Longitudinal assessment of reactivity and affinity profile of anti-Jo1 autoantibodies to distinct HisRS domains and a splice variant in a cohort of patients with myositis and anti-synthetase syndrome

Figures

• 

  Fig. 1. Anti-Jo1 autoantibodies display high reactivity against HisRS full-length/variant/domains at IIM/ASSD diagnosis. A Schematic figure of HisRS versions used in the experiments. B The reactivity towards HisRS-FL, WHEP, CD, ABD, and SV as conformational epitopes was measured by ELISA in IgG purified from serum of 19 anti-Jo1+, 25 anti-Jo1−, and 24 healthy controls (HC). High reactivity (OD 450 nm) corresponds to strong blue color. C Anti-HisRS reactivity (ng/mL concentration) of anti-Jo1+ total IgG against HisRS-FL, WHEP, CD, ABD, and SV. Antibody titers were calculated using a standard curve (Supplementary Fig. 3B) and titers were measured in the linear range between 5 and 100 ng/mL. Median values for each antigen are indicated by the red line. D Anti-Jo1 reactivity against denatured HisRS antigens was addressed by WB in IgG purified from serum of 12 representative anti-Jo1+, 1 anti-Jo1−, and 1 HC. Strong band intensity denotes higher anti-Jo1 reactivity. Arrows indicate the molecular weight of the different HisRS variants (Supplementary Table 2, Supplementary Fig. 1). Some patients also showed an additional band, corresponding to the HisRS dimer. Friedman’s test corrected for…

• 

  Fig. 2. Reactivity of serum and BALF-derived anti-Jo1 antibodies. A, C Anti-Jo1 IgG and B, D IgA reactivity in BALF and paired serum were measured by ELISA in 6 anti-Jo1+ patients. Autoantibody levels were normalized to total values of IgG and IgA (Y-axis). Friedman’s tests corrected for multiple comparisons by Dunn’s test was applied. p < 0.05 was assumed as significantly different

• 

  Fig. 3. Reactivity of anti-Jo1 autoantibodies towards HisRS variant and domains decreases over time but remains high against HisRS-FL. A Reactivity against HisRS-FL, HisRS splice variant (SV), and HisRS domains (WHEP, CD, and ABD) displayed by total IgG purified from the first available anti-Jo1+ sera close to diagnosis (T = −0.25–0 years), T = 0.1–1, T = 1.1–2, and T = 2.1–3 years after diagnosis. Additional graphs displaying anti-Jo1 reactivity against HisRS-FL, variant, and domains are displayed in Supplementary Fig. 7 upper panel. B Anti-Jo1 reactivity of 6 anti-Jo1+ patients (P2, P6, P7, P11, P16, P17) displayed by total IgG purified from sera collected longitudinally. The Y-axis represents anti-Jo1 antibody levels against HisRS, measured in the total IgG fraction isolated from anti-Jo1+ IIM/ASSD sera. The X-axis represents disease duration in years. Gray italic sentences provide information on interstitial lung disease outcome during follow-up. Improvement, stabilization, or worsening of ILD was based on the comparison of spirometry test results (5–10% absolute increase or decline of predicted FVC and/or 10–15% increase or decline of DLCO) and of HRCT reports made by experienced…

• 

  Fig. 4. IIM/ASSD patients diagnosed with ILD and arthritis and less skin involvement harbor more reactive anti-Jo1 autoantibodies. A–D Percentage of IIM/ASSD patients distributed according to anti-HisRS full-length (HisRS-FL) reactivity levels, clinical diagnosis, and clinical manifestations. The anti-HisRS-FL reactivity displayed was measured in total anti-Jo1+ IgG purified from serum. Numbers on top of the bars represent the percentage of patients in the group. Anti-HisRS-FL−, anti-Jo1 IIM/ASSD-negative patients; anti-HisRS-FL+ +, anti-Jo1 IIM/ASSD-positive patients with low to moderate reactivity (0.5–100 ng/mL); anti-HisRS-FL+ ++, anti-Jo1 IIM/ASSD-positive patients with high HisRS reactivity (> 100 ng/mL). Antibody titers were calculated from OD (450 nm) using the standard curve in Supplementary Fig. 3B. IIM, idiopathic inflammatory myopathies; ILD, interstitial lung disease; ASSD, anti-synthetase syndrome. Statistical differences among groups are displayed in Supplementary Table 3

• 

  Fig. 5. Multivariate data analysis shows a strong correlation between anti-Jo1 autoantibody reactivity profile, ILD and arthritis, and poor pulmonary function. A Principal component analysis (PCA) score plot of the baseline data obtained from the patients (cohort 1, n = 44) which demographics are listed in Table 1. B Shared and Unique Structures (SUS) plot obtained from two OPLS-DA models: (1) distinguishing the patients according to anti-Jo1 status (predictive component: R2 = 0.73, Q2 = 0.59) and (2) distinguishing the patients according to ILD status (predictive component: R2 = 0.74, Q2 = 0.60). The data processed is the same as in the PCA plot in cohort 1 (A). The factors that correlated most strongly with anti-Jo1+/ILD+ as well as anti-Jo1−/ILD− are highlighted. C PCA score plot of the baseline data obtained from the patients (cohort 2, n=10) which demographics are listed in Supplementary Table 1. D Shared and Unique Structures (SUS) plot obtained from two OPLS-DA models: (1) distinguishing the patients according to Jo1 status (predictive component: R2=0.86, Q2=0.61) and (2) distinguishing the patients according to ILD status (predictive component: R2=0.76, Q2=0.43). The data…

• 

  Fig. 6. Affinity binding profiles against HisRS-FL for IgG purified from sera at first available sample close to diagnosis measured by SPR in single cycle kinetics mode. Sensorgrams showing the obtained signal in response units (RU, y-axis) over time (seconds, s, x-axis) from four representative patients (P1, P5, P9, and P17). A Three patients did not show any binding in SPR, as indicated by a flat line (same observation for P4 and P13, data not shown). B P9 is an example of a patient having a biphasic off-rate (Avekd). This sort of heterogeneity was also observed for 6 other patients (P3, P7, P11, P12, P16, and P18). C, D A slower and more homogenous off-rate (Avekd < 0.0005 s-1) was observed for P5 and P17 (as well as P2, P8, P14, P15, P19), representing the other group (n=7) of patients. The biphasic binding profiles observed in approximately half the patients (n=7) could be explained by an avidity effect caused by the dimeric form of the antigen but can also be the result of the polyclonal nature of the patient sample and, thus, containing antibodies of different off-rates. Red lines represent the measured data values and dashed black lines represent the fit of the curve using the…