85–23, revised 1985) and the national laws on Protection of Animals were followed. A total of 109 female NOD mice were analysed. First, 62 female NOD mice were litter-matched and randomized after weaning into four groups (groups A1–D1; Fig. S1). As

a control group, female NOD mice in group A1 (n = 16) were not mated. In the remaining groups, female NOD were mated at age 10 weeks to male NOD mice (group B1, n = 15) representing MHC identical mating, male CByB6F1/J mice (group C1, n = 16) representing MHC haploidentical mating or male C57BL/6J mice (group D1, n = 15), representing fully MHC mismatched mating. For pairings, two male mice were used in each group. To analyse the effect of later gestation, a second set of 47 female NOD mice were litter-matched and randomized after weaning into three groups: control unmated females (n = 16, group A2); females mated at

age 13 weeks to three male Ulixertinib NOD mice (n = 16, group B2); and females mated at age 13 weeks to three male CByB6F1/J mice (n = 15, group C2). For the mating, two female and one male mouse were housed together in one cage for a median time of 14 days [interquartile range (IQR), 14–17 days]. Subsequently, 45 of 46 females mated at 10 weeks of age and 29 of 31 females mated at 13 weeks of age delivered altogether 610 pups. The number of pups per litter ranged between four and 13 (median, 9; IQR, 7–10), and the offspring numbers were distributed equally between the different mating groups (Fig. S1). All pups remained with their dam for the weaning period of median 21 days (IQR, 21–23 days). All female Adriamycin order NOD mice were followed to overt diabetes or until the age of 28 weeks. Inositol monophosphatase 1 Urine glucose levels were measured twice weekly using urine glucose sticks (Diastix, Bayer HealthCare LLC, Mishawaka, IN, USA), beginning at 10 weeks of age. The diagnosis of diabetes was defined as two consecutive urine glucose values > 5·5 mmol/l and blood glucose levels > 13·9 mmol/l (Glucometer Elite,

Bayer Diagnostics GmbH, Munich, Germany). Venous blood was obtained at age 10 weeks (prior to the mating) and 16 weeks (after weaning), and diabetes onset or 30 weeks for the measurement of insulin autoantibodies. In group C1, splenocytes from two diabetic mice at diabetes onset and two non-diabetic mice at the end of observation were collected to look for lymphocyte chimerism. Antibodies to insulin were detected using a radiobinding assay, as described previously [14]. All measurements were performed on coded samples that were operator-blinded. The upper limit of normal was determined from the 99th centile values obtained in sera from BALB/c and C57BL/6 female mice. The assay is represented as laboratory B in the animal models of Diabetes Workshop [15]. In order to analyse if cells with paternal genome alleles migrated during gestation from the fetus to the dam and persisted, staining and flow cytometry for MHC class I molecules was performed on the collected splenocytes.