2.4. Protein and Protein Fusions

Although recent studies have predominately used particle-based therapeutics, such as nanoparticles or cells to deliver antigen to B cells, soluble protein therapeutics are on the horizon. Benefits of these biologics include their small size, bioavailability, solubility, and increased specificity for targeted B cells. In a proof of concept study, Henry et al. used an anti-insulin monoclonal antibody (mab123) to target insulin-reactive B cells whose BCR is occupied by endogenous insulin using a VH125 heavy chain transgenic NOD mouse model of diabetes [35]. Insulin-reactive B cells were significantly reduced after injection of mab123 in vivo, likely through FcγR mediated depletion. Moreover, since mab123 does not recognize insulin bound to the insulin receptor [36], mice remained normoglycemic in nondiabetic mice. Importantly, this group demonstrated that systemic administration of mab123 to WT NOD mice resulted in protection from diabetes development [35].

While this method demonstrates the feasibility of using monoclonal antibodies to target antigen-specific B cells, it requires the BCR of autoreactive B cells to be occupied by antigen. In addition, it would not target B cells that recognize an insulin epitope bound by the 123 antibody, nor immediately adjacent epitopes due to steric hindrance. Moreover, there is concern that administration could result in formation of insulin antibody complexes and ultimately impair the action of physiologic action insulin. Hence to overcome these challenges, Akston Biosciences has created a designer Fc-insulin fusion protein, AKS-107, in which a hormonally inactive insulin is fused to an IgG Fc to specifically target insulin-reactive B cells for complement and FcγR mediated killing or induction of anergy (Figure 1D). This approach has the added advantage that the IgG Fc should extend the in vivo half-life of the drug by interaction with FcRN. Promising pre-clinical studies in mice demonstrate AKS-107 induces specific silencing of insulin-reactive B cells and can prevent autoimmune diabetes (unpublished), providing the basis for a first-in-human Phase 1 clinical trial of AKS-107. Results will greatly inform our understanding of the safety and potential applications of autoantigen-IgG Fc fusions to treat T1D, with implications for other autoimmune diseases in which pathogenic antigen-specific B cells are implicated.

While not targeting antigen specific B cells, antibody-based therapies that induce B cell unresponsiveness (anergy) through engagement of FcγRIIb have shown efficacy in treatment of autoimmunity and avoid the harmful side effects of total B cell depletion. FcγRIIb is an inhibitory surface receptor expressed primarily by B cells that when co-engaged with the BCR mediates recruitment of phosphatases that suppress B cell signaling [37]. Xencore Inc developed XmAb5871, which is a humanized mAb that binds to CD19 in the BCR complex and is engineered to have increased affinity for FcγRIIb. Preclinical studies have shown that co-engagement of CD19 and FcγRIIb by XmAb5871 suppresses B cell activation, inhibits upregulation of CD86, and suppresses antibody production [38,39]. Recently a Phase 2 clinical trial in SLE patients demonstrated that XmAb5871 is well tolerated and significantly suppressed disease recurrence [40]. Another antibody that exploits the inhibitory function of FcγRIIb is the bispecific antibody, PRV-3279 (formally known as MGD010), which binds to CD79b, a signaling component of the BCR, and FcγRIIb [41]. The first-in-human study of MGD010 demonstrated it is tolerated, does not cause B cell depletion but induces downregulation of surface BCR and suppression of BCR-mediated Ca²⁺ mobilization, indicative of B cell anergy [42]. Plans to address its potential use in autoimmunity are currently in development. Still, another approach emulates the ability of chronic low avidity antigen exposure, as occurs with soluble autoantigens in vivo, to induce B cell anergy. Antibodies lacking antibody-dependent cell-mediated cytotoxicity (ADCC) and complement fixing ability that are reactive with CD79 induce polyclonal anergy. As shown by the Cambier group and others, these antibodies are effective in mouse models of lupus, T1D, RA and MS (EAE) [43,44] (Cambier unpublished). While the above candidate therapeutics silence all B cells, the technology could be exploited to generate antibodies that engage the BCR in an antigen-specific manner, similar to AKS-107, which would result in tolerance induction of pathogenic B cells while sparing non-pathogenic B cells (Figure 1D).

Another carrier-free antigen-specific therapeutic on the horizon incorporates a drug or toxin directly conjugated to a self-antigen, similar to antibody-drug conjugates (ADCs) currently in use to treat some forms of cancer [45,46,47,48]. In principle, antigen-toxin conjugates would bind specifically to the BCR on pathogenic self-reactive B cells, inducing death by virtue of concentration and cellular uptake of the toxin (Figure 1D). In a proof-of-principle study, Pickens et al. demonstrated that conjugation of the self-antigen peptide, PLP_139-152, to the corticosteroid, dexamethasone, led to protection from development of EAE in mice with increased efficacy compared to dexamethasone treatment alone [49]. While this study utilized a self-antigen fragment and not full-length protein and did not analyze the effects on the immune system in depth, it demonstrates the potential of this type of therapeutic to prevent or delay autoimmunity. Preliminary studies in the Cambier lab have shown that insulin conjugated to the toxin, mitoxantrone, induces specific deletion of murine insulin-binding B cells in vitro (unpublished). Future studies are needed to fully address the therapeutic potential of antigen-toxin conjugates in the treatment of autoimmune disease.