Alternative Approach to Immunomodulation for Type 1 Diabetes: Antigen-Specific Immunotherapy In Utero

Type 1 diabetes (T1D) results from the T-cell–mediated destruction of insulin-secreting pancreatic β-cells (1). While we can treat this condition with insulin replacement, long-term outcomes for patients is suboptimal (2), and a “cure” for T1D is a very worthy goal. At a time when significant progress is being made with biological therapies, surely a cure has to be possible.

Many approaches to achieving this Holy Grail have acted through the suppression of T-cell function, often with drugs that we are familiar with in the context of solid organ transplantation (cyclosporine) or other organ-specific autoimmune diseases (rituximab) (3,4). These drugs come with risks associated with “off-target” immune suppression, such as reactivation of latent infection (5). A more elegant approach to immunotherapy is to suppress only the T cells with the potential to target the β-cell. This approach, termed antigen-specific immunosuppression, involves vaccinating with β-cell proteins so that β-cell–specific T cells do not subsequently respond to and attack them (6). While antigen-specific immunosuppression does not appear to be as powerful as pan-immunosuppression, their favorable side effect profile clearly makes them a more attractive option.

It has been over 30 years since the first immunotherapy trials for T1D. Not one has managed to achieve clinically meaningful benefit with a therapy that has an acceptable risk profile (reviewed by Skyler [7]). There are a number of potential reasons for this. One is that the immune process is very well entrenched by the time a patient is diagnosed with the condition. In many patients, pancreatic immune attack will have been in progress for over 10 years when hyperglycemic symptoms prompt the diagnosis of T1D. Therefore, at the time patients are entered into clinical trials, there will be a significant number of β-cell–reactive T cells, and they will have evolved to target multiple β-cell proteins. Perhaps most significantly, there will be T cells displaying the characteristics of memory. These cells survive for years, are able to travel across and survey multiple body compartments, and respond faster and with greater magnitude to β-cell proteins (8).

Therefore, the current belief is that we should be moving to stronger and combination therapies. A powerful example of this is autologous nonmyeloablative hematopoietic stem cell transplantation combined with immunosuppression, which achieves high rates of remission from T1D (9). However, this comes with significant risk of therapy, arguably comparable to that associated with another potential cure for T1D, a pancreas transplant.

Rather than going in harder with stronger combinatorial immunotherapy, an approach taken by Culina et al. (10) in this issue of the Diabetes was to go in early—very early. They used a murine model of T1D to demonstrate that antigen-specific immunotherapy delivered to the fetus is a very effective approach to preventing T1D.

Culina et al. (10) chose proinsulin as the vaccinating β-cell protein because proinsulin is an early target of the T-cell response against the β-cell. They chose to treat at embryonic day 16 because T-cell education starts in the thymus just after this age. They chose a CD8 T-cell–based murine model of T1D (where the T cells target a well-defined proinsulin peptide epitope) because CD8 T cells are the predominant infiltrating T cell within the inflamed T1D islet. A clever touch was the technique used to get the proinsulin into the embryo. Proinsulin was fused to the constant region of an antibody molecule (Fc) to facilitate the transport from the mother to the fetus through the neonatal Fc receptor (FcRn). The FcRn is highly expressed in the placenta (and in the neonatal gut) and is a route through which maternal antibodies are ferried to the embryo (or the neonate) as part of imparting immunity (11). Treatment reduced disease by threefold in the offspring (Fig. 1). This was associated with a change in the nature of the CD8 T cells: they expressed less granzymes, perforins, and Fas ligands (i.e., they were less toxic). Importantly, the immunoglobulin Fc molecule (PPI-Fc) was seen to accumulate in the thymus—the site of T-cell education and one that is defective in T-cell–mediated autoimmune diseases such as T1D. The thymic accumulation of PPI-Fc was associated with an increase in thymic-derived regulatory T cells, suggesting that treatment can educate thymic β-cell–reactive T cells.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Schematic diagram outlining the technique of in utero antigen-specific immunotherapy used by Culina et al. (10). A pregnant G9C8 mouse expressing a transgenic T-cell receptor for the B15–23 peptide of proinsulin was immunized with preproinsulin fused with PPI-Fc. PPI-Fc was transported to the fetus via the placental FcRn and then to the fetal thymus by migratory dendritic cells (DCs). The presence of thymic PPI-Fc was associated with an increase in thymic-derived regulatory T cells and with less cytotoxic preproinsulin-specific CD8 T cells. Following birth, induction of diabetes in the offspring through injection with proinsulin peptide B15–23 and adjuvant resulted in a threefold reduction of disease than found in the offspring of untreated female mice.

The work by Culina et al. (10) highlighted the effectiveness of treatment early in the natural history of T1D and the tolerogenic nature of the uterine environment. More importantly, where previous antigen-specific immunotherapy studies have targeted T cells after the autoreactive T-cell repertoire has been established, this study described a method whereby T-cell education can occur early in the thymus.

Important questions remain. Exactly what mechanisms are controlling β-cell–reactive T cells? T-cell deletion is a powerful approach to controlling self-reactive T cells. CD8 T-cell numbers actually increased following therapy in utero; can we assume deletional mechanisms are not at play in this model? Are nonthymic regulatory T cells activated with this treatment? Most importantly, can this therapy work in a disease model that more closely replicates human disease, where multiple peptide epitopes within multiple β-cell proteins are targeted? Early results presented with the NOD mouse are promising—treatment reduced β-cell inflammation and reduced the ability of these cells to transfer disease to another model of accelerated diabetes, but longer-term follow-up data are required to determine whether this translates to diabetes protection in the offspring of treated mothers. How will we translate this to patients? While T1D prediction strategies continue to improve (12), we are still some way off to accurately predicting clinical disease before the onset of β-cell autoimmunity and on the basis of genetic markers. Finally, the administration of therapies during pregnancy invokes understandable concerns. While these concerns relate to therapies that we do not want to pass on to the developing fetus, clear safety and efficacy needs to be demonstrated before we can take antigen-specific immunotherapy in utero to patients.

However, the work of Culina et al. (10) is important because it reminds us of another discussion that we should be having in the field of T1D immunotherapy—that of therapy that is early enough to educate developing T cells in the thymus. As the authors remind us, the FcRn is also present in the neonatal gut (a time where T-cell education is still ongoing), thus allowing therapy to be administered soon after birth. At a time when we are faced with the side effects associated with combinatorial immunotherapy administered after the onset of autoimmunity, this surely is worth exploring further.

• © 2015 by the American Diabetes Association. Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered.