ClockTutorial #3a - Clock Evolution

EVOLUTIONARY PHYSIOLOGY OF BIOLOGICAL CLOCKS

"Circadian clocks allow organisms to predict, instead of merely react to, cyclic (predictable) changes in the environment". A sentence similar to this one is the opening phrase of many a paper in the field of chronobiology. Besides becoming a truth by virtue of frequent repetition, such a statement appeals to common sense. It is difficult to imagine a universe in which it was not true. Yet, the data supporting the above statement are few and far-between. Believe it or not, the data are not always supporting it either.

This post will attempt to briefly review the literature on evolutionary and adaptive aspects of biological rhythmicity. Also, using the perspectives and the methodology of evolutionary physiology, I will try to suggest some ways to test the hypothesis stated in the first sentence above.

REASONING BEHIND THE ARGUMENT FROM COMMON SENSE

For outside observers of the field of chronobiology and its recent successes in molecular, neural and medical aspects of biological rhythmicity, it may come as a surprise that the field was founded by ecologists, ethologists and evolutionary biologists. When the statements about adaptive function of clocks were initially made, the authors were much more careful than is usually seen today. It was meant as a hypothesis to be tested, and elaborate reasoning was often offered to persuade the reader why it might be true (Daan 1981, Pittendrigh 1967,1993, Enright 1970).

One of the most common arguments that a clock must be adaptive (for one reason or another) was its ubiquity - all plants, fungi, protista, invertebrates and vertebrates (more recently cyanobacteria, too) tested by the pioneers in the field showed circadian rhythmicity. The way those rhythms behaved in the laboratory in various experimental treatments was surprisingly similar over all species. Thus, the reasoning goes, if a physiological mechanism is found in every living thing, and it seems to work in the same way in all of them, then it must have originated early due to natural selection and was preserved over eons due to natural selection.

Some of the earliest experimental work was designed to test the genetic basis of biological rhythmicity. Many generations of laboratory organisms were raised and spent all their lives in aperiodic environments, yet the rhythms persist (Sheeba et al. 1999). Period of the rhythm was species -specific, highly heritable, and very amenable to artificial selection. So, if it is in the genes, the clock must have evolved due to some kind of selective pressure.

When reviewing evolutionary literature on biological rhythms, it is often difficult to distinguish between hypotheses of current utility from hypotheses of origin. It was often assumed that same selective pressures which keep the clocks ticking all over biosphere today, are the pressures responsible for the initial discovery of timing mechanisms by early forms of life.

The current adaptive functions of biological rhythms are often divided into two, mutually not exclusive categories. The Internal Synchronization hypothesis stresses the need for temporal separation of incompatible biochemical and physiological processes within a body (or cell), and for temporal synchronization of processes which need to coincide. An example of the former would be temporal separation of photosynthesis from nitrogen fixation. For the latter, surge of a hormone and availability of its receptor need to be synchronized for the generation of the endocrine effect. Evolution of such timing control mechanisms would presumably alleviate energetic costs of constant production of enzymes and their substrates.

The External Synchronization hypothesis supposes existence of temporal niches. For instance it would be adaptive for an animal to forage at the time when food is available, to hide when the predators are hungry, and to find an individual of the opposite sex at the time conducive to mating. Notably, it was never explicitly stated that possession of a clock is adaptive only if everyone else has a clock, too. So, who had it first?

The two hypotheses are easily meshed. A bird will sleep instead of foraging while its body temperature is low. It will wake up, raise its temperature and corticosterone levels and start to forage when the conditions outside are most conducive to it - when it's warm, light and worms are crawling around. The old adage about the early bird getting the worm is often invoked (and, as R.A.Heinlein once quipped, "it just goes to say that the worm should have stayed in bed").

Coupled to the second hypothesis is also the notion that circadian clock is involved in temporal memory (Enright 1975, Biebach et al. 1991). Thus every event which is committed to memory will, along with information about "what" and "where", also have a time-stamp given by the clock - the "when" of memory. If doing something at one time of day yesterday resulted in survival it might be prudent to do it again today at the exactly same time (Daan 1995).

One important distinction between the origin of the first clock and its current use is in the size and complexity of the organisms in question. The first clock presumably appeared in a unicellular organism. Most of research today focuses on multicellular organisms. In the first case, we are talking about a single-cell clock, in the second about the circadian system, which may be composed of many clocks of different properties, all interacting with each other, with the environment, and with other functions of the body. There is only so much a single cell can do. A complex clockshop which exists at a higher level of organization can evolve complex new functions, e.g., photoperiodic time measurement, time-compensated sun-compass orientation, tidal, lunar and circannual rhythms. In the study of evolution of biological rhythms, are we talking about evolution of the cellular building blocks or evolution of higher-order systems?

CHRONOBIOLOGY IN THE CONTEXT OF EVOLUTIONARY PHYSIOLOGY

The 2000 review of Evolutionary Physiology by Feder, Bennett and Huey currently serves as a manifesto of the field (Feder et al. 2000). In this section, I will follow the organization of their paper to explore how circadian physiology might fit into the framework of evolutionary physiology, as well as how evolutionary approaches may benefit the study of biological rhythms.

Evolutionary physiology is a relatively recent natural outgrowth of the field of comparative physiology, which, in turn, developed out of general mechanistic physiology. The evolutionary physiologists are direct intellectual (and often academic) descendants of traditional comparative physiologists, so the topics remain similar: scaling, symmorphosis, as well as studies of metabolism, respiration, osmoregulation, thermoregulation and locomotor performance. However, other aspects of physiology have been tackled by other groups of scientists, e.g., comparative endocrinology, neurobiology, sensory physiology, digestive physiology and reproductive physiology are vibrant fields of their own. Comparative chronobiology has a long tradition, too (Horton 2001). The importance of comparative studies to understanding of underlying mechanisms can be seen from examples of medically driven areas of physiology (e.g., sleep). Such areas fumbled in the dark for decades due to lack of basic understanding of what the process is all about, which in turn was due to lack of comparative studies of the phenomenon within an adaptive and evolutionary context.

It would be erroneous to state that evolutionary thinking was not a part of physiological thinking in the past, yet evolutionary physiology explicitly imports evolutionary theories, paradigms, concepts, models and techniques into the study of physiological adaptation. Feder et al. (2000) recognize four major developments which led to the rise of evolutionary physiology out of its general and comparative precursors.

First, the critique of adaptationist thinking (Gould and Lewontin1979) had an enormous ripple effect through all of biology, including physiology. The naive assumption that every trait is an adaptation forged by natural selection gave way to a more sophisticated view of evolution that includes non-adaptive forces like constraints and drift. Here, the ubiquity of biological clocks and the inability to imagine a clock-less universe led to the common view that not just the sheer existence of the clock, but every aspect of its mechanism must be an adaptation to the environment which evolved through the agency of natural selection. The non-adaptive evolutionary mechanisms are difficult to fathom in this context, and no explicit statements were made in the literature along these lines.

Second, new awareness of non-independence of species as analytical units for comparative biology has resulted in development of new analytical tools, new approaches to comparative studies, and, most importantly, in the shift in emphasis to ancestor-descendant relationships. Comparative studies of biological clocks were the foundation on which the whole field of chronobiology was built. Yet, these studies were not explicitly phylogenetic. The clocks are so widespread that comparisons were rarely made between closely related species. It was, and still is, more common to compare representatives of plants to fungi to insects to mammals. On those rare occasions when a phylogenetic tree appears in chronobiological literature, it is most likely to be the Tree of Life and not just a tiny piece of it. Thus, the non-independence of species is a problem that did not have much impact on comparative work to date, but is one that will have to be tackled in the future when more analyses of closely related species are performed.

Third, incorporation of tools of evolutionary biology into physiology allows research to state evolutionary hypotheses a priori, instead of post hoc. Using these tools, one is able to monitor evolution in progress and predict future evolutionary trajectories, instead of just registering the results of past evolution. In chronobiology, a few studies have recently been performed along these lines, more concerned with photoperiodic than circadian time-measurement, though (Ben Saad and Maurel 2001, Heideman et al. 1999, Heideman and Bronson 1991).

Fourth, use of most recent tools of molecular biology allows one to escape from standard genetic laboratory models to many other species of interest to physiologists. Comparative molecular analyses of genes involved in circadian rhythmicity have been performed in Diptera and Lepidoptera, and work on other organisms is sure to follow in the near future (Piccin et al. 2000, Peixoto et al. 1998, Costa and Kyriacou 1998, Kyriacou et al. 1996, Saleem et al. 2001).

Evolutionary biology aims to answer two kinds of questions: about the process ("how evolution works") and the pattern ("what has evolution wrought so far in the history of life on this particular planet"). Evolutionary physiology, as its subset, is likewise interested in both process and pattern. After all, one informs the other and vice versa. In the same vein, mechanistic and evolutionary data inform each other and vice versa, too. While traditional comparative physiology explained mechanistically how the organisms adapt to their environments and added adaptive explanations post hoc, the modern evolutionary physiology starts with evolutionary predictions and tests them with comparative studies in the laboratory and in the field.

Of particular interest is the question of optimality - is there a close match between requirements of the environments and the physiological answers to these, or are safety margins and overdesign the rule. It would be difficult to define what would safety margins and overdesign mean in the field of circadian clocks. The clock is at the very core of physiological function of an organism. It acts as a relay station which controls practically every other aspect of biochemistry, physiology and behavior. It is not expected that the cellular clock at the core will be much affected by the environment, but the outputs of the clock, the signals from the clock to all the other functions, are much more likely to be the target of selection. Plasticity of the circadian system and of its coupling to downstream functions is a trait that can be expected to correlate to environmental parameters, e.g., stability vs. variability of the environment.

EVOLUTIONARY CRONOPHYSIOLOGY TO DATE

Several approaches to study of evolutionary pattern and process have been employed by students of biological rhythms. These include studies of variation in natural populations, including latitudinal clines (Skopik and Takeda 1987, Sawyer et al. 1997), tests of correlated traits, e.g., circadian vs. photoperiodic (Majoy and Heideman 2000) or circadian vs. developmental timing (Bloch et al. 2001), and studies of responses to artificial selection.

Comparative studies have been performed on levels from molecular to organismal. However, the choice of species is questionable. If relevant environmental parameter is the light-dark cycle, then most species studied so far occupied the same environmental niche - the earth surface in a temperate region. For a comparative study to be able to discriminate between ecological specialization and phylogenetic inertia, one needs to compare organisms occupying different environments. In the case of clocks, such environments would include polar regions (light-dark cycle of LD 6mo:6mo), equator (constant LD12h:12h), and life below the surface (constant darkness, DD). Although many species have been studied in such conditions in the lab, the study of organisms that live in such conditions in the wild is rare. Recent studies on subterranean and cave animals are a welcome change to the comparative work in chronobiology (Lee 1969, Hoenen and Gnaspini 1999, Riccio and Goldman 2000a, b, Avivi et al. 2001, Koilraj et al. 2000, Trajano and Menna-Barreto 2000).

Most recent data on adaptive function of circadian clocks employed measurements of fitness in altered light environments, as well as comparisons of fitness between genetic mutants of clock function. However, the only measure of fitness used is longevity (Pittendrigh and Minis 1972, Klarsfeld and Rouyer 1998, Hurd and Ralph 1998), and only two papers look at the trade-off between longevity and the lifetime reproductive output (Sheeba et al. 2000, Beaver et al. 2002).

Phenotypic engineering is the hallmark of studies by Patricia DeCoursey, one of the pioneers of the field (DeCoursey et al. 1997, 2000, DeCoursey and Krulas 1998). She measures survival of various species of rodents in the wild after she has removed their circadian pacemakers in the suprachiasmatic nuclei of the hypothalamus. The data are not as straightforward as one might expect - the survival is dependent on how bad was the year in regard to abundance of predators. Masking effects of light on behavior, i.e., burrowing during the day due to bright light, can protect clockless animals from predation in some cases, dependent on the species of rodent and the species of predator. There are two main problems with these kinds of studies. First, the effect of removal of the clock is observed in only one generation. If some clock-less individuals managed to survive and reproduce every year, and if it was engineered (genetically?) in such a way that the progeny were also clock-less, would such a population manage to establish itself in the wild? What alternative mechanisms it would evolve to decrease risk from predation? Would other correlated aspects of timing (e.g., photoperiodism, developmental timing, ultradian rhythms) additionally reduce their fitness? If we are removing a mechanism that was inherited for billions of years, why would we expect that an alternative mechanism would exist to protect the animals in the absence of the clock?

FINDING THE CORRECT ECOLOGICAL CONTEXT FOR EVOLUTIONARY CHRONOPHYSIOLOGY

Earlier in this post I noted that most of the thinking about adaptive function of biological clocks concentrated on adaptation to physical environment - the daily rhythm of light and darkness. That is quite reasonable assumption when one is concerned with the original appearance of a biological clock in the primordial soup, and might still be valid in studies of some unicellular organisms (Dvornyk et al. 2002). Yet, in a world of today, in which it seems that everyone's got a clock, it seems reasonable to assume that biotic aspects of the environment would be the key selective pressure for current uses of biological timing mechanisms. Thus, studies of temporal aspects of interactions between individuals, groups, populations and species can potentially provide important insights into the evolution of biological clocks. That kind of research would also be forced into further sophistication. Many biotic interactions are dependent on niche-constructing traits, as well as on traits that evolved due to simultaneous selective pressure at two or more levels of organization, e.g., genic, individual and group selection acting at the same time either in the same direction, the opposite direction, or at some angle between the vectors.

Several exciting research findings along these lines of thought have been reported recently. First, in 1998, the "resonance hypothesis" was tested in cyanobacteria by Carl Johnson's group at Vanderbilt University. The notion that intrinsic ("freerunning") period of the circadian clock needs to be similar to the period of the entraining cycle (24 hours in nature) was tested in period mutants of Synechococcus sp (Ouyang et a. 1998). Various mutants were exposed to different lengths of the light cycle. Two mutants were tested at the time in competitive assay protocols. In every case, the strain with intrinsic period more closely matching the entraining period won the competitive assay. Thus, being in sync with the environment confers fitness against conspecific competitors (Johnson et al. 1998).

A series of papers appeared recently on temporal niche exclusion between two con-generic species of golden spiny mice living in sympatry in Israel. Originally both nocturnal, in the area of sympatry one of the species adopted a diurnal lifestyle. Physiological adaptations to diurnality, including thermoregulation, osmoregulation and vision, were not altered. The circadian performance in the laboratory was not different either. The only difference was in the coupling between the clock and the behavioral output (sleep-wake cycle) in the field, i.e., detachment of the activity rhythm from circadian control under the ecological pressure of competitor's presence (Kornfeld-Schor et al. 2001).

Fleury et al (2000) tested the hypothesis that the least competitive parasitoids of fruitflies would gain adaptive advantage from parasitizing their hosts earlier in the day than their competitors. Comparison of three species of parasitoids in the lab and field revealed that this indeed is the case, as well as that the time-of-day of parasitism was directly driven by the properties of the circadian clock in each of the three species (Fleury et al. 2000).

Bolas spiders use chemical mimicry to lure their moth prey to within the reach of the bolas - their weapon for capturing prey. Two species of moths that are the favorite prey of this spider are active at different times of night and also produce different blends of pheromones. Bolas spiders produce different pheromone blends to match their prey at correct times of the night (Cesar Gemeno, pers.comm.).

Finally, probably the most impressive example of multiple species involved in an arms race around the circadian clock is the case of malaria (Garcia et al. 2001). Mosquitoes are vectors carrying malarial plasmodia. Mosquitoes in tropical regions are nocturnal and, furthermore, restrict their activity to short periods of night. The exact time of night during which they are active is different for different species of mosquitoes in different geographic regions, presumably to avoid the times when the local bat species are most active hunting. Mosquitoes are attracted to heat, carbon-dioxide and certain odorant molecules. A human (or animal) in fever emits heat, carbon-dioxide and, as components of sweat, the same odorant compounds. Malarial patients (human or veterinary) exhibit fever only during a restricted time of night - the time of night when the local mosquitoes are most likely to bite. Fever is induced by billions of malarial plasmodia simultaneously bursting out of red blood cells into the plasma.