Ancient Allelism at the Cytosolic Chaperonin--Encoding Gene of the Zebrafish

Corresponding author: Felipe Figueroa, Max-Planck-Institut für Biologie, Abteilung Immungenetik, Corrensstrasse 42, D-72076 Tübingen, Germany., felipe.figueroa{at}tuebingen.mpg.de (E-mail)

Communicating editor: N. TAKAHATA

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The T-complex protein 1, TCP1, gene codes for the CCT- subunit of the group II chaperonins. The gene was first described in the house mouse, in which it is closely linked to the T locus at a distance of ~11 cM from the Mhc. In the zebrafish, Danio rerio, in which the T homolog is linked to the class I Mhc loci, the TCP1 locus segregates independently of both the T and the Mhc loci. Despite its conservation between species, the zebrafish TCP1 locus is highly polymorphic. In a sample of 15 individuals and the screening of a cDNA library, 12 different alleles were found, and some of the allelic pairs were found to differ by up to nine nucleotides in a 275-bp-long stretch of sequence. The substitutions occur in both translated and untranslated regions, but in the former they occur predominantly at synonymous codon sites. Phylogenetically, the alleles fall into two groups distinguished also by the presence or absence of a 10-bp insertion/deletion in the 3' untranslated region. The two groups may have diverged as long as 3.5 mya, and the polymorphic differences may have accumulated by genetic drift in geographically isolated populations.

TO attain functional conformation during synthesis or recovery from a denaturated state, proteins require the assistance of molecular chaperones (ELLIS and VAN DER VIES 1991 ; FRYDMAN and HARTL 1996 ). This large family of related molecules includes the chaperonins, a class of proteins with a high degree of sequence conservation in eukarya, bacteria, and archaea (for review see KUBOTA et al. 1995 ). Chaperonins fall into two groups, characterized by the possession of ATPase- and substrate-binding domains. The group I or classical chaperonins (WILLISON and KUBOTA 1994 ) include the soluble M_r 10,000 GroEL (GEORGOPOULUS et al. 1973 ), the mitochondrial hsp60 (CHENG et al. 1989 ), and the Rubisco-subunit-binding protein of chloroplasts and other plastids (HEMMINGSEN et al. 1988 ). Group II includes archaeal chaperonins TF55 and thermosomes, as well as eukaryotic cytosolic chaperonins (CCT). In contrast to the archaeal group II chaperonins, which are composed of only two subunit species, the eukaryotic CCT chaperonins are composed of at least eight types of subunits: , ß, , , , , , and (KIM et al. 1994 ; KUBOTA et al. 1994 , KUBOTA et al. 1995 ).

The mouse CCT- subunit was identified as a t-complex-specific polypeptide, Tcp1, and the encoding gene was mapped to chromosome 17, at a distance of 11 cM from the major histocompatibility complex (Mhc) (see SILVER 1985 ). The CCT--encoding genes have been cloned and sequenced from a number of different species including the mouse (WILLISON et al. 1986 ), hamster (AHMAD and GUPTA 1990), rat (MORITA et al. 1991 ), human (KIRCHHOFF and WILLISON 1990 ), Drosophila (URSIC and GANETZKY 1988 ), Caenorhabditis elegans (LEROUX and CANDIDO 1995A , LEROUX and CANDIDO 1995B ), yeast (URSIC and CULBERTSON 1991 ), Schistosoma mansoni (KIM et al. 1994 ), and Arabidopsis thaliana (MORI et al. 1992 ), as well as other plants (AHNERT et al. 1996 ). Other subunits have also been sequenced and mapped to different chromosomes.

Our laboratory is engaged in the characterization of the chromosomal regions encoding the Mhc of the zebrafish (SULTMANN et al. 1994 ; BINGULAC-POPOVIC et al. 1997 ; TAKAMI et al. 1997 ; GONGORA et al. 1998 ). Linkage studies have shown that, in contrast to all other vertebrate groups characterized thus far, class I and class II Mhc genes are not linked in the teleosts, represented by the zebrafish (BINGULAC-POPOVIC et al. 1997 ). Furthermore, we found that the zebrafish homolog of the mouse Brachyury (T) locus, which in the mouse is located 14.5 cM centromeric to the H2-K class I locus, maps 21.9 cM from the zebrafish class I cluster, suggesting a long-term synteny conservation (BINGULAC-POPOVIC et al. 1997 ). Since the mouse Tcp1 locus is located in the middle of the region encompassing the T and the H2-K loci, we cloned and sequenced the zebrafish TCP1 gene and determined its linkage relationship to the Mhc loci. During these studies, we came across an interesting, but puzzling, case of two very old TCP1 allelic lineages.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Fish:
The laboratory strains of the zebrafish, Danio rerio (abbreviated Dare), were maintained at the Max-Planck Institute for Biology. The strains ORE, KON, and AB were obtained from the Universities of Oregon and Konstanz and the M.P.I. für Entwicklungsbiologie, Tübingen, respectively. Wild-derived strains were established in our laboratory from fishes collected in Singapore, Bengal, and Calcutta. Strains KOL, HPS, and KOC were established from fishes obtained from local dealers. Haploid embryos were produced by in vitro fertilization of zebrafish eggs with irradiated sperm from Pseudotropheus tropheops (see BINGULAC-POPOVIC et al. 1997 for details).

Preparation of genomic DNA:
Total genomic DNA was isolated from fresh or ethanol-preserved adult specimens, as previously described (SATO et al. 1995 ). For the isolation of DNA from haploid tissues, 4-day-old embryos were incubated for 3–4 hr in 50 µl of a solution containing 10 mM Tris-HCl, pH 8.0, 2 mM ethylenediaminetetraacetic acid (EDTA), 0.2% Triton X-100, and 200 µg/ml of proteinase K. The lysate was then incubated for 10 min at 95°, and the DNA was used for amplification by the polymerase chain reaction (PCR) without further purification.

Zebrafish cDNA library:
The library was prepared from the spleens and hepatopancreases of a pool of 20 fishes from the noninbred KOC strain. The cDNA fragments were cloned in the gt10 vector, as previously described (ONO et al. 1993 ).

PCR amplification:
Standard PCR amplification was carried out in the PTC-100 programmable thermal controller (MJR, Biozym, Oldendorf, Germany). A total of 1 µl of a solution containing 1 x 10⁷ plaque-forming units of the zebrafish cDNA library or 100 ng of genomic DNA was added to a reaction mixture consisting of 1x PCR buffer (50 mM KCl, 1.5 mM MgCl₂, 10 mM Tris-HCl, pH 8.3, 0.001% gelatin), 0.2 mM of each of the four deoxynucleoside triphosphates (Pharmacia, Freiburg, Germany), 1 mM of each of the sense and antisense primers (Table 1), and 1 unit of AmpliTaq DNA polymerase (Perkin-Elmer, Überlingen, Germany). In the amplifications, DNA denaturation for 1 min at 94° was followed by 35 cycles, each cycle consisting of 15 sec of denaturation at 94°, 15 sec of annealing at the required temperature depending on the primer combination (Table 1), and 2 min of extension at 72°. The reactions were completed by a final primer extension for 5 min at 72°. Hot-start PCR amplifications were performed as above, except that the 1x PCR buffer was used free of MgCl₂, and 2.5 mM HotWax Mg²⁺ beads (Invitrogen, BV, Groningen, The Netherlands) were added to the mixture.

DNA sequencing:
Selected PCR products were isolated from low-melting-point agarose gels (Life Technologies, Eggenstein, Germany) and cloned in the pUC18 plasmid vector with the aid of the SureClone ligation kit (Pharmacia). Double-stranded DNA prepared with the aid of the Qiagen Plasmid Kit (Qiagen, Hilden, Germany) was resuspended at a concentration of 1 µg/µl and sequenced by the dideoxy chain-termination method (SANGER et al. 1977 ) using the AutoRead sequencing kit (Pharmacia). Six microliters of each sequencing reaction was then loaded onto a 6.6% acrylamide/bisacrylamide (29:1) gel and run in the automated laser fluorescent DNA sequencer (Pharmacia).

Single-stranded conformational polymorphism (SSCP) analysis:
Four microliters of amplified DNA samples were denaturated by incubation for 10 min at 50° in the presence of 50 mM NaOH and 1 mM EDTA and immediately cooled in ice water. After the addition of 1.8 µl of loading buffer, containing 100% formamide and xylene cyanol, the samples were loaded into the wells of 10% minipolyacrylamide CleanGel (ETC elektrophorese-Technik, Kirchentellinsfurt, Germany) and were electrophoresed in the Multiphor II system (Pharmacia) for 10 min at 200 V and then for 3–4 hr at 375 V at a constant temperature of 15°, using the DELECT kit (ETC elektrophorese-Technik). Separated DNA fragments were visualized by silver staining (ORITA et al. 1989 ).

Southern DNA blotting and hybridization:
Genomic fish DNA (10 µg) was digested with restriction endonucleases for 10 hr under the conditions recommended by the supplier (Pharmacia). After the digestion, fragments were separated by agarose gel electrophoresis and blotted onto Hybond-N⁺ nylon filters (Amersham Buchler, Braunschweig, Germany). DNA probe labeling and hybridizations were performed as previously described (SATO et al. 1995 ). Blots were used to expose the Fuji imaging plates, type BAS-III (Raytest, Strauben, Germany), for 2–20 hr at room temperature, and the signals were detected with the Fuji BAS 1000 system (Raytest).

PAC clone screening:
Zebrafish PAC library filters (library BUSMP706) were obtained from the Ressourcenzentrum im Deutschen Humangenomprojekt am Max-Planck-Institut für Molekulare Genetik, Berlin. The filters were hybridized with a TCP1 probe encompassing exons 9–12 at 65° in 7% SDS, 0.5 M sodium phosphate, pH 7.2, 1 mM EDTA, and were washed twice in 40 mM sodium phosphate containing 0.1% SDS. Two TCP1-containing clones were obtained (nos. BUS-MP706H0274Q2 and BUSMP706023263Q2), and the presence of TCP1 in them was confirmed by PCR amplification.

Dendrogram construction:
Evolutionary relationships were evaluated with the help of the MEGA program (KUMAR et al. 1993 ) using Kimura-2-parameter distances and the neighbor-joining algorithm for phylogenetic tree construction (SAITOU and NEI 1987 ). Five hundred bootstrap replications were performed to determine the reliability of the branching order. Maximum parsimony analysis was carried out with the help of the PAUP 3.1.1 software (SWOFFORD 1993 ). Unrooted dendrograms were constructed using the heuristic search algorithm. Zero-length branches were collapsed, and all minimal trees were saved. A strict consensus tree was calculated and displayed as a midpoint-rooted tree.

The sequence data presented in this article have been submitted to the GenBank Data Library under accession nos. AF164028–AF164038.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Isolation of a PCR clone encompassing 334 bp of the zebrafish TCP1 gene:
To identify conserved nucleotide stretches of the TCP1 gene, an alignment of representative TCP1 sequences was prepared. It included -subunit CCT sequences from human, mouse, rat, Drosophila, and yeast (WILLISON et al. 1986 ; URSIC and GANETZKY 1988 ; KIRCHHOFF and WILLISON 1990 ; MORITA et al. 1991 ; URSIC and CULBERTSON 1991 ). Specific sense and antisense oligonucleotide primers were then synthesized and used to amplify a segment encompassing exons 3, 4, and 5 from the cDNA library. The combination Tu1269/Tu1270 (Table 1) produced a 334-bp fragment, which after subcloning and sequencing yielded a TCP1-like sequence encoding 33 codons of exon 3, the whole of exon 4, and 26 codons of exon 5. The clone was designated Dare TCP1-a (Fig 1).

View larger version (80K): In this window In a new window Download PPT slide   Figure 1. Alignment of the zebrafish, Dare cDNA TCP1 nucleotide sequences. Dashes indicate identity with the simple-majority consensus at the top; dots indicate unavailability of sequence information; and asterisks indicate indels introduced to achieve optimal sequence alignment. Exon borders are putatively assigned by inference from the mouse gene organization. The initiation and stop codons are underlined.

Determination of a complete Dare-TCP1 cDNA sequence:
To obtain the complete cDNA sequence derived from the expressed Dare-TCP1 gene, the primer P168, complementary to exon 3 codons 87–94 of the cDNA clone TCP1-a, was used in the PCR in combination with the vector primer to produce a fragment of ~1.6 kb, which was then cloned and sequenced. Two different sequences, T-3-5 and T-3-7, were obtained, which extend from the beginning of exon 4 to the 3' untranslated (UT) region (Fig 1). In the overlapping region, each of the two sequences differs from the cDNA-TCP1-a clone by a single nucleotide substitution. To obtain the 5' end of the gene, the antisense oligonucleotide P166 was used in PCR in combination with the vector primer. A product of 306 bp (primers included) was obtained, which upon cloning and sequencing revealed the presence of only one sequence, T-1-12, encompassing exons 1 and 2 and part of exon 3 up to codon 86 (Fig 1).

To obtain a clone encompassing most of the TCP1 transcript, two new oligonucleotide primers were used (P202 and P203) specific for exon 1 of clone T-1-12 and the 3' UT region of clones T-3-5 and T-3-7, respectively (Table 1). Sequencing of the amplification product yielded a new TCP1 sequence, TA2 (Fig 1). Clones TA2 and T-3-5 are 99.8% identical in the sequence from exon 3 through to exon 12, and they are completely identical in their 3' UT regions. Clone T-3-7 has a 10-bp deletion in the 3' UT region compared with the TA2/T-3-5 clones (Fig 1). A comparison with TCP1 genes of other species indicates, first, that all 12 exons of the consensus TCP1 sequence are apparently represented in the various Dare clones; and second, that the zebrafish TCP1 sequences are homologous to the CCT--encoding genes (Fig 2).

View larger version (30K): In this window In a new window Download PPT slide   Figure 2. Phylogenetic tree showing the relationship of the Dare TCP1 protein sequences with representatives of members of different chaperonin-encoding genes. The dendrogram was constructed by the neighbor-joining method (SAITOU and NEI 1987 ) from genetic distances estimated by Kimura's two-parameter method (KIMURA 1980 ). The numbers after the sequences indicate GenBank accession numbers. The numbers on branches indicate bootstrap values. Species abbreviations are as follows: Arth, A. thaliana; Caal, Candida albicans; Cael, C. elegans; Crgr, Cricetulus griceus; Drme, D. melanogaster; Hosa, Homo sapiens; Mumu, M. musculus; Nafo, Naegleria fowleri; Rano, Rattus norvegicus; Sace, Saccharomyces cerevisiae; Scma, Schistosoma mansoni; Stle, Stylonychia lemnae; Sush, Sulfolobus shibatae; Tepy, Tetrahymena pyriformis.

Allelism of T-3-5 and T-3-7 sequences:
The T-3-5 and T-3-7 clones differ by 16 nucleotide substitutions in the region encompassing exons 4–12. They also differ in the 3' UT region by 10 nucleotides and a 10-bp deletion at position 82-91 of clone T-3-7. To determine whether the two sequences are derived from alleles at a single locus or from two loci, gene-specific primers (P207 and P210; Table 1) were prepared for each clone and were used to amplify DNA from 17 mothers of sets of haploid embryos. Six DNA samples were amplified by both primer combinations, eight were amplified by either the T-3-5-specific or the T-3-7-specific primers, and three samples were not amplified by either of the two primer combinations. The typing of a 20-haploid progeny of one of the double-positive females revealed an antithetical segregation in which 12 embryos inherited the T-3-5 sequence and the remaining eight inherited the T-3-7 sequence. No embryo was typed as double-positive or negative. These results suggest that the two sequences are derived from alleles at a single locus.

To determine the number of TCP1 genes present in the genome of a single individual, we digested genomic DNA from several zebrafishes with either BamHI, EcoRI, HindIII, or TaqI restriction enzymes, blotted the digests onto filters, and hybridized the filters with a 600-bp probe encompassing exons 10–12, including 200 bp of the 3' UT region of the T-3-5 cDNA clone. The hybridization revealed the presence of a maximum of two hybridizing bands per individual (Fig 3), and thus suggested that the zebrafish genome contains a single TCP1 locus.

View larger version (60K): In this window In a new window Download PPT slide   Figure 3. Southern blot hybridizations of selected genomic DNAs derived from laboratory zebrafish stocks. The probe was the 600-bp fragment that encompasses exons 10–12. Digestions were as follows: lane 1–6, BamHI; lane 7, EcoRI; lane 8, HindIII; lane 9, TaqI. Marker sizes are indicated on the left.

DNA samples of six zebrafishes were then digested with BamHI, and the hybridizing bands (either 3.2 or 1.8 kb in length) were excised from the gel. The DNA was eluted from the bands and amplified with primers A14 and K5 (Table 1). The resulting PCR products of ~330 bp were then cloned and sequenced (sequences AB, HPS, and KOL-2 in Fig 4). Although several individual clones were sequenced from each PCR product, a maximum of two different TCP1 sequences per individual was obtained, suggesting again the presence of a single TCP1 locus in the zebrafish. Both the deleted and nondeleted 3' UT region forms of the TCP1 genes were obtained from the amplified genomic samples. The 3.2-kb band yielded only one sequence, which corresponded to the form without the deletion, whereas the 1.8-kb band yielded either the deleted or nondeleted forms of the gene (Fig 4).

View larger version (51K): In this window In a new window Download PPT slide   Figure 4. Nucleotide alignments of TCP1 exon 12-3' UT region sequences obtained from genomic zebrafish DNA. For comparison, the three cDNA sequences are also included. Identity with the simple-majority consensus is indicated by dashes; indels are indicated by asterisks. Allele designations are given in parentheses.

Testing of PAC clones:
Screening of a zebrafish PAC genomic library by using the TCP1 cDNA clone as a probe yielded two positive PAC clones (nos. 023263Q2 and H0274Q2). However, each of the two clones yielded only one type of sequence, that with the 10-bp deletion in the 3' UT region. Hybridization of the clones with a single exon probe revealed that each of the PAC clones contains a single copy of the gene (data not shown).

Linkage of the TCP1 locus:
Previously, we mapped the zebrafish class I and class II Mhc loci to different linkage groups in the Danio genome (BINGULAC-POPOVIC et al. 1997 ). The class I loci were mapped to linkage group XIX at a distance of 21.9 cM from the nontail gene, which is homologous to the mouse Brachyury (T) locus. Since in the mouse the T locus maps 14.5 cM centromeric to the class I H2-K locus and the TCP1 locus is located between T and H2-K, we investigated the possibility that this linkage relationship may have been conserved from fish to mammals. Haploid progenies derived from double-heterozygous females were tested for TCP1 and for class I or class II loci (see Table 1 for primers used). A total of 44 and 20 haploid embryos were tested for linkage of TCP1 with the class I or class II loci, respectively, but no indication of linkage was found (Table 2). Similarly, no linkage was found between TCP1 and the class I-linked PSMB9 locus (Table 2; TAKAMI et al. 1997 ).

Polymorphism of the TCP1 locus:
Primers specific for the Dare-TCP1 genes with (T-3-7) or without (T-3-5) the deletion were used to amplify genomic DNA from nine zebrafish (six caught in the wild and the other three from different unrelated laboratory stocks). The amplification using either the T-3-7-specific or T-3-5-specific primer combination produced a band of ~330 bp, which was tested by SSCP to distinguish homozygotes from heterozygotes and to differentiate the TCP1 alleles. Of the nine zebrafish DNA samples, two were amplified only by the T-3-5 combination, four only by the T-3-7 combination, two by both primer combinations, and one was not amplified at all. Four of the six samples amplified by one of the two pairs produced an SSCP pattern indicative of homozygosity (data not shown), while the remaining two samples appeared to be derived from TCP1 heterozygous fish. Alternatively, all the samples could have been heterozygous, but the null alleles were not amplified by the primer combinations used.

To investigate the extension of the polymorphism of the zebrafish TCP1 locus further, we sequenced the TCP1 alleles borne by the nine individuals using the primer combination A14-K5. Altogether the six wild-caught fishes (Ben, Cal, Sin in Table 3 and Fig 4), together with the nine laboratory fishes and the three cDNA clones, yielded 21 sequences, each encompassing 275 bp of parts of exon 12 and of the 3' UT region (Fig 4).

The 21 sequences represent 12 different alleles, designated by letters a–l (Table 3). Although the sequences TA2 (c) and T-3-5 (a) are identical in the 275-bp stretch, they differ in other parts of the cDNA sequence. Four of the alleles were represented by more than 1 sequence (i.e., alleles a, e, g, and j by 2, 6, 2, and 3 sequences, respectively). The alleles a, e, and j were present in both wild-caught and laboratory fishes. The high level of polymorphism of the TCP1 locus could be even higher than that detected if longer sequences were compared, as in the case of TA2 and T-3-5. In the 275-bp stretch, the total number of substitutions in pairwise comparisons ranges from one (0.36%), between alleles a and b, to nine (3.27%), between alleles b and f. In the coding part, the number ranges from one, between alleles a and b, to six (2.18%), between allele b and alleles f and h. However, all of these are synonymous substitutions suggesting, at least for this part, a strong purifying selection operating on the TCP1 gene.

The comparison of the cDNA sequences reveals that the TA2 and T-3-5 clones differ by 2 substitutions, both of which are in the coding part of the gene and are nonsynonymous at codon positions 297 (Met–Thr) and 437 (Gln–Arg). The TA2 clone differs from T-3-7 by 12 and 6 substitutions (and the 10-bp deletion) in the coding part and the 3' UT region, respectively, but only 1 of these is nonsynonymous at codon position 143 (Asp–Glu). Finally, clones T-3-5 and T-3-7 differ by 16 and 20 substitutions in the coding part and 3' UT region, respectively. Three of the substitutions are nonsynonymous at codon positions 143 (Asp–Glu), 297 (Met–Thr), and 437 (Gln–Arg). Of these 3 substitutions, 2 cause conservative amino acid replacement at positions 143 and 297.

Phylogenetic relationships among TCP1 sequences:
A maximum parsimony phylogenetic tree based on the sequenced 275-bp stretch shows two clusters supported by high bootstrap values (Fig 5). One cluster contains all sequences without the deletion in the 3' UT region; therefore, all of these sequences are derived presumably from the same ancestor. In this cluster, the alleles a (c), b, and d form a group separated from the allele e. The second well-defined cluster contains all alleles with the deletion in the 3' UT region, again all derived presumably from a common ancestor. Within this cluster, alleles f and g form a group, alleles i and j a second group, and both are separated from alleles h, k, and l. The age of the alleles estimated from the average genetic distance of 0.0156 using a rate of 3.5 x 10^-9 synonymous substitutions per synonymous site per year (LI 1997 ) is ~2.2 million years (my) with an upper limit of 5 my.

View larger version (22K): In this window In a new window Download PPT slide   Figure 5. Phylogenetic relationships among zebrafish TCP1 sequences from Fig 4. The dendrogram was constructed by the maximum parsimony method using the heuristic search option of the PAUP program (SWOFFORD 1993 ). Numbers indicate recovery of the branching orders in 100 bootstrap replications. The letters C, G, or T indicate a substitution; the subsequent number indicates site, 56, 71, 104, or 119.

Maximum parsimony trees of the entire 275-bp stretch (coding part and 3' UT region; Fig 5) and of the coding part alone (not shown) differ. In the former, all the sequences containing the 10-bp deletion in the 3' UT region group together, but the substitutions at sites 56 (C-T), 71 (C-T), 104 (C-T), and 119 (A-G) are scattered among all four clusters. In the latter, one branch of the tree contains all the sequences with T at site 56, a second branch clusters sequences with C at site 71, and a third branch contains sequences with T at site 104. The second branch is split into two subbranches, one with G and the other with A at site 119. In this tree, however, sequences with the 10-bp deletion are scattered among the different clusters. This discrepancy can be explained either by parallelism of the substitutions or by recombination.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The data described in this study indicate the presence of a single TCP1 locus in the haploid zebrafish genome. There are two highly divergent groups of alleles at this locus, one marked by the presence and the other by the absence of a 10-bp deletion in the 3' UT region. These conclusions are supported by several observations. First, in the progeny test, in the haploid embryos derived from a TCP1 heterozygous mother, the genes with and without the deletion segregate antithetically, as would be expected of alleles. Second, on a Southern blot, the TCP1 gene is found to be present in no more than two BamHI restriction fragments per individual. When two fragments are present, the 3.2-kb fragment carries the gene without the deletion only, whereas the 1.8-kb fragment bears either the gene with the deletion or the one without it. Third, no evidence for the presence of more than one TCP1 locus could be obtained by the screening of a PAC library and analysis of the positive clones. Fourth, although multiple alleles could be identified at the TCP1 locus, no individual was found with more than two alleles. Taken in isolation, none of these four observations provides conclusive evidence, but taken together they make a strong case in support of allelism and against pseudoallelism of the groups of TCP1 sequences.

The zebrafish TCP1 locus shows a surprisingly high degree of variability, in terms of both the number of alleles identified in the small sample of individuals tested and of the genetic distances between alleles. The collection of 15 individuals tested and the screening of the cDNA library yielded 12 different TCP1 sequences, even though only a 275-bp-long stretch of the gene was sequenced from most of the alleles. Some of the alleles differ by up to nine substitutions (and the 10-bp deletion with changes in its flanks) in the 275-bp-long stretch of sequence.

Since no frequency data are available for any of the alleles and since some of the sequences originated from laboratory stocks, strictly speaking, it is not possible to differentiate between ephemeral variation and polymorphism in any of the cases. However, those alleles that differ by more than one substitution have been present in the zebrafish population presumably for extended periods of time in order to accumulate multiple differences in a presumably stepwise fashion. These alleles apparently represent true polymorphism, especially in regard to the differences between the two groups of alleles distinguished by the presence or absence of the 10-bp deletion.

The existence of two groups of highly divergent alleles in the zebrafish is reminiscent of the situation in the house mouse, Mus musculus. In the latter species, two major Tcp1 alleles have been described by SILVER et al. 1979 and have been shown to be separated by a distance of 0.023 ± 0.010 at the synonymous sites, 0.014 ± 0.005 at the nonsynonymous sites, and 0.026 ± 0.006 in the intron region (MORITA et al. 1992 ). By calibrating the molecular clock by the mouse-rat divergence (17 mya), MORITA et al. 1992 estimated that the two alleles diverged 2.9 ± 0.7 mya. Apparently, the two alleles evolved separately from each other because one of them came to be included accidentally in an inversion that prevented recombination with wild-type chromosomes in this region and thus founded the so-called t-haplotypes (SILVER 1985 ). The presence in the t-haplotype of transmission ratio distorters maintained these haplotypes in the population. It is unlikely that a similar mechanism is responsible for the divergence of the two main zebrafish groups of alleles, for no indication of significant transmission ratio distortion was evident in crosses. Nevertheless, isolation of one kind or another is one explanation for the large distance separating the alleles. An alternative possibility is that balancing selection acts either at the TCP1 locus itself or at a linked locus. Since we did not find any nonsynonymous substitution consistently differentiating the two groups of TCP1 alleles, we consider positive selection targeting the TCP1 locus an unlikely explanation. As for selection at a linked locus, the Mhc class I genes would have been good candidates because in mammals, as shown by C. O'HUIGIN, A. HAUSMANN, R. DAWKINS and J. KLEIN (unpublished results) for humans, the regions flanking the functional class I loci exhibit higher polymorphism than other regions. The increase in polymorphism is attributed to a hitchhiking effect of the balancing selection operating on the class I loci. Since we have demonstrated here that the zebrafish Mhc class I and TCP1 loci are not linked, however, this explanation does not apply. Of course, there could exist another, as yet unidentified locus in the vicinity of TCP1, which evolves under balancing selection and hitchhikes the arising variants in its flanking regions, including the TCP1 locus. But in the absence of any evidence for the existence of such a hypothetical locus, the explanation based on the postulate of isolation is more attractive.

Of the various types of isolation that might come in question, the most reasonable seems to be geographical isolation. The natural habitats of the zebrafish encompass the fresh waters of much of the Indian subcontinent, in particular its eastern part, Nepal, Pakistan, Bangladesh, Myanmar, and Sri Lanka (BARMAN 1991 ). In such an enormous region, the opportunities for gene flow between the widely separated populations are limited and therefore some of the populations may evolve in relative isolation. In such a case, however, the divergence would be expected to be discernible at multiple loci. Indeed, there are indications that it is not restricted to the TCP1 locus. For example, ONO et al. 1993 described two sequences at the ß₂-microglobulin-encoding locus, one obtained from a cDNA and the other from a genomic clone. Although the ß₂-microglobulin locus, like the TCP1 locus, is highly conserved in its evolution, the two sequences differed by 12 (1.7%) nucleotide substitutions in the 3' UT region, and the authors estimated that they diverged from each other >0.7 mya. Zebrafish RAG1 (GREENHALGH and STEINER 1995 ; WILLETT et al. 1997 ) or rhodopsin (ROBINSON et al. 1995 ; VIHTELIC et al. 1999 ) alleles differ by 0.8% substitutions, indicating that they diverged 1.1 mya. In other cases, genes present in one zebrafish stock could not be found in another stock (e.g., the Mhc class II DCB, DBB, and DEB genes; see SULTMANN et al. 1994 ; BINGULAC-POPOVIC et al. 1997 ), which might be taken as an indication for the existence of different genomes in the zebrafish populations. Assuming that geographical isolation is the cause of the relatively large distances between some of the TCP1 alleles and taking the synonymous substitution rate of 3.5 x 10^-9 substitutions per synonymous site per year (LI 1997 ), we estimate that the two most different alleles diverged 2.2 mya (or 2.2 x 10⁶ generations, assuming that one generation corresponds to 1 year). This divergence time is comparable to that estimated by MORITA et al. 1992 for TCP1 genes of different mouse species. If the fish substitution rate is lower than the mammalian rate used in these computations, the divergence times of the zebrafish alleles might be even longer. Therefore, individuals from geographically distant populations might be regarded molecularly as representing different species.

The possibility that Danio rerio may comprise a heterogeneous collection of highly divergent genomes has two important practical implications. First, it underscores the need to work with defined stocks if incongruence of results is to be avoided; and second, it reveals a source of naturally occurring variation for segregation analyses and other genetic studies. With the increasing use of the zebrafish in developmental and genetic research, it is desirable to keep these possibilities in mind.

The observed absence of linkage between the TCP1 and class I Mhc loci has implications for the interpretation of the ancient synteny described by TRACHTULEC et al. 1997 . These authors pointed out that the loci PIM1, RXRB, TNX, PBX, PSMB, TBP, and NOTCH are part of a linkage group conserved among C. elegans, Drosophila melanogaster, and humans (mouse). Additional genes, including the Brachyury and TCP1 homologs, have since been shown to be part of the group (KLEIN and SATO 1998 ; H. SÜLTMANN and J. KLEIN, unpublished results). The genome of C. elegans contains at least five TCP1-related genes, three of which are on chromosome II and two on chromosome III (LEROUX and CANDIDO 1995A , LEROUX and CANDIDO 1995B ). It is the latter chromosome that bears the conserved syntenic group of TRACHTULEC et al. 1997 . The actual ortholog of the vertebrate TCP1 gene has been reported to be one of the three loci on chromosome II (LEROUX and CANDIDO 1995B ), but since the five genes are closely related, orthology relationships are difficult to establish with certainty. Whatever the case may be, the TCP1 synteny with the rest of the group has been disrupted in the zebrafish and the same is also true for the PIM1 gene (B. MURRAY, H. SÜLTMANN and J. KLEIN, unpublished data). This disruption may have been associated with the reorganization that placed the zebrafish class I and class II Mhc genes on different chromosomes (BINGULAC-POPOVIC et al. 1997 ). In mammals, the TCP1 gene overlaps in its 3' UT region with the ACAT gene, but as shown by SHINTANI et al. 1999 , this overlap is of relatively recent origin. Apparently, it arose during the transition from reptiles to mammals.

	ACKNOWLEDGMENTS

We thank Mr. Ryszard Lorenz for technical assistance and Ms. Jane Kraushaar for editorial assistance, as well as Dr. Holger Sültmann for critical reading of the manuscript.

Manuscript received May 31, 1999; Accepted for publication September 2, 1999.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AHMAN, S. and R. S. GUPTA, 1990  Cloning of a Chinese hamster protein homologous to the mouse t-complex protein TCP-1: structural similarity to the ubiquitous `chaperonin' family of heat-shock proteins. Biochim. Biophys. Acta 1087:253-255[Medline].

AHNERT, V., C. MAY, R. GERKE, and H. KINDL, 1996  Cucumber T-complex protein. Molecular cloning, bacterial expression and characterization within a 22-S cytosolic complex in cotyledons and hypocotyls. Eur. J. Biochem. 235:114-119[Medline].

BARMAN, R. P., 1991  A taxonomic revision of the Indo-Burmese species of Danio Hamilton Buchanan (Pisces: Cyprinidae). Rec. Zool. Surv. India 137:1-91.

BINGULAC-POPOVIC, J., F. FIGUEROA, A. SATO, W. S. TALBOT, and S. L. JOHNSON et al., 1997  Mapping of Mhc class I and class II regions to different linkage groups in the zebrafish, Danio rerio.. Immunogenetics 46:129-134[Medline].

CHENG, M. Y., F. U. HARTL, J. MARTIN, R. A. POLLOCK, and F. KALOUSEK et al., 1989  Mitochondrial heat-shock protein hsp60 is essential for assembly of proteins imported into yeast mitochondria. Nature 337:620-625[Medline].

ELLIS, R. J. and S. M. VAN DER VIES, 1991  Molecular chaperones. Annu. Rev. Biochem. 60:321-347[Medline].

FRYDMAN, J. and F. U. HARTL, 1996  Principles of chaperone-assisted protein folding: differences between in vitro and in vivo mechanisms. Science 272:1497-1502[Abstract].

GEORGOPOULUS, C. P., R. W. HENDRIX, S. R. CASJENS, and A. D. KAISER, 1973  Host participation in bacteriophage lambda head assembly. J. Mol. Biol. 76:45-60[Medline].

GONGORA, R., Z. ZALESKA-RUTCZYNSKA, K. TAKAMI, F. FIGUEROA, and J. KLEIN, 1998  Linkage of RXRB-like genes to class I and not class II Mhc genes in the zebrafish. Immunogenetics 48:141-143[Medline].

GREENHALGH, P. and L. A. STEINER, 1995  Recombination activating gene 1 (Rag1) in zebrafish and shark. Immunogenetics 41:54-55[Medline].

HEMMINGSEN, S. M., C. WOOLFORD, S. M. VAN DER VIES, K. TILLY, and D. T. DENNIS et al., 1988  Homologous plant and bacterial proteins chaperone oligomeric protein assembly. Nature 333:330-334[Medline].

KIM, S., K. R. WILLISON, and A. L. HORWICH, 1994  Cystosolic chaperonin subunits have a conserved ATPase domain but diverged polypeptide-binding domains. Trends Biochem. Sci. 19:543-548[Medline].

KIMURA, M., 1980  A simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. J. Mol. Evol. 16:111-120[Medline].

KIRCHHOFF, C. and K. R. WILLISON, 1990  Nucleotide and amino acid sequence of human testis-derived TCP1. Nucleic Acids Res. 18:4247[Free Full Text].

KLEIN, J. and A. SATO, 1998  Birth of the major histocompatibility complex. Scand. J. Immunol. 47:199-209[Medline].

KUBOTA, H., G. HYNES, A. CARNE, A. ASHWORTH, and K. WILLISON, 1994  Identification of six Tcp-1 related genes encoding divergent subunits of the Tcp-1 containing chaperonin. Curr. Biol. 4:89-99[Medline].

KUBOTA, H., G. HYNES, and K. WILLISON, 1995  The chaperonin containing t-complex polypeptide 1 (TCP-1). Eur. J. Biochem. 230:3-16[Medline].

KUMAR, S., K. TAMURA and M. NEI, 1993 MEGA: Molecular Evolutionary Genetic Analysis Version 1.0. The Pennsylvania State University, University Park, PA.

LEROUX, M. R. and E. P. CANDIDO, 1995a  Characterization of four new tcp-1-related cct genes from the nematode Caenorhabditis elegans.. DNA Cell Biol. 14:951-960[Medline].

LEROUX, M. R. and E. P. CANDIDO, 1995b  Molecular analysis of Caenorhabditis elegans tcp-1, a gene encoding a chaperonin protein. Gene 156:241-246[Medline].

LI, W.-H., 1997 Molecular Evolution. Sinauer Associates, Sunderland, MA.

MORI, M., K. MURATA, H. KUBOTA, A. YAMAMOTO, and A. MATSUSHIRO et al., 1992  Cloning of a cDNA encoding Tcp-1 (t-complex polypeptide 1) homologue of Arabidosis thaliana.. Gene 122:381-382[Medline].

MORITA, T., H. KUBOTA, G. GACHELIN, M. NOZAKI, and A. MATSUSHIRO, 1991  Cloning of cDNA encoding rat TCP-1. Biochim. Biophys. Acta 1129:96-99[Medline].

MORITA, T., H. KUBOTA, K. MURATA, M. NOZAKI, and C. DELARBRE et al., 1992  Evolution of the mouse t haplotypes: recent and worldwide introgression to Mus musculus.. Proc. Natl. Acad. Sci. USA 89:6851-6855[Abstract/Free Full Text].

ONO, H., F. FIGUEROA, C. O'HUIGIN, and J. KLEIN, 1993  Cloning of the ß₂-microblobulin gene in the zebrafish. Immunogenetics 38:1-10[Medline].

ORITA, M., H. IWAHANA, H. KANAZAWA, K. HAYASHI, and T. SEKIYA, 1989  Detection of polymorphisms of human DNA by gel electrophoresis as single-strand conformation polymorphisms. Proc. Natl. Acad. Sci. USA 86:2766-2770[Abstract/Free Full Text].

ROBINSON, J., E. A. SMITH, and J. E. DOWLING, 1995  Temporal and spatial patterns of opsin gene expression in zebrafish (Danio rerio). Vis. Neurosci. 12:895-906[Medline].

SAITOU, N. and M. NEI, 1987  The neighbor-joining method: a new method for reconstruction phylogenetic trees. Mol. Biol. Evol. 4:406-425[Abstract].

SANGER, F., S. NICKLEN, and A. R. COULSON, 1977  DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463-5467[Abstract/Free Full Text].

SATO, A., F. FIGUEROA, C. O'HUIGIN, D. N. REZNICK, and J. KLEIN, 1995  Identification of major histocompatibility complex genes in the guppy, Poecilia reticulata.. Immunogenetics 43:38-49.

SHINTANI, S., C. O'HUIGIN, S. TOYOSAWA, V. MICHALOVÁ, and J. KLEIN, 1999  Origin of gene overlap: the case of TCP1 and ACAT2.. Genetics 152:743-754[Abstract/Free Full Text].

SILVER, L. M., 1985  Mouse t haplotypes. Annu. Rev. Genet. 19:179-208[Medline].

SILVER, L. M., K. ARTZT, and D. BENNETT, 1979  A major testicular cell protein specified by a mouse T/t complex gene. Cell 17:275-284[Medline].

SÜLTMANN, 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 23:1-14[Medline].

SWOFFORD, D. D., 1993 PAUP: Phylogenetic Analysis Using Parsimony Version 3.1.1. Illinois Natural History Survey, Champaign, IL.

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. 159:6052-6060[Abstract].

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 42:77-84[Medline].

TRACHTULEC, Z., R. M. HAMVAS, J. FOREJT, H. R. LEHRACH, and V. VINCEK et al., 1997  Linkage of TATA-binding protein and proteasome subunit C5 genes in mice and humans reveals synteny conserved between mammals and invertebrates. Genomics 44:1-7[Medline].

URSIC, D. and M. R. CULBERTSON, 1991  The yeast homolog to mouse Tcp-1 affects microtubule-mediated processes. Mol. Cell. Biol. 11:2629-2640[Abstract/Free Full Text].

URSIC, D. and B. A. GANETZKY, 1988  A Drosophila melanogaster gene encodes a protein homologous to the mouse t complex polypeptide 1. Gene 68:267-274[Medline].

VIHTELIC, T. S., C. J. DORO, and D. R. HYDE, 1999  Cloning and characterization of six zebrafish photoreceptor opsin cDNAs and immunolocation of their corresponding proteins. Vis. Neurosci. 16:571-585[Medline].

WILLETT, C. E., J. J. CHERRY, and L. A. STEINER, 1997  Characterization and expression of the recombination activating genes (rag1 and rag2) of zebrafish and shark. Immunogenetics 45:394-404[Medline].

WILLISON, K. R., and H. KUBOTA, 1994 The structure, function and genetics of the chaperonin containing TCP-1 (CCT) in eukaryotic cytosol, pp. 299–321 in The Biology of Heat Shock Proteins and Molecular Chaperones, edited by R. I. MORIMOTO, A. TISSIERES and C. GEORGOPOULUS. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

WILLISON, K. R., K. DUDLEY, and J. POTTER, 1986  Molecular cloning and sequence analysis of a haploid expressed gene encoding t complex polypepetide 1. Cell 44:727-738[Medline].

This article has been cited by other articles:

				D. E. Neafsey, J. P. Blumenstiel, and D. L. Hartl Different Regulatory Mechanisms Underlie Similar Transposable Element Profiles in Pufferfish and Fruitflies Mol. Biol. Evol., December 1, 2004; 21(12): 2310 - 2318. [Abstract] [Full Text] [PDF]

				H. Sultmann, A. Sato, B. W. Murray, N. Takezaki, R. Geisler, G.-J. Rauch, and J. Klein Conservation of Mhc Class III Region Synteny Between Zebrafish and Human as Determined by Radiation Hybrid Mapping J. Immunol., December 15, 2000; 165(12): 6984 - 6993. [Abstract] [Full Text] [PDF]