Chopping down antibodies for a good cause

Main text

Patients are often excluded from adeno-associated viral (AAV) gene therapy because of pre-existing humoral immunity (stemming from natural infection and posing a particularly strong obstacle for intravenous vector administration). While various strategies to break through the barrier of neutralizing antibodies have been explored, perhaps the most promising method employs bacterial enzymes that degrade human immunoglobulins in vivo.^1,2 One example is the endopeptidase imlifidase (IdeS) derived from Streptococcus pyogenes, which degrades all 4 subtypes of human immunoglobulin G (IgG) (Figure 1A).³ In this issue of Molecular Therapy, Smith et al. introduce an engineered enzyme capable of cleaving both IgM and IgG, thereby making the approach substantially more powerful.⁴

Figure 1.

Differences between the effects of imlifidase and IceMG on anti-AAV antibodies

(A) Imlifidase cleaves IgG into F(ab) and Fc fragments but does not affect IgM.

(B) IceMG cleaves both IgG and IgM in F(ab) and Fc fragments, in addition to removing surface-bound monomeric IgM (BCR) from B cells. Created with BioRender.com.

While the focus thus far had mostly been on IgG formation, humans may also have circulating IgM with neutralizing activity against AAV capsid, a problem that is addressed with the dual-activity enzyme. Since immunoglobulins are cleaved regardless of their antigen specificity, the approach is serotype independent and could therefore be broadly applied. This feature is particularly attractive given the tendency of patients to have antibodies that react to multiple serotypes. Smith et al. mined bacterial papain-like proteases through phylogenetic analyses and structural modeling with molecular docking tools to identify prospective enzyme candidates that may recognize IgM constant domains as substrates.⁴ A candidate enzyme from Lachnoanaerobaculum saburreum was identified as a human IgM cleaving enzyme (IceM), cutting IgM into F(ab) and Fc fragments below Cμ2 of the heavy chain. IceM binds human IgM heavy chain at a pocket adjacent to the enzyme’s active cleavage site, homologous to the binding domain of IdeS for human IgG. Specificity of IceM for IgM and not IgG is due to amino acid differences in contact residues within the major binding pocket. Next the authors engineered an enzyme with dual proteolytic activity by using a rigid linker to fuse IceM with IdeZ, an IdeS homolog derived from Streptococcus equi with specificity to human IgG.⁵ This multi-cleaving enzyme, called IceMG, bound both IgM and IgG within their respective catalytic pockets (Figure 1B). Both IceM and IceMG not only cleave circulating IgM but also the cell surface-expressed B cell receptor (BCR), thereby downregulating IgM-mediated signaling. Surface IgM baseline levels are restored between 12 and 24 h post treatment, indicating that BCR inactivation is transient and reversible.

Antibody reduction with IceMG treatment is highly transient compared to B cell depletion with rituximab, which takes months for B cells and circulating antibodies to recover.^6,7 Post IceMG treatment, IgM and IgG are almost completely cleaved at 1 h and remain significantly lower for up to 72 h. However, circulating F(ab) fragments can continue to neutralize AAV even though their circulating half-life is significantly reduced in the absence of the Fc domain.⁸ Therefore, identifying the right window for effective AAV gene therapy following treatment with IceMG is critical. Besides interfering with gene transfer, antibodies against capsid may also cause immunotoxicities. At high vector doses, complement activation has been a source of serious complications in systemic AAV gene therapy, leading for example to thrombotic microangiopathy (TMA). A recent investigation documented the essential role of antibody formation against capsid in development of TMA, which is linked to classical complement pathway (and likely amplified by alternative pathway activation by the capsid).⁹ Because IgMs are potent activators of complement, IceMG may also have utility in preventing TMA and other toxicities related to complement activation.

Currently, patients might not receive life-saving gene therapy because of pre-existing immunity. Others may have alternative treatment options but desire gene therapy because they seek long-term treatment from a single drug administration. In addition, gene therapy can accomplish superior outcomes by providing consistent levels of the therapeutic protein. Pre-existing immunity, however, can take away this option. Clinical studies will reveal whether gene transfer in pre-immune patients treated with IdeS or IceMG could have undesired consequences that will have to be dealt with. For instance, IdeS is in clinical trial to evaluate safety of AAV-microdystrophin gene therapy in Duchenne muscular dystrophy patients with pre-existing antibodies to rAAVrh74 (NCT06241950). Concerns include reactivation of memory T cells that could potentially target AAV-infected cells and of memory B cells, which may produce high-titer antibodies soon after the enzyme is discontinued. Nonetheless, we are cautiously optimistic that immunoglobulin-degrading enzymes will be an effective tool, possibly in conjunction with other drugs, to circumvent pre-existing immunity and reduce the risk for immunotoxicities, so that more individuals with diverse diseases may benefit from AAV gene therapy. IceMG expands the power and utility of this approach.