Stress- and metabolic responses of Candida albicans require Tor1 kinase N-terminal HEAT repeats

Whether to commit limited cellular resources toward growth and proliferation, or toward survival and stress responses, is an essential determination made by Target of Rapamycin Complex 1 (TORC1) for a eukaryotic cell in response to favorable or adverse conditions. Loss of TORC1 function is lethal. The TORC1 inhibitor rapamycin that targets the highly conserved Tor kinase domain kills fungal pathogens like Candida albicans, but is also severely toxic to human cells. The least conserved region of fungal and human Tor kinases are the N-terminal HEAT domains. We examined the role of the 8 most N-terminal HEAT repeats of C. albicans Tor1. We compared nutritional- and stress responses of cells expressing a message for N-terminally truncated Tor1 from repressible tetO, with cells expressing a wild type TOR1 allele from tetO or from the native promoter. Some but not all stress responses were significantly impaired by loss of Tor1 N-terminal HEAT repeats, including those to oxidative-, cell wall-, and heat stress; in contrast, plasma membrane stress and antifungal agents that disrupt plasma membrane function were tolerated by cells lacking this Tor1 region. Translation was inappropriately upregulated during oxidative stress in cells lacking N-terminal Tor1 HEAT repeats despite simultaneously elevated Gcn2 activity, while activation of the oxidative stress response MAP kinase Hog1 was weak. Conversely, these cells were unable to take advantage of favorable nutritional conditions by accelerating their growth. While consuming oxygen more slowly than cells containing wild type TOR1 alleles during growth in glucose, cells lacking N-terminal Tor1 HEAT repeats were capable of utilizing non-fermentable as well as fermentable carbon sources and of growth during severe acid stress, but were uniquely unable to grow on lactate. Genome-wide expression analysis showed paradoxical simultaneous activation of anabolic- and starvation responses in cells lacking Tor1 N-terminal HEAT repeats, with misregulation of carbon metabolism and of translational machinery biosynthesis. Targeting fungal-specific Tor1 N-terminal HEAT repeats with small molecules might abrogate fungal viability, especially when during infection multiple stresses are imposed by the host immune system.


58
The Target of Rapamycin Complex 1 (TORC1) makes fundamental decisions in the life of a eukaryotic cell. 59 It collects information from numerous sources on conditions that affect the cell's chances of successful 60 since regulating translation initiation does not involve a 4EBP1 homolog in these fungi and the orthologous 152 role of Sch9 as an S6K1 homolog is still debated [27,28]. Therefore, despite higher-level structural 153 similarities between fungal and mammalian TORC1, Tor kinase activity is regulated differently in fungi 154 than in mammals and the roles of its distinct domains and their physical interaction partners are not 155 understood in detail. 156

157
The N-termini of fungal and human Tor kinase are their most widely divergent domains, suggesting they 158 may be sufficiently structurally and functionally distinct that differential chemical targeting could be 159 achievable. The large size and complexity of the N-and M-HEAT domains [18] indicates commensurate 160 complexity of interactions. A functional dissection of their individual roles may aid in further characterizing 161 afferent as well as efferent signal fluxes of Tor1. Responses to nitrogen source quality and quantity and 162 control of protein synthesis are central to TORC1 signaling [29][30][31], as recently reviewed e.g. in [32]. In S. 163 cerevisiae and C. albicans, TORC1 responds to carbon source quality and quantity [31] and it controls 164 carbon source-acquisition and -metabolism genes of S. cerevisiae [30]. Phosphate repletion is an afferent 165 signal to TORC1 in S. cerevisiae and C. albicans and TORC1 activity also affects the C. albicans phosphate 166 acquisition system, the PHO regulon [33]. 167 168 As mentioned above, C. albicans must adapt to rapidly shifting nutritional states as a commensal in the 169 human gastrointestinal tract, and to stresses imposed not only by competing flora but also by the human 170 immune system during commensalism, mucosal-and invasive infection. To maximize fitness in these 171 quickly changing environments, TORC1 of C. albicans must be finely tunable to an array of environmental 172 parameters in widely varying combinations. The challenges of adapting to varying host environments and 173 the importance of Target of Rapamycin complexes in meeting these challenges are exemplified by the 174 sleeping sickness parasite Trypanosoma brucei, whose complex life cycle comprising four cell types 175 alternates between its fly vector and its human host [34]: T. brucei has four TOR complexes [35,36]. C. 176 albicans TORC1 is involved in hyphal growth and adhesion [37][38][39][40], biofilm formation [26], secretion of 177 aspartyl protease [24], and it responds to nitrogen-, carbon-source [25,41] and phosphate [33] 178 availability. TORC1 is predicted to be required for C. albicans fitness in favorable conditions like repletion 179 of distinct nutrients, as well as in manifold stress conditions encountered in the host. 180 Given the wide array of stimuli to which TORC1 responds and the broad spectrum of responses it controls, 182 we set out to examine the processes regulated by the most N-terminal segment of the N-terminal HEAT 183 domain in C. albicans. We constructed two mutant genotypes in which the single remaining TOR1 alleles 184 were placed under the control of a repressible promoter: wild type full-length TOR1 or a 5'-truncated 185 TOR1 allele whose predicted protein product lacks the 8 most N-terminal HEAT repeats. Cells in which 186 these TOR1 alleles were overexpressed and partially or fully repressed were examined for their responses 187 to nutritional repletion or starvation, and to stresses to which C. albicans is exposed in the host. Aligning C. albicans Tor1 and the K. marxianus Tor kinase, whose structure has been elucidated in detail 195 [18], with human mTOR, we noted that the most divergent regions of these orthologs are in their N-196 termini while their catalytic domains are highly conserved (Fig. 1). Tor kinase N termini consist of arrays 197 of HEAT repeats [15]. In order to test the role of C. albicans Tor1 and of its N-terminal HEAT repeats in the 198 response to distinct nutrients and stressors, we constructed strains in which transcription of a single TOR1 199 allele is controlled by repressible tetO. In the absence of the repressing compound doxycycline, 200 transcription from this construct is known to be high [42]. We mutated the second TOR1 allele in three 201 independent heterozygous deletion mutants in order to detect and avoid artifacts arising from suppressor 202 mutations; strains were constructed to express either full-length Tor1 (Tor1-FL) or an N-terminally 203 truncated Tor1 whose Start methionine is residue 382 of the full length protein (Tor1-Del381). We used 204 the descriptor TOR1-Del381 for this TOR1 allele as well, for the sake of simplicity, though the number 381 205 refers to the truncated amino acid residues and not to the deleted nucleotides. The mutant protein lacks 206 the 8 N-terminal HEAT repeats, truncating the N-terminal HEAT domain, a predicted interaction site with 207 the regulator Kog1. We chose at least 2 strains from each genotype lineage whose phenotypes were 208 indistinguishable for further analysis of growth-and stress phenotypes. We also used two isogenic wild 209 type strains whose deleted ARG4 locus was independently restored to one wild type allele (resulting in 210 wild type TOR1 gene, respectively, were expressed from tetO; their respective genotypes were tor1/tetO-213 TOR1-Del381 or tor1/tetO-TOR1. 214 215 Growth of cells expressing full-length or truncated TOR1 from tetO decreased in increasing concentrations 216 of doxycycline, as expected ( Fig. 2A). When TOR1 expression was fully repressed at high concentrations 217 of doxycycline, 1 or 2 µg/ml (Fig. S1A, see time point 0), growth of cells containing the truncated allele 218 was nearly abolished ( Fig. 2A). In contrast, apparent residual expression of full-length TOR1 was sufficient 219 to permit substantial growth during tetO repression ( Fig. 2A), possibly due to inevitable leakiness of an 220 inhibited tetO construct. qRT-PCR showed rapid overexpression of TOR1-Del381 alleles from tetO in the 221 absence of doxycycline, while TOR1-FL mRNA levels recovered more slowly after tetO repression, hinting 222 at a possible role of N-terminal HEAT repeats in the half-life of the TOR1 mRNA itself (Fig. S1A). We assayed 223 TORC1 activity during exponential growth in rich complex medium (YPD) by the phosphorylation state of 224 ribosomal protein S6 (P-S6) [25], using a minimal doxycycline concentration of 5 ng/ml as a ceiling for the 225 expression level. Cells overexpressing full-length TOR1 or TOR1-Del381 from tetO showed comparable P-226 S6 signal intensity to wild type cells in rich medium (Fig. 2B), with a slightly weaker P-S6 signal in Del381 227 cells under these conditions of optimal growth. Levels of total Rps6 protein did not change under any of 228 the experimental conditions we examined and hence are shown only twice ( Fig. 2B and D). We concluded 229 that overexpression of TOR1 from unrepressed tetO does not per se increase TORC1 signaling, possibly 230 because the other TORC1 components required by the stoichiometry of the complex are not available at 231 higher levels. 232 233 Preferred nitrogen sources activate S. cerevisiae TORC1 [43][44][45], and this effect is conserved in C. albicans 234 [25]. To test the role of C. albicans Tor1 and its N-terminal HEAT repeats in cells' responses to rich versus 235 non-preferred nitrogen sources, wild type, tor1/TOR1, Del381 and FL cells were grown in different 236 nitrogen sources. Without tetO-repression, or during repression in moderate concentrations of 300 ng/ml 237 doxycycline, FL cells grew as well as wild type or tor1/TOR1 heterozygous cells (Fig. 2C). In synthetic YNB 238 medium with preferred nitrogen sources known to induce TORC1 signaling [25], ammonium sulfate or 239 glutamine, Del381 strains grew more slowly than wild type or heterozygotes (Fig. 2C). In contrast, in the 240 non-preferred nitrogen source proline [25], Del381 cells had no detectable specific growth defect while 241 wild type and all TOR1 genotypes grew more slowly than in preferred nitrogen sources (Fig. 2C). Cell 242 dilutions of the same strains spotted on solid agar media with these different nitrogen sources, showed 243 analogous phenotypes during overexpression of tetO-TOR1-Del381 or tetO-TOR1-FL alleles in the absence 244 of doxycycline, while during partial repression of these alleles with a moderate doxycycline concentration, 245 growth of Del381 cells was sharply diminished on all 3 nitrogen sources (Fig. S1B). Tor1 N-terminal HEAT 246 repeats were therefore specifically required for cells' growth acceleration during their use of preferred 247 nitrogen sources. 248

249
To examine the role of Tor1 N-terminal HEAT repeats in TORC1 signaling during growth in distinct nitrogen 250 sources, we assayed TORC1 activity by the P-S6 signal [25]. Expression of TOR1 was repressed for 4 hours 251 by exposure to 2 µg/ml doxycycline and then released for 45 minutes in media without doxycycline 252 containing glutamine or proline, respectively. Cell lysates were probed with antibody to P-S6 and total 253 Rps6. Del381 cells showed significantly elevated P-S6 signals above those of wild type or FL cells in both 254 conditions: in proline, where wild type and FL cells' P-S6 signal was undetectable while that of Del381 cells 255 was strong, and in glutamine, where the P-S6 signal from Del381 cells was more intense than that of the 256 other two strains (Fig. 2D). Hence, downregulation of TORC1 signaling in response to a non-preferred 257 nitrogen source required the complete Tor1 N-terminal HEAT domain. 258 259 Eukaryotic initiation factor 2 mediates translational control in response to starvation and environmental 260 stresses in S. cerevisiae as reviewed e.g. in [46]. During starvation for preferred nitrogen sources and for 261 glucose, the eIF2 alpha subunit (eIF2α) is phosphorylated by the kinase Gcn2 [47] and translation of many 262 anabolic messenger RNAs is inhibited. Using an antibody against the conserved phospho-serine 51 of 263 human eIF2α, we examined whether inappropriately increased TORC1 signaling in Del381 cells might 264 correspond to inappropriately weak translation inhibition signaling through Gcn2, as assayed by eIF2α 265 phosphorylation. To the contrary, we found that eIF2α phosphorylation was actually increased in Del381 266 cells (Fig. 2D), reflecting increased inhibitory signaling by Gcn2 and suggesting that TORC1-and Gcn2-267 signaling can become uncoupled when TORC1 lacks a function provided by N-terminal HEAT repeats. 268 269 Tor1 N-terminal HEAT repeats were required for growth acceleration in phosphate-replete conditions. 270

271
During invasion of the host, C. albicans apparently experiences starvation for inorganic phosphate, since 272 expression of the high-affinity inorganic phosphate transporters Pho84 and Pho89 is induced in models of 273 invasive disease [48][49][50][51], and since loss of Pho84 attenuates virulence [52]. We examined the role of Tor1 274 and its N-terminal HEAT repeats in distinct conditions of phosphate availability. Cells containing TOR1 275 genotypes wild type, heterozygous null (tor1/TOR1), Del381 and FL alleles were grown in distinct 276 phosphate concentrations. In low ambient phosphate, cells overexpressing TOR1-Del381 in the absence 277 of doxycycline grew at rates comparable to those overexpressing TOR1-FL (Fig. 2E); they showed mildly 278 increasing growth defects with increases in phosphate concentrations (Fig. 2E). During moderate 279 repression of tetO, the growth defect of Del381 cells became more pronounced in increasing ambient 280 phosphate concentrations compared with cells containing full length TOR1 alleles (Fig. 2E). We concluded 281 that the N-terminal HEAT repeats were required for growth acceleration in favorable phosphate-as well 282 as nitrogen source conditions. 283

284
The response of TORC1 signaling in wild type, Del381 and FL cells to external phosphate repletion or -285 starvation was examined. P-S6 was not decreased in low-phosphate (0.22 mM KH2PO4) versus phosphate-286 replete (11 mM KH2PO4) conditions in Del381 cells (Fig. S1C). Overall growth, as assayed by an increase in 287 the optical density of the culture, was therefore delinked from the intensity of TORC1 signaling in Del381 288

cells. 289 290
Tor1 N-terminal HEAT repeats contributed to acceleration of growth during glucose repletion and were 291 required for growth on lactate and for physiologic oxygen consumption. 292 293 Carbon source-and phosphate repletion are components of C. albicans cells' nutrient status monitored 294 by TORC1 [25,33]. In the host interaction, Candida cells compete with host phagocytes for glucose [53-295 55]. During growth in low concentrations of glucose, cells overexpressing TOR1-Del381 from tetO grew at 296 similar rates as wild type or cells expressing full length TOR1. In high glucose concentrations, Tor1-Del381 297 expressing cells showed a slower growth rate than those containing an intact N-terminal HEAT domain 298 ( Fig. 3A) and this effect again became more pronounced during partial repression of tetO with a moderate 299 doxycycline concentration of 300 ng/ml. This finding suggested that functions residing in the N-terminus 300 of the N-terminal HEAT domain contributed to growth acceleration during repletion of the preferred 301 carbon source, glucose. 302 303 C. albicans does not encounter high glucose concentrations during invasive infection: human bloodstream 304 glucose is 0.1%. Candida's ability to use a range of carbon sources contributes to fitness in the host [53, 305 56-63]. The contribution of Tor1 and its N-terminal HEAT repeats to utilization of another fermentable 306 carbon source, maltose, and of the non-fermentable carbon sources glycerol, ethanol, lactate, acetate, 307 and oleic acid were examined. During full induction of tetO in the absence of doxycycline, Del381 cells had 308 a significant growth defect on rich complex medium with lactate as the carbon source (Fig. 3B). On all 309 other carbon sources their growth defects during tetO induction compared with FL cells were minor (Fig.  310 3B). During moderate repression of tetO, growth defects of Del381 cells were more severe than those 311 expressing full length Tor1 (Fig. 3B). Del381 cells' growth defect on lactate was not due to acid stress, 312 because they grew equally well on glucose-or glycerol-containing medium buffered to pH 2, as on 313 unbuffered rich medium (Fig. S2). Physiologic expression of full-length TOR1 was required for appropriate 314 carbon source use. N-terminal HEAT repeats were specifically required for growth on lactate. 315

316
Since utilization of fermentable carbon sources like glucose and maltose was not dramatically impaired in 317 cells overexpressing TOR1-Del381 from tetO (in the absence of doxycycline), compared with non-318 fermentable carbon sources like glycerol and ethanol (Fig. 3B), we asked whether these cells were able to 319 grow anaerobically, i.e. under conditions requiring fermentation. Del381 cells again had no dramatic 320 growth defect under anaerobic (hypoxic) conditions compared with wild type during overexpression of 321 TOR1-Del381 in the absence of doxycycline (Fig. 3C); only when TOR1-Del381 transcription was 322 moderately repressed were these cells unable to grow anaerobically (Fig. 3C). We concluded that 323 fermentation does not require a regulatory activity residing in the Tor1 N-terminal HEAT repeats. 324 325 TORC1 activation and translational regulation through eIF2a were examined in cells growing in 2% glucose 326 (control), 0.25% glucose, 2% glycerol and in the absence of a carbon source. The P-S6 signal, indicating 327 TORC1 activation, was slightly reduced in the lower glucose concentration (Fig. 3D). Del381 cells showed 328 aberrantly increased P-S6 intensity in 0.25% glucose, similarly to their behavior during nitrogen starvation. 329 In the absence of a direct carbon source (0 glucose) and in 2% glycerol, the P-S6 signal was undetectable 330 for all strains. Phosphorylation of eIF2a did not change during provision of different glucose 331 concentrations or glycerol (Fig. 3D), but the P-eIF2a signal again was stronger in Del381 cells under these 332 conditions, including when they had no detectable P-S6 signal. C. albicans Gcn2 signaling apparently did 333 not respond to carbon source provision under these experimental conditions; it was uncoupled from 334 TORC1 activation in cells lacking N-terminal HEAT repeats. 335 336 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 We asked whether lagging growth of Del381 cells in high glucose concentrations corresponded to 337 decreased respiration. We measured oxygen consumption of wild type, tor1/TOR1, FL and Del381 cells. 338 Oxygen consumption was significantly decreased in Del381 cells in which tetO-TOR1-Del381 was not 339 repressed, consistent with a slower respiratory metabolism (Fig. 3E). We concluded that N-terminal HEAT 340 repeats contributed to increasing oxidative phosphorylation when glucose was abundant. Del381 cells might enable better endurance of oxidative stress; alternatively, perturbation of Tor1 could 348 impair the switch from growth-promoting to stress-enduring processes and increase sensitivity to 349 oxidative stress. We examined the role of Tor1 and its N-terminal HEAT repeats in the fungus's endurance 350 of oxidative stress by exposing cells to the superoxide-generating compound plumbagin and the peroxide-351 generating agent H2O2. tor1/TOR1 heterozygous cells, spotted in 5-fold dilutions on YPD plates containing 352 plumbagin or H2O2, were able to tolerate these compounds as well as wild type, but cells expressing tetO-353 TOR1-FL were hypersensitive to H2O2 in the absence and presence of tetO repression with doxycycline 354 (Fig. 4B). Del381 cells were strikingly hypersensitive to both sources of ROS (Fig. 4B). These findings 355 confirmed that TORC1 contributes to managing oxidative stress in C. albicans, and suggested that the N-356 terminal HEAT repeats of the Tor1 protein were critical for its role in oxidative stress endurance. conditions in which FL cells induced a strong phospho-Hog1 signal (Fig. 4C). This result suggested that Tor1 363 N-terminal HEAT repeats were required to induce a physiologic Hog1 oxidative stress response. 364

365
We tested whether during oxidative stress exposure, TORC1 signaling was physiologically downmodulated 366 in these tor1 mutants. Wild type and FL cells responded to plumbagin exposure as expected, by inhibiting 367 Rps6 phosphorylation as evinced by an absent P-S6 signal on Western blot in the first 20 minutes of the 368 time course. In contrast, Del381 cells failed to downmodulate P-S6, remaining in an abnormally activated 369 TORC1 state even immediately after exposure to plumbagin (Fig. 4D). To exclude an effect of plumbagin 370 on transcription from tetO, we examined TOR1 mRNA levels by qRT-PCR in each of the strains and found 371 no plumbagin effect (Fig. S3). 372 373 Since in S. pombe, translation initiation during oxidative stress is suppressed by Gcn2 kinase's 374 phosphorylation of eIF2α, we examined this response in C. albicans cells from the same cell lysates 375 exposed to plumbagin or vehicle that were assayed for P-S6. In vehicle the P-S6 and P-eIF2α signals of 376 Del381 cells were not substantially different from those of wild type or FL cells (Fig. 4D). In contrast, in 377 plumbagin-treated cells, the P-eIF2α signal of Del381 cells was stronger than in wild type or FL cells, 378 indicating stronger pro-inhibitory activity by Gcn2 (Fig. 4D). Hence in the absence of stress, Del381 cells' 379 TORC1-and Gcn2 signaling were aligned. During oxidative stress with plumbagin these pathways were 380 uncoupled in these cells, with inappropriately increased TORC1 signaling and increased counter-regulation 381 by inhibitory Gcn2 signaling. Translation initiation of most messages [29] is induced during TORC1 activation and repressed during 386 TORC1 inhibition, with specific regulatory mechanisms known for mTOR [67]. During oxidative stress, as 387 during TORC1 inhibition by rapamycin, translation initiation in S. cerevisiae is also inhibited through Gcn2-388 dependent phosphorylation of eIF2α, as well as through decreased ribosomal transit [68]; Gcn2 also 389 responds to oxidative stress to inhibit translation initiation in S. pombe [69]. Since TORC1 signaling as 390 reflected in the P-S6 signal was hyperactive in Del381 cells during oxidative stress, while paradoxically, 391 their translation inhibition through Gcn2 phosphorylation of eIF2a was also increased, we questioned 392 which signaling activity determined the final output, translational activity. We used a heterologous 393 message, GFP, whose transcription we could control from tetO and which presumably had no internal 394 sequences directing translational regulation. We examined whether appearance of the protein was 395 decreased in C. albicans during exposure to the superoxide generating compound plumbagin (Fig. 5A). As 396 expected, wild type cells expressing GFP from inducible tetO showed slower appearance of a GFP signal 397 during exposure to plumbagin than to vehicle (Fig. 5A). 398

399
The effect of plumbagin on GFP translation in cells containing distinct TOR1 alleles in which GFP was 400 expressed from the conditional MAL2 promoter (pMAL2) was then examined. When cells are shifted from 401 glucose to maltose, the MAL2 promoter is induced; we assayed appearance of a GFP protein signal after 402 this shift in cells containing TOR1 genotypes wild type, tor1/TOR1, Del381 and FL. Translation of GFP to 403 detectable levels occurred earlier in plumbagin-and vehicle-exposed Del381 cells than in FL cells (Fig. 5B). 404 This finding indicated that translation was aberrantly upregulated in Del381 cells and that the hyperactive 405 TORC1 signal, not simultaneously increased eIF2a phosphorylation and hence inhibition by Gcn2, 406 determined this final output. which induces cell wall stress, were strongly inhibitory to these cells on agar medium even when tetO-420 TOR1-Del381 expression was induced in the absence of doxycycline (Fig. 6A). The cell wall disrupting dye 421 Congo red, which binds to chitin fibrils [76], similarly had a very strong inhibitory effect on Del381 cells on 422 agar media, regardless of induction or inhibition of their tetO-TOR1-Del381 allele (Fig. 6A). Reexamining 423 this phenotype in liquid media, given the activation of cell wall integrity pathway signaling by cells' contact 424 with agar surfaces [73], we found that TOR1-Del381-expressing cells were extremely sensitive to 425 micafungin, whether the allele was overexpressed in 0 doxycycline or moderately repressed in 300 ng/ml 426 doxycycline; TOR1-FL-expressing cells were hypersensitive during tetO repression but to a lesser extent 427 ( Fig. 6B). To confirm that this strong phenotype was not specific to the drug or the cell wall component 428 inhibited (beta-1,3-glucan), we exposed the cells to the chitin synthase inhibitor nikkomycin. Since 429 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 nikkomycin competes with components of YPD for plasma membrane uptake through oligopeptide 430 transporters, we used synthetic complete medium (SC) for these experiments. Del381 and, to a lesser 431 extent, FL cells were also hypersensitive to nikkomycin during moderate tetO repression, though the 432 difference to wild type cells' inhibition was less stark (Fig. 6B). The Tor1 N-terminal HEAT domain hence 433 had a distinct role in responding to cell wall stressors. 434

435
Cell wall integrity signaling can be assayed in C. albicans by the Mkc1 phosphorylation state [73]. The P-436 Mkc1 signal intensity was minimally weaker in Del381 than in FL cells during micafungin exposure, though 437 not at baseline (Figs. 6C, S4A), suggesting that the strong growth defect of these cells during cell wall stress 438 was not related to a specific role of the Tor1 N-terminal HEAT domain in activating this pathway. 439 Micafungin exposure did not affect Rps-6 phosphorylation even in wild type cells (Fig. 6D failed to grow at this elevated temperature, regardless whether tetO was induced or repressed (Fig. 6E). response. This finding also suggests that it was not lack of translation inhibition that left Del381 cells 459 unable to tolerate elevated temperatures at which cells expressing a wild type TOR1 allele could grow. 460 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021  Rapamycin exposure induces the PKC pathway stress response that can be assayed by Mkc1 481 phosphorylation intensity in C. albicans [25]. Del381 cells responded more slowly to rapamycin with 482 increased P-Mkc1 signal intensity, and the increased signal receded more slowly, than in wild type and FL 483 cells (Fig. 7C), suggesting that N-terminal HEAT repeats may be involved in down-as well as upregulation 484 of the PKC pathway stress response during rapamycin exposure. 485 486 Plasma membrane stress induced translation inhibition but Tor1 HEAT repeats were not required for 487 membrane stress endurance. 488 489 C. albicans is exposed to cytoplasmic membrane stress in the host e.g. by bile [79,80] or by antimicrobial 490 peptides [81]. We used low concentrations of SDS (0.005-0.01%) to test whether Tor1 and its N-terminal 491 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

Lack of N-terminal HEAT repeats sensitized cells to rapamycin and delayed their PKC-dependent
The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 HEAT repeats are required in the response to membrane stress. During induction of tetO-TOR1-Del381 in 492 the absence of doxycycline, Del381 cells grew apparently normally on agar medium containing 0.005% 493 SDS; only during repression of tetO by doxycycline did residual low-level TOR1-Del381 expression fail to 494 support growth on this medium (Fig. 8A). 495 496 Even more striking than on solid media, cells overexpressing tetO-TOR1-Del381 had a smaller growth 497 defect, compared with wild type, in liquid SDS-containing medium than in vehicle (Fig. 8B), i.e. their 498 relative growth defect was partially rescued by SDS exposure. Membrane stress exposure by 0.01% SDS 499 in liquid medium induced downregulation of Rps6 phosphorylation in wild type and FL cells while Del381 500 cells were defective in this response (Fig. 8C). Repression of translation hence appeared to be a non-critical 501 component of the physiologic response to plasma membrane stress, since Del381 cells' defect in this 502 response did not correspond to a growth defect. known to affect C. albicans morphogenesis [37][38][39][40]. We examined hyphal growth on several filamentation-517 inducing media including M199 (Fig. 9A), Spider and RPMI (not shown); mutants in TOR1 invariably had 518 defective filamentation. Del381 and FL cells had decreased hyphal growth during overexpression of their 519 TOR1 alleles in 0 doxycycline, and during their moderate repression (Fig. 9A). On M199 medium of low pH 520 during anaerobic growth, the hyphal growth defect of Del381 > FL cells was partially rescued compared 521 to neutral pH (Fig. 9A). During anaerobic growth, cells overexpressing TOR1-Del381 and TOR1-FL from 522 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 tetO were hyperfilamentous, while moderate repression of tetO resulted in a hyperfilamentous 523 phenotype of TOR1-FL expressing cells (Fig. 9A). Together these results indicated that the Tor1 N-terminal 524 HEAT repeats, like the entire Tor1 protein, participated in a variety of signaling events whose final output 525 is hyphal growth and which are modulated depending on multiple external and internal cellular 526 conditions. 527 528 TORC1 inhibition with rapamycin was previously found to induce aggregation of C. albicans cells in liquid 529 Spider medium through induction of adhesin gene expression [38]. We questioned whether Tor1 N-530 terminal HEAT repeats play a role in aggregation. Del381 cells aggregated excessively even during 531 derepression of tetO-TOR1-Del381 in the absence of doxycycline (Fig. 9B); aggregation was somewhat 532 more pronounced during moderate tetO inhibition with 300 ng/ml doxycycline (Fig. 9B). Hence Tor1 N-533 terminal HEAT repeats contribute to repressing aggregation during physiological TORC1 activity. (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ; https://doi.org/10.1101/2021.11.05.467412 doi: bioRxiv preprint exhibited decreased expression [88]. While ribosome biogenesis genes were increased (NOP4 and 6, RRP8, 554 DIP2, C1_04710C), paradoxically, starvation responses were also increased, such as components of the 555 glyoxylate cycle (ICL1, MLS1), urea degradation (DUR1 and 3) and arginine degradation (CAR1 and CAR2) 556 pathways, import of glucose (HGT1,2,8,14 and 16), of amino acids and oligopeptides (CAN1 and 2, OPT1 557 and 2) and of nucleobases and their precursors (FUR4, XUT1) with a concomitant decrease in amino acid 558 biosynthetic genes like MET15 and 16, ARG3, LYS1, 9 and12, HIS4, 5 and 7 ( Supplementary Files 1 and 2). 559 Similar patterns were seen when examining genes whose expression increased during repression of tetO-560 TOR1-Del381 in doxycycline, again reflecting genes whose physiologic repression was abnormally released   [3][4][5][6]. In general, repressing the TOR1-Del381 allele with doxycycline (30 µg/ml) increased the number of 580 differentially regulated genes identified previously, while GSEA analysis indicated that the processes 581 affected following doxycycline treatment were similar to those perturbed in these cells without 582 doxycycline ( Supplementary Files 1-4). Many similar processes were affected in FL cells following 583 doxycycline suppression, including ribosome biogenesis, hypoxia and oxidative stress (Fig. 10B). 584 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 Expression of the genes encoding inducers of hyphal growth, the transcriptional regulator Ume6 [89] and 585 the hypha-promoting cyclin-like protein Hgc1 [90], was abnormally increased in Del381 cells during strong 586 tetO repression (Fig. 10C). This finding may help explain the hyperfilamentous phenotype exhibited by 587 Del381 cells under specific conditions, i.e. during anaerobic growth on acidic medium (Fig. 9A). Expression 588 of both UME6 and HGC1 decreased following suppression of the TOR1-FL allele (Fig. 10C). Uniquely, 589 Del381 cells exhibited increased expression of the hyphal wall protein-encoding gene HYR1 in addition to 590 the filamentation-inducing transcription factor UME6 (Fig. 10B), as well as cell wall protein-encoding 591 genes ALS2, ALS4, ALS6 and HWP1, which may explain these cells' hyperaggregation phenotype (Fig. 9B). We set out to examine whether specific roles could be assigned to a region of Tor kinase that is highly 607 divergent between fungal and human cells, comprising the 8 most N-terminal HEAT domains. Cells whose 608 only TOR1 allele was transcribed from repressible tetO revealed specific functions to which this region 609 contributes, when phenotypes of cells expressing an N-terminally truncated predicted protein were 610 compared to those expressing predicted full length Tor1. While we did not measure protein levels of Tor1 611 as we lacked a Tor1-specific antibody and do not yet have a functional epitope tagged Tor1, qRT-PCR 612 experiments showed that mRNA levels of both TOR1 alleles correlated with the concentrations of 613 doxycycline used to repress tetO (Fig. S1A). Similarly, growth of both tetO-controlled TOR1 genotypes 614 correlated with doxycycline concentrations in the medium, with 1 µg/ml doxycycline providing nearly 615 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 complete repression of Del381 cells' growth, while a moderate doxycycline concentration of 300 ng/ml 616 had a small growth-repressive effect in this medium ( Fig. 2A). For many experiments going forward, we 617 therefore chose 300 ng/ml doxycycline in order to compare phenotypes in cells that had diminished but 618 not ablated TOR1 expression. Overexpression of either TOR1 allele from unrepressed tetO in the absence 619 of doxycycline did not lead to TORC1 hyperactivation in rich medium (Fig. 2B), as determined from 620 phosphorylation of ribosomal protein S6, an established C. albicans TORC1 activity readout [25]. A simple 621 explanation for this finding is that Tor1 kinase depends on the other components of TORC 1 for its activity, 622 since presumably many of its activators, inhibitors and substrates are brought to their sites of physical 623 interaction by the other complex components; overexpressing only the catalytic moiety of the complex 624 appears to be insufficient to hyperactivate the complex (Fig. 2B). 625 626 A major input activating TORC1 signaling is availability and quality of nitrogen sources. Lack of N-terminal 627 HEAT repeats had little impact on cells' growth rates in the non-preferred nitrogen source proline, in the 628 absence or presence of a moderate repressing doxycycline concentration (Fig. 2C). However, in two 629 preferred nitrogen sources, glutamine and ammonium sulfate, cells with Tor1 lacking N-terminal HEAT 630 repeats grew more slowly, and during moderate tetO repression grew only to low densities (Fig. 2C). 631 Del381 cells similarly failed to maximize growth rates during optimal provision of phosphate-and carbon 632 sources (Figs. 2E and 3A). These findings together indicated that the TORC1 response to favorable 633 nutritional conditions, to coordinate anabolic processes like translation and DNA replication with provision 634 of metabolic intermediates and harvesting of energy from carbon sources, depends on at least one 635 function provided by the N-terminal HEAT repeats comprising amino acid residues 1-381. Since this failure 636 occurred in response to three central macronutrients, nitrogen-, carbon-and phosphate sources, we 637 speculate that the defective activity of N-terminally truncated Tor1 relates not to defective sensing of 638 individual nutrients, but to misregulation of a pro-anabolic output that normally results from TORC1's 639 integration of favorable nutritional inputs together with absent perception of unfavorable stressors. In 640 the host, inability to take advantage of favorable nutrient conditions will place a member of the mucous 641 membrane microbiome at a distinct disadvantage to other organisms that maximize their growth rates 642 when nutritional conditions permit. The ability to optimally accelerate growth during favorable nutrient 643 conditions is therefore a critical role of TORC1 for C. albicans. 644 645 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ; https://doi.org/10.1101/2021.11.05.467412 doi: bioRxiv preprint While Del381 cells were unable to appropriately accelerate growth in optimal conditions, they 646 inappropriately activated TORC1 signaling when nutrient conditions were suboptimal, as indicated by an 647 excessive P-S6 signal in the nonpreferred nitrogen source proline and in a lower glucose concentration of 648 0.25% (Figs. 2D and 3D). Rps6 phosphorylation was not simply indiscriminately activated, since Del381 649 cells growing in the absence of glucose (without an added carbon source) or in the nonfermentable carbon 650 source glycerol, had no detectable P-S6 signal, in line with wild type and FL cells (Fig. 3D). Hence Del381 651 cells seemed to be lacking a fine-tuning function of TORC1 activity, while "on" and "off" switches remained 652 functional. conditions, their slower growth may have been due to futile protein production with wasteful 657 consumption and premature exhaustion of scarce resources. We observed slower oxygen consumption of 658 cells overexpressing TOR1-Del381 compared to those overexpressing TOR1-FL, or to wild type cells, during 659 growth in glucose (Fig. 3E), suggesting that energy extraction of these cells from glucose was not 660 coordinated with translation. To examine translational control in these cells from another aspect, we 661 assayed the behavior of Gcn2 kinase through the phosphorylation state of its target, eIF2a. Strikingly, 662 Gcn2 appeared to be hyperactive in Del381 cells in most conditions (except for cell wall stress) compared 663 to wild type and FL cells: eIF2a was hyperphosphorylated, indicating inhibition of translation initiation by 664 Gcn2, in conditions where Del381 cells' P-S6 signal was hyperintense indicating excessive translation 665 activation e.g. during growth in proline (Fig. 2D). (Hyperphosphorylation of eIF2a additionally occurred in 666 Del381 cells in conditions where the P-S6 signal was undetectable e.g. in stationary phase and in medium 667 without an external carbon source or with glycerol ( Figs. 2D and 3D)). Experiments we then performed 668 with these cells during oxidative stress may have shed some light on this paradox, as described below. 669

670
We examined intracellular ROS of Del381 cells compared with FL and wild type cells, since slow oxygen 671 utilization of these cells (Fig. 3E) suggested that their putative decreased mitochondrial activity might 672 generate less intrinsic ROS, as was indeed the case (Fig. 4A). This alteration did not correspond to 673 increased tolerance of extrinsic oxidative stress, however: to the contrary, Del381 cells were strongly 674 hypersensitive both to peroxide-and to superoxide anion-stress, as induced by H2O2 and plumbagin, 675 respectively (Fig. 4B). Their oxidative stress signaling through the Hog1 kinase was weak (Fig. 4C), and their 676 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ; https://doi.org/10.1101/2021.11.05.467412 doi: bioRxiv preprint transcriptional oxidative stress responses were blunted (Figs. 10A and 11). During invasion of the host, C. 677 albicans is exposed to ROS from host phagocytes [49,52,92], hence an activity of Tor1 and its N-terminal 678 HEAT repeats in coordinating oxidative stress responses with HOG pathway signaling is predicted to be 679 indispensable for the host interaction. Gcn2 activity was more active than in wild type or FL cells, as assayed by eIF2a phosphorylation (Fig. 4D). 687 Since the final common effector of these two signaling pathways is the activity of the translational 688 machinery, we examined which of the two pathways that gave opposing signals -TORC1 an activating-689 and Gcn2 an inhibitory signal -had the decisive effect, or the "last word." We examined translation of a 690 heterologous message encoding GFP, which had no known intrinsic regulatory motifs responding to 691 oxidative stress, transcribed from regulatable promoters tetO or pMAL2. In wild type cells, translation of 692 a GFP message transcribed from inducible tetO was delayed during oxidative stress exposure (Fig. 5A). 693 GFP expressed from pMAL2 during oxidative stress was translated inappropriately early in Del381 cells 694 compared with wild type, tor1/TOR1 heterozygous, and FL cells (Fig. 5B), indicating that hyperactivation 695 of translation by TORC1 prevailed over translation inhibition by Gcn2. 696 697 Why was translation inhibition through Gcn2 upregulated in the same cells (we examined the same 698 protein extract samples for these comparisons of P-S6 and P-eIF2a signal intensities) in which translation 699 was hyperactivated as assayed through P-S6? Since in S. cerevisiae, Gcn2 directly phosphorylates the N-700 terminus of Kog1 [96] to downmodulate TORC1 activity, we surmised that the missing HEAT repeats in 701 Del381 cells normally participate in downregulation of translation by Gcn2 through Kog1 during oxidative 702 stress; when their activity is absent in these cells, Gcn2 develops compensatory hyperactivity. Excessive 703 GFP translation during oxidative stress in these cells showed that this compensatory upregulation of eIF2a 704 phosphorylation is not sufficient, in the end, to suppress hyperactive translation in Del381 cells. 705 706 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ; https://doi.org/10.1101/2021.11.05.467412 doi: bioRxiv preprint However, hyperactive translation did not invariably predict growth defects: Del381 cells exposed to the 707 plasma membrane stressor SDS had inappropriately high levels of P-S6 while enduring this stress well 708 when their TOR1-Del381 allele was not repressed (Fig. 8A-C). Strikingly, during TOR1-Del381 709 overexpression, these cells showed no growth defect in fluconazole, and even during moderate tetO 710 repression their growth defect in this drug was minor compared to their defect in vehicle, indicating 711 enhanced fluconazole endurance (Fig. 8D). These cells also endured amphotericin B exposure well while 712 their TOR1-Del318 allele was overexpressed (Fig. 8D). Conversely, growth defects of Del381 cells in a 713 specific stress condition were not strictly tied to defective signaling of the cognate stress response 714 pathway, as exemplified by phosphorylation of the cell wall integrity pathway component Mkc1 (Figs. 6C,  715   S4A). Instead these growth defects may be caused by inefficient respiration (Fig. 3E)  processes that respond to combinations of environmental conditions to which the fungus is exposed, as 726 (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 shown in Table 1. In the experiments we report here, single stressors were examined in isolation; actually 727 during infection, combinations of these stressors act on C. albicans as often emphasized by Brown and 728 colleagues e.g. in [100]. 729

730
Of particular interest will be investigation of roles of N-terminal HEAT repeats in metabolic regulation. 731 TORC1 inhibition by rapamycin is known to change the metabolome within minutes, suggesting that 732 TORC1 regulates metabolic enzymes by posttranslational modification [101]. When some signals to TORC1 733 cannot be received and others from TORC1 cannot be sent because the interaction domains for relevant 734 up-or downstream partners are absent, specific metabolic derangements would be expected. In addition 735 to decreased oxygen consumption of Del381 cells (Fig. 3E), our transcriptional analysis also shows 736 perturbation of carbon source metabolism in Del381 cells. 737

738
Del 381 cells' inability to grow on lactate is notable. They grew well on other non-fermentable carbon 739 sources (Fig 3B), indicating that defective growth on lactate is not simply attributable to defective 740 respiration. These cells also grew well on medium of pH2 with a fermentable-or non-fermentable carbon 741 source (Fig. S2), suggesting that a more specific defect than acid stress intolerance is responsible for this (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 We focused on TORC1 though truncation of Tor1 is likely to also affect its role in the rapamycin-insensitive 758 TOR complex 2 (TORC2) [107]. The prominence of inappropriately active anabolic processes in Del381 cells 759 during oxidative stress, like ribosome biosynthesis and translation, indicates that Tor1 N-terminal HEAT 760 repeats in stress responses we examined, act in the context of TORC1. Which TORC2 functions are 761 disrupted by loss of Tor1 N-terminal HEAT repeats will need to be studied in further investigations of this 762 complex's interaction partners. 763

764
Our findings highlight the importance of intact, physiologically modulated TORC1 signaling in specific 765 stress responses critical to C. albicans' interaction with the human host (Fig. 12). Suboptimal TORC1 766 reactivity hence is predicted to impair C. albicans' virulence; differences in their TORC1 signaling system 767 were previously discussed as underlying the virulence differences between the frequent pathogen C. 768 albicans and its close relative, the rarely invasive C. dubliniensis [108]. Our findings also demonstrate that 769 stepwise functional dissection of distinct segments among the large number of interaction domains of the 770 Tor1 kinase is possible. Together with identification of interaction partners for each segment, as well as 771 changes in the phosphoproteome and metabolic shifts during perturbation of distinct segments, the 772 anatomy of the afferent and efferent pathways to each element of the "brain of the cell" can be defined 773 in more detail than has so far been undertaken. Small molecules that disrupt some of these pathways 774 could become "nonimmunosuppressive rapamycin analogs" as envisioned by Cruz et al. [6]. In addition to 775 an approach that targets the highly conserved Tor kinase domain in a more fungal-selective manner [6], 776 the substantial structural differences between human and fungal TORC1 components including the Tor 777 kinase itself (Fig. 1) could be exploited for fungal-specific targeting of essential functions, particularly those 778 required during the host interaction. (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 shown in Table S2 and oligonucleotides shown in Table S3,  Cell lysis and Western blotting were performed as described in [25]. Antibodies used are listed in Table  807 S4. For densitometry, ImageJ (imagej.net/welcome) software (opensource) was used to quantitate signals 808 obtained either from KODAK Image Station 4000MM or from Azure biosystems c600. 809 810

Growth of cell dilution spots on solid media 811 812
Serial cell dilution spotting assays on agar media was performed as previously described [42]. Briefly, cells 813 recovered from glycerol stock were grown on YPD agar media for 36-48 h, then washed in 0.9% NaCl and 814 diluted with 0.9% NaCl in 5-fold steps from a starting OD600 of 0.5 in a microtiter plate, then pin transferred 815 to different agar media. Plates were imaged after 48 hours of incubation at 30 o unless stated otherwise. 816 817 Oxygen consumption measurement 818 819 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 Oxygen consumption was measured using a Clark-type electrode (dual digital-model 20; Rank Brothers. 820 Ltd., Cambridge, United Kingdom) at 25 o , according to the manufacturer's instructions. C. albicans cells 821 were grown to mid-logarithmic phase in YPD at 30°, washed twice in 0.9% NaCl and resuspended in the 822 same solution. Cell suspensions were prepared at 5x10 6 cells/ml. As a control for oxygen saturation, one 823 chamber was filled with 700 µl 0.9% NaCl; in the second chamber 650 µl of cell suspensions was added to 824 50 µl glucose (to achieve a 2% final glucose concentration). Oxygen saturation was recorded every 3 825 (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 Cells from YPD plates with 10 ng/ml doxycycline were collected and washed once in 0.9% NaCl, then 851 inoculated into YPD medium containing 100 ng/ml doxycycline with a starting OD600 of 0.2. Cells were 852 grown at 30 o till exponential phase (~4 h). Cells were washed twice in 0.9% NaCl and inoculated into 5.5 853 ml Spider medium (with additional 0.3 mM histidine) with or without 300 ng/ml doxycycline, with final 854 Transcriptomes of JKC1549 (tor1/tetO-TOR1) and JKC1441 (tor1/tetO-TOR1-Del381) were compared with 877 those of the heterozygous control JKC1347 (tor1/TOR1). Each strain was cultured for 24 h on solid YPD 878 containing 5 ng/ml of doxycycline. Cells were harvested from plates, washed twice in phosphate buffered 879 saline and inoculated into 50 ml of YPD to an OD600 of 0.1 in the absence or presence of doxycycline (30 880 µg/ml). Cells were harvested after 2 h and 8 h growth. Cell pellets were flash frozen in liquid N2 and RNA 881 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 extraction was performed using the Qiagen RNeasy mini-kit as described [110]. The microarrays used in 882 this study were designed from assembly 21 of the C. albicans genome using eArray from Agilent 883 Technologies (design ID 017942) [111]. A total of 6,101 genes (including 12 mitochondrial genes) are 884 represented by two sets of probes, both spotted in duplicate. Total RNA (100 ng) was labelled with Cy5 or 885 Cy3 using the Two-Color Low Input Quick Amp labelling Kit (Agilent Technologies) according to the 886 manufacturer's instructions. Array hybridization, washing, scanning and data extraction was carried out 887 as described by Haran et al. [110]. Each condition was examined with four biological replicates including 888 dye swapped replicates. Data for each feature was background corrected and expression normalized using 889 the Lowess transformation method in GeneSpring GX13 (Agilent Technologies). Differential expression 890 between pairs of conditions was examined by comparing relative expression ratios (Log2 values) among 891 genes that satisfied a post-hoc test (Storey with Bootstrapping) with a corrected Q value £0.05. Gene set 892 enrichment analysis (GSEA) was carried out as described by Haran et al. [110]. The full data set can be 893 accessed at the NCBI GEO repository (Accession GSE182186). 894

895
To compare the transcriptional responses of JKC1361 (wild-type), JKC1549 (tor1/tetO-TOR1) and JKC1441 896 (tor1/tetO-TOR1-Del381) to plumbagin, cells were grown as above and used to inoculate 200 ml YPD 897 supplemented with 5 ng/ml doxycycline at OD600 of 0.3. These cells were grown for 3.5 hours, harvested 898 and used to inoculate a fresh 50 ml YPD at OD600 of 0.4. These cultures were treated with plumbagin (10 899 µM) and incubated at 30˚ at 200 rpm and RNA was harvested at 30 min and 60 min incubation, as 900 described above. A wild-type JKC1361 culture was also treated with DMSO alone (50 µl) in a control 901 experiment. Three biological replicates were generated for each time point and RNA samples were pooled 902 for cDNA sequencing Using Oxford Nanopore Technologies (Oxford Nanopore Technologies, Oxfordshire, 903 UK) PCR-cDNA Sequencing Kit (SQK-PCS108) and barcoding kit (SQK-LWB001) according to the 904 manufacturer's instructions. Briefly, total RNA (50 ng) was reverse transcribed to cDNA using the PCR-905 cDNA Sequencing Kit (Oxford Nanopore Technologies) and SuperScript IV reverse transcriptase (Carlsbad,906 California, United States) according to the Oxford Nanopore Technologies protocol (version 907 PCB_9037_v108_revF_30Jun2017). Following this, cDNA was amplified and indexed using the PCR 908 Barcoding Kit (Oxford Nanopore Technologies) and LongAmp Taq 2X master mix (New England Biolabs). 909 The resulting cDNA libraries were sequenced on a MinION (MIN101B) device with a FLO-min106 SpotON 910 R9.4 Flowcell using the MinKNOWN software v1.7.10 (Oxford Nanopore Technologies). Fast5 reads that 911 passed filtering where basecalled and demultiplexed using the Albacore software v2.2.3 (Oxford 912 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 Nanopore Technologies). Porechop v0.2.4 (https://github.com/rrwick/Porechop) was used to further trim 913 reads and a total of 5.16 million reads were obtained. Sequence reads were aligned to the SC5314 genome 914 (Assembly 21) using BWA-MEM and the resulting SAM files were analysed in Strand NGS (Strand Life 915 Sciences). Differential gene expression between pairs of conditions was examined using genes that 916 satisfied a post-hoc test (Storey with Bootstrapping; Q £0.05) using GSEA [110]. Raw data can be accessed 917 at NCBI (Accession PRJNA753651). 918 919

Statistical analysis 920 921
For growth curves, the total area under the curve and its standard deviation were obtained and 922 normalized based on the mean area for the control group, within a 95% confidence interval, using XY 923 analysis in Prism 9 GraphPad (GraphPad Software, Inc., CA, USA). Statistical significance among groups 924 was determined using an unpaired t-test with Welch's correction, with two-tailed P values.  Domain annotation is adapted from Baretic et al. [18] according to the KmTor1 sequence. (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 from tetO were pre-grown in YPD medium for 4 h, and inoculated in YPD with increasing concentrations 944 of doxycycline (Doxy). OD600 was read every 15 minutes. B. Western blot of cells of indicated TOR1 945 genotypes, wild type (TOR1/TOR1), Del381 and FL, grown in YPD with 5 ng/ml doxycycline; protein extracts 946 probed with antibody to phosphorylated Rps6 (P-S6), total Rps6 (S6) and tubulin (Tub) as loading control. ng/ml doxycycline for 4 h were inoculated into YNB without Pi with 5 ng/ml doxycycline containing 969 nutrient concentrations equivalent to, or distinct from, standard YNB (2% Glu, 2% glucose, 10 mM Pi; 970 0.25% Glu, 0.25% glucose, 10 mM Pi; 0% Glu, no added direct carbon source, 10 mM Pi; 2% Gly, 2% 971 glycerol, 10 mM Pi; 0.05 mM Pi, 2% glucose, 0.05 mM Pi). Protein extracts were probed with antibody to 972 phosphorylated Rps6 (P-S6) or phosphorylated eukaryotic translation initiation factor 2A (P-eIF2a), and 973 tubulin (Tub) as loading control. Dens: signal intensity ratio of P-S6 or P-eIF2a to Tub. (TOR1/TOR1, 974 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021  genotypes were pre-grown in YPD medium with 5 ng/ml doxycycline for 3.5 h and diluted into fresh YPD 988 medium with 5 ng/ml doxycycline with either 10 µM Plumbagin (Plum) or DMSO as vehicle (Veh). Total 989 protein extract was probed with antibody to phosphorylated Hog1 (P-Hog1) and the PSTAIRE antigen of 990 Cdc28 as loading control (C), or with antibody to phosphorylated Rps6 (P-S6) and eIF2a (P-eIF2a ), and 991 tubulin (Tub) as loading control (D). Dens: signal intensity ratio of P-Hog1 to Cdc28 (C) or P-S6 or P-eIF2a 992 (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021  (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 A. Mkc1 phosphorylation during cell wall stress. Cells of indicated genotypes were pre-grown in YPD 1129 medium with 5 ng/ml doxycycline for 3.5 h, and inoculated into YPD medium with 5 ng/ml doxycycline 1130 and 10 ng/ml micafungin (Mica). Protein extract was probed with antibody to phosphorylated Mkc1 (P-1131 Mkc1) and the PSTAIRE antigen of Cdc28 as loading control. Dens: signal intensity ratio of P-Mkc1 to 1132 Genes showing 1.5 fold expression increase in D381 relative to heterozygous TOR1/tor1 2 Genes showing 1.5 fold expression decrease in D381 relative to heterozygous TOR1/tor1 3 Genes showing increased expression in D381 following 2h exposure to Doxycycline 4 Genes showing decreased expression in D381 following 2h exposure to Doxycycline 5 Genes showing increased expression in tetO-TOR1-FL following 2h exposure to Doxycycline 6 Genes showing decreased expression in tetO-TOR1-FL following 2h exposure to Doxycycline 7 Gene expression SC5314 exposed to 10 µM Plumbagin vs SC5314 exposed to DMSO 8 Gene expression in SC5314 vs tetO-TOR1-D381 in 10 µM Plumbagin 1151 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101 doi: bioRxiv preprint Fig. 1 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

Figures
The copyright holder for this preprint this version posted November 5, 2021. ; Fig. 2 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ; Fig. 3 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ; Fig. 4 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ; Fig. 6 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ; Fig. 7 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101 doi: bioRxiv preprint Fig. 8 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ; Fig. 9 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101 doi: bioRxiv preprint Fig. 10 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021.

Fig. 11
. CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ; Fig. 12 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 Fig. S1 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021  . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021  . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021  . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021  . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted November 5, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021