DIRECT C-11 FUNCTIONALISATION OF ANATOXIN-a

SYNTHESIS OF DANSYL-BASED FLUORESCENT LIGANDS

Laurent Ducry,^a Nicholas A. Magnus,^a Valérie Rolland,^a
Susan Wonnacott,^b and Timothy Gallagher^a*

^aSchool of Chemistry, University of Bristol, Bristol BS8 1TS, UK
^b School of Biology and Biochemistry, University of Bath, Bath BA2 7AY, UK

Abstract. N-Boc anatoxin-a (6) is efficiently converted to alpha-tosyloxyketone (7) in a single step using hydroxy(tosyloxy)iodobenzene. The electrophilic reactivity available in (7) has been used to synthesise the fluorescent anatoxin-a derivatives (2a) and (2b).

Anatoxin-a (1) is a potent nicotinic agonist¹ and offers significant potential as a basis for new molecular probes for the characterisation of neuronal nicotinic acetylcholine receptors. Fluorescent² and radiolabelled variants of anatoxin-a have obvious potential in this field and in this paper we describe the synthesis of two fluorescent anatoxin-a derivatives (2a) and (2b) incorporating a dansyl unit. The fluorescent moiety was attached to the ligand template via a thioether and a linear methylene spacer, incorporation of which has been achieved using an efficient and regioselective activation at C-11 of N-Boc anatoxin-a.^3,4

When contemplating structural modification of a potent ligand, such as anatoxin-a, two important issues must be addressed and balanced against one another. Firstly, what site of the ligand's skeleton will tolerate substitution and, in the process, retain effective biological potentcy? Secondly, are suitable synthetic methods available for achieving the chemical activation required to append a substituent or spacer unit at this site?

Both of these issues have now been evaluated. Anatoxin-a offers various options, but substitution of C-11 is especially attractive: homoanatoxin (3) and simple alkyl homologues (4) and (5) have been shown to retain high nicotinic potency.⁴ Modification via C-11 was also attractive in a synthetic sense because this position offers potential access to both nucleophilic (enolate) and electrophilic (e.g. via an -haloketone) reactivity profiles.

A variety of protocols for activating C-11 of N-Boc anatoxin-a (6) have been evaluated but the method of choice involved use of hydroxy(tosyloxy)iodobenzene (Koser's reagent)⁵ as shown in Scheme 1. Treatment of (±)-(6) with 5 equivalents of the Koser reagent in CH₂Cl₂ at room temperature gave the alpha-tosyloxy ketone (7) in 68 % yield based on recovered (6).⁶

Scheme 1

Other options for functionalisation at C-11 have also been evaluated. Silyl enol ether (8) was prepared as described by Rapoport,⁷ but we were unable to convert this intermediate to an alpha-haloketone (9) under various conditions. Enolate generation (by direct deprotonation of the C-11 methyl group) is achieveable; the enolate corresponding to (6) has been methylated in 40% yield as a route to (+)-homoanantoxin (3).⁸ However, this enolate is comparatively unreactive and we have not successfully trapped this species with other alkylating agents.

The process used to assemble the dansylated anatoxin-a ligands (2a) and (2b) is shown in Scheme 2. Using efficient methods, w-bromosulfonamides (10) and (11) were converted to disulfides (12a) and (12b) respectively. Reductive cleavage of the disulfide linkage provided the corresponding free thiols which were directly coupled with the alpha-tosyloxyketone (7) under basic conditions to give the N-Boc-protected adducts (13a) and (13b). N-Deprotection was then carried out under acidic conditions, followed by basic workup, to provide ligands (2a) and (2b), both in essentially quantitative yield.

Scheme 2

The biological activity of dansylated adducts (2a) and (2b) has been evaluated at the two major nicotinic radioligand binding sites in brain membranes.⁹ Somewhat surprisingly, neither ligand displayed significant potency at the [¹²⁵I]alpha-bungarotoxin binding site compared with anatoxin-a itself which has an IC₅₀ value of 0.5 µM. At the [³H]nicotine binding site (corresponding to alpha4beta2 receptors), anatoxin-a has a higher affinity and although the simpler dansyl ligand (2a) effectively competed for the binding site, this ligand was more than two orders of magnitude less active than anatoxin-a. It is not clear if this lack of potency is associated with the presence of the thioether linkage or the dansyl unit itself but work to resolve this issue is underway.

In summary, N-Boc anatoxin-a (6) may be effectively activated at C-11 in a single step using the Koser reagent. This provides alpha-tosyloxyketone (7) as a versatile unit for further elaboration of the side chain of anatoxin-a.

Acknowledgements. We thank SIBIA Neurosciences Inc. (part support to N.A.M.), the ERASMUS Programme (L.D. and V.R.) and EC (S.W.) and BBSRC (T.G. and S.W.) for financial support.

References and Notes

1. Thomas, P.; Stephens, M.; Wilkie, G.; Amar, M.; Lunt, G. G.; Whiting, P.; Gallagher, T.; Pereira, E.; Alkondan, M.; Albuquerque, E. X.; Wonnacott, S. J. Neurochemistry 1993, 60, 2308.
2. For the application of dansylated acetylcholine derivatives see Heidmann, T.; Changeux, J. P. Eur. J. Biochem. 1979, 94, 281. Heidmann, T.; Changeux, J. P. Biochem. Biophys. Res. Commun. 1980, 97, 889.
3. For details of the biological activity of anatoxin ligands based on modification⁷ of the methyl ketone side chain see: (a) Wonnacott, S.; Jackman, S.; Swanson, K. L.; Rapoport, H.; Albuquerque, E. X. J. Pharmacol. Exp. Ther. 1991, 259, 387. (b) Swanson, K. L.; Aronstam, R. S.; Wonnacott, S.; Rapoport, H.; Albuquerque, E. X. J. Pharmacol. Exp. Ther. 1991, 259, 377.
4. For the synthesis of the C-11 anatoxin-a analogues (3-5) and variants, see: (a) Wonnacott, S.; Swanson, K. L.; Albuquerque, E. X.; Huby, N. J. S.; Thompson, P.; Gallagher, T. Biochem. Pharmacol. 1992, 43, 419. (b) Thomas, P.; Brough, P.A.; Gallagher, T.; Wonnacott, S. Drug Dev. Res. 1994, 31, 147. (c) Huby, N. J. S.; Thompson, P.; Wonnacott, S.; Gallagher, T. J. Chem. Soc., Chem. Commun. 1991, 243. Homoanatoxin,^4a an obvious candidate on which to base a tritiated anatoxin-a ligand, was subsequently isolated from a natural source: Skulberg, O. M.; Carmichael, W.W.; Andersen, R. A.; Matsunaga, S.; Moore, R. E.; Skulberg, R.Environ. Toxicol. Chem. 1992, 11, 321.
5. Koser, F. G.; Relenyi, A. G.; Kales, A. N.; Rebrovic, L.; Wettach, R. H. J. Org. Chem. 1982, 47, 2487.
6. Racemic N-Boc Anatoxin-a (6) (120 mg, .45 mmol) in CH₂Cl₂ (8.5 mL) (under Ar) was treated with the Koser reagent and stirred for 17 h. After this time, water (5 mL) was added, the mixture extracted with CH₂Cl₂ (3 x 20 mL) and the combined extracts were dried (Na₂SO₄). Removal of solvent and chromatography gave (6) (56 mg) (15% EtOAc/85% hexanes) followed by (7) (71 mg, 36%,68% based on recovered (6)) (50% EtOAc/50% hexanes) as colourless foam. IR 2972, 1688, and 1630 cm^-1; ¹H NMR (270 MHz, CDCl₃) 1.20-2.42 (17 H, m), 2.45 (3 H, s), 4.28-4.42 (1 H, m), 4.84-5.30 (3 H, m), 6.71 (1 H, t, J 6.2 Hz), 7.36 (2 H, d, J 8 Hz) and 7.84 (2H, d, J 8 Hz); HRMS (CI) calcd forC₂₉H₂₉NO₆S+H 436.179. Found: m/e 436.178. If this reaction is allowed to continue for longer periods, decomposition of (6) becomes significant. All new compounds shown are racemic (based on use of (±)-(6)) and have been fully characterised.
7. The efficient synthesis of silyl enol ether (8) has been described and oxidation (MCPBA) provides the corresponding alpha-hydroxyketone. Howard, M. H.; Sardina, F. J.; Rapoport, H. J. Org. Chem. 1990, 55, 2829. This chemistry provided the basis of the synthesis of a series of side-chain analogues.³
8. Brough, P. A. PhD Thesis, University of Bristol, 1994. The potassium enolate derived from (6) was trapped by (PhS)₂ to give the alpha-SPh substituted ketone (62%) but we were unable to achieve satisfactory results using disulfides (12a) or (12b) as the electrophilic component.
9. Competition binding assays were carried out using rat brain P2 membranes and [¹²⁵I]bungarotoxin (1 nM) or [³H]nicotine (10 nM) as described previously.^3,4 Ligands (2a) and (2b) were assayed over the concentration range 10^-4 to 10^-9 M: the ligands were dissolved in DMSO to a concentration of 10 mM and diluted in assay buffer to yield the desired final concentration. DMSO, diluted to correspond to the highest concentration of (2a) and (2b) examined, itself inhibited the binding of both radioligands by about 20%. Anatoxin-a (1) was assayed in parallel as a positive control.

DIRECT C-11 FUNCTIONALISATION OF ANATOXIN-a.

SYNTHESIS OF DANSYL-BASED FLUORESCENT LIGANDS

Laurent Ducry, Nicholas A. Magnus, Valérie Rolland, Susan Wonnacott, and Timothy Gallagher

School of Chemistry, University of Bristol, Bristol and School of Biology and Biochemistry, University of Bath, Bath, UK

C-11 Functionalision of N-Boc anatoxin-a is achieved using PhI(OH)OTs (the Koser reagent) and the resulting alpha-tosyloxy ketone (1) has been used to synthesise dansylated ligands (2).