A fentanyl vaccine constructed upon opsonizing antibodies specific for the Galα1-3Gal epitope

Abstract

A double conjugation strategy was implemented to produce an anti-fentanyl vaccine, which was predicated upon preformed-antibody-assisted antigen presentation. The new vaccine was found to reduce the psychoactive effects of fentanyl without the addition of the immunostimulant CpG oligodeoxynucleotide.

Graphical Abstract

A double-conjugate fentanyl vaccine leveraging preformed anti-Gal antibodies significantly reduced the pain-relieving effect of fentanyl in mice.

The rapid increase of synthetic opioid abuse is a major driving force in the evolving public health crisis of opioid misuse and addiction.^1,2 In the United States, overdose deaths involving fentanyl and other synthetic opioids increased from 3105 in 2013 to 19413 in 2016, and synthetic opioids have for the first-time surpassed prescription opioids as the most common drug category related to fatal overdoses.² Fentanyl acts primarily by activating μ-opioid receptors (MORs) in the brain and is used as a schedule II drug for pain relieving.^3,4 Its superior potency (about 100-fold greater than morphine and 10-fold greater than heroin^5,6) poses a great risk of overdosing to people using it for recreational purposes, especially when it is mixed with other drugs such as heroin and cocaine.²

The current pharmacotherapies for fentanyl-related overdoses and opioid use disorder (OUD) have relied on MOR antagonists and agonists, which suffer from limited effectiveness and possible side effects.^1,7,8 One strategy to address this issue is to use immunoconjugates as vaccines to stimulate the generation of anti-fentanyl antibodies, thus sequestering the drug in peripheral blood and prohibiting it from entering the brain.^9–11 Specifically, our previous study with a fentanyl tetanus toxoid vaccine showed that this immunotherapeutic approach was able to blunt the effects of fentanyl class drugs and protect the mice from lethal drug doses.⁹

To improve the immunogenicity of anti-drug vaccines, adjuvants that act as vehicles or immunostimulants have played an important role in the vaccine formulation. Yet, despite the critical roles of adjuvants in augmenting immune response, the fact that greater adjuvant potency is often associated with higher risk of adverse effects means that it is not always possible to reach a satisfying benefit-risk balance.^12,13 In fact, new adjuvants approved for human use to date are rare. Therefore, in addition to continued searching for highly efficient adjuvants with low toxicity, we have also sought to develop effective anti-drug vaccine formulations with minimal use of an adjuvant.

To this end, we decided to make use of the targeting effect of opsonizing antibodies.¹⁴ Antigens attached by corresponding IgG can be directed to non-specific antigen-presenting cells (APCs), leading to enhanced antigen uptake and presentation by these cells.¹⁵ This opsonization process is mediated by the interaction of the antibody Fc fragment and Fcγ receptors (FcγRs) on APCs such as dendritic cells (DCs) and macrophages.¹⁶ Accordingly, this strategy of leveraging the formation of immune complexes to enhance the immune response to an anti-drug vaccine has been embraced herein.

Anti-Gal antibodies are natural IgG antibodies specific for the α-Gal epitope (Galα1–3Gal disaccharide), constituting ~1% of circulating IgG in humans.¹⁷ They represent one of the most abundant immunoglobulins in human serum and serve as good candidates as opsonizing antibodies for immune response enhancement. Previous studies with tumor cells,¹⁸ viral vaccines^19,20 and liposomes²¹ have shown that linking the α-Gal epitope to antigens, either covalently or non-covalently, could induce increased antigen transport to lymph nodes, elevated activation of CD4⁺ T helper cells and greater antibody response. Our work on an anti-cocaine vaccine has further proved the feasibility of this approach for immunization against non-protein or peptide like molecules.²² Using this enabled logic, we have developed an anti-fentanyl vaccine exploiting the opsonizing effect of anti-Gal, minimizing the need for adjuvant addition.

To empower anti-Gal-assisted active vaccination against fentanyl, an immunogenic protein conjugate equipped with both the fentanyl and α-Gal epitopes was synthesized (Scheme 1). Ovalbumin (OVA) was chosen as the carrier of choice based upon findings from our previous anti-cocaine vaccine.²² The Galα1–3Galβ1–4GlcNAc structure in the α-Gal hapten (1) has reduced conformational flexibility compared to the Galα1–3Gal disaccharide, and was assumed to be the major target of natural anti-Gal binding.²³ The squaric acid ester moiety permitted direct reaction with free amino groups of OVA in borate buffer (pH 9.0), resulting in 1 to 3 copies of the trisaccharide on the carrier protein, as determined by MALDI-TOF analysis. The remaining lysine side chains of the protein were used in the second conjugation step mediated by 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC) and N-hydroxysulfosuccinimide (Sulfo-NHS). As a result, the carrier protein was further functionalized with 11–14 copies of the fentanyl hapten (2).

Scheme 1.

Synthesis of OVA conjugated to two different haptens

In contrast to humans, rodents lack anti-Gal due to immune tolerance to the abundant α-Gal epitopes on their cells synthesized by α1,3galactosyltransferase (α1,3GalT).¹⁷ To mimic the immune characteristics of humans, α1,3GalT knockout (KO) mice²⁴ were used in our study and were immunized with tetanus toxoid (TT) conjugated to the α-Gal hapten as described previously²³ to stimulate anti-Gal immune response. Successful induction of anti-Gal antibody production was indicated by high antibody titers determined from enzyme-linked immunosorbent assay (ELISA, ESI^†). The resulting mice were vaccinated intraperitoneally (i.p.) following the schedule shown in Figure 1A. Notably, a previous study with TT suggested that the formation of antigen-antibody complexes may affect the primary and secondary immune responses differently.²⁵ While the primary response benefited from immune complex injection, the anti-TT antibody production was repressed when TT-antibody complex was injected in the secondary immunization. To evaluate the effect of anti-Gal-mediated opsonization on secondary immunization, two different vaccination strategies were applied in our study: (1) a Fent-Gal/Fent-Gal group was vaccinated with the fentanyl-(α-Gal)-OVA double conjugate in all three injections. (2) A Fent-Gal/Fent group was vaccinated with the double conjugate in the first injection (prime, week 0) and the fentanyl-OVA single conjugate in the subsequent two boosts. To serve as a control a cohort of mice were inoculated with phosphate buffered saline (PBS) in all three injections. Importantly, Al(OH)₃ (alum) was added as the sole adjuvant in all formulations due to its proved safety in humans. Aluminum salts work by providing an antigen depot and serve as relatively weak Th2 adjuvants that promote humoral immune response.^12,13 They have been used for decades and are the most widely used human vaccine adjuvants.

Figure 1.

Immunization with the anti-Gal-assisted vaccines induced the production of anti-fentanyl antibodies. (A) Vaccination schedule used in this study. Anti-Gal antibody induction was carried out before week 0. Antinoc. stands for antinociception assay. Drug Distrib. stands for drug distribution study. (B) Midpoint anti-fentanyl IgG titers of sera from the vaccinated groups (n = 7 or 8) shown as means ± SEM. Fentanyl-bovine serum albumin conjugate (Fent-BSA) was used as the coating antigen in ELISA. **P < 0.01, ***P < 0.001 versus control, Brown-Forsythe and Welch ANOVA test.

Sera collected from mice at weeks 3 and 5 (after the two boosts, respectively) were analysed by ELISA to determine the anti-fentanyl IgG antibody levels. As shown in Figure 1B, robust midpoint titers were observed. For both vaccination groups, titers from week 5 are higher, indicating the continuing accumulation of anti-fentanyl IgG antibodies. At week 5, the Fent-Gal/Fent group had higher titers than the Fent-Gal/Fent-Gal group, but the difference was not significant (P = 0.74, Brown-Forsythe and Welch ANOVA test). To evaluate the binding affinity of the serum antibodies for fentanyl, competitive binding assays based on surface plasmon resonance (SPR) was carried out (Table 1). At week 3, both the vaccination groups showed sub-micromolar IC₅₀ values, and the Fent-Gal/Fent-Gal group exhibited 2-fold higher affinity than the other group. At week 5, IC₅₀ value of the Fent-Gal/Fent-Gal group remained almost the same despite the boost at week 4, possibly reaching a plateau. The affinity of the Fent-Gal/Fent group increased, albeit still lower than the Fent-Gal/Fent-Gal group.

Table 1.

IC₅₀ values (μM) of serum antibodies for fentanyl^a

Group	Week 3	Week 5
Fent-Gal/Fent-Gal	0.43	0.42
Fent-Gal/Fent	0.85	0.59
Control	N.D.b	N.D.b

^aValues were determined using pooled serum samples from each group.

^bNo binding was detected.

To evaluate the efficacy of the vaccines in vivo, antinociception assays were carried out. Fentanyl, as with other synthetic opioids, increases pain thresholds in a dose-dependent manner. In the hot plate and tail flick antinociception assays, the pain-relieving effect of drugs are examined by testing animal behaviour in response to heat treatment. An effective anti-fentanyl vaccine should increase the fentanyl dose needed to induce the intended drug effect in these assays. As shown in Figures 2A and B, both vaccination strategies demonstrated a significant shift in fentanyl dose-response curves. Specifically, the ED₅₀ values of the vaccination groups are 3 to 5-fold larger than the control group (Figures 2C and D). Of the two vaccination groups, the Fent-Gal/Fent-Gal group showed better results in both assays, although in the tail flick assay the difference was not significant. The sequestering effect elicited by the vaccines was further supported by the results from fentanyl blood-brain distribution studies. Following administration of fentanyl, mice were sacrificed, and the drug concentration was measured via liquid chromatography-tandem mass spectrometry (LC-MS/MS). Compared to the control group, the vaccination groups showed higher fentanyl concentration in serum samples (Figure 2E, differences not significant). The differences of fentanyl concentration in brain samples, however, were marginal among the groups (Figure 2F). Notably, the ED₅₀ fold change promoted by anti-Gal-assisted vaccines studied here is comparable to that induced by some of our anti-fentanyl vaccines adjuvanted with the strong immunostimulant CpG oligodeoxynucleotide (ODN) 1826,^26,27 despite the differences in carrier proteins and mouse strains. Nevertheless, the fact that CpG-formulated vaccines were able to induce antibodies with nanomolar affinity indicates that the adjuvant is still preferred in pursuing optimal protection against fentanyl. Because CpG and anti-Gal antibodies promote immune response by different pathways (through binding to Toll-like receptor 9²⁸ and FcγR, respectively), the combination of both strategies is hypothesized to induce cross-talk effects and calls for further investigation.

Figure 2.

Anti-Gal-assisted vaccines are effective in blunting fentanyl effect and sequestering the drug in blood. Fentanyl citrate dose-response curves were measured in (A) hot plate and (B) tail flick antinociception tests (n = 7 or 8). %MPE = percentage of the maximum possible drug effect. (C) and (D), ED₅₀ values determined in the tests and fold change with respect to the control groups are shown. Fentanyl concentrations in (E) serum and (F) homogenized brains of mice sacrificed 15 min after drug administration are shown as individual points and means, n = 4. Error bars denote SEM. ****P < 0.0001 versus control, Brown-Forsythe and Welch ANOVA test.

The fact that the Fent-Gal/Fent-Gal vaccination strategy performed better in shifting fentanyl ED₅₀ than the Fent-Gal/Fent strategy suggests that in contradiction to the previous report, antibody opsonization could be beneficial in secondary immune response, at least in our case. It is known that the antigen processing pathways are different between the primary and secondary immune response. While antigen presentation in the primary immune response is mainly mediated by nonspecific immune cells such as DCs, in secondary response antigen-specific B cells become increasingly important.²⁹ Because antigen recognition of B cells is mediated by the B cell receptor (BCR), antibody opsonization will not promote B cell-mediated antigen presentation. On the contrary, excessive antibody binding may impede efficient antigen processing, which possibly led to immune response repression observed in previous studies. In the current study, the copy number of the α-Gal hapten on the immunoconjugate is low (1 to 3). Therefore, the masking effect of antibody binding can be minimized, resulting in the beneficial effect induced by anti-Gal opsonization in the secondary immune response.

In conclusion, we have developed a fentanyl conjugate vaccine equipped with the Galα1–3Gal epitope to leverage the opsonization effect of preformed anti-Gal antibodies. The vaccine induced robust immune response against fentanyl and proved effective in animal behaviour tests. The comparison between two immunization strategies revealed the importance of vaccination with the double conjugate for both priming and boosting. The current vaccine design will provide new tools in the arsenal to combat the ongoing opioid crisis driven by the rise in fentanyl class drug misuse.