These results indicate that measurement of basal LH beta is of poor value in the diagnosis of non-functioning pituitary adenomas and the identification of gonadotroph adenomas among non-functioning pituitary adenomas. Conversely, an abnormal response of free LH beta to TRH occurs in about a third of patients with low/normal basal gonadotrophins unresponsive to TRH stimulation.

Human nonfunctioning pituitary adenomas (NFPA) may produce CG in addition to the classical glycoprotein hormones (LH, FSH, and TSH). The aim of the present study was to localize LH beta, FSH beta, TSH beta, alpha-subunit (alpha SU), CG, and its beta-subunit (beta SU) in NFPA using a highly specific immunohistochemical technique. Nine NFPA, obtained at surgery, were processed for both electron microscopy and immunohistochemistry. Three tumors resulted oncocytomas, and six were null cell. Using an immunofluorescence technique, all tumors were positive for at least one glycoprotein; in particular, seven adenomas were markedly positive for CG beta, whereas only two were positive for the intact CG. No association among LH beta, FSH beta, and CG beta positivity could be demonstrated in the different adenomas. In the seven tumors positive for both CG beta and alpha SU, double fluorescence labeling demonstrated that six cases localized CG beta and alpha SU in different cells, but only one tumor showed the two subunits colocalized in the same cells. These data confirm that pituitary tumors synthesize both alpha SU and beta SU of glycoprotein hormones; in particular, the present study indicates that the majority of NFPA is able to synthesize CG, particularly its beta SU. Moreover, the localization of CG beta and alpha SU in different tumoral cells might account for the preferential expression of beta SU and alpha SU instead of the intact hormonal molecules in NFPA.

The aim of this study was to determine the effect of long-term in vitro exposure to high glucose on the release and content of proinsulin and insulin in human islets. After 48 h culture in CMRL medium at 5·5 mM (control islets) and 16·7 mM glucose (experimental islets), islets were perifused and acutely stimulated with 16·7 mM glucose, followed by 3·3 mM glucose. Compared with control islets, experimental islets showed a higher basal release of true insulin and proinsulin-like-molecules (PLM), with no increase of true insulin and PLM release in response to 16·7 mM glucose, and a paradoxical true insulin release in response to 3·3 mM glucose; the PLM/total insulin ratio increased significantly after 16·7 mM glucose. Moreover these islets showed a decreased true insulin content and an increased PLM/total insulin ratio. Quantitative ultrastructural analysis of granules, supported by double gold immunostaining with monoclonal antibodies against proinsulin and insulin, showed an increased proinsulin to insulin ratio in -cells from experimental islets. These data support in vitro what was recently shown in vivo, and further confirm that culture in high glucose is a useful tool to mimic the effect of in vivo chronic hyperglycemia on human -cell function.