Synthesis and biological evaluation of a lipopeptide-based methamphetamine vaccine

Article abstract of DOI:10.1016/j.cclet.2020.10.038

We describe an application of carrier protein-free strategy in constructing a fully synthetic methamphetamine (METH) vaccine that contains three components: Toll-like receptor 2 ligand, Th2 epitope, and METH hapten. The immunological evaluation in mice revealed high titers of METH-specific antibodies induced by the construct and the activation of humoral immunity that would be beneficial for neutralization and clearance of the METH molecule. Behavioral experiments indicated that the synthetic vaccine attenuated the acquisition of METH-induced conditioned place preference and inhibited the initiation and expression of METH-induced locomotor sensitization. These results demonstrate that the lipopeptide-based vaccine has invoked an immune response and showed the potential of preventing the rewarding and psychoactive effects of METH.

Full text of DOI:10.1016/j.cclet.2020.10.038

Journal Pre-proof
Synthesis and biological evaluation of a lipopeptide-based
methamphetamine vaccine
Fan Liao, Hanxuan Wang, Yuankun Dao, Kai Yuan, Jiazhen Lu, Jie
Shi, Ying Han, Suwei Dong, Lin Lu
PII:
S1001-8417(20)30621-5
https://doi.org/10.1016/j.cclet.2020.10.038
CCLET 5929
DOI:
Reference:
To appear in:
Chinese Chemical Letters
Received Date:
Revised Date:
Accepted Date:
21 August 2020
9 October 2020
23 October 2020
Please cite this article as: { doi: https://doi.org/
This is a PDF file of an article that has undergone enhancements after acceptance, such as
the addition of a cover page and metadata, and formatting for readability, but it is not yet the
definitive version of record. This version will undergo additional copyediting, typesetting and
review before it is published in its final form, but we are providing this version to give early
visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal
pertain.
© 2020 Published by Elsevier.

Chinese
Chemical
Letters
journal homepage: www.elsevier.com
Communication
Synthesis and biological evaluation of a lipopeptide-based methamphetamine vaccine
Fan Liaoa, Hanxuan Wangc, Yuankun Daoc, Kai Yuana, Jiazhen Lud,e, Jie Shib, Ying Hanb,*
yinghan@bjmu.edu.cn , Suwei Dongc, * dongsw@pku.edu.cn, Lin Lu a,d,* linlu@bjmu.edu.cn
aPeking University Sixth Hospital, Peking University Institute of Mental Health, NHC Key Laboratory of Mental Health (Peking University), National Clinical
Research Center for Mental Disorders (Peking University Sixth Hospital), Beijing 100191, China
bNational Institute on Drug Dependence and Beijing Key Laboratory of Drug Dependence, Peking University, Beijing 100191, China
cState Key Laboratory of Natural and Biomimetic Drugs, and Department of Chemical Biology, School of Pharmaceutical Sciences, Peking University, Beijing
100191, China
dPeking-Tsinghua Center for Life Sciences and PKU-IDG/McGovern Institute for Brain Research, Academy for Advanced Interdisciplinary Studies Peking
University, Beijing 100871, China
eSchool of Basic Medicine, Tongji Medical College, Huazhong University of Science and Technology, Wuhan 430030, China
Corresponding authors.
Graphical Abstract
A fully synthetic lipopeptide-based methamphetamine vaccine invoked an immune response and demonstrated the potential of
preventing the rewarding and psychoactive effects of methamphetamine.
ABSTRACT
We describe an application of carrier protein-free strategy in constructing a fully synthetic methamphetamine (METH) vaccine that contains three components:
Toll-like receptor 2 ligand, Th2 epitope, and METH hapten. The immunological evaluation in mice revealed high titers of METH-specific antibodies induced by
the construct and the activation of humoral immunity that would be beneficial for neutralization and clearance of the METH molecule. Behavioral experiments
indicated that the synthetic vaccine attenuated the acquisition of METH-induced conditioned place preference and inhibited the initiation and expression of
METH-induced locomotor sensitization. These results demonstrate that the lipopeptide-based vaccine has invoked an immune response and showed the potential
of preventing the rewarding and psychoactive effects of METH
Keywords: METH vaccine, Carrier protein-free, Fully synthesis, Lipopeptide, Behavioral model
The number of methamphetamine (METH) abusers has rapidly
grown worldwide in recent years [1]. The consecutive use of
METH usually produces direct effects on the central nervous
system, contributing to euphoria, intense emotions, alterations of
self-esteem, and greater alertness, aggression, and sexual drive
[2-4]. Methamphetamine use can also quickly lead to addiction,
increasing the risk of overdose, which is a significant source of
mortality across multiple populations. However, treatments for
METH addiction are limited and mainly consist of behavioral
and cognitive therapy [5], and no medications have been
approved to date. Thus, more effective treatments are essential to
prevent METH addiction and relieve associated symptoms.
addiction have also been investigated [7]. Methamphetamine-
specific antibodies that are induced by such vaccines may
neutralize circulating METH molecules and prevent their
penetration into the brain. Due to the highly simplified structure
of the METH molecule and its low immunogenicity, triggering a
sufficient immune response to generate the necessary specific
antibodies has been extremely difficult. Currently, the
development of METH vaccines relies heavily on the conjugation
of varied haptens to carrier proteins [7, 8], such as keyhole limpet
hemocyanin (KLH) [9]. However, the introduction of foreign
proteins elicits other concerns. First, the high immunogenicity
and complexity of the carrier protein and requisite linker moiety
may result in high titers of non-METH-specific antibodies, which
would lead to immune suppression of the selected antigen [10].
Second, the conjugation of a small molecule to the large protein
Immunotherapy has been applied for the treatment of various
types of diseases, including cancer [6]. Based on prominent
results from exploiting the immune system, vaccines for METH

usually makes equivalency difficult to control, which may lead to
issues with quality control. Third, heterogeneity of the carrier
protein may cause systemic toxicity [10,11]. Therefore, new
strategies are needed toward the development of novel METH
vaccines for clinical use. The present study describes our efforts
to use a carrier protein-free strategy [12,13] to construct and
immunologically and behaviorally evaluate a fully synthetic
METH vaccine.
compound 10 (41%). Hydrogenolysis of the obtained nitro-
substituted METH derivative, followed by amide coupling with
N-Boc-6-aminohexanoic acid, led to the formation of 12 in a
decent yield (57%) over two steps. After the removal of both
protecting groups, a properly functionalized hapten METH
derivative 7 was obtained, appending a linker that is ready for
connecting to the C-termini of the peptides. On the other hand,
Pam3Cys-OH (13) was synthesized based on the previously
reported methods using racemic glycidol as the starting materials
[25,26], as it was demonstrated that the stereochemical
configuration of the glycerol in a synthetic MUC1 antitumor
vaccine was inconsequential to its immunogenicity [26].
Fig. 1. (A) Proposed mechanism by which the conjugate induces
humoral immunity. (B) Design of the lipopeptide-based
methamphetamine vaccine.
Scheme 1. Synthesis of hapten 7.
Considering recent developments in multicomponent vaccines
that contain haptens, T-helper (Th) epitopes, and immunological
activators [14-17], we envisioned that the absence of carrier
protein might be beneficial for improving immune specificity for
the METH antigen and biosafety (Fig. 1A). Thus two three-
component carrier protein-free constructs 1a and 1b were
designed, each consisting of a ligand to the Toll-like receptor 2
(TLR2) [18, 19], a Th2 epitope-containing peptide sequence, and
a METH hapten (Fig. 1B). Pam3Cys has been utilized in several
vaccine candidates as a built-in adjuvant to activate the innate
immune system [16,18,20], presumably as stimuli to the
maturation of dendritic cells and subsequent activation of helper
T cells [21]. To further induce the production of different
subtypes of antibodies from plasma cells [22], the DRAHYNI
peptide sequence, deriving from the E7 open reading frame
protein of human papillomavirus (HPV) type 16, was
incorporated to the construct as the Th epitope. With the METH
hapten serving as a B cell epitope, we hoped that the designed
construct could trigger an immune response, in which T-cell
receptors (TCRs) on the T-helper cell could recognize the Th
epitope that was presented by MHCII and assist with the
differentiation and maturation of naive B cells into plasma cells
[23,24], leading to the production of specific anti-METH IgG
antibodies by strengthening humoral immunity (Fig. 1A).
Scheme 2. Synthesis of vaccine 1b.
Two Th2 epitope-containing peptide sequences, DRAHYNI
and SKKKKDRAHYNI (Scheme 2), were synthesized in a
straightforward manner using Fmoc-based solid-phase peptide
synthesis (SPPS). Notably, a glycine residue was installed at the
peptide C-terminus to avoid possible epimerization in the later
coupling step for connecting the hapten. Moreover, 2-chlorotrityl
chloride resin was utilized for peptide synthesis, which allowed
the cleavage from resin without removal of all other protecting
groups, and ensured the selective conjugation of compound 7 to
the C-terminus of lipopeptide 5. In practice, amide coupling
between amine 4 and Pam3Cys-OH (13) on the resin was
The synthesis of the designed vaccine was initiated by
preparing the METH hapten with a proper linker used to connect
the Th2 epitope-containing peptide. Accordingly, the synthesis of
compound
7
was carried out by first derivatize D-
methamphetamine (8) [9], whose secondary amine was protected
with a trifluoro acetyl (TFA) group using TFAA (Scheme 1),
leading to the generation of compound 9 in an 83% yield.
Nitration reaction with concentrated nitric acid gave rise to

achieved using HATU and NMM, followed by cleavage from the
resin using a DCM/TFE/AcOH cocktail to give compound 5 with
only the C-terminal carboxylate exposed for further reactions.
After amidation with hapten 7, global deprotection using the TFA
cocktail (TFA/TIS/water, 95:2.5:2.5) afforded the crude vaccine
construct. It was observed that crude 1a displayed poor solubility
and dissolved only in DMSO, which made it impossible to be
further confirmed the self-adjuvanting role of Pam3Cys in the
synthetic vaccine. On the other hand, the vaccine with Freund’s
adjuvant produced significantly higher METH-specific antibody
titers (1:10000 to 1:25000), whereas the vaccine with Alum or
PBS alone produced anti-METH antibodies at low titers,
indicating that Freund’s adjuvant may play an important role in
the process of the immune response induced by this METH
vaccine. By monitoring the mice immunized with vaccine 1b and
Freund’s adjuvant, the assay showed that METH-specific
antibody titers reached the highest level after the fourth
immunization and remained relatively stable for at least one and
half months (Fig. 2D). The persistent antibodies generated could
be beneficial for preventing addictive behaviors and relapse, as
they may sequester METH in the blood and reduce its
distribution to the brain [31].
purified
using reverse-phase
high-performance
liquid
chromatography (HPLC). In contrast, construct 1b with extra
amino acids SKKKK in the sequence exhibits significantly
improved solubility and could be purified on HPLC, affording
the desired product in good yield (21%). These results indicate
that the introduction of multiple lysines in the sequence is
necessary for the preparation of pure synthetic vaccines with
hydrophobic hapten and lipid chains [27].
With the pure vaccine construct in hand, we evaluated the
immunogenic capability of the vaccine in mice and the possible
combination with suitable adjuvants. Two widely studied
immunostimulants, Alum, and Freund’s adjuvant were chosen in
our study, as they have been utilized in the developments of
vaccines to drugs of abuse and addiction [28]. Female mice (6- to
7-week-old BALB/c) were divided into four groups (n = 6/group)
and subcutaneously injected with different sample solutions (30
μg of vaccine 1b with phosphate-buffered saline (PBS), the
vaccine formulated with Alum, the vaccine with Freund’s
adjuvant [29], and PBS control) on days 0, 14, 28, and 42. Serum
was collected on days 35, 49, 59, 69, 80, and 94 (Fig. 2A).
Antibody titers were analyzed using an enzyme-linked
immunosorbent assay (ELISA) coated with a BSA-METH hapten
conjugate [9] (Fig. 2B). The experimental results showed that the
synthetic vaccine itself was able to stimulate the secretion of IgG
antibodies, where the IgG titers elicited were tenfold higher than
that of the PBS control group (Fig. 2C). This observation is in
accord with the previous studies on the synthetic vaccine
constructs containing Pam3Cys, which induce strong humoral
immunity without external adjuvant existed [18, 30]. This data
In order to obtain more information on the possible types of
activation for different vaccine formulations, we further
investigated the isotypes and subtypes of the generated antibodies
(Fig. 2E). In the cases of the mice vaccinated with Freund’s
adjuvant, a higher level of IgG2a antibodies were elicited than
that of the IgG1 type (IgG2a ≥ IgG1), indicating that FcγRIIb
might be activated [32]. In contrast, the levels of IgG subtypes in
the group vaccinated without adjuvant were approximately the
same (IgG2a = IgG1), suggesting that the FcγRIIa signaling
pathway was likely activated, thus enhancing antigen processing
and presentation and inducing a T-cell response [32-34]. These
results are also in accord with the literature precedent showing
the use of Freund’s adjuvant promotes an increase in the level of
IgG2a [35]. Moreover, the IgG1 and IgG2 isotypes are believed
to play crucial roles in antibody-dependent humoral immunity
[23,36,37]. Therefore, the utilized Th2 epitope seemed to be
critical for our vaccine for
a class switch from other
immunoglobulins to IgG antibodies, consistent with previous
studies of carbohydrate vaccines without Th epitopes that led to
no or low titers of IgG antibodies [38-40].
Fig. 2. Immunogenic capability of the vaccine. (A) Timeline of active immunization. (B) BSA-linked methamphetamine immunogen used as
the coating antigen in ELISA. (C) Mean antibody titers on day 49. (D) Mean antibody titers at different time points after the third booster
immunization. (E) Antibody isotypes and subtypes of the conjugate vaccine. The data are expressed as mean ± SEM. *P < 0.05, n = 6/group.
To investigate the effects of this carrier protein-free vaccine on
METH-induced behaviors, conditioned place preference (CPP)
and locomotor sensitization models were utilized to assess
efficiency in mice in vivo. After active immunization, the mice

underwent CPP training (0.5 mg/kg METH, s.c.) for 8 days and
then were tested (Fig. 3A) [41]. The repeated-measures analysis
of variance (ANOVA) of CPP scores revealed a significant main
effect of test phase (F1,35 = 21.17, P < 0.0001). The mice that
received the emulsion of vaccine and Freund’s adjuvant had the
lowest CPP scores during test (Fig. 3B), suggesting that higher
anti-METH antibody titers may have slowed or reduced the entry
of METH molecules into the brain, thus demonstrating better
efficacy in preventing METH addiction. These results indicate
that this vaccine that was formulated with Freund’s adjuvant
prevented the acquisition of METH-induced CPP by neutralizing
the ability of METH to support contextual conditioning and
decreasing its rewarding effects.
In the animal study, we observed no significant difference in
animal states between mice with the non-adjuvanted vaccine and
the ones in the control group, such as their weight, vital signs,
and the capability of motion after immunization. The survival
rate of the vaccinated group is 100%. Moreover, none of the
animals with the non-adjuvanted vaccine experienced local
reactogenicity or adverse events in the following behavioral test,
indicating that this METH vaccine is generally safe for the
animal. These observations are in accord with the previous
studies showing the safety of full synthetic three-component
vaccine and lipopeptide [43-45].
Fig. 3. Immunization with the lipopeptide-based vaccine inhibited
methamphetamine-induced conditional place preference (CPP). (A)
Timeline of immunization, CPP training, and test. (B) Conditioned
place preference scores during baseline and test sessions. The data
are expressed as mean ± SEM. n = 12-13/group.
Fig. 4. Immunization with the lipopeptide-based vaccine prevented
the initiation and expression of methamphetamine-induced
locomotor sensitization. (A) Timeline of immunization and
locomotor activity test. (B) Locomotor activity during the training of
locomotor sensitization. (C) Locomotor activity after the
methamphetamine challenge on day 24. (D) Locomotor activity after
the methamphetamine challenge on day 34. The data are expressed
as mean ± SEM. **P < 0.01, ****P < 0.0001, n = 12-14/group.
A locomotor sensitization model was then used to evaluate the
effect of the vaccine on the psychomotor stimulant effect of
METH. The procedure for locomotor sensitization was based on
our previous studies [42], consisting of three phases: initiation,
withdrawal, and expression. During initiation, the mice received
METH injections (1 mg/kg, i.p.) for 14 consecutive days, and
locomotor activity was measured on days 1, 5, 10, and 14 (Fig.
4A). The repeated-measures ANOVA of locomotor activity
revealed significant effects of vaccine (F2,40 = 3.455, P < 0.05)
and test phase (F3,120 = 34.67, P < 0.0001) and a significant
vaccine × test phase interaction (F6,120 = 3.25, P < 0.01; Fig. 4B).
The post hoc analysis showed that control mice exhibited a
significant increase in locomotor activity after the METH
injection compared with the group that was vaccinated with
Freund’s adjuvant on the last day of the initiation phase (P <
0.0001). The group that was vaccinated without the adjuvant
exhibited a tendency toward a decrease in locomotor activity
compared with the control group. After 10 days of withdrawal,
locomotor activity was tested following a METH challenge (1
mg/kg, i.p.) on days 24 and 34 (Figs. 4C and 4D). The one-way
ANOVA revealed a significant main effect of vaccine (day 34:
F2,36 = 5.906, P < 0.01). The post hoc analysis showed a
significant difference between the PBS control and adjuvant-
vaccinated groups, indicating that the vaccine with Freund’
adjuvant inhibited the initiation, expression, and maintenance of
METH-induced locomotor sensitization. These behavioural
results suggest that our tri-component lipopeptide-based vaccine
effectively prevented the rewarding and psychoactive effects of
METH.
In summary, we have developed a METH vaccine candidate
that was composed of the TLR2 ligand Pam3Cys, a Th2 epitope,
and a METH hapten. Immunological and behavioural analyses of
this vaccine in combination with Freund’s adjuvant revealed that
a robust humoral immune response was induced, and behavioural
effects of METH were counteracted in mice. This carrier protein-
free strategy may be beneficial for amplifying the
immunogenicity of a hapten with a simple structure and
minimizing an undesirable immune response. The preparation of
this construct was also convenient and easily characterized, thus
ensuring homogeneity of the vaccine and demonstrating an
advantage with regard to quality control. Our findings provide
valuable information for the further development of anti-METH
vaccines, which may also be of use to the design of new vaccines
for other drugs of abuse that could be promising for the treatment
of addiction.
Declaration of competing interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to
influence the work reported in this paper.
Acknowledgments

[29] H. Cai, M.S. Chen, Z.Y. Sun, et al., Angew. Chem. Int. Ed. 52
(2013) 6106-6110.
[30] H. Cai, Z.Y. Sun, Z.H. Huang, et al., Chem. Eur. J. 19 (2012)
1962-1970.
This work was supported in part by the National Natural
Science Foundation of China (Nos. 81821092, and 81701312),
Beijing Municipal Science and Technology Commission (Nos.
Z181100001518005, and JQ18024), and National Key Research
and Development Program of China (Nos. 2017YFC0803608 and
2019YFC0118604). All of the experiments were conducted at
Peking University in accordance with National Institutes of
Health guidelines and were approved by the Institutional Animal
Care and Use Committee. We thank Drs. Xiaohui Zhang, Yuan
Wang, and Xiaomeng Shi (Peking University) for spectroscopic
assistance.
[31] M. Gooyit, P.O. Miranda, C.J. Wenthur, A. Ducime, K.D.
Janda, ACS Chem. Neurosci. 8 (2017) 468-472.
[32] R.J. Pleass, J.M. Woof, Trends Parasitol. 17 (2001) 545-551.
[33] K.G. Smith, M.R. Clatworthy, Nature Rev. Immunol. 10 (2010)
328-343.
[34] S. Bournazos, J.V. Ravetch, Immunity 47 (2017) 224-233.
[35] J.A. Fontes, J.G. Barin, M.V. Talor, et al., Immun. Inflamm.
Dis. 5 (2017) 163-176.
[36] K.B. Lorvik, C. Hammarström, M. Fauskanger, et al., Cancer
Res. 76 (2016) 6864.
[37] B. Pulendran, D. Artis, Science 337 (2012) 431.
[38] T. Toyokuni, B. Dean, S.P. Cai, et al., J. Am. Chem. Soc. 116
(1994) 395-396.
[39] S.J. Danishefsky, J.R. Allen, Angew. Chem. Int. Ed. 39 (2000)
836-863.
References
[40] S.F. Slovin, G. Ragupathi, C. Musselli, et al., J. Clin. Oncol. 21
(2003) 4292-4298.
[41] C.N. Haile, T.A. Kosten, X.Y. Shen, et al., Am. J. Addict. 24
(2015) 748-755.
[42] C.M. Xu, J. Wang, P.Wu, et al., J. Neurochem. 118 (2011) 126-
139.
[43] W.H. Li, Y.M. Li, Chem. Rev. (2020) doi: 10.1021/acs.
chemrev.9b00833.
[44] C. Durier, O. Launay, V. Meiffrédy, et al., AIDS 20 (2006)
1039-49.
[45] F. Steinhagen, T. Kinjo, C. Bode, D.M. Klinman, Vaccine 29
(2011) 3341-3355.
[1] A. Pabst, J.C. Castillo-Duque, A. Mayer, M. Klinghuber, R.
Werkmeister, Dent. J. 5 (2017) 29.
[2] A. Elkashef, F. Vocci, G. Hanson, et al., Subst. Abus. 29 (2008)
31-49.
[3] C.C. Cruickshank, K.R. Dyer, Addiction 104 (2009) 1085-1099.
[4] D.E. Rusyniak, Neurol. Clin. 29 (2011) 641-655.
[5] T. Brackins, N.C. Brahm, J.C. Kissack, J. Pharm. Pract. 24
(2011) 541-550.
[6] U. Sahin, O. Tureci, Science 359 (2018) 1355-1360.
[7] P.T.A Trends Guide Bremer, K.D. Janda, Pharmacol. Rev. 69
(2017) 298-315.
[8] R. Arora, C.N. Haile, T.A. Kosten, et al., Am. J. Addict. 28
(2019) 119-126.
[9] A.Y. Moreno, A.V. Mayorov, K.D. Janda, J. Am. Chem. Soc.
133 (2011) 6587-6595.
[10] P.W. Anderson, M.E. Pichichero, E.C. Stein, et al., J. Immunol.
(1989) 2464-2468.
[11] T. Buskas, Y. Li, G.J. Boons, Chem. Eur. J. 10 (2004) 3517-
3524.
[12] A.S. Cordeiro, M.J. Alonso, Pharm. Pat. Anal. 5 (2016) 49-73.
[13] L. Zeng, Z. Liao, W. Li, et al., Chin. Chem. Lett. 31 (2020)
1162-1164.
[14] H. Cai, Z.Y. Sun, M.S. Chen, et al., Angew. Chem. Int. Ed.
Engl. 53 (2014) 1699-1703.
[15] T. Buskas, P. Thompson, G.J. Boons, Chem. Commun. (2009)
5335-5349.
[16] S. Dziadek, A. Hobel, E. Schmitt, H. Kunz, Angew. Chem. Int.
Ed. 44 (2005) 7630-7635.
[17] A.S. Ashhurst, D.M. McDonald, C.C. Hanna, et al., J. Med.
Chem. 62 (2019) 8080-8089.
[18] B.L. Wilkinson, S. Day, L.R. Malins, V. Apostolopoulos, R.J.
Payne, Angew. Chem. Int. Ed. 50 (2011) 1635-1639.
[19] A.O. Aliprantis, R.B. Yang, M.R. Mark, et al., Science 285
(1999) 736.
[20] Z. Zhou, H. Lin, C. Li, Z. Wu, Chin. Chem. Lett. 29 (2018) 19-
26.
[21] T. Buskas, S. Ingale, G.J. Boons, Angew. Chem. Int. Ed. 44
(2005) 5985-5988.
[22] R.W. Tindle, G.J.P. Fernando, J.C. Sterling, I.H. Frazer, Proc.
Natl. Acad. Sci. U. S. A. 88 (1991) 5887-5891.
[23] G. Grødeland, E. Fossum, B. Bogen, Front. Immunol. 6 (2015)
367.
[24] R.N. Germain, Cell 76 (1994) 287-299.
[25] F. Reichel, A.M. Roelofsen, H.P.M. Geurts, et al., J. Am. Chem.
Soc. 121 (1999) 7989-7997.
[26] L. Shi, H. Cai, Z.H. Huang, et al., ChemBioChem 17 (2016)
1412-1415.
[27] W. Zeng, S. Ghosh, Y.F. Lau, L.E. Brown, D.C. Jackson, J.
Immunol. 169 (2002) 4905.
[28] C.R. Alving, G.R. Matyas, O. Torres, R. Jalah, Z. Beck,
Vaccine 32 (2014) 5382-5389.