Both positive and negative feedback on the (hypothalmus)- pituitary-gonad axis occur in salmonids. The aim of the present study was to investigate the role of different androgens, and in particular the involvement of aromatization of androgens to estrogens in feedback mechanisms. Previously mature Atlantic salmon, Salmo salar, male parr were studied in two experiments. In the first experiment, intact fish were implanted with Silastic capsules filled with the aromatase inhibitor 1,4,6-androstatriene-3,17-dione (ATD) in spring and sampled in the autumn when the rematuring males were starting to display running milt. In the second experiment, castrated males were implanted with capsules containing ATD, the androgens testosterone (T) and 11-ketoandrostenedione (11KA), or ATD and T combined in spring. These fish were sampled in the summer when rematuring fish were starting to show signs of gonadal growth. Pituitary and plasma gonadotropins (GTH I and GTH II) were studied using radioimmunoassay. In autumn, ATD treatment reduced pituitary and plasma GTH II levels. In summer, GTH II was consistently nondetectable in plasma. Castration diminished pituitary GTH II content. Treatment with T increased pituitary GTH II content, an effect that was attenuated when T treatment was combined with ATD. All these results are consistent with the presence of an aromatase-dependent positive feedback of T on GTH II. 11KA also had a stimulatory effect on GTH II, although weaker than that of T. Testicular size and spermiation was reduced by ATD in autumn; the latter of these results is likely to be due to the inhibitory effect of ATD on GTH II. Positive effects of ATD on plasma GTH I were found in autumn, indicative of an aromatase-dependent negative feedback in this phase. On the other hand, castration increased plasma and pituitary GTH I levels in summer, indicating that the gonads in this phase exert a predominantely negative control of GTH I. In summer, negative effects of T on GTH I pituitary levels were not suppressed, but were rather enhanced, by the combined treatment with ATD. Furthermore, plasma GTH I levels were lower after treatment with T in combination with ATD than with T or ATD separately. These negative effects of T were not diminished by ATD, so that they are nonaromatase-dependent. Furthermore, since they were actually more pronounced in the presence of ATD it is suggested that there is also a positive aromatase-dependent feedback component in this phase. In addition, 11KA had a negative effect on plasma and pituitary GTH I in castrated previously mature males. Thus, GTH I secretion is controlled by both aromatase-dependent and nonaromatase-dependent feedback effects, of which at least the former may be positive or negative depending on season. In summary, the feedback control of GTH I appears to be more complex than that of GTH II.
Copyright 1999 Academic Press.