A Myosin Light Chain Is Critical for Fungal Growth Robustness in Candida albicans

ABSTRACT In a number of elongated cells, such as fungal hyphae, a vesicle cluster is observed at the growing tip. This cluster, called a Spitzenkörper, has been suggested to act as a vesicle supply center, yet analysis of its function is challenging, as a majority of components identified thus far are essential for growth. Here, we probe the function of the Spitzenkörper in the human fungal pathogen Candida albicans, using genetics and synthetic physical interactions (SPI). We show that the C. albicans Spitzenkörper is comprised principally of secretory vesicles. Mutant strains lacking the Spitzenkörper component myosin light chain 1 (Mlc1) or having a SPI between Mlc1 and either another Spitzenkörper component, the Rab GTPase Sec4, or prenylated green fluorescent protein (GFP), are viable and still exhibit a Spitzenkörper during filamentous growth. Strikingly, all of these mutants formed filaments with increased diameters and extension rates, indicating that Mlc1 negatively regulates myosin V, Myo2, activity. The results of our quantitative studies reveal a strong correlation between filament diameter and extension rate, which is consistent with the vesicle supply center model for fungal tip growth. Together, our results indicate that the Spitzenkörper protein Mlc1 is important for growth robustness and reveal a critical link between filament morphology and extension rate.

where N is the number of vesicles released from the VSC per unit time, V is the rate of linear displacement of the VSC, and N V is the distance between the apical wall and the VSC. Support for this model predicting hyphal shape came with early studies of the plant fungal pathogen Rhizoctonia solani, in which morphology was followed upon disturbing growth. Specifically, dislocation of the Spitzenkörper, i.e., an increase in N V led to an increase in tip diameter (4). Furthermore, the hyphal tips of 32 fungal species were shown to approximate the hyphoid shape as described in equation 1 (7). Indeed, the position of the Spitzenkörper appears to anticipate a change in growth direction in Aspergillus niger and R. solani (2,7). When grown on surfaces, Candida albicans hyphal tips, as well as the Spitzenkörper, are asymmetrically positioned toward the substrate, and the position of this structure, which moves upon filament contact, was not an absolute predictor of growth direction (8). The studies of A. niger and R. solani have also analyzed the consequences of altering the position of the Spitzenkörper N V À Á with respect to filament morphology (2,7). From the hyphoid equation (equation 1), the maximum diameter of the hyphae was derived as follows: where D is the maximum hyphal diameter. This equation predicts that VSC links the diameter of the hyphae to its extension rate. As N V is the position of the vesicle supply center relative to the apical wall, this indicates that the hyphal diameter is proportional to the position of the vesicle supply center, i.e., the closer to the apical wall, the narrower the hyphal diameter and vice versa. Similarly, as the number of vesicles released from the vesicle supply center per unit time increases, i.e., increased extension rate, the hyphal diameter should increase, assuming the position of the vesicle supply center relative to the wall N V À Á does not change substantially. In filamentous fungi, a number of proteins have been localized to the Spitzenkörper, including Rab GTPases (Rab11 and Rab8 homologs, as well as Rab6 and Rab1 in Aspergillus nidulans) (9)(10)(11)(12)(13)(14)(15)(16), Rab guanine nucleotide exchange factors (GEFs) (e.g., Sec2) (17), the polarisome protein Spa2, the nuclear dbf2-related (NDR) kinase COT-1 (18), lipid flippases (19,20), the myosin V motor (21,22) and myosin light chain (9,23), the glycolysis enzyme GPI-1 (24), the coiled-coiled protein SPZ-1 (24), scaffold proteins (Leashin-2 and Janus-1) (24), chitin and glucan synthases (14,(25)(26)(27)(28)(29)(30)(31) and the formin Bni1 (23,32,33). In C. albicans, the Spitzenkörper has been visualized by fluorescence microscopy (10,15,17,23,33), and electron microscopy revealed that it is comprised of a homogeneous vesicle population of approximately 60 vesicles (16), in contrast to Neurospora crassa, where a layered structure of micro-and macrovesicles is observed (31). N. crassa Spitzenkörper mutants with altered composition of this structure (Dspz-1, Djns-1, Dspa-2, and Dmyo-5) exhibit reduced extension rates (24). In C. albicans, perturbation of actin cables (10,16,23) or secretion in a sec3 mutant disrupted the Spitzenkörper (16,34). Mutation of the C. albicans Ras-like GTPase Rsr1 also resulted in the position of the Spitzenkörper to meander during growth (8). Finally, two polarisome components, Spa2 and Bud6, have been shown to be required for the integrity of the C. albicans Spitzenkörper, and deletion of either component resulted in wider and less polarized filaments (23).
Here, we used mutants of a myosin light chain, Mlc1, which localizes to the Spitzenkörper, to probe the function of this structure using a quantitative analysis of C. albicans hyphal growth and morphology. We first demonstrated that, in the absence of Mlc1, filamentous growth still occurs and a Spitzenkörper is present, yet mlc1 deletion mutants exhibited wider filaments with increased extension rates. In addition to this deletion mutant, we generated two strains with constitutive synthetic physical interactions (SPI) (35) between Mlc1 and either the Rab8 GTPase Sec4 or prenylated green fluorescent protein (GFP). Strikingly, with these mutants, similar to the mlc1 deletion mutant, we observed a strong correlation between filament diameter and extension rate, as predicted by the vesicle supply center model. Together, our results reveal that the Spitzenkörper protein Mlc1 is essential for growth robustness.

RESULTS
First, we set out to examine the composition of the C. albicans Spitzenkörper, specifically if it was predominantly composed of secretory vesicles. Using a functional Scarlet-Sec4 fusion, present as the sole copy, we quantitated the signal in the smallest structures (ranging in size between 0.0068 and 0.027 mm 3 ) and compared it to the signal associated with the Spitzenkörper (Fig. 1A and B). We used the red fluorescent protein mScarlet, as there was less background signal in the cells compared to using GFP, which was critical for identifying individual secretory vesicles. The mean signal per secretory vesicle (3,059 6 1,630) was then used to determine the number of secretory vesicles in the Spitzenkörper, and we observed a good correlation between this number of secretory vesicles and the volume of the Spitzenkörper (Fig. 1C). The average number of secretory vesicles in the Spitzenkörper of 110 6 46 (Fig. 1D) is in good agreement with the number of total vesicles determined by electron microscopy of 60 to 75 (16). Hence, these results are consistent with the notion that the Spitzenkörper is comprised essentially of secretory vesicles. While we did observe some variation in the absolute number of secretory vesicles in the Spitzenkörper, the intensity of the Sec4 signal at this structure was constant over time, suggesting that the rate of vesicles arriving equals the rate of vesicles departing from it.
Investigating the function of the Spitzenkörper has been hampered as central components are required for viability. For example, in C. albicans, the Spitzenkörper components Sec4 (10) and its GEF Sec2 (17), as well as the Rab11 homolog Ypt31 (15) (see Fig. S1A in the supplemental material), are essential. One Spitzenkörper component that is not essential for viability in C. albicans is the formin Bni1, which is partially redundant with Bnr1 (33,36). We examined filamentous growth in a bni1 deletion mutant, and Fig. S2 shows that bni1 filaments were wider, yet their extension rate was substantially reduced. However, we did not observe a cluster of secretory vesicles in this mutant, using Scarlet-Sec4 (Fig. S2D), consistent with data on other fungi (24) that in the absence of a Spitzenkörper, filamentous growth is reduced. In N. crassa, mutants lacking the polarisome component SPA-2, the scaffolding proteins Janus-1 (JNS-1) and Leashin-2 (LAH-2), as well as the coiled-coil protein SPZ-1 have a reduced growth rate, with mutants lacking the myosin V motor protein exhibiting very little growth (24). Similarly, in A. nidulans, myosin V mutants are wider and extend slower (12,22,37). To probe the function of the C. albicans Spitzenkörper, we generated mutants of the myosin light chain 1 (Mlc1). Homozygous mlc1 deletion mutants were viable, with a doubling time comparable to that of the wild-type strain during budding growth, albeit with chains of cells that suggest a cytokinesis defect ( Fig. 2A). Of note, it was shown that the myosin V homolog, Myo2, is also not essential for viability in C. albicans (38). The mlc1 cells were slightly elongated and somewhat larger than control cells ( Fig. 2B and C). Colonies of the mlc1 mutant grew faster than the wild-type and complemented strains on rich medium, yet this mutant was unable to form invasive filaments in serum-containing agar medium ( Fig. 2D and E). The mlc1 mutant was nonetheless able to form filaments that resembled hyphae in liquid serum medium (Fig. 3), with lengths of the filament compartments identical to those of the control cells ( Fig. 4A and B) and a tip-localized cluster of Sec4 also similar to control cells ( Fig. 3 and Fig. 4A and C). However, we observed on average two nuclei per filament compartment in the mlc1 mutant filaments, compared to a single nucleus in the control cells ( Fig. 4C and D). To further probe Mlc1 function at the Spitzenkörper, we generated two mutants, each with a specific synthetic physical interaction (SPI). We sought to determine whether linking Mlc1 to prenylated GFP, which localizes to the plasma membrane, would alter the Spitzenkörper and perturb filament growth. We predicted that broadening the distribution of the Spitzenkörper should result in wider filaments that, according to the VSC model, should also extend faster. In Saccharomyces cerevisiae, Mlc1 has been shown to form a complex with Myo2 and/or Sec4 on secretory vesicles (39), and Myo2 has been shown to bind Sec4 (40). Given that Mlc1, Myo2, and Sec4 are likely to function together also in C. albicans, we examined whether stabilizing this interaction might increase the number of vesicles targeted to the site of fusion at the plasma membrane, thereby accelerating filament extension. To generate a Mlc1ÁGFP prenylated mutant, one copy of Mlc1 was fused to a GFP nanobody and expressed together with prenylated GFP (Fig. 5A). In contrast, to generate a Mlc1ÁSec4-stabilized mutant, the interaction between Mlc1 and Sec4 was stabilized, using a GFP nanobody fused to one copy of Mlc1 in cells expressing GFP-Sec4 (Fig. 5A). Both of these SPI mutants were viable and formed filaments that appeared larger than the control cells, in serum inducing medium (Fig. 5B). In both mutants, a cluster of Sec4 was observed at the filament tip (Fig. 5B). Furthermore, Ypt31 also localized to a cluster at the tip of mlc1 mutant filaments (Fig. S1B), which confirms that a Spitzenkörper still forms when Mlc1 is altered.
We next examined the cluster of Sec4 present at the filament tips of all three mlc1 mutants. Both the mlc1 deletion mutant and the Mlc1ÁSec4-stabilized mutant had increased Sec4 signal compared to control strains ( Fig. 6A and B), suggesting an increased number of secretory vesicles. In contrast, there was a reduction in Sec4 signal in the Mlc1ÁGFP prenylated mutant (Fig. 6B), and the length of the vesicle cluster long axis, visualized by Sec4 and Mlc1, increased as a function of filament diameter (Fig. 7A), consistent with a more spread-out distribution of secretory vesicles. We next examined the dynamics of Sec4 at the Spitzenkörper, using fluorescence recovery after photobleaching (FRAP). We observed a striking increase in the Sec4 immobile fraction in the Mlc1ÁSec4-stabilized mutant (Fig. 7B), but not in the mlc1 deletion mutant or the Mlc1ÁGFP prenylated mutant compared to their respective controls FIG 2 In the absence of Mlc1, cells are enlarged and grow faster during budding growth yet are defective for invasive growth. (A to C) Deletion of results in larger budding cells. DIC images of mlc1D/ MLC1 (Control) and mlc1D/mlc1D (mlc1) cells expressing Scarlet-Sec4 and either Nop1-GFP (PY5716) or Cdc10-GFP (PY5720), respectively, are shown. Bar, 5 mm. (B and C) The aspect ratios (long axis over short axis) and area (assuming a uniform ellipse) were determined (n = 250 to 500 cells) with **** indicating P value of ,0.0001. (D) Colonies of the mlc1 mutant grow faster than the wild type. The indicated strains, including wild-type (WT; PY4860), mlc1 (mlc1D/mlc1D; PY4754), and mlc1 1 MLC1 (mlc1D/mlc1D 1 MLC1; PY5658) were grown on rich medium at 30°C, and colony diameter (n = 25 to 50) was determined after 3 to 7 days incubation. Symbols are mean values of the normalized colony size at 3 days, with error bars indicating standard deviations; the lines are best fits with r 2 . 0.94. (E) Mlc1 is required for invasive growth. Indicated strains, expressing Scarlet-Sec4 (WT, PY4860; mlc1, PY5451; mlc1 1 MLC1, PY5661) were spotted on rich medium containing serum and incubated at 37°C for 7 days.
( Fig. 3A and B). We confirmed that the increase in Sec4 immobile fraction observed in the Mlc1ÁSec4-stabilized mutant was not due to a difference in the fraction of the total Sec4 bleached between this strain and a control strain (Fig. 3C). This Sec4 immobile fraction in the Mlc1ÁSec4-stabilized mutant is similar to that previously observed for Mlc1 in wild-type hyphal filaments (10). We did not observe a significant difference in the Sec4 FRAP half-life (t 1/2 ) values in the different mutants, although it was slightly lower in the Mlc1ÁGFP prenylated mutant ( Fig. 3A and B). As is the case with mammalian Rab1 (41), GDP dissociation inhibitor (GDI)-mediated recycling is likely to account for the recovery of Sec4 signal after photobleaching, which is substantially faster (10) than the turnover of vesicles at the Spitzenkörper (16). Together, these results indicate that a synthetic interaction between Sec4 and Mlc1 stabilizes the association between these two proteins and suggest that targeting Mlc1 to the plasma membrane results in a more spread-out Spitzenkörper.
Our initial examination of the three mlc1 mutants ( Fig. 3 and 5) showed that the filaments had morphological defects, and hence, we measured their diameters and extension rates from time-lapse microscopy. In all mutants, the filament diameters ( Fig. S4A to D) and extension rates were constant over time. The filaments of all three mutants exhibited increased diameters, with mean diameters significantly increased compared to that of the respective control strains (Fig. 8A and B); note that the coefficient of variation was 2 to 3 times higher compared to the control strains. Similarly, both the Mlc1ÁGFP prenylated and Mlc1ÁSec4-stabilized mutants exhibited a significant increase in the average extension rate compared to the control strains ( Fig. 8A and B), and while such an increase was not apparent in the mlc1 mutant, the coefficient of variation was increased in all three mutants. These data, color coded according to a purple-to-yellow gradient representing an increase in filament diameter, indicate that the mlc1 mutants with larger diameters extend the fastest. This is further illustrated by the strong correlation between the filament diameter and extension rate ( Fig. 8C and D), consistent with the vesicle supply center model (5), specifically D ¼ 2p N V (equation 2) predicts that when N V increases, so does the hyphal diameter. In the three mlc1 mutants, there was an average of two-to fivefold increase in the rate of cell volume change (Fig. S4E and F). These results indicate that Mlc1 is critical for growth robustness. Our results are also consistent with an increased region of vesicle fusion with the plasma membrane in the mlc1 mutant. Indeed, we examined the distribution of the key exocyst component Sec3 (34), and Fig. 9A and B show that Sec3 was found at the filament tip in the mlc1 mutant, similar to the control strain, but with a wider distribution in wider cells (average diameter of These results suggest that the Spitzenkörper regulates filament growth rate and morphology and that the filament extension rate is tightly linked to diameter. To investigate further this relationship, we examined C. albicans of different ploidy, as cell size has been shown to increase with increased ploidy (42). We compared two isogenic strains, and Fig. 10 shows a striking correlation between filament diameter and extension rate, both of which increase upon increased ploidy. These results further confirm that filament diameter and extension rate are linked and support the vesicle supply center model for hyphal growth.

DISCUSSION
Our results indicate that the Spitzenkörper plays a key role in regulating hyphal growth and morphology by linking these two processes. In particular, when Mlc1 interacts with prenylated GFP, we observed increased filament extension rates and diameters. In this Mlc1ÁGFP prenylated mutant, the vesicle cluster is more spread out, and exocytosis is likely to be targeted to a larger area. There is precedence for such a phenotype in C. albicans, as germ tubes of the Ras-like GTPase Rsr1 deletion mutant grow faster, are somewhat wider than wildtype cells, and have an increased Mlc1 signal at the Spitzenkörper (43,44). Intriguingly, in the Mlc1ÁSec4-stabilized strain, we also observed increased filament extension rates and diameters, and we speculate that stabilizing the Sec4-Mlc1 interaction increases the targeting of vesicles to the plasma membrane. In contrast to the two SPI mutants, the average extension rate of the mlc1 deletion mutant is not significantly increased compared to control strains, yet we observe an increased variation in the extension rate among cells, and filaments that extend faster are also wider. We speculate that this could be due to the removal of Mlc1 negative regulation on the myosin V Myo2.
Mutations in components of the Spitzenkörper alter filamentous growth. For example, deletion of C. albicans BNI1 (33,36), N. crassa and A. nidulans myosin V (21,22) and A. nidulans Rab8 (12) all resulted in reduced filament extension rates. In contrast, the C. albicans mlc1 deletion mutant extends faster than wild-type controls, consistent with this myosin light chain being a negative regulator of myosin V activity. As a Spitzenkörper was not observed in the C. albicans bni1 (this study) and N. crassa Dmyo-5 deletion mutants (21,24), this suggests that this structure is critical for or associated with efficient hyphal growth.
Studies of a number of fungi indicated that as a particular fungal species grows faster, its diameter also increases. This has been observed in Ashbya gossypii (45), where filament extension rates increase over time, concomitant with an increase in filament diameter and a change in Spitzenkörper shape. Similarly, an increase in C. albicans extension rates was observed between undifferentiated and differentiated mycelia concomitant with an increase in hyphal diameter (46). In addition, an increase in N. crassa hyphal extension rate was observed going from secondary to primary branches up to leading hyphae, and this correlated with an increase in diameter (47). Most strikingly, López-Franco et al. compared a range of fungi and oomyces, including N. crassa, Gilbertella persicaria, Pythium aphanidermatum, Trichoderma viride, Saprolegnia ferax, Fusarium culmorum, and R. solani, in which an increase in hyphal diameter (from 6.7 to 11.5 mm) appears to correlate with hyphal extension rate (from 0.1 to 0.7 mm/s) (48). This relationship between hyphal diameter and extension rate has been predicted by the vesicle supply center model (equation 2) (5), and our results establish that this relationship holds upon specific perturbation of the Spitzenkörper. Interestingly, it was recently shown that variation in hyphal width and extension rate increases in faster and wider growing species (49), suggesting that the regulation of growth may be less precise when growth speeds increase. Together, our results reveal that the Spitzenkörper component Mlc1 links hyphal growth and morphology, suggesting that this structure is critical for minimizing growth and morphology variation in a fungal population, which is likely to be important for mycelium development. Diameters are the averages of values measured every 5 to 10 min over a 120-min time-lapse experiment, and extension rates are from linear fits of filament length over at least 60 min (r 2 . 0.9). Diameters were normalized to the control strain mlc1D/MLC1, which had a mean value of 2.3 6 0.1 mm. Extension rates were normalized to the control strain mlc1D/MLC1 mean value of 0.37 6 0.06 mm/min. Values were sorted by filament diameter and color coded with a color gradient from purple (smallest diameter) to yellow (largest diameter) for each strain (Lookup Table [

MATERIALS AND METHODS
Strains, media, and genetic methods. Standard methods were used for C. albicans cell culture and molecular and genetic manipulations as described previously (50). Derivatives of the BWP17 strain (51) were used in this study and are listed in Table S1 in the supplemental material. Strains were grown in rich medium (yeast extract-peptone-dextrose [YEPD]) at 30°C for all experiments, and induction of filamentous growth was carried out with fetal calf serum (FCS) at 37°C. For doxycycline gene repression, cells were grown in the presence of 20 mg/ml doxycycline (52). Oligonucleotides and synthesized DNA used in this study are listed in Tables S2 and S3. The genes encoding GFP nanobody (GNB) (53) and monomeric far-red fluorescent protein CamiRFP670 (54) were codon optimized for C. albicans and commercially synthesized (BaseClear, Netherlands). The C. albicans GNB (CaGNB) gene was cloned into pFA-GFPg-URA3 (55) using PstI and AscI sites, resulting in pFA-CaGNB-URA3. The CamiRFP670 gene was PCR amplified and cloned into pFA-CaGNB-URA3 with unique 59 and 39 PstI sites (oligonucleotides CamiRFP670PstIp and CamiRFP670m-GSlink-PstI_noStop), which also introduced a GSGSGS linker between the CamiRFP670 gene and the CaGNB gene, resulting in pFA-CamiRFP670-CaGNB-URA3. The synthesized CamiRFP670 gene was PCR amplified with a unique 59 PstI site and GAGAGA linker sequence, and a unique  Single particles and clusters of mScarlet-Sec4 signals were identified using a signal intensity of 5.5 standard deviations above the mean. An average intensity for single particles of Sec4 was calculated as the average of 2 to 8 voxel objects (0.0068 to 0.027 mm 3 ). The intensity of the Spitzenkörper cluster was divided by the average single vesicle intensity to estimate the number of vesicles in the identified Spitzenkörper cluster. For the analyses of Sec4 Spitzenkörper intensity and long axis, images were acquired as described above, and Volocity software was used to quantitate z-stack sum projections. Spitzenkörpers were identified by different numbers of standard deviations above the mean signal intensity, 8 for mScarlet-Sec4 and 25 for GFP-Sec4.
Budding cell doubling time was determined from time-lapse microscopy, from bud emergence to bud emergence. The aspect ratio and area of budding cells were determined by fitting an ellipse region of interest (ROI) to cell maximum projection (0.4-mm z-steps), using FIJI to extract area and aspect ratio. Filamentous cell diameter and extension rates were determined by measuring the cell length and diameter at each time point (every 5 to 10 min) during a 120-min time lapse. The extension rate was determined as previously described (62). Briefly, length was plotted over time, and the slope of a best-fit line was used as the extension rate. Cell diameter was also measured at each time point at the center of the hyphal filament, and the numbers reported are the average over the time course. Volume was calculated at each time point using the cylindrical volume formula: where r is the radius and h is the cell length. Change in volume was determined by plotting volume over time and using the slope of a best-fit line.
Fluorescence recovery after photobleaching (FRAP) analysis was performed as described previously (15). Images were captured every 0.63 s at 0.7% or 0.2% maximum laser intensity for 488-nm or 561-nm laser lines, respectively. Photobleaching scans on a circular area of 1 to 2 mm 2 were carried out with five Myosin Light Chain Critical for Growth in C. albicans ® September/October 2021 Volume 12 Issue 5 e02528-21 mbio.asm.org 13 consecutive pulses at 80% or 50% laser intensity for 488-nm or 561-nm laser lines, respectively, using a 1.63-ms pixel dwell time. The average signal intensity of the bleach ROI was normalized to photobleaching during image acquisition, which was fit to a one-phase decay regression: Y = (Y 0 2 plateau)(e 2kx ) 1 plateau, of the average intensity elsewhere in the cell (using GraphPad Prism 8 software). Regression analysis to determine the FRAP t 1/2 was done using a one-phase exponential association function in GraphPad Prism 8 software as follows: Y = Y max (1 2 e 2kx ), where k is the rate constant and t 1/2 is 0.69/k. The immobile fraction was calculated using the equation, 1 -[(I final 2 I postbleach )/(I prebleach 2 I postbleach )], where I is the signal intensity.