PINA is essential for growth and positively influences NIMA function in Aspergillus nidulans.

The phospho-Ser/Thr-directed prolyl-isomerase Pin1 was originally identified in vertebrate systems as a negative regulator of NIMA, a Ser/Thr protein kinase that regulates the G(2)/M transition in Aspergillus nidulans. Here we explore the physiological roles of the Pin1 orthologue, PINA, in A. nidulans and evaluate the relevance of the interaction of PINA with NIMA in this fungus. We find pinA to be an essential gene in A. nidulans. In addition, when PINA levels are reduced 50-fold the cells grow at a reduced rate. Upon germination under conditions that repress PINA expression, the cells are delayed in the interphase activation of NIMX(cdc2), whereas they traverse the other phases of the cell cycle at a similar rate to controls. These results indicate that a marked reduction of PINA results in a lengthening of G(1). Additionally, PINA repression increases the rate at which the cells enter mitosis following release from a hydroxyurea arrest without altering the sensitivity of the fungus to agents that activate the replication or DNA damage checkpoints. In contrast to predictions based on Pin1, the physical interaction between PINA and NIMA is primarily dependent upon the prolylisomerase domain of PINA and the C-terminal 303 amino acids of NIMA. Finally, reduction of PINA levels exacerbates the nimA5 temperature-sensitive mutant, whereas overexpression of PINA decreases the severity of this mutation, results that are consistent with a positive genetic interaction between PINA and NIMA. Thus, although PINA is essential and positively regulates NIMA function, A. nidulans is most sensitive to a reduction in PINA concentration in G(1) rather than in G(2)/M.

The phospho-Ser/Thr-directed prolyl-isomerase Pin1 was originally identified in vertebrate systems as a negative regulator of NIMA, a Ser/Thr protein kinase that regulates the G 2 /M transition in Aspergillus nidulans.
Here we explore the physiological roles of the Pin1 orthologue, PINA, in A. nidulans and evaluate the relevance of the interaction of PINA with NIMA in this fungus. We find pinA to be an essential gene in A. nidulans. In addition, when PINA levels are reduced 50-fold the cells grow at a reduced rate. Upon germination under conditions that repress PINA expression, the cells are delayed in the interphase activation of NIMX cdc2 , whereas they traverse the other phases of the cell cycle at a similar rate to controls. These results indicate that a marked reduction of PINA results in a lengthening of G 1 . Additionally, PINA repression increases the rate at which the cells enter mitosis following release from a hydroxyurea arrest without altering the sensitivity of the fungus to agents that activate the replication or DNA damage checkpoints. In contrast to predictions based on Pin1, the physical interaction between PINA and NIMA is primarily dependent upon the prolylisomerase domain of PINA and the C-terminal 303 amino acids of NIMA. Finally, reduction of PINA levels exacerbates the nimA5 temperature-sensitive mutant, whereas overexpression of PINA decreases the severity of this mutation, results that are consistent with a positive genetic interaction between PINA and NIMA. Thus, although PINA is essential and positively regulates NIMA function, A. nidulans is most sensitive to a reduction in PINA concentration in G 1 rather than in G 2 /M.
Pin1 is an evolutionarily conserved member of the parvulin family of proline isomerases that selectively targets proteins containing proline preceded by phospho-Ser/Thr residues (reviewed in Refs. 1 and 2). The phospho-epitopes recognized by Pin1 show remarkable sequence similarity with that of the mitosis-specific antibody, MPM-2, and are frequently generated by cdk2 or mitogen-activate protein kinase-dependent phosphorylation (3). Based on Pin1 interaction studies as well as predictions based on substrate specificity and analysis of Pin1 function in multiple eukaryotic systems, Pin1 has been implicated in numerous signaling pathways critical for cell proliferation. However, the phenotypic consequences of Pin1 deletion, depletion, or inhibition vary greatly among organisms. Whereas Pin1 is essential for growth in Saccharomyces cerevisiae and Candida albicans, only subtle phenotypes are observed when Pin1 is deleted in Schizosaccharomyces pombe, and quite specific ones occur in Drosophila and mouse (4 -12).
As might be predicted from its substrate specificity, Pin1 has been demonstrated to play an important role in regulating both mitosis and G 1 . Indeed, temperature-sensitive mutants of the S. cerevisiae and C. albicans Pin1 homologue, Ess1p, arrest cells in mitosis (5,13). Furthermore, expression of Pin1 antisense RNA in HeLa cells has been reported to induce either a mitotic arrest (14) or an interphase arrest with high levels of apoptosis (15). However, at least in the case of S. cerevisiae, the mitotic role of Ess1p appears to result from transcriptional alterations rather than via direct interaction with critical mitotic regulatory proteins (13). Regardless, Pin1 has been demonstrated to interact with a number of signaling molecules involved in mitotic progression, including cdc25, wee1, plk1, and NIMA (14,16,17). In contrast to the mitotic arrest observed when Pin1 is depleted from HeLa cells, MEFs 1 derived from the Pin1 null mouse display no detectible mitotic phenotype but, rather, proliferate slightly more slowly than controls and are impaired in their ability to re-enter the cell cycle after serum starvation (9,18). In support of a G 1 role for Pin1, it is required for the proper timing of primordial germ cell proliferation due to a lengthening of G 1 and has been reported to positively regulate c-Jun, ␤-catenin, and cyclinD1, all of which are involved in G 1 progression (10, 11, 19 -21).
Although Pin1 has been demonstrated to interact with and regulate a number of phospho-proteins in numerous systems, it was initially identified as an interactor and suppressor of the lethality induced by overexpression of the Aspergillus nidulans essential G 2 /M regulatory protein kinase, NIMA, in a yeast twohybrid screen (14). The identification of Pin1 as a negative regulator of NIMA was supported by the observations that not only is the mitotic arrest induced by NIMA overexpression in HeLa cells abrogated by coexpression of Pin1, but overexpression of Pin1 in the absence of active NIMA arrests the cells in G 2 . As in HeLa cells, NIMA overexpression in Aspergillus induces a pseudomitotic state characterized by microtubule depolymerization and chromatin condensation. On the other hand, the repression of NIMA activity arrests cells in G 2 (22). Thus, the twohybrid and HeLa cell results are consistent with the notion that Pin1 could function as a negative regulator of NIMA and a NIMA-like signaling pathway in mammalian cells.
To understand the in vivo relevance of the interaction of NIMA with Pin1 we have evaluated the function of the A. nidulans Pin1 homologue, PINA. Similar to the results in the yeasts S. cerevisiae and C. albicans, disruption of PINA in A. nidulans is lethal. However, using an inducible expression system, we find that a 50-fold reduction of PINA expression is sufficient to allow growth, albeit at a markedly slower rate than the wildtype strain. In contrast to the mitotic arrest due to ess1ts mutants, yet similar to defects in Pin1 null cells, PINA depletion delays the timing of G 1 progression. Furthermore, we identify a potential role for PINA in regulating the recovery from a replication checkpoint arrest, because cells with reduced levels of PINA enter mitosis more quickly than controls following release from a hydroxyurea-induced arrest without sensitizing the cells to the drug. Finally, in contrast to predictions of earlier studies in S. cerevisiae and HeLa cells, PINA genetically interacts with and is a positive regulator of NIMA kinase function in A. nidulans.  (23). Benomyl (Sigma-Aldrich) and hydroxyurea (Sigma-Aldrich) were used at final concentrations of 5 g/ml and 20 mM, respectively. All genetic crosses were performed as described by Pontecorvo (24).
Generation of A. nidulans Vectors and Transformation-pPINAdis was generated by cloning a portion of the PINA gene containing nucleotides 55-552 (equivalent to 55-499 of cDNA, generated by PCR of the genomic PINA clone) into the EcoRI and SmaI sites of pRG1 (26). The pAlcPINA vector was generated by cloning the 5Ј end of the PINA gene into the SmaI site of pAL3 as a blunted HindIII/XmnI fragment (26). All vectors were sequenced prior to use.
A. nidulans GR5 cells were transformed as described by Lu and Means (23). Positive transformants were selected by growth on minimal medium in the absence of uracil and uridine to select for the presence of the pyr4 nutritional marker. Transformants of pPINAdis were maintained as heterokaryons by transfer of mycelia instead of spores. Southern analysis was performed on vegetatively growing mycelia as described by Rasmussen et al. (27). Transformants of pAlcPINA were streaked three times to ensure strain purity followed by Southern analysis and Western analysis to ensure homologous integration and protein expression.
Microscopy and Growth Assays-Nuclear number and staining were performed by growing spores in minimal medium on glass coverslips at a concentration of 10 4 spores/ml. Germlings were collected at various times after germination followed immediately by fixing and staining with 4Ј,6-diamidino-2-phenylindole or Hoechst 33258 as described by Harris et al. (28). Chromosome mitotic indices were determined using cells germinated in liquid medium at 30°C and 200 rpm for ϳ5 h followed by cell cycle arrest via temperature shift to 42°C or the addition of 20 mM hydroxyurea for 3 h. Cells were then released into fresh medium or 30°C, and samples were fixed and stained and scored for mitotic nuclei as described by Lu and Means (23).
Western Blotting and Immunoprecipitation-Rabbit anti-PINA polyclonal antibodies were generated using keyhole limpet hemocyaninconjugated, bacterially expressed PINA as the antigen. The coupled protein was injected into rabbits, and antiserum was collected using standard techniques (29). Western analysis of PINA was performed using a 1:5000 dilution of the anti-PINA antisera and detected using either 125 I-Protein A (Amersham Biosciences) at a concentration of 10 5 cpm/ml or HRP-conjugated anti-rabbit IgG. Western analysis of phospho-Ser-10 histone H3 was performed as described by DeSouza et al. (30). Western analysis of HA-NIMA was performed using the 12CA5 monoclonal antibody (Roche Applied Science) at a dilution of 1:1000 followed by detection using HRP-conjugated anti-mouse IgG. MPM-2 reactive epitopes were detected using 1 g/ml MPM2 monoclonal antibody (Upstate Cell Signaling) followed by detection via HRP-conjugated anti-mouse IgG.
HA-NIMA and PINA were immunoprecipitated from A. nidulans extracts prepared by grinding flash frozen germlings prepared as described by Dayton et al. (31). For immunoprecipitation of both PINA and NIMA 500 g of protein was brought to a total of 500 l with lysis buffer followed by the addition of the primary antibody (2 l of anti-PINA and 1 g anti-HA). Following rocking for 1 h at 4°C, ϳ20 l of protein A-or G-Sepharose (Amersham Biosciences) was added, and the mixture was incubated for an additional hour. Following binding, the beads were washed five times in lysis buffer. The bound proteins were eluted with the addition of 30 l of SDS-PAGE protein loading buffer and detected by Western analysis following SDS-PAGE electrophoresis and transfer to Immobilon P membrane (Millipore).
NIMA and NIMX Kinase Assays-Extracts for NIMX cdc2 histone H1 kinase were prepared and assays were performed as described previously (33). 32 P labeled phosphate incorporation of SDS-PAGE separated reactions was quantified using a Molecular Dynamics PhosphorImager.
Full-length, bacterially expressed NIMA used for the in vitro kinase assays was generated by first cloning the NIMA cDNA into pET30 expression vector (Novagen) as an NcoI/SalI fragment. NIMA was expressed in freshly transformed BLR(DE3)pLysS competent cells (Novagen) grown at 37°C to an A 600 nm of 0.6 at which point the cultures were shifted to room temperature, and protein expression was induced with the addition of 0.4 mM isopropyl-1-thio-␤-D-galactopyranoside for 3 h. Following centrifugation the bacteria were resuspended in lysis buffer (25 mM Tris HCl, pH 8.0, 1 mM DTT, 5 mM EDTA, 1 mg/ml Pefabloc) and lysed by sonication. Following sonication, Triton X-100 was added to a final concentration of 0.5%, and the mixture was incubated on ice for 30 min. The extract was then clarified by centrifugation and incubated with nickel nitrilotriacetic acid-agarose (Qiagen) for 1 h with rocking at 4°C. The bound resin was then washed with 20 ml of lysis buffer plus 0.5% Triton X-100, followed by 20 ml of lysis buffer plus 0.5% Triton X-100, 150 mM NaCl, and 10 mM imidazole. The bound protein was eluted in lysis buffer containing 250 mM imidazole and stored at Ϫ80°C in 40% glycerol. Recombinant PINA was expressed, purified, and cleaved from the glutathione S-transferase tag as described by Crenshaw et al. (17).
NIMA kinase assays were performed using 100 ng of recombinant NIMA or NIMA immunoprecipitated from 500 g of A. nidulans extract using either the anti-HA 12CA5 monoclonal antibody or anti-HA 3F10conjugated Sepharose. Immunoprecipitation was as described above with two additional washes with kinase buffer (50 mM Tris HCl, pH 8.0, 1 mM DTT, 0.1% Triton X-100, 10 mM MgCl 2 ). Conditions for the kinase assay were as follows: 50 mM Tris HCl, pH 8.0, 1 mM DTT, 0.1% Triton X-100, 10 mM MgCl 2 , 500 M ATP, 1 mM F-peptide (GRFRRSRRMI), 0.2 l/reaction [␥-32 P]ATP. Reactions were performed in 30 l (for recombinant NIMA) or 50 l (for immunoprecipitated NIMA) volumes for 10 min and terminated by spotting 20 l of the reaction onto p81 phosphocellulose filters followed by extensive washing in 75 mM phosphoric acid. Activities were determined followed scintillation counting of the dried filters. When applicable purified, cleaved PINA or BSA was added to the kinase, and the mixture was incubated for 20 min prior to the initiation of the kinase assay.
Phosphatase assays were performed at 30°C using purified PP2A (Calbiochem) in the 50 mM Tris HCl, pH 7.5, 1 mM EDTA, 0.1% ␤-mercaptoethanol, and 2 mM MnCl 2 at a final concentration of ϳ0.02 unit of PP2A/50 g of NIMA. Time courses of dephosphorylation were terminated either by the addition of protein loading buffer for observation of phosphorylation or the addition of 0.5 M okadaic acid followed by kinase assay for activity measurements. Either BSA or PINA was incubated with the recombinant NIMA in assay buffer 20 min prior to the initiation of reaction.
Hydroxyurea and UV Sensitivity Assay-Hydroxyurea sensitivity of the AlcPinA and control strains was determined by spotting 1 l of ϳ5 spores/ml onto glycerol or glucose minimal medium plates containing graded concentrations of hydroxyurea of up to 10 g/ml. Growth was compared 3 days after germination at 37°C. UV sensitivity of the AlcPinA and control strains was compared by UV irradiating germlings 4 h post-germination in minimal medium plus dextrose or glycerol using a Stratalinker UV cross-linker (Stratagene) at various intensities. Following UV irradiation the ϳ500 germlings/plate were plated on minimal medium plates containing either glucose or glycerol and 0.01% Triton X-100. Percent survival was determined after 4 days of growth at 32°C.

RESULTS
PINA Is Essential in Aspergillus nidulans-To address the in vivo function of PINA in A. nidulans, we attempted to disrupt the endogenous pinA gene by homologous recombination in the haploid fungus. A schematic representation of the disruption strategy is presented in Fig. 1A. Despite screening over 150 integrated strains by Southern analysis, we were unable to identify any strains in which the endogenous pinA gene was disrupted. Because A. nidulans normally grows in the haploid state the disruption of an essential gene would result in non-viable progeny, thus the inability to identify viable pinA disruptants is consistent with pinA being an essential gene.
To confirm that pinA is essential in A. nidulans we disrupted the gene in a heterokaryon in a manner similar to that described for cmkA, cmkB, CaM, and cnaA (27,(33)(34)(35). As diagramed in Fig. 1B, heterokaryon strains can be generated by the disruption of a gene following DNA synthesis or by the fusion of the parental strain with a strain containing the targeted disruption. By this method, heterokaryon strains in which essential genes are disrupted are maintained by the presence of the selectable nutritional marker in the disrupted nucleus and the presence of the wild-type targeted allele in the parental nucleus. The heterokaryon is then propagated by transfer of multinucleate mycelia. When pinA-targeted transformants were maintained as heterokaryons, multiple strains were identified in which pinA was disrupted as indicated by Southern hybridization bands at ϳ0.9 kbp (representing the wild-type pinA allele) and 3.0 kbp (representing the pinAdisrupted allele), three examples of which are presented in Fig.  1C. As represented in Fig. 1B the sporulation of a heterokaryon strain generates two populations of spores, one representing the maternal strain, which contains a wild-type allele of the targeted gene but no nutritional marker (in our case pyr4 Ϫ / pinA ϩ ), whereas the second contains both the disrupted gene and the selectable nutritional marker (pyr4 ϩ /pinA Ϫ ). Thus, if pinA is an essential gene, inoculation of the heterokaryon progeny in medium without uracil and uridine would yield no viable germlings, and in medium containing the required nutrients only the spores representing the maternal nuclei will germinate. When the pinA-disrupted heterokaryons were grown in the absence of uracil and uridine, Ͻ1% of the spores germinated. However, in the presence of uracil and uridine, ϳ54% of the spores underwent multiple rounds of nuclear division and initiated polarized growth within 10 h, and the remaining spores germinated but arrested with a single nuclei and failed to initiate polarized growth (Fig. 1D). The absence of the disrupted pinA allele in the fungus isolated following growth in the presence of uracil and uridine was confirmed by Southern analysis (data not shown). Thus, similar to its homologues in both S. cerevisiae and C. albicans, pinA is an essential gene in A. nidulans.
Reduction of PINA Levels Slows A. nidulans Growth-To generate a system in which we could evaluate the physiological roles of PINA in A. nidulans, we created a strain in which the endogenous pinA gene was placed under the control of the regulable alcA promoter. The homologous integration of the alcA promoter has proven useful for regulating the expression of numerous A. nidulans genes in a carbon source-dependent manner (26,36,37). When grown in dextrose as the sole carbon source transcription from the alcA promoter is actively repressed. This repression is relieved in the presence of glycerol, allowing low levels of transcriptional activity, and transcription from the alcA promoter is induced in the presence of glycerol plus threonine. A schematic diagram representing the homologous integration of pAlcPinA into the endogenous pinA locus is depicted in Fig. 2A. Following transformation we identified several strains by Southern analysis in which pAlcPinA had integrated into the pinA locus (data not shown). The multiple strains all displayed similar carbon source-dependent PINA expression, and a single strain was randomly chosen for analysis. The carbon source-dependent regulation of the AlcPinA strain in comparison to the nutritionally complemented control strain, generated by transformation of the parental GR5 strain with pAL5 vector containing no cDNA insert (Pal5) is demonstrated in Fig. 2B. When AlcPinA is grown in minimal medium plus dextrose PINA expression is reduced ϳ50-fold in comparison to controls. In glycerol expression is reduced ϳ3-fold. In glycerol plus threonine PINA expression is induced ϳ3-fold. The Pal5 strain displayed no variation of PINA expression in the three different carbon sources.
When germinated in the presence of dextrose the AlcPinA strain grows at a slower rate than that of the Pal5 control strain as determined by the increase in average number of nuclei per germling as a function of time (Fig. 2C). In contrast, in the presence of glycerol or glycerol plus threonine both strains grew at the same rate ( Fig. 2C and data not shown). Using the data presented in Fig. 2C, the exponential rate of nuclear division of the AlcPinA strain grown in dextrose is calculated to be ϳ95 min versus 75 min for the control. In glycerol, both strains have a doubling time of 125 min. Importantly, Fig. 2C clearly demonstrates that the growth delay due to PINA repression is not due to a lag in germination as the slope of the growth curve is decreased when plotted in a log format. Thus, although carbon source-mediated repression of PINA expression is not sufficient to mimic the effects of pinA disruption, the reduction of endogenous PINA levels to ϳ2% of wild-type significantly decreases the rate of nuclear division.
PINA Is Required for Normal G 1 Progression-Because Pin1 has been suggested to play an important role in multiple phases of the cell cycle, we evaluated whether reduced PINA expression caused a specific delay in cell cycle progression in A. nidulans. Because Pin1 was initially identified as a negative regulator of NIMA function, we first analyzed whether reduced PINA levels altered S-phase and G 2 progression. First we generated new strains of A. nidulans by crossing either the Al-cPinA or the control strain with a strain containing a temperature-sensitive mutation in the G 1 cyclin, nimG (31). When these new strains were arrested in G 1 at the restrictive temperature and released into a permissive temperature, the timing of mitotic entry as indicated by chromosome mitotic index (Fig. 3A) or histone H3 Ser-10 phosphorylation (Fig. 3B) were unaffected by reduced PINA levels, suggesting that the reduced rate of nuclear division observed in the AlcPinA strain is not due to a lengthening of S or G 2 phase.
Because a reduction in PINA levels has been reported to induce a mitotic arrest in both mammalian and fungal systems, we next asked whether the pinA-induced reduction in prolifer-ation was due to a lengthening of mitosis. To address this question we crossed the AlcPINA strain with a strain harboring a temperature-sensitive mutation in cyclinB (nimE) and followed mitotic progression when released from the G 2 arrest.
As seen in Fig. 3C, when AlcPINA and nutritionally complemented controls are released from the nimE arrest point under conditions that repress PINA expression, the cells progress through mitosis at the same rate as determined by following FIG. 1. pinA is essential for A. nidulans proliferation. A, shown is a schematic representation of the homologous integration of the pPINAdis disruption plasmid into the endogenous pinA locus. The approximate location of the HindIII restriction sites used for identification of the homologously integrated transformants are noted. B, diagram of the heterokaryon transformation strategy. Homologous integration of pPINAdis into the pinA locus generates a heterokaryon containing nuclei with the genotypes pyr4 ϩ pinA Ϫ and pyr4 Ϫ pinA ϩ . Following asexual reproduction, the uninucleate spores maintain the genotypes pyr4 ϩ pinA Ϫ and pyr4 Ϫ pinA ϩ . If PINA is essential, neither genotype will germinate in rich medium (YG) and only pyr4 Ϫ pinA ϩ will germinate in rich medium containing uracil and uridine. C, Southern analysis of ⌬pinA/pinA heterokaryons. Southern analysis of HindIII digested total genomic DNA derived from the parental (GR5), and heterokaryon demonstrates the presence of the predicted 0.9-kbp fragment representing the endogenous pinA locus, and the additional 3.0-kbp bands in the homologously integrated transformants. D, growth of spores derived from the ⌬pinA/pinA heterokaryon. Spores derived from a single heterokaryon strain were inoculated into either YG or YG plus uridine and uracil. The data shown are representative of results obtained from six heterokaryon strains. the chromosome mitotic index or the appearance of multinucleate germlings. Thus, because pinA repression does not appear to cause a reduction in the rate of progression through S, G 2 , or mitosis, it is likely that the observed increase in the nuclear division cycle due to pinA repression is due to a lengthening of G 1 .
To determine whether the reduced PINA levels did alter G 1 progression, we examined the changes in NIMX cdc2 kinase activity as a function of time after germination. AlcPinA and Pal5 were germinated in medium containing dextrose, and germinating conidia were collected for analysis every half hour from 2 through 4.5 h post-germination. As demonstrated in Fig.  3D the activation of NIMX cdc2 histone H1 kinase activity is delayed by ϳ30 min under conditions that repress PINA expression. The approximate 30-min delay in NIMX cdc2 activation correlates well with the approximate 20-min lengthening of the cell cycle under conditions that repress PINA expression. Thus, reduced PINA levels appear to cause a lengthening of the cell cycle by increasing the duration of G 1 prior to NIMX cdc2 activation.
PINA Functions in the Recovery from S-phase Arrest-Because PINA was isolated as a negative regulator of NIMA activity, we surmised it would likely play some role in G 2 /M progression in A. nidulans. To test this idea we arrested the AlcPinA and control strains in S-phase with hydroxyurea and then examined entry into mitosis following release from the drug by using the chromosome mitotic index or the temporal accumulation of histone H3 Ser-10 phosphorylation. In contrast to the results observed upon release from the nimG arrest presented earlier, the AlcPinA strain accumulated condensed chromatin and Ser-10 phosphorylated histone H3 ϳ10 -20 min earlier than the Pal5 strain following release from the hydroxyurea arrest (Fig. 4, A and B).
Because reduction of PINA levels accelerate mitotic entry following a hydroxyurea release, we evaluated if this apparent alteration in the replication checkpoint also resulted in an increased sensitivity to hydroxyurea. If PINA performs a critical role in establishing or maintaining the replication checkpoint in A. nidulans, it would likely be reflected by an increased sensitivity to hydroxyurea. As seen in Fig. 4C, AlcPinA and Pal5 appear to have the same sensitivity to increasing concentrations of hydroxyurea, indicating that PINA does not cause a defect in initiating the replication arrest induced by hydroxyurea. Reduced PINA levels also had no effect on sensitivity to other agents that result in various checkpoint arrests such as UV (Fig. 4D) or benomyl (data not shown). Thus, it appears that reduced PINA levels do not inhibit initiation of the replication checkpoint arrest but perhaps are important for recovery from the arrest.
PINA Physically Interacts with NIMA-Because PINA was originally identified as an NIMA interacting protein via the two-hybrid assay (17), we asked whether the two proteins interact in vivo. To address this question we performed coimmunoprecipitation assays utilizing a strain of A. nidulans expressing one extra copy of HA-tagged NIMA and harboring a temperature-sensitive mutation in nimT cdc25 (CDS46) (30). As demonstrated in Fig. 5A, when HA-NIMA was immunoprecipitated from extracts derived from asynchronous, benomyl-ar-rested, or nimT-arrested extracts, PINA can be coimmunoprecipitated. Interestingly, the amount of PINA coprecipitated is greatest in the benomyl-arrested extract indicating that, similar to Pin1, PINA associates most avidly with mitotically phosphorylated NIMA. Likewise, when PINA is immunoprecipitated from benomyl-arrested fungus, HA-NIMA is also present in the precipitate (Fig. 5A).
Because PINA appears to associate most avidly with mitotically phosphorylated NIMA we next examined what domains of both PINA and NIMA are required for this interaction. To address this question we utilized a traditional yeast two-hybrid assay and scored positive interactions by promotion of GAL4dependent transcription of HIS3, which compliments the his3-200 deletion and allows yeast proliferation on medium in the AlcPinA plus nimG10 and Pal5 plus nimG10 were inoculated into minimal medium plus dextrose. Exponentially growing cells were then arrested at the G 1 /S boundary by incubation at 42°C for 2 h. Following release, samples were collected, and the chromosome mitotic index was determined. B, PINA repression does not alter the time course of histone H3 Ser-10 phosphorylation following release from nimG arrest. As in A, germlings were arrested at the restrictive temperature and released at 30°C. Samples were collected and analyzed for histone H3 Ser-10 phosphorylation. C, PINA repression does not alter mitotic progression following nimE arrest. AlcPinA plus nimE6 (squares) and Pal5 plus nimE6 (circles) were inoculated into minimal medium plus dextrose and arrested in late G 2 by incubation at 42°C. Following release CMI (open symbols) and % germlings with Ͼ1 nuclei (closed symbols) were determined. D, PINA repression delays NIMX cdc2 activation upon germination. AlcPinA (squares) and Pal5 (circles) were germinated in minimal medium dextrose to repress PINA protein expression. Following 2 h of growth samples were taken every 30 min until 4.5 h post-germination. Extracts from the samples were assayed for NIMX cdc2 histone H1 kinase activity following p13-agarose precipitation as described under "Materials and Methods." The data are representative of three independent experiments. Between experiments the absolute timing of mitotic entry varied by as much as 20 min, but the differences between the AlcPINA and Pal5 strains were consistent. absence of histidine. As observed in the original interaction screen, full-length PINA fused to the GAL4 transcriptional activation domain interacts with full-length NIMA fused to the GAL4 DNA binding domain, because yeast expressing these two proteins grows equally well in the presence or absence of histidine (Fig. 5B). Additionally, mutation of the PINA WW FIG. 4. The absence of PINA accelerates recovery from a hydroxyurea-induced arrest without altering drug sensitivity. A, PINA repression accelerates entry into mitosis following a hydroxyurea-induced arrest. Exponentially growing AlcPinA and Pal5 germlings were arrested in hydroxyurea for 2 h and released into fresh medium. Following release samples were taken every 10 min for analysis of CMI as described under "Materials and Methods." B, PINA repression accelerates accumulation of histone H3 Ser-10 phosphorylation following release from hydroxyurea-induced arrest. Exponentially growing AlcPinA and Pal5 germlings were arrested in hydroxyurea for 2 h and released into fresh medium. Following release samples were taken every 10 min for analysis of histone H3 phosphorylation. C, reduced PINA expression does not alter the sensitivity of A. nidulans to hydroxyurea. AlcPinA and Pal5 were germinated on minimal medium plus dextrose or glycerol in the presence of hydroxyurea and allowed to germinate for 2 days at 37°C. D, PINA repression does not alter the sensitivity of A. nidulans to UV irradiation. Pal5 (circles) and AlcPinA (squares) germlings were irradiated 4 h post-germination, followed by plating on minimal medium plus glycerol or dextrose and the % survival after 3 days at 37°C was determined. All results are representative of those obtained in at least three independent experiments. domain (W33A) or expression of the prolyl-isomerase domain (PI domain) alone does not alter the ability of the yeast to grow on His medium. However, either mutation of the prolyl-isomerase active site (C126A) or removal of the prolyl-isomerase domain completely (WW domain) abolishes the ability of the yeast to grow in the absence of histidine without altering its proliferative properties in medium supplemented with histidine. All the proteins were expressed to similar levels within the yeast strains (data not shown). Thus, in contrast to the characterized interactions of Pin1 and its target proteins, the interaction of PINA with NIMA is dependent upon a functional prolylisomerase domain, and the WW domain appears to be dispensable for the interaction.
To determine the domain(s) of NIMA required for its interaction with PINA we performed two-hybrid analysis of wildtype PINA with various NIMA fragments. Consistent with results using Pin1, the catalytic domain of NIMA (NIMA-(1-292)) fused to the Gal4 DNA binding domain does not support yeast growth in the absence of histidine, whereas the C terminus (NIMA-(293-699)) does interact with PINA in the twohybrid assay. Because the Pin1 interaction domain of NIMA was originally identified to lie within amino acids 280 -396 (14), we tested whether PINA interacts with a similar fragment, NIMA-(293-396). Fig. 5C demonstrates that NIMA-(293-396) does not interact with PINA in the two-hybrid system, whereas the C-terminal fragment containing amino acids 397-699 demonstrates a positive interaction in this system. All NIMA fragments were expressed to similar levels in the yeast (data not shown). Thus, the PINA binding domain of NIMA appears to be confined to the C-terminal 302 amino acids of the protein.
PINA Positively Regulates NIMA Function in Vivo-Because Pin1 was originally identified as an NIMA interacting protein and demonstrated to negatively regulate NIMA function in cultured cells (14), we examined whether PINA regulated NIMA function in A. nidulans. Because PINA repression or overexpression does not appear to share any phenotypic consequences with NIMA repression or overexpression in this organism, we tested for a genetic interaction between PINA and NIMA. Such an interaction was examined by crossing the AlcPinA strain with a strain containing the nimA5 temperature-sensitive mutation and testing whether the repression or overexpression of PINA altered the temperature-sensitive growth of this new strain. As shown in Fig. 6, in the absence of the nimA5 mutation (Wt), the AlcPinA and Pal5 strains grow on the various carbon sources as described earlier and display little to no temperature-dependent growth variation at 31, 37, or 41°C. However, in the nimA5 background there are two significant growth differences. First, when germinated in the presence of dextrose to repress PINA expression, the nimA5/ AlcPinA strain did not grow even at the nimA5-permissive temperature of 31°C. In contrast, at 31°C in dextrose the nimA5/Pal5 strain grows similar to the wild-type/Pal5 control strain. Second, when the strains are germinated on glycerol plus threonine to induce PINA overexpression, the AlcPinA/ nimA5 strain shows enhanced growth compared with the Pal5/ nimA5 control strain when germinated at the partially permissive temperature of 37°C. Thus, PINA positively regulates NIMA function genetically. Specifically, reduced PINA expression exacerbated the nimA5 temperature-sensitive mutation while PINA overexpression partially relieves the temperature sensitivity of growth. Moreover, this genetic interaction is selective for NIMA, because no genetic interactions were seen with temperature-sensitive mutations of other genes involved in growth control, including NIMT cdc25 , NIMQ mcm2 , or NIME cyclinB (data not shown).
PINA Does Not Regulate NIMA Kinase Activity-Because PINA not only interacts with NIMA but also positively influences its function, we evaluated whether PINA altered the kinase activity of NIMA in vitro or in vivo. To assess the effect of PINA on NIMA kinase activity in vitro we tested whether the addition of purified PINA altered the kinase activity of bacterially expressed and purified full-length NIMA. As seen in Fig.  7A, in comparison to BSA, the addition of PINA has no effect on NIMA kinase activity, using a peptide substrate, either before or after an activating preincubation in the presence of ATP. Additionally, the presence of PINA did not alter the K m for either peptide or ATP substrates (data not shown).
Given that Pin1 has been demonstrated to promote PP2A function (38), and PP2A has also been shown to dephosphorylate and inactivate NIMA in vitro (39), we next examined whether the addition of PINA would facilitate the ability of PP2A to dephosphorylate and inactivate recombinant NIMA. To test whether PINA promotes PP2A dephosphorylation of NIMA we followed the dephosphorylation of active bacterially expressed NIMA in the presence of added PINA or BSA through the use of the MPM2 monoclonal antibody, which recognizes antigens similar to those that bind Pin1. As demonstrated in Fig. 7B, the addition of PINA or BSA alone had no effect on the phosphorylation state of recombinant NIMA over a 60-min incubation. However, the addition of purified PP2A resulted in a significant time-dependent reduction in MPM2 immunoreactivity independent of the presence of PINA. Similarly, in comparison to BSA, the presence of PINA did not alter the ability of PP2A to promote inactivation of NIMA kinase activity in vitro (Fig. 7C). Thus, PINA does not appear to alter either NIMA activity or PP2A-mediated dephosphorylation of NIMA in vitro.
Finally, although we could detect no deficits in the G 2 /M transition under conditions that repress PINA expression, we questioned whether PINA repression altered NIMA protein levels or kinase activity. To more easily detect and assay NIMA from A. nidulans extracts, we crossed the AlcPinA strain with the CDS46 strain, which possesses one extra copy of HA-tagged NIMA and used this new strain in addition to the parental CDS46 strain for the subsequent experiments. First, we assayed NIMA protein levels and kinase activity in extracts derived from mitotically arrested AlcPINA/CDS46 and CDS46 under conditions that repress, de-repress, or induce PINA expression. Fig. 7D demonstrates that in mitotically arrested FIG. 6. PINA displays a positive genetic interaction with NIMA. Pal5, AlcPinA, Pal5 plus nimA5, and AlcPinA plus nimA5 were inoculated onto minimal medium containing dextrose (repressing), glycerol (non-repressing), or glycerol plus threonine (inducing) and incubated at the indicated temperatures for 2 or 3 days. These data are representative of three experiments with three strains of each genotype.
germlings, HA-NIMA levels are constant regardless of carbon source or PINA levels. Additionally, there is very little carbon source or PINA-dependent variation in HA-NIMA kinase activity when HA immunoprecipitates are assayed in a peptide kinase assay (Fig. 7E). Therefore, under the conditions of the assays, PINA repression or overexpression has little or no effect on either NIMA protein level or kinase activity in vivo. DISCUSSION Although Pin1 has been implicated in numerous signaling pathways critical for cell proliferation in multiple eukaryotic systems, the phenotypic responses to Pin1 deletion or inhibi-tion vary greatly among these systems. Pin1 is essential for growth in S. cerevisiae and C. albicans, but its disruption in S. pombe, mouse, and Drosophila has less severe phenotypic consequences (4 -12). Here we report characterization of the in vivo roles of the A. nidulans Pin1 homologue, PINA.
Similar to S. cerevisiae and C. albicans, PINA is essential for A. nidulans cell proliferation (4,5). Although we were unable to characterize the terminal phenotype of the PINA null fungus, a 50-fold reduction of PINA protein increases the doubling time of A. nidulans by ϳ25%. Based on the consistent increase in cell cycle length between the first nuclear division and subsequent divisions and the lack of evidence that other phases of the cell cycle are altered in length, we believe that the reduced growth rate under conditions that repress pinA expression is due to an increase in the length of G 1 . However, we cannot exclude the possibility that the observed lengthening of the cell cycle when PINA is repressed could be due to a subtle lengthening of multiple cell cycle phases that are unable to be detected by the assays used. Nevertheless, our findings contrast with the observations that temperature-sensitive mutations in the S. cerevisiae and C. albicans Pin1 homologue, ess1, lead to a mitotic arrest (5,13). However, data suggest that the essential mitotic role of ess1p in S. cerevisiae is not direct (i.e. acting directly upon a regulator of mitosis) but instead is due to alterations in transcriptional regulation (13). Regardless, although we interpret our data to indicate that PINA function is critical for the proper timing of G 1 , they do not preclude the possibility that a rate-limiting amount of PINA (less than 2% of normal cells) is essential for progression through other phases of the cell cycle such as mitosis.
A G 1 role for PINA in A. nidulans is consistent with reported G 1 deficits observed in mammalian cell systems. Although expression of antisense Pin1 RNA in HeLa cells has been reported to induce a cell cycle arrest with a high percentage of cells containing condensed chromatin (14), the characterization of Pin1 null mice and fibroblasts has produced evidence for a critical role of Pin1 for G 1 progression. Pin1 null primordial germ cells and MEFs grow at a slower rate compared with controls, and MEFs are impaired in cell cycle re-entry following serum deprivation (9,11,12,18). At the biochemical level, Pin1 has been suggested to positively regulate numerous proteins important for or implicated in G 1 progression, including cy-clinD1, c-Jun, and ␤-catenin (10,19,20). Although the mechanism by which Pin1 modulates the function of these proteins varies from changes in activity to stability to localization, a consistent theme in the G 1 roles proposed for Pin1 is its involvement in transcriptional regulation (reviewed in Ref. 40). Thus, an obvious question is whether the G 1 phenotype we have observed in Aspergillus is reflective of a role Pin1 in the transcriptional regulation of genes important for G 1 control. If this is the case, the phenotype observed in A. nidulans could be more similar to the role ess1 performs in S. cerevisiae than might have initially appeared to be so. In both systems Pin1 may be altering the transcriptional regulation of genes critical for cell cycle progression, but in budding yeast this is manifest by a mitotic arrest while A. nidulans displays a G 1 deficit. Unfortunately, little is known about either transcriptional regulation or the genes involved in G 1 progression in A. nidulans, thus elucidation of the molecular target(s) of Pin1 responsible for or contributing to this phenotype remains to be addressed by future work.
Similar to results obtained in S. pombe but in contrast to Pin1 null fibroblasts, reduced PINA protein levels do not appear to markedly alter the response of A. nidulans to DNAdamaging agents (UV-irradiation or methyl methanesulfonate) or the replication inhibitor hydroxyurea (6,(41)(42)(43). However, following release from a hydroxyurea-induced replication arrest, the fungal cells enter mitosis more rapidly under conditions that repress PINA expression, a result reminiscent of our previous data that depletion of Pin1 from interphase Xenopus egg extracts accelerates the entry into mitosis upon addition of cyclinB (44). If, as observed in Xenopus egg extracts, PINA performs a critical role in establishing or maintaining the replication checkpoint in A. nidulans, it would likely be reflected by an increased sensitivity to hydroxyurea. However, because PINA does not alter the sensitivity to hydroxyurea, we interpret these results to suggest that PINA plays a role in the recovery from a checkpoint-induced cell cycle arrest rather than in the establishment or maintenance of the replication checkpoint. In A. nidulans the replication checkpoint has two described components (45)(46)(47). The first, initiated when DNA replication is slowed, results in the inhibitory tyrosine phosphorylation of NIMX cdc2 via ANKA wee1 , which is relieved by NIMT cdc25 (46,47). Although both wee1 and cdc25 have been demonstrated to interact with Pin1 (16,17), we believe this component of the checkpoint pathway to be intact, because the fungus shows no increased sensitivity to hydroxyurea. The second component of the checkpoint is enabled when replication is completely inhibited and is dependent upon the activity of BIME APC1 (45,46). The current hypothesis is that BIME APC1 prevents the lethal accumulation and activation of NIMA during the replication arrest. Thus, even though both components of the checkpoint pathway are functional, either arm of the pathway (or both arms) could be subject to PINA regulation yielding the observed acceleration into mitosis following release from a hydroxyurea arrest.
Consistent with previous reports of the interaction of Pin1 with NIMA, by coimmunoprecipitation PINA appears to associate most avidly with the mitotic form of NIMA. Upon mitotic entry NIMA not only becomes hyperphosphorylated on numerous S/TP motifs presumably via NIMX cdc2 , but it also changes subcellular localization from perinuclear to nuclear (30,48). In contrast to the defined interaction domain with Pin1 (14), we have mapped the PINA interaction domain of NIMA to the C-terminal 302 amino acids that contain the majority of the potential mitotically phosphorylated S/TP sites. Given that Pin1 has been demonstrated to interact with mitotic phosphoproteins on sites that can be generated by cdc2 kinase activity and that the primary cellular localization of PINA is nuclear, the increased association of PINA and NIMA in benomyl-arrested extracts is likely due to both an increased phosphorylation state of the protein as well as its transport into the nucleus where PINA is concentrated. However, although our data suggest that the NIMA-PINA interaction is phosphorylation-dependent, this question has yet to be addressed specifically.
Although the association of PINA with NIMA via coimmunoprecipitation came as no surprise to us and confirmed the original identification of the interaction via the two-hybrid system (17), the results from mapping of the domains responsible for the interaction were quite unexpected. Not only did the PINA binding site map to the C-terminal 302 amino acids of NIMA, but the primary domain in PINA responsible for its interaction with NIMA is the prolyl-isomerase domain. Based on the precedent that the domain responsible for the interaction of Pin1 with its target proteins is the WW domain, and the findings that the WW domain of PINA binds many phosphorylated peptides known to bind Pin1 and that PINA interacts primarily with the mitotic form of NIMA, we would have predicted that the WW domain of PINA would also be responsible for its interaction with NIMA. These results raise the question as to whether this type of interaction is specific to PINA or could also reflect the presence of a different class of Pin1interacting proteins. The isomerase domain-dependent interaction of PINA with NIMA is reminiscent of numerous characterized immunophilin complexes, including (but not limited to) interactions of FKBP12 with ryanodine, inositol 1,4,5-trisphosphate, and transforming growth factor ␤ receptors; the cyclo-philinA-human immunodeficiency virus capsid p24 and Gag protein complexes; and the cyclophilinB-CAML protein complex (49 -53). In instances such as these it has been proposed that the rotomase recognizes and binds with high affinity to a domain of the target protein that structurally resembles conformation of a suitable isomerization substrate with some subtle differences (reviewed in Ref. 54).
That PINA interacts with NIMA both physically and genetically is clear from our study, although the molecular consequences of this interaction remain enigmatic. Pin1 has been demonstrated to affect protein function by a number of mechanisms, including alteration of protein stability, localization, transcription, and enzymatic activity (reviewed in Refs. 1 and 2), all of which are known modes of NIMA regulation in vivo. Although measured statically and under conditions designed to produce maximal activity using a non-physiological peptide substrate, our data suggest that PINA does not significantly alter NIMA protein levels or kinase activity. Thus, it seems possible that PINA plays a non-essential role in targeting the subcellular localization of NIMA. For example, PINA could act as an adaptor molecule by targeting NIMA to mitotic phosphoproteins. In this case, the prolyl-isomerase domain could play a non-catalytic role acting as a phosphorylation state-specific binding domain, a role normally attributed to the WW domain, and the WW domain could act in trans to effect formation of a ternary, phosphorylation-dependent, protein complex. Although multidomain adaptor molecules targeting phosphoepitopes are quite common, this function has yet to be proposed for Pin1. Structurally, Pin1 has been proposed to be suitable for such a signal-dependent adaptor molecule, because the WW domain and prolyl-isomerase domain are connected by a highly flexible linker that can adapt different conformations depending upon the peptide bound (55). Thus, by this model PINA could influence NIMA function by facilitating the proper subcellular localization of the active kinase to sites that require phosphorylation by NIMA. Clearly, this mechanism has yet to be tested, and there are other potential mechanisms by which PINA could positively regulate NIMA function in vivo.
Pin1-mediated, phosphorylation-directed proline isomerization is now appreciated to be a conserved mechanism for modulation of cellular signaling events in organisms from S. cerevisiae to mammals. The demonstration that Pin1/Ess1p/PINA is essential for proliferation in S. cerevisiae, C. albicans, and A. nidulans emphasizes the emerging importance of phosphorylation-dependent conformational changes mediated via proline isomerization. Although the deletion of Pin1 is not universally lethal (mice, D. melanogaster, and S. pombe), there is evidence suggesting that other isomerases can substitute for Pin1 in its absence. Drugs targeting the catalytic activity of both Pin1 and a parvulin family member, par14, have been demonstrated to block proliferation of numerous mammalian cell lines (56). Also, in S. pombe the disruption of Pin1 sensitizes the yeast to the cyclophilin inhibitor, cyclosporin A (6,57). Given the difficulty presented by the apparent genetic redundancy in multicellular organisms and the multitude of potential Pin1 targets, the characterization of its essential roles in a genetically tractable system such as A. nidulans should yield mechanistic insight into the critical conserved cellular roles of Pin1-mediated proline isomerization in more complex metazoans.