nutshell: Assembling the Tree of Life Developing methods to analyze data and infer phylogenies Applying evolutionary models and methods to open problems The Tree of Lifedepicts the evolutionary relationships of Earth’s taxonomic diversity — including all living and extinct forms — over the past 3.5 billion years of its existence. The hierarchical arrangement of this phylogeny provides a comparative and predictive framework for all fundamental
different from the king brown, it is actually closely related and can be treated with the same antivenin. Conversely, the western brown looks very similar to the king brown, but it is only distantly related and thus responds best to different antivenin.” e poisonous snakes than any other many people die from snakebites each g effective antivenins is thus a high nown about the venins of most species. s helping with this task because venin rongly with evolutionary relationships. ed black snake looks very different from ctually closely related and can be treated n. Conversely, the western brown looks g brown, but it is only distantly related t to different antivenin. The phylogeny is ent demonstration that the poorly-known ed to the death adder (orange lineage) er is also highly dangerous and might lable death adder antivenin. etic knowledge to discover enins Black whip snake Talpan Fierce snake Common brown Western brown Dugte Collatt’s snake Spotted black Butler’s snake King brown Red-bellied black Death adder Barclick Small-eyed snake Australian copperhead Tiger snake Rough-scaled snake Broad-headed snake Recommended Antivenene Taipan Brownsnake Blacksnake Death adder Tiger snake Keogh et al. (2000), Journal of Zoology 252: 317–326.
Indo-European origin is con- gruent with the proposal that the family began to diverge with the spread of agriculture from Fig. 2. Map and maximum clade credibility tree showing the diversification of the major Indo-European subfamilies. The tree shows the timing of the emergence of the major branches and their subsequent diversification. The inferred location at the root of each subfamily is shown on the map, colored to match the corresponding branches on the tree. Albanian, Armenian, and Greek subfamilies are shown separately for clarity (inset). Contours represent the 95% (largest), 75%, and 50% HPD regions, based on kernel density estimates (15). likely than from land to land by a factor of 100 Landscape aware: Sailor 236.0 111.7 Bouckaert et al. (2012), Science 337: 957–960.
flower and a bee might slowly become, either simultaneously or one after the other, modified and adapted in the most perfect manner to each other, by the continued preservation of individuals presenting mutual and slightly favorable deviations of structure. —Charles Darwin, On the Origin of Species What is it? Reciprocal evolution between two or more organisms Symbioses—interactions between organisms—inherently involve coevolutionary processes Mutualisms, in particular, are critical to most aspects of life, at the macro- and microscopic levels Coevolution is believed to be a key driving force of biodiversity
335 SCIENCE www.sciencemag.org ks separated re combina- arwinian fi t- valleys are ness (see the n move from gher, peak, h gene com- aladaptive? er et al. ( 2) key innova- kly onto the . Ehrlich and ution could sequentially volving lin- neage leads ich, in turn, her lineage. een trying to ns coevolu- s and diver- ovements in coevolution is important. The authors study the coevolution of a virus, λ bacteriophage, and the bacterium Escherichia coli. Natural a bacterium and the ability of bacteria to pre- vent this adsorption. λ bacteriophages attach to E. coli bacteria using a tail protein called J, Biology, Univer- CA 95064, USA. LamB OmpF Resistance mutation in bacteria Four mutations in phages How to cross an adaptive valley. In an adaptive landscape, some genetic combinations have high Darwinian fi tness (adaptive peaks) and others have low fi tness (adaptive valleys). In the coevolving interaction between λ bacteriophages and E. coli bacteria reported by Meyer et al., the phages normally attach to the bacterial surface protein LamB. It takes the co-occurrence of four mutations to allow the phages to shift their attach- ment site to OmpF rather than LamB. Coevolution of the phage population with LamB fortuitously makes it possible for that unlikely combination of mutations to come together. This all-or-nothing combination of mutations creates, in effect, a gene combination that forms a narrow ridge between the two peaks, allow- ing the phage population to shift from using LamB to using OmpF. The ridge could eventually disappear as selection acts to favor resistant forms of the OmpF protein. The fi gure is intended heuristically; the full fi tness landscape was not measured in the study and cannot be captured in three dimensions. Downloaded from Thompson (2010), Science 335: 410–411.
alisms pecies s. As (e.g., other it the well- pecies es that on by oblem pecies onary Fig. 1 Diagrammatic representation of some of the major components of the geographic mosaic of coevolution between a pair of species. 11 Thompson (2012), Evo. Edu. Outreach 3: 340–346.
a coevolving host and symbiont compare? Perfectly matching phylogenies are not evidence for coevolution! Geographic vicariance events can also cause this phenomenon Ehrlich and Raven’s Escape-and-Radiate theory of coevolution Cophylogenetic mappings and Charleston’s concept of jungles a b c a b c
novel defense or counter-defense provides ecological opportunity for new species to form, the tempo of speciation is quick. Therefore, the coevolving plants and insects should have star-like phylogenies—trees that have short internal branch lengths typical of rapid speciation additional tes Many plants canals that re Farrell et al. ( that had spec paired lineag (sister groups latex and res defended. In diverse taxa species richn isons had gre resin and late Coupled w species-rich speciation sh should occur trait. Thus, th base of the s arises, the tim quickly, and m simultaneousl short internal Fig. 4 Comparison of plant and insect phylogenies consistent with escape and radiate coevolution. At time 1, the plant evolves a defense trait (orange circle) and then radiates into a number of new species that have the defense trait (green box). At a point later in time (time 2), the insect evolves a counter-defense trait (orange circle) and then Segraves (2012), Evo. Edu. Outreach 3: 62–70.
3. Tanglegram with a subtree match. In the above tanglegram the top three parasite and host taxa (unlabelled here, for clarity) form a subtree match, since ϕ preserves the isomorphism between those parts of H and P which are encircled by the light dashed line. Host Associate Codivergence Host Associate Duplication Host Associate Loss Host Associate Host switch Fig. 4. Four kinds of cophylogenetic events. (Time is from left to right.)
recently understood to be insect-pollinated Pollination syndrome involves a number of species-specific interactions where cycads attract weevils via volatile chemicals Interactions evolved independently on each continent Critical coevolutionary relationship between genetic drift, natural selection, and gene flow The cycad-weevil mutualisms can be an important model system How can the same mutualism evolve so many times independently?!
15 10 5 Miocene B Miocene Pli Neogene Q Cycas Lepidozamia Encephalartos Macrozamia Dioon Bowenia Stangeria Ceratozamia Microcycas Zamia Ginkgo etaceous Eoc P Mi Cenozoic O etaceous Eoc P PP H P H Mi Cenozoic O 100 50 0 Ma 2 3 4 Pli PP H P H Neogene Q ozamia ozamia nia d molecular imetree and constraints, d nodes (9). Geographic distributions were obtained from (2). (B) Enlarged view of timetree from (A) focusing on the Miocene–Recent. L. Paleoz, Late Paleozoic; P, Paleocene; Eoc, Eocene; O, Oligocene; Mi, Miocene; PPH, Pleistocene–Pliocene–Holocene; Q, Quaternary; Pli, Pliocene; PH, Pleistocene–Holocene. on February 15, 2012 www.sciencemag.org Downloaded from Nagalingum et al. (2011), Science (334): 796–799.
Predominant cycad-weevil mutualism of Africa Amorphocerini consists of two parasitic genera, Propor and Amorphocerous, and the pollinating genus, Porthetes Evidence for coevolution in the interaction The Porthetes mutualism is thought to have evolved from a parasitism Porthetes has olfactory globules and a grooved body that increase pollen retention
SIV in a number of independent introductions (host-shifts) from other primates to humans The two types of HIV, HIV-1 and HIV-2, originate from different primates HIV-1 originated from chimpanzees (SIVcpz) and is the pandemic strain, particularly group M HIV-2 originates from sooty mangabeys (SIVsmm) and is significantly less virulent than HIV-1 It should not be a surprise that HIV-1 is more virulent, as its predecessors infect chimpanzees, the closest relatives of humans
2. The two codivergence-optimal maps of the lentivirus phylogeny into the primate phylo es show the two alternative locations for the earliest host switch on to the chimpanzee lineage Charleston & Robertson (2002), Syst. Biol. 51: 528–535.
are the relationships between the organisms in the tree? The substitution model—are some kinds of mutations more likely than others? The clock model—what is the mutation rate, and does it vary across the tree? The population (a.k.a. coalescent) model—how does the organismic population change over time? Further abstractions A geographic model—how does the organism spread across space over time? A gene tree–species tree model—how do specific genes evolve in the context of organismic evolution?
Model cophylogenies parsimoniously: the simplest explanation is the best explanation Give each event type a penalizing score Find the mapping with the lowest score: this is your solution!
What score should I give each event type? Co-Re-Pa can “estimate” the appropriate scores There often exist hundreds of differing solutions at the minimum score Assumption that the given phylogenies are perfect Uncertainty is a fact of science, however we have probalistic ways to represent it Poor treatment of temporal information, and essentially ignorance of geographic data
P (B|A) P (A) P (B) (1) so we can test the probability of an evolutionary model with P (T, θ|D) = P (D|T, θ) P (T, θ) P (D) (2) where P (D|T, θ) := What is the probability of generating our exact data when simulating under this model? P (T, θ) := What do we already believe is the probability that this model actually exists?
coevolutionary parameters via Bayesian MCMC, e.g. host-shift rate What are the odds of a virus making a host-shift to humans? Simultaneous estimation host and symbiont phylogenies Existing methods for evaluating analytical convergence Systematic treatment of phylogenetic uncertainty because with Bayesian MCMC we integrate over the posterior distribution
and estimates to avoid unrealistic cophylogenies, i.e. time-traveling host-shift events Recognize the ecological perspective: coevolving organisms must be sympatric! Identify coevolutionary hot-spots using models of gene flow and speciation And of course, integration over uncertainties
existing Bayesian phylogenetics package No need to reinvent the wheel: all existing and future models integrated into BEAST are available for use Consideration of cospeciation as the null hypothesis Treatment of duplication, host-shift, and losses as poisson processes and estimate their rates Rejection of time-travel events Constraint of geographic correlation between host and symbiont, to force optimization of cophylogenetic mapping
Ahmad Baluch Ahmad Baluch X Tekke Yomut Shahsevan Qashqa’i Boyer Ahmad Bakhtiari Baluch (b) key: co-divergence Cophylogeny of populations and cultures J. J. Tehrani et al. 3871 on March 16, 2013 rstb.royalsocietypublishing.org m Tehrani et al. (2010), Phil. Trans. R. Soc. B 365: 3865–3874.
solved for simpler duplication-loss model Allowing transfer events (i.e., host-shifts) adds complexity 130 Michael A. Charleston Fig. 6. Untraceable “short stay” parasite. On the left the parasite switches host lineage, then switches back to the nearer host, leaving no trace on the m one. This is indistinguishable from simply switching to the nearer host lin Short stays: A short stay is a case where a parasite lineage switches t possibly after a duplication, and then switches away again leaving no d behind. It is clear to see why we cannot hope to reconstruct this from of H, P and ϕ, in Fig. 6. 3.4 Pandora’s box: Sampling, ghosts, sources and sinks Another spanner in the works is caused by the inevitable difficulties o all taxa: we cannot guarantee to find all cases of parasitological infestat all lineages of host, nor can we guarantee to find all host species at all. of the parasite life cycle may be on other hosts entirely, in different niches, and we may simply miss their presence by chance. If we are missing host taxa (either through extinction or lack of then our reconstruction of the past associations between host and p likely to contain deflated numbers of lineage sorting events: we shall h them simply because we are unaware of the ancient host species, no from our phylogeny.
cophylogenies present evidence supporting or contradicting a coevolutionary hypothesis? What effects do we expect coevolution to have on speciation processes?