John Maynard Smith Prize

Chro­mo­somal inver­sions have been repor­ted in many organ­isms, ran­ging from bac­teria to prim­ates,  and are increas­ingly found to be involved in spe­ci­ation, adap­tions to the envir­on­ment, many genet­ic dis­orders and agri­cul­tur­ally import­ant plant traits. Des­pite this appar­ent first-class bio­lo­gic­al importance,why and how inver­sions evolve remain poorly under­stood. In par­tic­u­lar, why inver­sions are often found at inter­me­di­ate fre­quency and asso­ci­ated with com­plex traits is a well-estab­lished puzzle in evol­u­tion­ary bio­logy. My cur­rent and future research aims to shed new light into this enigma that has major con­sequences in biology.

Dur­ing my PhD with Math­ieu Joron at the Uni­ver­sity of Mont­pel­li­er I used a com­bin­a­tion of gen­om­ics and mod­el­ing to study the evol­u­tion of supergenes, not­ably focus­ing on an intriguing but­ter­fly wing-pat­tern poly­morph­ism. Supergenes are tight clusters of loci con­trolling strik­ing poly­morph­isms, such as mor­pho­logy, beha­vi­or and sexu­al poly­morph­isms in numer­ous taxa. My work high­lighted the role played by chro­mo­somal inver­sions in the form­a­tion of such archi­tec­tures and we have espe­cially shown that com­plex poly­morph­isms could evolve simply because inver­sions are intrins­ic­ally prone to car­ry­ing recess­ive dele­ter­i­ous muta­tions. Dur­ing my postdoc with Tatiana Giraud at the Uni­ver­sity Par­is-Saclay, I stud­ied the con­sequences of the muta­tion load of inver­sions in the form­a­tion of non-recom­bin­ing sex chro­mo­somes and related archi­tec­tures, such as mat­ing-type chro­mo­somes in fungi. We not­ably shown that recom­bin­a­tion sup­pres­sion may evolve at sex chro­mo­somes simply because it provides the fit­ness advant­age of shel­ter­ing dele­ter­i­ous muta­tions. In brief, we sug­ges­ted that dele­ter­i­ous muta­tions could be a cause, and not only a con­sequence, of recom­bin­a­tion sup­pres­sion at sex chro­mo­somes and oth­er supergenes.

My cur­rent research, as a Human Fron­ti­er Sci­ence Pro­gram postdoc fel­low with Rasmus Nielsen at the Uni­ver­sit­ies of Copen­ha­gen and Berke­ley, aims to dis­sect the causes and con­sequences of the evol­u­tion of chro­mo­somal inver­sions through the study of thou­sands of present-day and ancient human gen­omes. I believe this could provide us long-awaited insights into gen­ome evol­u­tion, which may have import­ant reper­cus­sions in evol­u­tion­ary bio­logy and beyond.

Under­stand­ing the genet­ic basis of organis­mal adapt­a­tion remains a cent­ral goal of evol­u­tion­ary bio­logy. Struc­tur­al gen­om­ic vari­ants are muta­tions that influ­ence hun­dreds to mil­lions of DNA base-pairs and have the poten­tial to play a sig­ni­fic­ant role in adapt­a­tion. While struc­tur­al vari­ants have tra­di­tion­ally been chal­len­ging to detect with DNA sequen­cing, recent advance­ments in long-read DNA sequen­cing tech­no­lo­gies now facil­it­ate the molecu­lar detec­tion of struc­tur­al gen­om­ic changes, provid­ing excit­ing oppor­tun­it­ies to invest­ig­ate the effects of struc­tur­al vari­ants on organis­mal adaptation.

My research addresses the role of struc­tur­al gen­om­ic vari­ation in mam­mali­an adapt­a­tion and gen­ome evol­u­tion. Spe­cific­ally, I study chro­mo­somal inver­sions, a form of struc­tur­al gen­om­ic muta­tion in which an entire set of genes reverses ori­ent­a­tion along a chro­mo­some. Dur­ing my PhD in Dr. Hopi Hoekstra’s labor­at­ory at Har­vard Uni­ver­sity, I have invest­ig­ated chro­mo­somal inver­sions in deer mice, a mam­mali­an mod­el sys­tem for evol­u­tion­ary gen­om­ics. We found that deer mice har­bor many large, poly­morph­ic inver­sions which shape pat­terns of genet­ic diversity and recom­bin­a­tion across the gen­ome. Fur­ther, we dis­covered that inver­sions are asso­ci­ated with adapt­ive traits (such as tail length and coat col­or) in deer mice and facil­it­ated the deer mouse’s adapt­a­tion to dif­fer­ent loc­al hab­it­ats. Togeth­er, these res­ults sug­gest that chro­mo­somal inver­sions are an import­ant form of gen­om­ic muta­tion under­ly­ing mam­mali­an adaptation.

Evol­u­tion­ary and eco­lo­gic­al pro­cesses influ­ence one anoth­er. This real­iz­a­tion has led the dis­cip­lines of eco­logy and evol­u­tion­ary bio­logy to become increas­ingly integ­rated. Yet, some prob­lems in each dis­cip­line are still viewed through a mono-dis­cip­lin­ary lens. This was until recently the case of eco­lo­gic­al resi­li­ence the­ory. I will show how the intro­duc­tion of evol­u­tion­ary dynam­ics in the eco­lo­gic­al resi­li­ence the­ory can fun­da­ment­ally reshape our know­ledge about resi­li­ence and tip­ping points in eco­sys­tems. For example, it is tra­di­tion­ally
estab­lished that an eco­sys­tem tips to an altern­at­ive stable state when an envir­on­ment­al threshold is exceeded. On the con­trary, we have shown that eco-evol­u­tion­ary feed­backs can cause an eco­sys­tem to tip without exceed­ing an envir­on­ment­al threshold. I will con­clude high­light­ing the tight link between evol­u­tion­ary mech­an­isms, and pop­u­la­tion and com­munity process.

The human body har­bours tril­lions of bac­teri­al cells that live in diverse com­munit­ies. Much of what is known about the com­pos­i­tion and dynam­ics of these com­munit­ies comes from highthrough­put meta­ge­n­om­ic stud­ies which have yiel­ded a large invent­ory of bac­teri­al spe­cies and genes that are cor­rel­ated with health and dis­ease. Although these cata­logs under­score the rel­ev­ance of the micro­bi­o­me for human health, they offer lim­ited insights on how micro­bi­al com­munit­ies in the human body assemble, what determ­ines their func­tion­al prop­er­ties and how they evolve. My long-term goal is to build a mech­an­ist­ic under­stand­ing of the human micro­bi­o­me by estab­lish­ing how broad­er eco­lo­gic­al and evol­u­tion­ary pat­terns are linked to the under­ly­ing physiology and inter­ac­tions of indi­vidu­al microbes. In this talk, I will present two examples of the insights that can be gained by this com­pre­hens­ive approach. First, I will show
that in bac­teri­al pop­u­la­tions made of single spe­cies and even of single gen­o­types, vari­ation in how indi­vidu­als sense and respond to their envir­on­ment is per­vas­ive, and under­lies the evol­u­tion of phe­nom­ena like anti­bi­ot­ic tol­er­ance and res­ist­ance. Second, I will shift the focus to
multis­pe­cies com­munit­ies, and show how bac­teri­al physiology shapes the evol­u­tion of private and pub­lic resource util­iz­a­tion strategies in the gut micro­bi­o­me, and how this pro­cess can be mod­u­lated by the human host.

Cam­i­lo Barbosa, Andrew F. Read, Robert J. Woods

Evol­u­tion is the root of the anti­bi­ot­ic res­ist­ance crisis. Cla­ri­fy­ing the evol­u­tion­ary pro­cesses lead­ing to res­ist­ance, in par­tic­u­lar the determ­in­ants of chance and repeat­ab­il­ity, are thus pivotal to the goal of under­stand­ing how patho­gens adapt to drugs. Two import­ant open ques­tions are how repro­du­cible anti­bi­ot­ic res­ist­ance evol­u­tion is with­in human hosts and to what extent those evol­u­tion­ary paths seen in vivo can be recapit­u­lated in vitro. To address these two ques­tions, we ret­ro­spect­ively identi­fy changes in res­ist­ance against dap­to­my­cin and
linezol­id when these drugs are used as the main treat­ment in patients with blood stream infec­tions with Entero­coc­cus fae­ci­um. E. fae­ci­um isol­ates from blood cul­tures of hos­pit­al­ized patients are routinely stored in our lab. Twelve patients were iden­ti­fied with in vivo res­ist­ance evol­u­tion of E. fae­ci­um to dap­to­my­cin and 6 to linezol­id, each obtained from inde­pend­ent patients. We fully sequenced and assembled the gen­omes of 18 ini­tially sens­it­ive isol­ates. We then per­formed whole gen­ome sequen­cing of the sub­sequent isol­ates show­ing an increase in res­ist­ance against the cor­res­pond­ing drug with­in the same patient to identi­fy the repeat­ab­il­ity of gen­om­ic changes asso­ci­ated with res­ist­ance. Addi­tion­ally, each of the 18 ini­tially sens­it­ive isol­ates was used to found 20 inde­pend­ent bio­lo­gic­al rep­lic­ates from in vitro evol­u­tion against increas­ing con­cen­tra­tions of the cor­res­pond­ing drug to determ­ine wheth­er the same mech­an­isms iden­ti­fied in vivo emerge in vitro. This study cast light on the role of determ­in­ism and con­tin­gency in evol­u­tion, with implic­a­tions for med­ic­al treatment.

Explain­ing the main­ten­ance of genet­ic vari­ance in fit­ness remains one of the most long­stand­ing chal­lenges for evol­u­tion­ary bio­logy. Muta­tion-selec­tion bal­ance can­not account for all of the genet­ic vari­ance observed in nature, mean­ing that some form(s) of bal­an­cing selec­tion must com­monly ensue. Sexu­ally ant­ag­on­ist­ic (SA) selec­tion can gen­er­ate bal­an­cing selec­tion and main­tain poly­morph­isms for fit­ness through­out the gen­ome, wherein altern­at­ive alleles pose oppos­ite fit­ness effects in the sexes. Des­pite grow­ing evid­ence for SA genet­ic vari­ation, I argue that its role in main­tain­ing fit­ness vari­ance may still be under­es­tim­ated. I will review the rel­ev­ant his­tor­ic­al con­text, focus­ing on the evid­ence and per­spect­ives sur­round­ing the role of sex-spe­cif­ic dom­in­ance reversal (SSDR) – where the allele that bene­fits a giv­en sex is also dom­in­ant in that sex. I will then present evid­ence of SSDR under­ly­ing SA poly­morph­isms for fit­ness through­out the gen­ome of the seed beetle Cal­lo­sobruchus mac­u­lat­us, dis­cuss poten­tial molecu­lar mech­an­isms of SSDR, and provide examples of the phen­o­types that medi­ate SA genet­ic vari­ation for fit­ness. Lastly, I will describe an ongo­ing effort to identi­fy SA poly­morph­isms (and their dom­in­ance prop­er­ties) in Dro­so­phila melano­gaster, and dis­cuss why SA poly­morph­isms have proven so elusive.

Microbes are embed­ded with­in diverse com­munit­ies com­pris­ing highly com­plex net­works of social inter­ac­tions. Con­sequently, pre­dict­ing the evol­u­tion and eco­logy of a focal spe­cies in a single-spe­cies in vitro envir­on­ment may not be indic­at­ive of what is hap­pen­ing in nature. This conun­drum is dif­fi­cult to address – because 1) com­munit­ies are highly com­plex and inter­ac­tions are dif­fi­cult to decipher in situ, 2) obser­va­tions from nature are gen­er­ally cor­rel­at­ive and determ­in­ing caus­al­ity requires exper­i­ments, 3) it is dif­fi­cult to accur­ately recre­ate nat­ur­al con­di­tions in the lab and 4) many microbes are uncul­tur­able under stand­ard labor­at­ory con­di­tions. My research aims to over­come these dif­fi­culties, by attempt­ing to bridge the gap between single spe­cies in vitro exper­i­ments and data col­lec­ted from nat­ur­al pop­u­la­tions. I will dis­cuss how this approach has provided nov­el insights into our under­stand­ing of micro­bi­al com­munit­ies, in both envir­on­ment­al (heavy met­al con­tam­in­a­tion) and clin­ic­al (cyst­ic fibrosis lung infec­tions) contexts.

Over forty years ago, John Maynard Smith inspired one of the out­stand­ing prob­lems in evol­u­tion­ary bio­logy: the main­ten­ance of sexu­al repro­duc­tion. First, I’ll show that Maynard Smith’s simple mod­el, the two-fold cost of males, holds in a nat­ur­al sys­tem. I com­bined the­ory and exper­i­ment­al data to dir­ectly quanti­fy the cost of sex in the fresh­wa­ter snail Pot­amopyr­gus anti­podar­um. Con­sist­ent with Maynard Smith’s pre­dic­tion, the per-cap­ita birth rate of asexu­al lin­eages is at least twice that of sexu­al lin­eages. So, in Maynard Smith’s terms: what use is sex? Second, I’ll present tests of the Red Queen hypo­thes­is, which pro­poses that host-para­site coe­volu­tion main­tains sex. Obser­va­tions of a nat­ur­al pop­u­la­tion, paired with exper­i­ment­al manip­u­la­tions, show that coe­volu­tion can explain fine-scale spa­tial and tem­por­al vari­ation in the fre­quency of sexu­al snails. Field data span­ning a fif­teen-year peri­od reveal a dynam­ic coe­volu­tion­ary pro­cess, with para­sites switch­ing to select against sexu­al repro­duc­tion as asexu­al lin­eages become rare. Taken togeth­er, these res­ults sup­port coe­volving para­sites in main­tain­ing coex­ist­ence of repro­duct­ive modes.

One com­pon­ent of sex-alloc­a­tion the­ory pos­its that sons and daugh­ters are dif­fer­en­tially affected by early rear­ing con­di­tions, whereby the amount of par­ent­al care received has dif­fer­ing effects on the fit­ness of males and females. When this occurs, selec­tion is expec­ted to favor off­spring sex-ratio adjust­ment accord­ing to anti­cip­ated fit­ness returns. Here, I describe a series of ques­tions related to sex-by-envir­on­ment effects on the devel­op­ment, sur­viv­al, and future repro­duc­tion of off­spring and asso­ci­ated vari­ation in primary off­spring sex ratios. What emerges is a pat­tern of con­sist­ent, and per­sist­ent, sex-spe­cif­ic effects of nat­al envir­on­ment­al con­di­tions on off­spring, thus favor­ing the adjust­ment of off­spring sex ratios by moth­ers. How­ever, increased sens­it­iv­ity of males to envir­on­ment­al con­di­tions should also con­trib­ute to shap­ing an optim­al off­spring sex ratio, with implic­a­tions for the evol­u­tion of sex-ratio adjustment.

Math­em­at­ic­al mod­el­ling has always played an import­ant role in elu­cid­at­ing evol­u­tion­ary phe­nom­ena. Here, I present an over­view of vari­ous mod­els I’ve worked on dur­ing my career. First, it has been pos­tu­lated that sexu­al repro­duc­tion is bene­fi­cial by recom­bin­ing gen­omes, thus recre­at­ing optim­al gen­o­types. This hypo­thes­is has faced renewed interest due to the abil­ity to test hypo­theses using next gen­er­a­tion sequence data. I will show how strong selec­tion for recom­bin­a­tion, that can poten­tially main­tain costly sex, appears if act­ing over hun­dreds of sites sub­ject to selec­tion. Major bene­fits to sex and recom­bin­a­tion arise through dis­en­tangling bene­fi­cial muta­tions from dele­ter­i­ous back­grounds. Finally, I will also dis­cuss my research into dis­ease emer­gence. Spe­cific­ally, I invest­ig­ate how the spread of exist­ing strains hampers the abil­ity of mutated patho­gens to emerge, by lim­it­ing the avail­able pool of sus­cept­ible individuals.

We recently com­pleted three link­age dis­equi­lib­ri­um (LD) based recom­bin­a­tion maps gen­er­ated using whole gen­ome sequen­cing of 10 Nigeri­an chim­pan­zees, 13 bonobos, and 15 west­ern gor­il­las, col­lec­ted as part of the Great Ape Gen­ome Pro­ject (Pra­do Mar­tinez et al. 2013). We also iden­ti­fied spe­cies spe­cif­ic recom­bin­a­tion hot­spots in each group using a mod­i­fied LDhot frame­work, which greatly improves stat­ist­ic­al power to detect hot­spots at vary­ing strengths. Using spe­cies spe­cif­ic PRDM9 sequences to pre­dict poten­tial bind­ing sites in hot­spot regions as com­pared to match cold spot regions, we iden­ti­fied an import­ant role for PRDM9 in pre­dict­ing recom­bin­a­tion rate vari­ation in mul­tiple great ape spe­cies. While pre­vi­ous research showed that PRDM9 is not asso­ci­ated with recom­bin­a­tion in west­ern chim­pan­zees (Auton et al. 2012), we attrib­ute this lack of sig­nal to high­er pop­u­la­tion level diversity at the PRDM9 locus in this group. Addi­tion­ally, we show that few­er hot­spots are shared among chim­pan­zee sub­spe­cies than with­in human pop­u­la­tions, fur­ther nar­row­ing the time scale of com­plete hot­spot turnover. We quan­ti­fied the vari­ation in the biased dis­tri­bu­tion of recom­bin­a­tion rates towards recom­bin­a­tion hot­spots across great apes, high­light­ing sim­il­ar dis­tri­bu­tions across great apes with Europeans as an out­lier. Fur­ther, we found that pair­wise com­par­is­ons of broad scale recom­bin­a­tion rates decay more rap­idly than pair­wise nuc­le­otide diver­gence between spe­cies. We also com­pared the skew of recom­bin­a­tion rates at centromeres and telomeres between spe­cies and show a skew from chro­mo­some means extend­ing as far as 10–15 Mb from chro­mo­some ends. Our study is the first to ana­lyze with­in and between spe­cies gen­ome wide recom­bin­a­tion rate vari­ation in sev­er­al close relatives.

Spe­cies selec­tion – her­it­able trait-depend­ent dif­fer­ences in rates of spe­ci­ation or extinc­tion – may be respons­ible for vari­ation in both taxo­nom­ic and trait diversity among clades. While ini­tially con­tro­ver­sial, interest in spe­cies selec­tion has been revived by the accu­mu­la­tion of evid­ence of wide­spread trait-depend­ent diver­si­fic­a­tion. I will present sev­er­al meth­ods for invest­ig­at­ing spe­cies selec­tion by detect­ing the asso­ci­ation between spe­cies traits and spe­ci­ation or extinc­tion rates. These meth­ods are expli­citly phylo­gen­et­ic and incor­por­ate simple, but com­monly used, mod­els of spe­ci­ation, extinc­tion, and trait evol­u­tion. Using these meth­ods, I will present sev­er­al examples where traits are cor­rel­ated with spe­ci­ation or extinc­tion rates in plants and mam­mals. All meth­ods have assump­tions and lim­it­a­tions, and I will dis­cuss the pit­falls that arise when apply­ing these meth­ods (and the widely used meth­ods that they derive from) to messy bio­lo­gic­al data. Com­par­at­ive phylo­gen­et­ic meth­ods must be used with cau­tion, but allow test­ing of long-stand­ing hypo­theses about causes of vari­ation in bio­lo­gic­al diversity.

Phylo­gen­et­ic trees of present-day spe­cies allow infer­ence of the rate of spe­ci­ation and extinc­tion which led to the present-day diversity. Clas­sic­ally, infer­ence meth­ods assume a con­stant rate of diver­si­fic­a­tion, or neg­lect extinc­tion. I will present a new meth­od­o­logy which allows spe­ci­ation and extinc­tion rates to change through time (envir­on­ment­al-depend­ent diver­si­fic­a­tion) as well as with the num­ber of spe­cies (dens­ity-depend­ent diver­si­fic­a­tion). Par­tic­u­lar atten­tion is paid towards the spe­cif­ic spe­cies sampling schemes for incom­plete phylo­genies.
Using this new frame­work, I show that mam­mali­an diver­si­fic­a­tion rates are mainly determ­ined by envir­on­ment­al effects; how­ever, I reject the hypo­thes­is of accel­er­ated mam­mali­an evol­u­tion fol­low­ing the extinc­tion of dino­saurs at the KT-bound­ary. The oth­er two con­sidered data­sets, birds and ants, reveal dens­ity-depend­ence as the main factor determ­in­ing diver­si­fic­a­tion rates. In con­trast to pre­vi­ous res­ults, the new ana­lyses pre­dict high extinc­tion rates for birds, as well as no major envir­on­ment­al impact on diver­si­fic­a­tion for ants.
The meth­ods can eas­ily be applied to oth­er data­sets using the R pack­ages TreeP­ar and TreeS­im avail­able on CRAN.

Sum­mary state­ment: The genet­ics of adapt­a­tion: com­bin­ing the­ory, lab and field stud­ies to under­stand the mech­an­isms that drive eco­lo­gic­al and evol­u­tion­ary responses to chan­ging environments.

Human activ­it­ies are res­ult­ing in extens­ive world­wide changes to eco­sys­tems, with both eco­lo­gic­al and evol­u­tion­ary con­sequences. Under­stand­ing the pro­cess of adapt­a­tion to chan­ging envir­on­ments requires integ­rat­ive stud­ies that com­bine approaches from pop­u­la­tion genet­ics, evol­u­tion­ary eco­logy and molecu­lar bio­logy. Here, I present the­or­et­ic­al, labor­at­ory and field stud­ies with microbes and fish that help to determ­ine the genet­ic basis, eco­lo­gic­al mech­an­isms, and evol­u­tion­ary effects of rap­id adapt­a­tion to chan­ging envir­on­ments. The work involves dir­ect meas­ures of nat­ur­al selec­tion act­ing at the molecu­lar level, thus provid­ing cru­cial inform­a­tion on the func­tion­al links among gen­o­type, phen­o­type, and fit­ness. This research is help­ing to identi­fy some of the primary mech­an­isms that are likely to drive adapt­a­tion to glob­al envir­on­ment­al change.

Under­stand­ing how a single gen­ome can pro­duce a vari­ety of dif­fer­ent phen­o­types is of fun­da­ment­al import­ance in genet­ics and devel­op­ment­al bio­logy. One of the most strik­ing examples of phen­o­typ­ic plas­ti­city is the female caste sys­tem found in ants and oth­er euso­cial insects, where dif­fer­ent phen­o­types are asso­ci­ated with repro­duc­tion (queen caste) or help­ing beha­viour (work­er castes). A long-stand­ing paradigm for caste determ­in­a­tion was that female eggs are always toti­po­tent with the import­ant mor­pho­lo­gic­al and physiolo­gic­al dif­fer­ences between queens and work­ers stem­ming solely from a devel­op­ment­al switch dur­ing the lar­val stage under the con­trol of nutri­tion­al and oth­er envir­on­ment­al factors. How­ever, there are an increas­ing num­ber of examples show­ing genet­ic com­pon­ents to caste determ­in­a­tion as well as mater­nal effects influ­en­cing the devel­op­ment­al fate of females. I will present a broad over­view of the stud­ies provid­ing strong dir­ect and indir­ect evid­ence for a genet­ic com­pon­ent to caste dif­fer­en­ti­ation and dis­cuss factors that may have led to the evol­u­tion of genet­ic­ally hard­wired caste sys­tems. In addi­tion, I will argue that a purely envir­on­ment- con­trolled caste sys­tem is very dif­fi­cult to demon­strate and prob­ably unlikely to occur in genet­ic­ally het­ero­gen­eous soci­et­ies. Detailed molecu­lar ana­lyses and breed­ing exper­i­ments are likely to uncov­er addi­tion­al cases of genet­ic­ally-determ­ined queen and work­er determ­in­a­tion and vari­ous degrees of genet­ic pre­dis­pos­i­tion towards a par­tic­u­lar caste.

Spite, altru­is­m’s neg­lected ugly sis­ter, is the most mys­ter­i­ous and con­tro­ver­sial of the four social beha­viours. How can an indi­vidu­al be favoured to harm itself and its social part­ners? Hamilton’s rule, which was devised in order to explain altru­ist­ic beha­viours, has a dark­er side that reveals when spite will be favoured. Spe­cific­ally, it requires that the spite­ful act­or and its vic­tim be neg­at­ively related. I devel­op the­ory for the evol­u­tion of spite in com­pet­it­ive envir­on­ments, and show that increas­ingly strong loc­al com­pet­i­tion can favour spite­ful beha­viour. Applic­a­tion of the the­ory to chem­ic­al war­fare in microbes and sui­cid­al sib­ling rivalry in para­sit­oid wasps leads to nov­el pre­dic­tions for para­site vir­ulence and sex alloc­a­tion the­ory. I dis­cuss the semantics of spite and ambi­gu­ities in the stand­ard clas­si­fic­a­tion of social behaviours.

Spe­ci­ation occurs through the evol­u­tion of any of sev­er­al forms of repro­duct­ive isol­a­tion, includ­ing the intrins­ic ster­il­ity or invi­ab­il­ity of hybrids. These hybrid fit­ness prob­lems are caused by neg­at­ive epi­stat­ic inter­ac­tions – new alleles that evolve in one spe­cies are some­times incom­pat­ible with alleles at inter­act­ing loci from related spe­cies. Rel­at­ively little is known about the iden­tity and func­tion of such “spe­ci­ation genes” or about the evol­u­tion­ary forces driv­ing their diver­gence. I will present res­ults from a large, fine-scale genet­ic ana­lys­is of loci caus­ing hybrid invi­ab­il­ity in Dro­so­phila. The first of these genes to be iden­ti­fied encodes Nup96, an essen­tial pro­tein com­pon­ent of the nuc­le­ar pore com­plex (NPC). I will show that the func­tion­al diver­gence of Nup96, and sev­er­al oth­er inter­act­ing Nup pro­teins, was driv­en by adapt­ive evol­u­tion. I will then dis­cuss how this pos­it­ive selec­tion may be a con­sequence of genet­ic con­flict medi­ated by the NPC.

Her­it­able phen­o­typ­ic vari­ation is the “raw mater­i­al” of evol­u­tion by nat­ur­al selec­tion, and under­stand­ing the mech­an­isms that gen­er­ate such vari­ation has become a fun­da­ment­al chal­lenge for con­tem­por­ary evol­u­tion­ary bio­logy. In recent years, evol­u­tion­ary devel­op­ment­al bio­logy has encour­aged a change of focus from the sort­ing of phen­o­typ­ic vari­ation by selec­tion to the pro­duc­tion of that vari­ation through devel­op­ment. The col­our pat­terns dec­or­at­ing but­ter­fly wings provide ideal mater­i­al to study the recip­roc­al inter­ac­tions between evol­u­tion and devel­op­ment in this pro­cess. They are visu­ally com­pel­ling products of selec­tion, often with a clear adapt­ive value, and are also amen­able to a detailed devel­op­ment­al char­ac­ter­iz­a­tion at dif­fer­ent levels. We stud­ied dif­fer­ent aspects of the pro­cess of gen­er­a­tion of vari­ants in Bicyc­lus any­n­ana eye­s­pot pat­terns. Res­ults will be dis­cussed of exper­i­ments where we have used arti­fi­cial selec­tion to explore the poten­tial for changes in eye­s­pot size phen­o­types, which were thought to be con­strained by the prop­er­ties of but­ter­fly wing pat­tern devel­op­ment. We also report on exper­i­ments aimed at identi­fy­ing the actu­al genes involved in the response to selec­tion. Our res­ults show that a com­bin­a­tion of approaches from evol­u­tion­ary and devel­op­ment­al bio­logy used to study the pat­terns of col­our on but­ter­fly wings can greatly con­trib­ute to under­stand­ing how evol­u­tion­ar­ily rel­ev­ant vari­ation is generated.

In the sum­mer of 1939, a group of 40–50 house finches Car­po­da­cus mex­ic­anus col­lec­ted in south­ern Cali­for­nia was released from a pet store in New York City. In the sub­sequent 62 years, this intro­duced pop­u­la­tion has under­gone tre­mend­ous expan­sion, spread­ing across the east­ern U.S. and south-east­ern Canada and increas­ing to an estim­ated 1.3 bil­lion birds. This expan­sion of eco­lo­gic­al range was accom­pan­ied by rap­id diver­gence in sexu­al size dimorph­ism among new pop­u­la­tions. We show that the observed diver­gence in mor­pho­logy were caused by pop­u­la­tion dif­fer­ences in pat­terns of nat­ur­al selec­tion act­ing over the lifespan of both sexes. This rep­res­ents an appar­ent para­dox of rap­id inde­pend­ent evol­u­tion of each sex in traits for which there is no sex-biased genet­ic vari­ance in adults. We show that cor­rel­ated selec­tion on growth tra­ject­or­ies of males and females in com­bin­a­tion with per­sist­ent and strongly sex-biased mater­nal effects can account for the observed adapt­ive diver­gence in sexu­al dimorph­ism among newly-estab­lished pop­u­la­tions of the house finch.

Base com­pos­i­tions (A‑, C‑, G- and T‑percent) are highly vari­able among genes and gen­omes. Het­ero­gen­eous base com­pos­i­tions have been observed in most taxo­nom­ic groups sampled, for organ­el­lar or nuc­le­ar gen­omes, in cod­ing and non-cod­ing regions. Base com­pos­i­tion has some func­tion­al implic­a­tions: depend­ing on the organ­ism, it relates to codon usage, gene dens­ity, res­ist­ance to high tem­per­at­ure. Observing unequal base com­pos­i­tions between gen­omes or between homo­log­ous genes implies that dis­tinct lin­eages have under­gone dis­tinct evol­u­tion­ary pro­cesses. This raises sev­er­al inter­est­ing ques­tions. First, one may won­der about the robust­ness of DNA sequence ana­lys­is meth­ods – and espe­cially phylo­gen­et­ic infer­ence meth­ods when the base com­pos­i­tion var­ies between com­pared sequences. Secondly, the his­tory of diver­ging base com­pos­i­tions deserves atten­tion: what were the ances­tral states, which lin­eages exper­i­enced severe com­pos­i­tion­al changes? Finally, the mech­an­isms of com­pos­i­tion­al diver­gence are unknown in most cases: what are the evol­u­tion­ary forces that under­lie the observed changes in base com­pos­i­tion? Is nat­ur­al selec­tion act­ing to shape gen­om­ic base com­pos­i­tions, or is the vari­ation between gen­omes mainly due to vari­able muta­tion pro­cesses?
These prob­lem­at­ics instan­ti­ate the dual goal of molecu­lar evol­u­tion, namely (i) recov­er­ing the his­tory of spe­cies and pop­u­la­tions through that of their gen­omes, and (ii) under­stand­ing bet­ter the struc­ture and func­tion of gen­omes thanks to the evol­u­tion­ary per­spect­ive. The above ques­tions are addressed thanks to a non-homo­gen­eous, non-sta­tion­ary mod­el of DNA sequence evol­u­tion, allow­ing diver­ging GC-con­tent in time and between lin­eages (Gal­ti­er & Gouy 1995, 1998). Max­im­um-like­li­hood ana­lyses based on this mod­el allow to (i) cor­rectly estim­ate phylo­genies in case of vari­able GC-con­tent between sequences, and (ii) estim­ate ances­tral base com­pos­i­tions. The lat­ter pos­sib­il­ity is applied to ribosomal RNA sequences from spe­cies sampled in all three domains of life (Gal­ti­er et al. 1999), yield­ing evid­ence that the last uni­ver­sal com­mon ancest­or was not a ther­mo­phil­ic organism.

Gal­ti­er, N., and Gouy, M. 1995. Infer­ring phylo­genies from sequences of unequal base com­pos­i­tions. Proc. Natl. Acad. Sci. USA. 92: 11317–11321.
Gal­ti­er, N., and Gouy, M. 1998. Infer­ring pat­tern and pro­cess: max­im­um like­li­hood imple­ment­a­tion of a non-homo­gen­eous mod­el of DNA sequence evol­u­tion for phylo­gen­et­ic ana­lys­is. Mol. Biol. Evol. 15: 871–879
Gal­ti­er, N., Tour­as­se, N.J., and Gouy, M. 1999. A non-hyper­ther­mo­phil­ic ancest­or to extant life forms. Sci­ence 283: 220–221.

Para­sit­ism is the ances­tral state of most mutu­al­isms. What kinds of traits facil­it­ate the trans­ition from an ant­ag­on­ist­ic to a mutu­ally bene­fi­cial inter­ac­tion? The only well form­al­ised and tested scen­ario for the ori­gin of mutu­al­ism is based on the evol­u­tion of ver­tic­al trans­mis­sion of para­sites (from par­ents to off­spring Yamamura, 1996), which leads to reduced vir­ulence and some­times to the evol­u­tion of mutu­al­ism. How­ever, this scen­ario can apply only to sym­bi­ot­ic mutu­al­isms, and even these include examples in which ver­tic­al trans­mis­sion does not occur. For these, form­al­ised mod­els are lack­ing. What oth­er kinds of traits facil­it­ate the evol­u­tion of mutu­al­ism? Can we identi­fy traits, main­tained by selec­tion on oth­er func­tions that inde­pend­ently in dif­fer­ent lin­eages acquire the same nov­el func­tion in a par­tic­u­lar type of mutu­al­ism? If such “pre-adapt­a­tions” exist, what factors inter­vene to alter the selec­tion pres­sures act­ing on them and shape them as new adapt­a­tions? How recur­rent and pre­dict­able is the evol­u­tion of mutu­al­ism?
Small dif­fer­ences in traits already present at the ori­gin of the mutu­al­ism may lead to dif­fer­ences in how the mutu­al­ism func­tions and how it evolves. “Con­straint” is the flip side of “pre-adapt­a­tion.” While con­straints are usu­ally envis­aged to lim­it the range of evol­u­tion­ary pos­sib­il­it­ies, con­straints may also open evol­u­tion­ary path­ways that are oth­er­wise not pos­sible. For some mutu­al­isms, evol­u­tion­ary sta­bil­ity appears to be based on a co-evol­u­tion­ary equi­lib­ri­um between trait val­ues for the two mutu­al­ists. In oth­er cases, how­ever, the inter­ac­tion appears to be sta­bil­ised by con­straints imposed by pre-exist­ing traits of one spe­cies that the asso­ci­ated spe­cies can­not evolve to over­come. These points will be developed using as examples the fig/ fig wasp pol­lin­a­tion mutu­al­ism and pro­tect­ive ant/plant interactions.