Search

A significant increase in endometrial cancer incidence in tamoxifen-treated breast cancer patients has been reported in many recent studies. The major growth stimulators of endometrial tumors are estrogens, but paradoxically, tamoxifen, a known antiestrogen, also stimulates their growth. The mode of action of estrogen can be partially explained by the modulation of insulin-like growth factor (IGF) autocrine or paracrine action. The purpose of the present study was to examine the involvement of the IGF system in the tamoxifen-stimulated growth of Ishikawa endometrial cancer cells by quantitating the IGF-I receptors and their phosphorylation, as well as membrane associated and secreted IGF-binding proteins (IGFBPs). Tamoxifen did not affect the number or affinity of IGF-I receptors. On the other hand, tamoxifen, similar to estradiol, increased IGF-I-stimulated tyrosine phosphorylation of cellular substrates. In contrast, in MCF-7 mammary cancer cells, tamoxifen reduced IGF-induced tyrosine phosphorylation in the presence of estradiol. The pure antiestrogen LY156758 did not affect Ishikawa basal cell growth but inhibited estradiol- and tamoxifen-induced growth. Growth inhibition by LY156758 of tamoxifen and estradiol-stimulated cells was accompanied by a corresponding inhibition of IGF-stimulated tyrosine phosphorylation. Tamoxifen caused a 3-fold decrease in membrane-associated IGFBPs. Moreover, a reduction in soluble IGFBPs was also observed, making the IGF peptides more available to the receptors. A parallel decrease in IGFBP-3 mRNA was also detected. These experiments suggest that tamoxifen, like estradiol, directly sensitizes endometrial cancer cells to the effects of IGFs that act through the type I receptor. Furthermore, the decrease in IGFBPs and the increase in tyrosine phosphorylation in the presence of tamoxifen provides a molecular mechanism that accounts for the uterotropic effects that are seen with tamoxifen therapy.

Insulin-like growth factor I (IGF-I) receptors and membrane-associated IGF-binding proteins (IGFBPs) were examined in Ishikawa endometrial cancer cells. Our findings suggest that about 95% of [125I]IGF-I is bound to membrane-associated IGFBPs rather than to IGF-I receptors. Specifically, [125I]IGF-I binding to cell membranes could be completely displaced by cold IGF-I or IGF-II, but not by insulin, suggesting that binding was primarily due to IGFBPs. This was confirmed by using [125I]des-(1-3)IGF-I as the ligand. Des-(1-3) IGF-I binds with high affinity to IGF-I receptors, but with markedly lower affinity to IGFBPs. [125I]Des-(1-3)IGF-I bound to Ishikawa cells was displaced by IGF-I, IGF-II, and insulin. These results suggest that measuring IGF-I receptor levels using labeled IGF-I may be misleading. Accordingly, we evaluated the differential binding of [125I]IGF-I and [125I]des-(1-3)IGF-I to study the involvement of the IGF system in the stimulation of Ishikawa cell growth by estradiol. IGF-I stimulates Ishikawa cell proliferation, but at low concentrations, and this stimulation is largely dependent on the presence of estradiol. Estradiol caused a 2.5-fold increase in IGF-I receptor levels. Moreover, estradiol reduced soluble IGFBP levels, presumably increasing the availability of IGFs for their receptors. This elevation in IGF-I receptor levels and the decrease in IGFBP levels were accompanied by a 3.5-fold increase in IGF-I receptor messenger RNA and a 2.5-fold decrease in IGFBP messenger RNAs. These experiments suggest that estradiol sensitizes endometrial cancer cells to the effects of IGFs by simultaneously elevating receptor levels and decreasing (potentially inhibitory) IGFBP levels.

A protein kinase activity fraction was defined in cytosols and membranes of mammary tissue isolated from rats during pregnancy lactation, and weaning. By partial purification on DEAE-cellulose columns, it was shown that this protein kinase activity is cAMP independent and that its preferential substrate is casein and not histone. This protein kinase activity is inhibited by the bioflavonoid quercetin at doses that do not inhibit cAMP-dependent protein kinase activity. The enzyme requires Mg2+ and is inactive in the presence of 10 mM Ca+2; these properties distinguish this activity from casein kinase activity found in the Golgi fraction and involved in milk protein processing. By following the physiological cycle of mammary gland development during pregnancy, lactation, and weaning, we found a close correlation between proliferation, expressed as the DNA content per gland, and quercetin-inhibited cytosolic protein kinase activity. Moreover, changes in this phosphorylating activity preceded the glandular growth changes. There was a less significant correlation between the growth process and protein kinase activity in the membrane fraction. The cytosolic cAMP-dependent protein kinase activity showed (only partial) correlation with growth only during pregnancy. Cytosolic progesterone receptor levels in mammary tissue were used as an estrogenic marker. Tissue growth correlated with progesterone receptor levels during pregnancy, where estrogens are the predominant hormones affecting tissue proliferation. However, no such correlation was found during lactation and weaning, when PRL is the major hormone affecting mammary gland growth. These results suggest that quercetin-inhibitable protein kinase activity is not merely another estrogenic marker, but represents more general regulatory activity which might be connected to growth processes of breast tissue.

A protein kinase activity was defined in cytosols and membranes of rat uterus by using two types of enzyme inhibitors. The heat stable protein kinase inhibitor inhibited the cAMP-stimulated protein kinase but not the basal activity measured in the absence of the cyclic nucleotide. However, this basal activity was inhibited by quercetin in a dose-dependent manner, whereas the cAMP-stimulated protein kinase was not inhibited by this bioflavonoid. Thus, quercetin can be used as a tool to define a specific fraction of cAMP-independent protein kinase activity in rat uterine cytosols and membranes. The mature rat uterus is characterized by short (4 days) cyclic changes in tissue growth. Cell proliferation expressed as total uterine DNA content is observed on proestrus, reaches its peak at estrus, and declines sharply during metestrus and diestrus. The quercetin-inhibited protein kinase activity in the membranes of this tissue also changes cyclically and precedes by one phase the cyclic change in tissue proliferation. The clearest effect was seen when the protein substrates for this protein kinase activity were endogenous membrane proteins partially solubilized by Triton X-100 treatment. In contrast no change was observed in cytosolic, quercetin-inhibited protein kinase activity. The cAMP-stimulated protein kinase activity decreased slightly on diestrus. These results support our working hypothesis that tissue proliferation in the uterus and other tissues such as rat mammary gland and mammary tumors, correlates with cAMP-independent, quercetin-inhibited, protein kinase activity in these tissues.