Pharmacological hallmarks of allostery at the M4 muscarinic receptor elucidated through structure and dynamics

Ziva Vuckovic; Jinan Wang; Vi Pham; Jesse I Mobbs; Matthew J Belousoff; Apurba Bhattarai; Wessel AC Burger; Geoff Thompson; Mahmuda Yeasmin; Vindhya Nawaratne; Katie Leach; Emma T van der Westhuizen; Elham Khajehali; Yi-Lynn Liang; Alisa Glukhova; Denise Wootten; Craig W Lindsley; Andrew Tobin; Patrick Sexton; Radostin Danev; Celine Valant; Yinglong Miao; Arthur Christopoulos; David M Thal

doi:10.7554/eLife.83477

. 2023 May 30;12:e83477. doi: 10.7554/eLife.83477

Pharmacological hallmarks of allostery at the M4 muscarinic receptor elucidated through structure and dynamics

Ziva Vuckovic ^1,^†, Jinan Wang ^2,^†, Vi Pham ^1,^†, Jesse I Mobbs ^1,^3,^†, Matthew J Belousoff ^1,³, Apurba Bhattarai ², Wessel AC Burger ^1,³, Geoff Thompson ¹, Mahmuda Yeasmin ¹, Vindhya Nawaratne ¹, Katie Leach ^1,³, Emma T van der Westhuizen ¹, Elham Khajehali ¹, Yi-Lynn Liang ¹, Alisa Glukhova ^1,³, Denise Wootten ^1,³, Craig W Lindsley ⁴, Andrew Tobin ⁵, Patrick Sexton ^1,³, Radostin Danev ⁶, Celine Valant ^1,^✉, Yinglong Miao ^2,^✉, Arthur Christopoulos ^1,^3,^7,^✉, David M Thal ^1,^3,^✉

Editors: Sudarshan Rajagopal⁸, Kenton J Swartz⁹

PMCID: PMC10229135 PMID: 37248726

Abstract

Allosteric modulation of G protein-coupled receptors (GPCRs) is a major paradigm in drug discovery. Despite decades of research, a molecular-level understanding of the general principles that govern the myriad pharmacological effects exerted by GPCR allosteric modulators remains limited. The M₄ muscarinic acetylcholine receptor (M₄ mAChR) is a validated and clinically relevant allosteric drug target for several major psychiatric and cognitive disorders. In this study, we rigorously quantified the affinity, efficacy, and magnitude of modulation of two different positive allosteric modulators, LY2033298 (LY298) and VU0467154 (VU154), combined with the endogenous agonist acetylcholine (ACh) or the high-affinity agonist iperoxo (Ipx), at the human M₄ mAChR. By determining the cryo-electron microscopy structures of the M₄ mAChR, bound to a cognate G_i1 protein and in complex with ACh, Ipx, LY298-Ipx, and VU154-Ipx, and applying molecular dynamics simulations, we determine key molecular mechanisms underlying allosteric pharmacology. In addition to delineating the contribution of spatially distinct binding sites on observed pharmacology, our findings also revealed a vital role for orthosteric and allosteric ligand–receptor–transducer complex stability, mediated by conformational dynamics between these sites, in the ultimate determination of affinity, efficacy, cooperativity, probe dependence, and species variability. There results provide a holistic framework for further GPCR mechanistic studies and can aid in the discovery and design of future allosteric drugs.

Research organism: None

Introduction

Over the past 40 y, there have been major advances to the analytical methods that allow for the quantitative determination of the pharmacological parameters that characterize G protein-coupled receptor (GPCR) signaling and allosteric modulation (Figure 1A and B). These analytical methods are based on the operational model of agonism (Black and Leff, 1983) and have been extended or modified to account for allosteric modulation (Leach et al., 2007), biased agonism (Kenakin, 2012), and even biased allosteric modulation (Slosky et al., 2021). Collectively, these models and subsequent key parameters (Figure 1B) are used to guide allosteric drug screening, selectivity, efficacy, and ultimately, clinical utility, and provide the foundation for modern GPCR drug discovery (Wootten et al., 2013). Yet, a systematic understanding of how these pharmacological parameters relate to the molecular structure and dynamics of GPCRs remains elusive.

Figure 1. — (A) Schematic of the pharmacological parameters that define effects of orthosteric and allosteric ligands on a G protein-coupled receptor (GPCR). (B) A simplified schematic diagram of the Black–Leff operational model to quantify agonism, allosteric modulation, and agonist bias with pharmacological parameters defined (Black and Leff, 1983). (C) 2D chemical structures of the orthosteric and allosteric ligands used in this study. (**D–G**) Key pharmacological parameters for interactions between orthosteric and allosteric ligands in [³H]-N-methylscopolamine ([³H]-NMS) binding assays. (D) Equilibrium binding affinities (pK_i and pK_B) and (E) the degree of binding modulation (α) between the agonists and PAMs resulting in the modified binding affinities (F) α/K_A and (G) α/K_B. (**H–K**) Key pharmacological parameters relating to Gα_i1 activation for interactions between orthosteric and allosteric ligands measured with the TruPath assay (Figure 1—figure supplement 1). (H) The signaling efficacy (τ_A and τ_B) and (I) transduction coupling coefficients (log (τ/K)) of each ligand. (J) The functional cooperativity (αβ) between ligands and (K) the efficacy modulation (β) between ligands. All data are mean ± SEM of three or more independent experiments performed in duplicate or triplicate with the pharmacological parameters determined using a global fit of the data. The error in (**F, G, K**) was propagated using the square root of the sum of the squares. See Table 1. Concentration–response curves are shown in Figure 1—figure supplement 1.

Figure 1—source data 1. Related to Figure 1D–K.

elife-83477-fig1-data1.xlsx^{(12.4KB, xlsx)}

(A) Schematic of the pharmacological parameters that define effects of orthosteric and allosteric ligands on a G protein-coupled receptor (GPCR). (B) A simplified schematic diagram of the Black–Leff operational model to quantify agonism, allosteric modulation, and agonist bias with pharmacological parameters defined (Black and Leff, 1983). (C) 2D chemical structures of the orthosteric and allosteric ligands used in this study. (**D–G**) Key pharmacological parameters for interactions between orthosteric and allosteric ligands in [³H]-N-methylscopolamine ([³H]-NMS) binding assays. (D) Equilibrium binding affinities (pK_i and pK_B) and (E) the degree of binding modulation (α) between the agonists and PAMs resulting in the modified binding affinities (F) α/K_A and (G) α/K_B. (**H–K**) Key pharmacological parameters relating to Gα_i1 activation for interactions between orthosteric and allosteric ligands measured with the TruPath assay (Figure 1—figure supplement 1). (H) The signaling efficacy (τ_A and τ_B) and (I) transduction coupling coefficients (log (τ/K)) of each ligand. (J) The functional cooperativity (αβ) between ligands and (K) the efficacy modulation (β) between ligands. All data are mean ± SEM of three or more independent experiments performed in duplicate or triplicate with the pharmacological parameters determined using a global fit of the data. The error in (**F, G, K**) was propagated using the square root of the sum of the squares. See Table 1. Concentration–response curves are shown in Figure 1—figure supplement 1.

Figure 1—source data 1. Related to Figure 1D–K.

elife-83477-fig1-data1.xlsx^{(12.4KB, xlsx)}

The muscarinic acetylcholine receptors (mAChRs) are an important family of five Class A GPCRs that have long served as model systems for understanding GPCR allostery (Conn et al., 2009). The mAChRs have been notoriously difficult to exploit therapeutically and selectively due to high-sequence conservation within their orthosteric binding domains (Burger et al., 2018). However, the discovery of highly selective positive allosteric modulators (PAMs) for some mAChR subtypes has paved the way for novel approaches to exploit these high-value drug targets (Chan et al., 2008; Gentry et al., 2014; Marlo et al., 2009). X-ray crystallography and cryo-electron microscopy (cryo-EM) have been used to determine inactive state structures for all five mAChR subtypes (Haga et al., 2012; Kruse et al., 2012; Thal et al., 2016; Vuckovic et al., 2019) and active state structures of the M₁ and M₂ mAChRs (Maeda et al., 2019). For the M₂ mAChR, this includes structures co-bound with the high-affinity agonist iperoxo (Ipx) and the PAM LY2119620 in complex with a G protein mimetic nanobody (Kruse et al., 2013) and the transducers G_o (Maeda et al., 2019) and β-arrestin1 (Staus et al., 2020). These M₂ mAChR structures were foundational to validating the canonical mAChR allosteric site but are limited to only one agonist (iperoxo) and one PAM (LY2119620) and do not account for the vast pharmacological properties of ligands targeting mAChRs. A recent nuclear magnetic resonance (NMR) study of the M₂ mAChR revealed differences in the conformational landscape of the M₂ mAChR when bound to different agonists, but no clear link was established between the properties of the ligands and the conformational states of the receptor (Xu et al., 2019). The M₄ mAChR subtype is of major therapeutic interest due to its expression in regions of the brain that are rich in dopamine and dopamine receptors, where it regulates dopaminergic neurons involved in cognition, psychosis, and addiction (Bymaster et al., 2003; Dencker et al., 2011; Foster et al., 2016; Tzavara et al., 2004). Importantly, these findings have been supported by studies utilizing novel PAMs that are highly selective for the M₄ mAChR (Bubser et al., 2014; Chan et al., 2008; Leach et al., 2010; Suratman et al., 2011). Among these, LY2033298 (LY298) was the first reported highly selective PAM of the M₄ mAChR and displayed antipsychotic efficacy in a preclinical animal model of schizophrenia (Chan et al., 2008). Despite LY298 being one of the best characterized M₄ mAChR PAMs, its therapeutic potential has been limited by numerous factors, including its chemical scaffold, which has been difficult to optimize with respect to its molecular allosteric parameters (Figure 1C) and variability of response between species (Suratman et al., 2011; Wood et al., 2017b). In the search for better chemical scaffolds, the PAM, VU0467154 (VU154), was subsequently discovered. VU154 showed robust efficacy in preclinical rodent models; however, it also exhibited species selectivity that prevented its clinical translation (Bubser et al., 2014). Collectively, LY298 and VU154 are exemplar tool molecules that highlight the promises and the challenges in understanding and optimizing allosteric GPCR drug activity for translational and clinical applications.

Herein, by examining the pharmacology of the PAMs LY298 and VU154 with the agonists ACh and Ipx across radioligand binding assays and two different signaling assays and analyzing these results with modern analytical methods, we determined the key parameters that describe signaling and allostery for these ligands. To investigate a structural basis for these pharmacological parameters, we used cryo-EM to determine high-resolution structures of the M₄ mAChR in complex with a cognate G_i1 heterotrimer and ACh and Ipx. We also determined structures of receptor complexes with Ipx co-bound with the PAMs LY298 or VU154. Moreover, because protein allostery is a dynamic process (Changeux and Christopoulos, 2016), we performed all-atom simulations using the Gaussian accelerated molecular dynamics (GaMD) enhanced sampling method (Draper-Joyce et al., 2021; Miao et al., 2015; Wang et al., 2021a) on the M₄ mAChR using the cryo-EM structures. The structures and GaMD simulations, in combination with detailed molecular pharmacology and receptor mutagenesis experiments, provide fundamental insights into the molecular mechanisms underpinning the hallmarks of GPCR allostery. To further validate these findings, we investigated the differences in the selectivity of VU154 between the human and mouse receptors and established a structural basis for species selectivity. Collectively, these results will enable future GPCR drug discovery research and potentially lead to the development of next generation M₄ mAChR PAMs.

Results

Pharmacological characterization of M₄ mAChR PAMs with ACh and Ipx

The pharmacology of LY298 or VU154 interacting with ACh has been well characterized in binding and functional assays at the M₄ mAChR (Bubser et al., 2014; Chan et al., 2008; Gould et al., 2016; Leach et al., 2010; Suratman et al., 2011; Thal et al., 2016). However, their pharmacology with Ipx has not been reported. Therefore, we characterized both PAMs with ACh and Ipx in binding and in two different functional assays to provide a thorough foundational comparative characterization of the pharmacological parameters of these ligands from the same study.

We first used radioligand binding assays (Figure 1—figure supplement 1A) to determine the binding affinities (i.e., equilibrium dissociation constants) of ACh and Ipx (K_A) for the orthosteric site and of LY298 and VU154 (K_B) for the allosteric site of the unoccupied human M₄ mAChR (Figure 1D), along with the degree of binding cooperativity (α) between the agonists and PAMs when the two are co-bound (Figure 1E). Analysis of these experiments revealed that LY298 and VU154 have very similar binding affinities for the allosteric site with values (expressed as negative logarithms; pK_B) of 5.65 ± 0.07 and 5.83 ± 0.12, respectively (Table 1), in accordance with previous studies (Bubser et al., 2014; Leach et al., 2011). Both PAMs potentiated the binding affinity of ACh and Ipx (Figure 1E), with the effect being greatest between LY298 and ACh (~400-fold increase in binding affinity). Comparatively, the positive cooperativity between VU154 and ACh was only 40-fold. When Ipx was used as the agonist, the binding affinity modulation mediated by both PAMs was more modest, characterized by an approximately 72-fold potentiation for the combination of Ipx and LY298, and 10-fold potentiation for the combination of Ipx and VU154. These results indicate probe-dependent effects (Valant et al., 2012) with respect to the ability of either PAM to modulate the affinity of each agonist (Figure 1F and G). A probe-dependent effect was also observed with the radioligand, [³H]-NMS, evidenced by a reduction in specific radioligand binding due to negative cooperativity between the antagonist probe and LY298, which has been previously reported (Chan et al., 2008; Leach et al., 2010; Suratman et al., 2011; Thal et al., 2016). It is important to note that binding affinity modulation is thermodynamically reciprocal at equilibrium, and the affinities of LY298 and VU154 were thus also increased in the agonist bound state (Figure 1—figure supplement 1A). This results in LY298 having a fivefold higher binding affinity than VU154 when agonists are bound (Table 1).

Table 1. Pharmacological parameters from radioligand binding and functional experiments.

[³H]-NMS saturation binding on stable M₄ mAChR CHO cells
Constructs		Sites per cell^*			pK_D^†
Human WT M₄ mAChR		598,111 ± 43,067 (7)			9.76 ± 0.05 (7)
Mouse WT M₄ mAChR		21,027 ± 2188 (3)			9.76 ± 0.05 (3)
Human D432E M₄ mAChR		126,377 ± 10,066 (3)			9.60 ± 0.07 (3)
Human T433R M₄ mAChR		157,442 ± 36,658 (6)			9.64 ± 0.09 (6)
Human V91L, D432E, T433R M₄ mAChR		205,771 ± 20,975 (4)			9.58 ± 0.08 (4)
[³H]-NMS interaction binding assays between ACh or Ipx and LY298 or VU154 on stable M₄ mAChR constructs in Flp-In CHO cells
Constructs	PAM	pK_i ACh ^‡	pK_i Ipx ^‡	pK_B PAM ^‡	log α_ACh ^§	log α_Ipx ^§
Human WT M₄ mAChR	LY298	4.50 ± 0.06 (4)	8.30 ± 0.06 (4)	5.65 ± 0.07 (8) ^¶	2.59 ± 0.10 (4)	1.86 ± 0.10 (4)
Human WT M₄ mAChR	VU154	4.40 ± 0.09 (4)	8.19 ± 0.06 (8)	5.83 ± 0.11 (12) ^¶	1.61 ± 0.13 (4)	1.03 ± 0.10 (8)
Mouse WT M₄ mAChR	LY298	4.52 ± 0.07 (4)	8.55 ± 0.06 (4)	5.74 ± 0.07 (8) ^¶	1.78 ± 0.10 (4)	1.30 ± 0.11 (4)*
Mouse WT M₄ mAChR	VU154	4.59 ± 0.06 (4)	8.57 ± 0.06 (3)	6.07 ± 0.09 (7) ^¶	2.43 ± 0.10 (4)	1.75 ± 0.12 (3)*
Human D432E M₄ mAChR	LY298	N.T.	8.28 ± 0.04 (5)	5.86 ± 0.07 (5)	N.T.	1.59 ± 0.06 (5)
Human D432E M₄ mAChR	VU154	N.T.	8.27 ± 0.06 (6)	6.21 ± 0.12 (6)	N.T.	1.04 ± 0.09 (6)
Human T433R M₄ mAChR	LY298	N.T.	8.05 ± 0.08 (5)	5.04 ± 0.04 (5)*	N.T.	1.91 ± 0.11 (5)
Human T433R M₄ mAChR	VU154	N.T.	7.88 ± 0.04 (5)	5.50 ± 0.08 (5)	N.T.	1.67 ± 0.07 (5)*
Human V91L, D432E, T433R M₄ mAChR	LY298	N.T.	7.95 ± 0.10 (4)	5.29 ± 0.26 (4)	N.T.	1.80 ± 0.22 (4)
Human V91L, D432E, T433R M₄ mAChR	VU154	N.T.	7.89 ± 0.12 (4)	6.34 ± 0.16 (4)*	N.T.	1.35 ± 0.16 (4)
Gα_i1 activation (TruPath) interaction assays between ACh or Ipx and LY298 or VU154 on transiently expressed M₄ mAChR constructs in HEK293A cells
Constructs	PAM	log τ ACh^**	log τ Ipx^**	pK_B PAM ^‡	log τ PAM^**	log αβ_ACh^††	log αβ_Ipx^††
Human WT M₄ mAChR	LY298	2.71 ± 0.14 (4)	1.49 ± 0.12 (4)	= 5.65	1.02 ± 0.03 (8) ^¶	2.01 ± 0.14 (4)	1.96 ± 0.16 (4)
Human WT M₄ mAChR	VU154	2.71 ± 0.14 (4)	1.49 ± 0.12 (4)	= 5.83	–0.55 ± 0.08 (8) ^¶	1.22 ± 0.13 (4)	0.20 ± 0.13 (4)
pERK1/2 interaction assays between ACh or Ipx and LY298 or VU154 on stable M₄ mAChR constructs in Flp-In CHO cells
Constructs	PAM	log τ ACh^**	log τ Ipx^**	pK_B PAM ^‡	log τ_C PAM ^{‡ ‡}	log αβ_ACh^††	log αβ_Ipx^††
Human WT M₄ mAChR	LY298	3.27 ± 0.06 (8) ^¶	1.74 ± 0.03 (16) ^¶	= 5.65	1.19 ± 0.05 (12)^**	2.29 ± 0.22 (4)	1.08 ± 0.28 (8)
Human WT M₄ mAChR	VU154	3.27 ± 0.06 (8) ^¶	1.74 ± 0.03 (16) ^¶	= 5.83	0.11 ± 0.05 (12)^**	0.88 ± 0.23 (4)	0.66 ± 0.15 (8)
Mouse WT M₄ mAChR	LY298	N.T.	N.D.	= 5.74	1.32 ± 0.07 (5)	N.T.	1.24 ± 0.12 (4)
Mouse WT M₄ mAChR	VU154	N.T.	N.D.	= 6.07	1.47 ± 0.08 (5) ^{§ §}	N.T.	2.08 ± 0.15 (5) ^{§ §}
Human D432E M₄ mAChR	LY298	N.T.	N.D.	= 5.86	1.34 ± 0.08 (5)	N.T.	1.37 ± 0.28 (5)
Human D432E M₄ mAChR	VU154	N.T.	N.D.	= 6.21	0.78 ± 0.08 (5) ^{§ §}	N.T.	1.02 ± 0.15 (5)
Human T433R M₄ mAChR	LY298	N.T.	N.D.	= 5.04	1.73 ± 0.13 (5) ^{§ §}	N.T.	1.85 ± 0.28 (5)
Human T433R M₄ mAChR	VU154	N.T.	N.D.	= 5.50	0.95 ± 0.12 (5) ^{§ §}	N.T.	1.18 ± 0.14 (5)
Human V91L, D432E, T433R M₄ mAChR	LY298	N.T.	N.D.	= 5.29	1.62 ± 0.09 (5) ^{§ §}	N.T.	1.64 ± 0.30 (5)
Human V91L, D432E, T433R M₄ mAChR	VU154	N.T.	N.D.	= 6.34	0.68 ± 0.06 (5) ^{§ §}	N.T.	1.34 ± 0.11 (5) ^{§ §}

Open in a new tab

Values represent the mean ± SEM with the number of independent experiments shown in parenthesis.

N.T.: not tested; N.D.: not determined; Ach, acetylcholine; Ipx: iperoxo; PAM: positive allosteric modulator.

Number of [³H]-NMS binding sites per cell.

^†

Negative logarithm of the radioligand equilibrium dissociation constant.

^‡

Negative logarithm of the orthosteric (pK_i) or allosteric (pK_B) equilibrium dissociation constant.

^§

Logarithm of the binding cooperativity factor between the agonist (ACh or Ipx) and the PAM (LY298 or VU154).

^¶

Parameter was determined in a shared global analysis between agonists.

^**

Logarithm of the operational efficacy parameter determined using the Operational Model of Agonism.

^††

Logarithm of the functional cooperativity factor between the agonist (ACh or Ipx) and the PAM (LY298 or VU154).

^{‡ ‡}

logτ_C = logarithm of the operational efficacy parameter corrected for receptor expression (methods in Appendix 1).

^{§ §}

Values from pK_B PAM, log α_Ipx, log τ_C PAM, and log αβ_Ipx that are significantly different from human WT M₄ mAChR (p<0.05) calculated by a one-way ANOVA with a Dunnett’s post-hoc test.

We subsequently used the BRET-based TruPath assay (Olsen et al., 2020) as a proximal measure of G protein activation with Gα_i1 (Figure 1—figure supplement 1B). We also used a more amplified downstream signaling assay, extracellular signal-regulated kinases 1/2 phosphorylation (pERK1/2), that is also dependent on G_i activation (Figure 1—figure supplement 2A), to measure the cell-based activity of each PAM with each agonist. These signaling assays allowed us to determine the efficacy of the agonists (τ_A) and the PAMs (τ_B) (Figure 1H, Figure 1—figure supplement 2B). Importantly, efficacy (τ), as defined from the Black–Leff operational model of agonism (Black and Leff, 1983), is determined by the ability of an agonist to promote an active receptor conformation, the receptor density (B_max), and the subsequent ability of a cellular system to generate a response (Figure 1B). Notably, in both signaling assays, the rank order of efficacy was ACh > Ipx > LY298 > VU154. We subsequently calculated the transducer coupling coefficient (τ/K) (Figure 1I, Figure 1—figure supplement 2C), a parameter often used as a starting point to quantify biased agonism (Kenakin et al., 2012) and that is specific to the intact cellular environment in which a given response occurs. Thus the dissociation constant (K) in the transduction coefficient subsumes the affinity for the ground state (non-bound) receptor, in addition to any isomerization states of the receptor that ultimately yield cellular responses (Kenakin and Christopoulos, 2013). Consequently, in both assays, the rank order of transducer coupling was Ipx >> ACh ~ LY298 > VU154 due to Ipx having a higher binding affinity for the receptor. Overall, these results indicate that although ACh is a more efficacious agonist than Ipx, it has lower transducer coupling coefficient. In contrast, LY298 has both better efficacy and transducer coupling coefficient than VU154 (Table 1).

The signaling assays and use of an operational model of allosterism also allowed for the determination of the functional cooperativity (αβ) exerted by the PAMs (Figure 1J, Figure 1—figure supplement 2D), which is a composite parameter accounting for both binding (α) and efficacy (β) modulation. Notably, VU154 displayed lower positive functional cooperativity with ACh than LY298. Strikingly, VU154 had negligible functional modulation with Ipx, in contrast to the cooperativity observed with ACh in the TruPath assay. The tenfold difference in αβ values for VU154 between ACh and Ipx highlights the dependence of the orthosteric probe used in the assay (i.e. probe dependence); on this basis, VU154 would be classified as a ‘neutral’ allosteric ligand (not a PAM) with Ipx in the TruPath assay, that is, VU154 still binds to the allosteric site, but displays neutral cooperativity (αβ = 1) with Ipx (Table 1).

The degree of efficacy modulation (β) that the PAMs have on the agonists can be calculated by subtracting the binding modulation (α) from the functional modulation (αβ) (Figure 1K, Figure 1—figure supplement 2E). A caveat of this analysis is that errors for β are higher due to the error being propagated between calculations. Ideally, the degree of efficacy modulation would be determined in an experimental system where the maximal efficacy of system is not reached by the agonists alone (Berizzi et al., 2016). Nevertheless, our analysis shows the PAMs LY298 and VU154 appear to have a slight negative to neutral effect on agonist efficacy in the G_i1 TruPath and pERK1/2 assays (Table 1), suggesting that the predominant allosteric effect exerted by these PAMs is mediated through modulation of binding affinity.

Collectively, our extensive analysis on the pharmacology of LY298 and VU154 with ACh and Ipx offers detailed insight into the key differences between these ligands across a range of pharmacological properties: ligand binding, probe dependence, efficacy, agonist–receptor–transducer interactions, and allosteric modulation (Figure 1, Table 1). We hypothesized that structures of the human M₄ mAChR in complex with different agonists and PAMs combined with molecular dynamic simulations could provide high-resolution molecular insights into the different pharmacological profiles of these ligands.

Determination of M₄R-G_i1 complex structures

Similar to the approach used in prior determination of active-state structures of the M₁ and M₂ mAChRs (Maeda et al., 2019), we used a human M₄ mAChR construct that lacked residues 242–387 of the third intracellular loop to improve receptor expression and purification, and made complexes of the receptor with G_i1 protein and either the endogenous agonist, ACh, or Ipx. Due to the higher affinity of Ipx compared to ACh (Schrage et al., 2013), we utilized Ipx to form additional M₄R-G_i1 complexes with or without the co-addition of either LY298 or VU154. In all instances, complex formation was initiated by combining purified M₄ mAChR immobilized on anti-FLAG resin with detergent solubilized G_i1 membranes, a single-chain variable fragment (scFv16) that binds Gα_i and Gβ, and the addition of apyrase to remove guanosine 5′-diphosphate (Maeda et al., 2018). For this study, we used a G_i1 heterotrimer composed of a dominant negative form of human Gα_i1, and human Gβ₁ and Gγ_2. (Liang et al., 2018b). Vitrified samples of each complex were imaged using conventional cryo-TEM on a Titan Krios microscope (Danev et al., 2021).

The structures of ACh-, Ipx-, LY298-Ipx-, and VU154-Ipx-bound M₄R-G_i1 complexes were determined to resolutions of 2.8, 2.8, 2.4, and 2.5 Å, respectively (Figure 2A, Figure 2—figure supplement 1, Table 2). For the ACh-bound M₄R-G_i1 complex, an additional focus refinement yielded an improved map of the receptor and binding site (2.75 Å) for modeling (Figure 2—figure supplements 2 and 3). The cryo-EM density maps for all complexes were sufficient for confident placement of backbone and sidechains for most of the receptor, G_i1, and scFv16, and the bound ligands with exception of the alkyne bond of Ipx, which was consistent with prior cryo-EM studies (Maeda et al., 2019; Figure 2B, Figure 2—figure supplement 3).

(A) Cryo-EM maps of Ipx-bound M₄R-G_i1-scFv16 complex with views from the membrane and the extracellular surface. Cryo-EM maps of the other ligand-bound structures are shown in Figure 2—figure supplement 1. (B) Representative EM density around the ligands in this study. EM-maps of Ipx-, LY298-Ipx-, and VU154-Ipx were set to a contour level of 0.011 and the receptor-focused map of ACh- was set to 0.32. (**C–E**) Comparison of the receptor models with bound ligands and views from the (C) membrane, (D) extracellular surface, and (E) intracellular surface.

Table 2. Cryo-electron microscopy (cryo-EM) data collection, refinement, and validation statistics.

	M4R-G_i1-Ipx	M4R-G_i1-Ipx-LY298	M4R-G_i1-Ipx-VU154	M4R-G_i1-ACh
Data collection & refinement
EMD code	26,099	26,100	26,101	26,102
Micrographs	5056	5121	6021	5913
Electron dose (e^-/A²)	66	66	59.5	53.6
Voltage (kV)	300	300	300	300
Pixel size (Å)	0.83	0.83	0.83	0.83
Spot size
Exposure time	4	4	3	5
Movie frames	76	76	75	71
K3 CDS mode	No	No	No	Yes
Defocus range (µm)	0.5–1.5	0.5–1.5	0.5–1.5	0.5–1.5
Symmetry imposed	C1	C1	C1	C1
Particles (final map)	415,743	617,793	677,392	315,595
Resolution @0.143 FSC (Å)	2.8	2.4	2.5	2.8
Refinement
CC_map–model	0.87	0.87	0.88	0.82
Map sharpening B factor (Å²)	–80.9	–60.8	–46.6	–85.1
Model quality
PDB code	7TRK	7TRP	7TRQ	7TRS
R.M.S. deviations
Bond length (Å)	0.004	0.004	0.005	0.006
Bond angles (^o)	0.849	0.811	0.826	0.773
Ramachandran
Favored (%)	98.38	99.14	98.02	98.10
Outliers (%)	0	0	0	0
Rotamer outliers (%)	0.11	0.21	0	0
C-beta deviations (%)	0	0	0	0
Clashscore	2.69	2.62	2.26	4.08
MolProbity score	1.06	1.05	1.00	1.19

Open in a new tab

mAChR: muscarinic acetylcholine receptor; ACh: acetylcholine; Ipx: iperoxo; FSC: Fourier shell correlation.

In all four structures, EM density beyond the top of transmembrane helix 1 (TM1) and the third intracellular loop (ICL3) of the receptor was poorly observed and not modeled. Similarly, the EM density of the α-helical domain of Gα_i1 was poor and not modeled. These regions are highly dynamic and typically not modeled in many class A GPCR-G protein complex structures. Apart from these regions, most amino acid side chains were well resolved in the final EM density maps (Figure 2—figure supplement 3).

Structure and dynamics of agonist binding

Recently, cryo-EM structures of M₄R-G_i1 complexes bound to Ipx, Ipx, and the PAM, LY2119620, and a putative novel allosteric agonist, c110, were determined (Wang et al., 2022). Surprisingly, comparison of the M₄R-G_i1 complex structures revealed larger differences in the position of key orthosteric and allosteric site residues than the M₁R-G₁₁ and M₂R-G_oA complex structures (Figure 2—figure supplement 4). Unfortunately, the quality of density in the EM maps around the orthosteric and allosteric sites of these M₄R-G_i1 structures (Wang et al., 2022) was poor, resulting in several key residues being mismodeled in each site (Figure 2—figure supplement 5). Therefore, differences between the M₄R-G_i1 structures described herein and those by Wang et al., 2022 are highly likely to not be due to genuine differences and, as such, we compared the prior M₁R-G₁₁ and M₂R-G_oA complex structures (Maeda et al., 2019) in this study.

Overall, our M₄R-G_i1 complex structures are similar in architecture to that of other activated class A GPCRs, including the M₁R-G₁₁ and M₂R-G_oA complexes (Figure 2—figure supplement 4). Superposition of the M₄R-G_i1 complexes revealed nearly identical structures with root mean square deviations (RMSD) of 0.4–0.5 Å for the full complexes and 0.3–0.4 Å for the receptors alone (Figure 2C). The largest differences occur around the extracellular surface of the receptors (Figure 2D) along with slight displacements in the position of the αN helix of Gα_i1 and Gβ₁, Gγ₂, and scFv16 with respect to the receptor (Figure 2—figure supplement 1D). The EM density of side chains surrounding the ACh and Ipx binding sites (Figure 3A and B) was well resolved providing the opportunity to understand structural determinants of orthosteric agonist binding. The orthosteric site of the M₄ mAChR, in common with the other mAChR subtypes, is buried within the TM bundle in an aromatic cage that is composed of four tyrosine residues, two tryptophan residues, one phenylalanine residue, and seven other polar and nonpolar residues (Figure 3C). Notably, all 14 of these residues are absolutely conserved across all five mAChR subtypes, underscoring the difficulty in developing highly subtype-selective orthosteric agonists (Burger et al., 2018). Both ACh and Ipx have a positively charged trimethyl ammonium ion that makes cation-π interactions with Y113^3.33, Y416^6.51, Y439^7.39, and Y443^7.43 (Figure 3C; superscript refers to the Ballesteros and Weinstein scheme for conserved class A GPCR residues; Ballesteros and Weinstein, 1995). Likewise, both ACh and Ipx have a polar oxygen atom that can form a hydrogen bond to the indole nitrogen of W164^4.57 with the oxygen of Ipx also being in position to interact with the backbone of N117^3.37 (Figure 3D). Mutation of any of these contact residues reduces the affinity of ACh, validating their importance for agonist binding (Leach et al., 2011; Thal et al., 2016). The largest chemical difference between ACh and Ipx is the bulkier heterocyclic isoazoline group of Ipx that makes a π-π interaction with the conserved residue W413^6.48 (Figure 3D). The residue W413^6.48 is part of the CWxP motif, also known as the rotamer toggle switch, a residue that typically undergoes a change in rotamer between the inactive and active states of class A GPCRs (Shi et al., 2002).

Figure 3. — (**A, B**) Cryo-electron microscopy (cryo-EM) density of the (A) ACh- and (B) Ipx-bound structures. (**C, D**) Interactions at the orthosteric binding site comparing the active state ACh- and Ipx-bound structures with the inactive state tiotropium-bound structure (PDB: 5DSG). Arrows denote relative movement of residues between the inactive and active states. (D) Detailed interactions of ACh and Ipx. Hydrogen bonds are shown as black dashed lines. (**E, F**) Time courses from Gaussian accelerated molecular dynamics (GaMD) simulations of the ACh- and Ipx- bound M₄R-G_i1 cryo-EM structures, each performed with three separate replicates. Individual replicate simulations are illustrated with different colors. The heading of each plot refers to the specific model used in the simulations. Root mean square deviations (RMSDs) of (E) ACh and (F) Ipx from simulations of the cryo-EM structures. (**G, H**) Cross-sections through the ACh- and Ipx-bound structures denoting the relative size of the binding pockets outlined in black.

Figure 3—figure supplement 1. — (**A, B**) Cryo-electron microscopy (cryo-EM) density of the (A) ACh- and (B) Ipx-bound structures. (**C, D**) Interactions at the orthosteric binding site comparing the active state ACh- and Ipx-bound structures with the inactive state tiotropium-bound structure (PDB: 5DSG). Arrows denote relative movement of residues between the inactive and active states. (D) Detailed interactions of ACh and Ipx. Hydrogen bonds are shown as black dashed lines. (**E, F**) Time courses from Gaussian accelerated molecular dynamics (GaMD) simulations of the ACh- and Ipx- bound M₄R-G_i1 cryo-EM structures, each performed with three separate replicates. Individual replicate simulations are illustrated with different colors. The heading of each plot refers to the specific model used in the simulations. Root mean square deviations (RMSDs) of (E) ACh and (F) Ipx from simulations of the cryo-EM structures. (**G, H**) Cross-sections through the ACh- and Ipx-bound structures denoting the relative size of the binding pockets outlined in black.

To investigate the structural dynamics of the M₄ mAChR, we performed three independent 500 ns GaMD simulations on the ACh- and Ipx-bound M₄R-G_i1 cryo-EM structures (Table 3). GaMD simulations revealed that ACh undergoes higher fluctuations in the orthosteric site than Ipx (Figure 3E and F, Videos 1 and 2). Similarly, the interactions of N117^3.37, W164^4.57, and W413^6.48 with Ipx were more stable than those with ACh (Figure 3—figure supplement 1). In the ACh-bound structure, W413^6.48 was in a conformation matching the inactive-state tiotropium-bound structure (Figure 3C and D). GaMD simulations also showed that W413^6.48 sampled a larger conformational space in the ACh-bound structure than in the Ipx-bound structure (Figure 3—figure supplement 1C and G). The predominate χ₂ angle of W413^6.48 was approximately 60^◦ and 105^◦ in the ACh-bound and Ipx-bound simulations, respectively, corresponding to the cryo-EM conformations.

Table 3. Gaussian accelerated molecular dynamics (GaMD) simulations of the M₄ muscarinic acetylcholine receptor (mAChR).

System	Method
M4-G_i1-Ipx (cryo-EM structure)	GaMD (3 × 500 ns)
M4-G_i1-Ipx-VU154 (cryo-EM structure)	GaMD (3 × 500 ns)
M4-G_i1-Ipx-LY298 (cryo-EM structure)	GaMD (3 × 500 ns)
M4-G_i1-ACh (cryo-EM structure)	GaMD (3 × 500 ns)
M4-D432E-G_i1-Ipx-VU154	GaMD (3 × 500 ns)
M4-T433R-G_i1-Ipx-VU154	GaMD (3 × 500 ns)
M4-G_i1-ACh-VU154	GaMD (3 × 500 ns)
M4-G_i1-ACh-LY298	GaMD (3 × 500 ns)
M4-G_i1-VU154	GaMD (3 × 500 ns)
M4-G_i1-LY298	GaMD (3 × 500 ns)
M4-VU154	GaMD (3 × 1000 ns)
M4-LY298	GaMD (3 ×1000 ns)

Open in a new tab

Video 1. Movie from one Ipx-M₄R-G_i1 Gaussian accelerated molecular dynamics (GaMD) simulation.

Download video file^{(1.6MB, mp4)}

Open in a new tab

Video 2. Movie from one ACh-M₄R-G_i1 Gaussian accelerated molecular dynamics (GaMD) simulation.

Download video file^{(1.6MB, mp4)}

Open in a new tab

Located above ACh and Ipx is a tyrosine lid formed by three residues (Y113^3.33, Y416^6.51, and Y439^7.39) that separate the orthosteric binding site from an extracellular vestibule (ECV) at the top of the receptor and the bulk solvent (Figure 3C). In the inactive conformation, the tyrosine lid is partially open due to Y416^6.51 rotating away from the binding pocket to accommodate the binding of bulkier inverse agonists such as tiotropium. In contrast, mAChR agonists are typically smaller in size than antagonists and inverse agonists, and this is reflected in a contraction of the size of the orthosteric binding pocket from 115 Å³ when bound to tiotropium to 77 and 63 Å³ when bound to ACh and Ipx, respectively (Figure 3G and H; Tian et al., 2018). Together, the smaller binding pocket of Ipx and more stable binding interactions with nearby residues that include W413^6.48 likely explain why Ipx has greater than 1000-fold higher binding affinity than ACh.

Structure and dynamics of PAM binding and allosteric modulation of agonist affinity

The M₄R-G_i1 structures of LY298 and VU154 co-bound with Ipx are very similar to the Ipx- and ACh-bound structures, as well as to prior structures of the M₂ mAChR bound to Ipx and the PAM, LY2119620 (Figure 2—figure supplement 4; Kruse et al., 2013; Maeda et al., 2019). Both LY298 and VU154 bind directly above the orthosteric site in the ECV that is composed of a floor delineated by the tyrosine lid, and ‘walls’ formed by residues from TM2, TM6, TM7, ECL2, and ECL3 (Figure 4A and B). The EM density surrounding the PAM binding site and the ECV of the M₄ mAChR were clearly resolved with one exception; in the VU154-bound structure, the EM density begins to weaken around the trifluoromethylsulfonyl moiety (Figure 2B, Figure 4B). This was likely due to the moiety’s ability to freely rotate and a lack of strong interactions with the receptor.

Figure 4. — (**A, B**) Cryo-electron microscopy (cryo-EM) density of the (A) VU154- and (B) LY298-binding sites. (C) The root mean square deviations (RMSDs) between receptor models of the respective cryo-EM structures that were refined into the first and last frames of the EM maps from each principal component (PC1-PC3) of the 3D variability analysis. Values shown are mean ± SEM. (**D, E**) Top representative binding conformations of (D) VU154 and (E) LY298 obtained from structural clustering with frame populations ≥1% and time courses of the RMSDs of each positive allosteric modulator (PAM) relative to the cryo-EM structures. (**F, G**) Binding interactions of VU154 and LY298 with views from the (F) membrane and (G) extracellular surface. (H) Position and χ₂ angle of W435^7.35 in the tiotropium-, ACh-, Ipx-, VU154-Ipx-, and LY298-Ipx bound structures. (**I–K**) Time courses of the W435^7.35 χ₂ angle obtained from Gaussian accelerated molecular dynamics (GaMD) simulations on the (I) Ipx-, (J) VU154-Ipx-, and (K) LY298-Ipx-bound cryo-EM structures. See Table 3.

Figure 4—figure supplement 1. — (**A, B**) Cryo-electron microscopy (cryo-EM) density of the (A) VU154- and (B) LY298-binding sites. (C) The root mean square deviations (RMSDs) between receptor models of the respective cryo-EM structures that were refined into the first and last frames of the EM maps from each principal component (PC1-PC3) of the 3D variability analysis. Values shown are mean ± SEM. (**D, E**) Top representative binding conformations of (D) VU154 and (E) LY298 obtained from structural clustering with frame populations ≥1% and time courses of the RMSDs of each positive allosteric modulator (PAM) relative to the cryo-EM structures. (**F, G**) Binding interactions of VU154 and LY298 with views from the (F) membrane and (G) extracellular surface. (H) Position and χ₂ angle of W435^7.35 in the tiotropium-, ACh-, Ipx-, VU154-Ipx-, and LY298-Ipx bound structures. (**I–K**) Time courses of the W435^7.35 χ₂ angle obtained from Gaussian accelerated molecular dynamics (GaMD) simulations on the (I) Ipx-, (J) VU154-Ipx-, and (K) LY298-Ipx-bound cryo-EM structures. See Table 3.

Given the overall similarities revealed by our four cryo-EM structures, we examined whether there were further differences in the dynamics between the PAM-bound structures by performing a 3D multivariance analysis (3DVA) of the principal components of motion within the Ipx-, LY298-Ipx, VU154-Ipx, and ACh-bound M₄R-G_i1 cryo-EM data sets using Cryosparc (Punjani and Fleet, 2021); a similar analysis performed previously on cryo-EM structures of class A and class B GPCRs provided important insights into the allosteric motions of extracellular domains and receptor interactions with G proteins (Josephs et al., 2021; Liang et al., 2020; Mobbs et al., 2021; Zhang et al., 2020).

In the 3DVA of the Ipx-bound complex, the M₄ mAChR appeared less flexible than the receptor in the ACh-bound complex (Videos 3 and 4) consistent with Ipx having a higher binding affinity and more stable pose during the GaMD simulations (Figure 3E and F). The LY298-Ipx-bound complex appeared similar to the Ipx-bound complex with LY298 being bound in the ECV (Video 5). In contrast, the 3DVA of the VU154 structure had more dynamic movements in the allosteric pocket that could reflect partial binding of VU154 (Video 6). This observation was in line with our findings that VU154 had lower binding modulation (Figure 1E) and functional modulation with agonists than LY298 (Figure 1J, Figure 1—figure supplement 2D, Table 1). To quantify the differences from the 3DVA, we rigid body fitted and refined the respective M₄R-G_i1 models into the first and last frames of the EM maps from each principal component of the 3DVA and then calculated the RMSD between the receptor models (Figure 4C). In agreement with our prior observations, the VU154-Ipx-bound and ACh-bound complexes had greater RMSDs with values of 0.06 and 0.09 Å, respectively. Comparatively, the Ipx-bound and LY298-Ipx-bound complexes had lower RMSD values of 0.02 and 0.001 Å, respectively. The results of the 3DVA do not represent bona fide measures of receptor dynamics, rather they are suggestive of differences between the collected data sets that led to the structures. To support these findings, we compared the GaMD simulations of all four cryo-EM structures (Table 3). Notably, VU154 underwent considerably higher fluctuations than LY298 with RMSDs ranging from 1.5 to 15 Å for VU154 (Video 7) and 0.8–2.1 Å for LY298 (Video 8) relative to the cryo-EM structures (Figure 4D and E). Therefore, the GaMD simulations corroborate our 3DVA results and suggest that complexes bound to agonists with high affinity or co-bound with agonists and PAMs with high positive cooperativity will exhibit lower dynamic fluctuations.

Video 3. 3D variability analysis of the Ipx-M₄R-G_i1 cryo-electron microscopy (cryo-EM) structure.

Download video file^{(4.4MB, mp4)}

Open in a new tab

Video 4. 3D variability analysis of the ACh-M₄R- G_i1 cryo-electron microscopy (cryo-EM) structure.

Download video file^{(5.1MB, mp4)}

Open in a new tab

Video 5. 3D variability analysis of the LY298-Ipx-M₄R- G_i1 cryo-electron microscopy (cryo-EM) structure.

Download video file^{(4.3MB, mp4)}

Open in a new tab

Video 6. 3D variability analysis of the VU154-Ipx-M₄R- G_i1 cryo-electron microscopy (cryo-EM) structure.

Download video file^{(5.2MB, mp4)}

Open in a new tab

Video 7. Movie from one VU154-Ipx-M₄R-G_i1 Gaussian accelerated molecular dynamics (GaMD) simulation.

Download video file^{(1.7MB, mp4)}

Open in a new tab

Video 8. Movie from one LY298-Ipx-M₄R-G_i1 Gaussian accelerated molecular dynamics (GaMD) simulation.

Download video file^{(1.6MB, mp4)}

Open in a new tab

To investigate why the binding of LY298 was more stable than VU154, we examined the ligand interactions with the receptor. There are three key binding interactions that are shared between both PAMs and the M₄ mAChR: (1) a three-way π-stacking interaction between F186^45.51 (ECL2 residues have been numbered 45.X denoting their position between TM4 and TM5 with X.50 being a conserved cysteine residue), the aromatic core of the PAMs, and W435^7.35; (2) a hydrogen bond between Y439^7.39 of the tyrosine lid and the primary amine of the PAMs; and (3) a hydrogen bond between Y89^2.61 and the carbonyl oxygen of the PAMs (Figure 4F and G). While these interactions are conserved for both PAMs in the consensus cryo-EM maps, during GaMD simulations these interactions were more stable with LY298 than VU154 (Figure 4H–K, Figure 4—figure supplement 1). The importance of these interactions was validated pharmacologically (Figure 4—figure supplement 2, Table 4), whereby mutation of any of these residues completely abolished the binding affinity modulation mediated by LY298 and VU154 at the M₄ mAChR with both Ipx and ACh as agonists.

Table 4. Pharmacological parameters of LY298 and VU154 at key M₄ muscarinic acetylcholine receptor (mAChR) mutants.

[³H]-NMS saturation binding on stable M₄ mAChR Flp-In CHO cells
Constructs			Sites per cell^*				pK_D^†
Human WT M₄ mAChR (from Table 1)			598,111 ± 43,067 (7)				9.76 ± 0.05 (7)
Y89A^2.61			32,674 ± 4174 (4)				9.88 ± 0.06 (4)
Q184A^45.49			88,728 ± 3056 (3)				9.99 ± 0.06 (3)
F186A^45.51			36,907 ± 4170 (4)				9.75 ± 0.16 (4)
W435A^7.35			34,861 ± 3510 (3)				9.81 ± 0.22 (3)
Y439A^7.39			42,690 ± 4547 (3)				8.31 ± 0.14 (3)
[³H]-NMS interaction binding assays between ACh or Ipx and LY298 or VU154 on stable M₄ mAChR constructs in Flp-In CHO cells
Constructs	PAM	pK_i ACh^‡		pK_i Ipx^‡	pK_B PAM^‡	log α_ACh^§		log α_Ipx^§	log α_NMS^¶
Human WT M₄	LY298	5.09 ± 0.07 (7)		8.54 ± 0.04 (11)	= 5.65	1.57 ± 0.11		1.71 ± 0.09	= 0
Human WT M₄	VU154	5.06 ± 0.05 (7)		8.54 ± 0.03 (11)	= 5.83	1.44 ± 0.07		1.11 ± 0.06	= 0
Y89A^2.61	LY298	5.25 ± 0.05 (6)		8.48 ± 0.05 (6)	N.D.	N.D.		N.D.	N.D.
Y89A^2.61	VU154	5.27 ± 0.05 (6)		8.47 ± 0.05 (6)	N.D.	N.D.		N.D.	N.D.
Q184A^45.49	LY298	5.24 ± 0.06 (6)		8.74 ± 0.04 (10)	6.23 ± 0.06	1.28 ± 0.13		1.27 ± 0.11	–1.10 ± 0.07
Q184A^45.49	VU154	5.28 ± 0.05 (6)		8.69 ± 0.04 (10)	5.87 ± 0.17	1.07 ± 0.09		0.81 ± 0.07	= 0
F186A^45.51	LY298	4.91 ± 0.05 (6)		8.12 ± 0.05 (8)	N.D.	N.D.		N.D.	N.D.
F186A^45.51	VU154	4.91 ± 0.05 (6)		8.12 ± 0.05 (8)	N.D.	N.D.		N.D.	N.D.
W435^7.35	LY298	3.79 ± 0.07 (7)		6.88 ± 0.07 (7)	N.D.	N.D.		N.D.	N.D.
W435^7.35	VU154	3.79 ± 0.07 (7)		6.88 ± 0.07 (7)	N.D.	N.D.		N.D.	N.D.
Y439A^7.39	LY298	3.23 ± 0.22 (8)		5.36 ± 0.25 (8)	N.D.	N.D.		N.D.	N.D.
Y439A^7.39	VU154	3.23 ± 0.22 (8)		5.36 ± 0.25 (8)	N.D.	N.D.		N.D.	N.D.

Open in a new tab

Values represent the mean ± SEM with the number of independent experiments shown in parenthesis.

N.D.: not determined; ACh: acetylcholine; Ipx: iperoxo; PAM: positive allosteric modulator.

Number of [³H]-NMS binding sites per cell.

^†

Negative logarithm of the radioligand equilibrium dissociation constant.

^‡

Negative logarithm of the orthosteric (pK_i) or allosteric (pK_B) equilibrium dissociation constant. pK_i values for ACh and Ipx are shared at each M4 mAChR construct. pK_B values for the PAMs at Q184A are shared across the agonist data sets.

^§

Logarithm of the binding cooperativity factor between the agonist (ACh or Ipx) and the PAM (LY298 or VU154).

^¶

Logarithm of the binding cooperativity factor between the [³H]-NMS and the PAM (LY298 or VU154).

A potential fourth interaction was observed with residue Q184^45.49 and the amide nitrogen of the PAMs; however, the GaMD simulations suggest that this interaction is relatively weak (Figure 4—figure supplement 1D and H), consistent with the fact that mutation of Q184^45.49 to alanine had no effect on the binding affinity modulation of LY298 or VU154 (Figure 4—figure supplement 2, Table 4). In addition, each PAM had at least one potential unique binding interaction with the receptor (Figure 4F and G). For LY298, this is an interaction between the fluorine atom and N423^6.58 that appeared to be stable during simulation and, when mutated to alanine reduced the binding modulation of LY298 (Figure 4—figure supplement 1K, Figure 4—figure supplement 2, Table 4; Thal et al., 2016). For VU154, there were two additional possible hydrogen bonding interactions with residues Y92^2.64 and T433^7.33 (Figure 4G); however, these interactions were highly fluctuating during GaMD simulations, suggesting they were – at best – transient interactions (Figure 4—figure supplement 1I and J). Finally, W435^7.35 is a key residue in the ECV that changes from a planar rotamer in the agonist-bound structures to a vertical rotamer that π stacks against the PAMs (Figure 4H). In GaMD simulations of the Ipx-bound structure, W435^7.35 is predominantly in a planar conformation that corresponds to its conformation in the cryo-EM structure (Figure 4I). In contrast, the binding of LY298 stabilizes W435^7.35 into a vertical position (Figure 4K). However, in the VU154-bound receptor, W435^7.35 appears to alternate between the planar and vertical positions, consistent with VU154 having a less stable binding pose (Figure 4J). These results indicate that the binding of LY298 is more stable than VU154 due to LY298 being able to form stable binding interactions with key residues in the ECV. This provides a likely explanation for why LY298 was able to exert greater positive binding cooperativity on orthosteric agonists than VU154.

A molecular mechanism of probe dependence

As highlighted above, PAMs, LY298 and VU154, displayed stronger allosteric binding affinity modulation with ACh than Ipx, an example of probe dependence (Figure 1E, Table 1). These findings are in accord with previous studies where we identified probe dependence in the actions of LY298 when tested against other orthosteric agonists (Chan et al., 2008; Suratman et al., 2011). To investigate a mechanism for probe dependence at the M₄ mAChR, we performed GaMD simulations with LY298 and VU154 co-bound with ACh by replacing Ipx with ACh in the corresponding cryo-EM structures (Table 3, Figure 4—figure supplement 3). In the absence of PAM, ACh was more dynamic than Ipx with root-mean-square fluctuations (RMSF) of 2.13 Å versus 0.88 Å, reflective of the fact Ipx binds with higher affinity than ACh (Figure 4—figure supplement 3F). In the presence of LY298 or VU154, the dynamics of ACh binding was decreased, with RMSFs reduced to 1.23 Å and 1.82 Å, respectively, and with LY298 having the greatest effect (Figure 4—figure supplement 3F). This is in line with LY298 having more cooperativity with ACh than VU154 (Figure 1E). In comparison to ACh, there was a modest increase in the dynamics of Ipx with the addition of LY298 or VU154, likely reflecting the fact Ipx binding to the receptor was already stable (Figure 3—figure supplement 1I and J, Figure 4—figure supplement 3F). These results provide a plausible mechanism for probe dependence, at least with regard to differences in the magnitude of the allosteric effect depending on the ligand bound. Namely, PAMs manifest higher cooperativity when interacting with agonists, such as ACh, that are inherently less stable on their own when bound to the receptor, in contrast to more stable ligands such as Ipx.

Structural and dynamic insights into orthosteric and allosteric agonism

In addition to the ability to allosterically modulate the function of orthosteric ligands, it has become increasingly appreciated that allosteric ligands may display variable degrees of direct agonism in their own right, over and above any allosteric modulatory effects (Changeux and Christopoulos, 2016). Prior studies have established that the activation process of GPCRs involves conformational changes that extend from the extracellular domains through to the intracellular surface (Nygaard et al., 2009). Comparison of the active state ACh-, Ipx-, LY298-Ipx-, and VU154-Ipx-bound M₄R-G_i1 structures to the inactive state tiotropium-bound M₄ mAChR structure (Protein Data Bank accession 5DSG) (Thal et al., 2016) thus affords an opportunity to gain new insights into the activation process mediated by multiple orthosteric agonists in the presence and absence of two different PAMs that display high (LY298) and low (VU154) degrees of direct allosteric agonism (Figures 1H and 5A–C).

Figure 5. — (A) Cartoon of the receptor models indicating regions of interest for panels (**B, C**) shown within the red boxes. (B) View of the tiotropium-bound, agonist-bound, and positive allosteric modulator (PAM)-agonist-bound conformations from the extracellular surface. (C) Membrane view of residues and activation motifs involved in signaling. Residues colored red in (**B, C**) indicate residues of investigated in Gaussian accelerated molecular dynamics (GaMD) simulations. (**D–G**) Time course of the root mean square deviations (RMSDs) of the PAMs (**D, E**) from GaMD simulations of the M₄R bound to G protein and no orthosteric agonist, (**F, G**) and in the absence of both G protein and agonist. (**H–K**) Similar to (**D–G**) the time courses of (**H–K**) the W435^7.35 χ₂ angle, (**L–O**) the W413^6.48 χ₂ angle, and (**P–S**) the TM3-TM6 distance measured by distance between R130^3.50 and T399^6.34. See Table 3.

As discussed previously, agonist binding decreases the size of the orthosteric binding site (Figure 3G and H). The primary driver of this decrease was the tyrosine lid residue Y416^6.51, which underwent a large rotation toward Y113^3.33 creating a hydrogen bond that seals off the tyrosine lid (Figure 3C). The closure of the tyrosine lid was further reinforced by a change in the rotamer of W435^7.35 to a planar position that sits parallel to the tyrosine lid allowing for a π-π interaction with Y416^6.51 and a positioning of the indole nitrogen of W435^7.35 to potentially form a hydrogen bond with the hydroxyl of Y89^2.61 (Figure 5B). The contraction of the orthosteric pocket by the inward movement of Y416^6.51 also led to a contraction of the ECV with a 5 Å inward movement of the top of TM6 and ECL3. As a consequence, the top of TM5 was displaced outward by 4 Å forming a new interface between TM5 and TM6 that was stabilized by a hydrogen bond between T424^6.59 and the backbone nitrogen of P193^5.36 along with aromatic interactions between F197^5.40 and F425^6.60 (Figure 5B). These interactions were specific to the active state structures and appear to be conserved as they were also present in the M₁ and M₂ mAChR active state structures (Maeda et al., 2019). In addition to the movements of TM5 and TM6, there was a smaller 1 Å inward movement of ECL2 (Figure 5B). The binding of LY298 and VU154 had a minimal impact on the conformation of most ECL residues, implying that the reorganization of residues in the ECV by orthosteric agonists contributes to the increased affinity of the PAMs (Figure 1G). There was a slight further inward shift of ECL2 toward the PAMs to facilitate the 3-way π-stacking interaction with F186^45.51 and W435^7.35. In addition, in the PAM-bound structures, Y89^2.61 rotated away from its position in the ACh- and Ipx-bound structures either due to a loss of an interaction with W435^7.35 or to form a better hydrogen bond with the carbonyl oxygen of the PAMs (Figure 5B).

Below the orthosteric binding site are several signaling motifs that are important for the activation of class A GPCRs, including the PIF motif (Rasmussen et al., 2011; Wacker et al., 2013), the Na⁺ binding site (Liu et al., 2012a; White et al., 2018), the NPxxY motif (Fritze et al., 2003), and the DRY motif (Figure 5C; Ballesteros et al., 2001). The conformations of these activation motifs were very similar across all four active-state M₄ mAChR structures and were consistent with the position of these motifs across other active-state class A GPCR structures (Zhou et al., 2019). Collectively, all of the described activation motifs facilitate an 11 Å outward movement of TM6 that typifies GPCR activation and creation of the G protein binding site. In comparison to the ECV residues (Figure 5B), beyond the rotamer toggle switch residue W413^6.48, there are no discernible differences between the agonist and PAM-agonist-bound structures, suggesting a shared activation mechanism for residues below W413^6.48 (Figure 5C).

As indicated above, LY298 also displays robust allosteric agonism in comparison to VU154 (Figure 1H, Figure 1—figure supplement 2B). To probe whether the allosteric agonism of LY298 could be related to its ability to better stabilize the M₄ mAChR in an active conformation in comparison to VU154, we performed additional GaMD simulations on the LY298-Ipx- and VU154-Ipx-bound M₄R-G_i1 structures with the agonist Ipx removed (3 × 500 ns) and with both Ipx and the G protein removed (3 × 1000 ns) (Figure 5D–S, Table 3). In GaMD simulations, LY298 underwent lower RMSD fluctuations than VU154 before dissociating from the receptor (Figure 5D–G). Similarly, the conformations of W435^7.35 and W413^6.48 were better stabilized in the LY298-Ipx-bound systems, indicating that LY298 more strongly promotes an active receptor conformation (Figure 5H–K). In the presence of the G protein, both PAMs stabilized an active conformation of the receptor based on the distances between TM3 and TM6 (Figure 5P and Q). Upon removal of the G protein, the VU154-bound M₄ mAChR quickly transitioned toward the inactive conformation, while the LY298-bound M₄ mAChR was more resistant to deactivation in the GaMD simulations (Figure 5R and S). This observation supports LY298 having greater efficacy than VU154 (Table 1) as it better stabilizes the active conformation of the M₄ mAChR. Overall, the GaMD simulations show that in the absence of agonist alone, or agonist and G protein, LY298 better stabilizes activation motifs from the top of the receptor (W435^7.35) all the way down to the intracellular G protein binding pocket (DRY-TM6), providing mechanistic insights into the function of LY298 as a stronger PAM-agonist than VU154.

Structural insights into allosteric modulation of agonist signaling

In a previous study, we characterized over 40 distinct mutations of M₄ mAChR residues that span from the orthosteric site up to the extracellular surface (Table 5; Leach et al., 2011; Nawaratne et al., 2010; Thal et al., 2016). As expected, these studies revealed that mutation of residues around the orthosteric and allosteric sites often resulted in a reduction in the binding affinity of either ACh or LY298 at their respective binding sites, though the allosteric site was typically less affected (Figure 6A and B, Table 5). In contrast, the binding affinity modulation between ACh and LY298 was largely affected by mutation of aromatic residues that link the orthosteric and allosteric sites (Figure 6C), implying a network of residues that were responsible for transmitting binding cooperativity between these two sites (Thal et al., 2016). Analyzing unpublished data from prior studies allowed an examination of the signaling efficacy of ACh (τ_A) and LY298 (τ_B), but also the functional cooperativity (αβ) in the context of active state structures of the co-complexes (Figure 6D–F, Figure 6—figure supplement 1, Table 5).

Table 5. Pharmacological parameters of M₄ muscarinic acetylcholine receptor (mAChR) mutants.

	pERK1/2 interaction assays^*			[³H]-QNB interaction binding assays^†			Study
Constructs	log τ_C ACh ^‡	log τ_C LY298 ^‡	log αβ ^§	pK_i ACh ^¶	pK_B LY298 ^¶	log α ^**
WT M₄ mAChR	2.96 ± 0.14 (4)	1.10 ± 0.09	2.43 ± 0.14	4.51 ± 0.15	4.89 ± 0.12	1.97 ± 0.11	Current/Thal^††
S85A^2.57	3.15 ± 0.11(4)	0.91 ± 0.07	1.75 ± 0.09	4.09 ± 0.11	5.44 ± 0.14	1.43 ± 0.06	Current/Thal
Y89A^2.61	2.53 ± 0.17 (5)	= –3	–0.43 ± 0.27*	5.07 ± 0.45	5.36 ± 0.03	–0.13 ± 0.08 ^{‡ ‡}	Current/Thal
Y92A^2.64	2.26 ± 0.15 (4)	–0.06 ± 0.16*	2.25 ± 0.11	4.15 ± 0.25	4.53 ± 0.15	1.2 ± 0.19 ^{‡ ‡}	Current/Thal
I93T, I94V, K95I	2.57 ± 0.11	2.27 ± 0.19 ^{‡ ‡}	N.T.	4.69 ± 0.11	4.82 ± 0.36	2.14 ± 0.17 ^{‡ ‡}	Nawaratne ^{§ §}
I93T^2.65	2.34 ± 0.09	2.38 ± 0.22 ^{‡ ‡}	N.T.	4.97 ± 0.04	5.36 ± 0.09	2.42 ± 0.16 ^{‡ ‡}	Nawaratne
I94V^2.66	2.34 ± 0.09	1.24 ± 0.09	N.T.	4.71 ± 0.06	5.17 ± 0.08	1.74 ± 0.07	Nawaratne
K95I^2.67	2.00 ± 0.07	0.61 ± 0.09 ^{‡ ‡}	N.T.	4.86 ± 0.05	5.20 ± 0.14	1.24 ± 0.04 ^{‡ ‡}	Nawaratne
Y97A^23.49	2.94 ± 0.11 (4)	2.19 ± 0.07*	3.43 ± 0.12*	4.69 ± 0.17	4.25 ± 0.10 ^{‡ ‡}	2.33 ± 0.12	Current/Thal
W98A^23.50	1.45 ± 0.15* (5)	= –3	2.29 ± 0.10	3.65 ± 0.11 ^{‡ ‡}	4.39 ± 0.04	0.73 ± 0.07 ^{‡ ‡}	Current/Thal
G101A^23.53	2.58 ± 0.09 (4)	0.43 ± 0.08*	2.10 ± 0.07	4.37 ± 0.19	5.03 ± 0.17	1.47 ± 0.01	Current/Thal
D106A^3.26	1.24 ± 0.11	= –3	N.T.	3.95 ± 0.09 ^{‡ ‡}	5.29 ± 0.11	1.51 ± 0.15	Leach ^{¶ ¶}
W108A^3.28	1.49 ± 0.17	= –3	N.T.	4.01 ± 0.06 ^{‡ ‡}	4.24 ± 0.07 ^{‡ ‡}	1.23 ± 0.01 ^{‡ ‡}	Leach
L109A^3.29	1.17 ± 0.14	= –3	N.T.	3.11 ± 0.09 ^{‡ ‡}	4.28 ± 0.14 ^{‡ ‡}	2.54 ± 0.10 ^{‡ ‡}	Leach
D112E^3.32	–0.80 ± 0.16 ^{‡ ‡}	= –3	N.T.	<2	5.56 ± 0.13	0.39 ± 0.11 ^{‡ ‡}	Leach
D112N^3.32	N.D.	N.D.	N.T.	3.19 ± 0.02 ^{‡ ‡}	5.79 ± 0.2 ^{‡ ‡}	0.74 ± 0.08	Leach
Y113A^3.33	N.T.	N.T.	N.T.	2.98 ± 0.12 ^{‡ ‡}	4.97 ± 0.15	0.80 ± 0.10 ^{‡ ‡}	Leach
S116A^3.36	0.82 ± 0.17 ^{‡ ‡}	–0.35 ± 0.45 ^{‡ ‡}	N.T.	3.61 ± 0.10 ^{‡ ‡}	5.12 ± 0.08	1.54 ± 0.05	Leach
N117A^3.37	0.80 ± 0.27 ^{‡ ‡}	–0.27 ± 0.16 ^{‡ ‡}	N.T.	3.64 ± 0.04 ^{‡ ‡}	5.30 ± 0.15	1.57 ± 0.13	Leach
V120A^3.40	1.47 ± 0.11	1.20 ± 0.19	N.T.	5.63 ± 0.05 ^{‡ ‡}	5.41 ± 0.10	1.83 ± 0.11	Leach
D129E^3.49	1.45 ± 0.24	0.78 ± 0.16	N.T.	5.04 ± 0.07	5.59 ± 0.12	1.61 ± 0.16	Leach
D129N^3.49	2.56 ± 0.39	1.86 ± 0.12	N.T.	5.54 ± 0.10 ^{‡ ‡}	5.37 ± 0.20	1.86 ± 0.22	Leach
W164A^4.57	N.D. (3)	= –3	2.17 ± 0.64^***	3.95 ± 0.24	5.15 ± 0.28	ND	Current/Thal
F170A^4.63	3.13 ± 0.17 (5)	2.66 ± 0.12*	3.58 ± 0.17*	4.77 ± 0.2	4.53 ± 0.06	2.23 ± 0.13	Current/Thal
W171A^4.64	2.59 ± 0.17 (5)	1.31 ± 0.11	3.11 ± 0.13	3.91 ± 0.21	4.56 ± 0.15	2.00 ± 0.09	Current/Thal
Q172A^4.65	3.05 ± 0.33 (5)	1.18 ± 0.31	2.71 ± 0.16	4.02 ± 0.09	4.99 ± 0.03	1.54 ± 0.08	Current/Thal
F173A^4.66	3.39 ± 0.11(4)	2.03 ± 0.10*	3.31 ± 0.23*	4.09 ± 0.01	4.78 ± 0.19	1.90 ± 0.14	Current/Thal
Q184A^45.49	3.01 ± 0.10 (4)	1.05 ± 0.08	2.08 ± 0.12	4.25 ± 0.12	5.36 ± 0.04	1.70 ± 0.05	Current/Thal
F186A^45.51	1.99 ± 0.11	N.D.	N.T.	4.85 ± 0.06	NR	NR	Nawaratne
I187A^45.52	2.57 ± 0.10	0.62 ± 0.09	2.07 ± 0.15	3.71 ± 0.12	5.46 ± 0.29	1.07 ± 0.29 ^{‡ ‡}	Current/Thal
Q188A^45.53	2.48 ± 0.15	0.99 ± 0.11	2.35 ± 0.16	4.6 ± 0.22	4.94 ± 0.08	1.49 ± 0.04	Current/Thal
F189A^45.54	2.28 ± 0.11	1.25 ± 0.09	2.67 ± 0.11	4.65 ± 0.02	5.09 ± 0.13	1.99 ± 0.08	Current/Thal
L190A^45.55	2.50 ± 0.14	1.32 ± 0.12	2.81 ± 0.12	4.20 ± 0.06	4.92 ± 0.1	2.06 ± 0.09	Current/Thal
W413A^6.48	0.66 ± 0.12^**, (4)	0.61 ± 0.13^***	3.54 ± 0.09^**,	3.47 ± 0.06 ^{‡ ‡}	4.51 ± 0.37	2.45 ± 0.36	Current/Thal
Y416A^6.51	N.T.	N.T.	N.T.	2.85 ± 0.10 ^{‡ ‡}	NR	NR	Thal
N423A^6.58	3.44 ± 0.15 (3)	0.82 ± 0.10	1.43 ± 0.19*	4.41 ± 0.15	5.02 ± 0.06	1.18 ± 0.08 ^{‡ ‡}	Current/Thal
Q427A^6.62	3.15 ± 0.14 (3)	0.99 ± 0.12	1.64 ± 0.12	4.46 ± 0.03	5.43 ± 0.06	1.36 ± 0.04	Current/Thal
S428P^6.63	1.99 ± 0.09	1.40 ± 0.19	N.T.	5.14 ± 0.03 ^{‡ ‡}	5.17 ± 0.15	1.81 ± 0.11	Nawaratne
D432N^7.32	2.26 ± 0.12	1.25 ± 0.18	N.T.	5.19 ± 0.04 ^{‡ ‡}	5.21 ± 0.2	1.37 ± 0.04	Nawaratne
W435A^7.35	2.58 ± 0.17 (4)	= –3	N.R	3.37 ± 0.08 ^{‡ ‡}	NR	NR	Current/Thal
Y439A^7.39	0.60 ± 0.18 ^{‡ ‡}	N.D.	N.T.	3.33 ± 0.10 ^{‡ ‡}	5.84 ± 0.12	0.49 ± 0.03 ^{‡ ‡}	Nawaratne
W440A^7.40	3.69 ± 0.17^†††(4)	0.84 ± 0.11	1.52 ± 0.20^†††	4.29 ± 0.24	4.94 ± 0.06	0.96 ± 0.04 ^{‡ ‡}	Current/Thal
C442A^7.42	1.49 ± 0.16 ^{‡ ‡}	0.82 ± 0.31	N.T.	4.04 ± 0.07 ^{‡ ‡}	5.35 ± 0.06	1.81 ± 0.03	Nawaratne
Y443A^7.43	0.50 ± 0.16 ^{‡ ‡}	N.D.	N.T.	3.36 ± 0.01 ^{‡ ‡}	6.22 ± 0.05 ^{‡ ‡}	1.16 ± 0.01 ^{‡ ‡}	Nawaratne

Open in a new tab

Values represent the mean ± SEM from three or more independent experiments with the number of individual experimental replicates from the current study shown in parenthesis.

N.T.: not tested; N.D.: not determined; N.R.: no response; ACh, acetylcholine.

Data and analysis from pERK1/2 assays were generated in the current study, Nawaratne et al, J. Bio. Chem. 2010, and Leach et al, Mol. Pharm. 2011. logτ ACh were calculated from the operational model of agonism. log τ LY298 and log αβ were calculated using a simplified operational model of allosterism.

^†

Data and analysis from [³H]-QNB interaction binding assays were generated in Nawaratne et al, J. Bio. Chem. 2010, Leach et al, Mol. Pharm., and Thal et al, Nature 2016.

^‡

logτ_C = logarithm of the operational efficacy parameter corrected for receptor expression using the maximum number of receptor binding sites as previously determined from Nawaratne et al, J. Bio. Chem. 2010, Leach et al, Mol. Pharm., and Thal et al, Nature 2016.

^§

Logarithm of the functional cooperativity factor between ACh and LY298.

^¶

Negative logarithm of the orthosteric (pK_i) or allosteric (pK_B) equilibrium dissociation constant.

^**

Logarithm of the binding cooperativity factor between ACh and LY298.

^††

Values of logτ_C ACh, log τ_C LY298, and log αβ that were calculated in this study. Other parameters are from Thal et al., 2016.

^{‡ ‡}

Values are significantly different from WT M₄ mAChR as determined in previous studies.

^{§ §}

All values are from Nawaratne et al, J. Bio. Chem. 2010 with logτ corrected for receptor expression.

^{¶ ¶}

All values are from Leach et al, Mol. Pharm. 2011.

^***

Parameters determined from the full Operational Model of Allosterism.

^†††

Values are significantly different from WT M₄ mAChR (p<0.05) calculated by a one-way ANOVA with a Dunnett’s post-hoc test.

Figure 6—figure supplement 1. — (**A–F**) M₄ muscarinic acetylcholine receptor (mAChR) alanine point mutations that increase (green colored sticks) or decrease (pink colored sticks) (A) ACh binding, (B) LY298 binding, (C) binding modulation between ACh and LY298, (D) ACh efficacy, (E) LY298 efficacy, (F) and functional modulation by values more than tenfold. Efficacy values are corrected for receptor expression (Gregory et al., 2010) using receptor expression data from Thal et al., 2016. Quantitative data used to identify key residues are from both the current study and previous studies as summarized in Table 5 (Leach et al., 2011; Nawaratne et al., 2010; Thal et al., 2016). (**G–I**) pERK1/2 concentration response curves for interaction of ACh and LY298 at (G) WT and (H) W413A^6.48 M₄ mAChR with (I) values of efficacy and functional modulation. *Indicates statistical significance (p<0.05) relative to WT as determined by a one-way ANOVA with a Dunnett’s post-hoc test that includes the other M₄ mAChR mutants. Data shown are mean ± SEM from three or more experiments performed in duplicate with the pharmacological parameters determined from a global fit of the data.

Mutation of residues that directly surround ACh primarily decreased the efficacy of ACh (Figure 6D, Table 5). One exception was W98^23.50 (an ECL1 residue numbered 23.X denoting its position between TM2 and TM3 with X.50 denoting the most conserved residue), a residue that was recently identified in a deep scanning mutagenesis study as a conserved class A residue that is intolerant to mutation (Jones et al., 2020) and stabilizes the conserved disulfide bridge between ECL1 and TM3 that is important for the stability of the active state of many GPCRs including mAChRs (Hulme, 2013). Interestingly, residues that affect the efficacy of LY298 include nearly all of the residues that also affect ACh efficacy, along with residues that link to the allosteric site and surround the LY298 binding site (Figure 6E, Table 5). This suggests that the direct signaling of LY298 via the allosteric site is nonetheless linked through a similar network of residues and requires a functional orthosteric site for the transduction of signaling, and that mechanism involves equivalent closure of the orthosteric binding site, consistent with the thermodynamic reciprocity of cooperativity (Canals et al., 2011).

Residues Y89^2.61, N432^6.58, W435^7.35, and W440^7.40 were identified as residues that, when mutated to alanine, significantly decreased the functional modulation between ACh and LY298 (Figure 6F, Table 5). In prior work, all four residues were also shown to contribute to LY298 binding or affinity modulation (Thal et al., 2016). Surprisingly, three mutations resulted in increased functional modulation by LY298. Of particular interest was, again, the rotamer toggle switch residue W413^6.48. Mutation of W413^6.48 to alanine significantly impaired the efficacy of ACh but only reduced the efficacy of LY298 by twofold, such that ACh and LY298 had similar efficacy for this mutant (Figure 6G–I, Table 5). Interestingly, the functional modulation (αβ) between ACh and LY298 increased to over 3600 (a 20-fold increase vs. WT) at W413A^6.48. Similar results were observed in the TruPath assay with ACh, Ipx, and LY298 (Figure 6—figure supplement 2 [mutant], Figure 1—figure supplement 1B [WT]). However, with VU154, the functional modulation was considerably reduced with ACh and non-existent with Ipx, in line with our TruPath experiments at the WT M₄ mAChR. These results show that, at the M₄ mAChR, the rotamer toggle switch residue is important for the signaling efficacy of orthosteric agonists and PAM-agonists but does not impair the process of functional allosteric modulation. Thus, suggesting that the stability of LY298 co-binding with agonists can restore impaired function, while the less stable binding of VU154 does not. Together with the observation that most of the structural differences between the active-state M₄ mAChR structures occur at or above W413^6.48, we propose that this residue has a strong role in maintaining the conformational dynamics of the receptor and is a key trigger for robust signal transduction.

A molecular basis of species selectivity

One of the main advantages of allosteric modulators is the ability to selectivity target highly conserved proteins. The mAChRs are the prime example where allosteric modulators have been designed to selectively target specific subtypes. To date, the only PAM-bound mAChR structures are ones with LY2119620, a PAM that has activity at both the M₂ and M₄ mAChRs. Similarly, LY298 has activity at the M₂ mAChR. However, the allosteric properties of VU154 are differentially affected by the species of the receptor (Wood et al., 2017b; Wood et al., 2017a). At the human M₄ mAChR, LY298 displays robust binding affinity modulation, functional modulation, and allosteric agonism, while VU154 has comparatively weaker allosteric properties (Figure 1, Table 1). Conversely, at the mouse M₄ mAChR, VU154 has a high degree of positive binding modulation, functional modulation, and allosteric agonism that is comparable to LY298 at the human M₄ mAChR (Figure 7—figure supplements 1 and 2, Table 1). Therefore, we aimed to determine whether our prior findings could be used to explain the selectivity of VU154 between the human and mouse receptors.

The amino acid sequences of the human and mouse M₄ mAChRs are highly conserved, with most of the differences occurring between the long third intracellular loop and the N- and C- termini. As shown in Figure 7A, only three residues differ between the human and mouse M₄ mAChR with respect to the transmembrane domain. Specifically, residue V91 (L in mouse) at the top of TM2 points into the lipid bilayer, and D432 and T433 (E and R in mouse), which are located at the top of TM7 and form part of the allosteric binding site near VU154.

(A) Comparison of the cryo-electron microscopy (cryo-EM) structure of the human M4 muscarinic acetylcholine receptor (mAChR) bound to Ipx-VU154 with the AlphaFold model of the mouse M4 mAChR (Jumper et al., 2021; Varadi et al., 2022). The three residues that differ between species and within the core 7TM bundle from the human receptor (V91, D432, and T433) are shown as sticks along with the corresponding residues from the mouse receptor. (B) The binding affinity of VU154 for the Ipx-bound conformation (pK_B-Ipx = pK_B + α) determined from [³H]-NMS binding experiments. Values calculated with data from Figure 7—figure supplement 1 with propagated error. (C) Efficacy of VU154 (τ_B – corrected for receptor expression) of pERK1/2 signaling from data in Figure 7—figure supplement 2. (**D–K**) Time courses of obtained from Gaussian accelerated molecular dynamics (GaMD) simulations of the (**D–G**) D432E and (**H–K**) T433R mutant M₄R-Ipx-G_i1-VU154 systems with (**D, H**) Ipx RMSDs, (**E, I**), VU154 root mean square deviations (RMSDs), (**F, J**) W435^7.35 χ₂ angle, and (**G, K**) W413^6.48 χ₂ angle. Data shown are mean ± SEM from three or more experiments performed in duplicate with the pharmacological parameters determined from a global fit of the data. *Indicates statistical significance (p<0.05) relative to WT as determined by a one-way ANOVA with a Dunnett’s post-hoc test.

Figure 7—source data 1. Related to Figure 7.

elife-83477-fig7-data1.xlsx^{(11KB, xlsx)}

Figure 7—figure supplement 1. — (A) Comparison of the cryo-electron microscopy (cryo-EM) structure of the human M4 muscarinic acetylcholine receptor (mAChR) bound to Ipx-VU154 with the AlphaFold model of the mouse M4 mAChR (Jumper et al., 2021; Varadi et al., 2022). The three residues that differ between species and within the core 7TM bundle from the human receptor (V91, D432, and T433) are shown as sticks along with the corresponding residues from the mouse receptor. (B) The binding affinity of VU154 for the Ipx-bound conformation (pK_B-Ipx = pK_B + α) determined from [³H]-NMS binding experiments. Values calculated with data from Figure 7—figure supplement 1 with propagated error. (C) Efficacy of VU154 (τ_B – corrected for receptor expression) of pERK1/2 signaling from data in Figure 7—figure supplement 2. (**D–K**) Time courses of obtained from Gaussian accelerated molecular dynamics (GaMD) simulations of the (**D–G**) D432E and (**H–K**) T433R mutant M₄R-Ipx-G_i1-VU154 systems with (**D, H**) Ipx RMSDs, (**E, I**), VU154 root mean square deviations (RMSDs), (**F, J**) W435^7.35 χ₂ angle, and (**G, K**) W413^6.48 χ₂ angle. Data shown are mean ± SEM from three or more experiments performed in duplicate with the pharmacological parameters determined from a global fit of the data. *Indicates statistical significance (p<0.05) relative to WT as determined by a one-way ANOVA with a Dunnett’s post-hoc test.

Figure 7—source data 1. Related to Figure 7.

elife-83477-fig7-data1.xlsx^{(11KB, xlsx)}

Previous work suggested that residues D432 and T433 were important for differences in the species selectivity of LY298 (Chan et al., 2008). As such, we examined two single D432E and T433R mutants and a V91L/D432E/T433R triple mutant of the human receptor, along with the mouse M₄ mAChR in radioligand binding and pERK1/2 experiments using Ipx and both PAMs (Figure 7—figure supplements 1 and 2, Table 1). For LY298, there were no statistically significant differences in binding or function between species and across the mutants that were more than threefold in effect. In contrast, VU154 had a tenfold higher binding affinity for the Ipx-bound mouse M₄ mAChR (compare Figure 1G with Figure 7B). The affinity of VU154 increased by 2.5-fold at the D432E and T433R mutants and the triple mutant matched the affinity of the mouse receptor (Figure 7B). In functional assays, similar results were observed for VU154 with Ipx at the mouse M₄ mAChR, with significant increases in the efficacy (τ_B – corrected for receptor expression), transduction coefficients (τ_B/K_B), and functional modulation (αβ) (Figure 7B, Figure 7—figure supplements 1 and 2, Table 1). Relative to the WT M₄ mAChR, the efficacy (Figure 7C), transduction coefficients, and functional modulation of VU154 increased for all of the mutants (Figure 7—figure supplements 1 and 2, Table 1); however, none of the values fully matched the mouse receptor. Nevertheless, these results indicate that V91L, D432E, and T433R play a key role in mediating the species selectivity of VU154.

Our prior findings suggest the robust allosteric activity of LY298 at the human M₄ mAChR was due to stable interactions with the receptor. As a proof-of-principle, we questioned whether GaMD simulations would produce a stable binding mode for VU154 with D432E and T433R mutations to the VU154-Ipx-bound M₄R-G_i1 cryo-EM structure that was similar to our previously observed stable binding pose of LY298 (Figure 4). Excitingly, both the D432E and T433R mutants resulted in a dynamic profile of VU154 that matched our GaMD simulations of LY298 from the LY298-Ipx-bound M₄R-G_i1 cryo-EM structure, including stabilized VU154 binding, constrained χ₂ rotamer conformations of W435^7.35 and W413^6.48, and stable binding interactions with Y89^2.61, Y439^7.39, Q184^45.49, and F186^45.51 (Figure 7D–K, Figure 7—figure supplement 3, Videos 9 and 10). The GaMD simulations also suggest that a potential interaction between the mutant residue T433R and the sulfoxide group of VU154 was more stable (5.2 ± 1.5 Å; Figure 7—figure supplement 3I) versus the WT residue T433 (6.56 ± 2.1 Å, Figure 4—figure supplement 1J), albeit the distance of this interaction was far apart and would be better validated by structure determination of VU154 with the mouse M₄ mAChR.

Video 9. Movie from one VU154-Ipx-M₄R(D432E)-G_i1 Gaussian accelerated molecular dynamics (GaMD) simulation.

Download video file^{(1.7MB, mp4)}

Open in a new tab

Video 10. Movie from one VU154-Ipx-M4R(T433R)-G_i1 Gaussian accelerated molecular dynamics (GaMD) simulation.

Download video file^{(1.2MB, mp4)}

Open in a new tab

Collectively, these findings reiterate the importance of receptor dynamics in the determination of allosteric modulator selectivity as even subtle differences in amino acid residues between species may result in profound changes in overall stability of the same PAM-agonist-receptor complex.

Discussion

Major advances have been made in recent years in the appreciation of the role of GPCR allostery and its relevance to modern drug discovery (Changeux and Christopoulos, 2016; Wootten et al., 2013). Despite an increase in the number of reported high-resolution GPCR structures bound to allosteric ligands (Thal et al., 2018), there remains a paucity of molecular-level details about the interplay between the complex chemical and pharmacological parameters that define allostery at GPCRs. By combining detailed pharmacology studies, multiple high-resolution cryo-EM structures of the M₄ mAChR bound to two pharmacologically different agonists and PAMs, and GaMD simulations, we have now provided exquisite in-depth insights into the relationship between both structure and dynamics that govern multiple facets of GPCR allostery (Figure 8A).

Figure 8. — (A) A schematic cartoon illustrating the conformational states of the ligands and the M₄ mAChR when bound to different types of ligands and transducer, along with the resulting dynamic profiles. Pharmacological parameters related to each conformational change are shown. Stable ligand–receptor interactions are denoted by a straight line and less-stable (more dynamic) interactions are denoted by a wavy line. (B) Iperoxo (Ipx) bound the M₄ mAChR with a higher affinity and more stability than ACh but had lower efficacy. ACh being more loosely bound and coupled to G protein may facilitate more G protein turnover accounting for its higher efficacy. (C) LY298 and VU154 bound to the M₄ mAChR with similar affinity for the receptor, but LY298 was found to bind more stably. LY298 had a higher efficacy than VU154, suggesting that allosteric agonism at the M₄ mAChR is mediated by stabilization of the extracellular vestibule (ECV). (D) The positive allosteric modulators (PAMs) LY298 and VU154 display robust binding modulation at the M₄ mAChR with LY298 having a stronger allosteric effect. Both PAMs displayed stronger binding modulation with the agonist ACh versus Ipx, an example of probe dependence. Both PAMs also displayed a slight negative to neutral effect on the efficacy of the agonists, suggesting that their mechanism of action is largely through binding.

Comparison of the ACh- and Ipx-bound M₄ mAChR structures revealed that Ipx bound in a smaller binding pocket (Figure 3G and H), and GaMD simulations showed that Ipx formed more stable interactions with the receptor (Figure 3—figure supplement 1). These observations likely explained why Ipx exhibited greater than 1000-fold higher binding affinity than ACh (Figure 1D), being consistent with studies of other agonists at the β₁-adrenoceptor and the M₁ mAChR (Brown et al., 2021; Warne et al., 2019; Figure 8B). The observation that ACh was a more efficacious agonist than Ipx (Table 1) yet bound with lower affinity and less stable interactions than Ipx was paradoxical. Kenakin and Onaran, 2002 previously opined on the paradox between ligand binding affinity and efficacy and showed via simulations that, in general, there was a negative correlation between binding affinity and efficacy. One interpretation of these results was that the ACh-bound M₄ mAChR more readily sampled receptor conformations that engaged with the transducers (Manglik et al., 2015). Similarly, the ACh-bound M₄ mAChR may also have faster G protein turnover than Ipx due to Ipx-M₄R-G_i1 forming a more stable ternary complex (Furness et al., 2016; Figure 8B).

It is worth noting that structures of GPCRs bound to agonists with different pharmacological properties (full, partial, and biased agonists) have now been reported for some GPCRs (Liang et al., 2018a; Masureel et al., 2018; McCorvy et al., 2018; Ring et al., 2013; Wacker et al., 2013; Warne et al., 2012; Wingler et al., 2019). However, insights gained from such cryo-EM and X-ray crystallography structures may be limited due to the role that the bound transducer plays on the observed final receptor conformation, and not necessarily due solely to the properties of the ligand. The ultimate underlying conformational differences, therefore, are likely to be subtle and dynamic (Seyedabadi et al., 2022), requiring application of additional techniques such as NMR spectroscopy, single-molecule FRET and MD simulations for furthering our understanding (Cao et al., 2021; Cong et al., 2021; Gregorio et al., 2017; Huang et al., 2021; Katayama et al., 2021; Liu et al., 2012b; Solt et al., 2017; Sušac et al., 2018; Xu et al., 2023; Ye et al., 2016).

Indeed, if considering this issue from the perspective of allosteric modulators of GPCRs, our study highlights that two PAMs with distinctly different pharmacological profiles (Figure 1) may bind to and stabilize receptor conformations that were very similar when viewed as static structures (Figure 4). Yet, in contrast, the 3DVA analysis from our cryo-EM structures suggested differences in the dynamics of the cryo-EM structures that were explored further in GaMD simulations (Figure 4C) and revealed that LY298 had a more stable binding pose and interactions with the receptor than VU154 in the PAM-agonist–receptor–transducer-bound conformation. These observations were consistent with LY298 having greater positive binding cooperativity than VU154 (Figure 1E) and suggest that GaMD simulations of GPCRs bound to allosteric ligands could be an extremely valuable tool for drug discovery and optimization (Bhattarai and Miao, 2018).

Pharmacological analysis revealed that LY298 is a better PAM-agonist than VU154 with respect to efficacy (Figure 1H) in the G_i1 TruPath and pERK1/2 signaling assays (Figure 1—figure supplement 2B). GaMD simulations of the PAM–receptor–transducer and PAM–receptor bound complexes, again showed that LY298 more stably interacted with the receptor (Figure 4) and in the absence of G protein better stabilized the duration of the active conformation of the receptor (Figure 5). These findings were not contradictory to our above findings that ACh was more efficacious than Ipx despite having weaker interactions with the receptor because when the affinity of the ligands was accounted for in the transduction coupling coefficients, the rank order was Ipx >> ACh ~ LY298 > VU154 (Figure 1I). Furthermore, these results were in accordance with the observations of Kenakin and Onaran that ligands with the same binding affinity can also have differing efficacies (and vice versa). In addition, the mechanism of agonism for allosteric ligands that bind to the ECV may differ (Xu et al., 2021). Prior work by DeVree et al., 2016 established that allosteric coupling of G proteins to the unliganded active receptor conformation promoted closure of the ECV region. This allosteric coupling is reciprocal and stabilizing the ECV region by PAMs likely leads to increased efficacy (Figure 8).

The PAMs, LY298 and VU154, also displayed stronger allosteric effects with ACh than with Ipx, an observation known as probe dependence (Figure 1E–G). Probe dependence can have substantial implications on how allosteric ligands are detected, validated, and their potential therapeutic utility (Kenakin, 2005). Examples of probe dependence are not limited to studies on mAChRs and have been observed across multiple receptor families (Christopoulos, 2014; Gentry et al., 2015; Pani et al., 2021; Slosky et al., 2020; Wang et al., 2021b). GaMD simulations comparing the PAMs co-bound with either Ipx or ACh showed that the PAMs had a stabilizing effect on ACh, whereas the stability of Ipx was slightly reduced by the PAMs likely because the binding of Ipx was already stable. This is a sensible explanation from thermodynamic principles. Another explanation invokes the two-state receptor model (Canals et al., 2011), which stipulates that the degree of positive modulation for PAMs increases with an increase in the efficacy of the agonists. The pharmacology data support this model as ACh was more efficacious than Ipx and was better modulated by both PAMs (Figure 8D). These observations are also consistent with recent studies that suggest that conformational dynamics between agonist and receptor are important for functional signaling (Bumbak et al., 2020; Cary et al., 2022; Deganutti et al., 2022; O’Connor et al., 2015).

The findings presented here provide new insights into the allosteric signaling and allosteric modulation of GPCRs by combining the analytical analysis of multiple pharmacology assays with cryo-EM structures and GaMD simulations. Overall, these results provide a framework for future mechanistic studies and, ultimately, can aid in the discovery, design, and optimization of allosteric drugs as novel therapeutic candidates for clinical progression.

Limitations of the study

The complexities of GPCR signaling cannot be fully explained by any single receptor or set of experiments. This study was limited to the investigation of two agonists and two PAMs at the human M₄ mAChR. Future studies will be required to determine how these results extrapolate to other classes of ligand, mAChR subtypes, and GPCRs. For instance, this study determined the structures of the M₄ mAChR bound with the ligands ACh, Ipx, Ipx-LY298, and Ipx-VU154. It is possible that structures of the M₄ mAChR bound with ACh-LY298 and ACh-VU154 could reveal different receptor conformations (although GaMD simulations already performed on their docked complexes and the conformational differences between the Ipx-bound cryo-EM structures suggest otherwise). Similarly, structures of the M₄ mAChR bound in complex with either PAM alone may provide better insights into direct allosteric agonism. However, we note that our attempt at determining an LY298-bound complex did not have sufficient stability for the determination of a high-resolution structure, as also supported by our GaMD simulations. Additionally, our cryo-EM structures and MD-simulations utilized an M₄ mAChR sequence with a large portion of the third intracellular loop removed and were complexed with a dominant negative mutant of Gα_i1 and stabilized with the antibody scFv16. This contrasts with our pharmacological characterization of the ligands that were performed on the WT M₄ mAChR. Further investigation into the molecular determinants of species selectivity is also warranted, as is the need for future experiments that incorporate the combined interplay between dynamics/kinetics of ligands, receptor, transducer recruitment and activation.

Materials and methods

Bacterial strains

DH5α (New England Biolabs) and DH10bac (Thermo Fisher Scientific) Escherichia coli cells were grown in LB at 37°C.

Cell culture

Tni and Sf9 cells (Expression Systems) were maintained in ESF-921 media (Expression Systems) at 27°C. Flp-In Chinese hamster ovary (CHO) (Thermo Fisher Scientific) cells stably expressing human M₄ mAChR or mutant constructs were maintained in Dulbecco’s modified Eagle’s medium (DMEM, Invitrogen) containing 5% fetal bovine serum (FBS; ThermoTrace) and 0.6 μg/ml of Hygromycin (Roche) in a humidified incubator (37°C, 5% CO₂, 95% O₂). HEK293A cells were grown in DMEM supplemented with 5% FBS at 37°C in 5% CO₂. Cell lines were authenticated by vendor and confirmed negative for mycoplasma contamination using the Lonza MycoAlert Mycoplasma Detection Kit (#LT07-318).

Radioligand binding assays

Flp-In CHO cells stably expressing M₄ mAChR constructs were seeded at 10,000 cells/well in 96-well white clear bottom isoplates (Greiner Bio-one) and allowed to adhere overnight at 37°C, 5% CO₂, and 95% O₂. Saturation binding assay was performed to quantify the receptor expression and equilibrium dissociation constant of the radioligand [³H]-NMS (PerkinElmer, specific activity 80 Ci/mmol). Briefly, plates were washed once with phosphate-buffered saline (PBS) and incubated overnight at room temperature (RT) with 0.01–10 nM [³H]-NMS in Hanks’s balanced salt solution (HBSS)/10 mM HEPES (pH 7.4) in a final volume of 100 μl. For binding interaction assays, cells were incubated overnight at RT with a specific concentration of [³H]-NMS (pK_D determined at each receptor in saturation binding) and various concentrations of ACh or Ipx in the absence or presence of increasing concentrations of each allosteric modulator. In all cases, nonspecific binding was determined by the coaddition of 10 μM atropine (Sigma). The following day, the assays were terminated by washing the plates twice with ice-cold 0.9% NaCl to remove the unbound radioligand. Cells were solubilized in 100 μl per well of Ultima Gold (PerkinElmer), and radioactivity was measured with a MicroBeta plate reader (PerkinElmer).

G protein activation assay

Upon 60–80% confluence, HEK293A cells were transfected transiently using polyethylenimine (PEI, Polysciences) and 10 ng per well of each of pcDNA3.1-hM4 mAChR (WT or mutant), pcDNA5/FRT/TO-Gα_i1-RLuc8, pcDNA3.1-β₃, and pcDNA3.1-Gγ₉-GFP2 at a ratio of 1:1:1:1 ratio with 40 ng of total DNA per well. Cells were plated at 30,000 cells per well into 96-well Greiner CELLSTAR white-walled plates (Sigma-Aldrich). 48 hr later, cells were washed with 200 μl phosphate buffer saline (PBS) and replaced with 70 μL of 1× HBSS with 10 mM HEPES. Cells were incubated for 30 min at 37°C before addition of 10 μl of 1.3 μM Prolume Purple coelenterazine (Nanolight Technology). Cells were further incubated for 10 min at 37C° before BRET measurements were performed on a PHERAstar plate reader (BMG Labtech) using 410/80 nm and 515/30 nm filters. Baseline measurements were taken for 8 min before addition of drugs or vehicle to give a final assay volume of 100 μl and further reading for 30 min. BRET signal was calculated as the ratio of 515/30 nm emission over 410/80 nm emission. The ratio was vehicle corrected using the initial 8 min of baseline measurements and then baseline corrected again using the vehicle-treated wells. Data were normalized using the maximum agonist response to allow for grouping of results using an area under the curve analysis in Prism. Data were analyzed at timepoints of 4, 10, and 30 min yielding similar results.

Phospho-ERK1/2 assay

The level of phosphorylated extracellular signal-regulated protein kinase 1/2 (pERK1/2) was detected using the AlphaScreen SureFire Kit (PerkinElmer Life and Analytical Sciences). Briefly, FlpIn CHO cells stably expressing the receptor were seeded into transparent 96-well plates at a density of 20,000 cells/well and grown overnight at 37°C, 5% CO₂. Cells were washed with PBS and incubated in serum-free DMEM at 37°C for 4 hr to allow FBS-stimulated pERK1/2 levels to subside. Cells were stimulated with increasing concentrations of ACh or Ipx in the absence or presence of increasing concentrations of the allosteric modulator at 37°C for 5 min (the time required to maximally promote ERK phosphorylation for each ligand at each M₄ mAChR construct in the initial time-course study; data not shown). For all experiments, stimulation with 10% (v/v) FBS for 5 min was used as a positive control. The reaction was terminated by the removal of media and lysis of cells with 50 μl of the SureFire lysis buffer (TGR Biosciences). Plates were then agitated for 5 min and 5 μl of the cell lysate was transferred to a white 384-well ProxiPlate (Greiner Bio-one) followed by the addition of 5 μl of the detection buffer (a mixture of activation buffer:reaction buffer:acceptor beads:donor beads at a ratio of 50:200:1:1). Plates were incubated in the dark for 1 hr at 37°C followed by measurement of fluorescence using an Envision plate reader (PerkinElmer) with standard AlphaScreen settings. Data were normalized to the maximal response mediated by 10 μM ACh, Ipx, or 10% FBS.

Purification of scFv16

Tni insect cells were infected with scFv16 baculovirus at a density of 4 million cells per ml and harvested at 60 hr post infection by centrifugation for 10 min at 10,000 × g. The supernatant was pH balanced to pH 7.5 by the addition of Tris pH 7.5, and 5 mM CaCl₂ was added to quench any chelating agents, then left to stir for 1.5 hr at RT. The supernatant was then centrifuged at 30,000 × g for 15 min to remove any precipitates. 5 ml of EDTA-resistant Ni resin (Cytivia) was added and incubated for 2 hr at 4^oC while stirring. Resin was collected in a glass column and washed with 20 column volumes (CVs) of high salt buffer (20 mM HEPES pH 7.5, 500 mM NaCl, 20 mM imidazole) followed by 20 CVs of low salt buffer (20 mM HEPES pH 7.5, 100 mM NaCl, 20 mM imidazole). Protein was then eluted using 8 CV of elution buffer (20 mM HEPES pH 7.5, 100 mM NaCl, 250 mM imidazole) until no more protein was detected using Bradford reagent (Bio-Rad Laboratories). Protein was concentrated using a 10 kDa Amicon filter device (Millipore) and aliquoted into 1 mg aliquots for further use.

Expression and purification of M₄R-G_i1-scFv16 complexes

The human M₄ mAChR with residues 242–387 of the third intracellular loop removed and the N-terminal glycosylation sites (N3, N9, N13) mutated to D was expressed in Sf9 insect cells, and human DNG_αi1 and His6-tagged human G_β1γ2 were co-expressed in Tni insect cells. Cell cultures were grown to a density of 4 million cell per ml for Sf9 cells and 3.6 million per ml for Tni cells and then infected with either M₄ mAChR baculovirus or both G_αi1 and G_β1γ2 baculovirus, at a ratio of 1:1. M₄ mAChR expression was supplemented with 10 mM atropine. Cultures were grown at 27°C and harvested by centrifugation 60–72 hr (48 hr for Hi5 cells) post infection. Cells were frozen and stored at –80°C for later use. 1–2 l of the frozen cells were used for each purification.

Cells expressing M₄ mAChR were thawed at RT and then dounced in the solubilization buffer containing 20 mM HEPES pH 7.5, 10% glycerol, 750 mM NaCl, 5 mM MgCl₂, 5 mM CaCl₂, 0.5% LMNG, 0.02% CHS, 10 µM atropine, and cOmplete Protease Inhibitor Cocktail (Roche) until homogeneous. The receptor was solubilized for 2 hr at 4°C while stirring. The insoluble material was removed by centrifugation at 30,000 × g for 30 min followed by filtering the supernatant and batch-binding immobilization to M1 anti-flag affinity resin, previously equilibrated with high salt buffer, for 1 hr at RT. The resin with immobilized receptor was then washed using a peristaltic pump for 30 min at 2 ml/min with high salt buffer: 20 mM HEPES pH 7.5, 750 mM NaCl, 5 mM MgCl₂, 5 mM CaCl₂, 0.5% lauryl maltose neopentyl glycol (LMNG, Anatrace), 0.02% cholesterol hemisuccinate (CHS, Anatrace) followed by low salt buffer: 20 mM HEPES pH 7.5, 100 mM NaCl, 5 mM MgCl₂, 5 mM CaCl₂, 0.5% LMNG, 0.02% CHS, and an agonist (5 µM Ipx, 1 µM Ipx with 10 µM VU154, or 100 µM ACh). While the receptor was immobilized on anti-FLAG resin, the DNGα_i1 cell pellet was thawed, dounced, and solubilized in the solubilization buffer containing 20 mM HEPES pH 7.5, 100 mM NaCl, 5 mM MgCl₂, 5 mM CaCl₂, 0.5% LMNG, 0.02% CHS, apyrase (five units), and cOmplete Protease Inhibitor Cocktail. DNGα_i1 was solubilized for 2 hr at 4°C followed by the centrifugation at 30,000 × g for 30 min to remove the insoluble material. Supernatant was filtered through a glass fiber filter (Millipore) and then added to the receptor bound to anti-Flag resin. Apyrase (five units), scFv16, and agonist (either 1 µM Ipx, 1 µM Ipx with 10 µM VU154, or 100 µM ACh) were added and incubated for 1 hr at RT with gentle mixing. The anti-FLAG resin was then loaded onto a glass column and washed with approximately 20 CVs of washing buffer: 20 mM HEPES pH 7.4, 100 mM NaCl, 5 mM MgCl₂, 5 mM CaCl₂, 0.01% LMNG, 0.001% CHS, agonist (1 µM Ipx, 1 µM Ipx with 10 µM VU154, or 100 µM ACh). Complex was eluted with size-exclusion chromatography (SEC) buffer: 20 mM HEPES pH 7.5, 100 mM NaCl, 5 mM MgCl₂, 0.01% LMNG, 0.001% CHS and agonist (1 µM Ipx, or 1 µM Ipx with 10 µM VU154, or 100 µM ACh) with the addition of 10 mM EGTA and 0.1 mg/mL FLAG peptide. After the elution, an additional 1–2 mg of scFv16 was added and shortly incubated on ice before concentrating using a 100 kDa Amicon filter to a final volume of 500 µl. The sample was filtered using a 0.22 µm filter followed by SEC using a Superdex 200 increase 10/300 column (Cytivia) using SEC buffer. For the ACh- and VU154-Ipx-bound samples, the fractions containing protein were concentrated again and re-run over SEC using a buffer with half the amount of detergent in order to remove empty micelles. Samples were concentrated and flash frozen using liquid nitrogen. In case of the LY298-Ipx-bound sample, the sample was purified with 1 µM Ipx only. After SEC, the sample was then split in half, where one half was incubated with approximately 1.6 µM LY298 at 4°C overnight, and then concentrated and flash frozen in liquid nitrogen.

EM sample preparation and data acquisition

Samples (3 µl) were applied to glow-discharged Quantifoil R1.2/1.3 Cu/Rh 200 mesh grids (Quantifoil) (M4R-G_i1-Ipx and M4R-G_i1-Ipx-LY298) or UltrAuFoil R1.2/1.3 Au 300 mesh grids (Quantifoil) (M4R-G_i1-Ipx-VU154 and M4R-G_i1-Ach) and were vitrified on a Vitrobot Mark IV (Thermo Fisher Scientific) set to 4°C and 100% humidity and 10 s blot time. Data were collected on a Titan Krios G3i 300 kV electron microscope (Thermo Fisher Scientific) equipped with GIF Quantum energy filter and K3 detector (Gatan). Data acquisition was performed in EFTEM NanoProbe mode with a 50 µM C2 aperture at an indicated magnification of ×105,000 with zero-loss slit width of 25 eV. The data were collected automatically with homemade scripts for SerialEM performing a nine-hole beam-image shift acquisition scheme with one exposure in the center of each hole. Experimental parameters specific to each collected data set is listed in Table 2.

Image processing

Specific details for the processing of each cryo-EM data set are shown in Figure 2—figure supplement 2. Image frames for each movie were motion corrected using MotionCor2 (Zheng et al., 2017) and contrast transfer function (CTF)-estimated using GCTF (Zhang, 2016). Particles were picked from corrected micrographs using crYOLO (Wagner et al., 2019) or RELION-3.1 software Zivanov et al., 2018 followed by reference-free 2D and 3D classifications. Particles within bad classes were removed and remaining particles subjected to further analysis. Resulting particles were subjected to Bayesian polishing, CTF refinement, 3D auto-refinement in RELION, followed by another round of 3D classification and 3D refinement that yielded the final maps (Zivanov et al., 2018). Local resolution was determined from RELION using half-reconstructions as input maps. Due to the high degree of conformational flexibility between the receptor and G protein, a further local refinement was performed in cryoSPARC for the ACh-bound M₄R-complex. A receptor-focused map was generated (2.75 Å), which was used to generate a PDB model of the ACh-bound M₄R.

Model building and refinement

An initial M₄R template model was generated from our prior modeling studies of the M₄ mAChR that was based on an active state M₂ mAChR structure (PBD: 4MQT) (Kruse et al., 2013). An initial model for dominant negative Gα_i1Gβ₁Gγ₂ was from a structure in complex with Smoothend (PDB: 6OT0) (Qi et al., 2019) and scFv16 from the X-ray crystal structure in complex with heterotrimeric G protein (PDB: 6CRK) (Maeda et al., 2018). Models were fit into EM maps using UCSF Chimera (Pettersen et al., 2004), and then rigid-body-fit using PHENIX (Liebschner et al., 2019), followed by iterative rounds of model rebuilding in Coot (Casañal et al., 2020) and ISOLDE (Croll, 2018), and real-space refinement in PHENIX. Restrains for all ligands were generated from the GRADE server (https://grade.globalphasing.org). Model validation was performed with MolProbity (Williams et al., 2018) and the wwPDB validation server (Berman et al., 2003). Figures were generated using UCSF Chimera (Pettersen et al., 2004), Chimera X (Pettersen et al., 2021), and PyMOL (Schrödinger).

Cryo-EM 3D variability analysis

3D variability analysis (3DVAR) was performed to access and visualize the dynamics within the cryo-EM datasets of the M₄ mAChR complexes, as previously described using cryoSPARC (Punjani and Fleet, 2021). The polished particle stacks were imported into cryoSPARC, followed by 2D classification and 3D refinement using the respective low-pass-filtered RELION consensus maps as an initial model. 3DVA was analyzed in three components with 20 volume frames of data per component of motion. Output files were visualized using UCSF Chimera (Pettersen et al., 2004).

Gaussian accelerated molecular dynamics (GaMD)

GaMD enhances the conformational sampling of biomolecules by adding a harmonic boost potential to reduce the system energy barriers (Miao et al., 2015). When the system potential $V (\overset{⃑}{r})$ is lower than a reference energy E, the modified potential $V^{} (\overset{⃑}{r})$ of the system is calculated as

V^{} (\overset{⃑}{r}) = V (\overset{⃑}{r}) + ∆ V (\overset{⃑}{r})

Δ V (\vec{r}) = {\begin{cases} \frac{1}{2} k {(E - V (\vec{r}))}^{2}, V (\vec{r}) < E \\ 0, V (\vec{r}) \geq E, \end{cases}

(1)

where k is the harmonic force constant. The two adjustable parameters E and k are automatically determined on three enhanced sampling principles. First, for any two arbitrary potential values $v_{1} (\overset{⃑}{r})$ and $v_{2} (\overset{⃑}{r})$ found on the original energy surface, if $V_{1} (\vec{r}) < V_{2} (\vec{r})$ , $∆ V$ should be a monotonic function that does not change the relative order of the biased potential values; that is, $V_{1}^{} (\vec{r}) < V_{2}^{} (\vec{r})$ . Second, if $V_{1} (\vec{r}) < V_{2} (\vec{r})$ , the potential difference observed on the smoothened energy surface should be smaller than that of the original; i.e., $V_{2}^{} (\vec{r}) - V_{1}^{} (\vec{r}) < V_{2} (\vec{r}) - V_{1} (\vec{r})$ . By combining the first two criteria and plugging in the formula of $V^{} (\overset{⃑}{r})$ and $∆ V$ , we obtain

V_{m a x} \leq E \leq V_{m i n} + \frac{1}{k}

(2)

where $V_{m i n}$ and $V_{m a x}$ are the system minimum and maximum potential energies. To ensure that Equation 2 is valid, k has to satisfy $k \leq 1 / (V_{m a x} - V_{m i n})$ . Let us define $k = k_{0} ∙ 1 / (V_{m a x} - V_{m i n})$ , then $0 k_{0} \leq 1$ . Third, the standard deviation (SD) of $∆ V$ needs to be small enough (i.e. narrow distribution) to ensure accurate reweighting using cumulant expansion to the second order: $σ_{Δ V} = k (E - V_{a v g}) σ_{V} \leq σ_{0}$ , where $V_{a v g}$ and $σ_{V}$ are the average and SD of $∆ V$ with $σ_{0}$ as a user-specified upper limit (e.g. ${10 k}_{B} T$ ) for accurate reweighting. When E is set to the lower bound $E = V_{m a x}$ according to Equation 2, $k_{0}$ can be calculated as

k_{0} = m i n (1.0, k_{0}^{`}) = m i n (1.0, \frac{σ_{0}}{σ_{V}} ∙ \frac{V_{m a x} - V_{m i n}}{V_{m a x} - V_{a v g}})

(3)

Alternatively, when the threshold energy E is set to its upper bound $E = V_{m i n} + 1 / k$ , $k_{0}$ is set to

k_{0} = k_{0}^{` `} \equiv (1 - \frac{σ_{0}}{σ_{V}}) ∙ \frac{V_{m a x} - V_{m i n}}{V_{a v g} - V_{m i n}}

(4)

If $k_{0}^{` `}$ is calculated between 0 and 1. Otherwise, $k_{0}$ is calculated using Equation 3.

Energetic reweighting of GaMD simulations

For energetic reweighting of GaMD simulations to calculate potential of mean force (PMF), the probability distribution along a reaction coordinate is written as $p^{} (A)$ . Given the boost potential $∆ V (r)$ of each frame, $p^{} (A)$ can be reweighted to recover the canonical ensemble distribution $p (A)$ , as

p (A_{j}) = p^{} (A_{j}) \frac{{⟨e^{β ∆ V (r)}⟩}_{j}}{\sum_{i = 1}^{M} {⟨{p^{} (A_{i}) e}^{β ∆ V (r)}⟩}_{i}}, j = 1, \dots, M

(5)

where M is the number of bins, $β = k_{B} T$ , and ${⟨e^{β ∆ V (r)}⟩}_{j}$ is the ensemble-averaged Boltzmann factor of $∆ V (r)$ for simulation frames found in the jth bin. The ensemble-averaged reweighting factor can be approximated using cumulant expansion:

⟨e^{β ∆ V (r)}⟩ = e x p \{\sum_{k = 1}^{\infty} \frac{β^{k}}{k!} C_{k}\}

(6)

where the first two cumulants are given by

\begin{matrix} C_{1} = ⟨∆ V⟩, \\ C_{2} = ⟨∆ V^{2}⟩ - {⟨∆ V⟩}^{2} = σ_{v}^{2} . \end{matrix}

(7)

The boost potential obtained from GaMD simulations usually follows near-Gaussian distribution (Miao and McCammon, 2017). Cumulant expansion to the second order thus provides a good approximation for computing the reweighting factor (Miao et al., 2015; Miao et al., 2014). The reweighted free energy $F (A) = {- k}_{B} T l n p (A)$ is calculated as

F (A) = F^{} (A) - \sum_{k = 1}^{2} \frac{β^{k}}{k!} C_{k} + F_{c}

(8)

where $F^{} (A) = {- k}_{B} T l n p^{} (A)$ is the modified free energy obtained from GaMD simulation and $F_{c}$ is a constant.

System setup

The M₄R-ACh-G_i1, M₄R-Ipx-G_i1, M₄R-Ipx-G_i1-VU154, and M₄R-Ipx-G_i1-LY298 cryo-EM structures were used for setting up simulation systems. The scFv16 in the cryo-EM structures was omitted in all simulations. The initial structures of single mutant D432E and T433R mutant of M₄R-Ipx-G_i1-VU154 were obtained by mutating the corresponding residues in the M₄R-Ipx-G_i1-VU154 cryo-EM structure. The initial structures of M₄R-ACh-G_i1-VU154 and M₄R-ACh-G_i1-LY298 were obtained from M₄R-Ipx-G_i1-VU154 and M₄R-Ipx-G_i1-LY298 cryo-EM structures by replacing Ipx with ACh through alignment of receptors to the M4R-ACh-G_i1 cryo-EM structure. The initial structures of M₄R-G_i1-VU154 and M₄R-G_i1-LY298 were obtained by removing the corresponding Ipx agonist from the M₄R-Ipx-G_i1-VU154 and M₄R-Ipx-G_i1-LY298 cryo-EM structures. The initial structures of M₄R-VU154 and M₄R-LY298 were obtained by removing the corresponding Ipx agonist and G_i1 protein from the M₄R-Ipx-G_i1-VU154 and M₄R-Ipx-G_i1-LY298 cryo-EM structures. According to previous findings, intracellular loop (ICL) 3 is highly flexible and removal of ICL3 does not appear to affect GPCR function (Dror et al., 2015; Dror et al., 2011). The ICL3 was thus omitted as in the current GaMD simulations. Similar to a previous study, helical domains of the G_i1 protein missing in the cryo-EM structures were not included in the simulation models. This was based on earlier simulation of the β₂AR-G_s complex, which showed that the helical domain fluctuated substantially (Dror et al., 2015). All chain termini were capped with neutral groups (acetyl and methylamide). All the disulfide bonds in the complexes (i.e. Cys108^3.25-Cys185^45x50 and Cys426^ECL3-Cys429^ECL3 in the M4R) that were resolved in the cryo-EM structures were maintained in the simulations. Using the psfgen plugin in VMD (Humphrey et al., 1996), missing atoms in protein residues were added and all protein residues were set to the standard CHARMM protonation states at neutral pH. For each of the complex systems, the receptor was inserted into a palmitoyl-oleoyl-phosphatidyl-choline (POPC) bilayer with all overlapping lipid molecules removed using the membrane plugin in VMD. The system charges were then neutralized at 0.15 M NaCl using the solvate plugin in VMD (Humphrey et al., 1996). The simulation systems were summarized in Table 3.

Simulation protocol

The CHARMM36M parameter set (Huang et al., 2017; Klauda et al., 2010; Vanommeslaeghe and MacKerell, 2015) was used for the M₄ mAChRs, G_i1 proteins, and POPC lipids. Force field parameters of agonists ACh and Ipx, PAMs LY298 and VU154 were obtained from the CHARMM ParamChem web server (Vanommeslaeghe et al., 2012b; Vanommeslaeghe and MacKerell, 2012a). Force field parameters with high penalty were optimized with FFParm (Kumar et al., 2020). GaMD simulations of these systems followed a similar protocol used in previous studies of GPCRs (Draper-Joyce et al., 2021; Miao and McCammon, 2018; Miao and McCammon, 2016). For each of the complex systems, initial energy minimization, thermalization, and 20 ns cMD equilibration were performed using NAMD2.12 (Phillips et al., 2005). A cutoff distance of 12 Å was used for the van der Waals and short-range electrostatic interactions and the long-range electrostatic interactions were computed with the particle-mesh Ewald summation method (Darden et al., 1993). A 2-fs integration time step was used for all MD simulations, and a multiple-time-stepping algorithm was used with bonded and short-range non-bonded interactions computed every time step and long-range electrostatic interactions every two-time steps. The SHAKE algorithm (Ryckaert et al., 1977) was applied to all hydrogen-containing bonds. The NAMD simulation started with equilibration of the lipid tails. With all other atoms fixed, the lipid tails were energy minimized for 1000 steps using the conjugate gradient algorithm and melted with a constant number, volume, and temperature (NVT) run for 0.5 ns at 310 K. The 12 systems were further equilibrated using a constant number, pressure, and temperature (NPT) run at 1 atm and 310 K for 10 ns with 5 kcal/(mol. Å²) harmonic position restraints applied to the protein and ligand atoms. Final equilibration of each system was performed using a NPT run at 1 atm pressure and 310 K for 0.5 ns with all atoms unrestrained. After energy minimization and system equilibration, conventional MD simulations were performed on each system for 20 ns at 1 atm pressure and 310 K with a constant ratio constraint applied on the lipid bilayer in the X-Y plane.

With the NAMD output structure, along with the system topology and CHARMM36M force field files, the ParmEd tool in the AMBER package was used to convert the simulation files into the AMBER format. The GaMD module implemented in the GPU version of AMBER20 (Case et al. 2020) was then applied to perform the GaMD simulation. GaMD simulations of systems with G_i1 protein (M₄R-ACh-G_i1, M₄R-Ipx-G_i1, M₄R-Ipx-G_i1-VU154, M₄R-Ipx-G_i1-LY298, M₄R-ACh-G_i1-VU154, M₄R-ACh-G_i1-LY298, single mutant D432E and T433R mutants of M₄R-Ipx-G_i1-VU154) included an 8-ns short cMD simulation used to collect the potential statistics for calculating GaMD acceleration parameters, a 48-ns equilibration after adding the boost potential, and finally three independent 500-ns GaMD production simulations with randomized initial atomic velocities. The average and SD of the system potential energies were calculated every 800,000 steps (1.6 ns). GaMD simulations of M₄R-VU154 and M₄R-LY298 included a 2.4-ns short cMD simulation used to collect the potential statistics for calculating GaMD acceleration parameters, a 48-ns equilibration after adding the boost potential, and finally three independent 1000-ns GaMD production simulations with randomized initial atomic velocities. The average and SD of the system potential energies were calculated every 240,000 steps (0.48 ns). All GaMD simulations were run at the ‘dual-boost’ level by setting the reference energy to the lower bound. One boost potential is applied to the dihedral energetic term and the other to the total potential energetic term. The upper limit of the boost potential SD, σ₀ was set to 6.0 kcal/mol for both the dihedral and the total potential energetic terms. Similar temperature and pressure parameters were used as in the NAMD simulations.

Simulation analysis

CPPTRAJ (Roe and Cheatham, 2013) and VMD (Humphrey et al., 1996) were used to analyze the GaMD simulations. The RMSDs of the agonist ACh and Ipx, PAM VU154 and LY298 relative to the simulation starting structures, the interactions between receptor and agonists/PAMs, distances between the receptor TM3 and TM6 intracellular ends were selected as reaction coordinates. Particularly, distances were calculated between the Cα atoms of residues Arg^3.50 and Thr^6.30, N atom of residue N117^3.37 and carbon atom (C5) in the acetyl group of ACh or oxygen atom (O09) in the ether bond of Ipx, NE1 atom of residue W164^4.67 and carbon atom (C5) in the acetyl group of ACh or oxygen atom (O09) in the ether bond of Ipx, indole ring of residue W413^6.48 and acetyl group of ACh or heterocyclic isoazoline group of Ipx, OH atom of residue Y89^2.61 and oxygen atom in the amide group of VU154/LY298, benzene ring of residue F186^45.51 and aromatic core of the PAMs VU154/LY298, OH atom of residue Y439^7.39 and nitrogen atoms in the amine group of the PAMs VU154/LY298, CD atom of residue Q184^45.49 and nitrogen atom in the amide group of VU154/LY298, CG atom of residue N423^6.58 and chlorine atom in PAM LY298, OH atom of residue Y92^2.64 and nitrogen atom in the amide group of VU154, OG1 atom of residue T433^7.33 and sulfur atom in the trifluoromethylsulfonyl group of VU154. In addition, the χ₂ angle of residue W413^6.48 and W435^7.35 were calculated. Time courses of these reaction coordinates obtained from the GaMD simulation were plotted in the respective figures. The PyReweighting (Miao et al., 2014) toolkit was applied to reweight GaMD simulations to recover the original free energy or PMF profiles of the simulation systems. PMF profiles were computed using the combined trajectories from all the three independent 500 ns GaMD simulations for each system. A bin size of 1.0 Å was used for RMSD. The cutoff was set to 500 frames for 2D PMF calculations. The 2D PMF profiles were obtained for wildtype M₄R-Ipx-G_i1-LY298, M₄R-Ipx-G_i1-VU154, and the D432E and T433R single mutants of the M₄R-Ipx-G_i1-VU154 system regarding the RMSDs of the agonist Ipx and the RMSDs of the PAMs relative to the cryo-EM conformation.

Data analysis

All pharmacological data was fit using GraphPad Prism 9.2.0. Saturation binding experiments to determine B_max and pK_d values were determined as previously described (Leach et al., 2011; Nawaratne et al., 2010; Thal et al., 2016). Detailed equations and analysis details can be found in Appendix 1. Interaction inhibition binding curves between [³H]-NMS, agonists (ACh or Ipx), and PAMs (LY298 or VU154) were analyzed using the allosteric ternary complex model to calculate binding affinity values for each ligand (pK_A – for ACh/Ipx and pK_B for LY298/VU154) and the degree of binding modulation between agonist and PAM (log α) (Christopoulos and Kenakin, 2002). The pK_B values for LY298 and VU154 were determined from global fits of the ACh and Ipx curves to generate one pK_B value per ligand (Ehlert, 1988; Leach et al., 2011; Nawaratne et al., 2010; Thal et al., 2016). All pERK1/2 and TruPath assays were analyzed using the operational model allosterism and agonism to determine values of orthosteric (τ_A) or allosteric efficacy (τ_B) and the functional modulation (log αβ) between the agonists and PAMs (Leach et al., 2011; Nawaratne et al., 2010). Binding affinities of the agonists and the PAMs were fixed to values determined from equilibrium binding assays. The τ_B values for LY298 and VU154 were determined from global fits of the ACh and Ipx curves (when possible) to generate one value per ligand. For comparison between WT human M₄ mAChR and other M₄ mAChR constructs, the log τ values were corrected (denoted log τ_C) by normalizing to B_max values from saturation binding experiments (Leach et al., 2011; Nawaratne et al., 2010; Thal et al., 2016). All affinity, potency, and cooperativity values were estimated as logarithms, and statistical analysis between WT and mutant M₄ mAChR was determined by one-way ANOVA using a Dunnett’s post-hoc test with a value of p<0.05 considered as significant in this study.

Acknowledgements

This work was supported by a Wellcome Trust Collaborative Award (201529/Z/16/Z; PMS, ABT, AC), the National Health and Medical Research Council of Australia (1055134, 1150083, and 1138448), the Australian Research Council (DE170100152, DP190102950, and IC200100052), and the National Institutes of Health (GM132572). PMS is a Senior Principal Research Fellow (1154434), DW a Senior Research Fellow (1155302), DMT an Early Career Research Fellow (1196951), and KL a Future Fellow (160100075). RD was supported by Takeda Science Foundation 2019 Medical Research Grant and Japan Science and Technology Agency PRESTO (18069571). This work was partially supported by the Monash University Ramaciotti Centre for cryo-electron microscopy and the Monash University MASSIVE high-performance computing facility and supercomputing resources with the XSEDE allocation award TG-MCB180049, BIO220137 from the Advanced Cyberinfrastructure Coordination Ecosystem: Services & Support (ACCESS) program, and NERSC project M2874. We thank John Tesmer for discussion and the suggestion of calculating DAQ scores.

Appendix 1

Data analysis

Pharmacological parameters related to ligand binding

Interaction radioligand binding data were analyzed according to the following adapted form of an allosteric ternary complex model that accounts for the interaction of two orthosteric ligands and one allosteric ligand on a receptor (Christopoulos and Kenakin, 2002; Leach et al., 2007; Leach et al., 2011):

Y = \frac{B_{m a x} \cdot [A]}{[A] + [(K_{A} \cdot K_{B}) / (α_{A} [B] + K_{B})] \cdot {1 + ([I] / K_{I}) + ([B] / K_{B}) + [(α_{I} [I] [B] / K_{I} K_{B}])}}

[A], [B], and [I] represent the concentrations of the radioligand ([³H]-NMS), allosteric ligand, and orthosteric inhibitor, respectively.
K_A, K_B, and K_I represent their respective equilibrium dissociation constants. The value K_A was fixed to the value determined from saturation binding experiments.
B_max is the total number of receptors.
α_A and α_I represent the affinity cooperativity values between the allosteric ligand and the radioligand or orthosteric inhibitor, respectively. Values greater than 1 indicate positive cooperativity, values <1 (but >0) indicate negative cooperativity, and values of unity indicate neutral cooperativity.
All potency, affinity, and cooperativity parameters were estimated as logarithms.

Agonist binding affinity with a PAM bound:

Y = p K_{I} + l o g α_{I}

Δ Y = ((Δ p K_{I})^{2} + (Δ l o g α_{I})^{2})^{1 / 2}

PAM binding affinity for the agonist-bound state:

Y = p K_{B} + l o g α_{I}

Δ Y = ((Δ p K_{B})^{2} + (Δ l o g α_{I})^{2})^{1 / 2}

Global analysis for sharing the pK_B value between interaction experiments using two different agonists:

Define:

Ag1=agonist 1 (e.g. ACh)

Ag2=agonist 2 (e.g. Ipx)

alpha = α_A

betta = α_I

Data for agonist 1 in columns <A:F>

Data for agonist 2 in columns <G:L>

Prism equation:

K A = 10^{\land} L o g K A

K B = 10^{\land} L o g K B

I = 10^{\land} X

A = 10^{\land} L o g H o t A

K I A g 1 = 10^{\land} L o g K I A g 1 (K i f o r a g o n i s t 1)

K I A g 2 = 10^{\land} L o g K I A g 2 (K i f o r a g o n i s t 2)

a l p h a A g 1 = 10^{\land} L o g a l p h a A g 1

b e t t a A g 1 = 10^{\land} L o g b e t t a A g 1

a l p h a A g 2 = 10^{\land} L o g a l p h a A g 2

b e t t a A g 2 = 10^{\land} L o g b e t t a A g 2

P a r t 1 A g 1 = (K A^{*} K B) / (a l p h a A g 1^{*} B + K B)

P a r t 2 A g 1 = 1 + I / K I A g 1 + B / K B + (b e t t a A g 1^{*} I^{*} B) / (K I A g 1^{*} K B)

K A p p A g 1 = P a r t 1 A g 1^{*} P a r t 2 A g 1

P a r t 1 A g 2 = (K A^{*} K B) / (a l p h a A g 2^{*} B + K B)

P a r t 2 A g 2 = 1 + I / K I A g 2 + B / K B + (b e t t a A g 2^{*} I^{*} B) / (K I A g 2^{*} K B)

K A p p A g 2 = P a r t 1 A g 2^{*} P a r t 2 A g 2

< A : F > Y = (B m a x A g 1^{*} A) / (A + K A p p A g 1)

< G : L > Y = (B m a x A g 1^{*} A) / (A + K A p p A g 2)

Pharmacological parameters related to function

To determine efficacy values ( $τ_{A}$ ) of ACh and Ipx in TruPath and pERK1/2 assays, data were directly fit to the operational model of agonism (Black and Leff, 1983):

Y = B a s a l + \frac{E_{m} - B a s a l}{1 + \frac{K_{A} + [A]}{τ_{A} \times [A]}}

where $E_{m}$ is the maximal response of the system, Basal is the basal level of response in the absence of agonist, [A] is the concentration of agonist, K_A denotes the equilibrium dissociation constant of the agonist (A), and $τ_{A}$ is the index of the coupling efficiency/efficacy of the agonist. Values of K_A were constrained to the corresponding K_I values from interaction radioligand binding experiments.

The determination of $τ_{B}$ and $α β$ values for LY298 and VU154 in TruPath and pERK1/2 assays data was directly fit to the following operational model of allosterism and agonism (Leach et al., 2007):

Y = B a s a l + \frac{{(E_{m} - B a s a l) ([A] K_{B} + α β [A] [B] + τ_{B} [A] [B] {E C}_{50})}^{n}}{{([A] K_{B} + α β [A] [B] + τ_{B} [A] [B] {E C}_{50})}^{n} + {E C}_{50}^{n} {(K_{B} + [B])}^{n}}

where $E_{m}$ is the maximal response of the system, Basal is the basal level of response in the absence of agonist, [A] and [B] are the concentrations of agonist and the PAM, respectively. K_B denotes the equilibrium dissociation constant of the PAM. $τ_{B}$ is the index of the coupling efficiency/efficacy of the PAM. $α$ denotes the binding cooperativity between the agonist and PAM, whereas $β$ denotes a scaling factor that quantifies the allosteric effect of the PAM on the orthosteric ligand efficacy. This model assumes that the agonists (A) are full agonists. n is the transducer slope that describes the stimulus–response coupling of the ligand–receptor to the signaling pathway and was constrained to 1. The allosteric binding affinity ( $K_{B}$ ) was constrained to the value determined from radioligand binding experiments. All potency, affinity, efficacy, and cooperativity parameters were estimated as logarithms.

Transduction coupling coefficients were calculated as follows Kenakin et al., 2012:

Y = l o g \frac{τ_{A}}{K_{A}} = l o g τ_{A} - l o g K_{A}

Δ Y = ((Δ l o g τ_{A})^{2} + (Δ K_{A})^{2})^{1 / 2}

Efficacy modulation ( $β$ ) was calculated as follows:

Y = l o g β = l o g α β - l o g α

Δ Y = ((Δ l o g α β)^{2} + (Δ l o g α)^{2})^{1 / 2}

Global analysis for sharing $τ_{B}$ between interaction functional experiments using two different agonists:

Define:

Ag1=agonist 1 (e.g. ACh)

Ag2=agonist 2 (e.g. Ipx)

a b = α β

Data for agonist 1 in columns <A:F>

Data for agonist 2 in columns <G:L>

Prism equation:

K B = 10^{\land} L o g K B

t a u B = 10^{\land} L o g t a u B

A = 10^{\land} X

B = C o n c M o d

E C 50 A g 1 = 10^{\land} L o g E C 50 A g 1

E C 50 A g 2 = 10^{\land} L o g E C 50 A g 2

a b A g 1 = 10^{\land} L o g A l p h a B e t a A g 1

a b A g 2 = 10^{\land} L o g A l p h a B e t a A g 2

P a r t 1 A g 1 = (A^{*} (K B + a b A g 1^{*} B) + t a u B^{*} B^{*} E C 50 A g 1)^{\land} n

P a r t 1 A g 2 = (A^{*} (K B + a b A g 2^{*} B) + t a u B^{*} B^{*} E C 50 A g 2)^{\land} n

P a r t 2 A g 1 = (E C 50 A g 1^{\land} n)^{*} ((K B + B)^{\land} n)

P a r t 2 A g 2 = (E C 50 A g 2^{\land} n)^{*} ((K B + B)^{\land} n)

S p a n A g 1 = E m A g 1 ‐ B a s a l A g 1

S p a n A g 2 = E m A g 2 ‐ B a s a l A g 2

< A : F > Y = B a s a l A g 1 + (S p a n A g 1^{*} P a r t 1 A g 1) / (P a r t 1 A g 1 + P a r t 2 A g 1)

< G : L > Y = B a s a l A g 2 + (S p a n A g 2^{*} P a r t 1 A g 2) / (P a r t 1 A g 2 + P a r t 2 A g 2)

Correcting efficacy $τ_{X}$ values using receptor expression (Leach et al., 2011; Nawaratne et al., 2010):

B_max–WT = maximal number of receptors for the WT receptor

B_max–X = maximal number of receptors for the receptor construct X

$τ_{W T}$ = efficacy at the WT receptor

$τ_{X}$ = efficacy at receptor X

l o g τ_{X - c o r r} = l o g τ_{X} - (l o g B_{m a x - X} - l o g B_{m a x - W T})

Δ l o g τ_{X - c o r r} = {({(Δ l o g τ_{X})}^{2} + {(0.432 \times (\frac{Δ B_{m a x - X}}{B_{m a x - X}}))}^{2} + {(0.432 \times (\frac{Δ B_{m a x - W T}}{B_{m a x - W T}}))}^{2})}^{1 / 2}

Appendix 1—key resources table.

Reagent type (species) or resource	Designation	Source or reference	Identifiers	Additional information
Antibody	Anti-FLAG M1 (mouse polyclonal IgG2a)	Gift from Prof. Brian Kobilka (PMID::17962520)		Antibody was used to make anti-FLAG mAb resin that was used for the purification of FLAG-tagged M4 mAChR
Strain, strain background (Escherichia coli)	DH5α	New England Biolabs	C2987H
Strain, strain background (E. coli)	DH10bac	Thermo Fisher Scientific	10361012
Cell line (Spodoptera frugiperda)	Sf9	Expression Systems	94-001S
Cell line (Trichoplusia ni)	Tni	Expression Systems	94-002S
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR WT	PMID:26958838
Cell line (Chinese hamster ovary)	CHO K1 mouse M4 mAChR WT	PMID:21198541
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR D432E	This study		pEF5-FRT-V5-DEST plasmid
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR T433R	This study		pEF5-FRT-V5-DEST plasmid
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR V91L, D432E, T433R	This study		pEF5-FRT-V5-DEST plasmid
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR Y89A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR Q184A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR F186A	PMID:20406819
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR W435A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR W439A	PMID:20406819
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR S85A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR Y89A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR Y92A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR I93T, I94V, K95I	PMID:20406819
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR I93T	PMID:20406819
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR I94V	PMID:20406819
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR K95I	PMID:20406819
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR Y97A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR W98A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR G101A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR D106A	PMID:21300722
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR W108A	PMID:21300722
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR L109A	PMID:21300722
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR D112E	PMID:21300722
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR D112N	PMID:21300722
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR S116A	PMID:21300722
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR N117A	PMID:21300722
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR V120A	PMID:21300722
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR D129E	PMID:21300722
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR D129N	PMID:21300722
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR W164A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR F170A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR W171A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR Q172A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR F173A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR Q184A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR F186A	PMID:20406819
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR I187A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR Q188A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR F189A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR L190A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR W413A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR Y416A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR N423A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR Q427A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR S428P	PMID:20406819
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR D432N	PMID:20406819
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR W435A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR Y439A	PMID:20406819
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR Y440A	PMID:26958838
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR C442A	PMID:20406819
Cell line (Chinese hamster ovary)	Flp-In CHO human M4 mAChR Y443A	PMID:20406819
Cell line (Chinese hamster ovary)	Flp-In CHO cell line	Thermo Fisher Scientific	R75807
Cell line (Homo sapiens)	293A cell line	Thermo Fisher Scientific	R70507
Recombinant DNA reagent	Human FLAG-M4∆i3-His	This study		pVL1392 vector
Recombinant DNA reagent	Human Gαi1 dominant negative mutant	PMID:32193322		pFastBac vector
Recombinant DNA reagent	Human Gβ1γ2	Gift from Prof. Brian Kobilka (PMID:24256733)		pVL1392 vector
Recombinant DNA reagent	scFv16	PMID:30213947		pFastBac vector
Recombinant DNA reagent	pcDNA5/FRT/TO-GAlphai1-RLuc8	Gift from Prof. Bryan Roth (PMID:32367019)		TRUPATH assay
Recombinant DNA reagent	pcDNA3.1-Beta3	Gift from Prof. Bryan Roth (PMID:32367019)		TRUPATH assay
Recombinant DNA reagent	pcDNA3.1-GGamma9-GFP2	Gift from Prof. Bryan Roth (PMID:32367019)		TRUPATH assay
Chemical compound, drug	Acetylcholine	Sigma-Aldrich
Chemical compound, drug	Iperoxo	Sigma-Aldrich
Chemical compound, drug	LY2033298	Sigma-Aldrich
Chemical compound, drug	VU0467154	Gift from Prof. Craig Lindsley (PMID:25137629)		PMID:25137629
Chemical compound, drug	Prolume Purple	Nanolight Technology
Chemical compound, drug	[³H]-NMS	PerkinElmer
Chemical compound, drug	Polyethylenimine (PEI)	Sigma-Aldrich
Chemical compound, drug	Atropine	Sigma-Aldrich
Commercial assay or kit	AlphaScreen SureFire pERK 1/2 (Thr202/Tyr204) Assay Kits	PerkinElmer
Software, algorithm	Prism 8.0	GraphPad
Software, algorithm	PyMOL	Schrödinger
Software, algorithm	GaMD	PMID:26300708
Software, algorithm	AMBER20	https://ambermd.org
Software, algorithm	CPPTRAJ	PMID:26300708
Software, algorithm	PyReweighting	PMID:25061441
Software, algorithm	Phenix suite	PMID:20124702
Software, algorithm	Coot	PMID:31730249
Software, algorithm	Chimera	PMID:15264254
Software, algorithm	Chimera X	PMID:32881101
Software, algorithm	cryoSPARC	PMID:28165473
Software, algorithm	Relion 3.1	PMID:30412051
Software, algorithm	Motioncor2	PMID:28250466
Software, algorithm	GCTF	PMID:26592709
Software, algorithm	crYOLO	PMID:31240256
Software, algorithm	ISOLDE	PMID:29872003
Software, algorithm	MolProbity	PMID:29067766
Software, algorithm	DAQ score	PMID:35953671

Open in a new tab

Funding Statement

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication. For the purpose of Open Access, the authors have applied a CC BY public copyright license to any Author Accepted Manuscript version arising from this submission.

Contributor Information

Celine Valant, Email: celine.valant@monash.edu.

Yinglong Miao, Email: miao@ku.edu.

Arthur Christopoulos, Email: Arthur.Christopoulos@monash.edu.

David M Thal, Email: david.thal@monash.edu.

Sudarshan Rajagopal, Duke University Medical Center, United States.

Kenton J Swartz, National Institute of Neurological Disorders and Stroke, National Institutes of Health, United States.

Funding Information

This paper was supported by the following grants:

Wellcome Trust 201529/Z/16/Z to Andrew Tobin, Patrick Sexton, Arthur Christopoulos.
National Health and Medical Research Council 1055134 to Arthur Christopoulos.
National Health and Medical Research Council 1150083 to Arthur Christopoulos.
National Health and Medical Research Council 1138448 to David M Thal.
Australian Research Council DE170100152 to David M Thal.
Australian Research Council DP190102950 to Celine Valant.
Australian Research Council IC200100052 to Denise Wootten.
National Institutes of Health GM132572 to Yinglong Miao.
National Health and Medical Research Council 1154434 to Patrick Sexton.
National Health and Medical Research Council 1155302 to Denise Wootten.
National Health and Medical Research Council 1196951 to David M Thal.
Australian Research Council 160100075 to Katie Leach.
Takeda Science Foundation 2019 Medical Research Grant to Radostin Danev.
Japan Science and Technology Agency 18069571 to Radostin Danev.

Additional information

Competing interests

No competing interests declared.

P.M.S, D.W., and A.C. are shareholders of Septerna Inc.

Author contributions

Investigation, Methodology, Writing – review and editing.

Software, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing – review and editing.

Formal analysis, Investigation, Writing – review and editing.

Validation, Investigation, Visualization, Writing – review and editing.

Resources, Formal analysis.

Investigation, Methodology, Writing – review and editing.

Formal analysis, Investigation, Writing – review and editing.

Investigation.

Writing – review and editing.

Investigation, Writing – review and editing.

Investigation.

Investigation, Writing – review and editing.

Resources, Supervision, Project administration, Writing – review and editing.

Resources, Writing – review and editing.

Conceptualization, Resources, Funding acquisition, Project administration, Writing – review and editing.

Conceptualization, Resources, Supervision, Funding acquisition, Project administration, Writing – review and editing.

Resources, Funding acquisition, Investigation, Writing – review and editing.

Conceptualization, Formal analysis, Supervision, Validation, Visualization, Writing – original draft, Project administration, Writing – review and editing.

Conceptualization, Resources, Software, Formal analysis, Supervision, Funding acquisition, Validation, Visualization, Methodology, Writing – original draft, Project administration, Writing – review and editing.

Conceptualization, Resources, Supervision, Funding acquisition, Writing – original draft, Project administration, Writing – review and editing.

Conceptualization, Resources, Software, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing – original draft, Project administration, Writing – review and editing.

Additional files

MDAR checklist

elife-83477-mdarchecklist1.pdf^{(195.9KB, pdf)}

Data availability

All data generated or analysed during this study are included in the manuscript. Structural data has been deposited in the Protein Data Bank (PDB) and Electron Microscopy Data Bank (EMDB) under the following codes:(1) M4R-Gi1-Ipx PDB: 7TRK and EMD-26099 (2) M4R-Gi1-Ipx-LY298 PDB: 7TRP and EMD-26100 (3) M4R-Gi1-Ipx-VU154 PDB: 7TRQ and EMD-26101 (4) M4R-Gi1-ACh PDB: 7TRS and EMD-26102.

The following datasets were generated:

Vuckovic Z, Mobbs JI, Belousoff MJ, Glukhova A, Sexton PM, Danev R, Thal DM. 2023. Human M4 muscarinic acetylcholine receptor complex with Gi1 and the agonist iperoxo. RCSB Protein Data Bank. 7TRK

Vuckovic Z, Mobbs JI, Belousoff MJ, Glukhova A, Sexton PM, Danev R, Thal DM. 2023. Human M4 muscarinic acetylcholine receptor complex with Gi1 and the agonist iperoxo and positive allosteric modulator LY2033298. RCSB Protein Data Bank. 7TRP

Vuckovic Z, Mobbs JI, Belousoff MJ, Glukhova A, Sexton PM, Danev R, Thal DM. 2023. Human M4 muscarinic acetylcholine receptor complex with Gi1 and the endogenous agonist acetylcholine. RCSB Protein Data Bank. 7TRS

Vuckovic Z, Mobbs JI, Belousoff MJ, Glukhova A, Sexton PM, Danev R, Thal DM. 2023. Human M4 muscarinic acetylcholine receptor complex with Gi1 and the agonist iperoxo. Electron Microscopy Data Bank. EMD-26099

Vuckovic Z, Mobbs JI, Belousoff MJ, Glukhova A, Sexton PM, Danev R, Thanl DM. 2023. Human M4 muscarinic acetylcholine receptor complex with Gi1 and the agonist iperoxo and positive allosteric modulator VU0467154. Electron Microscopy Data Bank. EMD-26101

References

Ballesteros JA, Weinstein H. In: Methods in Neurosciences, Receptor Molecular Biology. Sealfon SC, editor. Academic Press; 1995. Integrated methods for the construction of three-dimensional models and computational probing of structure-function relations in G protein-coupled receptors; pp. 366–428. [DOI] [Google Scholar]
Ballesteros JA, Jensen AD, Liapakis G, Rasmussen SG, Shi L, Gether U, Javitch JA. Activation of the beta 2-adrenergic receptor involves disruption of an ionic lock between the cytoplasmic ends of transmembrane segments 3 and 6. The Journal of Biological Chemistry. 2001;276:29171–29177. doi: 10.1074/jbc.M103747200. [DOI] [PubMed] [Google Scholar]
Berizzi AE, Gentry PR, Rueda P, Den Hoedt S, Sexton PM, Langmead CJ, Christopoulos A. Molecular mechanisms of action of M5 muscarinic acetylcholine receptor allosteric modulators. Molecular Pharmacology. 2016;90:427–436. doi: 10.1124/mol.116.104182. [DOI] [PubMed] [Google Scholar]
Berman H, Henrick K, Nakamura H. Announcing the worldwide protein data bank. Nature Structural Biology. 2003;10:980. doi: 10.1038/nsb1203-980. [DOI] [PubMed] [Google Scholar]
Bhattarai A, Miao Y. Gaussian accelerated molecular dynamics for elucidation of drug pathways. Expert Opinion on Drug Discovery. 2018;13:1055–1065. doi: 10.1080/17460441.2018.1538207. [DOI] [PMC free article] [PubMed] [Google Scholar]
Black JW, Leff P. Operational models of pharmacological agonism. Proceedings of the Royal Society of London. Series B, Biological Sciences. 1983;220:141–162. doi: 10.1098/rspb.1983.0093. [DOI] [PubMed] [Google Scholar]
Brown AJH, Bradley SJ, Marshall FH, Brown GA, Bennett KA, Brown J, Cansfield JE, Cross DM, de Graaf C, Hudson BD, Dwomoh L, Dias JM, Errey JC, Hurrell E, Liptrot J, Mattedi G, Molloy C, Nathan PJ, Okrasa K, Osborne G, Patel JC, Pickworth M, Robertson N, Shahabi S, Bundgaard C, Phillips K, Broad LM, Goonawardena AV, Morairty SR, Browning M, Perini F, Dawson GR, Deakin JFW, Smith RT, Sexton PM, Warneck J, Vinson M, Tasker T, Tehan BG, Teobald B, Christopoulos A, Langmead CJ, Jazayeri A, Cooke RM, Rucktooa P, Congreve MS, Weir M, Tobin AB. From structure to clinic: design of a muscarinic M1 receptor agonist with the potential to treat Alzheimer’s disease. Cell. 2021;184:5886–5901. doi: 10.1016/j.cell.2021.11.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
Bubser M, Bridges TM, Dencker D, Gould RW, Grannan M, Noetzel MJ, Lamsal A, Niswender CM, Daniels JS, Poslusney MS, Melancon BJ, Tarr JC, Byers FW, Wess J, Duggan ME, Dunlop J, Wood MW, Brandon NJ, Wood MR, Lindsley CW, Conn PJ, Jones CK. Selective activation of M4 muscarinic acetylcholine receptors reverses MK-801-induced behavioral impairments and enhances associative learning in rodents. ACS Chemical Neuroscience. 2014;5:920–942. doi: 10.1021/cn500128b. [DOI] [PMC free article] [PubMed] [Google Scholar]
Bumbak F, Thomas T, Noonan-Williams BJ, Vaid TM, Yan F, Whitehead AR, Bruell S, Kocan M, Tan X, Johnson MA, Bathgate RAD, Chalmers DK, Gooley PR, Scott DJ. Conformational changes in tyrosine 11 of neurotensin are required to activate the neurotensin receptor 1. ACS Pharmacology & Translational Science. 2020;3:690–705. doi: 10.1021/acsptsci.0c00026. [DOI] [PMC free article] [PubMed] [Google Scholar]
Burger WAC, Sexton PM, Christopoulos A, Thal DM. Toward an understanding of the structural basis of allostery in muscarinic acetylcholine receptors. The Journal of General Physiology. 2018;150:1360–1372. doi: 10.1085/jgp.201711979. [DOI] [PMC free article] [PubMed] [Google Scholar]
Bymaster FP, Carter PA, Yamada M, Gomeza J, Wess J, Hamilton SE, Nathanson NM, McKinzie DL, Felder CC. Role of specific muscarinic receptor subtypes in cholinergic parasympathomimetic responses, in vivo phosphoinositide hydrolysis, and pilocarpine-induced seizure activity. The European Journal of Neuroscience. 2003;17:1403–1410. doi: 10.1046/j.1460-9568.2003.02588.x. [DOI] [PubMed] [Google Scholar]
Canals M, Sexton PM, Christopoulos A. Allostery in GPCRs: `` MWC'' revisited. Trends in Biochemical Sciences. 2011;36:663–672. doi: 10.1016/j.tibs.2011.08.005. [DOI] [PubMed] [Google Scholar]
Cao AM, Quast RB, Fatemi F, Rondard P, Pin JP, Margeat E. Allosteric modulators enhance agonist efficacy by increasing the residence time of a GPCR in the active state. Nature Communications. 2021;12:5426. doi: 10.1038/s41467-021-25620-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
Cary BP, Deganutti G, Zhao P, Truong TT, Piper SJ, Liu X, Belousoff MJ, Danev R, Sexton PM, Wootten D, Gellman SH. Structural and functional diversity among agonist-bound states of the GLP-1 receptor. Nature Chemical Biology. 2022;18:256–263. doi: 10.1038/s41589-021-00945-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
Casañal A, Lohkamp B, Emsley P. Current developments in coot for macromolecular model building of electron cryo-microscopy and crystallographic data. Protein Science. 2020;29:1069–1078. doi: 10.1002/pro.3791. [DOI] [PMC free article] [PubMed] [Google Scholar]
Chan WY, McKinzie DL, Bose S, Mitchell SN, Witkin JM, Thompson RC, Christopoulos A, Lazareno S, Birdsall NJM, Bymaster FP, Felder CC. Allosteric modulation of the muscarinic M4 receptor as an approach to treating schizophrenia. PNAS. 2008;105:10978–10983. doi: 10.1073/pnas.0800567105. [DOI] [PMC free article] [PubMed] [Google Scholar]
Changeux JP, Christopoulos A. Allosteric modulation as a unifying mechanism for receptor function and regulation. Cell. 2016;166:1084–1102. doi: 10.1016/j.cell.2016.08.015. [DOI] [PubMed] [Google Scholar]
Christopoulos A, Kenakin T. G protein-coupled receptor allosterism and complexing. Pharmacological Reviews. 2002;54:323–374. doi: 10.1124/pr.54.2.323. [DOI] [PubMed] [Google Scholar]
Christopoulos A. Advances in G protein-coupled receptor allostery: from function to structure. Molecular Pharmacology. 2014;86:463–478. doi: 10.1124/mol.114.094342. [DOI] [PubMed] [Google Scholar]
Cong X, Maurel D, Déméné H, Vasiliauskaité-Brooks I, Hagelberger J, Peysson F, Saint-Paul J, Golebiowski J, Granier S, Sounier R. Molecular insights into the biased signaling mechanism of the μ-opioid receptor. Molecular Cell. 2021;81:4165–4175. doi: 10.1016/j.molcel.2021.07.033. [DOI] [PMC free article] [PubMed] [Google Scholar]
Conn PJ, Jones CK, Lindsley CW. Subtype-Selective allosteric modulators of muscarinic receptors for the treatment of CNS disorders. Trends in Pharmacological Sciences. 2009;30:148–155. doi: 10.1016/j.tips.2008.12.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
Croll TI. ISOLDE: a physically realistic environment for model building into low-resolution electron-density maps. Acta Crystallographica. Section D, Structural Biology. 2018;74:519–530. doi: 10.1107/S2059798318002425. [DOI] [PMC free article] [PubMed] [Google Scholar]
Danev R, Belousoff M, Liang YL, Zhang X, Eisenstein F, Wootten D, Sexton PM. Routine sub-2.5 Å cryo-EM structure determination of GPCRs. Nature Communications. 2021;12:4333. doi: 10.1038/s41467-021-24650-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
Darden T, York D, Pedersen L. Particle mesh ewald: an N⋅log (N) method for ewald sums in large systems. The Journal of Chemical Physics. 1993;98:10089–10092. doi: 10.1063/1.464397. [DOI] [Google Scholar]
Deganutti G, Liang YL, Zhang X, Khoshouei M, Clydesdale L, Belousoff MJ, Venugopal H, Truong TT, Glukhova A, Keller AN, Gregory KJ, Leach K, Christopoulos A, Danev R, Reynolds CA, Zhao P, Sexton PM, Wootten D. Dynamics of GLP-1R peptide agonist engagement are correlated with kinetics of G protein activation. Nature Communications. 2022;13:92. doi: 10.1038/s41467-021-27760-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
Dencker D, Wörtwein G, Weikop P, Jeon J, Thomsen M, Sager TN, Mørk A, Woldbye DPD, Wess J, Fink-Jensen A. Involvement of a subpopulation of neuronal M4 muscarinic acetylcholine receptors in the antipsychotic-like effects of the M1/M4 preferring muscarinic receptor agonist xanomeline. The Journal of Neuroscience. 2011;31:5905–5908. doi: 10.1523/JNEUROSCI.0370-11.2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
DeVree BT, Mahoney JP, Vélez-Ruiz GA, Rasmussen SGF, Kuszak AJ, Edwald E, Fung JJ, Manglik A, Masureel M, Du Y, Matt RA, Pardon E, Steyaert J, Kobilka BK, Sunahara RK. Allosteric coupling from G protein to the agonist-binding pocket in GPCRs. Nature. 2016;535:182–186. doi: 10.1038/nature18324. [DOI] [PMC free article] [PubMed] [Google Scholar]
Draper-Joyce CJ, Bhola R, Wang J, Bhattarai A, Nguyen ATN, Cowie-Kent I, O’Sullivan K, Chia LY, Venugopal H, Valant C, Thal DM, Wootten D, Panel N, Carlsson J, Christie MJ, White PJ, Scammells P, May LT, Sexton PM, Danev R, Miao Y, Glukhova A, Imlach WL, Christopoulos A. Positive allosteric mechanisms of adenosine A1 receptor-mediated analgesia. Nature. 2021;597:571–576. doi: 10.1038/s41586-021-03897-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
Dror RO, Arlow DH, Maragakis P, Mildorf TJ, Pan AC, Xu H, Borhani DW, Shaw DE. Activation mechanism of the β2-adrenergic receptor. PNAS. 2011;108:18684–18689. doi: 10.1073/pnas.1110499108. [DOI] [PMC free article] [PubMed] [Google Scholar]
Dror RO, Mildorf TJ, Hilger D, Manglik A, Borhani DW, Arlow DH, Philippsen A, Villanueva N, Yang Z, Lerch MT, Hubbell WL, Kobilka BK, Sunahara RK, Shaw DE. Signal transduction. Structural basis for nucleotide exchange in heterotrimeric G proteins. Science. 2015;348:1361–1365. doi: 10.1126/science.aaa5264. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ehlert FJ. Estimation of the affinities of allosteric ligands using radioligand binding and pharmacological null methods. Molecular Pharmacology. 1988;33:187–194. [PubMed] [Google Scholar]
Foster DJ, Wilson JM, Remke DH, Mahmood MS, Uddin MJ, Wess J, Patel S, Marnett LJ, Niswender CM, Jones CK, Xiang Z, Lindsley CW, Rook JM, Conn PJ. Antipsychotic-Like effects of M 4 positive allosteric modulators are mediated by CB 2 receptor-dependent inhibition of dopamine release. Neuron. 2016;91:1244–1252. doi: 10.1016/j.neuron.2016.08.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
Fritze O, Filipek S, Kuksa V, Palczewski K, Hofmann KP, Ernst OP. Role of the conserved NPXXY (X) 5,6F motif in the rhodopsin ground state and during activation. PNAS. 2003;100:2290–2295. doi: 10.1073/pnas.0435715100. [DOI] [PMC free article] [PubMed] [Google Scholar]
Furness SGB, Liang YL, Nowell CJ, Halls ML, Wookey PJ, Dal Maso E, Inoue A, Christopoulos A, Wootten D, Sexton PM. Ligand-Dependent modulation of G protein conformation alters drug efficacy. Cell. 2016;167:739–749. doi: 10.1016/j.cell.2016.09.021. [DOI] [PubMed] [Google Scholar]
Gentry PR, Kokubo M, Bridges TM, Noetzel MJ, Cho HP, Lamsal A, Smith E, Chase P, Hodder PS, Niswender CM, Daniels JS, Conn PJ, Lindsley CW, Wood MR. Development of a highly potent, novel M5 positive allosteric modulator (PAM) demonstrating CNS exposure: 1- ((1H-indazol-5-yl) sulfoneyl) -N-ethyl-N- (2- (trifluoromethyl) benzyl) piperidine-4-carboxamide (ML380) Journal of Medicinal Chemistry. 2014;57:7804–7810. doi: 10.1021/jm500995y. [DOI] [PMC free article] [PubMed] [Google Scholar]
Gentry PR, Sexton PM, Christopoulos A. Novel allosteric modulators of G protein-coupled receptors. The Journal of Biological Chemistry. 2015;290:19478–19488. doi: 10.1074/jbc.R115.662759. [DOI] [PMC free article] [PubMed] [Google Scholar]
Gould RW, Nedelcovych MT, Gong X, Tsai E, Bubser M, Bridges TM, Wood MR, Duggan ME, Brandon NJ, Dunlop J, Wood MW, Ivarsson M, Noetzel MJ, Daniels JS, Niswender CM, Lindsley CW, Conn PJ, Jones CK. State-Dependent alterations in sleep/wake architecture elicited by the M4 PAM VU0467154-relation to antipsychotic-like drug effects. Neuropharmacology. 2016;102:244–253. doi: 10.1016/j.neuropharm.2015.11.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
Gregorio GG, Masureel M, Hilger D, Terry DS, Juette M, Zhao H, Zhou Z, Perez-Aguilar JM, Hauge M, Mathiasen S, Javitch JA, Weinstein H, Kobilka BK, Blanchard SC. Single-Molecule analysis of ligand efficacy in β2ar-G-protein activation. Nature. 2017;547:68–73. doi: 10.1038/nature22354. [DOI] [PMC free article] [PubMed] [Google Scholar]
Gregory KJ, Hall NE, Tobin AB, Sexton PM, Christopoulos A. Identification of orthosteric and allosteric site mutations in M2 muscarinic acetylcholine receptors that contribute to ligand-selective signaling bias. The Journal of Biological Chemistry. 2010;285:7459–7474. doi: 10.1074/jbc.M109.094011. [DOI] [PMC free article] [PubMed] [Google Scholar]
Haga K, Kruse AC, Asada H, Yurugi-Kobayashi T, Shiroishi M, Zhang C, Weis WI, Okada T, Kobilka BK, Haga T, Kobayashi T. Structure of the human M2 muscarinic acetylcholine receptor bound to an antagonist. Nature. 2012;482:547–551. doi: 10.1038/nature10753. [DOI] [PMC free article] [PubMed] [Google Scholar]
Huang J, Rauscher S, Nawrocki G, Ran T, Feig M, de Groot BL, Grubmüller H, MacKerell AD., Jr CHARMM36m: an improved force field for folded and intrinsically disordered proteins. Nature Methods. 2017;14:71–73. doi: 10.1038/nmeth.4067. [DOI] [PMC free article] [PubMed] [Google Scholar]
Huang SK, Pandey A, Tran DP, Villanueva NL, Kitao A, Sunahara RK, Sljoka A, Prosser RS. Delineating the conformational landscape of the adenosine A2A receptor during G protein coupling. Cell. 2021;184:1884–1894. doi: 10.1016/j.cell.2021.02.041. [DOI] [PMC free article] [PubMed] [Google Scholar]
Hulme EC. Gpcr activation: a mutagenic spotlight on crystal structures. Trends in Pharmacological Sciences. 2013;34:67–84. doi: 10.1016/j.tips.2012.11.002. [DOI] [PubMed] [Google Scholar]
Humphrey W, Dalke A, Schulten K. VMD: visual molecular dynamics. Journal of Molecular Graphics. 1996;14:33–38. doi: 10.1016/0263-7855(96)00018-5. [DOI] [PubMed] [Google Scholar]
Jones EM, Lubock NB, Venkatakrishnan AJ, Wang J, Tseng AM, Paggi JM, Latorraca NR, Cancilla D, Satyadi M, Davis JE, Babu MM, Dror RO, Kosuri S. Structural and functional characterization of G protein-coupled receptors with deep mutational scanning. eLife. 2020;9:e54895. doi: 10.7554/eLife.54895. [DOI] [PMC free article] [PubMed] [Google Scholar]
Josephs TM, Belousoff MJ, Liang YL, Piper SJ, Cao J, Garama DJ, Leach K, Gregory KJ, Christopoulos A, Hay DL, Danev R, Wootten D, Sexton PM. Structure and dynamics of the CGRP receptor in apo and peptide-bound forms. Science. 2021;372:eabf7258. doi: 10.1126/science.abf7258. [DOI] [PubMed] [Google Scholar]
Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, Tunyasuvunakool K, Bates R, Žídek A, Potapenko A, Bridgland A, Meyer C, Kohl SAA, Ballard AJ, Cowie A, Romera-Paredes B, Nikolov S, Jain R, Adler J, Back T, Petersen S, Reiman D, Clancy E, Zielinski M, Steinegger M, Pacholska M, Berghammer T, Bodenstein S, Silver D, Vinyals O, Senior AW, Kavukcuoglu K, Kohli P, Hassabis D. Highly accurate protein structure prediction with alphafold. Nature. 2021;596:583–589. doi: 10.1038/s41586-021-03819-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
Katayama K, Suzuki K, Suno R, Kise R, Tsujimoto H, Iwata S, Inoue A, Kobayashi T, Kandori H. Vibrational spectroscopy analysis of ligand efficacy in human M2 muscarinic acetylcholine receptor (M2R) Communications Biology. 2021;4:1321. doi: 10.1038/s42003-021-02836-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kenakin T, Onaran O. The ligand paradox between affinity and efficacy: can you be there and not make a difference? Trends in Pharmacological Sciences. 2002;23:275–280. doi: 10.1016/S0165-6147(02)02036-9. [DOI] [PubMed] [Google Scholar]
Kenakin T. New concepts in drug discovery: collateral efficacy and permissive antagonism. Nature Reviews. Drug Discovery. 2005;4:919–927. doi: 10.1038/nrd1875. [DOI] [PubMed] [Google Scholar]
Kenakin TP. Biased signalling and allosteric machines: new vistas and challenges for drug discovery. British Journal of Pharmacology. 2012;165:1659–1669. doi: 10.1111/j.1476-5381.2011.01749.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kenakin TP, Watson C, Muniz-Medina V, Christopoulos A, Novick S. A simple method for quantifying functional selectivity and agonist bias. ACS Chemical Neuroscience. 2012;3:193–203. doi: 10.1021/cn200111m. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kenakin T, Christopoulos A. Signalling bias in new drug discovery: detection, quantification and therapeutic impact. Nature Reviews. Drug Discovery. 2013;12:205–216. doi: 10.1038/nrd3954. [DOI] [PubMed] [Google Scholar]
Klauda JB, Venable RM, Freites JA, O’Connor JW, Tobias DJ, Mondragon-Ramirez C, Vorobyov I, MacKerell AD, Jr, Pastor RW. Update of the CHARMM all-atom additive force field for lipids: validation on six lipid types. The Journal of Physical Chemistry. B. 2010;114:7830–7843. doi: 10.1021/jp101759q. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kruse AC, Hu J, Pan AC, Arlow DH, Rosenbaum DM, Rosemond E, Green HF, Liu T, Chae PS, Dror RO, Shaw DE, Weis WI, Wess J, Kobilka BK. Structure and dynamics of the M3 muscarinic acetylcholine receptor. Nature. 2012;482:552–556. doi: 10.1038/nature10867. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kruse AC, Ring AM, Manglik A, Hu J, Hu K, Eitel K, Hübner H, Pardon E, Valant C, Sexton PM, Christopoulos A, Felder CC, Gmeiner P, Steyaert J, Weis WI, Garcia KC, Wess J, Kobilka BK. Activation and allosteric modulation of a muscarinic acetylcholine receptor. Nature. 2013;504:101–106. doi: 10.1038/nature12735. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kumar A, Yoluk O, MacKerell AD. FFParam: standalone package for CHARMM additive and drude polarizable force field parametrization of small molecules. Journal of Computational Chemistry. 2020;41:958–970. doi: 10.1002/jcc.26138. [DOI] [PMC free article] [PubMed] [Google Scholar]
Leach K, Sexton PM, Christopoulos A. Allosteric GPCR modulators: taking advantage of permissive receptor pharmacology. Trends in Pharmacological Sciences. 2007;28:382–389. doi: 10.1016/j.tips.2007.06.004. [DOI] [PubMed] [Google Scholar]
Leach K, Loiacono RE, Felder CC, McKinzie DL, Mogg A, Shaw DB, Sexton PM, Christopoulos A. Molecular mechanisms of action and in vivo validation of an M4 muscarinic acetylcholine receptor allosteric modulator with potential antipsychotic properties. Neuropsychopharmacology. 2010;35:855–869. doi: 10.1038/npp.2009.194. [DOI] [PMC free article] [PubMed] [Google Scholar]
Leach K, Davey AE, Felder CC, Sexton PM, Christopoulos A. The role of transmembrane domain 3 in the actions of orthosteric, allosteric, and atypical agonists of the M4 muscarinic acetylcholine receptor. Molecular Pharmacology. 2011;79:855–865. doi: 10.1124/mol.111.070938. [DOI] [PubMed] [Google Scholar]
Liang YL, Khoshouei M, Glukhova A, Furness SGB, Zhao P, Clydesdale L, Koole C, Truong TT, Thal DM, Lei S, Radjainia M, Danev R, Baumeister W, Wang MW, Miller LJ, Christopoulos A, Sexton PM, Wootten D. Phase-plate cryo-EM structure of a biased agonist-bound human GLP-1 receptor–gs complex. Nature. 2018a;555:121–125. doi: 10.1038/nature25773. [DOI] [PubMed] [Google Scholar]
Liang YL, Zhao P, Draper-Joyce C, Baltos JA, Glukhova A, Truong TT, May LT, Christopoulos A, Wootten D, Sexton PM, Furness SGB. Dominant negative G proteins enhance formation and purification of agonist-GPCR-G protein complexes for structure determination. ACS Pharmacology & Translational Science. 2018b;1:12–20. doi: 10.1021/acsptsci.8b00017. [DOI] [PMC free article] [PubMed] [Google Scholar]
Liang YL, Belousoff MJ, Fletcher MM, Zhang X, Khoshouei M, Deganutti G, Koole C, Furness SGB, Miller LJ, Hay DL, Christopoulos A, Reynolds CA, Danev R, Wootten D, Sexton PM. Structure and dynamics of adrenomedullin receptors AM1 and AM2 reveal key mechanisms in the control of receptor phenotype by receptor activity-modifying proteins. ACS Pharmacology & Translational Science. 2020;3:263–284. doi: 10.1021/acsptsci.9b00080. [DOI] [PMC free article] [PubMed] [Google Scholar]
Liebschner D, Afonine PV, Baker ML, Bunkóczi G, Chen VB, Croll TI, Hintze B, Hung LW, Jain S, McCoy AJ, Moriarty NW, Oeffner RD, Poon BK, Prisant MG, Read RJ, Richardson JS, Richardson DC, Sammito MD, Sobolev OV, Stockwell DH, Terwilliger TC, Urzhumtsev AG, Videau LL, Williams CJ, Adams PD. Macromolecular structure determination using x-rays, neutrons and electrons: recent developments in phenix. Acta Crystallographica Section D Structural Biology. 2019;75:861–877. doi: 10.1107/S2059798319011471. [DOI] [PMC free article] [PubMed] [Google Scholar]
Liu W, Chun E, Thompson AA, Chubukov P, Xu F, Katritch V, Han GW, Roth CB, Heitman LH, IJzerman AP, Cherezov V, Stevens RC. Structural basis for allosteric regulation of GPCRs by sodium ions. Science. 2012a;337:232–236. doi: 10.1126/science.1219218. [DOI] [PMC free article] [PubMed] [Google Scholar]
Liu JJ, Horst R, Katritch V, Stevens RC, Wüthrich K. Biased signaling pathways in β2-adrenergic receptor characterized by 19F-NMR. Science. 2012b;335:1106–1110. doi: 10.1126/science.1215802. [DOI] [PMC free article] [PubMed] [Google Scholar]
Maeda S, Koehl A, Matile H, Hu H, Hilger D, Schertler GFX, Manglik A, Skiniotis G, Dawson RJP, Kobilka BK. Development of an antibody fragment that stabilizes GPCR/G-protein complexes. Nature Communications. 2018;9:3712. doi: 10.1038/s41467-018-06002-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
Maeda S, Qu Q, Robertson MJ, Skiniotis G, Kobilka BK. Structures of the M1 and M2 muscarinic acetylcholine receptor/G-protein complexes. Science. 2019;364:552–557. doi: 10.1126/science.aaw5188. [DOI] [PMC free article] [PubMed] [Google Scholar]
Manglik A, Kim TH, Masureel M, Altenbach C, Yang Z, Hilger D, Lerch MT, Kobilka TS, Thian FS, Hubbell WL, Prosser RS, Kobilka BK. Structural insights into the dynamic process of β2-adrenergic receptor signaling. Cell. 2015;161:1101–1111. doi: 10.1016/j.cell.2015.04.043. [DOI] [PMC free article] [PubMed] [Google Scholar]
Marlo JE, Niswender CM, Days EL, Bridges TM, Xiang Y, Rodriguez AL, Shirey JK, Brady AE, Nalywajko T, Luo Q, Austin CA, Williams MB, Kim K, Williams R, Orton D, Brown HA, Lindsley CW, Weaver CD, Conn PJ. Discovery and characterization of novel allosteric potentiators of M1 muscarinic receptors reveals multiple modes of activity. Molecular Pharmacology. 2009;75:577–588. doi: 10.1124/mol.108.052886. [DOI] [PMC free article] [PubMed] [Google Scholar]
Masureel M, Zou Y, Picard L-P, van der Westhuizen E, Mahoney JP, Rodrigues JPGLM, Mildorf TJ, Dror RO, Shaw DE, Bouvier M, Pardon E, Steyaert J, Sunahara RK, Weis WI, Zhang C, Kobilka BK. Structural insights into binding specificity, efficacy and bias of a β2AR partial agonist. Nature Chemical Biology. 2018;14:1059–1066. doi: 10.1038/s41589-018-0145-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
McCorvy JD, Wacker D, Wang S, Agegnehu B, Liu J, Lansu K, Tribo AR, Olsen RHJ, Che T, Jin J, Roth BL. Structural determinants of 5-HT2B receptor activation and biased agonism. Nature Structural & Molecular Biology. 2018;25:787–796. doi: 10.1038/s41594-018-0116-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
Miao Y, Sinko W, Pierce L, Bucher D, Walker RC, McCammon JA. Improved reweighting of accelerated molecular dynamics simulations for free energy calculation. Journal of Chemical Theory and Computation. 2014;10:2677–2689. doi: 10.1021/ct500090q. [DOI] [PMC free article] [PubMed] [Google Scholar]
Miao Y, Feher VA, McCammon JA. Gaussian accelerated molecular dynamics: unconstrained enhanced sampling and free energy calculation. Journal of Chemical Theory and Computation. 2015;11:3584–3595. doi: 10.1021/acs.jctc.5b00436. [DOI] [PMC free article] [PubMed] [Google Scholar]
Miao Y, McCammon JA. Graded activation and free energy landscapes of a muscarinic G-protein–coupled receptor. PNAS. 2016;113:12162–12167. doi: 10.1073/pnas.1614538113. [DOI] [PMC free article] [PubMed] [Google Scholar]
Miao Y, McCammon JA. Gaussian accelerated molecular dynamics: theory, implementation, and applications. Annual Reports in Computational Chemistry. 2017;13:231–278. doi: 10.1016/bs.arcc.2017.06.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
Miao Y, McCammon JA. Mechanism of the G-protein mimetic nanobody binding to a muscarinic G-protein-coupled receptor. PNAS. 2018;115:3036–3041. doi: 10.1073/pnas.1800756115. [DOI] [PMC free article] [PubMed] [Google Scholar]
Mobbs JI, Belousoff MJ, Harikumar KG, Piper SJ, Xu X, Furness SGB, Venugopal H, Christopoulos A, Danev R, Wootten D, Thal DM, Miller LJ, Sexton PM. Structures of the human cholecystokinin 1 (CCK1) receptor bound to Gs and Gq mimetic proteins provide insight into mechanisms of G protein selectivity. PLOS Biology. 2021;19:e3001295. doi: 10.1371/journal.pbio.3001295. [DOI] [PMC free article] [PubMed] [Google Scholar]
Nawaratne V, Leach K, Felder CC, Sexton PM, Christopoulos A. Structural determinants of allosteric agonism and modulation at the M4 muscarinic acetylcholine receptor: identification of ligand-specific and global activation mechanisms. The Journal of Biological Chemistry. 2010;285:19012–19021. doi: 10.1074/jbc.M110.125096. [DOI] [PMC free article] [PubMed] [Google Scholar]
Nygaard R, Frimurer TM, Holst B, Rosenkilde MM, Schwartz TW. Ligand binding and micro-switches in 7TM receptor structures. Trends in Pharmacological Sciences. 2009;30:249–259. doi: 10.1016/j.tips.2009.02.006. [DOI] [PubMed] [Google Scholar]
O’Connor C, White KL, Doncescu N, Didenko T, Roth BL, Czaplicki G, Stevens RC, Wüthrich K, Milon A. Nmr structure and dynamics of the agonist dynorphin peptide bound to the human kappa opioid receptor. PNAS. 2015;112:11852–11857. doi: 10.1073/pnas.1510117112. [DOI] [PMC free article] [PubMed] [Google Scholar]
Olsen RHJ, DiBerto JF, English JG, Glaudin AM, Krumm BE, Slocum ST, Che T, Gavin AC, McCorvy JD, Roth BL, Strachan RT. TRUPATH, an open-source biosensor platform for interrogating the GPCR transducerome. Nature Chemical Biology. 2020;16:841–849. doi: 10.1038/s41589-020-0535-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
Pani B, Ahn S, Rambarat PK, Vege S, Kahsai AW, Liu A, Valan BN, Staus DP, Costa T, Lefkowitz RJ. Unique positive cooperativity between the β-arrestin-biased β-blocker carvedilol and a small molecule positive allosteric modulator of the β2-adrenergic receptor. Molecular Pharmacology. 2021;100:513–525. doi: 10.1124/molpharm.121.000363. [DOI] [PMC free article] [PubMed] [Google Scholar]
Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, Ferrin TE. UCSF chimera? A visualization system for exploratory research and analysis. Journal of Computational Chemistry. 2004;25:1605–1612. doi: 10.1002/jcc.20084. [DOI] [PubMed] [Google Scholar]
Pettersen EF, Goddard TD, Huang CC, Meng EC, Couch GS, Croll TI, Morris JH, Ferrin TE. UCSF chimerax: structure visualization for researchers, educators, and developers. Protein Science. 2021;30:70–82. doi: 10.1002/pro.3943. [DOI] [PMC free article] [PubMed] [Google Scholar]
Phillips JC, Braun R, Wang W, Gumbart J, Tajkhorshid E, Villa E, Chipot C, Skeel RD, Kalé L, Schulten K. Scalable molecular dynamics with NAMD. Journal of Computational Chemistry. 2005;26:1781–1802. doi: 10.1002/jcc.20289. [DOI] [PMC free article] [PubMed] [Google Scholar]
Punjani A, Fleet DJ. 3D variability analysis: resolving continuous flexibility and discrete heterogeneity from single particle cryo-EM. Journal of Structural Biology. 2021;213:107702. doi: 10.1016/j.jsb.2021.107702. [DOI] [PubMed] [Google Scholar]
Qi X, Liu H, Thompson B, McDonald J, Zhang C, Li X. Cryo-Em structure of oxysterol-bound human smoothened coupled to a heterotrimeric Gi. Nature. 2019;571:279–283. doi: 10.1038/s41586-019-1286-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
Rasmussen SGF, Choi HJ, Fung JJ, Pardon E, Casarosa P, Chae PS, DeVree BT, Rosenbaum DM, Thian FS, Kobilka TS, Schnapp A, Konetzki I, Sunahara RK, Gellman SH, Pautsch A, Steyaert J, Weis WI, Kobilka BK. Structure of a nanobody-stabilized active state of the β2 adrenoceptor. Nature. 2011;469:175–180. doi: 10.1038/nature09648. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ring AM, Manglik A, Kruse AC, Enos MD, Weis WI, Garcia KC, Kobilka BK. Adrenaline-activated structure of β2-adrenoceptor stabilized by an engineered nanobody. Nature. 2013;502:575–579. doi: 10.1038/nature12572. [DOI] [PMC free article] [PubMed] [Google Scholar]
Roe DR, Cheatham TE. PTRAJ and CPPTRAJ: software for processing and analysis of molecular dynamics trajectory data. Journal of Chemical Theory and Computation. 2013;9:3084–3095. doi: 10.1021/ct400341p. [DOI] [PubMed] [Google Scholar]
Ryckaert JP, Ciccotti G, Berendsen HJC. Numerical integration of the Cartesian equations of motion of a system with constraints: molecular dynamics of n-alkanes. Journal of Computational Physics. 1977;23:327–341. doi: 10.1016/0021-9991(77)90098-5. [DOI] [Google Scholar]
Schrage R, Seemann WK, Klöckner J, Dallanoce C, Racké K, Kostenis E, De Amici M, Holzgrabe U, Mohr K. Agonists with supraphysiological efficacy at the muscarinic M2 ACh receptor. British Journal of Pharmacology. 2013;169:357–370. doi: 10.1111/bph.12003. [DOI] [PMC free article] [PubMed] [Google Scholar]
Seyedabadi M, Gharghabi M, Gurevich EV, Gurevich VV. Structural basis of GPCR coupling to distinct signal transducers: implications for biased signaling. Trends in Biochemical Sciences. 2022;47:570–581. doi: 10.1016/j.tibs.2022.03.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
Shi L, Liapakis G, Xu R, Guarnieri F, Ballesteros JA, Javitch JA. Beta2 adrenergic receptor activation. Modulation of the proline kink in transmembrane 6 by a rotamer toggle switch. The Journal of Biological Chemistry. 2002;277:40989–40996. doi: 10.1074/jbc.M206801200. [DOI] [PubMed] [Google Scholar]
Slosky LM, Bai Y, Toth K, Ray C, Rochelle LK, Badea A, Chandrasekhar R, Pogorelov VM, Abraham DM, Atluri N, Peddibhotla S, Hedrick MP, Hershberger P, Maloney P, Yuan H, Li Z, Wetsel WC, Pinkerton AB, Barak LS, Caron MG. Β-Arrestin-Biased allosteric modulator of NTSR1 selectively attenuates addictive behaviors. Cell. 2020;181:1364–1379. doi: 10.1016/j.cell.2020.04.053. [DOI] [PMC free article] [PubMed] [Google Scholar]
Slosky LM, Caron MG, Barak LS. Biased allosteric modulators: new frontiers in GPCR drug discovery. Trends in Pharmacological Sciences. 2021;42:283–299. doi: 10.1016/j.tips.2020.12.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
Solt AS, Bostock MJ, Shrestha B, Kumar P, Warne T, Tate CG, Nietlispach D. Insight into partial agonism by observing multiple equilibria for ligand-bound and gs-mimetic nanobody-bound β1-adrenergic receptor. Nature Communications. 2017;8:1795. doi: 10.1038/s41467-017-02008-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
Staus DP, Hu H, Robertson MJ, Kleinhenz ALW, Wingler LM, Capel WD, Latorraca NR, Lefkowitz RJ, Skiniotis G. Structure of the M2 muscarinic receptor-β-arrestin complex in a lipid nanodisc. Nature. 2020;579:297–302. doi: 10.1038/s41586-020-1954-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
Suratman S, Leach K, Sexton P, Felder C, Loiacono R, Christopoulos A. Impact of species variability and `` probe-dependence'' on the detection and in vivo validation of allosteric modulation at the M4 muscarinic acetylcholine receptor. British Journal of Pharmacology. 2011;162:1659–1670. doi: 10.1111/j.1476-5381.2010.01184.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
Sušac L, Eddy MT, Didenko T, Stevens RC, Wüthrich K. A2A adenosine receptor functional states characterized by 19F-NMR. PNAS. 2018;115:12733–12738. doi: 10.1073/pnas.1813649115. [DOI] [PMC free article] [PubMed] [Google Scholar]
Terashi G, Wang X, Maddhuri Venkata Subramaniya SR, Tesmer JJG, Kihara D. Residue-wise local quality estimation for protein models from cryo-EM maps. Nature Methods. 2022;19:1116–1125. doi: 10.1038/s41592-022-01574-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
Thal DM, Sun B, Feng D, Nawaratne V, Leach K, Felder CC, Bures MG, Evans DA, Weis WI, Bachhawat P, Kobilka TS, Sexton PM, Kobilka BK, Christopoulos A. Crystal structures of the M1 and M4 muscarinic acetylcholine receptors. Nature. 2016;531:335–340. doi: 10.1038/nature17188. [DOI] [PMC free article] [PubMed] [Google Scholar]
Thal DM, Glukhova A, Sexton PM, Christopoulos A. Structural insights into G-protein-coupled receptor allostery. Nature. 2018;559:45–53. doi: 10.1038/s41586-018-0259-z. [DOI] [PubMed] [Google Scholar]
Tian W, Chen C, Lei X, Zhao J, Liang J. CASTp 3.0: computed atlas of surface topography of proteins. Nucleic Acids Research. 2018;46:W363–W367. doi: 10.1093/nar/gky473. [DOI] [PMC free article] [PubMed] [Google Scholar]
Tzavara ET, Bymaster FP, Davis RJ, Wade MR, Perry KW, Wess J, McKinzie DL, Felder C, Nomikos GG. M4 muscarinic receptors regulate the dynamics of cholinergic and dopaminergic neurotransmission: relevance to the pathophysiology and treatment of related CNS pathologies. FASEB Journal. 2004;18:1410–1412. doi: 10.1096/fj.04-1575fje. [DOI] [PubMed] [Google Scholar]
Valant C, Felder CC, Sexton PM, Christopoulos A. Probe dependence in the allosteric modulation of a G protein-coupled receptor: implications for detection and validation of allosteric ligand effects. Molecular Pharmacology. 2012;81:41–52. doi: 10.1124/mol.111.074872. [DOI] [PubMed] [Google Scholar]
Vanommeslaeghe K, MacKerell AD. Automation of the CHARMM General force field (cgenff) I: bond perception and atom typing. Journal of Chemical Information and Modeling. 2012a;52:3144–3154. doi: 10.1021/ci300363c. [DOI] [PMC free article] [PubMed] [Google Scholar]
Vanommeslaeghe K, Raman EP, MacKerell AD. Automation of the CHARMM General force field (cgenff) II: assignment of bonded parameters and partial atomic charges. Journal of Chemical Information and Modeling. 2012b;52:3155–3168. doi: 10.1021/ci3003649. [DOI] [PMC free article] [PubMed] [Google Scholar]
Vanommeslaeghe K, MacKerell AD. CHARMM additive and polarizable force fields for biophysics and computer-aided drug design. Biochimica et Biophysica Acta. 2015;1850:861–871. doi: 10.1016/j.bbagen.2014.08.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
Varadi M, Anyango S, Deshpande M, Nair S, Natassia C, Yordanova G, Yuan D, Stroe O, Wood G, Laydon A, Žídek A, Green T, Tunyasuvunakool K, Petersen S, Jumper J, Clancy E, Green R, Vora A, Lutfi M, Figurnov M, Cowie A, Hobbs N, Kohli P, Kleywegt G, Birney E, Hassabis D, Velankar S. AlphaFold protein structure database: massively expanding the structural coverage of protein-sequence space with high-accuracy models. Nucleic Acids Research. 2022;50:D439–D444. doi: 10.1093/nar/gkab1061. [DOI] [PMC free article] [PubMed] [Google Scholar]
Vuckovic Z, Gentry PR, Berizzi AE, Hirata K, Varghese S, Thompson G, Westhuizen ET, Burger WAC, Rahmani R, Valant C, Langmead CJ, Lindsley CW, Baell JB, Tobin AB, Sexton PM, Christopoulos A, Thal DM. Crystal structure of the m5 muscarinic acetylcholine receptor. PNAS. 2019;116:26001–26007. doi: 10.1073/pnas.1914446116. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wacker D, Wang C, Katritch V, Han GW, Huang XP, Vardy E, McCorvy JD, Jiang Y, Chu M, Siu FY, Liu W, Xu HE, Cherezov V, Roth BL, Stevens RC. Structural features for functional selectivity at serotonin receptors. Science. 2013;340:615–619. doi: 10.1126/science.1232808. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wagner T, Merino F, Stabrin M, Moriya T, Antoni C, Apelbaum A, Hagel P, Sitsel O, Raisch T, Prumbaum D, Quentin D, Roderer D, Tacke S, Siebolds B, Schubert E, Shaikh TR, Lill P, Gatsogiannis C, Raunser S. SPHIRE-cryolo is a fast and accurate fully automated particle picker for cryo-EM. Communications Biology. 2019;2:218. doi: 10.1038/s42003-019-0437-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wang J, Arantes PR, Bhattarai A, Hsu RV, Pawnikar S, Huang Y-MM, Palermo G, Miao Y. Gaussian accelerated molecular dynamics (gamd): principles and applications. Wiley Interdisciplinary Reviews. Computational Molecular Science. 2021a;11:e1521. doi: 10.1002/wcms.1521. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wang J, Pani B, Gokhan I, Xiong X, Kahsai AW, Jiang H, Ahn S, Lefkowitz RJ, Rockman HA. Β-Arrestin-Biased allosteric modulator potentiates carvedilol-stimulated β adrenergic receptor cardioprotection. Molecular Pharmacology. 2021b;100:568–579. doi: 10.1124/molpharm.121.000359. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wang J, Wu M, Chen Z, Wu L, Wang T, Cao D, Wang H, Liu S, Xu Y, Li F, Liu J, Chen N, Zhao S, Cheng J, Wang S, Hua T. The unconventional activation of the muscarinic acetylcholine receptor M4R by diverse ligands. Nature Communications. 2022;13:2855. doi: 10.1038/s41467-022-30595-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
Warne T, Edwards PC, Leslie AGW, Tate CG. Crystal structures of a stabilized β1-adrenoceptor bound to the biased agonists bucindolol and carvedilol. Structure. 2012;20:841–849. doi: 10.1016/j.str.2012.03.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
Warne T, Edwards PC, Doré AS, Leslie AGW, Tate CG. Molecular basis for high-affinity agonist binding in GPCRs. Science. 2019;364:775–778. doi: 10.1126/science.aau5595. [DOI] [PMC free article] [PubMed] [Google Scholar]
White KL, Eddy MT, Gao ZG, Han GW, Lian T, Deary A, Patel N, Jacobson KA, Katritch V, Stevens RC. Structural connection between activation microswitch and allosteric sodium site in GPCR signaling. Structure. 2018;26:259–269. doi: 10.1016/j.str.2017.12.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
Williams CJ, Headd JJ, Moriarty NW, Prisant MG, Videau LL, Deis LN, Verma V, Keedy DA, Hintze BJ, Chen VB, Jain S, Lewis SM, Arendall WB, Snoeyink J, Adams PD, Lovell SC, Richardson JS, Richardson DC. MolProbity: more and better reference data for improved all-atom structure validation. Protein Science. 2018;27:293–315. doi: 10.1002/pro.3330. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wingler L.M, Elgeti M, Hilger D, Latorraca NR, Lerch MT, Staus DP, Dror RO, Kobilka BK, Hubbell WL, Lefkowitz RJ. Angiotensin analogs with divergent bias stabilize distinct receptor conformations. Cell. 2019;176:468–478. doi: 10.1016/j.cell.2018.12.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wood MR, Noetzel MJ, Melancon BJ, Poslusney MS, Nance KD, Hurtado MA, Luscombe VB, Weiner RL, Rodriguez AL, Lamsal A, Chang S, Bubser M, Blobaum AL, Engers DW, Niswender CM, Jones CK, Brandon NJ, Wood MW, Duggan ME, Conn PJ, Bridges TM, Lindsley CW. Discovery of vu0467485/az13713945: an M4 PAM evaluated as a preclinical candidate for the treatment of schizophrenia. ACS Medicinal Chemistry Letters. 2017a;8:233–238. doi: 10.1021/acsmedchemlett.6b00461. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wood MR, Noetzel MJ, Poslusney MS, Melancon BJ, Tarr JC, Lamsal A, Chang S, Luscombe VB, Weiner RL, Cho HP, Bubser M, Jones CK, Niswender CM, Wood MW, Engers DW, Brandon NJ, Duggan ME, Conn PJ, Bridges TM, Lindsley CW. Challenges in the development of an M4 PAM in vivo tool compound: the discovery of VU0467154 and unexpected DMPK profiles of close analogs. Bioorganic & Medicinal Chemistry Letters. 2017b;27:171–175. doi: 10.1016/j.bmcl.2016.11.086. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wootten D, Christopoulos A, Sexton PM. Emerging paradigms in GPCR allostery: implications for drug discovery. Nature Reviews. Drug Discovery. 2013;12:630–644. doi: 10.1038/nrd4052. [DOI] [PubMed] [Google Scholar]
Xu J, Hu Y, Kaindl J, Risel P, Hübner H, Maeda S, Niu X, Li H, Gmeiner P, Jin C, Kobilka BK. Conformational complexity and dynamics in a muscarinic receptor revealed by NMR spectroscopy. Molecular Cell. 2019;75:53–65. doi: 10.1016/j.molcel.2019.04.028. [DOI] [PubMed] [Google Scholar]
Xu J, Hübner H, Hu Y, Niu X, Gmeiner P, Jin C, Kobilka B. An Allosteric Ligand Stabilizes Distinct Conformations in the M2 Muscarinic Acetylcholine Receptor. bioRxiv. 2021 doi: 10.1101/2021.02.14.431178. [DOI]
Xu J, Wang Q, Hübner H, Hu Y, Niu X, Wang H, Maeda S, Inoue A, Tao Y, Gmeiner P, Du Y, Jin C, Kobilka BK. Structural and dynamic insights into supra-physiological activation and allosteric modulation of a muscarinic acetylcholine receptor. Nature Communications. 2023;14:376. doi: 10.1038/s41467-022-35726-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ye L, Van Eps N, Zimmer M, Ernst OP, Prosser RS. Activation of the A2A adenosine G-protein-coupled receptor by conformational selection. Nature. 2016;533:265–268. doi: 10.1038/nature17668. [DOI] [PubMed] [Google Scholar]
Zhang K. Gctf: real-time CTF determination and correction. Journal of Structural Biology. 2016;193:1–12. doi: 10.1016/j.jsb.2015.11.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
Zhang X, Belousoff MJ, Zhao P, Kooistra AJ, Truong TT, Ang SY, Underwood CR, Egebjerg T, Šenel P, Stewart GD, Liang YL, Glukhova A, Venugopal H, Christopoulos A, Furness SGB, Miller LJ, Reedtz-Runge S, Langmead CJ, Gloriam DE, Danev R, Sexton PM, Wootten D. Differential GLP-1R binding and activation by peptide and non-peptide agonists. Molecular Cell. 2020;80:485–500. doi: 10.1016/j.molcel.2020.09.020. [DOI] [PubMed] [Google Scholar]
Zheng SQ, Palovcak E, Armache JP, Verba KA, Cheng Y, Agard DA. MotionCor2: anisotropic correction of beam-induced motion for improved cryo-electron microscopy. Nature Methods. 2017;14:331–332. doi: 10.1038/nmeth.4193. [DOI] [PMC free article] [PubMed] [Google Scholar]
Zhou Q, Yang D, Wu M, Guo Y, Guo W, Zhong L, Cai X, Dai A, Jang W, Shakhnovich EI, Liu ZJ, Stevens RC, Lambert NA, Babu MM, Wang MW, Zhao S. Common activation mechanism of class A GPCRs. eLife. 2019;8:e50279. doi: 10.7554/eLife.50279. [DOI] [PMC free article] [PubMed] [Google Scholar]
Zivanov J, Nakane T, Forsberg BO, Kimanius D, Hagen WJ, Lindahl E, Scheres SH. New tools for automated high-resolution cryo-EM structure determination in RELION-3. eLife. 2018;7:e42166. doi: 10.7554/eLife.42166. [DOI] [PMC free article] [PubMed] [Google Scholar]

eLife. doi: 10.7554/eLife.83477.sa0

Editor's evaluation

Sudarshan Rajagopal ¹

This important work advances our understanding of the structural basis of allosteric modulation of the M4 muscarinic receptor but has broad implications for GPCRs. The evidence supporting the conclusions is exceptional, with multiple cryo-EM structures that are complemented by excellent pharmacological and dynamics studies.

eLife. doi: 10.7554/eLife.83477.sa1

Decision letter

Editor: Sudarshan Rajagopal¹

Reviewed by: Sudarshan Rajagopal²

Our editorial process produces two outputs: (i) public reviews designed to be posted alongside the preprint for the benefit of readers; (ii) feedback on the manuscript for the authors, including requests for revisions, shown below. We also include an acceptance summary that explains what the editors found interesting or important about the work.

Decision letter after peer review:

Thank you for submitting your article "Pharmacological hallmarks of allostery at the M4 muscarinic receptor elucidated through structure and dynamics" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, including Sudarshan Rajagopal as the Reviewing Editor and Reviewer #1, and the evaluation has been overseen by Richard Aldrich as the Senior Editor.

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

1) Please address the comment from reviewer 1 regarding log tau vs log tau/KD for assessing the efficacy of ligands.

2) Please address the comments from reviewers 2-3, especially with the presentation of the data/figures for improved readability.

Reviewer #1 (Recommendations for the authors):

I have a concern about the use of a transducer coupling coefficient (log(tau/K)) in the manuscript. As the authors note, this is typically used to quantify agonist bias, usually by comparing it to a coupling coefficient for a different signaling pathway by the same agonist. But coupling coefficients really can't be used to compare different agonists because of different dissociation constants for each agonist – a reference is made to Kenakin, 2012 (which I went back to and didn't find this argument) that the coupling coefficient characterizes agonism of a specific pathway defined as the interaction between an agonist, receptor, and transducer. This leads to some questionable interpretations in lines 177-181. I believe the proper interpretation is that ACh is more efficacious as it has a larger log tau (which has been shown to be related to proportional efficacy by Onaran et al. – Sci Rep. 2017 Mar 14;7:44247. doi: 10.1038/srep44247.) – the transducer coupling coefficient doesn't matter as it is being driven by the tighter binding. Using the ternary complex model, efficacy is proportional to the differences in affinity between the transducer-bound and unbound states, while affinity largely reflects the interaction with the transducer-unbound state (depending on the experimental platform).

The same issue arises in lines 613-619 where an argument is made that "structures of GPCRs in a ternary complex… are better represented by their transducer coupling coefficient than the efficacy of the agonist…." Essentially, this is making the argument that efficacy/affinity is a better representation of the ternary complex than efficacy. Let's take an example of the b2 adrenergic receptor (which I know better than the M4) – isoproterenol has a log KD of -6 but is a full agonist hitting Emax. Pindolol is a very weak partial agonist but binds tightly with log KD -9.3 but with an efficacy of 10% of isoproterenol. If I calculate tau/KD for these two compounds, pindolol would have a much higher value – but it would be completely driven by the high binding affinity. It would not reflect the stability of the ternary complex. The interpretation that the transducer coupling coefficient is in contradiction to basic tenets of pharmacology.

I think the section on "Structural and dynamic insights into orthosteric and allosteric agonism" proposes some plausible ideas but probably overstates the insights that can be obtained from transducer-bound structures. As with other crystal or cryoEM structures, the observed conformation of the receptor is largely driven by the bound transducer and not by the pharmacological characteristics of the agonist (partial vs full, etc. – although I doubt you could get an antagonist bound to a transducer-bound receptor). This limitation has to be highlighted as it is a major one for these structural studies (and is theoretically less of an issue with solution-based studies).

Figure 6. It would be helpful to label some of the microswitches in the figure, as it may not be obvious to all readers, e.g., me, what specific residues they are looking at.

Figure S2. Any comment on the different populations of complexes with the PAMs? It is interesting that VU154 induces a large population of presumably a different conformation.

- Consider focusing on log tau as a readout of efficacy as opposed to the transduction coefficients, which have limited utility when comparing agonists.

- A sentence or two on limitations of the cryoEM approach in studying agonism due to the stabilization of the A-R-T complex.

Reviewer #2 (Recommendations for the authors):

1. The abstract focuses on the questions and the methods used to address them, with relatively little in terms of specific details about what is found. Simply saying the work offers "in-depth insights", while true, is a bit vague and leaves the reader guessing. Some more specific statements about key findings (e.g., the superior coupling to G protein induced by ACh) would be appreciated.

2. Figure 2 panels B-E and G-J might be made larger, and the data points for the other panels smaller, for clarity. The data points obscure error bars in all but a few cases.

3. Figure 3 shows a huge amount of structural data, but this means all panels in the figure are very small. The authors may want to consider a more focused presentation, with fewer essential components left for the SI. Panels A-C are somewhat redundant, for example, and a single representative example might suffice.

4. The discussion of Wang et al. 2022 and the corresponding figure S5 are very important. Figure S5 was difficult to interpret and see clearly, perhaps in part due to image compression by the journal. Even so, a more focused presentation would still be helpful, ideally with a finer map mesh. A transparent surface view may be even easier to interpret.

5. The discussion is clearly written and helpful, but reads as being targeted to a specialist pharmacology audience. Including more of a broad perspective and relating this to allostery in other GPCRs and other protein families could enhance the overall breadth of appeal.

6. The final sentence about signaling bias (line 624) is important, and the authors may want to state explicitly that the structural and mechanistic basis for biased signaling (and allostery as well) is likely to vary from one receptor to another. This is implied, but if the authors feel justified in making a stronger statement it may be helpful for clarity and emphasis.

Reviewer #3 (Recommendations for the authors):

1. I would propose merging Figures 1 and 2 to clarify which aspect is covered for each experiment in Figure 2. It would be great to use a color code (or any other sort of label) for each parameter (K, α, etc.) and highlight the respective plots (including supplements), for readers to follow the discussion.

2. It would also be helpful to show the structures of all four compounds in the main text, for clarity. These could be used to highlight which parts of the ligands were observed in the structure, vs. which ones were disordered (i.e. which bonds are rotating freely), as discussed for both Ipx and VU154.

3. All pharmacological experiments have been performed with a full-length wild type, however, the cryoEM structure contains a major deletion of ICL3. While I understand that previous experiments have used similar constructs for structure determination, I believe it would be helpful to confirm binding affinities, as well as the efficacy of the respective drugs with the truncated cryoEM construct. If this is not feasible, please highlight the caveat that all structures, and consequently all MD simulations, are based on data obtained from an engineered construct, rather than the wildtype sequence.

4. Line 190: What is a "NAL" (i presume it should spell 'NAM')? Please also discuss the impact of this observation.

5. The pharmacological experiments were conducted with all possible combinations of ligands (ACh, Ipx with LY298, VU154). I understand that it could have been merely a factor of resources, but were any attempts made to elucidate structures of the PAM-complexes with ACh?

6. It is not clear to the reader, whether the above pharmacological experiments (Figure 2) have been carried out for the very first time, or if others have attempted similar studies. Please clarify exactly what part of the work is novel.

7. Figure 2A: the fact that increasing concentrations of LY298 appear to block the overall binding of ACh is not described or discussed anywhere. Based on this plot LY298 would be a PAM-antagonist (see for example Figure 1, Grundmann et al. 2021, 10.3390/ijms22041763). This would be an important aspect, which would need to be addressed.

8. I would suggest placing some of the MD simulation traces into the supplementary materials, as these currently take up a large fraction of all figures. Alternatively, different complexes could be color coded and overlaid in one figure to highlight differences.

9. Figure 4 B and H: given the structure in Figure 4 H, it means that the binding pocket around Ipx leaves no room for any movement. Overall, I am not able to follow the discussion regarding the alkyne group/linker of Ipx not being visible. If the start and end points are fixed (visible densities), and have a rigid, planar triple bond involved (alkyne), I find it hard to imagine that the linker is flexible enough to wash out the signal for the linker. Also, the representative Figure S3F is not very convincing, as (A) all iperoxo densities seem to be of rather poor quality and (B) the average of all aligned structures would still likely result in an 'average-able' linker density. I would suggest either elaborating on or omitting this claim.

10. Line 301, The predominate χ2 angle of W413 was approximately 60◦ and 105◦ in the ACh-bound and Ipx-bound simulations, respectively, corresponding to the cryo-EM conformations. As depicted in Figure 4L, W413 when bound to ACh also samples angles close to 90 degrees or higher in the majority of Sim2 and part of Sim1. What is the significance of this conformational sampling?

11. On the note of ligands having flexible parts in VU154, and therefore no resolved densities in the maps, is there any pharmacological evidence (i.e. SAR) for these regions not contributing to binding/signaling? Analogously, are there any SAR data for replacing the alkyne bond in Ipx? It would be conceivable that a rigid replacement linker would affect (either positively or negatively) receptor binding and signaling.

12. According to the GaMD simulation results in Figure S7C, the minimum distance for the T433 and VU154 seems to be close to 4 Å while the predominant distance is around 7 Å. Therefore, I am wondering what the significance of this hydrogen bond is observed in the cryoEM structure. Additionally, T433R mutant showed increased binding to VU154 and showed the importance of T433 in species selectivity. Is this increased binding of VU154 with T433R mutant a result of a more "stable" hydrogen bond between receptor and VU154?

13. Regarding the species selectivity aspect, there is no mention of the V91L mutant in Figure 7, only as part of the triple mutant. It is hard to judge which mutations are responsible for species selectivity without either showing results for the D432E/T433R double mutant, or the additional V91L single mutant.

eLife. 2023 May 30;12:e83477. doi: 10.7554/eLife.83477.sa2

Author response

Essential revisions:

1) Please address the comment from reviewer 1 regarding log tau vs log tau/KD for assessing the efficacy of ligands.

2) Please address the comments from reviewers 2-3, especially with the presentation of the data/figures for improved readability.

We appreciate and thank the editors and reviewers for their time and effort providing positive and constructive feedback on the manuscript. For (1) we now have further clarified how and why we estimate both ligand efficacy and transducer coefficients of ligands in the manuscript, taking into account the reviewer concerns and suggestions (see detailed response below). For (2) we have made substantial revisions to the figures and text in accordance with the suggestions of the reviewers. These changes were done in accordance with the general editorial recommendations of eLife.

Reviewer #1 (Recommendations for the authors):

I have a concern about the use of a transducer coupling coefficient (log(tau/K)) in the manuscript. As the authors note, this is typically used to quantify agonist bias, usually by comparing it to a coupling coefficient for a different signaling pathway by the same agonist. But coupling coefficients really can't be used to compare different agonists because of different dissociation constants for each agonist – a reference is made to Kenakin, 2012 (which I went back to and didn't find this argument) that the coupling coefficient characterizes agonism of a specific pathway defined as the interaction between an agonist, receptor, and transducer. This leads to some questionable interpretations in lines 177-181. I believe the proper interpretation is that ACh is more efficacious as it has a larger log tau (which has been shown to be related to proportional efficacy by Onaran et al. – Sci Rep. 2017 Mar 14;7:44247. doi: 10.1038/srep44247.) – the transducer coupling coefficient doesn't matter as it is being driven by the tighter binding. Using the ternary complex model, efficacy is proportional to the differences in affinity between the transducer-bound and unbound states, while affinity largely reflects the interaction with the transducer-unbound state (depending on the experimental platform).

The same issue arises in lines 613-619 where an argument is made that "structures of GPCRs in a ternary complex… are better represented by their transducer coupling coefficient than the efficacy of the agonist…." Essentially, this is making the argument that efficacy/affinity is a better representation of the ternary complex than efficacy. Let's take an example of the b2 adrenergic receptor (which I know better than the M4) – isoproterenol has a log KD of -6 but is a full agonist hitting Emax. Pindolol is a very weak partial agonist but binds tightly with log KD -9.3 but with an efficacy of 10% of isoproterenol. If I calculate tau/KD for these two compounds, pindolol would have a much higher value – but it would be completely driven by the high binding affinity. It would not reflect the stability of the ternary complex. The interpretation that the transducer coupling coefficient is in contradiction to basic tenets of pharmacology.

I think the section on "Structural and dynamic insights into orthosteric and allosteric agonism" proposes some plausible ideas but probably overstates the insights that can be obtained from transducer-bound structures. As with other crystal or cryoEM structures, the observed conformation of the receptor is largely driven by the bound transducer and not by the pharmacological characteristics of the agonist (partial vs full, etc. – although I doubt you could get an antagonist bound to a transducer-bound receptor). This limitation has to be highlighted as it is a major one for these structural studies (and is theoretically less of an issue with solution-based studies).

We appreciate the reviewer’s concern about our use and description of transducer coupling coefficients in the manuscript (as originally written), but respectfully disagree with the comment that “…coupling coefficients really can't be used to compare different agonists because of different dissociation constants for each agonist…”. This is because the latter statement is dependent on the context in which the transducer coupling coefficient is used, and we apologise for not being clearer in terms of both its definition and the reason why we included this additional parameter in our analyses of our functional data.

In essence, we feel that there are two main concerns that are raised by the reviewer regarding the transducer coupling coefficient:

1. The purpose of this parameter when it comes to quantifying biased agonism.

2. The relevance of this parameter for interpreting structural biology data of the ternary complex of agonist (+/- PAM)-receptor-transducer.

In terms of the first concern, the transducer coupling coefficient is used as a starting point for the ultimate quantification of biased agonism in cellular signaling pathways; it is, in-and-of-itself, not the ‘end point’ for the comparison of agonists. As described in Kenakin and Christopoulos, 2013, Nat. Rev. Drug Discovery, 12: 205, this parameter is used specifically because bias may arise from different tau (efficacy) values for a given pathway linked to a single (common) affinity value and/or a different affinity value linked to a given pathway (in the intact cellular environment). The affinity (K) parameter in this operational model-based approach thus represents a ‘conditional’ or ‘apparent/observed’ affinity, even for the ‘ground state’, because it subsumes the potential for the receptor to isomerize between different states (transducer-bound or otherwise) prior to initiating a cellular response. To date (and to our knowledge), in most instances in the literature, the measured affinity of an agonist for a given receptor is essentially identical between pathways (irrespective of rapid isomerisation between more than one state, i.e., K_D = K_Observed) – as would normally be expected and, in which case, we completely agree with the reviewer that a comparison of efficacy (tau) values alone is sufficient for the quantification of biased agonism and for interpreting the classic ternary complex. However, there are some examples of where the estimated affinity of an agonist for a specific pathway is significantly different between biochemical and cell-based assays – in which case a comparison of tau values alone would yield errors in bias estimates if a single affinity value is assumed/used in calculations. For example, please see Gregory et al., 2010 JBC, 285: 7459 and Davey et al., 2012, Endocrinology, 153: 1232. In the former paper, the estimated affinity of the agonist, McN-A-343 for the M₂ muscarinic receptor based on biochemical (radioligand binding) assays is approx.10-fold lower than any functional estimate of its affinity across different pathways. In the latter paper, the estimated affinity (pK_B) of the PAM, cinacalcet, is identical when determined in cellular assays of intracellular calcium mobilization or downstream pERK1/2 (pK_B = 6.73), but significantly different (pK_B = 8.14) in an assay of plasma membrane ruffling – all determined in the same cellular background. Although the mechanisms underlying these observations remain unclear, they nonetheless suggest that GPCRs can sometimes adopt a different affinity state depending on the signaling pathway that is retained over the time course of a given experiment (e.g., via compartmentalisation), such that a single affinity value that would normally represent rapid isomerization of the ground state between any number of different states would fail to account for the data on the basis of tau values alone.

It is for this latter reason that we choose to use the tau/K_A ratio, rather than comparison of tau values alone, as the starting point for quantifying bias – while also using a reference agonist across different pathways. Based on the method for estimating bias as described in Kenakin et al., 2012, ACS Chem. Neurosci. 3: 193 and summarized in Kenakin and Christopoulos, 2013, Nat. Rev. Drug Discovery, 12: 205, if the affinity values of the agonists are identical between pathways (as would be expected in most instances according to classic receptor theory) then the method essentially becomes a comparison of tau values. However, if this assumption is not met, then errors will arise using tau values alone, but not when using tau/K_A ratios (Kenakin and Christopoulos, 2013, Nat. Rev. Drug Discovery, 12: 205).

Nonetheless, in order to better address this issue and the reviewer’s concern, we have re-written the original description in the Results of the transducer coupling coefficient (original lines 177 – 181) to more explicitly define the meaning of this parameter, as well as addressing the ambiguity in our original statement that this parameter ‘…accounts for the ground state binding affinity of the agonist….and characterises the agonism of a specific pathway defined as the interaction between agonist, receptor and transducer…etc.’ to state the following:

“These signaling assays allowed us to determine the efficacy of the agonists (τA) and the PAMs (τB) (Figure 1H, Figure 1 —figure supplement 2B). Importantly, efficacy (τ ), as defined from the Black-Leff operational model of agonism (Black and Leff, 1983), is determined by the ability of an agonist to promote an active receptor conformation, the receptor density (Bmax), and the subsequent ability of a cellular system to generate a response (Figure 1B). Notably, in both signalling assays, the rank order of efficacy was ACh > Ipx > LY298 > VU154. We subsequently calculated the transducer coupling coefficient (τ/K) (Figure 1I, Figure 1 —figure supplement 2C), a parameter often used as a starting point to quantify biased agonism (Kenakin et al., 2012) and that is specific to the intact cellular environment in which a given response occurs. Thus the dissociation constant (K) in the transduction coefficient subsumes the affinity for the ground state (non-bound) receptor, in addition to any isomerisation states of the receptor that ultimately yield cellular responses (Kenakin and Christopoulos, 2013).”

We also totally agree with the reviewer’s comment regarding the transduction coefficients of isoproterenol (high efficacy) and pindolol (low efficacy) but, nonetheless, using these values as a starting point in the ultimate quantification of bias (not efficacy) of these ligands between pathways would still yield the correct result.

In terms of the second concern, we agree with the reviewer that our suggestion that the transducer coupling coefficient is a ‘better’ representation of efficacy in an atomic resolution structure of the ternary complex than, e.g., the tau parameter alone is likely an overstatement. We have thus removed this statement altogether and re-written that section of the Discussion accordingly.

We also agree with the reviewer that most GPCR structures to date suggest that the observed conformation is largely driven by the bound transducer, but feel that the field is still ‘relatively’ nascent in a number of key areas: for instance, the fact that many important structures (e.g., unliganded states; antagonist-bound receptor-transducer states; PAM-bound, agonist-free states, etc.) have not been readily solved; the fact that (as shown in our MD simulations) the dynamics of the complex must also be incorporated into the understanding of the mechanisms governing stability of any observed structure and, ultimately, the need to strive to improve our ability to link solved structures to biochemistry and to intact cell pharmacology. It is for this reason that we have chosen to include transducer coupling coefficient estimates, in addition to independent affinity and efficacy (tau) estimates, because both parameters play different roles in ultimately defining the stability and signaling capacity of a given complex; the transducer coupling coefficient is another means of reinforcing this point.

Figure 6. It would be helpful to label some of the microswitches in the figure, as it may not be obvious to all readers, e.g., me, what specific residues they are looking at.

We thank the reviewer for this suggestion, and have labelled the microswitches in this figure accordingly.

Figure S2. Any comment on the different populations of complexes with the PAMs? It is interesting that VU154 induces a large population of presumably a different conformation.

Figure S2 shows the cryo-EM data processing workflow that was done for each complex. The 3D classification doesn’t necessarily reflect differences in conformations or populations of receptor complexes. The “different populations” were likely due to ‘bad particles’ (i.e., receptor complexes) that arise from differences in the purification, stability of the complexes, vitrification, imaging, and processing of the samples.

- Consider focusing on log tau as a readout of efficacy as opposed to the transduction coefficients, which have limited utility when comparing agonists.

In general, throughout the text we have now focused more on efficacy, and in Figure 7C we have replaced the transduction coefficients with efficacy estimates alone.

- A sentence or two on limitations of the cryoEM approach in studying agonism due to the stabilization of the A-R-T complex.

We have amended the discussion to include this limitation:

“It is worth noting that structures of GPCRs bound to agonists with different pharmacological properties (full, partial, and biased agonists) have now been reported for some GPCRs (Liang et al., 2018a; Masureel et al., 2018; McCorvy et al., 2018; Ring et al., 2013; Wacker et al., 2013; Warne et al., 2012; Wingler et al., 2019). However, insights gained from such cryo-EM and x-ray crystallography structures may be limited, due to the role that the bound transducer plays on the observed final receptor conformation, and not necessarily due solely to the properties of the ligand. The ultimate underlying conformational differences, therefore, are likely to be subtle and dynamic (Seyedabadi et al., 2022), requiring application of additional techniques such as NMR spectroscopy, single molecule FRET and MD simulations for furthering our understanding (Cao et al., 2021; Cong et al., 2021; Gregorio et al., 2017; Huang et al., 2021; Katayama et al., 2021; J. J. Liu et al., 2012; Solt et al., 2017; Sušac et al., 2018; Xu et al., 2023; Ye et al., 2016, p. 19).”

Reviewer #2 (Recommendations for the authors):

1. The abstract focuses on the questions and the methods used to address them, with relatively little in terms of specific details about what is found. Simply saying the work offers "in-depth insights", while true, is a bit vague and leaves the reader guessing. Some more specific statements about key findings (e.g., the superior coupling to G protein induced by ACh) would be appreciated.

We appreciate the reviewer’s concerns and have amended the abstract to more explicitly highlight the key findings while also trying to remain within the constraints of abstract length.

“Allosteric modulation of G protein-coupled receptors (GPCRs) is a major paradigm in drug discovery. Despite decades of research, a molecular-level understanding of the general principles that govern the myriad pharmacological effects exerted by GPCR allosteric modulators remains limited. The M4 muscarinic acetylcholine receptor (M4 mAChR) is a validated and clinically relevant allosteric drug target for several major psychiatric and cognitive disorders. In this study, we rigorously quantified the affinity, efficacy and magnitude of modulation of two different positive allosteric modulators, LY2033298 (LY298) and VU0467154 (VU154), combined with the endogenous agonist acetylcholine (ACh) or the high-affinity agonist iperoxo (Ipx), at the human M4 mAChR. By determining the cryo-electron microscopy (cryo-EM) structures of the M4 mAChR, bound to a cognate Gi1 protein and in complex with ACh, Ipx, LY298-Ipx, and VU154-Ipx, and applying molecular dynamics (MD) simulations we determine key molecular mechanisms underlying allosteric pharmacology. In addition to delineating the contribution of spatially distinct binding sites on observed pharmacology, our findings also revealed a vital role for orthosteric and allosteric ligand-receptor-transducer complex stability, mediated by conformational dynamics between these sites, in the ultimate determination of affinity, efficacy, cooperativity, probe-dependence and species variability. There results provide a holistic framework for further GPCR mechanistic studies and can aid the discovery and design of future allosteric drugs.”

2. Figure 2 panels B-E and G-J might be made larger, and the data points for the other panels smaller, for clarity. The data points obscure error bars in all but a few cases.

We have decreased the size of the data points and increased the thickness of the error bars. We have increased the size of panels (B-E, G-J) and moved the concentration response curves to supplemental: Figure 1 —figure supplement 1.

3. Figure 3 shows a huge amount of structural data, but this means all panels in the figure are very small. The authors may want to consider a more focused presentation, with fewer essential components left for the SI. Panels A-C are somewhat redundant, for example, and a single representative example might suffice.

We now only show the M4R-Gi-Ipx panel as a single representative example. The remaining structures and G protein comparison have been moved to Figure 2 —figure supplement 1.

4. The discussion of Wang et al. 2022 and the corresponding figure S5 are very important. Figure S5 was difficult to interpret and see clearly, perhaps in part due to image compression by the journal. Even so, a more focused presentation would still be helpful, ideally with a finer map mesh. A transparent surface view may be even easier to interpret.

We thank the reviewer for suggesting a transparent surface view. Indeed, this is easier to interpret. We have thus removed one set of the panels, allowing an increase in the size of the others.

5. The discussion is clearly written and helpful, but reads as being targeted to a specialist pharmacology audience. Including more of a broad perspective and relating this to allostery in other GPCRs and other protein families could enhance the overall breadth of appeal.

We have trimmed and amended the discussion such that it is balanced towards a broader audience.

6. The final sentence about signaling bias (line 624) is important, and the authors may want to state explicitly that the structural and mechanistic basis for biased signaling (and allostery as well) is likely to vary from one receptor to another. This is implied, but if the authors feel justified in making a stronger statement it may be helpful for clarity and emphasis.

We have added this statement to a section describing the limitations of the study.

Reviewer #3 (Recommendations for the authors):

1. I would propose merging Figures 1 and 2 to clarify which aspect is covered for each experiment in Figure 2. It would be great to use a color code (or any other sort of label) for each parameter (K, α, etc.) and highlight the respective plots (including supplements), for readers to follow the discussion.

We have merged Figures 1 and 2 together and included the chemical structures of the compounds into Figure 1. The concentration response curves have been moved to Figure 1 —figure supplement 1 and 2. We’ve experimented with colour coding, but have not found a way to do this with sufficient clarity. Instead, we’ve linked the parameters in panel B to their respective panels below in the text of the schematic.

2. It would also be helpful to show the structures of all four compounds in the main text, for clarity. These could be used to highlight which parts of the ligands were observed in the structure, vs. which ones were disordered (i.e. which bonds are rotating freely), as discussed for both Ipx and VU154.

Chemical structures have been included in Figure 1. The cryo-EM density surrounding the ligands at an appropriate contour level have been added to Figure 2.

3. All pharmacological experiments have been performed with a full-length wild type, however, the cryoEM structure contains a major deletion of ICL3. While I understand that previous experiments have used similar constructs for structure determination, I believe it would be helpful to confirm binding affinities, as well as the efficacy of the respective drugs with the truncated cryoEM construct. If this is not feasible, please highlight the caveat that all structures, and consequently all MD simulations, are based on data obtained from an engineered construct, rather than the wildtype sequence.

We agree that this is an important caveat, and have added a comment to this effect in a ‘limitations’ section at the end of the manuscript.

4. Line 190: What is a "NAL" (i presume it should spell 'NAM')? Please also discuss the impact of this observation.

We apologize for the use of this abbreviation without context. A ‘NAL’ is a ‘neutral allosteric ligand’ i.e., a ligand that occupies the allosteric site but has neutral cooperativity (ab = 1) with the orthosteric ligand (but can still compete with other allosteric ligands binding to the same site – see Christopoulos et al., 2014, Pharmacol. Rev.66: 918). We have thus amended the sentence to explicitly state:

“on this basis, VU154 would be classified as a ‘neutral’ allosteric ligand (not a PAM) with Ipx in the TruPath assay, i.e., VU154 still binds to the allosteric site, but displays neutral cooperativity (ab=1) with Ipx”.

5. The pharmacological experiments were conducted with all possible combinations of ligands (ACh, Ipx with LY298, VU154). I understand that it could have been merely a factor of resources, but were any attempts made to elucidate structures of the PAM-complexes with ACh?

Ipx was the initial ligand of choice for the structural studies, largely due to its high affinity and prior use in structural studies of the M₂ mAChR. Structures were not attempted with ACh and LY298/VU154 due to limited cryo-EM resources.

6. It is not clear to the reader, whether the above pharmacological experiments (Figure 2) have been carried out for the very first time, or if others have attempted similar studies. Please clarify exactly what part of the work is novel.

We apologize for lack of clarity. The pharmacology of LY298 with ACh has been relatively well characterized in binding and functional assays by us and other groups as has, albeit to a lesser extent, the pharmacology of VU154. As far as we are aware, however, the TruPath data are new for both ligands ,as well as the effects of VU154 on pERK1/2. Moreover, the pharmacological effects of LY298 and VU154 on Ipx have not been reported to our knowledge. We’ve added the following lines to the manuscript to make this issue clearer:

“The pharmacology of LY298 or VU154 interacting with ACh have been well characterized in binding and functional assays at the M4 mAChR (Bubser et al., 2014; Chan et al., 2008; Gould et al., 2016; Leach et al., 2010; Suratman et al., 2011; Thal et al., 2016). However, their pharmacology with Ipx has not been reported. Therefore, we characterized both PAMs with ACh and Ipx in binding and in two different functional assays to provide a thorough foundational comparative characterization of the pharmacological parameters of these ligands from the same study.”

7. Figure 2A: the fact that increasing concentrations of LY298 appear to block the overall binding of ACh is not described or discussed anywhere. Based on this plot LY298 would be a PAM-antagonist (see for example Figure 1, Grundmann et al. 2021, 10.3390/ijms22041763). This would be an important aspect, which would need to be addressed.

We apologise, but the reviewer has misinterpreted these data. The experiments in figure 2A describe a three-way interaction between a fixed concentration of radiolabeled orthosteric antagonist, [³H]-NMS, and increasing concentrations of unlabeled orthosteric agonist (ACh) and allosteric modulator, LY298. What is actually occurring, therefore, is that the reduced overall specific binding ‘window’ is due to negative cooperativity between LY298 and the radiolabel, [³H]NMS (because the antagonist prefers the inactive state, whereas LY298 prefers the active state). At the same time, the potency of the ACh competition curve is actually increasing because the modulator is increasing ACh affinity (positive cooperativity) i.e., it is not ‘blocking the overall binding of ACh’ but, rather, it is actually blocking the overall binding of [³H]NMS. This is an example of ‘probe dependence’, whereby the same modulator can have a different allosteric effect depending on the orthosteric ligand (e.g., antagonist vs agonist). Thus, LY298 is not a PAM-antagonist; it is a NAM of [³H]NMS and a PAM of ACh; a true ‘PAM-antagonist’ is a modulator that increases agonist affinity while concomitantly reducing agonist signaling efficacy (e.g., see Price et al., 2005, Mol. Pharm. 67: 1484), which is not the case here (indeed, binding assays alone cannot be used to determine whether a compound is a PAM-antagonist). We have added the following line to the main text:

“A probe-dependent effect was also observed with the radioligand, [3H]-NMS, evidenced by a reduction in specific radioligand binding due to negative cooperativity between the antagonist probe and LY298, which has been previously reported (Chan et al., 2008; Leach et al., 2010; Suratman et al., 2011; Thal et al., 2016).”

8. I would suggest placing some of the MD simulation traces into the supplementary materials, as these currently take up a large fraction of all figures. Alternatively, different complexes could be color coded and overlaid in one figure to highlight differences.

We agree and have moved some of the MD simulation traces to supplemental (e.g. Figure 3 —figure supplement 1; figure 4 —figure supplement 1).

9. Figure 4 B and H: given the structure in Figure 4 H, it means that the binding pocket around Ipx leaves no room for any movement. Overall, I am not able to follow the discussion regarding the alkyne group/linker of Ipx not being visible. If the start and end points are fixed (visible densities), and have a rigid, planar triple bond involved (alkyne), I find it hard to imagine that the linker is flexible enough to wash out the signal for the linker. Also, the representative Figure S3F is not very convincing, as (A) all iperoxo densities seem to be of rather poor quality and (B) the average of all aligned structures would still likely result in an 'average-able' linker density. I would suggest either elaborating on or omitting this claim.

We apologise for the ambiguity here. The trimethyl ammonium ion is making a cation-pi interaction with nearby residues. The relative orientation of the “ion” does not need to be specific relative to the receptor, as the interaction is driven by the position of the charge, which overlays nicely across the structures. There is enough room in the active site for the alkyne bond to rotate circularly around the position of the ion (similar to a ‘crankshaft’ rotating) and the effect of this would be a blurring of the density around this bond. Nevertheless, we acknowledge the esoteric nature of this point and that it is not necessarily relevant to the main text; as such we have removed most of this statement from the text and supplemental figure.

10. Line 301, The predominate χ2 angle of W413 was approximately 60◦ and 105◦ in the ACh-bound and Ipx-bound simulations, respectively, corresponding to the cryo-EM conformations. As depicted in Figure 4L, W413 when bound to ACh also samples angles close to 90 degrees or higher in the majority of Sim2 and part of Sim1. What is the significance of this conformational sampling?

Ipx stabilizes the active conformation/rotamer of W413 (~105◦) by directly interacting with the residue, whereas ACh is too small to directly interact with W413. As such, the conformation of W413 matches the inactive conformation (60◦) that was observed in the tiotropium-bound structure. As commented, during GaMD simulations with ACh, the conformation of W413 appears to fluctuate between the inactive and active conformations. One may speculate that this could relate to differences in efficacy between ACh and Ipx, as the residue is known to be important for biased signaling at some GPCRs (Cong et al., Molecular Cell 2021). Indeed, we have studied residue W413 in more detail in pERK1/2 signalling assays and the G_i1 TruPath assay –data that was not included in the original submitted version of this manuscript, but now added back to a new section: “Structural insights into allosteric modulation of agonist signaling” as we believe that these data are important to the overall manuscript.

11. On the note of ligands having flexible parts in VU154, and therefore no resolved densities in the maps, is there any pharmacological evidence (i.e. SAR) for these regions not contributing to binding/signaling? Analogously, are there any SAR data for replacing the alkyne bond in Ipx? It would be conceivable that a rigid replacement linker would affect (either positively or negatively) receptor binding and signaling.

We have not thoroughly investigated the SAR around VU154 and other M4-PAMs. We agree that this is an interesting idea, but beyond the scope of the current manuscript. Instead, we looked at the mouse receptor, where we knew VU154 had better PAM-agonist activity. It is plausible that the unresolved / flexible region of VU154, which is negatively charged, is able to interact with an Arg residue that is present at the mouse receptor and not at the human receptor (Figure 7) – but this would require a cryo-EM structure with the mouse receptor for proper validation. With respect to Ipx, we note that it is one of the most potent agonists discovered for mAChRs to date (Schrage et al. BJP 2013) and has ‘supraphysiological’ activity, i.e., higher intrinsic efficacy than the cognate agonist, ACh. As such, we would expect analogs that removed the linker to have lower potency and binding, but we have not investigated this to date.

12. According to the GaMD simulation results in Figure S7C, the minimum distance for the T433 and VU154 seems to be close to 4 Å while the predominant distance is around 7 Å. Therefore, I am wondering what the significance of this hydrogen bond is observed in the cryoEM structure. Additionally, T433R mutant showed increased binding to VU154 and showed the importance of T433 in species selectivity. Is this increased binding of VU154 with T433R mutant a result of a more "stable" hydrogen bond between receptor and VU154?

We apologize for not being clear here. We meant to imply there were potential interactions between VU154 and residue T433 based on inspection of the structure. However, during GaMD simulations these distances were too far apart and highly fluctuating suggesting the interaction would be transient at best. We have amended the manuscript as such:

“For VU154, there were two additional hydrogen bonding interactions were possible with residues Y922.64 and T4337.33 (Figure 4G); however, these interactions were highly fluctuating during GaMD simulations suggesting they were at best transient interactions (Figure 4 —figure supplement 1I,J).

With respect to the T433R mutant we apologize for not including that data. GaMD simulations suggest the interaction between residue R433 and the sulfoxide group of VU154 is more stable (5.2 ± 1.5 Å) vs the WT residue T433 (6.56 ± 2.1 Å), but the interaction distance is still a bit far. Ultimately, this is an interaction that would be better confirmed by structure as suggested above. The GaMD data for T433R has been added to X and included in the text as:

“The GaMD simulations also suggest that a potential interaction between the mutant residue T433R and the sulfoxide group of VU154 were more stable (5.2 ± 1.5 Å; Figure 7 —figure supplement 3I) vs the WT residue T433 (6.56 ± 2.1 Å, Figure 4 —figure supplement 1J), albeit the distance of this interaction far apart and would be better validated by structure determination of VU154 with the mouse M4 mAChR.”

13. Regarding the species selectivity aspect, there is no mention of the V91L mutant in Figure 7, only as part of the triple mutant. It is hard to judge which mutations are responsible for species selectivity without either showing results for the D432E/T433R double mutant, or the additional V91L single mutant.

We do apologize for not having data regarding either V91L or a D432E/T433R double mutant. This is a limitation of the study that we highlight in the manuscript text. However, V91 does not face the allosteric site and note that the V to L mutation is conserved; we thus do not expect this residue to have a strong role in species selectivity. Hence, the biggest contributors are likely to be D432E and T433R.

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Citations

Vuckovic Z, Mobbs JI, Belousoff MJ, Glukhova A, Sexton PM, Danev R, Thal DM. 2023. Human M4 muscarinic acetylcholine receptor complex with Gi1 and the agonist iperoxo. RCSB Protein Data Bank. 7TRK
Vuckovic Z, Mobbs JI, Belousoff MJ, Glukhova A, Sexton PM, Danev R, Thal DM. 2023. Human M4 muscarinic acetylcholine receptor complex with Gi1 and the agonist iperoxo and positive allosteric modulator LY2033298. RCSB Protein Data Bank. 7TRP
Vuckovic Z, Mobbs JI, Belousoff MJ, Glukhova A, Sexton PM, Danev R, Thal DM. 2023. Human M4 muscarinic acetylcholine receptor complex with Gi1 and the agonist iperoxo and positive allosteric modulator VU0467154. RCSB Protein Data Bank. 7TRQ
Vuckovic Z, Mobbs JI, Belousoff MJ, Glukhova A, Sexton PM, Danev R, Thal DM. 2023. Human M4 muscarinic acetylcholine receptor complex with Gi1 and the endogenous agonist acetylcholine. RCSB Protein Data Bank. 7TRS
Vuckovic Z, Mobbs JI, Belousoff MJ, Glukhova A, Sexton PM, Danev R, Thal DM. 2023. Human M4 muscarinic acetylcholine receptor complex with Gi1 and the agonist iperoxo. Electron Microscopy Data Bank. EMD-26099
Vuckovic Z, Mobbs JI, Belousoff MJ, Glukhova A, Sexton PM, Danev R, Thal DM. 2023. Human M4 muscarinic acetylcholine receptor complex with Gi1 and the agonist iperoxo and positive allosteric modulator LY2033298. Electron Microscopy Data Bank. EMD-26100
Vuckovic Z, Mobbs JI, Belousoff MJ, Glukhova A, Sexton PM, Danev R, Thanl DM. 2023. Human M4 muscarinic acetylcholine receptor complex with Gi1 and the agonist iperoxo and positive allosteric modulator VU0467154. Electron Microscopy Data Bank. EMD-26101
Vuckovic Z, Mobbs JI, Belousoff MJ, Glukhova A, Sexton PM, Danev R, Thal DM. 2023. Human M4 muscarinic acetylcholine receptor complex with Gi1 and the endogenous agonist acetylcholine. Electron Microscopy Data Bank. EMD-26102

Supplementary Materials

Figure 1—source data 1. Related to Figure 1D–K.

elife-83477-fig1-data1.xlsx^{(12.4KB, xlsx)}

Figure 1—figure supplement 1—source data 1. Related to Figure 1—figure supplement 1.

elife-83477-fig1-figsupp1-data1.xlsx^{(26KB, xlsx)}

Figure 1—figure supplement 2—source data 1. Related to Figure 1—figure supplement 2.

elife-83477-fig1-figsupp2-data1.xlsx^{(18.9KB, xlsx)}

Figure 4—figure supplement 2—source data 1. Related to Figure 4—figure supplement 2.

elife-83477-fig4-figsupp2-data1.xlsx^{(30.8KB, xlsx)}

Figure 7—source data 1. Related to Figure 7.

elife-83477-fig7-data1.xlsx^{(11KB, xlsx)}

Figure 7—figure supplement 1—source data 1. Related to Figure 7—figure supplement 1.

elife-83477-fig7-figsupp1-data1.xlsx^{(31.6KB, xlsx)}

Figure 7—figure supplement 2—source data 1. Related to Figure 7—figure supplement 2.

elife-83477-fig7-figsupp2-data1.xlsx^{(26KB, xlsx)}

MDAR checklist

elife-83477-mdarchecklist1.pdf^{(195.9KB, pdf)}

Data Availability Statement

The following datasets were generated:

Vuckovic Z, Mobbs JI, Belousoff MJ, Glukhova A, Sexton PM, Danev R, Thal DM. 2023. Human M4 muscarinic acetylcholine receptor complex with Gi1 and the agonist iperoxo. RCSB Protein Data Bank. 7TRK

[bib1] Ballesteros JA, Weinstein H. In: Methods in Neurosciences, Receptor Molecular Biology. Sealfon SC, editor. Academic Press; 1995. Integrated methods for the construction of three-dimensional models and computational probing of structure-function relations in G protein-coupled receptors; pp. 366–428. [DOI] [Google Scholar]

[bib2] Ballesteros JA, Jensen AD, Liapakis G, Rasmussen SG, Shi L, Gether U, Javitch JA. Activation of the beta 2-adrenergic receptor involves disruption of an ionic lock between the cytoplasmic ends of transmembrane segments 3 and 6. The Journal of Biological Chemistry. 2001;276:29171–29177. doi: 10.1074/jbc.M103747200. [DOI] [PubMed] [Google Scholar]

[bib3] Berizzi AE, Gentry PR, Rueda P, Den Hoedt S, Sexton PM, Langmead CJ, Christopoulos A. Molecular mechanisms of action of M5 muscarinic acetylcholine receptor allosteric modulators. Molecular Pharmacology. 2016;90:427–436. doi: 10.1124/mol.116.104182. [DOI] [PubMed] [Google Scholar]

[bib4] Berman H, Henrick K, Nakamura H. Announcing the worldwide protein data bank. Nature Structural Biology. 2003;10:980. doi: 10.1038/nsb1203-980. [DOI] [PubMed] [Google Scholar]

[bib5] Bhattarai A, Miao Y. Gaussian accelerated molecular dynamics for elucidation of drug pathways. Expert Opinion on Drug Discovery. 2018;13:1055–1065. doi: 10.1080/17460441.2018.1538207. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib6] Black JW, Leff P. Operational models of pharmacological agonism. Proceedings of the Royal Society of London. Series B, Biological Sciences. 1983;220:141–162. doi: 10.1098/rspb.1983.0093. [DOI] [PubMed] [Google Scholar]

[bib7] Brown AJH, Bradley SJ, Marshall FH, Brown GA, Bennett KA, Brown J, Cansfield JE, Cross DM, de Graaf C, Hudson BD, Dwomoh L, Dias JM, Errey JC, Hurrell E, Liptrot J, Mattedi G, Molloy C, Nathan PJ, Okrasa K, Osborne G, Patel JC, Pickworth M, Robertson N, Shahabi S, Bundgaard C, Phillips K, Broad LM, Goonawardena AV, Morairty SR, Browning M, Perini F, Dawson GR, Deakin JFW, Smith RT, Sexton PM, Warneck J, Vinson M, Tasker T, Tehan BG, Teobald B, Christopoulos A, Langmead CJ, Jazayeri A, Cooke RM, Rucktooa P, Congreve MS, Weir M, Tobin AB. From structure to clinic: design of a muscarinic M1 receptor agonist with the potential to treat Alzheimer’s disease. Cell. 2021;184:5886–5901. doi: 10.1016/j.cell.2021.11.001. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib8] Bubser M, Bridges TM, Dencker D, Gould RW, Grannan M, Noetzel MJ, Lamsal A, Niswender CM, Daniels JS, Poslusney MS, Melancon BJ, Tarr JC, Byers FW, Wess J, Duggan ME, Dunlop J, Wood MW, Brandon NJ, Wood MR, Lindsley CW, Conn PJ, Jones CK. Selective activation of M4 muscarinic acetylcholine receptors reverses MK-801-induced behavioral impairments and enhances associative learning in rodents. ACS Chemical Neuroscience. 2014;5:920–942. doi: 10.1021/cn500128b. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib9] Bumbak F, Thomas T, Noonan-Williams BJ, Vaid TM, Yan F, Whitehead AR, Bruell S, Kocan M, Tan X, Johnson MA, Bathgate RAD, Chalmers DK, Gooley PR, Scott DJ. Conformational changes in tyrosine 11 of neurotensin are required to activate the neurotensin receptor 1. ACS Pharmacology & Translational Science. 2020;3:690–705. doi: 10.1021/acsptsci.0c00026. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib10] Burger WAC, Sexton PM, Christopoulos A, Thal DM. Toward an understanding of the structural basis of allostery in muscarinic acetylcholine receptors. The Journal of General Physiology. 2018;150:1360–1372. doi: 10.1085/jgp.201711979. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib11] Bymaster FP, Carter PA, Yamada M, Gomeza J, Wess J, Hamilton SE, Nathanson NM, McKinzie DL, Felder CC. Role of specific muscarinic receptor subtypes in cholinergic parasympathomimetic responses, in vivo phosphoinositide hydrolysis, and pilocarpine-induced seizure activity. The European Journal of Neuroscience. 2003;17:1403–1410. doi: 10.1046/j.1460-9568.2003.02588.x. [DOI] [PubMed] [Google Scholar]

[bib12] Canals M, Sexton PM, Christopoulos A. Allostery in GPCRs: `` MWC'' revisited. Trends in Biochemical Sciences. 2011;36:663–672. doi: 10.1016/j.tibs.2011.08.005. [DOI] [PubMed] [Google Scholar]

[bib13] Cao AM, Quast RB, Fatemi F, Rondard P, Pin JP, Margeat E. Allosteric modulators enhance agonist efficacy by increasing the residence time of a GPCR in the active state. Nature Communications. 2021;12:5426. doi: 10.1038/s41467-021-25620-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib14] Cary BP, Deganutti G, Zhao P, Truong TT, Piper SJ, Liu X, Belousoff MJ, Danev R, Sexton PM, Wootten D, Gellman SH. Structural and functional diversity among agonist-bound states of the GLP-1 receptor. Nature Chemical Biology. 2022;18:256–263. doi: 10.1038/s41589-021-00945-w. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib15] Casañal A, Lohkamp B, Emsley P. Current developments in coot for macromolecular model building of electron cryo-microscopy and crystallographic data. Protein Science. 2020;29:1069–1078. doi: 10.1002/pro.3791. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib16] Chan WY, McKinzie DL, Bose S, Mitchell SN, Witkin JM, Thompson RC, Christopoulos A, Lazareno S, Birdsall NJM, Bymaster FP, Felder CC. Allosteric modulation of the muscarinic M4 receptor as an approach to treating schizophrenia. PNAS. 2008;105:10978–10983. doi: 10.1073/pnas.0800567105. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib17] Changeux JP, Christopoulos A. Allosteric modulation as a unifying mechanism for receptor function and regulation. Cell. 2016;166:1084–1102. doi: 10.1016/j.cell.2016.08.015. [DOI] [PubMed] [Google Scholar]

[bib18] Christopoulos A, Kenakin T. G protein-coupled receptor allosterism and complexing. Pharmacological Reviews. 2002;54:323–374. doi: 10.1124/pr.54.2.323. [DOI] [PubMed] [Google Scholar]

[bib19] Christopoulos A. Advances in G protein-coupled receptor allostery: from function to structure. Molecular Pharmacology. 2014;86:463–478. doi: 10.1124/mol.114.094342. [DOI] [PubMed] [Google Scholar]

[bib20] Cong X, Maurel D, Déméné H, Vasiliauskaité-Brooks I, Hagelberger J, Peysson F, Saint-Paul J, Golebiowski J, Granier S, Sounier R. Molecular insights into the biased signaling mechanism of the μ-opioid receptor. Molecular Cell. 2021;81:4165–4175. doi: 10.1016/j.molcel.2021.07.033. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib21] Conn PJ, Jones CK, Lindsley CW. Subtype-Selective allosteric modulators of muscarinic receptors for the treatment of CNS disorders. Trends in Pharmacological Sciences. 2009;30:148–155. doi: 10.1016/j.tips.2008.12.002. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib22] Croll TI. ISOLDE: a physically realistic environment for model building into low-resolution electron-density maps. Acta Crystallographica. Section D, Structural Biology. 2018;74:519–530. doi: 10.1107/S2059798318002425. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib23] Danev R, Belousoff M, Liang YL, Zhang X, Eisenstein F, Wootten D, Sexton PM. Routine sub-2.5 Å cryo-EM structure determination of GPCRs. Nature Communications. 2021;12:4333. doi: 10.1038/s41467-021-24650-3. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib24] Darden T, York D, Pedersen L. Particle mesh ewald: an N⋅log (N) method for ewald sums in large systems. The Journal of Chemical Physics. 1993;98:10089–10092. doi: 10.1063/1.464397. [DOI] [Google Scholar]

[bib25] Deganutti G, Liang YL, Zhang X, Khoshouei M, Clydesdale L, Belousoff MJ, Venugopal H, Truong TT, Glukhova A, Keller AN, Gregory KJ, Leach K, Christopoulos A, Danev R, Reynolds CA, Zhao P, Sexton PM, Wootten D. Dynamics of GLP-1R peptide agonist engagement are correlated with kinetics of G protein activation. Nature Communications. 2022;13:92. doi: 10.1038/s41467-021-27760-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib26] Dencker D, Wörtwein G, Weikop P, Jeon J, Thomsen M, Sager TN, Mørk A, Woldbye DPD, Wess J, Fink-Jensen A. Involvement of a subpopulation of neuronal M4 muscarinic acetylcholine receptors in the antipsychotic-like effects of the M1/M4 preferring muscarinic receptor agonist xanomeline. The Journal of Neuroscience. 2011;31:5905–5908. doi: 10.1523/JNEUROSCI.0370-11.2011. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib27] DeVree BT, Mahoney JP, Vélez-Ruiz GA, Rasmussen SGF, Kuszak AJ, Edwald E, Fung JJ, Manglik A, Masureel M, Du Y, Matt RA, Pardon E, Steyaert J, Kobilka BK, Sunahara RK. Allosteric coupling from G protein to the agonist-binding pocket in GPCRs. Nature. 2016;535:182–186. doi: 10.1038/nature18324. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib28] Draper-Joyce CJ, Bhola R, Wang J, Bhattarai A, Nguyen ATN, Cowie-Kent I, O’Sullivan K, Chia LY, Venugopal H, Valant C, Thal DM, Wootten D, Panel N, Carlsson J, Christie MJ, White PJ, Scammells P, May LT, Sexton PM, Danev R, Miao Y, Glukhova A, Imlach WL, Christopoulos A. Positive allosteric mechanisms of adenosine A1 receptor-mediated analgesia. Nature. 2021;597:571–576. doi: 10.1038/s41586-021-03897-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib29] Dror RO, Arlow DH, Maragakis P, Mildorf TJ, Pan AC, Xu H, Borhani DW, Shaw DE. Activation mechanism of the β2-adrenergic receptor. PNAS. 2011;108:18684–18689. doi: 10.1073/pnas.1110499108. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib30] Dror RO, Mildorf TJ, Hilger D, Manglik A, Borhani DW, Arlow DH, Philippsen A, Villanueva N, Yang Z, Lerch MT, Hubbell WL, Kobilka BK, Sunahara RK, Shaw DE. Signal transduction. Structural basis for nucleotide exchange in heterotrimeric G proteins. Science. 2015;348:1361–1365. doi: 10.1126/science.aaa5264. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib31] Ehlert FJ. Estimation of the affinities of allosteric ligands using radioligand binding and pharmacological null methods. Molecular Pharmacology. 1988;33:187–194. [PubMed] [Google Scholar]

[bib32] Foster DJ, Wilson JM, Remke DH, Mahmood MS, Uddin MJ, Wess J, Patel S, Marnett LJ, Niswender CM, Jones CK, Xiang Z, Lindsley CW, Rook JM, Conn PJ. Antipsychotic-Like effects of M 4 positive allosteric modulators are mediated by CB 2 receptor-dependent inhibition of dopamine release. Neuron. 2016;91:1244–1252. doi: 10.1016/j.neuron.2016.08.017. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib33] Fritze O, Filipek S, Kuksa V, Palczewski K, Hofmann KP, Ernst OP. Role of the conserved NPXXY (X) 5,6F motif in the rhodopsin ground state and during activation. PNAS. 2003;100:2290–2295. doi: 10.1073/pnas.0435715100. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib34] Furness SGB, Liang YL, Nowell CJ, Halls ML, Wookey PJ, Dal Maso E, Inoue A, Christopoulos A, Wootten D, Sexton PM. Ligand-Dependent modulation of G protein conformation alters drug efficacy. Cell. 2016;167:739–749. doi: 10.1016/j.cell.2016.09.021. [DOI] [PubMed] [Google Scholar]

[bib35] Gentry PR, Kokubo M, Bridges TM, Noetzel MJ, Cho HP, Lamsal A, Smith E, Chase P, Hodder PS, Niswender CM, Daniels JS, Conn PJ, Lindsley CW, Wood MR. Development of a highly potent, novel M5 positive allosteric modulator (PAM) demonstrating CNS exposure: 1- ((1H-indazol-5-yl) sulfoneyl) -N-ethyl-N- (2- (trifluoromethyl) benzyl) piperidine-4-carboxamide (ML380) Journal of Medicinal Chemistry. 2014;57:7804–7810. doi: 10.1021/jm500995y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib36] Gentry PR, Sexton PM, Christopoulos A. Novel allosteric modulators of G protein-coupled receptors. The Journal of Biological Chemistry. 2015;290:19478–19488. doi: 10.1074/jbc.R115.662759. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib37] Gould RW, Nedelcovych MT, Gong X, Tsai E, Bubser M, Bridges TM, Wood MR, Duggan ME, Brandon NJ, Dunlop J, Wood MW, Ivarsson M, Noetzel MJ, Daniels JS, Niswender CM, Lindsley CW, Conn PJ, Jones CK. State-Dependent alterations in sleep/wake architecture elicited by the M4 PAM VU0467154-relation to antipsychotic-like drug effects. Neuropharmacology. 2016;102:244–253. doi: 10.1016/j.neuropharm.2015.11.016. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib38] Gregorio GG, Masureel M, Hilger D, Terry DS, Juette M, Zhao H, Zhou Z, Perez-Aguilar JM, Hauge M, Mathiasen S, Javitch JA, Weinstein H, Kobilka BK, Blanchard SC. Single-Molecule analysis of ligand efficacy in β2ar-G-protein activation. Nature. 2017;547:68–73. doi: 10.1038/nature22354. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib39] Gregory KJ, Hall NE, Tobin AB, Sexton PM, Christopoulos A. Identification of orthosteric and allosteric site mutations in M2 muscarinic acetylcholine receptors that contribute to ligand-selective signaling bias. The Journal of Biological Chemistry. 2010;285:7459–7474. doi: 10.1074/jbc.M109.094011. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib40] Haga K, Kruse AC, Asada H, Yurugi-Kobayashi T, Shiroishi M, Zhang C, Weis WI, Okada T, Kobilka BK, Haga T, Kobayashi T. Structure of the human M2 muscarinic acetylcholine receptor bound to an antagonist. Nature. 2012;482:547–551. doi: 10.1038/nature10753. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib41] Huang J, Rauscher S, Nawrocki G, Ran T, Feig M, de Groot BL, Grubmüller H, MacKerell AD., Jr CHARMM36m: an improved force field for folded and intrinsically disordered proteins. Nature Methods. 2017;14:71–73. doi: 10.1038/nmeth.4067. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib42] Huang SK, Pandey A, Tran DP, Villanueva NL, Kitao A, Sunahara RK, Sljoka A, Prosser RS. Delineating the conformational landscape of the adenosine A2A receptor during G protein coupling. Cell. 2021;184:1884–1894. doi: 10.1016/j.cell.2021.02.041. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib43] Hulme EC. Gpcr activation: a mutagenic spotlight on crystal structures. Trends in Pharmacological Sciences. 2013;34:67–84. doi: 10.1016/j.tips.2012.11.002. [DOI] [PubMed] [Google Scholar]

[bib44] Humphrey W, Dalke A, Schulten K. VMD: visual molecular dynamics. Journal of Molecular Graphics. 1996;14:33–38. doi: 10.1016/0263-7855(96)00018-5. [DOI] [PubMed] [Google Scholar]

[bib45] Jones EM, Lubock NB, Venkatakrishnan AJ, Wang J, Tseng AM, Paggi JM, Latorraca NR, Cancilla D, Satyadi M, Davis JE, Babu MM, Dror RO, Kosuri S. Structural and functional characterization of G protein-coupled receptors with deep mutational scanning. eLife. 2020;9:e54895. doi: 10.7554/eLife.54895. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib46] Josephs TM, Belousoff MJ, Liang YL, Piper SJ, Cao J, Garama DJ, Leach K, Gregory KJ, Christopoulos A, Hay DL, Danev R, Wootten D, Sexton PM. Structure and dynamics of the CGRP receptor in apo and peptide-bound forms. Science. 2021;372:eabf7258. doi: 10.1126/science.abf7258. [DOI] [PubMed] [Google Scholar]

[bib47] Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, Tunyasuvunakool K, Bates R, Žídek A, Potapenko A, Bridgland A, Meyer C, Kohl SAA, Ballard AJ, Cowie A, Romera-Paredes B, Nikolov S, Jain R, Adler J, Back T, Petersen S, Reiman D, Clancy E, Zielinski M, Steinegger M, Pacholska M, Berghammer T, Bodenstein S, Silver D, Vinyals O, Senior AW, Kavukcuoglu K, Kohli P, Hassabis D. Highly accurate protein structure prediction with alphafold. Nature. 2021;596:583–589. doi: 10.1038/s41586-021-03819-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib48] Katayama K, Suzuki K, Suno R, Kise R, Tsujimoto H, Iwata S, Inoue A, Kobayashi T, Kandori H. Vibrational spectroscopy analysis of ligand efficacy in human M2 muscarinic acetylcholine receptor (M2R) Communications Biology. 2021;4:1321. doi: 10.1038/s42003-021-02836-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib49] Kenakin T, Onaran O. The ligand paradox between affinity and efficacy: can you be there and not make a difference? Trends in Pharmacological Sciences. 2002;23:275–280. doi: 10.1016/S0165-6147(02)02036-9. [DOI] [PubMed] [Google Scholar]

[bib50] Kenakin T. New concepts in drug discovery: collateral efficacy and permissive antagonism. Nature Reviews. Drug Discovery. 2005;4:919–927. doi: 10.1038/nrd1875. [DOI] [PubMed] [Google Scholar]

[bib51] Kenakin TP. Biased signalling and allosteric machines: new vistas and challenges for drug discovery. British Journal of Pharmacology. 2012;165:1659–1669. doi: 10.1111/j.1476-5381.2011.01749.x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib52] Kenakin TP, Watson C, Muniz-Medina V, Christopoulos A, Novick S. A simple method for quantifying functional selectivity and agonist bias. ACS Chemical Neuroscience. 2012;3:193–203. doi: 10.1021/cn200111m. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib53] Kenakin T, Christopoulos A. Signalling bias in new drug discovery: detection, quantification and therapeutic impact. Nature Reviews. Drug Discovery. 2013;12:205–216. doi: 10.1038/nrd3954. [DOI] [PubMed] [Google Scholar]

[bib54] Klauda JB, Venable RM, Freites JA, O’Connor JW, Tobias DJ, Mondragon-Ramirez C, Vorobyov I, MacKerell AD, Jr, Pastor RW. Update of the CHARMM all-atom additive force field for lipids: validation on six lipid types. The Journal of Physical Chemistry. B. 2010;114:7830–7843. doi: 10.1021/jp101759q. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib55] Kruse AC, Hu J, Pan AC, Arlow DH, Rosenbaum DM, Rosemond E, Green HF, Liu T, Chae PS, Dror RO, Shaw DE, Weis WI, Wess J, Kobilka BK. Structure and dynamics of the M3 muscarinic acetylcholine receptor. Nature. 2012;482:552–556. doi: 10.1038/nature10867. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib56] Kruse AC, Ring AM, Manglik A, Hu J, Hu K, Eitel K, Hübner H, Pardon E, Valant C, Sexton PM, Christopoulos A, Felder CC, Gmeiner P, Steyaert J, Weis WI, Garcia KC, Wess J, Kobilka BK. Activation and allosteric modulation of a muscarinic acetylcholine receptor. Nature. 2013;504:101–106. doi: 10.1038/nature12735. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib57] Kumar A, Yoluk O, MacKerell AD. FFParam: standalone package for CHARMM additive and drude polarizable force field parametrization of small molecules. Journal of Computational Chemistry. 2020;41:958–970. doi: 10.1002/jcc.26138. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib58] Leach K, Sexton PM, Christopoulos A. Allosteric GPCR modulators: taking advantage of permissive receptor pharmacology. Trends in Pharmacological Sciences. 2007;28:382–389. doi: 10.1016/j.tips.2007.06.004. [DOI] [PubMed] [Google Scholar]

[bib59] Leach K, Loiacono RE, Felder CC, McKinzie DL, Mogg A, Shaw DB, Sexton PM, Christopoulos A. Molecular mechanisms of action and in vivo validation of an M4 muscarinic acetylcholine receptor allosteric modulator with potential antipsychotic properties. Neuropsychopharmacology. 2010;35:855–869. doi: 10.1038/npp.2009.194. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib60] Leach K, Davey AE, Felder CC, Sexton PM, Christopoulos A. The role of transmembrane domain 3 in the actions of orthosteric, allosteric, and atypical agonists of the M4 muscarinic acetylcholine receptor. Molecular Pharmacology. 2011;79:855–865. doi: 10.1124/mol.111.070938. [DOI] [PubMed] [Google Scholar]

[bib61] Liang YL, Khoshouei M, Glukhova A, Furness SGB, Zhao P, Clydesdale L, Koole C, Truong TT, Thal DM, Lei S, Radjainia M, Danev R, Baumeister W, Wang MW, Miller LJ, Christopoulos A, Sexton PM, Wootten D. Phase-plate cryo-EM structure of a biased agonist-bound human GLP-1 receptor–gs complex. Nature. 2018a;555:121–125. doi: 10.1038/nature25773. [DOI] [PubMed] [Google Scholar]

[bib62] Liang YL, Zhao P, Draper-Joyce C, Baltos JA, Glukhova A, Truong TT, May LT, Christopoulos A, Wootten D, Sexton PM, Furness SGB. Dominant negative G proteins enhance formation and purification of agonist-GPCR-G protein complexes for structure determination. ACS Pharmacology & Translational Science. 2018b;1:12–20. doi: 10.1021/acsptsci.8b00017. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib63] Liang YL, Belousoff MJ, Fletcher MM, Zhang X, Khoshouei M, Deganutti G, Koole C, Furness SGB, Miller LJ, Hay DL, Christopoulos A, Reynolds CA, Danev R, Wootten D, Sexton PM. Structure and dynamics of adrenomedullin receptors AM1 and AM2 reveal key mechanisms in the control of receptor phenotype by receptor activity-modifying proteins. ACS Pharmacology & Translational Science. 2020;3:263–284. doi: 10.1021/acsptsci.9b00080. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib64] Liebschner D, Afonine PV, Baker ML, Bunkóczi G, Chen VB, Croll TI, Hintze B, Hung LW, Jain S, McCoy AJ, Moriarty NW, Oeffner RD, Poon BK, Prisant MG, Read RJ, Richardson JS, Richardson DC, Sammito MD, Sobolev OV, Stockwell DH, Terwilliger TC, Urzhumtsev AG, Videau LL, Williams CJ, Adams PD. Macromolecular structure determination using x-rays, neutrons and electrons: recent developments in phenix. Acta Crystallographica Section D Structural Biology. 2019;75:861–877. doi: 10.1107/S2059798319011471. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib65] Liu W, Chun E, Thompson AA, Chubukov P, Xu F, Katritch V, Han GW, Roth CB, Heitman LH, IJzerman AP, Cherezov V, Stevens RC. Structural basis for allosteric regulation of GPCRs by sodium ions. Science. 2012a;337:232–236. doi: 10.1126/science.1219218. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib66] Liu JJ, Horst R, Katritch V, Stevens RC, Wüthrich K. Biased signaling pathways in β2-adrenergic receptor characterized by 19F-NMR. Science. 2012b;335:1106–1110. doi: 10.1126/science.1215802. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib67] Maeda S, Koehl A, Matile H, Hu H, Hilger D, Schertler GFX, Manglik A, Skiniotis G, Dawson RJP, Kobilka BK. Development of an antibody fragment that stabilizes GPCR/G-protein complexes. Nature Communications. 2018;9:3712. doi: 10.1038/s41467-018-06002-w. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib68] Maeda S, Qu Q, Robertson MJ, Skiniotis G, Kobilka BK. Structures of the M1 and M2 muscarinic acetylcholine receptor/G-protein complexes. Science. 2019;364:552–557. doi: 10.1126/science.aaw5188. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib69] Manglik A, Kim TH, Masureel M, Altenbach C, Yang Z, Hilger D, Lerch MT, Kobilka TS, Thian FS, Hubbell WL, Prosser RS, Kobilka BK. Structural insights into the dynamic process of β2-adrenergic receptor signaling. Cell. 2015;161:1101–1111. doi: 10.1016/j.cell.2015.04.043. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib70] Marlo JE, Niswender CM, Days EL, Bridges TM, Xiang Y, Rodriguez AL, Shirey JK, Brady AE, Nalywajko T, Luo Q, Austin CA, Williams MB, Kim K, Williams R, Orton D, Brown HA, Lindsley CW, Weaver CD, Conn PJ. Discovery and characterization of novel allosteric potentiators of M1 muscarinic receptors reveals multiple modes of activity. Molecular Pharmacology. 2009;75:577–588. doi: 10.1124/mol.108.052886. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib71] Masureel M, Zou Y, Picard L-P, van der Westhuizen E, Mahoney JP, Rodrigues JPGLM, Mildorf TJ, Dror RO, Shaw DE, Bouvier M, Pardon E, Steyaert J, Sunahara RK, Weis WI, Zhang C, Kobilka BK. Structural insights into binding specificity, efficacy and bias of a β2AR partial agonist. Nature Chemical Biology. 2018;14:1059–1066. doi: 10.1038/s41589-018-0145-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib72] McCorvy JD, Wacker D, Wang S, Agegnehu B, Liu J, Lansu K, Tribo AR, Olsen RHJ, Che T, Jin J, Roth BL. Structural determinants of 5-HT2B receptor activation and biased agonism. Nature Structural & Molecular Biology. 2018;25:787–796. doi: 10.1038/s41594-018-0116-7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib73] Miao Y, Sinko W, Pierce L, Bucher D, Walker RC, McCammon JA. Improved reweighting of accelerated molecular dynamics simulations for free energy calculation. Journal of Chemical Theory and Computation. 2014;10:2677–2689. doi: 10.1021/ct500090q. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib74] Miao Y, Feher VA, McCammon JA. Gaussian accelerated molecular dynamics: unconstrained enhanced sampling and free energy calculation. Journal of Chemical Theory and Computation. 2015;11:3584–3595. doi: 10.1021/acs.jctc.5b00436. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib75] Miao Y, McCammon JA. Graded activation and free energy landscapes of a muscarinic G-protein–coupled receptor. PNAS. 2016;113:12162–12167. doi: 10.1073/pnas.1614538113. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib76] Miao Y, McCammon JA. Gaussian accelerated molecular dynamics: theory, implementation, and applications. Annual Reports in Computational Chemistry. 2017;13:231–278. doi: 10.1016/bs.arcc.2017.06.005. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib77] Miao Y, McCammon JA. Mechanism of the G-protein mimetic nanobody binding to a muscarinic G-protein-coupled receptor. PNAS. 2018;115:3036–3041. doi: 10.1073/pnas.1800756115. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib78] Mobbs JI, Belousoff MJ, Harikumar KG, Piper SJ, Xu X, Furness SGB, Venugopal H, Christopoulos A, Danev R, Wootten D, Thal DM, Miller LJ, Sexton PM. Structures of the human cholecystokinin 1 (CCK1) receptor bound to Gs and Gq mimetic proteins provide insight into mechanisms of G protein selectivity. PLOS Biology. 2021;19:e3001295. doi: 10.1371/journal.pbio.3001295. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib79] Nawaratne V, Leach K, Felder CC, Sexton PM, Christopoulos A. Structural determinants of allosteric agonism and modulation at the M4 muscarinic acetylcholine receptor: identification of ligand-specific and global activation mechanisms. The Journal of Biological Chemistry. 2010;285:19012–19021. doi: 10.1074/jbc.M110.125096. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib80] Nygaard R, Frimurer TM, Holst B, Rosenkilde MM, Schwartz TW. Ligand binding and micro-switches in 7TM receptor structures. Trends in Pharmacological Sciences. 2009;30:249–259. doi: 10.1016/j.tips.2009.02.006. [DOI] [PubMed] [Google Scholar]

[bib81] O’Connor C, White KL, Doncescu N, Didenko T, Roth BL, Czaplicki G, Stevens RC, Wüthrich K, Milon A. Nmr structure and dynamics of the agonist dynorphin peptide bound to the human kappa opioid receptor. PNAS. 2015;112:11852–11857. doi: 10.1073/pnas.1510117112. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib82] Olsen RHJ, DiBerto JF, English JG, Glaudin AM, Krumm BE, Slocum ST, Che T, Gavin AC, McCorvy JD, Roth BL, Strachan RT. TRUPATH, an open-source biosensor platform for interrogating the GPCR transducerome. Nature Chemical Biology. 2020;16:841–849. doi: 10.1038/s41589-020-0535-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib83] Pani B, Ahn S, Rambarat PK, Vege S, Kahsai AW, Liu A, Valan BN, Staus DP, Costa T, Lefkowitz RJ. Unique positive cooperativity between the β-arrestin-biased β-blocker carvedilol and a small molecule positive allosteric modulator of the β2-adrenergic receptor. Molecular Pharmacology. 2021;100:513–525. doi: 10.1124/molpharm.121.000363. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib84] Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, Ferrin TE. UCSF chimera? A visualization system for exploratory research and analysis. Journal of Computational Chemistry. 2004;25:1605–1612. doi: 10.1002/jcc.20084. [DOI] [PubMed] [Google Scholar]

[bib85] Pettersen EF, Goddard TD, Huang CC, Meng EC, Couch GS, Croll TI, Morris JH, Ferrin TE. UCSF chimerax: structure visualization for researchers, educators, and developers. Protein Science. 2021;30:70–82. doi: 10.1002/pro.3943. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib86] Phillips JC, Braun R, Wang W, Gumbart J, Tajkhorshid E, Villa E, Chipot C, Skeel RD, Kalé L, Schulten K. Scalable molecular dynamics with NAMD. Journal of Computational Chemistry. 2005;26:1781–1802. doi: 10.1002/jcc.20289. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib87] Punjani A, Fleet DJ. 3D variability analysis: resolving continuous flexibility and discrete heterogeneity from single particle cryo-EM. Journal of Structural Biology. 2021;213:107702. doi: 10.1016/j.jsb.2021.107702. [DOI] [PubMed] [Google Scholar]

[bib88] Qi X, Liu H, Thompson B, McDonald J, Zhang C, Li X. Cryo-Em structure of oxysterol-bound human smoothened coupled to a heterotrimeric Gi. Nature. 2019;571:279–283. doi: 10.1038/s41586-019-1286-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib89] Rasmussen SGF, Choi HJ, Fung JJ, Pardon E, Casarosa P, Chae PS, DeVree BT, Rosenbaum DM, Thian FS, Kobilka TS, Schnapp A, Konetzki I, Sunahara RK, Gellman SH, Pautsch A, Steyaert J, Weis WI, Kobilka BK. Structure of a nanobody-stabilized active state of the β2 adrenoceptor. Nature. 2011;469:175–180. doi: 10.1038/nature09648. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib90] Ring AM, Manglik A, Kruse AC, Enos MD, Weis WI, Garcia KC, Kobilka BK. Adrenaline-activated structure of β2-adrenoceptor stabilized by an engineered nanobody. Nature. 2013;502:575–579. doi: 10.1038/nature12572. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib91] Roe DR, Cheatham TE. PTRAJ and CPPTRAJ: software for processing and analysis of molecular dynamics trajectory data. Journal of Chemical Theory and Computation. 2013;9:3084–3095. doi: 10.1021/ct400341p. [DOI] [PubMed] [Google Scholar]

[bib92] Ryckaert JP, Ciccotti G, Berendsen HJC. Numerical integration of the Cartesian equations of motion of a system with constraints: molecular dynamics of n-alkanes. Journal of Computational Physics. 1977;23:327–341. doi: 10.1016/0021-9991(77)90098-5. [DOI] [Google Scholar]

[bib93] Schrage R, Seemann WK, Klöckner J, Dallanoce C, Racké K, Kostenis E, De Amici M, Holzgrabe U, Mohr K. Agonists with supraphysiological efficacy at the muscarinic M2 ACh receptor. British Journal of Pharmacology. 2013;169:357–370. doi: 10.1111/bph.12003. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib94] Seyedabadi M, Gharghabi M, Gurevich EV, Gurevich VV. Structural basis of GPCR coupling to distinct signal transducers: implications for biased signaling. Trends in Biochemical Sciences. 2022;47:570–581. doi: 10.1016/j.tibs.2022.03.009. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib95] Shi L, Liapakis G, Xu R, Guarnieri F, Ballesteros JA, Javitch JA. Beta2 adrenergic receptor activation. Modulation of the proline kink in transmembrane 6 by a rotamer toggle switch. The Journal of Biological Chemistry. 2002;277:40989–40996. doi: 10.1074/jbc.M206801200. [DOI] [PubMed] [Google Scholar]

[bib96] Slosky LM, Bai Y, Toth K, Ray C, Rochelle LK, Badea A, Chandrasekhar R, Pogorelov VM, Abraham DM, Atluri N, Peddibhotla S, Hedrick MP, Hershberger P, Maloney P, Yuan H, Li Z, Wetsel WC, Pinkerton AB, Barak LS, Caron MG. Β-Arrestin-Biased allosteric modulator of NTSR1 selectively attenuates addictive behaviors. Cell. 2020;181:1364–1379. doi: 10.1016/j.cell.2020.04.053. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib97] Slosky LM, Caron MG, Barak LS. Biased allosteric modulators: new frontiers in GPCR drug discovery. Trends in Pharmacological Sciences. 2021;42:283–299. doi: 10.1016/j.tips.2020.12.005. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib98] Solt AS, Bostock MJ, Shrestha B, Kumar P, Warne T, Tate CG, Nietlispach D. Insight into partial agonism by observing multiple equilibria for ligand-bound and gs-mimetic nanobody-bound β1-adrenergic receptor. Nature Communications. 2017;8:1795. doi: 10.1038/s41467-017-02008-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib99] Staus DP, Hu H, Robertson MJ, Kleinhenz ALW, Wingler LM, Capel WD, Latorraca NR, Lefkowitz RJ, Skiniotis G. Structure of the M2 muscarinic receptor-β-arrestin complex in a lipid nanodisc. Nature. 2020;579:297–302. doi: 10.1038/s41586-020-1954-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib100] Suratman S, Leach K, Sexton P, Felder C, Loiacono R, Christopoulos A. Impact of species variability and `` probe-dependence'' on the detection and in vivo validation of allosteric modulation at the M4 muscarinic acetylcholine receptor. British Journal of Pharmacology. 2011;162:1659–1670. doi: 10.1111/j.1476-5381.2010.01184.x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib101] Sušac L, Eddy MT, Didenko T, Stevens RC, Wüthrich K. A2A adenosine receptor functional states characterized by 19F-NMR. PNAS. 2018;115:12733–12738. doi: 10.1073/pnas.1813649115. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib102] Terashi G, Wang X, Maddhuri Venkata Subramaniya SR, Tesmer JJG, Kihara D. Residue-wise local quality estimation for protein models from cryo-EM maps. Nature Methods. 2022;19:1116–1125. doi: 10.1038/s41592-022-01574-4. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib103] Thal DM, Sun B, Feng D, Nawaratne V, Leach K, Felder CC, Bures MG, Evans DA, Weis WI, Bachhawat P, Kobilka TS, Sexton PM, Kobilka BK, Christopoulos A. Crystal structures of the M1 and M4 muscarinic acetylcholine receptors. Nature. 2016;531:335–340. doi: 10.1038/nature17188. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib104] Thal DM, Glukhova A, Sexton PM, Christopoulos A. Structural insights into G-protein-coupled receptor allostery. Nature. 2018;559:45–53. doi: 10.1038/s41586-018-0259-z. [DOI] [PubMed] [Google Scholar]

[bib105] Tian W, Chen C, Lei X, Zhao J, Liang J. CASTp 3.0: computed atlas of surface topography of proteins. Nucleic Acids Research. 2018;46:W363–W367. doi: 10.1093/nar/gky473. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib106] Tzavara ET, Bymaster FP, Davis RJ, Wade MR, Perry KW, Wess J, McKinzie DL, Felder C, Nomikos GG. M4 muscarinic receptors regulate the dynamics of cholinergic and dopaminergic neurotransmission: relevance to the pathophysiology and treatment of related CNS pathologies. FASEB Journal. 2004;18:1410–1412. doi: 10.1096/fj.04-1575fje. [DOI] [PubMed] [Google Scholar]

[bib107] Valant C, Felder CC, Sexton PM, Christopoulos A. Probe dependence in the allosteric modulation of a G protein-coupled receptor: implications for detection and validation of allosteric ligand effects. Molecular Pharmacology. 2012;81:41–52. doi: 10.1124/mol.111.074872. [DOI] [PubMed] [Google Scholar]

[bib108] Vanommeslaeghe K, MacKerell AD. Automation of the CHARMM General force field (cgenff) I: bond perception and atom typing. Journal of Chemical Information and Modeling. 2012a;52:3144–3154. doi: 10.1021/ci300363c. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib109] Vanommeslaeghe K, Raman EP, MacKerell AD. Automation of the CHARMM General force field (cgenff) II: assignment of bonded parameters and partial atomic charges. Journal of Chemical Information and Modeling. 2012b;52:3155–3168. doi: 10.1021/ci3003649. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib110] Vanommeslaeghe K, MacKerell AD. CHARMM additive and polarizable force fields for biophysics and computer-aided drug design. Biochimica et Biophysica Acta. 2015;1850:861–871. doi: 10.1016/j.bbagen.2014.08.004. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib111] Varadi M, Anyango S, Deshpande M, Nair S, Natassia C, Yordanova G, Yuan D, Stroe O, Wood G, Laydon A, Žídek A, Green T, Tunyasuvunakool K, Petersen S, Jumper J, Clancy E, Green R, Vora A, Lutfi M, Figurnov M, Cowie A, Hobbs N, Kohli P, Kleywegt G, Birney E, Hassabis D, Velankar S. AlphaFold protein structure database: massively expanding the structural coverage of protein-sequence space with high-accuracy models. Nucleic Acids Research. 2022;50:D439–D444. doi: 10.1093/nar/gkab1061. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib112] Vuckovic Z, Gentry PR, Berizzi AE, Hirata K, Varghese S, Thompson G, Westhuizen ET, Burger WAC, Rahmani R, Valant C, Langmead CJ, Lindsley CW, Baell JB, Tobin AB, Sexton PM, Christopoulos A, Thal DM. Crystal structure of the m5 muscarinic acetylcholine receptor. PNAS. 2019;116:26001–26007. doi: 10.1073/pnas.1914446116. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib113] Wacker D, Wang C, Katritch V, Han GW, Huang XP, Vardy E, McCorvy JD, Jiang Y, Chu M, Siu FY, Liu W, Xu HE, Cherezov V, Roth BL, Stevens RC. Structural features for functional selectivity at serotonin receptors. Science. 2013;340:615–619. doi: 10.1126/science.1232808. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib114] Wagner T, Merino F, Stabrin M, Moriya T, Antoni C, Apelbaum A, Hagel P, Sitsel O, Raisch T, Prumbaum D, Quentin D, Roderer D, Tacke S, Siebolds B, Schubert E, Shaikh TR, Lill P, Gatsogiannis C, Raunser S. SPHIRE-cryolo is a fast and accurate fully automated particle picker for cryo-EM. Communications Biology. 2019;2:218. doi: 10.1038/s42003-019-0437-z. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib115] Wang J, Arantes PR, Bhattarai A, Hsu RV, Pawnikar S, Huang Y-MM, Palermo G, Miao Y. Gaussian accelerated molecular dynamics (gamd): principles and applications. Wiley Interdisciplinary Reviews. Computational Molecular Science. 2021a;11:e1521. doi: 10.1002/wcms.1521. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib116] Wang J, Pani B, Gokhan I, Xiong X, Kahsai AW, Jiang H, Ahn S, Lefkowitz RJ, Rockman HA. Β-Arrestin-Biased allosteric modulator potentiates carvedilol-stimulated β adrenergic receptor cardioprotection. Molecular Pharmacology. 2021b;100:568–579. doi: 10.1124/molpharm.121.000359. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib117] Wang J, Wu M, Chen Z, Wu L, Wang T, Cao D, Wang H, Liu S, Xu Y, Li F, Liu J, Chen N, Zhao S, Cheng J, Wang S, Hua T. The unconventional activation of the muscarinic acetylcholine receptor M4R by diverse ligands. Nature Communications. 2022;13:2855. doi: 10.1038/s41467-022-30595-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib118] Warne T, Edwards PC, Leslie AGW, Tate CG. Crystal structures of a stabilized β1-adrenoceptor bound to the biased agonists bucindolol and carvedilol. Structure. 2012;20:841–849. doi: 10.1016/j.str.2012.03.014. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib119] Warne T, Edwards PC, Doré AS, Leslie AGW, Tate CG. Molecular basis for high-affinity agonist binding in GPCRs. Science. 2019;364:775–778. doi: 10.1126/science.aau5595. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib120] White KL, Eddy MT, Gao ZG, Han GW, Lian T, Deary A, Patel N, Jacobson KA, Katritch V, Stevens RC. Structural connection between activation microswitch and allosteric sodium site in GPCR signaling. Structure. 2018;26:259–269. doi: 10.1016/j.str.2017.12.013. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib121] Williams CJ, Headd JJ, Moriarty NW, Prisant MG, Videau LL, Deis LN, Verma V, Keedy DA, Hintze BJ, Chen VB, Jain S, Lewis SM, Arendall WB, Snoeyink J, Adams PD, Lovell SC, Richardson JS, Richardson DC. MolProbity: more and better reference data for improved all-atom structure validation. Protein Science. 2018;27:293–315. doi: 10.1002/pro.3330. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib122] Wingler L.M, Elgeti M, Hilger D, Latorraca NR, Lerch MT, Staus DP, Dror RO, Kobilka BK, Hubbell WL, Lefkowitz RJ. Angiotensin analogs with divergent bias stabilize distinct receptor conformations. Cell. 2019;176:468–478. doi: 10.1016/j.cell.2018.12.005. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib123] Wood MR, Noetzel MJ, Melancon BJ, Poslusney MS, Nance KD, Hurtado MA, Luscombe VB, Weiner RL, Rodriguez AL, Lamsal A, Chang S, Bubser M, Blobaum AL, Engers DW, Niswender CM, Jones CK, Brandon NJ, Wood MW, Duggan ME, Conn PJ, Bridges TM, Lindsley CW. Discovery of vu0467485/az13713945: an M4 PAM evaluated as a preclinical candidate for the treatment of schizophrenia. ACS Medicinal Chemistry Letters. 2017a;8:233–238. doi: 10.1021/acsmedchemlett.6b00461. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib124] Wood MR, Noetzel MJ, Poslusney MS, Melancon BJ, Tarr JC, Lamsal A, Chang S, Luscombe VB, Weiner RL, Cho HP, Bubser M, Jones CK, Niswender CM, Wood MW, Engers DW, Brandon NJ, Duggan ME, Conn PJ, Bridges TM, Lindsley CW. Challenges in the development of an M4 PAM in vivo tool compound: the discovery of VU0467154 and unexpected DMPK profiles of close analogs. Bioorganic & Medicinal Chemistry Letters. 2017b;27:171–175. doi: 10.1016/j.bmcl.2016.11.086. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib125] Wootten D, Christopoulos A, Sexton PM. Emerging paradigms in GPCR allostery: implications for drug discovery. Nature Reviews. Drug Discovery. 2013;12:630–644. doi: 10.1038/nrd4052. [DOI] [PubMed] [Google Scholar]

[bib126] Xu J, Hu Y, Kaindl J, Risel P, Hübner H, Maeda S, Niu X, Li H, Gmeiner P, Jin C, Kobilka BK. Conformational complexity and dynamics in a muscarinic receptor revealed by NMR spectroscopy. Molecular Cell. 2019;75:53–65. doi: 10.1016/j.molcel.2019.04.028. [DOI] [PubMed] [Google Scholar]

[bib127] Xu J, Hübner H, Hu Y, Niu X, Gmeiner P, Jin C, Kobilka B. An Allosteric Ligand Stabilizes Distinct Conformations in the M2 Muscarinic Acetylcholine Receptor. bioRxiv. 2021 doi: 10.1101/2021.02.14.431178. [DOI]

[bib128] Xu J, Wang Q, Hübner H, Hu Y, Niu X, Wang H, Maeda S, Inoue A, Tao Y, Gmeiner P, Du Y, Jin C, Kobilka BK. Structural and dynamic insights into supra-physiological activation and allosteric modulation of a muscarinic acetylcholine receptor. Nature Communications. 2023;14:376. doi: 10.1038/s41467-022-35726-z. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib129] Ye L, Van Eps N, Zimmer M, Ernst OP, Prosser RS. Activation of the A2A adenosine G-protein-coupled receptor by conformational selection. Nature. 2016;533:265–268. doi: 10.1038/nature17668. [DOI] [PubMed] [Google Scholar]

[bib130] Zhang K. Gctf: real-time CTF determination and correction. Journal of Structural Biology. 2016;193:1–12. doi: 10.1016/j.jsb.2015.11.003. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib131] Zhang X, Belousoff MJ, Zhao P, Kooistra AJ, Truong TT, Ang SY, Underwood CR, Egebjerg T, Šenel P, Stewart GD, Liang YL, Glukhova A, Venugopal H, Christopoulos A, Furness SGB, Miller LJ, Reedtz-Runge S, Langmead CJ, Gloriam DE, Danev R, Sexton PM, Wootten D. Differential GLP-1R binding and activation by peptide and non-peptide agonists. Molecular Cell. 2020;80:485–500. doi: 10.1016/j.molcel.2020.09.020. [DOI] [PubMed] [Google Scholar]

[bib132] Zheng SQ, Palovcak E, Armache JP, Verba KA, Cheng Y, Agard DA. MotionCor2: anisotropic correction of beam-induced motion for improved cryo-electron microscopy. Nature Methods. 2017;14:331–332. doi: 10.1038/nmeth.4193. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib133] Zhou Q, Yang D, Wu M, Guo Y, Guo W, Zhong L, Cai X, Dai A, Jang W, Shakhnovich EI, Liu ZJ, Stevens RC, Lambert NA, Babu MM, Wang MW, Zhao S. Common activation mechanism of class A GPCRs. eLife. 2019;8:e50279. doi: 10.7554/eLife.50279. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib134] Zivanov J, Nakane T, Forsberg BO, Kimanius D, Hagen WJ, Lindahl E, Scheres SH. New tools for automated high-resolution cryo-EM structure determination in RELION-3. eLife. 2018;7:e42166. doi: 10.7554/eLife.42166. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Pharmacological hallmarks of allostery at the M4 muscarinic receptor elucidated through structure and dynamics

Ziva Vuckovic

Jinan Wang

Vi Pham

Jesse I Mobbs

Matthew J Belousoff

Apurba Bhattarai

Wessel AC Burger

Geoff Thompson

Mahmuda Yeasmin

Vindhya Nawaratne

Katie Leach

Emma T van der Westhuizen

Elham Khajehali

Yi-Lynn Liang

Alisa Glukhova

Denise Wootten

Craig W Lindsley

Andrew Tobin

Patrick Sexton

Radostin Danev

Celine Valant

Yinglong Miao

Arthur Christopoulos

David M Thal

Roles

Abstract

Introduction

Figure 1. Pharmacological characterization of the positive allosteric modulators (PAMs), LY298 and VU154, with acetylcholine (ACh) and iperoxo (Ipx) at the human M4 muscarinic acetylcholine receptor (mAChR).

Figure 1—figure supplement 1. Concentration–response curves of interactions between the orthosteric ligands (acetylcholine [ACh], iperoxo [Ipx]) and the allosteric ligands (LY298, VU154) at the human M4 muscarinic acetylcholine receptor (mAChR).

Figure 1—figure supplement 2. Pharmacological characterization of the positive allosteric modulators (PAMs), LY298 and VU154, with acetylcholine (ACh) and iperoxo (Ipx) in pERK1/2 signaling assays.

Results

Pharmacological characterization of M4 mAChR PAMs with ACh and Ipx

Table 1. Pharmacological parameters from radioligand binding and functional experiments.

Determination of M4R-Gi1 complex structures

Figure 2. Cryo-electron microscopy (cryo-EM) structures of the M4R-Gi1-scFv16 complexes.

Figure 2—figure supplement 1. Cryo-electron microscopy (cryo-EM) structures of the M4R-Gi1-scFv16 complexes.

Figure 2—figure supplement 2. Cryo-electron microscopy (cryo-EM) data processing and analysis.

Figure 2—figure supplement 3. Cryo-electron microscopy (cryo-EM) density maps.

Figure 2—figure supplement 4. Comparison of active state muscarinic acetylcholine receptor (mAChR) structures.

Figure 2—figure supplement 5. Comparison of active state M4 muscarinic acetylcholine receptor (mAChR) structures.

Table 2. Cryo-electron microscopy (cryo-EM) data collection, refinement, and validation statistics.

Structure and dynamics of agonist binding

Figure 3. Interactions of acetylcholine (ACh) and iperoxo (Ipx) with the receptor.

Figure 3—figure supplement 1. Interactions of acetylcholine (ACh) and iperoxo (Ipx) with the receptor measured during Gaussian accelerated molecular dynamics (GaMD) simulations.

Table 3. Gaussian accelerated molecular dynamics (GaMD) simulations of the M4 muscarinic acetylcholine receptor (mAChR).

Video 1. Movie from one Ipx-M4R-Gi1 Gaussian accelerated molecular dynamics (GaMD) simulation.

Video 2. Movie from one ACh-M4R-Gi1 Gaussian accelerated molecular dynamics (GaMD) simulation.

Structure and dynamics of PAM binding and allosteric modulation of agonist affinity

Figure 4. Binding and dynamics of LY298 and VU154.

Figure 4—figure supplement 1. Gaussian accelerated molecular dynamics (GaMD) simulations of LY293 and VU154 binding.

Figure 4—figure supplement 2. Key residues for the binding of LY298 and VU154 at the human M4 muscarinic acetylcholine receptor (mAChR).

Figure 4—figure supplement 3. Gaussian accelerated molecular dynamics (GaMD) simulations of M4R complexes with acetylcholine (ACh).

Video 3. 3D variability analysis of the Ipx-M4R-Gi1 cryo-electron microscopy (cryo-EM) structure.

Video 4. 3D variability analysis of the ACh-M4R- Gi1 cryo-electron microscopy (cryo-EM) structure.

Video 5. 3D variability analysis of the LY298-Ipx-M4R- Gi1 cryo-electron microscopy (cryo-EM) structure.

Video 6. 3D variability analysis of the VU154-Ipx-M4R- Gi1 cryo-electron microscopy (cryo-EM) structure.

Video 7. Movie from one VU154-Ipx-M4R-Gi1 Gaussian accelerated molecular dynamics (GaMD) simulation.

Video 8. Movie from one LY298-Ipx-M4R-Gi1 Gaussian accelerated molecular dynamics (GaMD) simulation.

Table 4. Pharmacological parameters of LY298 and VU154 at key M4 muscarinic acetylcholine receptor (mAChR) mutants.

A molecular mechanism of probe dependence

Structural and dynamic insights into orthosteric and allosteric agonism

Figure 5. Structural and dynamic insights into orthosteric and allosteric agonism.

Structural insights into allosteric modulation of agonist signaling

Table 5. Pharmacological parameters of M4 muscarinic acetylcholine receptor (mAChR) mutants.

Figure 6. Residues involved in binding, agonism, and modulation of acetylcholine (ACh) and LY298.

Figure 6—figure supplement 1. Concentration–response curves between acetylcholine (ACh) and LY298 at M4 muscarinic acetylcholine receptor (mAChR) mutants.

Figure 6—figure supplement 2. Interaction assays of agonists and positive allosteric modulators (PAMs) at the W413A6.48 M4 muscarinic acetylcholine receptor (mAChR) in a TruPath assay.

A molecular basis of species selectivity

Figure 7. A molecular mechanism for the species selectivity for VU154.

Figure 7—figure supplement 1. Binding parameters of positive allosteric modulators (PAMs) at the human and mouse M4 muscarinic acetylcholine receptors (mAChRs).

Figure 7—figure supplement 2. Functional parameters of the positive allosteric modulators (PAMs) at the human and mouse M4 muscarinic acetylcholine receptors (mAChRs) in pERK1/2 signaling assays.

Figure 7—figure supplement 3. Gaussian accelerated molecular dynamics (GaMD) simulations of D432E and T433R human M4 muscarinic acetylcholine receptor (mAChR) mutants.

Video 9. Movie from one VU154-Ipx-M4R(D432E)-Gi1 Gaussian accelerated molecular dynamics (GaMD) simulation.

Video 10. Movie from one VU154-Ipx-M4R(T433R)-Gi1 Gaussian accelerated molecular dynamics (GaMD) simulation.

Discussion

Figure 8. Conformational dynamics of the allostery at M4 muscarinic acetylcholine receptor (mAChR) signaling complexes.

Limitations of the study

Materials and methods

Figure 1. Pharmacological characterization of the positive allosteric modulators (PAMs), LY298 and VU154, with acetylcholine (ACh) and iperoxo (Ipx) at the human M₄ muscarinic acetylcholine receptor (mAChR).

Figure 1—figure supplement 1. Concentration–response curves of interactions between the orthosteric ligands (acetylcholine [ACh], iperoxo [Ipx]) and the allosteric ligands (LY298, VU154) at the human M₄ muscarinic acetylcholine receptor (mAChR).

Pharmacological characterization of M₄ mAChR PAMs with ACh and Ipx

Determination of M₄R-G_i1 complex structures

Figure 2. Cryo-electron microscopy (cryo-EM) structures of the M₄R-G_i1-scFv16 complexes.

Figure 2—figure supplement 1. Cryo-electron microscopy (cryo-EM) structures of the M₄R-G_i1-scFv16 complexes.

Figure 2—figure supplement 5. Comparison of active state M₄ muscarinic acetylcholine receptor (mAChR) structures.

Table 3. Gaussian accelerated molecular dynamics (GaMD) simulations of the M₄ muscarinic acetylcholine receptor (mAChR).

Video 1. Movie from one Ipx-M₄R-G_i1 Gaussian accelerated molecular dynamics (GaMD) simulation.

Video 2. Movie from one ACh-M₄R-G_i1 Gaussian accelerated molecular dynamics (GaMD) simulation.

Figure 4—figure supplement 3. Gaussian accelerated molecular dynamics (GaMD) simulations of M₄R complexes with acetylcholine (ACh).

Video 3. 3D variability analysis of the Ipx-M₄R-G_i1 cryo-electron microscopy (cryo-EM) structure.

Video 4. 3D variability analysis of the ACh-M₄R- G_i1 cryo-electron microscopy (cryo-EM) structure.

Video 5. 3D variability analysis of the LY298-Ipx-M₄R- G_i1 cryo-electron microscopy (cryo-EM) structure.

Video 6. 3D variability analysis of the VU154-Ipx-M₄R- G_i1 cryo-electron microscopy (cryo-EM) structure.

Video 7. Movie from one VU154-Ipx-M₄R-G_i1 Gaussian accelerated molecular dynamics (GaMD) simulation.

Video 8. Movie from one LY298-Ipx-M₄R-G_i1 Gaussian accelerated molecular dynamics (GaMD) simulation.

Table 4. Pharmacological parameters of LY298 and VU154 at key M₄ muscarinic acetylcholine receptor (mAChR) mutants.

Table 5. Pharmacological parameters of M₄ muscarinic acetylcholine receptor (mAChR) mutants.

Figure 6—figure supplement 1. Concentration–response curves between acetylcholine (ACh) and LY298 at M₄ muscarinic acetylcholine receptor (mAChR) mutants.

Figure 6—figure supplement 2. Interaction assays of agonists and positive allosteric modulators (PAMs) at the W413A^6.48 M₄ muscarinic acetylcholine receptor (mAChR) in a TruPath assay.

Figure 7—figure supplement 1. Binding parameters of positive allosteric modulators (PAMs) at the human and mouse M₄ muscarinic acetylcholine receptors (mAChRs).

Figure 7—figure supplement 2. Functional parameters of the positive allosteric modulators (PAMs) at the human and mouse M₄ muscarinic acetylcholine receptors (mAChRs) in pERK1/2 signaling assays.

Figure 7—figure supplement 3. Gaussian accelerated molecular dynamics (GaMD) simulations of D432E and T433R human M₄ muscarinic acetylcholine receptor (mAChR) mutants.

Video 9. Movie from one VU154-Ipx-M₄R(D432E)-G_i1 Gaussian accelerated molecular dynamics (GaMD) simulation.

Video 10. Movie from one VU154-Ipx-M4R(T433R)-G_i1 Gaussian accelerated molecular dynamics (GaMD) simulation.

Figure 8. Conformational dynamics of the allostery at M₄ muscarinic acetylcholine receptor (mAChR) signaling complexes.

Expression and purification of M₄R-G_i1-scFv16 complexes