Modulating cocaine vaccine potency through hapten fluorination

Article abstract of DOI:10.1021/ja400356g

Cocaine addiction is a long-lasting relapsing illness characterized by cycles of abuse, abstinence, and reinstatement, and antibody-based therapies could be a powerful therapeutic approach. Herein, we explored the possibility of using halogenated cocaine haptens to enhance the immunological properties of anti-cocaine vaccines. Three fluorine-containing cocaine haptens (GNF, GNCF and GN5F) and one chlorine-containing cocaine hapten (GNCl) were designed and synthesized, based upon the chemical scaffold of the only hapten that has reached clinical trials, succinyl norcocaine (SNC). Hapten GNF was found to retain potent cocaine affinity, and also elicit antibodies in a higher concentration than the parent structure SNC. Our data suggests that not only could strategic hapten fluorination be useful for improving upon the current cocaine vaccine undergoing clinical trials, but it may also be a valuable new approach, with application to any of the vaccines being developed for the treatment of drugs of abuse.

Full text of DOI:10.1021/ja400356g

Communication
pubs.acs.org/JACS
Modulating Cocaine Vaccine Potency through Hapten Fluorination
Xiaoqing Cai, Kyoji Tsuchikama, and Kim D. Janda*
Department of Chemistry, Department of Immunology and Microbial Science, The Skaggs Institute for Chemical Biology, The Worm
Institute of Research and Medicine (WIRM), The Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, California
92037, United States
S
* Supporting Information
ABSTRACT: Cocaine addiction is a long-lasting relapsing
illness characterized by cycles of abuse, abstinence, and
reinstatement, and antibody-based therapies could be a
powerful therapeutic approach. Herein, we explored the
possibility of using halogenated cocaine haptens to
enhance the immunological properties of anti-cocaine
vaccines. Three fluorine-containing cocaine haptens
(GNF, GNCF and GN5F) and one chlorine-containing
cocaine hapten (GNCl) were designed and synthesized,
based upon the chemical scaffold of the only hapten that
has reached clinical trials, succinyl norcocaine (SNC).
Hapten GNF was found to retain potent cocaine affinity,
and also elicit antibodies in a higher concentration than
the parent structure SNC. Our data suggests that not only
could strategic hapten fluorination be useful for improving
upon the current cocaine vaccine undergoing clinical trials,
but it may also be a valuable new approach, with
application to any of the vaccines being developed for
the treatment of drugs of abuse.
Figure 1. Structures of cocaine, hapten SNC and halogen-containing
cocaine haptens.
ocaine abuse and addiction remains a major medical and
1
Cpublic health problem. To date, no pharmacotherapies
Fluorine substitution is an established tool in bioorganic and
medicinal chemistry due to the unique properties of fluorine,
such as the comparable size of fluorine to hydrogen, the
superhydrophobicity of fluorocarbons, and fluorine’s unique
inductive effect and “polar hydrophobicity”.6 Incorporation of
fluorine atoms or fluorine-containing substituents is often used
to enhance ligand-binding affinity, and has recently been found
to enhance immune recognition.7 Indeed, immune response in
part is based on the T cell receptor (TCR) recognition of
antigenic molecules bound and presented by the major
histocompatibility complex (MHC). A weak interaction of
TCRs with antigen-MHC may fail to evoke a significant
immune response. There is increasing evidence that rationally
modified antigens can boost TCR binding and thereby
overcome the poor antigenicity of native antigens. Antigen-
fluorination has become one means to enhance TCR affinity
without significantly perturbing the composition or structure of
the antigen.8 Thus, various laboratories have prepared fluorine-
modified carbohydrates, peptides or glycopeptides, and in some
cases, these variants showed significant improvement in the
immunogenicity of the vaccine.9
have been approved for the treatment of cocaine dependence.
However, a number of direct/indirect agonists and antagonists
aimed to modulate or disrupt the drug’s effect at its site of
action have been investigated, but these have achieved very
limited success.2 In view of these limitations, interest has turned
to strategies that target the drug molecule itself, aiming to keep
the drug below its effective concentration at its site of action.
One such tactic using this approach is termed active
vaccination, wherein a cocaine vaccine is used to elicit
antibodies (Abs) for drug neutralization.2,3 We view active
vaccine design to be dependent upon three basic elements: a
drug-like hapten, immunogenic carrier and adjuvant.3,4 During
the past two decades, a small but intensive effort has been
devoted to exploring cocaine-like haptens to produce cocaine-
specific antibodies; yet only one cocaine vaccine, termed TA-
CD, which consists of succinyl norcocaine (SNC, Figure 1)
coupled to a recombinant cholera toxin B subunit using an
aluminum hydroxide gel as adjuvant, has reached clinical trials.5
Furthermore, this vaccine has offered limited therapeutic
efficacy for cocaine abstinence as a result of the high subject-
to-subject variability in antibody titers among participants.
Thus, there is an unmet need to engineer vaccine formulations
with improved immunogenicity that will validate vaccination as
a therapeutic strategy to treat cocaine abuse and addiction.
Received: January 11, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja400356g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
An Achilles’ heel seen with vaccines against drugs of abuse
has been poor immunogenicity. Herein, we detail using cocaine
as a drug of abuse platform scaffold to show how hapten-
halogenation can be implemented as a new tool for modulating
vaccine potency. The succinyl norcocaine hapten was chosen as
a starting structure to examine this hypothesis, as modification
on this structure could be carefully controlled and immuno-
logical consequences readily accounted for. Three fluorine-
containing homologues of SNC termed GNF, GNCF and
GN5F were judiciously chosen and synthesized (Figure 1), in
which the anticipated benzoyl ester dominant epitope was
substituted with fluorine(s) or a trifluoromethyl group. Last, we
also prepared a chlorine-containing cocaine hapten, GNCl, to
probe the importance of the halogen atom itself.
efficiency using MALDI-TOF MS. Importantly, no significant
difference in hapten coupling efficiency was observed (see
Table S1).
The efficacy of SNC-KLH, GNF-KLH, GNCl-KLH, GNCF-
KLH and GN5F-KLH immunoconjugates was assessed by
vaccination of groups of n = 5 Swiss Webster mice. Test groups
included a KLH negative control and the five cocaine hapten-
KLH conjugates. Each unique vaccine was administered on
days 0, 21, and 42 with bleeds taken one week post-vaccination.
Success of active vaccination is contingent upon both the
magnitude of the response as well as the affinity and specificity
of the antibodies generated.12 The magnitude of the immune
response was initially assessed by ELISA on microtiter plates
coated with SNC-BSA conjugate. While higher titers to each
individual hapten could have been observed through examina-
tion of each hapten-BSA conjugate, we opted to use SNC-BSA
as “generic” hapten for all vaccines tested, as we felt this
recourse would provide a quick metric for assessing cocaine-
antibody fidelity.
The synthesis of the halogenated cocaine haptens is
illustrated in Scheme 1. SNC was synthesized using a similar
Scheme 1. Synthesis of Halogenated Cocaine Haptens
The KLH control group showed no significant titer to SNC,
and thus, it was discarded and not investigated further. As seen
in Figure 2, the titers of the second bleed increased significantly
Figure 2. Total SNC-specific antibody titers from the sera of the first
and second bleeds of mice immunized with SNC-KLH, GNF-KLH,
GNCF-KLH, GNCl-KLH and GN5F-KLH conjugates, respectively.
Each column represents the average antibody titer in the sera of five
replicate mice.
strategy as reported,10 while the synthesis of each halogenated
hapten commenced with ecgonine methyl ester 2, which was
prepared from (−)-cocaine hydrochloride in two steps.11
Benzoylation of 2 was achieved by the use of halogenated
benzoyl chloride, Et3N and DMAP in dichloromethane,
furnishing compounds 3a−3d in moderate to high yields.
The demethylation of each of these tropane alkaloids
compounds was accomplished by forming a carbamate
intermediate before treatment with zinc dust, providing
norcocaine derivative 4a−4d. Finally, the target compounds
GNF, GNCl, GNCF and GN5F were obtained through N-
acylation of amine 4a−4d with succinic anhydride in the
presence of Et3N in DMF. To generate the immunoconjugates
for vaccination, the obtained haptens GNF, GNCl, GNCF and
GN5F as well as SNC were coupled to the carrier protein
keyhole limpet hemocyanin (KLH) under standard protein
conjugation conditions.12 Prior to administration, the resultant
protein conjugates were formulated with SAS (Sigma Adjuvant
System), a stable oil-in-water emulsion derived from bacterial
and mycobacterial cell walls. Finally, each hapten was
conjugated to bovine serum albumin (BSA) for ELISA
microtiter plate coating and monitoring of the coupling
compared to that of the first bleed. As expected, the titers
elicited by SNC-KLH were significant; however, GNF-KLH
also induced a strong titer to SNC-BSA. Further, analysis of this
ELISA data presented us with additional observations that were
not readily predictable. Thus, fluorine appears to be a “guiding”
immune modulator as GNF and GNCF provided excellent
titers. GN5F titers were greatly diminished, and we suspect this
was due to hydrolytic lability of the pentafluorobenzoyl ester.
Interestingly, titers induced by GNCl were vastly reduced in
comparison to GNF. While steric effects could play a role (rw =
1.35 Å for fluorine versus rw = 1.75 Å for chlorine),13a this
appears not to be the central tenet controlling titer magnitude
as the titer induced by GNCF (rw = 2.20 Å)13b was not altered
dramatically. Furthermore, DFT calculations provided no
evidence that the confirmation of SNC was perturbed by any
of these halogen substitutions (see Figure S2). Intriguingly, this
lack of GNCl immunogenicity is in contrast to a recent
publication presenting sialylated tumor-associated carbohydrate
B
dx.doi.org/10.1021/ja400356g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
antigens containing unnatural phenylacetyl derivatives where a
p-chlorophenyl immunogen was found to be the most
promising candidate.9a
To obtain a more accurate account of each vaccine’s ability to
bind cocaine, we conducted a soluble radioimmunoassay (RIA),
which was then normalized thus allowing for direct comparison
of values between test groups. As shown in Table 1, polyclonal
aromatic ring or the aliphatic scaffold of cocaine structure, or
multiple-fluorination upon cocaine’s scaffold is worthy of
further investigation.
In summary, we have synthesized a series of halogenated
cocaine haptens, including GNF, GNCF, GN5F and GNCl, as
well as their KLH conjugates. The immunological properties of
the synthetic immunoconjugates as vaccines were evaluated and
compared to the SNC hapten undergoing clinical trials.
Monosubstitution with fluorine at the 4′-position on the
benzoyl ring of SNC resulted in a novel hapten GNF, which
was found to elicit antibody concentrations to cocaine greater
than the parent structure SNC and still retain potent cocaine
affinity. TCR recognition of fluorinated antigenic peptides has
been shown as a way to selectively modulate TCR binding
affinity. Whether such a basis of immune recognition is
operative for these fluorinated haptens will require additional
research to elucidate hapten-immune molecular recognition.
However, what is clear is that fluorine’s unique properties6
provide a promising new hapten design strategy for the
development of active vaccines against all drugs of abuse.
Table 1. Average Affinity of Antisera from Immunized Mice
a
against Cocaine As Determined by Equilibrium Dialysis
hapten
Kd
[Ab]
(mean SD, nM)
(mean SD, μg/mL)
SNC
5.81 0.48
7.97 0.22
148.47 3.61
181.53 6.95
107.31 14.69
75.86 1.06
GNF
GNCF
GNCl
9.86 0.074
101.91 12.97
a
All assays were performed in triplicate. SD, standard deviation.
Abs elicited by SNC, GNF and GNCF displayed comparable
affinity to cocaine between 5.8 and 9.8 nM. Unanticipated,
vaccination with GNCl produced Abs with a rather poor
cocaine-binding affinity (∼102 nM). To begin to rationalize
these findings, we note that GNF and GN5F haptens contain
fluorine, the most electronegative atom among our tested
compounds, followed by the GNCF hapten possessing the
trifluoromethyl group and finally GNCl possessing chlorine,
which has the least electron-withdrawing effect.14 These results,
while thought-provoking, do not establish a complete
physicochemical basis for the binding affinity of the induced
Abs. However, the correlation seen between the electro-
negativity of the substituents and the antibody binding affinity
suggests that the electronic properties of the 4′-substituent on
the benzoyl ring could be one critical determinant for affinity
seen with these unique cocaine vaccines. Last, for hapten
GN5F, the elicited antibodies were found to have no binding
affinity for cocaine. Again, we assume this was due to hapten
instability.
The benefits of active vaccination upon a drug’s pharmaco-
dynamics should be judged by not only antibody affinity, but
also antibody concentration. Cocaine-specific IgG concen-
tration can also be determined from RIA analysis, as shown in
Table 1.15 Excitingly, the highest cocaine-specific IgG
concentration in sera was observed with Abs induced by
GNF, which was found to be approximately 181 μg/mL. It is
noteworthy that this is an increase of about 20% in antibody
concentration compared with the clinical hapten SNC. Again,
GNCl-KLH demonstrated the lowest immunogenicity among
these four viable immunoconjugates; the antibody concen-
tration in sera was about 60% lower than GNF. Remarkably,
fluorination at the 4′-position of the benzoyl ring of SNC was
able to not only maintain cocaine affinity, but also improve
upon antibody concentration to cocaine. This latter finding is
extremely pertinent, as documented active immunization
regimens for abused substances have elicited poor or variable
immune stimulation, and it is this failure to generate
therapeutic antibody concentrations that has thwarted the
approval of anti-drug vaccines for pharmacotherapy. The fact
that monofluorination of hapten SNC provided equivalent
binding affinity, yet superior antibody concentrations, supports
the idea of using strategically fluorinated haptens for the
development of cocaine vaccines with robust immunogenicity.
In this regard, fluorination at alternative positions, either on the
ASSOCIATED CONTENT
* Supporting Information
Synthetic protocols, assay methods, additional biological assays
and characterization of new compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
kdjanda@scripps.edu
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We gratefully acknowledge the Skaggs Institute for Chemical
Biology and the National Institute on Drug Abuse (DA 08590
to K.D.J.) for financial support. We also thank Dr. Sacha Javor
for assistance with the modeling study and insightful
discussions.
REFERENCES
■
(1) National Institute of Drug Abuse; National Household Survey on
Drug Abuse, Population Estimates; U.S. Dept. of Health and Human
Services: Rockville, MD, 2001.
(2) Carrera, M. R. A.; Meijler, M. M.; Janda, K. D. Bioorg. Med. Chem.
2004, 12, 5019−5030.
(3) Moreno, A.; Janda, K. D. Pharmacol., Biochem. Behav. 2009, 92,
199−205.
(4) (a) Meijler, M. M.; Matsushita, M.; Wirsching, P.; Janda, K. D.
Curr. Drug Discovery Technol. 2004, 1, 77−89. (b) Moreno, A.; Janda,
K. D. Expert Rev. Vaccines 2011, 10, 1637−1639. (c) Janda, K. D.;
Treweek, J. B. Nat. Rev. Immunol. 2012, 12, 67−72.
(5) (a) Martell, B. A.; Orson, F. M.; Poling, J.; Mitchell, E.; Rossen,
R. D.; Gardner, T.; Kosten, T. R. Arch. Gen. Psychiatry 2009, 66,
1116−1123. (b) Haney, M.; Gunderson, E. W.; Jiang, H.; Collins, E.
D.; Foltin, R. W. Biol. Psychiatry 2010, 67, 59−65. (c) Kosten, T. R.;
Rosen, M.; Bond, J.; Settles, M.; Roberts, J. S. C.; Shields, J.; Jack, L.;
Fox, B. Vaccine 2002, 20, 1196−1204. (d) Martell, B. B.; Mitchell, E.;
Poling, J.; Gonsai, K.; Kosten, T. R. Biol. Psychiatry 2005, 58, 158−164.
(6) (a) Kirk, K. Org. Process Res. Dev. 2008, 12, 305−321. (b) Filler,
R.; Saha, R. Future Med. Chem. 2009, 1, 777−791. (c) Biffinger, J. D.;
Kim, H. W.; Dimagno, S. G. ChemBioChem 2004, 5, 622−627.
(d) Zhou, P.; Zou, J.; Tian, F.; Shang, Z. J. Chem. Inf. Model. 2009, 49,
2344−2355.
C
dx.doi.org/10.1021/ja400356g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
(7) (a) Muller, K.; Faeh, C.; Diederich, F. Science 2007, 317, 1881−
1886. (b) Gomez-Nunez, M.; Haro, K. J.; Dao, T.; Chau, D.; Won, A.;
Escobar-Alvarez, S.; Zakhaleva, V.; Korontsvit, T.; Gin, D. Y.;
Scheinberg, D. A. PLoS One 2008, 3, e3938.
(8) Piepenbrink, K. H.; Borbulevych, O. Y.; Sommese, R. F.;
Clemens, J.; Armstrong, K. M.; Desmond, C.; Do, P.; Baker, B. M.
Biochem. J. 2011, 423, 353−361.
(9) (a) Wang, Q.; Guo, Z. Med. Chem. Lett. 2011, 2, 373−378.
(b) Yang, F.; Zheng, X.-J.; Hua, C.-X.; Wang, Y.; Zhang, Y.; Ye, X.-S.
ACS Chem. Biol. 2011, 6, 252−259. (c) Wagner, S.; Mersch, C.;
Hoffmann-Roder, A. Chem.Eur. J. 2010, 16, 7319−7330. (d) Hoff-
̈
mann-Roder, A.; Kaiser, A.; Wagner, S.; Gaidzik, N.; Kowalczyk, D.;
̈
Westerlind, U.; Gerlitzki, B.; Schmitt, E.; Kunz, H. Angew. Chem., Int.
Ed. 2010, 49, 8498−8503.
(10) (a) Fox, B. S.; Kantak, K. M.; Edwards, M. A.; Black, K. M.;
Bollinger, B. K.; Alison, J. B.; French, T. L.; Thompson, T. L.; Schad,
V. C.; Greenstein, J. L.; Gefter, M. L.; Exley, M. A.; Swain, P. A.;
Briner, T. J. Nat. Med. 1996, 2, 1129−1132. (b) Lazer, E. S.; Aggarwal,
N. D.; Hite, G. J.; Nieforth, K. A.; Kelleher, R. T.; Speralman, R. D.;
Schuster, C. R.; Wolverton, W. J. Pharm. Sci. 1978, 67, 1656−1658.
(11) Matsushita, M.; Hoffman, T.; Ashley, J. A.; Zhou, B.; Wirsching,
P.; Janda, K. D. Bioorg. Med. Chem. Lett. 2011, 11, 87−90.
(12) Moreno, A. Y.; Mayorov, A. V.; Janda, K. D. J. Am. Chem. Soc.
2011, 133, 6587−6595.
(13) (a) Bondi, A. J. Phys. Chem. 1964, 68, 441−451. (b) Smart. B. E.
Chapter 6, Properties of Fluorinated Compounds. In Chemistry of
Organic Fluorine Compounds II, A Critical Review; Hundlicky, M.,
́
Pavlath, A. E, Eds.; American Chemical Society: Washington, DC,
1995; p 979.
(14) Well, P. R. Prog. Phys. Org. Chem. 1968, 6, 111−145.
(15) Muller, R. Methods Enzymol. 1983, 92, 589−601.
̈
D
dx.doi.org/10.1021/ja400356g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX