Homology modeling of MT₁ and MT₂ receptors

Abstract

Melatonin is a neurohormone synthesized and secreted mainly during the dark period of the circadian cycle by the pineal gland. It has already been proved to be involved in a number of chronobiological processes, most of them being mediated by its membranar receptors MT₁ and MT₂. Both are members of the GPCR class and, despite the interest they elicit, their 3D structure is still to be described. Models for both human MT₁ and MT₂ receptors have been constructed by homology modeling, using the X-ray structure of bovine rhodopsin as template. These models have been evaluated in terms of hydrophobic properties of the helices and refined to take into account the rearrangement of GPCRs necessary for their activation, thus leading to a putative activated model for each subtype.

Introduction

Melatonin, N-acetyl-5-methoxytryptamine (Fig. 1) was isolated and its structure described in 1959 [1]. It is principally synthesized by the pineal gland and secreted during the dark period of the circadian cycle [2], due to the suppressive effect of light on the stimulatory retinohypothalamic pathway [3], [4]. This light-regulated cycle gives to melatonin a major role in seasonal and chronobiological processes. It is already employed for veterinary use, such as stimulation of fur growth for the mink or regulation of the ovine reproduction [5], [6]. Among the possibilities under investigation for its use in human clinics, its chronobiotic properties are widely involved. Numerous works have been devoted to the resynchronization of disturbed biological rhythms such as sleep shift caused by jet lag or night work, insomnia and seasonal affective disorders [7], [8], [9], [10], [11]. Overall, two kinds of melatonin binding sites have been identified so far. They have been localized and characterized initially on the basis of their affinity for 2-[¹²⁵I]-iodomelatonin, giving rise to high affinity ML₁ sites and low affinity ML₂ sites [12]. Melatonin exerts the major part of its aforementioned effects by the activation of two high affinity binding sites denoted as MT₁ [13] and MT₂ [14], which belong to the G-protein coupled receptor superfamily, and a low affinity binding site, referred to as MT₃, according to the IUPHAR nomenclature [15], which was recently identified as quinone reductase 2 (QR2 EC 1.6.99.2) [16], [17], an enzyme closely related to the detoxifying enzyme quinone reductase 1. Moreover, there is also another high affinity binding site, the subtype Mel_1c that was first cloned from Xenopus laevis but not yet identified in mammals [18]. As the IUPHAR nomenclature concerns mammals only, Mel_1c has kept its original name. The MT₁ subtype is found in the retina, the kidneys and the brain, more particularly in the suprachiasmatic nuclei and thus should be implicated in the transmission of the circadian effects of melatonin and reproduction. This receptor could also be involved in peripheral vasoconstriction. The MT₂ subtype is found in the retina and the brain too, but contrary to MT₁, it does not appear in the suprachiasmatic nuclei. It could play a role in the physiology of the retina and in body temperature regulation [19], [20]. Structurally, MT₁ is a 350 amino acid protein, presenting a sequence homology of 60% with the 362 amino acid MT₂ [14]. The GPCR class is an important target for drugs due to the great diversity of their ligands, ranging from photons to peptides or proteins. Some studies suggest that 30–50% of known drugs actually target a GPCR to produce their effects [21], [22], [23]. These receptors display a common characteristic fold composed of seven transmembrane alpha-helices. Moreover, they present highly conserved residues, such as the DRY sequence of the intracellular end of transmembrane domain 3 (TM3) implied in the G-protein activation. This sequence is replaced by an NRY motif for melatonin receptors, which plays exactly the same role by coupling the receptor and a G_i or Gq protein, leading to a decrease in cAMP [24], [25]. However, GPCRs are extremely difficult to isolate and crystallize as their inclusion in the cell membrane contributes to their structural stability and renders them poorly purifiable [26]. This leads to the impossibility to determine directly their structure by standard means. The only way to cope with this problem is the construction of theoretical models based on the only GPCR crystallographic structure available, bovine rhodopsin, using a homology modeling approach [27], [28]. However, it is assumed that the high-resolution X-ray structure of bovine rhodopsin is in an inactive state and cannot stand for agonist ligand activated GPCRs. This GPCR activation process can be simulated by rotation of some transmembrane domains as suggested by experimental data [29], [30], [31]. We report here the construction of a 3D structure for both human MT₁ and MT₂ receptors and their careful modification toward an activated state using mutational and ligand SAR data in an attempt to propose a valid melatonin binding site. Such models may be very helpful to facilitate the design and development of new selective ligands.