IgE antibodies are the mainstay of allergic responses. They are monomeric glycoproteins composed of two light and heavy chains, the latter showing four constant Ig-like domains (Cϵ1–4), bound via disulfide bridges (38). Several factors are involved in the development of functional IgE antibodies, including specific affinity maturation, conformational/allosteric properties, and glycosylation patterns (38–40). IgE blood concentration in healthy individuals is very low (below 210  IU/ml) compared to normal levels of IgG (5.65–17.65 mg/ml) (41, 42). IgE are mostly sequestered in peripheral tissues, with an average half-life estimated of 16–20 days in the skin versus 2–4 days in blood (43). Given the high affinity of FcϵRI to IgE (Kd = 10¹⁰–10¹¹ M⁻¹) and the slow dissociation rate (44, 45), the majority of IgE are cell-bound to either FcϵRI or the low-affinity receptor FcϵRII (CD23) via the Cϵ3-4 Fc domains (46). FcϵRI is the high-affinity IgE receptor constitutively expressed on MCs and basophils, while inducible on monocytes, dendritic cells, eosinophils and neutrophils (47–50). A tight correlation between atopic status, circulating IgE levels and surface expression of FcϵRI on MCs, basophils and other cell types has been proven (45, 47, 51, 52). While peripheral blood-resident cells acquire IgE directly from the circulation, perivascular tissue-resident MCs, sensing changes in total IgE levels, probe IgE from blood vessels using endoluminal cell processes (53).

Furthermore, occupancy of the FcϵRI receptor is crucial to ensure its expression on the cell membrane by MCs and basophils, as shown by mechanistic studies demonstrating increased FcϵRI expression upon IgE binding due to decreased FcϵRI endocytosis and degradation (44, 54–56). IgE bound to FcϵRI persists as long as MCs are alive, thus indicating that MCs preferentially display rather than catabolize IgE. FcϵRI-mediated constitutive internalization of IgE by dendritic cells and monocytes promotes serum IgE clearance instead (57).

FcϵRI is constituted by one alpha and one beta chain on MCs and basophils, or a single alpha chain on monocytes and dendritic cells (45, 58, 59), complexed with two additional gamma chains with immunoreceptor tyrosine-based activation motif (ITAM) domains acting as docking and activation sites for the Spleen tyrosine kinase (Syk) pathway (60, 61). The activation of the Syk, phosphatidylinositol 3-OH kinase/protein kinase B (PI3K/Akt) and extracellular signal-regulated kinase (ERK) pathways leads to increased intracellular calcium flux, calcium-dependent release of preformed mediators stored in intracellular granules and activation of transcription factors for eicosanoids, cytokines and chemokines production (62).

MCs and basophils express the highest density of FcϵRI receptor [estimated 0.7 × 10⁵ molecules per cell measured on LAD2 MCs (63)], with a bell-shaped dose–response when exposed to increasing allergen concentrations (64). Degranulation is tightly regulated via mechanisms modulating the MC activation threshold, not limited to IgE–FcϵRI complex expression. In fact, the nature and dose of the eliciting allergen also play a modulatory role. For instance, simultaneous stimulation using multiple allergens shows an additive effect on MC activation when suboptimal allergen concentrations are used. Conversely, stimulation with supra-optimal allergen concentrations inhibits MC degranulation (64, 65).

Given the pivotal role of IgE in the initiation and maintenance of allergic responses, increasing evidence supports the use of anti-IgE molecules as therapeutic strategy to treat allergic diseases, including food allergy ( Tables 1 and 2 ). Anti-IgE therapy disrupts the IgE–FcϵRI axis via the active removal of circulating IgE and the downregulation of FcϵRI on MCs, basophils and dendritic cells (136–138). By removing circulating IgE, the turnover between circulating and cell-bound allergen-specific IgE (sIgE) slowly declines, ultimately reducing the amount of sIgE bound on the cell surface and decreasing the likelihood of allergen-IgE cross-linking and allergen-specific effector cell responses (139–141) ( Figure 1A ).

Furthermore, anti-IgE treatment induces FcϵRI down-regulation by interfering with the accumulation of IgE–FcϵRI complexes occurring at the cell surface due to reduced receptor occupancy by IgE (54–56, 136). The reduced availability of sIgE–FcϵRI complexes further inhibits the release of Th2 cytokines and allergic mediators upon allergen challenge by MCs, basophils and dendritic cells (136, 137, 141–144).

Some anti-IgE treatments also inhibit IgE binding to the CD23 receptor, the low affinity IgE receptor constitutively expressed on naïve B cells, exerting an inhibitory effect on IgE-mediated antigen presentation (145, 146), inducing anergy or apoptosis of membrane IgE-bearing B cells (147, 148) and in some cases modulating IgE production (146, 149). However, treatment discontinuation is followed by the quick restoration of pre-treatment IgE levels (150).

Omalizumab, a humanized anti-IgE monoclonal antibody, is the first and most studied biologic, currently used to treat severe asthma and chronic spontaneous urticaria ( Table 2 ). It binds to IgE Cε3 domains, outside of the FcεRI-binding site, and sterically disrupts binding to both FcϵRI and CD23 (151). Omalizumab does not affect pre-bound IgE-receptor interactions, due to conformational changes of receptor-bound IgE masking omalizumab binding sites, and does not induce IgE cross-linking on the cell surface (151, 152).

Omalizumab downregulates the surface expression of FcϵRI in both basophils and MCs (153). However, while FcϵRI expression declines rapidly in circulating basophils (less than 24 h), this process requires longer time in tissue resident MCs (estimated 10–20 days) (136, 154, 155).

The effects exerted by omalizumab on MCs are of clinical relevance also in non-IgE-mediated diseases such as inducible urticarias (156) and MC activation syndrome (105), thus suggesting a broad MC stabilizing function. In food allergy, several clinical trials and real-life evidence showed the safety and usefulness in inhibiting allergic responses of omalizumab as monotherapy (157–160) ( Table 1 ), or in association with allergen-specific immunotherapy, further discussed in Combination Treatments With Biologics section.

Designed Ankyrin Repeat Proteins (DARPins), genetically engineered antibody mimetic proteins, recognize IgE Cϵ3 domains with high specificity and affinity, and have been shown to be 10,000-fold more efficient than omalizumab in dissociating IgE complexes in vitro and in both ex vivo transgenic mouse models and human tissues. Thus, their rapid onset of action makes them of particular interest as treatment option to thwart pre-initiated anaphylaxis episodes (161–165) ( Table 3 ). Along with DARPins, other new generation high-affinity anti-IgE monoclonal antibodies like ligelizumab can actively bind IgE Cε3–4 fragments and efficiently disrupt IgE-FcεRI complexes without, however, interfering with CD23 binding, differently than omalizumab (146) ( Table 3 ). Thus, DARPins and ligelizumab might improve treatment efficacy in food allergy, albeit to date no trials on food allergy are ongoing ( Table 2 ).

New anti-IgE strategies involve self-assembled mRNA vaccines, that provide epitopes mimicking IgE Cε3 domains and stimulate the production of endogenous anti-IgE IgG antibodies, eventually modulating circulating IgE levels via the same mechanisms of omalizumab and other anti-IgE molecules (166, 167). These new treatments inhibited IgE-mediated anaphylaxis in animal models ( Table 3 ) and were tested in a Phase I trial conducted on allergic rhinitis patients (NCT01723254, Table 2 ); however their application in food allergy is still unclear.

Concerns over the long-lasting implications of irreversible IgE suppression might also arise, considering that, along with omalizumab and other high affinity molecules, broad anti-IgE agents bind also to IgE antibodies serving housekeeping functions, like protection against parasitic infections and tumor surveillance (185–187). Therefore, alternative strategies have been developed to specifically target IgE of interest. The use of covalent heterobivalent low affinity allergic response inhibitors (LARI) has been promising and showed to reduce the risk of anaphylaxis in experimental mouse models of peanut allergy (168) ( Table 3 ).

Alternatively to anti-IgE molecules, a recent approach using anti-FcϵRIα monoclonal antibodies strongly suppressed IgE-mediated MC activation in a humanized mouse model of food allergy and anaphylaxis, revealing another promising therapeutic option (169) ( Table 3 ).

Furthermore, inhibition of FcϵRI-mediated signaling using Bruton Tyrosin Kinase (BTK) inhibitors, significantly reduced degranulation and cytokine production in human MCs and basophils, decreased bronchoconstriction in isolated human bronchi, and proved effective in preventing anaphylaxis in a passive systemic anaphylaxis model using humanized mice (170, 188) ( Table 3 ). Although ibrutinib is well known for its gastrointestinal, cardiovascular and hematological side effects, newer generation molecules, like acalabrutinib, show better safety profile and could become effective, fast-acting oral treatments (189). To this date, however, no clinical trials using BTK inhibitors in food allergy are on-going.

Current evidence suggests that IgE from atopic individuals show an increased sialic acid content, contrary to subjects with no atopy, thus pointing at an important role of sialylation to determine IgE allergenicity (40). Neuraminidase-induced de-sialylation of IgE in a non-FcϵRI dependent manner also diminished downstream signaling in MCs (40). Therefore, de-sialylation of IgE promises to decrease IgE allergenicity, without disrupting non-allergenic IgE activity ( Table 3 ). However, sialidases are ubiquitously expressed in human tissues and play an important role in a variety of physiological and pathological processes, including tumor, infection and inflammation (190), hence the manipulation of the sialylation axis remains an ambitious goal. Notwithstanding, selective small molecule inhibitors of human sialidases hold a great potential for therapeutic development and warrants further investigation (191).

Cytokines involved in Th2 responses, such as IL-4 and IL-13, promote MC proliferation, FcϵRI expression, IgE-mediated degranulation and cytokine production, adhesion and chemotaxis (171, 192, 193). IL-4 and IL-13 receptors share a common alpha chain (IL-4Rα), broadly expressed on lymphocytes, granulocytes and MCs, forming different functional heterodimers according to the associated beta chain (i.e. IL-4R Type I and II, IL-13R), which ultimately activate the intracellular signal transducer and activator of transcription 6 (STAT6) via the phosphorylation of Janus Kinases (Jak1-3, Tyk2) (194, 195). In particular, the proliferation and chemotaxis of MCs induced by IL-4/IL-4R engagement in mucosal interfaces are crucial for the amplification of local allergen responses and responsible for augmented permeability in the intestines and enhanced sensitivity to food allergens and anaphylaxis in experimental mouse models (196–198).

Alongside classical Th2 cytokines, MCs respond rapidly to tissue damage signals such as IL-33 and thymic stromal lymphopoietin (TSLP), alarmins produced mostly by epithelia, innate lymphoid cells and, in some conditions, by MCs themselves (199, 200). IL-33 is known to promote maturation and survival of MCs, enhance the production of pre-formed mediators (e.g. tryptase, serotonin) (201), cytokines (e.g. IL-4, IL-6, IL-13, GM-CSF), and chemokines (e.g. CCL2, CCL17) (201–203), while inhibiting the expression of regulatory cytokines, such as IL-10 (204). Furthermore, IL-33 potentiates IgE-mediated degranulation (202). However, a long-lasting IL-33 stimulation downregulates FcϵRI expression in human MCs, thus inhibiting IgE-dependent MC activation (201), and generating a hyporesponsive phenotype in both mouse and human MCs (205).

TSLP shares common properties with IL-33. They both promote the proliferation and differentiation of MC progenitors (206), and the production of pro-inflammatory cytokines (IL-5, IL-6, IL-13, GM-CSF) and chemokines (CXCL8, CCL1) without inducing the release of pre-formed granule mediators (207). In a food allergy mouse model, TSLP participates in the skin sensitization to food antigens, promoting basophil recruitment and initiating Th2 responses, whereas IL-33 is essential for gut-mediated sensitization and effector responses, including anaphylaxis (208).

Several anti-cytokine treatments have shown promising results in food allergy. The monoclonal antibody dupilumab, blocking IL-4 and IL-13 from binding to the IL-4Rα chain, is currently approved for treatment of severe atopic dermatitis and asthma ( Table 2 ). IL4Rα blockade broadly reduces Th2-responses (171) while increasing Treg suppressive responses (98), reduces eosinophil infiltration (171) and MC proliferation in mucosal tissues of IL4Rα^−/− mice (198). Dupilumab potentially inhibits MC priming and enhancement of IgE-mediated responses by IL-4 (171) ( Table 3 ), while hampering B cell activation and IgE synthesis in mice (171, 209). In fact, recent evidence shows an important role of dupilumab in modulating B cell recall responses, as demonstrated by the reduction of peanut-specific IgE production by human B cells in vitro, and sustained inhibition after in vivo re-exposure in a peanut anaphylaxis mouse model (210) ( Figure 1B ). Albeit limited to a single case report, dupilumab is an efficient therapeutic option for multiple co-occurring food allergies (211), and under clinical trial as treatment for peanut allergy ( Table 1 ).

The upstream role of IL-33 and TSLP in promoting Th2 responses makes them interesting targets for the treatment of atopic conditions, including food allergy (36) ( Figure 1C ). In a Phase II study 73% of peanut allergic patients treated with the anti-IL-33 antibody etokimab achieved tolerance to target peanut dose, showing reduced IL-4, IL-5, IL-13 and IL-9 production after an in vitro T cell challenge with peanut extract, along with reduced peanut-specific IgE levels compared to the placebo arm (107) (NCT02920021, Table 1 ). As for TSLP blockade, mouse models suggest some efficacy, in combination with either IL-25 or IL-33 receptor monoclonal antibodies, in preventing sensitization to food allergens, and promoting tolerance in association with oral immunotherapy (172) ( Table 3 ). Anti-TSLP (tezepelumab, AMG 157, MEDI9929) has been successfully used in reducing allergen-induced bronchoconstriction and indexes of airway inflammation in patients with allergic asthma (NCT01405963) (134, 212) and is currently under investigation in a study combining tezepelumab with allergen-specific immunotherapy for the induction of tolerance in subjects with cat allergy (NCT02237196, Table 2 ). However, no clinical studies assessing the efficacy of anti-TSLP treatment in food allergy are currently on-going.

Allergen-specific immunotherapy (AIT) is the only disease-modifying intervention currently available to treat some allergic conditions, like insect venom allergy, allergic rhinitis and asthma due to respiratory allergy to pollens and house dust mites (234–237).

AIT consists in the repetitive exposure to escalating doses of native allergen extracts, which might induce generalized MC and basophil activation. The risk of eliciting an anaphylactic episode is mitigated by starting with very low allergen doses, by being performed only by trained professionals and in safe conditions under careful monitoring of potential early signs of systemic reaction (236, 238). The timing of dose increase depends on the protocol, ranging from weeks in conventional AIT to days/hours in rush/ultra-rush protocols (234, 238).

The concerted activity of cells from both innate and acquired immunity contributes to the efficacy of AIT (34–36), ultimately eliciting antigen-selective inhibition of MC and basophil activation and long-lasting suppression of IgE-mediated responses at large. In fact, AIT induces a pro-tolerogenic state, promoting allergen-specific IgG/IgG4 production opposed to sIgE by B cells (15). IgG and IgG4 not only selectively compete with IgE in allergen binding, but also the engagement of the FcγRIIb receptor by allergen-IgG complexes cross-linking with surface IgE-FcϵRI actively inhibits MC activation (218, 239–242). IgG-mediated inhibition also prevents further amplification of IgE production, by reducing Th2 cytokine release from activated MCs and basophils (242).

AIT also promotes the development of Tregs, which suppress MC activities, not only by secreting the anti-inflammatory cytokine IL-10, but also inducing MC cell anergy via OX40L receptor engagement (15, 27, 243). OX40-OX40L binding on MCs activates downstream signalling by C-terminal Src kinases, suppressing Fyn kinase activity and impairing microtubule rearrangement and degranulation (243) ( Figure 1E ).

Although effective, these events require time to induce a protective response, while exposure to incremental doses of allergen rapidly desensitizes MCs. However, the mechanism explaining such effect remains unclear. A study suggests that rapid incremental IgE receptor occupancy induces the depletion of cell surface IgE by internalization of IgE–FcϵRI complexes (244). Others find in desensitized anergic MCs an impaired internalization of allergen–IgE–FcϵRI molecules (245), and aberrant rearrangements of cytoskeleton actin fibers that inhibit FcϵRI-mediated calcium flux and intracellular vesicles trafficking (246).

Rapidly desensitized MCs, in turn, produce IL-2 that contribute to Treg survival and recruitment in the periphery, hence indirectly contributing to peripheral tolerance, as demonstrated in mice (247).

Both tolerance induction and MC desensitization are widely exploited to achieve long-term modulation and quick onset protection of allergic reactions with rush/ultra-rush protocols, respectively (248).

For both treatment and prevention of severe reactions upon accidental exposure to food allergens, increasing the maximum tolerated dose of allergen is necessary and can be achieved with AIT (249).

AIT in food allergy is performed using either native allergens (e.g. whole food, allergen extracts) administered via the oral, sublingual or epicutaneous routes, or baked allergens (alone or mixed with other ingredients creating a food matrix) via the oral route (250).

Recently, the first peanut allergen powder formulation (AR101) was approved for peanut AIT by the U.S. Food and Drug Administration and European Medicines Agency (251, 252), and numerous other trials using either whole peanut or peanut extracts promoted tolerance to varying doses of crude peanut in 60–80% of treated subjects (72, 81, 83, 85, 90, 92, 94) ( Table 1 ). However, the safety of AIT protocols in food allergy is still a matter of debate, since the risk of a severe allergic reaction during AIT cannot be completely abated (253). In fact, a 1–21% frequency of systemic adverse reactions and increased occurrence with higher peanut end goal doses were observed in peanut AIT trials (254). Furthermore, while long-term treatment is effective in preventing severe allergic reactions in AIT responders (79, 92), a fraction of subjects might still experience anaphylaxis with previously tolerated allergen doses when aggravating co-factors are present (i.e. physical exercise, use of non-steroidal anti-inflammatory drugs, infections, etc.) or due to poor AIT adherence (79, 253).

To increase AIT safety in food allergy, newer therapeutic strategies involve the combination of AIT with biologics. Evidence suggests that omalizumab administered during AIT reduces the risk of severe reactions and facilitates AIT (97, 99, 101) ( Tables 1 and 2 ). In fact, while AIT caused an increase in the levels of inhibitory allergen-specific IgG4, in the threshold for MC responsiveness and a reduction of Th2 cytokine production (83, 84, 92, 239), omalizumab decreased the likelihood of basophil degranulation, especially relevant during dose escalation (101). This omalizumab-induced protection is most likely dependent on basophil IgE–FcϵRI disengagement, as suggested by empirical evidence (159) and omalizumab pharmacokinetics.

However, studies on long-term use of omalizumab in cow’s milk AIT proved long-term omalizumab add-on treatment not being cost-effective, albeit the higher safety profile (255) ( Table 2 ). Further trials testing the utility of omalizumab adjunct to food AIT, or other biologics like dupilumab with AR101 (NCT03682770) are currently ongoing ( Tables 1 and 2 ).

Allergen-dependent strategies alternative to AIT are currently being tested. Among these, the use of hypoallergenic molecules, lacking key anaphylactogenic conformational epitopes, promises to obtain safer alternatives to AIT using native allergen extracts, as observed in fish and peanut allergy studies conducted in humans and mice, albeit still in early development (256, 257).

Other therapeutic approaches involve antibodies targeting major allergenic molecules, like a recently developed monoclonal anti-Ara h 2, preventing both local and systemic allergic reactions, as tested in a mouse model of peanut allergy (177) ( Table 3 ). The advantage of monoclonal treatment is not only given by their competition with IgE molecules in allergen binding, but also by sharing with endogenous allergen-specific IgG antibodies the same mechanisms, regardless of patients’ capacity to mount an effective anti-allergic immune response as in conventional allergen immunotherapy. However, subjects sensitized to multiple allergen epitopes might only partially benefit from such treatment, unless multiple monoclonal antibodies against different epitopes are used in combination.

The complexing of allergenic epitopes with molecules actively promoting a tolerogenic state (i.e. production of IL-10, induction of IgG4, generation of Tregs), such as Toll-like receptor ligands (i.e. CpG, LPS, R848), viral-attenuated molecules, Siglec-engaging tolerance-inducing antigenic liposomes (STALs) and nanoformulations, is used as adjuvant immunotherapy to elicit allergen-specific tolerance (258).

An alternative approach under study is the use of plasmid DNA-based vaccines. Such vaccines induce the production of specific exogenous target proteins via allergen-coding DNA particles, exploiting the natural immune pathways leading to the production of IgG to promote long-lasting tolerance (259). In addition, peptide vaccines aimed at eliciting IgG antibody production targeted against highly allergenic epitopes are also currently under scrutiny (260).

Several recent studies on nanoformulations and adjuvant immunotherapy candidates for cow’s milk and peanut allergy have been conducted, showing promising results in mouse allergy models (183, 184, 261, 262) ( Table 3 ). In humans, few ongoing clinical trials on DNA-based vaccines (ASP0892, NCT03755713; ASP0892, NCT02851277) and modified allergen proteins (HAL-MPE1 subcutaneous AIT, NCT02991885) are currently in Phase I, while a previous attempt with attenuated E. Coli Ara h 1-2-3 recombinant vaccine candidate failed to promote tolerance, inducing severe adverse reactions in 20% of participants (96) ( Table 1 ).