Failure to control inflammation as a driving force of symptoms and phenotype in the NSG mouse model of ulcerative colitis

Background:

Specific autoantibody signatures are characteristics of chronic inflammatory disease such as lupus erythematosus or multiples sclerosis. However, no comprehensive analysis of autoantibodies in patients with ulcerative colitis (UC) has been conducted.

Methods:

To analyze the spectrum of autoantibodies, serum-IgG from UC patients and non-UC donors were screened using a human protein microarray. To elucidate the role of autoantigen in vitro, peripheral blood mononuclear cells (PBMC) were incubated with the purified antigen CD99 and TSLP and subjected to flow cytometric analysis. To elucidate the role of CD99 in vivo, we used a humanized mouse model (NSG-UC). To validate the results sirolimus and anti CCR4 antibodies were tesed in this model.

Results:

We discovered 697 differentially-reactive antigens, most expressed on immune cells. CD99 emerged as a biomarker to discriminate between non-UC and UC patients. The majority (81%) of UC patients exhibited anti-CD99 autoantibodies, while the majority (82%) of non-UC patients tested negative (p = 1e-04). The levels of anti-CD99 autoantibodies did not correlate with age, gender, duration of disease, SCCAI (Simple Clincal Activity Score) or responsiveness to medication, but was inversely correlated with the serum level of TSLP (Thymic stromal lymphopoietin) and TSLPR expressing macrophages which are modulators of immune responses. In vitro, incubation of human peripheral blood mononuclear cells with purified CD99 increased frequencies of CD4⁺CCR4⁺ T cells, TSLPR⁺CD11b⁺ macrophages and CD14⁺ monocytes, and promoted differentiation of Th2 cells. X Addition of TSLP to these cultures reversed the effect of CD99. In vivo, rectal and intraperitoneal challenge with CD99 aggravated disease symptoms and pathological phenotype in a humanized mouse model of UC. Treatment of these mice with sirolimus decreased clinical and histological scores and IFNγ mRNA levels, and increased frequencies of CCR4⁺CD4⁺ T-cells. In contrast, treatment with anti-CCR4 antibody depleted CCR4⁺CD4⁺ T-cells and aggravated inflammation.

Conclusion:

These results suggest that an altered balance between CD99 and TSLP drive disease in the humanized mouse model of UC and human UC. This balance can be restored by promoting regulatory T cells. Furthermore, this animal model serves as a tool to examine the role of autoantigens and autoantibodies in the pathogenesis of UC.