| Background and purpose:Lung cancer is the leading cause of cancer death in global range,including developed and developing countries,with over a million death tolls worldwide annually.Due to the absence of effective diagnostic tools to detect lungs cancer at early stages,the 5-year mortality rate of lung cancer reached 80%.Traditional pathophysiology studies based on single gene and its transcriptional and protein functional characteristics,explaining life mechanism from the molecular level,but it is only limited to local.The biological network can directly show the interconnections between various functional departments in the biological system,and conduct a thorough and comprehensive study of the biological system as a whole.The weighted gene co-expression network analysis(Weighted gene co-expression network analysis,WGCNA)uses the idea of system biology to find the similarity between genes and determine the highly related genes as a gene module.At present,some studies have supported that the damage of the airway epithelium of lung cancer patients and even some smokers extends to the nose,which suggests that nasal gene expression may be used as a noninvasive biomarker for lung cancer detection.Therefore,identifying co-expressed gene clusters from nasal epithelial cells may be helpful for screening lung cancer patients.Based on the common microarray data set(data set GSE80796,downloaded from GEO),WGCNA was used to explore and calculate the correlation of gene module characteristics,and qRT-PCR technology was used to quantify the expression of gene expression at the top level.Materials and methods:WGCNA analysis based on public microarray dataset GSE80796 and correlation computation of module properties.The dataset contained 505 samples and a total of 32321 genes.5000 genes from 196 samples of nasal mucosa epithelial cells of lung cancer were selected,out of which 96 were samples of nasal mucosa tissue of lung cancer and the other 100 were samples of the normal nasal mucosa.RMA,SVA and T-test from Limma software package were used to screen differentially expressed genes(DEGS),and 3600 differentially expressed genes were obtained.Four different co-expression networks were constructed for the four differentially expressed genes,which were mixed data group Ⅰ,mixed data group Ⅱmale specific data groups and female specific data groups.These were calculated and extracted from the feature gene module of 15,25,10 and 10 key genes,respectively.UseqRT-PCR technology was selected to verify the 12 top hub genes based on mixed data set II WGCNA analysis.The SYBR GREEN I method was used to quantify the expression of 12 related genes in 80 cases of nasal swabs.The experiment was divided into three groups:30 cases in normal contrast group,32 in lung cancer group and 18 in a postoperative group of lung cancer.Fluorescent quantitative PCR technology was used to mark and track the PCR products with fluorescent dyes or fluorescent labeled probes and monitor the reaction process in real time.With the progress of PCR reaction,the reaction products continued to accumulate,and the intensity of fluorescence signal was also increased.After a cycle,a fluorescence intensity signal was collected.In this way,the changes in the amount of the product were monitored through the change of fluorescence intensity.The fluorescence amplification curve was obtained by the corresponding software analysis.The amount of the initial template for the sample to be measured was calculated.When the internal reference gene was introduced and the target gene was detected in the fluorescence quantitative PCR detection system,the relative change of the same sequence was calculated by the 2-△△Ct method.The experimental results were analyzed and compared with the T test(Welch correction and Tamhane’s T2 test)to determine the difference in gene expression of nasal mucosa in lung cancer(P<0.05).Results:1.WGCNA constructed 9 co-expression modules from the differentially expressed genes of mixed data group I,and 5 modules with the strongest association with the state of cancer were black,yellow,green,blue and light blue.The GO pathway enrichment results showed that the genes of the black module were mainly enriched in the GO:0006955(immune response),GO:0034341(to interferon gamma reaction)and GO:0008009(chemokine activity),while the blue module gene mainly enriched the protein transport process GO:0005654(nuclear substance),GO:0005813(centrosome)and GO:00468 72(metal ion binding).Whereas,the genes of the light blue module were mainly enriched in GO:0042384(ciliary assembly)and GO:0060271(cilium form)and the genes of the yellow module were mainly enriched in receptor activation,including G0:0004984(olfactory receptor activation)and GO:0004930(activated by G-protein coupled receptor).IFI44L,THOK2,NEK11 and CCDC144CP were four modules of the top hub genes studied.2.WGCNA constructed four specific LC modules in mixed data group Ⅱ,including blue,brown,yellow and light blue,and a strong association between the blue and brown module genes in their modules,from the differentially expressed genes of the mixed data group two.Results of the current analysis showed that HCK,NCF1,TLR8,EMR3,CSF2RB and DYSF genes were the most important central genes in the brown module,while SPEF2,ANKFN1,HYDIN,DNAH5,C12orf55,and CCDC113 genes were the top key genes in the hierarchy.3.Nasal mucosa fluorescent quantitative PCR showed that out of the 12 top junction gene expression experiments,8 gene expressions were significantly different between lung cancer and control group(0.05%).There were 6 genes(HCK,NCF1,TLR8,EMR3,CSF2RB,DYSF)confirmed between lung cancer and control groups.There were 4 genes(HCK,CSF2RB)showed after lung cancer surgery and control groups.There were 2 genes(EMR3 and C12orf55)detected in lung cancer group and lung cancer groups after operation(SPEF2,C12orf55).The distinct changes detected in gene expression between lung cancer and control group were mucin like hormone receptor gene EMR3,Tamhane’s T2/P value 3.051/0.000,followed by neutrophil cytoplasmic factor 1(NCF1)gene,Tamhane’s T2/P value 2.937/0.001.4.When the cluster tree of male and female specific data sets was analyzed,it was found that based on the specific data set of male and female sample group,the consensus module characteristic gene and clinical characteristics showed significant correlation(P<0.05),and the common understanding module characteristic gene and clinical feature were based on the mixed data set of men and women.The relationship was rnot significantly(P>0.05)correlated.WGCNA of male and female specific expression profiles,black and green modules,which were significantly related to lung cancer status,were detected in male specific data.Similarly,the brown module was detected as a useful module in the specific gene expression profiles of women with lung cancer.Further analysis revealed that 10 key genes were identified from the green gene module that may play a key role in the occurrence of lung cancer.The correlation from strong to weak was as following:SP100,XAF1,EPSTL1,PARP9,APOL6,SAMD9L,MX1,BST2,GBP 1,CMPK2.10 key genes that may play a key role in the occurrence of lung cancer are identified from the brown module as well.PNPLA2 showed the strongest correlation,followed by GAK,RNF31,ATG9A,EPS8L2,LLGL2,RABGGTA,TMEM63B,BRPF1 and AP3D1.The enrichment analysis showed that the enrichment terms in the male data centralized annotation system were mainly related to the immune response and infection,and the enrichment terms in the female data centralized annotation system were mainly related to the olfactory transduction.Conclusion:1.WGCNA can detect the gene module with biological significance,and can dig into the regulation of gene and the effect on the development of the tumor.2.Abnormal expression of HCK,NCF1,TLR8,EMR3,CSF2RB,DYSF,SPEF2,C12orf55 of nasal mucosa epithelial cells may be a noninvasive biomarker for detection of lung cancer,and the abnormal expression of two genes in EMR3 and C12orf55 may be used to observe the therapeutic effect of postoperative lung cancer treatment.3.Based on the specific data set of male and female sample groups,the consensus module characteristic gene and clinical characteristics showed significant correlation.And the relationship between the common module characteristic gene and the clinical features based on the male and female mixed sample data set showed no correlation and thus indicated the heterogeneity of the characteristic module and sex.4.Weighted gene co-expression network analysis was consistent with the results of PCR gene expression. |
PDF Full Text Request