A chemically contiguous hapten approach for a heroin-fentanyl vaccine

Article abstract of DOI:10.3762/bjoc.15.100

Background: Increased death due to the opioid epidemic in the United States has necessitated the development of new strategies to treat addiction. Monoclonal antibodies and antidrug vaccines provide a tool that both aids addiction management and reduces the potential for overdose. Dual drug vaccines formulated by successive conjugation or by mixture have certain drawbacks. The current study examines an approach for combatting the dangers of fentanyl-laced heroin, by using a hapten with one epitope that has domains for both fentanyl and heroin. Results: We evaluated a series of nine vaccines developed from chemically contiguous haptens composed of both heroin- and fentanyl-like domains. Analysis of the results obtained by SPR and ELISA revealed trends in antibody affinity and titers for heroin and fentanyl based on epitope size and linker location. In antinociception studies, the best performing vaccines offered comparable protection against heroin as our benchmark heroin vaccine, but exhibited attenuated protection against fentanyl compared to our fentanyl vaccine. Conclusion: After thorough investigation of this strategy, we have identified key considerations for the development of a chemically contiguous heroin-fentanyl vaccine. Importantly, this is the first report of such a strategy in the opioid-drug-vaccine field.

Full text of DOI:10.3762/bjoc.15.100

A chemically contiguous hapten approach for a
heroin–fentanyl vaccine
Yoshihiro Natori‡1,2, Candy S. Hwang‡1,3, Lucy Lin1, Lauren C. Smith1, Bin Zhou1
and Kim D. Janda*1
Full Research Paper
Open Access
Address:
Beilstein J. Org. Chem. 2019, 15, 1020–1031.
1Departments of Chemistry, Immunology and Microbial Science,
Skaggs Institute for Chemical Biology; The Scripps Research
Institute, 10550 N Torrey Pines Rd, La Jolla, CA, 92037, USA,
2Faculty of Pharmaceutical Sciences, Tohoku Medical and
Pharmaceutical University, Komatsushima 4-4-1, Aoba-ku, Sendai,
981-8558, Japan and 3Department of Chemistry, Southern
Connecticut State University, 501 Crescent St, New Haven, CT,
06515, USA
doi:10.3762/bjoc.15.100
Received: 28 February 2019
Accepted: 23 April 2019
Published: 03 May 2019
Associate Editor: I. R. Baxendale
© 2019 Natori et al.; licensee Beilstein-Institut.
License and terms: see end of document.
Email:
Kim D. Janda* - kdjanda@scripps.edu
* Corresponding author ‡ Equal contributors
Keywords:
antinociception; fentanyl; hapten; heroin; vaccine
Abstract
Background: Increased death due to the opioid epidemic in the United States has necessitated the development of new strategies to
treat addiction. Monoclonal antibodies and antidrug vaccines provide a tool that both aids addiction management and reduces the
potential for overdose. Dual drug vaccines formulated by successive conjugation or by mixture have certain drawbacks. The cur-
rent study examines an approach for combatting the dangers of fentanyl-laced heroin, by using a hapten with one epitope that has
domains for both fentanyl and heroin.
Results: We evaluated a series of nine vaccines developed from chemically contiguous haptens composed of both heroin- and
fentanyl-like domains. Analysis of the results obtained by SPR and ELISA revealed trends in antibody affinity and titers for heroin
and fentanyl based on epitope size and linker location. In antinociception studies, the best performing vaccines offered comparable
protection against heroin as our benchmark heroin vaccine, but exhibited attenuated protection against fentanyl compared to our
fentanyl vaccine.
Conclusion: After thorough investigation of this strategy, we have identified key considerations for the development of a chemical-
ly contiguous heroin–fentanyl vaccine. Importantly, this is the first report of such a strategy in the opioid–drug–vaccine field.
1020

Beilstein J. Org. Chem. 2019, 15, 1020–1031.
Introduction
In 2016, the Centers for Disease Control (CDC) estimated that of drug exposed to µ-opioid receptors that induce respiratory
approximately 20.4% of the U.S. population (e.g., 50 million depression. Indeed, anti-drug vaccines for individual conjugate
people) was suffering from chronic pain [1]. While using vaccines of fentanyl and heroin have already exhibited protec-
opioids to treat chronic pain is exceedingly effective in miti- tion in several animal models [3-8].
gating pain, the over prescription of opioids has substantially
contributed to a rise in opioid abuse and a growing overdose Antidrug vaccines targeting multiple, structurally distinct
crisis in the United States (CDC). In 2015, 63% of all drug species are not common and have utilized two tactics: 1) mixing
overdose deaths involved an opioid drug (CDC). Although the two individual vaccines together (admixture vaccine) [9,10]; or
overall rate of opioid prescription has declined in the past few 2) successive conjugation steps of one hapten and then the
years, heroin has become an increasingly common substitute for second to a carrier protein [11,12]. To specifically address the
prescription opioids due to the accessibility and inexpensive- escalating incidence of fentanyl-contamination in heroin, we in-
ness of the illicit drug. This upsurge is also attributed to an vestigated the efficacy of an admixture of individual heroin and
increase in illicitly manufactured fentanyl, a powerful synthetic fentanyl conjugate vaccines against challenges of 10–20%
opioid commonly used for treatment of pain in cancer and post- fentanyl in heroin [13,14]. Gratifyingly, we found these admix-
operative patients (CDC). The recent emergence of fentanyl ture vaccines were able to produce antibodies with excellent
resulted in a 264% increase in synthetic opioid deaths within the affinity toward both drugs, as well as six other analogues of
U.S. from 2012 to 2015, due to the fact that users are often fentanyl. Moreover, the heroin–fentanyl admixture vaccine was
unaware of the presence of fentanyl contaminating the supply of able to blunt nociception incurred from either drug, however,
heroin and cocaine to enhance their overall potencies. This we also realized the challenges of navigating through FDA
trend is particularly disconcerting due to the observation that approval of a combination vaccine. As such we pondered
the combinatorial consumption of 10% fentanyl in heroin whether singular antigen presentation of both drugs in close
significantly enhances the risk of overdose by prolonging the proximity would still be able to entice a competent
detrimental effects of respiratory depression and brain hypoxia antibody–drug response and memory generation. To test this
[2].
notion, we envisioned two drug haptens chemically joined to
produce a colocalized epitope by chemical design. While
Respiratory depression is a common side effect of opioid over- success of such an approach would have undeniable challenges,
dose and can lead to serious complications, including death. all data gathered would at the very least provide insight into
The current treatment for overdose is naloxone (NARCAN®), a several immunochemical questions including: (1) the impact of
competitive antagonist of µ-opioid receptors. Other treatments multivalent hapten scaffolding; (2) the influence on increasing
for opioid substance abuse focus on mitigating cravings and hapten density as compared to successive hapten-conjugations
managing long-term addiction. However, these treatments have and (3) how concomitant hapten-presentation would impact
an inherent risk for continued abuse. Combatting the complex immune response to each drug [15]. Herein, we describe the
nature of addiction requires a multifaceted strategy – several el- synthesis of nine unique chemically contiguous heroin–fentanyl
ements of the opioid crisis that need to be addressed include, haptens and their evaluation as conjugate vaccines to induce a
but are not limited to, new non-addictive opioid drugs, innova- polyclonal antibody response, as measured by ELISA, surface
tive treatments to prevent relapse, and preventing and/or plasmon resonance (SPR), and in a behavioral assay (Figure 1).
reversing overdose.
Results
Monoclonal antibodies and antidrug vaccines are immunophar- As a starting point for our exploration of chemically contiguous
macological treatments that address two of these highlighted drug–hapten conjugates, we envisioned generating a series of
needs: addiction management and a potential for reducing over- structurally diverse heroin–fentanyl haptens with a focus on
dose. By chemically attaching a non-immunogenic drug-like linker placement and bond distance between each drug’s chemi-
molecule to an immunogenic carrier protein, a vaccine is cal connections in order to probe the effect of the drug-epitope
formed that has the capacity to stimulate an immune response, structure on eliciting an immune response. Figure 2 illustrates
generating antibodies capable of binding the targeted drug. the nine haptens, as well as the original individual heroin and
Moreover, when a user consumes the drug, the antibodies circu- fentanyl haptens. To facilitate the preparation of a singular
lating in the blood stream can block the “high” by forming a hapten that preserves the structure of each parent drug, the
drug–antibody complex incapable of passing through the blood heroin- and fentanyl-like haptens were chemically joined by a
brain barrier. Ancillary, the vaccine’s immune response can series of what we hoped to be silent spacers. In sum, the feasi-
alleviate the potential for overdose by reducing the total amount bility of a singular heroin–fentanyl vaccine would be evaluated
1021

Beilstein J. Org. Chem. 2019, 15, 1020–1031.
Figure 1: Graphical summary of chemically contiguous opioid vaccine approach. A) Illustration of chemically contiguous heroin-fentanyl concept and
immunoconjugation strategy. B) Description of vaccine immunization schedule and metrics used to evaluate each vaccine’s efficacy.
through five spatial presentations, differentiated in Figure 2 by heroin domain was the primary constraint in constructing the
background color.
epitope itself, while the protein carrier linker locations necessi-
tated different syntheses of the fentanyl-like portion of the
The second consideration in our chemically contiguous hapten singular epitope. For example, the heroin-like domains of HF-3,
design was the location of the linker to the carrier protein. We HF-5, and HF-8 could share an intermediate, while the
chose three arrangements: 1) extending the linker off the fentanyl-like domains of HF-1, HF-2, HF-3, and HF-7
phenylamide ring of fentanyl; 2) extending the linker from the (Figure 2) might be accessed from the same synthesis. Our
phenethyl ring of fentanyl; or 3) anchoring the linker in be- other consideration was the general strategy for joining the two
tween the two drug-like portions of our singular epitope. Again, drug domains to form our chemically contiguous epitope. We
Figure 2 provides a visual guide to our approach, with the envisioned synthesizing HF-1, HF-2, and HF-4 by N-alkyl-
heroin and fentanyl chemical-epitopes in each drug–hapten ation of the heroin domain via reductive amination. For
highlighted in red and blue, respectively. Based on this logic, the remaining drug–haptens, the use of repeated amide
we posit that the centrally linked haptens (Figure 2D) might be bonds naturally led us to employ EDC-mediated coupling
the most successful at producing an equivalent response to both to join the heroin and fentanyl domains. For the sake
drugs, as they are equidistant from the carrier protein, and of brevity, detailed synthetic procedures, and structures
should reduce structural masking of fentanyl.
of all intermediates can be found in Supporting Information
File 2.
In synthesizing the planned series of haptens, it was useful to
consider each in terms of carrier-protein linker attachment as Heroin domain syntheses
well as the size of the epitope as delineated in Figure 2. Based All nine haptens in the series were constructed from one of the
upon these restrictions, retention of a tertiary amine in the three heroin intermediates shown in Figure 3. Haptens to be
1022

Beilstein J. Org. Chem. 2019, 15, 1020–1031.
Figure 2: The chemically contiguous heroin–fentanyl haptens designed in this study. Grouping was based on the linker-carrier protein attachment.
A) The individual drug haptens with an acid moiety for hapten-carrier protein attachment. B) Haptens to be linked to the carrier protein at the 4-posi-
tion on the phenyethyl group. C) Drug haptens to be attached to the carrier protein para to the phenylamide. D) Haptens conjoined centrally between
each drug hapten. Background color indicates haptens with structurally identical epitopes. Heroin and fentanyl molecules are highlighted in red and
blue, respectively, in the hapten structures.
prepared via reductive amination (HF-1, HF-2, HF-4, Figure 2 cording to literature procedure and purified by recrystallization
yellow background) required the secondary amine form, [16]. Alkylation of 1 with tert-butyl bromoacetate and subse-
norheroin (1, Figure 3). Thus, norheroin (1) was synthesized ac- quent deprotection with trifluoroacetic acid yielded the
1023

Beilstein J. Org. Chem. 2019, 15, 1020–1031.
Figure 3: Heroin intermediates used to synthesize HF-1 through HF-9.
N-glycine derivative of norheroin, 16. This critical intermediate Reductive amination of 4 with commercially available
was used to construct HF-3, HF-5, and HF-8 (Figure 2, gray phenethylpiperidin-4-one furnished intermediate 5 in good yield
background). Finally, the lengthy HF-6, HF-7, and HF-9 (92%, Scheme 2).
haptens (Figure 2, blue background) were synthesized using
intermediate 36, prepared according to literature procedures With intermediate 5 in hand, further elaboration was necessary
[3,17].
to obtain the aldehyde required for the proposed reductive
amination to access HF-1. Intermediate 5 was acylated with
pentenoyl chloride to give alkene 6, which was then oxidative-
Fentanyl domain syntheses
The scaffold variations in the fentanyl domain necessitated a ly cleaved using potassium osmate(IV) hydrate followed by so-
more methodical approach to mapping out essential intermedi- dium periodate to obtain aldehyde 7 (Scheme 1) [18,19].
ates. We first considered the phenylamide-linked haptens HF-1, Buoyed by this success, we tried a similar approach to obtain
HF-2, HF-3, and HF-7 (Figure 2B) and determined that they the fentanyl-like domain aldehyde required for HF-2. To our
could be synthesized from intermediate 5 (Scheme 1).
surprise, oxidative cleavage both by ozonolysis and osmium
tetroxide on the corresponding alkene resulted in a mixture of
To prepare this intermediate, the tert-butyl ester protected products. We opted to try the synthesis of HF-2 by 1,4-addition,
derivative 4 of the commercially available 4-(4-amino- rather than reductive amination. Thus, intermediate 5 was
phenyl)butanoic acid was accessed via protection and deprotec- acylated with acryloyl chloride to give the α,β-unsaturated inter-
tion of the amine with a phthaloyl group (intermediates 2, 3). mediate 10 (Scheme 1).
Scheme 1: General outline of HF-1, HF-2, HF-3, HF-7 synthesis from fentanyl intermediate 5 and heroin intermediates 1, 16, and 36.
1024

Beilstein J. Org. Chem. 2019, 15, 1020–1031.
Scheme 2: Synthesis of fentanyl intermediate 5. Reaction conditions: a) phthalic anhydride, AcOH, reflux, 81%; b) t-BuOH, EDC, Et3N, DMAP,
CH2Cl2, 0 °C to rt, 86%; c) H2NNH2, EtOH, 89%; d) 1-phenethylpiperidin-4-one, NaBH(OAC)3, AcOH, 0 °C to rt, 92%.
Condensation with a tert-butyl-protected glycine residue to As with HF-1 and HF-2, the goal for the synthesis of the HF-4
intermediate 5 yielded 15 (Scheme 1), which we used en route fentanyl-like domain was to obtain an aldehyde that could then
to HF-3. To reach HF-7 in a similar manner, a homoglycine be used in a reductive amination. Beginning with the commer-
residue was condensed with 5 to give 40 (Scheme 1).
cially available 4-(4-bromophenyl)butanoic acid, an alkene was
installed with a Suzuki reaction, which in turn underwent ozon-
We then considered systematic preparation of the branched olysis to give 21, shown in Scheme 4. In parallel, a commercial-
haptens HF-5, HF-8, and HF-9 (Figure 2). It appeared that ly available piperidine was acylated and Boc-deprotected to
N-acylated fentanyl intermediate 28 could be used to access yield the amine that was then reductively aminated with 21 to
HF-5 and HF-8, while the related intermediate 46 could be give 24, completing the fentanyl-like domain (Scheme 4). Oxi-
used for HF-9 (Scheme 3). Both of these N-acylated fentanyl dative cleavage of 24 yielded the aldehyde 25, ready for reduc-
intermediates were prepared according to published procedures tive amination with the heroin domain intermediate.
[20,21]. Coupling 28 to either a four- or eight-carbon linker
yielded the desired intermediates 29 and 43, for HF-5 and While HF-6 was designed with the same linker as HF-4, its
HF-8, respectively (Scheme 3). Alkylation of 46 with an eight- lengthy epitope actually required a different, but parallel syn-
carbon linker gave intermediate 47, procuring the fentanyl-like thetic strategy. The same commercially available piperidine
domain of HF-9 (Scheme 3).
used to obtain 24 was instead acylated with phthaloyl-protected
Scheme 3: General outline of HF-5, HF-8, HF-9 synthesis from fentanyl intermediates 28 and 46, and heroin intermediates 16, and 36.
1025

Beilstein J. Org. Chem. 2019, 15, 1020–1031.
Scheme 4: Parallel synthesis of fentanyl domains 25 and 34, for HF-4 and HF-6, respectively.
aminopropanoyl chloride, deprotected, and reductively
aminated with 21 to give 33 (Scheme 4). Intermediate 33 was
then deprotected to give 34, leading to facile EDC-mediated
coupling with 36 to give HF-6 (Scheme 4).
Heroin and fentanyl epitope coupling
With the various drug-like domains prepared, we set to work
forging them into a single epitope by employing our proposed
strategies. HF-1 and HF-4 were obtained from reductive amina-
tion of heroin intermediate 1 to key fentanyl intermediates 7 and
25, respectively (Scheme 5). In synthesizing HF-2, the afore-
mentioned failure to obtain the necessary aldehyde for reduc-
tive amination necessitated a strategic modification to 1,4-addi-
tion. After screening acid catalysts (Figure S1, Supporting
Information File 1), successful 1,4-addition of heroin intermedi-
ate 1 to fentanyl intermediate 10 yielded HF-2 (Scheme 5).
Heroin intermediate 16 gave access to the phenylamide-linked
haptens HF-3, HF-5, HF-8 via EDC-mediated coupling to their
respective fentanyl-like domains. Similarly, HF-6, HF-7, HF-9
were obtained by coupling heroin domain 36 to 34, 40, or 47,
respectively (Scheme 5). All haptens were synthesized with the
carrier protein linker protected as a tert-butyl ester. Deprotec-
tion in the presence of trifluoracetic acid unmasked the
carboxylic acids, allowing direct coupling of haptens to the
carrier protein.
Scheme 5: General strategy and coupling partners for the chemically
contiguous series. aGeneral conditions for reductive amination of 1
with indicated aldehyde: NaBH(OAc)3 (2.5 equiv), AcOH, 0 °C → rt.
bConditions for 1,4-addition of 1 to 10: Zn(OTf)2 (0.2 equiv), CH3CN, rt.
cGeneral conditions for amide coupling: EDC·HCl (1 equiv), Et3N
(3 equiv), DMAP (0.1 equiv), CH2Cl2, 0 °C → rt.
Hapten conjugation reaction with carrier
proteins BSA and KLH
eliminating the possibility of conjugation. To mitigate this
Test reactions on submilligram scale (0.1 mg hapten:0.1 mg precipitation problem, glycerol was added to the soluble KLH
BSA, 1 mg/mL protein) were performed with activated HF-1 or BSA solution (Figures S2–S7, Supporting Information
and BSA, resulting in a moderate number of haptens coating the File 1). Concerned by the possibility that the alcohols on glycer-
protein’s surface (i.e., 10.0 copies). However, increasing the ol would outcompete amidation, we tested several concentra-
conjugation to a 1 mg scale caused the protein to precipitate tions of glycerol in the range of 10–50% (w/w). Based on
immediately upon addition of HF-1 in 9:1 DMF/H2O solution, MALDI–TOF data and as expected, the conjugation with the
1026

Beilstein J. Org. Chem. 2019, 15, 1020–1031.
least amount of glycerol had the highest number of hapten Figure 4. Mice were bled at weeks 3, 5, and 8 using
copies (i.e., 11.1 copies). Addition of 50% w/w glycerol submandibular puncture with a single-use lancet to collect
reduced the hapten density by one-half (i.e., 5.2 copies). After 100–500 µL whole blood. All sera were tested against both Her-
dialysis, the solutions remained translucent, but the protein BSA and Fent-BSA antigens, and sera from each chemically
conjugates did not crash out of the solution. As a reference, contiguous vaccine group were tested against their own antigen.
copy number for the individual heroin and fentanyl haptens For heroin analysis, 6-acetylmorphine (6-AM) was used as the
were 7.3 and 12.4, respectively.
primary analyte in surface plasmon resonance (SPR) analysis as
it has a longer half-life than heroin at pH 7.4.
Another trend observed for the chemically contiguous haptens
was the increased amount of time needed to activate the To begin epitope interpretation, we ran the haptens in numeri-
carboxylic acids. Typically, fentanyl and heroin haptens reach cal order, and observed midpoint titers, which were highest for
optimal threshold activation (above 80%) within thirty minutes all groups at week 5. In contrast to titer, IC50 affinity continued
to several hours. However, these haptens required a minimum to improve up to week 8. Individual conjugate vaccines had the
of four to twelve hours for at least 50% activation. In addition, highest level of midpoint titers of all groups as well as the best
bulkier haptens, including HF-4, had solubility issues upon ad- affinity values (Figures S8–S11, Supporting Information
dition to carrier protein and therefore needed higher glycerol File 1). Of the chemically contiguous vaccines only a handful of
content to ameliorate protein precipitation and conjugation them elicited single-digit nanomolar affinity to either 6-AM or
issues. The linker placement at the phenethyl position greatly fentanyl. Interestingly, the centrally linked haptens (HF-5,
improved conjugation compared to the other phenylamide HF-8, and HF-9) had affinity only toward one of the two drugs,
group; HF-4 had a hapten density of 21.1, drastically higher not both as initially anticipated. Also of note was that, while
than HF-2 (hapten density: 12.6) and HF-3 (hapten density: HF-4 and HF-6 had excellent affinity toward 6-AM, this did
8.3). Gratifyingly, we observed copy numbers on BSA for every not translate to antibody-fentanyl affinity, which was 40 and
synthesized hapten. Immunoconjugates were then prepared with 70-fold poorer. As expected this poor affinity to fentanyl was
keyhole limpet hemocyanin (KLH) for murine immunization recapitulated as a nonstarter in behavior as seen in the tail flick
studies.
antinociception assay (Figure S12, Supporting Information
File 1). Finally, from all nine vaccine designs, HF-7 provided
most compelling antinociception efficacy in terms of the 10%
fentanyl in heroin challenge (Figure 5).
Vaccination schedule, SPR and nociception
analysis
The nine haptens were formulated with Alhydrogel (alum) and
murine CpG ODN as previously described [13,17]. Briefly, Discussion
vaccines were formulated with 25% (v/v) glycerol, 50 µg of Unfortunately, analyzing data by chemical epitope, i.e., each
CpG/dose per mouse and stored at −80 °C. On the day of vacci- drug, did not reveal any foretelling trends for immune response
nation, 0.8 mg of alum/dose per mouse was added and mixed against said drug. With that said, it seems that the chemical
for at least thirty minutes after thawing.
linkage providing continuity between the heroin and fentanyl-
like portions of the hapten had less impact on vaccine perfor-
Female BALB/cByJ mice were immunized and bled using the mance. Rather, the regiochemical placement of the hapten-
protocol detailed in our previous study, except this study incor- linker when coupled to the carrier protein seemed to provide a
porated an extra boost at week 7, bleed at week 8, and antinoci- greater influence on the immune response (grouping in Figure 2
ception at week 9 [13]. The vaccination schedule is shown in and Figure S13, Supporting Information File 1).
Figure 4: Vaccination, titer assessment, and bleed schedule.
1027

Beilstein J. Org. Chem. 2019, 15, 1020–1031.
Figure 5: Summary of behavioral data for most promising chemically contiguous vaccine HF-7, compared to singular hapten-vaccines and controls.
A) Structures of haptens; B) ED50 values for heroin and fentanyl; C) potency values. Bars represent means + SEM.
We were taken aback to observe that the centrally-linked processing, so regardless of affinity for 6-AM and/or fentanyl,
vaccines comprised of HF-5, HF-8, HF-9, Figure 2D per- these vaccines were not efficacious as too little antibody
formed so poorly in ELISA analysis (Figure S9, Supporting was produced. Indeed, in the tail flick antinociceptive experi-
Information File 1), given that they most closely mirror the ment against 10% fentanyl in heroin, mice vaccinated with
individual conjugate hapten vaccines. It is possible that the these haptens were not significantly better than the control
branched structure is not ideal for MHC-presentation or group.
1028

Beilstein J. Org. Chem. 2019, 15, 1020–1031.
The phenethyl-linked hapten vaccines (HF-4, HF-6, Figure 2) the Fent-KLH vaccine when examined in the 10% fentanyl in
performed similarly or better than the Her-KLH vaccine as seen heroin.
in the antinociception assay when mice were challenged with
heroin, but performed poorly against fentanyl challenged mice We also sought to determine if there was any predictive correla-
(Figure 6). In line with this result is the SPR data, which by tion between the in vitro SPR and ELISA results, and the
week 8 showed that HF-4 and HF-6 displayed the best IC50s behavioral results. To this end, nociception ED50 values ob-
for 6-AM (Figure S8, Supporting Information File 1). The tained from each vaccine challenged with either heroin,
phenethyl-linked conjugate-vaccines also produced the highest fentanyl, or 10% fentanyl in heroin were plotted against either
titers against Her-BSA and the lowest against Fent-BSA (Figure IC50 or midpoint titers obtained for that analyte. Linear regres-
S9, Supporting Information File 1). We hypothesize that these sion analysis was used to determine correlation. Initially we
phenethyl-linked haptens mask the fentanyl-like domain plotted ED50 versus SPR IC50s, but we found no correlation.
epitope, resulting in poor titers. Poor titers against fentanyl as Similarly, no correlation was found between IC50 and midpoint
expected translated to little protection against 10% fentanyl in titers. However, when we plotted ED50 against midpoint titers
heroin.
(Figure 7), we saw a positive correlation in the heroin group and
this correlation improved by week 8. Because there was no rela-
It is difficult to extrapolate any definitive conclusions regarding tionship between antibody affinity for 6-AM and in vivo effi-
phenylamide-linked vaccines, (Figure 2C) other than that their cacy, it is tempting to conclude that the quantity of antibody is
ability to enhance heroin’s nociception effects was modest, and more important for protection against a heroin challenge. How-
only worked marginally better against fentanyl. That being said, ever, the interplay between antibody concentration and affinity
overall it appears that the HF-7 vaccine had the most for this series is difficult to delineate. That being said, the two
“balanced” protection against heroin and fentanyl. Possibly, vaccines most successful against heroin challenge, HF-4 and
fentanyl’s epitope domain is better displayed in HF-7, com- HF-6, had the highest titers for Her-BSA and displayed the
pared to HF-6, which is identical except for protein carrier greatest affinity for 6-AM. Interestingly, it appears that when
attachment. In the antinociceptive studies, the HF-7 vaccine compared to fentanyl or 10% fentanyl in heroin groups, the
provided protection against heroin comparable to Her-KLH week 5 midpoint titers have no correlation with efficacy (R2 <
vaccine (Figure 5B), but did relatively poorly against a fentanyl 0.03). Week 8 titers show a slight correlation, although it is was
challenge compared to the Fent-KLH vaccine. Although the not found to be significant. Moreover, nearly all of the vaccines
HF-7 vaccine demonstrated limited protection against fentanyl produced very low titers and affinity to fentanyl compared to
challenge, its superior immune response against heroin fentanyl-KLH vaccine (Figures S8 and S9, Supporting Informa-
permitted it to be as efficacious in nociception examination as tion File 1).
Figure 6: Summary of behavioral data for phenethyl-linked haptens HF-4 and HF-6. Bars represent mean ± SEM.
1029

Beilstein J. Org. Chem. 2019, 15, 1020–1031.
Figure 7: Correlation plots of dual hapten vaccines comparing week 5 and 8 ELISA midpoint titers to ED50 values. Linear regression analysis was
performed and R2 values for week 8 are A) heroin week 8 R2 = 0.7803; B) fentanyl week 8 R2 = 0.3459; C) 10% fentanyl in heroin week 8
R2 = 0.3715.
Conclusion
the fentanyl moiety. These studies will be reported in due
Our chemically contiguous hapten campaign was initiated to in- course.
vestigate the possibility of taking two structurally different
opioids, heroin and fentanyl, and combining them into a
Supporting Information
singular, unique opioid hapten that, when formulated with an
adjuvant, would present us with a vaccine capable of gener-
ating antibodies to both drugs. Based on this idea, a series of
nine vaccines were prepared that enabled us to detect subtle
influences on antibody response to each drug. In sum, we found
little traction with each hapten’s vaccine efficacy from the
standpoint of the chemical connection between the two drugs.
Interestingly, immune response seemed to be more influenced
by linker-carrier protein regiochemical placement. Although
none of the nine haptens outperformed the individual conjugate
vaccines against their assigned drug, the HF-7 conjugate
vaccine did show promising results as its heroin response was
on par with Her-hapten immune response and it provided some
efficacy against the fentanyl in heroin mixture. As linker-carrier
protein regiochemical placement seemed more controlling for a
balanced immune response against both drugs, it will be of
interest to explore linker-carrier-protein display off the 3 or
6-position of heroin, thereby more prominently displaying
Supporting information features SPR and ELISA
experimental results, and additional experimental data. It
also includes detailed procedures for all reported
intermediates and final compounds, as well as their
1H NMR, 13C NMR, and high-resolution
mass-spectrometry signals.
Supporting Information File 1
Supporting experimental results.
[https://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-15-100-S1.pdf]
Supporting Information File 2
Experimental procedures for compounds 1–53.
[https://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-15-100-S2.pdf]
1030

Beilstein J. Org. Chem. 2019, 15, 1020–1031.
12.Jacob, N. T.; Anraku, K.; Kimishima, A.; Zhou, B.; Collins, K. C.;
Lockner, J. W.; Ellis, B. A.; Janda, K. D. Chem. Commun. 2017, 53,
8156–8159. doi:10.1039/c7cc04055e
Supporting Information File 3
NMR spectra of compounds 1–53.
13.Hwang, C. S.; Smith, L. C.; Natori, Y.; Ellis, B.; Zhou, B.; Janda, K. D.
ACS Chem. Neurosci. 2018, 9, 1269–1275.
[https://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-15-100-S3.pdf]
doi:10.1021/acschemneuro.8b00079
14.Hwang, C. S.; Smith, L. C.; Natori, Y.; Ellis, B.; Zhou, B.; Janda, K. D.
ACS Omega 2018, 3, 11537–11543. doi:10.1021/acsomega.8b01478
15.Collins, K. C.; Janda, K. D. Bioconjugate Chem. 2014, 25, 593–600.
doi:10.1021/bc500016k
Acknowledgements
This work was supported by funding from the National Insti-
tutes of Health NIDA (UH3DA041146). This is manuscript #
29814 from The Scripps Research Institute.
16.Stowe, G. N.; Vendruscolo, L. F.; Edwards, S.; Schlosburg, J. E.;
Misra, K. K.; Schulteis, G.; Mayorov, A. V.; Zakhari, J. S.; Koob, G. F.;
Janda, K. D. J. Med. Chem. 2011, 54, 5195–5204.
doi:10.1021/jm200461m
17.Hwang, C. S.; Bremer, P. T.; Wenthur, C. J.; Ho, S. O.; Chiang, S.;
Ellis, B.; Zhou, B.; Fujii, G.; Janda, K. D. Mol. Pharmaceutics 2018, 15,
1062–1072. doi:10.1021/acs.molpharmaceut.7b00933
18.VanRheenen, V.; Kelly, R. C.; Cha, D. Y. Tetrahedron Lett. 1976, 17,
1973–1976. doi:10.1016/s0040-4039(00)78093-2
ORCID® iDs
Candy S. Hwang - https://orcid.org/0000-0003-2826-4078
Lucy Lin - https://orcid.org/0000-0001-9546-6919
Lauren C. Smith - https://orcid.org/0000-0002-4290-3262
Bin Zhou - https://orcid.org/0000-0001-5779-9085
Kim D. Janda - https://orcid.org/0000-0001-6759-4227
19.Yu, W.; Mei, Y.; Kang, Y.; Hua, Z.; Jin, Z. Org. Lett. 2004, 6,
3217–3219. doi:10.1021/ol0400342
20.Petrov, R. R.; Lee, Y. S.; Vardanyan, R. S.; Liu, L.; Ma, S.-w.;
Davis, P.; Lai, J.; Porreca, F.; Vanderah, T. W.; Hruby, V. J.
Bioorg. Med. Chem. Lett. 2013, 23, 3434–3437.
References
1. Dahlhamer, J.; Lucas, J.; Zelaya, C.; Nahin, R.; Mackey, S.; DeBar, L.;
doi:10.1016/j.bmcl.2013.03.065
Kerns, R.; Von Korff, M.; Porter, L.; Helmick, C.
21.Jacquot, C.; McGinley, C. M.; Plata, E.; Holman, T. R.;
van der Donk, W. A. Org. Biomol. Chem. 2008, 6, 4242–4252.
doi:10.1039/b808003h
MMWR Morb. Mortal. Wkly. Rep. 2018, 67, 1001–1006.
doi:10.15585/mmwr.mm6736a2
2. Solis, E., Jr.; Cameron-Burr, K. T.; Kiyatkin, E. A. eNeuro 2017, 4,
e0323-17.2017. doi:10.1523/eneuro.0323-17.2017
3. Bremer, P. T.; Schlosburg, J. E.; Banks, M. L.; Steele, F. F.; Zhou, B.;
Poklis, J. L.; Janda, K. D. J. Am. Chem. Soc. 2017, 139, 8601–8611.
doi:10.1021/jacs.7b03334
License and Terms
This is an Open Access article under the terms of the
Creative Commons Attribution License
4. Bremer, P. T.; Kimishima, A.; Schlosburg, J. E.; Zhou, B.;
Collins, K. C.; Janda, K. D. Angew. Chem., Int. Ed. 2016, 55,
3772–3775. doi:10.1002/anie.201511654
(http://creativecommons.org/licenses/by/4.0). Please note
that the reuse, redistribution and reproduction in particular
requires that the authors and source are credited.
5. Schlosburg, J. E.; Vendruscolo, L. F.; Bremer, P. T.; Lockner, J. W.;
Wade, C. L.; Nunes, A. A. K.; Stowe, G. N.; Edwards, S.; Janda, K. D.;
Koob, G. F. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 9036–9041.
doi:10.1073/pnas.1219159110
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
6. Matyas, G. R.; Mayorov, A. V.; Rice, K. C.; Jacobson, A. E.; Cheng, K.;
Iyer, M. R.; Li, F.; Beck, Z.; Janda, K. D.; Alving, C. R. Vaccine 2013,
31, 2804–2810. doi:10.1016/j.vaccine.2013.04.027
(https://www.beilstein-journals.org/bjoc)
7. Sulima, A.; Jalah, R.; Antoline, J. F. G.; Torres, O. B.; Imler, G. H.;
Deschamps, J. R.; Beck, Z.; Alving, C. R.; Jacobson, A. E.; Rice, K. C.;
Matyas, G. R. J. Med. Chem. 2018, 61, 329–343.
The definitive version of this article is the electronic one
which can be found at:
doi:10.1021/acs.jmedchem.7b01427
doi:10.3762/bjoc.15.100
8. Ohia-Nwoko, O.; Kosten, T. A.; Haile, C. N. Int. Rev. Neurobiol. 2016,
126, 263–291. doi:10.1016/bs.irn.2016.02.009
9. Pravetoni, M.; Raleigh, M. D.; Le Naour, M.; Tucker, A. M.;
Harmon, T. M.; Jones, J. M.; Birnbaum, A. K.; Portoghese, P. S.;
Pentel, P. R. Vaccine 2012, 30, 4617–4624.
doi:10.1016/j.vaccine.2012.04.101
10.Torres, O. B.; Matyas, G. R.; Rao, M.; Peachman, K. K.; Jalah, R.;
Beck, Z.; Michael, N. L.; Rice, K. C.; Jacobson, A. E.; Alving, C. R.
Vaccine 2017, 2, No. 13. doi:10.1038/s41541-017-0013-9
11.Anraku, K.; Sato, S.; Jacob, N. T.; Eubanks, L. M.; Ellis, B. A.;
Janda, K. D. Org. Biomol. Chem. 2017, 15, 2979–2992.
doi:10.1039/c7ob00448f
1031