Overexpression of DNA Polymerase @Jsensitizes Mammalian Cells to 2',3'

ICANCERRESEARCH57.I10-I 16,JanuaryI. I997J Overexpression of DNA Polymerase @JSensitizesMammalian Cells to 2',3'-Deoxycytidine and 3'-Azido-3' .deoxythymidine'

Khalil Bouayadi, Jean-SÃ©bastien Hoffmann, Pascal Fons, Michele Tiraby, Jean-Paul Reynes, and Christophe Cazaux2

Laboratoire de Microbiologic et de GÃ©nÃ©tique,UniversitÃ©Paul Sabatier, 118 route de Narbonne, 31062 Toulouse cÃ©dex(K. B., P. F., C. C.]; Institut de Pharmacologic et Biologic Structurale. Centre National de Ia Recherche ScienujIque, UPR 9062, 205 route de Narbonne, 31077 Toulouse cÃ©dexfl. S. H.!; and Laboratoire CAYLA, ZI. Montaudran, 5 rue J. Rodier, 3/400 Toulouse (M. T., J. P. R.J

ABSTRACT nator AZT-MP was also observed (10, 11). In vivo, AZT-MP is incorporated into cellular DNA (12), and it has been suggested that Mammalian DNA polymerase @3is a DNA repair enzyme expressed polymerase 13may play a role in this process (1 1). @ constitutively at a low level. In vitro, purified DNA polymerase (Pol) incorporates the nucleotide analogues 2'-3' deoxycytidine (ddC)-triphos AZT and ddC are antiviral agents currently used in the treatment phate and 3'-azido-3'-deoxythymidine (AZT)-triphosphate Into DNA, of AIDS. These chemotherapeutic drugs target HIV reverse tran causing chain termination. We have tested the possibility ofenhancing the scriptase which incorporates them into HIV DNA, terminating cytotoxicity of these chain terminators against mammalian cells by in DNA polymerization (13). In this study, we hypothesized that creasing the level of Pol (I. Chinese hamster ovary AA8 and murine overexpression of Pol f3, a cellular target of ddC and AZT, might melanoma B16 cell lines were stably transfected with rat pol fi cDNA render these analogues effective also against tumor cells. To ex underthe controlof a viral enhancer/promoter.Wefoundthat overex plore this possibility, we stably transfected CHO AA8 and murine pression of Pol @3sensitized the cells to ddC and AZT. To confirm the role of this polymerasein this process,we preparedcell extractsfrom the melanoma B 16 cells with rat polf3 cDNA. We found an increased control and Pol @3overexpressingChinese hamster ovary cell lines and sensitivity to ddC and AZT of the transfected cells, and we tested in vitro their capacity to incorporate ddC-triphosphate and AZT confirmed the involvement of Pol (3 in this process using an in vitro triphosphate into DNA. We found that inhibition of DNA replication by replication assay with cell extracts. To our knowledge, this is the both chain terminators was more pronounced when extracts from pol first report of a kill or â€œsuicideâ€•associationinvolving a polymerase @3-transfectedcellswere used, providing a direct evidence of the involve activity. ment of Pol @3inthe sensitizationprocess.In addition,we showedthat The conversion of nucleoside analogues to nucleotide triphosphates cotransfection with bacterial or viral thymidine/thymidylate kinase genes enhanced the Pol fl-mediated cytotoxicity of AZT, suggesting that phos. is necessary before their incorporation into DNA by DNA poly phorylationandpolymerizationactivitiesmightbecombinedtopotentiate merases. Analogues readily enter cells but are often poor substrates their respective effects. These observations may be useful for improving for the sequential conversion to the phosphorylated forms by cellular therapeutic efficiency of DNA chain terminators. TK and TMK which specifically convert thymidine (14). For instance, the concentration of AZT-MP is more than 50-fold higher than that of AZT-TP in human cell lines, suggesting that it is a very poor substrate INTRODUCTION for human thymidylate kinase (13). Here, we present data showing Pol j3@,a single polypeptide of 39 kDa, is one of the five known that cells coexpressing both Pol f3 and bacterial or herpes thymidine/ mammalian DNA polymerases and is highly conserved among higher thymidylate kinases are more sensitive to AZT than cells expressing eukaryotes (I). It is believed to function primarily in the repair of only kinases or polymerase. These data may have particular impor damaged DNA (2) but may also have a role in DNA replication (3, 4) tance in the perspective of new therapeutic combinations for cancer and in DNA damage processing (5). Pol /3 is expressed at a constant gene therapy. level throughout the cell cycle (6) and exposure of cells to certain DNA-damaging agents induces Pol @3expression (7, 8). Features that distinguish Pol @3fromother cellular polymerases are its small size, MATERIALS AND METHODS the lack of associated exonuclease and endonuclease activities, and its high infidelity in replicating DNA (9). Materials. All oligonucleotides were synthesized on a Cyclone Plus In vitro, it has been demonstrated that purified Pol (3 efficiently DNA synthesizer from MilliGenlBiosearch and purified on 20% polyacryl amide gel. For DNA synthesis assays, the 60-mer templates were hybrid incorporates the deoxynucleotide analogue ddCMP, an inhibitor of ized to the 5' 32P-labeled 17-mer primer. Pol 13cDNA harboring plasmid DNA synthesis, with efficiency comparable to that of dCMP when an pRSET was kindly furnished by Dr. Wilson (Galveston, TX). Highly oligonucleotide template is used (10). Under similar conditions, but purified calf thymus DNA polymerase a-primase, @3,8, and a were gener using a higher amount of enzyme, incorporation of the chain termi ous gifts from Dr. HUbscher (Zurich, Switzerland) and polyclonal antibod ies against Pol @3were provided by Dr. Sweasy (Yale University, New Received 7/I 5196;accepted 11/1/96. Haven, CT). DNA plasmids pUT526& pUT599 (VECT 4991), pZEOSGO, The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with and pZEOSG1 were obtained from Cayla (Toulouse, France). Analogues 18 U.S.C. Section 1734 solely Io indicate this fact. were obtained from Weilcome (AZT), Moravek Biochemicals, Inc. (AZT @ This research was supported by the Region Midi-PyrÃ©nÃ©esandthe Ligue Nationale TP), Sigma (ddC), and Pharmacia (ddCTP). Restriction and DNA modifi Contre Ic Cancer (axe thdrapie gÃ©nique). cation enzymes were obtained from New England Biolabs and Boehringer 2 To whom requests for reprints should be addressed, at Institut de Pharmacologic et de Biologic Structurale, Centre National de Ia Recherche Scientifique UPR A9062, 205 Mannheim. route de Narbonne, 31077 Toulouse cÃ©dex,France.Fax: 33-5-61-17-59-94; E-mail: Bacteria and Mammalian Cell Culture. Bacteria were grown in LB broth [email protected]. complemented when required with zeocin (Cayla) at 20 p.g/ml. CHO AA8 and 3 The abbreviations used are: Pol (3, DNA polymerase @;HSV tic, herpes simplex virus type-I thymidine kinase gene; TK, thymidine kinase; TMK, thymidylate kinase; Sh, ble Bl6 BL/6 parental cells were grown in L-glutamine containing a-MEM and Sh encoded protein conferring resistance to zeocin; ddC, 2'-3' deoxycytidine; AZT, RPMI 1640 (BioWhittaker), respectively, complemented with 10% FCS (Bio 3'-azido-3'-deoxythymidine; AZT-MP, AZT monophosphate; AZT-TP, AZT triphos Whittaker), penicillin G/streptomycin sulfate (BioWhittaker), and amphoteri phate; ddCMP, ddC monophosphate; GCV. ganciclovir [(2-amino-1,9-(2-hydroxy-l-hy droxymethyl)ethoxy)methyl)-6H-purine-6-one]; CHO, Chinese hamster ovary cell; HSV, cm B (Sigma). Media for growth of stably transfected CHO and Bl6 cells were @ herpes simplex virus: HSV-l, herpes simplex virus type I ddCTP, ddC triphosphate. supplemented by 100 gxg/mland 20 g@g/m1zeocin(Cayla), respectively. 110

Construction of Expression Plasmids. Pol @3-overexpressing plasmid mm and were disrupted in a Dounce homogenizer. Nuclei were harvested by pUTpolIJ was constructed from a pUT687 plasmid4 carrying the thymidine centrifugation for 10 mm at 3300 X g at 0Â°C.The nuclei pellet was suspended kinase gene of Escherichia coli fused in frame with the bacterial Sh ble gene in hypotonic buffer containing 350 mtvi NaCI. After 30 mm of extraction at conferring resistance to the broad-spectral zeocin xenobiotic of the phleomycin 0Â°C,thenuclear extracts were centrifuged at 15,000 X g for 20 mm. Proteins family (15). The pol @3genegenerated primers TA'VFCCATGOCACTCGTG from cytosol and nuclear extracts were precipitated by addition of ammonium GAACFCGCAAAC1TF and UAAGCTAGCTCACTCCTGTCmGGGC sulfate, resuspended in dialysis buffer [50 mt@iTris-HC1 (pH 7.5), 1 mM DTT, TC. Upon restriction enzyme digestion, a NcoI-NheI fragment containing pol 100 mt@imono-K glutamic acid, and 10% glycerol], and dialyzed for 2 h at 0Â°C. @3wasobtainedandthenrecombinedwiththeNcoI-AvrIIfragmentofpUT687 Extracts were frozen in liquid nitrogen and stored at â€”70Â°C.Theywere used to give pUTpolf3(Fig. IA). for both Western blotting and in vitro DNA replication assays. The E. coli tmk gene was fused with E. coli tk and Sh ble by using a 4-kb Western Blot Analysis. Sixty @igof cell extracts were electrophoresed in PCR-generated fragment obtained as already described (16) by amplification a 12% SDS-polyacrylamide gel. Proteins in the gel were transferred electro of the acpP-hoIB intergenic region of the Kohara phage AE9G1(17). First, this phoretically to nitrocellulose membrane. Pol (3 was revealed by incubating the tmk-containing PCR product digested with NcoI and MluI (treated with Kle membrane with purified rabbit anti-PoI (3 polyclonal antibody followed by now enzyme) was inserted into pUT633 (VECT 6331; Cayla) opened by NcoI incubation with antibody to rabbit IgG conjugated to horseradish peroxidase. and PvuII to give the E. coli tmk::Sh ble harboring pUT832 (Fig. 1B).PUT633 Immobilized horseradish peroxidase activity was detected using enhanced is similar to pUT687 but contains a PvuII site into the genetic link of a chemiluminescence (Amersham Corp.). luciferase::Sh-encoding fusion. Second, the MscI-RsrII fragment from the 4-kb in Vitro Replication by Cell-free Extracts. Standard15-pi reactionmix PCR-generated fragment containing the E. coli tmk gene was inserted in place tures contained 5 ng of labeled primed oligonucleotide substrates and 3 p@gof of the Sh ble gene-containing MscI-RsrII fragment of pUT687 to give pUT833 Sh or Pol f3::Shcell extract proteins in reaction buffer J45 mMHEPES-KOH which harbors the E. coli tk::tmk fusion (Fig. 1B). Finally the EcoRV-NotI (pH 7.8), 7 mM MgCl2, 1 mM Dli', 0.4 mM EDTA, 34% glycerol, 50 mM fragment from pUT832 was inserted into pUT833 cut with same enzymes to mono-K glutamic acid, and 18 @xgofBSAI. The concentrations of the deoxyri give pUT834 (Fig. 1B). bonucleotides and the amount of ddCTP and AZT-TP used in the reactions are The pZETkTmk@plasmid carrying polfi and E. coli tk/tmk genes was the indicated in the legends of Figs. 1â€”6.At the end of the reaction, 5 pi of result of intermolecular recombination between Bbrpl-SmaI-digested frag stopping buffer (90% formamide/0.l% xylene cyanol/0.l% bromphenol blue/ ments from pUT834 and pZHTk(3DNA (Fig. 1C)carrying pol @3andthe HSV 0.1 mMEDTA) were added. Samples were denatured for 10 mm at 70Â°Cand tk::ble Sh fusion. the pZHTk/3 plasmid was constructed by PCR amplifying a loaded on sequencing gels. EM7-pol 13 fragment from pUTpolfJ with primers TFCTCACGTGACCG Quantitative Determination of Protein Overexpression and Replication GCGCCTAGT and GGGAGCCCAAGGACAGGAGTGAATGAUCGAA Products Generated by Pol @l.The autoradiograms were scanned using an C1TF to create a Bbrpl-BstBI fragment that was inserted into a pZEOSGO Omnimedia scanner 6cx (Bioimage) and analyzed by a Whole Band Analyser plasmid that had been opened with a SpeI-BstBI double cut and then treated version 3.2.2, and the bands were quantified with the Bioimage application with Klenow polymerase at the SpeI extremity (Fig. 10. PZEOSGO is derived (Roissy-France). from pZEO SV1 (VECT 2001; Cayla). It harbors the HSV tk gene and a sequence from the human T-cell lymphotrophic virus ( I8) which encodes a RESULTS transcriptional activator (Fig. 1C). The SC18-l2 temperature-sensitive mutant of B/r E. coil polA (from E. Overexpression of Pol @3inPol @3::ShCHOCells. We have Witkin, Piscataway, NJ) was transformed with pUTpol@ and pZETkTmkj3 to transfected CHO and melanoma Bl6 cells with the DNA expression verify the functionality of the polymerase activity. The TK function encoded vector pUTpolf3. The plasmid harbors the cDNA encoding the rat Pol by pZETkTmk@3DNAwas assessed by using a i/C E. coli mutant,4 TMK @3fusedin frame with the ble S/i gene conferring resistance to zeocin activity being revealed as described previously (16, 19) by complementing a (Fig. 1A). This fusion was driven by a strong, constitutive promoter @,ik temperature-sensitive E. coli mutant named TD205 (gift from J. A. Fuchs, unit, which is the viral HSV TK promoter coupled to the viral St. Paul, MN). DNA Transfection.Parentalcellswere transfectedwithplasmidDNA polyoma pYF44l enhancer. To validate the construct, we showed that using the DMSO/polybrene shock procedure (20). After 48â€”62 h of pheno thermosensitive polAl2 E. coli bacteria mutated in DNA polymerase typic expression, cells were split into the appropriate selective medium. One I activity were able to grow at 42Â°Cwhen transformed by pUTpol@ hundred @.tWmland 20 @xWmlzeocin were added, respectively, to CHO and (data not shown). To assess the expression of Pol @3protein in BI6 cells. When they appeared, clones were picked, transferred to 35-mm mammalian cells, extracts were prepared from cells transfected with dishes, and incubated for 5 days in selective medium before testing for their pUTpol@3 or pUT 526@, the latter carrying the ble Sh gene alone. sensitivity. Extracts were electrophoresed and a Western blotting was carried out In Vitro Sensitivity Assay. Toxicity of AZT and ddC againstparentaland using polyclonal antibodies against Pol (3. As seen in Fig. 2, an transfected tumor cells was assessed by inoculating 24-well plates with 3000 immunopositive band corresponding to Pol (.3(39 kDa) was observed cells/ml of serum-supplemented medium containing different concentrations of when 60 @gofproteins of Pol f3::Sh cell extracts were loaded onto a nucleoside analogues. Cytotoxic effects were estimated after 5 days of incu polyacrylamide gel. We could not see any Pol 13-related immunore bation by slightly modifying a survival assay described previously (21). Cells were washed three times with PBS buffer (BioWhittaker), and plated cells active band after electrophoresis of up to 240 @gofproteins from the were harvested by trypsinization, counted, and plated in serial dilutions. Ten control Sh cell extracts (data not shown). By scanning the autoradio days later, colonies in formation were counted by fixing cells with ice-cold graph of the Western blot, the bands corresponding to Pol f3 in the Sh methanol, washing them twice with PBS, and then staining surviving cells with extract (240 ,.Lg)and Pol @::Shextract (60 pg), we determined a Pol filtered 10% Giemsa for 15 mm at room temperature. Survival was expressed 13overexpressionratioof9.8.In addition,nobandat60kDacorre as a percentage relative to untreated cells. sponding to the Pol @::Shprotein was detected by using Pol @3 Preparation of Cell Extracts. CHO cells were grown to subconfluence as antibodies, suggesting that the fusion protein was cleaved either in monolayer cultures on 145-mm plates at 37Â°Cin a 5% CO2 humidified vivo or during protein extraction. Indeed, Western blotting with anti incubator. Cells were washed with ice-cold PBS, scraped off the plates, and bodies against Sh protein showed the same amount of an immunore harvested by centrifugation. The cell pellet was suspended in hypotonic buffer active band around 20 kDa corresponding to Sh protein in both Pol [10 mi@tTris-HCI(pH 7.5), 10 mM KC1, 10 mist MgCl2, and I nmi Dill @::Shand Sh CHO extracts (data not shown). containing protease inhibitors. The cells were allowed to swell on ice for 10 Pol @3OverexpressionEnhances Cytotoxicity to ddC and AZT. Fig. 3 shows the effect of overexpression of Pol f3 on CHO cell 4 C. Cazaux, M. Tiraby, L. Loubieres, L. Haren, D. Klatzmann, and G. Tiraby. Differential activities toward therapeutic nucleosides of a- and y-herpes virus thymidine survival when exposed to the DNA chain terminators ddC and AZT, kinases expressed in Escherichia coli, submitted for publication. respectively. Survival of cells containing the control plasmid Ill

@ poll3 pUT pol@3

Nhe I

Fig. I. Construction of DNA plasmid expression vectors. A, cDNA of NcoI Po.II pol 13was amplified by PCR using restriction site-generating primers poly A and subcloned into the pUT687 plasmid in place of the E. coli tk gene \ I @1 B @ to overexpress only Pol f3.pUT68l harbors E. coli ik- and ble S/i-coding pUT633 I Apr E@h@k@luc J@zeorlfOnI genes separated by a flexible polypeptide link-encoding sequence; this hybrid gene is driven by both a synthetic strong EM7 bacterial promoter / and the viral HSV TK promoter coupled to the viral polyoma pYF441 1 E. co/i enhancer (28). B, construction of plasmid pUT834 harboring E. co/i tk, pUT 832 trnk I tmk, and ble Sh genes fused in frame. C, for coexpressing HSV TK and Pol f3, we used plasmid pZeoSGO, which is an eukaryotic expression NcoI MscI RsrII vector carrying the ble S/i resistance and the HSV tk genes. respectively. poly A under control of enhancer/promoter sequences from the immediate early @ gene of human cytomegalovirus and the SV4Oearly region. cDNA of pol pUT 687 I Apr E@[email protected] Zeo@'( @I p [email protected] expressing Pol 13and E. coli TKTI'MKwas constructed by recombining pUT835andpZHTh/3(seetextfordetails).AllPCR-amplifiedregions were sequenced to ensure that no errors were generated by the PCR pUT833 1E.co/ionk process. MCS is a multicloning site: OR! is the replication origin of the @ E. coli colEl plasmid; ZEO'@ and Apâ€• the zeocin and ampicillin @ resistance genes; PoIvA is the late polyadenylation signal of SV4O;TX,,, E. co/i E. co/i SV4O,,. and C/PIts' @yenhancer-promoter sequences from herpes tk gene pUT834 I tk tink @or1 and SV4Oand cytomegalovirus (CMV)viruses; and HTLVis a transac tivating sequence. Bb,pI

@@@ C polo

@@ pZHTk@

13si81

EM7

pUT526& which possesses the pUTpolf3 backbone but lacks the pol 13gene,wasslightlyaffectedbybothdrugs,AZT beingmoretoxic (Fig. 3B). In contrast, pUTpol@3-transfected cells were 30â€”SOtimes more sensitive to ddC and AZT than the cell line expressing the zeocin resistance function alone. Murine melanoma B 16 cells were also transfected by plasmids pUTS2fiz@and pUTpol/3 and analyzed as described with CHO cells. We found that survival was very similar to (Kd)â€•@@ that obtained with CHO (data not shown). Namely, IC50 and lethal dose of nucleoside analogues (Table I) were comparable to those determined for CHO (Fig. 3), AZT being more toxic than ddC. These 43 results suggest that the data presented here by using the well-known CHO model could be extrapolatedto tumorcell lines. In Vitro Inhibition of DNA Synthesis by Pol @::ShCell Extracts in the Presence of ddCTP. To comparethe effect of ddCTPon DNA synthesis catalyzed by Sh and Pol j3::Sh DNA replication machineries, we prepared cell extracts from both cell lines and investigated their 30 capacity to extend a 5' 32P-labeled phosphorylated 17-mer oligonu cleotide primer on a G-rich 60-mer template (Fig. 4A). Reactions were Fig. 2. Western blot analysis showing Pol (3protein level in Sh and Pol fl::Sh CHO performed in the presence of ddCTP at a ratio of ddCTP:dCTP equal cells. Samples of cell extracts (60 @gofprotein) or I @sgofpurified rat Pol fi were subjected to a 12% SDS-polyacrylamide gel, transferred electrophoretically to nitrocel to 1. After the substrate was replicated, the newly synthesized DNA lulose membrane, and then probed with purified rabbit anti-Pol @3polyclonalantibody (see products were resolved by a 15% polyacrylamide gel and visualized text for details). Left, size markers. 112

Co Co

1@ Ia CID Cl)

1 i04

ddC(@ M) AZT (11 M) Fig. 3. Toxicity of ddC (A) and AZ'f (B) to CHO cells transfected with DNA plasmid pUT526i@-Sh(0) or pUTpol/3(â€¢).Cytotoxicitywas determined from three isolated colonies. Data given are the means of these three experiments and error is less than 20%. A pool of more than five other independent clones was tested and showed comparable cytotoxicity results. with autoradiography (Fig. 4B). With Pol @::Shextracts, most of the with 40 p.MdATP, 40 @MdGTP,and 50 @tMor100 /LMAZT-TP on reaction products (61%) migrated as a population of 18-mer, 19-mer, an A-rich 60-mer DNA template (Fig. 5A). After the copying of this and 20-mer DNA fragments, showing that DNA synthesis stopped at template for 30 mm at 37Â°C,the reaction products were analyzed on the first dG base encountered by the cellular enzymes after ddCTP a 20% polyacrylamide gel (Fig. SB). DNA chains migrating as 20- incorporation. In contrast to Sh cell extracts, the majority (77%) of the mer, arrested opposite to the first dA encountered, were detected for reaction products migrated after this first dG encountered from the both Sh and Pol @::Shcellextracts. By scanning autoradiographs, we primer, indicating that no or few ddCTPs were incorporated during evaluated that the band corresponding to this arrest was 2.3 times elongation. Therefore, overexpression of Pol (3 in the cell extracts more intense for Pol f3-overproducing cell extracts. These observa inhibited in vitro DNA replication by efficient incorporation of tions suggest that, in the absence of dTFP, AZT was preferentially ddCTP. We then performed primer extension experiments by adding incorporated at this site by extracts containing a high level of rat Pol to the Sh cell extracts an excess of highly purified calf thymus DNA @3and that cell sensitivity to AZT (Fig. 3B) was due to its incorpo polymerases a, f3, @,orâ‚¬(Fig. 4B). Severe inhibition of replication ration into DNA by this DNA polymerase. was observed only when Pol @3waspresent, confirming its ability to Enhancement ofthe Cytotoxic Effect of Pol fi by Cotransfection incorporate ddCTP with a high efficiency in the presence of cellular proteins. Taken together, these results strongly suggest that sensitiza with Nucleoside/Nucleotide Kinases. We investigated the effect of tion of the Pol (3::Sh cells to ddC treatment (Fig. 3A) was due to an increasing intracellular pools of phosphorylated analogues on the enhanced incorporation of the analogue into cellular DNA by DNA sensitivity of cells overexpressing Pol @3.CHOcells were transfected Pol @3. with pZETkTmkfJ or pZHTk@ vectors coexpressing Pol @3andbac Capacity of Pol @::Sh Extracts to Incorporate AZT-TP into terial or viral TKITMK, respectively, which sequentially promote DNA in Vitro. To study the effect of AZT-TP on DNA synthesis monophosphorylation and diphosphorylation of the analogues. The catalyzed by Sh and Pol f3::Sh cell extracts, reactions were conducted expression of each activity was assessed by complementing E. coli bacteria deficient in DNA polymerase I (mutant SC18l2), TMK (mutant TD205) or TK.4 Cotransfection of CHO cells by pol @3and E. Table 1 Cytotoxic effect of ddC and AZT on murine melanoma B16 cells coli tkltmk genes led to an increasing sensitization to AZT compared IC50a () Minimal cytotoxic dose'@ (SM) to cells transfected with each gene alone (Fig. 6A). Similar experi DNA plasmid ddC AZT ddC AZT ments carried out by using plasmid pZHTk@3coexpressing Pol @3and pUTS29@ 900 500 >i04 >i0'@ HSV TKITMK instead of bacterial TK-TMK fusion led also to an pUTpolf3 80 70 600 200 additive effect, even to a lesser extent (Fig. 6B). ddCMP did not give aInhibitoryconcentrationofanaloguerequiredtoreducecellsurvivedby50%. a comparable additive effect, probably because HSV TK and E. coli b Analogue concentration for which no viable cell was detectable with the Giemsa coloration. TK and/or TMK do not efficiently interact with this drug. 113

Downloaded from cancerres.aacrjournals.org on October 5, 2021. © 1997 American Association for Cancer Research. CYTOTOXIC ACTION OF DNA POLYMERASE /3 A DNAsynthesis 17-mer @ 3'-CATAcGAGAAccAAcAT.s' 5'[email protected]' 60-mer

Â°., Â°*, @, c@ Ta â€¢: dQ@'@ S @ D .@( ,c. t â€˜

a'. a@

@ 17

Fig. 4. Effect of ddCTP on in vitro replication by Sh and Pol @::ShCHO cell extracts. A, the G-rich 60-mer substrate annealed to the 17-mer primer used for the primer extension assays. B, 5' 32P-labeled primed 60-mer template was replicated at 37Â°Cfor 1 h by 3 @gofextract from the indicated cell lines in the presence of ddCTP at a ratio of ddC:dC equal to I. Two units of each purified DNA polymerase were added to the reaction mixture when indicated. Arrows, position of the 17-mer (primer) and 60-mer (full-size product). To determine the exact location of the arrest sites, the 60-mer substrate was sequenced using the dideoxy method and migrated on the gel next to the reaction products.

DNA synthesis 17-mer .4â€” 3'-CATACGAGAACCAACAT-5' 5'-AATAATAATA@CACCAACAATATAGGT1@CAACGCTAACCATCGTATGCI@CTTGGTTGTA-3' 60-mer

AZT (jiM) 50 100 III

@ ;l) III 17 F. I Fig. 5. Effect of AZT-TP on in vitro replication by Sh and Pol @::ShCHOcell extracts. A, the A-rich 60-mer substrate annealed to the l7-mer primer used for the primer extension assays. B, 5' 32-labeledprimed 60-mer template was replicated at 37Â°Cfor30 mm by 3 @gofcell extract in the presence of 40 p@dATP. 40 j.@MdGTP,and 50 /LMor 100,SMAZT-TP. The reactions were conducted without dTFP. The printed sequence on the left corresponds to the template. 114

Co Co

Ia Ia Co Cl) Cl)

io@ io@

AZT ( R M) AZT ( l@M)

Fig. 6. A, toxicity of AZT to CHO cells transfected with DNA plasmid pUTpol@ (â€¢),pZEOSG 1 similar to pZEOSGO but expressing both E. co/i TK and TMK (U), and pZETkTmkf3 coexpressing E. co/i TK!FMK and Pol (3(A). B, toxicity of AZT to CHO cells transfected with DNA plasmid pUTpolf3(â€¢),pUTS99expressing HSVTK (U), and pZHTk@3coexpressing HSV TK and Pol @3(A).Cytotoxicity was determined from three isolated colonies and a pool of more than five clones. Data given are the means of these four experiments; coors are less than 20%.

DISCUSSION metabolization of the drugs by the cellular kinases; (c) the effect of DNA polymerase a, which is also able to incorporate AZT-MP but The antiviral action of chain terminators comes from the capacity of not ddCMP into DNA (10); and (d) higher excision efficiency of ddC viral polymerases or reverse transcriptases to incorporate these nude from DNA by the exonuclease proofreading activities, AZT-mono oside analogues into viral DNA, causing termination of their replica phosphate termini being 100- and 5-fold more resistant to exonucleo tion (13). The most useful drugs for the treatment of AIDS are lytic degradation by the 3' to 5' exonuclease of DNA polymerase â‚¬ nucleoside analogues, such as ddC and AZT, which target the viral and t5,respectively, than normal base-paired 3' termini.5 reverse transcriptase. The poor affinity of these drugs to the cellular To our knowledge, we present here the first reported kill association replicative DNA polymerases may explain why this category of involving a DNA polymerase activity. Thus far, previous suicide agents has much less effect on cell survival. Because of its high associations using base analogues have involved TK activities be infidelity in vitro (9) and its ability to catalyze highly error-prone cause of their efficiency in metabolizing nucleotide analogues into bypass replication of bulky adducts in vitro (5), Pol @3isprobably their active forms (22). The HSV-1 TK has been used in different cell mutagenic in vivo. This may explain the constitutive low level of this types, including lymphoma, hepatocellular carcinoma, glioma fibro enzyme. However, the temporary induction of its expression after sarcoma, and melanoma (23â€”26),to render targeted cells susceptible exposure to certain DNA-damaging agents (7, 8) might minimize this to normally nontoxic chemotherapeutic agents. The viral TK is able to mutagenicity. It has been shown that Pol @3,which is believed to toxify into the replicating cancer cells systemically administered function primarily in DNA repair (2), is capable of incorporating antiviral drugs such as GCV, a guanosine analogue normally metab chain terminators ddC and AZT into DNA (10), but its concentration olized at very low levels by mammalian enzymes. The association of in mammalian cells may be too low for efficient incorporation. HSV-1 TK/GCV is presently undergoing clinical trials. The HSV-1 To investigate a potentialcytotoxic association of Pol (3and chain TK/AZT combination was also used as part of a strategy of using terminators, we analyzed the effect of overexpression of rat Pol (3 in suicide genes for eradicating TK-expressing CD4@ cells in the design CHOandmurinemelanomaB16 cell lines treatedwith ddCandAZT. of gene therapy of HIV infection (27). We found here that overex We showed thatthese agents were about50 times more toxic toward pression of Pol f3improved the sensitization effect of exogenous HSV Pol f3overexpressing cells than control cells (Fig. 3). Using an in vitro TKI1'MK after treatment with AZT (Fig. 6B). It is worth noting that cell extract-based DNA replication assay, we demonstrated a direct the toxic effect of Pol @3ismore important than that of HSV TK when role of the overexpressed Pol (3 in the incorporation of ddC and AZT associated with this agent. No additive effect was observed with GCV, into DNA (Figs. 4 and 5). However, AZT-TP was incorporated with a very low efficiency compared to ddC, whereas the cellular toxic 5 W. Nickel, S. Austermann, G. Bialek, and F. Grosse. Interactions of azido-thymidine effect of the two drugs was similar. This discrepancy might be due to triphosphate with the cellular DNA polymerases a, b, and a, and with DNA primases. J. several factors: (a) differential uptake of the analogues; (b) different Biol. Chem., 267: 848â€”854,1992. 115

probably because GCV-TP is a poor substrate for Pol @3(data not DNA polymerase 13bypass in vitro d(GpG)-cisplatin adduct placed on codon 13 of shown). H-rat gene. Proc. NatI. Sci. USA, 92: 5356â€”5360,1995. 6. Zmudzka, B. Z., Fornace, A., Collins, J., and Wilson, S. H. Characterisation of DNA @ In the search for a microorganism which would be naturally polymerase mRNA: cell-cycle and growth response in cultured human cells. sensitive to low concentrations of analogue added in the growth Nucleic Acids Res., 16: 9589â€”9596,1988. medium, which could indicate an efficient drug metabolization, we 7. Srivastava,D. K., Rawson,T. Y., Showalter,S. D., and Wilson, S. H. Phorbolester @ abrogates up-regulation of DNA polymerase by DNA-alkylating agents in Chinese have recently found that E. coli K12 was the most sensitive to AZT hamster ovary cells. J. Biol. Chem.. 270: 16402â€”16408,1995. of several bacterial and fungal strains tested.4 The E. coli tmk gene 8. Fornace, A. J., Zmudzka, B., Hollander, M. C., and Wilson, S. H. Induction of @-polymerasemRNAbyDNAdamagingagentsinChinesehamsterovarycells.Mol. was isolated (16) in such a way that an active gene fusion could be Cell. Biol., 9: 851â€”853,1989. made with AZT-E. coli tk-metabolizing gene to efficiently promote 9. Kunkel,T. A. FrameshiftmutagenesisbyeukaryoticDNApolymerasesin vitro. the first two steps of AZT phosphorylation, final phosphorylation J.Biol.Chem.,261:13581-13587,1986. 10. Copeland, W. C., Chen, M. S., and Wang, T. S. F. Human DNA polymerases a and of AZT-diphosphate to AZT-TP being carried out by a nonspecific 13areableto incorporateanti-HIVdeoxynucleotidesintoDNA.J. Biol.Chem.,267: cellular nucleotide diphosphokinase. We show here that the genetic 21459â€”21464,1992. fusion of E. coli tic and tmk genes encodes a fusion protein that led 11. Parker, W. B., White, E. L., Shaddix, S. C., Ross, L J., Buckheit, R. W., Germany, J. M., Secrist, J. A., Vince, R., and Shannon, W. M. Mechanisms of inhibition of to an additive cytotoxic effect of AZT when associated with Pol @3 human immunodeficiency virus type 1 reverse transcriptase and human DNA poly expression (Fig. 6A) compared to phosphorylation or DNA polym merases a, @,and y by the 5'-triphosphates of carbovir: 3'-azido-3'-deoxythymi dine,2',3'-dideoxyguanosine, and 3'-deoxythymidine. J. Biol. Chem., 266: 1754â€” erization activity taken alone. This effect is much more marked 1762, 1991. than with HSV TK (Fig. 6B), the suicide effect ofthe kinases being still 12. Sommadossi, J. P., Carlisle, R., and Thou, Z. Cellular pharmacology of 3'-azido-3'- more marked than that induced by Pol (3. TK and TMK activities were deoxythymidine with evidence of incorporation into DNA of human bone marrow cells. Mol. Pharmacol., 36: 9â€”14,1989. investigated by measuring the amounts ofanabolic products in extracts of 13. Furman, P. A., Fyfe, J. A., St Gale, M. H., Weinhold, K., Rideout, J. L, Freeman, radiolabeled AZT-treated CHO cells transfected by E. coli tkltmk or HSV 0. A.,Lehrman,S.N.,Bolognesi,D.P.,Broder,S.,andMitsuya,H.Phosphorylation tk (data not shown). In both cases, we found a higher amount of mono of 3'-azido-3'-deoxythymidineandselectiveinteractionof5'-triphosphatewithhu man immunodeficiency virus reverse transcriptase. Proc. Ned. Aced. Sci. USA, 83: phosphorylated and diphosphorylated forms ofAZT. Similar results were 8333â€”8337, 1986. obtained with B16 cell extracts expressing HSVTK.6 Taken together, 14. Neuhard, D. J., and Nygaard, P. Biosynthesis and conversions of nucleotides: purines these assays strengthen the claim that the increased cytotoxicity noted in andpyrimidines.in:F. C. Neidhart,J.C. Ingraham,K.B. Low,B. Magasanik,M. Schaecter, and H. E. Umbarger (eds.), Escherichia co/i and Salmonella typhimurium: cells transfected by nucleoside/nucleotide kinases was due to analogue CellularandMolecularBiology,pp.445-473.Washington,DC:AmericanSociety anabolism. for Microbiology, 1987. These data suggest that the use of new suicide molecular combi 15. Drocourt, D., Calmels, T., Reynes, J. P., Baron, M., and Tiraby, G. Cassettes of the Streptoal/oteichus hindustanus b/c gene for transformation of lower and higher nations should be investigated. Pol @3mightalso be used to potentiate eukaryotes to phleomycin resistance. Nucleic Acids Res., 18: 4009, 1990. the efficacy of novel or existing suicide associations involving TK 16. Reynes, J. P., Tiraby, M., Baron, M., Drocourt, D., and Tiraby, 0. Escherichia co/i thymidylate kinase: molecular cloning, nucleotide sequence and genetic organization activities. This combination ofphosphorylating and replicative suicide of the corresponding test locus. J. Bacteriol., 178: 2804â€”2812,1996. genes may create new opportunities for using chain terminators in 17. Kohara, Y., Akiyama, K., and Isono, K. The physical map of the whole E. co/i protocols for gene therapy of cancer. chromosome: application of a new strategy for rapid analysis and sorting of a large genomic library. Cell, 50: 495â€”508,1987. 18. Takebe, Y., Seiki, M., Fujisawa, J., Hoy, P., Yokota, K., Arai, K., Yoshida, M., and Arai, N. SRalpha promoter: an efficient and versatile mammalian cDNA expression ACKNOWLEDGMENTS system composed of the simian virus 40 early promoter and the R-U5 segment of human T-cell leukemia virus type 1 long terminal repeat. Mol. Cell. Biol., 8: Discussions with Dr. D. Drocourt from Cayla Laboratory and Professor 466â€”472, 1988. G. Tiraby are acknowledged. We gratefully thank Drs. G. Villani and N. 19. Daws, T. D., and Fuchs, J. A. Isolation and characterization of an Escherichia co/i Johnson for the careful review of the manuscript, Dr. S. Wilson for mutantdeficientindl'MPkinaseactivity.J. Bacteriol.,157:440â€”444,1984. providing us with polf3 cDNA, Dr. J. Sweasy for polyclonal antibodies 20. Kawai, S., and Nishizawa, M. New procedure for DNA transfection with polycation and dimethyl sulfoxide. Mol. Cell. Biol., 4: 1172â€”1174,1984. against Pol f3, and Cayla Laboratory for DNA plasmids pUT526& 21. Pressacco, J., Mitrovski, B., Erlichman, C., and Hedley, D. W. Effects of thymidylate pZEOSGO, and pZEOSG1. We also thank N. Tanguy le Gac for scanning synthase inhibition on thymidine kinase activity and nucleoside transporter expres analysis at the Ecole Nationale VÃ©tÃ©rinaire(Servicede Biochimie, Tou sion. Cancer Rca., 55: 1505â€”1508,1995. louse, France). 22. Culver, K. W., Van Gilder, 3., Link, C. J., Carlstrom, T., Buroker. T., Yuh, W., Koch, K., Schabold, K., Doornbas, S., and Wetjen, B. Gene therapy for the treatment of malignantbraintumorswith in vivotumortransductionwiththe herpessimplex REFERENCES thymidine kinase gene/ganciclovir system. Hum. Gene Ther., 5: 343-379, 1994. 23. Moolten, F. L., and Wells, J. M. Curability oftumors bearing herpes thymidine kinase 1. Kornberg, A., and Baker, T. Eukaryotic DNA polymerases. DNA Replication, 2nd ed, genes transferred by retroviral vectors. J. Nail. Cancer Inst., 82: 297â€”300,1990. pp. 197-225. New York: W. H. Freeman, 1992. 24. Huber, B. E., Richards, C. A., and Krenitsky, T. A. Retroviral-mediated gene therapy 2. Sobol, R. W., Horton, J. K., Kuhn, R., Gu, H., Singhal, R. K., Prasad, R., Rajewsky, for the treatment of hepatocellular carcinoma: an innovative approach for cancer K., and Wilson, S. H. Requirement of mammalian DNA polymerase-@in base therapy. Proc. Nail. Aced. Sci. USA, 88: 8039â€”8043,1991. excision repair. Nature (Land.), 379: 183â€”186,1996. 25. Vile, R. 0., and Hart, I. A. Use of tissue-specific expression of the of established 3. Jenkins, T. M., Saxena, J. K., Kumar, A., Wilson, S. H., and Ackerman, E. J. DNA murine melanomas following direct intratumoral injection of DNA. Cancer Rca., 53: polymerase 13and DNA synthesis in Xenopus oocytes and in a nuclear extract. 3860â€”3864,1993. Science (Washington DC), 258: 475â€”478,1992. 26. Amundson, S. A., and Liber, H. L. A comparison of induced mutation at homologous @ 4. Sweasy, J. B., and Loeb, L Mammalian DNA polymerase can substitute for DNA alleles of the tIclocus in human cells. II. Molecular analysis of mutants. Mutat. Res., polymerase I during DNA replication in Escherichia co/i. J. Biol. Chem., 267: 267: 89â€”95,1992. 1407â€”1410,1992. 27. Caruso, M., and Klatzmann, D. Selective killing of CD4@ cells harboring a human 5. Hoffmann, J. S., Pillaire, M. J., Maga, G., Podust, V., Hlibscher, U., and Villani, G. immunodeficiency virus-inducible suicide gene prevents viral spread in an infected cell population. Proc. Natl. Acad. Sci. USA, 89: 182â€”186,1992. 28. Thomas, K. R., and Capecchi, M. R. Site-directed mutagenesis by gene targeting in 6 D. Klatzmann, personal communication. mouse embryo-derived stem cells. Cell, 51: 503â€”512,1987.

116

Downloaded from cancerres.aacrjournals.org on October 5, 2021. © 1997 American Association for Cancer Research. Overexpression of DNA Polymerase β Sensitizes Mammalian Cells to 2 ′,3′-Deoxycytidine and 3′-Azido-3′-deoxythymidine

Khalil Bouayadi, Jean-Sébastien Hoffmann, Pascal Fons, et al.

Cancer Res 1997;57:110-116.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/57/1/110

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/57/1/110. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.