1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Ustilago maydis, the causal agent of corn smut disease, displays dimorphic growth in which it alternates between a budding haploid saprophyte and a filamentous dikaryotic pathogen. We are interested in identifying the genetic determinants of filamentous growth and pathogenicity in U. maydis. To do this, we have taken a forward genetic approach. Previously, we showed that haploid adenylate cyclase (uac1) mutants display a constitutively filamentous phenotype. Mutagenesis of a uac1 disruption strain allowed the isolation of a large number of budding suppressor mutants. These mutants are named ubc, for Ustilagobypass of cyclase, as they no longer require the production of cAMP to grow in the budding morphology. Complementation of one of these suppressor mutants led to the identification of ubc3, which is required for filamentous growth and encodes a MAP kinase most similar to those of the yeast pheromone response pathway. In addition to filamentous growth, the ubc3 gene is required for pheromone response and for full virulence. Mutations in the earlier identified fuz7 MAP kinase kinase also suppress the filamentous phenotype of the uac1 disruption mutant, adding evidence that both ubc3 and fuz7 are members of this same MAP kinase cascade. These results support an important interplay of the cAMP and MAP kinase signal transduction pathways in the control of morphogenesis and pathogenicity in U. maydis.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Ustilago maydis, the causal agent of corn smut disease, has evolved a life cycle dependent upon a dimorphic switch between a saprophytic haploid yeast phase and an infectious dikaryotic filamentous form. To generate the filamentous dikaryon, two compatible haploid yeast strains, differing at two mating-type loci (termed a and b), must mate (for a recent review, see Banuett, 1995; Kahmann et al., 1995; Kronstad and Staben, 1997). Generally, the a and b loci are involved in pre- and post-mating events respectively. Thus, if the dikaryon, formed by virtue of mating between haploid strains of opposite a mating specificity, possesses nuclei with different b specificities, it grows in a filamentous morphology, is pathogenic (Banuett, 1995; and references therein) and further mating is precluded (Laity et al., 1995). Because of its relevance to the work presented here, the a locus function is described in greater detail below.

The a locus possesses mfa and pra, two tightly linked genes that encode secreted pheromone and membrane-spanning pheromone receptors respectively (Bölker et al., 1992). In addition to its function as a mating attraction system, dikaryon heterozygosity at the a locus also contributes to the production (together with heterozygosity at b) of the filamentous growth form through an autocrine response in which the pheromones and receptors of opposite allelic specificity are present within the same cell and therefore may continually interact (Banuett and Herskowitz, 1989; Spellig et al., 1994). Genetic and biochemical data indicate that the interactions between the U. maydis pheromones and receptors are similar to the events in the Saccharomyces cerevisiae paradigm (Banuett, 1998). The pheromone encoded by the mfa gene is thought to interact directly with the pheromone receptor product encoded by the pra gene of the opposite a mating specificity (Spellig et al., 1994). The downstream events generating the final response to pheromone presumably involve components similar to those encountered in S. cerevisiae. In S. cerevisiae, signal transduction from the pheromone–receptor interaction to the final cellular responses involves trimeric G proteins and a MAP kinase cascade with the final phosphorylation of the Ste12p transcription factor, which in turn regulates transcription of target genes (Banuett, 1998).

In addition to the a locus, several U. maydis genes orthologous to those in the pheromone response pathway of S. cerevisiae have been described. Four G protein α subunit genes have been identified and one of these genes, designated gpa3, is required for the pheromone response (Regenfelder et al., 1997). Additionally, Δgpa3 mutants display an elongated morphology (Regenfelder et al., 1997) reminiscent of adenylate cyclase (uac1) mutants (Gold et al., 1994a). Both the elongate morphology and the sterility phenotypes of the Δgpa3 mutant are remedied by addition of exogenous cAMP, which was interpreted to indicate that cAMP signalling is a prerequisite for pheromone response (Kruger et al., 1998). The fuz7 gene, cloned by virtue of homology to the yeast ste7 gene, encodes a MAP kinase kinase required for U. maydis pheromone-responsive morphogenesis (Banuett and Herskowitz, 1994). The high mobility group (HMG) protein encoded by prf1 appears to play the role of the downstream transcription factor and is required for expression of the genes at the a and b loci (Hartmann et al., 1996).

In addition to the triggering of filamentous growth via the action of the a and b loci, we have shown that a cAMP signal transduction pathway is critical for the maintenance of budding growth of the U. maydis wild-type haploid (Gold et al., 1994a). Mutants defective in the gene encoding adenylate cyclase (uac1) are converted to a constitutive filamentous but non-pathogenic phenotype (Barrett et al., 1993; Gold et al., 1994a). Several genes playing roles in dimorphism, in addition to uac1, have been identified by complementation of mutations that suppress the filamentous phenotype of the uac1 mutant (Gold et al., 1994a; Mayorga and Gold, 1998). Thus, these genes are characterized as influencing a morphogenetic response to cAMP signal transduction in U. maydis. These suppressor mutations are called ubc for Ustilagobypass of cyclase. Four ubc genes have been isolated and sequence homology data have suggested the biochemical functions for three of these genes. The ubc1 gene encodes the regulatory subunit of the cAMP-dependent protein kinase (PKA) and is required for gall formation (Gold et al., 1994a, 1997). The ubc3 and ubc4 genes encode components of a MAP kinase cascade, whereas ubc2 has an as yet undetermined function (Mayorga and Gold, 1998; M. E. Mayorga and S. E. Gold, unpublished).

Here, we describe the discovery that the ubc3 gene encodes a MAP kinase from U. maydis that plays a critical role in the dimorphic switch. Additionally, evidence is presented that the ubc3 mutation reduces the fungus's ability to form a visible filamentous dikaryon in plate mating assays. Cytoduction assays indicated that the ubc3 mutation reduced, but did not negate, cytoplasmic fusion in mating reactions. Drop mating assays suggested that ubc3 mutants were compromised in both responding to and transmitting mating pheromone signals. Pathogenicity assays with uac1 ubc3 double mutants indicated a synergistic reduction in pathogenicity compared with single mutants. However, dikaryons mutant only at ubc3 were pathogenic at near normal levels. This work establishes a strong link between the cAMP and MAP kinase pathways in the control of dimorphism in U. maydis. Finally, mutations in fuz7 were found to act to suppress the uac1::ble filamentous phenotype in a way that was indistinguishable from a ubc3 mutation and thereby provide corroborating evidence that ubc3 and fuz7 are members of the same MAP kinase pathway.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Cloning and analysis of the ubc3 gene

The original uac1::ble ubc3-1 mutant, strain 3/25, was isolated by screening for spontaneous suppressor mutants of the constitutively filamentous uac1 disruption mutant, strain 1/9, as previously described (Mayorga and Gold, 1998). The mutation was then complemented using a previously constructed U. maydis cosmid library (Barrett et al., 1993) and the complementing cosmid (cosubc3) was recovered (Mayorga and Gold, 1998). Two additional uac1 ubc3 double mutant strains (4/14, uac1::ble ubc3-2; 4/25, uac1::ble ubc3-3) were similarly obtained as spontaneous mutants complemented by cosubc3. Fig. 1 details the morphology of liquid-grown cells of the uncomplemented uac1::ble ubc3-1 mutant strain 3/25 (Fig. 1A) as well as the complemented mutant (Fig. 1B).


Figure 1. . Morphological complementation of the filamentation suppressor mutation ubc3-1. A and B. Cells grown in YEPS supplemented with 150 μg ml−1 of hygromycin B. Bar represents 50 μm. A. uac1::ble ubc3-1 mutant strain 3/25 transformed with selectable phyg101 vector plasmid control. B. uac1::ble ubc3-1 mutant strain 3/25 transformed with the minimal ubc3 complementing clone pTP4F4a.

Download figure to PowerPoint

The ubc3 gene was subcloned from cosubc3 by a functional assay in which the multiple budding uac1::ble ubc3-1 double mutant was converted to the filamentous phenotype of the single uac1::ble mutant (Mayorga and Gold, 1998). Co-transformation experiments in which cosubc3 was digested to completion with various restriction enzymes indicated that the complementing activity was localized to a 7.5 kb XbaI fragment (Fig. 2). After a series of nested deletions, a 3.5 kb deletion clone, designated pTP4F4a (Fig. 2), was found to be the minimal fragment complementing the ubc3-1 mutation and was chosen for sequencing.


Figure 2. . Subcloning and deletion series to determine the minimal complementing fragment of cosubc3 in cosmid vector pJW42. A 7.5 kb fragment containing the functional ubc3 gene from cosubc3 was subcloned into the selectable vector plasmid phyg101, pSMBUBC3-1 (+). A deletion series was carried out to obtain the minimal complementing clone pTP4F4a (+). The putative ubc3 open reading frame is shown as the dark box. The translational initiation codon (ATG) is shown and the putative orientation of the ORF is indicated by the arrow. The region of ubc3 deleted and replaced with the nourseothricin (NAT) resistance gene in Δubc3 strains is shown as a light box. Plasmid pCLAD1, generated by digestion of pSMBUBC3-1 with ClaI followed by self-ligation, was no longer capable of complementation (−). Note that in pSMBUBC3-1 the 0.5 kb XbaI/ClaI fragment is not co-linear with the 3.5 kb ClaI/XbaI fragment as it is after the deletion in pCLAD1. Vector restriction sites shown are: B, BamHI; K, KpnI; and N, NotI. cosubc3 and ubc3 restriction sites are: X, XbaI; and C, ClaI.

Download figure to PowerPoint

A 354-amino-acid open reading frame was identified. A TATA box-like element (TATTAA) was found 916 bases upstream from the putative translational initiator ATG. No polyadenylation signal was detected in the 210 bp of sequence 3′ of the putative stop codon. A blast search (Altschul et al., 1990) with the putative amino acid sequence of ubc3 indicated very strong homology to members of the MAP kinase family. Fig. 3 compares the putative amino acid sequence encoded by the ubc3 gene with several other MAP kinases. The gene with the highest similarity was the Nectria haematococca FsMAPK gene (Li et al., 1997). The shared amino acid sequence identities between U. maydis Ubc3 and the sequences of several other MAP kinases are: 74% for the Nectria haematococca FsMAPK (Li et al., 1997), 72% for the Magnaporthe grisea PMK1 (Xu and Hamer, 1996), 71% for the Pneumocystis carinii MKP2 (A. G. Smulian, unpublished; GenBank accession no. AF077548), 59% for the Candida albicans CEK1 (Whiteway et al., 1992), and 56% for the Saccharomyces cerevisiae FUS3 (Elion et al., 1990). Phylogenetic analysis of the ubc3 encoded protein and yeast MAP kinases places Ubc3 on the same clade as yeast MAP kinases involved in the pheromone response (Fig. 3B).


Figure 3. . Comparison of Ubc3 to other MAP kinases. A. Alignment of the putative amino acid sequence encoded by the ubc3 gene to several MAP kinases. The sources of these sequences are as follows: Nectria haematococca, FsMAPK (Li et al., 1997); Magnaporthe grisea, PMK1 (Xu and Hamer, 1996); Candida albicans, CEK1 (Whiteway et al., 1992); Pneumocystis carinii, MKP2 (A. G. Smulian, unpublished; GenBank accession no. AF077548); and Saccharomyces cerevisiae FUS3 (Elion et al., 1990). B. Phylogenetic comparison of Ubc3 to MAP kinases of various functions from S. cerevisiae and S. pombe.

Download figure to PowerPoint

Deletion of ubc3 suppresses filamentation of uac1 mutants

Plant inoculations with the spontaneous uac1 ubc3 double mutants crossed with the wild type were carried out to attempt to genetically separate the ubc3 from the uac1 mutation as previously carried out with the uac1 ubc1 double mutants (Gold et al., 1997). These crosses, however, generated very weak reactions and no progeny were recovered. Deletion mutants in the ubc3ubc3) gene were generated. Two Δubc3 strains of opposite mating type were designated 2/58 and 2/59 (Table 1). Crosses of Δubc3 mutant strain 2/59 (a2b2 ubc3::nat ) with strain 1/9 (a1b1 uac1::ble) generated double mutant progeny resistant to both nourseothricin and phleomycin (Gold et al., 1994b). Southern blotting data (not shown) confirmed that these strains possess both disruption mutations. Double uac1 ubc3 mutants display a ‘frosty’ phenotype (Mayorga and Gold, 1998) that interferes with the determination of their mating type by standard mating reactions. PCR and Southern analyses (data not shown) were used to determine the allelic specificity at the a and b loci of these progeny. These uac1::bleΔubc3 progeny displayed a phenotype that is indistinguishable from the original spontaneous uac1::ble ubc3-1 mutant (compare Fig. 4D with Fig. 1A). Thus, the ubc3 gene is required for the filamentation of haploid strains lacking the ability to synthesize cAMP.


Figure 4. . Mutation of ubc3 or of fuz7 acts as a specific suppressor of filamentous growth. A–F. Cells grown in liquid YEPS overnight at 30°C with constant shaking. Bar represents 50 μm. A. Wild-type haploid (1/2). B. uac1::ble haploid (1/9). C. Δubc3 haploid (2/58). D. Δubc3 uac1::ble haploid (2/60). E. Δfuz7 haploid (2/38). F. Δfuz7 uac1::ble haploid (2/76).

Download figure to PowerPoint

The Δubc3 single mutant was morphologically identical to the wild-type haploid (compare Fig. 4C with Fig. 4A). This was as expected because previous results showed that the introduction of a functional uac1 gene into the uac1::ble ubc3-1 double mutant (3/25) generated a strain morphologically indistinguishable from the wild type (Mayorga and Gold, 1998). In addition, growth rates of the Δubc3 mutants are indistinguishable from the wild type. Interestingly, the fungus's ability to grow in the filamentous morphology at low pH (pH 3) (Ruiz-Herrera et al., 1995) was abolished in a Δubc3 mutant background (A. D. Martinez and S. E. Gold, unpublished).

The role of the ubc3 MAP kinase gene in pheromone response

Because its structure suggested a role in pheromone response, the effect of the ubc3 mutation on cell fertility was examined using several methods. Fig. 5 presents plate mating assays that were carried out on charcoal mating medium (Holliday, 1974). The visible white filamentous growth indicative of the dikaryon was significantly reduced in a ubc3 gene dose-dependent manner. In matings between compatible wild-type strains (1/2 × 2/9), very strong white filamentous growth was observed 24 h after co-spotting. However, when one strain was deleted for ubc3 (1/2 × 2/59 or 2/9 × 2/58), an easily visible but significantly reduced reaction was observed. Finally, if both mating partners were deleted for ubc3 (2/58 × 2/59), a weak visible mating reaction was apparent after 24 h. This result indicates that ubc3 mutants are blocked in efficient formation of filamentous dikaryons by defects in either pre- or post-mating events.


Figure 5. . Effect of the ubc3 mutation on mating. Spots of 20 μl mating mixtures were placed on charcoal-containing medium after mixing equal volumes of paired cell types that had been grown overnight at 30°C in YEPS. Each spot is a combination of two strains, one listed in the row and the second in the column. Mating reactions were photographed 24 h after plating.

Download figure to PowerPoint

To distinguish between pre- and post-mating points of blockage to dikaryon filament formation, two assays for defects in premating events were carried out. First, a cytoduction assay (Trueheart and Herskowitz, 1992; Laity et al., 1995) was used to determine whether the dikaryon was unable to form because of an inability to fuse. This assay measures cytoplasmic exchange by simultaneous selection of a nuclear marker in the recipient strain and a mitochondrial marker from the donor. Thus, only recipient cells that have fused with the donor and have obtained mitochondria by cytoplasmic exchange are obtained on selection plates. Table 2 depicts the relative effect on fusion in matings with one or both strains possessing the ubc3 null allele. When both the recipient and donor strains contain the Δubc3 allele (Table 2, 2/72 × 2/69), cytoplasmic fusion was reduced approximately 20-fold compared with the wild-type fusion (Table 2, 2/18 × 1/8). No cytoduction was detected when recipient and donor strains had the same a alleles (Table 2, 2/72 × 2/71). When a Δubc3 strain was mated to a ubc3+ strain (Table 2, 2/72 × 1/8), fusion was only reduced about twofold. These results indicate that mutations in ubc3 decreased but did not negate fusion.

Table 2. . Δubc3 cytoduction. a. Proportion of benomyl-resistant colonies that are oligomycin resistant.b. Positive control.c. Negative control.d.Δ indicates a deletion mutation.e. No cytoductants detected.Thumbnail image of

Figure 6 shows the results of a second assay for premating defects. In these experiments, a drop mating assay was used to determine whether ubc3 mutants produced and/or responded to mating pheromone. It should be noted that although this assay has been used reliably to address the role of pheromone in U. maydis mating reactions (Snetselaar et al., 1996) pheromone is not directly measured in these experiments and morphological responses are thus simply interpreted as a result of pheromone production and/or response. To distinguish from which strain filaments originated, we used a wild-type a2 strain expressing green fluorescent protein (GFP) (Fig. 6A and B). Wild-type strains possessing opposite a mating-type specificity (b is irrelevant in this assay) responded to each other by the production of copious filaments (Fig. 6A). As reported by Snetselaar et al. (1996), a2 strains (Fig. 6A–D, bottom strain in all images) responded more rapidly than did a1 strains (Fig. 6A–D, top strain in all images). Apparently, Δubc3 strains secrete less pheromone than wild-type strains as indicated by the greatly reduced abundance of mating hyphae produced by the GFP-expressing a2 strain when paired with the Δubc3 strain (compare Fig. 6A and B). Even more dramatic is the essentially complete loss of response to secreted pheromone in Δubc3 strains (Fig. 6C). No visible response by either mating partner was witnessed when two ubc3 deletion strains were paired (Fig. 6D). This probably means that mating between two Δubc3 strains can only occur if the colonies grow large enough to bring mating partners in direct physical contact so that mating filament production is not required.


Figure 6. . Drop mating assays for pheromone response in Δubc3 deletion strains. A–D. Drops of 0.5 μl of 24 h PDB cultures of appropriate strains were spotted in close proximity on microscope slides covered with water agar and observed after 24 h (Snetselaar et al., 1996). In all cases, an a1 b1 strain is at the top and an a2 b2 strain is at the bottom of the image. Bar represents 50 μm. A. Top, wild type (1/2); bottom, wild-type FB2 GFP (2/77). B. Top, FB1Δubc3 (2/58); bottom, wild-type FB2 GFP (2/77). C. Top, wild type (1/2); bottom, FB2Δubc3 (2/59). D. Top, FB1Δubc3 (2/58); bottom, FB2Δubc3 (2/59).

Download figure to PowerPoint

Mutation in the fuz7 MAP kinase kinase gene suppresses filamentation of uac1 mutants

The U. maydis fuz7 gene encodes a MAP kinase kinase that was identified earlier based on conserved sequences (Banuett and Herskowitz, 1994). Null fuz7 haploid mutants in pure culture have a morphological phenotype similar to wild type (Fig. 4E). Defects in fuz7, however, were reported to cause haploid cells to lose their ability to respond to pheromone and to thus generate a reduced plate mating reaction similar to that described above for ubc3 mutants (Banuett and Herskowitz, 1994). We carried out spot mating assays with fuz7 mutants (data not shown) and they gave results indistinguishable from those of the ubc3 mutants described above. Additionally, dikaryons homozygous for fuz7 mutations are significantly reduced in virulence (Banuett and Herskowitz, 1994). Because of the potential role of the putative Fuz7 MAP kinase kinase as an activator of the Ubc3 MAP kinase, we determined whether the genes are probably members of the same or disparate signal transduction pathways. The fuz7 disruption mutant was crossed to the uac1::ble strain 1/9. Progeny were selected that displayed both a hygromycin- and a phleomycin-resistant phenotype and were observed for morphological characteristics in liquid and solid media. The phenotype of a uac1 fuz7 double mutant was found to be indistinguishable from the phenotype of the uac1 ubc3 double mutants (compare Fig. 4F with Fig. 1A and Fig. 4D). Thus, the fuz7 gene, like ubc3, is required for filamentous growth of haploid strains lacking the ability to synthesize cAMP. These results are consistent with the probable involvement of both Fuz7 and Ubc3 in the same MAP kinase cascade.

The role of ubc3 in virulence of U. maydis on maize

To determine the effects of mutations in the ubc3 gene on pathogenicity, plants were inoculated with various pairwise combinations of wild-type, ubc3 single mutant and uac1 ubc3 double mutant compatible mating partners. Plant inoculations to assess the pathogenicity and virulence of the ubc3 mutants were conducted as previously described (Gold et al., 1997). Plant inoculation experiments were performed at least three times with the typical results of one such test shown in Table 3. The results indicate that the ubc3 gene does not act as a pathogenicity factor but functions as a relatively weak virulence factor (Table 3). This interpretation is supported by the finding that mutant dikaryons are capable of infecting 100% of inoculated plants but those plants show slightly less severe symptoms when compared with wild-type infections. The uac1 and ubc3 mutations act synergistically to reduce pathogen virulence. This was evident from the observation that strains mutated in ubc3 alone, when crossed to a wild-type strain (Table 3, treatment 2), caused disease typical of dikaryons formed by mating two wild-type strains (Table 3, treatment 1); crosses involving one strain mutated in uac1 were moderately reduced in virulence (Table 3, treatment 5); crosses generating dikaryons heterozygous for functional and non-functional alleles of uac1 and ubc3 had an even greater reduction in virulence (Table 3, treatments 6 and 7). Dikaryons formed in which both mutations are within the same nucleus consistently appear more reduced in virulence (Table 3, compare treatment 7 with treatment 6). No large stem galls were ever observed in any of the trials of treatment 7. In addition, dikaryons homozygous for non-functional ubc3 alleles and heterozygous for functional and non-functional uac1 alleles were even more reduced in virulence and slightly less pathogenic with some plants escaping infection (Table 3, treatment 8). Finally, plants inoculated with two uac1::bleΔubc3 double mutants generally showed no symptoms (Table 3, treatment 9). A few plants receiving this treatment displayed localized chlorosis, but no galls were ever formed.

Table 3. . Pathogenicity of ubc3-1 and ubc3 disruption/deletion mutations.a a. Results of one of three tests. All test results were similar.b. Treatments were inoculations of 106 cells ml−1 for each of the paired strains as follows: 1 = (1/2 × 2/9); 2 = (1/2 × 2/59); 3 = (2/9 × 2/58); 4 = (2/58 × 2/59); 5 = (2/9 × 1/9); 6 = (1/9 × 2/59); 7 = (2/9 × 3/25); 8 = (2/59 × 3/25); 9 = (2/60 × 2/61).c. Recorded 10 days after inoculation.d. Disease index is calculated as Σ disease ratings divided by number of plants.e.Δ indicates a deletion mutation.Thumbnail image of


  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The ubc3 gene encodes a MAP kinase

The cosmid clone cosubc3 was identified previously in a functional complementation assay (Mayorga and Gold, 1998). The active fragment has now been subcloned and sequenced. A single long open reading frame lacking introns was identified and the putative translation product possessed strong homology to other MAP kinases. In addition, the open reading frame is co-linear with and approximately the same size as the other MAP kinases. The putative polypeptide has the canonical Cdc2-related kinase domain sequences HRDLKP and WYRAPE for subdomains VIB and VIII respectively (Hardie, 1993). Phylogenetic analysis indicated that Ubc3 was more similar to Saccharomyces pombe and S. cerevisiae MAP kinases involved in pheromone response than to MAP kinases from these organisms involved in responses to various other signals.

The ubc3 MAP kinase gene is required for haploid filamentation and for typical strong mating reactions

To confirm its requirement in filamentation of adenylate cyclase mutant haploid strains, a ubc3 deletion mutant was used. Progeny of crosses between single Δubc3 and uac1::ble mutants were identified that were defective in both genes. These mutants display a morphological phenotype that is indistinguishable from the original uac1::ble ubc3-1 strain 3/25. Thus, the function of the ubc3 MAP kinase is required for the filamentation displayed by adenylate cyclase mutants.

Strains deficient for ubc3 generate greatly reduced filamentation in plate mating reactions (Fig. 5). There was a clear ubc3 dose-dependent intensity gradient in mating reactions (ubc3+ × ubc3+ ubc3+ × ubc3 ubc3 × ubc3). As described below, this is in part probably due to reduced pheromone production and response by ubc3 strains. It is, however, impossible to conclude from this data that there are no post-fusion defects in Δubc3/Δubc3 dikaryons. In fact, although not tested here, it would be expected that there should be a post-fusion defect in Δubc3/Δubc3 dikaryons as well as a premating defect due to an inability of the dikaryon to respond to pheromone in the autocrine response (Spellig et al., 1994).

The ubc3 gene is required for a normal response to pheromone

Reduced filamentation in mating reactions with Δubc3 strains was observed, indicating a likely role for Ubc3 in pheromone response. Thus, two assays were used to determine the function of the ubc3 gene in the pheromone response pathway.

A cytoduction assay measured the frequency of cell fusion. The frequency of cell fusion was reduced in matings between strains when one or both mating partner(s) was mutated in the ubc3 gene, albeit if both mating partners were defective in the ubc3 gene mating competence was more compromised than if only one strain is a ubc3 mutant.

A drop mating assay clearly demonstrated that Δubc3 strains are compromised for receipt of pheromone signals. These strains are also evidently reduced in pheromone production as they trigger a much weaker reaction in mating partners than do wild-type strains. This could be due to a reduction in the constitutive and/or induced level of pheromone. Although at this point we cannot distinguish between these possibilities, it seems likely that the reduction is primarily due to a defect in response.

The ubc3 and fuz7 genes appear to be in the same MAP kinase cascade

The sequence similarities to pheromone response pathway MAP kinase cascade members coupled with the direct evidence that mutation in either gene caused interference with the response to pheromone produced by a mating partner of opposite a locus specificity (Banuett and Herskowitz, 1994; Fig. 6D) suggests a common role for ubc3 and fuz7 in pheromone response. Additionally, mutations in the fuz7 gene, like mutations in the ubc3 gene, were found to suppress specifically the filamentous phenotype of an adenylate cyclase mutant, providing another piece of evidence that ubc3 and fuz7 are probably involved in the same signalling pathway. However, although both of these genes appear to be members of the same pheromone-responsive MAP kinase pathway (see below), there is a large difference in the relative importance of the ubc3 and fuz7 genes in virulence on inoculated maize. As proposed by Banuett and Herskowitz (1994), this discrepancy in significance for virulence may imply that there is a second gene at the level of the MAP kinase that is responsive to signals from the plant. In addition to the ubc3 and fuz7 genes, we identified another gene, ubc4, that encodes a MAP kinase kinase kinase (Mayorga and Gold, 1998; M. E. Mayorga and S. E. Gold, unpublished) that also is a suppressor of uac1::ble filamentous growth. Thus, members of all three levels of the kinase cascade for pheromone response in U. maydis have been identified.

The ubc3 gene is a weak virulence factor

In contrast to the fuz7 MAP kinase kinase gene which is important for the development of typical large galls containing numerous teliospores, the sexual spores of Ustilago maydis (Banuett and Herskowitz, 1994), the ubc3 gene plays only a minor role in virulence. Compatible strains mutant in the ubc3 gene often produced large galls that were blackened by the presence of abundant teliospores. Fewer plants were killed by inoculation with compatible ubc3 mutants, indicating a reduction in virulence. This reduction in virulence is evident in the reduced disease index of 3.1 calculated for treatment 4 compared with the disease index of 4.1 for the wild-type infection (Table 3). It was observed that the germination rate of the few teliospores produced by infections with fuz7 mutants was dramatically reduced (Banuett and Herskowitz, 1994). This also appears to be the case with regard to the more abundant teliospores produced in Δubc3 infections. We noted an ≈ 100-fold reduction in the germination rate for teliospores derived from Δubc3 ×Δubc3 inoculations compared with Δubc3 × wild type or wild type × wild type.

Integration of the pheromone response MAP kinase and the cAMP-dependent protein kinase signal transduction pathways

Several genes have been identified in the U. maydis cAMP signalling pathway. The gpa3 gene encodes a stimulatory Gα subunit of a trimeric GTPase (Regenfelder et al., 1997; Kruger et al., 1998). The uac1 gene encodes adenylate cyclase (Gold et al., 1994a) and is epistatic to gpa3 (Kruger et al., 1998). The ubc1 gene encodes the regulatory subunit of PKA and is epistatic to uac1 both for morphology (Gold et al., 1994a) and for pathogenicity (Gold et al., 1997). Finally, the principal catalytic subunit of PKA (adr1) has been identified (Dürrenberger et al., 1998). Mutations in adr1 are epistatic to mutations in ubc1. Thus, the members of the pathway from G-protein to catalytic subunit are defined. The signal(s) and receptor(s) responsible for activating Gpa3 as well as the substrates of Adr1 are yet to be identified. The conservation of structure and function of orthologous genes in other smut fungi is implicated by the identification of the fil1 (McCluskey et al., 1994) gene of U. hordei as a Gα that when mutated generates strains that, like gpa3 mutants, are filamentous (Lichter and Mills, 1997).

The interaction of MAP kinase and cAMP signalling pathways in fungi has recently been demonstrated in a number of fungal systems including saprophytes and both plant and animal pathogens (for recent reviews, see Kronstad et al., 1998; Madhani and Fink, 1998). In diploid S. cerevisiae pseudohyphal growth is controlled by two of three PKA catalytic subunits genes (Robertson and Fink, 1998) and involves the pheromone-responsive MAP kinase cascade (Roberts and Fink, 1994). In S. cerevisiae haploid cells, the two MAP kinases at the level of Ubc3 are Fus3p, which regulates mating, and the Kss1p, which is involved in filamentous invasive growth (Madhani et al., 1997). The MAP kinase cascades of both diploid and haploid S. cerevisiae cells transmit their signals to the STE12p transcription factor (Liu et al., 1993). In Neurospora crassa, a MAP kinase pathway and a kinase similar in structure to a cAMP-dependent protein kinase were found to regulate conidiation (Kothe and Free, 1998). In Magnaporthe grisea, the rice blast fungus, the pheromone response MAP kinase gene PMK1 acts downstream from, and co-operatively with, cAMP in appressorium formation (Xu and Hamer, 1996). In Candida albicans, conserved components of the pheromone response MAP kinase cascade are required for filamentous growth (Liu et al., 1994; Kohler and Fink, 1996), which is in turn required for pathogenicity (Lo et al., 1997). In the present work, we show results that add additional support for a similar interaction of the two pathways in dimorphism and virulence in U. maydis.

Our results show a direct interplay of the pheromone response MAP kinase cascade and the cAMP-dependent protein kinase pathway. This is evident from the result that the ubc3 and fuz7 genes act as specific suppressors of the filamentous phenotype of haploid strains mutant in the adenylate cyclase gene. The roles of these two kinase pathways appear to be antagonistic with respect to filamentous growth. Active PKA encourages budding growth (Gold et al., 1994a) and inactivated PKA results in filamentous growth (Gold et al., 1994a; Dürrenberger et al., 1998). The filamentous growth of strains with inactive PKA requires a functional ubc3 (also fuz7 and ubc4) MAP kinase cascade. Double adenylate cyclase–MAP kinase mutants were observed to have a budding morphology, but with a tendency to form clusters of cells (Mayorga and Gold, 1998; Fig. 4D). No obvious mutant phenotype in the budding morphology was apparent in MAP kinase single mutants. This suggests that the MAP kinase does not play a role in the normal haploid budding morphology and that cAMP is involved in transmitting a signal for cell separation that is independent of MAP kinase function. High levels of cAMP were previously shown to suppress filament formation in mating reactions and particularly in diploids heterozygous at both the a and b mating-type loci (Gold et al., 1997). This suggests that pheromone response can be overpowered by PKA activity, again indicating antagonistic roles for the two kinase pathways. Kruger et al. (1998) showed that the cAMP pathway is also required for pheromone response.

As shown in the model presented by Dürrenberger et al. (1998), the best candidate currently available for the integration of the cAMP and MAP kinase signal transduction pathways in U. maydis is the pheromone response factor 1 (prf1) gene product (Hartmann et al., 1996). Prf1 is an HMG class transcription factor (Hartmann et al., 1996) and has consensus sequences for phosphorylation by both MAP kinase and cAMP-dependent protein kinase (R. Kahmann, personal communication). Therefore, the variable phosphorylation of specific amino acid residues on the Prf1 transcription factor by the cAMP-dependent and/or mitogen-activated protein kinases may be the integration point controlling mating, morphogenesis and virulence.

Experimental procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Fungal and bacterial strains and culture conditions

Ustilago maydis strains were grown at 30°C in liquid YEPS (1% yeast extract, 2% peptone, 2% sucrose), in liquid potato dextrose broth (PDB, Difco) or on solid potato dextrose agar (PDA). Transformed U. maydis strains were plated on double complete medium with sorbitol containing 300 μg ml−1 hygromycin B as previously described (Barrett et al., 1993). The U. maydis strains used in this study are described in Table 1.

Escherichia coli strain DH5α (Bethesda Research Laboratories) was used for all DNA manipulations. E. coli was grown in/on Luria–Bertani (LB) medium with appropriate antibiotics.

Nucleic acid manipulations

The 7.5 kb XbaI fragment from cosubc3 was first subcloned into the XbaI site of pBluescript KS (Stratagene) (Fig. 2). Subsequently, the 7.5 kb XbaI fragment was transferred to the directly selectable autonomously replicating plasmid phyg101 (Mayorga and Gold, 1998). This was accomplished by removal of the 7.5 kb fragment with the flanking BamHI and NotI polylinker sites followed by ligation into phyg101 digested with the same enzymes (pSMBUBC3-1, Fig. 2). The Erase-a-base kit was used according to the manufacturer's instructions (Promega) to generate an exonuclease-III-mediated nested deletion series from this fragment. A 3.5 kb deletion clone, designated pTP4F4a, containing the minimal fragment for complementing the ubc3-1 mutation was chosen for sequencing (Fig. 2).

For construction of the Δubc3 deletion mutants, the 717 bp MluI/BglII fragment of ubc3 was replaced with a 1.5 kb AscI/BglII fragment containing the nourseothricin (NAT) resistance gene (see Fig. 2) (Gold et al., 1994b). The portion of the MAP kinase gene deleted extends from −239 (MluI) to +478 (Bglll) with respect to the translational start. The Bglll site cuts within subdomain VII of the conserved catalytic domain, thus deleting all N-terminal portions of this domain. It is therefore expected that this mutation yields a non-functional ubc3 allele.

To confirm the functional status of the ubc3 and uac1 genes in the double mutant progeny, Southern blotting was performed using the non-radioactive DIG/Genius system (Boehringer Mannheim) according to the manufacturer's instructions (data not shown). Genomic DNA of the double mutant progeny (phleomycin and nourseothricin resistant) as well as genomic DNA from the parental strains 1/9 and 2/59 were digested with HindIII or BamHI and blotted. The HindIII-digested blot was probed with the ubc3 deletion construct. The BamHI-digested blot was probed with pFuz60C (containing a 4.7 kb BamHI fragment of uac1 cloned into pUC13; Barrett et al., 1993).

Because standard mating assays were not possible with some progeny strains, the allelic specificity of the a and b mating-type loci was determined by molecular means. To determine the b mating-type allelic specificity (b1 or b2) of the double mutant progeny, primer 1A (5′-AACGGATCCTCA TAAGCCTCCTCGTAT-3′), primer 10 (5′-AAGGATCCATAG CGTGAGCTGATGA-3′) and primer 3W (5′-ATGGATCCTC ATACACTCGTCGTAG-3′), kindly provided by J. Kronstad, were used (Barrett et al., 1993). For a idiomorph determination, plasmid pA1.1 (Froeliger and Leong, 1991), kindly provided by J. Kronstad, was used as a probe in the BamHI blotting described above.

PCR amplifications as well as cloning procedures were carried out by standard techniques (Ausubel et al., 1987; Sambrook et al., 1989).

Gene sequencing and analysis

Sequencing was carried out in part by the University of Georgia Molecular Genetics Instrumentation Facility (MGIF). The ubc3 sequence was completed in our laboratory and samples were run on the IBI 310 capillary automated sequencer in the shared Genome Analysis Facility. The ubc3 gene nucleotide sequence has been submitted to GenBank (accession no. AF170532).

macdnasis pro V3.5 (Hitachi Software Engineering) was used for DNA and protein sequence analysis.

Plant inoculations

Trucker's Favorite (Imperial Garden Seed, Athens Seed Co.) seedlings were grown in UGA soil mix (Gold et al., 1997), and at 7 days they were injected into the culm just above the soil line with cell suspensions of mixtures of 106 cells ml−1 of each of the mating strains. Double uac1 ubc3 mutants used in this study commonly possess a multicellular phenotype (see Fig. 1A), thus when quantifying inoculum with a haemocytometer each cell was counted whether or not dehiscence from the mother cell had occurred. Plants were maintained in a Conviron E 15 growth chamber (Conviron, Inman) with 14 h day/10 h night cycles at a constant temperature of 26°C. Disease symptom data were collected at 7, 10 and 14 days after inoculation. Disease ratings were as described previously (Gold et al., 1997). Briefly, ratings 0–5 are: 0, no disease; 1, anthocyanin or chlorosis; 2, leaf galls; 3, small stem galls; 4, large stem galls; 5, plant death due to disease. Experiments to analyse pathogenicity were carried out a minimum of three times.


Photographs of fungal cell morphology were taken using Kodak Ektachrome 64T film on an Olympus BH-2 microscope with differential interference optics as previously described (Mayorga and Gold, 1998). For fluorescence microscopy, preparations were visualized with a Zeiss Axioplan Universal microscope (Carl Zeiss) and photographed with a Zeiss MC100 microscope camera system using Kodak Elite 100 film.

Photographic slides were scanned with a Nikon Coolscan (Nikon) and photographic figures were compiled with the software application photoshop 4.0 (Adobe Systems) and printed with a Sony UP-D7000 dye sublimation printer with Power Macintosh text editing.

Plate mating and cytoduction assays

A plate mating assay was used to determine the effect of the ubc3 mutations on the filamentous phenotype associated with the mating reaction. Strains were grown overnight in YEPS and then plated onto complete medium containing 1% charcoal at room temperature (Holliday, 1974). Compatible wild-type haploid control strains 1/2 and 2/9 form yeast-like colonies; upon mixing, they produce the dikaryotic filaments typical of the successful mating reaction. Pairs of mutant strains possessing opposite alleles at both the a and b mating-type loci were first mixed, then spotted and subsequently observed for production of filaments typical of the successful mating reaction. These mating reactions were photographed on a copy stand with Kodak DMP 5360 direct positive film.

The role of the ubc3 gene in the fusion interaction was determined using a cytoduction assay (Trueheart and Herskowitz, 1992; Laity et al., 1995). Strains were grown overnight in YEPS and approximately the same number of cells, based on optical densities, were co-spotted onto charcoal plates. Plates were then wrapped and incubated at 30°C for 24 h. After incubation, the colony was transferred to a microfuge tube containing 1 ml sterile water. These were then vortexed for 15 min at full speed to ensure complete resuspension. The suspensions were then diluted from 1 × 100 to 1 × 10−6. Dilutions from 1 × 10−3 to 1 × 10−6 were plated on minimal medium containing 1.5 μg ml−1 benomyl to determine the concentration of potential recipient strains. Dilutions from 1 × 100 to 1 × 10−3 were plated on minimal medium containing 3 μg ml−1 oligomycin (Sigma) and 1.5 μg ml−1 benomyl to determine the concentration of cytoduced recipient cells that were now oligomycin resistant. The reduction in fusion was calculated as the ratio of cytoduction frequency of the control cross to each experimental cross (control/experimental). Experiments to analyse cell fusion were carried out four times.

Drop mating assays

Drop mating assays were modified from Snetselaar et al. (1996). Cells were grown for 24 h in PDB and 0.5 μl drops of undiluted cultures of compatible mating-type strains spotted in close proximity to each other on water agar-covered slides. Those drops that were sufficiently close were covered with 3 μl drops of mineral oil (Sigma). Mating reactions were observed over a 24 h period. To aid in differentiation of the mating filaments produced by the two compatible strains, a haploid strain (2/77) that is a2 b2 and expresses GFP constitutively as a result of insertion of plasmid pOTEF-SG (Spellig et al., 1996) was used.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

This work was supported by US Department of Agriculture National Research Initiative Grant nos 95-37303-1706 and 97-35303-4871 to S.E.G. We acknowledge Flora Banuett for providing strain 2/38. We acknowledge R. Kahmann for providing the Δkpp2ubc3) strains and for strain 2/77. We would like to thank Ms. Gretel Abramowsky, Mr John Duick, Dr John Sherwood and Dr David Andrews for critical review of the manuscript.

  1. *Present address: Microbia Inc., 840 Memorial Drive, Cambridge, MA 02139, USA.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  • 1
    Altschul, S.F., Gish, W., Miller, W., Myers, E.W., Lipman, D.J. (1990) Basic local alignment search tool. J Mol Biol 215: 403410.
  • 2
    Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A., Struhl, K. (1987) Current Protocols in Molecular Biology. New York: Greene Publishing Associates/Wiley-Interscience.
  • 3
    Banuett, F. (1995) Genetics of Ustilago maydis, a fungal pathogen that induces tumors in maize. Annu Rev Genet 29: 179208.
  • 4
    Banuett, F. (1998) Signalling in the yeasts: an informational cascade with links to the filamentous fungi. Microbiol Mol Biol Rev 62: 249274.
  • 5
    Banuett, F. & Herskowitz, I. (1989) Different a-alleles of Ustilago maydis are necessary for maintenance of filamentous growth but not for meiosis. Proc Natl Acad Sci USA 86: 58785882.
  • 6
    Banuett, F. & Herskowitz, I. (1994) Identification of fuz7, a Ustilago maydis MEK/MAPKK homolog required for a-locus-dependent and -independent steps in the fungal life cycle. Genes Dev 8: 13671378.
  • 7
    Barrett, K.J., Gold, S.E., Kronstad, J.W. (1993) Identification and complementation of a mutation to constitutive filamentous growth in Ustilago maydis. Mol Plant Microbe Interact 6: 274283.
  • 8
    Bölker, M., Urban, M., Kahmann, R. (1992) The a mating type locus of U. maydis specifies cell signaling components. Cell 68: 441450.
  • 9
    Dürrenberger, F., Wong, K., Kronstad, J.W. (1998) Identification of a cAMP-dependent protein kinase catalytic subunit required for virulence and morphogenesis in Ustilago maydis. Proc Natl Acad Sci USA 95: 56845689.
  • 10
    Elion, E.A., Grisafi, P.L., Fink, G.R. (1990) FUS3 encodes a cdc2+/CDC28-related kinase required for the transition from mitosis into conjugation. Cell 60: 649664.
  • 11
    Froeliger, E.H. & Leong, S.A. (1991) The a mating-type alleles of Ustilago maydis are idiomorphs. Gene 100: 113122.
  • 12
    Gold, S., Duncan, G., Barrett, K., Kronstad, J. (1994 a) cAMP regulates morphogenesis in the fungal pathogen Ustilago maydis. Genes Dev 8: 28052816.
  • 13
    Gold, S.E., Bakkeren, G., Davies, J.E., Kronstad, J.W. (1994 b) Three selectable markers for transformation of Ustilago maydis. Gene 142: 225230.
  • 14
    Gold, S.E., Brogdon, S.M., Mayorga, M.E., Kronstad, J.W. (1997) The Ustilago maydis regulatory subunit of a cAMP-dependent protein kinase is required for gall formation in maize. Plant Cell 9: 15851594.
  • 15
    Hardie, D.G. (1993) Protein Phosphorylation: A Practical Approach. Rickwood, D. (ed.). Oxford: IRL Press.
  • 16
    Hartmann, H.A., Kahmann, R., Bölker, M. (1996) The pheromone response factor coordinates filamentous growth and pathogenicity in Ustilago maydis. EMBO J 15: 16321641.
  • 17
    Holliday, R. (1974) Ustilago maydis. In Handbook of Genetics. King, R.C. (ed.). New York: Plenum, pp. 575595.
  • 18
    Kahmann, R., Romeis, T., Bölker, M., Kämper, J. (1995) Control of mating and development in Ustilago maydis. Curr Opin Genet Dev 5: 559564.
  • 19
    Kohler, J.R. & Fink, G.R. (1996) Candida albicans strains heterozygous and homozygous for mutations in mitogen-activated protein kinase signaling components have defects in hyphal development. Proc Natl Acad Sci USA 93: 1322313228.
  • 20
    Kothe, G.O. & Free, S.J. (1998) The isolation and characterization of nrc-1 and nrc-2, two genes encoding protein kinases that control growth and development in Neurospora crassa. Genetics 149: 117130.
  • 21
    Kronstad, J.W. & Staben, C. (1997) Mating type in filamentous fungi. Annu Rev Genet 31: 245276.
  • 22
    Kronstad, J., De Maria, A.D., Funnell, D., Laidlaw, R.D., Lee, N., De Sa, M.M., Ramesh, M. (1998) Signaling via cAMP in fungi: interconnections with mitogen-activated protein kinase pathways. Arch Microbiol 170: 395404.
  • 23
    Kruger, J., Loubradou, G., Regenfelder, E., Hartmann, A., Kahmann, R. (1998) Crosstalk between cAMP and pheromone signaling pathways in Ustilago maydis. Mol Gen Genet 260: 193198.
  • 24
    Laity, C., Giasson, L., Campbell, R., Kronstad, J. (1995) Heterozygosity at the b mating-type locus attenuates fusion in Ustilago maydis. Curr Genet 27: 451459.
  • 25
    Li, D., Rogers, L., Kolattukudy, P.E. (1997) Cloning and expression of cDNA encoding a mitogen-activated protein kinase from a phytopathogenic filamentous fungus. Gene 195: 161166.
  • 26
    Lichter, A. & Mills, D. (1997) Fil1, a G-protein alpha-subunit that acts upstream of cAMP and is essential for dimorphic switching in haploid cells of Ustilago hordei. Mol Gen Genet 256: 426435.
  • 27
    Liu, H., Styles, C.A., Fink, G.R. (1993) Elements of the yeast pheromone response pathway required for filamentous growth of diploids. Science 262: 17411744.
  • 28
    Liu, H., Kohler, J., Fink, G.R. (1994) Suppression of hyphal formation in Candida albicans by mutation of a STE12 homolog (published erratum appears in Science 1995 267: 17). Science 266: 17231726.
  • 29
    Lo, H.J., Kohler, J.R., DiDomenico, B., Loebenberg, D., Cacciapuoti, A., Fink, G.R. (1997) Nonfilamentous C. albicans mutants are avirulent. Cell 90: 939949.
  • 30
    Madhani, H.D. & Fink, G.R. (1998) The control of filamentous differentiation and virulence in fungi. Trends Cell Biol 8: 348353.
  • 31
    Madhani, H.D., Styles, C.A., Fink, G.R. (1997) MAP kinases with distinct inhibitory functions impart signaling specificity during yeast differentiation. Cell 91: 673684.
  • 32
    Mayorga, M.E. & Gold, S.E. (1998) Characterization and molecular genetic complementation of mutants affecting dimorphism in the fungus Ustilago maydis. Fungal Genet Biol 24: 364376.
  • 33
    McCluskey, K., Agnan, J., Mills, D. (1994) Characterization of genome plasticity in Ustilago hordei. Curr Genet 26: 486493.
  • 34
    Regenfelder, E., Spellig, T., Hartmann, A., Lauenstein, S., Bolker, M., Kahmann, R. (1997) G proteins in Ustilago maydis: transmission of multiple signals? EMBO J 16: 19341942.
  • 35
    Roberts, R.L. & Fink, G.R. (1994) Elements of a single MAP kinase cascade in Saccharomyces cerevisiae mediate two developmental programs in the same cell type: mating and invasive growth. Genes Dev 8: 29742985.
  • 36
    Robertson, L.S. & Fink, G.R. (1998) The three yeast A kinases have specific signaling functions in pseudohyphal growth. Proc Natl Acad Sci USA 95: 1378313787.
  • 37
    Ruiz-Herrera, J., León, C.G., Guevara-Olvera, L., Carabez-Trejo, A. (1995) Yeast-mycelial dimorphism of haploid and diploid strains of Ustilago maydis. Microbiology 141: 695703.
  • 38
    Sambrook, J., Fritsch, E.F., Maniatis, T. (1989) Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor NY: Cold Spring Harbor Laboratory Press.
  • 39
    Snetselaar, K.M., Bölker, M., Kahmann, R. (1996) Ustilago maydis mating hyphae orient their growth toward pheromone sources. Fungal Genet Biol 20: 299312.
  • 40
    Spellig, T., Bölker, M., Lottspeich, F., Frank, R.W., Kahmann, R. (1994) Pheromones trigger filamentous growth in Ustilago maydis. EMBO J 13: 16201627.
  • 41
    Spellig, T., Bottin, A., Kahmann, R. (1996) Green fluorescent protein (GFP) as a new vital marker in the phytopathogenic fungus Ustilago maydis. Mol Gen Genet 252: 503509.
  • 42
    Trueheart, J. & Herskowitz, I. (1992) The a locus governs cytoduction in Ustilago maydis. J Bacteriol 174: 78317833.
  • 43
    Whiteway, M., Dignard, D., Thomas, D.Y. (1992) Dominant negative selection of heterologous genes: isolation of Candida albicans genes that interfere with Saccharomyces cerevisiae mating factor-induced cell cycle arrest. Proc Natl Acad Sci USA 89: 94109414.
  • 44
    Xu, J.R. & Hamer, J.E. (1996) MAP kinase and cAMP signaling regulate infection structure formation and pathogenic growth in the rice blast fungus Magnaporthe grisea. Genes Dev 10: 26962706.