The molecular genealogy of sequential overlapping inversions implies both homologous chromosomes of a heterokaryotype in an inversion origin

Abstract

Cytological and molecular studies have revealed that inversion chromosomal polymorphism is widespread across taxa and that inversions are among the most common structural changes fixed between species. Two major mechanisms have been proposed for the origin of inversions considering that breaks occur at either repetitive or non-homologous sequences. While inversions originating through the first mechanism might have a multiple origin, those originating through the latter mechanism would have a unique origin. Variation at regions flanking inversion breakpoints can be informative on the origin and history of inversions given the reduced recombination in heterokaryotypes. Here, we have analyzed nucleotide variation at a fragment flanking the most centromere-proximal shared breakpoint of several sequential overlapping inversions of the E chromosome of Drosophila subobscura -inversions E₁, E₂, E₉ and E₃. The molecular genealogy inferred from variation at this shared fragment does not exhibit the branching pattern expected according to the sequential origin of inversions. The detected discordance between the molecular and cytological genealogies has led us to consider a novel possibility for the origin of an inversion, and more specifically that one of these inversions originated on a heterokaryotype for chromosomal arrangements. Based on this premise, we propose three new models for inversions origin.

Figure 1

Schematic representation of chromosomal arrangements E_st, E₁₊₂, E₁₊₂₊₉, E_1+2+9+3, and E_1+2+9+12 of Drosophila subobscura. Inversions originating these arrangements are indicated by crossed lines. Breakpoint regions including the A fragment that is shared by various inversions are highlighted. The cytological location of these breakpoints on the Kunze-Mühl and Müller map is given. A_p and A_d refer, respectively, to the proximal and distal section of the A fragment (see text). The A_d section was duplicated when inversion E₉ originated. Not at scale.

Figure 2

Functional annotation of breakpoint regions spanning fragment A. Annotation extracted from our previous work^,. Breakpoint regions are named as in Fig. 1 except for the GAL region that is here presented in reverse orientation (and therefore named LAG) to facilitate its comparison with the other regions. Chromosomal arrangements presenting each breakpoint region are indicated on the rightmost part of the figure. A_p and A_d refer, respectively, to the proximal and distal sections of the A part (see text). Small black arrowheads indicate the location of amplification primers. The sequenced A fragment at each breakpoint is rose shadowed. B_p refers to the proximal fragment of the ancestral B part of the E_st arrangement with the arrow indicating its orientation relative to the breakpoint. Colored arrowed boxes represent protein-coding regions. Green-striped boxes represent the presence of multiple snoRNA genes. SGM indicates an SGM element. α indicates an alpha element exhibiting some similarity to the SGM element.

Figure 3

Neighbor-joining tree of the A fragment sequences corresponding to the breakpoint regions of different E chromosomal arrangements. Bootstrap values >70% (based on 1000 replicates) are shown on the tree. Positions with over 5% alignment gaps, missing data, or ambiguous bases were not considered. D. guanche was used as outgroup.

Figure 4

Schematic representation of the NHEJ-4-chromosome model for the origin of inversion E₉. The sequential steps of how arrangement E₁₊₂₊₉ could have originated from an E_st/E₁₊₂ heterokaryotypic individual through inversion E₉ are graphically represented in the central part of the figure. Fragments flanking the different breakpoint regions are labeled as in Fig. 1. Initial state: pairing of the E homologous chromosomes of an E_st/E₁₊₂ heterokaryotypic individual with discontinuous arrows indicating the location of future breaks. Parts flanking future breaks are labeled in E_st and E₁₊₂ homologues by ordinal numbers and roman numerals, respectively. Upper left corner inset, image of an E_st/E₁₊₂ polytene chromosome preparation. First step: a total of four breaks considering both homologous chromosomes, with the two breaks in the proximal region occurring at different sites in both homologues —between sections B_p and B_d of the E_st homologue and between sections A_p and A_d of the E₁₊₂ homologue—, and those in the distal region (KL) occurring at the same site. Discontinuous lines indicate the location of breaks. Second step: inversion of the central fragment of the E₁₊₂ homologue and resolution of the double-strand breaks. Insets on both sides of the central scheme highlight the resolution phase. Final state: result of the inversion process with the generation of the E₁₊₂₊₉ arrangement. Also shown within a grey-shaded box are the chromosomal fragments that might have resulted —highlighted by a question mark,?— in an evolutionary unsuccessful arrangement.

Figure 5

Schematic representation of the NHEJ-3-chromosome model for the origin of inversion E₉. The sequential steps of how arrangement E₁₊₂₊₉ could have originated from an E_st/E₁₊₂ heterokaryotypic individual through inversion E₉ are graphically represented in the central part of the figure. Fragments flanking the different breakpoint regions are labeled as in Fig. 1. Initial state: pairing of the E homologous chromosomes of an E_st/E₁₊₂ heterokaryotypic individual with discontinuous arrows indicating the location of future breaks. Parts flanking breakpoints are labeled as in Fig. 4. First step: a total of three breaks considering both homologous chromosomes, with the two breaks in the proximal region occurring at different sites in both homologues —between sections B_p and B_d of the E_st homologue and between sections A_p and A_d of the E₁₊₂ homologue—, and that in the distal region (KL) occurring at the E₁₊₂ homologue. Discontinuous lines indicate the location of breaks. Second step: inversion of the central fragment of the E₁₊₂ homologue and resolution of the double-strand breaks. Insets on both sides of the central scheme highlight the resolution phase. Final state: result of the inversion process with the generation of the E₁₊₂₊₉ arrangement. Also shown within a grey-shaded box are the chromosomal fragments that might have resulted —highlighted by a question mark,?— in an E_st chromosome lacking sections A_d and B_p.

Figure 6

Schematic representation of the BIR-NHEJ-chromosome model for the origin of inversion E₉. The sequential steps of how arrangement E₁₊₂₊₉ could have originated from an E_st/E₁₊₂ heterokaryotypic individual through inversion E₉ are graphically represented in the central part of the figure. Fragments flanking the different breakpoint regions are labeled as in Fig. 1. Initial state: pairing of the E homologous chromosomes of an E_st/E₁₊₂ heterokaryotypic individual with discontinuous arrows indicating the location of future breaks. Parts flanking breakpoints are labeled as in Fig. 4. First step: two breaks in the E₁₊₂ homologue, with that in the proximal region occurring between sections A_p and A_d, and that in the distal region occurring between the K and L parts of the KL breakpoint region. Discontinuous lines indicate the location of breaks. Second step: inversion of the central fragment of the E₁₊₂ homologue and resolution of the double-strand break of the proximal region through the BIR pathway and that of the distal region through the NHEJ mechanism. Insets on both sides of the central scheme highlight the different steps of the BIR and NHEJ pathways, respectively. Final state: result of the inversion process with the generation of the E₁₊₂₊₉ arrangement. Also shown is the E_st chromosome that did not undergo any break.