Prolactin (PRL) promotes tumor growth in various experimental models and leads to prostate hyperplasia and mammary neoplasia in PRL transgenic mice. Increasing experimental evidence argues for the involvement of autocrine PRL in this process. PRL receptor antagonists have been developed to counteract these undesired proliferative actions of PRL. However, all forms of PRL receptor antagonists obtained to date exhibit partial agonism, preventing their therapeutic use as full antagonists. In the present study, we describe the development of new human PRL antagonists devoid of agonistic properties and therefore able to act as pure antagonists. This was demonstrated using several in vitro bioassays, including highly sensitive assays able to detect extremely low levels of receptor activation. These new compounds also act as pure antagonists in vivo, as assessed by analyzing their ability to competitively inhibit PRL-triggered signaling cascades in various target tissues (liver, mammary gland, and prostate). Finally, by using transgenic mice expressing PRL specifically in the prostate, which exhibit constitutively activated signaling cascades paralleling hyperplasia, we show that these new PRL analogs are able to completely revert PRL-activated events. These second generation human PRL antagonists are good candidates to be used as inhibitors of growth-promoting actions of PRL.

Androgen-independent recurrence is the major limit of androgen ablation therapy for prostate cancer. Identification of alternative pathways promoting prostate tumor growth is thus needed. Stat5 has been recently shown to promote human prostate cancer cell survival/ proliferation and to be associated with early prostate cancer recurrence. Stat5 is the main signaling pathway triggered by prolactin (PRL), a growth factor whose local production is also increased in high-grade prostate cancers. The first aim of this study was to use prostate-specific PRL transgenic mice to address the mechanisms by which local PRL induces prostate tumorogenesis. We report that (i) Stat5 is the major signaling cascade triggered by local PRL in the mouse dorsal prostate, (ii) this model recapitulates prostate tumorogenesis from precancer lesions to invasive carcinoma, and (iii) tumorogenesis involves dramatic accumulation and abnormal spreading of p63-positive basal cells, and of stem cell antigen-1-positive cells identified as a stem/progenitor-like subpopulation. Because basal epithelial stem cells are proposed to serve as tumor-initiating cells, we challenged the relevance of local PRL as a previously unexplored therapeutic target. Using a doubletransgenic approach, we show that Δ1-9-G129R-hPRL, a competitive PRL-receptor antagonist, prevented early stages of prostate tumorogenesis by reducing or inhibiting Stat5 activation, cell proliferation, abnormal basal-cell pattern, and frequency or grade of intraepithelial neoplasia. This study identifies PRL receptor/Stat5 as a unique pathway, initiating prostate tumorogenesis by altering basal-/ stem-like cell subpopulations, and strongly supports the importance of further developing strategies to target locally overexpressed PRL in human prostate cancer.P rostate cancer is the most common cancer affecting men in the Western world, and the second leading cause of male cancer-related deaths (1). After initial response to androgen ablation most patients relapse, and hormone-refractory prostate cancer develops to very aggressive stages that are frequently lethal. Because prostate cancer stem cells do not express androgen receptor, they offer a theoretical explanation for the failure of androgen-based therapies (2). These cells are presumably located in a niche-like site within the basal epithelial layer (2), which is identified by p63-positive staining (3). The essential role of basal epithelial stem cells in prostate cancer initiation has been recently demonstrated (4). Identification of pathways regulating basal and stem cell compartment is thus required to develop novel therapeutic strategies.Prostate cancer develops through well-defined stages, from prostate intraepithelial neoplasia (PIN, considered as preneoplastic lesions), to localized (in situ), invasive, and finally metastatic cancer. The role of the PI3K/Akt/mTOR (mammalian target of rapamycin) pathway in progressing through these stages has been well characterized using various genetically modified mouse models (5, 6). One hallmar...

The GH receptor (GHR) is a member of the cytokine receptor superfamily; its signaling involves the activation of Janus tyrosine kinases (JAK2) and Stat (signal transducers and activators of transcription) transcription factors. Using truncated and tyrosine mutants of the receptor, we show that different receptor domains are essential for the activation of Stat3 and Stat5. GH-dependent phosphorylation of JAK2, Stat3, and Stat5, as well as transactivation studies with reporter genes containing Stat3 and Stat5 DNA-binding elements, was performed in cells expressing the various GHR mutants. The membrane-proximal region of the receptor necessary for JAK2 activation is sufficient for Stat3 activation. In contrast, C-terminal tyrosine residues of GHR are absolutely required for Stat5 activation. The same residues are also involved in the regulation of JAK2 dephosphorylation, possibly through the activation of a phosphatase. Using in vitro experiments with glutathione-S-transferase fusion proteins, we demonstrate that the SH2 domain of Stat5 binds to the carboxy-terminal tyrosine-phosphorylated residues of GHR. Our results show that a cytokine receptor can mediate differently the activation of distinct Stat proteins that could be involved in cytokine-specific effects.

Experimental, clinical, and epidemiological data support the growth-promoting role of endocrine prolactin (PRL) in mammary tumors. PRL is also produced by the breast, where it is now recognized to act as a growth/survival factor via autocrine/paracrine mechanisms. Recent transgenic (Tg) mouse models have revealed the pro-oncogenic effect of PRL over-expression in virgin mammary glands. To address the question whether PRL tumorigenicity was maintained on differentiated mammary glands, we generated mammaryspecific Tg mice expressing human (h)PRL under the control of the milk whey acidic protein promoter, which directs autocrine hPRL over-expression in late gestation throughout lactation. Minimal levels of transgene expression were detected in the mammary glands of virgin animals, which at best induced partial ductal branching and lobulo-alveolar structures in older nulliparous females. As expected, expression of mammary hPRL dramatically increased at the end of first pregnancy, and from this point it never returned to baseline, although it peaked at each gestation/lactation cycle. Over-expression of hPRL that starts when the gland is already well into the differentiation process led to various morphological mammary alterations, including abnormally differentiated epithelium, atropy of the myoepithelial layer, dilated ducts, cysts, and lymphocytic infiltrates. These phenotypes tended to worsen with successive pregnancies, also reflecting cumulative damage of failure of involution. Although some older, multiparous females developed benign tumors (papillomas and metaplasias), none of the animals studied developed mammary carcinomas. In addition, we noticed that half of the Tg females exhibited lactation defects, leading to significantly increased pup mortality. This phenotype was due neither to failure of milk production nor to modification of its protein content, but rather it was correlated to lipid enrichment of the milk, which, in combination with profoundly altered morphology of the gland, led to impaired milk extrusion through the nipple. In summary, these data show that over-expression of autocrine hPRL in a differentiating mammary gland induces dramatic functional and morphological defects, but not carcinoma. This deserves further investigations on the emerging concept that autocrine PRL may have different effects on pathological development of the mammary gland depending on the differentiation state of the latter.

Human livers, obtained from donors at the time of transplant, were homogenized in 0.25 M sucrose and fractionated by differential centrifugation. The specific binding of [125I] human (h) GH to total particulate fractions from 18 livers varied from 0.4-5.1% of the total radioactivity/100 micrograms protein. Binding affinity was 2.0 +/- 0.3 X 10(9) M-1, and binding capacity ranged from 14-53 fmol/mg protein. A different proportion of receptors occupied by endogenous hGH did not explain the large variation in binding. Binding sites were specific for hGH. Dissociation of the hormone-receptor complex was extremely slow. No specific binding of [125I]hPRL was observed. Specific binding of insulin was found in fractions from all livers and varied less than hGH binding. Cross-linking of [125I]hGH to plasma membrane and microsome receptors yielded two major autoradiographic bands corresponding to an estimated mol wt of 103,000 for the receptor, with a possible subunit of 54,000. Human liver primary fractions were characterized. The binding of hGH and insulin displayed a nucleo-microsomal distribution pattern in the primary fractions; 54.2% and 27.9% of the hGH-binding activity were found in the microsomes and the nuclear fraction, respectively, whereas insulin binds equally to nuclear and microsomal elements. Our findings suggest that hGH-binding sites are present in the plasma membrane and also in one or more intracellular compartments, whereas a high proportion of insulin receptors is associated with the plasma membrane.

Two distinct soluble proteins that specifically bind 125I-labeled human growth hormone (GH) are identified in the supernatant of ultracentrifuged rabbit milk, using HPLC gel filtration. The higher molecular weight protein is GH specific, whereas the other one is specific for prolactin (PRL). 125I-labeled human GH, the M, of the GH-binding protein and of the PRL-binding protein were estimated to be 51,000 and 33,000, respectively. The binding proteins identified in the present work are probably responsible for the transport oftheir specific ligands in the milk. It is also conceivable that they have a role in the effects of GH and PRL in the mammary gland and/or the intestine of the young.

[125I]Iodohuman GH ([125I]iodo-hGH) was injected iv to female rats, and its subcellular distribution was studied with time using fractionation techniques. Uptake in the total particulate fraction was maximal by 15 min after injections; at that time, 7% of the radioactivity injected was recovered per g liver. Liver uptake of [125I]iodo-hGH was markedly inhibited by coinjection of native hGH. [125I]Iodo-hGH taken up by the liver underwent a time-dependent translocation process. In the first 5 min, the radioactivity was associated with membranes of high density sedimenting in the nuclear and microsomal fractions. Later on, it was progressively associated with microsomal subfractions of lower density, the Golgi fractions; labeling of these was maximum at 15 min. Fifteen and 30 min after injection, labeled material was recovered in the mitochondrial-lysosomal fraction. Further fractionation of the latter by centrifugation on a metrizamide gradient showed that all of the radioactivity was associated with lysosomal subfractions, with virtually no radioactivity associated with mitochondria. On linear sucrose density gradients, the radioactivity exhibited a broad, somewhat bimodal distribution; the component of highest density coincided with the lysosomal enzyme acid phosphatase. Triton WR 1339 treatment of rats resulted in a shift of the radioactivity and acid phosphatase toward lower densities, indicating that a high proportion of the former was associated with lysosomes. The labeled material eluted from the subcellular fractions prepared at different times appeared as intact hGH, ad judged by trichloroacetic acid precipitability and binding to membranes. Upon in vivo interaction with liver cells, [125I]iodo-hGH is internalized, with a sequential association with plasma membranes, Golgi elements, and lysosomes.

GH receptors (GHRs) and PRL receptors (PRLRs) were studied in human peripheral blood mononuclear cells (PBMC) using flow cytometry, biotinylated anti-GH receptor monoclonal antibody 10B8, and biotinylated human PRL. Variations of GHR and PRLR expression and the relationship of plasma GHBP and GH receptor in PBMC subsets were examined as a function of age and sex. By double immunofluorescence staining, we show that about 30% of total cells express GH receptors, with a low expression in T cells, whereas almost all B cells and monocytes are GH receptor positive. Four age groups were defined among the 64 normal volunteers, aged 12 to 85 yr, who were included in the study. The percentage of PBMC expressing GH receptors is significantly lower in group 2 (20-40 yr) than in group 1 (12-20 yr) and group 4 (>60 yr). In T cells, monocytes and B cells, no significant changes are detected in either the percentage of GH receptor positive cells or in the GH receptor level per cell. The level of PRLRs expressed in PBMC is significantly higher in age group 2 than in age group 4. A negative correlation is observed between plasma GHBP and the percentage of PBMC expressing GH receptors. These results suggest that regulation of GH receptors in lymphocytes and in other target cells could be different.