Diverse mechanisms control amino acid‐dependent environmental alkalization by Candida albicans

Candida albicans has the capacity to neutralize acidic growth environments by releasing ammonia derived from the catabolism of amino acids. The molecular components underlying alkalization and its physiological significance remain poorly understood. Here, we present an integrative model with the cytosolic NAD+‐dependent glutamate dehydrogenase (Gdh2) as the principal ammonia‐generating component. We show that alkalization is dependent on the SPS‐sensor‐regulated transcription factor STP2 and the proline‐responsive activator Put3. These factors function in parallel to derepress GDH2 and the two proline catabolic enzymes PUT1 and PUT2. Consistently, a double mutant lacking STP2 and PUT3 exhibits a severe alkalization defect that nearly phenocopies that of a gdh2‐/‐ strain. Alkalization is dependent on mitochondrial activity and in wild‐type cells occurs as long as the conditions permit respiratory growth. Strikingly, Gdh2 levels decrease and cells transiently extrude glutamate as the environment becomes more alkaline. Together, these processes constitute a rudimentary regulatory system that counters and limits the negative effects associated with ammonia generation. These findings align with Gdh2 being dispensable for virulence, and based on a whole human blood virulence assay, the same is true for C. glabrata and C. auris. Using a transwell co‐culture system, we observed that the growth and proliferation of Lactobacillus crispatus, a common component of the acidic vaginal microenvironment and a potent antagonist of C. albicans, is unaffected by fungal‐induced alkalization. Consequently, although Candida spp. can alkalinize their growth environments, other fungal‐associated processes are more critical in promoting dysbiosis and virulent fungal growth.


| INTRODUC TI ON
Candida albicans is the primary cause of human mycoses presenting a spectrum of pathologies ranging from superficial lesions to life-threatening invasive infections.Recently, the World Health Organization (WHO) listed C. albicans under its "critical priority" group of fungal pathogens alongside the newly characterized multidrug-resistant Candida auris (Fisher & Denning, 2023;WHO, 2022).Despite its ill repute, C. albicans is normally a harmless commensal, thriving as a benign member of human microbiota in healthy individuals.Amino acids are among the most abundant nutrients in human hosts capable of supporting fungal growth.
However, the utilization of amino acids must be controlled due to the risk of accumulating ammonia (NH 3 ), a weak base that becomes toxic when in excess (Vylkova, 2017).The release of ammonia (NH 3 ) into the extracellular space alkalinizes the growth environment as it converts to ammonium (NH 4 + ) (Vylkova et al., 2011).The interest in studying this process in C. albicans is linked to findings that alkaline pH induces filamentous growth in vitro; morphological switching is a known virulence characteristic.Extracellular alkalization via ammonia release is not exclusive to C. albicans, other members of the pathogenic Candida species complex are also capable of alkalinizing their growth environments, albeit to varying degrees and dependent on the medium used (Kasper et al., 2014;Vylkova et al., 2011).For example, Candida glabrata, a WHO "high priority" fungal pathogen, can only alkalinize a medium containing amino acids as a sole source of carbon and nitrogen (Kasper et al., 2014); the addition of other carbon sources like glycerol abrogates alkalization (Vylkova et al., 2011).
We recently reported that the NAD + -dependent glutamate dehydrogenase (Gdh2; EC:1.4.1.2),which catalyzes the deamination of glutamate to α-ketoglutarate (Han et al., 2019;Silao et al., 2019Silao et al., , 2020)), is the key enzyme responsible for ammonia generation during growth on amino acids (Silao et al., 2020).Deletion of GDH2 completely abrogated both the ability of strains to generate ammonia and alkalinize amino acid-based media, even during extended periods of incubation (Silao et al., 2020).The alkalization defect of a gdh2 mutant surpassed the prominent alkalization deficiency observed in a strain lacking STP2 (Silao et al., 2020), which has been extensively referenced as an alkalization-deficient strain (Danhof et al., 2016;Hollomon et al., 2022;Miramon et al., 2020;Miramon & Lorenz, 2016;Todd et al., 2019;Vylkova et al., 2011;Vylkova & Lorenz, 2014).STP2 encodes a transcription factor required to derepress the expression of several amino acid permease genes and a subset of oligopeptide transporters (Martinez & Ljungdahl, 2005;Miramon & Lorenz, 2016).Stp2 is produced as a latent precursor that is proteolytically activated by the concerted action of the SPS (Ssy1-Ptr3-Ssy5) sensor in response to the presence of extracellular amino acids.The processed form of Stp2, lacking an Nterminal cytoplasmic retention domain, efficiently translocates to the nucleus where it binds to promoters of target genes (Martinez & Ljungdahl, 2005;Miramon & Lorenz, 2016;Silao et al., 2019).In addition to the alkalization deficiency, stp2 mutants were found to exhibit reduced virulence in mouse bone marrow-derived macrophages and/or in murine systemic infection models (Amorim-Vaz et al., 2015;Danhof & Lorenz, 2015;Vesely et al., 2017;Vylkova & Lorenz, 2014).These findings suggested a link between alkalization and virulence.Specifically regarding macrophages, ammonia extrusion was postulated to provide C. albicans (Danhof & Lorenz, 2015;Vylkova & Lorenz, 2014) and C. glabrata (Kasper et al., 2014) the means to increase the pH within the phagosomal compartment, enhancing their survival and their ability to evade macrophages.This notion is inconsistent with recent data.The inactivation of GDH2 does not affect C. albicans virulence in a murine systemic infection model, nor prevents morphological switching or survival of cells phagocytized by macrophages (Silao et al., 2020).Furthermore, we and others have shown using independent imaging-based methods (i.e., pHrodo Silao et al., 2020 andratiometric Westman et al., 2018) that phagosomes engulfing wildtype C. albicans cells remain acidic and that the observed phagosomal alkalization, if any, is a result of membrane distention as hyphae extend (Westman et al., 2018).
Thus, Stp2-regulated gene expression must be linked to other processes affecting virulence.
Mitochondrial function is key to effect alkalization, which occurs only when cells are grown under respiratory conditions, i.e., medium with low (<0.2%)glucose or alternative carbon sources, such as glycerol and/or lactate (Silao et al., 2020;Vylkova et al., 2011).
Here, we present an integrative model that accounts for how C. albicans generates ammonia and initiates the neutralization of acidic environments.Our work was based on testing the hypothesis that Gdh2 is the critical component, which enabled us to evaluate the contributions and connections of other known factors linked to amino acid uptake and mitochondrial function.Consistent with our finding in C. albicans, we show that although GDH2 is essential for amino acid dependent-alkalization in other pathogenic Candida species, including C. glabrata and C. auris, it remains dispensable for virulence as assessed by survival in whole blood culture.Furthermore, by examining the proliferation of Lactobacillus crispatus, a potent antagonist of C. albicans that normally thrives in the acidic vaginal microenvironment, we found that environmental alkalization does not directly limit the growth of this competing microorganism.

| Gdh2 level is sensitive to extracellular pH
Gdh2 levels decrease as the pH of the growth medium increases toward neutrality (Silao et al., 2020).Previously, we have shown that the rate of alkalization is highly dependent on the starting cell density; dense cultures (OD 600 ≥ 2) alkalinize the medium faster than diluted cultures (Silao et al., 2020).Based on the more rapid and stringent output, we have used high-density cultures to assess the alkalization potential of C. albicans strains.To directly examine the effect of pH on Gdh2 stability, strain CFG433 was pre-grown in YPD to fully repress Gdh2 expression and cells were subsequently shifted to a synthetic medium containing casamino acids (CAA) as the sole carbon/nitrogen/ energy source (YNB + CAA; pH = 4, unbuffered) for 2 h to induce the expression of Gdh2.Following induction, the cultures were added with buffers at defined pHs ranging from 4 to 8, spiked with cycloheximide (CHX) to arrest translation, and then the levels of Gdh2 were monitored after 1 h (Figure 1a).Relative to T = 0h immediately after adding CHX, there was no significant change in Gdh2 levels at pH = 4, indicating that Gdh2 is relatively stable at acidic pH (Figure 1b).Interestingly, a significant drop in the levels of Gdh2 was observed at a pH of 5.
Similarly, the levels of Gdh2 in the unbuffered control culture, where the pH increased to ~5 (see wells in Figure 1a), significantly decreased.
The strain CFG433 also co-expresses Put1-RFP and Put2-HA, which enabled us to simultaneously assess the levels of three proteins in a single blot.In contrast to Gdh2, both Put1 and Put2 were stable under all the pH conditions tested (Figure 1b), suggesting that the capacity to generate glutamate in the mitochondria from proline remained unchanged.Consequently, we posited that the reduced levels of cytosolic Gdh2 may reflect an active control mechanism that engages to limit the production of ammonia as the external pH increases.To directly test this notion, we examined the capacity of cells, grown at different starting pHs to alkalinize the culture media.Cells were pregrown in YNB + CAA medium (pH = 4) for 2 h prior to shifting them to YNB + CAA with an initial pH (pH i ) of 4 or 6.The pH of the media were assessed after 3 h (pH 3h ) (Figure 1c).Gdh2 levels were clearly lower in cells shifted to pH 6; however, detectable levels of Gdh2 remained (Figure 1b).Consistent with this, the change in pH (ΔpH = pH 3h -pH i ) was significantly lower in cells shifted to pH i of 6 than 4, ΔpH = 1.3 versus ΔpH = 2.8, respectively (Figure 1c).This is despite the fact that the resulting pH (pH 3h ) of the growth medium was significantly higher at pH i of 6 (pH 3h = 7.3) compared to that of cells shifted to pHi of 4 (pH 3h = 6.8; Figure 1c).These results are consistent with the existence of a pH-sensitive regulatory mechanism that constrains the alkalization potential of C. albicans.

| High glucose and mitochondrial activity regulate glutamate availability for Gdh2
Previous studies have shown that high (2%) levels of glucose in CAAbased media inhibit extracellular alkalization (Vylkova et al., 2011).
Consistently, we have found that cells grown in the presence of high glucose express significantly lower levels of Gdh2 (Figure 2a).
Consequently, cells grown in the presence of high glucose are expected to have a limited capacity to produce ammonia.We have previously shown that the components of the proline utilization (PUT) pathway, Put1 and Put2, generate glutamate in the mitochondria and are also repressed in cells grown with high glucose (Silao et al., 2019(Silao et al., , 2020)).Immunoblot analysis of Gdh2 and Put2 confirms this (Figure 2a), and consistently, the level of intracellular glutamate was significantly lower in cells grown in a medium containing 2% glucose (Figure 2b).
In addition to limiting the levels and deamination of glutamate, high glucose pleiotropically represses mitochondrial activity, which can affect the movement of metabolites between the cytosol and the mitochondria (Silao & Ljungdahl, 2021).Consistently, and supporting the notion that mitochondrial activity is essential to alkalization, we have shown that alkalization is blocked by the potent mitochondrial complex III inhibitor antimycin A (AntA) (Silao et al., 2020).Based on the assumption that the bulk of cytoplasmic glutamate originates from the mitochondria, we considered the possibility that acute inhibition of mitochondrial function would reduce the cytoplasmic glutamate levels, presumably due to an impaired capacity to transport glutamate out of the mitochondria.To test this, we analyzed the level of glutamate in the cytosol of the reporter strain (CFG441) grown in YNB + CASG (i.e., YNB + CAA medium supplemented with 38 mM ammonium sulfate and 1% glycerol) 30 min after inhibition by AntA (1 μg/mL) (Figure 2c, top panel).Having found that alkalization reduces Gdh2 levels (Figure 1b), we deliberately limited the analysis of expressed enzyme levels to 30 min after the addition of AntA.The cytosol fraction was extracted using the procedure reported by Ohsumi et al. (1988) using dilute CuCl 2 (0.2 mM) to gently The level of Gdh2 is sensitive to extracellular pH.(a) Scheme for the alkalization experiment.Gdh2 was induced for 2 h in YNB + CAA with bromocresol purple (BCP) as pH indicator and then spiked with concentrated buffer to 50 mM final concentration and cycloheximide (CHX) to arrest translation.Cultures were incubated for another 1 h followed by whole cell lysis and immunoblotting.(Right panel) Aliquots of spiked cultures after 1 h were placed in microplate wells to show the change in color corresponding to the indicated pH.(b) Immunoblot and quantification of Gdh2-GFP, Put1-RFP, and Put2-HA levels in cells grown at different culture pH.The signals from the target proteins were first normalized to α-tubulin and then compared relative to T = 0 h (set to 1).Data presented are from 4 biological replicates (mean with 95% CI; analyzed by one way ANOVA followed by Dunnett's posthoc test, ****p < 0.0001).(c) Alkalization rate is slower at elevated pH.Gdh2 was first induced to maximal expression in YNB + CAA and then shifted to YNB + CAA with initial pH (pHi) of 4 or 6 followed by incubation for 3 h.Resulting pH values were deducted from pHi to obtain ΔpH.Box and Whiskers plot with minimum and maximum values derived from 6 biological replicates analyzed by unpaired student t-test, **p < 0.01, ****p < 0.0001).
permeabilize the plasma membrane.We observed a significant ~3fold reduction in the level of cytosolic glutamate following treatment by AntA (Figure 2c).As expected, alkalization was arrested by AntA as inferred by the lack of color change of the pH indicator (Figure 2c).
Consistent with glutamate being limiting, the level of α-ketoglutarate was also significantly reduced (Figure 2c).Under these conditions, the levels of Gdh2, Put1, and Put2 remained unchanged (Figure 2d).
To further test the role of mitochondria in alkalization, we examined a strain lacking NUO1, which encodes a subunit of mitochondrial NADH:ubiquinone oxidoreductase (complex I).The nuo1Δ/Δ manifests the classical mitochondrial-deficient phenotypes, including reduced respiration (oxygen consumption), low mitochondrial membrane potential, and extremely poor growth on non-fermentable carbon sources (Huang et al., 2017;She et al., 2015).A dense culture of wildtype cells in YNB + CASG (OD 600 of 5) becomes alkaline within 3 h.In contrast, similar cultures of a nuo1Δ/Δ strain fail to alkalinize the medium even after 8 h, although the cultures eventually became alkaline after 24 h incubation (Figure 2e).Finally, a simple classical experiment, based on incubating dense cultures statically at room temperature, revealed that alkalization occurs only within the upper layer exposed to the atmosphere, clearly demonstrating that alkalization is linked to respiratory growth (Figure 2f).Together, these results support the scheme shown in Figure 2g in which mitochondria actively contribute to the activity of Gdh2 by supplying the cytoplasmic localized Gdh2 with glutamate produced via proline catabolism (PUT) and/or mitochondrial-localized aminotransferases (AT; Silao & Ljungdahl, 2021).

| Glutamate excretion enables cells to transiently dispose of excess nitrogen
Based on our findings that the levels of Gdh2, but not that of PUT enzymes, decrease as the extracellular pH rises above 5 (Figure 1b), we posited that the size of the intracellular glutamate pool would increase in parallel with increasing pH.We considered the possibility that excess glutamate could be metabolized to support the biosynthesis of other amino acids, e.g., glutamine (Silao & Ljungdahl, 2021), or alternatively, be excreted to the extracellular environment.Either of these processes could individually or together decrease the intracellular glutamate pools and reduce the possibility to generate ammonia.Amino acid excretion has been reported to occur in S. cerevisiae (Velasco et al., 2004); however, this has not been adequately addressed in C. albicans.Amino acid excretion has only been observed in studies analyzing spent growth medium from C. albicans biofilm cultures (Bottcher et al., 2022).
We measured the level of extracellular glutamate 3 and 5 h after growth in a glutamate-devoid minimal medium supplemented with amino acids that can be metabolized to glutamate (Silao & Ljungdahl, 2021); this YNB + PALAAG medium contains 0.5 g/L of Proline, Arginine, Leucine, Alanine, and Aspartate, and 1% Glycerol as primary carbon source.Extensive washing (4×) of the pre-cultured YPD grown cells was crucial to remove contaminating amino acids, including glutamate, prior to shifting cells to YNB + PALAAG.We chose 3 h as the earliest time point as this correlates with the dramatic reduction in Gdh2 levels in cells growing in the unbuffered YNB + CAA medium (Figure 1).Consistent with expectations, we detected substantial amounts of glutamate in the spent medium from wildtype cultures (Figure 3).Strikingly, 2 h later, i.e., 5 h post shift, we observed a ~20-fold reduction of extracellular glutamate, suggesting that glutamate is transported back into cells.The spent medium from cultures of the gdh2−/− mutant also contained glutamate, however, at significantly lower levels than wildtype (p < 0.0001; Figure 3).
In clear contrast to wildtype, the extracellular glutamate levels increased in the gdh2−/− strain, indicating continued excretion.
Together, these results indicate that glutamate excretion provides C. albicans cells the alternative to dispose of excess nitrogen without the negative effects of ammonia-dependent alkalization.
To critically assess the action of these factors, we inactivated STP2 or/and PUT3 in strain CFG441 (Figure S1A) and tested the capacity of the mutant strains to alkalinize YNB + CASG medium (Figure 4a, upper panel).The application of CRISPR/Cas9 enabled the simultaneous inactivation of all three STP2 alleles (Figure S1A; Martinez & Ljungdahl, 2005); from here onward, stp2−/−/− is designated as stp2-/-or stp2.As in previous reports, the stp2 strain showed a clear defect in alkalization even at high density (OD ≈ 5; Figure 4a, upper panel); however, media alkalization was observed upon prolonged incubation (24 h).The put3 strain showed a modest but significant delay in alkalization (Figure 4a, lower panel).Strikingly, the stp2 put3 double mutant showed a severe alkalization defect that clearly surpassed that of stp2 mutant and that almost phenocopied the clear alkalization defective phenotype of the gdh2 mutant (Figure 4a).
Relative to YNB + CASG, the stp2 or stp2 put3 mutants showed an even more prominent defect in YNB + CAA after 24 h (Figure S1B), likely due to the tight dependency of amino acid uptake on Stp2.We inactivated PUT3 in the previously reported stp2Δ/Δ strain (Vylkova & Lorenz, 2014) and obtained identical results.
These findings prompted us to directly examine Gdh2 levels.
Importantly, since extracellular alkalization negatively affects Gdh2 levels (Figure 1b), we limited the analysis to cultures grown only for in its regulation compared to Put3 (Figure 4b).Surprisingly, Stp2 also regulates Put1 and Put2 as indicated by their reduced levels in the stp2 strain.Interestingly, Put1 expression exhibited a tighter requirement for Stp2, whereas Put2 expression was more Put3dependent.Put3 levels were not altered in a stp2 mutant (Figure 4c).In all cases, the difference between stp2 and stp2 put3 were always significant (Figure 4b, red asterisks), suggesting that Stp2 and Put3 operate independently and in parallel to regulate the expression of GDH2, PUT1, and PUT2.

| The role of Gdh2 is conserved in other pathogenic Candida species
The capacity to neutralize the environment in the presence of amino acids was reported previously in other members of the pathogenic Candida species complex (Vylkova et al., 2011), including the multidrug resistant C. auris (Smith et al., 2022) and C. glabrata (Kasper et al., 2014).To test whether Gdh2-dependent alkalization is conserved in these species, we created GDH2 deletions in C. auris and C. glabrata by homology-directed recombination using an HA-tagging cassette (pFA6a-3HA-SAT-flipper) derived from the original SAT1flipper cassette (Reuss et al., 2004).We created reconstituted strains following the procedure described previously in C. albicans (Silao et al., 2020;Figure 5a).We compared the alkalization capacity of each strain grown in dense cultures (OD ≈ 5).Interestingly and consistent with a previous report (Vylkova et al., 2011), C. glabrata does

F I G U R E 3 Glutamate extrusion in
Candida albicans as a mechanism to regulate intracellular glutamate pool.(Upper panel) Glutamate levels in YNB + PALAAG medium grown with WT (SC5314) and gdh2 (CFG279) cells taken at the indicated timepoints.Data shown were obtained from seven (WT) and four (gdh2) biological replicates (mean with 95% CI; ****p < 0.0001 and **p < 0.01 by student t-test; red asterisks denote comparison between WT and gdh2 at 3 h timepoint).(Lower panel) Schematic diagram of glutamate extrusion in C. albicans as a function of time.Cytosolic glutamate (Glu cyto ) generated from aminotransferases (AT) reactions or proline catabolism (PUT) are converted by Gdh2 to ammonia (NH 3 ), which neutralizes the extracellular space.As the pH increases, Gdh2 level decreases which provide a surplus of Glu cyto that are then extruded to the medium and then reassimilated.
not readily alkalinize the extracellular medium (YNB + CASG) even after 72 h of prolonged growth (Figure 5b).However, when we grew the wildtype or reconstituted strain (gdh2Δ::GDH2) in YNB + CAA, where amino acids are used as sole carbon/nitrogen/energy source, alkalization was evident after ~5-6 h of growth being more pronounced after 8 h (Figure 5b).As expected, the Cggdh2 strain is completely unable to alkalinize the medium even after prolonged incubation (32 h).Unlike C. glabrata, the wildtype C. auris strain robustly neutralized the YNB + CASG medium, and as expected, the Caugdh2 strain did not, even after extended incubation beyond 24 h (Figure 5c).The results suggest that consistent with the role of Gdh2 in C. albicans (CaGdh2), the Gdh2 orthologues in C. auris (CauGdh2) and C. glabrata (CgGdh2) are also essential for amino acid-dependent alkalization.Analysis of the N-termini of CauGdh2 (B9j08_004192p) or CgGdh2 (CAGL0G05698g) by MitoFates (Fukasawa et al., 2015) revealed no mitochondrial pre-sequence (probability ~0.00) suggesting that these enzymes, as in C. albicans, localize to the cytoplasm.
Gdh2 is dispensable for the virulence of C. albicans (Silao et al., 2020).To assess the role of Gdh2 in virulence of C. auris and C. glabrata, we first assayed the survival of the wildtype and gdh2 strains using a human whole blood infection model (Kammer et al., 2020).Consistent with our findings in C. albicans, Gdh2 is dispensable for virulence of these two Candida species as there was no significant difference in the survival of gdh2 mutant in either species relative to their respective wildtypes (Figure 5d).We also tested the importance of GDH2 in the capacity of C. glabrata to specifically  1) was deleted by homology-directed recombination using a cassette amplified from a plasmid derivative of the SAT1-flipper cassette (2) and correct integration was verified by PCR using one of the primer pairs indicated in the scheme (3).The corresponding gels on the right used the primer pairs in blue.For reconstitution, the deleted region was replaced by WT allele bearing restriction sites (RE) (4) that were inserted in the primers (5, green font).Transformation was made on YNB + CAA plate and colonies pointed with arrows (4) were analyzed by PCR and restriction digest using either XhoI (C.glabrata) or ApaI (C.auris) (5).The reconstituted strains lost the nourseothricin resistance marker as indicated by the lack of growth on YPD with 200 μg/mL nourseothricin (+Nou).(b, c) Wildtype (+), gdh2 (−), and reconstituted (−::+) strains of C. glabrata (b) and C. auris (c) were grown in the indicated alkalization medium at OD ≈ 5 and photographed at the indicated timepoints.(d) GDH2 is dispensable for survival of C. glabrata and C. auris in a human blood infection model.Cells (~10 5 CFU) were added into a blood aliquot and survival was assessed 1 h after by plating.Data shown were obtained from six independent donors (mean with 95% CI; not significant by student t-test).(e) GDH2 is not required for immediate survival of C. glabrata following interaction with macrophages.Survivals of antibody-opsonized fungal cells were calculated by comparing the recovered CFU after 2 h of incubation with macrophages to the initial CFU.Data (n = 4) were analyzed by student t-test (ns = not significant).(f) GDH2 is not required for the cytotoxic effects of C. auris.Cytotoxicity was assessed using resazurin reduction assay on a confluent monolayer of A375 cells infected with wildtype and gdh2 strains for 24 h.Data (n = 4) between wildtype and gdh2 strains were analyzed by student t-test.evade fungal killing through modulation of phagosomal pH (Kasper et al., 2014).Like in our previous data in C. albicans, GDH2 is also dispensable for survival of C. glabrata during co-culture with macrophages (Figure 5e).As for C. auris, we also tested the capacity of wildtype and gdh2 cells to kill cells in an epithelial monolayer but we are unable to observe any difference between the two (Figure 5f), corroborating previous data that C. auris is not cytotoxic to intact skin epithelial cells or epidermis tissue (Brown et al., 2020).In addition, despite amino acid metabolism being implicated in antifungal tolerance (McCarthy & Walsh, 2018), the loss of GDH2 did not affect the fluconazole susceptibility of either C. auris or C. glabrata (Figure 5g).In summary, although Gdh2 is the key enzyme that endows Candida spp. the capacity to neutralize the environment when grown in amino acids, it remains dispensable for properties associated with virulence.

| Environmental alkalization does not antagonize growth of Lactobacillus
The intriguing finding that inactivation GDH2 does not affect C. albicans virulence (Kasper et al., 2014) motivated us to question the role of environmental pH modulation as a means to alter the composition of the host microbiome.In line with its opportunistic character, it is possible that C. albicans-dependent alkalization contributes to create imbalances in the host microflora antagonizing the growth of other competing, and potentially antagonistic microorganisms.For example, Lactobacillus species normally inhabit the acidic vaginal microenvironment in most healthy females, and these bacteria are thought to provide protection against C. albicans (Wang et al., 2017).Acute vulvovaginal candidiasis (VVC) occurs at least once in the lifetime of most women (75%) and represents a prime example of dysbiosis within a complex host niche (Sobel et al., 1998).Among the Lactobacillus species, L. crispatus is one of the predominant species in the vaginal tract (Antonio et al., 1999) and is thought to exert the most potent antimicrobial activity against C. albicans (Wang et al., 2017), making it an ideal representative of this genus.To experimentally address the possibility that C. albicans-dependent alkalization inhibits the proliferation of L. crispatus we used a transwell (0.4 μm) culture system schematically depicted in Figure 6a.The C. albicans strains were seeded into the transwell at a relatively high starting cell density (~1.6 × 10 7 CFU/transwell) to minimize growth-dependent effects.The high fungal cell density used was intended to simulate a worst-case scenario where there is a Candida overgrowth in VVC.
The high fungal density was also used to ensure that the alkalization process would take effect within a reasonable timeframe.
We used a modified alkalization medium (CNY; see Methods) containing ~5.6 mM (0.1%) glucose, which is comparable to the level observed in vaginal secretions (Ehrstrom et al., 2006).Under these conditions, CNY supports the growth of L. crispatus and C. albicans.
We were unable to use MRS medium, commonly used in culturing lactobacilli, due to its high level of glucose (2%), which inhibits alkalization in C. albicans, i.e., precisely the parameter under study.
The initial pH (pH = 4) is comparable to the acidic vaginal pH, the physiological consequence of lactobacilli-dependent fermentation.
Under aerobic conditions, L. crispatus failed to grow without C. albicans growing in the transwells; C. albicans apparently depletes oxygen in the media, a requisite for L. crispatus growth (Figure S3A).
As expected, and in contrast to the cph1Δ/Δ efg1Δ/Δ strain, the strain lacking GDH2 (cph1Δ/Δ efg1Δ/Δ gdh2-/-) was unable to alkalinize the media (Figure 6b; Silao et al., 2020).The tight dependency of alkalization on respiration was clear (Figure 6c); under aerobic and microaerophilic conditions, the medium became alkaline in a Gdh2dependent manner, whereas consistent with the requirement of mitochondrial function, alkalization did not occur under anaerobic conditions.Contrary to our expectations and despite obvious difference in extracellular pH, there were no significant differences in the number of L. crispatus cells recovered in the wells with transwells seeded with either GDH2 or gdh2 C. albicans strains (Figure 6d).Also unexpectedly, L. crispatus failed to counteract the alkalization exerted by GDH2 strain grown under aerobic and microaerophilic conditions, which is surprising given that lactobacilli are thought to play a prominent role in maintaining the acidic vaginal pH (Miller et al., 2016) and that L. crispatus inhibits growth of C. albicans (Wang et al., 2017).Taken together, these results clearly suggest that active pH modulation exerted by C. albicans may not effectively antagonize the growth of competing microorganism such as lactobacilli.

| DISCUSS ION
The interest in studying environmental alkalization as a consequence of amino acid metabolism in C. albicans is due its purported link facilitating virulent growth (reviewed in : Fernandes et al., 2017;Miramon & Lorenz, 2017;Vylkova, 2017).A precise understanding of the mechanisms underlying alkalization has remained elusive due to the lack of information regarding how ammonia is generated.C. albicans lacks the enzyme urease, which in many pathogenic microorganisms catalyzes the breakdown of urea to ammonia and carbon dioxide (Navarathna et al., 2010).
Consequently, urea amidolyase (Dur1,2), which catalyzes the analogous reaction in C. albicans, but in two steps, was initially considered to be a primary source of ammonia.However, the inactivation of DUR1,2 was found to have a modest effect on alkalization (Silao et al., 2019;Vylkova et al., 2011).We recently discovered that the cytoplasmic glutamate dehydrogenase (Gdh2) is the key enzyme catalyzing the release of ammonia and functions as the major driver of environmental alkalization (Silao et al., 2020).With this In parallel, the transcription factor Stp2 is proteolytically activated by the SPS sensor (Ssy1-Ptr3-Ssy5) in response to extracellular amino acids (4).The processed form of Stp2 efficiently targets the nucleus and binds UAS aa (5).The coordinated activities of these factors determine the transcriptional output of the key alkalization genes (6).Uptake of extracellular amino acids is facilitated by amino acid permeases (AAP) (7).Many amino acids are converted to glutamate in the cytosol (Glu cyto ) by transamination reactions (8) that require α-ketoglutarate (α-KG) as a substrate.Proline, enters the mitochondria via an unidentified (?) transporter ( 9) where it is converted by the proline catabolic pathway (Put1 and Put2) to glutamate (Glu mito ) (10).It is postulated that Glu mito is directly or indirectly (?) exported out of the mitochondria (11) and contributes to the Glu cyto pool.This multi-step path involves conversion of Glu mito to aspartate in the mitochondria (Asp mito ), predicted to be catalyzed by the mitochondrial aspartate aminotransferase (Aat1).Asp is exported to the cytosol where it gets converted to Glu cyto by cytosolic aspartate aminotransferase.Glu cyto is the substrate of Gdh2 forming α-KG and ammonia (NH 3 ) (12).Due to the cytosolic pH, maintained at around 6.5 (Rane et al., 2019), the bulk of ammonia rapidly converts to ammonium (NH 4

+
). Ammonium is a substrate of central nitrogen metabolism (Gdh3 and Gln1) (13).NH 3 , although a minor species, can efficiently diffuse across the cell membrane ( 14) where it forms NH 4 + and thereby effectively contributes to neutralizing the extracellular space (15).As the extracellular pH increases, Gdh2 level decrease, and a fraction of Glu cyto is extruded transiently (Glu ext ) by a still uncharacterized (?) exporter ( 17).Glu ext is reassimilated after transport into cells via AAP (18).
knowledge, we assessed the contribution of other factors that impinge on this process and have defined several control points.
Based on our findings, we have developed an integrative model that accounts for how the catabolism of amino acids leads to environmental alkalization (Figure 7).Despite its role in alkalization, GDH2 is dispensable for virulence of C. albicans in Drosophila and murine models of candidemia (Silao et al., 2020) and in human whole blood (our unpublished results).Our results challenge several commonly accepted dogmas regarding the role of alkalization in fungal virulence.
Based on the pK a of ammonium/ammonia (NH 4 + /NH 3 ) relationship in aqueous medium (Figure 7a), the ammonia generated by Gdh2 in the cytosol (pH ~ 6.5, Rane et al., 2019) is expected to rapidly form ammonium.Ammonium can be reassimilated by NADPH-dependent glutamate dehydrogenase (Gdh3; α-ketoglutarate to glutamate) and/or glutamine synthetase (Gln1; glutamate to glutamine; Silao & Ljungdahl, 2021).However, it is unlikely that these anabolic reactions can accommodate the amount of excess ammonium generated when cells are metabolizing amino acids as energy sources.Rather, our data are consistent with the notion that ammonia, although present in small amounts (~0.4%), passively diffuses across the cell membrane into the extracellular environment where it contributes to neutralizing the acidic pH.This occurs without needing a dedicated exporter (Antonenko et al., 1997;Cueto-Rojas et al., 2017;Labotka et al., 1995).We tested the role of Ato5, a putative ammonia transport protein, by creating an ato5−/− strain.Contrary to previous reports (Danhof & Lorenz, 2015;Vylkova et al., 2011), we were unable to confirm a major effect on alkalization (Figure S2A,B); thus, we excluded Ato5 from our model.
The passive diffusion of ammonia fully accounts for the observed density-dependent alkalization; higher cell densities result in more ammonia released and faster alkalization.This also explains why liquid cultures, rather than colonies growing on solid substrates, exhibit more rapid alkalization.The volatile ammonia (NH 3 ) formed by colonies diffuses in multiple directions limiting the amount that is directly captured by the solid growth medium (Figure 7b).The use of dense liquid cultures enables the assessment of the alkalization capacity of strains in a manner that is less dependent on growth, e.g., the nuo1Δ/Δ strain (Figure 2e).In fact, dense liquid cultures (OD ≈ 5) provides the most conclusive test to rigorously define the involvement of components thought to contribute to alkalization.To our knowledge, Gdh2 is the only component that clearly passes this test, i.e., strains lacking Gdh2 activity exhibit an absolute inaptitude to alkalinize amino acidbased media.
Although Gdh2 is a cytosolic component (Silao et al., 2020(Silao et al., , 2023)), our model accounts for the tight dependence of alkalization on mitochondrial function (Figure 7c).The substrate glutamate originates primarily from metabolic events in the mitochondria (Figure 2b,c).
Consistently, conditions that downregulate mitochondrial function, such as growth in high glucose (2%), limiting oxygen (Figures 2f and   5c), the lack of mitochondrial respiratory subunits (nuo1Δ/Δ), or inhibition by Antimycin A, all result in alkalization deficiency.Indeed, regardless of how amino acids are used by the cells, i.e., as sole carbon/nitrogen/energy source (YNB + CAA), as sole nitrogen source (YNB + CAA supplemented with 1% glycerol, 1% lactate, or <0.2% glucose as main carbon sources), or as a supplementary carbon/nitrogen/energy sources (YNB + CASG), alkalization readily occurs as long as the conditions permit respiratory growth, which enables the steady supply of mitochondrial glutamate to the cytosol to support the intrinsic catalytic activity of Gdh2.High glucose represses mitochondrial function and Put1 and Put2 expression, thus lowering intracellular glutamate (Figure 2a).The coupling of reduced intracellular glutamate with the downregulation of mitochondrial function explains why alkalization is virtually abolished in high glucose.
Strains lacking PUT1 and PUT2 exhibit delayed alkalization even in dense cultures (Silao et al., 2019(Silao et al., , 2020)), but alkalization is ultimately achieved presumably due to multiple transamination reactions that generate glutamate (Silao & Ljungdahl, 2021).We note that high glucose-mediated repression of the mitochondrial function can result in the generation of fermentation by-products and CO 2 , which can contribute to maintaining acidic pH (Morales et al., 2013;Silao et al., 2019).We predict (Figure 7c) that glutamate can be directly exported out of the mitochondria by a dedicated transporter, or alternatively, after being converted to aspartate (or other amino acids via transamination).These represent two distinct mechanisms, and dissecting them will require efforts that go beyond the scope of this work.
We traced the prominent alkalization defect of strains carrying stp2 to the inability to fully derepress GDH2, PUT1, and PUT2 (Figure 4).This is consistent with, but adds to our recent findings (Silao et al., 2023) that the proline-dependent expression of Put1, Put2, and Gdh2 is partially independent on Put3.The additive effect of combining stp2 and put3 mutations indicate that Stp2 and Put3 function in parallel to regulate the enzymes contributing to alkalization in C. albicans.This relationship explains some of the interesting stp2 phenotypes that are relevant to C. albicans pathogenesis, such as the defective filamentation in the phagosomes of macrophages (Danhof & Lorenz, 2015;Vylkova & Lorenz, 2014) and virulence in murine systemic infection model (Amorim-Vaz et al., 2015;Vylkova & Lorenz, 2014), which we have also shown to be dependent on proline catabolism (Silao et al., 2019(Silao et al., , 2023)).
The data presented here also help explain other findings related to the increase in proline uptake in stp2 mutant (Bottcher et al., 2022), which is likely a compensatory mechanism to generate more energy when intracellular amino acids become limiting.
The tight dependency of Gdh2 expression on Stp2 represented an apparent conundrum given that Gdh2 expression is extremely low in YPD (Silao et al., 2020) even though Stp2 is constitutively expressed and activated (Martinez & Ljungdahl, 2005;Miramon & Lorenz, 2016).However, this can be understood since GDH2 expression is apparently subject to glucose repression by Mig1 and Mig2; these factors have been shown to negatively regulate GDH2, PUT1, and PUT2 (Lagree et al., 2020).Shifting cells from YPD to YNB + CAA medium with low glucose relieves the repression, presumably enabling Stp2 and Put3 to activate expression of target genes by binding to putative UAS aa and UAS Put3 in their promoters.Also, the level of Gdh2 is tightly regulated; its levels rapidly decrease as pH increases, whereas the levels of Put1 and Put2 do not (Figure 1).Further work is required to experimentally define the precise nature of the promoter regions in these genes.For example, since Gdh2 is weakly expressed in YNB + CAA with 2% glucose (Figure 2b), glucose repression cannot fully explain the lack of Gdh2 expression in YPD grown cells (Silao et al., 2020).In addition, whether Stp2 regulates GDH2, PUT1, and PUT2 by directly binding to their indicated promoter regions (Figure 4d) or indirectly through regulation of an intermediate transcription factor needs to be further examined.
The observation that Gdh2 levels decrease as the extracellular pH increases and cells extrude glutamate suggests that cells do this to limit ammonia production and thereby actively respond by reducing the negative consequences of creating an alkaline growth environment.A similar strategy has been demonstrated in S. cerevisiae, which limits the production of ammonia by excreting amino acids, including glutamate via transporters that belong to the multidrug resistance transporter family postulated to function as H + antiporters (e.g., Aqr1) (Velasco et al., 2004) which reduces the alkalization potential, suggests that as other fungi, they prefer acidic growth conditions that provide a competitive advantage over many bacteria, is more favorable to facilitating nutrient uptake powered by the proton gradient, and minimizes the potential loss of nitrogen.Finally, maintaining an acidic growth environment likely contributes to keeping intracellular ammonium levels below toxic levels.Ammonia, not ammonium, readily diffuses across the plasma membrane (Antonenko et al., 1997;Cueto-Rojas et al., 2017;Labotka et al., 1995) in a direction toward a lower pH, effectively pulling the ammonium/ammonia equilibrium to drive excess ammonium out of cells.
As in C. albicans, Gdh2 is essential for the growth of C. glabrata and C. auris in media with amino acids as the sole carbon/nitrogen/ energy source (YNB + CAA).We exploited this phenotype during strain constructions (Figure 5a).Consistent with our finding that Gdh2 is dispensable for C. albicans virulence (Silao et al., 2020),  (Silao et al., 2020).This is entirely consistent to the V-ATPase pumping rate calculated by Westman et al., which is several orders of magnitude higher than the capacity of C. albicans to extrude ammonia (Westman et al., 2018).It is unlikely that C. glabrata can exert any significant change in phagosomal pH through ammonia production.The fact that Gdh2 is dispensable for virulence indicates that the metabolism of other carbon sources suffices for growth of gdh2 strains within the model host systems tested.The apparent difference in the virulence of put and gdh2 mutants can be attributed to the toxicity exerted by proline in strains that are unable to catabolize it, i.e., proline inhibits mitochondrial function in put1 and put2 mutants (Silao et al., 2023), a phenotype that is not observed in gdh2 mutants.et al., 2014).We initially considered applying a murine vaginal infection model to critically test prevailing notions regarding the importance of fungal-induced alkalization.However, contrary to the acidic human vaginal pH, the murine vaginal pH is near neutral (Miao et al., 2021).Consequently, the murine vaginal model is not well-suited to assess the significance of alkalization.This realization prompted us to develop a transwell assay (Figure 6) to specifically test whether the capacity of C. albicans to increase the extracellular pH negatively affects the growth of lactobacilli that comprise roughly 70% of the human vaginal tract microbiome.This contrasts to non-human primates, including rodents, where lactobacilli comprise <1% of the vaginal tract microbiome (Miller et al., 2016).The use of L. crispatus as a representative Lactobacillus species was deliberate given its potent antifungal character (Jang et al., 2019;Wang et al., 2017).The observation, that despite robust growth under aerobic and microaerophilic conditions, L. crispatus was unable to counteract the alkalization by C. albicans in a culture medium containing low but physiological levels of glucose (0.1%) requires comment.First, fungal-induced alkalization did not inhibit L. crispatus growth, making it unlikely that fungal-induced alkalization, per se, significantly impairs the growth of resident lactobacilli in the vaginal tract.Second, C. albicans supported the growth of L.
crispatus, a strict anaerobe, under aerobic conditions, presumably by sequestering oxygen.These data are consistent with reports of vaginal yeast colonization within Lactobacillus-dominant vaginal microbiomes (Eastment et al., 2021;van de Wijgert et al., 2014).Third, although this experimental set-up does not take into account alternative carbon sources, it is reasonable to expect that Lactobacillus spp.ferment other carbohydrate sources, e.g., glycogen, to maintain an acidic vaginal pH (Wang et al., 2017).Fourth, based on our finding that fungal-driven alkalization is strictly coupled to respiration (Figures 2f and 6c), the vaginal microenvironment is likely to be oxygen-limiting.Thus, even if there are transient increases in oxygen tension and pH, and also the influx of nutrients associated with increased sexual activity and menstruation (Alvarez et al., 2015;Ng et al., 2018;Wagner et al., 1984;Wagner & Levin, 1978), it is unlikely that C. albicans contributes directly to long-term alkalization of the vaginal tract.Our findings align with clinical observations that in contrast to vaginitis caused by bacteria or Trichomonas, vaginal candidiasis is not associated with an increase in vaginal pH (Lin et al., 2021).Rather it is likely that other fungal-associated processes are more critical in diminishing the efficacy of the mucosal barrier and of the vaginal microbiome to restrict fungal growth and limit access to epithelial cells.

| Organisms and culture
All key materials (organisms, chemicals, oligonucleotides, and software) used in this work are summarized in Table S1 (Key resource table).Yeast strains were routinely cultivated on YPD agar (1% yeast extract, 2% peptone, 2% glucose, and 2% Bacto agar) following recovery from glycerol stock stored in a −80°C freezer.Where needed, YPD is supplemented with nourseothricin (Nou) at the required concentrations (200-, 100-, and 25μg/mL) from filtered stock (200 mg/ mL).Overnight YPD broth cultures were prepared by picking single colonies from a YPD plate and grew in a shaking incubator (Infors HT Multitron Incubator Shaker) set at 30°C and 150-180 rpm.
This medium and other alkalization media were adjusted to pH = 4.0 using 1 M HCl before sterile-filtered (0.45 μm).The base medium was used without or with 2% glucose depending on the experiment.Where indicated, YNB + CAA is supplemented with 38 mM ammonium sulfate (5 g/L ≈ 38 mM) and 1% glycerol (YNB + CASG) to support the growth of some mutants, or used without BCP.For glutamate extrusion analysis, YNB was supplemented with 0.5 g/L each of Proline, Arginine, Leucine, Alanine, and Aspartate, and added with 1% Glycerol as the main carbon source, which is referred elsewhere in the text as YNB + PALAAG medium and used without BCP.For the alkalization experiment with Lactobacillus crispatus, the CNY medium was composed of YNB (0.085%), CAA (0.5%), proteose peptone no.
This tagging cassette has a very good recombination efficiency even if the homology flanking regions to the target genes are short (<100-bp).Amplified cassettes were purified and transformed into

| Alkalization assay
For routine assays, single colonies from YPD plates were picked and grown overnight (16-20 h) in YPD broth and then collected the following day by centrifugation at 4000 g for 3-5 min.

| Enzyme expression analysis
A 1 mL aliquot of the culture at 2 h timepoint prepared from a starting OD ≈ 2 was immediately mixed with 250 μL of cold 2 M NaOH with 7% β -mercaptoethanol (βME) to lyse the cells.After 15 min, an equal volume of 50% trichloroacetic acid (TCA) was added to precipitate the proteins, which was carried out overnight at 4°C.Protein pellets were harvested by high-speed centrifugation (17,000 g, 10 min, 4°C) and then the residual liquid in the tube completely removed by a quick high-speed spin.Protein precipitates were solubilized completely in a 50 μL 2X SDS sample buffer containing 5% βME and 167 mM of Tris Base (pH ≈ 11) and then boiled at 95°C for 5 min.

| Cycloheximide (CHX) treatment
Cells collected from YPD cultures were inoculated into 20 mL of YNB + CAA at OD ≈ 2 and then grown continuously with shaking at 37°C.After 2 h, 3 mL culture aliquots were added to separate tubes containing concentrated buffers (500 mM, pH = 4-8) diluted by cultures to 50 mM final concentration.After brief mixing, CHX was added to each tube at 200 μg/mL final concentration and then incubated another 1 h at 37°C before taking an aliquot for cell lysate preparation using the NaOH/TCA method (see preceding section).For the unbuffered control, an equal volume of ddH 2 O was added to the tube, and immediately after adding CHX, a 1 mL culture aliquot was mixed with NaOH/βME solution to prepare cell lysate, which was then used as reference (T = 0 h) for all comparisons.The selection of pH buffers used was based on their buffering capacities: sodium acetate (pH = 4, 5), MES (pH = 6), and HEPES (pH = 7, 8).

| Alpha-Ketoglutarate analysis
The cytosolic extracts obtained from antimycin-treated cells (see preceding section) were also analyzed for α-ketoglutarate using the Alpha Ketoglutarate (alpha KG) Assay Kit (Cat.# ab83431) following the manufacturer's instruction.Briefly, aliquots of the extracts were first deproteinated by TCA and then later neutralized with KOH.
Neutralized extracts were added with HEPES (to 50 mM final concentration) to ensure the extract pH was around 7.4 during analysis.

| Fluconazole susceptibility assay
Broth microdilution assay was performed in a microplate format

| Ex vivo human whole blood infection
Freshly extracted anonymized blood samples (4 mL) were purchased from Blodcentralen (Odenplan, Stockholm, Sweden), and all investigations were conducted according to the principles expressed in the Declaration of Helsinki.The well-established donation protocol for collecting human blood by the Karolinska University Laboratory for purposes other than medical treatment provided samples of peripheral blood collected from healthy volunteers with explicit consent for use in this study.According to legal requirements, the samples were labeled with donation numbers for traceability, but no names or other personal data was provided.Upon receipt in our laboratory, the samples are deidentified and can no longer be traced to an individual.This study does not require specific ethical approval.
For infection, fungal cells were first grown overnight in YPD and then refreshed the following day in a fresh medium.Exponentially growing cells were collected, washed twice in PBS, and then adjusted to 1 × 10 7 CFU/mL.Around 1 × 10 5 cells (10 μL) were added to 400 μL of whole blood (+EDTA), mixed briefly, and then incubated for 1 h in a shaking heat block (Eppendorf) set at 37°C and 400 rpm.
After incubation, tubes were vortexed vigorously, serially diluted in Plates were spun down (300 g for 3 min) to collect the Candida cells at the bottom of the well.The use of low MOI and the extra opsonization step were to ensure that all fungal cells would be phagocytized by macrophages.Co-cultures were performed for 2 h in a humidified chamber at 37°C with 5% CO 2 , after which the wells were treated with Triton X-100 to a final concentration of 0.1% to lyse the macrophages and release the fungal cells.Lysates were serially diluted and then plated onto YPD.
Survival of Candida cells following co-culture (% Survival) were determined by comparing the recovered CFU to the initial CFU.

| Epithelial cell cytotoxicity assay
Epithelial cells (A375; ATCC CRL-1619) growing to 80%-90% confluency in D10 in a T-25 flask at 37°C with 5% CO 2 were harvested by digestion with 0.5 mL of TrypLE for 5 min at 37°C.Detached cells were mixed with 6 mL of fresh D10 medium to deactivate the enzyme.Cells were then transferred into a 50 mL tube, harvested by centrifugation at 300 g for 5 min at 4°C, and then resuspended in 6 mL of fresh D10 medium.Viable cells were counted by trypan blue exclusion, diluted to 1 × 10 6 cells/mL in D10, seeded into 96 well plate at 1 × 10 5 cells/well, and then allowed to adhere overnight in the humidified incubator.The following day, C. auris wildtype (CFG552) and gdh2 (CFG586) cells from exponential phase YPD cultures were harvested, washed twice with PBS, and then diluted in D10 medium at 3 × 10 6 cells/mL.To start the infection, the growth medium of A375 was removed using a multi-channel pipette followed by addition of 100 μL aliquots of fungal cell suspensions containing 3 × 10 5 cells (MOI 3:1, C:M).Plates were centrifuged briefly at 300 g for 3 min to collect the fungal cells to the bottom of the wells before incubating it for 24 h in the humidified chamber.After incubation, wells were washed with gentle pipetting up and down with sterile PBS repeated five times (5×) to remove fungal cells.After the final wash, 120 μL of complete D10 medium containing resazurin dye (0.025 mg/mL) was added into each well.The plate was incubated in a humidified chamber for 2 h and then the fluorescence measured using a TECAN microplate reader at excitation and emission of 535 and 595 nm, respectively.To measure cytotoxicity, raw resazurin signals from infected and non-infected wells were first subtracted with signals from no cell control wells (with D10 medium only) and then the normalized signal from infected well (IW) were subtracted with signal from non-infected well (NIW) (i.e., cytotoxicity = IW-NIW).To ensure the exact processing across all treated and control wells, a multi-channel pipette was used for all steps.Experimental data were derived from two independently grown A375 cells each infected with at least two independent colonies of the indicated C. auris strains, all performed in duplicates.

| Transwell assay
Lactobacillus crispatus (LC100) vaginal isolate from a healthy female was recovered on MRS agar medium from glycerol stocks and incubated anaerobically for 48 h at 37°C in anaerobic jar (GasPak™) with an Anaerogen sachet (Oxoid).For broth culture, isolated colonies were grown anaerobically in MRS broth for 48 h at 37°C.Cells were then collected by centrifugation 6000 g and then washed three times to remove the media.Cells were inoculated into CNY medium at ~3 × 10 7 CFU/mL, and then a 500-μL aliquot (~1.5 × 10 7 CFU/well) of the suspension was gently dispensed in the well of a 24-well plate 2 h post-YPD shift.The results clearly show that Stp2 and Put3 coregulate Gdh2 expression, with Stp2 having a more dominant role F I G U R E 2 High glucose and mitochondrial activity regulate glutamate availability for Gdh2.(a) Glucose availability inversely affects cytoplasmic glutamate levels.Immunoblot analysis of Gdh2-GFP and Put2-HA levels in lysates prepared from cells (CFG404) grown in YNB + CAA without or with 2% glucose.(b) Intracellular glutamate levels in lysates prepared from (a).Data shown were obtained from 4 to 5 biological replicates (mean with 95% CI; ***p < 0.0001 by student t-test).(c) Mitochondrial function is required to maintain cytoplasmic glutamate levels.Cells were grown for 2 h in YNB + CAA medium to induce Gdh2 expression and then spiked with Antimycin A (AntA) or vehicle control (ethanol, −).Tubes were photographed at the indicated timepoints (Right) or sampled after 30 min to determine the levels of cytosolic glutamate and α-ketoglutarate (Left); results are presented as percentage (%) of vehicle control derived from 5 biological replicates (mean with 95% CI; ****p < 0.0001 by student t-test).(d) Immunoblot analysis of cells grown and sampled as in (c).Enzymes levels are not altered after exposing the cells to AntA.Enzymes were simultaneously detected in the same lysate and quantified using Tdh3 as loading control.Data is presented as mean with 95% CI (n = 5).(e) Strain with mitochondrial dysfunction exhibits striking alkalization defect.(Left) Wildtype NUO1+/+ (SN250) and nuo1Δ/Δ (CB342) strains were grown in YNB + CASG at OD ≈ 5 and photographed at the indicated timepoints.(Right) pH values of cultures after 24 h (mean with 95% CI; ****p < 0.0001 by student t-test).(f) CFG441 cells were inoculated into 20 mL of YNB + CAA at OD ≈ 5 and then kept upright at room temperature for 2 days.Only the top layer with the highest oxygen concentration resulted to alkalization.(g) Schematic diagram of AntA or high glucose inhibits alkalization.AntA or high glucose inhibits mitochondrial function and prevent the export of Glu mito generated by either proline catabolism (PUT) or aminotransferases (AT) to the cytosol.

F
Environmental alkalization does not antagonize growth of Lactobacillus crispatus.(a) Schematic diagram showing the in vitro set-up to investigate the effect of pH modulation in growth of L. crispatus.(b) cph1Δ/Δ efg1Δ/Δ GDH2+/+ (CASJ041) and cph1Δ/Δ efg1Δ/Δ gdh2−/− (CFG352) were grown in CNY alkalization medium at OD ≈ 5 and photographed at the indicated timepoints.(c) Alkalization is dependent on oxygen tension.Photographs of the wells after 48 h of growth under aerobic, microaerophilic, and anaerobic growth conditions.(d) Fungal-driven pH modulation does not antagonize L. crispatus.Viable cell counts of L. crispatus recovered in the medium indicated in (c); Results (n = 4) between GDH2+/+ and gdh2−/− in each oxygen tension condition were not significant by student t-test.F I G U R E 7 Integrative model of amino acid-dependent alkalization of the extracellular growth environment by Candida albicans.A) A plot showing the relative concentrations of ammonia (NH 3 ) and ammonium (NH 4 + ) in aqueous solution based on pH set at 37°C.Values were calculated using pKa = 8.89 (Erickson, 1985).The ratio of NH 4 + to NH 3 in this mixture is highly pH-dependent.(b) Schematic diagram showing the difference in the ammonia release in liquid and solid media.In comparison to solid media, volatile ammonia is more efficiently captured by the acidic liquid culture to form ammonium resulting in a more rapid alkalization of the media.(c) Regulatory control of alkalization gene expression and glutamate-dependent metabolism determines the alkalization potential of C. albicans.When cells are shifted from high glucose medium (e.g., YPD) to alkalization medium (e.g., YNB + CAA), C. albicans cells are relieved of glucose repression (1) transcriptionally mediated by Mig1/2 repressors binding upstream repressing sequences (URS glu ) (2).Cytosolic proline (Pro cyto ) enters the nucleus (Pro nuc ) where it binds and activates the transcription factor Put3 at upstream activating sequence (UAS pro ) (3).
Alkalization has been speculated to generate conditions that favor disease-associated vaginal microbes, providing opportunities for Candida overgrowth(Vazquez-Munoz & Dongari- Bagtzoglou, 2021) and access to host epithelial cells (van de Wijgert specificity tyrosine phosphorylation-regulated kinase (DYRK) family (MacAlpine et al., 2021).Consistent with current understanding, we have observed that spent media from L. crispatus inhibits C. albicans growth (data not shown).However, in the context of our transwell co-culture experiments, although L. crispatus grew in the microaerophilic conditions, and presumably secrete normally, C. albicans retained the capacity to alkalinize the media, indicating that mitochondrial functions remain intact.The failure to counteract the alkalization by C. albicans indicates that the inhibitory effects of L.crispatus toward C. albicans does not appear to involve mitochondrial repression in a manner analogous to the effect of phenazine produced by Pseudomonas aeroginosa(Morales et al., 2013).To conclude, although alkalization is not directly required for virulence of C. albicans and other Candida spp., the results reported here warrant the reassessment of virulence processes reportedly linked to or associated with pH modulation by fungal pathogens.Our future efforts will examine how Candida cells adjust their metabolism in the context of infection, which we posit involves transcriptional and/or translational arrest.
dium and NYCIII medium (without glucose) and then supplemented with glucose and BCP at the indicated concentrations.Other specific media modifications are indicated elsewhere in the text.Stock solutions of different carbon and nitrogen sources used are as follows: selected on YPD + 200 μg/mL Nou, and further clonal purification was made on YPD + 100 μg/mL Nou.Transformants were purified and verified several by colony PCR using the appropriate primers to amplify the mutated gene and then digested with XhoI to identify the knockouts.For pV1524-derived plasmids, cassette excision was done by growing purified colonies in liquid YPM and then plating on YPD + 25 μg/mL Nou.Nourseothricin-sensitive (Nou S ) popouts were verified by streaking on YPD + 100 μg/mL Nou.For the construction of stp2−/− put3−/− double knockout strain, the Nou s stp2−/− strains were co-transformed with the digested pFS084 and RT, and transformants were selected accordingly.In most cases, single preparations of purified digested cassettes and RT sufficed for 10-15 independent transformations and were kept at −20°C until used.4.3 | Genetic construction of gdh2 and reconstituted C. glabrata and C. auris strainsSince C. glabrata and C. auris are haploids, a homology-directed recombination gene deletion strategy was used to delete GDH2 in these strains.Briefly, a knockout cassette was amplified from an HAtagging plasmid (pFS069 = pFA6a-3HA-SAT flipper;Silao et al., 2019) using primer pairs p21/p22 for C. glabrata GDH2 (CAGL0G05698g)

C
. glabrata(ATCC 2001)  and C. auris (CFG552) wildtype strains and Nou R transformants selected on YPD + 200 μg/mL Nou.Purified colonies were checked for correct integration, and verified knockouts were stored at −80°C as glycerol stock without removing the NAT R marker, as it would be used as a marker for reconstitution.The following primers were used to verify junctions for correct knockout cassette integration: Cg-gdh2Δ (p27/p24; p23/p34) and Cau-gdh2Δ (p27/p30; p33/p34).For genetic reconstitution of knockout strains, a region of the GDH2 gene was amplified by PCR from the respective wildtype strains using primers p25/p26 for C. glabrata GDH2 and p31/p32 for C. auris GDH2.The amplicons were purified and then transformed into the respective knockout strains.Electroporated cells were recovered in YNB + CAA medium for 6-8 h and then plated on YNB + CAA agar.Plates were incubated for 2-3 days at 30°C until colonies appeared.These colonies were purified on YPD, and then 3-5 independent colonies per transformation were tested for correct integration by PCR (C.glabrata, p23/p24; C. auris, p33/ p30) and restriction digest using either XhoI (C.glabrata) or ApaI (C.auris).The verified reconstituted strains were tested for growth on YPD agar with 200 μg/mL nourseothricin to determine the loss of the Nou R marker, indicating replacement by the reconstituted allele.
according to the protocol by EUCAST method (Subcommittee on Antifungal Susceptibility Testing (AFST) of the ESCMID, 2008) with minor modifications that are limited to dissolving the fluconazole powder in 2% DMSO solution in ddH 2 O and cell density adjustment by OD (1 OD ≈ 3 × 10 7 CFU/mL).The minimum inhibitory concentration (MIC) is the lowest drug concentration that gives rise to ≥50% growth inhibition of the drug-free control.
ddH 2 O (to lyse human cells), and plated on YPD agar.Plates were incubated at 30°C for 2 days and colonies (CFU) counted.% Survival of fungal cells was calculated by comparing the CFU after the 1 h incubation to the CFU of the inoculum.4.11 | Macrophage co-culture assayPrimary bone marrow-derived macrophages (BMDM) were used to compare the survival of C. glabrata wildtype and gdh2 mutant in macrophages.Briefly, bone marrows collected from mouse femurs of C57BL/6 wildtype mice (7-to 9-week-old) were differentiated in differentiation medium (DM) composed of complete DMEM medium (referred to as D10) supplemented with 20% L929 conditioned medium (LCM) in a humidified chamber set at 37°C with 5% CO 2 .D10 is composed of DMEM medium (high glucose) supplemented with 10% fetal bovine serum (FBS), and 100 U/mL penicillin and 100 μg/mL streptomycin.Differentiation was carried out initially in 10 mL of DM for 3 days before boosting the cells with another dose (10 mL) of DM until macrophages are fully differentiated (about 7-10 days after plating).Macrophages were collected by PBS-EDTA treatment followed by gentle scraping.Macrophages resuspended in D10 were seeded in a 48-well dish at 3.5 × 10 5 cells/well and allowed to adhere overnight.The following day, exponentially growing C. glabrata wildtype (ATCC 2001) and gdh2 (CFG670) strains in YPD cultures were harvested, washed 2X with PBS, and then adjusted to OD ≈ 1 (~3 × 10 7 CFU/well).A 200 μL aliquot of the cell suspension was removed and then incubated with 1 μL of α-Candida albicans antibody (1:200 dilution) for 30 min at 37°C in a tabletop shaker to opsonize the cells.Opsonized cells were first diluted in pre-warmed D10 to ~2.35 × 10 5 CFU/mL and then a 500 μL aliquot was used to replace the macrophage growth medium to obtain an MOI of 1:3 (C:M).
containing a sterile insert (Millicell, 0.4 μm PCF, Cat.# PIHP01250; Merck Millipore, Ltd.) placed at the center of the well.For the processing of fungal cells, cph1Δ/Δ efg1Δ/Δ (CASJ041) or cph1Δ/Δ efg1Δ/Δ gdh2−/− (CFG352) cells were harvested from overnight YPD broth, washed twice in ddH 2 O, and then resuspended in CNY medium at OD ≈ 5.A 200 μL aliquot (OD ≈ 1; ~1.6 × 10 7 CFU/well) was carefully pipetted into the center of the insert.Plates were then incubated in aerobic, microaerophilic, and anaerobic conditions for 48 h at 37°C . We note that the C. albicans We found that the extrusion of glutamate by C. albicans is transient as the excreted glutamate is eventually reassimilated.It is unclear whether other amino acids are also excreted in addition to glutamate, but it would not be surprising if there are others.At this point, defining the precise transport processes will involve more experimental work.The fact that C. albicans extrudes glutamate,