Vaccine therapy for rheumatoid arthritis: A step forward

Main text

Autoimmune diseases result from a favorable interplay of environmental factors in genetically predisposed individuals. Standard treatments for most autoimmune disorders target inflammation but also trigger systemic immunosuppression without treating the underlying problem—a lack of immune tolerance to self-antigens. In this issue of Molecular Therapy, Romero-Castillo et al. publish their findings on how peptide-based vaccination may open new avenues for treating rheumatoid arthritis (RA) by inducing tolerance without the need for systemic immunosuppression.¹ The authors report the development of a novel RA vaccine composed of recombinant HLA-DRB1 04:01 bound to type II collagen (COL2) peptide COL2_259-273 that was chemically modified at glycine position 265 and could effectively suppress arthritis in human DRB1 04:01-expressing mice, thus highlighting its therapeutic potential.¹

Antigen-specific therapies can potentially target pathogenic autoreactive lymphocytes while preserving protective immune responses. Such therapies include approaches involving direct depletion of pathogenic autoreactive cells or altering their function and inducing anergy by establishing peripheral tolerance via interventions such as delivery of antigen-specific regulatory T cells (Tregs) or their induction by tolerogenic vaccination. Establishing peripheral tolerance via delivery of autoantigen in the presence of tolerogenic signals (e.g., presence of tolerogenic dendritic cells [DCs] or lack of costimulatory molecules) allows for selective silencing of autoreactive T cells while preserving protective immune responses. Treg-based therapies, tolerogenic DC-based therapies, chimeric antigen receptor (CAR) T cell therapies (such as antigen-specific chimeric autoantibody receptor [CAAR]), and tolerogenic vaccine therapies are the main antigen-specific therapies that have been used to induce tolerance in autoimmune diseases. Tolerogenic DCs express low levels of costimulatory molecules (i.e., CD80, CD86, and CD40), thus depriving T cells of the necessary second activation signal, leading to a hyporesponsive/anergic state and the expression of anti-inflammatory molecules, e.g., transforming growth factor β (TGF-β), interleukin (IL)-10, and indoleamine 2,3-dioxygenase (IDO), which are conducive to establishing a tolerant environment and the induction or expansion of Tregs. CAR T cell-, Treg-, and DC-based therapies are generally more costly than autoantigen vaccine approaches, as they require good manufacturing practice (GMP)-compliant manufacturing and must meet stringent regulatory standards for cell production.

RA is genetically associated with HLA-DRB1 ∗04:01, ∗04:04, ∗04:05, and ∗01:01 and ∗01:02 haplotypes which show strong affinity to citrullinated (cit) peptides that act as the autoantigens in 70%–90% of clinical cases.² Common cit autoantigens include fibrinogen/fibrin, vimentin, enolase-I, histone 4, Epstein-Barr virus nuclear antigen (EBNA1), and the immunodominant COL2.³ Early clinical trials for RA treatment involved the use of autologous tolerogenic DCs that were either loaded with cit peptides⁴ or exposed to synovial fluid,⁵ which increased the Treg/effector ratio and reduced pro-inflammatory cytokines. In subsequent studies, liposomes loaded with COL2_259–273 peptide together with calcitriol (nuclear factor κB [NF-κB] inhibitor) were shown to induce exhaustion in COL2-specific T cells as well as cit-vimentin-specific T cells (via the bystander effect).⁶ Furthermore, in SKG and CIA (collagen-induced arthritis) models of RA, all-trans-retinoic acid (ATRA) encapsulated in biodegradable poly-(lactic-co-glycolic) acid (PLGA) microparticles was retained in arthritic mouse joints after intra-articular (IA) injection and could enhance the differentiation of naive T cells into Tregs, promote the stability of pre-existing natural Tregs via epigenetic modulation, and induce tolerization of local DCs.⁷ In addition to liposomes and microparticles, polymer-based nanoparticles consisting of antigenic peptide-major histocompatibility complex (MHC) class II complexes have also been used. Notably, in a HLA-DR4-IE-transgenic C57BL/6 I10.M mouse model of CIA (MHC class II knockout mice expressing a transgenic hybrid MHC class II composed of a peptide-binding domain of human HLA-DR4 and a membrane-proximal domain of mouse IE [DR4-IE]), nanoparticles displaying the mouse collagen II (mCII)_259–273/DR4-IE complex could reduce joint inflammation in association with systemic expansion of cognate IL-10- and TGF-β-producing Tr1 (type 1 Treg)-like T cells.⁸

The Holmdahl group previously described a collagen-based vaccine therapy comprising a complex of a mouse MHC class II (A^q) molecule bound to a galactosylated COL2_259-273 peptide (at position 264), which interacts with antigen-specific T cell receptors (TCRs) and protects against arthritis in a model of CIA.⁹ However, because glycosylated vaccines are hard to produce and scale, and since patients with RA have T cells that recognize non-glycosylated COL2, the authors adapted their design for humans by creating a version called “DR4-AL179.” This consists of HLA-DRB1∗04:01 linked to a non-glycosylated COL2 peptide (residues 259–273), where position 265 was replaced with glycine and modified with a hydrophobic group. Given that position 265 is a key position for TCR recognition, this modification allowed for enhanced interaction of the peptide with the TCR without compromising MHC class II binding. Compared to the native non-glycosylated DR4-COL2_259-273 vaccine, which had no effect on arthritis, the DR4-AL179 vaccine reduced both the incidence and severity of arthritis. It led to lower disease scores, less immune cell infiltration and bone damage, and better preservation of collagen fibers. Additionally, there was a significant reduction in COL2-specific T cells, anti-COL2 immunoglobulin (Ig)G antibody levels, and arthritogenic U1 antibody levels in the DR4-AL179 group as compared to the DR4-COL2_259-273 group upon vaccination. T cells from the DR4-AL179-vaccinated mice had higher levels of CD73, FR4, PD-1, and VISTA—usually associated with a tolerogenic phenotype. To show that VISTA-expressing Tregs are important for the vaccine’s effect, the authors used anti-VISTA antibodies, which reduced the ability of these Tregs to treat disease after transfer into Parker (DRB1∗04:01-transgenic) mice immunized with COL2/CFA intradermally. This confirmed the key role of VISTA⁺ Tregs in reducing arthritis after vaccination.

This work incorporates multiple progressive approaches for developing an effective RA vaccine for humans. Firstly, it uses a simplistic design overriding the requirement of carrier molecules (like nanoparticles or polymers) and adjuvants, thus improving safety (by avoiding unknown carrier-specific immune responses) and minimizing manufacturing costs. Second, the improvised design without the need for galactosylation would significantly benefit the scalability and clinical translation, allowing better reproducibility for vaccine purity. Third, the vaccine was evaluated in DRB1 ∗04:01-expressing mice which display physiological levels of human MHC class II with allele-restricted immune responses and arthritis susceptibility, thus closely mirroring the genetic associations observed in human RA. Finally, the vaccine was administered via three subcutaneous injections rather than continuous delivery through osmotic pumps, providing a more practical method of delivery that aligns more closely to human clinical protocols. Collectively, the results show that the vaccine has immense potential for translatability to the clinic.

Mechanistically, the study attributes vaccine efficacy mainly to VISTA⁺ non-conventional T(regs). While this subset is clearly involved, immune tolerance is multifactorial. Sole focus on VISTA⁺ COL2-specific T(regs) may overlook other contributors in the broader regulatory landscape. The observed increase in CD73^highFR4^high and PD-1⁺ T cells suggests potential roles for anergy or exhaustion pathways. Although the role of PD-1 was interrogated, it is now increasingly seen more as a marker of Treg exhaustion than suppressive function.¹⁰ Therefore, it would be valuable to assess bona fide suppressive molecules like CTLA-4. Additionally, the involvement of myeloid suppressor cells or tDCs should be considered for future studies. Finally, it remains unclear whether the vaccine expands the overall number of COL2-specific FOXP3⁺ Tregs and whether there is an expansion of Tregs specific to other RA-relevant autoantigens.

Given the growing understanding of molecular mechanisms of Treg suppressive functions, future studies should also examine whether the vaccine induces “infectious tolerance” (a mechanism by which Tregs induce other naive or effector T cells to acquire regulatory function) and “linked suppression” (a phenomenon where Tregs specific for a given antigen can suppress immune responses to other antigens presented by the same antigen-presenting cell [APC]) (Figure 1). Specifically, future works should investigate whether tolerance induced to a dominant autoantigen like COL2 can extend to other RA-associated epitopes, such as cit vimentin or fibrinogen, thereby broadening the therapeutic impact through antigen spreading of regulatory function.

Figure 1.

Mechanism of infectious tolerance mediated by Tr1 cells after tolerogenic vaccination

The vaccine delivers collagen-derived antigen, leading to the induction of antigen-specific Tr1 cells. These Tr1 cells promote infectious tolerance by creating a local immunoregulatory environment that enables the suppression of immune responses against other rheumatoid arthritis-related autoantigens, such as vimentin. This results in the transfer of tolerance from the initial vaccine antigen to additional disease-relevant targets. This presentation was created in BioRender.

Acknowledgments

The research reported in this publication was supported by the following grant from the National Institute of Allergy and Infectious Diseases (NIAID): R21AI175813 (M.K.-M.).

Declaration of interests

The authors declare that they have no conflicts of interest to disclose.