Evolution of Beta Satellite DNA Sequences: Evidence for Duplication-Mediated Repeat Amplification and Spreading
http://www.100md.com
分子生物学进展 2004年第9期
* Dipartimento di Anatomia Patologica e Genetica, Sezione di Genetica, Bari, Italy
Dipartimento di Biologia e Genetica per le Scienze Mediche, Milano, Italy
Dipartimento di Medicina Sperimentale, Ambientale e Biotecnologie Mediche, Monza, Italy
E-mail: raffaella.meneveri@unimib.it.
Abstract
In this article, we report studies on the evolutionary history of beta satellite repeats (BSR) in primates. In the orangutan genome, the bulk of BSR sequences was found organized as very short stretches of approximately 100 to 170 bp, embedded in a 60-kb to 80-kb duplicated DNA segment. The estimated copy number of the duplicon that carries BSR sequences ranges from 70 to 100 per orangutan haploid genome. In both macaque and gibbon, the duplicon mapped to a single chromosomal region at the boundary of the rDNA on the marker chromosome (chromosome 13 and 12, respectively). However, only in the gibbon, the duplicon comprised 100 bp of beta satellite. Thus, the ancestral copy of the duplicon appeared in Old World monkeys (25 to 35 MYA), whereas the prototype of beta satellite repeats took place in a gibbon ancestor, after apes/Old World monkeys divergence (25 MYA). Subsequently, a burst in spreading of the duplicon that carries the beta satellite was observed in the orangutan, after lesser apes divergence from the great apes–humans lineage (18 MYA). The analysis of the orangutan genome also indicated the existence of two variants of the duplication that differ for the length (100 or 170 bp) of beta satellite repeats. The latter organization was probably generated by nonhomologous recombination between two 100-bp repeated regions, and it likely led to the duplication of the single Sau3A site present in the 100-bp variant, which generated the prototype of Sau3A 68-bp beta satellite tandem organization. The two variants of the duplication, although with a different ratios, characterize the hominoid genomes from the orangutan to humans, preferentially involving acrocentric chromosomes. At variance to alpha satellite, which appeared before the divergence of New World and Old World monkeys, the beta satellite evolutionary history began in apes ancestor, where we have first documented a low-copy, nonduplicated BSR sequence. The first step of BSR amplification and spreading occurred, most likely, because the BSR was part of a large duplicon, which underwent a burst dispersal in great apes' ancestor after the lesser apes' branching. Then, after orangutan divergence, BSR acquired the clustered structural organization typical of satellite DNA.
Key Words: beta satellite ? duplication ? primate evolution
Introduction
About 5% of our genome is composed of duplicated sequences (duplicons), which consist of DNA stretches ranging from a few to hundreds of kb in length, with 90% to 100% sequence identity (Samonte and Eichler 2002). Studies on the evolutionary history of duplicons have shown that the vast majority have emerged recently during primate evolution, after the split of Old World monkeys (OWMs) and hominoid lineage (<35 MYA) (Bailey et al. 2002a; Bailey, Liu, and Eichler 2003).
Duplicons can be grouped as intrachromosomal and interchromosomal, and the latter tends to be enriched in pericentrometic and subtelomeric regions. Pericentromeric duplications usually span the transition from euchromatic genes on the short/long arm to pericentromeric satellites/centromeric alpha satellites (Horvath et al. 2000). However, despite their location, the evolutionary history of alpha satellite appears distinct from that of the pericentromeric duplicons (Willard 1991).
Beta satellite repeats (BSR) represent another family of sequences that show a predominant heterochromatic distribution, which includes the short arm of acrocentric chromosomes and chromosomes 1q12, 3q12, 9q12, and Yq11 (Meneveri et al. 1985,1993; Agresti et al. 1987, 1989). In these regions, beta satellite arrays are in linkage with LSau repeats, which are part of complex repetition units of 3.3 kb in length (D4Z4-like repeats) (Agresti et al. 1989; Hewitt et al. 1994; Lyle et al. 1995; Ballarati et al. 2002). D4Z4 repetitions and BSRs are also localized on chromosomes 4q35 and 10q26 (Bakker et al. 1995; Lemmers et al. 2002).
Evolutionary studies suggest that 4q35 represents the ancestral locus for the D4Z4 sequence and that an extensive radiation of the region occurred after the divergence of Old World monkeys and hominoids (Clark et al. 1996; Winokur et al. 1996; Ballarati et al. 2002). In this respect, a few studies have been devoted to investigating the evolution of beta satellite sequences. The origin of this class of repeats can be tentatively traced back to the orangutan (PPY), where the beta satellite organization appears different from humans (HSA), chimpanzee (PTR), and gorilla (GGO) (Meneveri et al. 1995; Hirai, Taguchi, and Godwin 1999).
To further investigate this point, we screened two different orangutan genomic libraries in search of BSR-positive DNA sectors. All the gathered results indicate that BSR originated in an early hominoid ancestor and that the initial step of BSR amplification and spreading was mediated by the insertion of very short stretches of the repeat into a long genomic region, which underwent duplicative transpositions.
Materials and Methods
Genomic Library Screening
Genomic libraries were obtained from RZPD, German Resource Center for Genome Research (PPY cosmid library n.141) and BACPAC Resources, Children's Hospital, Oakland Research Institute (PPY, CHORI-253 BAC library, segment 1; MMU, CHORI250 BAC library, segment 1). High-density arrayed cosmid and BAC filters were hybridized at 55°C in standard hybridization solution for 14 to 16 h with 32P-labeled A17a human probe that contained five repetition units of beta satellite (Agresti et al. 1987). Filters were washed twice in 2xSSC (1xSSC = 150 mM NaCl, 15 mM sodium citrate, pH 7.0) at room temperature and twice in 0.1xSSC at 65°C for 20 min. Positive cosmid and BAC clones were grown in LB medium supplemented with kanamycin or chloramphenicol, and DNA was purified by the Qiagen miniprep kit (Qiagen).
Southern Blot Hybridization
Cosmid and BAC DNAs were digested with Sau3A restriction enzyme (Biolabs), fractionated by agarose gel electrophoresis, blotted onto nylon filters (Hybond N+; Amersham), and hybridized with 32P-labeled A17a probe. Molecular hybridizations were in 2xSSC at 60°C overnight, and filters were washed at 60°C in 1xSSC, twice for 20 min. Hybridization signals were quantified by phosphoimaging (Typhoon; Amersham).
Polymerase Chain Reaction and Sequencing
Genomic DNA was obtained from human placenta, lymphoblastoid, and fibroblast primate cell lines (chimpanzee, Pan troglodytes [PTR]; gorilla, Gorilla gorilla [GGO]; orangutan, Pongo pygmaeus [PPY]; lar gibbon, Hylobates lar [HLA]; rhesus monkey, Macaca mulatta [MMU]) (http://www.biologia.uniba.it/rmc/) and from a human monochromosomal somatic cell hybrid panel (http://www.biologia.uniba.it/rmc/) by standard methods.
Polymerase chain reaction (PCR) amplifications with the different DNA templates were carried out by the primer pairs BSR forw (5'-AGGGGCTTTATCCTCATTTCACAA-3') and rev (5'-GGCCTCCATATTCCCTAACTTCC-3'); PPY2 forw (5'-TGAATTTTAGCACCCACAA-3') and rev (5'-ATTCGATTCAACCCCAGTTA-3'); PPY4 forw (5'-AGTCACTGTGGCGATTCTTCTA-3') and rev (5'-GTGTGATGTCCCCTTTCCTG-3'); PPY14 forw (5'-ATGGGCCAGGAGCTATTCACAG-3') and rev (5'-CGCAACGCCCCCTTCAACC-3') and by Alu-PCR (Breen et al. 1992).
Primer pairs BSR and PPY2, PPY4, and PPY14 were derived, respectively, from the human genomic DNA sequence AC006987 (Yp11.2) and from sequences obtained by Alu-PCR (cosmids O05112Q2 and H1653Q2) and T3 end (cosmid N1981Q2) of BSR-positive PPY cosmids. For inter-Alu and T3/T7 ends, only one strand was sequenced.
PCR was carried out in a final volume of 25 μl that contained 0.2 mM dNTPs, 1 μM each primer, and 1U Taq polymerase in standard reaction buffer (Sigma). For amplification denaturation was at 95°C for 1 min, and annealing was as follows: BSR (60°C), PPY2 (46°C), PPY4 (54°C), and PPY18 (54°C), for 1 min, with extension 72°C for 1 min. The total number of cycles was of 30 to 35. PCR products were analyzed by 1% agarose gel electrophoresis, subcloned into the pGEM-T easy vector (Promega), and sequenced on both strands by the Big Dye terminator system (Perkin-Elmer Applied Biosystems) following the instructions of the vendor.
All fluorescent traces were analyzed on the Applied Biosystem Model 3100 DNA Sequencing System (PerkinElmer Applied Biosystems). DNA sequence analysis was performed by DNASTAR software and by NCBI facilities (www.ncbi.nlm.nih.gov). New sequence data from this article are available at http://www.ncbi.nlm.nih.gov/, with accession numbers AY546195 to AY546245.
Fluorescence in situ Hybridization
Metaphase spreads were obtained by standard methods from peripheral blood lymphocytes of normal human donors and from lymphoblastoid or fibroblast primate cell lines. The probes were directly labeled with Cyanine 3-dUTP by nick-translation, according to the protocol of the vendor (Perkin-Elmer), and hybridized overnight to chromosomal preparations. Hybridizations were in 50% formamide (v/v), 10% dextran sulfate, SSC at 37°C, in the presence of Cot1 human DNA (Gibco-BRL). Posthybridization washing was in 50% formamide, SSC at 42°C, followed by three washes in SSC at 60°C (HSA), and in 50% formamide, SSC at 37 C, followed by three washes in SSC at 42°C (PTR, GGO, PPY, HLA, and MMU). Chromosomes were stained with DAPI (4',6-diamidino-2-phenylindole). Digital images were captured by a Leica DMRXA epifluorescence microscope equipped with a cooled CCD camera (Princeton Instruments). Cy3 (red) and DAPI (blue) fluorescence signals were recorded separately as gray-scale images. Pseudocoloring and merging of images were performed by use of Adobe Photoshop software.
Results
Organization of Beta Satellite Sequences in the Orangutan Genome
Cosmid and BAC orangutan genomic libraries (see Materials and Methods) were screened using a human BSR probe (A17a). Approximately 100 and 70 positive recombinant clones per haploid genome were, respectively, isolated (not shown). After Sau3A digestion and hybridization with the BSR probe, 35 randomly chosen BSR-positive clones (22 cosmids and 13 BACs) showed a single Sau3A-hybridizing band of approximately 0.6 kb (fig. 1A), as shown in total orangutan genomic DNA (Meneveri et al. 1995). None of the analyzed clones showed the release of the ladder of bands, peculiar of the bulk of human BSR clustered organization (Meneveri et al. 1995).
FIG. 1. Analysis of BSR-positive PPY genomic clones. (A) Examples of Southern blot hybridization of the human BSR–labeled probe (A17a) to Sau3A-digested DNA from cosmids (N17130Q2, N1981Q2, G1478Q3; a–c) and BACs (162K8, 1O23, 25F18; d–f) PPY genomic clones, after fractionation by 1% agarose gel electrophoresis. (B) Schematic diagram of the sequence similarity between a 0.6-kb BSR-positive PPY fragment (AY546195) of figure 1A and human entries AC006987 (chromosome Yp11.2) and AC011850 (chromosome 15). Rectangles identify the BSR region and the heavy line flanking BSR indicates unrelated sequences. Arrows indicate the position of a primer pair (BSR) flanking the BSR region in the human genome, which by PCR gives amplification bands of 450 and 520 bp on chromosomes 15 and Yp11.2, respectively. (C) Examples of PCR amplification by the BSR primer pair of BSR-positive PPY cosmids O05112Q2 and N17130Q2 that contain, respectively, the long (a) and short (b) BSR region, and of PPY genomic DNA (c). PCR products were fractionated by 1% agarose gel electrophoresis
The 0.6-kb Sau3A DNA fragment from cosmid N17130Q2 was subcloned and sequenced (AY546195). The sequence analysis revealed a 70-bp BSR sequence in one side and unrelated DNA on the other side. Data bank comparison of the 0.6-kb DNA sequence disclosed similarity to several human clones that included AC079860 (1p36.3), AC006987 (Yp11.2), and unmapped clones from chromosomes 4, 15, and 16 (AC129664, AC011850, and AC137488, respectively). Interestingly, in these entries 170 bp (chromosomes 16 and Yp11.2) or 100 bp (chromosomes 1p36.3, 4, and 15) of BSR are flanked at both sides by unrelated DNA (fig. 1B).
A primer pair (BSR forw and rev) from outside both repeat regions and defining 520 and 450 bp of DNA on chromosomes Yp11.2 and 15, respectively (fig. 1B), was then used to analyze a total of 51 BSR-positive orangutan genomic clones (27 cosmids and 24 BACs) and orangutan genomic DNA. Only two clones gave a band of 520 bp (long BSR [fig. 1C]), whereas the great majority of the analyzed clones were positive for the 450-bp band (short BSR [fig. 1C]). The same PCR reaction on orangutan genomic DNA showed the simultaneous presence of both PCR bands, with a ratio in favor of the light band (fig. 1C).
The alignment of one long BSR sequence with the consensus derived from nine short BSR DNA sequences (figure 2A and Supplementary Materials online) revealed an organization comparable to that derived from chromosomes Yp11.2 and 15: 170 and 100 bp of BSR were embedded within unrelated DNA. Furthermore, only within the 170-bp BSR region, two Sau3A sites define a 68-bp beta satellite unit. The phylogenetic analysis of the nine short BSR sequences clearly indicated the occurrence of sequence divergence, which suggests their derivation from different orangutan genomic regions (figure 2B and Supplementary Materials online). The 100-bp and 170-bp orangutan BSR regions showed a similarity of approximately 94% and 92% with a human BSR consensus sequence (not shown).
FIG. 2. Sequence analysis of the BSR region in PPY genomic clones. (A) Alignment of the PPY genomic regions carrying 170 bp (long BSR) and 100 bp (short BSR) of beta satellite repeats (in bold) and 245 bp of unrelated flanking DNA. The sequences were derived by PCR with the BSR primer pair on PPY cosmid and BAC genomic clones. The short BSR sequence represents the consensus derived by aligning a core of 345 bp out of 450 bp from four cosmids N17130Q2 (AY546197), G1478Q3 (AY546202), H1653Q2 (AY546206), and N1981Q2 (AY546211), and five BACs 162K8 (AY546219), 1O23 (AY546223), 25F18 (AY546227), 60F17 (AY546231), and 98A13 (AY546235), whereas the long BSR sequence (416 bp) was derived from a single 520-bp PCR product (cosmid O05112Q2; accession number AY546215). Periods and hyphens indicate, respectively, base identities and deletions. Within the 170-bp BSR region, Sau3A sites are underlined and the 68-bp Sau3A unit defined by arrows. (B) Phylogenetic analysis of nine DNA sequences carrying the 100 bp of BSR. The sequences from cosmids (cos) and BACs (BAC) PPY genomic clones were those utilized to derive the consensus of the short BSR region shown in (A)
Further insights into the sequence environment in which orangutan beta satellite sequences are embedded were derived by the partial sequencing (inter-Alu and T3/T7 end sequences) of five cosmid clones (fig. 3A). Besides the BSR region, the different clones contained other redundant as well as unique sequences. In particular, 15 nonredundant sequences out of 39 aligned to a region of approximately 50 kb on chromosomes Yp11.2 (fig. 3A) and 15 (not shown). Database analysis revealed that the identified DNA region is part of a larger duplicated sequence of about 80 kb on human chromosomes Yp11.2, 15, and 16 and of approximately 60 and 6 kb, respectively, on chromosomes 4 and 1p36.3 (not shown).
FIG. 3. BSR-unrelated sequence analysis of PPY clones carrying the BSR region. (A) Alignment of 15 nonredundant inter-Alu (IA) and T3/T7 end sequences (T3, or T7) and of the BSR region from five BSR-positive PPY cosmids to a DNA sequence (AC006987) from chromosome Yp11.2. PPY sequences were derived from the cosmids N1981Q2 (AY546207 to AY546210), O05112Q2 (AY546212 to AY546214), G1478Q3 (AY546198 to AY546201), H1653Q2 (AY546203 to AY546205), and N17130Q2 (AY546196). PPY2, PPY4, PPY14, and BSR identify, respectively, the location of sequences used to derive three primer pairs that span the DNA region and of the region carrying 170 bp of BSR. (B) Paralogous PCR carried out on PPY BAC 1O23 (1), 162K8 (2), 98A13 (3), 25F18 (4), and 60F17 (5) with the three (PPY2, PPY4, and PPY14) primer pairs identified in (A) and with the BSR primer pair (BSR). PCR products were fractionated by 1% agarose gel electrophoresis
Using three primer pairs derived from the orangutan sequences that spanned the duplication (labeled as PPY2, PPY4, and PPY14 in figure 3A) and the BSR primer pair, we then analyzed by paralogous PCR five BACs (fig. 3B). All the BACs exhibited the same set of PCR products. Sequencing (AY546216 to AY546218, AY546220 to AY546222, AY546224 to AY546226, AY546228 to AY546230, and AY546232 to AY546234) and alignment of the amplification bands (not shown) demonstrated that the BACs, as the analyzed cosmids, are characterized by the same repertoire of divergent sequences (for the BSR region, [see figure 2B]). These results strongly indicated the occurrence in the orangutan genome, as in man, of duplications of this nucleotide region.
Evolution of the Duplication Carrying the BSR Region
Five BACs (162K8, 1O23, 25F18, 60F17, and 98A13) were used as probes on primate chromosome spreads (HSA, PTR, GGO, PPY, HLA, and MMU). An example of comparative FISH experiments is shown in figure 4, and a summary of the chromosome location obtained by all clones is reported in table 1. On the macaque and the gibbon, the probes highlighted only the "marker chromosome," at the level of the secondary constriction bearing the rDNA array (fig. 4A [MMU and HLA]). Furthermore, the intensity of the hybridization signals was constantly higher in the gibbon than in the macaque. Conversely, the other analyzed species (HSA, PTR, GGO, and PPY) showed a multiple chromosome location that included pericentromeric regions and the short arm of acrocentric chromosomes (fig. 4A and table 1). In addition to the macaque and the gibbon, the other analyzed primates also showed almost all the chromosomal regions bearing the rDNA labeled by the duplication. The above FISH experiments, however, although clearly demonstrating the spreading of the duplication, are not proof that BSR sequences were also involved. A human clone that contained only BSR sequences was then used in FISH experiments on orangutan, gibbon, and macaque chromosome spreads. The negative results (not shown) suggested that in these species, BSR sequences were either absent or below FISH resolution. To further investigate this point, we performed a PCR assay with the BSR primer pair on primate genomic DNAs (fig. 4B). A band of 450 bp was detected in great apes and in the gibbon but not in the macaque. In addition, as already observed for the orangutan (see figure 1C), humans, chimpanzee, and gorilla also showed an additional band of approximately 520 bp. The ratio of the two bands varied in the different species. In the orangutan, the shorter band was predominant, whereas the opposite was found in the gorilla. Conversely, chimpanzee and humans showed comparable amounts of the 450-bp and 520-bp bands. The 450-bp and 520-bp bands yielded positive results when hybridized with a human BSR probe (not shown). Furthermore, we performed with the human BSR probe at low-stringency conditions, the Southern blot analysis of macaque genomic DNA and the screening of a macaque BAC library (not shown). In both cases, no hybridization signals were detected, which further supports the absence of BSR sequences in the macaque genome.
FIG. 4. Evolution of the duplicon carrying the BSR region. (A) Comparative FISH on chromosome spreads from human (HSA), great apes (PTR, GGO, and PPY), lower ape (HLA), and Old World monkey (MMU) by one BSR-positive PPY BAC (60F17) as a probe. Except for MMU and HLA, chromosomes are indicated by the phylogenetic nomenclature (roman numerals). (B) PCR assay with the BSR primer pair on genomic DNA from primate species (HSA, PTR, GGO, PPY, HLA, and MMU). PCR products were fractionated by 1% agarose gel electrophoresis
Table 1 Summary of FISH Experiments Performed on Chromosome Spreads from Primate Species with Several PPY Clones Carrying Beta Satellite Repeats Embedded into a Duplication.
Human Chromosome Distribution of the Short and Long BSR Regions
A panel of human monochromosomal somatic cell hybrids was amplified with the BSR primer pair (fig. 5A). Chromosomes 1, 13, 15, and 21 showed the shorter version of the region (450 bp), whereas chromosomes 14, 22, X, and Y exhibited the longer version (520 bp). Only chromosome 20 showed the simultaneous occurrence of the two amplification bands.
FIG. 5. Human chromosome distribution of the short and long BSR regions. (A) PCR assay with the BSR primer pair on DNA from a panel of human monochromosomal cell hybrids. Human chromosomes are identified by Arabic numerals (1, 13 to 15, 20 to 22), and X and Y identifies the sex chromosomes; gen identifies total human genomic DNA. Only hybrids giving an amplification product are shown. (B) Phylogenetic analysis of the short (chr 1, 13, 15, 20B, and 21; AY546236, AY546237, AY546239, AY546241, and AY546242) and long (chr 14, 20A, 22, X, and Y; AY546238, AY546240, AY546244, and AY546245) BSR regions from different human chromosomes. The short and long BSR regions comprise, respectively, 100 and 170 bp of beta satellite and 261 and 259 bp of flanking DNA
Sequencing (AY546236 to AY546245) and alignment of the PCR products for the two groups of chromosomes confirmed that the 450-bp and 520-bp bands differ for the addition of one BSR unit (see Supplementary Material online). Figure 5B shows the phylogenetic analysis of both chromosome groups. Chromosomes 1 and Y were the most divergent, and the acrocentric chromosomes showed a very good level of similarity. In particular, chromosomes 14 and 22 shared the same sequence, and chromosomes 13, 15, and 21 showed a percentage of similarity ranging from 97.5 to 99.4.
Discussion
We report studies on the evolutionary history of beta satellite DNA in primates. The analysis of a substantial number of orangutan genomic inserts gave no evidence of large arrays of beta satellite repeats. Instead, the bulk of orangutan BSR sequences were found organized as very short stretches of approximately 100 to 170 bp, embedded in a longer DNA segment duplicated in the orangutan genome. These results give support to previous data that showed absence of FISH signals on orangutan chromosome spreads hybridized with a human beta satellite probe. BSR sequences are only detectable by FISH in African great apes and humans (Hirai, Taguchi, and Godwin 1999). Indeed, we have detected in these species clusters of BSR sequences organized in tandemly repeated 68-bp units (Meneveri et al. 1995). We have now documented an unpredicted step-by-step evolutionary history of a satellite sequence, in which the first step, in the orangutan, consisted in the spreading of an ancestral duplicon that contained the BSR as unique or as a few tandemly repeated copies. In the orangutan, which, as stated, is the first primate species in which we were able to document the spreading of BSR, the estimated copy number of the duplicon carrying short stretches of the repeat ranges from 70 to 100 per haploid genome, distributed on at least seven chromosomes. In the orangutan, as well as in humans, the duplication spans for approximately 60 to 80 kb.
The ancestral locus of the duplicon, found in a chromosomal region at the boundary of the rDNA on the marker chromosome (chromosome 13) of the macaque, does not carry any BSR subsequently found in the gibbon and orangutan. Indeed, Southern analysis and BAC library screening of the macaque were negative for the BSR sequence. Conversely, beta satellite sequences were detected within the duplicon in the genome of the gibbon. Similarly to the macaque, also in the gibbon, the duplicon mapped to a single chromosomal location, at the boundary of the secondary constriction bearing rDNA on chromosome 12 (marker chromosome). Furthermore, the comparison between the FISH signal intensity in the macaque and the gibbon suggested the occurrence in the latter species of the initial duplication of the sequence.
Thus, the ancestral copy of the duplicon appeared in OWMs (<35 MYA), whereas the prototype of beta satellite repeats and their initial amplification took place in a gibbon ancestor, after apes/OWM divergence (25 MYA). Subsequently, a burst in spreading of the duplicon carrying the beta satellite was observed in the orangutan, after lesser apes divergence from the great apes–humans lineage (18 MYA).
The analysis of the orangutan genome also indicated the existence of two variants of the duplication, differing for the length of inserted beta satellite repeats. The shorter repeat organization is 100 bp, whereas the longer one is 170 bp. The latter organization was probably generated by nonhomologous recombination between two 100-bp repeated regions. Likely, this event led to the duplication of the single Sau3A site present in the 100-bp variant, generating, in this way, the Sau3A 68-bp repetition unit of beta satellite sequences. The longer beta satellite organization seems to be absent in the gibbon, but it characterizes the hominoid genomes from orangutan to humans. In the orangutan, the ratio between the 100-bp and 170-bp repeated stretches is in great favor of the short one. Conversely, this ratio is reversed in the gorilla, which is also characterized by two large arrays of beta satellite repeats on chromosomes IV and Y (Hirai, Taguchi, and Godwin 1999). The large arrays of beta satellite originated, therefore, in the African apes–humans ancestor, after the divergence from the orangutan (14 MYA). The generation of the beta satellite clustered organization can be hypothesized to have occurred by unequal crossing-over during meiosis or by distinct mechanisms acting internally to the duplicons carrying the initial BSR seeds. If this hypothesis is correct, one would expect in African great apes a certain extent of colocalization of clustered BSR and these duplicons. Data from the gorilla, however, does not support this scenario. The analysis of DNA sequences flanking the beta satellite clusters in the gorilla could allow the definition of the more plausible evolutionary hypothesis. Conversely, the comparison of chimpanzee/human location of BSR showed a consistent distribution of both clustered and interspersed repeats, which further supporting their greater closeness in respect to the gorilla (Caccone and Powell 1989; Ruvolo 1997).
During primate evolution, the distribution of BSR duplicons involved the marker chromosome in the macaque and the gibbon, and preferentially the acrocentrics in the hominoid species. This distribution strongly paralleled the evolutionary dispersal of rDNA clusters (Hirai, Taguchi, and Godwin 1999). Conversely, human chromosomes 20 and Y acquired the BSR duplication, respectively, before gorilla and chimpanzee divergence. Location consistency between duplications bearing BSR and rDNA clusters, with potential functionality implications, is intriguing (Horvath et al. 2001). Evolutionary comparison among sequences (when available) (Eichler and DeJong 2002) of these chromosomal regions will probably elucidate the possible role of these duplications.
As derived by the comparison between FISH and PCR data, human chromosomes 1 and X carry a very reduced length of the duplication, characterized by the occurrence of 100 and 170 bp of BSR, respectively. The two BSR variants showed also distinct patterns of distribution in human acrocentric chromosomes. Chromosomes 13, 15, and 21 contain a very similar 100-bp stretch of beta satellite, whereas chromosomes 14 and 22 carry the same 170-bp version. Acrocentric short arm sequence homogenization has been hypothesized to be a consequence of their physical association at meiotic prophase and somatic interphase (Schweizer and Loidl 1987). However, this does not seem to occur for the identified duplication, as well as for other subclasses of repetitive sequences (i.e., satIII), proposed to be at the basis of preferential dicentric Robertsonian translocations among chromosomes 13, 14, and 21 (Sullivan et al. 1996; Bandyopadhyay et al. 2001). These results suggest the existence of constrains against a full homogenization of the DNA sequences within the short arm of acrocentrics. In this respect, chromosomes 13 and 21, as well as 14 and 22 share high homologous subsets of alphoid sequences (Choo, Vissel, and Earle 1989). In addition, large paralogous regions are shared between these same couples of chromosomes (13/21 and 14/22) (Bailey et al. 2002b; Golfier et al. 2003).
In conclusion, we have delineated the evolutionary history of beta satellite sequences, which began in a lesser apes ancestor as a low-copy, or nonduplicated, BSR sequence. After lesser apes/great apes divergence, we observed a burst of BSR spreading as part of a duplicon. In a second burst dispersal, which occurred after orangutan/gorilla, chimpanzee, and human divergence, the BSR acquired the basic features of classical satellite DNA. The specific causes that triggered the two kinds of events are not clear. The only hypothesized mechanism of duplicon spreading points to Alu sequences because they have been seen significantly over-represented at duplicon junctions (Bailey et al. 2003). The preferential location of BSR in pericentromeric regions suggests that centromere-specific mechanisms are involved. This conclusion is indirectly supported by the fact that inactivated centromeres tend to rapidly loose all satellite DNAs (Eder et al. 2003; Ventura et al. 2003). On the contrary, newly seeded centromeres have rapidly acquired the complex organization of normal centromeres that feature large blocks of satellite DNA (Ventura et al. 2003). At moment, however, we have a very poor understanding of forces that drive these processes.
Acknowledgements
This work was supported by grants from Ministero della Pubblica Istruzione (Cofin 2002 and Cluster C03, Prog. L.488/92), CEGBA (Centro di Eccellenza Geni in campo Biosanitario e Agroalimentare), and European Commission (INPRIMAT, QLRI-CT-2002–01325).
Literature Cited
Agresti, A., G. Rainaldi, A. Lobbiani, I. Magnani, R. Di Lernia, R. Meneveri, A. G. Siccardi, and E. Ginelli. 1987. Chromosomal location by in situ hybridization of the human Sau3A family of DNA repeats. Hum. Genet. 75:326-332.
Agresti, A., R. Meneveri, A. G. Siccardi, A. Marozzi, G. Corneo, S. Gaudi, and E. Ginelli. 1989. Linkage in human heterochromatin between highly divergent Sau3A repeats and a new family of repeated DNA sequences (HaeIII family). J. Mol. Biol. 205:625-631.
Bandyopadhyay, R., C. McQuillan, S. L. Page, K. A. H. Choo, and L.G. Shaffer. 2001. Identification and characterization of satellite III subfamilies to the acrocentric chromosomes. Chrom. Res. 9:223-233.
Bailey, J. A., Z. Gu, R. A. Clark, K. Reinert, R. V. Samonte, S. Schwartz, M. D. Adams, E. W. Myers, P. W. Li, and E. E. Eichler. 2002a. Recent segmental duplications in the human genome. Science 297:1003-1007.
Bailey, J. A., G. Liu, and E. E. Eichler. 2003. An Alu transposition model for the origin and expansion of human segmental duplications. Am. J. Hum. Genet. 73:823-834.
Bailey, J. A., A. M. Yavor, L. Viggiano, D. Misceo, J. E. Horvath, N. Archidiacono, S. Schwartz, M. Rocchi, and E. E. Eichler. 2002b. Human-specific duplication and mosaic transcripts: the recent paralogous structure of chromosome 22. Am. J. Hum. Genet. 70:83-100.
Bakker, E., C. Wijmenga, R. H. Vossen, G. W. Padberg, J. E. Hewitt, M. van der Wielen, K. Rasmussen, and R. R. Frants. 1995. The FSHD-linked locus D4F104S1 (p13E-11) on 4q35 has a homologue on 10qter. Muscle Nerve 2:S39-S44.
Ballarati, L., I. Piccini, L. Carbone, N. Archidiacono, A. Rollier, A. Marozzi, R. Meneveri, and E. Ginelli. 2002. Human genome dispersal and evolution of 4q35 duplications and interspersed LSau repeats. Gene 296:21-27.
Breen, M., B. Arveiler, I. Murray, J. R. Gosden, and D. J. Porteous. 1992. YAC mapping by FISH using Alu-PCR-generated probes. Genomics 13:726-730.
Caccone, A., and J. R. Powell. 1989. DNA divergence among hominoids. Evolution 43:926-942.
Choo, K. H., B. Vissel, and E. Earle. 1989. Evolution of alpha-satellite DNA on human acrocentric chromosomes. Genomics 5:332-344.
Clark, L. N., U. Koehler, D. C. Ward, J. Wienberg, and J. E. Hewitt. 1996. Analysis of the organization and localization of the FSHD-associated tandem array in primates: implications for the origin and evolution of the 3.3 kb repeat family. Chromosoma 105:180-189.
Eder, V., M. Ventura, M. Ianigro, M. Teti, M. Rocchi, and N. Archidiacono. 2003. Chromosome 6 phylogeny in primates and centromere repositioning. Mol. Biol. Evol. 20:1506-1512.
Eichler, E. E., and P. J. DeJong. 2002. Biomedical applications and studies of molecular evolution: a proposal for a primate genomic library resource. Genome Res. 12:673-678.
Golfier, G., F. Chibon, A. Aurias, X. N. Chen, J. Korenberg, J. Rossier, and M. C. Potier. 2003. The 200-kb segmental duplication on human chromosome 21 originates from a pericentromeric dissemination involving human chromosomes 2, 18 and 13. Gene 312:51-59.
Hewitt, J. E., R. Lyle, and L. N. Clark, et al. (12 co-authors). 1994. Analysis of the tandem repeat locus D4Z4 associated with facioscapulohumeral muscular dystrophy. Hum. Mol. Genet. 3:1287-1295.
Hirai, H., T. Taguchi, and A. K. Godwin. 1999. Genomic differentiation of 18S ribosomal DNA and beta satellite DNA in the hominoid and its evolutionary aspects. Chromosome Res. 7:531-540.
Horvath, J. E., J. A. Bailey, B. J. Loftus, D. P. Locke, and E.E. Eichler. 2001. Lessons from the human genome: transition between euchromatin and heterochromatin. Hum. Mol. Genet. 10:2215-2223.
Horvath, J. E., L. Viggiano, B. J. Loftus, M. D. Adams, N. Archidiacono, M. Rocchi, and E. E. Eichler. 2000. Molecular structure and evolution of an alpha satellite/non-alpha satellite junction at 16p11. Hum. Mol. Genet. 9:113-123.
Lemmers, R. J., P. de Jievit, L. Sandkuijl, G. W. Padberg, G. B. van Ommen, R. R. Frants, and S. M. van der Maarel. 2002. Facioscapulohumeral muscular dystrophy is uniquely associated with one of the variants of the 4q subtelomeres. Nat. Genet. 32:235-236.
Lyle, R., T. J. Wright, L. N. Clark, and J. E. Hewitt. 1995. The FSHD-associated repeat, D4Z4, is a member of a dispersed family of homeobox-containing repeats, subsets of which are clustered on the short arms of the acrocentric chromosomes. Genomics 28:389-397.
Meneveri, R., A. Agresti, G. Della Valle, D. Talarico, A. G. Siccardi, and E. Ginelli. 1985. Identification of a human clustered GC-rich DNA family of repeats (Sau 3A family). J. Mol. Biol. 186:483-489.
Meneveri, R., A. Agresti, A. Marozzi, S. Saccone, M. Rocchi, N. Archidiacono, G. Corneo, G. Della Valle, and E. Ginelli. 1993. Molecular organization and chromosomal location of human GC-rich heterochromatic blocks. Gene 123:227-234.
Meneveri, R., A. Agresti, M. Rocchi, A. Marozzi, and E. Ginelli. 1995. Analysis of GC-rich repetitive nucleotide sequences in great apes. J. Mol. Evol. 40:405-412.
Ruvolo, M. 1997. Molecular phylogeny of the hominoids: inference from multiple independent DNA sequences data sets. Mol. Biol. Evol. 14:248-265.
Samonte, R. V., and E. E. Eichler. 2002. Segmental duplications and the evolution of the primate genome. Nat. Rew. Genet. 3:65-72.
Schweizer, D., and J. Loidl. 1987. A model for heterochromatin dispersion and evolution of C-band patterns. Chromosome Today 9:61-74.
Sullivan, B. A., L. S. Jenkins, E. M. Karson, J. Leana-Cox, and S. Schwartz. 1996. Evidence for structural heterogeneity from molecular cytogenetic analysis of dicentric Robertsonian translocations. Am. J. Hum. Genet. 59:167-175.
Ventura, M., J. M. Mudge, and V. Palumbo, et al. (11 co-authors). 2003. Neocentromeres in 15q24–26 map to duplicons which flanked an ancestral centromere in 15q25. Genome Res. 13:2059-2068.
Willard, H. F. 1991. Evolution of alpha satellite. Curr. Opin. Genet. Dev. 1:509-514.
Winokur, S. T., U. Bengtsson, J. C. Vargas, J. J. Wasmuth, and M. R. Altherr. 1996. The evolutionary distribution and structural organization of the homeobox-containing repeat D4Z4 indicates a functional role for the ancestral copy in the FSHD region. Hum. Mol. Genet. 10:1567-1575.(M. F. Cardone*, L. Ballar)
Dipartimento di Biologia e Genetica per le Scienze Mediche, Milano, Italy
Dipartimento di Medicina Sperimentale, Ambientale e Biotecnologie Mediche, Monza, Italy
E-mail: raffaella.meneveri@unimib.it.
Abstract
In this article, we report studies on the evolutionary history of beta satellite repeats (BSR) in primates. In the orangutan genome, the bulk of BSR sequences was found organized as very short stretches of approximately 100 to 170 bp, embedded in a 60-kb to 80-kb duplicated DNA segment. The estimated copy number of the duplicon that carries BSR sequences ranges from 70 to 100 per orangutan haploid genome. In both macaque and gibbon, the duplicon mapped to a single chromosomal region at the boundary of the rDNA on the marker chromosome (chromosome 13 and 12, respectively). However, only in the gibbon, the duplicon comprised 100 bp of beta satellite. Thus, the ancestral copy of the duplicon appeared in Old World monkeys (25 to 35 MYA), whereas the prototype of beta satellite repeats took place in a gibbon ancestor, after apes/Old World monkeys divergence (25 MYA). Subsequently, a burst in spreading of the duplicon that carries the beta satellite was observed in the orangutan, after lesser apes divergence from the great apes–humans lineage (18 MYA). The analysis of the orangutan genome also indicated the existence of two variants of the duplication that differ for the length (100 or 170 bp) of beta satellite repeats. The latter organization was probably generated by nonhomologous recombination between two 100-bp repeated regions, and it likely led to the duplication of the single Sau3A site present in the 100-bp variant, which generated the prototype of Sau3A 68-bp beta satellite tandem organization. The two variants of the duplication, although with a different ratios, characterize the hominoid genomes from the orangutan to humans, preferentially involving acrocentric chromosomes. At variance to alpha satellite, which appeared before the divergence of New World and Old World monkeys, the beta satellite evolutionary history began in apes ancestor, where we have first documented a low-copy, nonduplicated BSR sequence. The first step of BSR amplification and spreading occurred, most likely, because the BSR was part of a large duplicon, which underwent a burst dispersal in great apes' ancestor after the lesser apes' branching. Then, after orangutan divergence, BSR acquired the clustered structural organization typical of satellite DNA.
Key Words: beta satellite ? duplication ? primate evolution
Introduction
About 5% of our genome is composed of duplicated sequences (duplicons), which consist of DNA stretches ranging from a few to hundreds of kb in length, with 90% to 100% sequence identity (Samonte and Eichler 2002). Studies on the evolutionary history of duplicons have shown that the vast majority have emerged recently during primate evolution, after the split of Old World monkeys (OWMs) and hominoid lineage (<35 MYA) (Bailey et al. 2002a; Bailey, Liu, and Eichler 2003).
Duplicons can be grouped as intrachromosomal and interchromosomal, and the latter tends to be enriched in pericentrometic and subtelomeric regions. Pericentromeric duplications usually span the transition from euchromatic genes on the short/long arm to pericentromeric satellites/centromeric alpha satellites (Horvath et al. 2000). However, despite their location, the evolutionary history of alpha satellite appears distinct from that of the pericentromeric duplicons (Willard 1991).
Beta satellite repeats (BSR) represent another family of sequences that show a predominant heterochromatic distribution, which includes the short arm of acrocentric chromosomes and chromosomes 1q12, 3q12, 9q12, and Yq11 (Meneveri et al. 1985,1993; Agresti et al. 1987, 1989). In these regions, beta satellite arrays are in linkage with LSau repeats, which are part of complex repetition units of 3.3 kb in length (D4Z4-like repeats) (Agresti et al. 1989; Hewitt et al. 1994; Lyle et al. 1995; Ballarati et al. 2002). D4Z4 repetitions and BSRs are also localized on chromosomes 4q35 and 10q26 (Bakker et al. 1995; Lemmers et al. 2002).
Evolutionary studies suggest that 4q35 represents the ancestral locus for the D4Z4 sequence and that an extensive radiation of the region occurred after the divergence of Old World monkeys and hominoids (Clark et al. 1996; Winokur et al. 1996; Ballarati et al. 2002). In this respect, a few studies have been devoted to investigating the evolution of beta satellite sequences. The origin of this class of repeats can be tentatively traced back to the orangutan (PPY), where the beta satellite organization appears different from humans (HSA), chimpanzee (PTR), and gorilla (GGO) (Meneveri et al. 1995; Hirai, Taguchi, and Godwin 1999).
To further investigate this point, we screened two different orangutan genomic libraries in search of BSR-positive DNA sectors. All the gathered results indicate that BSR originated in an early hominoid ancestor and that the initial step of BSR amplification and spreading was mediated by the insertion of very short stretches of the repeat into a long genomic region, which underwent duplicative transpositions.
Materials and Methods
Genomic Library Screening
Genomic libraries were obtained from RZPD, German Resource Center for Genome Research (PPY cosmid library n.141) and BACPAC Resources, Children's Hospital, Oakland Research Institute (PPY, CHORI-253 BAC library, segment 1; MMU, CHORI250 BAC library, segment 1). High-density arrayed cosmid and BAC filters were hybridized at 55°C in standard hybridization solution for 14 to 16 h with 32P-labeled A17a human probe that contained five repetition units of beta satellite (Agresti et al. 1987). Filters were washed twice in 2xSSC (1xSSC = 150 mM NaCl, 15 mM sodium citrate, pH 7.0) at room temperature and twice in 0.1xSSC at 65°C for 20 min. Positive cosmid and BAC clones were grown in LB medium supplemented with kanamycin or chloramphenicol, and DNA was purified by the Qiagen miniprep kit (Qiagen).
Southern Blot Hybridization
Cosmid and BAC DNAs were digested with Sau3A restriction enzyme (Biolabs), fractionated by agarose gel electrophoresis, blotted onto nylon filters (Hybond N+; Amersham), and hybridized with 32P-labeled A17a probe. Molecular hybridizations were in 2xSSC at 60°C overnight, and filters were washed at 60°C in 1xSSC, twice for 20 min. Hybridization signals were quantified by phosphoimaging (Typhoon; Amersham).
Polymerase Chain Reaction and Sequencing
Genomic DNA was obtained from human placenta, lymphoblastoid, and fibroblast primate cell lines (chimpanzee, Pan troglodytes [PTR]; gorilla, Gorilla gorilla [GGO]; orangutan, Pongo pygmaeus [PPY]; lar gibbon, Hylobates lar [HLA]; rhesus monkey, Macaca mulatta [MMU]) (http://www.biologia.uniba.it/rmc/) and from a human monochromosomal somatic cell hybrid panel (http://www.biologia.uniba.it/rmc/) by standard methods.
Polymerase chain reaction (PCR) amplifications with the different DNA templates were carried out by the primer pairs BSR forw (5'-AGGGGCTTTATCCTCATTTCACAA-3') and rev (5'-GGCCTCCATATTCCCTAACTTCC-3'); PPY2 forw (5'-TGAATTTTAGCACCCACAA-3') and rev (5'-ATTCGATTCAACCCCAGTTA-3'); PPY4 forw (5'-AGTCACTGTGGCGATTCTTCTA-3') and rev (5'-GTGTGATGTCCCCTTTCCTG-3'); PPY14 forw (5'-ATGGGCCAGGAGCTATTCACAG-3') and rev (5'-CGCAACGCCCCCTTCAACC-3') and by Alu-PCR (Breen et al. 1992).
Primer pairs BSR and PPY2, PPY4, and PPY14 were derived, respectively, from the human genomic DNA sequence AC006987 (Yp11.2) and from sequences obtained by Alu-PCR (cosmids O05112Q2 and H1653Q2) and T3 end (cosmid N1981Q2) of BSR-positive PPY cosmids. For inter-Alu and T3/T7 ends, only one strand was sequenced.
PCR was carried out in a final volume of 25 μl that contained 0.2 mM dNTPs, 1 μM each primer, and 1U Taq polymerase in standard reaction buffer (Sigma). For amplification denaturation was at 95°C for 1 min, and annealing was as follows: BSR (60°C), PPY2 (46°C), PPY4 (54°C), and PPY18 (54°C), for 1 min, with extension 72°C for 1 min. The total number of cycles was of 30 to 35. PCR products were analyzed by 1% agarose gel electrophoresis, subcloned into the pGEM-T easy vector (Promega), and sequenced on both strands by the Big Dye terminator system (Perkin-Elmer Applied Biosystems) following the instructions of the vendor.
All fluorescent traces were analyzed on the Applied Biosystem Model 3100 DNA Sequencing System (PerkinElmer Applied Biosystems). DNA sequence analysis was performed by DNASTAR software and by NCBI facilities (www.ncbi.nlm.nih.gov). New sequence data from this article are available at http://www.ncbi.nlm.nih.gov/, with accession numbers AY546195 to AY546245.
Fluorescence in situ Hybridization
Metaphase spreads were obtained by standard methods from peripheral blood lymphocytes of normal human donors and from lymphoblastoid or fibroblast primate cell lines. The probes were directly labeled with Cyanine 3-dUTP by nick-translation, according to the protocol of the vendor (Perkin-Elmer), and hybridized overnight to chromosomal preparations. Hybridizations were in 50% formamide (v/v), 10% dextran sulfate, SSC at 37°C, in the presence of Cot1 human DNA (Gibco-BRL). Posthybridization washing was in 50% formamide, SSC at 42°C, followed by three washes in SSC at 60°C (HSA), and in 50% formamide, SSC at 37 C, followed by three washes in SSC at 42°C (PTR, GGO, PPY, HLA, and MMU). Chromosomes were stained with DAPI (4',6-diamidino-2-phenylindole). Digital images were captured by a Leica DMRXA epifluorescence microscope equipped with a cooled CCD camera (Princeton Instruments). Cy3 (red) and DAPI (blue) fluorescence signals were recorded separately as gray-scale images. Pseudocoloring and merging of images were performed by use of Adobe Photoshop software.
Results
Organization of Beta Satellite Sequences in the Orangutan Genome
Cosmid and BAC orangutan genomic libraries (see Materials and Methods) were screened using a human BSR probe (A17a). Approximately 100 and 70 positive recombinant clones per haploid genome were, respectively, isolated (not shown). After Sau3A digestion and hybridization with the BSR probe, 35 randomly chosen BSR-positive clones (22 cosmids and 13 BACs) showed a single Sau3A-hybridizing band of approximately 0.6 kb (fig. 1A), as shown in total orangutan genomic DNA (Meneveri et al. 1995). None of the analyzed clones showed the release of the ladder of bands, peculiar of the bulk of human BSR clustered organization (Meneveri et al. 1995).
FIG. 1. Analysis of BSR-positive PPY genomic clones. (A) Examples of Southern blot hybridization of the human BSR–labeled probe (A17a) to Sau3A-digested DNA from cosmids (N17130Q2, N1981Q2, G1478Q3; a–c) and BACs (162K8, 1O23, 25F18; d–f) PPY genomic clones, after fractionation by 1% agarose gel electrophoresis. (B) Schematic diagram of the sequence similarity between a 0.6-kb BSR-positive PPY fragment (AY546195) of figure 1A and human entries AC006987 (chromosome Yp11.2) and AC011850 (chromosome 15). Rectangles identify the BSR region and the heavy line flanking BSR indicates unrelated sequences. Arrows indicate the position of a primer pair (BSR) flanking the BSR region in the human genome, which by PCR gives amplification bands of 450 and 520 bp on chromosomes 15 and Yp11.2, respectively. (C) Examples of PCR amplification by the BSR primer pair of BSR-positive PPY cosmids O05112Q2 and N17130Q2 that contain, respectively, the long (a) and short (b) BSR region, and of PPY genomic DNA (c). PCR products were fractionated by 1% agarose gel electrophoresis
The 0.6-kb Sau3A DNA fragment from cosmid N17130Q2 was subcloned and sequenced (AY546195). The sequence analysis revealed a 70-bp BSR sequence in one side and unrelated DNA on the other side. Data bank comparison of the 0.6-kb DNA sequence disclosed similarity to several human clones that included AC079860 (1p36.3), AC006987 (Yp11.2), and unmapped clones from chromosomes 4, 15, and 16 (AC129664, AC011850, and AC137488, respectively). Interestingly, in these entries 170 bp (chromosomes 16 and Yp11.2) or 100 bp (chromosomes 1p36.3, 4, and 15) of BSR are flanked at both sides by unrelated DNA (fig. 1B).
A primer pair (BSR forw and rev) from outside both repeat regions and defining 520 and 450 bp of DNA on chromosomes Yp11.2 and 15, respectively (fig. 1B), was then used to analyze a total of 51 BSR-positive orangutan genomic clones (27 cosmids and 24 BACs) and orangutan genomic DNA. Only two clones gave a band of 520 bp (long BSR [fig. 1C]), whereas the great majority of the analyzed clones were positive for the 450-bp band (short BSR [fig. 1C]). The same PCR reaction on orangutan genomic DNA showed the simultaneous presence of both PCR bands, with a ratio in favor of the light band (fig. 1C).
The alignment of one long BSR sequence with the consensus derived from nine short BSR DNA sequences (figure 2A and Supplementary Materials online) revealed an organization comparable to that derived from chromosomes Yp11.2 and 15: 170 and 100 bp of BSR were embedded within unrelated DNA. Furthermore, only within the 170-bp BSR region, two Sau3A sites define a 68-bp beta satellite unit. The phylogenetic analysis of the nine short BSR sequences clearly indicated the occurrence of sequence divergence, which suggests their derivation from different orangutan genomic regions (figure 2B and Supplementary Materials online). The 100-bp and 170-bp orangutan BSR regions showed a similarity of approximately 94% and 92% with a human BSR consensus sequence (not shown).
FIG. 2. Sequence analysis of the BSR region in PPY genomic clones. (A) Alignment of the PPY genomic regions carrying 170 bp (long BSR) and 100 bp (short BSR) of beta satellite repeats (in bold) and 245 bp of unrelated flanking DNA. The sequences were derived by PCR with the BSR primer pair on PPY cosmid and BAC genomic clones. The short BSR sequence represents the consensus derived by aligning a core of 345 bp out of 450 bp from four cosmids N17130Q2 (AY546197), G1478Q3 (AY546202), H1653Q2 (AY546206), and N1981Q2 (AY546211), and five BACs 162K8 (AY546219), 1O23 (AY546223), 25F18 (AY546227), 60F17 (AY546231), and 98A13 (AY546235), whereas the long BSR sequence (416 bp) was derived from a single 520-bp PCR product (cosmid O05112Q2; accession number AY546215). Periods and hyphens indicate, respectively, base identities and deletions. Within the 170-bp BSR region, Sau3A sites are underlined and the 68-bp Sau3A unit defined by arrows. (B) Phylogenetic analysis of nine DNA sequences carrying the 100 bp of BSR. The sequences from cosmids (cos) and BACs (BAC) PPY genomic clones were those utilized to derive the consensus of the short BSR region shown in (A)
Further insights into the sequence environment in which orangutan beta satellite sequences are embedded were derived by the partial sequencing (inter-Alu and T3/T7 end sequences) of five cosmid clones (fig. 3A). Besides the BSR region, the different clones contained other redundant as well as unique sequences. In particular, 15 nonredundant sequences out of 39 aligned to a region of approximately 50 kb on chromosomes Yp11.2 (fig. 3A) and 15 (not shown). Database analysis revealed that the identified DNA region is part of a larger duplicated sequence of about 80 kb on human chromosomes Yp11.2, 15, and 16 and of approximately 60 and 6 kb, respectively, on chromosomes 4 and 1p36.3 (not shown).
FIG. 3. BSR-unrelated sequence analysis of PPY clones carrying the BSR region. (A) Alignment of 15 nonredundant inter-Alu (IA) and T3/T7 end sequences (T3, or T7) and of the BSR region from five BSR-positive PPY cosmids to a DNA sequence (AC006987) from chromosome Yp11.2. PPY sequences were derived from the cosmids N1981Q2 (AY546207 to AY546210), O05112Q2 (AY546212 to AY546214), G1478Q3 (AY546198 to AY546201), H1653Q2 (AY546203 to AY546205), and N17130Q2 (AY546196). PPY2, PPY4, PPY14, and BSR identify, respectively, the location of sequences used to derive three primer pairs that span the DNA region and of the region carrying 170 bp of BSR. (B) Paralogous PCR carried out on PPY BAC 1O23 (1), 162K8 (2), 98A13 (3), 25F18 (4), and 60F17 (5) with the three (PPY2, PPY4, and PPY14) primer pairs identified in (A) and with the BSR primer pair (BSR). PCR products were fractionated by 1% agarose gel electrophoresis
Using three primer pairs derived from the orangutan sequences that spanned the duplication (labeled as PPY2, PPY4, and PPY14 in figure 3A) and the BSR primer pair, we then analyzed by paralogous PCR five BACs (fig. 3B). All the BACs exhibited the same set of PCR products. Sequencing (AY546216 to AY546218, AY546220 to AY546222, AY546224 to AY546226, AY546228 to AY546230, and AY546232 to AY546234) and alignment of the amplification bands (not shown) demonstrated that the BACs, as the analyzed cosmids, are characterized by the same repertoire of divergent sequences (for the BSR region, [see figure 2B]). These results strongly indicated the occurrence in the orangutan genome, as in man, of duplications of this nucleotide region.
Evolution of the Duplication Carrying the BSR Region
Five BACs (162K8, 1O23, 25F18, 60F17, and 98A13) were used as probes on primate chromosome spreads (HSA, PTR, GGO, PPY, HLA, and MMU). An example of comparative FISH experiments is shown in figure 4, and a summary of the chromosome location obtained by all clones is reported in table 1. On the macaque and the gibbon, the probes highlighted only the "marker chromosome," at the level of the secondary constriction bearing the rDNA array (fig. 4A [MMU and HLA]). Furthermore, the intensity of the hybridization signals was constantly higher in the gibbon than in the macaque. Conversely, the other analyzed species (HSA, PTR, GGO, and PPY) showed a multiple chromosome location that included pericentromeric regions and the short arm of acrocentric chromosomes (fig. 4A and table 1). In addition to the macaque and the gibbon, the other analyzed primates also showed almost all the chromosomal regions bearing the rDNA labeled by the duplication. The above FISH experiments, however, although clearly demonstrating the spreading of the duplication, are not proof that BSR sequences were also involved. A human clone that contained only BSR sequences was then used in FISH experiments on orangutan, gibbon, and macaque chromosome spreads. The negative results (not shown) suggested that in these species, BSR sequences were either absent or below FISH resolution. To further investigate this point, we performed a PCR assay with the BSR primer pair on primate genomic DNAs (fig. 4B). A band of 450 bp was detected in great apes and in the gibbon but not in the macaque. In addition, as already observed for the orangutan (see figure 1C), humans, chimpanzee, and gorilla also showed an additional band of approximately 520 bp. The ratio of the two bands varied in the different species. In the orangutan, the shorter band was predominant, whereas the opposite was found in the gorilla. Conversely, chimpanzee and humans showed comparable amounts of the 450-bp and 520-bp bands. The 450-bp and 520-bp bands yielded positive results when hybridized with a human BSR probe (not shown). Furthermore, we performed with the human BSR probe at low-stringency conditions, the Southern blot analysis of macaque genomic DNA and the screening of a macaque BAC library (not shown). In both cases, no hybridization signals were detected, which further supports the absence of BSR sequences in the macaque genome.
FIG. 4. Evolution of the duplicon carrying the BSR region. (A) Comparative FISH on chromosome spreads from human (HSA), great apes (PTR, GGO, and PPY), lower ape (HLA), and Old World monkey (MMU) by one BSR-positive PPY BAC (60F17) as a probe. Except for MMU and HLA, chromosomes are indicated by the phylogenetic nomenclature (roman numerals). (B) PCR assay with the BSR primer pair on genomic DNA from primate species (HSA, PTR, GGO, PPY, HLA, and MMU). PCR products were fractionated by 1% agarose gel electrophoresis
Table 1 Summary of FISH Experiments Performed on Chromosome Spreads from Primate Species with Several PPY Clones Carrying Beta Satellite Repeats Embedded into a Duplication.
Human Chromosome Distribution of the Short and Long BSR Regions
A panel of human monochromosomal somatic cell hybrids was amplified with the BSR primer pair (fig. 5A). Chromosomes 1, 13, 15, and 21 showed the shorter version of the region (450 bp), whereas chromosomes 14, 22, X, and Y exhibited the longer version (520 bp). Only chromosome 20 showed the simultaneous occurrence of the two amplification bands.
FIG. 5. Human chromosome distribution of the short and long BSR regions. (A) PCR assay with the BSR primer pair on DNA from a panel of human monochromosomal cell hybrids. Human chromosomes are identified by Arabic numerals (1, 13 to 15, 20 to 22), and X and Y identifies the sex chromosomes; gen identifies total human genomic DNA. Only hybrids giving an amplification product are shown. (B) Phylogenetic analysis of the short (chr 1, 13, 15, 20B, and 21; AY546236, AY546237, AY546239, AY546241, and AY546242) and long (chr 14, 20A, 22, X, and Y; AY546238, AY546240, AY546244, and AY546245) BSR regions from different human chromosomes. The short and long BSR regions comprise, respectively, 100 and 170 bp of beta satellite and 261 and 259 bp of flanking DNA
Sequencing (AY546236 to AY546245) and alignment of the PCR products for the two groups of chromosomes confirmed that the 450-bp and 520-bp bands differ for the addition of one BSR unit (see Supplementary Material online). Figure 5B shows the phylogenetic analysis of both chromosome groups. Chromosomes 1 and Y were the most divergent, and the acrocentric chromosomes showed a very good level of similarity. In particular, chromosomes 14 and 22 shared the same sequence, and chromosomes 13, 15, and 21 showed a percentage of similarity ranging from 97.5 to 99.4.
Discussion
We report studies on the evolutionary history of beta satellite DNA in primates. The analysis of a substantial number of orangutan genomic inserts gave no evidence of large arrays of beta satellite repeats. Instead, the bulk of orangutan BSR sequences were found organized as very short stretches of approximately 100 to 170 bp, embedded in a longer DNA segment duplicated in the orangutan genome. These results give support to previous data that showed absence of FISH signals on orangutan chromosome spreads hybridized with a human beta satellite probe. BSR sequences are only detectable by FISH in African great apes and humans (Hirai, Taguchi, and Godwin 1999). Indeed, we have detected in these species clusters of BSR sequences organized in tandemly repeated 68-bp units (Meneveri et al. 1995). We have now documented an unpredicted step-by-step evolutionary history of a satellite sequence, in which the first step, in the orangutan, consisted in the spreading of an ancestral duplicon that contained the BSR as unique or as a few tandemly repeated copies. In the orangutan, which, as stated, is the first primate species in which we were able to document the spreading of BSR, the estimated copy number of the duplicon carrying short stretches of the repeat ranges from 70 to 100 per haploid genome, distributed on at least seven chromosomes. In the orangutan, as well as in humans, the duplication spans for approximately 60 to 80 kb.
The ancestral locus of the duplicon, found in a chromosomal region at the boundary of the rDNA on the marker chromosome (chromosome 13) of the macaque, does not carry any BSR subsequently found in the gibbon and orangutan. Indeed, Southern analysis and BAC library screening of the macaque were negative for the BSR sequence. Conversely, beta satellite sequences were detected within the duplicon in the genome of the gibbon. Similarly to the macaque, also in the gibbon, the duplicon mapped to a single chromosomal location, at the boundary of the secondary constriction bearing rDNA on chromosome 12 (marker chromosome). Furthermore, the comparison between the FISH signal intensity in the macaque and the gibbon suggested the occurrence in the latter species of the initial duplication of the sequence.
Thus, the ancestral copy of the duplicon appeared in OWMs (<35 MYA), whereas the prototype of beta satellite repeats and their initial amplification took place in a gibbon ancestor, after apes/OWM divergence (25 MYA). Subsequently, a burst in spreading of the duplicon carrying the beta satellite was observed in the orangutan, after lesser apes divergence from the great apes–humans lineage (18 MYA).
The analysis of the orangutan genome also indicated the existence of two variants of the duplication, differing for the length of inserted beta satellite repeats. The shorter repeat organization is 100 bp, whereas the longer one is 170 bp. The latter organization was probably generated by nonhomologous recombination between two 100-bp repeated regions. Likely, this event led to the duplication of the single Sau3A site present in the 100-bp variant, generating, in this way, the Sau3A 68-bp repetition unit of beta satellite sequences. The longer beta satellite organization seems to be absent in the gibbon, but it characterizes the hominoid genomes from orangutan to humans. In the orangutan, the ratio between the 100-bp and 170-bp repeated stretches is in great favor of the short one. Conversely, this ratio is reversed in the gorilla, which is also characterized by two large arrays of beta satellite repeats on chromosomes IV and Y (Hirai, Taguchi, and Godwin 1999). The large arrays of beta satellite originated, therefore, in the African apes–humans ancestor, after the divergence from the orangutan (14 MYA). The generation of the beta satellite clustered organization can be hypothesized to have occurred by unequal crossing-over during meiosis or by distinct mechanisms acting internally to the duplicons carrying the initial BSR seeds. If this hypothesis is correct, one would expect in African great apes a certain extent of colocalization of clustered BSR and these duplicons. Data from the gorilla, however, does not support this scenario. The analysis of DNA sequences flanking the beta satellite clusters in the gorilla could allow the definition of the more plausible evolutionary hypothesis. Conversely, the comparison of chimpanzee/human location of BSR showed a consistent distribution of both clustered and interspersed repeats, which further supporting their greater closeness in respect to the gorilla (Caccone and Powell 1989; Ruvolo 1997).
During primate evolution, the distribution of BSR duplicons involved the marker chromosome in the macaque and the gibbon, and preferentially the acrocentrics in the hominoid species. This distribution strongly paralleled the evolutionary dispersal of rDNA clusters (Hirai, Taguchi, and Godwin 1999). Conversely, human chromosomes 20 and Y acquired the BSR duplication, respectively, before gorilla and chimpanzee divergence. Location consistency between duplications bearing BSR and rDNA clusters, with potential functionality implications, is intriguing (Horvath et al. 2001). Evolutionary comparison among sequences (when available) (Eichler and DeJong 2002) of these chromosomal regions will probably elucidate the possible role of these duplications.
As derived by the comparison between FISH and PCR data, human chromosomes 1 and X carry a very reduced length of the duplication, characterized by the occurrence of 100 and 170 bp of BSR, respectively. The two BSR variants showed also distinct patterns of distribution in human acrocentric chromosomes. Chromosomes 13, 15, and 21 contain a very similar 100-bp stretch of beta satellite, whereas chromosomes 14 and 22 carry the same 170-bp version. Acrocentric short arm sequence homogenization has been hypothesized to be a consequence of their physical association at meiotic prophase and somatic interphase (Schweizer and Loidl 1987). However, this does not seem to occur for the identified duplication, as well as for other subclasses of repetitive sequences (i.e., satIII), proposed to be at the basis of preferential dicentric Robertsonian translocations among chromosomes 13, 14, and 21 (Sullivan et al. 1996; Bandyopadhyay et al. 2001). These results suggest the existence of constrains against a full homogenization of the DNA sequences within the short arm of acrocentrics. In this respect, chromosomes 13 and 21, as well as 14 and 22 share high homologous subsets of alphoid sequences (Choo, Vissel, and Earle 1989). In addition, large paralogous regions are shared between these same couples of chromosomes (13/21 and 14/22) (Bailey et al. 2002b; Golfier et al. 2003).
In conclusion, we have delineated the evolutionary history of beta satellite sequences, which began in a lesser apes ancestor as a low-copy, or nonduplicated, BSR sequence. After lesser apes/great apes divergence, we observed a burst of BSR spreading as part of a duplicon. In a second burst dispersal, which occurred after orangutan/gorilla, chimpanzee, and human divergence, the BSR acquired the basic features of classical satellite DNA. The specific causes that triggered the two kinds of events are not clear. The only hypothesized mechanism of duplicon spreading points to Alu sequences because they have been seen significantly over-represented at duplicon junctions (Bailey et al. 2003). The preferential location of BSR in pericentromeric regions suggests that centromere-specific mechanisms are involved. This conclusion is indirectly supported by the fact that inactivated centromeres tend to rapidly loose all satellite DNAs (Eder et al. 2003; Ventura et al. 2003). On the contrary, newly seeded centromeres have rapidly acquired the complex organization of normal centromeres that feature large blocks of satellite DNA (Ventura et al. 2003). At moment, however, we have a very poor understanding of forces that drive these processes.
Acknowledgements
This work was supported by grants from Ministero della Pubblica Istruzione (Cofin 2002 and Cluster C03, Prog. L.488/92), CEGBA (Centro di Eccellenza Geni in campo Biosanitario e Agroalimentare), and European Commission (INPRIMAT, QLRI-CT-2002–01325).
Literature Cited
Agresti, A., G. Rainaldi, A. Lobbiani, I. Magnani, R. Di Lernia, R. Meneveri, A. G. Siccardi, and E. Ginelli. 1987. Chromosomal location by in situ hybridization of the human Sau3A family of DNA repeats. Hum. Genet. 75:326-332.
Agresti, A., R. Meneveri, A. G. Siccardi, A. Marozzi, G. Corneo, S. Gaudi, and E. Ginelli. 1989. Linkage in human heterochromatin between highly divergent Sau3A repeats and a new family of repeated DNA sequences (HaeIII family). J. Mol. Biol. 205:625-631.
Bandyopadhyay, R., C. McQuillan, S. L. Page, K. A. H. Choo, and L.G. Shaffer. 2001. Identification and characterization of satellite III subfamilies to the acrocentric chromosomes. Chrom. Res. 9:223-233.
Bailey, J. A., Z. Gu, R. A. Clark, K. Reinert, R. V. Samonte, S. Schwartz, M. D. Adams, E. W. Myers, P. W. Li, and E. E. Eichler. 2002a. Recent segmental duplications in the human genome. Science 297:1003-1007.
Bailey, J. A., G. Liu, and E. E. Eichler. 2003. An Alu transposition model for the origin and expansion of human segmental duplications. Am. J. Hum. Genet. 73:823-834.
Bailey, J. A., A. M. Yavor, L. Viggiano, D. Misceo, J. E. Horvath, N. Archidiacono, S. Schwartz, M. Rocchi, and E. E. Eichler. 2002b. Human-specific duplication and mosaic transcripts: the recent paralogous structure of chromosome 22. Am. J. Hum. Genet. 70:83-100.
Bakker, E., C. Wijmenga, R. H. Vossen, G. W. Padberg, J. E. Hewitt, M. van der Wielen, K. Rasmussen, and R. R. Frants. 1995. The FSHD-linked locus D4F104S1 (p13E-11) on 4q35 has a homologue on 10qter. Muscle Nerve 2:S39-S44.
Ballarati, L., I. Piccini, L. Carbone, N. Archidiacono, A. Rollier, A. Marozzi, R. Meneveri, and E. Ginelli. 2002. Human genome dispersal and evolution of 4q35 duplications and interspersed LSau repeats. Gene 296:21-27.
Breen, M., B. Arveiler, I. Murray, J. R. Gosden, and D. J. Porteous. 1992. YAC mapping by FISH using Alu-PCR-generated probes. Genomics 13:726-730.
Caccone, A., and J. R. Powell. 1989. DNA divergence among hominoids. Evolution 43:926-942.
Choo, K. H., B. Vissel, and E. Earle. 1989. Evolution of alpha-satellite DNA on human acrocentric chromosomes. Genomics 5:332-344.
Clark, L. N., U. Koehler, D. C. Ward, J. Wienberg, and J. E. Hewitt. 1996. Analysis of the organization and localization of the FSHD-associated tandem array in primates: implications for the origin and evolution of the 3.3 kb repeat family. Chromosoma 105:180-189.
Eder, V., M. Ventura, M. Ianigro, M. Teti, M. Rocchi, and N. Archidiacono. 2003. Chromosome 6 phylogeny in primates and centromere repositioning. Mol. Biol. Evol. 20:1506-1512.
Eichler, E. E., and P. J. DeJong. 2002. Biomedical applications and studies of molecular evolution: a proposal for a primate genomic library resource. Genome Res. 12:673-678.
Golfier, G., F. Chibon, A. Aurias, X. N. Chen, J. Korenberg, J. Rossier, and M. C. Potier. 2003. The 200-kb segmental duplication on human chromosome 21 originates from a pericentromeric dissemination involving human chromosomes 2, 18 and 13. Gene 312:51-59.
Hewitt, J. E., R. Lyle, and L. N. Clark, et al. (12 co-authors). 1994. Analysis of the tandem repeat locus D4Z4 associated with facioscapulohumeral muscular dystrophy. Hum. Mol. Genet. 3:1287-1295.
Hirai, H., T. Taguchi, and A. K. Godwin. 1999. Genomic differentiation of 18S ribosomal DNA and beta satellite DNA in the hominoid and its evolutionary aspects. Chromosome Res. 7:531-540.
Horvath, J. E., J. A. Bailey, B. J. Loftus, D. P. Locke, and E.E. Eichler. 2001. Lessons from the human genome: transition between euchromatin and heterochromatin. Hum. Mol. Genet. 10:2215-2223.
Horvath, J. E., L. Viggiano, B. J. Loftus, M. D. Adams, N. Archidiacono, M. Rocchi, and E. E. Eichler. 2000. Molecular structure and evolution of an alpha satellite/non-alpha satellite junction at 16p11. Hum. Mol. Genet. 9:113-123.
Lemmers, R. J., P. de Jievit, L. Sandkuijl, G. W. Padberg, G. B. van Ommen, R. R. Frants, and S. M. van der Maarel. 2002. Facioscapulohumeral muscular dystrophy is uniquely associated with one of the variants of the 4q subtelomeres. Nat. Genet. 32:235-236.
Lyle, R., T. J. Wright, L. N. Clark, and J. E. Hewitt. 1995. The FSHD-associated repeat, D4Z4, is a member of a dispersed family of homeobox-containing repeats, subsets of which are clustered on the short arms of the acrocentric chromosomes. Genomics 28:389-397.
Meneveri, R., A. Agresti, G. Della Valle, D. Talarico, A. G. Siccardi, and E. Ginelli. 1985. Identification of a human clustered GC-rich DNA family of repeats (Sau 3A family). J. Mol. Biol. 186:483-489.
Meneveri, R., A. Agresti, A. Marozzi, S. Saccone, M. Rocchi, N. Archidiacono, G. Corneo, G. Della Valle, and E. Ginelli. 1993. Molecular organization and chromosomal location of human GC-rich heterochromatic blocks. Gene 123:227-234.
Meneveri, R., A. Agresti, M. Rocchi, A. Marozzi, and E. Ginelli. 1995. Analysis of GC-rich repetitive nucleotide sequences in great apes. J. Mol. Evol. 40:405-412.
Ruvolo, M. 1997. Molecular phylogeny of the hominoids: inference from multiple independent DNA sequences data sets. Mol. Biol. Evol. 14:248-265.
Samonte, R. V., and E. E. Eichler. 2002. Segmental duplications and the evolution of the primate genome. Nat. Rew. Genet. 3:65-72.
Schweizer, D., and J. Loidl. 1987. A model for heterochromatin dispersion and evolution of C-band patterns. Chromosome Today 9:61-74.
Sullivan, B. A., L. S. Jenkins, E. M. Karson, J. Leana-Cox, and S. Schwartz. 1996. Evidence for structural heterogeneity from molecular cytogenetic analysis of dicentric Robertsonian translocations. Am. J. Hum. Genet. 59:167-175.
Ventura, M., J. M. Mudge, and V. Palumbo, et al. (11 co-authors). 2003. Neocentromeres in 15q24–26 map to duplicons which flanked an ancestral centromere in 15q25. Genome Res. 13:2059-2068.
Willard, H. F. 1991. Evolution of alpha satellite. Curr. Opin. Genet. Dev. 1:509-514.
Winokur, S. T., U. Bengtsson, J. C. Vargas, J. J. Wasmuth, and M. R. Altherr. 1996. The evolutionary distribution and structural organization of the homeobox-containing repeat D4Z4 indicates a functional role for the ancestral copy in the FSHD region. Hum. Mol. Genet. 10:1567-1575.(M. F. Cardone*, L. Ballar)