the biological one because chemical systems are inherently simpler. That greater simplicity allows us to further break down the composite chemical replication reaction into the individual reaction steps that go to make it up. The chemical analysis can tell you how long it will take for one molecular replicator to out-replicate the other. It will even tell you under what circumstances the two replicators may coexist. Coexistence between competing molecular replicators can also be observed under appropriate circumstances.
Biological systems, on the other hand, are many orders of magnitude more complex, and are therefore less amenable to a detailed chemical analysis. That is why the two subjects are typically discussed at their different hierarchical levels. No matter, the recognition that natural selection has its roots within a fundamentally chemical phenomenon, one that is well understood, provides an important link connecting the two sciences of chemistry and biology.
Fitness and its chemical roots
What about that central biological term ‘fitness’? What is the chemical analogue of that term and what new insights does the translation of that central biological term offer? According to Darwin, fitness is just the capacity to survive and reproduce, and its optimization is deemed the ultimate goal of the evolutionary process. Yet that concept, conceived by Darwin in strictly qualitative terms, has become a source of endless confusion due to continuing attempts to formally quantify it. The large number of fitness types that have been proposed and discussed—absolute fitness, relative fitness, inclusive fitness, ecological fitness, to mention some key ones—clearly attest to th
e inherent difficulties in this venture. The problem of fitness is a highly complex one, and one that has been troubling leading biologists for the better part of the past half-century, so a detailed discussion is well beyond the scope of this book. In the present context our goal is a more limited one: to explore how the merging of chemistry and biology can assist in clarifying at least some aspects of the troublesome ‘fitness’ issue.
In our earlier discussion of replicating systems we identified a fundamental characteristic of those systems—their dynamic kinetic stability, DKS. The ability of a replicating system to maintain itself over time reflects its stability, but a stability kind that differs from the conventional thermodynamic one. Our discussion now reveals that ‘fitness’ is actually the biological expression of that more general and fundamental chemical concept, so let us state that explicitly:
fitness = dynamic kinetic stability (DKS)
When we classify a biological entity as ‘fit’ we are really specifying that it is stable—stable in the sense of being persistent. However, as we explained previously in some detail, that stability kind only applies to a population, not to individual replicators within the population. Specifying that a population is fit (or stable) just means that the population is able to maintain itself through ongoing replication/reproduction. The immediate consequence of relating fitness and DKS is that it indicates more explicitly that fitness is best viewed as a population characteristic, not an individual one. The concept of DKS has no real meaning at the individual level. A stable population of some replicating system is the reality that comes about through individual replicators being formed and then decaying, like the water droplets turning over in a fountain. In the context of life, if you focus on the individual entity, tempting as it may be, you are missing the essence of what defines life—its dynamic nature, the continual turnover of the individual entities that make up a particular replicating population. Bottom line: in order to understand life’s essence one should focus on life’s population aspect, not its individual aspect. Life is an evolutionary phenomenon and evolution does not operate on individuals, only on populations. Individuals are just born and then die. Focus on the individual and you will miss much of what life entails. In actual fact the difficulty in individual thinking goes deeper than the above comments might suggest. What is an individual living entity, and do they actually exist? The answer to this question is more complex than we might imagine, but I will defer this aspect of the discussion to chapter 8.
The fact that a population perspective is crucial for a proper understanding of replicator dynamics received considerable impetus from important theoretical work carried out in the 1970s by Manfred Eigen, the eminent Nobel prize-winning German chemist, together with Peter Schuster, the distinguished Austrian chemist, on what is termed quasispecies theory.58 In order to understand that theory in simplest terms we first need to describe what is meant by a ‘fitness landscape’. As already discussed in chapter 4, when a replicating molecule, say an RNA of some particular sequence, proceeds to replicate, occasional errors in the replication reaction will result in the formation of RNA mutants. Mutants that are faster replicators will tend to drive the slower replicating sequences to extinction. That process of sequence modification can be represented by what is termed a fitness landscape—a three-dimensional topographical map. In that three-dimensional representation, the horizontal axes represent sequence changes (that come about through mutations) while the vertical axis represents the fitness of the particular sequence. The higher the value on that vertical axis, the greater the fitness. Accordingly, the fitness landscape resembles a three-dimensional topological map of mountain ranges and valleys in between. High points on the landscape—mountain peaks—represent RNA sequences of high fitness (fast replicators) and low points—valleys—represent RNA sequences of lower fitness (slower replicators). What that means is that some initial RNA sequence of a particular fitness, a point on that topology map, will tend to explore the fitness landscape in search of the highest point on the fitness landscape, representing the sequence of highest fitness, much like a hiker in the mountains seeking to climb to the top of the highest peak.
But here’s the important point. Manfred Eigen and Peter Schuster discovered that the population of replicating RNAs that is generated by this exploration of the fitness landscape does not consist of one single sequence, but rather a population of RNAs of differing sequences, centred around the most successful sequence (termed the wild type) within that population. This population of varied sequences was termed a quasispecies, and an analogy that may help capture the essence of a quasispecies would be a flock of birds as it moves in concert over the sequence landscape in search of ever higher peaks. Eigen and Schuster discovered through their computer modelling of evolutionary changes in the RNA sequences that it is not the fittest sequence that is selected for but the fittest population of sequences—the fittest quasispecies—that is selected for. In other words, evolution operates by seeking out improved fitness in a population sense rather than in an individual sense. In fact one can see in Eigen and Schuster’s seminal work the importance of population heterogeneity. A mutation leading to a particularly successful replicator is as likely to come from a slower RNA as from a faster one. Counter-intuitively, the road to a fitter population may actually pass through a ‘less fit’ individual replicator within the existing population. Population heterogeneity opens up more possibilities for evolution to carry out its magic—heterogeneous populations evolve more effectively than homogeneous ones. The message is clear: the essence of stability in the world of replicators is rooted in populations, not individuals. Evolution is a process that populations undergo, not individuals. In the evolutionary scheme of things the individual is but a fleeting event, a transient water droplet in the fountain of life.
We have discussed the concept of DKS in some detail and the pertinent question now arises: can DKS be quantified? The short answer—only to a limited extent. We have identified DKS as a distinct stability kind in nature and stated that evolution operates so that DKS tends to increase over time. That fact alone suggests the term is quantifiable. Surely, if we say that evolution leads to greater DKS, that means that DKS is measurable. Yes and no. Take two RNA molecules in Sol Spiegelman’s experiment competing for building blocks during self-replication and we see one replicates faster than the other. So the relative rates at which the two RNAs replicate may be taken as a quantitative measure of the relative DKSs of the two RNA populations. Recall, it is precisely because of this rate difference that a population of one replicator drives the other population to extinction. However that attempt at quantification is only of limited value for two reasons. First, attempts at quantification are only relevant for two populations that feed off a common resource. That means it can be applied to two RNA populations competing for the same set of activated nucleotides. But asking whether an E. Coli bacterium or a camel is more stable is meaningless, even in a population sense—there is no common frame of reference; it would be like comparing apples and oranges.
But the difficulty in quantifying DKS goes deeper. Despite the above comments implying that relative DKS can be estimated for competing replicators, another problem arises. The relative rates of replication depend on the reaction conditions. Let’s return to Sol Spiegelman’s classic RNA test-tube experiments. If some extraneous material is introduced into the test tube that inhibits the replication rates, as in fact was done by Spiegelman, then the winner of the replication race can switch. Change the reaction conditions and the evolutionary course changes as well; the winner of the Darwinian race is likely to be an entirely different set of RNA molecules. When Sol Spiegelman added a substance, ethidium bromide, to the reaction mixture, the winner of the Darwinian race turned out to be a different sequence.59 Why is that? Since the extraneous material that had been added affected the mechanism of replication, certain sequences that initially facilitated rapid replication were inhibited, while other RNA sequences, initially slower, wer
e favoured. In other words DKS for populations of RNAs is circumstantial. Its magnitude depends on the particular materials that are present in the reaction. But that means that DKS is quite different from that other kind of stability we speak about in chemistry, thermodynamic stability. The thermodynamic stability of water is a defined quantity regardless of what else is present (though to be precise, it does depend to some extent on the physical conditions, temperature, pressure, etc.). Thermodynamic stability is an intrinsic property of any system and is measured in closed systems. Dynamic kinetic stability depends on rates of reaction, is highly sensitive to reaction conditions, and can only be assessed in open systems, in which energy and resources are continually supplied. That makes comparisons of DKS highly problematic.
To clarify the point further, let us consider a biological example—a population of bacteria in a pool of water. Such a population may well be highly stable—billions of bacteria are busy replicating, resulting in the establishment of a dynamic population of those bacteria. But add chlorine to the pool and the bacteria simply die—their stability vanishes. The DKS of that bacterial population has dropped to zero. Same bacteria, different circumstances. The thermodynamic stability of the water molecules in that pool, however, does not depend on the environment. It is measured relative to that of a hydrogen-oxygen gas mixture (the materials from which water is formed), and that difference does not depend on the location of the water and the presence of other materials in the water (at least to any significant extent). Attempting to quantify DKS for a particular system is like preparing for an exam where the answers to the questions keep changing!
