A null mutant of ATP5 (atp1Δ/Δ) was constructed by PCR-mediated homologous recombination as described previously (Sasse and Morschhäuser, 2012). C. albicans SC5314 (wild type; Gillum et al., 1984) was used to generate the atp1Δ/Δ mutant strain (atp1Δ::FRT/atp1Δ::FRT) and atp1Δ/ATP1 gene-reconstituted strain (atp1Δ::FRT /ATP1::FRT). The primers used for gene deletion in this study are listed in Supplementary Table S1. The confirmation of the atp1Δ/Δ mutant and reconstituted strain (atp1Δ/ATP1) are performed by PCR amplicons using primer pairs as shown in Supplementary Figures S1 and S2. Strains were routinely grown in YPD broth or on YPD agar (1% yeast extract, 2% peptone, and 2% glucose) with or without compounds as indicated.

A mouse model of disseminated candidiasis was used to evaluate the virulence of the strains (Spellberg et al., 2003, 2005). Female BALB/c mice (18–22 g; Guangdong Medical Laboratory Animal Center, Foshan, Guangdong, China) were used for all experiments. Mice were injected via the lateral tail vein with either a suspension of 1 × 10⁵ cells or 1 × 10⁶ cells from each strain. Survival rate was calculated from 10 infected mice per strain. For determination of fungal burden, another three mice from each group were euthanized after 1, 24, 48, and 72 h infection. Kidney, spleen and liver were harvested, weighed, homogenized, and quantitatively cultured. In addition, at day 1 of infection, mice were killed and organs removed to fix in 10% buffered formalin, then embed in paraffin, sectioned and stained with Periodic Acid-Schiff for histological study. Mortality was represented with Kaplan–Meier survival curves and quantitative tissue burdens were marked in the log scale and compared in the Mann–Whitney test.

The animal experiments were performed under the guidance of a protocol approved by the Animal Study Committee of the Institute of Dermatology, CAMS, according to the National Guidelines for Animal Care. All animal experiments were carried out with permission from the Ethical Committee of Institute of Zoonosis, Jinan University, Guangdong, China (Ref no. 20080101).

The mouse macrophage cell line RAW264.7 (ATCC) was used for all assays. To investigate the susceptibility of C. albicans strains to the macrophages, macrophage cells were seeded in a 96-well plate at 5 × 10⁴ cells/ml in 150 μl of DMEM with 10% FBS overnight at 37°C in 5% CO₂. The C. albicans were diluted into DMEM at 2 × 10⁶ cells/ml, and 50 μl was added into the first column wells, mixed and then serially diluted 1:4 a total of six times. Plates were incubated at 37°C and 5% CO₂ for 24 h. Colonies were counted. The survival rate was calculated as the numbers of colonies with macrophages divided by the number of colonies in the absence of macrophages. P-values were determined using the unpaired Student’s t-test (She et al., 2015). For evaluating macrophage survival, macrophage cells were seeded in 24-well plate at 1 × 10⁶ cells in 1 ml of DMEM with 10% FBS, and cultured at 37°C and 5% CO₂ for 24 h. The C. albicans cells were suspended in DMEM with 10% FBS to 1 × 10⁶ cells/ml, then added into the wells. After 1 h of co-incubation, supernatant was removed from each wells, washed and stained with Hoecheset33258 for 1 h, and observed by a fluorescent microscope (Olympus, Japan) with a TRIT-C/Texas red filter set. Hoecheset-positive (damaged) and total macrophages were counted, and the percentage of damaged macrophages was calculated as the number of damaged macrophages divided by that of total macrophages (Yu et al., 2014).

The extent of damage of endothelial cells by the C. albicans was quantified by measuring lactate dehydrogenase (LDH) activity (Mayer et al., 2012). The human umbilical vein endothelial cell line HUVEC (ATCC CRL-1730, LGC Standards, Promocell) cells were seeded in a 96-well plate at 1 × 10⁵ cells/ml in 200 μl of DMEM with 10% FBS overnight at 37°C and 5% CO₂ for 2 days. Cells were washed and 100 μl DMEM with 2% FBS were added. The C. albicans were diluted into DMEM at 5 × 10⁵ cells/ml, and 100 μl was added to infect endothelial cells. Incubation was carried out at 37°C and 5% CO₂ for 15 or 24 h. Measurement of LDH activity with the Cytotoxicity Detection Kit was performed according to the manufacturer’s manual. Absorbance of the samples was measured at 490 nm. Medium only and low control values were subtracted from all sample values. Damage was expressed as percentage of the control strain. Each experiment was performed at least in triplicate.

The adhesion was observed visually (Nobile et al., 2006). The 24-well flat bottomed pre-sterilized microtiter plates were incubated in fetal bovine serum overnight at 37°C. A suspension of 1 × 10⁷ cells of C. albicans in Spider media was added to the wells and incubated at 37°C for 2 h. Non-adherent cells were removed by washing with PBS. Fresh Spider media was added to the corresponding wells and the plates were incubated at 37°C for 24 h. The wells were washed and photographed.

A 1 × 10⁶ cells/ml suspension of each strain in YPD, YPD+10% FBS, Spider and Lee’s liquid media was added to the wells of a 12-well flat-bottomed pre-sterilized microtiter plate and incubated at 37°C for 2 h. The plates were visualized under inverted microscope and photographed. In addition, cells of each strain were serial diluted and spotted on YPD, YPD+10% FBS, Spider, Lee’s and SLAD agar, and incubated at 37°C for 7 days. The colonies were photographed.

The biofilms were observed by Confocal laser scanning microscopy (CSLM) (Nobile et al., 2006). Briefly, all strains were grown in Glass Bottom Cell Culture Dish. After 24 h of incubation at 37°C, the resulting biofilms were washed and stained with 25 μg/ml Concanavalin A-Alexa Fluor 594 conjugate (C-11253; Molecular Probes, Eugene, OR) at 37°C for 1 h. CSLM was performed with a Zeiss LSM 510 upright confocal microscope (Carl Zeiss, Thornwood, NY, USA), using a Zeiss Achroplan 40×, 0.8-W objective, and a HeNe1 laser with an excitation wavelength of 543 nm. Moreover, biofilms activity was assessed by XTT reduction assay and the crystal violet assay, accordingly to previously described protocols (Peeters et al., 2008; Pierce et al., 2008; Krom and Willems, 2016). Each experiment was performed at least in triplicate.

Overnight cultures of tested strains in 5 ml of YPD at 30°C were collected, then diluted into 100 ml of YPD or Spider medium to obtain an OD₆₀₀ of 0.2 and incubated at 37°C for additional 6 h. Total RNA was extracted using the E.Z.N.A. Yeast RNA kit (Omega Bio-tek) following the manufacturer’s instruction. cDNA was synthesized and qRT-PCR was done as previously described (Guo et al., 2014).

To investigate the sensitivity of C. albicans strains to different stress agents, cells were overnight cultured, washed and suspended in PBS with an initial OD₆₀₀ of 1.0. Five microliters of 10-fold dilutions was spotted onto YPD agar without or with different stress agents. For testing the sensitivity of C. albicans strains to heat stress, plates were cultured at indicated temperatures for 2 days and then photographed. The minimum inhibitory concentrations (MICs) were determined for each strain according to the broth microdilution method of the Clinical and Laboratory Standards Institute (CLSI) M27-A3 protocol (Method M27-A3 CLSI, 2008).

All strains were routinely grown in 5 ml of YPD at 30°C overnight, washed with PBS, then inoculated in 100 ml of YPD with an OD₆₀₀ of 0.02. Shake cultures were grown at 30°C and OD₆₀₀ of each strain was measured every two hours.

All strains were grown in YPD overnight at 30°C. Mitochondrial complex V activity was measured by an assay kit according to the manufacturer’s protocol. For determinations of intracellular ATP concentrations, an aliquot of 1 × 10⁶ cells from each strain was mixed with the same volume of BacTiter-GloTM reagent (Promega Corporation, Madison, WI, USA) and incubated for 5 min at room temperature as described previously (Zhang et al., 2013). ROS measurement was performed by using a oxidation-sensitive fluorescent dye DCFDA (Guo et al., 2014). Briefly, a suspension of 2.5 × 10⁶ cells was stained with DCFDA (20 μg/ml) at 37°C for 20 min. The emission spectra at 488 nm and excitation spectra at 595 nm were determined by FACScan flow cytometer (Becton Dickinson). The cyanine dye JC-1 was used for determination of mitochondrial membrane potential of each strain (She et al., 2015). A suspension of 2.0 × 10⁶ cells was incubated with 5 μM JC-1 at 37°C for 15 min and excitation spectra at 595 nm was determined by FACScan flow cytometer with emission spectra at 488 nm (Becton Dickinson).

The atp1Δ/Δ mutant displayed moderately reduced growth rate in vitro even when grown in rich medium (Figure 1A, Supplementary Table S3), which is similar to other C. albicans mutants with reduced virulence (Murad et al., 2001; Hwang et al., 2003; Chauhan et al., 2005; Lopes da Rosa et al., 2010; Chang et al., 2015; Lee et al., 2015). We speculate that such slow growth of atp1Δ/Δ in vitro may affect its virulence. A murine model of hematogenously disseminated candidiasis was first used to measure virulence changes in atp1Δ/Δ. All of the mice infected intravenously with either 1 × 10⁵ cells or 1 × 10⁶ cells of atp1Δ/Δ survived post 30 days infection (Figure 1B). By contrast, mice infected with WT began to die on days 7 and 1 post infection, and all mice were moribund on days 14 and 3, respectively. On the other hand, mice infected with 1 × 10⁵ cells and 1 × 10⁶ cells of atp1Δ/ATP1 began to die on days 8 and 4 and all mice were moribund on days 15 and 8, respectively. Clearly, ATP1 is required for C. albicans virulence, and the loss of virulence in atp1Δ/Δ is correlated with is not slow growth in vitro.

Next, to correlate the avirulence of atp1Δ/Δ with its growth in host tissues, the fungal load was evaluated in kidney, liver and spleen (Figure 1C). The colony forming unit (CFU) was the same for all strains 1 h post infection. However, tissue loads in either WT or atp1Δ/ATP1 increased significantly (by about 2 orders of magnitude) in kidney while atp1Δ/Δ load gradually reduced post infection and dropped by about one order of magnitude at 48 h post infection. The fungal loads in liver and spleen reached their peaks at 48 h for all three strains but reduced fungal cell populations were observed in mutant during 72 h post infection when compared to the control strains. At 72 h post infection with atp1Δ/Δ, fungal cells continued to be cleared from the kidney, liver and spleen. The significant reduction of fungal burden in the kidneys infected with the atp1Δ/Δ within 3 days of infection suggests that ATP1 is required for kidney invasion and is necessary for the organism to persist in other tissues at later time points.

The low fungal burden in the kidneys of mice infected with the atp1Δ/Δ was verified by histopathology. After 24 h of infection, the kidneys infected with atp1Δ/Δ were scant to detect in tissue sections, as demonstrated by a absence of fungal hyphae and inflammatory cells, even though the fungal load was still positive according to CFU count. By contrast, kidneys infected with either WT or atp1Δ/ATP1 displayed the expected micro-abscesses consisting of numerous inflammatory cells surrounded by abundant hyphae or pseudohyphae (Figure 1D), suggesting that the avirulence of the atp1Δ/Δ is likely due to defects in tissue invasion and colonization ability.

Macrophages play a key role in host defense against C. albicans infections (Mayer et al., 2012). In order to invade the host organs and begin colonization, blood-borne C. albicans must evade the attacks from the immune system prior to crossing the endothelial cell barrier of blood vessels (Grubb et al., 2008). We therefore investigate whether ATP1 is required for C. albicans sensitivity to host killing events. The sensitivity of the atp1Δ/Δ to macrophages and its capability to damage macrophages were both measured. Macrophage killing tests revealed that atp1Δ/Δ is more vulnerable to macrophages, as demonstrated by the survival of fewer atp1Δ/Δ cells after an overnight incubation with macrophages than those from either WT or atp1Δ/ATP1 (Figure 2A). On the other hand, far fewer macrophages were damaged when incubated with atp1Δ/Δ cells than when incubated with the control strains (Figure 2B). Evanescence of atp1Δ/Δ cells and insufficient killing of macrophages suggest that the ATP1 is required for C. albicans to evade macrophage attacks.

The capability of C. albicans to utilize non-glucose carbon sources and thereby counteract elevated reactive oxygen species (ROS) within macrophages are believed to be two elemental factors for the establishment of candidiasis (Frohner et al., 2009). We find that growth of the atp1Δ/Δ on spot assay is slightly less pronounced than those of WT and atp1Δ/ATP1 on YPD agar, but show no growth on YPG (glycerol), YPE (ethanol), YPO (oleic acid), or YPA (acetate). The last four media are YP-based media integrated with a non-glucose carbon source (Figure 3A). For oxidative stress response, the atp1Δ/Δ showed a reduced growth when exposed to H₂O₂, but no response to menadione (Figure 3B). These data indicate a vulnerable response of the atp1Δ/Δ to macrophage attacks may be due largely to an inability to consume non-glucose carbons as well as poor management of oxidative stress.

Endothelial cells of blood vessels provide a physical barrier for C. albicans to penetrate and invade the internal organs. We find that the atp1Δ/Δ has a significantly reduced capacity to destroy endothelial cells. The cell lysis rates were 77% less and 24% less at 15 and 24 h of infection (Figure 4), respectively, when compared to WT. These results indicate that ATP1 is required for endothelial cell penetration.

The contribution of ATP1 to adhesion events has also been tested in microtiter plates in vitro in order to understand whether less lysis of macrophages or endothelial cells may be due to the inefficiency of adhesion (Filler et al., 1991; Park et al., 2005; Wächtler et al., 2011). Adhesion assay reveals that atp1Δ/Δ fails to adhere to the bottom of the microtiter plate when compared with WT (Figure 5A). In qRT-PCR assay, the gene expression levels of SSA1 and ALS3 were markedly down-regulated in the atp1Δ/Δ when compared with WT (Figure 5B). These genes have long been known for their prominent roles in host tissue adhesion and invasion (Phan et al., 2007; Sun et al., 2010).

Filamentous growth enables C. albicans to penetrate host immune cells and tissue more efficiently after endocytosis (Phan et al., 2000; Dalle et al., 2010; Zhu and Filler, 2010). The requirement of ATP1 for filamentation has been tested under hyphal inducing conditions. In each hyphal inducing medium (10% FBS, Spider and Lee’s medium), the atp1Δ/Δ displays shorter or even no filamentous growth at 2 h growth point compared with WT, which has massive and long filamentous growth under microscopy (Figures 6A,B, Supplementary Figure S3). Similar results were also present on each agar medium. The atp1Δ/Δ displays a more severe defective phenotype in filamentous growth and forms small downy or even smooth colonies which lack the peripheral and invasive filaments, while WT produced large colonies with florid and invasive filaments on the edges (Figure 6C). The atp1Δ/ATP1 has intermediate effects in these matters between WT and atp1Δ/Δ. The filamentation defect in the atp1Δ/Δ is consistent with marked down-regulation of hyphal-specific genes such as HWP1, ALS3, ECE1, HGC1 in qRT-PCR test (Figure 6D). These results indicate that ATP1 is required for the filamentation process in C. albicans.

Aside from filament propensity, Candida biofilm also carries important clinical consequences because it will encase organism from antifungal therapy and withstand host immune defense (Douglas, 2003; Nett et al., 2010). Confocal scanning laser microscopy (CSLM) imaging was used to visualize biofilm formation in the atp1Δ/Δ and meanwhile, the biomass and cell metabolic activity within biofilm were measured. After 24 h culture, the atp1Δ/Δ forms a rudimentary biofilm with a thickness of less than 20 μm, while WT produces biofilm with thickness of over 50 μm, and atp1Δ/ATP1 produces biofilm of intermediate thicknesses (>20 μm and <50 μm) (Figure 7A). Under CSLM, the rudimentary atp1Δ/Δ biofilm consists mainly of yeast cells, with few filaments, whereas the biofilms of WT and atp1Δ/ATP1 have abundant filaments (Figure 7A). Consistently, the metabolic activity and total biomass of the atp1Δ/Δ are obviously reduced by XTT assay (Figure 7B) and by crystal violet assay, respectively when compared with WT and atp1Δ/ATP1 (Figure 7C). These results suggest that ATP1 deletion may be responsible for the failure of C. albicans to form biofilms.

The association of mitochondrial activity with C. albicans filamentation has been noted previously (Bambach et al., 2009; Grahl et al., 2015). Enzymatic assay of mitochondrial CV and intracellular ATP content were used to evaluate mitochondrial activity in the atp1Δ/Δ mutant. In the absence of the α subunit of Complex V, we find that the activity of Complex V is only 50% of the WT level (Figure 8A) and intracellular ATP drops to 1/3 of WT levels (Figure 8B). Meanwhile, the membrane potential of the atp1Δ/Δ is significantly decreased to nearly 50% of WT (Figure 8D). These data therefore confirm a critical role of ATP1 on mitochondrial function. However, in spite of these suppressed mitochondrial functions, we find that atp1Δ/Δ has a decreased ROS when compared with the WT (Figure 8C). Unlike other mitochondrial CI mutants with high ROS (Li et al., 2011; She et al., 2015), such reduced ROS levels in atp1Δ/Δ may be due to the fact that CV is not a direct electron acceptor. However, the contribution of ROS scavengers in atp1Δ/Δ needs to be clarified with further study.

The fate of invasive Candida cells also depends on how well they handle stress in the host environment. To evaluate these adaptive skills, we measured the susceptibility of the atp1Δ/Δ to various stresses. As shown in Figure 9A, atp1Δ/Δ is susceptible to plasma membrane stress (SDS), but the susceptibilities to cell wall (Calcofluor White, Congo Red), heat, calcium (CaCl₂) and osmotic stress (NaCl) are less pronounced in mutant when compared to its growth on YPD agar (Supplementary Figure S4). Interestingly, atp1Δ/Δ shows resistance to FLC, but not to ITR, KCZ and VRC (Figure 9B, Supplementary Table S2) even though the membrane of this mutant is more susceptible to SDS. The FLC resistance in the absence of ATP1 highlights the possible link between ATP1 function and FLC target or uptake process.

In C. albicans, the α subunit has been found at the surface of hyphae but not in yeast cells (Hernández et al., 2004). To better understand the pathogenic roles of ATP1 in this organism, we also measured the gene expression levels of ATP1 at the yeast and hyphal stages of WT. We find ATP1 expression levels at the hyphal form are similar to those of yeast form (Supplementary Figure S5). Likely, one possible explanation is that Atplp is regulated at the post-translational level rather than transcriptional level. Whether ATP1 is hyphal-specific requires further investigation.