Introduction

The concept of function, both natural and unnatural, is central to our understanding of the workings of the world around us. Natural function, meaning function that has come about by natural means, is ubiquitous, as it permeates all of biology. We recognize that the heart’s function is to pump blood, that of the kidneys to remove waste products, the ribosome to synthesizes proteins, the mitochondria to provide a source of energy, and so on. The living cell, the basic unit of life, is characterized and understood, first and foremost, as a functional entity. In fact, Ernst Mayr, the noted evolutionary biologist, considers that biology covers two largely separate areas—functional biology and evolutionary biology (Mayr 1988). The functional biologist continually asks the “how” question. How does something function? How does it work? Indeed, thanks to the extraordinary advances in molecular biology over the past half century, much of the recent effort in functional biology has been dedicated to unravel the intimate details of cellular function at the molecular level. The central role that function plays in biology is beyond dispute.

Unnatural function also abounds. Our world is filled with non-biological functional entities—chairs, cars, computers, screwdrivers, toothbrushes, and so on, all created by humans for functional reasons. And then there are birds’ nests, beaver dams, termite mounds, biofilms, and so on, created by biological organisms other than humans, also for their function. But, of course, these non-biological functional entities, all created by biological organisms, are only functional within a biological context, so one can really think of them as biological appendages. Ultimately all unnatural function is associated with, and derived from biological natural function.

But the very existence of function raises an immediate question, one that has attracted surprisingly scant attention: how did function come about? Or, more fundamentally, how could function have come about? In the inanimate world there is no function (putting aside, of course, those unnatural functional entities associated with biological organisms that were mentioned above). Within the inanimate world all material entities obey the laws of physics and chemistry, and, in particular, the Second Law of Thermodynamics. Within the inanimate world there is no purpose, and hence no function. If, however, as we now believe, life emerged from inanimate matter by some physicochemical process, the question necessarily arises: how could function have emerged from an objective universe in which function was non-existent? Did function somehow evolve gradually over evolutionary time, or, did some abrupt transition take place, what one might call a phase transition? The question is an important one since, if we want to understand the connection between chemistry and biology, we must understand how function was at all able to emerge from a chemical world originally devoid of function. Offering answers to these questions could also shine new light on a related issue that has troubled biologists and philosophers of biology since Darwin—how to reconcile Darwinian materialism with the undeniable teleological character of living systems (see, for example: Nagel 2012; Ghiselin 1994; Lennox 1993; Mayr 1988) (for a related question, the emergence of teleonomy, see Pross 2008).

A second longstanding question in biology is the nature and origin of biological complexity. The topic is highly contentious since, though we intuitively understand what is intended by the term, the term itself is notoriously difficult to define. Biological complexity is multifaceted, reflecting structural (McShea 1996), functional (McShea 2000), informational (Gell-Mann and Lloyd 1996; Adami and Cerf 2000), even ecological (Arthur 1994) aspects, and, as a consequence, its practical quantification remains a source of unending debate (Edlund et al. 2011; Adami 2002; Gell-Mann 1995). Nonetheless, despite those imponderables, there seems to be broad agreement that complexity did increase over evolutionary time, unambiguously so during abiogenesis, when simple molecular systems complexified into (so-called) simplest life, but also during biological evolution, when the process was characterized by several major transitions (Schuster 1996; Maynard Smith and Szathmáry 1995). But the reasons for that process of complexification remain far from clear. Was it to increase fitness, to enhance function, or possibly to accommodate increasing information (Edlund et al. 2011; Adami 2002; Adami et al. 2000)? Or did it arise through random drift (McShea and Brandon 2010; Gould 1996)? In any case, with regard complexification during abiogenesis, biological explanations based on the concepts of fitness, function, information, and the like, seem less applicable, even inappropriate. After all, such biological terms are not a regular part of the chemical lexicon. As we have discussed in previous publications (Pross 2011, 2012), the scientific hierarchy within the natural sciences requires us to explain biological phenomena in chemical terms, not the other way around.

We have previously argued that if we want to understand the essence of any biological phenomenon, we need to trace that phenomenon right back to its initiation, to the physicochemical process of abiogenesis, which is presumed to have led to the emergence of life on earth (Pross 2011, 2012). What is needed is a fundamental understanding of the process by which life on earth emerged. Carl Woese, the iconic 20th century biologist phrased it most pointedly: “the essence of biology lies not in things as they are, but in things coming into existence. Biology is a study, not in being, but in becoming” (Woese and Goldenfeld 2009). While some might take exception to that extreme position, it is increasingly clear that if we wish to uncover the fundamental nature of biological phenomena, such as function and complexity, we will need to discover how these biological phenomena could have emerged naturally from chemical systems. We need to understand the physicochemical process by which chemistry became biology.

Till recently the process of abiogenesis remained highly speculative and shrouded in uncertainty (Pross and Pascal 2013; Pross 2012; Lazcano 2008; Forterre and Gribaldo 2007; Shapiro 2006; Luisi 2006; Cleland and Chyba 2002; Kauffman 2000; Fry 2000; Eigen 1992) so that uncovering the chemical roots of any biological phenomenon by seeking its origin in abiogenesis, seemed a distant goal. In the last few years, however, thanks to significant advances in systems chemistry (von Kiedrowski et al. 2010; Ludlow and Otto 2008; Dadon et al. 2008) that situation has changed. There is now definitive evidence to suggest that the two seemingly discrete processes of abiogenesis and Darwinian evolution are actually one single continuous process, and, significantly, that the process can be associated with an identifiable driving force—the drive toward greater stability. However, the relevant stability kind is not thermodynamic stability, the one which traditionally governs physicochemical process, but rather one that is associated solely with replicating entities, and termed dynamic kinetic stability (DKS) (Pross 2009, 2011, 2012; Pascal 2012a, b; Pross and Khodorkovsky 2004). Being able to identify this “other” stability kind is of fundamental importance, as it enables the biological phenomenon of evolution to be placed squarely within a physicochemical framework. No longer need chemistry and biology be viewed as disparate sciences with different scientific methodologies, as was claimed in the past (Mayr 1988). The merging of abiogenesis and evolution reaffirms the intimate chemistry-biology connection and opens up new means for uncovering the chemical underpinnings of the entire gamut of biological phenomena, including central ones such as function and complexity.

The aim of this article, therefore, is to build on the DKS concept and the conceptual merging of abiogenesis and biological evolution to: (a) understand how function, a phenomenon so specific and so characteristic of living things, could have emerged naturally from a world without function, (b) consider the mechanistic implications of the emergence of function, in particular with regard the problematic issue of complexity, and (c) explore the possibility of a simple underlying relationship linking these two central biological concepts, function, and complexity. The discussion that follows builds on some of our earlier ideas on biological complexity (Pross 2005).

Discussion

The Emergence of Function

As noted above, it is now generally accepted that life emerged from inanimate matter, and that the seeds of that transformation began with the emergence of some self-replicating chemical system, most likely based on a limited network of oligomeric structures (Pross and Pascal 2013; Pross 2012; Lazcano 2008; Luisi 2006; Fry 2000; Eigen 1992; Joyce 1989). But, even presuming that the emergence of life was initiated by such a system, how would that process have led to the emergence of function, natural function? Let us now clarify how this could have come about.

A fundamental, even axiomatic, principle in nature, alluded to by Dawkins, is that less stable systems tend to become transformed into more stable ones (Dawkins 1989). The principle is axiomatic because it is implicit in the definition of the term “stable”—stable meaning “persistent, unchanging over time.” Given that matter is not immutable, that it can undergo change, it follows logically that less stable forms will tend to be transformed into more stable forms. But within the physicochemical world, as noted above, there are two discrete stability kinds—thermodynamic stability, a stability kind that reflects a system’s potential to undergo reaction, and DKS, a stability kind associated solely with persistent replicating entities, and which reflects the system’s ability, through replication, to maintain a continuing presence over time. As we will now describe, the emergence of function is a direct outcome of the existence of this second, less familiar, stability kind.

The reason the replication reaction gives rise to an alternative stability kind, DKS, stems from its unique kinetic characteristics. Any persistent replicating system composed of a population of individual replicators is a dynamic far-from-equilibrium steady-state system in which the component replicators are continually turning over. One consequence of this dynamic state is that competing replicators within that population may not co-exist over time. As Lifson pointed out some years ago, when two replicators, 1 and 2, compete for the same building blocks according to the simple kinetic scheme:

$$ {\text{d}}X_{1} /{\text{d}}t = k_{1} M \, X_{1} - \, g_{1} X_{1} $$
(1)
$$ {\text{d}}X_{2} /{\text{d}}t = k_{2} M \, X_{2} - \, g_{2} X_{2} $$
(2)

(X 1 and X 2 are the concentrations of the two replicators, 1 and 2, respectively; M is the concentration of building blocks from which the replicating system is composed; k 1 and g1 are rate constants for replicator 1 formation and decay respectively; and k 2 and g 2 are the corresponding rate constants for replicator 2 formation and decay respectively), the steady-state solution is one in which the more effective replicator (reflecting larger k, smaller g) drives the poorer one into extinction. In other words, through a process of kinetic selection, a population of less stable replicators will over time be transformed into a population of more stable replicators (Lifson 1997). Though this pattern was inferred on the basis of the simple kinetic scheme depicted in Eqs. 1 and 2, it appears to be quite general. It is observed through detailed kinetic modeling (Wattis and Coveney 1999) and, more importantly, in competitive RNA replication experiments, such as those conducted by Spiegelman almost 50 years ago (Mills et al. 1967) (though the precise selection patterns expected in particular mechanistic circumstances has been the source of some dispute, see Scheuring and Szathmáry 2001; Lifson and Lifson 2001). In other words, once a population of replicating systems emerges, one in which all members compete for the same limited resources, there is a general tendency for more effective replicators to displace less effective ones. Crucially then, this biological-like pattern is manifest not just at the biological level through natural selection, but at the chemical level as well, through kinetic selection (Pross 2011, 2012).

But now to the central point: we have seen there exists a natural law that tends to convert poorer chemical replicators into better ones, expressing the general drive toward greater replicator DKS. But the very existence of this natural law, based on the existence of a replicative stability kind, leads directly to the emergence of function. The fact that the replication process tends to improve naturally leads to the creation of function, replicative function. In contrast to the “regular” chemical world, where it is the Second Law that governs the direction of irreversible chemical processes, in the replicative world there is a Second Law analog (Pross 2009) that is effectively operative—from less DKS stable to more DKS stable (though consistent, of course, with the requirements of the Second Law). The operation of this distinct selection rule in the world of replicators means that through the abiotic emergence of some persistent replicating system, a natural function comes into being.

Note, no other reaction type exhibits such behavior. There is no natural law that transforms a less combustible material into a more combustible material, or a weaker oxidizing agent into a stronger oxidizing agent. Both combustion and oxidation reactions are governed by the Second Law and no function is associated with these or any other chemical processes. It is that distinct stability kind—replicative stability, DKS—that leads to the replication reaction acquiring the characteristic of function. Note, also, that the appearance of a replicating entity as the unit of selection (sensu Dawkins 1989), would, in itself, be insufficient to explain the emergence of function. The fact that a replicating molecule or molecular system may emerge, replicate sporadically, undergo occasional variation and then undergo sporadic kinetic selection, will not result in an evolutionary process that leads toward life. For an evolutionary process to lead to living systems—systems that are characterized by their dynamic, far-from-equilibrium state—selection must take place between persistent replicators, i.e., populations of replicators that have accessed that dynamic, far-from-equilibrium state, and which must be maintained by an on-going supply of material and energy resources.

Till now the discussion has focused on the replicative function, but, of course, biology is replete with functions other than replication—from ribosomes that churn out polypeptide chains through to pumping hearts. All such functions are, however, secondary functions, functions that necessarily derive, directly or indirectly, from the primary function, the fundamental drive toward increasingly stable persistent replicators. A moment’s consideration though suggests that such secondary functions have come about through a process of complexification. Secondary functions are almost invariably associated with highly complex supramolecular entities. Let us therefore now consider the nature of the complexification process to discover the manner in which function and complexity are related.

The Relationship Between Function and Complexity

Some 70 years ago, Erwin Schrödinger puzzled over life’s extraordinary complexity. In his words: “Now, why are atoms so small? Clearly, the question is an evasion. For it is not really aimed at the size of the atoms. It is concerned with the size of organisms, more particularly with the size of our own corporeal selves.” (Schrödinger 1944). Simply put, Schrödinger was asking “why is life complex?”. His tentative answer was to point to the randomness of atomic motion and the need for systems to be composed of a large number of atoms to allow for appropriate “statistical laws” to operate. Let us now address this issue through a functional perspective.

We mentioned previously that the world manifests an incomprehensibly large range of functional entities, some entirely natural—hearts, kidneys, ribosomes, etc., and some man-made—cars, computers, chairs, toothbrushes, whatever. But even superficial inspection of those functional entities reveals a clear and unambiguous underlying pattern—the existence of a general relationship between functionality and complexity. Improved function is almost invariably associated with increased complexity. While this relationship is also one that cannot be readily quantified, there can be little doubt as to its veracity. Consider the Wright Brothers 1903 airplane and any modern aircraft. The modern aircraft is orders of magnitude more complex than the early airplane and that complexity has been introduced for one single reason—to improve airplane function. Step by step, model by model, airplanes have become more complex to enhance reliability, to extend airplane range and speed, to facilitate navigation, to increase passenger comfort, and so on. That process of complexification has involved both an increase in the number of components as well as the diversity of those components. To clarify the importance of diversity in improving function, consider a tool box with ten different tools—hammer, pliers, screwdriver, chisel, saw, etc. Such a tool box would be more functional to a tradesman than one with, say, ten screwdrivers. Diversity, then, tends to facilitate function. The bottom line is clear: almost any man-made entity—cars, computers, houses, chairs, whatever—that has undergone evolutionary change to improve its function, will invariably have been found to have become more complex, both in the number and the diversity of its component parts.

Given this fundamental connection between function and complexity, it is not surprising that nature has discovered and exploited that very principle, and the process of abiogenesis illustrates that clearly. Life appears to have been initiated by the emergence of some relatively simple autocatalytic system, likely limited in molecular numbers and types (though its precise nature remains a subject of uncertainty and dispute). But the end result of that long evolutionary process toward life is unambiguous—a bacterial cell, highly complex both with respect to the number and the diversity of its molecular components—nucleic acids, proteins, carbohydrates, lipids, metal ions, etc. And the driving force for that evolutionary process? The drive toward greater DKS. The primordial system, being simple, would have been functionally poor, of relatively low DKS. Indeed, in order for a simple molecular replicating system to replicate, skilled chemists must toil in well-equipped laboratories under carefully monitored conditions, and even then replicative success is not assured. In contrast, the bacterial cell, that exquisitely complex network of interacting molecules and molecular aggregates, which emerged from that long process of abiogenesis, is able to replicate unassisted just about anywhere it can survive, whether in the soil in one’s backyard, in a hydrothermal vent under the sea, or even in the cooling waters of a nuclear reactor, despite the high levels of radiation. The evolutionary process of complexification came about solely to facilitate replicator function, thereby leading to an increase in replicator stability, DKS. Note, then, that the moment natural function emerged was the moment that nature began to act much like an engineer, when functional design, so ubiquitous throughout biology, began to appear naturally.

In recent years studies in systems chemistry (von Kiedrowski et al. 2010) have provided valuable empirical support for this view. Following on from earlier work by Sievers and von Kiedrowski (1994), recent studies on RNA replicating systems indicate that network formation—an expression of chemical complexification—enhances replicative capability. Lincoln and Joyce (2009) observed that from an extensive RNA library, a molecular network based on two cross-catalyzing RNAs replicated rapidly and could be sustained indefinitely, whereas the most effective single molecule RNA replicator replicated slowly and its self-replication could not be sustained. But, in an even more striking example, Vaidya et al. (2012) recently discovered that a set of three self-replicating RNAs was able to generate a cooperative replicative cycle, which could out-compete those same RNAs when acting as individual replicators. Their study reaffirmed the replicative advantage of complexification, as manifested in network establishment, emphasizing the importance of autocatalytic sets, as envisaged by Stuart Kauffman several decades ago (Kauffman 1986, 2000). Thus, the above experimental studies lend support to the proposed general relationships between complexity, functionality, and (dynamic kinetic) stability. These relationships may be expressed most simply as follows:

Within the world of persistent replicators:

More functional is more (DKS) stable

More functional is more complex

More (DKS) stable is more complex

These general relationships provide a conceptual framework for understanding the evolutionary trend of complexification during evolution, both in its early prebiotic phase, abiogenesis, as well as during its later biotic phase and offer a simple answer to Schrödinger’s provocative question. Thus, living systems are large (in atomic terms) because of the extensive molecular complexity required to generate high replicative stability. While the introduction of a concept foreign to physics and chemistry—that of function—might be of initial concern, the manner in which that concept is able to link between two chemically familiar terms—stability and complexity—appears to justify its introduction into what is fundamentally a chemical analysis.

Finally, an important point regarding the function-complexity relationship should be made to dispel possible misunderstanding. It is eminently clear that function is not always improved by complexification, that complexification is just an underlying trend. For example, Spiegelman’s classic experiments on molecular evolution led to a process of simplification, not complexification (Mills et al. 1967). In the artificial resource-rich environment provided in those test-tube experiments, viral Qβ RNA oligomers shortened over time, discarding those segments of the genome that proved redundant in that artificial environment. But, of course, this apparent exception to the rule only serves to reaffirm the DKS concept. When simplification acts to enhance DKS, then simplification is what will take place. More generally though, the underlying evolutionary trend would be DKS enhancement through complexification. Interestingly, the evolutionary process of simplification, noted above in a chemical context, is also well-established in biology, for example, in cave-dwelling animals which lose their eyesight as they adapt to life in the dark. But that similar pattern, found in both chemical and biological systems, only serves to strengthen the chemistry-biology connection, the unity of the evolutionary process, and further illustrates the explanatory power of the DKS concept.

Concluding Remarks

This paper has sought to reveal the chemical origin of function, a term seemingly out of place within the physicochemical world, yet one that is ubiquitous within the biological world. By demonstrating that the chemical roots of function can be found hidden within the kinetic nature of persistent replicators, our analysis has attempted to shed further light on the chemical nature of abiogenesis as well as its intimate connection with biological evolution. Significantly, the analysis offers a physicochemical explanation for the fact that the evolutionary process of abiogenesis has fundamentally been one of complexification—from some fragile and relatively simple autocatalytic molecular system, through to robust and highly complex life forms. The conclusion: the process of complexification in persistent replicating systems is just the response of those systems to that most fundamental of driving forces in nature—the drive toward greater stability, in this case replicative stability, the stability kind that is manifest through improved replicative function. The analysis thus serves to uncover the underlying connection linking function, complexity, and stability within the world of persistent replicators. That connection leads to a further strengthening of the chemistry-biology union and reaffirms that abiogenesis is just one phase of an extended evolutionary process, which incorporates both animate and inanimate replicating systems.

Needless to say the relationships between function, complexity, and stability described above, is qualitative in form. No explicit mathematical relationships were proposed, when it is clear that quantification, as predicated by the scientific method, would have been preferable. However, as those who have struggled to capture biological complexity in mathematical form have learned the hard way, the nature of physical reality is such that not all relationships are amenable to explicit quantification. The difficulty in this case lies in the fact that biological complexity encompasses not just the individual living entity, not even the population of entities to which the individual belongs, but must also reflect the material diversity of the environment in which the population is embedded. The inherent connectivity of living systems to one another and to their ecological milieu, suggests that all attempts to usefully quantify biological complexity will likely remain tantalizingly out of reach. No matter. As Darwin’s revolutionary concept of natural selection has demonstrated so incontrovertibly, the power and significance of qualitative relationships should not be underestimated. Ultimately, the insight that there is a stability kind in nature quite distinct from the traditional thermodynamic kind, one associated solely with persistent replicating systems far-from-equilibrium, opens new doors to addressing many of biology’s most fundamental questions, including the ones addressed in this paper: how function would have been able to emerge from a universe initially devoid of function, and why that evolutionary process largely manifested itself through increasing (but unquantifiable) complexification.