ATG plus recombinant probiotic expressing proinsulin prevents type 1 diabetes (T1D) in NOD mice

Background: In previous studies of NOD mice, anti-CD3 plus oral delivery of live Lactococcus lactis genetically modified to secrete the autoantigen proinsulin (PINS) with the tolerogenic cytokine IL-10 (L. lactis?PINS/IL-10) reversed T1D. We hypothesized that a short course of low-dose murine anti-thymocyte globulin (ATG) plus daily intragastric administration of L. lactis?PINS or ?PINS/IL-10 could prevent T1D in NOD mice.

Methods: 12-week-old female NOD mice (n=15/group) received two intraperitoneal injections of 250?g ATG or IgG on days 1 and 4 ± L. lactis (2x109 CFU/day, five days/week): 1) ATG+L. lactis?PINS/IL-10, 2) ATG+L. lactis?PINS, 3) ATG+L. lactis?IL-10, 4) ATG+L. lactis, 5) ATG, 6) IgG+L. lactis?PINS/IL-10, 7) IgG+L. lactis?PINS, 8) IgG+L. lactis?IL-10, 9) IgG+L. lactis, and 10) no treatment. Animals were followed for T1D through 32-weeks-of-age. Separately, mice were sacrificed at 13 (n=5) and 15 (n=5) weeks-of-age to study therapeutic mechanism.

Results: Preliminary data confirmed ATG-mediated depletion of CD4+ and CD8+ T-cells (p<0.05), and low-dose ATG was selected to provide room for synergy. Indeed, 9/15 (60%) ATG alone and 13/15 (87%) of IgG+L. lactis?PINS treated animals developed diabetes versus 4/15 (27%) ATG+L.lactis?PINS, indicating a synergistic requirement for both reagents for prevention (p<0.01). The addition of IL-10 did not improve outcomes in this model: 7/14 (50%) ATG+L. lactis?PINS/IL-10-treated animals developed T1D versus 10/15 (67%) untreated controls, unlike mice treated with anti-CD3 mAb, where a clear synergy was observed between anti-CD3, IL-10, and PINS.  Interestingly, mesenteric lymph node CD103+ DCs (p<0.01) and pancreatic lymph node CD4+Foxp3+Helios- peripheral Tregs were elevated for ATG+L. lactis?PINS animals versus controls (p<0.05). Together, this suggests ATG-mediated T-cell modulation plus delivery of PINS to the gut halted autoimmunity and induced protective innate and adaptive immune responses in murine T1D and should therefore be considered for human therapy.