Structural analysis of docking studies for the 56 selected CB2R modulators revealed several compounds with key structural characteristics compatible with the M1R allosteric binding site, indicating their potential as dual CB2R-M1R modulators.

Among these, compound 29, featuring an oxoindolinylidene benzohydrazide scaffold, showed promising results due to its optimal bent conformation, which fits snugly within the M1R PAM binding site (Compound 29, Figures fig8 and fig9A). Docking studies indicated that the oxoindolinylidene moiety aligns parallel to W400^7.35, enabling aromatic stacking; the benzohydrazide carbonyl’s proximity to K392^ECL3 facilitates an ion–dipole interaction, and its benzene ring engages in π–π stacking with Y179^ECL2. Concurrently, the benzyl tail projects toward the aromatic-rich region between TMHs 2 and 7, forming additional stacking interactions that further stabilize the binding pose.

A second promising scaffold class includes the cyclopentapyrazole carboxamides, which were particularly abundant and featured structural variations at the carboxamide cap (Compounds 30–35, Figure fig8 ). Remarkably, the R,R-cyclopropyl isomers demonstrated superior accessibility and fit within the M1R PAM pocket, suggesting a favorable conformational bias. In this series, the pyrazole cores are stacked with W400^7.35, which is applicable only when the cyclopropyl group is oriented away from this residue in the R,R isomers. The carboxamide head groups extended toward the extracellular vestibule and could form key hydrogen bonds with Q177^ECL2, K392^ECL3, and/or D393^ECL3, and in some cases engage in π-cation interactions with K392^ECL3 depending on the nature of the cap. Meanwhile, the difluorophenyl moiety further stabilizes the pose via stacking not only with Y85^2.63 but also with Y404^7.39, the latter being a residue known for its role in π-cation interactions with orthosteric agonists and potentially contributing to receptor stabilization. Among this series, compounds 30–33 emerged as the most promising derivatives (Figures fig9 B,C and S3A,B , respectively). In contrast, compounds featuring bulky noncyclic caps (e.g., compound 34) or lacking branched polar groups on their cyclic caps (e.g., compound 35) failed to establish effective interactions. Overall, the structural diversity at the cap positions provides a compelling foundation for systematic SAR exploration and offers strong potential for fine-tuning dual activity toward the development of optimized modulators.

Indole and pyrrolopyridine carboxamide derivatives also showed promising characteristics (Compounds 36–38, Figure fig8 ). These compounds incorporate an additional carboxamide group at the head moiety, paired with a benzylic substituent, permitting extra hydrogen bonds and potential π-cation interactions with ECL2 or ECL3 residues. Docking indicated that the indole or pyrrolopyridine carboxamide promotes aromatic stacking with W400^7.35 and/or Y179^ECL2, along with hydrogen bonding with Q177^ECL2 (Figures S3C,D and fig9D, respectively). The appended benzyl carboxamide could further contribute by forming additional hydrogen bonds with Q177^ECL2, E393^ECL3, and/or K392^ECL3. Despite the fact that these derivatives have hydrocarbon tails rather than rigid aromatic systems, docking results suggested that the aliphatic chains are accommodated within the hydrophobic pocket between TMHs 2 and 7. This compatibility highlights their good potential for further structural optimization.

Due to their strong structural resemblance to well-known M1R PAMs, especially compound 39, the naphthyridinone carboxamide scaffold stood out as one of the most promising dual modulators (Compounds 39–41, Figure fig8 ). Its bent conformation enabled the bicycle core to engage in π–π stacking with W400^7.35 and/or Y179^ECL2, while the carbonyl oxygens formed hydrogen bonds with K392^ECL3 and Y85^2.63, and the aromatic tail extended toward the TMHs 2 and 7 region (Figure fig9 E). Derivatives lacking aromatic tails, such as compounds 40 and 41, still displayed fair binding potential, where the former showed more favorable hydrogen bond interactions than did the latter compound (Figure S3E,F, respectively).

Similarly, a prominent core shared by M1R PAMs and CB2R agonists is the pyrazoloquinolinone scaffold. The primary differences lie in the tail region: M1R-directed analogs possess aromatic tails (compounds 12–20, Figure fig6 ), whereas CB2R counterparts feature flexible pentyl chains (compounds 42–45, Figure fig8 ). Docking studies revealed that compounds 42 and 43 engage in key interactions, including aromatic stacking with W400^7.35 and/or Y179^ECL2, hydrogen bonding with Q177^ECL2, and favorable orientation of the pentyl tail within the TMHs 2 and 7 hydrophobic pocket (Figures fig9 F and S3G, respectively). Conversely, compounds 44 and 45, which incorporate an extended linker and a bulky adamantanyl head, respectively, display significantly lower interaction quality, likely due to steric hindrance disrupting optimal positioning within the binding site.

The pyrid-4-one carboxamide scaffold was highly represented (compounds 46–56, Figure fig8 ). Top binders, such as compounds 46–48, combine cyclohexyl or tetralin heads with a 6-phenyl substitution. These compounds effectively engaged in stacking with W400^7.35 and/or Y179^ECL2 while simultaneously forming hydrogen bonds with K392^ECL3 and/or Q177^ECL2 without encountering steric hindrance (Figures fig9 G and S3H,I, respectively). Despite the fact that these derivatives possess monocyclic aromatic cores, the 6-phenyl group seems to compensate by permitting additional stacking. Superposition of these ligand docking poses with those of bi/tricyclic analog docking showed favorable spatial overlaps, suggesting that the phenyl group is well accommodated when oriented orthogonally to the pyridone ring. Nevertheless, their flexible hydrocarbon tails may limit the binding efficacy by favoring van der Waals interactions over robust aromatic stacking. While derivatives featuring adamantanyl or adamantanyl-methyl head groups or branched substituents at position 6 were too bulky and lacked compatibility with the pocket.

Similarly, quinolinone carboxamide derivatives (Compounds 57–60, Figure fig8 ), featuring a fused bicyclic core this time, were evaluated for binding. Among them, only compound 60, which incorporates a tetralin headgroup, showed a favorable binding profile, outperforming analogs bearing bulkier headgroups (Figure fig9 H).

Another scaffold was the indole carboxamide, characterized by a single carboxamide group (Compounds 61–69, Figure fig8 ), in contrast to the dicarboxamide structure seen in compound 36. Notably, derivatives bearing a naphthyl or cyclohexyl substituent at the carboxamide group exhibited the best fit within the binding site, striking an intriguing pose reminiscent of the selective M2R PAM LY2119620, as observed in M2R–IXO–PAM complex structures (PDB: 4MQT ref25 and 6OIK ref63). In this orientation, the indole core is preserved for π–π stacking with W400^7.35 and Y179^ECL2, while the carboxamide moiety is directed toward the TMHs 2 and 7 region this time (Figures fig9 I and S3J). Here, the carbonyl oxygen forms a hydrogen bond with Y82^2.60, and the amide nitrogen is positioned to interact with E397^ECL3, potentially via a water-mediated hydrogen bond. Additionally, the naphthalene ring in compound 61 further stabilized the binding via aromatic stacking with Y85^2.63. While other derivatives that possess a bulky adamantanyl headgroup were unable to fit within the binding site and demonstrated limited binding.

A similar binding orientation was observed for the indolyl-phenyl methanone derivatives (Compounds 70–71, Figure fig8 ). Within this series, compound 70 outperformed compound 71, likely due to the presence of a hydrophilic N-substituent in compound 70 that extended toward ECL3, enabling an extra ion–dipole interaction with K392^ECL3 (Figures fig9 J and S3K, respectively). Nonetheless, while these results are encouraging, further validation will be necessary to validate the proposed binding modes for both scaffold classes.

Conversely, several scaffold series demonstrated poor suitability for the M1R PAM site due to insufficient size, overly short structures, or excessive steric bulk. The dicyclohexylpicolinamide derivatives (Compounds 72–73, Figure fig8 ), featuring monocyclic aromatic cores and short cyclohexyl tails, result in shallow surface-level binding with weak interaction profiles. Similarly, the adamantanyl indolinyl or quinolinyl methanone derivatives (Compounds 74–76, Figure fig8 ) were unable to form critical contacts due to their diminutive size and the steric impediment created by the bulky adamantanyl group. Their structural resemblance to dopaminergic modulators may also raise concerns about off-target effects. The pyrazole carboxamide scaffold (Compound 77, Figure fig8 ), represented by a single Y-shaped molecule bearing a bulky trimethylbicycloheptanyl head, was excessively sterically hindered, preventing effective access to the binding pocket and precluding the formation of critical hydrogen bonds. Lastly, pyrid-2-one carboxamide derivatives (Compounds 78–80, Figure fig8 ) suffered from steric congestion, particularly in the tetrasubstituted analogue, leading to poor accommodation within the M1R PAM site.