Transfer of PBMC From SSc Patients Induces Autoantibodies and Systemic Inflammation in Rag2-/-/IL2rg-/- Mice

Abstract

Objective: The contribution of sustained autologous autoantibody production by B cells to the pathogenesis of systemic sclerosis (SSc) and granulomatosis with polyangiitis (GPA) is not fully understood. To investigate this, a humanized mouse model was generated by transferring patient-derived peripheral blood mononuclear cells (PBMC) into immunocompromised mice.

Methods: PBMC derived from patients with SSc and GPA as well as healthy controls (HD) were isolated, characterized by flow cytometry, and infused into Rag2^-/-/IL2rg^-/- mice. In addition, PBMC from SSc patients treated with rituximab were transferred into mice. Twelve weeks later, human autoantibodies were determined in blood of the recipient mice and affected tissues were analyzed for pathological changes by histology and immunohistochemistry.

Results: Mice engrafted with PBMC derived from SSc patients developed autoantibodies such as antinuclear antibodies (ANA) mimicking the pattern of the respective donors. Moreover, cellular infiltrates dominated by B cells were observed in lung, kidney and muscles of the recipient mice. By contrast, PBMC derived from HD or GPA patients survived in recipient mice after transfer, but neither human autoantibodies nor inflammatory infiltrates in tissues were detected. Furthermore, these pathological changes were absent in mice transferred with PBMC from rituximab-treated SSc patients.

Conclusion: This humanized mouse model is indicative for cross-reactivity of human lymphocytes to murine autoantigens and argues for a pivotal role of B cells as well as of sustained autoimmunity in the pathogenesis of SSc. It provides a powerful tool to study interstitial lung disease and so far, under-recognized disease manifestations such as myositis and interstitial nephritis.

Keywords: B cells; autoantibodies; autoimmune diseases; granulomatosis with polyangiitis; humanized mouse model; peripheral blood mononuclear cells; systemic inflammation; systemic sclerosis.

Figure 1

Composition and location of human PBMC in recipient mice. (A) Relative amounts of various lymphocyte subsets in PBMC of patients and healthy donors. The percentages of CD4+, CD8+ and CD20+ cells were detected by FACS after isolation of PBMC from healthy donors (HD, n = 5), SSc patients (n = 4), and patients with GPA (n = 3). (B) Relative amounts of human leukocytes detected in peripheral blood of recipient mice. Human leukocytes were quantified in murine blood 4 weeks (W4) and 12 weeks (W12) after transfer by CD45 expression using flow cytometry. The levels of human leukocytes were defined as number of human leukocytes divided by the respective total leukocyte number in murine blood (HD: n = 8, SSc: n = 5, GPA: n = 5) and expressed as percentages. (C) Representative micrographs of H&E stainings (bar = 500µm) and immunohistochemical stainings of human CD4⁺ T cells, human CD20⁺ B cells and human CD138⁺ cells (bar = 100 µm) in splenic sections of mice receiving PBMC from HD, SSc and GPA patients. The splenic samples were collected 12 weeks after transfer of human PBMC and prepared for histological evaluation. All data are presented as mean ± SD. Statistical analyses were performed using ANOVA (*p < 0.05, **p < 0.01).

Figure 2

Survival and functionality of human PBMC in recipient mice. (A) Production of total human IgG in sera of recipient mice. The amount of total human IgG was quantified in a sandwich ELISA using human IgG-specific antibodies for capture and detection. (B) ANA production in sera of recipient mice. The ANA binding patterns of the mice were detected using immunofluorescence staining on human HEp-2 cells. (C) Levels of human IgG autoantibodies against AT1R (C) and ETAR (D) in sera of recipient mice. The autoantibodies were detected by ELISA coated with membrane extracts from CHO cells overexpressing hAT1R or hETAR. All data are presented as mean ± SD and statistical analysis was performed using ANOVA (*p < 0.05, **p < 0.01).

Figure 3

Systemic inflammation in PBMC-transferred mice. Inflammatoin in lung (A), kidney (B), and muscle (C) was kidney and muscle was evaluated by H&E staining in mice transferred with PBMC from healthy donors (HD), SSc patients and GPA patients. Representative histologic images are shown (bar = 100µm) in the left panel, while the right panel shows the quantitative score of inflammation evaluated in a blinded manner based on the ratio of infiltrated area to total area of the organ. The data are presented as mean ± SD and statistical analysis was performed using ANOVA (*p < 0.05, **p < 0.01).

Figure 4

Cellular infiltrates in mice transferred with SSc-PBMC are dominated by human B cells. Human lymphocytes in lung, kidney and muscle of recipient mice were examined by immunohistochemistry staining with anti-hCD4 and anti-hCD20 antibodies. Representative micrographs of the staining are shown (bar = 100µm).

Figure 5

PBMC derived from Rituximab-treated SSc patients do not induce disease symptoms in mice. (A) Relative amounts of CD20⁺ cells in PBMC derived from untreated and Rituximab-treated SSc patients. Percentages of CD20⁺ cells in PBMC were detected by FACS after the PBMC isolation from blood samples of SSc patients (SSc) and SSc patients treated with Rituximab (SSc_Rituximab). (B) Relative amounts of human leukocytes in peripheral blood of recipient mice. (C) Representative micrographs of immunohistochemistry stainings of human CD4⁺ T cells, CD20⁺ B cells and CD138⁺ cells (bar = 100 µm) of spleens of mice receiving PBMC from SSc patients treated with Rituximab. (D) Production of total human IgG in sera of recipient mice. Levels of human IgG autoantibodies against AT1R (E) and ETAR (F) were determined in sera of recipient mice. (G) Inflammation of lung, kidney and muscle of recipient mice. Representative histologic images are shown (bar = 100µm) in the left panel, while the right panel shows the quantitative scores of inflammation evaluated in a blinded manner based on the ratio of infiltrated areas to total areas of the organ. All data are presented as mean ± SD. Statistical analyses were performed using ANOVA (*p < 0.05, **p < 0.01).