DIFFERENTIAL CONTROL OF CONNEXIN-32 AND CONNEXIN-43 EXPRESSION IN THYROID EPITHELIAL-CELLS - EVIDENCE FOR A DIRECT RELATIONSHIP BETWEEN CONNEXIN-32 EXPRESSION AND HISTIOTYPIC MORPHOGENESIS

Thyroid epithelial cells cultured either as a monolayer or in the form of follicles, rapidly reconstitute functional gap junctions (Gj). We previously reported that the thyroid Gj gating is regulated by TSH. We have now performed molecular analyses of Gj proteins 1) to detect the connexin(s) (Cx) that is expressed in thyroid epithelial cells, 2) to determine whether the expression of Ox is hormonally regulated, and 3 ) to analyze the relationship between Cx expression and histiotypic mo rphogenesis, i.e. folliculogenesis. Studies were carried out on thyroc ytes freshly isolated from the gland and on corresponding thyrocytes a fter 1-7 days in culture as monolayers or in the form of reconstituted follicles. The Cx gene transcription products were analyzed by Northe rn blot using specific complementary DNA probes for Cx26, Cx32, and Cx 43. Cx proteins were identified and estimated by Western blot and indi rect immunofluorescence using polyclonal antipeptide antibodies. Cx32 and Cx43 proteins and their corresponding messenger RNA (mRNA) were de tected in thyrocytes freshly isolated from the gland. Thyrocytes conta ined a high amount of the 1.8-kilobase Cx32 mRNA and only traces of th e 3-kilobase Cx43 transcript. No Cx26 transcripts could be detected. T hyrocytes cultured at a density of 0.2-0.5 x 10(6) cells/cm(2) in the absence of TSH formed monolayers. Surprisingly, monolayer cells lost C x32 protein within 24 h, and their Cx32 mRNA content decreased from hi gh to barely detectable levels; Cx32 protein was no longer detected th roughout the 1-week culture period. On the contrary, Cx43 mRNA and Cx4 3 protein rapidly increased in monolayer cells to reach very high leve ls within 2-4 days. Thyrocytes cultured at the same density, but in th e presence of TSH also rapidly lost Cx32, but as soon as they reorgani zed into follicular structures, reexpressed Cx32 at a level (in terms of protein and mRNA) comparable to that found in cells freshly extract ed from the gland. As observed for cell monolayers, reconstituted foll icles overexpressed Cx43. The Cx43 protein and Cx43 mRNA contents of c ultured thyrocytes were 20- to 50-fold higher than those found in isol ated thyrocytes at the outset of culture. When thyrocytes were culture d with TSH, but at a low density (<0,2 x 10(6) cells/cm(2)) to prevent follicle formation, a TSH-dependent increase in Cx43 was observed in monolayer cells. However, TSH did not cause any reexpression of Cx32. When thyrocytes were cultured without TSH but at a high density (>0.8 X 10(6) cells/cm(2)), follicular structures called spontaneous follicl es were obtained, and these structures were found to reexpress Cx32. T he Cx32 expression level, however, remained lower than that observed i n TSH-induced follicles. In conclusion, thyroid epithelial cells coexp ress Cx32 and Cx43. We demonstrate the existence of distinct regulator y mechanisms for Cx32 and Cx43. Cx43 is expressed in thyroid cells in culture undergoing, or not, histiotypic morphogenesis, whereas the exp ression of Cx32 is directly related to the establishment of the thyroi d-specific morphological phenotype, i.e. folliculogenesis. TSH exerts a positive regulatory action on the expression of Cx43 and Cx32.