Candida albicans Hyphal Expansion Causes Phagosomal Membrane Damage and Luminal Alkalinization

C. albicans is the most common cause of nosocomial fungal infection, and over 3 million people acquire life-threatening invasive fungal infections every year. Even if antifungal drugs exist, almost half of these patients will die. Despite this, fungi remain underestimated as pathogens. Our study uses quantitative biophysical approaches to demonstrate that yeast-to-hypha transition occurs within the nutrient-deprived, acidic phagosome and that alkalinization is a consequence, as opposed to the cause, of hyphal growth.

C andida albicans is a commensal yeast of humans but is frequently the source of mucosal infections and can, in severe cases, cause life-threatening systemic infections (1). It colonizes the epithelial surfaces of 30 to 70% of healthy individuals, and superficial infections are usually transient (2). Due to an aging population, an increased use of antibiotics, and immunocompromising drug treatments, nosocomial C. albicans infections have increased dramatically over the last few decades (3). Unlike many other pathogenic microbes, C. albicans is polymorphic and grows as budding yeast, pseudohyphae, or true filamentous hyphae. The yeast-to-hypha transition is initiated as a response to various environmental stimuli. These include increased pH or temperature, nutrient deprivation, contact with immune cells, and exposure to serum proteins (4,5). C. albicans yeast cells are associated with commensal growth (but also with dissemi-of change of pH (ΔpH/Δt) induced by addition of the potent and specific inhibitor concanamycin A (CCA) to the murine macrophage cell line RAW 264.7 (here referred to as RAW cells). Such measurements are based on the notion that, in the steady state, the rate of pumping by the V-ATPases is identical to the rate of proton (equivalent) leakage (30). To measure the phagosomal pH, C. albicans yeast cells were allowed to bind fluorescein isothiocyanate (FITC)-labeled concanavalin A and a C. albicans-specific IgG prior to phagocytosis. Such labeled and opsonized yeast cells were centrifuged onto macrophages grown on glass coverslips to initiate phagocytosis synchronously, and at the desired times, phagosomal pH was measured by dual-wavelength ratiometric fluorescence imaging as detailed in Materials and Methods.
As illustrated in Fig. 1B and when added 1 h after phagocytosis-when the phagosomes are acidic, averaging a pH of 5.03 Ϯ 0.14 (mean Ϯ standard error of the mean [SEM] from 32 determinations in 4 experiments)-CCA elicited a rapid alkalinization at an average rate of 0.54 pH/min. The amount of protons pumped per unit time can be calculated by multiplying this rate by the phagosomal buffering capacity. The latter was measured by pulsing the cells with known amounts of membrane-permeant weak electrolytes (see Materials and Methods and reference 31). In 4 independent experiments, the phagosomal buffering power averaged 91.2 Ϯ 3.3 mmol/liter/pH. The rate of proton pumping at the steady state was therefore calculated to be 49.2 Ϯ 15.5 mmol/ liter/min. We proceeded to compare the rate of pumping with the reported rate of NH 3 production by C. albicans. Vylkova and Lorenz (15) reported a production of Ϸ35 ppm over 24 hours, which is equivalent to 1.28 ϫ 10 Ϫ8 nmol/yeast/min. Similar rates have been reported by others (13,14,(16)(17)(18). This rate is 2 orders of magnitude lower than the rate of proton pumping at the steady state (Fig. 1C). It should be noted that the activity of the V-ATPase decreases markedly as the pH becomes more acidic (32), so that the disparity between the rates of pumping and NH 3 production would become even greater at more alkaline pH. At such pH values, the rate of leakage of proton equivalents by other (endogenous) pathways decreases, which would further offset the rates of acidification and alkalinization. We conclude that NH 3 production by C. albicans is unlikely to account for the reported phagosomal alkalinization.
C. albicans-containing phagosomes are permeable to NH 3 . Not only is the rate of NH 3 production insufficient to overcome the rate of proton pumping, but sustained alkalinization would require retention of NH 3 within the phagosome. Because, as illustrated diagrammatically in Fig. 2A, the protonation of NH 3 is a rapidly reversible reaction, a fraction of the NH 3 /NH 4 ϩ will always exist inside phagosomes as the unprotonated weak base. Because the extracellular space and the cytoplasm are nominally free of NH 3 /NH 4 ϩ , the prevailing outward gradient would promote ongoing loss of NH 3 from the phagosome, provided that the phagosomal membrane is permeable to the weak base. While most mammalian membranes are permeable to NH 3 , the permeability of the C. albicans-containing phagosome has not been ascertained. We assessed permeability to NH 3 by measuring the phagosomal pH while pulsing the medium with extracellular NH 3 /NH 4 ϩ , as illustrated in Fig. 2B. Addition of NH 3 /NH 4 ϩ to the medium caused an immediate and pronounced phagosomal alkalinization ( Fig. 2C to E), ostensibly due to permeation by NH 3 and protonation to NH 4 ϩ . Of note, instantaneous removal of extracellular NH 3 /NH 4 ϩ resulted in rapid restoration of the acidic pH, implying very rapid conversion of NH 4 ϩ to NH 3 and efflux of the latter (Fig. 2C to E). Such experiments enabled us to estimate the rate at which NH 3 permeates the membrane of C. albicans-containing phagosomes. Considering the ΔpH elicited by NH 3 /NH 4 ϩ (from 4.85 Ϯ 0.22 to 6.0 Ϯ 0.03, n ϭ 4) (Fig. 2E) after 1 s-the fastest time measurable using our experimental setup-and the buffering power determined earlier, we estimated that NH 3 can enter/exit phagosomes at a rate of 2.42 ϫ 10 Ϫ4 nmol/phagosome/min. As illustrated graphically in Fig. 2F, this rate is several orders of magnitude greater than the reported rate of NH 3 production by C. albicans. These additional data reinforce our conclusion that NH 3 production by C. albicans is unlikely to account for the reported phagosomal alkalinization.
C. albicans hyphal expansion drives phagosomal alkalinization. We next analyzed the time course of the pH changes undergone by the C. albicans-containing phagosomes (Fig. 3A). While NH 3 is presumably produced continuously by the yeast, the phagosome initially becomes acidic and remains so for nearly 2 hours (Fig. 3A, open boxes). These observations not only argue once again against a role for NH 3 production in the pH changes but suggest that an alternative, time-dependent mechanism is involved. Significant alkalinization was detectable only after Ϸ3 hours. Notably, marked hyphal growth was clearly apparent at this stage: on average, the hyphae reached 24.7 Ϯ 2 m in length (Fig. 3A, solid black boxes). We therefore hypothesized that hyphae can form inside acidic phagosomes and that phagosomal alkalinization was a consequence of hyphal growth.
To test this hypothesis, we investigated whether the yeast-locked cph1Δ/efg1Δ C. albicans mutant (33), which is unable to form hyphae, alters the phagosomal pH as does the wild type. Remarkably, over a period of 4 hours yeast-locked C. albicans cells failed to dissipate the phagosomal acidification (Fig. 3B), despite the fact that this strain generates NH 3 at rates comparable to the wild-type C. albicans (see Fig. S2A in the supplemental material). These data are consistent with the notion that the pH change is a consequence of hyphal growth.
During hypha formation, the expression of several genes is activated (4,34). One such gene is ECE1, the product of which is processed into the pore-forming toxin candidalysin (28). Because of the association between hyphal growth and phagosomal pH changes, we tested whether candidalysin contributes to the alkalinization. The pores formed by the toxin could conceivably cause leakage of proton (equivalents) through the phagosomal membrane. As illustrated in Fig. 3C, a mutant lacking Ece1 (ece1Δ), the precursor required for candidalysin generation, caused phagosomal alkalinization at a rate that was only slightly lower than that induced by the wild type. The difference between the two strains was small but statistically significant (Fig. S2B). Of Phagosomal Alkalinization by C. albicans note, the ece1Δ mutant formed hyphae that grew inside the phagosome at rates that were similar to the wild type (Fig. S2C).
These results indicate that, while candidalysin aids in alkalinizing the phagosome, its contribution is comparatively small and other factors are involved. To validate this conclusion, we constructed a strain of yeast-locked C. albicans that expresses ECE1 at levels comparable to those recorded during hyphal growth of wild-type C. albicans and that produced similar quantities of the candidalysin peptide (Data Set S1 and Fig. S1). Despite the continuous production of candidalysin, this strain had negligible effects on phagosomal pH over a 4-h period (Fig. 3D).
The results described above suggest that the phagosomal pH change is a direct consequence of hyphal growth. Remarkably, the opposite has previously been proposed: namely, that alkalinization precedes and is required for hyphal growth (15,35). The latter concept derives from analyses of the pH dependence of C. albicans hyphal growth, which show higher growth rates at more alkaline pH (35)(36)(37)(38). Indeed, it has been shown that neutral pH is required for full virulence of C. albicans, as many effectors are activated by proteolytic cleavage at neutral pH (39). However, while we could readily replicate the faster growth of C. albicans at more alkaline pH values, we also noted that hyphal growth does occur in vitro at the pH normally attained by phagosomes, i.e., pH 4.5 to 5.0 ( Fig. 1 and 3 and Fig. S2).
C. albicans hyphal expansion drives phagosomal membrane rupture. As the hypha expands over time within the phagosome, increasing mechanical tension must be applied on the phagosomal membrane. This could conceivably alter the permeability of the membrane to proton (equivalents) and potentially even cause its rupture.
We assessed the phagosomal membrane integrity using sulforhodamine B (SRB), a fluorescent dye that is nominally impermeant across biological membranes. SRB was delivered to phagosomes via fusion with lysosomes, which had been previously loaded with the dye using a pulse-chase protocol (see Materials and Methods for details). As illustrated in Fig. 4A and Movie S1, SRB contained within phagolysosomes can be initially observed lining the yeast and hyphae. Over time, however, the SRB contained within phagosomes formed by wild-type C. albicans was lost progressively ( Fig. 4A and B). In line with phagosomal alkalinization, the number of SRB-positive phagosomes decreased over a period of 4 hours (Fig. 4B). SRB was lost at a similar rate from the ece1Δ mutant, which grows at a comparable rate as the wild-type C. albicans strain (Fig. 4C). Importantly, SRB was retained during the same period by the yeast-locked mutants, even when they were engineered to produce Ece1 ( Fig. 4D and E). The latter data suggest that SRB leakage is a consequence of hyphal growth and not permeation of the dye via candidalysin. Our data are most consistent with a growth-induced change in permeability, likely manifested as rupture of the membrane. Note that, as documented below, in some instances rupture may have been transient, followed by membrane repair.

FIG 4
Growth-induced phagosomal leakage demonstrated using sulforhodamine B (SRB). To load endosomes and lysosomes with SRB (red), cells were bathed in 150 g/ml of the dye 60 min prior to phagocytosis of wild-type C. albicans yeast. (A) During phagosome maturation, endosomes and lysosomes fuse with the phagosome, delivering SRB to its lumen. Cells were imaged over 4 hours following phagocytosis by spinning disk confocal microscopy, and representative images are depicted. C. albicans is shown in white (left images for each time); the yeasts were omitted from the right panels to more clearly visualize the SRB outline. The borders of the macrophages are outlined using a dotted white line. Bar, 5 m. (B to E) Quantitation of the fraction of phagosomes retaining SRB as a function of time after phagocytosis. Results obtained using wild-type C. albicans (B) or Ece1-null (ece1Δ) (C), yeast-locked (cph1⌬/efg1⌬) (D), or Ece1-expressing yeast-locked (cph1⌬/efg1⌬ ϩ pENO1-ECE1) (E) strains are illustrated. Data are means Ϯ SEMs from at least 100 determinations in 3 independent experiments for each type. Significance was calculated using one-way ANOVA, with Tukey's test. *, P Յ 0.05; **, P Յ 0.01; ***, P Յ 0.001; n.s., not significantly different.
Phagosomal Alkalinization by C. albicans ® Phagosomal expansion using GPN causes alkalinization and membrane rupture similar to hyphal expansion. To verify that mechanical tension (like that induced by hyphal growth) suffices to produce phagosomal alkalinization and membrane rupture, we treated phagosomes containing the yeast-locked C. albicans cph1Δ/efg1Δ mutant with Gly-Phe-␤-naphthylamide (GPN). GPN is a membrane-permeant dipeptide and a substrate for cathepsin C. Phagosomal cathepsin C can cleave GPN, generating the membrane-impermeant Phe-␤-naphthylamide. Accumulation of Phe-␤-naphthylamide drives osmotically obliged water into the phagosome, which consequently expands, exerting hydrostatic pressure that distends the membrane in a manner akin to that occurring during hyphal growth. As illustrated in Fig. 5 (top, left panel) and Movie S2, following addition of GPN, phagosomes containing the yeastlocked C. albicans expanded rapidly and alkalinized within 20 min, while untreated phagosomes remained acidic (Fig. 5, top, right panel, and Movie S2). Even faster onset of alkalinization was observed using higher concentrations of GPN (data not shown). GPN also induced leakage of phagosomal SRB within minutes (Fig. 5, bottom, left panel, and Movie S3), while untreated phagosomes remained SRB positive (Fig. 5, bottom, right panel, and Movie S3). We concluded that mechanical tension is sufficient to cause phagosome alkalinization and membrane rupture.
Phagosomes undergo transient changes in pH before irreversible rupture. Biological membranes have effective means of repairing occasional tears, thereby maintaining homeostasis (40)(41)(42)(43)(44). It was therefore conceivable that phagosomes containing growing hyphae would attempt to maintain their integrity, at least until the repair mechanisms were exhausted. To investigate this possibility, macrophages were infected with C. albicans, and after hyphae were allowed to grow for 3 hours, we repeatedly measured phagosomal pH every 5 s over a period of 10 to 60 min. As depicted in Fig. 6A, acidic phagosomes frequently underwent a series of rapid changes in pH. We refer to this phenomenon as a pH flash. While phagosomes often underwent  Fig. 6B, where the time courses of pH recordings from multiple phagosomes are overlaid. The rate of pH flashes was similar when the C. albicans ece1Δ mutant was phagocytosed, suggesting that pH flashes are a consequence of hyphal expansion rather than candidalysin ( Fig. 6C and Movie S4). The flashing pattern is consistent with the notion that, during hyphal expansion, phagosomes undergo tears that can in some instances be repaired, before irreversible rupture occurs, leading to sustained alkalinization and loss of SRB. It is noteworthy that neither the transient nor the permanent phagosomal breaks were associated with macrophage lysis, as the macrophages remained largely impermeable to propidium iodide (PI) 4 hours postinfection (Fig. S3) (total cell lysis was less than 10%).

DISCUSSION
In this study, we reached three main conclusions: first, that NH 3 generation by C. albicans and its retention by phagosomes cannot be responsible for the observed alkalinization; second, that initiation of hyphal growth occurs in acidic phagosomes; and third, that hyphal growth drives phagosomal alkalinization by stretching and eventually rupturing the phagosomal membrane. These conclusions are discussed in turn below.
It is generally agreed that the mature phagosome containing C. albicans is highly acidic, and low pH is known to inhibit yeast-to-hypha transition. Hence, it was recently suggested that C. albicans produces NH 3 to alkalinize the phagosome prior to hyphal growth (13,(15)(16)(17)(18). Our findings suggest that intracellular NH 3 production by C. albicans is not directly responsible for the alkalinization. The rate of NH 3 generation is much too low to overcome the ability of the V-ATPase to acidify the phagosome (Fig. 1), Phagosomal Alkalinization by C. albicans ® and more importantly, NH 3 cannot be retained within the phagosomes, which are extremely permeable to the uncharged weak base (Fig. 2).
Our findings indicate that, rather than being the cause of hyphal growth, phagosomal alkalinization is its consequence. How then is hyphal extension initiated? Several environmental factors influence yeast-to-hypha transition, including high temperature (37°C), adherence to host cells, high CO 2 , and nutrient deprivation (4,5). All these factors are experienced by C. albicans cells within the mammalian phagosome and may suffice to induce hyphal formation despite the acidic phagosomal pH, albeit at a reduced rate.
Among the factors listed above, CO 2 deserves special mention. CO 2 is known to contribute to the yeast-to-hypha transition as part of the process involving conversion of arginine to urea that is secondarily decomposed to NH 3 and CO 2 (9,45). Indeed, after phagocytosis by macrophages C. albicans changes its transcriptional program, stimulating biosynthesis of L-arginine, the L-arginase Car1p, and several related arginase genes (34) to generate urea. In turn, urea is converted to NH 3 and CO 2 by the amidolyases Dur1,2p. Accordingly, L-arginine and urea induced hyphal formation in a C. albicans wild-type strain but not in a dur1,2⌬ (amidolyase-deficient) strain (9). The dur1,2⌬ strain also failed to escape from mouse macrophages and was less virulent after intravenous injections in mice (46). While this has been interpreted as supporting a requirement for NH 3 formation, it may instead reflect a role for CO 2 or its conjugated weak base, bicarbonate. Bicarbonate, which serves a signaling role in other systems, would accumulate in the fungal cytoplasm supported by the active alkalinization promoted by the plasma membrane H ϩ -ATPase (Pma1) (47). Accumulation of bicarbonate would be further aided by alkalinization of the phagosomal lumen. Accordingly, we find that C. albicans hyphae grow faster in phagosomes treated with CCA, a V-ATPase inhibitor that dissipates the lysosomal and phagosomal acidification, which was verified using the acidotrophic dye cresyl violet (see Fig. S2D and E in the supplemental material). Thus, while capable of growing inside acidic phagosomes (Fig. 3), C. albicans hyphae indeed extend more rapidly at more alkaline pH (Fig. S2E).
Some of the evidence supporting the involvement of NH 3 stemmed from experiments using mutants with defective amino acid permeases, which ostensibly lacked the substrates to generate sufficient NH 3 . One such mutant, the stp2⌬ strain, was found to be unable to alkalinize phagosomes (15). In the presence of amino acids, Stp2 induces the transcription of genes leading to amino acid uptake and catabolism. This in turn produces urea, which subsequently leads to the production of NH 3 and CO 2 by urea amidolyases. It is noteworthy, however, that Stp2 affects the expression of several genes (48,49) and, as a result, suffers from growth defects, particularly in environments where nutrient availability is restricted, such as the phagosomal lumen. Thus, it is impossible to distinguish whether the effects caused by deletion of the gene are due to lack of NH 3 production or to impaired growth. Other mutants such as the ahr1Δ strain used to buttress the NH 3 hypothesis suffer from similar shortcomings, which is why we opted not to include them in our analyses (14,17,18). Since NH 3 generation appeared unlikely to account for the observed alkalinization, we sought for an alternative mechanism. Our data are consistent with the notion that hyphal growth distends the phagosomal membrane, causing leakage of proton equivalents and even larger molecules like SRB. In the early stages, the ruptures are transient, possibly reflecting the activation of repair mechanisms; indeed, we have preliminary evidence that reacidification is associated with additional fusion of lysosome-associated membrane protein (LAMP)-positive compartments with the phagosome (data not shown). The progressive membrane tears become irreversible thereafter, judged by the impossibility of reestablishing the acidic pH. Of note, the sudden and initially reversible increases in pH cannot be readily explained by NH 3 production, which is anticipated to be continuous, producing a gradual and sustained pH change. That mechanical stretching of the phagosomal membrane is the cause of the permeability change is supported by the observation that outward hydrostatic pressure established by osmotic means-using GPN-resulted in a similar disruption of the phagosomal membrane, with dissipation of the pH gradient and leakage of SRB (Fig. 5). Whether transient or more permanent, phagosomal membrane rupture exposes the fungus and its products to the cytosolic milieu. Little is known regarding the means whereby C. albicans-secreted effectors activate signaling pathways within the cytosol. We therefore speculate that discontinuities in the phagosomal membrane associated with hyphal growth could contribute to inflammasome activation and pyroptosis (50)(51)(52).
In conclusion, we propose that hyphal growth is initiated inside acidic phagosomes (albeit at a reduced rate) and that alkalinization results from excessive membrane distension, which either activates mechanosensitive channels and/or causes outright rupture of the phagosomal lining. The discontinuities may initially be transient, as the membrane is repaired by fusion with other organelles (likely LAMP-positive late endosomes-lysosomes), but eventually become permanent, leading to sustained alkalinization and granting the fungus access to the richer cytosolic environment.

MATERIALS AND METHODS
Strains and reagents. Experiments were carried out using mouse RAW 264.7 macrophages (ATCC). RAW cells were plated sparsely in 12-well tissue culture plates (Corning Inc.) and grown overnight at 37°C in an air-CO 2 (19:1) environment in RPMI 1640 (Wisent Inc.) supplemented with 5% (vol/vol) fetal bovine serum (FBS). The C. albicans wild-type strain was the prototrophic strain BWP17/CIp30 (53). Other strains used are listed in Table S2 in the supplemental material. C. albicans cultures were grown in YPD medium (1% yeast extract, 2% peptone, 2% dextrose) at 30°C overnight. Cultures were washed in sterile phosphate-buffered saline (PBS) and adjusted to the required cell density. Concanavalin A labeled with FITC and sulforhodamine B were from Invitrogen. Nigericin, monensin, and Gly-Phe-␤-naphthylamide were from Sigma. Concanamycin A was from Abcam. Recombinant pneumolysin was a kind gift from John Brumell.
Phagocytosis of C. albicans. After overnight incubation at 30°C, C. albicans yeast was washed twice in PBS and incubated in the dark with concanavalin A-FITC (1:100) and rabbit anti-C. albicans IgG (1:167, for 60 min at room temperature with rotation). After labeling, yeast was washed twice in PBS and diluted to an optical density at 600 nm (OD 600 ) of 1.0 in PBS.
Five microliters of concanavalin A-FITC-labeled yeast in fresh RPMI-FBS was added to RAW cells grown on glass coverslips, which were then centrifuged at 1,500 ϫ g for 1 min at room temperature to synchronize phagocytosis. After 20 min of incubation at 37°C, yeast not associated with the macrophages was washed away with PBS and yeast that had adhered to the macrophages but had not been internalized was labeled with donkey anti-rabbit Cy3 (1:1,000) for 10 min at 37°C. Cells were imaged live or, where indicated, fixed with 4% paraformaldehyde for subsequent analysis. All phagocytosis experiments were imaged in Hanks' balanced salt solution (HBSS).
Buffering capacity (␤). To determine buffering capacity (␤) of the C. albicans-containing phagosome, phagosomes were allowed to acidify for 1 h before phagosomal pH was measured ratiometrically, as described below. The bathing solution was switched to HBSS containing 15 mM NH 4 Cl, and the pH was measured again immediately. At the end of each experiment, a standard calibration was performed as described below and fluorescence ratios were converted to pH. The intraphagosomal NH 4 ϩ concentration ([NH 4 ϩ ]) was calculated using the Henderson-Hasselbalch equation, and the intrinsic buffer capacity (in millimoles/liter/pH) was calculated as Δ[NH 4 ϩ ]/ΔpH. V-ATPase pumping rate. Phagosomes were generated as described above, and after 1 h, the steady-state phagosomal pH was measured prior to addition of 2 M CCA to the bathing solution. Thereafter, the phagosomal pH was measured every minute for 15 min at 37°C. At the end of each experiment, a standard calibration was performed and fluorescence values were converted to pH. The rate of change of the luminal pH (ΔpH) measured during the first minute after CCA treatment was used to estimate the vacuolar H ϩ -ATPase (V-ATPase) pumping rate, which was assumed to be identical to the proton leakage rate at steady state. Proton pumping rates were calculated as (ΔpH ϫ ␤ ϫ phagosome volume)/time. To quantify the phagosomal volume, the radius of phagosomes containing C. albicans yeast was measured microscopically. Volume was calculated assuming that the phagosomes were spherical (volume ϭ 4/3 r 3 ). Phagosomal and lysosomal alkalinization was confirmed using the acidotrophic dye cresyl violet as previously described (54).
Determination of NH 3 leakage rate. Phagosomes were generated as described above, and after 1 h, the steady-state phagosomal pH was measured prior to the addition of 15 mM NH 4 Cl. Phagosomal pH was then measured every second for 10 s. A standard calibration was generated as described below, to determine the change in pH (ΔpH) induced by NH 3 addition or withdrawal. Leakage of NH 3 was calculated as (ΔpH ϫ ␤ ϫ phagosome volume)/time.
Dual-wavelength ratiometric fluorescence measurements. RAW cells were allowed to ingest C. albicans as described above, and the coverslips were then mounted in a Chamlide magnetic chamber and overlaid with HBSS. The chamber was placed in a Leiden microincubator maintained at 37°C on the stage of an inverted microscope (DM IRB; Leica Biosystems) equipped with a 40A/1.25-numericalaperture (NA) oil objective (Leica Biosystems), a lamp (X-Cite 120; EXFO Life Sciences Group), and filter wheels (Sutter Instrument) that control excitation and emission filters. For experiments using concanavalin A-FITC-labeled C. albicans, excitation wavelengths were alternated between 485 Ϯ 10 nm and 438 Ϯ 12 nm, with emitted light selected through a 520-nm filter. Light was captured by a cooled Conversion of dual-wavelength fluorescence ratio to pH. To convert measured phagosomal fluorescence ratios to pH, samples were sequentially bathed in isotonic K ϩ solutions containing 10 M nigericin and 5 M monensin and calibrated to pH 7.5, 6.0, 5.5, 5.0, and 4.5, respectively. Samples were imaged 5 min after addition of each solution to ensure pH equilibration across all compartments. After background subtraction at each wavelength, measured fluorescence ratios at defined pH were plotted into a calibration curve that was fitted with least squares. The measured phagosome fluorescence ratios were transformed into intracellular pH by using the equation describing the curve generated above. For pH measurements, fluorescence values from each phagosome were obtained using the freehand tool in Fiji (version 1.0) to select regions of interest (ROIs) delimiting the phagosome. The fluorescence intensity was corrected by subtracting the background fluorescence at each wavelength and converted to pH by using the equation described above. For pseudocolor display, a RatioPlus plug-in in Fiji was used to depict the 490/440-nm ratio.
C. albicans hyphal growth. The length of C. albicans was measured from differential inference contrast (DIC) images acquired at different time points after phagocytosis. Total length included the yeast head and hyphal projection.
Sulforhodamine B loading and leakage during phagocytosis. To load endosomes and lysosomes with SRB, RAW cells were bathed for 60 min at 37°C in RPMI-FBS containing 150 g SRB/ml, prior to phagocytosis. After loading, phagocytosis was initiated as described above. Leakage of SRB from phagosomes was measured using spinning disk confocal microscopy, and the percentage of SRB-positive phagosomes was calculated as (SRB-positive phagosomes/total phagosomes) ϫ 100. For GPN-induced membrane rupture, 200 M GPN was added to phagosomes containing yeast-locked C. albicans 1 h postinfection. GPN-treated phagosomes were visualized for 30 min by DIC and fluorescence microscopy.
Construction of an ECE1-expressing yeast-locked strain (cph1⌬/efg1⌬ ؉ pENO1-ECE1) of C. albicans. The cph1Δ/efg1Δ ϩ pENO1-ECE1 strain expressing ECE1 under the control of the C. albicans enolase (ENO1) promoter was generated from the parent strain HLC54 (cph1Δ/efg1Δ) (33). A cassette containing the C. albicans NAT1 (CaNAT1) marker and the ENO1 promoter was amplified from pNAT-ENO1 (50) using the primers ECE1_ENO1_PF and ECE1_ENO1_PR containing homology to the ECE1 gene, allowing targeted integration to replace the native ECE1 promoter (Table S1). Correct integration was confirmed by PCR with the ENO1-specific primer ENO1-S in conjunction with a reverse primer internal to the ECE1 gene (ECE1-IR) (Table S1). Positive transformants yielded a specific product approximately 450 bp in size (data not shown).
Quantification of ECE1 expression by reverse transcriptase-quantitative PCR (qRT-PCR). Total RNA was extracted from yeast-phase cells grown in YPD medium at 30°C or hyphae grown in RPMI 1640 according to the method described in reference 51. RNA (500 ng) was treated with DNase (Epicentre), and cDNA was synthesized using Superscript III reverse transcriptase (Invitrogen). cDNA samples were used for quantitative PCR with EvaGreen mix (Bio&SELL). Primers (ACT1-F/R and ECE1-F/R, Table S1) were used at a final concentration of 500 nM. qPCR amplifications were performed using a CFX96 thermocycler (Bio-Rad). ECE1 expression was calculated using the threshold cycle (ΔΔC T ) method, with ACT1 as the reference gene and C. albicans reference strain SC5314 (yeast morphology) as the control sample.
Results of three biological replicates were analyzed in Prism version 6 (GraphPad, San Diego, CA). LC-MS/MS analysis of hypha-secreted Ece1 peptides. Analysis of hypha-secreted Ece1 peptides was optimized for the detection of candidalysin and performed as previously described (28). Briefly fragmentation (30 V). Resulting MS/MS spectra were monitored at R ϭ 17.5k (FWHM). Proteome Discoverer 1.4 (Thermo) and the Sequest HT algorithm were used for protein database searching against C. albicans SC5314 (Candida Genome Database [http://www.candidagenome.org]). Mass spectra were searched for both unspecific cleavages (no enzyme) and tryptic peptides up to 4 missed cleavages. The precursor mass tolerance was 10 ppm, and the fragment mass tolerance was 0.02 Da. At least two unique peptides per protein, a false-discovery rate of Ͻ1%, and X corr validation (from 2.0 at z ϭ 2 up to 3.0 at z ϭ 6) were required for positive protein hits.
Statistical analysis. Unless otherwise indicated, data are presented as means Ϯ SEMs from the number of determinations shown in parentheses. Statistical significance was determined using unpaired t test, one-way analysis of variance (ANOVA) (Tukey's test or Dunnett's test), and two-way ANOVA (multiple comparisons) with Prism 7 (GraphPad Software), with P Ͻ 0.05 considered significant.