|
Advertisement | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
J. Biol. Chem., Vol. 281, Issue 37, 27052-27062, September 15, 2006
Analysis of Strains with Mutations in Six Genes Encoding Subunits of the V-ATPaseEUKARYOTES DIFFER IN THE COMPOSITION OF THE V0 SECTOR OF THE ENZYME*From the Department of Molecular, Cell, and Developmental Biology, Sinsheimer Laboratories, University of California, Santa Cruz, California 95064
Received for publication, April 24, 2006 , and in revised form, June 26, 2006.
To address questions about the structure of the vacuolar ATPase, we have generated mutant strains of Neurospora crassa defective in six subunits, C, H, a, c, c', and c''. Except for strains lacking subunit c', the mutant strains were indistinguishable from each other in most phenotypic characteristics. They did not accumulate arginine in the vacuoles, grew poorly at pH 5.8 with altered morphology, and failed to grow at alkaline pH. Consistent with findings from Saccharomyces cerevisiae, the data indicate that subunits C and H are essential for generation of a functional enzyme. Unlike S. cerevisiae, N. crassa has a single isoform of the a subunit. Analysis of other fungal genomes indicates that only the budding yeasts have a two-gene family for subunit a. It has been unclear whether subunit c', a small proteolipid, is a component of all V-ATPases. Our data suggest that this subunit is present in all fungi, but not in other organisms. Mutation or deletion of the N. crassa gene encoding subunit c' did not completely eliminate V-ATPase function. Unlike other V-ATPase null strains, they grew, although slowly, at alkaline pH, were able to form conidia (asexual spores), and were inhibited by concanamycin, a specific inhibitor of the V-ATPase. The phenotypic character in which strains differed was the ability to go through the sexual cycle to generate mature spores and viable mutant progeny. Strains lacking the integral membrane subunits a, c, c', and c'' had more severe defects than strains lacking subunits C or H.
Vacuolar ATPases (V-ATPases)2 are ATP-driven proton pumps that acidify intracellular compartments in eukaryotic cells, including the vacuoles of fungi and plants, lysosomes, endosomes, the Golgi, and secretory vesicles. Acidification of these organelles is important for numerous cellular processes, including receptor-mediated endocytosis, zymogen activation, intracellular trafficking, protein degradation, coupled transport of small molecules, and viral entry (13). V-ATPases are also located at the plasma membrane in a subset of cells in which they function in processes such as renal acidification, bone resorption, and pH homeostasis (46).
Much of what is known about V-ATPases comes from studies of fungal and bovine enzymes. The V-ATPase is composed of two sectors that form a ball and stalk structure (7, 8). The V1 sector protrudes above the surface of the membrane. It has a molecular mass of In several ways the V-ATPase has a more complex structure than the F-ATPase. It is significantly larger and has subunits (e.g. C, H and d) with no obvious counterparts in the F-ATPase (1, 2). However, genetic experiments with Saccharomyces cerevisiae have provided strong evidence that these subunits are essential components of the V-ATPase (1517). The integral membrane part of the rotor is composed of two or three different types of proteolipids (the c, c', and c'' subunits). The a subunit, which interacts with the proteolipids, is nearly 4-fold larger than its counterpart in the F-ATPase. Characterization of S. cerevisiae strains that lack each of the 14 subunits has provided strong evidence that every subunit is essential for the synthesis of a functional V-ATPase (1). Such a comprehensive study of V-ATPase mutants has not been possible in metazoans because the loss of the V-ATPase causes embryonic lethality (1820). Strains with a null mutation in a V-ATPase gene are viable in the filamentous fungus Neurospora crassa. Loss of the V-ATPase gives rise to a distinctive morphological change in the hyphae, loss of vacuolar function, and defects in development (21). In this report we have used this phenotypic complexity to ask whether strains of N. crassa lacking different V-ATPase subunits have identical phenotypes and to address specific questions concerning the subunit composition of the V-ATPase. Subunits C and H have sometimes been viewed as accessory subunits that are not part of the core catalytic mechanism (2225). Subunit H interacts with other proteins such as Nef in human immunodeficiency virus, type 1 (26) and yeast ectoapyrase (27), perhaps facilitating the association of the V-ATPase with other enzymes or membranes. Subunit C appears to readily dissociate from the V-ATPase. Purified preparations of the enzyme from N. crassa were initially reported to lack this subunit (28). Here we ask, do strains lacking the C and H subunits have a typical V-ATPase null phenotype? In S. cerevisiae, vph1 mutants (i.e. lacking subunit a) are able to grow, although slowly, in medium that has high calcium or alkaline pH, conditions that completely prevent the growth of strains lacking other V-ATPase subunits (29). An isoform of the a subunit, encoded by the STV1 gene, can partially substitute for the loss of Vph1p. Later work reported that in the wild type the STV1-encoded subunit functions in the Golgi, whereas the VPH1-encoded subunit is targeted to the vacuole (30). Our preliminary findings suggested that N. crassa has only one gene for this protein. In this report we ask, are multiple genes for the a subunit characteristic of fungal species? Some investigators have suggested that V-ATPases from animals lack the c' subunit, shown to be essential in S. cerevisiae (31). In this yeast c and c' are similar proteins with 53% sequence identity (32, 33). Identification of the c' subunit and characterization of strains with mutations in the gene encoding it have not been reported from other organisms. We have found a candidate gene for subunit c' in N. crassa and have characterized strains with mutations in this gene. We have also analyzed genome sequence data bases to ask what types of organisms have genes that may encode a c' subunit?
Protein Alignment and Phylogenetic AnalysisSequences were obtained from the Neurospora Genome Project (34). Protein sequences were aligned with ClustalW (www.ebi.ac.uk/clustalw/) (35). Phylogenetic analyses were performed using the program PAUP (Phylogenetic Analysis Using Parsimony; Sinauer, Sunderland, MA). Phylogenetic trees were constructed with JALVIEW (36). Mutation of vma Genes by Repeat-induced Point MutationThe procedure of repeat-induced point mutation (RIP) is an efficient means of inactivating genes in N. crassa (37). Briefly, an ectopic copy of the gene to be mutated is introduced by transformation. The strain is crossed, and before nuclear fusion occurs a DNA silencing mechanism changes many of the C nucleotides to T in both the endogenous and ectopic copies of the introduced gene. If the size of the duplicated gene is 2 kb or larger, 550% of the progeny are typically mutated (38). Progeny from the cross are screened either by direct analysis of the DNA or by the appearance of a new phenotype. We used PCR to amplify the vma-5, vma-11, vma-13, and vma-16 genes. The primers used and the sizes of the fragments generated are shown in Table 1. Each of the genes was then subcloned into the plasmid pCR2.1 TOPO from the TA cloning kit (Invitrogen). Gene fragments were excised from the pCR2.1 plasmids by digestion with either EcoRI (vma-5 and vma-13) or NotI and SpeI (vma-11 and vma-16) and subcloned into the vector pBM61 (39).
The resulting plasmids, pBM61/vma5, 11, 13, and 16, were used to transform the his-3 strain of N. crassa (FGSC#6103) by electroporation. The pBM61 vector contains a truncated form of the his-3 gene designed to facilitate the targeted integration of plasmid DNA at the his-3 locus (39). We selected transformants that were histidine prototrophs. Because conidia typically have multiple nuclei, these primary transformants were usually heterokaryons. By crossing the transformants to the his-3 strain and selecting his+ progeny, we obtained homokaryotic strains. Integration of the plasmid was confirmed by PCR using the T3 and T7 primer pair that flanks the cloning site in pBM61. The homokaryotic his+ transformants were then crossed again to the wild type strain 74A. For the vma-11 and vma-16 genes, we isolated transformants of both mating types and crossed them to each other. Thus, all of the nuclei in these crosses had a gene duplication and were susceptible to the RIP process. The ascospores produced from each cross were germinated on agar plates containing Vogel's salts (40), 2% sorbose, 0.05% glucose, and 0.05% fructose. After 24 h at 30 °C colonies with defective V-ATPase were identified by the morphological abnormality shown in Fig. 1. For all putative mutant strains, genomic DNA was prepared from 20 mg of lyophilized mycelia, using the DNeasy Plant kit (Qiagen), and amplified by PCR, using the primers shown in Table 1. Sequencing was performed by the DNA Sequencing Facility at the University of California, Berkeley.
The vph-1 gene was originally cloned in 1993, before sequences were available from many organisms. Briefly, we obtained from Dr. Morris Manolson (University of Toronto) 2 primers designed for the S. cerevisiae VPH1 gene (41). With these we used PCR to generate a 141-bp fragment of the vph-1 gene from N. crassa genomic DNA. We screened a N. crassa We used the RIP procedure to inactivate the vph-1 gene. A 3.9-kb SpeI fragment with the vph-1 coding region and flanks from pFO2B#1A was subcloned into the pBM61 vector and transformed into the his-3 strain. By genetic crosses to a his-3 strain and then to the wild type 74A, we obtained one rare isolate that was vph-1 null with no ectopic copy of the vph-1 gene.
Inactivation of vma Genes by Homologous Gene ReplacementDespite numerous attempts, we were not able to isolate a strain with a null mutation in the vma-3 gene by using the RIP procedure. Although much less efficient, we decided to use homologous integration to replace the vma-3 protein-coding region with the inl gene, which encodes an enzyme required for inositol synthesis. Using primers that contained BamHI sites (Table 1, pair A), we amplified a 4.9-kb fragment that contained the vma-3 protein-coding region plus a 2-kb flanking region on each side. This fragment was digested with BamHI and cloned into the BamHI site in the Bluescript SK+ vector (Stratagene) to produce pVMA3/5kb. We then used primers containing MluI and StuI sites (Table 1, pair B) to initiate replication on each side of the open reading frame, oriented such that the flanking regions and the vector were amplified, but not the open reading frame. Phosphate was added to the ends of the PCR product with T4 polynucleotide kinase, and subsequent ligation produced pVMA3noORF. Primers containing an MluI site (Table 1) were used to amplify the inl gene, and the gene was cloned into the MluI site in pVMA3noORF. This plasmid, named pVMA3/inl, had a functional inl gene flanked by the 5' and 3' noncoding regions of vma-3. It was used to transform the inl mutant strain by electroporation. Prototrophic transformants were selected. Essentially all of the transformants grew with normal hyphal morphology, indicating that the plasmid had integrated at an ectopic location, as is typical for N. crassa. However, one transformant had the same growth defects we had observed in other vma null strains. PCR analysis indicated that the inl gene had replaced the vma-3 gene. The strain with the vma-3::inl+ gene, had no detectable vma-3 protein-coding region by Southern blot analysis. Introduction of the pVMA35kb plasmid, which contains a functional copy of vma-3, restored growth (data not shown).
As described under "Results" analysis of the phenotype of vma-11-RIP strains suggested they contained some residual V-ATPase activity. For this reason we used homologous integration to construct a strain in which the coding region of the vma-11 gene was replaced by the bar gene, which confers resistance to the fungicide bialaphos (42). We used the same strategy as with pVMA3/inl described above. A 3-kb fragment of the vma-11 gene containing pCR2.1vma-11bar was introduced into the strain mus-52 inl a by electroporation. Strains with the mus-52 mutation have recently been shown to have higher rates of homologous integration (43). Colonies that grew in the presence of 200 µg of bialophos/ml (Duchefa Biochemie, Haarlem, The Netherlands) were selected. PCR analysis of a colony with a morphology similar to that observed in other vma null strains indicated that the vma-11 gene had been replaced by the bar gene. This strain, named vma-11::bar, had no detectable vma-11 protein-coding region by Southern blot analysis (data not shown).
Generation of Anti-subunit H AntibodyA cDNA for vma-13 was generated by PCR and cloned into the EcoRI-HindIII sites of pQE-9 (Qiagen), generating the plasmid pVMA-136His. The plasmid was transformed into the E. coli strain DH5 Methods for Measuring Growth Rates, Spore Germination, and Arginine Content of CellsThe procedures used for composition of media, growth of strains, and genetic analysis were those of Davis (44). Vogel's medium contains salts (40) plus 2% sucrose. The rate of growth was determined by inoculating at the edge of agar plates containing Vogel's medium. The plates were incubated at 30 °C, and the position of the growth front of the colony was marked at 8- and 16-h intervals. To assess the production of viable progeny from crosses, ascospores were spread on plates containing Vogel's salts, 2% agar, 2% sorbose, 0.05% glucose 0.05% fructose, and 0.15 mg of myo-inositol/ml. Germination was induced by heat shock (30 min at 60 °C). The spores were either left on the plates or transferred individually to tubes with 1 ml of medium (2% Vogel's salts, 2% sucrose, 0.15 mg of inositol/ml). The arginine content in whole cell extracts was assayed as described (21).
Comparison of Genes Encoding V-ATPase Subunits in N. crassa and Other FungiWe have previously described the N. crassa V-ATPase subunits A, B, D, E, F, G, c, and d (4550). Using the amino acid sequences of S. cerevisiae V-ATPase subunits, we searched the N. crassa genome and identified subunits C, H, a, c', c'', and e (Table 2). The N. crassa vma genes are scattered randomly throughout the genome. We did no further analysis of subunit e.
One interesting finding was that, unlike S. cerevisiae, N. crassa appeared to contain only a single gene encoding subunit a. We examined the genomes of other fungi for which complete sequences are available and constructed a phylogenetic tree of all genes encoding subunit a (Fig. 2). Other filamentous ascomycetes, such as Magnaporthe grisea and Aspergillus nidulans, had a single gene, closely related to the N. crassa gene. Cryptococcus neoformans, a basidomycete, and Schizosaccharomyces pombe, a fission yeast, also had only one gene. Only S. cerevisiae and other budding yeasts had two genes. The VPH1 and STV1 genes of S. cerevisiae are less similar to each other than either one is to its apparent ortholog in other budding yeasts. Both mouse and human cells appear to have four genes encoding subunit a, which are equally divergent in sequence from the fungal VPH1 and STV1 genes. As in the budding yeasts, each human isoform is more similar to an isoform from mouse than to other human isoforms. A second interesting finding concerned subunits c' and c''. Like yeast, N. crassa contains orthologs of both the vma-11 and vma-16 genes, encoding subunits c' and c'' (Table 2). These are small, hydrophobic proteins (proteolipids) with four or five transmembrane helices. We found orthologs to these genes in all of the completed fungal genomes. Subunits c and c' have similar sequences, but within fungi they are clearly distinct types of proteins. The N. crassa c' subunit has a higher degree of protein sequence identity to the S. cerevisiae c' subunit (66.5%) than it does to the N. crassa c subunit (59.5%). Fig. 3 shows the alignment of amino acid sequences of c and c' subunits from two filamentous fungi, two budding yeasts, and one fission yeast. The subunit c from mouse is also included. The helical regions of the proteins are most highly conserved, especially the fourth helix, which contains the glutamate (Glu138 in N. crassa subunit c) that is the probable proton-binding site. However, a few regions are unique to subunit c or c'. The loop connecting helices 2 and 3 is longer in the c' subunits and has conserved proline and histidine residues (marked with arrows in Fig. 3). Four residues in the loop between helices 3 and 4 (SFML in N. crassa) are consistently different in c' subunits. The subunit from mouse is more similar to subunit c than to c'.
In a phylogenetic tree the c and c' subunits of fungi cluster in two distinct groups (Fig. 4). The tree clearly suggests a closer relationship of the mammalian orthologs to fungal subunit c than to c'. For Caenorhabditis elegans the relationship is less clear. This organism has two "c-like" subunits that appear to have diverged from subunits in the other organisms but are still closer to subunit c (51). The plant Arabidopsis thaliana has five genes (not shown) encoding c subunits, all with nearly identical protein sequences that differ only in the first four residues (52). Overall, the analysis of the genome data bases was consistent with the conclusion that a distinct c' subunit is found only in the fungi.
Although it is typical for N. crassa genes to contain introns, the c and c' subunit genes have more than average. The vma-3 and vma-11 genes encode very small proteins yet contain 4 and 5 introns, respectively. Between the start and stop codons In contrast to the c' subunit, the c'' subunit is highly conserved among N. crassa, other fungi and mammals (Fig. 5). This subunit differs from c and c' by virtue of a putative additional helical region at the N terminus. The S. cerevisiae protein has at least 10 more residues in this region than the filamentous fungi; however, the N terminus of all the proteins is long enough for a transmembrane helix.
Analysis of Strains with Mutations in Genes Encoding Subunits C H, a, c, c', and c''We previously reported that disruption of vma-1, encoding subunit A of the N. crassa vacuolar ATPase, gives rise to a distinctive phenotype (21). The strain cannot grow on alkaline medium, growth on acidic medium is much slower than for the wild type, and the morphology of hyphae is severely altered. The strain does not form conidia (asexual spores) but is fertile in a cross if the other parent has a functional V-ATPase. When hyphae of opposite mating type fuse they form a multicellular fruiting body, the perithecium. Several hundred cells are generated, each with one nucleus from each parent. These nuclei fuse, immediately undergo meiosis, and then divide again to yield eight nuclei. Each nucleus develops into a separate heavily melanized spore. If a nucleus contains a mutant gene that disrupts development of that spore, an immature white spore is produced. The viability of ascospores that contain a vma-1 null gene is low, less than 1%. In the mycelia, levels of cellular arginine are reduced by 85% in the vma-1 null strain. In the wild type strain almost all of the cellular arginine is stored in the vacuole (53). Concanamycin, a specific inhibitor of the V-ATPase and a potent inhibitor of growth of the wild type strain (54, 55), has no effect on the growth of the null strain. We have used this complex phenotype to evaluate the roles of previously uncharacterized V-ATPase subunits in N. crassa. Strains with Mutations in Genes Encoding Subunits C and HAs described under "Experimental Procedures," the RIP procedure was used to inactivate the vma-5 and vma-13 genes. Sequencing showed that stop codons had been introduced into the first half of both genes (Table 3). The phenotypes of these mutant strains were nearly indistinguishable from those of the vma-1-RIP strain characterized previously. They were unable to grow at pH 7.2 and had the same severe morphological changes as the vma-1-RIP strain when grown at pH 5.8, which is optimal for N. crassa (Fig. 6). The mutant strains did not form conidia (asexual spores). The vma-5 and vma-13 mutant strains had only 15% the amount of arginine seen in the wild type (Table 4), the same as in the vma-1-RIP strain. Thus, the vma-5-RIP and vma-13-RIP strains had the V-ATPase null phenotype, indicating that subunits C and H were essential for generating a functional enzyme in vivo.
We observed one difference in the phenotypes of vma-5-RIP and vma-13-RIP strains as compared with the pvn1 strain (vma-1-RIP). Like the vma-1-RIP strain, vma-5-RIP and vma-13-RIP strains were fertile in a cross if the other parent had a functional V-ATPase. The crosses produced a high proportion, 2230%, of white spores. However, the crosses also produced viable vma-5-RIP and vma-13-RIP progeny at significantly higher rates than the cross with the pvn-1 strain (Table 5).
Does N. crassa Contain Multiple Forms of Subunit H?We previously reported that preparations of the N. crassa V-ATPase often appeared to have two polypeptides of 54 kDa, the size of subunit H (47). The V-ATPase of clathrin-coated vesicles from bovine cells contains two polypeptides of 50 and 57 kDa that co-purify with the enzyme. These proteins, termed SFD (sub-fifty-eight-kDa dimer), activate ATPase activity of the enzyme and couple ATPase activity to proton pumping (25). The proteins turned out to be two forms of the H subunit with the 57-kDa species containing a 54-base pair insert in the open reading frame corresponding to an 18-amino acid insert in the protein (56). We generated a polyclonal antibody against subunit H and used Western blots to visualize the protein in different cell fractions. The subunit H protein was concentrated in the vacuolar membrane fraction, where it appeared as a single band of 54 kDa (data not shown). Although it is possible that there are two subunits of nearly identical mobility, it appears that N. crassa, like S. cerevisiae, contains only one form of the subunit H protein. Strains with Mutations in the vph-1 Gene, Encoding Subunit aThe RIP procedure yielded one viable strain with mutations in the vph-1 gene. Sequencing showed 177 nucleotide changes and 12 in-frame stop codons in the first 2221 bp. No protein was detected in cell fractions on Western blots (data not shown), consistent with its being a null mutant strain. In S. cerevisiae, VPH1 deletion strains give a partial vma phenotype because STV1 can partially compensate for the absence of VPH1 (29). By contrast, the phenotype of the N. crassa vph-1-RIP strain was indistinguishable from that of the vma-1-RIP strains. On pH 5.8 medium the vph-1-RIP strain was slow growing and had altered morphology, and it failed to grow on alkaline medium (Fig. 6). It was defective in sequestering arginine in the vacuole (Table 4). Cell fractions had no detectable V-ATPase activity (data not shown). Like the pvn1 strain (vma-1-RIP) fewer than 1% of the viable progeny from crosses contained the vph-1 RIP gene (Table 5). These data confirmed that N. crassa has a single gene encoding subunit a, in agreement with the analysis of the genome data base discussed above. Strains with Mutations in the vma-3 Gene, Encoding Subunit cWe were unable to generate a strain with a null mutation in the vma-3 gene using the RIP procedure. However, we were able to delete the protein-coding region of vma-3 by homologous replacement with the inl gene as described under "Experimental Procedures." The vma-3::inl+ strain had the same phenotype as vma-1-RIP with regard to growth, morphology, and cellular arginine (Fig. 6 and Table 4). It differed from other vma null strains in one characteristic. Although vma-3::inl+ was fertile when mated with a vma+ strain and produced vma+ progeny, we never observed viable progeny that inherited the vma-3::inl+ gene. Disruption of the vma-3 gene appeared to completely prevent the development of viable spores (Table 5).
Strains with Mutations in the vma-16 Gene, Encoding Subunit c''With the RIP procedure we generated a strain mutated in the vma-16 gene. The phenotype was indistinguishable from that of the pvn1 (vma-1-RIP) strain (Fig. 6 and Table 4). Mutant progeny were obtained from crosses, but at very low rates, fewer than 1% of the viable progeny. Strains with Mutations in the vma-11 Gene, Encoding Subunit c'The RIP procedure readily generated strains with mutations in the vma-11 gene. To our surprise the growth phenotypes of these strains were subtly different from the strains mutated in any of the other vma genes. Three vma-11-RIP strains had 518 amino acid changes, and the vma-11-RIP-16 strain had a stop codon at residue 82 (Fig. 7). Despite these extensive changes in the c' subunit, the vma-11-RIP strains grew slightly better than other vma mutant strains on pH 5.8 medium; they also grew on medium at pH 7.2 (Fig. 6). These results suggested that the vma-11-RIP strains were not completely null for V-ATPase function.
To generate a strain that completely lacked the vma-11 gene, we used homologous replacement as described under "Experimental Procedures." The resulting mutant strain, vma-11::bar, had a more severe growth defect than the vma-11-RIP strains but still grew marginally better than other vma null strains on pH 5.8 medium. Like the other vma-11 mutant strains, it produced more aerial hyphae and grew slowly but significantly on alkaline medium (Fig. 8). To compare growth rates we measured radial growth on solid medium, pH 5.8 (Fig. 9). The wild type strain grew much faster than any of the vma mutant strains, completely covering the agar plate in 48 h. The vma-11-RIP-25 strain, which had 5 amino acid changes, grew at less than half the rate of wild type but faster than any other mutant strain. The vma-11::bar strain grew slightly faster than the strains with mutations in other vma genes. For example, at 100 h vma-11::bar had grown To determine whether the vma-11 mutant strains retained residual V-ATPase activity, we tested the effects of concanamycin, a specific V-ATPase inhibitor, on growth. At pH 5.8, 0.2 µM concanamycin causes slow, morphologically abnormal growth of the wild type strain (55), similar to that observed in vma null strains. All of the vma mutant strains, except vma-11, grew the same in the presence or absence of concanamycin (Fig. 10). The vma-11 mutants, including vma-11::bar, grew more slowly in the presence of concanamycin and failed to produce aerial hyphae. In the presence of concanamycin the vma-11 strains became indistinguishable from the other vma mutant strains. In general the amount of arginine in vma-11 mutant strains was similar to that of other vma mutants. Strain vma-11-RIP-25, which grew significantly better than the others (Fig. 9), had approximately twice as much arginine as the other vma-11 mutant strains.
The ability of the vma-11 mutant strains to produce viable ascospores corresponded roughly to the severity of the mutation (Table 5). vma-11-RIP-25 produced fewer white spores and a higher proportion (36%) of viable mutant progeny. vma-11::bar produced more white spores and only a few viable mutant progeny per thousand spores. vma-11-RIP-16 and vma-11-RIP-27, which are intermediate in phenotype, also produced many white spores, but The differences in the phenotypes of vma-11 mutant strains prompted us to try another test for the ability to produce conidia. In our previous investigation we found that the pvn-2 allele of vma-1, which has only 4 altered amino acids, produces abundant conidia if high concentrations of sorbitol are added to the medium (21). We repeated this experiment, using 1.5 M sorbitol, with the new vma mutant strains. We found that the vma-11::bar strain also produced conidia, whereas strains mutated in other vma genes did not conidiate (data not shown).
The inability of the vma null strains to produce conidia makes it more difficult to store these strains and to transform them. For most aconidial stains of N. crassa this problem can be overcome by forming a heterokaryon with a "helper strain." For example, if a vma-1-RIP strain is co-inoculated with an adenine-requiring strain (ad3-B), the two can fuse, and each will complement the deficiency of the other. Most of the vma-RIP strains formed vigorously growing heterokaryons with helper strains. However, strains with null mutations in genes encoding subunits c, c', and c'' did not (Table 5).
Disruption of a gene encoding a subunit of the N. crassa vacuolar ATPase gives rise to a distinctive phenotype (21). The strain cannot grow on alkaline medium, growth on acidic medium is much slower than for the wild type, and the morphology of hyphae is severely altered. The strains do not form conidia. Levels of cellular arginine, which is normally stored in vacuoles, are reduced by 85%. With the exception of strains mutated in vma-11, all of the strains we tested with putative null mutations in a V-ATPase subunit had this phenotype. The vma-5-RIP (subunit C) and vma-13-RIP (subunit H) strains were indistinguishable from the other vma null strains. Although these subunits do not appear to have counterparts in the F-ATPase, our data agree with observations made with S. cerevisiae that subunits C and H are essential for formation of a functional V-ATPase (1, 2). A significant difference between N. crassa and S. cerevisiae is the presence of one versus two genes encoding subunit a. Analysis of fungal genomes for which a complete sequence is available finds a two-gene family, VPH1 and STV1, only in the budding yeasts. Disruption of the vph-1 gene in N. crassa gives a complete null phenotype. These results argue strongly that N. crassa, and probably most other fungi, have only one gene encoding subunit a. This raises the interesting question of whether different forms of subunit a, or some other V-ATPase subunit, occur generally in Golgi versus vacuoles as observed in S. cerevisiae (30). Plants and animals have multiple genes for subunit a, expressed at different levels in different types of cells (3, 57, 58). In animals there is good evidence that specialized proton-secreting cells have a specific isoform of the a subunit in the plasma membrane (3, 59), and cells in the kidney have another isoform targeted to the endosomes (60). In eukaryotic cells V-ATPases have at least two types of proteolipid subunits, c (vma-3) and c'' (vma-16), both similar to the c subunit in F-ATPases. The c'' subunit is distinguished by additional residues at the N terminus that form an extra transmembrane helix. The function of this region is not known, and it can be deleted in S. cerevisiae without affecting the function of the V-ATPase (61). As expected, N. crassa and other fungi have genes encoding orthologs of c and c''. Disruption of these genes in N. crassa gave rise to strains with a typical vma null phenotype. The N-terminal region of c'' is shorter in N. crassa and other filamentous fungi than in S. cerevisiae but still of sufficient length to encode a fifth membrane helix.
Whether V-ATPases of animals and plants have a c' subunit orthologous to that encoded by the VMA11 gene in S. cerevisiae is unclear. All of the sequenced fungal genomes have this subunit. The human and mouse genomes have only two "c-like" subunits, which appear to correspond to c and c'' (Fig. 3). Plants have multiple proteolipid subunits, but they also are most similar to either c or c'' (52). Thus, our analysis suggests that subunit c' may be unique to fungi. Therefore, it was interesting to observe that strains with mutations in the N. crassa vma-11 gene had different phenotypes from any of the other vma mutant strains. The N. crassa V-ATPase was unusually tolerant of mutations in the vma-11 gene. We isolated strains with 518 amino acid changes, most in highly conserved positions. These mutant strains retained some V-ATPase function. They grew better than any other vma mutant strain, and the growth was inhibited by concanamycin, which is specific for the V-ATPase. The strain in which the vma-11 gene was deleted also grew better than the other vma null strains and showed some sensitivity to concanamycin. This strain was also able to produce conidia. These results suggest that a small amount of functional V-ATPase can be made in the absence of subunit c'. Malkus et al. (62) reported that subunit c' has a unique role in the assembly of the V-ATPase in S. cerevisiae. It mediates the binding of Vma21p, which appears to coordinate the assembly of V0 subunits in the endoplasmic reticulum and to help package them into transport vesicles. A possible explanation of the phenotype of the N. crassa mutant strains is that the V-ATPase in vma-11 RIP strains retains some ability to interact with vma-21p and allows the assembly of a small amount of enzyme. In the vma-11::bar strain a smaller amount of enzyme may assemble in the absence of subunit c'. Subunits c and c' are so similar that vma21p could possibly recognize subunit c, although with lower affinity. In summary, we speculate that other organisms make the V-ATPase without subunit c' and that an N. crassa strain lacking subunit c' can also generate some functional V-ATPase. Except for vma-11, the only phenotypic difference in strains lacking different V-ATPase subunits was in the ability to develop viable ascospores. Strains with the pvn2 allele of vma-1 form mature viable spores almost as well as the wild type (21). vma-5-RIP, vma-13-RIP, and the vma-11-RIP strains produced more immature spores and had a lower germination rate than pvn2; nevertheless, 411% of the viable progeny were vma mutants. Strains with the pvn1 allele of vma-1, vph-1-RIP, vma-16-RIP, and vma-11::bar produced only a few viable mutant progeny /thousand spores, and vma-3::inl+ produced none. We have not been able to formulate a hypothesis to explain all of these differences. The pvn2 strain has only 4 altered amino acids and, conceivably, has V-ATPase activity not detectable in our assays. It is also possible that strains lacking subunits C or H retain some function associated with the V-ATPase. These subunits appear to be on the periphery of the enzyme and have no counterparts in F-type ATPases. Subunit C easily dissociates from the N. crassa enzyme. In S. cerevisiae VMA13 deletion mutants (lacking H) are able to assemble the rest of the enzyme (17). The most severe effect on N. crassa spore development was observed in strains lacking the A, a, c, c', and c'' subunits. Because the a, c, and c'' subunits form the core of the V0 sector, the severity of the phenotype is not surprising. Likewise, subunit A forms most of the ATP-binding site and is thus a key component of the V1 sector (63). The nearly complete loss of ability to form spores in the vma-11::bar strain is somewhat surprising because that strain retained some V-ATPase activity. We can only speculate that the amount of enzyme in this strain is too small to form viable spores.
* This work was supported in part by National Institutes of Health Grants GM28703 and GM58903. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 To whom correspondence should be addressed. Tel.: 831-459-2245; E-mail: bowman{at}biology.ucsc.edu.
2 The abbreviations used are: V-ATPase, vacuolar ATPase; F-ATPase, F1F0-ATP synthase; RIP, repeat-induced point mutation.
We thank Ryan Kendle, Forest O'Neill, Tri Vo, Amber Rivera, and Julio Escobar for technical assistance.
This article has been cited by other articles:
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Advertisement | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||