The Essential Phosphoinositide Kinase MSS-4 Is Required for Polar Hyphal Morphogenesis, Localizing to Sites of Growth and Cell Fusion in Neurospora crassa

Fungal hyphae and plant pollen tubes are among the most highly polarized cells known and pose extraordinary requirements on their cell polarity machinery. Cellular morphogenesis is driven through the phospholipid-dependent organization at the apical plasma membrane. We characterized the contribution of phosphoinositides (PIs) in hyphal growth of the filamentous ascomycete Neurospora crassa. MSS-4 is an essential gene and its deletion resulted in spherically growing cells that ultimately lyse. Two conditional mss-4-mutants exhibited altered hyphal morphology and aberrant branching at restrictive conditions that were complemented by expression of wild type MSS-4. Recombinant MSS-4 was characterized as a phosphatidylinositolmonophosphate-kinase phosphorylating phosphatidylinositol 4-phosphate (PtdIns4P) to phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5)P2). PtdIns3P was also used as a substrate. Sequencing of two conditional mss-4 alleles identified a single substitution of a highly conserved Y750 to N. The biochemical characterization of recombinant protein variants revealed Y750 as critical for PI4P 5-kinase activity of MSS-4 and of plant PI4P 5-kinases. The conditional growth defects of mss-4 mutants were caused by severely reduced activity of MSS-4(Y750N), enabling the formation of only trace amounts of PtdIns(4,5)P2. In N. crassa hyphae, PtdIns(4,5)P2 localized predominantly in the plasma membrane of hyphae and along septa. Fluorescence-tagged MSS-4 formed a subapical collar at hyphal tips, localized to constricting septa and accumulated at contact points of fusing N. crassa germlings, indicating MSS-4 is responsible for the formation of relevant pools of PtdIns(4,5)P2 that control polar and directional growth and septation. N. crassa MSS-4 differs from yeast, plant and mammalian PI4P 5-kinases by containing additional protein domains. The N-terminal domain of N. crassa MSS-4 was required for correct membrane association. The data presented for N. crassa MSS-4 and its roles in hyphal growth are discussed with a comparative perspective on PI-control of polar tip growth in different organismic kingdoms.


Introduction
Establishing polarity and maintaining cell shape are fundamental processes for cellular growth and the development of multicellular structures [1,2]. Polar growth is mediated by polarized transport of vesicles to the apical plasma membrane of the cells. Membrane at the apex is expanded and extracellular components are secreted, while in subapical regions the cell recycles excess membranes. This cyclic movement of secretory vesicles and endosomes in polar growing cells is achieved by a complex protein machinery employing motor proteins and a dynamic actin-cytoskeleton controlled by regulatory small GTPases [3,4,5,6]. Fungal hyphae share with neurons and pollen tubes the distinction of being amongst the most highly polarized cells known in biology [7,8,9]. Nevertheless, the relative impact of evolutionary conserved mechanisms and convergent traits in shaping these cell types is yet unresolved. This is partly a consequence of the fact that studies in different model systems have focused on different aspects of polar tip growth, and the available information has not been integrated to elucidate common as well as distinct features of cell polarity in plants, fungi and animals [1,10,11,12].
While it is assumed that PtdIns(4,5)P 2 -functions in filamentous fungi is similar to those in other eukaryotic models, several aspects complicate such a simplified view. First, in contrast to animals and plants where multiple isoforms of PI4P 5-kinases are common [30,32,43], the genomes of filamentous ascomycetes encode only one gene for a putative PI4P 5-kinase, MSS-4 [32,44]. For instance, in plant pollen tubes several PI4P 5-kinases produce PtdIns(4,5)P 2 with distinct functionalities [33,34,35,40], and such alternative functions would have to be performed by a single enzyme in filamentous fungi. A second aspect refers to possible alternative modes of action of N. crassa MSS-4, which might be mediated by specialized regulatory protein domains. N. crassa MSS-4 contains sequence extensions N-and C-terminal of the kinase domain, which distinguishes the enzyme from mammalian homologs. The additional protein domains of N. crassa MSS-4 are not similar in sequence to those of other PI4P 5-kinases and their functions remain unclear, resulting in a need to characterize the role of N. crassa MSS-4 in the control of polar tip growth as a base for further comparative studies.
Our working hypothesis was that apical localization of a functional enzyme capable of forming PtdIns(4,5)P 2 in fungal hyphae contributes to the establishment of lipid microdomains guiding the machinery for polarized growth. Evidence for a role of N. crassa MSS-4 in the control of polar tip growth derived from a large-scale screen for mutants defective in hyphal growth [45]. However, the function of the mss-4 lesions has not been further characterized. In order to delineate the morphogenetic role of PtdIns(4,5)P 2 in fungal hyphae, experiments with the predicted N. crassa phosphatidylinositolphosphate kinase MSS-4 were performed in vivo in N. crassa and plant pollen tubes as a heterologous expression system as well as in vitro on recombinant protein expressed in E. coli.

Experimental Procedures
Strains, Media and Growth Conditions for N. crassa N. crassa strains used in this study are listed in Table 1 (also see [46]). General genetic procedures and media for N. crassa are available through the Fungal Genetics Stock Center (www.fgsc. net). An mss-4 deletion strain (Dmss-4) having the full-length open reading frame replaced by a hygromycin resistance cassette was generated by the N. crassa genome project [47] and was verified by Southern analysis. Growth rates of fungal strains were determined by measuring radius of colonies on agar plates starting with a wellestablished colony to exclude the lag phase of germination and the initial slow growth phase of a developing colony.
Plasmid Construction and Expression of Tagged Proteins in N. crassa, Pollen Tubes and E. Coli The ORF of N. crassa mss-4 was annotated by the MIPS N. crassa database (http://mips.helmholtz-muenchen.de/genre/proj/ ncrassa) and obtained by PCR from genomic DNA of wild type N. crassa using the oligonucleotide combination 59-GAT CGA ATT CAT GAC CTC TTC GCC AAA TGG AG-39 and 59-GAT CGC GGC CGC TTG ACA CGC CTT GGG GCC GA-39. The PCR-fragment was subcloned into the vector pGEM-Teasy (Promega). The resulting plasmid was designated NcMSS4-pGEM-Teasy. The insert was cloned into the vector pENTR-EYFP [34] as an EcoRI-NotI fragment and transferred by Gateway technology (Invitrogen) into the vector pLatGW (a kind gift from Dr. Wolfgang Dröge-Laser; Julius-Maximilians-University Würzburg, Germany) to generate the mss-4-EYFP fusion construct for expression in tobacco pollen tubes. Pollen tubes were transformed by particle bombardment and cultured as previously described [33]. For the expression of a GFP-MSS4 fusion protein in N. crassa, the mss-4 ORF was amplified from NcMSS4-pGEM-Teasy using the oligonucleotide combination 59-GAT CGG CGC GCC CAT GAC CTC TTC GCC AAA TGG AG-39 and 59-GAT CTC TAG ACT ATG ACA CGC CTT GGG GCC GA-39 and transferred as an XbaI-AscI fragment into pCCG1-N-GFP [48], creating the plasmid pCCG1-N-GFP-MSS4(86-1012). The extended N-terminus as annotated by the N. crassa BROAD database (http://www.broadinstitute.org/annotation/genome/ neurospora/MultiHome.html) was amplified by PCR from genomic DNA of N. crassa with the oligonucleotide combination 59-GAT CGG CGC GCC CAT GCC CTC CTT CGC ACC CGA-39 and 59-GAT CTA CGT AGC ATC GGA GCC GTC C-39 and moved as XbaI-Eco105I fragment into pCCG1-N-GFP-MSS4(86-1012) to generate pCCG1-N-GFP-MSS4(1-1012). For the expression of a GFP-HsPLCd1-fusion protein in N. crassa, the ORF was amplified from HsPLCd1-pGEM-Teasy [33] using the oligonucleotide combination 59-GAT CGG CGC GCC CAT GGA TGA GGA TCT ACA GGC GC-39 and 59-GAT CTC TAG ACT AGA TCT TGT GCA GCC CCA GCA-39 and transferred as an XbaI-AscI fragment into pCCG1-N-GFP [48], creating the plasmid pCCG1-N-GFP-HsPLCd1. N. crassa was transformed as previously described [49]. For the expression of MSS-4 in E. coli, the ORF was amplified from N. crassa cDNA using the oligonucleotide combination 59-GAT CCC ATG GGT ACC TCT TCG CCA AAT GGA G -39 and 59-GAT CCC ATG GCT GAC ACG CCT TGG GGC CGA -39 and transferred as an NcoI-NcoI fragment into the expression plasmid pETM-41 (EMBL Protein Expression and Purification Facility). A. thaliana PIP5K2 was cloned into pETM-41 as previously described [40].

Lipid kinase Assays
Lipid kinase activity was assayed by monitoring the incorporation of radiolabel from c[ 32 P]ATP into defined lipid substrates (Avanti Polar Lipids) as described [50], using total extracts of BL-21-AI expression cultures. For assays using exogenous substrates, 5 mg of lipids were used. Each assay contained 10 mg of total bacterial protein, as determined by the Bradford method [51]. Recombinant expression levels were adjusted between individual cultures according to immunodetection of expressed MBP-tagged proteins. Radiolabeled lipid reaction products were separated by thin-layer-chromatography using silica S60 plates (Merck) and CHCl 3 :CH 3 OH:NH 4 OH:H 2 O (57:50:4:11, by volume) as a developing solvent [52], and visualized using a Fuji HLA-3000 phosphoimager. Substrate conversion was quantified by phosphoimager analysis using a standard curve of c[ 32 P]ATP for calibration of relative phosphate incorporation into lipids.

Microscopy and Imaging
Low magnification documentation of fungal hyphae or colonies was performed as described [53] using an SZX16 stereomicroscope, equipped with a Colorview III camera and Cell D imaging software (Olympus). Images were further processed using Photoshop CS2 (Adobe). Fluorescence microscopy with N. crassa was performed as described [54,55]. An inverted Axio Observer. Z1 (Zeiss) microscope equipped with a QuantEM 512SC camera (Photometrics) and the slidebook 5.0 software (Intelligent Imaging Innovations) were used for image acquisition. Pollen tubes were recorded using either an Olympus BX51 epifluorescence microscope or a Zeiss LSM 510 confocal microscope. BX51: Images were obtained using an F41-028 HQ-Filterset for Yellow GFP (Olympus), an Olympus ColorView II camera and analySIS Docu 3.2 software (Soft-Imaging-Systems GmbH). LSM 510: EYFP was excited at 514 nm and imaged using an HFT 405/514/633 nm major beam splitter (MBS) and a 530-600 nm band pass filter. Fluorescence and transmitted light images were contrast-enhanced by adjusting brightness and c-settings using image-processing software (Photoshop; Adobe Systems).

MSS-4 is Required for Cell Polarization and Polar Tip Extension in N. crassa
As basis for subsequent experiments, N. crassa wild type, mss-4(18-2) and the complemented mutant mss-4(18-2) compl were phenotypically characterized in detail. When grown at restrictive conditions of 37uC, wild type N. crassa formed colonies of approx. 60 mm in diameter within 24 h (Figure 1 A, B). In contrast, the temperature-sensitive mutant mss-4(18-2) did not exhibit substantial growth at 37uC (Figure 1 A, B). mss-4(18-2) compl , which contained an ectopic integration of the MSS-4 ORF and 1 kb of 59 and 39 regulatory sequence, exhibited substantial growth. However, it was clearly slower than that of the wild type control (Figure 1 A, B). The hyphal morphology of mss-4(18-2) shifted to restrictive conditions was characterized by an abnormal cell shape, increased branching of hyphae, and subsequently cessation of tip extension and cell lysis (Figure 1 C). mss-4(18-2) compl exhibited only weakly increased branching of hyphae (Figure 1 C). mss-4(18-2) conidiospores germinating at 37uC displayed contorted and swollen cell morphology and increased branching, which resulted in highly restricted growth and the formation of compact colonies. mss-4(18-2) was still able to establish hyphae (Figure 1 D). In contrast, a N. crassa strain in which the full-length MSS-4 ORF was replaced by a hygromycin resistance cassette (Dmss-4) could only be maintained as a heterokaryon, having the mutation sheltered by the presence of nuclei containing a wild type copy of mss-4. Most uninucleate ascospores of Dmss-4 germinated only in an apolar manner (Figure 1 E), while those few that were able to establish germ tubes formed highly apolar and contorted growing cells that ultimately lysed (Figure 1 F), indicating that mss-4 is an essential gene.

N. crassa MSS-4 is a PI4P 5-kinase
In order to characterize the biochemical properties of MSS-4, we expressed recombinant protein in E. coli and then tested extracts in vitro for PI-kinase activity. In the course of a phylogenetic comparison of fungal, plant and animal PI4P 5kinases [32], we had noted that the gene model presented for NCU02295/MSS-4 in the MIPS database might be ambiguous and that a possible alternative start codon exists 255 bp upstream of the originally proposed one. As in vivo data obtained later in the course of this study suggest the earlier start codon as the relevant one, all numbers identifying residues in MSS-4 in this study refer to the extended gene product of 1012 amino acids in length. Because MSS-4(1-1012) expressed poorly in E. coli, the biochemical characterization was carried out with MSS-4(86-1012), the variant originally proposed in the MIPS database. This approach is in line with previous reports, which confirm that PI4P 5-kinases do not require additional domains for catalytic activity and phospholipid specificity [34,37,40]. MSS-4(86-1012) was capable of phosphorylating both PtdIns3P and PtdIns4P to PtdIns(3,5)P 2 and PtdIns(4,5)P 2 , respectively (Figure 2 A, B). PtdIns5P was not a substrate for MSS-4(86-1012), but was converted to PtdIns(4,5)P 2 with high efficiency by a recombinant human PI5P 4-kinase [56] used as a positive control (Figure 2 A, B). When recombinant MSS-4(86-1012) was presented with equimolar amounts of PtdIns3P and PtdIns4P, the enzyme clearly preferred PtdIns4P, resulting in the predominant formation of PtdIns(4,5)P 2 (Figure 2 C). These data indicate that the N. crassa MSS-4-gene encodes an active PI4P 5-kinase.
Our work was based on the hypothesis that MSS-4 generates PtdIns(4,5)P 2 required for polar tip growth of hyphae. To test this hypothesis, the intrinsic PI4P 5-kinase activity was determined in membranes purified from wild type, mss-4(18-2) and mss-4(18-2) compl grown at permissive and at restrictive conditions. While we detected PtdIns(4,5)P 2 formation in membrane extracts of all three strains grown at 16uC, the mss-4(18-2) mutant was devoid of PI4P 5-kinase activity at 37uC (Figure 3 A). The data indicate that the cell polarity defects of mss-4(18-2) correlated with defective PI4P 5-kinase activity, and that ectopic expression of a genomic fragment encoding for wild type MSS-4 partially restored both hyphal growth and PI4P 5-kinase activity. Based on these observations, we conclude that the formation of PtdIns(4,5)P 2 is necessary for cell polarization and tip growth in N. crassa and is mediated by MSS-4. Nevertheless, membranes purified from mss-4(18-2) compl grown at 37uC displayed substantial reduced PI4P 5kinase activity compared the wild type control (2167 versus 4268 fmol min 21 mg 21 protein, respectively; n = 3). This may indicate that the ectopic integration of MSS-4 may affect the expression level of the kinase. Alternatively, a dominant negative effect of the mutant MSS-4 variant is also possible.

A highly Conserved Tyrosine is Required for the Function of Plant and Fungal PI4P 5-kinases
The conditional phenotype of mss-4(18-2) suggested that the mutant encoded for a temperature-sensitive variant of MSS-4, which enabled growth at permissive temperature and precluded growth at prohibitive temperatures. Currently available conditional alleles encoding PI4P 5-kinases carry multiple point mutations, which are not readily applicable to enzymes from different biological sources. For instance, a temperature-sensitive variant of Mss4p from Saccharomyces cerevisiae [57] carries the changes Q351L, M392T, H532Q and K615R, of which only H532 is conserved in PI4P 5-kinases from plants, but none in human or animal sequences. A new temperature-sensitive PI4P 5kinase allele would be a powerful tool to enable future studies on the roles of PtdIns(4,5)P 2 in other biological systems. Therefore, we sequenced the genomic loci of mss-4(18-2) and of mss-4(34-10), another previously identified conditional mss-4 strain [45]. Both mutants contained only a single point mutation in nucleotide position 2248 (T to A) that resulted in the substitution of tyrosine 750 to asparagine (Figure 4). Position 750 of MSS-4 resides in the C-terminal portion of the catalytic domain (Figure 4 A). An alignment of MSS-4 with phosphoinositide kinases from human, Drosophila melanogaster, A. thaliana and S. cerevisiae (Figure 4 B) indicated that this tyrosine residue is strictly conserved among PI4P 5-kinases. To our knowledge, the invariant tyrosine residue has not been reported as critical for catalytic activity of PI4P 5kinases. To determine if this conserved tyrosine might be part of Figure 1. Phenotypic characterization of N. crassa mss-4. The growth of N. crassa wild type, the mss-4(18-2) mutant [45], and a complemented strain mss-4(18-2) compl ectopically expressing wild type MSS-4 on solid media was monitored by macroscopic examination after 24 h at 37uC (A). The increase in colony diameters of growing cultures over time was determined (B). Errors (SD) are too small to be seen at the scale given. The hyphal morphologies of wild type, the mss-4(18-2) mutant or mss-4(18-2) compl grown at room temperature or 2h and 8 h after shifting to 37uC, were microscopically characterized, as indicated (C). Germination of wild type (D) and mss-4(18-2) conidia was documented after 18 h at 37uC (E). Note the highly contorted and abnormal morphology of the mss-4(18-2)-strain. A strain carrying a deletion of the entire MSS-4 locus (Dmss-4) failed to establish hyphae upon germination from ascospores (black) and, in contrast to wild type (F) was not viable (G, H). Experiments were repeated independently five times with similar results. doi:10.1371/journal.pone.0051454.g001 the catalytic center of PI4P 5-kinases, the corresponding position of the human PIP-kinase IIb (Y403) was highlighted in the available 3D-structure [56] (Figure 4 C). Based on this structural information, Y403 is located in immediate proximity of the substrate-binding pocket, opposing the position of the ATP cosubstrate. When we assume that MSS-4 shares the overall organization of the catalytic center with the human PIP-kinase IIb, the sequencing of two mss-4 mutant loci identified a conserved tyrosine residue in the substrate-binding pocket as important for PI4P 5-kinase function.
To verify this notion, we analyzed the involvement of this conserved tyrosine in the temperature-sensitivity of N. crassa MSS-4 or other PI4P 5-kinases in more detail. Thus, the Y750N exchange resulted in drastically reduced catalytic activity rather than in temperature-dependent functionality of the mutated enzyme. In order to test whether the conserved tyrosine defined by the two N. crassa mss-4 mutants was also important for the catalytic activity of PI4P 5-kinases from other organisms, an analogous variant of the A. thaliana PI4P 5kinase was generated. When the PI4P 5-kinase activity of recombinant AtPIP5K2(Y738N) was compared with recombinant parental AtPIP5K2 at either 16uC or at 37uC, AtPIP5K2(Y738N) was found to be fully inactive at both temperatures (Figure 4 D), whereas wild type AtPIP5K2 was active at 16uC and at 37uC (759695 and 332612 fmol min 21 mg 21 protein, respectively; n = 3). Overall, these data indicate that the temperature-sensitive growth behavior of the N. crassa mss-4 mutants is a result of minimal residual activity of MSS-4(Y750N), which seems sufficient to support growth at 16uC, but not at 37uC.

N. crassa MSS-4 Localizes as a Subapical Membraneassociated Ring and Along Constricting Septa
The data obtained so far suggested that MSS-4 regulates PtdIns(4,5)P 2 -levels required for apical tip growth. In line with this assumption, we determined the subcellular distribution of PtdIns(4,5)P 2 in vivo using the specific reporter PLCd1-PH-EYFP [33,58]. In germinating conidia, the reporter fluorescence was associated with the plasma membrane (  Images represent phosphoimager signals of radiolabeled lipids. Migration of lipid products was interpreted according to comigration with unlabeled authentic standards as previously described [71]. Formation of PtdIns(4,5)P 2 in the indicated N. crassa strains (A). The image is representative for three independent experiments. Phosphoimages presented in (A) were quantified (B). Specific activities were calculated using total protein content in the recombinant extracts. Data represent means 6 SD from three independent experiments. doi:10.1371/journal.pone.0051454.g003 consistently observed, and reporter-fluorescence more often appeared evenly distributed along the entire plasma membrane. In addition to general plasma membrane association of the PtdIns(4,5)P 2 -reporter, PLCd1-PH-EYFP appeared to stronger label emerging branches and constricting septa (Figure 5 D, E). In subapical regions of the hypha, the PtdIns(4,5)P 2 -reporter still associated with the plasma membrane and with septa and was also found associated with cytosolic puncta (Figure 5 E). Reassured by the localization of PtdIns(4,5)P 2 at the plasma membrane and its increased abundance at sites of active growth (i.e. septum and hyphal tip), we determined the subcellular localization of MSS-4 by expressing a GFP-MSS-4 fusion protein in N. crassa. GFP-MSS-4 localized in a ring-like membrane domain immediately behind the apex of growing hyphae (Figure 6 A). Plasma membrane localization appeared in spots rather than a continuous, smooth signal, possibly indicating association with membrane microdomains (insert in Figure 6 A). GFP-MSS-4 also decorated the plasma membrane in regions distal of the apical ring, but fluorescence was much weaker than at the apical domain. Moreover, we observed cytosolic fluorescence and association with cytosolic filaments or endomembranes of unknown nature in addition to the apical plasma membrane localization, indicative of a portion of the enzyme present in a soluble state and associated with intracellular structures (Figure 6 A). This observation is in line with previous reports on the dynamic subcellular distribution of PI4P 5-kinases, which peripherally associate with membranes [56]. Subapical plasma membrane association of PI4P 5-kinases of the subfamily B from A. thaliana requires the presence of N-terminal regulatory domains other than the catalytic domain [36,37,40]. As N. crassa MSS-4 can also be categorized as a PI4P 5-kinase of subfamily B [32] we tested whether a deletion of the first 85 amino acids, as proposed by the alternative gene model for MSS-4, would impair membrane localization of MSS-4. When GFP-MSS-4(86-1012) was expressed in N. crassa, the fusion protein accumulated primarily in the cytoplasm and associated with intracellular filaments and/or endomembranes (Figure 6 B). The plasma membrane localization was strongly reduced. Reciprocally, we also expressed two fragments of MSS-4 that contained the Nterminal region, but neither GFP-MSS-4(1-85) nor GFP-MSS-4(1-355) displayed membrane association or apical localization (data not shown). The failure of such smaller N-terminal fragments to localize has previously been reported from equivalent experiments using plant PI4P 5-kinases and might be a result of misfolding or the requirement of additional domains for proper membrane association of PI4P 5-kinases [40]. The requirement for the Nterminal domain (Figure 6 B) indicated similarities in the modes of plasma membrane recruitment of N. crassa MSS-4 with A. thaliana PI4P 5-kinases [36,37,40]. To further substantiate this notion, N. crassa MSS-4-EYFP was transiently expressed in tobacco pollen tubes (Figure 6 C). The observed subapical membrane association of the heterologously expressed kinase was similar to the localization in the native fungal context, but also patterns reported for plant PI4P 5-kinases from A. thaliana [33,35] or tobacco [40] (Figure 6 D, A. thaliana PIP5K5 as an example). The data indicate that localization of PI4P 5-kinases, such as MSS-4 and its plant homologs as a subapical membrane-associated collar is conserved across kingdoms and can be found in plant pollen tubes and root hairs as well as in fungal hyphae.
Besides similarities of GFP-MSS-4 localization in tip growing cells, some hyphal-specific localization patterns were observed. Fungal septa represent cell walls and associated plasma membrane that partition growing hyphae into smaller cytosolic sections. GFP-MSS-4 localized to sites initiating septum formation, decorating the inner edges of the constricting ring of cell wall ( Figure 6 E; Movie S1). Another example for a process known in filamentous ascomycetes, but not for tip-growing plant cells, is intercellular communication and subsequent fusion to establish a mycelial network [59,60,61,62]. N. crassa germling fusion involves the fusion of specialized cell protrusions that have been termed conidial anastomosis tubes (CATs). CATs are characterized by positive tropic growth. When cell communication of N. crassa CATs was investigated, GFP-MSS-4 was strongly enriched at both tips of mutually attracting CATs that concentrated at the contact sites of two fusing cells (Figure 7;. Movie S2). Moreover, upon initiation of cell fusion, the accumulated GFP-MSS-4 expanded concentrically around the expanding fusion pore. Together, these localization patterns observed for GFP-MSS-4 in N. crassa suggest a central role for PtdIns(4,5)P 2 -formation at sites of polar growth at the hyphal apex and constricting septa. During hyphal fusion MSS-4 might contribute to membrane rearrangements required for the establishment of cytosolic contacts between fusing cells.

Discussion
In order to delineate what perturbations of PI-metabolism might contribute to the phenotypes of mss-4(18-2) and Dmss-4, the biochemical properties of MSS-4 must be considered. When PtdIns3P or PtdIns4P were tested individually as substrates for recombinant MSS-4(86-1012), the enzyme converted both substrates equally well. This contrasts the situation of plant PI4P 5kinases, which exhibit high specificity towards PtdIns4P [33,34,37,52,63]. PtdIns5P was not a substrate for MSS-4(86-1012), in line with the presence of a glutamate residue in position 723, which is implicated in specifying PI4P 5-kinases [64]. The notion that MSS-4(86-1012) can produce PtdIns(3,5)P 2 or PtdIns(4,5)P 2 from PtdIns3P and PtdIns4P, respectively, indicates a low degree of substrate discrimination during the production of PtdIns-bisphosphates in vitro. Therefore, we cannot rule out that PtdIns(3,5)P 2 , PtdIns(4,5)P 2 , or both lipids are formed by MSS-4 in vivo and that both lipids might be required for hyphal growth. However, a functional role for PtdIns(4,5)P 2 rather than PtdIns(3,5)P 2 is supported by the observations that (i) PtdIns4P was preferentially used in in vitro competition experiments, resulting in the formation of PtdIns(4,5)P 2 , (ii) membrane fractions purified from mss-4(18-2) growth at 37uC displayed drastically reduced PI4P 5-kinase activity, and (iii) growth defects manifested despite the presence of a putative PI3P 5-kinase encoded by the locus NCU02083, which is similar to the yeast PI3P 5-kinase, Fab1p, [32] and likely capable of generating PtdIns(3,5)P 2 in the mss-4(18-2) background. Thus, our data strongly suggest that the reduced formation of PtdIns(4,5)P 2 is the reason for the cell polarity defects of the N. crassa mss-4 mutants.
The conditional N. crassa mss-4 strains allowed the identification of an invariant tyrosine, which is critical for catalytic activity of MSS-4 and related PI4P 5-kinases. Mapping of the homologous Y403 in the 3D-structure of the human PIP-kinase IIb [56] places this residue in close proximity to the substrate-binding pocket within the active site. The position of this conserved Y750 suggests that it is likely involved in catalysis by binding or orienting the substrate in MSS-4. An involvement in catalysis rather than conditional activity of this substitution is also consistent with the Figure 5. Distribution of a high-affinity fluorescent probe for PtdIns(4,5)P 2 in N. crassa hyphae. The fluorescence distribution of PLCd1-PH-EYFP was imaged in transgenic N. crassa hyphae. The reporter decorated the plasma membrane at different developmental stages, including germinating conidia (A), and accumulated at sites of hyphal branching or emergence (B). Reporter fluorescence was also present in the plasma membrane of hyphal tips (C). The reporter associated strongly with constricting septa (D). Decoration of septa (open arrowheads) and emerging hyphal tips (closed arrowhead) could be observed in the same cell (D). Overall plasma membrane association of the reporter, strong association with septa and cytosolic puncta were observed also in older, larger hyphae (E). Bars, 5 mm (A-C; F); 20 mm (E). doi:10.1371/journal.pone.0051454.g005 fact that a conditional MSS-4 variant in S. cerevisiae contained multiple, non-conserved amino acid substitutions [57]. The conclusion is also in line with our finding of close-to-normal PtdIns(4,5)P 2 levels in extracts of mss-4(18-2) grown at 16uC, while PtdIns(4,5)P 2 at 37uC was drastically reduced. In summary, these data suggest that poor catalytic activity of MSS-4(Y750N) in combination with increased turnover of PtdIns(4,5)P 2 at 37uC may be the cause for the observed cell polarity defects in mss-4 .
The morphological defects observed in the two mss-4 mutants partially resemble phenotypic characteristics of mutant plant root hairs [36,37] or pollen tubes [33,35,65]. For instance, pollen tubes of A. thaliana pip5k6 single or pip5k4 pip5k5 and pip5k10 pip5k11 double mutants exhibit impaired germination, reduced tip expansion and increased sensitivity to the actin depolymerizing drug, latrunculin B [33,34,35,66]. Since plasma membrane domains containing PtdIns(4,5)P 2 as well as other lipids, such as sphingolipids or sterols [15,16] define the sites of polarized expansion, the functionality of such membrane domains, and thus, polar tip extension are compromised if one or more lipid constituents are missing. The increased branching rates of N. crassa mss-4(18-2) may be a consequence of mislocalized plasma membrane domains, leading to recruitment of the machinery for cell expansion to multiple sites rather than one apical domain as in wild type. Alternatively, increased hyphal branching in mss-4  might be a result of altered biophysical properties of the cell wall or effects on intracellular signaling [32].
PtdIns(4,5)P 2 was associated with the plasma membrane in N. crassa hyphae. Reporter fluorescence was found along the entire plasma membrane and was, thus, different from that previously observed in yeast challenged by mating pheromone [29]. Increased reporter-fluorescence at new branching sites (cf.  PtdIns(4,5)P 2 resembled that of MSS-4, which was also associated with sites of active growth, with the notable exception that PtdIns(4,5)P 2 was detected also in the extreme apex of the cells (cf. Figure 5 C), whereas MSS-4 was abesnt from the very apex (cf. Figure 6 A). A possible explanation for the extended localization of PtdIns(4,5)P 2 is that it diffuses laterally from sites of its biogenesis either as a free lipid or bound to target proteins in or at the membrane. The subcellular distribution of MSS-4 in the hyphal context closely resembles patterns reported for PI4P 5-kinases in plant pollen tubes. The membrane domain decorated by MSS-4 in N. crassa hyphae corresponds to the zone of endocytosis in fungal hyphae [67,68,69] and in pollen tubes [70]. Expression of GFP-MSS-4(86-1012) resulted in loss of subapical plasma membrane association and accumulation in the cytosol and at undefined endomembranes, indicating a role for the N-terminal 85 amino acids in mediating membrane recruitment of MSS-4. Interestingly, both variants of MSS-4 (MSS-4(1-1012) and MSS-4(86-1012)) complemented the mss-4(18-2) defects (data not shown). At this point it remains unclear whether MSS-4(1-1012) and MSS-4(86-1012) represent alternative gene products of MSS-4 that differ in their subcellular distribution, possibly as part of mechanisms that control subcellular levels of PtdIns(4,5)P 2 . Alternatively, complementation of nonfunctional MSS-4(86-1012) in mss-4(18-2) might be the result of heterodimerization with endogenous MSS-4(Y750N) and correct localization by a piggyback mechanism.
Recently, a non-conserved N-terminal domain (NTD) was characterized as essential for membrane association of plant PI4P 5-kinases of subfamily B [40]. The similar localization patterns of native and heterologously expressed N. crassa MSS-4-EYFP and plant PI4P 5-kinases [33,35,37] in tobacco pollen tubes suggest a similar mechanism for controlling membrane recruitment of N. crassa MSS-4(1-1012) and plant PI4P 5-kinases via N-terminal protein domains. It has previously been proposed that PI4P 5kinases are recruited to their target membranes by protein-protein interactions, possibly through their N-terminal domains. While Nterminal domains of N. crassa MSS-4 and plant PI4P 5-kinases are not similar in sequence, they may nonetheless enable interaction with other partners of a conserved recruitment machinery in the respective heterologous system. Alternatively, it is possible that heterodimerization of MSS-4 with plant-endogenous PI4P 5kinases recruited the N. crassa enzyme to the apical plasma membrane in tobacco pollen tubes.
Overall, the data presented in this study indicate that tip growth in the filamentous ascomycete N. crassa requires the formation of PtdIns(4,5)P 2 by the PI4P 5-kinase MSS-4 at the hyphal tip and, possibly, during septation to support polar growth. The patterns of MSS-4 localization in living N. crassa cells suggest a mode of action for PtdIns(4,5)P 2 that is similar to that proposed for polar growing cells from other biological models, such as plant pollen tubes or root hairs. Structural similarities between plant root hairs or pollen tubes and fungal hyphae are, thus, accompanied by similarities in regulatory membrane lipids contributing to the control of polar tip growth.

Supporting Information
Movie S1 Time-course of GFP-MSS-4 localization during septum formation. GFP-MSS-4 formed cortical rings at incipient septation sites that constricted during septum formation and accumulated around the septal pore of the completed septum.