Cla4, but not Rac1, regulates the filamentous response of Ustilago maydis to low ammonium conditions

Abstract

Ustilago maydis, the fungal pathogen of maize, undergoes a dimorphic transition from budding yeast-like growth to filamentous growth, both as part of its program for pathogenesis and distinctly, in response to environmental cues, such as acid pH or low nitrogen availability. Smu1 is a p21-activated protein kinase (PAK) with roles in both the mating response required for the former function, as well as for the nutrient response. Hsl7 may be a negative regulator of Smu1 and appears to play a role in cell length and cell cycle.  Additional proteins that participate in cell polarity and filamentation pathways include the small G protein, Rac1, and its effector PAK kinase, Cla4. Here we describe further experiments that explore the roles of Cla4 and Rac1 in the response to nitrogen availability. While deletion of rac1severely delays filamentous growth on solid media low in ammonium (SLAD), we found that deletion of cla4 does not abolish filamentous cell morphology on solid SLAD. Unexpectedly, however, the Dcla4 mutants also filament in liquid SLAD. The filamentous cell morphology of the cla4 mutant in liquid SLAD has only been seen previously for one other mutant, a strain deleted for hsl7 that simultaneously over-expresses smu1. 

Rho / Rac GTPases have been implicated in regulating cell polarity, cytoskeletal organization, and cytokinesis;^1-5 when activated, they modulate localization and activity of several downstream effectors, including p21-activated protein kinases (PAKs). The PAKs comprise a large family of serine/threonine protein kinases that regulate many cell processes including cell cycle, cytoskeletal organization, and polar growth.^4-6 As potential downstream effectors of the Rho / Rac GTPases, PAKs contribute to many of the GTPase specific roles.^4-11

Ustilago maydis is a basidiomycete fungus for which unicellular haploid cells reproduce by budding. Haploid cells of opposite mating backgrounds are able to fuse and form a diploid filamentous dikaryon. This dikaryon can subsequently infect maize (Zea mays). Cell fusion and pathogenic development are controlled by two separate loci, a and b. The a locus encodes a pheromone and pheromone receptor involved in cell recognition and cell fusion. Subsequent pathogenic development is dependent upon the b locus which encodes two homeodomain proteins, bE and bW.¹² Differentiation into the infectious dikaryon is dependent upon the activity of this b heterodimer. However, U. maydis cells can also undergo similar filamentous differentiation in response to environmental conditions. Lipids, acidic pH, and low ammonium conditions can trigger the filamentous response in a b locus-independent manner.^13-15

The signaling components responsible for such filamentous responses have been the subject of a number of investigations. In U. maydis, three small GTPases have been well-characterized: Rac1 establishes polar growth, while Cdc42 regulates cell separation, and Rho1 is required for cell polarity and cytokinesis.^2,16 Cla4, a Ste20-like PAK kinase, has been identified as a downstream effector of Rac1, while downstream effectors of Cdc42 and Rho1 have yet to be identified, although Cla4 is a potential candidate for these as well.¹⁶ Rac1 plays a role in polar growth and normal bud formation by its localization to the polar tip; Rac1 activity promotes localized cell wall formation in a Cla4 dependent manner.2The Δcla4 mutant strains display a dramatic disruption in chitin deposition; however the Δrac1 mutant strain did not display any defects in chitin deposition. Though both Cla4 and Rac1are involved in polar growth, the role of Cla4 in chitin deposition is independent of Rac1. Ultimately, Rac1 and Cdc42 have separate roles in U. maydis: Rac1, through Cla4, regulates septin apparatus assembly necessary for bud formation, and then promotes polarized cell growth while Cdc42 is required to promote mother daughter cell separation, but is dispensable for bud formation and polar growth.² Thus it appears that Rac1, through Cla4, acts as the master regulator of filament formation.^2,16

However, very little work has been performed on Rac1, Cdc42, Rho1, and Cla4 under conditions of low ammonium. In contrast the Ste20 homolog, Smu1, is known to have a role in the filamentous response to low ammonium conditions.⁹ Smu1 appears to regulate filament formation under low ammonium conditions in a positive fashion. In this study, we provide evidence that Cla4 also plays a role in the filamentous response to low ammonium conditions. In addition, Rac1 may not be the only upstream activator of Cla4. While neither Cla4 nor Rac1 are absolutely required for filament formation as a response to low ammonium, this ability is greatly reduced in rac1 deletion strains.

When grown in rich conditions, the Δcla4 mutant cells are morphologically distinct from the cells disrupted for the U. maydis PAK-like Ste20 homolog, Smu1. As seen previously,⁸ Δcla4 cells were septated, containing several independent nuclei having failed to correctly separate (Fig. 1), unlike the Δsmu1 mutant strains which are affected only in cell length¹⁷ (Table 1). Moreover, the Δcla4 cells appear fatter than the wild type strains.⁸ In the current study, examination of the Δcla4 mutant strains identified a significant increase in cell length (Table 1). This increase in cell length stands in sharp contrast to the decrease in cell lengths observed in Δsmu1 mutant strains.¹⁷ These results are consistent with the observed roles of Cla4 in cytokinesis and polar growth in S. cerevisiae and C. albicans.^10,18 In addition, the phenotypes observed in the Δcla4 mutants were similar to those exhibited in Cdk1, Cdk5, and Wee1 mutants, indicating that Cla4 may play a role in cell cycle similar to Cla4 from S. cerevisiae.^19-21 The increased cell length of the Δcla4 mutant strains (Table 1) could indicate a delay in the G2-M transition in these U. maydis cells. Another possibility is that deletion of cla4 indirectly promotes cell elongation by failing to promote cell separation (via binding to Rac1), thus creating a cell cycle delay. Other pathways could then promote increased polar growth leading to a filamentous response. On low ammonium agar (SLAD), both the Δcla4 mutant and overexpression cla4^otef mutants were filamentous, albeit less than the wild type (Fig. 2). Strikingly, in contrast to wild type, in liquid SLAD the Δcla4 mutant produced elongated filamentous cells (Fig. 3).

Figure 1.

Cla4 and Rac1 have roles in cytokinesis and cell length. Δcla4 cells display an increase in cell length, while for cla4^Otef cells increase in cell length was not statistically significant (Table 1). Δcla4 cells also exhibited cytokinesis defects. Δrac1 cells displayed cell separation defects. rac1^Otef cells exhibit a dramatic increase in cell length. Strains were grown in YEPS liquid medium. Scale bars, 10 µm.

Table 1. Measures of cell length across all strains^A.

Strain	n	LengthB (in μm)	Comparison	pc
WT	213	19.14 +/− 0.31
Δcla4	111	24.05 +/− 0.83	WT v. Δcla4	> 0.001
Δsmu1	281	17.64 +/− 0.18	WT v. Δsmu1	> 0.05
cla4Otef	505	19.69 +/− 0.20	WT v. cla4Otef	N.S.
smu1Otef	224	19.66 +/− 0.30	WT v. smu1Otef	N.S.
AAll mutants and wild type (WT) compared here were in the same genetic background, that of strain FB224.B Cell length values are averages, +/- S.E. CStatistical analysis was performed using a one way ANOVA with a Dunnett’s Multiple Comparison Test. N.S., Not Significant.

Figure 2.

Colony morphology of Cla4 and Rac1 mutant strains grown on SLAD for 3 and 6 d post inoculation (dpi). (A) Examination of Cla4 mutants at 3 dpi indicates that the Δcla4 and cla4^Otef mutant strains were reduced in filament formation compared with wild type strains. The rac1^Otef strain was increased in filamentation on SLAD at 3 dpi . The Δrac1 strain exhibited bent, shorter, thickened cells that ended with a rounded structure at the apical tip (arrow, inset). A few cells were beginning to form filaments (see inset; scale bar, 40 µm). (B) At 6 dpi, the Δcla4 and cla4^Otef mutant strains were again reduced in filament formation compared with the wild type strain. The Δrac1 strain began to form filaments, but they were severely reduced compared with the wild type strain. The rac1^Otef strain was still increased in filamentation on SLAD at 6 dpi.

Figure 3.

Δcla4 strains produce filaments in liquid SLAD and display defects in cell wall localization. Δcla4 cells produce filaments in liquid SLAD, unlike WT or other mutant cells. The Δcla4 cells were highly branched displaying very few septa, with branch points occurring at cross wall septa (See arrows). The Δcla4 strains display massive chitin delocalization (WGA) but no defects in β-glucan localization (CFW). Neither cla4^Otef , nor either rac1 mutant type produced filaments; none of these latter mutants displayed defects in cell wall localization. Staining of U. maydis cells was obtained by treating 5 µL of cells with 1 µL of 10 µg / mL of Calcofluor white (CFW, 2 µg / mL final concentration; Fluorescent Brightener 28, Sigma, St. Louis, MO, specific for β-glucan and chitin of cell wall) or 100 µg / mL wheat germ agglutinin (WGA, 17 µg / mL final concentration; Tetramethylrhodamine conjugate, [Invitrogen], specific for chitin of cell wall). Nucleic acid staining used 10 µM Syto 11 (5 nM green fluorescent nucleic acid stain). Scale bars, 10 µm.

The role Rac1 plays in the filamentous response to low ammonium conditions is more complicated. Deletion of rac1 dramatically delayed filament formation, but did not eliminate the ability to form filaments completely (Fig. 2). Conversely, overexpression of rac1 increased the filamentous response; however the rac1^Otef mutant strain did not exhibit filament formation when grown in liquid SLAD, though the mutant cells were elongated. The exaggeration of the filamentous response in the Δcla4 mutant cells, along with the lack of such a response in either rac1 mutant strain suggest that the role Cla4 plays in the filamentous response to low ammonium conditions may be independent of Rac1. These results imply that Cla4 may not be the only downstream effector of Rac1, or conversely, Cla4 may have other upstream activators regulating septin assembly and proper cell separation. Potential candidates include Cdc42 and Rho1.

In sum, the PAK homolog, Cla4, differs from Ste20 homologs in one major regard: Cla4 homologs contain a pleckstrin homology (PH) domain, while Ste20 homologs do not. This domain allows Cla4 to bind to the membrane and localize to polar growth sites as well as regulate cell morphogenesis and mitotic exit.²² In U. maydis, Cla4 has been identified as the downstream effector of Rac1, regulating bud formation and morphology and polar growth.^2,8 Moreover, deletion of either cla4 or rac1 eliminates b-dependent filamentation. However, on SLAD media, Δcla4 strains were still able to produce filaments, being the only other mutant strain to filament in liquid SLAD besides the Δ_810–2832hsl7 smu1^Otef strains.¹⁷ Interestingly this filamentous response to SLAD appears to be independent of Rac1 activity. Deletion of rac1 created a dramatic delay in the filament formation, while overexpression of rac1 leads to hyphal growth. Yet, neither mutant strain formed filaments in liquid SLAD. It could be that Cla4 limits the filamentous response to low ammonium conditions by either sequestering a component in the filamentous response pathway (potentially Cdc42 or some PAK/MAPK activator) or by promoting bud emergence and normal cell cycle progression preventing cell elongation and hyphal growth. S. cerevisiae Cla4 binds Cdc42 promoting septin ring formation and proper bud emergence. Subsequently Cla4 is involved in the hyper-phosphorylation of Swe1 promoting G₂-M transition.^19,23 It is striking that the two U. maydis PAKs, Cla4 and Smu1, have opposite roles in the filamentous response pathway, where deletion of cla4 alone or overexpressing smu1 in a Δ_810–2832hsl7 background, leads to a hyper-filamentous response to low ammonium.

The role of Hsl7 in morphology and pathogenicity and its interaction with other signaling components in the plant pathogen, Ustilago maydis. Lovely C Ben, Aulakh Kavita Burman, Perlin Michael H. Eukaryot Cell. 2011;11:869–83. doi: 10.1128/EC.00237-10.