ZOOL 304, Class Notes

Chapter 5, Feb. 21 - 26.

Notes for chapter 1 / 2 / 3 / 4 / 5 / 6 / 7 / 8 / 9 / 10 / 11 / 12 / 13 / 14 / 15 / 16 / 17

304 index page

Introductory comments.

Variation is the raw material for evolution.  Neither genetic drift nor selection can operate in the absence of genetic variation.  

To appreciate the actual extent of variation, we must understand that variation exists as a balance among processes which introduce variation, those which reduce variation, and those which maintain variation.

In this chapter, we shall apply our understanding of mutation, genetic drift, and selection to explain particular patterns of variation.

304 index page

About the "wild type".

In introductory genetics classes, alleles are often labelled as if they came in two distinct types, a single wild type allele and one or more mutant alleles.  In this scheme, the wild type is "normal", and the mutant is "abnormal" (and usually defective).  Historically, this terminology has two sources, one from experimental genetics and one from an early, simplistic understanding of evolution.

For many decades, geneticists have use alleles that are associated with conspicuous phenotypic effects.  Most of these useful, easy-to-recognize markers were originally produced by exposing animals to radiation or mutagenic chemicals.  The process yields damaged, nonfunctional genes that are commonly called "mutant" alleles.  The gene's natural condition (as found in specimens collected from the wild, prior to mutagenesis) is then called the "wild type".  That the "mutant" alleles were unnatural, and that multiple alleles with inconspicuous differences could also be found within the "wild type", were well-known facts but of little concern to the study of inheritance patterns.  

It was also once commonly presumed that natural selection would favor the single "best" allele.  This would be the "wild type" allele, the allele that would predominate in natural, wild populations.  Mutations would occur rarely.  Most newly arising (i.e., "mutant") alleles would be deleterious and hence eliminated by selection.  A rare beneficial mutation would quickly become the new "wild type" as selection eliminated the previously favored allele.

However, when studying evolutionary variation, the terms "wild type allele" and "mutant allele" can be misleading.  All alleles were once new "mutant alleles".  And genes normally found in "wild" populations need not be limited to a single "wild type".  Indeed, the more carefully one looks, the more allelic variation (polymorphism) one finds.  

Note on terminology:  "Polymorphism" refers to distinctly differerent structures (such as alleles or amino-acid sequences), while "variation" refers more generally to any type of difference, including continuously variable quantitative differences.  Genes are commonly described as "polymorphic" if there are two or more allelic forms.  Measurable traits (height, weight, etc.) are called "polymorphic" only if the distribution of variation in a population is bi- or multi-modal (i.e., not bell-shaped but instead with two or more peaks).

Protein polymorphism, revealed by protein electrophoresis or by amino-acid sequence, is common.  Even more common is DNA polymorphism.  However, in contrast with the damaged genes used as markers in experimental studies, all of the various alleles which occur in a natural population may be phenotypically indistinguishable, at least to casual inspection.  Such natural variation is commonly lumped together into the so-called "wild type".  

In evolutionary biology, it can be useful to distinguish between an ancestral allele of a gene and subsequent variants formed by mutation.  But it can become awkward, and potentially confusing, to refer to the ancestor as "wild type" or to apply the label "mutant" beyond the generation in which the new allele first appeared.  Once a new allele becomes established within a population, even at low frequency, it is no longer a "mutant".  It simply becomes one more allele within the gene pool.

304 index page

Mutation rates

"Mutation rate" is an important quantity in evolutionary biology.  However, there are many different mutation rates depending on how the rate is measured (i.e., the particular type of mutation, the species, the specific site in the genome, etc.).  These rates can differ by several orders of magnitude.  Most reported "mutation rates" are estimates, based either on crude phenotypic assessment or on measurement of DNA base-pair substitution rate in a few protein-coding genes.  Most evolutionary-significant mutation rates remain largely unknown.

In fact, our modern understanding of mutation remains severely limited.  Although the evolutionary significance of mutation is beyond question, the "causes and laws of variation" remain in many ways as they were for Darwin (1859, Origin of Species, p. 486), "a grand and almost untrodden field of inquiry".

Mutation has been widely regarded as accidental, as an incidental but inevitable consequence of imperfect gene replication.  

"In short, mutations are accidents, and accidents will happen" (Sturtevant 1937, Quarterly Review of Biology 12:464-467).  

"It is common sense that most mutations that alter fitness at all will lower it" (John Maynard Smith 1989, Evolutionary Genetics, p. 55).

"In other words, natural selection of mutation rates has only one possible direction, that of reducing the frequency of mutation to zero.  That mutations should continue to occur ... is merely a reflection of the unquestionable principle that natural selection can often produce mechanisms of extreme precision, but never of perfection" (George Williams 1966, Adaptation and Natural Selection, p. 138).

The word "mutation" commonly carries negative connotations.  Yet evolution depends on mutation to provide the raw variation for adaptation by natural selection.  In spite of the view expressed by George Williams in the above quote, the mechanisms of DNA metabolism have probably been shaped by indirect selection to yield mutation rates for various sites and classes of mutation that are consistent both with short-term reproductive success and also long-term adaptive evolution in response to changing circumstances.

There is no guarantee that mutation will always supply adequate variation, at the appropriate time, for any particular "need".  This is proven by the simple fact of extinction.  Nevertheless, the existence of adaptation itself demonstrates that mutation can provide (and often has provided) sufficient raw material for adaptive evolution to proceed.  

More:

304 index page

Recombination

Because so many familiar organisms reproduce sexually, the effects of recombination during sexual reproduction must be appreciated in any study of variation.  

Recombination during sexual reproduction increases the amount of variation in a population by creating new combinations of pre-existing alleles.  Recombination entails two mechanisms which occur during meiosis -- reciprocal crossing-over within homologous chromosomes, which creates new chromosomal haplotypes, and independent segregation of chromosomes into gametes.  (More on haplotypes.)

Through sexual recombination, various alleles which originated at different times and in different individuals can come together in the same gamete.  This permits the process of adaptative evolution to proceed much faster and more efficiently that would occur without some such mechanism.  Indeed, the variation produced by recombination may be the primary reason why sexual reproduction is so prevalent.  As Darwin wrote (The Origin of Species, 1859, p. 131), "Some authors believe it to be as much the function of the reproductive system to produce individual differences, or very slight deviations of structure, as to make the child like its parents".  (For more on this function, see Chapter 7.)

Reciprocal crossing over is a prolific source of new DNA sequences, including even novel alleles formed when crossing-over occurs within a gene.  Furthermore, crossing-over yields additional sequence novelty when the exchange of chromosome segments is not quite equal (an occurrence which is quite common at some sites).  However, even though "mutation" is typically defined as any heritable change in DNA sequence, the novel products of crossing-over are seldom considered to be "mutants.  Indeed, recombination is often explicitly excluded from this definition of mutation (e.g., Futuyma 1998, Evolutionary Biology, 3rd edition. Sinauer), presumably because the process is "normal"and seems too orderly to fit the "accidental" connotation often associated with mutation.

Other mechanisms of recombination and gene rearrangement, in addition to crossing over and independent segregation, are known to occur.  In some organisms (e.g., the trypanosome which causes sleeping sickness), gene rearrangement is part of a sophisticated adaptive mechanism for evading a host organism's immune system by facilitating very rapid evolution within the host.

304 index page

Amount of variation.

"[T]he common thread throughout [population genetics] is what Gillespie calls 'the main obsession of our field,' the recurring question, 'Why is there so much genetic variation in natural populations?'"  (from an announcement for Population Genetics: A Concise Guide, by John H. Gillespie, Johns Hopkins Univ. Press).

A couple quotes from our textbook (pp. 98-99) convey the central problem of variation.

Basically, there is lots of variation out there, to serve as raw material for natural selection and for drift.  

Heterozygosity is a measure of allelic variation.  Heterozygosity can be defined several ways. At a single locus, heterozygosity is the proportion of a population which heterozygous (or, equivalently, the probability that a randomly-sampled individual will be heterozygous).  For many loci, heterozygosity is the proportion of loci which are, on average, heterozygous.  

See notes for Chapter 5 end-of-chapter questions for some additional discussion of heterozygosity and haplotypes.

Heterozygosity is affected not only by the number of alleles and the frequency of alleles but also by the extent of inbreeding.  This why h, the measure of "genetic diversity", may be different from H, the measure of "heterozygosity".  A population with several inbred local populations may contain high genetic diversity overall, but the inbreeding within each local population creates much-reduced heterozygosity.

For more on the evolution of molecular variation, see neutral evolution and genetic drift.

Next we will consider several mechanisms which can maintain variation within a population.

304 index page

Mechanisms which maintain variation.

Equilibrium models.  Equilibrium models are quantitative expressions of the common experience that opposing processes often establish a balance, especially when the action of one process produces an increase in the magnitude of the opposing process.  

High levels of genetic variation (as measured by heterozygosity) can be maintained in several ways.  These include:

304 index page

Mutation-drift balance is simply a quantification of a fairly intuitive idea.  IF mutation is constantly introducing new neutral "mutant" alleles into a population (for which there is much evidence), and IF drift is constantly reducing variation (by eliminating some neutral alleles from a population and fixing others), THEN the more variation that accumulates, the faster it will be eliminated.  There must thus be some equilibrium value at which the rate at which new variation is introduced by mutation balances the rate at which that variation is reduced by drift.  For more, see genetic drift.

304 index page

Mutation-selection balance is also a quantification of a fairly intuitive idea.  IF mutation is constantly introducing new deleterious "mutant" alleles into a population (for which there is much evidence), and IF selection reduces the frequency of deleterious alleles at a faster rate when those alleles become more common (remember, slow-fast-slow), the more variation that accumulates, the faster it will be eliminated.  There must thus be some equilibrium value at which the rate at which new variation is introduced by mutation balances the rate at which that variation is reduced by selection.  

For more on allele frequencies at mutation-selection balance, see notes for Chapter 5 end-of-chapter questions.  

You might also want to review the use of Punnett squares to figure allele frequencies.   

304 index page

Migration-selection balance.  Migration, or the equivalent idea of gene flow, is not discussed in Chapter 5.  But the topic was introduced, with a nice example, in Chapter 1, pp. 15-16.

For populations which are not quite isolated from one another, but which are adapted to different circumstances, the movement of individuals can provide a continuing supply of not-so-well-adapted alleles from one population into another.  Individual organisms don't even have to move very far, since sequential matings can allow genes to move much farther than individual organisms ever move.  This can be conceptualized as a "flow" of alleles from one population to the other, hence the term gene flow.

Immigrant individuals can carry alleles that have been selected in a different location under (presumably) different circumstances.  Such alleles are therefore unlikely to be optimal for the local population and will be selected against.  Thus, from the perspective of a local population, a balance between migration and selection is analogous to a balance between mutation and selection.  In both processes, "new" alleles are introduced at some rate into the population and eliminated at some rate by local selection.  An equilibrium will be attained when those two rates balance.  

Unlike mutation-selection balance, migration-selection balance has a strong geographic component.  The amount of gene flow depends on movement of individuals and the extent of variation introduced by migration depends on geographic differences in allele composition among populations.  

The extent of migration is critical for the process of speciation.  Complete isolation guarantees eventual speciation; too much gene flow can prevent speciation.  At in-between rates, the process of speciation depends on several factors, but especially on the strength of selection against in-flowing alleles.

304 index page

Selection.  There are several common circumstances in which selection can actively maintain two (or more) alleles within a population with fixing (or eliminating) any of them.  These include:

Heterosis (also called heterozygote advantage or overdominance).  If a heterozygote has greater fitness than either homozygote, selection guarantees that both alleles will be maintained in the population in a stable equilibrium.  This can be readily computed from the standard population-genetic model of selection (for more see models).  Also see question 5.4 at the end of Chapter 5 (notes on end-of-chapter questions).

Frequency-dependent selection.  In the simple population-genetic model of selection, the single "selection coefficient" is a constant attribute of each genotype.  However, in many real-world circumstances, the fitness of one genotype depends on how many individuals within the population share that genotype.  In some cases, the more common genotypes are less fit because they are more common.  The outcome, then, is the maintainence of two (or several, or many) different alleles in a population.

Perhaps the most intuitively-obvious examples of frequency-dependent selection involve alleles which confer disease resistance.  Many microbes (bacteria, viruses, other pathogens) can evolve very quickly.  Once a microbe has acquired the ability to overcome the defense offered by a particular host-allele, the bearers of that allele will be vulnerable.  That vulnerability is greater for alleles at high frequency (the microbe can spread more readily) than for alleles at low frequency (the microbe will have a hard time finding another suitable host).  In fact, immunoglobin genes are noted both for their exceptionally high allelic diversity (hence the need for careful tissue-typing for organ transplants) and for their high mutation rates (rapidly creating new alleles).

Other examples:

Stabilizing selection.  Stabilizing selection is selection favoring the current value for a quantitative trait (as opposed to directional selection, which drives change toward a higher or lower value).  One might simply expect stabilizing selection to eliminate genetic variation which caused deviation from this optimal value.  However, the effectiveness of artificial selection on quantitative traits suggests that extensive genetic variation exists for most quantitative traits in natural populations.

On the presumption that many (most) quantitative traits are influenced by multiple genetic loci and that each such QTL (quantitative trait locus) may have multiple alleles (see Chapter 4, pp. 81ff.), the existance of extensive genetic variation can be explained by a combination of all the mechanisms (above) known to maintain variation at individual loci.  

As long as there enough loci and alleles to generate very many different haplotypes (see discussion of haplotypes), the genetic background against which an individual allele appears is extremely various.  This, in turn, creates conditions which can foster heterozygote advantage (i.e., a heterozygote at any particular locus works better with a variety of haplotypes than either of the more extreme homozygotes) and frequency-dependent selection (i.e., advantageous heterozygotes are most common when no single allele occurs at high frequency).  Against such a various background, mutations of small effect may also accumulate to fairly high frequency by mutation-selection balance, because against such a various background any particular mutation may be beneficial almost as often as it is deleterious.  Finally, there is some reason to believe that mutations which generate small-scale quantitative genetic variation may occur at substantially higher rates than intragenic point mutations (see, for example, Kashi, Y., D.G. King, and M. Soller, 1997, Simple sequence repeats as a source of quantitative genetic variation.  Trends in Genetics 13:74-78).

Fluctuating environmental conditions.  This mechanism is not emphasized in Chapter 5 but is alluded to elsewhere (e.g., p. 88).  The conditions of selection can (and commonly do) vary from generation to generation.  If different alleles are favored under different conditions, and if conditions change before selection can fix or eliminate any of these varying alleles, then these alleles will be maintained at frequencies which change from generation to generation.

304 index page

Concluding commentary -- History and paradox

Allelic polymorphism is pervasive in natural populations.  Quantitative variation is extensive and apparently inexhaustible.  Historically, these observations have been unexpected and give an appearance of paradox.

Paradox is when two ideas or facts seem to be inconsistent with one another, when the truth of one appears to preclude the other, but both are nevertheless true.

Darwin knew about variation.  Darwin documented variation. But Darwin also accepted variation as a given.  Variation was a puzzle for Darwin -- he didn't know where it came from -- but it was not a paradox.  Variation was simply the way life works.

Then came the modern synthesis.  Our basic understanding of evolution then went something like this.  By selection, a population becomes more-or-less well adapted to its environment.  Mutations (random variants) occur.  Most are deleterious and do not endure for long.  Very occasionally (but often enough, over geological time), a beneficial mutant appears.  That mutant allele then increases in frequency until it becomes fixed in the population, becoming the new "wild type" by replacing the previous wild type.  

As we could see in our population genetics model, selection can move a beneficial allele from moderately low frequency to high frequency in a relatively short time.  Thus we might expect the history of adaptation, when observed at the scale of an individual locus, to look like a long wait, a beneficial mutation, a quick period of selection, another long wait, etc. This understanding carried with it a couple expectations.

However, by the mid-20th century, two observations had become well established.

Thus we went from Darwin's position of taking variation as a given, through the development of a well-supported theory with the expectation that variation should be low, to the discovery that variation is very abundant.

The next stage in this process was the development of explanations for the presence of so much variation. Rather quickly, it was realized that the modern synthesis include several different tools for explaining variation. Now we are in the somewhat embarassing situation of having so many explanations, some yielding rather similar expectations, that we cannot satisfactorily decide which explanation applies most strongly in any particular case.

These several explanations are discussed above, and listed below.

Explanations for the maintenance of genetic variation in a population.

  1. mutation - selection balance
  2. mutation - drift balance
  3. migration ("gene flow")
  4. heterozygote advantage
  5. frequency-dependent selection
  6. fluctuating environmental conditions

Notes for chapter 1 / 2 / 3 / 4 / 5 / 6 / 7 / 8 / 9 / 10 / 11 / 12 / 13 / 14 / 15 / 16 / 17

304 index page

Comments and questions: dgking@siu.edu
Department of Zoology e-mail: zoology@zoology.siu.edu
Comments and questions related to web server: webmaster@science.siu.edu

SIUC / College of Science / Zoology / Faculty / David King / ZOOL 304
URL: http://www.science.siu.edu/zoology/king/304/ch05.htm
Last updated:  28 February 2003 / dgk