| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 279, Issue 20, 21271-21281, May 14, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
¶¶¶
From the
Department of Molecular Pharmacology, St. Jude Children's Research Hospital, Memphis, Tennessee 38105, the ¶Department of Biochemistry and Molecular Pharmacology, Thomas Jefferson University, Philadelphia, Pennsylvania 19107, **Institute of Cell Biology, CNR, Monterotondo, Italy, and the Department of Biology, University of Padua, 35131 Padua, Italy
Received for publication, November 11, 2003 , and in revised form, February 24, 2004.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
cells was not detected in the doa4-10 mutant. Mutations in SLA1 or SLA2, which alter actin cytoskeleton architecture, induced a conditional synthetic lethal phenotype in combination with doa4-10 in the absence of DNA damage. Here actin cytoskeleton defects coincided with the enhanced fragility of large-budded cells. In contrast, the enhanced sensitivity of doa4-10 mutant cells to Top1T722Ap was unrelated to alterations in endocytosis and was selectively suppressed by increased dosage of the ribonucleotide reductase inhibitor Sml1p. Additional studies suggest a role for Doa4p in the Rad9p checkpoint response to Top1p poisons. These findings indicate a functional link between ubiquitin-mediated proteolysis and cellular resistance to CPT-induced DNA damage. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
DNA topoisomerase I is the cellular target of the antitumor drug, camptothecin (CPT),1 which reversibly stabilizes the covalent Top1p-DNA complex by inhibiting DNA religation (3-5). The ternary CPT-Top1p-DNA complexes per se are insufficient to induce a cytotoxic response. Rather, collisions with advancing replication forks convert the drug-stabilized complexes into the irreversible DNA lesions that trigger cell cycle arrest and cell death. This model is consistent with the S-phase specificity of CPT and the observation that inhibition of DNA synthesis by aphidicolin abrogates the cytotoxic action of the drug (6).
In the yeast Saccharomyces cerevisiae, DNA topoisomerase I is nonessential (6, 7). top1 strains are resistant to CPT, although drug sensitivity may be restored by expressing plasmid-borne yeast or human TOP1. Thus, the cytotoxic action of CPT does not derive from the inhibition of Top1p catalytic activity; rather the drug converts this nonessential enzyme into a cellular toxin.
Biochemical and structural data have begun to reveal the domainal organization of Top1p critical for DNA binding, enzyme catalysis, and CPT sensitivity (8-12). Single amino acid substitutions in Top1p have also been defined that alter enzyme sensitivity to CPT and other Top1p poisons or that alter the DNA cleavage-religation equilibrium of the enzyme in the absence of drug (13-25). In yeast Top1T722Ap, substitution of Ala for Thr-722 (five residues N-terminal to the active site tyrosine, Tyr-727) mimics the action of CPT by reducing the rate of DNA religation (26, 27). Similar effects on human enzyme activity were obtained with the same substitution at the corresponding position in hTop1T718Ap (24). These studies highlight the similarities in eukaryotic Top1p mechanism, structure, and sensitivity to CPT.
Considerably less is known of the cellular processes that regulate cellular responses to drug- or top1 mutation-induced DNA lesions. Rad9p, Mec1p, and Rad53p checkpoints modulate cell sensitivity to CPT, consistent with studies in mammalian cells demonstrating the activation of the ATR/CHK1 checkpoint in response to drug treatment (26, 28-34). Rad9p also protects cells from self-poisoning top1 mutants, including Top1T722Ap (26). The induction of double-stranded DNA breaks has, in part, been inferred by the enhanced sensitivity of rad52 strains to Top1p-DNA lesions (35). More direct evidence for alterations in DNA replication potentiating the cytotoxic action of CPT comes from studies demonstrating the enhanced drug sensitivity of yeast strains mutated for SGS1, MUS81, TRF4, CDC45, and DPB11 (29, 30, 36-40). In the case of the Sgs1 helicase or Mus81p, alterations in fork regression might preclude repair of Top1-DNA lesions (36-38). Trf4p (DNA polymerase 
) activity is required to establish sister chromatid cohesion and for top1 cell viability (40). Hypomorphic alleles of CDC45 and DBP11, isolated in our screen for top1T722A- hypersensitive (tah) mutants, induce a transient accumulation of Okazaki-sized DNA fragments and enhance cell sensitivity to Top1p-induced DNA damage (39). These results link processive DNA replication with cellular resistance to CPT.
Here we report the characterization of three additional tah mutants, doa4-10, sla1-10, and sla2-10, that exhibit enhanced sensitivity to the self-poisoning Top1T722Ap at 36 °C. We provide evidence that the Doa4 deubiquitinating enzyme and the cortical actin proteins, Sla1p and Sla2p, function to protect cells from Top1p-induced DNA damage. Doa4p is a ubiquitin-specific protease that physically associates with the 26 S proteasome and recycles ubiquitin from proteins targeted for proteolytic degradation (41, 42). Doa4p also functions in the endocytic pathway to remove ubiquitin from plasma membrane proteins targeted for vacuolar degradation (43). Sla1p and Sla2p function in the assembly of cortical actin patches and the actin cytoskeleton (44, 45) and in endocytosis (46).
We also report that the ribonucleotide reductase inhibitor, Sml1p (47), functions as a specific dosage suppressor of doa4-10 sensitivity to Top1T722Ap-induced lethality. Further studies of checkpoint mutants, such as rad9, mec1, and tel1, suggest doa4-10-mediated alterations in the Rad9p DNA damage checkpoint. These results are discussed in terms of a functional link between ubiquitin-mediated proteolysis, cortical actin organization, and cellular responses to Top1p poisons.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
|
cells, transformed with YCpSctop1T722A, were subjected to ethyl methonesulfonate mutagenesis and screened for colonies exhibiting temperature-sensitive (ts) growth at 36 °C. Plating on 5-FOA eliminated mutants where ts growth was not linked to plasmid-encoded Top1T722Ap. Backcrossing to TAH+, top1 strains identified 10 tah mutants, where a single recessive gene defect was linked to top1T722A hypersensitivity. Complementation analysis determined that tah6 and tah20 were allelic. Cloning DOA4, SLA1, and SLA2 by Complementation—tah22, tah20, and tah14 strains, transformed with YCp-FY250 library DNA, were screened for plasmid-dependent growth at 36 °C on SC-uracil media containing 10 mg/ml HU. The smaller of the overlapping TAH22 clones contained 5.9 kbp of chromosome IV including YDR068, DOA4, and YDR070. A TAH20 clone contained 5.7 kbp of chromosome II encompassing SLA1 and YBL006. This clone also complemented the HU hypersensitivity of the tah6 strain. TAH14 clone contained 8.4 kbp of chromosome XIV with only one gene, SLA2. Subcloning confirmed that DOA4, SLA1, and SLA2 complemented the HU and top1T722A hypersensitivity of tah22, tah6/tah20, and tah14, respectively. A YEpDOA4 vector was made by ligating a 5.4-kbp SpeI DNA fragment, excised from TAH22 clone, into YEp24-PL.
To confirm the genetic identity of TAH22 as DOA4, TAH6/20 as SLA1, and TAH14 as SLA2, a selectable URA3 (or LEU2) marker was integrated into the genomic sequences flanking DOA4, SLA1, and SLA2, respectively. After mating with respective tah22, tah6, and tah14 mutants, meiotic products were assessed for segregation of uracil (or leucine) prototrophy and the tah phenotype. Integration constructs were as follows: a 2.7-kbp NotI DNA fragment from TAH22 clone was ligated into pRS406 to yield YIpDOA4·U or pRS405 to make YIpDOA4·L. Transformation of EKY2 cells with MscI-digested DNA targeted integration 5' to DOA4. A LEU2-based YIpSLA1·L vector was made by ligating a 2.1-kbp PstI DNA fragment from TAH20 clone, into pRS405. AvrII digestion targeted URA3 integration 3' to SLA1. To target URA3 integration to sequences 3' to SLA2, a 2.4-kbp NotI/XbaI DNA fragment excised from the TAH14 clone was ligated into pRS406. EKY2 cells were transformed with HpaI-digested YIpSLA2.
Mutant doa4-10 and sla1-10 alleles were recovered after targeted integration of YIpDOA4·L and YIpSLA1·L, respectively. Purified genomic DNA was restricted with HindIII (doa4-10) or ClaI (sla1-10). Size-selected DNA fragments were ligated and transformed into E. coli. DNA sequencing defined the mutations in doa4-10 and sla1-10.
A C-terminal HA tag was introduced into wild type DOA4 using a URA3 marker flanked by 3x HA tags (50). A C-terminal HA tag and stop codon were introduced after residue 387 in DOA4 to generate doa4-10-HA. HA-URA3-HA was integrated into a single DOA4 locus in diploid EKY23 cells. Meiotic products were screened for uracil prototrophy and sensitivity to HU. Excision of URA3 was selected on 5-FOA plates, and the generation of doa4-10-HA was confirmed by DNA sequencing. DOA4-HA and doa4-10-HA cell sensitivity to top1T722A, CPT, and HU was indistinguishable from that of DOA4 and doa4-10 strains, respectively.
Cell Viability Assays—Exponentially growing cells, transformed with YCpScTOP1, YCpSctop1T722A, or pRS416, were 10-fold serially diluted, and 5-µl aliquots were spotted on SC-uracil plates. Cell viability was assessed after incubation at 26 °C or 36 °C. To assay CPT sensitivity, YCpScTOP1-transformed cells were spotted onto SC-uracil media plus 25 mM HEPES, pH 7.2, 0.125% Me2SO and 0 or 5 µg/ml CPT. To quantitate colony formation, serial dilutions of transformed cells were plated onto SC-uracil media (+ or - CPT) and incubated at 26 or 36 °C. Cell sensitivity to methyl methanesulfonate (0.0125 or 0.025%), HU (5 or 10 mg/ml), or UV (10 or 20 µJ/M2) was assayed as described (39). Sensitivity to high salt was assayed on YPD, 6% NaCl plates. Cell growth on nonfermentable carbon sources was assessed on yeast extract, peptone, 3% glycerol plates.
To quantitate HU sensitivity, cultures exponentially growing at 26 °C were divided, and half was treated with 15 mg/ml HU. After 30 min, the cultures were shifted to 36 °C(t = 0). At various times, aliquots were diluted and plated onto YPD media at 26 °C. Colony formation was assessed at 26 °C. Aliquots of cells were fixed with 70% ethanol and stored at -20 °C for subsequent DAPI staining and microscopy.
Actin Localization and DAPI Staining—Exponentially growing DOA4, doa4-10, and sla1-10 cells, in YPD at 26 and 36 °C, were fixed with 3.7% formaldehyde. After 2 h, the pelleted cells were washed with phosphate-buffered saline. For actin staining (51), cells were resuspended in phosphate-buffered saline, stained with 0.66 µM rhodamine-phalloidin (Molecular Probes), and then mixed with mounting media (Molecular Probes) on polylysine-coated Teflon slides. For DAPI staining of DNA, cells spotted on polylysine-coated Teflon slides were suspended in 0.2 µg/ml DAPI and mounting media. Cells were viewed with a Zeiss Axioskop 2 microscope equipped with DIC, epifluorescence, and UV blocking filter sets. Images were acquired with a Micromax CCD camera and IP lab software (Scanalytics).
Electron Microscopy—Cells were fixed, embedded, and stained as described (52-55). Briefly, cells grown at 26 °C were divided, and half was shifted to 36 °C for 10 h. Cells, pelleted by centrifugation, were resuspended in freshly prepared 40 mM sodium phosphate, pH 6.7, 1 M sorbitol, 0.5% glutaraldehyde, and 3% formaldehyde for 1 h, washed three times with 40 mM sodium phosphate, pH 6.7, and resuspended in 1% NaIO4 for 10 min. After quenching free aldehyde groups with 50 mM NH4Cl, cells were washed with water and successively dehydrated in EtOH (50, 70, 80, 85, 90, and 95% and two washes in 100% EtOH) at 4 °C, followed by 100% EtOH at room temperature. Infiltration and polymerization of LR White embedding media were as per manufacturer's instructions (Polysciences). Thin sections (50 nm) were cut with a Leica UltraCut ultramicrotome, mounted on copper grids coated with a 2% collodium film, stained with 2% uranyl acetate and Reinold lead citrate, and observed in a Jeol Jem 10-10 microscope at 80 kV. Magnifications were corrected using a tropomyosin crystal standard.
DNA Topoisomerase I Activity and Immunoprecipitation—Top1 protein levels in DOA4 and doa4-10 cells transformed with YCpScTOP1 or YCpSctop1T722A were assessed in crude extracts prepared from exponential cultures grown at 26 °C or shifted to 36 °C for 4-24 h. As described (27), cells harvested by centrifugation were resuspended in 2 ml/g cells of TEEG buffer (20 mM Tris, pH 7.5, 10 mM EDTA, 10 mM EGTA, 10% glycerol) plus 0.2 M KCl and protease inhibitors and were stored at -80 °C. Thawed cells, lysed by vortexing with glass beads, were clarified by centrifugation. Extracts were also prepared from DOA4 or doa4-10 transformants replica-plated onto selective media and incubated at 26 or 36 °C for 24 or 48 h. Cells were collected and processed as described above. After correcting for total protein, DNA topoisomerase I catalytic activity was assessed in a plasmid DNA relaxation assay (14, 27). Top1p levels and integrity were assessed by SDS-PAGE and immunoblotting with a polyclonal antibody specific for yeast Top1p and chemiluminescence (Amersham Biosciences).
Extracts of exponentially growing DOA4-HA and doa4-10-HA cells, prepared as above, were immunoblotted with a monoclonal HA-specific antibody. Alternatively, cells were harvested by centrifugation and resuspended in SDS-PAGE sample buffer, and the proteins were resolved by SDS-PAGE. Although full-length Doa4p was readily detected, truncated Doa4-10-HA protein was sporadically observed at 36 °C. The inclusion of phosphatase inhibitors, N-ethylmaleimide, additional protease inhibitors, or osmotic stabilizers or extraction with high salt failed to enhance Doa4-10-HA stability.
Isolation of Dosage Suppressors—To isolate dosage suppressors, doa4-10 cells, co-transformed with YCpSctop1T722A·L and YEp-FY250 DNA library, were incubated on SC-uracil, -leucine plates at 36 °C. To ensure that suppression of doa4-10 cell sensitivity to top1T722A was linked to the YEp-FY250 vector, individual transformants were plated on SC-leucine, 5-FOA plates at 26 °C and re-screened for top1T722A-induced lethality at 36 °C. Transformants exhibiting YEp-FY250-dependent viability at 36 °C were selected and the YEp-DNAs isolated. Data base queries and subcloning of putative suppressors into YEp24-PL confirmed the identity of UBI2, UBI4, and SML1 as dosage suppressors of doa4-10.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
In backcrosses, the ts phenotype of individual tah mutants segregated as a recessive, single gene defect. As the strains were top1, cell viability could be assessed in the presence or absence of plasmid-encoded Top1p or Top1T722Ap. Wild type cells were viable at 26 and 36 °C, independent of TOP1 or top1T722A expression (Fig. 1). Furthermore, low levels of vector expressed Top1p were insufficient to confer CPT sensitivity (Tables II and III). In contrast, the viability of tah22 cells expressing top1T722A was diminished at 36 °C, which was further exacerbated by CPT (Fig. 1 and Tables II and III). Similar patterns of conditional sensitivity to top1T722A and CPT were observed with tah6, tah14, and tah20 cells (Tables II and III; not shown). Consistent with other tah mutants, such as cdc45-10 and dpb11-10 (7, 39), these mutants exhibited ts hypersensitivity to the DNA replication inhibitor, HU (Table II). However, the lack of tah22 and tah14 enhanced sensitivity to DNA-damaging agents, such as the alkylating agent methyl methanesulfonate and UV light (Table II), suggest selective alterations in these mutant strains in response to DNA replication-induced lesions.
|
|
|
DOA4 encodes a deubiquitinating enzyme (DUB) that associates with the 26 S proteasome to maintain ubiquitin homeostasis by removing poly-ubiquitin chains from proteins targeted for proteasomal degradation (41, 42). Doa4p also functions at a late stage in endocytosis to recycle ubiquitin from plasma membrane proteins targeted for vacuolar degradation (43). One mechanism then, of potentiating the cytotoxic activity of Top1T722Ap or CPT, would be increased Top1p levels either as a direct consequence of defective ubiquitin-mediated proteolysis or as indirect effects on protein turnover. To address this, extracts were prepared of DOA4 and doa4-10 cells transformed with YCpScTOP1 or YCpSctop1T722A, grown at 26 °C, and then shifted to 36 °C for 4-24 h. Western blot analysis revealed no variation in Top1p or Top1T722Ap levels (data not shown), whereas the specific catalytic activity of wild type and mutant enzymes in plasmid DNA relaxation assays was unaffected by temperature shift (Fig. 2). As reported previously (26), Top1T722Ap activity was 
5-fold lower than comparable levels of Top1p. However, the extent of TopT722Ap-catalyzed plasmid DNA relaxation was identical in extracts prepared from doa4-10 and DOA4 strains grown at 26 or 36 °C (compare lanes highlighted with a single and double asterisk, respectively, in Fig. 2). As the ts phenotype of doa4-10 cells was more pronounced on solid media, cells were also harvested from plates, again with no detectable difference in Top1p or Top1T722Ap levels or activity (data not shown). Thus, the enhanced sensitivity of doa4-10 cells to Top1p poisons was not due to alterations in enzyme stability.
|
|
strains revealed considerable phenotypic differences (Table II). doa4 cells alone exhibited a slow growth phenotype at 36 °C (as reported in Ref. 58) and were hypersensitive to HU at all temperatures. In contrast, only doa4-10 cell growth was impaired in the presence of high salt. These data suggest that the rhodanese domain induces a distinct set of cellular responses from that observed in doa4 cells. However, as both strains exhibited a tah ts phenotype (Table II), the loss of the DUB catalytic residues sufficed to sensitize yeast to top1T722A-induced DNA damage. This appeared to be restricted to Doa4p-regulated events, as deletion of the closely related Ubp5p, or more distinct Ubp11p and Ubp12p, failed to enhance the cytotoxic activity of Top1T722Ap or CPT (data not shown).
SLA1 and SLA2 were initially identified in a genetic screen for mutants that were synthetically lethal with abp1 (45). Both gene products function in polarized cell growth and are required for normal cortical cytoskeletal organization. Sla1p is a component of cortical actin patches and plays a role in endocytosis (45, 59). The interaction of Sla1p with Sla2p also affects endocytosis by regulating actin dynamics (46). The sla1-10 allele (Fig. 3) contained a nonsense mutation at Trp-500 and a Glu-188 to Lys substitution. Comparisons of sla1-10 and sla1 strains failed to discern any phenotypic differences (Table II, data not shown), and neither strain was ts for growth at temperatures up to 36 °C. SLA1 mutations in tah20 were not defined nor were SLA2 alterations in sla2-10.
Genetic Interactions between DOA4, SLA1, SLA2, and CDC45—To ascertain whether Doa4p was functionally linked to the diverse set of tah mutants isolated in our screen, doa4-10,tah double mutants were examined. In contrast to the single mutants, double sla1-10,sla2-10, doa4-10,sla1-10, and doa4-10,sla2-10 mutants were all inviable at 36 °C in the absence of DNA damage (Fig. 4 and data not shown). A similar synthetic lethal phenotype was observed with doa4-10 or doa4 in combination with sla1. Genetic and functional interactions between SLA1 and SLA2 have been reported (45); however, our results suggest Doa4p also shares an essential function with Sla1p and Sla2p.
|
cells (39).
Taken together, these data indicate that the defects in DNA replication induced by cdc45-10 may be exacerbated by the loss of Doa4p function. As with rad9, this may be a consequence of inappropriate DNA damage/S-phase checkpoint responses leading to alterations in cell cycle arrest. Given the genetic interactions between DOA4, SLA1, and SLA2, such events might arise from alterations in cell morphology and actin cytoskeleton reorganization as cells progress through the cell cycle. To address these issues, doa4-10 and sla1-10 cell morphology was examined.
doa4-10 and sla1-10 Mutants Exhibit Alterations in Cell Morphology—Wild type and mutant strains, grown at 26 °C or shifted to 36 °C, were fixed and stained with rhodamine-phalloidin to visualize the distribution of F-actin. In representative images of wild type DOA4 cells (Fig. 5, a-h), cortical actin patches concentrate at the incipient bud site (a and b). As the bud emerges (c and d), cortical actin patches are mostly confined to the apical end of the bud (single arrow), with actin cables aligned with the axis of bud growth (double arrows in e and f). A switch to isotropic bud growth is accompanied by a redistribution of actin patches over the bud surface (g), and following cytokinesis, the patches relocalize to the mother-bud junction (h).
|
strains, sla1-10 cells exhibit gross defects in cortical actin patch organization, with a few large patches visible per cell (data not shown). The alterations in doa4-10 cell morphology were more subtle. First, doa4-10 cells cultured at 36 °C were larger in size, particularly the large-budded cells, with evidence of autophagic bodies and cell lysis in the DIC images (data not shown). A survey of >400 DAPI-stained cells indicated 45% of doa4-10 cells were large-budded, 60% of which had a single nuclear mass. Corresponding values for DOA4 cells cultured under the same conditions were 27 and 30%, respectively. Moreover, in contrast to wild type cells, a significant percentage of doa4-10 and sla1-10 cells were lysed (6 and 10%, respectively). Most of the ghosts appeared to have large buds. Double doa4-10,sla1-10 mutants were particularly fragile, with extensive cell lysis and clumping evident shortly after shift to 36 °C. The pattern of actin staining in doa4-10 cells was consistent with these alterations in cell cycle distribution (Fig. 5, i-p). The cortical patches were appropriately localized to the bud site and emerging bud (i-l). Actin cables radiating into the mother cell were apparent (double arrows, l), although not as well organized as in wild type DOA4 cells. As bud size increased (n-p), the actin patches became isotropically distributed and the cables more diffuse. In many large-budded cells with segregated nuclear masses, the actin patches failed to relocalize to the mother-bud junction (compare n and p with h), consistent with a defect in cytokinesis in doa4-10 cells. Whether this is caused by actin cytoskeletal defects or is simply reflected by the isotropic distribution of actin patches and cables has yet to be addressed.
Electron microscopy was also used to examine doa4 and sla1 mutant cell morphology. In contrast to wild type cells (Fig. 6A), enlarged and fragmented vacuoles are evident in 30% of the doa4-10 cells grown at 36 °C. As reported for sla1 cells (44, 46, 59), thick cell walls were also evident in sla1-10 mother cells but not in the bud. Significant cell lysis was apparent, with discontinuities in cell wall structure visible at the mother-bud juncture (Fig. 6A). Lysis was more pronounced with the double doa4-10,sla-10 mutant, coincident with more severe cell wall defects. At 26 °C, double or triple cell wall layers were evident in 40% of the double mutant cells. At 36 °C, 80% of the cells had multiple cell wall layers; 40% of the cells were lysed ghosts, and 30% of these were 2-3 times larger than wild type cells. As seen in Fig. 6A (doa4-10,sla1-10), some layers appeared to have pealed away from the cell wall structure. Whether this induces, or results from, cell lysis remains unclear.
|
and doa4-10 mutants further distinguished doa4 cells from cells expressing the N-terminal rhodanese domain of Doa4p. In addition to alterations in vacuolar architecture, 25-30% of doa4 cells accumulated stacks of curved cisternal membranes, independent of temperature (Fig. 6B). These structures resemble pre-vacuolar class E compartments reported in some vps mutants (52). Doa4p functions late in endocytosis to remove ubiquitin from plasma membrane proteins targeted for vacuolar degradation; however, class E endosomal compartments have not been reported in doa4 mutants (43). In contrast, stacked membranes were not detected in doa4-10 cells at either temperature, suggesting that the loss of Doa4p catalytic activity is insufficient to elicit this phenotype.
Given the role of Doa4p in endocytosis and vacuolar mediated degradation, the possibility remained that defects in endocytosis, and not ubiquitin homeostasis per se, enhanced doa4-10 cell sensitivity to Top1p poisons. However, deletion of VPS4, which functions immediately after Doa4p in the formation of vesicles that fuse with the vacuole (61), had no effect on cell sensitivity to top1T722A or CPT (data not shown). Instead, vps4 complemented the tah phenotype of doa4-10 cells, supporting the notion that other DUBs recycle ubiquitin from plasma membrane proteins that accumulate in the pre-vacuolar compartment. Thus, defects in endocytosis and vacuolar mediated degradation did not modulate cell sensitivity to Top1p-induced DNA lesions.
Dosage Suppressors of doa4-10 top1T722Ap-Hypersensitivity—To address more directly the mechanism of Doa4p protection against Top1p poisons, dosage suppressors of top1T722A- induced doa4-10 cell death at 36 °C were isolated from a YEp-FY250 genomic DNA library (detailed under "Experimental Procedures"). Consistent with the vps4,doa4-10 results, increased dosage of the ubiquitin genes UBI4 and UBI2 suppressed doa4-10 hypersensitivity to top1T722A and HU at 36 °C (data not shown, Fig. 7A). YEpUBI2 and YEpUBI4 also rescued doa4-10 temperature sensitivity to high salt and glycerol (data not shown).
|
or rad53 cell viability but not checkpoint function. The Mec1p/Rad53p kinase cascade also targets Sml1p for degradation following checkpoint activation (63). sml1 cells were not hypersensitive to top1T722A or CPT (Table IV, data not shown); thus, the activity of SML1 as a dosage suppressor could not be attributed to a direct effect on Sml1p stability or turnover. An alternative possibility is that increased Sml1p levels reinforce activation of the Mec1p-Rad53p checkpoint, thereby slowing S-phase progression and/or enhancing replication fork stability and diminishing Top1p poison-induced DNA lesions. This model requires a functional MEC1/RAD53 checkpoint in doa4-10 cells and may derive from the inactivation of other DNA damage checkpoint functions. Indeed, the latter point would be consistent with the ts slow growth phenotype of the cdc45-10 mutant in combination with either doa4-10 (Fig. 4) or rad9 (39) and the accumulation of doa4-10 cells with a large-budded phenotype in the absence of DNA damage.
|
|
), and RAD53 (sml1) were hypersensitive to top1T722A. However, the rad9, rad17, and rad24 data obtained here indicate Top1T722Ap-induced damage was more cytotoxic than lesions induced by CPT. Tel1p and Tof1p also protected against Top1p poisons. In contrast to Schizosaccharomyces pombe (67), chk1 cells were no more sensitive to Top1T722Ap or CPT than wild type cells. Although Chk1p may function in response to higher concentrations of CPT, these data preclude a significant role for Chk1p at low levels of Top1p poisons.
In combination with doa4-10, all checkpoint mutants were either more sensitive to Top1p poisons or as sensitive as the single checkpoint null strain (Table IV). The exception was the doa4-10,rad9 double mutant, where the cold sensitivity of rad9 cells to top1T722A and CPT was partially suppressed by the doa4-10 mutation. Similarities in doa4-10 and doa4-10,rad9 sensitivity to top1T722A and CPT suggest doa4-10 is epistatic to rad9, although in a manner distinct from Rad24p/Rad17p, Chk1p, or Mec1p/Rad53p.
Doa4 and Tdp1 Act in the Distinct Pathways in Response to Top1p-DNA Damage—To extend these findings, tyrosyl DNA phosphodiesterase I (Tdp1p) function was also examined. Tdp1p specifically hydrolyzes the 3'-phosphotyrosyl linkage between Top1p and DNA to yield a 3'-phosphoryl DNA end (68). Increasing evidence suggests this enzyme plays an important role in the repair of lesions induced by Top1p-DNA intermediates (37, 38). Pouliet et al. (68) demonstrated the enhanced sensitivity of tdp1 cells to Top1p poisons when the RAD9 gene was deleted, positing that Rad9p and Tdp1p constitute parallel pathways that protect cells from CPT. Consistent with this view, the enhanced sensitivity of doa4-10 cells to low constitutive levels of Top1T722Ap or CPT-induced DNA damage at 36 °C is exacerbated in doa4-10,tdp1 double mutants (Fig. 9 and Table III). Moreover, the double mutant is unable to tolerate Top1T722Ap at any temperature. In contrast, tdp1 cells were CPT-resistant at all temperatures and only exhibit enhanced sensitivity to Top1T722Ap at 36 °C when co-treated with CPT. Thus, defects in Doa4p function potentiate the cytotoxic activity of low doses of Top1p poisons in cells deleted for TDP1.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Although detection of the truncated Doa4-10-HA protein proved difficult, phenotypic differences between doa4 and doa4-10 cells provided indirect evidence of rhodanese domain function. Perhaps most compelling was the selective accumulation of pre-vacuolar E class vesicles evident in doa4 cells but absent in congenic doa4-10 strains (Fig. 6B). Differences in growth rate at 36 °C, temperature-sensitive growth in the presence of high salt, and the enhanced sensitivity of doa4 to HU at all temperatures were also suggestive of rhodanese domain function (Table II). However, the DUB activity of Doa4p appeared critical for cellular resistance to Top1p poisons, as both doa4-10 and doa4 strains exhibited enhanced sensitivity to Top1T722Ap and CPT.
Deletion of VPS4 inhibits a late step in endocytosis, immediately downstream of Doa4p-catalyzed recycling of ubiquitin from membrane proteins in multivesicular bodies that are targeted for degradation following vesicle fusion with the vacuole (reviewed in Ref. 69). The ability of vps4 to complement the tah phenotype of doa4-10, yet eliminate a critical step in vacuolar trafficking, suggests defects in endocytosis per se do not alter cell sensitivity to Top1 poisons. Rather the ability of other DUB activities to recycle ubiquitin from membrane proteins targeted for vacuolar degradation and the isolation of UBI2 and UBI4 as dosage suppressors of doa4-10 argue for a direct effect of ubiquitin homeostasis on cellular responses to these S-phase toxins. Yet this phenotype appears restricted to Doa4p-catalyzed events, as deletion of the closely related Ubc5p or the more distantly related Ubp12 or Ubp11 enzymes had no adverse effects on cell sensitivity to Top1p poisons.
In mammalian cells, transcription-dependent degradation of ubiquitinated Top1p by the 26 S proteasome has been reported in response to high concentrations of CPT (70, 71). Whereas this may constitute a mechanism of CPT resistance, the converse does not appear to be true, i.e. the absence of ubiquitin-mediated down-regulation of Top1p enhancing drug sensitivity. The lack of detectable alterations in Top1p or Top1T722Ap protein levels or catalytic activity in extracts of DOA4 and doa4-10 cells cultured under a variety of experimental conditions precludes a direct effect of ubiquitin-mediated effects on Top1 protein stability or activity. Indeed, our previous studies of pGAL1-promoted expression of top1T722A indicate substantially elevated levels of protein expression are required to induce cell lethality in wild type cells (26). The results reported here are inconsistent with the up-regulation of Top1p levels or activity in doa4-10 mutant cells. Furthermore, alterations in HU sensitivity were Top1p-independent. Whether putative human orthologs of DOA4, such as UBP8 or UBP2, play a similar role in modulating human cell sensitivity to Top1p poisons is currently being investigated with small interfering RNA approaches.
Genetic interactions between SLA1 and SLA2 have been reported (45, 46). However, the ts synthetic lethality of doa4 mutants in combination with sla1 or sla2 mutants in the absence of DNA damage was surprising. Morphological inspection of the single and double mutants indicated that the defects in cell wall structure, reported for sla1 mutants (44, 46, 59), was exacerbated by the loss of Doa4p DUB activity. The result was a pronounced increase in cell lysis, particularly at the largebudded stage of the cell cycle. The obvious prediction that accumulation of cells with large buds would enhance cell fragility was also borne out by the dramatic drop in doa4-10,sla1-10 double mutant cell viability in response to HU activation of the S-phase checkpoint. Cell fragility could be a consequence of defects in cell wall structure due to Doa4p-mediated effects on vesicle trafficking. However, in a recent proteomics approach to identify ubiquitinated proteins in yeast, Sla1p and Sla2p were among the proteins identified (72). This raises the interesting possibility that ubiquitin conjugation may directly affect Sla1p and Sla2p function. Whether cell lysis is restricted to the late S/G2 cell cycle compartment or would also be evident in response to mitotic poisons, such as nocodazole, has yet to be determined. Nevertheless, our results establish a functional link between ubiquitin homeostasis, actin cytoskeletal organization, and cellular resistance to Top1p poisons.
Insights into potential mechanisms of Doa4p-mediated protection from Top1p-induced DNA damage came from the isolation of SML1 as a selective albeit weak dosage suppressor of doa4-10 temperature sensitivity to Top1T722A. Sml1p inhibition of ribonucleotide reductase is antagonistic to the induction of this enzyme by the Mec1p/Rad53p/Dun1p signaling pathway in response to replicative stress (47, 62). Indeed, Zhao et al. (63) demonstrated that Sml1p is also targeted for ubiquitin-mediated degradation following activation of the Mec1p/Rad53p checkpoint. As sml1 cells were not hypersensitive to CPT or Top1T722Ap, it is unlikely that Sml1p down-regulation is the basis for the doa4-10 tah phenotype. Furthermore, treatment with HU suffices to induce doa4-10 cell cycle arrest, providing evidence that at least some aspects of Mec1p/Rad53p signaling pathways remain functional. Taken together, this raises the possibility that increased dosage of Sml1p reinforces activation of the Mec1p/Rad53p signaling pathway, enhancing the DNA damage response and possibly repair.
A survey of replication and DNA damage checkpoint mutants provided further support of genetic interactions between RAD9 and DOA4, as doa4-10 partially suppressed the enhanced sensitivity of rad9 cells to top1T722A at 26 °C. In contrast, all other doa4-10 checkpoint mutant combinations examined potentiated the cytotoxic action of Top1p poisons. Several lines of evidence indicate the functional interaction of Doa4p and some aspect of DNA damage signaling. First is the accumulation of large-budded doa4-10 cells with a single DNA mass at 36 °C, even in the absence of DNA damage. Singer et al. (73) reported previously a role for Doa4p in regulating the coordination of DNA replication. However, in their studies, the re-replication of DNA was only observed at 38 °C and was not suppressed by increased levels of UBI4. As the tah phenotype of doa4-10 cells was evident at 36 °C and was fully suppressed by increased dosage of UBI2 or UBI4, such a re-replication mechanism seems unlikely. An alternative explanation may be a defect in replication fork stability, which induces DNA repair pathways. Indeed, several recent studies (65, 74, 75) demonstrate Mrc1p and Tof1p function to stabilize stalled replication forks and may prevent the processing of blocked forks to form alternative structures that trigger components of the DNA damage checkpoint, such as Rad9p. Analogous to Rad9p in response to DNA damage, Mrc1p acts as a mediator to activate Rad53p in response to replication stress. Thus, independent mechanisms may enhance signaling through a common kinase cascade. Increased dosage of SML1 may reinforce the Mrc1p/Tof1p pathway and diminish signaling/repair by Rad9p.
The synthetic slow growth phenotype induced in rad9,cdc45-10 cells upon shift to 35 °C (39) suggests that alterations in processive DNA replication induced by cdc45-10 trigger the Rad9p-dependent DNA damage checkpoint. The fact that doa4-10,cdc45-10 cells exhibit a similar phenotype is consistent with the genetic interaction of DOA4 and RAD9 in altering cellular responses to Top1T722Ap (Table IV). The enhanced CPT/top1T722A sensitivity of doa4-10 cells deleted for tyrosyl phosphodiesterase I (Table III) also suggests doa4-10-induced alterations in repair pathways that act in parallel with Tdp1p to effect the repair of Top1-DNA lesions, such as the Rad9p DNA damage checkpoint (68). Whether this is due to specific alterations in Rad9p function or indirect effects on checkpoint-induced repair activities has yet to be investigated. More thorough analyses of checkpoint signaling and extragenic suppressors of doa4-10 sensitivity to Top1T722Ap are also being pursued to elucidate the functional link between alterations in Doa4p-mediated ubiquitin homeostasis and cellular response to Top1p poisons.
|
FOOTNOTES |
|---|
Present address: Dept. of Biology, University of Rome Tor Vergata, 00133 Rome, Italy.
|| Present address: Dept. of Genetics and Development, Columbia University, College of Physicians and Surgeons, New York, NY 10032.
Present address: EMBL Heidelberg, D-69117 Heidelberg, Germany.
¶¶ To whom correspondence should be addressed: Dept. of Molecular Pharmacology, St. Jude Children's Research Hospital, 332 N. Lauderdale, Memphis, TN 38105. Tel.: 901-495-2315; Fax: 901-495-4290; E-mail: Mary-Ann.Bjornsti{at}stjude.org.
1 The abbreviations used are: CPT, camptothecin; ts, temperature-sensitive; HU, hydroxyurea; DAPI, 4,6-diamidino-2-phenylindole; DUB, deubiquitinating enzyme; FOA, fluoroorotic acid; DIC, differential interference contrast.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
