Heterogeneity of pituitary folliculo-stellate cells: implications for interleukin-6 production and accessory function in vitro.

The population of folliculo-stellate (FS) cells of the rat anterior pituitary has been shown to be ultrastructurally and immunohistochemically heterogeneous. Based on the overlap of ultrastructural characteristics, the localization in the anterior pituitary and the co-expression within the same cel of the S-100 protein (a marker for FS cells) and MHC-class II determinants (an immune marker) we concluded that a partial overlap exists between the population of FS cells and the monocyte-derived dendritic cells (DC). In this report we describe that interleukin-6 (IL-6) immunoreactivity was found in situ in stellate cells of the rat, mouse and human anterior pituitary at a very low density (< 1% of the cells); the topography was reminiscent of the distribution of FS cells. In the present study we also analyse three different pituitary cell separation methods, in order to study the functional heterogeneity of the FS cells in vitro, and to verify whether functionally distinct subpopulations exist within the FS cell group. Production of bioactive IL-6 was measured in conditioned media of rat anterior pituitary cells separated by (i) bovine serum albumin (BSA) gradient sedimentation at 1 g, (ii) Nycodenz gradient and (iii) a magnetic cell separation (MACS) technique. Production of bioactive IL-6 by cell cultures of 1 to 4 days was correlated with the proportional number of S100 immunoreactive and S100 producing cells, but was not correlated with the proportional number of MHC-class II expressing (OX6-positive) dendritic cells (DC). The distribution pattern of OX6-positive DC was found to partly overlap with the distribution pattern of S100-positive cells in the BSA gradient. Co-sedimentation of S100-positive FS cells and MHC-class II-expressing DC was not restricted to the top fractions of the BSA gradient, but was also found in the low density Nycodenz fraction. MACS separation of the rat anterior pituitary cells resulted into a population enriched in OX6 and OX62 positive DC and a population devoid of such cells, while S100+ cells were equally divided into these two subpopulations. Although there was a significantly decreased production of IL-6 as compared to that of an original pituitary cell population, both MACS separated populations were equal in IL-6 production. The diminution in IL-6 production in both populations may be the result of an impediment of paracrine communication due to the MACS separation into these two populations. Our data also show that a subpopulation of FS cells was capable of stimulating T cell proliferation in vitro. Concomitantly with the distribution pattern of S100- and OX6-immunoreactive cells in the BSA and Nycodenz gradient fractions, we found a similar pattern of stimulation of T cell proliferation. Unlike the IL-6 production pattern, the T cell stimulating capacity was present in the MHC-class II-enriched cell population but absent in the MHC-class II-depleted cell population. These findings-together with earlier in situ histochemical data-suggest that there is an OX6+ S100- subpopulation of FS cells in the anterior pituitary that in itself is capable of stimulating T cell proliferation in vitro, and acts as lymphoid DC. There is also an S100+ OX6- population that is unable to stimulate T cell proliferation in vitro. Both populations are able to produce IL-6, but probably need stimuli from other subpopulations of pituitary cells (or exogeneous stimuli) to produce maximal amounts of IL-6.