Surgical Oncology Research Laboratories, Massachusetts General Hospital; and Department of Surgery, Harvard Medical School, Boston, Massachusetts 02114
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
glutamine metabolism; mitochondrial proteins; gene expression; mRNA processing; CAG repeat polymorphisms; brain; kidney; muscle; mammary adenocarcinoma
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
To date, two isoforms of mitochondrial glutaminase have been characterized, liver type (LGA) and kidney type (KGA) (5). LGA is expressed only in periportal hepatocytes of the postnatal liver (9), whereas KGA has been reported to be abundant in the kidney, brain, intestine, fetal liver, lymphocytes, and tumors (5). The two isoenzymes have different structural and kinetic properties and are thought to be the products of different, but related, genes (5). KGA helps to maintain the acid-base balance of the host via ammoniagenesis in kidney tissue (4, 12). In brain tissue, KGA is instrumental in the production of glutamate and -aminobutyric acid (GABA) for neurotransmission (32). In intestinal epithelium, KGA is thought to initiate the catabolism of Gln, which serves as a major respiratory fuel source (31).
The present work was initiated to obtain a cDNA of the human KGA, in order to study its expression and its relevance to the growth and Gln utilization of human cancer cells. At the inception of this project, a nearly full-length cDNA and partial cDNA sequences of the KGA had been cloned only from rat and pig libraries, respectively (18, 23). [Recently, however, a full-length human KGA homolog was identified by random cloning as part of a large-scale human cDNA-sequencing project (15).] We used a rat KGA (rKGA) cDNA probe to screen a human cDNA library and obtain a partial cDNA clone of human KGA (hKGA). In addition, a partial cDNA clone containing an open reading frame corresponding to the rKGA with a different COOH-terminal sequence was identified and termed hGAC. Analysis of another cDNA clone (HSAAD20) isolated by Imbert and colleagues (8) revealed that this represents a third human glutaminase homolog, hGAM. These isoforms exhibited unique tissue-specific patterns of expression. Evidence is presented suggesting that these three hGA isoforms originate from a common gene. Therefore, expression of these human glutaminase isoforms is presumably controlled by tissue-specific alternative splicing of a common pre-mRNA. Furthermore, hGAC was found to be the predominant glutaminase splice form expressed by a human breast cancer cell line with an extraordinarily high rate of Gln utilization and glutaminase activity.
![]() |
MATERIALS AND METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Cloning and Analysis
A cDNA library in LambdaGEM-2 vector (Promega) constructed with poly(A)+ RNA isolated from the human colon carcinoma cell line HT-29 cultured on 2.5 mM inosine and harvested during log phase (30) was kindly provided by Dr. Burton Wice (Washington Univ. School of Medicine, St. Louis, MO) via Dr. Daniel K. Podolsky (Dept. of Gastroenterology, Massachusetts General Hospital). An rKGA probe, termed rGA, was generated by random priming of a 1.3-kb template isolated from pGA104 (see Materials) by Cla I-Acc I digestion. After screening the library, positive clones were plaque-purified, and cDNA inserts were separated from the phage fragments by Spe I or partial EcoR I-Xba I digestion and subcloned in pBluescript II-SK+/- (Stratagene). Subcloned inserts were sequenced in both directions by the dideoxynucleotide chain termination reaction method (20) using single-stranded DNA, specific primers, and the Sequenase version 2.0 sequencing kit (U.S. Biological, Swampscott, MA) according to the vendor's protocol. Single-stranded phagmid DNA was obtained by infecting a single bacterial colony carrying plasmid of interest with a helper phage VCS-M13 (Stratagene) and isolating single-stranded DNA from the phage particles according to the vendor's protocol. The sequence data were assembled with assistance from MacVector software (Oxford Molecular Group, Campbell, CA). cDNA sequence information was used to perform a computer-aided comparison to sequences deposited in GenBank, using BLASTN (DNA database comparisons).
Probe templates for analysis of hGA isoform expression were generated as follows: probes 111RR (0.5 kb) and 101RR (1.4 kb) were generated by EcoR I digestion of hGA111 cDNA and hGA101 cDNA, respectively; probe 111XX (0.5 kb) was generated by Xba I digestion of hGA111 cDNA; and probe 101NX (1.2 kb) was generated by Hinc II-Xba I digestion of hGA101 cDNA. Probe AD20HS (0.6 kb) was generated by Hind III-Sca I digestion of HSAAD20. Probe AD20EB (0.3 kb) was generated by EcoR I-BstE II digestion of HSAAD20.
Cloning of the 5' end of hGAC
The 5' end of hGAC was PCR cloned using 5',3'-rapid amplification of cDNA ends (RACE)-ready Marathon human heart cDNA library (Clontech, San Francisco, CA) and Advantage-GC cDNA PCR kit (Clontech). PCR was performed using hGAC-specific backward primer (hGA MP3) (5'-CGGGACTGAATTTGGCCAGTTGAGG-3') and forward adapter primer (AP1) (5'-CCATCCTAATACGACTCACTATAGGGC-3'). PCR was performed using 5 µl of human heart cDNA (0.5 ng), 1 µl of each backward and forward primer (10 µM each), 1 M GC-Melt (Clontech), 0.2 mM of each dNTP, and 1 µl of Advantage-GC cDNA polymerase mixture in a final volume of 50 µl. PCR was performed as follows: 95°C for 1 min, followed by 35 cycles of 94°C for 30 s and 66°C for 3 min, followed by a final extension step at 66°C for 5 min. A nested PCR reaction was performed using backward primer hGA MP2 (5'-ACCTTTCCTCCAGACTGCTTTTTAGC-3'), forward primer AP1, and 5 µl of a 1:50 dilution of the initial PCR reaction product in 10 mM Tricine-KOH (pH 9.2), 0.1 mM EDTA. PCR products ~1 kb in size from both the original and the nested reactions were cut with restriction enzymes Hinc II and Not I. These DNA restriction fragments were gel-purified and cloned into the pBluescript KS+ vector (Promega). Positive clones were isolated and sequenced.
Northern Blot Analysis
Northern blot analysis of human kidney RNA was performed as described previously (1), using 2 µg of poly(A) RNA from human kidney tissue (Clontech). cDNA fragments representing the homologous and nonhomologous regions of the cloned fragments were used as templates to generate 32P-labeled probes by random hexamer priming. To analyze the tissue-specific expression of hGA mRNA isoforms, a human tissue poly(A) RNA blot (Clontech) was prehybridized with ExpressHyb hybridization solution (Clontech) for 1 h at 65°C and then incubated with 32P-radiolabeled, random hexamer-primed cDNA probes for 2 h at 65°C. After each hybridization, membranes were washed at high stringency (0.1x SSPE, 1% SDS, 65°C, where 1x SSPE is 0.15 M NaCl, 0.01 M Na2HPO4, and 0.001 M EDTA) and exposed to X-ray film (Fuji Medical Systems, Stamford, CT) at -80°C for 824 h.
RT-PCR Analysis
Poly(A)+ RNA from human kidney tissue (Clontech) was used as substrate for RT-PCR reactions performed using a GeneAmp RNA PCR Kit (Perkin Elmer Cetus, Branchburg, NJ) according to the manufacturer's protocol. Reverse transcription reactions (10 µl) were prepared by mixing 0.025 µg of RNA, 25 pmol of oligo(dT), 1 µl of 10x PCR buffer (500 mM KCl, 100 mM Tris-HCl), 2 µl of 25 mM MgCl2, 1 µl of 10 mM dNTPs, 10 U of RNase inhibitor, diethyl pyrocarbonate-treated water, and 25 U of Moloney Murine leukemia virus reverse transcriptase. This mixture was incubated at room temperature for 10 min, at 42°C for 15 min, and at 99°C for 5 min and then put on ice. This reaction was used as a substrate for 30 cycles of PCR (2 min at 95°C followed by 30 cycles of 45 s at 95°C and 45 s at 55°C and a final extension step of 7 min at 72°C) in a 50-µl reaction with 1.25 U of Taq polymerase in 1x PCR buffer containing 0.2 mM dNTPs, 2 mM MgCl2, and 0.3 µM of each primer. Primers used for PCR were common forward primer F1 (5'-TGATGGCTGCGACACTGGCTAATG-3'), common nested forward primer F2 (5'-GGTCTCCTCCTCTGGATAAGATGG-3'), hGA111-specific backward primer B1 (5'-CCCGTTGTCAGAATCTCCTTGAGG-3'), and hGA101-specific backward primer B2 (5'-GATGTCCTCATTTGACTCAGGTGAC-3'). The PCR products were separated on 1% agarose gels.
RNase H Analysis
TSE cells were grown overnight to a subconfluent stage in 150-mm culture plates at 37°C under a humidified atmosphere of 5% CO2-95% air. Cells were maintained in DMEM supplemented with 4 mM L-Gln, 10% FBS, 100 U/ml penicillin, 100 µg/ml streptomycin, 0.25 µg/ml amphotericin B, and 10 µg/ml bovine insulin. Total RNA was extracted using RNA Stat-60 (Tel-Test, Friendswood, TX) as described previously (1). Ten micrograms of total RNA was preincubated with 0.5 µg of antisense primer (B2) at 70°C for 5 min in a mixture containing 4 µl of 2.5x Universal RiboClone second-strand buffer [Promega, 100 mM Tris-HCl, pH 7.2, 225 mM KCl, 7.5 mM MgCl2, 7.5 mM dithiothreitol, 0.125 mg/ml bovine serum albumin (BSA)] and diethyl pyrocarbonate-treated water (total vol 9 µl). RNase H (1.5 U; Promega) was added to the mixture, which was then incubated at 37°C for 20 min and put on ice. The reaction was analyzed by Northern blotting as described above.
Southern Analysis of Genomic DNA and Genomic Glutaminase Clones
Human genomic DNA was extracted from a culture of human renal proximal tubule epithelial cells (RPTEC). Briefly, cells were grown to near confluence, harvested and separated by centrifugation, and resuspended in 4.5 ml of DNA extraction buffer (10 mM Tris, pH 8.0, 100 mM EDTA) with 250 µl 10% SDS and 100 µl of 10 mg/ml proteinase K. The mixture was incubated at 55°C for 16 h. DNA was separated by phenol-chloroform extraction and ethanol precipitation. The pellet was then resuspended in TE (10 mM Tris, pH 8.0, 1 mM EDTA) overnight and treated with DNase-free RNase (Boeringer-Mannheim, 500 µg/ml) at 37°C for 1 h, phenol-chloroform extracted, and ethanol precipitated. After resuspension in TE, 5-µg samples of genomic DNA were digested overnight with 50 U of restriction endonuclease (Hind III, BamH I, EcoR I, Pst I, or Not I) in a total volume of 50 µl of appropriate restriction buffer, fractionated in a 0.8% agarose gel, and transferred to nylon membrane.
Five human glutaminase gene (gls) clones were identified and isolated from a bacterial artificial chromosome (BAC) human genomic DNA library by a commercial screening service (Genome Systems, St. Louis, MO) using probe 101RR. For Southern analysis, 5 µg of each clone was digested with EcoR I, fractionated in a 0.8% agarose gel, and transferred to nylon membrane. Southern blots were hybridized with differential probes, washed at high stringency (0.1x SSPE, 1% SDS, 65°C) and exposed to X-ray film, as described above for Northern blotting.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The 1,285-base sequence of hGA111 is presented in Fig. 1. The first 28 bases of this clone are not homologous to rKGA. After the first 28 bases, clone hGA111 exhibits 91% sequence homology with a portion of the coding sequence of rKGA [bases 11102080 of rKGA (23)] and then an 82% homology with a portion of the 3'-UTR of rKGA [bases 20812346 of rKGA (23)]. hGA111 contains a 22-base poly(A) tail beginning at base 1264. This tail is not substantially longer than the oligo(dT) primer used in library construction. However, a polyadenylation signal (AATAAA) that is absent in rKGA is located at bases 11801185 of hGA111. Clone hGA111 contained a 999-bp open reading frame (ORF). The translation of the ORF of hGA111 has complete homology with the 323 COOH-terminal amino acids of rKGA, excluding the first 9 NH2-terminal amino acids encoded by hGA111 (NVFMSLIFL) and three 3 acids within the hGA111 ORF (2 of which are not specified because of 2 unidentified nucleic acids). Thus hGA111 was deemed a partial cDNA clone of the human kidney-type glutaminase isoform, hKGA. This was confirmed by the isolation and sequencing of a full-length hKGA cDNA by Nagase and co-workers (15). Bases 12872522 of this 4,198-bp hKGA cDNA (clone hk03864, GenBank accession no. AB020645) share complete homology with bases 291269 of hGA111, except for the two unidentified nucleic acids in the hGA111 sequence.
|
|
The complete sequence of hGA101 is presented in Fig. 2 (bases 3473549 of hGAC). hGA101 has a very high homology with the 5' end of the rKGA but diverges completely at the 3' end. The 5' end of the mRNA corresponding to hGA101 was PCR cloned from a human heart 5'-,3'-RACE-ready cDNA library using hGA101-specific primers (complementary to bases 9951019 and 996981 in hGAC; Fig. 2). An 879-bp cDNA fragment was cloned and found to overlap with the 5' 533 bases of hGA101. The 3' portion of hGA101 sequence, which is not homologous to rKGA, was subjected to a BLASTN DNA database search, which identified an EST clone, dbEST247733 (GenBank accession no. R70875). This EST clone was purified from a library constructed from poly(A)+ RNA isolated from full-term placental tissue (Soare's placenta Nb2HP; The WashU-Merck EST Project, L. Hillier et al., 1995, unpublished). dbEST247733 was obtained and fully sequenced. The 5' 938 bases of this 1,847-bp cDNA overlap the 3' end of hGA101 with near-perfect homology (1 base deletion compared to hGA101). The combined sequence (4,438 bases) of these three clones is presented in Fig. 2. The hGA mRNA species represented by these three partial cDNA clones was termed hGAC.
|
Imbert and co-workers (8) reported the isolation and partial sequence of a cDNA clone, HSAAD20, which exhibited partial homology to the rKGA (EMBL accession no. Y08264). This cDNA was one of several clones containing CAG repeats isolated from a spinocerebellar ataxia patient lymphoblastoid cell line library (8). The 2,010 bases of HSAAD20 were fully sequenced (GenBank accession no. AF097495) and compared with hKGA and hGAC. Bases 1696 are completely homologous to hGAC except for the presence of 14 CAG repeats in HSAAD20 (nucleotides 546). At this position, the 5'-UTR of hGAC (obtained from the RACE-ready human heart cDNA) contains only seven GAC repeats. Beyond base 696, HSAAD20 contains a stretch of 356 nucleotides that shares no homology with either hGAC, hKGA, or rKGA. A BLASTN DNA database search confirmed that this stretch is a unique sequence. Bases 10531173 of HSAAD20 are again completely homologous to hGAC, but the final 836 bases of HSAAD20 share no homology with hGAC or hKGA. Thus HSAAD20 may represent an hGA mRNA species that shares a 5' end with hGAC but contains a unique internal exon and a unique 3' end. HSAAD20 does not contain a complete 3' end, for this clone was isolated from a random hexamer-primed cDNA library (8). HSAAD20 contains two polyadenylation signals, but these are distal to the 3' end (nucleotides 741746 and 973978). Thus HSAAD20 was considered a partial clone of another human glutaminase isoform and termed hGAM. hGAM contains an ORF from base 213 to 719. This ORF encodes a 169-amino acid peptide identical to that encoded by hGAC up to amino acid 161, at which point the hGAM peptide diverges and encodes a unique eight-amino acid COOH terminus (Fig. 3). As the deduced NH2-terminal amino acids of hGAM are the same as those encoded by hGAC, this protein also contains a potential mitochondrial signal peptide.
Figure 4 schematically represents the relationships between rKGA, hKGA, hGAC, and hGAM cDNAs, as well as the probe templates and oligonucleotides used to further analyze these species.
|
|
Human Tissue Distribution of hGA Isoforms
The tissue distribution of the hKGA, hGAC, and hGAM mRNA transcripts was determined by probing a human tissue poly(A)-enriched RNA blot with differential cDNA probes. As shown in Fig. 6, 111RR (containing sequences common to hKGA and hGAC) detected a 4.8-kb band in all but liver tissue and a 3.5-kb band in kidney tissue. The intensity of the 4.8-kb band was relatively high in brain, heart, kidney, and pancreas. 111XX (specific for hKGA) also detected a 4.8-kb band in all but liver tissue, with a high intensity in brain and kidney and a weak signal in other tissues. The tissue distribution observed with probe 111XX is nearly identical to that determined by Nagase and co-workers (15) by RT-PCR with oligonucleotide primers specific to clone hk03864. Probe 111XX also detected a 3.5-kb band in kidney and a slightly smaller species (~3.4 kb) in pancreatic tissue. Probe 101NX (specific to hGAC) clearly detected a 4.8-kb band in all but brain and liver tissue. This signal was strongest in heart and pancreas, followed by placenta, kidney, and lung. Thus, in contrast to 111XX, 101NX did not detect the 4.8-kb mRNA species present in brain and produced a relatively weak signal with the kidney sample. Probe 101NX also weakly detected a 3.5-kb mRNA species in kidney tissue. By probing a fresh blot, it was determined that this band was not a ghost left from hybridization with probe 111NX. Unlike both hKGA- and hGAC-specific probes, AD20HS (specific for hGAM) detected mRNA species in cardiac and skeletal muscle only, which were single bands of 2.6-kb size. These results suggest that three nonhepatic isoforms of hGA exist and are expressed in unique tissue-specific fashions. hKGA is expressed predominantly in brain and kidney. hGAC is expressed primarily in cardiac muscle and pancreas but also appreciably in placenta, kidney, and lung. hGAM is expressed exclusively in cardiac and skeletal muscle tissues.
|
Analysis of hGA Isoform Expression in a Human Breast Carcinoma Cell Line
The human breast carcinoma cell line TSE was previously found to exhibit a high rate of Gln utilization and an extraordinarily high glutaminase activity (2). It was previously observed that this cell line exhibits a relatively high hGA mRNA content of exclusively 4.8-kb size (Abcouwer, unpublished observation). In light of the identification of multiple isoforms of hGA, it was of interest to determine which isoform was expressed by this cell line. Oligonucleotide-targeted RNAse H digestion of total RNA from subconfluent TSE cells was utilized to determine definitively which isoform of hGA was expressed by these tumor cells. To differentiate the two possible 4.8-kb mRNA species, total RNA was treated with RNase H in the presence of an antisense oligonucleotide (B2) specific for the nonhomologous coding region of hGAC. RNase H has endoribonuclease activity specific for RNA-DNA duplexes and therefore should cleave only that RNA hybridized to B2. The result of this targeted digestion was analyzed by Northern blotting using differential probes. The mRNA band corresponding to hGAC was expected to shift to a lower molecular weight, separating it from the hKGA band. The Northern blots were hybridized with the differential cDNA probes for hGA111 and hGAC (Fig. 7). Ethidium bromide staining showed intact ribosomal RNA bands, indicating the absence of any gross nonspecific digestion of RNA. Both 101RR and 101NX probes detected one 4.8-kb band in the control sample, whereas the same probes detected only 2-kb and 3-kb bands, respectively, in the digested sample. No signal was obtained when this blot was probed with the hKGA-specific probe, 111XX (data not shown). This result confirms the existence of an mRNA transcript corresponding to hGAC and also suggests that this is the predominant isoform expressed by TSE cell line. However, the RT-PCR assay (as described in Expression of hGA Isoform in Human Kidney Tissue) established that an amplifiable amount of hKGA mRNA was present in TSE RNA (data not shown).
|
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Using Southern blotting analysis of human x hamster somatic cell hybrids, the gls gene was previously mapped to human chromosome 2 (13, 15). Human gls was further mapped to the region 2q322q34 by in situ hybridization (14). These studies utilized cDNA probes that corresponded to hKGA or rKGA. In the present study, Southern analysis with differential probes indicated that hKGA, hGAC, and hGAM isoforms originate from a single genetic locus contained within 14-kb BamH I and 16.5-kb Not I restriction fragments. In addition, these three hGA mRNA isoforms share appreciable regions of exact sequence homology, suggesting that they contain common exons. Thus it is most likely that these isoforms are produced by alternative splicing of a common pre-mRNA species. This was confirmed by cloning and Southern analysis of the gls BAC clones, which again indicated that exons of the three hGA mRNA isoforms are present in the same genomic locus. However, these BAC clones have not yet been mapped to a chromosomal location. Nonetheless, we propose that human glutaminase represents an example of regulated expression by tissue-specific alternative splicing. This splicing results in the generation of coding sequences with different COOH-terminal amino acid sequences. The implications of these differing COOH termini on the activity and function of these glutaminase enzymes are not yet known.
The 5'-UTRs of hKGA, hGAC, and hGAM contain CAG repeats at the same location but of different lengths (15 repeats in hKGA, 7 repeats in hGAC, and 14 repeats in hGAM). Assuming that these sequences originate from the same exon, these three cDNAs demonstrate a trimorphism in the length of this CAG repeat. CAG repeat-length polymorphisms are associated with several genetic diseases, including myotonic dystrophy and Huntington's (25). Whether the CAG repeat found in the 5'-UTRs of hGA isoforms has an impact on hGA expression is yet to be evaluated. There is an apparent interspecies difference in the location of CAG repeats in GA mRNAs. A CAG repeat is present within the coding sequence of rKGA (nucleotides 329355), and a BLASTN search of mouse dbEST database identified a cDNA clone similar to rKGA (dbEST2110859, GenBank accession no. AI327299; The WashU-HHMI Mouse EST Project, M. Marra et al., 1996, unpublished), which has a CAG repeat at the same location as the rat clone.
The NH2-terminal amino acid sequence of the rKGA contains a 16-residue mitochondrial signaling sequence that directs the proenzyme to the mitochondria (23, 26, 27). The NH2 terminus of the deduced amino acid sequence of hKGA, hGAC, and hGAM contains a putative mitochondrial signaling sequence that nearly matches that of the rat kidney isoform, suggesting that these human isoforms also encode mitochondrial proteins.
The expression of hGA mRNA isoforms was analyzed in a TSE human breast carcinoma, which was chosen because it exhibits a high rate of Gln consumption (2) and expresses an extraordinarily large amount of glutaminase mRNA and activity (2). In fact, the glutaminase activity measure in these cells (44.8 µmol · mg protein-1 · h-1) is ~4 times higher than that reported in human intestinal epithelial cells(22), 18 times higher than human fibroblasts (21), 19 times higher than rat brain (16), and 12 times higher than rat kidney (3). Moreover, hGAC was found to be the predominant isoform expressed by TSE cells. The fact that these cells exhibit a high glutaminase activity in the absence of detectable hKGA expression strongly suggests that hGAC encodes a functional enzyme. Expression of hGAC by TSE cells should not be interpreted to indicate that this isoform is exclusively associated with neoplastic transformation or with breast tissue, for Northern analysis of several human breast cell lines demonstrated that hGAC expression varied greatly (S. F. Abcouwer, unpublished observation).
In contrast, no data regarding functionality of hGAM can be offered at this time. It is probable that the small peptide encoded by hGAM does not retain glutaminase activity. hGAM encodes only 169 amino acids, 72 of which represent a putative targeting sequence removed during mitochondrial processing. The predicted size of the remaining peptide (<11 kDa) is smaller than any known glutaminase, including the glutaminase subunit of bacterial amidotransferase (33). Because heart and muscle tissue contain relatively low glutaminase activity, it is conceivable that splicing of the primary glutaminase transcript to form hGAM mRNA represents a means to reduce expression of functional glutaminase activity in these tissues. Such a mechanism has been hypothesized to explain the discrepancy between glucokinase mRNA expression and lack of glucokinase activity in the pituitary (11).
By comparison of brain and kidney hybridization signals using probes 111RR and 111XX, it was estimated that hKGA is approximately six times more abundant than hGAC in kidney (data not shown). Similarly, it was estimated that hGAC is slightly more abundant than hKGA in placenta, lung, cardiac, and skeletal muscle. Expression of these isoforms in normal intestine has not yet been tested. The levels of both hKGA and hGAC mRNA in the HT-29 human colon carcinoma cell line (from which the cDNA library was obtained) were found to be very low (data not shown). Although mRNA quantity is not a direct indicator of protein expression or activity, the results suggest that the expression of the kidney-type glutaminase isoform is indeed dominant in kidney as well as brain. However, it is now apparent that the expression of glutaminase-type C may be considerable in several other tissues. Differential antisera are needed to test this presumption. No information regarding the differential functions of these glutaminase isoforms has yet been obtained, and very little data on the properties of glutaminase enzyme in tissues other than kidney, brain, and liver are available. Therefore, the reason for tissue-specific expression of different glutaminase isoforms remains to be determined.
No analysis of the developmental, hormonal, or environmental regulation of hKGA, hGAC, or hGAM has been presented. In the rat, glutaminase expression is upregulated in kidney in response to acidosis (28). This response has been attributed to an increase in rKGA mRNA stability under acidic conditions that is orchestrated by several AU-rich, pH-responsive instability elements in its 3'-UTR (7). The 3'-UTR of hKGA does not contain any of these elements. The 3'-UTR of hGAC does contain two eight-base sequences (UUUAAAUA) that match the first half of the pH-responsive instability element and two other eight-base sequences (UUAAAAUA) that match the second half of this element (10). Although the results of Laterza and colleagues (10) suggested that an intact 16-base element is needed for full function of this element, the presence of these half-elements raises the possibility that the stability of hGAC mRNA is regulated at the posttranscriptional level by a similar pH-dependent mechanism. However, neither hKGA nor hGAC mRNA levels responded when RPTEC or TSE cells were subjected to acidotic conditions (K. M. Elgadi, unpublished results).
A recent report by Roberg and colleagues (19) claimed that distinct soluble and membrane-bound forms of phosphate-activated glutaminase are present in rat and pig kidney. It is conceivable that these isozymes are the products of KGA and GAC mRNAs. However, reports from the same laboratory, as well as others, also described the presence of distinct soluble and membrane-bound forms of phosphate-activated glutaminase in pig brain (17, 29). If, like human brain tissue, pig brain does not express glutaminase type C, then the origin of these distinct forms of glutaminase enzyme cannot be attributed to the two mRNA isoforms. Instead, soluble and membrane-bound forms of glutaminase may be due to altered mitochondrial processing or represent monomeric and polymeric forms of the same protein (5).
In summary, this report describes a cDNA clone representing the human kidney-type glutaminase isoform, hKGA, and two novel human glutaminase isoforms, hGAC and hGAM. The existence of hGAC mRNA was not suspected because it shares common sizes with the kidney-type mRNA species. This seems to be true for both human and rat isoforms (data for the rat not shown). The existence of hGAM mRNA was not suspected because glutaminase mRNA expression in muscle has not been previously published. Differential probes and a differential RT-PCR assay were used to confirm the existence of the additional isoforms and examine their expression. The RT-PCR assay used represents a particularly convenient method for analysis of hKGA and hGAC mRNA expression. The three nonhepatic glutaminase mRNA isoforms exhibit unique tissue-specific expression patterns. The functionality of one of these novel isoforms, type-C glutaminase, was evidenced by its predominant expression in TSE cells. However, comparison of the enzymatic properties and subcellular localization of the separate isoforms will require further expression studies and the development of isozyme-specific antibodies. It would be of interest to determine whether expression of certain glutaminase isoforms occurs during normal development or is associated with pathological conditions, such as malignant transformation.
![]() |
ACKNOWLEDGMENTS |
---|
This work was supported by National Cancer Institute Grant R29-CA-72772 to S. F. Abcouwer and by the Clinical Nutrition Research Center at Harvard University (Pilot Feasibility Grant P30-DK-40561 to S. F. Abcouwer).
The complete nucleotide sequences of hGAC, hGA101, hGA111, dbEST247733, and hGAM have been deposited in the GenBank database under accession nos. AF158555, AF097492, AF097493, AF097494, and AF097495, respectively.
Address for reprint requests and other correspondence (as of Oct 1, 1999): S. F. Abcouwer, Dept. of Biochemistry and Molecular Biology, The Univ. of New Mexico Health Science Ctr., School of Medicine, Basic Medical Sciences Bldg.-249, Albuquerque, NM 87131-5221 (E-mail: SFAbcouwer{at}salud.unm.edu).
![]() |
FOOTNOTES |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Visit Other APS Journals Online |