Link to home

Population Genetics of Plant Pathogens

Mutation is a change in the DNA at a particular locus in an organism. Mutation is a weak force for changing allele frequencies, but is a strong force for introducing new alleles. Mutation is the ultimate source of new alleles in plant pathogen populations. It also is the source of new alleles that create new genotypes (such as new pathotypes) within clonal lineages. Small populations have fewer alleles due to genetic drift and also because fewer mutations are generated in a small population. The number of effective alleles in a population (i.e. the number of equally frequent alleles in an ideal population that is required to produce the same homozygosity or gene diversity as in an actual population) is 4Neu + 1 (for diploid organisms), where Ne is the effective population size (i.e. an ideal population of given size in which all parents have an equal expectation of being the parents of any progeny individual) and u is the mutation rate. Old populations have more neutral alleles than new populations when Ne is equal. Thus the center of gene diversity for a species is most often also the center of origin for a species. Plants and pathogens have coevolved for the longest time at the center of coevolution, leading to selection for a diversity of resistance alleles in the plant population. This is why plant breeders seek resistant germplasm at centers of diversity. If the pathogen coevolved with its plant host at the center of origin, we predict that the pathogen population also will exhibit maximum diversity at the center of origin.

Mutation plays an important role in evolution. The ultimate source of all genetic variation is mutation. Mutation is important as the first step of evolution because it creates a new DNA sequence for a particular gene, creating a new allele. Recombination also can create a new DNA sequence (a new allele) for a specific gene through intragenic recombination. Mutation acting as an evolutionary force by itself has the potential to cause significant changes in allele frequencies over very long periods of time. But if mutation were the only force acting on pathogen populations, then evolution would occur at a rate that we could not observe.

In plant pathology, we are most often concerned with mutations that affect pathogen virulence or sensitivity to fungicides or antibiotics. In pathogens that show a gene-for-gene interaction with plants, we are especially interested in the mutation from avirulence to virulence because this is the mutation that leads to a loss of genetic resistance in both agroecosystems and natural ecosystems. But mutations from fungicide sensitivity to fungicide resistance also are important in agroecosytems, as are any mutations that affect fitness.

A Simple Mutation Model

To demonstrate how mutation can lead to changes in allele frequencies, let's consider a simple model of mutation. Assume that we have two alleles at a single locus, call them A1 and A2, where A1 can mutate to become A2, and A2 can undergo the reverse mutation to become A1. Let A1 mutate to A2 at a frequency of u per generation. We will call u the forward mutation rate. Let A2 mutate to A1 at a frequency of v per generation. We will call v the backward mutation rate. Let the frequency of allele A1 be pt at time t in the population, and let the frequency of allele A2 be qt at time t. In every generation, a proportion of the A1 alleles will mutate to A2 alleles. This proportion will be the forward mutation rate (u) times the frequency of allele A1 (p), up. In every generation, a proportion of the A2 alleles will mutate to A1 alleles. This proportion will be the backward mutation rate (v) times the frequency of allele A2 (q).

f(A1) = pt       f(A2) = qt 
A1 = avirulence allele, A2= virulence allele

forward (u)

A1 A2

backward (v)


q = upt - vqt
gain  loss

What happens to the frequency of the A2 allele under these conditions? In every generation, the frequency of the A2 allele (q) will increase by up due to forward mutation. At the same time, the frequency of A2 will decrease by vq due to the backward mutation. The net change in A2 will depend on the difference between the gain in A2 and the loss in A2. Delta () q = up - vq for one generation. To calculate the frequency of A2 at generation t + 1, add the change in q ( = delta q) to the original frequency q at time t.

q(t+1) = qt + q

q(t+1) = qt + (upt - vqt)

Application of the Mutation Model

Assume that u = 1 x 10-5 and v = 1 x 10-6 per generation. These are typical forward and backward mutation rates. Let the frequency of A1 (call it p) = 0.99, and the frequency of A2 (q) = 0.01. What is the new frequency of A2 after one generation of mutation?

= 0.01 + [(10-5)(0.99) - (10-6)(0.01)] = 0.01 + 9.89 x 10-6 approximately = 0.01001.

We find that there is not much change in the frequency of A2 after one generation of mutation.

In general, after t generations, the frequency of the A1 (wild-type) allele will be

pt = p0(1-u)t

To calculate the number of generations required to change allele frequencies by a given amount, solve for t, which gives:

t = log(pt/p0)/log(1-u)

We can use this formula to calculate the number of generations needed to change allele frequencies under the assumption that mutation is the only evolutionary force acting on a population. To change the allele frequency by 1%, or 0.01, will take a long time. Assume that u = 10-5 per generation (a high mutation rate). To move the frequency of A1 from 1.00 to 0.99 will take 2000 generations. To move it from 0.10 to 0.09 will take 10,000 generations. In general, as the frequency of the wild-type allele decreases, it takes longer to accomplish the same amount of change.

This simple model should convince you that mutation is a very weak force when it comes to changing allele frequencies.

But mutation is very important for introducing new alleles (new DNA sequences) into populations. The number of alleles in a population will be related to the size of the population. Mutation rates are calculated in units of generations, either per individual, per base pair, or per spore. A mutation rate of 1 x 10-6 can mean that a mutation for a particular gene will occur once every million cells per generation, or once in every million base pairs of DNA per generation. The only mutations that are passed to progeny are those that occur in reproductive cells, such as fungal spores or virus particles or sperm or eggs. A mutation rate of 1 x 10-6 also implies that the mutation occurs at a frequency of one in every million individuals in a population. Mutation rates vary across genes and organisms, but they are usually low and can be considered rare events in most cases (Flor 1958, Zimmer 1963, Gassman et al. 2000).

Assume a mutation rate of u = 1 out of a million spores per generation or 1 x 10-6. This means that, on average, in a population of one million individuals (spores, bacterial cells, or virus particles), you can expect to find one mutant for any given locus per generation. In a population of 10 million individuals, you would expect to find 10 mutants for any locus. And in a population of 1 billion individuals, you expect to find 1000 mutants for any locus. 

Mutations in Plant Pathology

Consider the barley powdery mildew pathogen Blumeria graminis f. sp. hordei. A mature sporulating powdery mildew lesion produces ~104 conidia per day. If 10% of the barley leaf area in a field is infected, there are approximately 105 lesions per square meter in the field, and the daily spore production is approximately 109 spores per square meter or 1013 spores per hectare per day. With a mutation rate of 10-6 at avirulence loci, there would be approximately 107 virulent mutant spores produced in each hectare each day. These virulent mutants can travel out of a field planted to a susceptible barley cultivar and infect a neighboring field planted to a resistant barley cultivar. The virulent mutants that have lost the elicitor encoded by the avirulence allele can infect the resistant cultivar and produce a new generation of virulent progeny. This process appears to have happened many times with powdery mildew and rust fungi in agricultural ecosystems, leading eventually to boom-and-bust cycles. Thus mutation is the critical first stage in producing the "bust."

In general, large populations are expected to have more alleles than small populations because there are more mutants present for selection or genetic drift to operate on. This is one reason to keep pathogen population sizes as low as possible in agroecosystems. In theory, if the pathogen population size is kept low (< 106), you would not expect to find many mutant alleles for any particular gene, including avirulence genes.

In addition, large populations usually contain more alleles because they experience less genetic drift. Genetic drift leads to a reduction in the number of alleles in a population.

Finally, the diversity of alleles at a locus will be affected by the length of time a population occupies a particular area. Over thousands of generations, many mutations will be introduced into a population and some of these will increase to a detectable frequency as a result of selection or genetic drift. Both of these processes may take a long time to make a measurable increase in allele diversity. This concept of a "center of genetic diversity" is used to identify the center of origin of a host plant and its pathogens. The center of genetic diversity is usually the center of origin for both host and parasite and it marks the place where coevolution has likely occurred for the longest period of time. As a result of coevolution, the center of origin is expected to have the largest diversity of plant resistance alleles, as well as the largest diversity of pathogen virulence and avirulence alleles.

Mutations and Plant Breeding Strategies

Consider the effects of creating resistance gene pyramids. If two new resistance genes are introduced simultaneously into a host genotype, then the pathogen will need two simultaneous (or sequential) mutations from avirulence to virulence to overcome those two resistance genes. If we assume a typical mutation rate of 10-6, then the probability of two mutations occurring in the same pathogen strain is (10-6) x (10-6) = 10-12. So in the theoretical mildew population described earlier, only about 10 double mutants would be produced each day. If three resistance genes were pyramided into the same host genotype, then the pathogen would require three mutations, at a probability of 10-18, to overcome the resistance gene pyramid. In this case, we expect to find one triple mutant per 105 hectares of host. This is why plant breeders are interested in utilizing resistance gene pyramids. Our theoretical prediction is that R-gene pyramids are likely to be very effective for asexual pathogens like bacteria and some asexual fungi, such as Fusarium spp. (Mundt 1990).

But given this scenario, and knowing that it is relatively common to find 106 hectares of a plant host planted in a region for some cereals, you should wonder why resistance gene pyramids don't break down immediately since most "pyramids" include only 2 or 3 resistance genes. The answer is that the very rare mutant spore is extremely unlikely to land on a suitable host plant and then encounter favorable conditions for infection. The majority (probably 99.99%) of spores produced by biotrophic pathogens fall to the soil, are lost to the air, land on a non-host, or do not encounter an environment favorable for infection when they land on a suitable host. Thus most of the rare mutants never have an opportunity to infect and reproduce. This highlights the fact that both epidemiology (in this case spore numbers) and population genetics (allele frequencies) are needed to explain observed phenomena in plant pathology.

Go to Knowledge Test for Interactions/Genetic Structure

Go to References

Next Section