1. Evidence for Positive Epistasis in HIV-1 Sebastian Bonhoeffer, Colombe Chappe, Neil T. Parkin, Jeanette M. Whitcomb, Christos J. Petropoulos

2. Evolutionary basis for sexual reproduction and recombination Population theories based on effects of fitness interactions between alternate allelles at different loci to control a single phenotype (epistasis) In a two-locus, two-allele model, it is defined as: E = w ab + w AB – w aB – w Ab where a/A and b/B are the alternative alleles at the two loci and w ** is the log fitness of the corresponding genotype Measures deviations from multiplicativity of the fitness effect caused by individual mutations

3. Positive and negative epistasis Michalakis and Roze, Science 306: 1492-1493, 2004 Synergistic beneficial mutations; antagonistic deleterious mutations Antagonistic beneficial mutations; synergistic deleterious mutations

4. Evolutionary basis for sexual reproduction and recombination In negative epistasis, beneficial mutations interact antagonistically (increase fitness less than multiplicatively) and detrimental mutations act synergistically (decrease fitness more than multiplicatively) Increases efficiency in selection Appeal of negative epistasis is that recombination can efficiently eliminate deleterious mutations, thus providing a strong selective advantage No clear evidence for an overall predominance of positive or negative epistasis due to inherent difficulties

5. Why use HIV-1 as a model organism to study epistasis Reproductive cycle is a primitive form of sexual reproduction Retroviruses pack two full-length copies of the RNA genome After infection, RT engages one copy and converts it to DNA RT carrying nascent DNA may disengage from the first template and switch over to the next If the RNA genomes are heterozygous, this may lead to a recombinant proviral DNA Common occurrence when host cells are infected by distinct proviruses RT switches RNA templates approximately 10 times per replication (about 1 recombination event for 1000 NBP)

6. HIV-1 as a model organism to study epistasis 9466 virus samples from HIV-1 patients Only PR and most of RT inserted into a HIV-1 clone that can undergo only one round of replication Fitness assay quantifies the total production of infectious progeny virus, and measured in absence of drug, relative to ability of the clone

7. HIV-1 as a model organism to study epistasis Plot log fitness as a function of the number of mutations Less than linear decrease suggests positive epistasis Could be due to bias in data set against sequences with low fitness

8. HIV-1 as a model organism to study epistasis Measure fitness interactions between pairs of alternative amino acids (“alleles”) at different positions (“loci”) where all four possible combinations are observed (103,826 pairs) Average log fitness values used to calculate E Mean of distribution is 0.052; different from random; mean for those that significantly affect fitness is 0.109

9. What’s going on here Most isolates are lower in fitness than the clone, suggesting antagonism between deleterious mutations (positive epistasis) Not consistent with current genetic theories of sexual reproduction and recombination May not be extrapolatable to higher eukaryotes Might be a mechanism to repair single stranded breaks in the RNA genome that is susceptible to degradation

10. What’s going on here – a structural perspective Analysis done on two enzymes; positive epistasis makes sense for isolates that have high fitness from a protein structure stability perspective Complementary beneficial mutations that increase stability more than multiplicatively, like two near oppositely charged residues on the surface replacing polar ones Strong evidence for positive epistasis of beneficial mutations:

11. What’s going on here – a structural perspective Positive epistasis makes less sense for isolates that have a low fitness from a protein structure stability perspective, but still possible Deleterious mutations that decrease stability due to environment compensate each other Strong evidence for positive epistasis of deleterious mutations: