Christos Matsingos: Elucidating the activation mechanism of the proton-sensing GPR68 receptor

This receptor is coupled with different types of G-proteins 1, 3, 4 and is involved in a variety of biological processes including cellular differentiation in bone tissue 5, 6, regulation of insulin levels [7], mechanosensing 8, 9, hippocampal neurogenesis [10], memory [11] and hormone release from the pituitary gland 12, 13. GPR68 has also been shown to have a role in tumour growth 4, 14, 15, 16, 17, 18 and metastasis 19, 20, and has been found in many neoplastic tissues [21], possibly associated with the microacidic environment involved in solid tumours [22].

No experimental structure of GPR68 or closely related protein has been published so far and the molecular details of the proton-sensing mechanism in GPCRs are still unclear. After the discovery of GPR4 and GPR68 as proton-sensing GPCRs in 2003 [1], homology models and mutagenesis studies initially indicated that a cluster of histidine residues located on the extracellular side of these receptors (green in Figure 1) was mostly responsible for pH sensitivity 1, 11, 23. Upon a decrease in environmental pH, histidines in the cluster would become protonated, initiating activation through a series of unknown conformational changes. Mutagenesis showed that these histidines have different degrees of involvement in proton-sensing [23], ranging from H201.31 and H169EL2, which have the most consistent effects on pH sensitivity, to H89EL1, H1594.63 and H175EL2, which show a significant effect only when mutated in combination with other histidines. The two transmembrane H2456.52 and H2697.36 residues (purple in Figure 1), while significantly affecting proton sensitivity when mutated to phenylalanine, were found to lead to modest effects when mutated to alanine, suggesting they have a structural role rather than a direct involvement in proton-sensing through changes in their protonation state.

Subsequent studies placed emphasis on the role of acidic residues in proton sensitivity. Evolutionary analyses [2] identified an ‘acidic triad’ (red in Figure 1) composed of residues D672.50 and D2827.49 (collectively named as the DyaD site), together with E1494.53 (the apEx site). This triad is conserved in the proton-sensing GPR4, GPR65 and GPR68 receptors and was shown to affect their activity upon mutation [2]. Moreover, a recent computational study predicted the existence of a pH-dependent network of water-mediated hydrogen bond interactions connecting histidine and acidic residues in the extracellular and transmembrane regions of GPR68 [24].

While various lines of evidence are available on the activation mechanism of GPR68, clear atomistic detail is missing. The molecular events triggered by histidine protonation, how they are propagated through allosteric communication and if elements found to be important in class A GPCR activation are involved in such communication are unknown. In this work, we provide new insight into the activation mechanism of GPR68 through Molecular Dynamics simulations of modelled inactive and active states of the receptor and through mutagenesis studies. A sequence of protonation events is shown to lead to the beginning of the transition towards the active state. Our findings not only confirm the importance of the acidic triad in the activation mechanism but also lead to the identification of key pairings between acidic and histidine residues and of a previously unknown hydrophobic lock in the extracellular region. Elucidating the activation mechanism of GPR68 at the molecular level is essential to understand the role of the receptor in diseases such as cancer 4, 14, 15, 16, 17, 18, 19, 20 and to develop compounds that can modulate its function.

RESULTS AND DISCUSSION

Protonation of key residues affects inter-residue distances

MD simulations were used to model the structural changes triggered by protonation of the inactive state of GPR68. Simulations were first carried out in three replicas starting from an inactive state model (see Methods section) with all the residues in their standard protonation state (‘U’ trajectories in Table S1). The predicted pKa values of histidine and acidic amino acids were calculated for each trajectory frame in order to estimate the propensity of these residues to become protonated within the activation window of GPR68. Interestingly, the pKa values of H171.28, H842.67 and H169EL2 were found to be ≥ 6.8 (proton EC50 in experiments measuring cAMP accumulation in transiently transfected HEK293T cells [23]) for at least part of the trajectories (Figures 2A and S1), indicating an increase in their basicity and therefore in their ability to become protonated within the activation window (as a reference, the model PROPKA pKa value for histidine in water is 6.5 [26]). These three residues are part of the group of histidines considered to be involved in proton-sensing (Table S2). Our data show that among the extracellular histidines, H169EL2 seems to be the most likely to get protonated, at least in the initial stages of the activation. The two transmembrane residues H2456.52 and H2697.36 have pKa values consistently low (< 6) for all (H2456.52) or most (H2697.36) of the simulated time, confirming the indication from previous experiments [23] that these histidines have more a structural role than a proton-sensing one. We cannot exclude however that a raise in the pKa of these residues could be observed during longer simulations. Among the acidic residues, the DyaD residues D672.50 and D2827.49 and the apEx residue E1494.53 were all found to have pKa values shifted upwards from their standard reference values (3.8 for aspartates and 4.5 for glutamates [26]), with D672.50 having a pKa ≥ 6.8 for at least 60% of the trajectory in all the replicas (Figure 2A).

Read full publication here