A Contig Map of the Mhc Class I Genomic Region in the Zebrafish Reveals Ancient Synteny

Věra Michalová, Brent W. Murray, Holger Sültmann and Jan Klein
J Immunol May 15, 2000, 164 (10) 5296-5305; DOI: https://doi.org/10.4049/jimmunol.164.10.5296

Abstract

In contrast to the human and mouse Mhc, in which the clusters of class I and class II loci reside in close vicinity to one another, in the zebrafish, Danio rerio, they are found in different linkage groups. Chromosome walking using BAC (bacterial artificial chromosome) and PAC (P1 artificial chromosome) clones reveals the zebrafish class I region to occupy a segment of ∼450 kb and to encompass at least 19 loci. These include three class I (Dare-UDA, -UEA, -UFA), five proteasome subunit β (PSMB8, -9A, -9C, -11, -12), two TAPs (TAP2A, TAP2B), and one TAP binding protein (TAPBP). This arrangement contrasts with the arrangements found in human and mouse Mhc, in which the orthologues of the PSMB, TAP, and TAPBP loci reside within the class II region. In addition to this main zebrafish class I contig, a shorter contig of about 150 kb contains two additional class I (UBA, UCA) and at least five other loci. It probably represents a different haplotype of part of the class I region. The previously identified UAA gene shares an identical 5′ part with UEA, but the two genes differ in their 3′ parts. One of them is probably the result of an unequal crossing over. The described organization has implications for the persistence of syntenic relationships, coevolution of loci, and interpretation of the origin of the human/mouse Mhc organization.

The Mhc derives its name from the observation that tissue transplants mismatched at its loci are rapidly destroyed by the recipient’s immune system (1). The loci governing this histoincompatibility (allograft) reaction fall into two categories: the class I and class II loci, which diverged evolutionarily from each other shortly after the emergence of their common ancestor >400 million years ago (2). The allograft reaction is an artifactual manifestation of the true function of the class I and class II loci, which is the presentation of antigenic peptides for recognition by T lymphocytes and thus the initiation of the adaptive immune response (3). Histoincompatibility and the allograft reaction occur because the donor’s class I and class II molecules of the graft present peptides to the host’s T lymphocytes and because the host’s class I and class II molecules present peptides derived from the Mhc molecules of the donor (4). The former situation is the consequence of the TCR’s plasticity; although selected to recognize peptides presented by self class I and class II molecules (5), the receptor’s degeneracy enables it to interact also with a range of nonself (allogeneic) Mhc molecules (6).

During their evolution, the class I and class II genes continuously suffer duplications and deletions (7, 8) so that their numbers vary not only between, but also within species. Often they differentiate into families that persist in various evolutionary lineages for different lengths of time and then disappear, only to be replaced by new families. Despite their long evolutionary history checkered by genomic upheavals, however, most of the human Mhc (HLA) class I loci and all the class II loci are found together in a single chromosomal segment, ∼3.6 million bp in length (9). Similar conservation of class I-class II linkage has been documented for all other mammals investigated (10) and is indicated also for two other tetrapod groups, birds (11) and amphibians (12). Furthermore, in all these taxa, the Mhc segment also contains two other groups of loci. In one group are loci that are structurally unrelated to but functionally involved with the class I loci; in the other are loci that are structurally and functionally unrelated to both the class I and class II loci (13). The former group comprises the proteasome subunit β type (PSMB),4 the TAP (TAP), and the TAP binding (TAPBP) loci (14). The PSMB8 and PSMB9 code for the low molecular mass polypeptide 7 and 2 subunits, which replace the X and Y subunits of a housekeeping proteasome, thereby converting it into an immunoproteasome-producing peptide particularly suited for loading onto class I molecules (15). The TAP1 and TAP2 genes each encode two polypeptides of a heterodimer that spans the membrane of the endoplasmic reticulum (ER) multiple times and transports proteasome-generated peptides from the cytosol onto the lumenal surface of the ER (16). The TAPBP gene specifies a protein that links newly synthesized class I molecules to the TAP and thus assists in peptide loading onto them (17). The second group of unrelated loci is a potpourri of genetic elements, some of which are scattered among the class I or class II loci, while others are aggregated in the so-called class III region (9).

In contrast to the situation in the tetrapods, in the zebrafish, a representative of the teleost (bony) fishes, the class I loci are not linked to the class II loci (18), the class II loci are split among three linkage groups (18), and the TAP, as well as the PSMB loci are closely linked to the class I loci (19); in mammals they are located in the class II region (20, 21). A survey of different teleost orders indicates that a similar situation to that in the zebrafish exists in other bony fishes (22). Because bony fishes comprise, in terms of species numbers, at least one half of living vertebrates (23), the zebrafish arrangement is not an esoteric deviation from the norm, but, rather, an important variant of the norm. For this reason it is essential for understanding of the origin and evolution of the Mhc to identify the loci in the zebrafish class I and class II regions, to determine their organization, and to establish the positions of the orthologues of the mammalian class III genes in the zebrafish genome.

Materials and Methods

Genomic clones

The zebrafish BAC clones 716, 913, and 42 were described by Graser et al. (24). They were obtained from a library produced by Genome Systems (St. Louis, MO) from a fish of the AB stock. The PAC and YAC clones were obtained from libraries produced at the Resource Center of the German Human Genome Project (25). DNA was isolated from 20-ml overnight cultures in the Luria-Bertoni medium by the alkaline lysis method (26) in the case of PAC clones and the Plasmid Midi Kit (Qiagen, Hilden, Germany) method in the case of the BAC clones. YAC clones were grown at 30°C for 48 h in the YNB medium (Difco, Detroit, MI) supplemented with 2% dextrose, nucleotides, and amino acids, except uracil and tryptophan. YAC clone DNA was isolated with the Genomic DNA preparation kit (Qiagen, Chatsworth, CA). The DNA was single, double, or partially digested with restriction enzymes (New England Biolabs, Schwalbach, Germany) according to the supplier’s instructions.

Polymerase chain reaction

PCR amplifications (27) were conducted in 25 or 50 μl of reaction mixture using the PTC-100 Programmable Thermal Controller (MJ Research, Watertown, MA) and 1× reaction buffer containing 1.5 mM MgCl2, 100 μM of each of the four deoxynucleoside triphosphates, 1 μM of each of the sense and antisense primers, 1 U of the Taq DNA polymerase (Amersham Pharmacia, Arlington Heights, IL), and 20–100 ng of template DNA.

Vectorette PCR

To obtain the ends of the BAC and PAC clone inserts, the clone DNA was digested with a frequently cutting, blunt end-producing restriction enzyme (HaeIII or RsaI). Partially double-stranded adaptors consisting of the “bubble-top” oligonucleotide and the “bubble-bottom” oligonucleotide were ligated at 16°C overnight to the ends of the digested DNA using T4 DNA ligase (New England Biolabs). The vectorette PCR was conducted with the bubble primer (which is identical with a sequence stretch of the bubble-bottom oligonucleotide) in combination with either the PAC-SP6 or the PAC-T7 end-specific primers. For amplifications of BAC clone inserts the BAC-SP6 or the BAC-T7 primer was used. The resulting PCR products, representing the ends of the inserts and a small part of the vector, were subcloned and sequenced. Hybridization probes were PCR amplified from the subclones with the bubble primer and either the PAC-SP6 or the PAC-T7 primer. The PCR products were separated by electrophoresis on low melting point agarose gels (Biozym, Hessisch-Oldendorf, Germany).

Subcloning and hybridization

The PCR products were purified from agarose gel slices with the QIAEX II kit (Qiagen), ligated to the pGEM-T vector (Promega, Mannheim, Germany) under conditions recommended by the supplier, and used to transform electrocompetent Escherichia coli DH10B cells (Life Technologies, Eggenstein, Germany). For hybridizations, DNA was labeled with [α-32P]dCTP and the Ready-to-Go or the AlkPhos DNA labeling kits (all materials from Amersham Pharmacia). Radioactive hybridizations were conducted overnight at 42°C in 40% formamide (26) in the presence of 100 μg/ml yeast transfer RNA. PAC libraries from the RZPD (25) were screened with radioactively labeled probes and washed according to the supplier’s instructions. The filters were washed and used to expose either an x-ray film (Kodak, Stuttgart, Germany) or the Fuji BAS 1000 x-ray image plates (Raytest, Straubenhardt, Germany). Nonradioactive hybridizations were performed according to the supplier’s instructions.

Sequencing and sequence analysis

Plasmid DNA was isolated from 10-ml overnight cultures with the Qiagen Plasmid Mini kit. The DNA was sequenced using AutoRead or cycle sequencing kits (Amersham Pharmacia) with fluorescent primers annealing to multiple cloning sites of the pGEM-T vector. All clones were sequenced from both ends. The sequencing products were separated on an A.L.F. (Amersham Pharmacia) or an LI-COR DNA sequencer (MWG Biotech, Ebersberg, Germany). The sequences were analyzed with the GCG software package (28). The new sequences reported here are available under GenBank accession numbers AF181248, AF182155–AF182161, and AF192763.

Linkage analysis by PCR of YAC pools

Pools of the zebrafish YAC library Y932 (25) were screened for the presence of genes by PCR using the primers listed in Table I⇓. Following identification of the positive primary pools, the secondary pools were rescreened by PCR, and the individual YAC clones carrying the genes of interest were identified.

View this table:
Table I.

PCR primers used in the analysis of BAC/PAC clones

Results

Clone identification and contig assembly

The starting point of the analysis was the earlier identification of two class I gene-bearing BAC clones, 42 and 716 (24). The T7 and SP6 ends of these two clones were amplified by vectorette PCR and used as probes to screen a PAC library (no. 706) (25) by hybridization. The screening yielded four clones: P02, I01, B20, and K16. The SP6 end of the I01 clone was then amplified by vectorette PCR and used to screen the PAC library again. This second round of screening yielded nine clones: IS1, 2, 3, 5, 6, 9, 11, 15, and 16. In a similar manner, a probe obtained by vectorette PCR amplification of the SP6 end of the B20 clone hybridized with three PAC clones in the library: BS1, 2, and 9. Additional clones were obtained by screening the library with probes specific for genes previously mapped into the zebrafish class I region. Thus, hybridization of the PAC library with the RXRD intron 6 and RXRE intron 8 probes (29) yielded three clones (RX2, 3, and 4) and one clone (RX1), respectively. Probes obtained from the T7 and SP6 ends of the RX1 clone identified two additional clones (R12 and R15) and one clone (R32) in the PAC library, respectively. Hybridization of the PAC library with the full-length cDNA probe of the class I gene Dare-UBA*01 (30) yielded seven clones: CI1, 6, 12, 13, 14, 19, and 20. Finally, screening of the library with the zebrafish RING3 probe (cDNA fragment encompassing exons 3 and 4) (19) produced six PAC clones: R32 through R37. The full clone designations are available on request.

The various clones were arranged in a contig (Fig. 1⇓), and the locations of genes on the clones were determined by application of three methods: hybridization with gene- or end-specific probes (Table II⇓), PCR amplification and sequencing using gene-specific primers (especially in the case of the class I genes; Fig. 2⇓), and restriction fragment analysis. The restriction maps, based on single, double, and partial digestions with the enzymes MluI, NotI, SalI, and XhoI, were constructed for the clones RX1, R32, R34, BAC716, CI19, P02, B20, BAC42, BS2, IS1, K16, CI12, and I01. The clone order and the restriction maps were confirmed by sequencing the T7 ends of the clones RX1, R12, BAC7 (24), B20, BAC42, CI12, and IS11 and the SP6 ends of the clones P02, BAC913 (24), B20, K16, BS2, and I01.

FIGURE 1.
FIGURE 1.

Contig map of the zebrafish Mhc class I gene region. The two clusters (A and B) and the singleton clones are boxed. The locations of the genes are symbolized by filled rectangles, and the direction of transcription (if known) is indicated by arrows. Asterisks above the clones indicate that the presence of the corresponding gene was confirmed by PCR amplification. Clones extending the contig (long arrows) are not shown in full. Letters with bars indicate restriction sites shared by all clones covering the particular segment. Abbreviations of the restriction sites: m, MluI; n, NotI; s, SalI; x, XhoI.

FIGURE 2.
FIGURE 2.

Alignment of class Ia nucleotide sequences obtained from the genomic clones. Identities with the consensus sequence are indicated by dashes, insertions/deletions (indels) are shown by asterisks, and lack of sequence information is indicated by dots.

View this table:
Table II.

Probes used for restriction mapping of the clones

Altogether 34 clones (2 BAC and 32 PAC clones) were characterized (Fig. 1⇑). Of these, 29 comprised a single cluster of overlapping clones (cluster A), three formed another cluster (B), and the remaining two (CI14 and BS1) behaved as singletons. Neither the B cluster nor the singletons could be integrated into the A cluster. Of the 29 A cluster clones, 10 were restriction-mapped; the positions of the remaining 19 clones in the contig were determined on the basis of PCR amplification using primers specific for gene fragments or nongenic regions (mostly clone ends). Orthologous relationships between corresponding regions of different clones were established from both restriction fragment size comparisons and sequences of the amplified fragments. The cluster A contig is ∼450 kb long and it contains 19 identified genes or gene fragments. The orientation of some of the genes was determined from the distribution of exons in different clones or clone fragments (31, 32).

Cluster A analysis

Cluster A contains three previously undescribed class I loci, which we designate, in accordance with the accepted convention (33), Dare-UEA, -UFA, and -UDA (Fig. 10) (32). The three loci occupy a segment ∼130 kb in length and are all oriented in the same direction. Each of the UEA and UFA loci is spread over ∼25 kb of sequence, which, at least in the case of UEA, is in part due to an unusually long (∼10 kb) intron 2. The three class I loci mingle with PSMB and TAP loci; the segment between UEA and UFA contains the PSMB9C, and that between UFA and UDA contains the TAP2B loci. cDNA clones derived from the transcripts of the three class I loci were found in a cDNA library, and both the genomic and cDNA sequences contained no obvious defect that might render them nonfunctional (Fig. 2⇑). Genes at these loci were found in different stocks of the zebrafish (in contrast to other class I genes discussed below) and may therefore represent the main (class Ia) genes of this species. A curious relationship was found between UEA and the previously described (30) UAA locus. Genes at these two loci have identical exon 2 sequences, but differ considerably in all downstream exon sequences (overall sequence divergence of 34%; Fig. 2⇑). The relationship is not the result of a cloning artifact, because both types of genes have been found in cDNA libraries, different genomic libraries, and genomic DNA. However, because screening of the PAC library with several class I probes failed to yield any UAA exon 3-positive clones, the UAA gene is apparently present in only some zebrafish stocks. It may have arisen by unequal crossing over that brought together the 5′ part of a gene at one locus with the 3′ part of a gene at another locus. The recombination may have taken place in intron 2, which is not only long in the UEA gene, but also rich in repeats (Fig. 1⇑). Since on phylogenetic analysis the UAA/UEA exon 2 clusters with UBA exon 2 (not shown) and the two are considerably divergent from the exon 2 sequences of the other class I genes (Fig. 2⇑), the recombination might have involved a UBA-like gene.

Cluster A contains five PSMB genes (PSMB9A, PSMB11, PSMB12, PSMB8, and PSMB9C) and two TAP genes (TAP2A and TAP2B; Fig. 1⇑); these were described by Murray et al. (31) and Sültmann et al. (32). Four of the PSMB genes and one TAP gene (TAP2A) are clustered on a segment of <30 kb, at a distance of about 10 kb from the nearest class I locus (UEA). One PSMB gene (PSMB9C) and one TAP gene (TAP2B) are inserted between the UEA-UFA and UFA-UDA genes, respectively. Analysis of promoter regions suggests that expression of the PSMB and TAP genes in the main cluster might be regulated by IFN-γ and that the clustering and orientation of the genes may be related to this observation (31). In mammals, the PSMB-TAP clusters contain two different TAP loci, TAP1 and TAP2 (9). Extensive search of the PAC clones and of genomic DNA has, however, failed to produce any evidence for the existence of the TAP1 locus in the zebrafish. The single TAP2B gene is at a distance of about 50 kb from the nearest class I gene. Thus, in the zebrafish, the genes functionally involved with the class I loci are near or intermingled with these loci.

In a search for additional zebrafish orthologues of HLA-associated genes, we used PCR primers based on available vertebrate sequences of the collagen α-2 (type XI, COL11A2) gene to screen zebrafish genomic DNA. The human COL11A2 gene maps about 50 kb centromeric of the class II DPB gene region (9). Using the primer pair ColXIa2*e57f3/ColXIa2*e57r3, we amplified a 231-bp zebrafish genomic DNA fragment that corresponds to the human COL11A2 exon 57. This fragment was most similar to the mouse and human COL11A2 genes in a Blastx search. A phylogenetic analysis (not shown) with different collagen sequences identified the zebrafish COL11A2 gene fragment as the closest relative of the human and mouse Mhc-linked COL11A2 genes. Thus, the zebrafish COL11A2 represents the orthologue of these genes. By using the genomic PAC clone DNAs as templates, the zebrafish COL11A2 gene fragment was PCR amplified from the clones R32, R36, R35, and R37, but not R34. However, by using the primers PAC-SP6 and ColXIa2*e57r3 we were able to obtain the COL11A2 fragment by PCR from the PAC clone R34. Sequencing confirmed the identity of the two fragments and revealed the reason for the amplification failure from the R34 clone with the former pair of primers: the absence of one priming site in the R34 clone. It also enabled us to determine the transcriptional orientation of the COL11A2 gene (Fig. 1⇑).

All nine loci identified in cluster A whose protein products are not known to be involved in Ag processing are Mhc associated in humans, albeit eight of them (RXRE, RING3, KE6, TAPBP, DAXX, BING1, COL11A2, and KNSL2) with the class II region (9, 34) and only one (REGGIE2) with the class I region (35). Of these, the retinoic acid receptor gene deserves comment. Two zebrafish RXR genes, RXRE and RXRD (36), are believed to have arisen by gene duplication from an ancestral gene closely related to the mammalian RXRB gene, which in humans is at a distance of about 100 kb from the nearest class II locus (HLA-DPB2) (37). By using haploid zebrafish embryos segregating for the RXR loci, Gongora et al., (29) reported that both RXRD and RXRE genes are linked to the class I UCA gene. This observation prompted us to screen the zebrafish PAC library with a pool of RXRD and RXRE probes. Of the four positive clones thus identified, three (RX2, -3, and -4) were shown to carry the RXRD locus, and one (RX1) the RXRE was shown to carry the locus. The RX1 clone could subsequently be linked to the rest of cluster A and mapped to a position ∼150 kb from the UEA locus (Fig. 1⇑). Attempts to link the class I contig with the group of RXRD-bearing clones by chromosome walking failed, however. In fact, in a study still in progress (B. W. Murray and R. Geisler, unpublished observations), the RXRD locus could be mapped, using a panel of zebrafish radiation hybrids, to linkage group 16, whereas the class I loci are known to be in linkage group 19 (18). Therefore, the mapping of the RXRD locus by Gongora et al. (29) must have been in error.

The RXRE gene marks one end of the zebrafish class I region as currently known; the opposite end is marked by the REGGIE2 gene. Attempts to extend the region on either side of cluster A failed for lack of markers orthologous to genes mapped to the Mhc in other vertebrate taxa. The REGGIE2 homologue was identified by Blastx and Fasta searches using the T7 end sequence of the PAC clone IS11. The searches revealed a significant similarity between the zebrafish sequence and the goldfish growth-associated protein reggie-2 (38) (GenBank accession no. U33556) and the mouse flotillin protein (39). The zebrafish end sequence of the IS11T7 clone is 252 bp long and shares 55% overall nucleotide similarity with the goldfish cDNA sequence (Fig. 3⇓). The similarity between the two sequences is concentrated in the region between sites 4 and 96 of the zebrafish sequence. Because sites 4/5 and 104/105 are occupied by the eukaryotic consensus splicing sites (AG and GT, respectively), it is possible that the similarity encompasses a short exon. The genomic organization of the fish REGGIE2 gene is not known and no further attempt was made to characterize the zebrafish gene. Phylogenetic analysis (not shown) places the zebrafish gene fragment and the goldfish REGGIE2 gene in a clade that contains the flotillin (ESA1) genes of Drosophila melanogaster, mouse, human, and yeast. Recently, a flotillin gene has been found in the class I region of the human Mhc (35). The REGGIE2 gene may therefore mark the end of synteny between the TAPBP, DAXX, BING1, and KNSL2 gene-containing class II regions in zebrafish and mammals.

FIGURE 3.
FIGURE 3.

Nucleotide sequence alignment of the zebrafish REGGIE2 gene fragment (upper sequence) with the goldfish REGGIE2 sequence (lower sequence; accession code U33556). Identities are indicated by vertical lines.

By cDNA selection, the clones BAC716, PAC P02, BAC42, and PAC I01 were shown to contain several repetitive elements (32). Hybridization analysis places the LINE elements on BAC716 and -42, the SATA repeat on BAC716 and PAC P02, the Tc1 element on PAC P02, and the DANA repeat on BAC716 and 42 (Fig. 1⇑). None of the repeats was found to hybridize to the PAC I01 clone.

Cluster B and singleton analysis

The second contig, cluster B, was identified by three clones. This contig is ∼150 kb long and contains two class I and five other genes. Based on sequence similarity (Fig. 2⇑), the two class I genes could be identified as Dare-UBA and -UCA, characterized in an earlier study (30). The UBA gene is at the end of the K16 clone, the end interrupting the gene’s exon 4, but preserving the 5′ part of the locus. A complete UBA cDNA sequence was described by Takeuchi et al. (30). The UCA gene is probably complete, at least on clone K16 (for K16 and CI12 only the presence of exons 1, 2, 3, and 4 and introns 1 and 3 was established). A cDNA derived from the UCA gene was found in some libraries, but not in others. The organization of the eight loci on contig B corresponds to a part of contig A. The differences between the two contigs are, first, the lack of allelism between the class I genes and, second, the sequence divergence among non-class I loci; in the ∼1 kb of intron sequence of the DAXX gene (Fig. 4⇓) and ∼370-bp exon sequence of the BING1 gene, the two contigs showed an average sequence divergence of 2%. Furthermore, while the sequences derived from the cluster A clones BS2, BS9, and IS1 were all identical, as were those of cluster B clones K16 and I01, the overall divergence between the CI12 and I01 clones of cluster B was 5%. Because of the high sequence similarity at the shared UCA locus (Fig. 2⇑) and the similarity of the restriction maps (Fig. 1⇑), we consider CI12 to be allelic with K16/IO1.

FIGURE 4.
FIGURE 4.

Alignment of the zebrafish DAXX intron sequences obtained from six PAC clones. Positions of the variant sites in the intron are indicated by vertical numbers. Identities with the consensus sequence are indicated by dashes; indels are indicated by asterisks.

As for the two singletons, the BS1 clone could be shown to contain a fragment of a class I gene (exon 2; the rest of the gene could not be found), as well as the PSMB9B and TAPBP genes. The class I gene exon 2 sequence resembled UDA, but its identity could not be established with certainty. The CI14 clone contained a fragment of a class I gene, as well as the TAP2A, PSMB9A, PSMB11, PSMB12, and PSMB8 genes (Fig. 1⇑). In the class I gene, exons 3 and 4, but not exon 2, could be found; the 5′ part had apparently been lost during generation of the clone. The gene appears to be an allele at the UEA locus, a contention also supported by the presence of the TAP2 and PSMB genes on the clone. The class I gene has, however, diverged considerably from the corresponding gene of the A cluster (average nucleotide difference of 11%).

YAC pool analysis

In an attempt to provide additional support for the organization deduced from the study of the PAC clones, primary and secondary YAC pools of the library Y932 (25) were screened by PCR using primer pairs for the class I (exon 4), PSMB, BING1, and RXRE (29) genes and the I01SP6 clone end (Table I⇑). Three YAC clones were identified that carried most of the genes examined. The YAC clone MGH y932E0951 yielded PCR fragments with primers for class I exon 4, PSMB, BING1, and RXRE genes, and I01SP6. The YAC clone MGH y932E0962 was positive with the class I exon 4, PSMB, BING1, and I01SP6, but not the RXRE primers. Finally, the MGH y932G07196 YAC clone was positive with the class I exon 4, PSMB, BING1, and RXRE, but not the I01SP6 primers. Given the average YAC clone size of 500 kb (25), these results confirmed the close physical linkage of the genes as well as the order of the markers RXRE and I01SP6 with respect to the class I/PSMB/BING1 region. The class I UBA and UCA loci could not be amplified from the YAC pools. By contrast, UAA-specific exon 2 primers yielded positive PCR results with all three YAC clones tested.

Discussion

The present study maps the genes residing in the zebrafish Mhc class I region (24, 31, 32) and extends the region by chromosome walking. The analysis establishes the existence of one major contig of about 450 kb, one minor contig of about 150 kb, and two extra PAC clones. Attempts to link these four genomic fragments failed, so the first point that needs to be addressed is their mutual relationship. The interpretation of clone CI14 is relatively straightforward. The class I gene it carries is apparently an allele (although a highly divergent one) at the UEA locus of the main contig. This conclusion is supported by the observation that the gene on CI14 clone is flanked by the same genes that flank the UEA locus of the main contig. The interpretation of the BS1 clone is less straightforward. Although its class I gene could be an allele at the UDA locus, this assignment is equivocal. Similarly, although the clone carries the TAPBP locus as does the main contig, other genes (TAP2B and UFA) one might expect to be present if the clone were strictly allelic to part of the A contig are apparently absent in the BS1 clone. A similar situation occurs in the B contig. Here, the righthand part of the contig corresponds well to part of the A contig, but the lefthand part does not. The latter contains the UBA and UCA genes, which are clearly not alleles at any of the three class I loci of the main contig, and it lacks the non-class I genes mingled in the A contig with the class I loci. The BS1 clone and the B contig could be, therefore, either paralogues of a contig A segment or contig A haplotypes with rearranged genes. Arguing against the former possibility is the observation that even though genetic tests indicate that the UAA/UEA and the UBA/UCA loci are very closely linked (18), the UBA/UCA loci are absent from a YAC clone library in which three YAC clones were found that bear the UEA/UDA loci (together encompassing >500 kb of sequence). The B contig and possibly also the BS1 clone may therefore represent haplotypic variants of part of the A contig. The existence of haplotypes differing in composition and number of class I or class II genes is well documented for the Mhcs of different taxa (40, 41, 42). Because the PAC library used in the present study was prepared from ∼200 fish, the occurrence of different class I alleles and haplotypes on the clones isolated from it is not surprising.

In addition to the identifiable orthologous genes, the HLA complex and the zebrafish class I region share several paralogues at the class I loci at which, because of their complicated history, the orthologues can no longer be identified. Taken together, no fewer than 12 closely linked loci have been retained in synteny since the divergence of the zebrafish and human lineages from their last common ancestor ∼400 million years ago (Fig. 5⇓A). Most of the old syntenic relationships reported to date concerned gene families such as the Hox (43) and globin (44) genes. The synteny described here is remarkable in that it involves several genes that are clearly not members of the same gene family. Interestingly, an even older synteny reported by Trachtulec et al. (45), predating the protostome-deuterostome split, involved loci residing on the Mhc chromosome in both zebrafish and humans (RXR, Brachyury) or at least in humans (NOTCH, PBX, PIM1, TUBB, GLO1, PSMB). The reasons for the retention of the syntenic relationships are unclear. Among the members of the syntenic group shared by zebrafish and humans are genes such as PSMB, TAP, and TAPBP, which are functionally involved with the class I genes. For these it can be argued that their synteny with one another and with the class I genes has been maintained by selection facilitating their coevolution (46). Indeed, there are some data that can be interpreted as supporting the idea of PSMB (47) or TAP (48) coevolution with the class I loci; there are, however, no data supporting the coevolution of TAPBP with any of these three loci. Similarly, there is no evidence for coevolution of the remaining loci with one another or with the class I genes. In the absence of functional involvement, it can be argued that the reason why the loci have remained linked for so long is perhaps that they are all part of a single transcription loop or some other three-dimensional chromatin structure responsible for their coordinated expression (49). It is, of course, also possible that the linkage has been conserved for no apparent reason, simply for lack of an incentive to disperse (32).

FIGURE 5.
FIGURE 5.

Alignment of the human with the zebrafish Mhc class I genomic region. A, Unedited depiction. Orthologous genes are connected by vertical lines; other genes are listed on top of (human) or below (zebrafish) the horizontal lines. Not all human genes are shown. The transcriptional orientations of the genes, if known, are indicated by horizontal arrows. B, Alignment under the assumption of an inverted translocation of the gene cluster KNSL2, BING1, DAXX, and TAPBP following divergence of the teleost-tetrapod lineages (see Discussion).

The zebrafish-human comparison reveals substantial differences in the organization of the zebrafish class I and the HLA class II regions. The three main differences are, first, linkage of the class I and class II loci in the HLA compared with their independent segregation in the zebrafish (18); second, the tight association of several non-class I/II loci with the HLA class II region and the equally tight association of the orthologous loci with the class I region in the zebrafish (Refs. 19, 29 , and 32 and the present study); and third, the different order of loci in the zebrafish class I region relative to the HLA class II region (the present study). Accumulating evidence indicates that the separation of class I and class II loci onto different chromosomes is not restricted to the zebrafish, but is most likely a characteristic feature of all Euteleostei and perhaps of all bony fishes and thus of more than one-half of living vertebrates (22) (unpublished observations reported at meetings by Wiegertjes et al. (50) and Stet et al. (51)). Although the class I and class II loci presumably arose from a single ancestral gene by duplication (2), this presumption does not automatically imply that they were originally together in the same region. Tandem duplication is certainly one possible explanation of their origin, but another is the duplication of the ancestral chromosome bearing the undifferentiated Mhc gene (13). The former possibility is the more parsimonious of the two, because separating into two different parts of the genome is more likely than coming together in the same part of previously separated genes. It also appears to be supported by preliminary data providing evidence for linkage of class I and class II genes in cartilaginous fishes (unpublished observations reported at a meeting by Ohta and Flajnik (52)). However, because the traditional interpretation of the relationship of osteichthyans, chondrichthyans, and tetrapods has recently been questioned (53), the choice between the two possibilities is not as straightforward as it may seem.

Even if an origin of class I and class II loci by tandem duplication is assumed, the question of the original position in the Mhc region of the PSMB/TAP and possibly also of other loci functionally involved with the class I genes remains unanswered. Nonaka et al. (54) published preliminary evidence that in the African clawed frog Xenopus laevis, the order of loci is class II, PSMB, class I, and class III and proposed that this is the primordial organization of the Mhc. In fact, however, even if the two recombinants on which this interpretation is based had revealed the true organization of the Xenopus Mhc region, an alternative interpretation of the primordial Mhc would be possible. Our own unpublished observations indicate that in the zebrafish most of the so-called class III region loci are not linked to either the class I or the class II genes (H. Sültmann and J. Klein, manuscript in preparation). Any interpretation of the primordial Mhc must therefore take this result into account. On the other hand, the data described in the present study reveal that not only loci functionally involved with class I genes, but also other loci for which there is no evidence for such an involvement and which in humans are part of the class II region reside in the class I region in the zebrafish.

Unedited comparison of the zebrafish class I with the human class II region seems to reveal a complete scrambling of the order of loci in the two species (Fig. 5⇑A). If, however, a simple inverted translocation of a segment encompassing the KNSL2, BING1, DAXX, and TAPBP loci is introduced in either of the two species, reasonably good correspondence in the organization of the two regions is obtained (Fig. 5⇑B). Only minor rearrangements are then required to explain the remaining inconsistencies. Rearrangements apparently occur frequently during evolution of the Mhc. Examples of rearrangements documented in the few species that have been studied include inverted translocation involving part of the class II region together with a large piece of the chromosome in cattle (reviewed in Ref. 55), apparent removal of the class III region from the Mhc of the rabbit (56), translocation of two class I loci to the class II region in the mouse (reviewed in 1), and the contraction of much of the Mhc region to a segment of <100 kb and a pericentric duplication of the Mhc in the domestic fowl (11). These local rearrangements may be of interest in comparisons between Mhcs of closely related taxa, but they are largely uninformative when it comes to questions of Mhc’s origin and early evolution. For this reason our present effort concentrates on elucidation of organization of the zebrafish class II and III regions and organization of the Mhc in a representative of cartilaginous fishes.

Acknowledgments

We thank Sabine Jantschek for technical assistance, Jane Kraushaar for editorial assistance, Robert Graser and Kimitaka Takami for the clones and primers, the staff at the Resource Center of the German Human Genome Project for the PAC and YAC clones, and Robert Geisler (Max Planck Institut für Entwicklungsbiologie, Tübingen, Germany) for sharing the zebrafish RH panel and the Institute’s facilities.

Footnotes

  • ↵1 V.M. and B.W.M. contributed equally to this publication.

  • ↵2 Address correspondence to Dr. Holger Sültmann, Deutsches Krebsforschungszentrum, Im Neuenheimer Feld 506, D-69120 Heidelberg, Germany. E-mail address: h.sueltmann{at}dkfz-heidelberg.de

  • ↵3 Address reprint requests to Prof. Dr. Jan Klein, Max Planck Institut für Biologie, Abteilung Immungenetik, Corrensstrasse 42, D-72076 Tübingen, Germany. E-mail address: jan.klein{at}tuebingen.mpg.de

  • ↵4 Abbreviations used in this paper: PSMB, proteasome subunit β type; TAPBP, TAP binding protein; ER, endoplasmic reticulum; indel, insertion/deletion; BAC, bacterial artificial chromosome; PAC, P1 artificial chromosome; YAC, yeast artificial chromosome.

  • Received December 20, 1999.
  • Accepted March 2, 2000.

References

  1. ↵
    Klein, J.. 1986. Natural History of the Major Histocompatibility Complex John Wiley & Sons, New York.
  2. ↵
    Klein, J., C. O’hUigin. 1993. Composite origin of major histocompatibility complex genes. Curr. Opin. Genet. Dev. 3: 923
    OpenUrlCrossRefPubMed
  3. ↵
    Colvin, R. B.. 1990. Cellular and molecular mechanisms of allograft rejection. Annu. Rev. Med. 41: 361
    OpenUrlCrossRefPubMed
  4. ↵
    Gould, D. S., H. J. Auchincloss. 1999. Direct and indirect recognition: the role of MHC antigens in graft rejection. Immunol. Today 20: 77
    OpenUrlCrossRefPubMed
  5. ↵
    Marrack, P., J. W. Kappler. 1993. How the immune system recognizes the body. Sci. Am. 269: 80
    OpenUrlCrossRefPubMed
  6. ↵
    Garcia, K. C., M. Degano, L. R. Pease, M. Huang, P. A. Peterson, L. Teyton, I. A. Wilson. 1998. Structural basis of plasticity in T cell receptor recognition of a self peptide-MHC antigen. Science 279: 1166
    OpenUrlAbstract/FREE Full Text
  7. ↵
    Klein, J., H. Ono, D. Klein, C. O’hUigin. 1993. The accordion model of Mhc evolution. J. Gergely, and G. Petranyi, eds. Progress in Immunology 137-143. Springer-Verlag, Heidelberg.
  8. ↵
    Nei, M., X. Gu, T. Sitnikova. 1997. Evolution by the birth-and-death process in multigene families of the vertebrate immune system. Proc. Natl. Acad. Sci. USA 94: 7799
    OpenUrlAbstract/FREE Full Text
  9. ↵
    Campbell, R. D., J. Trowsdale. 1993. Map of the human MHC. Immunol. Today 14: 349
    OpenUrlCrossRefPubMed
  10. ↵
    Kasahara, M., M. F. Flajnik, T. Ishibashi, T. Natori. 1995. Evolution of the major histocompatibility complex: a current overview. Transplant. Immunol. 3: 1
    OpenUrlCrossRefPubMed
  11. ↵
    Kaufman, J., S. Milne, T. W. F. Göbel, B. A. Walker, J. P. Jacob, C. Auffray, R. Zoorob, S. Beck. 1999. The chicken B locus is a minimal essential major histocompatibility complex. Nature 401: 923
    OpenUrlCrossRefPubMed
  12. ↵
    Flajnik, M. F., Y. Ohta, C. Namikawa-Yamada, M. Nonaka. 1999. Insight into the primordial MHC from studies in ectothermic vertebrates. Immunol. Rev. 167: 59
    OpenUrlCrossRefPubMed
  13. ↵
    Klein, J., A. Sato. 1998. Birth of the major histocompatibility complex. Scand. J. Immunol. 47: 199
    OpenUrlCrossRefPubMed
  14. ↵
    Pamer, E., P. Cresswell. 1998. Mechanisms of MHC class I-restricted antigen processing. Annu. Rev. Immunol. 16: 323
    OpenUrlCrossRefPubMed
  15. ↵
    Driscoll, J., M. G. Brown, D. Finley, J. J. Monaco. 1993. MHC-linked LMP gene products specifically alter peptidase activities of the proteasome. Nature 365: 262
    OpenUrlCrossRefPubMed
  16. ↵
    Kelly, A., S. H. Powis, L. A. Kerr, I. Mockridge, T. Elliot, J. Bastin, B. Uchanska-Ziegler, J. Trowsdale, A. Townsend. 1992. Assembly and function of the two ABC transporter proteins encoded in the human major histocompatibility complex. Nature 355: 641
    OpenUrlCrossRefPubMed
  17. ↵
    Elliot, T.. 1997. How does TAP associate with MHC class I molecules?. Immunol. Today 14: 349
    OpenUrl
  18. ↵
    Bingulac-Popovic, J., F. Figueroa, A. Sato, W. S. Talbot, S. L. Johnson, M. Gates, J. H. Postlethwait, J. Klein. 1997. Mapping of Mhc class I and class II regions to different linkage groups in the zebrafish, Danio rerio. Immunogenetics 46: 129
    OpenUrlCrossRefPubMed
  19. ↵
    Takami, K., Z. Zaleska-Rutczynska, F. Figueroa, J. Klein. 1997. Linkage of LMP, TAP, and RING3 with Mhc class I rather than class II genes in the zebrafish. J. Immunol. 159: 6052
    OpenUrlAbstract
  20. ↵
    Glynne, R., S. H. Powis, S. Beck, A. Kelly, L.-A. Kerr, J. Trowsdale. 1991. A proteasome-related gene between the two ABC transporter loci in the class II region of the human MHC. Nature 353: 357
    OpenUrlCrossRefPubMed
  21. ↵
    Kelly, A., S. H. Powis, R. Glynne, E. Radley, B. S., and J. Trowsdale. 1991. Second proteasome-related gene in the human MHC class II region. Nature 353:667.
  22. ↵
    Sato, A., F. Figueroa, B. W. Murray, E. Málaga-Trillo, Z. Zaleska-Rutczynska, H. Sültmann, S. Toyosawa, C. Wedekind, N. Steck, J. Klein. 2000. Nonlinkage of the major histocompatibility complex of class I and class II loci in bony fishes. Immunogenetics 51: 108
    OpenUrlCrossRefPubMed
  23. ↵
    Eschmeyer, W. N.. 1998. Catalog of Fishes California Academy of Science, San Francisco.
  24. ↵
    Graser, R., V. Vincek, K. Takami, F. Figueroa, J. Klein. 1998. Analysis of zebrafish Mhc using BAC clones. Immunogenetics 47: 318
    OpenUrlCrossRefPubMed
  25. ↵
    Resource-Center/Primary-Database. RZPD of the German Human Genome Project at the Max-Planck-Institute for Molecular Genetics (former Reference Library Database). URL: http://www.rzpd.de, Berlin-Charlottenburg, Germany.
  26. ↵
    Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular Cloning: A Laboratory Manual. 2nd. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
  27. ↵
    Saiki, R. K., D. H. Gelfand, S. Stoffel, S. J. Scharf, R. Higuchi, G. T. Horn, K. B. Mullis, H. A. Erlich. 1988. Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239: 487
    OpenUrlAbstract/FREE Full Text
  28. ↵
    Devereux, J., and P. Haeberli. 1991. Genetics Computer Group, Program Manual for the GCG Package, Version 7. Genetics Computer Group, Madison.
  29. ↵
    Gongora, R., Z. Zaleska-Rutczynska, K. Takami, F. Figueroa, J. Klein. 1998. Linkage of RXRB-like genes to class I and not class II Mhc genes in the zebrafish. Immunogenetics 48: 141
    OpenUrlCrossRefPubMed
  30. ↵
    Takeuchi, H., F. Figueroa, C. O’hUigin, J. Klein. 1995. Cloning and characterization of class I Mhc genes of the zebrafish, Brachydanio rerio. Immunogenetics 42: 77
    OpenUrlPubMed
  31. ↵
    Murray, B. W., H. Sültmann, J. Klein. 1999. Analysis of a 26-kilobase region linked to the Mhc in zebrafish: genomic organization of the proteasome component β/transporter associated with antigen processing-2 gene cluster and identification of five new proteasome β subunit genes. J. Immunol. 163: 2657
    OpenUrlAbstract/FREE Full Text
  32. ↵
    Sültmann, H., B. W. Murray, and J. Klein. 2000. Identification of seven genes in the Mhc class I region of the zebrafish. Scand. J. Immunol. In press.
  33. ↵
    Klein, J., R. E. Bontrop, R. L. Dawkins, H. A. Erlich, U. B. Gyllensten, E. R. Heise, P. P. Jones, P. Parham, E. K. Wakeland, D. I. Watkins. 1990. Nomenclature for the major histocompatibility complexes of different species: a proposal. Immunogenetics 31: 217
    OpenUrlPubMed
  34. ↵
    Herberg, J. A., S. Beck, J. Trowsdale. 1998. TAPASIN, DAXX, RGL2, HKE2 and four new genes (BING1, 3 to 5) form a dense cluster at the centromeric end of the MHC. J. Mol. Biol. 277: 839
    OpenUrlCrossRefPubMed
  35. ↵
    MHC sequencing consortium. 1999. Complete sequence and gene map of a human major histocompatibility complex. Nature 401: 921
    OpenUrlCrossRefPubMed
  36. ↵
    Endo, T., T. Imanishi, T. Gojobori, H. Inoko. 1997. Evolutionary significance of intra-genome duplications on human chromosomes. Gene 205: 19
    OpenUrlCrossRefPubMed
  37. ↵
    Nagata, T., E. H. Weiss, K. Abe, K. Kitagawa, A. Ando, Y. Yara-Kikuti, M. F. Seldin, K. Ozato, H. Inoko, M. Taketo. 1995. Physical mapping of the retinoid receptor B gene in mouse and human. Immunogenetics 41: 83
    OpenUrlPubMed
  38. ↵
    Schulte, T., K. A. Paschke, U. Laessing, C. A. O. Stuermer. 1997. Reggie-1 and reggie-2, two cell surface proteins expressed by retinal ganglion cells during axon regeneration. Development 124: 577
    OpenUrlAbstract
  39. ↵
    Bickel, P. E., P. E. Scherer, J. E. Schnitzer, P. Oh, M. P. Lisanti, H. F. Lodish. 1997. Flotillin and epidermal surface antigen define a new family of caveolae-associated integral membrane proteins. J. Biol. Chem. 272: 13793
    OpenUrlAbstract/FREE Full Text
  40. ↵
    Böhme, J., M. Andersson, G. Andersson, E. Möller, P. A. Peterson, L. Rask. 1985. HLA-DRβ genes vary in number between different DR specificities, whereas the number of DQβ genes is constant. J. Immunol. 135: 2149
    OpenUrlAbstract
  41. ↵
    Schönbach, C., V. Vincek, W. E. Mayer, M. Golubic, C. O’hUigin, J. Klein. 1993. Multiplication of Mhc-DRB5 loci in the orangutan: implications for the evolution of DRB haplotypes. Mamm. Genome 4: 159
    OpenUrlCrossRefPubMed
  42. ↵
    Málaga-Trillo, E., Z. Zaleska-Rutczynska, B. McAndrew, V. Vincek, F. Figueroa, H. Sültmann, J. Klein. 1998. Linkage relationships and haplotype polymorphism among cichlid Mhc class II B loci. Genetics 149: 1527
    OpenUrlAbstract/FREE Full Text
  43. ↵
    Kappen, C., F. H. Ruddle. 1993. Evolution of a regulatory gene family: HOM/HOX genes. Curr. Opin. Genet. Dev. 3: 931
    OpenUrlCrossRefPubMed
  44. ↵
    Fritsch, E. F., R. M. Lawn, T. Maniatis. 1980. Molecular cloning and characterization of the human β-like globin gene cluster. Cell 19: 959
    OpenUrlCrossRefPubMed
  45. ↵
    Trachtulec, Z., R. M. Hamvas, J. Forejt, H. R. Lehrach, V. Vincek, J. Klein. 1997. Linkage of TATA-binding protein and proteasome subunit C5 genes in mice and humans reveals synteny conserved between mammals and invertebrates. Genomics 44: 1
    OpenUrlCrossRefPubMed
  46. ↵
    Trowsdale, J., S. H. Powis. 1992. The MHC: relationship between linkage and function. Curr. Opin. Genet. Dev. 2: 492
    OpenUrlCrossRefPubMed
  47. ↵
    Namikawa, C., L. Salter-Cid, M. F. Flajnik, Y. Kato, M. Nonaka, M. Sasaki. 1995. Isolation of Xenopus LMP7 homologues: striking allelic diversity and linkage to the MHC. J. Immunol. 155: 1964
    OpenUrlAbstract
  48. ↵
    Joly, E., A.-F. Le Rolle, A. L. Gonzalez, B. Mehling, J. Stevens, J. W. Coadwell, T. Hünig, J. C. Howard, G. W. Butcher. 1998. Co-evolution of rat TAP transporters and MHC class I RT1-A molecules. Curr. Biol. 8: 169
    OpenUrlCrossRefPubMed
  49. ↵
    Lundin, L. G.. 1993. Evolution of the vertebrate genome as reflected in paralogous chromosome regions in man and in the house mouse. Genomics 16: 1
    OpenUrlCrossRefPubMed
  50. ↵
    Wiegertjes, G. F., E. Egbert, R. J. M. Stet. 1997. Segregation of MHC class I and class II genes in the common carp (Cyprinus carpio L.). Dev. Comp. Immunol. 21: 105
    OpenUrl
  51. ↵
    Stet, R. J. M., R. Johnstone, P. Parham, and G. F. Wiegertjes. 1997. The unMHC of teleostean fish: segregation analyses in common carp and Atlantic salmon. In Fifth International Workshop on Mhc Evolution. (Abstract).
  52. ↵
    Ohta, Y., and M. Flajnik. 1999. Linkage analysis of MHC genes in cartilaginous fish. In Sixth International Workshop on MHC Evolution. (Abstract).
  53. ↵
    Rasmussen, A.-S., U. Arnason. 1999. Molecular studies suggest that cartilaginous fishes have a terminal position in the piscine tree. Proc. Natl. Acad. Sci. USA 96: 2177
    OpenUrlAbstract/FREE Full Text
  54. ↵
    Nonaka, M., C. Namikawa, Y. Kato, M. Sasaki, L. Salter-Cid, M. Flajnik. 1997. Major histocompatibility complex gene mapping in the amphibian Xenopus implies a primordial organization. Proc. Natl. Acad. Sci. USA 94: 5789
    OpenUrlAbstract/FREE Full Text
  55. ↵
    Lewin, H. A., G. C. Russel, E. J. Glass. 1999. Comparative organization and function of the major histocompatibility complex of domesticated cattle. Immunol. Rev. 167: 145
    OpenUrlCrossRefPubMed
  56. ↵
    Chouchane, L., T. J. Kindt. 1992. Mapping of the rabbit MHC reveals that class I genes are adjacent to the DR subregion and defines an insertion/deletion-related polymorphism in the class II region. J. Immunol. 149: 1216
    OpenUrlAbstract
Back to top

In this issue

The Journal of Immunology: 164 (10)
The Journal of Immunology
Vol. 164, Issue 10
15 May 2000
Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section