The Book of Works

In physics, chemistry, geology, climatology and even some of the social sciences, theories and laws are born from models of systems, generally based on observations or experimental data. Molecular biologists also use the term model in their papers and presentations.  These so-called models or cartoons of mechanistic pathways tend to resemble roadmaps with lots of arrows pointing to and from long lists of gene products. But such cartoons are not really models; they are summaries of data. Models have specific purposes, one of which is to simplify and focus the experimental work; another is to lead to new hypotheses for testing. Building a real model is usually difficult and requires a good deal of creative effort.

And what a model needs, by definition, is a mathematical treatment of the idea being modeled. Mathematics has been successfully applied to certain areas of biological science, such as ecology, enzymology, and genetics. But there are surprising gaps in how mathematical models are used in biology. One of the most surprising to me is the lack of any mathematical laws of evolution.

The usefulness of a mathematical theoretical approach to evolution becomes clear when one considers the enormous complexity of the systems being studied. Evolutionary biology has become very broad and now covers most aspects of molecular biology, developmental biology, genomics, physiology, systematics, ecology and genetics. There has been a great deal of theoretical work done on many of these areas, and the overall theory has been described, elaborated and modified continuously.

There have even been a number of mathematical treatments of various aspects of evolutionary theory, such as the use of fitness differences in the Hardy-Weinberg law to predict allele frequencies over time. A search of the literature for math in biology will reveal a plethora of equations related to theoretical treatments of evolution. But despite a wealth of formulas (many of which are not comprehensible to non-mathematicians), there are none that rise to the level of a law of nature,  because they are either speculative, or too specialized in their application, or because some terms (fitness being notorious for this) are impossible to define.

On the face of it, it does seem strange (at least to me) that no universal law has been stated for the process of evolution. The components of a possible model are clearly known. Organisms inherit genes and resulting phenotypes from their parents. There is a diversity of such phenotypes due to mutation, recombination, and other factors. Some phenotypes increase the probability for survival and reproduction, and others don’t. These simple statements should, one would imagine, allow for some form of mathematical treatment that could be simply (and perhaps elegantly) stated as a law (or more likely laws) of evolutionary biology.

The problem, of course, is in the details. And biologists are quite keen to drill down as far as they can to uncover more and more details about how life works. Most of these details cannot be accounted for in any general law or theory. So the argument goes that any simple law would fail to incorporate various modes of genetic variation, or neutral drift, or some other aspect of how evolution actually happens. The hard part therefore is come up with such laws that are general enough to be fairly simple and universal, but at the same time, are able to at least acknowledge, if not account for, the unending complexity of living creatures, how they live, and how they evolve.

But while the task is hard, the reward for success is quite tangible. With  good theoretical laws, certain doubts about evolution as the central basis for biological science might be laid to rest, and hopefully a good deal of disparate and confusing information could be integrated into a unified context. And biology would become another of the sciences governed by the rule of law.

No, the title of this post doesn’t mean that I think biologists are criminals. It means that biology is a science without laws (or with very few of them). And by law, of course I mean a construct, generally written in symbolic or mathematical terms, that expresses a theory. Despite the few well known examples of biological theory, such as evolution by natural selection, and some ecological paradigms, biology is notoriously short on theory. I have heard colleagues dismiss any idea of theory in biology as impossible, or non-scientific. And that, I believe, is a serious problem.

An argument could be made that despite the unquestionably rapid pace of scientific progress in biology, our actual level of understanding of the mechanisms that operate in living cells is less profound, less integrated and less easily communicated than it should be given the enormity of the data that has been generated. We are often faced with fragmented fields and subfields of research into an explosive number of subcellular factors, proteins, genes and pathways, each studied in isolation or in association with a very limited subset of the bewildering complexity that is the reality of the cell.

I believe that the reason for the discrepancy between the weight of facts and the breadth of knowledge that we have about cell biology stems from a paucity of theoretical efforts in the field. In all other fields of science, theoretical and experimental work are mutually tied together and tend to progress together, each driving the other, and leading to some degree of an integrated framework of understanding. The absence of any theoretical framework in molecular biology has severely hampered attempts to make sense of the data in an integrative manner.

Biological scientists are not as familiar with theory and its importance as are physicists, and a large part of the problem is the relunctance of molecular biologists to acknowledge  the nature of the problem. One response to a plea for more theory in biology is that since biological systems are far more complex than the systems physicists deal with, biologists are forced to focus on relatively narrow areas and cannot hope to address global questions with comprehensive theories. This response betrays the typical biologist’s ignorance of the role of theory in science, which has always been precisely to take large masses of uncoordinated information, which alone seem to suggest unfathomable complexity, and use them to create an orderly, logical, testable and ultimately satisfying understanding of nature that we call a theory.

The advantage of a theory is that it allows for a simplified picture of nature and makes large numbers of difficult experiments unnecessary. Without a theory of gravity, for example, one could imagine one or more journals devoted to the topic of how things fall. Papers would be published describing experiments in which investigators observed the fall of bricks, rocks, oranges, apples, bags filled with various materials, all from different heights, with measurements taken on the speed of descent, force of impact, the effects of climate, altitude , temperature etc. While we might now smile at the absurdity of such experiments, we do so only because we know that Newton’s theory of gravity holds true and makes  all these experiments  superfluous.

There are examples of this process in biology. Both Darwin and Wallace, and many other biologists before them, collected vast amounts of data about the morphology and comparative anatomy of species. These data now seem boring and trivial when compared to the elegance of the theory of evolution by natural selection that emerged from them. In the 18th and 19th centuries biologists performed hundreds of experiments trying to prove or disprove the idea of spontaneous generation of life. Reams of data were published comparing the growth of “animalcules” in infusions of hay, vs. turnips, vs. potatoes vs. chicken broth and so on. The idea that any organic medium exposed to a source of microorganisms will allow for their growth obviated the need for all these redundant experiments.

It is the theories and ideas that allow for real scientific progress. Of course such theories require experimental data in order to be formulated. But if all we do is collect data, without any effort at theoretical formulation, we may be fooling ourselves that we are learning more about the secrets of life when in fact we are simply generating noise.

(Part 2 will discuss biological theories in more detail, with an emphasis on possible laws of evolution). 

(This series of three posts is taken from a review article I am planning to submit for publication. I have removed references from the posts, but will be happy to supply them on request.)

Natural Genetic Engineering (cont.)

There is strong evidence that a massive genomic alteration event occurred at about the time of the origin of the vertebrates.  At some point between the origin of chordates and that of jawed vertebrates, an entire genome was duplicated, at two different times. Whole genome duplication (WGD) is an extremely useful (and rare) event in evolutionary terms, because it allows for a great deal of genomic trial and error in organisms, without interference from purifying or balancing selection. By providing an extra, non-essential copy of every gene, WGD allows for very rapid and dramatic evolutionary leaps, such as the development of new structures and functions (cartilage and bony skeletons being the relevant story here). There is also good evidence that similar WGD events have occurred in flowering plants, at the origin of teleost fishes, and probably at many other critical evolutionary transition points.

Duplication of individual genes or smaller groups of genes (such as a chromosome) can also lead to major evolutionary changes. Gene duplication is often mediated by a mechanism called retrotransposition, whereby a gene is duplicated at a new location thanks to the action of genetic elements called retrotransposons. Such events were found to occur during primate evolution, when the common ancestor of gorillas, chimps and humans split from the orangutan line. Gene amplification leads to a similar result – the production of many copies of a single gene, that is also likely to be induced by stress. Exon shuffling, repetitive elements play an important role in gene duplication and new gene creation in files.

Another mechanism for rapid large-scale genomic change is horizontal gene transfer, whereby one organism transfers a large chunk of genetic material to another organism. This is a well-known phenomenon in bacteria. It now appears that such genetic transfers have taken place between prokaryotes like bacteria and eukaryotes, like parasites and sponges. Horizontal gene transfer could also explain the origin of animal like alpha amylase from animals and plants to bacteria. Horizontal gene transfer from bacteria to eukaryotes has been linked to the origin of mineralization in sponges, which led to the eventual development of skeletons in modern animals.

All of the mechanisms described in this series are forming part of what has been called the Extended Evolutionary Synthesis (EES), which many evolutionary biologists believe should and will replace the standard neo Darwinian synthesis. There are many more aspects of the EES than have been covered here, including niche selection, epigenetics, and the control of gene expression during development (Evo Devo). These will be described in future posts.

(This series of three posts is taken from a review article I am planning to submit for publication. I have removed references from the posts, but will be happy to supply them on request.)

Stress Directed Mutations

In 1988, a paper by John Cairns and his colleagues data showed that bacteria could produce beneficial mutations targeted specifically to relieve severe stress. Cairns’ paper took a major step away from the “purely random” concept for mutation.  These beneficial mutations (now called stress-directed mutations or SDM) are produced at rates up to 5 times higher than other mutations with neutral effects.  Numerous researchers have confirmed this phenomenon, and have found a number of molecular mechanisms to account for it. Hypermutability of specific stress-related genes can result from a variety of environmental stressors, like starvation, or changes in osmolarity, temperature or anaerobiosis.

Stress leads to derepression of specific genes, whose function are related to the stress. The resulting higher level of transcription of these genes allows for unpaired and exposed bases in loop structures that are more susceptible to mutation. Several investigators have found  evidence that mutants arising from SDM in starving bacteria arise from different molecular mechanisms than the ordinary mutational events.  Most mutations due to SDM occur in newly derepressed genes. Derepression of genes can lead to supercoiling and much higher mutation rates in the genes affected. Supercoliing of DNA during selective gene transcription is one of the leading molecular precursors of  SDM in bacteria. Such changes in supercoiling that can lead to hypermutability can result from a variety of environmental stressors, like changes in  osmolarity, temperature or anaerobiosis.

Natural Genetic Engineering

Over the past decades, microbiologist James Shapiro has applied many findings on how cells can accomplish major genomic alterations  to develop a model he calls Natural Genetic Engineering (NGE). His view is that the cell can control the genome as much as the genome controls the cell.  When applied to evolution, these sources of genetic variation do not fit the neo-Darwinian model of slow progressive changes,  but are rapid, dramatic, and involve grand molecular events such as whole genome duplication, transposition of DNA sections leading to massive re-engineering of proteins, horizontal transfer of coding regions from plastids, viruses and other organisms.

One such revolutionary event was the huge evolutionary step taken  when a  cell engulfed a bacterium  that remained alive and functional within its host, giving rise to eukaryotic  cells with mitochondria. Nobody thinks that event was a slow stepwise process. Dawkins has described it as a one-time incredibly lucky accident, more or less equivalent to the origin of life.  (In fact it happened at least twice, since chloroplasts  also started out as bacteria swallowed by an ancient plant cell.)

(This series of three posts is taken from a review article I am planning to submit for publication. I have removed references from the posts, but will be happy to supply them on request.)

The idea that the source of variation in individuals of a species is random and not in any way directed did not come from Darwin. In the Origin of Species, he states:

“I have hitherto sometimes spoken as if the variations…were due to chance. This, of course is a wholly incorrect expression, but it serves to acknowledge plainly our ignorance of the cause of each particular variation.”

As the quote demonstrates, Darwin simply had no idea, and more importantly, the distinction between chance and purpose really had no direct consequence on the general theory.

The issue of randomness or chance is closely tied in with one of the most essential questions in biology, is there a purpose or direction to evolution? While the standard answer among most biologists is no, there is an increasing amount of evidence that the standard dogma might be wrong. The existence of teleology in nature, and in evolution has not been ruled out, and the work of Simon Conway Morris on convergence, and his demonstration that evolution in fact follows fairly narrow pathways, restricted by biological constraints supports the idea of reexamining this question. Others, such as Francisco Ayala have found evidence for teleology in the very nature of adaptive change.

The Neo Darwinian Synthesis, and Alternatives

In the middle of the 20^th century, even before the discovery of DNA as the genetic molecule, biologists were examining mutations in experimental systems of bacteria, to answer questions about purpose and chance in mutation production. Do bacteria tend to specifically mutate those genes that would help them survive an environmental stress, such as starvation or exposure to toxic drugs, or do they simply generate random mutations, and then selection chooses which ones allow for increased fitness? Luria and Delbruck seemed to have answered this question in the 1940s  with an elegant system called fluctuation analysis . The results of these experiments were clear: mutations were random, and then selected for their relative fitness. This finding contributed to the emerging  neo-Darwinian synthesis, with molecular genetics playing the key role in the production of phenotypic variation, and with a confirmation of the idea that purpose is replaced by chance in the mechanism of the first stage of evolution.

This idea became engrained in the biological dogma, and as more and more data regarding the nature of genes and how they operate and change became available, the prevailing consensus grew stronger. Evolution became a theory that neither required, nor admitted to any degree of purpose or design. . The  proponents of Neo-Darwinism maintain that evolution is a slow, continuous process wherein the effects of small scale point mutations accumulate gradually over vast stretches of time to produce changes in the phenotype of species, as well as lead to the emergence of new species and even new genera and phyla.

But there now exist  a great deal of new data and new theoretical constructs that are chipping away at the standard neo-Darwinian paradigm. The eventual acceptance of the neutral drift theory led to a modification of the role of positive adaptationist mutations as the only drivers of evolutionary change, especially as related to population genetics and microevolution.

The ideas of Mayr and Gould on punctuated equilibrium, first roundly rejected by dogmatic neo-Darwinians such as Richard Dawkins, have been debated for decades, but there is now emerging a body of solid evidence that provides strong molecular mechanisms for them. The fossil record seems to show long periods (on the scale of hundreds of millions of years) of very little change, “punctuated” by remarkable brief “moments” (in geological time) of explosions of new forms. The Cambrian explosion is the best known of these, but there are many other examples. While the paleontological data  is consistent with brief moments of very dramatic changes, no molecular mechanisms were put forward to explain how this could  have happened. That has now changed.