- Email: [email protected]
Ustilago maydisMating Hyphae Orient Their Growth toward Pheromone Sources
Fungal Genetics and Biology 20, 299–312 (1996) Article No. 0044
Ustilago maydis Mating Hyphae Orient Their Growth toward Pheromone Sources
Karen M. Snetselaar,*,1 Michael Bo ¨ lker,† and Regine Kahmann† *Department of Biology, St. Joseph’s University, 5600 City Avenue, Philadelphia, Pennsylvania 19131; and †Institut fu¨r Genetik und Mikrobiologie, Universita¨t Mu¨nchen, Maria-Ward-Str. 1a, 80638 Munich, Germany
Accepted for publication October 18, 1996
Snetselaar, K. M., Bo¨lker, M., and Kahmann, R. 1996. Ustilago maydis mating hyphae orient their growth toward pheromone sources. Fungal Genetics and Biology 20, 299–312. When small drops of Ustilago maydis sporidia were placed 100–200 mm apart on agar surfaces and covered with paraffin oil, sporidia from one drop formed thin hyphae that grew in a zig-zag fashion toward the other drop if it contained sporidia making the appropriate pheromone. For example, a2b2 mating hyphae grew toward a1b1 and a1b2 mating hyphae, and the filaments eventually fused tip to tip. Time-lapse photography indicated that the mating hyphae can rapidly change orientation in response to nearby compatible sporidia. When exposed to pheromone produced by cells in an adjacent drop, haploid sporidia with the a2 allele began elongating before sporidia with the a1 allele. Sporidia without functional pheromone genes responded to pheromone although they did not induce a response, and sporidia without pheromone receptors induced formation of mating hyphae although they did not form mating hyphae. Diploid sporidia heterozygous at b but not at a formed straight, rigid, aerial filaments when exposed to pheromone produced by the appropriate haploid sporidia. Again, the a2a2b1b2 strain formed filaments more quickly than the a1a1b1b2 strain. Taken together, these results suggest that the a2 pheromone diffuses less readily or is degraded more quickly than the a1 pheromone. r 1996 Academic Press
1
To whom reprint requests should be addressed.
1087-1845/96 $18.00 Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
Index Descriptors: Ustilago; pheromone; mating; chemotaxis. Maize smut disease is caused by the fungus Ustilago maydis (DC.) Corda. The disease cycle begins when compatible haploid sporidia cease budding growth and form mating hyphae (also called conjugation tubes) that fuse to form a dikaryotic filament. The obligately parasitic dikaryon can enter the plant in any meristematic area. After limited hyphal proliferation around the infection site, the fungus induces abnormal cell division and enlargement in the host, resulting in gall formation. Hyphal walls of dikaryotic fungal cells in the galls swell, gelatinize, and separate at the crosswalls to form teliospores that contain a single diploid nucleus at maturity. These thick-walled spores are disseminated when the galls break open, and they can overwinter in soil or plant debris. Meiosis occurs when the teliospores germinate and form saprobic haploid sporidia that can begin the infection cycle anew (see Christensen, 1963, for further life history details). Classical genetic studies showed that U. maydis has an unusual tetrapolar mating system (see Holliday, 1974, for review). The a locus, which has two alleles, controls sporidial fusion (Rowell, 1955) while the b locus, which has many alleles, regulates the ability of the fungus to cause disease on the host (Holliday, 1961). Alleles from both the a and the b loci have been cloned, sequenced, and characterized. The a1 and a2 alleles each contain a pheromone gene (mfa) and a pheromone receptor gene (pra). mfa and pra mutants, which have nonfunctional pheromones or pheromone receptors, were constructed and used to show that these genes are required for mating
299
300
(Bo¨lker et al., 1992; Spellig et al., 1994). In addition the a2 allele encodes two genes with unknown functions (Urban et al., 1996b). Each b allele encodes two homeodomain proteins able to form heterodimers if derived from different alleles (Gillissen et al., 1992; Ka¨mper et al., 1995). Morphological and developmental information about sporidial mating and infection in U. maydis has come more slowly. Indeed, the genetic evidence for pheromones and pheromone receptors preceded any morphological evidence of a role for chemical attractants in the mating process (Bo¨lker and Kahmann, 1993). Recently, however, assays for formation of mating hyphae (Banuett and Herskowitz, 1994) and for mating and infection hyphae (Snetselaar and Mims, 1992) have been described. These assays allowed microscopic confirmation that haploid sporidia with unlike a alleles fuse regardless of b, but unlike b alleles are required for development of infection hyphae (Snetselaar, 1993), confirming earlier observations made by Rowell (1955). The behavior of mating and infection hyphae have also been observed on host cells (Snetselaar and Mims, 1993). When induced by starvation, diploid sporidia heterozygous at both a and b formed infection hyphae without mating, but diploid sporidia heterozygous at a but not at b mated repeatedly unless fusion occurred between cells with different b alleles, at which time infection hyphae formed (Snetselaar, 1993). In addition, pheromone isolated from culture filtrates has now been used to induce formation of mating hyphae on plates (Spellig et al., 1994). This assay was used to purify the secreted form of both pheromones. Peptide sequencing and mass analysis showed that mature a1 pheromone consists of 13 amino acids and a2 pheromone of 9 amino acids, and both pheromones are farnesylated at their C-terminal cysteine residue (Spellig et al., 1994). The mating assays described so far have been useful, but they involve combining sporidia in solutions so it is not possible to distinguish one genotype from another or to determine whether mating behavior differs between partners. The assay described here allowed us to observe sporidia of known genotypes as they interacted, providing new information about the roles of pheromones and pheromone receptors in haploid and diploid U. maydis cells.
MATERIALS AND METHODS All strains (Table 1) were maintained on Difco potato dextrose agar and transferred into flasks containing 10 ml
Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
Snetselaar, Bo¨lker, and Kahmann
TABLE 1 Strains Used Strain
Genotype
FB1
alb1
FB2
a2b2
FB6a
a2b1
FB6b
a1b2
FBD11-7
a1/a1 b1/b2
FBD12-17
a2/a2 b1/b2
FBD12-3
a1/a2 b1/b1
FBD11-21
a1/a2 b2/b2
RK1785 (derived from FB2) RK1786 (derived from FB2) MU161 (derived from FB2) MU162 (derived from FB2) MU18 (derived from FB2)
a1 mfal
Source Banuett and Herskowitz, 1989 Banuett and Herskowitz, 1989 Banuett and Herskowitz, 1989 Banuett and Herskowitz, 1989 Banuett and Herskowitz, 1989 Banuett and Herskowitz, 1989 Banuett and Herskowitz, 1989 Banuett and Herskowitz, 1989 Bo¨lker et al., 1992 (see below) Bo¨lker et al., 1992 (see below) M. Urban, unpublished (see below) M. Urban, unpublished (see below) M. Urban, unpublished (see below)
Note. Strains RK1785 and RK1786 carry mutant a1 alleles in the FB2 genetic background. MU161 carries the hygromycin resistance cassette isolated as 3-kb PvuII fragment of pCM54 (Tsukuda et al., 1988) inserted into the MscI site in the open reading frame of mfa2; in MU162 the same cassette is inserted into the SnaBI within the open reading frame of pra2. MU18 carries a Tn5H insertion in the pan1 gene that is located adjacent to the a locus. The insertion was generated as described (Bo¨lker et al., 1992) in the 10-kb BamHI fragment that contains the complete a1 allele with flanking sequences. The mutation was introduced into FB2 by transformation, and transformants were checked for replacement of the resident a2 allele by the a1 allele by Southern analysis.
of Difco potato dextrose broth to grow sporidia for experiments. Flasks were shaken at 200 rpm for 12–18 h at 28°C under constant illumination. To set up mating experiments, small aliquots of cells from cultures in mid to late log phase were centrifuged in a clinical centrifuge at 4000 rpm for 3 min, and the supernatant was withdrawn and replaced with sterile water, in which the sporidia were resuspended to reach a concentration of approximately 105/ml. Mating experiments were performed by placing drops of sporidia very close together on an agar surface. The following procedure for preparing cultures was found to be efficient and suitable for subsequent microscopic observa-
301
Ustilago maydis Pheromone Responses
tions. Microscope slides were placed into large petri dishes and covered with hot 2% water agar to a depth of approximately 2 mm. After the agar cooled, slides with their agar coating were removed from the dishes by slicing the agar around them with a razor blade, and a small section of agar at one edge of the slides was removed to facilitate handling and labelling. A 10-µl micropipettor was used to place 0.5-µl drops of sporidial suspensions on the agar films. About 15 drops of one mating type were placed in rows on the slide. After approximately 5 min, the fluid from the drop had been absorbed into the agar, making it easier to place a second drop within 100–200 µm of the first. After the water in the second drop had been absorbed into the agar, those pairs of drops that were very close together but not touching were covered with 3 µl of paraffin oil (Merck). Generally, about 10 of the 15 pairs of drops would be unsuitable, either because the drops ran together or because they were farther apart than 200 µm. Two different combinations could be prepared on a single microscope slide, and five slides could be prepared in less than an hour. For some experiments, sporidia were combined in a single drop. Plastic dishes containing the slides were covered to prevent drying out and incubated at 28°C until they were observed microscopically. Mating interactions were observed with a Zeiss Axiophot microscope with differential interference contrast capability. Objectives of 103 and 203 were used without coverslips, so observations could be made over a period of time without disturbing the sporidia. For higher magnifications, coverslips were simply placed over the oil-covered drops. Growth continued for some time after coverslips were added, but generally slowed appreciably after about an hour. In order to observe nuclear condition in mating sporidia, the paraffin oil was removed from the slide and replaced with a DAPI2 solution prepared as previously described (Snetselaar, 1993). Although observations could be made directly on the agar-coated slide, usually it was more convenient to remove the sporidia and DAPI solution to a clean slide.
RESULTS Microscopic observations of sporidia combined in drops provided more detail than the plate-mating assays commonly used to assess mating, and examination of sporidia in adjacent drops allowed observation of previously unde2
Abbreviation used: DAPI, 48,6-diamidino-2-phenylindole dihydrochloride.
scribed mating behavior. All experiments described here were repeated from two to several times. There was some variation among experiments in the length of time from when the sporidial crosses were set up until the first mating hyphae were observed. Therefore, positive and negative controls for mating behavior (FB1 3 FB2 and FB1 3 FB6b) were included with each experiment, and rates of hyphal formation and elongation were only compared within a particular experiment. Note that in all figures showing interaction between sporidia in separate drops, the a2 strain is in the top drop.
Haploid Sporidia with Unlike a Alleles Form Mating Hyphae Regardless of b No mating hyphae formed when drops of FB6a (a2b1) sporidia were placed next to FB2 (a2b2) sporidia (Fig. 1) or when these sporidia were combined in a single drop (Fig. 2). DAPI-stained sporidia from these drops were yeast-like and uninucleate (Fig. 3). When a drop of FB1 (a1b1) sporidia was placed next to a drop of FB6a (a2b1) sporidia, both types of sporidia had mating hyphae after approximately 8 h (Fig. 4). When combined in a single drop, these sporidia fused and produced short, contorted, multinucleate hyphae (Figs. 5 and 6) but no straight aerial filaments. When a drop of FB1 (a1b1) sporidia was placed near a drop of FB2 (a2b2) sporidia, again both types of sporidia formed mating hyphae after approximately 8 h (Fig. 7). However, straight, rapidly growing, aerial filaments eventually formed in the area where the mating hyphae met, as well as in drops where FB1 and FB2 were combined (Fig. 8). The filaments were dikaryotic (Fig. 9). All combinations of the four tester strains gave similar results; mating hyphae formed and fused in all combinations involving unlike a alleles, while unlike alleles at both a and b were required for the formation of the aerial dikaryotic filaments. In addition, sporidia with the a2 allele began forming mating hyphae earlier than sporidia with the a1 allele, regardless of which b allele was present.
Mutants with Disrupted Mating Factor Genes Respond to Pheromone but They Cannot Induce a Pheromone Response, while Mutants with Disrupted Pheromone Receptor Genes Induce Pheromone Responses although They Cannot Respond to Pheromone Strains with nonfunctional pheromones or pheromone receptors were used to further investigate the formation of
Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
FIGS. 1–9. Mating of wild-type U. maydis sporidia. FIG. 1. No mating hyphae form between drops of sporidia with the same a allele [FB2 (top) and FB6a]. FIG. 2. No filaments form when FB2 and FB6a sporidia are in a single drop. FIG. 3. DAPI-stained cells from FB2 3 FB6a drop are uninucleate. FIG. 4. Mating hyphae form between adjacent drops of FB2 (top) and FB6b sporidia. FIG. 5. Convoluted filaments form (arrows) when FB2 and FB6b are combined in a single drop. FIG. 6. DAPI-stained filaments from FB2 3 FB6b are multinucleate. FIG. 7. Mating hyphae form between drops of FB1 and FB2 (top) sporidia. FIG. 8. Long straight filaments form (arrows) when FB1 and FB2 sporidia are combined in a single drop. FIG. 9. Filaments from the FB1 3 FB2 cross are dikaryotic (arrows). Magnification: Figs. 1, 2, 4, 5, 7, 8 approx 2503; Figs. 3, 6, 9 approx 8003.
302
303
Ustilago maydis Pheromone Responses
mating hyphae. Strain MU161, which produces an a2 receptor but no pheromone, did not induce formation of mating hyphae in FB1 sporidia, although MU161 sporidia responded to FB1 sporidia by producing mating hyphae (Fig. 10). Strain RK1786, which produces a1 pheromone but has a disrupted receptor allele, did not respond to pheromone produced by FB2, but RK1786 sporidia induced formation of mating hyphae in FB2 sporidia (Fig. 11). Similar results were observed with RK1785 (which has a disrupted a1 pheromone gene) and MU162 (which has a disrupted a2 receptor) except that the response of a1 mating hyphae was always less vigorous than the a2 response (Figs. 12 and 13). The initial responses of MU161 to FB1 and of FB2 to RK1786 seemed to occur as rapidly as in FB1 3 FB2. Therefore, it seems that basal levels of pheromone production are sufficient to induce at least the early stages of the mating program in compatible strains.
Haploid a1 and a2 Strains Respond to Pheromone Differently In all tests of wild-type and mutant sporidia, mating hyphae with the a2 allele elongated more quickly than mating hyphae with the a1 allele, especially when drops of sporidia were not very close together or sporidial concentrations were low. Several tests were performed in order to determine whether this response difference was inherent in the a alleles or due to the genetic background. Six a1 strains and six a2 strains isolated from three different maize fields near Munich were tested against FB1, FB2, and each other. Although there was some variation in how quickly the mating hyphae formed, in no case did the a1 strains respond more strongly than the a2 strains. Strain MU18, an FB2 derivative in which the entire a2 locus has been replaced with an a1 locus, was placed in a drop near a1 and a2 strains. MU18 behaved as an a1 strain, inducing a more rapid response in the a2 strain (Fig. 14). These results indicate that the difference in response to pheromone is inherent in the different alleles of the a locus.
paraffin oil (not shown). In similar drops containing six-fold dilutions of the same sporidial strains, there were no mating hyphae after 5 h (Fig. 17) and only a few after 10 h (Fig. 18). Mating hyphae also formed later when drops were farther apart (data not shown). A likely explanation for this is that it takes longer for pheromone to diffuse from a small source or from a greater distance in quantities adequate to cause a response. Therefore, in all comparative experiments described here, drops containing approximately equivalent numbers of sporidia were placed about the same distance apart.
Mating Hyphae Orient Their Growth toward the Source of Pheromone In order to ascertain whether mating hyphae orient to a pheromone gradient rather than forming randomly in response to pheromone stimulation, the RK1786 strain (which lacks a1 pheromone receptors but still produces basal levels of a1 pheromone and thus induces FB2 strains to form mating hyphae, see Fig. 11) was used as pheromone donor (Fig. 19). When FB2 sporidia were combined with an equal number of RK1786 sporidia and placed in a drop adjacent to a drop containing only RK1786 sporidia, some of the FB2 sporidia still formed mating hyphae directed toward the RK1786 drop (Fig. 20), presumably because the pheromone concentration diffusing from the RK1786 drop became higher than the concentration produced in the drop containing both FB2 and RK1786 sporidia. This indicates that the FB2 mating hyphae were following a pheromone gradient, not randomly elongating in response to pheromone. When each drop contained equal numbers of RK1786 and FB2 sporidia, no mating hyphae formed at the edges of either drop (Fig. 21), although there were mating hyphae within the drops (not shown). The FB2 sporidia were surrounded by pheromone at a concentration as high or higher than that diffusing from the adjacent drop, so there was no pheromone gradient to induce growth of mating hyphae out of the drops.
The Rate at which Mating Hyphae Elongate Depends on the Size and Location of the Pheromone Source
Wild-Type Mating Hyphae Can Quickly Reorient Their Direction of Growth toward Compatible Mating Partners
When FB1 and FB2 sporidia at a concentration of approximately 105/ml were placed in drops approximately 100 µm apart, mating hyphae began to form within 5 h (Fig. 15) and by 10 h many had formed, grown together (Fig. 16), and formed aerial filaments that grew up into the
A time-lapse series (Figs. 22–25) shows interactions between FB1 and FB2 mating hyphae that are consistent with extreme sensitivity to changes in pheromone concentrations. The small arrows show how precisely the mating hyphae orient prior to fusion; the high magnification inset
Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
FIGS. 10–18. Interactions between U. maydis sporidia in adjacent drops. The top drop in all figures contains a2 sporidia. FIG. 10. MU161 3 FB1, note mating hyphae formed by the MU161 strain. FIG. 11. FB2 3 RK1786, note mating hyphae formed by FB2. FIG. 12. FB2 3 RK1785. The RK1785 strain forms mating hyphae, but fewer than the MU161 strain (see Fig. 10) even in these drops that are very close together. FIG. 13. MU162 3 FB1. Again, the response of the FB1 strain to the pheromone produced by MU162 is much weaker than the FB2 response (Fig. 11). FIG. 14. Cross between FB2 and MU18 (bottom), an FB2 strain which has had the entire a2 allele replaced with an a1 allele. This strain now behaves as an a1 strain and induces the a2 strain to form longer filaments. FIG. 15. FB1 3 FB2 sporidia at approximately 105/ml showed mating hyphae beginning to form after 5 hr. FIG. 16. Same sporidia as Fig. 14, 5 h later. FIG. 17. FB1 3 FB2 sporidia, same cells as for Fig. 14 but diluted sixfold. No mating hyphae had formed after 5 h. FIG. 18. Same as Fig. 17, 5 h later. A few mating hyphae had formed. Magnification: All figures approx 2503.
304
Ustilago maydis Pheromone Responses
305
FIGS. 19–21. Interactions between drops of U. maydis sporidia. FIG. 19. FB2 (top) 3 FB1 cross. FIG. 20. Top drop contains equal numbers of FB2 and RK1786 sporidia, bottom is entirely RK1786. Note that some FB2 sporidia are still attracted to the RK1786 drop, where the pheromone concentration is presumably higher. FIG. 21. Both drops contain equal amounts of FB2 and RK1786 sporidia; there is no pheromone gradient attracting the FB2 sporidia from either drop. Magnification: All figures approx 2503.
provides a closer look at the cells fusing. As the FB2 hyphae approached them, the FB1 hyphae began to grow, and they often seemed to respond dramatically to nearby FB2 hyphae. The large arrows in Figs. 22–25 show two examples of how FB1 hyphae that had been turning toward approaching FB2 hyphae changed direction of growth after the FB2 hypha fused with another FB1 partner. In both cases, the change in direction occurred in less than 10 min.
Although mfa Strains Are Attracted to a Pheromone Source, They Are Unable to Precisely Locate and Fuse with Compatible Partners To further investigate the roles of pheromones and pheromone receptors on later stages in the mating process,
a drop containing equal numbers of RK1785 and RK1786 sporidia (both mutants in a1 alleles) was placed near a drop containing MU161 and MU162 sporidia (both mutants in a2 alleles). Therefore, half of the cells in each drop produced pheromone but could not sense pheromone, while the other half could respond to pheromone but could not produce it. A time-lapse sequence lasting 45 min is shown in Figs. 26–28. At first, mating hyphae grew relatively normally. However, as the MU161 hyphae approached the RK1785 hyphae, they did not fuse with them as wild-type mating hyphae generally do, but instead they grew past the potential fusion partners as though they could not detect their presence. Likewise, the RK1785 mating hyphae did not turn toward nearby MU161 mating hyphae the way the wild-type a1 strains did. No cell fusions were observed in these crosses.
Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
FIGS. 22–25. Time-lapse series of mating interactions between FB2 (top) and FB1 (bottom) U. maydis sporidia. Time elapsed between Figs. 22 and 23 is 12 min, between Figs. 23 and 24 is 6 min, between Figs. 24 and 25 is 24 min. Arrows and arrowheads point out the same area in each frame. Small arrows follow compatible mating hyphae as they orient precisely before fusing at their tips. Large arrows show two examples of FB1 mating hyphae changing growth direction away from a newly formed dikaryon. Arrowheads and inset show compatible hyphae fusing at the tips. Magnification: Figs. 22–25 approx 4003; insets approx 12003.
306
307
Ustilago maydis Pheromone Responses
Diploid Strains Heterozygous at a and Homozygous at b Do Not Exhibit Directional Growth but Can Fuse with Compatible Haploid Strains Strains heterozygous at a but not at b formed short, convoluted filaments when placed in drops alone (not shown) that looked much like those seen in afib5 crosses between haploids (Fig. 5). When the diploid strain FBD12-3 (a1a2b1b1) was combined in a drop with FB1, the filaments sometimes became quite long, but they remained convoluted and did not become aerial (Fig. 29). DAPI staining showed that these contorted filaments often contained multiple nuclei (Fig. 30). When the two strains were placed in adjacent drops, there was little evidence of attraction on either side, and the diploid strain continued its highly contorted growth (Fig. 31). When FB2 was combined in a single drop with FBD12-3, straight, rapidly growing filaments resulted (Fig. 32). DAPI staining showed that these straight filaments had two and sometimes more nuclei (Fig. 33).
Diploid Strains FBD11-7 (a1a1b1b2) and FBD12-17 (a2a2b1b2) Become Filamentous in the Presence of Compatible Haploid Cells, but FBD12-17 Cells Respond More Quickly When placed in drops alone, diploid strains that were heterozygous at b but not at a remained almost entirely yeast-like (not shown) and had the same appearance as combined a5 haploid strains (Figs. 2 and 3). When one of these diploid strains was combined in a drop with a haploid strain carrying the complementary a allele, aerial filaments formed. These filaments were straight and rigid, unlike the mating filaments formed by haploid strains, and unlike the convoluted filaments formed by the afib5 diploid strains. DAPI staining showed the filaments were uninucleate (Fig. 34). Filaments formed by the FBD12-17 3 FB1 combination were only slightly longer than those formed by the FBD11-7 3 FB2 combination (Figs. 35 and 37), but differences were more dramatic when sporidia were placed in adjacent drops (Figs. 36 and 38). FBD12-17 rapidly became quite filamentous when placed next to FB1, while FBD11-7 elongated slowly when placed next to FB2; indeed under these circumstances the FB2 usually formed mating hyphae, indicating that the diploid strain was secreting a1 pheromone.
DISCUSSION The responses described here occurred consistently when sporidia were covered with paraffin oil, but they do not require the paraffin oil. Weak interactions between mating hyphae were previously observed in drops of sporidia placed close together on water agar, and a2 strains elongated more rapidly than a1 strains (Snetselaar, unpublished). We envision that diffusion of the amphipathic pheromone is greatly facilitated at the interface between the hydrophilic water agar and the hydrophobic paraffin oil. It is interesting to note that similar interfaces may occur between hydrophobic plant surfaces and the saturated atmosphere in the leaf whorl or ear, common sites of infection in maize. These results show for the first time that wild-type U. maydis cells locate and approach each other in a highly coordinated fashion as a prelude to cell fusion. The responses of the pra and mfa mutants indicate that pheromone responses act in at least two ways before the cells fuse. Expression of the pheromone genes is induced in certain complex media and by starvation (Spellig et al., 1994; M. Bo¨lker, unpublished), and we have shown here that this level of expression is sufficient to cause mating hyphae to form in haploid sporidia. Transcription of the mfa and pra genes is more strongly stimulated by the appropriate pheromone (Urban et al., 1996a). These induced levels of pheromone are not essential for the formation of mating hyphae, because mutants without receptors can induce mating hyphae even though they cannot detect pheromone. However, the induced pheromone levels may become important in helping mating hyphae ‘‘fine-tune’’ growth as they closely approach a potential mating partner. Mating hyphae that produce no pheromone do not induce increased pheromone production in compatible strains, and under these conditions they apparently cannot detect nearby possible mating partners against a background of pheromone. There are many similarities between the premating behavior of U. maydis mating hyphae and what has been called ‘‘courtship’’ in Saccharomyces cerevisiae (Jackson and Hartwell, 1990a). Potential S. cerevisiae mating partners communicate by producing and receiving pheromone, and polarized growth of mating cells in response to pheromone gradient was suggested to be the outcome of courtship (Jackson and Hartwell, 1990b). Subsequent work demonstrated that yeast cells do indeed respond to a pheromone gradient by orienting mating structures toward a pheromone source over a period of several hours (Segall,
Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
FIGS. 26–28. Time-lapse series of interactions between U. maydis sporidia. Top drop contains equal numbers of MU161 and MU162 sporidia, bottom drop contains equal numbers of RK1785 and RK1786 sporidia. Thus, individual sporidia in each drop make either pheromone or receptors, but not both. Note absence of fusions between mating hyphae. Inset in Fig. 28 is higher magnification of the boxed area. Magnification: Figs. 26–28 approx 4003; inset approx 1003.
308
Ustilago maydis Pheromone Responses
309
FIGS. 29–33. Mating interactions between diploid afib5 sporidia and haploid sporidia. FIG. 29. When FBD12-3 (a1/a2b1/b1) sporidia were combined with FB1 sporidia, flat, convoluted filaments formed; compare with Fig. 5. FIG. 30. DAPI staining showed that some of the filaments were multinucleate. FIG. 31. When FBD12-3 sporidia (bottom drop) were placed near FB2 sporidia, little attraction between drops was seen. Note the extremely curved FBD12-3 sporidia (arrows). FIG. 32. When FBD12-3 and FB2 were placed in the same drop, straight, aerial, rapidly growing filaments formed. FIG. 33. Filaments formed by FBD12-3 and FB2 combination usually had two nuclei (arrows) but occasionally more (arrowheads). Magnification: Figs. 29, 31, 32 approx 3003; Figs. 30, 33 approx 8003.
Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
310
Snetselaar, Bo¨lker, and Kahmann
FIGS. 34–38. Mating interactions between diploid a5bfi sporidia and haploid sporidia. FIG. 34. DAPI-stained hyphae from a drop containing diploid FBD12-17 sporidia and FB1 sporidia were uninucleate (arrows). FIG. 35. When FBD12-17 sporidia were combined with FB1 sporidia, straight filaments quickly grew away from the agar surface. FIG. 36. When FBD12-17 sporidia were placed next to FB1 sporidia, the diploid sporidia formed straight aerial filaments (arrows), and the FB1 sporidia formed short mating hyphae (arrowheads). FIG. 37. When FBD11-7 sporidia were combined in a drop with FB2 sporidia, the straight aerial filaments elongated slightly more slowly than those found in the FBD12-17 3 FB1 combination. FIG. 38. When FBD11-7 sporidia were placed near drops of FB2 sporidia, the diploid strain formed filaments (arrows) much more slowly than the FBD12-17 strain. Note the mating hyphae forming on the FB2 sporidia (arrowhead). Magnification: Fig. 34 approx 8003; Figs. 35–38 approx 3003.
1993). Remarkably, results presented here indicate that U. maydis mating hyphae can respond to changes in pheromone concentration by reorienting their direction of growth in only 10 min. The rapid growth of the mating hyphae and their sensitivity to pheromone indicate that
Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
this system could be a fertile one for investigating hyphal tip growth in response to external stimuli. In all tests involving paired drops of sporidia, cells with a2 alleles produced mating hyphae before cells with a1 alleles did. As the distance between compatible drops was
311
Ustilago maydis Pheromone Responses
increased or the concentration of sporidia was decreased, the effect became more pronounced. This difference in response to pheromone is clearly due to the a alleles, because MU18 cells, in which the a2 allele was replaced with an a1 allele, behaved as a1 cells in that they induced long mating hyphae in an a2 strain (Fig. 14). The same faster response of a2 strains was observed for formation of aerial filaments in the a5bfi diploids upon pheromone stimulation. It may be that the a2 pheromone is produced in lesser quantities, diffuses more slowly, is inactivated more quickly, has a lower affinity for its receptor, or some combination of these possibilities. The a2 pheromone has also been more difficult to isolate from culture filtrates (Spellig et al., 1994). The speed at which the a1 cells turned away from newly formed dikaryotic hyphae (Figs. 22–25) makes it seem likely that secreted a2 pheromone is inactivated quickly. Transcriptional rates of pheromone and receptor genes drop in the presence of the active b heterodimer (Urban et al., 1996a), which presumably forms immediately when cells fuse to form the dikaryon. If secreted a2 pheromone is rapidly internalized or degraded, extracellular pheromone levels could then drop quickly so that the unmated a1 hyphae would not be attracted to the dikaryon. In S. cerevisiae, a cells secrete a protein known as BAR1 that inactivates the pheromone produced by a cells (Hicks and Herskowitz, 1976). BAR1 encodes a protease (Mackay et al., 1988) that is expressed only in a cells and is negatively regulated in a cells and diploid cells by the Mata2 gene product (Sprague and Herskowitz, 1981). Although it is possible that a protease is also involved in degrading the mfa2 protein in U. maydis, if this is so the regulation must be quite different. The difference in pheromone response between a1 and a2 cells is apparently located in the a locus, which does not encode regulatory proteins. At present, it is not clear whether the differences in response of a1 and a2 cells occur by chance or confer some selective advantage. Based on observations of numerous mating interactions, however, we speculate that a mating system whereby one pheromone is most effective from long distances and the other is most effective as compatible cells approach each other might lead to more efficient mating interactions. If both mating types were equally sensitive to pheromone and responded by growing very quickly toward the pheromone source, mating hyphae could overshoot potential partners. This type of response has been observed when sporidial concentrations (and thus presumably pheromone concentrations) were very high (Snetselaar, unpublished observations). Differences in sen-
sitivity to pheromone could occur in several different ways, some of which have been mentioned here. It will be interesting to decipher the language by which these cells communicate as more mutants and techniques for studying them become available.
ACKNOWLEDGMENTS This work was partially supported by the Deutsche Forschungsgemeinschaft through SFB 369. We are grateful to M. Urban for helpful discussion and for making many of the mutants available. Thanks to M. McCann for critically reading the manuscript.
REFERENCES Banuett, F., and Herskowitz, I. 1994. Morphological transitions in the life cycle of Ustilago maydis and their genetic control by the a and b loci. Exp. Mycol. 18: 247–267. Banuett, F., and 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: 5878–5882. Bo¨lker, M., and Kahmann, R. 1993. Sexual pheromones and mating responses in fungi. Plant Cell 5: 1461–1469. Bo¨lker, M., Urban, M., and Kahmann, R. 1992. The a mating type locus of U. maydis specifies cell signaling components. Cell 68: 441–450. Christensen, J. J. 1963. Corn Smut Caused by Ustilago maydis. Monogr. No. 2. Am. Phytopathol. Soc., St. Paul, MN. Gillissen, B., Bergemann, J., Sandmann, C., Schroeer, B., Bo¨lker, M., and Kahmann, R. 1992. A two-component regulatory system for self/nonself recognition in Ustilago maydis. Cell 68: 647–657. Hicks, J. B., and Herskowitz, I. 1976. Evidence for a new diffusible element of mating pheromones in yeast. Nature 260: 246–248. Holliday, R. 1961. Induced mitotic crossing over in Ustilago maydis. Genet. Res. 2: 231–248. Holliday, R. 1974. Ustilago maydis. In Handbook of Genetics (R. C. King, Ed.), Vol. 1, pp. 575–595. Plenum, New York. Jackson, C., and Hartwell, L. 1990a. Courtship in Saccharomyces cerevisiae: An early cell–cell interaction during mating. Mol. Cell. Biol. 10: 2202–2213. Jackson, C., and Hartwell, L. 1990b. Courtship in S. cerevisiae: Both cell types choose mating partners by responding to the strongest pheromone signal. Cell 63: 1039–1051. Ka¨mper, J., Reichmann, M., Romeis, T., Bo¨lker, M., and Kahmann, R. 1995. Multiallelic recognition: Nonself-dependent dimerization of the bE and bW homeodomain proteins in Ustilago maydis. Cell 81: 73–83. MacKay, V. L., Welch, S. K., Insley, M. Y., Manney, T. R., Holly, J., Saari, G. C., and Parker, M. L. 1988. The Saccharomyces cerevisiae BAR1 gene encodes an exported protein with homology to pepsin. Proc. Natl. Acad. Sci. USA 85: 55–59. Rowell, J. B. 1955. Functional role of compatibility factors and an in vitro
Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
312 test for sexual compatibility with haploid lines of Ustilago zeae. Phytopathology 45: 370–374. Segall, J. 1993. Polarization of yeast cells in spatial gradients of a mating factor. Proc. Natl. Acad. Sci. USA 90: 8332–8338. Snetselaar, K. M. 1993. Microscopic observation of Ustilago maydis mating interactions. Exp. Mycol. 17: 345–355. Snetselaar, K. M., and Mims, C. W. 1993. Infection of maize stigmas by Ustilago maydis: Light and electron microscopy. Phytopathology 83: 843–850. Snetselaar, K. M., and Mims, C. W. 1992. Sporidial fusion and infection of maize seedlings by the smut fungus Ustilago maydis. Mycologia 84: 193–203. Spellig, T., Bo¨lker, M., Lottspeich, F., Frank, R., and Kahman, R. 1994.
Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
Snetselaar, Bo¨lker, and Kahmann
Pheromones trigger filamentous growth in Ustilago maydis. EMBO J. 13: 1620–1627. Sprague, G. F., and Herskowitz, I. 1981. Control of yeast cell type by the mating type locus. I. Identification and control of expression of the a-specific gene BAR1. J. Mol. Biol. 153: 305–321. Tsukuda, T., Carleton, S., Fotheringham, S., and Holloman, W. K. 1988. Mol. Cell. Biol. 8: 3703–3709. Urban, M., Kahmann, R., and Bo¨lker, M. 1996a. Identification of the pheromone response element in Ustilago maydis. Mol. Gen. Genet. 251: 31–37. Urban, M., Kahmann, R., and Bo¨lker, M. 1996b. The biallelic a mating type locus of Ustilago maydis: Remnants of an additional pheromone gene indicate evolution from a multiallelic ancestor. Mol. Gen. Genet., 250: 414–420.
Ustilago maydis Mating Hyphae Orient Their Growth toward Pheromone Sources
Karen M. Snetselaar,*,1 Michael Bo ¨ lker,† and Regine Kahmann† *Department of Biology, St. Joseph’s University, 5600 City Avenue, Philadelphia, Pennsylvania 19131; and †Institut fu¨r Genetik und Mikrobiologie, Universita¨t Mu¨nchen, Maria-Ward-Str. 1a, 80638 Munich, Germany
Accepted for publication October 18, 1996
Snetselaar, K. M., Bo¨lker, M., and Kahmann, R. 1996. Ustilago maydis mating hyphae orient their growth toward pheromone sources. Fungal Genetics and Biology 20, 299–312. When small drops of Ustilago maydis sporidia were placed 100–200 mm apart on agar surfaces and covered with paraffin oil, sporidia from one drop formed thin hyphae that grew in a zig-zag fashion toward the other drop if it contained sporidia making the appropriate pheromone. For example, a2b2 mating hyphae grew toward a1b1 and a1b2 mating hyphae, and the filaments eventually fused tip to tip. Time-lapse photography indicated that the mating hyphae can rapidly change orientation in response to nearby compatible sporidia. When exposed to pheromone produced by cells in an adjacent drop, haploid sporidia with the a2 allele began elongating before sporidia with the a1 allele. Sporidia without functional pheromone genes responded to pheromone although they did not induce a response, and sporidia without pheromone receptors induced formation of mating hyphae although they did not form mating hyphae. Diploid sporidia heterozygous at b but not at a formed straight, rigid, aerial filaments when exposed to pheromone produced by the appropriate haploid sporidia. Again, the a2a2b1b2 strain formed filaments more quickly than the a1a1b1b2 strain. Taken together, these results suggest that the a2 pheromone diffuses less readily or is degraded more quickly than the a1 pheromone. r 1996 Academic Press
1
To whom reprint requests should be addressed.
1087-1845/96 $18.00 Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
Index Descriptors: Ustilago; pheromone; mating; chemotaxis. Maize smut disease is caused by the fungus Ustilago maydis (DC.) Corda. The disease cycle begins when compatible haploid sporidia cease budding growth and form mating hyphae (also called conjugation tubes) that fuse to form a dikaryotic filament. The obligately parasitic dikaryon can enter the plant in any meristematic area. After limited hyphal proliferation around the infection site, the fungus induces abnormal cell division and enlargement in the host, resulting in gall formation. Hyphal walls of dikaryotic fungal cells in the galls swell, gelatinize, and separate at the crosswalls to form teliospores that contain a single diploid nucleus at maturity. These thick-walled spores are disseminated when the galls break open, and they can overwinter in soil or plant debris. Meiosis occurs when the teliospores germinate and form saprobic haploid sporidia that can begin the infection cycle anew (see Christensen, 1963, for further life history details). Classical genetic studies showed that U. maydis has an unusual tetrapolar mating system (see Holliday, 1974, for review). The a locus, which has two alleles, controls sporidial fusion (Rowell, 1955) while the b locus, which has many alleles, regulates the ability of the fungus to cause disease on the host (Holliday, 1961). Alleles from both the a and the b loci have been cloned, sequenced, and characterized. The a1 and a2 alleles each contain a pheromone gene (mfa) and a pheromone receptor gene (pra). mfa and pra mutants, which have nonfunctional pheromones or pheromone receptors, were constructed and used to show that these genes are required for mating
299
300
(Bo¨lker et al., 1992; Spellig et al., 1994). In addition the a2 allele encodes two genes with unknown functions (Urban et al., 1996b). Each b allele encodes two homeodomain proteins able to form heterodimers if derived from different alleles (Gillissen et al., 1992; Ka¨mper et al., 1995). Morphological and developmental information about sporidial mating and infection in U. maydis has come more slowly. Indeed, the genetic evidence for pheromones and pheromone receptors preceded any morphological evidence of a role for chemical attractants in the mating process (Bo¨lker and Kahmann, 1993). Recently, however, assays for formation of mating hyphae (Banuett and Herskowitz, 1994) and for mating and infection hyphae (Snetselaar and Mims, 1992) have been described. These assays allowed microscopic confirmation that haploid sporidia with unlike a alleles fuse regardless of b, but unlike b alleles are required for development of infection hyphae (Snetselaar, 1993), confirming earlier observations made by Rowell (1955). The behavior of mating and infection hyphae have also been observed on host cells (Snetselaar and Mims, 1993). When induced by starvation, diploid sporidia heterozygous at both a and b formed infection hyphae without mating, but diploid sporidia heterozygous at a but not at b mated repeatedly unless fusion occurred between cells with different b alleles, at which time infection hyphae formed (Snetselaar, 1993). In addition, pheromone isolated from culture filtrates has now been used to induce formation of mating hyphae on plates (Spellig et al., 1994). This assay was used to purify the secreted form of both pheromones. Peptide sequencing and mass analysis showed that mature a1 pheromone consists of 13 amino acids and a2 pheromone of 9 amino acids, and both pheromones are farnesylated at their C-terminal cysteine residue (Spellig et al., 1994). The mating assays described so far have been useful, but they involve combining sporidia in solutions so it is not possible to distinguish one genotype from another or to determine whether mating behavior differs between partners. The assay described here allowed us to observe sporidia of known genotypes as they interacted, providing new information about the roles of pheromones and pheromone receptors in haploid and diploid U. maydis cells.
MATERIALS AND METHODS All strains (Table 1) were maintained on Difco potato dextrose agar and transferred into flasks containing 10 ml
Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
Snetselaar, Bo¨lker, and Kahmann
TABLE 1 Strains Used Strain
Genotype
FB1
alb1
FB2
a2b2
FB6a
a2b1
FB6b
a1b2
FBD11-7
a1/a1 b1/b2
FBD12-17
a2/a2 b1/b2
FBD12-3
a1/a2 b1/b1
FBD11-21
a1/a2 b2/b2
RK1785 (derived from FB2) RK1786 (derived from FB2) MU161 (derived from FB2) MU162 (derived from FB2) MU18 (derived from FB2)
a1 mfal
Source Banuett and Herskowitz, 1989 Banuett and Herskowitz, 1989 Banuett and Herskowitz, 1989 Banuett and Herskowitz, 1989 Banuett and Herskowitz, 1989 Banuett and Herskowitz, 1989 Banuett and Herskowitz, 1989 Banuett and Herskowitz, 1989 Bo¨lker et al., 1992 (see below) Bo¨lker et al., 1992 (see below) M. Urban, unpublished (see below) M. Urban, unpublished (see below) M. Urban, unpublished (see below)
Note. Strains RK1785 and RK1786 carry mutant a1 alleles in the FB2 genetic background. MU161 carries the hygromycin resistance cassette isolated as 3-kb PvuII fragment of pCM54 (Tsukuda et al., 1988) inserted into the MscI site in the open reading frame of mfa2; in MU162 the same cassette is inserted into the SnaBI within the open reading frame of pra2. MU18 carries a Tn5H insertion in the pan1 gene that is located adjacent to the a locus. The insertion was generated as described (Bo¨lker et al., 1992) in the 10-kb BamHI fragment that contains the complete a1 allele with flanking sequences. The mutation was introduced into FB2 by transformation, and transformants were checked for replacement of the resident a2 allele by the a1 allele by Southern analysis.
of Difco potato dextrose broth to grow sporidia for experiments. Flasks were shaken at 200 rpm for 12–18 h at 28°C under constant illumination. To set up mating experiments, small aliquots of cells from cultures in mid to late log phase were centrifuged in a clinical centrifuge at 4000 rpm for 3 min, and the supernatant was withdrawn and replaced with sterile water, in which the sporidia were resuspended to reach a concentration of approximately 105/ml. Mating experiments were performed by placing drops of sporidia very close together on an agar surface. The following procedure for preparing cultures was found to be efficient and suitable for subsequent microscopic observa-
301
Ustilago maydis Pheromone Responses
tions. Microscope slides were placed into large petri dishes and covered with hot 2% water agar to a depth of approximately 2 mm. After the agar cooled, slides with their agar coating were removed from the dishes by slicing the agar around them with a razor blade, and a small section of agar at one edge of the slides was removed to facilitate handling and labelling. A 10-µl micropipettor was used to place 0.5-µl drops of sporidial suspensions on the agar films. About 15 drops of one mating type were placed in rows on the slide. After approximately 5 min, the fluid from the drop had been absorbed into the agar, making it easier to place a second drop within 100–200 µm of the first. After the water in the second drop had been absorbed into the agar, those pairs of drops that were very close together but not touching were covered with 3 µl of paraffin oil (Merck). Generally, about 10 of the 15 pairs of drops would be unsuitable, either because the drops ran together or because they were farther apart than 200 µm. Two different combinations could be prepared on a single microscope slide, and five slides could be prepared in less than an hour. For some experiments, sporidia were combined in a single drop. Plastic dishes containing the slides were covered to prevent drying out and incubated at 28°C until they were observed microscopically. Mating interactions were observed with a Zeiss Axiophot microscope with differential interference contrast capability. Objectives of 103 and 203 were used without coverslips, so observations could be made over a period of time without disturbing the sporidia. For higher magnifications, coverslips were simply placed over the oil-covered drops. Growth continued for some time after coverslips were added, but generally slowed appreciably after about an hour. In order to observe nuclear condition in mating sporidia, the paraffin oil was removed from the slide and replaced with a DAPI2 solution prepared as previously described (Snetselaar, 1993). Although observations could be made directly on the agar-coated slide, usually it was more convenient to remove the sporidia and DAPI solution to a clean slide.
RESULTS Microscopic observations of sporidia combined in drops provided more detail than the plate-mating assays commonly used to assess mating, and examination of sporidia in adjacent drops allowed observation of previously unde2
Abbreviation used: DAPI, 48,6-diamidino-2-phenylindole dihydrochloride.
scribed mating behavior. All experiments described here were repeated from two to several times. There was some variation among experiments in the length of time from when the sporidial crosses were set up until the first mating hyphae were observed. Therefore, positive and negative controls for mating behavior (FB1 3 FB2 and FB1 3 FB6b) were included with each experiment, and rates of hyphal formation and elongation were only compared within a particular experiment. Note that in all figures showing interaction between sporidia in separate drops, the a2 strain is in the top drop.
Haploid Sporidia with Unlike a Alleles Form Mating Hyphae Regardless of b No mating hyphae formed when drops of FB6a (a2b1) sporidia were placed next to FB2 (a2b2) sporidia (Fig. 1) or when these sporidia were combined in a single drop (Fig. 2). DAPI-stained sporidia from these drops were yeast-like and uninucleate (Fig. 3). When a drop of FB1 (a1b1) sporidia was placed next to a drop of FB6a (a2b1) sporidia, both types of sporidia had mating hyphae after approximately 8 h (Fig. 4). When combined in a single drop, these sporidia fused and produced short, contorted, multinucleate hyphae (Figs. 5 and 6) but no straight aerial filaments. When a drop of FB1 (a1b1) sporidia was placed near a drop of FB2 (a2b2) sporidia, again both types of sporidia formed mating hyphae after approximately 8 h (Fig. 7). However, straight, rapidly growing, aerial filaments eventually formed in the area where the mating hyphae met, as well as in drops where FB1 and FB2 were combined (Fig. 8). The filaments were dikaryotic (Fig. 9). All combinations of the four tester strains gave similar results; mating hyphae formed and fused in all combinations involving unlike a alleles, while unlike alleles at both a and b were required for the formation of the aerial dikaryotic filaments. In addition, sporidia with the a2 allele began forming mating hyphae earlier than sporidia with the a1 allele, regardless of which b allele was present.
Mutants with Disrupted Mating Factor Genes Respond to Pheromone but They Cannot Induce a Pheromone Response, while Mutants with Disrupted Pheromone Receptor Genes Induce Pheromone Responses although They Cannot Respond to Pheromone Strains with nonfunctional pheromones or pheromone receptors were used to further investigate the formation of
Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
FIGS. 1–9. Mating of wild-type U. maydis sporidia. FIG. 1. No mating hyphae form between drops of sporidia with the same a allele [FB2 (top) and FB6a]. FIG. 2. No filaments form when FB2 and FB6a sporidia are in a single drop. FIG. 3. DAPI-stained cells from FB2 3 FB6a drop are uninucleate. FIG. 4. Mating hyphae form between adjacent drops of FB2 (top) and FB6b sporidia. FIG. 5. Convoluted filaments form (arrows) when FB2 and FB6b are combined in a single drop. FIG. 6. DAPI-stained filaments from FB2 3 FB6b are multinucleate. FIG. 7. Mating hyphae form between drops of FB1 and FB2 (top) sporidia. FIG. 8. Long straight filaments form (arrows) when FB1 and FB2 sporidia are combined in a single drop. FIG. 9. Filaments from the FB1 3 FB2 cross are dikaryotic (arrows). Magnification: Figs. 1, 2, 4, 5, 7, 8 approx 2503; Figs. 3, 6, 9 approx 8003.
302
303
Ustilago maydis Pheromone Responses
mating hyphae. Strain MU161, which produces an a2 receptor but no pheromone, did not induce formation of mating hyphae in FB1 sporidia, although MU161 sporidia responded to FB1 sporidia by producing mating hyphae (Fig. 10). Strain RK1786, which produces a1 pheromone but has a disrupted receptor allele, did not respond to pheromone produced by FB2, but RK1786 sporidia induced formation of mating hyphae in FB2 sporidia (Fig. 11). Similar results were observed with RK1785 (which has a disrupted a1 pheromone gene) and MU162 (which has a disrupted a2 receptor) except that the response of a1 mating hyphae was always less vigorous than the a2 response (Figs. 12 and 13). The initial responses of MU161 to FB1 and of FB2 to RK1786 seemed to occur as rapidly as in FB1 3 FB2. Therefore, it seems that basal levels of pheromone production are sufficient to induce at least the early stages of the mating program in compatible strains.
Haploid a1 and a2 Strains Respond to Pheromone Differently In all tests of wild-type and mutant sporidia, mating hyphae with the a2 allele elongated more quickly than mating hyphae with the a1 allele, especially when drops of sporidia were not very close together or sporidial concentrations were low. Several tests were performed in order to determine whether this response difference was inherent in the a alleles or due to the genetic background. Six a1 strains and six a2 strains isolated from three different maize fields near Munich were tested against FB1, FB2, and each other. Although there was some variation in how quickly the mating hyphae formed, in no case did the a1 strains respond more strongly than the a2 strains. Strain MU18, an FB2 derivative in which the entire a2 locus has been replaced with an a1 locus, was placed in a drop near a1 and a2 strains. MU18 behaved as an a1 strain, inducing a more rapid response in the a2 strain (Fig. 14). These results indicate that the difference in response to pheromone is inherent in the different alleles of the a locus.
paraffin oil (not shown). In similar drops containing six-fold dilutions of the same sporidial strains, there were no mating hyphae after 5 h (Fig. 17) and only a few after 10 h (Fig. 18). Mating hyphae also formed later when drops were farther apart (data not shown). A likely explanation for this is that it takes longer for pheromone to diffuse from a small source or from a greater distance in quantities adequate to cause a response. Therefore, in all comparative experiments described here, drops containing approximately equivalent numbers of sporidia were placed about the same distance apart.
Mating Hyphae Orient Their Growth toward the Source of Pheromone In order to ascertain whether mating hyphae orient to a pheromone gradient rather than forming randomly in response to pheromone stimulation, the RK1786 strain (which lacks a1 pheromone receptors but still produces basal levels of a1 pheromone and thus induces FB2 strains to form mating hyphae, see Fig. 11) was used as pheromone donor (Fig. 19). When FB2 sporidia were combined with an equal number of RK1786 sporidia and placed in a drop adjacent to a drop containing only RK1786 sporidia, some of the FB2 sporidia still formed mating hyphae directed toward the RK1786 drop (Fig. 20), presumably because the pheromone concentration diffusing from the RK1786 drop became higher than the concentration produced in the drop containing both FB2 and RK1786 sporidia. This indicates that the FB2 mating hyphae were following a pheromone gradient, not randomly elongating in response to pheromone. When each drop contained equal numbers of RK1786 and FB2 sporidia, no mating hyphae formed at the edges of either drop (Fig. 21), although there were mating hyphae within the drops (not shown). The FB2 sporidia were surrounded by pheromone at a concentration as high or higher than that diffusing from the adjacent drop, so there was no pheromone gradient to induce growth of mating hyphae out of the drops.
The Rate at which Mating Hyphae Elongate Depends on the Size and Location of the Pheromone Source
Wild-Type Mating Hyphae Can Quickly Reorient Their Direction of Growth toward Compatible Mating Partners
When FB1 and FB2 sporidia at a concentration of approximately 105/ml were placed in drops approximately 100 µm apart, mating hyphae began to form within 5 h (Fig. 15) and by 10 h many had formed, grown together (Fig. 16), and formed aerial filaments that grew up into the
A time-lapse series (Figs. 22–25) shows interactions between FB1 and FB2 mating hyphae that are consistent with extreme sensitivity to changes in pheromone concentrations. The small arrows show how precisely the mating hyphae orient prior to fusion; the high magnification inset
Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
FIGS. 10–18. Interactions between U. maydis sporidia in adjacent drops. The top drop in all figures contains a2 sporidia. FIG. 10. MU161 3 FB1, note mating hyphae formed by the MU161 strain. FIG. 11. FB2 3 RK1786, note mating hyphae formed by FB2. FIG. 12. FB2 3 RK1785. The RK1785 strain forms mating hyphae, but fewer than the MU161 strain (see Fig. 10) even in these drops that are very close together. FIG. 13. MU162 3 FB1. Again, the response of the FB1 strain to the pheromone produced by MU162 is much weaker than the FB2 response (Fig. 11). FIG. 14. Cross between FB2 and MU18 (bottom), an FB2 strain which has had the entire a2 allele replaced with an a1 allele. This strain now behaves as an a1 strain and induces the a2 strain to form longer filaments. FIG. 15. FB1 3 FB2 sporidia at approximately 105/ml showed mating hyphae beginning to form after 5 hr. FIG. 16. Same sporidia as Fig. 14, 5 h later. FIG. 17. FB1 3 FB2 sporidia, same cells as for Fig. 14 but diluted sixfold. No mating hyphae had formed after 5 h. FIG. 18. Same as Fig. 17, 5 h later. A few mating hyphae had formed. Magnification: All figures approx 2503.
304
Ustilago maydis Pheromone Responses
305
FIGS. 19–21. Interactions between drops of U. maydis sporidia. FIG. 19. FB2 (top) 3 FB1 cross. FIG. 20. Top drop contains equal numbers of FB2 and RK1786 sporidia, bottom is entirely RK1786. Note that some FB2 sporidia are still attracted to the RK1786 drop, where the pheromone concentration is presumably higher. FIG. 21. Both drops contain equal amounts of FB2 and RK1786 sporidia; there is no pheromone gradient attracting the FB2 sporidia from either drop. Magnification: All figures approx 2503.
provides a closer look at the cells fusing. As the FB2 hyphae approached them, the FB1 hyphae began to grow, and they often seemed to respond dramatically to nearby FB2 hyphae. The large arrows in Figs. 22–25 show two examples of how FB1 hyphae that had been turning toward approaching FB2 hyphae changed direction of growth after the FB2 hypha fused with another FB1 partner. In both cases, the change in direction occurred in less than 10 min.
Although mfa Strains Are Attracted to a Pheromone Source, They Are Unable to Precisely Locate and Fuse with Compatible Partners To further investigate the roles of pheromones and pheromone receptors on later stages in the mating process,
a drop containing equal numbers of RK1785 and RK1786 sporidia (both mutants in a1 alleles) was placed near a drop containing MU161 and MU162 sporidia (both mutants in a2 alleles). Therefore, half of the cells in each drop produced pheromone but could not sense pheromone, while the other half could respond to pheromone but could not produce it. A time-lapse sequence lasting 45 min is shown in Figs. 26–28. At first, mating hyphae grew relatively normally. However, as the MU161 hyphae approached the RK1785 hyphae, they did not fuse with them as wild-type mating hyphae generally do, but instead they grew past the potential fusion partners as though they could not detect their presence. Likewise, the RK1785 mating hyphae did not turn toward nearby MU161 mating hyphae the way the wild-type a1 strains did. No cell fusions were observed in these crosses.
Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
FIGS. 22–25. Time-lapse series of mating interactions between FB2 (top) and FB1 (bottom) U. maydis sporidia. Time elapsed between Figs. 22 and 23 is 12 min, between Figs. 23 and 24 is 6 min, between Figs. 24 and 25 is 24 min. Arrows and arrowheads point out the same area in each frame. Small arrows follow compatible mating hyphae as they orient precisely before fusing at their tips. Large arrows show two examples of FB1 mating hyphae changing growth direction away from a newly formed dikaryon. Arrowheads and inset show compatible hyphae fusing at the tips. Magnification: Figs. 22–25 approx 4003; insets approx 12003.
306
307
Ustilago maydis Pheromone Responses
Diploid Strains Heterozygous at a and Homozygous at b Do Not Exhibit Directional Growth but Can Fuse with Compatible Haploid Strains Strains heterozygous at a but not at b formed short, convoluted filaments when placed in drops alone (not shown) that looked much like those seen in afib5 crosses between haploids (Fig. 5). When the diploid strain FBD12-3 (a1a2b1b1) was combined in a drop with FB1, the filaments sometimes became quite long, but they remained convoluted and did not become aerial (Fig. 29). DAPI staining showed that these contorted filaments often contained multiple nuclei (Fig. 30). When the two strains were placed in adjacent drops, there was little evidence of attraction on either side, and the diploid strain continued its highly contorted growth (Fig. 31). When FB2 was combined in a single drop with FBD12-3, straight, rapidly growing filaments resulted (Fig. 32). DAPI staining showed that these straight filaments had two and sometimes more nuclei (Fig. 33).
Diploid Strains FBD11-7 (a1a1b1b2) and FBD12-17 (a2a2b1b2) Become Filamentous in the Presence of Compatible Haploid Cells, but FBD12-17 Cells Respond More Quickly When placed in drops alone, diploid strains that were heterozygous at b but not at a remained almost entirely yeast-like (not shown) and had the same appearance as combined a5 haploid strains (Figs. 2 and 3). When one of these diploid strains was combined in a drop with a haploid strain carrying the complementary a allele, aerial filaments formed. These filaments were straight and rigid, unlike the mating filaments formed by haploid strains, and unlike the convoluted filaments formed by the afib5 diploid strains. DAPI staining showed the filaments were uninucleate (Fig. 34). Filaments formed by the FBD12-17 3 FB1 combination were only slightly longer than those formed by the FBD11-7 3 FB2 combination (Figs. 35 and 37), but differences were more dramatic when sporidia were placed in adjacent drops (Figs. 36 and 38). FBD12-17 rapidly became quite filamentous when placed next to FB1, while FBD11-7 elongated slowly when placed next to FB2; indeed under these circumstances the FB2 usually formed mating hyphae, indicating that the diploid strain was secreting a1 pheromone.
DISCUSSION The responses described here occurred consistently when sporidia were covered with paraffin oil, but they do not require the paraffin oil. Weak interactions between mating hyphae were previously observed in drops of sporidia placed close together on water agar, and a2 strains elongated more rapidly than a1 strains (Snetselaar, unpublished). We envision that diffusion of the amphipathic pheromone is greatly facilitated at the interface between the hydrophilic water agar and the hydrophobic paraffin oil. It is interesting to note that similar interfaces may occur between hydrophobic plant surfaces and the saturated atmosphere in the leaf whorl or ear, common sites of infection in maize. These results show for the first time that wild-type U. maydis cells locate and approach each other in a highly coordinated fashion as a prelude to cell fusion. The responses of the pra and mfa mutants indicate that pheromone responses act in at least two ways before the cells fuse. Expression of the pheromone genes is induced in certain complex media and by starvation (Spellig et al., 1994; M. Bo¨lker, unpublished), and we have shown here that this level of expression is sufficient to cause mating hyphae to form in haploid sporidia. Transcription of the mfa and pra genes is more strongly stimulated by the appropriate pheromone (Urban et al., 1996a). These induced levels of pheromone are not essential for the formation of mating hyphae, because mutants without receptors can induce mating hyphae even though they cannot detect pheromone. However, the induced pheromone levels may become important in helping mating hyphae ‘‘fine-tune’’ growth as they closely approach a potential mating partner. Mating hyphae that produce no pheromone do not induce increased pheromone production in compatible strains, and under these conditions they apparently cannot detect nearby possible mating partners against a background of pheromone. There are many similarities between the premating behavior of U. maydis mating hyphae and what has been called ‘‘courtship’’ in Saccharomyces cerevisiae (Jackson and Hartwell, 1990a). Potential S. cerevisiae mating partners communicate by producing and receiving pheromone, and polarized growth of mating cells in response to pheromone gradient was suggested to be the outcome of courtship (Jackson and Hartwell, 1990b). Subsequent work demonstrated that yeast cells do indeed respond to a pheromone gradient by orienting mating structures toward a pheromone source over a period of several hours (Segall,
Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
FIGS. 26–28. Time-lapse series of interactions between U. maydis sporidia. Top drop contains equal numbers of MU161 and MU162 sporidia, bottom drop contains equal numbers of RK1785 and RK1786 sporidia. Thus, individual sporidia in each drop make either pheromone or receptors, but not both. Note absence of fusions between mating hyphae. Inset in Fig. 28 is higher magnification of the boxed area. Magnification: Figs. 26–28 approx 4003; inset approx 1003.
308
Ustilago maydis Pheromone Responses
309
FIGS. 29–33. Mating interactions between diploid afib5 sporidia and haploid sporidia. FIG. 29. When FBD12-3 (a1/a2b1/b1) sporidia were combined with FB1 sporidia, flat, convoluted filaments formed; compare with Fig. 5. FIG. 30. DAPI staining showed that some of the filaments were multinucleate. FIG. 31. When FBD12-3 sporidia (bottom drop) were placed near FB2 sporidia, little attraction between drops was seen. Note the extremely curved FBD12-3 sporidia (arrows). FIG. 32. When FBD12-3 and FB2 were placed in the same drop, straight, aerial, rapidly growing filaments formed. FIG. 33. Filaments formed by FBD12-3 and FB2 combination usually had two nuclei (arrows) but occasionally more (arrowheads). Magnification: Figs. 29, 31, 32 approx 3003; Figs. 30, 33 approx 8003.
Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
310
Snetselaar, Bo¨lker, and Kahmann
FIGS. 34–38. Mating interactions between diploid a5bfi sporidia and haploid sporidia. FIG. 34. DAPI-stained hyphae from a drop containing diploid FBD12-17 sporidia and FB1 sporidia were uninucleate (arrows). FIG. 35. When FBD12-17 sporidia were combined with FB1 sporidia, straight filaments quickly grew away from the agar surface. FIG. 36. When FBD12-17 sporidia were placed next to FB1 sporidia, the diploid sporidia formed straight aerial filaments (arrows), and the FB1 sporidia formed short mating hyphae (arrowheads). FIG. 37. When FBD11-7 sporidia were combined in a drop with FB2 sporidia, the straight aerial filaments elongated slightly more slowly than those found in the FBD12-17 3 FB1 combination. FIG. 38. When FBD11-7 sporidia were placed near drops of FB2 sporidia, the diploid strain formed filaments (arrows) much more slowly than the FBD12-17 strain. Note the mating hyphae forming on the FB2 sporidia (arrowhead). Magnification: Fig. 34 approx 8003; Figs. 35–38 approx 3003.
1993). Remarkably, results presented here indicate that U. maydis mating hyphae can respond to changes in pheromone concentration by reorienting their direction of growth in only 10 min. The rapid growth of the mating hyphae and their sensitivity to pheromone indicate that
Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
this system could be a fertile one for investigating hyphal tip growth in response to external stimuli. In all tests involving paired drops of sporidia, cells with a2 alleles produced mating hyphae before cells with a1 alleles did. As the distance between compatible drops was
311
Ustilago maydis Pheromone Responses
increased or the concentration of sporidia was decreased, the effect became more pronounced. This difference in response to pheromone is clearly due to the a alleles, because MU18 cells, in which the a2 allele was replaced with an a1 allele, behaved as a1 cells in that they induced long mating hyphae in an a2 strain (Fig. 14). The same faster response of a2 strains was observed for formation of aerial filaments in the a5bfi diploids upon pheromone stimulation. It may be that the a2 pheromone is produced in lesser quantities, diffuses more slowly, is inactivated more quickly, has a lower affinity for its receptor, or some combination of these possibilities. The a2 pheromone has also been more difficult to isolate from culture filtrates (Spellig et al., 1994). The speed at which the a1 cells turned away from newly formed dikaryotic hyphae (Figs. 22–25) makes it seem likely that secreted a2 pheromone is inactivated quickly. Transcriptional rates of pheromone and receptor genes drop in the presence of the active b heterodimer (Urban et al., 1996a), which presumably forms immediately when cells fuse to form the dikaryon. If secreted a2 pheromone is rapidly internalized or degraded, extracellular pheromone levels could then drop quickly so that the unmated a1 hyphae would not be attracted to the dikaryon. In S. cerevisiae, a cells secrete a protein known as BAR1 that inactivates the pheromone produced by a cells (Hicks and Herskowitz, 1976). BAR1 encodes a protease (Mackay et al., 1988) that is expressed only in a cells and is negatively regulated in a cells and diploid cells by the Mata2 gene product (Sprague and Herskowitz, 1981). Although it is possible that a protease is also involved in degrading the mfa2 protein in U. maydis, if this is so the regulation must be quite different. The difference in pheromone response between a1 and a2 cells is apparently located in the a locus, which does not encode regulatory proteins. At present, it is not clear whether the differences in response of a1 and a2 cells occur by chance or confer some selective advantage. Based on observations of numerous mating interactions, however, we speculate that a mating system whereby one pheromone is most effective from long distances and the other is most effective as compatible cells approach each other might lead to more efficient mating interactions. If both mating types were equally sensitive to pheromone and responded by growing very quickly toward the pheromone source, mating hyphae could overshoot potential partners. This type of response has been observed when sporidial concentrations (and thus presumably pheromone concentrations) were very high (Snetselaar, unpublished observations). Differences in sen-
sitivity to pheromone could occur in several different ways, some of which have been mentioned here. It will be interesting to decipher the language by which these cells communicate as more mutants and techniques for studying them become available.
ACKNOWLEDGMENTS This work was partially supported by the Deutsche Forschungsgemeinschaft through SFB 369. We are grateful to M. Urban for helpful discussion and for making many of the mutants available. Thanks to M. McCann for critically reading the manuscript.
REFERENCES Banuett, F., and Herskowitz, I. 1994. Morphological transitions in the life cycle of Ustilago maydis and their genetic control by the a and b loci. Exp. Mycol. 18: 247–267. Banuett, F., and 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: 5878–5882. Bo¨lker, M., and Kahmann, R. 1993. Sexual pheromones and mating responses in fungi. Plant Cell 5: 1461–1469. Bo¨lker, M., Urban, M., and Kahmann, R. 1992. The a mating type locus of U. maydis specifies cell signaling components. Cell 68: 441–450. Christensen, J. J. 1963. Corn Smut Caused by Ustilago maydis. Monogr. No. 2. Am. Phytopathol. Soc., St. Paul, MN. Gillissen, B., Bergemann, J., Sandmann, C., Schroeer, B., Bo¨lker, M., and Kahmann, R. 1992. A two-component regulatory system for self/nonself recognition in Ustilago maydis. Cell 68: 647–657. Hicks, J. B., and Herskowitz, I. 1976. Evidence for a new diffusible element of mating pheromones in yeast. Nature 260: 246–248. Holliday, R. 1961. Induced mitotic crossing over in Ustilago maydis. Genet. Res. 2: 231–248. Holliday, R. 1974. Ustilago maydis. In Handbook of Genetics (R. C. King, Ed.), Vol. 1, pp. 575–595. Plenum, New York. Jackson, C., and Hartwell, L. 1990a. Courtship in Saccharomyces cerevisiae: An early cell–cell interaction during mating. Mol. Cell. Biol. 10: 2202–2213. Jackson, C., and Hartwell, L. 1990b. Courtship in S. cerevisiae: Both cell types choose mating partners by responding to the strongest pheromone signal. Cell 63: 1039–1051. Ka¨mper, J., Reichmann, M., Romeis, T., Bo¨lker, M., and Kahmann, R. 1995. Multiallelic recognition: Nonself-dependent dimerization of the bE and bW homeodomain proteins in Ustilago maydis. Cell 81: 73–83. MacKay, V. L., Welch, S. K., Insley, M. Y., Manney, T. R., Holly, J., Saari, G. C., and Parker, M. L. 1988. The Saccharomyces cerevisiae BAR1 gene encodes an exported protein with homology to pepsin. Proc. Natl. Acad. Sci. USA 85: 55–59. Rowell, J. B. 1955. Functional role of compatibility factors and an in vitro
Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
312 test for sexual compatibility with haploid lines of Ustilago zeae. Phytopathology 45: 370–374. Segall, J. 1993. Polarization of yeast cells in spatial gradients of a mating factor. Proc. Natl. Acad. Sci. USA 90: 8332–8338. Snetselaar, K. M. 1993. Microscopic observation of Ustilago maydis mating interactions. Exp. Mycol. 17: 345–355. Snetselaar, K. M., and Mims, C. W. 1993. Infection of maize stigmas by Ustilago maydis: Light and electron microscopy. Phytopathology 83: 843–850. Snetselaar, K. M., and Mims, C. W. 1992. Sporidial fusion and infection of maize seedlings by the smut fungus Ustilago maydis. Mycologia 84: 193–203. Spellig, T., Bo¨lker, M., Lottspeich, F., Frank, R., and Kahman, R. 1994.
Copyright r 1996 by Academic Press All rights of reproduction in any form reserved.
Snetselaar, Bo¨lker, and Kahmann
Pheromones trigger filamentous growth in Ustilago maydis. EMBO J. 13: 1620–1627. Sprague, G. F., and Herskowitz, I. 1981. Control of yeast cell type by the mating type locus. I. Identification and control of expression of the a-specific gene BAR1. J. Mol. Biol. 153: 305–321. Tsukuda, T., Carleton, S., Fotheringham, S., and Holloman, W. K. 1988. Mol. Cell. Biol. 8: 3703–3709. Urban, M., Kahmann, R., and Bo¨lker, M. 1996a. Identification of the pheromone response element in Ustilago maydis. Mol. Gen. Genet. 251: 31–37. Urban, M., Kahmann, R., and Bo¨lker, M. 1996b. The biallelic a mating type locus of Ustilago maydis: Remnants of an additional pheromone gene indicate evolution from a multiallelic ancestor. Mol. Gen. Genet., 250: 414–420.