Evidence for Follicle-stimulating Hormone Receptor as a Functional Trimer

Abstract

Follicle-stimulating hormone receptor (FSHR), a G-protein coupled receptor, is an important drug target in the development of novel therapeutics for reproductive indications. The FSHR extracellular domains were observed in the crystal structure as a trimer, which enabled us to propose a novel model for the receptor activation mechanism. The model predicts that FSHR binds Asnα(52)-deglycosylated FSH at a 3-fold higher capacity than fully glycosylated FSH. It also predicts that, upon dissociation of the FSHR trimer into monomers, the binding of glycosylated FSH, but not deglycosylated FSH, would increase 3-fold, and that the dissociated monomers would in turn enhance FSHR binding and signaling activities by 3-fold. This study presents evidence confirming these predictions and provides crystallographic and mutagenesis data supporting the proposed model. The model also provides a mechanistic explanation to the agonist and antagonist activities of thyroid-stimulating hormone receptor autoantibodies. We conclude that FSHR exists as a functional trimer.

Keywords: Allosteric Regulation; Arrestin; Cysteine-knot Growth Factor; G Protein-coupled Receptors (GPCR); Glycoprotein Hormones; Receptor Structure-function; Reproduction.

FIGURE 1.

Effect of FSH glycosylation at Asnα⁵² to its receptor binding. A, spatial consideration of Asnα⁵² glycosylation on FSH binding to its receptor. Left panel, a molecular model of a single fully glycosylated FSH molecule binding to an FSHR trimer, viewing from top. For clarity, glycosylations except at the Asnα⁵² site have been omitted. The receptor trimer is shown as a magenta surface, FSH amino acids as a blue surface, and carbohydrates as yellow balls. Right panel, crystal structure of deglycosylated FSH bound to FSHR_ED trimer. B, experimental validation of the trimeric model prediction. Left panel, saturation curves of FSH binding to FSHR. The curve represents experiments performed in duplicate samples. Right panel, receptor binding ratio of αN52D-FSH mutant versus fully glycosylated FSH. An equal amount (5 μg) of cell membrane from the same preparation was used for each derived binding ratio to minimize FSHR count difference. The data reflect the ¹²⁵I-FSH receptor binding assays in four independent assays, each with a different membrane preparation.

FIGURE 2.

Effect of LMW allosteric modulators on the FSH/FSHR binding stoichiometry. A, saturation curves of FSH binding to FSHR in the absence or presence of Compound 5 (at indicated concentrations). The curve represents experiments performed in duplicate samples. Right panel, the FSH K_d and B_max values at the specified Compound 5 concentration calculated from the saturation curves. B, relative FSH binding sites of FSHR at different concentrations of Compound 5 where the B_max value in the absence of the modulator is normalized to 100%. C, effect of Compound 5 on ¹²⁵I-FSH αN52D mutant binding to FSHR. The curve represents experiments performed in duplicate samples.

FIGURE 3.

Effects of Compound 5 on receptor activation. A, consideration of spatial compatibility between a 7-TM domain and β-arrestin. Each 7-TM domain is represented as a blue circle and each arrestin as a magenta-like rectangle. The three panels represent three representative orientations of β-arrestins in relative to the 7-TM domains, assuming a 3-fold rotational symmetry in the 7-TM trimer. It can be concluded that only one β-arrestin can bind to the FSHR trimer due to the steric hindrance along the elongated dimension. B, the relative amount of β-arrestin recruited to the activated FSHR inside the CHO cell upon stimulation of FSH alone (left panel) or Compound 5 plus FSH at the EC₁₀₀ concentration (right panel). The amount of recruited β-arrestin is normalized to 100% for the maximum response of FSH. Data represent experiments performed in duplicate samples. C, the relative amount of recruited β-arrestin upon stimulation of Compound 5 alone (left panel) or FSH at the EC₂₀, EC₅₀, EC₇₀, and EC₁₀₀ concentrations mixed with 1 μm Compound 5 (right panel). Data represent experiments performed in duplicate samples.

FIGURE 4.

Crystallographic and mutagenesis studies of the FSH-FSHR complex. A, superimposition of the P1 and P3₁ trimer structures. P1, green; P3₁, magenta. Of 1449 common Cα atom pairs, 1378 pairs were superimposed, resulting in an root mean square deviation of 0.57 Å between the trimers of two space groups. B, top view of the trimer observed in the crystal structures. The inset shows a close-up view of the potential exosite originating from the FSH-FSHR_ED complex oligomerizations. The magenta ribbons are for the receptor trimer; green and blue ribbons are for the FSH α- and β-chains, respectively. The FSH Asnα⁵² glycan is shown as yellow balls. C, validation of the roles of the exosite in FSHR activation by FSH mutagenesis. Left panel, relative amount of β-arrestin recruited for binding to the activated FSHR upon stimulation by FSH or its mutant. The amount of recruited β-arrestin is normalized to 100% for the maximum response of FSH. Data represent experiments performed in duplicate samples. Right panel, relative amount of estradiol production inside primary granulosa cells from immature rats on stimulation by FSH or its mutant. The amount of estradiol production is normalized to 100% for the maximum response of FSH. Data represent experiments performed in triplicate samples.

FIGURE 5.

Proposed mechanism of glycoprotein hormone receptor activation. The extracellular leucine-rich repeats of the receptor are represented as purple blocks with the flexible loops as hairpins and 7-TM domains as cylinders (inactivated and activated forms are colored as gray and green, respectively). The other key receptor elements are also shown, including the sulfate group at Tyr³³⁵ depicted as yellow balls, residues Ser²⁷¹ as green stars, and disulfide bonds as thin yellow lines. G-protein or β-arrestin are shown as an ellipsoid. The GPH heterodimer is shown in blue, carbohydrates at Asnα⁵² as yellow sticks, and LMW modulators as yellow hexagons.

FIGURE 6.

Explanation of TSHR autoantibody agonist and antagonist activities. A, theoretical model of the TSHR extracellular domain (TSHR_ED) in complex with TSHR autoantibodies, M22 and K1–70. The molecules are shown as color-coded ribbons, marked by their names in the corresponding colors. Left panel, the TSHR trimer:autoantibody model. For clarity, only one TSHR protomer of the trimer is shown to bind the antibodies. Right panel, the TSHR monomer-autoantibody model. The N- and C-terminals of TSHR ectodomain are marked by their respective letters. The hinge sulfated tyrosine side chain is shown as colored balls. B, M22 agonist autoantibody clashes with its neighboring TSHR. Left panel, same orientation as in the left panel of A. Right panel, a rotated orientation to show the clashed surface (∼300 Ų). TSHR₍₂₎ is shown in a magenta surface. C, same representation as in B except the autoantibody is K1–70. Note that there is no clash between the autoantibody and its neighboring TSHR.