CIRCULATING HEPARAN SULFATE PROFILES IN PEDIATRIC ACUTE RESPIRATORY DISTRESS SYNDROME

Abstract

Introduction: Sepsis-induced degradation of endothelial glycocalyx heparan sulfate (HS) contributes to the pulmonary microvascular endothelial injury characteristic of acute respiratory distress syndrome (ARDS) pathogenesis. Our objectives were to (1) examine relationships between plasma indices of HS degradation and protein biomarkers of endothelial injury and (2) identify patient subgroups characterized by distinct profiles of HS degradation in children with ARDS. Methods: We analyzed prospectively collected plasma (2018-2020) from a cohort of invasively mechanically ventilated children (aged >1 month to <18 years) with ARDS. Mass spectrometry characterized and quantified patterns of HS disaccharide sulfation. Protein biomarkers reflective of endothelial injury (e.g., angiopoietin-2, vascular cell adhesion molecule-1, soluble thrombomodulin) were measured with a multiplex immunoassay. Pearson correlation coefficients were used to construct a biomarker correlation network. Centrality metrics detected influential biomarkers (i.e., network hubs). K-means clustering identified unique patient subgroups based on HS disaccharide profiles. Results: We evaluated 36 patients with pediatric ARDS. HS disaccharide sulfation patterns, 6S, NS, and NS2S, positively correlated with all biomarkers of endothelial injury (all P < 0.05) and were classified as network hubs. We identified three patient subgroups, with cluster 3 (n = 5) demonstrating elevated levels of 6S and N-sulfated HS disaccharides. In cluster 3, 60% of children were female and nonpulmonary sepsis accounted for 60% of cases. Relative to cluster 1 (n = 12), cluster 3 was associated with higher oxygen saturation index (P = 0.029) and fewer 28-day ventilator-free days (P = 0.016). Conclusions: Circulating highly sulfated HS fragments may represent emerging mechanistic biomarkers of endothelial injury and disease severity in pediatric ARDS.

FIG. 1.. Heparan sulfate disaccharide sulfation patterns.

HS is anchored to the endothelial cell surface by transmembrane (i.e., syndecan-1) proteoglycans. HS GAGs are linear polysaccharides composed of repeating hexuronic acid (glucuronic acid or its epimer iduronic acid)-glucosamine disaccharide units. These disaccharide units are characterized by differential sulfation at the 2-O position (2S) of the iduronic acid residue and/or the N-acetyl (NS) and 6-O (6S) positions of the glucosamine residue. The patterns of sulfation along HS polysaccharides allow for specific electrostatic interactions with positively charged protein residues. Note: Not depicted is the 3-O-sulfo group on the glucosamine residue, thought to be rare in HS. Future advances in glycomics techniques may allow better appreciation of this sulfation pattern. Adapted from Oshima et al. (20). GAGs, glycosaminoglycans; HS, heparan sulfate.

FIG. 2.

A, Heatmap of Pearson correlations. Pearson correlation coefficients were used to construct the biomarker correlation network using a FDR of P < 0.05. B, Network analysis topology. Nodes (biomarkers) were connected by edges, where the thickness of the edges represented the strength of correlation between nodes. Centrality varied according to node size. Nodes with the largest size (sdc1; NS; 6S; NS2S) indicated network hubs (or influential biomarkers with the highest centrality metrics, see Supplemental Table 2, http://links.lww.com/SHK/B1000). SPD, a biomarker of alveolar-epithelial injury, had the lowest centrality. Two modules were detected: one module (red) included all HS disaccharides and endocan (a specific marker of endothelial injury); another module (blue) included sdc-1 and inflammation-related biomarkers previously validated in ARDS. Ang-2, angiopoetin-2; ARDS, acute respiratory distress syndrome; FDR, false discovery rate; HS, heparan sulfate; ICAM, intercellular adhesion molecule-1; RAGE, receptor for advanced glycation end-products; Sdc1, syndecan-1; SPD, surfactant protein D; TM, thrombomodulin; TNFRII, tumor necrosis factor receptor II; TriS, Tri-Sulfated HS (2SNS6S); VCAM, vascular cell adhesion molecule-1; vWF, von Willebrand factor.

FIG. 3.

A, Gap statistic. The gap statistic supported a 3-cluster solution. B, Standardized means for HS disaccharide sulfation patterns by patient cluster. Cluster 3 (green, n = 5) was characterized by significantly elevated levels of all HS disaccharide sulfation patterns relative to cluster 1 (red, n = 12) and 2 (blue, n = 19). All differences were statistically significant at P < 0.05. HS, heparan sulfate.

FIG. 4.

A, Cluster differences in peak OSI. Clusters 2 and 3 were associated with significantly higher peak OSI relative to cluster 1. B, Cluster differences in 28-day VFDs. Cluster 3 was associated with significantly fewer 28-day VFDs relative to cluster 1. P < 0.05 was statistically significant. OSI, oxygen saturation index; VFD, ventilator-free days.