EMC3 regulates trafficking and pulmonary toxicity of the SFTPCI73T mutation associated with interstitial lung disease

Abstract

The most common mutation in surfactant protein C gene (SFTPC), SFTPCI73T, causes interstitial lung disease with few therapeutic options. We previously demonstrated that EMC3, an important component of the multiprotein endoplasmic reticulum membrane complex (EMC), is required for surfactant homeostasis in alveolar type 2 epithelial (AT2) cells at birth. In the present study, we investigated the role of EMC3 in the control of SFTPCI73T metabolism and its associated alveolar dysfunction. Using a knock-in mouse model phenocopying the I73T mutation, we demonstrated that conditional deletion of Emc3 in AT2 cells rescued alveolar remodeling/simplification defects in neonatal and adult mice. Proteomic analysis revealed that Emc3 depletion reversed the disruption of vesicle trafficking pathways and rescued the mitochondrial dysfunction associated with I73T mutation. Affinity purification-mass spectrometry analysis identified potential EMC3 interacting proteins in lung AT2 cells, including valosin containing protein (VCP) and its interactors. Treatment of SftpcI73T knock-in mice and SFTPCI73T-expressing iAT2 cells derived from SFTPCI73T patient-specific iPSCs with the VCP inhibitor CB5083 restored alveolar structure and SFTPCI73T trafficking, respectively. Taken together, the present work identifies the EMC complex and VCP in the metabolism of the disease-associated SFTPCI73T mutant, providing therapeutical targets for SFTPCI73T-associated interstitial lung disease.

Keywords: Cell biology; Protein traffic; Pulmonology.

Figure 1. Generation of Sftpc^I73T knock-in mouse model.

(A) Design of the Sftpc^I73T knock-in allele. Founder mice carrying the knock-in allele were crossed to EIIA-Cre deleter mice to excise the neomycin resistance cassette. (B) Representative H&E-stained sections from 6–8 week-old WT/WT and I73T/I73T mice. Scale bars: 200 μm. (C) Quantitative morphometry of B using ImageJ expressed as the volume density of alveolar septa (V_Vsep), mean linear intercept of the airspaces (Lm), the mean transsectional wall length (Lmw), and the surface area density of the air spaces (S_Vair). I73T/I73T lungs showed significantly reduced volume density of alveolar septa and increased mean linear intercept of the airspaces. Mean ± SEM; **P < 0.01 using student t test, n = 5. (D) Reduced Sftpc mRNA in isolated AT2 cells from 8-week-old I73T/I73T mice compared with WT/WT mice. Levels of the Sftpc transcript were normalized to that of 18S by qPCR. Mean ± SEM; **P < 0.01 using unpaired, 2-tailed Student’s t test, n = 4/group. (E) 8-week-old I73T/I73T and WT/WT lung sections were stained for SP-C proprotein (proSP-C, green) and DAPI (blue). WT proSP-C is detected in a punctate pattern while proSP-C(I73T) is detected as a dense staining stripe near the cell surface. Red signal is autofluorescence of erythrocytes. Scale bars: 20 μm. (F and G) Western blot of lysates of isolated AT2 cells (F) or whole lung homogenates (G) from 6–8 week-old WT/WT and I73T/I73T mice. Processing of SP-C(I73T) proprotein into mature peptide (mSP-C) is decreased and processing intermediates are increased. (H) Western blot of bronchoalveolar lavage fluid (BALF) detected secreted proteins from 6–8 week-old WT/WT and I73T/I73T mice. Decreased mature SP-C and increased proSP-C(I73T) processing intermediates were present in BALF of I73T/I73T mice. Lysozyme was used as a loading control.

Figure 2. Deletion of Emc3 rescued I73T-associated alveolar simplification in the neonatal mice.

(A) AT2-specific deletion of Emc3 was induced by injection of tamoxifen to neonatal mice on P6, 7 and 8. Lung tissue was analyzed on P21. Representative H&E-stained lung sections are shown. Scale bars: 200 μm. (B) ImageJ was used to quantify the H&E staining results in A. Both defects of volume density of alveolar septa (V_Vsep) and mean linear intercept of the airspaces (Lm) in I73T/CreERT lungs were rescued by deletion of Emc3. Mean ± SEM; *P < 0.05, **P < 0.01, ***P < 0.001 using 1-way ANOVA multiple comparisons test, n = 6 (WT/CreERT), n = 5 (I73T/CreERT), n = 6 (I73T/CreERT;Emc3^Δ/Δ). (C) Whole lung homogenates from mice treated as in A were prepared on P21 for Western blotting. Mature SP-B (mSP-B) levels were similar among 3 groups. Reduced levels of mSP-C were detected in I73T/CreERT and I73T/CreERT; Emc3^Δ/Δ lungs. (D) Lung sections were prepared from P21 mice treated as in A and stained for mSP-B (green), mSP-C (red) and DAPI (blue). While mSP-B was unchanged, mSP-C was decreased in I73T/CreERT and I73T/CreERT; Emc3^Δ/Δ lungs. Scale bars: 20 μm. (E) Western blot of AT2 cell lysates prepared from P21 mice treated as in A. (F) Lung sections were prepared from P21 mice treated as in A and stained for proSP-C (green), ABCA3 (red), LAMP1 (white), and DAPI (blue). Deletion of Emc3 did not change ABCA3 staining or its colocalization with LAMP1-positive vesicles, while it decreased staining of proSP-C(I73T) near the plasma membrane and caused diffuse intracellular staining. Scale bars: 20 μm. (G) Western blot of AT2 cell lysates prepared from P21 mice treated as in A. (H) Quantification of Western blots in G. ABCA3 protein levels were normalized to that of Actin.

Figure 3. Proteomic changes caused by expression of SP-C(I73T).

(A) Protein sequencing data were obtained from AT2 cells sorted from P21 control (WT/CreERT) and Sftpc^I73T heterozygous (I73T/CreERT) mice treated with tamoxifen as in Figure 2A. Protein concentrations in AT2 cells from I73T/CreERT were compared with those in WT/CreERT mice. Proteins significantly altered were unbiasedly clustered and visualized in a z score–normalized heatmap. (B) Functional enrichment analyses of the significantly changed proteins in A were performed using Toppfun. A subset of significant relationships was represented by graphing the corresponding –log₁₀ (P value). Blue represents proteins decreased and red represents those increased in AT2 cells from I73T/CreERT mice compared with WT/CreERT.

Figure 4. Loss of Emc3 rescues trafficking defects and mitochondria dysfunction caused by SP-C(I73T).

(A) Protein sequencing data were obtained from AT2 cells sorted from control (WT/CreERT), Sftpc^I73T heterozygous (I73T/CreERT), and I73T/CreERT;Emc3^Δ/Δ mice on P21 after treatment with tamoxifen as in Figure 2A. Protein levels in I73T/CreERT;Emc3^Δ/Δ AT2 cells were compared with that in the other 2 genotypes and proteins significantly altered in 1 or more conditions were unbiasedly clustered and visualized in a z score–normalized heatmap. Group 1, proteins induced in AT2 cells from I73T/CreERT;Emc3^Δ/Δ lungs; Group 2, proteins rescued by Emc3 deletion (induced in I73T/CreERT and reversed in I73T/CreERT;Emc3^Δ/Δ); Group 3, proteins only repressed in I73T/CreERT;Emc3^Δ/Δ AT2 cells. Functional enrichment analyses of these proteins were performed using Toppfun. A subset of significant relationships was represented by graphing the corresponding –log₁₀ (P value). (B) Lung sections were prepared from P21 mice treated with tamoxifen as in Figure 2A and stained for proSP-C (green), an early endosome marker, EEA1 (red), and DAPI (blue). Diffuse staining of EEA1 in AT2 cells is shown in WT/CreERT controls. SP-C(I73T) accumulated with EEA1 in close proximity to plasma membranes. Emc3 deletion restored the diffuse intracellular distribution of both proSP-C(I73T) and EEA1. (C) Lung sections were prepared from P21 mice treated with tamoxifen as in Figure 2A and were stained for proSP-C (green), TOM20 (red), and DAPI (blue). Mitochondrial outer membrane protein TOM20 accumulated in SP-C(I73T)–expressing AT2 cells and was restored by deletion of Emc3. Scale bars: 20μm.

Figure 5. Loss of Emc3 rescues I73T-associated alveolar simplification and mitochondrial dysfunction in adult mice.

(A) AT2-specific Emc3 deletion was induced by administration of tamoxifen chow to control (WT/CreERT), Sftpc^I73T heterozygous (I73T/CreERT), and I73T/CreERT;Emc3^Δ/Δ adult mice (6–8 weeks of age). Lung tissues were obtained 14 days later. Representative H&E-stained lung tissue is shown. Scale bars: 200 μm. (B) ImageJ was used to quantify the H&E staining results in A. Both defects of volume density of alveolar septa (V_Vsep) and mean linear intercept of the airspaces (Lm) in I73T/CreERT lungs were rescued by deletion of Emc3. Mean ± SEM; **P < 0.01, ***P < 0.001 using 1-way ANOVA multiple comparisons test, n = 6. (C) Lung sections from adult mice treated as in A were stained for proSP-C (green), LAMP1 (red) and DAPI (blue). Emc3 deletion restored the intracellular distribution of SP-C(I73T). Scale bar: 20 μm. (D) Oxygen consumption rate (OCR) and extracellular acidification rates (ECAR) of AT2 cells isolated from adult mice treated as in A were tested with the Seahorse XF96 analyzer using 1.2 × 10⁵ cells for each sample. OCR was measured under basal conditions, followed by addition of oligomycin (Oligo; 2 μM), FCCP (2 μM), and antimycin A (Ant A; 2.5 μM), as indicated. ECAR was measured under basal conditions followed by addition of glucose (10 nM), oligomycin (Oligo; 2 μM), and 2-DG (50 mM), as indicated. Emc3 deletion restored maximal respiration and glycolytic capacity of AT2 cells expressing SP-C(I73T). Mean ± SEM; **P < 0.01, ****P < 0.0001 using 2-way ANOVA multiple comparisons test, n = 6 (WT/CreERT), n = 3 (I73T/CreERT), n = 5 (I73T/CreERT;Emc3^Δ/Δ).

Figure 6. EMC3 interacts with VCP to influence SP-C(I73T) trafficking.

(A) Mass spectrometry of proteins isolated from EMC3 coimmunoprecipitates in MLE-15 cells identified 26 proteins as potential EMC3 binding partners. For each experimental pair, MLE-15 cells were transfected with empty vector or vector encoding Myc-tagged EMC3 (Myc-EMC3). Myc antibody coimmunoprecipitates were isolated from cell lysates using μMACS c-myc Isolation Kit (Miltenyi Biotec) and analyzed by mass spectrometry. Five independent pairs of co-IP assays were performed and analyzed. The R package apmsWAPP, sub package TSPM, was used to determine significant EMC3 interacting partners, P < 0.05. Significant EMC3 PPI partners were visualized in a z score–transformed heatmap of the normalized spectral counts. (B) Functional enrichment analyses of potential EMC3 interaction partners were performed using Toppfun. A subset of significant relationships was represented by graphing the corresponding –log₁₀ (P value). (C) Normalized VCP protein levels were measured from the proteomic analysis in Figure 4A. Mean ± SEM; **P < 0.01 using 2-way ANOVA multiple comparisons test, n = 4/group. (D) VCP inhibitor, CB5083, was given every other day (q.o.d.) by oral gavage at 50 mg/kg/dose to 6–8 week-old I73T/I73T mice for 14 days. Representative H&E-stained lung sections are shown. Scale bar: 200 μm. (E) ImageJ was used to quantify the H&E staining results in D. Mean ± SEM; **P < 0.01, ***P < 0.001, ****P < 0.0001 using 1-way ANOVA multiple comparisons test, n = 6. (F) Lung sections from adult mice treated in D were stained for proSP-C and cell-surface marker WGA. Scale bars: 20 μm. (G) Quantification of the ratio of cell-surface proSP-C stained in F by ImageJ. For each group, 20 AT2 cells from 4 mice were quantified. (H) Western blot of lysates from AT2 cells isolated from CB5083 or vehicle treated adult mice as in D. Emc3 deletion did not change levels of proSP-C(I73T).

Figure 7. Accumulation and localization of SP-C(I73T) and EMC4 in patients with SFTPC^I73T-related ILDs.

Representative confocal immunofluorescence microscopy of sections of explanted lungs from normal donors and patients with SFTPC^I73T-associated ILD. Scale bars: 20 μm. (A) Lung sections were stained for proSP-C (green), EEA1(red), WGA (white), and DAPI (blue). WGA staining was used to mark the plasma membrane. Accumulation and colocalization of proSP-C(I73T) and EEA1 was detected in lung tissues from the patients with SFTPC^I73T. (B) Lung sections were stained for proSP-C (green), EMC4 (red), and DAPI (blue). Accumulation and distinct distribution of EMC4 was detected in AT2 cells of patients with SFTPC^I73T. (C) Quantification of the ratio of subcellular proSP-C stained in A by ImageJ. Fifteen AT2 cells were quantified for each individual. Note that AT2 cells in the 2 patients accumulate more proSP-C(I73T) on the cell surface and in EEA1-positive endosomes.

Figure 8. Rescue of I73T-iAT2 cells by EMC3 shRNA and VCP inhibition.

(A) I73T/tdT and WT/tdT iAT2 cells were transfected with control shRNA (shControl) or EMC3 shRNA (shEMC3). Fourteen days after transfection, cells were fixed and paraffin sections were stained for proSP-C (green), EMC4 (red), and DAPI (blue). EMC4 staining was used to identify sites of Emc3 inhibition. Scale bars: 20 μm. (B) Western blots were performed on lysates of iAT2 cells in A. In I73T/tdT iAT2 cells, shEMC3 transfection reduced EMC3 and TOM20 expression. (C) WT/tdT and I73T/tdT iAT2 cells were treated with DMSO (vehicle) or 0.25 μM CB5083 for 14 days and stained for proSP-C (green), proSP-B (red), LAMP1 (white), and DAPI (blue). Scale bars: 20 μm. (D) Western blots on lysates of iAT2 cells treated with DMSO or CB5083 as in C. (E) Quantification of Western blots in D. Total proSP-C (n = 3), EMC3 (n = 5), and TOM20 (n = 4) were normalized to that of actin. Mean ± SEM; *P < 0.05, **P < 0.01, ****P < 0.0001 using 2way ANOVA multiple comparisons test. (F) I73T/tdT iAT2 cells were treated with CB5083 as in C and stained for proSP-C (green), TOM20 (red), and DAPI (blue). Scale bars: 20 μm. (G) Oxygen consumption rate (OCR) of WT/tdT or I73T/tdT iAT2 cells treated with vehicle (DMSO) or CB5083 were tested with the Seahorse XF96 analyzer using 2 × 10⁴ cells for each sample. OCR was measured under basal conditions, followed by addition of oligomycin (Oligo; 2 μM), FCCP (2 μM), and antimycin A (Ant A; 2.5 μM), as indicated. Emc3 deletion restored basal respiration, proton leak, and maximal respiration of I73T/tdT iAT2 cells. Mean ± SEM; **P < 0.01, ***P < 0.001, ****P < 0.0001 using 1-way ANOVA multiple comparisons test, n = 5 (WT/tdT+DMSO), n = 6 (WT/tdT+CB5083), n = 5 (I73T/tdT+DMSO), n = 6 (I73T/tdT+CB5083).