Marker and model for predicting pulmonary fibrosis degree and prognosis condition and application

Abstract

The invention belongs to the technical field of biomedicine, and particularly relates to a marker for predicting the pulmonary fibrosis degree and prognosis condition, a model and application. The invention firstly provides a marker for predicting the degree of pulmonary fibrosis and / or evaluating the prognosis of pulmonary fibrosis. The marker comprises CD163, DNAJB11, EIF6, ITGAX, REEP5, RNASE3, SFTPB and / or TIMM8B. And then based on the marker, carrying out single-factor COX, LASSO and multi-factor COX regression analysis by utilizing IPF expression data and prognosis data to establish a pulmonary fibrosis prognosis risk prediction model. The pulmonary fibrosis prognosis risk prediction model has good prediction efficiency, can perform prognosis evaluation on the pulmonary fibrosis degree of new sudden infectious diseases and viral pneumonia, assists clinical identification of potential critical patients, and further provides help for making treatment strategies more accurately and individually.

Description

Technical Field

[0001] The present invention belongs to the field of biomedical technology, and specifically relates to a marker, a model and an application for predicting the degree and prognosis of pulmonary fibrosis. Background Art

[0002] Pulmonary fibrosis is a characteristic pathological change and the final pathological outcome of a variety of interstitial lung diseases that are not properly treated, mistreated, or prolonged. The pathological characteristics of pulmonary fibrosis are mainly manifested in the abnormal accumulation of extracellular matrix in the lungs, as well as the transdifferentiation of fibroblasts into myofibroblasts and their continuous proliferation. This will not only lead to a decrease in lung compliance and pulmonary ventilation dysfunction; the remodeling of the lung structure will also destroy the normal alveolar structure, leading to a reduction in alveolar area and thickening of the respiratory membrane, causing pulmonary ventilation dysfunction, and thus affecting the patient's prognosis and treatment. Long-term pulmonary fibrosis will also increase the patient's right heart load, causing pulmonary hypertension and even right heart failure, increasing the complexity of the disease and the difficulty of treatment. At present, the diseases that cause pulmonary fibrosis mainly include idiopathic pulmonary fibrosis (IPF), various types of viral pneumonia such as COVID-19, and various interstitial lung diseases. Treatments for pulmonary fibrosis include oxygen therapy, drug therapy, surgical treatment, etc. However, there are very limited drugs for the treatment of pulmonary fibrosis. Currently, only two drugs, Nintedanib and Pirfenidone, have been approved by the U.S. Food and Drug Administration for the treatment of idiopathic pulmonary fibrosis, but these two drugs can only delay the progression of pulmonary fibrosis. Lung transplantation is still the only way to cure IPF.

[0003] Since pulmonary fibrosis progresses slowly and its clinical manifestations are similar to those of bronchitis, asthma and other diseases, most patients are diagnosed in the middle and late stages of the disease, which is not only difficult to treat but also has a poor prognosis. Therefore, it is necessary to conduct in-depth research on the pulmonary fibrosis process, especially IPF, and find new methods to achieve early diagnosis and treatment of pulmonary fibrosis and prognosis risk assessment, which is of great clinical significance for improving the prognosis of patients with pulmonary fibrosis.

[0004] Some prior arts disclose methods for predicting the occurrence of idiopathic pulmonary fibrosis or methods for predicting prognosis. For example, patent publication number CN114026429A discloses a method for predicting the prognosis of idiopathic pulmonary fibrosis, which determines the prognosis of idiopathic pulmonary fibrosis by detecting the amount of at least one protein among S100A4, CIRP and 14-3-3γ. And the patent with publication number CN107099581A discloses a method for predicting, diagnosing and treating idiopathic pulmonary fibrosis, which predicts idiopathic pulmonary fibrosis by detecting the expression of any one or more of the following genes: MUCL1, MUC4, MUC20, PRR7, PRR15, SPRR1B, SPRR2D, KRT5, KRT6B, KRT13, KRT14, KRT15, ​​KRT17, SERPINB3, SERPINB4, SERPINB5, SERPINB13, CLCA2, TRPV4, BBS5, MMP3 and SAA4; CXCR3, CXCR5, CXCL13, CCR6, CCR7, CD19, MS4A1 (CD20), BLK, BLNK, FCRLA, FCRL2, FCRL5, CD79A, CD79B, CD27, CD28, CD1A, CD1B, CD1C, CD1E, IGHV1-69, IGLJ3, IGJ, IGHV3-48, IGLV3-21, IGKV1-5, IGHG1, IGKC, IGLV6-57, IGK@, ​​IGHA1, IGKV2- 24. IGKV1D-8 and IGHM; COL1A1, COL1A2, COL5A2, COL12A1, COL14A1, COL15A1, COL16A1, COL18A1, CTHRC 1, HGF, IGFBP7, SCGF (CLEC11A), LOXL1, LOXL2; GLI1, GLI2, SMO, SFRP2, DIO2, CDH11, POSTN and TGFB3; CHI3L1 (YKL-40), CCL11, CCL13, CCL17, CCL18, COMP, CXCL13, MMP3, MMP7, SAA4 (constitutive SAA), POSTN and AND SPP1 (OPN). Summary of the invention

[0005] In view of this, one of the objects of the present invention is to provide a marker for predicting the degree of pulmonary fibrosis and / or evaluating the prognosis of pulmonary fibrosis, so as to provide support for the construction of a subsequent pulmonary fibrosis prognosis risk prediction model.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] Markers for predicting the degree of pulmonary fibrosis and / or assessing the prognosis of pulmonary fibrosis, wherein the markers include any one or more of CD163, DNAJB11, EIF6, ITGAX, REEP5, RNASE3, SFTPB, and TIMM8B.

[0008] Furthermore, the markers consist of CD163, DNAJB11, EIF6, ITGAX, REEP5, RNASE3, SFTPB and TIMM8B.

[0009] Furthermore, the pulmonary fibrosis includes pulmonary fibrosis caused by emerging infectious diseases and viral pneumonia.

[0010] Preferably, the pulmonary fibrosis comprises idiopathic pulmonary fibrosis and / or COVID-19 pulmonary fibrosis.

[0011] A second object of the present invention is to provide an application of the aforementioned markers in preparing a pulmonary fibrosis prognosis risk prediction model.

[0012] To achieve the above object, the present invention adopts the following technical solutions:

[0013] Application of the aforementioned markers in the preparation of a prognostic risk prediction model for pulmonary fibrosis.

[0014] Preferably, the pulmonary fibrosis comprises idiopathic pulmonary fibrosis and / or COVID-19 pulmonary fibrosis.

[0015] The third object of the present invention is to provide a reagent for detecting the expression amount of the aforementioned markers for use in preparing a product for predicting the degree of pulmonary fibrosis and / or evaluating the prognosis of pulmonary fibrosis.

[0016] To achieve the above object, the present invention adopts the following technical solutions:

[0017] Use of a reagent for detecting the expression of the aforementioned markers in the preparation of a product for predicting the degree of pulmonary fibrosis and / or evaluating the prognosis of pulmonary fibrosis.

[0018] Furthermore, a reagent for detecting the expression level of CD163 is used in the preparation of a product for predicting the degree of pulmonary fibrosis.

[0019] Preferably, the pulmonary fibrosis comprises idiopathic pulmonary fibrosis and / or COVID-19 pulmonary fibrosis.

[0020] A fourth object of the present invention is to provide a pulmonary fibrosis prognosis risk prediction model.

[0021] To achieve the above object, the present invention adopts the following technical solutions:

[0022] A pulmonary fibrosis prognosis risk prediction model is constructed based on the aforementioned markers.

[0023] The present invention first uses laser microdissection technology and DIA quantitative proteomics technology to obtain protein expression data of 8 cases of COVID-19 and 8 cases of normal lung tissue. Then, the gene expression data and prognosis data of IPF are obtained using the GEO database, and the differential analysis method is used to screen the differentially expressed genes / proteins of IPF and COVID-19 relative to normal lung tissue. Then, the intersection of the two is taken to screen 8 common differentially expressed genes / proteins: CD163, DNAJB11, EIF6, ITGAX, REEP5, RNASE3, SFTPB and TIMM8B. Finally, based on these 8 prognosis-related genes, univariate COX, LASSO and multivariate COX regression analysis are performed using IPF expression data and prognosis data to establish a pulmonary fibrosis prognosis risk model. Through this pulmonary fibrosis risk prediction model, the degree of pulmonary fibrosis in patients with new emerging infectious diseases and viral pneumonia can be predicted, and groups that are more prone to pulmonary fibrosis can be screened from a genetic level, providing a basis for early precision and personalized treatment.

[0024] Furthermore, the formula of the pulmonary fibrosis prognosis risk prediction model is: Riskscore=0.821×Exp(CD163)+1.49×Exp(DNAJB11)-0.937×Exp(EIF6)+0.682×Exp(ITGAX)-0.828×Exp(REEP5)+0.295×Exp(RNASE3)+0.125×Exp(SFTPB)+1.143×Exp(TIMM8B); wherein, "Riskscore" represents the risk value, and "Exp" represents the expression value of the corresponding gene.

[0025] Furthermore, the determination of the Riskscore value is based on the median of a group of samples. If the Riskscore value is greater than or equal to the median of the samples in this group, it is judged to be a high-risk state; conversely, if the Riskscore value is less than the median of the samples in this group, it is judged to be a low-risk state; the higher the risk value, the more severe the patient's pulmonary fibrosis, the worse the prognosis, and the shorter the survival time.

[0026] Furthermore, the median risk value is 2.21E-08.

[0027] Furthermore, the test sample includes any one or more of a blood sample and a tissue sample, preferably a blood sample.

[0028] The fifth object of the present invention is to provide an application of the aforementioned pulmonary fibrosis prognosis risk prediction model in the preparation of a product for predicting the degree of pulmonary fibrosis, assessing the prognosis of pulmonary fibrosis, and assessing the severity of COVID-19.

[0029] To achieve the above object, the present invention adopts the following technical solutions:

[0030] The application of the aforementioned pulmonary fibrosis prognosis risk prediction model in the preparation of products for predicting the degree of pulmonary fibrosis, assessing the prognosis of pulmonary fibrosis, and assessing the severity of COVID-19.

[0031] Furthermore, the expression level of CD163 was positively correlated with the degree of COVID-19 pulmonary fibrosis.

[0032] The beneficial effects of the present invention are:

[0033] 1. The present invention provides a marker that can predict the degree of pulmonary fibrosis and / or evaluate the prognosis of pulmonary fibrosis, including CD163, DNAJB11, EIF6, ITGAX, REEP5, RNASE3, SFTPB and TIMM8B; then a pulmonary fibrosis risk prediction model is constructed based on the marker. The pulmonary fibrosis risk prediction model has good predictive efficiency, can predict the degree of pulmonary fibrosis in patients with new emerging infectious diseases and viral pneumonia (such as IPF patients, COVID-19 pulmonary fibrosis patients, etc.), and make a prognostic assessment of their pulmonary fibrosis degree; the higher the risk value obtained by the prognostic risk model, the more severe the patient's pulmonary fibrosis, the worse the prognosis, and the shorter the survival time.

[0034] 2. The pulmonary fibrosis prognosis risk model constructed by the present invention can evaluate the degree of COVID-19 pulmonary fibrosis. The higher the risk value, the more severe the pulmonary fibrosis.

[0035] 3. The present study found that CD163 is correlated with the degree of COVID-19 pulmonary fibrosis. The higher the CD163 expression, the higher the degree of pulmonary fibrosis in COVID-19 patients and the more severe the condition.

[0036] 4. The pulmonary fibrosis prognosis risk model of the present invention can screen the population that is more susceptible to pulmonary fibrosis at the genetic level, which can effectively assist in the clinical identification of potential severe patients and provide a basis for early precise and personalized treatment.

[0037] 5. Compared with relying on tissue biopsy to obtain transcriptome data of the patient's lungs, the pulmonary fibrosis prognosis risk model of the present invention can predict the prognosis of pulmonary fibrosis patients by collecting blood samples; blood tests are easier to carry out, and the operability and feasibility of this prediction model are better. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a statistical diagram of the TOP10 significantly up-regulated and significantly down-regulated genes in IPF (IPF:Control=176:20) relative to normal lung tissue, where red indicates up-regulated and blue indicates down-regulated;

[0039] Figure 2 The statistical chart shows the top 10 significantly upregulated and downregulated genes in COVID-19 compared with normal lung tissue, where red indicates upregulated and blue indicates downregulated; COVID-19: Control = 8: 8;

[0040] Figure 3 The Venn diagram of the number of genes / proteins that are significantly upregulated and downregulated in IPF and COVID-19, where: Figure 3 -A is a Venn diagram of the number of genes / proteins that are significantly upregulated in IPF and COVID-19. Figure 3 -B is a Venn diagram of the number of genes / proteins that are significantly downregulated by IPF and COVID-19;

[0041] Figure 4 KM survival curve for the training set;

[0042] Figure 5 This is the ROC curve of the training set;

[0043] Figure 6 KM survival curve for the test set;

[0044] Figure 7 This is the ROC curve of the test set;

[0045] Figure 8 This is the KM survival curve of the validation set;

[0046] Fig. 9 This is the ROC curve of the validation set;

[0047] Fig.10 This is the GO-BP enrichment result diagram of differentially up-regulated genes between high-risk and low-risk groups in GSE70866;

[0048] Fig.11 This is a graph showing the difference in mean fibrosis percentage between high- and low-risk groups, where low:high = 7:7;

[0049] Fig.12 is a graph of the risk value and the correlation risk with the mean lung fibrosis percentage, n = 14;

[0050] Fig.13 This is a graph showing the difference in median fibrosis percentage between high- and low-risk groups, where low:high = 7:7;

[0051] Fig.14 is a graph of the risk value and the median lung fibrosis percentage, n = 14;

[0052] Fig.15 HE staining of COVID-19 lung tissue, Fig.15 -A and Fig.15 -B are representative images of HE staining of different patients;

[0053] Fig.16 This is a graph showing the GO-BP enrichment results of differentially upregulated genes between high- and low-risk groups for pneumonia caused by novel coronavirus infection;

[0054] Fig.17 This is a graph showing the difference in risk values ​​between COVID-19 lung tissue and normal lung tissue, where Control:COVID-19=8:8;

[0055] Fig.18 This is a graph showing the difference in risk values ​​between severe / non-severe COVID-19 patients, where COVID-19-No-ICU:COVID-19-ICU=50:50;

[0056] Fig.19 Heat map of the correlation between 8 genes and COVID-19 pulmonary fibrosis, n = 14;

[0057] Fig. 20 is a graph showing the correlation between CD163 and the mean lung fibrosis percentage of COVID-19, n=14;

[0058] Fig.21 is a graph showing the correlation between CD163 and median lung fibrosis percentage in COVID-19, n=14;

[0059] Fig. 22 This is a graph showing the difference in risk values ​​between COVID-19 lung tissue and normal lung tissue, where Control:COVID-19=8:8;

[0060] Fig.23 This is a graph showing the difference in risk values ​​between severe / non-severe COVID-19 patients, where COVID-19-No-ICU:COVID-19-ICU=50:50;

[0061] Fig.24 This is the CD163 immunohistochemical staining of COVID-19 lung tissue and normal lung tissue, among which, Fig.24 -A is the CD163 immunohistochemical staining of normal lung tissue, Fig.24 -B is the CD163 immunohistochemical staining of COVID-19 lung tissue;

[0062] Fig.25 This is a graph showing the difference in the percentage of CD163 immunohistochemical staining positive cells between COVID-19 lung tissue and normal lung tissue, where Control:COVID-19=8:12. DETAILED DESCRIPTION

[0063] The technical solution of the present invention will be further described clearly and completely in conjunction with specific embodiments below. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Therefore, based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work belong to the protection scope of the present invention.

[0064] In the embodiment of the present invention, the hematoxylin and eosin staining (HE) method is as follows:

[0065] (1) Paraffin tissue sections: After 16 paraffin-fixed samples were obtained from the biological sample bank, they were first placed in a -20°C refrigerator for 30 minutes, and then sliced, fished, mounted, and baked. The tissues were attached to MMI membrane slides and adhesive slides. The slices were 5 μm thick and stored at room temperature.

[0066] (2) Dewaxing: Bake the paraffin tissue sections in an oven at 60°C for 30 min. After baking, quickly place the sections in xylene I and xylene II in sequence for 15 min each time.

[0067] (3) Hydration: The sections were placed in 100%, 100%, 95%, 85%, and 75% ethanol in sequence for 5 min each time, and then washed with ultrapure water three times for 5 min each time;

[0068] (4) Hematoxylin staining: Place the sections in hematoxylin stain for 15 seconds, then rinse with tap water for 1 minute, place in 1% hydrochloric acid alcohol for 3 seconds, and then rinse with tap water to turn blue;

[0069] (5) Eosin staining: Place the sections in 0.5% eosin staining solution for 30 seconds;

[0070] (6) Dehydration: Place the sections in 75%, 85%, 95%, 100%, and 100% ethanol in sequence for 5 min each time;

[0071] (7) Transparency: Place the sections in xylene I and xylene II in turn for 5 min each time, and then place them in a ventilated place for 60 min. For the sections attached to the adhesive slides, add 20 μl of neutral gum to each section and seal them with coverslips (no need to seal if attached to MMI membrane slides), and then store them at room temperature.

[0072] In the embodiment of the present invention, the method for laser microdissection of the index proteomics sample is as follows:

[0073] (1) Start the laser microdissection instrument: remove the machine dust cover and start the laser microdissection instrument in the order of the host, control unit, laser generator and software;

[0074] (2) Sample sectioning and collection tube installation: Install HE-stained laser microdissection membrane slides (MMI membrane slides) with tissue attached (the side with the tissue facing up and an adhesive slide on the other side) on the stage, and install the collection tube into the centrifuge tube loading rack and then install it in the designated position of the laser microdissection instrument;

[0075] (3) Correction: Under the microscope, move the sample to a blank area, click the Correct button to correct the white balance, then click the Laser button of the software to correct the laser position, then select the appropriate texture in the area without tissue, click the Cut button to perform a trial cut, and adjust the laser power value according to the cutting situation so that the sample can be completely cut and collected by the collection tube;

[0076] (4) Cutting and collection: Install the laser microdissection membrane with tissue (after HE staining) on ​​the stage in the order of samples, complete white balance and laser position correction, select the lung tissue area (do not collect blood vessels and tracheal tissue), click cut to cut and collect, and collect the cutting picture; after cutting and collecting, keep the collection tube cover upward to close the collection tube and mark it, replace the sample slices and collection tubes, repeat this operation until all samples are cut and collected, and store the samples at room temperature;

[0077] (5) Turn off the laser microdissection instrument: Turn off the laser microdissection instrument in the order of software, laser generator, control unit, and host, and cover the machine with a dust cover.

[0078] In the embodiment of the present invention, quantitative proteomic analysis: the samples collected by laser microdissection are handed over to the staff of the technical service company to complete the identification and quantification of proteins according to the standard process of DIA quantitative proteomic analysis.

[0079] In the embodiment of the present invention, the data set is downloaded: the bulk RNA sequencing data and prognostic information (GSE70866 and GSE93606) of IPF are downloaded from the GEO database (https: / / www.ncbi.nlm.nih.gov / geo / ). GSE70866 contains 176 lung tissue samples of IPF (176 with prognostic information) and 20 lung tissue samples of normal controls; GSE93606 contains 174 whole blood samples of IPF (57 with prognostic information) and 20 whole blood samples of normal controls.

[0080] In the embodiment of the present invention, the construction and evaluation method of the prognostic risk model is as follows:

[0081] (1) A prognostic risk model for idiopathic pulmonary fibrosis was constructed based on the regression coefficients and gene expression levels of multivariate regression analysis. The formula is as follows:

[0082]

[0083] Among them, “Riskscore” represents the risk value, “n” represents the number of genes, “βi” represents the multivariate COX regression coefficient of gene i, and “Expi” represents the expression value of gene i.

[0084] (2) The KM survival curve was drawn using the R language package Survival package and the ROC curve was drawn using the R language package Survival ROC (version 1.0.3.1) package. The above risk model was evaluated on 124 samples in GSE70866 (training set), 52 samples in GSE70866 (test set), and 57 samples in GSE93606 (validation set).

[0085] In the embodiment of the present invention, the gene enrichment analysis method is as follows:

[0086] (1) The gene expression data of 176 cases in the GSE70866 dataset were divided into high-risk and low-risk groups according to the median risk value. The R language package Limma package was used for differential analysis, and the genes that were significantly upregulated in the high-risk group compared with the low-risk group were screened (p < 0.05);

[0087] (2) The R language package ClusterProfiler (version 4.8.2) was used to perform GO-BP pathway enrichment analysis (p<0.05) using the significantly differentially expressed genes screened as input genes.

[0088] In the present embodiment, statistical analysis: KM survival analysis uses Log-rank test to test the survival difference between groups. The specificity and sensitivity of the risk model in predicting survival time are evaluated by ROC curve, and the area under the ROC curve (AUC) is quantified using R language package SurvivalROC. p<0.05 is considered to be statistically significant.

[0089] Example 1. Screening of common differential genes / proteins between IPF and COVID-19

[0090] In order to screen the common differential genes / proteins between IPF and COVID-19 and understand the similarities between the two diseases at the molecular level, the present invention performed differential analysis on the transcriptome sequencing data of GSE70866 and the proteomics data of COVID-19 lung tissues (p<0.05). The specific method is as follows:

[0091] (1) The R language package Sva (version 3.48.0) was used to perform batch correction on the expression profile data of GSE70866 (from three sampling locations, with batch effects), and then the R language package Limma (version 3.56.2) was used to perform differential analysis on the transcriptome sequencing data, and the significantly differentially expressed genes between IPF and normal lung tissue were screened (R version 4.3.1) (p < 0.05);

[0092] (2) The R language package Limma was used to perform differential analysis on COVID-19 protein expression data and screen out the significantly differential proteins between COVID-19 and normal lung tissue (p < 0.05);

[0093] (3) The significantly upregulated proteins in the protein expression profile of COVID-19 were intersected with the significantly upregulated genes in GSE70866, and the significantly downregulated proteins in the proteomics of COVID-19 were intersected with the significantly downregulated genes in GSE70866.

[0094] The results showed that in the GSE70866 dataset, IPF lung tissue significantly upregulated 7745 genes and significantly downregulated 849 genes relative to normal lung tissue (IPF: Control = 176: 20). COVID-19 lung tissue significantly upregulated 671 proteins and significantly downregulated 362 proteins relative to normal lung tissue (COVID-19: Control = 8: 8). There are 157 genes / proteins that are significantly upregulated and 16 genes / proteins that are significantly downregulated in both IPF and COVID-19, as shown in Tables 1-2. Figure 1-Figure 3 .

[0095] Table 1. Summary of differentially upregulated genes / proteins in IPF and COVID-19 (157 genes)

[0096]

[0097]

[0098] Table 2. Summary of common differentially downregulated genes / proteins in IPF and COVID-19 (16 genes)

[0099] ADD1 ATP1B1 DMD EPN1 HLA-DRB1 PECAM1 TJP1 ADD3 CES1 DNAJB2 F11R LGALS3BP REEP5 TMEM123 AOC3 CSPG4

[0100] Example 2. Screening of IPF prognosis-related genes

[0101] In order to screen differentially expressed genes related to the prognosis of IPF patients, the present invention performed a single factor COX regression analysis on 157 significantly up-regulated genes / proteins and 16 significantly down-regulated genes / proteins. The specific method is as follows:

[0102] (1) The intersection of significantly up-regulated genes / proteins and the intersection of significantly down-regulated genes / proteins were combined as candidate genes; the 176 samples of the GSE70866 dataset were randomly divided into two groups (124:52) in a ratio of 7:3 (random number: 64475), 124 samples were used as the training set to screen genes and construct a prognostic risk model, and the remaining 52 samples were used as the test set;

[0103] (2) The R language package Survival (version 3.5.5) was used to perform univariate COX regression analysis (p < 0.05) to screen the candidate genes mentioned above. The R language package Glmnet (version 4.1.7) was then used to perform LASSO regression analysis to further screen the screened genes (Lambda.min = 0.0667). Finally, the remaining genes were subjected to multivariate COX regression analysis (p < 0.05) to screen out genes related to the prognosis of IPF patients.

[0104] The results are shown in Table 3. There are 53 genes positively correlated with the prognosis of patients (HR>1, p<0.05), and 5 genes negatively correlated with the prognosis of patients (HR<1, p<0.05). Then LASSO regression analysis was performed to screen out 16 genes further associated with the prognosis of IPF / COVID-19 patients. Finally, multivariate COX regression analysis was performed to screen out the eight genes most associated with the prognosis of IPF patients, namely, CD163 (Cluster of differentiation antigen 163), DNAJB11 (DNAJ heat shock protein family menber Bl1), EIF6 (Eukaryotic translation initiation factor 6), ITGAX (Integrinsubunit alpha X), REEP5 (Receptor accessory protein 5), RNASE3 (Ribonuclease A family member 3), SFTPB (Surfactant protein B), and TIMM8B (Translocase of innermitochondrial membrane 8homolog B).

[0105] Table 3. LASSO and COX regression analysis of common differentially expressed genes / proteins between IPF and COVID-19

[0106]

[0107] Note: β in the table refers to Regression coefficient; p refers to P value.

[0108] Example 3. Construction and evaluation of a prognostic risk model for pulmonary fibrosis

[0109] The present invention further constructs a pulmonary fibrosis prognosis risk model based on the regression coefficient and gene expression of the multifactor COX regression analysis in Example 2, and its formula is: Riskscore = 0.821 × Exp (CD163) + 1.49 × Exp (DNAJB11) - 0.937 × Exp (EIF6) + 0.682 × Exp (ITGAX) - 0.828 × Exp (REEP5) + 0.295 × Exp (RNASE3) + 0.125 × Exp (SFTPB) + 1.143 × Exp (TIMM8B). In order to evaluate the value of the prognostic risk model, the present invention randomly divides the GSE70866 data set into a ratio of 7: 3 (124 samples as a training set and 52 samples as a test set). And select 57 samples in the GSE93606 data set as a validation set to evaluate the prognostic value of the model. First, the sample risk value was calculated according to the above formula, and the patients were divided into high-risk and low-risk groups according to the median risk value (2.21E-08). The KM survival curve was used to evaluate the difference in survival time between the two groups, and the area under the curve (AUC) of the ROC curve was used to evaluate the accuracy of the prediction.

[0110] Results: Figure 4-Figure 5 As shown in the figure, in the training set, there are 62 samples belonging to the high-risk group and 62 samples belonging to the low-risk group. There is a significant survival difference between the two groups (p<0.0001), and the ROC curve shows AUC=0.845. Figure 6-Figure 7 As shown in , in the test set, there are 26 samples belonging to the high-risk group and 26 samples belonging to the low-risk group. There is a significant survival difference between the two groups (p = 0.0065), and the ROC curve shows AUC = 0.836. Figure 8-Figure 9 As shown, in the validation set, there were 28 samples belonging to the high-risk group and 29 samples belonging to the low-risk group. There was a significant survival difference between the two groups (p=0.0062), and the ROC curve showed AUC=0.701.

[0111] The above results show that the pulmonary fibrosis prognostic risk model can effectively assess the prognostic risk of IPF patients.

[0112] Example 4. Biological process enrichment analysis of significantly upregulated genes in the IPF high-risk group

[0113] In order to study the biological processes with significant differences between the high-risk group and the low-risk group judged by the model, the present invention divided the 176 patients of GSE70866 into a high-risk group and a low-risk group (88:88) according to the median risk value, analyzed the differentially expressed genes in the high-risk group compared with the low-risk group (p<0.05), and found 3448 differentially up-regulated genes and 1094 differentially down-regulated genes. GO biological process (BP) enrichment analysis of the differentially up-regulated genes revealed that these genes were mainly enriched in positive regulation of cell adhesion, positive regulation of cytokine production, regulation of cell-cell adhesion, leukocyte migration, cellular response to chemical stress, cellular response to biotic stimulus, cellular response to molecule of bacterial origin, leukocyte chemotaxis, cellular response to lipopolysaccharide, and apoptotic mitochondrial changes (high:low=88:88). For details, see Fig.10 The results suggest that these biological processes may be associated with the prognosis of IPF patients.

[0114] Example 5. Application of prognostic risk model in assessing COVID-19 pulmonary fibrosis

[0115] The bulk RNA sequencing data of 14 COVID-19 patients were extracted, and the pulmonary fibrosis prognostic risk model constructed in Example 3 was used to calculate the risk value of each sample; then the 14 patients were divided into two groups (7:7) according to the median risk value, as shown in Table 4. Then, all lung tissue HE staining sections of the 14 COVID-19 patients were used, and the lung lesion identification and quantification system was used to quantify the degree of pulmonary fibrosis in the 14 COVID-19 patients. The degree of pulmonary fibrosis was expressed as the percentage of the pulmonary fibrosis area in the section to the lung tissue area in the entire section, see Table 4 for details. Difference analysis and correlation analysis were then performed. The results showed that the average percentage of pulmonary fibrosis in the high-risk group was significantly higher than that in the low-risk group (Wilcoxon, p=0.0012) (low:high=7:7), see Fig.11 ; The risk value was correlated with the average percentage of lung fibrosis (Spearman, R = 0.66, p = 0.012) (n = 14), see Fig.12 The median lung fibrosis percentage in the high-risk group was significantly higher than that in the low-risk group (Wilcoxon, p = 0.00058) (low:high = 7:7), see Fig.13 ; The risk value was correlated with the median percentage of pulmonary fibrosis (Spearman, R = 0.7, p = 0.0069) (n = 14), see Fig.14 HE staining results of COVID-19 lung tissue Fig.15 The results show that there is a significant difference in the degree of pulmonary fibrosis between the high-risk and low-risk groups judged by the prognostic model. The higher the risk value, the higher the degree of pulmonary fibrosis; the risk value of the prognostic risk model is positively correlated with the degree of pulmonary fibrosis. The pulmonary fibrosis prognostic risk model can be used to assess the degree of pulmonary fibrosis in COVID-19.

[0116] Table 4. COVID-19 lung fibrosis percentage and prognostic risk model

[0117] ID Average Fibrosis Ratio(%) Median Fibrosis Ratio(%) RiskScore Group 1 45.78 45.97 2.81E-08 High 2 51.01 49.22 1.90E-08 High 3 45.4 44.48 8.75E-09 High 4 43.9 44.02 7.54E-09 High 5 46.35 46.72 2.56E-09 High 6 55.28 55.34 2.45E-09 High 7 45.8 45.08 9.41E-10 High 8 40.89 38.48 8.12E-10 Low 9 42.49 42.97 5.04E-10 Low 10 31.85 31.73 3.24E-10 Low 11 44.08 43.91 2.54E-10 Low 12 39.33 38.62 4.65E-11 Low 13 43.68 43.25 3.15E-11 Low 14 42.06 42.57 1.34E-12 Low

[0118] Note: Average Fibrosis Ratio (%): The average percentage of pulmonary fibrosis area in all lung tissue HE-stained sections of the patient; Median Fibrosis Ratio (%): The median percentage of pulmonary fibrosis area in all lung tissue HE-stained sections of the patient.

[0119] Example 6. Biological process enrichment analysis of significantly upregulated genes in the COVID-19 high-risk group

[0120] In order to study the biological processes with significant differences between the high-risk group and the low-risk group of COVID-19 determined by the model, this example further analyzed the differential genes between the high-risk group and the low-risk group (p<0.05), and found 3423 differentially up-regulated genes and 27 differentially down-regulated genes. GO biological process (BP) enrichment analysis of differentially upregulated genes revealed that these genes were mainly enriched in glycoprotein metabolic process, extracellular matrix organization, extracellular structure organization, external encapsulating structure organization, Golgi vesicle transport, cytoplasmic translation, antigen processing and presentation, proteoglycan metabolic process, collagen fibril organization, and antigen processing and presentation of peptide antigen. This indicates that the biological process of extracellular matrix tissue synthesis in the high-risk group was significantly upregulated, which further illustrates that the model can predict the degree of COVID-19 pulmonary fibrosis (high: low = 7: 7), see for details. Fig.16 .

[0121] Example 7. Prognostic risk model can assess the severity of COVID-19 disease

[0122] To further verify the effectiveness of the prognostic risk model in COVID-19, this example uses the lung proteomics dataset of COVID-19 (Control: COVID-19 = 8:8) and the plasma expression profile data of GSE157103 (COVID-19-No-ICU: COVID-19-ICU = 50:50), and then calculates the risk values ​​of 16 samples in the proteomics data and 100 samples in GSE157103, and finally performs a differential analysis. The results show that the risk value of the COVID-19 group is significantly higher than that of the control group (Wilcoxon, p = 0.00062) (Control: COVID-19 = 8:8), see for details. Fig.17 ; The risk value of severe patients (COVID-19-ICU) was significantly higher than that of mild patients (COVID-19-No-ICU) (Wilcoxon, p = 0.0012) COVID-19-No-ICU: COVID-19-ICU = 50:50), see Fig.18 . The results show that the risk values ​​calculated by the prognostic risk model are significantly different between COVID-19 patients and normal people; there are significant differences between severe COVID-19 (COVID-19-ICU) and mild COVID-19 (COVID-19-No-ICU). That is, the model is conducive to identifying severe patients with pneumonia caused by new coronavirus infection, and may also indicate that patients with severe pneumonia caused by new coronavirus infection may face a higher risk of pulmonary fibrosis. At the same time, GSE157103 is plasma bulk RNA sequencing data, which means that blood samples can be collected to predict the severity of the disease in COVID-19 patients.

[0123] Example 8. CD163 is associated with the degree of COVID-19 pulmonary fibrosis

[0124] In order to clarify the relationship between each of the 8 genes in the prognostic risk model and COVID-19 pulmonary fibrosis, this example uses Spearman correlation to analyze the relationship between the 8 genes and the degree of COVID-19 pulmonary fibrosis (n = 14). The results showed that CD163 was significantly correlated with the degree of COVID-19 pulmonary fibrosis, DNAJB11 was possibly correlated with COVID-19 pulmonary fibrosis, and EIF6, ITGAX, REEP5, RNASE3, SFTPB and TIMM8B were not found to be correlated based on existing data (n = 14), see Fig.19 The expression of CD163 was significantly correlated with the average percentage of lung fibrosis in COVID-19 (Spearman, R=0.71, p=0.0064) (n=14), see Fig. 20CD163 was significantly correlated with the median lung fibrosis percentage in COVID-19 (Spearman, R=0.73, p=0.0045) (n=14), see Fig.21 .

[0125] Example 9. Differential expression of CD163 in COVID-19

[0126] To further verify the differential expression of CD163 in COVID-19, this example performed a differential analysis of the expression of CD163 in two datasets: the COVID-19 lung proteomics dataset (Control: COVID-19 = 8: 8) and the GSE157103 plasma bulk RNA sequencing dataset (COVID-19-No-ICU: COVID-19-ICU = 50: 50). The results showed that the expression of CD163 in COVID-19 lungs was significantly higher than that in normal lung tissues (Wilcoxon, p = 0.00016) (Control: COVID-19 = 8: 8), see Fig. 22 The expression of plasma CD163 in patients with severe COVID-19 (COVID-19-ICU) was significantly higher than that in patients with mild COVID-19 (COVID-19-No-ICU) (Wilcoxon, p=0.0023) (COVID-19-No-ICU:COVID-19-ICU=50:50), see Fig.23 Immunohistochemistry results Fig.24 and Fig.25 As shown, the expression of CD163 in COVID-19 lung tissue was significantly higher than that in the normal lung group (Wilcoxon, p=0.003) (Control:COVID-19=8:12).

[0127] In summary, the present invention first screens out genes / proteins with significant differential expression in both IPF and COVID-19 diseases, and then further screens to obtain 8 prognosis-related genes, including CD163, DNAJB11, EIF6, ITGAX, REEP5, RNASE3, SFTPB and TIM8B8, and constructs a pulmonary fibrosis risk model based on the multifactor regression coefficient and gene expression of these 8 genes. The model can calculate the patient's risk value based on the patient's gene expression, and the higher the risk value, the worse the patient's prognosis. To verify the effectiveness of the model, the present invention verifies the effectiveness of the model in an external independent data set GSE93606 (blood transcriptomic sequencing data of IPF patients), and finds that the model is effective. The pulmonary fibrosis risk model of the present invention provides preliminary evidence for assessing the prognosis risk of IPF patients from a molecular level. And because GSE93606 is blood transcriptomic sequencing data, compared to relying on tissue biopsy to obtain transcriptome data of the patient's lungs, blood tests are easier to carry out, and the operability and feasibility of this model are better.

[0128] The eight prognosis-related genes also provide theoretical support for the present invention to carry out targeted research on these genes and pulmonary fibrosis. The present invention conducted an enrichment analysis of biological processes on the significantly upregulated genes in the IPF high-risk group compared with the low-risk group judged by the risk model. It was found that the main enrichment was in pathways such as positive regulation of cell adhesion, positive regulation of cytokine production, leukocyte migration, and mitochondrial apoptosis, suggesting that these biological processes may be related to the prognosis of IPF patients, but their specific biological mechanisms need further study. Subsequently, the present invention conducted correlation analysis on the above-mentioned eight common differential genes of IPF and COVID-19 with the degree of pulmonary fibrosis. It was found that CD163 was positively correlated with the degree of pulmonary fibrosis of COVID-19. The expression of CD163 was analyzed in the COVID-19 proteomics data, and it was found that the expression of CD163 in the COVID-19 group was significantly higher than that in the control group; the expression of CD163 was analyzed in the GSE157103 (COVID-19) plasma bulk RNA sequencing data, and it was found that the expression of plasma CD163 in patients with severe COVID-19 was significantly higher than that in patients with mild COVID-19. This shows that CD163 is related to the severity of COVID-19. This further illustrates the necessity of conducting targeted research on CD163 and pulmonary fibrosis.

[0129] In conclusion, the pulmonary fibrosis prognostic risk model can identify the degree of pulmonary fibrosis, providing molecular support for a more accurate and timely understanding of the degree of pulmonary fibrosis in patients with emerging infectious diseases and viral pneumonia. The pulmonary fibrosis prognostic risk model can then predict the severity of disease in patients with emerging infectious diseases and viral pneumonia by collecting blood samples for bulk RNA sequencing data, providing a new basis for further early intervention measures, and ultimately playing an important role in improving patient prognosis and prolonging survival.

Claims

1. A marker for predicting the degree of pulmonary fibrosis and / or assessing the prognosis of pulmonary fibrosis, characterized in that: The markers include any one or more of CD163, DNAJB11, EIF6, ITGAX, REEP5, RNASE3, SFTPB, and TIMM8B.

2. The marker according to claim 1, characterized in that The pulmonary fibrosis includes idiopathic pulmonary fibrosis and / or COVID-19 pulmonary fibrosis.

3. Use of the marker described in any one of claims 1-2 in preparing a prognosis risk prediction model for pulmonary fibrosis.

4. Use of a reagent for detecting the expression amount of the marker described in any one of claims 1 to 2 in the preparation of a product for predicting the degree of pulmonary fibrosis and / or evaluating the prognosis of pulmonary fibrosis.

5. A pulmonary fibrosis prognosis risk prediction model, characterized in that: The pulmonary fibrosis prognosis risk prediction model is constructed based on the markers described in any one of claims 1-2.

6. The pulmonary fibrosis prognosis risk prediction model according to claim 5, characterized in that: The formula of the pulmonary fibrosis prognosis risk prediction model is: Riskscore=0.821×Exp(CD163)+1.49×Exp(DNAJB11)-0.937×Exp(EIF6)+0.682×Exp(ITGAX)-0.828×Exp(REEP5)+0.295×Exp(RNASE3)+0.125×Exp(SFTPB)+1.143×Exp(TIMM8B).

7. The pulmonary fibrosis prognosis risk prediction model according to claim 6, characterized in that: The determination of Riskscore value is based on the median of a group of samples. If the Riskscore value is greater than or equal to the median of the group of samples, it is judged as a high-risk state; conversely, if the Riskscore value is less than the median of the group of samples, it is judged as a low-risk state. The higher the risk value, the more severe the patient's pulmonary fibrosis, the worse the prognosis, and the shorter the survival time.

8. The pulmonary fibrosis prognosis risk prediction model according to claim 5, characterized in that: The test sample includes any one or more of a blood sample and a tissue sample.

9. Use of the pulmonary fibrosis prognosis risk prediction model according to any one of claims 5 to 8 in the preparation of a product for predicting the degree of pulmonary fibrosis, assessing the prognosis of pulmonary fibrosis, and assessing the severity of COVID-19.

10. The use according to claim 9, characterized in that: The expression level of CD163 was positively correlated with the degree of COVID-19 pulmonary fibrosis.

Images