A functional transmembrane complex: the luteinizing hormone receptor with bound ligand and G protein

Abstract

The luteinizing hormone receptor (LHR) is one of eight members in a cluster of the rhodopsin family of the large G protein-coupled receptor (GPCR) superfamily that contains some 800-900 genes in the human genome. LHR, along with its paralogons, follicle stimulating hormone receptor (FSHR) and thyroid stimulating hormone receptor, form one of the three classes in this cluster; the two other classes contain the relaxin-binding GPCRs and orphan GPCRs. These GPCRs are characterized by a relatively large ectodomain (ECD) containing leucine-rich-repeats (LRRs); in the class of glycoprotein hormone receptors, the LRR region is capped by N-terminal and C-terminal cysteine-rich regions. Binding of human chorionic gonadotropin (hCG) or luteinizing hormone to the LHR-ECD triggers a conformational change of the transmembrane region of the receptor facilitating binding and activation of Gs, followed by effector enzyme activation and subsequent intracellular signaling. Viewing LHR as a transmembrane anchoring protein that sequentially binds hCG and Gs to give the hCG-LHR-Gs complex, numerous interactions and conformational changes must be considered. There is, unfortunately, a paucity of structural data on LHR, but crystal structures exist for hCG, the homologous FSH-FSHR-ECD (N-terminal fragment) complex, rhodopsin (in the inactive state), an active form of Galphas (transducin), and the betagamma heterodimer. Using a combined experimental (site-directed mutagenesis followed by characterization in transfected cells) and computational (homology modeling and molecular dynamics simulations) approach, good working models are being developed for the protein-protein interaction faces and, in some cases, the ensuing conformational changes induced by complex formation. hCG binding to the LHR-ECD appears to involve several LRRs; LHR activation can be described in terms of disrupting a network of H-bonds in the cytosolic halves of helices 1-3, 6, and 7; and binding of LHR to Gs involves, in large part, intracellular loop 2 binding, presumably to Gsalpha at its C-terminus. Major gaps exist in our understanding at the molecular level of the six-polypeptide chain complex, hCG-LHR-Gs, but considerable progress has been made in the past few years.

Fig. 1

Schematic representation of the hCG-LHR-Gs protein complex. This illustration depicts a complex containing six distinct proteins, the heterodimeric hormone, hCG (with one α and one β subunit, colored in yellow and orange, respectively), bound to the LHR-ECD (visualized in pink) on the extracellular face of the plasma membrane and the transmembrane portion of LHR (helices colored green) bound to the heterotrimeric G protein (with one each of α, β, and γ subunits, represented in royal blue, light blue, and red, respectively; these α and β subunits are unrelated to those comprising hCG) on the intracellular face of the plasma membrane. As discussed in the text, the structure shown for hCG is based on crystallographic data, that for LHR on homology modeling, and that for Gα is actually the crystal structure of the homologous Gα_t subunit complexed with βγ. The N-linked oligosaccharides on hCG and the LRR domain of the LHR ECD were modeled as biantennary complex forms (except that on αN52 as a hybrid form) and presented in a ball-and-stick representation. While there is some confidence that the hCG-LHR-ECD complex is reasonably accurate, the positioning of this complex with the transmembrane portion of LHR is simply pictorial, as is the interaction surface of LHR and the G protein. This figure was generated with the molecular visualization programs, MOLSCRIPT (Kraulis, 1991) and Raster3D (Merritt and Bacon, 1997).

Fig. 2

Structure of a typical LHR in the LHR-ECD. The amino acid sequences of LxxLxLxxNxxLxxLPxxaFxxLxx (see text for definition) adopts a structure in which the xxLxLxx forms a β-strand and the remainder adopts a loop structure, with the side chains of the conserved L, N, and F residues pointing inside to form a hydrophobic core.

Fig. 3

Predicted structure of the LHR-ECD. A. Based on homology modeling with the structures of the Nogo receptor (He et al., 2003) and glycoprotein Ibα (Huizinga et al., 2002) as outlined in the text. B. A large N-terminal fragment (residues 1 – 241) of the LHR-ECD based on homology modeling with the structure of the FSH-ECD. C. Comparison of the models of the LHR-ECD N-terminal fragment based on homology modeling with the Nogo receptor/glycoprotein Ibα and the FSH-ECD N-terminal fragment. The structures were drawn with MOLSCRIPT (Kraulis, 1991).

Fig. 3

Predicted structure of the LHR-ECD. A. Based on homology modeling with the structures of the Nogo receptor (He et al., 2003) and glycoprotein Ibα (Huizinga et al., 2002) as outlined in the text. B. A large N-terminal fragment (residues 1 – 241) of the LHR-ECD based on homology modeling with the structure of the FSH-ECD. C. Comparison of the models of the LHR-ECD N-terminal fragment based on homology modeling with the Nogo receptor/glycoprotein Ibα and the FSH-ECD N-terminal fragment. The structures were drawn with MOLSCRIPT (Kraulis, 1991).

Fig. 3

Predicted structure of the LHR-ECD. A. Based on homology modeling with the structures of the Nogo receptor (He et al., 2003) and glycoprotein Ibα (Huizinga et al., 2002) as outlined in the text. B. A large N-terminal fragment (residues 1 – 241) of the LHR-ECD based on homology modeling with the structure of the FSH-ECD. C. Comparison of the models of the LHR-ECD N-terminal fragment based on homology modeling with the Nogo receptor/glycoprotein Ibα and the FSH-ECD N-terminal fragment. The structures were drawn with MOLSCRIPT (Kraulis, 1991).

Fig. 4

Predicted and experimental structures of the FSHR-ECD N-terminal fragment (residues 1 – 241). A. Molecular modeling with the Nogo receptor (He et al., 2003) and glycoprotein Ibα (Huizinga et al., 2002) as templates. B. Crystal structure shown without the bound FSH (Fan and Hendrickson, 2005a). C. Comparison of the predicted structure of FSH-ECD 1 – 241 (from A) with that of the crystal structure (from B). The structures were drawn with MOLSCRIPT (Kraulis, 1991).

Fig. 4

Predicted and experimental structures of the FSHR-ECD N-terminal fragment (residues 1 – 241). A. Molecular modeling with the Nogo receptor (He et al., 2003) and glycoprotein Ibα (Huizinga et al., 2002) as templates. B. Crystal structure shown without the bound FSH (Fan and Hendrickson, 2005a). C. Comparison of the predicted structure of FSH-ECD 1 – 241 (from A) with that of the crystal structure (from B). The structures were drawn with MOLSCRIPT (Kraulis, 1991).

Fig. 4

Predicted and experimental structures of the FSHR-ECD N-terminal fragment (residues 1 – 241). A. Molecular modeling with the Nogo receptor (He et al., 2003) and glycoprotein Ibα (Huizinga et al., 2002) as templates. B. Crystal structure shown without the bound FSH (Fan and Hendrickson, 2005a). C. Comparison of the predicted structure of FSH-ECD 1 – 241 (from A) with that of the crystal structure (from B). The structures were drawn with MOLSCRIPT (Kraulis, 1991).

Fig. 5

Predicted and experimental structures of gonadotropin-receptor-ECD complexes. A. Predicted structure of the hCG-LHR-ECD complex using rigid-body docking. B. Predicted structure of the FSH-FSHR-ECD complex using rigid-body docking. C. Comparison of the predicted and experimentally determined (Fan and Hendrickson, 2005a) structure of the FSH-FSHR-ECD complex.

Fig. 5

Predicted and experimental structures of gonadotropin-receptor-ECD complexes. A. Predicted structure of the hCG-LHR-ECD complex using rigid-body docking. B. Predicted structure of the FSH-FSHR-ECD complex using rigid-body docking. C. Comparison of the predicted and experimentally determined (Fan and Hendrickson, 2005a) structure of the FSH-FSHR-ECD complex.

Fig. 5

Predicted and experimental structures of gonadotropin-receptor-ECD complexes. A. Predicted structure of the hCG-LHR-ECD complex using rigid-body docking. B. Predicted structure of the FSH-FSHR-ECD complex using rigid-body docking. C. Comparison of the predicted and experimentally determined (Fan and Hendrickson, 2005a) structure of the FSH-FSHR-ECD complex.

Fig. 6

Predicted structures for extracellular motifs of representative members of class B and class C GPHRs. A. The orphan receptor, hLGR4, contains about twice the number of LRRs as other members of the GPHR cluster. B. The relaxin-binding receptor, hLGR8, contains multiple LRRs and one LDL receptor “class A” motif. These figures were generated with the molecular visualization programs MOLSCRIPT (Kraulis, 1991) and Raster3D (Merritt and Bacon, 1997).

Fig. 6

Predicted structures for extracellular motifs of representative members of class B and class C GPHRs. A. The orphan receptor, hLGR4, contains about twice the number of LRRs as other members of the GPHR cluster. B. The relaxin-binding receptor, hLGR8, contains multiple LRRs and one LDL receptor “class A” motif. These figures were generated with the molecular visualization programs MOLSCRIPT (Kraulis, 1991) and Raster3D (Merritt and Bacon, 1997).

Fig. 7

Average minimized structures of the WT LHR (A) and of the D564G constitutively active mutant (B). Only the cytosolic halves of helices 3 and 6 are shown, seen in a direction parallel to the putative membrane surface and colored in green and cyan, respectively. The side chains of E463, R464, and D564 are also shown, represented by liquorice models and colored according to their polarity (i.e. blue for cations and red for anions). The SAS computed over R464, T467, I468, and K463 is represented by gray dots. This index is 28 Ǻ² for WT LHR and 120 Ǻ² for the D564G mutant (Fanelli et al., 2004).

Fig. 8

Comparison of WT LHR and gain-of-function LHR mutants in response to exogenous hCG. WT LHR and constitutively activating mutants were expressed in HEK 293 cells, and the intracellular concentrations of cAMP were determined in the absence and presence of a saturating concentration of hCG (with an added inhibitor of cyclic nucleotide phosphodiesterase).