The IRE1α-XBP1s Arm of the Unfolded Protein Response Activates N-Glycosylation to Remodel the Subepithelial Basement Membrane in Paramyxovirus Infection

Abstract

Respiratory syncytial virus (RSV) causes severe lower respiratory tract infections (LRTI) associated with decreased pulmonary function, asthma, and allergy. Recently, we demonstrated that RSV induces the hexosamine biosynthetic pathway via the unfolded protein response (UPR), which is a pathway controlling protein glycosylation and secretion of the extracellular matrix (ECM). Because the presence of matrix metalloproteinases and matricellular growth factors (TGF) is associated with severe LRTI, we studied the effect of RSV on ECM remodeling and found that RSV enhances the deposition of fibronectin-rich ECM by small airway epithelial cells in a manner highly dependent on the inositol requiring kinase (IRE1α)-XBP1 arm of the UPR. To understand this effect comprehensively, we applied pharmacoproteomics to understand the effect of the UPR on N-glycosylation and ECM secretion in RSV infection. We observe that RSV induces N-glycosylation and the secretion of proteins related to ECM organization, secretion, or proteins integral to plasma membranes, such as integrins, laminins, collagens, and ECM-modifying enzymes, in an IRE1α-XBP1 dependent manner. Using a murine paramyxovirus model that activates the UPR in vivo, we validate the IRE1α-XBP1-dependent secretion of ECM to alveolar space. This study extends understanding of the IRE1α-XBP1 pathway in regulating N-glycosylation coupled to structural remodeling of the epithelial basement membrane in RSV infection.

Keywords: IRE1α; N-glycosylation; XBP1; extracellular matrix; hexosamine biosynthetic pathway; unfolded protein response.

Figure 1

RSV induces ECM remodeling via the IRE1α–XBP1 arm of UPR. hSAECs were treated with solvent control (DMSO) or KIRA8 (10 µM) and mock- or RSV infected (MOI = 1, 24 h). Total RNA was extracted and analyzed by Q-RT-PCR for (A) XBP1 splicing; (B) GFPT2; and (C) FN1. For each graph, fold change mRNA relative to solvent-treated mock-infected cells is shown. ***, p < 0.001; n.s., not significant. (D), hSAECs were cultured on PDL-gelatin coated coverslips until confluent, which was followed by treatment with solvent or KIRA8. Cells were then mock or RSV-infected (MOI = 1, 24 h). The cells were fixed and stained for extracellular FN1 without permeabilization. Nuclei were then stained with DAPI. Red, FN1. Blue, DAPI. Scale bar 50 µm shown. (E). Identically treated plates were decellularized and stained for FN1 and imaged. (F,G), Quantitation of the FN1 fluorescence intensity by FIJI. The data points and mean from three independent experiments are presented. **, p < 0.01; ***, p < 0.001; ****, p < 0.0001; n.s., not significant.

Figure 2

Proteomics analysis of hSAECs infected with RSV in the presence or absence of KIRA8. hSAECs were infected with RSV at 1.0 MOI for 24 h in the presence or absence of KIRA8 (10 µM). The proteins were analyzed with label-free LC-MS/MS. (A) Principal component analysis of significant proteins (ANOVA with permutation-based FDR < 0.01). Green circle, controls; red square, RSV infection; blue diamond, RSV infection + KIRA8 treatment. (B) Unsupervised hierarchical cluster analysis of 813 significant proteins. The colors of the heatmap represent the z-scored normalized log2 LFQ intensity of each protein. (C) Volcano plot of proteins (RSV+KIRA8 vs. RSV). Significantly proteins. Red circle, proteins upregulated by KIRA8; green square, proteins downregulated by KIRA8. (D) Top Panther Reactome pathways activated by RSV infection but blocked by KIRA8 (FDR < 0.05%). (E) Protein expression of IRF3-mediated type I IFN genes. (F) Expression of proteins involved in the interaction of viral structure protein NS2 with the cellular export machinery. Student’s t-test with permutation correction, *, q < 0.05.

Figure 3

Proteomics analysis of N-glycosylation in hSAECs infected with RSV in the presence or absence of KIRA8. hSAECs were infected with RSV at 1.0 MOI for 24 h in the presence or absence of KIRA8 (10 µM). The N-glycosylated peptides were enriched with lectins and then analyzed with label-free LC-MS/MS. (A) Volcano plot of N-glycosylated peptides (RSV vs. Control). Red circle, N-glycoproteins upregulated by RSV; green square, N-glycoproteins downregulated by RSV infection. (B,C) Some N-glycosylated peptides strongly induced by RSV infection and regulated by the IRE1α–XBP1 arm of UPR are shown (Student’s t-test with permutation FDR < 0.05). (D) Panther Reactome pathways activated by RSV infection (FDR < 0.05%). (E) Panther Reactome pathways activated by RSV infection and attenuated by KIRA8 (FDR < 0.05%). (F) N-glycosylation of proteins involved neutrophil degranulation, which was regulated by the IRE1α–XBP1 arm of UPR. Student’s t-test with Permutation correction, *, q < 0.05, **, q < 0.01, ***, q < 0.001.

Figure 4

Proteomics analysis of secretome in hSAECs infected with RSV in the presence of KIRA8. hSAECs were infected with RSV at MOI 1.0 for 24 h in the presence or absence of KIRA8 (10 µM). The secretome was analyzed with label-free LC-MS/MS. (A) Volcano plot of secretome (RSV vs. Control). Dark red square, cytokines and growth factors; green circle, extracellular matrix proteins; and red open square, RSV proteins. (B) Examples of proteins whose secretion were significantly induced by RSV and blocked by KIRA8, including cytokines, growth factors, proteases, protease inhibitors, ECM, and ECM-modifying enzymes. (C) Pathways are strongly induced by RSV infection and regulated by the IRE1α–XBP1 arm of UPR (FDR < 0.05). (D) Correlation between proteome and secretome profiles (RSV vs. Control). (E) Correlation between proteome and secretome profiles (RSV-KIRA8 vs. RSV). (F) Protein expression of some proteins that were shown in Figure 4B. Student’s t-test with Permutation correction, *, q < 0.05, **, q < 0.01, ***, q < 0.001.

Figure 5

Histological analysis of IRE1α signaling in SeV infection. Masson’s trichrome staining was performed on paraffin-embedded sections from uninfected, SeV infected, or SeV+KIRA8 treated animals. Shown is a small airway. Images were taken at 40×; scalebar indicates 90 µm. Note the subepithelial accumulation of cells (nuclei) and expansion of ECM (blue) in the SeV infected mice that was reduced by KIRA8.

Figure 6

Proteomics analysis of BALF of mice infected with Sandi virus (SeV) in the presence or absence of KIRA8. The mice were infected with SeV in the presence or absence of KIRA8 (n = 6 in each group). The BALF was analyzed with label-free LC-MS/MS. (A) Unsupervised hierarchical cluster analysis of significant proteins (ANOVA with permutation-based FDR < 0.01). The colors of the heatmap represent the z-scored normalized log2 LFQ intensity of each protein. (B) GO annotation enrichment analysis of protein in each cluster (FDR < 0.02). The colors of heatmap represent log2 enrichment factors. Red, enrichment; green, depletion; and gray, not significant. (C) ER proteins in Cluster 1, whose secretion was induced by SeV and blocked by KIRA8. (D) Proteins related to innate immunity. (E) Serine proteases and peptidases. (F) Profiles of BALF glycoproteins in Cluster 1. (G) Profile of mucin-4 in BALF. (H) Profiles of protease inhibitors in Cluster 4, whose secretion was upregulated by KIRA8. (I) Pathways are strongly induced by RSV infection and regulated by the IRE1α–XBP1 arm of UPR (FDR < 0.05). Student’s t-test with Permutation correction, *, q < 0.05, **, q < 0.01, ***, q < 0.001.

Figure 7

RSV induced N-glycosylation is mediated by the IRE1α–XBP1 arm of the UPR. A schematic view of the relationship between the IRE1α–XBP1 pathway of the unfolded protein response, accumulation of UDP-GlcNAc, protein N glycosylation, and remodeling of the basal lamina. IRE1 activated in the ER induces alternative splicing and produces the formation of activated XBP1s, which is a transcription factor controlling the expression of the hexosamine biosynthetic pathway, integrin (ITG), and ECM components, including fibronectin 1 (FN1). UDP-GlcNAc is a rate-limiting enzyme for protein N-glycosylation. After processing through the Golgi, glycosylated ECM components are presented on the cell surface and contribute to remodeling of the basal lamina.