Temporal transcriptional response of Candida glabrata during macrophage infection reveals a multifaceted transcriptional regulator CgXbp1 important for macrophage response and drug resistance

Candida glabrata can thrive inside macrophages and tolerate high levels of azole antifungals. These innate abilities render infections by this human pathogen a clinical challenge. How C. glabrata reacts inside macrophages and what is the molecular basis of its drug tolerance are not well understood. Here, we mapped genome-wide RNA polymerase II (RNAPII) occupancy in C. glabrata to delineate its transcriptional responses during macrophage infection in high temporal resolution. RNAPII profiles revealed dynamic C. glabrata responses to macrophage with genes of specialized pathways activated chronologically at different times of infection. We identified an uncharacterized transcription factor (CgXbp1) important for the chronological macrophage response, survival in macrophages, and virulence. Genome-wide mapping of CgXbp1 direct targets further revealed its multi-faceted functions, regulating not only virulence-related genes but also genes associated with drug resistance. Finally, we showed that CgXbp1 indeed also affects azole resistance. Overall, this work presents a powerful approach for examining host-pathogen interaction and uncovers a novel transcription factor important for C. glabrata’s survival in macrophages and drug tolerance.

The overall transcriptional response was highly diverse with each group exhibiting a unique 86 temporal transcriptional pattern ( Figure 1C). Interestingly, while some genes were induced 87 immediately (0.5 h, Group 1, n = 181) upon internalisation by macrophages, transcriptional 88 induction of over 80% of genes (Group 2-6, n = 897) did not happen until later (2-8 h). 89 Besides, their expression patterns were highly variable, illustrating the complex and dynamic 90 nature of C. glabrata transcriptional response during macrophage infection. 91 processes during the infection process ( Figure 1D, Supplementary File 2). In the immediate 93 response (0.5 h), genes (Group 1, n = 181) were significantly enriched in processes such as 94 adhesion, responses to copper ion and nitrogen compound, positive regulation of nuclear 95 export in response to glucose starvation, lipid oxidation, and ATP synthesis ( Figure 1D). This 96 indicated that C. glabrata experiences nutrient and energy deprivation immediately upon 97 entry to macrophages ( Figure 1E). Alternatively, the induction of ATP biosynthesis genes 98 may reflect a strong demand for energy by C. glabrata to deal with the host's attacks and/or 99 to adapt to the host microenvironment. Subsequently (2 h post phagocytosis), C. glabrata 100 underwent a major metabolic remodelling presumably to prepare for growth and generate 101 energy, as reflected by the next wave of transcriptional induction for genes (Group 2, n = 102 171) involved in the TCA cycle, biosynthesis of inosine 5' monophosphate (IMP), carboxylic 103 acid, amino acid, nucleotide, and precursor for metabolite and energy ( Figures 1C&D). In 104 addition, cell cycle arrest and DNA damage checkpoint genes (CgMEC3, CgGLC7, 105 CAGL0G07271g, and CAGL0A04213g) were also strongly induced at this early stage, and C. were markedly induced at the early stages (0.5 and 2 h). Therefore, virulence-centric 112 biological processes were among the most immediate C. glabrata responses upon 113 macrophage phagocytosis ( Figure 1E), implying the importance of the early transcriptional 114 response towards its adaptation and survival in macrophages. 115 formation involves cell growth and proliferation, this observation potentially suggests that the 143 cells are preparing for growth, and this is consistent with the concomitant induction of the 144 iron homeostasis genes that are also necessary for proliferation. Altogether, the overall results 145 revealed details into the dynamic stage-wise responses of C. glabrata during macrophage 146 infection ( Figure 1E). 147 148

Identification of potential transcriptional regulators of early temporal response 149
We next attempted to identify the potential transcriptional regulators for the 150 chronological transcriptional response. Remarkably, more than 25% of C. glabrata 151 transcription factor (TF) genes (n = 53) were expressed during macrophage infection (Table  152 1), with 39 TF genes showing a temporal induction pattern ( consistently slower mortality rate by ~20-30% compared to larvae infected by wildtype cells, 231 suggesting that the loss of Xbp1 function attenuated the virulence ( Figure 4B). The 232 attenuated virulence was rescued in the complemented strain ( Figure 4B). Therefore The direct functions of CgXbp1 were determined using GO analysis and were 273 significantly associated with major biological processes important for host infection such as 274 "transport", "response to stress", "carbon and nitrogen metabolism", and "biofilm formation" 275 (Table 2 antifungal fluconazole. Serial dilution spotting assay on solid media showed that Cgxbp1∆ 290 mutant had higher resistance to fluconazole compared to wildtype ( Figure 6B), and the 291 resistance was restored to the wildtype level in the complemented strain ( Figure 6B). 292 Importantly, the altered resistance is not due to an intrinsic difference in growth rate between 293 the two strains, as demonstrated by their indistinguishable growth rates in the absence of drug 294 in liquid media ( Figure 6C). However, in presence of fluconazole (64 µg/mL), the Cgxbp1∆ 295 mutant was able to grow faster and to a higher density than wildtype ( Figure 6C). It is 296 interesting to note that a biphasic growth curve was observed for both strains in the presence 297 of fluconazole, suggesting the existence of two populations of cells (sensitive versus 298 resistant). Consistently, we also noted from the spotting assay relatively more resistant 299 colonies in the Cgxbp1∆ mutant as compared to wildtype. To confirm this, we performed a 300 CFU assay by plating an equal number of exponentially growing wildtype, Cgxbp1∆ mutant 301 and complemented cells on YPD medium with or without fluconazole (64 µg/mL). The 302 Cgxbp1∆ mutant displayed ~8-fold higher CFUs on fluconazole compared to that of wildtype 303 and the complemented strain ( Figure 6D homo-dimer. Interestingly, the latter motif (STVCN7TCT) was found at a higher frequency 379 (~3 fold) than the common TCGAG motif from the CgXbp1 MYC binding sites, suggesting that 380 CgXbp1 can also form a dimer with another transcription factor that recognizes the 381 STVCN7TCT sequence and that this hetero-dimer controls a larger number of genes than by 382

CgXbp1 alone. 383
The CgXbp1 MYC ChIP-seq data also confirms the transcriptional phenotype of the 384 Cgxbp1∆ mutant demonstrating that CgXbp1 is a pivotal regulator of many processes that Chromatin immuno-precipitation was carried out using 2 µL of a commercially available 437 anti-RNA polymerase II subunit B1 phospho-CTD Ser-5 antibody (Millipore, clone 3E8, cat. were then added. The mixture was further incubated at room temperature for another 1. Macrophage fungi infection assays were done as described earlier (Rai et al., 2013). 536 To prepare macrophages for infection assay, THP-1 monocytes were grown till 80% 537 confluence, harvested, and resuspended to a cell density of 10 6 cells/ml in complete RPMI medium. Phorbol-13-myrstyl-acetate (PMA) was added to the cell suspension to 16 nM final 539 concentration, mixed well, and 1 million cells were seeded in each well of a 24 well cell 540 culture plate. Cells were incubated for 12 hours in a cell culture incubator, the medium was 541 replaced with fresh pre-warmed complete RPMI medium, cells were allowed to recover from 542

Galleria mellonella infection assay for virulence analyses 555
Indicated C. glabrata strains were grown in YPD medium overnight, washed with 556 PBS thrice, and resuspended in PBS to a final cell density of 10 8 cells/ml. Next, 20 µL of this 557 cell suspension carrying 2 x 10 6 C. glabrata cells were used to infect G. mellonella larvae. 558 The infection was carried out three independent times, each on 16 to 20 larvae. An equal 559 volume of PBS was injected into the control set of larvae. Infected larvae were transferred to 560 a 37°C incubator, and monitored for melanin formation, morbidity and mortality for the next 561 seven days at every 24 hours. The number of live and dead larvae was noted for seven days, 562 and the percentage of G. mellonella larvae survival was calculated.
shaking at 200 rpm. Cells were harvested from 1 ml culture, washed with PBS, and were 567 diluted to an OD600 of 1. Next, five ten-fold serial dilutions were prepared from an initial 568 culture of 1 OD600. Subsequently, 3 μL of each dilution was spotted on YPD plates with or 569 without fluconazole (32 & 64 μg/mL). Plates were incubated at 30°C and images were 570 captured after 2-8 days of incubation. 571 572

Growth curve analyses 573
A single colony of the indicated strains was inoculated to liquid YPD medium and 574 grown for 14-16 h. The overnight grown culture was used to inoculate to YPD medium with 575 or without 64 µg/mL fluconazole at an initial OD600 of 0.1 in a 96-well culture plate. The 576 culture plate was transferred to a 96 well-plate reader, Cytation3, set at 30°C and 100 rpm. 577 The absorbance of cultures was recorded at OD600 nm at regular intervals of 30 minutes over 578 a period of 48 h. Absorbance values were used to plot the growth curve. 579 580

Protein extraction and western blotting 581
For protein extraction from macrophage-internalized C. glabrata cells, THP-1 582 macrophages were infected as described above. At the indicated time post-infection, 583 macrophages were lysed in sterile chilled water, and phagocytosed C. glabrata cells were 584 recovered and washed with 1X TBS buffer, transferred into 1.5 ml microcentrifuge tubes, and 585 stored at -80°C until use. C. glabrata cell pellets were resuspended in 1X lysis buffer (50 mM 586 HEPES, pH 7.5; 200 mM NaOAc, pH 7.5; 1 mM EDTA, 1 mM EGTA, 5 mM MgOAc, 5% 587 microcentrifuge tubes and resuspended cells were lysed by 6 rounds of bead beating on a 590 bullet blender. The sample was centrifuged at 12,000g at 4°C for 10 minutes. Supernatant 591 was carefully transferred to a new tube, and the resultant protein sample was quantified using 592 Biorad protein assay kit (DC protein assay kit, cat. no. 5000116), and stored in -80°C freezer. 593 For western analysis, 25 µg of protein samples were resolved on 12% SDS-PAGE gel and 594 blotted on methanol activated PVDF membrane (350 mA, 75 minutes in cold room). PVDF 595 membrane was transferred to 5% fat-free milk prepared in 1X TBST for blocking and 596   The colour scale represents the Z-score of the normalized RNAPII ChIP-seq signal. 804