Creation of a Fully Functional Human Chimeric DNA Repair Protein

A dose-limiting toxicity of certain chemotherapeutic alkylating agents is their toxic effects on nontarget tissues such as the bone marrow. To overcome the myelosuppression observed by chemotherapeutic alkylating agents, one approach is to increase the level of DNA repair proteins in hematopoietic stem and progenitor cells. Toward this goal, we have constructed a human fusion protein consisting of O 6-methylguanine DNA methyltransferase coupled with an apurinic endonuclease, resulting in a fully functional protein for bothO 6-methylguanine and apurinic/apyrimidinic (AP) site repair as determined by biochemical analysis. The chimeric protein protected AP endonuclease-deficient Escherichia coli cells against methyl methanesulfonate and hydrogen peroxide (H2O2) damage. A retroviral construct expressing the chimeric protein also protected HeLa cells against 1,3-bis(2-chloroethyl)-1-nitrosourea and methyl methanesulfonate cytotoxicity either when these agents were used separately or in combination. Moreover, as predicted from previous analysis, truncating the amino 150 amino acids of the apurinic endonuclease portion of theO 6-methylguanine DNA methyltransferase-apurinic endonuclease protein resulted in the retention ofO 6-methylguanine DNA methyltransferase activity but loss of all AP endonuclease activity. These results demonstrate that the fusion of O 6-methylguanine DNA methyltransferase and apurinic endonuclease proteins into a combined single repair protein can result in a fully functional protein retaining the repair activities of the individual repair proteins. These and other related constructs may be useful for protection of sensitive tissues and, therefore, are candidate constructs to be tested in preclinical models of chemotherapy toxicity.

DNA alkylating agents are an important part of most doseintensified chemotherapy protocols. Despite the increased use of myeloid growth factor and stem cell support, myelosuppression continues to be a dose-limiting toxicity of many alkylating agents. An example of this would be the severe bone marrow toxicity seen with the use of 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) 1 that is commonly used to treat brain tumors, lymphomas, breast, lung, and gastrointestinal cancers (1). The toxicity to bone marrow cells is most likely due to low levels of existing DNA repair activities that would otherwise help to protect cellular DNA from the damaging consequences of BCNU treatment (2). One strategy to overcome this limited DNA repair capacity is to transduce bone marrow cells with specific genes that encode repair enzymes that act on the DNA lesions produced by BCNU or other alkylating agents. This has recently been accomplished with murine and human bone marrow cells by retroviral-mediated transduction of human O 6methylguanine DNA methyltransferase (MGMT), an activity that repairs BCNU-generated chloroethyl groups at the O 6 position of guanine (3)(4)(5)(6).
Alkylating agents are known to generate multiple DNA adducts by reacting with cellular DNA (7). These agents can alkylate all four bases of DNA at the nitrogens or oxygens as well as the sugar phosphates of the DNA backbone. However, the distribution of the adducts at the various sites depends on both the chemical structure of the alkylating agent and the alkyl group itself. One of the sites, O 6 -methylguanine, preferentially pairs with thymine during DNA replication rather than cytosine, resulting in a GC to AT transition (7,8). Furthermore, BCNU-induced DNA adducts such as a chloroethyl group at the O 6 -position can initiate the subsequent formation of an interstrand cross-link by rearranging to produce an ethyl bridge between N 1 of guanine and N 3 of cytosine in the opposite strand (9). Interstrand DNA cross-links are particularly cytotoxic because they disrupt DNA replication (10). Repair of this lesion is distinct, since it involves the direct reversal of the damaged adduct by MGMT (11,12).
Another important product of attack on DNA by alkylating agents is N 3 -methyladenine, which is cytotoxic (13). In addition, N-alkylpurines are indirectly mutagenic because their removal, either as a spontaneous chemical reaction or by the action of DNA glycosylases, results in the formation of apurinic/apyrimidinic (AP) sites (14,15). Although AP sites normally prevent DNA replication, they can also lead to muta-tions (16). The accumulation of N-alkylpurines may also contribute to other biological effects such as induction of chromosomal aberrations as well as the aging process (17).
The major AP endonuclease in humans (18,19) has been identified to contain two nonoverlapping domains of activity (20). One is for the DNA repair activity associated with APE, where amino acid sequences at the C terminus have been shown to be essential for AP endonuclease activity (21)(22)(23). Amino acid sequences at the N terminus are required for the redox regulation of different transcription factors such as Fos, Jun, Myb, activating transcription factor 1 (ATF-1), ATF-2, cAMP-responsive element-binding protein (CREB), nuclear factor B (NF-B), and p53 (20,24,25) and in its interaction with thioredoxin (26,27).
Since alkylating agents are known to generate many different types of DNA modifications, we hypothesized that added protection to nontarget tissues could be achieved via linking human MGMT together with a number of other DNA repair enzymes to form a protein that recognized a broad spectrum of DNA lesions. As a model of this general approach we first linked the human APE to MGMT, thereby providing for the repair not only of O 6 modifications of guanine but also an activity directed toward baseless sites and modified 3Ј termini in DNA. All of these lesions are a consequence of chloroethylnitrosourea (CNU)-generated DNA damage. We have also demonstrated, contrary to initial reports (28), that deletion of the first 150 amino acids results in the complete loss of AP endonuclease activity in fusion or nonfusion APE proteins.
The data generated provides evidence that by combining domains of DNA base excision repair (BER), enzymes can be used to generate multiactive, highly efficient DNA repair protein(s) for gene therapy. This approach may allow the use of gene transfer to modulate DNA repair in the setting of high dose chemotherapy, with multiple chemotherapeutic agents, thus diminishing dose-limiting cytotoxicity. Molecular Biology and Biochemistry Techniques-DNA sequencing was performed in the macromolecular facility in the Department of Biochemistry and Molecular Biology using an Applied Biosystems automated sequencing system and fluorescent labeling. DNA isolation, RNA isolation, Northern and Western blot analysis, SDS-PAGE, and glutathione S-transferase (GST) fusion protein production and purification were performed as has been previously described (29 -32).

Materials-Enzymes
O 6 -Methylguanine DNA methyltransferase activity using the 18-mer oligonucleotide assay was performed as described (5, 29 -32), whereas AP assays were performed using our standard procedure (32). Briefly, the abasic assay utilized a 37-bp 5Ј 32 P-end-labeled duplex DNA fragment (AP-37mer) as described previously (32), and reaction mixtures (10 l) containing 1 pmol of 5Ј end-labeled AP-37mer, 50 mM HEPES, pH 7.5, 50 mM KCl, 1 g/ml bovine serum albumin, 10 mM MgCl 2 , 0.05% Triton X-100, and the protein of interest. The DNA reaction products were separated on a 20% polyacrylamide gel containing 7 M urea. Dried gels were subjected to autoradiography for visualization and densitometric analysis using the Sigma Scan software package (Jandel Scientific). Cell culture and cell protection assays for Escherichia coli or HeLa cells has been described (5,32,33).
Construction of the MGMT-APE or MGMT-dl151APE Chimeric Molecules in pGEX and MSCV Vectors-THE MGMT-APE or MGMT-dl151APE (deletion of the first 150 amino acids of APE fused with full-length MGMT) fusion constructs were constructed using the overlapping thermocycle amplification technique (34,35). Bacterial cDNA clones of human MGMT and human APE were used as templates for PCR (1 ϫ Mg 2ϩ -free Tfl buffer, 1.5 mM MgSO 4 , 0.1 mM dNTPs, 15 pmols each of 5Ј primer and 3Ј primer, 1 unit of Tfl DNA polymerase (Pro-mega)) in a thermocycler (MJ Research, Inc., Watertown, MA) using the human MGMT (5Ј primer, 5Ј-CCGGAATTCATGGACAAGGATTGT-3Ј, and 3Ј primer, 5Ј-CTTTTTCCCACGCTTCGGGTTTCGGCCAGCAGGC-GG-3Ј) and human APE (5Ј primer, 5Ј-CCGCCTGCTGGCCGAAACCC-GAAGCGTGGGAAAAAG-3Ј, and 3Ј primer, 5Ј-GGCCGTCGACATCA-CAGTGCTAGG-3Ј) primer sequences. The PCR products were electrophoresed through a 1% agarose gel, and fragments (652 bp for MGMT and 986 bp for APE) were gel-purified by centrifugation through glass wool. The MGMT and APE fragments were combined in a second PCR reaction to amplify a 1,603-bp MGMT-APE product utilizing the MGMT 5Ј primer with the APE 3Ј primer under the following cycling conditions (95°C for 10 min, 72°C for 3 min, 94°C for 30 s, 60°C for 1 min, 72°C for 3 min for 30 cycles to step 3 and a final 72°C for 10 min). The PCR products were electrophoresed through a 1% agarose gel. The 1.6-kb MGMT-APE fragment was gel-purified, double-digested with EcoRI and SalI, and ligated into the EcoRI/SalI site in pGEX4T-1 (Pharmacia) using T4 DNA ligase (Life Technologies). After transformation into HB101 competent cells, the colonies containing pGEX4T-MGMT-APE were confirmed by PCR and restriction endonuclease digestion. DNA sequencing was performed to confirm the integrity of the MGMT and APE sequences.
The other construct, MGMT-dl151APE, was constructed in a similar manner. The deletion of the first 150 amino acids of the human APE was performed using PCR and primers that contained the carboxyl end of MGMT and APE nucleotide sequences starting at amino acid 151 of the APE protein sequence (5Ј primer starting at amino acid 151 of APE; 5Ј-CCGCCTGCTGGCCGAAACCATGATCAGGAAGGCCGG-3Ј).
The amplified products were purified and combined, and overlapping PCR was performed as described above. The 1.2-kb human MGMT-dl151APE-purified fragment was double-digested with EcoRI and SalI and ligated into the EcoRI/SalI cloning site in pGEX4T. Positive colonies were confirmed by PCR and restriction digestion and sequenced as described above.
Construction of Retroviral Constructs-The MGMT-APE and MGMT-dl151APE chimeric sequences were removed from the pGEX4T constructs by XhoI and EcoRI, gel-purified, and ligated into the EcoRI/ XhoI cloning site of the retroviral vector MSCV2.1 (generous gift of Dr. Robert Hawley). HB101 cells were transformed with MSCV2.1 MGMT-APE, and MGMT-dl151APE ligation products and positive clones were identified by PCR and restriction digestion.
Retroviral producer cells were generated by the transfection of 2 g of purified plasmid DNA (Qiagen, Chatsworth, CA) added to the Lipofectin transfection reagent (Life Technologies) into GPϩAM12 (an amphotropic retrovirus packaging cell line) cells (36,37) following the protocol provided by the manufacturer. Clones were selected using 0.75 mg/ml G418 (dry powder; Life Technologies), and individual clones were titered on NIH3T3 cells. High titer clones were used to infect HeLa cells in ␣-modified Eagles's medium (Life Technologies) supplemented with 10% fetal bovine serum (Hyclone, Logan, UT) and 10 g/ml polybrene. After infection, HeLa cells were selected for G418 resistance, as above. Individual clones were isolated and examined for expression using Northern and Western blot analysis.
Survival Assays-HeLa cells containing each construct were plated into a 6-well plate (Corning Costar, Cambridge, MA) and cultured overnight at 37°C at 5% CO 2 . The next day, the cells were washed and treated for 1 h with 0 -150 M BCNU (Drug Synthesis and Chemistry Branch, Developmental Therapeutics Program, Division of Cancer Treatment, NCI, NIH, Bethesda, MD) and 0 -2 mM methyl methanesulfonate (MMS; Aldrich). Seven days later the viability of the cells was determined using trypan blue stain and compared with untreated cells and cells with vector alone. Experiments were performed in triplicate and repeated three times. Statistical analysis was performed using SigmaStat (Jandel Scientific) software package (t test and ANOVA).

RESULTS
Chimeric MGMT-APE Construction-The MGMT-APE fusion was constructed using the overlapping PCR technique that has been previously used in our laboratory (35). Briefly, we separately amplified the human MGMT and human APE cDNA sequences (Fig. 1). The 5Ј primer of the MGMT included EcoRI sequences for subsequent cloning into pGEX4T, and the 3Ј MGMT primer contained an additional 18 nucleotides of the 5Ј end of the APE coding region. The 5Ј primer for the APE PCR included 18 nucleotides from the 3Ј coding region of the MGMT cDNA, with the stop codon removed and the 3Ј primer-included sequences for the restriction enzyme SalI (Fig. 1). These PCR products (653 bp for MGMT and 986 bp for APE) were purified and combined in a second PCR reaction to amplify a 1603-bp human MGMT-APE product by utilizing the MGMT 5Ј primer with the APE 3Ј primer. All amplifications were kept under 30 cycles, and large amounts of template were used to decrease the possibility of PCR-induced nucleotide changes. The PCR products were purified, and the 1.6-kilobase human MGMT-APE fragment was double-digested with EcoRI and SalI and ligated into the EcoRI/SalI cloning site in pGEX 4T-1. After transformation into competent cells, the colonies containing pGEX 4T-1 MGMT-APE were confirmed by PCR and restriction digest, and DNA sequencing was used to confirm the integrity of the human MGMT and APE sequence.
Another construct containing the full-length MGMT coding region, but only amino acids 151-318 of the APE coding region, was made in a similar fashion. This construct was made following the initial reports of AP endonuclease activity residing in the carboxyl region downstream of amino acid 150 (20) (Fig.   FIG. 1. Construction of the human MGMT-APE fusion construct. The PCR product was subcloned into pGEX for sequencing and protein overproduction for biochemical analysis. The chimeric MGMT-APE fragment was then excised and cloned into the MSCV2.1 retroviral backbone. The MGMT-dl151APE is a chimeric fusion protein with MGMT activity but no APE activity. h, human; LTR, long terminal repeat; pgk, phosphoglycerate kinase. 2). However, subsequent to this report, it is clear that only the first 60 amino acids can be deleted without significant loss of AP endonuclease activity (21-23); our results are presented below.
MGMT Activity Assay of the Chimeric Proteins-The fulllength human chimeric MGMT-APE pGEX4T construct (MGMT-APE) and MGMT with the deleted APE (MGMT-dl151-APE) were transfected into E. coli ada Ϫ ogt Ϫ (GWR111) cells, and expression of the GST fusion proteins was induced with isopropyl-1-thio-␤-D-galactopyranoside. The 18-mer oligonucleotide assay was employed on the cell extract to determine the activity of the MGMT portion of the chimeric constructs (35,38). As shown in Fig. 3, the GWR111 cells are devoid of O 6 -methylguanine DNA methyltransferase activity as expected, whereas the selected full-length chimeric clone was as active as the nonchimeric human MGMT clone. The selected MGMT-dl151APE clone was also fully active for O 6 -methylguanine repair (Fig. 3). Additional clones were assayed, and similar results were found (data not shown).
AP Endonuclease Activity of Chimeric Clones-The chimeric MGMT-APE and MGMT-dl151APE constructs were transfected into E. coli RPC501 (xth Ϫ , nfo-1 Ϫ ), which is deficient for the two major AP endonucleases in E. coli, exonuclease III (xth) and endonuclease IV (nfo-1). The fusion protein was overexpressed, and the soluble supernatant was applied to a glutathione-agarose column and washed, and the purified MGMT-APE protein was eluted with glutathione. This resulted in homogenous preparations of fusion proteins as judged by SDSpolyacrylamide gel electrophoresis. Concomitantly, APE and dl151APE constructs were prepared in a similar fashion as we have previously reported (32). A 37-mer oligonucleotide with a uracil at position 21 in the 32 P-labeled strand was annealed with the complementary unlabeled oligonucleotide and treated with uracil glycosylase to create an AP site in place of the uracil. This assay is similar to the one used for the 8-oxoguanine repair analysis (Yacoub et al. (32)). As can be seen from Fig. 4A, dilutions of the chimeric MGMT-APE protein (lanes 6 -8) were equally as effective on the AP substrate as APE alone (lanes 2-4). We did not detect any APE activity using this assay with the MGMT-dl151APE construct (data not shown). To confirm that the deleted APE was inactive due to the deleted amino acids and not due to the addition of the MGMT moiety to the carboxyl region, we compared nonfusion APE and dl151APE using the AP oligonucleotide assay (Fig. 4B). We did not see any activity with the dl151APE protein (Fig. 4B). Activity in this assay is demonstrated by the concentration of the 37-mer (upper band) to a 21-mer (lower band). Subsequently, the MGMT-dl151APE served as a negative control for chimeric AP endonuclease function. In additional experiments, we pretreated the MGMT-APE chimeric protein with unlabeled O 6methylguanine oligo substrate that was used in the MGMT assay and then performed the AP assay to ascertain whether the stoichiometric transfer of the methyl group from the DNA to the MGMT portion of the chimeric protein would hinder AP activity. We found no diminution of APE function in this assay (data not shown).
Protection of E. coli AP Endonuclease-deficient Mutant Cells with MGMT-APE Chimera-To verify the activity of the chimeric protein in cells and not just in biochemical analyses, two experiments were performed using E. coli cells that are deficient in AP endonuclease activity. E. coli RPC501(xth Ϫ , nfo Ϫ ) cells with the MGMT-APE and MGMT-dl151APE chimeric constructs in pGEX, as described above, were used on gradient plates with either MMS or H 2 O 2 (32). The length of the cell growth along the gradient is a measure of the resistance or the strain to the agent. Using MMS, the MGMT-APE fully protects the AP endonuclease-deficient cells when compared with wildtype levels (Fig. 5A, lane 3), whereas the MGMT-dl151APE or the dl151APE shows no protection (Fig. 5A, lanes 4 and 5). From the data, it is clear that the chimeric MGMT-APE affords as much protection against MMS as APE alone. APE has previously been shown to protect cells against H 2 O 2 damage (15,18). Using the gradient plate assay but with the DNA damaging agent H 2 O 2 (Fig. 5B), the MGMT-APE chimera was shown to protect to nearly wild-type levels. Once again, the MGMT-dl151APE is deficient in its ability to protect from damage requiring AP endonuclease activity.
In the biochemical and E. coli protection studies described, the chimeric MGMT-APE protein was attached at the amino end to GST, which did not affect either MGMT or APE activity. This suggests that additional repair proteins may be added in lieu of the GST moiety to our construct in future experiments. Protection of Mammalian Cells with the MGMT-APE Chimera-Having concluded that the chimeric MGMT-APE protein was fully functional for MGMT and APE activity in biochemical and E. coli complementation assays, we proceeded to determine the functionality of the chimera in mammalian cells. The chimeric construct was transfected using Lipofectin transfection reagent into GPϩAM12 cells and selected using G418. Individual clones were isolated, expression of the chimeric gene was assessed using Northern blot analysis (data not shown), and virus supernatant was used to infect HeLa cells. After infection, HeLa cells were selected for G418, individual clones were analyzed for RNA (Northern) and protein expression (Western), and two HeLa cell clones were selected for survivability assays in HeLa cells. HeLa cells were treated for 1 h with either 75 or 150 M 1,3-bis(2-chloroethyl)-nitrosourea, 1 or 2 mM MMS, or a mixture of 0.5 mM MMS and 75 M BCNU. 7-10 days later the viability of the cells was determined using trypan blue stain and compared with untreated cells.
As shown in Fig. 6, the chimerics have a 2-fold (83 and 92% versus 44%) survival enhancement over the HeLa cells alone at the lower BCNU dose, whereas at 150 M, there is a 4 -8-fold enhancement (43 and 86% versus 11%). For the MMS protection, there was roughly a 10-fold protection level at both the 1 mM (39 and 47% versus 4%) and 2 mM (12 and 10% versus 1%). The protection afforded by the chimeric construct against a dual exposure of MMS and BCNU was, again, roughly 10-fold (26 and 31% versus 3%) (Fig. 6).

DISCUSSION
Chemotherapeutic agents used in cancer treatments, including dose intensification protocols, typically result in dose-limiting damage to nontarget tissues including bone marrow. A number of studies have utilized MGMT for protection against alkylating DNA damage, particularly in protocols using chloroethylnitrosoureas, which lead to modifications of the O 6methylguanine residue leading to cross-linking and cytotoxicity. However, there exist two problems with this protective gene therapy paradigm; first, alkylating agents create damage at numerous other sites in DNA besides the mutagenic O 6alkylguanine and these other sites, such as N 3 -alkyl-adenine or N 3 -alkyl-guanine, can be cytotoxic. Second, although MGMT may afford protection against cytotoxicity resulting from chloroethylnitrosourea cross-linking via the O 6 -methylguanine lesion, the lack of corrective repair of the other DNA lesions after alkylation damage may have long term effects on the cell, resulting in mutations that could lead to subsequent cell transformations. Moreover, the overexpression of MGMT in cells with little or reduced levels of endogenous MGMT may allow cells to escape the cytotoxicity from agents causing cross-linking, but these cells may then have an increased mutational load due to the lack of repair of the other mutagenic lesions. Therefore, the data presented in this manuscript represents our initial attempts to couple direct reversal repair (MGMT) with members of the BER pathway. We have begun these long term investigations coupling MGMT with APE for several reasons. The first step in the BER pathway involves the removal of the damaged base by a glycosylase such as MPG, which catalyzes the hydrolysis of the N-glycosylic bond that links a base to the deoxyribose-phosphate backbone of DNA. This enzymatic activity results in the generation of an AP site. Although overexpression of MPG might be a logical starting point for coupling MGMT with the BER pathway, we have decided not to initially study the mammalian MPG gene since data by other investigators clearly demonstrate that the overexpression of the MPG gene, by itself, may lead to genetic instability. For example, studies have indicated that the overexpression of MPG results in an imbalance in the BER pathway, leading to a large number of AP sites and gapped intermediates. These alterations lead to the formation of chromosomal aberrations, sister chromatid exchanges, and can block the entry of cells into replication after the exposure to alkylating agents (39,40). Other studies have shown that the overexpression of MPG does not lead to increased resistance to alkylating agents (39) and MPG activity does not appear to be the rate-limiting step in the BER path-way (41). Given this data on MPG, we elected to move one step downstream in the BER pathway and use an APE. This choice was also supported by previous antisense studies that targeted the level of APE in mammalian cells and have demonstrated that these APE Ϫ cells become more sensitive to DNA damaging agents such as MMS and H 2 O 2 (42) as well as bleomycin, menadione, and paraquat (43). These results suggest that APE or downstream members of the BER pathway (DNA ␤-polymerase or ligase) are the rate-limiting steps for base damage repair (44).
We also chose to fuse MGMT with APE, in that order, due to previous data demonstrating the carboxy 25 amino acids of MGMT and approximately 60 amino acids of APE are dispensable for the DNA repair functions of these proteins (21)(22)(23)35). We demonstrated and confirmed these properties using biochemical analyses on the fused MGMT-APE protein (Figs. 3 and 4). Furthermore, we have shown that the chimeric protein can rescue AP endonuclease-deficient E. coli cells back to wildtype levels after alkylating and oxidizing mutagen administration (Figs. 4 and 5). We have also demonstrated the protective ability of our constructs in mammalian cells (Fig. 6). Although very promising, we feel the real advantage of our chimeric constructs will not be fully appreciated until we conduct long term animal studies showing bone marrow protection and mutation reduction following chloroethylnitrosourea and other alkylating agent administration.
Two other points are worth discussing. First, previous studies by us have demonstrated that APE levels are higher in primary bone marrow progenitor cells (CD34 ϩ ) or hematopoietic cell lines that represent progenitors (HL-60) and decline as these cells mature and undergo apoptosis (Kelley and co-workers (33)). 2 Furthermore, we have shown that APE levels also decline in HL-60 cells as they differentiate but can be maintained using retroviral constructs of APE (data not shown). Therefore, even though APE already exists in progenitor cells, gene therapy may allow protective effects in the maturing and developing blood cells. This is important as it demonstrates the ability to overcome endogenous regulatory events leading to decreased APE levels in the cells (33,45).
A second important element of these studies demonstrates the ability of an enhanced protective effect of the chimeric MGMT-APE when cells were challenged with two distinctly different alkylating agents, BCNU and the classic alkylator, MMS. Although we 3 and others (46) have been unable to use expression constructs of APE alone to demonstrate an en- hanced protective effect in mammalian cells, surprisingly we see protection when APE is expressed as a chimera. (However, APNI, the major yeast AP endonuclease and member of the EndoIV family has been used successfully in cell protection studies (46)). This may be attributable to the chimeric protein escaping normal protein-protein interactions due to the fusion of the MGMT moiety onto the amino end of APE. APE has been suggested to form a complex with other BER proteins (47), and the amino region of APE has been shown to interact with other proteins, either on a functional level (25,48,49) or through binding (26,27). Therefore, the addition of MGMT onto the amino end of APE may disrupt these interactions via altered folding or other similar mechanisms. This disruption of interactions hypothesis will be tested using a chimeric construct with MGMT fused onto APE missing its NH 2 -terminal 60 amino acids in which DNA repair activity is reduced by only 20%, but the loss of redox activity is complete. 4 Because of the multifunctional nature of APE, this latter construct will be of importance to compare with the full-length MGMT-APE chimera.
The data presented here demonstrate the feasibility of combining DNA repair proteins into multifunctional repair enzymes. We have shown that the fusion of MGMT with APE results in a fully functional chimeric protein in which the individual activities of both MGMT and APE remain unaffected as shown in biochemical tests and E. coli and mammalian cell protection studies. The fusion construct may be useful in protection of primary bone marrow cells in vivo from simultaneous sequential exposure to different alkylating agents used in doseintensive chemotherapy protocols.