Mode of action of yeast toxins: energy requirement for Saccharomyces cerevisiae killer toxin

Bioprotection with Saccharomyces cerevisiae: A Promising Strategy

Bioprotection in winemaking refers to the use of naturally occurring microorganisms-mainly non-Saccharomyces yeasts-to inhibit the growth of spoilage microbes and reduce the need for chemical preservatives like sulfur dioxide (SO2). Numerous studies have demonstrated the benefits of non-Saccharomyces as bioprotectants. However, the use of Saccharomyces cerevisiae as a bioprotectant has been studied very little. Furthermore, it can offer many advantages for the production of sulfite-free wines. To test if S. cerevisiae could be used in bioprotection, we compared the ability of different strains to inhibit the growth of Brettanomyces bruxellensis and Hanseniaspora uvarum. Among the strains tested, the S. cerevisiae Sc54 strain isolated from the vineyard of the Bekaa plain was selected. To investigate its mechanisms of action, we analyzed its metabolite production, including acetic acid and ethanol. Taking into account the low levels of these metabolites and the lack of similar inhibition patterns in media supplemented with acetic acid and ethanol, it appears that other factors contribute to its antagonistic properties. Nutrient competition was ruled out as a factor, as the growth inhibition of B. bruxellensis and H. uvarum occurred rapidly within the first 24 h of co-culture. In this study, we explored the role of the S. cerevisiae killer toxin (Sc54Kt) as a bioprotective agent against H. uvarum and B. bruxellensis spoilage yeasts. Purification procedures with ethanol allowed the extraction of Sc54Kt, yielding two concentrations (0.185 and 0.5 mg/mL). Remarkably, semi-purified Sc54Kt exhibited inhibitory effects at both concentrations under winemaking conditions, effectively controlling the growth and metabolic activity of the target spoilage yeasts. Overall, these findings demonstrate that S. cerevisiae Sc54 not only exerts a strong bioprotective effect but also contributes to improving the quality of wine. The results suggest that S. cerevisiae Sc54 is a promising bioprotective agent for mitigating spoilage yeasts in winemaking, offering a natural and effective alternative to conventional antimicrobial strategies.

The Killer Saccharomyces cerevisiae Toxin: From Origin to Biomedical Research

The killer systems of S. cerevisiae are defined by the co-infection of two viral agents, an M virus and a helper virus. Each killer toxin is determined by the type of M virus (ScV-M1, ScV-M2, ScV-M28, and ScV-Mlus), which encodes a specific toxin (K1, K2, K28, and Klus). Since their discovery, interest in their potential use as antimicrobial agents has driven research into the mechanisms of action of these toxins on susceptible cells. This review provides an overview of the key aspects of killer toxins, including their origin and the evolutionary implications surrounding the viruses involved in the killer system, as well as their potential applications in the biomedical field and as a biological control strategy. Special attention is given to the mechanisms of action described to date for the various S. cerevisiae killer toxins.

The Viral K1 Killer Yeast System: Toxicity, Immunity, and Resistance

Killer yeasts, such as the K1 killer strain of S. cerevisiae, express a secreted anti-competitive toxin whose production and propagation require the presence of two vertically-transmitted dsRNA viruses. In sensitive cells lacking killer virus infection, toxin binding to the cell wall results in ion pore formation, disruption of osmotic homeostasis, and cell death. However, the exact mechanism(s) of K1 toxin killing activity, how killer yeasts are immune to their own toxin, and which factors could influence adaptation and resistance to K1 toxin within formerly sensitive populations are still unknown. Here, we describe the state of knowledge about K1 killer toxin, including current models of toxin processing and killing activity, and a summary of known modifiers of K1 toxin immunity and resistance. In addition, we discuss two key signaling pathways, HOG (high osmolarity glycerol) and CWI (cell wall integrity), whose involvement in an adaptive response to K1 killer toxin in sensitive cells has been previously documented but requires further study. As both host-virus and sensitive-killer competition have been documented in killer systems like K1, further characterization of K1 killer yeasts may provide a useful model system for study of both intracellular genetic conflict and counter-adaptation between competing sensitive and killer populations.

Substitution of cysteines in the yeast viral killer toxin K1 precursor reveals novel insights in heterodimer formation and immunity

The killer toxin K1 is a virally encoded fungal A/B toxin acting by disrupting plasma membrane integrity. The connection of α and β constitutes a critical feature for toxin biology and for decades the formation of three disulphide bonds linking the major toxin subunits was accepted as status quo. Due to the absence of experimental evidence, the involvement of each cysteine in heterodimer formation, K1 lethality and immunity was systematically analysed. Substitution of any cysteine in α led to a complete loss of toxin dimer secretion and toxicity, whereas K1 toxin derivatives carrying mutations of C248, C312 or the double mutation C248-312 were active against spheroplasted cells. Importantly, substitution of the C95 and C107 in the toxin precursor completely abolished the mediation of functional immunity. In contrast, K1 toxicity, i.e. its ionophoric effect, does not depend on the cysteine residues at all. In contrast to the literature, our data imply the formation of a single disulphide bond involving C92 in α and C239 in β. This finding not only refines the current model stated for decades but also provides new opportunities to elucidate the mechanisms underlying K1 toxicity and immunity at the molecular level.

Transcriptome Kinetics of Saccharomyces cerevisiae in Response to Viral Killer Toxin K1

The K1 A/B toxin secreted by virus-infected Saccharomyces cerevisiae strains kills sensitive cells via disturbance of cytoplasmic membrane functions. Despite decades of research, the mechanisms underlying K1 toxicity and immunity have not been elucidated yet. In a novel approach, this study aimed to characterize transcriptome changes in K1-treated sensitive yeast cells in a time-dependent manner. Global transcriptional profiling revealed substantial cellular adaptations in target cells resulting in 1,189 differentially expressed genes in total. Killer toxin K1 induced oxidative, cell wall and hyperosmotic stress responses as well as rapid down-regulation of transcription and translation. Essential pathways regulating energy metabolism were also significantly affected by the toxin. Remarkably, a futile cycle of the osmolytes trehalose and glycogen was identified probably representing a critical feature of K1 intoxication. In silico analysis suggested several transcription factors involved in toxin-triggered signal transduction. The identified transcriptome changes provide valuable hints to illuminate the still unknown molecular events leading to K1 toxicity and immunity implicating an evolutionarily conserved response at least initially counteracting ionophoric toxin action.

K2 killer toxin-induced physiological changes in the yeast Saccharomyces cerevisiae

Saccharomyces cerevisiae cells produce killer toxins, such as K1, K2 and K28, that can modulate the growth of other yeasts giving advantage for the killer strains. Here we focused on the physiological changes induced by K2 toxin on a non-toxin-producing yeast strain as well as K1, K2 and K28 killer strains. Potentiometric measurements were adjusted to observe that K2 toxin immediately acts on the sensitive cells leading to membrane permeability. This correlated with reduced respiration activity, lowered intracellular ATP content and decrease in cell viability. However, we did not detect any significant ATP leakage from the cells treated by killer toxin K2. Strains producing heterologous toxins K1 and K28 were less sensitive to K2 than the non-toxin producing one suggesting partial cross-protection between the different killer systems. This phenomenon may be connected to the observed differences in respiratory activities of the killer strains and the non-toxin-producing strain at low pH. This might also have practical consequences in wine industry; both as beneficial ones in controlling contaminating yeasts and non-beneficial ones causing sluggish fermentation.

Cell cycle arrest and apoptosis, two alternative mechanisms for PMKT2 killer activity

Pichia membranifaciens CYC 1086 secretes a unique 30kDa killer toxin (PMKT2) that inhibits a variety of spoilage yeasts and fungi of agronomical interest. The cytocidal effect of PMKT2 on Saccharomyces cerevisiae cells was studied. Metabolic events associated with the loss of S. cerevisiae viability caused by PMKT2 were qualitatively identical to those reported for K28 killer toxin activity, but different to those reported for PMKT. At higher doses, none of the cellular events accounting for the action of PMKT, the killer toxin secreted by P. membranifaciens CYC 1106, was observed for PMKT2. Potassium leakage, sodium influx and the decrease of intracellular pH were not among the primary effects of PMKT2. We report here that this protein is unable to form ion-permeable channels in liposome membranes, suggesting that channel formation is not the mechanism of cytotoxic action of PMKT2. Nevertheless, flow cytometry studies have revealed a cell cycle arrest at an early S-phase with an immature bud and pre-replicated 1n DNA content. By testing the sensitivity of cells arrested at different stages in the cell cycle, we hoped to identify the execution point for lethality more precisely. Cells arrested at the G1-phase by α-factor or arrested at G2-phase by the spindle poison methyl benzimidazol-2-yl-carbamate (MBC) were protected against the toxin. Cells released from the arrest in both cases were killed by PMKT2 at a similar rate. Nevertheless, cells released from MBC-arrest were able to grow for a short time, and then viability dropped rapidly. These findings suggest that cells released from G2-phase are initially able to divide, but die in the presence of PMKT2 after initiating the S-phase in a new cycle, adopting a terminal phenotype within that cycle. By contrast, low doses of PMKT and PMKT2 were able to generate the same cellular response. The evidence presented here shows that treating yeast with low doses of PMKT2 leads to the typical membranous, cytoplasmic, mitochondrial and nuclear markers of apoptosis, namely, the production of reactive oxygen species, DNA strand breaks, metacaspase activation and cytochrome c release.

Pichia anomala DBVPG 3003 secretes a ubiquitin-like protein that has antimicrobial activity

The yeast strain Pichia anomala DBVPG 3003 secretes a killer toxin (Pikt) that has antifungal activity against Brettanomyces/Dekkera sp. yeasts. Pikt interacts with beta-1,6-glucan, consistent with binding to the cell wall of sensitive targets. In contrast to that of toxin K1, secreted by Saccharomyces cerevisiae, Pikt killer activity is not mediated by an increase in membrane permeability. Purification of the toxin yielded a homogeneous protein of about 8 kDa, which showed a marked similarity to ubiquitin in terms of molecular mass and N-terminal sequences. Pikt is also specifically recognized by anti-bovine ubiquitin antibodies and, similar to ubiquitin-like peptides, is not absorbed by DEAE-cellulose. However, Pikt differs from ubiquitin in its sensitivity to proteolytic enzymes. Therefore, Pikt appears to be a novel ubiquitin-like peptide that has killer activity.

Biology of the pathogenic yeast Candida glabrata

The yeasts, being favorite eukaryotic microorganisms used in food industry and biotechnologies for production of biomass and various substances, are also used as model organisms in genetic manipulation, molecular and biological research. In this respect, Saccharomyces cerevisiae is the best-known species but current situation in medicine and industry requires the use of other species. Here we summarize the basic taxonomic, morphological, physiological, genetic, etc. information about the pathogenic yeast Candida glabrata that is evolutionarily very closely related to baker's yeast.

Kluyveromyces phaffii killer toxin active against wine spoilage yeasts: purification and characterization

The killer toxin secreted by Kluyveromyces phaffii (KpKt) is active against spoilage yeast under winemaking conditions and thus has potential applications in the biocontrol of undesired micro-organisms in the wine industry. Biochemical characterization and N-terminal sequencing of the purified toxin show that KpKt is a glycosylated protein with a molecular mass of 33 kDa. Moreover, it shows 93% and 80% identity to a beta-1,3-glucanase of Saccharomyces cerevisiae and a beta-1,3-glucan transferase of Candida albicans, respectively, and it is active on laminarin and glucan, thus showing a beta-glucanase activity. Competitive inhibition of killer activity by cell-wall polysaccharides suggests that glucan (beta-1,3 and beta-1,6 branched glucans) represents the first receptor site of the toxin on the envelope of the sensitive target. Flow cytometry analysis of the sensitive target after treatment with KpKt and K1 toxin of S. cerevisiae, known to cause loss of cell viability via formation of pores in the cell membrane, suggests a different mode of action for KpKt.

Killing of Candida albicans by human salivary histatin 5 is modulated, but not determined, by the potassium channel TOK1

Salivary histatin 5 (Hst 5), a potent toxin for the human fungal pathogen Candida albicans, induces noncytolytic efflux of cellular ATP, potassium, and magnesium in the absence of cytolysis, implicating these ion movements in the toxin's fungicidal activity. Hst 5 action on Candida resembles, in many respects, the action of the K1 killer toxin on Saccharomyces cerevisiae, and in that system the yeast plasma membrane potassium channel, Tok1p, has recently been reported to be a primary target of toxin action. The question of whether the Candida homologue of Saccharomyces Tok1p might be a primary target of Hst 5 action has now been investigated by disruption of the C. albicans TOK1 gene. The resultant strains (TOK1/tok1) and (tok1/tok1) were compared with wild-type Candida (TOK1/TOK1) for relative ATP leakage and killing in response to Hst 5. Patch-clamp measurements on Candida protoplasts were used to verify the functional deletion of Tok1p and to provide its first description in Candida. Tok1p is an outwardly rectifying, noisily gated, 40-pS channel, very similar to that described in Saccharomyces. Knockout of CaTOK1 (tok1/tok1) completely abolishes the currents and gating events characteristic of Tok1p. Also, knockout (tok1/tok1) increases residual viability of Candida after Hst 5 treatment to 27%, from 7% in the wild type, while the single allele deletion (TOK1/tok1) increases viability to 18%. Comparable results were obtained for Hst-induced ATP efflux, but quantitative features of ATP loss suggest that wild-type TOK1 genes function cooperatively. Overall, very substantial killing and ATP efflux are produced by Hst 5 treatment after complete knockout of wild-type TOK1, making clear that Tok1p channels are not the primary site of Hst 5 action, even though they do play a modulating role.

Isolation and characterization of Saccharomyces cerevisiae mutants with a different degree of resistance to killer toxins K1 and K2

Killer toxin K1 of Saccharomyces cerevisiae kills sensitive cells of the same species by disturbing the ion gradient across the plasma membrane after binding to the receptor at cell wall beta-1,6-glucan. Killer protein K2 is assumed to act by a similar mechanism. To identify the putative plasma membrane receptors for both toxins we mutagenized three sensitive S. cerevisiae strains and searched for clones with killer-resistant spheroplasts. The well diffusion assay identified three phenotypically different groups of clones: clones resistant simultaneously to both toxins, clones with lowered sensitivity to only K1 toxin and those with strongly lowered sensitivity to K2 and partially lowered sensitivity to K1 toxin. These phenotypes are controlled by recessive mutations that belong to at least four different complementation groups. This indicates certain differences at the level of interaction of K1 and K2 toxin with sensitive cells.

A molecular target for viral killer toxin: TOK1 potassium channels

Killer strains of S. cerevisiae harbor double-stranded RNA viruses and secrete protein toxins that kill virus-free cells. The K1 killer toxin acts on sensitive yeast cells to perturb potassium homeostasis and cause cell death. Here, the toxin is shown to activate the plasma membrane potassium channel of S. cerevisiae, TOK1. Genetic deletion of TOK1 confers toxin resistance; overexpression increases susceptibility. Cells expressing TOK1 exhibit toxin-induced potassium flux; those without the gene do not. K1 toxin acts in the absence of other viral or yeast products: toxin synthesized from a cDNA increases open probability of single TOK1 channels (via reversible destabilization of closed states) whether channels are studied in yeast cells or X. laevis oocytes.

Anti-Candida activity of a novel killer toxin from the yeast Williopsis mrakii

A screening of putative killer yeast strains showed that spore-forming ascomycetous yeasts of the genera Pichia and Williopsis displayed the broadest range of activity against sensitive strains of Candida spp. and Saccharomyces cerevisiae. Williopsis mrakii (NCYC 500) showed extensive anti-Candida activity against strains isolated from clinical specimens. W. mrakii killer factor was produced in minimal media as a function of growth and its activity reached constant levels as cells entered stationary phase. The proteinaceous killer toxin was found to be unstable without a specific range of temperature and pH (above 30 degrees C and pH 4.0), and further analysis showed that the active toxin molecule was an acidic polypeptide with a relative molecular mass between 1.8-5.0 kDa. At critical concentrations the killer factor exerted a greater effect on stationary phase cells of Candida than cells from an exponential phase of growth. At low concentrations, the killer toxin produced a fungistatic effect on sensitive yeasts but at higher concentrations there was evidence to suggest that membrane damage accounted for the zymocidal effects of the killer factor. the cidal nature of the toxin was reflected in a rapid decrease in sensitive cell viability. Findings presented suggest that W. mrakii killer toxin has potential as a novel antimycotic agent in combatting medically important strains of Candida.

Significance of the lag phase in K1 killer toxin action on sensitive yeast cells

The minimum period between the addition of killer toxin K1 to sensitive yeast cells and the appearance of first cells stained with bromocresol purple indicating membrane damage, is approximately 20 min. The length of this lag phase depends strongly on toxin concentration, extending sharply at toxin levels lower than 60 lethal units (LU) per cell (about one-tenth of the toxin concentration necessary for saturating all surface receptors). As the binding of the toxin to the cell is virtually complete within 1 min, the rest of the lag phase reflects processes different from actual binding, e.g. combination of several toxin molecules to form a membrane ion channel or pore.

The characterization and crystallization of a virally encoded Ustilago maydis KP4 toxin

KP4 is a virally encoded and highly specific toxin that kills fungi closely related to the fungus Ustilago maydis. The toxin was purified and crystals were formed using ammonium sulfate as precipitant. The crystals belong to the space group P6(1)(5)22 and diffracted to approximately 2.2 A resolution. Circular dicroism spectroscopy suggests that the protein is predominantly comprised of beta-strands.

Kinetic studies of killer toxin K1 binding to yeast cells indicate two receptor populations

A recently described new method for determination of killer toxin activity was used for kinetic measurements of K1 toxin binding. The cells of the killer sensitive strain Saccharomyces cerevisiae S6 were shown to carry two classes of toxin binding sites differing widely in their half-saturation constants and maximum binding rates. The low-affinity and high-velocity binding component (KT1 = 2.6 x 10(9) L.U./ml, Vmax1 = 0.19 s-1) probably reflects diffusion-limited binding to cell wall receptors; the high-affinity and low-velocity component (KT2 = 3.2 x 10(7) L.U./ml, Vmax2 = 0.03 s-1) presumably indicates the binding of the toxin to plasma membrane receptors. Adsorption of most of the killer toxin K1 to the surface of sensitive cells occurred within 1 min and was virtually complete within 5 min. The amount of toxin that saturated practically all cell receptors was about 600 lethal units (L.U.) per cell of S. cerevisiae S6.

Factors affecting the susceptibility of sensitive yeast cells to killer toxin K1

Optimum conditions for action of the killer toxin K1 on sensitive strain S. cerevisiae S6 were established. Maximum killing was reached in a very narrow pH range of 4.5-4.6. Maximum susceptibility to toxin was displayed by highly energized fresh cells from the early exponential phase in the presence of an external energy source (at least 200 mmol/L glucose). Further, maintenance of maximum membrane potential was necessary for killer action, as documented by decreasing toxin activity in the presence of increasing concentrations of KCl. The killing was strongly stimulated in the presence of millimolar concentrations of Ca2+ and Mg2+.

Fluorescent staining with bromocresol purple: a rapid method for determining yeast cell dead count developed as an assay of killer toxin activity

A method is described for detecting yeast cells with plasma membrane damage, based on cell staining with bromocresol purple (BCP) which has a convenient fluorescence after entering the cells at pH below 5.2. The method was used to determine the activity of Saccharomyces cerevisiae pore-forming killer toxin K1 in commonly used lethal units. The BCP test is rapid and as precise as the plating test.

Determination of killer yeast activity in fermenting grape juice by using a marked Saccharomyces wine yeast strain

The Escherichia coli beta-glucuronidase gene has been used as a marker gene to monitor a killer Saccharomyces cerevisiae strain in mixed-culture ferments. The marked killer strain was cured of its M-dsRNA genome to enable direct assessment of the efficiency of killer toxin under fermentation conditions. Killer activity was clearly evident in fermenting Rhine Riesling grape juice of pH 3.1 at 18 degrees C, but the extent of killing depended on the proportion of killer to sensitive cells at the time of inoculation. Killer activity was detected only when the ratio of killer to sensitive cells exceeded 1:2. At the highest ratio of killer to sensitive cells tested (2:1), complete elimination of sensitive cells was not achieved.

Interfaces of the yeast killer phenomenon

A new prophylactic and therapeutic antimicrobial strategy based on a specific physiological target that is effectively used by killer yeasts in their natural ecological competition is theorized. The natural system exploited is the yeast killer phenomenon previously adopted as an epidemiological marker for intraspecific differentiation of opportunistic yeasts, hyphomycetes, and bacteria. Pathogenic microorganisms (Candida albicans) may be susceptible to the activity of yeast killer toxins due to the presence of specific cell wall receptors. On the basis of the idiotypic network, we report that antiidiotypic antibodies, produced against a monoclonal antibody bearing the receptor-like idiotype, are in vivo protecting animals immunized through idiotypic vaccination and in vitro mimicking the antimicrobial activity of yeast killer toxins, thus acting as antibiotics.

Detection by immunofluorescent anti-idiotypic antibodies of yeast killer toxin cell wall receptors of Candida albicans

Yeast killer toxin cell wall receptors of Candida albicans were observed by indirect immunofluorescence using an affinity purified rabbit anti-idiotypic antiserum. The antiserum had been raised against a monoclonal antibody neutralizing the in vitro activity of a killer toxin produced by a selected strain of Hansenula anomala UCSC 25F. This simple procedure permitted the location of killer toxin cell wall receptors in various morphological phases of the yeast cells. The use of the indirect immunofluorescence technique with anti-idiotypic antibodies may have potential value in determining the occurrence of killer toxin receptors in other microbial systems.

Yeast K1 killer toxin forms ion channels in sensitive yeast spheroplasts and in artificial liposomes

The patch-clamp technique was used to examine the plasma membranes of sensitive yeast spheroplasts exposed to partially purified killer toxin preparations. Asolectin liposomes in which the toxin was incorporated were also examined. Excised inside-out patches from these preparations often revealed at 118 pS conductance appearing in pairs. The current through this conductance flickered rapidly among three states: dwelling mostly at the unit-open state, less frequently at the two-unit-open state, and more rarely at the closed state. Membrane voltages from -80 to 80 mV had little influence on the opening probability. The current reversed near the equilibrium potential of K+ in asymmetric KCl solutions and also reversed near O mV at symmetric NaCl vs. KCl solutions. The two levels of the conductance were likely due to the toxin protein, as treatment of spheroplasts or liposomes with extracellular protein preparations from isogenic yeasts deleted for the toxin gene gave no such conductance levels. These results show that in vivo the killer-toxin fraction can form a cation channel that seldom closes regardless of membrane voltage. We suggest that this channel causes the death of sensitive yeast cells.

Genetic and molecular approaches to synthesis and action of the yeast killer toxin

The K1 killer toxin of Saccharomyces cerevisiae is a secreted, virally-coded protein lethal to sensitive yeasts. Killer yeasts are immune to the toxin they produce. This killer system has been extensively examined from genetic and molecular perspectives. Here we review the biology of killer yeasts, and examine the synthesis and action of the protein toxin and the immunity component. We summarise the structure of the toxin precursor gene and its protein products, outline the proteolytic processing of the toxin subunits from the precursor, and their passage through the yeast secretory pathway. We then discuss the mode of action of the toxin, its lectin-like interaction with a cell wall glucan, and its probable role in forming channels in the yeast plasma membrane. In addition we describe models of how a toxin precursor species functions as the immunity component, probably by interfering with channel formation. We conclude with a review of the functional domains of the toxin structural gene as determined by site-directed mutagenesis. This work has identified regions associated with glucan binding, toxin activity, and immunity.

The double-stranded RNA in Trichomonas vaginalis may originate from virus-like particles

A linear 5.5-kilobase double-stranded RNA, identified in many strains and isolates of the parasitic protozoan Trichomonas vaginalis in a previous study, is found largely intact in ribonuclease-treated homogenates of the parasite. It can be pelleted with membranes from the homogenate at 12,500 X g and further purified in CsCl buoyant density-gradient centrifugations. The purified sample contains the double-stranded RNA as well as one major protein with an estimated molecular mass of 85 kDa in NaDodSO4/PAGE. Electron microscopic examinations indicated the presence of icosahedral virus-like particles of 33-nm diameter in the purified preparation. The exact location of the virus in T. vaginalis is not clear, except that it is not found in the nuclear fraction and is probably membrane-bound. No free virus can be recovered from the culture medium of T. vaginalis, and no successful infection of virus-free T. vaginalis strains by purified virus has yet been accomplished. There is no viral genomic sequence identifiable in host DNA. So far as we know, it is the first time a double-stranded RNA virus has been identified in a protozoan.

Double-stranded ribonucleic acid killer systems in yeasts

Images

Characterization of a novel killer toxin encoded by a double-stranded linear DNA plasmid of Kluyveromyces lactis

A novel killer toxin, encoded by a double-stranded linear DNA plasmid pGK l-1 (5.4 MDa) in Kluyveromyces lactis IFO 1267 was purified 320 000-fold from the culture broth of yeast. The toxin was obtained in an electrophoretically homogeneous state with a yield of 24% by hydroxyapatite column chromatography, chromatofocusing and polyacrylamide gel electrophoresis. The purified toxin was dissociated into two subunits with molecular masses of 27 kDa and above 80 kDa, as estimated by Laemmli's sodium dodecylsulfate gel electrophoresis; the exact composition ratio of the two subunits remains unestablished. The isoelectric point was between 4.4 and 4.8. As compared with the reported narrow pH range of action and instability of k1 killer toxin encoded by a double-stranded RNA plasmid of Saccharomyces cerevisiae, the K. Lactis toxin was effective with sensitive strains of S. cerevisiae in a relatively wider pH range between 4 and 8; it was stable for several months at pH 6.0 when stored below -20 degrees C. In contrast to the simple protein nature of the k1 killer toxin with a molecular mass of 11.47 kDa, the K. lactis toxin maintained a mannoprotein nature, as it was absorbed by a ConA-Sepharose column and eluted by methyl alpha-D-mannoside. The growth inhibitory activity of K. lactis toxin was enhanced 2-35-fold by the presence of 4-60% glycerol.

Comparison of the killer toxin of several yeasts and the purification of a toxin of type K2

A total of 13 killer toxin producing strains belonging to the genera Saccharomyces, Candida and Pichia were tested against each other and against a sensitive yeast strain. Based on the activity of the toxins 4 different toxins of Saccharomyces cerevisiae, 2 different toxins of Pichia and one toxin of Candida were recognized. The culture filtrate of Pichia and Candida showed a much smaller activity than the strains of Saccharomyces. Extracellular killer toxins of 3 types of Saccharomyces were concentrated and partially purified. The pH optimum and the isoelectric point were determined. The killer toxins of S. cerevisiae strain NCYC 738, strain 399 and strain 28 were glycoproteins and had a molecular weight of Mr = 16,000. The amino acid composition of the toxin type K2 of S. cerevisiae strain 399 was determined and compared with the composition of two other toxins.

Kluyveromyces lactis killer toxin inhibits adenylate cyclase of sensitive yeast cells

K1 killer toxin secreted by the K1 strain of Saccharomyces cerevisiae, has been well characterized. It is a simple protein of molecular weight (MW) 11,470 (ref. 3), encoded by a double-stranded, linear RNA plasmid, called M RNA, of MW 1.1-1.7 x 10(6) (refs 4-6). It is lethal to sensitive Saccharomyces cerevisiae which does not carry M RNA. Leakage of K+ and ATP is the first distinct response in sensitive cells, and the toxic action is thought to be due to its action as a protonophore or K+ ionophore. Recently, a further killer toxin has been found in Kluyveromyces lactis IFO 1267, and it is associated with the presence of the double-stranded linear DNA plasmids, pGK1-1 (MW 5.4 x 10(6)) and pGK1-2 (MW 8.4 x 10(6)). It has been shown, by curing pGK1-1 or deletion mapping, that the structural gene for the killer toxin and immunity-determining gene reside on the smaller plasmid. Moreover, the plasmids could be transferred from K. lactis to S. cerevisiae by protoplast fusion and protoplast transformation. As the K. lactis toxin is encoded by a DNA plasmid and has a relatively wider action spectrum than K1 killer toxin, the mode of action of the toxin is highly interesting. Here we report that K. lactis toxin inhibits adenylate cyclase in sensitive yeast cells and brings about arrest of the cells at the G1 stage.

Mode of action of yeast killer toxins: channel formation in lipid bilayer membranes

The toxic action of yeast killer proteins seems to involve selective functional damage to the plasma membrane of the sensitive cell. Physiological effects include leakage of K+ (refs 1, 2), inhibition of active transport of amino acids and acidification of the cell interior. These effects are strikingly similar to the effects of certain bacterial colicins which have been demonstrated previously to form channels in membranes. Proposed mechanisms of action have usually postulated a limited permeability change induced by the toxin in the plasma membrane. We report here that a killer toxin from the yeast Pichia kluyveri forms ion-permeable channels in phospholipid bilayer membranes, and we propose that the in vitro electrophysiological properties of these channels account for the morbid effects observed in intoxicated cells. A preliminary account of this work has appeared elsewhere.

A glycosylated protoxin in killer yeast: models for its structure and maturation

The type 1 killer phenotype in S. cerevisiae, mediated by secretion of an 11.5 kilodalton (kd) protein toxin, is cytoplasmically determined by the 1.9 kb M1-dsRNA plasmid. Maintenance of M1-dsRNA is dependent on the 4.5 kb L1-dsRNA because L1 encodes the capsid protein of the virus-like particles that separately encapsidate both dsRNA species. We have shown that in vitro translation of denatured M1-dsRNA produces M1-P1, a 32 kd protein containing the toxin peptides. We now demonstrate the presence of an unstable, 42 kd, membrane-associated, glycosylated protoxin in killer cells, probably derived from M1-P1 by cotranslational processing, and glycosylation. In vitro cotranslational processing of M1-P1, derived both from in vivo mRNAs and from denatured M1-dsRNA, produces a product resembling protoxin. Processing involves loss of 1.6 kd of protein, presumably an N-terminal leader peptide, and glycosylation. This information, together with data on in vitro expression of suppressive deletion mutants of M1-dsRNA, allows construction of testable models for the functional sequence of M1-P1 and for its maturation to toxin.

Yeast killer plasmid mutations affecting toxin secretion and activity and toxin immunity function

M double-stranded RNA (MdsRNA) plasmid mutants were obtained by mutagenesis and screening of a diploid killer culture partially heat cured of the plasmid, so that a high proportion of the cells could be expected to have only on M plasmid. Mutants with neutral (nonkiller [K-], immune [R+]) or suicide (killer [K+], sensitive [R-] phenotypes were examined. All mutants became K- R- sensitives on heat curing of the MdsRNA plasmid, and showed cytoplasmic inheritance by random spore analysis. In some cases, M plasmid mutations were indicated by altered mobility of the MdsRNA by agarose gel electrophoresis or by altered size of in vitro translation products from denatured dsRNA. Neutral mutants were of two types: nonsecretors of the toxin protein or secretors of an inactive toxin. Of three neutral nonsecretors examined, one (NLP-1), probably a nonsense mutation, made a smaller protoxin precursor in vitro and in vivo, and two made full-size protoxin molecules. The in vivo protoxin of 43,000 molecular weight was unstable in the wild type and kinetically showed a precursor-product relationship to the processed, secreted 11,000-molecular-weight toxin. In one nonsecretor (N1), the protoxin appeared more stable in a pulse-chase experiment, and could be altered in a recognition site required for protein processing.

Yeast killer plasmid mutations affecting toxin secretion and activity and toxin immunity function

M double-stranded RNA (MdsRNA) plasmid mutants were obtained by mutagenesis and screening of a diploid killer culture partially heat cured of the plasmid, so that a high proportion of the cells could be expected to have only on M plasmid. Mutants with neutral (nonkiller [K-], immune [R+]) or suicide (killer [K+], sensitive [R-] phenotypes were examined. All mutants became K- R- sensitives on heat curing of the MdsRNA plasmid, and showed cytoplasmic inheritance by random spore analysis. In some cases, M plasmid mutations were indicated by altered mobility of the MdsRNA by agarose gel electrophoresis or by altered size of in vitro translation products from denatured dsRNA. Neutral mutants were of two types: nonsecretors of the toxin protein or secretors of an inactive toxin. Of three neutral nonsecretors examined, one (NLP-1), probably a nonsense mutation, made a smaller protoxin precursor in vitro and in vivo, and two made full-size protoxin molecules. The in vivo protoxin of 43,000 molecular weight was unstable in the wild type and kinetically showed a precursor-product relationship to the processed, secreted 11,000-molecular-weight toxin. In one nonsecretor (N1), the protoxin appeared more stable in a pulse-chase experiment, and could be altered in a recognition site required for protein processing.