Cell cycle-dependent initiation of hepatocarcinogenesis in rats by (+/-)-7r,8t-dihydroxy-9t,10t-epoxy-7,8,9,10-tetrahydrobenzo(a)py rene.

Male F344 rats were treated once with (+/-)-7r,8t-dihydroxy-9t,10t-epoxy-7,8,9,10-tetrahydrobenzo(a)pyre ne (BPDE) at various times after a two-thirds partial hepatectomy. Diet containing 0.05% phenobarbital was fed subsequently to promote the development of hepatocellular neoplasms from initiated cells. This combined treatment caused an apparently continuous emergence of hepatocellular neoplasms between 31 and 60 weeks after initiation. Yields of hepatocellular neoplasms (adenomas and carcinomas) seen at 45 weeks after initiation varied according to the time of treatment with BPDE. Rats treated at 15 h after partial hepatectomy, when maximal fractions of proliferating hepatocytes were entering the S phase, displayed the greatest incidence and yield of hepatocellular neoplasms. Rats treated when hepatocytes were early in the prereplicative G1 phase demonstrated a significantly lower incidence and yield of hepatocellular neoplasms. No neoplasms were seen in livers of rats treated with BPDE without prior partial hepatectomy; this indicates the insensitivity or resistance of nonproliferating G0 hepatocytes to initiation by this chemical. The variation in yields of neoplasms could not be attributed to variation in binding of carcinogen to DNA or to rates of DNA repair. The risk of initiation of hepatocytes appeared to be correlated (r = 0.95) with the fraction of hepatocytes that were entering the S phase at the time of treatment with BPDE. These results reveal a significant cell cycle dependency for initiation of hepatocarcinogenesis by BPDE and show that proliferating hepatocytes are at greatest risk when early in the S phase. Quantitation of islands of hepatocellular alteration suggested a similar pattern of cell cycle-dependent sensitivity, although the cycle-dependent variation in island frequencies (1.4-fold) was less than the variation in yields of neoplasms (5-fold). However, 1300-5000 islands were estimated to appear for every one neoplasm that appeared. When viewed within the context of multistep carcinogenesis, and assuming that islands of altered hepatocytes represent a step in this process, then these results suggest that islands may represent populations of cells at least one step removed from neoplasia.