Salvinorin B methoxymethyl ether

The title compound [MOM-SalB; systematic name: methyl (2S,4aR,6aR,7R,9S,10aS,10bR)-2-(3-furyl)-9-methoxymethoxy-6a,10b-dimethyl-4,10-dioxo-2,4a,5,6,7,8,9,10a-octahydro-1H-benzo[f]isochromene-7-carboxylate], C23H30O8, is a derivative of the κ-opioid salvinorin A with enhanced potency, selectivity, and duration of action. Superimposition of their crystal structures reveals, surprisingly, that the terminal C and O atoms of the MOM group overlap with the corresponding atoms in salvinorin A, which are separated by an additional bond. This counter-intuitive isosterism is possible because the MOM ether adopts the ‘classic anomeric’ conformation (gauche–gauche), tracing a helix around the planar acetate of salvinorin A. This overlap is not seen in the recently reported structure of the tetrahydropyranyl ether, which is less potent. The classic anomeric conformation is strongly favoured in alkoxymethyl ethers, but not in substituted acetals, which may contribute to their reduced potency. This structure may prove useful in evaluating models of the activated κ-opioid receptor.

The title compound [MOM-SalB; systematic name: methyl (2S,4aR,6aR,7R,9S,10aS,10bR)-2-(3-furyl)-9-methoxymethoxy-6a, 10b-dimethyl-4,10-dioxo-2,4a,5,6,7,8,9,10a-octahydro-1H-benzo[f]isochromene-7-carboxylate], C 23 H 30 O 8 , is a derivative of the -opioid salvinorin A with enhanced potency, selectivity, and duration of action. Superimposition of their crystal structures reveals, surprisingly, that the terminal C and O atoms of the MOM group overlap with the corresponding atoms in salvinorin A, which are separated by an additional bond. This counter-intuitive isosterism is possible because the MOM ether adopts the 'classic anomeric' conformation (gauche-gauche), tracing a helix around the planar acetate of salvinorin A. This overlap is not seen in the recently reported structure of the tetrahydropyranyl ether, which is less potent. The classic anomeric conformation is strongly favoured in alkoxymethyl ethers, but not in substituted acetals, which may contribute to their reduced potency. This structure may prove useful in evaluating models of the activated -opioid receptor.   (1) is among the most potent and selective κ (kappa) opioids known, with subnanomolar affinity and potency (Wang et al., 2008). A semisynthetic derivative of the naturally occurring κ opioid salvinorin A (2), (1) was the first derivative reported to be more potent than (2) in vitro, and also showed greater potency and duration of action in mice (Wang et al., 2008). The extreme potency of (1) has been confirmed both in vitro (Munro et al., 2008, Prevatt-Smith et al., 2011) and in rats (Baker et al., 2009, Peet & Baker, 2011. The name MOM-SalB is widely used; the incorrect name `2-methoxymethylsalvinorin B′, implying that the substituent is directly attached to C2, should be avoided.

Related literature
In Figure 1, the structures of (1) and (2) have been drawn to emphasize their similarity, with O2, C21 and O3 superimposable. The terminal methyl group C22 is attached to O3 in (1) but C21 in (2), and might be expected to interact with different regions of the receptor. Extensive research has been done into the structure-activity relationships of (2), especially the role of the C2 acetate. The deacetyl analogue salvinorin B is at least 60-fold less potent than (2). Deoxygenation or demethylation of the acetate causes smaller reductions in potency (Cunningham et al., 2011). This suggests that the two extremities of the acetate (O3 and C22) engage in separate, synergistic interactions with the binding pocket (Munro et al., 2008). The structure-activity relationships of (1) have also been explored. Potency is dramatically reduced by replacement of O3 with sulfur or carbon (Munro et al., 2008); this similarity to (2) is consistent with the proposed common binding pose. The ethoxymethyl ether (3) appears to be even more potent and selective than (1), both in vitro (Munro et al., 2008, Prevatt-Smith et al., 2011) and in vivo (Baker et al., 2009, Peet & Baker, 2011 reportedly exhibits higher affinity than 12-epi-(1) (Béguin et al., 2012). Further extension or branching of the terminal alkyl chain reduces affinity (Munro et al., 2008). Thus, the ethoxymethyl substituent appears to confer optimal affinity and potency. Like (1), (3) is also metabolized more slowly than (2) (Hooker et al., 2009). Based on the above hypothesis that the C22 methyl groups in (1) and (2) address different regions of the binding pocket, ethoxyethyl ether (4) was designed in that hope that it would interact with both of these regions, maximizing affinity. However, upon testing, (4) proved to have much lower affinity and potency than (3) (Munro et al., 2008, Prevatt-Smith et al., 2011. Indeed, all derivatives tested to date featuring substituted acetals, such as (5), exhibit reduced affinity and potency (Munro et al., 2008, Prevatt-Smith et al., 2011. These surprising and disappointing results cast doubt on the proposed binding model. We therefore determined the structure of (1) by single-crystal X-ray diffraction to obtain conformational information ( Figure 2).
Other than the disordered furan ring, the neoclerodane scaffold is almost perfectly superimposable (r.m.s. < 0.1 Å) upon that of (2), as expected (Ortega et al., 1982). However, the resulting relationship between the acetate and the MOM ether was unexpected. Both O3 and C22 in (1) overlap with their counterparts in (2), being separated by just 0.9 Å (O3) and 1.2 Å (C22) -less than their atomic radii. The overlapping van der Waals surfaces of O3 and C22 are shown in Figure 3.
This result was surprising, given the different point of attachment of C22 in these two compounds ( Figure 1). This supplementary materials counterintuitive result occurs because both bonds to the acetal carbon C21 in (1) are gauche (torsion angles: 69.8° (O2-C22) and 76.5° (C2-O3)), allowing the ether to trace a part helix around the planar acetate in (2). This is known as the `classic anomeric′ conformation (Anderson, 2000, Brameld et al., 2008. Generally, solid-state conformations coincide closely with the bioactive conformation of the protein-bound ligand (Brameld et al., 2008). This is because both solidstate and bound conformations tend toward the free energy minimum. The similarity is greatest in high-affinity ligands, since any change in conformation during binding requires energy, and this `energetic penalty′ reduces affinity (Brameld et al., 2008). As discussed above, structure-activity studies indicate that O3 and C22 contribute substantially to binding of both (2) and (1). The near-superimposability of these atoms in the crystal structures of these two high-affinity ligands suggests that they may represent similar bioactive conformations. Alkoxymethyl ethers invariably adopt the classic anomeric conformation, due to strong anomeric interactions involving both O atoms (Anderson, 2000, Brameld et al., 2008. Interestingly, however, substitution of the acetal carbon introduces steric interactions which greatly reduce this preference. With a methyl substituent, as in (4), the classic anomeric conformation predominates, but is not exclusive.
With larger substituents this conformation is strongly disfavoured, and rarely occurs (Anderson, 2000). If the classic anomeric conformation seen in (1) is optimal for binding, acetal substitution would therefore be expected to reduce affinity by this conformational influence, even if the substituents do not themselves interact unfavourably with the receptor. This may contribute to the dramatic reductions in affinity and potency seen even with small acetal substituents (Munro et al., 2008, Prevatt-Smith et al., 2011. The recently reported crystal structure of the tetrahydropyranyl (THP) ether (5) illustrates this point (Prevatt-Smith et al., 2011). The cyclic acetal does not adopt the classic anomeric conformation, and superimposition on (2) gives much poorer overlap than seen with (1) (Figure 4). Acetal oxygen O3 is separated from its counterpart in (2) by 2.5 Å, and is instead almost coincident with C22 (<0.2 Å). Furthermore, the THP ring is disordered, consisting of a mixture of two interconvertible chair conformations. Thus, the THP ether exhibits weaker conformational preferences than the MOM ether, and much poorer overlap with (2). This may partly account for its lower potency. Our results suggest a possible conformational basis for the high binding affinity of salvinorin B alkoxymethyl ethers such as (1) and (3), and for the reduced affinity of substituted acetal derivatives such as (4) and (5). As a structurally atypical and extremely potent agonist, the structure of (1) reported here may prove useful in modelling the activation of the κ opioid receptor.

Experimental
Compound (1) was prepared as described previously, by treatment of salvinorin B with CH 3 OCH 2 Cl and i-Pr 2 NEt in anhydrous CH 2 Cl 2 , and purified by flash chromatography on silica gel (Béguin et al., 2009). Amended characterization data have been reported elsewhere (Munro et al., 2008). Dissolution of 200 mg in minimal boiling methanol (~3 ml) and slow cooling gave colourless needles, mp 165-167 °C (438-440 K).

Figure 1
Structures of compounds discussed.

Figure 2
Crystal structure of (1) with 50% probability thermal displacement ellipsoids. Atom numbering follows the crystal structure of (2)

Special details
Refinement. Refinement of F 2 against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F 2 , conventional R-factors R are based on F, with F set to zero for negative F 2 . The threshold expression of F 2 > 2σ(F 2 ) is used only for calculating R-factors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F 2 are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger.

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å 2 )
x y z U iso */U eq Occ. (