Glycobiology Advance Access originally published online on February 13, 2007
Glycobiology 2007 17(5):516-528; doi:10.1093/glycob/cwm016
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Long-term evolution of the CAZY glycosyltransferase 6 (ABO) gene family from fishes to mammals—a birth-and-death evolution model
2 INSERM U601, Université de Nantes, Institut de Biologie, 9 Quai Moncousu, 44093 Nantes Cedex, France
3 Laboratoire d'Immunogénétique Moléculaire, Université Paul Sabatier, CHU de Toulouse, Laboratoire d'Immunologie, Hôpital Rangueil, TSA 50032, 31059 Toulouse Cedex 9, France
4 L'Etablissement Français Sang Centre Atlantique, 2 Boulevard Tonnellé, Box 52009, 37020 Tours Cedex 1, France
5 Shemyakin Institute for Bioorganic Chemistry, 117871 Moscow, Russia
1 To whom correspondence should be addressed; Tel: +33 5 61 32 34 34; Fax: +33 5 61 32 34 24; e-mail: blancher.a{at}chu-toulouse.fr
Received on December 15, 2006; revised on February 3, 2007; accepted on February 3, 2007
 |
Abstract |
|---|
 Top Abstract IntroductionResultsDiscussionMaterials and methodsSupplementary dataConflict of interest statementAcknowledgmentsReferences |
|---|
Functional glycosyltransferase 6 (GT6) family members catalyze the transfer of galactose or N-acetylgalactosamine in
1,3 linkage to various substrates and synthesize structures related to the A and B histo-blood group antigens, the Forssman antigen, Gal epitope, and iGb3 glycolipid. In rat, mouse, dog, and cow genomes, we have identified three new mammalian genes (GT6m5, GT6m6, and GT6m7) encoding putative proteins belonging to the GT6 family. Among these, GT6m6 protein does not display major alterations of the GT6 motifs involved in binding of the divalent cation and the substrate. Based on protein sequence comparison, gene structure, and synteny, GT6 homologous sequences were also identified in bird, fish, and amphibian genomes. Strikingly, the number and type of GT6 genes varied widely from species to species, even within phylogenetically related groups. In human, except ABO functional alleles, all other GT6 genes are either absent or nonfunctional. Human, mouse, and cow have only one ABO gene, whereas rat and dog have several. In the chicken, the Forssman synthase-like is the single GT6 family member. Five Forssman synthase-like genes were found in zebrafish, but are absent from three other fishes (fugu, puffer fish, and medaka). Two iGb3 synthase-like genes were found in medaka, which are absent from zebrafish. Fugu, puffer fish, and medaka have an additional GT6 gene that we termed GT6m8, which is absent from all other species analyzed here. These observations indicate that individual GT6 genes have expanded and contracted by recurrent duplications and deletions during vertebrate evolution, following a birth-and-death evolution type. Key words: ABO / birth-and-death / evolution / glycosyltransferase / histo-blood group
|
Introduction |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataConflict of interest statementAcknowledgmentsReferences |
|---|
Glycosyltransferases form a functional family of intracellular membrane-bound enzymes. They participate coordinately in the biosynthesis of carbohydrate moieties of glycoproteins and glycolipids, which play an important role in various cellular functions (Varki 1993
). Glycosyltransferases have been classified into families, on the basis of sequence similarities, in the CAZY database (http://www.cazy.org) and, most often, within a given family, the various enzymes transfer the same sugar residue. Enzymes of the glycosyltransferase 6 (GT6) family contribute to the synthesis of histo-blood group-related antigens. These enzymes are type II transmembrane proteins localized in the Golgi complex, with a cytoplasmic tail, a transmembrane domain of about 20 amino acids, and a large luminal catalytic domain. All glycosyltransferases within the CAZY database GT6 family transfer either a galactose (Gal) or an N-acetylgalactosamine (GalNAc) in 1,3 linkage from the uridyl-diphosphate (UDP) nucleotide sugar, retaining the configuration of the transferred Gal or GalNAc (Hennet 2002). Their activity requires the presence of the divalent cation, Mn2 +. It has been shown that a conserved DVD (aspartic acid, valine, aspartic acid) motif interacts with both the nucleotide–sugar and the divalent cation (Boix et al. 2001, 2002; Gastinel et al. 2001; Patenaude et al. 2002). Conserved cysteines have additionally been identified within the GT6 family (Shetterly et al. 2001). Moreover, analysis of the crystal structure of the Ggta1 enzyme and of the A and B histo-blood group enzymes together with molecular modeling allowed characterization of conserved amino acids regions involved in the substrate-binding sites. These regions were called ligand-binding regions (LBRs) (Heissigerova et al. 2003). Additionally, the carboxylic terminal amino acids have to be acidic and seem to play a role in the enzyme function (Heissigerova et al. 2003).
Among the presently known enzymes in the family, the histo-blood group A and B enzymes transfer either a GalNAc or a Gal synthesizing A or B histo-blood group antigens, respectively. In human, they are products of the A and B alleles at the ABO locus, and both use as acceptor substrate the H histo-blood group antigen characterized by the presence of a fucose molecule in the C2 position of the Gal residue of the acceptor. The products of the various O alleles of the ABO gene are inactive, leaving the H antigen unchanged, thus characteristic of the O blood group (Yamamoto 2000). The 1,3galactosyltransferase enzyme or Ggta1, which is responsible for the synthesis of the -Gal epitope found in most mammalian species with the exception of all apes and Old World monkeys, transfers a Gal on an N-acetyllactosamine unit (Galili 2001). The Forssman synthase (FS) transfers a GalNAc on a glycolipid, the globotetraosylceramide (Haslam and Baenziger 1996). From species to species, the Forssman gene seems to be active or inactive. Mouse and chicken are Forssman positive, whereas rat and pigeon are Forssman negative. The last functional member of the GT6 family isogloboside 3 synthase (iGb3S) transfers a Gal on lactosylceramide. This gene is a pseudogene in human, but is active in several other mammals (Keusch et al. 2000).
Searching for new GT6 family members by BLAST in human, rat, and mouse genomes, we identified three putative genes encoding proteins displaying significant homology with GT6 functionally identified enzymes and observed that in the human, rat, and mouse genomes, the corresponding genes were either present or absent. We thus sought to explore the variation in GT6 gene members in various vertebrate species for which significant genome sequence data are available in databanks. For this aim, we searched databanks for genomic GT6 segments and putative GT6 enzyme sequences. Although it has not been possible to define the enzyme activity of all proteins displaying significant homology with GT6 functional members, the analysis allows us to propose a scheme of long-term evolution for this glycosyltransferase gene family.
|
Results |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataConflict of interest statementAcknowledgmentsReferences |
|---|
New mammalian GT6 family members
A blast search in the rat genome with the mouse Ggta1 copy DNA (cDNA) sequence as a query revealed the existence of putative genes displaying homology with the GT6 family. They were termed GT6m5, GT6m6, and GT6m7 for GT6 member 5, 6, and 7, respectively since four members of the family were already known (ABO, FS, Ggta1, and iGb3S). The complete rat cDNA sequences of GT6m5 and GT6m6 were obtained by reverse transcription and rapid amplification of cDNA ends-polymerase chain reaction (RACE-PCR). Transcripts for GT6m5 and GT6m6 were detected in small amounts in some rat tissues: olfactory bulb, lung, kidney, and salivary glands. The highest expression was found in the olfactory bulb (see Supplementary data Tables I and II). The human and macaque GT6m7 cDNAs were also cloned. The human putative protein is shorter than its macaque counterpart due to the presence of a point mutation responsible for a premature stop codon. The latter mutation was observed at homozygosity in all 40 human DNA samples studied by direct sequencing of amplified fragments (data not shown).
The coding sequences of both GT6m5 and GT6m6 are comprised within five exons, whereas that of GT6m7 is comprised within seven exons. One and two additional transcribed noncoding exons were identified upstream of the start codon for GT6m5 and GT6m6, respectively. Three GT6m5 mRNA isoforms were characterized during the amplification of the complete coding sequence: a long 966-bp isoform and two shorter isoforms of 828 and 744 bp, which result from alternative splicing since they lack either only exon 4 or exons 2, 3, and 4, respectively. The reading frame is conserved in the splice variants. Sequence comparison with known members of the GT6 family indicates that exon 4 encodes part of the catalytic domain. The short, putative GT6m5 isoforms are thus expected to lack enzyme activity. At the protein level, rat GT6m5, GTm6, and GT6m7 show 49, 57, and 51% of identity with rat Ggta1, respectively. Their general structure correspond to that of most glycosyltransferases with a short N-terminal cytoplamic domain, a transmembrane domain followed by a stem region and a catalytic domain. However, comparison with other members of the GT6 family reveals significant characteristics. Neither the DxD motif, which interacts with the divalent cation and is conserved in almost all glycosyltransferases, nor the typical cysteine residues of the GT6 family, identified by Shetterly et al. (2001), is present in GT6m5 and GT6m7. Yet, they are present in GT6m6. Likewise, the LBR characterized by Heissigerova et al. (2003) is either clearly altered or absent from both GT6m5 and GT6m7 putative proteins, but not from GT6m6 (Figure 1). Nine LBR have been identified and named from A–I. All are well conserved in the GT6m6 putative protein, except for the arginine or lysine residue of LBR-A, which is replaced by a serine in the rat sequence. The arginine or lysine of LBR-A makes contact with the uridine or ribose of the donor in the human A and B enzymes, as well as in the bovine Ggta1. Notably, in other mammalian species, the arginine at the corresponding position of GT6m6 is conserved (Figure 1). Likewise, LBR-A, as well as LBR-D and LBR-E, which interact with the Gal or GalNAc of the donor substrate or the oligosaccharide acceptor, is partly conserved in GT6m5 and GT6m7. In contrast, the LBR-B, LBR-C (the DxD motif), LBR-F, LBR-G, LBR-H, and LBR-I are completely absent from the GT6m5 and GT6m7 sequences. Amino acids of these motifs are crucial for interactions with the divalent cation, the nucleotide, the pyrophosphate, or the sugar of the donor substrate.
|
Assay of the enzyme activities
The complete coding sequence of rat GT6m5 and GT6m6 were cloned in a eukaryotic expression vector and transfected into Chinese hamster ovary (CHO) cells either expressing an
1,2fucosyltransferase or not. Analysis of cell surface glycosylation using lectins and monoclonal antibodies (mAbs) following transfection did not reveal any change. In contrast, control transfection with the A rat enzyme cDNA of 1,2fucosyltransferase-expressing CHO cells led to A antigenic expression. Likewise, control transfection with the rat Ggta1 cDNA allowed expression of GS1-B4 lectin binding sites (see Supplementary data Figure 1). Soluble forms of the rat GT6m5 and GT6m6 enzymes lacking the intracytoplasmic and transmembrane domains were produced in COS (African green monkey cell line derived from kidney cell transformed by SV40) cells using the pSecTag2 vector. The acceptor substrate specificity of both potential enzymes was tested on a series of oligosaccharides. Since the known enzymes of the GT6 family catalyze the transfer of either Gal or GalNAc, UDP-Gal and UDP-GalNAc were used as donor substrates. No enzyme activity could be detected in conditions were the soluble rat A enzyme prepared in the same manner showed clear activity (see Supplementary data Table III). Thus, we were unable to demonstrate any glycosyltransferase activity for these two new members of the GT6 family.
|
Chromosome localization and structure of the GT6 genes
The three new genes are localized on rat chromosome 3, like most members of the gene family. The most telomeric member of the GT6 family is GT6m7, localized in 3p13 at 1.25 Mb and 30 loci apart from the rat Abo genes. Depending on the strain, two or more Abo genes are present in rat genomes, representing paralogues located at the border between 3p13 and 3p12 (Iwamoto et al. 2002
; Turcot et al. 2003). Further centromeric, 1.7 Mb and 25 loci apart, lies an FS pseudogene in 3p12. In rat, this gene lacks all small 5' exons and retained the two last exons only, in line with the absence of Forssman glycolipid in that species. Much further centromeric, 6.9 Mb apart, in 3p11, lies GT6m5 separated by only one locus (similar to spermidine synthase) from GT6m6 and Ggta1, which are contiguous (Figure 2). The iGb3S locus is on rat chromosome 5. In addition, two fragments of probable processed Abo pseudogenes are found at 1q21 and 15q11 in the rat genome. A comparison with the human genome revealed a general, similar organization of the GT6 genes. The human genome region orthologous to rat chromosome 3p11–3p13 is chromosome 9q34. The most telomeric GT6 family member on this chromosome is GT6m7 like in the rat genome, then come the Abo and FS loci and further centromeric the GGTA1 locus. However, the GT6m5 and GT6m6 genes are not found in the human genome. A further difference between the rat and human genomes is the existence of a single Abo gene in human (Figure 2). The human iGb3S gene is localized on chromosome 1 in a region orthologous to that of the rat genome where the rat iGb3S gene lies. Two processed pseudogenes are found in the human genome, a pseudo GGTA1 on chromosome 12 and a fragment of Abo on chromosme 19. The organization of the GT6 family is similar in the mouse genome, with most members in a genomic region of chromosome 2, orthologous to rat 3p11–3p13 and to human 9q34. Like in the rat and human genomes, the iGb3S locus is unlinked to the other members, located on chromosome 4 in a region orthologous to rat 5q36 and human 1p35. A processed Abo pseudogene is present on chromosome 7. Similar to the rat genome, the mouse genome contains the GT6m5 and GT6m6 genes and, like the human genome, it presents one Abo gene only (not shown). Surfeit gene cluster (Surf1, Surf2, Surf4, Surf5, and Surf6) is closely linked to Abo in mammalian genomes. In Xenopus, Surf6 is found on the same scaffold about 50 kb upstream of Abo-L3. In zebrafish, it is on the same chromosome and there are about 13.7 Mb between Surf5 and the Abo cluster. This shows a conserved synteny between Surfeit genes and the GT6 family members, except iGb3 synthase, throughout vertebrates' evolution.
|
The organization of the coding regions in the rat GT6 genes is shown in Figure 3A. The two last exons are the largest and code for the catalytic domain. The intracytoplasmic domain, the transmembrane domain, and the stem region are encoded by three to five small exons depending upon the gene. The size of introns is also variable from gene to gene. These characteristics are found in all GT6 family members of other species as shown for a few selected examples on Figure 3B. Sequence comparisons indicate that nearly all homologies are concentrated in the two last exons that correspond to the known or potential catalytic domains, almost no homologies exist between sequences of the 5' exons.
|
Evolution of the GT6 gene family
In order to get insights into the long-term evolution of the GT6 family, databanks were searched for sequences presenting clear homologies to the known members described in Figure 2. Sequences clearly belonging to the family were retrieved from various mammalian species, a bird, a frog, and four bony fish species. A related gene could also be found in bacteria, but none was obtained from invertebrates. The GT6 bacterial member belongs to the Escherichia coli strain O86, which is well known to express B antigen on its lipopolysaccharide O-chains (Yi et al. 2005
). In order to make comparisons possible among different species, we used only sequences retrieved from genomes from which a high coverage, assembled or not, was available. Individual GT6 family members can be found from various other species. Yet, the absence of other GT6 sequences in those species may simply indicate a lack of information rather than an absence of the corresponding gene. The retained species were: for mammals, Homo sapiens, Rattus norvegicus, Mus musculus, Canis familiaris, Bos taurus, and Monodelphis domestica; for bird species, only Gallus gallus (the domestic fowl); for frog, Xenopus tropicalis; and for fishes, Takifugu rubripes (fugu), Tetraodon nigroviridis (puffer fish), Danio rerio (zebrafish), and Oryzias latipes (medaka). Sequence alignments generate two major subgroups as shown in Figure 1A and B, respectively. Figure 1A contains the Abo- and FS-related sequences and Figure 1B contains the Ggta1-, iGb3S-, GT6m5-, GT6M6-, and GT6m7-related sequences. Many primate GT6 sequences can be found and the chimpanzee genome is now fully sequenced, but since our aim was to analyze the long-term evolution of the family, we decided to use only human sequences in the present analysis as representative of the primate group. With the exception of human, chimpanzee, and opposum, GT6m5 and GT6m6 were found in all mammalian species for which sufficient genomic data are available. GT6m7 is present in all mammals, except opposum. Yet, the human sequence contains a premature stop codon in the last exon. If the corresponding protein has a biological activity, it is certainly lost in humans. Orthologues to GT6m5, GTm6, and GT6m7 could not be found outside the mammalian group.
A phylogenetic analysis was performed using the alignment of amino acid sequences corresponding to the catalytic domain of known functional proteins (Gastinel et al. 2001; Patenaude et al. 2002), Figure 4. The two major branches with the Abo- and FS-related sequences on one side and the Ggta1, iGb3S, GT6m5, GT6m6, and GT6m7 on the other side are clearly visible. As expected, among mammals, the opposum sequences are more distant, but can be clearly ascribed to an orthologous group. The only exception is a sequence that lies between Ggta1 and GT6m6. Since GT6m5, GT6m6, and GT6m7 could not be found in opposum, this gene could be their ancestor. Alternatively, it could result from an independent duplication of Ggta1 in the opposum lineage. Within the first main branch, mammalian Abo sequences are grouped. Some species, such as man, mouse, pig, and cow, present a single Abo gene, while others, such as rat and dog, present several. Eight Abo related sequences are found in xenope (. tropicalis). Since it is not yet known whether they correspond to active enzymes with either an A or B activity, they have been termed Abo like. The other subgroup of this main branch corresponds to Forssman or FS like. The only GT6 member found in the chicken genome belongs to that subgroup. The zebrafish genome has five distinct Forssman-like sequences, at variance with the other fishes which have none. The second main branch of the tree shows an iGb3S subgroup with all the mammalian sequences and two sequences from each of the three fishes fugu, tetraodon, and medaka. The latter were termed iGb3S-like since the enzyme activity has not been described for any of these fish sequences. Surprisingly, no such sequence was found in the zebrafish genome. Of the three new genes GT6m5, GT6m6, and GT6m7 described in the first paragraph of the Results section, GT6m6 is the most closely related to Ggta1. In addition to the iGb3S-like sequences, each of the three fish species, medaka, fugu, and tetraodon possesses one more sequence of the GT6 family that cannot be unambiguously ascribed to one of the two main branches. These new sequences cannot be considered orthologous to one of the other GT6 family members and we called them GT6m8 for GT6 family member 8. Finally, the E. coli sequence is used as outgroup. Its region corresponding to the catalytic domain shows 11% identity with rat Abo1 and clearly belongs to the GT6 family as judged from its homology with all other members of the family. This is unambiguously confirmed by the recent demonstration of its B blood group enzyme activity (Yi et al. 2006).
|
It appears that irrespective of whether they correspond to active enzymes or not, the number of GT6 members largely differ from genome to genome even within phylogenetically related subgroups like mammals or bony fishes. Individual genes of the GT6 family have thus expanded or contracted by duplications and deletions in the course of their evolution.
|
Discussion |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataConflict of interest statementAcknowledgmentsReferences |
|---|
Three new mammalian GT6 members are described, but it is not clear whether they are functional in any species. Both GT6m5 and GT6m7 lack several motifs that appear essential for catalytic activity, in particular for binding to the donor substrate and the divalent cation. Not so surprisingly, we were unable to demonstrate any enzyme activity for rat GT6m5 and preliminary assays using a limited number of substrates on macaque GT6m7 were negative as well (data not shown). GT6m5 and GT6m7 could thus correspond to pseudogenes. Phylogenetic analysis indicated that GT6m5, GT6m6, and GT6m7 evolved faster than the other mammalian GT6 members since they formed longer branches within the mammalian group. This would suggest that they benefit from lower structural constraints to maintain a biological function or more likely, that they are easily dispensable so that less negative selection was exerted in the past on these three GT6 genes as compared to the others. Yet, their conservation without gross alterations in distant mammalian genomes is surprising. It contrasts with the accumulation of nonsense mutations or deletions in Ggta1 of apes on a much shorter period of time (Koike et al. 2002
). Whether they have any function at all thus remains to be seen. Unlike the two previous sequences, GT6m6 shows a good conservation of all essential motifs. Yet, we were unable to demonstrate enzyme activity of rat GT6m6. Our data indicate that it does not possess a redundant activity with the previously known GT6 enzymes. If it has an enzyme activity, it may not correspond to the addition of a Gal or a GalNAc in 1,3-linkage like other GT6 members. Even though within a glycosyltransferase family, the various enzymes generally transfer the same sugar residue, there are exceptions such as ß3GnT1, ß3GnT3, and ß3GnT5, which belong to a galactosyltransferase family but catalyze the transfer of a GlcNAc or 4GalT whose closest homologue is an 4GlcNAc transferase (Zhou et al. 1999; Steffensen et al. 2000; Shiraishi et al. 2001; Togayachi et al. 2001). It is thus possible that GT6m6 transfers a sugar other than Gal and GalNAc. Alternatively, it may be nonfunctional in rat, but functional in other species like human FS, that is, inactive even though it does not show obvious sequence alterations (Xu et al. 1999). Phylogenetic studies are limited by the large divergence of orthologous sequences when a long phylogenetic scale is considered. They are also limited by the lack or partial nature of data from many important species and by the difficulty in genome reconstruction from crude sequences. Thus, distinguishing allelic sequences from paralogues is not necessarily straightforward. Considering these limitations, we concentrated on sequences unambiguously belonging to the GT6 family because of important similarities with well-known GT6 members characterized by the presence of at least some of the previously defined functional motifs, by the overall conservation of the gene structure and by their synteny in the genome.
Sequences with clear features characteristic of the GT6 family were found in all vertebrate species for which data are available, but could not be unambiguously detected in invertebrates.
Peptide sequences showing weak homologies with GT6 members were found in anopheles, but it is not clear whether they correspond to GT6 proteins since they do not present any of the conserved motifs characteristic of the family in vertebrates. Nevertheless, some invertebrate species express carbohydrate antigens related to histo-blood group antigens. The A antigen or A-like antigen has recently been detected in oysters with anti-A mAbs or the HPA lectin. It is recognized by human norovirus strains that use it as a ligand both in oysters, concentrating the virus, and in human where the virus replicates, causing gastroenteritis (Le Guyader et al. 2006). Likewise, HPA-binding sites have been found in species belonging to other invertebrate orders, indicating the presence of terminal GalNAc residues (Fang and Welsch 1995; Evangelista and Leite 2002). The enzymes involved in the synthesis of these invertebrate carbohydrate structures may be phylogenetically too distant to allow unambiguous classification in the GT6 family. Alternatively, they may be unrelated enzymes or they may be absent from databanks. The presence of bacterial sequences such as the protein with B blood group enzyme activity from E. coli strain 086 (Yi et al. 2005), clearly belonging to the GT6 family is probably due to lateral transfer or possibly to convergent evolution (Yi et al. 2006).
Although it was not possible to find an orthologue of the common GT6 ancestor, a scenario of the GT6 family evolution in vertebrates can be proposed (Figure 5). Three GT6 subfamilies, likely deriving from a common ancestor, are already present in teleost fishes, FS-like, iGb3S-like, and a member with no clear mammalian orthologue, GT6m8. The latter has probably been lost from all terrestrial vertebrates. In this scenario, FS-like would be the ancestor of both FS and ABO through a duplication that occurred in an ancestor of amphibians. IGb3S-like would have been duplicated in a mammalian ancestor giving rise to iGb3S and Ggta1. Since then, the latter has been duplicated twice, first to generate an ancestor of GT6m5 and GT6m7 and more recently to generate GT6m6. Considering the long time scale, chromosome localizations in mammals are compatible with this model since Abo and FS are close from each other, Ggta1, GT6m6, and GT6m7 are next to each other, whereas iGb3S is on a distinct chromosome in human, chimpanzee, and rodents.
|
According to the species considered, some gene members have been further duplicated or inversely have been lost. The chicken genome, which is the only bird species for which near complete genomic information is available, has only one GT6 member with high similarity to FS. Consistent with the absence of A or B antigens as well as of the
Gal xenoantigen in chicken and the presence of Forssman glycolipid in chicken and ostrich (Bouhours et al. 1999; Oriol et al. 1999), it indicates that the evolutionary branch leading to birds has lost the genes of at least two GT6 subfamilies.
The single amphibian genome (X. tropicalis) available for the analysis presents eight genes related to the ABO subfamily and none to the other subfamilies. Since these genes are located in tandem in the genome, they probably occurred by repeated duplications from a common ancestor. Surprisingly, none of the other GT6 subfamilies are represented in Xenopus. The large number of ABO-like genes in X. tropicalis is compatible with the large number of A and B blood group typical and unusual structures characterized from the jelly coat of various frogs eggs (Strecker et al. 1995; Delplace et al. 2002; Coppin et al. 2003).
Opposum is a representative of marsupial mammals, which diverged from placental mammals about 170 million years ago. Four GT6 members were found in its genome. An ABO gene, a FS gene, a Ggta1 gene, and a gene related to Ggta1 that could correspond either to an opossum-specific duplication or that could be an ancestor of GT6m5, GT6m6, or GT6m7.
All GT6 members derived from the ancestors of FS and iGb3S are represented in mammals. Yet, important differences exist among species in this group with the expansion of the Abo gene in some species and the complete loss or the loss of function of other genes in other species. Thus, of all GT6 family members, only the products of the A and B alleles at the ABO locus are functional in human. Ggta1, FS, and iGb3S appear inactivated because of promoter alterations and of aberrant splicing (Taylor et al. 2003; Milland et al. 2006), GT6m7 is inactivated and both GT6m5 and GT6m6 are lacking. This situation of expansion or deletion of genes of a given family in distinct species is typical of the model of birth-and-death evolution described by Nei et al. (2005) at variance with models of divergent evolution or concerted evolution. Birth-and-death evolution of a multigene family is characterized by the creation of new genes by duplication that can be maintained in the genome for long periods of time, whereas some duplicates are deleted or inactivated through mutations. This model of evolution seems to apply particularly well to gene families involved in interactions with the environment since the selective pressures exerted on species living in different environments will not be identical. They include gene families of the immune system, such as the major histocompatibility complex, the natural killing receptors, and defensins, or of the sensory systems, such as the taste and olfactory receptors (Nei et al. 1997; Conte et al. 2003; Niimura and Nei 2003; Hao and Nei 2004, 2005; Lynn et al. 2004).
Although the biological roles of the GT6 family members are not well defined as yet, evidence has accumulated suggesting that they play roles in interactions with pathogens and in innate immunity. The A and B antigens are ligands for some pathogens and this correlates with sensitivity to the pathogen as recently shown for some norovirus strains (Tan and Jiang 2005; Le Pendu et al. 2006). The Forssman antigen is a ligand of pathogenic E. coli strains. Its absence from some species prevents cross-species transmission of these pathogenic bacteria (Xu et al. 1999). Natural antibodies against the A and B antigens as well as against the Gal xenoantigen synthesized by Ggta1 may, respectively, restrict the intra-species or trans-species spread of viruses expressing the corresponding antigens (Rother et al. 1995; Takeuchi et al. 1996; Seymour et al. 2004; Neil et al. 2005). The iGb3 glycolipid was recently shown to be an endogenous ligand for mouse iNKT lymphocytes (Zhou et al. 2004). These nonconventional lymphocytes recognize glycolipids presented by the CD1d molecule to an invariant T-cell receptor. They play essential roles at the interface between the innate and adaptative immune responses and are involved in anti-bacterial responses in the control of autoimmunity as well as in the immunity against cancer cells (Godfrey and Kronenberg 2004). They can directly recognize some bacterial glycolipids or regulate immune responses after recognition of an endogenous glycolipid (Mattner et al. 2005). The B blood group antigen is a developmentally regulated marker of some sensitive neurons, suggesting that it could be involved in hearing and olfaction (Mollicone et al. 1985; Astic et al. 1989). In this respect, it is interesting to note that a main site of expression of GT6m5 and GT6m6 was the olfactory bulb. It has also been suggested that oligosaccharides synthesized by members of the GT6 family could participate in gamete interactions, although this is debated (Litscher et al. 1995; Johnston et al. 1998). Amphibians show a large repertoire of such oligosaccharides on their jelly coat that differ from species to species. They could play a role in species-specific gamete interactions (Coppin et al. 2003). Such a role would be highly compatible with the birth-and-death evolution character of the GT6 family since a distinct set of genes in different species would be required.
In conclusion, we have shown that the GT6 family has more members than previously appreciated although it is not clear if the new genes described are functional in any species. Likewise, the biological activities of the ABO, FS, and iGb3 synthase found in fish, frog, or bird remain to be demonstrated. A long-term phylogenetic analysis of the GT6 family in vertebrates shows a clear mode of birth-and-death evolution consistent with the putative biological functions of the various members of the family.
|
Materials and methods |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataConflict of interest statementAcknowledgmentsReferences |
|---|
Cloning of rat GT6m5, GT6m6, and
3GalT (Ggta1)Cloning of these cDNAs was performed before the release of the rat genome sequences in databanks, and the sequence was deposited in the NCBI database (http://www.ncbi.nlm.nih.gov).
Identification and cloning of GT6m5
Briefly, a BLAST search using the mouse 3GalT sequence (M 85153) revealed the existence of unknown rat expressed sequence tags (EST). Screening by real time-PCR of various rat cell lines and tissues showed that a corresponding sequence could be amplified from the mammary carcinoma cell line MT450 and from testis. The complete coding sequence was obtained by RACE-PCR using the Marathon cDNA Amplification kit (Clontech, Saint-Germain-en-Laye, France) from the MT450 and the testis cDNA libraries. It was then amplified from the rat testis cDNA library. Three isoforms were identified (966, 828, and 744 bp). The longest isoform was then cloned in the pCR3.1 + vector (Invitrogen, Cergy Pontoise, France) and sequenced.
Identification and cloning of GT6m6
In order to search for new members of the GT6 family, degenerated primers (MotAB2 and MotAB3i), deduced from the available sequences of the ABO and Ggta1 genes in various mammalian species, were used to amplify a fragment of 200 bp. New primers were designed from this fragment. Elongation in the 3' direction was performed on rat Genome Walker cDNA library (Clontech). Elongation in the 5' direction was performed on a rat stomach cDNA. The complete coding sequence was obtained from the rat testis cDNA library and the product was cloned in pcDNA3.1+.
Cloning of the rat Ggta1
The degenerated primers used to obtain the GT6m6 sequence allowed to isolate a second fragment of 200 bp. Both 3' and 5' direction elongations were performed on a rat stomach cDNA library by RACE-PCR using the Marathon cDNA Amplification kit (Clontech). The complete coding sequence was obtained from the rat testis cDNA library. The product was cloned in the pCR3.1+ vector.
Sequence analysis, chromosome localization, and phylogenic analysis
Sequences belonging to the GT6 family were searched using BLASTN, TBLASTN, and BLAT with default parameters from all genomic and EST sequences available in the NCBI, ENSEMBL (http://www.ensembl.org/index.html), MEDAKA project (http://dolphin.lab.nig.ac.jp/medaka/index.php), and FlyBase (http://flybase.bio.indiana.edu/blast/) databases. Human and rat sequences were used as queries in the first round of search. The retrieved chicken, Xenopus, and fish sequences were used as queries in a second round of search to ensure completeness of the search process. Since we aimed at analyzing the long-term evolution of GT6 family, only human sequences were considered for primates. All genomic sequences capable to generate a complete catalytic domain, thus potential functional enzymes, were considered. The only exception was the human GT6m7 gene that contains a stop codon that should lead to a truncated catalytic domain. Since this nonsense codon is not present in the chimpanzee orthologous sequence and that a single nucleotide change could generate a complete coding peptide sequence, the corrected human sequence was used. The human FS sequence was also included since it corresponds to a full-length protein although it is known to be a nonfunctional enzyme. Processed pseudogenes were excluded from the analysis. Since no sequence fulfilling the above criteria could be retrieved from invertebrate genomic data, the study was restricted to vertebrates. All sequences and their accession numbers used in the study are listed in Table I. Approximate localizations of transmembrane regions were determined using the TMHMM program available from CBS prediction servers (http://www.cbs.dtu.dk/services). Multiple alignments were performed with ClustalW 1.8 (http://searchlauncher.bcm.tmc.edu/seq-search/nucleic_acid-search.html). Phylogenetic trees were constructed by Neighbor-Joining with p distances and 1000 bootstrap replicates using MEGA3 (http://www.megasoftware.net).
|
Supplementary data |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataConflict of interest statementAcknowledgmentsReferences |
|---|
Supplementary data are available at Glycobiology online (http://glycob.oxfordjournals.org/).
|
Conflict of interest statement |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataConflict of interest statementAcknowledgmentsReferences |
|---|
None declared.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataConflict of interest statementAcknowledgmentsReferences |
|---|
This work was supported in part by the Ministère Français de la Recherche (contrat E.A. 3034), funds from EFS (Etablissement Français du Sang), the Association for International Cancer Research (AICR), by the Association pour la Recherche sur le Cancer (ARC), and by the Institut National de la Santé et de la Recherche Médicale. We also thank Dr. G. Strecker (Université des Sciences et des Technologies de Lille, France) for his gift of purified milk oligosaccharides.
|
Abbreviations |
|---|
cDNA, copy DNA; CHO, Chinese hamster ovary; COS, African green monkey cell line derived from kidney cell transformed by SV40; DVD, aspartic acid, valine, aspartic acid; EST, expressed sequence tags; FS, Forssman synthase; Gal, galactose; GalNAc, N-acetylgalactosamine; Ggta1,
1,3galactosyltransferase; GT6, glycosyltransferase 6 family; GT6m5, glycosyltransferase family member 5; GT6m6, glycosyltransferase family member 6; GT6m7, glycosyltransferase family member 7; iGb3S, isogloboside 3 synthase; LBR, ligand binding region; mAb, monoclonal antibody; RACE-PCR, rapid amplification of cDNA ends-polymerase chain reaction; UDP, uridyl-diphosphate. |
References |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataConflict of interest statementAcknowledgmentsReferences |
|---|
Astic L, Le Pendu J, Mollicone R, Saucier D, Oriol R. Cellular expression of H and B antigens in the rat olfactory system during development. J Comp Neurol (1989) 289:386–394.[CrossRef][Web of Science][Medline]
Boix E, Swaminathan GJ, Zhang Y, Natesh R, Brew K, Acharya KR. Structure of UDP complex of UDP-galactose:ß-galactoside--1,3-galactosyltransferase at 1.53-A resolution reveals a conformational change in the catalytically important C terminus. J Biol Chem (2001) 276:48608–48614.
Boix E, Zhang Y, Swaminathan GJ, Brew K, Acharya KR. Structural basis of ordered binding of donor and acceptor substrates to the retaining glycosyltransferase, -1,3-galactosyltransferase. J Biol Chem (2002) 277:28310–28318.
Bouhours D, Liaigre J, Richard C, Oriol R, Bouhours JF. Forssman penta- and tetraglycosylceramide are xenoantigens of ostrich kidney and liver. Glycobiology (1999) 9:875–886.
Conte C, Ebeling M, Marcuz A, Nef P, Andres-Barquin PJ. Evolutionary relationships of the Tas2r receptor gene families in mouse and human. Physiol Genomics (2003) 14:73–82.
Coppin A, Florea D, Maes E, Cogalniceanu D, Strecker G. Comparative study of carbohydrate chains released from the oviducal mucins of the two very closely related amphibian species Bombina bombina and Bombina variegata. Biochimie (2003) 85:53–64.[Medline]
Delplace F, Maes E, Lemoine J, Strecker G. Species specificity of O-linked carbohydrate chains of the oviducal mucins in amphibians: structural analysis of neutral oligosaccharide alditols released by reductive beta-elimination from the egg-jelly coats of Rana clamitans. Biochem J (2002) 363:457–471.[CrossRef][Web of Science][Medline]
Evangelista LG, Leite AC. Histochemical localization of N-acetyl-galactosamine in the midgut Lutzomyia longipalpis (Diptera: Psychodidae). J Med Entomol (2002) 39:432–439.[Web of Science][Medline]
Fang YO, Welsch U. A histochemical study of the distribution of lectin binding sites in the developing oocytes of the lancelet Branchiostoma belcheri. Cell Tissue Res (1995) 280:427–434.[Web of Science][Medline]
Galili U. The -gal epitope (Gal 1–3Gal ß 1–4GlcNAc-R) in xenotransplantation. Biochimie (2001) 83:557–563.[Medline]
Gastinel LN, Bignon C, Misra AK, Hindsgaul O, Shaper JH, Joziasse DH. Bovine 1,3-galactosyltransferase catalytic domain structure and its relationship with ABO histo-blood group and glycosphingolipid glycosyltransferases. EMBO J (2001) 20:638–649.[CrossRef][Web of Science][Medline]
Godfrey DI, Kronenberg M. Going both ways: immune regulation via CD1d-dependent NKT cells. J Clin Invest (2004) 114:1379–1388.[CrossRef][Web of Science][Medline]
Hao L, Nei M. Genomic organization and evolutionary analysis of Ly49 genes encoding the rodent natural killer cell receptors: rapid evolution by repeated gene duplication. Immunogenetics (2004) 56:343–354.[Web of Science][Medline]
Hao L, Nei M. Rapid expansion of killer cell immunoglobulin-like receptor genes in primates and their coevolution with MHC Class I genes. Gene (2005) 347:149–159.[CrossRef][Web of Science][Medline]
Haslam DB, Baenziger JU. Expression cloning of Forssman glycolipid synthetase: a novel member of the histo-blood group ABO gene family. Proc Natl Acad Sci USA (1996) 93:10697–10702.
Heissigerova H, Breton C, Moravcova J, Imberty A. Molecular modeling of glycosyltransferases involved in the biosynthesis of blood group A, blood group B, Forssman, and iGb3 antigens and their interaction with substrates. Glycobiology (2003) 13:377–386.
Hennet T. The galactosyltransferase family. Cell Mol Life Sci (2002) 59:1081–1095.[CrossRef][Web of Science][Medline]
Iwamoto S, Kumada M, Kamesaki T, Okuda H, Kajii E, Inagaki T, Saikawa D, Takeuchi K, Ohkawara S, Takahashi R, et al. Rat encodes the paralogous gene equivalent of the human histo-blood group ABO gene. Association with antigen expression by overexpression of human ABO transferase. J Biol Chem (2002) 277:46463–46469.
Johnston DS, Wright WW, Shaper JH, Hokke CH, Van den Eijnden DH, Joziasse DH. Murine sperm-zona binding, a fucosyl residue is required for a high affinity sperm-binding ligand. A second site on sperm binds a nonfucosylated, ß-galactosyl-capped oligosaccharide. J Biol Chem (1998) 273:1888–1895.
Keusch JJ, Manzella SM, Nyame KA, Cummings RD, Baenziger JU. Expression cloning of a new member of the ABO blood group glycosyltransferases, iGb3 synthase, that directs the synthesis of isoglobo-glycosphingolipids. J Biol Chem (2000) 275:25308–25314.
Koike C, Fung JJ, Geller DA, Kannagi R, Libert T, Luppi P, Nakashima I, Profozich J, Rudert W, Sharma SB, et al. Molecular basis of evolutionary loss of the 1,3-galactosyltransferase gene in higher primates. J Biol Chem (2002) 277:10114–10120.
Le Guyader F, Loisy F, Atmar RL, Hutson AM, Estes MK, Ruvoen-Clouet N, Pommepuy M, Le Pendu J. Norwalk virus-specific binding to oyster digestive tissues. Emerg Infect Dis (2006) 12:931–936.[Web of Science][Medline]
Le Pendu J, Ruvoen-Clouet N, Kindberg E, Svensson L. Mendelian resistance to human norovirus infections. Semin Immunol (2006) 18:375–386.[CrossRef][Web of Science][Medline]
Litscher ES, Juntunen K, Seppo A, Penttila L, Niemela R, Renkonen O, Wassarman PM. Oligosaccharide constructs with defined structures that inhibit binding of mouse sperm to unfertilized eggs in vitro. Biochemistry (1995) 34:4662–4669.[CrossRef][Medline]
Lynn DJ, Lloyd AT, Fares MA, O'Farrelly C. Evidence of positively selected sites in mammalian alpha-defensins. Mol Biol Evol (2004) 21:819–827.
Mattner J, Debord KL, Ismail N, Goff RD, Cantu C 3rd, Zhou D, Saint-Mezard P, Wang V, Gao Y, Yin N, et al. Exogenous and endogenous glycolipid antigens activate NKT cells during microbial infections. Nature (2005) 434:525–529.[CrossRef][Medline]
Milland J, Christiansen D, Lazarus BD, Taylor SG, Xing PX, Sandrin MS. The molecular basis for gal(1,3)gal expression in animals with a deletion of the 1,3galactosyltransferase gene. J Immunol (2006) 176:2448–2454.
Mollicone R, Trojan J, Oriol R. Appearance of H and B antigens in primary sensory cells of the rat olfactory apparatus and inner ear. Brain Res (1985) 349:275–279.[Medline]
Nei M, Gu X, Sitnikova T. Evolution by the birth-and-death process in multigene families of the vertebrate immune system. Proc Natl Acad Sci USA (1997) 94:7799–7806.
Nei M, Rooney AP. Concerted and birth-and-death evolution of multigene families. Annu Rev Genet (2005) 39:121–152.[CrossRef][Web of Science][Medline]
Neil SJ, McKnight A, Gustafsson K, Weiss RA. HIV-1 incorporates ABO histo-blood group antigens that sensitize virions to complement-mediated inactivation. Blood (2005) 105:4693–4699.
Niimura Y, Nei M. Evolution of olfactory receptor genes in the human genome. Proc Natl Acad Sci USA (2003) 100:12235–12240.
Oriol R, Candelier JJ, Taniguchi S, Balanzino L, Peters L, Niekrasz M, Hammer C, Cooper DK. Major carbohydrate epitopes in tissues of domestic and African wild animals of potential interest for xenotransplantation research. Xenotransplantation (1999) 6:79–89.[CrossRef][Web of Science][Medline]
Patenaude SI, Seto NO, Borisova SN, Szpacenko A, Marcus SL, Palcic MM, Evans SV. The structural basis for specificity in human ABO(H) blood group biosynthesis. Nat Struct Biol (2002) 9:685–690.[CrossRef][Web of Science][Medline]
Rother RP, Fodor WL, Springhorn JP, Birks CW, Setter E, Sandrin MS, Squinto SP, Rollins SA. A novel mechanism of retrovirus inactivation in human serum mediated by anti--galactosyl natural antibody. J Exp Med (1995) 182:1345–1355.
Seymour RM, Allan MJ, Pomiankowski A, Gustafsson K. Evolution of the human ABO polymorphism by two complementary selective pressures. Proc Biol Sci (2004) 271:1065–1072.
Shetterly S, Tom I, Yen TY, Joshi R, Lee L, Wang PG, Macher BA. 1,3 galactosyltransferase: new sequences and characterization of conserved cysteine residues. Glycobiology (2001) 11:645–653.
Shiraishi N, Natsume A, Togayachi A, Endo T, Akashima T, Yamada Y, Imai N, Nakagawa S, Koizumi S, Sekine S, Narimatsu H, Sasaki K. Identification and characterization of three novel ß 1,3-N-acetylglucosaminyltransferases structurally related to the ß 1,3-galactosyltransferase family. J Biol Chem (2001) 276:3498–3507.
Steffensen R, Carlier K, Wiels J, Levery SB, Stroud M, Cedergren B, Nilsson Sojka B, Bennett EP, Jersild C, et al. Cloning and expression of the histo-blood group Pk UDP-galactose: Ga1ß-4G1cß1-cer 1, 4-galactosyltransferase. Molecular genetic basis of the p phenotype. J Biol Chem (2000) 275:16723–16729.
Strecker G, Wieruszeski JM, Plancke Y, Boilly B. Primary structure of 12 neutral oligosaccharide-alditols released from the jelly coats of the anuran Xenopus laevis by reductive beta-elimination. Glycobiology (1995) 5:137–146.
Takeuchi Y, Porter CD, Strahan KM, Preece AF, Gustafsson K, Cosset FL, Weiss RA, Collins MK. Sensitization of cells and retroviruses to human serum by ( 1–3. galactosyltransferase. Nature (1996) 379:85–88.[CrossRef][Medline]
Tan M, Jiang X. Norovirus and its histo-blood group antigen receptors: an answer to a historical puzzle. Trends Microbiol (2005) 13:285–293.[CrossRef][Web of Science][Medline]
Taylor SG, McKenzie IF, Sandrin MS. Characterization of the rat (1,3)galactosyltransferase: evidence for two independent genes encoding glycosyltransferases that synthesize Gal(1,3)Gal by two separate glycosylation pathways. Glycobiology (2003) 13:327–337.
Togayachi A, Akashima T, Ookubo R, Kudo T, Nishihara S, Iwasaki H, Natsume A, Mio H, Inokuchi J, Irimura T, et al. Molecular cloning and characterization of UDP-GlcNAc:lactosylceramide ß 1,3-N-acetylglucosaminyltransferase (ß 3Gn-T5), an essential enzyme for the expression of HNK-1 and Lewis X epitopes on glycolipids. J Biol Chem (2001) 276:22032–22040.
Turcot AL, Blancher A, Le Moullac-Vaidye B, Despiau S, Rocher J, Roubinet F, Szpirer C, Le Pendu J. Cloning of a rat gene encoding the histo-blood group B enzyme: rats have more than one Abo gene. Glycobiology (2003) 13:919–928.
Varki A. Biological roles of oligosaccharides: all of the theories are correct. Glycobiology (1993) 3:97–130.
Xu H, Storch T, Yu M, Elliott SP, Haslam DB. Characterization of the human Forssman synthetase gene. An evolving association between glycolipid synthesis and host-microbial interactions. J Biol Chem (1999) 274:29390–29398.
Yamamoto F. Molecular genetics of ABO. Vox Sang (2000) 78(Suppl 2):91–103.[Web of Science][Medline]
Yi W, Shao J, Zhu L, Li M, Singh M, Lu Y, Lin S, Li H, Ryu K, Shen J, et al. Escherichia coli O86 O-antigen biosynthetic gene cluster and stepwise enzymatic synthesis of human blood group B antigen tetrasaccharide. J Am Chem Soc (2005) 127:2040–2041.[CrossRef][Web of Science][Medline]
Yi W, Zhu L, Guo H, Li M, Li J, Wang PG. Formation of a new O-polysaccharide in Escherichia coli O86 via disruption of a glycosyltransferase gene involved in O-unit assembly. Carbohydr Res (2006) 341:2254–2260.[CrossRef][Web of Science][Medline]
Zhou D, Dinter A, Gutierrez Gallego R, Kamerling JP, Vliegenthart JF, Berger EG, Hennet T. A ß-1,3-N-acetylglucosaminyltransferase with poly-N-acetyllactosamine synthase activity is structurally related to ß-1,3-galactosyltransferases. Proc Natl Acad Sci USA (1999) 96:406–411.
Zhou D, Mattner J, Cantu C 3rd, Schrantz N, Yin N, Gao Y, Sagiv Y, Hudspeth K, Wu YP, Yamashita T, et al. Lysosomal glycosphingolipid recognition by NKT cells. Science (2004) 306:1786–1789.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. S. Schaefer, G. M. Richter, M. Nothnagel, T. Manke, H. Dommisch, G. Jacobs, A. Arlt, P. Rosenstiel, B. Noack, B. Groessner-Schreiber, et al. A genome-wide association study identifies GLT6D1 as a susceptibility locus for periodontitis Hum. Mol. Genet., February 1, 2010; 19(3): 553 - 562. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Tumbale and K. Brew Characterization of a Metal-independent CAZy Family 6 Glycosyltransferase from Bacteroides ovatus J. Biol. Chem., September 11, 2009; 284(37): 25126 - 25134. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Casals, A. Ferrer-Admetlla, M. Sikora, A. Ramirez-Soriano, T. Marques-Bonet, S. Despiau, F. Roubinet, F. Calafell, J. Bertranpetit, and A. Blancher Human pseudogenes of the ABO family show a complex evolutionary dynamics and loss of function Glycobiology, June 1, 2009; 19(6): 583 - 591. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Tumbale, H. Jamaluddin, N. Thiyagarajan, K R. Acharya, and K. Brew Screening a limited structure-based library identifies UDP-GalNAc-specific mutants of {alpha}-1,3-galactosyltransferase Glycobiology, December 1, 2008; 18(12): 1036 - 1043. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||