Isolation, characterization and expression analysis of a hypoxia-responsive glucose transporter gene from the grass carp, Ctenopharyngodon idellus Ziping Zhang, Rudolf S. S. Wu, Helen O. L. Mok, Yilei Wang, Winnie W. L. Poon, Shuk H. Cheng and Richard Y. C. Kong Department of Biology and Chemistry and Centre for Coastal Pollution and Conservation, City University of Hong Kong, Kowloon Tong, Hong Kong Special Administrative Region, People’s Republic of China Glucose transporters (GLUTs) have been implicated in adaptive and survival responses to hypoxic stress in mam- mals. In fish, the expression and regulation of GLUT in relation to hypoxia remains unexplored. Here we describe the identification of a hypoxia-responsive glucose transpor- ter gene (gcGLUT) and the corresponding full-length cDNA from the grass carp. The gene spans  11 kb of genomic sequence and consists of 12 exons and 11 introns, and an open reading frame (ORF) of 1599 bp encoding a poly- peptide of 533 amino acids, with a predicted molecular mass of  57 kDa and a pI of 8.34. BLASTX analysis showed that the ORF shared high sequence identity with the GLUT1 (57–59%), GLUT3 (59–60%) and GLUT4 (55–59%) pro- teins from different vertebrates. Comparative analysis of GLUT genomic structures showed that the arrangement of exons and position of split codons are highly conserved amongst members of the class I GLUTs suggesting that these genes share a common ancestor. Phylogenetic ana- lysis indicated that gcGLUT is most closely related to the GLUT3 proteins. Northern blot analysis showed that the 3.1-kb gcGLUT transcript was most abundantly expressed and responsive to hypoxia in kidney. Up-regulated expression by hypoxia was also evident in eye and gill, but differential patterns of expression were observed. Low expression levels detected in brain, heart, liver and muscle were not responsive to hypoxic stress. Keywords: Ctenopharyngodon idellus; glucose transporter; grass carp; hypoxia; split codon. Glucose transporters (GLUTs) are members of a structur- ally related family of membrane glycoproteins that facilitate cellular uptake of glucose and are ubiquitiously expressed in mammalian cells in a tissue-specific manner [1]. At least 13 different GLUT isoforms have been described in vertebrates to date, and based on amino acid sequence similarities they can be grouped into three subclasses [2]. Class I members include GLUTs 1–4; class II members include GLUTs 5, 7, 9 and 11; and class III members include GLUTs 6, 8, 10, 12 and the proton/myoinositol cotransporter [3]. Although structurally very similar, these isoforms have different tissue distribution, subcellular localizations, kinetic characteristics, and regulatory properties [1,4] and may be attributed to a different glucose requirement by various embryonic [5] and adult [6] tissues. Expression of the GLUT1 and GLUT3 genes in mammals is known to be induced by hypoxic stress and is mediated by the basic helix-loop-helix transcription factor, hypoxia- inducible factor-1 [7], presumably via its binding to the hypoxia-responsive DNA elements in these genes [8]. Increased expression of these genes in hypoxic tissues has been associated with enhanced glucose utilization to facili- tate the supply of metabolic energy [9]. While much is known about the distribution and regulation of these genes and their responses to hypoxia in mammals, the corres- ponding information in fish is not known. Although a number of GLUT genes have recently been described in various fish species [10–13], nothing is known about the hypoxia responsiveness of these genes. Here, we describe the cloning and genomic structure of a hypoxia-responsive glucose transporter gene, gcGLUT from the grass carp and the characterization of its in vivo expression and response pattern to short- and long-term hypoxia. Experimental procedures Animals Grass carp, Ctenopharyngodon idellus, weighing around 500 g, were obtained from a commercial hatchery and acclimated in 300-L fibreglass tanks with circulating, filtered and well-aerated tap water at 20 °C for 1 week prior to experimentation. Fish were fed daily with lettuce that amounted to  1% of body weight. Fish were then divided into two groups, one group was reared under normoxia Correspondence to R. Y. C. Kong, Department of Biology and Chemistry, City University of Hong Kong, 83 Tat Chee Avenue, Kowloon Tong, Hong Kong. Fax: + 852 2788 7406, Tel.: + 852 2788 7794, E-mail: bhrkong@cityu.edu.hk Abbreviation: GLUT, glucose transporter. (Received 28 February 2003, revised 3 May 2003, accepted 19 May 2003) Eur. J. Biochem. 270, 3010–3017 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03678.x (7.0 ± 0.2 mg O 2 ÆL )1 ) and the other under hypoxia (0.5 ± 0.3 mg O 2 ÆL )1 ) in a continuous flow system des- cribed by Zhou et al. [14]. The levels of dissolved oxygen were monitored continuously using a YSI Model 580 dissolved oxygen meter. After the exposure period, fish were anaesthetized by immersion in 2-phenoxyethanol (0.05% v/v) for 5 min, and killed by a blow to the head. Tissues were then dissected out and snap-frozen in liquid nitrogen, and stored at )80 °C. Animal care and experiments were conducted in accordance with the City University of Hong Kong animal care guidelines. RNA isolation and cloning of full-length cDNAs Total RNA was prepared from grass carp tissues using the Trizol reagent (Invitrogen) according to the manufacturer’s instructions. Poly(A) + RNA was purified from total RNA using the PolyATract System kit (Promega). The primers GLUT1-F (5¢-ATGAGCAGAAATCGAGGGCTCTC-3¢) and GLUT1-R (5¢-ACAGCCCTCAGAGGAGCCCTT- 3¢) derived from the common carp GLUT1 sequence (AF247730), were used to amplify a 0.2-kb GLUT1-like cDNA fragment by RT-PCR with total RNA from grass carp kidney. PCR in a 100-uL mixture was performed on first-strand cDNAs that were reverse transcribed from total RNA by use of Superscript II reverse transcriptase (Invi- trogen) and consisted of 20 ng of first strand cDNA, 1 · PCR buffer (20 m M Tris/HCl pH 8.4, 50 m M KCl), 1 l M of each primer, 0.2 m M of dNTPs, 1.5 m M MgCl 2 and 5U of Taq DNA polymerase (Invitrogen). The PCR program consisted of predenaturation at 94 °Cfor3min, followed by 35 cycles of amplification (denaturation at 94 °C for 20 s, annealing at 55 °C for 1 min, and extension at 72 °C for 1 min) and a final extension at 72 °Cfor10 min in a Gene Cycler (Bio-Rad, USA). The amplified DNA fragment was subcloned into a pGEM-T vector (Promega) and DNA sequencing showed that it shared 100% nucleo- tide similarity to common carp GLUT1. The 0.2-kb cDNA subclone (designated as RK-1) was used as a probe to screen a grass carp kidney cDNA library that was prepared in kTriplEx2 in our laboratory using the Smart cDNA library construction kit (Clontech). A single 1.6-kb cDNA clone (RK-2) was obtained and DNA sequencing showed that it shared 76% nucleotide similarity with RK-1. 5¢-RACE and 3¢-RACE were performed using the Marathon cDNA amplification kit (Clontech) to obtain the full-length cDNA sequence for RK-2 with poly(A) + RNA purified from the kidney of a hypoxic grass carp. The adaptor primers AP1 and AP2 were purchased from Clontech. Gene-specific nested primers for 5¢-RACE were: Primer A, 5¢-TGTC AGTCCTGTACAAAGAC-3¢ and Primer B, 5¢-CAT CAGGCTTCCCCATA-3¢. Gene-specific nested primers for 3¢-RACE were: Primer C, 5¢-CCAGTGTCCCCAT CATCAG-3¢ and Primer D, 5¢-GGCAATTTTAAA GTCATTATGGCGCAAA-3¢. First strand cDNAs were synthesized using Superscript II RNase H – reverse tran- scriptase (Invitrogen) according to the manufacturer’s instructions. PCRs were performed using 1 · Advantage2 Taq polymerase mix (Clontech) in 50-uL reactions which contained 20 ng of first strand cDNA, 1 · PCR buffer (20 m M Tris/HCl pH 8.4, 50 m M KCl), 0.2 l M of each primer, 1.5 m M MgCl 2 and 0.2 m M of dNTPs. PCR amplification was performed in a Gene Cycler (Bio-Rad, USA) under the following conditions: 94 °C, 30 s followed by 94 °C, 5 s; 72 °C, 3 min (5 cycles); 94 °C, 5 s, 70 °C, 3 min (5 cycles); 94 °C, 5 s; 68 °C, 3 min (25 cycles). RACE products were cloned into a pGEM-T vector (Promega) for DNA sequencing. Full-length cDNAs were obtained by reverse-transcription PCR using gene-specific primers: GT1-F, 5¢-CCTGATCGACGCACGAGT-3¢ and GT1-R, 5¢-TTTTGCAAGTCATAGTAATCAGTTT-3¢ for GT- cDNA1 (2150 bp); and GT2-F, 5¢-CACCAGCAACTAC CTGATCGA-3¢ and GT2-R, 5¢-CACAAAATATGCTT CCAAGTGC-3¢ for GT-cDNA2 (3043 bp). Construction and screening of a grass carp genomic DNA library Genomic DNA was extracted from grass carp liver by the use of Genomic-tips (Qiagen) according to the manufac- turer’s instructions. Genomic DNA was partially digested with Sau3AI and fragments larger than 9.5 kb were ligated into BamHI-digested EMBL3 arms (Stratagene) and pack- aged into Escherichia coli XL1-Blue MR cells using Gigapack Gold Packaging Extract (Stratagene). Approxi- mately 30 000 plaque forming units were screened with the 3-kb GT-cDNA2 fragment radiolabeled with [a- 32 P]dCTP by random priming. Hybridization was performed in ExpressHyb solution (Clontech) at 65 °C for 2 h and one strongly hybridizing phage clone (kgH-1) was identified and further characterized by restriction enzyme digestion with BstXI, HindIII, PstIandXbaI, and southern hybridization analysis using GT-cDNA2 as a probe. Northern blot analysis Total RNA (20 lg) from different tissues was electropho- resed on 1% agarose/formaldehyde gels and blotted onto nylon membrane (Hybond-XL, Amersham Biosciences). DNA probes were radiolabeled by the random priming method and 2.0 · 10 6 c.p.m.ÆmL )1 were used in northern hybridizations and were carried out at 65 °Cfor2hin ExpressHyb solution (Clontech). Blots were washed thrice in 2 · NaCl/Cit (1 · NaCl/Cit ¼ 0.15 M NaCl, 0.015 M sodium citrate), 0.05% (w/v) SDS at room temperature for 10 min, and twice in 0.1 · NaCl/Cit, 0.1% SDS at 50 °C for 20 min. Blots were exposed on a phosphor screen (Kodak-K) at room temperature for 20 h, and the signals were captured using the Molecular Imager FX System (Bio- Rad). A 115-bp 28S rDNA fragment was amplified from grass carp genomic DNA using primers 28S-F (5¢-GAT CCTTCGATGTCGGCTCT-3¢) and 28S-R (5¢-CTAA CCTGTCTCACGACGGT-3¢), and was used as an internal control probe in Northern hybridization for normalization of gcGLUT expression. Phylogenetic analysis Phylogenetic analysis was performed by maximum parsi- mony using the PROTPARS program of the PHYLIP package version 3.57c [15]. Support for the inferred clades was obtained by bootstrap analysis from 1000 replications of the data set using the SEQBOOT and CONSENSE programs. Phylogenetic tree was displayed using TREEVIEW [16]. Ó FEBS 2003 A hypoxia-responsive glucose transporter gene from grass carp (Eur. J. Biochem. 270) 3011 Sequence analyses and homology searches were performed using the online BLAST suite of programs (NCBI, USA). Statistical analysis A nonparametric v 2 test was used to test the hypothesis that the ratio of expression level in the hypoxic treatment group (n ¼ 4) was not significantly different from the normoxic control (n ¼ 4) at different time intervals. A one-way ANOVA was used to test if there was any significant difference in gene expression levels between different exposure periods. Where the null hypothesis was rejected, a Tukey’s test was performed to identify significant difference between indi- vidual means; a ¼ 0.05 was used in all statistical tests. Results Isolation of two gcGLUT cDNAs generated from alternative polyadenylation sites In an attempt to identify GLUT-like cDNA sequences, a grass carp kidney cDNA library was screened with a 0.2-kb grass carp GLUT1-likecDNA(RK-1)fragmentthatwas derived by RT-PCR. A single 1.6-kb cDNA clone (desig- nated as RK-2) that showed strong positive hybridization was identified and DNA sequencing showed that it shared 76% nucleotide sequence similarity with the RK-1 DNA probe. Further analysis indicated that RK-2 contained an incomplete GLUT ORF and lacked the start codon. Using 5¢-and3¢-RACE PCR, two overlapping full-length cDNA clones of 2.1 kb (GT-cDNA1) and 3.1 kb (GT-cDNA2) that shared 100% nucleotide sequence identity were obtained. Sequence analysis showed that both GT-cDNA1 and GT-cDNA2 contain a 5¢-UTR of 203 bp and a coding region of 1599 bp (excluding the stop codon), but the 3¢-UTR of GT-cDNA1 is 345 bp while that of GT- cDNA2 is 1238 bp (Fig. 1A; GenBank accession number AY231475). The results indicated that GT-cDNA1 and GT-cDNA2 are derived from the same GLUT gene from alternative use of polyadenylation sites. This was corrob- orated by Northern blot analysis in which two transcripts  2.1and3.1kbinsizeweredetectedinthetotalRNAof grass carp kidney; the larger transcript showed a 30-fold higher expression level than the former (data not shown). Further analysis of the ORF showed that it encodes a putative polypeptide of 533 amino acids, with a predicted molecularmassof 57 kDa and a pI of 8.34. A database search using BLASTX showed that the ORF shared high sequence identity with the GLUT1 (57–59%), GLUT3 (59–60%) and GLUT4 (55–59%) proteins, and moderate Fig. 1. Organization of the gcGLUT gene. (A) The 12 exons of the gcGLUT gene are shown as boxes. Filled and open boxes indicate translated and untranslated regions, respect- ively. The position of the start (ATG) and stop (TAA) codons are indicated by inverted arrows (fl).The 5¢-end of exon 1 was inferred by 5¢-RACE and the two alternate 3¢-ends of exon 12 were deduced by 3¢-RACE, and are delineated by the full-length cDNA clones, GT-cDNA1 and GT-cDNA. The two puta- tive polyadenylation sites (ATTAAA) are indicated by an upright arrow (›). (B) The exon/intron boundaries of the split codons for arginine (between exon 4 and exon 5) and valine (between exons 6 and exon 7) are shown. Exonic regions are shown in uppercase and intronic regions are in lowercase. The split codons are boxed and highlighted in gray. 3012 Z. Zhang et al. (Eur. J. Biochem. 270) Ó FEBS 2003 (GLUT2; 48–50%) to low (GLUTs 5–13; < 40%) sequence identity with other GLUT types from various vertebrate species. Characteristics of the deduced amino acid sequence of gcGLUT Analysis of the deduced amino acid sequence of GT- cDNA1 with the HMMTOP program (http://www.enzim.hu/ hmmtop/) [17] indicated the presence of 12 putative transmembrane helices, and alignment to human GLUT1, GLUT3 and GLUT4 (with which it shares high sequence identity) indicated a high degree of structural conservation and the presence of typical sugar transporter motifs common to all members of the class I GLUTs [2]. These include: a putative N-glycosylation site in extracellular loop 1; the STSIF motif in loop 7 (the third S residue is substituted by an E); the PESPR/PETKGR motifs after transmembrane helix 6 and 12; GRR motifs in intracellular loops 2 and 8; glutamate and arginine residues in intracel- lular loops 4 and 10; the QL motif in transmembrane helix 5; the QLS motif in transmembrane helix 7; and the three tryptophan residues in transmembrane helix 2, 6 and 11 (motifs are highlighted in bold type in Fig. 2). These features suggest that GT-cDNA1 encodes for a class I glucose transporter and is hereupon designated as gcGLUT. A striking feature of gcGLUT is the presence of a relatively longer loop 9 sequence that contains a putative N-glycosy- lation site that is not present in GLUT1–GLUT4. Moreover, the FGY motif which is highly conserved in transmembrane helix 1 is changed to FGF in gcGLUT; a change which was alsoobservedinmammalianGLUT8[18]. In an attempt to ascertain the phylogenetic affinity of gcGLUT, a phylogenetic tree consisting of GLUT1, GLUT2, GLUT3 and GLUT4 proteins was constructed by maximum parsimony ( PROTPARS ) and bootstrapped with 1000 replications using the PHYLIP package version 3.57c [15]. As shown in Fig. 3, gcGLUT was found to cluster in the same clade with the GLUT3 proteins, although it was supported by a bootstrap value of only 41%. Phylogenetic analysis using the neighbor-joining method also produced a similar tree of the same topology (data not shown). Genomic structure of gcGLUT GT-cDNA2 was used to screen a kEMBL-3 grass carp genomic library from which a phage clone (kgH-1) was obtained and was characterized by restriction mapping and Southern blot analyses (data not shown). Appropriate fragments that showed positive hybridization were cloned into pBluescript and sequenced on both strands, and gaps in the sequences were filled by primer walking. A contiguous stretch of  14 kb of genomic sequence was obtained. The exon/intron boundaries were identified by comparing the genomic sequence with the full-length cDNA sequence and conform to the invariant gt/ag sequences at the 5- and 3-splice sites, respectively (data not shown). As shown in Fig. 1A, the gcGLUT gene (GenBank accession number AY231476) spans  11 kb of genomic DNA and contains 12 exons and 11 introns. The 5¢-UTR (203 bp) is contained in exons 1 and 2; the coding region (1599 bp) is distributed across exons 2–12 and the 3¢-UTR is located within a 1238-bp stretch of sequence (corresponding to 3¢-UTR of Fig. 2. Alignment of gcGLUT with the human GLUT1, GLUT3 and GLUT4 proteins. hGLUT1 (accession number AAA52571), hGLUT3 (accession number AAB61083), hGLUT4 (accession number AAA59189). Amino acids are designated by single-letter codes. Dashes (–) indicate gaps inserted for improved alignment. Amino acid positions are indicated on the right. Boxshade is used to highlight regions with different levels of sequence identity: identical amino acids in at least three sequences are in black, and similar amino acids in at least two sequences are in gray. Functional motifs described in other GLUTs are highlighted in boldtype (see text). The putative 12-transmembrane helices are boxed and labeled transmembrane helix 1–12. Upright triangles (n) indicate the exon splice sites and corresponding exon domains. Filled triangles (m) indicate codon splitting positions of arginine-96 and valine-231. Ó FEBS 2003 A hypoxia-responsive glucose transporter gene from grass carp (Eur. J. Biochem. 270) 3013 GT-cDNA2) at the 3¢ end of exon 12. Examination of the 3¢-flanking genomic sequence revealed two putative poly- adenylation (ATTAAA) signals: one is located 18 bp upstream from the poly(A) of GT-cDNA1 and another is located 11 bp upstream from the poly(A) of GT-cDNA2 (data not shown). Of particular interest is the presence of eight AUUUA motifs in the 3¢-UTR of GT-cDNA2 compared to only one in GT-cDNA1 (data not shown). This sequence motif is a potential adenosine-uridine-binding factor site that has been identified as important for regulating mRNA stability [19], as has been also reported for GLUT1 [20] and GLUT3 [21]. Comparative analysis of gcGLUT to members of the class I (human GLUT1, GLUT2, GLUT3 and GLUT4), class II (human GLUT5) and class III (human GLUT10) extended GLUT family [2] revealed marked structural similarities in genomic organization amongst members of the class I subfamily (Table 1). Six of the exons of gcGLUT (exons 4–9) encoding for the region spanning transmem- brane helix 2 to transmembrane helix 9 (Fig. 2) are identical in size to six respective exons (exons 3–8) in human GLUT1 and GLUT3 (exons 3–8), and four in human GLUT2 (exons 7–10) and GLUT4 (exons 6–9) (Table 1). The codons for arginine (96) and valine (231) in gcGLUT (Fig. 2) are split between exons 4 and 5, and exons 6 and 7, respectively (Fig. 1B). Whilst computer analysis indicated that codon splitting at the first site is also conserved in human GLUT1 and GLUT3, codon splitting at the second site is conserved in all four class I human GLUTs (data not shown). These observations therefore suggest that gcGLUT and human GLUTs1–4 arose by duplication of a common ancestral gene encoding these specific domains. In vivo expression and response pattern of gcGLUT to short and long-term hypoxia To study the in vivo expression and response pattern of gcGLUT to hypoxia, grass carp were exposed to normoxic (7 mg O 2 ÆL )1 ) and hypoxic (0.5 mg O 2 ÆL )1 ) conditions and fish (n ¼ 4) were sampled from each treatment group and control after 4, 96 and 170 h. Total RNA was isolated from seven different tissues of each of four fish from the normoxic and hypoxic groups at each time point for Northern blot analysis. A representative autoradiogram is given in Fig. 4. Under normoxic conditions, the 3.1-kb gcGLUT mRNA transcript was most abundantly expressed in kidney; however, lower levels of expression were also detected in all other tissues examined; brain, eye, gill, heart, liver and muscle. Exposure to hypoxia for 4, 96 and 170 h resulted in a marked and persistent increase in gcGLUT expression in kidney, while hypoxic induction was only observed in gill at 4 h, and eye at 4 and 170 h. In vivo expression of 3.1-kb gcGLUT transcript was seemingly unaffected by both short and long-term hypoxia in brain, heart, liver and muscle of grass carp at all time points examined. Interestingly, the less abundant 2.1-kb gcGLUT transcript also showed promin- ent expression and hypoxia up-regulation ( threefold) in kidney; however, it was barely detectable in all other tissues examined under both normoxic and hypoxic conditions (data not shown). Expression levels of gcGLUT in all replicates of each tissue under normoxic and hypoxic conditions were nor- malized against 28S rRNA and were found to vary considerably within each tissue type as well as each time point. A Chi square test was used to identify whether expression level was significantly different between each hypoxic treatment and the respective normoxic control. One way analysis of variance was performed to test the hypothesis that there was no significant difference in expression level between different time points within each tissue type. Where significant differences were identified (P<0.05), pairwise comparisons were carried out using a Dunnett’s test. All statistical analyses were carried out using Graphpad PRISM (version2). The analysis showed that statistically significant increases in gcGLUT expression levels were observed only in eye (1.5 ± 0.2 fold at 4 and 170 h; P < 0.05), gill (1.7 ± 0.13 and 1.4 ± 0.19 fold at Fig. 3. Phylogenetic analysis of gcGLUT. An unrooted tree depicting the phylogenetic relatedness of gcGLUT to the known GLUT1– GLUT4 proteins from fish, avian or mammalian sources. The protein sequences obtained from the GenBank/EMBL/Swissprot databases include: common carp ccGLUT1 (AAF75683); rainbow trout rtGLUT1A (AAF75681); chicken GLUT1 (AAB02037); mouse GLUT1 (AAA37752); rat GLUT1 (P11167) rabbit GLUT1 (P13355); bovine GLUT1 (P27674); human GLUT1 (AAA52571); rtGLUT2 (AAK09377); chicken GLUT2 (Q90592); human GLUT2 (AAA59514); mouse GLUT2 (P14246); rat GLUT2 (P12336); chicken GLUT3 (AAA48662); mouse GLUT3 (AAH34122); rat GLUT3 (Q07647); rabbit GLUT3 (Q9XSC2); human GLUT3 (AAB61083); dog GLUT3 (P47842); bovine GLUT3 (AAK70222); sheep GLUT3 (P47843); brown trout btGLUT4 (AAG12191); bovine GLUT4 (Q27994); rat GLUT4 (P19357); mouse GLUT4 (P14142); and human GLUT4 (AAA59189). The bootstrap support (SEQBOOT program, PHYLIP package) for each branch (1000 replications) is shown. 3014 Z. Zhang et al. (Eur. J. Biochem. 270) Ó FEBS 2003 4 and 170 h, respectively; P < 0.05) and kidney (2 ± 0.35 fold at 4 h, 2.7 ± 0.9 fold at 96 h and 2.2 ± 0.2 fold at 170 h; P < 0.05). Discussion In the present study, we have isolated and characterized the structure and expression pattern of a hypoxia-responsive glucose transporter gene, gcGLUT, from the grass carp. Computer analysis of the deduced amino acid sequence predicted that gcGLUT is a 12-transmembrane spanning protein and that it possesses all the major structural features and sequence motifs characteristic of a functional class I glucose transporter, and include: (1) the QLS residues in transmembrane helix 7 which is required for high-affinity transport of glucose [22]; (2) the two arginine residues (336/ 337) in the conserved GRR motif in intracellular loop 8 [23] and proline residues in transmembrane helix 6 and trans- membrane helix 10 [24] which are essential for glucose transport activity; and (3) the serine/threonine residues at positions 298 and 299 in loop 7 (Fig. 2) that are involved in conformational change of the GLUT protein during transport [25]. To date, cDNAs of five GLUT isotypes have been described in fish and sequence comparison showed that gcGLUT shares a sequence identity of 58% with GLUT1 of common carp [13], 57% with GLUT1A of rainbow trout [12], 59% with GLUT4 of brown trout [10], 58% with the GLUT4-like protein (accession number AAM22227) of coho salmon, and 50% with GLUT2 of rainbow trout [11]. No report of GLUT3 has yet been described in fish. Although we were unable to predict the actual isoform of gcGLUT based on sequence identity scores, maximum parsimony (Fig. 3) and neighbor-joining (data not shown) analyses both indicated that gcGLUT is phylogenetically more similar to GLUT3 than to other class I GLUTs. Comparative analysis of the genomic organization of gcGLUT with different human GLUT genes showed that exons 4–9 of the gcGLUT gene, that encode for the region spanning transmembrane helix 2 to transmembrane helix 9 (Fig. 2), share strong structural homology with six of the respective exons in the hGLUT1 and hGLUT3 Table 1. Exonic structure conservation in gcGLUT and selected human GLUT genes. Values are shown as the exon size (bp) distribution. The stretch of homologous exons that are conserved amongst different GLUT genes are highlighted in bold type. Accession numbers of the respective GLUT genes are in parentheses. Exon Class I (human) Class II (human) GLUT5 (NT_028054) Class III (human) GLUT10 (NT_011362) gcGLUT (AY231476) GLUT1 (NT_004852) GLUT2 (NT_034563) GLUT3 (NT_024397) GLUT4 (NT_010823) 1 86 197 279 256 233 207 253 2 141 96 45 93 117 99 1284 3 102 161 93 161 173 161 123 4 161 241 263 241 125 125 136 5 241 163 125 163 116 153 2590 6 163 188 116 188 163 126 7 188 105 163 105 188 188 8 105 102 188 102 105 111 9 102 204 105 204 102 102 10 57 1398 102 219 204 76 11 201 2566 2615 657 128 12 1468 845 Fig. 4. In vivo expression and response pattern to hypoxia of gcGLUT. A representative Northern blot derived from the tissues of one normoxic and one hypoxic fish from a total of four in each group is shown. Total RNA (20 lg) samples from different tissues of fish subjected to normoxia (N) and hypoxia (H) for 4 h, 96 h and 170 h were analysed by Northern hybridization using GT-cDNA2 (gcGLUT) and a 115-bp grass carp 28S rDNA fragment as probes. Quantitation was performed by normalizing gcGLUT expression levels against the 28S rRNA. Ó FEBS 2003 A hypoxia-responsive glucose transporter gene from grass carp (Eur. J. Biochem. 270) 3015 genes, and four of the respective exons in hGLUT2 and hGLUT4 (homologous exons are shown in bold type in Table 1). Moreover, whilst the nature and position of the split codon for arginine 96 (divided between exons 4 and 5; Fig. 2) is conserved in gcGLUT, hGLUT1 and hGLUT3; the position of the split codon for valine-231 (divided between exon 6 and exon 7) is conserved in gcGLUT, hGLUT1, hGLUT2, hGLUT3 and hGLUT4. Overall, the analysis indicated that the genomic organiza- tion of gcGLUT is structurally more similar to the hGLUT1 and hGLUT3 genes. Computer analysis of the mouse GLUT1–GLUT4 genes, which are highly homo- logous to the human counterparts, also showed conserved homology in these stretch of exons (data not shown). Moreover, when version 3 of the Fugu rubripes genome (http://genome.jgi-psf.org/fugu6/fugu6.home.html) was queried with the gcGLUT coding sequence, four candidate Fugu GLUT genes that share  56–70% sequence identity with gcGLUT were obtained, and all showed a pattern of exon sizes similar to gcGLUT. Overall, these observations strongly indicate that members of the class I GLUT subfamily may have arisen by duplication of a common ancestral gene encoding these domains and that there is a high selective pressure to maintain the arrangement of these exons. In mammals, hypoxic stress is known to increase GLUT1 and GLUT3 expression in specific tissues to enhance the uptake rate of glucose both to facilitate the supply of metabolic energy [9] and to protect cells from hypoxic injury [26]. Fish often have to contend with low and variable oxygen levels in the aquatic environment. Although, it has been reported that hypoxia causes significant changes in plasma glucose level [27,28] and glucose flux [29] in fish, until now, nothing has been known about the regulation or tissue-specific expression pattern of fish GLUTs in response to hypoxic stress. Here, we have demonstrated for the first time a hypoxia- responsive GLUT gene that is both most prominently expressed and responsive to hypoxia in the carp kidney. Moreover, although gcGLUT expression was markedly lower in eye and gill, up-regulated differential expression patterns during short (4 h) and long-term (96 and 170 h) hypoxia were also evident in these two organs (Fig. 4), implying a difference in glucose transport and regulation in different tissues during hypoxic stress. It is conceivable that modulation of glucose transport for a continuous supply of energy is important in the fish kidney (for osmoregulatory activities), an organ which is known to show high glucose uptake rates [30]. So far, no GLUT gene that shows predominant expression in kidney has been reported, and further studies have yet to be done to determine the functional characteristics and regulation of this apparent kidney-specific GLUT, in particular its physiological role(s) in relation to hypoxia adaptation and tolerance in fish. Acknowledgements This work was supported by a Central Earmarked Research Grant (Project No. CityU1057/99 M ) from the Research Grants Council of Hong Kong Special Administrative Region, People’s Republic of China. References 1. Mueckler, M. (1994) Facilitative glucose transporters. Eur. J. Biochem. 219, 713–725. 2. Joost, H.G. & Thorens, B. (2001) The extended GLUT-family of sugar/polyol transport facilitators: nomenclature, sequence char- acteristics, and potential function of its novel members. Mol. Memb. Biol. 18, 247–256. 3. Uldry, M., Ibberson, M., Horisberger, J.D., Chatton, J.Y., Rie- derer, B. & Thorens, B. (2001) Identification of a mammalian H + - myoinositol symporter expressed predominantly in the brain. EMBO J. 20, 4467–4477. 4. Baldwin, S.A. (1993) Mammalian passive glucose transporters: members of an ubiquitous family of active and passive transport proteins. Biochim. Biophys. Acta. 1154, 17–49. 5. Carver, F.M., Shibley, I.A. Jr, Pennington, J.S. & Pennington, S.N. (2001) Differential expression of glucose transporters during chick embryogenesis. Cell. Mol. Life. Sci. 58, 645–652. 6. Bell, G.I., Burant, C.F., Takeda, J. & Gould, G.W. (1993) Structure and function of mammalian facilitative sugar transpor- ters. J. Biol. Chem. 268, 19161–19164. 7. Iyer,N.V.,Kotch,L.E.,Agani,F.,Leung,S.W.,Laughner,E., Wenger, R.H., Gassmann, M., Gearhart, J.D. & Lawler, A.M., Yu, A.Y. & Semenza, G.L. (1998) Cellular and developmental control of O 2 homeostasis by hypoxia-inducible factor 1a. Genes Dev. 12, 149–162. 8. Ebert, B.L., Firth, J.D. & Ratcliff, P.J. (1995) Hypoxia and mitochondrial inhibitors regulate expression of glucose transporter-1 via distinct cis-acting sequences. J. Biol. Chem. 270, 29083–29089. 9. Bunn, H.F. & Poyton, R.O. (1996) Oxygen sensing and molecular adaptation to hypoxia. Physiol. Rev. 76, 839–885. 10. Planas, J.V., Capilla, E. & Gutierrez, C.J. (2000) Molecular identification of a glucose transporter from fish muscle. FEBS Lett. 481, 266–270. 11. Krasnov, A., Teerijoki, H. & Molsa, H. (2001) Rainbow trout (Onchorhynchus mykiss) hepatic glucose transporter. Biochim. Biophys. Acta. 1520, 174–178. 12. Teerijoki, H., Krasnov, A., Pitkanen, T.I. & Molsa, H. (2000) Cloning and characterization of glucose transporter in teleost fish rainbow trout (Oncorhynchus mykiss). Biochim. Biophys. Acta. 1494, 290–294. 13. Teerijoki, H., Krasnov, A., Pitkanen, T.I. & Molsa, H. (2001) Monosaccharide uptake in common carp (Cyprinus carpio)EPC cells is mediated by a facilitative glucose carrier. Comp. Biochem. Physiol. (Part B). 128, 483–491. 14. Zhou, B.S., Wu, R.S.S., Randall, D.J. & Lam, P.K.S. (2001) Bioenergetics and RNA/DNA ratios in the common carp (Cyprinus carpio) under hypoxia. J. Comp. Physiol. B. 171, 49–57. 15. Felsenstein, J. (1995) PHYLIP (Phylogeny Inference Package), Version 3.572. Distributed over the World Wide Web, Seattle. 16. Page, R.D.M. (1996) TREEVIEW: An application to display phylogenetic trees on personal computers. Comp. Appl. Biosci. 12, 357–358. 17. Tusnay, G.E. & Simon, I. (1998) Principles governing amino acid composition of integral membrane proteins: Applications to topology prediction. J. Mol. Biol. 283, 489–506. 18. Doege, H., Schurmann, A., Bahrenberg, G., Brauers, A. & Joost, H.G. (2000) GLUT8, a novel member of the sugar transport facilitator family with glucose transport activity. J. Biol. Chem. 21, 16275–16280. 19. Asson-Batres, M., Spurgeon, S., Diaz, J., DeLoughery, T. & Bagby, G. (1994) Evolutionary conservation of the AU-rich 3¢-untranslated region of messenger RNA. Proc.NatlAcad.Sci. USA 91, 1318–1322. 3016 Z. Zhang et al. (Eur. J. Biochem. 270) Ó FEBS 2003 20. Boada, R.J. & Pardridge, W.M. (1993) Glucose deprivation causes post-transcriptional enhancement of brain capillary endothelial glucose transporter gene expression via GLUT1 mRNA stabili- zation. J. Neurochem. 60, 2290–2296. 21. Borson, N.D., Salo, W.L. & Drewes, L.R. (1996) Canine brain glucose transporter 3: gene sequence, phylogenetic comparisons and analysis of functional sites. Gene 168, 251–256. 22. Seatter, M.J., De La Rue, S.A., Porter, L.M. & Gould, G.W. (1998) QLS motif in transmembrane helix VII of the glucose transporter family interacts with the C-1 position of D -glucose and is involved in substrate selection at the exofacial binding site. Biochemistry 37, 1322–1326. 23. Wandel, S., Schuermann, A., Becker, W., Summers, S.A., Shanahan, M.F. & Joost, H.G. (1995) Mutation of two conserved arginine residues in the glucose transporter GLUT4 suppresses transport activity, but not glucose-inhibitable binding of inhibitory ligands. Naunyn-Schmiedeberg’s Arch. Pharmacol. 353, 36–41. 24. Wellner, M., Monden, I., Mueckler, M.M. & Keller, K. (1995) Functional consequences of proline mutations in the putative transmembrane segments 6 and 10 of the glucose transporter GLUT1. Eur. J. Biochem. 227, 454–458. 25. Doege, H., Schurmann, A., Ohnimus, H., Monser, V., Holman, G.D. & Joost, H.G. (1998) Serine-s94 and threonine-295 in the exofacial loop domain between helixes 7 and 8 of glucose trans- porters (GLUT) are involved in the conformational alterations during the transport process. Biochem. J. 329, 289–293. 26. Lin, Z., Weinberg, J.M., Malhotra, R., Merritt, S.E., Holzman, L.B. & Brosius, F.C.I.I.I. (2000) GLUT-1 reduces hypoxia- induced apoptosis and JNK pathway activation. Am.J.Physiol. Endocrinol. Metab. 278, E958–E966. 27. White, A. & Fletcher, T.C. (1989) The effect of physical dis- turbance, hypoxia and stress hormones on serum components of the plaice, Pleuronectes platessa L. Comp. Biochem. Physiol. 93A, 455–461. 28. Kakuta, I., Namba, K., Uemaisu, K. & Murachi, S. (1992) Effects of hypoxia on renal function in carp, Cyprinus carpio. Comp. Biochem. Physiol. 101A, 769–774. 29. Haman, F., Zwingelstein, G. & Weber, J.M. (1997) Effects of hypoxia and low temperature on substrate fluxes in fish: plasma metabolite concentrations are misleading. Am. J. Physiol. 273, R2046–R2054. 30. Blasco, J., Fernandez-Borras, J., Marimon, I. & Requena, A. (1996) Plasma glucose kinetics and tissue uptake in brown trout in vivo: effect of an intravascular glucose load. J. Comp. Physiol. 165B, 534–541. Ó FEBS 2003 A hypoxia-responsive glucose transporter gene from grass carp (Eur. J. Biochem. 270) 3017 . 5¢-CCTGATCGACGCACGAGT-3¢ and GT1-R, 5¢-TTTTGCAAGTCATAGTAATCAGTTT-3¢ for GT- cDNA1 (2150 bp); and GT2-F, 5¢-CACCAGCAACTAC CTGATCGA-3¢ and GT2-R, 5¢-CACAAAATATGCTT CCAAGTGC-3¢. Isolation, characterization and expression analysis of a hypoxia-responsive glucose transporter gene from the grass carp, Ctenopharyngodon idellus Ziping