Lung transcriptomics of radiologic emphysema reveal barrier function impairment and macrophage M1-M2 imbalance

Emphysema, Imaging, Transcriptomics, Inflammation, Pathways 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 5

(250/250 words) 1 Rationale: While many studies have examined gene expression in lung tissue, the gene 2 regulatory processes underlying emphysema are still not well understood. Finding efficient 3 non-imaging screening methods and disease-modifying therapies has been challenging, but 4 knowledge of the transcriptomic features of emphysema may help in this effort. 5

Our goals were to identify emphysema-associated biological pathways through 6 transcriptomic analysis of bulk lung tissue, to determine the lung cell types in which these 7 emphysema-associated pathways are altered, and to detect unique and overlapping 8 transcriptomic signatures in blood and lung. 9

Using RNA-sequencing data from 456 samples in the Lung Tissue Research 10 Consortium and 2,370 blood samples from the COPDGene study, we examined the 11 transcriptomic features of computed tomography quantified emphysema. We also queried 12 lung single-cell RNA-sequencing data to identify cell types showing COPD-associated 13 differential expression of the emphysema pathways found in the bulk analyses. 14 Measurements and Main Results: In the lung, 1,055 differentially expressed genes and 29 15 dysregulated pathways were significantly associated with emphysema. We observed 16 alternative splicing of genes regulating NF-κ B and cell adhesion and increased activity in the 17 TGF-β and FoxO signaling pathways. Multiple lung cell types displayed dysregulation of 18 epithelial barrier function pathways, and an imbalance between pro-inflammatory M1 and 19 anti-inflammatory M2 macrophages was detected. Lung tissue and blood samples shared 251 20 differentially expressed genes and two pathways (oxidative phosphorylation and ribosomal 21 function). 22

This study identified emphysema-related changes in gene expression and 23 alternative splicing, cell-type specific dysregulated pathways, and instances of shared 24 pathway dysregulation between blood and lung. 25 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 6

1 Chronic obstructive pulmonary disease (COPD) is a major source of morbidity and 2 mortality (1). Emphysema, an important COPD phenotype, has been shown to be 3 independently associated with elevated risk for cardiovascular disease, lung cancer, and 4 mortality (2-4). Finding efficient non-imaging screening methods and targeted therapies may 5 be aided by knowledge of the transcriptomic characteristics of emphysema. Although 6 emphysema has been linked to genes involved in transforming growth factor beta (TGF- β ) 7 signaling (5, 6), B-cell mediated immunity (5, 7, 8), and hypoxia (8-10), our understanding of 8 emphysema-associated alternative splicing mechanisms and cell-type specific biological 9 pathways from human lung tissue is still limited. 10 Earlier transcriptomic studies of emphysema were limited in their sample sizes or the 11 scope of the panel of genes evaluated and have not investigated alternative splicing 12 extensively. In addition, although oxidative stress (11, 12) and cellular senescence (13, 14) 13 have been associated with emphysema, it is unclear whether other biological processes may 14 also be at play. It is also less obvious which cell types exhibit pathway dysregulation in the 15 lungs of subjects with emphysema compared to controls, although it is widely recognized that 16 cell types such as neutrophils (15, 16) and T-lymphocytes (17, 18) are drivers of the disease. 17 Comparing dysregulated genes and pathways in blood and lung tissues is also necessary to 18 better understand the utility of non-invasive diagnostic and prognostic tools for emphysema. 19 In the present study, we used genome-wide RNA-sequencing (RNA-seq) data from 20 lung tissue samples from the Lung Tissue Research Consortium (LTRC) to identify the genes 21 and alternative splicing mechanisms that are associated with computed tomography (CT)-22 quantified emphysema. We then queried a previously published single-cell RNA-seq atlas of 23 lung tissues of COPD patients and controls to determine which cell types show significant 24 associations to the identified emphysema pathways. We lastly compared emphysema-25 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 7 associated transcriptomic associations in LTRC lung tissue samples to emphysema-1 associations from whole blood samples in the COPD Genetic Epidemiology (COPDGene) 2 study. We hypothesized that there would be significant emphysema-associated transcriptomic 3 biomarkers and pathways identified from lung tissue, cell-type specific signatures, and 4 important similarities and differences between lung tissue and whole blood. 5 6

7 8 Study description 9 We obtained lung tissue samples from the NHLBI Lung Tissue Research Consortium 10 (LTRC) (https://www.nhlbi.nih.gov/science/lung-tissue-research-consortium-ltrc , 11 https://biolincc.nhlbi.nih.gov/studies/ltrc/). Details regarding subject recruitment has been 12 previously published (19). Subjects included smokers and non-smokers over the age of 21 13 who had undergone surgical lung biopsy, lung volume reduction surgery, lung transplantation, 14 or lung nodule/mass resection. These subjects cover the entire spectrum of the Global 15 Initiative for Chronic Obstructive Lung Disease (GOLD) spirometric grading system (20). 16 We excluded subjects missing the following clinical data: CT-quantified emphysema, CT 17 scanner model, forced expiratory volume in one second (FEV 1), body mass index (BMI), 18 current smoking status, and pack-years of smoking. We also excluded subjects with 19 idiopathic pulmonary fibrosis (IPF). IPF was defined according to the American Thoracic 20 Society/European Respiratory Society guidelines as a consensus clinical diagnosis of IPF or a 21 pathologic diagnosis of usual interstitial pneumonia or honeycomb lung in the absence of a 22 clinical diagnosis of another interstitial lung disease (21). 23 All analyses conducted on LTRC were also completed in previously published 24 analyses using whole blood RNA-seq data from the COPDGene study. The COPDGene study 25 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 8 is a longitudinal study investigating the epidemiologic and genomic characteristics of COPD. 1 This study included 10,371 smokers from 21 U.S. clinical institutions centers, representing 2 the full spectrum of lung health, who were between the ages of 45 and 80 and had at least ten 3 pack-years of lifetime cigarette smoking history (NCT00608764, www.copdgene.org) (22). 4 COPDGene continues to collect longitudinal data on study participants at five-year intervals. 5 Each study visit collected spirometry data, questionnaires, and chest CT scans using a 6 standard protocol. Whole blood RNA-seq was obtained at Visit 2. 7 All subjects provided informed consent, and all clinical sites received institutional 8 review board approval. 9 10 Emphysema quantification 11 The Analyze 8.1 (www.analyzedirect.com ) (23) and Thirona (www.thirona.eu ) 12 softwares quantified emphysema in LTRC and COPDGene, respectively. Emphysema was 13 measured as the Hounsfield units (HU) at the 15 th percentile of the CT density histogram at 14 end-inspiration (Perc15 density) (24, 25). In COPDGene, the Perc15 density values were 15 corrected for the inspiratory depth variations (adjusted Perc15 density). Both Perc15 and 16 adjusted Perc15 are given as the HU + 1000. The lower the Perc15 or adjusted Perc15 values 17 are, i.e., the closer to -1,000 HU, the more CT-quantified emphysema is present. 18 19 Differential expression and usage analyses 20 We used the limma-voom linear modeling approach (as implemented in limma 21 v3.46.0) to test for the associations between emphysema and whole blood RNA transcripts 22 (26, 27). While differential expression refers to the change in the absolute expression levels 23 of a feature, differential usage captures alternative splicing and refers to the change in the 24 relative expression levels of the isoforms/exons within a given gene. The concepts of 25 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 9 differential expression and usage with a discrete variable are naturally extended to a 1 continuous variable where the changes in expression and usage mean changes in the absolute 2 and relative rate of association with emphysema. All models were adjusted for age, race, sex, 3 pack-years of smoking, current smoking status, FEV 1, CT scanner model, and library 4 preparation batch. The LTRC analysis was also adjusted for BMI. The COPDGene analysis 5 was also adjusted for CBC cell count proportions. 6 7 Pathway analysis 8 We used the egsea software (v1.18.1) to perform gene set enrichment analysis (GSEA) 9 on the gene sets obtained from the differential gene expression (DGE) analysis (28). The 10 KEGG pathway gene sets were utilized as the reference for annotated gene sets, and the 11 CAMERA base approach was employed (29, 30). CAMERA is a competitive gene set test 12 approach that employs estimated inter-gene correlation to adjust the gene set test statistic (30). 13 The directionality of a pathway was ascertained by counting the number of upregulated and 14 downregulated genes in the gene set and taking the direction of the majority (28). The genes 15 corresponding to these pathways were mapped on the pathway maps given by the KEGG 16 database using the pathview and gage libraries in R. Twenty-nine pathways in total were 17 found to be dysregulated in our data, but five (Parkinson’s disease, Huntington’s disease, 18 dorso-ventral axis formation, Alzheimer’s disease, and proteasome) were not mapped to the 19 KEGG database. For easier visualization and interpretation, the reported gene log fold 20 changes in the pathway diagrams were multiplied by -100 so that positive log fold changes 21 represent upregulated genes and negative log fold changes represent downregulated genes. 22 Thus, the fold change estimates correspond to expression change per 100 HU decrease in 23 lung density (increasing emphysema). 24 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 10 1 Cell-type specificity analysis 2 We re-analyzed data from a previously published lung tissue single-cell RNA-seq 3 experiment to identify enriched pathways from the bulk lung DGE analyses at cell type 4 resolution (31). The dataset consists of 312,928 single cells from 78 human samples (28 5 control, 18 end-stage COPD, 32 IPF). The control samples in this cohort represented allograft 6 rejected donors, and only control and end-stage COPD samples were used for this analysis, 7 while keeping the author-provided cell type annotations. 8 A composite gene expression signature for the pathway members was summarized 9 using the AddModuleScore function from the Seurat package (32). Briefly, the function 10 calculates a normalized average expression for a set of genes belonging to a given pathway 11 subtracted by an aggregate expression of control genes. All genes in the dataset were binned 12 based on their expression levels. For each gene in the pathway, control genes were randomly 13 selected from bins with matching expressions. This ensured similar distribution of the gene 14 expression values in the control set. Finally, the activity of the pathway was defined by 15 subtracting the averaged expression of control genes from the average expression of the 16 member genes for each cell. We employed this background-corrected gene expression 17 signature approach to calculate pathway activity scores for the 24 KEGG pathways identified 18 using GSEA across every single cell in the single-cell dataset. We compared the pathway 19 activity scores between controls and end-stage COPD for each cell type separately using the 20 two-sample t-tests (33). 21 Next, we scored the specificity of a pathway across cell types using the pathway 22 activity scores. The pathway activity scores showed a normal unimodal distribution. We, 23 therefore, used an arbitrary cutoff of the mean plus two times the standard deviation to assess 24 the activity of a given pathway in all cells. This stringent positive cutoff value allows us to 25 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 11 look at cell types that strongly enrich a given pathway activity. The odds ratio was computed 1 for the cells showing the pathway activity versus not showing activity in a given cell type. 2 The P-values were calculated using the Fisher’s Exact Test (34). The negative log10 of the 3 corrected P-values were capped at 100 and plotted in a heatmap using the ComplexHeatmap 4 package (35). 5 6 Statistical analyses 7 Data were reported as mean with standard deviations or counts with percentages. 8 Upregulated versus downregulated genomic features or KEGG pathways were provided with 9 respect to their relationships with emphysema (i.e., they have opposite directions for their 10 associations with Perc15 or adjusted Perc15 density). Because Perc15 and adjusted Perc15 11 decrease with more severe emphysema, negative log fold change values represent features 12 upregulated with emphysema, and positive log fold change values represent features 13 downregulated with emphysema. The Benjamini-Hochberg method corrected for multiple 14 comparisons using a threshold of significance of a false discovery rate (FDR) of 10% (36). 15 16

17 18 Study subjects 19 The study flow diagram is depicted in Figure 1, and the missingness of the pertinent 20 covariates is shown in Figure E1. A total of 456 LTRC samples coming from 446 participants 21 were included in this analysis. Eight subjects had multiple lung tissue samples taken on the 22 same date. Of the 1,589 LTRC samples for which RNA-seq data were available, 508 had 23 comprehensive clinical information including quantitative CT emphysema. We further 24 excluded 52 samples with IPF. Table 1 provides an overview of the demographics and 25 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 12 clinical characteristics of the included subjects. The majority of these subjects were non-1 Hispanic whites with a mean age of 64, a mean BMI of 28.1, and a balanced sex 2 representation. 8.1% were current smokers with an average of 36 pack-years of smoking. 3 With the exception of more smoking and slightly less spirometric impairment in the included 4 subjects, there were no significant differences found between the included and excluded 5 subjects (Table E1). The baseline characteristics of the LTRC and the COPDGene individuals 6 were also comparable (Table E2), as well as the distributions of their respective Perc15 and 7 adjusted Perc15 values (Figure E2). 8 9 Differential gene expression in lung tissue 10 A total of 1,055 of the 17,353 genes evaluated in lung tissue samples achieved 11 statistical significance at FDR 10% (Table E3). Of these significant genes, 632 were 12 upregulated and 423 were downregulated with increasing emphysema. The top 20 most 13 significant differentially expressed genes (DEGs) are listed in Table 2. Interestingly, the 14 FOXL1 and F2RL2 genes, which had previously been found to be overexpressed in IPF lungs 15 (37, 38), were downregulated in lung tissues with emphysema. Table E4 presents the 29 16 significantly dysregulated pathways found by performing gene set enrichment analysis on the 17 identified DEGs. Table 3 reports ten noteworthy pathways that were selected for inclusion 18 based on their biological relevance. There was enhanced activity in pluripotency pathways 19 (TGF-β and forkhead box O (FoxO) signaling) and cell barrier function pathways (adherens 20 junction, Rap1 signaling, and gap junction). Figure 2 displays KEGG pathway maps that 21 show coordinated expression changes for TGF- β signaling, FoxO signaling, adherens 22 junction, and Rap1 signaling, which have been previously found to be dysregulated in COPD 23 or emphysema (39-44). Additional KEGG pathway maps are included in the supplemental 24

for the remaining six significantly enriched pathways (Figure E3). 25 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 13 1 Differential isoform and exon usage in lung tissue 2 Differential isoform usage (DIU) and differential exon usage (DEU) analyses were 3 performed to understand emphysema-associated alternative splicing changes in lung tissue. A 4 summary of the top twenty DUIs and DUEs can be found in Table 2. 5 Of the 41,891 isoforms evaluated, 730 isoforms were significantly associated with 6 emphysema (FDR 10%) (Table E5). Of these significant isoforms, 271 were upregulated and 7 459 were downregulated. Mapping these isoforms to their corresponding genes revealed that 8 6.9% (41/598) of them were significant in the DGE analysis (Figure 3). These corresponding 9 genes included MYB binding protein 1a ( MYBBP1A) and protein kinase C zeta ( PRKCZ), 10 regulators of nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κ B) (45, 46). 11 Additionally, there were significant isoforms mapping to CD63 and paladin ( PALLD), which 12 are involved in cell adhesion and cell-cell junctions (47, 48). An isoform of interferon gamma 13 receptor 1 (IFNGR1) was also significantly associated with emphysema. 14 Of the 162,747 exons evaluated, 285 were significantly associated (FDR 10%) with 15 emphysema with 24 being upregulated and 261 being downregulated (Table E6). After 16 mapping these exons to their respective genes, we found that 5.6% (12/216) of the genes 17 were differentially expressed (Figure 3). One of the top significantly associated exons 18 corresponded to TRAPPC9 , which codes for a protein that regulates NF- κ B activation (49). 19 We also identified exons within SNX1, NBAS, and COG2 , which are genes involved in 20 intracellular transport (50-52). 21 22 Identification of cell-type specific pathways 23 To identify lung cell types showing the highest expression levels of emphysema 24 associated pathways, we computed the odds ratio for each pathway and cell type using the 25 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 14 pathway activity scores. The odds ratio was computed for the cells showing pathway activity 1 versus not showing activity in a given cell type and for the remainder of the data. The 2 heatmap revealed distinct clusters of the most strongly enriched pathways (Figure 4A). The 3 first cluster included the phosphatidylinositol signaling, spliceosome, long term potentiation, 4 and ribosome biogenesis pathways which were enriched in T cells and cytotoxic T cells. In 5 the second cluster, the Fc-epsilon RI signaling pathway, platelet activation, Rap1 signaling 6 pathway, and FoxO signaling pathways were enriched in both classical and non-classical 7 monocytes. Adherens junction and TGF- β signaling pathway which were enriched in 8 multiple cell types, including pulmonary cells, vascular endothelial cells, alveolar cells, and 9 epithelial cells, made up the third major cluster. 10 To identify cell types showing differential expression of emphysema pathways in 11 subjects with severe COPD, we compared the pathway activity scores between controls and 12 subjects with end-stage COPD for each lung cell type (Figure 4B). Alveolar macrophages had 13 multiple highly dysregulated pathways. The Rap1 signaling, adherens junction, gap junction, 14 and FoxO signaling pathways were more active in alveolar macrophages from subjects with 15 COPD compared to controls. The oxidative phosphorylation and spliceosome pathways were 16 less active in alveolar macrophages from COPD than control subjects. Macrophages, alveolar 17 type (AT) II cells, and B cells clustered together, showing similar changes in pathway activity 18 in COPD compared to controls. Cytotoxic T cells, T cells, NK cells, and non-classical 19 monocytes also clustered together and had decreased activity in COPD samples for nearly 20 every pathway, except for ribosome-associated pathways. When comparing the activity of 21 oxidative phosphorylation in macrophages and monocytes, we observed that macrophages, 22 alveolar macrophages, and classical monocytes (M1 macrophages) had decreased activity, 23 while non-classical monocytes (M2 macrophages) had increased activity. 24 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 15 We then specifically compared the pathway activity scores for the TGF- β signaling 1 pathway, adherens junction, and Rap1 signaling pathway between controls and end-stage 2 COPD for each lung cell type (Figure 4C-E). For all three pathways, there was a disruption in 3 multiple cell types, and cell types were split between having increased or decreased activity. 4 Alveolar macrophages, ATI, ATII, and B cells were among the cell types with the greatest 5 increase in activity in the end-stage COPD compared to controls for all three pathways. Non-6 classical monocytes and cytotoxic T cells were among the cell types with the greatest 7 decrease in activity in end-stage COPD versus controls. TGF- β signaling was more active in 8 T cells and goblet cells from end-stage COPD samples, while Rap1 and adherens junction 9 signaling were not significantly altered. A list of abbreviations of the cell types can be found 10 in Table E7 in the supplements. 11 12 Comparison of lung and blood transcriptomics 13 To compare the transcriptomic signature of emphysema in the lung to that in the 14 blood, we obtained data from previously published DGE, DIU, and DEU analyses in whole 15 blood data from COPDGene. There were 247 shared genes among the 1,055 DEGs and 4,913 16 DEGs from the lung and blood, respectively (Figure 5A). Additionally, 25 DUIs and 2 DUEs 17 were shared between lung and blood samples (Figure 5B-C). GSEA was also run in both lung 18 tissue data from LTRC and whole blood data from COPDGene. The full results from 19 COPDGene can be found in Table E8. When comparing the significant pathways from the 20 blood and lung, there were four shared pathways, two of which were biologically relevant: 21 oxidative phosphorylation and ribosomal RNAs and proteins (Figure 5D and Table E8). Both 22 pathways had negative log-fold changes in both lung and blood, indicating that the majority 23 of genes in the respective gene sets were downregulated with emphysema. The beta 24 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 16 coefficients (log-fold changes) of the genes that were significant in either blood or lung were 1 mostly different overall (Figure 5E). 2 3

4 In the present study, we identified dysregulated pathways associated with emphysema, 5 which included oxidative phosphorylation, ribosomal function, and pluripotency TGF- β and 6 FoxO signaling pathways. We also found significant associations with pathways involved in 7 epithelial barrier function in multiple cell types. We additionally discovered cell-type specific 8 enrichments, such as decreased and increased oxidative phosphorylation in the pro-9 inflammatory M1 and anti-inflammatory M2 macrophages, respectively. While there are 10 some shared transcriptomic signals of emphysema in blood and lung tissue, these tissue-11 specific expression profiles are mostly different. 12 While prior reports have evaluated the differential gene expression characteristics of 13 emphysema, the knowledge of the alternative splicing mechanisms underlying this disease is 14 limited. Studies have shown that the SERPINA1 gene, which encodes for the alpha-1 15 antitrypsin protein involved in the pathophysiology of emphysema in individuals with alpha-16 1 antitrypsin deficiency, has significant splicing signatures in this disease (53-55). 17 Additionally, the p53/hypoxia-related genes NUMB and PGFA were significantly associated 18 with alternative splicing signatures in both emphysema and IPF (56). In this transcriptome-19 wide study of alternative splicing, we identified hundreds of isoforms significantly associated 20 with emphysema. Only 7% of the identified isoforms corresponded to significant genes from 21 the gene-level analysis, indicating the DIU analysis did uncover novel transcriptional events 22 and emphysema-related pathways relative to the standard gene differential expression 23 analysis. 24 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 17 NF-κ B is a transcription factor that promotes innate immune and T cell differentiation, 1 suppresses apoptosis, and enhances pro-inflammatory genes (57). Higher amounts of the NF-2 B p65 subunit protein were found in sputum samples and bronchial biopsies of COPD 3 patients compared to controls (58, 59). We add to these findings by revealing significant 4 associations between emphysema and three NF- κ B signaling genes ( PRKCZ, MYBB1A, 5 TRAPPC9) (60-62). Exposure to cigarette smoke was shown to increase the activity of 6 pathways related to protein kinase C in rat bronchial tissues and mouse alveolar epithelial 7 cells (63, 64). We are the first to demonstrate a relationship between emphysema in human 8 lungs and the PRKCZ gene, which generates protein kinase C zeta (65). Reduced MYBBP1A 9 protein levels were previously associated with head and neck squamous cell carcinoma, but 10 no study to date has implicated MYBB1A in emphysema (60, 61). We have previously shown 11 an association of TRAPPC9 with apico-basal emphysema distribution (66), and here we 12 provide more evidence for this association. 13 Our analysis identified particularly strong and widespread changes in emphysema-14 associated pathways in alveolar macrophages. Macrophages exposed to oxidants develop an 15 imbalance in the protease/anti-protease activity, particularly in subjects with emphysema who 16 have alpha-1 antitrypsin deficiency (67-69). However, we are the first to show that anti-17 inflammatory (M2) macrophages (non-classical monocytes) have increased oxidative 18 phosphorylation activity, while pro-inflammatory (M1) macrophages (classical monocytes) 19 have lower activity. M1 macrophages cause tissue damage and block cell division, while M2 20 macrophages have the opposite effect (70). Cornwell et al. have demonstrated elevated M2 21 macrophages in COPD (71) but not specifically in emphysema. Together, these findings can 22 be reflective of a process whereby the pro-inflammatory M1 macrophages shift energy 23 metabolism from oxidative phosphorylation to lactic acidosis, and the M2 macrophages 24 predominate in oxygen-dependent metabolism in COPD and emphysema. One study which 25 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 18 used single-cell RNA-seq data from explanted human lung tissue showed an elevated 1 expression of the anti-inflammatory and antioxidant-related gene HMOX1 in advanced COPD 2 (72). Another recent study found that in severe emphysema, macrophages may be responsible 3 for the destruction of bronchiolar and alveolar tissue, as well as the invasion of tertiary lymph 4 nodes (73). This highlights the potential multifaceted nature of macrophages in emphysema, 5 which may warrant future research into this cell type as it relates to emphysema 6 pathophysiology. 7 Our pathway analysis also revealed enrichment for the pluripotency FoxO signaling 8 and TGF-β signaling pathways. The increased pluripotency function likely preserves cellular 9 homeostasis following lung injury in emphysema, implying a possible role for pluripotent 10 stem cell therapy in emphysema patients (74-76). Both FoxO and TGF- β have been linked to 11 cancer pathogenesis, with the former being a positive mediator of cell growth (77) and 12 vascular remodeling (41) and the latter having both tumor suppressing and promoting 13 functions (78). Studies have shown decreased levels of the FoxO protein in COPD (41, 79-14 81). Di Stefano et al. found that, compared to controls, COPD peripheral airway samples 15 have decreased TGF- β 1+ and TGF- β 3+ bronchial epithelial cells (82). We build on these 16 findings by showing through our lung tissue pathway analysis that TGF- β signaling is 17 dysregulated in emphysema. This finding supports our earlier emphysema integrative 18 genomics and functional investigations, in which we identified functional genetic variations 19 in the TGFB2 and ACVR1B loci in lung fibroblasts and airway epithelial cells (83). The 20 TGFB2 gene encodes a protein isoform of TGF- β . The ACVR1B gene is a transducer of 21 activin-like ligands that are growth and differentiation factors belonging to the TGF- β 22 superfamily of signaling proteins. In another project, we used CRISPR gene editing to 23 discover that the genomic region spanning rs1690789, one of the genetic variants identified 24 by the genome-wide association analysis (GWAS) of emphysema, contains an active 25 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 19 enhancer element that increases the expression of TGFB2 in human lung fibroblasts, 1 providing additional evidence of the link between emphysema and TGF-β (84). 2 With regards to the cell barrier function, damage to the alveolar respiratory epithelium 3 has been shown to be linked to emphysema and COPD (85, 86). We showed that a wide array 4 of cell types is significantly associated with TGF- β signaling, adherens junction, and Rap1 5 signaling pathways, all related to epithelial barrier homeostasis, in emphysema and end-stage 6 COPD. For example, alveolar macrophages, B cells, plasma cells, pulmonary ATI and ATII 7 cells are significantly downregulated, while natural killer and T lymphocytes are significantly 8 upregulated for all three pathways. 9 The current study has a number of strengths. It is the largest study to date to 10 investigate the relationship of emphysema to gene expression and alternative splicing. 11 Pathway enrichment analysis revealed novel biological processes and confirmed the results of 12 previously published COPD and emphysema studies. In addition, by integrating our bulk 13 RNA-seq analysis with another single-cell dataset from subjects with severe COPD, we were 14 able to identify specific lung cell types showing COPD-associated dysregulation of the 15 pathways identified in the bulk tissue analysis. 16 There are also limitations to this study. One limitation is that we did not include a 17 sizable number of LTRC participants in our analyses. Most samples were excluded due to 18 missing CT emphysema values. An additional limitation is that the single-cell dataset we 19 queried was made up of patients with end-stage COPD rather than emphysema specifically. 20 However, emphysema is typically present in end-stage COPD patients (87). Future research 21 should examine enriched pathways using bulk and single-cell RNA-seq data from the same 22 cohorts to shed further light on the various cell types implicated in emphysema pathobiology. 23 Lastly, the large sample size for our primary analysis reduces the risk of false positive 24 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 20 associations, but further validation of these results in comparable cohorts will provide greater 1 confidence in these associations. 2 3

4 CT-quantified emphysema exhibits distinctive alternative splicing and transcriptomic 5 patterns, as well as associations to biological processes connected to oxidative 6 phosphorylation and pluripotency pathways. In addition to other cell-type-specific markers, 7 the M1/M2 macrophage ratio was found to be decreased in relation to oxidative 8 phosphorylation, and a number of innate and adaptive immune cell types are enriched for 9 epithelial barrier function. These novel molecular and cellular findings in emphysema may 10 help in the development of effective screening methods and therapeutic strategies. 11 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 21

1 1. Lindberg A, Lindberg L, Sawalha S, Nilsson U, Stridsman C, Lundbäck B, Backman H. 2 Large underreporting of COPD as cause of death-results from a population-based 3 cohort study. Respir Med 2021; 186: 106518. 4 2. Sharafkhaneh A, Hanania NA, Kim V. Pathogenesis of emphysema: from the bench to the 5 bedside. Proc Am Thorac Soc 2008; 5: 475-477. 6 3. Mohamed Hoesein FAA, Zanen P, Boezen HM, Groen HJM, van Ginneken B, de Jong PA, 7 Postma DS, Lammers JJ. Lung function decline in male heavy smokers relates to 8 baseline airflow obstruction severity. Chest 2012; 142: 1530-1538. 9 4. Carolan BJ, Hughes G, Morrow J, Hersh CP, O'Neal WK, Rennard S, Pillai SG, Belloni P, 10 Cockayne DA, Comellas AP, Han M, Zemans RL, Kechris K, Bowler RP. The 11 association of plasma biomarkers with computed tomography-assessed emphysema 12 phenotypes. Respir Res 2014; 15: 127. 13 5. Zuo Q, Wang Y, Yang D, Guo S, Li X, Dong J, Wan C, Shen Y, Wen F. Identification of 14 hub genes and key pathways in the emphysema phenotype of COPD. Aging (Albany 15 NY) 2021; 13: 5120-5135. 16 6. Xu L, Bian W, Gu XH, Shen C. Genetic polymorphism in matrix metalloproteinase-9 and 17 transforming growth factor-β 1 and susceptibility to combined pulmonary fibrosis and 18 emphysema in a Chinese population. Kaohsiung J Med Sci 2017; 33: 124-129. 19 7. Faner R, Cruz T, Casserras T, López-Giraldo A, Noell G, Coca I, Tal-Singer R, Miller B, 20 Rodriguez-Roisin R, Spira A, Kalko SG, Agustí A. Network Analysis of Lung 21 Transcriptomics Reveals a Distinct B-Cell Signature in Emphysema. Am J Respir Crit 22 Care Med 2016; 193: 1242-1253. 23 8. Obeidat M, Nie Y, Fishbane N, Li X, Bossé Y, Joubert P, Nickle DC, Hao K, Postma DS, 24 Timens W, Sze MA, Shannon CP, Hollander Z, Ng RT, McManus B, Miller BE, 25 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 22 Rennard S, Spira A, Hackett TL, Lam W, Lam S, Faner R, Agusti A, Hogg JC, Sin 1 DD, Paré PD. Integrative Genomics of Emphysema-Associated Genes Reveals 2 Potential Disease Biomarkers. Am J Respir Cell Mol Biol 2017; 57: 411-418. 3 9. Yoo S, Takikawa S, Geraghty P, Argmann C, Campbell J, Lin L, Huang T, Tu Z, Foronjy 4 RF, Spira A, Schadt EE, Powell CA, Zhu J. Integrative analysis of DNA methylation 5 and gene expression data identifies EPAS1 as a key regulator of COPD. PLoS Genet 6 2015; 11: e1004898. 7 10. Yasuo M, Mizuno S, Kraskauskas D, Bogaard HJ, Natarajan R, Cool CD, Zamora M, 8 Voelkel NF. Hypoxia inducible factor-1 α in human emphysema lung tissue. Eur 9 Respir J 2011; 37: 775-783. 10 11. Tan LH, Bahmed K, Lin CR, Marchetti N, Bolla S, Criner GJ, Kelsen S, Madesh M, 11 Kosmider B. The cytoprotective role of DJ-1 and p45 NFE2 against human primary 12 alveolar type II cell injury and emphysema. Sci Rep 2018; 8: 3555. 13 12. Bahmed K, Boukhenouna S, Karim L, Andrews T, Lin J, Powers R, Wilson MA, Lin CR, 14 Messier E, Reisdorph N, Powell RL, Tang HY, Mason RJ, Criner GJ, Kosmider B. 15 The effect of cysteine oxidation on DJ-1 cytoprotective function in human alveolar 16 type II cells. Cell Death Dis 2019; 10: 638. 17 13. Baker JR, Donnelly LE, Barnes PJ. Senotherapy: A New Horizon for COPD Therapy. 18 Chest 2020; 158: 562-570. 19 14. Tsuji T, Aoshiba K, Nagai A. Alveolar cell senescence in patients with pulmonary 20 emphysema. Am J Respir Crit Care Med 2006; 174: 886-893. 21 15. Jasper AE, McIver WJ, Sapey E, Walton GM. Understanding the role of neutrophils in 22 chronic inflammatory airway disease. F1000Res 2019; 8. 23 16. Wang Y, Xu J, Meng Y, Adcock IM, Yao X. Role of inflammatory cells in airway 24 remodeling in COPD. Int J Chron Obstruct Pulmon Dis 2018; 13: 3341-3348. 25 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 23 17. Lee SH, Goswami S, Grudo A, Song LZ, Bandi V, Goodnight-White S, Green L, 1 Hacken-Bitar J, Huh J, Bakaeen F, Coxson HO, Cogswell S, Storness-Bliss C, Corry 2 DB, Kheradmand F. Antielastin autoimmunity in tobacco smoking-induced 3 emphysema. Nat Med 2007; 13: 567-569. 4 18. Majo J, Ghezzo H, Cosio MG. Lymphocyte population and apoptosis in the lungs of 5 smokers and their relation to emphysema. Eur Respir J 2001; 17: 946-953. 6 19. Yang IV, Pedersen BS, Rabinovich E, Hennessy CE, Davidson EJ, Murphy E, Guardela 7 BJ, Tedrow JR, Zhang Y, Singh MK, Correll M, Schwarz MI, Geraci M, Sciurba FC, 8 Quackenbush J, Spira A, Kaminski N, Schwartz DA. Relationship of DNA 9 methylation and gene expression in idiopathic pulmonary fibrosis. Am J Respir Crit 10 Care Med 2014; 190: 1263-1272. 11 20. Vestbo J, Hurd SS, Agustí AG, Jones PW, Vogelmeier C, Anzueto A, Barnes PJ, Fabbri 12 LM, Martinez FJ, Nishimura M, Stockley RA, Sin DD, Rodriguez-Roisin R. Global 13 strategy for the diagnosis, management, and prevention of chronic obstructive 14 pulmonary disease: GOLD executive summary. Am J Respir Crit Care Med 2013; 15 187: 347-365. 16 21. Raghu G, Collard HR, Egan JJ, Martinez FJ, Behr J, Brown KK, Colby TV, Cordier JF, 17 Flaherty KR, Lasky JA, Lynch DA, Ryu JH, Swigris JJ, Wells AU, Ancochea J, 18 Bouros D, Carvalho C, Costabel U, Ebina M, Hansell DM, Johkoh T, Kim DS, King 19 TE, Jr., Kondoh Y, Myers J, Muller NL, Nicholson AG, Richeldi L, Selman M, 20 Dudden RF, Griss BS, Protzko SL, Schunemann HJ, Fibrosis AEJACoIP. An official 21 ATS/ERS/JRS/ALAT statement: idiopathic pulmonary fibrosis: evidence-based 22 guidelines for diagnosis and management. Am J Respir Crit Care Med 2011; 183: 23 788-824. 24 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 24 22. Regan EA, Hokanson JE, Murphy JR, Make B, Lynch DA, Beaty TH, Curran-Everett D, 1 Silverman EK, Crapo JD. Genetic epidemiology of COPD (COPDGene) study design. 2 COPD 2010; 7: 32-43. 3 23. Karwoski R, Bartholmai B, Zavaletta V, Holmes Iii D, Robb R. Processing of CT images 4 for analysis of diffuse lung disease in the lung tissue research consortium. Progress in 5 Biomedical Optics and Imaging - Proceedings of SPIE 2008; 6916. 6 24. Quanjer PH, Stanojevic S, Cole TJ, Baur X, Hall GL, Culver BH, Enright PL, Hankinson 7 JL, Ip MS, Zheng J, Stocks J, Initiative ERSGLF. Multi-ethnic reference values for 8 spirometry for the 3-95-yr age range: the global lung function 2012 equations. Eur 9 Respir J 2012; 40: 1324-1343. 10 25. Parr DG, Sevenoaks M, Deng C, Stoel BC, Stockley RA. Detection of emphysema 11 progression in alpha 1-antitrypsin deficiency using CT densitometry; methodological 12 advances. Respir Res 2008; 9: 21. 13 26. Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, Smyth GK. limma powers 14 differential expression analyses for RNA-sequencing and microarray studies. Nucleic 15 Acids Res 2015; 43: e47. 16 27. Law CW, Chen Y, Shi W, Smyth GK. voom: Precision weights unlock linear model 17 analysis tools for RNA-seq read counts. Genome Biol 2014; 15: R29. 18 28. Alhamdoosh M, Ng M, Wilson NJ, Sheridan JM, Huynh H, Wilson MJ, Ritchie ME. 19 Combining multiple tools outperforms individual methods in gene set enrichment 20 analyses. Bioinformatics 2017; 33: 414-424. 21 29. Kanehisa M, Goto S. KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids 22 Res 2000; 28: 27-30. 23 30. Wu D, Smyth GK. Camera: a competitive gene set test accounting for inter-gene 24 correlation. Nucleic Acids Res 2012; 40: e133. 25 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 25 31. Adams TS, Schupp JC, Poli S, Ayaub EA, Neumark N, Ahangari F, Chu SG, Raby BA, 1 DeIuliis G, Januszyk M, Duan Q, Arnett HA, Siddiqui A, Washko GR, Homer R, Yan 2 X, Rosas IO, Kaminski N. Single-cell RNA-seq reveals ectopic and aberrant lung-3 resident cell populations in idiopathic pulmonary fibrosis. Sci Adv 2020; 6: eaba1983. 4 32. Hao Y, Hao S, Andersen-Nissen E, Mauck WM, 3rd, Zheng S, Butler A, Lee MJ, Wilk 5 AJ, Darby C, Zager M, Hoffman P, Stoeckius M, Papalexi E, Mimitou EP, Jain J, 6 Srivastava A, Stuart T, Fleming LM, Yeung B, Rogers AJ, McElrath JM, Blish CA, 7 Gottardo R, Smibert P, Satija R. Integrated analysis of multimodal single-cell data. 8 Cell 2021; 184: 3573-3587 e3529. 9 33. Welch BL. The generalisation of student's problems when several different population 10 variances are involved. Biometrika 1947; 34: 28-35. 11 34. Fisher RA. On the Interpretation of χ 2 from Contingency Tables, and the Calculation of P. 12 Journal of the Royal Statistical Society 1922; 85: 87-94. 13 35. Gu Z, Eils R, Schlesner M. Complex heatmaps reveal patterns and correlations in 14 multidimensional genomic data. Bioinformatics 2016; 32: 2847-2849. 15 36. Benjamini Y, Hochberg Y. Controlling the False Discovery Rate: A Practical and 16 Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society: 17 Series B (Methodological) 1995; 57: 289-300. 18 37. Miyashita N, Horie M, Suzuki HI, Saito M, Mikami Y, Okuda K, Boucher RC, Suzukawa 19 M, Hebisawa A, Saito A, Nagase T. FOXL1 Regulates Lung Fibroblast Function via 20 Multiple Mechanisms. American journal of respiratory cell and molecular biology 21 2020; 63: 831-842. 22 38. Wygrecka M, Didiasova M, Berscheid S, Piskulak K, Taborski B, Zakrzewicz D, 23 Kwapiszewska G, Preissner KT, Markart P. Protease-activated receptors (PAR)-1 and 24 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 26 -3 drive epithelial-mesenchymal transition of alveolar epithelial cells - potential role 1 in lung fibrosis. Thromb Haemost 2013; 110: 295-307. 2 39. Hersh CP, Hansel NN, Barnes KC, Lomas DA, Pillai SG, Coxson HO, Mathias RA, 3 Rafaels NM, Wise RA, Connett JE, Klanderman BJ, Jacobson FL, Gill R, Litonjua 4 AA, Sparrow D, Reilly JJ, Silverman EK. Transforming growth factor-beta receptor-3 5 is associated with pulmonary emphysema. Am J Respir Cell Mol Biol 2009; 41: 324-6 331. 7 40. Koli K, Myllärniemi M, Keski-Oja J, Kinnula VL. Transforming growth factor-beta 8 activation in the lung: focus on fibrosis and reactive oxygen species. Antioxid Redox 9 Signal 2008; 10: 333-342. 10 41. Hwang JW, Rajendrasozhan S, Yao H, Chung S, Sundar IK, Huyck HL, Pryhuber GS, 11 Kinnula VL, Rahman I. FOXO3 deficiency leads to increased susceptibility to 12 cigarette smoke-induced inflammation, airspace enlargement, and chronic obstructive 13 pulmonary disease. J Immunol 2011; 187: 987-998. 14 42. Nishida K, Brune KA, Putcha N, Mandke P, O'Neal WK, Shade D, Srivastava V, Wang 15 M, Lam H, An SS, Drummond MB, Hansel NN, Robinson DN, Sidhaye VK. 16 Cigarette smoke disrupts monolayer integrity by altering epithelial cell-cell adhesion 17 and cortical tension. Am J Physiol Lung Cell Mol Physiol 2017; 313: L581-l591. 18 43. Ye R, Wang C, Sun P, Bai S, Zhao L. AGR3 Regulates Airway Epithelial Junctions in 19 Patients with Frequent Exacerbations of COPD. Front Pharmacol 2021; 12: 669403. 20 44. Eriksson Ström J, Kebede Merid S, Pourazar J, Blomberg A, Lindberg A, Ringh MV, 21 Hagemann-Jensen M, Ekström TJ, Behndig AF, Melén E. Chronic Obstructive 22 Pulmonary Disease Is Associated with Epigenome-Wide Differential Methylation in 23 BAL Lung Cells. Am J Respir Cell Mol Biol 2022; 66: 638-647. 24 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 27 45. Owen HR, Elser M, Cheung E, Gersbach M, Kraus WL, Hottiger MO. MYBBP1a is a 1 novel repressor of NF-kappaB. J Mol Biol 2007; 366: 725-736. 2 46. Diaz-Meco MT, Moscat J. The atypical PKCs in inflammation: NF-kappaB and beyond. 3 Immunol Rev 2012; 246: 154-167. 4 47. Berditchevski F. Complexes of tetraspanins with integrins: more than meets the eye. J 5 Cell Sci 2001; 114: 4143-4151. 6 48. Parast MM, Otey CA. Characterization of palladin, a novel protein localized to stress 7 fibers and cell adhesions. J Cell Biol 2000; 150: 643-656. 8 49. Hu WH, Pendergast JS, Mo XM, Brambilla R, Bracchi-Ricard V, Li F, Walters WM, 9 Blits B, He L, Schaal SM, Bethea JR. NIBP, a novel NIK and IKK(beta)-binding 10 protein that enhances NF-(kappa)B activation. J Biol Chem 2005; 280: 29233-29241. 11 50. Cozier GE, Carlton J, McGregor AH, Gleeson PA, Teasdale RD, Mellor H, Cullen PJ. 12 The phox homology (PX) domain-dependent, 3-phosphoinositide-mediated 13 association of sorting nexin-1 with an early sorting endosomal compartment is 14 required for its ability to regulate epidermal growth factor receptor degradation. J Biol 15 Chem 2002; 277: 48730-48736. 16 51. Aoki T, Ichimura S, Itoh A, Kuramoto M, Shinkawa T, Isobe T, Tagaya M. Identification 17 of the neuroblastoma-amplified gene product as a component of the syntaxin 18 18 complex implicated in Golgi-to-endoplasmic reticulum retrograde transport. Mol Biol 19 Cell 2009; 20: 2639-2649. 20 52. Ungar D, Oka T, Brittle EE, Vasile E, Lupashin VV, Chatterton JE, Heuser JE, Krieger M, 21 Waters MG. Characterization of a mammalian Golgi-localized protein complex, COG, 22 that is required for normal Golgi morphology and function. J Cell Biol 2002; 157: 23 405-415. 24 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 28 53. Matamala N, Martínez MT, Lara B, Pérez L, Vázquez I, Jimenez A, Barquín M, 1 Ferrarotti I, Blanco I, Janciauskiene S, Martinez-Delgado B. Alternative transcripts of 2 the SERPINA1 gene in alpha-1 antitrypsin deficiency. J Transl Med 2015; 13: 211. 3 54. Corley M, Solem A, Phillips G, Lackey L, Ziehr B, Vincent HA, Mustoe AM, Ramos 4 SBV, Weeks KM, Moorman NJ, Laederach A. An RNA structure-mediated, 5 posttranscriptional model of human α -1-antitrypsin expression. Proc Natl Acad Sci U 6 S A 2017; 114: E10244-e10253. 7 55. Seixas S, Mendonça C, Costa F, Rocha J. alpha1-Antitrypsin null alleles: evidence for the 8 recurrence of the L353fsX376 mutation and a novel G-->A transition in position +1 of 9 intron IC affecting normal mRNA splicing. Clin Genet 2002; 62: 175-180. 10 56. Kusko RL, Brothers JF, 2nd, Tedrow J, Pandit K, Huleihel L, Perdomo C, Liu G, Juan-11 Guardela B, Kass D, Zhang S, Lenburg M, Martinez F, Quackenbush J, Sciurba F, 12 Limper A, Geraci M, Yang I, Schwartz DA, Beane J, Spira A, Kaminski N. Integrated 13 Genomics Reveals Convergent Transcriptomic Networks Underlying Chronic 14 Obstructive Pulmonary Disease and Idiopathic Pulmonary Fibrosis. Am J Respir Crit 15 Care Med 2016; 194: 948-960. 16 57. Liu T, Zhang L, Joo D, Sun SC. NF- κ B signaling in inflammation. Signal Transduct 17 Target Ther 2017; 2: 17023-. 18 58. Brown V, Elborn JS, Bradley J, Ennis M. Dysregulated apoptosis and NFkappaB 19 expression in COPD subjects. Respir Res 2009; 10: 24. 20 59. Di Stefano A, Caramori G, Oates T, Capelli A, Lusuardi M, Gnemmi I, Ioli F, Chung KF, 21 Donner CF, Barnes PJ, Adcock IM. Increased expression of nuclear factor-kappaB in 22 bronchial biopsies from smokers and patients with COPD. Eur Respir J 2002; 20: 23 556-563. 24 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 29 60. Acuña Sanhueza GA, Faller L, George B, Koffler J, Misetic V, Flechtenmacher C, 1 Dyckhoff G, Plinkert PP, Angel P, Simon C, Hess J. Opposing function of 2 MYBBP1A in proliferation and migration of head and neck squamous cell carcinoma 3 cells. BMC Cancer 2012; 12: 72. 4 61. George B, Horn D, Bayo P, Zaoui K, Flechtenmacher C, Grabe N, Plinkert P, 5 Krizhanovsky V, Hess J. Regulation and function of Myb-binding protein 1A 6 (MYBBP1A) in cellular senescence and pathogenesis of head and neck cancer. 7 Cancer Lett 2015; 358: 191-199. 8 62. Mbimba T, Hussein NJ, Najeed A, Safadi FF. TRAPPC9: Novel insights into its 9 trafficking and signaling pathways in health and disease (Review). Int J Mol Med 10 2018; 42: 2991-2997. 11 63. Hou HH, Cheng SL, Chung KP, Wei SC, Tsao PN, Lu HH, Wang HC, Yu CJ. PlGF 12 mediates neutrophil elastase-induced airway epithelial cell apoptosis and emphysema. 13 Respir Res 2014; 15: 106. 14 64. Chang WC, Lee YC, Liu CL, Hsu JD, Wang HC, Chen CC, Wang CJ. Increased 15 expression of iNOS and c-fos via regulation of protein tyrosine phosphorylation and 16 MEK1/ERK2 proteins in terminal bronchiole lesions in the lungs of rats exposed to 17 cigarette smoke. Arch Toxicol 2001; 75: 28-35. 18 65. Seto KK, Andrulis IL. Atypical protein kinase C zeta: potential player in cell survival and 19 cell migration of ovarian cancer. PLoS One 2015; 10: e0123528. 20 66. Boueiz A, Lutz SM, Cho MH, Hersh CP, Bowler RP, Washko GR, Halper-Stromberg E, 21 Bakke P, Gulsvik A, Laird NM, Beaty TH, Coxson HO, Crapo JD, Silverman EK, 22 Castaldi PJ, DeMeo DL. Genome-Wide Association Study of the Genetic 23 Determinants of Emphysema Distribution. Am J Respir Crit Care Med 2017; 195: 24 757-771. 25 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 30 67. Barnes PJ, Shapiro SD, Pauwels RA. Chronic obstructive pulmonary disease: molecular 1 and cellular mechanisms. Eur Respir J 2003; 22: 672-688. 2 68. Cantin A, Crystal RG. Oxidants, antioxidants and the pathogenesis of emphysema. Eur J 3 Respir Dis Suppl 1985; 139: 7-17. 4 69. Belchamber KBR, Walker EM, Stockley RA, Sapey E. Monocytes and Macrophages in 5 Alpha-1 Antitrypsin Deficiency. Int J Chron Obstruct Pulmon Dis 2020; 15: 3183-6 3192. 7 70. Mills CD. M1 and M2 Macrophages: Oracles of Health and Disease. Crit Rev Immunol 8 2012; 32: 463-488. 9 71. Cornwell WD, Kim V, Fan X, Vega ME, Ramsey FV, Criner GJ, Rogers TJ. Activation 10 and polarization of circulating monocytes in severe chronic obstructive pulmonary 11 disease. BMC Pulm Med 2018; 18: 101. 12 72. Sauler M, McDonough JE, Adams TS, Kothapalli N, Barnthaler T, Werder RB, Schupp 13 JC, Nouws J, Robertson MJ, Coarfa C, Yang T, Chioccioli M, Omote N, Cosme C, Jr., 14 Poli S, Ayaub EA, Chu SG, Jensen KH, Gomez JL, Britto CJ, Raredon MSB, 15 Niklason LE, Wilson AA, Timshel PN, Kaminski N, Rosas IO. Characterization of 16 the COPD alveolar niche using single-cell RNA sequencing. Nat Commun 2022; 13: 17 494. 18 73. Suzuki M, Sze MA, Campbell JD, Brothers JF, 2nd, Lenburg ME, McDonough JE, Elliott 19 WM, Cooper JD, Spira A, Hogg JC. The cellular and molecular determinants of 20 emphysematous destruction in COPD. Sci Rep 2017; 7: 9562. 21 74. Sueblinvong V, Weiss DJ. Stem cells and cell therapy approaches in lung biology and 22 diseases. Transl Res 2010; 156: 188-205. 23 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 31 75. Prakash YS, Tschumperlin DJ, Stenmark KR. Coming to terms with tissue engineering 1 and regenerative medicine in the lung. Am J Physiol Lung Cell Mol Physiol 2015; 309: 2 L625-638. 3 76. Torres-Durán M, Lopez-Campos JL, Barrecheguren M, Miravitlles M, Martinez-Delgado 4 B, Castillo S, Escribano A, Baloira A, Navarro-Garcia MM, Pellicer D, Bañuls L, 5 Magallón M, Casas F, Dasí F. Alpha-1 antitrypsin deficiency: outstanding questions 6 and future directions. Orphanet J Rare Dis 2018; 13: 114. 7 77. Jiramongkol Y, Lam EW. FOXO transcription factor family in cancer and metastasis. 8 Cancer Metastasis Rev 2020; 39: 681-709. 9 78. Katz LH, Li Y, Chen JS, Muñoz NM, Maju mdar A, Chen J, Mishra L. Targeting TGF- β 10 signaling in cancer. Expert Opin Ther Targets 2013; 17: 743-760. 11 79. Mumby S, Adcock IM. Recent evidence from omic analysis for redox signalling and 12 mitochondrial oxidative stress in COPD. J Inflamm (Lond) 2022; 19: 10. 13 80. Huang HH, Liang Y. Integrating molecular interactions and gene expression to identify 14 biomarkers and network modules of chronic obstructive pulmonary disease. Technol 15 Health Care 2022; 30: 135-142. 16 81. Barnes PJ. Oxidative stress-based therapeutics in COPD. Redox Biol 2020; 33: 101544. 17 82. Di Stefano A, Sangiorgi C, Gnemmi I, Casolari P, Brun P, Ricciardolo FLM, Contoli M, 18 Papi A, Maniscalco P, Ruggeri P, Girbino G, Cappello F, Pavlides S, Guo Y, Chung 19 KF, Barnes PJ, Adcock IM, Balbi B, Caramori G. TGF- β Signaling Pathways in 20 Different Compartments of the Lower Airways of Patients With Stable COPD. Chest 21 2018; 153: 851-862. 22 83. Boueiz A, Pham B, Chase R, Lamb A, Lee S, Naing ZZC, Cho MH, Parker MM, 23 Sakornsakolpat P, Hersh CP, Crapo JD, Stergachis AB, Tal-Singer R, DeMeo DL, 24 Silverman EK, Zhou X, Castaldi PJ. Integrative Genomics Analysis Identifies 25 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 32 ACVR1B as a Candidate Causal Gene of Emphysema Distribution. Am J Respir Cell 1 Mol Biol 2019; 60: 388-398. 2 84. Parker MM, Hao Y, Guo F, Pham B, Chase R, Platig J, Cho MH, Hersh CP, Thannickal 3 VJ, Crapo J, Washko G, Randell SH, Silverman EK, San José Estépar R, Zhou X, 4 Castaldi PJ. Identification of an emphysema-associated genetic variant near TGFB2 5 with regulatory effects in lung fibroblasts. Elife 2019; 8. 6 85. Carlier FM, de Fays C, Pilette C. Epithelial Barrier Dysfunction in Chronic Respiratory 7 Diseases. Front Physiol 2021; 12: 691227. 8 86. Hadzic S, Wu CY, Avdeev S, Weissmann N, Schermuly RT, Kosanovic D. Lung 9 epithelium damage in COPD - An unstoppable pathological event? Cell Signal 2020; 10 68: 109540. 11 87. Devine JF. Chronic obstructive pulmonary disease: an overview. Am Health Drug 12 Benefits 2008; 1: 34-42. 13 14 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 33 Table 1. Baseline characteristics of the included Lung Tissue Research Consortium (LTRC) subjects. Characteristics Mean (SD) or n (%) n 456 Age (years) 64 (11.2) Sex, %male 241 (52.9%) Race Non-Hispanic white 397 (87.1%) African American 46 (10.1%) Hispanic 11 (2.4%) Other 2 (0.4%) BMI (kg/m2) 28.1 (6.0) Current smoker 37 (8.1%) Smoking pack-years 36 (36.4) FEV1, % predicted 71 (25.9) GOLD grade PRISm 59 (12.9%) GOLD 0 172 (37.7%) GOLD 1 56 (12.3%) GOLD 2 88 (19.3%) GOLD 3 45 (9.9%) GOLD 4 36 (7.9%) Total Perc15 density -918 (45.3) Data presented as mean (standard deviation (SD)) or number (%). BMI: Body mass index. FEV1: Forced expiratory volume in 1 second. GOLD: Global Initiative for Chronic Obstructive Lung Disease. GOLD 0: Normal spirometry (defined as post-bronchodilator FEV 1/FVC ≥ 0.7 and FEV1 ≥ 80% predicted); GOLD 1: FEV1/FVC < 0.70 and post-bronchodilator FEV1 ≥ 80% predicted; GOLD 2: FEV1/FVC < 0.70 and post-bronchodilator FEV1 50-79% predicted; GOLD 3: FEV1/FVC < 0.70 and post-bronchodilator FEV1 30-49% predicted; GOLD 4: FEV1/FVC < 0.70 and post-bronchodilator FEV1 < 30% predicted. PRISm: Preserved ratio impaired spirometry. Total Perc15 density: Hounsfield units at the 15 th percentile of CT density histogram at total lung capacity. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 34 Table 2. Top 20 differentially expressed or used features in lung tissue. logFC = log fold change (expression/usage change per Hounsfield unit); FDR = false discovery rate. Differentially Expressed Genes Ensembl Gene ID Gene logFC Mean Log Expression FDR ENSG00000176678 FOXL1 0.005 1.75 0.003 ENSG00000120332 TNN 0.006 0.85 0.003 ENSG00000164220 F2RL2 0.008 0.21 0.003 ENSG00000225972 MTND1P23 0.015 -0.38 0.005 ENSG00000224877 NDUFAF8 0.003 1.96 0.014 ENSG00000155313 USP25 -0.002 6.02 0.014 ENSG00000196549 MME -0.006 6.52 0.016 ENSG00000156052 GNAQ -0.002 7.31 0.016 ENSG00000137819 PAQR5 -0.004 4.60 0.016 ENSG00000243509 TNFRSF6B 0.013 -1.36 0.016 ENSG00000173706 HEG1 -0.004 8.49 0.016 ENSG00000120162 MOB3B -0.002 5.01 0.016 ENSG00000226950 DANCR 0.002 3.54 0.016 ENSG00000012061 ERCC1 0.001 5.52 0.016 ENSG00000138669 PRKG2 -0.005 2.86 0.016 ENSG00000106628 POLD2 0.002 5.17 0.016 ENSG00000102763 VWA8 -0.001 4.87 0.017 ENSG00000242114 MTFP1 0.003 2.59 0.017 ENSG00000094963 FMO2 -0.004 8.36 0.017 ENSG00000171421 MRPL36 0.002 2.92 0.018 Differentially Used Isoforms Ensembl Transcript ID Gene logFC Mean Log Expression FDR ENST00000589296 CYTH1 0.023 -3.49 7.74 x 10 -17 ENST00000523282 DNPEP 0.019 -1.92 2.10 x 10 -13 ENST00000441627 ZMIZ2 0.022 -1.47 1.32 x 10 -10 ENST00000546939 CD63 0.022 -2.75 1.03 x 10 -8 ENST00000507699 PALLD 0.020 -1.92 1.03 x 10 -8 ENST00000528996 SERPING1 0.009 1.86 1.09 x 10 -8 ENST00000537526 USP22 0.026 -0.79 1.82 x 10 -8 ENST00000373266 KIAA0319L 0.018 -3.21 1.82 x 10 -8 ENST00000479806 DYNC1I2 0.013 0.08 9.06 x 10 -8 ENST00000566130 ALDOA 0.017 -1.26 9.06 x 10 -8 ENST00000571368 MYBBP1A 0.017 -1.51 2.92 x 10 -7 ENST00000243562 LTBP4 0.016 -0.92 4.10 x 10 -7 ENST00000479263 PRKCZ 0.022 -3.32 4.10 x 10 -7 ENST00000681161 P4HB 0.012 -0.43 1.18 x 10 -6 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 35 ENST00000679119 BAG6 0.013 -1.47 1.64 x 10 -6 ENST00000265637 PPP6R3 0.014 -0.53 2.47 x 10 -6 ENST00000429192 ELN 0.017 -0.49 2.60 x 10 -6 ENST00000587311 LGALS3BP 0.021 -3.09 3.08 x 10 -6 ENST00000645753 IFNGR1 0.015 -1.64 9.14 x 10 -6 ENST00000311172 FCHSD2 0.016 -0.91 1.16 x 10 -5 Differentially Used Exons Chromosome Strand Left Right Gene logFC Mean Log Expression FDR chr16 + 72112640 72112650 DHX38 0.002 0.61 0.001 chr16 + 72112651 72112720 DHX38 0.002 1.12 0.002 chr15 + 64138098 64139875 SNX1 -0.002 2.67 0.002 chr8 - 139730345 139730891 TRAPPC9 0.002 0.67 0.002 chr2 - 15166917 15167323 NBAS 0.002 1.34 0.002 chr6 - 31639227 31639499 BAG6 0.002 -0.07 0.002 chr11 + 67286373 67286392 GRK2 0.003 -0.05 0.002 chr16 + 72112625 72112639 DHX38 0.002 0.69 0.003 chr5 - 64718148 64724431 SREK1IP1 -0.002 3.49 0.004 chrX - 154348648 154348690 FLNA 0.004 2.24 0.004 chr9 - 111361886 111362053 ECPAS 0.002 1.46 0.004 chr10 - 109864766 109865053 XPNPEP1 0.002 1.02 0.004 chr11 - 61299470 61299555 DDB1 0.002 1.55 0.004 chr7 - 1470277 1470477 INTS1 0.003 -0.02 0.004 chr16 - 27460676 27460980 GTF3C1 0.002 1.57 0.006 chrX - 154348535 154348647 FLNA 0.004 1.81 0.006 chr17 + 7514180 7514616 POLR2A 0.003 2.58 0.006 chr3 + 184309001 184309048 PSMD2 0.002 0.17 0.006 chr16 + 72112721 72112903 DHX38 0.002 0.87 0.009 chr1 + 230693626 230693981 COG2 0.002 0.21 0.009 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 36 Table 3. Select significant pathways (FDR 10%) from the gene set enrichment analyses results using differentially expressed genes from lung tissue. Term Name Number of genes annotated Average logFC FDR Oxidative phosphorylation 125 0.001 1.18 x 10-4 Signaling pathways regulating pluripotency of stem cells 110 -0.002 2.36 x 10-2 Adherens junction 70 -0.002 5.30 x 10-2 FoxO signaling pathway 120 -0.001 6.03 x 10-2 Regulation of actin cytoskeleton 183 -0.002 6.73 x 10-2 Jak-STAT signaling pathway 109 -0.002 7.57 x 10-2 Rap1 signaling pathway 186 -0.003 7.58 x 10-2 Fc epsilon RI signaling pathway 62 -0.001 7.58 x 10-2 TGF-beta signaling pathway 76 -0.002 8.61 x 10-2 Gap junction 75 -0.003 9.41 x 10-2 Upregulated versus downregulated pathways are provided with respect to their relationships with Perc15 density which is opposite to emphysema. A negative logFC value indicates a pathway upregulated with emphysema, and a positive logFC value indicates a pathway downregulated with emphysema. Pathways were reported based on relevance to lung disease. Average logFC = mean log fold change of genes in the pathway (gene expression change per Hounsfield unit); FDR = false discovery rate. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 37 FIGURE LEGENDS Figure 1 . Study flow diagram. Abbreviations: BMI: Body mass index. FEV 1: Forced expiratory volume in 1 second. IPF: Idiopathic pulmonary fibrosis. LTRC: Lung Tissue Research Consortium. QC: Quality control. RNA-seq: RNA sequencing. Total Perc15 density: Hounsfield units at the 15 th percentile of CT density histogram at total lung capacity. Figure 2 . Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway maps of the ( A) TGF-β signaling pathway, (B) adherens junction, (C) Rap1 signaling pathway, and (D) FoxO signaling pathway reporting the effect size (log fold change) of all genes within each pathway in the 456 lung tissue samples from subjects in the Lung Tissue Research Consortium (LTRC). Emphysema was quantified by Hounsfield units at the 15 th percentile of chest CT density histogram at full inspiration (Perc15). The lower the Perc15 values are, i.e. the closer to -1,000 HU, the more CT-quantified emphysema is present. Gene log fold changes were multiplied by -100 so that positive log fold changes represented upregulated genes and negative log fold changes represented downregulated genes. Red represents upregulated genes and blue represents downregulated genes. Figure 3 . Venn diagram of the number of significant emphysema-associated genes from the differential gene expression, differential isoform usage, and differential exon usage models from lung tissue. Figure 4 . Heatmaps of ( A) pathway specificity across lung cell types and ( B) pathway T statistic values of COPD versus control lung tissue samples. T statistic values of COPD versus control lung tissue samples for ( C) TGF-β signaling pathway, ( D) adherens junction, All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 38 and (E) Rap1 signaling pathway. Pathway activity scores were generated for each cell in the single-cell dataset. The odds ratio was computed for the cells showing the pathway activity versus not in a given cell type and for the remainder of the data. The P-values were calculated using Fisher’s Exact Test. Additionally, the Welch Two Sample t-test was used to compare the pathway activity scores between controls and end-stage COPD for each cell type separately. A negative T statistic (red) indicates that the pathway has higher activity in COPD compared to controls. A positive T statistic (blue) indicates that the pathway has lower activity in COPD compared to controls. A star indicates that there was a significant difference in activity in COPD versus controls. Figure 5 . Number of emphysema-associated features shared between blood and lung. ( A) Differentially expressed genes. ( B) Differentially used isoforms. ( C) Differentially used exons. ( D) Dysregulated pathways. ( E) Log fold change values of differentially expressed genes associated with emphysema in blood and lung tissue. Genes were included if they were significant in blood, lung tissue, or both (FDR 10%). Emphysema was quantified by Hounsfield units at the 15 th percentile of chest CT density histogram at full inspiration (Perc15). The lower the Perc15 values are, i.e. the closer to -1.000 HU, the more CT- quantified emphysema is present. Upregulated versus downregulated pathways are provided with respect to their relationships with Perc15 density which is opposite to emphysema. Negative log fold change values represent upregulated genes and positive log fold change values represent downregulated genes. When comparing the significant pathways from the blood and lung, there were four shared pathways, two of which were biologically relevant: oxidative phosphorylation and ribosomal RNAs and proteins. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 39

COPDGene Investigators - Core Units: Administrative Center: James D. Crapo, MD (PI); Edwin K. Silverman, MD, PhD (PI); Barry J. Make, MD; Elizabeth A. Regan, MD, PhD Genetic Analysis Center: Terri H. Beaty, PhD; Peter J. Castaldi, MD, MSc; Michael H. Cho, MD, MPH; Dawn L. DeMeo, MD, MPH; Adel Boueiz, MD, MMSc; Marilyn G. Foreman, MD, MS; Auyon Ghosh, MD; Lystra P. Hayden, MD, MMSc; Craig P. Hersh, MD, MPH; Jacqueline Hetmanski, MS; Brian D. Hobbs, MD, MMSc; John E. Hokanson, MPH, PhD; Wonji Kim, PhD; Nan Laird, PhD; Christoph Lange, PhD; Sharon M. Lutz, PhD; Merry- Lynn McDonald, PhD; Dmitry Prokopenko, PhD; Matthew Moll, MD, MPH; Jarrett Morrow, PhD; Dandi Qiao, PhD; Elizabeth A. Regan, MD, PhD; Aabida Saferali, PhD; Phuwanat Sakornsakolpat, MD; Edwin K. Silverman, MD, PhD; Emily S. Wan, MD; Jeong Yun, MD, MPH Imaging Center : Juan Pablo Centeno; Jean-Paul Charbonnier, PhD; Harvey O. Coxson, PhD; Craig J. Galban, PhD; MeiLan K. Han, MD, MS; Eric A. Hoffman, Stephen Humphries, PhD; Francine L. Jacobson, MD, MPH; Philip F. Judy, PhD; Ella A. Kazerooni, MD; Alex Kluiber; David A. Lynch, MB; Pietro Nardelli, PhD; John D. Newell, Jr., MD; Aleena Notary; Andrea Oh, MD; Elizabeth A. Regan, MD, PhD; James C. Ross, PhD; Raul San Jose Estepar, PhD; Joyce Schroeder, MD; Jered Sieren; Berend C. Stoel, PhD; Juerg Tschirren, PhD; Edwin Van Beek, MD, PhD; Bram van Ginneken, PhD; Eva van Rikxoort, PhD; Gonzalo Vegas SanchezFerrero, PhD; Lucas Veitel; George R. Washko, MD; Carla G. Wilson, MS PFT QA Center, Salt Lake City, UT : Robert Jensen, PhD All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 40 Data Coordinating Center and Biostatistics, National Jewish Health, Denver, CO : Douglas Everett, PhD; Jim Crooks, PhD; Katherine Pratte, PhD; Matt Strand, PhD; Carla G. Wilson, MS Epidemiology Core, University of Colorado Anschutz Medical Campus, Aurora, CO: John E. Hokanson, MPH, PhD; Erin Austin, PhD; Gregory Kinney, MPH, PhD; Sharon M. Lutz, PhD; Kendra A. Young, PhDVersion Date: March 26, 2021 Mortality Adjudication Core: Surya P. Bhatt, MD; Jessica Bon, MD; Alejandro A. Diaz, MD, MPH; MeiLan K. Han, MD, MS; Barry Make, MD; Susan Murray, ScD; Elizabeth Regan, MD; Xavier Soler, MD; Carla G. Wilson, MS Biomarker Core : Russell P. Bowler, MD, PhD; Katerina Kechris, PhD; Farnoush BanaeiKashani, PhD COPDGene Investigators - Clinical Centers: Ann Arbor VA: Jeffrey L. Curtis, MD; Perry G. Pernicano, MD Baylor College of Medicine, Houston, TX : Nicola Hanania, MD, MS; Mustafa Atik, MD; Aladin Boriek, PhD; Kalpatha Guntupalli, MD; Elizabeth Guy, MD; Amit Parulekar, MD Brigham and Women’s Hospital, Boston, MA : Dawn L. DeMeo, MD, MPH; Craig Hersh, MD, MPH; Francine L. Jacobson, MD, MPH; George Washko, MD Columbia University, New York, NY : R. Graham Barr, MD, DrPH; John Austin, MD; Belinda D’Souza, MD; Byron Thomashow, MD Duke University Medical Center, Durham, NC : Neil MacIntyre, Jr., MD; H. Page McAdams, MD; Lacey Washington, MD All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 41 HealthPartners Research Institute, Minneapolis, MN : Charlene McEvoy, MD, MPH; Joseph Tashjian, MD Johns Hopkins University, Baltimore, MD: Robert Wise, MD; Robert Brown, MD; Nadia N. Hansel, MD, MPH; Karen Horton, MD; Allison Lambert, MD, MHS; Nirupama Putcha, MD, MHS Lundquist Institute for Biomedical Innovation at Harbor UCLA Medical Center , Torrance, CA: Richard Casaburi, PhD, MD; Alessandra Adami, PhD; Matthew Budoff, MD; Hans Fischer, MD; Janos Porszasz, MD, PhD; Harry Rossiter, PhD; William Stringer, MD Michael E. DeBakey VAMC, Houston, TX: Amir Sharafkhaneh, MD, PhD; Charlie Lan, DO Minneapolis VA: Christine Wendt, MD; Brian Bell, MD; Ken M. Kunisaki, MD, MS Morehouse School of Medicine, Atlanta, GA: Eric L. Flenaugh, MD; Hirut Gebrekristos, PhD; Mario Ponce, MD; Silanath Terpenning, MD; Gloria Westney, MD, MS National Jewish Health, Denver, CO : Russell Bowler, MD, PhD; David A. Lynch, MB Reliant Medical Group, Worcester, MA: Richard Rosiello, MD; David Pace, MD Temple University, Philadelphia, PA : Gerard Criner, MD; David Ciccolella, MD; Francis Cordova, MD; Chandra Dass, MD; Gilbert D’Alonzo, DO; Parag Desai, MD; Michael Jacobs, PharmD; Steven Kelsen, MD, PhD; Victor Kim, MD; A. James Mamary, MD; Nathaniel Marchetti, DO; Aditi Satti, MD; Kartik Shenoy, MD; Robert M. Steiner, MD; Alex Swift, MD; Irene Swift, MD; Maria Elena Vega-Sanchez, MD University of Alabama, Birmingham, AL : Mark Dransfield, MD; William Bailey, MD; Surya P. Bhatt, MD; Anand Iyer, MD; Hrudaya Nath, MD; J. Michael Wells, MD All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint 42 University of California, San Diego, CA : Douglas Conrad, MD; Xavier Soler, MD, PhD; Andrew Yen, MD University of Iowa, Iowa City, IA : Alejandro P. Comellas, MD; Karin F. Hoth, PhD; John Newell, Jr., MD; Brad Thompson, MD University of Michigan, Ann Arbor, MI : MeiLan K. Han, MD MS; Ella Kazerooni, MD MS; Wassim Labaki, MD MS; Craig Galban, PhD; Dharshan Vummidi, MD University of Minnesota, Minneapolis, MN : Joanne Billings, MD; Abbie Begnaud, MD; Tadashi Allen, MD University of Pittsburgh, Pittsburgh, PA : Frank Sciurba, MD; Jessica Bon, MD; Divay Chandra, MD, MSc; Joel Weissfeld, MD, MPH University of Texas Health, San Antonio, San Antonio, TX : Antonio Anzueto, MD; Sandra Adams, MD; Diego Maselli-Caceres, MD; Mario E. Ruiz, MD; Harjinder Singh All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint LTRC samples with available RNA-seq data N = 1,589 Samples with available RNAseq and relevant clinical data N = 508 Samples with available RNAseq data that pass QC N = 1,554 LTRC samples with available clinical data Current smokers and non-smokers (21-91 years old) N = 1,993 Samples with complete relevant clinical data N = 609 Samples without IPF pathology N = 456 Samples with missing data N = 1,384 • Total Perc15 density and CT scanner model: N = 704 • Current smoking status: N = 364 • FEV1: N = 7 • 2 or more variables: N = 309 See Figure E1 for a summary of all missing variables Association analyses Adjusted Perc15 density DGE, DIU, and DEU Figure 1 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint Figure 2 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint A All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint B All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint C All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint D All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint Figure 3 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint Figure 4 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint A All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint B All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint C All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint D All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint E All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint Figure 5 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint A B C D All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint E All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted October 22, 2022. ; https://doi.org/10.1101/2022.10.21.22281369doi: medRxiv preprint