Rose bengal dye is thought to highlight corneal lesions induced by herpes simplex virus type 1 (HSV-1) by virtue of its binding to dead or dying HSV-1-infected epithelial cells. However, whether rose bengal binds specifically to damaged versus normal corneal epithelial cells is unclear. To determine the binding properties of rose bengal, the authors compared binding of the dye to HSV-1-infected and uninfected cells, determined the cellular binding sites of the dye, and investigated the effects of rose bengal on HSV-1 replication in dye-treated cells in vitro. Spectrophotometric analysis revealed that uninfected and infected Vero cells bound equivalent amounts of dye at several times post inoculation, indicating that rose bengal does not preferentially bind to HSV-1-infected cells. By light microscopy, rose bengal was found to bind to the cell nuclei and perinuclear region of human corneal epithelial cells (HCEC) and Vero cells. Pretreatment of Vero and HCEC with different concentrations of rose bengal and exposure to 148 microW/cm2 of white light for 2 min reduced the ability of both cell types to support HSV-1 replication. Vero cells, in the absence of rose bengal, supported HSV-1 replication, whereas pre-treatment with 0.05% rose bengal reduced the yield of HSV-1 by 99.99% (P less than 0.000001) and 1% rose bengal completely prevented HSV replication. HCEC supported HSV-1 replication in the absence of rose bengal, but pre-treatment with 1% or 0.05% rose bengal completely prevented HSV-1 replication (P less than 0.000001).(ABSTRACT TRUNCATED AT 250 WORDS)