The PHOA and PHOB Cyclin-Dependent Kinases Perform an Essential Function in Aspergillus nidulans

Corresponding author: Stephen A. Osmani, The Ohio State University, 804 Riffe Bldg., 496 W. 12th Ave., Columbus, OH 43210. E-mail osmani.2@osu.edu

Communicating editor: B. ANDREWS

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Unlike Pho85 of Saccharomyces cerevisiae, the highly related PHOA cyclin-dependent kinase (CDK) of Aspergillus nidulans plays no role in regulation of enzymes involved in phosphorous acquisition but instead modulates differentiation in response to environmental conditions, including limited phosphorous. Like PHO85, Aspergillus phoA is a nonessential gene. However, we find that expression of dominant-negative PHOA inhibits growth, suggesting it may have an essential but redundant function. Supporting this we have identified another cyclin-dependent kinase, PHOB, which is 77% identical to PHOA. Deletion of phoB causes no phenotype, even under phosphorous-limited growth conditions. To investigate the function of phoA/phoB, double mutants were selected from a cross of strains containing null alleles and by generating a temperature-sensitive allele of phoA in a phoB background. Double-deleted ascospores were able to germinate but had a limited capacity for nuclear division, suggesting a cell cycle defect. Longer germination revealed morphological defects. The temperature-sensitive phoA allele caused both nuclear division and polarity defects at restrictive temperature, which could be complemented by expression of mammalian CDK5. Therefore, an essential function exists in A. nidulans for the Pho85-like kinase pair PHOA and PHOB, which may involve cell cycle control and morphogenesis.

THE PHOA cyclin-dependent kinase (CDK) of Aspergillus nidulans (BUSSINK and OSMANI 1998 ) is a member of the Pho85 family of cyclin-dependent kinases. PHO85 was isolated in Saccharomyces cerevisiae due to its involvement in regulation of phosphate-scavenging enzymes (UESONO et al. 1987 ; TOH-E et al. 1988 ) and has been implicated in numerous other cellular processes, including stress adaptation, glycogen storage, and cell cycle progression (see ANDREWS and MEASDAY 1998 ; NISHIZAWA et al. 2001 ; CARROLL and O'SHEA 2002 for reviews and further references). Pho85 has been shown to bind to 10 cyclin partners (MEASDAY et al. 1997 ), which are thought to target the kinase activity of Pho85 to numerous substrates involved in these processes. However, PHO85 is not an essential gene nor is the highly related Pef1 kinase of the fission yeast Schizosaccharomyces pombe (TANAKA and OKAYAMA 2000 ).

PHOA of A. nidulans is not apparently involved in regulation of phosphate-scavenging enzymes but, when compared to isogenic wild-type strains, lack of PHOA causes a switch from asexual to sexual development (at pH 8.0) or the absence of development altogether (at pH 6.0) under limiting phosphate growth conditions. A. nidulans PHOA is therefore required to integrate environmental conditions with developmental fate to allow ordered differentiation of asexual and sexual cell types under varying growth conditions (BUSSINK and OSMANI 1998 ). Deletion of the kinase does not cause lethality or developmental defects under nonrestrictive growth conditions.

We have isolated a second CDK with 77% identity to PHOA and report the consequences of its deletion in combination with a lack of PHOA in recombinant germinating ascospores. A temperature-sensitive allele of phoA was also generated, which caused growth defects in combination with deletion of phoB. The ability of human Cdk5 to complement loss of PHOA/PHOB function was also determined. The data demonstrate that PHOA and PHOB function redundantly and have an essential function involved in cell growth and nuclear division.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Strains and medium:
The A. nidulans strains used in this study are listed in Table 1. Minimal, yeast extract glucose (YG), and malt extract glucose (MAG) media with supplements were prepared as described previously (PONTECORVO 1953 ; BUSSINK and OSMANI 1998 ).

Dominant-negative phoA mutagensis:
Plasmid pPAP (M47) containing phoA under control of the alcohol dehydrogenase promoter was mutated using the QuickChange site-directed mutagenesis kit (Stratagene, La Jolla, CA). Lysine 39 of phoA was replaced by an arginine codon using oligonucleotides PHOALU-1(5'-GAA CTT GTC GCC CTG AGG GAA ATC CAC CTC-3') and PHOALU-2 (5'-GAG GTG GAT TTC CCT CAG GGC GAC AAG TTC-3'). The plasmid with mutant phoA was transformed into A. nidulans strain GR5 and transformants were streaked to single colonies three times before replica plating on both inductive (ethanol) and repressive (glucose) media to identify any strains inhibited by dominant-negative PHOA.

phoB deletion:
A phoB bacterial artificial chromosome (BAC) clone was obtained by screening an A. nidulans genomic BAC library (obtained from W. Choi, H. Zhu and R. A. Dean, Genomics Institute, Clemson University). Escherichia coli strain DY380 (LEE et al. 2001 ; SWAMINATHAN et al. 2001 ) was transformed with BAC DNA harboring the phoB gene. A pyroA/Zeo gene cassette was amplified from plasmid pPZD with primers PPA (5'-GAG ACC CTG TTT TCT TGT ATT CGT ATC AAT CAT ATT TTT ATT CAT TTT CTG AAT TCT CAG TCC TGC TCC T-3') and PZP (5'-TTC TTT GTT CTC ACC TGG CTA CGA ATC ACA CAG GGA CTG GCT GAG CCC CAC AGT TGA GCC TGA GAC CAA T-3'). Plasmid pPZD was generated by cloning the Zeo gene from pCDA21 as an EcoRI-BamHI fragment into plasmid p14 containing the pyroA gene of A. nidulans (OSMANI et al. 1999 ). The amplified pyroA/Zeo gene cassette was transformed into the above BAC-transformed E. coli DY380 competent cells to promote in vivo homologous recombination as described (LEE et al. 2001 ; SWAMINATHAN et al. 2001 ). The successful replacement of phoB in the BAC by the pyroA/Zeo cassette was checked with PCR amplification using primers anchored out of the phoB gene deletion area. The phoB-deleted BAC DNA was prepared and used to transform A. nidulans strain GR5, as described (OSMANI et al. 1987 ). The deletion of phoB was confirmed by both PCR amplification as above and Western blot analysis as described (BUSSINK and OSMANI 1998 ) using antibodies raised against PHOA that also detect PHOB. The phoB null allele was named phoB.

A. nidulans crossing and growth conditions:
Crosses were completed as previously described (PONTECORVO 1953 ) and mature cleistothecia were collected and cleaned on 4% agar plates and broken into 0.2% Tween 80 solution. Ascospore germination was checked on minimal medium with supplements as indicated and grown at 37°. The test of sexual differentiation on phosphate-limited media, growth sensitivity or resistance to hydroxyurea, high NaCl and sucrose media, and the activity test of three phosphate-regulated extracellular phosphatases (alkaline phosphatase, acid phosphatase, and phosphodiesterase) were as described (BUSSINK and OSMANI 1998 ).

The construction of the phoA^TS+phoB strain:
phoA was mutated in plasmid pRG3 using oligonucleotides phoAW337H (5'-GGC GCT CTG CAG CAT CCA CAC TTC CAT GAC CTT CCG CAG) and phoAW337H-R (5'-CTG CGG AAG GTC ATG GAA GTG TGG ATG CTG CAG AGC GCC-3') as described above to generate the W337H mutation. The resulting plasmid DNA was used to transform a phoB strain (AZ16). Transformants were streaked to single colonies three times and conidial spores were harvested and spread on minimal medium uridine and uracil (UU) plates plus 1% 5-fluoroorotic acid (5-FOA) at a density of 4000–6000 spores/plate. Viable colonies on FOA plates were streaked to single colonies three times on minimal media plus uridine and uracil and checked for growth at 32° and 42°. The mutations in those colonies showing temperature sensitivity were confirmed by sequencing genomic DNA amplified using PCR.

Human CDK5 complementation:
Human CDK5 gene was a kind gift from B. Andrews, which was cloned into the BamHI site of the pAL5 expression plasmid under the control of the alcA regulatable promoter. The phoA^TS+phoB strain XD1 was transformed with the CDK5 expression plasmid and successful transformation was confirmed by PCR amplification using CDK5-specific primers.

4',6-Diamidino-2-phenylindole staining and determination of the spindle mitotic index:
Nuclei were visualized using 4',6-diamidino-2-phenylindole (DAPI) staining and the spindle mitotic index was determined after immunofluorescence staining of microtubules as described (DE SOUZA et al. 2000 ) using mouse anti--tubulin antibody (TAT1, Amersham, Buckinghamshire, UK) and a second antibody (AlexA Fluro 595 goat anti-mouse IgG, Molecular Probes, Eugene, OR). A total of 500 ascospore germlings were counted for each strain to determine its mitotic index. The spindle mitotic index of phoA^TS strain at 32° or 42° was obtained by directly counting spindle bars in strain XG1. XG1 strain was obtained by crossing XD1 (phoA^TS+phoB) with LO1022 [-tubulin-green fluorescent protein (GFP) tag].

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Identification of a second nonessential Pho85-like kinase in A. nidulans:
Two lines of evidence suggest that A. nidulans may encode a second Pho85-like kinase in addition to the previously isolated PHOA kinase. First, polyclonal antiserum raised against recombinant PHOA detects not only PHOA but also an additional band of a size predicted for a CDK (BUSSINK and OSMANI 1998 ; Fig 2B). Second, expression of a kinase negative form of PHOA (lysine 39 was converted into arginine) using the inducible alcA promoter severely restricts growth. This indicates that it acts in a dominant-negative fashion (data not shown), suggesting that phoA has an essential but redundant function in addition to its role during development. We therefore searched the A. nidulans genome sequence for an additional phoA-like gene at the Monsanto Microbial Sequence Database (http://microbial.cereon.com/), which is now publicly available at http://www-genome.wi.mit.edu/annotation/fungi/aspergillus/index.html. Such a sequence was identified and termed phoB. Sequence analysis of both genomic and cDNA clones revealed an open reading frame encoding a CDK with 77% identity to PHOA and 68% identity to Pho85 of S. cerevisiae. Other highly related CDKs include S. pombe Pef1 (69%) and human Cdk5 (56%; Fig 1).

View larger version (101K): [in this window] [in a new window] [Download PPT slide]   Figure 1. Similarity between PHOA and PHOB with other cyclin-dependent kinases. The protein sequence alignment of PHOA and PHOB (A. nidulans), Pho85 (S. cerevisiae), Pef1 (S. pombe), and CDK5 (Homo sapiens) was done using Clustal W (http://www.ebi.ac.uk/clustalw/). The PSTAIRE motif is underlined. Completely identical amino acids are indicated by an asterisk and conserved amino acids by a colon.

View larger version (93K): [in this window] [in a new window] [Download PPT slide]   Figure 2. Deletion of phoB. (A) A PCR approach was used to analyze DNA isolated from 17 strains transformed with a phoB-specific deletion cassette. If phoB is deleted, the resulting PCR band is shifted higher (indicated by MT). If no deletion has occurred, then a WT-sized fragment is amplified. Of the 17 strains analyzed, 4 deleted strains are indicated (1–4). (B) Western blot analysis of wild-type strains (R153 and GR5), strains deleted for phoA (HB9 and 17), and strains with deleted phoB (GZ21 and AZ16). All strains were grown for 18 hr in minimal medium and protein extracts were analyzed by Western blotting using a polyclonal antibody raised against E. coli expressed PHOA. The position of PHOAM1, PHOAM47 (BUSSINK and OSMANI 1998 ), and PHOB is indicated to the left. An asterisk indicates the position of missing proteins in deleted strains.

To analyze the function of phoB, a deletion construct was generated using recombination in E. coli (see MATERIALS AND METHODS) in which phoB was replaced with the A. nidulans pyroA gene (OSMANI et al. 1999 ). This construct was used to transform A. nidulans to pyridoxine prototrophy and, using PCR analysis, four strains were identified with phoB deleted (Fig 2A).

Because additional putative CDK bands were previously detected using antiserum raised against PHOA, cross-reactive bands were investigated in proteins isolated from phoA- and phoB-deleted strains along with wild-type strains (Fig 2B). Strains with phoA deleted (HB9 and 17) lacked the PHOA-specific PHOA^M1 and PHOA^M47 bands as expected. However, another 34-kD cross-reactive band was still present. Conversely, in the phoB-deleted strains GZ21 and AZ16, the 34-kD band was missing and the PHOA proteins were still present. This demonstrates that the 34-kD protein represents PHOB and that phoB is a nonessential CDK in A. nidulans. The deleted alleles of phoB were designated phoB.

Deletion of phoB does not affect growth or differentiation:
Previous studies have shown that lack of phoA allows accumulation of an undefined brown pigment when grown on low-P_i medium while little pigment is formed in the wild-type strain (BUSSINK and OSMANI 1998 ). phoB-deleted strains were therefore tested for pigment production under these conditions and they responded exactly like the wild-type control strain with no visible pigment being formed (Fig 3A). A phoB strain was also tested for sensitivity or resistance to hydroxyurea, high NaCl, and sucrose media with no defects being observed (data not shown).

View larger version (64K): [in this window] [in a new window] [Download PPT slide]   Figure 3. Phenotypic analysis of phoB strains. (A) Strains with the indicated mutations were grown on 0.1 mM Pi medium (pH 8.0) for 4 days along with a WT control. Under these conditions pigment production in the media is apparent in the phoA strain but not in phoB strains or in the WT control. (B) Conidia of strains WT, phoA (HB9), and phoB (GZ3121) were inoculated in top agar at densities (spores per plate) and on media (with 0.1 or 11 mM phosphate, at pH 8.0 or pH 6.0) as indicated, and the plates were photographed at low magnification after 4 days of growth. Asexual development results in formation of conidiophores carrying chains of conidia that give the green color, whereas sexual development is indicated by the yellow color of Hulle cells occurring in clusters that surround nascent cleistothecia. Little developmental difference can be seen between the wild-type and the phoB strain whereas the development of the phoA strain can be seen to switch to sexual development or no development, depending on growth conditions.

Because Pho85 is involved in regulation of phosphate-scavenging enzymes in S. cerevisiae and a similar system appears to function in Neurospora crassa (KANG and METZENBERG 1993 ; PELEG et al. 1996A , PELEG et al. 1996B ), we considered the possibility that phoB may have a similar function. A phoB strain was therefore tested for activity of three phosphate-regulated extracellular phosphatases (alkaline phosphatase, acid phosphatase, and phosphodiesterase) by activity staining of colonies grown in the presence or absence of inorganic phosphate (P_i). No obvious differences from wild type were observed (data not shown). This indicates that phoB, like phoA, plays no obvious or important role in the regulation of enzymes involved in phosphorous acquisition.

It has also been shown that lack of phoA affects development in confluent plate cultures depending on pH, inoculation density, and phosphate concentration. A phoB strain was therefore checked under similar growth conditions and its development compared to a wild-type and a phoA-deleted strain (Fig 3B). It is clear that deletion of phoB caused no defects in the differentiation programs of A. nidulans in response to these environmental conditions. This is in marked contrast to the defects caused by lack of phoA (Fig 3B), a CDK with 77% identity to PHOB.

Deletion of both phoA and phoB causes lethality:
To see if deletion of both phoA and phoB would cause lethality or other marked phenotypes, strains deleted for phoA or phoB were crossed together to see if viable phoA phoB progeny could be generated. To do this, two strains were crossed in which phoA and phoB were deleted using two different nutritional marker genes. The genetic makeup of the cross was arranged so that plating progeny on selective media would allow only the double-deleted recombinants the chance to germinate and grow. However, if the double mutant were inviable, no colony formation would occur on selective media.

For this analysis we used strain HB9, which contains phoA deleted with pyrG⁺ (previously termed phoA1 by BUSSINK and OSMANI 1998 , but referred to here as phoA). HB9 also carries the pyroA4 nutritional marker and therefore requires pyridoxine (pyro). HB9 has no spore color mutations and produced wild-type green asexual spores. The pertinent genotype for HB9 is phoA; pyroA4. HB9 was crossed to strain AZ16, which has phoB deleted with pyroA⁺ termed phoB. This strain contains the pyrG89 nutritional marker and so requires UU for growth and carries the white spore color marker wA3. The pertinent genotype for AZ16 is phoB; pyrG89. Crosses were completed using the forcing nutritional markers pyroA4 and pyrG89. Ascospores isolated from fruiting bodies (cleistothecia) were plated on nonselective media (+pyridoxine +uridine +uracil) to allow colony formation. Crossed HB9 x AZ16 cleistothecia were identified because they generated ascospores that formed both green and white recombinant colonies. Self-crossed cleistothecia produced green (HB9) or white (AZ16) colonies.

Ascospores (500) from HB9 (phoA) and AZ16 (phoB) self-crosses and from HB9 x AZ16 (phoA x phoB) crosses were spotted onto supplemented media (+Pyro +UU) and all generated colonies (Fig 4A). Media selective for either the pyrG⁺ (-UU) or pyroA⁺ (-Pyro) nutritional markers allowed the expected strains to grow. As expected, recombinants from the phoA x phoB cross could also grow on these media. Most importantly, however, no recombinants from the phoA x phoB cross were able to form colonies on media selective for both pyrG89 and pyroA4 markers (Fig 4A, - -). This indicates that the phoA phoB double mutants, even though prototrophic (phoA:pyrG⁺ phoB:pyroA⁺), are not viable (Fig 4A, - -). This genetic analysis demonstrates that phoA and phoB have a redundant but essential function.

View larger version (74K): [in this window] [in a new window] [Download PPT slide]   Figure 4. PHOA and PHOB play an essential function. (A) Ascospores of (a) a cross between a strain with deletion of phoA to a strain with deletion of phoB (phoA:pyrG; pyroA4 x phoB:pyroA; pyrG89), (b) self-cross of phoA:pyrG; pyroA4, (c) self-cross of wild type, (d) self-cross of phoB:pyroA; pyrG89 plated on minimal media with supplements as indicated. pyro, pyridoxine that complements the pyroA4 mutation; UU, uridine and uracil that complement the pyrG89 mutation. Plates were incubated for 2 days at 37°. (B) Ascospores from a wild-type (GR5; left, top) self-cross plated on nonselective media or ascospores from a phoA:pyrG; pyroA4 x phoB:pyroA; pyrG89 cross (left, bottom) plated on selective media for pyrG and pyroA for the time indicated. (Right) Ascospores from self-crossed phoA:pyrG; pyroA4 and phoB:pyroA; pyrG89 strains. The media employed were minimal media supplemented with uridine/uracil and pyridoxine unless otherwise indicated. Insets show higher magnification of arrowed cells. Bar, 50 µm. (C) The growth rates of wild-type strain ascospores (GR5, dashed line) and double phoA phoB mutant ascospores (916, solid line). (D) Nuclear division in germinating ascospores from a wild-type self-cross and a phoA x phoB cross after incubation at 37° for 6.5 or 24 hr.

Defects in the phoA phoB double mutant:
Although the phoA phoB mutant was unable to form visible colonies after 2 days, we examined very early growth of the mutant to investigate the defect caused by lack of phoA and phoB. Ascospores from a phoA x phoB cross were grown over a 24-hr period along with ascospores of a control GR5 (pyrG89; pyroA4) self-cross. To allow growth of the GR5 spores, they were plated on media with added pyridoxine, uridine, and uracil. To identify the double phoA phoB recombinants, the ascospores from the phoA x phoB cross were plated on minimal media lacking pyridoxine, uridine, and uracil. As phoA is deleted with pyrG⁺, and phoB is deleted with pyroA⁺, these conditions will allow germination of only the double phoA:pyrG⁺ phoB:pyroA⁺ recombinants. Over the 24-hr period of growth, GR5 ascospores germinated and formed small colonies visible without magnification. Over the same time period no visible colonies were formed from the phoA x phoB cross.

GR5 germlings could be observed after 6 hr of incubation and had grown to form a mat of cells by 12 hr (Fig 4B). Only 20% of the ascospores from the phoA x phoB cross were able to germinate and form germlings. Taking spore viability into account, this number fits into the theoretic expectation that one-fourth of the spores are phoA and phoB wild type, one-fourth are phoA wild type and phoB, one-fourth are phoA and phoB wild type, and one-fourth are phoA and phoB. Although the phoA phoB ascospores could geminate, they failed to continue to grow and arrested with cells 50 µm in length (Fig 4B, two germlings indicated by arrows and quantified in Fig 4C). This indicates that the phoA phoB ascospores are able to break dormancy and start normal growth processes but are unable to sustain growth.

A potential problem with the analysis is that other recombinant progeny from the phoA x phoB cross could perhaps geminate and form germlings. Three other classes of progeny are expected from this cross: (1) phoA; pyroA4, (2) phoB; pyrG89, and (3) pyrG89; pyroA4. We therefore allowed PHO17 (phoA; pyroA4) and AZ16 (phoB; pyrG89) to undergo self-crosses and inoculated the resulting ascospores on selective media (Fig 4B, right two panels). Neither set of ascospores was able to form germlings although the phoB; pyrG89 spores from the AZ16 self-cross swelled enough to split the two shells of the ascospore casings (CHAMPE and SIMON 1992 ). However, in no instance were germ tubes similar to those seen in the phoA x phoB cross observed (Fig 4B). We can therefore say with confidence that the ascospores that are able to form germ tubes from the phoA x phoB cross are indeed the phoA phoB mutant recombinants.

One potential factor involved in lack of continued growth of the phoA phoB ascospores could be defects in nuclear division as previously observed for temperature-sensitive cell cycle mutants (MORRIS 1976 ). The number of nuclear divisions of phoA phoB and control ascospores was therefore determined during 24 hr of growth (Fig 4D). Unlike uninucleated asexual spores, ascospores of A. nidulans are binucleate (CHAMPE and SIMON 1992 ). After 6.5 hr of germination, the majority of wild-type germlings had undergone at least one nuclear division, whereas the phoA phoB germlings remained largely at the two-nuclear-division stage, having undergone no nuclear divisions (Fig 4D). The rate of initial germ-tube extension for the phoA phoB mutant compared to that for GR5 controls was not markedly different (Fig 4C) although their abilities to complete the first cell cycle were notably different. This indicates a potential cell-cycle-specific defect in the phoA phoB mutant. Even after 24 hr of growth, half the phoA phoB germlings were still arrested with two nuclei. Only 38% underwent one mitosis and 12% underwent more than one nuclear division. In contrast, the control cells had grown to visible colonies at this time with uncountable numbers of nuclei (Fig 4D).

Another defect of the phoA phoB cells was a marked effect on nuclear structure. As revealed by DAPI staining, A. nidulans nuclei are normally oval in appearance and contain a domain occupied by the nucleolus with diminished DAPI staining. Typical nuclear morphology can be seen in Fig 5A in a germling of a wild-type (WT) strain after growth of ascospores for 6.5 hr at 37°. In contrast, nuclei in germlings of phoA phoB mutants at the same time of growth appear more compact and lacked the area of diminished DAPI staining occupied by the nucleolus in normal strains. This defect becomes more pronounced after 12 hr of growth with nuclear DNA becoming quite condensed and sometimes fragmented (Fig 5A, MT). We considered that these defects could be associated with defective mitosis, but virtually no spindles could be detected in the phoA phoB germlings for the first 9 hr of germination. In contrast, the wild-type spindle mitotic index increased to 4% after 5 hr of growth (Fig 5B). Therefore, as no spindles could be seen at a time when the nuclear DNA of the phoA phoB germlings was condensed, and nuclei lacked a visible nucleolus, these changes are unlikely to result from defective mitosis.

View larger version (35K): [in this window] [in a new window] [Download PPT slide]   Figure 5. (A) Nuclear morphology in phoA phoB germinating ascospores. Germinating ascospores from a WT were incubated for 6.5 hr and a phoA phoB ascospore (MT) was incubated at 37° for 12 hr. Cells were stained with DAPI to visualize nuclear number and morphology. Bar, 5 µm. (B) The percentage of cells in mitosis as determined by staining microtubules and scoring of the spindle mitotic index. A total of 500 cells of each strain at each time point were counted.

Defects in a phoA^TS allele in a phoB strain:
Because of potential endowment of PHOA and PHOB to ascospores when crossing phoA to phoB strains to look at the phenotypes caused by lack of phoA and phoB, we tried to generate a temperature-sensitive (ts^-) allele of phoA. Previously, ts^- alleles of nimX^cdc2 of A. nidulans were generated on the basis of mutations causing temperature sensitivity in cdc2 of S. pombe (OSMANI et al. 1994 ). On the basis of protein sequence alignment among NIMX^cdc2, CDK5, Pho85, PHOA, and PHOB, three similar point mutations were introduced into phoA and used to replace the wild-type allele of phoA in a phoB strain. In vitro mutagenesis was used to generate the individual mutations in phoA in a plasmid vector. A two-step gene replacement was completed and ts^- strains were identified after replica plating at permissive and restrictive temperatures. Of the three mutations introduced to phoA (F255 was replaced by L, G257 by S, and W337 by H), only the W337H mutation caused a ts^- phenotype (Fig 6). This mutation was called phoA^TS and was recessive to phoA.

View larger version (68K): [in this window] [in a new window] [Download PPT slide]   Figure 6. (A) Sequence similarity among PHOA, PHOB, Pho85 CDK5, and NIMX. The regions in which the three conserved amino acids F255, G257, and W337 were replaced by L, S, and H, respectively, to get the phoATS mutant are shown. (B) The phoATS/phoB (W337H) substitution conidia spores were point-inoculated on MAGUU plates for 24 hr. The wild-type strain SO51 and the ts- strain SO53 were also inoculated as controls. XD1 and XD2 are different phoATS/phoB transformants.

The phenotypes caused by phoA^TS were investigated by germinating phoA^TS/phoB strains in liquid minimal media at both permissive and restrictive temperatures. Even after 48 hr of incubation at 42°, very few conidia were able to send out a germ tube (Fig 7A) and 99% were unable to undergo any mitotic division arresting with a single nucleus per cell (Fig 7B). At the permissive temperature of 25° the vast majority of cells had undergone germ-tube emergence (Fig 7A) and mitotic divisions after 13 hr of growth.

View larger version (66K): [in this window] [in a new window] [Download PPT slide]   Figure 7. Phenotypes caused by phoATS/phoB. (A) XD1 (phoATS/phoB) conidia were germinated in the media for the time and at the temperatures indicated before photography using differential interference contrast (DIC) illumination. Bar, 20 µm. (B) Strains with the indicated mutations were grown and photographed using DIC illumination or epifluorescence to visualize nuclei stained with DAPI. Bar, 5 µm.

Surprisingly, the terminal phenotype of the phoA^TS/phoB strains was modified on richer YG media at restrictive temperature, even though the cells were still markedly temperature sensitive. After 48 hr of germination in YG media, only 22% of phoA^TS/phoB conidia were able to send out a germ tube and, even though the rest were able to germinate and undergo some nuclear divisions, they were incapable of establishing polarized growth but instead expanded to form large, round-shaped cells (Fig 7A and Fig B). Those cells that were able to form a germ tube were abnormal compared to wild-type cells at 42° or the phoA^TS/phoB cells grown at 32°, typically being thicker (Fig 7B). The ability of the YG media to modify the phenotype of the phoA^TS/phoB strain was found to reside in the yeast extract component of this medium, which was active even when yeast extract was added to minimal media at a 1/50 dilution (Fig 7A) and even down to a 1/200 dilution (data not shown).

It has previously been shown that the polarity defect caused by the mpkA mitogen-activated protein kinase was remediated by growth on high-osmolarity media (BUSSINK and OSMANI 1999 ). However, unlike deletion of mpkA, the polarity defect caused by phoA^TS/phoB was not remediated by growth in 1 M sucrose-supplemented YG media (data not shown).

Expression of mammalian CDK5 can complement the lethality caused by phoA^TS/phoB:
The generation of the phoA^TS/phoB temperature-sensitive strain allowed us the opportunity to test for similarities in function between higher eukaryotic CDK5 and phoA/B function in A. nidulans by complementation. Expression of mammalian CDK5 was placed under control of the inducible alcA promoter and introduced into a phoA^TS/phoB strain. Representative transformants were replica plated onto alcA-repressing and -inducing media and incubated at restrictive temperature (Fig 8). Expression of CDK5 was found to suppress the ts^- caused by phoA^TS/phoB, most notably when expression of CDK5 was induced (Fig 8). The lower level of complementation on alcA-repressing media likely results from low level CDK5 expression expected under these conditions (SOM and KOLAPARTHI 1994 ). We conclude that there is a conserved function among mammalian CDK5 and A. nidulans phoA and phoB, which is essential in both mammals and A. nidulans although PHO85/pef1 is nonessential in yeast.

View larger version (75K): [in this window] [in a new window] [Download PPT slide]   Figure 8. Mammalian CDK5 can complement the phoATS/phoB mutations. Strains were replica plated and incubated for 2 days before photography. (A) alcA-repressing media at 32°. (B) alcA-inducing media at 32°. (C) alcA-repressing media at 42°. (D) alcA-inducing media at 42°. Strains: 1, wild type (SO51); 2, control ts- strain (SO53); 3 and 4, phoATS/phoB strain XD1; 5–8, XD1 transformed with mammalian Cdk5 under control of the inducible alcA promoter.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Previous work has identified two cyclin-dependent kinases in the filamentous fungus A. nidulans. The first, nimX^cdc2, is a functional homolog of the cell-cycle-specific Cdc2 kinase of fission yeast (OSMANI et al. 1994 ). Temperature-sensitive inactivation or deletion of nimX^cdc2 causes cell cycle arrest and lack of colony formation, demonstrating that it is an essential gene. The second CDK, PHOA, is not essential but plays a role in developmental fate in response to environmental conditions, including phosphorous concentrations. PHOA is most similar to the Pho85 class of CDKs, which includes PHO85 of S. cerevisiae and Pef1 of S. pombe. This particular class of CDKs performs a wide range of functions in different species, but neither of these yeast CDKs are essential and strains carrying null alleles are viable. Neither yeast genomes encode a second CDK in the Pho85 family.

Although A. nidulans phoA is a nonessential gene, we were surprised to find that expression of a dominant-negative version of phoA caused inhibition of growth. This indicated that phoA could potentially have a redundant essential function shared by a second unknown CDK. Previous Western blot analysis, using antiserum raised against PHOA, also indicated that a second PHOA-like kinase may be present in A. nidulans (BUSSINK and OSMANI 1998 ). This potential was confirmed using BLAST searches of the genome of A. nidulans for a gene encoding a phoA-related kinase. This kinase, with 77% identity to PHOA, was isolated and termed phoB.

To analyze the function of phoB, a null allele was generated using homologous recombination to replace the coding domain of phoB with the pyroA nutritional marker. Analysis of protein from deleted strains demonstrated that phoB does encode the second PHOA-related kinase previously seen by use of Western blotting (BUSSINK and OSMANI 1998 ). No phenotype could be attributed to the deletion of phoB under any growth conditions tested, including those that markedly affect development in a phoA-deleted strain. Thus, like phoA, phoB is a nonessential gene.

phoB apparently has no functions that cannot be fulfilled by phoA. However, because deletion of phoA causes marked developmental defects under specific growth conditions, and lack of phoB does not cause such defects, it is clear that phoA has some functions during development that cannot be fulfilled by phoB. These two kinases, however, do have some common functions because deletion of both caused lethality. This was revealed by crossing haploid strains having deletions of either phoA or phoB using complementing genetic markers. In this way we could ask if germination and growth of the phoA phoB mutant recombinant ascospores could occur on media selective for the two nutritional markers used to delete the genes. In the phoA phoB mutant ascospores, germination was able to take place but no colony formation was observed. This is because the double mutants had a limited capacity to undergo continued hyphal growth and nuclear division. These cells also displayed abnormal nuclear morphology and thinner germ tubes compared to controls at later times of germination. This indicates that neither phoA nor phoB is required for breaking ascospore dormancy or short-term growth, but they are required to undergo the first mitosis during germination, suggestive of a cell-cycle-specific defect. However, unlike mutations in genes that are specifically required for cell cycle progression (MORRIS 1976 ), the phoA phoB germlings did not continue normal growth while preventing nuclear divisions. Instead, germlings became thin and nuclear structure was compromised.

To further characterize the role of phoA/B, a temperature-sensitive allele of phoA was generated in a phoB background. This approach also allowed characterization of the phoA phoB null phenotype in rich media as we did not need to impose nutritional limitations on cells to identify the double mutant as dictated by crossing the phoA to phoB strains. The phenotype caused by temperature inactivation of phoA^TS/phoB in germinating conidia on minimal media was more dramatic than that seen in phoA phoB ascospores germinated in similar media. These differences could be attributed to differences in the germination properties of conidia vs. ascospores and/or to the potential for endowment of wild-type phoA function during ascospore formation. However, in both cases inactivation of phoA phoB function allows some features of germination (such as swelling of spores and germ-tube emergence for phoA phoB ascospores), but not cell cycle progression.

Surprisingly, the phenotype of phoA^TS/phoB was significantly affected by low levels of yeast extract (down to 1/200 of YG medium that contains 0.5% yeast extract). The main effect of the yeast extract was to allow germination and limited cell cycle progression. Importantly, the majority of conidia that were able to germinate were unable to switch to polarized growth but instead continued isotropic growth. This phenotype can also be caused by actin depolymerization using cytochalasin, suggesting that perhaps phoA phoB function may be required for normal actin function, as has been proposed for Pho85 (LEE et al. 1998 ) and CDC5 (SMITH and TSAI 2002 ).

The fact that very low levels of yeast extract can modify the terminal phenotype caused by lack of PHOA/B function further implicates these kinases as mediators of responses to extracellular conditions. At this time, however, it is unclear what particular component of yeast extract is responsible for modifying the phenotype.

The lethality caused by lack of both phoA and phoB demonstrates that in A. nidulans an essential function exists for this class of CDK, whereas in both budding and fission yeast neither PHO85 nor pef1 is essential. In S. cerevisiae six cyclin-dependent kinases are known. Of these, CDC28, KIN28, and BUR1 are essential genes whereas PHO85, CTK1, and SRB10 are nonessential. It is therefore possible that phoA phoB may fulfill functions carried out by one of the essential CDKs in S. cerevisiae (CDC28, KIN28, and BUR1). It is known that nimX and CDC28 are functional homologs. We therefore searched by BLAST analysis for potential CDKs in the genomes of both A. nidulans (http://www.genome.wi.mit.edu/annotation/fungi/aspergillus/index.html), which has not yet been annotated, and another fully sequenced filamentous fungus, N. crassa (GALAGAN et al. 2003 ; http://www.genome.wi.mit.edu/annotation/fungi/neurospora/). This analysis indicates that in A. nidulans and in N. crassa, there is a complement of six CDKs similar to that encoded in S. cerevisiae. It is therefore unlikely that the reason that the PHOA-PHOB pair is essential is due to lack of one of the essential CDKs found in S. cerevisiae. Interestingly, unlike A. nidulans, only one PHOA/B-like kinase is found in N. crassa and it will be interesting to determine if this is an essential gene.

The question therefore remains, Why are PHO85 in S. cerevisiae and pef1 in S. pombe nonessential genes whereas phoA and phoB together play an essential role in A. nidulans? In vertebrates the nearest CDK to PHOA/PHOB based upon primary amino acid sequence identity is CDK5. Interestingly, CDK5 is an essential gene in mice where homozygous nulls die in utero or soon after birth (OHSHIMA et al. 1996 ). This is due to lack of normal neuronal migration and subsequent failure of normal brain development. It has been recently hypothesized that perhaps CDK5 plays a role in intracellular trafficking (SMITH and TSAI 2002 ). Our results showing that a high percentage of phoA^TS/phoB conidia germinated at restrictive temperature are unable to establish polarized growth is consistent with phoA/phoB perhaps having a role in regulating intracellular trafficking. Further suggesting a similar role for phoA/phoB and CDK5 is the ability of CDK5 to complement the lethality caused by lack of phoA/phoB function.

	FOOTNOTES

Sequence data from this article have been deposited with the EMBL/GenBank Data Libraries under accession no. AY350425.

	ACKNOWLEDGMENTS

We thank Ralph Dean and colleagues for generating the A. nidulans BAC library, Brenda Andrews for mammalian CDK5, and members of the Osmani laboratory for their help and interest. This work was supported by a grant from the National Institutes of Health (GM-42564).

Manuscript received June 2, 2003; Accepted for publication August 6, 2003.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ANDREWS, B. and V. MEASDAY, 1998  The cyclin family of budding yeast: abundant use of a good idea. Trends Genet. 14:66-72.[Medline]

BUSSINK, H. J. and S. A. OSMANI, 1998  A cyclin-dependent kinase family member (PHOA) is required to link developmental fate to environmental conditions in Aspergillus nidulans.. EMBO J. 17:3990-4003.[Medline]

BUSSINK, H. J. and S. A. OSMANI, 1999  A mitogen-activated protein kinase (MPKA) is involved in polarized growth in the filamentous fungus, Aspergillus nidulans.. FEMS Microbiol. Lett. 173:117-125.[Medline]

CARROLL, A. S. and E. K. O'SHEA, 2002  Pho85 and signaling environmental conditions. Trends Biochem. Sci. 27:87-93.[Medline]

CHAMPE, S. P., and L. D. SIMON, 1992 Cellular differentiation and tissue formation in the fungus Aspergillus nidulans, pp. 63–91 in Morphogenesis: An Analysis of the Development of Biological Structure, edited by E. F. ROSSOMANDO and S. ALEXANDER. Marcel Dekker, New York.

DE SOUZA, C. P., A. H. OSMANI, L. P. WU, J. L. SPOTTS, and S. A. OSMANI, 2000  Mitotic histone H3 phosphorylation by the NIMA kinase in Aspergillus nidulans.. Cell 102:293-302.[Medline]

GALAGAN, J. E., S. E. CALVO, K. A. BORKOVICH, E. U. SELKER, and N. D. READ et al., 2003  The genome sequence of the filamentous fungus Neurospora crassa.. Nature 422:859-868.[Medline]

KANG, S. and R. L. METZENBERG, 1993  Insertional mutagenesis in Neurospora crassa: cloning and molecular analysis of the preg+ gene controlling the activity of the transcriptional activator NUC-1. Genetics 133:193-202.[Abstract]

LEE, E. C., D. YU, J. MARTINEZ DE VELASCO, L. TESSAROLLO, and D. A. SWING et al., 2001  A highly efficient Escherichia coli-based chromosome engineering system adapted for recombinogenic targeting and subcloning of BAC DNA. Genomics 73:56-65.[Medline]

LEE, J., K. COLWILL, V. ANELIUNAS, C. TENNYSON, and L. MOORE et al., 1998  Interaction of yeast Rvs167 and Pho85 cyclin-dependent kinase complexes may link the cell cycle to the actin cytoskeleton. Curr. Biol. 8:1310-1321.[Medline]

MEASDAY, V. K., L. MOORE, R. RETNAKARAN, J. LEE, and M. DONOVIEL et al., 1997  A family of cyclin-like proteins that interact with the Pho85 cyclin-dependent kinase. Mol. Cell. Biol. 17:1212-1223.[Abstract]

MORRIS, N. R., 1976  Mitotic mutants of Aspergillus nidulans.. Genet. Res. 26:237-254.

NISHIZAWA, M., M. TANABE, N. YABUKI, K. KITADA, and E. TOH, 2001  Pho85 kinase, a yeast cyclin-dependent kinase, regulates the expression of UGP1 encoding UDP-glucose pyrophosphorylase. Yeast 18:239-249.[Medline]

OHSHIMA, T., J. M. WARD, C. G. HUH, G. LONGENECKER, and G. LONGENECKERVEERANNA et al., 1996  Targeted disruption of the cyclin-dependent kinase 5 gene results in abnormal corticogenesis, neuronal pathology and perinatal death. Proc. Natl. Acad. Sci. USA 93:11173-11178.[Abstract/Free Full Text]

OSMANI, A. H., N. VAN PEIJ, M. MISCHKE, M. J. O'CONNELL, and S. A. OSMANI, 1994  A single p34^cdc2 protein kinase (nimX^cdc2) is required at G1 and G2 in Aspergillus nidulans.. J. Cell Sci. 107:1519-1528.[Abstract]

OSMANI, A. H., G. S. MAY, and S. A. OSMANI, 1999  The extremely conserved pyroA gene of Aspergillus nidulans is required for pyridoxine synthesis and is required indirectly for resistance to photosensitizers. J. Biol. Chem. 274:23565-23569.[Abstract/Free Full Text]

OSMANI, S. A., G. S. MAY, and N. R. MORRIS, 1987  Regulation of the mRNA levels of nimA, a gene required for the G2-M transition in Aspergillus nidulans.. J. Cell Biol. 104:1495-1504.[Abstract/Free Full Text]

PELEG, Y., R. ADDISON, R. AMAMAYO, and R. L. METZENBERG, 1996a  Translocation of Neurospora crassa transcription factor NUC-1 into the nucleus is induced by phosphorous limitation. Fungal Genet. Biol. 20:186-191.

PELEG, Y., R. ARAMAYO, S. KANG, J. G. HALL, and R. L. METZENBERG, 1996b  NUC-2, a component of the phosphate-regulated signal transduction pathway in Neurospora crassa, is an ankyrin repeat protein. Mol. Gen. Genet. 252:709-716.[Medline]

PONTECORVO, G., 1953 The genetics of Aspergillus nidulans, pp. 141–238 in Advances in Genetics, edited by M. DEMEREC. Academic Press, New York.

SMITH, D. S. and L. H. TSAI, 2002  Cdk5 behind the wheel: A role in trafficking and transport? Trends Cell Biol. 12:28-36.[Medline]

SOM, T. and V. KOLAPARTHI, 1994  Developmental decisions in Aspergillus nidulans are modulated by ras activity. Mol. Cell. Biol. 14:5333-5348.[Abstract/Free Full Text]

SWAMINATHAN, S., H. M. ELLIS, L. S. WATERS, D. YU, and E. C. LEE et al., 2001  Rapid engineering of bacterial artificial chromosomes using oligonucleotides. Genesis 29:14-21.[Medline]

TANAKA, K. and H. OKAYAMA, 2000  A pcl-like cyclin activates the Res2p-Cdc10p cell cycle "start" transcriptional factor complex in fission yeast. Mol. Biol. Cell 11:2845-2862.[Abstract/Free Full Text]

TOH-E, A., K. TANAKA, Y. UESONO, and R. B. WICKNER, 1988  PHO85, a negative regulator of the PHO system, is a homolog of the protein kinase gene, CDC28, of Saccharomyces cerevisiae. Mol. Gen. Genet. 214:162-164.[Medline]

UESONO, Y., K. TANAKA, and A. TOH-E, 1987  Negative regulators of the PHO system in Saccharomyces cerevisiae: isolation and structural characterization of PH085. Nucleic Acids Res. 15:10299-10309.[Abstract/Free Full Text]

This article has been cited by other articles:

				S. Castillo-Lluva, I. Alvarez-Tabares, I. Weber, G. Steinberg, and J. Perez-Martin Sustained cell polarity and virulence in the phytopathogenic fungus Ustilago maydis depends on an essential cyclin-dependent kinase from the Cdk5/Pho85 family J. Cell Sci., May 1, 2007; 120(9): 1584 - 1595. [Abstract] [Full Text] [PDF]

				K. Bomeke, R. Pries, V. Korte, E. Scholz, B. Herzog, F. Schulze, and G. H. Braus Yeast Gcn4p Stabilization Is Initiated by the Dissociation of the Nuclear Pho85p/Pcl5p Complex Mol. Biol. Cell, July 1, 2006; 17(7): 2952 - 2962. [Abstract] [Full Text] [PDF]

				D. Wu, X. Dou, S. B. Hashmi, and S. A. Osmani The Pho80-like Cyclin of Aspergillus nidulans Regulates Development Independently of Its Role in Phosphate Acquisition J. Biol. Chem., September 3, 2004; 279(36): 37693 - 37703. [Abstract] [Full Text] [PDF]

				M. R. V. Z. K. Fagundes, J. F. Lima, M. Savoldi, I. Malavazi, R. E. Larson, M. H. S. Goldman, and G. H. Goldman The Aspergillus nidulans npkA Gene Encodes a Cdc2-Related Kinase That Genetically Interacts With the UvsBATR Kinase Genetics, August 1, 2004; 167(4): 1629 - 1641. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Dou, X. Articles by Osmani, S. A. Search for Related Content PubMed PubMed Citation Articles by Dou, X. Articles by Osmani, S. A.