Carboxy derivatives of isoflavones that exhibit oestrogenic/anti-oestrogenic properties were used as carriers for affinity drug targeting to H295R adrenocortical carcinoma cells that express transcripts of oestrogen receptor (ER) alpha and beta. These derivatives were prepared by introducing a carboxymethyl group at the 6-position of genistein and of biochanin A, yielding 6CG and 6CB respectively. In transactivation assays, 6CG displayed mixed agonist/antagonist activity for ERalpha, whereas 6CB displayed only weak antagonist activity. Low concentrations of oestrogen, 6CG and 6CB were capable of inducing proliferation in H295R cells and of stimulating creatine kinase (CK) specific activity, suggesting that these cells were sensitive to oestrogenic compounds. In in vivo experiments, both 6CG and 6CB were capable of inhibiting oestrogen-induced CK specific activity in rat tIssues. For affinity drug targeting, the cytotoxic drug daunomycin was coupled to 6CB and 6CG, yielding 6CB-Dau and 6CG-Dau respectively. These conjugates were tested for their antiproliferative ability to inhibit DNA synthesis as assessed by incorporation of [(3)H]thymidine in H295R cells. A dose-dependent cytoxicity was observed with both conjugates. At 0.3-3 nM, both conjugates were 10 to 30 times more potent than daunomycin. At 30 nM these conjugates were two to three times more potent than daunomycin. At concentrations ranging between 300 and 3000 nM, no difference in cytotoxicity was observed between the conjugates and daunomycin. When the cells were treated over a wide range of concentrations with a combination of 6CG plus daunomycin, the observed cytotoxicity was less than with daunomycin alone. When non-transformed rat enterocytes, which do not express ER, were treated with 6CG-Dau or daunomycin, the antiproliferative effect of 6CG-Dau was the same as that of daunomycin over the concentration range tested. These pilot studies suggest that the ready availability of oestrogenic binding sites in H295R cells can be exploited for site-directed chemotherapy.

We have reported previously that dihydrotestosterone (DHT) induces a biphasic effect on DNA synthesis in human vascular smooth muscle cells (VSMC), i.e. stimulation at low concentrations and inhibition at high concentrations. In contrast, DHT dose-dependently stimulated [(3)H]thymidine incorporation in a human endothelial cell line (ECV304). Additionally, DHT increased the specific activity of creatine kinase (CK) in both vascular cell types. In the present study, we have determined whether some of these effects are exerted via membrane-binding sites. We measured changes in DNA synthesis and CK after treatment with DHT and the membrane-impermeant testosterone-3-carboxymethyl oxime conjugated to bovine serum albumin (BSA) (T-BSA). High concentrations of either DHT or T-BSA inhibited VSMC proliferation (by 52+22% and 51+25% respectively). DHT as well as T-BSA increased DNA synthesis in ECV304 cells dose-dependently. In contrast, T-BSA did not affect CK in either cell type. In both cell types, DHT as well as T-BSA increased mitogen-activated protein kinase (MAPK) kinase activity as measured by total phosphorylated MAPK. Further, the inhibitory effect of either the free or protein-bound androgens on DNA synthesis was blocked by UO126, an inhibitor of MAPK kinase activity. T-BSA conjugate labeled with Europium showed binding to whole VSMC, which could be displaced by excess T-BSA, but not by estradiol-BSA or the free hormones. Finally, using T-BSA linked to the fluorescent dye Cy3.5, we directly demonstrated the presence of membrane-binding sites for androgen in VSMC. Hence, the inhibitory effects of testosterone on DNA synthesis in VSMC are apparently exerted by membrane-binding sites for androgen, do not require intracellular entry of the hormone and its binding to the classical nuclear receptors and are linked to MAPK activation.

The novel genistein (G) derivative, 6-carboxymethyl genistein (CG) was evaluated for its biological properties in comparison with G. Both compounds showed oestrogenic activity in vitro and in vivo. On the other hand G and CG differed in the following parameters: (i) only CG displayed mixed agonist-antagonist activity for oestrogen receptor (ER) alpha in transactivation assays and (ii) only CG was capable of attenuating oestrogen (E(2))-induced proliferation in vascular smooth muscle cells and of inhibiting oestrogen-induced creatine kinase (CK) specific activity in rat tissues. On the other hand only G enhanced the stimulatory effect on CK specific activity in the uterus. In comparison to the selective oestrogen receptor modulator (SERM) raloxifene (RAL), CG showed the same selectivity profile as RAL in blocking the CK response to E(2) in tissues derived from both immature and ovariectomized female rats. Molecular modelling of CG bound to the ligand binding domain (LBD) of ERbeta predicts that the 6-carboxymethyl group of CG almost fits the binding cavity. On the other hand, molecular modelling of CG bound to the LBD of ERalpha suggests that the carboxyl group of CG may perturb the end of Helix 11, eliciting a severe backbone change for Leu 525, and consequently induces a conformational change which could position Helix 12 in an antagonist conformation. This model supports the experimental findings that CG can act as a mixed agonist-antagonist when E(2) is bound to its receptors. Collectively, our findings suggest that CG can be considered a novel SERM with unique effects on the vasculature, bone and uterus.