Transcriptomic analysis reveals pathophysiological relationship between chronic obstructive pulmonary disease (COPD) and periodontitis

Abstract

Background: The aim of this study was to detect potential crosstalk genes, pathways and immune cells between periodontitis and chronic obstructive pulmonary disease (COPD).

Methods: Chronic periodontitis (CP, GSE156993) and COPD (GSE42057, GSE94916) datasets were downloaded. Differential expressed genes (DEGs; p < 0.05) were assessed and screened for overlapping results, following functional pathway enrichment analyses (p < 0.05). The xCell method was used to assess immune cell infiltration relationship between CP and COPD. Features of the detected cross-talk genes were revealed using conventional Recursive Feature Elimination (RFE) algorithm in R project. Receiver-operating characteristic curves were applied to evaluate the predictive value of the genes. Furthermore, Pearson correlation analysis was performed on crosstalk markers and infiltrating immune cells in CP and COPD, respectively.

Results: A total of 904 DEGs of COPD and 763 DEGs of CP were acquired, showing 22 overlapping DEGs between the two diseases. Thereby 825 nodes and 923 edges were found in the related protein-protein-interaction network. Eight immune cell pairs were found to be highly correlated to both CP and COPD (|correlation coefficients |> 0.5 and p-value < 0.05). Most immune cells were differently expressed between COPD and CP. RFE identified three crosstalk genes, i.e. EPB41L4A-AS1, INSR and R3HDM1. In correlation analysis, INSR was positively correlated with Hepatocytes in CP (r = 0.6714, p = 0.01679) and COPD (r = 0.5209, p < 0.001). R3HDM was positively correlated with Th1 cells in CP (r = 0.6783, p = 0.0153) and COPD (r = 0.4120, p < 0.01).

Conclusion: EPB41L4A-AS1, INSR and R3HDM1 are potential crosstalk genes between COPD and periodontitis. R3HDM was positively correlated with Th1 cells in both diseases, while INSR was positively correlated with Hepatocytes in periodontitis and COPD, supporting a potential pathophysiological relationship between periodontitis and COPD.

Keywords: Bioinformatics; COPD; Inflammation; Periodontitis.

Fig. 1

The framework of the current study

Fig. 2

Principal component analysis (PCA) cluster plot before (A) and after (B) sample merge between GSE42057 and GSE94916

Fig. 3

(A) and (B) show the Volcano map of deregulated genes (DEGs) for periodontitis (CP) and chronic obstructive pulmonary disease (COPD), respectively. Red represents up-regulated differentially expressed genes, grey represents not significantly different genes, and blue represents down-regulated differentially expressed genes. C The overlapped DEGs between CP and COPD

Fig. 4

The significant enriched biological processes (A) and pathways B of 22 potential crosstalk genes. C The protein–protein interaction network for 22 potential crosstalk genes. In the network, the size of node indicated the higher degree of node

Fig. 5

Cell type enrichment analysis in periodontitis (CP) and chronic obstructive pulmonary disease (COPD) for 22 potential crosstalk genes. (A) and (B) Heatmap of immune cell analysis for CP and COPD, respectively. The color legend represented the xCell scores. (C) and (D) the correlation among immune cells in CP and COPD, respectively. Both color and pie chart were corresponding to average Pearson coefficients

Fig. 6

(A) and (B) different immune cell infiltration between disease and normal control samples for periodontitis (CP) and chronic obstructive pulmonary disease (COPD). C Different immune cell infiltration between CP disease and COPD disease samples

Fig. 7

Feature genes selected by using RFE algorithm for periodontitis (CP) (A) and chronic obstructive pulmonary disease (COPD) (B). The abscissa of the figure is the variable of the number of genes, and the ordinate is the exact value of the whole data set measured under this variable. The results show that when the minimum variable is 6, the score is high, which means that 6 features could map the entire dataset. (C) and (D) the receiver operation curve (ROC) of 3 crosstalk marker genes and the combination for the 3 genes based on the average expression value in CP and COPD, respectively

Fig. 8

Correlation between EPB41L4A-AS1, INSR and R3HDM, and immune cells in periodontitis (CP) and chronic obstructive pulmonary disease (COPD). A Correlation between EPB41L4A-AS1 and infiltrating immune cells in CP. B Correlation between INSR and infiltrating immune cells in CP. C Correlation between R3HDM and infiltrating immune cells in CP. D Correlation between EPB41L4A-AS1 and infiltrating immune cells in COPD. E Correlation between INSR and infiltrating immune cells in COPD. F Correlation between R3HDM and infiltrating immune cells in COPD