New enzyme for reductive cancer chemotherapy, YieF, and its improvement by directed evolution Yoram Barak, 1 Stephen H. Thorne, 1,2 David F. Ackerley, 1 Susan V. Lynch, 1 Christopher H. Contag, 1,2 and A. Matin 1 Departments of 1 Microbiology and Immunology and 2 Pediatrics and Molecular Imaging Program, Stanford University School of Medicine, Stanford, California Abstract Reductive prodrugs, mitomycin C and 5-aziridinyl-2,4- dinitrobenzamide (CB 1954), are nontoxic in their native form but become highly toxic upon reduction. Their effectiveness in cancer chemotherapy can be enhanced by delivering to tumors enzymes with improved prodrug reduction kinetics. We report the discovery of a new prodrug-reducing enzyme, YieF, from Escherichia coli , and the improvement of its kinetics for reducing mitomy- cin C and CB 1954. A YieF-derived enzyme, Y6, killed HeLa spinner cells with z5-fold efficiency than the wild- type enzymes, YieF and NfsA, at a variety of drug and enzyme concentrations and incubation times. With ad- hered HeLa cells and Salmonella typhimurium SL 7838 bacteria as enzyme delivery vehicle, at least an order of magnitude less of Y6-producing bacteria were required to kill >90% of tumor cells compared with bacteria expressing the wild-type enzymes, which at a comparable level killed <5% of the cells. Thus, Y6 is a promising enzyme for use in cancer chemotherapy, and Salmonella strain SL 7838, which specifically targets tumors, may be used to deliver the prodrug-activating enzymes to tumors. [Mol Cancer Ther 2006;5(1):97 – 103] Introduction Mitomycin C and 5-aziridinyl-2,4-dinitrobenzamide (CB 1954), which are reductive prodrugs, are nontoxic in their native form but produce a highly toxic species when reduced. These drugs kill by generating DNA adducts and can target both growing and nongrowing tumor cells, which is advantageous because in human tumors, gener- ally only a small fraction of cells are actively replicating at a given time (1, 2). Reductive prodrug cancer chemotherapy owes its rationale to the fact that the concentration of the enzymes that reduce them, such as mammalian DT- diaphorase (NQO1; refs. 3 and 4), increases in tumor cells. This makes the tumor cells more potent reducers of these drugs and, therefore, more susceptible to their killing effect. However, these enzymes are present also in normal mammalian cells and although their activity is lower in such cells than in tumor cells, it is high enough to produce severe side effects. A potential solution is to use suicide gene therapy (also known as gene-delivered enzyme prodrug therapy; ref. 5), which aims to deliver genes to tumors to selectively enhance their prodrug reducing activity. A class of enzymes that has been well studied in gene-delivered enzyme prodrug therapy is bacterial nitro- reductases, such as NfsA and NfsB from Escherichia coli (6 – 8). These enzymes can reduce several nitrosubstituted organic compounds, including CB 1954, which is an antitumor reductive prodrug (2). Previous studies of CB 1954/bacterial nitroreductase in gene-delivered enzyme prodrug therapy have indicated the need for improving the specific activity of bacterial nitroreductases for CB 1954 (7). In our ongoing work on bioengineering of bacteria with superior capacity to remediate Cr(VI) (chromate), which is a widespread carcinogenic environmental pollutant, we identified a new E. coli enzyme, YieF, which is particularly attractive toward this end (9, 10). By applying directed evolution through the technique of error-prone PCR, we have improved the capacity of this enzyme to reduce chromate by over 200-fold. 3 Based on the general characteristics of YieF (11), we tested for and confirmed that this enzyme can also reduce CB 1954 as well as another prodrug, mitomycin C. Furthermore, we discov- ered that improvement in chromate reductase activity of the evolved enzymes also led to improvement in their capacity to reduce the two prodrugs. Here, we report that compared with the wild-type YieF, our evolved enzymes exhibit >5-fold increased capacity to kill HeLa cells. We also show that this enzyme can be effectively delivered to tumor cells using an attenuated strain of Salmonella typhimurium (SL 7838). These bacteria target both the aerobic and anaerobic zones of tumors and do not infect normal tissues. 4 Received 9/12/05; revised 10/13/05; accepted 10/25/05. Grant support: Grants DE-FG03-97ER-624940 and DE-FG02-96ER20228 from the Natural and Accelerated Bioremediation Program of the U.S. Department of Energy; Small Animal Imaging Research Program (R24) grant CA 92862; Lady Davis postdoctoral fellowship (Y. Barak); STAX0101 fellowship from Foundation for Research, Science, and Technology New Zealand (D.F. Ackerley); and Stanford Medical School Dean’s Fellowship (S.V. Lynch). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Requests for reprints: A. Matin, Department of Microbiology and Immunology, Stanford University School of Medicine, Sherman Fairchild Science Building, 299 Campus Drive, Stanford, CA 94305. Phone: 650-725-4745; Fax: 650-725-6757. E-mail: a.matin@stanford.edu Copyright C 2006 American Association for Cancer Research. doi:10.1158/1535-7163.MCT-05-0365 3 Y. Barak, D.F. Ackerley, B. Lal, A. Matin, unpublished data. 4 S.H. Thorne, J. Hardy, B. Stoker, C.H. Contag, D.H. Kirn, unpublished data. 97 Mol Cancer Ther 2006;5(1). January 2006 Downloaded from http://aacrjournals.org/mct/article-pdf/5/1/97/1871140/97.pdf by guest on 09 June 2022