Functions of Saccharomyces cerevisiae 14-3-3 Proteins in Response to DNA Damage and to DNA Replication Stress

Corresponding author: Maria Pia Longhese, Università di Milano-Bicocca, Piazza della Scienza 2, 20126 Milano, Italy., mariapia.longhese{at}unimib.it (E-mail)

Communicating editor: M. ROSE

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

Two members of the 14-3-3 protein family, involved in key biological processes in different eukaryotes, are encoded by the functionally redundant Saccharomyces cerevisiae BMH1 and BMH2 genes. We produced and characterized 12 independent bmh1 mutant alleles, whose presence in the cell as the sole 14-3-3 source causes hypersensitivity to genotoxic agents, indicating that Bmh proteins are required for proper response to DNA damage. In particular, the bmh1-103 and bmh1-266 mutant alleles cause defects in G₁/S and G₂/M DNA damage checkpoints, whereas only the G₂/M checkpoint is altered by the bmh1-169 and bmh1-221 alleles. Impaired checkpoint responses correlate with the inability to maintain phosphorylated forms of Rad53 and/or Chk1, suggesting that Bmh proteins might regulate phosphorylation/dephosphorylation of these checkpoint kinases. Moreover, several bmh1 bmh2 mutants are defective in resuming DNA replication after transient deoxynucleotide depletion, and all display synthetic effects when also carrying mutations affecting the pol-primase and RPA DNA replication complexes, suggesting a role for Bmh proteins in DNA replication stress response. Finally, the bmh1-169 bmh2 and bmh1-170 bmh2 mutants show increased rates of spontaneous gross chromosomal rearrangements, indicating that Bmh proteins are required to suppress genome instability.

UNREPAIRED or inaccurately repaired DNA damage can lead to mutations, genomic instability, and ultimately cancer in multicellular organisms. Eukaryotic cells have evolved protective signal-transduction pathways, known as DNA damage checkpoints, which are specialized in detecting abnormal DNA structures and in coordinating cell cycle progression with DNA repair. Their activation delays the G₁/S and G₂/M transitions and slows down DNA synthesis when DNA is damaged in G₁, G₂, or S phases, respectively, thus preventing replication or segregation of damaged DNA molecules (reviewed in LONGHESE et al. 1998 ; ZHOU and ELLEDGE 2000 ; NYBERG et al. 2002 ). Central to the checkpoint-mediated response to DNA damage and to incomplete DNA replication are components of a highly conserved phosphatidylinositol protein kinase family, among which are Saccharomyces cerevisiae Mec1 (WEINERT et al. 1994 ) and Tel1 (GREENWELL et al. 1995 ; MORROW et al. 1995 ), Schizosaccharomyces pombe Rad3 (BENTLEY et al. 1996 ), Drosophila melanogaster Mei-41 (HARI et al. 1995 ), and human ATM (SAVITSKY et al. 1995 ) and ATR (BENTLEY et al. 1996 ). S. cerevisiae Mec1, as well as human ATM and ATR and S. pombe Rad3, is required to phosphorylate different substrates in response to DNA insults, and several data suggest a pivotal role for these proteins in both sensing and transducing the checkpoint signal (reviewed in ZHOU and ELLEDGE 2000 ; NYBERG et al. 2002 ). Mec1 functions are supported by its interacting factor Ddc2 (also called Lcd1 or Pie1; PACIOTTI et al. 2000 ; ROUSE and JACKSON 2000 ; WAKAYAMA et al. 2001 ), functionally related to Rad26 and ATRIP, which interact with Rad3 and ATR in S. pombe and human cells, respectively (EDWARDS et al. 1999 ; CORTEZ et al. 2001 ). Other checkpoint factors, such as the PCNA-like complex Mec3/Ddc1/Rad17 and the RFC-like complex Rad24/Rfc2-5, participate in the DNA-damage-sensing step and are thought to modulate Mec1 activation (reviewed in NYBERG et al. 2002 ).

Once DNA perturbations are sensed, checkpoint signals are propagated through the protein kinases Rad53 and Chk1, which undergo Mec1-dependent phosphorylation in response to DNA damage (SANCHEZ et al. 1996 , SANCHEZ et al. 1999 ). While Rad53 is required for cell cycle arrest after genotoxic treatments in all the cell cycle phases, Chk1 contributes only to G₂/M checkpoint activation by a Rad53-independent pathway (SANCHEZ et al. 1999 ). The DNA-damage-sensing functions are linked with the downstream effectors by Mec1-dependent phosphorylation of Rad9 (EMILI 1998 ; SUN et al. 1998 ; VIALARD et al. 1998 ), which triggers Rad9 interaction with Rad53 and consequent release of active Rad53 kinase (GILBERT et al. 2001 ).

How the checkpoint kinases are targeted to their substrates is presently unknown. A possible role in this process has been envisaged for proteins of the 14-3-3 family, highly conserved from yeast to mammals, which regulate cellular activities by binding and sequestering phosphorylated proteins (reviewed in FU et al. 2000 ; TZIVION and AVRUCH 2002 ). The 14-3-3 proteins participate in the regulation of diverse biological processes, including signal transduction, cell cycle regulation, apoptosis, stress response, cytoskeleton organization, and malignant transformation (reviewed in FU et al. 2000 ; VAN HEMERT et al. 2001 ; TZIVION and AVRUCH 2002 ). There are at least 7 distinct 14-3-3 genes in vertebrates, >10 in plants, and 2 in S. cerevisiae, S. pombe, D. melanogaster, and Caenorhabditis elegans (reviewed in FU et al. 2000 ). Although their roles are still obscure, they appear to be involved in the G₂/M DNA damage checkpoint in both S. pombe and human cells, perhaps by targeting checkpoint regulators to specific effectors. In fact, the 14-3-3- protein is required to prevent mitotic catastrophe after DNA damage in human cells (CHAN et al. 1999 ). Moreover, the lack of the Rad24 14-3-3 isoform causes mild sensitivity to DNA-damaging agents and a defective G₂/M DNA damage checkpoint in S. pombe (FORD et al. 1994 ). Both 14-3-3 fission yeast isoforms, Rad24 and Rad25, physically interact with the checkpoint effector kinase Chk1 (reviewed in CARR 1997 ). This association is stimulated by DNA damage (CHEN et al. 1999 ), suggesting that Chk1/14-3-3 association may direct the kinase to particular substrates. The DNA damage checkpoint arrests both S. pombe and mammalian cells in G₂ through inhibitory phosphorylation of the cyclin-dependent kinase Cdc2/Cdk1 on tyrosine-15 by members of the Wee1/Mik1 tyrosine kinase family (reviewed in RUSSELL 1998 ). Phosphorylated tyrosine-15 is in turn dephosphorylated by the Cdc25 phosphatase, and both Cdc25 and Wee1 have been shown to be potential in vivo targets of Chk1 (reviewed in MOSER and RUSSELL 2000 ). Although it is not clear whether Wee1 regulation is important for checkpoint function in vivo, experiments on Xenopus extracts demonstrate that Chk1 is responsible for Wee1 phosphorylation at serine 549 and that 14-3-3 binds this Wee1 phosphorylated form, thus enhancing its inhibitory kinase activity toward Cdc2 (O'CONNELL et al. 1997 ; LEE et al. 2001 ). Moreover, Chk1 can phosphorylate mammalian Cdc25C on serine 216 in vitro (PENG et al. 1997 ; ZENG et al. 1998 ). This phosphorylation event allows 14-3-3 binding, which in turn renders Cdc25 either sequestered in the cytoplasm or catalytically less active in both human and S. pombe cells (FURNARI et al. 1997 ; PENG et al. 1997 ; BLASINA et al. 1999 ; DALAL et al. 1999 ; KUMAGAI and DUNPHY 1999 ; LOPEZ-GIRONA et al. 1999 , LOPEZ-GIRONA et al. 2001 ; YANG et al. 1999 ). Thus, although the precise role of 14-3-3 proteins in the DNA damage checkpoint is still undefined, altogether these data suggest that DNA-damage-induced Chk1/14-3-3 interaction may regulate Chk1 catalytic activity and/or its phosphorylation state.

In S. cerevisiae, the two members of the 14-3-3 protein family, which share 93% amino acid identity, are encoded by the BMH1 and BMH2 genes. Both genes are required for Ras/mitogen-activated protein kinase (MAPK) signaling during pseudohyphal development, but are not essential for the mating MAPK pathway (ROBERTS et al. 1997 ). They are also involved in the Ras/protein kinase A (PKA) signaling, since overexpression of TPK1, encoding the catalytic subunit of cAMP-dependent PKA, is able to partially suppress cell lethality caused by Bmh depletion (GELPERIN et al. 1995 ). Moreover, high Bmh levels counteract defects in the GDP/GTP exchange protein Cdc25 (GELPERIN et al. 1995 ). Despite these findings, the fact that BMH1 and BMH2 appear to be functionally interchangeable, while their simultaneous disruption is lethal in most S. cerevisiae backgrounds (ROBERTS et al. 1997 ), has until now limited the in vivo functional characterization of Bmh1 and Bmh2.

In this study, we show that BMH1 overexpression impairs DNA damage checkpoint response and describe a variety of bmh1 mutant alleles that confer temperature sensitivity and/or hypersensitivity to genotoxic agents in strains lacking other Bmh sources. Characterization of these alleles allows us to establish that the S. cerevisiae 14-3-3 proteins function in both DNA damage response and DNA metabolism.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

Plasmids:
Plasmid pML35.20, carrying a 2630-bp yeast DNA fragment containing the BMH1 gene, was found as a suppressor of the pri1-2 pip1 synthetic lethal phenotype (LONGHESE et al. 1996 ) in a yeast genomic DNA library constructed in plasmid YCplac111 (CVRCKOVA and NASMYTH 1993 ).

To obtain plasmids pML309 and pML48, containing the 801-bp BMH1 open reading frame (ORF), flanked by 429 bp upstream and 499 bp downstream, the 1732-bp BamHI-NcoI fragment from plasmid pML35.20 was cloned into the BamHI-SmaI sites of plasmids YCplac33 and YCplac111 (GIETZ and SUGINO 1988 ), respectively. To construct plasmid pML319, carrying the GAL1-BMH1 fusion, a BMH1 BamHI-PstI fragment was obtained by PCR amplification using plasmid pML309 as a template and oligonucleotides PRP247 (5'-CGC GGA TCC ATA TGT CAA CCA GTC GTG AAG ATT CT-3') and PRP248 (5'-AAC TGC AGC GCG TTT GAA TGA AGC AGC AAG CAT TG-3') as primers and then cloned into the BamHI-PstI sites of plasmid pML95 (LONGHESE et al. 1997 ).

Yeast strains and media:
The relevant genotypes of all the yeast strains used in this study are listed in Table 1. All the strains but YLL839 (CLERICI et al. 2001 ) were constructed during this study and all were derivatives of W303 (MATa or MAT, ade2-1, can1-100, his3-11, -15, leu2-3, -112, trp1-1, ura3), with the exception of strains YLL1238, YLL1290, YLL1295, and YLL1296, which were generated from strain RDKY3615 (MATa, ura3-52, leu21, trp163, his3200, lys2Bgl, hom3-10, ade21 ade8), kindly provided by R. Kolodner (La Jolla, CA; see below).

To generate the BMH1 chromosomal deletion, a bmh1::HIS3 cassette was constructed by PCR using plasmid pfl39 (STRUHL and DAVIS 1980 ) as template and oligonucleotides PR1-BMH1 (5'-GTC AAC CAG TCG TGA AGA TTC TGT GTA CCT AGC CAA GTT GGC TGA ACT CTT GGC CTC CTC TAG-3') and PR2-BMH1 (5'-CAG AAT ACT TAC TTT GGT GCT TCA CTT CGG CGG CAG CAG GTG GCT GTC GTT CAG AAT GAC ACG-3') as primers, followed by transformation of strain K699, giving rise to strain YLL138, where 716 bp of the BMH1 coding region was replaced by the HIS3 gene. Similarly, a bmh2::KANMX4 cassette was constructed by PCR using the pFA6a-KANMX4 plasmid (WACH et al. 1994 ) as template and oligonucleotides PRP230 (5'-CAA CAA AAA GTA CCC GTT ACA ACA AAA AAA ATG TCC CAA ACC GTA CGC TGC AGG TCG AC-3') and PRP231 (5'-TTG TAT TTC TCA GCG CTC TTA TTT GGT TGG TTC ACC TTG AAT CGA TGA ATT CGA GCT CG-3') as primers, followed by transformation of strain K699, giving rise to strain YLL908, where 789 bp of the BMH2 coding region was replaced by the KANMX4 gene. Strain DMP3566/1B was a meiotic segregant from a cross between strains YLL908 and K700. Strain YLL921 was obtained by transformation of strain YLL138 with plasmid pML309. The bmh1 bmh2 strain DMP3593/7C, kept alive by the centromeric plasmid pML309 (BMH1 URA3), was a meiotic segregant from a cross between strains YLL921 and DMP3566/1B. Strain YLL962.2, carrying two copies of the GAL1-BMH1 fusion at the LEU2 locus, was obtained by transforming strain K699 with EcoRV-digested plasmid pML319.

Strain DMP3444/4D was a meiotic segregant from a cross between strains YLL839 and K700. Strains DMP4058/9A, DMP3817/1A, DMP3818/5C, DMP3820/6A, DMP3821/1B, and DMP3855/2D were meiotic segregants from crosses between strain DMP3444/4D and strains YLL138, YLL908, YLL1082, YLL1080, YLL1081, and YLL1120, respectively.

Strains YLL1238 and YLL1290 were derivatives of strain RDKY3615 (MYUNG et al. 2001 ), where deletions of the BMH1 and BMH2 genes were performed as described above. Strains carrying the bmh2 allele and different bmh1 alleles were obtained by transforming strain YLL1238 with PmlI-digested integrative plasmids carrying the bmh1 alleles (see next paragraph), followed by deletion of the BMH2 gene.

The accuracy of all gene replacements and integrations was verified by Southern blot analysis or PCR. Standard yeast genetic techniques and media were according to ROSE et al. 1990 . Cells were grown in YEP medium (1% yeast extract, 2% bactopeptone, 50 mg/liter adenine) supplemented with 2% glucose (YEPD) or 2% raffinose (YEP+raf) or 2% raffinose and 1% galactose (YEP+raf+gal). Transformants carrying the KANMX4 cassette were selected on YEPD plates containing 400 µg/ml G418 (U.S. Biological). Selective plates for gross chromosomal rearrangements (GCR) were made by adding 5-fluoroorotic acid (5-FOA) and canavanine (Can) to final concentrations of 60 µg/ml and 1 mg/ml, respectively, to SC medium without arginine (SC-arg).

Mutagenesis of the BMH1 gene and search for bmh1 mutants:
Primers PRP237 (5'-CGT GTG TAC ATG CAC ATG TTA ATG-3') and PRP240 (5'-TTT TGC GGA AGC TAC TTT ATT CCG-3') were used to amplify by PCR under mutagenic conditions a BMH1 region spanning from 335 bp upstream of the translation start codon to 116 bp downstream of the stop codon, using plasmid pML35.20 as the template. Fifty independent PCR reaction mixtures (50 µl) were prepared, each containing 1.25 units of Taq DNA polymerase, 10 ng of template DNA (pML35.20), 250 ng of each primer, 500 µM each of deoxynucleoside triphosphate (dNTP: dATP, dTTP, dCTP), 100 µM dGTP, 500 µM MnCl2, 10 mM ß-mercaptoethanol, 10 mM Tris-HCl (pH 9), 50 mM KCl, and 1.5 mM MgCl₂. The bmh1 bmh2 strain DMP3593/7C, kept alive by wild-type BMH1 on the URA3 centromeric plasmid pML309, was cotransformed with the PCR products and the MscI-MscI fragment of plasmid pML48 (YCplac111 CEN4 LEU2 BMH1), carrying a 485-bp gap in the BMH1 ORF from +104 to +589 with respect to the first ATG codon to obtain reconstruction of the whole BMH1 coding region on the plasmid by gap repair with the PCR products. Transformants were selected on SC-plates lacking both uracil and leucine and assayed for the ability to lose the URA3 BMH1 pML309 plasmid after growth under nonselective conditions at 25°. A total of 3550 Leu⁺ Ura^- clones, derived from 4700 independent transformants and possibly containing bmh1 non-null mutant alleles on the pML48 derivative plasmids, were then assayed by drop tests for the ability to grow on YEPD plates after UV irradiation (60 J/m²) or in the presence of the alkylating agent methyl methanesulfonate (MMS; 0.01%) or the DNA synthesis inhibitor hydroxyurea (HU; 100 mM) at 25° and on YEPD plates at 37°.

Of the 147 putative bmh1 mutant alleles identified by the above procedure, 12 were studied further. To obtain stable bmh1 mutants, the 1745-bp BamHI-EcoRI fragments from the pML48 derivative plasmids pFM42, pFM98, pFM103, pFM127, pFM167, pFM168, pFM169, pFM170, pFM221, pFM266, pFM280, and pFM342, containing the bmh1-42, bmh1-98, bmh1-103, bmh1-127, bmh1-167, bmh1-168, bmh1-169, bmh1-170, bmh1-221, bmh1-266, bmh1-280, and bmh1-342 alleles, respectively, were cloned into the BamHI-EcoRI sites of the YIplac128 (LEU2) integrative plasmid to generate plasmids pML378, pML379, pML380, pML381, pML382, pML383, pML384, pML385, pML387, pML389, pML391, and pML392, respectively. PmlI digestion was then used to direct integration of these plasmids into the BMH1 promoter region of a bmh1 bmh2 strain, kept alive by the URA3 BMH1 plasmid pML309. Leu+ Ura- clones were selected on 5-FOA plates from the Leu+ Ura+ transformants containing the different bmh1 alleles, giving rise to strains YLL1085, YLL1079, YLL1082, YLL1088, YLL1083, YLL1086, YLL1080, YLL1087, YLL1081, YLL1090, YLL1092, and YLL1120, carrying the different bmh1 alleles as the sole 14-3-3 source (see Table 1) at the BMH1 chromosomal locus in single copy, with the exception of strain YLL1120, which carries two tandem copies of bmh1-266.

Genetic interactions and cell viability:
Strains YLL1085, YLL1079, YLL1082, YLL1088, YLL1083, YLL1086, YLL1080, YLL1087, YLL1081, YLL1120, YLL1090, and YLL1092 (see previous section) were crossed to W303-derivative strains carrying the pol1-1 (PIZZAGALLI et al. 1988 ), pri2-1 (LONGHESE et al. 1993 ), rfa1-M2 (LONGHESE et al. 1994 ), rfa2-2 (SANTOCANALE et al. 1995 ), or cdc6-1 (HARTWELL 1971 ) mutations concomitantly with the bmh2::KANMX4 allele. All diploid strains were therefore homozygous bmh2/bmh2 and each of them was heterozygous for one bmh1 allele and one DNA replication mutation. For each sporulated diploid, spores from 12 to 20 dissected tetrads were scored for the ability to form colonies on YEPD plates at 25°, and segregation of the bmh1 and DNA replication mutations in the bmh2 viable spores was assayed by complementation tests. Synthetic lethality was inferred when no viable double-mutant spores were observed from a diploid and when there was no significant deviation from the 1 parental ditype:4 tetratypes:1 nonparental ditype ratio of tetrad types predicted from segregation of two unlinked genes. When segregants containing both mutations were viable, serial dilutions of three independent cultures of two single-mutant and two double-mutant segregants for each cross were spotted on YEPD plates with or without MMS (0.005%) or HU (10 mM). YEPD plates were prepared in triplicate and one of them was UV irradiated (40 J/m²). One YEPD plate was incubated at 32° for 3 days, while all the other plates were incubated at 25° for 4 days before determining the number of colony forming units for each strain under the different conditions. No significant differences were found among strains with the same genotypes.

To obtain strains for epistasis analysis, strains YLL157, YLL244 (LONGHESE et al. 1997 ), and YLL490 (PACIOTTI et al. 2000 ), carrying the rad9, ddc1, and mec1 alleles, respectively, were crossed with strain DMP4168/1A, carrying the bmh1-266 allele, followed by sporulation and tetrad dissection. UV dose-response killing curves for two double-mutant and two single-mutant meiotic segregants from each cross were then determined as described in the Fig 6 legend. No significant differences were found among strains with the same genotypes.

View larger version (43K): [in this window] [in a new window] [Download PPT slide]   Figure 1. Overexpression of BMH1 causes DNA damage checkpoint defects. (A and B) Cultures of wild-type (K699) and GAL1-BMH1 (YLL962.2) strains, logarithmically growing in YEP+raf, were synchronized in G1 with -factor, UV irradiated (45 J/m2), and released from -factor at time zero in YEP+raf+gal. Galactose was added 30 min prior to the release. Samples were collected at the indicated times after -factor release to analyze the DNA content by fluorescence-activated cell sorter (FACS) in unirradiated (A, top) and UV-irradiated (A, bottom) cultures and the pattern of Rad53 phosphorylation by Western analysis of protein extracts of the UV-irradiated cultures using anti-Rad53 antibodies (B). The histogram color in A changes when >50% of the cells have completed S phase, as measured by CELLQuest software. (C and D) Cultures of wild-type (K699) and GAL1-BMH1 (YLL962.2) strains, logarithmically growing in YEP+raf, were synchronized in G2 with nocodazole (5 µg/ml), UV irradiated (50 J/m2), and released from nocodazole at time zero in YEP+raf+gal. Galactose was added 30 min prior to the release. Samples were collected at the indicated times after nocodazole release to analyze the percentage of binucleate cells (C) by propidium iodide staining in unirradiated (open symbols) and UV-irradiated (solid symbols) cultures and the pattern of Rad53 phosphorylation in the UV-irradiated cultures as in A (D). exp, exponentially growing cells.

View larger version (83K): [in this window] [in a new window] [Download PPT slide]   Figure 2. Phenotypes of the bmh1 mutants. Serial dilutions of wild-type (wt; K699), bmh1 (YLL138), bmh2 (YLL908), bmh1-42 bmh2 (YLL1085), bmh1-98 bmh2 (YLL1079), bmh1-103 bmh2 (YLL1082), bmh1-127 bmh2 (YLL1088), bmh1-167 bmh2 (YLL1083), bmh1-168 bmh2 (YLL1086), bmh1-169 bmh2 (YLL1080), bmh1-170 bmh2 (YLL1087), bmh1-221 bmh2 (YLL1081), bmh1-266 bmh2 (YLL1120), bmh1-280 bmh2 (YLL1090), and bmh1-342 bmh2 (YLL1092) strains, exponentially growing in YEPD at 25°, were spotted on YEPD plates with or without MMS and HU at the indicated concentrations and incubated at 25° for 3 days. YEPD plates were made in triplicate and one of them was UV irradiated before incubation at 25°, while another was incubated at 37°.

View larger version (57K): [in this window] [in a new window] [Download PPT slide]   Figure 3. Amino acid residues changed by bmh1 mutations. The predicted amino acid sequence of wild-type Bmh1 is shown (top line), as well as the amino acid changes caused by the different bmh1 alleles, which are listed on the left. Numbers on the right refer to the amino acid sequence of S. cerevisiae Bmh1. Asterisks and dots below the sequence indicate identical and similar amino acid residues, respectively, found by aligning the whole amino acid sequence of Bmh1, Bmh2, S. pombe Rad24 and Rad25, and human 14-3-3- and 14-3-3- using the ClustalW program. Solid circles and lines above the sequence indicate the amino acids that interact with the substrate and the -helix regions (A–I), respectively (reviewed in FU et al. 2000 ).

View larger version (65K): [in this window] [in a new window] [Download PPT slide]   Figure 4. G1/S DNA damage checkpoint. Strains were as follows: wild type (wt; K699), bmh1 (YLL138), bmh2 (YLL908), bmh1-103 bmh2 (YLL1082), and bmh1-266 bmh2 (YLL1120). Cell cultures, logarithmically growing in YEPD at 25°, were synchronized in G1 with -factor, UV irradiated (45 J/m2), and released from -factor at time zero in YEPD at 25°. Samples were withdrawn at the indicated times after -factor release to analyze the kinetics of bud emergence (A) in unirradiated (open symbols) and UV-irradiated (solid symbols) cultures, the DNA content by FACS in unirradiated (B, top) and UV-irradiated (B, bottom) cell cultures, and Rad53 phosphorylation in UV-irradiated cell cultures as in Fig 1A (C). The histogram color in B changes when >50% of the cells have completed S phase, as measured by CELLQuest software. Time zero corresponds to cell samples withdrawn immediately before UV irradiation and release from -factor. exp, exponentially growing cells. Percentages of cell viability measured 10 min after UV irradiation were the following: wild type, 63%; bmh1, 62%; bmh2, 61%; bmh1-103 bmh2, 30%; bmh1-266 bmh2, 28%.

View larger version (64K): [in this window] [in a new window] [Download PPT slide]   Figure 5. G2/M DNA damage checkpoint. Strains were as follows: wild type (wt; YLL839), bmh1 (DMP4058/9A), bmh2 (DMP3817/1A), bmh1-103 bmh2 (DMP3818/5C), bmh1-169 bmh2 (DMP3820/6A), bmh1-221 bmh2 (DMP3821/1B), and bmh1-266 bmh2 (DMP3855/2D). Cell cultures, logarithmically growing in YEPD at 25°, were synchronized in G2 with nocodazole (5 µg/ml), UV irradiated (60 J/m2), and released from nocodazole at time zero in YEPD at 25°. Samples were collected at the indicated times after nocodazole release to determine the percentage of binucleate cells (A) by propidium iodide staining in unirradiated (open symbols) and UV-irradiated (solid symbols) cultures. Protein extracts from the UV-treated cell cultures were analyzed by Western blots using anti-Rad53 (B) and anti-HA (Chk1) antibodies (C). Time zero corresponds to cell samples withdrawn immediately before UV irradiation and release from nocodazole. exp, exponentially growing cells.

View larger version (19K): [in this window] [in a new window] [Download PPT slide]   Figure 6. Epistasis analysis. Dose-response killing curves were determined by plating serial dilutions of YEPD exponentially growing cell cultures of strains with the indicated genotypes (see MATERIALS AND METHODS) on YEPD plates, which were exposed to the indicated UV doses. Plates were incubated at 25° and colony-forming units were counted after 3 days. All the analyzed mec1 strains also carried the sml1 allele, which suppresses the lethal effects of MEC1 deletion (PACIOTTI et al. 2000 ), and all the analyzed bmh1 mutant strains also carried the bmh2 allele.

Gross chromosomal rearrangements:
GCR were detected, as in MYUNG et al. 2001 , by selection of Can^R-FOA^R colonies from wild-type (RDKY3615), bmh1 (YLL1238), bmh2 (YLL1290), and bmh1 bmh2 cell cultures. Spontaneous GCR rates were calculated by fluctuation analysis using the method of the median, as previously described (LEA and COULSON 1948 ), with sets of five independent cultures.

Other techniques:
Synchronization experiments, total protein extract preparation, and Western blot analysis were performed as described in PACIOTTI et al. 2000 . Flow cytometric DNA quantitation was determined on a Becton-Dickinson FACScan. The percentage of cells with 2C DNA contents was measured using CELLQuest software.

	RESULTS AND DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

BMH1 overexpression impairs DNA damage checkpoint response:
We previously identified BMH1 as a high-dosage suppressor of the synthetic lethality caused by the combination of the pri1-2 allele altering the DNA primase subunit of the pol-primase complex with the mec1-14 DNA damage checkpoint mutation (LONGHESE et al. 1996 ; our unpublished observation), suggesting that Bmh1 may be involved in DNA replication and/or DNA damage response. To explore this possibility further, we first analyzed the effects of BMH1 overexpression on cell cycle progression in the absence or presence of DNA damage. To this end, exponentially growing cell cultures of a BMH1 strain, carrying two copies of a galactose-inducible GAL1-BMH1 fusion at the LEU2 locus, were arrested in G₁ with -factor or in G₂ with nocodazole and then released into cell cycle under galactose-induced conditions, either without any genotoxic treatment or after UV irradiation. As shown in Fig 1A-phase progression, which occurred with similar kinetics in unirradiated GAL1-BMH1 and wild-type cell cultures after release from the G₁ block, was faster in BMH1-overexpressing cells that were UV-irradiated before release than in wild-type cells under the same conditions. In fact, most GAL1-BMH1 cells completed DNA replication within 120 min after UV irradiation, whereas UV-treated wild-type cells were still mostly in S phase at 150 min (Fig 1A, bottom). Galactose-induced GAL1-BMH1 cells failed not only to properly arrest DNA replication, but also to maintain phosphorylated forms of the key checkpoint kinase Rad53 after DNA damage in G₁. In fact, Rad53 phosphorylation, which correlates with checkpoint activation and is detectable as changes in Rad53 electrophoretic mobility (SANCHEZ et al. 1996 ), started to decrease 90–105 min after UV irradiation in G₁ in GAL1-BMH1 cells, whereas it persisted until the end of the experiment in similarly treated wild-type cells (Fig 1B).

As shown in Fig 1C and Fig D, BMH1-overexpressing cells were also impaired in the G₂/M DNA damage checkpoint. In fact, galactose-induced GAL1-BMH1 cells started to divide nuclei, with concomitant decrease of Rad53 phosphorylation, 75 min after UV irradiation in G₂, while similarly treated wild-type cells underwent the same processes 15–30 min later (Fig 1C, solid symbols, and Fig 1D). Conversely, kinetics of nuclear division were comparable in GAL1-BMH1 and wild-type cell cultures released from G₂ arrest without genotoxic treatment (Fig 1C, open symbols). Thus, GAL1-BMH1-overexpressing cells are defective in arresting cell cycle progression and in maintaining phosphorylated Rad53 after UV irradiation in either G₁ or G₂, suggesting that increasing the levels of Bmh1 might unbalance its association with different targets and this, in turn, may alter DNA damage checkpoint response.

Random mutagenesis of the BMH1 gene:
The above results prompted us to analyze in more detail the possible role of Bmh proteins in DNA metabolism and DNA damage response. We noted that bmh1 cells were slightly more sensitive to the DNA synthesis inhibitor HU than were isogenic bmh2 cells, which were instead undistinguishable from wild type (Fig 2). Since bmh1 HU sensitivity was fully complemented by wild-type BMH1 on a centromeric plasmid (data not shown), this suggests that Bmh1 may have a slightly more important role than Bmh2 in DNA metabolism and/or response to DNA damage, although Bmh1 and Bmh2 functions likely largely overlap.

We then used a bmh1 bmh2 strain kept alive by an URA3 BMH1 centromeric plasmid to screen for bmh1 alleles conferring temperature sensitivity and/or sensitivity to HU, UV radiations, or MMS, when present as the sole 14-3-3 source in the cell (see MATERIALS AND METHODS). We integrated 12 bmh1 alleles found to cause different mutant phenotypes in the preliminary screening at the BMH1 chromosomal promoter region of a bmh1 bmh2 strain (see MATERIALS AND METHODS). As shown in Fig 2, a colony-forming-unit assay under different conditions showed that the obtained stable bmh1 bmh2 strains displayed different phenotypes. In fact, the bmh1-42, bmh1-103, bmh1-167, bmh1-169, bmh1-170, bmh1-221, and bmh1-266 mutants all showed different degrees of temperature sensitivity and hypersensitivity to all the genotoxic treatments. The bmh1-98, bmh1-127, and bmh1-280 mutants, which grew well at 37°, were hypersensitive to HU and MMS but not to UV. Finally, the bmh1-168 and bmh1-342 mutants were temperature sensitive and hypersensitive to MMS and UV radiations, while they showed a very limited sensitivity to HU. All of the above bmh1 alleles did not cause any mutant phenotype in the presence of functional BMH2 (data not shown), indicating that Bmh2 was able to substitute for the altered Bmh1 functions.

Since the structure of 14-3-3 proteins is evolutionarily conserved and Bmh1 shows 67–71% amino acid sequence identity with human 14-3-3 isoforms (reviewed in FU et al. 2000 ), we verified whether determination and comparison of the nucleotide sequences of the whole wild-type and mutant BMH1 coding region allowed us to ascribe significance to some mutations on the basis of what is known about human 14-3-3 structure. As shown in Fig 3, most bmh1 alleles carried multiple base-pair substitutions causing changes of different amino acid residues. Only the bmh1-167 and bmh1-280 alleles carried single base-pair substitutions, resulting in the amino acid changes D129N and E136G, respectively. Although our analysis so far has not allowed us to easily ascribe specific phenotypes to specific amino acid changes, alignment of the amino acid sequence of Bmh1, Bmh2, S. pombe Rad24 and Rad25, and human 14-3-3- and 14-3-3- indicated that most bmh1 mutations changed highly conserved residues (Fig 3).

Checkpoint response after UV irradiation in G₁ or G₂:
All the analyzed bmh1 bmh2 mutants were more sensitive than wild type to all or some genotoxic treatments, indicating that Bmh proteins are required for proper response to DNA damage. We therefore asked whether this was related to checkpoint defects.

Since the lack of the Rad24 S. pombe 14-3-3 isoform causes mild sensitivity to genotoxic agents and defective DNA damage checkpoint (FORD et al. 1994 ), we first analyzed the checkpoint response in bmh1 and bmh2 single mutants. To this end, exponentially growing cultures of wild type, bmh1, and bmh2 cells were synchronized in G₁ with -factor (Fig 4) or in G₂ with nocodazole (Fig 5) and then released into the cell cycle, either untreated or after UV irradiation. The kinetics of budding (Fig 4A, top) and S-phase progression (Fig 4B) after release from G₁, as well as that of nuclear division after release from G₂ (Fig 5A, top), of both unirradiated and UV-irradiated bmh1 and bmh2 cells were identical to those of wild-type cells and were properly delayed in the irradiated cultures. Thus, unlike single rad24 deletion in S. pombe, single deletion of either BMH1 or BMH2 does not seem to affect DNA damage checkpoint response.

When the 12 bmh1 bmh2 mutants were analyzed under the above conditions, most bmh1 mutants did not show checkpoint defects, since neither their budding and DNA replication kinetics after UV irradiation in G₁ nor their nuclear division kinetics after UV treatment in G₂ were significantly different from those of the control strains (data not shown).

On the other hand, bmh1-103 bmh2 and bmh1-266 bmh2 cells, UV treated in G₁ before release into the cell cycle, failed to properly delay budding (Fig 4A, bottom, solid symbols) and S-phase progression (Fig 4B, bottom). In fact, most of the UV-irradiated mutant cells budded and reached 2C DNA content within 105 min after UV irradiation in G₁, whereas most bmh2 cells under the same conditions completed budding and DNA replication only 120 and 150 min after -factor release, respectively (Fig 4A, bottom, solid symbols, and Fig 4B, bottom). Thus, a G₁/S checkpoint defect was detectable in bmh1-103 bmh2 and bmh1-266 bmh2 cells, in spite of a slight delay in both bud emergence and S-phase progression in untreated mutant cultures compared to bmh2 (Fig 4A, bottom, open symbols, and Fig 4B, top).

The bmh1-103 bmh2 and bmh1-266 bmh2 strains also turned out to be defective in the G₂/M checkpoint, as did the bmh1-169 bmh2 and bmh1-221 bmh2 strains, which did not show G₁/S checkpoint defects. In fact, when exponentially growing cell cultures were arrested in G₂ with nocodazole and UV irradiated before release into the cell cycle, the percentage of binucleate cells started to increase within 60 min in bmh1-266 bmh2 cells and within 75 min in bmh1-103 bmh2, bmh1-169 bmh2, and bmh1-221 bmh2 cells (Fig 5A, middle and bottom, solid symbols), whereas it increased only 90 min after nocodazole release in similarly treated wild-type, bmh1, and bmh2 cell cultures (Fig 5A, top, solid symbols). It is worth noting that, when parts of the above cultures were released from the block under unperturbed conditions, the same bmh1 bmh2 mutants (Fig 5A, middle and bottom, open symbols) underwent nuclear division more slowly than the control strains (Fig 5A, top, open symbols), indicating that the corresponding Bmh1 mutant proteins delayed nuclear division in the absence of exogenous DNA damage, which might partially mask the G₂/M checkpoint defects.

G₁/S checkpoint activation requires the Rad53 checkpoint kinase and correlates with its phosphorylation, while both Chk1 and Rad53 kinases are phosphorylated after DNA damage in G₂ and are required for the G₂/M checkpoint (SANCHEZ et al. 1996 , SANCHEZ et al. 1999 ). As shown in Fig 4C, Rad53 was phosphorylated immediately after UV irradiation in G₁ in all the mutant and control cultures described above, but the bmh1 bmh2 mutants were defective in maintaining this phosphorylation. In fact, Rad53 phosphorylated forms persisted until the end of the experiment in wild-type, bmh1, and bmh2 cells, whereas their amount started to decrease in bmh1-266 bmh2 and bmh1-103 bmh2 mutants 90–105 min after UV irradiation, concomitantly with the increase in the amount of the unphosphorylated form (Fig 4C).

As shown in Fig 5B and Fig C, both Rad53 and Chk1 phosphorylated forms appeared immediately in all cell cultures also after UV irradiation in G₂ and persisted until the end of the experiment in wild-type, bmh1, and bmh2 cells. However, the amount of phosphorylated Rad53 started to decrease prematurely, 75–90 min after UV irradiation, in bmh1-103 and bmh1-266 cells, whereas it was only slightly reduced in bmh1-169 cells and did not seem to be affected in the bmh1-221 mutant (Fig 5B). Chk1 phosphorylation was instead abolished in all the mutants 90 min after UV treatment (Fig 5C).

Thus, the Bmh1-103 and Bmh1-266 proteins, impairing both the G₁/S and the G₂/M checkpoints, lead to defects in maintaining high levels of phosphorylated Rad53 after DNA damage in both G₁ and G₂, as well as phosphorylated Chk1 after DNA damage in G₂. Conversely, the Bmh1-169 and Bmh1-221 proteins, causing only G₂/M checkpoint defects, appeared to be specifically impaired in their ability to maintain DNA-damage-induced phosphorylation of the G₂/M checkpoint kinase Chk1.

The above findings suggest that Bmh proteins may act by ensuring persistence of DNA damage checkpoint activation until repair events allow it to turn off. Consistent with this hypothesis, somatic cells lacking 14-3-3- can initiate G₂/M arrest following DNA damage, but are unable to maintain it (CHAN et al. 1999 ). S. cerevisiae Bmh proteins may protect the phosphorylated forms of the key checkpoint kinases Rad53 and Chk1 from the action of phosphatases and/or regulate the accessibility of these proteins to factors acting upstream in the checkpoint cascade, such as Mec1 and Rad9. A role for 14-3-3 proteins in protecting against dephosphorylation events has been previously described for a proapoptotic member of the BCL-2 family of proteins, BAD, which plays a role in regulating apoptosis in cytokine-dependent hematopoietic cells. In fact, its dissociation from human 14-3-3 was shown to be a prerequisite for BAD dephosphorylation in vitro (CHIANG et al. 2001 ). Moreover, 14-3-3 proteins block inactivation of Raf1, a member of the serine/threonine kinase Raf family, by inhibiting its protein tyrosine phosphatase-1B-dependent dephosphorylation (JELINEK et al. 1996 ).

To further analyze the role of Bmh1 in DNA damage checkpoints, we performed epistasis analysis of the four checkpoint-defective bmh1 alleles with mutations in the DDC1, RAD9, and MEC1 checkpoint genes by comparing the UV sensitivity of bmh2 strains carrying each single mutation to that of isogenic strains carrying the double-mutant combinations. As shown in Fig 6, the bmh1-266 ddc1 bmh2 strain was more sensitive to UV treatment than the corresponding single-mutant strains. Conversely, both the bmh1-266 rad9 bmh2 and the bmh1-266 mec1 bmh2 strains were as sensitive as the rad9 or the mec1 single mutants, respectively, which in turn were more sensitive than the bmh1-266 bmh2 strain (Fig 6). Similar results were obtained also with the bmh1-103, bmh1-169, and bmh1-221 alleles (data not shown), indicating that Bmh1 acts in the same pathway of DNA damage response as Rad9 and Mec1. Since phosphorylation of Rad53 and Chk1 in response to UV irradiation is totally dependent on Mec1, which triggers their activation through Rad9 (EMILI 1998 ; SUN et al. 1998 ; VIALARD et al. 1998 ), this finding further supports the hypothesis that Bmh proteins may modulate Rad53 and Chk1 phosphorylation activities. Also consistent with this hypothesis is the observation that the same bmh1 alleles increase UV sensitivity in the absence of the DDC1 gene, which contributes to full Mec1 activation, but belongs to an epistasis group different from that of RAD9 (LONGHESE et al. 1997 ).

bmh1 bmh2 mutants fail to complete DNA replication after transient nucleotide depletion:
Hydroxyurea blocks the progression of replication forks from early-firing origins, thus activating the DNA replication checkpoint pathway. This in turn blocks initiation at late replication origins, stabilizes stalled forks, and links S-phase completion with mitotic onset by inhibiting spindle elongation (DESANY et al. 1998 ; SANTOCANALE and DIFFLEY 1998 ; SHIRAHIGE et al. 1998 ; LOPES et al. 2001 ). We found that, although several bmh1 mutants were hypersensitive to HU, all of them arrested with a single nucleus and short spindles after an S-phase block by HU (data not shown), indicating that their HU sensitivity is not due to inappropriate mitotic entry.

So, while the 12 characterized bmh1 bmh2 mutants all showed hypersensitivity to at least some of the used genotoxic treatments, 8 of them did not have detectable DNA damage or DNA replication checkpoint defects. This indicates that Bmh proteins are likely required for cellular responses other than checkpoints to these DNA-damaging agents.

Since HU inhibits DNA synthesis by depleting the dNTP pool, while MMS and UV profoundly reduce the rate of DNA replication fork progression (PAULOVICH and HARTWELL 1995 ; NEECKE et al. 1999 ; TERCERO and DIFFLEY 2001 ), the hypersensitivity of bmh1 bmh2 mutants to these genotoxic agents might be due to the inability to properly respond to stresses on progressing replication forks. We therefore asked whether these mutants were defective in the recovery from DNA replication stress by analyzing S-phase progression in bmh1 bmh2 mutants released from a HU-induced replication block. To this end, exponentially growing cultures of our 12 mutants and of bmh1, bmh2, and wild-type control strains were treated with HU (150 mM) for 3.5 hr, thus arresting cells as large-budded cells with mostly 1C DNA content (Fig 7, time zero). Cells were then released from HU in YEPD medium containing nocodazole, which allows completion of S phase and avoids entering in the subsequent cell cycle. As shown in Fig 7, wild-type, bmh1, and bmh2 strains behaved very similarly to each other and completed S phase within 60 min after release, indicating that deletion of BMH1 or BMH2 does not affect recovery from the DNA replication block. Strains carrying the bmh1-127, bmh1-168, and bmh1-342 alleles in a bmh2 background, which showed a very weak sensitivity to HU (Fig 2), were also undistinguishable from the control strains (data not shown). Conversely, bmh1-170 bmh2 cells were still mostly in S phase 240 min after HU release, and the bmh2 cells carrying the bmh1-42, bmh1-98, bmh1-169, bmh1-266, and bmh1-280 alleles completed DNA replication after 180–240 min (Fig 7). A less dramatic, but still significant, effect was observed in the bmh1-103 bmh2, bmh1-167 bmh2, and bmh1-221 bmh2 strains, which completed DNA replication within 120–150 min after release (Fig 7). The inability of bmh1 bmh2 mutants to recover from HU-induced arrest of DNA replication did not correlate with possible defects in cell cycle progression caused by the mutations per se. In fact, bmh2 strains carrying the bmh1-42, bmh1-98, bmh1-167, and bmh1-280 alleles did not show any delay in S-phase entry and progression after release from -factor under unperturbed conditions (data not shown).

View larger version (92K): [in this window] [in a new window] [Download PPT slide]   Figure 7. Recovery from HU arrest in bmh1 mutants. Strains were as follows: wild type (wt; K699), bmh1 (YLL138), bmh2 (YLL908), bmh1-103 bmh2 (YLL1082), bmh1-167 bmh2 (YLL1083), bmh1-221 bmh2 (YLL1081), bmh1-42 bmh2 (YLL1085), bmh1-98 bmh2 (YLL1079), bmh1-169 bmh2 (YLL1080), bmh1-266 bmh2 (YLL1120), bmh1-280 bmh2 (YLL1090), and bmh1-170 bmh2 (YLL1087). Cell cultures, exponentially growing in YEPD at 25°, were arrested with HU (150 mM) for 3.5 hr (time zero) and then released from HU at 25° in YEPD containing nocodazole (15 µg/ml). Samples were collected at the indicated times after release to analyze the DNA content by FACS. The histogram color changes when >50% of the cells have completed S phase, as measured by CELLQuest software. Cell viability was measured by plating serial dilutions of the above cultures on YEPD immediately before HU treatment and after release from HU. Percentages of cell survival to HU treatment were the following: wild type, 97%; bmh1, 88%; bmh2, 89%; bmh1-103 bmh2, 55%; bmh1-167 bmh2, 51%; bmh1-221 bmh2, 52%; bmh1-42 bmh2, 41%; bmh1-98 bmh2, 39%; bmh1-169 bmh2, 41%; bmh1-266 bmh2, 34%; bmh1-280 bmh2, 41%; bmh1-170 bmh2, 3%.

Thus, Bmh1 is involved in maintaining the ability to resume DNA synthesis when replication stress is removed or overcome, suggesting that Bmh proteins may be required to maintain the integrity of replication forks under stress conditions.

Synthetic effects between bmh1 and DNA replication mutations:
To investigate further the involvement of these proteins in ensuring DNA synthesis, we analyzed the in vivo interactions of our bmh1 alleles with mutations in DNA replication genes. Among several DNA replication mutants available, for this analysis we chose mutants defective in the POL1, PRI2, RFA1, and RFA2 genes, which play different roles in both the initiation and elongation steps of DNA replication (WAGA and STILLMAN 1998 ) and in CDC6, which has an essential role in prereplicative complex assembly and subsequent licensing of DNA synthesis initiation (PIATTI et al. 1995 ; COCKER et al. 1996 ). None of the chosen DNA replication mutations showed any synthetic effects with either the bmh1 or the bmh2 alleles (data not shown). On the other hand, as shown in Table 2, in a bmh2 background all the analyzed bmh1 alleles caused synthetic effects with mutations altering either the pol-primase complex (pol1-1 and pri2-1) or the single-strand DNA-binding RPA complex (rfa1-M2 and rfa2-2), although to different extents. In particular, viable double mutants could not be recovered when the bmh1-98, bmh1-167, bmh1-169, bmh1-170, bmh1-221, bmh1-266, and bmh1-280 alleles were introduced as the sole 14-3-3 source in the pol1-1 temperature-sensitive mutant altering DNA polymerase- (PIZZAGALLI et al. 1988 ), indicating that these combinations were synthetic lethal. The bmh1-221 allele was also lethal in combination with both the pri2-1 temperature-sensitive mutation, affecting the p58 subunit of the pol-primase complex (LONGHESE et al. 1993 ), and the rfa2-2 or rfa1-M2 conditional mutations, altering the p70 and p34 subunits of the RPA complex, respectively (LONGHESE et al. 1994 ; SANTOCANALE et al. 1995 ). Moreover, the bmh1-170 and bmh1-280 alleles were lethal for both the pri2-1 bmh2 and rfa2-2 bmh2 mutants, and introduction in the latter of bmh1-167 was also lethal. Finally, the bmh1-98 allele was lethal in the rfa1-M2 bmh2 background. When not causing synthetic lethality, all the analyzed bmh1 alleles caused increased growth defects and/or temperature sensitivity and/or sensitivity to HU, MMS, and UV when combined with pol1-1, pri2-1, rfa1-M2, or rfa2-2 (Table 2).

The extent of these synthetic effects largely correlates with the defects caused by the same bmh1 alleles in resuming DNA synthesis after transient nucleotide depletion. In fact, the bmh1-42, bmh1-98, bmh1-169, bmh1-170, bmh1-266, and bmh1-280 alleles, all causing severe defects in recovery from HU treatment in a bmh2 background, resulted in either synthetic lethality or dramatic growth defects when introduced as the sole 14-3-3 source in the above conditional mutants. The presence in the same mutants of the Bmh1-127, Bmh1-168, and Bmh1-342 variants, which do not affect cell cycle recovery from HU in the absence of other Bmh forms, and of Bmh1-103, which only weakly affected it, instead caused less severe synthetic effects. Since both HU and DNA replication mutants interfere with replication fork progression, it is tempting to propose that Bmh proteins may be required to properly carry out DNA replication under stress conditions.

This hypothesis does not fully explain the behavior of the bmh1-167 and bmh1-221 alleles, which have severe synthetic effects on DNA replication mutants while only weakly affecting recovery from HU arrest (Fig 7 and Table 2). More specific interactions should account for the bmh1-167 synthetic effects, while impairment of the G₂/M DNA damage checkpoint caused by the bmh1-221 allele might be responsible for its synthetic lethality with all the analyzed DNA replication mutations.

These genetic interactions seem to be specific for mutations altering proteins required for replication fork progression. In fact, introduction of any of the bmh1 alleles in a bmh2 strain carrying the cdc6-1 allele (HARTWELL 1971 ), which interferes with proper initiation of DNA replication, resulted in viable bmh1 bmh2 cdc6-1 spores, whose growth at 25° was not significantly affected, with the exception of the strains carrying the bmh1-167 cdc6-1, bmh1-221 cdc6-1, and bmh1-266 cdc6-1 combinations, which grew more slowly compared to each single mutant (Table 2). We could not yet rule out the possibility that this lack of synthetic effects might be due to the specific cdc6-1 allele. On the other hand, the finding that 14-3-3 alterations may cause more dramatic effects when combined with mutations directly affecting DNA synthesis than with a mutation impairing DNA replication licensing raises the possibility that these proteins might have some functions in DNA replication fork progression.

Bmh proteins and genetic stability:
Since defects in repair and checkpoint mechanisms, as well as faulty DNA replication or aberrant chromosome segregation, all lead to genomic instability, we studied whether bmh1 bmh2 mutants were affected in genome stability. We used a genetic assay to measure the rate of rearrangements on the left arm of chromosome V, containing the telomere-proximal CAN1 gene and the URA3 gene integrated 7.5 kb telomeric to CAN1 (MYUNG et al. 2001 ). Rearrangements that simultaneously delete that region of chromosome V can be detected in haploid cells, as loss of both CAN1 and URA3 leads to resistance to both canavanine and 5-fluoroorotic acid. As shown in Table 3, we found that deletion of either BMH1 or BMH2 repeatedly slightly increased the GCR rate 2-fold compared to wild type, and a similar effect was observed in most bmh1 bmh2 mutants (Table 4). Strikingly, GCR rates were 32- and 12-fold higher in the bmh1-169 bmh2 and bmh1-170 bmh2 strains, respectively, than in wild type (Table 3), indicating that Bmh proteins are required to suppress genome instability even in the absence of genotoxic treatments. Bmh functions in maintaining genetic integrity are therefore conserved throughout evolution. In fact, inactivation of human 14-3-3- causes sensitivity to -irradiation, frequent chromosome breakage with complete loss of the telomeric repeats in some chromosomes, as well as chromosome end-to-end fusions, ring chromosome formation, and translocations (DHAR et al. 2000 ).

View this table: [in this window] [in a new window]   Table 4. Summary of the main phenotypes of bmh1 bmh2 mutants

Altogether, the data presented in this article and summarized in Table 4 indicate that Bmh proteins are involved in different aspects of DNA damage response and DNA metabolism, and open the possibility of using S. cerevisiae mutants for functional in vivo characterization of 14-3-3 proteins. Moreover, further analysis of the roles of Bmh proteins in DNA replication and checkpoint controls should yield interesting insights into the regulation of these processes.

	ACKNOWLEDGMENTS

We thank J. Diffley for antibodies against Rad53, R. D. Kolodner for strain RDKY3615, S. Piatti for critical reading of the manuscript, G. Santarossa for assistance in the analysis of 14-3-3 structure, and all the members of our laboratory for useful discussions and criticisms. This work was supported by grants from Associazione Italiana per la Ricerca sul Cancro, Telethon-Italy (E.1247) and Cofinanziamento 2003 MIUR/Università di Milano-Bicocca to M.P.L., and Consorzio Interuniversitario Biotecnologie, Fondo per gli investimenti della Ricerca di Base and Progetto Strategico MIUR-Legge 449/97 to G.L. F. Lottersberger was partially supported by a fellowship from Fondazione Confalonieri and V. Baldo was supported by a fellowship from Fondazione Italiana per la Ricerca sul Cancro.

Manuscript received July 31, 2003; Accepted for publication September 2, 2003.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

BENTLEY, N. J., D. A. HOLTZMAN, G. FLAGGS, K. S. KEEGAN, and A. DEMAGGIO et al., 1996  The Schizosaccharomyces pombe rad3 checkpoint gene. EMBO J. 15:6641-6651.[Medline]

BLASINA, A., I. V. DE WEYER, M. C. LAUS, W. H. LUYTEN, and A. E. PARKER et al., 1999  A human homologue of the checkpoint kinase Cds1 directly inhibits Cdc25 phosphatase. Curr. Biol. 9:1-10.[Medline]

CARR, A. M., 1997  Control of cell cycle arrest by the Mec1^sc/Rad3^sp DNA structure checkpoint pathway. Curr. Opin. Genet. Dev. 7:93-98.[Medline]

CHAN, T. A., H. HERMEKING, C. LENGAUER, K. W. KINZLER, and B. VOGELSTEIN, 1999  14-3-3 is required to prevent mitotic catastrophe after DNA damage. Nature 401:616-620.[Medline]

CHEN, L., T. H. LIU, and N. C. WALWORTH, 1999  Association of Chk1 with 14-3-3 proteins is stimulated by DNA damage. Genes Dev. 13:675-685.[Abstract/Free Full Text]

CHIANG, C. W., G. HARRIS, C. ELLIG, S. C. MASTERS, and R. SUBRAMANIAN et al., 2001  Protein phosphatase 2A activates the proapoptotic function of BAD in interleukin-3-dependent lymphoid cells by a mechanism requiring 14-3-3 dissociation. Blood 97:1289-1297.[Abstract/Free Full Text]

CLERICI, M., V. PACIOTTI, V. BALDO, M. ROMANO, and G. LUCCHINI et al., 2001  Hyperactivation of the yeast DNA damage checkpoint by TEL1 and DDC2 overexpression. EMBO J. 20:6485-6498.[Medline]

COCKER, J. H., S. PIATTI, C. SANTOCANALE, K. NASMYTH, and J. F. X. DIFFLEY, 1996  An essential role for the Cdc6 protein in forming the pre-replicative complexes of budding yeast. Nature 379:180-182.[Medline]

CORTEZ, D., S. GUNTUKU, J. QIN, and S. J. ELLEDGE, 2001  ATR and ATRIP: partners in checkpoint signaling. Science 294:1713-1716.[Abstract/Free Full Text]

CVRCKOVA, F. and K. NASMYTH, 1993  Yeast G1 cyclins CLN1 and CLN2 and a GAP-like protein have a role in bud formation. EMBO J. 12:5277-5286.[Medline]

DALAL, S. N., C. M. SCHWEITZER, J. GAN, and J. A. DECAPRIO, 1999  Cytoplasmic localization of human cdc25C during interphase requires an intact 14-3-3 binding site. Mol. Cell. Biol. 19:4465-4479.[Abstract/Free Full Text]

DESANY, B. A., A. A. ALCASABAS, J. B. BACHANT, and S. J. ELLEDGE, 1998  Recovery from DNA replicational stress is the essential function of the S-phase checkpoint pathway. Genes Dev. 12:2956-2970.[Abstract/Free Full Text]

DHAR, S., J. A. SQUIRE, M. P. HANDE, R. J. WELLINGER, and T. K. PANDITA, 2000  Inactivation of 14-3-3 influences telomere behavior and ionizing radiation-induced chromosomal instability. Mol. Cell. Biol. 20:7764-7772.[Abstract/Free Full Text]

EDWARDS, R. J., R. N. BENTLEY, and A. M. CARR, 1999  A Rad3-Rad26 complex responds to DNA damage independently of other checkpoint proteins. Nat. Cell Biol. 1:393-398.[Medline]

EMILI, A., 1998  MEC1-dependent phosphorylation of Rad9p in response to DNA damage. Mol. Cell 2:183-189.[Medline]

FORD, J. C., F. AL-KHODAIRY, E. FOTOU, K. S. SHELDRICK, and D. J. F. GRIFFITHS et al., 1994  14-3-3 protein homologs required for the DNA damage checkpoint in fission yeast. Science 265:533-535.[Abstract/Free Full Text]

FU, H., R. R. SUBRAMANIAN, and S. C. MASTERS, 2000  14-3-3 proteins: structure, function, and regulation. Annu. Rev. Pharmacol. Toxicol. 40:617-647.[Medline]

FURNARI, B., N. RHIND, and P. RUSSELL, 1997  Cdc25 mitotic inducer targeted by chk1 DNA damage checkpoint kinase. Science 277:1495-1497.[Abstract/Free Full Text]

GELPERIN, D., J. WEIGLE, K. NELSON, P. ROSEBOOM, and K. IRIE et al., 1995  14-3-3 proteins: potential roles in vesicular transport and Ras signaling in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 92:11539-11543.[Abstract/Free Full Text]

GIETZ, R. D. and A. SUGINO, 1988  New yeast-Escherichia coli shuttle vectors constructed with in vitro mutagenized yeast genes lacking six base-pair restriction sites. Gene 74:527-534.[Medline]

GILBERT, C. S., C. M. GREEN, and N. F. LOWNDES, 2001  Budding yeast Rad9 is an ATP-dependent Rad53 activating machine. Mol. Cell 8:129-136.[Medline]

GREENWELL, P. W., S. L. KRONMAL, S. E. PORTER, J. GASSENHUBER, and B. OBERMAIER et al., 1995  TEL1, a gene involved in controlling telomere length in S. cerevisiae, is homologous to the human ataxia telangiectasia gene. Cell 82:823-829.[Medline]

HARI, K. L., A. SANTERRE, J. J. SEKELSKY, K. S. MCKIM, and J. B. BOYD et al., 1995  The mei-41 gene of D. melanogaster is a structural and functional homolog of the human ataxia telangiectasia gene. Cell 82:815-821.[Medline]

HARTWELL, L. H., 1971  Genetic control of the cell-division in yeast. II. Genes controlling DNA replication and its initiation. J. Mol. Biol. 59:183-194.[Medline]

JELINEK, T., P. DENT, T. W. STURGILL, and M. J. WEBER, 1996  Ras-induced activation of Raf-1 is dependent on tyrosine phosphorylation. Mol. Cell. Biol. 16:1027-1034.[Abstract]

KUMAGAI, A. and W. G. DUNPHY, 1999  Binding of 14-3-3 proteins and nuclear expor