O R I G I NA L A RT I C L E doi:10.1111/j.1558-5646.2009.00821.x

THE EFFECT OF ELEVATED MUTATION RATES ON THE EVOLUTION OF COOPERATION AND VIRULENCE OF PSEUDOMONAS AERUGINOSA Daniel Racey,1,2 Robert Fredrik Inglis,2 Freya Harrison,2,3 Antonio Oliver,4 and Angus Buckling2,5 1

Peninsula Medical School, Plymouth, PL6 8BU, United Kingdom

2

Department of Zoology, University of Oxford, Oxford OX1 3PS, United Kingdom

3

Department of Biology & Biochemistry, University of Bath, Bath BA2 7AY, United Kingdom

4

´ Hospital Son Dureta, Instituto Universitario de Investigacion ´ en Servicio de Microbiologia and Unidad de Investigacion,

Ciencias de la Salud (IUNICS), 07014, Palma de Mallorca, Spain 5

E-mail: [email protected]

Received July 25, 2008 Accepted August 3, 2009 Within-host competition between parasite genotypes can play an important role in the evolution of parasite virulence. For example, competition can increase virulence by imposing selection for parasites that replicate at a faster absolute rate within the host, but may also decrease virulence by selecting for faster relative growth rates through social exploitation of conspecifics. For many parasites, both outcomes are possible. We investigated how competition affected the evolution of virulence of the opportunistic pathogen Pseudomonas aeruginosa in caterpillar hosts, over the course of an approximately 60 generation selection experiment. We initiated infections with clonal populations of either wild-type bacteria or an isogenic mutant with an approximately 100fold higher mutation rate, resulting in low and high between-genotype competition, respectively. We observed the evolution of increased virulence, growth rate, and public goods cheating (exploitation of extracellular iron scavenging siderophores produced by ancestral populations) in mutator but not wild-type, populations. We conclude increases in absolute within-host growth rates appear to be more important than social cheating in driving virulence evolution in this experimental context.

KEY WORDS:

Competition, parasite, public goods, mutator, siderophore.

Within-host competition between parasite genotypes can theoretically play an important role in the evolution of virulence (parasite-mediated reductions in host fitness) (Bremermann and Pickering 1983; Nowak and May 1994; van Baalen and Sabelis 1995; Frank 1996; Brown et al. 2002; West and Buckling 2003; Buckling and Brockhurst 2008). However, the effect of competition on virulence will depend on the details of parasite interactions (Buckling and Brockhurst 2008). If competitive interactions are limited to simple resource competition, then within-host competition is expected to result in selection for increased virulence (Bremermann and Pickering 1983; Nowak and May 1994; van Baalen and Sabelis 1995; Frank 1996), as “imprudent” parasites with the fastest absolute within-host growth rate come to dom-

1

 C 2009 The Author(s). Evolution

inate (Hamilton 1964; Maynard Smith 1964; Bremermann and Pickering 1983; Nowak and May 1994; van Baalen and Sabelis 1995; Frank 1996; Grafen 2006). By contrast, if the growth rate is largely determined by the production of metabolically costly public goods, such as toxins, iron-scavenging siderophores or the matrix required for biofilm formation, then competition will select for decreased virulence (Brown et al. 2002; Buckling and Brockhurst 2008). This is because public goods cooperation is open to invasion by “social cheats” that reap the rewards of the public good without paying the cost of its production (Hamilton 1964; Maynard Smith 1964; Brown et al. 2002; West and Buckling 2003; Grafen 2006; Buckling and Brockhurst 2008), and population growth rate will decrease as these social cheats increase in

DA N I E L R AC E Y E T A L .

frequency. Finally, intermediate levels of competition will minimize virulence in situations where competition is mediated by costly intraspecific toxins, because it is under these circumstance that toxin production will be maximized (Gardner et al. 2004; Massey et al. 2004; Inglis et al. 2009). Empirical support has been obtained for all three predictions (e.g., Herre 1993; Ebert 1998; de Roode et al. 2005, 2008; Harrison et al. 2006; Inglis et al. 2009). It is inevitable that multiple types of intraspecific interactions will sometimes (if not always) occur within hosts. Many pathogenic bacteria, for example, produce public goods to obtain nutrients (Brown et al. 2002) and manipulate the immune system (Brown et al. 2006), release anticompetitor toxins (Riley and Wertz 2002), as well as competing directly for resources (de Roode et al. 2005, 2008). The relative importance of different parasite interactions which operate simultaneously in driving virulence evolution is, however, unclear because they have only been studied in isolation. Our primary aim is to determine how competition between genotypes of the pathogenic bacterium Pseudomonas aeruginosa affects the evolution of virulence in insects where both absolute (asocial) within-host growth rates and public goods cooperation are free to evolve, and can, to some extent, be quantified. P. aeruginosa is not a specialized parasite, but rather an opportunist that can colonize most hosts given appropriate transmission opportunities. It is a significant cause of morbidity in cystic fibrosis sufferers and patients who acquire infections as a result of medical intervention (Smith et al. 2006). Experimental studies have shown that natural selection results in increased growth rates of P. aeruginosa (e.g., Perron et al. 2007) in simple nutrient-rich media where public goods production is probably relatively unimportant. Competition between genotypes in vivo could therefore increase virulence. However, P. aeruginosa, like the majority of microbes, produces public goods that affect population growth rates when particular nutrients are limiting. One well-studied example of such a public good is the extracellular iron-scavenging siderophore, pyoverdin (Ratledge and Dover 2000; Buckling et al. 2007). Comparison between siderophore producing and nonproducing isogenic mutants demonstrates that siderophore production provides a growth rate advantage and results in higher virulence in both mouse (Meyer et al. 1996) and caterpillar (Harrison et al. 2006) models. Siderophore production is open to exploitation both in vitro (Griffin et al. 2004) and in vivo (Harrison et al. 2006) by social cheats which pay none of the costs of siderophore production, but can still take up iron-siderophore complexes produced by their neighbors. Clinical data are consistent with the experimental data, with high frequencies of pyoverdin-deficient genotypes observed within the lungs of cystic fibrosis patients (De Vos et al. 2001; Smith et al. 2006). Competition between genotypes could

2

EVOLUTION 2009

therefore also result in a decrease in virulence, because of the selective advantage of public goods cheats. For many parasites, the extent of within-host genetic diversity (and hence competition between genotype diversity) is determined by the number of genotypes that infect a host (Frank 1996). However, the massive population sizes and short generation times of bacterial infections means that within-host mutation is likely to be more important in generating within-host diversity (Andre and Godelle 2006), as recently reported in both acute and chronic P. aeruginosa infections of humans (Smith et al. 2006; Kohler et al., 2009). The large (1000-fold) genetic variation in mutation rates in natural populations of bacteria (LeClerc et al. 1996; Matic et al. 1997) suggests that mutation rate may play a key role in the evolution of bacterial virulence. Recent experimental studies using a range of bacterial species (including P. aeruginosa) report mutators sometimes displaying reduced virulence (Merino et al. 2002; Mena et al. 2007; Montanari et al. 2007), and sometimes equal or increased virulence (Zahrt et al. 1999; Picard et al. 2001), relative to wild-types. These inconsistencies may well reflect differences in the relative importance of different competitive interactions, because mutation rate has been shown to both increase the evolution of growth rates in Eschirichia coli (Giraud et al. 2001) and the rate at which pyoverdin social cheats invade cooperating populations in P. aeruginosa (Harrison and Buckling 2005, 2007). We carried out a simple study to determine how the extent of genetic diversity, mediated by mutation rate, affects the evolution of P. aeruginosa virulence in caterpillar hosts. Elevated mutation rates will increase within-population genetic diversity on which selection acts, potentially resulting in both general adaptations for growth in the host and adaptation to the social environment (i.e., increased social cheating). A wild-type clone and an isogenic mutant with an approximately 100-fold increase in mutation rate were passaged through caterpillars for approximately 60 generations. We then determined the interrelationships between virulence, growth rate and siderophore-mediated cooperation in the ancestral and evolved populations.

Materials and Methods BACTERIAL STRAINS AND INSECT HOST

Strain PAO1 (ATCC 15692) was used as the wild-type, and strain PAOmutS (Oliver et al. 2004), an isogenic mutant of PAO1 which has a deletion of the mismatch repair gene mutS, was used as the mutator. This strain has a spontaneous mutation rate more than two orders of magnitude higher than PAO1. Strains were grown overnight on a rotary shaker (0.9 g) at 37◦ C in 6 mL King’s medium B (KB) (Harrison et al. 2006) to provide cultures for initial inoculation. Fifth instar waxmoth (Galleria mellonella) larvae (from livefood.co.uk) were used as the insect host (Harrison

M U TAT I O N R AT E A N D V I RU L E N C E E VO L U T I O N

et al. 2006). Note that the growth of pyoverdin mutants is not limited in King’s Medium B (Griffin et al. 2004).

Siderophore production of evolved populations was scaled by that of their respective ancestors.

SELECTION EXPERIMENT

VIRULENCE BIOASSAY

Fresh overnight cultures of PAO1 and PAOmutS were diluted in 0.8% NaCl solution, and 103 cells of each bacterial genotype was inoculated into six larvae in 10 μl volumes via injection between the second posterior legs (Harrison et al. 2006). Larvae were kept on ice and surface sterilized with ethanol prior to inoculation. They were placed into separate wells in 6-well titre plates, and left for 24 h at 37◦ C. They were then separately homogenized in 200 μl of 0.8% NaCl solution using a plastic pestle in a centrifuge tube. Following centrifugation at 13,000 rpm for 3 min in a microcentrifuge, the supernatant was further diluted 103 -fold, and 10 μl of this solution (approximately 103 cells) inoculated into fresh larvae. A total of 10 such inoculations were carried out, so that for the mutator and wild-type there were 6 replicates of 10 serial inoculations (approximately 60 generations).

Bioassays were carried out using populations of ancestral bacteria and the 12 selection lines after transfer 5 and 10. Frozen samples were grown overnight in KB broth and diluted in 0.8% NaCl solution. Larvae were swabbed with 70% ethanol to prevent contamination of the injection site and injected with 102 cells (with densities equalized across all replicates using OD 600 measures (optical densities) using a BioTek microplate reader), as above. Fifteen larvae were assigned to each bacterial population, 30 for each of the ancestral populations. A further 15 larvae were injected with 10 μl of the NaCl solution as negative controls: their mortality rate was negligible. The injections were necessarily staggered, with the timing of injections for a particular bacterial population completely randomized. Larvae were incubated at 37◦ C and monitored for death over a 24-h period. Larvae were scored as dead if they failed to respond to mechanical stimulation of the head. The mean time to death for each population was determined and used for subsequent analysis.

SIDEROPHORE ASSAYS

Following homogenization of caterpillars at transfers 5 and 10, a sample of supernatant was plated on KB agar at different dilutions. Agar plates were supplemented with 15 μg/mL ampicillin to select against growth of native larval gut bacteria (Harrison et al. 2006). The plates were incubated overnight at 37◦ C and numbers of green and white colonies scored (out of a total of approximately 200 colonies per plate); green colonies indicate pyoverdin (the primary siderophore of P. aeruginosa) production (Griffin et al. 2004). Each plate was then extensively sampled by repeatedly scraping a sterilized loop across the plate, and then frozen in 20% v:v glycerol:KB broth at −86◦ C for subsequent virulence assays. Each of the frozen samples (and the ancestral genotypes) was grown overnight in KB broth. An aliquot of this wholepopulation mix was centrifuged to pellet the cells, and the supernatant (containing siderophores) was stored at −20◦ C. The total siderophore content of these supernatants was later determined using the chrome azurol S (CAS) method described by Schwyn and Neilands (1987), with the modification that we diluted the CAS recipe 1:1 with ddH 2 O (Harrison and Buckling 2005). The relative absorbance at 630 nm of a mixture of 50 μl supernatant, and 100 μl CAS solution (all chemicals from Sigma) decreases linearly as siderophore concentration rises. Thus a measure of mean siderophore production per (natural log) colony-forming units (CFU) in the ith microcosm is given by [1 − (Ai/Aref)]/[ln(Densityi)] where Ai = absorbance of the ith sample, Aref = absorbance of a reference solution comprising 50 mL sterile growth medium plus 100 μl CAS, and density = CFU in 50 μl of the population sample.

GROWTH RATE ASSAY

Fresh overnight cultures of PAO1 and PAOmutS and the twelve lines that had been passaged 10 times were diluted in 0.8% NaCl solution. Fourteen groups of twenty larvae were inoculated as above, and incubated at 37◦ C for 8 h. Larvae were then dipped in 70% ethanol to kill surface contaminants and homogenized in 500 μl M9 minimal medium using a plastic pestle. Homogenates were centrifuged at 13,000 rpm for 3 min to pellet solid material, and aliquots of diluted homogenate plated on KB agar. Agar plates were supplemented with 15 μg/mL ampicillin to select against growth of native larval gut bacteria (PAO1 being resistant to this concentration of ampicillin). The plates were incubated overnight at 37◦ C and numbers of colonies scored. The mean number of doublings was then determined for each strain. This measure of growth rate may also reflect differences in carrying capacity between strains, assuming logistic growth. However, the densities reached in the current study area at least an order of magnitude less than densities we have determined in a previous study using the ancestral nonmutator strain (Harrison et al. 2006), so it is likely that our estimates are reasonable approximations of exponential growth rates. STATISTICAL ANALYSES

We wanted to determine how the proportion of cheats, in vitro siderophore production, and mean time to death varied through time for both wild-type and mutator populations separately. We used General Linear Mixed Models (with proportion data arcsin square root–transformed to meet model assumptions), fitting line

EVOLUTION 2009

3

DA N I E L R AC E Y E T A L .

(a 6 level random factor) and time (a covariate). Note that all wild-type lines share the single ancestral value for each response variables at time point zero, and likewise for the mutator lines. We extended the above analyses by including proportion of cheats, growth rate and genetic background (fixed effect) as additional explantory variables in models investigating virulence, and additionally looked at the impact of proportion of cheats on growth rates. GenStat 7 was used for analyses.

Results We determined how the virulence of wild-type and an isogenic mutator strain of P. aeruginosa evolved during approximately 60 generations in caterpillar hosts. The virulence (as measured by the rate at which they killed their insect hosts) of the wild-type population did not change through evolutionary time (Fig. 1; F 1,11 = 3.14, P = 0.1), whereas there was a significant increase in virulence through time in the mutator populations (Fig. 1; F 1,11 = 11.85, P = 0.005). Virulence did not however differ (P > 0.2) at the final time point (transfer 10) between mutator and wild-type populations, probably because virulence of the ancestral mutator was less than the ancestral wild-type due to unknown pleiotropic effects of the inactivation of the mutS gene. The increase in virulence of the mutator genotype through time was associated with an increase in within-host growth rate between the evolved and ancestral populations (Fig. 2; 1 sample t-test of evolved lines versus the ancestor: t = 10.34, P < 0.001). As with virulence, no such increase in growth rate occurred in the wild-type lines (Fig. 2; P > 0.2). However, there was no relationship between growth rate and virulence when measured across across the whole data

Figure 2. The growth rate, measured as mean number of doublings (±SEM; n = 6), of ancestral and evolved lines during 8 h

growth in caterpillars.

set (P > 0.2). This is partly explained by the mutator lines having evolved higher growth rates than the wild-type lines by transfer 10 (Fig. 2; 2 sample t-test: t = 3.85, P < 0.01), despite there being no difference in virulence. We also followed the evolution of siderophore production in populations of wild-type and mutator P. aeruginosa during 10 passages through caterpillars. The frequency of pyoverdin-deficient mutants increased in both wild-type (Fig. 3A; F 1,11 = 5.72, P = 0.04) and mutator (Fig. 3A; F 1,11 = 21.43, P = 0.001) populations through time. Mutants were present at frequencies of at least 15% in all 6 mutator lines by transfer 5 and showed no apparent further increase in frequency by transfer 10. By contrast, mutants were only apparent in one wild-type line by passage 5, and in three by transfer 10. Unsurprisingly, mean frequencies of pyoverdindeficient mutants were significantly higher in mutator than wildtype populations at the final passage (t = 3.67, P = 0.006). Consistent with the above results, in vitro siderophore production (per colony forming unit) of the evolved wild-type populations showed no detectable change through time (Fig. 3B; F 1,11 = 0.08, P = 0.8), whereas it decreased by passage 5 and showed no further change at passage 10 in mutator populations (Fig. 3B; F 1,11 = 9.73, P = 0.01). No significant relationships were found between siderophore production and virulence or growth rate (P > 0.2, in all cases).

Discussion Figure 1.

The mean (±SEM; n = 6) time to kill of evolved wild-

type (closed symbols) and mutator populations (open symbols). Note that SEMs of the ancestral populations (time point 0) are not shown, as n = 1 in each case.

4

EVOLUTION 2009

We investigated how mutation rate (and hence the extent of between-genotype competition) affected the evolution of virulence of the opportunistic pathogen, P. aeruginosa, infecting caterpillars. To identify if competition between genotypes resulted

M U TAT I O N R AT E A N D V I RU L E N C E E VO L U T I O N

Figure 3. The mean (±SEM; n = 6) proportion of siderophore (pyoverdin) cheats (A), and population siderophore production (B) through time, in wild-type (closed symbols) and mutator popula-

tions (open symbols). Note that SEMs of the ancestral populations (time point 0) are not shown, as n = 1 in each case.

in selection for greater absolute within-host growth rates, or the breakdown of public goods cooperation, or both, we also followed the evolution of population growth rates and the production of a public good (siderophores). Our key findings are that virulence, growth rate, and public goods cheating all increased (relative to the ancestor) in the mutator but not wild-type populations. Reductions in siderophore production are typically associated with reduced growth rate (Ratledge and Dover 2000; Griffin et al. 2004; Harrison et al. 2006) and virulence (Meyer et al. 1996; Harrison et al. 2006), hence changes in other traits more than compensated for any reductions in population growth and virulence resulting from loss of siderophore production. We think it is unlikely that changes in other social traits are responsible for this increased growth, because most Pseudomonas social traits, like siderophores, involve the production of extracellular, diffusible public goods (West et al. 2007), and thus seem likely to evolve similarly to siderophore production. These results therefore sug-

gest that selection on asocial traits is more important that selection on public goods production in determining growth rate and virulence in this experimental context. Despite the consistency between growth rate and virulence through evolutionary time (i.e., when comparing ancestors with evolved genotypes), no relationship was found between growth and virulence across the whole data set. Given the relatively large degree of replication to obtain each independent data point for both assays, this is unlikely explained by measurement error. Instead, it is likely that other factors, such as toxin production (Gooderham and Hancock 2009), contribute to virulence over and above growth rate. Stronger relationships have been found between virulence and growth rate in other studies using P. aeruginosa and the caterpillar system (Harrison et al. 2006; Inglis et al. 2009), but in these cases the strains used were genetically defined mutants. In the current study, populations were evolved independently for approximately 60 generations, and hence, would inevitably differ from each other in undefined ways. An important result from this study is that siderophoredeficient mutants can rapidly emerge from initially isogenic populations, and have a large selective advantage, in vivo. We have previously shown that siderophore production provides a population growth rate advantage in this animal host (Harrison et al. 2006) (and a similar advantage has been shown in mice (Meyer et al. 1996) and iron-limited media (Griffin et al. 2004)), and that siderophore cheats can exploit wild-type siderophore (Griffin et al. 2004; Harrison et al. 2006). The fitness advantage of siderophore mutants in the current study therefore almost certainly results from their ability to exploit wild-type siderophore without paying a metabolic cost. The advantage of siderophore cheats seems to plateau with increasing cheat frequency in the mutator population (Fig. 3A). This frequency-dependent fitness of cheats can theoretically arise because (1) the caterpillar is a somewhat spatially structured environment, and (2) that the carrying capacity of the population increases with increasing cooperator frequency (Ross-Gillespie et al. 2007). Both explanations probably apply to some extent. This result further suggests that such social exploitation may help to explain the high frequency of public goods mutants in P. aeruginosa infections (De Vos et al. 2001; Smith et al. 2006). Note that results from other studies (e.g., Diggle et al. 2007; Kohler et al. 2009) suggest that selection is also likely to have favored exploitation of other public goods traits that enhance growth rate in this experiment (for example, public goods whose expression is mediated by density dependent “Quroum-Sensing”), although here we only measured siderophore production as a representative public good. Here, we found that mutator populations evolved increased virulence because asocial growth is probably more important than social exploitation in determining population growth rates. However, the relative importance of asocial growth and social

EVOLUTION 2009

5

DA N I E L R AC E Y E T A L .

exploitation in driving virulence evolution is likely to be crucially dependent on biological details. For example, in the present study, the host represents a novel environment for the bacteria, and hence there is likely to be numerous beneficial mutations that will result in increased fitness both in terms of asocial growth rate and social exploitation (Giraud et al. 2001; Orr 2005). By contrast, if pathogens are well adapted to their host, mutators may evolve reduced virulence relative to wild-types. Mutations that increase asocial growth rate are likely to be rare, whereas there still might be numerous mutations that allow social exploitation, and, furthermore, mutators are likely to have a relatively high deleterious mutation load (Taddei et al. 1997). Consistent with this view, virulence evolved to be lower than the wild-type following multiple passages of the specialized parasitic bacterium, Listeria monocytogenes, through a mouse host (Merino et al. 2002). More generally, the balance between changes that increase and decrease virulence is likely to vary during the course of within-host adaptation for a wide range of parasites. We emphasize two further details that we have ignored in our experiments that are likely to alter the association with mutators and virulence. First, between-host selection (our design imposed only within-host selection) can play a major role in driving the evolution of pathogen social interactions and virulence (Bremermann and Pickering 1983; Nowak and May 1994; van Baalen and Sabelis 1995; Frank 1996; Brown et al. 2002; West and Buckling 2003; Buckling and Brockhurst 2008). Second, very different selection pressures may operate in chronic versus acute (as is the case in the present study) infections (Yahr and Greenberg 2004). We have argued previously that the benefits of social cheating within a host may contribute to the high frequency of P. aeruginosa mutators in clinical infections (Harrison and Buckling 2005, 2007), and the current data provide strong support for this hypothesis. Specifically, mutators have a higher propensity to evolve cheat mutations than do nonmutators in vivo, establishing a positive genetic correlation between mutators and cheating. Our results also suggest mutators will more readily adapt to novel host environments, although as discussed above, it is unclear if this advantage of mutators would be realized in chronic infections of Cystic Fibrosis patients, where P. aeruginosa mutators are found at particularly high frequencies (Oliver et al. 2000). Further work is required to understand the links between bacterial virulence, fitness and mutation rates, particularly because mutation rates could potentially be manipulated by external selection pressures such as antibiotics and viral parasites (bacteriophage) (Pal et al. 2007). ACKNOWLEDGMENT We thank NERC (UK), the Royal Society and the Leverhulme Trust for funding this work.

6

EVOLUTION 2009

LITERATURE CITED Andre, J. B., and B. Godelle. 2006. The evolution of mutation rate in finite asexual populations. Genetics 172:611–626. Bremermann, H. J., and J. Pickering. 1983. A game-theoretical model of parasite virulence. J. Theor. Biol. 100:411–426. Brown, S. P., M. E. Hochberg, and B. T. Grenfell. 2002. Does multiple infection select for raised virulence? Trends Microbiol. 10:401–405. Brown, S. P., L. Le Chat, M. De Paepe, and F. Taddei. 2006. Ecology of microbial invasions: amplification allows virus carriers to invade more rapidly when rare. Curr. Biol. 16:2048–2052. Buckling, A., and M. A. Brockhurst. 2008. Kin selection and the evolution of virulence. Heredity 100:484–488. Buckling, A., F. Harrison, M. Vos, M. A. Brockhurst, A. Gardner, S. A. West, and A. Griffin. 2007. Siderophore-mediated cooperation and virulence in Pseudomonas aeruginosa. FEMS Microbiol. Ecol. 62:135– 141. de Roode, J. C., R. Pansini, S. J. Cheesman, M. E. H. Helinski, S. Huijben, A. R. Wargo, A. S. Bell, B. H. K. Chan, D. Walliker, and A. F. Read. 2005. Virulence and competitive ability in genetically diverse malaria infections. Proc. Natl Acad. Sci. USA 102:7624–7628. de Roode, J. C., A. J. Yates, and S. Altizer. 2008. Virulence-transmission trade-offs and population divergence in virulence in a naturally occuring butterfly parasite. Proc. Natl Acad. Sci. USA 105:7489– 7494. De Vos, D., M. De Chial, C. Cochez, S. Jansen, B. Tummler, J. M. Meyer, and P. Cornelis. 2001. Study of pyoverdine type and production by Pseudomonas aeruginosa isolated from cystic fibrosis patients: prevalence of type II pyoverdine isolates and accumulation of pyoverdine-negative mutations. Arch. Microbiol. 175:384–388. Diggle, S. P., A. S. Griffin, G. S. Campbell, and S. A. West. 2007. Cooperation and conflict in quorum-sensing bacterial populations. Nature 450:411– 414. Ebert, D. 1998. Evolution—experimental evolution of parasites. Science 282:1432–1435. Frank, S. A. 1996. Models of parasite virulence. Q. Rev. Biol. 71:37–78. Gardner, A., S. A. West, and A. Buckling. 2004. Bacteriocins, spite and virulence. Proc. R. Soc. Lond. B 271:1529–1535. Giraud, A., I. Matic, O. Tenaillon, A. Clara, M. Radman, M. Fons, and F. Taddei. 2001. Costs and benefits of high mutation rates: adaptive evolution of bacteria in the mouse gut. Science 291:2606–2608. Gooderham, W. J., and R. E. W. Hancock. 2009. Regulation of virulence and antibiotic resistance by two-component regulatory systems in Pseudomonas aeruginosa. FEMS Microbiol. Rev. 33:279–294. Grafen, A. 2006. Optimisation of inclusive fitness. J. Theor. Biol. 238:541– 563. Griffin, A. S., S. A. West, and A. Buckling. 2004. Cooperation and competition in pathogenic bacteria. Nature 430:1024–1027. Hamilton, W. D. 1964. The genetical evolution of social behaviour, I & II. J. Theor. Biol. 7:1–52. Harrison, E. F., and A. Buckling. 2005. Hypermutability impedes cooperation in pathogenic bacteria. Curr. Biol. 15:1968–1971. ———. 2007. High relatedness selects against hypermutability in bacterial metapopulations. Proc. R. Soc. Lond. B. 274:1341–1347. Harrison, F., L. E. Browning, M. Vos, and A. Buckling. 2006. Cooperation and virulence in acute Pseudomonas aeruginosa infections. BMC Biol. 4, 21. Herre, E. A. 1993. Population structure and the evolution of virulence in nematode parasites of fig wasps. Science 259:1442–1445. Inglis, R. F., A. Gardner, P. Cornelis, and A. Buckling. 2009. Spite and virulence in Pseuomonas aeruginosa. Proc. Natl. Acad. Sci. USA 106:5703– 5707.

M U TAT I O N R AT E A N D V I RU L E N C E E VO L U T I O N

Kohler, T., A. Buckling, and C. van Delden. 2009. Cooperation and virulence of clinical Pseudomonas aeruginosa populations. Proc. Natl. Acad. Sci. USA 106:6339–6344. LeClerc, J. E., B. G. Li, W. L. Payne, and T. A. Cebula. 1996. High mutation frequencies among Escherichia coli and Salmonella pathogens. Science 274:1208–1211. Massey, R. C., A. Buckling, and R. Ffrench-Constant. 2004. Interference competition and parasite virulence. Proc. R. Soc. Lond B 271:785–788. Matic, I., M. Radman, F. Taddei, B. Picard, C. Doit, E. Bingen, E. Denamur, and J. Elion. 1997. Highly variable mutation rates in commensal and pathogenic Escherichia coli. Science 277:1833–1834. Maynard Smith, J. 1964. Group selection and kin selection. Nature 201:1145– 1147. Mena, A., M. D. Macia, N. Borrell, B. Moya, T. de Francisco, J. L. Perez, and A. Oliver. 2007. Inactivation of the mismatch repair system in Pseudomonas aeruginosa attenuates virulence but favors persistence of oropharyngeal colonization in cystic fibrosis mice. J. Bact. 189:3665– 3668. Merino, D., H. Reglier-Poupet, P. Berche, A. Charbit, and C. European Listeria Genome. 2002. A hypermutator phenotype attenuates the virulence of Listeria monocytogenes in a mouse model. Mol. Micro. 44:877–887. Meyer, J. M., A. Neely, A. Stintzi, C. Georges, and I. A. Holder. 1996. Pyoverdin is essential for virulence of Pseudomonas aeruginosa. Infect. Imm. 64:518–523. Montanari, S., A. Oliver, P. Salerno, A. Mena, G. Bertoni, B. Tummler, L. Cariani, M. Conese, G. Doring, and A. Bragonzi. 2007. Biological cost of hypermutation in Pseudomonas aeruginosa strains from patients with cystic fibrosis. Microbiology 153:1445–1454. Nowak, M. A., and R. M. May. 1994. Superinfection and the evolution of parasite virulence. Proc. R. Soc. Lond. B 255:81–89. Oliver, A., R. Canton, P. Campo, F. Baquero, and J. Blazquez. 2000. High frequency of hypermutable Pseudomonas aeruginosa in cystic fibrosis lung infection. Science 288:1251–1253. Oliver, A., B. R. Levin, C. Juan, F. Baquero, and J. Blazquez. 2004. Hypermutation and the preexistence of antibiotic-resistant Pseudomonas aeruginosa mutants: implications for susceptibility testing and treatment of chronic infections. Antimicrob. Agents Chemother. 48:4226–4233. Orr, H. A. 2005. The genetic theory of adaptation: a brief history. Nature Rev. Genetics 6:119–127.

Pal, C., M. D. Macia, A. Oliver, I. Schachar, and A. Buckling. 2007. Coevolution with viruses drives the evolution of bacterial mutation rates. Nature 450:1079–1081. Perron, G. G., A. Gonzalez, and A. Buckling. 2007. Source-sink dynamics shape the evolution of antibiotic resistance and its pleiotropic fitness cost. Proc. R. Soc. Lond. B. 274:2351–2356. Picard, B., P. Duriez, S. Gouriou, I. Matic, E. Denamur, and H. Taddei. 2001. Mutator natural Escherichia coli isolates have an unusual virulence phenotype. Infect. Imm. 69:9–14. Ratledge, C., and L. G. Dover. 2000. Iron metabolism in pathogenic bacteria. Ann. Rev. Microbiol. 54:881–941. Riley, M. A., and Wertz, J. E. 2002. Bacteriocins: evolution, ecology and application Ann. Rev. Microbiol. 56:117–137. Ross-Gillespie, A., A. Gardner, S. A. West, and A. S. Griffin. 2007. Frequency dependence and cooperation: theory and a test with bacteria. Am. Nat. 170:331–342. Schwyn, B., and J. B. Neilands. 1987. Universal chemical-assay for the detection and determination of siderophores. Analyt. Biochem. 160:47–56. Smith, E. E., D. G. Buckley, Z. N. Wu, C. Saenphimmachak, L. R. Hoffman, D. A. D’Argenio, S. I. Miller, B. W. Ramsey, D. P. Speert, S. M. Moskowitz, et al. 2006. Genetic adaptation by Pseudomonas aeruginosa to the airways of cystic fibrosis patients. Proc. Natl Acad. Sci. USA 103:8487–8492. Taddei, F., M. Radman, J. Maynard Smith, B. Toupance, P. H. Gouyon, and B. Godelle. 1997. Role of mutator alleles in adaptive evolution. Nature 387:700–702. van Baalen, M., and M. W. Sabelis. 1995. The dynamics of multiple infection and the evolution of virulence. Am. Nat. 146:881–910. West, S. A., and A. Buckling. 2003. Cooperation, virulence and siderophore production in bacterial parasites. Proc. R. Soc. Lond. B. 270:37–44. West, S. A., S. P. Diggle, A. Buckling, A. Gardner, and A. S. Griffins. 2007. The social lives of microbes. Ann. Rev. Ecol. Syst. 38:53–77. Yahr, T. L., and E. P. Greenberg. 2004. The genetic basis for the commitment to chronic versus acute infection in Pseudomonas aeruginosa. Mol. Cell 16:497–498. Zahrt, T. C., N. Buchmeier, and S. Maloy. 1999. Effect of mutS and recD mutations on Salmonella virulence. Infect. Imm. 67:6168–6172.

Associate Editor: M. Travisano

EVOLUTION 2009

7