A novel Lactococcus lactis-based antigen-specific platform for the treatment of type 1 diabetes: Essential steps towards successful clinical translation.

In type 1 diabetes (T1D) the insulin-producing beta cell is destroyed by the aberrant immune system leading to blood glucose dysregulation. To date, patients are dependent on external insulin for survival however, limitations such as hypoglycemia unawareness and suboptimal glycemic control leading to micro- and macrovascular complications still exist. Furthermore, although a combination of serum islet autoantibodies, genetic factors, and metabolic features can predict those who will be affected with high accuracy, current pharmacological therapies do not prevent or slow down disease progression. Moreover, interventions in patients with newly diagnosed T1D using clinically acceptable doses of a single systemic immunomodulatory drug, like aCD3, or antigen-specific approaches, like oral insulin, have not lived up to expectations. In addition, their efficacies are not uniform among T1D patients, highlighting on the one hand the complexity of this multifactorial disease and on the other hand the heterogeneity of this patient population. Oral autoAg delivery as a means to induce mucosal tolerance is hindered by practical challenges related to gastric protein degradation. Recombinant biologically contained live L. lactis bacteria have proven to be safe and well tolerated in pre-clinical animal models as well as several clinical trials making these bacteria ideal biofactories to synthesize and deliver autoAgs to the gastrointestinal tract. GM L. lactis secreting proinsulin (PINS) along with the biologically active immunoregulatory cytokine IL-10 (LL-PINS+IL-10) were able to successfully restore glucose homeostasis in newly diagnosed NOD mice when combined with low-dose aCD3. This supports the idea that has been growing for several years now in the field of T1D where preclinical research evidence has encouraged experts to propose combinations of general immunomodulation and Ag-specific approaches as the best strategy to stably cure or prevent T1D. However, clinical translation has failed in the past due to a lack of biomarkers, dosing, ill-suited end points, timing of intervention and a lack of insights in the mechanism of action. Here, we wanted to further explore these potential pitfalls for the CT of low-dose aCD3 combined with LL-PINS+IL-10. First, we generated a clinical-grade recombinant L. lactis strain appropriate for human application secreting human pro-insulin and IL-10. When combined with low-dose aCD3 this combination therapy (CT) was able to reverse diabetes in 66% of newly diagnosed diabetic NOD mice, comparable to plasmid-driven L. lactis. Initial blood glucose concentrations (<350 mg/dL) and insulin autoantibody (IAA) positivity were predictors of stable reversal of hyperglycemia and decline in IAA positivity was an immune biomarker of therapeutic outcome. Assessment of the immune changes induced by the L. lactis-based CT revealed elevated frequencies of CD4+Foxp3+ T cells in the pancreatic draining lymph nodes, pancreas, and peripheral blood of all treated mice, independent of metabolic outcome. However, CD4+Foxp3+ T cells from CT-cured mice had higher CTLA-4 expression, alluding to a better functionality in vivo. Neutralization of CTLA-4 and TGF-β partially abrogated the suppressive function of therapy-induced Tregs. Ablation or functional impairment of Foxp3+ Tregs in vivo pre- or post-therapy impaired immune tolerance, highlighting the dependence of the CT-induced tolerance in mice with newly diagnosed diabetes on the presence and functionality of CD4+Foxp3+ T cells. Of note, no difference was seen in the degree of insulitis between therapeutic responder and non-responder mice, suggesting also alterations in other immune cell subsets apart from Tregs. Biomarkers identified in this study can potentially be used in the future to tailor the L. lactis-based CT for individual patients. We further explored the potential of this CT outside the window of acute diabetes diagnosis. To do so, we substituted autoimmune diabetic NOD mice, with disease durations varying between 4 and 53 days, with syngeneic islets at the time of therapy initiation. Untreated islet recipients consistently showed disease recurrence after 8.2 ± 0.7 days, while 32% of aCD3-treated and 48% of CT-treated mice remained normoglycemic until 6 weeks after therapy initiation. However, mice that were diabetic for more than 2 weeks before treatment initiation were less efficient at maintaining normoglycemia than those treated within 2 weeks of diabetes diagnosis, particularly in the aCD3-treated group. The efficiency of glycemic control and the rate of graft acceptance in relation to disease duration is in part related to the therapy-induced protection of any residual endogenous islets as reflected by the pancreatic insulin content at different time points after disease diagnosis. The complete elimination of endogenous beta cell mass with alloxan at the time of diabetes diagnosis pointed towards the significance of continuous feeding of the islet antigen PINS together with aCD3 therapy for treatment success. The CT providing PINS protected 60% of mice, compared to only 17% when an irrelevant antigen (ovalbumin) was combined with aCD3, or to 17% with aCD3 therapy alone. Sustained tolerance by aCD3 therapy alone or the CT was accompanied with a reduction of IGRP+CD8+ autoreactive T cells and an increase in Foxp3+CD4+ Tregs. However, after CT with PINS a specific accumulation of Foxp3+ Tregs was observed around the insulin-containing islet grafts. Taken together these data show that the combination of proinsulin and IL-10 via oral L. lactis with low-dose aCD3 therapy can restore tolerance to beta cells in autoimmune diabetic mice, also when therapy is started outside the window of acute diabetes diagnosis. As a general conclusion, the current data presented in this thesis helped design a phase Ia/IIb clinical trial to assess the safety and tolerability clinical-grade LL-PINS+IL-10 alone or in combination with Teplizumab in patients with newly diagnosed T1D. We hope these findings can be corroborated in a human setting and will help fulfil its clinical potential.

Jaar van publicatie:2019

Toegankelijkheid:Open