LC and PJdJ conceived and designed the experiments. LC, AK, and JW performed the experiments. GMV, CS, RAH, and AM analyzed the data. BFHtH, BZ, KO, AM, JW, JR, and SH contributed reagents/materials/analysis tools. LC, GMV, and PJdJ wrote the paper.
The authors have declared that no competing interests exist.
Gibbons are part of the same superfamily (Hominoidea) as humans and great apes, but their karyotype has diverged faster from the common hominoid ancestor. At least 24 major chromosome rearrangements are required to convert the presumed ancestral karyotype of gibbons into that of the hominoid ancestor. Up to 28 additional rearrangements distinguish the various living species from the common gibbon ancestor. Using the northern white-cheeked gibbon (2n = 52)
It is commonly accepted that mammalian chromosomes have undergone a limited number of rearrangements during the course of more than 100 million years of evolution. Surprisingly, some species have experienced a large increase in the incidence of rearrangements, including translocations (exchange between two non-homologous chromosomes), inversions (change of orientation of one chromosomal segment), fissions, and fusions. Within the primate order, gibbons exhibit the most strikingly unstable chromosome pattern. Gibbon chromosomal structure greatly differs from that of their most recent common ancestor with humans from which they diverged over 15 million years ago. The authors are interested in the mechanisms causing this extraordinary instability. In this study, they employed modern techniques to compare the human and white-cheeked gibbon chromosomes and to localize all the regions of disrupted homology between the two species. Their findings indicate that the molecular mechanism of gibbon chromosomal reshuffling is based on the same principles as in other mammalian species. To explain the 10-fold higher incidence of gibbon chromosomal rearrangements, it will be necessary to pursue future studies into other biological factors such as inbreeding and population dynamics.
During recent years, great progress has been made in understanding the evolutionary processes governing mammalian chromosomal organization. It is now commonly accepted that the mammalian karyotype has undergone a limited number of major rearrangements over the course of more than 100 million years [
Recent studies describing the dynamics of mammalian genome evolution indicate a “reuse” of genomic regions for independent evolutionary breakpoints in different lineages as well as the presence of hotspots and fragile sites more prone to rearrangements. These fragile loci frequently coincide with regions enriched for segmental duplications (SDs) in primates and involved in human genomic disorders [
Gibbon karyotypic changes have previously been investigated by cytogenetic banding analysis [
Here, we compared the karyotypes of
The three main resources used in this study were 1) high-resolution microarray slides containing about 32,000 BAC clones spanning the entire human genome (“32K set”), 2) a genomic BAC library for the northern white-cheeked gibbon (CHORI-271) described in more detail at
To better understand karyotype instability, gibbon-specific sequences at the break of synteny regions (BOSRs) need to be analyzed. To determine if it is possible to attribute breakpoints to specific sequence elements or to the genomic architecture of these regions, we sequenced a preliminary set of 11 gibbon BAC clones spanning BOSRs.
We employed array painting (see [
Optimal experimental conditions were found by hybridizing three gibbon chromosomes separately (NLE13, NLE20, and NLE10) in a preliminary array-painting experiment. Array-painting experiments were then economized by pooling the sorted gibbon chromosomes so that the gibbon chromosomes in each pool detected distinct human chromosomes. This “smart pooling” was possible because the gibbon–human synteny regions have previously been identified through chromosome painting [
Single test and reference hybridization results revealed that the signal-to-noise ratio was too low for accurate detection of BOSR coordinates. After determining that the hybridization noise was systematic, we developed a noise-reduction method (
(A and B) The plotted value of Log2 ratio/chromosome length for human Chromosome 2 after hybridization with sorted gibbon chromosomes NLE14 and NLE19, respectively.
(C) Results of the application of the difference method to the datasets in (A) and (B). After canceling out the systematic variation, it is possible to discern three different regions from left to right, one amplified (1), one deleted (2), and one at the baseline (3).
Altogether, 64 BOSRs have been mapped to the human genome (Hg17, UCSC May 2004) (
Array-Painting Results
Ideogram of human chromosomes with orthologous gibbon chromosomes identified by array painting represented by colored bars to the left of each chromosome. Each segment is named after the gibbon chromosome followed by a small letter that refers to its mapping order in the gibbon chromosome. The BOSRs have been defined for convenience by numbers (
In order to confirm the accuracy of array-painting results, ten BOSRs were validated by FISH experiments where BAC clones from the “32K set” were hybridized to NLE metaphase preparations. In most of the cases, a single BAC hybridized to two disparate locations as expected. Only BOSR-32 and BOSR-11 demonstrated a single hybridization signal, suggesting that the breakpoint was located between two BACs or in a small region of overlap between them. A few examples of FISH experiments are shown in
(A) FISH experiments to validate breakpoints identified by array painting. Images 1 and 2 show hybridization on NLE metaphase preparations with human BACs spanning breakpoints identified by array painting. The yellow color in image 1 is due to the overlap of red and green spots as both BACs map on the same chromosome. Image 3 shows a similar experiment done on HLA metaphase preparations. The reciprocal position of the BACs used in each experiment is shown in the boxes below the images.
(B) FISH validation experiments on six gibbon BAC clones spanning three reciprocal breakpoints for the same rearrangement. In the diagrams, the rearrangements are illustrated starting from the ancestral chromosome form. Abbreviation: AC, ancestral chromosome.
(C) Gibbon BACs spanning inversion breakpoints were tested by FISH on human and gibbon metaphases. A BAC spanning an inversion in gibbon is expected to give a split signal on the human chromosome and a single signal on the corresponding gibbon chromosome.
Additionally, six of these BOSRs, thought to be common to all gibbons [
Both ends of 5,376 clones from the gibbon genomic BAC library CHORI-271 were sequenced and mapped onto the human genome using the BLAT program [
Analysis Using the Pairing Criteria of Gibbon BES Mapped on the Human Genome
For the purpose of visualizing the gibbon clones mapped onto the human genome, we created a software tool that graphically depicts the chromosomal position of gibbon BAC mappings on the human genome. This software allows for full chromosome views as well as views showing a user-configurable window size (
Gibbon clones identified as putatively spanning interchromosomal rearrangements were mapped by FISH on gibbon metaphase preparations. Clones giving signals on more than two gibbon chromosomes were considered possible clone artifacts (chimeric clones) or duplicated regions in the gibbon and were removed from further analysis. We also screened the gibbon BAC CHORI-271 library (Materials and Methods) in order to identify at least two additional clones spanning identified breakpoints.
Gibbon BAC Clones Spanning Breakpoints Identified by Mapping End Sequences
Based on our pairing criteria, most of the clones identified as spanning intrachromosomal rearrangements resulted from insertion/deletions (indels) in either the gibbon or the human genomes (74 out of 97). Indels cause discrepancies between paired BES mapping distances on the human genome compared to the average gibbon BAC insert size. We formally defined this as BES mappings at a distance less than or greater than three standard deviations from the 172-kb-clone insert size. We verified the insert size of the 74 clones spanning putative indels using NotI digestion and pulsed-field electrophoresis. Based on the pulsed-field electrophoresis, the 60 gibbon BACs with BES mapping distances of 40–60 kb relative to the human genome were found to not be indels, as they had actual insert sizes in the 40–60 kb range. We presumed that the remaining 14 clones with BES distances exceeding 300 kb represent actual insertions in the human genome or deletions in the gibbon genome. Results are summarized in
Gibbon BAC Clones Spanning Indels
Clones putatively spanning inversion breakpoints were validated by comparisons with BOSRs previously defined by array painting (
The goals of our study were 1) to obtain a map of the BOSRs between human and gibbon at high resolution and 2) to identify species-specific clones spanning chromosomal rearrangements for use in further molecular analysis. In pursuit of our second goal, we selected for further analysis 38 gibbon BAC clones corresponding to BOSRs identified by array painting human BACs. We constructed probes across these BOSRs at 75-kb intervals based on the human genome sequence and used these probes to screen the gibbon library. Using this approach, we identified an additional 26 gibbon clones containing breakpoint loci (15 inversions and 11 translocations) (
Gibbon BAC Clones Isolated Using the Array-Painting Map Combined with Filter Screening and BES Mapping
To ensure that we identified rearrangements that occurred in the gibbon lineage, we mapped the BES of gibbon clones identified as spanning rearrangement breakpoints onto the latest genome assemblies of rhesus macaque (UCSC Build rheMac2) and chimpanzee (UCSC Build panTro1) using BLAT. We removed ambiguous mappings and classified the remaining mappings using the same pairing criteria applied to the human genome mappings. We classified putative rearrangements as 1) gibbon specific if gibbon was rearranged relative to human, chimpanzee, and macaque; 2) great ape specific if gibbon was not rearranged relative to macaque, but was rearranged relative to human and chimpanzee; and 3) human specific if gibbon was not rearranged relative to macaque and chimpanzee but was rearranged relative to human (
One of the events classified as great ape–specific is the inversion of human Chromosome 3, with breakpoints at 3p25 and 3q21, which are regions already known as rearrangement hotspots in primates [
In total, we identified 110 breakpoints between human and gibbon chromosomes due to intra- and interchromosomal rearrangements. Of those, 100 occurred during the evolution of the gibbon.
It is widely accepted that regions of chromosomal instability and SDs colocalize more frequently than expected by random chance. SDs are blocks of DNA 1–400 kb in length, repeated in the genome with a high level of sequence identity (>90%) [
We first analyzed the overlap between the human regions orthologous to gibbon-specific BOSRs and the human SDs reported in the UCSC browser [
To statistically validate the significance of these data, a simulation was run in which the 100 breakpoint regions were randomly relocated 1,000 times in their original chromosome, emulating a random-breakage model (
(A) The figure shows the sampling distribution of the overlap between SDs and a random set of regions obtained by relocating our original sample 1,000 times in the corresponding chromosomes. The original sample fell more than three standard deviations away from the mean of the simulated distribution (red arrow).
(B) The regions from the original sample were expanded in 100 kb increments. The number of regions overlapping with SDs at each step is shown.
(C) We measured the amount of overlap (in base pairs) of our 100 regions, while shifting each of them up to 5 Mb left and right of their original positions in 100 kb increments. The strong correlation between the original position (red arrow) and SD content is shown.
Measuring overlap alone does not sufficiently express the proximity of these regions to SDs; thus, the breakpoint regions were expanded in 100 kb increments and monitored for variation in the number of regions overlapping with SDs (
Of the 46 BOSRs overlapping with human SDs, 27 are covered by at least one gibbon clone. We used these clones for interphase FISH experiments on NLE. In 22 cases, multiple signals were evident on NLE nuclei, suggesting duplicated regions. The remaining five clones showed no indication of duplications at the cytogenetic level. Additionally, 10 out of these 22 clones were duplicated in two other species of gibbon (
We cannot assume that these duplications were responsible for the chromosomal rearrangement events in NLE, as we have insufficient data to indicate the duplications predated the breakage events. However, this correlation is consistent with a well-established model in which duplications are indicative of the “plasticity” of a region [
We analyzed the finished sequence of 11 gibbon clones comprising a representative sample of the clones identified in this study. Although the rearrangement events occurred in ancestral chromosomal sequences that are in part altered in the current genome, the study of orthologous sequences can nevertheless still provide us with information about the nature of the genomic instability present in these regions. The breakpoint spanned by each clone was localized to the break of synteny between the gibbon clone and the human genome.
The first interesting discovery to emerge from the analysis of sequenced clones was the presence of “micro-rearrangements” that fell below the resolution of the BES mappings. Micro-rearrangements were observed in two clones, CH271-372B11 and CH271-236L11, in which the complete sequence revealed regions orthologous to human chromosomes other than those predicted by BES or array painting (
Two sequenced gibbon clones spanning rearrangement breakpoints revealed the presence of additional segments of synteny not observed by other methods. In both cases, the first break of synteny was found to contain SDs and the second to contain interspersed repeats.
Five BOSRs, including the two BOSRs mentioned above, were within 5 kb of SDs. One example is the clone CH271-262E11 mapping to NLE Chromosome 19 and spanning a breakpoint between human Chromosomes 2 and 17. The BOSR in this clone was identified at the base-pair level and was found to be adjacent to the growth-hormone gene cluster. The breakpoint is 20 bp away from a duplicated segment containing the ortholog of gene
Six of the BOSRs coincided with interspersed repeats (SINEs, LINEs, and LTRs). In clone CH271-236L11, an alpha-satellite was identified due to the proximity of the BOSR to a centromere (
The presence of the AT-rich repeat in relation to these BOSRs may indicate a different breakpoint-inducing mechanism. Recently, Gotter and colleagues showed that the propensity to form secondary structures such as stem-loops can confer fragility to DNA [
This study describes the mapping and validation of a large number of syntenic breakpoints between homologous chromosomes of human and NLE. All the translocation breakpoints previously identified by chromosome-painting studies were mapped to the human genome at a greatly increased resolution. About 20 additional rearrangements were discovered as a result of the higher sensitivity of our approaches. Overall, our research identified about 100 breakpoints occurring in the gibbon lineage. The study also yielded gibbon BACs containing breakpoint sites. In 11 sequenced gibbon BACs, we found elements near the breakpoints previously shown to play a key role in primate chromosome plasticity and evolution. Within the sequenced BACs three different patterns were evident. First, two BACs contained additional breakpoints that may have resulted from a complex, nonreciprocal translocation event or from subsequent chromosomal rearrangements. Recent high-resolution breakpoint analyses on human translocations thought to be balanced showed various “microtranslocations” [
No generalized pattern unique to gibbon breakpoints is evident from the present molecular data. It remains to be determined if the greater number of chromosomal rearrangements in the small apes is due to an enhanced frequency of chromosomal breakages or an increased ability to rescue derivative chromosomes in comparison to other mammals, possibly due to mating behavior or inbreeding. We believe that these questions may be answered by examining additional aspects of small ape biology such as behavioral factors and population dynamics.
BACs (32,855) spanning 95% of the human euchromatic genome were assembled and rearrayed into 384-well microtiter dishes [
FISH was used as a validation method for BOSRs identified uniquely by array painting or for gibbon BACs spanning inversions. Metaphase preparations of NLE were obtained from the same cell line used for chromosome sorting and previously described by Müller et al. [
BAC DNA extraction was performed as reported by Ventura et al. [
To identify BACs spanning putative breakpoint regions, overgo probes of 40 bp [
The images were analyzed with the software ArrayVision Ver6.0 (Imaging Research Inc.,
The results of array painting experiments done with different pools were combined for each human chromosome. After applying the difference method for noise reduction (see text) we identified all 64 BOSRs at a resolution of 300 kb (average) with the employment of a limited number of experiments. The figure shows the results obtained for all human chromosomes.
(240 MB PPT)
This tool was developed in order to easily localize gibbon clones spanning a translocation or inversion breakpoint in human.
(493 KB PPT)
A sample of gibbon BAC clones overlapping human SDs was hybridized on NLE nuclei. The presence of duplications was revealed by the presence of either multiple signals or a single but broadened signal. The figure shows the results obtained with four clones also tested on HLA and
(122 KB JPG)
(41 KB DOC)
The gibbon chromosomes were divided into four pools in order to minimize the number of array-painting experiments. A smart-pooling strategy was used, taking advantage of the data available in the literature. Through this approach the repetition of one human chromosome in the same pool was avoided. Additionally, three gibbon chromosomes were hybridized in individual experiments.
(45 KB DOC)
The table reports the outcome of the mapping of gibbon clones spanning BOSRs on the latest genome assembly of Rhesus macaque (reMach2) and chimpanzee (panTro1). Depending on the result, clones were classified into three evolutionary groups: 1) gibbon specific, 2) great ape specific, and 3) human specific.
Mapping results not consistent with human are in italics.
(134 KB DOC)
The National Center for Biotechnology Information (NCBI) Entrez Gene (
We are extremely grateful to Dr. Ulli Weier and his group at Lawrence Berkeley National Laboratory for making their facilities available to us. We are also thankful to Dr. Yuko Yoshinaga for help in managing communications with the sequencing agency for the gibbon BAC end sequencing. We finally thank BACPAC Resources for their help with the production of the rearrayed filters employed in the secondary screenings.
bacterial artificial chromosome
BAC end sequence
break of synteny region
degenerate oligonucleotide-primed PCR
fluorescence in situ hybridization
insertion/deletion
segmental duplication