Downloaded from genome.cshlp.org on September 27, 2021 - Published by Cold Spring Harbor Laboratory Press Article Alu Elements Support Independent Origin of Prosimian, Platyrrhine, and Catarrhine Mhc-DRB Genes Karin Kriener, Colm O’hUigin, and Jan Klein1 Max-Planck-Institut fu¨r Biologie, Abteilung Immungenetik, D-72076 Tu¨bingen, Germany The primate major histocompatibility complex (Mhc) genes fall into two classes and each of the classes into several families. Of the class II families, the DRB family has a long and complex evolutionary history marked by gene turnover, rearrangement, and molecular convergence. Because the history is not easily decipherable from sequences alone, Alu element insertions were used as cladistic markers to support the surmised phylogenetic relationships among the DRB genes. Intron 1 segments of 24 DRB genes from five platyrrhine species and five DRB genes from three prosimian species were amplified by PCR and cloned, and the amplification products were sequenced or PCR–typed for Alu repeats. Three Alu elements were identified in the platyrrhine and four in the prosimian DRB genes. One of the platyrrhine elements (Alu50J) is also found in the Catarrhini, whereas the other two (Alu62Sc, Alu63Sc) are restricted to the New World monkeys. Similarly, the four prosimian elements are found only in this taxon. This distribution of Alu elements is consistent with the phylogeny of the DRB genes as determined from their intron 1 sequences in an earlier and the present study. It contradicts the exon 2-based phylogeny and thus corroborates the conclusion that the evolution of DRB exon 2 sequences is, to some extent, shaped by molecular convergence. Taken together, the data indicate that each of the assemblages of DRB genes in prosimians, platyrrhines, and catarrhines is derived from a separate ancestral gene. [The sequence data described in this paper have been submitted to the GenBank data library under accession nos. AF197226–AF197240.] The major histocompatibility complex (Mhc)isamul- In earlier publications (Kriener et al. 2000a,b), we ticomponent assemblage comprised of genes of differ- provided evidence that exon 2 sequences, on which ent age (Parham 1999). All jawed vertebrates possess previous phylogenies of primate DRB genes were based two classes of Mhc loci and in each class there are sev- (Trtkova´et al. 1993; Figueroa et al. 1994; Gyllensten et eral families of genes whose divergence times differ al. 1994), are providing misleading phylogenetic sig- depending on the taxonomical position of the animal nals. The evolution of the exon is strongly affected by (Klein 1986; Klein and Figueroa 1986; Kasahara et al. positive selection (Hughes and Nei 1989b), which cre- 1995). In primates, a few of the class I loci diverged ates repeatedly and independently similar sequence prior to the emergence of this order, but most are of motifs (O’hUigin 1995; Kriener et al. 2000a,b). These much more recent origin (Hughes and Nei 1989a). The motifs make genes appear more closely related than primate class II gene families DO, DP, DQ, and DR,on they are in reality. This, at least, is the message ex- the other hand, diverged before the radiation of the tracted from comparisons of the DRB intron with the eutherian mammals (Carson and Trowsdale 1986). DRB exon 2 sequences. Specifically, whereas the exon 2 Within the families, the loci can vary considerably in sequences suggest that most primate DRB genes derive their ages (Satta et al. 1996a,b). Loci of the DRB sub- from a common ancestor that existed prior to the di- family in particular appear to have undergone frequent vergence of prosimians, Platyrrhini, and Catarrhini, in- expansions and contractions (Klein et al. 1993) that tron sequences support the origin of DRB genes in each considerably obscured their evolutionary history. The of the three taxa from a distinct ancestor (Kupfermann deciphering of their history is further complicated by et al. 1999; Kriener et al. 2000a,b). Analysis of the exon the fact that parts of the genes are subject to conver- 2 similarities implies molecular convergence as an ex- gent evolution (Andersson et al. 1991; Kriener et al. planation and thus indicates that the introns and not 2000a,b). Interpreting the evolution of these genes is exon 2 are reflecting the true DRB gene phylogeny. therefore a daunting task, which can succeed only if However, as both the exon 2 and intron data are se- based on a combination of different approaches and quence–based, an independent source of information utilization of a variety of marker systems. corroborating these conclusions was needed. We 1Corresponding author. sought this source in the Alu elements inserted into the E-MAIL [email protected]; FAX 49 7071/600437. DRB genes. 634 Genome Research 10:634–643 ©2000 by Cold Spring Harbor Laboratory Press ISSN 1088-9051/00 $5.00; www.genome.org www.genome.org Downloaded from genome.cshlp.org on September 27, 2021 - Published by Cold Spring Harbor Laboratory Press Alu elements in Mhc genes The use of short interspersed repetitive elements ence, and because several Alu elements were identified (SINEs) in phylogenetic analysis is widespread. They in it in catarrhine DRB genes (Andersson et al. 1987; have been used successfully to resolve phylogenies of a Mnˇukova´et al. 1994; Satta et al. 1996a), including one variety of mammals and other vertebrates (Batzer et al. old element (Alu50J). To identify Alu elements in plat- 1994; Shimamura et al. 1997; Stoneking et al. 1997; yrrhine DRB genes, we selected seven genomic New Hamdi et al. 1999; Nikaido et al. 1999). They offer the World Monkeys (NWM) DNA samples bearing previ- advantages of ubiquity, uniqueness, and stability. In- ously identified DRB exon 2 sequences (Trtkova´et al. sertion occurs often enough to provide an array of use- 1993). Using exon 1- and exon 2-based primers, we ful cladistic markers. The chance of independent inser- then attempted to amplify the entire intron 1 and most tions at identical positions is small. Finally, SINEs are of exon 2 of the different DRB genes by PCR. We suc- rarely removed without leaving evidence of their pre- ceeded in amplifying segments from 24 different DRB vious presence. genes and failed with 5, perhaps because of a too large Alu repeats constitute one of several families of intron length. We confirmed the identity of the 24 SINEs found in the mammalian genome (Deininger amplification products by cloning them and sequenc- and Batzer 1993; Jurka 1995). They are believed to be ing their ends, including exon 2. For 23 of the 24 derived from the 7SL RNA, which is a part of an 11S clones, the exon 2 sequences have already been de- cytoplasmic ribonucleoprotein involved in targeting scribed (Trtkova´ et al. 1993; Gyllensten et al. 1994; secretory proteins across the membranes of the endo- Antunes et al. 1998; Kriener et al. 2000a, b), whereas plasmic reticulum (Ullu and Tschudi 1984). The deri- one sequence identified a new exon 2, which we des- vation occurred in multiple steps in which two vari- ignate Sasc–DRB*W3401. Ten of the 24 amplification ants were first produced by deletions in different parts products were chosen for restriction digest and hybrid- of the 7SL RNA gene — the free left Alu monomer or ization analysis. FLAM and the free right Alu monomer or FRAM — and Samples of each of the 10 clones were divided into these monomers then fused to form the dimeric Alu three parts and each part was digested with a different elements (Quentin 1992a,b). All three forms are still pair of restriction enzymes (BamHI–HindIII, HindIII– found in the primate genome. The dimeric family of HincII, and BamHI–EcoRI). The digests were separated Alu elements is divided into three major subfamilies by gel electrophoresis, blotted, and the blots were hy- that are of different age, and which, in the standard- bridized with an Alu-specific probe (Fig. 1). The probe ized nomenclature of Batzer et al. (1996), are desig- was obtained by PCR amplification of human genomic nated J (∼80-million-years [my] old), S (∼35–44-my DNA using primers Alu1 and Alu2 (Table 1). It was ∼ old), and Y (< 5-my old). Each subfamily is further di- 250-bp long and was shown in control experiments to vided into sub-subfamilies (e.g., the S subfamily is dif- hybridize with members of the J, Sb, Sc, and Sq sub- ferentiated into Sx, Sy, Sp, and Sc branches). The sub- families of Alu elements. By use of this probe, one or families and the sub-subfamilies are distinguished by two hybridizing fragments could be identified on the diagnostic substitutions shared by all members of a blots of the digested NWM clones. The positive frag- given group. The Alu elements are retroposons owing ments were then subcloned and sequenced. The se- their mobility to the possession of sequences enabling quences were aligned and the Alu elements in them their transcription by RNA polymerase III. identified (Fig. 2). As Alu elements are ubiquitous in the primate ge- This approach revealed the presence of three dis- nome, they can be identified relatively easily in any tinct Alu elements in platyrrhine DRB intron 1 se- genomic region of interest. In earlier studies, we iden- quences. One of these three elements, Alu50, was iden- tified a series of > 60 Alu elements in the catarrhine tified previously in catarrhine DRB genes (Mnˇukova´et Mhc–DRB genes (Scho¨nbach and Klein 1991; Mnˇukova´ al. 1994), the other two are new and so we designated et al. 1994; Satta et al. 1996a) and designated them them Alu62 and Alu63. The identification of the first Alu1–Alu61. The aim of the present study was to iden- element as Alu50 is based on sequence similarity and tify platyrrhine and prosimian DRB gene-associated sharing of flanking direct repeats, as well as its position Alu elements and use them to resolve the incongru- and orientation in intron 1 (Figs.