Acidic Calcium Stores of Saccharomyces cerevisiae

Abstract

Fungi and animals constitute sister kingdoms in the eukaryotic domain of life. The major classes of transporters, channels, sensors, and effectors that move and respond to calcium ions were already highly networked in the common ancestor of fungi and animals. Since that time, some key components of the network have been moved, altered, relocalized, lost, or duplicated in the fungal and animal lineages and at the same time some of the regulatory circuitry has been dramatically rewired. Today the calcium transport and signaling networks in fungi provide a fresh perspective on the scene that has emerged from studies of the network in animal cells. This review provides an overview of calcium signaling networks in fungi, particularly the model yeast Saccharomyces cerevisiae, with special attention to the dominant roles of acidic calcium stores in fungal cell physiology.

Calcium Signaling Pathways in the Fungal Cytoplasm

Dozens of genome sequencing projects on phylogenetically diverse fungal species have revealed a basic toolkit of Ca²⁺-binding proteins and Ca²⁺ pumps, exchangers, and channels and the remarkable conservation of the signaling network across the fungal kingdom [1]. The central Ca²⁺ sensor calmodulin can be easily spotted in the genomes of all sequenced fungi. Despite some lineage-specific drift in the amino acid sequence of yeast calmodulins, many targets of Ca²⁺/calmodulin are also well preserved throughout the kingdom. Examples include two families of serine/threonine protein kinases and a family of serine/threonine protein phosphatases known as calcineurin. As in animal cells, these kinases and phosphatases in the bakers yeast S. cerevisiae and in other fungi become activated upon binding of Ca²⁺/calmodulin to conserved sequences within their autoregulatory tails and displacement of autoinhibitory motifs from their active sites (reviewed in [2, 3]). Thus, the rise and fall of free Ca²⁺ concentrations in the cytoplasm can be directly sensed, decoded, and retransmitted to cellular targets through regulated protein phosphorylation and dephosphorylation. Myosins and other well-known targets of calmodulin have also been described in S. cerevisiae and many other fungi. Additionally, the genomes contain a spectrum of conserved proteins that bear EF-hand and C2 domains, which bind Ca²⁺ and may respond to fluctuation Ca²⁺ concentrations in their microenvironments. The emerging picture from these accounts is one where a multitude of Ca²⁺-responsive regulatory pathways exist in fungal cells.

One of the best-studied Ca²⁺-responsive signaling pathways in fungi (see Fig. 1) involves the calcineurin-dependent dephosphorylation of Crz1, a zinc-finger transcription factor first described in S. cerevisiae [4, 5]. Crz1 is not related to the NFAT family of calcineurin-sensitive transcription factors that famously control many calcineurin-dependent processes in mammals. Similar to the NFAT story, activated calcineurin binds to canonical PxIxIT-like motifs and dephosphorylates several residues in Crz1, resulting in a conformational change that hides a nuclear export signal and exposes a nuclear localization signal (respectively recognized by the β-importins Msn5 and Nmd5) [6–8]. After transport of dephosphorylated Crz1 into the nucleus, the Zn-finger domain binds specific DNA sequences (termed CDREs for Crz1-Dependent Response Elements) present within the promoter regions of target genes and significantly increases expression above the resting basal level [6–12]. Inducible targets of Crz1 include CRZ1 that encodes Crz1 [5], RCN1 and RCN2 that encode positive and negative regulators of calcineurin [13, 14], CMK2 that encodes a Ca²⁺/calmodulin-dependent protein kinase [15], PMC1, PMR1, and ENA1 that encode several P-type cation pumps responsible for efflux of Ca²⁺, Mn²⁺, Na⁺, and Li⁺ [5] and 60 to 100 other genes that control other processes [16, 17]. Mutants of S. cerevisiae that lack Crz1 are hypersensitive to high environmental concentrations of these cations as a consequence of failed induction of the cation pumps and they exhibit other phenotypes that can be attributed to failed induction of other targets. Calcineurin-deficient mutants of S. cerevisiae exhibit an even larger set of phenotypes due to defects in the regulation of phosphoproteins other than Crz1. For example, calcineurin-dependent dephosphorylation of Hph1/Hph2 and Slm1/Slm2 protein pairs can alter sensitivity to high pH medium and alter trafficking of secretory and endocytic cargo proteins [18, 19]. Calcineurin-dependent feedback regulation of Ca²⁺ channels and Ca²⁺ transporters will be discussed more fully later in this review. An overview of the known and suspected calcineurin targets in S. cerevisiae is given in Figure 1.

Figure 1. Ca²⁺ transport and signaling in S. cerevisiae cells exposed to high environmental Ca²⁺.

Ca²⁺ enters the cell through unknown pathways and elevates [Ca²⁺]_CYT, which results in activation of calmodulin, calcineurin, and a transcription factor (Crz1) that induces numerous genes. Activated calmodulin and calcineurin also regulate other cellular factors such as protein kinases (Cmk1, Cmk2), feedback regulators of calcineurin (Rcn1, Rcn2), membrane trafficking factors (Slm1, Slm2, Hph1, Hph2), and a vacuolar Ca²⁺/H+ exchanger (Vcx1). Ca²⁺ pumps in the vacuole (Pmc1) and the Golgi complex (Pmr1) become up-regulated and help to lower [Ca²⁺]_CYT to non-toxic levels.

In vegetatively growing S. cerevisiae cells, Crz1 is fully phosphorylated, localized to the cytoplasm, and transcriptionally inactive [16, 20]. The simple inference from all these findings is that cytosolic free Ca²⁺ concentrations are maintained at low non-signaling levels in vegetatively growing S. cerevisiae cells and that [Ca²⁺]_CYT levels rise to levels capable of activating calcineurin and Crz1 in response to specific stimuli or stresses. Thus, cells of S. cerevisiae and its relatives probably behave very much like mammalian cells in their ability to dynamically control [Ca²⁺]_CYT and the downstream signaling pathways. A series of conserved Ca²⁺ pumps, exchangers, transporters, and channels accomplish that important task in fungal and animal cells. They also control intracellular pools of Ca²⁺ that have many important functions.

Ca²⁺ in secretory organelles and store-operated Ca²⁺ influx

The increasingly acidic lumens of the nuclear envelope, endoplasmic reticulum, Golgi complex, and trans-Golgi/endosomal network contain an array of conserved Ca²⁺ dependent enzymes that are involved in various aspects of protein secretion (reviewed in [21]). For instance, S. cerevisiae retains homologs of BiP, calnexin, UDP-glucose-glucosyltransferase, glucosidase II, and ERGIC-53 (termed Kar2, Cne1, Kre6, Rot2, and Emp46/47, respectively) in its nuclear envelope and endoplasmic reticulum [22–24]. In S. cerevisiae, many of these enzymes no longer retain an ability to bind Ca²⁺ and can function independently of Ca²⁺. Such adaptations are probably related to the loss of SERCA-family Ca²⁺ pumps that normally supply the endoplasmic reticulum with sufficient Ca²⁺ for secretory functions. SERCA was probably present in the common ancestor of fungi and animals and subsequently lost several different times independently in the evolution of Ascomycetes (moulds, yeasts), Basidiomycetes (mushrooms, smuts, rusts), and other fungal phyla. A SERCA-family Ca²⁺ pump is expressed in the endoplasmic reticulum of the mould Neurospora crassa [25] but this enzyme has not yet been characterized biochemically or genetically. Though S. cerevisiae and other budding yeasts do not retain a SERCA-family Ca²⁺ pump and may have secretory machinery with reduced Ca²⁺ dependence, Ca²⁺ starvation of S. cerevisiae cells still causes activation of the so-called Unfolded Protein Response (UPR) signaling pathway that emanates from the endoplasmic reticulum upon its accumulation of misfolded or unassembled secretory proteins [26]. Inhibitors of SERCA elicit similar UPR responses in animal cells. Ca²⁺ starvation can also decrease the retention of foreign proteins expressed in yeasts [27] and thereby enhance the yield of recombinant protein preparations. Thus, luminal Ca²⁺ performs important secretory functions in the fungal endoplasmic reticulum, even in S. cerevisiae where both the supply and the demand seem greatly diminished.

The endoplasmic reticulum of S. cerevisiae concentrates Ca²⁺ approximately 100-fold relative to the cytoplasm largely through the Ca²⁺ transport activity of Pmr1 [28], the prototypical member of the SPCA-family of Ca²⁺/Mn²⁺ pumps that are widely distributed among fungi, animals, and other eukaryotic kingdoms. Pmr1 localizes primarily to the Golgi complex of S. cerevisiae [27, 29], like its homologs in mammals [30], and therefore supplies Ca²⁺ and Mn²⁺ to the endoplasmic reticulum during its early biogenesis or through vesicle-mediated trafficking in the retrograde direction from the Golgi complex. Mutants of S. cerevisiae that lack the Pmr1 exhibit a range of secretion defects that can be largely attributed to underperformance of late secretory pathway enzymes. A homolog of the Ca²⁺-dependent pro-protein convertases or furins (termed Kex2) is highly dependent on Pmr1 function [29, 31]. A Ca²⁺-dependent lectin-like protein involved in sorting of specific cargo proteins to the lysosome-like vacuole (termed Vps10) also depends on Pmr1 for proper function [32]. The normal retention of foreign secretory proteins that are expressed heterologously in S. cerevisiae also depends on Pmr1 [27, 33]. These defects of Pmr1-deficient mutants can be suppressed by elevating Ca²⁺ salts in the culture medium or by expressing SERCA in the endoplasmic reticulum, suggesting they are specifically a consequence of luminal Ca²⁺ insufficiency [32]. On the other hand, defects in N-glycosylation and O-glycosylation of secretory cargo are attributable to Mn²⁺ insufficiency in the Golgi complex [32]. All these findings are consistent with the hypothesis that the SPCA-family pump Pmr1 supplies the majority of the Ca²⁺ and Mn²⁺ that is crucial for normal processing functions in both the endoplasmic reticulum and Golgi complex of S. cerevisiae.

Many cell types in animals are known to employ store-operated Ca²⁺ entry mechanisms in which Ca²⁺ influx channels in the plasma membrane become activated in response to depletion of Ca²⁺ from the lumen of the endoplasmic reticulum. In animals, depletion of Ca²⁺ stores occurs physiologically through the repetitive activation of IP3-receptors or other Ca²⁺ channels located in the endoplasmic reticulum. IP3-receptors are not evident in any of the fungal genomes sequenced to date though they are clearly present in animals and amoebas such as Dictyostelium discoideum, which probably diverged before the fungal/animal bifurcation. Moreover, the CRAC/OraI Ca²⁺ influx channels and the Stim1 Ca²⁺ sensors in the endoplasmic reticulum that are both essential for store-operated Ca²⁺ influx in animals are completely absent in the known fungal genomes. In spite of these key differences, the Pmr1-deficient mutants of S. cerevisiae exhibit a much higher rate of Ca²⁺ influx, elevated [Ca²⁺]_CYT, and activated calcineurin relative to wild-type cells [34], analogous to the situation in animal cells when luminal Ca²⁺ in the endoplasmic reticulum is depleted.

The mechanism of store-operated Ca²⁺ influx in S. cerevisiae has been partially unraveled (see Fig. 2). One consequence of Pmr1-deficiency is the up-regulation and mislocalization of Pmc1 [35], a PMCA-type Ca²⁺ pump located in the vacuole that can partially suppress the secretory defects of Pmr1-deficient mutants [36]. The up-regulation of Pmc1 is a consequence of increased Ca²⁺ influx through a high-affinity Ca²⁺ influx system (termed HACS) that activates calcineurin and Crz1 [34]. HACS requires three interacting proteins: Cch1, a homolog of the catalytic α-subunit of voltage-gated Ca²⁺ channels in animals [37, 38], Mid1, which bears some similarity to the α2δ-subunit of voltage-gated Ca²⁺ channels in animals [39, 40], and Ecm7, a homolog of regulatory γ-subunits of voltage-gated Ca²⁺ channels in animals [41]. The Cch1 subunit of HACS becomes phosphorylated in response to depletion of secretory Ca²⁺ stores in S. cerevisiae by a process that involves activation of a MAP kinase (termed Slt2) and a cascade of upstream protein kinases that couple to rho-type GTPase activation [42]. Though Slt2 is required for HACS activation in these conditions [42], it is not yet clear whether the phosphorylation of Cch1 is necessary or sufficient for HACS activation. Nevertheless, the store-operated Ca²⁺ entry mechanism in S. cerevisiae is strikingly different from the STIM/Orai mechanism of animal cells.

Figure 2. Ca²⁺ transport and signaling in S. cerevisiae cells exposed to membrane-active toxins and antifungal agents.

Toxins (e.g. tunicamycin) and antifungals (e.g. fluconazole) that disrupt secretory protein biogenesis or membrane biogenesis lead to membrane stresses and activation of a MAP kinase (Slt2) that activates a high-affinity Ca²⁺ influx system (HACS; Cch1, Mid1, Ecm7). Downstream signaling pathways help to alleviate the stress by boosting secretory performance. Mutants that lack the secretory pathway Ca²⁺ pump (Pmr1) or other factors important in vesicle-mediated trafficking pathways (e.g. Kex2) result in membrane stresses that constitutively activate Slt2 and HACS. Cells containing calcineurin survive these stresses whereas cells lacking calcineurin die by poorly understood necrosis-like mechanism.

The HACS of S. cerevisiae can also be activated by other deficiencies in the secretory pathway that do not result in Ca²⁺ depletion. Misfolded proteins in the endoplasmic reticulum can activate Slt2 and HACS by a process that is independent of signaling through the UPR signaling pathway [26]. Defects in an ER-localized phospholipid flipase also promote Ca²⁺ alterations [43]. Defects in protein trafficking to the vacuole similarly activate HACS [44]. A recent screen of all the viable gene knockout mutants of S. cerevisiae demonstrated HACS can be strongly activated by deficiencies in several dozen different enzymes of the vacuole protein sorting and protein secretion pathways, including the Kex2-deficiency that arises in Pmr1-deficient cells [41]. Thus, a wide variety of stresses in the vesicle mediated trafficking system seem to be capable of activating HACS in S. cerevisiae. Depletion of secretory Ca²⁺ stores is but one of many ways to generate a HACS-activating stress. The induction of Ca²⁺ pumps (Pmc1 and Pmr1) by Crz1 and calcineurin is probably one of many compensatory responses controlled by HACS.

Membrane stresses may be common in fungal and animal cells as a consequence of interactions with the environment and toxins that attack them. Microbes often secrete compounds that are toxic to eukaryotic competitors and predators through their effects on key enzymes of secretory pathway. One such compound secreted by Bacillus lysosuperficus is tunicamycin, which enters the eukaryotic cell and potently inhibits an essential enzyme necessary for N-glycan biogenesis [45]. The resulting blockade in protein N-glycosylation in the endoplasmic reticulum of S. cerevisiae causes the accumulation of misfolded and unassembled secretory proteins. While the UPR pathway becomes rapidly activated in response to tunicamycin exposure, activation of the Slt2 MAP kinase cascade, HACS, and calcineurin proceeds much more slowly [26]. The faster response induces molecular chaperones and other factors that repair the damage and help the S. cerevisiae cells adapt and recover from a brief exposure to the toxin. But what are the benefits of the slower response pathway involving Ca²⁺ ? Those benefits are not completely clear but they seem to be essential to survival of the stressed cell. Mutants of S. cerevisiae and other yeasts that lack HACS or calcineurin rapidly die upon exposure to tunicamycin [15]. In contrast, mutants that lack the UPR pathway remain alive indefinitely in the presence of tunicamycin but cannot repair enough of the damage to proliferate once the compound has been removed [26, 46]. Thus, the calcium signaling pathways of S. cerevisiae may be a life-saving countermeasure against microbial assaults. The invention of novel toxins that target calcineurin, such as FK506 from Streptomyces tsukubaensis [47], may represent a counter-countermeasure in a chemical arms race between bacterial and fungal competitors.

Common antifungal antibiotics use to combat fungal infections in humans seem to mimic the effects of the natural toxins described above. For example, the antifungals that target sterol biosynthetic enzymes in the endoplasmic reticulum of fungal pathogens (fluconazole, miconazole, terbinafine, fenpropimorph, and several others) probably activate essential functions of HACS and calcineurin because the co-administration of these agents together with either FK506 or cyclosporine (another potent inhibitor of calcineurin) usually results in potent fungicidal effects rather than simple fungistatic effects when used alone [26, 48–54]. Inhibitors of sterol biosynthesis are not known to activate or inhibit the UPR, so they probably trigger membrane stresses akin to those produced by tunicamycin or mutations in the non-essential genes of the secretory pathway described earlier. The fungicidal synergism of calcineurin inhibitors and sterol biosynthesis inhibitors is broadly conserved in diverse fungal pathogens (reviewed in [55, 56]), which suggests that calcineurin activation in response to membrane stresses may be a widespread occurrence in the fungal kingdom.

To summarize the key differences between fungi and animals with regard to Ca²⁺ in the secretory pathway, animals possess a unique store-operated Ca²⁺ entry mechanism that replenishes the endoplasmic reticulum and other secretory compartments after IP3 signals their depletion. Fungi completely lack the IP3 receptors, luminal Ca²⁺ sensors, and Ca²⁺ influx channels that constitute the core of this system. On the other hand, depletion of Ca²⁺ from secretory organelles in fungi can produce membrane stress and activation of a MAP kinase cascade that promotes Ca²⁺ influx through HACS-type channels. A wide array of toxins, mutations, and antifungal drugs may create similar types of membrane stress and similar effects on HACS. The resulting rise in [Ca²⁺]_CYT not only resupplies the secretory organelles but signals through calmodulin, calcineurin, and Crz1 to regulate processes that may restore luminal Ca²⁺, mitigate membrane stress, and promote cell survival.

The vacuole and Ca²⁺ sequestration/release mechanisms

Organelles of the secretory pathway in fungi, like those of mammalian cells, are expected to become increasingly acidic as a consequence of increasing activity of H⁺-pumping V-ATPases. Fungi also contain very acidic organelles termed vacuoles that are acidified by V-ATPases and, like lysosomes in animals, are constructed using well-conserved vesicle-mediated trafficking pathways [57]. The bread mould Neurospora crassa contains several different types of vacuoles that can be distinguished by morphological and compositional criteria [25]. It is not yet clear if those vacuoles represent different intermediates in a pathway of vacuole maturation or different endpoints. The yeast S. cerevisiae contains one to several vacuoles that can undergo cycles of invagination, fusion, and fission in response to cell cycle and environmental triggers. A remarkable feature of fungal vacuoles is their ability to sequester large amounts of Ca²⁺ and release it in response to particular stimuli (see Fig. 3).

Figure 3. Ca²⁺ release from acidic vacuoles in S. cerevisiae cells.

The vacuolar TRPC-family ion channel (Yvc1) becomes rapidly activated in response to hydrogen peroxide (H2O2) or tert-butylhydroperoxide (tBOOH). Hypertonic shock (e.g. high salt or sugar in the environment) triggers a slower activation of Yvc1, possibly through stretching of the vacuole membrane. Both stimuli result in transient elevation of [Ca²⁺]_CYT that is reversed by the actions of vacuolar Ca²⁺ transporters (Pmc1, Vcx1). The activation of downstream signaling pathways has unknown functions.

In the yeast S. cerevisiae, more than 90% of total cell-associated Ca²⁺ is immobilized in the vacuole largely in complexes with inorganic polyphosphate [58]. The vacuole probably receives a small amount of Ca²⁺ through the fusion of vesicles derived from the Golgi complex that received its Ca²⁺ directly from SPCAs or indirectly from the environment through endocytosis. A large majority of vacuolar Ca²⁺ comes directly from the cytoplasm through the action of Ca²⁺ pumps and Ca²⁺/H⁺ exchangers that are specifically localized to the vacuole membrane.

The vacuolar Ca²⁺ pump of S. cerevisiae (termed Pmc1) is the major contributor of vacuolar Ca²⁺ [36]. Pmc1 is conserved in nearly all fungi and is closely related to the PMCA-family of plasma membrane Ca²⁺ ATPases found in animals, plants, and other kingdoms of eukaryotes. However, Pmc1 has not been observed in the plasma membrane of S. cerevisiae and only has been observed in the vacuole membrane, though Pmr1-deficient mutants induce Pmc1 expression and accumulate some fraction in earlier secretory compartments [35]. Overexpression of Pmc1 can partially suppress the secretory defects associated with Pmr1-deficient mutants [36], suggesting some ability to transport Ca²⁺ while en route to the vacuole. The simultaneous loss of both Pmc1 and Pmr1 is lethal to S. cerevisiae, though lethality can be avoided in conditions that increase activity of the vacuolar Ca²⁺/H⁺ exchanger (Vcx1) [59]. The loss of Pmc1 alone is not lethal in ordinary growth conditions, though vacuoles contain only ~10% of the normal level of Ca²⁺. Pmc1-deficient mutants of S. cerevisiae are strongly hypersensitive to supplemental Ca²⁺ salts in the culture medium. Wild-type S. cerevisiae respond to high Ca²⁺ environments with a large induction of Pmc1 expression and a small induction of Pmr1 expression, both mediated by the activation of calcineurin and Crz1 [4, 5]. When taken altogether, these findings suggest that Pmc1 and the vacuole play a major role in the detoxification of cytoplasmic Ca²⁺ and perhaps a minor role in supplying the secretory pathway, where Pmr1 normally predominates.

A vacuolar Ca²⁺/H⁺ exchanger of S. cerevisiae (termed Vcx1) was identified genetically based on its ability to confer tolerance to high Ca²⁺ and high Mn²⁺ salts in the culture medium when the enzyme is overexpressed [59, 60]. Earlier experiments had predicted that Vcx1 was responsible for most Ca²⁺ transport into the vacuole in vivo in normal growth conditions [58]. However, Vcx1-deficient knockout mutants exhibited wild-type levels of vacuolar Ca²⁺ and wild-type tolerance to supplemental Ca²⁺ salts in the culture medium [59, 60], in striking contrast to the aforementioned Pmc1-deficient mutants. On the other hand, the contributions of Vcx1 to vacuolar Ca²⁺ uptake and to Ca²⁺ tolerance are markedly increased when calcineurin is inactivated by either mutations or inhibitors [59]. These findings suggest that calcineurin effectively blocks Vcx1 activity in vivo. Consistent with this view, several different single amino-acid substitutions in Vcx1 have been recovered that greatly increase its activity even when calcineurin is functioning [59, 61, 62]. Such hyperactive variants of Vcx1 can often transport Mn²⁺ in addition to Ca²⁺ and compete with Pmr1 in the Golgi complex for substrates, thus depriving the secretory pathway of essential minerals and triggering membrane stress described in the previous section [59]. The mechanism by which calcineurin inhibits Vcx1 function is not yet known. Crz1 and the other known targets of calcineurin are not required for the calcineurin-dependent inhibition of Vcx1. Vcx1 abundance and gel mobility also do not change in a calcineurin-dependent fashion. Therefore the regulation of Vcx1 may be indirect, for example through a calcineurin-dependent inhibition of the V-ATPase or some other factor that impinges on Vcx1 but not Pmc1. It will be important to define the molecular basis of this interaction because Vcx1 also inhibits calcineurin by removing Ca²⁺ from the cytoplasm that is necessary for calcineurin activation by calmodulin. Thus, Vcx1 and calcineurin form a double-negative feedback interaction. Double-negative feedback loops are often capable of producing bi-stability, a phenomenon where the network switches rapidly between two stable states but does not significantly populate the intermediate states [63, 64]. For instance, the Vcx1-on/calcineurin-off state may rapidly switch to the Vcx1-off/calcineurin-on state, and vice versa, which might then contribute to the dramatic “spikes” of [Ca²⁺]_CYT elevation and “bursts” of Crz1 localization to the nucleus that have been recently observed in S. cerevisiae cells [20]. For several minutes following exposure to high Ca²⁺ salts, Vcx1 remains very effective at lowering [Ca²⁺]_CYT [65], but what happens after calcineurin activation has not been thoroughly investigated. Obviously, more work needs to be completed before the roles of Vcx1 in Ca²⁺ homeostasis and signaling are fully understood.

Vcx1 is a member of the CAX family of Ca²⁺/H⁺ exchangers, which are not found in animals but are expressed widely in plants and many other eukaryotes [66]. Vcx1 also retains some weak Na⁺/H⁺ exchange activity, at least in vitro [67]. The VNX family of cation exchangers (formerly called type II CAX exchangers) from S. cerevisiae and the zebrafish Danio rerio function as Na⁺/H⁺ exchangers in the vacuole membrane [68, 69]. Fungi lack members of NCX and NCKX families of Ca²⁺/Na⁺ exchangers that have been well studied in animal cells [70]. Fungi also express members of the CCX family of cation/cation exchangers that are expressed in animals, but very poorly characterized. The sole CCX protein of mammals (termed NCLX or NCKX6) functions as a Ca²⁺/Na⁺ exchanger in the plasma membrane and mitochondria [71–73]. No fungal CCX proteins have been characterized to date, so their potential contributions to Ca²⁺ efflux, storage, and signaling are wholly unknown at present.

In S. cerevisiae cells grown in standard conditions, vacuolar free Ca²⁺ concentration ([Ca²⁺]_VAC) has been estimated at ~30 μM whereas total vacuolar Ca²⁺ has been estimated at ~3 mM [58], suggesting that 99% of vacuolar Ca²⁺ is buffered by inorganic polyphosphate. In spite of complete releasability of vacuolar Ca²⁺ by the ionophore A23187, little or no vacuolar Ca²⁺ is released into the cytoplasm or the culture medium when S. cerevisiae cells are cultivated in standard culture medium [36, 58]. Therefore, very little Ca²⁺ is released from the vacuole during proliferation in normal conditions. The possibility that Ca²⁺ stored in the vacuole can be released and reutilized during Ca²⁺ starvation has not been explored. In conditions of Mg²⁺ starvation, Ca²⁺ influx via unknown transporters in the plasma membrane is dramatically increased [74] and the resulting increases in vacuolar Ca²⁺ probably serve to displace Mg²⁺ from binding sites on inorganic polyphosphate and promote Mg²⁺ reutilization elsewhere in the cell [75]. Vacuolar H⁺ probably affects this process too. The possibility that Vcx1 releases Ca²⁺ from the vacuole in some conditions also has not been explored fully. Vcx1 and the other cation/cation exchangers are generally reversible if the substrate concentrations are favorable on both sides of the membrane. A sudden loss of vacuole acidity (or uncoupling) will not only prevent movement of Ca²⁺ through Vcx1 into the vacuole but will potentially increase the reverse-mode transport activity of Vcx1, effectively allowing Ca²⁺ release and reutilization [76]. To date, the clearest examples of regulated Ca²⁺ release from the vacuole occurs via via TRPC-family ion channels (termed Yvc1).

Yvc1 is well conserved among fungi and the family is most closely related to the TRPC-family of Ca²⁺ channels that are well characterized in animals [77]. Yvc1 is localized almost exclusively to the vacuole membrane of S. cerevisiae [78]. Yvc1-dependent Ca²⁺ release occurs within ~2 seconds of injecting hydrogen peroxide or tert-butylhydroperoxide into the culture medium [79]. These compounds are membrane permeable oxidants that, over time, cause production of reactive oxygen species and a range of oxidative damage, in addition to compensatory responses. Yvc1 is detrimental to S. cerevisiae cell growth in the presence of hydrogen peroxide or tert-butylhydroperoxide [79], which suggests that the release of vacuolar Ca²⁺ or other cations is somehow toxic in the presence of these oxidants.

Yvc1-dependent Ca²⁺ release also occurs with much slower and more prolonged kinetics after exposure of S. cerevisiae cells to hypertonic conditions such as high salinity and high sugar [78]. A broad survey of genes that alter Yvc1 function in S. cerevisiae did not reveal any candidates for direct regulators of Yvc1 [80]. The screen revealed instead a broad correlation between vacuolar Ca²⁺ content and the amount of Ca²⁺ released upon hyperosmotic shock. However, a recent study identified Fab1, a PI(3)P 5-kinase that synthesizes phosphatidylinositol-3,5-bisphosphate, as essential for Yvc1 activation in response to hyperosmotic shock [81]. Electrophysiological studies of Yvc1 in isolated vacuoles show that channel gating can be activated by membrane stretch, suggesting Yvc1 is directly mechanosensitive [82]. Extensive electrophysiological characterizations also suggest activation by reducing agents and an ability to pass Na⁺ and K⁺ in vitro in addition to Ca²⁺ [77, 82–85]. These properties suggest that Yvc1 may be capable of rapidly releasing cationic osmolytes into the cytoplasm upon hypertonic shock, which may provide temporary relief against cytoplasmic dehydration and osmotic imbalance. This prediction has not yet been tested with viability experiments, so it remains possible that Yvc1 activation is neutral or even harmful in these conditions. Clearly, much remains to be learned about the physiological roles of Yvc1 in S. cerevisiae and other fungi.

An interesting, yet counterintuitive, property of Yvc1 is its dependence on cytosolic Ca²⁺ for maximal gating (reviewed in [86]). This finding may indicate a role for auto-activation in Yvc1 physiology by a process commonly known as Ca²⁺-induced Ca²⁺ release. In animals, Ca²⁺-induced Ca²⁺ release contributes to the coordinated release of Ca²⁺ from the endoplasmic or sarcoplasmic reticulum for the purpose of generating coherent waves of [Ca²⁺]_CYT elevation. The existence of this phenomenon in S. cerevisiae is surprising because of the well-established role of the vacuole in detoxifying Ca²⁺ after exposure to high Ca²⁺ salts in the environment. If Yvc1 becomes activated by [Ca²⁺]_CYT elevation and remains active in the vacuole membrane for long periods of time, it is difficult to imagine how Pmc1 and Vcx1 can effectively diminish and detoxify [Ca²⁺]_CYT. Strains of S. cerevisiae that lack Yvc1 are not detectably hypersensitive to high Ca²⁺ environments and do not exhibit an increased activation of calcineurin. However, strains that overexpress Yvc1 exhibit hypersensitivity to environmental Ca²⁺ [78] but the potential deleterious effects of Na⁺ and K⁺ release have not been discriminated from Ca²⁺ release in this instance. Thus, similar to the confusing state of affairs with Vcx1, the potential feedback regulation of Yvc1 is both fascinating and crucial for understanding global Ca²⁺ homeostasis and signaling in fungi.

Ca²⁺ has been shown to leak from purified vacuoles at a defined step in the process of homotypic (vacuole to vacuole) fusion [87–90], which is one of several processes that dynamically synthesize, reshape, and redistribute vacuoles in S. cerevisiae. Originally this Ca²⁺ leak was thought to be important for triggering the formation of protein complexes and fusion pores near the end of the homotypic fusion process through effects on calmodulin. However, subsequent work showed that the Ca²⁺ binding sites in calmodulin were not required for homotypic fusion of vacuoles [90] and that Ca²⁺ chelators blocked homotypic fusion in vitro not by buffering Ca²⁺ but by changing the ionic strength of the reaction buffer [91]. Furthermore, vacuolar Ca²⁺ was not required for homotypic fusion because vacuoles that lack both Pmc1 and Vcx1 fuse as efficiently as normal vacuoles [88]. The release of Ca²⁺ correlates with the formation of inter-vacuolar SNARE complexes [90] but does not require Yvc1 [77]. Though it remains possible that some other vacuolar Ca²⁺ channel or transporter and some other targets of Ca²⁺ in the cytoplasm will be discovered during the process, the evidence currently favors a model where Ca²⁺ release from vacuoles during homotypic fusion is coincidental rather than purposeful.

Finally, there has been a report of inositol-1,4,5-trisphosphate (IP3)-dependent release of Ca²⁺ from purified S. cerevisiae vacuoles [92]. The IP3-sensitive channel or transporter has not yet been identified and, as mentioned earlier, homologs of the known IP3-receptors found in animals and amoebas are not evident in fungi. Yvc1 also fails to respond to IP3 [77]. Evidence for IP3-sensitive Ca²⁺ influx pathways, on the other hand, is growing.

IP3-sensitive Ca²⁺ influx

Two different experimental regimens have been shown to trigger Ca²⁺ influx into S. cerevisiae cells in a fashion that depends on phospholipase C (see Fig. 4). The first regimen involves the use of protonophores (carbonyl cyanide m-chlorophenylhydrazone [CCCP]), which are thought to promote influx of H⁺ in acidic media but may also uncouple vacuolar and mitochondrial membranes. Within seconds of CCCP addition, wild-type S. cerevisiae cells exhibit robust Ca²⁺ influx and elevation of [Ca²⁺]_CYT [93]. Extracellular Ca²⁺ and phospholipase C (termed Plc1) were absolutely required for these effects. Additionally, the rise of [Ca²⁺]_CYT is augmented by elimination of an IP3-kinase (termed Arg82) that consumes IP3. These findings suggest H⁺ influx may activate Plc1 and the resulting rise of IP3 production may activate Ca²⁺ influx pathways. Within a minute of [Ca²⁺]_CYT elevation, the plasma membrane H⁺-ATPase (termed Pma1) primarily responsible for maintenance of cytoplasmic pH becomes significantly activated [93]. Consistently, Pma1 activation in response to protonophores is dampened by the activities of Arg82 (IP3-kinase) and Pmc1 (vacuolar Ca²⁺ pump) and is dependent on the activities of Plc1 and Pkc1 (protein kinase C) which may directly phoshorylate the C-terminus of Pma1 [93]. It is not yet known how H⁺ influx might activate Plc1. Identification of the proposed IP3-sensitive Ca²⁺ influx channel or transporter will be a key advance in this field.

Figure 4. IP3-dependent Ca²⁺ influx and signaling in S. cerevisiae cells in response to protonophores and sugar refeeding.

Protonphores (CCCP) stimulate H+ influx and subsequent Ca²⁺ influx via unknown pathways that depend on phospholipase C (Plc1) and are sensitive to IP3-kinase (Arg82). Sugar (glucose) refeeding to starved cells induces similar Plc1-dependent Ca²⁺ influx that also depends on glucose transporters, hexokinases, phosphoglucomutases, and glucose sensors (Snf3; Gpr1-Gpa2). Both stimuli result in transient elevation of [Ca²⁺]_CYT and activation of a plasma membrane H+ pump (Pma1) through a process that may depend on protein kinase C (Pkc1).

The second regimen that triggers Plc1-dependent Ca²⁺ influx and Pma1 activation in S. cerevisiae cells is sugar starvation and refeeding [94–96]. Glucose refeeding to starved cells induces a rise of [Ca²⁺]_CYT within ten seconds and an increase of Pma1 activity within minutes. Hexose transporters, hexokinases/glucokinases, and phosphoglucomutases are all required for the [Ca²⁺]_CYT elevation upon glucose refeeding [65, 95, 97], indicating that elevated concentrations of glucose-1-phosphate (or a derivative) may be crucial for the effect. The glucose metabolite may be sensed by receptors (termed Snf3 and Gpr1), the latter of which couples to a heterotrimeric G-protein (termed Gpa2) that may somehow activate Plc1 [95, 98]. Alternatively, glucose metabolites may trigger H⁺ influx similar to the effects of protonophores described earlier or glucose metabolism in general may acidify the cytoplasm directly. In any case, Plc1 is required for waves of IP3 production and the subsequent [Ca²⁺]_CYT rise whereas Arg82 dampens these effects [96]. Though the molecular targets of intracellular IP3 is not yet known, observations suggest that HACS plays either a large role [97] or a small role [98] in the Ca²⁺ influx that occurs after glucose refeeding. The resulting rise of [Ca²⁺]_CYT is more strongly dissipated by Vcx1 than by Pmc1 [99]. Though glucose refeeding can activate calcineurin and Crz1 [100], the primary role of the [Ca²⁺]_CYT elevation may be the activation of Pma1 and the maintenance of cellular pH during bursts of glycolytic activity [94].

Metabolic engineering experiments that cause constitutively high levels of glucose-1-phosphate accumulation cause long-term increases in Ca²⁺ influx, [Ca²⁺]_CYT, and calcineurin signaling [65, 97, 101–103]. Again, the relevant sensors of intracellular glucose metabolites have not yet been identified. The regulatory network operating in these conditions may be more challenging to unravel because the growth conditions (galactose utilization in phosphoglucomutase-deficient cells) also generate stress in the endoplasmic reticulum, which activates the UPR signaling pathway [103] and probably activates HACS via the MAP kinase cascade described earlier. Because Pmc1 contributes to the stress in these long-term experiments [103], it seems possible that the constitutively elevated glucose-1-phosphate somehow diminishes Pmr1 function or enhances Ca²⁺ efflux from secretory compartments, thus generating secretory stresses. Therefore, the acute accumulation of glucose-1-phosphate through glucose refeeding and the chronic accumulation through metabolic engineering may utilize distinct mechanisms in the coupling to Ca²⁺ influx and signaling.

Much remains to be learned about the linkages between glucose metabolism, H⁺ influx, IP3 production, Ca²⁺ influx, sequestration, and signaling in S. cerevisiae and other fungi. The identification of the IP3 sensors, the glucose-1-phosphate sensors, and the relevant ion channels will be novel and critical for determining a much more realistic picture of this signaling network, which can then be compared and contrasted to related networks that exist in animal cells.

Ca²⁺ signaling during the mating response

Haploid cells of S. cerevisiae and many other fungi that are of complimentary mating types often undergo developmental programs that allow for their union and formation of stable diploid cells or heterokaryons. These programs are generally referred to as mating. In S. cerevisiae each of the two mating cell types secretes a peptide pheromone that arouses and attracts members of the other cell type and coordinates a process of pair-wise mate selection that culminates with cell and nuclear fusion. The pheromones, receptors, and response pathways in this model fungus are extremely well characterized as a consequence of several decades of intensive study. In terms of Ca²⁺ influx and signaling during mating, it is first necessary to distinguish the cells that successfully mate from those that respond to mating pheromones and fail to mate.

Different concentrations of mating pheromones are known to evoke different developmental programs in the responding cells. For example, low concentrations cause a cell-cycle arrest whereas moderate concentrations induce polarized growth and changes in cell morphology. Moderate concentrations of mating pheromones are required to stimulate Ca²⁺ influx via HACS (involving Cch1, Mid1, and Ecm7 proteins) and to elevate [Ca²⁺]_CYT [37–39, 104, 105]. High concentrations of mating pheromones are required to stimulate a low-affinity Ca²⁺ influx system (LACS) that involves Fig 1, a member of the claudin superfamily of four-spanner transmembrane proteins that includes γ-subunits of voltage-gated Ca²⁺ channels as well as regulatory subunits of ionotropic glutamate receptors in animals [106]. As depicted in Figure 5, HACS and LACS function independently of each other but the potential for cross regulation becomes evident when considering that calcineurin may dephosphorylate Cch1 and specifically inhibit HACS in vivo [34]. Regardless of the system utilized in any particular condition, the influx of extracellular Ca²⁺ and the elevation of [Ca²⁺]_CYT begin to occur after 45–60 minutes of pheromone exposure, which is a period of time that is sufficient to allow changes in cell morphology. It is not yet known whether these slow responses involve glucose-1-phosphate, IP3, membrane stresses, H⁺ fluxes, or other possible mediators. Nevertheless, they do result in the eventual activation of calcineurin and expression of Crz1-dependent target genes. Mutants that lack Crz1 seem to survive as long as wild-type S. cerevisiae cells in the continuous presence of mating pheromones [4, 5]. In contrast, mutants that lack calcineurin, calmodulin, or the upstream Ca²⁺ channels slowly die when continuously exposed to mating pheromones in the absence of mates [39, 104, 107, 108]. The manner of cell death was originally proposed as apoptosis [109] but subsequent work refutes apoptosis and suggests a necrosis-like process instead [110]. Thus, calcineurin activation is essential for long-term survival of cells that are responding to mating pheromones but are unable to find mating partners or mate.

Figure 5. Ca²⁺ transport and signaling in S. cerevisiae cells responding to mating pheromones in the absence of mates.

Mating pheromones elicit mating responser in haploid S. cerevisiae cells that include induction of a low-affinity Ca²⁺ influx system (LACS) that depends on Fig 1 and a high-affinity Ca²⁺ influx system (HACS) that depends on Cch1, Mid1, and Ecm7. The delayed elevation of [Ca²⁺]_CYT results in activation of downstream signaling pathways that are essential for long-term survival of cells that do not successfully mate. Rapidly mated cells (not shown) become desensitized to mating pheromones and do not require LACS, HACS, calmodulin, or calcineurin for efficient mating.

In mixtures of the two haploid mating types of S. cerevisiae, nearly all the cells respond to secreted mating pheromones but only a small fraction of them successfully fuse to form a diploid zygote (that becomes insensitive to both mating pheromones). Removal of extracellular Ca²⁺ has little effect on mating efficiency, but some fusion-incompetent mutants of S. cerevisiae become more fertile when environmental Ca²⁺ is elevated [111]. Quantitative assays of mating efficiency show that calcineurin and calmodulin are not required in either cell type for efficient mating [107, 112, 113], so most mating is probably completed prior to the time at which the calcineurin-deficient mutant cells typically die from pheromone exposure. Mutants that lack Pmc1, Vcx1, or Yvc1 also seem to mate efficiently, suggesting the vacuolar Ca²⁺ stores are not critical for the mating process. Mutants that lack Pmr1 mate poorly because one of the mating pheromones, α-factor, depends on the Ca²⁺-dependent protease Kex2 in the Golgi complex for proteolytic activation [29]. Mutants that lack the high- or low-affinity Ca²⁺ channels also mate efficiently in standard conditions, but the zygotes sometimes fail to remove some of the cell wall material at the septum between mating partners [41, 106, 114]. These studies have revealed little or no role for intracellular Ca²⁺ elevation in the process of mating, if indeed such elevations actually occur in mating cells. No direct observations of either [Ca²⁺]_CYT increases or calcineurin activation have been reported on mating S. cerevisiae cells. Toward this goal, it may be possible to adapt the GFP-based sensors of [Ca²⁺]_CYT and calineurin activity used in the analysis of spike and burst phenomena in single S. cerevisiae cells [20] to S. cerevisiae cells engaged in mating. Until real-time assays can be applied, we will not know with certainty if [Ca²⁺]_CYT regulates the mating process or if it just promotes survival of the cells that cannot find mates. How acidic Ca²⁺ stores of S. cerevisiae and other fungi shape developmental processes such as this is yet another important question to be addressed.

Acidic Ca²⁺ stores in fungi, animals, and beyond

The discoveries of Pmc1, Vcx1, and Yvc1 as major Ca²⁺ importers and exporters in the acidic vacuoles of S. cerevisiae has opened new frontiers in the study of acidic Ca²⁺ stores in other fungi and eukaryotes. The few studies of these proteins in other fungi are generally consistent with the S. cerevisiae paradigms: localization to the vacuole membrane, regulation by calcineurin, and contributions to Ca²⁺ homeostasis and signaling. But the other fungal species allow investigations of acidic Ca²⁺ stores in wholly new contexts. For example, the human pathogen Cryptococcus neoformans requires Vcx1 in its vacuole for full virulence in a mouse model of inhalation cryptococcosis [115]. The human pathogen Candida albicans utilizes Pmc1 to promote hypersensitivity to common antifungal medications [116] in a process that can be studied in S. cerevisiae. The plant pathogen Magnaporthe oryzae employs homologs of both Pmc1 and Yvc1 for virulence in rice and wheat and for several developmental processes [117]. These effects of vacuolar Ca²⁺ transporters and channels on pathogenicity may result from their impacts on signaling by calcineurin or other sensors of [Ca²⁺]_CYT. Alternatively, the stored Ca²⁺ itself may contribute in unknown ways to the function of acidic compartments and their roles in fungal virulence.

The sequenced animal genomes so far indicate a complete absence of CAX-type Ca²⁺/H⁺ exchangers (Vcx1 homologs) and a re-localization of PMCA-type Ca²⁺ pumps and TRPC-type Ca²⁺ channels (Pmc1 and Yvc1 homologs) to the plasma membranes of most cell types. The acidic Ca²⁺ stores of animal cells may instead be supplied by SPCA-type Ca²⁺ pumps with some additional contributions by PMCA- and SERCA-type pumps and may instead be tapped by TPC-type Ca²⁺ channels, which are not found in the available fungal genomes. A Golgi-localized V-type H⁺ pump may also facilitate Ca²⁺ uptake and release through coupling effects on Na⁺/H⁺ and Ca²⁺/Na⁺ exchangers in animals. Such differences between fungi and animals probably reflect the somewhat distinct functions of the major acidic organelles in fungi and animals. Fungal vacuoles seem larger and more oriented toward storage of nutrients and metabolites than animal lysosomes, which seem more specialized for digestive and degradative functions given their smaller size and much greater acidity. Such rationalizations should be taken with due skepticism, however, especially in light of the growing recognition that these organelles exist as a spectrum of vacuole-like and lysosome-like structures even within a single cell type.

From the more distant branches of the tree of life, the core toolkit responsible for Ca²⁺ influx, sequestration, release, and signaling become even more apparent. The CAX-type Ca²⁺/H⁺ exchangers are common in plants, apicomplexan parasites, and other protozoans that can impact human health and wellbeing. Given their prevalence, why would this useful class of Ca²⁺ transporter have been lost in the earliest animals only to be replaced by more elaborate systems? The answers to questions like this one may provide a very useful framework for viewing and understanding the acidic Ca²⁺ stores of humans.