Adenoviral receptor expression of normal bladder and transitional cell carcinoma of the bladder

OBJECTIVE: The insertion of absent or underexpressed genes into cancer cells to alter their malignant phenotype is an important potential application of available gene therapy technology. One of the more common viral vector systems that has been extensively studied for this purpose are the replication-deficient adenoviruses (Ad). Adenoviral infection of cells is mediated through a complex pathway, initiated following viral-cell attachment. Adenoviral-cell attachment occurs following interactions with a 46-kDa transmembrane protein with high affinity for both the Coxsackie and adenovirus, designated the CAR (Coxsackie and adenoviral receptor). Additional important cell-viral interactions that occur involve the alpha(v)-based integrins, specifically alpha(v)beta3 and alpha(v)beta5. The purpose of the present study was to determine the extent of expression and localization of the known Ad receptor proteins (CAR, alpha(v)beta3, and alpha(v)beta5) in normal and cancerous human bladders. MATERIAL AND METHODS: Frozen tissue samples of normal bladder and invasive transitional cell cancers of the bladder were evaluated. Tissue blocks containing muscle-invasive transitional cell carcinoma (TCC) were obtained following radical cystectomy, which were performed at our institution. Thirty-two invasive transitional cell bladder tumors were evaluated, each with a matched sample of histologically normal-appearing bladder used as a control. Four additional samples of normal bladder were obtained from patients with no evidence of disease of the bladder and served as further controls. Three additional cases of invasive bladder cancer with no matching normal tissue were also evaluated. Identification of the CAR receptor was performed using the anti-CAR mouse monoclonal antibody designated RmBC. The integrins alpha(v)beta3 and alpha(v)beta5 were identified using the mouse monoclonal antibodies designated LM609 and P1F6 respectively. All slides were evaluated by two of the authors (M.B., B.B.) without knowledge of the clinical and pathological data. RESULTS: Normal bladder: Normal bladder mucosa demonstrated a marked positivity for CAR in 29/35 (82.8%) cases. In contrast, normal transitional epithelial cells were uniformly negative when tested for the integrins alpha(v)beta3 and alpha(v)beta5. Subepithelial tissues, specifically the connective tissue components of the lamina propria and deep muscle wall of the bladder, were positive for alpha(v)beta3 and for alpha(v)beta5 in 61 and 75% of samples, respectively. Endothelial cells associated with the various layers throughout the bladder uniformly expressed both integrins and served as a consistent internal control for both antibodies. An almost identical staining pattern of the endothelium was observed using LM609 and P1F6 in all samples tested. Bladder transitional cell carcinoma: CAR immunoreactivity against TCC cells was uniformly decreased compared to normal transitional cells. Nine tumors exhibited a weak positivity for CAR while the remaining samples were negative. In some cases, the absence of CAR positivity was associated with histological evidence of carcinoma in situ. In 6 cases, it led to the identification of small regions of carcinoma in situ that were not noted on primary pathological evaluation. Peritumoral connective tissue expressed both integrins in the majority of cases, similar to the pattern described above for normal bladder. Transitional cell cancers demonstrated a similar pattern of expression of alpha(v)beta5, in which all tumor cells exhibited minimal or no staining. CONCLUSIONS: The success of all viral-mediated gene therapy strategies relies on the ability of the vector to efficiently deliver its genetic material to a target cell population. In the current study, we demonstrate that the bladder epithelial layer consistently expresses high levels of CAR. Deeper layers of the epithelium also express CAR, including the basal layer cells. A decrease in the expression of CAR appears as an early event in bladder carcinogenesis. We observed that both alpha(v)beta3 and alpha(v)beta5 are strongly expressed in muscle cells surrounding the neoplastic cells, as well as within the peritumoral connective tissue. In cases of invasive bladder cancer that have lost CAR expression, an adenoviral vector may still be utilized through the less efficient interactions with the integrins. Bladder tumor tissue may be less susceptible to an adenoviral-mediated gene therapy approach in which a significant percentage of tumor cells require transduction. Adenoviral uptake by tumor or peritumoral cells with subsequent gene transfer could be predicted by the level of CAR and alpha(v)-based integrin expression. This would enhance our ability to identify those patients whose tumors would be more susceptible to Ad-mediated gene delivery as part of an antitumor treatment