Prostate epithelial cell growth is under the control of both steroid a
nd peptide factors. Human prostate cancer cell lines have been used to
investigate similar agents in malignancy. Activins are dimeric peptid
es structurally related to transforming growth factor-beta and produce
d in the gonads and a nide array of extragonadal tissues. The activins
act at the pituitary to regulate the synthesis and secretion of FSH.
At other sites, such as bone marrow, liver, and gonads, activin may pl
ay an important role in the regulation of cell growth and differentiat
ion. It was the purpose of the current study to determine whether acti
vin had similar actions on prostate cancer cells, specifically the and
rogen-responsive LNCaP and the androgen-resistant PC-3 cell lines. Usi
ng reverse transcription-PCR, messenger RNAs for type I and type II ac
tivin receptor subunits as well as the activin-binding protein follist
atin were detected in both cell lines. Activin treatment rapidly (<24
h) inhibited LNCaP, but not PC-3, cell growth. The effects of activin
were evident at low levels, with a concentration of 5 ng/ml being effe
ctive at 24 h, and a concentration of 0.5 ng/ml being effective at 48
h. These results contrasted with the actions of transforming growth fa
ctor-beta, which inhibited only PC-3 cells and required a greater trea
tment duration (96 h) to be effective. To determine whether these pros
tate cancer cell lines were also producing activin, LNCaP and PC-3 cel
ls were treated with follistatin. Again, only the LNCaP cells responde
d, with growth acceleration noted by 24 h. As PC-3 cell responses to a
ctivin could be independent of cell proliferation, we transfected LNCa
P and PC-3 cells with a known activin-responsive promoter/reporter gen
e construct (p3TP-Lux) and treated cells with activin. Only LNCaP cell
s produced a measurable response in luciferase activity. Finally, we a
ttempted to determine whether the PC-3 cell resistance to activin was
mediated via a transferable factor. PC-3 conditioned medium was added
to LNCaP cells in the absence or presence of exogenous activin and had
a small, but statistically nonsignificant (P < 0.09), action to blunt
the actions of activin. We conclude that activin is a potent growth i
nhibitor of LNCaP cell growth. Moreover, these cells also produce acti
vin, suggesting that locally derived activin may play a role in regula
ting cell proliferation. Despite expressing messenger RNAs for activin
receptors, PC-3 cells are resistant to activin, perhaps the result of
the production of an activin-blocking factor or a defective activin r
esponse system. These cell lines will thus serve as useful models in w
hich to further study the cellular basis of activin action.