Applied molecular evolution of O6-benzylguanine-resistant DNA alkyltransferases in human hematopoietic cells

Efficient and stable MGMT-mediated selection of long-term repopulating stem cells in nonhuman primates

HSC transplantation using genetically modified autologous cells is a promising therapeutic strategy for various genetic diseases, cancer, and HIV. However, for many of these conditions, the current efficiency of gene transfer to HSCs is not sufficient for clinical use. The ability to increase the percentage of gene-modified cells following transplantation is critical to overcoming this obstacle. In vivo selection with mutant methylguanine methyltransferase (MGMTP140K) has been proposed to overcome low gene transfer efficiency to HSCs. Previous studies have shown efficient in vivo selection in mice and dogs but only transient selection in primates. Here, we report efficient and stable MGMTP140K-mediated multilineage selection in both macaque and baboon nonhuman primate models. Treatment consisting of both O6-benzylguanine (O6BG) and N,N'-bis(2-chloroethyl)-N-nitroso-urea (BCNU) stably increased the percentage of transgene-expressing cells from a range of initial levels of engrafted genetically modified cells, with the longest follow-up after drug treatment occurring over 2.2 years. Drug treatment was well tolerated, and selection occurred in myeloid, lymphoid, and erythroid cells as well as platelets. Retrovirus integration site analysis before and after drug treatments confirmed the presence of multiple clones. These nonhuman primate studies closely model a clinical setting and should have broad applications for HSC gene therapy targeting human diseases of malignant, genetic, and infectious nature, including HIV.

Long-term polyclonal and multilineage engraftment of methylguanine methyltransferase P140K gene-modified dog hematopoietic cells in primary and secondary recipients

Overexpression of methylguanine methyltransferase P140K (MGMTP140K) has been successfully used for in vivo selection and chemoprotection in mouse and large animal studies, and has promise for autologous and allogeneic gene therapy. We examined the long-term safety of MGMTP140K selection in a clinically relevant dog model. Based on the association of provirus integration and proto-oncogene activation leading to leukemia in the X-linked immunodeficiency trial, we focused our analysis on the distribution of retrovirus integration sites (RIS) relative to proto-oncogene transcription start sites (TSS). We analyzed RIS near proto-oncogene TSS before (n = 157) and after (n = 129) chemotherapy in dogs that received MGMTP140K gene-modified cells and identified no overall increase of RIS near proto-oncogene TSS after chemotherapy. We also wanted to determine whether in vivo selected cells retained fundamental characteristics of hematopoietic stem cells. To that end, we performed secondary transplantation of MGMTP140K gene-modified cells after in vivo selection in dog leukocyte antigen (DLA)-matched dogs. Gene-modified cells achieved multilineage repopulation, and we identified the same gene-modified clone in both dogs more than 800 and 900 days after transplantation. These data suggest that MGMTP140K selection is well tolerated and should allow clinically for selection of gene-corrected cells in genetic or infectious diseases or chemoprotection for treatment of malignancy.

Gene therapy approaches for stem cell protection

Cytotoxic exposure of bone marrow and other non-hematopoietic organs containing self-renewing stem cell populations is associated with damage to the supportive microenvironment. Recent evidence indicates that radical oxygen species resulting from the initial oxidative stress persist for months after ionizing irradiation exposure of tissues including oral cavity, esophagus, lung and bone marrow. Antioxidant gene therapy using manganese superoxide dismutase plasmid liposomes has provided organ-specific radiation protection associated with delay or prevention of acute and late toxicity. Recent evidence has suggested that manganese superoxide dismutase transgene expression in cells of the organ microenvironment contributes significantly to the mechanism of protection. Incorporating this knowledge into designs of novel approaches for stem cell protection is addressed in the present review.

Differential competitive resistance to methylating versus chloroethylating agents among five O6-alkylguanine DNA alkyltransferases in human hematopoietic cells

P140K-MGMT and G156A-MGMT genes encode two O(6)-benzylguanine-resistant O(6)-alkylguanine DNA alkyltransferase proteins that confer a high degree of O(6)-benzylguanine and 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) or O(6)-benzylguanine and temozolomide resistance to primary hematopoietic cells. In this study, we directly compared these and three other O(6)-benzylguanine-resistant MGMT genes for their ability to protect the human erythroleukemia cell line, K562, using a direct competitive selection strategy to identify the mutation that conferred the greatest degree of protection from O(6)-benzylguanine and either BCNU or temozolomide. MFG retroviral vector plasmids for each of these mutants [G156A-MGMT (ED(50) for O(6)-benzylguanine, 60 micromol/L); and P140K-MGMT, MGMT-2 (S152H, A154G, Y158H, G160S, L162V), MGMT-3 (C150Y, A154G, Y158F, L162P, K165R), and MGMT-5 (N157T, Y158H, A170S; ED(50) for benzylguanine, >1,000 micromol/L)] were mixed, and the virus produced from Phoenix cells was transduced into K562 cells. Stringent selection used high doses of O(6)-benzylguanine (800 micromol/L) and temozolomide (1,000 micromol/L) or BCNU (20 micromol/L) administered twice, and following regrowth, surviving clones were isolated, and the MGMT transgene was sequenced. None of the mutants was lost during selection. Using temozolomide, the enrichment factor was greatest for P140K-MGMT (1.7-fold). Using BCNU selection, the greatest enrichment was observed with MGMT-2 (1.5-fold). G156A-MGMT, which is the least O(6)-benzylguanine-resistant MGMT gene of the mutants tested, was not lost during selection but was selected against. The optimal mutant MGMT useful as a drug resistance gene may depend on whether a methylating or chloroethylating agent is used for drug selection.

Polyclonal chemoprotection against temozolomide in a large-animal model of drug resistance gene therapy

Incorporation of drug resistance genes into gene vectors has 2 important roles in stem cell gene therapy: increasing the proportion of gene-corrected cells in vivo (ie, in vivo selection) and marrow protection to permit higher or more tightly spaced doses of chemotherapy in the treatment of malignant diseases. We studied in a clinically relevant canine model of gene therapy the P140K mutant of the drug resistance gene methylguanine methyltransferase (MGMT), which encodes a DNA-repair enzyme that confers resistance to the combination of the MGMT inhibitor O(6)-benzylguanine (O(6)BG) and nitrosourea drugs such as carmustine and methylating agents such as temozolomide. Two dogs received MGMT(P140K)-transduced autologous CD34(+)-selected cells. After stable engraftment, gene marking in granulocytes was between 3% and 16% in the 2 animals, respectively. Repeated administration of O(6)BG and temozolomide resulted in a multilineage increase in gene-modified repopulating cells with marking levels of greater than 98% in granulocytes. MGMT(P140K) overexpression prevented the substantial myelosuppression normally associated with this drug combination. Importantly, hematopoiesis remained polyclonal throughout the course of the study. Extrahematopoietic toxicity was minimal, and no signs of myelodysplasia or leukemia were detected. These large-animal data support the evaluation of MGMT(P140K) in conjunction with O(6)BG and temozolomide in clinical trials.

Directed enzyme evolution and selections for catalysis based on product formation

Enzyme engineering by molecular modelling and site-directed mutagenesis can be remarkably efficient. Directed enzyme evolution appears as a more general strategy for the isolation of catalysts as it can be applied to most chemical reactions in aqueous solutions. Selections, as opposed to screening, allow the simultaneous analysis of protein properties for sets of up to about 10(14) different proteins. These approaches for the parallel processing of molecular information 'Is the protein a catalyst?' are reviewed here in the case of selections based on the formation of a specific reaction product. Several questions are addressed about in vivo and in vitro selections for catalysis reported in the literature. Can the selection system be extended to other types of enzymes? Does the selection control regio- and stereo-selectivity? Does the selection allow the isolation of enzymes with an efficient turnover? How should substrates be substituted or mimicked for the design of efficient selections while minimising the number of chemical synthesis steps? Engineering sections provide also some clues to design selections or to circumvent selection biases. A special emphasis is put on the comparison of in vivo and in vitro selections for catalysis.

Transfer of drug resistance genes in hematopoietic progenitors for chemoprotection: is it still an option?

For numerous malignancies a relationship between the intensity of antineoplastic chemotherapy and tumor response has been demonstrated. Myelotoxicity is the main cause of chemotherapy-associated morbidity and of treatment delays. The concept of myeloprotective cytostatic drug resistance gene transfer to normal hematopoietic stem cells (HSC) therefore sparks great enthusiasm. While initial studies using murine retroviral vectors on murine HSC showed that the concept works, a number of clinical studies in the last decade were not informative because of limitations in transduction efficiency and transgene expression.Furthermore, possible side effects such as unforeseen transgene activity and vector integration-based leukemogenesis have been reported. Among others, these developments raised some scepticism against the feasibility of myeloprotective gene transfer. Recently, considerable improvements have been achieved in vector design, HSC manipulation, selection protocols and risk assessment methods which are discussed in detail here. Based on these experimental studies successful clinical trials can now be anticipated.

Advances in gene transfer into haematopoietic stem cells by adenoviral vectors

Until recently, the cells of haematopoietic origin were not considered good adenoviral (Adv) targets, primarily because they lacked the specific Adv receptors required for productive and efficient Adv infections. In addition, because of limitations inherent in Adv infections, such as short-term expression and a non-integrating nature, their application has been precluded from haematopoietic stem cell (HSC) and bone marrow transduction protocols where long-term expression has been required. Therefore, limited research utilising Adv-mediated gene transfer into haematopoietic cells had been conducted. With recent insights into the critical interactions between adenovirus (Adv) and cells, new Adv-mediated gene transduction strategies have now been reported that may overcome these limitations. These new strategies include Adv possessing synthetic polymer coatings, genetically modified capsid proteins or antibody-redirected fibres that can efficiently redirect and retarget Adv to transfer genes into HSC. Additionally, new hybrid Advs, engineered with both modified capsid proteins and novel cis-acting integration sequences, are also being developed which can efficiently deliver and integrate Adv delivered genes into HSC. This is an area of research that is now rapidly gaining momentum in terms of techniques and applications. Here we review the current status of adenovirus-based vectors as a means to achieve high-level gene transfer into haematopoietic cell types.

Cancer gene therapy: scientific basis

Gene therapy of cancer has been one of the most exciting and elusive areas of therapeutic research in the past decade. Critical developments have occurred in gene therapy targeting cancer cells, cancer vasculature, the immune system, and the bone marrow, itself often the target for severe toxicity from therapeutic agents. We review some recent developments in the field. In each instance, clear preclinical models validated the therapeutic approach and efforts have been made to evaluate the target impact in both preclinical and early clinical trials. Although no cures can consistently be expected from today's cancer gene therapy, the rapid progress may imply that such cures are a few short years away.

Clinical relevance of MGMT in the treatment of cancer

Anumber of DNA-damaging chemotherapeutic agents attack the O(6) position on guanine, forming the most potent cytotoxic DNA adducts known. The DNA repair enzyme O(6)-alkylguanine DNA alkyltransferase (AGT), encoded by the gene MGMT, repairs alkylation at this site and is responsible for protecting both tumor and normal cells from these agents. Cells and tissues vary greatly in AGT expression, not only between tissues but also between individuals. AGT activity correlates inversely with sensitivity to agents that form O(6)-alkylguanine DNA adducts, such as carmustine (BCNU), temozolomide, streptozotocin, and dacarbazine. The one exception is those tumors lacking mismatch repair, which renders them resistant to methylating agents. A recent study in patients with gliomas confirmed the correlation between low-level expression of the MGMT gene and response and survival after BCNU. An inhibitor to AGT, O(6)-benzylguanine (BG), depletes AGT in human tumors without associated toxicity and is now in phase II clinical trials. Finally, mutations within the active site region of the MGMT gene render the AGT protein resistant to BG inactivation. As a result, mutant MGMT gene transfer into hematopoietic stem cells has been shown to selectively protect the marrow from the combination of an alkylating agent and BG, while at the same time sensitizing tumor cells. MGMT remains a paradigm for development of new agents that modulate known mechanisms of drug resistance in cancer cells and raise the spectra of combinatorial therapies that encompass known drug resistance mechanisms.

Interaction of mammalian O(6)-alkylguanine-DNA alkyltransferases with O(6)-benzylguanine

Human O(6)-alkylguanine-DNA alkyltransferase (hAGT) activity is a major factor in providing resistance to cancer chemotherapeutic alkylating agents. Inactivation of hAGT by O(6)-benzylguanine (BG) is a promising strategy for overcoming this resistance. Previous studies, which have focused on the region encompassed by residues Pro138 to Gly173, have identified more than 100 individual mutations located at 23 discrete sites at which alterations can render AGT less sensitive to BG. We have now extended the examination of possible sites in hAGT at which alterations might lead to BG resistance to include the residues from Val130 to Asn137, which also make up part of the binding pocket into which BG is postulated to fit. A further 21 mutations located at positions Gly132, Met134, Arg135, and Gly136 were found to lower sensitivity to BG. Mutants R135L, R135Y, and G136P were the most strikingly resistant, with a 50-fold increase in the amount of BG needed to obtain 50% inactivation. These results therefore increase the number of sites at which BG resistance can occur in response to a single amino acid change to 27. Although mammalian AGTs are very similar in amino acid sequence, mouse AGT (mAGT) is significantly less sensitive to BG than rat AGT (rAGT) or hAGT. Construction of chimeric proteins in which portions came from the rAGT and the mAGT indicated that the difference in inactivation resided solely in the amino acids located in the sequence from residues 150 to 188. Individual mutations of the three residues where rAGT and mAGT differ in this region showed that the principal reason for the reduced ability of the mAGT to react with BG was the presence of a histidine residue at position 161, which is occupied by asparagine in rAGT and hAGT. These experiments indicate that many minor changes in amino acids forming all parts of the nucleoside binding pocket of AGT can alter its ability to react with BG and that the possibility that polymorphisms or variants may occur reducing the effectiveness of combination therapy with BG and alkylating agents must be considered.