Transcriptional tumor-selective promoter targeting of E. coli purine nucleoside phosphorylase for pancreatic cancer suicide gene therapy

J Gene Med. 2005 May;7(5):672-80. doi: 10.1002/jgm.701.


Background: Pancreatic cancer remains a rapidly fatal disease. Suicide gene therapy has been shown to be an effective tool for pancreatic tumor cell destruction, but a cell-specific gene delivery is required to limit host toxicity. The objective of this study was both to design recombinant vectors in which the suicide gene E. coli purine nucleoside phosphorylase (ePNP) is under the control of either CEA or MUC1 promoter sequences and to investigate on experimental pancreatic carcinomas the selective killing effects of the conditional ePNP/prodrug (MePdR) system.

Methods: Transcriptional activities of CEA and MUC1 promoter sequences were analyzed using luciferase reporter gene constructions. Thereafter, recombinant vectors expressing ePNP under control of the most promising pCEA and pMUC1 sequences were designed and used to establish stable tumor cell transfectants from two human pancreatic cell lines, respectively tumor-marker positive (BxPc3) or negative (Panc-1), then applied for in vitro and in vivo experiments.

Results: Transient experiments indicated that CEA and MUC1 promoter sequences confer specificity while preserving high transcriptional activities. The MePdR treatment induced a high in vitro cytotoxicity on the sole CEA- and MUC1-producing cell lines (i.e. BxPc3-CEA and -MUC1/ePNP). In the same way, prodrug treatment induced a significant tumor regression on the sole tumor-marker-positive BxPc3 xenografts, whilst the Panc1-CEA and -MUC1/ePNP tumors were not affected.

Conclusions: These data confirm and extend the antitumor efficacy of the ePNP/MePdR killing system and demonstrate the feasibility of the transcriptional targeting strategy under tumor marker promoter control and thereby a preferential killing of CEA- and MUC1-producing pancreatic tumor cells. Thus, efficient in vivo gene delivery and transcriptional targeting constitute the major future clinical challenge for a selective pancreatic cancer suicide gene strategy.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Animals
  • Carcinoembryonic Antigen / genetics
  • Carcinoembryonic Antigen / metabolism
  • Cell Proliferation
  • Escherichia coli / enzymology*
  • Female
  • Genes, Transgenic, Suicide*
  • Genetic Therapy*
  • Humans
  • Luciferases / metabolism
  • Mice
  • Mice, Nude
  • Mucin-1 / genetics
  • Mucin-1 / metabolism
  • Pancreatic Neoplasms / enzymology
  • Pancreatic Neoplasms / genetics
  • Pancreatic Neoplasms / therapy*
  • Prodrugs / therapeutic use*
  • Promoter Regions, Genetic / genetics*
  • Purine Nucleosides / therapeutic use
  • Purine-Nucleoside Phosphorylase / genetics*
  • Purine-Nucleoside Phosphorylase / metabolism
  • Transcription, Genetic
  • Tumor Cells, Cultured
  • Xenograft Model Antitumor Assays


  • Carcinoembryonic Antigen
  • Mucin-1
  • Prodrugs
  • Purine Nucleosides
  • 6-methylpurine 2'-deoxyriboside
  • Luciferases
  • Purine-Nucleoside Phosphorylase