Chapter Zebrafish vtg family Chapter Characterization of the zebrafish vtg family: sequencing, mapping and phylogenetic analysis 27 Chapter Zebrafish vtg family Abstract Vitellogenins (Vtgs) are precursors of yolk proteins in oviparous species and are cleaved into three portions upon uptake by oocytes, lipovitellin I (LVI), phosvitin (PV) and liopovitellin II (LVII) In the present study, we found that the zebrafish genome contains at least seven vtg genes (vtg1-7) encoding heterogeneous vitellogenins with three distinct groups: group A Vtgs (Vtg1, 4-7) contain all three major portions but lack the C-terminal half of LVII; group B Vtg (Vtg2) is the only one including intact three portions; group C Vtg (Vtg3) lacks both PV and the LVII C-terminal half The seven vtgs were located in two different chromosomes: one (vtg3) in LG11 and the rest closely linked in LG22 Phylogenetic analysis indicated that the expansion of group A vtgs in zebrafish is lineage specific, whereas gene duplication forming precursors of three groups of vtg may occur before the radiation of teleost fish and may through whole chromosome or genome duplication 28 Chapter Zebrafish vtg family 2.1 Introduction In oviparous vertebrates, yolk is critical for embryonic development as it is a rich source of nutrients, including amino acids, phosphate, carbohydrates, lipids and vitamins In addition, maternally derived hormones such as thyroxine and T3 have also been detected in embryonic yolk (Kobuke et al., 1987) Early studies on the composition of yolk revealed two classes of egg yolk proteins: the phosphoserine rich glycosylated phosvitin (PV) and lipid-binding lipovitellin (LV) (Wallace and Jared, 1968; 1969) Both types of proteins are derived from a common precursor protein termed as vitellogenin (Vtg), which was first coined by Pan et al (1969) Vtgs are calcium and zinc-binding phospholipoglycoproteins synthesized in hepatic parenchymal cells under the influence of female sex steroid hormone, estrogen (E2) (Wallace and Jared, 1969; Montorzi et al., 1995) Vtgs are ancient proteins belonging to a multiple member family that includes a variety of lipoproteins (Bryne et al., 1989) Mammalian apolipoprotein B, the large subunit of mammalian microsomal triglyceride transfer protein and insect apolipophorin II/I precursor are suspected to share a common ancestor with Vtgs (Baker, 1988a&b; Babin et al., 1999) Vtgs are encoded by multigene families in essentially all oviparous species examined For example, Wahli et al (1979) first reported that there were four vtgs in Xenopus laevis In the nematode Caenorhabditis elegans, six vtgs have been identified, which are located on different chromosomes (Spieth et al., 1991 and references within) The number of vtgs reported in different fish varies with species It is believed that a teleost genome contains 2-20 copies of vtgs depending on species The presence of several 29 Chapter Zebrafish vtg family distinct vtg EST clones in the zebrafish (Danio rerio) also indicated that multiple copies of vtgs exist in this fish species (Gong et al., 1997) The genes or cDNAs encoding Vtgs have been described in many vertebrate species including chicken, Xenopus, and several fish species (Chen et al., 1997 and references within) Vtgs are among the largest proteins and contain up to 2,139 amino acids with a predicted molecular weight of 250 kDa before post-translational modification (Chen et al., 1994a) After a Vtg precursor is internalized into the ovary by receptor-mediated endocytosis (RME), it is cleaved into LVI, PV and LVII (Wahli, 1988) PV is a serine rich domain containing one or more stretches of serine residues In Xenopus, the PV can be further cleaved into two smaller phosphoproteins, phosvettes I and II (Wahli, 1988) In genomic organization, despite similar length of Vtg proteins, C elegans vtgs have 4-5 exons (Spieth et al., 1991) whereas Xenopus and chicken vtgs have 35 exons (Nardelli, et al., 1987) The size of the corresponding exons between Xenopus and chicken vtgs was highly conserved except for exon 23, which generally codes for the PV moiety in vertebrates Preliminary studies on Vtgs in invertebrates suggested that these organisms not have the PV domain (Spieth et al., 1985; Trewitt et al., 1992) However, a contradictory report from Chen et al (1994a) stated that two polyserine rich regions were discovered in mosquito (Aedes aegypti) Vtg Whether Vtgs without phosvitin also exist in vertebrates is unknown Therefore, it is imperative to investigate vtgs from a wide diversity of organisms for a better understanding of the evolutionary relationships among vtgs 30 Chapter Zebrafish vtg family The purpose of the present study was to characterize the zebrafish (Danio rerio) vtg multigene family, including identification of individual vtg members, elucidation of the primary structures of Vtg proteins, mapping of vtg genes and inferring the evolutionary relationships among Vtgs of various fish species Based on this characterization, an appropriate Vtg candidate will be selected for preparation of a DNA carrier, which will be used in the development of a novel gene transfer method, receptor-mediated gene transfer (see Chapter 4) 31 Chapter Zebrafish vtg family 2.2 Materials and Methods 2.2.1 vtg cDNA clones and DNA sequencing All vtg cDNA clones were obtained from our previous EST clones isolated from an adult cDNA library (Gong et al., 1997) Longest clones representing vtg1 to vtg7 were sequenced completely from both ends by automatic sequencing The deduced amino acid sequences of Vtg1-7 were determined by DNAMAN V 4.15 (Lynnon BioSoft) and putative signal peptide cleavage sites were determined by a computer program SignalP V1.1, accessible at http://www.cbs.dtu.dk/services/SignalP/ The deduced amino acid sequences of Vtg1-7 were used in Fasta search (http://www.ebi.ac.uk/fasta33/) to determine the most homologous Vtgs in other species for phylogenetic analysis For manual sequencing, SK primer (5’-CGCTCTAGAACTAGGATC-3’) was used for direct sequencing of 5’ ends of cDNA inserts using the T7 Sequencing Kit (Pharmacia) and rapid denaturation and annealing were performed according to the manufacturer's instructions For automatic sequencing, primers of SK, T7 (5’- GTAATACGACTCACTATAGGGC-3’) and various gene specific primers (not shown) were used for complete sequencing of whole inserts of representative vtg cDNA clones ABI PRISM BigDye Terminator Cycle Sequencing Ready Reaction Kit (Perkin Elmer Applied Biosystems) was used to prepare sequencing reactions and the cycle sequencing reactions were performed on a GeneAmp PCR system 9600 (Perkin Elmer) Briefly, each 20 µl of reaction was composed of µl of Terminator Ready Reaction Mix, 3.2 pmol of primer and 200-500 ng of plasmid DNA The reaction was performed for 25 cycles with the following parameters for each cycle: denaturing at 96 ºC for 10 sec, annealing at 50 ºC for sec and extension at 60 ºC for The sequencing reaction products were then 32 Chapter Zebrafish vtg family loaded onto polyacrylamide gel, followed by gel electrophoresis and data process using an automatic sequencer (ABI Prism 377, Perkin Elmer) 2.2.2 Sequence alignment and phylogenetic analysis Putative Vtg amino acid sequences were aligned by a multiple sequence alignment program Clustal W (http://www.ebi.ac.uk/clustalw/) with default parameters Well aligned regions were chosen from each sequence and used in phylogenetic analysis by the parsimony method using the phylogenetic program PAUP v3.1 (Swofford, 1993) The input file was in NEXUS format following the PROTPARS example included in the PAUP program One-hundred bootstrap replicates were attempted using the heuristic type of search The mosquito (Aedes aegypti) Vtg (GenBank accession No AAA18221) or C elegans Vtg1 (GenBank accession No U37430) was used as an outgroup in construction of the phylogenetic tree Sequence alignment was also performed by DNAMAN V 4.15 (Lynnon Biosoft) 2.2.3 5' RACE PCR and partial genomic region amplification of vtg3 Because of the relatively high sequence divergence of vtg3, full length vtg3 cDNA was isolated for further characterization Briefly, a Marathon cDNA Amplification Kit (Clontech) was used to construct an adaptor-ligated double stranded cDNA library from total liver RNA of female zebrafish based on the manufacturer’s instructions One genespecific primer (primer 1: 5'-GGTAACTCAAGTGGCCAAGT-3', Figs 2-3 and 2-12) was designed for the 5' RACE-PCR using a diluted adaptor-ligated liver cDNA library as templates The 5’ missing cDNA sequence of vtg3 was obtained by performing PCR using primer and an adaptor primer AP1 supplied by the manufacturer In 50 µl of PCR 33 Chapter Zebrafish vtg family reaction, there was 29.8 µl of dH2O, µl of 10 x PCR buffer, µl of 10 x dNTPs (2 mM each), µl of 25 mM MgCl2, µl each of primer and AP1 (10 µM each), µl of 1/5 diluted adaptor-ligated liver cDNA library and 0.2 µl of Taq DNA polymerase (5 U/µl) PCR was performed on a DNA Thermal Cycler 480 (Perkin Elmer) with the following parameters: 94 oC for min; 35 cycles of 94 oC for 30 sec, 60 oC for and 72 oC for min; finally 72 oC for A PCR fragment of about 2-kb was recovered from agarose gel and ligated into pT7Blue vector (Novagen) Subsequent transformation was carried out using competent cells DH5α and resulting colonies were screened by PCR The plasmid harboring the retrieved 5’ sequence of vtg3 was sequenced from the multiple cloning site by automatic sequencing To determine whether the vtg3 genomic sequence comprises a PV region, zebrafish genomic DNA was extracted from a single fish according to a modified protocol reported by Xu et al (1999), which was based on a standard protocol by Sambrook et al (1989) Two primers (primer 2: 5' -TGCACACTATCTTCACGAA-3' and primer 3: 5' GCTGATGTATGAGTCCTAT-3', Figs 2-3 and 2-12) flanking the putative missing PV region were designed and used in PCR amplification of partial genomic sequence of vtg3 Briefly, 250 ng of zebrafish genomic DNA was used as templates in 50 µl of PCR reaction, which contained the same reagents as described for the 5’ RACE PCR except for the template PCR was also performed on the DNA Thermal Cycler 480 with the following parameters: 94 oC for min; 30 cycles of 94 oC for 30 sec, 54 oC for and 72 oC for min; finally 72 oC for A fragment of ~ 3.0-kb was amplified from the genomic DNA and cloned into pT7Blue vector (Novagen) The 3.0 kb insert was then sequenced completely by automatic sequencing 34 Chapter Zebrafish vtg family 2.2.4 Genome mapping of zebrafish vtgs To map the seven vtgs, mapping PCR was carried out using the T51 radiation hybrid panel, which consists of 94 radiation hybrids of zebrafish-hamster hybrid cell lines (Research Genetics) (Kwok et al., 1998; Geisler et al., 1999) For vtg1, vtg4, vtg5 and vtg7, their forward and reverse mapping primers were designed based on nucleotide sequences in the 3’end coding regions and 3’ UTRs, respectively (Figs 2-1, 2-4, 2-5 and 2-7) For vtg2, mapping primers were designed based on the sequence located in the unique 3’ end coding region of cDNA clone A183 (Fig 2-2) vtg3 mapping primers were targeted to an intron of the vtg3 gene (Fig 2-14B) Mapping PCR was performed using Taq PCR Core Kit (Qiagen) µl of 25 ng DNA (of radiation hybrids or controls) was used in each 20 µl of reaction which contained 6.5 µl of dH2O, µl of 10 x PCR buffer, µl of x Q-solution, 0.4 µl of dNTPs (10 mM each), µl each of forward and reverse primers (20 µM each) and 0.1 µl of Taq DNA polymerase (5 U/µl) The PCR machine was programmed with the following parameters: one cycle of initial denaturation at 95 °C for min, 45 cycles of amplification at 94 °C for 30 sec, 55 °C for 15 sec and 72 °C for and a final extension at 72 °C for PCR products were resolved on a 1.5% agarose gel and images were recorded under UV illumination Most of the reactions were performed in duplicate and some in triplicate Consistent results from at least two separate reactions were scored according to SAMapper codes ("1" – specific band amplified, "0" – no band amplified, "R" – undecided) and submitted for analysis via e-mail to http://wwwmap.tuebingen.mpg.de:8082/rh/ 35 Chapter Zebrafish vtg family 2.2.5 Zebrafish databases used Data on mapped vtg EST clones was retrieved from the zebrafish EST database at http://zfish.wustl.edu/ (Washington University Zebrafish Genome Resources Project) Searching of the UniGene database (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=unigene) and the Ensembl zebrafish genome database v19.3.2 (http://www.ensembl.org/Danio_rerio/) was carried out for determining vtg clusters and potential novel vtgs in the zebrafish genome As a comparison, fugu(Fugu rubripes) genome database v2.0 (http://www.ensembl.org/Fugu_rubripes/) was also searched to determining homologous vtgs in this new model fish 36 Chapter Zebrafish vtg family Table 2-3 The number of mismatching nucleotides in specific and non-specific mapping primer annealing regions vtg1 vtg2 vtg3 vtg4 vtg5 vtg6 vtg7 vtg1MF1† (25 mer) 10 1 vtg1MR1 (26 mer) 19 –* 11 10 12 15 vtg4MF1 (25 mer) 16 vtg4MR2 (28 mer) 20 23 – 22 21 25 vtg5MF1 (25 mer) 11 1 vtg5MR1 (25 mer) 10 17 – 13 12 15 vtg7MF2 (26 mer) 16 1 vtg7MR1 (28 mer) 15 24 – 16 19 19 † For mapping primer sequences, see Figs 2-1, 2-4, 2-5 and 2-7 * No homologous region was found by sequence alignment 70 Chapter Zebrafish vtg family B A M Hybrids 1-24 M Hybrids 49-72 vtg1 M M Hybrids 25-48 Hybrids 73-94 M M Hybrids 1-24 M P NM M Hybrids 49-72 M 00110010000001001111100010010100000000010010010100000 0000000010000010000000001100101000000000110 C Hybrids 1-24 M M Hybrids 49-72 M Hybrids 25-48 PN vtg2 001100100000010011111000100101000000000101100101000000 000100010000010000000001000001000000000110 Hybrids 73-94 M PN 00010000100000100000010110010000000000000000100001000 0010000000000010100000001100000000000010010 E Hybrids 1-24 M M Hybrids 49-72 M vtg4 Hybrids 25-48 Hybrids 73-94 PN 00110000000001001111100010010100000000010010000000000 0000000010000000000000001100001000000000110 F M Hybrids 1-24 M M Hybrids 49-72 M Hybrids 25-48 Hybrids 73-94 M Hybrids 1-24 M M Hybrids 49-72 M Hybrids 25-48 Hybrids 73-94 PN PN vtg7 vtg5 Hybrids 73-94 D M vtg3 Hybrids 25-48 00110010000001001111100010010100000000010110010100000 0000100010000010000000001100001000000000110 00110010000001001111100010010100000000010110010100000 0000100010000010000000001100101000000000110 Fig 2-16 Genome mapping of zebrafish vtgs by PCR based on the T51 radiation hybrid (RH) panel Gel electrophoresis of mapping PCR products indicated that fragments of ~ 400-bp (A), ~ 800-bp (B), ~ 590-bp (C), ~ 200-bp (D), ~ 200-bp (E) and ~ 1000-bp (F) were amplified from certain radiation hybrids in the panel when mapping primers of vtg1, vtg2, vtg3, vtg4, vtg5 and vtg7 were used, respectively SAMapper codes representing the results of mapping PCR are listed below Serial number of the 94 radiation hybrids is marked briefly above the wells P, positive control (zebrafish genomic DNA); N, negative control (hamster genomic DNA); M, DNA ladder 71 Chapter Zebrafish vtg family A Markers cM LG 11 cR vtg3 B Markers cM LG 22 cR vtg4 vtg1 vtg5 vtg7 vtg2 Fig 2-17 Chromosome locations of the six zebrafish vtgs A: Partial linkage map of the linkage group (LG) 11, showing the map location of vtg3 B: Partial linkage map of LG 22, showing the map locations of vtg1, vtg2, vtg4, vtg5 and vtg7 Mapped vtg ESTs are underlined (map releasing date: September 8, 2003) 72 Chapter Zebrafish vtg family and vtg7, the lengths of the PCR products exceeded the estimated lengths from cDNA sequences, indicating that one or more genomic introns were amplified; for vtg3-5, the lengths of amplified genomic fragments were close to those estimated based on cDNA sequences (Fig 2-16) The SAMapper codes representing the results of mapping PCR are also shown in Fig 2-16 Unfortunately, mapping of the zebrafish vtg6 was not successful, since the mapping PCR products were very faint and not distinguishable After linkage analysis between the zebrafish vtgs and other markers, the zebrafish vtg3 was mapped to linkage group 11 (LG 11) and the remaining five vtgs, including vtg1, vtg2, vtg4, vtg5 and vtg7, were all mapped to LG 22 (Fig 2-17A,B) Strikingly, all the five zebrafish vtgs on LG22 are distributed in a close region within 97 centiRays (cR) 2.4 Discussion 2.4.1 Heterogenicity of fish vtg genes In many species, it has been demonstrated that vtgs belong to a multigene family For example, four vtgs have been isolated from Xenopus (Germond et al., 1984), three from chicken (van het Schip et al., 1987; Silva et al., 1989 and GenBank access number D83747) and two from killifish (Fundulus heteroclitus) (Lafleur et al., 1995a&b) In rainbow trout (Oncorhynchus mykiss), up to twenty complete vtgs were identified in the genome (Trichet et al., 2000) Searching of the UniGene database showed that in the zebrafish, there are two vtg clusters, vtg1 and vtg3, with 2585 sequences (including mRNAs and 2582 ESTs) and 97 sequences (including mRNA and 96 ESTs), respectively However, no further information is available about vtg subgroups in each cluster in the database The present study indicated that zebrafish have at least seven 73 Chapter Zebrafish vtg family distinct vtgs (vtg1-7), which can be further divided into three groups, group A (vtg1 and vtg4-7), group B (vtg2) and group C (vtg3) Whereas vtg3 encodes a protein without the PV domain, the other vtgs encode proteins with the PV domain except for Vtg7, whose PV sequence is not available Furthermore, Vtg2 has homologous subdomains IV and Vand likely I-III as well while other Vtgs may only have homologous subdomains I-III Thus, the zebrafish vtg family not only contains multiple vtg members, but also has three different types of vtg genes which encode three distinct types of Vtg proteins Although the PV-encoding vtgs and vtg3 coexist in the zebrafish, the expression levels of PV-encoding vtgs appear to be much higher than those of vtg3 For example, the expression levels of the zebrafish vtg1 and vtg2 mRNAs are 1000 and 10 times, respectively, higher than those of vtg3 in the liver of control female fish (see Chapter 3, Section 3.3.2.1.2 for detailed description) The difference in vtg mRNA concentrations is also reflected by the frequency of respective EST clones in the adult zebrafish cDNA library where among 42 vtg EST clones, 24 clones were identified as vtg1, clones as vtg2 and only one clone as vtg3 Based on the presence of conserved subdomains I-V (Chen et al., 1997) and the PV domain, Vtgs from other fish species also fall into three groups, A, B and C (Table 2-4) By definition, group A Vtgs have homologous subdomains I-III and the PV domain but without subdomains IV and V; group B Vtgs have all the five subdomains (I-V) and the PV domain; group C Vtgs have subdomains I-III but lack the PV domain (Table 2-4) This classification is basically in agreement with the phylogenetic analysis of fish Vtgs (Fig 2- 74 Chapter Zebrafish vtg family Table 2-4 Classification of fish Vtgs based on the homologous subdomains I-V and PV domain Species Gene name Abbreviation zebrafish (Danio rerio) vtg1 vtg2 vtg3 vitellogenin vitellogenin vitellogenin precursor common carp (Cyprinus carpio) fathead minnow (Pimephales promelas) rainbow trout (Oncorhynchus mykiss) Japanese sillago (Sillago japonica) medaka (Oryzias latipes) haddock (Melanogrammus aeglefinus) killifish (Fundulus heteroclitus) tilapia (Oreochromis aureus) Japanese common goby (Acanthogobius flavimanus) white sturgeon (Acipenser transmontanus) silver lamprey (Ichthyomyzon unicuspus) Fugu rubripes zfVtg1 zfVtg2 zfVtg3 ccVtg ccVtg2 fmVtg Homologous subdomains presented† I-III I-V†† I-III I-III III, IV (partial) I-III Phosvitin domain presented Yes Yes No Yes Yes Yes A B C A B A vitellogenin rtVtg1 I-V Yes B vitellogenin jsVtg I-V Yes B Ol-vit1 (vitellogenin 1) vitellogenin II vitellogenin A vitellogenin B mVtgI I-V Yes B mVtgII hVtgA hVtgB I-V I-V I-V Yes Yes Yes B B B vitellogenin I vitellogenin II vitellogenin kVtgI kVtgII tVtg1 I-V I-V I-V Yes Yes Yes B B B Vg-530 (Vitellogenin530 ) Vg-320 (Vitellogenin320) vitellogenin gVtg-530 I-V Yes B gVtg-320 I-III No C wsVtg I-IV Yes B‡ vitellogenin slVtg I-V Yes B VIT_ONCMY Q9DFT9 frVtgB frVtgC I-V I (partial), II, III Yes No B C Vtg group †Homologous subdomains I-V in Vtgs were defined by Chen et al (1997) ††Near full-length Vtg2 was encoded by a vtg2-like transcript ENSDART00000025283 in the zebrafish genome database ‡White sturgeon Vtg may represent an intermediate Vtg, since it only contains subdomains I-IV 75 Chapter Zebrafish vtg family 10B), as Vtgs of the same group are generally clustered together with the exception of Vtgs from white sturgeon and silver lamprey From the overall structure of Vtg protein, both Vtgs of white sturgeon and silver lamprey belong to group B, though each of which shares relatively high sequence similarity with that of Vtg3 This discrepancy may be due to the fact that only one region in LVII was chosen for phylogenetic analysis and thatregion may not be representative Interestingly, the group B Vtg in white sturgeon contains only subdomains I-IV and may represent an intermediate Vtg between groups A and B The current situation is that not all three groups of Vtgs have been identified in other fish species For example, in common carp (Cyprinus carpio), one Vtg each from groups A and B has been identified In Japanese common goby (Acanthogobius flavimanus) and Fugu rubripes, one Vtg each from groups B and C has been identified On the other hand, two Vtgs of group B have been found in the following species, medaka (Oryzias latipes), haddock (Melanogrammus aeglefinus) and killifish (Fundulus heteroclitus) (Table 2-4) In the rainbow trout (Oncorhynchus mykiss), twenty complete vtg genes have been identified (Trichet et al., 2000) The 20 vtgs in rainbow trout may all belong to group B based on the fact that the rainbow trout vtg1 belongs to group B and there are 97.4-99.9% sequence identities between exons and of the rainbow trout vtg1 and other 19 vtg sequences The assembly of the genome sequence of the new model species Fugu rubripes is also nearly finished Recent blast-search of the Ensembl Fugu rubripes genome database revealed that there are three vtgs in the fugu genome, SINFRUG00000140063 (tentatively named as frvtgB), SINFRUG00000124251 (frvtgC) and SINFRUG00000133882 (frvtgX) 76 Chapter Zebrafish vtg family The cDNA sequences of the above vtgs are not complete yet especially for SINFRUG00000133882, of which only 207-bp cDNA sequence is available The deduced protein of frVtgB is most similar to the zebrafish Vtg2 (with amino acid sequence identity of 47% in 596 amino acid overlap) and it also has homologous subdomains I-V The deduced protein of frVtgC is most similar to Vtg3, sharing 46% identity in 1146 amino acid overlap Furthermore, no PV domain could be found in the predicated protein of frVtgC The relationship between frVtgX and zebrafish Vtgs cannot be determined, since the overlapping sequence is too short Thus, the vtg multigene family in fugu rubripes is also heterogeneous, since homologous genes of the zebrafish vtg2 and vtg3 are present in this fish species Based on the current genome information of fugu, it seems that the number of vtgs in fugu is less than that in the zebrafish 2.4.2 Non-conserved domains of Vtgs: the PV domain and the homologous subdomains IV and V The variable parts of Vtg proteins are the PV domain and the homologous subdomains IV and V The exact role of the PV domain is unknown, but it has been proposed to be involved in receptor binding during receptor-mediated endocytosis of Vtg by the oocytes (Miller et al., 1982; Woods and Roth, 1984; Sharrock et al., 1992) Lack of the PV domain in zebrafish Vtg3 and the variability of this domain observed in different Vtgs suggest that this domain is less likely to play an essential role in receptor binding unless there are multiple receptor binding sites present in Vtg Nevertheless, the discovery of the phosvitinless Vtg (Vtg3) provides a natural "mutant" to test the function of this region such as receptor interaction and other potential functions Because all characterized insect and nematode Vtgs lack the PV domain and all vertebrate Vtgs so far characterized 77 Chapter Zebrafish vtg family contain the PV domain, a primitive Vtg in the ancestor of both invertebrates and vertebrates was likely a phosvitinless Vtg Since the PV domain is present in Vtg of the lamprey, a primitive jawless vertebrate (Sharrock et al., 1992), its acquisition must have occurred quite early during vertebrate evolution or even before the appearance of vertebrates The polyserine regions in the PV domain are heavily phosphorylated and may be important in maintaining tertiary structure or in carrying calcium phosphate in support of vertebrate embryonic bone formation (Wahli, 1988) Both functions may increase the fitness of living organisms, resulting in the predominant expression of Vtgs containing the PV domain It is interesting to note that some insect Vtgs also independently evolved polyserine regions near the N- or C-terminals (Chen et al., 1997) This may be a good example of convergent evolution at the molecular level Another variable part is the C-terminal homologous subdomains IV and V (Fig 2-9) or the von Willebrand factor D (VWD) module identified in most members of the large lipid transfer protein (LLTP) superfamily (Babin et al., 1999) In the zebrafish, Vtgs belonging to groups A (Vtg1 and Vtg4-7) and C (Vtg3) lack subdomains IV and V, which are present only in the zebrafish Vtg2 (group B Vtg) (Fig 2-9) Based on the observation that the zebrafish group A vtgs are more similar to group B vtg than to group C, it is speculated that dropping out of the subdomains IV and V in group A Vtgs may be a recent event, which took place during the diversification of vtgs in teleosts The identification of an intermediate group B Vtg only with subdomain IV in the sturgeon probably indicates a progressive missing of the C-terminal part in Vtg during evolution Absence of the homologous subdomains IV and V reflects an ancestral state is further supported by the observation that one of the potential paralogues of Vtg, the large subunit of mammalian 78 Chapter Zebrafish vtg family microsomal triglyceride transfer protein (MTP), contains only one N-terminal large lipid transfer (LLT) module but without the C-terminal VWD module (Babin et al., 1999) Wallace et al (1990) reported that approximately 200 C-terminal residues (including the majority of subdomain IV and the complete subdomain V) are lost in the conversion of Xenopus Vtgs to yolk lipovitellin complexes Similar loss of the C-terminal 199 residues was also implied during the formation of lipovitellin complexes in the lamprey (Sharrock et al., 1992) Thus, it was suggested that the missing polypeptide segments could be involved in receptor-binding and uptake (Sharrock et al., 1992) In addition, a “CGLC” motif was identified at the C-terminal of several vertebrate Vtgs and it was suggested to be involved in protein dimmer formation (Mouchel et al., 1996) In the zebrafish Vtg2, the above motif is located in the homologous subdomain IV (Fig 2-10) and whether a similar loss of the C-terminal part occurred with zebrafish Vtg2 in oocytes is unclear This extended C-terminal region in Vtg2 may be involved in receptor-binding and protein dimmerization If this is the case, then, multiple receptor-binding sites and dimmerization sites must exist in the zebrafish Vtgs, since all other six Vtgs lack the subdomains IV and V 2.4.3 Multiple zebrafish vtg genes are located in two different chromosomes Our mapping results indicated that the zebrafish vtg3 is located in linkage group 11 (chromosome 11), and vtg1, vtg2, vtg4, vtg5 and vtg7 are all located in linkage group 22 (chromosome 22) Searching of the Washington University Zebrafish EST database (http://zfish.wustl.edu/) revealed that eleven vtg ESTs had been mapped to four linkage 79 Chapter Zebrafish vtg family Table 2-5 Percentage identities of nucleotide sequences between mapped vtg EST clones and our seven vtg cDNAs Mapped vtg EST clones† fb62f05 x1 fb62e09 x1 fb59f05 x1 fb59e10 x1 fb59d07 x1 fb64h06 y1 fb63d10 y1 fb62h11 y1 fi49a01 x1 fd52a08 x1 fb59c05 x1 Mapping position‡ 22:1512:GF 22:193:LN 22:1503:GF 22:186:LN 11:463:GF vtg1 80.1% (346) 80.3% (295) 44.4% (374) 22:1517:GF 82.4% 22:192:LN (364) 22:1506:GF 77.4% 22:187:LN (393) 22:176:LN 91.2% (468) 22:176:LN 89.8% (530) 19:152:HS 86.3% (212) 22:1503:GF 90.2% (265) 8:280:LN 98.2% (722) 22:186:LN 39.9% (343) Percentage identity (%) vtg4 vtg5 vtg2 vtg3 60.4% (346) 56.5% (292) 42.4% (403) 61.4% (376) 60.1% (396) 79.1% (459) 82.4% (510) 47.5% (183) 65.6% (250) 70.9% (698) 99.1% (344) 43.4% (339) 44.1% (279) 96.0% (401) 42.2% (372) 42.1% (385) 38.0% (458) 38.4% (524) 37.5% (200) 49.0% (257) 40.6% (719) 38.4% (344) 77.5% (342) 87.5% (295) 44.1% (390) 99.5% (378) 99.0% (396) 94.6% (386) 93.6% (314) 38.8% (206) 89.5% (267) 85.5% (709) 35.8% (335) 83.6% (342) 80.3% (295) 49.2% (366) 79.0% (376) 79.1% (393) 99.5% (214) 100% (141) 43.4% (203) 100% (266) 87.7% (709) 35.2% (335) vtg6 vtg7 81.7% (338) 97.6% (295) 44.3% (361) 90.1% (363) 82.2% (393) 90.4% (468) 89.8% (510) 42.0% (200) 93.6% (267) 87.5% (710) 44.1% (313) 99.7%* (348) 76.8% (280) 48.2% (351) 78.9% (350) 78.3% (369) 95.6% (180) 32.1% (533) 39.3% (206) 87.2% (266) 85.2% (722) 38.6% (321) † Information about mapped vtg EST clones was from WashU-Zebrafish Genome Resources Project (http://zfish.wustl.edu/) ‡ Mapping position is shown as following: Chromosome : centiRay Position : Panel GF, Goodfellow T51 RH panel; LN, LN54 RH panel; HS, heat shock homozygous diploids * ESTs falling into one of the seven vtg cDNA categories are indicated by bold numbers Alignment range (bp) is listed in brackets 80 Chapter Zebrafish vtg family groups, LG8, LG11, LG19 and LG22 After sequence comparison, it was found that these vtg ESTs (except for fb62h11.y1) share 96-100% sequence identities with our seven vtg cDNAs (Table 2-5) Thus, most of them are assumed to be identical to the vtgs identified by us The origin of EST fb62h11 cannot be determined It may represent a novel vtg or vtg-like gene, since it shares the highest sequence identity of only 86.3% (in 214 bp overlap) with vtg1 and was mapped on LG19 by using the HS panel However, it also could derive from one of our identified vtg genes due to the fact that this short EST clone has no overlap region with the available cDNA sequence of either vtg2, vtg4, vtg6 or vtg7 As shown in Fig 2-17A, fb59f05 (identical to vtg3) was mapped to a location very close to that of vtg3 on LG11 by using the same T51 RH mapping panel In contrast, most of the other vtg ESTs, including fb59c05 (identical to vtg2); fb59e10 and fb59d07 (both identical to vtg4); fb64h06, fb63d10 and fi49a01 (all identical to vtg5); fb62e09 (identical to vtg6) and fb62f05 (identical to vtg7), were mapped to LG22 by using either the T51 or LN54 RH panel (Table 2-5) As further illustrated in Fig 2-17B, four vtg ESTs, fb59e10 and fb59d07 (both identical to vtg4), fi49a01 (identical to vtg5) and fb62f05 (identical to vtg7) were mapped in the same genomic region as those for vtg4, vtg5 and vtg7, respectively, although the exact map position is different except for fi49a01, which has the same location as vtg5 Based on the map location of a highly homologous EST clone fb62e09 (with 97.6% sequence identity in 295 nucleotides with vtg6), it was tentatively believed that the zebrafish vtg6 may be also mapped to LG22 Thus, Our mapping result is consistent with the chromosome locations of most mapped vtg ESTs One discrepancy is the map location of fd52a08 (98.2% identity with vtg1), which was mapped to LG8 by 81 Chapter Zebrafish vtg family using the LN54 RH panel The reason for the discrepancy remains unknown and it could be a mapping error Our mapping results of vtgs are also in agreement with the results of genome database searching in that zebrafish vtg2 (group B) is linked with vtgs of group A, while vtg3 (group C) is separated from vtgs of both groups A and B In addition, the possible presence of tandemly located multiple vtg1-like genes in the region spanning 166 kb in chromosome 22 also echoes our mapping result that group A vtgs (vtg1 and vtg4- vtg7) were all mapped in a relatively close region in chromosome 22 The results from gene mapping and genome organization of group A vtgs indicated that the amplification resulting in group A vtgs probably involved multiple rounds of tandem duplication of a vtg1-like precursor gene A similar mechanism was also suggested for the formation of most of the twenty tandemly arrayed vtgs in rainbow trout and for the formation of another reproduction related gene cluster, zona pellucida (zp) 2a, zp2b and zp2c in the zebrafish (Trichet et al., 2000; Mold et al., 2001) 2.4.4 Lineage-specific amplification of teleost vtg genes It has been reported that there are two paralogous vtg systems, Vtg-A and Vtg-B, present in salmonid fishes and the gene copy number varies from to 31 and from to 13 in the above two systems, respectively, depending on genus (Buisine et al., 2002) Based on the observation that about 20% sequence divergence presents between Vtg-A and Vtg-B related genes (Buisine et al., 2002), the relationship between the zebrafish group A and group B vtgs is similar to that between Vtg-A and Vtg-B in salmonid fish, since there is an 18% sequence divergence between the full-length cDNAs of the zebrafish vtg1 and vtg2 82 Chapter Zebrafish vtg family (completed from the Ensembl vtg2-like transcript ENSDART-00000025283) The salmonid Vtg-A related genes are high likely to be categorized as group B vtgs since rainbow trout contains only a Vtg-A system (Trichet et al., 2000) and one of its vtgs (vtg1) belongs to group B (see Section 2.4.1.) Whether the salmonid Vtg-B belongs to group A vtg cannot be determined, since only partial sequences (exons 1-7) of the VtgB genes are available in GenBank from three species: Thymallus thymallus, Coregonus lavaretus and Salmo salar In nine salmonid species studied, five (including the rainbow trout) have undergone preferential amplification of Vtg-A cluster genes, two have undergone preferential amplification of Vtg-B cluster genes and the remaining two have equally low copy numbers of vtgs of both clusters (Buisine et al., 2002) The salmonid species were believed to have undergone a more recent tetraploidization One homologous redundant vtg locus might be lost in the rainbow trout since all 30 copies of the rainbow trout vtgs (20 vtgs plus 10 vtg pseudogenes) are contained in a single 1500-kb region localized at the end of a small acrocentric chromosome (Trichet et al., 2000) It has been proposed that the presence of two retrotransposons (Omr2 and Omr3) and a trace of a HIV-like retroelement in a 4.5-kb rainbow trout vtg1-vtg2 intergenic region renders this region with high recombination potential and facilitated the amplification of juxtaposed ancestral vtgs (Trichet et al., 2000) Subsequent amplification of the whole vtg cluster led to a 3-fold increase in the number of genes (20 copies of group B vtgs) and pseudogenes in rainbow trout (Buisine et al., 2002) In contrast, the present study showed that in zebrafish, only group A vtgs were preferentially amplified, resulting in a relatively high copy number of this group of vtgs (5 copies) when compared with that of group B (1 copy) or C (1 copy) 83 Chapter Zebrafish vtg family It is intriguing to examine whether similar retrotransposon sequences also present in the intergenic regions of the zebrafish group A vtgs Based on the above analysis, it seems that among the three groups of ancestral vtgs, at least two of them (ancestors of groups A and B) have undergone independent amplifications in the zebrafish and rainbow trout, respectively The underlying mechanisms may be due to the random insertion of some retrotransposon elements in the proximal promoter regions of vtgs as in the case of rainbow trout, though evidences are not available from the zebrafish at present The amplification of vtgs may be related to the high demand for yolk proteins during reproduction of these fish species For example, female trout lay an average of 2000 eggs per kg of body weight in a single spawning, which requires the synthesis of tens of grams of vitellogenins (Trichet et al., 2000) Having a shorter spawning cycle, a female zebrafish can produce more than 1000 eggs per week in the laboratory In addition, the zebrafish can produce eggs year round rather than seasonally which inevitably requires the supply of large amount of yolk proteins more frequently 84 ... full-length cDNA sequences of vtg1-7 A8** A17 A 22 A30 A67 A80 A87 A119 A 126 A139 A183 A186 A1 92 A193 A 220 A 227 A248 A250 A2 52 A253 A256 A257 A259 A269 A2 72 A290 A295 A296 A300 A306 A3 42 A349 A368 A371... (24 0) 44% (28 8) 43% (23 5) 47% (24 7) 40% ( 325 ) 49% (22 2) 39% (316) 48% (26 5) 40% (20 8) 47% (25 9) 43% (24 9) 46% ( 328 ) 52% (20 0) 43% (20 5) 43% (21 8) 46% (25 2) 43% (24 4) 48% (21 6) 41% (25 6) 53% (21 7)... 47% (24 4) 43% (318) 78% (22 2) 42% (318) 80% (26 1) 44% (20 6) 44% (25 6) 78% (24 4) 46% ( 323 ) 79% (20 0) 42% (20 7) 79% (21 8) 99% (25 3) 45% (24 4) 44% (21 7) 43% (25 6) 81% (22 1) 88% (28 9) 98% (26 5) 43%