Ion-channel blocker sensitivity of voltage-gated calcium-channel homologue Cch1 in Saccharomyces cerevisiae

Mammalian voltage-gated Ca²⁺ channels (VGCCs) are classified electrophysiologically into five classes on the basis of Ca²⁺ currents: L-, P/Q-, N-, R- and T-types (Catterall, 2000; Yamakage & Namiki, 2002). All but one class (T-type) of VGCCs is inhibited by specific blockers, and these blockers are frequently used to classify VGCCs of interest. VGCCs are composed of four or five subunits: α₁, α₂/δ, β and γ; α₁ is the most prominent subunit, and is responsible for Ca²⁺ permeation, Ca²⁺ selectivity and voltage sensing (Catterall, 2000). The structure of the α₁ subunit consists of four homologous domains (termed I–IV), and each domain contains six transmembrane segments (termed S1–S6), and a membrane-associated loop (termed the pore loop or P-loop) between S5 and S6 (see Fig. 1). The P-loop has a motif, T-X-E/D-X-W, responsible for Ca²⁺ selectivity, and the S5–P–S6 region forms the pore domain. The S4 segment contains several positively charged amino acid residues, each of which is followed by two hydrophobic residues; and the segment acts as a voltage sensor.

(19K): Fig. 1. Predicted membrane topology of Cch1, a homologue of animal VGCC α1 subunits. I–IV, putative domains; 1–6, putative transmembrane segments (S1–S6); P, putative pore loop located between the S5 and S6 segments. In the P-loop, there is a sequence homologous to the consensus T-X-E/D-X-W motif, in which E or D participates in Ca2+ selectivity in animal VGCC α1 subunits. The P-loops of domains II, III and IV of Cch1 contain E, but the P-loop of domain I has N instead of E or D. + (located in the S4 segment) represents a positively charged amino acid residue R or K that serves for voltage sensing in animal VGCC α1 subunits. Cch1 is composed of 2039 aa residues.

The yeast Saccharomyces cerevisiae has one orthologue, termed Cch1, of the pore-forming α₁ subunit of mammalian VGCCs (Fischer et al., 1997; Paidhungat & Garrett, 1997). Although the organization of domains and transmembrane segments of Cch1 is similar to those of α₁ subunits, overall amino acid sequence identity between Cch1 and α₁ subunits is low: e.g. 24 % for Cch1 versus an L-type mammalian VGCC (Paidhungat & Garrett, 1997). It is noteworthy that one of the four S4 segments of Cch1 lacks most of the positively charged residues crucial for voltage sensing. In addition, each of the remaining three S4 segments of Cch1 has four positively charged residues, while most mammalian S4 segments have five positively charged residues. Therefore, from a structural viewpoint, it is uncertain whether Cch1 functions in response to changes in membrane potential. Cch1 constitutes a high-affinity Ca²⁺-influx system with another protein, Mid1 (Muller et al., 2001). Mid1 does not have any structural homologues in higher eukaryotes, but has a stretch-activated Ca²⁺-channel activity when expressed in mammalian cultured cells (Kanzaki et al., 1999). Mutant cells lacking either CCH1 or MID1, or both, show the same phenotypes: low Ca²⁺-uptake activity, and mating-pheromone-induced death (Iida et al., 1994; Fischer et al., 1997; Paidhungat & Garrett, 1997; Muller et al., 2001; Iida et al., 2004). The molecular nature of a high-affinity Ca²⁺-influx system composed of Cch1 and Mid1 is unknown. In this paper, we characterize the kinetic property of a putative Cch1–Mid1 Ca²⁺ channel, and present experimental evidence to suggest that yeast Cch1 is pharmacologically similar to L-type VGCCs. Yeast strains and plasmids.
Yeast strains and plasmids used in this study are listed in Table 1.

Table 1. Yeast strains and plasmids used in this study

Media.
Low-Ca²⁺ medium SD.Ca100 containing 100 µM CaCl₂ as the sole Ca²⁺ source, and SD-Ca medium containing no CaCl₂, were prepared as described previously (Iida et al., 1990).

Blockers and agonist.
Stock solutions of the following compounds were prepared as follows: 50 mM nifedipine (in DMSO; Sigma-Aldrich Japan, cat. no. N7634), 100 mM verapamil hydrochloride (in H₂O; Wako Jyunyaku, 222-00781), 100 mM (+)-cis-diltiazem hydrochloride (in SD.Ca100 medium; Sigma-Aldrich Japan, D2521), 25 mM (S)-(–)-Bay K8644 (in DMSO; Sigma-Aldrich Japan, B133), 50 mM amiloride hydrochloride hydrate (in DMSO; Sigma-Aldrich Japan; A7410), 10 µM ω-agatoxin IVA (in H₂O; Sigma-Aldrich Japan, A6719), 100 µM ω-conotoxin GIVA (in H₂O; Sigma-Aldrich Japan, C9915), 20 µM SNX-482 (in H₂O; Sigma-Aldrich Japan, S1818), 10 mM tetrodotoxin (TTX; in 10 mM phosphate buffer, pH 7.0; Wako Jyunyaku, 207-15901), and 10 M tetraethylammonium (TEA) chloride (in H₂O; Sigma-Aldrich Japan, T2265).

Ca²⁺-accumulation assay.
Exponentially growing cultures (approx. 2x10⁶ cells ml^–1) of yeast strains were incubated for 2 h at 30 °C with ⁴⁵CaCl₂ (PerkinElmer Japan; 185 kBq ml^–1; 1.81 kBq nmol^–1) with or without an appropriate blocker or agonist, and an aliquot (100 µl; duplicate) was taken, filtered on a Millipore filter (type HA; 0.45 µm) presoaked in 5 mM CaCl₂, and washed five times with 5 ml 5 mM CaCl₂. Radioactivity retained on the filter was counted with the scintillation cocktail ReadyProtein (Beckman Coulter K.K.) in a liquid scintillation counter (Beckman Coulter K.K.).

Determination of viability.
The method described by Iida et al. (1990) was followed. Aliquots (100 µl) of cultures (approx. 2x10⁶ cells ml^–1; at 30 °C) were taken and mixed with an equal volume of a solution containing 0.01 % (w/v) methylene blue and 2 % (w/v) sodium citrate. The mixture was sonicated briefly with a Sonifier (Branson model W-200P) to dissociate cell clumps. The number of methylene-blue-negative (viable) and -positive (non-viable) cells was determined by using a differential interference contrast microscope (Olympus model BX50). Viability was expressed as the percentage of the methylene-blue-negative cells in the total number of cells.

Statistics.
Data are given as means±SD. Statistical comparisons were made with Student's t test.

Kinetic characterization of Cch1–Mid1
Molecular cloning of the CCH1 gene has been unsuccessful for more than a decade since the discovery of the gene in the S. cerevisiae genome (Paidhungat & Garrett, 1997), but we have recently succeeded in cloning it, using a plasmid whose copy number is kept low in Escherichia coli (Iida et al., 2007). In the present study, we used the cloned gene to examine the in vivo properties of Cch1 produced from a strong promoter, TDH3p, on the plasmid pBC111 (Table 1). Western blotting has shown that the resulting plasmid pBCT-CCH1H ensures more than a 50-fold overproduction of Cch1 (Iida et al., 2007). Since Cch1 requires Mid1 to function as a Ca²⁺-influx system (Fischer et al., 1997; Paidhungat & Garrett, 1997; Iida et al., 2004; Muller et al., 2001), Mid1 was also designed to be produced from the same promoter on a different plasmid, YCplac33 (Table 1). Again, the plasmid thus constructed, YCpT-MID1, ensured more than 50-fold overproduction of Mid1 (data not shown). The resulting co-transformant which overexpresses both Cch1 and Mid1 (designated CCH1ox MID1ox) was used to assess the properties of the putative Cch1–Mid1 Ca²⁺ channel. This co-overexpression system has the advantage of reducing the perturbation effect caused by unidentified Ca²⁺-influx systems of S. cerevisiae cells on Ca²⁺ permeation through Cch1–Mid1.

Using cells of CCH1ox MID1ox and the control strains, we first estimated the affinity for Ca²⁺ (K_m), and the maximum Ca²⁺-accumulation rate (V_max) of the putative Cch1–Mid1 channel. To do this, the cells were grown in SD.Ca100, which discloses the Ca²⁺-uptake activity of the putative Cch1–Mid1 channel (Muller et al., 2001). Ca²⁺-accumulation assays were conducted at 30 °C for 2 h, which was within the linear range for the cells assayed. The concentration of ⁴⁵CaCl₂ was varied within single experiments, and Ca²⁺ accumulation in the cells was measured. As shown in Fig. 2, CCH1ox MID1ox, CCH1–MID1 (wild-type) and cch1Δ mid1Δ cells (lacking both Cch1 and Mid1) displayed Michaelis–Menten kinetics. Table 2 shows the apparent K_m and V_max values deduced from the Hanes–Woolf plots for each strain. The K_m value of CCH1ox MID1ox (12 µM) was approximately 2.6-fold lower than that of CCH1–MID1 (31 µM), and 24-fold lower than that of cch1Δ mid1Δ (290 µM). The V_max/K_m ratio, which measures the efficiency of Ca²⁺-channel function, and thus can be used to compare the channel function of different strains, was approximately 12 times greater in CCH1ox MID1ox than in CCH1–MID1, and 290 times greater in CCH1ox MID1ox than in cch1Δ mid1Δ. Therefore, it was concluded that the contribution of putative Ca²⁺ channels and/or transporters other than Cch1–Mid1 was negligible in CCH1ox MID1ox cells under the conditions tested.

(15K): Fig. 2. Michaelis–Menten plot of Ca2+-accumulation rate versus CaCl2 concentration. Exponentially growing cells cultured in SD.Ca100 medium were harvested, washed twice with SD-Ca medium, and resuspended in SD-Ca medium containing varying concentrations (4.85, 9.7, 48.5, 97, 194 and 291 µM) of 45CaCl2 (1.81 kBq nmol–1). The cell suspension was incubated at 30 °C, with shaking. The Ca2+-accumulation rate was determined for the initial 2 h after the start of the incubation, as described in Methods. The data were fitted by nonlinear regression of Michaelis–Menten curves. The mean of three independent experiments (±SD) is shown for each point. , CCH1ox MID1ox cells (strain H207 bearing pBCT-CCH1H and YCpT-MID1); , CCH1–MID1 cells (strain H207 bearing empty vectors pBC111 and YCplac33); , cch1Δ mid1Δ cells (strain H315 bearing empty vectors pBC111 and YCplac33).

Table 2. Km and Vmax of various strains with altered copy numbers of CCH1 and MID1 The results presented here are mean (±SD) values based on the Hanes–Woolf plots (Segel, 1976) of the data in Fig. 2.

Sensitivity of Cch1–Mid1 to ion-channel blockers
To address the question of whether Cch1 is a functional homologue of mammalian VGCCs, we employed a pharmacological approach using channel blockers typical of L-, P/Q-, N-, R- or T-types of VGCCs. Since the yeast cell wall is known to provide a barrier that inhibits reagents accessing the plasma membrane (Schindler & Davies, 1975; Gorenstein et al., 1978), the channel blockers were applied in concentrations that were almost at the upper limit of their solubility in SD.Ca100 medium. CCH1ox MID1ox cells were grown overnight to exponential growth phase in SD.Ca100 medium, and further incubated for 2 h with one of the Ca²⁺-channel blockers and ⁴⁵CaCl₂, and then Ca²⁺ accumulation was measured. As shown in Fig. 3, among the Ca²⁺-channel blockers tested, 200 µM nifedipine (L-type) and 1 mM verapamil (L-type) blocked Ca²⁺ accumulation to levels of 55 % and 71 %, respectively, while ω-agatoxin IVA (P/Q-type), ω-conotoxin GIVA (N-type), SNX-482 (R-type) and amiloride (T-type) did not affect Ca²⁺ accumulation, even at concentrations near their solubility limit in SD.Ca100 medium. Those blockers did not affect cell viability during the course of the experiments at any concentration used (data not shown). Based on the data presented in Fig. 3, the K_i values of nifedipine and verapamil on Cch1 were estimated from Dixon plots (Segel, 1976) to be 248 and 1750 µM, respectively (Table 3). These values were much higher than those for nifedipine and verapamil on mammalian VGCC α₁ subunits (Table 3).

(24K): Fig. 3. Effect of VGCC blockers on Cch1 activity. Exponentially growing CCH1ox MID1ox cells grown in SD.Ca100 medium at 30 °C were incubated for 2 h with 100 µM 45CaCl2 (185 kBq ml–1; 1.81 kBq nmol–1) and an appropriate blocker, and Ca2+ accumulation was determined as described in Methods. The mean of three independent experiments (±SD) is shown for each point. L, P/Q, N, R and T represent the types of VGCC based upon Ca2+ current. P versus control cells: *P=0.00369, **P=0.0103, ***P=0.00008, ****P=0.0036, *****P=0.0010.

Table 3. Ki values for nifedipine and verapamil on Cch1 The values were estimated based on the Dixon plot for non-competitive blockers (Segel, 1976). Examples of Ki values for the blockers on mammalian VGCC α1 subunits are also shown.

Surprisingly, the third L-type blocker diltiazem enhanced Ca²⁺ accumulation, depending upon its concentration tested, although measurement errors were large (Fig. 4a). At the highest concentration tested (12 mM), diltiazem enhanced Ca²⁺ accumulation 1.8-fold. At the same concentration, this compound resulted in a decrease in cell viability to 68 % (Fig. 4b). Microscopic observation indicated that diltiazem-treated cells were morphologically indistinguishable from untreated cells, suggesting that the increase in Ca²⁺ accumulation was not due to cell deterioration.

(29K): Fig. 4. Effect of the L-type VGCC blocker diltiazem on Cch1 activity (a) and cell viability (b). The experimental conditions used here were the same as those described in the legend to Fig. 3. Cell viability was determined as described in Methods. The mean of three independent experiments (±SD) is shown for each measurement. The cch1Δ mid1Δ strain was H315 bearing empty vectors pBC111 and YCplac33. The CCH1ox mid1Δ strain was H315 bearing pBCT-CCH1H and YCplac33. The cch1Δ MID1ox strain was H315 bearing pBC111 and YCpT-MID1. P versus control cells: *P=0.005, **P=0.0478, ***P=0.0005, ****P=0.0118, *****P=0.0068.

To examine whether the effects of diltiazem observed were mediated through Cch1–Mid1, Ca²⁺ accumulation and viability were measured for this compound in cch1Δ mid1Δ, CCH1ox mid1Δ and cch1Δ MID1ox strains. The results showed that Ca²⁺ accumulation was not increased by diltiazem in the three strains (Fig. 4a), indicating that the diltiazem-induced increase in Ca²⁺ accumulation in CCH1ox MID1ox cells was mediated through Cch1–Mid1. Viability assays indicated that this compound significantly decreased the viability of cch1Δ mid1Δ cells to 82 % at 12 mM (Fig. 4b). This observation suggests that the decrease in the viability of CCH1ox MID1ox cells cannot be ascribed solely to the increase in Ca²⁺ accumulation which potentially disturbs Ca²⁺ homeostasis.

The VGCC agonist (S)-(–)-Bay K8644 (5–100 µM) did not affect Ca²⁺ accumulation (data not shown). Since the secondary structure of VGCC α₁ subunits and Cch1 is similar to that of Na⁺ channels and six-transmembrane K⁺ channels (Catterall, 1995), we examined the effect of the Na⁺-channel blocker TTX, and the K⁺-channel blocker TEA, on Ca²⁺ accumulation, and found that the blockers had no effect on Ca²⁺ accumulation (10–10 000 nM TTX; 1–100 mM TEA) (data not shown).

The results in the present study, especially those for the effect of nifedipine and verapamil, suggest that Cch1 is pharmacologically similar to L-type VGCCs. However, this suggestion is not clear-cut because another L-type blocker, diltiazem, enhanced Ca²⁺ accumulation unexpectedly (Fig. 4a). A possible explanation for this unexpected observation is that diltiazem may specifically bind to Cch1 to activate Ca²⁺ entry, instead of blocking it. The second possibility is that diltiazem, at high concentrations, may be a stress for yeast cells, and that it indirectly stimulates the Ca²⁺-channel activity of Cch1, because, at 12 mM, this compound kills about 32 % of CCH1ox MID1ox cells and 18 % of cch1Δ mid1Δ cells (Fig. 4b). Several stresses, including endoplasmic reticulum stress caused by tunicamycin (an inhibitor of N-glycosylation) or dithiothreitol (an inhibitor of disulfide bond formation), hyperosmotic stress and alkaline stress, have been reported to activate the Cch1–Mid1 channel (Bonilla et al., 2002; Matsumoto et al., 2002; Viladevall et al., 2004). The third possibility is that diltiazem at high concentrations (7.5–12 mM) causes non-specific fluxes of ions without activating Cch1. It has been reported that diltiazem increases the cation and anion permeability of biological membranes, such as bovine photoreceptor membranes, at high concentrations (12.5–200 µM; Caretta et al., 1991). However, as mentioned above, our study showed that diltiazem, even at 12 mM, did not increase Ca²⁺ accumulation in cch1Δ mid1Δ cells (Fig. 4a), indicating that the diltiazem-induced increase in Ca²⁺ accumulation depends on the Cch1–Mid1 channel. Therefore, the third possibility seems unlikely.

It should be noted that the half-blocking concentrations of organic Ca²⁺-channel blockers, including nifedipine and verapamil, are in the 20 nM to 50 µM range, depending upon the type of vertebrate cells (Triggle, 1990; Hille, 1992; Catterall et al., 2005). These concentrations are much lower than those used for S. cerevisiae cells in this study. One reason for the necessity of high concentrations of the blockers could be the presence of the cell wall in S. cerevisiae cells. In the case of actinomycin D, which binds to DNA to inhibit RNA synthesis, it has been reported that spheroplasts lacking the cell wall are more sensitive to this antibiotic than intact cells of S. cerevisiae (Gorenstein et al., 1978). Since lytic enzymes that degrade the cell wall to generate spheroplasts contain proteases as well as endoglucanases (Scott & Schekman, 1980), proteins protruding from the plasma membrane, including Ca²⁺ channels, would be damaged during spheroplast preparation. Therefore, careful examination is necessary to assess Ca²⁺-uptake activity in spheroplasts.

Another reason for the need for high concentrations of nifedipine and verapamil (Fig. 3) could be a difference in the binding sites of these L-type blockers, in addition to the presence of the cell wall. Although the three L-type blockers used in this study are classified into the same type, their binding sites are not completely identical (Mitterdorfer et al., 1998; Striessnig, 1999). Nifedipine, verapamil and diltiazem are members of the chemically unrelated classes of L-type blockers dihydropyridines (DHPs), phenylalkylamines (PAAs) and benzothiazepines (BTZs), respectively. As shown in Fig. 5, the three classes bind to two or three of the transmembrane segments IIIS5, IIIS6 and IVS6 in mammalian L-type VGCCs, and the binding of each class is unique in each segment. Alignment of the corresponding transmembrane segments of Cch1 with a mammalian VGCC shows that the binding sites of Cch1 for the three classes are limited: there is no identical amino acid residue in IIIS5; there is one for DHPs, and two for PAAs in IIIS6; and one for DHPs in IVS6 (Fig. 5). This limitation may account for the low affinities of nifedipine and verapamil to Cch1.

(30K): Fig. 5. Amino acid sequences of the IIIS5, IIIS6 and IVS6 segments of a mammalian VGCC and Cch1. The upper sequences (written by Mitterdorfer et al., 1998) are according to a rat brain Cav1.2 α1 subunit (rbCII) (GenBank accession number M67515) and the lower sequences Cch1 (GenBank accession number Z73002). Italicized lower-case letters indicate DHP (d), PAA (p), and BTZ (b). Nifedipine is a member of the DHP group, verapamil is a member of the PAA group, and diltiazem is a member of the BTZ group. The position of these letters on the upper sequences shows the amino acid residues that contribute to blocker binding, as summarized for mammalian VGCCs (Mitterdorfer et al., 1998). Asterisks indicate identical amino acids, and colons indicate conserved amino acid substitutions determined according to the CLUSTAL W program (Chenna et al., 2003).

A BLAST search () with the amino acid sequence of the S. cerevisiae Cch1 found homologous proteins for over 30 fungal species, including pathogenic fungi. Modulation of fungal Cch1 function has been suggested to be useful in the development of cures for fungal diseases in humans caused by Candida albicans (Bonilla et al., 2002), Candida glabrata (Kaur et al., 2004), Cryptococcus neoformans (Liu et al., 2006) and Trichophyton rubrum (Yu et al., 2007), and for those in plants caused by Magnaporthe grisea (Zelter et al., 2004). This suggestion should facilitate further study on Cch1 and related Ca²⁺-signalling mechanisms, and expand it to other pathogenic fungi. Our study with diltiazem indicates that when assessing the effect of a blocker of interest on any fungal species, one should always measure Ca²⁺ accumulation or influx in terms of Ca²⁺ signalling and curing the fungal disease. We thank Dr Masami Yoshino for his comment on channel blockers, and Ms Yumiko Higashi for her secretarial work. This work was supported by a Grant-in-Aid for Scientific Research on Priority Areas (No. 15031212 to H. I), a Grant-in-Aid for Scientific Research B (No. 16370072 to H. I.), and a Grant-in-Aid for Scientific Research C (No. 20570158 to K. I.) from the Ministry of Education, Culture, Sports, Science and Technology in Japan, and a grant for the Frontiers of Membrane Protein Research (the Joint Project between the Institute for Protein Research, Osaka University and Okazaki Institute for Integrative Bioscience; to H. I.).

Edited by: J. M. Becker

Footnotes

^,†, Rika Goto¹ †These authors contributed equally to this work.