Osmotic stress-induced increase of phosphatidylinositol 3,5-bisphosphate requires Vac14p, an activator of the lipid kinase Fab1p

Abstract

Phosphatidylinositol 3,5-bisphosphate (PtdIns[3,5]P(2)) was first identified as a non-abundant phospholipid whose levels increase in response to osmotic stress. In yeast, Fab1p catalyzes formation of PtdIns(3,5)P(2) via phosphorylation of PtdIns(3)P. We have identified Vac14p, a novel vacuolar protein that regulates PtdIns(3,5)P(2) synthesis by modulating Fab1p activity in both the absence and presence of osmotic stress. We find that PtdIns(3)P levels are also elevated in response to osmotic stress, yet, only the elevation of PtdIns(3,5)P(2) levels are regulated by Vac14p. Under basal conditions the levels of PtdIns(3,5)P(2) are 18-28-fold lower than the levels of PtdIns(3)P, PtdIns(4)P, and PtdIns(4,5)P(2). After a 10 min exposure to hyperosmotic stress the levels of PtdIns(3,5)P(2) rise 20-fold, bringing it to a cellular concentration that is similar to the other phosphoinositides. This suggests that PtdIns(3,5)P(2) plays a major role in osmotic stress, perhaps via regulation of vacuolar volume. In fact, during hyperosmotic stress the vacuole morphology of wild-type cells changes dramatically, to smaller, more highly fragmented vacuoles, whereas mutants unable to synthesize PtdIns(3,5)P(2) continue to maintain a single large vacuole. These findings demonstrate that Vac14p regulates the levels of PtdIns(3,5)P(2) and provide insight into why PtdIns(3,5)P(2) levels rise in response to osmotic stress.

Figure 1.

Proteins with identity to Vac14p exist in higher eukaryotes. Identical amino acids (black) and similar amino acids (gray) are highlighted. (Left) The NH₂-terminal sequence of S. cerevisiae VAC14 and related ORFs were aligned using ClustalW (http://searchlauncher.bcm.tmc.edu:9331/multi-align/multi-align.html). Sequences were identified by searching the indicated databases via the BLAST algorithm (Altschul et al., 1990). The C. albicans sequence was found in the C. albicans database (http://sequence-www.stanford.edu/group/candida/search.html). ORFs from S. pombe (EMBL/GenBank/DDBJ accession no. CAB08779.1), A. thaliana (EMBL/GenBank/DDBJ accession no. AAD12702.1), C. elegans (EMBL/GenBank/DDBJ accession no. CAB00043.1) and D. melanogaster (EMBL/GenBank/DDBJ accession no. AAF54829.1) were in the GenBank database (http://www.ncbi.nlm.nih.gov/BLAST). The M. musculus sequence was identified in the mouse EST database (http://www.ncbi.nlm.nih.gov/BLAST). The sequence shown is a consensus of two similar ESTs (EMBL/GenBank/DDBJ accession nos. BE573148 and BF162275). The H. sapiens sequence was in the human EST database (http://www.ncbi.nlm.nih.gov/BLAST). The consensus sequence of 14 similar ESTs from chromosome 16 (EMBL/GenBank/DDBJ accession nos. AL527155, AL535971, AL555680, AL556062, BE409891, BE696780, BE728471, BE893810, BE901196, BE937614, BF081182, BF091052, BF325708, and BG107035) is shown. The sequences contain at least 25% global identity and 42% global similarity to S. cerevisiae VAC14. (Right) The COOH-terminal sequence of yeast Vac14p and similar ORFs were identified and aligned as in the left sequence. The M. musculus sequence is a consensus of 12 similar ESTs (EMBL/GenBank/DDBJ accession nos. AA036005, AA050423, AA058300, AA276168, AA497446, AA670618, BE862623, BF023070, BF237130, BF720417, BG079707, W09660). The H. sapiens sequence is hypothetical protein, FLJ10305 found on chromosome 16, deposited in the human genome database (http://www.ncbi.nlm.nih.gov/genome/seq). One of the mouse ESTs (clone ID 468926) had been mapped to chromosome VIII (106 cM offset) with an inferred position on human chromosome 16 (16q22.1-qter).

Figure 1.

Proteins with identity to Vac14p exist in higher eukaryotes. Identical amino acids (black) and similar amino acids (gray) are highlighted. (Left) The NH₂-terminal sequence of S. cerevisiae VAC14 and related ORFs were aligned using ClustalW (http://searchlauncher.bcm.tmc.edu:9331/multi-align/multi-align.html). Sequences were identified by searching the indicated databases via the BLAST algorithm (Altschul et al., 1990). The C. albicans sequence was found in the C. albicans database (http://sequence-www.stanford.edu/group/candida/search.html). ORFs from S. pombe (EMBL/GenBank/DDBJ accession no. CAB08779.1), A. thaliana (EMBL/GenBank/DDBJ accession no. AAD12702.1), C. elegans (EMBL/GenBank/DDBJ accession no. CAB00043.1) and D. melanogaster (EMBL/GenBank/DDBJ accession no. AAF54829.1) were in the GenBank database (http://www.ncbi.nlm.nih.gov/BLAST). The M. musculus sequence was identified in the mouse EST database (http://www.ncbi.nlm.nih.gov/BLAST). The sequence shown is a consensus of two similar ESTs (EMBL/GenBank/DDBJ accession nos. BE573148 and BF162275). The H. sapiens sequence was in the human EST database (http://www.ncbi.nlm.nih.gov/BLAST). The consensus sequence of 14 similar ESTs from chromosome 16 (EMBL/GenBank/DDBJ accession nos. AL527155, AL535971, AL555680, AL556062, BE409891, BE696780, BE728471, BE893810, BE901196, BE937614, BF081182, BF091052, BF325708, and BG107035) is shown. The sequences contain at least 25% global identity and 42% global similarity to S. cerevisiae VAC14. (Right) The COOH-terminal sequence of yeast Vac14p and similar ORFs were identified and aligned as in the left sequence. The M. musculus sequence is a consensus of 12 similar ESTs (EMBL/GenBank/DDBJ accession nos. AA036005, AA050423, AA058300, AA276168, AA497446, AA670618, BE862623, BF023070, BF237130, BF720417, BG079707, W09660). The H. sapiens sequence is hypothetical protein, FLJ10305 found on chromosome 16, deposited in the human genome database (http://www.ncbi.nlm.nih.gov/genome/seq). One of the mouse ESTs (clone ID 468926) had been mapped to chromosome VIII (106 cM offset) with an inferred position on human chromosome 16 (16q22.1-qter).

Figure 2.

FAB1 supresses the vacuole inheritance, morphology, and acidification defects of vac14-Δ1. vac14-Δ1 mutants expressing pRS426 (a and e), VAC14 on a low copy plasmid (b and f), FAB1 on a multicopy plasmid (c and g), or FAB1 from a low copy plasmid (d and h) were labeled with either FM4-64 to visualize vacuole membranes (A) or with 200 μM quinacrine to assess vacuole acidification (B). The photographs were taken with both fluorescence and a low level of transmitted light. Bar, 5 μm.

Figure 3.

Vac14p associates with a membrane fraction. (A) Wild-type and vac14-Δ1 yeast were lysed with glass beads. Equivalent amounts of the cell extracts were separated by SDS-PAGE and transferred to nitrocellulose. Western blot analysis was performed with a 1:5,000 dilution of anti-Vac14p antibody. (B) Whole cell extracts of wild-type were subjected to differential centrifugation. Equivalent amounts were loaded in each lane. Lane 2, total crude extract; lane 3, P13, 13,000 g spin pellet; lane 4, S13, supernatant from the 13,000 g spin; lane 5, P100, 100,000 g spin pellet; and lane 6, S100, the supernatant from the 100,000 g spin. Western blot analysis with anti-Vac14 antibodies was performed. The nitrocellulose membrane was reprobed with anti-Vac8 polyclonal rabbit antibodies (Wang et al., 1998) at a 1:5,000 dilution.

Figure 4.

Vac14p is peripherally associated with isolated vacuoles. (A) Vacuoles from wild-type, vac14-Δ1, vac7-Δ1, and fab1-Δ1 were isolated on Ficoll flotation gradients. Equivalent protein amounts were loaded in each lane. Western blot analysis was performed with anti-Vac14 and anti-Vac8p antibodies. The nitrocellulose blot was then reprobed with anti-Pho8p antibodies at a 1:1,000 dilution. (B) Vacuoles were isolated from LWY7235. 35 μg of the total cell extract (lane 2) and increasing amounts of the P13 fraction (lanes 3–7), as well as increasing amounts of vacuoles (lanes 8–13), were loaded onto a 7.5% SDS-polyacrylamide gel. Western blot analyses were performed using polyclonal anti-Vac14p, anti-Pho8p, and anti-Kar2p antibodies. (C) Vacuoles were prepared, divided into six equal aliquots, and treated on ice with one of the following conditions: 0.1 M Na₂CO₃, pH 11.5, 1 M NaCl, 0.8 M NH₂OH, or left untreated. Samples were centrifuged at 100,000 g for 1 h at 4°C. The resultant supernatant fractions (S100) were separated and the pellets (P100) were resuspended in 100 μl cytosol cocktail. Equal amounts of each were separated on a 7.5% SDS-PAGE, transferred to nitrocellulose, and probed with anti-Vac14 antibody. The nitrocellulose blots were then reprobed with monoclonal anti-Pho8p antibodies. (D) Equal amounts of protein from total cell extracts were loaded into each lane of a 7.5% SDS-polyacrylamide gel. The proteins were then transferred to nitrocellulose. Western blot analyses were performed using anti-Vac14 antibodies and anti-Vac8p antibodies.

Figure 5.

GFP-Fab1p is localized to the vacuole membranes in vac14-Δ1 and vac7-Δ1. Wild-type (A), fab1-Δ1 (B), vac14-Δ1 (C), and vac7-Δ1 (D) strains expressing GFP-Fab1p from a 2μ plasmid (pRS426) were visualized using fluorescence light combined with low levels of transmitted light. Bar, 8 μm.

Figure 6.

Use of 4.5% perchloric acid results in a significantly higher extraction of phosphatidylinositol monophosphates and bisphosphates. Cells were labeled with myo-[2-³H]inositol for 12 h at 24°C and exposed to 0.9 M NaCl for 10 min. (A) Cells were lysed with glass beads in 1.5 ml chloroform/methanol/1 N HCl (1:1:1) at room temperature. Phospholipids were extracted in the organic phase and dried under a N₂ stream. (B) Cells were lysed with glass beads in 0.8 ml of 4.5% perchloric acid. Cell extracts were centrifuged at 14,000 rpm at 4°C to obtain a phospholipid-containing pellet. The phospholipids from both protocols were then deacylated and analyzed by HPLC. An online scintillation counter was used to detect the tritiated compounds. Approximately 3.2 × 10⁶ cpm (A, chloroform/methanol/HC1) and 3.6 × 10⁶ cpm (B, 4.5% perchloric acid) were injected, respectively.

Figure 7.

Simultaneous overexpression of VAC7 and VAC14 results in cells with a larger number of vacuole lobes and lobes of smaller size. Wild-type cells simultaneously expressing VAC7 and VAC14 from multicopy plasmids (A) or containing empty vectors, pRS424 and pRS423 (B). Vacuoles were labeled with 80 μM FM4-64. Several fields are shown for each. Bar, 8 μm.

Figure 8.

PtdIns(3,5)P ₂ is required for osmotic stress induced fragmentation of the vacuole. Wild-type (a and e), vac14-Δ1 (b and f), fab1-Δ1 (c and g), vac7-Δ1 (d and h) cells were labeled with FM4-64 to visualize vacuole membranes. Cells were incubated in YEPD alone (a–d) or YEPD containing 0.4 M NaCl (e–h) for 10 min. Bar, 8 μm.

Figure 9.

Proposed model for regulation of PtdIns(3,5)P₂ levels in yeast. PtdIns(3)P is produced by Vps34p (Schu et al., 1993; Stack et al., 1993) and is phosphorylated by Fab1p to produce PtdIns(3,5)P₂ (Cooke et al., 1998; Gary et al., 1998). Both Vac7p (Gary et al., 1998) and Vac14p are required for normal levels of PtdIns(3,5)P₂. Although Vac7p is required for conversion of PtdIns(3)P to PtdIns(3,5)P₂, it is unclear whether Vac7p plays a specific role in the osmotic stress response. Vac14p regulates Fab1p to produce basal levels of PtdIns(3,5)P₂ during vegetative growth and further stimulates Fab1p activity during osmotic stress. Although osmotic stress activates both Vsp34p and Fab1p, Vac14p activates Fab1p but not Vps34p. Increased levels of PtdIns(3,5)P₂ may protect cells from osmotic stress by modulating water efflux and ion influx across the vacuole membrane, regulating vacuole membrane fission and retrograde traffic, and possibly via modulation of other downstream targets of PtdIns(3,5)P₂.