The gene for the bacterial enzyme carboxypeptidase G2 (CPG2) wasexpressed internally in mammalian cells. Mammalian-expressed CPG2had kinetic properties Indistinguishablefrom bacterially expressed CPG2.Human tumor cell lines A2780, SK-OV-3 (ovarian adenocarcinomas),LS174T, and WLDr(colon carcinomas) were engineered to express constitutively either CPG2 or bacterial 13-galactosidase. These cell lines weresubjected to a gene-directed enzyme prodrug therapy regime, using theprodrug 4-[(2-chloroethyl)(2-mesyloxyethyl)amino]benzoyl-L-glutamicacid (CMDA). The lines which expressed CPG2 had enhanced sensitivityto CMDA. Comparing IC@s, WIDr-CPG2 and SK-OV-3-CPG2 were11-16-fold more sensitive, whereas A2780-CPG2 and LS174T-CPG2 were—95-foldmore sensitive than the corresponding control lines. CPG2-expressing cells and control cells were mixed in differing proportions andthen treated with prodrug. Total kill occurred when only —12%of cellsexpressed CPG2 with the W1Dr and SK-OV-3 lines and when only 4-5%of cells expressed CPG2 with the LS174T and A2780 lines, indicating asubstantial bystander effect. These results establish this CPG2 enzyme!
CMDA prodrug system as an effective combination for the gene-directedenzyme prodrug therapy approach.
INTRODUCTION
One of the major problems with current cancer therapies is the lackof specificity of treatment, which leads to harmful side effects innormal tissues, especially the gut lining and bone marrow (1). Currentresearch is thus focused on the development of more selective methods for the delivery of toxic compounds to cancer cells. Gene therapymay be broadly defined as a genetic technology aimed at modifyingcells for therapeutic gain that has been proposed as one such methodto achieve greater selectivity (2).
In cancer gene therapy, both malignant and nonmalignant cells maybe targeted for therapeutic benefit. The possibility to render cancercells more sensitive to chemotherapy or toxins, either by suppressingthe expression of resistance genes (e.g., multidrug resistance gene) orby introducing “suicidegenes―has been considered (2). The latterincludes two approaches: (a) toxin gene therapy, whereby transfectedgenes are able to generate toxins; and (b) enzyme-activating prodrugtherapy whereby transfected genes express foreign enzymes that canactivate prodrugs inside the cancer cells. This latter approach istermed VDEPT (for virally directed enzyme prodrug therapy; Ref. 3),
or GDEPT3 (4, 5).GDEVF is a two-step approach to targeted chemotherapy of human
cancer. In the first step, the gene for a foreign enzyme is delivered tothe tumor in a form that directs tumor-specific expression of theforeign protein. In the second step, a nontoxic prodrug is administeredthat is converted to a cytotoxic drug by the action of the expressed
Received 5/22/96; accepted 8/15/96.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 with18U.S.C.Section1734solelyto indicatethisfact.
I This work was supported by The Cancer Research Campaign, United Kingdom.
2 To whom requests for reprints should be addressed. Phone: 44 1 81 643 8901, ext.
4214. Fax: 44 1 81 770 7899.3 The abbreviations used are: GDEPT, gene-directed enzyme prodrug therapy; AD
enzyme. Since the foreign enzyme is expressed only in the tumor, thecytotoxic drug will be restricted to the tumor. Theoretically, such anapproach should enhance the therapeutic index of chemotherapeuticagents by minimizing systemic toxicity (3, 5).
A variety of different methods for gene delivery has been considered. These include retrovirus (6), naked DNA (7), liposomes (8), andadenovirus (9) delivery. Tissue-specific promoters such as those oftyrosinase (10, 11), a-fetoprotein (12), and carcinoembryonic antigengenes (13, 14) have been used to effect selectivity of expression intarget tissues. Since expression of the foreign enzyme with thesemethods is unlikely to occur in all cells of a targeted tumor in vivo, abystander cytotoxic effect is required whereby the prodrug is cleaved
to an active drug that kills not only tumor cells expressing the foreignenzyme but also neighboring nonexpressing tumor cells (3). In animalmodels, when as few as 2% of the tumor cells express foreign enzymeafter subsequent treatment with a suitable prodrug, long-term survivors can be obtained (3). Therefore, an expression efficiency of10—20%should be enough to achieve 100% cell kill in tumors, andefficiencies of 1—5%are considered sufficient for a therapeutic response (3).
A number of different enzyme/prodrug systems has been designedfor GDEPT. These include purine nucleoside prodrugs, which areactivated by viral thymidine kinase (6, 7, 15—24)or thymidine phosphorylase (11, 25, 26), 5-fluorocytosine activated by bacterial cytosine deaminase (3, 27—31),cyclophosphamide and isofosphamideactivated by rat liver cytochrome P-450 isoenzyme (32), and CB 1954activated by bacterial nitroreductase (4).
In all the prodrug/enzyme examples cited above, the expressedenzyme converts the prodrug to an intermediate metabolite, whichrequires further catalysis by cellular enzymes within the tumor beforethe active drug is formed. If the cellular enzymes responsible for thissecond phase of activation become defective or deficient in the tumorcells, this would lead to resistance of the tumor to the prodrug (5).Thus, it is preferable that the active moiety be released directly fromthe prodrug cleavage by the expressed enzyme.
For maximal benefit, the released drug should be effective againstboth cycling and noncycling cells. Most drugs are active only againstcycling cells, whereas mustard alkylating drugs are also cytotoxic tononcycling cells (33, 34). They have the added advantage that theircytotoxicity is dose related, and their use is less prone than otherclasses of drugs to induce resistance (33, 34). We have previouslyused the bacterial enzyme CPG2 (which has no mammalian homologue) to activate a glutamic acid prodrug derivative of a benzoic acidmustard in an ADEPT context (35, 36). We have demonstrated theefficacy of an antibody-CPG2 conjugate that effected CMDA prodrugactivation in tumor xenograft models in nude mice, such as choriocarcinoma (37) and ovarian (38), colorectal (39, 40), and breastcarcinomas (41). The CPG2 enzyme removes the glutamic acid moiety from the prodrug releasing the active mustard drug (Fig. 1; Ref.42). No further enzymatic processing is required to activate the drug(43, 44).
These properties suggested that the CPG2 enzyme could be a goodcandidate for GDEPT if the corresponding gene could be expressed inmammalian cells. The present study was designed to investigate the
4735
Gene-directed Enzyme Prodrug Therapy with a Mustard Prodrug/Carboxypeptidase G2 Combination'
Richard Marais, Robert A. Spooner, Yvonne Light, Janet Martin, and Caroline J. Springer@CRC Centrefor CancerTherapeuticsat the Institute of CancerResearch,15 CotswoldRoad,Sutton,SurreySM2 5NG [R. A. S., J. M., C.J. S.], and CRC Centrefor Cell andMolecular Biology at the Institute of Cancer Research, 237 Fulham Road, London SW3 6JB (R.M., R.A.S., Y.LJ, United Kingdom
Oligonucleotides. The following oligonucleotides were used. For oligonucleotides nos. 1 and 2, differences from the CPG2 sequence are italicized andcontain the engineered restriction sites; sequences are 5'—+3':no. 1, CGC GGATCC GTF GCA CAC TFG GGT OCT C; no. 2, CGC GM 7712GAC fTGGAT GAC AAG GGC; no. 3, C ATG CCG CAT CAC CAC CAT CAT CACGC; and no. 4, CAT GGC GTG ATG ATG GTG GTG ATG COG.
Prodrug Synthesl@ The CMDA prodrug was synthesized and characterized as described previously (53).
CPG2 Specific Antisera. Rabbit antisera were raised to H@CPG2* cxpressed in Sf9 insect cells by the methods described previously by Marais et a!.(54). For purification of H6CPO2*, 1.5 X 108 cells were infected with virus
particles, incubated for 48 h, and extracted as reported previously (54), exceptthe extraction and dilution buffers contained only 0.5 mMEDTA and 0.06%(vlv) 2-mercaptoethanol. The H@CPG2* protein was purified by Ni2@NTA
agarose affinity chromatography (Qiagen Ltd.) according to the manufacturer'sinstructions, eluting the protein with 90 nmi imidazole. The eluted protein wasjudged to be pure by silver-stained SDS polyacrylamide gels (data not shown).Rabbits were inoculated with —150 @gof purified protein, and subsequentbleeds were examined for production of antibodies by immunoprotcin blottingusing crude insect cell extracts expressing H@CPO2*.
Cell Culture and Transfection. The humancarcinomatumorcell linesA2780 and SK-OV-3 (ovarian adenocarcinomas), LS174T and WiDr (coloncarcinomas), and the COS-7 cell lines were obtained from American TypeCulture Collection. All cell lines were maintained in DMEM supplementedwith 10% FCS (DMEMIFCS; 37°C,10% C02). Transient transfection ofCOS-7 cells was performed as described previously for NIH3T3 cells usingLipofectAMlNE (47). For stable line construction, the human tumor cell lineswere transfected with either pMCEFcpg2* or pMCEF!acZ using LipofectAMINE. Forty eight h following transfection, the cells were recultured intomedium supplemented with 2 mgml' neomycin (0418, Geneticin; LifeTechnologies, Inc.). Individual 0418-resistant colonies were cloned by limiting dilution, and all clones that could degrade MTX were selected for furthercharacterization.
Mammalian Cell Extraction and Kinetic Analysis of CPG2. Cells weregrown to confluence, and extracts were prepared by washing twice with 5 mlofPBSA (137 mMNaCI, 3.4 mMKC1,10 mMNa@HP04,1.8 [email protected]), followed by the addition of200 @lofextraction buffer (250 mM Tris-HC1,
10% v/v glycerol, and 1% v/v Triton X-100, pH 7.5). The cells were lysed insitu (5 mm at room temperature), and the extracts were collected and clarified
by centrifugation in a microfuge (5 mm at 14,000 rpm); the supernatantfractions were stored at —70°C.Cell extracts were subjected to kinetic analyses for CPG2 activity by a modification of the method described previously(37). Briefly, cell extract containing —200ng of CPG2* was added to 1 ml ofCPG2 assay buffer (100 mt@iTris-HC1, and 260 @ts@ZnCl2, pH 7.3) containingMDCatconcentrationsfrom0.5—100pM,andtherateof changeof absorbanceat 320 nm was measured. Km5 were calculated from standard regressionanalysis.
Cytotoxicity Assays. To determine the sensitivity of the cell lines to theCMDA prodrug in vitro, cells were seeded into 6-well tissue culture plates at3 X 10@cells/well (SK-OV-3, A2780, and WiDr) or 2 X l0@cells/well(LS174T) and allowed to grow to confluence. CMDA was prepared in DMSOimmediately prior to use and diluted in DMEM/FCS; then 225 @.dwere used toreplace the medium in wells at 6—4000 @Mconcentrations. After incubation (1h), the medium was replaced with 1 ml of medium containing CMDA at thesame concentration and incubated for an additional 18 h. The cells werewashed and trypsinized; then —3%of the cells was reseeded into fresh dishes.After 4 days of further growth, cell viability was assessed by incorporation of[3H]thymidine(0.4 @Ci/mlfor 6 h). The cells were washed twice with PBSA,fixed in 5% trichloroacetic acid (4°Cfor 20 mm), washed twice with methanol,and air dried. The fixed cells were solubilized (1 ml, 1% SDS/0.2 M NaOH)
Fig. 1. Structures of the prodrug CMDA and itscorresponding drug, 4-[(2-chloroethyl)(2-mesyloxyethyl)amino]benzoic acid formed by the actionof CPG2 on CMDA are shown.
feasibility of this proposal within COS cells (monkey kidney) and infour human tumor epithelial cell lines. Kinetic parameters were measured for all of the modified lines. Cytotoxicity assays in combinationwith prodnig and the bystander effects were studied for GDEPTprotocols.
MATERIALS AND METHODS
Plasmid Construction. All cloning procedures were performed usingstandard molecular biology techniques (45). The cpg2 gene was manipulated
to delete the first 22 amino acids that code for the signal peptide. This wasachieved by PCR, in which the PCR primers were designed to change codons21 and 22 to a BamHI site and the stop codon to an EcoRI site. The 5' cpg2primer (no. 1, see below) was used in conjunction with the 3' CPG2 primer(no. 2, see below) to amplify the cpg2 gene using the plasmid pNM&3O(46).The PCR product was digested with the restriction endonucleases BamHI andEcoRI and cloned into those sites in the polylinker of the mammalian expression vector pEFplink.2 (47). An initiator ATG is supplied by the vector, whichis in frame with the cpg2 reading frame and is in a highly conserved Kozakmotif to ensure efficient translation (48). This plasmid is referred to aspEFcpg2* and encodes a protein, the predicted structure of which is MAGS(CPG2 residues 23—415)-EFLEID (single letter amino acid code; Ref. 49).
The vector pEFlacZ, which codes for bacterial /3-galactosidase, was constructed by cloning the NcoIJXbaIfragment from the plasmid pMLVf3lacZ(50)into those sites in the plasmid pEFplink.2. The plasmid pMCEF—was createdby cloning the end-repaired Hindffl fragment from the plasmid pEFplink4intothe end-repaired XhoI site of a modified version of pMClNeo Poly(A) (Stratagene). The pEFplink fragment contains the elongation factor icr promoter(51) todirectefficientexpressionof theforeignproteinanda multiplecloningsite flanked by the 5' and 3' untranslated regions from the human f3-globingene, which are provided to give efficient mRNA processing, polyadenylation,and translation. The pMCEF—vector, therefore, contains the elongation factorha promoter, which directs expression of the foreign gene, and a polyomaenhancer which directs expression of the Neo'@gene. These two promoters arejuxtaposed on opposite strands of the plasmid, giving divergent expression inopposite directions. The cpg2* coding sequence was cloned from pEFcpg2* asan NcoIJXbaI fragment into the NcoIJSpeI sites of the plasmid pMCEF— tocreate the plasmid pMCEFcpg2*. The lacZ gene was cloned into pMCEFfrom the plasmid pEFf3lacZ, using the same strategy, to create the vectorpMCEFIacZ.
To express CPG2* in Sf9 insect cells, cpg2* was cloned into the insect cellvector pVLH6, a derivative of pVL94l (PharMingen). The EcoRI site inpVL941 was destroyed by digesting the plasmid with EcoRI, end-repairing theoverhangs and then religating the plasmid. The EcoRV/BamHI fragment frompAcYl (52) was then cloned into those sites in pVL941 to create the plasmidpVLplink.2. The oligonucleotide adaptors formed by oligonucleotides nos. 3and 4 were then cloned into the NcoI site of pVLplink.2 to form the plasmidpVLH6. This plasmid contains the multicloning site (4080) CCATG CCG CATCAC CAC CAT CAT CAC 0CC ATG GCG 0CC CGGGTAC CFG CAGATC TAG AAT TCG GAT CC (4135) and encodes an oligohistidine tag(numbers refer to base numbering in pVL941; the initiator codon is in italics).cpg2* was cloned into pVLH,@ as a NcoI/XbaI fragment, and baculoviruses
were prepared using the BaculoGold insect cell virus system (PharMingen). Incloning cpg2* into pVLH6, the resultant plasmid contains the cpg2* sequencefused in frame with the histidine tag. Thus, the expressed protein contains ahistidine tag located at its NH2 terminus and has the structure MPHHHHHHAMAGS-(CPG2 residues 23—415)-EFLEID;it is referred to as H@CPG2*.
Fig. 2. Expression of CPG2@ in COS-7 cells. A,immunoprotein blot analysis. COS-7 cells weretransfected with either pEFlacZ (Lane 1) or pEFcpg2@ (Lane 2), and detergent extracts were prepared. For each cell line, 5 @gof soluble proteinswere analyzed on an 8% polyacrylamide gel andprobed with the rabbit CPG2-specific polyclonalantiserum. The position of standard molecular massmarker proteins is indicated (X 10@),and theCPG25 band is indicated by the arrowhead. B,CPG2 enzymatic activity in transfected cells. DetergentextractsfromcellstransfectedwitheitherpEFIOCZ (Lane 1) or pEFcpg2@ (Lane 2) were testedforCPO2activity,using5 @gofsolubleproteinandCMDA as a substrate. The degradation of CMDAwas monitored at 305 em, and the results are cxpressed in the form of amount CMDA degraded. C,kinetic analysis. Lineweaver-Burk plots, where theX-axis intercept of a linear regression equals —1/Km,@ shown for CPG2@ expressed in mammalian
cells(•)andfor bacteriallyexpressedCPG2(0).
200—
96—
68—
45
31lIES](M-1)
and the soluble fraction was added to 6 ml of scintillation fluid to determine thethymidine incorporation; the results are expressed as percentage growth ofcontrol cells that were treated with vehicle alone. For bystander cytotoxicityassays, mixtures of CPG2*@expressingcell lines and their appropriate lacZ cellline were seeded as above and treated with either 1 mM(A2780 and LS174T)or 2 sims(SK-OV-3 and WiDr) CMDA.
Miscellaneous. Immunoproteinblottingwas performedby standardtechniques using ‘25I-labeledprotein A to detect specific antibody interactions(55). The primary antiserum was used at a concentration of 1:2000. SDS
PAGEwasperformedunderstandardconditions(56).
RESULTS
Expression of CPG2 in COS Cells. Our initial experiments weredesigned to address whether the bacterial enzyme (CPG2) could beexpressed in mammalian cells in an active form. Wild-type CPG2is a secreted enzyme found in the bacterial periplasm (46). Themature protein is a homodimer consisting of subunits, the molec
ular weight of which is —42,000, and the monomers are inactive.We were concerned that if CPG2 were secreted from tumor cells,it could escape from the tumor, giving rise to nonspecific toxicityowing to production of active drug at distant sites. Therefore, weexpressed CPG2 internally in a form that could not be secretedfrom mammalian cells. We also wished to establish whether theCPG2 in mammalian cells was active, since problems due toincorrect folding of the protein, lack of formation of dimers, or tosequestration into a hostile cell subcompartment might render thisenzyme inactive. To this end, we chose to use a transient transfection system based on COS-7 cells to examine whether CPG2could be expressed in mammalian cells.
Our first priority was to create a protein that would not be secretedfrom mammalian cells. The first 22 codons of the cpg2 gene encodea signal peptide that is responsible for targeting the protein to theperiplasm of bacteria and which is removed by proteolysis followingmembrane translocation (57). We used PCR-directed mutagenesis toremove the sequences encoding the signal peptide and cloned thealtered gene into the mammalian expression vector pEFplink.2, whichuses the promoter from the elongation factor liz gene to directexpression of foreign proteins in mammalian cells (47). This promoterwas chosen because it is active in a wide variety of cell types (51); theexpression construct is referred to as pEFcpg2*.
We used immunoproteinblotting to determinewhetherCPG2* proteinwas expressed in COS-7 cells. COS-7 cells were transfected with either
pEFcpg2* or pEF!acZ (which codes for the bacterial (3-galactosidaseenzyme), and detergent extracts were prepared. These extracts wereexamined using a CPG2-specific rabbitpolyclonal antiserum.The resultsshow that an immunoreactive band with an apparent molecular weight of—42,000was detected in cells transfected with pEFcpg2*, which wasabsent in the cells transfected with pEFlacZ (Fig. 2A). We also found thatwhen CPG2* was analyzed in nondenaturing gels, it migrated with amobility that is consistent with the protein being a dimer (data not
shown). The detergent extracts were then analyzed for CPG2 activity,using CMDA as a substrate. The extract from the cells transfected withpEFcpg2* was able to degrade CMDA, whereas there was no suchactivity in the control f3-galactosidase extract (Fig. 2B). Conversely, using
O-nitrophenyl f3-o-galactopyranoside as a substrate, we could detect
j3-galactosidase activity in the extract from the cells transfected withpEF!acZ but not in the extract from the cells transfected with pEFcpg2*(data not shown).
Using MTX as a substrate, the mammalian-expressed CPG2* wasfound to have a Km of 7 @tM,which is in close agreement with our ownfmdings for the bacterially expressed wild-type protein (Fig. 2C) andwith those published previously for CPG2 purified from the periplasmof Escherichia co!i and Pseudonwnas sp. strain RS16 (46). To determine whether CPG2* expressed in COS-7 cells was secreted, weexamined the tissue culture supematant from the pEFcpg2*@trans@
fected cells for CPG2 activity, and none was detectable (data notshown). These data, taken together with the lack of the signal peptideand the fact that CPG2* can be released from the transfected cellswith the detergent Triton X-lOO, suggest that CPG2* is not secretedby mammalian cells.
Establishment of Cell Lines Constitutively Expressing CPG2*.The results presented above show that when CPG2* is expressedinternally in mammalian cells, it is soluble and fully active, indicatingthat neither the alterations that have been introduced into the codingsequence nor the intracellular location affect the enzymic properties ofCPG2. Because we wished to create a model system to examine thepotential 0fCPG2* in a GDEPT approach, we established mammaliantumor cell lines that constitutively expressed CPG2*. This wasachieved by cloning the gene encoding CPG2* into the mammalianexpression vector pMCEF— to create the plasmid pMCEFcpg2*. Thevector pMCEF— uses the elongation factor lea promoter to directforeign gene expression, but it also contains the Neo't gene; therefore,the vector can be used to select cells with the cytotoxic drug G418.We chose the human colon carcinoma cell lines LS174T (58) and
The clone from each celi line type picked forfurther study.
12345678
A2780 LS174T SK-OV-3 WiDr* * * *C'4 C@4 e'1 N
N Q N C., N Q N QV@ V@ V@ U
.@ U ,@ LI ,@ LI ,@ U
GDEPT WITH A MUSTARDPRODRUG
WiDr (59) and the human ovarian adenocarcinoma cell lines SKOV-3 (60) and A2780 (61) for this analysis.
For each tumor line, cells were transfected with either pMCEFcpg2* or the plasmid pMCEF1acZ (which directs expression of bactenal j3-galactosidase), and G4l8-resistant colonies were selected.G418-resistant colonies were examined for CPG2* expression byenzymic assay, using MTX as a substrate, and for j3-galactosidaseactivity using O-nitrophenyl @-D-galactopyran05ide as a substrate(data not shown). All the cell lines were found to express >0.4 HiCPG2*/mg soluble protein (data not shown). We then examined thesusceptibility of the cell lines to prodrug treatment in tissue culture byincubating each line with increasing concentrations of the prodrugCMDA. For these experiments, the cells were treated with CMDAtwice, and then the medium was replaced with fresh prodrug-freemedium. Thus, the cells were incubated with the prodrug for only19 h. The rate of survival of the cells was determined by [3H]thymidine incorporation.
The results show that the cell lines fall into two different categories(Table 1): those that are highly susceptible to CMDA, with IC50s inthe 5—50/.LMrange (LS174T and A2780); and those that are lesssensitive to CMDA, with IC50s in the range 250—600p.M (WiDr andSK-OV-3). For each tumor cell line, one clone was selected to expresseither CPG2* or @3-galactosidase for further study. The CPG2*@ex@pressing lines were selected to express similar amounts of CPG2activity (—1 unit/mg of detergent-soluble proteins; Fig. 3A and Table2). Each selected line was then subjected to immunoprotein blottinganalysis. An immunoreactive band with Mr @42,OOOwas detected inthe cell lines that contained CPG2 activity but not in those thatcontained @-galactosidase activity, confirming the presence of the
cpg2* gene (Fig. 3B).Following G4l8 selection, the selected cell lines were transferred
into medium lacking G418 to determine the stability ofexpression andto assess whether CPG2* was toxic to mammalian cells. No differences in the levels of CPG2* enzyme activity were detected, evenwhen the cells were maintained in the absence of G418 selection for4 months, and no significant differences were detected in the rate ofcell growth of the CPG2*@expressing cells compared to control (3-galactosidase-expressing cells of the same lineage (data not shown).Because we did not observe any reduction in the expression of theprotein with time or major differences in the rate of growth rate ofcells, we conclude that the expression of CPG2* in these cell lines isstable, and CPG2* is not toxic to mammalian cells.
The IC50s for the selected CPG2*@expressing cell lines were compared to their corresponding (3-galactosidase lines. The IC50 for theselected WiDr@CPG2* line was found to be >11 fold lower than the
A. A2780LS174TSK-OV-3WiDr* * * *C'4 Cl @4
NQ NQ NQ NQ
1.5 @b@ @b
@1.o
4@ 0.5c'4Q
0
B.
200
96 —
68—
45 —
31—
1 2 3 4 5 6 7 8
Fig. 3. Constitutive expression of CPG2* in human tumor cell lines. A, CPG2* enzymeactivity in human tumor cell lines. The levels of CPG2 enzyme activity in the pEFlacZtransfected cell lines (Lanes 1, 3, 5, and 7) or the pEFcpg2@-transfectedcell lines (Lanes2. 4, 6. and 8), for A2780 cells (Lanes I and 2); LS174T cells (Lanes 3 and 4); SKOV-3cells (Lanes 5 and 6); and WiDr cells (Lanes 7 and 8) were determined by standard kineticanalysis and are expressed as a function of activity with respect to the soluble protein inthe cell extracts. B, immunoprotein blot analysis. Immunoblot analysis was performedwith the CPG2-specific rabbit antiserum on 5 @gof detergent-extracted protein for eachof the cell lines stably transfected with either pMCEFIacZ (Lanes I, 3, 5, and 7) orpMCEFcpg2* (Lanes 2, 4, 6, and 8), for A2780 cells (Lanes I and 2); LS174T cells(Lanes 3 and 4); SKOV-3 cells (Lanes 5 and 6); and WiDr cells (Lanes 7 and 8). Theposition of migration of standard molecular mass markers (X l0@) is indicated, and theCPG2* band is indicated by the arrowhead.
.4
IC50for the WiDr-lacZ line (Fig. 4; Table 1); the IC50of the selectedSK@OV@3@CPG2*line was >16 fold lower than the IC50 for theSK-OV-3-lacZ line; the IC50 for the selected A278O@CPG2* line was>92-fold lower than the IC50 for the A2780-lacZ line, and the IC50 ofthe selected LSl74T@CPG2* line was —95-fold lower than the IC50for the LS174T-lacZ line. These data show that for each each tumorcell line, expression of CPG2* significantly increases the sensitivityof the cells to the CMDA prodrug.
Table2 Expressionof CPG2*in selectedhumantumorcell linesFor each cell line, the levels ofCPO2 activity have been determined by standard kinetic
analysisandareexpressedas a functionof unitsof activity/mgof cellproteinin thecellextracts (units/mg). The sensitivity ofthe cell lines to CMDA was determined by treatingthe cells with varying concentrations of prodrug. The IC5@sare shown and are defmed astheconcentrationofprodrugrequiredtokill50%ofthe cells,comparedtovehicle-treatedcontrols.
results with the SK-OV-3 and the WiDr clones were similar, 50% ofthe cells in the culture mixture could be killed when only 1—2%of thecells were expressing CPG2*; 90% of the cells were killed when —8%of the cells expressing CPG2*, and 100% cell kill was achieved whenabout 12% of the cells expressed CPG2* (Fig. 5; Table 3). The A2780and LS174T clones were able to direct an even greater bystandereffect; 50% cell kill occurred when only 0.1 and 1.6%, respectively,of the cells were CPG2* expressors; 90% of the cells were killedwhen 2.0 and 3. 1%, respectively, of the cells were CPG2* expressors.Total cell kill occurred when only an estimated 5 and 3.7%, respectively, of the cells were CPG2* expressors (Fig. 5 and Table 3).
DISCUSSION
We have analyzed the potential use of the enzyme CPG2 togetherwith the prodrug CMDA as a mammalian GDEVF model. To preventCPG2 secretion, we removed the signal peptide from the cpg2 gene tomake CPG2* and cloned the altered gene into mammalian expressionvectors. The altered protein had the same kinetic properties as wildtype CPG2 purified from bacterial cells but was located intracellularly. The altered gene was cloned into a variety of mammalian tumorcells for constitutive expression. The stable cell lines expressingCPG2* were subjected to GDEPT protocols, using the CMDA prodrug. The sensitivity of the CPG2*@expressing lines fell into twocategories: the more susceptible A2780 and LS174T lines; and theless sensitive WiDr and SK-OV-3 lines. Clearly, this is not lineagedependent because A2780 and SK-OV-3 are derived from ovarianadenocarcinomas, whereas WiDr and SK-OV-3 are derived fromcolon carcinomas.
We found that in each case, the selected cell lines expressingCPG2* were significantly more sensitive to the prodrug than the(3-galactosidase controls, with IC50 differentials between 11- and95-fold. Bystander assays of mixtures of the CPG2* and control lines
150
100
50
0150
V
0
a)1.4
E0V
01.4
@150a)
11:
@150
11:
[CMDAI(jiM)Fig. 4. Susceptibility to CMDA of human tumor cell lines expressing [email protected] of
the cell lines was treated with increasing concentrations of CMDA. and survival wasdetermined by [tmHlthymidineincorporation. The results are expressed as the proportion ofsurviving cells relative to the vehicle-treated controls. For each cell line, the control celllines expressing @-galactosidase(0) are shown with their corresponding CPG2@-expressing line(I). Bars,SD.
The Bystander Effect. The modified cell lines were examined fortheir ability to mount a bystander effect in vitro. The bystander effectis defined as the ability of CPG2*@expressing cells to kill neighboringcells that do not express CPG2* in the presence of the prodrug. Wedetermined this effect by growing mixtures of each of the clonesexpressing CPG2* with their corresponding (3-galactosidase control,followed by treatment of the cells with the CMDA prodrug. Theconcentration of the prodrug used was determined by the sensitivity of
the (3-galactosidase-expressing lines to CMDA and was chosen to beapproximately one-half the IC50 concentration of those lines. Thus,for the A2780 and LS174T clones, the concentration used was 1 nmiCMDA, and for the SK-OV-3 and the WiDr clones, the concentrationwas 2 mM CMDA (Table 3). At these concentrations, between 80 and100% of the lacZ-expressing cells survive.
Mixtures of cells were treated with the prodnig, and cell survival
was determined by [3H]thymidine incorporation. The results areshown in Fig. 5 and expressed in Table 3. They indicate that all of theclones expressing CPG2* were able to direct a bystander effect. The
100
0 25 50 75 1000 25 50 75 100
Proportion of CPG2*@ expressing cells (%)Fig. 5. Bystander effect of cells expressing CPG2*. For each of the tumor models, the
cell line expressing @-galactosidase and the cell line expressing CPG2* were mixed invarious proportions in tissue culture and treated with the appropriate concentration ofCMDA (1 mMfor A2780 and LS174T; 2 mat for SK-OV-3 and WiDr). The proportion ofcells surviving was determined by [3H]thymidine incorporation and is expressed as theproportion of surviving cells relative to the cells expressing 100% (3-galactosidasetreatedwith the prodrug. The dashed line predicts cell survival if there were no bystander effect.Bars, SD.
Table 3 The bystander effect in vitroFor each tumor model, the line expressing CPG2* was mixed with the line expressing @-galactosidaseand treated with the indicated concentration
cells required to kill 50, 90. and 100% of the cells was determined by logarithmic regression of the data presented in Fig. 5.ofCMDA. The proportionofProportion
of CPG2*@expressing cells required for:CMDA conc.'@
Cell line (mM) 50% cell death 90% cell death100% celldeathA2780
demonstrated that total cell kill could be obtained when only 3.7 to12% of the cells were expressing CPG2*.
To obtain selective toxicity in GDEPT, there should be no endogenous enzyme capable of catalyzing the prodrug to the cytotoxicmoiety. In previous ADEPT clinical trials, when the CMDA prodrugwas administered to patients without prior injection of an antibodyCPG2 conjugate, there was no CMDA related toxicity (62), nor wasthere any conversion of the prodrug to the active drug, as monitoredin plasmaby high-performanceliquidchromatographyand the moresensitive technique of liquid chromatography-mass spectrometry (63).
Taken together, these data demonstrate that there is no human enzymecapable of converting CMDA to the active drug when it is administered systemically. Thus, the required selectivity of GDEPT can beattained by the CPG2 enzyme.
In clinical trials with CMDA in ADEPT, doses of 2.2—5.5mmols/m2/days of prodrug were administered to patients (62), and concentrations in excess of 3 mt@ihave been measured in the plasma ofpatients.5 The concentrations of 1—2mr@iCMDA used in the in vitrobystander cytotoxicity studies described herein, therefore, fall withinthe range of concentrations that could be achieved in patients. At theseconcentrations, the CMDA was not cytotoxic to the cells in vitro thatwere not expressing CPG2. However, in GDEPT this level of prodrugconcentration would be effective to kill those cells that express theenzyme and would be sufficient to direct a substantial bystanderresponse.
Activated CMDA exerts its cytotoxicity by cross-linking DNA;thus, it is able to kill both cycling and noncycling cells. This contrastswith the most commonly used GDEPT enzyme/prodrug system, thymidine kinase/ganciclovir, which is cytotoxic only during the S phaseof the cell cycle (64). It has been proposed that resistance to ganciclovir in GDEPT is due to the proportion of cells in G0 at the time ofganciclovir administration (65). Resistant tumor outgrowth occurred,
despite up to 30 days of continuous ganciclovir administration, mdicating that some tumor cells can remain in G0 for long periods.Tumors that grew out remained sensitive to ganciclovir on additionaladministrations of the prodrug, showing that acquired resistance wasnot the cause for the lack of response. In contrast, mustard alkylatingagents are not cell cycle dependent and are able to exert their cytotoxic effects in a cell cycle independent manner (33, 66). The alkylating agent prodrug CMDA can be converted directly by CPG2 to themustard drug, without intermediate metabolites (42). We found thatcells expressing CPG2* intracellularly were between 10- and 100-foldmore sensitive to CMDA than cells of the same lineage expressingbacterial (3-galactosidase. Thus, CMDA used in combination withCPG2* should provide advantages over the existing prodrug enzymesystemsused in GDEPT.
It has been shown that gene delivery to tumors in vivo does not leadto modification of 100% cells (3, 67). Therefore, an essential requirement for GDEPT is that the activated prodrug should have a bystandereffect, whereby conversion of prodrug to the active form in the
5 j. Martin and C. Springer,unpublisheddata.
enzyme-modified cells leads to killing of adjacent unmodified cells.Herein, the bystander effect was examined by mixing cells expressingCPG2* with those of the same lineage that did not express CPG2,followed by treatment with CMDA. We have demonstrated thateffective bystander effects occurred when we mixed CPG2*@express@ing and nonexpressing cells in combination with CMDA. Total cell
kill was obtained when as few as 3.7% of the cells were expressingCPG2*. This compares very favorably with the data on in vitrobystander cytotoxicity of other enzyme/prodrug systems: (a) total celldeath in combination with ganciclovir required expression of thymidine kinase in 60—90%of the cell population (7, 68); (b) expressionof cytosine deaminase in 33% of cells in vitro resulted in 100% cellkill with 5-fluorocytosine (3); (c) in a cytochrome P-450-activatingenzyme GDEPT system, >50% cells were required to express theenzyme to obtain total cell kill with cyclophosphamide (69); and (d)only 1% (25) to 2% (11) of cells expressing purine nucleoside phosphorylase were required to effect almost total cell kill with purinenucleoside prodrugs. However, in each case with the purine nucleoside phosphorylase, there was an extended 6-day incubation withprodrug required to obtain such an effect; in order to obtain greaterthan 90% cell kill with the prodrug CB1954, 30—50%of cells wererequired to express nitroreductase (4).
In summary, CPG2 represents an effective enzyme for use inGDEPT, in combination with the CMDA prodrug, providing a novelGDEPT system with advantages over the other enzyme/prodrug combinations. We are currently investigating the utility of this combination in vivo in CPG2*@modified tumor xenografts of the ovarianadenocarcinoma and the colon carcinoma cell lines described here.
ACKNOWLEDGMENTS
We thank Professors K. R. Harrap and C. J. Marshall for support andProfessor I. Niculescu-Duvaz for helpful discussions. We are grateful to Dr. R.Treisman for the pMLV(3lacZ plasmid, Dr. N. Minton for the pNM83Oplasmid, Dr. R. Sherwood for bacterially expressed CPG2, and 0. Patel for thepAcYl plasmid.
REFERENCES
1. Calabresi, P., and Welch, A. D. Cytotoxic drugs, hormones and radioactive isotopes.In: L. S. Goodman and A. Oilman (eds.), Cytotoxic drugs, hormones, and radioactiveisotopes, pp. 1345—1393.New York: The Macmillan Company, 1994.
2. Thang, W. W., Fujiwara, T., Grimm, E. A., and Roth, J. A. Advances in cancer genetherapy. Adv. Pharmacol., 12: 289—341, 1995.
3. Huber, B. E., Austin, E. A., Richards, C. A., Davis, S. T., and Good, S. S. Metabolismof 5-fluorocytidine to 5-fluorouracil in human colorectal tumor cells transduced withthe cytosine deaminase gene: significant antitumor effects when only a small percentage of tumor cells express cytosine deaminase. Proc. Nail. Acad. Sci. USA, 91:8302—8306,1994.
4. Bridgewater, G., Springer, C. J., Knox, R., Minton, N., Michael, P., and Collins, M.Expression of the bacterial nitroreductase enzyme in mammalian cells renders themselectively sensitive to killing by the prodrug CB1954. Eur. J. Cancer, 31A: 2362—2370,1995.
5. Niculescu-Duvaz, I., and Springer, C. J. Gene-directed enzyme prodrug therapy(GDEPT): choice of prodrugs. In: K. D. Bagshawe (ed.) Gene-directed EnzymeProdrug Therapy (ODEPT): Choice of Prodrugs, in press. Amsterdam: ElsevierScience B. V., 1996.
6. Culver, K. W., Ram, Z., Walbridge, S., Ishii, H., Oldfield, E. H., and Blaese, M. R.
in vivo gene transfer with retroviral vector-producer cells for treatment of experimental brain tumors. Science (Washington DC). 256: 1550—1552,1992.
7. Vile, R. G., and Hart, I. R. Use of tissue-specific expression of the herpes simplexvirus thymidine kinase gene to inhibit growth of established murine melanomasfollowing direct intratumoral injection of DNA. Cancer Res., 53: 3860—3864,1993.
8. Nabel, 0. J., Chang, A. E., Nabel, E. G., Plautz, 0. E., Ensminger, W., Fox, B. A.,FeIgner,P., Shu,S., andChoK. Immunotherapyfor cancerby directgenetransferinto tumors. Hum. Gene Ther., 5: 57—77,1994.
9. O'Malley,B. W.,Jr., Chen,S. H.,Schwartz,M.R.,andWoo,S. L. C. Adenovirusmediated gene therapy for human head and neck squamous cell cancer in a nudemouse model. Cancer Rca., 55: 1080—1085, 1995.
10. Vile, R., Miller, N., Chemayovsky, Y., and Hart, I. A comparison of the properties ofdifferent retroviral vectors containing the murine tyrosinase promoter to achievetranscriptionally targeted expression of the HSVtk or IL-2 genes. Gene Ther., 1:307—316,1994.
I 1. Hughes, B. W., Wells, A. H., Bebok, Z., Gadi, V. K., Garver, R. I., Parker, W. B., andSorscher, E. J. Bystander killing of melanoma cells using the human tyrosinasepromoter to express the Escherichia coli purine nucleoside phosphorylase gene.Cancer Res., 55: 3339—3345, 1995.
12. Ido, A., Nakata, K., Kato, Y., Nakao, K., Fujita, M., Ishii, N., Tamaoki, T., Shiku, H..andNagataki,S. Genetherapyforhepatomacellsusinga retrovirusvectorcarryingherpes simplex virus thymidine kinase under the control of human a-fetoprotein genepromoter. Cancer Res., 55: 3105—3109,1995.
13. Trinh, Q. T., Austin, E. A., Murray, D. M., Knick, V. C., and Huber, B. E.Enzyme/prodrug gene therapy: comparison of cytosine deaminase/5-fluorocytosineversus thymidine kinase/ganciclovir enzyme/prodrug systems in a human colorectalcarcinoma cell line. Cancer Res., 55: 4808—4812, 1995.
14. Tanaka, T., Kanai, F., Okabe, S., Yoshida, Y., Wakimoto, H., Hamada, H., Shiratori,Y., Lan, K. H., Ishitobi,M., and Omata,M. Adenovirus-mediatedprodruggenetherapy for carcinoembryonic antigen-producing human gastric carcinoma cells invitro. Cancer Res., 56: 1341—1345,1996.
15. Moolten, F. L. Tumor chemosensitivity conferred by inserted herpes thymidine kinasegenes: paradigm for a prospective cancer control strategy. Cancer Res.. 46: 5276—5281, 1986.
16. Borelli,E., Heyman,R., Hsi,M.,and Evans,R. M.Targetingof an inducibletoxicphenotype in animal cell. Proc. Natl. Acad. Sci. USA, 85: 7572—7576,1988.
17. Moolten, F. L., and Wells, J. M. Curability of tumors bearing herpes virus thymidinekinase genes transferred by retroviral vectors. J Natl. Cancer Inst., 82: 297—300,1990.
18. Huber, B. A., Richards, C. A., and Krenitsky, T. A. Retroviral-mediated gene therapyfor the treatment of hepatocellular carcinoma: an innovative approach for cancertherapy. Proc. Natl. Acad. Sci., USA, 88: 8039—8043,1991.
19. Caruso,M.,Panis,Y.,Gagandeep,S.,Houssin,D.,Salzmann,J. L.,andKlatzmann,D.Regressionofestablishedmacroscopiclivermctastasesafterinsitutransductionofa suicide gene. Proc. Natl. Acad. Sci. USA, 90: 7024—7028, 1993.
20. Oldfield, E. H., Ram, Z., Culver, K. W., Blaese, R. M., and DeVroom, H. L. Genetherapy for the treatment of brain tumors using intra-tumoral transduction withthymidine kinase gene and intravenous ganciclovir. Hum. Gene Ther., 4: 39—69,1993.
21. Barba, D., Hardin, J., Ray, J., and Gage. F. H. Thymidine kinase-mediated killing ofrat brain tumors. J. Neurosurg., 79: 729—735,1993.
22. Vile, R. G., and Hart, I. R. in vitro and in vivo targeting of gene expression tomelanoma cells. Cancer Rca., 53: 962—967,1993.
23. Barbs, D., Hardin, J., Sadelain, M., and Gage, F. H. Development of anti-tumorimmunity following thymidine kinase-mediated killing of experimental brain tumors.Proc. Natl. Acad. Sci. USA, 91: 4348—4352, 1994.
24. Chen, H., Boiziau, J., Parker, F., Mailliet, P., Commercon, A., Tocque, B., Le Pecq,J. B., Roques,B. P., and Garbay,C. Structure-activityrelationshipsin a seriesof5-[(2, 5-dihydroxybenzyl)amino] salicylate inhibitors of EGF-receptor-associated tyrosine kinase: importance of additional hydrophobic aromatic interactions. J. Med.Chem., 37: 845—859,1994.
25. Sorscher, E. J., Peng, S., Bebok, Z., Allan, P. W., Bennett, L. L., and Parker, W. B.Tumor cell bystander killing in colonic carcinoma utilizing the Escherichia coli DeoDgene to generate toxic purines. Gene Ther., 1: 233—238,1994.
26. Patterson, A. V., Zhang, H., Moghaddam, A., Bicknell, R., Talbot, D. C., Stratford,I. J., and Harris, A. L. Increased sensitivity to the prodrug 5'-deoxy-S-fluorouridineand modulationof 5-fluoro-2'-deoxyuridinesensitivityin MCF-7cells transfectedwith thymidine phosphorylase. Br. J. Cancer, 72: 669—675,1995.
27. Mullen, C. A., Kilstrup, M., and Blaese, R. M. Transfer of the bacterial gene forcytosine deaminase to a mammalian cells confers lethal sensitivity to 5-fluorocytosine: a negative selection system. Proc. NatI. Acad. Sd. USA, 89: 33—37,1992.
28. Austin, E. A., and Huber, B. E. A first step in the development of gene therapy forcolorectal carcinoma: cloning, sequencing and expression of Escherichia coli cytosine deaminase. Mol. Pharmacol., 43: 380—387,1993.
29. Huber,B. E., Austin,E. A., Good,V. C., Knick,V. C., Tibbels,S., and Richards,C. A. in vivo antitumor activity of 5-fluorocytosine on human colorectal carcinomacells genetically modified to express cytosine deaminase. Cancer Res., 53: 4619—4626, 1993.
30. Mullen, C. A., Coale, M. M., Lowe, R., and Blaese, R. M. Tumor expressing thecytosine deaminase suicide gene can be eliminated in vivo with 5-fluorocytosine andinduce protective immunity to wild type tumor. Cancer Res., 54: 1503—1506,1994.
31. Harris, J. D., Gutierrez, A. A., Hurst, H. C., Sikora, K., and Lemoine, N. R. Genetherapy for cancer using tumor-specific prodrug activation. Gene Ther., 1: 1—6,1994.
32. Chen, L., and Waxman, D. J. Intratumoral activation and enhanced chemotherapeuticeffect of oxazaphosphorines following cytochrome P-450 gene transfer: developmentofa combined chemotherapy/cancer gene therapy strategy. Cancer Res., 55: 581—589,1995.
33. Teicher. B. A.. and Frei, E., III. Development of alkylating agent resistant humantumor cell lines. Cancer Chemother. Pharmacol., 21: 292—298,1988.
34. Frei, E., III, Teicher, B. A., Holden, S. A., Cathcart, K. N. S., and Wang, Y.Preclinical studies and clinical correlation of the effect of alkylating drugs. CancerRes., 48: 6417—6423, 1988.
35. Bagshawe, K. D. ADEPT and related concepts. Cell Biophys., 24/25: 83—91, 1994.36. Springer, C. J., Dowell, R. L., Burke, P. J., Hadley, E., Davies, D. H., Blakey, D. C.,
Melton, R. G., and Niculescu-Duvaz, I. Optimization of alkylating agent prodrugsderived from phenol and aniline mustards: a new clinical candidate prodrug (ZD2767)for ADEPT. J. Med. Chem., 2: 687—706,1995.
37. Springer, C. J.. Bagshawe. K. D.. Sharma, S. K., Searle, F., Boden, J. A., Antoniw,P., Burke,P. J., Rogers,G. T., Sherwood,R. F., and Melton,R. 0. Ablationofhuman choriocarcinoma xenografts in rode mice by antibody-directed enzyme prodrug therapy (ADEPT) with three novel compounds. Eur. J. Cancer, 27: 1361—1366,I991.
38. Sharma, S. K.. Boden, J. A., Springer, C. J., Burke, P. J., and Bagshawe, K. D.Antibody-directed prodrug therapy (ADEPT): a three-phase study in ovarian tumourxenografts. Cell Biophys., 24125: 219—228,1994.
39. Antoniw, P.. Springer, C. J., Bagshawe, K. D., Searle, F., Melton, R. 0., Rogers,0. T.. Burke, P. J., and Sherwood, R. F. Disposition of the prodrug 4-[bis(2-chloroethyl)aminolbenzoyl-L-glutamicacid and its active parent drug in mice. Br. J.Cancer, 62: 909—914,1990.
40. Sharma, S. K., Bagshawe, K. D., Springer, C. J., Burke, P. J., Rogers, 0. T.,Boden J. A., Antoniw, P., Mellon, R. G., and Sherwood, R. F. Antibody-directedenzyme prodrug therapy (ADEPT): a three phase system. Disease Markers, 9:225—231,1991.
41. Eccles, S., Court, W. J., Box, G. A., Dean, C. J., Mellon, R. G., and Springer, C. J.Regression of established breast carcinoma xenografts with antibody-directed enzymeprodrug therapy against c-erbB2 p185. Cancer Res., 54: 5171—5177, 1994.
42. Springer, C. J., Antoniw, P., Bagshawe, K. D., Searle, F., Bisset, 0. M. F., andJarman, M. Novel prodrugs which are activated to cytotoxic alkylating agents bycarboxypeptidase G2. J. Med. Chem., 33: 677—681, 1990.
43. Springer, C. J., Antoniw, P., Bagshawe, K. D., and Wilman, D. E. V. Comparison ofhalf-lives and cytotoxicity of N-mesyloxyethyl- and N-chloroethyl-4-amino benzoylcompounds, products of prodrugs in antibody-directed enzyme prodrug therapy(ADEPT). Anticancer Drug Design, 6: 467—479,1991.
44. Springer, C. J. Aspects of antibody directed prodrug therapy (ADEPT). in: J. Menon(ed), Current Topics in Medicinal Chemistry, pp. 93—100.India: Council of ScientificResearch, 1993.
45. Sambrook, J., Fritsch, E. F., and Maniatis, T. Molecular Cloning: A LaboratoryManual, Ed. 2. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory, 1989.
46. Minton, N. P., Atkinson, T., and Sherwood, R. F. Molecular Cloning of the Pseudomonas carboxypeptidase G2 gene and its expression in Escherichia coli and Pseudomonas putida. J. Bacteriol., 156: 1222—1227,1983.
47. Marais, R., Light, Y., Paterson, H. F., and Marshall, C. J. Ras recruits Raf-l to theplasma membrane for activation by tyrosine phosphorylation. EMBO J., 14: 3136—3145,1995.
48. Kozak, M. Point mutations define a sequence flanking the AUG initiator codon thatmodulates translation by eukaryotic ribosomes. Cell, 44: 283—292,1986.
49. IUPAC-LUB. Commission on biochemical nomenclature. A one-letter notation foramino acid sequences: tentative rules. Biochem. J., 113: 1—4,1969.
50. Hill, C. S., Marais, R., John, S., Wynne, J., Dalton, S., and Treisman, R. Functionalanalysis of a growth factor responsive transcription factor complex. Cell, 73: 395—406, 1993.
5 1. Mizushima, S., and Nagata, S. pEF-BOS, a powerful mammalian expression vector.Nucleic Acids Res., 18: 5322, 1990.
52. Patel, G., Nasmyth, K., and Jones, N. A new method for the isolation of recombinantbaculoviruses. Nucleic Acids Research, 20: 97—104,1992.
53. Springer. C. J. CMDA, an antineoplastic prodrug. Drugs Future, 18: 212—215,1993.54. Marais, R. M., Hsuan, J. J., McGuigan, C., Wynne, J., and Treisman, R. Casein kinase
II phosphorylation increases the rate of serum response factor-binding site exchange.EMBO J.,11:97—105,1992.
55. Towbin, H., and Gordon, J. Immunoblotting and dot immunobinding: current statusand outlook. J. lmmunol. Methods, 72: 313—340,1984.
56. Laemmli, U. K. Cleavage of structural proteins during the assembly of the head ofbacteriophage T4. Nature (Lond.), 227: 680—685, 1970.
57. Minton, N. P., Atkinson, T., Bruton, C. J., and Sherwood, R. F. The completenucleotide sequence of the Pseudomonas gene coding for carboxypeptidase G2. Gene(Amst.), 31: 31—38,1984.
58. Tom, B. H., Rutzky, L. H., Jakstys, M. M., Oyasu, R., Kaye, C., and Kahan, B. D.Human colonic adenocarcinoma cells. I . Establishment and description of a new cellline. In Vitro, 12: 180—181,1976.
59. Noguchi, P., Wallace, R., Johnson, J., Earley, E. M., O'Brien, S., Ferrone, S.,Pellegrino, M. A., Milstien, J., Needy, C., Browne, W., and Petricciani, J. Characterization of the WIDR: a human colon carcinoma cell line. In Vitro, 15: 401—408,1979.
60. Hill, B. T., Whelan, R. D., Gibby, E. M., Sheer, D., Hosking, L. K., Shellard, S. A.,and Rupniak, H. T. Establishment and characterisation of three new human ovariancarcinoma cell lines and initial evaluation of their potential in experimental chemotherapy studies. mt. J. Cancer, 39: 219—225,1987.
61. Louie,K. G., Behrens,B. C., Kinsella,T. J., Hamilton,T. C., Grotzinger,K. R.,McKoy, W. M., Winker, M. A., and Ozols, R. F. Radiation survival parameters ofantineoplastic drug-sensitive and -resistant human ovarian cancer cell lines and theirmodification by butathionine sulfoximine. Cancer Res., 45: 21 10—2115, 1985.
62. Bagshawe, K. D., Sharma, S. K., Springer, C. J., and Antoniw, P. Antibody directedenzyme prodrug therapy: a pilot-scale clinical trial. Tumor Targeting, I: 17—30,1995.
63. Springer, C. J., Poon, 0. K., Sharma, S. K., and Bagshawe, K. D. Analysis ofantibody-enzyme conjugate clearance by investigation of prodrug and active drug inADEPT clinical study. Cell Biophys., 24/25: 193—207,1994.
64. Mar, E. C., Chiou, J. F., Cheng, Y. C., and Huang, E. S. Inhibition of cellular DNApolymerase-a and human cytomegalovinis-induced DNA polymerase by the triphosphates of 9-(2-hydroxyethoxymethyl)guanine and 9-(l , 3-dihydroxy-2-propoxymethyl)guanine. J. Virol., 53: 776—780, 1985.
65. Golumbek, P. T., Hamzeh, F. M., Jaffec. E. M., Levitsky, H., Leitman, P., andPardoll, D. M. Herpes simplex-l virus thymidine kinase gene is unable to completelyeliminatelive,nonimmunogenictumorcell vaccines.J. Immunother.,12:224—230,1992.
66. Frei, E., III. Curative cancer chemotherapy. Cancer Res., 45: 6532—6538, 1985.67. Ram,Z., Culver,K. W., Walbridge,S., Blaese,R. M., and Oldfield,E. H. in situ
retroviral-mediated gene transfer for the treatment of brain tumors in rats. CancerRes., 53: 83—88,1993.
68. Freeman, S. M.. AbbOud,C. N., Whartenby, K. A., Packman, C. H., Koeplin, D. S.,Moolten, F. S., and Abraham, 0. N. The “bystandereffect―:tumor regression whena fraction of the tumor mass is genetically modified. Cancer Res., 53: 5274—5283.1993.
69. Chen, L, Waxman, D. J., Chen, D., and Kufe, D. W. Sensitization of human breastcancer cells to cyclophosphamide and ifosphamide by transfer of a liver cytochromeP450 gene. Cancer Res., 56: 1331—1340, 1996.
1996;56:4735-4742. Cancer Res Richard Marais, Robert A. Spooner, Yvonne Light, et al. Prodrug/Carboxypeptidase G2 CombinationGene-directed Enzyme Prodrug Therapy with a Mustard
![Stimuli-responsive oligonucleotides in prodrug-based ...the oligonucleotide field. Based on the definition of a prodrug given by Albert in 1958 [12], a prodrug is an agent that under-goes](https://static.documents.pub/doc/80x56/5e9fe1c20dd6ff22d727d93b/stimuli-responsive-oligonucleotides-in-prodrug-based-the-oligonucleotide-field.jpg)
