INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

The vacuolar-type H⁺-ATPase (V-ATPase)¹ is a ubiquitous proton pump found in a variety of intracellular organelles including the endosome, lysosome, Golgi apparatus, and clathrin-coated vesicle (1-3). The V-ATPase-mediated acidification of organellar lumens is critical for protein sorting, zymogen activation, and receptor-mediated endocytosis (4-7). In addition, V-ATPase is found in the plasma membranes of kidney and bladder where the enzyme has a key role in pH regulation (8, 9) of osteoclasts for bone resorption (10) and of seminal duct for spermatogenesis (11).

V-ATPase has two distinct sectors called V₁ and V₀ which are analogous to F₁ and F₀ of F-type ATPases (12), respectively. V₁ is the peripheral membrane sector and is composed of eight subunits. A and B subunits form the catalytic complex, and six other subunits (C-H) link the A-B hexamer complex to V₀ (13). V₀ is the integral membrane sector making up the proton pathway and consists of the proteolipid (16 and 23 kDa) and 100- and 36-kDa subunits (13).

A hydrophobic 16-kDa protein called the proteolipid is a major component of the V₀ sector and has a N,N'-dicyclohexylcarbodiimide-reactive glutamate that is essential for proton transport (14, 15). The cDNA clones for proteolipids have been identified from a number of species. Three proteolipid genes are independently required for yeast V-ATPase function (14-17): VMA3 and VMA11, both of which code for 16-kDa proteolipids, and VMA16, which codes for a 23-kDa isoform. Higher eukaryotic cells also have multiple proteolipid genes. Previously, we reported that Caenorhabditis elegans has three genes (vha-1, vha-2, and vha-4; Ref. 18). At this time, the number of proteolipid genes in mammalian cells is unclear. Although four genomic clones were isolated from a human library, only one corresponded to the cDNA and the other three clones appeared to be of pseudogenes (19).

In this study, we demonstrate that a third 16-kDa proteolipid gene (vha-3) is present in the C. elegans genome. The vha-3 and vha-2 genes exhibit strong similarity and encode identical polypeptides but are expressed differently in a cell-specific manner. Another newly described gene, vha-11, is located just downstream of vha-3 and codes for a hydrophilic polypeptide very similar to the V-ATPase C subunits of yeast (20, 21), bovine (22), and human (23). Expression of the vha-11 cDNA restores growth of the yeast vma5 mutant that lacks the C subunit gene (20). The results demonstrate that the vha-11 gene product is the C subunit.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Culture and Transformation of Worms-- Wild-type strain Bristol N2 was cultured as described (24). Transformation was carried out using the selectable marker plasmid pRF4 (25).

Sequencing cDNA Clones-- The C. elegans expressed sequence tag (EST) clones, yk413h1 (vha-3) and yk434g8 (vha-11), were kindly provided by Y. Kohara and sequenced using the Dye Terminator DNA sequencing kit (Applied Biosystems). The 5' regions were amplified from total RNA taken from a mixed population of worms by RT-PCR using spliced leader primers and gene-specific primers (vha-3, 5'-gcacttgccgacgtcacttt-3' and vha-11, 5'-caccttcccattggaatttcg-3') as described (18). The resulting PCR products were sequenced and ligated with the EST cDNAs to construct full-length cDNA clones. The nucleotide sequence data reported in this paper will appear in the DDBJ, EBI, and GenBank^TM nucleotide sequence data bases with the following accession numbers: vha-3, AB009566; and vha-11, AB009567.

Northern Blot Analysis-- Synchronously growing worms were prepared using the alkaline hypochlorite method (26). Total RNA was prepared from C. elegans of mixed growth stages using TriZOL^TM LS Reagent (Life Technologies, Inc.) according to the recommendations of the manufacturer. 15 µg of total RNA was electrophoresed on a 1.5% agarose, 6% formaldehyde gel and transferred to a Hybond-N+ membrane (Amersham Pharmacia Biotech). Hybridizations were carried out using probes labeled by the random primed DNA labeling kit (Boehringer Mannheim) and QuikHyb solution (Stratagene) according to the recommendations of the manufacturer. The blot was also hybridized with a probe from the ribosomal protein gene rp21 as a loading control (27). The filter was scanned by the Image-Analyzer BAS-1000 (Fuji Film) for semi-quantitative estimation of transcript levels.

Genomic Southern Blots-- Genomic DNA was prepared from a mixed population of worms and subjected to Southern blot analysis. For low stringency conditions, the filters were washed at room temperature for 10 min with 5× SSC containing 0.1% SDS, followed by a further wash at 60 °C for 10 min (27).

Constructions of lacZ Reporter Plasmids-- The upstream region of vha-3 was amplified directly from C. elegans genomic DNA by PCR using primers (5'-gcgacaggtactgcaggatattg-3' and 5'-gcggtttccaagtcgtcgacattt-3'). The maximum amount of upstream region was used but without introducing the regulatory regions of neighboring genes. The 2.2-kb product was digested with PstI and SalI and inserted into pPD89.03, a lacZ reporter plasmid, to create pCV3-03. pCV02 was constructed by inserting the 1.3-kb BamHI to SalI fragment, which contained the regulatory region of vha-1 and vha-2 (18), into pPD89.03. Transgenic worms were fixed and stained for -galactosidase using 5-bromo-4-chloro-3-indolyl--D-galactoside (28). Identification of cell types has been previously described using Nomarski optics (18, 29).

Rescue of Yeast vma5 Null Mutant with C. elegans vha-11-- vha-11 cDNA was ligated downstream of the TDH3 promoter of pKT10, a yeast expression vector (30). The resulting plasmid was introduced into the yeast vma5 mutant SF838-1D (MAT leu2-3,112 ura3-52 his4-519 ade6 gal2 pep4-3 vma5::LEU2, Ref. 20). Transformants were streaked on YPD plates (2% peptone, 1% yeast extract, and 2% glucose), including either 50 mM sodium succinate, pH 5.0, or 50 mM potassium phosphate, pH 7.5, and were incubated at 30 °C for 3 days to score the growth phenotype.

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

A Third V-ATPase 16-kDa Proteolipid Gene, vha-3, in C. elegans-- Previously, we identified two proteolipid genes, vha-1 and vha-2, in C. elegans which had 60% amino acid identity (18). We searched for additional genes using RT-PCR. One PCR product was highly similar to the vha-2 gene, and analysis of the C. elegans genome data base indicated that the PCR product matched the Y38F2 YAC genomic clone from chromosome IV. We named the corresponding gene vha-3 and searched the EST data base for a full-length clone. We found only one clone (yk413h1) that contained a part of the sequence of the PCR product.

View larger version (52K): [in this window] [in a new window]   Fig. 1.   Nucleotide alignment of vha-2 and vha-3 and gene structure of vha-3 and vha-11. A, nucleotide alignment of vha-2 and vha-3, which code for identical 16-kDa proteolipids. Identical residues are indicated by boxes. Amino acid residues are shown by one-letter symbols. B, gene structure of vha-3 and vha-11. Closed and shaded boxes represent coding and untranslated regions, respectively. The identified spliced leaders (SL1 or SL2) are indicated upstream of the first exon of each gene. The 2.2-kb fragment containing the 5' upstream regulatory region was fused with a lacZ reporter gene in pCV3-03 for testing expression pattern.

The V-ATPase C Subunit Gene, vha-11, of C. elegans Is Linked to vha-3-- Sequence comparison with the Y38F2 YAC genomic clone revealed that clone yk434g8 mapped immediate downstream of vha-3. An open reading frame was found in yk434g8 that shared high sequence identity with the V-ATPase C subunits of other species. We named the gene vha-11 and obtained the entire vha-11 cDNA by RT-PCR. Comparison with genomic DNA and cDNA sequences indicated that the vha-11 gene consisted of seven exons (Fig. 1A). The 1639-bp cDNA contained a coding region for a 384-amino acid polypeptide without any putative transmembrane regions. The VHA-11 protein exhibits 37%, 56%, and 56% sequence identity with the C subunits of yeast (Vma5p) (20, 21), bovine (22), and human (23), respectively (Fig. 2A).

View larger version (64K): [in this window] [in a new window]   Fig. 2.   Alignment of V-ATPase C subunit amino acid sequences and rescue of yeast vma5 by C. elegans vha-11. A, alignment of the V-ATPase C subunits from nematode, yeast (Vma5p) (20, 21), bovine (22), and human (23). The deduced amino acid sequences were aligned for maximal homology. Conserved residues among at least three species are indicated by boxes. B, rescue of the yeast vma5 mutation by vha-11. Yeast cells harboring vector with no insert (Vector), the vha-11 expression plasmid (vha-11), or the VMA5 gene on a single-copy plasmid (VMA5) were streaked on either YPD, pH 5.0, or YPD, pH 7.5, and incubated at 30 °C for 3 days before phenotypic scoring.

Identification of vha-3 and vha-11 Transcripts-- To confirm that vha-3 and vha-11 genes are expressed, Northern blot analysis was performed using RNA from a mixed-stage culture. The vha-3 probe hybridized with three transcripts (0.9, 1.0, and 1.4 kb; see Fig. 3A). The 1.4-kb transcript corresponded to vha-3 because it was consistent with the length of the cDNA. Two weaker bands (0.9 and 1.0 kb) showed the same mobility as vha-2 transcripts and are likely because of cross-hybridization. The amount of the vha-3 transcript was roughly five percent of vha-1 or vha-2. A single band (1.5 kb) was detected for the vha-11 transcript whose size was compatible with that of the cDNA (Fig. 3B).

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Northern blot analysis of vha-3 and vha-11 transcripts. Total RNA from a mixed stage population was electrophoresed, blotted, and hybridized to probes from the coding regions of each gene. Transcript sizes were estimated using RNA standards run on the same gel. A, arrowheads show positions of the vha-3 (closed) and vha-2 (open) transcripts. B, arrowhead represents the vha-11 transcript.

Presence of Four vha Proteolipid Genes in C. elegans-- To search for other C. elegans proteolipid genes, Southern blot analysis was carried out under low stringency conditions. The vha-1 probe hybridized with single bands, the sizes of which were consistent with the length predicted from the genomic sequence (Fig. 4). No other bands were detected even under low stringency conditions, thus indicating that vha-1 is a single gene and no other homologue exists in this organism.

View larger version (62K): [in this window] [in a new window]   Fig. 4.   Genomic Southern blot analysis of the four vha genes. C. elegans genomic DNA (15 µg) was digested with EcoRI (lane 1), EcoRV (lane 2), HincII (lane 3), or PvuII (lane 4), subjected to electrophoresis, blotted to a filter, and hybridized with probes from the coding regions of each gene. Blots were washed under the low stringency condition as detailed under "Experimental Procedures." Arrowheads represent hybridization with indicated genes. DNA standards were stained with ethidium bromide, and their sizes are shown in kilobase pairs.

Expression of Proteolipid Genes during the Worm Life Cycle-- The obvious but interesting question is what distinguishes vha-2 and vha-3 since they code for the identical protein. We first assessed the expression of each gene as a function of life cycle. Northern blot analysis was carried out using RNA from different developmental stages. All four proteolipid genes were highly expressed in the embryo and much lower during larval stages (Fig. 5). The results suggest that the V-ATPase activities may be important during embryogenesis and cell proliferation (32). The expression of all genes gradually increased again leading up to the adult stage. When in the dauer stage, an obligate stage of diapause, all four transcripts were present but at approximately 2-fold higher levels than the L2 or L3 stage as judged from quantitation of the blot scans.

View larger version (83K): [in this window] [in a new window]   Fig. 5.   Stage-specific Northern blot analysis of the four vha genes. Total RNA (15 µg) from each synchronous population of worms was subjected to electrophoresis, blotted to a filter, and hybridized to probes from the coding regions of the indicated genes. The blot was also hybridized with a probe from a ribosomal protein gene (rp21) cDNA as a loading control. Lanes represent egg (E); larvae stages 1 (L1), 2 (L2), 3 (L3), and 4 (L4); adult (A); dauer (D); and mixed population (M).

Preferential Expression of vha-3 in Gastric and Hypodermal Cells-- Although V-ATPase genes are ubiquitously expressed, vha-1, vha-2, and vha-4 were expressed at much higher levels in the H-shaped excretory cell (18). To determine the expression pattern of the vha-3 gene, a fusion gene was constructed by inserting a genomic fragment containing the vha-3 5' transcriptional regulatory sequences through to the first two codons of the reading frame into a lacZ expression vector (Fig. 1B). Surprisingly, expression of the vha-3 fusion gene was significantly different in the adult worm from the other 16-kDa proteolipid genes. vha-1 and vha-2 were predominantly expressed in the H-shaped excretory cell and rectum (Fig. 6C), which confirmed previous results obtained with GFP (green fluorescent protein) fusion genes (18). In contrast, vha-3 was mainly expressed in gastrointestinal and hypodermal cells, in addition to the H-shaped excretory cell (Fig. 6, A and B). After prolonged staining, the canals of the H-shaped excretory cell also became visible (arrowheads). Similar results were obtained in transgenic worms carrying the reporter gene fused to the 5' upstream region of vha-11 (data not shown), supporting that vha-3 and vha-11 are transcribed as a polycistronic unit. These results demonstrate that the transcriptional regulation of vha-3 is different from that for vha-1 and vha-2.

View larger version (80K): [in this window] [in a new window]   Fig. 6.   Expression of vha-3 in gastric and intestinal cells. Transgenic worms carrying different expression plasmids were permeabilized and stained with 5-bromo-4-chloro-3-indolyl--D-galactoside (28). A, a transgenic worm harboring pCV3-03, the lacZ reporter gene under control of the 2.2-kb upstream regulatory region of vha-3. Staining is seen in the H-shaped excretory cell, intestine, and hypodermis. B, a shorter staining of the transgenic worm carrying pCV3-03. The staining of cells in the head and tail were clearly detectable. C, an adult worm carrying pCV02, the lacZ gene under control of the upstream regulatory region of vha-1 and vha-2. The lacZ fusion gene was expressed in the H-shaped excretory cell and rectum as reported previously with GFP constructs (18). Dark-stained spots correspond to cell nuclei. Nuclei and canals of the H-shaped excretory cell are indicated by arrows and arrowheads, respectively.

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

We have identified two genes (vha-3 and vha-11) coding for V-ATPase subunits of C. elegans. The 1421-bp vha-3 cDNA encodes the 16-kDa proteolipid. This is the fourth gene coding for isoforms of the proteolipid. vha-11 consists of seven exons and codes for a 384-amino acid hydrophilic polypeptide. The VHA-11 protein exhibited significant similarities to the V₁ sector C subunits of yeast (20, 21), bovine (22), and human (23), and the vha-11 cDNA complemented a yeast vma5 (C subunit) mutation, thus indicating that vha-11 is a functional counterpart of VMA5.

The vha-3 and vha-11 genes are tandemly located on chromosome IV as a single polycistronic unit. Transgenic analysis showed that vha-3 was most strongly expressed in intestine, hypodermis, and the H-shaped excretory cell (Fig. 6). Parallel expression patterns of a reporter gene fused to the 5' upstream region of vha-11 support that vha-3 and vha-11 are transcribed as a polycistronic unit. As shown previously, vha-1 and vha-2, which are two other genes that code for the 16-kDa proteolipid, are also transcribed as a unit (18). Furthermore, the first gene of both units (vha-1 and vha-3) were found to have unusually long 3'-untranslated regions and no introns. In contrast to the vha-1/vha-2 unit, the vha-3/vha-11 consists of genes for one V₀ and one peripheral V₁ subunit. Genes for other V-ATPase subunits, B, D, E, F, G, H, a and d, have been identified from the C. elegans genome project data base (ACeDB WS2 4-16) based on sequence similarities to known yeast subunits (13); however, none of these are in a polycistronic unit. Thus the transcriptional units of vha-3/vha-11 and vha-1/vha-2 are unique among V-ATPase genes in C. elegans. Finally, we note that vha-3/vha-11 and vha-1/vha-2 gene pairs are functionally related. Most C. elegans gene clusters consist of unrelated genes, although lin-15A/lin-15B and fib-1/rps-16 pairs are clearly functionally related (31, 33).

Southern blot analysis of the C. elegans genome showed that additional homologues of the proteolipid genes are not present, therefore suggesting that only four proteolipid genes (vha-1, vha-2, and vha-3, 16-kDa subunit; vha-4, 23 kDa) are present. This is the first report of four proteolipid genes in the same genome. Among these genes, vha-3 and vha-2 share very high nucleotide similarity within the coding region and encode identical polypeptides. Although 38% of the codons used for vha-3 and vha-2 are different, the codon usage is not significantly different from other C. elegans open reading frames (34). Two cDNAs for very similar proteolipids (16 kDa) were also isolated from cotton and shared 95% identity within the coding region (35). The functions of the two similar genes now found in plants and animals remains to be understood. We should note our attempts to isolate the knock-out mutants of vha-2 and vha-3 by the target-selected mutagenesis using transposon Tc1 (36). This procedure is known to delete more than a few kilobase pairs of DNA from the Tc1 target site. As expected, the null mutations were accompanied by large deletions of flanking genes. It is very difficult to isolate the desired knock-out mutants, possibly because the lengths of the two vha coding regions are less than 600 bp.

We might presume that vha-2 and vha-3 are transcribed differently during the worm life cycle or have different transcriptional regulation in a cell-specific manner; however, Northern analysis of worms at each developmental stage indicated that vha-2 and vha-3 are transcribed similarly. Instead, we assessed the transcription of vha-2 and vha-3 in different cell types. Analysis of adult worms carrying the lacZ gene under control of the vha-3 regulatory sequence indicated that vha-3 was strongly expressed in the gastrointestinal and hypodermal cells, as well as the H-shaped excretory cell. Similar results were obtained using the GFP reporter plasmid. In contrast, vha-2 was highly expressed in the H-shaped excretory cell and rectum (Fig. 6C and Ref. 18) and was not detected in gastrointestinal and hypodermal cells even after prolonged staining. These results strongly suggest that expression of the two vha genes are differentially regulated according to cell type.

From the above results, we suggest there is a subunit isoform difference in the V₀ sectors in different cell types. Similarly, cell-specific expression of mammalian V₁ subunits is also known. There are brain and kidney forms of the B subunit that have been identified in human and bovine. These B subunits share 84% amino acid identity (37, 38). The possibility that different isoforms define functional differences of the V-ATPases is under investigation.