Conformation-stabilizing ELISA and cell-based assays reveal patient subgroups targeting three different epitopes of AGO1 antibodies

Abstract

Autoantibodies (Abs) are biomarkers for many disease conditions and are increasingly used to facilitate diagnosis and treatment decisions. To guarantee high sensitivity and specificity, the choice of their detection method is crucial. Via cell-based assays, we recently found 21 patients with neurological diseases positive for antibodies against argonaute (AGO), 10 of which having a neuropathy (NP). Here, we established a simple and conformation-sensitive ELISA with the aim to distinguish between AGO1 Abs against conformational epitopes and non-conformational epitopes and to reveal further characteristics of AGO1 antibodies in NP and autoimmune disease (AID). In a retrospective multicenter case/control and observational study, we tested 434 patients with NP, 274 disease controls with AID, and 116 healthy controls (HC) for AGO1 Abs via conformation-stabilizing ELISA. Seropositive patients were also tested for conformation-specificity via comparative denaturing/stabilizing ELISA (CODES-ELISA), CBA positivity, AGO1 titers and IgG subclasses, and AGO2 reactivity. These parameters were statistically compared among different epitope-specific patient groups. We found Abs in 44 patients, including 28/434 (6.5%) NP, 16/274 (5.8%) AID, and 0/116 (0%) HC. Serum reactivity was consistently higher for AGO1 than AGO2. Globally among the 44 AGO1 Abs-positive patients, 42 were also tested in CBA for AGO1 Abs positivity and 15 (35.7%) were positive. Furthermore, 43 were tested for conformation-specificity and 32 (74.4%) bound a conformational epitope. Among the subgroups of highly positive patients (ELISA z-score >14) with sera binding conformational epitopes (n=23), 14 patient sera were also CBA positive and 9 bound a second conformational but CBA-inaccessible epitope. A third, non-conformational epitope was bound by 11/43 (15.6%). Among the epitope-specific patient subgroups, we found significant differences regarding the Abs titers, IgG subclass, and AGO2 reactivity. When comparing AGO1 Abs-positive NP versus AID patients, we found the conformation-specific and CBA inaccessible epitope significantly more frequently in AID patients. We conclude that 1) conformational ELISA was more sensitive than CBA in detecting AGO1 Abs, 2) serum reactivity is higher for AGO1 than for AGO2 at least for NP patients, 3) AGO1 Abs might be a marker-of-interest in 6.5% of NP patients, 4) distinguishing epitopes might help finding different patient subgroups.

Keywords: Anti-Su antibodies; argonaute antibodies; autoimmune neuropathy; conformation-stabilizing ELISA; ganglionopathie; sensory neuronopathy.

Figure 1

Comparison of AGO1 and AGO2 Abs and establishment of a denaturation assay to assess conformation-specificity of AGO1 Abs. (A) Correlation analysis between ELISA ΔODs for AGO2 and AGO1 Abs. (B) Correlation analysis between normalized ELISA ΔODs (z-scores) for AGO2 and AGO1 Abs. (C) Scheme of CODES-ELISA to distinguish antibodies with conformational and linear epitopes. Size and line width of schematic antibodies represent their binding reactivity. (D) ELISA denaturation assay using sera from 19 AGO Abs-positive patients, 16 healthy controls, and 1 commercial anti-AGO1 antibody. For each of the three conditions a respective positivity threshold was defined based on the ODs of the 16 healthy controls under the respective conditions (horizontal bars = arithmetic mean plus 3 standard deviations).

Figure 2

Frequencies in different disease groups and characteristics of AGO1 Abs. (A) ELISA ΔODs plotted for different disease groups comprising 824 subjects. Each data point represents one subject. Each green data point (44 in total) above the horizontal line at ΔOD = 0.386 (i.e., z-score = 4) represents a AGO1 Abs-positive subject. NP: neuropathies, AID: autoimmune disease, HC: healthy controls. (B) Correlation analysis between normalized ELISA ΔODs (z-scores) for AGO2 and AGO1 Abs. Dotted line: z-score AGO1 = z-score AGO2; Dashed line: linear correlation curve. Numbers above and below the dashed curve represent the proportion of AGO1/AGO2 Abs-positive NP sera. *** p ≤ 0.001. (C) ELISA denaturation assay using sera from 43 AGO Abs-positive patients (27 with NP and 16 with AID). Green curves: patients with conformation-specific reactivity pattern; red curves: patients with non-conformation-specific reactivity pattern.

Figure 3

Proportions, ELISA reactivities and disease characteristics of sera with conformational or CBA-positive AGO1 Abs. (A) Bar charts representing proportion of subjects with conformation-specific AGO1 Abs among all AGO1-ELISA-positive subjects, on the right side split up in strongly (++) and moderately (+) ELISA-positive subjects together with another bar chart showing the proportion of AGO1-CBA-positive subjects among those with conformation-specific AGO1 Abs. (B) ELISA-reactivity of all AGO1-ELISA-positive subjects plotted depending on conformation specificity. For moderately positive (+) subjects, colors of single spots represent disease group. (C) ELISA-reactivity of all strongly (++) AGO1-ELISA-positive subjects with conformation-specific AGO1 Abs plotted depending on CBA reactivity. Colors of single spots represent the disease group of the corresponding subject. *** p < 0.001; ns = not significant; Kruskal-Wallis test.