505 Effects of proinflammatory cytokines on anterior pituitary 5-deiodinase type I and type II

A Baur, K Bauer1, H Jarry2 and J Köhrle Abteilung für Molekulare Innere Medizin und Klinische Forschergruppe der Medizinischen Poliklinik der Universität Würzburg, Röntgenring 11, D-97070 Würzburg, Germany 1 Max-Planck-Institut für Experimentelle Endokrinologie, Feodor-Lynen-Str. 7, 30625 Hannover, Germany 2Universitäts-Frauenklinik, Robert-Koch-Str. 40, 37075 Göttingen, Germany (Requests for offprints should be addressed to J Köhrle, Medizinische Poliklinik, Abteilung für Molekulare Innere Medizin und Klinische Forschergruppe, Röntgenring 11, D-97070 Würzburg, Germany)

Abstract Cytokines are locally produced in the anterior pituitary treated with these cytokines for 24 h. When TtT/Gf cells and act through para-/autocrine mechanisms to modulate in the coculture model were treated with IL-1, TNF cell growth and hormone production. 5-Deiodinases type I and IL-6, no effect on D1 or D2 activities in GH3 cells (D1) and type II (D2) are also expressed in the anterior was observed. pituitary and play an integrative role in the regulation of In male, adult rats a single LPS injection (i.p.) stimu- hormone production and pituitary feedback. D1 activity is lated D2 and D1 activities in the anterior pituitary, and known to be regulated by proinflammatory cytokines in decreased liver D1 activities and serum TSH levels. In liver and . Therefore, we examined the effects of vitro, LPS stimulation of the coculture model of GH3 and IL-1, IL-6 and TNF on 5-deiodinase activities in FS cells also increased D1 activity. reaggregates of rat anterior pituitaries and rat somato- Electrophoretic mobility shift assays (EMSAs) revealed mammotroph GH3 cells cultured alone, or in a bicameral that IL-1 and TNF activate the transcription factor culture system together with the murine folliculo-stellate NFB in reaggregates of rat anterior pituitaries and in (FS) cell line TtT/Gf. In reaggregate cultures of rat TtT/Gf cells cultured alone or cocultured with GH3 cells. anterior pituitaries IL-1 stimulated D1 and D2 dose- Taken together, these findings imply that in anterior dependently and D2 activity was increased by TNF. pituitary cells 5-deiodinase activities are stimulated by When GH3 cells were cocultured with TtT/Gf cells, D2 locally produced cytokines in a para-/autocrine manner activities were 2·3-fold lower than in GH3 cells cultured but cell types other than FS cells seem to mediate some of alone. TNF (50 ng/ml) and IL-6 (100 U/ml) stimulated the effects. D2 in GH3 cells when the cells were cultured alone and Journal of Endocrinology (2000) 167, 505–515

Introduction an important integrative role in the regulation of hormone production and pituitary feedback regulatory mechanisms Cytokines are known to be multipotent mediators of the (Köhrle 1996, 1999, Köhrle et al. 1995, Braverman 1994). bidirectional interaction between the immune and endo- It is known that D1 is modulated by cytokines but the crine system. In the anterior pituitary a cytokine network reports are controversial depending on species, method of of agonists, receptors and antagonists exists and pituitary cytokine application and organs concerned (Ozawa et al. functions are modulated by cytokines (for review see 1988, Pang et al. 1989, Tang et al. 1995, Pekary et al. Besedovsky & Ray 1996, Ray & Melmed 1997). Thyroid 1994, Ongphiphadhanakul et al. 1994, Davies et al. 1997). hormones also play an important role for various functions Therefore, we examined the effects of interleukin-1 of the anterior pituitary. Tri-iodothyronine (T3), the main (IL-1), interleukin-6 (IL-6) and tumor necrosis factor  active form of , is locally generated (TNF) on pituitary deiodinase activities in reaggregates of   by 5 -deiodination of -thyroxine (T4) via two distinct rat APs since maintenance of cell–cell communication and deiodinase isoenzymes (D1 and D2) (Ko¨hrle 1994). D2 is interaction is known to affect their functional and regu- involved in thyroid-stimulating hormone (TSH) feedback latory properties in comparison to monolayer cultures regulation especially under hypothyroid conditions (Denef et al. 1989, Rawlings & Hezareh 1996). Folliculo- (Köhrle 1999, Larsen et al. 1981). Apart from D2, D1 is stellate (FS) cells were proposed to modulate pituitary also expressed in the anterior pituitary (AP) and may play hormone release in a paracrine manner and thus to act as

Journal of Endocrinology (2000) 167, 505–515 Online version via http://www.endocrinology.org 0022–0795/00/0167–505  2000 Society for Endocrinology Printed in Great Britain

Downloaded from Bioscientifica.com at 09/25/2021 04:33:35AM via free access 506 A BAUR and others · Effects of cytokines on anterior pituitary

a link between the hormone and the cellular immune 300 W) in ice-cold homogenization buffer (250 mM system (Baes et al. 1987, Ferrara et al. 1987, Allaerts & sucrose, 20 mM HEPES, 1 mM ethylenediamine tetra- Denef 1989). To study the effects of proinflammatory acetic acid, 1 mM dithiothreitol (DTT), pH 7·0). For cytokines we established cocultures of rat somatomammo- EMSAs, reaggregates from male adult Sprague–Dawley troph GH3 cells which express functional 5-deiodinase rats were incubated with IL-1 (10 ng/ml) and TNF (Baur et al. 1997) with TtT/Gf cells, a murine (50 ng/ml) for 2·5 h and nuclear extracts were prepared as cell line (Inoue et al. 1992) that shares typical character- described by Grandison et al. (1994). istics of FS cells, including the ability to produce IL-6 (Renner et al. 1997). As systemically administered LPS stimulates the expression of IL-1 and TNF in the Cell culture pituitary of rats (Whiteside et al. 1999) we examined GH3 cells (ATCC No.CCL. 82·1) and TtT/Gf cells the effects of a single LPS injection (i.p.) on pituitary (Inoue et al. 1992) were routinely cultured in Ham’s F10 deiodinase expression in male, adult rats. medium supplemented with 15% horse serum, 2·5% fetal In the pituitary, the exact mechanisms of cytokine calf serum and antibiotics (penicillin 10 U/ml, strepto- action on the expression of hormones and pituitary-specific mycin 0·1 mg/ml). Cells were maintained at 37 Cina genes are still unclear. IL-1 and TNF are known to humidified atmosphere of 5% CO and 95% air. For mediate some of their effects by activating the intracellular 2 ‘coculture’ experiments GH3 cells were plated in six- transcription factor NFB (Baeuerle & Baltimore 1996, well tissue culture plates and TtT/Gf cells were plated Hill & Treisman 1995). Activated transcription factors are in cell culture inserts (0·4 µm pore size, PET ‘track- translocated to the nucleus and can directly alter gene etched’ membrane, Falcon, Becton Dickinson GmbH, expression by binding to distinct DNA elements. There- Heidelberg, Germany). The inserts were then placed in fore, we used electrophoretic mobility shift assays (EMSAs) the six-well plates containing the GH3 cells. Thus, the to examine whether NFB is also activated by these cells were grown in a two compartment system with GH3 cytokines in reaggregates and in TtT/Gf cells. cells at the bottom and TtT/Gf cells on top, separated by a permeable membrane. As control, inserts with TtT/Gf cells were placed in six-well plates containing only Materials and Methods medium. The cells were incubated until TtT/Gf cells achieved confluence with maximal transepithelial resist- Chemicals ance. Then the medium of both compartments was All chemicals were of biochemical or analytical grade. Cell replaced by fresh medium supplemented with the respec- culture media, fetal calf serum, horse serum and antibiotics tive test reagents. At the indicated time points cells were were purchased from Gibco BRL-Life Technologies washed once with ice-cold PBS, GH3 cells were scraped GmbH (Eggenstein, Germany). Additives for the into ice-cold homogenization buffer and sonicated as serum-free culture medium and other chemicals were described above. Nuclear cell extracts of TtT/Gf cells purchased from Boehringer (Mannheim, Germany), were prepared as described by Grandison et al.(1994). For Merck (Darmstadt, Germany), Sigma (Munich, Germany) some nuclear extracts TtT/GF cells were cultivated in and Baker Chemikalien (Gro-Gerau, Germany). Re- 25 cm2 culture flasks and for experimental studies under combinant human (rh) cytokines were purchased from serum-free conditions cells were maintained in serum-free PBH (Hannover, Germany). A NFB consensus-site medium (DMEM-F12 (1:1), 15 mM HEPES, 15 mM oligonucleotide was purchased from Promega (Madison, TES, 20 µM ethanolamine, 25 nM sodium selenite, WI, USA). Antibodies were purchased from Santa Cruz 5 mg/l insulin, 0·9 g/l BSA, 40 mg/l transferrin, 1 mg/l Biotechnology (Santa Cruz, CA, USA). , 10 mM ethanol, 35 mg/l penicillin, 50 mg/l streptomycin) for 24 h. On the day of the experiment medium was replaced by fresh medium supplemented Reaggregate cultures of anterior pituitary cells with test reagents as indicated. Anterior pituitaries were removed from adult male Cara rats after decapitation and processed according to the Animal treatment protocol of Denef et al. (1989). After 2 days in culture reaggregates from two 35 mm petri dishes were transferred Male, adult Sprague–Dawley rats were kept under stan- to one 60 mm petri dish and half the serum-free medium dard housing conditions (12 h day/night cycle) and re- was replaced by 2 volumes of fresh medium. After 5 days, ceived a single injection of LPS (0·5 mg/kg or 2·5 mg/kg half the medium was removed and substituted by fresh BW) or vehicle (0·9% NaCl) i.p. and were killed by medium containing test reagents as indicated. After 48 h decapitation at the indicated time points (1 h, 4 h, 6 h, the reaggregates were harvested, washed once with ice- 24 h and 48 h ). Blood was collected from every animal for cold PBS and homogenized by sonication (10 times, 0·5s, serum preparation. Serum was stored at 80 C until use.

Journal of Endocrinology (2000) 167, 505–515 www.endocrinology.org

Downloaded from Bioscientifica.com at 09/25/2021 04:33:35AM via free access Effects of cytokines on anterior pituitary · A BAUR and others 507

Preparation of pituitary and liver homogenates antibody (1 mg/ml) was added to the nuclear extracts prior to the addition of labeled oligonucleotide. After decapitation pituitaries were removed, AP and posterior pituitary (including intermediate lobe) were separated and immediately frozen on dry ice blocks. Livers Statistical analysis were removed and immediately frozen in liquid nitrogen. Data were analyzed by multiple ANOVA. When the Tissues were stored at 80 C until use. Single APs were main effect was significant, the Mann–Whitney U test homogenized by sonication as described above, fragments was applied as the posthoc test to determine individual of single livers were homogenized using a glass–teflon differences between means. potter (Braun, Melsungen, Germany). The protein con- tents of the homogenates were determined by a modified Bradford protein assay (Biorad, Munich, Germany) using Results gamma-globulin as protein standard (Bradford 1976). Effects of IL-1, IL-6 and TNF on 5-deiodinase activities in reaggregate cultures of rat anterior pituitaries Biochemical assays In AP reaggregates of male adult rats the incubation with Specific activities of D1 and D2 were determined in 100 ng/ml IL-1 for 48 h led to a 1·8-fold stimulation of 125 -   125 parallel by the release of I from 3,3 ,5 -[ I]-tri- D1. The activities of D1 after incubation with IL-6 and iodothyronine (rT3, DuPont, Bad Homburg, Germany) TNF were always higher than control but this increase in the absence or presesence of 1 mM PTU (6-n-propyl- did not reach values of significance due to variations 2-thio-uracil) using 10 nM non-radioactive rT3 for pitu- between the individual wells (Fig. 1a). The activity of D2 itary and cell-culture homogenates and 10 µM for liver was significantly enhanced both by 100 ng/ml IL-1 homogenates (Leonard & Rosenberg 1980). The fraction (5·9-fold over basal levels) and by 25 ng/ml TNF (4·6- of iodide release blocked by 1 mM PTU was assigned to fold over basal levels) (Fig. 1b). The effect of TNF was D1, the residual activity not inhibited by PTU was biphasic. 100 ng/ml TNF did not stimulate D2 activity   ascribed to D2. 5 -Deiodinase activities of each sample and IL-6 exerted no effect at either concentration. were determined in triplicates and expressed as fmol 125I- released per min, per mg protein. Effects of IL-1, IL-6 and TNF on deiodinase activities in rat somatomammotroph GH3 cells cocultured with TtT/Gf Determination of TSH cells Serum TSH levels were determined by RIA using Coculture of hormone-producing cells with FS-cells is materials supplied by the National Hormone and Pituitary known to modify their response to cytokines. Treatment Program (Rockville, MD, USA). The rat-TSH antiserum with IL-1 for 24 h did not alter D1 activity in GH3 cells was rTSH-S-5. For iodination rTSH-I-9 was used. TSH (Fig. 2a) cultured alone (filled bars) or together with FS levels were expressed in relation to the reference prep- cells (shaded bars) in the presence of serum, irrespective of aration rTSH-RP-2. Cross-reactivity with other pituitary whether IL-1 was added directly to GH3 cells alone hormones was less then 0·1%. Sensitivity of the RIA (third and fifth column) or to FS cells in the coculture was 0·2 ng/ml with an intra- and interassay variance of model (fourth and sixth column). However, coculturing 6·5% and 10·9% respectively. The RIA was performed GH3 cells with TtT/Gf cells led to a 2·3-fold inhibition of according to the instructions of the provider. D2 activity compared with the activity in GH3 cells cultured without TtT/Gf cells (Fig. 2b). In GH3 cells that were cultured alone, no direct significant effect of either of Electrophoretic mobility shift assays (EMSA) the IL-1 concentrations on D2 was observed (Fig. 2b, The protein concentrations of the nuclear extracts were fourth and sixth column). In a second series of exper- measured as described above with 6–15 µg being used for iments, coculturing GH3 cells with TtT/Gf cells again led EMSAs (equal amounts of protein for a single experiment). toa2·3-fold inhibition of D2 in GH3 cells (Fig. 3b, second 32 Oligonucleotides were labeled with [ P]-ATP using T4 column) compared with activities in GH3 cells cultured polynucleotide kinase. Binding reactions were performed alone (Fig. 3b, first column). No effect on D1 in GH3 for 30 min at room temperature in reaction buffer de- cells was observed after TNF (50 ng/ml) and IL-6 scribed by Grandison et al. (1994). Labeled oligonucleotide (100 U/ml) treatment of TtT/Gf cells or direct treatment (20 000 c.p.m.) was used for each binding reaction. The of GH3 cells for 24 h (Fig. 3a). However, TNF as well as sequence of the NFB oligonucleotide is as follows (Sen IL-6 stimulated D2 activities (2-fold and 1·5-fold respect- & Baltimore 1986): 5-AGTTGAGGGGACTTTCCC ively) compared with the control when GH3 cells were AGGC-3. Supershifts were performed with an antibody cultured alone, and directly treated with these cytokines directed against the NFB subunit p65. 1 µl of the for 24 h (Fig. 3b). D2 activity of GH3 cells was not altered www.endocrinology.org Journal of Endocrinology (2000) 167, 505–515

Downloaded from Bioscientifica.com at 09/25/2021 04:33:35AM via free access 508 A BAUR and others · Effects of cytokines on anterior pituitary

Figure 1 Effects of IL-1, IL-6 and TNF on D1 (a) and D2 (b) activities in reaggregates of rat anterior pituitaries 48 h after Figure 2 Effects of IL-1 (10 ng/ml and 100 ng/ml) on D1 (a) and  application. Values are means S.E.M. of duplicates of two D2 (b) activity in GH3 cells cultured alone (solid bars) or < independent experiments (n=4). *P 0·05. cocultured with TtT/GF cells (shaded bars) as described in Materials and Methods 24 h after application. In the coculture in the coculture model either when the TtT/Gf cells were model TtT/Gf cells were treated with IL-1. Values are meansS.E.M. of triplicates of three independent experiments treated with TNF, or with IL-6, for 24 h (Fig. 3b). (n=9). *P<0·005. Effects of LPS on deiodinase activities in rat somatomammotroph GH3 cells co-cultured with TtT/Gf cells cantly stimulated compared with control (Fig. 4a, solid When GH3 cells were cultured alone and treated with bars). However, LPS treatment of the TtT/Gf cells that LPS (2 µg/ml) for 6 h D1 activity was not signifi- were cocultured with the GH3 cells led to a significant

Journal of Endocrinology (2000) 167, 505–515 www.endocrinology.org

Downloaded from Bioscientifica.com at 09/25/2021 04:33:35AM via free access Effects of cytokines on anterior pituitary · A BAUR and others 509

activities in GH3 cells that were cocultured with TtT/Gf cells for another 24 h, were not changed by LPS treatment (Fig. 4d).

Effects of a single LPS injection in vivo in male, adult rats A single injection of LPS (2·5 mg/kg BW) i.p. in male, adult rats led to a transient stimulation of D1 in the anterior pituitary. One hour after LPS application, specific D1 activities were 1·5-fold higher than in APs of untreated controls but as soon as 4 h after LPS treatment, D1 activities in the AP reached basal levels (Fig. 5a). Four hours after a single LPS injection (2·5 mg/kg BW) serum TSH levels and D1 activities in the liver were decreased (to 37% and 64% of untreated controls, respectively). TSH levels were still slightly, but not significantly decreased 6 h after LPS treatment (data not shown) and reached basal levels after 24 h (Fig. 5c). Specific D2 activities in the APs were stimulated by LPS (2·2-fold compared with un- treated controls) 24 h after application (Fig. 5b). At 1 h and 4 h after LPS application D2 activities in the AP were not changed compared with control. 6 h after treatment, D2 activities were slightly increased, but this stimulation was not significant (data not shown). 48 h after treatment D2 activities reached basal levels again (data not shown). The effects of LPS were dose dependent. A single injection of 0·5 mg/kg BW LPS had similar effects but to a lesser extent (data not shown).

EMSAs with nuclear extracts of reaggregates EMSAs were performed in order to identify DNA- binding proteins which, after activation by cytokines, might transduce these signals in reaggregates and TtT/Gf cells. Nuclear extracts of reaggregates formed a specific protein–DNA complex with a radiolabeled NFB oligo- nucleotide (Fig. 6, lane 1). Incubation of reaggregates with IL-1 (10 ng/ml) (Fig. 6, lane 2) and TNF (50 ng/ml) (Fig. 6, lane 3) for 2·5 h led to an activation of NFB compared with control. NFB activation induced by TNF was stronger than when induced by IL-1. The addition of an antibody directed against the NFB subunit p65 resulted in the formation of a supershifted band (Fig. 6, lane 4). Addition of excess unlabeled oligonucleotide led Figure 3 Effects of TNF (50 ng/ml) and IL-6 (100 U/ml) on D1 to the complete disappearance of the retarded band (Fig. 6, (a) and D2 (b) activity in GH3 cells cultured alone (solid bars) or cocultured with TtT/GF cells (shaded bars) as described in lane 5). Supershifting and competition by unlabeled oligo- Materials and Methods 24 h after application. In the coculture nucleotide ensured specificity of the retarded band. model TtT/Gf cells were treated with the cytokines. Values are meansS.E.M. of triplicates of two independent experiments (n=6). *P<0·05; **P<0·005; ***P<0·0005. EMSAs with nuclear extracts of TtT/Gf cells In TtT/Gf cells, incubation with IL-1 (100 ng/ml) for 90 min activated NFB-DNA binding (Fig. 7, lane 2) stimulation of D1 activity in GH3 cells (Fig. 4a, shaded compared with untreated controls (Fig. 7, lane 1) in the bars). No effect on D2 activities was observed after presence of serum as well as under serum-free conditions treatment with LPS for 6 h (Fig. 4b). The LPS effect on (Fig. 7, lane 3: control, lane 4: IL-1). Activation of NFB D1 activity was transient – 24 h after LPS treatment, D1 by IL-1 under serum-free conditions was as strong as activity reached basal levels (Fig. 4c). Decreased D2 in the presence of serum. Addition of excess unlabeled www.endocrinology.org Journal of Endocrinology (2000) 167, 505–515

Downloaded from Bioscientifica.com at 09/25/2021 04:33:35AM via free access 510 A BAUR and others · Effects of cytokines on anterior pituitary

Figure 4 Effects of LPS (2 g/ml) on D1 (a, c) and D2 (b, d) in GH3 cells cultured alone (solid bars) or cocultured with TtT/Gf cells (shaded bars) 6 h and 24 h after application. In the coculture model TtT/Gf cells were treated with LPS. Values are meansS.E.M. of triplicates of three independent experiments (n=9). *P<0·05; **P<0·005.

oligonucleotide led to disappearance of the retarded band Discussion (Fig. 7, lane 5) and incubation with an antibody against the NFB subunit p65 produced a supershifted band (Fig. 7, Proinflammatory cytokines exert effects on liver and lane 6). thyroid D1, although the effects reported are controversial. TNF, a well-known activator of NFB, was effective In rat FRTL-5 thyroid cells, basal and TSH-induced D1 in TtT/Gf cells. When the cells were incubated with activity is inhibited by IL-1, IL-6 and TNF, and D1 50 ng/ml TNF for 24 h, activation of NFkB was ob- mRNA levels are reduced by IL-1 and IL-6, but not by served compared with the control (Fig. 8, lanes 1, 2: TNF (Hashimoto et al. 1995). In mouse liver D1 is control; lanes 3, 4: TNF 25 ng/ml). Again the addition of stimulated by IL-1 (Fujii et al. 1989) but is not affected by an antibody against the NFB subunit p65 resulted in the TNF (Ozawa et al. 1988). However, in rat liver D1 is formation of a supershift (Fig. 8, lane 5) but not the inhibited by TNF (Pang et al. 1989, Tang et al. 1995) addition of an antibody against AP1 (Fig. 8, lane 6). This in vivo but stimulated by TNF, IL-1 and IL-6 in vitro in ensures specificity of the p65 antibody. rat liver 1 cells (Davies et al. 1997). These findings imply When TtT/Gf cells that were cocultured with GH3 that proinflammatory cytokines exert different effects on cells were treated with TNF for 24 h, the same activation D1 in different organs, different species and different of NFB occurred (Fig. 9, lanes 1,2: control; lanes 3, 4: models. In this study we now demonstrate that D1 as well TNF 50 ng/ml) as in TNF-treated TtT/Gf cells cul- as D2 activities are modulated by proinflammatory cyto- tured alone. There were also no differences in NFB kines in reaggregates of rat APs and rat somato- activation by IL-1 between cocultured TtT/Gf cells and mammotroph GH3 cells. In reaggregates and GH3 cells TtT/Gf cultured alone (data not shown). Treatment with D1 and D2 are expressed, although the activities of D2 are IL-6 (100 U/ml) for 24 h had no effect on NFB an order of magnitude lower. Moreover in GH3 cells that activation, either in coculture experiments or in TtT/Gf were cocultured with TtT/Gf cells, D2 activity was cells cultured alone. further decreased.

Journal of Endocrinology (2000) 167, 505–515 www.endocrinology.org

Downloaded from Bioscientifica.com at 09/25/2021 04:33:35AM via free access Effects of cytokines on anterior pituitary · A BAUR and others 511

Figure 5 Effects of a single LPS injection (i.p.) on D1 activities in the anterior pituitary (AP) and the liver, serum TSH levels and D2 activities in the AP of male, adult rats at the indicated time points after injection. LPS dose was 2·5 mg/kg body weight. Values are meansS.E.M. (n=6).*P<0·05.

www.endocrinology.org Journal of Endocrinology (2000) 167, 505–515

Downloaded from Bioscientifica.com at 09/25/2021 04:33:35AM via free access 512 A BAUR and others · Effects of cytokines on anterior pituitary

Figure 7 EMSA of nuclear extracts of TtT/Gf cells with a 32P-labeled oligonucleotide representing a NFB consensus site. Lane 1: control, TtT/Gf cells cultured in the presence of serum; lane 2: IL-1 (100 ng/ml) 90 min., TtT/Gf cells cultured in the presence of serum; lane 3: control, TtT/Gf cells cultured under serum-free conditions; lane 4: IL-1 (100 ng/ml) 90 min., TtT/Gf cells cultured under serum-free conditions; lane 5: IL-1 (100 ng/ml) 90 min., TtT/Gf cells cultured under serum-free conditions, nuclear extract incubated with an excess of unlabeled oligonucleotide; lane 6: IL-1 (100 ng/ml) 90 min., TtT/Gf cells cultured under serum-free conditions, nuclear extract incubated with an antibody directed against NFB subunit p65.

Figure 6 EMSA of nuclear extracts of reaggregates with a 32P-labeled oligonucleotide representing a NFB consensus site. Lane 1: control; lane 2: IL-1 (10 ng/ml, 2·5 h); lane 3: TNF (50 ng/ml, 2·5 h); lane 4: TNF (50 ng/ml, 2·5 h), cell extract incubated with an antibody directed against NFB subunit p 65; lane 5: TNF (50 ng/ml, 2·5 h), nuclear extract incubated with an excess of unlabeled oligonucleotide.

The effect of TNF on D2 activity in reaggregates was biphasic, producing stimulation at lower, but not at higher concentrations. A biphasic response to TNF has also been described for several endpoints in vitro, e.g. superoxide generation in chondrocytes (Ahmadzadeh et al. 1990), arachidonic acid release of monocytes (Nokta et al. 1995) and cell proliferation of osteoblasts (Frost et al. 1997). Folliculo-stellate cells that do not produce hormones are known to influence hormone-producing AP cells by Figure 8 EMSA of nuclear extracts of TtT/Gf cells with a secreting humoral factors (Baes et al. 1987) including IL-6 32P-labeled oligonucleotide representing a NFB consensus site. (Vankelecom et al. 1989, 1993), and to mediate the effects TtT/Gf cells were cocultured with GH3 cells as described in materials and methods. Lanes 1 and 2: control; lanes 3 and 4: of INF on hormone secretion in rat AP cells (Vankelecom TNF (50 ng/ml) 24 h; lane 5: TNF (50 ng/ml) 24 h, nuclear et al. 1992). Murine TtT/Gf cells (Inoue et al. 1992) share extract incubated with an antibody directed against NFB subunit many typical characteristics of FS cells and therefore p 65; lane 6: TNF (50 ng/ml) 24 h, nuclear extract incubated with provide a good model for in vitro studies of paracrine an antibody directed against AP1. interactions between hormone-producing cells of the AP and the FS cells. In addition, the reaggregate culture model of the rat AP combines several hormone-producing cell–cell interactions since in the bicameral coculture and FS cells. The effects of TtT/Gf cells on D2 activity is model used, direct contact between the two cell lines is mediated by humoral, paracrine factors and not by direct not possible. TtT/Gf cells in our model produced IL-6,

Journal of Endocrinology (2000) 167, 505–515 www.endocrinology.org

Downloaded from Bioscientifica.com at 09/25/2021 04:33:35AM via free access Effects of cytokines on anterior pituitary · A BAUR and others 513

on deiodinase activities. Reaggregates of the AP provide a more physiological model since the three-dimensional, tissue-like organization provides cell–cell communication and para-/autocrine interactions between different cell types of the AP are facilitated. Cell–cell contact between different AP cell types has recently been shown to enable communication by gap junctions and rapid signal trans- duction between cells (Morand et al. 1996). Furthermore, reponsiveness of AP cells to various signals and gene expression is modulated by cells in their neighbourhood (Abraham et al. 1996). LPS treatment of male, adult Sprague–Dawley rats showed that components involved in thyroid hormone metabolism were changed. D1 activites in the liver were decreased. It is thought that decreased D1 activities in the liver play an important role in the pathogenesis of the low T3 syndrome observed in nonthyroidal illness (Docter et al. 1993). The first change after LPS treatment, however, was a transient stimulation of D1 activities in the anterior pituitary, preceeding changes in serum TSH levels and liver D1 activities. Changes in D1 activity in the AP could be due to the action of locally produced cytokines and these changes might be involved in subsequent TSH suppression. IL-1 mRNA is rapidly induced in anterior pituitaries of rats, with maximal levels at 0·5–1 h after LPS application followed by the induction of mRNA of IL-1 Figure 9 EMSA of nuclear extracts of TtT/Gf cells with a converting , IL-1 receptor antagonist and TNF 32P-labeled oligonucleotide representing a NFB consensus site. 2 h after LPS treatment (Whiteside et al. 1999, Quan et al. TtT/GF cells were cultured alone in membrane inserts as 1998). Stimulation of the D2 activities 24 h after LPS described in material and methods. Lanes 1 and 2: control; lanes 3 and 4: TNF (50 ng/ml) 24 h. application could be either due to direct stimulation by locally or systemically produced cytokines, or caused by reduced thyroid hormone levels as a consequence of with IL-1 being a stimulator of IL-6-production (data not inhibitory actions of cytokines on thyroid hormone secre- shown), but it is unlikely that IL-6 produced by TtT/Gf tion. The in vitro experiments on the regulation of de- cells is responsible for the inhibition of D2 activity in GH3 iodinases by proinflammatory cytokines were done with cells, because direct treatment of GH3 cells with IL-6 cell-lines or primary cell culture of APs of rats. Therefore, revealed stimulatory effects. Stimulation of D2 activity by we also used rats for the in vivo experiments, but the IL-1 was observed in reaggregates but not in the cocul- observed effects of a single i.p. LPS injection were not very ture experiments, when either GH3 cells directly received prominent and only transient. Sprague–Dawley rats seem IL-1 or when TtT/Gf cells were treated with the to be quite insensitive towards a single LPS application. cytokine. This implies that IL-1 exerts its effects on D2 Despite the large amount of LPS used, no changes in activity either in cell types of the anterior pituitary other selenoprotein P mRNA levels (Dreher et al. 1997) that are than somato- and mammotroph cells or when the presence decreased in livers of LPS treated mice (I Dreher, personal of other cell types, apart from FS cells, is needed to communication), were observed in rat livers (data not mediate the effects in a paracrine manner. Alternatively shown) and the animals did not show signs of severe illness direct cell–cell contact may be required. in their behaviour. Nevertheless, 5-deiodinase activities Wassen et al. (1996) recently reported that IL-1,as and TSH levels were changed by LPS treatment, suggest- well as TNF, decreases basal but not TRH-induced TSH ing that components involved in thyroid hormone metab- release from AP cells cultured as monolayers in the olism are more sensitive to LPS treatment than other presence of serum. These effects were not due to elevated possible targets. The presence of folliculo-stellate cells ff thyroid hormone uptake or increased T3 nuclear occu- seems to be necessary for the stimulating e ect of LPS on pancy. Moreover, effects on the release of other pituitary D1 activity as D1 activity in GH3 cells is only stimulated hormones were not observed and no clear dose-dependent by LPS if the cells are cocultured with TtT/Gf cells but effects were reported. Our findings imply that reaggre- not when the cells are cultured alone. gates are a more suitable model than monolayer primary Proinflammatory cytokines are locally produced in cultures to examine cytokine-mediated effects in the AP the anterior pituitary and IL-1 is able to induce the www.endocrinology.org Journal of Endocrinology (2000) 167, 505–515

Downloaded from Bioscientifica.com at 09/25/2021 04:33:35AM via free access 514 A BAUR and others · Effects of cytokines on anterior pituitary

production of IL-6 (Koenig et al. 1990, Spangelo et al. Baeuerle PA & Baltimore D 1996 NF-kB: Ten years after. Cell 87 1990, Vankelecom et al. 1989). The IL-6-gene contains a 13–20. Baur A, Bauer K, JarryH&Köhrle J 1997 3,5-diiodo-L-thyronine NFB binding site in its promoter region and IL-6 gene stimulates type 1 5-deiodinase activity in rat anterior pituitaries in expression is activated by NFB (Libermann & Baltimore vivo and in reaggregate cultures and GH3 cells in vitro. Endocrinology 1990). By EMSAs we demonstrated that IL-1 and TNF 138 3242–3248. led to an activation of the transcription factor NFBin Besedovsky HO & Del Rey A 1996 Immune-neuro-endocrine reaggregates of rat AP and in TtT/Gf cells. These findings interactions: facts and hypotheses. Endocrine Reviews 17 6. Bradford MM 1976 A rapid and sensitive method for the quantitation suggest that gene expression induced by IL-1 and TNF of microgram quantities of protein utilizing the principle of is in part mediated via the activation of NFBinthe protein-dye binding. Analytical Biochemistry 72 248–254. anterior pituitary. Braverman LE 1994 Deiodination of thyroid hormones. A 30 year In the present study we demonstrate that in an- perspective. Experimental and Clinical Endocrinology 102 355–363.  Davies PH, Sheppard MC & Franklyn JA 1997 Inflammatory terior pituitary cells, including somatomammotrophs, 5 -  cytokines and type I 5 -deiodinase expression in rat 1 liver cells. deiodinase activities are stimulated by proinflammatory Molecular and Cellular Endocrinology 129 191–198. cytokines in a para-/autocrine manner and the activation Denef C, Maertens P, Allaerts W, Mignon A, Robberecht W, of cytoplasmatic transcription factors suggests that tran- SwennenL&Carmeliet P 1989 Cell-to-cell communication in scriptional modulation is involved in this regulation. peptide target cells of anterior pituitary. Methods in Enzymology 168 Nevertheless, stimulation of D1 activity in the AP after 47–71. Docter R, Krenning EP, De Jong M & Hennemann G 1993 The LPS treatment was very rapid (1 h) and only transient sick euthyroid syndrome: changes in thyroid hormone serum (basal levels were reached after 4 h), suggesting that parameters and hormone metabolism. Clinical Endocrinology 39 mechanisms other than transcriptional regulation, e.g. the 499–518. ff Dreher I, Jakobs TC & Köhrle J 1997 Cloning and characterization of e ects of T3 on D1 in the AP (Köhrle et al. 1995) or new protein synthesis, e.g. the effects of IL-1, TNF and IL-6 the human Selenoprotein P promoter. Response of Selenoprotein P expression to cytokines in liver cells. Journal of Biological Chemistry in human liver cells, are responsible for this stimulation. 272 29364–29371. Up to now the exact mechanism of this stimulation is not Ferrara N, Schweigerer L, Neufeld G, Mitchell R & Gospodarowicz known and remains to be investigated. D 1987 Pituitary follicular cells produce basic fibroblast growth The stimulation of both 5-deiodinase-isoenzymes (D1 factor. PNAS 84 5773–5777. and D2) in the AP by proinflammatory cytokines may play Frost A, Jonsson KB, NilssonO&Ljunggren O 1997 Inflammatory cytokines regulate proliferation of cultured human osteoblasts. Acta a causal role for decreased TSH levels observed during Orthopaedica Scandinavica 68 91–96. non-thyroidal illness. Whether other cell types, such as Fujii T, Sato K, Chol Han D, Kawakami M, TsushimaT&Shizume tanycytes, recently identified as D2 expressing cells (Tu K1989Effects of interleukin 1 (IL-1) on thyroid hormone et al. 1997, Guadano-Ferraz et al. 1997) are additionally metabolism in mice: stimulation by IL-1 of iodothyronine  involved in TSH feedback regulation of thyrotroph cells 5 -deiodination activity (type 1) in the liver. Endocrinology 124 167–174. remains to be studied. Grandison L, Nolan GP & Pfaff DW 1994 Activation of the transcription factor NF-KB in GH3 pituitary cells. Molecular and Cellular Endocrinology 106 9–15. Guadano-Ferraz A, Obregon MJ, St Germain DL & Bernal J 1997 Acknowledgements The type 2 is expressed primarily in glial cells in the neonatal rat brain. PNAS 94 10391–10396. This work was supported by a grant from the Deutsche Hashimoto H, Igarashi N, MiyawakiT&SatoT1995Effects of Forschungsgemeinschaft (Kö 922/3–1/2/3). tumor necrosis factor-, interleukin-1, and interleukin-6 on type I iodothyronine 5-deiodination in rat thyroid cell line, FRTL-5. Journal of Interferon and Cytokine Research 15 367–375. Hill CS & Treisman R 1995 Trancriptional regulation by extracellular References signal: mechanisms and specificity. Cell 80 199–211. Inoue K, Matsumoto H, Koyama C, Shibata K, Nakazato Y & Ito A Abraham EJ, Faught WJ & Frawley LS 1996 Intercellular 1992 Establishment of a folliculo-stellate-like cell line from murine 131 communication: Relative importance of cellular adhesion and thyrotropic pituitary tumor. Endocrinology 3110–3116. paracrine signaling to hormonal gene expression. Endocrinology 137 Koenig JI, Snow K, Clark BD, Toni R, Cannon JG, Shaw AR, 4050–4053. Dinarello CA, Reichlin S, Lee SL & Lechan RM 1990 Intrinsic Ahmadzadeh N, Shingu M & Nobunaga M 1990 The effect of pituitary interleukin-1 is induced by lipopolysacharide. recombinant tumor necrosis factor alpha on superoxide and Endocrinology 126 3053–3058. metalloproteinase production by synovial cells and chondrocytes. Köhrle J 1994 Thyroid hormone deiodination in target tissues - a Clinical and Experimental Rheumatology 8 387–391. regulatory role for the trace element ? Experimental and Allaerts W & Denef C 1989 Regulatory activity and topological Clinical Endocrinology 102 63–89. distribution of folliculo-stellate cells in rat anterior pituitary cell Köhrle J 1996 Thyroid hormone deiodinases - a selenoenzyme family aggregates. Neuroendocrinology 49 409–418. acting as gate keepers to thyroid hormone action. Acta Medica Baes M, Allaerts W & Denef C 1987 Evidence for functional Austriaca 23 17–30. communication between folliculo-stellate cells and hormone Köhrle J 1999 Local activation and inactivation of thyroid hormones: secreting cells in perifused anterior pituitary cell aggregates. the deiodinase family. Molecular and Cellular Endocrinology 151 Endocrinology 120 685–691. 103–119.

Journal of Endocrinology (2000) 167, 505–515 www.endocrinology.org

Downloaded from Bioscientifica.com at 09/25/2021 04:33:35AM via free access Effects of cytokines on anterior pituitary · A BAUR and others 515

Köhrle J, Schomburg L, Drescher S, FeketeE&Bauer K 1995 Rapid Renner U, Gloddek J, Arzt E, Inoue K & Stalla GK 1997 stimulation of type I 5-deiodinase in rat pituitaries by 3,3,5- Interleukin-6 is an autocrine growth factor for folliculostellate-like triiodo-L-thyronine. Molecular and Cellular Endocrinology 108 17–25. TtT/Gf mouse pituitary tumor cells. Experimental and Clinical Larsen PR, Silva JE & Kaplan MM 1981 Relationships between Endocrinology and Diabetes 105 345–352. circulating and intracellular thyroid hormones: physiological and SenR&Baltimore D 1986 Multiple nuclear factors interact with the clinical implications. Endocrine Reviews 2 87–102. immunoglobulin enhancer sequences. Cell 46 705–716. Leonard JL & Rosenberg IN 1980 Iodothyronine 5-deiodinase from Spangelo BL, Isakson PC & MacLeod RM 1990 Production of rat kidney: substrate specificity and the 5-deiodination of reverse interleukin-6 by anterior pituitary cells is stimulated by increased . Endocrinology 107 1376–1383. intracellular adenosine 3,5-monophosphate and vasoactive Libermann TA & Baltimore D 1990 Activation of interleukin-6 gene intestinal peptide. Endocrinology 127 403–409. expression through the NFB transcription factor. Molecular and Tang K, Braverman LE & DeVito WJ 1995 Tumor necrosis factor- Cellular Biology 10 2327–2334. and interferon- modulate gene expression of type I 5-deiodinase, Morand I, Fonlupt P, Guerrier A, Truillas J, Calle A, Remy C, thyroid , and thyroglobulin in FRTL-5 rat thyroid cells. Rousset B & Munari-Silem Y 1996 Cell-to-cell communication Endocrinology 136 881–888. in the anterior pituitary: Evidence for gap junction-mediated Tu HM, Kim S, Salvatore D, Bartha T, Legardi G, Larsen PR & exchanges between endocrine cells and folliculostellate cells. Lechan RM 1997 Regional distribution of type 2 thyroxine Endocrinology 137 3356–3367. deiodinase messenger ribonucleic acid in rat hypothalamus and Nokta MA, Hassan MI, Loesch KA & Pollard RB 1995 HIV-induced pituitary and its regulation by thyroid hormone. Endocrinology 138 TNF- regulates arachidonic acid and PGE2 release from HIV- 3359–3368. infected mononuclear phagocytes. Virology 208 590–600. Vankelecom H, Carmeliet P, Van Damme J, BilliauA&Denef C Ongphiphadhanakul B, Fang SL, Tang K, Patwardhan NA & 1989 Production of interleukin-6 by folliculo-stellate cells of the Braverman LE 1994 Tumor necrosis factor- decreases thyrotropin- anterior pituitary gland in histotypic cell aggregate culture system. induced 5-deiodinase activity in FRTL-5 thyroid cells. European Neuroendocrinology 49 102–106. Journal of Endocrinology 130 502–507. Vankelecom H, Andries M, BilliauA&Denef C 1992 Evidence that Ozawa M, Sato K, Chol Han D, Kawakami M, Tsushima T & folliculo-stellate cells mediate the inhibitory effect of interferon- Shizume K 1988 Effects of tumor necrosis factor-/cachechtin on on hormone secretion in rat anterior pituitary cell cultures. thyroid hormone metabolism in mice. Endocrinology 123 1461–1467. Endocrinology 130 3537–3546. Pang XP, Hershman JM, Mirell CJ & Kary AE 1989 Impairment of Vankelecom H, Matthys P, Van Damme J, Heremans H, Billiau A & hypothalamic-pituitary-thyroid function in rats treated with human Denef C 1993 Immunocytochemical evidence that S-100-positive recombinant tumor necrosis factor- (cachectin). Endocrinology 125 cells of the mouse anterior pituitary contain interleukin-6 76–84. immunoreactivity. Journal of Histochemistry and Cytochemistry 41 PekaryAE,BergL,SantiniF,ChopraI&Hershman JM 1994 151–156. Cytokines modulate type I iodothyronine deiodinase mRNA levels Wassen FWJS, Moerings EPCM, Van Toor H, De Vrey EA, and enzyme activity in FRTL-5 rat thyroid cell. Molecular and HennemannG&Everts ME 1996 Effects of interleukin-1 on Cellular Endocrinology 101 31–35. thyrotropin secretion and thyroid hormone uptake in cultured rat Quan N, WhitesideM&Herkenham M 1998 Time course and anterior pituitary cells. Endocrinology 137 1591–1598. localization patterns of interleukin-1 beta messenger RNA Whiteside MB, QuanN&Herkenharn M 1999 Induction of expression in brain and pituitary after peripheral administration of pituitary cytokine transcripts by peripheral lipopolysaccharide. lipopolysaccharide. Neuroscience 83 281–293. Journal of Neuroendocrinology 11 115–120. Rawlings SR & Hezareh M 1996 Pituitary adenylate cyclase- activating polypeptide (PACAP) and PACAP/vasoactive intestinal polypeptide receptors: Actions on the anterior pituitary gland. Endocrine Reviews 17 4–29. Received 20 April 2000 RayD&MelmedS1997Pituitary cytokine and growth factor Revised manuscript received 11 August 2000 expression and action. Endocrine Reviews 18 206–228. Accepted 22 August 2000

www.endocrinology.org Journal of Endocrinology (2000) 167, 505–515

Downloaded from Bioscientifica.com at 09/25/2021 04:33:35AM via free access