Distinct retroelement classes define evolutionary breakpoints demarcating sites of evolutionary novelty

Abstract

Background: Large-scale genome rearrangements brought about by chromosome breaks underlie numerous inherited diseases, initiate or promote many cancers and are also associated with karyotype diversification during species evolution. Recent research has shown that these breakpoints are nonrandomly distributed throughout the mammalian genome and many, termed "evolutionary breakpoints" (EB), are specific genomic locations that are "reused" during karyotypic evolution. When the phylogenetic trajectory of orthologous chromosome segments is considered, many of these EB are coincident with ancient centromere activity as well as new centromere formation. While EB have been characterized as repeat-rich regions, it has not been determined whether specific sequences have been retained during evolution that would indicate previous centromere activity or a propensity for new centromere formation. Likewise, the conservation of specific sequence motifs or classes at EBs among divergent mammalian taxa has not been determined.

Results: To define conserved sequence features of EBs associated with centromere evolution, we performed comparative sequence analysis of more than 4.8 Mb within the tammar wallaby, Macropus eugenii, derived from centromeric regions (CEN), euchromatic regions (EU), and an evolutionary breakpoint (EB) that has undergone convergent breakpoint reuse and past centromere activity in marsupials. We found a dramatic enrichment for long interspersed nucleotide elements (LINE1s) and endogenous retroviruses (ERVs) and a depletion of short interspersed nucleotide elements (SINEs) shared between CEN and EBs. We analyzed the orthologous human EB (14q32.33), known to be associated with translocations in many cancers including multiple myelomas and plasma cell leukemias, and found a conserved distribution of similar repetitive elements.

Conclusion: Our data indicate that EBs tracked within the class Mammalia harbor sequence features retained since the divergence of marsupials and eutherians that may have predisposed these genomic regions to large-scale chromosomal instability.

Figure 1

Fluorescence in situ hybridization (FISH) of eight BACs identified in KERV-1 screen. (A) Tammar karyotype depicting marsupial syntenic segments and cytological localization of BAC clones (as per [22]). (B-D) BACs FISH mapped to tammar metaphase chromosomes localizing to (B) 1q evolutionary breakpoint (EB), (C) centromeres (CEN), and (D) euchromatin (EU).

Figure 2

Phylogenetic trajectory of the chromosome segments participant in the derivation of Meu1q (segments C8 and C9). Ancestral orientation is derived from [7] and key species representing 65 million years of marsupial evolution are derived from [17,19,21]. Key is shown to left.

Figure 3

Annotation of tammar BACs. BACs for the (A) evolutionary breakpoint (EB: I6, G7, B9) on tammar 1q and (B) pericentric (CEN: B18, G17, M7) regions were annotated to obtain a visual representation of the genomic landscape of each region. Annotations include all predicted interspersed repeats and coding regions. (C) Included, in contrast, are two representations of BACs from euchromatic regions of the tammar genome (A8 and J6). An enrichment of LINE1s (dark blue) and ERVs (red) is seen in both the pericentric and EB with relatively few SINE (orange) elements present. Key to annotated elements is shown at the bottom.

Figure 4

Quantification of interspersed repeats. Percentage of sequence predicted to be (A) LINE1s, (B) endogenous retroviruses (ERVs), (C) SINE elements, (D) CR1s, and (E) Simple Repeats in the tammar by region – EU, CEN, and EB. (** statistically significant difference from EU)

Figure 5

Number of intact repeats (95% or more of consensus sequence length) estimated for every 100 kb by region (EU, CEN, EB). (** statistically significant difference from EU, # statistically significant difference from EU and CEN)

Figure 6

Multipipmaker alignment of tammar BAC G7 (1q EB) aligned to the 5 clones that make up the human IGHv contig (bottom). Alignments performed with repeats masked. Above, map of G7 from Figure 3 showing repeat distribution relative to regions of 14q32.33 orthology.

Figure 7

Map of Hsa14q32 and Meu 1q. (A) Map of M. eugenii 1q compared to the orthologous human 14q32.33 showing tammar BACs FISH mapped to the EB region of 1q and their relative position on 14q. BACs O12 and H21 have been identified by the Sanger Institute to be orthologous to Hsa14q32. BAC A8 and G7 were identified by screening the M. eugenii BAC library with KERV. BAC A8 contains a predicted protein with high homology to human TMEM179. BAC G7 contains regions with high identity to the entire region of the human immunoglobulin heavy chain variable region (IGHv). (B) The density of ERVs, LINE1s, SINEs, and simple repeats in the most distal 3.4 Mb of 14q32 in increments of 200 kb, including the IGH region and TMEM179.

Figure 8

The diversity of (A) LINEs, (B) ERV/LTRs, (C) SINEs and (D) DNA transposons in Hsa 14q32.33. Shown is the number of different types of elements from each class identified by Censor spanning Hsa14q32.33 in increments of 200 kb.

Figure 9

The percent of ERV2 type repeats identified by Censor spanning Hsa14q32.33 in increments of 200 kb showing the relative contribution of each specific ERV identified.