Functional MICA Variants Are Differentially Associated with Immune-Mediated Inflammatory Diseases

Abstract

As the principal ligand for NKG2D, MICA elicits the recruitment of subsets of T cells and NK cells in innate immunity. MICA gene variants greatly impact the functionality and expression of MICA in humans. The current study evaluated whether MICA polymorphisms distinctively influence the pathogenesis of psoriasis (PSO), rheumatoid arthritis (RA), and systemic lupus erythematosus (SLE) in Taiwanese subjects. The distributions of MICA alleles and levels of serum soluble NKG2D were compared between healthy controls and patients with PSO, RA, and SLE, respectively. The binding capacities and cell surface densities of MICA alleles were assessed by utilizing stable cell lines expressing four prominent Taiwanese MICA alleles. Our data revealed that MICA*010 was significantly associated with risks for PSO and RA (P_FDR = 1.93 × 10^-15 and 0.00112, respectively), while MICA*045 was significantly associated with predisposition to SLE (P_FDR = 0.0002). On the other hand, MICA*002 was associated with protection against RA development (P_FDR = 4.16 × 10^-6), while MICA*009 was associated with a low risk for PSO (P_FDR = 0.0058). MICA*002 exhibited the highest binding affinity for NKG2D compared to the other MICA alleles. Serum concentrations of soluble MICA were significantly elevated in SLE patients compared to healthy controls (p = 0.01). The lack of cell surface expression of the MICA*010 was caused by its entrapment in the endoplasmic reticulum. As a prevalent risk factor for PSO and RA, MICA*010 is deficient in cell surface expression and is unable to interact with NKG2D. Our study suggests that MICA alleles distinctively contribute to the pathogenesis of PSO, RA, and SLE in Taiwanese people.

Keywords: MICA alleles; NKG2D; psoriasis; rheumatoid arthritis; systemic lupus erythematosus.

Figure 1

Analysis of MICA allele expression. (A). Three independent FACS analyses of surface expression levels of MICA alleles with APC-conjugated anti-MICA (clone 159227, left row panels) or anti-MICA/B (clone 6D4, right row panels) monoclonal antibodies (mAbs). Anti-MICA mAb (clone 159227, R&D, FAB1300A-100) detected a low level of cell surface MICA*002 in small percentage of cells, while anti-MICA/B mAb (clone 6D4, BioLegend, 320907) detected MICA*002 in majority of cells. Both mAbs detected MICA*008 and MICA*019 with similar capacity, while they failed to detect the expression of MICA*010 on cell surface. (B). Western blot analysis of MICA allele proteins treated by Endo H. The total lysates from C1R cells expressing vector pCDH-copGFP, MICA*002, MICA*008, MICA*010, or MICA*019 were untreated (−Endo H) or treated with Endo H (+Endo H) before immunoblot with anti-MICA antibodies, as described in “Materials and Methods”. Endo H treatment led to a significant decrease in MICA*010’s molecular weight, while the treatment did not cause molecular weight changes in most MICA*002, MICA*008, and MICA*019 proteins. (C). Western blot analysis of MICA alleles treated with PNGase F. PNGase F treatment led to similar molecular weights for MICA*002, MICA*010, and MICA*019. MICA*008 without transmembrane segment and cytoplasmic domain had a smaller molecular weight than other MICA alleles after PNGase F treatment. Both Western blots (B,C) are representatives of two independent experiments.

Figure 2

Binding capacity of MICA alleles to recombinant human NKG2D-Fc fusion protein (rhNKG2D-Fc). (A). FACS analysis of the binding of MICA alleles to rhNKG2D-Fc. MICA on LCL stable cell lines expressing vector control, MICA*002, MICA*008, MICA*010, and MICA*019 were detected by anti-MICA/B (6D4) (left column panels). Binding capacity of cell lines expressing MICA alleles at 4 °C (middle column panels) and 37 °C (right column panels) was determined by flow cytometry using rhNKG2D-Fc and APC-conjugated anti-human IgG Fc, as described in “Materials and Methods”. Representative histograms of the cell surface staining of isotype controls (filled) and MICA (black lines) with the MFI values of MICA are shown (left column panels). The staining of the cells with APC-conjugated anti-human IgG Fc was used as negative control (filled). The binding of human rhNKG2D-Fc fusion protein to the cells at 4 °C (middle column panels) and 37 °C (right column panels) was detected by APC-conjugated anti-human IgG Fc (black lines). The MFI (mean fluorescent intensity) values of rhNKG2D-Fc surface staining are indicated. (B). Bar graph showing the average MFI of three independent experiments. The MFI values of MICA staining are shown with dark grey bars indicating the expression levels of different surface MICA alleles; MFI values of rhNKG2D-Fc staining are shown with light grey bars representing the binding ability of NKG2D to the respective surface MICA. MICA*008 expression level on cell surface was highest among MICA alleles. MICA*019 expression level was higher than MICA*002. MICA*002 had highest binding capacity among MICA variants. Cells expressing MICA*010 did not express surface MICA and had background staining by rhNKG2D-Fc. MICA*008 and MICA*019 had similar binding capacity at 4 °C and 37 °C.

Figure 3

Analyses of soluble MICA (sMICA) in psoriasis, RA, and SLE patients. (A). The concentrations of sMICA were significantly higher (p = 0.01) in SLE patients (N = 139, 46.66 ± 12.15 pg/mL) as compared to healthy controls (N = 92, 6.224 ± 5.417 pg/mL). (B). The concentrations of sMICA were not significantly different between healthy controls and RA patients (N = 80, 18.42 ± 10.55 pg/mL). (C). The concentrations of sMICA were not significantly different between healthy controls and psoriasis patients (N = 63, 11.08 ± 5.461 pg/mL). *, p < 0.05.