Differential Regulation of Growth and Checkpoint Control Mediated by a Cdc25 Mitotic Phosphatase from Pneumocystis carinii *

Pneumocystis carinii is an opportunistic fungal pathogen phylogenetically related to the fission yeast Schizosaccharomyces pombe. P. carinii causes severe pneumonia in immunocompromised patients with AIDS and malignancies. Although the life cycle of P. carinii remains poorly characterized, morphologic studies of infected lung tissue indicate that P. carinii alternates between numerous small trophic forms and fewer large cystic forms. To understand further the molecular mechanisms that regulate progression of the cell cycle ofP. carinii, we have sought to identify and characterize genes in P. carinii that are important regulators of eukaryotic cell cycle progression. In this study, we have isolated a cDNA from P. carinii that exhibits significant homology, but unique functional characteristics, to the mitotic phosphatase Cdc25 found in S. pombe. P. carinii Cdc25 was shown to rescue growth of the temperature-sensitive S. pombe cdc25-22 strain and thus provides an additional tool to investigate the unique P. carinii life cycle. AlthoughP. carinii Cdc25 could also restore the DNA damage checkpoint in cdc25-22 cells, it was unable to restore fully the DNA replication checkpoint. The dissociation of checkpoint control at the level of Cdc25 indicates that Cdc25 may be under distinct regulatory control in mediating checkpoint signaling.

Despite advances in prophylaxis and treatment, Pneumocystis carinii remains an important cause of life-threatening pneumonia in patients with impaired immunity. Patients with AIDS, organ transplantation, and those receiving chemotherapy are particularly vulnerable to P. carinii pneumonia. The case fatality rate of P. carinii pneumonia ranges from ϳ10 to 40%, being substantially higher in infected immunocompromised patients without AIDS (1)(2)(3). Unfortunately, a considerable number of patients are intolerant of the currently available agents used to prevent and treat this infection, such as sulfamethoxazole and pentamidine. The development of new chemotherapeutic agents to treat P. carinii pneumonia has been hampered by a limited understanding of the P. carinii life cycle.
Morphological and ultrastructural studies of infected lung tissues indicate that P. carinii alternates between numerous diminutive trophic forms and fewer larger cyst forms (4). Trophic forms are known to bind preferentially type I alveolar epithelial cells and eventually develop into mature cysts characterized by a thickened cell wall (5,6). The mature cyst, containing eight intracystic bodies, ruptures releasing these bodies as haploid trophic forms (4). Although the proliferation of P. carinii in the lung is thus enhanced by the attachment of trophic forms to lung epithelium, only recently have there been initial reports of limited success with in vitro cultivation of P. carinii in the absence of feeder cells (7)(8)(9). However, long term in vitro propagation P. carinii has perennially remained a key obstacle in the investigations of the life cycle of the organism.
Based on DNA sequence analysis, P. carinii has been classified as a member of the fungi and is phylogenetically related to the fission yeast Schizosaccharomyces pombe (10 -12). The life cycle and the cell cycle of S. pombe have been well characterized and utilized to identify cell division cycle (cdc) 1 genes in other organisms with obligate roles in cell proliferation (13)(14)(15). Despite the fact that these cell division cycle genes are well conserved, it is known that diverse eukaryotic organisms differentially regulate their cell cycle machinery in response to environmental stimuli and other internal signaling pathways (16).
To understand better the molecular mechanisms that regulate progression of the cell cycle of P. carinii, we have sought to identify and characterize genes in P. carinii that are important regulators of cell cycle progression in related organisms. Accordingly, we began our investigations by demonstrating that P. carinii utilizes a cyclin-dependent kinase, Cdc2, over its life cycle (17). Activated Cdc2, in association with cyclin B (cdc13), phosphorylates a variety of targets, such as histones and nuclear laminins, to initiate mitosis (18,19). The kinase activity of Cdc2 is regulated by both inhibitory and stimulatory phosphorylations (20). One important positive regulator is the Cdc25 phosphatase that activates Cdc2 by removing the inhibitory phosphoryl group on Tyr 15 of Cdc2 (21,22). In fission yeast, Cdc25 protein levels and activity rise in the G 2 phase of the cell cycle (23). Therefore, at the G 2 /M border Cdc2 is able to phosphorylate and activate Cdc25, initiating an autofeedback loop resulting in a rapid entry into mitosis (24,25). In addition to mitotic control, Cdc25 has roles in regulating meiosis and imposing a checkpoint arrest in the DNA repair and DNA replication processes (26 -31). Thus, Cdc25 represents an integral component of eukaryotic cell cycle regulation.
Current models state that the phosphorylation of the Tyr 15 residue of Cdc2 is a key component of maintaining the S/M and G 2 /M checkpoints (32,33). DNA damage or unreplicated DNA activates the appropriate checkpoint pathway that propagates signals that lead to the removal of Cdc25 from the nucleus and/or inactivation of Cdc25 activity, thereby maintaining the Tyr 15 -phosphorylated state of Cdc2 (26, 31, 34 -37). In this study, we have identified and characterized a cDNA and corresponding genomic clone from P. carinii that is structurally and functionally homologous to the essential Cdc25 mitotic regulator. In the context of a heterologous fungal system, this homolog rescues growth in Cdc25 temperature-sensitive mutants and restores the DNA damage checkpoint. However, it is unable to restore fully the DNA replication checkpoint following hydroxyurea treatment. These findings (i) provide evidence for distinct molecular mechanisms regulating G 2 /M and S/M checkpoints previously thought to be under similar control, and (ii) identify a key regulatory molecule involved in P. carinii life cycle progression.

EXPERIMENTAL PROCEDURES
Materials-Nitrocellulose membranes containing separated P. carinii chromosomes were kindly provided by Dr. Melanie T. Cushion, University of Cincinnati College of Medicine (38). A P. carinii cDNA library was obtained from the National Institutes of Health AIDS Research and Reference Reagent program. The temperature-sensitive S. pombe cdc25-22 mutant strain and the pREP41 vector were the kind gifts of Dr. Kathleen Gould, Vanderbilt University. Dr. Barbara Painter generously provided Ciprofloxacin (Miles Pharmaceuticals, Inc., West Haven, CT). All other reagents were from Sigma unless stated otherwise.
P. carinii Isolation-All animal studies were approved by the Mayo Institutional Animal Care and Utilization Committee. Harlan Sprague-Dawley rats were immunosuppressed with dexamethasone and subsequently challenged with P. carinii sp. f. (special form) carinii via transtracheal injection to induce P. carinii pneumonia (39,40). P. carinii were purified by differential filtration through 10-m filters as we have previously reported (17).
Cloning of the P. carinii Cdc25 Mitotic Phosphatase-P. carinii genomic DNA was prepared (17,41,42), and degenerate oligonucleotide primers were synthesized to the conserved CH2A and CH2B domains of Cdc25 family members. The following primers were used for amplification, 5Ј-ATW ATW GAT TGT CGS TTG-3Ј and 5Ј-WGG ATA ATA SAA AAA WGG ATA-3Ј. PCR was performed as follows: 94°C for 2 min, followed by 30 cycles of 94°C for 1 min, 55°C for 1 min, and 72°C for 2 min, and a final extension at 72°C for 5 min. A single 280-bp fragment was amplified, subcloned into the pGEM-Teasy vector (Promega), and sequenced. The amplicon was radiolabeled with [␣-32 P]dATP (Amersham Pharmacia Biotech) with the RadPrime DNA Labeling System (Life Technologies, Inc.) and hybridized to a nitrocellulose membrane containing P. carinii chromosomes. The P. carinii chromosomes were separated by contour-clamped homogenous field electrophoresis (CHEF) and transferred to nitrocellulose (38). After 30 min of prehybridization in ExpressHyb (CLONTECH Laboratories, Inc., Palo Alto, CA), the CHEF blot was incubated with probe (1 ϫ 10 6 cpm/ml) at 68°C for 60 min, washed four times at room temperature in 2ϫ SSC, 0.05% SDS buffer, washed twice at 50°C for 40 min in 0.1ϫ SSC, 0.1% SDS, and visualized by autoradiography. In addition, the probe was used to screen a rat-derived P. carinii cDNA library under similar conditions. Clones were isolated after plaques were purified to homogeneity. A 2.1-kb clone was purified and sequenced (GenBank TM accession number AF098935). The genomic sequence (GenBank TM accession number AF098936) was determined by PCR amplification using freshly isolated P. carinii genomic DNA as the template. Primers made from the 5Ј end and the 3Ј end of the cDNA used for amplification are as follows: 5Ј-CATATATGGATACTTCACCTCTTG-3Ј and 5Ј-CGG-ACGTTACTCACATC-TTTTTGCAG-3Ј. PCRs were carried out as previously mentioned.
Production of Cdc25 Proteins in Bacteria and p-Nitrophenyl Phosphate Assays-The full-length cDNAs for the P. carinii Cdc25 homolog, an inactive mutant (C432A), and the S. pombe cdc25 were subcloned into the pGEX-2T vector (Amersham Pharmacia Biotech) and transformed into competent Escherichia coli BL21 cells. The bacteria were grown to an A 600 of 0.5 at 27°C in LB containing 100 g/ml ampicillin. Isopropylthio-␤-galactosidase was added for 4 h at a final concentration of 1 mM for BL21 cells harboring P. carinii Cdc25 and C432A expression plasmids and 0.2 mM for BL21 cells containing the S. pombe Cdc25 expression plasmid. The pelleted cells were suspended in phosphatebuffered saline containing 10 g/ml leupeptin, pepstatin, aprotinin, and phenylmethylsulfonyl fluoride. A final concentration of 100 g/ml of lysozyme and 1% Triton was added, incubated on ice for 15 min, lysed by mild sonication, and centrifuged to remove insoluble debris. The GST fusion proteins were purified from the soluble fraction using glutathione-Sepharose beads and eluted with 10 mM reduced glutathione, 50 mM Tris, pH 8.0. The inactivated P. carinii Cdc25 was created by mutating the Cys 432 residue in the active site to an Ala residue with the Stratagene Quikchange Site-directed Mutagenesis Kit ® . The following primers were used: C432A-5Ј (5Ј-GTTTGATTATTTTTCATGCTGAAT-ATAGTTCACATCGTGC-GCCA-3Ј) and C432A-3Ј (5Ј-GGCGCACGAT-GTGAACTATATTCAGCATG-AAAAATAATCAAAC-3Ј) with the mutated codon underlined. DNA sequencing verified the presence of the C432A mutation and the absence of other mutations. Reactions for p-nitrophenyl phosphate assays were performed as described by Dunphy and Kumagai (43). For GST-PcCdc25 purification, the eluted protein was loaded onto a Superose 12 (Amersham Pharmacia Biotech) gel filtration column, and fractions were collected and visualized by Coomassie-stained SDS-polyacrylamide gel electrophoresis. Fractions containing the fusion protein were subsequently loaded onto a Mono Q (Amersham Pharmacia Biotech) anion exchange column. Two buffers consisting of 25 mM Tris, pH 8.2, 1 mM dithiothreitol with and without 1 M NaCl were used to create a 40-ml gradient to elute the protein.
Complementation of a S. pombe cdc25 Temperature-sensitive Mutant Using P. carinii cdc25-An NdeI-BamHI fragment encoding the entire 1614-bp open reading frame encoding the P. carinii Cdc25 homolog was subcloned into NdeI-BamHI-digested pREP41 (containing a leu2 marker). The resulting construct contained the P. carinii gene in sense orientation under control of the thiamine-repressible nmt promoter. A 7-kb SalI genomic clone of the S. pombe Cdc25 was subcloned into SalI-digested pREP41 in an analogous fashion. The constructs, including pREP-C432A and pREP alone, were transformed into the S. pombe h ϩ cdc25-22 leu1-32 ura4-218 ade6-M210 strain as described (44) and selected on leucine-and thiamine-deficient media. Leucine-positive clones that grew at the permissive temperature (25°C) were then tested for growth at the restrictive temperature (35°C). The expression of the P. carinii Cdc25 was repressed by the addition of 25 M thiamine.
DNA Damage and DNA Replication Checkpoint Analysis with Pc-Cdc25-complemented cdc25-22 cells-For the DNA replication checkpoint studies, 12 mM hydroxyurea was added to asynchronous yeast cultures (1 ϫ 10 6 /ml) and fixed with 3% formaldehyde at selected time points and then stained with DAPI (45). Cells were analyzed by fluorescence microscopy and counted for cells containing mitotic cuts. For the DNA damage checkpoint studies, cells were synchronized in G 1 in media lacking nitrogen, released, and irradiated with 200 Gy from a 137 Cs source. Following fixation and DAPI staining, images were taken on an Olympus AX-70 fluorescence microscope and analyzed using the Metamorph software program (Universal Imaging Corp., West Chester, PA).

RESULTS
Cloning of the P. carinii Cdc25 Mitotic Phosphatase-A degenerative PCR approach was taken to determine whether P. carinii contained a Cdc25 homolog. Because the active site of Cdc25 family members is highly conserved across species, degenerate primers were designed from the CH2A and CH2B domains flanking the catalytic site. To optimize the efficiency of the degenerate primers, codon bias was used to reflect the adenine/thymine-rich (Ͼ65%) P. carinii DNA (46). A 280-base pair product was amplified from P. carinii genomic DNA. Sequencing the product revealed that the 280-base pair fragment was unique but homologous to cdc25 genes from other organisms (data not shown). To confirm the P. carinii origin of the amplicon, it was hybridized to rat-derived P. carinii chromosomes separated by contour-clamped homogenous field electrophoresis (CHEF). Under high stringency, the amplicon hybridized to a single chromosome from P. carinii and to a 2.3-kb P. carinii RNA transcript ( Fig. 1 and data not shown). Subsequent screening of a P. carinii cDNA library identified a 2.1-kb fragment that was isolated and characterized. Based on Northern data, this cDNA clone appears to contain the entire coding region and most of the surrounding regulatory regions. The genomic clone was isolated by PCR amplification from freshly isolated P. carinii genomic DNA. Comparison of the cDNA and genomic sequences indicated that the coding region is interrupted by five introns that span 40 -50 base pairs in length and contain the GT/AG intron/exon junction sequence (GenBank TM accession numbers AF098935 and AF098936, respectively). Fig. 2 shows the predicted amino acid alignment of P. carinii Cdc25 with Cdc25 proteins from other eukaryotic species. Consistent with other Cdc25 family members, PcCdc25 exhibits a highly conserved C terminus containing the catalytic site and a variable N terminus. The PcCdc25 protein was found to be most closely related to fission yeast, being 61.2% homologous and 40.3% identical to the S. pombe Cdc25. The PcCdc25 homolog contained the HCXXXXXR consensus sequence (amino acids 431-438) and the conserved DCR motif at amino acids 389 -391 unique to Cdc25 phosphatases (47). Although the homolog does not contain an LIGD motif found in higher eukaryotes, it has three putative Cdc2 consensus phosphorylation sites on the N terminus at amino acids 91-93, 112-114, and 141-143 (47). These results demonstrate that the isolated cDNA is from P. carinii and exhibits significant homology to other Cdc25 species.
The P. carinii cdc25 Transcript Is Specific for P. carinii-One of the major concerns in cloning genes from Pneumocystis is host cell or other microbial contamination during the isolation of the organism. Whereas the CHEF blot provides substantial evidence for P. carinii origin, these preparations likely contain minor amounts of host material. As such, to ensure specificity, a PCR-based approach was taken to determine whether specific primers designed from one cdc25 gene would amplify a cdc25 gene from another organism. Primers were designed near the 5Ј end of the putative P. carinii cdc25 ϩ homolog, rat cdc25 ϩ , and S. pombe cdc25 ϩ genes, reflecting the region of the proteins that are the least conserved (see "Experimental Procedures"). RT-PCR was performed with RNA isolated from normal rat lung, P. carinii-infected rat lung, isolated P. carinii, and S. pombe. The P. carinii cdc25 ϩ primers amplified products from only P. carinii-infected lungs and isolated P. carinii (Fig. 3A). Rat cdc25 ϩ primers only generated a product from rat lung and P. carinii-infected rat lung, whereas S. pombe cdc25 ϩ primers specifically produced a transcript from S. pombe cDNA. By having verified the origin of the cDNA, we wished to determine the gene expression pattern in the two major life forms of P. carinii. From Northern analysis, the P. carinii cdc25 ϩ transcript was expressed in both the trophic and cyst form, with a moderate increase in expression in the cysts (Fig. 3B). These results, in addition to the CHEF hybridization and nucleotide and amino acid sequence differences, clearly demonstrate the Cdc25 gene product is from P. carinii and is expressed in both life cycle forms.
PcCdc25 Exhibits in Vitro Phosphatase Activity-Cdc25 proteins are dual-specific phosphatases known to catalyze the transition from various cell cycle checkpoints (25,28,48,49). To determine whether the P. carinii Cdc25 homolog had similar enzymatic activity as that seen for other Cdc25 proteins, we expressed the protein fused to a glutathione S-transferase (GST) domain at the N terminus in E. coli. Phosphatase activity was measured by the ability to dephosphorylate p-nitrophenol phosphate and compared with GST fusion proteins of human Cdc25C and S. pombe Cdc25. As shown in Fig. 4, the kinetic parameters of the partially purified GST-PcCdc25, GST-SpCcdc25, and the GST-HuCdc25C were similar, with K m values of 17, 26, and 35 mM, and V max values of 6, 3, and 31 nmol⅐min Ϫ1 ⅐mg Ϫ1 , respectively. Similar kinetics have been re- To ensure that the phosphatase activity observed in Fig. 4 reflected an enzymatic activity of GST-PcCdc25 and not some bacterial contaminant, a Cys 432 3 Ala mutation was introduced in the predicted catalytic domain of GST-PcCdc25. This mutation abolishes phosphatase activity by deletion of the active site nucleophile (43). As shown in Fig. 4, the introduced mutation abolished P. carinii Cdc25 activity. In addition, the GST-PcCdc25 fusion protein was purified to Ͼ80% purity using a gel filtration and ion exchange (Mono Q) chromatography (data not shown). The resulting purification resulted in a 7-fold increase in the specific activity of the protein. Moreover, the K m of the Mono Q-purified enzyme (average of two preparations) remained very close to the partially purified protein (15 versus 17 mM, respectively), but the V max increased from 6 to 15 nmol⅐min Ϫ1 ⅐mg Ϫ1 . These data not only documented that the phosphatase activity was due to the added fusion protein, but strongly suggested that GST-PcCdc25 has a similar active site to other Cdc25 proteins.
P. carinii Cdc25 Rescues cdc25-22 Temperature-sensitive Mutants Thereby Restoring Growth-The absence of defined genetics and culture systems makes it difficult to do genetic manipulations in P. carinii. As such, it is presently not possible to examine directly the role of the P. carinii Cdc25 homolog in P. carinii proliferation. However, to overcome that limitation, we determined whether the P. carinii Cdc25 cDNA would complement a temperature-sensitive Cdc25 S. pombe mutant (cdc25-22). This strain grows normally at a permissive temperature of 25°C, but at the restrictive temperature of 35°C the thermo-labile endogenous Cdc25 is no longer active and the strain arrests at the G 2 /M phase border (48). The PcCdc25 cDNA, the inactive PcCdc25(C432A) mutant, and the SpCdc25 genomic clone were subcloned downstream of a thiamine-re- pressible nmt promoter (50) and transformed into cdc25-22 cells. Individual clones containing the pREP expression vector were then selected on leucine-deficient minimal media plates and transformants shifted to the restrictive temperature to measure their ability to restore growth. As shown in Fig. 5, although the pREP vector alone and the PcCdc25(C432A) mutant did not restore growth at the restrictive temperature of 35°C, complementation was observed in the S. pombe clone containing the PcCdc25 homolog (Fig. 5C). Moreover, clones grown in the presence of thiamine, which represses expression of the PcCdc25 cDNA, were unable to grow (Fig. 5D). The growth of pREP-SpCdc25 in the presence of thiamine (Fig. 5D, panel 2) reflects the intact promoter activity upstream of the genomic S. pombe Cdc25 sequence. Therefore, in the context of a heterologous fungal system, the P. carinii Cdc25 homolog is able to initiate mitosis and support fungal growth.
PcCdc25 Restores the DNA Damage Checkpoint but Not the DNA Replication Checkpoint in cde25-22 Cells-Checkpoint pathways in eukaryotic cells ensure that genomic integrity is maintained in response to environmental and genotoxic stress (51). In fission yeast, the entry into mitosis is blocked when DNA synthesis is incomplete or when DNA is damaged by such agents as ionizing radiation (52). The G 2 /M checkpoint is imposed by a signal transduction system that ultimately leads to the sequestration of the Cdc25 protein into the cytoplasm (mediated by Rad24, a 14-3-3 protein) and/or inactivation of Cdc25 activity. These events prevent the phosphatase from activating Cdc2 (34 -36, 49). S. pombe Cdc25 has three canonical 14-3-3binding sites, Arg-Ser-Leu-Ser 99 -Cys-Thr, Arg-Arg-Thr-Gln-Ser 359 -Met-Phe, and Arg-Ser-Arg-Ser 192 -Ser-Gly, that are phosphorylated by Chk1 and Cds1 in response to DNA damage and incomplete DNA replication (26,36,49). In contrast to S. pombe Cdc25, we observed that the P. carinii Cdc25 contained only one similar consensus 14-3-3-binding site at serine 314 (Arg-Arg-Thr-Gln-Ser 314 -Leu-Tyr). As such, it was unclear whether a complete checkpoint response could be provided by PcCdc25. To that end, we subjected PcCdc25-complemented cdc25-22 cells to ionizing radiation (IR) to determine whether the P. carinii Cdc25 restored the DNA damage checkpoint. The DNA damage checkpoint was measured by determining the number of cells passing mitosis following irradiation. G 1 -synchronized cells were exposed to 200 Gy of ionizing radiation and harvested over the next 12 h. As shown in Fig. 6A, SpCdc25 and PcCdc25-complemented yeast show the characteristic G 2arrested elongated phenotype following exposure to IR. When FIG. 3. Pc cdc25 ؉ transcript is specific for P. carinii. A, RT-PCR was performed with primers made near the 5Ј end of P. carinii cdc25 ϩ , rat cdc25 ϩ , and S. pombe cdc25 ϩ . Templates are as follows: U, uninfected rat lung; I, P. carinii-infected rat lung; Pc, isolated P. carinii; Sp, S. pombe. The cDNA templates were made from 1 g of total RNA primed with oligo(dT) primers. The same template was used for each of the different primer sets, and the appropriate size fragment was amplified from each condition. P. carinii cdc25 ϩ , 640 bp; rat cdc25 ϩ , 652 bp; and S. pombe cdc25 ϩ , 604 bp. B, Northern blot analysis of 5 g of total RNA isolated from the trophic (Troph) and cyst forms of P. carinii. The ethidium bromide-stained gel (EtBr) shows the ribosomal subunits of P. carinii. The blot was hybridized with the radiolabeled 280-bp amplicon. the number of septated cells (i.e. cells that have past mitosis and thus nonarrested) were quantified, SpCdc25-complemented cultures showed a significant decrease in the number of septated cells starting at 4 h and recovering around 12 h (Fig.  6B). A similar result held true for the PcCdc25-complemented yeast in that a delay into mitosis was observed. Although the kinetics of the delay were somewhat slower, the results demonstrate that in response to ionizing radiation, PcCdc25 complemented yeast arrest similar to that observed with the endogenous S. pombe Cdc25 protein. Thus, the DNA damage checkpoint is intact in cells harboring PcCdc25.
Since the DNA damage and DNA replication checkpoints are proposed to signal through Cdc25 by Chk1 and Cds1 phosphorylation, we determined whether PcCdc25-complemented clones would similarly maintain a checkpoint arrest in response to inhibitors of DNA replication. DNA synthesis was inhibited by treating cultures with hydroxyurea (HU), an agent that inhibits the function of ribonucleotide reductase. Yeast that cannot maintain the DNA replication checkpoint enter mitosis with incompletely replicated DNA. This results in an unequal distribution of DNA as cells pass through mitosis and septate, therefore generating cells that contain little or no DNA (referred to as mitotic "cut" cells). As shown in Fig. 7A, SpCdc25-complemented yeast grown in 12 mM HU maintained the replication checkpoint and generated very few cells exhibiting the mitotic cut phenotype. However, unlike HU-treated SpCdc25-complemented yeast, a significant fraction of the Pc-Cdc25-complemented cells grown in 12 mM HU bypassed the S/M checkpoint and entered into mitosis. Quantification of mitotic cut cells showed that SpCdc25-complemented cells grown in HU maintained the DNA replication checkpoint in that only 1% of cells exhibited the mitotic cut phenotype at 6 h (Fig. 7B). This is contrasted by the PcCdc25-complemented cells grown in HU where the number of cut cells increased to 8.0% at 8 h. This result is similar to the 9-fold induction of mitotic cut cells observed with SpCdc25-S3 yeast that have a diminished replication checkpoint response due to serine mutations in the three 14-3-3-binding sites (36). We therefore conclude that PcCdc25 does not restore the DNA replication checkpoint. Although the SpCdc25-S3 mutation is unable to FIG. 6. The DNA damage checkpoint is restored in PcCdc25-complemented cdc25-22 cells. A, photomicrographs of the cdc25-22: SpCdc25 cells exposed to 200 Gy of radiation, unexposed cdc25-22:PcCdc25 cells, and cdc25-22:PcCdc25 cells exposed to 200 Gy of radiation. After G 1 release, cells were exposed to IR versus unexposed, fixed after 8 h, and stained with DAPI for fluorescence microscopy analysis. Each panel was taken at ϫ 1000 magnification, and the white bars represent 10 microns. The arrows indicate septated cells that represent dividing, nonarrested cells. B, quantitative analysis measuring the number of septated cells following radiation exposure or no treatment. At indicated time points, the cells were fixed and mounted. The graph on the right represents PcCdc25-complemented clones, and the graph on the left represents SpCc25complemented clones. The % septated cells were determined in unexposed cultures (f) and IR-exposed cultures (q). Numbers were derived from two clones analyzed in four separate experiments with at least 100 cells counted per time point. complement either the DNA replication or DNA damage checkpoints (34,36), we demonstrate that the PcCdc25 protein differentially responds to these genotoxic assaults. The data indicate that the pathways that eukaryotic organisms employ in response to DNA damage and incomplete DNA synthesis can be dissociated at the level of Cdc25 regulation. DISCUSSION Our laboratory has previously characterized roles for P. carinii Cdc2 and its cognate partner Cdc13 in cell cycle regulation. In this report, we have continued our approach to define key mediators of this system in P. carinii. A 2.1-kb cDNA and corresponding genomic clone has been isolated from P. carinii that shows greatest homology to the S. pombe cdc25 gene ( Figs.  1 and 2). The predicted amino acid sequence contains a consensus HCXXXXXR motif in the active site and many other consensus sites found in the Cdc25 family (47). We show that the cdc25 ϩ transcript is specific for P. carinii and is expressed in both the trophic and cyst life cycle forms (Fig. 3). The Cdc25 homolog contains in vitro phosphatase activity and demonstrates similar kinetic parameters to other Cdc25 homologs (Fig. 4). Since molecular genetics in P. carinii are currently not feasible, we examined the function of PcCdc25 in a heterologous fungal system with a temperature-sensitive deficiency of endogenous Cdc25. The results demonstrate that the PcCdc25 cDNA rescues the defect and supports growth of the cdc25-22 strain (Fig. 5). Therefore, in fission yeast, PcCdc25 provides the necessary function(s) for complete cell cycle progression. Interestingly, yeast cells containing the P. carinii Cdc25 homolog demonstrated differential responses to the G 2 /M and S/M checkpoints. Although the PcCdc25-complemented cdc25-22 yeast maintained the G 2 /M checkpoint in response to damaged DNA by IR (Fig. 6), the S/M DNA replication checkpoint, as tested with HU, was impaired (Fig. 7). Thus, PcCdc25 is capable of dissociating the signals regulating normal proliferation and checkpoint control.
Although the P. carinii life cycle has been well characterized morphologically, the molecular events regulating life cycle progression have only recently begun to be defined (17,(53)(54)(55). It is evident that whereas cell division cycle molecules have key roles in the proliferation of P. carinii, the unique requirements of P. carinii for life cycle progression suggest other regulatory mechanisms. For instance, P. carinii attachment to type I pneumocytes promotes proliferation, whereas Candida albicans, Saccharomyces cerevisiae, and S. pombe propagate independent of binding to a substrate. Furthermore, P. carinii progresses through a cyst form that is critical for survival and necessary for life cycle progression (54,56). As such, when P. carinii-infected rats are treated with ␤-glucan synthesis inhibitors to prevent cyst formation, the infection is eliminated (56). This is in contrast to budding and fission yeast haploid forms that utilize a mitotic cycle independent of forming an ascus, unless environmental conditions are unfavorable (57). Although these differences need to be considered as the P. carinii life cycle is investigated, a common feature found in all eukaryotic proliferation is a central regulatory role for cell division cycle gene products (16). Therefore, it is very likely that these molecules are intimately involved in similarly coordinating P. carinii life cycle progression.
In response to environmental stimuli, signal transduction pathways are activated that eventually impact on the cell cycle machinery. Cdc25 is a key regulator of several cellular processes including regulating entry into mitosis, meiotic phase transitions, and maintaining G 2 /M and S/M checkpoints in the response to DNA damage and incomplete DNA replication (26 -31). The protein complexes that sense DNA damage or stalled DNA replication forks transmit signals that ultimately lead to the inactivation of Cdc2 (32, 33, 37). One such mechanism includes the phosphorylation of Cdc25 by Chk1 and Cds1 ki- nases. Phosphorylation of Cdc25 results in the inactivation and/or sequestration of Cdc25 to the cytoplasm thereby maintaining Cdc2 in the inactive Tyr 15 -phosphorylated state (26, 31, 34 -36).
The observation that PcCdc25 restored the DNA damage ( Fig. 6) but not the DNA replication checkpoint (Fig. 7) indicates a dissociation of checkpoint pathways at the level of Cdc25 regulation. In response to IR, PcCdc25-complemented yeast exhibited a decrease in the number of cells passing mitosis and an increase of cells in G 2 (elongated phenotype). Although the DNA damage checkpoint is dampened in fission yeast harboring a S99A mutation in Cdc25 (34), PcCdc25 restored the checkpoint even though it lacks the appropriate context of this regulatory site found in S. pombe Cdc25. Although PcCdc25 can restore the checkpoint response to IR (Fig.  6), following treatment with HU, PcCdc25-complemented yeast exhibited an 8-fold increase in mitotic cut cells when compared with untreated cells. The impaired S/M checkpoint elicited by PcCdc25 in response to inhibitors of DNA replication is consistent with that observed in yeast containing a Cdc25 protein in which the 14-3-3-binding/Chk1 and Cds1 phosphorylation sites have been mutated (SpCdc25-S3) (31). Although Chk1 appears to be the major kinase involved in the response to DNA damage and Cds1 in the incomplete DNA replication checkpoint, they both appear to phosphorylate Cdc25 at the same residues (Ser 99 , Ser192, and Ser 359 ) in fission yeast (34).
The definitive role of the three Chk1/Cds1 phosphorylation sites (Ser 99 , Ser 192 , and Ser 359 ) has yet to be clarified. Whereas Zeng et al. (36) analyzed the HU-induced S/M checkpoint with a Cdc25 containing Ser 3 Ala mutations of the three sites (SpCdc25-S3), they did not dissect the role of each site individually in maintaining the arrest nor was the response to IR examined. Although a SpCdc25-S99A mutant generated by Furnari et al. (34) was found to impair both the S/M and G 2 checkpoints, the role of the other two mutations was not determined. Moreover, results from both studies suggest that there may be other Cdc25 sites weakly phosphorylated by Cds1 or Chk1. It remains unclear whether these additional sites have functional significance. Finally, the findings that (i) these Cdc25 mutations do not completely abolish the checkpoints and (ii) Mik1 and Wee1 kinases phosphorylate and inhibit Cdc2 activity in response to DNA damage and unreplicated DNA (36,58,59) indicate that multiple mechanisms cooperate with Cdc25 to maintain genomic integrity.
It is surprising that PcCdc25 would only restore one checkpoint pathway since cell cycle proteins are highly conserved and can function normally in heterologous systems. Since Pc-Cdc25 complements growth in fission yeast, the host machinery must recognize sites of control in the primary sequence and secondary structure of PcCdc25. Furthermore, this differential control cannot easily be explained by the fact that PcCdc25 is overexpressed in yeast (due to the nmt promoter) in that cdc25-22 yeast containing overexpressed SpCdc25 maintain both checkpoints. The data provide initial evidence for distinct domains on Cdc25 family members or other unknown checkpoint regulatory sites that differentially regulate the response to IR and HU. Further studies will identify the sequence(s) controlling this differential response in checkpoint control and determine the manner which Chk1, Cds1 and/or Rad 24 interact with PcCdc25.
Our current studies continue to be focused on the role of the P. carinii Cdc25 homolog in regulating the cyst/trophic form transitions. The data presented (Fig. 3B) show that PcCdc25 is expressed in both life forms. We hypothesize that PcCdc25 will have a vital role in regulating organism replication and life cycle transitions. Furthermore, cdc25 gene expression in the trophic forms provides evidence for the model in which trophic forms might undergo their own mitotic cycle. Although these studies are limited by the ability to manipulate genetically the organism, the further development of an axenic culture will be necessary to address this problem. It is anticipated that further studies of the P. carinii Cdc25 homolog in the context of life cycle progression will provide new insights to understanding P. carinii pathogenesis and Cdc25 biology in checkpoint control.