Skip to main content
  • Loading metrics

The Species-Specific Acquisition and Diversification of a K1-like Family of Killer Toxins in Budding Yeasts of the Saccharomycotina


Killer toxins are extracellular antifungal proteins that are produced by a wide variety of fungi, including Saccharomyces yeasts. Although many Saccharomyces killer toxins have been previously identified, their evolutionary origins remain uncertain given that many of these genes have been mobilized by double-stranded RNA (dsRNA) viruses. A survey of yeasts from the Saccharomyces genus has identified a novel killer toxin with a unique spectrum of activity produced by Saccharomyces paradoxus. The expression of this killer toxin is associated with the presence of a dsRNA totivirus and a satellite dsRNA. Genetic sequencing of the satellite dsRNA confirmed that it encodes a killer toxin with homology to the canonical ionophoric K1 toxin from Saccharomyces cerevisiae and has been named K1-like (K1L). Genomic homologs of K1L were identified in six non-Saccharomyces yeast species of the Saccharomycotina subphylum, predominantly in subtelomeric regions of the genome. When ectopically expressed in S. cerevisiae from cloned cDNAs, both K1L and its homologs can inhibit the growth of competing yeast species, confirming the discovery of a family of biologically active K1-like killer toxins. The sporadic distribution of these genes supports their acquisition by horizontal gene transfer followed by diversification. The phylogenetic relationship between K1L and its genomic homologs suggests a common ancestry and gene flow via dsRNAs and DNAs across taxonomic divisions. This appears to enable the acquisition of a diverse arsenal of killer toxins by different yeast species for potential use in niche competition.

Author summary

Antifungal killer toxins produced by Saccharomyces yeasts can be found mostly encoded by cytoplasmic double-stranded RNAs (dsRNAs) rather than the DNA genome. A survey of Saccharomyces yeasts has identified a new dsRNA-encoded killer toxin that has a unique antifungal activity and is related in structure and function to the canonical K1 killer toxin. This “K1-like” (K1L) killer toxin was identified to be part of a larger family of DNA-encoded “K1 killer toxin-like” (KKT) genes in more distantly related yeasts. KKT genes encode active killer toxins and appear to have been acquired sporadically during the evolution of each yeast species, with evidence of ongoing gene duplication. The common ancestry of K1L and homologous killer toxins suggests the transfer of these genes via dsRNAs and DNAs between different yeast species that inhabit similar natural and artificial environments. The potential evolutionary advantage of acquiring killer toxins for niche competition rationalizes the ongoing acquisition and diversification of these genes in yeasts.


Many different species of fungi have been observed to produce proteinaceous killer toxins that inhibit the growth of competing fungal species [110]. The killer phenotype was reported in the budding yeast Saccharomyces cerevisiae in 1962, when Bevan et al. observed that spent culture medium had antifungal properties [11]. The potential future application of killer toxins as novel fungicides has led to the discovery of many different killer yeasts with varying toxicities and specificities [12,13]. In the Saccharomyces yeasts, including commonly used laboratory strains, it is estimated that 3–10% are able to produce killer toxins [710]. Despite the number of known killer yeasts that have been identified, a complete understanding of the diversity of killer toxins and their evolutionary history is lacking, even within the model yeast S. cerevisiae. The horizontal transfer of genetic material between different species of yeast is well appreciated [1416] and the acquisition of killer toxins can be associated with the horizontal transfer of DNAs between related species that occupy similar ecological niches [17,18]. There is also evidence to support the sporadic genomic capture of killer toxins that are encoded by DNAs and RNAs that are derived from plasmids, viruses, and satellites [19,20]. These examples suggest that mobilization of killer toxins can occur by different mechanisms and that extrachromosomal elements play a role in gene flow between species.

In general, killer toxin production by S. cerevisiae is most often enabled by infection with double-stranded RNA (dsRNA) totiviruses of the family Totiviridae [2123]. Totiviruses that infect Saccharomyces yeasts are approximately 4.6 kbp in length and encode two proteins, Gag and Gag-Pol (the latter by a programmed -1 frameshift). These proteins are essential for the assembly of viral particles and the replication of viral RNAs [24,25]. Totiviruses act as helper viruses for the replication and encapsidation of ‘M’ satellite dsRNAs, which often encode killer toxins. These satellite dsRNAs are not limited to Saccharomyces yeasts, as they have been identified within other yeasts of the phylum Ascomycota (i.e. Zygosaccharomyces bailii, Torulaspora delbrueckii, and Hanseniaspora uvarum [2628]) and the phylum Basidiomycota (Ustilago maydis and Malassezia sympodialis [2931]). In the Ascomycota, the organization of dsRNA satellites is similar, with all encoding a 5’ terminal sequence motif with the consensus of G(A)4-6, one or more central homopolymeric adenine (poly(A)) tracts, and a 3’ untranslated region (UTR) containing packaging and replication cis-acting elements [24]. In all known satellite dsRNAs, killer toxin genes are positioned upstream of the central poly(A) tract and encode a single open reading frame. Identifying and characterizing dsRNAs is challenging, as the sequencing of dsRNAs currently requires specialized techniques for nucleic acid purification and conversion to cDNAs [3234]. This has limited our understanding of the diversity of dsRNA-encoded killer toxins within fungi.

In Saccharomyces yeasts, there are eight known satellite dsRNA-encoded killer toxins that have been identified (K1, K2, K28, Klus, K21/K66, K45, K62, and K74), with the majority found in the species S. paradoxus [3539]. Owing to their early identification and distinct mechanisms of action, most functional studies have focused on the S. cerevisiae killer toxins K1 and K28. Saccharomyces-associated killer toxin genes appear to be evolutionarily diverse and unrelated by nucleotide and amino acid sequence. Despite the lack of homology, there are similarities in the posttranslational modifications that occur during killer toxin maturation prior to extracellular export of the active toxin. Each killer toxin is expressed as a pre-processed toxin (preprotoxins) with hydrophobic signal peptides that are required for extracellular secretion [40]. These signal peptides are cleaved by a signal peptidase complex in the endoplasmic reticulum. In the case of K1 and K28 toxins, the resulting protoxins are glycosylated and then crosslinked by disulfide bonds in the endoplasmic reticulum. These disulfide bonds are critical for both stability and toxicity [4143]. The disulfide-linked protoxins are further cleaved by carboxypeptidases in the Golgi network to yield mature toxins that are secreted via exocytosis [44]. The mature K1 and K28 toxins can be described as α/β heterodimers that are linked by interchain disulfide bonds. Once outside of the producer cell, mature killer toxins can exert their antifungal activities upon competing fungi.

The K1 toxin was the first to be discovered in S. cerevisiae and the mechanism of action has been studied extensively (reviewed in [45]). K1 is an ionophoric toxin that attacks the cell membrane of susceptible yeast cells and is mechanistically similar to the K2 toxin [46]. Interaction of K1 with a susceptible cell occurs in a two-step process that involves the initial energy-independent binding of the toxin to the β-1,6-D-glucan polysaccharide of the yeast cell wall [47]. After binding surface glucans, K1 translocates to the cell membrane where it interacts with a secondary receptor, Kre1p [48]. Although there is still uncertainty on the exact mechanism of action of K1, it is likely that intoxication is caused by α-domain dependent formation of membrane channels and the subsequent selective leakage of monovalent cations from the cytoplasm [49]. Importantly, killer toxin immunity is provided by the immature protoxin by a poorly understood mechanism. For K1, this activity has been mapped to the α-subunit and 31 amino acids of the adjacent γ-subunit, which is usually removed during toxin maturation [50,51].

This manuscript describes two novel findings related to killer toxin biology. The first is the identification of a novel killer toxin encoded by a satellite dsRNA in S. paradoxus by screening a large collection of Saccharomyces yeasts. This killer toxin has low sequence identity to the canonical K1 toxin produced by S. cerevisiae but has a similar secondary structure and domain organization. Due to its relatedness to K1, we have named this new killer toxin K1-like (K1L). The second major finding is that this is the first example of a dsRNA-encoded killer toxin from Saccharomyces yeasts that has significant homology to DNA-encoded killer toxins that we have named “K1 killer toxin-like” (KKT) within multiple species of the Saccharomycotina. These proteins represent a family of K1-like killer toxins that are both diverse and show possible signs of rapid protein evolution. This work provides insights into the expansion and horizontal transfer of killer toxin genes in yeasts, whether they are encoded by DNAs or mobilized and replicated as dsRNAs by viruses.


Killer phenotype assays

Killer toxin production by yeasts was measured using killer yeast agar plates (yeast extract, peptone, and dextrose (YPD) agar plates with 0.003% w/v methylene blue buffered at pH 4.6), as described previously [32]. Toxin production was identified by either a zone of growth inhibition or methylene blue-staining of the susceptible lawn yeasts. The pH optima of killer toxins produced by different yeasts was measured on killer yeast agar plates adjusted to pH values of 4.0, 4.5, 5.0 and 5.5.

Killer toxin enrichment

Strains of killer yeast were grown in 2 mL of YPD medium (pH 4.6) overnight at room temperature with vigorous shaking (250 rpm). The culture was centrifuged at 3,100 × g for 5 min followed by filtration of the culture medium through a 0.22 μm filter. Filtered growth medium was added 1:1 with 4°C supersaturated ammonium sulfate solution and mixed by inversion and incubated on ice for 3 h. The precipitated proteins were collected by centrifugation at 20,800 × g for 10 min at 4°C. The supernatant was then removed, and the precipitated proteins suspended in 10 μL of YPD pH 4.6. Killer toxins were incubated either at room temperature, or heat-inactivating at 98°C for 2 min before treating lawns of susceptible yeasts spread onto killer assay agar plates.

RNA extraction and purification

Double-stranded RNAs were extracted as described previously [32], with the following modifications: The extracted dsRNAs were not incubated with oligo d(T)25 magnetic beads and 2× LTE buffer was replaced with 2× STE (500 mM NaCl; 20 mM Tris-HCl, pH 8.0; 30 mM EDTA, pH 8.0). Single-stranded RNAs were extracted from yeasts using the RNeasy RNA extraction kit (Qiagen) with the bead beating method and an on column DNase I digestion. After column elution, an additional incubation with DNase I (Omega Bio-Tek) was performed at 37°C for 10 min followed by 75°C for 10 min when the removal of additional contaminating DNAs was required.

Curing Saccharomyces yeasts of dsRNAs

Cycloheximide, anisomycin, or high temperatures were used to create strains of yeasts that lacked satellite dsRNAs. Yeasts were first grown overnight in 2 mL of liquid YPD medium before 1 μL of cell suspension was transferred to YPD agar with either cycloheximide (0.4–5.0 μM) or anisomycin (0.8 μM). Yeast cultures were incubated for 2–5 days at 23°C to recover surviving cells. Curing satellite dsRNAs using temperature involved incubating yeast cultures on YPD agar for 2–5 days at 30°C, 37°C, or 40°C. Growing cells were streaked onto YPD agar plates from the heat-treated agar plates and were incubated for an additional 2–3 days at 23°C. The colonies resulting from chemical or temperature treatment were analyzed for the loss of killer toxin production by replica plating onto killer assay agar plates seeded with a killer toxin sensitive yeast strain.

Short read sequencing of satellite dsRNAs

The preparation of dsRNAs, cDNAs, Nextera Illumina libraries, and sequence analysis was the same as previously reported, with several amendments detailed below [32]. Reads were cleaned with fastp or HTStream ( and de novo contigs were assembled with SPAdes assembler v3.11.1 using default parameters [52,53].

Sanger sequencing of SpV-M1L

Reverse transcriptase PCR (RT-PCR) was used to generate overlapping DNAs that represented the genetic sequence of the satellite dsRNA named SpV-M1L. The approximate molecular weight of these DNAs was determined by agarose gel electrophoresis and capillary electrophoresis (Fragment Analyzer, Agilent Technologies Inc). DNAs were cloned using the pCR-Blunt II-TOPO vector and subjected to Sanger sequencing. 5’ and 3’ RACE were used to determine the terminal ends of the dsRNA molecules using the protocol provided by the manufacturer (Invitrogen) (see S4 File for the full list of primers).

Cloning of genome-encoded killer toxin genes

Genomic DNAs were extracted from K. africana, N. dairenensis, N. castellii, T. phaffii, and P. membranifaciens using the method of Hoffman and Wilson (1987) and were used as templates for PCR (see S4 File for the full list of primers) [54]. Killer toxin genes were cloned into pCR8 by TOPO-TA cloning (Thermo Fisher) and the DNA sequences were confirmed by Sanger sequencing. The K1L gene was commercially synthesized (GeneArt by Thermo Fisher) and used as a PCR template to amplify K1L. The PCR-derived K1L gene was cloned into pCR8 by TOPO-TA cloning and confirmed via Sanger sequencing. All killer toxin genes were sub-cloned using Gateway technology into the destination vector pAG426-Gal-ccdB for ectopic expression in either S. cerevisiae or S. paradoxus (see S5 file for all plasmid sequences) [55,56].

Expression of K1L and homologs from the Ascomycota

For ectopic expression of killer toxins, plasmids encoding toxin genes were used to transform either S. cerevisiae BY4741 or a non-flocculant derivative of S. paradoxus A12 (named A12C) [57]. Transformants were selected on complete medium lacking uracil. To assay toxin expression, a single colony of each transformed strain was used to inoculate a series of consecutive overnight cultures in 2 mL of complete medium lacking uracil, first with dextrose, then raffinose, and finally galactose at 30°C with shaking (250 rpm). The optical density of the final 2 mL culture was normalized to an OD600 of 1 and 1 mL was centrifuged at 3,000 × g for 5 min. The supernatant was removed, and the cell pellet was disrupted by gentle agitation. 2.5 μL of the resulting cell slurry was used to inoculate YPD and YPG plates (with 0.003% w/v methylene blue, pH 4.6) seeded with a killer toxin-susceptible yeast strain. Inoculated plates were incubated for 48–72 h at 25°C until killer toxin production was visible.

Phylogenetic analyses

Killer toxin gene sequences were aligned using MUSCLE and manually trimmed to represent the most confident alignment of the α-domain. MEGA (version 7) was used for phylogenetic analysis using neighbor-joining and maximum likelihood methodologies. The optimal model for amino acid substitution was determined as the Whelan and Goldman model with a gamma distribution. 500 bootstrap replicates were used to construct a phylogenetic model with the highest log-likelihood.


The identification of new strains of killer toxin-producing yeasts

A total of 110 strains of Saccharomyces yeasts were obtained from the USDA Agricultural Research Service (ARS) culture collection and screened to identify the production of novel killer toxins. The first screen used eight yeasts from four different species to indicate toxin production on “killer assay medium”, which is a complex YPD medium buffered to pH 4.6, the optimal pH for most Saccharomyces killer toxins. Methylene blue was also added to the medium as a redox indicator of cell death. This assay found that 22% (24/110) of yeasts could inhibit the growth of at least one strain of yeast (S1 File). To identify the types of killer toxins based on their unique spectrum of activities, 13 killer yeasts were further screened against 46 indicator lawn yeasts. Four strains of S. cerevisiae that have been previously described to produce killer toxins of unknown types were also included (NCYC1001, NCYC190, CYC1058, and CYC1113) [5860]. To facilitate the classification of different toxin types, yeasts that produce canonical K1, K28, and K74 killer toxins, and a non-killer yeast were included for comparison (Fig 1). The degree of growth inhibition by each killer yeast was scored qualitatively based on the appearance of zones of growth inhibition and methylene blue staining of the surrounding indicator strain on agar plates (Fig 1 and S2 File).

Fig 1. The strain- and species-specificity of killer toxins produced by Saccharomyces yeasts.

A total of 21 killer yeasts were assayed for killer toxin production on agar plates seeded with 46 different lawn yeasts. Yeast from the USDA (ARS) culture collection are designated by either ‘Y’ or ‘YB’. Killer toxin activity was qualitatively assessed based on the presence and size of zones of growth inhibition and/or methylene blue staining around killer yeasts as diagrammed (bottom left). Darker colors on the cluster diagram represent a more prominent killer phenotype with yellow indicating no detectable killer phenotype. The non-killer yeast strain S. cerevisiae 1116 was used as a negative control (bottom right). Results were analyzed using the R package gplots to cluster killer yeasts and lawn yeasts based on killer toxin production and susceptibility, respectively. S. cerS. cerevisiae, S. par–S. paradoxus, S. bay–S. bayanus, S. kud–S. kudriavzevii, S. mik–S. mikatae, N. dair–N. dairenensis, N. cas–N. castellii, and K. afr–K. africana.

The known strain- and species-specificity of killer toxins have been previously used to biotype different yeasts [6165] and to identify groups of unknown killer toxins [3,66]. Therefore, cluster analysis was used to group killer yeasts based on their ability to inhibit growth of lawn strains and revealed that no two strains had the exact same spectrum of antifungal activity. S. cerevisiae CYC1113, CYC1058, and NCYC1001 had the broadest antifungal activities but did not cluster with any of the known killer toxins. Clustering found that that the antifungal specificity of S. cerevisiae NCYC190 was most closely correlated with that of a K1 killer yeast, with 85% identical interactions with competing lawn strains. The killer toxin activity of S. paradoxus Y-63717 also clustered with K1 (50% identical interactions), showing both gain and loss of activity. This result was particularly intriguing because evidence of K1 production by S. paradoxus is inconsistent between different research groups with recent studies suggesting that K1 toxins are unique to S. cerevisiae [8,67,68]. The remaining killer yeasts were identified to have weaker inhibitory activities with no clustering with K1, K28, or K74 killer yeasts.

Killer yeasts harbor satellite dsRNAs that encode proteinaceous killer toxins

The production of killer toxins by yeasts is often associated with the presence of satellite dsRNAs that are maintained by totiviruses [25,45]. To determine whether satellite dsRNAs were responsible for the observed killer phenotypes, killer yeasts were treated with either cycloheximide, anisomycin, or incubated at elevated temperatures to select for the loss of killer toxin production [69,70]. The majority of killer yeast (12/19) lost their killer phenotype after exposure to chemical or thermal insult (Fig 2A). The same treatments were unable to select for the loss of the killer phenotype in the remaining seven strains, which suggests that their killer toxins were genome-encoded (S1 Fig). To determine if the killer phenotype was correlated with the presence of dsRNAs, cellulose chromatography was used to selectively purify dsRNAs from the killer yeasts [32]. The analysis of extracted dsRNAs revealed that the loss of the killer phenotype was 100% correlated with the loss of satellite dsRNAs (Fig 2A).

Fig 2. Analysis of dsRNAs present within killer yeasts and the properties of their killer toxins.

(A) The extraction and analysis of dsRNAs from totiviruses and associated satellites in Saccharomyces yeasts that produced killer toxins (top panel) or lost the killer phenotype after exposure to cycloheximide, anisomycin, or elevated temperatures (bottom panel). S. cerevisiae Y-27788 is included as a representative killer yeast that lacks a satellite dsRNA. (B) The relative pH optimum of killer toxins against three different indicator lawn yeasts S. cerevisiae YB-4237, DSM70459, and NBRC10991 using YPD buffered to pH 4.0, 4.5, 5.0, and 5.5. (C) Enrichment and concentration of killer toxins from spent growth medium by ammonium sulfate precipitation and their loss of inhibitory activity after incubation at 98°C.

The dsRNA-encoded killer toxins identified have a pH optimum mostly between 4.5 and 5, with no inhibitory activity at pH 5.5 (Fig 2B). To confirm that the identified killer toxins are proteinaceous, each was purified by ammonium sulfate precipitation and used to challenge susceptible yeasts. Zones of inhibition were clearly visible on confluent lawns of yeast cells for all of the killer toxins tested but not for a non-killer yeast or a strain cured of its dsRNA satellite (Fig 2C). The inhibitory activities of these precipitates were heat-labile, and the toxicity was lost after incubation at 98°C for 2 minutes. The identified killer toxins have similar biochemical characteristics to known proteinaceous killer toxins despite their differing inhibitory effects towards yeasts.

The discovery of a new killer toxin produced by S. paradoxus

To identify the unknown killer toxins produced by killer yeasts, dsRNAs were purified and subjected to a short-read sequencing pipeline for dsRNAs [32]. BLASTn analysis of de novo assembled contigs revealed that the satellite dsRNAs within strains CYC1058 and NCYC1001 encode canonical K2 toxins and NCYC190 a canonical K1 toxin (S2 Fig and S1 Table). The sequence reads derived from the dsRNAs of Y-63717 assembled into 125 different contigs, with six >750 bp in length and a coverage score >1,000 (Fig 3A). BLASTn analysis of these high-quality contigs identified the dsRNA genome of the totivirus L-A-45 from S. paradoxus N-45 with 100% coverage and 95.5% nucleotide identity [36]. However, the remaining contigs did not match the nucleotide sequence of any known Saccharomyces yeast killer toxin or dsRNA satellite. Short-read sequencing was augmented with a combination of 5’ and 3’ RACE, RT-PCR, Sanger sequencing, and capillary electrophoresis to assemble the complete sequence of the dsRNA satellite from Y-63717 (Figs 3B and S3). The novel satellite dsRNA is approximately 2,371 bp in length with a single open reading frame (ORF) that encodes a protein of 340 amino acids. The 5’ ORF is positioned upstream of a central poly(A) tract of ~220 bp (Figs 3B and S3). The 5’ terminus has a nucleotide sequence of 5’-GAAAAA that is found in many satellite dsRNAs (Figs 3B and S3) and is predicted to fold into a large stem-loop structure (S4 Fig). Downstream of the poly(A) tract in the 3’ UTR there are elements of secondary structure that are indicative of replication (terminal recognition element; TRE) and packaging signals (viral binding site; VBS) that have been well characterized in the canonical M1 satellite dsRNA from S. cerevisiae (ScV-M1) (Figs 3B and S4) [7173]. In addition to RNA secondary structures, there are also two direct repeats of the sequence motif named “Downstream of Poly(A)” (DPA; 5’-CTCACCYTGAGNHTAACTGG-3’) that is found in different satellite dsRNAs isolated from S. paradoxus (M45, M74, and M62), S. cerevisiae (M1 and Mlus), Zygosaccharomyces bailii (MZb), and Torulaspora delbrueckii (Mbarr) (Figs 3B and S3) [27,74].

Fig 3. Short-read sequencing and analysis of the K1L killer toxin from S. paradoxus Y-63717.

(A) Sequence contigs after de novo assembly of sequence reads represented by contig coverage score and contig length. BLASTx analysis was used to annotate contigs as similar to totiviruses or satellite dsRNAs. (B) Schematic of the organization of the SpV-M1L satellite dsRNA (C) Jpred secondary structure prediction of the K1 and K1L killer toxins with confidence score plotted as a histogram. Predicted domain boundaries are drawn below the secondary structure prediction [76]. Signal peptidase cleavage site were predicted by SignalP and TargetP (S3 File). (D) Ectopic expression of K1L from a plasmid in the non-killer yeast S. paradoxus A12 induced by the presence of galactose in the growth medium.

The length and positioning of the 5’ ORF of the satellite dsRNA in S. paradoxus Y-63717 suggested that it encoded a killer toxin (Fig 3B). A PSI-BLAST search of the NCBI database with two iterations found that this putative killer toxin had weak homology to the canonical K1 toxin from S. cerevisiae (99% coverage, 18% amino acid identity, e-value 4 × 10−17) (Fig 4A and S2 Table). Based on this homology, the putative killer toxin was named K1L (K1-like) and the dsRNA satellite was named Saccharomyces paradoxus virus M1-like (SpV-M1L). The organization of the functional domains of K1L appears to be similar to K1 based on the secondary structure, conserved cysteine residues, and the predicted signal peptidase cleavage sites (Fig 3C) [75]. K1L contains ten cysteine residues, two of which are likely important for interchain disulfide linkage (Cys91) and killer toxin immunity (Cys257) based on their alignment with cysteines from K1 [41]. To confirm that the K1L is an active killer toxin, it was ectopically expressed by the non-killer strain S. paradoxus A12C using a galactose inducible promoter. A well-defined zone of growth inhibition was visible when the strain was grown on galactose-containing medium (Fig 3D). No K1L toxin expression was observed when cells were plated on dextrose-containing growth medium. Together, these data confirm the identification of a new dsRNA satellite in S. paradoxus and a novel killer toxin related to K1.

K1L homologs are found in yeasts of the Saccharomycotina

The PSI-BLAST search that identified K1L as a homolog of K1 also identified 24 hypothetical “K1-like Killer Toxin” (KKT) genes in six diverse species of non-Saccharomyces yeasts from the subphylum Saccharomycotina (Figs 4A and S5). The species Kazachstania africana, Naumovozyma castellii, Naumovozyma dairenensis, Tetrapisispora phaffii, and Pichia membranifaciens encode KKT genes that are homologous to K1L (aligned >300 amino acids, e-value <10−80) and represent yeasts from the families Saccharomycetaceae and Pichiaceae (Fig 4A and S3 Table) [77]. Importantly, genomic KKT genes appear to be unique to these particular species and absent from other related yeasts (i.e. Kazachstania naganishii, Tetrapisispora blattae, and Pichia kudriavzevii). The length of all Kkt proteins are between 153–390 amino acids with 11 being similar in length (340 amino acids) (Fig 4A) and amino acid identity (16–28%) to K1L (S2 Table). In addition, BLASTn was used to identify 14 additional pseudogenes that, in some species, outnumber intact KKT genes (Fig 4B and S3 Table). All KKT genes and related pseudogenes are found in multiple copies that vary in frequency between different yeast species and are mostly located within subtelomeric regions (within ~20 kb of the assembled chromosome ends) (Fig 4B). Of the 38 KKT genes and pseudogenes found within six different species, only two are positioned away from the subtelomeric regions in the yeasts N. dairenensis and N. castellii (Fig 4B). Analysis of the chromosomal position of these two genes revealed that their insertions are unique to each species, absent from other related species at the syntenic chromosomal location, and inserted close to tRNAs (S6 Fig). It was noted that six of the KKT genes and pseudogenes from K. africana contained a characteristic “GAAAAA” sequence motif close to the start codon of each ORF, similar to K1L and other killer toxins (S6 Fig).

Fig 4. KKT genes are distributed across the subphylum Saccharomycotina and are primarily subtelomeric in their genomic location.

(A) PSI-BLAST analysis of the killer toxin encoded upon the satellite dsRNA from S. paradoxus Y-63717. (B) Location of KKT genes and pseudogenes on the linear representation of the chromosomes of four yeasts of the Saccharomycetaceae. Chromosome I of N. castellii did not contain any KKT genes and was truncated in the diagram due to its large size. (C) Unrooted maximum likelihood phylogeny of the aligned α-domain of 21 killer toxins from six species of yeast and one dsRNA satellite. Numerical values represent the bootstrap support for the placement of each node. The domain organization of each protein is illustrated and annotated based on the four-domain structure of K1. Signal peptidase cleavage site predicted by SignalP and TargetP (S3 File). Underlined killer toxin names represent those that were cloned and functionally tested for antifungal activity. *Killer toxins with confirmed antifungal activities.

To ascertain the evolutionary history and relatedness of K1L and its homologs, a multiple amino acid sequence alignment was constructed. The most confident alignment was achieved between the putative α-domain of each protein and included eight truncated proteins with premature stop codons. Phylogenetic analysis was performed using maximum likelihood (Fig 4C) and neighbor-joining (S7 Fig) methodologies with 500 bootstrap iterations as implemented by MEGA [78]. The tree topologies of the two phylogenetic models are in general agreement and show the distinct clustering of killer toxins from each species (Figs 4C and S7). Importantly, K1L is most closely related to the Kkt1 proteins from T. delbrueckii and P. membranifaciens. The Kkt proteins from K. africana form a monophyletic clade that contrasts those from T. phaffii and N. dairenensis that appear to be polyphyletic. There also appears to be some support for a common ancestry of killer toxins within the Naumovozyma yeasts. BLASTn analysis and nucleotide alignment of the 5’ and 3’ UTRs of the KKT genes from K. africana indicates that flanking nucleotide sequence is between 83–94% identical over ~2,000 bp, suggesting gene duplication. There is also evidence of gene duplication of KKT genes within N. dairenensis based on their close phylogenetic relationships and similar untranslated regions. To ascertain the evolutionary trajectory of KKT paralogs, the rate of accumulation of nonsynonymous (dN) and synonymous (dS) mutations were calculated for KKT paralogs that could be confidently aligned by their nucleotide sequences. When dN/dS was calculated for all codons over the evolutionary history of the sequence pairs, all three pairs of paralogs have evolved under purifying selection since their duplication (dN/dS = 0.72, 0.41, 0.23) [79]. A domain-resolution approach using a sliding window to calculate dN/dS supports the predominance of purifying selection with a few localized peaks of possible weak positive selection (dN/dS >1) in the α- and/or γ-domains (S8 Fig).

A new family of K1-like antifungal killer toxins

The relatedness of the KKT genes to the killer toxin K1L suggests that they encode antifungal killer toxins. To assay for killer toxin production by the yeasts that encode KKT genes, P. membranifaciens, N. dairenensis, N. castellii, T. delbrueckii, T. phaffii, and K. africana were used to challenge 19 different yeast on killer assay plates. With the exception of N. castellii, all of the KKT-encoding yeast species produced killer toxins that caused growth inhibition of at least one other yeast (Fig 5A and 5B). Each of these killer yeasts was also immune to its own killer toxin, but susceptible to those produced by other KKT-encoding yeasts. The production of killer toxins by these species is consistent with the previously reported killer activity of P. membranifaciens, T. delbrueckii, and T. phaffii [27,80,81]. There was no evidence of satellite dsRNAs in any of the KKT-encoding yeasts, except for the detection of an unknown high molecular weight dsRNA within P. membranifaciens NCYC333 (S9 Fig). The differences in killer toxin production by five strains of P. membranifaciens suggested that there could be strain-specific polymorphisms in KKT genes (S10 Fig). The published genome sequence of P. membranifaciens Y-2026 revealed a large central deletion in the γ-domain of its KKT gene (S10 Fig). Sanger sequencing of the same KKT gene from P. membranifaciens Y-2026 acquired directly from the NRRL culture collection failed to identify the same deletion, instead there was an indel within the γ-domain that caused the truncation of the killer toxin gene (S10 Fig). Sequencing of the KKT gene from P. membranifaciens NCYC333 confirmed a full-length gene that correlated with robust killer toxin production by the strain. However, P. membranifaciens Y-2026 was still able to express killer toxins, suggesting the production of other alternative antifungal molecules.

Fig 5. The antifungal activity of Kkt toxins from K. africana, N. dairenensis, and P. membranifaciens.

(A) KKT-encoding yeasts were assayed for killer toxin production on killer assay agar plates seeded with indicator strains. Killer toxin production was judged by the presence of zones of growth inhibition or methylene blue staining (see the key in Fig 1). (B) Representative pictures of killer toxin production by different species of yeast from the Saccharomycotina. (C) PCR detection of KKT gene DNA and RT-PCR detection of RNA transcripts under laboratory conditions. S. paradoxus Y-63717 served as a dsRNA control. (D) Galactose-dependent ectopic expression of KKT genes from K. africana, P. membranifaciens and N. dairenensis can inhibit the growth of K. africana. N. dai–N. dairenensis, N. cas–N. castellii, K. afr–K. africana, P. mem–P. membranifaciens, T. pha–T. phaffii, S. cer–S. cerevisiae, S. mik–S. mikatae, S. par–S. paradoxus, and T. del–T. delbrueckii.

Although P. membranifaciens, N. dairenensis, T. delbrueckii, T. phaffii, and K. africana are killer yeasts (Fig 5A and 5B), it was unclear whether KKT genes were directly responsible for the observed production of killer toxins. Indeed, T. phaffii has been reported to express an antifungal glucanase and K2 killer toxin-related genes have been found in the genome of K. africana [74,80]. To confirm the expression of KKT genes in killer yeasts, total genomic DNA and RNA were extracted from six species of killer yeasts and used as templates for PCR and RT-PCR with primers targeting ten different KKT genes (S11 Fig). Due to the similarity of the KKT genes it was not possible to unambiguously distinguish between transcripts from TpKKT1 and TpKKT2, or KaKKT1 and KaKKT2. Of the six species assayed, KKT RNA transcripts were detected in T. phaffii, N. dairenensis, and K. africana indicating active gene transcription under laboratory conditions (Fig 5C). To demonstrate that KKT genes are active killer toxins with antifungal activities, a subset was cloned into a galactose inducible expression vector (labelled in Fig 4C). Active killer toxin production from a non-killer strain of S. cerevisiae transformed with KKT genes was assayed against eight lawns of yeasts using galactose containing agar plates. The majority of KKT genes did not cause any noticeable growth inhibition or methylene blue staining of competing yeasts when expressed from S. cerevisiae (S12 Fig). However, killer toxins from P. membranifaciens NCYC333 (PmKKT1), K. africana (KaKKT2), and N. dairenensis (NdKKT1) were able to create visible zones of growth inhibition (Fig 5D). No growth inhibition was observed when gene expression was repressed by plating cells on dextrose. Altogether, these data show that KKT genes are transcriptionally active and encode active killer toxins. This confirms the discovery of a new family of K1-like genome-encoded antifungal proteins in the Saccharomycotina.


The most significant finding of this study is the discovery of a novel satellite dsRNA that encodes a killer toxin related to K1 and a larger family of DNA-encoded homologs in Saccharomycotina yeasts. The relatedness of killer toxins encoded on dsRNAs and DNAs suggests that the origins of K1L are outside of the Saccharomyces genus, with killer toxin gene mobilization and interspecific transfer by dsRNAs. These killer toxins have been shown to be biologically active and are diverse in their amino acid sequences with evidence of their rapid evolution by gene duplication and elevated rates of non-synonymous mutations. This demonstrates the likely benefits of killer toxin acquisition and the ongoing mobilization of these genes between divergent species of yeasts. The more specific implications of our findings are discussed below.

Horizontal acquisition and copy number expansion of killer toxins in fungi

KKT genes have most likely been acquired by horizontal gene transfer because of their sporadic distribution and lack of common ancestry in closely related yeast species. Moreover, the lack of relatedness of KKT genes in these species suggests that they were acquired independently. Fungi are known to acquire foreign DNAs from other species of fungi [1418,8284] as well as bacteria [14,85,86]. The interspecific capture of DNAs derived from retrotransposons, viruses, and plasmids has also been observed [19,20,87,88]. Specifically, genome integrated copies of dsRNA-encoded killer toxins homologous to KP4, K1, K2, Klus, and Kbarr have been found within bacteria and fungi, indicating gene flow between dsRNAs and DNAs across taxonomic divisions [74,89]. However, the vast majority of these putative killer toxins are uncharacterized, and it remains unclear as to whether they are biologically active.

Phylogenetic evidence suggests that cross-species transmission of viruses between fungi has occurred on multiple independent occasions [90,91]. Laboratory experiments have also successfully demonstrated extracellular [92,93] and interspecific virus transfer [36]. In particular for Saccharomyces yeasts, mating and hybridization between different species has been observed, and is a mechanism for gene introgression, as well as for the acquisition of retrotransposons and plasmids. The close association of many yeast species in natural and anthropic habitats may increase the likelihood of horizontal gene transfer or invasion by dsRNA viruses and satellites [9496]. Specifically, the satellite dsRNA that was identified in this study (named SpV-M1L) and an unrelated satellite dsRNA (SpV-M45) are both found within sympatric Far Eastern yeast strains of S. paradoxus [36,97]. The parasitism of L-A-45 by both of these satellite dsRNAs in different strains of yeast strongly implies that they were acquired by horizontal gene transfer. The apparent structural similarities of the satellite dsRNA-encoded VBS elements suggest that, despite the evolutionary distance between SpV-M1L and SpV-M45, each has evolved a similar mechanism to enable the hijack of the replication and packaging machinery from the L-A-45 totivirus. Unlike Saccharomyces yeasts, the presence of active RNAi within the KKT-encoding yeasts would hinder the horizontal acquisition of dsRNA viruses from other yeasts, which could explain the abundance of genome-encoded killer toxins and paucity of dsRNA viruses and satellites (S5 Fig) [98,99]. However, T. delbrueckii can support the replication of the putative totivirus TdV-LAbarr and an associated satellite [27], which suggests that this strain has lost its RNAi machinery or the totivirus has the ability to evade or disrupt RNAi similar to other mycoviruses [100102]. The apparent invasion of these yeasts by viruses could enable the acquisition of killer toxins encoded upon satellite dsRNAs.

To capture killer toxin genes from dsRNAs, the erroneous reverse transcription of mRNAs by endogenous retrotransposons would allow their insertion into a yeast genome by the retrotransposon integrase protein [103]. The ancestral KKT gene in K. africana could have been captured directly from dsRNAs as all extant paralogs encode a 5’UTR “GAAAAA” motif, which is characteristic of satellite dsRNAs. The 5’ UTR is often conserved during mRNA capture by retrotransposons [103]. While the majority of KKT genes are subtelomeric in their location, we have observed that non-subtelomeric KKT genes from N. dairenensis and N. castellii have been uniquely inserted into genomic loci near tRNA genes. The genomic integration of retrotransposon cDNAs is selectively targeted to tRNA genes and many extrachromosomal nucleic acids are identified at loci adjacent to tRNAs [19]. This means the potential mobilization of these genes by retrotransposons and direct insertion by integrase or cellular DNA repair mechanisms [104]

The potential benefits of killer toxin acquisition

Acquisition of foreign nucleotide sequences can be associated with adaptation to a specific environmental niche, including genes associated with nutrient acquisition, virulence, the stress response, and interference competition (e.g. allelopathy) [18,19,84]. Killer toxin-like genes have been identified within the genomes of many different species of fungi and could represent selection for functional killer toxins to improve competitive fitness. Indeed, the production of killer toxins by different species of yeasts has been consistently shown to provide a competitive advantage, particularly in a spatially structured environment at an optimal pH for killer toxin activity [105110]. However, competition between different killer yeasts likely selects for locally adapted populations that are immune to the predominant killer toxins in a specific environmental niche [6]. The laboratory evolution of killer yeast populations has also shown that killer toxin exposure increases the prevalence of killer toxin resistance [111]. High levels of killer toxin resistance perhaps account for the low prevalence of killer yeasts within natural yeast populations and that killer toxin production might not always be advantageous. However, the rise of killer toxin resistance within a population could perhaps drive the acquisition of new killer toxins or the subfunctionalization of existing killer toxins to maintain a selective advantage. In the context of the multicopy dsRNA satellites and expanded killer toxin gene families in yeasts, it would seem logical to assume that natural variation would provide a pool of genetic diversity on which natural selection could act.

The majority of KKT genes are found to have been acquired within the subtelomeric regions of chromosomes that would facilitate gene expansion due to elevated rates of homologous recombination between telomeric repeat sequences. As has been noted for the subtelomeric MAL gene family, gene duplication enables evolutionary innovation that is also evident in KKT paralogs by possible weak signatures of positive selection [112]. Other genome-encoded fungal killer toxin families have also undergone copy number expansion and are experiencing elevated rates of non-synonymous substitutions [89,113]. Both KP4-like and Zymocin-like killer toxin families appear to have roles during antagonistic interactions with plants and fungi, respectively, which could drive the continued evolution of novel killer toxins. In addition to the expansion of KKT genes in yeasts, there is also gene loss and pseudogenization. KKT gene inactivation is biased towards truncations that leave the α-domain and a small portion of the γ-domain. This same α/γ region of K1 is the minimal sequence required for functional killer toxin immunity that could provide a selective advantage by protecting yeasts from exogenous killer toxins related to KKT [50,51]. KKT yeasts all encode C-terminally truncated KKT genes and are mostly resistant to other KKT killer yeasts. However, despite encoding several truncated KKT genes, K. africana appears to be naturally susceptible to many of the killer toxins produced by other KKT encoding yeasts, including its own killer toxin when ectopically expressed by S. cerevisiae. Subtelomeric Sir-dependent gene silencing could account for this apparent susceptibility under laboratory conditions [114,115]. This would prevent the expression of KKT genes and associated immunity by K. africana, although transcripts from full-length KKT genes were detected in K. africana.

K1L and its homologs represent a unique example of the mobilization and subtelomeric expansion of DNA- and dsRNA-encoded killer toxin genes in different species of yeasts. The diversity of dsRNAs and the known prevalence of killer yeasts suggests that more killer toxins await future discovery and characterization. This will enable a better understanding of their role in fungal biology and provide greater insights into the mechanisms of horizontal gene transfer in eukaryotes.

Supporting information

S1 Fig. Double-stranded RNAs extracted from killer yeasts that retained their killer phenotype after exposure to cyclohexamide (CHX).

Agarose gel electrophoresis was used to show dsRNAs present in killer yeasts. Satellite dsRNAs are labeled as dsRNAs that are smaller than the associated totivirus dsRNAs. The high molecular weight DNAs are predicted to be mitochondrial in origin.


S2 Fig. Coverage and contig lengths of dsRNAs from different strains of killer yeasts.

The scatter plots represent all contigs generated after de novo assembly of sequence reads. M satellites are labeled in red according to their relatedness to other previously described sequences.


S3 Fig. Sequence analysis of the dsRNA satellite M1L from S. paradoxus Y-63717.

(A) Coverage of the assembled sequence reads for M1L and the positioning of the expected products from three RT-PCR reactions to amplify portions of the K1L ORF (reaction 1), the 3’ UTR (reaction 2), and across the internal poly(A) tract (reaction 3). PCR products shown by agarose gel electrophoresis and their estimated sizes as determined from fragment analysis are represented. (B) The positioning of the repeated DPA element is represented relative to the genomes of eight dsRNA satellites. (C) The consensus sequence derived from 15 DPA elements is shown as a sequence logo and multisequence alignment.


S4 Fig. RNA secondary structure predictions of M1L, M1, and M45 (+) strand 5’ and 3’ termini.

(A) Secondary structure prediction of the 5’ terminal structures. Start codons for the translation of preprotoxin synthesis are highlighted by a grey outline. Numbers represent nucleotides from the 5’ terminus. (B) Putative replication signal represented as a stem-loop at the 3’ end of M1L and M1 satellite. Numbers represent distance from the 3’ terminal nucleotide. (C) Putative viral binding sites (VBS) with a 5’ facing ‘A’ bulge present in the stem-loops (indicated by an arrow). Numbers represent distance from the 3’ terminal nucleotide. The mFold server was used to calculate the change in free energy for each structure [116].


S5 Fig. Cladogram of selected budding yeasts of the Saccharomycotina.

The panel details the presence or absence of totiviruses, satellite dsRNAs, the killer phenotype, KKT genes, and RNAi within 15 species. ‘?’ denotes the uncertainty of the putative dsRNA virus detected in P. membranifaciens in S8 Fig. WGD indicates the ancestral yeast species that underwent a whole genome duplication.


S6 Fig. The unique genomic insertion of KKT genes in N. dairenensis and N. castellii.

(A) N. dairenensis KKT2 inserted into chromosome III and N. castellii KKT3 inserted into chromosome II. Genes flanking KKT insertions are represented as black arrows and demonstrate synteny between related genomes. Single red triangles represent tRNA genes. Broken lines represent gaps in synteny. (B) 5’ UTR sequence from KKT genes and one pseudogene identified within K. africana.


S7 Fig. Phylogenetic model of the evolutionary relationship between Kkt proteins using the neighbor-joining method.

Unrooted neighbor-joining phylogeny of the aligned α-domain of 21 KKT proteins from six species of yeast and K1L from one dsRNA satellite (SpV-M1L). Numerical values represent the bootstrap support for the placement of each node.


S8 Fig. Paralogous KKT genes are potentially evolving under weak positive selection in different species of yeasts.

Three sliding window dN/dS calculations are shown for the comparison of three pairs of closely related genes in three yeast species. The x-axis represents the nucleotide (nt) number of each gene and is shown in the context of the predicted domain organization of each pair of genes. Omega values represent the whole gene dN/dS value for each gene pair. Sliding windows analysis was performed using a window of 90 nucleotides with a 30 nucleotide overlap. Asterisks in the T. phaffii dN/dS plot represent instances where dS = 0 and dN > 0.


S9 Fig. The absence of satellite dsRNAs within killer yeasts of the Saccharomycotina.

Agarose gel electrophoresis of dsRNAs extracted from different killer yeasts. Stained bands in lane 2 represent a canonical totivirus and satellite dsRNA from S. cerevisiae.


S10 Fig. Strain-specific production of killer toxins by P. membranifaciens.

(A) Killer toxin production and partial purification from P. membranifaciens. (B) Mutations within KKT1 in the context of the proteins secondary structure organization (as predicted by Jpred) from strain Y-2026 compared to a full-length active killer toxin sequenced from strain NCYC333. Arrows represent β-sheets and cylinders represent α-helices. “c” represents cysteine residues.


S11 Fig. The specificity of primers targeting KKT genes.

Where possible, primers pairs were designed to recognize different full-length KKT paralogs from T. phaffii, N. dairenensis, and K. africana. DNA plasmids were used as templates to test their specificity. Due the similarities in DNA sequences, specific primer pairs were not identified for TpKKT1, TpKKT2, and KaKKT2.


S12 Fig. Ectopic expression of KKT genes causes the inhibition of K. africana and is dependent on the induction of KKT expression.

Galactose-dependent ectopic expression of KKT genes from a multicopy plasmid by S. cerevisiae on agar plates seeded with different species of yeasts. Key: 1. pUI114 (NcKKT1), 2. pUI109 (TpKKT2), 3. pUI110 (TpKKT1), 4-pUI111 (TpKKT3), 5. pUI112 (NdKKT3), 6. pUI113 (NdKKT2), 7. pML115 (NdKKT1), 8. pML117 (KaKKT2), 9. pML118 (KaKKT1), 10. pML116 (PmKKT1), Nc. N. castellii NCYC2898, Nd. N. dairenensis NCYC777, Tp. T. phaffii Y-8282, Pm. P. membranifaciens NCYC333.


S1 Table. Nucleotide and amino acid identity of killer toxins from different strains of S. cerevisiae compared to canonical K1 and K2.


S2 Table. Amino acid identity and similarity between K1, K1L and Kkt proteins.


S3 Table. Species that encode genomic killer toxins that are homologous to K1 and K1L.


S1 File. A large-scale screen to identify killer yeasts in the Saccharomyces genus.


S2 File. Image data illustrating the susceptibility of yeast to a selection of potent killer toxins produced by Saccharomyces yeasts.


S3 File. SignalP and TargetP predictions for K1, K1L and Kkt proteins.


S4 File. Supplementary file listing all primers, plasmids, and yeast strains used in this study.


S5 File. Supplementary FASTA file with the DNA sequences of all plasmids used in this study.



We would like to acknowledge the NRRL Agricultural Research Service culture collection for providing the Rowley laboratory with a diverse collection of Saccharomyces yeasts. We would also like to thank Dr. Antonia Santos (Complutense Yeast Collection, Complutense University of Madrid) for the CYC yeasts, Prof. Manfred J. Schmitt (Saarland University, Saarbrücken, Germany) for S. cerevisiae MS300c, Dr. Gianni Liti (University of Nice) for the genetically tractable strains of S. paradoxus and S. cerevisiae. We would also like to acknowledge Dr. Gianni Liti and Dr. Scott Minnich for helpful comments on the draft manuscript. Opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the funders.


  1. 1. Stumm C, Hermans J, Middelbeek EJ, Croes AF, Vries G de. Killer-sensitive relationships in yeasts from natural habitats. Antonie van Leeuwenhoek. 1977;43: 125–128. pmid:596861
  2. 2. Rosini G. The occurrence of killer characters in yeasts. Can J Microbiol. 1983;29: 1462–1464. pmid:6661706
  3. 3. Young TW, Yagiu M. A comparison of the killer character in different yeasts and its classification. Antonie van Leeuwenhoek. 1978;44: 59–77. pmid:655699
  4. 4. Buzzini P, Martini A. Biodiversity of killer activity in yeasts isolated from the Brazilian rain forest. Can J Microbiol. 2000;46: 607–611. pmid:10932353
  5. 5. Kandel JS, Stern TA. Killer phenomenon in pathogenic yeast. Antimicrob Agents Chemother. 1979;15: 568–571. pmid:572656
  6. 6. Starmer WT, Ganter PF, Aberdeen V, Lachance MA, Phaff HJ. The ecological role of killer yeasts in natural communities of yeasts. Can J Microbiol. 1987;33: 783–796. pmid:3690423
  7. 7. Nakayashiki T, Kurtzman CP, Edskes HK, Wickner RB. Yeast prions [URE3] and [PSI+] are diseases. Proc Natl Acad Sci U S A. 2005;102: 10575–10580. pmid:16024723
  8. 8. Chang S-L, Leu J-Y, Chang T-H. A population study of killer viruses reveals different evolutionary histories of two closely related Saccharomyces sensu stricto yeasts. Mol Ecol. 2015;24: 4312–4322. pmid:26179470
  9. 9. Pieczynska MD, Visser JAGM de, Korona R. Incidence of symbiotic dsRNA ‘killer’ viruses in wild and domesticated yeast. FEMS Yeast Res. 2013;13: 856–859. pmid:24028530
  10. 10. Philliskirk G, Young TW. The occurrence of killer character in yeasts of various genera. Antonie van Leeuwenhoek. 1975;41: 147–151. pmid:239627
  11. 11. Bevan EA, Makower M. The physiological basis of the killer character in yeast. Proc. Int. Congr. Genet; 1963.
  12. 12. Schaffrath R, Meinhardt F, Klassen R. Yeast killer toxins: fundamentals and applications. In: Anke T, Schüffler A. (eds) Physiology and Genetics. The Mycota (A Comprehensive Treatise on Fungi as Experimental Systems for Basic and Applied Research), vol 15. Springer, Cham.
  13. 13. Spadaro D, Droby S. Development of biocontrol products for postharvest diseases of fruit: The importance of elucidating the mechanisms of action of yeast antagonists. Trends Food Sci Tech. 2016;47: 39–49.
  14. 14. Shen X-X, Opulente DA, Kominek J, Zhou X, Steenwyk JL, Buh KV, et al. Tempo and mode of genome evolution in the budding yeast Subphylum. Cell. 2018;175: 1533–1545.e20. pmid:30415838
  15. 15. Peter J, Chiara MD, Friedrich A, Yue J-X, Pflieger D, Bergström A, et al. Genome evolution across 1,011 Saccharomyces cerevisiae isolates. Nature. 2018;556: 339–344. pmid:29643504
  16. 16. Dujon BA, Louis EJ. Genome diversity and evolution in the budding yeasts (Saccharomycotina). Genetics. 2017;206: 717–750. pmid:28592505
  17. 17. Cheeseman K, Ropars J, Renault P, Dupont J, Gouzy J, Branca A, et al. Multiple recent horizontal transfers of a large genomic region in cheese making fungi. Nat Commun. 2014;5: 2876. pmid:24407037
  18. 18. Novo M, Bigey F, Beyne E, Galeote V, Gavory F, Mallet S, et al. Eukaryote-to-eukaryote gene transfer events revealed by the genome sequence of the wine yeast Saccharomyces cerevisiae EC1118. Proc Natl Acad Sci U S A. 2009;106: 16333–8. pmid:19805302
  19. 19. Frank AC, Wolfe KH. Evolutionary capture of viral and plasmid DNA by yeast nuclear chromosomes. Eukaryot Cell. 2009;8: 1521–1531. pmid:19666779
  20. 20. Satwika D, Klassen R, Meinhardt F. Repeated capture of a cytoplasmic linear plasmid by the host nucleus in Debaryomyces hansenii. Yeast. 2012;29: 145–154. pmid:22434608
  21. 21. Herring AJ, Bevan EA. Virus-like particles associated with double-stranded-RNA species found in killer and sensitive strains of yeast Saccharomyces cerevisiae. J Gen Virol. 1974;22: 387–394. pmid:4594342
  22. 22. Vodkin MH, Fink GR. A nucleic acid associated with a killer strain of yeast. Proc Natl Acad Sci. 1973;70: 1069–1072. pmid:4577789
  23. 23. Berry EA, Bevan EA. A new species of double-stranded RNA from yeast. Nature. 1972;239: 279–280. pmid:4562032
  24. 24. Wickner RB, Fujimura T, Esteban R. Viruses and prions of Saccharomyces cerevisiae. Adv Virus Res. 2013;86: 1–36. pmid:23498901
  25. 25. Rowley PA. The frenemies within: viruses, retrotransposons and plasmids that naturally infect Saccharomyces yeasts. Yeast. 2017;49: 111–292. pmid:28387035
  26. 26. Schmitt MJ, Neuhausen F. Killer toxin-secreting double-stranded RNA mycoviruses in the yeasts Hanseniaspora uvarum and Zygosaccharomyces bailii. J Virol. 1994;68: 1765–1772. pmid:8107238
  27. 27. Ramìrez M, Velázquez R, Maqueda M, López-Piñeiro A, Ribas JC. A new wine Torulaspora delbrueckii killer strain with broad antifungal activity and its toxin-encoding double-stranded RNA virus. Front Microbiol. 2015;6: 403–12. pmid:26042092
  28. 28. Zorg J, Kilian S, Radler F. Killer toxin producing strains of the yeasts Hanseniaspora uvarum and Pichia kluyveri. Arch Microbiol. 1988;149: 261–267.
  29. 29. Puhalla JE. Compatibility reactions on solid medium and interstrain inhibition in Ustilago maydis. Genetics. 1968;60: 461–474. pmid:5731752
  30. 30. Kandel J, Koltin Y. Killer phenomenon in Ustilago maydis. Comparison of the killer proteins. Genetics; 1978.
  31. 31. Clancey SA, Ruchti F, LeibundGut-Landmann S, Heitman J, Ianiri G. A Novel mycovirus evokes transcriptional rewiring in the fungus malassezia and stimulates beta interferon production in macrophages. Mbio. 2020;11. pmid:32873760
  32. 32. Crabtree AM, Kizer EA, Hunter SS, Leuven JTV, New DD, Fagnan MW, et al. A rapid method for sequencing double-stranded RNAs purified from yeasts and the identification of a potent k1 killer toxin isolated from Saccharomyces cerevisiae. Viruses. 2019;11: 70. pmid:30654470
  33. 33. Potgieter AC, Page NA, Liebenberg J, Wright IM, Landt O, Dijk AA van. Improved strategies for sequence-independent amplification and sequencing of viral double-stranded RNA genomes. J Gen Virol. 2009;90: 1423–1432. pmid:19264638
  34. 34. Decker CJ, Steiner HR, Hoon-Hanks LL, Morrison JH, Haist KC, Stabell AC, et al. dsRNA-Seq: Identification of viral infection by purifying and sequencing dsRNA. Viruses. 2019;11: 943–21. pmid:31615058
  35. 35. Rodríguez-Cousiño N, Maqueda M, Ambrona J, Zamora E, Esteban R, Ramìrez M, et al. A new wine Saccharomyces cerevisiae killer toxin (Klus), encoded by a double-stranded RNA virus, with broad antifungal activity is evolutionarily related to a chromosomal host gene. Applied and Environmental Microbiology Appl Environ Microbiol. 2011;77: 1822–1832. pmid:21239561
  36. 36. Rodríguez-Cousiño N, Gómez P, Esteban R. Variation and distribution of L-A helper totiviruses in Saccharomyces sensu stricto yeasts producing different killer toxins. Toxins. 2017;9: 313–20. pmid:29019944
  37. 37. Bostian KA, Elliott Q, Bussey H, Bum V, Smith AI, Tipper DJ. Sequence of the preprotoxin dsRNA gene of type I killer yeast: Multiple processing events produce a two-component toxin. Cell. 1984;36: 741–751. pmid:6697395
  38. 38. Schmitt MJ, Tipper DJ. Sequence of the M28 dsRNA: preprotoxin is processed to an α/β heterodimeric protein toxin. Virology. 1995;213: 341–351. pmid:7491759
  39. 39. Dignard D, Whiteway M, Germain D, Tessier D, Thomas DY. Expression in yeast of a cDNA copy of the K2 killer toxin gene. Mol Gen Genetics. 1991;227: 127–136. pmid:2046653
  40. 40. Sturley SL, Elliot Q, LeVitre J, Tipper DJ, Bostian KA. Mapping of functional domains within the Saccharomyces cerevisiae type 1 killer preprotoxin. EMBO J. 1986;5: 3381–3389. pmid:3545818
  41. 41. Gier S, Lermen M, Schmitt MJ, Breinig F. Substitution of cysteines in the yeast viral killer toxin K1 precursor reveals novel insights in heterodimer formation and immunity. Sci Rep-uk. 2019;9: 13127. pmid:31511600
  42. 42. Suzuki Y, Schwartz SL, Mueller NC, Schmitt MJ. Cysteine residues in a yeast viral A/B toxin crucially control host cell killing via pH-triggered disulfide rearrangements. Mol Biol Cell. 2017;28: 1123–1131. pmid:28228551
  43. 43. Zhu H, Bussey H. Mutational analysis of the functional domains of yeast K1 killer toxin. Mol Cell Biol. 1991;11: 175–181. pmid:1702512
  44. 44. Bussey H, Saville D, Greene D, Tipper DJ, Bostian KA. Secretion of Saccharomyces cerevisiae killer toxin: processing of the glycosylated precursor. Mol Cell Biol. 1983;3: 1362–1370. pmid:6353202
  45. 45. Schmitt MJ, Breinig F. Yeast viral killer toxins: lethality and self-protection. Nat Rev Microbiol. 2006;4: 212–221. pmid:16489348
  46. 46. Novotná D, Flegelová H, Janderová B. Different action of killer toxins K1 and K2 on the plasma membrane and the cell wall of Saccharomyces cerevisiae. FEMS Yeast Res. 2004;4: 803–813. pmid:15450187
  47. 47. Hutchins K, Bussey H. Cell wall receptor for yeast killer toxin: involvement of (1,6)-beta-D-glucan. J Bacteriol. 1983;154: 161–169. pmid:6300031
  48. 48. Breinig F, Tipper DJ, Schmitt MJ. Kre1p, the plasma membrane receptor for the yeast K1 viral toxin. Cell. 2002;108: 395–405. pmid:11853673
  49. 49. Martinac B, Zhu H, Kubalski A, Zhou XL, Culbertson M, Bussey H, et al. Yeast K1 killer toxin forms ion channels in sensitive yeast spheroplasts and in artificial liposomes. Proc Natl Acad Sci. 1990;87: 6228–6232. pmid:1696721
  50. 50. Boone C, Bussey H, Greene D, Thomas DY, Vernet T. Yeast killer toxin: Site-directed mutations implicate the precursor protein as the immunity component. Cell. 1986;46: 105–113. pmid:3521889
  51. 51. Zhu YS, Kane J, Zhang XY, Zhang M, Tipper DJ. Role of the gamma component of preprotoxin in expression of the yeast K1 killer phenotype. Yeast. 1993;9: 251–66. pmid:8488726
  52. 52. Bankevich A, Nurk S, Antipov D, Gurevich AA, Dvorkin M, Kulikov AS, et al. SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing. J Comput Biol. 2012;19: 455–477. pmid:22506599
  53. 53. Chen S, Zhou Y, Chen Y, Gu J. fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics. 2018;34: i884–i890. pmid:30423086
  54. 54. Hoffman CS, Winston F. A ten-minute DNA preparation from yeast efficiently releases autonomous plasmids for transformation of Escherichia coli. Gene. 1987;57: 267–272. pmid:3319781
  55. 55. Alberti S, Gitler AD, Lindquist S. A suite of Gateway cloning vectors for high-throughput genetic analysis in Saccharomyces cerevisiae. Yeast. 2007;24: 913–919. pmid:17583893
  56. 56. Walhout AJ, Temple GF, Brasch MA, Hartley JL, Lorson MA, Heuvel S van den, et al. GATEWAY recombinational cloning: application to the cloning of large numbers of open reading frames or ORFeomes. Methods Enzymol. 2000;328: 575–592. pmid:11075367
  57. 57. Cubillos FA, Louis EJ, Liti G. Generation of a large set of genetically tractable haploid and diploid Saccharomyces strains. FEMS Yeast Res. 2009;9: 1217–1225. pmid:19840116
  58. 58. Santos A, Sanchez A, Marquina D. Yeasts as biological agents to control Botrytis cinerea. Microbiol Res. 2004;159: 331–338. pmid:15646379
  59. 59. Young TW. The genetic manipulation of killer character into brewing yeast. J I Brewing. 1981;87: 292–295.
  60. 60. Walker GM, McLeod AH, Hodgson VJ. Interactions between killer yeasts and pathogenic fungi. FEMS Microbiol Lett. 1995;127: 213–222. pmid:7758935
  61. 61. Polonelli L, Morace G, Dettori G, Cattel C. Biotyping of micelial fungus cultures by the killer system. Eur J Epidemiol. 1987;3: 237–242. pmid:3653352
  62. 62. Buzzini P, Turchetti B, Vaughan-Martini AE. The use of killer sensitivity patterns for biotyping yeast strains: the state of the art, potentialities and limitations. FEMS Yeast Res. 2007;7: 749–760. pmid:17425671
  63. 63. Buzzini P, Martini A. Discrimination between Candida albicans and other pathogenic species of the genus Candida by their differential sensitivities to toxins of a panel of killer yeasts. J Clin Microbiol. 2001. pmid:11526179
  64. 64. Polonelli L, Morace G, Archibusacci C, Sestito M. Killer system: a simple method for differentiating Candida albicans strains. J Clin Microbiol. 1983;17: 774–780. pmid:6345575
  65. 65. Fuentefria AM, Faganello J, Pazzini F, Schrank A, Valente P, Vainstein M. Typing and patterns of cellular morphological alterations in Cryptococcus neoformans and Cryptococcus gattii isolates exposed to a panel of killer yeasts. Med Mycol. 2007;45: 503–512. pmid:17710619
  66. 66. Rogers D, Bevan EA. Group classification of killer yeasts based on cross-reactions between strains of different species and origin. Microbiology. 1978;105: 199–202.
  67. 67. Rodríguez-Cousiño N, Esteban R. Relationships and evolution of double-stranded RNA totiviruses of yeasts inferred from analysis of L-A-2 and L-BC variants in wine yeast strain populations. Cullen D, editor. Appl Environ Microb. 2017;83: e02991–16–18. pmid:27940540
  68. 68. Pieczynska MD, Wloch-Salamon D, Korona R, Visser JAGM de. Rapid multiple-level coevolution in experimental populations of yeast killer and nonkiller strains. Evolution. 2016;70: 1342–1353. pmid:27168531
  69. 69. Wickner RB. “Killer character” of Saccharomyces cerevisiae: curing by growth at elevated temperature. J Bacteriol. 1974;117: 1356–7. pmid:4591956
  70. 70. Fink GR, Styles CA. Curing of a killer factor in Saccharomyces cerevisiae. Proc Natl Acad Sci. 1972;69: 2846–. pmid:4562744
  71. 71. Esteban R, Fujimura T, Wickner RB. Site-specific binding of plus replicase-containing particles of yeast. Proc Natl Acad Sci. 1988;85: 4411–4415. pmid:3288994
  72. 72. Esteban R, Fujimura T, Wickner RB. Internal and terminal cis-acting sites are necessary for in vitro replication of the L-A double-stranded RNA virus of yeast. EMBO J. 1989;8: 947–954. pmid:2656262
  73. 73. Fujimura T, Esteban R, Esteban LM, Wickner RB. Portable encapsidation signal of the L-A double-stranded RNA virus of S. cerevisiae. Cell. 1990;62: 819–828. pmid:2117501
  74. 74. Ramìrez M, Velázquez R, López-Piñeiro A, Naranjo B, Roig F, Llorens C. New insights into the genome organization of yeast killer viruses based on “atypical” killer strains characterized by high-throughput sequencing. Toxins. 2017;9. pmid:28925975
  75. 75. Petersen TN, Brunak S, Heijne G von, Nielsen H. SignalP 4.0: discriminating signal peptides from transmembrane regions. Nat Methods. 2011;8: 785–786. pmid:21959131
  76. 76. Drozdetskiy A, Cole C, Procter J, Barton GJ. JPred4: a protein secondary structure prediction server. Nucleic Acids Res. 2015;43: W389–W394. pmid:25883141
  77. 77. Shen X-X, Zhou X, Kominek J, Kurtzman CP, Hittinger CT, Rokas A. Reconstructing the backbone of the Saccharomycotina yeast phylogeny using genome-scale data. G3 (Bethesda). 2016;6: 3927–3939. pmid:27672114
  78. 78. Hall BG. Building phylogenetic trees from molecular data with MEGA. Mol Biol Evol. 2013;30: 1229–1235. pmid:23486614
  79. 79. Hurst LD. The Ka/Ks ratio: diagnosing the form of sequence evolution. Trends Genet. 2002;18: 486. pmid:12175810
  80. 80. Comitini F, Mannazzu I, Ciani M. Tetrapisispora phaffii killer toxin is a highly specific β-glucanase that disrupts the integrity of the yeast cell wall. Microb Cell Fact. 2009;8: 55. pmid:19860881
  81. 81. Santos A. Killer toxin of Pichia membranifaciens and its possible use as a biocontrol agent against grey mould disease of grapevine. Microbiology (Reading). 2004;150: 2527–2534. pmid:15289549
  82. 82. Kellis M, Patterson N, Endrizzi M, Birren B, Lander ES. Sequencing and comparison of yeast species to identify genes and regulatory elements. Nature. 2003;423: 241–254. pmid:12748633
  83. 83. Marsit S, Mena A, Bigey F, Sauvage F-X, Couloux A, Guy J, et al. Evolutionary advantage conferred by an eukaryote-to-eukaryote gene transfer event in wine yeasts. Mol Biol Evol. 2015;32: 1695–1707. pmid:25750179
  84. 84. Ropars J, Rodríguez de la Vega RC, López-Villavicencio M, Gouzy J, Sallet E, Dumas É, et al. Adaptive horizontal gene transfers between multiple cheese-associated fungi. Curr Biol. 2015;25: 2562–2569. pmid:26412136
  85. 85. Marcet-Houben M, Gabaldón T. Acquisition of prokaryotic genes by fungal genomes. Trends Genet. 2010;26: 5–8. pmid:19969385
  86. 86. Dujon B, Sherman D, Fischer G, Durrens P, Casaregola S, Lafontaine I, et al. Genome evolution in yeasts. Nature. 2004;430: 35–44. pmid:15229592
  87. 87. Carr M, Bensasson D, Bergman CM. Evolutionary genomics of transposable elements in Saccharomyces cerevisiae. Plos One. 2012;7: e50978. pmid:23226439
  88. 88. Strope PK, Kozmin SG, Skelly DA, Magwene PM, Dietrich FS, McCusker JH. 2 micron plasmid in Saccharomyces species and in Saccharomyces cerevisiae. FEMS Yeast Res. 2015;15: fov090–8. pmid:26463005
  89. 89. Brown DW. The KP4 killer protein gene family. Curr Genet. 2011;57: 51–62. pmid:21116630
  90. 90. Myers JM, Bonds AE, Clemons RA, Thapa NA, Simmons DR, Carter-House D, et al. Survey of early-diverging lineages of fungi reveals abundant and diverse mycoviruses. Mbio. 2020;11. pmid:32900807
  91. 91. Liu H, Fu Y, Xie J, Cheng J, Ghabrial SA, Li G, et al. Evolutionary genomics of mycovirus-related dsRNA viruses reveals cross-family horizontal gene transfer and evolution of diverse viral lineages. Bmc Evol Biol. 2012;12: 91. pmid:22716092
  92. 92. El-Sherbeini M, Bostian KA. Viruses in fungi: infection of yeast with the K1 and K2 killer viruses. Proc Natl Acad Sci. 1987;84: 4293–4297. pmid:3295880
  93. 93. Yu X, Li B, Fu Y, Xie J, Cheng J, Ghabrial SA, et al. Extracellular transmission of a DNA mycovirus and its use as a natural fungicide. Proc Natl Acad Sci. 2013;110: 1452–1457. pmid:23297222
  94. 94. González SS, Alcoba-Flórez J, Laich F. Lachancea lanzarotensis sp. nov., an ascomycetous yeast isolated from grapes and wine fermentation in Lanzarote, Canary Islands. Int J Syst Evol Micr. 2013;63: 358–363. pmid:23125316
  95. 95. Zhang B-Q, Luan Y, Duan C-Q, Yan G-L. Use of Torulaspora delbrueckii co-fermentation with two Saccharomyces cerevisiae strains with different aromatic characteristic to improve the diversity of red wine aroma profile. Front Microbiol. 2018;9: 606. pmid:29674999
  96. 96. Sampaio JP, Gonçalves P. Natural populations of Saccharomyces kudriavzevii in Portugal are associated with oak bark and are sympatric with S. cerevisiae and S. paradoxus. Appl Environ Microb. 2008;74: 2144–2152. pmid:18281431
  97. 97. Naumov GI, Naumova ES, Sniegowski PD. differentiation of European and far east Asian populations of Saccharomyces paradoxus by allozyme analysis. Int J Syst Bacteriol. 1997;47: 341–344. pmid:9103619
  98. 98. Drinnenberg IA, Fink GR, Bartel DP. Compatibility with killer explains the rise of RNAi-deficient fungi. Science. 2011;333: 1592–1592. pmid:21921191
  99. 99. Wolfe KH, Armisén D, Proux-Wera E, ÓhÉigeartaigh SS, Azam H, Gordon JL, et al. Clade- and species-specific features of genome evolution in the Saccharomycetaceae. FEMS Yeast Res. 2015;15: fov035. pmid:26066552
  100. 100. Segers GC, Wezel R van, Zhang X, Hong Y, Nuss DL. Hypovirus papain-like protease p29 suppresses RNA silencing in the natural fungal host and in a heterologous plant system. Eukaryot Cell. 2006;5: 896–904. pmid:16757737
  101. 101. Yaegashi H, Yoshikawa N, Ito T, Kanematsu S. A mycoreovirus suppresses RNA silencing in the white root rot fungus, Rosellinia necatrix. Virology. 2013;444: 409–416. pmid:23896640
  102. 102. Lau SKP, Lo GCS, Chow FWN, Fan RYY, Cai JJ, Yuen K-Y, et al. Novel partitivirus enhances virulence of and causes aberrant gene expression in Talaromyces marneffei. mBio. 2018;9: 800–18. pmid:29895639
  103. 103. Derr LK, Strathern JN, Garfinkel DJ. RNA-mediated recombination in S. cerevisiae. Cell. 1991;67: 355–364. pmid:1655280
  104. 104. Moore JK, Haber JE. Capture of retrotransposon DNA at the sites of chromosomal double-strand breaks. Nature. 1996;383: 644–646. pmid:8857544
  105. 105. McBride R, Greig D, Travisano M. Fungal viral mutualism moderated by ploidy. Evol Int J Org Evol. 2008;62: 2372–80. pmid:18564376
  106. 106. Wloch-Salamon DM, Gerla D, Hoekstra RF, Visser JAGM de. Effect of dispersal and nutrient availability on the competitive ability of toxin-producing yeast. Proc Biol Sci. 2008;275: 535–541. pmid:18182371
  107. 107. Greig D, Travisano M. Density-dependent effects on allelopathic interactions in yeast. Evolution. 2008;62: 521–527. pmid:17983463
  108. 108. Pintar J, Starmer WT. The costs and benefits of killer toxin production by the yeast Pichia kluyveri. Antonie van Leeuwenhoek. 2003;83: 89–97. pmid:12755485
  109. 109. Ganter PF, Starmer WT. Killer factor as a mechanism of interference competition in yeasts associated with cacti. Ecology. 1992;73: 54–67.
  110. 110. Maule AP, Thomas PD. Strains of yeast lethal to brewery yeasts. J I Brewing. 1973;79: 137–141.
  111. 111. Buskirk SW, Rokes AB, Lang GI. Adaptive evolution of a rock-paper-scissors sequence along a direct line of descent. Biorxiv. 2019; 700302.
  112. 112. Brown CA, Murray AW, Verstrepen KJ. rapid expansion and functional divergence of subtelomeric gene families in yeasts. Curr Biol. 2010;20: 895–903. pmid:20471265
  113. 113. Tzelepis G, Karlsson M. Killer toxin-like chitinases in filamentous fungi: Structure, regulation and potential roles in fungal biology. Fungal Biol Rev. 2018;33: 123–132.
  114. 114. Hecht A, Strahl-Bolsinger S, Grunstein M. Spreading of transcriptional repressor SIR3 from telomeric heterochromatin. Nature. 1996;383: 92–6. pmid:8779721
  115. 115. Ai W, Bertram PG, Tsang CK, Chan T-F, Zheng XFS. Regulation of subtelomeric silencing during stress response. Mol Cell. 2002;10: 1295–1305. pmid:12504006
  116. 116. Zuker M. Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res. 2003;31: 3406–3415. pmid:12824337