Fig 2: The impact of pathogen components on the trans-binding of EPH and EFN(A) Real-time EPH-EFN trans-binding measurement was performed during the treatment of PAMPs and heat-killed bacteria. We presented relative changes in EPH-EFN binding compared to the control (CON) at the 4-h after stimulation in a heatmap. All the experimental results over the entire time course are presented in Figure S4 (n = 3).(B) Time course analysis of the changes in EPHA2-EFNA1 trans-binding following treatment with specified pathogen molecules. We presented the statistical significance of the results at the 4-h after treatment (n = 3). RM one-way ANOVA, Dunnett’s test.(C) Comparison of the time course of inflammatory response and the changes in EPHA2-EFNA1 trans-binding during treatment with 100 ng/mL FLA-PA in BEAS-2B cell. The inflammatory response was assessed by analyzing IL-8 mRNA level by RT-qPCR (n = 3).(D) The correlation between EPHA2-EFNA1 trans-binding (Figure 4A) and IL-8 or IL-6 mRNA induction (Figure S4B) in BEAS-2B cells 4 h after each stimulation. Data are presented as mean ± SEM (B (+SEM only), C, D).

Fig 3: Cleavage of EFNA1 induces inflammatory response in AECs(A) The effect of soluble EFNA1 on cytokine mRNA expression in AECs. BEAS-2B cells were treated with 50 ng/mL EFNA1-Fc for 0–24 h, and the inflammatory response was analyzed by RT-qPCR. n = 3, RM one-way ANOVA, Dunnett’s test.(B) TEV protease-induced specific cleavage of SF-a1-TEV in cells. BEAS-2B cells expressing SF-a1 or SF-a1-TEV were treated with the indicated concentrations of TEV protease in CO2-independent medium and TEV buffer at 37°C for 3 h. The released SF-a1 and SF-a1-TEV in the supernatant were measured. n = 3, RM two-way ANOVA, Sidak’s test.(C) TEV protease-induced specific dissociation of EPHA2 and SF-a1-TEV in cells. iLg-EPHA2-HA expressing cells were co-cultured with SF-a1 or SF-a1-TEV expressing cells, and the EPHA2-EFNA1 trans-binding level was measured during treatment of 200 U/ml TEV protease. n = 3.(D) The impact of specific cleavage of SF-a1-TEV on inflammatory response in AECs. SF-a1 or SF-a1-TEV expressing cells were treated with the indicated concentration of TEV protease for 3 h, and the inflammatory response was analyzed by RT-qPCR. n = 3, two-way ANOVA, Tukey’s test.(E) The effect of MMPs/ADAMs KD on FLA-PA-induced inflammatory cytokine up-regulation. BEAS-2B cells expressing Sm-Flag-EFNA1 were transfected with 20 nM siRNA targeting ADAMs and MMPs, and the inflammatory response was analyzed by RT-qPCR 4 h after treatment with 100 ng/mL FLA-PA. siNT: n = 8, siADAM9: n = 5, Other groups: n = 3, RM two-way ANOVA, Sidak’s test. (F) Correlation analysis between the impact of ADAM/MMP KD on FLA-PA-dependent IL-8 and IL-6 mRNA induction (Figure 7E) and EFNA1 release (Figure S6C). Data are presented as mean ± SEM (A, B, C (-SEM only), D, E, F).

Fig 4: Establishment of EPH-EFN trans-binding real-time measurement in live AECs(A) Schematic of established iLgBiT-EPH-HA and SmBiT-Flag-EFN(-V5) constructs. Original plasmid information of EPH and EFN genes used in this study is listed in Table 2. SP: signal peptide, LBD: ligand binding domain, FNIII: fibronectin-type III, TM: transmembrane domain, SAM: sterileαmotif, PDZ: PDZ-binding, RBD: receptor binding domain, GPI-anc: GPI anchor propeptide region.(B) Western blotting analysis of iLg-EPH-HA and Sm-Flag-EFN protein expression in BEAS-2B cells with stable expression. iLg-EPH-HA expression cells were treated with 1 μg/mL Dox for 72 h to induce their expression. The exogenous EPHs and EFNs were detected by anti-HA antibody and anti-Flag antibody, respectively. The expected molecular weight (MW: kDa) were shown below each band.(C) Plasma membrane expression analysis of iLg-EPH-HA in BEAS-2B cells by using HiBiT-Halo protein. iLg-EPH-HA expression was induced by treatment of Dox (A2, A4: 50 ng/mL, B1: 500 ng/mL, B3: 150 ng/mL, B4, B6: 1000 ng/mL) for 48 h (n = 3).(D) PM expression of Sm-Flag-EFN in BEAS-2B cells was measured by cell-surface ELISA using anti-Flag antibody and HRP-conjugated anti-mouse antibody (n = 3).(E) Schematic of EPH-EFN trans-binding measurement using iLg-EPH-HA-expressing cells (EPH, black) and Sm-Flag-EFN-expressing cells (EFN, red) co-culturing system. EPH-EFN binding is formed in cytonemes contact site between EPH or EFN expressing cells, and the signal of complemented NanoLuc reflects the amount of EPH-EFN interaction. Membrane-permeable substrate enables to detect EPH-EFN interactions inside cells.(F) Immunostaining of the cell-cell contact site during the co-culturing of iLg-EPHA2-HA and Sm-Flag-EFNA1 or iLg-EPHA4-HA and Sm-Flag-EFNB2-V5 expressing cells. The iLg-EPH-HA were labeled in green and the Sm-Flag-EFN were labeled in red. Outline of iLg-EPHA4-HA expressing cell (Green) or Sm-Flag-EFNB2-V5 (Red) expressing cells was shown in broken line. The right panel shows an enlarged image of the white box area of left panel, and white arrowhead shows co-localization.(G) The quantitative results of EPH-EFN trans-binding measurement between iLg-EPH and Sm-Flag-EFN in BEAS-2B cells were displayed as a heatmap, with the trans-binding amount relative to EPHA2-EFNA1 pair (n = 3). Data are presented as mean ± SEM (C, D).

Fig 5: FLA-PA induces TLR5 dependent EFNA1 release from the PM in AECs(A) WT or TLR5 KO BEAS-2B cells were transfected with 20 nM EPHA2 siRNA and the FLA-PA induced inflammatory response was analyzed by RT-qPCR. n = 3, RM two-way ANOVA, Tukey’s test.(B) WT or TLR5 KO BEAS-2B cells were treated with 100 ng/mL FLA-PA for 0 to 120 min, and the total EPHA2, phospho-S897 (pS897) and phosphor-Y772 (pY772) EPHA2 levels were analyzed by WB. Ponceau staining was used as loading control.(C) The ratio of pS897 or pY772/total EPHA2 at 1 h after FLA-PA treatment was calculated from the intensity of the WB result. n = 10, 10, 8, 8 (pS897, from left to right) and n = 7, 7, 6, 6 (pY772, from left to right), unpaired t-test, two-tailed.(D) Analysis of whether the TLR5-dependent dissociation of EPHA2-EFNA1 by FLA-PA results in an effect on EPHA2 or EFNA1. WT or TLR5 KO BEAS-2B cells expressing iLg-EPHA2-HA or Sm-Flag-EFNA1 were co-cultured in indicated combinations. The EPHA2-EFNA1 trans-binding relative to CON was quantified at 150 min after 100 ng/mL FLA-PA treatment. n = 3, unpaired multiple t-test, Bonferroni-Dunn test.(E) The analysis of EFN-RBD dependency in FLA-PA induced EPHA2-EFNA1 dissociation. BEAS-2B cells expressing Sm-Flag fused WT-EFNA1, a5 RBD-EFNA1 or a1 RBD-EFNA5 were co-cultured with iLg-EPHA2-HA-expressing cells. The EPHA2 and WT-EFNA1, a5 RBD-EFNA1 or a1 RBD-EFNA5 trans-binding was measured during 100 ng/mL FLA-PA treatment. The interaction ratio relative to CON was quantified at 180 min after FLA-PA-treatment. n = 3, unpaired multiple t-test, Bonferroni-Dunn test.(F) The FLA-PA impact on EPHA2 p.m. level. iLg-EPHA2-HA expressing BEAS-2B cell were treated with 100 ng/mL FLA-PA for 0–3 h and PM expression level of iLg-EPHA2-HA was analyzed by using HiBiT-Halo protein. n = 3, RM one-way ANOVA, Dunnett’s test.(G) The FLA-PA impact on EFNA1 p.m. level. SmBiT-Flag-EFNA1 p.m. levels in BEAS-2B cells under 100 ng/mL FLA-PA treatment for 0–4 h was analyzed by cell surface ELISA with anti-Flag antibodies. n = 3, RM one-way ANOVA, Dunnett’s test.(H) Proteasomal or lysosomal degradation inhibitor effect on FLA-PA induced EFNA1 loss from PM. Sm-Flag-EFNA1 expressing BEAS-2B cells were treated with 10 μM MG132 or 1 μM Bafilomycin A1 (Baf A1) for 2 h before and during treatment with 100 ng/mL FLA-PA. PM level of Sm-Flag-EFNA1 was measured at 1 h after FLA-PA treatment by cell surface ELISA and shown as relative to CON. n = 3, two-way ANOVA, Tukey’s test.(I and J) FLA-PA effect on endogenous EFNA1 protein level. BEAS-2B parental cells were treated with 100 ng/mL FLA-PA for 4 h, and the endogenous EFNA1 levels in the supernatant (Sup.) and cell lysate (Lysate) were immunoblotted (IB) by using anti-EFNA1 antibody.(J) The intensity of EFNA1 band was quantified. n = 4, unpaired t-test. Data are presented as mean ± SEM (A, C, D, E, F, G, H, J).