The effects of in v i w exposure to phenobarbital (PB) on hepatic gap junctional intercellular communication (GJIC) and conncxin protein expression in Sprague-Dawley rats were examined by in vivolirn virro dye-transfer assay, immunohistochemical staining, and by Western blot analysis. PB (50 mgkg) was administered orally once a day for up to 6 wk. The average size of the dye spread after injection of Lucifer Yellow decreased at week 1 and remained at the same level until \veek 6. The area and number of connexin 32 (Cx32) spots pcr hcpatocyte in the central zone of liver lobules decreased from week 1 to week 6, but no change of Cx32 spots in thc pcriphcral zone was obscrvcd. Thc average area and number of connexin 26 (Cx26) spots per hepatocytes showed no clear change through thc cxpcrimcntal pcriods. Thc dccrcascd level of Cx32 protein in plasma membranes was observed in the PB group. These results suggest that PB, a liver tumor-promoting agent, inhibits hepatic GJIC in vivo in rats and that aberrant Cx32 protein expression and/or localization may be responsible for this effect.Kejwords. Liver; gap junction; connexin; tumor promoter INTRODUCTION Gap junctional intercellular communication (GJIC) is considered to play an important role in carcinogenesis. A recent review of the ability of chemicals with tumor-promoting or tumor-inhibiting activity to modulate GJIC reports that more than 100 substances have been shown to inhibit GJIC (1). The inhibition of GJIC appears to be the most common property of chemicals with tumor-promoting activity; however, phenobarbital (PB), a liver tumor promoter (2, lo), has shown mixed results (negative andlor positive) in iit vitro assays (1). It was reported that PB (0.05% in diet) diminished the total area and size of rat hepatocyte gap junctions by freeze-fracture analysis in vivo (12). The level of gap junction protein mRNA in rat liver was observed to be markedly reduced at 4 and 11 wk of PB exposure (0.1% in drinking water) (7). However, there have been few studies that examine the hepatic gap junctional functions of rats exposed to PB iii vivo (5). We have studied the effect of DDT on hepatic GJIC in rats and clearly demonstrated that DDT inhibited hepatic GJIC and caused aberrant connexin 32 (Cx32) and connexin 26 (Cx26) protein expression iii vivo (13). This study was conducted to demonstrate the inhibition of hepatic GJIC in rats exposed to PB-iiz vivo and compare these results with the effects of DDT In the present study, the effects of PB on hepatic GJIC of rats treated orally with PB (50 mglkgiday) for up to 6 wk were examined by iit vivoiiit vitro dye-transfer assay using freshly removed liver slices. Localization of the hepatic gap junction proteins, Cx32 and Cx26, and protein levels of