the rfx protein rfxa is an essential ... - eukaryotic cell · multicellular hyphal growth at 25°c...

EUKARYOTIC CELL, Apr. 2010, p. 578–591 Vol. 9, No. 41535-9778/10/$12.00 doi:10.1128/EC.00226-09Copyright © 2010, American Society for Microbiology. All Rights Reserved.

The RFX Protein RfxA Is an Essential Regulator of Growthand Morphogenesis in Penicillium marneffei�

Hayley E. Bugeja, Michael J. Hynes, and Alex Andrianopoulos*Department of Genetics, University of Melbourne, Victoria 3010, Australia

Received 6 August 2009/Accepted 23 January 2010

Fungi are small eukaryotes capable of undergoing multiple complex developmental programs. The oppor-tunistic human pathogen Penicillium marneffei is a dimorphic fungus, displaying vegetative (proliferative)multicellular hyphal growth at 25°C and unicellular yeast growth at 37°C. P. marneffei also undergoes asexualdevelopment into differentiated multicellular conidiophores bearing uninucleate spores. These morphogeneticprocesses require regulated changes in cell polarity establishment, cell cycle dynamics, and nuclear migration.The RFX (regulatory factor X) proteins are a family of transcriptional regulators in eukaryotes. We sought todetermine how the sole P. marneffei RFX protein, RfxA, contributes to the regulation of morphogenesis.Attempts to generate a haploid rfxA deletion strain were unsuccessful, but we did isolate an rfxA�/rfxA�heterozygous diploid strain. The role of RfxA was assessed using conditional overexpression, RNA interference(RNAi), and the production of dominant interfering alleles. Reduced RfxA function resulted in defectivemitoses during growth at 25°C and 37°C. This was also observed for the heterozygous diploid strain duringgrowth at 37°C. In contrast, overexpression of rfxA caused growth arrest during conidial germination. The datashow that rfxA must be precisely regulated for appropriate nuclear division and to maintain genome integrity.Perturbations in rfxA expression also caused defects in cellular proliferation and differentiation. The datasuggest a role for RfxA in linking cellular division with morphogenesis, particularly during conidiation andyeast growth, where the uninucleate state of these cell types necessitates coupling of nuclear and cellulardivision tighter than that observed during multinucleate hyphal growth.

We are interested in examining the regulatory networkscontrolling cell-type specification and development in theopportunistic fungal pathogen Penicillium marneffei and therole these processes play in pathogenicity. P. marneffei is athermally dimorphic fungus capable of causing disseminatedinfection in immunocompromised individuals. Dimorphismis a common morphological process for many fungal patho-gens and has been clearly linked to pathogenicity. At roomtemperature (25°C), P. marneffei exhibits mycelial growth inwhich multinucleate cells are connected in long hyphal fil-aments. This growth form is also capable of asexual devel-opment (conidiation), in which single-celled uninucleatespores (conidia) are produced on specialized aerial hyphae(conidiophores). The transition from a multicellular hyphalgrowth form to a unicellular growth form occurs upon trans-fer to 37°C. During this process, known as arthroconidia-tion, cellular and nuclear division become coupled and dou-ble septa are deposited between cells. The subsequentfragmentation of these filaments leads to the production ofuninucleate yeast cells that divide by fission. It is this growthform of P. marneffei that presents as an intracellular patho-gen in phagocytic cells during infection (10, 25).

Transition between multicellular and unicellular morpholog-ical states is common to most fungi and serve as an importantprocess within developmental programs such as conidiationand mating. The process of arthroconidiation in P. marneffei isanalogous to the transition from a hyphal growth form to a

unicellular spore form in Acremonium chrysogenum and Coc-cidioides immitis, suggesting there may be common mecha-nisms underlying these events. The cpcR1 gene of A. chrysoge-num, initially identified as a regulator of cephalosporin Cbiosynthesis genes, was subsequently shown to regulate arthro-sporulation, whereby the filamentous mycelium undergoesfragmentation into unicellular arthrospores (31, 56, 58).

CPCR1 is a member of the regulatory factor X (RFX)family of transcriptional regulators that have been impli-cated in the regulation of both developmental and cell cycleevents. To date, 17 members of this protein family that arehighly conserved from yeast to humans have been isolated(21). The defining feature of these proteins is the novel RFXDNA-binding domain, a member of the winged-helix sub-family of helix-turn-helix DNA-binding domains (24). Inaddition, most of the RFX proteins contain a highly con-served dimerization domain mediating the formation of ho-mo- and/or heterodimers (21, 52). While specific roles havebeen assigned to individual RFX proteins, a unified under-standing of the processes regulated by these proteins acrossspecies has not been forthcoming.

Variation in tissue- and cell-type-specific expression of thefive (RFX1 to -5) RFX genes identified in mammals has beenobserved (33, 53). The prototypical RFX protein, RFX1, isubiquitously expressed and can act as both an activator and arepressor (34). Potential targets of RFX1 include cell prolif-eration and DNA damage genes, such as c-myc and the PCNA(proliferating cell nuclear antigen), MAP1A (microtubule-as-sociated protein), IL-5R� (interleukin-5 receptor �), and RNR(ribonucleotide reductase) genes (33, 36, 42, 59, 72). Addition-ally, the role of RFX5 in the regulation of major histocompat-

* Corresponding author. Mailing address: Department of Genetics,University of Melbourne, Victoria 3010, Australia. Phone: 61 3 83445164. Fax: 61 3 8344 5139. E-mail: [email protected].

� Published ahead of print on 29 January 2010.

578

on August 29, 2020 by guest

http://ec.asm.org/

Dow

nloaded from

https://crossmark.crossref.org/dialog/?doi=10.1128/EC.00226-09&domain=pdf&date_stamp=2010-04-01

http://ec.asm.org/

ibility complex (MHC) class II gene expression is well estab-lished (60).

The Caenorhabditis elegans and Drosophila melanogasterRFX proteins, DAF-19 and dRFX, respectively, have recentlybeen assigned roles in the development of ciliated sensoryneurons (17, 61). D. melanogaster dRFX2 appears to be in-volved in the regulation of cell cycle progression, a theme alsoevident to various extents in its fungal counterparts (47). InSaccharomyces cerevisiae, the RFX homologue Crt1 preventsthe expression of the DNA damage-inducible RNR genes inthe absence of DNA damage through recruitment of the Tup1-Ssn6 corepressor complex (32). The sak1� protein of Schizo-saccharomyces pombe has been shown to function downstreamof protein kinase A (PKA), where it promotes mitotic exit andthereby allows the onset of sexual development or entry intostationary phase (67). Deletion of sak1� in S. pombe results inlethality, with transient phenotypes indicative of severe mitoticdefects. More recently, Rfx2 of the dimorphic pathogen Can-dida albicans was found to regulate not only elements of theDNA damage response (DDR), presumably by a mechanismsimilar to that of Crt1 in S. cerevisiae, but also morphogenesisand virulence (27).

Here, the role of the RFX protein RfxA was investigatedduring the growth and morphogenesis of P. marneffei. The rfxAgene appears to be essential for the viability of P. marneffei,and studies involving overexpression, RNA interference(RNAi), and the production of dominant interfering alleleshave shown that the levels of functional RfxA must be preciselymaintained for growth and morphogenesis, suggesting thatRfxA participates in the regulation of cell division events.As such, RfxA may be required for linking cell cycle regu-lation with cellular differentiation during morphogenesis inP. marneffei.

MATERIALS AND METHODS

Molecular techniques. Plasmid DNA was isolated using the Wizard Plus SVDNA Purification System (Promega). Genomic DNA was prepared from frozenmycelia of P. marneffei as previously described (7). For the extraction of RNA,fungal cultures were grown as previously described (6). RNA was extracted from0.1 to 0.2 g of biomass using the FastRNA Pro Red kit (Bio101). Southern blotswere prepared with Hybond N� membranes (Amersham) using standard pro-cedures (55). For screening of the P. marneffei genomic DNA lambda library,plaque lifts and the isolation of positive clones were performed according to theinstructions for the �BlueSTAR vector system kit (Novagen). Hybridizationswere performed with [�-32P]dATP-labeled DNA probes using standard methods(55). The oligonucleotides used for PCR are listed in Table 1. Reverse trans-criptase (RT)-PCR was performed using the rfxA-specific primers L13 and Q50,the benA-specific primers F58 and F59, and the mobA-specific primers HH21 andHH22 on 100 ng of total RNA using the Superscript one-step RT-PCR kit withPlatinum Taq (Invitrogen) according to the manufacturer’s directions. The num-ber of amplification cycles was optimized for each primer pair to ensure thatproduct synthesis was in the exponential phase of amplification. Product yieldswere estimated from ethidium bromide-stained gel images using MacBAS ver2.1(Fuji PhotoFilm Co. and Kohshin Graphic Systems Inc.). PCR screening ofputative rfxA deletion transformants was performed using three primers: L31,specific for the wild-type rfxA locus; L29, a pyrG�-specific primer; and L30, anrfxA genomic-locus-specific primer. A 1.5-kb product generated using the prim-ers L31 and L30 was expected for strains containing the wild-type rfxA locus,while the presence of the rfxA�::pyrG� deletion locus would give rise to a 1.9-kbproduct using the primers L29 and L30. For quantitative real-time RT-PCR, 2 �gof total RNA was subjected to DNase treatment using RQ1 RNase-free DNase(Promega) prior to cDNA synthesis, performed with the Reverse TranscriptionSystem (Promega) according to the manufacturer’s specifications. Real-timePCR was performed in a Rotor-Gene RG-3000 instrument (Corbett Research)on �20 ng of cDNA using the SensiMix Plus SYBR detection kit (Quantace)

with an initial denaturation for 10 min at 95°C, followed by 45 cycles using thefollowing parameters: 95°C for 20 s, 60°C for 20 s, and 72°C for 15 s. The relativefold expression of the sldA (primers FF20 and FF21) and bimB (primers FF18and FF19) transcripts was determined using the comparative threshold cycle(CT) method (37), where samples were normalized to the actin (primers GG2and GG3) transcript abundance to adjust for variations in sample-loadingamounts.

Cloning and plasmid construction. A 650-bp fragment of rfxA was isolatedfrom degenerate PCR performed on genomic DNA of the wild-type strain 2161using the primers J72 and J73, designed to amplify the region encoding theputative DNA-binding domain, which is highly conserved in RFX proteins fromthe filamentous fungi A. chrysogenum, Neurospora crassa, Aspergillus fumigatus,and Penicillium chrysogenum. This PCR fragment was used to probe a P. marnef-fei genomic DNA � library at high stringency (50% formamide, 0.1� SSC [1�SSC is 0.15 M NaCl plus 0.015 M sodium citrate], 65°C) to clone the full-lengthrfxA gene, contained within the plasmid pHS5584. A 5.9-kb KpnI/SacII subcloneof P. marneffei rfxA (pHS6520) was generated by combining a 1.4-kb KpnI/SpeIfragment containing the 5� noncoding region of rfxA with a 4.5-kb SpeI/SacIIfragment containing the entire rfxA coding region and an additional 3� noncodingregion, both derived from pHS5584.

A two-step cloning strategy was used to generate the rfxA gene deletionconstruct pHS5595. First, a 1.1-kb XhoI/EcoICRI fragment of pHS5584, con-taining the 5� region of rfxA, including the first 149 bp of coding sequence, wasinserted between the XhoI/EcoRV sites upstream of the Aspergillus nidulanspyrG blaster cassette in pAB4626. Subsequently, a 1.4-kb SmaI/SacII fragment ofpHS5584, containing the region 3� of rfxA, including the last 132 bp of codingsequence, was inserted downstream of the A. nidulans pyrG blaster cassette.

To generate the xylPp::rfxA-RNAi construct pHS6521, the 5� coding region ofrfxA (bp 1 to 1256) was amplified by PCR using the primers Q34 and L6 andblunt cloned into the SmaI site of pBluescript II SK(�) (pHS6098). Subse-quently, a 1.3-kb ClaI/SmaI and a 1.3-kb EcoRV/BamHI fragment of pHS6098were inserted on either side of green fluorescent protein (GFP) in the vectorpAA4111 in opposite orientations. A 3.4-kb NcoI fragment containing GFPflanked by the inverted repeats of the rfxA 5� coding sequence was then inserteddownstream of the xylP promoter in the plasmid pHS6103, containing a 2.6-kbEcoRI fragment of areA for targeting to the areA locus (H. Bugeja, M. J. Hynes,and A. Andrianopoulos, unpublished data).

To create a xylPp::rfxA construct, a 2.7-kb PCR product containing rfxA wasgenerated using the primers L32 and L33 and blunt cloned into the SmaI site ofpBluescript II SK(�). An NcoI fragment was isolated by partial digestion andcloned downstream of the xylP promoter in pXylNOM (70) (pHS5705). Sincethis construct lacked the correct ATG for rfxA, plasmid pHS5705 was modifiedusing partial digestion with NcoI to remove 0.7 kb of 5� rfxA coding sequence. A

TABLE 1. Oligonucleotides used in this study

Name Sequence

F58........................................5�-GCTCCGGTGTCTACAATGGC-3�F59........................................5�-AGTTGTTACCAGCACCGGAC-3�J72.........................................5�-GGMGARTCMAARTAYCAYTA-3�J73.........................................5�-RRTCRCAYTCCTYKATCCA-3�L13........................................5�-AAGGTTTGGGTGAATAAGCAG-3�L29........................................5�-GGACTTTGAGTGTGAGTGGAA-3�L30........................................5�-ATTCTGTCCCGTAGATGAAGA-3�L31........................................5�-GCAAACCCAGGAAATGACAC-3�L31........................................5�-GCAAACCCAGGAAATGACAC-3�L32........................................5�-ATGGAGCGCCTGCTAATCCG-3�L33........................................5�-ATGGCTAATACCACCCAAGA-3�L6..........................................5�-CGACTGGTGGTTGAGATATGC-3�Q34 .......................................5�-ATCGATCCATGGCTCCTGAAG-3�Q36 .......................................5�-CATCTGCTTGAGCGTGTAAC-3�Q37 .......................................5�-CAGGATCAAAATACGGCGAAC-3�Q50 .......................................5�-GTCTTTGCGACCTCGCAGTAT-3�FF18 .....................................5�-GGTAAATGCCAATGAACTCG-3�FF19 .....................................5�-CAAGAGACTCACGAGTAGCC-3�FF20 .....................................5�-ATGGACTCAAGCAGAGGAGG-3�FF21 .....................................5�-GCACACTTGATTGTGACTCA-3�GG2......................................5�-CTTCCAGCCTTCCGTCATC-3�GG3......................................5�-GCATACGGTCGGAGATACCA-3�HH21....................................5�-TGGTTGGCTGTTAATGTGGTC-3�HH22....................................5�-AACTGGTGTTGAGGTGTGGCT-3�

VOL. 9, 2010 REGULATION OF GROWTH AND MORPHOGENESIS BY RfxA 579


http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

second partial NcoI digestion and end fill reaction was performed to destroy the3� NcoI site (pHS6097). A 0.8-kb NcoI fragment from pHS6098 was inserted intothe 5� NcoI site of pHS6097 to generate a construct in which the entire rfxAcoding sequence was downstream of the xylP promoter (pHS6099). In order totarget this construct to the areA locus, a 4.5-kb XhoI/SphI fragment of pHS6099,containing part of xylPp along with the rfxA coding sequence and trpCt, was usedto replace an equivalent 1.7-kb fragment of pHS6103, giving rise to pHS6529.

Inverse PCR using the xylPp::rfxA construct (pHS6099) as a template and theprimers Q36 and Q37 was used to delete the region encoding the conservedDNA-binding domain (721 to 1099). The rfxA-DIM� construct pHS6530 wasgenerated by inverse PCR using the xylPp::rfxA construct (pHS6099) templateand the primers L31 and L13, thereby removing the region encoding the putativedimerization domain (1673 to 2566). To facilitate targeted integration of theseconstructs at the areA locus in the areA strain 41.2.14-3, the constructs containeda 2.6-kb EcoRI fragment of areA.

Fungal strains and media. The P. marneffei strains used in this study are listedin Table 2. The isolation and transformation of P. marneffei protoplasts wereperformed as previously described (7). For selection of pyrG� transformants ofstrain SPM4, protoplasts were regenerated on osmotically stabilized protoplastmedium (PM) containing 1.2 M sucrose and 10 mM ammonium tartrate[(NH4)2T]. For niaD� selection, 10 mM sodium nitrate (NaNO3) was used as asole nitrogen source, whereas 10 mM sodium nitrite (NaNO2) was used as anitrogen source for selection of areA� transformants from the areA strain41.2.14-3. The strains were grown on 1% glucose minimal medium (ANM) with10 mM -aminobutyric acid (GABA) (14), yeast synthetic dextrose medium (SD)with 10 mM ammonium sulfate [(NH4)2SO4] (5), or brain heart infusion (BHI)medium (Oxoid). When required, the medium was supplemented with 10 mMuracil to allow the growth of pyrG strains. For induction of the xylP promoter,0.5% xylose and 0.5% sucrose were used in place of 1% glucose (noninduced).The DNA replication inhibitor hydroxyurea (HU) was used at a final concen-tration of either 2 mM, 5 mM, or 10 mM. To assess the nuclear division arrest,spores were germinated on slides coated with solid medium under inducingconditions for 18 h in the presence or absence of 2 mM HU before beingprocessed for microscopic analysis (see below). Nuclear counts (approximately150 germlings) were determined (n � 2).

Microscopy. Strains were grown on slides coated with a thin layer of solidmedium with one end submerged in liquid medium, as described previously (6).The slides were fixed by soaking them in a solution of 4% para-n-formaldehydein PME {PIPES [50 mM piperazine-N,N�-bis(2-ethanesulfonic acid)], pH 6.7, 1mM MgSO4, 20 mM EGTA} for 30 min, followed by two 5-min PME washes.Samples were stained using fluorescent brightener 28 (calcofluor white [CAL]),Hoechst 33258, or 4�6�-diamino-2-phenylindole (DAPI) and visualized using aReichart Jung Polyvar II microscope with either differential interference contrast(DIC) or epifluorescence optics. Images were captured using a SPOT charge-coupled device (CCD) camera (Diagnostic Instruments) and processed usingAdobe Photoshop software.

Sequencing and bioinformatics. DNA sequencing was performed at the Aus-tralian Genome Research Facility (AGRF) on purified plasmid DNA. DNAsequence was analyzed using Sequencher 3.1.1 (Gene Codes). All sequenceanalyses, including database searches, were done using the Australian NationalGenomic Information Service (ANGIS). Pairwise sequence comparisons wereperformed using the GAP program available through ANGIS, and multiple-sequence alignments were generated using ClustalW (62) and MacBoxShade. Insome instances, sequence data were obtained from the following fungal genomedatabases: A. nidulans (http://www.broad.mit.edu/annotation/fungi/aspergillus/),Neurospora crassa (http://www.broad.mit.edu/annotation/fungi/neurospora/), As-

pergillus fumigatus (http://www.tigr.org/tdb/e2k1/afu1/), and S. cerevisiae (http://www.yeastgenome.org/).

Upstream sequences (1,000 bp) of annotated genes from A. nidulans, A. fumigatus,Aspergillus terreus, and Aspergillus oryzae were downloaded (http://www.broad.mit.edu/annotation/genome/aspergillus_group/Downloads.htm) and searched for theputative RfxA recognition sequence RTHNYYN0-3RGNAAC using the DNApattern search available at RSAT (http://rsat.ulb.ac.be/rsat/) (65). Common genesets from Aspergillus containing putative RfxA binding sites were then identifiedfrom protein clusters available from the Aspergillus comparative database (http://www.tigr.org/sybil/asp/index.html) based on preliminary sequence data ob-tained from The Institute for Genomic Research website (http://www.tigr.org).Putative RfxA target genes were functionally assigned based on the Gene On-tology Consortium (GO) classification and description available at http://www.geneontology.org (4).

Nucleotide sequence accession number. The sequence of the 5.8-kb KpnI/SacII fragment of P. marneffei rfxA (pHS6520) sequence was deposited in Gen-Bank under accession number DQ666366.

RESULTS

Isolation of P. marneffei rfxA containing the highly conservedRFX DNA-binding domain. A degenerate PCR approach wasused to isolate a region of P. marneffei rfxA encoding theconserved DNA-binding domain, and this was subsequentlyused to clone the full-length rfxA gene from a P. marneffeigenomic DNA � library. A 5.8-kb KpnI/SacII subclone of P.marneffei rfxA (pHS6520) was sequenced, and the predictedrfxA gene encoded a protein of 861 amino acids. Orthologuesof rfxA were detected in the genome sequences of many fila-mentous fungi, and the P. marneffei RfxA protein containsmany sequence features previously characterized in other eu-karyotic RFX factors (21) (Fig. 1A). Most importantly, P.marneffei RfxA contains the RFX DNA-binding domain, whichis highly conserved among all family members (Fig. 1B), inaddition to a putative dimerization domain (Fig. 1C). It alsocontains regions common to many transcriptional activators,such as a glutamine (Q)-rich region (RfxA residues 70 to 117;21% Q) and a highly acidic aspartate (D)- and glutamate(E)-rich region (RfxA residues 782 to 856; 24.3% D/E). Ho-mology regions B and C, which have not been functionallycharacterized in RFX proteins, are also present within P.marneffei RfxA.

The expression of rfxA is upregulated during conidiationand yeast growth. To examine rfxA expression during thegrowth and morphogenesis of P. marneffei, semiquantitativeRT-PCR was performed using RNA isolated from vegetativehyphal cells, cells undergoing asexual development (conidia-tion) at 25°C, and yeast cells cultured at 37°C. The abundanceof rfxA transcript was found to be relatively low during vege-

TABLE 2. P. marneffei strains used in this study

Strain Genotype Source or reference

2161 Wild type J. Pitt (CSIRO Food industries), ATCC 18224SPM4 niaD1 pyrG1 733.4.2 niaD1 pyrG1 �AnpyrG S. Zuber and A. Andrianopoulos, unpublished data41.2.14-3 niaD1 pyrG1 �areA H. E. Bugeja, M. J. Hynes, and A.

Andrianopoulos, unpublished data44.1.2 niaD1 pyrG1 rfxA/�rfxA::AnpyrG This study49.3.1 niaD1 pyrG1 �areA areA�::xylPp::rfxA::trpCt This study49.4.1 niaD1 pyrG1 �areA areA�::xylPp::rfxA�1673-2566::trpCt This study55.2.1 niaD1 pyrG1 �areA areA�::xylPp::rfxA1-1256::GFP::rfxA1256-1::trpCt This study56.4.4 niaD1 pyrG1 �areA areA�::xylPp::rfxA�721-1099::trpCt This study

580 BUGEJA ET AL. EUKARYOT. CELL


http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

tative hyphal growth at 25°C and was approximately 2-fold and5-fold higher during yeast growth at 37°C and conidiation at25°C, respectively (Fig. 2). Consistent with the increased ex-pression of rfxA during conidiation, potential binding sites

were identified for the BrlA (5), StuA (1), and AbaA (1)transcriptional regulators within the rfxA promoter region (ref-erences 3, 11, and 18 and data not shown). These proteins areinvolved in mediating the correct temporal and spatial expres-sion of genes involved in conidiation in P. marneffei (6, 8; A. R.Borneman, K. Tan, M. J. Hynes, and A. Andrianopoulos, un-published data).

Potential RfxA target genes include cell cycle regulators. Tounderstand the cellular process(es) that RfxA may regulate, weused a bioinformatics approach to identify putative RfxA tar-get genes in the sequenced and annotated genomes of the P.marneffei-related fungi A. fumigatus, A. nidulans, A. oryzae, andA. terreus. These fungi are sufficiently diverged to reveal con-served functional regulatory elements, such as RFX bindingsites. This unbiased approach took advantage of the highlyconserved DNA-binding domain across all the RFX proteinsand used the consensus RFX binding sequence RTHNYYN0-3

RGNAAC to search the 5� untranslated regions (UTRs) of allgenes in these fungi. Approximately 2,500 to 3,300 genes thatcontained at a least one potential RFX binding site were iden-tified in each species. The presence of a single site has beenfound to be sufficient to confer regulation in other systems (19,71). Given that a large proportion of these may representchance occurrences of the consensus target sequence, the dataset was restricted to include only those sites that were presentin the promoter sequences of orthologous genes in all fourspecies. This resulted in the identification of 75 orthologous

FIG. 1. P. marneffei RfxA contains protein motifs characteristic of RFX proteins. (A) Protein structure map of RfxA. The regions representedinclude the glutamine-rich region (Q; residues 70 to 117), the RFX DNA-binding domain (DBD; residues 243 to 317), the highly conserved B(residues 449 to 484) and C (residues 511 to 550) regions, the putative dimerization domain (DIM; residues 567 to 741), and the acidic aspartate-and glutamate-rich region (DE; 782 to 856). aa, amino acids. (B) Alignment of the regions containing the DNA-binding domain from allcharacterized eukaryotic RFX proteins. The protein sequences included in the alignment are Homo sapiens RFX1 (HsRFX1; P22670; 429 to 515),Mus musculus RFX1 (MmRFX1; P48377; 414 to 500), C. elegans DAF-19 (CeDAF19; Q09555; 250 to 335), D. melanogaster RFX (DmRFX;Q9U1K2; 362 to 448), D. melanogaster RFX2 (DmRFX2; Q65YQ9; 379 to 465), A. chrysogenum CPCR1 (AcCPCR1; Q9P8F6; 214 to 303), S.pombe Sak1 (SpSak1; P48383; 91 to 178), S. cerevisiae Crt1 (ScCrt1; P48743; 275 to 362), and P. marneffei RfxA (PmRfxA; this study; ABG56532;233 to 319). (C) Alignment of the dimerization domains from RFX proteins. The protein sequences included in the alignment are H. sapiens RFX1(HsRFX1; P22670; 743 to 909), M. musculus RFX1 (MmRFX1; P48377; 727 to 893), C. elegans DAF-19 (CeDAF19; Q09555; 605 to 769), D.melanogaster RFX (DmRFX; Q9U1K2; 686 to 851), A. chrysogenum CPCR1 (AcCPCR1; Q9P8F6; 547 to 704), S. pombe Sak1 (SpSak1; P48383;507 to 674), and P. marneffei RfxA (PmRfxA; this study; ABG56532; 567 to 741).

FIG. 2. Expression of rfxA is upregulated during asexual develop-ment. Semiquantitative RT-PCR was performed on total RNA ex-tracted during vegetative hyphal growth at 25°C (25°C V), conidiationat 25°C (25°C C), and yeast growth at 37°C (37°C), using primersspecific for rfxA (26 cycles) or benA (22 cycles), encoding �-tubulin, asan RNA abundance standard. The standardized amounts of rfxA prod-uct (�SEM) relative to benA product for the three RNA samples were0.35 (0.05), 1.7 (0.33), and 0.59 (0.17), respectively.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

genes containing at least one putative RFX binding site. In sixinstances, the identified binding sites were located in overlap-ping divergent promoters present in all four species. In thesecases, it is possible that only one member of the gene pair mayrepresent an RfxA target gene.

Of these 75 genes, no predicted cellular function could befound for 29 (39%) based on either GO annotation or conser-vation with previously characterized genes. Of the remaining46 genes, 19 (41%) represent genes involved in cell cycle reg-ulation, in particular, mitotic-spindle dynamics and exit frommitosis, and the rest (27 genes, or 59%) are divided among avariety of cellular processes, including cellular metabolism,protein synthesis and degradation, chromosome metabolism,and cytoskeleton dynamics. In many of the genes examined,the positions of the putative RFX binding sites relative to theinitiator codon were also found to be conserved across all fourspecies and in the P. marneffei orthologues (Table 3).

In an attempt to understand the regulatory role of RfxA,preliminary expression analysis of nine putative rfxA targetgenes representing a range of cellular processes related to celldivision (Table 3) was performed using either semiquantitativeor real-time RT-PCR. The expression of these genes in hyphal(25°C) and/or yeast (37°C) cells with either reduced or in-creased rfxA expression was examined, using rfxA-RNAi andrfxA OE strains, respectively (Fig. 3) (see below). Of these, sixgenes orthologous to S. pombe cdc4, cdc15, and src1 and A.nidulans kipB, nimA, mpsA, and sldA showed no consistentchange in expression level when RfxA was either over- orunderexpressed, indicating that they are not targets of RfxAunder the conditions examined. The remaining genes areknown to function in spindle dynamics, chromosome cohesion,and mitotic exit. P. marneffei mobA, an orthologue of the S.cerevisiae MOB1 gene encoding a protein kinase required formaintenance of ploidy, showed increased expression in the

TABLE 3. Conserved genes from filamentous fungi containing RFX consensus binding sites

Organism Gene Position of binding sitea Position in P. marneffei homologa Putative function

A. fumigatus AFUA_7G05950 609 () 542 () Cytokinesis EF hand protein (Cdc4)A. nidulans AN6732.3 388 (), 534 ()A. oryzae AO090005000459 577 ()A. terreus ATEG_063406 483 ()

A. fumigatus AFUA_3G10250 348 (), 967 () 318 () Similarity to protein kinase (Cdc15)regulating the mitotic exit networkA. nidulans AN4963.3 291 (�/)

A. oryzae AO090003000564 374 (�/)A. terreus ATEG_04556 338 (�/)

A. fumigatus AFUA_2G03150 326 (�/) 305 () Related to kinesin protein (KipB) involved inmitotic-spindle dynamicsA. nidulans AN4513.3 (kipB) 272 (�/), 570 ()

A. oryzae AO090120000272 327 (�)A. terreus ATEG_09578 262 (�), 316 (�/)

A. fumigatus AFUA_6G08120 246 (�) 250 (�/) Similarity to protein kinase (SldA/Bub1)regulating the spindle assembly checkpointA. nidulans AN3946.3 253 (�), 311 (�)

A. oryzae AO090003000118 230 ()A. terreus ATEG_05847 207 (�)

A. fumigatus AFUA_5G09710 147 () 150 (�) Related to separin protein (BimB) involvedin chromatid segregation and DNA repairA. nidulans AN8783.3 (BimB) 131 ()

A. oryzae AO090005000276 156 ()A. terreus ATEG_06440 159 ()

A. fumigatus AFUA_2G12390 619 (�) 352 (), 367 (), 481 (�) Maintenance of ploidy protein kinase (Mob1)A. nidulans AN6288.3 195 (�), 491 (�)A. oryzae AO090026000371 497 (�)A. terreus ATEG_01172 525 (�)

A. fumigatus AFUA_3G08100 234 (�), 668 () 222 () Spindle checkpoint kinase (Mph1)A. nidulans AN2927.3 247 ()A. oryzae AO090005001468 145 (), 206 (�)A. terreus ATEG_01659 200 (�)

A. fumigatus AFUA_6G02670 262 (�/) 766 () G2-specific protein kinase (NimA)A. nidulans AN9504.3 242 (�/), 328 ()A. oryzae AO090120000152 255 (�/)A. terreus ATEG_07188 268 (�/)

A. fumigatus AFUA_6G08530 244 () 267 (), 639 (�) Sister chromatid separation protein (Src1)A. nidulans AN3910.3 208 ()A. oryzae AO090001000513 167 ()A. terreus ATEG_05798 181 ()

a Relative to ATG. �, plus strand; , minus strand.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

rfxA-RNAi strain at 25°C (Fig. 3A). P. marneffei bimB, a ho-mologue of genes encoding separin/separase involved in cleav-age of cohesin and sister chromatid separation, showed in-creased expression in both the rfxA-RNAi and rfxA OE strainsunder induced conditions during filamentous growth at 25°C,but not during yeast growth at 37°C (Fig. 3B and data notshown). While these results indicate that P. marneffei RfxAmay play a role in spindle dynamics and mitotic exit, the smallfold changes observed may also be a consequence of the ex-perimental conditions required for induction of the dominantalleles in these strains and the inherent asynchrony producedor an indirect consequence of perturbed cell division (see be-low). Isolation of a conditional allele of RfxA would be impor-tant for further investigation. In addition, direct DNA-bindingstudies would be required to distinguish primary RfxA targetsfrom changes in gene regulation due to secondary effects re-sulting from the perturbed growth observed in strains withreduced or enhanced RfxA function.

rfxA is essential for viability in P. marneffei. To characterizethe role of rfxA during growth and development in P. marneffei,we attempted to isolate a strain containing a deletion of thesingle rfxA gene. Strain SPM4 (pyrG niaD) was transformedwith a linear 4.7-kb XhoI/SacII fragment of the rfxA genedeletion construct pHS5595 (Fig. 4A) to replace the majorityof the rfxA coding region with the Aspergillus nidulans pyrG�

blaster cassette (8). A total of 246 pyrG� transformants wereisolated and subsequently screened for integration of the rfxAdeletion construct at the rfxA locus using either PCR or South-ern blot analysis. Despite the large number of transformantsscreened, a strain lacking the rfxA coding sequences was notidentified (data not shown).

One transformant (44.1.2) was identified which was found byPCR screening to contain both rfxA� and rfxA�::pyrG� alleles(data not shown), and this was confirmed by Southern blotanalysis (Fig. 4B). As P. marneffei is a haploid organism, thisresult suggested that this represented either a heterokaryoticstrain containing nuclei of different genotypes or a heterozy-gous diploid. The isolation of such strains is not uncommon inmany fungi and is often used to study uncharacterized, poten-tially essential genes (45). While actively growing vegetativehyphal cells are predominantly multinucleate, during conidia-tion, single nuclei are partitioned into the conidium-producing

FIG. 3. Relative expression of putative rfxA target genes in strainswith reduced and increased rfxA expression. Total RNA was extractedfrom the P. marneffei rfxA� (SPM4), xylPp::rfxA-RNAi (55.2.1), andxylPp::rfxA (49.3.1) strains during vegetative hyphal growth at 25°Cunder conditions of delayed induction. Cultures were grown for 3 daysin liquid ANM containing 1% glucose and supplemented with 10 mMGABA before being harvested and transferred to ANM containing0.5% xylose and 0.5% sucrose for 6 h (A) or 1 day (B) of growth.(A) RT-PCR was performed on the genes rfxA (27 cycles), mobA (28cycles), and benA (25 cycles). The rfxA-specific primers bound outsidethe coding region included in the xylPp::rfxA-RNAi construct. (B) Theexpression of the bimB transcript was determined using quantitativereal-time RT-PCR. NI, noninduced (1% glucose); I, induced (0.5%xylose and 0.5% sucrose). The values were normalized to actin as aloading control and are shown relative to the rfxA� strain under non-inducing conditions. The error bars represent standard errors of themean (n � 3).

FIG. 4. Strain 44.1.2 is a heterozygous diploid containing rfxA� andrfxA�::pyrG� alleles. (A) Partial restriction map for the wild-type rfxAgenomic locus (top) and the expected rfxA deletion construct pHS5595(bottom) after integration. The gray box represents the rfxA codingregion, consisting of four exons, which has been replaced by the A.nidulans pyrG blaster cassette (pyrG�; white box) in the deletion con-struct. The hatched box indicates the 1 kb KpnI/XhoI fragment ofpHS6313 used as a probe for the Southern blot hybridization experi-ment in panel B. (B) Southern blot hybridization of genomic DNA(gDNA) isolated from the wild-type strain 2161 (WT) and strain 44.1.2(�/�), containing both rfxA� and rfxA�::pyrG� alleles. The restrictionenzymes used for digestion are indicated above the appropriate panels.The probe used for hybridization is shown in panel A (hatched box).(C) The rfxA� (33.4.2) and rfxA�/rfxA� (44.1.2) strains were pointinoculated onto slides containing a thin layer of 0.1% glucose ANMplus GABA and incubated for 4 days at 25°C. Microscopic images ofconidiophores were taken using DIC and epifluorescence optics toobserve nuclei stained with DAPI. The scale bars represent 20 �m.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

sterigmata cells and subsequently into conidia. Genotyping ofcolonies isolated from purified uninucleate conidia usinggrowth tests and PCR screening showed that the uninucleateconidia of strain 44.1.2 were genetically identical and con-tained both rfxA� and rfxA�::pyrG� alleles, showing that theseisolates represented heterozygous diploids (data not shown).Microscopic examination of DAPI-stained conidia of strain44.1.2 verified that each conidium contained only a single nu-cleus and that these conidia were larger than those of thecontrol haploid strain (Fig. 4C). The average volume of 44.1.2conidia was 159.2 � 23 �m3 (� standard deviation; n � 30),approximately 2 to 3 times that of the haploid rfxA� controlstrain (33.4.2), 63.6 � 10.6 �m3. It is well established thatconidial size increases with ploidy in fungi (12), and therefore,these data are fully consistent with the hypothesis that 44.1.2represents a heterozygous rfxA�/rfxA�::pyrG� diploid strain.The isolation of a diploid strain heterozygous for deletion ofrfxA, and the inability to obtain a haploid rfxA deletion strain,strongly suggests that rfxA is an essential gene of P. marneffei.

Haploinsufficiency of rfxA leads to reduced growth, nu-clear division defects, and genomic instability at 37°C. Itwas anticipated that deletion of a single copy of rfxA in therfxA�/rfxA�::pyrG� heterozygous diploid strain (44.1.2) wouldhave no effects on growth, since this strain also contains awild-type allele of rfxA and is therefore able to produce afunctional RfxA protein. However, while the ability of therfxA�/rfxA�::pyrG� strain (44.1.2) to undergo filamentousgrowth at 25°C was indistinguishable from that of the wild-typecontrol (2161), growth of this strain was significantly reducedunder conditions of yeast growth at 37°C (Fig. 5A). Micro-scopic examination of the rfxA�/rfxA�::pyrG� strain duringgrowth at 37°C revealed multiple cellular defects, including thepresence of swollen and irregularly shaped arthroconidial hy-phae and cell lysis. Using Hoechst staining of DNA, very fewintact nuclei were observed, and instead, enlarged nuclearmasses, characteristic of endoreplication (DNA rereplicationin the absence of nuclear division), and small nuclear frag-ments of irregular shape were prevalent (Fig. 5B).

The genetic stability of the rfxA�/rfxA�::pyrG� strain wasalso compromised during growth at 37°C. This was initiallyapparent upon isolation of single cells (protoplasts) from themulticellular arthroconidial hyphae of strain 44.1.2 grown at37°C, as approximately 90% of viable protoplasts had revertedto uracil auxotrophy (data not shown). Genotyping of theseputative haploid isolates by PCR confirmed loss of therfxA�::pyrG� allele associated with the onset of uracil auxot-rophy, and therefore, these strains represented wild-type rfxA�

haploids derived from the rfxA�/rfxA�::pyrG� strain (data notshown). The remaining protoplasts displayed uracil prototro-phy and contained both rfxA� and rfxA�::pyrG� alleles, asverified by PCR; however, many of these isolates displayed anabnormal morphology (data not shown). These strains werehighly unstable and frequently reverted to a wild-type growthphenotype, which was subsequently stable (data not shown).No haploids containing only the rfxA�::pyrG� allele were iso-lated. In fungi, the generation of haploid isolates from a dip-loid progenitor (haploidization) occurs via random chromo-some loss due to nondisjunction of chromosomes, as a result ofeither genetic mutation or the presence of genotoxic com-pounds (30, 64). The uracil prototrophic colonies displaying

the abnormal growth morphology phenotype likely repre-sented aneuploid strains that had not undergone completehaploidization, since both wild-type chromosomes and chro-mosomes with rfxA deleted were being maintained. Reversionto the wild-type growth phenotype may result from imbalancedchromosome segregation in the aneuploid strains, leading torestoration of the diploid state. Both the observed nuclearfragmentation and haploidization of the heterozygous diploidduring growth at 37°C, but not at 25°C, are suggestive ofnuclear division defects, possibly due to haploinsufficiency ofrfxA. In addition, the observation that only wild-type haploidscontaining the rfxA� allele, and not the rfxA�::pyrG� allele,

FIG. 5. The heterozygous rfxA�/rfxA� strain displays cellular divi-sion defects during yeast morphogenesis at 37°C. (A) The rfxA� pyrG�

(33.4.2), rfxA� pyrG (SPM4), and rfxA�/rfxA�::pyrG� (44.1.2) strainswere grown on either ANM containing 10 mM GABA at 25°C or onSD containing 10 mM (NH4)3SO4 at 37°C, with (�) and without ()the addition of 10 mM uracil. The plates were photographed after 6days of incubation. (B) Microscopic examination of wild-type rfxA�

(2161) and rfxA�/rfxA�::pyrG� (44.1.2) strains after 4 days of growth at37°C on slides coated with a thin layer of SD medium containing 10mM (NH4)2SO4. The images were captured using both DIC and epi-fluorescence optics to observe nuclei stained with Hoechst. The scalebars represent 20 �m.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

were isolated provides additional evidence that a functionalrfxA gene is required for the viability of P. marneffei.

Reduced RfxA levels result in cell division defects. To studythe consequences of reduced rfxA expression, an RNAi strat-egy was developed in order to silence the endogenous rfxAtranscript by driving high-level expression of an rfxA hairpintranscript. For this purpose, the construct pHS6521 was gen-erated, in which a spacer fragment (GFP) was flanked byinverted repeats of the first 1,256 bp of rfxA 5� coding sequenceand placed downstream of the xylP promoter of Penicilliumchrysogenum, which is effective in inducing high-level geneexpression in P. marneffei in the presence of xylose (51, 70).This construct was targeted to the areA locus, giving rise tostrain 55.2.1. Using semiquantitative RT-PCR, the abundanceof the endogenous rfxA transcript was found to be significantlyreduced upon induced expression of the rfxA-RNAi hairpintranscript in strain 55.2.1 (xylPp::rfxA-RNAi) at both 25°C and37°C (Fig. 3A and data not shown).

At 25°C, the induced expression of the rfxA-RNAi hairpintranscript in strain 55.2.1 resulted in severe growth inhibition.Microscopic examination of colonies revealed multiple growthdefects under inducing conditions, including aseptate andanucleate hyphae and deformed apical hyphal cells and lateralbranches, often with multiple short swollen hyphal tips exhib-iting cell lysis (Fig. 6A). In the older hyphal cells at the centerof the colony, the few nuclei that were observed displayed anabnormal highly elongated appearance.

Due to the severe growth inhibition resulting from overex-pression of the rfxA-RNAi hairpin transcript, the effects onconidiation could be examined only under conditions of de-layed induction. Abnormal conidiophore morphogenesis wasapparent upon induced expression of the rfxA-RNAi hairpintranscript, with a range of phenotypes observed (Fig. 6B). Theyincluded sterigmata cells and conidia that appeared engorgedand contained large numbers of nuclei, demonstrating defectsin nuclear partitioning. At the most severe end of the spec-trum, conidiophores lacked appropriate differentiation of thesterigmata cell types and consisted of conidiophore stalks withaberrant swollen and misshapen sterigmata cells. These cellslacked intact nuclei and instead contained fragmented nuclearstructures. This suggests that reduced rfxA function leads todefects in nuclear division and segregation during differentia-tion of the uninucleate cell types of the conidiophore.

The xylPp::rfxA-RNAi strain also displayed reduced growthunder conditions required for yeast morphogenesis at 37°C.Microscopic examination revealed growth phenotypes similarto those of the heterozygous rfxA�/rfxA�::pyrG� diploid strain.They included severe growth phenotypes, such as cell lysis andthe presence of swollen hyphae, which lacked intact nuclearstructures and instead contained small nuclear fragments (Fig.6C). Although reminiscent of arthroconidial hyphae, thesecells were aseptate, and no hyphal fragments or yeast cells,indicative of arthroconidiation, were observed.

Overexpression of rfxA results in cell division defects andgrowth arrest. To further examine the role of rfxA duringgrowth and morphogenesis, the rfxA coding region was in-serted downstream of the xylP promoter in the plasmidpHS6259 in order to drive high levels of rfxA expression in axylose-inducible manner. This construct was targeted to theareA locus in single copy. Growth of the xylPp::rfxA strain

49.3.1 was almost completely inhibited under inducing condi-tions at 25°C. Growth arrest was observed at an early stage ofgermination, coinciding with germ tube elongation and thepresence of either one (no nuclear division) or two (one nu-clear division) nuclei (data not shown). A small proportion ofthe inoculum was able to commence growth, and microscopicexamination of these hyphal cells during overexpression of rfxArevealed a drastic reduction in cell size and an increase in thenumber of nuclei partitioned into each subapical cell compart-ment (Table 4). Additional phenotypes identified included anincreased occurrence of lateral branches, some resemblingshort undifferentiated conidiophore stalks (Fig. 7A). Thesephenotypes were also observed under conditions of delayedinduction (see below).

To assess conidiophore morphogenesis upon overexpressionof rfxA, strain 49.3.1 was examined microscopically under con-ditions of delayed induction. The induced overexpression ofrfxA did not lead to any major conidiation defects; however, anincrease in the frequency of conidiophores, particularly towardthe periphery of the colony, was observed. Furthermore,many of these conidiophores appeared rudimentary, withshorter stalks and reduced abundance of sterigmata celltypes (Fig. 7B).

Under inducing conditions at 37°C, a complete lack ofgrowth was observed for the xylPp::rfxA strain (data notshown). Microscopic examination revealed that while manyconidia appeared to have initiated germination, as shown bythe presence of enlarged cells with some protruding short germtubes, the growth of these cells appeared to be arrested, withonly a single nucleus observed (Fig. 7C).

Both the DNA-binding and dimerization domains of RfxAare required to modulate its function. The predicted P.marneffei RfxA protein contains the highly conserved RFXDNA-binding domain, as well as a putative dimerization do-main. In some cases, these domains have been shown to func-tion independently, where the formation of either homo- orheterodimers is not a prerequisite for DNA binding in vitro andprotein dimers can be formed when the dimerization domain isfused to a heterologous DNA-binding domain (21, 35, 52). Toassess the codependence of these domains for the normalfunctioning of RfxA, mutant alleles of rfxA encoding RfxAproteins lacking either the DNA-binding domain (DBD�) orthe dimerization domain (DIM�) were inserted downstreamof the xylP promoter in the constructs pHS6522 and pHS6530,respectively. These constructs were targeted to the areA locus.

During filamentous hyphal growth at 25°C, overexpressionof either the rfxA-DBD� (56.4.4) or rfxA-DIM� (49.4.1) allelehad no effect on growth (data not shown). During yeast mor-phogenesis at 37°C, overexpression of rfxA-DIM� resulted inslightly reduced growth; however, no morphological defectswere apparent upon microscopic examination. In comparison,overexpression of the rfxA-DBD� allele caused a severe re-duction in growth at 37°C, with cell lysis apparent uponmicroscopic examination, as well as nuclear division defectscharacterized by enlarged nuclear masses, possibly due to en-doreplication, and nuclear fragmentation (data not shown).These phenotypes recapitulated those previously observed for therfxA�/rfxA� strain 44.1.2 and the induced xylP::rfxA-RNAi strainduring yeast morphogenesis at 37°C.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

Reduced rfxA function leads to defective checkpoint regula-tion. In light of the cellular division defects observed inresponse to a reduced level of rfxA expression in thexylPp::rfxA-RNAi strain, we assessed the response of thisstrain to known inhibitors of the cell cycle. HU, a directinhibitor of RNR function, results in activation of the S-

phase checkpoint by preventing completion of DNA repli-cation (16). In addition, the microtubule-destabilizing com-pound Benomyl causes a block in late mitosis due toactivation of the spindle checkpoint pathway in response tothe formation of a defective mitotic spindle (49). Mutantsdefective in the activation of checkpoints in response to a

FIG. 6. Silencing of the rfxA transcript by RNAi leads to cell division defects and poor growth. Microscopic examination of wild-type rfxA�

(SPM4) and xylPp::rfxA-RNAi (55.2.1) strains are shown. (A) Strains after 4 days of growth at 25°C on slides coated with a thin layer of solid ANMplus 10 mM GABA in the presence of either 1% glucose (noninducing) or 0.5% xylose and 0.5% sucrose (inducing). Nuclei (single arrowheads)and septa (double arrowheads) are shown. (B) Strains after 2 days of growth at 25°C on slides coated with a thin layer of solid 0.1%glucose-containing ANM plus 10 mM GABA, followed by an additional 2 days of growth at 25°C after the addition of 5 ml of liquid mediumcontaining either 1% glucose (noninduced) or 0.5% xylose and 0.5% sucrose (induced) to the bottom of the slide. The white and black arrowheadsrepresent misshapen sterigmata cells with fragmented nuclei and enlarged sterigmata cells and conidia containing multiple nuclei, respectively.(C) Strains after 4 days of growth at 37°C on slides coated with a thin layer of SD plus 10 mM (NH4)2SO4 solid medium under noninducing orinducing conditions. The images were taken using both DIC and epifluorescence optics to observe calcofluor and Hoechst (C/H) or Hoechst(H) staining. The scale bars represent 20 �m. The rightmost image in panel C represents a further 2� enlargement of the boxed area in the imageto its left.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

transient exposure to HU or Benomyl display reduced viabilitydue to the completion of defective mitotic divisions (20, 23).Both the wild-type rfxA� and xylPp::rfxA-RNAi strains weregerminated for 24 h in the presence of HU (2.5 or 10 mM) orBenomyl (0.01, 0.05, 0.2, or 0.5 �g ml1) under both inducingand noninducing conditions prior to being assessed for viabil-ity. The wild-type (rfxA�) and xylPp::rfxA-RNAi strains dis-played similar viability after transient exposure to Benomyland under noninducing conditions in the presence of HU (datanot shown). The reduced expression of rfxA in the xylPp::rfxA-RNAi strain under inducing conditions caused a signifi-cant reduction in viability after treatment with HU (Fig. 8). Asimilar reduction in viability was observed for thexylPp::rfxA-RNAi strain when germlings were exposed to HUfor only 6 h. However, the ability to arrest nuclear division inresponse to HU was not affected in the xylPp::rfxA-RNAistrain. Approximately 15% of germlings of the xylPp::rfxA-RNAi (17.9% � 8.6% [standard error of the mean {SEM}])and wild-type control (15% � 5.5%) strains underwent nucleardivision in the presence of HU, in contrast to 50% of germlingsin the absence of HU (49.2% � 7.9% and 47.4% � 6.8% forthe xylPp::rfxA-RNAi and wild-type control strains, respec-tively). Thus, although mitosis is inhibited, presumably via ac-tivation of the S-phase checkpoint, reduced RfxA functionprevents the appropriate response to DNA replication inhibi-tion to ensure survival. Given that the exposure to HU wastransient in nature, this may include the ability to slow S-phaseprogression and to stabilize and subsequently reactivate stalledreplication forks, thus preventing double-stranded DNAbreaks and facilitating DNA repair.

DISCUSSION

Alterations in the expression of rfxA lead to cellular divisionand growth defects. The rfxA gene of P. marneffei encodes amember of the RFX transcription factor family involved in theregulation of cellular differentiation events in eukaryotes. Thedata presented here show that rfxA is essential for viability andthat alterations in the levels of its expression lead to cell divi-sion defects with dramatic consequences for growth and mor-phogenesis. Despite numerous attempts, a haploid rfxA dele-tion strain could not be isolated. Instead the isolation of aheterozygous diploid strain from which only wild-type haploidscould be recovered suggests that rfxA is essential for survival.Examination of strains with reduced or enhanced rfxA expres-sion demonstrated that rfxA might be involved in cell cycle

regulation, consistent with the essential nature of the geneproduct. In A. nidulans, numerous genes involved in cell cycleregulation are essential for viability and were initially identifiedas temperature-sensitive lethal mutants (46).

The overexpression of rfxA resulted in growth arrest duringconidial germination. Although the nature of this growth arrestis unclear, it suggests a role for RfxA in the reactivation ofgrowth in dormant conidia. The observation that hyphal cellsdisplay increased nuclear division kinetics (increased numbersof nuclei per cell and shorter cell length) when rfxA is overex-pressed implicates RfxA in the temporal regulation of genesaffecting cell cycle events.

In contrast, reduced levels of rfxA caused by expression ofthe RNAi transcript, resulted in terminal degenerative pheno-types at 25°C characterized by hyphal cells that were aseptateand either produced aberrant elongated nuclei, indicative of ablock in mitosis, or were anucleate. Similar phenotypes areobserved in cell cycle mutants of A. nidulans affected in chro-mosome metabolism (sepB and sepJ) and cohesion (bimB andbimD), spindle dynamics (bimC), telomere maintenance(nimU/pot1), and mitotic exit (bimA) (15, 22, 28, 29, 40, 50, 66).In these mutants, the nuclear phenotypes result from prema-ture entry into mitosis prior to completion of DNA synthesisand/or failure to completely separate chromosomes during di-vision, preventing cells from exiting mitosis. The lack of sep-tation observed for many of these mutants is a secondaryconsequence of errors in DNA metabolism causing inhibitionof cytokinesis via activation of the DNA damage checkpointpathway (29, 66). The aseptate phenotype of the RNAi strainis also consistent with this hypothesis.

During growth at 37°C, reduced rfxA expression in the RNAistrain resulted in an inability to undergo yeast morphogenesis.Septation, fundamental to the process of arthroconidiation andthe liberation of yeast cells, was not observed as either a directconsequence of reduced RfxA function or a secondary effect ofthe terminal degenerative phenotypes observed (cell lysis andnuclear division defects). Mitotic catastrophe, as evidenced byenlarged nuclear masses, suggestive of endoreplication, andnuclear fragmentation was readily apparent. These phenotypeswere recapitulated in the heterozygous diploid strain specifi-cally during growth at 37°C, presumably as a consequence ofgene dosage effects (haploinsufficiency). In support of this hy-pothesis, haploinsufficiency has also been observed for CRT1/crt1� and RFX2/rfx2� heterozygous diploids in S. cerevisiae andC. albicans, respectively (27, 32). Such nuclear defects may alsoaccount for the apparent loss of genomic stability resulting inthe spontaneous haploidization of the heterozygous diploidstrain. Mutations affecting components of the cell cycle appa-ratus, such as in the genes sepB, sepJ, bimA, nimU (pot1), andhfaB of A. nidulans, are known to result in loss of genomicstability (28, 29, 50, 64, 66).

Given the role of the related RFX proteins in mediating theDDR in closely related organisms, it is interesting to speculatewhether RfxA is involved in a similar mechanism in P. marnef-fei, particularly in light of the increased sensitivity to HU ob-served under conditions of rfxA RNAi. Importantly, RFX actsa negative regulator in its role as an effector of the DDR andthus leads to derepression of the target genes, including theRNR-encoding genes, in the presence of HU (27, 32, 38).Thus, reduced RFX function would be expected to promote

TABLE 4. Average cell lengths and nuclear abundances duringoverexpression of rfxA

Strain Growthconditions

Avg cell length(�m; �SEM;

50 cells; n � 3)

Avg no. of nuclei/cell(�SEM; 50 cells;

n � 3)

rfxA� Noninducinga 66.8 � 0.61 1.17 � 0.04Inducingb 66.7 � 1.3 1.17 � 0.05

xylPp::rfxA� Noninducing 63.3 � 3.2 1.2 � 0.02Inducing 34.7 � 2.1 2.07 � 0.08

a Noninducing, 1% glucose.b Inducing, 0.5% xylose and 0.5% sucrose.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

survival in the presence of HU, whereas the results presentedhere show that in P. marneffei the converse is true. PutativeRFX binding sites have recently been identified in the promot-ers of both RNR-encoding genes in P. marneffei (data notshown), and the results suggest that RfxA might positivelyregulate these genes during DNA replication arrest.

If RfxA is involved in the DNA damage response in P.marneffei, this is unlikely to be its sole function, as componentsexclusive to the DDR in closely related fungi are rarely essen-tial for survival under non-DNA damage conditions (26). Al-most half of the potential RfxA target genes to which cellularroles could be assigned represent gene products involved in

FIG. 7. Overexpression of rfxA leads to cell division defects and growth arrest. (A) Microscopic examination of wild-type rfxA� (SPM4) andxylPp::rfxA (49.3.1) strains after 4 days of growth at 25°C on slides coated with a thin layer of solid ANM plus 10 mM GABA in the presence ofeither 1% glucose (noninduced) or 0.5% xylose and 0.5% sucrose (induced). The images were taken using both DIC and epifluorescence opticsto observe calcofluor and Hoechst (C/H) staining. The scale bars represent 20 �m. (B) Microscopic examination of wild-type rfxA� (SPM4) andxylPp::rfxA (49.3.1) strains after 2 days of growth at 25°C on slides coated with a thin layer of solid 0.1% glucose-containing ANM plus 10 mMGABA followed by an additional 2 days of growth at 25°C after the addition of 5 ml of liquid medium containing either 1% glucose (noninduced)or 0.5% xylose and 0.5% sucrose (induced) to the bottom of the slide. The images were taken using DIC optics. The scale bars represent 50 �m,and conidiophores are indicated by arrowheads. (C) Microscopic examination of wild-type rfxA� (SPM4) and xylPp::rfxA (49.3.1) strains after 4 daysof growth at 37°C on slides coated with a thin layer of solid SD plus 10 mM (NH4)2SO4 in the presence of either 1% glucose (noninduced) or 0.5%xylose and 0.5% sucrose (induced). The images were taken using both DIC and epifluorescence optics to observe Hoechst staining. The scale barsrepresent 20 �m.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

cell cycle events, with particular emphasis on chromosomalmechanics, spindle organization, and mitotic exit. Two of theseputative targets (mobA and bimB) displayed increased expres-sion in strains with altered RfxA activity. The maintenance ofthe ploidy kinase MobA, upregulated during rfxA RNAi,causes defects in mitotic progression and septation when over-expressed in fission yeast (54). The altered expression of theseparase BimB suggests that RfxA may influence chromosomedynamics by perturbing sister chromatid cohesion (40, 63).

RfxA is important for linking cellular division with morpho-genesis. The data suggest that the function of RfxA in theregulation of cell cycle progression is strongly influenced bythe cellular context. While the overexpression of rfxA increasesthe rate of cell division in hyphal cells, overexpression of rfxAin conidia caused growth arrest at an early stage of germina-tion. In addition, nuclear division defects were observed for theheterozygous diploid and xylP::rfxA-DBD� strains only duringyeast growth at 37°C, and not during filamentous growth at25°C. During both conidiation and yeast morphogenesis in P.marneffei, nuclear and cellular division are tightly coupled toensure the maintenance of the uninucleate state of these celltypes. In contrast, hyphae are coenocytic/multinucleate, and inA. nidulans, aside from the initial nuclear divisions duringgermination, nuclear division occurs in a parasynchronouswave (13). Thus, hyphal growth may be less sensitive to per-turbation in RfxA activity than the tightly regulated nucleardivision events occurring during development and yeast mor-phogenesis, and possibly even during germination.

Regulated changes in cell cycle dynamics are often associ-ated with cellular differentiation events in fungi, including an

increased rate of cell cycling during conidiation in A. nidulans(43, 68). Effects on conidiophore morphogenesis were ob-served in P. marneffei when the levels of rfxA expression wereperturbed. Reduced rfxA expression in the rfxA-RNAi strainled to the production of defective conidiophore structures,reflecting disturbances in the coupling of mitosis with cytoki-nesis. Similar defects resulting in inappropriate conidiophoremorphogenesis result from aberrant checkpoint regulation inA. nidulans due to the NimXcdc2-AF mutation, preventing neg-ative regulation of the cyclin-dependent kinase by inhibitoryTyr-15 phosphorylation (68). Interestingly, while overexpres-sion of rfxA caused an increase in mitotic cycling in basalhyphal cell compartments, the conidiophores produced weremorphologically normal, although increased in frequency. Thismay reflect the normal upregulation of rfxA occurring duringconidiation in wild-type cells. It is possible that RfxA directlyinfluences the onset of development through regulation of thetranscriptional regulator BrlA, a key component of the centralregulatory pathway involved in conidiophore development,since conserved RFX consensus binding sites were identified inthe promoters of brlA genes from the aspergilli and P. marnef-fei, and induction of brlA expression has previously been shownto be sufficient for initiation of conidiation (1, 41; Borneman etal., unpublished).

Regulation of RfxA activity. The underlying question of howRfxA integrates the regulation of cellular division with differ-entiation events in the different cell types produced by P.marneffei remains. The possibility that RfxA mediates its cell-type-specific roles through interaction with additional regula-tory proteins in a cell-type-dependent manner is supported bythe analysis of strains overexpressing rfxA alleles lacking theregions encoding the DNA-binding or dimerization domains.Overexpression of the rfxA-DBD� allele, containing a func-tional dimerization domain, resulted in a dominant-negativeeffect, suggesting that regulation of RfxA function occurs viaprotein interactions. This may involve one or more bindingpartners required for appropriate regulation of target genes,which are sequestered by the RfxA-DBD� protein. A similarmode of action has been proposed to account for the dominantinterfering effects resulting from a version of human RFX1with the DNA-binding domain deleted on the regulation ofRFX1 target genes (38). In contrast, the putative dimerizationdomain appears to be required for protein function, sinceoverexpression of a mutant allele lacking this domain has noobvious phenotypic effects, contrary to overexpression of thewild-type allele. While the stability of the RfxA-DIM� proteinhas not been assessed in this strain and could account for thelack of phenotype, this finding is consistent with observationsin A. chrysogenum, where the formation of CPCR1 dimers isessential for both DNA binding and the interaction withAcFKH1 (57, 58). These data suggest that the formation ofRfxA homodimers and the interaction with accessory proteins,possibly involving a forkhead protein, are necessary for appro-priate regulation of RfxA target genes.

Recent evidence in fission yeast also supports a role for theinteraction of RFX and forkhead proteins in cell cycle regula-tion. In S. pombe, the forkhead proteins Fkh2 and Sep1 arerequired for the G2/M-specific regulation of genes required forlate mitotic events and cytokinesis (9, 44). Interestingly, manyof these forkhead-regulated genes have orthologues identified

FIG. 8. Silencing of the rfxA transcript by RNAi leads to reducedviability after DNA replication inhibition imposed by HU. Approxi-mately 2 � 105 conidia of the wild-type rfxA� (SPM4) andxylPp::rfxA-RNAi (55.2.1) strains were inoculated into 1 ml of liquidANM plus 10 mM GABA containing 0.5% xylose and 0.5% sucrose(inducing conditions) and incubated at 25°C for 24 h in the presenceHU (2, 5, or 10 mM) or were pregerminated for 18 h in the absence ofHU followed by a transient 6-h exposure to HU (2 or 10 mM). Theconidial suspensions were diluted twice (1/10 dilution) to remove theHU, and 100 �l was spread onto plates containing 1% glucose ANMplus 10 mM GABA (noninducing). The CFU were counted after 2days of incubation at 25°C, and survival is expressed as the percentviability in the absence of HU.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

as putative RfxA targets in this study. Both Fkh2 and Sep1 areassociated with the PCB (pombe cell cycle box) site in thepromoters of these genes as components of a multiproteintranscription factor, PBF (PCB-binding factor) (2, 48). Notonly does the PCB sequence reflect the 3� half site of the RFXconsensus sequence, and not a prototypical forkhead consen-sus sequence, but additional consensus RFX and FKH bindingsequences are also present in the promoters of PCB-regulatedgenes (H. Bugeja and A. Andrianopoulos, unpublished data).PBF has negative (mediated by Fkh2) and positive (mediatedby Sep1) regulatory roles at the promoters of PCB-containinggenes (48), consistent with the context-dependent regulation ofRFX target genes in other organisms. The S. pombe RFXprotein Sak1 has an essential yet biochemically undefined rolein the regulation of cell cycle events. Recently, synthetic inter-actions between Sak1 and anaphase-promoting complex/cyclo-some (APC/C) components, including separase, have been de-scribed (39, 69). Thus, its involvement as a possible componentof PBF is worthy of further investigation.

RFX proteins as regulators of cell division and differentia-tion in eukaryotes. In P. marneffei, the RFX protein RfxA isrequired for integrating the mitotic cell cycle with cellulardifferentiation events in a cell context-dependent manner. Inother organisms where RFX proteins have been found to beinvolved in cell cycle regulation, such as humans (RFX1), Dro-sophila (dRFX2), and S. pombe yeast (Sak1), reduced RFXprotein function also has dramatic consequences for growth.This is not observed in S. cerevisiae, where crt1 mutants areviable, reflecting the underlying differences in checkpoint con-trol mechanisms in these organisms. Aside from the role ofRFX proteins in regulating cell cycle events or the DNA dam-age response, these factors are also involved in the temporal orspatial regulation of cellular differentiation events, includingciliated sensory neuron development in D. melanogaster(dRFX) and C. elegans (DAF-19) and arthrosporulation in A.chrysogenum (CPCR1). Thus, not only do RFX proteins havea fundamental role in cell cycle regulation, this is also mani-fested in the altered patterns of cell division required for mor-phogenesis in developmentally complex organisms.

ACKNOWLEDGMENTS

This work was supported by grants to A.A. from the AustralianResearch Council, the National Health and Medical Research Coun-cil, and the Howard Hughes Medical Institute. A.A. is a HowardHughes Medical Institute International Scholar. H.E.B. was supportedby an Australian Postgraduate Award.

We thank H. Robertson and K. J. Boyce for critical comments on themanuscript.

REFERENCES

1. Adams, T. H., M. T. Boylan, and W. E. Timberlake. 1988. brlA is necessaryand sufficient to direct conidiophore development in Aspergillus nidulans.Cell 54:353–362.

2. Anderson, M., S. S. Ng, V. Marchesi, F. H. MacIver, F. E. Stevens, T. Riddell,D. M. Glover, I. M. Hagan, and C. J. McInerny. 2002. Plo1(�) regulatesgene transcription at the M-G(1) interval during the fission yeast mitotic cellcycle. EMBO J. 21:5745–5755.

3. Andrianopoulos, A., and W. E. Timberlake. 1994. The Aspergillus nidulansabaA gene encodes a transcriptional activator that acts as a genetic switch tocontrol development. Mol. Cell. Biol. 14:2503–2515.

4. Ashburner, M., C. A. Ball, J. A. Blake, D. Botstein, H. Butler, J. M.Cherry, A. P. Davis, K. Dolinski, S. S. Dwight, J. T. Eppig, M. A. Harris,D. P. Hill, L. Issel-Tarver, A. Kasarskis, S. Lewis, J. C. Matese, J. E.Richardson, M. Ringwald, G. M. Rubin, and G. Sherlock. 2000. Geneontology: tool for the unification of biology. The Gene Ontology Con-sortium. Nat. Genet. 25:25–29.

5. Ausubel, F. M., R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, andJ. A. Smith. 1994. Current protocols in molecular biology. John Wiley &Sons, Inc., New York, NY.

6. Borneman, A. R., M. J. Hynes, and A. Andrianopoulos. 2000. The abaAhomologue of Penicillium marneffei participates in two developmental pro-grammes: conidiation and dimorphic growth. Mol. Microbiol. 38:1034–1047.

7. Borneman, A. R., M. J. Hynes, and A. Andrianopoulos. 2001. An STE12homolog from the asexual, dimorphic fungus Penicillium marneffei comple-ments the defect in sexual development of an Aspergillus nidulans steAmutant. Genetics 157:1003–1014.

8. Borneman, A. R., M. J. Hynes, and A. Andrianopoulos. 2002. A basic helix-loop-helix protein with similarity to the fungal morphological regulators,Phd1p, Efg1p and StuA, controls conidiation but not dimorphic growth inPenicillium marneffei. Mol. Microbiol. 44:621–631.

9. Buck, V., S. S. Ng, A. B. Ruiz-Garcia, K. Papadopoulou, S. Bhatti, J. M.Samuel, M. Anderson, J. B. Millar, and C. J. McInerny. 2004. Fkh2p andSep1p regulate mitotic gene transcription in fission yeast. J. Cell Sci. 117:5623–5632.

10. Chan, Y. F., and T. C. Chow. 1990. Ultrastructural observations on Penicil-lium marneffei in natural human infection. Ultrastruct. Pathol. 14:439–452.

11. Chang, Y. C., and W. E. Timberlake. 1993. Identification of Aspergillus brlAresponse elements (BREs) by genetic selection in yeast. Genetics 133:29–38.

12. Clutterbuck, A. J. 1969. Cell volume per nucleus in haploid and diploidstrains of Aspergillus nidulans. J. Gen. Microbiol. 55:291–299.

13. Clutterbuck, A. J. 1970. Synchronous nuclear division and septation in As-pergillus nidulans. J. Gen. Microbiol. 60:133–135.

14. Cove, D. J. 1966. The induction and repression of nitrate reductase in thefungus Aspergillus nidulans. Biochim. Biophys. Acta 113:51–56.

15. Denison, S. H., E. Kafer, and G. S. May. 1993. Mutation in the bimD geneof Aspergillus nidulans confers a conditional mitotic block and sensitivity toDNA damaging agents. Genetics 134:1085–1096.

16. Desany, B. A., A. A. Alcasabas, J. B. Bachant, and S. J. Elledge. 1998.Recovery from DNA replicational stress is the essential function of theS-phase checkpoint pathway. Genes Dev. 12:2956–2970.

17. Dubruille, R., A. Laurencon, C. Vandaele, E. Shishido, M. Coulon-Bublex, P.Swoboda, P. Couble, M. Kernan, and B. Durand. 2002. Drosophila regulatoryfactor X is necessary for ciliated sensory neuron differentiation. Develop-ment 129:5487–5498.

18. Dutton, J. R., S. Johns, and B. L. Miller. 1997. StuAp is a sequence-specifictranscription factor that regulates developmental complexity in Aspergillusnidulans. EMBO J. 16:5710–5721.

19. Efimenko, E., K. Bubb, H. Y. Mak, T. Holzman, M. R. Leroux, G. Ruvkun,J. H. Thomas, and P. Swoboda. 2005. Analysis of xbx genes in C. elegans.Development 132:1923–1934.

20. Efimov, V. P., and N. R. Morris. 1998. A screen for dynein synthetic lethalsin Aspergillus nidulans identifies spindle assembly checkpoint genes andother genes involved in mitosis. Genetics 149:101–116.

21. Emery, P., B. Durand, B. Mach, and W. Reith. 1996. RFX proteins, a novelfamily of DNA binding proteins conserved in the eukaryotic kingdom. Nu-cleic Acids Res. 24:803–807.

22. Enos, A. P., and N. R. Morris. 1990. Mutation of a gene that encodes akinesin-like protein blocks nuclear division in A. nidulans. Cell 60:1019–1027.

23. Fagundes, M. R., J. F. Lima, M. Savoldi, I. Malavazi, R. E. Larson, M. H.Goldman, and G. H. Goldman. 2004. The Aspergillus nidulans npkA geneencodes a Cdc2-related kinase that genetically interacts with the UvsBATRkinase. Genetics 167:1629–1641.

24. Gajiwala, K. S., H. Chen, F. Cornille, B. P. Roques, W. Reith, B. Mach, andS. K. Burley. 2000. Structure of the winged-helix protein hRFX1 reveals anew mode of DNA binding. Nature 403:916–921.

25. Garrison, R., and K. Boyd. 1973. Dimorphism of Penicillium marneffei asobserved by electron microscopy. Can. J. Microbiol. 19:1305–1309.

26. Goldman, G. H., and E. Kafer. 2004. Aspergillus nidulans as a model systemto characterize the DNA damage response in eukaryotes. Fungal Genet.Biol. 41:428–442.

27. Hao, B., C. J. Clancy, S. Cheng, S. B. Raman, K. A. Iczkowski, and M. H.Nguyen. 2009. Candida albicans RFX2 encodes a DNA binding proteininvolved in DNA damage responses, morphogenesis, and virulence. Eu-karyot. Cell 8:627–639.

28. Harris, S. D., and J. E. Hamer. 1995. sepB: an Aspergillus nidulans geneinvolved in chromosome segregation and the initiation of cytokinesis. EMBOJ. 14:5244–5257.

29. Harris, S. D., and P. R. Kraus. 1998. Regulation of septum formation inAspergillus nidulans by a DNA damage checkpoint pathway. Genetics 148:1055–1067.

30. Hastie, A. C. 1970. Benlate-induced instability of Aspergillus diploids. Nature226:771.

31. Hoff, B., E. K. Schmitt, and U. Kuck. 2005. CPCR1, but not its interactingtranscription factor AcFKH1, controls fungal arthrospore formation in Acre-monium chrysogenum. Mol. Microbiol. 56:1220–1233.

32. Huang, M., Z. Zhou, and S. J. Elledge. 1998. The DNA replication anddamage checkpoint pathways induce transcription by inhibition of the Crt1repressor. Cell 94:595–605.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

33. Iwama, A., J. Pan, P. Zhang, W. Reith, B. Mach, D. G. Tenen, and Z. Sun.1999. Dimeric RFX proteins contribute to the activity and lineage specificityof the interleukin-5 receptor � promoter through activation and repressiondomains. Mol. Cell. Biol. 19:3940–3950.

34. Katan, Y., R. Agami, and Y. Shaul. 1997. The transcriptional activation andrepression domains of RFX1, a context-dependent regulator, can mutuallyneutralize their activities. Nucleic Acids Res. 25:3621–3628.

35. Katan-Khaykovich, Y., I. Spiegel, and Y. Shaul. 1999. The dimerization/repression domain of RFX1 is related to a conserved region of its yeasthomologues Crt1 and Sak1: a new function for an ancient motif. J. Mol. Biol.294:121–137.

36. Liu, M., B. H. Lee, and M. B. Mathews. 1999. Involvement of RFX1 proteinin the regulation of the human proliferating cell nuclear antigen promoter.J. Biol. Chem. 274:15433–15439.

37. Livak, K. J., and T. D. Schmittgen. 2001. Analysis of relative gene expressiondata using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method.Methods 25:402–408.

38. Lubelsky, Y., N. Reuven, and Y. Shaul. 2005. Autorepression of rfx1 geneexpression: functional conservation from yeast to humans in response toDNA replication arrest. Mol. Cell. Biol. 25:10665–10673.

39. Matsumura, T., T. Yuasa, T. Hayashi, T. Obara, Y. Kimata, and M.Yanagida. 2003. A brute force postgenome approach to identify tempera-ture-sensitive mutations that negatively interact with separase and securinplasmids. Genes Cells 8:341–355.

40. May, G. S., C. A. McGoldrick, C. L. Holt, and S. H. Denison. 1992. ThebimB3 mutation of Aspergillus nidulans uncouples DNA replication from thecompletion of mitosis. J. Biol. Chem. 267:15737–15743.

41. Mirabito, P. M., T. H. Adams, and W. E. Timberlake. 1989. Interactions ofthree sequentially expressed genes control temporal and spatial specificity inAspergillus development. Cell 57:859–868.

42. Nakayama, A., H. Murakami, N. Maeyama, N. Yamashiro, A. Sakakibara, N.Mori, and M. Takahashi. 2003. Role for RFX transcription factors in non-neuronal cell-specific inactivation of the microtubule-associated proteinMAP1A promoter. J. Biol. Chem. 278:233–240.

43. O’Connell, M. J., A. H. Osmani, N. R. Morris, and S. A. Osmani. 1992. Anextra copy of nimEcyclinB elevates pre-MPF levels and partially suppressesmutation of nimTcdc25 in Aspergillus nidulans. EMBO J. 11:2139–2149.

44. Oliva, A., A. Rosebrock, F. Ferrezuelo, S. Pyne, H. Chen, S. Skiena, B.Futcher, and J. Leatherwood. 2005. The cell cycle-regulated genes of Schizo-saccharomyces pombe. PLoS Biol. 3:e225.

45. Osmani, A. H., B. R. Oakley, and S. A. Osmani. 2006. Identification andanalysis of essential Aspergillus nidulans genes using the heterokaryon rescuetechnique. Nat. Protoc. 1:2517–2526.

46. Osmani, S. A., D. B. Engle, J. H. Doonan, and N. R. Morris. 1988. Spindleformation and chromatin condensation in cells blocked at interphase bymutation of a negative cell cycle control gene. Cell 52:241–251.

47. Otsuki, K., Y. Hayashi, M. Kato, H. Yoshida, and M. Yamaguchi. 2004.Characterization of dRFX2, a novel RFX family protein in Drosophila.Nucleic Acids Res. 32:5636–5648.

48. Papadopoulou, K., S. S. Ng, H. Ohkura, M. Geymonat, S. G. Sedgwick, andC. J. McInerny. 2008. Regulation of gene expression during M-G1-phase infission yeast through Plo1p and forkhead transcription factors. J. Cell Sci.121:38–47.

49. Pinsky, B. A., and S. Biggins. 2005. The spindle checkpoint: tension versusattachment. Trends Cell Biol. 15:486–493.

50. Pitt, C. W., E. Moreau, P. A. Lunness, and J. H. Doonan. 2004. The pot1�

homologue in Aspergillus nidulans is required for ordering mitotic events.J. Cell Sci. 117:199–209.

51. Pongsunk, S., A. Andrianopoulos, and S. C. Chaiyaroj. 2005. Conditionallethal disruption of TATA-binding protein gene in Penicillium marneffei.Fungal Genet. Biol. 42:893–903.

52. Reith, W., C. Herrero-Sanchez, M. Kobr, P. Silacci, C. Berte, E. Barras, S.Fey, and B. Mach. 1990. MHC class II regulatory factor RFX has a novelDNA-binding domain and a functionally independent dimerization domain.Genes Dev. 4:1528–1540.

53. Reith, W., C. Ucla, E. Barras, A. Gaud, B. Durand, C. Herrero-Sanchez, M.Kobr, and B. Mach. 1994. RFX1, a transactivator of hepatitis B virus en-hancer I, belongs to a novel family of homodimeric and heterodimericDNA-binding proteins. Mol. Cell. Biol. 14:1230–1244.

54. Salimova, E., M. Sohrmann, N. Fournier, and V. Simanis. 2000. The S.pombe orthologue of the S. cerevisiae mob1 gene is essential and functionsin signalling the onset of septum formation. J. Cell Sci. 113:1695–1704.

55. Sambrook, J., and D. W. Russell. 2001. Molecular cloning: a laboratorymanual, 3rd ed. Cold Spring Harbor Laboratory Press, Cold Spring Har-bor, NY.

56. Schmitt, E. K., A. Bunse, D. Janus, B. Hoff, E. Friedlin, H. Kurnsteiner, andU. Kuck. 2004. Winged helix transcription factor CPCR1 is involved inregulation of �-lactam biosynthesis in the fungus Acremonium chrysogenum.Eukaryot. Cell 3:121–134.

57. Schmitt, E. K., B. Hoff, and U. Kuck. 2004. AcFKH1, a novel member of theforkhead family, associates with the RFX transcription factor CPCR1 in thecephalosporin C-producing fungus Acremonium chrysogenum. Gene 342:269–281.

58. Schmitt, E. K., and U. Kuck. 2000. The fungal CPCR1 protein, which bindsspecifically to �-lactam biosynthesis genes, is related to human regulatoryfactor X transcription factors. J. Biol. Chem. 275:9348–9357.

59. Siegrist, C. A., B. Durand, P. Emery, E. David, P. Hearing, B. Mach, and W.Reith. 1993. RFX1 is identical to enhancer factor C and functions as atransactivator of the hepatitis B virus enhancer. Mol. Cell. Biol. 13:6375–6384.

60. Steimle, V., B. Durand, E. Barras, M. Zufferey, M. R. Hadam, B. Mach, andW. Reith. 1995. A novel DNA-binding regulatory factor is mutated in pri-mary MHC class II deficiency (bare lymphocyte syndrome). Genes Dev.9:1021–1032.

61. Swoboda, P., H. T. Adler, and J. H. Thomas. 2000. The RFX-type transcrip-tion factor DAF-19 regulates sensory neuron cilium formation in C. elegans.Mol. Cell 5:411–421.

62. Thompson, J. D., F. Plewniak, J. Thierry, and O. Poch. 2000. DbClustal:rapid and reliable global multiple alignments of protein sequences detectedby database searches. Nucleic Acids Res. 28:2919–2926.

63. Uhlmann, F., D. Wernic, M. A. Poupart, E. V. Koonin, and K. Nasmyth.2000. Cleavage of cohesin by the CD clan protease separin triggers anaphasein yeast. Cell 103:375–386.

64. Upshall, A., and I. D. Mortimore. 1984. Isolation of aneuploid-generatingmutants of Aspergillus nidulans, one of which is defective in interphase of thecell cycle. Genetics 108:107–121.

65. van Helden, J. 2003. Regulatory sequence analysis tools. Nucleic Acids Res.31:3593–3596.

66. Wolkow, T. D., P. M. Mirabito, S. Venkatram, and J. E. Hamer. 2000.Hypomorphic bimA (APC3) alleles cause errors in chromosome metabolismthat activate the DNA damage checkpoint blocking cytokinesis in Aspergillusnidulans. Genetics 154:167–179.

67. Wu, S. Y., and M. McLeod. 1995. The sak1� gene of Schizosaccharomycespombe encodes an RFX family DNA-binding protein that positively regu-lates cyclic AMP-dependent protein kinase-mediated exit from the mitoticcell cycle. Mol. Cell. Biol. 15:1479–1488.

68. Ye, X. S., S. L. Lee, T. D. Wolkow, S. L. McGuire, J. E. Hamer, G. C. Wood,and S. A. Osmani. 1999. Interaction between developmental and cell cycleregulators is required for morphogenesis in Aspergillus nidulans. EMBO J.18:6994–7001.

69. Yuasa, T., T. Hayashi, N. Ikai, T. Katayama, K. Aoki, T. Obara, Y. Toyoda,T. Maruyama, D. Kitagawa, K. Takahashi, K. Nagao, Y. Nakaseko, and M.Yanagida. 2004. An interactive gene network for securin-separase, conden-sin, cohesin, Dis1/Mtc1 and histones constructed by mass transformation.Genes Cells 9:1069–1082.

70. Zadra, I., B. Abt, W. Parson, and H. Haas. 2000. xylP promoter-basedexpression system and its use for antisense downregulation of the Penicilliumchrysogenum nitrogen regulator NRE. Appl. Environ. Microbiol. 66:4810–4816.

71. Zaim, J., E. Speina, and A. M. Kierzek. 2005. Identification of new genesregulated by the Crt1 transcription factor, an effector of the DNA damagecheckpoint pathway in Saccharomyces cerevisiae. J. Biol. Chem. 280:28–37.

72. Zajac-Kaye, M., N. Ben-Baruch, E. Kastanos, F. J. Kaye, and C. Allegra.2000. Induction of Myc-intron-binding polypeptides MIBP1 and RFX1 dur-ing retinoic acid-mediated differentiation of haemopoietic cells. Biochem. J.345:535–541.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

the rfx protein rfxa is an essential ... - eukaryotic cell · multicellular hyphal growth at 25°c...

Documents