The Evolution of Chromosome Numbers: Mechanistic Models and Experimental Approaches

Abstract

Chromosome numbers have been widely used to describe the most fundamental genomic attribute of an organism or a lineage. Although providing strong phylogenetic signal, chromosome numbers vary remarkably among eukaryotes at all levels of taxonomic resolution. Changes in chromosome numbers regularly serve as indication of major genomic events, most notably polyploidy and dysploidy. Here, we review recent advancements in our ability to make inferences regarding historical events that led to alterations in the number of chromosomes of a lineage. We first describe the mechanistic processes underlying changes in chromosome numbers, focusing on structural chromosomal rearrangements. Then, we focus on experimental procedures, encompassing comparative cytogenomics and genomics approaches, and on computational methodologies that are based on explicit models of chromosome-number evolution. Together, these tools offer valuable predictions regarding historical events that have changed chromosome numbers and genome structures, as well as their phylogenetic and temporal placements.

Keywords: chromosome numbers; cytogenomics; dysploidy; genome evolution; phylogenetic models; polyploidy.

Fig. 1

Mechanisms of descending dysploidy in plants. (A) Robertsonian translocation. (B) End-to-end translocation. (C) Nested chromosome insertion. The blue lightning symbols denote the location of double-strand breaks (breakpoints), the black sandglass symbols represent centromeres, and the small white/gray rectangles stand for (sub)telomeric/pericentromeric repeats.

Fig. 2

Complementary approaches of comparative plant genomics. (A and B) BAC-based chromosome painting (CP). (A) CP using chromosome-1-specific BAC clones of A. thaliana on pachytene chromosomes of this species. The set of 66 BACs (∼6.7 Mb) was arbitrarily divided into ten alternatively labeled BAC contigs. (B) Comparative BAC-based CP. Pachytene chromosome 6 of Noccaea caerulescens (Alpine Penny-cress) painted using chromosome-specific BAC clones of A. thaliana. Capital letters refer to ancestral genomic blocks. cen, centromere. (C and D) Oligo painting. (C) Multiplex PCR-based oligo painting (MP-OP) using eight oligo probes specific for cucumber (Cucumis sativus) chromosome 4 on pachytene chromosomes of this species. (D) Comparative MP-OP using the same probes as in (C) revealing two homeologous chromosomes (M7 and M8) in the melon genome (Cucumis melo). (E) Chromosome-scale genome comparison among two strawberry (Fragaria) and one black raspberry (Rubus) species revealing the conserved versus corrupted intergenome chromosome collinearity. All three genomes were sequenced and assembled using the PacBio single-molecule sequencing technique (Edger et al. 2018, 2020; VanBuren et al. 2018). (F) High-throughput chromosome conformation capture (Hi-C) map of the Rubus occidentalis genome. Putative locations of centromeres are visible for some of the seven chromosomes. Figures were contributed by the authors of Mandáková and Lysak 2016 (A), Mandáková et al. 2015 (B), Bi et al. (2020) (C and D), Hardigan et al. (2020) (E), and VanBuren et al. (2018) (F).

Fig. 3

The possible transitions allowed in the chromEvol model. The models implemented in chromEvol allow for several types of transitions from a genome with i haploid chromosomes to a genome with j chromosomes: ascending dysploidy (j = i + 1), descending dysploidy (j = i − 1), genome duplication (j = 2×i), demi-ploidy (j = 1.5×i), and base-number transitions, in which the addition of any multiplication of a basic number b is allowed (j = i + c × b; here c = 1 and b = 6).

Fig. 4

Heterogeneous models of chromosome-number evolution. (A) The standard chromEvol model assumes that chromosome-number dynamics are similar throughout the phylogeny. In this example, however, the subtree on the left is characterized by low rate of polyploidization and high dysploidy rates, whereas the subtree on the right is representative of a hot spot of polyploidizations. If this clade is analyzed using a single rate matrix (represented by a single color to all branches) the model would not fit the data well and would possibly result in erroneous inferences. (B) The use of a split model allows examining whether distinct patterns of chromosome-number change are exhibited in different taxonomic clades. Here, the subtree on the right is a priori classified to have a rate matrix that is distinct from that assumed for the rest of the tree, as represented by gray and black branches, respectively. (C and D) The effect of a character trait on rates of chromosome-number change. (C) A hypothetical mapping describing the evolution of a character trait (here, growth form). W and H denote woody and herbaceous states, respectively, whereas the black and green branches represent the corresponding times spent in each state. In (D), this mapping induces distinct patterns of chromosome-number change, represented by black and gray lines, respectively. This could capture a scenario where the polyploidization rate of herbaceous lineages is higher than that of woody lineages and could be modeled using a joint model for the evolution of both chromosome numbers and discrete character traits. Numbers at the tips represent chromosome-number assignments.