8. Population genetics of prokaryotes

Transcription

1 8. Population genetics of prokaryotes Henrik Christensen, Department of Veterinary Disease Biology Faculty of Health Sciences, Copenhagen University Prokaryotic populations Mutation Selection Genetic drift Migration Biological consequences Practical use Prokaryotic Populations and population genetics. Population genetics is the study of evolutionary change in the genetic composition of populations (Whittam 1995). According to Whittan (1995) population genetics applies both to how the mechanisms (mutation, natural selection, migration, genetic drift) influence the evolutionary rate of change in the populations (Fig. 1) as well as to historical investigation of how and when pathogens have evolved. The outcomes of such investigations will be definition of population structures, knowledge of the nature of allelic variation and the role of different modes of recombination in generating genotypic variation (Milkman 1973, Selander & Levin 1980, Whittam 1995). In animals, a population is defined as a group of individuals with a potential for sexual reproduction often limited to a certain geographic region at a certain time point. However, this definition can not be used for prokaryotes due to the lack of sexual reproduction. Groupings of prokaryotes are better reflected by clonal relationships than physical barriers. A prokaryotic clone was originally defined as a group of prokaryotic isolates showing so many identical phenotypic and genetic traits that the most likely explanation for this identity is a common origin (Ørskov and Ørskov 1983). Bioinformatics in Microbiology 1

2 Populations make up species and species are thought to originate by differences in populations that have lead to isolation and diversification. For 99% of eukaryotes, the isolation mechanism between populations is prevention of production of fertile hybrids between these populations. In eukaryotes without sexual reproduction and all prokaryotes populations become isolated eventually leading to speciation by other means. Population genetics of prokaryotes for this as well as other reasons is completely different from population genetics of eukaryotes. Already the way species are defined demonstrates this difference. If we used the prokaryotic criteria (DNA-DNA hybridization, 16S rrna sequence comparison) (Box 1), for genotypic separation of species in the two groups then the bacterial species Escherichia coli would compare to the group of all primates (Staley 1997). Speciation in eukaryotes has classically been related to geographic separation with allopatric species formed because of physical barriers. In prokaryotes this mechanism has not been documented except for recent reports for a thermoacidophilic archaeon (Staley 2009) and sympatric speciation with formation of new species in the same place as the parent species is the most probable mechanism. The difficult task with prokaryotes is to set up the limits for the population. A practical solution might be to define the population borders according to the role of their members in causing disease or other properties useful for edidemiology and investigation of pathogenesis. However, populations need to be determined by certain genotyping methods and their principles are very much determining for how we observe the populations. These methodological problems with identification of clonal populations are illustrated in Fig. 2. Method 1 is optimal to link clusters with isolate properties in relation to host and disease association. Method 2 has slightly lower resolution that method 1 and results in the lack of separation of populations 1 and 2. The non-disease related isolates 3 and 4 are therefore included in this population which is confusing compared to method 1 that allowed the separation of populations associated disease from the non-pathogenic. Method 3 has higher resolution than method 1 and this will not contribute with any error as long as it is taken care of that some clusters might belong to the same population. Method 4 has the same resolution as method 1, however, isolates are clustered differently and will give another interpretation about their role in disease. In the mid 1990s it was found that Mannheimia haemolytica can spread from domestic sheep to wild bighorn sheep in Montana, USA. The spread was verified by one of the most sensitive genotyping methods, pulse field gel electrophoresis (PFGE) (Ward et al. 1997; 1999). Bioinformatics in Microbiology 2

3 When similar comparison were done between M. haemolytica found to cause disease in wild - and domestic animals in the Alps no such clonal relationship was found by PFGE, however, when the isolates were compared by MLST it was found that the same sequence type was involved telling that the resolution of PFGE was just too high (Kodjo et al; Petersen et al. 2009). To compare with Fig. 2, PFGE would compare to method 3 and MSLT to method 1 in this example. The term strain is sometimes misused to equal a population or a clone. In the strict sense a strain is an isolate that has been further characterized, archived and documented. The misuse is related to the fact that two isolates that share phenotype and genotype in principle must be considered as belonging to the same strain Mutation The most likely starting point for a population genetic event is a mutation (Fig. 1). The progeny of a prokaryotic cell should in principle be genetically identical to its ancestor. Two genetic mechanisms, point mutations and horizontal gene transfer (HGT) tend to abolish this identity over time. Mutations will increase genetic diversity between individuals. If such events will lead to evolutionary changes will also depend on the other population genetic processes including the potential to increase the potential for survival of the organism (Fig. 1). Point mutations accumulate with a nearly constant rate at random positions in the sequence and we can analyse them by sequence comparison until a certain level of divergence. HGT, however, is more problematic to analyse, since large fragments of sequences are exchanged between individuals and events of transfer can not easily be traced since the origin of such fragments are often unknown. Recombination in prokaryotes has a completely different meaning than in eukaryotes. In eukaryotes, recombination refers to the results of crossing over in a symmetrical way between chromosomes during the zygotene stage of meiosis. This happens between members of the same species and even with members of the same population. HGT between prokaryotes can take place by transformation, transduction and conjugation and usually occur a as an asymmetrical exchange events between partners. Some bacteria like Haemophilus influenzae are natural transformable meaning that they can take up DNA directly from the environment. In bacteria with a double wall structure, DNA is taken up as double Bioinformatics in Microbiology 3

4 stranded across the outer and as a linear single stranded across the inner membrane and uptake signal sequences (USS) have been found to favor uptake (Maughan et al. 2008). Transduction is the transfer of genetic material by bacteriophages. There is high variability in the ability of bacteriophages to cross-react with strains within the same species, between different species of the same genus and between different genera of the same family (Jones and Sneath 1970). Not all bacteria can be transformed, for example four different bacteriophages have been identified in H. influenzae, however none of them seemed suitable for transduction (Fink and Geme 2001). Conjugation is when genetic material is transferred between two prokaryotes on a special type of conjugational plasmid. Luria & Delbruck (1943) pioneered the investigation of prokaryotic population genetics by investigating how bacteria become resistant to bacteriophages. They found resistance to develop as a random process since resistance to bacteriophages developed independent on the presence of bacteriophages in cultures of bacteria sensitive to bacteriophages. Furthermore, they analyzed mutations in a statistical framework and found that the distribution of mutational events followed a special distribution (Luria & Delbruck distribution). Mutations rates are still measured based on the principles laid down by Luria & Delbruck (1943). The most common procedure is the fluctuation test. In this test, the distribution of mutants in a number of parallel cultures is used to estimate the mutation rate based on knowledge of the expected number of mutation events, the number of cultures and the size of initial inoculum. A range of assumptions are taken in this calculation: constant probability of mutations per cell cycle, that the mutation rate is independent of growth phase, no cell death occur and revertants are not formed, only single mutants arise, the same growth rate occur for wild type and mutants, that negligible number of mutant cells initially are present compared to final numbers in cultures investigated and that we are able to detect all mutants (Pope et al. 2008). The mutation rate is the probability of a mutation occurring per cell division. Another measure is the mutation frequency being the proportion of mutant prokaryotes present in a culture. The mutation rate is independent on the age of the prokaryotic culture and a more accurate measure of mutations compared to the mutation frequency that is related to the age of the prokaryotic culture. In principle we should be able to trace mutations directly to the DNA sequence level (Bell 2008). In practice this will require an enormous DNA sequencing efford, Bioinformatics in Microbiology 4

5 however, next generation sequencing technologies might be able to do this job in order to measure mutation rates directly and not via fluctuation tests Selection Periodic selection is a prominent population genetic mechanism of prokaryotes. If a specific allele is favored in the population, and this allele results in higher fitness, then all members of the population with this allele will replace other populations without the allele in the given environment by expansion of this clone. Periodic selection is a form of bottleneck (Levin 1981). The higher rate of HGT the less is the effect of periodic selection (Levin 1981). The degree of selection acting on a specific coding gene can in theory be predicted by calculation of the so-called dn/ds ratio (Nei 2005). The ratio is the ratio of non-synonymous substitutions per non-synonymous site to that of synonymous substitutions per synonymous site. A non-synonymous substitution at DNA level is a change in nucleotide leading to a change in the amino-acid translated from the codon where it occurs. It follows that a synomynous substitution is not affecting the amino-acid translated from the codon. The theory is based on the neutral theory : without any positive selection, dn/ds = 1. Negative selection (dn/ds < 1) occurs when deleterious alleles (recognized at amino acid level) are eliminated from the population by purifying selection and leaving only the synonymous changes to be observed. In positive selection (dn/ds > 1), polymorphism is assumed to be maintained at the amino acid level and the changes here will be relatively higher than those observed at synonymous sites. However, high polymorphism at the amino acid level might not necessarily be observed for positive selection to occur (Nei 2005). Textbook examples of positive selection is with peptide-binding sites of MHC genes from humans and mice as well as the antigenic genes of influenzae virus (Nei 2005). To do the calculations, at least two closely related nucleotide sequences need to be compared pair-wise. Computer programs can evaluate the substitutions in regards to all combinations of nucleotides that are legal for each codon Genetic drift Genetic drift is related to sampling error and sampling error will be most pronounced in small populations. The random distribution of certain genotypes within a population may lead to genetic drift if the populations is small. This way clones developed by periodic selection may be Bioinformatics in Microbiology 5

6 wiped out more or less by chance. The effect is linked to selection in the way that the weaker and more unpredictable periodic selection, the higher the effect of genetic drift. If the selection coefficient s is less that 1/N, where N is the population size, then genetic drift is predicted to be a serious hindrance to selection Migration For migration to take effect on evolution, the introduction of individuals should have consequences on population genetic processes like allelic frequencies and mutation rates. Investigation of the spread of populations including bacterial spread between animals is not population genetics as long as such spread is not affecting the evolution of the organisms. Investigation of such spread is part of epidemiology. Box 1. The genotypic prokaryotic species concept The species represents the main taxonomic unit in classification. Species include prokaryotes with 16S rrna gene sequence similarities higher than 97.0 %. If 16S rrna sequence similarity higher than 97 % between different species candidates exists then DNA-DNA hybridization or comparable methods should be applied to confirm the diversity of such species since members of a species also need to be related with more than 70 % DNA reassociation (Stackebrandt et al. 2002). The comparison of housekeeping gene sequences has recently been used for genotypic definition of species. Based on the work of Zeigler (2003), Kuhnert & Korczak (2006) proposed that sequence comparison of the three genes recn, rpoa and thdf could replace DNA-DNA hybridizations. They further showed that even sequencing of a single gene, recn, might serve this purpose. Further investigations are needed before final conclusions can be made as to the possibility of fully substituting DNA-DNA hybridizations with this method. Whole genomic sequence comparison to group species at the genotypic level is now in progress ( e. g. Bisgaard et al. 2012) The biological consequences of population genetics of prokaryotes. Bioinformatics in Microbiology 6

7 In theory, the population structure of a prokaryote species is determined by the ratio of genetic changes caused by recombination relative to de novo mutation (Spratt and Maiden, 1999) meaning that if HGT is relatively higher than point mutation rates, the population structure will be very complex and diverse, whereas low degree of recombination relative to point mutation will result in well defined populations at the sequence level. Only in the last case are we able to recognize clonal populations (Fig. 3). The distribution of alleles investigated by MLST can be used to predict if the population structure is clonal or non-clonal (panmictic). With equal assortment of alleles there should be an equal random distribution of alleles between populations of a species. The expected variance V E should equal the observed variance V O. If the populations have evolved like clones, the alleles will be identical or highly related within clones and very different between clones and V O will be higher than V E. To compare V E and V O, the index of association is calculated (I A = ((V O /V E ) - 1) and it follows that I A is not significantly different from 0 with a panmictic population structure but significantly different with a clonal population structure. The clonal population structure is observed for most bacteria. The calculation can be performed on the MLST web applications. According to one theory, the evolution of virulent prokaryotes may take place by clones with high level of natural competence allowing uptake of foreign DNA from other prokaryotes by HGT (Maughan et al. 2008) The importance of genetic drift within animal populations and in ecological niches outside animals remains to be investigated. In larger animal populations, the population structures of prokaryotes are predicted to be mostly influenced by selection. Some clonal lineages of a species seem to have adapted to specific hosts. In addition, these lineages often cause disease to a higher extent than other lineages, and for epidemiological investigations it is therefore of great importance to identify and understand prokaryotes at the population level. One outcome of population genetics investigations is more realistic diagnostic methods for prokaryotic populations involved in disease. If population structures reflect disease patterns and hosts, we will be able to determine the populations that are really responsible for disease and not their commensal sister groups (Fig. 2). Ecotypes are groups of prokaryotes playing ecological distinct roles defined based on DNA sequences. Ecotypes are analysed by construction of phylogenetic trees based on Bioinformatics in Microbiology 7

8 housekeeping genes of isolates representing populations of a species. At certain level of depth of a cluster in the tree, a group of populations that equal an ecotype is defined. Simulations are used to select this level of cutoff for the ecotype as well as to estimate periodic selection and genetic drift. Ecotypes are as a consequence one or more clonal populations if we refer to the current species consept of prokaryotes. Ecotypes can be regarded as species if we redefined the prokaryotic species consept, however, multiple ecotypes are usually recovered within the traditional species (Koeppel et al. 2008) Practical use of population genetics to investigate animal and human infections. As a rule of thump, a primary infection will be caused by only one clonal population of the bacterial agent involved. Often this clone will manifest itself as massive monocultural growth on primary isolation plates. It is necessary to determine the population structure by analyzing a representation number of isolates from the outbreak for example by MLST. If infections are recurrent the choice of control measure will also often depend on the primary (vaccination) or secondary (antibiotics) nature of infection. References Bell, G Selection. The mechanism of evolution. 2 nd ed. Oxford Univ. Press. Bisgaard, M., Norskov-Lauritsen, N., de Wit, S.J., Hess, C., Christensen, H., 2012, Multilocus sequence phylogenetic analysis of Avibacterium. Microbiology 158, Fink, D.L., and Geme, J.W.St. III. (2001). The genus Haemophilus. In The Prokaryotes: an evolving electronic resource for the microbiological community. Ver. 3.13, M. Jones, D. and Sneath, P. H Genetic transfer and bacterial taxonomy. Bacteriol Rev. 34, Kodjo, A., Villard, L., Bizet, C., Martel, J.L., Sanchis, R., Borges, E., Gauthier, D., Maurin, F., Richard, Y., Pulsed-field gel electrophoresis is more efficient than ribotyping and random amplified polymorphic DNA analysis in discrimination of Pasteurella haemolytica strains. J. Clin. Microbiol. 37, Koeppel A, Perry EB, Sikorski J, Krizanc D, Warner A, Ward DM, Rooney AP, Brambilla E, Connor N, Ratcliff RM, Nevo E, Cohan FM Identifying the fundamental units of bacterial diversity: a paradigm shift to incorporate ecology into bacterial systematics. Proc Natl Acad Sci U S A. 105: Kuhnert, P. & Korczak, B. M. (2006). Prediction of whole genome DNA-DNA similarity, determination of G+C content and phylogenetic analysis within the family Pasteurellaceae by multilocus sequence analysis (MLSA). Microbiology 152, Levin, B. R Periodic selection, infectious gene exchange and the genetic structure of E. coli populations. Genetics 99, Bioinformatics in Microbiology 8

9 Luria, S.E. & Delbrück M Mutations of Bacteria from Virus Sensitivity to Virus Resistance. Genetics 28, Milkman R Electrophoretic variation in Escherichia coli from natural sources. Science 182, Maughan, H., Sinha, S., Wilson L. and Redfield, R Competence, DNA Uptake and Transformation in the Pasteurellaceae. In Kuhnert, P and Christensen, H. Pasteurellaceae, Biology, genomics and molecular aspects. Caister Acad. Press. In press. Nei, M. (2005). Selectionism and neutralism in molecular evolution. Mol. Biol. Evol. 22, Nei, M. & Kumar, S Molecular evolution and phylogenetics. Oxford Univ. Press. Petersen, A., Christensen, H., Kodjo, A., Weiser, G. & Bisgaard, M Development of a multilocus sequence typing (MLST) scheme for Mannheimia haemolytica and assessment of the population structure of isolates obtained from cattle and sheep. Infection, Genetics and Evolution 9, Pope et al A practical guide to measuring mutation rates in antibiotic resistance. Antimicrob. Agents Chemotherapy 52, Selander RK, Levin BR. Genetic diversity and structure in Escherichia coli populations Science 210, Spratt, B. G., M. C. J. Maiden (1999): Bacterial population genetics, evolution and epidemiology. Phil. Trans. R. Soc. Lond. B, 354, Stackebrandt, E., Frederiksen, W., Garrity, G. M., Grimont, P. A., Kampfer, P., Maiden, M. C., Nesme, X., Rossello- Mora, R., Swings, J., Truper, H. G., Vauterin, L., Ward, A. C. & Whitman, W. B. (2002). Report of the ad hoc committee for the re-evaluation of the species definition in bacteriology. Int J Syst Evol Microbiol 52, Staley, J. T Biodiversity: are microbial species threatened? Current Opinion in Biotechnology 8, Staley, J The phylogenomic species concept. Microbiology Today 09, Ward, A.C., D. L. Hunter, M. D. Jaworski, P. J. Benolkin, M. P. Dobel, J. B. Jeffress, G. A. Tanner (1997): Pasteurella spp. in sympatric bighorn and domestic sheep. J. Wildlife Dis. 33, Ward, A.C., N. W. Dyer, B. W. Fenwick (1999): Pasteurellaceae isolated from tonsillar samples of commerciallyreared American bison (Bison bison). Can. J. Vet. Res. 63, Whittam, T. S Genetic population structure and pathogenicity in enteric bacteria. In Population genetics of bacteria. Eds. Baumberg, S., Young, J. P. W., Wellington, E. M. H. and Saunders, J. R nd Symposium of the Society for General Microbiology. pp Cambridge Univ. Press. Zeigler, D. R. (2003). Gene sequences useful for predicting relatedness of whole genomes in bacteria. Int J Syst Evol Microbiol 53, Ørskov, F., I. Ørskov (1983): Summary of a workshop on the clone concept in the epidemiology, taxonomy, and evolution of the Enterobacteriaceae and other bacteria. J. Infect. Dis. 148, Bioinformatics in Microbiology 9

10 Fig. 1. Population genetics of prokaryotes Selection (Factors: host, vaccination, medication, disinfectants, environmental tolerance, nutrients etc.) Mutation (point mutation and horizontal gene transfer) Evolution Genetic drift (limit 10 8 cells?) Migration (potential for horizontal gene transfer and change in allelic frequency)

11 Fig. 2. Model illustrating dilemmas between methodological and biological relevance of populations given one species. Clonal population are defined according to method 1 providing optimal biological interpretation (hosts A, B, disease associations X, Y) for the isolates investigated in the table M e t h o d 2 : r e d u c e d r e s o lu t io n H o s t A D is e a s e X P o p u la t io n 1 ( is o la te s 1,2,3,4 ) H o s t B D is e a s e Y P o p u la t io n 3 ( is o la te s 5,6 ) P o p u la t io n 4 ( is o la te 7 ) P o p u la t io n 5 ( is o la te 8 ) M ethod 1: used to define populations o th e r s p e c ie s Host A Host B Disease X Disease Y Population 1 (isolates 1,2) Population 2 (isolates 3,4) Population 3 (isolates 5,6) Population 4 (isolate 7) Population 5 (isolate 8) M e t h o d 3 : in c r e a s e d r e s o lu t io n D is e a s e X H o s t A D is e a s e Y H o s t B o th e r s p e c ie s P o p u la t io n 1 ( is o la te 2 ) P o p u la t io n 1 ( is o la te 1 ) P o p u la t io n 2 ( is o la te 3 ) P o p u la t io n 2 ( is o la te 4 ) P o p u la t io n 3 ( is o la te s 5,6 ) P o p u la t io n 4 ( is o la te 7 ) P o p u la t io n 5 ( is o la te 8 ) other species M e t h o d 4 : lo s s o f a c c u r a c y Isolate Host Disease 1 A X 2 A X 3 A 4 A 5 B Y 6 B Y 7 B 8 A and B o th e r s p e c ie s H o s t A H o s t B D is e a s e X D is e a s e Y D is e a s e Y P o p u la t io n 1 ( is o la te s 1, 2 ) P o p u la t io n 2 ( is o la te s 3,4 ) P o p u la t io n 3 ( is o la te s 5 ) P o p u la t io n 4 ( is o la te 6, 7 ) P o p u la t io n 5 ( is o la te 8 )

12 Fig. 3. Clonal: horizontal gene transfer << point mutations Non-clonal: horizontal gene transfer >> point mutations I A > 0 I A = 0