641. Cancer Detection Via Expression-Targeted Gene Delivery in Transitional Cell Carcinoma

Expression-targeted gene delivery was used to detect cancer cells by delivering exogenous secretory reporter genes into them under the control of cancer-specific promoters. Almost all cancer biomarkers currently in use are secreted proteins or cell surface proteins. Our strategy takes advantage of cancer-specific promoters, even those associated with exons encoding proteins that are not secreted. Plasmids that coupled such promoters with exons encoding secretable (and excretable) reporters were constructed.This strategy has yielded relatively high reporter expression in cancer cells as compared to normal cells and tissues, both in vitro in cell supernatants and in vivo, as detected in mouse urine. This technology could be used eventually for early cancer detection or the detection of cancer cells remaining following tumorectomy.Gaussia luciferase (LUC) is a secreted protein that has been used for monitoring bioprocesses. Plasmids carrying luc, driven by the promoters for cyclooxygenase type 2 (Pcox2), osteopontin (Popn), or the universally strong cytomegalovirus promoter (Pcmv) (used as a positive control) were constructed and delivered to cells using the branched polycation poly(ethyleneimine). The empty pUC19 vector was used as a negative control. The reporter levels detected in cell supernatants via a Gaussia luciferase assay showed that Pcox2- or Popn-driven luc expression was significantly higher in the bladder cancer cell line MB49 versus normal murine fibroblasts. In vivo experiments in an orthotopic model of transitional cell carcinoma showed that delivery of Pcox2-luc into tumor-bearing mice yielded detectable LUC excretion that was significantly higher versus negative controls.Interestingly, tumor-bearing mice excreting LUC did so with a 24-hour periodicity. This seemingly circadian phenomenon was counterintuitive. The implanted cells did not have synchronous cell cycles upon instillation 8 days prior, and had not developed organized vascular networks (data not shown). Urine concentration also was not seen as a factor, since total LUC per sample was quantified, and such values are independent of concentration. Circadian products known to be excreted in the urine were administered to transfected cells in vitro with no effects upon LUC expression. It was ultimately determined that urine collected at the onset of mouse morning (6:00 pm, mice being nocturnal), contained an agent that masked the detection of LUC. It was also found that serum shock could also induce increased LUC detection in vitro.Additional human reporters, including secreted alkaline phosphatase, human chorionic gonadotropin, and prostate-specific antigen were examined in this study for the purpose of developing a non-invasive urinary lateral flow assay for cancer detection which would detect more than one reporter protein simultaneously by delivering multiple plasmids containing different cancer-specific promoters.