Antibodies to Enteroviruses in Cerebrospinal Fluid of Patients with Acute Flaccid Myelitis

Abstract

Acute flaccid myelitis (AFM) has caused motor paralysis in >560 children in the United States since 2014. The temporal association of enterovirus (EV) outbreaks with increases in AFM cases and reports of fever, respiratory, or gastrointestinal illness prior to AFM in >90% of cases suggest a role for infectious agents. Cerebrospinal fluid (CSF) from 14 AFM and 5 non-AFM patients with central nervous system (CNS) diseases in 2018 were investigated by viral-capture high-throughput sequencing (VirCapSeq-VERT system). These CSF and serum samples, as well as multiple controls, were tested for antibodies to human EVs using peptide microarrays. EV RNA was confirmed in CSF from only 1 adult AFM case and 1 non-AFM case. In contrast, antibodies to EV peptides were present in CSF of 11 of 14 AFM patients (79%), significantly higher than controls, including non-AFM patients (1/5 [20%]), children with Kawasaki disease (0/10), and adults with non-AFM CNS diseases (2/11 [18%]) (P = 0.023, 0.0001, and 0.0028, respectively). Six of 14 CSF samples (43%) and 8 of 11 sera (73%) from AFM patients were immunoreactive to an EV-D68-specific peptide, whereas the three control groups were not immunoreactive in either CSF (0/5, 0/10, and 0/11; P = 0.008, 0.0003, and 0.035, respectively) or sera (0/2, 0/8, and 0/5; P = 0.139, 0.002, and 0.009, respectively).IMPORTANCE The presence in cerebrospinal fluid of antibodies to EV peptides at higher levels than non-AFM controls supports the plausibility of a link between EV infection and AFM that warrants further investigation and has the potential to lead to strategies for diagnosis and prevention of disease.

Keywords: VirCapSeq-VERT; acute flaccid myelitis; antibodies; enterovirus; enterovirus D-68; peptide array; serology.

FIG 1

Identification of an immunoreactive peptide sequence region in VP1 protein of reference sequence entries for EV-A, EV-B, EV-C, and EV-D from the National Center for Biotechnology Information (NCBI). VP1 protein models of RCSB Protein Data Bank (RPD) accession no. 4N53, 1COV, E3J48, and 6CSG were used to annotate EV-A, -B, -C, and -D, respectively, for the beta sheets (yellow arrows) and alpha helix (purple tubes). Approximate locations of BC and DE loops are based on analyses by Liu et al. (14) and Imamura et al. (23). Conserved amino acids are highlighted by color. The EV-D68-specific peptide shared less than 70% amino acid identity to other EVs, including EV-D70 and EV-D94.

FIG 2

Immunoreactivity against VP1 conserved peptide sequences of EV-A, EV-B, EV-C, and EV-D in cerebrospinal fluid samples of patients with AFM, non-AFM controls (NAC), Kawasaki disease controls (KDC), and adults with CNS diseases (AC). All AFM and NAC specimens were from 2018, except NC-P_76.

FIG 3

Immunoreactivity against an EV-D68-specific 22-aa VP1 capsid peptide in patients with AFM, non-AFM controls (NAC), Kawasaki disease controls (KDC), and adult CNS disease controls (AC). Respective immunoreactivity intensity measured by the high-density peptide microarrays is shown in heat maps of overlapping 12-mer peptides in the 22-aa EV-D68-specific VP1. Results are shown for cerebrospinal fluid (CSF) in the upper panel and serum in the lower panel. The heat map colors indicate descending reactivity from red, to yellow, to blue. Serum samples not available are indicated in gray. All AFM and NAC specimens were from 2018, except NC-P_76.