Multidrug-Resistant Gene (MDR)

Multidrug-resistant protein 1 (MDR-1) was encoded by the human ABCB1 gene, also called the MDR1 gene. The protein product is P-glycoprotein and belongs to ATP-binding cassette transporters, subfamily B member 1. It is a 170-kDa transmembrane protein and acts as an ATP-dependent efflux pump to remove cytotoxic drugs from cytoplasm ( Figure 26.1 ). It is multidrug resistance because it has broad substrate specificity, and it can decrease the cytosolic accumulation of vinca alkaloids, taxanes, epipodophyllotoxins, and anthracyclines . In various types of cancer, such as breast cancer, lung cancer, gastric carcinoma, and chronic lymphocytic leukemia, overexpression of MDR1 has been linked to multidrug resistance in vitro and treatment resistance in clinics . Because of its effectiveness in eliminating various types of chemotherapeutic drugs from cell cytoplasm, the ABCB1 gene is one of the earliest genes considered for drug-resistance gene therapy. The strategy is to transfer the ABCB1 gene into normal hematopoietic stem or progenitor cells, allowing those gene-modified cells and their progenies to overexpress MDR1 protein and become more resistant to chemo drugs than cancer cells during treatment ( Figure 26.1 ).

Mechanism of multidrug-resistance protein 1 (MDR1).

After adding alkylating agents to normal cells, there is not enough MDR1 to export those agents, causing DNA damage and eventually leading to apoptosis. On the other hand, overexpression of MDR1 on the cell surface can prevent the cytosolic accumulation of cytotoxic agents, resulting in cell survival.

Soon after the link between multidrug resistance and MDR1 was discovered, Guild et al . used retroviral vector to introduce murine mdr cDNA into NIH 3T3 cells to show that transfected cells were highly resistant to colchicine, vinblastine, or doxorubicin even with one copy of the vector after colchicine selection or two copies of the vector without prior selection . Similar studies with human MDR1 cDNA were performed in bone marrow cells of the DBA/2J mice . A high percentage of transduced bone marrow progenitor cells showed resistance to both colchicine and vinblastine treatments by an in vitro colony-forming assay. The first i n vivo selection of drug-resistant bone marrow cells after retroviral gene transfer of human MDR1 was done in 1992 . Human MDR1 cDNA was subsequently transferred into human CD34 ⁺ cells from cord blood or bone marrows and showed enrichment after chemoselection and improvement of chemotherapy tolerance , paving the way for human MDR1 clinical trials. Many drug-resistant cancer cells also have overexpressed P-glycoprotein, which can be blocked by trans -(E)-flupentixol, an inhibitor of P-glycoprotein, and become sensitive to cancer drugs again. A mutant form of P-glycoprotein (MDR1-F983A) was developed to protect hematopoietic cells from chemotherapeutic drugs in the presence of trans -(E)-flupentixol . In recent studies, lentiviral vectors have been used to efficiently transfer MDR1 to human CD34 ⁺ hematopoietic progenitor cells and protect them against radiotherapy and chemotherapy . MDR1 has also been introduced to keratinocyte stem cells by lentiviral vector for the selection of transduced cells and chemoprotection against topical application of colchicine in the treatment of skin disease .

O ⁶ -Methylguanine Methyltransferase (MGMT)

One of the most popular and successful drug-resistance genes is O ⁶ -methylguanine methyltransferase, also known as alkylguanine DNA-alkyltransferase (AGT). It is a 21-kDa DNA repair protein, encoded by human MGMT gene, that is mainly responsible for repairing DNA damage caused by alkylating agents in a suicide reaction . Alkylating agents, such as temozolomide (TMZ) and 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU), are often used as chemotherapeutic agents in cancer treatment because they add an alkyl group to the O ⁶ position of guanine, which is a highly genotoxic lesion, resulting in mutation, double strain break, and interstrand cross-link . Endogenous MGMT protein can irreversibly transfer the alkyl group from alkylated guanine to a cysteine residue on itself and undergo the degradation process. Various types of cancer cells have increased the level of MGMT protein to make themselves more resistant to the treatment of alkylating agents . The methylation status of MGMT promoter has been studied in patients as a potential predictive marker for prognosis and response to alkylating treatment in various types of cancer, particularly in high-grade glioblastoma . There is also evidence linking TMZ resistance to cancer stem cells in gliomas . Therefore, treatment that targets endogenous MGMT has been suggested as a valid approach to enhance chemo treatment in treatment-resistant patients . To resensitize resistant cancer cells, an MGMT inhibitor, O ⁶ -benzylguanine (BG), is often used to permanently inactivate wild-type MGMT protein in cancer cells to impair their cellular repair mechanism against alkylating agents . However, in a recent clinical trial, it was found that different TMZ-resistant high-grade gliomas may respond differently to O ⁶ -BG treatment. Anaplastic glioma was resensitized to TMZ after treatment with O ⁶ -BG, whereas glioblastoma multiforme was not, indicating different or additional TMZ-resistant mechanisms .

One of the most severe side effects of cancer chemotherapy is myelosuppression. In the early days of MGMT gene therapy, because of its myeloprotection, whole length MGMT cDNA was introduced to hematopoietic cells of mouse and human . However, in the application of drug-resistance gene therapy for cancer, to provide a unique drug-resistance mechanism in normal cells that is different from that in cancer cells and to reduce resistance in cancer cells, mutant forms of MGMT were screened and developed , and the two most effective mutations are P140K and G156A (proline-to-lysine mutation at amino acid 140 and glycine-to-alanine mutation at amino acid 156). Both mutations are shown to be highly resistant to BG-mediated degradation but are still functional in repairing DNA damage caused by alkylating agents . After introducing mutant forms of MGMT to normal stem and progenitor cells by gene transfer, a combination treatment of BG and alkylating agents will inactivate endogenous MGMT protein inside cancer cells and make them more sensitive to alkylating agents. At the same time, functional mutant MGMT can repair damaged DNA by removing DNA adducts and protect gene-modified cells against alkylating agents, allowing an in vivo selection of transduced cells ( Figure 26.2 )—an approach first characterized by Gerson and coworkers .

Mechanism of drug resistance by MGMT(P140K) and strategy for enrichment of transduced cells.

(A) Mechanism of drug resistance. (a) Wild-type MGMT is a DNA repair protein that is able to remove alkyl adduct on the O ⁶ position of guanine on damaged DNA. This irreversible repair process will lead WT-MGMT to degradation. (b) O ⁶ -benzylguanine is an inhibitor of WT-MGMT, and it can bind to MGMT, leading to degradation without the repairing of the damaged DNA. (c) Mutant MGMT, such as MGMT(P140K), will not bind to O ⁶ -BG, but it can still repair damaged DNA. (B) In vivo selection strategy. By adding O ⁶ -benzylguanine first before alkylating agents, wild-type MGMT will be rapidly depleted from both transduced and nontransduced cells. However, only transduced cells were able to sustain the subsequent challenge of alkylating agents, leaving a higher percentage of gene-modified cells.

Preclinical studies with mutant MGMT have been performed extensively in the past two decades. Stable gene transfer into both murine and human hematopoietic cells was achieved by using retroviruses and resulted in resistance against O ⁶ -BG and BCNU treatment . In vivo enrichment of transduced cells under drug selection and chemoprotection against the alkylating agents BCNU and TMZ was first investigated in murine models . It has been shown to protect mice against lethal treatment of BCNU, and it provided significant enrichment of transduced cells after O ⁶ -BG and BCNU or TMZ treatments in non-myeloablated recipients. In one study, as high as 93% transgene-positive cells were achieved in mice from the initial 12% at the time of transduction and transplantation through in vivo drug selection in primary followed by secondary recipients, indicating reconstitution of hematopoiesis from gene-modified hematopoietic stem cells .

In vivo selection of gene-modified cells has also been studied to enhance treatment of murine β-thalassemia, murine protoporphyria, low-frequency lung engraftment from transduced bone marrow cells, and murine model of hemophilia . In recent small animal studies, a novel imaging technology provided us the tool to monitor bone marrow transplantation and in vivo drug selection with mutant MGMT . From these imaging experiments, there is clear evidence of dynamic engraftment of transduced hematopoietic stem cells after in vivo BG+BCNU selection ( Figure 26.3 ). Not only was focal long-term gene expression identified, and increased after drug treatment, but also individual areas of hematopoietic stem cell engraftment and expansion were observed for up to 9 months at individual sites, suggesting a stochastic rather than systemic process of hematopoiesis under these conditions.

Bioluminescence imaging (BLI) showing the dynamic process of in vivo selection of MGMT(P140K)-modified bone marrow cells.

Nonmyeloablated recipients received different numbers of MGMT(P140K) lentiviral transduced murine bone marrow cells. Mice A and B both received 1×10 ⁵ transduced bone marrow cells, whereas mouse C received 6×10 ⁵ transduced cells. For mouse A, BG+BCNU treatments were given on days 23 and 50 after bone marrow transplantation (T1+5 means 5 days after first treatment). After BG+BCNU treatment, initial BLI foci, indicated by green arrows, rapidly expanded and engrafted. However, mouse B never developed full engraftment in the absent of BG+BCNU treatment, as did mouse C, which received six times more transduced bone marrow cells.

However, there was a huge disparity between successful gene therapy study in mice and its application in humans. It has been difficult to achieve sufficient gene transferring in human trials to observe its clinical benefit. Large animal and nonhuman primate studies were needed to pave the way for successful clinical trials. Successful MGMT gene transfer and myeloprotection have been reported in a canine model and nonhuman primates . In several of these studies, stable and multilineage selections in hematopoietic system have been achieved through several-round BG/TMZ or BG/BCNU treatments with a longest observation time of 2.2 years.

To expand drug-resistance capability against various types of treatments, combinational drug resistance was achieved by gene transferring two different mechanisms of drug resistance into the same targeted cell. MDR1 and mutant MGMT have been integrated into one bicistronic vector to provide modified cells with protection against BG/ACNU or BG/TMZ and vincristine or paclitaxel treatments . Cells transduced with dual drug-resistance genes showed more resistance to combination treatments and better selection outcome.

Cytidine Deaminase (CDA or CDD)

Cytidine deaminase is a key enzyme involved in the pyrimidine salvage pathway and maintains the cellular pyrimidine supply. It is encoded by the human CDA gene. In the 1960s, cytosine nucleoside analogs, such as cytarabine, were reported to contain antitumor effects. Cytidine analogs, such as azacitidine, can also induce cell differentiation and change methylation status of newly synthesized DNA . Nucleoside analogs are mainly used to treat acute myeloid leukemia and non-Hodgkin’s lymphoma . However, an elevated level of cytidine deaminase has been identified as one of the contributing factors in cellular resistance to nucleoside analogs ; thus, CDA was selected as a candidate for drug-resistance gene therapy.

In early studies, human CDA was introduced into murine fibroblast and hematopoietic cells by retroviral gene transfer, and the results conferred enhanced resistance to nucleoside analogs . Transplantation of gene-modified bone marrow cells in lethally irradiated recipients showed long-term expression of CDA; however, ex vivo selection with nucleoside analogs has been challenging due to the release of CDA in the culture media .

In recent studies, in vitro selection of human hematopoietic progenitor cells and protected myelopoiesis by gene transferring of CDA in murine transplantation model have been reported . However, lymphotoxicity with the overexpression of CDA was also reported. To minimize the side effect of overexpressed CDA, Lachmann et al . generated an inducible transgene expression system to regulate CDA expression and showed robust myeloprotection with minimal lymphotoxicity .

Dihydrofolate Reductase (DHFR)

Dihydrofolate reductase is an enzyme that converts dihydrofolate to tetrahydrofolate and is involved in purines and thymidylate synthesis. It is encoded by the human DHFR gene. Antifolate drugs, methotrexate (MTX) and trimetrexate, can tightly bind to DHFR and inhibit DNA synthesis and cell proliferation. On that account, antifolate drugs have been used as potent antitumor drugs. A recent study showed that MTX can also induce oxidative DNA damage, and this kind of treatment is particularly detrimental to MSH2 defective cancers, such as hereditary non-polyposis colon cancer . However, cancer cells can develop resistance to MTX through several mechanisms, such as DHFR overexpression, altered drug transporter, and DHFR promoter polymorphism . Mutant forms of murine and human DHFR has been discovered in methotrexate-resistant cancer cells .

One of the major side effects of methotrexate treatment is also myelosuppression, which makes mutant DHFR another candidate for drug-resistance gene therapy. During the early phase of gene therapy, DHFR was used as a model system for myeloprotection and in vivo selection. It has been successfully introduced into murine, canine, and human hematopoietic cells and showed protection against MTX cytotoxicity . One of the more potent mutations in DHFR that confers most resistance is a Leu-to-Tyr mutation at the codon number 22, particularly when used with trimetrexate treatment . When transplanting DHFR modified murine bone marrow cells to recipients with only mild preconditioning, even low-level engraftment (1%) of transduced murine bone marrow cells can confer methotrexate resistance after MTX treatment . In addition to mutant DHFR gene transfer, efficient DHFR-mediated selection may also require the treatment of MTX or TMTX with a nucleoside transport inhibitor, nitrobenzylmercaptopurine ribose phosphate . In recent studies, mutant DHFR gene transfer effectively protected hematopoietic stem cells and facilitated the combined treatment of trimetrexate and anti-CD137 immunotherapy for large solid tumors, and similar in vivo selection was also observed in human embryonic stem cell-derived cells when transplanted into NSG mice and treated with methotrexate .

Glutathione S -Transferase (GSTP1 and MGST2)

The glutathione S -transferase superfamily proteins can detoxify many endogenous compounds and break down xenobiotic substrates through conjugation of reduced glutathione with various substrates. GST proteins exist throughout the cells, from cytosol to mitochondria to microsomes. GST-pi (GSTP1) has been shown to be resistant to cisplatin, BCNU, cyclophosphamide, busulfan, and melphalan in cancer cell lines . Subsequently, GST-pi overexpression has been linked to treatment resistance in lung cancer, acute myeloid leukemia, head and neck cancer, and breast cancers .

In drug-resistance gene therapy studies, GST-pi transduced human CD34 ⁺ hematopoietic cells have been shown to be resistant to alkylating agents, doxorubicin, cisplatin, melphalan, and cyclophosphamide . Recently, overexpression of MGST2 and treatment of busulfan have been shown to increase cell survival in the fibroblast cell line. Because MGST2 was not highly expressed in murine bone marrow CD34 ⁺ cells, it can be used as a potential drug-resistance gene in hematopoietic stem cell gene therapy .

Hypoxanthine Phosphoribosyltransferase (HPRT)

Hypoxanthine phosphoribosyltransferase is a human enzyme involved in the purine salvage pathway. It recycles guanine to guanosine monophosphate during DNA degradation. It is encoded by the human HPRT1 gene and has been widely studied since the 1960s. The deficiency of HPRT in humans has been linked to Lesch–Nyhan syndrome and gout due to a high level of uric acid in the blood . Ultraviolet and radiotherapy have been shown to increase the HPRT mutation frequency in cancer cells , which used the mutation to develop resistance to 6-thioguanine, an antimetabolite for treating cancer . Interestingly, a recent study has also shown that 6-thioguanine resistance in melanoma cells is linked to increased levels of MGMT .

As a nucleotide analog, 6-thioguanine is relatively less toxic. Thus, knockdown HPRT gene expression in combination with 6-thioguanine can be a potential in vivo selection strategy. Porter and DeGregori demonstrated the use of a lentiviral vector, containing interfering RNA against hypoxanthine phosphoribosyltransferase, to transduce murine bone marrow cells and transplant them into recipients. After 6-thioguanine treatments in recipients, they observed increased percentages of transduced peripheral blood cells and hematopoietic progenitor cells in the bone marrow compared with control or untransduced cells. Thus, they suggested that interfering RNA-mediated purine analog resistance (iPAR) plus 6-thioguanine can be used for in vivo selection . Hacke et al . showed that 6-thioguanine alone could be used as a non-myeloablative preconditioning and in vivo selection agent in HPRT-deficient murine bone marrow cell transplantation . By using 6-thioguanine alone, they were able to achieve consistently 95% of engraftment from HPRT-deficient donor bone marrow cells, resulting in normal hematopoiesis in both primary and secondary recipients.