Identification of endoplasmic reticulum stress-related biomarkers in idiopathic pulmonary fibrosis based on machine learning

doi:10.21203/rs.3.rs-2453387/v1

Download PDF

Research Article

Identification of endoplasmic reticulum stress-related biomarkers in idiopathic pulmonary fibrosis based on machine learning

https://doi.org/10.21203/rs.3.rs-2453387/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Endoplasmic reticulum stress (ERS) is critical in the development and progression of idiopathic pulmonary fibrosis (IPF). The aim of this study was to explore ERS-related biomarkers in IPF using a bioinformatics approach and to further investigate their relationship with immune cells. Three IPF datasets (GSE10667, GSE24206 as the training set and GSE53845 as the validation set) were obtained based on the Gene Expression Omnibus (GEO). In the training set, ERS-related differentially expressed genes(DEGs) between IPF and normal tissues were screened and subjected to GO and KEGG enrichment analysis. Key ERS-related DEGs were further screened using Weighted Gene Co-expression Network Analysis (WGCNA) and three machine learning algorithms (Least Absolute Shrinkage and Selection Operator (LASSO), Random Forest (RF), and Support Vector Machine - Recursive Feature Elimination with Local Feature Selection (SVM-RFE)) and validated in a validation set. Then the CIBERSORT method was used to calculate the immune cell infiltration abundance and investigate the relationship between immune cells and key markers.

Results

we obtained 65 ERS-related DEGs from the training set, and 2 key ERS-related DEGs (COMP, GPX8) were screened by WGCNA and machine learning and validated in the validation set. COMP and GPX8 showed high diagnostic value (AUC > 0.8). The results of immune cell infiltration studies showed substantial associations between these two key markers and T-cell CD8, neutrophils, monocytes, macrophage M2 and plasma cells.

Conclusion

COMP and GPX8 are valuable potential biomarkers for IPF and provide a basis for future studies on the early diagnosis and treatment of IPF.

idiopathic pulmonary fibrosis

endoplasmic reticulum stress

machine learning

biomarkers

immunity

Idiopathic pulmonary fibrosis (IPF) is a chronic, progressive, fibrotic interstitial lung disease of unknown origin[1], characterized by damage to alveolar structures and epithelium, activation and proliferation of pulmonary fibroblasts, and excessive extracellular matrix accumulation, resulting in increased interstitial lung mass and decreased lung function[2]. According to statistics, the incidence of IPF in North America and Europe is 3–9 cases/100,000 people/1 year, and in South America and East Asia is 0–4 cases/100,000 people/1 year [3]. However, the incidence and prevalence of IPF is currently increasing due to factors such as increased disease awareness, improved diagnosis, and an aging population[1, 4]. IPF mostly affects adults over the age of 60 and has an extremely dismal prognosis, with an average lifespan of just 3–5 years following diagnosis[5]. Although many studies have been conducted on IPF in the past, the exact mechanisms underlying its development remain unclear. Therefore, it is essential to explore the potential pathogenic mechanisms of IPF and identify new biomarkers to guide the diagnosis and treatment of patients with IPF.

The endoplasmic reticulum (ER) is an important organelle in eukaryotic cells and plays a key role in protein synthesis, folding, modification and quality control[6]. Impairment of endoplasmic reticulum protein-folding capacity under the interference of oxidative stress, nutritional deficiency, viral infection, hypoxia and impaired calcium homeostasis leads to accumulation of misfolded proteins in the ER, a condition known as endoplasmic reticulum stress (ERS)[7, 8]. In response to ERS, cells simultaneously activate the unfolded protein response (UPR) to repair and protect cells[9]. However, when ERS is excessive or prolonged, UPR can also trigger apoptosis. ERS/UPR has been proven to have a significant role in a variety of disorders, including IPF[10–15]. The onset of familial idiopathic pulmonary fibrosis(FIPF) was discovered to be strongly correlated with mutations in surfactant protein C (SFTPC), which is generated by type II AECs (AECIIs)[16, 17]. The SFTPC precursor (proSP-C) has a mutation in the BRICHO structural domain that causes an excessive buildup of misfolded proteins in the ER, inducing ERS and activating the UPR, and ultimately causing AECII apoptosis[18] and ultimately causing lung fibrosis[19, 20]. The same severe ERS and apoptotic manifestations were found in AECIIs of patients with sporadic idiopathic pulmonary fibrosis (SIPF) [21]. These studies suggest that IPF is more sensitive to ERS. ERS can damage AECIIs and accelerate the development of pulmonary fibrosis. In addition to its effects on cell survival, ERS can also affect cell differentiation. It has been shown that in IPF, ERS can affect the differentiation of fibroblasts to myofibroblasts[22, 23]. In AECs ERS induces the epithelial-mesenchymal transition (EMT), at least in part through an Src-dependent pathway, thereby promoting the development of pulmonary fibrosis[24]. In addition, ERS is able to drive M2 macrophage polarization through a CHOP-dependent mechanism [25], exacerbating pulmonary fibrosis[26]. In conclusion, these studies suggest a critical role of ERS in IPF. Therefore, ERS-related genes may be potential markers for the diagnosis and treatment of IPF.

In this study, after exploring ERS-related DEGs in normal subjects and IPF patients, we used WGCNA analysis with three machine learning algorithms (Least absolute shrinkage and selection operator (LASSO), Random Forest (RF), and Support Vector Machine-Recursive Feature Elimination for localized feature selection (SVM-RFE)) to further screen for closely related disease ERS-related DGEs, which were subsequently validated in the validation set to identify potential biomarkers of IPF. Finally, potential associations between key genes and immune cells were investigated in depth.

The flow chart of the study is shown in Fig. 1

Processing of data and Screening of ERS-related DGEs

A total of 117 samples (normal = 29, IPF = 88) from three datasets GSE10667, GSE24206, and GSE53845 were obtained from the GEO database. GSE10667 and GSE24206 were combined and used as training sets after removing the batch effect. Before removing the batch effect, the two dataset samples showed different clustering patterns (Fig. 2A). After removing the batch effect they were able to mix together evenly (Fig. 2B). A total of 1929 genes were screened and the volcano and heat maps were drawn (Fig. 3C, Fig. 3D). DEGs were intersected with ERS-related genes to obtain 65 ERS-related DGEs and plotted Venn diagrams (Fig. 3E).

Functional enrichment analysis of ERS-related DGEs

We explored the biological functions and potential functional pathways of 65 ERS-related DGEs in IPF patients (Fig. 3). In biological processes, ERS-related DGEs are mainly involved in “response to endoplasmic reticulum stress” and “negative regulation of transport”; molecular function showed that ERS-related DGEs were mainly focused on “ion channel regulator activity”, “channel regulator activity”; cellular component showed that ERS-related DGEs were mainly focused on “endoplasmic reticulum lumen”, “membrane microdomain” (Fig. 3A). The KEGG enrichment analysis showed that ERS-related DGEs were mainly focused on “Pathways of neurodegeneration − multiple diseases”, “cAMP signaling pathway and HIF − 1 signaling pathway” (Fig. 3B).

WGCNA

A weighted gene co-expression network was constructed using the “WGCNA” R package in the training set. Among them, the optimal soft threshold was chosen as 4 and the scale independence reached as 0.88 (Fig. 4A, Fig. 4B). Then, MEDissThres was set to 0.25, and a total of 10 modules were obtained by the dynamic tree cutting algorithm and merging similar modules, and their relationships were displayed in a clustered dendrogram (Fig. 4C). The heat map of module-clinical relationship is shown in Fig. 4D, in which blue and red modules had the highest correlation with IPF, with correlation coefficients of 0.64 and 0.67, respectively (p < 0.05). After filtering by GS and MM, 599 hub genes in blue and red modules were finally extracted. The 599 hub genes obtained from the WGCNA analysis were intersected with 65 ERS-related DGEs to obtain 17 key ERS-related DGEs (Fig. 4E).

Screening of key ERS-related DGEs

The 17 key ERS-related DGEs were subsequently filtered using 3 algorithms (LASSO, random forest, SVM-RFE). Lasso regression screened 5 genes (Fig. 5A, Fig. 5B); SVM-RFE algorithm screened 7 genes (Fig. 5C); random forest selected 5 genes with the highest importance ranking (Fig. 5D, Fig. 5E). The genes screened by the 3 algorithms were intersected to obtain 2 DGEs associated with ERS (COMP, GXP8) (Fig. 5F). The expression of COMP and GPX8 was significantly upregulated in the validation group (adjusted p-value < 0.05) (SUPPLEMENTARY DATA Table 1). The AUC of COMP in the training set was 0.984 (95% CI: 0.955-1.000) (Fig. 6A). The AUC of COMP in the validation set was 0.953 (95% CI: 0.884-1.000) (Fig. 6B). The AUC of GPX8 in the training set was 0.938 (95% CI: 0.865–0.987) (Fig. 6C). The AUC of GPX8 in the validation set was 0.825 (95% CI: 0.634–0.966) (Fig. 6D).

Analysis of immune cells

To further explore the correlation between immune cells and IPF, ERS-related DGEs, this section analyzes the level of immune cell infiltration in the samples using the CIBERSORT method (Fig. 7A). The analysis showed that Plasma cells, Macrophages M0 were highly expressed in IPF species and NK cells activated, Monocytes, Eosinophils were low expressed in IPF (Fig. 7B). Analysis of the correlation between COMP and immune cells: COMP was correlated with T cells CD4 memory resting, Neutrophils, Monocytes, and positively with Plasma cells (p < 0.05) (Fig. 7C). Analysis of the correlation between GPX8 and immune cells: GPX8 expression was negatively correlated with T cells CD8, Neutrophils, Monocytes, and positively correlated with Macrophages M2, Plasma cells (p < 0.05) (Fig. 7D).

IPF This is a group of diffuse parenchymal lung diseases of unknown cause, and the mechanisms responsible for the fibrotic process and structural disorders of lung tissue are still unclear[27]. The fibrotic process of IPF is irreversible and there is no effective treatment. Statistically, nearly one third of patients die within 3 years of disease diagnosis[28]. Due to its extremely high mortality rate, there is an urgent need for an effective biomarker for early diagnosis and treatment of the disease. The endoplasmic reticulum (ER), a key organelle in eukaryotic cells, can induce endoplasmic reticulum stress (ERS) when there is an accumulation of misfolded proteins in the endoplasmic reticulum. ERS can activate the unfolded protein response (UPR), a cytoprotective response that removes misfolded proteins to protect the cell, but it can also trigger cell death when the ERS is overloaded[29]. In addition, ERS can cause immune and inflammatory reactions through some signal pathways[30, 31]. It has been shown that IPF is closely associated with ERS. Therefore, finding ERS-related biomarkers and revealing their correlation with IPF could provide important information for early diagnosis and treatment.

In this study, we comprehensively evaluated the expression of ERS-related genes in normal and IPF tissues and identified a total of 65 ERS-related DEGs. we then explored the biological functions of the 65 ERS-related DEGs. GO analysis showed a strong link between DEGs and biological processes connected to ERS. KEGG analysis similarly demonstrated the involvement of ERS in the pathophysiological pathway of IPF. For example, prostacyclin (IP) receptor agonist (ACT-333679) is able to suppress YAP/TAZ-dependent fibrosis gene transcription by elevating cAMP[32], thereby inhibiting pulmonary fibrosis. In addition, local hypoxia can lead to ER stress, induced CHOP expression, increased apoptosis of type II AEC cells and enhanced pulmonary fibrosis[33].

Machine learning is an emerging discipline. It is able to find patterns in observed data and use these patterns to make predictions about unknown data. Machine learning has now been applied in a large number of clinical studies. It was reported that a random forest algorithm was used to screen for iron death-related biomarkers in IPF by He J, et al[34]. However, different machine learning algorithms can produce different computational results, so this study uses three algorithms, SVM-RFE, LASSO, and RF, together to screen genes with excellent predictive power. SVM-RFE is an SVM-based machine learning method that generates optimal choices by subtracting the feature vectors generated by SVM [35]. LASSO regression is another machine learning method that selects variables by picking the value of λ with the smallest classification error [36]. The random forest algorithm can create binary subtrees from the training samples generated by bootstrap, and then merge the prediction results of multiple decision trees to finally obtain the importance ranking of each predictor [37]. In this study, two key ERS-related DGEs (COMP, GPX8) were further screened by WGCNA analysis and the three machine learning algorithms mentioned above and validated in the validation set.

Cartilage oligomeric matrix protein (COMP) is an extracellular matrix (ECM) protein that has been found to be expressed in a variety of tissues: fibroblasts, tendon, cartilage, etc[38–41]. Currently, COMP has been shown to play a role in a variety of fibrotic diseases. In the liver, COMP increases type 1 collagen synthesis in hepatic stellate cells via CD36 receptor signaling and activation of the MEK1/2-per k1/2 pathway, promoting liver fibrosis[42]. And in IPF, COMP has the same fibrogenic effect[43]. COMP was highly expressed in IPF tissues, which is consistent with our findings. In an in vitro assay, stimulation of normal human lung fibroblasts with TGF-β1 induced an increase in COMP expression, and silencing of COMP reduced the expression of TGF-β1 target genes and inhibited fibroblast proliferation. In addition, Vuga LJ et al. found that COMP concentrations in the blood of IPF patients correlated with the severity of disease in IPF patients[43]. GPX8 is the last member of the glutathione peroxidase (GPx) protein family to be identified, a type II transmembrane protein located in the ER[44]. GPx is able to break down peroxides into alcohols and limit intracellular deposition of reactive oxygen species (ROS)[45]. However, GPX8 exhibits low GPx activity [44]. The physiological function of GPX8 is unknown. It was found that the expression level of GPX8 could affect Ca2 + homeostasis and signaling, in which the transmembrane structural domain (TMD) of GPX8 was critical for Ca2 + homeostasis[46]. Interestingly, disruption of Ca2 + homeostasis can promote pulmonary fibrosis[47]. In addition, in mesenchymal-like cells (MDA-MB-231), GPX8 knockdown was able to lead to epithelial-like morphological changes, downregulation of EMT characteristics[48]. This further reveals the potential connection between GPX8 and IPF.

Immune cells have been shown to be associated with the development and progression of IPF [49]. In this study, Plasma cells (PCs), Macrophages M0 were more infiltrated in IPF tissues, NK cells activated, Monocytes, Eosinophils were low infiltrated. Plasma cells, also known as effector B cells, specifically express CD138 on their surface. Prele CM et al. also found high infiltration of CD138 + cells in lung tissue from IPF patients and mouse models of pulmonary fibrosis, and the use of bortezomib reduced PCs infiltration and inhibited bleomycin-induced pulmonary fibrosis[50]. Macrophages are the highest proportion of lung immune cells (about 70%) and play an important role in the development of pulmonary fibrosis[51]. However, the role of M0 macrophage infiltration has not been fully demonstrated. This may provide a new direction for further study of pulmonary fibrosis. PF543, a specific sphingosine kinase 1 (SPHK1) inhibitor, was found to reduce lung fibrosis by reducing mtDNA damage and monocyte recruitment in lung epithelial cells[52]. Similarly, monocyte infiltration was found to be less in our study, which could be a potential pathogenic mechanism for IPF. Schwartz DA et al. found more eosinophils in bronchoalveolar lavage fluid (BALF) in the IPF patient group than in the control group[53]. This phenomenon differs from the results of the present study and may be due to differences in samples and detection techniques, which require further validation. In the analysis of the correlation between immune cells and key markers, Comp and GPX8 were negatively correlated with monocytes and neutrophils and positively correlated with plasma cells. It's interesting to note that these 2 genes did not always exhibit the same link with various immune cell subpopulations; for example, COMP was inversely connected with T cells CD4 memory resting. In contrast, GPX8 was negatively correlated with T cells CD8 and positively correlated with Macrophages M2. Although our findings on the correlation of COMP and GPX8 with immune infiltration suggest that these genes may be involved in the regulation of immune cells, the exact mechanisms remain unclear and need to be further investigated.

Our research still has certain shortcomings, however. First, in order to increase the dependability of the study's findings, the sample size must be increased further. Second, relevant in vivo and in vitro research are still needed to support the findings of this study.

We discovered two biomarkers (COMP, GPX8) connected to ERS using machine learning and WGCNA analysis, and we also indicated possible connections between these two genes and the immunological microenvironment. The relationship between ERS and IPF is further explained by our work, which also offers fresh perspectives on how to identify and treat IPF.

Acquisition and processing of data

Three datasets (GSE10667, GSE24206, GSE53845) containing idiopathic pulmonary fibrosis and normal lung tissue and their corresponding platform annotation information were downloaded from the Gene Expression Omnibus (GEO) (https://www.ncbi.nlm.nih.gov/geo/). Convert probe ids to gene symbols based on annotation information from the platform. GSE10667 (n = 46) and GSE24206 (n = 23) were merged and corrected to be used as training sets. GSE53845(n = 48) as a validation set. ERS-related genes (correlation score ≥ 8) obtained from the GeneCards database (https://www.genecards.org).

Screening for differentially expressed genes (DGEs) in the training set The following criteria were used to identify DEGs: |log2 Fold Change| > 0.5 and adjusted p-value < 0.05. Ploting the “heat map” and “volcano map” of the DEGs. The DEGs were intersected with ERS-related genes and Venn diagram was plotted. This part is done using the “limma”, “sva”, “pheatmap” and “ggplot2” packages from the R software (version 4.2.1).

Enrichment analysis of ERS-related DEGs

Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) functional enrichment analysis was performed on all ERS-related DEGs using the “clusterProfiler, org.Hs.eg.db, enrichplot” package in R software. GO analysis shows the biological process (BP), molecular function (MF), and cellular components (CCs) of ERS-related DEGs. KEGG analysis demonstrates the important pathways of ERS-related DEGs. The threshold was set at a p-value < 0.05 and the results were visualized using the “ggplot” package.

Weighted gene co-expression network analysis

Weighted gene co-expression network analysis (WGCNA) of all genes using the “WGCNA” package. First, the samples are clustered and outliers are removed. The best value soft threshold was selected using the pickSoftThreshold function to build a topological overlap matrix (TOM) to calculate the expression relationship between genes. The hierarchical clustering nodes of the modules were determined based on the similarity of the TOM, and the related gene modules were constructed by the dynamic tree cutting algorithm (Minimum number of genes in a module: 60, threshold for merging similar modules: 0.25). Gene Significance (GS) and Module Membership (MM) were calculated to correlate modules with clinical features. Finally, the genes of key modules were extracted. The criteria were as follows: MM > 0.8 and GS > 0.5. To further screen genes, key module genes from WGCNA analysis were crossed with ERS-associated DEGs. The intersecting genes were plotted in a Venn diagram.

Screening and validation of key markers

ERS-related DEGs were further screened using 3 algorithms (LASSO, RF, SVM-RFE). LASSO regression analysis was performed using a 10-fold cross-validation method. Using random forest analysis, the top 5 genes with the highest importance scores were selected. Using the SVM-RFE algorithm, 17 genes were selected for model construction, and the least error node was selected to extract the significant genes. The ERS-related DEGs obtained by the three algorithms were intersected to obtain the key genes. This part is done using the packages “glmnet”, “randomForest”, “e1071”, “caret “, “kernlab”, “msvmRFE.R”, “tidyverse”, et al. Differential expression analysis of COMP and GPX8 was performed in the validation set. The receiver operating characteristic curve (ROC curve) of key genes was plotted in the training set and validation set.

Analysis of immune cells

We used the CIBERSORT method to analyze the level of immune cell infiltration between IPF and normal samples in the training set (parameter “PERM” set to 1000, threshold p < 0.05) and to calculate the proportion of each type of immune cells in the samples. Correlation analysis of differential immune cells and key genes using Spearson's correlation analysis method. This section was calculated and plotted using the “reshape2, ggplot2, vioplot, corrplot” packages.

Statistical Analysis

All statistical tests were performed using R software (version 4.2.1). p < 0.05 were considered statistically significant.

IPF

idiopathic pulmonary fibrosis

ERS

Endoplasmic reticulum stress

GEO

Gene Expression Omnibus

DEGs

Differentially Expressed Genes

WGCNA

Weighted Gene Co-expression Network Analysis

LASSO

Least Absolute Shrinkage and Selection Operator

Random Forest

SVM-RFE

Support Vector Machine - Recursive Feature Elimination with Local Feature Selection

Endoplasmic reticulum

UPR

Unfolded protein response

FIPF

Familial idiopathic pulmonary fibrosis

SIPF

Sporadic idiopathic pulmonary fibrosis

EMT

Epithelial-mesenchymal transition

Gene Ontology

KEGG

Kyoto Encyclopedia of Genes and Genomes

Biological Process

Cellular Component

Molecular Function

Gene Significance

Module Membership

BALF

Bronchoalveolar Lavage Fluid

ROC

Receiver Operating Characteristic

TOM

Topological Overlap Matrix.

Acknowledgements

None.

Availability of data and materials

Research data are reasonably available from the authors

Funding

Authors received no specific funding for this work.

Authors’ contributions

Haishan Huang was in charge of data analysis and drafting, and Zhiqi Wang was in charge of data acquisition. Tianyang Liu were responsible for study design and supervision. Final manuscript read and approved by all authors.

Ethics approval and consent to participate

This study does not contain any studies with human participants or animals.

Consent for publication

Not applicable.

Competing interests

Authors declare no conflicts of interest.

Author details

¹Department of Emergency, The Affiliated Suqian Hospital of Xuzhou Medical University, Suqian 223800, China. ²Department of Respiratory and Critical Care Medicine, Wuxi No.5 People's Hospital Wuxi 214000, China. ³Department of Lung Transplantation, The affiliated Wuxi People's Hospital of Nanjing Medical University, Wuxi 214000, China.

Supplementary Materials

Supplementary 1.Table 1: The expression difference analysis of COMP and GPX8.

Sgalla G, Biffi A, Richeldi L. Idiopathic pulmonary fibrosis: Diagnosis, epidemiology and natural history. Respirology. 2016;213:427–37.
Lederer DJ, Martinez FJ. Idiopathic Pulmonary Fibrosis. N Engl J Med. 2018;37819:1811–23.
Hutchinson J, Fogarty A, Hubbard R, McKeever T. Global incidence and mortality of idiopathic pulmonary fibrosis: a systematic review. Eur Respir J. 2015;463:795–806.
Sauleda J, Nunez B, Sala E, Soriano JB. Idiopathic Pulmonary Fibrosis: Epidemiology, Natural History, Phenotypes.Med Sci (Basel)2018;64
Raghu G, Remy-Jardin M, Myers JL, Richeldi L, Ryerson CJ, Lederer DJ et al. Diagnosis of Idiopathic Pulmonary Fibrosis. An Official ATS/ERS/JRS/ALAT Clinical Practice Guideline. American Journal of Respiratory and Critical Care Medicine 2018; 1985:e44-e68.
Schroder M. Endoplasmic reticulum stress responses. Cell Mol Life Sci. 2008;656:862–94.
Zhang HM, Ye X, Su Y, Yuan J, Liu Z, Stein DA, et al. Coxsackievirus B3 infection activates the unfolded protein response and induces apoptosis through downregulation of p58IPK and activation of CHOP and SREBP1. J Virol. 2010;8417:8446–59.
Ma Y, Hendershot LM. ER chaperone functions during normal and stress conditions. J Chem Neuroanat. 2004;281–2:51–65.
Walter P, Ron D. The unfolded protein response: from stress pathway to homeostatic regulation. Science. 2011;3346059:1081–6.
Yang S, Wu M, Li X, Zhao R, Zhao Y, Liu L et al. Role of Endoplasmic Reticulum Stress in Atherosclerosis and Its Potential as a Therapeutic Target. Oxid Med Cell Longev 2020; 2020:9270107.
Song M, Cubillos-Ruiz JR. Endoplasmic Reticulum Stress Responses in Intratumoral Immune Cells: Implications for Cancer Immunotherapy. Trends Immunol. 2019;402:128–41.
Jeong JS, Kim SR, Cho SH, Lee YC. Endoplasmic Reticulum Stress and Allergic Diseases. Curr Allergy Asthma Rep. 2017;1712:82.
Hsu HS, Liu CC, Lin JH, Hsu TW, Hsu JW, Su K, et al. Involvement of ER stress, PI3K/AKT activation, and lung fibroblast proliferation in bleomycin-induced pulmonary fibrosis. Sci Rep. 2017;71:14272.
Lawson WE, Cheng DS, Degryse AL, Tanjore H, Polosukhin VV, Xu XC, et al. Endoplasmic reticulum stress enhances fibrotic remodeling in the lungs. Proc Natl Acad Sci U S A. 2011;10826:10562–7.
Salminen A, Kauppinen A, Suuronen T, Kaarniranta K, Ojala J. ER stress in Alzheimer's disease: a novel neuronal trigger for inflammation and Alzheimer's pathology. J Neuroinflammation. 2009;6:41.
Thomas AQ, Lane K, Phillips J 3rd, Prince M, Markin C, Speer M, et al. Heterozygosity for a surfactant protein C gene mutation associated with usual interstitial pneumonitis and cellular nonspecific interstitial pneumonitis in one kindred. Am J Respir Crit Care Med. 2002;1659:1322–8.
Nogee LM, Dunbar AE 3rd, Wert SE, Askin F, Hamvas A, Whitsett JA. A mutation in the surfactant protein C gene associated with familial interstitial lung disease. N Engl J Med. 2001;3448:573–9.
Mulugeta S, Nguyen V, Russo SJ, Muniswamy M, Beers MF. A surfactant protein C precursor protein BRICHOS domain mutation causes endoplasmic reticulum stress, proteasome dysfunction, and caspase 3 activation. Am J Respir Cell Mol Biol. 2005;326:521–30.
Cheresh P, Kim SJ, Tulasiram S, Kamp DW. Oxidative stress and pulmonary fibrosis. Biochim Biophys Acta. 2013;18327:1028–40.
Cui Y, Robertson J, Maharaj S, Waldhauser L, Niu J, Wang J, et al. Oxidative stress contributes to the induction and persistence of TGF-beta1 induced pulmonary fibrosis. Int J Biochem Cell Biol. 2011;438:1122–33.
Korfei M, Ruppert C, Mahavadi P, Henneke I, Markart P, Koch M, et al. Epithelial endoplasmic reticulum stress and apoptosis in sporadic idiopathic pulmonary fibrosis. Am J Respir Crit Care Med. 2008;1788:838–46.
Ghavami S, Yeganeh B, Zeki AA, Shojaei S, Kenyon NJ, Ott S, et al. Autophagy and the unfolded protein response promote profibrotic effects of TGF-beta1 in human lung fibroblasts. Am J Physiol Lung Cell Mol Physiol. 2018;3143:L493–L504.
Baek HA, Kim DS, Park HS, Jang KY, Kang MJ, Lee DG, et al. Involvement of endoplasmic reticulum stress in myofibroblastic differentiation of lung fibroblasts. Am J Respir Cell Mol Biol. 2012;466:731–9.
Zhong Q, Zhou B, Ann DK, Minoo P, Liu Y, Banfalvi A, et al. Role of Endoplasmic Reticulum Stress in Epithelial–Mesenchymal Transition of Alveolar Epithelial Cells. Am J Respir Cell Mol Biol. 2011;453:498–509.
Yao Y, Wang Y, Zhang Z, He L, Zhu J, Zhang M, et al. Chop Deficiency Protects Mice Against Bleomycin-induced Pulmonary Fibrosis by Attenuating M2 Macrophage Production. Mol Ther. 2016;245:915–25.
Gharib SA, Johnston LK, Huizar I, Birkland TP, Hanson J, Wang Y, et al. MMP28 promotes macrophage polarization toward M2 cells and augments pulmonary fibrosis. J Leukoc Biol. 2013;951:9–18.
Nalysnyk L, Cid-Ruzafa J, Rotella P, Esser D. Incidence and prevalence of idiopathic pulmonary fibrosis: review of the literature. Eur Respir Rev. 2012;21126:355–61.
Blackwell TS, Tager AM, Borok Z, Moore BB, Schwartz DA, Anstrom KJ, et al. Future directions in idiopathic pulmonary fibrosis research. An NHLBI workshop report. Am J Respir Crit Care Med. 2014;1892:214–22.
Ozcan L, Tabas I. Role of endoplasmic reticulum stress in metabolic disease and other disorders. Annu Rev Med. 2012;63:317–28.
Sprenkle NT, Sims SG, Sanchez CL, Meares GP. Endoplasmic reticulum stress and inflammation in the central nervous system. Mol Neurodegener. 2017;121:42.
Bettigole SE, Glimcher LH. Endoplasmic reticulum stress in immunity. Annu Rev Immunol. 2015;33:107–38.
Zmajkovicova K, Menyhart K, Bauer Y, Studer R, Renault B, Schnoebelen M, et al. The Antifibrotic Activity of Prostacyclin Receptor Agonism Is Mediated through Inhibition of YAP/TAZ. Am J Respir Cell Mol Biol. 2019;605:578–91.
Burman A, Kropski JA, Calvi CL, Serezani AP, Pascoalino BD, Han W et al. Localized hypoxia links ER stress to lung fibrosis through induction of C/EBP homologous protein.JCI Insight2018;316
He J, Li X, Yu M. Bioinformatics Analysis Identifies Potential Ferroptosis Key Genes in the Pathogenesis of Pulmonary Fibrosis. Front Genet. 2021;12:788417.
Sanz H, Valim C, Vegas E, Oller JM, Reverter F. SVM-RFE: selection and visualization of the most relevant features through non-linear kernels. BMC Bioinformatics. 2018;191:432.
Cheung-Lee WL, Link AJ. Genome mining for lasso peptides: past, present, and future.J Ind Microbiol Biotechnol2019; 469 – 10:1371–1379.
Holodinsky JK, Yu AY, X, Kapral MK, Austin PC. Using random forests to model 90-day hometime in people with stroke. BMC Med Res Methodol. 2021;211:102.
Riessen R, Fenchel M, Chen H, Axel DI, Karsch KR, Lawler J. Cartilage oligomeric matrix protein (thrombospondin-5) is expressed by human vascular smooth muscle cells. Arterioscler Thromb Vasc Biol. 2001;211:47–54.
Muller G, Michel A, Altenburg E. COMP (cartilage oligomeric matrix protein) is synthesized in ligament, tendon, meniscus, and articular cartilage. Connect Tissue Res. 1998;394:233–44.
Hummel KM, Neidhart M, Vilim V, Hauser N, Aicher WK, Gay RE, et al. Analysis of cartilage oligomeric matrix protein (COMP) in synovial fibroblasts and synovial fluids. Br J Rheumatol. 1998;377:721–8.
DiCesare P, Hauser N, Lehman D, Pasumarti S, Paulsson M. Cartilage oligomeric matrix protein (COMP) is an abundant component of tendon. FEBS Lett. 1994;3542:237–40.
Magdaleno F, Arriazu E, Ruiz de Galarreta M, Chen Y, Ge X, de la Conde L, et al. Cartilage oligomeric matrix protein participates in the pathogenesis of liver fibrosis. J Hepatol. 2016;655:963–71.
Vuga LJ, Milosevic J, Pandit K, Ben-Yehudah A, Chu Y, Richards T, et al. Cartilage oligomeric matrix protein in idiopathic pulmonary fibrosis. PLoS ONE. 2013;812:e83120.
Nguyen VD, Saaranen MJ, Karala AR, Lappi AK, Wang L, Raykhel IB, et al. Two endoplasmic reticulum PDI peroxidases increase the efficiency of the use of peroxide during disulfide bond formation. J Mol Biol. 2011;4063:503–15.
Brigelius-Flohe R, Maiorino M. Glutathione peroxidases. Biochim Biophys Acta. 2013;18305:3289–303.
Yoboue ED, Rimessi A, Anelli T, Pinton P, Sitia R. Regulation of Calcium Fluxes by GPX8, a Type-II Transmembrane Peroxidase Enriched at the Mitochondria-Associated Endoplasmic Reticulum Membrane. Antioxid Redox Signal. 2017;279:583–95.
Mukherjee S, Ayaub EA, Murphy J, Lu C, Kolb M, Ask K, et al. Disruption of Calcium Signaling in Fibroblasts and Attenuation of Bleomycin-Induced Fibrosis by Nifedipine. Am J Respir Cell Mol Biol. 2015;534:450–8.
Khatib A, Solaimuthu B, Ben Yosef M, Abu Rmaileh A, Tanna M, Oren G, et al. The glutathione peroxidase 8 (GPX8)/IL-6/STAT3 axis is essential in maintaining an aggressive breast cancer phenotype. Proc Natl Acad Sci U S A. 2020;11735:21420–31.
Kolahian S, Fernandez IE, Eickelberg O, Hartl D. Immune Mechanisms in Pulmonary Fibrosis. Am J Respir Cell Mol Biol. 2016;553:309–22.
Prele CM, Miles T, Pearce DR, O'Donoghue RJ, Grainge C, Barrett L et al. Plasma cell but not CD20-mediated B cell depletion protects from bleomycin-induced lung fibrosis.Eur Respir J2022
Cai Y, Sugimoto C, Arainga M, Alvarez X, Didier ES, Kuroda MJ. In vivo characterization of alveolar and interstitial lung macrophages in rhesus macaques: implications for understanding lung disease in humans. J Immunol 2014; 1926:2821–2829.
Cheresh P, Kim SJ, Huang LS, Watanabe S, Joshi N, Williams KJN et al. The Sphingosine Kinase 1 Inhibitor, PF543, Mitigates Pulmonary Fibrosis by Reducing Lung Epithelial Cell mtDNA Damage and Recruitment of Fibrogenic Monocytes.Int J Mol Sci2020;2116
Schwartz DA, Helmers RA, Dayton CS, Merchant RK, Hunninghake GW. Determinants of bronchoalveolar lavage cellularity in idiopathic pulmonary fibrosis. J Appl Physiol (1985) 1991; 715:1688–1693.

SupplementaryMaterial.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Identification of endoplasmic reticulum stress-related biomarkers in idiopathic pulmonary fibrosis based on machine learning

Status:

Version 1

Abstract

Background

Results

Conclusion

Figures

Introduction

Results

Processing of data and Screening of ERS-related DGEs

Functional enrichment analysis of ERS-related DGEs

WGCNA

Screening of key ERS-related DGEs

Analysis of immune cells

Discussion

Conclusion

Materials And Methods

Acquisition and processing of data

Enrichment analysis of ERS-related DEGs

Weighted gene co-expression network analysis

Screening and validation of key markers

Analysis of immune cells

Statistical Analysis

Abbreviations

Declarations

References

Supplementary Files

Status:

Version 1