¹Department of Molecular Virology and Microbiology, Baylor College of Medicine, Houston, Texas, TX 77030, USA

²Program in Cell and Molecular Biology, Baylor College of Medicine, Houston, Texas TX 77030, USA

Correspondence to: S J Marriott, Department of Molecular Virology and Microbiology, Baylor College of Medicine, Houston, Texas, TX 77030, USA

Transient HTLV-1 Tax expression suppresses cellular nucleotide excision repair, and this effect correlates with Tax transactivation of the proliferating cell nuclear antigen promoter. The inability to repair DNA damage typically induces apoptotic cell death. Therefore, we investigated the effect of Tax-mediated suppression of DNA repair on apoptosis in stable Tax-expressing cells. Constitutive Tax expression reduced cellular nucleotide excision repair activity compared with parental and control cells. Tax-expressing cells were also more sensitive to apoptosis induced by DNA damaging agents than control cells. Even though Tax-expressing cells displayed reduced DNA repair, they showed increased DNA replication following UV damage. These results suggest that Tax suppresses the cell's ability to repair DNA damage and stimulates DNA replication even in the presence of damage. The inability to repair DNA damage is likely to stimulate apoptotic cell death in the majority of Tax-expressing cells while the ability to promote DNA replication may also allow the survival of a small population of cells. We propose that together these effects contribute to the monoclonal nature and low efficiency of HTLV-1 transformation. Oncogene (2000) 19, 2240-2248

HTLV-1; Tax; apoptosis; nucleotide excision repair; DNA damage

Introduction

Human T cell leukemia virus type 1 (HTLV-1) is the causative agent of at least two human diseases, adult T-cell leukemia (ATL) (Poiesz et al., 1980) and tropical spastic paraparesis/HTLV-1 associated myelopathy (TSP/HAM) (Gessain et al., 1985). The onset of both diseases is preceded by a long period of clinical latency, frequently lasting three to four decades. In addition, less than 5% of all infected individuals develop an HTLV-1 associated disease. Comparison of polyclonal integration patterns of the HTLV-1 provirus among various patients suggests that transformation does not occur by insertional mutagenesis (Seiki et al., 1984). However, transformed cells from individual ATL patients display monoclonal or oligoclonal integration of the viral genome suggesting that infection precedes transformation and that transformation is a rare event among HTLV-1 infected cells (Yoshida et al., 1984). Although cellular transformation by HTLV-1 is a multistage process, the molecular basis of this process is not well understood.

The HTLV-1 genome contains a unique pX region that encodes two important regulatory proteins, Tax and Rex. Tax is essential for viral replication and potently activates transcription of HTLV-1 genes. In addition, Tax activates transcription of several cellular genes involved in cell proliferation (for review, see Franklin and Nyborg, 1996; Ressler et al., 1996). Transactivation of cellular gene expression by Tax is thought to play an important role in HTLV-1 transformation. In fact, Tax has been identified as the key viral mediator of human T-cell transformation, and transformation of other cell types (for review, see Franchini, 1995).

An accumulation of DNA damage has been associated with Tax expression in several studies. This evidence includes induction of micronuclei in Tax transfected cells (Majone et al., 1993; Saggioro et al., 1994) and enhanced mutation frequency of the cellular genome in Tax-expressing cells (Miyake et al., 1999). These effects could result either directly from Tax stimulating DNA damage, or indirectly from Tax inhibiting repair of naturally occurring DNA damage. Most existing evidence supports the latter possibility; that Tax impairs the cell's ability to repair DNA damage. For example, it has been shown that Tax represses the expression of human -polymerase, an important cellular DNA repair enzyme (Jeang et al., 1988). In addition, the ability of Tax to inhibit cellular nucleotide excision repair (NER) was the first report of a direct Tax effect on DNA repair (Kao and Marriott, 1999). Subsequently, suppression of base excision repair (BER) by Tax was also demonstrated (Philpott and Buehring, 1999). Thus, Tax protein may facilitate malignant progression by decreasing the cell's ability to repair genetic damage in infected cells. Indeed, cytogenetic studies have shown that leukemic cells from ATL patients contain clonal chromosomal abnormalities, most commonly including deletions and translocations (Fujita et al., 1989; Kamada et al., 1992). Although chromosomal changes are common in transformed ATL cells, no consistent abnormalities have been found in all ATL patients (Sandberg, 1991), suggesting that a generalized deregulation of host DNA replication and repair machinery occurs during the development of ATL.

The presence of unrepaired genomic damage typically induces apoptotic cell death. Therefore, we wanted to investigate whether the inhibition of DNA repair by Tax induced apoptotic cell death. There have been conflicting reports about the effect of Tax on apoptosis. Tax has been shown to induce apoptosis in a variety of systems (Chen et al., 1997; Chlichlia et al., 1995,1997; Fujita and Shiku, 1995; Hall et al., 1998; Kitajima et al., 1996; Yamada et al., 1994), consistent with its ability to reduce DNA repair. However, in other reports Tax has been shown to inhibit apoptosis (Brauweiler et al., 1997; Copeland et al., 1994; Tsukahara et al., 1999; Mulloy et al., 1998), supporting its role as a transforming protein and inducer of T cell proliferation. The ability of Tax to activate or inhibit apoptosis may depend on the method used to induce apoptosis. None of the studies reported to date have directly examined the effect of Tax on apoptosis induced by DNA damaging agents.

Previously, the ability of Tax to inhibit nucleotide excision repair was investigated using transient expression assays (Kao and Marriott, 1999). To more closely mimic the expression of Tax in transformed cells, the current study was designed to investigate the effects of constitutive Tax expression on cellular DNA repair activity and apoptosis. Consistent with previous results, cells stably expressing Tax were defective in cellular NER. Since DNA damage is a common trigger of apoptosis, the effect of Tax on apoptosis following exposure to DNA damaging agents was tested. Tax expressing cells displayed increased cell death following DNA damage. This cell death was due to apoptosis as demonstrated by nuclear condensation and Annexin V staining. In addition to effects on DNA repair and apoptosis, Tax expressing cells displayed elevated DNA replication. The ability of Tax to induce apoptosis and DNA replication may contribute the clonal nature of HTLV-1 transformation.

Results

Cellular NER in normal and Tax-expressing fibroblasts

In order to determine the effect of constitutive Tax expression on DNA repair, a pool of CREF cells that had been selected for stable Tax expression were used. CREF cells were chosen because we have previously demonstrated suppression of DNA repair following transient transfection of Tax into CREF cells, and rat fibroblast cells have been used extensively to study Tax-mediated transformation (Matsumoto et al., 1997; Smith and Greene, 1991; Tanaka et al., 1990; Yamaoka et al., 1996). Western blot analysis was performed to examine Tax expression in these cells prior to the repair assays. Tax expression was detected in CREF-Tax cells but not in control CREF-neo and parental CREF cells (Figure 1a). We have previously shown that Tax activates expression from the proliferating cell nuclear antigen (PCNA) promoter (Kao and Marriott, 1999; Ressler et al., 1997). Therefore, CREF-Tax cells would be expected to express increased levels of endogenous PCNA when compared to CREF and CREF-neo cells. Western blots were performed to examine the level of PCNA expression in asynchronized cells. The expression of PCNA in CREF-neo and parental CREF cells was similar, while PCNA expression in CREF-Tax cells was approximately eightfold higher (Figure 1b). The PCNA band in CREF-Tax cells runs slightly faster than PCNA in CREF or CREF-neo cells. PCNA is a phosphoprotein (Prosperi et al., 1994) and we have previously observed a faster migrating form of PCNA in Tax-expressing cells which we attribute to differential phosphorylation (Kao and Marriott, 1999).

In a previous report, we observed suppression of DNA repair in cells expressing Tax transiently (Kao and Marriott, 1999) therefore, it was important to determine whether stable Tax expression also reduced DNA repair activity, and to investigate the effects of suppressed DNA repair on cellular apoptosis. To characterize the effect of constitutive Tax expression on DNA repair activity, a reporter plasmid (pMSV-Luc) was UV-irradiated, or mock-treated, and then transfected into CREF-Tax, CREF-neo, or parental CREF cells (HCR assay). Since UV induced lesions provide a strong block to transcription, expression of the irradiated plasmid was greatly diminished. Recovery of luciferase expression reflects the extent of repair activity in these cells. CREF and CREF-neo cells showed similar abilities to repair the transfected plasmid, and express luciferase (Figure 2a). However, luciferase activity in CREF-Tax cells was less than 50% of that in cells without Tax, indicating that CREF-Tax cells were defective in cellular DNA repair activity.

To further examine DNA repair activity, Tax-expressing cells were arrested in the cell cycle by serum starvation and addition of hydroxyurea to prevent DNA replication. Cells were then UV irradiated and labeled with ³H-thymidine, which will only be incorporated in cells actively undergoing DNA repair, as DNA replication is blocked (UDS assay). The extent of [³H]thymidine incorporation measures the cell's ability to repair DNA damage, and is referred to as 'repair activity'. Repair activity of parental CREF cells was set to 100%, and was used to determine the relative repair activity of CREF-neo and CREF-Tax cells. The CREF and CREF-neo cells had similar repair activities while the repair activity of CREF-Tax cells was reduced to about 60% of that observed in CREF cells (Figure 2b). Thus, results of the HCR assay (Figure 2a) and the UDS assay (Figure 2b) demonstrate that stable Tax expression reduces cellular DNA repair activity.

Replication of damaged DNA in Tax-expressing cells

We have previously proposed that Tax may induce a mutator phenotype in which the mutation rate within a population of cells is increased, and replication of the damaged DNA allows for mutations to become fixed in the genome (Kao and Marriott, 1999). Induction of such a phenotype would likely be a critical early step in Tax-mediated transformation. Establishment of a mutator phenotype would require Tax suppression of DNA repair as demonstrated above. However, an important additional aspect of this model is that cells must be capable of replicating damaged DNA. Therefore, the ability of Tax-expressing cells to replicate DNA following damage was examined using a [³H]thymidine incorporation assay.

To measure DNA replication, cells were labeled with [³H]thymidine in the absence of hydroxyurea therefore, both DNA replication and repair were measured. Since DNA repair activity is minimal compared to replication activity, thymidine incorporation predominantly reflects DNA replication. The replication activity of CREF-neo and CREF-Tax cells was determined and compared to parental CREF cell activity. In the absence of UV irradiation, CREF, CREF-neo, and CREF-Tax cells incorporated similar levels of [³H]thymidine (Figure 3, left). Following UV irradiation, replication activity of CREF-Tax cells was about 1.5-fold higher than parental CREF and CREF-neo cells (Figure 3, right). In related studies, we have found that CREF-Tax cells progress through the cell cycle more rapidly following UV irradiation, than CREF-neo cells (F Lemoine and SJ Marriott, unpublished data). Thus, in the presence of DNA damage, Tax-expressing cells suppress DNA repair, and induce DNA synthesis, resulting in fixation of mutations in the genome. These features are consistent with the classic mutator phenotype.

Effect of Tax on cell death induced by DNA damaging agents

Cells that are unable to repair DNA damage typically undergo apoptotic cell death if the damage is sufficient. Since we found that Tax suppresses cellular DNA repair activity and allows cells to replicate in the presence of DNA lesions, we wanted to determine whether these events would sensitize cells to apoptotic death. CREF-neo and CREF-Tax cells were treated with different doses of UV irradiation (254 nm), and, 24 h later, cell viability was determined by trypan blue staining. Although both cell types showed a dose dependent decrease in cell viability following UV irradiation (Figure 4a), CREF-Tax cells exhibited more sensitivity to UV damage (survival rate: 50% for CREF-Tax vs 75% for CREF-neo at 10 J/m²). The cell death observed in CREF-neo cells likely resulted from their inability to completely repair large amounts of DNA damage. This effect was accentuated in CREF-Tax cells where DNA repair is compromised.

A time course of CREF-neo and CREF-Tax cell death following 10 J/m² UV irradiation was performed to further characterize the effects of Tax on cell viability. Both cell types showed a time dependent decrease in cell viability (Figure 4b) and CREF-Tax cells were more sensitive to UV irradiation than CREF-neo cells (survival rate: 50% for CREF-Tax vs 75% for CREF-neo at 24 h).

Two additional agents were used to determine whether Tax has similar effects on cell death induced by other DNA damaging agents. Cisplatin and 5-fluorouracil (5-FU) are important chemotherapeutic agents used to treat human cancers (Kelland, 1993). The primary target of these drugs is DNA. Cisplatin is a DNA cross-linking agent, and 5-FU is a pyrimidine analog. Like UV irradiation, both drugs induce DNA adducts which are typically repaired by the cellular NER pathway. Similar to results shown above with UV induced damage, Tax inhibited DNA repair following exposure to these two drugs (data not shown). If these lesions are not repaired, cells may undergo apoptosis (Barry et al., 1990).

To determine whether Tax could sensitize cells to death following treatment with 5-FU or cisplatin, CREF-neo and CREF-Tax cells were treated with these agents, and cell survival was measured at various time points (Figure 5a,b). As observed following UV irradiation, CREF-Tax cells were more sensitive to 5-FU induced DNA damage than CREF-neo cells (survival rate: 68% for CREF-Tax vs 90% for CREF-neo at 48 h post-treatment). CREF-Tax cells were also more sensitive to cisplatin induced DNA damage than CREF-neo cells (survival rate: 50% for CREF-Tax vs 74% for CREF-neo at 48 h post-treatment).

Tax sensitizes cells to apoptosis induced by DNA damaging agents

We next wanted to determine whether the cell death observed in Tax expressing cells occurred as a result of apoptosis. Apoptosis is an important mechanism for elimination of cells that have suffered extensive DNA damage (Thompson, 1995). Some characteristic features of apoptosis include cell shrinkage, plasma membrane blebbing, presence of phosphatidylserine on the outer leaflet of the plasma membrane, and a series of nuclear changes including nuclear condensation, nuclear breakdown, and DNA fragmentation (Kerr et al., 1972). To determine whether the cell death observed in Tax-expressing cells following UV irradiation was due to apoptosis, nuclear condensation assays and Annexin V staining were performed.

Nuclear condensation, a late-stage marker of apoptosis, was examined in CREF-neo and CREF-Tax cells treated with UV irradiation, 5-FU, cisplatin, or mock treated. Twenty-four hours after treatment, cells were fixed and nuclei were stained with Hoechst 33342, a fluorescent DNA-binding dye. Untreated CREF-neo and CREF-Tax cells showed normal nuclear morphology with no detectable nuclear condensation (Figure 6a). This result demonstrated that sustained Tax expression does not independently sensitize cells to apoptosis. Condensed nuclei were infrequently observed in CREF-neo cells treated with UV (<1%), 5-FU (8%) or cisplatin (11%) (Figure 6c,e,g, respectively), indicating few apoptotic cells in these cultures. Increased numbers of condensed nuclei were detected in CREF-Tax cells treated with UV (31%), 5-FU (25%), or cisplatin (33%) (Figure 6d,f,h, respectively). These results are shown graphically in Figure 6b, and suggested that the cell death observed in Tax-expressing cells following DNA damage is due to apoptosis.

To investigate cells at early stages of apoptosis, a more sensitive and quantitative flow cytometric analysis using Annexin V staining was also performed. Annexin V is a 35 kD, Ca²⁺ dependent, phospholipid-binding protein that has high affinity for phosphatidylserine. In intact cells, Annexin V binds only to rare phosphatidylserines exposed to the external milieu (Raynal and Pollard, 1994). Early in apoptosis, phosphatidylserine is translocated from the inner to outer surface of the plasma membrane (Martin et al., 1995). Therefore increased Annexin V staining of intact cells identifies cells in early stages of apoptosis.

CREF-Tax and CREF-neo cells were treated with the three DNA damaging agents described above, UV, cisplatin, and 5-FU. Adherent cells were collected, stained with Annexin V-FITC and propidium iodide (PI), and analysed by flow cytometry. Annexin V positive and PI negative staining indicates cells that are undergoing apoptosis. A histogram showing Annexin V staining of PI negative cells is shown in Figure 7. Approximately 30% of CREF-Tax cells were sensitive to apoptosis induced by DNA damaging agents (UV, cisplatin, 5-FU) while less than 15% of CREF-neo cells were Annexin V positive. The effect of Tax on Taxol treated cells was also investigated to determine whether the stimulation of apoptosis was specific for DNA damaging agents. Taxol is an apoptosis-inducing agent that binds to tubulin, retards microtubule depolymerization, impairs mitosis, blocks progression through cells cycle, and facilitates apoptosis independent of DNA damage (Schiff and Horwitz, 1980). Although Tax sensitized cells to apoptosis following exposure to UV irradiation, 5-FU and cisplatin, CREF-Tax cells did not show increased sensitivity to Taxol induced apoptosis (Figure 7).

Since human T cells are the natural hosts of HTLV-1, the effect of Tax on apoptosis induced by DNA damaging agents in the human T cell line, Jurkat was investigated. Jurkat cells were transfected with Tax expression plasmid or mock transfected. Twenty-four hours after transfection, cells were treated with 5-FU, cisplatin or mock-treated, and Annexin V staining was performed 24 h after treatment. Similar levels of Annexin V positive cells were observed in mock and Tax transfected Jurkat cells without DNA damage. Higher levels of apoptosis were observed in Tax transfected Jurkat cells exposed to the DNA damaging agents, 5-FU and cisplatin, than mock transfected Jurkat cells (Figure 8). The difference between apoptotic cells in Tax transfected Jurkat cells was not as great as observed in CREF-Tax cells. This result is most likely due to the fact that the CREF-Tax cells were selected for stable Tax expression while only a portion of the Jurkat cells were successfully transfected to yield Tax expression.

Discussion

This study set out to determine the effect of constitutive HTLV-1 Tax expression on cellular NER, and cell death induced by DNA damaging agents. In a previous report, we showed that transient Tax expression interferes with cellular DNA repair activity (Kao and Marriott, 1999). However, it was not clear whether, upon long term expression of Tax, cells could adapt to overcome the interference with DNA repair. This report extends our earlier findings by demonstrating that cells stably expressing Tax are also defective in DNA repair activity (Figure 2). This is an important feature of Tax's effect on DNA repair because, if Tax suppression of DNA repair leads to an accumulation of DNA mutations, as our model predicts, it will likely require long term Tax expression to see this effect. This finding is consistent with other reports suggesting that Tax expression promotes an accumulation of DNA damage (Majone et al., 1993; Saggioro et al., 1994; Miyake et al., 1999), and that HTLV-1 transformed cells exhibit various chromosomal abnormalities (Fujita et al., 1989; Kamada et al., 1992). It is not likely that Tax directly induces mutations, but rather that Tax suppresses repair of DNA damage resulting from other sources.

The inability to repair DNA damage suggests that Tax expressing cells may be more prone to cell death. Recent studies have supported a role for Tax and/or HTLV-1 infection in the induction of apoptosis. Using a hormone inducible Tax expression system, Tax expressing cells were found to be sensitive to Fas ligand mediated cell death (Chen et al., 1997; Chlichlia et al., 1995), and ICE-proteases were shown to be involved in this Tax-mediated apoptosis (Chlichlia et al., 1997). It has also been shown that Rat-1 cells transformed by Tax undergo apoptotic cell death after serum deprivation and constitutive Bcl-2 expression blocks this Tax-mediated apoptosis (Yamada et al., 1994). Another study showed that Tax expression promotes an enhanced prooxidant state, followed by apoptotic cell death, and that this Tax-mediated cell death can be suppressed by anti-oxidant (Los et al., 1998). Tax also activates the NF-B transcription factor and subsequently TNF- expression, leading to apoptosis of osteoblasts (Kitajima et al., 1996). Using animal models, it has been reported that thymuses from mature outbred rabbits lethally inoculated with an HTLV-1 infected T cell line showed morphological and biochemical evidence of apoptosis (Leno et al., 1995). In all of these studies, the inducer of apoptosis was not known to be a DNA damaging agent. The current study is the first to demonstrate apoptotic cell death of Tax expressing cells induced by DNA damaging agents. The Taxol studies reported here suggest that the effect of Tax on apoptosis in our system is a result of its suppression of DNA repair, since Taxol mediated apoptosis, which occurs through a pathway other than DNA damage, was not affected by Tax.

Although evidence suggests that Tax either induces apoptosis or sensitizes cells to apoptotic death, there are also several reports showing an inhibition of apoptosis by Tax or HTLV-1 infection (Brauweiler et al., 1997; Copeland et al., 1994; Tsukahara et al., 1999; Mulloy et al., 1998). It has been reported that apoptosis induced by UV irradiation is inhibited in HTLV-1 infected cells (Brauweiler et al., 1997). These studies conflict with our results however, since infected cell lines were used in those studies, the inhibition of apoptosis may not have been Tax-specific. Tax also inhibited apoptosis induced by anti-APO-1 treatment (Copeland et al., 1994) or IL-2 withdrawal (Tsukahara et al., 1999). Tax has been shown to repress bax expression in vitro (Brauweiler et al., 1997), activate bcl-xl expression (Tsukahara et al., 1999) and inactivate p53 function (Mulloy et al., 1998), each of which suggests a role for Tax in inhibition of apoptosis. It remains unclear why some studies report stimulation of apoptosis by Tax while others report inhibition. The discrepancy among different studies may be due to differences in cell type or differences in methods used to induce or detect apoptosis.

Transformation by Tax is a slow process likely involving the deregulation of many cellular pathways (Franklin and Nyborg, 1996; Ressler et al., 1996). Although we propose that inhibition of DNA repair plays an important role in Tax-mediated transformation, other steps are certainly required. The activation of cellular genes dependent on CRE, SRE and NF-B transcription factors are likely to play important roles at other stages of transformation either to initiate or maintain the transformed phenotype. The exact mechanism by which Tax transforms cells is still unclear and several important clinical features related to Tax transformation remain to be explained at the molecular level.

In this study, we have demonstrated that Tax induces apoptotic cell death following UV-induced DNA damage and that Tax expressing cells show increased DNA replication following DNA damage. These results may explain the monoclonal nature of HTLV-1 transformation, the small percentage of infected individuals who progress to disease, the long period of clinical latency, and the presence of chromosomal abnormalities in HTLV-1 transformed cells.

Materials and methods

Cells, plasmids and antibodies

CREF (cloned rat embryonic fibroblast) cell pools selected for Tax expression (CREF-Tax) were established by transfecting CREF cells with Tax and neomycin resistance expression plasmids, and selecting for G418 resistance. Resistant cells were pooled and expanded. Control CREF cell pools (CREF-neo) were transfected with the neomycin-resistance gene alone, and selected as described above. The parental CREF cell line was maintained in Dulbecco's minimal essential medium with 10% fetal calf serum. CREF-Tax and CREF-neo cells were maintained in the same medium. Jurkat cells were obtained from the American Type Culture Collection (ATCC), and were grown in RPMI medium. Antibody against Tax was described previously (Lydy et al., 1998). Monoclonal antibody against PCNA (PC-10) was purchased from Santa Cruz Biotechnology.

Transfection

Transfection of CREF-Tax and CREF-neo cells was performed using calcium phosphate precipitation (Ressler et al., 1997). Transfection of Jurkat cells was performed by lipofection using DMRIE-C reagent according to the manufacturer's manual (Life Technologies).

Host cell reactivation (HCR) assay

The pMSV-Luc plasmid was exposed to 1000 J/m² of UV-C light (254 nm) using a Stratalinker (Stratagene). Approximately 2´10⁵ CREF, CREF-neo, or CREF-Tax cells in duplicate 60 mm dishes were transfected with 1 g of UV-damaged or control non-irradiated pMSV-Luc plasmid together with 1 g of pSV2-CAT to control for transfection efficiency. Forty-eight hours after transfection, cells were lysed in 400 l reporter lysis buffer (Promega) by freeze thawing, and transferred to 1.5 ml microfuge tubes. A 50 l aliquot of the lysate was removed for analysis of luciferase activity, and a 25 l aliquot was removed for analysis of CAT activity. Luciferase activity was normalized to CAT activity of the same plate. The repair activity of each cell line was determined by dividing it's normalized luciferase activity by the normalized luciferase activity of a matched transfection containing undamaged pMSV-Luc. Repair activity of CREF cells was set to 100%, and repair activity of other cell lines was reported as a per cent of the parental CREF cell line.

Unscheduled DNA synthesis (UDS) and DNA replication assays

For UDS assays, confluent (approximately 10⁶) CREF, CREF-Tax, and CREF-neo cells in duplicate 60 mm dishes were cultured in DMEM containing 0.5% serum for 3 h prior to irradiation, and 10 mM hydroxyurea was added 1 h before irradiation. Cells were irradiated with UV-C light (254 nm, 20 J/m²) using a Stratalinker (Stratagene) and labeled with ³H-thymidine (5 Ci/ml; ICN Radiochemicals) for 2 h in the presence of 10 mM hydroxyurea. The cells were then lysed in 400 l reporter lysis buffer (Promega) by freeze-thawing, and transferred to 1.5 ml microfuge tubes. A 50 l aliquot of the lysate was removed for analysis of luciferase activity. The remaining DNA was precipitated by adding cold 100% trichloroacetic acid (TCA) to a final concentration of 5%, and incubating at 4°C overnight. The [³H]thymidine incorporation was measured by spotting precipitated DNA onto glass fiber filters (GF/C; Whatman), washing sequentially with 10% and 5% TCA, and counting in a Beckman scintillation counter. [³H]thymidine incorporation represented cellular DNA repair synthesis. The [³H]thymidine incorporation in irradiated cells was calculated as a percentage of the incorporation in matched, non-irradiated cells. The adjusted [³H]thymidine incorporation in irradiated CREF cells is referred to as 100% repair activity, and the repair activity of other cell types is reported as a percent of the CREF cell repair activity.

For DNA replication assays, confluent 60 mm dishes of cells were labeled with ³H-thymidine in media containing 0.5% serum for 2 h. For this experiment, the cells were not incubated in hydroxyurea, which blocks DNA replication.

Western blot analysis

One million CREF, CREF-Tax or CREF-neo cells were lysed in SDS sample buffer. The lysates were electrophoresed on a 10% SDS polyacrylamide gel. The proteins were electroblotted to a polyvinylidene fluoride membrane (Immobilon-P, Millipore), and probed with anti-Tax or anti-PCNA antibodies. Immunoreactivity was detected with the enhanced chemiluminescence detection kit (ECL, Pharmacia). Bands were quantitated using a Molecular Dynamics densitometer.

Apoptosis studies

For viability studies, cells were treated with either 10 J/m²UV irradiation, 5 g/ml cisplatin, 25 g/ml 5-FU, or 1 mM Taxol (Paclitaxel from Taxus brevifolia, Sigma) for 24 h. Cell viability was assessed by trypan blue staining of both attached and detached cells at sequential time points following treatment with DNA damaging agents. Cisplatin and 5-FU were obtained from Sigma (St Louis, MO, USA).

For Annexin V staining, asynchronized cells were treated with UV irradiation, cisplatin, and 5-FU to induce apoptosis. After treatment, adherent cells were pooled, and stained with fluorescein isothiocyanate conjugated Annexin V (Annexin V-FITC) and PI according to manufacturer's manual (Apo Target, BioSource International, Inc. CA, USA). Cells were analysed by flow cytometry (Epic Profile, Coulter Co, USA). Apoptotic cells were defined as Annexin V positive and PI negative.

For nuclear condensation assays, cells were cultured on glass cover slips in 24-well plates. After treatment with DNA damaging agents, cells were washed three times with PBS and fixed with cold acetone-methanol (1 : 1). Apoptotic cells were detected by staining with 1 g/ml Hoechst 33342 (Sigma) for 5 min and observing by fluorescence microscopy (Olympus).

Statistic calculations

For analyses by two-sample t-test, results were considered statistically significant when P<0.05 (95% confidence interval). Statistic calculations were performed with MINITAB for Windows software (Minitab Inc.).

Acknowledgements

We thank Drs Robert Weiss and Ronald Javier for providing CREF, CREF-Tax and CREF-neo cells. We gratefully acknowledge Dr Betty Slagle and Yahua Chen for critical evaluation of the manuscript. We also thank Caroline Petrin for secretarial support, Michael Pastorello for technical assistance, and Luna Chang for statistical assistance. These studies were supported by Public Health Service grant CA-77371 to SJ Marriott from the National Cancer Institute. FJ Lemoine was supported, in part, by NIH training grant CA09197.

Barry MA, Behnke CA and Eastman A. (1990). Biochem Pharmacol 40, 2353-2362. MEDLINE

Brauweiler A, Garrus JE, Reed J and Nyborg JK. (1997). Virology 231, 135-140. Article MEDLINE

Chen XL, Zachar V, Zdravkovic M, Guo M, Ebbesen P and Liu XD. (1997). J Gen Virol 78, 3277-3285. MEDLINE

Chlichlia K, Busslinger M, Peter ME, Walczak H, Krammer PH, Schirrmacher V and Khazaie K. (1997). Oncogene 14, 2265-2272. MEDLINE

Chlichlia K, Moldenhauer G, Daniel PT, Busslinger M, Gazzolo L, Schirrmacher V and Khazaie K. (1995). Oncogene 10, 269-277. MEDLINE

Copeland KFT, Haaksma AGM, Goudsmit J, Krammer PH and Heeney JL. (1994). AIDS Res Hum Retroviruses 10, 1259-1268. MEDLINE

Franchini G. (1995). Blood 86, 3619-3639. MEDLINE

Franklin AA and Nyborg JK. (1996). J Biomed Sci 2, 17-29.

Fujita K, Yamasaki Y, Sawada H, Izumi Y, Fukuhara S and Uchino H. (1989). Leuk Res 13, 535-543. MEDLINE

Fujita M and Shiku H. (1995). Oncogene 11, 15-20. MEDLINE

Gessain A, Barin F, Vernant JC, Gout O, Maurs L, Calander A and DeThe G. (1985). Lancet 2, 407-409. MEDLINE

Hall AP, Irvine J, Blyth K, Cameron ER, Onions DE and Campbell ME. (1998). J Rheumatol 186, 209-124.

Jeang K-T, Widen SG, Semmes OJ and Wilson SH. (1988). Science 247, 1082-1084.

Kamada N, Sakurai M, Miyamoto K, Sancar A, Sadamori N, Fukuhara S, Abe S, Shiraishi Y, Abe T, Kaneko Y and Shimoyama M. (1992). Cancer Res 52, 1482-1493.

Kao S-Y and Marriott SJ. (1999). J Virol 73, 4299-4304. MEDLINE

Kelland LR. (1993). Crit Rev Oncol Hematol 15, 191-219. MEDLINE

Kerr JF, Willie AH and Currie AR. (1972). Br J Cancer 26, 239-257. MEDLINE

Kitajima I, Nakajimam T, Imamura T, Takasaki I, Kawahara K, Okano T, Tokioka T, Soejima Y, Abeyama K and Maruyama I. (1996). J Bone Miner Res 11, 200-210. MEDLINE

Leno M, Simpson RM, Bowers FS and Kindt TJ. (1995). J Exp Med 81, 1575-1580.

Los M, Khazaie K, Schulze-Osthoff K, Baeuerle PA, Schirrmacher V and Chlichlia K. (1998). J Immunol 161, 3050-3055. MEDLINE

Lydy SL, Conner ME and Marriott SJ. (1998). Virology 250, 60-66. MEDLINE

Majone F, Semmes OJ and Jeang K-T. (1993). Virology 193, 456-459. MEDLINE

Martin SJ, Reutelingsperger CP, McGadon AJ, Rader JA, van Schie RC, LeFace DM and Green DR. (1995). J Exp Med 182, 1545-1556. MEDLINE

Matsumoto K, Shibata H, Fujisawa J-I, Inoue H, Hakura A, Tsukahara T and Fujii M. (1997). J Virol 71, 4445-4451. MEDLINE

Miyake H, Suzuki T, Hirai H and Yoshida M. (1999). Virology 253, 155-161. MEDLINE

Mulloy JC, Kislyakova T, Cereseto A, Casareto L, LoMonico A, Fullen J, Lorenzi MV, Cara A, Nicot C, Giam CZ and Franchini G. (1998). J Virol 72, 8852-8860. MEDLINE

Philpott SM and Buehring GC. (1999). J Natl Cancer Inst 91, 933-942. MEDLINE

Poiesz BJ, Ruscetti FW, Gadzar AF, Bunn PA, Minna JD and Gallo RC. (1980). J Virol 64, 5682-5687.

Prosperi E, Scovassi AI, Stivala LA and Bianchi L. (1994). Exp Cell Res 215, 257-262. MEDLINE

Raynal P and Pollard HB. (1994). Biochim Biophys Acta 1197, 63-93. MEDLINE

Ressler S, Connor LM and Marriott SJ. (1996). FEMS Microbiol Lett 140, 99-109. MEDLINE

Ressler S, Morris GF and Marriott SJ. (1997). J Virol 71, 1181-1190. MEDLINE

Saggioro D, Majone F, Forino M, Turchetto L, Leszl A and Chieco-Bianchi L. (1994). Leuk Lymphoma 12, 281. MEDLINE

Sandberg AA. (1991). In The Chromosomes in Human Cancer and Leukemia 2nd edition. Elsevier Amsterdam,

Schiff PB and Horwitz SB. (1980). Proc Natl Acad Sci USA 77, 1561-1565. MEDLINE

Seiki M, Eddy R, Shows TB and Yoshida M. (1984). Nature 309, 640-642. MEDLINE

Smith MR and Greene WC. (1991). J Clin Invest 88, 1038-1042. MEDLINE

Tanaka A, Takahashi G, Yamaoka S, Nosaka T, Maki M and Hatanaka M. (1990). Proc Natl Acad Sci USA 87, 1071-1075. MEDLINE

Thompson CB. (1995). Science 267, 1456-1462. MEDLINE

Tsukahara T, Kannagi M, Ohashi T, Kato H, Arai M, Nunez G, Iwanaga Y, Yamamoto N, Ohtani K, Nakamura M and Fujii M. (1999). J Virol 73, 7981-7987. MEDLINE

Yamada T, Yamaoka S, Goto T, Nakai M, Tsujimoto Y and Hatanaka M. (1994). J Virol 68, 3374-3379. MEDLINE

Yamaoka S, Inoue H, Sakurai M, Sugiyama T, Hazama M, Yamada T and Hatanaka M. (1996). EMBO J 15, 873-887. MEDLINE

Yoshida M, Seiki M, Yamaguchi K and Takatsuki K. (1984). Proc Natl Acad Sci USA 81, 2534-2537. MEDLINE

Figure 1 Western blot analysis of Tax and PCNA expression. Extracts from CREF, CREF-neo, and CREF-Tax cells (lanes 1-3, respectively) were harvested and equivalent amounts of protein were analysed for (a) Tax expression using a rabbit anti-Tax polyclonal antibody, and (b) PCNA expression using an anti-PCNA monoclonal antibody (PC10)

Figure 2 Inhibition of cellular DNA repair by Tax. (a) HCR assay. The pMSV-Luc reporter plasmid was UV irradiated and transfected into CREF, CREF-neo or CREF-Tax cells. Duplicate dishes received non-irradiated plasmid. Luciferase activity from the plate transfected with non-irradiated plasmid was set to 100% and repair activity was calculated as a percentage of activity observed with non-irradiated plasmid. (b) UDS assay. CREF, CREF-neo, and CREF-Tax cells were plated in 60 mm dishes and UV irradiated for UDS assays as described in Materials and methods. Repair activity of CREF cells was set to 100%, and repair activity of CREF-neo and CREF-Tax cells was calculated as a percentage of this value. The data displayed for each repair assay is an average of three independent experiments. P<0.05 is statistically significant

Figure 3 Effect of Tax on DNA replication following DNA damage. CREF, CREF-neo, and CREF-Tax cells were UV irradiated (right) or not irradiated (left), and labeled with ³H-thymidine to determine DNA replication activity. The data displayed for non-irradiated cells was a single experiment and irradiated cells is a representative experiment from three repetitions

Figure 4 Effect of Tax on cell death following UV irradiation. (a) Viability of CREF-neo and CREF-Tax cells following treatment with different doses of UV irradiation. Viability was measured by Trypan blue exclusion staining. Cell survival was calculated by dividing the number of live cells by the total number of cells. Error bars were calculated from three independent experiments. (b) Time course of CREF-neo and CREF-Tax cell survival upon UV irradiation. At different times after UV treatment (10 J/m²), cell survival was calculated as described in (a) from three independent experiments

Figure 5 Sensitivity of CREF-Tax cells to DNA damaging agents. CREF-neo and CREF-Tax cells were incubated with either (a) 25 g/ml of 5-FU or (b) 5 g/ml of cisplatin. Survival of CREF-neo and CREF-Tax cells was measured at the indicated time points after treatment. Cell survival was calculated as described in Figure 4. This is a representative result from two independent experiments

Figure 6 Nuclear condensation in CREF-Tax cells following DNA damage. (a) CREF-neo (left panel) and CREF-Tax (right panel) cells were either incubated without treatment (no damage, panels a and b), or were treated with UV (panels c and d), 5-FU (panels e and f), or cisplatin (panels g and h). Cells were fixed and nuclei were detected by Hoechst stain and fluorescence microscopy. Arrows indicate the location of representative condensed nuclei. (b) The percentage of condensed nuclei in CREF-neo and CREF-Tax cells following treatment with DNA damaging agents

Figure 7 Effect of Tax on apoptosis following DNA damage. CREF-neo (top) and CREF-Tax (bottom) cells were treated with 10 J/m² UV, 5 g/ml cisplatin, 25 g/ml 5-FU, 1 mM Taxol, or were not treated (mock). After 24 h of treatment, cells were harvested and stained with Annexin V and PI. A histogram showing Annexin V staining of PI negative cells is shown. (-apoptotic cells). The percentage of apoptotic cells is shown at the upper right corner. This data is from a single experiment that is representative of two repetitions

Figure 8 Effect of Tax on apoptosis following DNA damage in Jurkat cells. Jurkat cells were transfected with 10 g of pCMV-Tax or mock transfected. Cells were then treated with 5 g/ml cisplatin, 25 g/ml 5-FU or mock-treated. Annexin V, PI staining and FACS analysis were performed 24 h after treatment. This data is from a single experiment that is representative of two repetitions. The percentage of apoptotic cells was shown at the upper right corner. ---: apoptotic cells