Abstract
Zymocin is a Kluyveromyces lactis protein toxin composed of αβγ subunits encoded by the cytoplasmic virus-like element k1 and functions by αβ-assisted delivery of the anticodon nuclease (ACNase) γ into target cells. The toxin binds to cells' chitin and exhibits chitinase activity in vitro that might be important during γ import. Saccharomyces cerevisiae strains carrying k1-derived hybrid elements deficient in either αβ (k1ORF2) or γ (k1ORF4) were generated. Loss of either gene abrogates toxicity, and unexpectedly, Orf2 secretion depends on Orf4 cosecretion. Functional zymocin assembly can be restored by nuclear expression of k1ORF2 or k1ORF4, providing an opportunity to conduct site-directed mutagenesis of holozymocin. Complementation required active site residues of α's chitinase domain and the sole cysteine residue of β (Cys250). Since βγ are reportedly disulfide linked, the requirement for the conserved γ C231 was probed. Toxicity of intracellularly expressed γ C231A indicated no major defect in ACNase activity, while complementation of k1ΔORF4 by γ C231A was lost, consistent with a role of β C250 and γ C231 in zymocin assembly. To test the capability of αβ to carry alternative cargos, the heterologous ACNase from Pichia acaciae (P. acaciae Orf2 [PaOrf2]) was expressed, along with its immunity gene, in k1ΔORF4. While efficient secretion of PaOrf2 was detected, suppression of the k1ΔORF4-derived k1Orf2 secretion defect was not observed. Thus, the dependency of k1Orf2 on k1Orf4 cosecretion needs to be overcome prior to studying αβ's capability to deliver other cargo proteins into target cells.
INTRODUCTION
The protein toxin zymocin, produced by the yeast Kluyveromyces lactis, was identified as the first known anticodon nuclease toxin from a eukaryote, selectively cleaving tRNA within the anticodon loop due to highly specific anticodon nuclease (ACNase) activity (1). Zymocin production is correlated with the presence of a pair of linear cytoplasmic genetic double-stranded DNA (dsDNA) elements, termed pGKL1 (k1 in short) and pGKL2 (k2 in short) (2). The larger k2 is required for cytoplasmic maintenance of k1, which encodes the toxin as well as an immunity determinant (2, 3). k2 provides essential functions for cytoplasmic replication, transcription, and transcript processing, and several of these components show phylogenetic proximity to viruses (4, 5). Since the proposed mode of replication via protein priming is typically found in viruses, the cytoplasmic linear dsDNA elements were termed virus-like elements (VLE) (6). The zymocin toxin is a heterotrimeric αβγ complex, the smallest subunit of which (γ) exhibits the cytotoxic ACNase activity (1, 7–9). The α and β subunits are generated from the 128-kDa k1ORF2 gene product, which is first translocated to the endoplasmic reticulum (ER), where it becomes glycosylated and cleaved by signal peptidase and then travels to the Golgi apparatus, where it is processed by the Kex1/2 endopeptidase internally at the N-terminal end to produce mature α (99-kDa) and β (30-kDa) subunits (7, 9–11). In the active, secreted holotoxin, β appears to be linked via a disulfide bond to the toxic γ subunit (7, 9). The latter is unable to act on target cells on its own since it relies on the chitin binding α subunit and the hydrophobic β subunit to transit into the target cell (8, 9). The process of ACNase cargo dropoff by the αβ carrier is poorly understood, but it possibly involves chitin binding and degradation mediated by the chitinase active site located within the α subunit (12, 13). The existence of conserved αβ-related carrier subunits in other VLE-encoded protein toxins, which shuttle cargo proteins that are dissimilar in primary sequence and/or target RNA, may suggest a general protein translocase ability of αβ (14–16). However, mechanistic studies of zymocin action have been largely restricted to the isolated γ subunit, since no system to generate altered variants of holozymocin was available, which is in part due to the complex genetic basis of cytoplasmic k1/k2 elements, which cannot be manipulated by standard approaches applicable to nuclear genes. To overcome this problem, we have now generated k1/k2-carrying zymocin expression strains in which either the gene encoding the αβ subunits (k1ORF2) or the gene encoding the γ subunit (k1ORF4) is deleted and altered variants of the missing genes can be supplied from standard nuclear expression vectors. Since this system enables the rapid generation of holozymocin variants containing site-specific exchanges, functional studies can be conducted. We demonstrate the usefulness of this mutagenesis system by analyzing the interdependence of αβ and γ subunits for efficient toxin secretion and the importance of chitinase active sites and potential disulfide bond-forming cysteine residues for holotoxin function.
MATERIALS AND METHODS
Strains and media.
Yeast strains employed in this study are listed in Table 1. Strains were grown in YPD (2% glucose, 2% peptone, and 1% yeast extract) or YNB (0.67% yeast nitrogen base without amino acids and carbohydrate and with ammonium sulfate and 2% glucose) supplemented with l-leucine (30 μg/ml), l-histidine (20 μg/ml), l-tryptophan (20 μg/ml), or uracil (20 μg/ml) at 30°C.
TABLE 1.
Strains used in this study
| Strain | Genotype | Reference |
|---|---|---|
| Kluyveromyces lactis AWJ137 | pGKL1+ (k1), pGKL2+ (k2) | 37 |
| Pichia acaciae NRRL Y-18665 | pPac1-1+, pPac1-2+ | 38 |
| Saccharomyces cerevisiae BY4741 | MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 | EUROSCARF |
| Saccharomyces cerevisiae CEN.PK2-1c | MATa ura3-52 leu2-3,112 his3Δ1 trp1-289 MAL-2-8c SUC2 | 39 |
| Saccharomyces cerevisiae CEN.PK2-1c pYEX-BX | Like S. cerevisiae CEN.PK2-1c, with pYEX-BX | This work |
| Saccharomyces cerevisiae CEN.PK2-1c PGAL1::ORF4 | Like S. cerevisiae CEN.PK2-1c, with pKL-BX | This work |
| Saccharomyces cerevisiae CEN.PK2-1c PGAL1::ORF4C231A | Like S. cerevisiae CEN.PK2-1c, with pKL-BX-C231A | This work |
| Saccharomyces cerevisiae CEN.PK2-1c PGAL1::ORF4A231C | Like S. cerevisiae CEN.PK2-1c, with pKL-BX-A231C | This work |
| Saccharomyces cerevisiae 301 (F102-2 ura3) | MATα his4-519 leu23,112 can1 ura3, k1 (pGKL1+), k2 (pGKL2+) | 29 |
| Saccharomyces cerevisiae 301 ΔpGKL | Like S. cerevisiae 301, plasmid cured, k1−, k2− | 29 |
| Saccharomyces cerevisiae 301 k1ΔORF4 | Like S. cerevisiae 301, k1ΔORF4, k2 | This work |
| Saccharomyces cerevisiae 301 k1ΔORF4 YEplac195 | Like S. cerevisiae 301 k1ΔORF4, with YEplac195 | This work |
| Saccharomyces cerevisiae 301 k1ΔORF4 PADH1::ORF4 | Like S. cerevisiae 301 k1ΔORF4, with p195-PADH1::ORF4 | This work |
| Saccharomyces cerevisiae 301 k1ΔORF4 PADH1::ORF4C231A | Like S. cerevisiae 301 k1ΔORF4. with p195-PADH1::ORF4–C231A | This work |
| Saccharomyces cerevisiae 301 k1ΔORF4 PADH1::ORF4A231C | Like S. cerevisiae 301 k1ΔORF4, with p195-PADH1::ORF4–A231C | This work |
| Saccharomyces cerevisiae 301 k1ΔORF4 PO4PO2 | Like S. cerevisiae 301 k1ΔORF4, with TU-PO4PO2-T | This work |
| Saccharomyces cerevisiae 301 k1ΔORF4 PO4PO2* | Like S. cerevisiae 301 k1ΔORF4, with TU-PO4PO2*-T | This work |
| Saccharomyces cerevisiae 301 k1ΔORF2 (MS1608) | Like S. cerevisiae 301, k1ΔORF2, k2 | 29 |
| Saccharomyces cerevisiae 301 k1ΔORF2 YEplac195 | Like S. cerevisiae 301 k1ΔORF2, with YEplac195 | This work |
| Saccharomyces cerevisiae 301 k1ΔORF2 PADH1::ORF2 | Like S. cerevisiae 301 k1ΔORF2, with p195-PADH1::ORF2 | This work |
| Saccharomyces cerevisiae 301 k1ΔORF2 PADH1::ORF2D462A | Like S. cerevisiae 301 k1ΔORF2, with p195-PADH1::ORF2–D462A | This work |
| Saccharomyces cerevisiae 301 k1ΔORF2 PADH1::ORF2A462D | Like S. cerevisiae 301 k1ΔORF2, with p195-PADH1::ORF2–A462D | This work |
| Saccharomyces cerevisiae 301 k1ΔORF2 PADH1::ORF2D464A | Like S. cerevisiae 301 k1ΔORF2, with p195-PADH1::ORF2–D464A | This work |
| Saccharomyces cerevisiae 301 k1ΔORF2 PADH1::ORF2A464D | Like S. cerevisiae 301 k1ΔORF2, with p195-PADH1::ORF2–A464D | This work |
| Saccharomyces cerevisiae 301 k1ΔORF2 PADH1::ORF2E466A | Like S. cerevisiae 301 k1ΔORF2, with p195-PADH1::ORF2–E466A | This work |
| Saccharomyces cerevisiae 301 k1ΔORF2 PADH1::ORF2A466E | Like S. cerevisiae 301 k1ΔORF2, with p195-PADH1::ORF2–A466E | This work |
| Saccharomyces cerevisiae 301 k1ΔORF2 PADH1::ORF2D464AE466A | Like S. cerevisiae 301 k1ΔORF2, with p195-PADH1::ORF2–D464A-E466A | This work |
| Saccharomyces cerevisiae 301 k1ΔORF2 PADH1::ORF2A464D-A466E | Like S. cerevisiae 301 k1ΔORF2, with p195-PADH1::ORF2–A464D-A466E | This work |
| Saccharomyces cerevisiae 301 k1ΔORF2 PADH1::ORF2C250A | Like S. cerevisiae 301 ΔKO2, with p195-PADH1::ORF2–C250A | This work |
| Saccharomyces cerevisiae 301 k1ΔORF2 PADH1::ORF2A250C | Like S. cerevisiae 301 ΔKO2, with p195-PADH1::ORF2–A250C | This work |
Isolation of DNA and linear plasmids.
Bulk DNA and linear plasmids were isolated as previously described (17) or by the minilysate method, which includes a proteinase K treatment (18).
DNA manipulation, cloning, and transformation.
Restriction and DNA ligations were performed with enzymes obtained from New England BioLabs GmbH (Frankfurt am Main, Germany) according to the supplier's recommendations. Escherichia coli was transformed, following standard procedures (19). Yeast transformation was performed following the polyethylene glycol (PEG)/lithium acetate method (20).
Curing.
Since yeast cells carrying linear cytoplasmic plasmids can efficiently be cured by UV irradiation (21, 22), approximately 2 × 103 cells of Saccharomyces cerevisiae 301 grown overnight in YPD at 30°C were plated on YPD agar and exposed to UV light essentially as described previously (23, 24). Following incubation for 24 h at 30°C, arising colonies of S. cerevisiae 301 ΔpGKL were analyzed for the presence of linear elements by gel electrophoresis and Southern analysis.
Killer toxin assays.
For eclipse assays (25), killer strains were point inoculated on YPD at pH 6.5. After incubation for 16 h at room temperature, an overnight culture of a sensitive yeast strain was diluted with sterile water to yield an optical density at 600 nm (OD600) of 0.1, from which a 10-μl sample was spotted onto the medium directly at the rim of the colony of the putative killer strain. After incubation for 16 h at 30°C, growth inhibition became evident with the formation of clear halos.
Microtiter assays, which are more sensitive than the eclipse assays, were performed as described previously (26). Briefly, yeasts were cultured in 200 ml YPD at pH 6.5 and 30°C. Partial purification of toxins was done by ultrafiltration using concentrators (Vivaspin 20; Sartorius Stedim Biotech GmbH, Göttingen, Germany). Since the calculated molecular mass of the chitinases encoded by the VLEs is approximately 129 kDa, centrifugal units with a 30-kDa-cutoff membrane were used. In the case of the P. acaciae toxin (PaT) (∼180 kDa), centrifugal concentrators with an exclusion size of 100 kDa were applied. Sterile toxin preparations were stored at 4°C prior to use. The concentrated samples were diluted in YPD medium to give final concentrations ranging from 0.1 up to 10-fold with respect to the original supernatants. Thus, the relative concentration factor (RCF) of 1 corresponds to the toxin amount in nonconcentrated culture supernatants (27, 28). After incubation for 16 h at 30°C, relative growth was determined spectrometrically at 620 nm (Multiscan FC; Thermo Fisher Scientific Oy, Vantaa, Finland) and refers to the OD value of strains incubated in toxin-free YPD medium.
Southern analysis.
Bulk DNA preparations from S. cerevisiae 301, the k1ΔORF4 mutant, and the VLE-cured ΔpGKL strain were separated on 0.8% agarose gels. A probe was generated from k1ORF4 using the primers k1ORF4-for and k1ORF4-rev (see Table 3). Labeling was performed using the DIG-High Prime kit (Roche Diagnostics, Mannheim, Germany), following the manufacturer's recommendations.
TABLE 3.
Primers used in this study
| Function and primer | Sequence (5′–3′)a |
|---|---|
| Southern analysis | |
| k1ORF4-for | TATATTTAGTGTTTGTTATC |
| k1ORF4-rev | AATTAAATCATCATGACCTTTATC |
| Construction of vector pARS | |
| RKF-pARS-fw-SacI | GCTAGCATGTGAGCTCGGATCTTTTTCTAATAAATATATAC |
| RKF-pARS-rv-KpnI/NheI | CGTACGATCGGGTACCGCTAGCCTGTAGATTATTCATACTATC |
| pSK-RKF-SacII-fw | GTATATAACAAAATAGCACCGCGGTCCAATGAAAGAAATAAATTTG |
| pSK-RKF-SacII-rv | CTTTCATTGGACCGCGGTGCTATTTTGTTATATACAAGTTCCATATAC |
| XbaI-stop-UCS-LEU2*-fw | GCATGCTACGTCTAGATAAATATGATATTTTTATTTTAAATAATAATGCATGCCCCTAAGAAGATCGTCG |
| LEU2*-rev-MCS-XbaI | CGATCGTAATTCTAGACCGCGGGCGGCCGCACTAGTGGATCCCCCGGGCTGCAGAAGCTTATCGATCTCGAGGGCCCGTGGTGCCCTCCTCCTTGTC |
| Expression of γ subunits and immunity protein | |
| TIR2 | AAAGTTGGGTTTTTAAGCTAATAAAAGTTG |
| PADH1-fw | CCGGGTGTACAATATGGAC |
| PADH1-NdeI-rv | GGGATAGACATATGATATGAGATAGTTGATTGTATGC |
| PO4-fw | GACCTTAGTGATGTATCAAAATTGAATGG |
| PO2-rv | TCCAGGATTAACCGAACAAG |
| k1ORF4-NdeI-for | GAATTCATATGAAGATATATCATATATTTAG |
| k1ORF4-EcoRI-rev | TAAGTCGAATTCTTATACACATTTTCCATTCTGTAGATTATTC |
| PO4-HAsc-fw | CTAAGCATAATCTGGAACATCATAAGGATAAATATTGTTAAAATAAGGATTAAGCTCATCCC |
| PO2-MYCsc-rev | CTACAAATCTTCTTCAGAAATCAACTTTTGTTCAACCTTACATGTAATACTTTTGATTTTACTGTC |
| PO4-3HA Extender for | CTAGCCAGCATAATCAGGAACATCATAAGGATAGCCAGCATAATCTGGAACATCATAAGGATAAGCATAATCTGGAACATC |
| PO2–3myc-Extender rev | CTAGTTCAAGTCTTCTTCTGAGATTAATTTTTGTTCACCGTTCAAGTCTTCCTCGGAGATTAGCTTTTGTTCACCGTTCAAATCTTCTTCAGAAAT |
| HA-extender | TTAAGAAGCGTAATCTGGAACGTCATACGGATAGGATGCATAGTCCGGGACGTCATAGGGATACAAAGCATAATCTGGAAC |
| PO4-rev | CCCCAACAGAGGGCAATCAAG |
| Expression of αβ-like subunits | |
| k1ORF2-NdeI-fw | GCATCATATGAATATATTTTACATATTTTTGTTTTTGCTGTCATTC |
| k1ORF2-PstI-rv | ATACTGCAGAAAAAGAAGGAGGTATGTGTCAAC |
| Site-directed mutagenesis | |
| k1ORF2–D462A-for | AATCTTGATGGTATAGCTTTAGATTGGGAATATC |
| k1ORF2–D462A-rev | CCAATCTAAAGCTATACCATCAAGATTATATTTA |
| k1ORF2–E464A-for | GATTTAGCTTGGGAATATCCAGGTGCTCCTGATATTC |
| k1ORF2–E464A-rev | CTGGATATTCCCAAGCTAAATCTATACCATCAAG |
| k1ORF2–E466A-for | GATTGGGCATATCCAGGTGCTCCTGATATTC |
| k1ORF2–E466A-rev | CTGGATATGCCCAATCTAAATCTATACCATCAAG |
| k1ORF2–D464A-E466A-for | GATTTAGCTTGGGCATATCCAGGTGCTCCTGATATTC |
| k1ORF2–D464A-E466A-rev | CTGGATATGCCCAAGCTAAATCTATACCATCAAG |
| k1ORF2–A462D-A464D-A466E-for | GATTTAGATTGGGAATATCCAGGTGCTCCTGATATTC |
| k1ORF2–A462D-A464D-A466E-rev | CTGGATATTCCCAATCTAAATCTATACCATCAAG |
| k1ORF2–C250A-for | GTTAAGATGGCTGGCTCTTAAAAGTAATGG |
| k1ORF2–C250A-rev | CTTTTAAGAGCCAGCCATCTTAACTTTCCC |
| k1ORF2–A250C-for | GTTAAGATGTGTGGCTCTTAAAAGTAATGG |
| k1ORF2–A250C-rev | CTTTTAAGAGCCACACATCTTAACTTTCCC |
| k1ORF4–C231A-for | GAATGGAAAAGCTGTATAAGAATTCACTGG |
| k1ORF4–C231A-rev | CTTATACAGCTTTTCCATTCTGTAGATTATTC |
| k1ORF4–A231C-for | GAATGGAAAATGTGTATAAGAATTCACTGG |
| k1ORF4–A231C-rev | CTTATACACATTTTCCATTCTGTAGATTATTC |
| RT-PCR | |
| k1ORF2s1-fw | AAGGTTTGGAGCATACTCATC |
| k1ORF2s1-rv | ACATCCTTTCCATCCATAATTAC |
Underlining alone indicates restriction sites; boldface together with underlining indicates base changes.
Generation of a γ-toxin-deficient k1ΔORF4 mutant.
For disruption of k1ORF4 (γ subunit) in S. cerevisiae 301 (F102-2 ura3) (29), the vector pARS, which harbors a recombination cassette consisting of a LEU2* selectable marker gene (30) governed by a cytoplasmic promoter (UCS [upstream conserved sequence]) and flanked by sequences of k1ORF3′ and/or k1ORF4, respectively, was constructed. The recombination flank (k1ORF3′/k1ORF4) was amplified via PCR, applying primers RKF-pARS-fw-SacI and RKF-pARS-rv-KpnI/NheI. The flank was subcloned by making use of KpnI and SacI sites and ligated into the likewise-cut vector pBluescript SK(−), resulting in vector pSK-RKF. By site-directed mutagenesis (31), a SacII restriction site was generated in k1ORF4 with the primers pSK-RKF-SacII-fw and pSK-RKF-SacII-rv.
LEU2* was amplified from vector pAR3 (32) using the primers XbaI-stop-UCS-LEU2*-fw and LEU2*-rev-MCS-XbaI, along with the UCS. The PCR product was subcloned and ligated with EcoRV-linearized pBluescript SK(−), resulting in pSK-LEU2*. LEU2* was subsequently cloned, by making use of SacII and XbaI sites, into pSK-RKF, resulting in vector pARS (Table 2). Prior to transformation into S. cerevisiae 301 (2, 29), harboring the K. lactis killer plasmids k1 and k2, the in vivo recombination cassette was cut out of this last vector with SacI and NheI. Transformants were subcultivated on YNB uracil agar lacking l-leucine. Linear elements of S. cerevisiae 301 k1ΔORF4 were verified by gel electrophoresis and Southern analysis.
TABLE 2.
Plasmids used in this study
| Plasmid | Genotype | Reference |
|---|---|---|
| pYEX-BX | E. coli ori, 2μ, Ampr, leu2-d URA3 | 15 |
| pKL-BX | pYEX-BX with PGAL1::ORF4 | 15 |
| pKL-BX-C231A | pKL-BX with PGAL1::ORF4C231A | This work |
| pKL-BX-A231A | pKL-BX with PGAL1::ORF4A231C | This work |
| pGTIRUTIR | E. coli ori, TIR, URA3 from S. cerevisiae, TIR, Ampr | 33 |
| pGT-PO4PO2 | pGTIRUTIR with P. acaciae pPac1-2 ORF2 and ORF4 | This work |
| pGT-PO4PO2* | pGTIRUTIR with P. acaciae pPac1-2 ORF4::3HA and ORF2::3myc | This work |
| pBluescript SK(−) | ColE1ori, Ampr, lacZ | Stratagene, Heidelberg, Germany |
| pSK-RKF | pBluescript SK(−) with recombination flank k1ORF3/k1ORF4 | This work |
| pSK-LEU2* | pBluescript SK(−) with UCS::LEU2 from S. cerevisiae | This work |
| pSK-PADH1 | pBluescript SK(−) with S. cerevisiaei ADH1 promoter | This work |
| pSK-PO4PO2 | pBluescript SK(−) with P. acaciae pPac1-2 ORF2 and ORF4 | This work |
| pSK-PO4::HA-PO2::myc | pBluescript SK(−) with P. acaciae pPac1-2 ORF4::HA and ORF2::myc | This work |
| pSK-PO4PO2* | pBluescript SK(−) with P. acaciae pPac1-2 ORF4::3HA and ORF2::3myc | This work |
| pSK-PADH1::ORF2 | pBluescript SK(−) with PADH1::ORF2 | This work |
| pSK-PADH1::ORF4 | pBluescript SK(−) with PADH1::ORF4 | This work |
| YEplac195 | 2μ, URA3, Ampr, E. coli ori | 40 |
| p195-PADH1::ORF4 | YEplac195 with PADH1::ORF4 | This work |
| p195-PADH1::ORF4C231A | YEplac195 with PADH1::ORF4C231A | This work |
| p195-PADH1::ORF4A231C | YEplac195 with PADH1::ORF4A231C | This work |
| p195-PADH1::ORF2 | YEplac195 with PADH1::ORF2 | This work |
| p195-PADH1::ORF2D462A | YEplac195 with PADH1::ORF2D462A | This work |
| p195-PADH1::ORF2A462D | YEplac195 with PADH1::ORF2A462D | This work |
| p195-PADH1::ORF2E464A | YEplac195 with PADH1::ORF2E464A | This work |
| p195-PADH1::ORF2A464E | YEplac195 with PADH1::ORF2A464E | This work |
| p195-PADH1::ORF2E466A | YEplac195 with PADH1::ORF2E466A | This work |
| p195-PADH1::ORF2A466E | YEplac195 with PADH1::ORF2A466E | This work |
| p195-PADH1::ORF2D464–E466A | YEplac195 with PADH1::ORF2D464–E466A | This work |
| p195-PADH1::ORF2A464D-A466E | YEplac195 with PADH1::ORF2A464D-A466E | This work |
| p195-PADH1::ORF2C250A | YEplac195 with PADH1::ORF2C250A | This work |
| p195-PADH1::ORF2A250C | YEplac195 with PADH1::ORF2A250C | This work |
| pARS | ColE1 ori, Ampr, k1ORF4-LEU2*-k1ORF4 | This work |
| pARS-PO4PO2 | pARS with P. acaciae pPac1-2 ORF2 and ORF4 | This work |
| pAR3 | k1ORF2′-LEU2*-k1ORF2″, Ampr, E. coli ori | 32 |
Construction of artificial linear elements.
For complementation of the k1ΔORF4 defect with a heterologous ACNase, the P. acaciae pPac1-2 ORF2 (PO2) and ORF4 (PO4), encoding the toxic γ subunit and immunity protein, respectively, were coexpressed in S. cerevisiae 301 k1ΔORF4. Both genes were amplified by PCR using primers PO4-fw and PO2-rv, along with their UCS, as well as the signal peptide-encoding region of ORF2. The PCR product was ligated into the SmaI-linearized vector pBluescript SK(−), yielding pSK-PO4PO2. PO4PO2 was subsequently subcloned via SpeI and XhoI digests into pARS, resulting in pARS-PO4PO2, and finally cloned via PspXI and BamHI into pGTIRUTIR (33), yielding pGT-PO4PO2.
In parallel, to analyze protein secretion, 3-hemagglutinin (3-HA) and 3-Myc epitopes were added to the C termini encoded by the respective P. acaciae ORF4 (PaORF4) and PaORF2 genes. First, single epitopes were added to these genes by PCR, amplifying from vector pSK-PO4PO2 using the primers PO4-HAsc-fw and PO2-MYCsc-rev. The PCR product was cloned and ligated (blunt end) into the SmaI-linearized vector pBluescript SK(−), resulting in vector pSK-PO4::HA-PO2::myc. In a second step, two further epitopes were added by PCR using the primers PO4-3HA-extender-for and PO2-3myc-extender-rev. The PCR product was again subcloned into a SmaI-linearized vector pBluescript SK(−), yielding pSK-PO4PO2*. The tagged genes were subsequently cloned by PspXI and BamHI digestion into vector pGTIRUTIR (33), yielding pGT-PO4PO2*.
Prior to transformation into S. cerevisiae 301 k1ΔKORF4, the artificial linear elements TU-PO4PO2-T and TU-PO4PO2*-T were amplified from vectors pGT-PO4PO2 and pGT-PO4PO2*, respectively, using the primer TIR2. Transformants were selected on YNB medium lacking l-leucine and uracil. Linear elements were verified by gel electrophoresis and Southern analysis.
Construction of nuclear expression vectors.
For nuclear-based expression of k1ORF4, the ADH1 promoter from S. cerevisiae BY4741 was PCR amplified using the primers PADH1-fw and PADH1-NdeI-rv and subcloned (blunt end) into SmaI-linearized pBluescript SK(−), yielding pSK-PADH1. k1ORF4 was amplified by PCR using the primers k1ORF4-NdeI-for and k1ORF4-EcoRI-rev and subsequently subcloned and ligated via NdeI and EcoRI restriction into the likewise-cut vector pSK-PADH1. The cassette PADH1::ORF4 was cloned and ligated via XmaI and EcoRI restrictions sites into the yeast expression vector YEplac195, yielding p195-PADH1::ORF4. The vector was transformed into S. cerevisiae 301 k1ΔORF4 and selected on YNB medium devoid of l-leucine and uracil. Transformants were verified by PCR analysis.
Site-directed mutagenesis.
Amino acid exchanges in the genes encoding the αβ (k1ORF2) or γ subunit (k1ORF4) of zymocin were achieved by site-directed mutagenesis using the Phusion site-directed mutagenesis kit from Thermo Fisher Scientific GmbH (Dreieich, Germany). Plasmid p195-PADH1::ORF4, p195-PADH1::ORF2, or pKL-BX was used as a template, and the corresponding primers are listed in Table 3. Sequencing of the mutated vectors (Table 2) was done with fluorescence-labeled dideoxynucleotide triphosphates (ddNTPs) using the BigDye Terminator v3.1 cycle sequencing kit and an ABI Prism 3730 capillary sequencer (Applied Biosystems, Foster City, CA).
Western analysis.
For analysis of protein secretion, cells grown in liquid YPD at 30°C were harvested, and toxin-containing supernatants were concentrated 200-fold using ultrafiltration units with 30-kDa-cutoff membranes (Sartorius Stedim Biotech GmbH, Göttingen, Germany). Concentrated supernatants were separated by discontinuous SDS-PAGE using 4% stacking, and 10% polyacrylamide gels and proteins were blotted onto polyvinylidene difluoride (PVDF) membranes. Immunological detection of the α subunit of zymocin was achieved by applying rabbit polyclonal anti-α-specific antibodies (13), followed by an anti-rabbit IgG-alkaline phosphatase (AP) secondary antibody (Sigma, Munich, Germany). Detection of the γ subunit of PaT (ORF2::3-Myc) was carried out using mouse monoclonal anti-c-Myc (clone 9E10; Roche Diagnostics Deutschland GmbH, Mannheim, Germany) and goat anti-mouse IgG-AP secondary antibody (Sigma, Munich, Germany).
RT-PCR.
Transcription of k1ORF2 in the parental strain S. cerevisiae 301, the k1ΔORF4 mutant, and the PADH1::ORF4 complemented strain was verified by reverse transcription-PCR (RT-PCR). Total RNA was isolated as described previously (34), and DNase I digestion was achieved employing the RNase-Free DNase I set (New England BioLabs GmbH, Frankfurt am Main, Germany). Reverse transcription was accomplished using the RevertAid H Minus First Strand cDNA synthesis kit (Fermentas, St. Leon-Rot, Germany), following the manufacturer's recommendations. For this purpose, 1 μg RNA and random hexamer primers were used. As a control, identical reactions were carried out without adding reverse transcriptase or with DNA. The cDNA of the 5′ end of k1ORF2 was detected by PCR using the primers k1ORF2s1-fw and k1ORF2s1-rv (Table 3).
RESULTS AND DISCUSSION
Disruption of k1ORF4 prevents zymocin production and αβ secretion.
To generate a platform allowing for systematic mutagenesis of αβγ holozymocin, we generated an S. cerevisiae 301 strain carrying the cytoplasmic k1/k2 pair with a disruption in k1ORF4 (encoding the ACNase subunit γ toxin). Disruption was facilitated by the use of an in vivo recombination vector (pARS) carrying a cytoplasmic expressible LEU2 gene (LEU2*) nested between long recombination flanks (500 to 550 bp) targeting the selection marker to k1ORF4 (Fig. 1A and B). The recombination cassette was removed from pARS and transformed into the k1/k2-carrying S. cerevisiae 301 strain (29). Correct integration and elimination of native k1 were verified by Southern analysis (Fig. 1C).
FIG 1.
Disruption of k1ORF4 prevents zymocin activity. (A) Recombination vector pARS for k1ORF4 disruption: “ampR,” ampicillin resistance (Ampr) gene; ColE1 ori, E. coli origin of replication; ORF3, immunity gene of k1; ORF4, γ toxin gene of k1; LEU2*, cytoplasmic expressible LEU2 gene of S. cerevisiae. (B) Scheme for targeted gene disruption of k1ORF4 by in vivo recombination. The genetic organization of k1 is represented prior to and after integration of the respective recombination cassette harboring the cytoplasmic LEU2* gene flanked by ORF3 and/or ORF4, ultimately yielding k1ΔORF4. (C) Agarose gel electrophoresis and Southern analysis performed with DNA from the parental strain (wt), the k1ΔORF4 mutant, and the VLE-cured strain (cured) using a k1ORF4 probe. The linear elements k1 (8.9 kb), k2 (13.8 kb), and k1ΔORF4 (10.3 kb), hmw (high-molecular-weight DNA), and dsRNA-L (4.6 kb) are indicated. M, GeneRuler 1-kb DNA ladder (Thermo Fisher Scientific, Dreieich, Germany). (D) Zymocin killer assay using partially purified and concentrated culture supernatants of the above strains against the susceptible S. cerevisiae CEN.PK2-1c strain.
To investigate the consequence of k1ORF4 disruption for zymocin production, we partially purified and concentrated culture supernatants from the k1/k2 S. cerevisiae parental strain, from the k1ΔORF4 mutant, and from a VLE-cured strain by ultrafiltration and analyzed these preparations for zymocin killer activity using the microdilution method. The supernatant of the parental strain proved growth inhibitory to the S. cerevisiae tester strain at dilutions equivalent to ∼1% of the original culture fluid (relative concentration factor, 0.01). Consistent with the notion that zymocin toxicity relies on the cytotoxic ACNase Orf4, we found that deletion of k1ORF4 results in a complete loss of toxin activity in the strains' supernatant (Fig. 1D).
To analyze whether removal of k1ORF4 from the zymocin-encoding plasmid system affects secretion or processing of k1Orf2 (encoding α and β subunits of zymocin), we checked supernatants from the k1/k2 parental strain and the k1ΔORF4 strain for the presence of protein bands detectable by a polyclonal antibody raised against the α subunit (13). The 129-kDa k1Orf2 protein is processed in the original host, K. lactis, by the KEX protease during secretion, which generates the 99-kDa α and 30-kDa β subunit from the precursor (7, 9–11). The size of the protein species detectable with anti-α sera therefore allows conclusions about whether or not Orf2 is processed correctly. We detected an ∼100-kDa signal in the supernatant of the k1/k2-carrying S. cerevisiae strain, showing that Orf2 secretion and processing occur similarly to those for K. lactis, which is consistent with earlier reports (9). However, upon removal of k1ORF4, not only is the ACNase-dependent killer activity lost, but also the k1Orf2 gene product, either unprocessed (129 kDa) or processed (99 kDa), is undetectable (Fig. 2A). This suggests that k1Orf2 secretion itself or the stability of the protein in the supernatant is severely compromised in the absence of k1Orf4. To check whether the obtained results are due to a possible coregulation of k1Orf2 and k1Orf4 expression, transcription of k1ORF2 was analyzed by RT-PCR. As depicted in Fig. 2B, transcription of the 5′ end of k1ORF2 takes place in the k1/k2 parental S. cerevisiae strain and in the k1ΔORF4 strain. Such results suggest a cosecretion dependency of k1Orf2 on k1Orf4, rather than a coregulation to ensure the equimolecular production of αβ and γ subunits.
FIG 2.
Compromised secretion of k1Orf2 in the absence of k1Orf4. (A) Concentrated supernatants of the parental strain S. cerevisiae 301 (wt), the k1ΔORF4 mutant, and the PADH1::ORF4 complemented strain (k1ΔORF4 ORF4) were tested by Western analysis using polyclonal antibodies raised against the α subunit and anti-rabbit secondary antibodies. Loading control, Coomassie-stained supernatants used for detection of anti-α. (B) RT-PCR analysis of k1ORF2 transcription. Reverse transcription of the 5′ end of k1ORF2 mRNA and detection of cDNA by PCR were carried out. DNA, genomic DNA used as the template; −RT or +RT, reaction with DNA-free RNA and without or with reverse transcriptase. (C) Microtiter assay with partially purified and concentrated supernatants of the above strains tested against S. cerevisiae CEN.PK2-1c.
In trans complementation of k1ΔORF4 requires Orf4C231.
We utilized the established k1ΔORF4 strain to check if (i) functional zymocin production can be restored by providing wild-type k1ORF4 in trans and (ii) whether this can be exploited to identify functional regions of γ toxin that are essential for αβγ holotoxin assembly but dispensable for the primary ACNase function. First, k1ORF4 was uncoupled from the cytoplasmic promoter, fused to the constitutive ADH1 promoter, and subsequently expressed from a standard nuclear vector (YEplac195). Since transcription of k1ORF2 from the cytoplasm was verified by RT-PCR (Fig. 2B), we then analyzed whether the k1Orf2 secretion defect could be restored by nuclear expression of k1ORF4. Indeed, a band of ∼100 kDa is detected by the anti-α antibody in the k1ΔORF4 strain carrying a nuclear PADH1::ORF4 construct (Fig. 2A). Thus, the secretion or stability defect of k1Orf2 associated with the loss of k1Orf4 could be restored by introduction of a nuclear expression construct for k1Orf4. In parallel, microtiter assays indicated an efficient restoration of zymocin presence in the supernatant (Fig. 2C). Hence, Orf4 can be provided in trans by nuclear expression of the corresponding gene, which efficiently restores functional zymocin production; therefore, such system can be utilized to identify residues in γ toxin that are essential for holotoxin function.
Multiple sequence alignments of VLE-encoded yeast killer toxins PaT, PiT, and DrT, which use a cargo import complex similar to that for zymocin (14–16, 29), revealed the presence of a single conserved Cys residue close to the C terminus of the cargo subunit (Fig. 3A). Since it was shown previously that βγ are disulfide linked and treatment of zymocin preparations with disulfide reducing agents abolishes activity, we chose to analyze the importance of this residue for both ACNase and holotoxin function. We created a k1ORF4C231A allele including the signal peptide coding region and expressed it under the control of the ADH1 promoter in the nucleus of the k1ΔORF4 strain. As controls, we included the empty vector, unmodified PADH1::ORF4, and also a back mutation of PADH1::ORF4A231C. Culture supernatants were partially purified and analyzed for the presence of killer activity. While wild-type Orf4 provided in trans again restored zymocin production, Orf4C231A completely lost this ability (Fig. 3B). Since zymocin production could also be restored by back mutation of A231 to C, loss of activity in Orf4C231A is due only to the exchange of C231 (Fig. 3B). To check whether the same exchange affects the killing efficiency of the intracellular form of γ toxin, we expressed k1ORF4 and its variant, constructed by replacing its promoter with the conditional GAL1 promoter of S. cerevisiae and by excluding the signal peptide coding region to achieve intracellular accumulation in order to mimic the γ toxin imported from outside the cell. Induced intracellular expression of k1ORF4, k1ORF4C231A, or k1ORF4A231C indistinguishably induced full growth arrest (Fig. 3C), in agreement with a previous study (35), thereby proving that the mutated variant of Orf4 is indeed translated. Thus, C231 is essential for functional zymocin secretion, but it does not affect the growth-inhibitory competency of the intracellular form of γ toxin (Fig. 3).
FIG 3.
Cys231 mutational effects on holotoxin and primary ACNase function. (A) Multiple sequence alignments of VLE-encoded γ toxin subunits. Black shading with reverse lettering, 100% conserved; gray shading with reverse lettering, 80% or more conserved; gray shading with black lettering, 60% or more conserved; no shading, less than 60% conserved. GenBank accession numbers: Kluyveromyces lactis pGKL1 ORF4, YP_001648060; Pichia inositovora pPin1-3 ORF4, CAD91887.1; Pichia acaciae pPac1-2 ORF2, CAE84960.1; Debaryomyces robertsiae pWR1A ORF3, CAE84956.1. The conserved Cys231 is boxed. (B) Effects of Cys231 mutations on holotoxin function. Microtiter assays were performed with partially purified and concentrated supernatants of k1ΔORF4, the PADH1::ORF4 complemented strain (PADH1::ORF4), and the PADH1::ORF4C231A (C231A) or PADH1::ORF4A231C (A231C) mutant. Killer assays were executed against S. cerevisiae CEN.PK2-1c. (C) Effects of Cys231 mutations on intracellular γ toxin activity. Ten-fold serial dilutions of S. cerevisiae CEN.PK2-1c cells expressing intracellularly the wild-type γ toxin (PGAL1::ORF4) or the C231A and A231C toxin mutants under a galactose-inducible promoter were tested on YNB supplemented with glucose or galactose for induction. As a control, a strain harboring an empty vector (control) was used.
Role of the sole cysteine residue in zymocin's β subunit.
To test whether, similar to k1ORF4, k1ORF2 (encoding α and β subunits of zymocin) can also be deleted in the cytoplasmic k1 and subsequently provided in trans from the nucleus, we utilized a strain carrying k1ΔORF2 that is unable to form active zymocin (29). For complementation, we removed the cytoplasmic promoter from the k1ORF2 gene, replaced it with the ADH1 promoter, and introduced the PADH1::ORF2 fusion in the nuclear vector YEplac195. The k1ΔORF2 strain carrying nuclear PADH1::ORF2 regained the ability to secrete zymocin (Fig. 4), demonstrating the usefulness of such a system to analyze the requirement of individual sites of k1ORF2 encoding α and β subunits for holotoxin function. It should be noted, however, that the complementation efficiency of PADH1::ORF2 is reduced compared to that of PADH1::ORF4 (Fig. 4), which is due to transcript instability in the former case and is accompanied by the lack of detection of alpha when expressed from the nucleus in k1ΔORF2 complemented with PADH1::ORF2 (data not shown).
FIG 4.
Functional complementation of k1ΔORF2 and k1ΔORF4 mutations. Microtiter assays were carried out on YPD with partially purified and concentrated supernatants of the k1/k2 parental strain S. cerevisiae 301 (wt), the k1ΔORF4 mutant, the PADH1::ORF4 complemented strain (k1ΔORF4 ORF4), the k1ΔORF2 mutant, and the PADH1::ORF2 complemented strain (k1ΔORF2 ORF2). Eclipse assays were performed on YPD with the above strains against S. cerevisiae CEN.PK2-1c.
Our above-described results revealed an essential function of the conserved γ C231 in the holotoxin context but not for the intracellular active form of the tRNase, suggesting that C231 could be involved in the formation of the disulfide bridge reported to exist between β and γ (7, 9). Interestingly, β contains only one cysteine residue, located at the very C terminus of the protein (Orf2C250), which is also conserved among other VLE-encoded killer toxins (Fig. 5A). Assuming a general requirement for covalent attachment of the cargo subunit (γ) to one of the carrier subunits (β), we predicted an absolute requirement for Orf2C250 to form active zymocin. To test this, we replaced Orf2C250 via an Orf2C250A change, expressed the PADH1::ORF2C250A gene in parallel to PADH1::ORF2 from the nucleus of the k1ΔORF2 strain, and analyzed supernatants for the presence of functional zymocin. As shown in Fig. 5B, the PADH1::ORF2 wild type and the A250C back mutant but not the C250A mutant could restore zymocin production in k1ΔORF2, indeed revealing an essential role of C250, the sole cysteine in the β subunit. These data underscore the assumption that the covalent attachment of the cargo subunit γ to the carrier subunit β is essential for functionality of the complex. This assumption is further supported by previous evidence that treatment of purified zymocin with disulfide reducing agents completely abolishes its activity (9, 13). In the C250A mutant, no such disulfide bridge can be formed since β becomes entirely devoid of cysteine, and consistently, loss of zymocin function is observed (Fig. 5B). Since γ C231 is similar to β C250 in locating at the C terminus of the protein, conserved among other VLE-encoded toxins and required for holotoxin function (Fig. 4), it appears likely that the cargo subunit is linked to β C250 via γ C231 (Fig. 5C) and such a covalent junction is key for secretion and/or toxicity.
FIG 5.
Cys250 mutational effects on holotoxin function. (A) Multiple sequence alignments of VLE-encoded β-like subunits. Black shading with reverse lettering, 100% conserved; gray shading with reverse lettering, 80% or more conserved; gray shading with black lettering, 60% or more conserved;, no shading, less than 60% conserved. GenBank accession numbers are as follows: K. lactis pGKL1 ORF2, YP_001648058; P. inositovora pPin1-3 ORF3, CAD91890.1; P. acaciae pPac1-2 ORF1, CAE84958.1; D. robertsiae pWR1A ORF2, CAE84954.1. The conserved Cys250 is boxed. (B) Effects of Cys250 mutations on holotoxin function. Microtiter assays were performed with partially purified and concentrated supernatants of k1ΔORF2, the PADH1::ORF2 complemented strain (PADH1::ORF2), and the PADH1::ORF4C250A (C250A) or PADH1::ORF4A250C (A250C) mutant. Killer assays were executed against S. cerevisiae CEN.PK2-1c. (C) Scheme of the heterotrimeric killer toxin zymocin. Structural integrity is maintained by intramolecular disulfide bonds within the α subunit and intermolecular disulfide bonds between β C250 and γ C231. The active center of the chitinase is indicated. LysM, chitin binding, and glycosyl hydrolase family 18 motifs are differently gray shaded. Transmembrane domains are depicted as hatched boxes.
The chitinase active site is essential for zymocin function.
The α subunit of zymocin carries a chitinase domain equipped with a chitinase family 18 active site (Fig. 6A). It was shown previously that the α subunit exhibits chitinase activity, which can be inhibited by allosamidin (12). Since increased allosamidin doses in bioassays with zymocin led to decreased toxin activity, the involvement of chitin degradation in the process of cargo (γ) import was assumed (12). However, even at the highest concentration of allosamidin, which led to the complete inhibition of the chitinase activity in vitro, some growth-inhibitory activity in vivo was observed, leaving some doubt on the essentiality of the chitinase activity for toxin function. Since our in trans complementation system for k1ΔORF2 provides a very sensitive readout for the functionality of mutated variants of α and β and since the chitinase family 18 active site is well characterized, we checked the relevance of chitinase catalytic residues within the α subunit for functionality of the complex. Thus, the three highly conserved residues of the chitinase active site motif (DXDXE) of the α subunit were converted to alanine (D462A, D464A, E466A, and D464A-E466A), and the genes were expressed by fusion to the ADH1 promoter in the k1ΔORF2 mutant. Supernatants from the strain complemented with the wild-type k1ORF2 gene and all active site substitutions were concentrated and analyzed for zymocin activity. Only the wild-type control displayed detectable zymocin activity in the supernatant, whereas all substitutions analyzed were unable to produce functional zymocin, as also verified by eclipse assays (Fig. 6B). Thus, rather than contributing to the toxin's efficiency, the chitinase active site of the α subunit is essential for toxin function. Since zymocin activity could be restored by back mutations, loss of activity only is due to the exchange of the corresponding residues (data not shown). However, it cannot be excluded that the chitinase active site mutations may affect the stability of the protein. The results support the assumption that chitin degradation is a prerequisite for import of the toxic γ subunit. Interestingly, chitin is localized as a thin layer on top of the plasma membrane, suggesting that its degradation could be intimately linked to membrane perforation and passage of γ into the cytoplasm. The complementation system established in this work will provide a valuable tool to further study early steps in zymocin action, for example, by generating immunologically detectable or fluorescently tagged versions of individual zymocin subunit.
FIG 6.
Chitinase active site residues are essential for zymocin function. (A) Multiple sequence alignments of VLE-encoded chitinase like subunits. Black shading with reverse lettering, 100% conserved; gray shading with reverse lettering, 80% or more conserved; gray shading with black lettering, 60% or more conserved; no shading, less than 60% conserved. GenBank accession numbers are as follows: K. lactis pGKL1 ORF2, YP_001648058; P. inositovora pPin1-3 ORF3, CAD91890.1; P. acaciae pPac1-2 ORF1, CAE84958.1; D. robertsiae pWR1A ORF2, CAE84954.1. Family 18 chitinase active sites (DXXDXDXE; [DN]-G-[LIVMF]-[DN]-[LIVMF]-[DN]-X-E) are highlighted by boxes and shown above in bold letters. (B) Microtiter assays were performed in YPD with partially purified and concentrated supernatants of the k1ΔORF2 mutant, the PADH1::ORF2 complemented strain, and the D462A, D464A, E466A, and D464A-E466A toxin mutants. Eclipse assays were done on YPD agar. Killer assays were executed against S. cerevisiae CEN.PK2-1c. Mutations of D462, D464, and E466 in the chitinase active site (DXDXE motif, α subunit) completely abolished killing activities.
The k1Orf2 secretion defect in k1ΔORF4 cannot be restored by a heterologous ACNase.
Since the αβ-like subunits of the known killer VLEs are highly conserved with respect to their chitin binding and chitinase domains (Fig. 6A), as well as the conserved cysteine residues at the C terminus of the interacting β and γ subunits (Fig. 3A and 5A), we wondered if the αβ subunits carry and deliver not only their cognate cargo protein but a heterologous subunit of another known killer system. We utilized the k1ΔORF4 strain to introduce the toxic ACNase subunit (PaOrf2) from the related Pichia acaciae killer system along with its immunity gene (PaOrf4) (15, 36). The last two were expressed cytoplasmically using the in vitro-constructed element TU-PO4PO2-T (see Materials and Methods), and the presence of the linear elements was verified by gel electrophoresis (Fig. 7A). In addition, PaOrf2 was expressed as a C-terminal 3-Myc-tagged variant to follow secretion of the protein in the k1ΔORF4 and k1ΔORF2 strains. Western analysis of concentrated supernatants revealed that PaOrf2-Myc was efficiently secreted in the k1ΔORF4 strain (Fig. 7D); however, no toxin activity could be detected (Fig. 7B), while PaOrf4 provided immunity toward exogenously applied PaT (Fig. 7C). Since we have shown that k1Orf2 normally requires k1Orf4 for efficient secretion, lack of toxicity might be attributed to an inability to form an active hybrid toxin complex or to the absence of k1Orf2 secretion. To test these two alternatives, we analyzed whether PaOrf2-Myc is capable of restoring k1Orf2 secretion in the k1ΔORF4 strain. As a control, we introduced the PADH1::ORF4 construct, shown before to restore k1Orf2 secretion. As shown in Fig. 7D, k1Orf4 but not PaOrf2-Myc is capable of restoring k1Orf2 secretion. Thus, absence of functional hybrid toxin activity can likely be attributed to the complete absence of k1Orf2 secretion and is not necessarily due to a general inability to form such a hybrid toxin. However, it cannot be excluded that the lack of proper assembly renders secretion interdependent.
FIG 7.
The k1ΔORF4 defect cannot be restored by a heterologous ACNase from P. acaciae. (A) Schematic diagram of the in vitro-constructed TU-PO4PO2-T element encoding the γ toxin and immunity factor from P. acaciae. ScURA3*, cytoplasmic expressible URA3 gene of S. cerevisiae; TIR, terminal inverted repeats; ●, 5′-terminal protein. Agarose gel electrophoresis showing different genetic materials in the k1/k2 parental strain S. cerevisiae 301 (wt), the k1ΔORF4 mutant, and the k1ΔORF4 PO4PO2 strain expressing the αβ subunits from K. lactis and the γ toxin and immunity protein from P. acaciae. hmw, high-molecular-weight DNA. (B) Microtiter assays were performed on YPD with partially purified and concentrated supernatants of the above strains. As a sensitive test strain, S. cerevisiae CEN.PK2-1c was applied. (C) In parallel, the strains were tested against the P. acaciae toxin (PaT). (D) Concentrated supernatants of k1ΔORF4, the PADH1::ORF4 complemented mutant, the k1ΔORF4 PO4PO2* strain, the k1ΔORF2 mutant, and the k1ΔORF2 PO4PO2* strain were tested by Western analysis using antibodies raised against the α subunit of zymocin or the 3-Myc epitope of the tagged γ toxin of PaT (γ PaT-Myc). Anti-rabbit or anti-mouse secondary antibodies were used. Loading control, Coomassie-stained concentrated supernatants.
Future work will be required to define the dependency of k1Orf2 secretion on k1Orf4 cosecretion. It has also been shown that k1Orf4 secretion is severely impaired in the absence of k1Orf2 (10), which might suggest that an interaction between k1Orf2 and k1Orf4 is required at a specific step during the secretory pathway for efficient secretion of either protein. Final zymocin assembly, however, is apparently dispensable for subunit secretion, since kex1/kex2 mutations inhibit k1Orf2 processing in the Golgi apparatus and secretion of αβ but do not affect the secretion of the γ subunit (8). The ACNase subunit PaOrf2 apparently differs from k1Orf4 in that it is secreted efficiently in the k1ΔORF2 strain without the need for an αβ precursor (Fig. 7D). It remains to be studied, however, whether the αβ-like PaOrf1 depends on PaOrf2 for cosecretion. Interestingly, fusion of the α mating factor pre-pro sequence, which harbors a Kex1/2 processing site, to the N terminus of k1Orf4 enabled secretion of the γ toxin independently of k1Orf2 cosecretion (8). Thus, modification of the k1Orf2 N terminus may represent an analogous strategy to overcome the secretion blockade of k1Orf2 in the absence of k1Orf4. Restoring k1Orf4-independent secretion of the αβ subunits will likely be a requirement to further investigate the specificity of the carrier subunit for cognate or alternative cargo proteins.
Footnotes
Published ahead of print 15 August 2014
REFERENCES
- 1.Lu J, Huang B, Esberg A, Johansson MJ, Byström AS. 2005. The Kluyveromyces lactis γ-toxin targets tRNA anticodons. RNA 11:1648–1654. 10.1261/rna.2172105 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Gunge N, Sakaguchi K. 1981. Intergeneric transfer of deoxyribonucleic acid killer plasmids, pGKl1 and pGKl2, from Kluyveromyces lactis into Saccharomyces cerevisiae by cell fusion. J. Bacteriol. 147:155–160 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Tokunaga M, Wada N, Hishinuma F. 1987. A novel yeast secretion vector utilizing secretion signal of killer toxin encoded on the yeast linear DNA plasmid pGKL1. Biochem. Biophys. Res. Commun. 144:613–619. 10.1016/S0006-291X(87)80010-4 [DOI] [PubMed] [Google Scholar]
- 4.Meinhardt F, Klassen R. 2007. Microbial linear plasmids. In Meinhardt F, Klassen R. (ed), Microbiology monographs, vol 7 Springer-Verlag, Berlin, Germany [Google Scholar]
- 5.Satwika D, Klassen R, Meinhardt F. 2012. Anticodon nuclease encoding virus-like elements in yeast. Appl. Microbiol. Biotechnol. 96:345–356. 10.1007/s00253-012-4349-9 [DOI] [PubMed] [Google Scholar]
- 6.Jeske S, Meinhardt F, Klassen R. 2007. Extranuclear inheritance: virus-like DNA-elements in yeast, p 98–129 In Esser K, Lüttge U, Beyschlag W, Murata J. (ed), Progress in botany, vol 68 Springer, Berlin, Germany [Google Scholar]
- 7.Stark MJR, Boyd A. 1986. The killer toxin of Kluyveromyces lactis—characterization of the toxin subunits and identification of the genes which encode them. EMBO J. 5:1995–2002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Tokunaga M, Kawamura A, Hishinuma F. 1989. Expression of pGKL killer 28K subunit in Saccharomyces cerevisiae: identification of 28K subunit as a killer protein. Nucleic Acids Res. 17:3435–3446. 10.1093/nar/17.9.3435 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Stark MJ, Boyd A, Mileham AJ, Romanos MA. 1990. The plasmid-encoded killer system of Kluyveromyces lactis: a review. Yeast 6:1–29. 10.1002/yea.320060102 [DOI] [PubMed] [Google Scholar]
- 10.Tokunaga M, Kawamura A, Kitada K, Hishinuma F. 1990. Secretion of killer toxin encoded on the linear DNA plasmid pGKL1 from Saccharomyces cerevisiae. J. Biol. Chem. 265:17274–17280 [PubMed] [Google Scholar]
- 11.Wesolowski-Louvel M, Tanguy-Rougeau C, Fukuhara H. 1988. A nuclear gene required for the expression of the linear DNA-associated killer system in the yeast Kluyveromyces lactis. Yeast 4:71–81. 10.1002/yea.320040108 [DOI] [PubMed] [Google Scholar]
- 12.Butler AR, O'Donnell RW, Martin VJ, Gooday GW, Stark MJ. 1991. Kluyveromyces lactis toxin has an essential chitinase activity. Eur. J. Biochem. 199:483–488. 10.1111/j.1432-1033.1991.tb16147.x [DOI] [PubMed] [Google Scholar]
- 13.Jablonowski D, Fichtner L, Martin VJ, Klassen R, Meinhardt F, Stark MJ, Schaffrath R. 2001. Saccharomyces cerevisiae cell wall chitin, the Kluyveromyces lactis zymocin receptor. Yeast 18:1285–1299. 10.1002/yea.776 [DOI] [PubMed] [Google Scholar]
- 14.Kast A, Klassen R, Meinhardt F. 2014. rRNA fragmentation induced by a yeast killer toxin. Mol. Microbiol. 91:606–617. 10.1111/mmi.12481 [DOI] [PubMed] [Google Scholar]
- 15.Klassen R, Paluszynski JP, Wemhoff S, Pfeiffer A, Fricke J, Meinhardt F. 2008. The primary target of the killer toxin from Pichia acaciae is tRNA(Gln). Mol. Microbiol. 69:681–697. 10.1111/j.1365-2958.2008.06319.x [DOI] [PubMed] [Google Scholar]
- 16.Klassen R, Teichert S, Meinhardt F. 2004. Novel yeast killer toxins provoke S-phase arrest and DNA damage checkpoint activation. Mol. Microbiol. 53:263–273. 10.1111/j.1365-2958.2004.04119.x [DOI] [PubMed] [Google Scholar]
- 17.Stam JC, Kwakman J, Meijer M, Stuitje AR. 1986. Efficient isolation of the linear DNA killer plasmid of Kluyveromyces lactis: evidence for location and expression in the cytoplasm and characterization of their terminally bound proteins. Nucleic Acids Res. 14:6871–6884. 10.1093/nar/14.17.6871 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Schaffrath R, Stark MJ, Gunge N, Meinhardt F. 1992. Kluyveromyces lactis killer system: ORF1 of pGKL2 has no function in immunity expression and is dispensable for killer plasmid replication and maintenance. Curr. Genet. 21:357–363. 10.1007/BF00351695 [DOI] [PubMed] [Google Scholar]
- 19.Sambrook J, Fritsch EF, Maniatis T. 1989. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY [Google Scholar]
- 20.Gietz RD, Schiestl RH, Willems AR, Woods RA. 1995. Studies on the transformation of intact yeast cells by the LiAc/SS-DNA/PEG procedure. Yeast 11:355–360. 10.1002/yea.320110408 [DOI] [PubMed] [Google Scholar]
- 21.Gunge N, Takahashi S, Fukuda K, Ohnishi T, Meinhardt F. 1994. UV hypersensitivity of yeast linear plasmids. Curr. Genet. 26:369–373. 10.1007/BF00310503 [DOI] [PubMed] [Google Scholar]
- 22.Niwa O, Sakaguchi K, Gunge N. 1981. Curing of the killer deoxyribonucleic acid plasmids of Kluyveromyces lactis. J. Bacteriol. 148:988–990 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Klassen R, Jablonowski D, Schaffrath R, Meinhardt F. 2002. Genome organization of the linear Pichia etchellsii plasmid pPE1A: evidence for expression of an extracellular chitin-binding protein homologous to the α-subunit of the Kluyveromyces lactis killer toxin. Plasmid 47:224–233. 10.1016/S0147-619X(02)00014-8 [DOI] [PubMed] [Google Scholar]
- 24.Klassen R, Meinhardt F. 2002. Linear plasmids pWR1A and pWR1B of the yeast Wingea robertsiae are associated with a killer phenotype. Plasmid 48:142–148. 10.1016/S0147-619X(02)00101-4 [DOI] [PubMed] [Google Scholar]
- 25.Kishida M, Tokunaga M, Katayose Y, Yajima H, Kawamura-Watabe A, Hishinuma F. 1996. Isolation and genetic characterization of pGKL killer-insensitive mutants (iki) from Saccharomyces cerevisiae. Biosci. Biotechnol. Biochem. 60:798–801. 10.1271/bbb.60.798 [DOI] [PubMed] [Google Scholar]
- 26.Klassen R, Meinhardt F. 2007. Linear protein-primed replicating plasmids in eukaryotic microbes, p 187–226 In Meinhardt F, Klassen R. (ed), Microbiology monographs, vol 7 Springer, Berlin, Germany [Google Scholar]
- 27.Muccilli S, Wemhoff S, Restuccia C, Meinhardt F. 2013. Exoglucanase encoding genes from three Wickerhamomyces anomalus killer strains isolated from olive brine. Yeast 30:33–43. 10.1002/yea.2935 [DOI] [PubMed] [Google Scholar]
- 28.Klassen R, Wemhoff S, Krause J, Meinhardt F. 2011. DNA repair defects sensitize cells to anticodon nuclease yeast killer toxins. Mol. Genet. Genomics 285:185–195. 10.1007/s00438-010-0597-5 [DOI] [PubMed] [Google Scholar]
- 29.Paluszynski JP, Klassen R, Meinhardt F. 2007. Pichia acaciae killer system: genetic analysis of toxin immunity. Appl. Environ. Microbiol. 73:4373–4378. 10.1128/AEM.00271-07 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Kämper J, Meinhardt F, Gunge N, Esser K. 1989. New recombinant linear DNA-elements derived from Kluyveromyces lactis killer plasmids. Nucleic Acids Res. 17:1781. 10.1093/nar/17.4.1781 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Weiner MP, Costa GL, Schoettlin W, Cline J, Mathur E, Bauer JC. 1994. Site-directed mutagenesis of double-stranded DNA by the polymerase chain reaction. Gene 151:119–123. 10.1016/0378-1119(94)90641-6 [DOI] [PubMed] [Google Scholar]
- 32.Schickel J, Helmig C, Meinhardt F. 1996. Kluyveromyces lactis killer system: analysis of cytoplasmic promoters of the linear plasmids. Nucleic Acids Res. 24:1879–1886. 10.1093/nar/24.10.1879 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Bennett AM, Norris AR, Limnander de Nieuwenhove A, Russell PJ. 2002. Replication of a linear mini-chromosome with terminal inverted repeats from the Kluyveromyces lactis linear DNA plasmid k2 in the cytoplasm of Saccharomyces cerevisiae. Plasmid 48:13–23. 10.1016/S0147-619X(02)00019-7 [DOI] [PubMed] [Google Scholar]
- 34.Klassen R, Fricke J, Pfeiffer A, Meinhardt F. 2008. A modified DNA isolation protocol for obtaining pure RT-PCR grade RNA. Biotechnol. Lett. 30:1041–1044. 10.1007/s10529-008-9648-y [DOI] [PubMed] [Google Scholar]
- 35.Keppetipola N, Jain R, Meineke B, Diver M, Shuman S. 2009. Structure-activity relationships in Kluyveromyces lactis gamma-toxin, a eukaryal tRNA anticodon nuclease. RNA 15:1036–1044. 10.1261/rna.1637809 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Klassen R, Kast A, Wünsche G, Paluszynski JP, Wemhoff S, Meinhardt F. 2014. Immunity factors for two related tRNA(Gln) targeting killer toxins distinguish cognate and non-cognate toxic subunits. Curr. Genet. 60:213–222. 10.1007/s00294-014-0426-1 [DOI] [PubMed] [Google Scholar]
- 37.Kämper J, Esser K, Gunge N, Meinhardt F. 1991. Heterologous gene expression on the linear DNA killer plasmid from Kluyveromyces lactis. Curr. Genet. 19:109–118. 10.1007/BF00326291 [DOI] [PubMed] [Google Scholar]
- 38.Worsham PL, Bolen PL. 1990. Killer toxin production in Pichia acaciae is associated with linear DNA plasmids. Curr. Genet. 18:77–80. 10.1007/BF00321119 [DOI] [PubMed] [Google Scholar]
- 39.van Dijken JP, Bauer J, Brambilla L, Duboc P, Francois JM, Gancedo C, Giuseppin ML, Heijnen JJ, Hoare M, Lange HC, Madden EA, Niederberger P, Nielsen J, Parrou JL, Petit T, Porro D, Reuss M, van Riel N, Rizzi M, Steensma HY, Verrips CT, Vindelov J, Pronk JT. 2000. An interlaboratory comparison of physiological and genetic properties of four Saccharomyces cerevisiae strains. Enzyme Microb. Technol. 26:706–714. 10.1016/S0141-0229(00)00162-9 [DOI] [PubMed] [Google Scholar]
- 40.Gietz RD, Sugino A. 1988. New yeast-Escherichia coli shuttle vectors constructed with in vitro mutagenized yeast genes lacking six-base pair restriction sites. Gene 74:527–534. 10.1016/0378-1119(88)90185-0 [DOI] [PubMed] [Google Scholar]