Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Pneumocystis carinii represents an opportunistic fungus that continues to cause severe pneumonia in immunocompromised patients with acquired immunodeficiency syndrome (AIDS), organ transplantation, or hematologic or solid malignancies (1). Despite currently available antimicrobial agents, the mortality of an acute episode of P. carinii pneumonia persistently ranges between 5 and 40% (2- 5). Accordingly, patients who are chronically immunosuppressed require ongoing prophylaxis to prevent primary or recurrent P. carinii pneumonia (6). Significant side effects continue to limit the use of many anti-Pneumocystis agents in a substantial number of these individuals (7). Furthermore, the discovery of new therapeutic agents has been substantially hampered by a limited understanding of life-cycle control in this species (8).

Recent investigations have begun to define the regulation of the life cycle of P. carinii. P. carinii is of fungal origin on the basis of ribosomal RNA gene sequencing and other molecular homology studies (9). Phylogenetic investigations reveal that P. carinii is a member of the Ascomycota, located on a branch separating it from other members of this phylum (12). As such, P. carinii is phylogenetically related to the fission yeast Schizosaccharomyces pombe, a useful model organism for studies of the molecular basis of cell-cycle control (9, 13).

Ultrastructural studies of infected lung tissue and materials obtained from bronchoalveolar lavage reveal that P. carinii consists predominantly of diminutive trophic forms (formerly termed trophozoites; 1 to 2 µ), with varying numbers of the larger thick-walled cysts (8 µ) (14, 15). Although P. carinii can occasionally be found at other sites of the body in advanced disease, it preferentially proliferates within the alveolar spaces (16). The interactions of trophic forms with alveolar epithelial cells appear to be a central component of the organism's life cycle that modulates cellular proliferation (17, 18). Attachment of P. carinii to alveolar epithelial cells impairs cell-cycle progression of host cells while simultaneously enhancing proliferation of the organism (17, 19). Despite substantial efforts, consistent in vitro cultivation of P. carinii has been difficult to achieve (8, 20).

The expression and activation of a number of cell division cycle molecules precisely control the cell cycles of fungi related to P. carinii (21). Of particular importance are the cyclins and their associated cyclin-dependent kinases. The B-type cyclins are conserved molecules encoded by cdc13 genes essential for eukaryotic cell mitosis (22, 23). Cdc13 proteins rise and fall over the cell cycle, peaking in late G2 and early M phase, and falling to baseline by the beginning of G1. Cdc13 proteins bind to the catalytic Cdc2 kinase, conferring conformational changes essential to the enzymatic activity of Cdc2 (21). Originally termed mitosis promoting factor, the Cdc2/Cdc13 complex provides essential activity for the completion of eukaryotic mitosis. We have recently shown that P. carinii possesses a catalytically active Cdc2 molecule whose activity is regulated over the life cycle of the organism. We have further shown that P. carinii Cdc2 mediates fungal proliferation (24). To date, no information has been yet uncovered with respect to the Cdc13 partner necessary for the Cdc2 activity of P. carinii. Herein, we report the identification, molecular cloning, and functional characterization of the P. carinii Cdc13 B-type cyclin molecule. These investigations reveal that the Cdc2/Cdc13 cyclin-dependent kinase complex has variable activity over the life cycle of P. carinii. Furthermore, the P. carinii Cdc13 cyclin is capable of restoring fungal proliferation when transformed into mutant strains of the related fungus S. pombe, which are otherwise deficient in endogenous Cdc13 cyclin protein. Better understanding of Cdc13 activity should provide essential new insight into proliferative control of this important pathogen.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Materials

Reagents were from Sigma Chemical Co. (St. Louis, MO), unless otherwise noted. [-³²P]adenosine triphosphate (ATP) was obtained through New England Nuclear (Boston, MA). A mouse antibody that interacts with Cdc13 (cyclin B₁) proteins from a variety of species (clone CB169) was purchased from Upstate Biotechnology Inc. (Lake Placid, NY) (25). Nonimmune mouse immunoglobulin (Ig) G was from Sigma Chemical Co. The rat-derived P. carinii genomic DNA library in gt11 bacteriophage was kindly provided by Dr. James R. Stringer, University of Cincinnati College of Medicine (Cincinnati, OH) (26). This rat P. carinii genomic library was derived from P. carinii f. sp. (formae specialis) carinii. A rat P. carinii complementary DNA (cDNA) library cloned into Uni-ZAP XR  phage as EcoRI-XhoI fragments was obtained from the AIDS Research and Reference Reagent Program (McKesson BioServices, Rockville, MD). Nitrocellulose membranes containing P. carinii chromosomes separated by contour-clamped homogenous field electrophoresis were generously donated by Dr. Melanie T. Cushion, University of Cincinnati College of Medicine (27). S. pombe cdc13 mutants containing temperature-sensitive deficiencies of S. pombe Cdc13 were obtained from the American Type Culture Collection (ATCC strain 96086; Rockville, MD) (28). The pREP41 vector was the kind gift of Dr. Kathy Gould, Vanderbilt University (Nashville, TN) (29, 30). Ciprofloxacin was the generous gift of Dr. Barbara Painter of Miles Pharmaceuticals, Inc. (West Haven, CT).

Preparation of P. carinii

All animal studies were approved by the Mayo Clinic Institutional Animal Care and Utilization Committee. P. carinii pneumonia was induced in Harlan Sprague-Dawley rats by immunosuppression with dexamethasone and transtracheal injection with P. carinii, as we have reported (31, 32). Pathogen-free rats were provided freely with drinking water containing dexamethasone (2 mg/liter), tetracycline (500 mg/liter), and nystatin (200,000 U/liter). Each week, the animals additionally received oral ciprofloxacin (0.45 g/liter) for two consecutive days to further reduce bacterial infections. After 5 d of immunosuppression, rats were transtracheally inoculated with P. carinii (~ 500,000 cysts) prepared by homogenizing infected rat lung. The animals were killed after an additional 6 to 8 wk of immunosuppression. P. carinii were purified by homogenization and exhaustive filtration through 10-µM filters that retain lung cells but allow passage of P. carinii (33). P. carinii filtrates were centrifuged (1,500 × g for 30 min), and the pellets were resuspended in Hanks' balanced salt solution. Duplicate 10-µl aliquots of suspension were spotted onto slides, stained with modified Wright-Giemsa (Diff Quik), and P. carinii quantified as described (34). If other microorganisms were noted on microbiologic examination, the material was discarded.

As a regulatory protein, the activity of Cdc13 should be controlled over the course of the organism's life cycle. To address this, differential filtration was used to separate P. carinii cysts and trophic forms (24, 33). In brief, P. carinii cysts are retained by a 3-µm Nucleopore filter and resuspended after exhaustive washing. P. carinii trophic forms pass through the device and are recovered by centrifugation. This separation procedure yielded trophic form populations containing 99.5% trophic forms and cyst preparations, which were 40-fold enriched (24).

Immunoprecipitation of Cdc2/Cdc13 Complex and Its Associated Kinase Activity from P. carinii

To initially determine whether P. carinii contains a Cdc13-like molecule, P. carinii extracts were incubated with a mouse antibody known to interact with Cdc13 proteins from various host species (25). Cdc2/Cdc13 complexes were immunoprecipitated from these P. carinii extracts and assayed for their ability to phosphorylate the target protein histone H1, an activity related to the active Cdc2 serine-threonine protein kinase being bound to its essential Cdc13 cyclin (24). To accomplish this, P. carinii organisms (~ 5 × 10⁸) were suspended in kinase lysis buffer (250 mM NaCl, 50 mM Tris-HCl, pH 7.4, 0.1% Triton X-100, 5 mM ethylenediaminetetraacetic acid, 5 mM NaVanadate, 5 mM NaF, 50 µg/ml phenylmethylsulfonyl fluoride, 1 µg/ml leupeptin, 0.1 TIU aprotinin) and sonicated three times at 150 W for 10 s on ice to disrupt the cell walls. Soluble proteins were clarified by centrifugation at 12,000 × g for 10 min, and the protein concentrations in the supernatants were determined using the bichinchoninic acid method (Pierce Chemical Company, Rockford, IL). Identical quantities of soluble proteins were preabsorbed with 50% protein A-sepharose, centrifuged at 12,000 × g for 10 min, and the supernatants were incubated with either anti-Cdc13 antibody or nonimmune IgG (50 µg/ml each; Upstate Biotechnologies) overnight at 4°C. Cdc2/ Cdc13 antibody complexes were collected using 50% protein A-sepharose and centrifugation at 12,000 × g for 10 min. The precipitated products were washed with lysis buffer and subsequently with kinase buffer (50 mM Tris-HCl, pH 7.4, 10 mM MgCl₂, 1 mM dithiothreitol). The pellets were resuspended in kinase buffer containing 5 µM ATP, 100 µg/ml histone H1, and 0.1 µCi/µl [-³²P]ATP, and incubated for 10 min at 30°C. The reaction was stopped using Laemmli buffer containing 5% 2-mercaptoethanol and resolved by 12% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and autoradiography. To further confirm that the protein kinase activity was derived from the purified P. carinii organisms and was not the result of host cell contamination, an equal volume of lung from uninfected rat was processed in an identical fashion. The resulting filtrates were lysed, immunoprecipitated with anti-Cdc13 antibody, and evaluated in the histone H1 kinase assay.

Molecular Cloning of the P. carinii cdc13 Gene and cDNA Sequences

P. carinii were isolated by homogenization and differential filtration from immunosuppressed rats, and the genomic DNA was extracted as reported (24, 35, 36). Polymerase chain reaction (PCR) amplification of P. carinii genomic DNA was performed using 1 µM of paired degenerate primers derived from conserved amino-acid sequences contained in other fungal Cdc13 proteins (37, 38). The coding bias was used to limit the degeneracy of the third position of each codon as previously reported (39, 40). The following primers were employed 5'-GCG(A/T)GG(A/T)AT(A/ T)(C/T)T(A/T/C/G)GT(A/T) GA(C/T)TGG-3' and 5'-GGACACAT(A/T)AC(C/T)TC(C/T)TC(A/G)TA(C/T)TT-3'. P. carinii genomic DNA was used as a template. An initial 5-min hot start at 94°C was followed by 30 cycles of 94°C for 60 s, 56°C for 60 s, and 72°C for 60 s. After the last cycles, a final extension at 72°C was performed for 15 min. A single 191-bp amplicon was generated, subcloned into the pGEM-T Easy vector (Promega Inc., Madison, WI), and sequenced. Sequence comparisons to GenBank were performed using the BLAST Genetic Analysis Program (National Center for Biotechnology Information [NCBI]) (41). To further verify that this PCR-derived product was derived from P. carinii and not the result of amplification of contaminating host or other DNA, the 191-bp amplicon was hybridized to a nitrocellulose membrane containing P. carinii chromosomes. Separation of P. carinii chromosomes by contour-clamped homogenous field electrophoresis was performed as previously reported (35). The 191-bp P. carinii amplification product was tagged with [-³²P]deoxyadenosine triphosphate by the random primer method (Rediprime System; Amersham). After prehybridization for 30 min (ExpressHyb solution; Clontech Laboratories), the contour-clamped homogeneous electrical field (CHEF) nitrocellulose membrane was incubated with the probe (1.5 × 10⁶ cpm/ml) at 60°C for 1 h, washed four times at 25°C for 40 min in 2× saline sodium citrate (SSC) buffer containing 0.05% SDS, washed twice at 50°C for 40 min in 2× SSC buffer containing 0.1% SDS, and visualized by autoradiography. To obtain the full-length P. carinii cdc13 genomic DNA sequence, a rat-derived P. carinii (P. carinii carinii) genomic DNA library in gt11 was next screened by hybridization to the 191-bp amplicon. Clones were plaque-purified to homogeneity. A 3.1-kb insert was identified, subcloned into pGEM-7Zf() (Promega, Inc.), and both strands sequenced. Subsequently, the P. carinii cdc13 cDNA was cloned from a rat-derived P. carinii cDNA library in the bacteriophage Uni-ZAP XR (Stratagene, Inc.). This library was screened again using the 191-bp probe. Hybridizing clones were identified and excision of the phagemid was accomplished using the M13 helper phage. A 2,072-bp insert was identified by restriction endonuclease digestion with EcoRI and XhoI and completely sequenced. Nucleotide comparisons to the GenBank database were performed using the BLAST protocol (NCBI).

Generation of Specific Antibody and Determination of the Relative Specific Activity of the P. carinii Cdc2/Cdc13 Complex in Cysts and Trophic Forms

The amino-acid sequence NDENKLHGQISRVKQ, representing predicted amino acids 17 through 31, was found to be antigenic and unique to P. carinii by MacVector analysis (IBI-Kodak, Inc.). This peptide was synthesized and coupled to keyhole limpet hemocyanin (KLH) through addition of an amino-terminal cysteine residue (~ 15 mol of peptide per mole of KLH). Rabbits received a primary subcutaneous injection (500 µg) of the antigen suspended in Freund's complete adjuvant. This was followed by subcutaneous boosts of the adjuvant in Freund's incomplete adjuvant at 2, 4, and 8 wk (500 µg each). Reactivity of the immune serum was verified by enzyme-linked immunosorbent assay, using previously reported methods (40). This P. carinii Cdc13 antiserum was used to evaluate the relative specific activity of Cdc2/Cdc13 complexes over the life cycle of P. carinii. To accomplish this, P. carinii populations were separated into trophic forms and enriched cyst preparations, and the organisms were lysed and extracted, as described previously. Parallel aliquots containing identical concentrations of total cellular protein were precleared with 50% protein A-sepharose at 4°C for 45 min, transferred to a clean tube, and immune and nonimmune sera added (1:75 dilution) and mixed end-over-end for 2 h at 4°C. The activity of precipitated Cdc2/Cdc13 complexes was determined using histone H1 substrate in a fashion identical to that described previously.

Immunoblotting of P. carinii Cdc13 Cyclin Protein

To further characterize the relative molecular mass of P. carinii Cdc13, P. carinii organisms (2 × 10⁶/lane) were isolated as described, and lysed in 2% SDS containing 5% -mercaptoethanol. P. carinii components were separated over a 12% polyacrylamide gel, electroblotted onto a nitrocellulose membrane, and cut into strips (42). The membrane was blocked in 3% dry milk and parallel membrane strips were incubated with either anti-P. carinii Cdc13 serum or nonimmune serum (1:250). Bound antibodies were localized using a commercial immunodetection system (Bio-Rad, Hercules, CA). The relative molecular mass was derived by comparison to the migration of rainbow molecular weight standards (Amersham, Inc.).

Northern Hybridization of P. carinii cdc13

Northern hybridizations were performed to additionally determine the extent of P. carinii cdc13 steady-state messenger RNA (mRNA) expression in P. carinii removed from the infected lung. P. carinii were isolated, purified, and identically aliquoted (in tubes of RPMI containing 10% fetal calf serum, 10,000 U/liter penicillin, 1 mg/liter streptomycin, and 25 µg/liter amphotericin. In general, P. carinii organisms were isolated within 30 min after the animals were killed. At selected time points after isolation, total RNA was extracted from the P. carinii aliquots using a monophasic solution of phenol and guanidine isothiocyanate followed by chloroform extraction and isopropyl alcohol precipitation (Trizol Reagent; GIBCO BRL, Grand Island, NY). Equal amounts of total RNA (5.0 µg) were loaded and separated by electrophoresis through a 1.2% agarose gel in the presence of 2.2 M formaldehyde. Equal loading of the RNA was verified by ethidium bromide staining of 18S and 28S ribosomal RNA, and the separated RNA was transferred to nitrocellulose membranes and prehybridized (ExpressHyb; Clontech Laboratories). The full-length cdc13 cDNA was excised from the pGEM-T Easy vector using EcoRI and labeled with -[³²P]deoxycytidine triphosphate (Dupont) by a random primer method (Rediprime; Amersham). The radiolabeled probe was added to hybridization solution (2 × 10⁶ cpm/ml) and incubated with the membranes for 1 h at 68°C. After hydridization, the membranes were washed four times with 2× SSC solution (where 1× solution contained 150 mM NaCl and 15 mM sodium citrate; pH 7.0) with 0.05% SDS at room temperature for 40 min followed by 0.1× SSC with 0.1% SDS solution at 50°C for 40 min. The blots were visualized by autoradiography. For comparison to the P. carinii cdc13 transcript, the membranes were stripped in 1% SDS and rehybridized with a [³²P]-labeled probe complementary to the constituitively expressed actin gene product of P. carinii (43).

Expression of Recombinant P. carinii Cdc13 Cyclin Protein

In vitro transcription and translation of P. carinii Cdc13 cDNA was accomplished as follows. A modified construct of the P. carinii cdc13 cDNA was generated by PCR amplification using a 5' primer containing an NdeI site encoding the initiating methionine of Cdc13 and a 3' primer containing a SalI site just beyond the cDNA coding region. This amplicon was subcloned in frame downstream of the T7 RNA polymerase promoter into the pGEM-T Easy vector (Promega, Inc.). The recombinant plasmid was introduced into Escherichia coli DH5, amplified, and extracted (Plasmid Miniprep Kit; Promega). In vitro transcription and translation reactions were conducted in a 50-µl reaction using 1 µg of the uncut plasmid with the T7 Coupled Reticulocyte Lysate System (Promega, Inc.) in the presence of 40 µCi of [³⁵S]methionine (1,000 Ci/mmol). The reaction mixtures were immunoprecipitated with P. carinii Cdc13 antiserum or nonimmune serum, and visualized by 12% SDS-PAGE and autoradiography.

Determination of P. carinii cdc13 Activity in Fungal Cell-Cycle Progression

As noted, studies of P. carinii have been hampered by our inability to continuously cultivate this fungus. Thus, we analyzed the functional activity of P. carinii Cdc13 by heterologously expressing this cyclin in mutant strains of S. pombe, which conditionally exhibit deficiencies of their endogenous Cdc13 protein. These S. pombe temperature-sensitive cdc13 mutants grow at the permissive temperature of 30°C, but undergo cell-cycle arrest at the nonpermissive temperature of 37°C (28). These Cdc13-deficient S. pombe were transformed with the P. carinii cdc13 cDNA and growth was assessed under conditions in which endogenous S. pombe Cdc13 was nonfunctional. In specific, P. carinii cdc13 cDNA was excised from pGEM-T Easy vector by digestion with NdeI and SalI and directionally subcloned into the yeast expression vector pREP41. The pREP41 plasmid has a leu2 gene, permitting growth of transformants on media lacking leucine, and a S. pombe nmt1 promoter, which is expressed in the absence of thiamine but repressed in the presence of this nutrient (30). S. pombe temperature-sensitive cdc13 mutants were grown to mid-log phase in YES broth at 30°C (optical density 595 = 0.5) and transformed by electroporation (44) using 1 µg of pREP41/ P. carinii cdc13 cDNA or 1 µg pREP41 vector alone without insert. Transformed S. pombe were plated on leucine- and thiamine-deficient plates at 30°C and 37°C, and assessed for growth. Transformed yeast colonies proliferating on leucine- and thiamine-deficient plates at 37°C were cultured to mid-log phase in leucine- and thiamine-deficient broth, and plasmid DNA was extracted and sequenced to confirm the presence of P. carinii cdc13. Additional incubations of the transformants were also performed in the presence of thiamine (10 µM), which represses the pREP41 nmt promoter and hence inhibits expression of the P. carinii cdc13 cDNA.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

P. carinii Contains a Putative Cdc13 B-Type Cyclin That Coprecipitates Cyclin-Dependent Kinase Activity from Extracts of the Organism

As a first means to evaluate whether P. carinii expresses a putative Cdc13 B-type cyclin, P. carinii extracts were immunoprecipitated with an antibody recognizing B-type cyclin molecules from a variety of host species (CB169; Upstate Biotechnologies Inc.) (Figure 1). P. carinii organisms were purified, extracted, and immunoprecipitated with this antibody (Figure 1, lane A). The anti-cyclin B₁ antibody, but not nonimmune control IgG (Figure 1, lane B), precipitated cyclin-dependent kinase complex-like activity, as evidenced by the ability of precipitated complexes to phosphorylate histone H1. At the present time, biochemical analyses of P. carinii are limited to P. carinii derived from infected rats. Accordingly, extensive controls were conducted to verify that the observed cyclin-dependent kinase activity was derived from P. carinii and was not due to rat lung contamination. To verify this, lungs from uninfected rats were homogenized, filtered, and the filtrates lysed in a manner identical to P. carinii-infected rats. These extracts were then analyzed in parallel by immunoprecipitation and protein kinase assay (Figure 1, lanes C and D). An equal volume of uninfected rat lung processed in an identical manner failed to demonstrate any residual cyclin-dependent kinase activity after filtration through the 10-µ filter, extraction, and immunoprecipitation with antibody to cyclin B₁. Because B-type cyclins are intracellular proteins and because the antibody used would have recognized rat cyclins, these controls document that the 10-µ filter purification procedure effectively eliminates intact rat host cells. Therefore, the cyclin-dependent kinase activity observed in the P. carinii extracts (Figure 1, lane A) is related to a P. carinii Cdc13-like activity and is not the result of rat cell contamination. These observations strongly suggest that P. carinii possesses a Cdc13 B-type cyclin that is capable of forming enzymatically active complexes with its Cdc2 kinase partner.

View larger version (33K): [in this window] [in a new window]   Figure 1.   P. carinii extracts contain Cdc13 cyclin-like activity. Shown is histone H1 phosphorylated with immunoprecipitated Cdc2/Cdc13 protein complexes generated using an antibody recognizing B-type cyclins (Cdc13) from a variety of species (50 µg/ ml). The precipitated complexes were reacted in the presence of [-32P]ATP. P. carinii was isolated from infected rat lung by filtration through a 10-µm filter. To ensure that the observed protein kinase activity was from P. carinii and not the result of host cell contamination, equal volumes of uninfected rat lung were identically filtered, and the filtrate was extracted. (Lane A) Histone H1 reaction products after anti-Cdc13 precipitation demonstrate that P. carinii contains a Cdc13-like cyclin molecule capable of binding and coprecipitating the Cdc2 kinase, thus phosphorylating the histone H1 substrate. (Lane B) Nonimmune IgG (50 µg/ml) immunoprecipitation of P. carinii extract generates minimal kinase activity. (Lane C ) Minimal host cell protein kinase activity related to Cdc2/ Cdc13 complexes was present from uninfected rat lung after filtration through a 10-µ filter, extraction, and immunoprecipitation with antibody to Cdc13. Therefore, the kinase activity observed from the P. carinii extracts in lane A is related to P. carinii organisms and is not the result of residual host cells passing through the filter during the purification. (Lane D) Nonimmune IgG immunoprecipitates of uninfected lung extracts also had minimal kinase activity.

Characterization of the P. carinii cdc13 Genomic and cDNA Sequences

To further characterize the P. carinii Cdc13 B-type cyclin, we performed molecular cloning of a cognate P. carinii cdc13 gene. As an initial step, a 191-bp partial clone specific for P. carinii cdc13 was obtained by degenerate PCR amplification of P. carinii genomic DNA using oligonucleotide primers predicted from conserved amino-acid sequences of Cdc13 cyclins utilized by S. pombe and other fungi. The degree of degeneracy in the third position of each codon was restricted by incorporating the A+T-rich (> 65%) coding bias of P. carinii during primer design (39). The 191-bp amplification product was found to be unique on GenBank analysis, but similar to the respective cdc13 region from S. pombe.

Because P. carinii are derived from infected lungs and because amplification could have potentially resulted from contaminating host or other foreign DNA, extensive controls were undertaken to verify that the amplicon was indeed of P. carinii origin. This is generally accomplished by confirming that the PCR product is specifically present within the P. carinii genome. To accomplish this, hybridization of the 191-bp P. carinii cdc13 partial clone was conducted under high stringency conditions to a CHEF nitrocellulose blot of P. carinii chromosomal DNA. The 191-bp fragment hybridized to a single location on the chromosomal blot indicating the cdc13 gene is represented on a single chromosome within P. carinii (Figure 2). Furthermore, these chromosomal hybridization studies strongly indicate that the 191-bp partial clone of P. carinii cdc13 is derived from the organism's genome and is not the result of amplifying sequences from the host or other contaminants.

View larger version (79K): [in this window] [in a new window]   Figure 2.   The 191-bp P. carinii cdc13 probe obtained by PCR hybridizes to a single P. carinii chromosome under high stringency conditions. (Lanes 1-3) Shown are P. carinii chromosomes resolved by CHEF and stained with ethidium bromide. Each lane represents P. carinii chromosomes isolated from a single rat. (Lanes 4-6 ) The separated P. carinii chromosomes were transferred to nitrocellulose and hybridized to the 191-bp [32P]-labeled P. carinii cdc13 probe, which was obtained by degenerate PCR amplification. Note that the cdc13 probe hybridizes specifically to one P. carinii chromosome (arrowhead ).

Subsequently, the 191-bp amplicon was used to identify a 1.8-kb, full-length P. carinii genomic cdc13 clone by screening a rat-derived P. carinii sp. f. carinii gt11 genomic library (Figure 3A; GenBank accession number AF097334). In parallel, a complete cDNA for P. carinii cdc13 was derived by screening a rat-derived P. carinii cDNA library cloned into the bacteriophage Uni-ZAP XR (GenBank accession number AF097476). Comparison of the genomic and cDNA sequences indicated that the P. carinii cdc13 gene is composed of three exons and two introns, which contain an open reading frame encoding 459 amino acids with a predicted molecular mass of 52,630 D (Figure 3B). The predicted protein structure is compatible with a hydrophilic soluble cytoplasmic protein. The predicted P. ca-rinii Cdc13 contains 31.37% polar amino-acid residues, 13.07% acidic residues, 15.68% basic residues, and 39.9% nonpolar amino-acid residues. The estimated isoelectric point is 7.04. The protein contains three cysteine residues capable of intrachain disulfide bond formation.

View larger version (7K): [in this window] [in a new window]   View larger version (39K): [in this window] [in a new window]   Figure 3.   Characterization of the P. carinii cdc13 gene. (A) Nucleotide and predicted amino-acid sequence of the P. carinii cdc13 B-type cyclin derived from P. carinii genomic DNA. The predicted peptide used to generate the specific P. carinii Cdc13 antiserum is underlined. (B) P. carinii cdc13 genomic and cDNA clones were sequenced using overlapping primers. Shown are three exons (boxes) determined by comparison of the genomic and cDNA clones and using known characteristics of P. carinii intron-exon splicing. The exons encode an open reading frame of 459 amino acids with a predicted mass of 52,630 D.

Computer analysis of the GenBank database revealed the P. carinii cdc13 gene to be unique. The closest sequence similarity was found at the nucleotide level (BlastN comparison) to S. pombe (71% identity), at the amino-acid level (BlastP) to S. pombe (49% identity), and following translation into six reading frames (BlastX analysis) also to S. pombe (54% identity). The cloned sequences were used to probe freshly isolated P. carinii mRNA (Figure 4A). The transcript was 1.4 kb in size, in comparison to the 1.1-kb P. carinii actin mRNA. The level of steady-state mRNA for P. carinii cdc13 appeared stable over 2.0 h after the organisms were removed from the host. Subsequently, the messages for both Cdc13 and actin deteriorated, presumably as the organisms lost viability.

View larger version (43K): [in this window] [in a new window]   View larger version (21K): [in this window] [in a new window]   Figure 4.   The cloned P. carinii cdc13 DNA sequences recognize RNA from P. carinii organisms and can be translated into an appropriately sized protein. (A) P. carinii were freshly purified, total RNA extracted, and probed with the cloned cdc13 gene sequences in Northern hybridization assays. The nitrocellulose blot was subsequently stripped and rehybridized with P. carinii actin (lower panel ). Each lane contains 5 µg of total RNA obtained under those steady-state conditions. Total RNA obtained from healthy rat lung exhibited no hybridization to either P. carinii cdc13 or to P. carinii actin (negative controls). However, the total RNA obtained from freshly isolated P. carinii revealed a transcript of ~ 1.4 kb in length, in contrast to the 1.1 kb P. carinii actin RNA. (B) Next, the cDNA was used to recombinantly express P. carinii Cdc13 cyclin protein. Shown are reticulocyte lysate transcription-translation reactions of P. carinii cdc13 cDNA followed by immunoprecipitation with the specific P. carinii Cdc13 antiserum. This antiserum was able to immunoprecipitate the 50.5-kD P. carinii Cdc13 from reaction mixtures. In contrast, the nonimmune serum did not precipitate any products from the reaction.

Variations in proteins derived from different sources of P. carinii organisms have been previously reported (24, 45). A 0.2% discrepancy in nucleotide sequence was observed comparing the genomic clone, derived from the P. carinii gt11 genomic library (University of Cincinnati), with the cDNA sequences cloned from the P. carinii Uni-ZAP XR bacteriophage cDNA library (AIDS Research and Reference Reagent Program). The genomic and the cDNA P. carinii cdc13 sequences differ by only 3 bp. At genomic nucleotide 692, guanine is replaced with adenine in the cDNA. At genomic nucleotide 883, guanine is again replaced with adenine. Lastly, at genomic nucleotide 962, cytosine is replaced by thymine in the cDNA clone. These differences likely represent variations in these two P. carinii sources. The amino-acid sequence of the proteins predicted from the genomic and cDNA are largely identical, with only the first substitution noted previously resulting in one amino-acid difference of a glutamate¹²³ in the genomic clone compared with a lysine¹²³ predicted from the cDNA clone. Because Cdc13 is an essential cell-cycle regulatory molecule, it is not surprising that distinct isolates of P. carinii from independent sources contain Cdc13 molecules whose protein sequences are nearly completely conserved.

Generation of a P. carinii Cdc13 Antiserum and Recombinant Expression of Cdc13 B-Type Cyclin Protein

To further characterize the Cdc13 cyclin protein found within P. carinii, we generated an antibody directed against a unique 15 amino-acid synthetic peptide sequence predicted from the P. carinii Cdc13 cDNA. The cdc13 cDNA was further used to recombinantly express the P. carinii Cdc13 cyclin protein in reticulocyte lysate-based in vitro transcription and translation experiments. The recombinant Cdc13 protein was specifically immunoprecipitated with the P. carinii Cdc13 antiserum, but not with nonimmune serum, and exhibited a relative molecular mass of 50.5 kD (Figure 4B). Thus, the synthetic peptide antiserum recognizes an exposed epitope on the intact P. carinii Cdc13 cyclin molecule.

We further used the P. carinii Cdc13 antiserum to immunoblot fresh extracts of P. carinii to assess the presence and molecular mass of the native P. carinii cyclin. The antiserum specifically reacted predominantly with a 50-kD protein extracted from purified P. carinii organisms (Figure 5, lane A). The molecular mass of this native protein is notably similar to that of our recombinant Cdc13 (Figure 4B). In contrast, nonimmune serum exhibited no specific reactivity (Figure 5, lane B). Thus, the unique amino-acid sequence deduced from the cloned cDNA is contained both within the recombinantly expressed Cdc13 and within a native P. carinii protein found in extracts of the organism. This Cdc13 protein exhibits an apparent molecular mass comparable to other Cdc13 proteins found in related fungi (37, 38).

View larger version (46K): [in this window] [in a new window]   Figure 5.   The P. carinii Cdc13 antiserum recognizes a 50-kD protein present in P. carinii extracts. A synthetic peptide antiserum recognizing a unique 15 amino-acid sequence of the predicted P. carinii Cdc13 cyclin protein was generated in rabbits and used to immunoblot P. carinii extracts. (Lane A) The immune antiserum to P. ca-rinii Cdc13 localizes predominantly to a 50.5-kD protein in P. carinii extracts. (Lane B) In contrast, nonimmune serum exhibited no specific reactivity with separated P. carinii proteins.

Activity of the P. carinii Cdc13-Like Protein Kinase Is Differentially Regulated during the Organism's Life Cycle

The P. carinii Cdc13 antiserum was further used to evaluate the specific activity of Cdc2/Cdc13 cyclin-dependent kinase complexes present in P. carinii. Extracts derived from unfractionated populations of P. carinii containing both cysts and trophic forms exhibited typical protein kinase activity against the target histone H1 substrate when precipitated with Cdc13 antiserum, but not when treated with nonimmune serum (Figure 6A). To further examine the relative specific activity of these Cdc2/Cdc13 complexes over the life cycle of P. carinii, freshly isolated P. carinii were separated into cyst and trophic form populations by differential filtration, extracted, and equal quantities (150 µg) of cellular protein immunoprecipitated with the P. carinii Cdc13 antiserum (Figure 6B). Interestingly, substantially greater specific activity per microgram of cytoplasmic protein was present in cellular extracts derived from P. carinii cysts compared with P. carinii trophic forms. No significant activity was detected in an equal amount (150 µg) of uninfected lung extract. Thus, the net overall specific activity of the Cdc2/Cdc13 cyclin-dependent kinase complexes appears to be differentially regulated over the life cycle of P. carinii.

View larger version (28K): [in this window] [in a new window]   View larger version (25K): [in this window] [in a new window]   Figure 6.   P. carinii exhibits differential activity of Cdc13 over its life cycle. Shown is histone H1 phosphorylated with immunoprecipitated Cdc2/Cdc13 protein complexes generated using the specific synthetic peptide antiserum to P. carinii Cdc13 (50 µg/ml). The precipitated complexes were reacted in the presence of [-32P]ATP. (A) Histone H1 reaction products from mixed P. ca-rinii populations (trophic forms and cysts) using the specific anti-Cdc13 precipitation demonstrate that P. carinii contain a Cdc13 cyclin capable of binding and coprecipitating the Cdc2 kinase, thereby phosphorylating the histone H1 substrate. In contrast, no active Cdc2/Cdc13 complexes are precipitated using nonimmune serum. (B) Next, P. carinii were separated into trophic forms and cysts by differential filtration. Equal amounts of cellular protein (150 µg) were immunoprecipitated with the anti-Cdc13 serum and reacted in the histone H1 kinase assay. Cyst extracts consistently showed greater specific activity than trophic form extracts. In addition, an equal amount of cellular protein from uninfected lung (150 µg) exhibited no significant level of kinase activity.

P. carinii cdc13 cDNA Can Function to Restore Proliferation in S. pombe Strains Containing Temperature-Sensitive Deficiencies of S. pombe Cdc13 Activity

Finally, we sought to determine the critical activity of P. carinii cdc13 cDNA in mediating fungal cell-cycle progression. At present, P. carinii cannot be genetically manipulated. Therefore, direct testing of cdc13 in P. carinii itself, with the goal of assessing impact on P. carinii proliferation, is not yet technically feasible. Instead, we evaluated the biologic activity of P. carinii Cdc13 by studying its ability to complement S. pombe mutants exhibiting temperature-sensitive deficiencies in endogenous Cdc13 activity (Figure 7). We have previously exploited a similar approach to identify the function of cdc2 in S. pombe (24). S. pombe mutants have been previously identified that grow at the permissive temperature of 30°C, but fail to enter mitosis at the nonpermissive temperature of 37°C (28). P. carinii cdc13 cDNA sequences were subcloned in the S. pombe expression vector pREP41 (30). S. pombe mutants with temperature-sensitive Cdc13 were transformed with P. carinii cdc13 cDNA in pREP41 and selected by proliferation at 37°C on solid medium without leucine and thiamine. The P. carinii cdc13 cDNA restored proliferation of these mutant S. pombe containing temperature-sensitive deficiencies of endogenous S. pombe Cdc13 activity under the nonpermissive 37°C temperature conditions. Isolated colonies were propagated at 37°C and plasmid DNA from P. carinii cdc13-complemented colonies was isolated to verify the presence of the P. carinii cdc13 gene in transformants growing at 37°C. S. pombe cdc13 mutant strains expressing temperature-sensitive Cdc13 transformed with the pREP41 vector alone failed to proliferate at the nonpermissive 37°C temperature. In contrast, wild-type S. pombe with normal Cdc13 expression also exhibited normal growth at both 30°C and 37°C. Furthermore, the addition of thiamine, which represses expression of the pREP41 plasmid, resulted in complete suppression of growth of the S. pombe mutants transformed with P. carinii cdc13 at the nonpermissive 37°C temperature. These experiments strongly indicate that the P. carinii cdc13 cDNA encodes a fully functional Cdc13 B-type cyclin protein, which can mediate fungal cell-cycle completion and organism proliferation.

View larger version (35K): [in this window] [in a new window]   Figure 7.   P. carinii cdc13 cDNA complements the DNA of mutant S. pombe, which lacks endogenous Cdc13 cyclin activity at 37°C, restoring fungal growth. The orientations of S. pombe streaks are the same on all plates. (1) S. pombe mutants expressing temperature-sensitive Cdc13 transformed with P. carinii cdc13 cDNA cloned into pREP41. (2) S. pombe mutants transformed with pREP41 vector alone. (3) Wild-type S. pombe containing wild-type S. pombe cdc13 (active at all temperatures). (A) Plates incubated at the permissive temperature of 30°C in the absence of thiamine show normal growth of all S. pombe after transformation with these plasmids. (B) Incubation at 30°C in the presence of thiamine (10 µM) also demonstrates normal growth of these S. pombe, verifying that thiamine is not toxic. (C ) Identical plates at the nonpermissive temperature of 37°C in the absence of thiamine reveals no growth of S. pombe mutants expressing temperature-sensitive Cdc13 transformed with pREP41 vector alone. However, normal growth of the S. pombe mutants occurs when transformed with P. carinii cdc13 cDNA and in wild-type S. pombe containing wild-type cdc13. (D) Incubation at 37°C in the presence of thiamine demonstrates that repression of P. carinii cdc13 in pREP41 results in no growth of the mutant S. pombe. Thus, P. carinii Cdc13 is functional in permitting proliferation of the fungus even under the nonpermissive temperature. Note that the wild-type S. pombe with wild-type cdc13, unlike the P. carinii cdc13 in pREP41, is equally expressed in the presence or absence of thiamine.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Despite effective prophylaxis and available antimicrobial therapy, P. carinii remains among the most important causes of life-threatening pneumonia in immunocompromised patients (2). Although the life cycle of this fungus has not been fully delineated, recent studies from our laboratory indicate that P. carinii uses a Cdc2 cyclin-dependent kinase-based regulatory system to regulate traverse through its cell cycle (24, 46). Cdc2 serine-threonine kinases must interact with cognate cyclins in order to maintain a conformation capable of enzymatic activity (47). In the current investigation, we demonstrate that P. carinii expresses a functional Cdc13 B-type cyclin. Analyses of the P. carinii Cdc13 genomic sequence demonstrate that this molecule is homologous to mitotic cyclins in related fungi (48). Furthermore, the activity of Cdc2/Cdc13 complexes is differentially regulated over the life cycle of the fungus, being higher in cystic forms of P. carinii compared with trophic forms. Finally, the functional capacity of P. carinii Cdc13 to promote cell-cycle traverse and fungal proliferation was demonstrated by heterologous expression of P. carinii Cdc13 in S. pombe. Mutant S. pombe that lack Cdc13 activity under nonpermissive temperature conditions were complemented by P. carinii cdc13 cDNA, and fungal proliferation was restored.

Most of what has been learned of the life cycle of P. carinii has been deduced from electron microscopic images and short-term in vitro studies of P. carinii interacting with lung cells (14, 15, 19, 49). P. carinii organisms found in infected lung tissues are mostly present as numerous trophic forms containing a single nuclear body. Fewer numbers of organisms are present as thick-walled cystic forms, which possess up to eight daughter nuclei (13). Fluorescence-based quantitation of P. carinii nuclear content suggests that the trophic forms are haploid in nature (1C), whereas the cyst contains up to 8C DNA content (50, 51). Binding of trophic forms to alveolar epithelial cells is a universal feature of P. carinii infections in both animals and humans, and is also consistently present in short-term in vitro experiments of P. carinii maintained on cultured lung epithelial cells (13, 49). Available evidence indicates that binding of P. carinii trophic forms to lung epithelial cells promotes proliferation of the organism (17, 49).

Eukaryotic cell-cycle progression is carefully regulated by the coordinate activation of cell division cycle (Cdc) control proteins (21, 23). Much of our understanding of fungal cell-cycle control has been learned by studying cdc gene mutants of S. pombe, an easily cultivated organism with notable phylogenetic homology to P. carinii (9). Cdc genes encode for critical components of the cyclin-dependent kinase system, including the Cdc2 kinase and its essential Cdc13 cyclin partner. The Cdc2 kinase represents the predominant cyclin-dependent kinase present in most fungi (21, 24). Whereas the level of protein expression of Cdc2 proteins generally remains constant over most eukaryotic cell cycles, the levels of their obligatory cyclins, such as Cdc13, rise and fall throughout the cell cycle. In related fungi, Cdc13 B-type cyclins normally peak in amount in late G2 and early M phase (21). The associated formation of activated Cdc2/Cdc13 complexes phosphorylate critical target proteins such as nuclear lamin, leading to disassembly of the nucleus and subsequent cellular division (52). In addition, B-type cyclins have been shown to specifically activate the Cdc25 phosphatase, a protein that promotes mitosis by removal of an inhibitory Tyr15 phosphate modification on Cdc2 (53, 54).

In the present investigation, we have identified a Cdc13 B-type cyclin molecule from P. carinii, a clinically significant human and animal pathogen. Although current methodologies do not yet permit precise identification of cell-cycle phases in this species, we have observed significant differences in the specific activity of Cdc2/Cdc13 complexes in the currently separable distinct life cycle forms of P. carinii. Specifically, by immunoprecipitation of these complexes using specific antibodies directed against the Cdc13 cyclin, we have observed substantially greater specific activity of Cdc2/Cdc13 kinase complexes per microgram of cellular protein in P. carinii cysts compared with trophic forms. It is of interest that our earlier assays conducted using antibodies against the Cdc2 PSTAIR amino-acid sequence suggested relatively greater activity in trophic forms (24). This raises important possibilities that P. carinii Cdc2 may also interact with additional, yet unidentified, cyclin partners during cell-cycle regulation of this organism. Genetic and functional searches for other regulatory elements are underway. It should also be noted that the PSTAIR amino-acid sequence is additionally present in other kinases. The immunoprecipitation of other protein kinases may have confounded the specific measurement of Cdc2-related activity in P. carinii (55). In contrast, in the present study, computer-assisted searches verified that the peptide used to generate the P. carinii Cdc13 antiserum was completely unique compared with available databases. Thus, these immunoprecipitation kinase activity assays strongly indicate that Cdc2/Cdc13 complex activity is differentially regulated over the life cycle of P. carinii.

To test the function of Cdc13 in P. carinii one would ideally seek to directly modify Cdc13 expression or activity through molecular genetic means and observe the effect on P. carinii growth, morphology, and nuclear content. Such an approach is currently not feasible in light of the unavailability of a consistently reliable transformation system that provides high levels of transgene expression in these organisms. To circumvent these laboratory obstacles, we used the functional conservation of Cdc13 molecules by testing the ability of P. carinii cdc13 cDNA to complement growth of S. pombe mutant strains that lack endogenous Cdc13 activity under nonpermissive temperature conditions (28). By selectively restoring growth of these S. pombe strains under conditions where S. pombe Cdc13 is deficient and P. carinii Cdc13 is expressed, we can strongly infer that P. carinii Cdc13 is functionally capable of interacting with the Cdc2 serine-threonine kinase, which promotes cell-cycle completion and organism proliferation. Unlike complementation with respect to Cdc2 catalytic subunits, cross-species complementation of Cdc13 molecules has not been widely reported in the literature (13, 24, 28). This likely reflects important regulatory functions of Cdc13 molecules, which are distinctly controlled by the host. Our observation of P. carinii Cdc13 complementing S. pombe growth further supports a fair level of phylogenetic similarity between these species.

Thus, the current study provides additional strong evidence that important mechanisms of cell-cycle regulation, which have been gleaned from studies of S. pombe, may be applied and tested to enhance our understanding of the life cycle of P. carinii. The heterologous expression of P. carinii cell-cycle control proteins in S. pombe additionally represents a convenient laboratory approach to broaden our understanding of the biochemical regulation of these P. carinii life-cycle control molecules using this conveniently culturable organism. The identification of S. pombe strains that proliferate under the control of P. carinii cdc2 (24) and Cdc13 (this work) further opens the possibility of testing compounds that inhibit the Cdc2/Cdc13 complex as potential therapeutic agents for treating P. carinii infection (56). Despite the presence of Cdc2/Cdc13 complexes in human and other mammalian cells, accumulating evidence indicates that small molecule inhibitors of the Cdc2/ Cdc13 complex exhibit substantially different activity against cell-cycle regulation between fungi and mammalian hosts (57). Thus, sufficient divergence of the structure of Cdc2/ Cdc13 complexes in the fungus and the human host might be exploited in the development of potential therapeutic agents that may suppress P. carinii growth with minimal host cell toxicity.

It is likely that a better understanding of molecular regulation of the P. carinii life cycle will provide important clues as to the mechanism of cell-cycle arrest that occurs when P. carinii is removed from the host lung, as well as the manner by which binding of P. carinii trophic forms to alveolar epithelial cells promotes organism proliferation. It is intriguing that experimental disruption of cyclin B activity from the cig1 and cig2 genes in S. pombe results in the formation of large cells containing multiple nuclei each with 1C DNA content, a phenotype not dissimilar from that observed in cystic forms of P. carinii (28). Future studies aimed at manipulating the activity of the cyclin- cyclin-dependent kinase regulatory system of P. carinii may provide additional new approaches to stimulate organism proliferation under cell-free culture conditions.

In summary, these investigations demonstrate that P. carinii possesses a functional Cdc13 B-type cyclin, which interacts with its Cdc2 kinase partner, yielding variable activity over the life cycle of the organism, and which is capable of promoting fungal proliferation when expressed in the related fungus S. pombe. Together the Cdc2/Cdc13 cyclin-cyclin-dependent complex represents a principal component of cell-cycle regulation. Molecular definition of these molecules in P. carinii provides important new tools for better understanding the life-cycle regulation of this intractable organism. In addition, such molecules may provide novel molecular targets for the prevention and treatment of P. carinii pneumonia in immunocompromised patients.