Chitin Synthesis in a gas1 Mutant of Saccharomyces cerevisiae

Strains, growth conditions, and genetic methods.

The yeast strains used here are listed in Table 1. Standard techniques were used for diploid construction, sporulation, and tetrad dissection. Cells were grown in batches at 30°C in YNB-glucose (Difco yeast nitrogen base without amino acids at 6.7 g/liter, 2% glucose, and the required supplements) or in YEPD (1% yeast extract, 2% Bacto-Peptone, 2% dextrose). For solid media, 2% agar was added. Diploids were sporulated in New Sporulation Medium (8.2 g of sodium acetate, 1.9 g of KCl, 0.35 g of MgSO₄, 1.2 g of NaCl, and 15 g of agar per liter) at 24°C. Spore germination was carried out at 24°C on YEPDAT plates (YEPD, 2% agar, and 100 mg of adenine and 50 mg of tryptophan per liter) containing 0.5 M KCl.

The following Escherichia coli strains were used: JM101 [Δ (lac-proAB) thi strA supE endA sbcB hsdR (F′ traD36 proAB laqI^q lacZΔM15)], DH5α [F′ endA1 hsdR17(r_k⁻m_k⁻) supE44 thi-1 recA1 gyrANaI^r) relA1 Δ(lacZYA-orgF)U169 deoR(φ80 dlacΔ(lacZ)M15)], and TOP10 [F′ mcrA Δ(mrr hsdRMS mcrBC) Φ80lacZΔM15 ΔlacX74 recA1 deoR araD139 Δ(ara-leu)7697 galU galK rpsL (Str^r) endA1 nupG] (Invitrogen, Carlsbad, Calif.).

DNA manipulations.

Recombinant DNA manipulations were performed using standard techniques (30). Transformation of yeast cells was carried out using the lithium acetate procedure (14) or the Saccharomyces cerevisiae EasyComp Transformation Kit (Invitrogen).

Plasmid and strain construction.

The S. cerevisiae haploid strain WB2d was generated from the wild-type strain W303-1B by gene replacement (37). In order to obtain the gas1::HIS3 mutation we used PCR to synthesize a linear fragment which lacked almost the whole of the GAS1 open reading frame (ORF). The upstream primer (5′-TGC GGA CGA TGT TCC AGC GAT TGA AGT TGT TGG TAA TAA GGT CCT GTT CCC TAG CAT GTA-) was designed to join 40 bp corresponding to the region from nucleotides 63 to 102 of the GAS1 ORF with the 5′ end from nucleotides 66 to 83 of HIS3. The downstream primer (5′-AGA CTT GGA AGA AGA CCC CGA AGC GTT AGA AGA GGC AGT ACT TGC CAC CTA TCA CCA CCA-) was designed to join 40 bp corresponding to the region from nucleotides 1470 to 1509 of the GAS1 ORF with the 3′ end from nucleotides 1446 to 1426 of HIS3. These primers were used to amplify a ∼1.2-kbp fragment from YEp6. This PCR product was transformed into W303-1A and Y1306. His⁺ transformants were selected, and correct substitution was tested by PCR analysis. Immunoblot analysis further confirmed the absence of the GAS1 gene product.

To construct W303-chs1Δ and WAH-chs1Δ, the BamHI/BglII fragment (∼2.4 kbp) of pHV149 (32), which carries the chs1::URA3 allele, was used to transform strains W303-1A and WAH. Correct substitution at the CHS1 locus was verified by PCR analysis and the absence of CSI activity.

CHS1 and CHS2 radiolabeled RNA probes were obtained using the TOPO TA Cloning Kit Dual Promoter (Invitrogen). Plasmids pCHS1 and pCHS2 were constructed by inserting the following PCR fragments, covering the ORF regions, into the pCRII-TOPO vector: a 1.3-kbp fragment from pMS1 (3) for CHS1 (from nucleotides 531 to 1821 from ATG), and a 1.5-kbp fragment from pUC19-CHS2 (M. H. Valdivieso, unpublished data) for CHS2 (nucleotides 787 to 2253 from ATG). ACT1 and CHS3 probes were obtained from pACT and pCAL1 plasmids. pACT was obtained by cloning the 1.5-kbp HindIII-BamHI fragment of the ACT1 gene into the HindIII-and BamHI-digested pGEM-3Zf(+). Plasmid pCAL1 was constructed by inserting the BglII-HindIII fragment of ca. 0.9 kbp of the CHS3 gene into the pGEM-Blue plasmid cut with BamHI and HindIII.

RNA extraction and Northern analysis.

Total RNA was prepared according to the method of selective precipitation with LiCl (12). Northern analysis was performed as previously described, using nonradioactive or ³²P-radiolabeled single-stranded RNA probes generated by in vitro transcription (25). After hybridization at 50°C, blots were twice washed at 50°C in 5× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate) for 15 min, once in 1× SSC–0.1% sodium dodecyl sulfate (SDS) for 30 min, and twice in 0.1× SSC–0.1% SDS for 30 min. Two final washes were carried out at 68°C in 0.2× SSC–0.1% SDS for 15 min and 0.2× SSC for 2 min. Densitometric quantification of mRNA was performed by using a computer program. mRNA loading was normalized using the hybridization signal of ACT1.

Preparation of total extracts and membrane fractionation.

For total extracts, 2 × 10⁸ cells were collected by filtration, washed, and resuspended in ice-cold deionized water supplemented with a Protease Inhibitor Cocktail (Boehringer Mannheim), one capsule in 25 ml, and 1 mM phenylmethylsulfonyl fluoride. After a 2-min centrifugation at 4°C, the pellets were frozen in dry ice-acetone and stored at −80°C. After thawing, 400 μl of SB-minus buffer (0.0625 M Tris-HCl pH 6.8; 5% SDS), supplemented with the protease inhibitors, was added to each pellet. After the addition of glass beads, cells were broken by shaking on a vortex for four cycles of 1 min alternated with 1 min in ice. After 5 min of centrifugation, the clarified lysate was withdrawn, quickly frozen, and stored at −80°C until use. For the determination of protein concentration, 15-μl aliquots of lysates in duplicate and the DC Protein Assay (Bio-Rad) were used. For SDS-polyacrylamide gel electrophoresis (PAGE) analysis, appropriate amounts of a concentrated solution were added to the lysate in order to bring the samples to a final concentration of 10% glycerol, 5% β-mercaptoethanol, and 0.002% bromophenol blue (BFB). Before the loading, samples were denatured at 100°C for 2 min.

Membranes were prepared as described by Orlean (22), except that the protease inhibitors were present during the whole procedure. Then, 2 × 10⁹ cells were collected by centrifugation, washed once with cold distilled water and then again with TM buffer (50 mM Tris-HCl, pH 7.5; 2.5 mM MgCl₂), and finally resuspended in 1.5 ml of TM buffer. After mechanical breakage, a pooled cell-wall-free extract was obtained as described above and then centrifuged at 60,000 × g for 45 min at 4°C, yielding supernatant (S) and pellet (P) fractions. The P fraction was resuspended in 200 to 400 μl of SB-minus buffer containing protease inhibitors, whereas the S fraction was concentrated with Centricon-10. After we determined the protein concentrations, aliquots of the P fraction were supplemented with the appropriate amounts of glycerol, β-mercaptoethanol, and BFB, whereas an equal volume of double-strength SDS sample buffer (0.0625 M Tris-HCl, pH 6.8; 2.3% SDS; 5% β-mercaptoethanol; 10% glycerol) was added to aliquots of the S fraction. Samples were denatured at 100°C for 2 min.

Electrophoresis and immunoblotting.

Proteins were resolved by SDS-PAGE on 7 or 8% polyacrylamide slab gels. Immunodecoration was carried out as previously described (26). Mouse anti-hemagglutinin (HA) monoclonal HA.11 antibodies (Babco) were used at a 1:1,000 dilution in TBS (0.01 M Tris–0.9% NaCl, pH 7.4) containing 5% bovine serum albumin (BSA) and 0.5% Tween 20 and anti-Gas1p rabbit polyclonal antibodies at a 1:3,000 dilution in TBS, 5% BSA, and 0.1% Tween 20. Horseradish peroxidase-conjugated anti-mouse antibodies (Amersham) or anti-rabbit antibodies (Zymed) were diluted to 1:5,000 and 1:10,000, respectively, in TBS–5% BSA–0.2% Tween 20. Binding was visualized with the ECL Western Blotting Detection Reagent (Amersham) according to the manufacturer's instructions.

Measurement of chitin levels.

Pellets corresponding to 5 × 10⁹ cells were collected, resuspended in 4.5 ml of H₂O, and divided into three equal aliquots (one was used for the determination of dry weight, and the other two were centrifuged), and the pellets stored at −20°C until use. After three extractions with 3% NaOH at 75°C, the alkali-insoluble pellet was neutralized and treated for 16 h with 4 mg of Zymolyase 100T per ml at 37°C. The chitin present in the indigestible material of the alkali-insoluble fraction was measured as described previously (24). The micrograms of glucosamine were normalized to the milligrams of dry weight.

Measurement of CS and chitinase activities.

For CS activity measurements, cell extracts were obtained according to the protocol described previously (2). CS assays were performed in Tris (pH 7.5) in the presence of Mg²⁺, Co²⁺, or Co²⁺ plus Ni²⁺ in order to discriminate among the three different activities, as described earlier (7). For CSI activity, MES buffer at pH 6.3 was also used (7). The determination of CS activity in cells permeabilized with digitonin was accomplished as described previously (13), except that CSIII activity was measured in the presence of 50 mM Tris (pH 7.5), Co²⁺, and Ni²⁺. Chitinase activity assays were performed as described previously (17).

Microscopic techniques.

Chitin was visualized by fluorescence microscopy after being stained with 2 mg of Calcofluor White (CF) per ml (24).

CSI activity and CHS1 mRNA increases are not responsible for the hyperaccumulation of chitin in gas1Δ cells.

CS activities were measured in the wild-type (W303-1B) and gas1Δ (WB2d) strains in the presence or absence of trypsin. The results (Table 2) reveal that in the absence of trypsin, the CSI activity doubles in the gas1Δ mutant with respect to the isogenic strain and that trypsin treatment elicits a six- to sevenfold increase in CSI activity in both strains, a finding in agreement with the zymogenic properties of this enzyme. On the contrary, the CSII and CSIII activities were similar in both strains.

We then analyzed the expression of the CHS1, -2, and -3 genes in the gas1Δ mutant. The level of CHS1 mRNA undergoes an increase of about 70% in gas1Δ cells (Fig. 1A), whereas the expression of the CHS3 and CHS2 genes was unchanged (data not shown). The increase in the CHS1 mRNA level is in good agreement with the increase of CSI activity in the gas1Δ mutant.

In order to determine whether the induction of CSI activity is involved in the increase in cell wall chitin levels in the gas1Δ mutant, we analyzed the tetrads obtained after sporulation of the chs1Δ gas1Δ heterozygous diploid LF4 (see Table 1). The spores obtained after dissection of 19 asci germinated normally, and all of them were viable. Two tetratype tetrads were analyzed in greater depth. No additional detrimental effect on the growth rate or morphological modification of gas1Δ cells were brought about by the chs1Δ mutation. Staining with CF revealed that the surface fluorescence of gas1Δ cells was very intense and was not changed by introduction of the chs1 null mutation (data not shown).

The chitin present in the zymolyase-undigestible material of the alkali-insoluble fraction was measured. The amounts of chitin were 3.2 ± 0.5, 2.8 ± 0.15, 35 ± 5.8, and 28 ± 4.7 μg of glucosamine/mg (dry weight) of cells for the wild-type, chs1Δ, gas1Δ, and gas1Δ chs1Δ spores, respectively. In the gas1Δ chs1Δ mutant the chitin level was about 20% lower than in gas1Δ, but it was still 10-fold higher than in chs1Δ.

This result and the genetic analysis indicate that despite the increase in the in vitro CSI activity, Chs1p is not the major enzyme responsible for in vivo chitin synthesis in the gas1 null mutant.

The gas1 null mutation suppresses the lysed-bud phenotype of chs1 null mutants.

When chs1Δ cells are grown in unbuffered minimal medium, numerous small refractile buds can be observed by phase-contrast microscopy (3, 4). The refractile cells have been shown to be lysed cells because of the lack of the Chs1p repair function, which does not counterbalance the acid-induced chitinase activity after cell separation (4, 5). Surprisingly, in the chs1Δ gas1Δ double mutant this phenotype could not be observed (Fig. 1B).

We wondered whether the suppression in the lysed-bud phenotype of the chs1 gas1 null mutant might be a consequence of a decrease in chitinase activity induced by the gas1Δ mutation. Thus, we measured chitinase activity in the spores of a tetrad. Unexpectedly, the secreted chitinase activity was stimulated in the presence of the gas1Δ mutation, with activities of 0.12 ± 0.01 nmol/10⁷ cells/min for the wild-type spores and 0.18 ± 0.01, 0.4 ± 0.15, and 0.32 ± 0.12 nmol/10⁷ cells/min for the chs1Δ, gas1Δ, and chs1Δ gas1Δ spores, respectively. Thus, suppression of the lysis-bud phenotype cannot be due to a decrease in chitinase activity. Moreover, this result strongly points to the notion that the increase in chitin accumulation in gas1Δ cells must be due to an increase of the synthesis of this polymer and not to a reduction in its degradation.

Chs3p is responsible for the increase in cell wall chitin levels in the gas1 null mutant.

We checked whether deletion of the CHS3 gene might have any effect on the chitin synthesis process in the gas1Δ mutant. A modification of the CHS3 locus by plasmid targeting had provided the first indication that the phenotype of gas1Δ cells is severely affected in the double mutant (24). In order to avoid any residual activity of Chs3p, we used a construct in which a large portion of the CHS3 gene had been replaced by the LEU2 gene. A heterozygous chs3Δ gas1Δ diploid (LF3) was sporulated, dissection of 20 asci was carried out, and 76 spores were examined. Growth was scored at different times during incubation of the spores at 24°C. Most of the spores gave rise to visible colonies after 48 to 72 h, whereas microcolonies appeared after a further 72 h of incubation (Fig. 2A). The scoring of the phenotype revealed that all of the microcolonies belonged to the His⁺ Leu⁺ class and were therefore chs3Δ gas1Δ double mutants. Inoculation of chs3Δ gas1Δ spores in liquid YNB-glucose medium had detrimental consequences. The double-mutant cells appeared to be greatly damaged, exhibited an aberrant morphology with many irregularly shaped cells, and were unable to grow. After a prolonged incubation of 4 to 5 days, the double-mutant spores started to grow weakly and in stationary phase rapidly lost viability compared to the wild-type or chs3Δ spores (Fig. 2B). After CF staining of chs3Δ gas1Δ cells only a faint fluorescence was detectable where the primary septum is produced (Fig. 2C). This finding is consistent with the specific loss of Chs3p function. Moreover, many cells were dead and became permeable to the dye (data not shown). The double-mutant cells progressively adapted to growth in liquid medium, and the growth rate defect was gradually suppressed, although not completely. This adaptation did not occur through restoration of the cell wall chitin and is probably due to the selection of second-site suppressors.

Inactivation of CHS3 was found to dramatically reduce the level of chitin in the double mutant compared to the single gas1 null mutant (Fig. 2D). This result indicates that Chs3p is indeed responsible for the increase in chitin accumulation induced by inactivation of the GAS1 gene. The CSI, CSII, and CSIII activities of the spores of two tetrads were analyzed; no changes in in vitro CSIII or CSII activities were detected in the gas1Δ mutant spores compared to the GAS1 ones. As previously shown (Table 2), the CSI activity increased in the spores carrying the gas1 null mutation. Moreover, a further twofold increase in CSI activity was found in the chs3 gas1 double mutant compared to the gas1 single mutant (results not shown).

Analysis of Chs3p level in gas1Δ cells.

We studied whether Chs3p might be differentially expressed in gas1Δ cells with respect to controls. We introduced the gas1::HIS3 mutation into the Y1306 strain, which expresses Chs3p fused at the C-terminal with three HA epitopes (31). By Western blot analysis using anti-HA monoclonal antibodies, we detected the HA-tagged Chs3 protein of 150 kDa in total extracts. Its level slightly decreased in gas1Δ cells compared to those of the isogenic strain (Fig. 3A). Upon overexposure of the blots additional smeared fragments of about 66 and 60 kDa were detected, and one of about 47 kDa was present only in the mutant (Fig. 3B). The extent of recovery of these fragments was not always the same across a series of experiments, but the pattern was qualitatively reproducible. These results suggest that Ha-tagged Chs3p is more susceptible to proteolytic degradation in the gas1Δ mutant than in the wild type.

Membrane and supernatant fractions were obtained with the same protocol used to determine the CSIII activity. In the membrane fraction (P), the 150-kDa polypeptide was the prominent band and its level did not show any appreciable differences between wild-type and gas1Δ strains (Fig. 3A). No enrichment of the 47- or 60- to 66-kDa polypeptides was observed, suggesting that they are probably unstable. Analysis of the supernatant fractions did not reveal any relevant species recognized by the antibodies used (data not shown).

CSIII activity increases in gas1 mutant cells permeabilized with digitonin.

In order to measure CSIII activity under more physiological conditions, whole cells permeabilized with digitonin were used (13). To avoid the high CSI activity level detected under these conditions, we constructed a set of strains that were deleted of the CHS1 gene and were either wild type or mutant for the CHS3 and/or GAS1 genes. A heterozygous diploid (LF5) was constructed by crossing chs1Δ cells with the suppressed LF3-13D spore (chs3Δ gas1Δ) and then sporulated. Using this method we were able to detect an increase in CSIII activity in gas1Δ cells (30.4 ± 5.2 mU/10⁷ cells in the chs1Δ gas1Δ mutant spore compared to 9.6 ± 2.3 mU/10⁷ cells in the chs1Δ spore). No activity was detected in the chs1Δ chs3Δ or the chs1Δ chs3Δ gas1Δ control strains.