Cch1 Restores Intracellular Ca2+ in Fungal Cells during Endoplasmic Reticulum Stress*

Pathogens endure and proliferate during infection by exquisitely coping with the many stresses imposed by the host to prevent pathogen survival. Recent evidence has shown that fungal pathogens and yeast respond to insults to the endoplasmic reticulum (ER) by initiating Ca2+ influx across their plasma membrane. Although the high affinity Ca2+ channel, Cch1, and its subunit Mid1, have been suggested as the protein complex responsible for mediating Ca2+ influx, a direct demonstration of the gating mechanism of the Cch1 channel remains elusive. In this first mechanistic study of Cch1 channel activity we show that the Cch1 channel from the model human fungal pathogen, Cryptococcus neoformans, is directly activated by the depletion of intracellular Ca2+ stores. Electrophysiological analysis revealed that agents that enable ER Ca2+ store depletion promote the development of whole cell inward Ca2+ currents through Cch1 that are effectively blocked by La3+ and dependent on the presence of Mid1. Cch1 is permeable to both Ca2+ and Ba2+; however, unexpectedly, in contrast to Ca2+ currents, Ba2+ currents are steeply voltage-dependent. Cch1 maintains a strong Ca2+ selectivity even in the presence of high concentrations of monovalent ions. Single channel analysis indicated that Cch1 channel conductance is small, similar to that reported for the Ca2+ current ICRAC. This study demonstrates that Cch1 functions as a store-operated Ca2+-selective channel that is gated by intracellular Ca2+ depletion. The inability of cryptococcal cells that lacked the Cch1-Mid1 channel to survive ER stress suggests that Cch1 and its co-regulator, Mid1, are critical players in the restoration of Ca2+ homeostasis.

Store-operated Ca 2ϩ (SOC) 2 entry is a process whereby Ca 2ϩ influx across the eukaryotic plasma membrane results from the depletion of Ca 2ϩ from endoplasmic reticulum (ER) stores (1-6). Stresses imposed on the ER can promote a prolonged depletion of Ca 2ϩ that will lead to cell death unless ER Ca 2ϩ levels are replenished (1-3). All lower eukaryotes including the model fungal pathogen, Cryptococcus neoformans, express the Cch1-Mid1 channel (CMC) in their plasma membrane (7)(8)(9)(10)(11). C. neoformans is a primary human pathogen because it causes life-threatening meningitis primarily in immunocompromised patients (12,13). Its ability to survive in low [Ca 2ϩ ] environments is dependent on the combined activity of Cch1, the predicted pore of the channel complex, and Mid1, a binding partner of Cch1 (7,14,20). Genetic analysis has indicated that strains of C. neoformans lacking CCH1 or MID1 were not viable in conditions of limiting extracellular [Ca 2ϩ ], consistent with its role as the only high affinity Ca 2ϩ channel in the plasma membrane (7,14).
Treatment of fungal cells with azole antifungals led to an influx of extracellular Ca 2ϩ that promoted the survival of fungi exposed to azole-induced stress (16 -19). Because azoles promote ER stress by blocking ergosterol biosynthesis, it had been proposed that the depleted state of Ca 2ϩ stores might activate CMC where it could then function to refill Ca 2ϩ stores (16,17). This notion was reinforced by experiments demonstrating that changes in secretory Ca 2ϩ levels promoted the influx of extracellular 45 Ca 2ϩ possibly through CMC in yeast (20); however, the gating mechanism of Cch1 channel activity has yet to be established.
Resolving channel gating is best accomplished by patch clamp techniques because these techniques can successfully determine specific kinetic parameters of ion channels such as channel selectivity, channel conductance, and activation kinetics (21). We had previously applied patch clamp techniques directly to isolated spheroplasts of C. neoformans to resolve Cch1 function by measuring Cch1 channel activity straightaway; however, this approach was burdened with many technical difficulties that led to inconsistencies in Cch1 channel measurements. Consequently, the expression of Cch1 in a heterologous expression system such as the HEK293 (human embryonic kidney) mammalian cell line appeared to be the best approach that would permit direct functional analysis of the Cch1 channel. Unfortunately, the measurement of Cch1 channel activity in a heterologous expression system had been impossible until recently because of the significant challenge posed by the molecular cloning of CCH1 as a result of the toxicity of the CCH1 gene or gene product in Escherichia coli cells (8). For this reason, standard molecular biological techniques that used E. coli as a host could not be used, and therefore we devised a new approach that ultimately led to the successful cloning of CCH1 and its functional expression (22).
In the present study we report the first electrophysiological characterization and mechanistic study of Cch1 channel activity. The functional expression of Cch1 along with its subunit Mid1 revealed that the Cch1 channel is directly activated by ER-stressing agents that deplete ER Ca 2ϩ levels, suggesting that Cch1 is gated by the passive depletion of intracellular Ca 2ϩ and that it functions primarily to restore Ca 2ϩ homeostasis. Consequently, cryptococcal cells that lacked the Cch1 channel could not survive ER stress, suggesting a critical physiological requirement for CMC.

EXPERIMENTAL PROCEDURES
Cell Culture and Reagents-All C. neoformans var. grubii strains (H99 MATa serotype A) were recovered from 15% glycerol stocks stored at Ϫ80°C prior to use in these experiments. The cch1⌬ null mutant strain was constructed previously (7). Strains were maintained on YPD (1% yeast extract, 2% peptone, and 2% dextrose) medium. Cells of C. neoformans were cultured in YPD medium at 30°C for 24 h. Cells from the HEK293 cell line were cultured in Dulbecco's modified Eagle's medium with 10% calf serum and antibiotics in a 5% CO 2 incubator at 37°C. HEK293 cells were purchased from ATCC (CRL-1573). Where indicated, a cell-impermeable form of BAPTA (Invitrogen) was added to YPD. Tunicamycin, fluconazole, and ketoconazole (Sigma-Aldrich) were dissolved in dimethyl sulfoxide at the concentrations noted in the figure legends. Thapsigargin and BAPTA-AM were purchased from Calbiochem and BAPTA from Invitrogen.
Construction of the Null Mutant Strains-Gene disruption cassettes were generated to knock down Mid1 and create the mid1⌬ single mutant and the cch1⌬mid1⌬ double mutant. The URA5 (uracil) nutritional marker was amplified by PCR (with primers: Ura5-Ssp1-F: CCC AATATT gatcttgggg atggtattga and Ura5R-Ssp1-R: CCC AATATT gatcccagtactacccgctc) and cloned into the middle of pCR2.1-MID1 vector to disrupt the MID1 open reading frame. This disruption cassette was transferred to a wild type C. neoformans strain by biolistic transformation as described previously (23). Colonies appeared after 3 days on ϪURA dropout medium (0.01% adenine, 0.2% yeast synthetic dropout medium without uracil, 0.5% ammonium sulfate, 0.17% yeast nitrogen base without amino acids and ammonium sulfate, 2% Bacto Agar) supplemented with 1 M sorbitol. The mid1⌬ mutant strain was confirmed by PCR. Southern blot analysis was performed to confirm a single integration in the knock-out strain.
A similar approach was taken to construct the cch1⌬mid1⌬ double mutant strain; however, in this case the NAT marker was used to disrupt the MID1 coding region. The disruption cassette was transformed into a cch1⌬ single mutant strain by biolistic transformation. Colonies appeared after 3 days on YPD ϩ 1 M sorbitol agar plates supplemented with 0.1 mg/ml NAT plates. The cch1⌬mid1⌬ mutant strain was confirmed by PCR and Southern blot analysis.
Plasmids and Transfection-We had determined previously that the CCH1 gene from C. neoformans could not be cloned by standard ligation-mediated techniques that used E. coli as a host because the gene encoding Cch1 or possibly the gene product is toxic to E. coli. We have devised an alternative approach to clone CCH1 cDNA into the pcDNA3.1/CT-GFP-topo expression plasmid under the control of the constitutive mammalian cytomegalovirus promoter (Invitrogen) that bypassed the conventional enzyme-mediated ligation reactions in E. coli. The approach we used to clone CCH1 successfully has been described elsewhere (22). The Cch1 channel protein was tagged with a C-terminal GFP instead of an N-terminal GFP to avoid interfering with the localization of Cch1 to the plasma membrane of HEK293 cells. A nontagged version of a Cch1 plasmid was also constructed and used as a control to assess the effect of the GFP fusion on the kinetics of Cch1. MID1 was cloned into the pcDNA3.1/CT-GFP-topo expression plasmid using conventional molecular biological techniques. The MID1 amplicon was generated from cDNA that was synthesized by the SuperScript III First-Strand synthesis system for RT-PCR (Invitrogen) using total RNA as template with the following primers: Mid-1F: atgccagcga gagaggtgta, Mid-1R: ctatccgttacaccatctat. RNA was isolated from a culture (ϳ5 ϫ 10 7 cells) of a wild type strain of C. neoformans var. grubii (H99). Cells were lysed with acid-washed glass beads, and total RNA was isolated and purified according to the manufacturer's instructions (RNeasy kit; Qiagen). Following blue/white colony selection, positive transformants were isolated and sequenced.
HEK293 cells were counted and trypsinized 24 h before transfection. Approximated 1.25 ϫ 10 5 in 1 ml of complete medium were plated/well such that the cell density was ϳ50 -80% confluent on the day of transfection. Transient transfection of HEK293 cells with CCH1-GFP plasmid DNA and MID1 was performed with Lipofectamine 2000 and Plus reagents according to the manufacturer's instructions (Invitrogen). Approximately 1 g of CCH1-GFP and MID1 plasmid DNA was added to 100 l of Opti-MEM I-reduced serum medium and mixed with 1.75 ml of Lipofectamine 2000, and the mixture was incubated for 30 min at room temperature. The mixture was added directly to HEK293 cells that had been grown in a culture dish and maintained in Dulbecco's modified Eagle's medium supplemented with 4 mM L-glutamine, 10% fetal bovine serum. HEK293 cells were maintained at 37°C with 5% CO 2 18 -24 h after transfection before assaying cells for transgene expression.
Protein Analysis-Plasma membranes from HEK293 cells expressing Cch1-GFP or Mid1-GFP were isolated according to the protocol outlined in the Qproteome plasma membrane protein kit (Qiagen). Plasma membrane preparations were separated on a 6.5% SDS-polyacrylamide gel. Western blotting for Mid1 and Cch1 was performed using a rabbit polyclonal antibody to GFP (Abcam) (1:5000) as the primary antibody followed by detection with a goat polyclonal antibody to rabbit immunoglobulin G (IgG) (1:5000) (H&L (horseradish peroxidase); Abcam).
Biotinylation of Mid1 and Cch1 was performed according to the protocol described in the Pierce cell surface biotin protein isolation kit (ThermoScientific). HEK293 cells that had been transfected with Mid1-GFP and Cch1-GFP were biotinylated according to the specifications outlined in the kit. Protein samples were separated on a 6.5% polyacrylamide gel. Western blotting for biotinylated Mid1 and Cch1 was performed as explained above.
Confocal Microscopy-To examine the localization of Mid1 and Cch1 in HEK293 cells, cells expressing plasmids containing Mid1-GFP or Cch1 were fixed in 4% paraformaldehyde for 20 min at room temperature, washed in phosphate-buffered saline and subsequently visualized with a confocal microscope. Immunofluorescence of Cch1 was performed in HEK293 cells that had been fixed in 4% paraformaldehyde and permeabilized. A primary peptide antibody against the C terminus of Cch1 (Antibodies Inc., Davis CA) was added to HEK293 cells at a dilution of 1:500. Cch1 was visualized with a secondary antibody (1:1000) conjugated with Texas Red (Abcam). Fluorescence was examined using a Carl Zeiss LSM-5 inverted confocal microscope. Mid1 and Cch1 were examined with a 488-nm and a 568-nm laser line at 40 ϫ objective. The images were scanned and captured at a resolution of 1024 ϫ 1024 pixels.
Patch Clamp Measurements-Experiments were performed using conventional whole cell and single channel patch clamp techniques (21). To measure currents through Cch1, HEK293 cells expressing Cch1 and Mid1 were visually selected based on bright GFP fluorescence. Currents were stimulated by passive Ca 2ϩ store depletion with a pipette solution containing 10 M BAPTA-AM or by activation with thapsigargin (100 M). Recording electrodes were pulled with a vertical puller (Adams and List, New York) from borosilicate glass capillaries, coated with Sylgard, fire-polished and filled with a solution containing either BAPTA-AM or thapsigargin, 130 mM Cs ϩ methane sulfonate, 5 mM MgCl 2 , 500 M Mg-ATP, 20 mM HEPES (pH 7.2) with CsOH. Recording electrodes had a tip resistance of ϳ5 M⍀ when placed in the following external solution: 2, 5, or 10 mM CaCl 2 or BaCl 2 , 140 mM N-methyl-D-glucamine (140 mM Na ϩ methane sulfonate in divalent-free solution and free of N-methyl-D-glucamine), 10 mM glucose, 10 mM HEPES (pH 7.4) with NaOH. Whole cell currents were measured by voltage ramps (Ϫ120 to ϩ60 mV lasting 200 ms) from a holding potential of Ϫ60 mV. All voltages have been corrected for liquid junction potentials. Currents were filtered at 2 kHz with a four-pole Bessel filter (Dagan) contained in the Dagan amplifier and sampled at 5 kHz. Amplifier was interfaced with a Digidata 1322A (Axon Instruments, Foster City, CA) to digitize data. All data were corrected for leak currents. For single channel measurements, the outside-out patch configuration was used. Both the extracellular and pipette solutions used for single channel recordings were the same as those used in whole cell experiments. Single channel records were obtained at voltages of Ϫ60 mV and Ϫ80 mV and sampled at 25 kHz. Currents were analyzed with pCLAMP 9.0 software (Axon) and graphed with Sigmaplot 8.0 software on an IBM 3GHz computer.

Cch1 and Mid1
Are Expressed in the Plasma Membrane of HEK293 Cells-To resolve Ca 2ϩ currents through CMC by patch clamp techniques, HEK293 cells were co-transfected with CCH1 and MID1 (21). Prior to initiating the patch clamp studies, the expression and localization of CMC was examined in HEK293 cells. This heterologous expression system was chosen because previous attempts to resolve CMC-mediated Ca 2ϩ currents directly in intact cells of C. neoformans and in yeast had failed. To express CMC in HEK293 cells, cDNA of MID1 or CCH1 was cloned into mammalian expression vectors, which were subsequently transfected into HEK293 cells. Detection of mRNA transcripts of Cch1 and Mid1 confirmed the expression of the channel in HEK293 cells that were previously transfected with CCH1 and MID1 (Fig. 1A). Transcripts of Cch1 and Mid1 were not detected in nontransfected HEK293 cells.
To assess the localization of CMC in HEK293 cells, confocal microscopy was performed. Cells expressing a Mid1-GFP fusion protein revealed a cell surface distribution of Mid1 that co-localized with Cch1 (Fig. 1B) (24). Western blot analysis of isolated plasma membranes from HEK293 cells expressing Cch1 or Mid1 revealed the presence of polypeptides corresponding to the predicted size of Cch1 and Mid1, suggesting that both proteins were indeed sufficiently expressed in HEK cells (Fig. 1C). Interestingly, under nonreducing conditions (the Tubulin is shown as a loading control. B, Mid1 was tagged with a C-terminal GFP tag and visualized by confocal microscopy (center panel). Mid1 is expressed predominantly on the cell surface of HEK293 cells, and Mid1 colocalizes with Cch1 (right panel, merge). Immunofluorescence (Texas Red) of Cch1 (nontagged) revealed a surface distribution in HEK cells (left panel) similar to Mid1. A primary peptide antibody to the C terminus of Cch1 and a Texas Red-conjugated secondary antibody were used to visualize Cch1. Scale bars, 10 M. C, Western blot analysis of isolated plasma membranes from HEK293 cells expressing Cch1 or Mid1 and revealed polypeptides corresponding to the predicted sizes for Cch1 (ϳ220 kDa) and Mid1 (monomer, ϳ100 kDa; complex, ϳ200 kDa). D, Western blot analysis of surface biotinylation of HEK293 cells expressing Cch1 or Mid1 revealed two distinct bands unique to the lane corresponding to Mid1 (monomer, ϳ100 kDa; complex, ϳ200 kDa). These bands were not detected in nontransfected HEK cells (NT). A protein band corresponding to biotinylated Cch1 could not be detected.
absence of ␤-mercaptoethanol) two distinct protein bands were detected for Mid1 similar to that reported for Mid1 in yeast (40). The 100 kDa band represented the Mid1 monomer, and the 200 kDa band likely corresponded to a Mid1 complex (Fig.  1C) (40). This result suggested that Mid1 likely forms an oligomer through sulfide bonding similar to Mid1 in yeast (40).
To confirm further the expression of Cch1 and Mid1 on the surface of HEK293 cells, biotinylation was performed. Western blot analysis of biotin-labeled HEK293 cells expressing Cch1 or Mid1 revealed two distinct bands corresponding to the monomer and the complex forms of Mid1 (Fig. 1D) similar to that observed in Western blots of isolated plasma membranes from HEK cells expressing Mid1 (Fig. 1C). These distinct bands were not detected in biotinylated HEK cells that did not express Mid1 or Cch1 (nontransfected HEK cells). Unfortunately, Cch1 could not be detected in Western blots from biotinylated HEK cells expressing Cch1 (Fig. 1D). Because Cch1 is such a large and bulky protein based on its predicted structure (24 transmembrane domains), it is conceivable that steric hindrance may have masked surface lysine residues thus preventing sufficient biotin from attaching. Nevertheless, the biotinylation of Mid1 confirms its expression on the surface of HEK293 cells, and the co-localization of Mid1 with Cch1 strongly suggests that Cch1 is also expressed on the surface of HEK cells.
Cch1 Is Activated by the Depletion of Intracellular Ca 2ϩ Stores-To assess whether Cch1-mediated Ca 2ϩ currents were activated by passive Ca 2ϩ store depletion, the Ca 2ϩ chelator BAPTA-AM was added to the patch pipette. This uncharged molecule permeates cell membranes, and once inside the cell the lipophilic blocking groups are cleaved by nonspecific esterases resulting in a charged form that remains trapped within the cytosol (3). Consequently, ER Ca 2ϩ stores are eventually depleted because the chelator captures released Ca 2ϩ . Ramp protocols were used to measure the time-dependent development of Ca 2ϩ currents through Cch1 in HEK293 cells expressing Cch1, Mid1, or both. Ca 2ϩ store depletion resulted in the development of robust, time-dependent Cch1 channel currents with either Ca 2ϩ or Ba 2ϩ as the charge carrier. In this case, Ba 2ϩ was used as a surrogate for Ca 2ϩ , a common practice when measuring Ca 2ϩ channels (25). Notably, Ba 2ϩ currents developed slower than Ca 2ϩ currents. This was probably due to secondary effects of Ba 2ϩ blocking K ϩ conductances and affecting the electrical potential difference across the plasma membrane thus reducing the driving force for Ba 2ϩ influx (Fig. 2, A and B) (25). Inward Ca 2ϩ currents through Cch1 developed promptly in response to BAPTA-AM and spontaneously declined to resting levels (Fig. 2C).
The addition of the well characterized Ca 2ϩ channel blockers such as nifedipine and verapamil did not block Cch1-mediated currents at physiological concentrations (data not shown). Some inhibition of Cch1 Ca 2ϩ currents was observed in the presence of nifedipine but only at very high concentrations (10 mM), suggesting that CnCch1 probably lacks the classic binding site for dihydropyridines and phenylalkyamines. In contrast, the addition of 10 M La 3ϩ resulted in a 75% reduction in the time-dependent inward Ca 2ϩ currents, indicating that La 3ϩ blocked Cch1 channel activity (Fig. 2D). The La 3ϩ -induced block of Ca 2ϩ currents is a pharmacological hallmark of SOC channels (3,4), further strengthening the functional similarity between CMC and SOC channels in mammalian cells.
Consistent with the stimulation of Ca 2ϩ currents by chelation of Ca 2ϩ from the ER, time-dependent inward Ca 2ϩ currents through Cch1 also developed in response to thapsigargin (Fig. 2E). Thapsigargin, an inhibitor of the ER Ca 2ϩ -ATPase pump, leads to the depletion of Ca 2ϩ stores because as Ca 2ϩ passively leaks from the ER, Ca 2ϩ levels are no longer balanced by reuptake (26). Store depletion-mediated inward Ca 2ϩ currents were not detected in HEK cells that expressed either Cch1

Cch1 Restores Intracellular Ca 2؉
or Mid1 alone or an empty plasmid, indicating that the activation of these Ca 2ϩ currents was completely dependent on the co-expression of Cch1 and Mid1 and that Cch1 activity required Mid1, consistent with previous genetic evidence (Fig.  2, F-H) (12,14,20,21). Importantly, Mid1 did not conduct any Ca 2ϩ currents on its own during passive Ca 2ϩ store depletion. The mechanism underlying the action of Mid1 on Cch1 function remains unknown.
Cch1 Is Permeable to Both Ca 2ϩ and Ba 2ϩ -Macroscopic I-V curves demonstrated that depletion of Ca 2ϩ stores by BAPTA-AM produced Cch1-mediated currents with either Ca 2ϩ or Ba 2ϩ as the charge carrier. Surprisingly, although Ca 2ϩ and Ba 2ϩ were both permeable, Ba 2ϩ currents were steeply voltage-dependent. As membrane potentials became decreasingly hyperpolarized, Ba 2ϩ currents fell to a greater extent than Ca 2ϩ currents, resulting in a more pronounced inward rectification (Fig. 3A). The Ca 2ϩ or Ba 2ϩ channel currents of Cch1 were not affected by the presence of the C-terminal GFP tag on the Cch1 protein as indicated by the I-V plot (Fig. 3E). In addition, currents were not detected in HEK cells expressing empty plasmids, strongly suggesting that the Ca 2ϩ currents detected were mediated by Cch1 (Fig. 3E). The amplitude of Cch1-mediated currents at Ϫ40 mV with Ca 2ϩ as the charge carrier was ϳ50% less than the amplitude measured at Ϫ80 mV (Fig. 3B). However, with Ba 2ϩ the current amplitude at Ϫ40 mV was only ϳ10% of the current levels detected at Ϫ80 mV, indicating that Ba 2ϩ permeation through Cch1 was voltage-dependent. We found that Ba 2ϩ currents were negligible at membrane potentials depolarized beyond Ϫ60 mV, whereas Ca 2ϩ currents were detected at these same potentials, reflecting the voltage independence of Ca 2ϩ currents through Cch1 under physiological conditions.
The voltage dependence of Ba 2ϩ currents through Cch1 was reflected in the ensemble average of single channel records consistent with the macroscopic I-V curves (Fig. 3C). Single channel events were not detected at Ϫ60 mV with Ba 2ϩ as the charge carrier; however, channel activity increased dramatically at Ϫ80 mV, indicating the voltage dependence of single channel Ba 2ϩ currents through Cch1 (Fig. 3C). Single channels exhibited long duration openings with a conductance of ϳ0.08 picosiemens in 10 mM Ba 2ϩ , consistent with the small channel conductance reported for CRAC channels (27).
Cch1 Maintains High Ca 2ϩ Selectivity in the Presence of Monovalent Ions-One of the distinguishing properties of SOC channels is their highly permeability to sodium ions and other monovalent ions in the absence of divalent cations (27). We sought to determine whether CMC shared this feature with SOC channels, and therefore we tested whether Cch1 was permeable to monovalent ions in the absence or presence of divalent cations. The reversal potential of the Ba 2ϩ inward current (E Ba ) in 5 mM Ba 2ϩ was ϩ42 Ϯ 2 mV and ϩ58 Ϯ 2 mV for the Ca 2ϩ inward current (E Ca ) in 5 mM Ca 2ϩ (Fig. 3A). These values were far less positive than the predicted E Ba and E Ca values for experimental conditions where extracellular [Ca 2ϩ ] and [Ba 2ϩ ] were about 10 5 times higher than the [Ca 2ϩ ] and [Ba 2ϩ ] inside the patch pipette. These results indicated that Cs ϩ carried some outward current through Cch1 and in doing so made a significant contribution to the reversal potential of Cch1 by shifting the E Ba and E Ca to less positive values. However, Cch1 was more selective for Ca 2ϩ than Ba 2ϩ because E rev was more positive and closer to the equilibrium potential for Ca 2ϩ . Interestingly, Na ϩ conductance through Cch1 in the absence of divalent cations was negligible, and the addition of Ca 2ϩ in the presence of Na ϩ resulted in large inward currents with a shift in E rev to positive potentials approaching E Ca (Fig. 3D). These results indicated that unlike some SOC channels in higher eukaryotes that are highly permeable to Na ϩ in the absence of divalent cations, Cch1 is weakly permeable to Na ϩ and maintains a strong Ca 2ϩ selectivity even in the presence of high [Na ϩ ] (4,27).
Cryptococcal Cells Require Cch1 Activity to Survive ER Stress-To test the physiological role of the channel, strains of C. neoformans lacking a functional CMC were constructed and tested for survival with ER-stressing agents. To single out the requirement for CMC in low [Ca ϩ ], ER-stressing agents were added to YPD medium (free [Ca 2ϩ ] ϳ0.140 mM) where the free [Ca 2ϩ ] was reduced to ϳ500 nM with a cell-impermeable form of a Ca 2ϩ chelator (BAPTA) (28,29). Lowering [Ca 2ϩ ] further rendered the CMC mutants inviable, but at ϳ500 nM, low affinity Ca 2ϩ channels and/or transporters allow for Ca 2ϩ influx that supports viable CMC mutants (7,20). In the presence of the ER-stressing agent tunicamycin, a wild type strain of C. neoformans showed no significant growth defect with the exception of the mid1⌬ mutant (Fig. 4, A and E); however, the growth of C. neoformans strains lacking a functional CMC was severely compromised. The cch1⌬ and mid1⌬ single null mutants and the cch1⌬mid1⌬ double mutants were not viable when exposed to tunicamycin in low extracellular [Ca 2ϩ ] (Fig. 4, B and F). A similar phenotype was observed in the presence of fluconazole or ketoconazole, drugs belonging to the azole class of antifungals (Fig. 4, C, D, G, and H) (30,31). These results indicated that cryptococcal cells required CMC to survive ER stress conditions and supported the notion that the Cch1 channel plays an essential role in reestablishing Ca 2ϩ homeostasis.

DISCUSSION
The results presented here demonstrate that the Cch1-Mid1 proteins constituted a functional Ca 2ϩ channel in HEK293 cells that exhibited some features inherent to the non-voltage-gated SOC channels. This first mechanistic study of Cch1 indicates that it is a Ca 2ϩ -selective SOC channel that is gated by the depletion of intracellular Ca 2ϩ based on the following observations: (i) inward Ca 2ϩ currents were activated upon the passive depletion of Ca 2ϩ with BAPTA-AM and thapsigargin at negative membrane potentials; (ii) Cch1 was permeable to Ba 2ϩ ions; and (iii) Ca 2ϩ currents were blocked by La 3ϩ .
It is known that in eukaryotes the ER constitutes a finite reservoir of Ca 2ϩ primarily mobilized for signaling that must be replenished to sustain the duration of signaling and ultimately restore Ca 2ϩ homeostasis in the ER and the cytosol (1-3, 32). In C. neoformans mediators of the Ca 2ϩ signaling pathway, such as calcineurin, calmodulin, and the sarco (endo)plasmic reticulum Ca 2ϩ -ATPase are central to maintaining intracellular Ca 2ϩ homeostasis while withstanding the multitude of stresses imposed by the host (33). ER stress can be detrimental to fungal cells if ER function is not restored because in addition to its FIGURE 3. Divalent and monovalent inward currents through the Cch1 channel in HEK293 cells. A, whole cell currents revealed that Cch1 is permeable to both Ca 2ϩ (5 mM) and Ba 2ϩ (5 mM). Arrows indicate E rev (reversal potential) for E Ca and E Ba . Inward current is carried by Ca 2ϩ or Ba 2ϩ , and outward current above ϩ40 mV is carried by Cs ϩ . B, Ba 2ϩ currents are voltage-dependent, whereas Ca 2ϩ currents are voltage-independent. Current levels for Ca 2ϩ and Ba 2ϩ measured at Ϫ80 mV and Ϫ40 mV were plotted (n ϭ 5). Data are expressed as means Ϯ S.D. (error bars). C, representative single channel record of single channel Ba 2ϩ currents through Cch1. Single channels exhibited long duration openings and a small channel conductance (ϳ0.08 picosiemens). D, weak Na ϩ current through the Cch1 channel in the absence of divalent cations. Cch1 maintains a high Ca 2ϩ selectivity in the presence of Ca 2ϩ and high [Na ϩ ]. E, Ca 2ϩ or Ba 2ϩ currents were not detected in HEK cells that had been transfected with empty plasmids as shown in the current-voltage (I-V) plot. In addition, the presence of a C-terminal GFP tag on Cch1 did not alter the channel kinetics of Cch1 as shown from the I-V plot measured from HEK cells expressing Cch1 without a GFP tag. E, Ca 2ϩ or Ba 2ϩ channel currents of Cch1 were not affected by the presence of the C-terminal GFP tag on the Cch1 protein as indicated by the I-V plot. Currents were not detected in HEK cells expressing empty plasmids (2) or in nontransfected HEK cells (indicated by flat I-V plots).
crucial role in the posttranslational processing of proteins, the ER is also a key player in Ca 2ϩ signaling (32). Here, we show that Cch1 activity plays an essential role in the survival of cryptococcal cells in response to ER stress by coupling the depleted state of ER Ca 2ϩ stores to the influx of Ca 2ϩ across the plasma membrane. The presence of a specific Ca 2ϩ chelator and an inhibitor of the Ca 2ϩ -ATPase, both capable of depleting intracellular ER Ca 2ϩ stores, led to the development of Ca 2ϩ inward currents at negative membrane potentials, indicating that Cch1 functions primarily to replenish these stores and reestablish Ca 2ϩ homeostasis. This is supported by studies in yeast that have suggested that the Cch1-Mid1 channel likely promotes the influx of Ca 2ϩ during depletion of secretory Ca 2ϩ and that this activity was necessary for yeast survival during ER-stressing conditions (17,18,20). The electrophysiological evidence provided here strongly supports a Ca 2ϩ depletion gating mechanism for Cch1 and further substantiates a role for Cch1 in the process of SOC entry in lower eukaryotes (20). Similar to other Ca 2ϩ channels in higher eukaryotes, Cch1 was also highly permeable to barium ions. Unexpectedly, Ba 2ϩ movement through Cch1 was voltage-dependent; however, Ca 2ϩ influx was not. This suggested that very hyperpolarized potentials were probably needed to displace Ba 2ϩ bound to the Cch1 pore with an incoming Ba 2ϩ ion likely due to weak electrostatic repulsion between Ba 2ϩ ions, thus resulting in the voltage dependence of Ba 2ϩ influx (25). Interestingly, CRAC channels, a subtype of the voltage-independent SOC channels in mammalian cells, also exhibit a similar voltage-dependent influx of Ba 2ϩ at negative membrane potentials (34). Our results indicate that Cch1 is not voltage-gated, and this is further supported by our previous findings demonstrating that the protein sequence of Cch1 in C. neoformans lacks the highly conserved voltage sensor motif, the hallmark of all voltage-gated channels in eukaryotic cells (7,(35)(36)(37). Moreover, the recently characterized NALCN neuronal Na ϩ channel, which like CnCch1 has fewer positively charged amino acids (lysine and arginine) within its voltage-sensing S4 segment and thus lacks a bona fide voltage sensor, was found to be voltage-independent (38).
The activation of Cch1-mediated Ca 2ϩ currents required the presence of Mid1, indicating that Cch1 cannot promote Ca 2ϩ influx independently of Mid1. This is consistent with previous genetic data that demonstrated identical phenotypic defects of strains lacking either Mid1 or Cch1 including nonviability in low [Ca 2ϩ ] medium (7,11,15,20). Despite the noted requirement for Mid1, its role in the functional regulation of the Cch1 channel remains unknown. Our results rule out the possibility that Mid1 is involved in the trafficking of Cch1 to the plasma membrane because Cch1 was clearly expressed in the plasma membrane of HEK293 cells independently of Mid1. In addition, recent evidence has demonstrated that the localization of Cch1 to the plasma membrane in cryptococcal cells is dependent on an elongation factor (EF3), further suggesting that Mid1 is excluded from this process (14). Under Ca 2ϩ -depleted conditions, Mid1 did not promote the influx of divalent cations when expressed on its own, indicating that Mid1 does not have any independent channel activity under these particular conditions (39). It is conceivable that Mid1 may be coupling the depleted state of intracellular Ca 2ϩ to the activity of Cch1. This is a particularly appealing notion based on the pivotal role of STIM1 in the activation of Orai1, a non-voltage-gated, storeoperated Ca 2ϩ -selective channel in mammalian cells (5). However, although Mid1 appears to have a predicted ER retention signal it does not contain any predicted Ca 2ϩ sensor domains thus if Mid1 did function in this role, other proteins would also likely be involved.
In summary, the SOC gating mechanism of Cch1 indicates that the Cch1 channel functions primarily to replenish depleted Ca 2ϩ levels and reestablish Ca 2ϩ homeostasis. This enables C. neoformans to overcome ER stress-mediated events that are likely rampant during infection and colonization of the host.