Microbial in action Richard E. Lenski

In his introduction Most of us first encountered evolution as child- more fit than its ancestor under one set of conditions does to this issue on ren, when we saw the fossil remains of not imply that it would be more fit elsewhere. microbial evolution, and other extinct organisms in museums. As Some years ago, I began a long-term experiment in Richard Lenski biologists, we also see the grand sweep of evolution which 12 populations of began from the recorded in the genomes of living organisms and, as micro- same ancestral strain and have evolved in identical, describes his own biologists, we see on-going evolution in the emergence of defined environments for more than 20,000 bacterial research into E. coli microbes resistant to antimicrobial agents. Until recently, generations (see Fig. 1). My two main objectives were to populations and few viewed evolution as an experimental science. This examine the reproducibility of evolution and to explore puts the other is now changing as microbes are increasingly used in the coupling between phenotypic and genomic changes. articles into context. designed experiments to test various hypotheses about In short, all the populations have become much more fit evolutionary dynamics, patterns and mechanisms. in the glucose-limited environment in which they For decades, some microbes served as model organisms evolved; at the end of 20,000 generations they grow about in genetics and molecular biology. The same advantages – 75 % faster than the ancestor when they compete head-to- ease of culture, rapid generations and large populations – head for glucose. There is much work still to be done on that served those fields also make micro-organisms ideally the genetic front, but one exciting result has been that Further reading suited to . The advances in those global gene-expression profiles show strikingly similar Cooper, T.F., Rozen, D.E. fields, from technical approaches to whole-genome sequ- evolution across replicate populations, yet sometimes & Lenski, R.E. (2003). ences, provide a powerful toolkit and a wealth of knowl- these parallel changes involved in different Parallel changes in gene edge for analysing and interpreting results of evolution genes. I have reviewed our findings elsewhere. expression after 20,000 experiments with microbes. Many evolutionary questions In this issue, Paul Rainey reviews his experiments generations of evolution in are being addressed: the dynamics of by showing the rapid diversification of Pseudomonas fluorescens E. coli. Proc Natl Acad Sci , the genetic changes underpinning from a single genotype into several lineages adapted to U S A 100, 1072–1077. that adaptation, tradeoffs between different aspects of different ecological niches that stably coexist even within Elena, S.F. & Lenski, R.E. performance, the specificity of adaptation with respect to a simple microcosm. He concludes that all microbiol- (2003). Evolution environmental variables, the causes and consequences of ogists should realize that any experiment involving experiments with micro- hypermutability, the effects of population size, the bacterial growth opens the door to evolutionary change. organisms: the dynamics and maintenance of , conflict and cooperation Of course, are not the only microbes that evolve: genetic bases of adaptation. Nature Rev Gen 4, 457–469. in social interactions, effects of sexual versus asexual Peter Simmonds describes the tremendous speed with reproduction and co-evolution of hosts and parasites. which viruses, especially RNA viruses, can evolve and Lenski, R.E. (2004). Evolution experiments with microbes are often quite adapt to different host environments. Frank Odds Phenotypic and genomic simple, at least in concept. Populations start from an examines the troubling problem of antimicrobial resist- evolution during a 20,000- generation experiment with ancestral strain and are propagated in defined environ- ance in Candida, an opportunistic fungal . We the bacterium Escherichia coli. ments for many generations. The ancestor is frozen away, can take some comfort from evidence that fungal mutants Breed Rev 24, 225–265. as are samples taken from the evolving populations at suffer a physiological cost of evolved resistance, which various generations. Later, ancestral and derived types can reduces their fitness in the absence of antifungal be revived and compared to determine what phenotypic compounds and thereby slows their spread. changes occurred and to identify the genetic bases of those Peter Williams describes rapid plasmid-mediated BELOW: changes. One can even perform competitions between the evolution of bacteria, which allows cells to degrade syn- Fig. 1. Twelve E. coli ancestral and descendant types to measure the net change thetic compounds that were never encountered in their populations, evolving in and in Darwinian fitness that occurred during an experiment. prior evolutionary history. This article reminds us that adapting to identical environments, provide a test of the repeatability (In terms, it is as though we could bring back many bacteria are friends, not foes, and evolution may of evolutionary dynamics and fossil hominids – not just their bones or bits of DNA, but sometimes work to our advantage in helping solve outcomes. the living beings – and challenge them to some environmental problems. Finally, Lynn Margulis des- COURTESY R.E. LENSKI competition – say, football or chess – to test whether cribes major evolutionary transitions in the history of and how much we have that emerged from endosymbiotic associations between improved on our distant micro-organisms. Remarkably, many of the hypothesized ancestors.) A genetic intermediate stages in these relationships can still be marker is often introduced found among living microbes and subjected to investiga- to distinguish more readily tion. Her perspective reminds those of us fascinated by the ancestral and evolved studying evolution in action that our experiments are a competitors. It is import- mere drop in the deep ocean of evolutionary time. ant to realize that the rela- tive fitness of any two types Professor Richard E. Lenski, Department of depends on the environ- & Molecular Genetics, Michigan ment, so the finding that State University, East Lansing, MI 48824, USA. a population has become email [email protected]

158 MICROBIOLOGYTODAY VOL 31/NOV04