Endogenous betaglycan is essential for high-potency inhibin antagonism in gonadotropes

Abstract

Inhibins are endocrine hormones that regulate gametogenesis and reproduction through a negative feedback loop with FSH. Inhibin action involves antagonism of signaling by activin or other TGFbeta family ligands. In transfection assays, antagonism by inhibin can be potentiated by betaglycan, a coreceptor for selected TGFbeta family ligands. We tested whether betaglycan is an obligate inhibin coreceptor through disruption of betaglycan function by RNA interference-mediated knockdown and immunoneutralization. Betaglycan knockdown and anti-betaglycan IgG each independently prevented inhibin-A binding to betaglycan and reversed functional effects of transfected betaglycan. Neither betaglycan immunoneutralization nor knockdown affected activin responsiveness in cell lines or in rat anterior pituitary cultures. Betaglycan knockdown decreased the potency of inhibin antagonism of activin-induced FSH secretion in primary gonadotropes. Similarly, anti-betaglycan IgG decreased the potency of inhibin antagonism in primary gonadotropes in a dose-dependent manner, with a reduction in the sensitivity to inhibin-A of greater than 1000-fold. These data establish that betaglycan is an endogenous inhibin coreceptor required for high-sensitivity inhibin antagonism of activin signaling in rat anterior pituitary gonadotropes.

Figure 1

Betaglycan shRNA suppresses betaglycan expression and reduces inhibin-A binding to betaglycan but not ActRIIB. 293T cells were mock infected or infected with lentivirus containing betaglycan, scramble shRNA, or GFP lentivirus. Cells were transfected with myc-tagged betaglycan DNA, protein extracts were prepared, and betaglycan expression was detected by Western blot as described in Materials and Methods (A). Bands corresponding to betaglycan core protein and sizes of molecular weight markers are shown. For binding, mock and shRNA 293T stable cells were transfected with ActRIIB (B), betaglycan (C), or ActRIIB and betaglycan (D), and incubated with [¹²⁵I]inhibin-A with or without 100 nm of unlabeled inhibin-A as described in Materials and Methods. Triplicate wells were washed, and total binding was measured in a γ-counter. Data were analyzed and graphed using the Prism program, and statistical significance was calculated by ANOVA, using the Tukey post test: *, P < 0.05; **, P > 0.05. con, Control.

Figure 2

Anti-BG IgG reduces inhibin binding and cross-linking to betaglycan through interaction with the inhibin-binding domain. A, A schematic of the fusion protein used to generate anti-BG IgG and wild-type myc-betaglycan or betaglycan mutants expressed in 293T cells with or without deletion of the inhibin-binding domain. B, 293T cells were transfected with empty vector, full-length myc-betaglycan, or myc-betaglycan truncations as indicated, and protein lysates were prepared and blotted against anti-BG IgG (B), or anti-myc (9E10) (C). For binding and cross-linking, 293T cells were transfected with empty vector or myc-ActRIIB (D and E), empty vector or betaglycan (F and G), or ActRIIB and betaglycan (H and I). For binding (D, F, and H), triplicate wells were preincubated with 250 μg/ml of anti-BG IgG, 250 μg/ml control IgG, 100 nm of unlabeled inhibin-A, or binding buffer without competitor, and 2 h later incubated with [¹²⁵I]inhibin-A as described in Materials and Methods. Cells were washed, and total binding was measured in a γ-counter. Data were analyzed and graphed using the Prism program, and statistical significance was calculated by ANOVA, using the Tukey post test: *, P < 0.05; **, P > 0.05. For cross-linking (E, G, and I), cells were incubated with [¹²⁵I]inhibin-A in the absence of competitor or with 100 nm of unlabeled inhibin-A, 250 μg/ml of anti-BG IgG, or 250 μg/ml control IgG as indicated and subjected to covalent cross-linking and immunoprecipitation with an anti-myc (9E10) antibody as described in Materials and Methods. Bands corresponding to [¹²⁵I]inhibin-A-cross-linked complexes with ActRIIB, betaglycan core protein, and GAG-betaglycan are indicated, and sizes of molecular weight markers are shown. Inh-A, Inhibin-A; GAG, glycosaminoglycan.

Figure 3

Betaglycan shRNA and anti-BG IgG decrease functional inhibin-A potency in betaglycan-transfected 293T cells. A, 293T cells were mock infected or infected with lentivirus containing betaglycan or scramble shRNA. Mock infected 293T cells (circles), BG shRNA stable cells (diamonds) or scramble shRNA stable cells (triangles) were transfected with either empty vector (open symbols) or betaglycan DNA (solid symbols) with A₃-luciferase/FAST-2/CMV-βgal as described in Materials and Methods. Cells were treated with 100 pm activin-A and the indicated doses of inhibin-A, and luciferase activity was measured and normalized relative to β-galactosidase activity in individual wells. B, 293T cells were transfected with either vector/A₃-luciferase/FAST-2/CMV-βgal DNA (open symbols) or betaglycan/A₃-luciferase/FAST-2/CMV-βgal DNA (solid symbols) as described in Materials and Methods. Transfected cells were incubated with 750 μg/ml of anti-BG IgG (diamonds), or 750 μg/ml control IgG (triangles), or media (circles) then treated 8 h later with the indicated doses of activin-A and inhibin-A such that final doses of anti-BG IgG or control IgG were 500 μg/ml. Luciferase activity was measured 16 h later and normalized relative to β-galactosidase activity in individual wells. In all conditions, treatments were performed in triplicate and presented as normalized luciferase activity with error bars representing standard error of the triplicate wells. Data were analyzed and curve fitted, including calculation of IC₅₀ with 95% confidence intervals, and graphed using the Prism program. Con, Control; RLU, relative light units; Vec, vector; NT, not treated.

Figure 4

Decreasing betaglycan levels with shRNA decreases inhibin antagonism of FSH secretion in gonadotropes. Primary cultures of male RAP were prepared and plated in 48-well plates as described in Materials and Methods. Cultures were incubated 72 h later at a M.O.I. of 5 with BG shRNA lentivirus (diamonds), scramble shRNA lentivirus (triangles), or media without lentivirus (circles) for 120 h as described in Materials and Methods. Triplicate wells were washed twice and incubated with new media containing doses of activin-A (A) or activin-A and inhibin-A (B) as indicated. After 72 h, media was collected and assayed for both FSH or LH in independent RIAs as described in Materials and Methods. Data are presented as total hormone levels detected in nanograms/ml with error bars representing the standard error of the triplicate wells. Data were analyzed and curve fitted, including calculation of EC₅₀ and IC₅₀ with 95% confidence intervals, and graphed using the Prism program. Act, Activin; Con, control.

Figure 5

Blocking inhibin binding to betaglycan interferes with inhibin antagonism in LβT2 cells. LβT2 cells were plated in 12-well plates and treated with 300 μg/ml of anti-BG IgG (diamonds), 300 μg/ml control IgG (triangles), or media without IgG (circles) and doses of activin-A alone (A) or activin-A and inhibin-A (B) as indicated in triplicate wells. Extracellular media was collected 72 h later and hormone secretion was assayed for both FSH and LH in independent RIAs as described in Materials and Methods. FSH data are presented as total hormone levels detected in nanograms/ml with error bars representing the standard error of the triplicate wells. Data were analyzed and curve fitted, including calculation of EC₅₀ and IC₅₀ with 95% confidence intervals, and graphed using the Prism program. Act-A, Activin-A; Con, control.

Figure 6

Blocking inhibin binding to betaglycan decreases inhibin antagonism of FSH secretion from gonadotropes. Primary cultures of male RAPs were prepared and plated in 48-well plates as described in Materials and Methods. After 72 h of recovery, cells were washed twice and treated in triplicate wells with fresh media without IgG (circles), or media containing anti-BG IgG (diamonds), or media containing control IgG (triangles), and doses of activin-A (A) or activin-A and inhibin-A (B) as indicated. For panel C, doses of anti-BG IgG are shown. After 72 h of incubation, media were collected and assayed for both FSH and LH in independent RIAs as described in Materials and Methods. Data are presented as total hormone levels detected in nanograms/ml with error bars representing the standard error of the triplicate wells. Data were analyzed and curve fitted, including calculation of EC₅₀ and IC₅₀ with 95% confidence intervals, and graphed using the Prism program. Experiments were repeated at least three times with similar results. Act-A, Activin-A; Con, control.