Novel Serology Testing for Sporadic Inclusion Body Myositis

Disease-Specificity and Diagnostic Utility

Megan K. Herbert; Ger J.M. Pruijn


Curr Opin Rheumatol. 2015;27(6):596-600. 

In This Article

Anti-cytosolic 5'-nucleotidase 1A Auto-antibodies: A Novel Serological Test for Sporadic Inclusion Body Myositis

To date, several different methods have been used for the detection of anti-cN-1A auto-antibodies. These include immunoprecipitation,[16,17] immunoblotting using human skeletal muscle extracts[15–17] or lysates from cN-1A-expressing human embryonic kidney 293 (HEK293) cells,[19,20] and ELISAs with a recombinant full-length cN-1A protein[20,21] or synthetic peptides derived from cN-1A.[22] Three recent studies,[20–22] used ELISA to detect anti-cN-1A antibodies in sIBM patient sera and two of these also addressed the prevalence of anti-cN-1A antibodies in other neuromuscular and autoimmune diseases.[21,22] The full-length cN-1A protein used in ELISA was either recombinant cN-1A[21] or lysate taken from ectopically cN-1A-expressing HEK293 cells. In the synthetic peptide ELISA,[22] a set of three peptides was used, which was based on the results of epitope mapping studies, with an array of overlapping peptides covering the complete cN-1A polypeptide. The sequences of these three peptides represent the three major epitopes targeted by anti-cN-1A-positive sIBM sera.[16] If optical density (OD) values for any one of the three peptides were above the determined cut-off, samples were considered anti-cN-1A-positive, and the highest of these values was used for further comparisons. As secondary antibodies, either IgG,[20] IgG, IgA and IgM-specific,[21] or a mixture of anti-human immunoglobulins (IgG, IgA, IgM)[22] were used. As with the earlier immunoblotting studies,[16,17] all of these ELISAs detected a high prevalence of anti-cN-1A antibodies in sIBM patients' sera, ranging from 37%[22] to 76%.[21] The differences might be due to the use of different antigens, the analysis of different cohorts and to the determination of cut-off values. In several studies, it was shown that the use of a canonical cut-off (average OD + 2 times SD of healthy control sera) resulted in the detection of low levels of anti-cN-1A antibodies in a relatively large number of sera from patients with other neuromuscular and autoimmune diseases. The cause for this phenomenon remains to be elucidated, but these observations suggest that it might be more appropriate to use more stringent cut-off values to discriminate anti-cN-1A-positive from negative sera. Table 1 provides a summary of the results of all studies in which the prevalence of anti-cN-1A auto-antibodies in various diseases, and detected by various methods, was addressed.

Although the synthetic peptide ELISA was based on the major epitopes identified by nine anti-cN-1A-positive sIBM sera, the possibility exists that this approach fails to detect auto-antibodies directed against other epitopes, particularly conformational epitopes. Alternatively, although less likely, linear sequences identified using the peptide ELISA might be missed using a full-length ELISA set-up, if they are inaccessible to auto-antibodies due to folding of the full-length protein. To investigate these possibilities, we recently developed an ELISA using the recombinant full-length human cN-1A protein expressed in bacteria to compare the reactivity to the full-length auto-antigen with the reactivity to the cN-1A peptides. In a subset (n = 55) of our sIBM cohort, only a moderate correlation in seropositivity between the full-length and peptide ELISAs was observed (r 2 = 0.538; Fig. 2), and of the 28 sera detected as seropositive using the peptide ELISAs, only 15 (54%) were congruently identified as seropositive using the full-length protein ELISA. This supports the idea that linear epitopes make an important contribution to the detection of anti-cN-1A auto-antibodies. Of the remaining sera, five additional sera were identified as seropositive using the full-length protein ELISA. It should be noted that sample selection in this pilot work was biased towards the selection of samples tested as seropositive using our peptide ELISAs. Therefore, additional work is required to determine the full utility of the full-length protein ELISA in conjunction with the peptide ELISA. In our opinion, it is likely that a combination of ELISAs utilizing both peptide and full-length protein targets will prove necessary to optimize the recognition of anti-cN-1A auto-antibodies in sIBM sera in future studies.

Figure 2.

Correlation between full-length and synthetic peptide-based cN-1A ELISA. Coloured symbols represent samples detected by both the full-length and peptide ELISAs (red), only by the peptide ELISAs (green) or only by the full-length ELISA (blue). Values below the cut-off in both assays are indicated in black. pMax=highest OD450 value obtained for any of the three cN-1A peptide ELISAs. cN-1A, cytosolic 5'-nucleotidase 1A.