Covalent attachment of antagonists to the α7 nicotinic acetylcholine receptor: Synthesis and reactivity of substituted maleimides

Article abstract of DOI:10.1039/c2cc32442c

The 3-methylmaleimide congeners of the natural product methyllycaconitine (MLA) and an analogue covalently attach to functional cysteine mutants of the α7 nicotinic acetylcholine receptor (nAChR).

Full text of DOI:10.1039/c2cc32442c

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 6699–6701
www.rsc.org/chemcomm
COMMUNICATION
Covalent attachment of antagonists to the a7 nicotinic acetylcholine
receptor: synthesis and reactivity of substituted maleimidesw
Joseph I. Ambrus,a Jill I. Halliday,a Nicholas Kanizaj,a Nathan Absalom,b Kasper Harpsøe,c
Thomas Balle,d Mary Chebibb and Malcolm D. McLeod*a
Received 5th April 2012, Accepted 9th May 2012
DOI: 10.1039/c2cc32442c
The 3-methylmaleimide congeners of the natural product
methyllycaconitine (MLA) and an analogue covalently attach
to functional cysteine mutants of the a7 nicotinic acetylcholine
receptor (nAChR).
structures of wildtype and mutant AChBPs.11 In this study
we employ cysteine mutagenesis in combination with thiophilic
ligands to explore the binding of MLA 1 and a simple analogue
2 at functional a7 nAChRs.
To develop thiophilic modifications of MLA 1 and analogue
2 the exchange of the side chain methylsuccinimide with a
substituted maleimide to afford compounds like 3 and 4
suggested itself as a suitably conservative substitution. Maleimides
react rapidly, selectively and irreversibly12 with thiolate-bearing
residues such as the amino acid cysteine, even in the presence
of other nucleophiles such as lysine.3 This reaction irreversibly
forms a thioether linkage between the maleimide and the
cysteine residue.6 Unsubstituted maleimides are routinely
employed, and bromo-substituted maleimides13 have recently
been investigated, however other substituents have not been
explored. Consequently we undertook preliminary studies to
synthesise some substituted maleimides and explore their
reactivity with thiol containing amino acids.
Bioconjugation provides a powerful tool to interrogate bio-
logical systems through the chemoselective and site-specific
covalent attachment of ligands to biomolecules.1–3 A powerful
example of this is the Substituted Cysteine Accessibility
Method (SCAM),4–6 which employs cysteine mutagenesis in
combination with thiophilic reagents to investigate and modify
proteins. The technique is most commonly used with water
soluble methanethiosulfonate reagents to identify the solvent
accessible residues of a protein, but related cysteine muta-
genesis techniques have also been used to append affinity tags
and chromophores, or to identify ligand binding sites using a
range of electrophilic functionality including maleimides,
haloacetamides or mustards.7 We recently employed a thio-
philic mustard derivative to establish the channel pore binding
site of a novel azabicyclic ligand at the a4b2 nAChR.8
To further explore the potential of cysteine mutagenesis as a
technique to probe ligand binding we elected to explore the
norditerpenoid alkaloid methyllycaconitine (MLA) 1 which is
widely used as a potent and subtype selective antagonist of the
a7 nAChR. Despite the rapidly emerging structural data from
related full length receptors9 and X-ray crystal structures of
mollusc-derived acetylcholine binding proteins (AChBPs) as
models of the nAChR N-terminal region,10 limited direct
evidence is available on the nature of ligand binding at
functional nAChRs. The binding mode of MLA 1 at the a7
nAChRs has recently been suggested from X-ray crystal
The reaction of two model thiol amino acids, N-Boc-Cys-
OMe and N-Ac-Cys-OH, was investigated with a range of
substituted maleimides appended to a model sidechain ester
(5–9, Scheme 1) under pseudo first order conditions. The
maleimides were selected to afford a range of steric and
electronic influences on reactivity and included the unsubsti-
tuted maleimide 5 together with the bromo 6, methyl 7,
trifluoromethyl 8 and dimethyl 9 maleimides. Two synthetic
routes were devised to access these compounds. The synthesis
of maleimides 7 and 9 was accomplished by the DCC coupling
of the appropriately substituted maleimide 10 with adamant-
1-ylmethanol to give the desired esters. An alternate route
was employed to access maleimides 5, 6 and 8. A DCC
a Research School of Chemistry, Australian National University,
Canberra, ACT 0200, Australia.
E-mail: malcolm.mcleod@anu.edu.au
b Faculty of Pharmacy, The University of Sydney, Sydney,
NSW 2006, Australia
c NNF Centre for Protein Research, Faculty of Health and Medical
Sciences, University of Copenhagen, 2100 Copenhagen, Denmark
d Department of Drug Design and Pharmacology,
Faculty of Health and Medical Sciences, University of Copenhagen,
Universitetsparken 2, 2100 Copenhagen, Denmark
w Electronic supplementary information (ESI) available: experimental and
computational procedures and data, PDB files for Fig. 2, 1H and 13C NMR
spectra for compounds 3 and 4. See DOI: 10.1039/c2cc32442c
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 6699–6701 6699

View Article Online
oocytes expressing these receptors, a cation-selective current
was elicited that was inhibited by MLA 1, demonstrating that
these receptors were functional.
Incubation of 1 mM MLA maleimide 3 with the L90T
receptor for 6 min did not alter currents elicited by ACh
(I(N) = 1.09 ꢀ 0.08, n = 4, Fig. 1). After 6 min incubation of
10 nM MLA maleimide 3, the current elicited by ACh was
significantly and irreversibly reduced in oocytes expressing
S188C : L90T compared to L90T alone (I(N) = 0.24 ꢀ 0.09,
n = 3, p o 0.01). In contrast, the ACh response of S189C : L90T
was not significantly reduced by MLA maleimide 3 compared
to the L90T (I(N) = 0.85 ꢀ 0.06, 1.12 ꢀ 0.18 respectively, n = 4).
The first order reaction rate constant for the reaction of MLA
maleimide 3 to S188C : L90T was estimated to be k = 0.0058 ꢀ
0.0014 sꢁ1. This rate of binding and reaction is comparable to
the estimated kon = 0.0205 sꢁ1 for binding of 10 nM [3H]MLA
to the a7 nAChR from rat membranes.17
Scheme 1 Synthesis of model maleimides.
Incubation of 1 mM analogue maleimide 4 with the L90T
receptor for 4 min reduced the currents elicited by ACh
(I(N) = 0.88 ꢀ 0.06, n = 3, p o 0.05 z-test cf. 1, Fig. 1). This
reduction may result from slower wash out of 4 compared to
MLA maleimide 3, or alternatively, from reaction of 4 with the
only other unpaired but sterically hindered cysteine in the
extracellular domain, C138. After 4 min incubation of 200 nM
analogue maleimide 4, the current elicited by ACh was signifi-
cantly and irreversibly reduced in oocytes expressing S189C :
L90T compared to L90T alone (I(N) = 0.60 ꢀ 0.04, n = 3,
p o 0.05). In contrast, the ACh responses of S188C : L90T
were not significantly reduced by analogue maleimide 4 com-
pared to L90T alone (I(N) = 0.86 ꢀ 0.09, n = 3). The first
order rate constant for the reaction of analogue maleimide 4 to
S189C : L90T was estimated to be k = 0.011 ꢀ 0.001 sꢁ1. These
results suggest that MLA 1 and the analogue 2 bind with
similar but not identical binding modes or alternatively, that
Loop-F on which S188 and S189 lie may have different
conformational preferences depending on which ligand is
bound. In the rat a7 nAChR homology model (Fig. 2, cyan),
S189 is hydrogen bonding to the succinimide moiety, with Ca
3.8 A from the unsubstituted carbon of the succinimide moiety
of MLA, and thus is positioned to react with maleimide
analogue 4 after introduction of a cysteine mutation. In
coupling/deprotection sequence from acid 11 afforded anthra-
nilate ester 12. Subsequent condensation with the appropriate
maleic anhydride gave the desired maleimides, albeit in a
lower yield.
Reaction of each maleimide with a 10-fold excess of thiol-
containing amino acids was conducted and maleimide con-
sumption was monitored by integration of the 1H NMR
spectrum. Confirmation of conjugate formation was obtained
by MS. Reaction with N-Boc-Cys-OMe as the thiol in CD3OD
established a relative rate of reaction as 5/6 > 7 c 9
(no reaction). The reactions with 5 and 6 were judged complete
prior to obtaining the NMR spectrum (3 min), while 9 showed
no reaction after 24 h by 1H NMR or MS. Reaction of N-Ac-
Cys-OH as the thiol in d6-DMSO established the relative rate
of reaction as 5/8 > 6 B 7 c 9 (no reaction). For 5 and 8 the
reaction was judged complete prior to obtaining the first NMR
spectrum (3 min), while 9 showed no reaction by NMR or MS
after 24 h. The bromo 6 and methyl 7 maleimides required 6 h
before complete consumption of the alkene proton was
observed. The relative rates are comparable to other reports13
and open avenues for the controlled deployment of substituted
maleimides as Michael acceptors in bioconjugation chemistry.
Based on the reactivity profile, ease of synthesis and close
structural resemblance to the methylsuccinimide present in
MLA 1 and analogue 2, the methylmaleimide was selected for
the development of electrophilic probes. Methyllycaconitine
maleimide 3 and analogue maleimide 4 were prepared by
carbodiimide coupling in one step from lycoctonine14 and
azabicyclic alcohol15 respectively in moderate yields.
From a homology model of the extracellular domain of the
rat a7 nAChR built with MLA 1 in the binding site (see
the ESI), two residues located on the complementary face of
the subunit interface S188 and S189 were selected for cysteine
mutagenesis. Further, to ensure that rapid desensitisation of
the a7 nAChR did not result in erroneous measurements,
a pore mutation changing the central leucine (L90T) was
introduced.16 This mutant increased the sensitivity of the
receptor to ACh (EC50 = 94 mM wt, 1.3 mM L90T) but did
not significantly alter the IC50 of MLA (IC50 = 238 pM wt,
228 pM L90T). The S188C and S189C mutants were created
from this L90T background. When ACh was applied to
Fig. 1 Inhibition of ACh-elicited current by MLA maleimide 3 and
analogue maleimide 4 at L90T (1 mM each, 1 mM ACh, black bars) and
mutant nAChRs (10 nM 3 and 200 nM 4 respectively, 5 mM ACh);
S188C : L90T, (grey bars) and S189C : L90T (white bars). Inhibition is
shown as the mean ꢀ sem relative current elicited by ACh after
incubation with the compound is complete.
c
6700 Chem. Commun., 2012, 48, 6699–6701
This journal is The Royal Society of Chemistry 2012

View Article Online
explore the nature of ligand binding to the a7 nAChR and
other receptor subtypes. Our results detailing the interaction of
MLA 1 with the a4b2 nAChR will be reported in due course.
We thank the Australian Research Council (DP0986469) for
financial support. MC thanks the Drug Research Academy,
University of Copenhagen, Denmark and the Australian
Academy of Science for travel support.
Notes and references
1 G. T. Hermanson, Bioconjugate Techniques, Academic Press,
San Diego, CA, 1996.
2 J. Kalia and R. T. Raines, Curr. Org. Chem., 2010, 14, 138.
3 J. M. Chalker, G. J. L. Bernardes, Y. A. Lin and B. G. Davis,
Chem.–Asian J., 2009, 4, 630.
4 G. G. Wilson and A. Karlin, Proc. Natl. Acad. Sci. U. S. A., 2001,
98, 1241.
5 J. G. Newell and C. Czajkowski, in Handbook of Neurochemistry
and Molecular Neurobiology, ed. A. Lajtha, G. Baker, S. Dunn and
A. Holt, Springer, Boston, MA, 2007, ch. 21, p. 439.
6 A. Karlin and M. H. Akabas, in Methods in Enzymology, Vol. 293,
Ion Channels Part B, ed. P. M. Conn, Academic Press, New York,
NY, 1998, ch. 8, p. 123.
7 K. Alarcon, A. Martz, L. Mony, J. Neyton, P. Paoletti,
M. Goeldner and B. Foucaud, Bioorg. Med. Chem. Lett., 2008,
18, 2765; C. Tahtaoui, M.-N. Balestre, P. Klotz, D. Rognan,
C. Barberis, B. Mouillac and M. Hibert, J. Biol. Chem., 2003,
278, 40010; Y. Yu, L. Shi and A. Karlin, Proc. Natl. Acad. Sci.
U. S. A., 2003, 100, 3907.
Fig. 2 Homology model of MLA 1 (pale cyan) bound rat a7 nAChR
(cyan) with S189 to succinimide hydrogen bond indicated (dotted line),
and loop sampled model (blue) after in-silico S188C mutation and
reaction with MLA maleimide 3 (light blue).
contrast, the S188 Ca is placed 7 A from the succinimide
moiety and the side chain points away from the binding site.
Rather than assuming a binding mode for MLA maleimide 3
different than that of MLA 1 in the X-ray structures,11 the
reactivity of the S188C: L90T mutant may suggest that Loop-F
is flexible and can change conformation so that S188 comes in
close proximity of the succinimide/maleimide moiety.
8 G. X. J. Quek, D. Lin, J. I. Halliday, N. Absalom, J. I. Ambrus,
A. J. Thompson, M. Lochner, S. C. R. Lummis, M. D. McLeod
and M. Chebib, ACS Chem. Neurosci., 2010, 1, 796.
9 R. E. Hibbs and E. Gouaux, Nature, 2011, 474, 54; R. J. C. Hilf
and R. Dutzler, Nature, 2008, 452, 375; N. Bocquet, H. Nury,
M. Baaden, C. Le Poupon, J.-P. Changeux, M. Delarue and
P.-J. Corringer, Nature, 2008, 457, 111; N. Unwin, J. Mol. Biol.,
2005, 346, 967.
10 A. B. Smit, N. I. Syed, D. Schaap, J. van Minnen, J. Klumperman,
K. S. Kits, H. Lodder, R. C. van der Schors, R. van Elk,
B. Sorgedrager, K. Brejc, T. K. Sixma and W. P. M. Geraerts,
Nature, 2001, 411, 261; S. B. Hansen, T. T. Talley, Z. Radic and
P. Taylor, J. Biol. Chem., 2004, 279, 24197.
11 A. Nemecz and P. Taylor, J. Biol. Chem., 2011, 286, 42555;
S. B. Hansen, G. Sulzenbacher, T. Huxford, P. Marchot,
P. Taylor and Y. Bourne, EMBO J., 2005, 24, 3635.
12 J. L. Vanderhooft, B. K. Mann and G. D. Prestwich, Biomacro-
molecules, 2007, 8, 2883.
13 M. E. B. Smith, F. F. Schumacher, C. P. Ryan, L. M. Tedaldi,
D. Papaioannou, G. Waksman, S. Caddick and J. R. Baker, J. Am.
Chem. Soc., 2010, 132, 1960; L. M. Tedaldi, M. E. B. Smith,
R. I. Nathani and J. R. Baker, Chem. Commun., 2009, 6583.
14 M. S. Yunusov, E. M. Tsyrlina, E. D. Khairitdinova,
L. V. Spirikhin, A. Y. Kovalevsky and M. Y. Antipin, Russ. Chem.
Bull., 2000, 49, 1629.
15 D. Barker, M. D. McLeod, M. A. Brimble and G. P. Savage,
Tetrahedron Lett., 2001, 42, 1785.
16 D. Bertrand, A. Devillers-Thiery, F. Revah, J. L. Galzi, N. Hussy,
C. Mulle, S. Bertrand, M. Ballivet and J. P. Changeux, Proc. Natl.
Acad. Sci. U. S. A., 1992, 89, 1261.
17 A. R. L. Davies, D. J. Hardick, I. S. Blagbrough, B. V. L. Potter,
A. J. Wolstenholme and S. Wonnacott, Neuropharmacology, 1999,
38, 679.
18 M. P. Jacobson, D. L. Pincus, C. S. Rapp, T. J. F. Day, B. Honig,
D. E. Shaw and R. A. Friesner, Proteins: Struct., Funct., Bioinf.,
2004, 55, 351.
To illustrate this possibility without violating common
geometric restraints for the protein backbone, Loop-F was
sampled using the Prime18 extended loop sampling protocol
(see the ESI) with a 3 A distance constraint between the
oxygen of S188 and the C4 of the succinimide moiety. As a
result, a loop sampled model with S188 stabilised by hydrogen
bonding to the oxygen of the succinimide moiety was obtained
with a Ramachandran plot of a quality comparable to that of
the original model. The loop-sampled model is shown in Fig. 2
(blue) for the S188C mutant following addition to MLA
maleimide 3 and energy minimisation. While the exact position
and functional importance of Loop-F in nAChRs is uncertain,
there is substantial experimental support for Loop-F flexibility,
e.g. different loop conformations in AChBP structures depos-
ited in the RCSB Protein Data Bank,19 high B-values and low
electron densities.20 Further, the length of Loop-F is different in
the a7 nAChR receptor compared to the templates. Thus the
loop-sampled model suggests an alternate F-Loop conforma-
tion supported by reaction data that is distinct from that
observed in the original model and X-ray crystal structures.11
In conclusion, we have synthesised and studied the reactivity
of five alkene-substituted maleimides (5–9) as Michael
acceptors. From this work, thiophilic reagents 3 and 4 were
developed from the known a7 nAChR antagonists MLA 1 and
the simplified analogue 2, and these were shown to covalently
attach to cysteine mutants selected based on a structural
model. This study of functional nAChRs using cysteine muta-
genesis in combination with thiophilic ligands complements
and confirms information afforded from recently reported
AChBP crystal structures. The study reveals differences in
reactivity that suggest different binding modes or more likely,
different Loop-F conformations depending on which ligand is
bound. The reactive probes 3 and 4 provide further scope to
19 H. M. Berman, J. Westbrook, Z. Feng, G. Gilliland, T. N. Bhat,
H. Weissig, I. N. Shindyalov and P. E. Bourne, Nucleic Acids Res.,
2000, 28, 235.
20 L. A. H. Rohde, P. K. Ahring, M. L. Jensen, E. Ø. Nielsen,
D. Peters, C. Helgstrand, C. Krintel, K. Harpsøe, M. Gajhede,
J. S. Kastrup and T. Balle, J. Biol. Chem., 2012, 287, 4248.
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 6699–6701 6701