Functional characterization and structural modeling of obesity associated mutations in the melanocortin 4 receptor

Abstract

Mutations in the melanocortin 4 receptor (MC4R) gene are the most common known cause of monogenic human obesity. The MC4R gene was sequenced in 2000 subjects with severe early-onset obesity. We detected seven different nonsense and 19 nonsynonymous mutations in a total of 94 probands, some of which have been reported previously by others. We functionally characterized the 11 novel obesity associated missense mutations. Seven of these mutants (L54P, E61K, I69T, S136P, M161T, T162I, and I269N) showed impaired cell surface trafficking, reduced level of maximal binding of the radioligand [125I]NDP-MSH, and reduced ability to generate cAMP in response to ligand. Four mutant MC4Rs (G55V, G55D, S136F, and A303T) displayed cell surface expression and agonist binding similar to the wild-type receptor but showed impaired cAMP production, suggesting that these residues are likely to be critical for conformational rearrangement essential for receptor activation. Homology modeling of these mutants using a model of MC4R based on the crystal structure of the beta2-adrenoreceptor was used to provide insights into the possible structural basis for receptor dysfunction. Transmembrane (TM) domains 1, 3, 6, 7, and peripheral helix 8 appear to participate in the agonist-induced conformational rearrangement necessary for coupling of ligand binding to signaling. We conclude that G55V, G55D, S136F, and A303T mutations are likely to strengthen helix-helix interactions between TM1 and TM2, TM3 and TM6, and TM7 and helix 8, respectively, preventing relative movement of these helices during receptor activation. The combination of functional studies and structural modeling of naturally occurring pathogenic mutations in MC4R can provide valuable information regarding the molecular mechanism of MC4R activation and its dysfunction in human disease.

Figure 1

Mutations in MC4R. A, A schematic representation of human MC4R with the location of 11 new mutations identified in this study. B, Amino acid sequence alignment of selected melanocortin receptors and two GPCRs with experimentally obtained 3D structures: bovine rhodopsin (RHOD_BOVIN) and human β2-adrenergic receptor (B2ADR_HUMAN). GenBank accession numbers of the aligned melanocortin receptors are as follows: human MC4R AAG35602, mouse MC4R AAI16958, chicken MC4R BAA25252, zebrafish MC4R AAO24745, human MC5R AAH95531, human MC3R AAH69105, human MC1R AAK58525, and human MC2R AAH69074. Residues where disease-causing mutations occurred are marked by gray. In human β2-adrenoreceptor, gray shading indicates mutated sites that cause constitutive activation of receptor uncoupling with G proteins (23). Mutated sites characterized in the current study are boxed and marked by arrows. Residues from TM helices in crystal structures and receptor models are underlined.

Figure 2

cAMP accumulation of MC4R mutants. Each point represents the mean ± se of at least three independent experiments performed in duplicate. A, Mutants that were completely unresponsive to αMSH. B, Mutants that displayed partial response to αMSH. WT, Wild type.

Figure 3

Cell surface expression of MC4R mutants. FACS analysis of MYC-tagged wild-type (WT) and mutant MC4Rs. The total cell receptor expression levels were determined using permeabilized cells measuring both cell surface and intracellular protein expression. The cell surface expression levels were determined using nonpermeabilized cells. Expression levels are presented relative to the wild-type MC4R control, means ± se of three to four independent experiments.

Figure 4

Competitive binding of MC4R mutants (A–F). HEK293 cells transiently transfected with wild-type (WT) or mutant MC4Rs were incubated with [¹²⁵I]NDP-MSH in the presence of increasing concentrations of NDP-αMSH. The ordinate is expressed as a percentage of total specific binding (B_max). Curves are fitted using a nonlinear regression analysis and a one-site competition model (GraphPad Prism). All curves are representative of three or four independent experiments, and each point is the mean of quadruplicate values.

Figure 5

Positioning of 11 mutated residues on the models of human MC4R. A, Model of the active conformation of MC4R in complex with natural agonist α-MSH. Mutated residues from class I are yellow, from class II are green, from class III are blue, and from class IV are red. Position of hydrophobic membrane boundaries was shown in accordance with Orientations of Proteins in Membranes database (http://opm.phar.umich.edu). B–I, Packing interactions around mutated residues in the inactive conformation are shown for the following MC4R mutants: G55D (B), G55V (C), S136F (D), I69T (E), M161T and T162I (F), I269N (G), L54P (H), and E61K and A303T (I). Figures were prepared by PyMOL (http://www.pymol.org).