In order to clarify whether the ovarian tumors induced in a long-term carci
nogenicity study of nitrofurazone (NF) in mice can be also produced in a sh
ort-term model using transgenic (Tg) mice carrying the human c-Ha-rns gene
(rasH2 mice), the following 3 experiments were performed. In experiment 1,
both rasH2 mice and their wild CB6F(1) littermates carrying no c-Ha-ms gene
(non-Tg mice) that were fed a diet containing 500 to 1,000 ppm NF for 7 we
eks demonstrated ovarian atrophy characterized by decreased labeling indice
s (LIs) for proliferating cell nuclear antigen (PCNA) in granulosa cells. I
n experiment 2, increased numbers of atretic follicles and decreased PCNA L
Is in granulosa cells were recognized in rasH2 mice given diets containing
250 or 500 ppm NF for 26 weeks. but no tumor induction was grossly observed
. In experiment 3, similar ovarian atrophy was observed in association with
increased serum luteinizing hormone (LH) levels in both rasH2 and non-Tg m
ice given diet containing 1,000 ppm NF for ii days. These results indicate
that long-term NF treatment induces ovarian tumors in mice, possibly by con
tinuous stimulation with gonadotropins such as LH via a negative-feedback p
henomenon secondary to ovarian atrophy (as the tumor-induction mechanism),
although we could not completely rule out a genotoxic mechanism.