Determined by Radiation Hybrid Mapping Synteny Between Zebrafish ...

Conservation of Mhc Class III Region Synteny Between Zebrafish and Human as Determined by Radiation Hybrid Mapping1 Holger Su¨ltmann,2* Akie Sato,* Brent W. Murray,* Naoko Takezaki,† Robert Geisler,‡ Gerd-Jo¨rg Rauch,‡ and Jan Klein3* In the HLA, H2, and other mammalian Mhc, the class I and II loci are separated by the so-called class III region comprised of ⬃60 genes that are functionally and evolutionarily unrelated to the class I/II genes. To explore the origin of this island of unrelated loci in the middle of the Mhc 19 homologues of HLA class III genes, we identified 19 homologues of HLA class III genes as well as 21 additional non-class I/II HLA homologues in the zebrafish and mapped them by testing a panel of 94 zebrafish-hamster radiation hybrid cell lines. Six of the HLA class III and eight of the flanking homologues were found to be linked to the zebrafish class I (but not class II) loci in linkage group 19. The remaining homologous loci were found to be scattered over 14 zebrafish linkage groups. The linkage group 19 contains at least 25 genes (not counting the class I loci) that are also syntenic on human chromosome 6. This gene assembly presumably represents the pre-Mhc that existed before the class I/II genes arose. The pre-Mhc may not have contained the complement and other class III genes involved in immune response. The Journal of Immunology, 2000, 165: 6984 – 6993.

A

lthough all jawed vertebrates possess an Mhc, our views of it have been forged by two mammalian systems, the human HLA and the mouse H2 complexes (1, 2). The remarkable similarity of the HLA and H2 complexes (3) has kindled the expectation that the Mhc of all vertebrates would be organized in a similar way to those of the human and the mouse. For some time, this expectation seemed to be borne out by studies of other species, albeit mostly other mammals (4). Like the HLA and the H2 complexes, the Mhcs of these other species could be shown to constitute a single chromosomal segment divisible, rather arbitrarily, into three regions, ensconced by three different classes of loci, the class I and class II regions taking up the flanks and the class III region the center. The class I and II loci comprise the core of the Mhc and, with their involvement in peptide presentation (5), provide functional and evolutionary identity to the chromosomal segment they occupy. Their organization and sequence, as well as the structure of the molecules they encode, leave no doubt that the genes in each of the two classes are related to one another by their origin (6). Nor has there ever been any doubt expressed that the class I and class II genes are derived from a common ancestor (7, 8). The existing differences between the genes and their products are generally interpreted as being the result of an adaptation to slightly different functions, the class I molecules to the presenta*Max-Planck-Institut fu¨r Biologie, Abteilung Immungenetik, Tu¨bingen, Germany; † The Graduate University for Advanced Studies, Department of Biosystems Science, Hayama, Kanagawa, Japan; and ‡Max-Planck-Institut fu¨r Entwicklungsbiologie, Tu¨bingen, Germany Received for publication July 6, 2000. Accepted for publication September 12, 2000. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported in part by a grant from the Deutsches Humangenomprojekt (to R.G. and G.-J.R.). 2 Address correspondence requests to Dr. Holger Su¨ltmann, Deutsches Krebsforschungszentrum, Im Neuenheimer Feld 506, D-69120 Heidelberg, Germany. E-mail address: [email protected] 3

Address reprint requests to Dr. Jan Klein, Max-Planck-Institut fu¨r Biologie, Abteilung Immungenetik, Corrensstrasse 42, D-72076 Tu¨bingen, Germany. E-mail address: [email protected]

Copyright © 2000 by The American Association of Immunologists

tion of peptides acquired in the endoplasmic reticulum (ER),4 and the class II molecules to the presentation of peptides procured in the endosomal compartment (5). However, the presence of the class III genes in the region separating the class I and class II parts of the complex has always been somewhat puzzling (3, 8). Most of the class III genes are neither functionally nor evolutionarily related to one another. They are a variegated assortment of elements that do not seem to have any particular reason to be together with one another or with the class I and class II loci. A justification for their presence in the vicinity of class I and II loci has been sought in the involvement of some of them in immunity. However, this argument always seemed rather weak for three reasons. First, the class III region contains several loci for which there is no evidence for involvement in immunity. Second, for the loci that are involved in immunity, no compelling reason has been provided for their linkage to the class I and II loci. And third, with the large number of loci involved in immune responses, it is to be expected that almost any region of the genome will, by chance, contain some of them. The existence and composition of the class III region are not the only seemingly illogical features of the HLA/H2 organization. Another such feature concerns the genes participating in the production of peptides for loading onto the class I molecules. The bulk of these peptides is derived from cytosolic proteins processed by proteasomes (5). The peptides thus produced are then transported across the ER membrane and on the lumenal side loaded into the peptide-binding groove of the newly synthesized class I molecules. Some of the proteasome components, the ER transporters, and the molecules involved in the loading are encoded in genes (PSMB, ABCB or TAP, and TAPBP, respectively) that reside in the HLA/H2 complexes. This arrangement makes sense, for one can argue that the genes must either coevolve with the class I genes or their expression must be coordinated with that of these genes (4). The illogical aspect of this arrangement is that the PSMB, ABCB, 4 Abbreviations used in this paper: ER, endoplasmic reticulum; EST, expressed sequence tag; PAC, P1-derived artificial chromosome; YAC, yeast artificial chromosome.

0022-1767/00/$02.00

The Journal of Immunology

6985

and TAPBP genes are found in the class II rather than the class I region. More recently, evidence has accumulated that the HLA/H2 paradigm of Mhc organization has many exceptions, even in mammals. For example, in cattle, a chromosomal segment bearing a part of the class II region has been inverted so that some of the class II genes are now at a distance of ⬎17 cM from the main body of the Mhc (9 –11). In the rabbit, to give another example, the class III region has apparently been transposed to an unidentified location, and the class I and class II regions have become directly adjacent (12). Even greater departures from the HLA/H2 norm have been found when studies of Mhc organization have been extended to nonmammalian gnathostomes. Thus, in the domestic fowl, the entire Mhc segment has been reduced in length to ⬃100 kb and part of it translocated transcentrically (13). In Xenopus laevis, a separate Mhc region has been generated by tetraploidization (14). In the zebrafish (15) and the majority of, if not all, teleost fishes (16), the class I and class II regions are on different chromosomes, the latter apparently on more than one chromosome. These puzzling aspects call for an explanation that can come only from an understanding of Mhc evolution. The latter can be obtained from a detailed knowledge of the Mhc in lower vertebrates. To provide such knowledge, we have embarked on a systematic study of the zebrafish Mhc. Being a widely used model in developmental and genomic research, the zebrafish seems an appropriate choice for the stated purpose. In earlier publications, we described the zebrafish class I (17) and class II (18) loci, and the organization of the class I region (19). We have shown that in the zebrafish, the Mhc-associated PSMB, ABCB, and TAPBP genes are located in the class I rather than the class II region (20, 21). This

finding seems to explain one illogical aspect of the HLA/H2 organization. The present study has been aimed at shedding light on the puzzle of the class III region and so on the other seemingly illogical feature of the Mhc organization.

Materials and Methods PCR, subcloning, and sequencing PCR amplifications (22) were conducted in 25 or 50 ␮l of reaction mixture using the MJ Research PTC-100 Programmable Thermal Controller (MJ Research, Watertown, MA), 1⫻ reaction buffer containing 1.5 mM MgCl2, 100 ␮M of each of the four deoxynucleosidtriphosphates, 0.5–1 ␮M of each of the sense and antisense primers (Table I), 1 U of Taq DNA polymerase (Amersham Pharmacia Biotech, Freiburg, Germany) or HotStar Taq DNA polymerase (Qiagen, Hilden, Germany), and 20 –100 ng of template DNA. The PCR products were separated on agarose gels and purified with the help of the QIAEX II kit (Qiagen). For cloning, the DNA fragments were ligated to the pGEM-T vector (Promega, Mannheim, Germany) under conditions recommended by the supplier, and transformed into electrocompetent Escherichia coli DH10B cells (Life Technologies, Karlsruhe, Germany). The plasmid DNA was isolated from 10-ml overnight cultures with the Qiagen Plasmid Mini kit (Qiagen) and sequenced using the AutoRead or the cycle sequencing kits (Amersham Pharmacia Biotech) with fluorescent primers annealing to the multiple cloning sites of the pGEM-T vector. The sequencing products were separated on an A.L.F. (Amersham Pharmacia Biotech) or the LI-COR DNA sequencer (MWG Biotech, Ebersberg, Germany).

Phylogenetic analysis Homologous sequences for constructing phylogenetic trees were procured by Blastx searches (23, 24) in the nonredundant parts of the GenBank/ EMBL/DDBJ databases. For genes identified from expressed sequence tag (EST) sequences, Blastp and Blastn searches were also conducted with the nonzebrafish sequences identified as the highest match in the initial Blastx

Table I. Primers used for radiation hybrid panel typing Gene

MUT NFYA RDS HMGIY PIM1 BING3⌿ RAB2L SACM2L RING1B KE4 DBB DCB PBX2 PPT2 STK19 SK12W BF G9A HSPA1A SMX5 VARS1 CLIC1 BAT5 CSNK2B APOM BAT3 BAT2 AIF1 BAT1 TUBB EDN1 TCP1 ACAT2 a

Primer Sequence (5⬘33⬘)

TGAGAACATGGGAGGTATGGCTCG/GCGTGCTTGTCTTCGGGCAGCG ATGGTTCAAGTCAGCGGCGGTC/CAGCCCCTGTGTACCAGATACC CAACMTTCTMATYYKKGTGGGAC/TTGTTGCCACAGCAKCGGAACTC CCACAGCAGGAGGCCAGTGGATC/CGAGACACAGTAGAGCTCGGC GGCATCGGATTTCAGACGGACA/GACCACTGTTGGATTCTGTCAC GAGTTCAARGAGGCSTTYAACATGAT/CCAAACATGGTNAGGAACATGGTG AGGAGCATGTTGGTCAGTCAGAG/GCTCTGCAGCGCTGAGAGGATG TTGGTGATCTGGACATCACTACAG/TCCACCTGCCTAGAGTACTGCC CACACCACCCAGCAGGCGCTCAG/CTGTGCACTGATGACTTGCTGCA GGTTATCTGAACCTGGCTGCTGA/TTGGTGCAGCCTGATTGGACAAG CATGGACATTATGGATTT/CCTTAATGACCGCACA GGCTATCTTCAACGTCAA/TTCAGTAACTGACCGAAT TCATTGGAGCACTCCGATTATAAAAG/GTGCGAGACTGCTCCCTCAGC GTGTTTCTGATGGACTTGTTTGG/GCCATTTTCAATGCAGGTTTTAAAC GTGTTTGCTGAAGATTACAGAGCC/ATCTCAGAGTCGGAGAAGAGGAAC CCCACATGGACTGCCTCCTAC/GCCGGTGGTTGGATCTCTGACC AAGTCCGGGTTTGCCAACCTAATG/ACAGTGGTACGTGACTTCATCATC ATCTTCGAGTGTAACATGGCCTG/AGGAATATCTTGAAGAGCTCGAAC GACCGCAGGTGGAGTCATGACG/TGATCTTGCCTTTCAGGTTGTCG GATATCAGTGTTACTGATCCAGAG/ACTTCATCTGCAGGAAGCTGAAC ACCTAGCTGTCACTCAAGTGGC/TCGGCTCACTCAGTTTCTTCCTC GCTCCAGAGGTATTCAAGGACCTG/TTATGAAAGATGTCATCGCCTGCC GGCCATGAGACCAATGTTACAGC/GAATGACACCTATCACACTCCAG GACATGATCCTCGATCTGGAACC/GACAGTAGCCAAAGTCTCCCTG CTGTTTAACTCTAAATGGGAGATG/ACATACAGCATGATGATGCATGT AGGAGCACATTTCACCATCTGTTG/GCATCACATAGCTATTGGCGTTG TCAGACCAAGCCAATGAGGAATG/TGGCCGCTGACTTGAGAAGCTC GATGGGCTTGAAGCGAATGATGG/TTGAGAACAGCTGATCGCTTGCC ATCGCCATCCACAGAGGAATGG/CGGCATGTCGTAGTTGAAGACG AGGAGGTGGACGAGCAGATGCT/ATGCAGGAAAGCCTTGCGCCTG CCAGCACGTCACTCCAGGAATAAG/AGCGTTTCAAGTCCATATGACACC TGGAAAGCCAAGGGATAACAAGC/AAACATGACAGACAGCCATGCTAA AGTCTGTGTGCCTTGGTGCTC/CAGTCAGTGGCATCTCACCTAT

Ta, Annealing temperature.

Taa

PCR Product Size (bp)

60°C 60°C 63°C 60°C 60°C 63°C 60°C 60°C 58°C 60°C 45°C 45°C 60°C 60°C 60°C 60°C 63°C 60°C 60°C 63°C 63°C 60°C 60°C 60°C 60°C 60°C 63°C 60°C 50°C 63°C 60°C 58°C 58°C

88 123 327 90 180 458 330 300 400 200 379 400 370 140 160 320 350 (550) 450 476 800 400 345 184 550 300 400 250 155 400 200 282 320 600

6986 search. Alignments of amino acid sequences were generated by CLUSTALW (25), and phylogenetic trees were constructed by the neighbor-joining method (26) with Poisson-corrected distances; significance of branching patterns was assessed by 1000 bootstrap replications (27).

Linkage analysis by PCR on zebrafish radiation hybrid panels Two zebrafish radiation hybrid panels constructed by Kwok et al. (28) (T51) and Hukriede et al. (29) (LN54) were used at the Max-Planck-Institut fu¨r Entwicklungsbiologie (Tu¨bingen, Germany). The PCR, agarose gel electrophoresis, and data analysis were conducted as recommended on the web site (http://wwwmap.tuebingen.mpg.de). The current T51-based map, including the markers reported in this work, is available at the same address.

Genomic clones P1-derived artificial chromosome (PAC) clones from the zebrafish genomic library 706 were obtained from the Resource Center of the German Human Genome Project (30). PAC DNA was isolated from 20-ml overnight cultures in the Luria-Bertani medium using the plasmid mini or the large construct kit (Qiagen). The phage genomic library was screened and DNA was isolated and hybridized, as described by Su¨ltmann et al. (18). Primary and secondary yeast artificial chromosome (YAC) pool DNA of the zebrafish genomic library HACHy914 (30) was screened by PCR with the primers listed in Table I.

Southern blot hybridization A total of 7 ␮g of zebrafish genomic DNA or 5 ␮g of the radiation hybrid cell line DNA (Research Genetics, Huntsville, AL) was digested with 100 U of the restriction enzymes HindIII, BamHI, and MspI (Roche Diagnostics, Mannheim, Germany) overnight. The recovered DNA was loaded onto 0.8% agarose gel and run overnight. Alkali blots were prepared using the Hybond N⫹ nylon membrane (Amersham Pharmacia Biotech). Prehybridization, hybridization, and probe labeling were conducted using the AlkPhos DIRECT kit (Amersham Pharmacia Biotech). One hundred nanograms of DNA was used for the labeling of the probe. After the overnight hybridizations, the DNA was washed according to the AlkPhos DIRECT protocol. Following the application of the chemiluminescent detection reagent CDP-Star (from the kit), Hyperfilm ECL (Amersham Pharmacia Biotech) was exposed to the blot for 6 h and developed.

Map construction The position of genes on the zebrafish map was determined by using the SAMapper program (31) (K. McKusick and D. R. Cox, unpublished data) on a DECstation 3000-600, following the standard procedure described in the SAMapper manual. Logarithm of the odds score limits and other parameters were set as described in Geisler et al. (32).

Nomenclature The zebrafish genes are designated by the same symbols as their human homologues. Where required by the context, the zebrafish symbols are prefixed by Dare, for Danio rerio. The human gene symbols are according to the Online Mendelian Inheritance in Man (OMIM) homepage (33). Symbols of genes not yet entered in OMIM are according to The MHC Sequencing Consortium (34).

Results Identification of HLA class III region homologues The specific aims of the study were to identify zebrafish genes homologous to the HLA/H2 class III region genes and to determine their positions in the zebrafish genome, in particular their linkage relationship to the previously identified Danio rerio (Dare) Mhc genes in linkage groups 19 (class I), 4 (class II), and 8 (class II). Selected loci more distantly linked to the HLA/H2 complexes were also targeted by the study to assess the limits of the expected synteny. In the search for the homologous loci, four approaches were applied (Table II). First, for each HLA class III locus, nucleotide sequences of known orthologs in other species were obtained from the database (GenBank), the sequences were aligned, conserved segments identified, and degenerate oligonucleotide primers based on them were synthesized. The primers were used to amplify zebrafish genomic DNA, a cDNA library (35), or a PAC library (30) by PCR. Candidate amplification products were cloned

EVOLUTION OF Mhc CLASS III REGION and sequenced to establish their identity. Second, amino acid sequences of proteins encoded in HLA and linked genes were used to conduct tBlastn searches (36) of the EST database (http://www. ncbi.nlm.nih.gov/blast/). If fish homologues identified by a high score were found, the identified ESTs were compared with the nonredundant compartment of the GenBank, EMBL, and DDBJ databases. The EST sequences were then used to design specific PCR primers. Third, sequences of previously identified zebrafish genes deposited in GenBank were used for designing PCR primers. Fourth, two DNA probes from other species, winter flounder RAB2L (37) and X. laevis HSPA1A (38), were used to screen zebrafish cDNA libraries (30) and a genomic (phage) DNA library (18), respectively. One of the four positive clones identified after rescreening was digested with restriction enzymes and the digest was hybridized with the same probe. The two hybridizing fragments were cloned, sequenced, and assembled into one intronless HSPA1A gene (GenBank accession number AF210640). In all four approaches, the identity of the PCR fragments obtained with specific primers was confirmed by cloning and sequencing. The homology of the identified candidate genes with corresponding HLA genes was assessed by conducting a Blastx search of the nonredundant GenBank/EMBL/DDBJ databases with the candidate gene. The HLA class III region contains a total of ⬃60 loci, of which more than half were targeted by the present study. Of the 31 loci analyzed, we failed to identify the homologues of 12 loci (NOTCH4, AGER, AGPAT1, CREBL1, CYP21A2, C4, TNXA, C2, NEU1, LTA, LTB, and NFKBIL1). The failure does not mean that these loci are absent in the zebrafish; they have just not yielded to our identification efforts to date. The 19 successfully identified zebrafish homologues of the HLA class III loci are listed in Table II. In addition to these, we have been able to identify six zebrafish homologues of loci that reside in the HLA class II region, but are neither class II or class I loci (BING3, RAB2L, RPS18, SACML2, RING1B, and KE4), and one nonclass I gene (TUBB) homologous to a gene in the HLA class I region. Of the loci on the centromeric flank of the HLA complex, we identified the homologues of the MUT, NFYA, RDS, HMGIY, and PIM1 (39) loci. Similarly, of the loci on the telomeric flank of the HLA complex, we identified in this study the zebrafish homologue of the EDN1 locus, and in an earlier study (40) the zebrafish homologues of the TCP1 and ACAT2 loci, which are located on human chromosome 6q26-q27. The Dare-DBB and -DCB class II genes were described by Su¨ltmann et al. (18). Altogether, we assembled a collection of 36 zebrafish homologues of HLA or linked loci suitable for a mapping study. Of the additional four genes listed in Table II, three (EF1A, GTF2H4, and DSP) were identified on a public website (http:// wwwmap.tuebingen.mpg.de) as already mapped, and one (FLOT1) was described by Michalova´ et al. (19). Tests of orthology Blastx searches established that the identified zebrafish sequences were homologous to genes on human chromosome 6p, in the HLA complex or in its vicinity. However, homology can be of two kinds: two genes can be either orthologous (divergent by a speciation event) or paralogous (derived by a gene duplication) (41). To be able to compare the zebrafish and human Mhc-associated genes, it is necessary to determine of what kind the homology of the genes from the two species is. We approached this issue from two different angles. The first approach consisted of phylogenetic reconstructions based on the gene or protein sequences. If the reconstruction revealed the existence of a clade containing both the HLA-linked gene (protein) and its zebrafish homologue, the two genes were considered to be orthologous. In cases in which the analysis yielded only one major clade, but the branching pattern


6987

Table II. Zebrafish genes and their positions on linkage groups as determined by typing radiation hybrid panels

Genea

LGb ⫹

MUT NFYA RDS HMGIY PIM1 BING3⌿ RAB2L RPS18 SACM2L RING1B KE4 DBB DCB PBX2 PPT2 STK19 SK12W BF G9A HSPA1A

[20 ] 11⫹ 12⫹ 23⫹ 8⫹ 7⫹ 16⫹ 19⫹ 19⫹ 2⫹ 19⫹ 18⫹ 8⫹ 19⫹ 19⫹ 3⫹ 19⫹ 21⫹ 19⫹ 3⫹

SMX5 VARS1 CLIC1 DDAH BAT5 CSNK2B

[6⫹] 16⫹ 21⫹ ND 16⫹ 19⫹

APOM BAT3 BAT2 AIF1 ATP6G BAT1 FLOT1e TUBB

2⫹ 15⫹ 19⫹ 5⫹ ND 1ⴱ 19⫹ –

EDN1 TCP1 ACAT2 EF1A f GTF2H4 f DSP f

⫹

19 23⫹ 20⫹ 19⫹ 19⫹ 19⫹

PAC Clone (Resource Center)

ND ND ND ND ND B12152, C0485, J15123, J15193, J24117, O2379 ND D12166, I17148, K01108 D12166, I17148, K01108 ND ND ND ND B02262, C1662, G12114, J08121, N24131, P0619 N1074 H06241 ND L04161, O16162 ND D2039, D1966, P1471, D05179, N21190 O23193, C19222, D12222, J15263, G17264 C04161, C13269, G0467, L14231, O0745, P0682 E0983, M12264, P05106 B14149, H13195, K18259, E09106 ND P13230, E15135 B1220, D1057, D1775, E0846, F0778, F17180, I0942, L1023, M1288 ND ND B1220, D1775, F0778, F17180, I0942, M1288 A22199, M05190 ND G12245, N1942 ND C09194, D04241, H16228, I24239, L22201, P17250 ND O23263 G1276, G23214, N20142 ND ND ND

GenBank Accession Number

Idc

Phylogenyd

Location of Human Homolog

AA605843 AI496965 AF210643/4 AI658166 AF062643 AF210638 AF202722 AF210641 AA494847 AF196346 AF196345 U08869 U08873 AF210642 AI545830 AF210646 AI545057 AF047412 AI497334 AF210640

2 2 1 2 3 1 2 1 1 1 2 3 3 1 2 1 2 3 2 4

O O O O O ND O O O P O O O O O O O O O O

6p21.2–6p21.1 6p21.1 6p21.2 6p21 6p21 Class II region Class II region Class II region Class II region Class II region Class II region Class II gene Class II gene Class III region Class III region Class III region Class III region Class III region Class III region Class III region

33 33 33 33 33 34, 34, 34, 34 34, 34, 18 18 34, 34, 34, 34, 34, 34, 34,

AF210645 AF210648 AA497337 AI477385 AI477821 S76877

1 1 2 1 2 1

O O O P O O

Class Class Class Class Class Class

III III III III III III

region region region region region region

34, 34, 34, 34, 34, 34,

64 58, 65 56, 60 56 66 67

AI497429 AW127905 AI588507 AA495202 AF210636 AI477441 AF182161 AF210647

2 2 2 2 1 2

O O O O P O?3P O ND

Class Class Class Class Class Class Class Class

III region III region III region III region III region III region I region I region

34, 34, 34, 34, 34, 34, 34 34,

66 66 56, 58, 66 56 56 56, 66

AI396807 AF143493 AF143488 X77689 AA497366 AI396965

2 3 3

1

O O O O O O

6p24 6q26–6q27 6q26–6q27 6q14 6p21.3 6p24

References

42, 55 42, 55, 56 42, 55, 56 56 42, 55, 57 56, 56, 60 61 62, 56, 56,

58 59 63 62 58, 62

56

33 33, 40 33, 40 33 33 33

a

Genes for which paralogs on other human chromosomes have been reported are indicated in bold. LG, Linkage group (brackets indicate that the support for the assignment is not, as yet, significant). ⫹, T51 radiation hybrid panel used; *, LN54 radiation hybrid panel used. Method of identification (see text). d O, Ortholog; P, paralog. e Michalova´ et al. (19). f The location of these genes is taken from the http://www.map.tuebingen.mpg.de. b c

was consistent with the known vertebrate phylogeny, the zebrafish gene was also considered orthologous to its human counterpart. If the zebrafish gene grouped in a clade with a human gene known to be located on a different chromosome than 6p, the HLA-linked and zebrafish genes were assumed to be paralogous. The analysis revealed the identified zebrafish RING1B, DDAH, ATP6G, and possibly BAT1 genes to be paralogs of the corresponding HLA-linked genes. Attempts to identify the zebrafish orthologs of the four human genes failed. The HLA-associated BING3 is a pseudogene whose relationship to the identified zebrafish gene was difficult to establish because of its accelerated evolution. It appears to be a late acquisition to the HLA region (42). The TUBB gene did not cluster with the mammalian Mhc-linked genes, but because only 60 aa residues were available for the analysis and most of the bootstrap values of the tree were low, the results of the analysis must be regarded as inconclusive. All other identified zebrafish genes behaved as orthologs of their human counterparts in the phylogenetic

analysis (Table II). Examples of the analysis are shown in Fig. 1. Phylogenetic analysis of the zebrafish BF gene is described in Gongora et al. (43), and that of TCP1 and ACAT2 genes in Takami et al. (44) and Shintani et al. (40). The BF and RDS genes are each found in two copies in the zebrafish genome, both copies behaving as orthologs of the single HLA-linked gene by the above criterion (43). We assume that they are the result of a recent gene duplication in the lineage leading to the zebrafish. Therefore, strictly speaking, in each of these two cases, it was the ancestor of each gene pair that was orthologous to the HLA-linked gene, whereas the two copies are paralogous. The recent origin of the two BF copies is supported by their high sequence similarity and by the fact that the copies are closely linked to each other (43). In the second approach, we used short stretches of coding sequence from the candidate genes (obtained by PCR amplification) as probes for hybridization of Southern blots. The target of the hybridization was genomic DNA isolated from either the whole

6988

EVOLUTION OF Mhc CLASS III REGION

FIGURE 1. Examples of phylogenetic analyses conducted to determine homology between zebrafish and other metazoan proteins. The PBX2 and FLOT1 are examples of proteins encoded in orthologous genes on human chromosome 6 and zebrafish linkage group 19. RAB2L is an example of a protein encoded by orthologous genes in the HLA region, but in zebrafish linkage group 16. RING1B is an example of a protein encoded by a zebrafish gene (linkage group 2) paralogous to the HLA-RING1 gene. Trees were produced by the neighbor-joining method of Saitou and Nei (26). Bootstrap values indicating the percentage of times a node could be recovered in 1000 replications are indicated. Sequences are identified by their accession codes in the GenBank/ EMBL/DDBJ and Swissprot databases.

body of the fish or from selected radiation hybrid cell lines, and digested with the restriction enzymes BamHI, MspI, and in some cases also HindIII. Because of their short length (100 – 600 bp), the probes used could be expected to hybridize to a single fragment in most cases, unless the genome contained more than one hybridizing gene. The presence of additional fragments could indicate that the candidate gene might not be orthologous to the HLA-linked gene. Examples of Southern blot hybridizations are shown in Fig. 2. We concentrated on genes that in the phylogenetic analysis were identified as orthologs of HLA class III genes, but that in linkage analyses (see next section) did not show evidence of close proximity to either Mhc class I or class II loci. Of these, the probes derived from the BAT3, CLIC1, and APOM genes hybridized with a single fragment in both the BamHI and MspI digests (Fig. 2A). Therefore, these three zebrafish loci, unlike their human orthologs, are not linked to either the class I or the class II loci. However, hybridization with probes derived from the remaining genes in this set revealed two (STK19, VARS1), three (AIF1, BAT5), four (BAT1), or multiple (HSPA1A) positive fragments (Fig. 2A). Therefore, the possibility existed that a gene borne by one of these additional fragments was not amplified by the PCR primers, but did hybridize with the probe. This gene might be linked to the class

I/II loci, but would not be detected by the linkage test. To examine this possibility, we hybridized the probes with DNA isolated from the zebrafish-hamster radiation hybrid cell lines 47, 53, and 56, which were part of the panel used in the linkage tests described in the next section and which together covered the entire class I region. These tests have revealed the presence of the class I and linked genes in the three cell lines, but the absence of genes from many other linkage groups. Therefore, absence of hybridization with the DNA samples from the three cell lines would support the conclusion that none of the multiple fragments carries a gene from the vicinity of the zebrafish class I loci. An example of this test is shown in Fig. 2B. Here, the HSPA1A probe hybridized with multiple fragments in the positive control (zebrafish genomic DNA), cross-hybridized with an unidentified fragment in the negative control (hamster genomic DNA), but did not hybridize with any of the zebrafish DNA fragments present in the cell lines 47, 53, and 56. In the positive control, the cell line DNAs hybridized with probes specific for zebrafish genes shown by linkage tests to reside in linkage group 19 (Fig. 2C). Similar results were obtained also with the other genes. Therefore, in all these cases, it can be concluded that none of the homologues of the HLA class III genes is near the class I genes of the zebrafish.


FIGURE 2. Examples of Southern blot hybridization with probes specific for zebrafish homologues of HLA class III genes. A, Zebrafish genomic DNA was digested by BamHI (B), MspI (M), or HindIII (H) restriction enzymes, and the blot was hybridized with a probe derived specific for the gene indicated at the top. B, DNA extracted from hamsterzebrafish radiation hybrid cell lines 47, 53, and 56 and digested with BamHI was hybridized with the HSPA1A probe. C, DNA extracted from cell lines 47 and 56 and digested with MspI was hybridized with the PBX2 probe. The positive and negative controls were hamster (HA) and zebrafish (ZF) genomic DNAs, respectively.

Mapping of genes In an earlier study (15), the zebrafish class I genes were mapped to linkage group 19 and the class II genes to linkage groups 8 (Mhc-DAA, -DAB, -DDB) and 4 (Mhc-DFB). To map the zebrafish homologues of the HLA class III and selected flanking genes on human chromosome 6p, primers specific for the individual genes (Table I) were used to amplify DNA samples isolated from the 94

6989 cell lines of the T51 zebrafish-hamster radiation hybrid panel (45). In the few cases in which the localization of genes to a linkage group was not statistically significant, another radiation hybrid panel, LN54 (29), was also tested. Only those PCR amplification results were considered in which amplification of genomic zebrafish DNA (positive control), but not of genomic hamster DNA (negative control), yielded the expected fragment. The results of the PCR amplifications were evaluated by logarithm of the odds score analysis, as described in Geisler et al. (32). Altogether the map positions of 37 genes were determined (Table II, Fig. 3). The TUBB-specific primers cross-amplified hamster DNA and, for this reason, the position of this gene in the zebrafish linkage group maps could not be determined. The ATP6G and DDAH homologues were not mapped because they turned out to be paralogous to the HLA-linked genes. Of the 37 genes, 10 mapped to the zebrafish linkage group 19 (Fig. 4) at differing distances from the class I region contig defined by Michalova´ et al. (19). The remaining 27 genes were scattered among 15 of the 25 zebrafish linkage groups (1, 2, 3, 5, 6, 7, 8, 11, 12, 15, 16, 18, 20, 21, 23). Because only a few of the tested genes that were not in linkage group 19 were found to be together in the same linkage group, we assume that whatever was responsible for the difference in their location between zebrafish and human was a random process. Of the 17 zebrafish genes that could be mapped and that were homologous to HLA class III region genes, six (PBX2, PPT2, SKI2W, G9A, CSNK2B, and BAT2) could be assigned to linkage group 19 (Fig. 4). The remaining 11 genes were scattered among eight different linkage groups (1, 2, 3, 5, 6, 15, 16, and 21). Only the genes STK19 and HSPA1A, VARS1, and BAT5, as well as BF and CLIC1 mapped to the same linkage groups (3, 16, and 21, respectively). Of the six tested zebrafish homologues of HLA class II region genes, three (RPS18, SACM2L, and KE4) mapped to the linkage group 19, in the vicinity of the class I region contig. The two new zebrafish class II genes tested (Dare-DBB and -DCB) mapped to linkage groups 18 and 8, respectively. The former is a third linkage group harboring class II genes; the other two are linkage groups 4 and 8 (15). The human BING3, as mentioned earlier, is a pseudogene, apparently a late acquisition to the HLA region (42), and so it is not surprising that its zebrafish homologue does not map to the linkage group 19. Similarly, the zebrafish RING1B is apparently a paralog of the corresponding HLA complex gene, so here too, its position outside linkage group 19 might have been expected. Of the eight zebrafish genes homologous to genes on human chromosome 6p outside of the HLA complex, only one (EDN1) mapped to linkage group 19. These results lead to the conclusion that there is a partial conservation of synteny between the HLA class III region and part of the zebrafish linkage group 19. In the HLA class III region, the conserved synteny genes are interspersed with genes whose homologues in the zebrafish are located in various other linkage groups. However, in the zebrafish, four of the six conserved class III synteny genes (SKI2W, CSNK2B, BAT2, and PPT2) are part of a single cluster that contains the DSP gene, whose human homologue is on chromosome 6p, but not in the HLA region. The six genes that comprise the conserved synteny group are scattered over the entire length of the HLA class III region and are interspersed among loci that are found in different linkage groups in the zebrafish. The conserved synteny loci thus form a core of the HLA class III region to which genes have either been added during the evolution of the mammalian Mhc, or from which loci have been deleted during the evolution of bony fishes, or both. The finding that the orthologs of the remaining 11 HLA class III region loci are scattered over eight zebrafish linkage groups suggests that the movement into and out of the core class III region occurred on

6990

EVOLUTION OF Mhc CLASS III REGION

FIGURE 3. Zebrafish linkage groups to which genes discussed in the present study have been mapped. Markers most closely related to the mapped genes are shown in small print.

multiple occasions and each time affected either a single locus or a small group of loci on a short chromosomal segment. It should be possible to determine the length of the transferred chromosomal segments by physical mapping. We have taken the first step toward this goal by screening zebrafish PAC and YAC libraries available at the Resource Center of the Max-Planck-Institute for Molecular Genetics (Berlin, Germany). Filters containing the PAC library 706 were screened with probes for the class III genes PBX2, PPT2, STK19, HSPA1A, SMX5, VARS1, CLIC1, BAT1, BAT2, BAT5, and AIF1 (Table II). The digested DNA of the positive PAC clones was then hybridized again with probes that gave positive signals with these clones during the screening. Only one pair of probes (CSNK2B and BAT2) was found to hybridize to the same six PAC clones. When this approach was extended to other homologues of the HLA genes, another pair of genes (RPS18 and SACM2L in the HLA class II region) was found to hybridize to a set of three PAC clones (Table II). Therefore, genes in these two pairs are closely linked in both the zebrafish and humans, and have apparently been moved together during the remodeling of the Mhc region. PCR screening of the HACHy914 YAC library with primers listed in Table I confirmed the close linkage of CSNK2B to BAT2 and of RPS18 to SACM2L, but did not reveal any additional linkages.

Discussion The present data must be interpreted in the context of earlier studies on the zebrafish class I region in linkage group 19. The core of this region is a chromosomal segment covered by a PAC clone contig of about 450 kb (19). It contains a variable, haplotypedependent number of class I loci (Dare-UAA through -UFA), as well as loci concerned with the production (PSMB8, PSMB9A, PSMB9B, PSMB9C, PSMB11, PSMB12) and transport (ABCB2, ABCB3) of peptides and their loading onto nascent class I molecules (TAPBP) (17, 19 –21, 46). It also contains additional loci not known to be involved with either class I molecules or with immunity in general (FLOT1, KNSL2, BING1, DAXX, KE6, FSRG1,

COL11A2, RXRE) (19, 46). The present study adds to this region the loci EDN1, G9A, and PBX2, which flank the contig on one side, as well as the loci RPS18, SACM2L, KE4, PPT2, DSP, BAT2, CSNK2B, SKI2W, and GTF2H4, which flank the other side of the contig. Of these, six loci (G9A, PBX2, NG3, BAT2, CSNK2B, SKI2W) are orthologs of HLA class III region loci; three (RPS18, SACM2L, KE4) are orthologs of HLA class II region genes; and the human orthologs of the remaining three loci (EDN1, DSP, and GTF2H4) are on the human 6p chromosome, but at some distance from the HLA complex. Taken together, the conserved synteny between the human chromosome 6p and the zebrafish linkage group 19 encompasses at least 27 loci (not counting the class I loci and obvious duplicates such as the PSMB9A and PSMB9C loci in the zebrafish). This is the largest conserved synteny between fishes and mammals recorded to date. The synteny is in the composition (gene content) of the conserved chromosomal segments and much less in their organization (gene order). To facilitate the comparison of the conserved segments, it is convenient to divide them somewhat arbitrarily into the four blocks depicted in Fig. 5 and designated A through D. The blocks have been rearranged relative to one another (Fig. 5E) and, to a lesser extent, internally, during the evolution of bony fishes and mammals from their common ancestor. Block A contains four genes, two of which appear to be in inverted order in the zebrafish relative to the human. The orientation of the block, too, is reversed in the two species, as is that of block B. The latter consists of six loci arranged in the same order in the zebrafish and the human, but the fish ABCB and PSMB9 loci have duplicated, and one locus (KE6) has been transposed from block C in the zebrafish or to block C in the human. It is this block that may have been the original integration site of the class I and class II loci when the Mhc arose in the ancestor of the jawed vertebrates or earlier. The zebrafish RXRE is one of two closely related loci (47), which presumably arose by duplication of the RXRB ortholog (43); the other gene, RXRD, is in linkage group 16, which, perhaps significantly, contains three other orthologs of


6991

FIGURE 4. Comparison of syntenic human and zebrafish groups. The data on the class I genomic region were reported previously (19).

HLA-associated genes (VARS1, BAT5, and RAB2L). Block C is comprised of four loci in the human and three loci in the zebrafish, the KE6 locus presumably having moved out of the latter or moved into the former. In both species, the loci are arranged in the same order, but the block’s orientation has been reversed in one species relative to the other. In the human, blocks A, B, and C encompass the HLA class II region. Block D is the largest and the most rearranged of the four. In the human, it covers the HLA class III region in the band 6p21.3, but it also includes a subblock of genes in bands 6p24.1 (EDN1) and 6p25 (DSP). In the human chromosome 6, these two subblocks are thus separated by a long genetic and physical distance; in the zebrafish, in contrast, they are not only closer together, but also intermixed. When suitable markers become available, it will be interesting to find out whether the region of synteny conservation also includes the segment between the two subblocks in human chromosome 6. The order of five block D loci is conserved between the human and the fish chromosomes, while five other loci are rearranged in one species relative to the other. The human block D contains at least 12 additional loci (not counting the class I genes) that are apparently absent in the zebrafish block D. Whether the zebrafish block D similarly contains loci absent in the human counterpart remains to be determined. There may exist another block of synteny conservation at the telomeric end of the long arm of chromosome 6 (band 6q27). The putative block is to date marked only by the Brachyury (T) homologue, which in the zebrafish is at a distance of ⬃4 –7 cM from the class I region (15, 48). However, the TCP1 and ACAT2 loci, which in the human reside in band 6q26-q27 and are partially overlapping, are in different linkage groups in the zebrafish (40; and the present study). Interestingly, TCP1 in the zebrafish linkage group 23 is closely linked to HMGIY, and ACAT2 in zebrafish linkage group 20 is loosely linked to MUT; the human HMGIY and MUT genes are closely linked to the HLA complex. Based on these findings, we suggest that many of the genes that are now part of the vertebrate Mhc or of the flanking segments are part of an ancient synteny group that existed before the divergence

FIGURE 5. Blocks of conserved synteny between human (Hosa) chromosome 6p and zebrafish (Dare) linkage group 19 genes. The zebrafish class I genomic region is represented by the blocks A and B (19); the blocks C and D refer to the other genes of the conserved synteny between the human chromosome 6p and the zebrafish linkage group 19. Relative orientation of the blocks in the two species is indicated by long arrows. Arrowheads indicate gene insertions (I and II, Mhc class I and class II genes, respectively). Relative arrangement of the blocks in the two species is shown in the lower part of the figure.

of the bony fish and tetrapod lineages more than 400 million years ago. That the group might, in fact, be much older than 400 million years is suggested by the existence of a region on human chromosome 6 displaying conserved synteny with the genomes in Caenorhabditis elegans and Drosophila melanogaster (49). As in that synteny, in the conserved synteny described in this study, there are ancient genes whose homologues have been found in vertebrates, as well as nonvertebrates. In contrast, the class I and class II Mhc loci are apparently absent in nonvertebrates and may have originated after the divergence of jawless and jawed vertebrates (50). Thus, before the emergence of the Mhc, there apparently existed a pre-Mhc region with many of the non-class I and non-class II loci already in place, including some of the class III loci, but lacking the class I and class II genes. Whether the class I and class II loci arose in situ from genes already present in the pre-Mhc region, or from genes elsewhere in the genome and were then transposed into the pre-Mhc region, is unclear. Abi Rached and his colleagues (51) have pointed out the presence of Ig superfamily genes (exons) in the vicinity of the vertebrate Mhc. Some of these genes may have donated the Ig-like exons of the class I (␣3 domain-encoding) and class II (␣2 and ␤2 domain-encoding) genes. However, there are

6992 no known genes in the Mhc region or elsewhere in the genome that could be considered good candidates for donors of the peptidebinding region-encoding exons of the class I and class II genes (7). The reasons for the conservation of synteny between the preMhc segments remain obscure. The gene composition of the preMhc segment does not give any indication for specialization in nonadaptive immune response. None of the genes suggested as forming a functional immunological supercluster in the mammalian Mhc (3) are part of the conserved synteny group in the zebrafish. There is to date no evidence for the existence of complement factor 2 (C2) gene in bony fishes (52); this gene may have arisen in tetrapods. The complement factor B (BF) gene is in the zebrafish linkage group 21, and in the medaka fish, too, it is in a different linkage group than any of the class I and class II loci (53). The zebrafish complement factor 4 (C4) locus could not be identified in the present study, but in medaka, it is again in a different linkage group than the class I and class II loci (53). The zebrafish HSPA1A gene is in linkage group 3, which apparently lacks class I and class II loci. Finally, all efforts to identify the zebrafish orthologs of the human TNF (TNF) and the lymphotoxin (LT) genes, either in genomic DNA or in PAC and YAC clones bearing class III region homologues, have failed to date (unpublished data). Therefore, we conclude that if there is any significance in the clustering of all these genes in the class III region of the mammalian Mhc, it apparently does not extend beyond tetrapods or possibly only some mammals. Teleost fishes, which comprise more than one-half of jawed vertebrates in terms of the number of identified species, apparently do not suffer any disadvantage from not having some of the genes of nonadaptive immune response linked to the class I or class II loci. However, there may be a selective advantage in having genes for protein processing and for transport, as well as for the loading of peptides, linked to the class I loci. The rest of the conserved synteny genes may have remained together perhaps for reasons having to do with the organization of chromatin loops (54). However, even this explanation must make generous allowance for the busy traffic of genes in and out of the conserved regions.

Acknowledgments We thank Dr. Susan Douglas for the RAB2L and Dr. Vladimir Vincek for the HSPA1A gene probes, Dr. Colm O’hUigin for critical reading of the manuscript, Sabine Jantschek and Silke Geiger-Rudolph for technical assistance, and Jane Kraushaar for editorial assistance. We are grateful to the staff of the Resource Center for their support.

References 1. Klein, J. 1986. Natural History of the Major Histocompatibility Complex. John Wiley, New York. 2. Kasahara, M. 2000. Major Histocompatibility Complex: Evolution, Structure, and Function. Springer-Verlag, Tokyo. 3. Trowsdale, J. 1995. “Both man & bird & beast”: comparative organization of MHC genes. Immunogenetics 41:1. 4. Parham, P., ed. 1999. Genomic organization of the MHC: structure, origin and function. Immunological Reviews, Vol. 167. Munksgaard International Publishers, Copenhagen. 5. Parham, P., ed. 1999. Pathways of antigen processing and presentation. Immunological Reviews, Vol. 172. Munksgaard International Publishers, Copenhagen. 6. Kasahara, M., M. F. Flajnik, T. Ishibashi, and T. Natori. 1995. Evolution of the major histocompatibility complex: a current overview. Transplant. Immunol. 3:1. 7. Klein, J., and C. O’hUigin. 1993. Composite origin of major histocompatibility complex genes. Curr. Opin. Genet. Dev. 3:923. 8. Klein, J., and A. Sato. 1998. Birth of the major histocompatibility complex. Scand. J. Immunol. 47:199. 9. Skow, L. C., S. N. Snaples, S. K. Davis, J. F. Taylor, B. Huang, and D. H. Gallagher. 1996. Localization of bovine lymphocyte antigen (BoLA) DYA and class I loci to different regions of chromosome 23. Mamm. Genome 7:388. 10. Band, M., J. H. Larson, J. E. Womack, and H. A. Lewin. 1998. A radiation hybrid map of BTA23: identification of a chromosomal rearrangement leading to separation of the cattle MHC class II subregions. Genomics 53:269.

EVOLUTION OF Mhc CLASS III REGION 11. Lewin, H. A., G. C. Russel, and E. J. Glass. 1999. Comparative organization and function of the major histocompatibility complex of domesticated cattle. Immunol. Rev. 167:145. 12. Chouchane, L., and T. J. Kindt. 1992. Mapping of the rabbit MHC reveals that class I genes are adjacent to the DR subregion and defines an insertion/deletionrelated polymorphism in the class II region. J. Immunol. 149:1216. 13. Kaufman, J., J. Jacob, I. Shaw, B. Walker, S. Milne, S. Beck, and J. Salomonsen. 1999. Gene organization determines evolution of function in the chicken MHC. Immunol. Rev. 167:101. 14. Flajnik, M. F., Y. Ohta, C. Namikawa-Yamada, and M. Nonaka. 1999. Insight into the primordial MHC from studies in ectothermic vertebrates. Immunol. Rev. 167:59. 15. Bingulac-Popovic, J., F. Figueroa, A. Sato, W. S. Talbot, S. L. Johnson, M. Gates, J. H. Postlethwait, and J. Klein. 1997. Mapping of Mhc class I and class II regions to different linkage groups in the zebrafish. Danio rerio. Immunogenetics 46:129. 16. Sato, A., F. Figueroa, B. W. Murray, E. Ma´laga-Trillo, Z. Zaleska-Rutczynska, H. Su¨ltmann, S. Toyosawa, C. Wedekind, N. Steck, and J. Klein. 2000. Nonlinkage of the major histocompatibility complex of class I and class II loci in bony fishes. Immunogenetics 51:108. 17. Takeuchi, H., F. Figueroa, C. O’hUigin, and J. Klein. 1995. Cloning and characterization of class I Mhc genes of the zebrafish Brachydanio rerio. Immunogenetics 42:77. 18. Su¨ltmann, H., W. E. Mayer, F. Figueroa, C. O’hUigin, and J. Klein. 1994. Organization of Mhc class II B genes in the zebrafish (Brachydanio rerio). Genomics 23:1. 19. Michalova´, V., B. W. Murray, H. Su¨ltmann, and J. Klein. 2000. A contig map of the Mhc class I genomic region in the zebrafish reveals ancient synteny. J. Immunol. 164:5296. 20. Takami, K., Z. Zaleska-Rutczynska, F. Figueroa, and 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. 21. Murray, B. W., H. Su¨ltmann, and J. Klein. 1999. Analysis of a 26-kilobase region linked to the Mhc in zebrafish: genomic organization of the proteasome component ␤/transporter associated with antigen processing-2 gene cluster and identification of five new proteasome ␤ subunit genes. J. Immunol. 163:2657. 22. Saiki, R. K., D. H. Gelfand, S. Stoffel, S. J. Scharf, R. Higuchi, G. T. Horn, K. B. Mullis, and H. A. Erlich. 1988. Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239:487. 23. Altschul, S. F., W. Gish, W. Miller, E. W. Myers, and D. J. Lipman. 1990. Basic local alignment search tool. J. Mol. Biol. 215:403. 24. Zhang, J., and T. L. Madden. 1997. PowerBLAST: a new network BLAST application for interactive or automated sequence analysis and annotation. Genome Res. 7:649. 25. Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. Improved sensitivity of profile searches through the use of sequence weights and gap excision. Comput. Appl. Biosci. 10:19. 26. Saitou, N., and M. Nei. 1987. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4:406. 27. Felsenstein, J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39:783. 28. Kwok, C., R. Critcher, and K. Schmitt. 1999. Construction and characterization of zebrafish whole genome radiation hybrids. Methods Cell Biol. 60:287. 29. Hukriede, N. A., L. Joly, M. Tsang, J. Miles, P. Tellis, J. A. Epstein, W. B. Barbazuk, F. N. Li, B. Paw, J. H. Postlethwait, et al. 1999. Radiation hybrid mapping of the zebrafish genome. Proc. Natl. Acad. Sci. USA 96:9745. 30. In Resource-Center/Primary-Database. (RZPD of the German Human Genome Project at the Max-Planck-Institute for Molecular Genetics (former Reference Library Database.) Heubnerweg, Berlin-Charlottenburg, Germany. URL: http:// www.rzpd.de. 31. Stewart, E. A., K. B. McKusick, A. Aggarwal, E. Bajorek, S. Brady, A. Chu, N. Fang, D. Hadley, M. Harris, S. Hussain, et al. 1997. An STS-based radiation hybrid map of the human genome. Genome Res. 7:422. 32. Geisler, R., G. J. Rauch, H. Baier, F. van Bebber, L. Brobeta, M. P. Dekens, K. Finger, C. Fricke, M. A. Gates, H. Geiger, et al. 1999. A radiation hybrid map of the zebrafish genome. Nat. Genet. 23:86. 33. In National Center for Biotechnology Information: Omim: Online Mendelian Inheritance in Man. http://www.ncbi.nlm.nih.gov/Omim/. National Library of Medicine, Bethesda, MD 34. The MHC Sequencing Consortium. 1999. Complete sequence and gene map of a human major histocompatibility complex. Nature 401:921. 35. Ono, H., D. Klein, V. Vincek, F. Figueroa, C. O’hUigin, H. Tichy, and J. Klein. 1992. Major histocompatibility complex class II genes of zebrafish. Proc. Natl. Acad. Sci. USA 89:11886. 36. Altschul, S. F., T. L. Madden, A. A. Schaffer, J. Zhang, W. Miller, and D. J. Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25:3389. 37. Douglas, S. E., J. W. Gallant, C. E. Bullerwell, C. Wolff, J. Munholland, and M. E. Reith. 1999. Winter flounder expressed sequence tags: establishment of an EST database and identification of novel fish genes. Mar. Biotechnol. 1:458. 38. Bienz, M. 1984. Xenopus hsp70 genes are constitutively expressed in injected oocytes. EMBO J. 3:2477. 39. Icard-Liepkalns, C., R. N. Haire, S. J. Strong, and G. W. Litman. 1999. Cloning of a cDNA encoding a Pim1 homologue in zebrafish. Danio rerio. Immunogenetics 49:351. 40. Shintani, S., C. O’hUigin, S. Toyosawa, V. Michalova´, and J. Klein. 1999. Origin of gene overlap: the case of TCP1 and ACAT2. Genetics 152:743.

The Journal of Immunology 41. Li, W.-H. 1997. Molecular Evolution. Sinauer Associates, Sunderland. 42. Herberg, J. A., S. Beck, and 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. 43. Gongora, R., F. Figueroa, and J. Klein. 1998. Independent duplications of Bf and C3 complement genes in zebrafish. Scand. J. Immunol. 48:651. 44. Takami, K., F. Figueroa, W. E. Mayer, and J. Klein. 2000. Ancient allelism at the cytosolic chaperonin-encoding gene of the zebrafish. Genetics 154:311. 45. Kwok, C., R. M. Korn, M. E. Davis, D. W. Burt, R. Critcher, L. McCarthy, B. H. Paw, L. I. Zon, P. N. Goodfellow, and K. Schmitt. 1998. Characterization of whole genome radiation hybrid mapping resources for non-mammalian vertebrates. Nucleic Acids Res. 26:3562. 46. Su¨ltmann, H., B. W. Murray, and J. Klein. 2000. Identification of seven genes in the Mhc class I region of the zebrafish. Scand. J. Immunol. 10:577. 47. Jones, B. B., C. K. Ohno, G. Allenby, M. B. Boffa, A. A. Levin, J. F. Grippo, and M. Petkovich. 1995. New retinoid X receptor subtypes in zebrafish (Danio rerio) differentially modulate transcription and do not bind 9-cis retinoic acid. Mol. Cell. Biol. 10:5226. 48. Johnson, S. L., M. A. Gates, M. Johnson, W. S. Talbot, S. Horne, K. Baik, S. Rude, J. R. Wong, and J. H. Postlethwait. 1996. Centromere-linkage analysis and consolidation of the zebrafish genetic map. Genetics 142:1277. 49. Trachtulec, Z., R. M. Hamvas, J. Forejt, H. R. Lehrach, V. Vincek, and 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. 50. Klein, J., A. Sato, and W. E. Mayer. 2000. Jaws and AIS. In Major Histocompatibility Complex: Evolution, Structure, and Function. M. Kasahara, ed. Springer-Verlag, Tokyo. 51. Abi Rached, L., M. F. McDermott, and P. Pontarotti. 1999. The MHC big bang. Immunol. Rev. 167:33. 52. Seeger, A., W. E. Mayer, and J. Klein. 1996. A complement factor B-like cDNA clone from the zebrafish (Brachydanio rerio). Mol. Immunol. 33:511. 53. Nonaka, M., N. Kuroda, K. Naruse, and A. Shima. 1998. Molecular genetics of the complement C3 convertases in lower vertebrates. Immunol. Rev. 166:59. 54. 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. 55. Stephens, R., R. Horton, S. Humphray, L. Rowan, J. Trowsdale, and S. Beck. 1999. Organization, sequence variation and isochore structure at the centromeric boundary of the human MHC. J. Mol. Biol. 291:789.

6993 56. Kasahara, M. 1999. Genome dynamics of the major histocompatibility complex: insights from genome paralogy. Immunogenetics 50:134. 57. Hanson, I., and J. Trowsdale. 1991. Colinearity of novel genes in the class II regions of the MHC in mouse and man. Immunogenetics 34:5. 58. Katsanis, N., J. Fitzgibbon, and E. M. C. Fisher. 1996. Paralogy mapping: identification of a region in the human MHC triplicated onto human chromosomes 1 and 9 allows the prediction and isolation of novel PBX and NOTCH loci. Genomics 35:101. 59. Soyombo, A. A., W. Yi, and S. L. Hofmann. 1999. Structure of the human palmitoyl-protein thioesterase-2 gene (PPT2) in the major histocompatibility complex on chromosome 6p21.3. Genomics 56:208. 60. Sargent, C. A., I. Dunham, and R. D. Campbell. 1989. Identification of multiple HTF-island associated genes in the human major histocompatibility complex class III region. EMBO J. 8:2305. 61. Dangel, A. W., L. Shen, A. R. Mendoza, L. Wu, and C. Y. Yu. 1995. Human helicase gene SKI2W in the HLA class III region exhibits striking structural similarities to the yeast antiviral gene SKI2 and to the human gene KIAA0052: emergence of a new family. Nucleic Acids Res. 23:2120. 62. Spies, T., M. Bresnahan, and J. Strominger. 1989. Human major histocompatibility complex contains a minimum of 19 genes between the complement cluster and HLA-B. Proc. Natl. Acad. Sci. USA 86:8955. 63. Postlethwait, J. H., Y.-L. Yan, M. A. Gates, S. Horne, A. Amores, A. Brownlie, A. Donovan, E. S. Egan, A. Force, Z. Gong, et al. 1998. Vertebrate genome evolution and the zebrafish gene map. Nat. Genet. 18:345. 64. Olavesen, M. G., M. Snoek, and R. D. Campbell. 1993. Localization of a new gene adjacent to the HSP70 genes in human and mouse MHCs. Immunogenetics 37:394. 65. Hsieh, S.-H., and R. D. Campbell. 1991. Evidence that gene G7a in the human major histocompatibility complex encodes valyl-tRNA synthetase. Biochem. J. 278:809. 66. Spies, T., G. Blanck, M. Bresnahan, J. Sand, and J. Strominger. 1989. A new cluster of genes within the human major histocompatibility complex. Science 243:214. 67. Albertella, M., H. Jones, W. Thomson, M. G. Olavesen, and R. D. Campbell. 1996. Localization of eight additional genes in the human major histocompatibility complex, including the gene encoding the casein kinase II ␤ subunit (CSNK2B). Genomics 36:240.