Review 3: "Prevalent and Persistent New-Onset Autoantibodies in Mild to Severe COVID-19"

RR:C19 Evidence Scale rating by reviewer:

• Potentially informative. The main claims made are not strongly justified by the methods and data, but may yield some insight. The results and conclusions of the study may resemble those from the hypothetical ideal study, but there is substantial room for doubt. Decision-makers should consider this evidence only with a thorough understanding of its weaknesses, alongside other evidence and theory. Decision-makers should not consider this actionable, unless the weaknesses are clearly understood and there is other theory and evidence to further support it.

***************************************

Review: The authors of this manuscript have longitudinally investigated new-onset autoantibodies in healthcare workers and hospitalized COVID-19 patients. This study convincingly shows elevation of autoantibodies against a wide range of antigens following SARS-CoV-2 infection. Such polyreactivity after COVID-19 has been well documented and this study is a methodologically solid addition to the literature. 

However, it is unclear whether these autoantibodies are truly 'functional' in the sense that they can impair the function of the targeted proteins. Other studies have found autoantibodies that do not impair function of their target epitopes (for example ACE2). As the linear epitopes for some of these autoantibodies have been worked out, it might be possible to answer this question if functional domains are known. 

Although the authors demonstrate an association between some of these autoantibodies and neuropsychiatric symptoms after COVID-19, the link remains tentative mechanistically given that some targets like CALU are not known to be expressed in neural tissue and others like SNURF are of unknown function. Similarly, the clinical relevance of anti-SNURF IgG after vaccination is unclear. It is also unknown to what extent such polyreactivity is unique to COVID-19; other viral infections such as cytomegalovirus, hepatitis E virus etc are well-associated with autoreactivity. 

Molecular mimicry is an attractive hypothesis for such autoantibodies, but it is doubtful whether the short identical fragment would be sufficient to elicit cross-reactive autoantibodies without structural confirmation. Conserved amino acid sequence does not equal similar binding epitopes. It is not uncommon to have short fragments of homology between microbial and host peptides, but this usually doesn't result in breakdown of tolerance (especially to intracellular antigens). Significantly, the authors confirmed that there was no association between these autoantibodies and the strength of the SARS-CoV-2 IgG response. One might expect a clear correlation in case of true molecular mimicry. Further work is needed to understand mechanisms of such autoimmunity.