Meiotic drive shapes rates of karyotype evolution in mammals

Abstract

Chromosome number is perhaps the most basic characteristic of a genome, yet generalizations that can explain the evolution of this trait across large clades have remained elusive. Using karyotype data from over 1000 mammals, we developed and applied a phylogenetic model of chromosome evolution that links chromosome number changes with karyotype morphology. Using our model, we infer that rates of chromosome number evolution are significantly lower in species with karyotypes that consist of either all bibrachial or all monobrachial chromosomes than in species with a mix of both types of morphologies. We suggest that species with homogeneous karyotypes may represent cases where meiotic drive acts to stabilize the karyotype, favoring the chromosome morphologies already present in the genome. In contrast, rapid bouts of chromosome number evolution in taxa with mixed karyotypes may indicate that a switch in the polarity of female meiotic drive favors changes in chromosome number. We do not find any evidence that karyotype morphology affects rates of speciation or extinction. Furthermore, we document that switches in meiotic drive polarity are likely common and have occurred in most major clades of mammals, and that rapid remodeling of karyotypes may be more common than once thought.

Keywords: Chromosome number; meiosis; meiotic drive; probabilistic phylogenetic models.

Figure 1

Schematic representation of the model for joint evolution of a binary character and chromosome number. The model depicts the possible transitions of a lineage that is in character state 1 or 2 and has i haploid number of chromosomes. For each binary state, there can be as many as four transition rates: descending dysploidy δ leading to $i-1$ chromosomes, ascending dysploidy γ leading to $i+1$ chromosomes, demi‐polyploidy ν leading to 1.5i chromosomes, and polyploidy ρ leading to 2i chromosomes (with subscripts on those rates indicating they can differ for binary states 1 and 2). Transition rates between the binary character states are q ₁₂ and q ₂₁. The chromePlus package can implement this model of state changes, and it can include state‐dependent speciation and extinction effects. In this case, the binary character states 1 and 2 have speciation rates λ₁ and λ₂ and extinction rates μ₁ and μ₂.

Figure 2

Simulation analysis of chromePlus model. In (A–C) the vertical axis is the proportion of simulations in which the states of the binary character were inferred to have different rates of chromosome number evolution. In (A) and (B), results are shown for test one (symmetric transitions in the binary trait) and test two (asymmetric transitions in the binary trait) with tree sizes of 50, 100, 200, and 500 taxa. The rates ratio describes the difference in the rates of chromosome number evolution in the two states of the binary character. A rates ratio of 1, bottom lines, illustrate the false‐positive rate. The three other lines in each plot represent power with different magnitudes of rate differences. In (C) are results for test three where speciation in one binary state was three times higher than in the other, and trees contained 200 taxa. Rates of chromosome number evolution are equal (left), greater for the state with higher diversification (center), or greater for the state with lower diversification (right). Labels on the horizontal axis indicate whether the inference model was with or without state‐dependent diversification. In (D) are results for test four where simulations with equal rates of chromosome evolution were performed on an empirical phylogeny. Labels on the horizontal axis indicate whether the simulation was for low or high rates of chromosome number and binary trait evolution. The vertical axis indicates the proportion of simulations in which the states of the binary character were inferred to have different rates of diversification or chromosome number evolution.

Figure 3

Distribution of karyotype morphology in mammals. The horizontal axis indicates the proportion of chromosomes that are monobrachial, and the vertical axis provides the total count of haploid chromosomes in a female genome. The margins show histograms for each of these characters, emphasizing the unimodal distribution of chromosome number and the bimodal distribution of karyotype morphology. Values for each species are plotted, and where they coincide, point size indicates the number of species. The apparent banding pattern results from the fewer possible values of the proportion as the haploid number decreases.

Figure 4

Differences in rates of chromosome number evolution for taxa with mismatched and matched karyotypes. The difference, $\mathrm{\Delta }r$, is calculated as the mean rate in taxa with mismatched karyotypes minus the mean rate in taxa with matched karyotypes. The 95% credible interval of each distribution is indicated by the matching line beneath the distribution. Panel (A) shows results for all mammals, and each subsequent panel shows the results from the independent analysis of a subclade, as indicated on top of each panel.

Figure 5

Meiotic drive polarity switching. (A) The rate of transition from matched karyotype to mismatched karyotype. (B) The relationship between rate of karyotype switching and the proportion of taxa with mismatched karyotypes.