Genomes and Evolutionary Biology Group Isabel Gordo Evolutionary

Genomes and Evolutionary Biology Group Isabel Gordo

Evolutionary change: -Rate of mutation -Strenght of natural selection -level of interaction between mutations - epistasis Understanding the evolution of any biological system depends upon understanding each of these mechanisms Estimates of these parameters in bacteria will have a profound impact on our understanding of their biology, their diversity, their rate of speciation and in our health. Experimental Evolution with bacteria allows us to directly measure these parameters and to test theoretical predictions about the genetic basis of adaptive evolution.

Mutation, selection and epistasis in Escherichia coli Is there a law governing its adaptation?

R. A. Fisher and a microscope A random mutation of large effect has much higher change of being deleterious than of being beneficial Most beneficial mutations have small effects on fitness

Study of adaptation in E. coli while it is occurring Genetically modify the bacteria to measure adaptive mutations as they get incorporated in the population time One adaptive event- selective sweep Imhof M, Schlotterer C. PNAS 2001

The rate of mutation to beneficial alleles can be as high as 2 x 10 -5 1 in 150 new mutations can be advantageous Ua=2 x 10 -5 per genome per generation E(Sa)=0. 01 Perfeito et al. Science 2007 Fig. 1. Distribution of fitness effects measured in the populations of Ne=2 x 104. The grey bars show the distribution of the measured beneficial mutations Compatible with Fisher: adaptive mutations of large effect are rarer

Epistasis and evolution of antibiotic resistance • Mutations confering antibiotic resistance have a benefit when the drug exists in the environments • Mutations confering antibiotic resistance have a cost in drug free environments If a pathogenic strain is resistant to antibiotic X, which antibiotic should be administered as a second treatment? Best combination is that which leads to the higher cost.

What is the cost of multiple antibiotic resistance? Cost of mutation c 1 X Resistant to antibiotic 1 Cost of mutation c 2 X Resistant to antibiotic 2 Cost of mutation 1 & 2 ? X X Resistant to both antibiotics If c 12 = c 1+ c 2 No epistasis, no interaction e=0 If c 12 > c 1+ c 2 Negative epistasis, high cost e<0 If c 12 < c 1+ c 2 Positive epistasis, low cost e>0

1) Select resistant clones: rps. L K 43 N X Sequencing Antibiotics used: LB with drug (i) nalidixic acid, which inhibits DNA replication by binding to DNA gyrase; (ii) rifampin, binds to the b-subunit of RNA polymerase thereby inhibiting transcription; (iii) streptomycin, binds to the ribosome and inhibits elongation of protein synthesis 2) Make all possible combinations of double resistant clones

Major finding 1 # combinations with positive epistasis > # with negative epistasis The cost of double resistance is lower than expected

Major finding 2: Resistance mutations to a new antibiotic can compensate the cost of resistance to another antibiotic

"What is true for E. coli is true for the elephant, " Jacques Monod E. coli Low cost for resistance mutations X and Y M. tuberculosis High frequency of multiple resistance involving X and Y

Conclusions: • We estimated that 1 in 150 mutations can be adaptive • This is the highest estimate ever obtained in a bacteria • The mean effect of each new beneficial mutations is about 1%. • The data supports the Fisherian hypothesis that most beneficial mutations have small effects and those that will fix follow a gamma distribution

Conclusions: • We estimated that 1 in 150 mutations can be adaptive • This is the highest estimate ever obtained in a bacteria • The mean effect of each new beneficial mutations is about 1%. • The data supports the Fisherian hypothesis that most beneficial mutations have small effects and those that will fix follow a gamma distribution • Double antibiotic resistance is less costly than we could a priori predict • Very difficult to eliminate resistant bacteria • Some resistance mutations which are deleterious when in a wildtype background are beneficial (compensatory) when in a genetic background that contains another resistance => Sign epistasis

Conclusions: Together these results are the worst nightmare for the host and the best dream for the microbe.

Thanks to: -Lisete Fernandes (IGC) -Ana Margarida Sousa(IGC) -Francisco Dionisio (FCUL) -Karina Xavier(IGC/ITQB) -Miguel Godinho Ferreira(IGC) Lilia Perfeito & Sandra Trindade Thank you all!!!

Evolutionary questions on mutation • Is there an optimal mutation rate? • What is the optimal mutation rate? R. A. Fisher 1930 “The genetical theory of natural selection”: optimal U must have an intermediate value

Drake’s Rule Genomic mutation rate in DNA microbes is ~ constant Drake’s U=0. 003 (Drake et al. 1998) Review by Sniegowski et al (2000)- Pink: RNA viruses (rv, rhinovirus; pv, poliovirus; vsv, vesicular stomatitis virus; mv, measles virus). Red: DNA phages M 13, T 2, T 4, and λ. E. coli (Ec) Saccharomyces cerevisiae (Sc) Neurospora crassa (Nc) Ce: C. elegans; Dm: Drosophila melanogaster; Mm, Mus musculus; Hs, Homo sapiens