Absence of antibodies against KIR4.1 in multiple sclerosis: A three-technique approach and systematic review

Abstract

Introduction: Antibodies targeting the inward-rectifying potassium channel KIR4.1 have been associated with multiple sclerosis (MS) but studies using diverse techniques have failed to replicate this association. The detection of these antibodies is challenging; KIR4.1 glycosylation patterns and the use of diverse technical approaches may account for the disparity of results. We aimed to replicate the association using three different approaches to overcome the technical limitations of a single technique. We also performed a systematic review to examine the association of anti-KIR4.1 antibodies with MS.

Methods: Serum samples from patients with MS (n = 108) and controls (n = 77) were tested for the presence of anti-KIR4.1 antibodies using three methods: 1) by ELISA with the low-glycosylated fraction of recombinant KIR4.1 purified from transfected HEK293 cells according to original protocols; 2) by immunocytochemistry using KIR4.1-transfected HEK293 cells; and 3) by immunocytochemistry using the KIR4.1.-transfected MO3.13 oligodendrocyte cell line. We developed a systematic review and meta-analysis of the association of anti-KIR4.1 antibodies with MS according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.

Results: We did not detect anti-KIR4.1 antibodies in the MS patients or in controls using ELISA. Neither did we detect any significant reactivity against the antigen on the cell surface using the KIR4.1-transfected HEK293 cells or the KIR4.1-transfected MO3.13 cells. We included 13 prospective controlled studies in the systematic review. Only three studies showed a positive association between anti-KIR4.1 and MS. Clinical and statistical heterogeneity between studies precluded meta-analysis of their results.

Conclusion: We found no association between anti-KIR4.1 antibody positivity and MS. Although this lack of replication may be due to technical limitations, evidence from our study and others is mounting against the role of KIR4.1 as a relevant MS autoantigen.

Fig 1. Systematic review flowchart.

Thirteen studies addressing the prevalence of anti-KIR4.1 antibodies in MS and controls were identified. These 13 studies plus our own results were used for the systematic review.

Fig 2. IgG reactivity against KIR4.1 does not differ in MS and controls by ELISA.

Low-glycosylated fraction of the KIR4.1 whole protein complex was purified and used to detect anti-KIR4.1 antibodies by ELISA. None of the 108 MS patients or disease controls reached a ΔOD signal above healthy control average ΔOD plus 5 SD. The average ΔOD signal did not show significant differences between groups (A). The average ΔOD signal from patients with MMN was significantly higher (p<0.0001, Kruskal-Wallis) than that from any other group except CIDP. The CIDP average ΔOD was higher than the average in the ALS and MG groups (p<0.0001, Kruskal-Wallis; Fig 2B).

Fig 3. Lack of identification of KIR4.1 antibodies in serum samples using ICC.

MS patients or disease controls did not show antibodies against KIR4.1 tested by ICC in KIR4.1 transfected HEK293 (A-F) or MO3.13 (G-L). Anti-KIR4.1 commercial antibody reactivity is shown in the upper row (A, D, G, J), serum IgG reactivity from an MS patient (B and H) and from a disease control (E and K) in the middle row, and merge images (C, F, I, L) in the lower row. The MS patient and disease control showing highest ΔOD were chosen to assemble this Figure.

Fig 4. Forest plot.

Thirteen studies were included in the systematic review. Three of them showed evidence of an association of KIR4.1 with MS patients compared to healthy controls (A) or all controls (B). One showed a weak association of KIR4.1 antibodies with healthy controls. None of the other reports found any association of KIR4.1 antibodies with MS, healthy or disease controls. The study by Probstel[19] was excluded from Fig 4A forest plot because it lacked comparison with healthy controls, as specified in the inclusion criteria (A). The study by Malyavantham[15] was excluded from Fig 4B forest plot because it lacked comparison with disease controls (B).