CAC3 (MSI1) Suppression of RAS2^G19V Is Independent of Chromatin Assembly Factor I and Mediated by NPR1

Abstract

Cac3p/Msi1p, the Saccharomyces cerevisiae homolog of retinoblastoma-associated protein 48 (RbAp48), is a component of chromatin assembly factor I (CAF-I), a complex that assembles histones H3 and H4 onto replicated DNA. CAC3 overexpression also suppresses the RAS/cyclic AMP (cAMP) signal transduction pathway by an unknown mechanism. We investigated this mechanism and found that CAC3 suppression of RAS/cAMP signal transduction was independent of either CAC1 or CAC2, subunits required for CAF-I function. CAC3 suppression was also independent of other chromatin-modifying activities, indicating that Cac3p has at least two distinct, separable functions, one in chromatin assembly and one in regulating RAS function. Unlike Cac1p, which localizes primarily to the nucleus, Cac3p localizes to both the nucleus and the cytoplasm. In addition, Cac3p associates with Npr1p, a cytoplasmic kinase that stablizes several nutrient transporters by antagonizing a ubiquitin-mediated protein degradation pathway. Deletion of NPR1, like overexpression of Cac3p, suppressed the RAS/cAMP pathway. Furthermore, NPR1 overexpression interfered with the ability of CAC3 to suppress the RAS/cAMP pathway, indicating that extra Cac3p suppresses the RAS/cAMP pathway by sequestering Npr1p. Deletion of NPR1 did not affect the quantity, phosphorylation state, or localization of Ras2p. Consistent with the idea that Npr1p exerts its effect on the RAS/cAMP pathway by antagonizing a ubiquitin-mediated process, excess ubiquitin suppressed both the heat shock sensitivity and the sporulation defects caused by constitutive activation of the RAS/cAMP pathway. Thus, CAC3/MSI1 regulates the RAS/cAMP pathway via a chromatin-independent mechanism that involves the sequestration of Npr1p and may be due to the increased ubiquitination of an Npr1p substrate.

Chromatin assembly factor I (CAF-I) is a complex of three proteins that has been purified from both mammalian and yeast cells that assembles histones H3 and H4 onto newly replicated DNA (24, 38). The chromatin assembly complex (CAC) is the combined complex of CAF-I with histones H3 and H4 (48). The three CAF-I proteins from Saccharomyces cerevisiae are designated Cac1p, Cac2p, and Cac3p and correspond to the human CAF-I proteins p150, p60, and p48, respectively. Deletion of any one of the three CAC genes in S. cerevisiae results in multiple phenotypes, including the derepression of telomere-adjacent genes, mislocalization of the telomere-binding protein Rap1p, and an increase in sensitivity to UV radiation (11, 24). However, deletion of any one or all three of the CAC genes is not lethal, indicating that there must be other activities in S. cerevisiae capable of chromatin assembly. Loss of one of the CAC genes coupled with loss of one of the HIR genes (which control histone H2A and H2B function) leads to synergistic defects in chromatin structure and decreased growth rates (22, 33).

CAC1 and CAC2 mRNAs are coordinately regulated, with expression peaking in the G₁ phase of the cell cycle (39). In contrast, CAC3 mRNA levels do not change through the cell cycle (39). All three Cac proteins copurify (23, 24, 28), but the majority of p48 in human cells is found in a large complex that does not include p150 or p60 (28). Vertebrate p48 also copurifies with the histone deacteylase HDAC1 (43) and with pRb, the product of the retinoblastoma susceptibility gene (32), which acts as a tumor suppressor. The closest Cac3p homolog in S. cerevisiae is Hat2p, the subunit of histone acetyltransferase I that is necessary for association with histones H3 and H4 (31). Viewed together, these data suggest that Cac3p/p48 plays multiple roles in the deposition and modification of histones (34).

Cac3p appears to have at least one role unrelated to its function in chromatin assembly and/or histone modification. CAC3 was originally isolated as MSI1, a multicopy suppressor of IRA1 (35), which is a negative regulator of the RAS/cyclic AMP (cAMP) pathway. High-copy CAC3 reduces cAMP levels in ira1 and RAS2^G19V strains, mitigating the heat shock sensitivity and sporulation deficiency of these strains. High-copy CAC3 also suppresses snf1 and snf4 mutations by decreasing cAMP levels (18). Overexpression of human p48, like CAC3/MSI1, can suppress the RAS/cAMP pathway in S. cerevisiae (32, 35).

The RAS/cAMP pathway has been well characterized in the yeast Saccharomyces. Two genes in S. cerevisiae, RAS1 and RAS2, are structural and functional homologs of human ras genes (8, 20), oncogenic mutations of which are found in 90% of pancreatic, 50% of colon, and 30% of lung adenocarcinomas as well as in 50% of thyroid tumors and 30% of myeloid leukemias (3). In yeast cells, Ras controls the metabolic state and alters the stress response of the cell through modulation of cAMP levels (5). Cdc25p stimulates Ras activity by promoting the exchange of GDP for GTP through the stabilization of Ras in a nucleotide-free state (16). Ira1p and Ira2p negatively regulate Ras proteins by stimulating their intrinsic GTPase activity (29). The RAS2^G19V allele, which is analogous to the most common oncogenic mutation in human cancers (3), encodes a protein that binds GTP normally but fails to hydrolyze it (30), resulting in constitutively active Ras protein. Activated Ras protein stimulates adenylyl cyclase (Cyr1p/Cdc35p), thereby promoting production of cAMP. cAMP binds to Bcy1p, the negative regulatory subunit of protein kinase A (PKA), and disassociates it from the catalytic subunit (encoded by any one of the three genes TPK1, -2, or -3) to yield enhanced kinase activity. PKA activation leads to glycogen utilization, increased glycolysis, the induction of many growth-related genes (5), and reduced expression of genes encoding heat shock proteins Hsp72, Hsp41 (37), and Hsp12 (47). Cells containing an activated RAS/cAMP pathway (those containing RAS2^G19V, bcy1Δ, or ira1Δ mutations) mate poorly, contain low levels of storage carbohydrates, and are sensitive to transient heat shock.

The mechanism by which Cac3p suppresses the heat shock sensitivity and the sporulation defect of cells with an activated RAS/cAMP pathway is not known. High-copy CAC3 suppresses phenotypes caused by the constitutively active RAS2^G19V allele or by deletion of the negative regulator IRA1 but does not suppress these phenotypes when the pathway is activated by deletion of BCY1 (35). We investigated the mechanism by which Cac3p overproduction suppresses the Ras2p^G19V oncoprotein. We found that the ability of CAC3 overexpression to suppress the RAS/cAMP pathway is independent of its role in CAF-I-mediated chromatin assembly and its putative role in histone modification activities. This result is similar to the recently published results of Zhu et al. (51). Consistent with this, a significant proportion of Cac3p resides in the cytoplasm, in contrast to Cac1p, which is primarily nuclear. We identified Npr1p as a protein that physically interacts with Cac3p and found that loss of Npr1p function suppresses the RAS/cAMP pathway in a manner indistinguishable from CAC3 overexpression. The Npr1p kinase is known to antagonize the ubiquitin-mediated inactivation of several transporter proteins; similarly, we have found that overexpression of polyubiquitin is also capable of suppressing RAS2^G19V-induced heat shock sensitivity. Our results suggest that CAC3 sequesters and thereby inactivates the Npr1p kinase, which effectively reduces the activity of the RAS/cAMP pathway.

MATERIALS AND METHODS

Strains and plasmids.

Escherichia coli strains XL1-Blue and DH5α were used for all standard plasmid preparations and manipulations (1). pML9 for the disruption of NPR1 was provided by J. Heitman (26). pJW192, encoding a Ras2p-green fluorescent protein (GFP) fusion, was provided by J. Rine (4). pCUP1-myc-UBI4 was provided by M. Hochstrasser (9). pbcy1::URA3 has been described (45). Triple-hemagglutinin (HA) epitope-tagged Npr1p carrying a mutation that presumably blocks its kinase activity was a generous gift from Yu Jiang. This mutation, D579E, was created on the basis of homology to mutations known to inactivate other kinases, although the effect of the mutation on Npr1 kinase activity has not been demonstrated.

pGBT9-CAC3 was amplified from yeast strain Y294 by PCR amplification using the oligonucleotides 5′-GGCCGGGGATCCATGAATCAGTGCGCGAAGG-3′ and 5′-GGGCCCGTCGACTCACGAATGTCCAACAAGGTTTCC-3′. The PCR product was cloned into pGBT9C which had been digested with BamHI and SalI. pGBT9C is the pGBT9 vector which has been altered in the reading frame of the multiple cloning site by the addition of two additional guanine residues before the EcoRI site and was a generous gift from Clint McDonald. YEp55-CAC3 was constructed by amplifying the CAC3 gene from pGBT9-CAC3 by PCR using the oligonucleotides 5′-GGCCGGGTCGACATGAATCAGTGCGCGAAGG-3′ and 5′-GGGCCCGGATCCTCACGAATGTCCAACAAGGTTTCC-3′ and was cloned into YEp55S that had also been digested with SalI and BamHI. The multiple cloning site of this vector had been modified through the introduction of a SalI restriction site by Corey Davis and was a gift from him. pRS406-Ras2Val19 was constructed by isolating the 2.1-kb genomic EcoRI-HindIII fragment containing the RAS2^G19V allele and cloned into pRS406 that had been digested with EcoRI and HindIII. pGAD1-CAC1 was cloned from a yeast genomic library constructed by Stan Fields and was a generous gift from Mark Rose.

NPR1 (with its start and stop codons) was amplified from W303 genomic DNA by PCR and cloned into pCR-II (Invitrogen). The XhoI-HindIII fragment containing the NPR1 gene was subcloned into pRSET-B (Invitrogen) to yield pRSET-NPR1. pRSET-NPR1 was digested with XbaI and the fragment was cotransformed into S. cerevisiae with pGalSET984 (10) linearized with XhoI. In vivo recombination between these two DNA fragments yielded pGalSET-NPR1 (10). Similarly, CAC3 was amplified from W303 genomic DNA by PCR and cloned into pCR-II. The BamHI-HindIII fragment of this plasmid containing the CAC3 gene was subcloned into pRSET-B. The XbaI fragment containing CAC3 and XhoI-linearized pGalSET985 were cotransformed into yeast cells, selecting for in vivo recombination. Immunoblotting confirmed that galactose induction of strains carrying either pGalSET-NPR1 or pGalSET-CAC3 generated an epitope-tagged protein of the expected size (data not shown). The protein encoded by YEp55-CAC3 was tagged at the carboxy terminus with GFP, using the PCR-mediated technique described by Longtine et al. (25). URA3 was integrated adjacent to the left telomere of chromosome VII as described (12).

The yeast strains used in this study are listed in isogenic groups in Table 1. Strains were grown in standard laboratory SD complete (SDC) medium with the appropriate amino acid dropouts (15). Genetic crosses, sporulation, dissection, and transformation were performed as described (15). Auxotrophic markers were swapped as described (7). The [rho⁻] strain was made by growth of the parental strain in the presence of ethidium bromide (25 μg/ml).

TABLE 1.

Yeast strains used

Isogenic group and strain	Relevant characteristic(s)	Source or reference
W303
YJB195	MATaura3-1 ade2-1 his3-11 leu2-3,112 can1-100 trp1-1	Berman lab
YJB209	MATα ura3-1 ade2-1 his3-11 leu2-3,112 can1-100 trp1-1	Berman lab
YJB334	MATa/α ura3-1/ura3-1 ade2-1/ ade2-1 his3-11/his3-11 leu2-3,112/leu2-3,112 can1-100/can1-100 trp1-1/trp1-1	Berman lab
YJB1358	YJB209 cac1-1	11
YJB1599	YJB209 cac2::TRP1	24
YJB1786	YJB209 cac3::hisG VIIL::URA3-TEL	This study
YJB2235	YJB195 URA3::RAS2G19V	This study
YJB2237	YJB209 cac1-1 URA3::RAS2G19V	This study
YJB2320	YJB195 URA3::RAS2G19V YEp55-CAC3	This study
YJB2322	YJB209 cac1-1 URA3::RAS2G19V YEp55-CAC3	This study
YJB2543	YJB195 sir3::TRP1	Berman lab
YJB2546	YJB195 gcn5::URA3	Berman lab
YJB3135	YJB195 gcn5::ura3::TRP1 URA3::RAS2G19V	This study
YJB3141	YJB195 gcn5::ura3::TRP1 URA3::RAS2G19V YEp55-CAC3	This study
YJB3554	YJB195 npr1::LEU2 bcy1::URA3	This study
YJB3563	YJB209 cac2::TRP1 URA3::RAS2G19V	This study
YJB3610	YJB209 VIIL::URA3-TEL pGalSET	This study
YJB3670	YJB195 sch9::TRP1 YCP50-RAS2G18A G19V	50a
YJB3674	YJB195 sch9::TRP1 YCP50-RAS2G18A G19V YEP55-CAC3	This study
YJB3712	YJB195 npr1::LEU2	This study
YJB3721	YJB334 ura3-1/URA3::RAS2G19V YEp55-CAC3	This study
YJB3723	YJB195 npr1::LEU2 URA3::RAS2G19V	This study
YJB3724	YJB2235 pCUP1-myc-UB14	This study
YJB3737	YJB334 ura3-1/URA3::RAS2G19V	This study
YJB3739	YJB334 YEp55-CAC3	This study
YJB3763	YJB195 sir3::TRP1 URA3::RAS2G19V	This study
YJB3773	YJB195 sir3::TRP1 URA3::RAS2G19V YEp55-CAC3	This study
YJB3785	MATahis3 Leu2-3, 112 trp1 ura3-1	4
YJB3791	YJB334 ura3-1/URA3::RAS2G19V npr1::LEU2/npr1::LEU2	This study
YJB3792	YJB334 npr1::LEU2/npr1::LEU2	This study
YJB3863	YJB195 URA3::RAS2G19V pGalSET-NPR1	This study
YJB3875	YJB195 URA3::RAS2G19V pGalSET-NPR1 pGalSET-CAC3	This study
YJB3876	YJB209 cac2::TRP1 URA3::RAS2G19V YEp55-CAC3	This study
YJB3896	YJB3785 pGFP-RAS2	This study
YJB3897	YJB3785 npr1::LEU2 pGEP-RAS2	This study
YJB3898	YJB3785 YEp55-CAC3 pGFP-RAS2	This study
YJB3954	YJB195 pGalSET-NPR1	This study
YJB4034	YJB3785 YEp55-CAC3	This study
YJB4317	YJB4034 YEp55-CAC3-GEP	This study
YJB4506	YJB4317 YEp55-CAC3-GFP [rho−]	This study
YJB4632	YJB209 gpa2::TRP1	50a
YJB4635	YJB209 gpr1::HIS3	50a
YJB5235	YJB195 URA3::RAS2G19V gpa2::TRP1 YEp55-CAC3	This study
YJB5243	YJB195 URA3::RAS2G19V gpa2::TRP1	50a
YJB5237	YJB195 URA3::RAS2G19V gpr1::HIS3 YEp55-CAC3	This study
YJB5245	YJB195 URA3::RAS2G19V gpr1::HIS3	50a
YJB5522	YJB3723 pCUP1-myc-UB14	This study
YJB5523	YJB3737 pCUP1-myc-UB14	This study
YSJ328	YJB3954 VIIL::URA3-TEL	This study
Y2799	MATα ura3-1 ade2-1 his3-11 leu2-3,112 can1-100 trp1-1 npr1::LEU2 pRS424 pGalSET-CAC3	This study
Y2800	MATα ura3-1 ade2-1 his3-11 leu2-3,112 can1-100 trp1-1 npr1::LEU2 pRS424- 3HA-NPR1 pGalSET-CAC3	This study
Y2801	MATα ura3-1 ade2-1 his3-11 leu2-3,112 can1-100 trp1-1 npr1::LEU2 pRS424-3HA- NPR1-KD pGalSET-CAC3	This study
YPH49/50
YJB1583	MATaura3-52 lys2-801 ade2-101 trp1-ars1 his3-200 leu2Δ1	31
YJB1584	YJB1583 hat1::HIS3	31
YJB1585	YJB1583 hat2::TRP1	31
YJB1586	YJB1583 hat1::HIS3 hat2::TRP1	31
YJB2349	YJB1583 URA3::RAS2G19V	This study
YJB2350	YJB1583 hat1::HIS3 URA3::RAS2G19V	This study
YJB2366	YJB1583 URA3::RAS2G19V YEp55-CAC3	This study
YJB2367	YJB1583 hat1::HIS3 URA3::RAS2G19V YEp55-CAC3	This study
YJB2648	YJB1583 hat2::TRP1 URA3::RAS2G19V	This study
YJB2658	YJB1583 hat2::TRP1 URA3::RAS2G19V YEp55-CAC3	This study
YJB2672	YJB1583 hat1::HIS3 hat2::TRP1 URA3::RAS2G19V	This study
YJB2673	YJB1583 hat1::HIS3 hat2::TRP1 URA3::RAS2G19V YEp55-CAC3	This study
YJB2697	YJB1583 bcy1::URA3	This study
YJB2781	YJB1583 bcy1::URA3 YEp55-CAC3	This study
YDS3
YJB2431	MATα ade2-1 can1-100 his3-11,15 leu2-3,112 trp1-1 ura3-1	35a
YJB2432	YJB2431 hda1::TRP1	35a
YJB2433	YJB2431 rpd3::LEU2	35a
YJB2434	YJB2431 hda1::TRP1 rpd3::LEU2	35a
YJB2474	YJB2431 URA3::RAS2G19V	This study
YJB2475	YJB2431 hda1::TRP1 URA3::RAS2G19V	This study
YJB2476	YJB2431 rpd3::leu2::HIS3 URA3::RAS2G19V	This study
YJB2474	YJB2431 hda1::TRP1 rpd3::leu2::HIS3 URA3::RAS2G19V	This study
YJB2529	YJB2431 URA3::RAS2G19V YEp55-CAC3	This study
YJB2530	YJB2431 rpd3::leu2::HIS3 URA3::RAS2G19V YEp55-CAC3	This study
YJB2531	YJB2431 hda1::TRP1 URA3::RAS2G19V YEp55-CAC3	This study
YJB2532	YJB2431 hda1::TRP1 rpd3::leu2::HIS3 URA3::RAS2G19V YEp55-CAC3	This study
HF7c
YJB1325	MATaura3-52 his3-200 ade2-101 lys2-801 trp1-901 leu2-3,112 gal4-542 gal180-538 LYS2::GAL1-HIS3
	URA3::(GAL4 17-mers)3-CYC1-lacZ	Clontech Inc.
YJB3197	YJB1325 pGBT9-Cac3p pGAD1-CAC1	This study
YJB3438	YJB1325 pGBT9-Cac3p pGAD1-NPR1561–605	This study
L1356
YJB2136	MATaura3-52 his3-200 leu2-1 met8-1 ilv1-1 trp1-901 cac3-878 lys2::his3Δ4	33
YJB2137	YJB2136 hir3::LEU2	33
YJB2998	YJB2136 hir3::leu2::TRP1 URA3::RAS2G19V	This study
YJB2999	YJB2136 hir3::leu2::TRP1 URA3::RAS2G19V YEp55-CAC3	This study
YJB3016	YJB2136 URA3::RAS2G19V	This study
YJB3017	YJB2136 URA3::RAS2G19V YEp55-CAC3	This study
Σ1278b
YJB2444	MATa/α ura3-52/ura3-52	24a
YJB3514	YJB2444 pGalSET	This study
YJB3515	YJB2444 pGalSET-CAC3	This study
YJB5723	YJB5724 npr1::LEU2/npr1::LEU2	26
YJB5724	MATa/α ura3-52/ura3-52	26

Heat shock and sporulation assays.

Yeast cells were grown to saturation on appropriate dropout medium with either 2% glucose or 2% galactose present as the carbon source. Cells were collected by centrifugation, washed once, and resuspended in water. Aliquots were incubated at 55°C for the time periods indicated. After cooling to room temperature, cells were serially diluted 10-fold, plated, and incubated at 30°C for 2 days. Resulting colonies were counted and normalized to the number of viable cells in an aliquot that was not exposed to 55°C. To determine sporulation efficiency, diploid strains were grown overnight in appropriate dropout medium supplemented with 2% galactose to induce the expression of genes under control of the GAL1 promoter, as appropriate. Cells were collected by centrifugation, washed twice, and allowed to sporulate for 3 days in 1% potassium acetate at room temperature, with shaking. Diploids and tetrads (at least 300 cells of each strain) were counted by microscopic observation.

Two-hybrid screen.

An S. cerevisiae genomic DNA two-hybrid library in pGAD1, -2, or -3 (6) was kindly provided by P. Siliciano (University of Minnesota) and transformed into HF7c containing pGBT9-CAC3 as bait. Transformed cells which contained interacting fusion proteins were selected by plating on SDC lacking Leu, Trp, and His and supplemented with 5 mM 3-amino-1,2,4-triazole (3AT). Transformants that contained GAL4 were identified by PCR and discarded. Surviving yeast strains were cured of either plasmid to ensure that growth on the selective medium was dependent on the presence of both plasmids. Inserts in pGAD were amplified directly from the yeast strain by PCR using primers flanking the multiple cloning site. Amplified DNA was sequenced using a nested primer.

Coimmunoprecipitation.

Yeast strains YSJ401, YSJ402, and YSJ403 were grown in SC lacking Leu and Trp containing 2% galactose for 2 days. Cells were collected by centrifugation and washed once with water and once with 10mM sodium azide. Cell pellets were then frozen at −70°C. Cells were then resuspended in 200 μl of buffer C (20 mM HEPES-KOH [pH 7.4], 150 mM NaCl, 1 mM EDTA, 5 mM MgCl₂, 2 mM phenylmethylsulfonyl fluoride plus 1 μg of pepstatin A, 0.5 μg of leupeptin, and 2 μg of aprotinin per ml) and lysed by vortexing with glass beads. Buffer C (200 μl) was then added, and unlysed cells, cellular debris, and glass beads were removed by centrifugation. Lysate (70 μg in 450 μl) was added to 20 μl of anti-HA affinity matrix (Roche) that had been blocked with bovine serum albumin (BSA, 2 mg/ml) in buffer C. Loading buffer (5 × sodium dodecyl sulfate [SDS]) was added to an aliquot of the input as a control. The lysate was allowed to bind to the affinity matrix for 3 h at 4°C. The affinity beads and bound proteins were pelleted by centrifugation, and the supernatant was retained. The 5 × SDS loading buffer was added to an aliquot of the supernatant as a control. Beads were then washed four times with buffer C. After the final wash, beads were boiled for 3 min in 50 μl of 1 × SDS sample buffer, and an aliquot (15 μl) of each sample was loaded onto an 8% polyacrylamide–SDS gel and subjected to polyacrylamide gel electrophoresis (PAGE). Samples were then transferred onto a polyvinylidene difluoride membrane and probed with monoclonal antibodies against the T7 epitope (Novagen) or the HA epitope (12CA5). Immunoreactivity was detected using ECL (Amersham).

Microscopy and pseudohyphal assays.

To determine the subcellular localization of Cac3p-GFP, YJB4506 was grown in SDC lacking Trp and supplemented with 2% raffinose, 0.5% galactose, and 10 ng of DAPI (4′,6′-diamidino-2-phenylindole) for 4 h at 30°C. Cells were viewed using a Nikon Eclipse E800 photomicroscope equipped with differential interference contrast and fluorescence optics using a 100 × 1.3-numerical-aperture plan apo objective. Digital images were collected using a CoolCam liquid-cooled, three-chip color charge-coupled device camera (Cool Camera Company, Decatur, Ga.) and captured to a Pentium II 300-MHz personal computer using Image Pro Plus version 4.0 software (Media Cybernetics, Silver Spring, Md.). Pseudohyphal growth was assayed as described by Lorenz and Heitman (26). Σ1278b strains carrying either a CAC3 overexpression plasmid or the parental plasmid were grown overnight in either glucose or galactose. Cells were then plated on limiting nitrogen medium with either glucose (SLAD) or raffinose and galactose (SLADG) as a carbon source (26) and grown for 3 days at 30°C before being photographed.

Metabolic labeling and immunoprecipitation.

³²P metabolic labeling and Ras2p immunoprecipitation were performed essentially as described by Whistler and Rine (50). In brief, wild-type or npr1Δ cells were grown overnight in YPAD medium, washed in SDC low-phosphate medium, and grown in 10 ml of this medium at 30°C for 2 h. Then 2 mCi of H₃³²PO₄ (ICN) was added, and the culture was incubated at 30°C for an additional 3 h. Cells were collected by centrifugation, washed in NLB buffer (50 mM Tris [pH 7.5], 20 mM MgCl₂, 100 mM NaCl, 0.1% Nonidet P-40, 1 mM dithiothreitol, 1% aprotinin, 1 mM phenylmethylsulfonyl fluoride) and stored at −70°C overnight. Cells were resuspended in 0.5 ml of NLB and lysed by bead beating. Unlysed cells and cell debris were removed by centrifugation. BSA-treated charcoal was added to the lysates, which were vortexed and centrifuged several times to remove all traces of the charcoal. Then 20 μl of either protein A-agarose or anti-Ras2-agarose (clone Y13-259; Calbiochem) was added to the lysate and mixed at 4°C for 2 h. Beads were washed three times with NLB and three times with NLB without detergent. Proteins were eluted from the beads by incubating at 70°C for 10 min in 20 μl of SDS loading buffer. Proteins were separated by SDS-PAGE through a 9% gel, which was fixed and dried, and and bands were quantitated by PhosphoImager analysis (Molecular Dynamics).

RESULTS

CAC3 suppression of the RAS/cAMP pathway is independent of CAF-I.

To understand the molecular mechanism by which CAC3 overexpression affects the RAS/cAMP pathway, we first established a quantitative assay for heat shock sensitivity, a phenotype caused by activation of the RAS/cAMP pathway. Cells were grown to saturation, aliquoted and incubated at 55°C for different lengths of time, cooled, serially diluted, and plated on nonselective medium. Resulting colonies were counted after 2 days at 30°C. Yeast cells carrying the dominant RAS2^G19V allele have a constitutively activated RAS/cAMP pathway (30) and are sensitive to heat shock (Fig. 1A). Overexpression of CAC3 suppressed the heat shock sensitivity of a strain carrying the RAS2^G19V allele (35) (Fig. 1A), restoring wild-type heat shock resistance. In contrast, heat shock sensitivity caused by deletion of BCY1, which encodes the negative regulator of PKA, was not suppressed by CAC3 overexpression (35) (Fig. 1B). Thus, our quantitative heat shock sensitivity assay confirmed the reported ability of CAC3 overexpression to suppress the RAS/cAMP pathway when it is activated by a RAS2^G19V allele but not when it is activated by BCY1 deletion.

FIG. 1.

CAC3 overexpression suppresses the RAS/cAMP signal transduction pathway between RAS and PKA. The heat shock resistance of (A) isogenic wild-type (YJB195), RAS2^G19V (YJB2235), and RAS2^G19V CAC3-over expressing (OE) (YJB2320) strains and (B) isogenic wild-type (YJB1583), bcy1Δ (YJB2697), and bcy1Δ CAC3-overexpressing (YJB2781) strains was determined by incubation at 55°C for the indicated time periods as described in Materials and Methods.

To determine if the CAF-I complex mediates CAC3 suppression of the RAS/cAMP pathway, perhaps by affecting the expression of one or more genes that alter RAS/cAMP signal transduction, we asked if CAC3 overexpression could suppress the RAS/cAMP pathway in the absence of functional CAF-I. We found that loss of CAC1 function did not affect the heat shock sensitivity of RAS2^G19V cells and CAC3 overexpression suppressed the heat shock sensitivity of cac1-1 RAS2^G19V cells (Fig. 2A). Similarly, deletion of CAC2 had no effect on the heat shock sensitivity of RAS2^G19V cells or on the ability of CAC3 overexpression to suppress the heat shock sensitivity (Fig. 2B). Thus, the ability of CAC3 to affect the RAS/cAMP pathway did not require the presence of active CAF-I and must be independent of CAF-I-mediated chromatin assembly function.

FIG. 2.

CAC1 and CAC2 are not required for CAC3 to suppress the RAS/cAMP pathway. The heat shock sensitivity of (A) cac1-1 (YJB1358), RAS2^G19V cac1-1 (YJB2237), RAS2^G19V cac1-1 CAC3-overexpressing (OE) (YJB2322) or (B) cac2Δ (YJB1599), RAS2^G19V cac2Δ (YJB3563), and RAS2^G19V cac2Δ CAC3-overexpressing (YJB3876) was determined as described for Fig. 1. (C) Cells carrying Cac3p-GFP (YJB4034) were stained with DAPI and analyzed by fluorescence microscopy to observe both GFP (top panel) and DNA (middle panel). The merged image (bottom panel) shows that Cac3p-GFP localizes throughout the cytoplasm and nucleus. Darker regions correspond to the vacuoles.

The ability of Cac3p to function independently of Cac1p and Cac2p is supported by the previous observations that the cell cycle regulation of CAC3 transcripts is different from that of CAC1 and CAC2 transcripts (39) and that Cac3p/p48 does not always copurify with Cac1p and Cac2p (28). Consistent with this, we found that Cac3p-GFP had a localization pattern different from that of Cac1p: Cac3p-GFP appeared to be diffuse and localized throughout the nucleus and cytoplasm (but not in the vacuoles) (Fig. 2C) during all stages of the cell cycle (S. Enomoto and J. Berman, unpublished data). In contrast, epitope-tagged Cac1p localized primarily to large foci within the nucleus, even when expressed from a high-copy-number vector (11).

CAC3 suppression of the RAS/cAMP pathway is independent of several histone-modifying activities.

Cac3p and other p48-related proteins are involved in several chromatin-related functions. For example, CAC3 has been shown to antagonize the Sin3p-Rpd3p histone deacetylation complex (42). Additionally, Cac3p has high sequence similarity to RbAp48p and RbAp46, which are subunits of histone modification enzymes such as histone deacetylase I (HDAC1) and histone acetylases (43, 49). Therefore, we asked if Cac3p affects the RAS/cAMP pathway through its association with histone deacetylases such as Rpd3p and Hda1p. The heat shock sensitivity of RAS2^G19V strains lacking RPD3 and/or HDA1 did not differ from that of otherwise wild-type RAS2^G19V strains (Table 2). Furthermore, there was no difference in the degree to which CAC3 overexpression suppressed the heat shock sensitivity of these strains (Table 2).

TABLE 2.

Effect of CAC3 overexpression on RAS/cAMP pathway activation in different mutant strains^a

Notation type	Strain	Relevant genotype	Heat shock response
RAS/cAMP pathway	YJB195	Wild type	Resistant
	YJB2235	RAS2G19V	Sensitive
	YJB2320	RAS2G19V YEp55-CAC3	Resistant
	YJB2697	bcy1Δ	Sensitive
	YJB2781	bcy1Δ YEp55-CAC3	Sensitive
Chromatin assembly	YJB2237	RAS2G19V cac1-1	Sensitive
	YJB2322	RAS2G19V cac1-1 YEp55-CAC3	Resistant
	YJB3563	RAS2G19V cac2Δ	Sensitive
	YJB3876	RAS2G19V cac2Δ YEp55-CAC3	Resistant
Histone acetyltransferases	YJB2350	RAS2G19V hat1Δ	Sensitive
	YJB2367	RAS2G19V hat1Δ YEp55-CAC3	Resistant
	YJB2648	RAS2G19V hat2Δ	Sensitive
	YJB2658	RAS2G19V hat2Δ YEp55-CAC3	Resistant
	YJB2672	RAS2G19V hat1Δ hat2Δ	Sensitive
	YJB2673	RAS2G19V hat1Δ hat2Δ YEp55-CAC3	Resistant
	YJB3135	RAS2G19V gcn5Δ	Sensitive
	YJB3141	RAS2G19V gcn5Δ YEp55-CAC3	Resistant
Histone deacetylases	YJB2475	RAS2G19V hda1Δ	Sensitive
	YJB2531	RAS2G19V hda1Δ YEp55-CAC3	Resistant
	YJB2476	RAS2G19V rpd3Δ	Sensitive
	YJB2530	RAS2G19V rpd3Δ YEp55-CAC3	Resistant
	YJB2474	RAS2G19V had1Δrpd3Δ	Sensitive
	YJB2532	RAS2G19V had1Δrpd3Δ YEp55-CAC3	Resistant
Histone transcription	YJB2998	RASG19V hir3Δ	Sensitive
	YJB2999	RAS2G19V hir3Δ YEp55-CAC3	Resistant
			Resistant
Silencing maintenance	YJB3763	RAS2G19V sir3Δ	Sensitive
	YJB3773	RAS2G19V sir3Δ YEp55-CAC3	Resistant
Cell signaling	YJB3670	RAS2G18A G19V sch9Δ	Sensitive
	YJB3674	RAS2G18A G19V sch9Δ YEp55-CAC3	Resistant
	YJB5245	RAS2G19V gpr1Δ	Sensitive
	YJB5237	RAS2G19V gpr1Δ YEp55-CAC3	Resistant
	YJB5243	RAS2G19V gpa2Δ	Sensitive
	YJB5235	RAS2G19V gpa2Δ YEp55-CAC3	Resistant

^aSensitive indicates that the heat shock response was indistinguishable from that of an isogenic RAS2^G19V strain (YJB2235, -2349, -2474, and -3016). Resistant indicates that the heat shock response was indistinguishable from that of an isogenic RAS2 strain (YJB195, -1583, -2431, and -2136). 

Based on the similarity between Cac3p and Hat2p, a factor that facilitates the association of Hat1p with histones H3 and H4 (31), we investigated the ability of CAC3 to suppress RAS2^G19V-induced heat shock sensitivity in strains lacking genes encoding histone acetyltransferase components. The heat shock sensitivity of RAS2^G19V strains carrying deletions in HAT1, HAT2, HAT1 and HAT2, or GCN5 was measured in the presence and absence of a plasmid overexpressing CAC3. In all of these strains, heat shock sensitivity and CAC3 suppression of this sensitivity were not significantly different from what was found in the isogenic wild-type RAS2^G19V strain (Table 2). In addition, we asked if deletion of HIR3, which encodes a histone transcriptional regulator that has synergistic effects with CAC mutants, or of SIR3, which encodes a component of silent chromatin, would affect RAS2^G19V-induced heat shock sensitivity in the presence and absence of CAC3 overexpression. Again, the heat shock sensitivity and the ability of CAC3 overexpression to suppress this heat shock sensitivity were not significantly different from the isogenic wild-type strain (Table 2). Thus, CAC3 overexpression does not appear to affect the RAS/cAMP pathway by functioning as a component of a histone modification complex.

CAC3 suppression of the RAS/cAMP pathway is independent of the GPA2/GPR1 signaling pathway.

A parallel pathway for the activation of PKA utilizes the membrane proteins Gpa2p and Gpr1p. Gpa2p is a G_α-like protein which activates adenylyl cyclase in response to extracellular glucose (44). Sch9p is a protein kinase that contributes to the heat shock response independently of the RAS/cAMP and GPA2/GPR1 pathways (27). Thevelein and de Winde (44) hypothesized the existence of a Saccharomyces protein containing WD40 repeat motifs which could function as a G_β-like protein, suppressing the function of Gpa2p. As CAC3 contains WD40 repeat motifs, we hypothesized that CAC3 overexpression might decrease the levels of intracellular cAMP by suppressing the activity of Gpa2p. To test this possibility, we asked if components of the GPA2 pathway were required for CAC3-mediated suppression of RAS2^G19V-induced heat shock sensitivity. We found that RAS2^G19V strains lacking GPA2 or GPR1 remained sensitive to heat shock and that CAC3 overexpression effectively suppressed the heat shock sensitivity of these strains (Table 2). Similarly, cells lacking functional SCH9 were also sensitive to heat shock, and this sensitivity was still suppressed by CAC3. Thus, signal transduction through the GPA2/GPR1 or SCH9 pathway is not required for suppression of RAS2^G19V by CAC3 overexpression and Cac3p is not acting as a G_β-like protein to directly suppress Gpa2p.

Identification of Npr1p as a Cac3p-interacting protein.

To identify factors that interact with Cac3p and that may be required for CAC3 suppression of the RAS/cAMP pathway, we isolated genes encoding proteins that interact with Cac3p using the yeast two-hybrid system (6). Cac3p was fused to the Gal4p DNA-binding domain and used to screen a library of S. cerevisiae genes fused to the Gal4p activation domain. One gene identified in this screen, CAC1, was subsequently used as a positive control in the screen. Further screening identified a clone containing the codons for amino acids 561 to 605 of NPR1. NPR1 encodes a nitrogen permease reactivator, a putative serine/threonine protein kinase required to regulate the posttranslational stability of several permeases, including Mep2p (26), Gap1p (41), and Tat2p (36). Although the Cac3p-Npr1p interaction is weaker than the Cac3p-Cac1p interaction, it was consistently and reproducibly detected. To biochemically confirm this protein-protein interaction, we constructed strains expressing T7 epitope-tagged Cac3p and either HA-tagged Npr1p or an HA-tagged, kinase-dead version of Npr1p. We immunoprecipitated HA-tagged Npr1 using anti-HA antibodies under nondenaturing conditions, fractionated the immunoprecipitates, and then probed for Npr1p and Cac3p by immunoblotting. As shown in Fig. 3B, Cac3p was coimmunoprecipitated with the putative kinase-dead version of Npr1p, although not with the wild-type version of the protein. This confirms the interaction between Npr1p and Cac3p detected by two-hybrid analysis but suggests that the interaction with the wild-type protein may be relatively transient. Loss of Npr1p kinase activity evidently stabilizes this transient interaction. To the best of our knowledge, neither interaction between Npr1p and the CAF-I complex (or other aspects of chromatin metabolism) nor any connection between Npr1p and the RAS/cAMP pathway has been reported previously.

FIG. 3.

Cac3p interacts with Npr1p in a two-hybrid assay. (A) Gal4_DBD-Cac3p/Gal4_AD (YJB3523), Gal4_DBD-Cac3p/Gal4_AD-Cac1p (YJB3197), and Gal4_DBD-Cac3p/Gal4_AD-Npr1p_561–605 (YJB3438) strains carrying the HIS3 gene under control of the GAL1 promoter were plated on SDC lacking Leu, Trp, and His and with 5 mM 3AT. Growth indicates a positive interaction between the Gal4 DNA-binding domain (DBD) and Gal4 activation domain (AD) fusion proteins. (B) Cac3p coimmunoprecipitates with Npr1p. Yeast strains overexpressing T7 epitope-tagged Cac3p either without (lane —) or with the cooverexpression of HA epitope-tagged wild-type Npr1p (WT) or kinase-dead Npr1p (KD) were immunoprecipitated under native conditions with anti-HA antiserum. Levels of Npr1p and Cac3p were determined from the input, supernatants (sup), and pellets by immunoblotting with either anti-HA or anti-T7 antiserum. Lane MW, size markers.

Deletion of NPR1 yields the same phenotype as does CAC3 overexpression.

To determine if the interaction of Cac3p with Npr1p was responsible for suppression of the RAS/cAMP pathway, NPR1 was deleted in a wild-type or RAS2^G19V background, and the resulting strains were assayed for heat shock sensitivity. Deletion of NPR1 had no affect on the heat shock sensitivity of a wild-type strain but fully suppressed the heat shock sensitivity induced by the RAS2^G19V allele (Fig. 4A). In contrast, deletion of NPR1, similar to overexpression of CAC3, had no effect on the heat shock sensitivity of a bcy1Δ strain (Fig. 4B).

FIG. 4.

Deletion of NPR1 suppresses the RAS/cAMP pathway between RAS and PKA. The heat shock resistance of (A) wild-type (YJB195), npr1Δ (YJB3712), RAS2^G19V (YJB2235), RAS2^G19V npr1Δ (YJB2723), and RAS2^G19V npr1Δ CAC3-overexpressing (OE) (YJB4017), and (B) bcy1Δ (YJB2697) and bcy1Δ npr1Δ (YJB3554) strains was determined. (C) The heat shock resistance of isogenic wild-type (YJB195), RAS2^G19V (YJB2235), and RAS2^G19V strains carrying high-copy CAC3 (YJB2320), NPR1 (YJB3863) and both CAC3 and NPR1 (YJB3875) was determined.

Activation of the RAS/cAMP pathway causes a sporulation defect (21), which can be suppressed by overexpression of CAC3 (35) (Table 3). Similar to CAC3 overexpression, deletion of both copies of NPR1 suppressed the sporulation defect caused by the RAS2^G19V allele (Table 3). The fact that CAC3 overexpression or NPR1 deletion can suppress two distinctly different phenotypes of an activated RAS/cAMP pathway indicates that this effect is specific to this pathway and does not reflect a generalized increase in heat tolerance. Thus, the phenotypes of strains lacking NPR1 are indistinguishable from the phenotypes of strains overexpressing CAC3: both mutations suppress the RAS/cAMP signal transduction pathway when it is activated by a RAS2^G19V allele but not by BCY1 deletion.

TABLE 3.

Deletion of NPR1, like overexpression of CAC3, suppresses the sporulation defect of RAS2^G19V strains

Strain	Relevant genotype	Sporulation efficiency (%)
YJB334	Wild type	14.9
YJB3737	RAS2G19V	4.8
YJB3792	npr1Δ/npr1Δ	23.0
YJB3791	RAS2G19V npr1Δ/npr1Δ	18.1
YJB3739	YEp55-CAC3	47.2
YJB3721	RAS2G19V YEp55-CAC3	34.0
YJB5523	RAS2G19V pCUP1-myc-UBI4	18.1

These data suggest a model in which excess Cac3p interacts with Npr1p, sequestering the kinase so that it is not available for its role in the RAS/cAMP pathway. The sequestration model predicts that (i) CAC3 overexpression in an npr1Δ strain would confer no additional resistance to heat shock, (ii) the ability of an npr1Δ mutation to suppress RAS2^G19V phenotypes would not depend on the presence of a functional CAC3 gene, and (iii) elevated levels of NPR1 would provide excess copies of Npr1p and negate the effect of CAC3 overexpression. Consistent with the first prediction, no additional heat shock resistance was observed when CAC3 was overexpressed in an npr1Δ strain (Fig. 4A), which is consistent with the idea that CAC3 overexpression and NPR1 deletion are suppressing the RAS/cAMP pathway by a common mechanism. Consistent with the second prediction, RAS2^G19V cells carrying deletions of both NPR1 and CAC3 remained heat shock resistant like the RAS2^G19V npr1Δ cells (data not shown). Consistent with the third prediction, RAS2^G19V cells remained sensitive to heat shock when both CAC3 and NPR1 were overexpressed from the GAL1 promoter (Fig. 4C).

In addition, the sequestration model predicts that if excess Cac3p sequesters Npr1p, then the phenotypes of cells overexpressing CAC3 should resemble the known phenotypes of cells lacking NPR1. Npr1p is required for the stable expression of the Mep2p ammonium transporter, which is necessary for pseudohyphal growth in diploid strains in the Σ1278b strain background. Homozygous deletion of NPR1 in the diploid Σ1278b strain background results in cells that produce fewer pseudohyphae (Fig. 5A, bottom panels), apparently due to the loss of Mep2p function (26). As predicted by the sequestration model, overexpression of CAC3 in a diploid Σ1278b strain, like deletion of NPR1 in these strains, greatly reduced the number and length of pseudohyphae formed (Fig. 5A, upper panels). Thus, CAC3 overexpression and NPR1 deletion both result in identical phenotypes for suppression of the RAS/cAMP pathway and for pseudohyphal growth.

FIG. 5.

CAC3 and NPR1 are mutally antagonistic. (A) CAC3 overexpression suppresses pseudohyphal growth. Σ1278B strains carrying either pGalSet (YJB3514) or pGalSet-CAC3 (YJB3515) were grown in either galactose (to induce CAC3 expression) or glucose (to repress CAC3 expression). The cells were then plated on either SLAD or SLADG medium to induce pseudohyphal growth either with or without CAC3 expression. Representative colonies were photographed after 3 days of growth at 30°C. For comparison, wild-type (YJB5724) and npr1/npr1 (YJB5723) Σ1278B strains are also shown. (B) NPR1 overexpression decreases telomeric silencing. Serial dilutions (1:10) of wild-type (YJB3610), cac3 (YJB1786), and pGalSET-NPR1 (YSJ328) strains were plated on either complete medium or medium with 5-fluoroorotic acid (FOA). Yeast cells were grown in medium containing either glucose (gluc) or galactose (gal), as indicated. Plates were photographed after 2 days of growth at 30°C.

Furthermore, the sequestration model predicts that overexpression of NPR1 should produce the same phenotypes as deletion of CAC3. Kaufman et al. (24) showed that deletion of CAC3 leads to a decrease in transcriptional silencing of a URA3 reporter gene at the telomere on chromosome VII by about 25-fold. The NPR1 gene was placed on a plasmid under control of the GAL1, 10 promoter to be induced in the presence of galactose and repressed in the presence of glucose. NPR1-overexpressing cells, like cac3Δ cells, have a modest decrease in telomeric silencing of approximately 16-fold (Fig. 5B). Taken together, these data are consistent with our proposed model that excess Cac3p binds to Npr1p and sequesters the kinase, resulting in a set of phenotypes that are indistinguishable from those caused by npr1Δ.

Npr1p does not affect the phosphorylation state or localization of Ras2p.

Ras2p is a membrane-associated phosphoprotein whose activity is increased by phosphorylation of Ser-214 (50) by an unknown kinase. Npr1p is a putative serine/threonine kinase that presumably phosphorylates several membrane proteins (26, 36, 41). To test the hypothesis that Npr1p may be the kinase that phosphorylates Ras2p, we metabolically labeled wild-type and npr1Δ cells with ³²P, immunoprecipitated Ras2p, performed SDS-PAGE, and detected phospho-Ras2p by autoradiography. The amount and electrophoretic mobility of phospho-Ras2p were not affected by deletion of NPR1 (Fig. 6A), indicating that Npr1p does not affect the degree of Ras2p phosphorylation either directly or indirectly. Immunoblot experiments did not detect any change in the quantity of total Ras2p in RAS2^G19V cells overexpressing Cac3p or missing Npr1p (data not shown). Furthermore, recombinant Npr1p did not phosphorylate recombinant human Ha-Ras (data not shown). Thus, it appears unlikely that Npr1p modulates the RAS/cAMP pathway by affecting the phosphorylation of Ras. We cannot rule out the formal possibility that Npr1p is one of several kinases that can phosphorylate Ras in vivo, although this scenario cannot explain how loss of Npr1p function alone could suppress the RAS/cAMP pathway while not causing any detectable change in the degree of Ras phosphorylation.

FIG. 6.

Npr1p does not affect the phosphorylation or localization of Ras2p. (A) Wild-type (WT) (YJB195) and npr1Δ (YJB3712) cells were grown in the presence of [³²P]orthophosphate. Lysates were immunoprecipitated (IP) with protein A-agarose (mock) or anti-Ras2p conjugated to agarose. Bound proteins were eluted by denaturation, separated by SDS-PAGE, and detected by autoradiography. The arrow indicates phospho-Ras2p. (B) Localization of GFP-Ras2p was determined by fluorescence microscopy in wild-type (YJB3896) and npr1Δ (YJB3897) strains and cells expressing extra copies of CAC3 (YJB3898).

In order to function correctly, Ras2p must localize to the plasma membrane. Mislocalization of Ras2p decreases its signaling activity and effectively suppresses the constitutively active RAS2^G19V allele (2, 4). Thus, we asked if deletion of NPR1 or overexpression of CAC3 altered the membrane localization of Ras2p. When the RAS2 product is fused to GFP, the plasma membranes of yeast cells fluoresce, indicating that Ras2p is located at the membrane (4) (Fig. 6B). Deletion of NPR1 or overexpression of CAC3 did not alter the peripheral localization pattern of Ras2p-GFP (Fig. 6B), suggesting that suppression of the RAS2^G19V allele by excess Cac3p is not due to an alteration in the subcellular distribution of Ras2p.

CAC3 overexpression and npr1Δ lead to indistinguishable phenotypes, leading us to propose a model in which Cac3p binds and sequesters Npr1p. Another possibility is that overexpression of CAC3 decreases the quantity of Npr1p, perhaps by targeting Npr1p for destruction. To test this possibility, we used a series of yeast strains containing an epitope-tagged NPR1 gene (36). The overexpression of CAC3 in either the presence or absence of the RAS2^G19V allele had no effect on the amount of Npr1p present in the cell (data not shown). Thus, consistent with the sequestration model, excess copies of Cac3p did not destabilize Npr1p, yet still caused the same phenotypes as in an npr1Δ strain.

Ubiquitin overexpression can suppress the RAS/cAMP pathway.

Npr1p is thought to stabilize Mep2p and other permeases by phosphorylating the protein, which then reduces the protein's ubiquitination and subsequent inactivation (26, 41). We hypothesized that, in a similar manner, Npr1p may stabilize another substrate protein that is important for the full function of the RAS/cAMP pathway by inhibiting its ubiquitination. Thus, deletion of NPR1 would be expected to increase the ubiquitination of putative substrate proteins. Furthermore, if this hypothesis is correct, overexpression of UBI4, the gene encoding polyubiquitin, would be expected to increase the ubiquitination of this substrate protein. Consistent with this expectation, overexpression of Myc epitope-tagged UBI4 from the copper-inducible CUP1 promoter on a high-copy plasmid suppressed RAS2^G19V-induced heat shock sensitivity (Fig. 7). Additionally, overexpression of UBI4 in a RAS2^G19V npr1Δ strain conferred no additional heat shock resistance, suggesting that these two genes suppress the RAS/cAMP signal transduction pathway by a common mechanism. Furthermore, overexpression of UBI4 suppressed the sporulation defect of RAS2^G19V cells (Table 3). Thus, overexpression of polyubiquitin suppressed activated RAS2 phenotypes in a manner indistinguishable from either CAC3 overexpression or NPR1 deletion. This result is consistent with the hypothesis that NPR1 and ubiquitin have antagonistic roles in the suppression of the RAS/cAMP pathway.

FIG. 7.

Overexpression of polyubiquitin (UBI4) suppresses the RAS2^G19V-induced heat shock sensitivity. The heat shock resistance of RAS2^G19V (YJB2235), RAS2^G19V UBI4-overexpressing (OE) (YJB27334), RAS2^G19V npr1Δ (YJB3723), and RAS2^G19V npr1 UBI4-overexpressing (YJB5522) strains in the presence and absence of NPR1 and UB14 overexpression were determined as for Fig. 1.

DISCUSSION

Cac3p has at least two separable functions.

Cac3p is the smallest subunit of CAF-I, which assemble histones H3 and H4 tetramers onto newly replicated DNA (24, 38). In addition, excess Cac3p suppresses the activated RAS/cAMP pathway (35) (Fig. 1) in a CAF-I independent manner (Fig. 2), a result consistent with a recent report (51). Because CAC3 acts antagonistically with the Sin3p-Rpd3p complex (42) and the mammalian Cac3p homolog (RbAp48) is associated with chromatin-modifying enzymes such as histone deacetylases and histone acetylases (43, 49), we asked if the CAC3 suppression of RAS2^G19V required histone-modifying activities such as the histone deacetylases Rpd3p and Hda1p or the histone acetyltransferases Hat1p, Hat2p, and Gcn5p. Interestingly, we found that CAC3 suppression of RAS2^G19V was not dependent upon any of the histone deacetylases or histone acetylases tested, nor was it dependent upon the histone regulator Hir3p or the silent chromatin component Sir3p (Table 2). Furthermore, several histone deacetylase complexes, including those containing Rpd3p and Hda1p, have been characterized in biochemical fractionation studies, and Cac3p has not been isolated as a component of these complexes (M. Grunstein, personal communication). Therefore, we conclude that the role of Cac3p in the RAS/cAMP signal transduction pathway is distinct and separable from the role of Cac3p in CAF-I-dependent chromatin assembly and histone modification. Consistent with this conclusion, CAC3 mRNA does not share the cell cycle-dependent expression pattern identified for CAC1 and CAC2 (39), not all Cac3p/p48 copurifies with the CAF-I complex (28, 32, 43), and the subcellular localization of Cac3p is distinct from that of Cac1p (Fig. 2C).

Cac3p suppression of the RAS/cAMP pathway is mediated by Npr1p.

A C-terminal fragment of Npr1p, which includes a portion of the kinase domain, interacted with Cac3p in a yeast two-hybrid screen and biochemically (Fig. 3). Consistent with the hypothesis that Cac3p overexpression reduces the effective activity of Npr1p, the suppression phenotypes of npr1Δ strains (e.g., heat shock resistance, sporulation, and loss of pseudohyphal growth) were indistinguishable from the phenotypes of strains overexpressing CAC3 (Fig. 4 and 5). Both mutations were capable of suppressing two different phenotypes caused by the RAS2^G19V allele but not the same phenotypes resulting from deletion of the negative regulator BCY1. This demonstrates that overexpression of CAC3 and deletion of NPR1 suppress the pathway in an indistinguishable manner, suggesting a common suppression mechanism which is dependent on the RAS/cAMP pathway. Furthermore, overexpression of CAC3 in an npr1Δ strain did not enhance the suppression phenotype, supporting the idea that Cac3p affects the RAS/cAMP pathway by reducing the level of active, available Npr1p. Finally, cooverexpression of NPR1 blocked the ability of CAC3 overexpression to reduce the heat shock sensitivity of a RAS2^G19V strain (Fig. 4C), supporting the model that excess Cac3p binds and sequesters Npr1p.

NPR1 encodes a nitrogen permease reactivator, a putative serine/threonine kinase that affects the activity of several nutrient transporters (13, 14), including Gap1p, the general amino acid permease (46); Mep2p, an ammonium permease (26); Pcp1p, a spermidine transporter (19); and Tat2p, the tryptophan transporter (36). NPI1/RSP5, encoding a ubiquitin-protein ligase, antagonizes the activity of NPR1 in many cases (14, 17). The currrent model postulates that phosphorylation of a transporter by Npr1p affects ubiquitination and subsequent proteolysis of that transporter, stabilizing nutrient-repressible permeases such as Gap1p (40, 41) but promoting degradation of constitutive permeases such as Tat2p (36). In its interaction with the RAS/cAMP pathway, NPR1 acts antagonistically with ubiquitin. We found that overexpression of polyubiquitin suppressed RAS2^G19V-induced heat shock sensitivity in a manner that was indistinguishable from the suppression observed in npr1Δ strains (Fig. 7). The target of rapamycin (TOR) nutrient signaling pathway leads to the phosphorylation and subsequent inhibition of Npr1p (36). For more than 15 years, we have known that NPR1 is a key regulator of nitrogen metabolism (14) and that RAS/cAMP pathway is the principal regulator of carbon metabolism in Saccharomyces (5). This work is the first reported example of cross-talk between these two metabolic regulatory pathways. We have demonstrated that Npr1p affects the activity of the RAS/cAMP signal transduction pathway, providing a heretofore unrecognized connection between the carbon and nitrogen signaling pathways.

The precise mechanism by which Npr1p affects the RAS/cAMP pathway remains unknown. Neither the TOR1-1 or TOR2-1 mutation nor treatment with rapamycin had any effect on the ability of either CAC3 overexpression or NPR1 deletion to suppress the RAS2^G19V phenotype (data not shown). This indicates that the TOR-dependent phosphorylation state of Npr1p does not affect the role of Npr1p in RAS/cAMP signaling. Since Npr1p affects the stability of a number of proteins and since overexpression of polyubiquitin yields the same phenotype as loss of Npr1p, we surmise that Npr1p activates or potentiates the RAS/cAMP pathway by stabilizing one or more intermediates in the pathway. Furthermore, the fact that overexpression of CAC3 or deletion of NPR1 suppresses RAS2^G19V-induced phenotypes but not the same phenotypes resulting from deletion of BCY1 suggests that Npr1p acts between Ras-induced synthesis of cAMP and cAMP-mediated activation of the A kinase. Deletion of NPR1 does not affect the phosphorylation, ubiquitination, localization, or abundance of Ras2p (Fig. 6 and data not shown), nor does deletion of NPR1 affect the levels of Cdc25p, the guanine nucleotide exchanger for Ras (L. Schneper and J. R. Broach, unpublished observations).

Zhu and colleagues (51) recently reported that CAC3 overexpression suppressed the RAS/cAMP pathway when that pathway was activated by deletion of PDE1 and PDE2 or by an activated TPK2 mutation. However, CAC3 overexpression did not affect the total level of extractable PKA kinase activity. Their data indicated that CAC3 suppressed the RAS/cAMP pathway in a CAC1-independent and BCY1-dependent fashion which is not fully understood (51). The results presented here support and extend their conclusions by indicating that Npr1p is the target of Cac3p that modulates the RAS/cAMP signal transduction pathway.