Inhibition of classical and alternative modes of respiration in C. albicans leads to cell wall remodelling and increased macrophage recognition

The human fungal pathogen C. albicans requires respiratory function for normal growth, morphogenesis and virulence. As such the mitochondria represent an enticing target for the development of new antifungal strategies. This possibility is further bolstered by the presence of fungal specific characteristics. However, respiration in C. albicans, as is the case in many fungal organisms, is facilitated by redundant electron transport mechanisms that makes direct inhibition a challenge. In addition, many chemicals known to target the electron transport chain are highly toxic. Here we make use of chemicals with low toxicity in mammals to efficiently inhibit respiration in C. albicans. We find that use of the Nitric Oxide donor, Sodium Nitroprusside (SNP), and the alternative oxidase inhibitor, SHAM, prevent respiration, lead to a loss in viability and to cell wall rearrangements that increase the rate of uptake by macrophages in vitro and in vivo. We propose that SNP+SHAM treatment leads to transcriptional changes that drive cell wall re-arrangement but which also prime cells to activate transition to hyphal growth. In line with this we find that pre-treatment of C. albicans with SNP+SHAM leads to an increase in virulence. Our data reveals strong links between respiration, cell wall remodelling and activation of virulence factors. Our findings also demonstrate that respiration in C. albicans can be efficiently inhibited with chemicals which are not damaging to the mammalian host, but that we need to develop a deeper understanding of the roles of mitochondria in cellular signalling if they are to be developed successfully as a target for new antifungals. Author Summary Current approaches to tackling fungal infections are limited and new targets must be identified to protect against the emergence of resistant strains. We investigate the potential of targeting mitochondria, organelles required for energy production, growth and virulence, in the yeast human fungal pathogen Candida albicans. Our findings suggest that mitochondria can be targeted using drugs that can be tolerated by humans and that this treatment enhances their recognition by immune cells. However release of C. albicans cells from mitochondrial inhibition appears to activate a stress response that increases traits associated with virulence. Our results make it clear that mitochondria are a valid target for the development of anti-fungal strategies but that we must determine the mechanisms by which they regulate stress signalling and virulence ahead of successful therapeutic advance.

(NO) which inhibits respiration at the level of Complex IV and has a safe record for 105 use as a vasodilator [19] and in dermatologic applications [20]. NO is a molecule 106 produced by phagocytes as an antimicrobial and is known to inhibit respiration in 107 bacteria and fungi. NO has been evaluated as therapy in a variety of bacterial and 108 fungal infections, including Pseudomonas aeruginosa in cases of cystic fibrosis and 109 infections caused by dermatophytes [21][22][23]. NO works by the inhibition of 110 Cytochrome c Oxidase at low concentrations but is rapidly reversible by oxygen.  We find that C. albicans cells are highly adaptive to classical respiration inhibition but 123 that a combination of SHAM and the NO donor Sodium Nitroprusside (SNP) leads to 124 fitness defects and loss of viability. In addition, SNP+SHAM treatment leads to cell 125 wall organisation defects that unmask C. albicans cells leading to increased immune 126 cell recognition in cell culture and animal models. However, release of cells from 127 SNP+SHAM treatment leads to a rapid activation of the hyphal transition programme 128 and increased virulence in a mouse model. Our data suggest that mitochondria form 134 Our goal was to identify conditions that would allow for the reproducible inhibition of 135 classical and alternative respiration in C. albicans. To achieve this oxygen 136 consumption was measured upon exposure of cells to the nitric oxide donor SNP 137 which inhibits cytochrome c oxidase, and the alternative oxidase (Aox) inhibitor SHAM. 138 Initial titration experiments suggested that the concentrations of 1 mM SNP and 0.5 139 mM SHAM were suitable for our purpose. Recovery from NO induced inhibition of 140 classical respiration by SNP addition was rapid, with full restoration achieved within 141 30 min (Fig 1B). Further application of SNP did not lead to a reduction in oxygen 142 consumption suggesting that cells had switched to alternative respiration upon SNP 143 treatment ( Fig 1B). This was confirmed by subsequent addition of SHAM, which 144 decreased the respiration level significantly (Fig 1B). Addition of 2mM cyanide, which 145 has been shown to inhibit the parallel pathway in C. parapsilosis [44] was sufficient to 146 inhibit the remaining respiration suggesting the presence of a parallel pathway in C. 147 albicans. Addition of SNP+SHAM simultaneously led to significant loss of respiration 148 that was comparable to that observed upon cyanide addition ( Fig 1C). In support of a 149 rapid transition to alternative respiration, an increase in Aox2 protein levels was 150 detectable within 20 min following exposure to SNP (Fig 1D). Given the effects of SNP+SHAM on respiration we investigated whether co-treatment 153 led to effects on growth and viability. Incubation of C. albicans cells with SNP+SHAM 154 led to a decrease in growth and final biomass in a dose dependent manner (Fig 1E). 155 Treatment with SNP+SHAM for 3 hours did not have a significant effect on viability, 9 156 however, prolonged SNP+SHAM exposure resulted in decreased viability when 157 compared to untreated controls ( Fig 1F). These data suggest that C. albicans 158 possesses a robust and adaptable respiratory system that can be targeted using a 159 SNP + SHAM co-treatment approach.

SNP+SHAM inhibits respiration, growth and reduces viability in C. albicans
160 161 SNP+SHAM treatment induces structural re-arrangements within the cell wall 162 Previous studies suggest a link between respiration and cell wall integrity [8,16]. Our 163 goal was to determine whether chemical inhibition of respiration using SNP+SHAM 164 could negatively impact cell wall integrity. SNP+SHAM exposure led to a reduction in 165 growth and sensitivity to the cell wall damaging agents Calcoflour White (CFW) and 166 Congo Red (CR) (Fig 2A). We employed transmission electron microscopy to examine 167 cell wall structure following SNP+SHAM treatment ( Fig 2B). This analysis revealed 168 that the outer wall, composed of mannans and cell wall proteins, but not the inner cell 169 wall, was reduced in thickness when compared to untreated cells (Fig 2B and C). The 170 density of mannan fibres in this layer was also found to be increased by SNP+SHAM 171 treatment. These data suggest that SNP+SHAM causes cell wall changes which result 172 in an altered organisation of the outer cell wall. HPLC analysis of acid-hydrolysed cell 173 wall components showed that SNP+SHAM treatment caused no significant changes 174 in the bulk levels of chitin, glucan or mannan (Fig 2D). This suggests cell wall changes 175 induced by SNP+SHAM are due to changes in wall organisation rather than gross 176 changes in the levels of cell wall components. Interestingly we also observed the 177 presence of a large structure adjacent to the vacuole within SNP+SHAM treated cells 178 (Fig 2E). We confirmed this to be a lipid droplet using the neutral lipid stain LD540 (Fig

187
We also determined whether SNP+SHAM treatment led to the activation of Hog1, a 188 major oxidative stress response regulator that also influences cell wall biosynthesis  To identify potential regulators of the SNP+SHAM response that may be important in 195 cell wall organisation we treated C. albicans with a combination of SNP+SHAM and 196 caspofungin, which targets the cell wall ( Fig 3B). We observed that SNP+ SHAM 197 treatment led to caspofungin resistance, presumably a result of the alteration in cell 198 wall structure (Fig 3B) increasing it as in the wild-type strain ( Fig 3D). These data may suggest a role for 211 Upc2 in regulating cell wall changes in response to SNP+SHAM treatment.

213
Analysis of transcriptional changes in response to SNP+SHAM treatment 214 To further examine the changes caused by SNP+SHAM we examined the 215 transcriptional response of C. albicans to this treatment using RNAseq. RNA was 216 extracted from log-phase yeast cells exposed to 1 mM SNP, 0.5 mM SHAM, or both 217 for 30 minutes (Fig 4A and B and Supplementary Tables S2-S4). A short exposure 218 time was selected in order to capture early alterations as opposed to profiles from the 219 expansion of adapted cell lineages. SNP and SNP+SHAM treatment led to the 220 differential expression of over 1500 genes, whereas SHAM treatment led to a smaller 221 response with only 131 genes differentially expressed (Fig 4A and supplementary   222 tables S2-S4). A significant overlap between differentially expressed genes within SNP 223 and SNP+SHAM treatment data sets was observed ( Fig 4A). As expected the 224 alternative oxidases AOX2 and AOX1 and genes required for a response to nitric 225 oxide, such as YHB1, were upregulated in both SNP only and SNP+SHAM groups.   Analysis of SNP+SHAM RNAseq data showed that several genes involved in chitin 247 synthesis and organisation were downregulated, including the chitinases CHT1, CHT2 248 and CHT4 (Fig 4B). Mannan biosynthesis and organisation genes were upregulated, 249 as were genes involved in β-glucan synthesis and organisation. As no significant 250 changes in relative glucan, mannan or chitin levels were detected by HPLC in acid-13 251 hydrolysed cell walls our data may suggest that SNP+SHAM induced transcriptional 252 changes lead to alterations in cell wall organisation as opposed to bulk synthesis.

253
Supporting this hypothesis we observed that several GPI-anchored cell wall protein 254 genes that play a role in crosslinking cell wall components were also differentially 255 expressed ( Fig 4B).    structures such as bud scars where chitin is exposed, and staining of the lateral cell 287 wall indicates that chitin is exposed at the surface of the cell wall. SNP+SHAM treated 288 cells exhibited a greater lateral wall staining compared to untreated cells (Fig 5A, B). 289 These results suggest that SNP+SHAM treatment causes exposure of the normally 290 hidden chitin in the cell wall.    Fig S4A). An increase in uptake of ndh51Δ cells was also observed 341 but not to the same degree as SNP+SHAM pre-treated cells ( Supplementary Fig S4B). combination of these factors led to a significantly higher outcome score for 365 SNP+SHAM pre-treated cells compared to untreated cells ( Fig 7A). This data 366 suggests that pre-treatment with SNP+SHAM does not reduce fitness of C. albicans 367 in the host, but may in fact prime a stress response that leads to increased virulence. 368 One possibility is that pre-treatment of C. albicans with SNP+SHAM may have primed 369 cells to activate traits associated with virulence when injected into the host. We 370 observed that cells pre-treated with SNP+SHAM exhibited a higher incidence of 18 371 filamentation than untreated cells. The rapid induction of filamentation is generally 372 associated with increased escape from macrophages and virulence [54,55]. However 373 it was not clear whether this effect was dependent on the presence of macrophages.

374
To determine this the incidence of hyphal formation was assessed after 90 min   to lead to an increase in the incidence of lipid droplets [66], which we also observed in  were also seen in goa1Δ by TEM, similar to SNP+SHAM treated cells (Fig 2B, C). 463 Mutations that affect mannan composition have also been shown to be important for    Table S1.    Microdilution assays were carried out in 96-well plates with caspofungin, with or 586 without 1 mM SNP + 0.5 mM SHAM, using synthetic complete medium, 2 % glucose, 28 587 50 mM MOPS pH 6. A dilution series of caspofungin was made using the appropriate 588 diluent between 3.9 ng/ml -8 µg/ml. C. albicans cells from an overnight YPD culture 589 were washed three times in PBS. The OD 600 was measured and adjusted to 2.0. This 590 cell suspension was diluted 1:100 into the appropriate diluent. One hundred microliters     Cells were treated with 0.5 mM SNP and 1.0mM SHAM for 30 min and processed for 1223 RNA Sequencing and downstream analysis as described in materials and methods.

1224
A full list of differentially expressed genes is provided.