Shedding of Glycocalyx Components in Neuromyelitis Optica: Initial Predictor of Immune Attack?


 Background: Neuromyelitis optica (NMO), multiple sclerosis (MS) and autoimmune glial fibrillary acidic protein (GFAP) astrocytopathy are idiopathic inflammatory demyelinating diseases (IIDDs) that mainly present as encephalomyelitis. Heparan sulfate (HS) and hyaluronic acid (HA) are two components of glycocalyx, a carbohydrate-rich layer on the surface of blood vessels that mediates interaction with blood. Degradation of glycocalyx in IIDDs is poorly understood. Purpose: To detect the serum and cerebrospinal fluid (CSF) levels of shed HS and HA and to correlate these levels with disease severity to determine their diagnostic value. Methods: We obtained serum and CSF samples from 24 NMO patients, 15 MS patients, 10 autoimmune GFAP astrocytopathy patients, and 18 controls without non-inflammatory neurological diseases. Soluble HS and HA, and IFNγ, IL17A, and matrix metalloproteinase (MMP) 1 were detected via ELISA. Results: Serum and CSF levels of HS, HA and related cytokines but not of plasma MMP1 were significantly elevated in these diseases. Notably, HS and HA levels were positively correlated with Expanded Disability Status Scale scores. Conclusions: Our results indicate glycocalyx degradation and inflammation in NMO, MS and autoimmune GFAP astrocytopathy. Moreover, increased shedding of HS or HA may indicate a worse clinical situation. Furthermore, therapeutic strategies that protect glycocalyx may be effective in these diseases.


Introduction
Idiopathic in ammatory demyelinating diseases (IIDDs) are devastating neurological diseases that affect both neurological and psychiatric functions and lead to poor quality of life [1] . Neuromyelitis optica (NMO) and multiple sclerosis (MS) are two representative immune-mediated IIDDs [2] . Recently, a new IIDD with unclear pathogenesis and no accurate diagnostic criteria was identi ed and de ned as autoimmune GFAP astrocytopathy [3] . Typically, the pathological changes of these antibody-mediated diseases involve lymphocyte in ltration, astrocytopathy, and autoantibodies targeting receptors in the brain or spinal cord [4][5][6] .
The blood-brain barrier (BBB) has long been accepted to prevent entry of immune cells and antibodies into the CNS [7] . In IIDDs, neuro-in ammation occurs causing breakdown of the BBB [8] and the interplay between neuroin ammation and BBB dysregulation can result in serious neurological disturbance [9][10][11] . The condition of the BBB in NMO, MS and autoimmune GFAP astrocytopathy is not well described, and investigation of this issue may facilitate understanding of these diseases and indicate novel therapies. Glycocalyx, a delicate membrane-bound network attached to the luminal side of the BBB endothelium, is the rst barrier between blood and the BBB [12] . Thus, levels of glycocalyx shedding can quickly re ect the extent of BBB damage.
HS and HA are two representative glycosaminoglycans present in glycocalyx that can indicate endothelial glycocalyx degradation [12][13][14] . Under pathological conditions, such as in ammation, hypertension or edema, glycocalyx components can be rapidly fragmented and released into the CSF or bloodstream [15][16][17] . Indeed, previous studies of non-IIDDs, such as sepsis, brain edema, stroke, and trauma, have con rmed the shedding of HS and HA in the acute stage [18][19][20] . In response to pathological conditions, extracellular proteases are rapidly activated and mediate the shedding of glycocalyx components, resulting in vascular permeability barrier breakdown, mechanotransduction impairment and endothelial cell dysfunction. These changes further enhance glycocalyx degradation in a feed-forward manner [20] . To date, most studies that have assayed glycocalyx have focused on vascular diseases and cancer. Some studies, however, have linked glycocalyx components with encephalitis in NMO, experimental autoimmune encephalomyelitis, and anti-NMDA receptor encephalitis [21][22][23] , although a holistic understanding of glycocalyx degradation in IIDDs is lacking.
In the early stage of an immune response, T cells secrete pro-in ammatory cytokines that stimulate the generation and maturation of matrix metalloproteinases (MMPs), which can act on glycocalyx [24][25][26] . Thus, plasma and CSF concentrations of HS and HA can re ect glycocalyx damage and, indirectly, disruption of the BBB in the acute stage of a disease. No studies have detailed the shedding of glycosaminoglycans, particularly HS and HA, into CSF or serum in patients with NMO, MS or autoimmune GFAP astrocytopathy.
Herein, we focus on the role of glycocalyx in NMO, MS and autoimmune GFAP astrocytopathy, three antibodydependent astrocyte disorders. To assess the severity of BBB injury on the severity of these neurological disorders, and to identify potential factors affecting the integrity of BBB glycocalyx, we detected the levels of two glycocalyx molecules, HS and HA, in both CSF and serum. We then correlated these levels with CSF levels of pro-in ammatory factors IFNγ and IL17A, the sheddase, MMP1, and Expanded Disability Status Scale (EDSS)·scores. CSF and plasma concentrations of HS and HA, and levels of CSF IFNγ, IL17A, and MMP1 were signi cantly elevated in these three idiopathic in ammatory demyelinating diseases. More importantly, the levels of shed glycocalyx molecules in CSF positively affected the severity of NMO, MS and autoimmune GFAP astrocytopathy, and the degree of in ammation may aggravate the disruption of the blood-brain barrier. However, the dynamic concentrations of HS and HA and their relationships with the severity of these disorders remain unknown. Of note, CSF HS and HA may be reliable markers for the diagnosis of NMO, MS and autoimmune GFAP astrocytopathy. Furthermore, therapeutic strategies focused on preservation of glycocalyx may improve outcomes of these neurological disorders.

Patients and clinical assessments
All subjects were enrolled from the Department of Neurology of Nanfang Hospital, Southern Medical University, China, including 24 NMO patients, 15 MS patients, 10 patients with autoimmune GFAP astrocytopathy and 18 controls without in ammatory or autoimmune neurological diseases (peripheral neuropathy=10, movement disorder=6, Alzheimer's disease=2). Diagnoses of autoimmune encephalitis were con rmed by two doctors on the basis of diagnostic criteria [3,[27][28] . All controls were negative for speci c CSF and serum antibodies. EDSS scores were used to evaluate disease severity. Relevant demographic and medical data were also collected and are shown in Table 1. This study was approved by the Ethics Committee of the Nan Fang Hospital and all subjects provided informed consent.

Measurement of HS, HA and related cytokines
We obtained CSF and serum samples from all subjects within 3 days of admission and before immunotherapy therapies were commenced. All samples were centrifuged at 1,000 g for 10 min. Then the supernatant was packed into polypropylene tubes and stored at -80℃ until the detection. Take one small tube for each test to avoid repeated freezing and thawing. Enzyme-Linked Immunosorbent Assay (ELISA) kits were used to measure the concentrations of HS (ELH-CD44-1, RayBiotech, Atlanta, USA), HA (DHYAL0, R&D, Minnesota, USA), IFNγ, IL17A and MMP1 ( IFNγ: KSC4021, IL17A: BMS2017, MMP1: EHMMP1, ThermoFisher, Massachusetts, USA). All detections were performed in accordance with the manufacturer's instructions and every standard and sample were assayed in duplicate.

Statistical analyses
All statistical analyses were conducted using SPSS version 24.0 (IBM, Armonk, NY, US). Data are displayed as the mean ± SEM or the median with interquartile range according to normality test results. Kruskal-Wallis plus Dunn's test or one-way analysis of variance (ANOVA) with Tukey's post hoc analysis were performed as required. p<0.05 was taken as statistically signi cant. Graphs were plotted using GraphPad Prism 8 (GraphPad, La Jolla, CA, US). All analyses were performed in a blinded manner.

Demographic and clinical characteristics
The data from patients with NMO (n = 24), MS (n = 15), autoimmune GFAP astrocytopathy (n = 10) and 18 controls are presented in Table 1. The median EDSS score (with IQR) was 4.0 (3.13, 4.88) for the NMO group, 2.5 (2.00, 3.00) for the MS group and 2.75 (2.38, 3.50) for the autoimmune GFAP astrocytopathy group (Table 1). There were no statistically signi cant differences in sex or age among groups.
3.2 Increased HS, HA and related cytokine levels in serum and CSF in patients with NMO, MS and autoimmune GFAP astrocytopathy As shown in Table 1, the median concentrations of IFNγ and IL17A in CSF and serum of NMO, MS and autoimmune GFAP astrocytopathy patients were higher than those of the control group (All p < 0.001). For MMP1, CSF concentrations but not serum were elevated in patients with NMO, MS and autoimmune GFAP astrocytopathy compared with those of the control group(CSF: p < 0.01; Seurm: p > 0.05). To further assess the severity of BBB injury in these autoimmune encephalitis patients, we compared the levels of HS and HA among patients with NMO, MS and autoimmune GFAP astrocytopathy and controls. Concentrations of HS and HA in plasma and CSF were elevated in the autoimmune encephalitis group compared with those in the control group (serum HS: NMO p < 0.001, MS p = 0.022, GFAP p < 0.001, respectively, Fig. 1A; CSF HS: NMO p < 0.001, MS p < 0.001, GFAP p < 0.001, respectively, Fig. 1B. serum HA: NMO p < 0.001, MS p < 0.001, GFAP p < 0.001, respectively, Fig. 1C; CSF HA: NMO p < 0.001, MS p < 0.001, GFAP p = 0.057, respectively, Fig. 1D). However, no signi cant differences in levels of serum or CSF glycocalyx molecules or other cytokine parameters were found among NMO, MS and autoimmune GFAP astrocytopathy patients. To assess relationships between the levels of glycocalyx molecules in CSF and serum, correlation tests were performed in autoimmune encephalitis subgroups. Positive correlation between CSF and serum levels was only present for HS in NMO and HA in MS (NMO HS: p = 0.029, r = 0.447; MS HA: p = 0.005, r = 0.696).

Correlations between CSF and plasma HS and HA levels and EDSS scores
To assess possible links between CSF and plasma glycocalyx levels and the severity of these three disorders, we examined correlations among them (Fig. 2

Demographic and clinical characteristics
The data from patients with NMO (n = 24), MS (n = 15), autoimmune GFAP astrocytopathy (n = 10) and 18 controls are presented in Table 1. The median EDSS score (with IQR) was 4.0 (3.13, 4.88) for the NMO group, 2.5 (2.00, 3.00) for the MS group and 2.75 (2.38, 3.50) for the autoimmune GFAP astrocytopathy group (Table 1). There were no statistically signi cant differences in sex or age among groups.

Increased HS, HA and related cytokine levels in serum and CSF in patients with NMO, MS and autoimmune GFAP astrocytopathy
As shown in Table 1, the median concentrations of IFNγ and IL17A in CSF and serum of NMO, MS and autoimmune GFAP astrocytopathy patients were higher than those of the control group (All p < 0.001). For MMP1, CSF concentrations but not serum were elevated in patients with NMO, MS and autoimmune GFAP astrocytopathy compared with those of the control group(CSF: p < 0.01; Seurm: p > 0.05). To further assess the severity of BBB injury in these autoimmune encephalitis patients, we compared the levels of HS and HA among patients with NMO, MS and autoimmune GFAP astrocytopathy and controls. Concentrations of HS and HA in plasma and CSF were elevated in the autoimmune encephalitis group compared with those in the control group (serum HS: NMO p < 0.001, MS p = 0.022, GFAP p < 0.001, respectively, Fig. 1A; CSF HS: NMO p < 0.001, MS p < 0.001, GFAP p < 0.001, respectively, Fig. 1B. serum HA: NMO p < 0.001, MS p < 0.001, GFAP p < 0.001, respectively, Fig. 1C; CSF HA: NMO p < 0.001, MS p < 0.001, GFAP p = 0.057, respectively, Fig. 1D). However, no signi cant differences in levels of serum or CSF glycocalyx molecules or other cytokine parameters were found among NMO, MS and autoimmune GFAP astrocytopathy patients. To assess relationships between the levels of glycocalyx molecules in CSF and serum, correlation tests were performed in autoimmune encephalitis subgroups. Positive correlation between CSF and serum levels was only present for HS in NMO and HA in MS (NMO HS: p = 0.029, r = 0.447; MS HA: p = 0.005, r = 0.696).

Correlations between CSF and plasma HS and HA levels and EDSS scores
To assess possible links between CSF and plasma glycocalyx levels and the severity of these three disorders, we examined correlations among them (Fig. 2

Availability of data and material
All data generated or analyzed during this study are included in this published article.

Competing interests
The authors declare that they have no competing interests.  a. Data were presented as mean ± SD. SD: standard deviations.
b. Data were presented as medians (IQRs-interquartile ranges). P values were calculated from the difference between CLTs and NMO, MS or autoimmune GFAP astrocytopathy.* P < 0.05, ** P < 0.01, *** P < 0.001.   In MS, the levels of HS and HA in CSF were signi cantly correlated with EDSS scores, while no correlations were found between plasma HS or HA levels and EDSS scores. (C) In autoimmune GFAP astrocytopathy, the CSF levels of HS and HA were signi cantly correlated with EDSS scores, while no correlations were found between plasma HS or HA levels and EDSS scores. The p and r-values are indicated within each analysis.