The cytokine interleukin (IL)-18 is a crucial amplifier of natural killer (NK) cell function. IL-18 signaling is regulated by the inhibitory effects of IL-18 binding protein (IL-18BP). Using mice deficient in IL-18BP (IL-18BPKO), we investigated the impact of mismanaged IL-18 signaling on NK cells. We found an overall reduced abundance of splenic NK cells in the absence of IL-18BP. Closer examination of NK cell subsets in spleen and bone marrow using CD27 and CD11b expression revealed that immature NK cells were increased in abundance, while the mature population of NK cells was reduced. Also, NK cells were polarized to greater production of TNF-α, while dedicated IFN-γ producers were reduced. A novel subset of IL-18 receptor α− NK cells contributed to the expansion of immature NK cells in IL-18BPKO mice. Splenocytes cultured with IL-18 resulted in alterations similar to those observed in IL-18BP deficiency. NK cell changes were associated with significantly reduced levels of circulating plasma IL-18. However, IL-18BPKO mice exhibited normal weight gain and responded to LPS challenge with a >10-fold increase in IFN-γ compared to wild type. Finally, we identified that the source of splenic IL-18BP was among dendritic cells/macrophage localized to the T cell-rich regions of the spleen. Our results demonstrate that IL-18BP is required for normal NK cell abundance and function and also contributes to maintaining steady-state levels of circulating IL-18. Thus, IL-18BP appears to have functions suggestive of a carrier protein, not just an inhibitor.

Multiple environmental triggers have been proposed to explain the increased incidence of type 1 diabetes (T1D). These include viral infections, microbiome disturbances, metabolic disorders, and vitamin D deficiency. Here, we used ELISA to examine blood plasma from juvenile T1D subjects and age-matched controls for the abundance of several circulating factors relevant to these hypotheses. We screened plasma for sCD14, mannose binding lectin (MBL), lipopolysaccharide binding protein (LBP), c-reactive protein (CRP), fatty acid binding protein 2 (FABP2), human growth hormone, leptin, total adiponectin, high molecular weight (HMW) adiponectin, total IgG, total IgA, total IgM, endotoxin core antibodies (EndoCAbs), 25(OH) vitamin D, vitamin D binding protein, IL-7, IL-10, IFN-γ, TNF-α, IL-17A, IL-18, and IL-18BPa. Subjects also were tested for prevalence of antibodies targeting adenovirus, parainfluenza 1/2/3, Coxsackievirus, cytomegalovirus, Epstein-Barr virus viral capsid antigen (EBV VCA), herpes simplex virus 1, and Saccharomyces cerevisiae. Finally, all subjects were screened for presence and abundance of autoantibodies targeting islet cell cytoplasmic proteins (ICA), glutamate decarboxylase 2 (GAD65), zinc transporter 8 (ZNT8), insulinoma antigen 2 (IA-2), tissue transglutaminase, and thyroid peroxidase, while β cell function was gauged by measuring c-peptide levels. We observed few differences between control and T1D subjects. Of these, we found elevated sCD14, IL-18BPa, and FABP2, and reduced total IgM. Female T1D subjects were notably elevated in CRP levels compared to control, while males were similar. T1D subjects also had significantly lower prevalence of EBV VCA antibodies compared to control. Lastly, we observed that c-peptide levels were significantly correlated with leptin levels among controls, but this relationship was not significant among T1D subjects. Alternatively, adiponectin levels were significantly correlated with c-peptide levels among T1D subjects, while controls showed no relationship between these two factors. Among T1D subjects, the highest c-peptide levels were associated with the lowest adiponectin levels, an indication of insulin resistance. In total, from our examination we found limited data that strongly support Harms et al. Confirmation and Identification of Biomarkers any of the hypotheses investigated. Rather, we observed an indication of unexplained monocyte/macrophage activation in T1D subjects judging from elevated levels of sCD14 and IL-18BPa. These observations were partnered with unique associations between adipokines and c-peptide levels among T1D subjects.

We analyzed T cell subsets from cryopreserved PBMC obtained from the TrialNet Pathway to Prevention archives. We compared subjects who had previously seroconverted for one or more autoantibodies with non-seroconverted, autoantibody negative individuals. We observed a reduced frequency of MAIT cells among seroconverted subjects. Seroconverted subjects also possessed decreased frequencies of CCR4-expressing CD4 T cells, including a regulatory-like subset. Interestingly, we found an elevation of CD57+, CD28–, CD127–, CD27– CD8 T cells (SLEC) among seroconverted subjects that was most pronounced among those that progressed to disease. The frequency of these SLEC was strongly correlated with CMV IgG abundance among seroconverted subjects, associated with IA-2 levels, and most elevated among CMV+ seroconverted subjects who progressed to disease. Combined, our data indicate discrete, yet profound T cell alterations are associated with islet autoimmunity among at-risk subjects.

Mucosal-associated invariant T (MAIT) cells have been implicated in various forms of autoimmunity, including type 1 diabetes (T1D). Here, we tested the hypothesis that CD8 and double negative (DN) MAIT cell frequencies were altered among diagnosed T1D subjects compared to controls. To do this, we analyzed cryopreserved peripheral blood mononuclear cells (PBMCs) from age-matched T1D and control children using flow cytometry. We observed that CD8 and DN MAIT cell frequencies were similarly abundant between the two groups. We tested for associations between MAIT cell frequency and T1D-associated parameters, which could reveal a pathogenic role for MAIT cells in the absence of changes in frequency. We found no significant associations between CD8 and DN MAIT cell frequency and levels of islet cell autoantibodies (ICA), glutamate decarboxylase 65 (GAD65) autoantibodies, zinc transporter 8 (ZNT8) autoantibodies, and insulinoma antigen 2 (IA-2) autoantibodies. Furthermore, CD8 and DN MAIT cell frequencies were not significantly associated with time since diagnosis, c-peptide levels, HbA1c, and BMI. As we have examined this cohort for multiple soluble factors previously, we tested for associations between relevant factors and MAIT cell frequency. These could help to explain the broad range of MAIT frequencies we observed and/or indicate disease-associated processes. Although we found nothing disease-specific, we observed that levels of IL-7, IL-18, 25 (OH) vitamin D, and the ratio of vitamin D binding protein to 25 (OH) vitamin D were all associated with MAIT cell frequency. Finally, previous cytomegalovirus infection was associated with reduced CD8 and DN MAIT cells. From this evaluation, we found no connections between CD8 and DN MAIT cells and children with T1D. However, we did observe several intrinsic and extrinsic factors that could influence peripheral MAIT cell abundance among all children. These factors may be worth consideration in future experimental design.