Original Article | Published:

Human–yeast chimeric repair protein protects mammalian cells against alkylating agents: enhancement of MGMT protection


Chemotherapeutic DNA alkylating agents are common weapons employed to fight both pediatric and adult cancers. In addition to cancerous cells, nontarget tissues are subjected to the cytotoxicity of these agents, and dose-limiting toxicity in the form of myelosuppression is a frequent result of treatment. One approach to prevent myelosuppression that results from the use of chemotherapeutic agents is to increase the levels of DNA repair proteins in bone marrow cells. Here we report our second successful attempt to create a fusion protein that possesses both direct reversal and base excision repair pathway DNA repair activities. The chimeric protein is composed of the human O6-Methylguanine-DNA Methyltransferase (MGMT) and the yeast Apn1 proteins and retains both MGMT and AP endonuclease activities as determined by biochemical analysis. We have also demonstrated that the chimeric protein is able to protect mammalian cells from the DNA alkylating agents 1,3-bis (2-chloroethyl)-1-nitrosourea (BCNU) and methyl methanesulfonate (MMS). The protection by the chimeric protein against BCNU is even greater than MGMT alone, which has potential translational significance given that MGMT is currently in clinical trials. Additionally, we show that the chimeric MGMT-Apn1 protein can protect mammalian cells from dual treatments of BCNU and MMS and that this effect is greater than that provided by MGMT alone. The data support our previous finding that a protein with multiple DNA repair activities can be constructed and that this and other constructs may play an important clinical role in guarding against dose-limiting effects of chemotherapy, particularly in situations of multiple drug use.


Dose-escalation studies using chemotherapeutic agents are a common situation in the clinical arena in an attempt to boost survival rates of both adult and pediatric cancers.1 Chemotherapeutic alkylating agents continue to play a crucial role in most of these dose-intensified chemotherapy protocols.2 However, despite the use of growth factors and stem cell support, myelosuppression continues to be a dose-limiting toxicity of many alkylating agents. A number of years ago we, and a number of other investigators, began a series of investigations using DNA repair cDNAs, particularly those involved in direct reversal or DNA base excision repair (BER), to protect bone marrow cells during dose-intensified chemotherapy.3,4 This avenue of research is ongoing with a number of individual cDNAs beginning to make their way through preclinical trials.4 However, it is becoming evident that the most successful approach may involve the use of more than one repair cDNA, either introduced in succession or at the same time in the cells of interest.4,5 One approach to dual repair protein expression has been to combine repair proteins, either through the use of multicistronic transcripts or direct fusion of repair proteins with different activities to enhance cellular protection against single and combination chemotherapeutic agents.3,6 Toward this goal, one avenue we are systematically investigating concerns which DNA repair enzymes are most effective when expressed as a single fusion protein.

Alkylating agents are commonly used in chemotherapy for a wide variety of cancers and are known to generate multiple DNA adducts by reacting with cellular DNA.1,7 These agents can attack all the bases of DNA; however, some alkylating agents demonstrate different adduct specificities that may allow for the proper pairing of the agents used and the repair protein selected to protect cells. For example, although 1,3-bis (2-chloroethyl)-1-nitrosourea (BCNU) produces alkylations at a variety of sites in DNA, one of the sites, O6-methylguanine, preferentially pairs with thymine during DNA replication resulting in a GC to AT transition.7,8 Furthermore, BCNU-induced DNA adducts, such as a chloroethyl group at the O6-position, can initiate the subsequent formation of an interstrand crosslink by rearranging to produce an ethyl bridge between N1 of guanine and N3 of cytosine in the opposite strand.9 Interstrand DNA crosslinks 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 However, even if the O6-guanine is corrected before crosslink formation or mutation induction, BCNU also alkylates a number of other sites in DNA that could lead to cell death or an increased mutational burden in that cell if not corrected.

As mentioned above, alkylating agents are known to generate many different types of DNA modifications; therefore, we hypothesized that added protection to nontarget tissues could be achieved by means of linking DNA BER proteins and MGMT to form a protein that recognized a broader spectrum of DNA lesions. As a model of this general approach, we first linked the human apurimic (AP) endonuclease Ape1 to MGMT, thereby providing for the repair, not only of O6 modifications of guanine, but also an activity directed toward baseless sites in DNA. However, in retrospect, Ape1 may not have been the best choice, since it is also one of the major redox factors in cells. Therefore, we decided to combine MGMT with another AP endonuclease, the yeast Apn1 protein. Apn1 was chosen for a number of reasons: first, it is another AP endonuclease, but one that does not have other associated activities besides its DNA repair capabilities. Second, we wanted to demonstrate that the “chimeric repair protein” concept is applicable to more than one construct and, third, it is a eucaryotic protein whose AP activity can be analyzed in the presence of endogenous Ape1 activity since, in contrast to endogenous Ape1, it is ethylenediaminetetraacetic acid (EDTA) resistant and Mg2+ independent.

In these studies, we have demonstrated that the MGMT-Apn1 protein is fully functional, both biochemically and by providing cellular protection against the alkylating agents BCNU and methyl methanesulfonate (MMS) in mammalian cells. Additionally, the MGMT-Apn1 construct is even more effective against BCNU treatments than MGMT alone. This corroborates our previous studies using MGMT-Ape1 in a similar manner and suggests that this approach could be applicable for other repair proteins leading to multifunctional repair enzymes and could be an alternative to, or combined with, the internal ribosome entry site (IRES) coexpression approach.


Chimeric MGMT-Apn1 construction

The MGMT-Apn1 chimeric molecule was created using the overlapping polymerase chain reaction (PCR) technique (Fig 1).3 Initially, the MGMT and Apn1 cDNA sequences were amplified individually from bacterial plasmid clones using the standard PCR procedure. The 5′ MGMT primer included an EcoRI restriction site for subcloning into the pcDNA3 vector, and the 3′ MGMT primer included the first 20 nucleotides of the 5′ end of the Apn1 cDNA. The 5′ Apn1 primer included the final 20 nucleotides of the MGMT 3′ end, with the stop codon removed, and the 3′ Apn1 primer included an XhoI restriction site for subcloning into the pcDNA 3.0 vector. The individual PCR products were separated by 1% agarose gel electrophoresis and the 644 bp MGMT and the 1123 bp Apn1 bands were excised and purified. The MGMT and Apn1 DNA fragments were combined in another PCR reaction with the 5′ MGMT primer and the 3′ Apn1 primer to amplify the 1747 bp MGMT-Apn1 chimeric cDNA. Large amounts of template DNA were used and all amplifications were performed with a high-fidelity, heat-stable DNA polymerase and kept under 30 cycles to decrease the likelihood of PCR-induced mutations. The MGMT-Apn1 chimeric cDNA molecule was subjected to double DNA restriction digest and subcloned into the EcoRI and XhoI cloning sites of the pcDNA3 neovector. Following transformation into competent DH5α cells, positive MGMT-Apn1 clones were confirmed by PCR and restriction digest. DNA sequencing was performed to confirm the integrity of the MGMT and Apn1 sequences.

Figure 1

Construction of the chimeric MGMT-Apn1 construct. The MGMT and Apn1 cDNA sequences were amplified from bacterial clones and the chimeric molecule was created using the overlapping PCR technique.3 The fusion product was subcloned into the pcDNA 3.0 vector, sequenced and used to transfect K562 cells.

Transfection and characterization of K562 cells with the MGMT-Apn1 chimera

K562 cells (human myeloid leukemia) were stably transfected with the pcDNA 3.0 neo MGMT-Apn1 plasmid and positive clones were selected by culturing in 0.75 mg/ml G418. Positive clones were initially characterized by Northern blot analysis. Both the radiolabeled MGMT and Apn1 cDNA molecules hybridized with a 1.75 kb transcript from total RNA of the G418 resistant clone (Fig 2a), while neither probe hybridized to total RNA of vector-transfected K562 cells, which are negative for both MGMT and Apn1 expression. The lower level of hybridization observed with the MGMT probe is only due to a weaker labeling of that reagent. The radiolabeled glyceraldehydes-3-phosphate dehydrogenase (GAPDH) cDNA hybridized with a 1 kb transcript from total RNA of both the chimeric G418 resistant clone and the vector-only transfected K562 cells. Total protein was collected from both the G418 chimeric resistant clone and vector-only transfected K562 cells in order to confirm expression of the MGMT-Apn1 chimera by Western blot analysis (Fig 2b). Chemiluminescent detection revealed that the primary monoclonal MGMT antibody detected an approximately 64 kDa protein from the chimeric clone, but the MGMT antibody did not detect any protein from vector-transfected K562 cells, which are MGMT negative. The 64 kDa protein of the chimeric clone is approximately the molecular weight of the sum of MGMT (21 kDa) and Apn1 (40 kDa). The results of Northern and Western blot analyses suggested that the chimeric G418 resistant clone expresses the MGMT-Apn1 chimera.

Figure 2

Analysis of K562 cells containing chimeric transgene. (a) The MGMT-Apn1 K562 clone and pcDNA vector-transfected control cells were analyzed by Northern blot analysis. Both the radiolabeled MGMT and Apn1 cDNAs hybridized with a 1.75 kb message from total RNA of MGMT-Apn1-transfected cells, but not of vector-control cells. The radiolabeled GAPDH cDNA hybridized to a 1 kb transcript from total RNA of both control cells and MGMT-Apn1-transfected cells. (b) Western blot analysis: The MGMT monoclonal antibody detected the 64 kDa MGMT-Apn1 fusion protein of transfected K562 cells, but not of control cells (lane 3). The antibody also detected the 21 kDa MGMT protein from MGMT-transfected K562 cells (lane 2). (c) MGMT activity analysis: The 18-mer O6-methyl guanine oligonucleotide assay was performed with 50 μg of protein from sonicated K562 cell extract. MGMT activity was detected in both MGMT- and MGMT-Apn1-transfected K562 cells. Presence of the 8-mer fragment indicates MGMT activity. (d) AP endonuclease activity assay: The 26-mer oligonucleotide assay was performed with 5 μg of total protein from sonicated cell extract in the presence of 20 mM EDTA. Apurinic endonuclease activity was detected in MGMT-Apn1-transfected cells, but not control or MGMT-transfected cells. Presence of the 14-mer indicates cleavage of the artificial AP site. Apn1 is Mg2+ independent, EDTA resistant whereas the endogenous mammalian Ape1 protein requires Mg2+ and is inactive in the presence of EDTA.

In addition to the data shown here, we have analyzed a number of other MGMT-Apn1 clones with similar results (data not shown).

MGMT activity of K562+MGMT-Apn1 cells

The 18-mer methylguanine DNA methyltransferase oligonucleotide assay was performed with K562 cells and K562 cells expressing the MGMT-Apn1 transgene (Fig 2c). If the cells possess O6-methylguanine DNA methyltransferase activity, the methyl group was removed from the guanine residue allowing PvuII digestion of the 18-mer into an 8-mer. The presence of the 8-mer in Figure 2 demonstrated that the cell extract of K562 cells that expressed the chimeric MGMT-Apn1 as well as the extract of K562 cells that expressed MGMT possessed O6-methylguanine DNA methyltransferase activity (Fig 2a, c). Additionally, the amount of MGMT activity, when normalized to unreacted substrate, was the same for the K562 cells with MGMT versus those with MGMT-Apn1 construct. As expected, the cell extract of vector-transfected K562 cells, which are MGMT negative, possessed no O6-methylguanine DNA methyltransferase activity.

Apurinic endonuclease activity of K562+MGMT-Apn1 cells

To confirm the AP endonuclease activity of the MGMT-Apn1 chimera, the 26-mer oligonucleotide assay was used (Fig 2d). In this assay, a radiolabeled 26 bp DNA fragment with a THF residue, which mimics an AP, was incubated with cell extract in the presence of 20 mM EDTA.13 EDTA ensures that the K562 cells’ endogenous AP endonuclease, which requires Mg2+ as a cofactor, remains inactive while Saccharomyces cerevisiae Apn1, which does not require Mg2+ as a cofactor, maintains its AP endonuclease activity.13,14 This reduces background and allows for specific detection of the transgene expression of AP endonuclease. Additionally, the THF oligonucleotide is resistant to cleavage by other types of glycosylase/lyase enzymes found in mammalian cells.27,33 The reaction products were separated by polyacrylamide gel electrophoresis and visualized by autoradiography. If the cell extract possessed AP endonuclease activity, the THF residue was recognized as an AP site and the DNA fragment was cleaved resulting in a 14-mer fragment. The presence of the 14-mer in Figure 2 demonstrated that the cell extract of K562+MGMT-Apn1 cells possessed AP endonuclease activity due to the presence of Apn1. As expected, the cell extract of pcDNA vector-transfected K562 cells had no AP endonuclease activity when the AP endonuclease assay was performed in the presence of 20 mM EDTA.

Protection of K562 cells with MGMT-Apn1

Cell survival assays were used to determine if the biochemically active MGMT-Apn1 chimeric protein was able to protect transduced cells from various DNA-damaging agents. K562 cells were chosen for the cell survival assays, as they are negative for both MGMT and Apn1 (since Apn1 is a yeast enzyme and no mammalian homologue has been identified, K562 cells are obviously devoid of this protein) and as a model system for bone marrow cells as previously described.1,3,4 K562 cells that expressed MGMT-Apn1 had greater than three-fold enhancement (90 versus 25%) at 0.10 mM MMS and greater than eight-fold survival enhancement (71 versus 8%) at 0.15 mM MMS when compared to K562 cells that did not express MGMT-Apn1 (P<.01). In contrast, we found that K562 cells that solely express MGMT were as equally sensitive as vector-transfected K562 cells to relatively low doses of MMS (0.01 mM), while at higher doses (0.15 and 0.20 mM MMS), we determined that MGMT expressing K562 cells had a greater than a two- and four-fold survival enhancement (15 versus 8%, 14 versus 3%, respectively) when compared to control cells (P<.01). Expression of Apn1 alone produced moderate protection against MMS, particularly at the higher doses. However, the fusion protein was clearly higher than the protection afforded by either MGMT or Apn1 alone.

Additionally, the chimeric MGMT-Apn1 construct also protected K562 cells from the cytotoxic effects of the simple alkylator BCNU (Fig 3, panel b). K562 cells were grown in the presence of various concentrations of BCNU for 6 days, and cell viability was determined using trypan blue stain and compared with untreated cells (Fig 3, panel b). K562 cells that expressed the MGMT-Apn1 chimeric protein had nearly a seven-fold (69 versus 11%) survival enhancement at 50 μM BCNU and nearly a 20-fold (39 versus 2.5%) survival enhancement at 75 μM BCNU when compared to K562 cells that did not express the transgene (P<.01). Apn1 did not protect against BCNU (data not shown).

Figure 3

MGMT-Apn1 cell survival assays. A selected MGMT-Apn1 clone was grown in increasing concentrations of MMS (panel a) or BCNU (panel b). After 6 days, cell survival was determined and compared to untreated and vector-alone control cells. Additionally, the cell survival of an MGMT K562 clone was also analyzed, as well as an Apn1 clone (panel a). Experiments were performed in triplicate and repeated three times. Data points represent mean % cell survival and error bars represent the standard error of each data point. Vertical axis scaling has been adjusted for visual presentation. Statistical significance difference data are presented in the Results section.

Cell survival assays were also performed to determine if MGMT-Apn1 enhanced K562 cell survival when DNA-damaging agents were combined (data not shown). K562 cells were grown in either 0.2 mM MMS, 75 μM BCNU, or a combination of either 10 μM BCNU and 0.10 mM MMS, 20 μM BCNU and 0.10 mM MMS, or 20 μM BCNU and 0.05 mM MMS and cell survival determined as described in Materials and methods. In all instances, we found that K562 cells expressing MGMT-Apn1 were significantly protected from the combination of DNA alkylators. At 10 μM BCNU and 0.10 mM MMS, K562 cells expressing MGMT-Apn1 had a nine-fold enhanced survival (74 versus 8.2%, P<.01) compared to K562 cells that do not express MGMT-Apn1 and greater than a nine-fold (52 versus 5.6%) enhanced survival compared to K562 cells that do not express the transgene when grown in 20 μM BCNU and 0.10 mM MMS (P<.01). When K562 cells expressing MGMT-Apn1 are grown in 20 μM BCNU and 0.05 mM MMS, they have a 4.6-fold (79 versus 17%) enhanced survival compared to vector-transfected K562 cells (P<.01).

Vector-transfected K562 cells were approximately three-fold more sensitive than K562 cells expressing MGMT to combinations of both 10 mM BCNU and 0.10 mM MMS, and 20 mM BCNU and 0.10 mM MMS (P<.01). At 20 mM BCNU and 0.05 mM MMS, however, there was no significant difference in growth between these two cell lines (data not shown).

We have also performed the same cell protection experiments with three other MGMT-Apn1 clones and have obtained similar results as those presented here (data not shown).


Chemotherapeutic agents that damage DNA are routinely used in cancer treatment protocols. However, one of the main problems encountered with chemotherapy is the collapse of the hematopoietic compartment during treatment, particularly when high doses of agents are used.1 Clearly, if the hematopoietic and immune systems could be protected, this would be beneficial and also allow for more aggressive approaches toward tumor treatment. The target for gene transfer is the hematopoietic stem cells as they have been shown to reconstitute the hematopoietic and immune systems permanently in vivo.15,16 By genetically manipulating the hematopoietic stem cell compartment with genes that confer resistance to chemotherapeutic agents, the dose escalation that is necessary to treat the tumor could potentially be achieved.17 One approach to achieve this goal is to perform a bone marrow transplant on the patient in an attempt to provide bone marrow stem cell support. However, this approach can be compromised due to the lack of appropriately matched donors. As an alternative to this approach, it has been suggested that the donors’ own bone marrow cells, particularly CD34+stem cells, be isolated, expanded, and either reintroduced or reintroduced following the insertion of a cDNA or cDNAs that code for drug resistance or DNA repair proteins.4 For example, cellular detoxification genes, such as glutathione S-transferase (GST), aldehyde dehydrogenase (ALDH), and multiple drug resistance (MDR1), have all been shown to be involved in drug resistance and have shown some moderate, but mixed success in protecting targeted cells.18,19,20 For example, Wang et al combined the expression of MDR and MGMT in a bicistronic vector and found that this particular combination produced CD34+cells that were resistant to P-glycoprotein-effluxed drugs, as well as BCNU.21 Also, Suzuki et al have shown protection in HeLa cells and murine bone marrow cells from nitrosourea damage after retroviral transfer of MGMT and human MDR1 connected with an IRES.22 The approach using DNA repair genes has more recently been investigated for protection of hematopoietic cells.23 The MGMT DNA repair protein has been studied extensively in gene therapy experiments.2,8 Human and murine hematopoietic cells expressing an MGMT transgene gain resistance to BCNU.2,8,24,25 MGMT has been used in hematopoietic cells to protect against nitrosoureas given that these cells have a low or absent endogenous levels of MGMT,26 and clinical trials in humans are ongoing. However, there are two main problems with using MGMT: first, it only repairs O6-methylguanine lesions, which in the case of nitrosoureas are cytotoxic due to their ability to form crosslinks, or if they cause a futile-cycle processing by the MMR system. These lesions are mainly mutagenic. Second, if cells containing MGMT afford protection, it is hypothesized that they will carry an increased mutational burden since MGMT can only process one DNA adduct and not the others that occur when DNA is attacked by alkylating agents.4 Therefore, although MGMT may protect against the initial O6-methylguanine and subsequent crosslinking toxicity, there is a high likelihood that secondary site mutations could result leading to other cancers, such as leukemia. This has led us to investigate other DNA repair genes that can either enhance protection and/or repair damaged sites that may become fixed mutations if not correctly repaired.

As mentioned, we have previously demonstrated that human MGMT could be fused to the human AP endonuclease, Ape1, which creates a functional chimeric protein.3 However, given the multifunctional activities of Ape1 in mammalian cells,27 we wanted to create another chimeric protein to demonstrate the effectiveness of this approach. We chose the major yeast AP endonuclease Apn1. Apn1 is also a Class II endonuclease, but it is part of the endonuclease IV family in Escherichia coli (Ape1 is a member of the exonuclease III family) and does not contain the redox function of Ape1.27 Apn1 cleaves 5′ of the AP site, and exhibits 3′-phosphodiesterase activity, but lacks 3′ to 5′ exonuclease activity found in the exonuclease III family.14 Apn1 has been overexpressed in CHO cells demonstrating an increase in AP endonuclease activity in the transfected cells, and demonstrated increased resistance to MMS and hydrogen peroxide as well as fewer chromosomal aberrations than control CHO cells.28 Additionally, in human epithelial lung cell lines, we have demonstrated that Apn1 can protect against the chemotherapeutic agent bleomycin.13

Using the MGMT-Apn1 constructs, we demonstrated and confirmed these properties using biochemical analyses on the fused MGMT-APN1 protein (Fig 2). We also demonstrated the protective ability of our constructs in mammalian cells (Fig 3), not only with enhanced protection against BCNU, but also in experiments where we combine BCNU and MMS treatments. The enhanced protection seen with the Apn1 chimeric protein as compared to the Ape1-MGMT protein could be due to the more robust 3′ diesterase activity of Apn1 versus Ape1. However, given the agents used to see this enhanced cellular protection, it is more likely that this is not the activity that is responsible for the enhanced protection observed; more likely it is due to Apn1 acting more as a pure AP endonuclease without the other functions seen with Ape1 (redox, etc.) Although very promising, we feel that 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 clinically relevant alkylating agent administration.

The data presented here further our continuing “proof-of-concept” studies, supporting the hypothesis that selected DNA repair proteins can be fused such that they perform either multifunction, act upon a wider variety of DNA lesions/adducts than each individually, and/or they comprise rate-limiting components of various steps of the repair pathways. It is interesting that the fusion of the two proteins does not affect each of the individual activities. Although this might seem surprising when first observed, it may not be all that unexpected given that each has very different substrate specificities and one, MGMT, is a stoichiometric acting protein while the other, Apn1, is enzymatic.

With the rapid proliferation of new studies and data on the structure–function of DNA repair proteins,29,30,31,32 future constructions may include multiple proteins, or selected regions of the various repair proteins in order to build a highly efficient or multifunctional repair protein. These constructs and approach may eventually be useful in protection of primary bone marrow cells, in vivo, from simultaneous sequential exposure to different chemotherapeutic agents. Finally, it was somewhat of a surprise that the chimeric MGMT-Apn1 fusion protein conferred much better protection against BCNU than just MGMT alone. Although we anticipated additional protection in the two-drug model, this enhanced protection to BCNU was not fully anticipated. This finding may be of usefulness in current studies and clinical trials that are currently using MGMT alone8 and warrants further investigation. It also raises questions about using MGMT alone such that it may not afford full protection to the protected cells without the potential accumulation of unrepaired DNA damage. This, too, warrants further investigation.

Materials and methods

Molecular biology and biochemistry techniques

DNA sequencing was performed at the University of Davis (Davis, CA) DNA sequencing facility. DNA isolation, RNA isolation, Northern and Western blot analyses, and sodium dodecyl sulfate-polyacrylamide eletrophoresis (SDS-PAGE) were performed as has been previously described.3 Western blot analysis was performed with the mouse anti-MGMT monoclonal antibody (MT 23.2; Novus Biologicals, Littleton, CO) and the goat anti-mouse IgG antibody (Chemicon International, Inc., Temecula, CA).

Construction of the MGMT-Apn1 chimeric molecule in the pcDNA 3.0 vector

The MGMT-Apn1 chimeric construct was synthesized using the overlapping thermocycle amplification technique as previously described.3 Oligonucleotides were used as primers to amplify human MGMT and S. cerevisiae Apn1 cDNA from bacterial clone templates. The 5′ MGMT primer (IndexTermIndexTermCCGGAATTCATGGACAAGGATTGT) and 3′ Apn1 primer included an EcoRI restriction site and an XhoI restriction site, respectively, so that the PCR product could be subcloned into the pcDNA3.0 vector. The 3′ MGMT primer (IndexTermIndexTermCAAAGCTAGGTGTCGAAGGGTTTCGGCCAGCAGGCGG) was designed so that the resulting PCR product included the first 20 base pairs of the Apn1 cDNA sequence. The 5′ Apn1 primer (IndexTermIndexTermCCGCCTGCTGGCCGAAACCCTTCGACACCTAGCTTTG) was designed so that the resulting PCR product included the final 20 bp of the MGMT cDNA at the 5′ end. PCR (1 × Mg2+-free Gibco Platinum buffer, 1.5 mM MgSO4, 0.1 mM dNTPs, 15 pmol each of 5′ primer and 3′ primer, 1 U Platinum DNA polymerase (Life Technologies, Rockville, MD)) was performed in a thermocycler (MJ Research, Inc., Watertown, MA) for 30 cycles. The PCR products were subjected to electrophoresis in a 1% agarose gel and the 644 bp MGMT and the 1123 bp Apn1 fragments were excised. The DNA fragments were purified of agarose and ethidium bromide using the Gene Clean kit (QBIOgene, Carlsbad, CA) and combined with the 5′ MGMT primer and the 3′ Apn1 primer in another PCR reaction (95°C for 10 minutes, 72°C for 3 minutes, 94°C for 30 s, 60°C for 1 minutes, 72°C for 3 minutes for 30 cycles to step 3 and a final 72°C for 10 minutes) to yield the 1.75-kb MGMT-Apn1 cDNA fragment. The fragment was gel-purified as before, subjected to double restriction digest with EcoRI and XhoI, and ligated into the EcoRI/XhoI site of the pcDNA 3.0 neo vector (Invitrogen, Carlsbad, CA) using T4 DNA ligase (Takara, Madison, WI). DH5α competent cells were transformed with the ligation product and colonies containing the MGMT-Apn1 construct were confirmed by PCR and restriction endonuclease digestion. DNA sequencing was performed to confirm the integrity of the MGMT and Apn1 sequences.

Transfection and culture of K562 cells

Purified plasmid DNA (Qiagen, Chatsworth, CA) was used to transfect K562 cells using the Lipofectin transfection reagent (Life Technologies, Rockville, MD) as per the manufacturer's protocol. Individual clones were selected using 0.75 mg/ml G418 (Life Technologies, Rockville, MD), and expression of the MGMT-Apn1 fusion protein was characterized using Northern and Western blot analyses. Cells were maintained at 37°C, 5% CO2 in RPMI media supplemented with 5% fetal bovine serum, 100 U/ml penicillin G sodium, and 100 μg/ml streptomycin sulfate (Life Technologies, Rockville, MD).

O6-methylguanine DNA methyltransferase activity

MGMT activity using the HEX-labeled 18-mer oligonucleotide assay was performed as previously described.3,33 Briefly, cells were collected, washed in 1 × PBS, resuspended in 0.5 ml chilled assay buffer (50 mM Tris (pH 7.5), 5% glycerol, 0.5 mM EDTA, 1 mM DTT) and lysed by sonication. Whole cell lysate was obtained by recovering the supernatant following centrifugation at 14,000 × g for 30 minutes at 4°C, and 50 μg of protein was incubated with 0.2 pmol of the HEX-labeled 18-mer for 2 hours in a 37°C water bath (150 μl reaction volume). Following phenol–chloroform extraction and ethanol precipitation, the reaction was subjected to PvuII digestion for 2 hours in a 37°C water bath (20 μl reaction volume). The restriction digest product was denatured in the presence of 33% formamide and 0.3 mM EDTA by incubating at 100°C for 2 minutes. Reaction products were separated by electrophoresis on a 33% polyacrylamide gel containing 7 M urea and visualized using the Hitachi FmBio II Flourescence Imaging System (Hitachi Genetic Systems, South San Francisco, CA). The presence of an 8-mer following electrophoresis indicates that PvuII was able to recognize and cleave the 18-mer following repair of the methylated guanine by MGMT activity present in the cell extract.

Apurinic endonuclease activity

AP activity was detected using the 26-mer oligonucleotide assay as previously described.3,33 Briefly, the AP endonuclease assay utilized a 26-bp 5′ 32P-end-labeled or fluorometric duplex DNA molecule that contains a tetrahydrofuranyl residue, a synthetic AP site. The reaction (10μl) contained 50 mM HEPES (pH 7.5), 50 mM KCl, 1 μg/ml bovine serum albumin, 0.05% Triton X-100, 20 mM EDTA, and 5 μg total protein from the cell lysate of interest, and was incubated in a 37°C water bath for 15 minutes. The reaction products were separated by electrophoresis on a 20% polyacrylamide gel containing 7 M urea. The gel was dried and subjected to autoradiography for visualization.

Cell survival assays

K562 cells expressing the MGMT-Apn1 fusion protein were plated into each well of a six-well plate (Corning Costar, Cambridge, MA) at a density of 50,000 cells/ml and cultured for 1 hour at 37° at 5% CO2. The cells were then treated with the specified drug(s) (see Results) and cultured for 6 days at 37°, 5%CO2. Six days later, the viability of the cells was determined using trypan blue stain and compared with untreated cells and vector-transfected cells. Experiments were performed in triplicate and repeated three times. Statistical analysis was performed using SigmaStat (Jandel Scientific) software package (t-test and ANOVA).


  1. 1

    Hansen WK, Kelley MR . Review of mammalian DNA repair and translational implications. J Pharmacol Exp Ther. 2000;295:1–9.

  2. 2

    Wadhwa PD, Zielske SP, Roth JC, Ballas CB, Bowman JE, Gerson SL . Cancer gene therapy: scientific basis. Annu Rev Med. 2002;53:437–452.

  3. 3

    Hansen WK, Deutsch WA, Yacoub A, Xu Y, Williams DA, Kelley MR . Creation of a fully functional human chimeric DNA repair protein. Combining O6-methylguanine DNA methyltransferase (MGMT) and AP endonuclease (APE/redox effector factor 1 (Ref 1)) DNA repair proteins. J Biol Chem. 1998;273:756–762.

  4. 4

    Limp-Foster M, Kelley MR . DNA repair and gene therapy: implications for translational uses. Environ Mol Mutagen. 2000;35:71–81.

  5. 5

    Jelinek J, Rafferty J, Cmejla R, et al. A novel dual function retrovirus expressing multidrug resistance 1 and O6-alkylguanine-DNA-alkyltransferase for engineering resistance of haemopoietic progenitor cells to multiple chemotherapeutic agents. Gene Ther. 1999;6:1489–1493.

  6. 6

    Frosina G . Overexpression of enzymes that repair endogenous damage to DNA. Eur J Biochem. 2000;267:2135–2149.

  7. 7

    Friedberg EC, Walker GC, Siede W . DNA Repair and Mutagenesis. Washington, DC.: ASM Press; 1995.

  8. 8

    Gerson SL . Clinical relevance of MGMT in the treatment of cancer. J Clin Oncol. 2002;20:2388–2399.

  9. 9

    Gonzaga PE, Potter PM, Niu TQ, et al. Identification of the cross-link between human O6-methylguanine-DNA methyltransferase and chloroethylnitrosourea-treated DNA. Cancer Res. 1992;52:6052–6058.

  10. 10

    Erickson LC, Laurent G, Sharkey NA, Kohn KW . DNA cross-linking and monoadduct repair in nitrosourea-treated human tumour cells. Nature. 1980;288:727–729.

  11. 11

    Robins P, Harris AL, Goldsmith I, Lindahl T . Cross-linking of DNA induced by chloroethylnitrosourea is presented by O6-methylguanine-DNA methyltransferase. Nucleic Acids Res. 1983;11:7743–7758.

  12. 12

    Samson L, Derfler B, Waldstein EA . Suppression of human DNA alkylation-repair defects by Escherichia coli DNA-repair genes. Genetics. 1986;83:5607–5610.

  13. 13

    He YH, Wu M, Kobune M, Xu Y, Kelley MR, Martin WJ . Expression of yeast apurinic/apyrimidinic endonuclease (APN1) protects lung epithelial cells from bleomycin toxicity. Am J Respir Cell Mol Biol. 2001;25:692–698.

  14. 14

    Ramotar D . The apurinic–apyrimidinic endonuclease IV family of DNA repair enzymes. Biochem Cell Biol. 1997;75:327–336.

  15. 15

    Bregni M, Magni M, Siena S, Di Nicola M, Bonadonna G, Gianni AM . Human peripheral blood hematopoietic progenitors are optimal targets of retroviral-mediated gene transfer. Blood. 1992;80:1418–1422.

  16. 16

    Bodine DM, Moritz T, Donahue RE, et al. Long-term in vivo expression of a murine adenosine deaminase gene in rhesus monkey hematopoietic cells of multiple lineages after retroviral mediated gene transfer into CD34+bone marrow cells. Blood. 1993;82:1975–1980.

  17. 17

    Moritz T, Williams DA . Transfer of drug resistance genes to hematopoietic precursors. In: Bertino, J. R. ed. Encyclopedia of Cancer, Vol. III. San Diego: Academic Press; 1997: 1765–1776.

  18. 18

    Dirven HA, Dictus EL, Broeders NL, van Ommen B, van Bladeren PJ . The role of human glutathione S-transferase isoenzymes in the formation of glutathione conjugates of the alkylating cytostatic drug thiotepa. Cancer Res. 1995;55: 1701–1706.

  19. 19

    Rekha GK, Sreetama L, Sladek NE . Intrinsic cellular resistance to oxazaphosphorines exhibited by a human colon carcinoma cell line expressing relatively large amounts of a class-3 aldehyde dehydrogenase. Biochem Pharmacol. 1994;48:1943–1952.

  20. 20

    Ueda K, Cardarelli C, Gottesman MM, Pastan I . Expression of a full-length cDNA for the human “MDR1” gene confers resistance to colchicine, doxorubicin, and vinblastine. Proc Natl Acad Sci USA. 1987;84:3004–3008.

  21. 21

    Wang J, Chen Z, Xia X, Lu D, Xue J, Ruan C . Improvement of combination chemotherapy tolerance by introduction of polycistronic retroviral vector drug resistance genes MGMT and MDR1 into human umbilical cord blood CD34(+) cells. Leuk Res. 2002;26:281–288.

  22. 22

    Suzuki M, Sugimoto Y, Tsuruo T . Efficient protection of cells from the genotoxicity of nitrosoureas by the retrovirus-mediated transfer of human O-6-methylguanine-DNA methyltransferase using bicistronic vectors with human multidrug resistance gene 1. Mutat Res Fundam Mol Mech Mut. 1998;401:133–141.

  23. 23

    Kleibl K, Margison GP . Increasing DNA repair capacity in bone marrow by gene transfer as a prospective tool in cancer therapy. Neoplasma 1998;45:181–186.

  24. 24

    Maze R, Kapur R, Kelley MR, Hansen WK, Oh SY, Williams DA . Reversal of 1,3-bis(2-chloroethyl)-1-nitrosourea-induced severe immunodeficiency by transduction of murine long-lived hemopoietic progenitor cells using O6-methylguanine DNA methyltransferase complementary DNA. J Immunol. 1997;158:1006–1013.

  25. 25

    Maze R, Kurpad C, Pegg AE, Erickson LC, Williams DA . Retroviral-mediated expression of the P140A, but not P140A/G156A, mutant form of O6-methylguanine DNA methyltransferase protects hematopoietic cells against O6-benzylguanine sensitization to chloroethylnitrosourea treatment. J Pharmacol Exp Ther. 1999;290:1467–1474.

  26. 26

    Gerson SL, Miller K, Berger NA . O6 alkylguanine-DNA alkyltransferase activity in human myeloid cells. J Clin Invest. 1985;76:2106–2114.

  27. 27

    Evans AR, Limp-Foster M, Kelley MR . Going APE over ref-1. Mutat Res. 2000;461:83–108.

  28. 28

    Tomicic M, Eschbach E, Kaina B . Expression of yeast but not human apurinic/apyrimidinic endonuclease renders Chinese hamster cells more resistant to DNA damaging agents. Mutat Res. 1997;383:155–165.

  29. 29

    Mol CD, Izumi T, Mitra S, Tainer JA . DNA-bound structures and mutants reveal abasic DNA binding by APE1 and DNA repair coordination [corrected]. Nature. 2000;403:451–456.

  30. 30

    Mol CD, Hosfield DJ, Tainer JA . Abasic site recognition by two apurinic/apyrimidinic endonuclease families in DNA base excision repair: the 3′ ends justify the means. Mutat Res. 2000;460:211–229.

  31. 31

    Tainer JA, Friedberg EC . Dancing with the elephants: envisioning the structural biology of DNA repair pathways. Mutat Res. 2000;460:139–141.

  32. 32

    Daniels DS, Tainer JA . Conserved structural motifs governing the stoichiometric repair of alkylated DNA by O(6)-alkylguanine-DNA alkyltransferase [In Process Citation]. Mutat Res. 2000;460:151–163.

  33. 33

    Kreklau EL, Limp-Foster M, Liu N, Xu Y, Kelley MR, Erickson LC . A novel fluorometric oligonucleotide assay to measure O(6)-methylguanine DNA methyltransferase, methylpurine DNA glycosylase, 8-oxoguanine DNA glycosylase and abasic endonuclease activities: DNA repair status in human breast carcinoma cells overexpressing methylpurine DNA glycosylase. Nucleic Acids Res. 2001;29:2558–2566.

Download references


This work was supported by National Institutes of Health grants CA76643, NS38506, ES05865, ES03456, P01-CA75426, and P30 DK49218 supporting MRK, as well as DOD grants OC00113 and OC990085 to MRK. The Riley Memorial Association also supported these studies. We would also like to acknowledge the help of Emi Kreklau for MGMT activity assistance.

Author information

Correspondence to Mark R Kelley.

Rights and permissions

Reprints and Permissions

About this article


  • Apn1
  • MGMT
  • base excision repair
  • translational

Further reading