Bioinformatic analyses hinted at augmented T helper 17 cell differentiation and cytokine response as the central mechanism of COVID‐19–associated Guillain‐Barré syndrome

Abstract Objectives Guillain‐Barré syndrome (GBS) results from autoimmune attack on the peripheral nerves, causing sensory, motor and autonomic abnormalities. Emerging evidence suggests that there might be an association between COVID‐19 and GBS. Nevertheless, the underlying pathophysiological mechanism remains unclear. Materials and Methods We performed bioinformatic analyses to delineate the potential genetic crosstalk between COVID‐19 and GBS. Results COVID‐19 and GBS were associated with a similar subset of immune/inflammation regulatory genes, including TNF, CSF2, IL2RA, IL1B, IL4, IL6 and IL10. Protein‐protein interaction network analysis revealed that the combined gene set showed an increased connectivity as compared to COVID‐19 or GBS alone, particularly the potentiated interactions with CD86, IL23A, IL27, ISG20, PTGS2, HLA‐DRB1, HLA‐DQB1 and ITGAM, and these genes are related to Th17 cell differentiation. Transcriptome analysis of peripheral blood mononuclear cells from patients with COVID‐19 and GBS further demonstrated the activation of interleukin‐17 signalling in both conditions. Conclusions Augmented Th17 cell differentiation and cytokine response was identified in both COVID‐19 and GBS. PBMC transcriptome analysis also suggested the pivotal involvement of Th17 signalling pathway. In conclusion, our data suggested aberrant Th17 cell differentiation as a possible mechanism by which COVID‐19 can increase the risk of GBS.


| INTRODUC TI ON
As of 13 January 2021, over 91 million patients have been diagnosed of coronavirus disease 2019 (COVID- 19), causing more than 1.96 million deaths worldwide. The causative agent, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), was found to mediate host cell entry by binding to the host receptors ACE2 and then propagate intracellularly, causing the symptoms of COVID-19. [1][2][3] Male gender, advanced age, cigarette smoking and the presence of comorbidities (eg hypertension, diabetes) have been associated with a heightened risk of unfavourable outcomes. 4,5 Although the exact molecular pathogenesis of COVID-19 is still being investigated, deranged immune response rather than SARS-CoV-2 itself seems to cause the worst damage in some severe cases. 6 Aside from respiratory symptoms, increasing amount of evidence suggested that COVID-19 might influence both the peripheral and central nervous systems, increasing the risks of ischaemic stroke, meningoencephalitis, acute necrotizing encephalopathy, spinal cord dysfunction and Guillain-Barré syndrome (GBS). 7,8 Radiological manifestations of COVID-19-associated neurological defects include nerve root enhancement, microhaemorrhages or infarcts.
Guillain-Barré syndrome encompasses a spectrum of autoimmune diseases of the peripheral nervous system, namely acute inflammatory demyelinating polyradiculoneuropathy (AIDP), acute motor axonal neuropathy (AMAN), acute motor-sensory axonal neuropathy (AMSAN), Miller Fisher syndrome and acute pandysautonomic neuropathy. AIDP is the most common clinical subtype of GBS (85% of cases) and is characterized pathologically by T cell-mediated activation of macrophages that invade the myelin sheaths, causing denudation of the axons (ie demyelination). 9,10 In contrast, B cell-mediated humoral immunity is strongly associated with AMAN and Miller Fisher syndrome, in which autoantibodies are produced to bind to gangliosides GM1 and GD1a at the nodes of Ranvier and/or anti-GQ1b in the paranodal region. 11,12 Depending on the type of peripheral nerve involved, patients of GBS could present with paraesthesia or pain, acute symmetric ascending motor weakness (the most severe manifestation is respiratory failure due to respiratory muscle weakness) or autonomic abnormalities (eg orthostatic hypotension, disturbances of heart rate and rhythm). 13,14 Infections with Campylobacter jejuni and cytomegalovirus are two major risk factors of GBS. 15,16 In C jejuni-associated GBS cases, molecular mimicry between the ganglioside-like structures in lipo-oligosaccharides of the bacterium and gangliosides on the axonal surface contributes to the cross-reactivity of autoantibodies. 17,18 Nevertheless, it is noteworthy that not all C jejuni-infected individuals develop GBS, suggesting that other mechanisms are at work. 19 In this regard, derangement of other components of the immune system has been suggested to play a pivotal role in the molecular pathogenesis of GBS. For instance, circulating T helper (Th) 17, Th22, and Th1 cells were elevated, 20 whereas regulatory T cells (Tregs) were reduced in patients with GBS. 21 Polymorphisms in genes encoding cytokines, such as tumour necrosis factorα and interleukin (IL)-10, were also associated with susceptibility to GBS. 22 Emerging studies have indicated that there is an epidemiological linkage between COVID-19 and autoimmune/autoinflammatory diseases, such as paediatric inflammatory multisystemic syndrome. 23 In particular, a recent systematic review supported that GBS is emerging as a disease that may appear in COVID-19-infected patients 23 that has a strong autoimmune/autoinflammatory component. Electromyographic findings of axonal injury and demyelination have been observed in some COVID-19-associated patients with GBS. 10,24,25 As the pandemic of COVID-19 continues, clinical features of COVID-19-associated GBS will continue to be clarified. However, the exact pathophysiological mechanisms underlying this disease entity remain unknown. In this study, we performed bioinformatic analyses to delineate the potential crosstalk between COVID-19 and GBS-associated genes, which can illuminate the pathogenic mechanisms of COVID-19-associated GBS to derive potential therapeutics for its treatment.

| Identification of COVID-19-and GBSassociated genes
Information of genes associated with these two disease entities (COVID-19, disease id: C000657245, version 4, N.PMIDs ≥ 4; GBS, disease id: C0018378, Gene-Disease Association Score ≥ 0.02) was retrieved from DisGeNET, a knowledge platform for disease genomics. These gene sets were subject to functional protein association networks analysis using STRING to look for missing genes in the networks to produce the final gene sets.

| Protein-protein interaction networks of COVID-19-and GBS-associated genes
Protein-protein interaction network analysis using STRING was pivotal involvement of Th17 signalling pathway. In conclusion, our data suggested aberrant Th17 cell differentiation as a possible mechanism by which COVID-19 can increase the risk of GBS.

| Transcriptome analysis
RNA sequencing data sets of peripheral blood mononuclear cells

| Function and pathway enrichment analyses
Kyoto Encyclopedia of Genes Genomes (KEGG) enrichment analysis and the false discovery rate (FDR) calculation were performed with the built-in function of STRING to infer the potential biological functions of COVID-19-and GBS-associated gene networks.

| Protein-protein interaction networks of both COVID-19-and GBS-associated genes showed enrichment in immune/inflammation-related pathways
We first defined the list of genes involved in the pathogenesis of COVID-19 and GBS using a disease-based pathway analysis. High-confident genes associated with COVID-19 (disease id: C000657245, version 4; N.PMIDs ≥ 4) and GBS (disease id: C0018378; Gene-Disease Association Score ≥ 0.02), respectively, were identified with DisGeNET. We then reconstructed the functional protein-protein association networks among these genes using STRING and identified additional connecting genes (Table 1; Figure 1A Table 2).

| Protein-protein interaction networks of differentially expressed genes in PMBCs isolated from COVID-19 and GBS patients
To consolidate our findings, we downloaded publicly available  Table 3).

| D ISCUSS I ON
SARS-CoV-2 was found to bind to ACE2 receptor through subunit of N-terminal S1, which was cleaved with TMPRSS2 to expose Cterminal S2 subunit that causes cell-virus fusion. 26 Guillain-Barré syndrome is a spectrum of diseases affecting peripheral nervous system, which commonly involve axonal and demyelinating neuropathy. AIDP is the most prevalent type of GBS.
One hallmark of AIDP pathogenesis is a significantly elevated cerebrospinal fluid protein level and is characterized by activated Tcell and antibody responses to myelin epitopes in peripheral nerves.
Although the exact pathophysiology of COVID-19-associated GBS is less understood, recent studies have shown similar clinical patterns and electrophysiological subtypes with that of classic GBS.
Different subtypes have been observed, although there was a preponderance for the demyelinating polyneuropathy subtype. 10,25,29 Furthermore, the interval between the onset of COVID-19 and GBS symptoms was similar to that of other established infections, such as C jejuni and cytomegalovirus. 24 These observations support our approach of studying classic GBS, as a surrogate to understand the pathogenesis of COVID-19-associated GBS, which occurs in a significant subset of patients. M1 macrophages were also found to aggravate EAN via boosting Th17 response by secreting exosomes. 35 In COVID-19, augmented Th17 response has been observed during the cytokine storm. 36 Lymphocytes isolated from patients with COVID-19 were also found to produce more IL-17. 37 These findings together with our bioinformatic analyses support that the augmented Th17 cell differentiation and cytokine response in COVID-19 may hasten GBS development.
A major limitation of the current study is that the identified

| CON CLUS ION
We first defined the genes involved in GBS and COVID-19, followed by delineating the protein-protein interaction networks. Augmented Th17 cell differentiation and cytokine response was identified as an important process connecting COVID-19 to GBS. PBMC transcriptome analysis also suggested the pivotal involvement of Th17 signalling pathway. In conclusion, our data suggested aberrant Th17 cell differentiation as a central mechanism by which COVID-19 heightened the risk of GBS.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
Research data will be shared upon request.