AURKA Suppresses Ferroptosis And Promotes Cisplatin Resistance In Oral Squamous Cell Carcinoma

Objective:Considering the high incidence and malignant degree,oral squamous cell carcinoma（OSCC）with poor prognosis has gradually become a major public health problem affecting human health in the world.Cisplatin（CDDP）-based chemotherapy is currently the most commonly used palliative treatment for advanced OSCC.However,many patients develop resistance to cisplatin therapy and even die from drug-related toxicity.As a form of regulated cell death,ferroptosis plays an important role in tumor suppression.In addition,the activation of ferroptosis may also contribute to the treatment of various cancers.At present,there are few molecular studies on the simultaneous regulation of ferroptosis and cisplatin resistance in OSCC,and the mechanisms of action of many potential molecules is also unknown.Aurora kinase A（Aurora A,AURKA）,a serine/threonine kinase,regulates the function of the mitotic spindle in cells.Current studies have found that AURKA promotes the progression of various tumors and plays a potential regulatory role in ferroptosis and chemotherapy resistance.Methods:Part I:1.Downloaded the RNA sequencing data of OSCC from TCGA and GEO databases,and screened out the differentially expressed ferroptosis-related genes（DE-FRGs）between OSCC and normal oral tissues.Meanwhile,DE-FRGs between OSCC cisplatin-resistant and sensitive cells were screened from the GSE117872 cohort from the GEO database.Ferroptosis-related gene modules highly associated with OSCC and OSCC cisplatin resistance were identified from the three cohorts using weighted co-expression network analysis,respectively.The genes common to the above steps were defined as DE-FRGs highly correlated with both OSCC and OSCC cisplatin resistance.2.Kaplan-Meier survival analysis further screened out 4 prognosis related differentially expressed ferroptosis-related genes（PR-DE-FRGs）.In the internal cohort（TCGA）,the time-dependent receiver operating characteristic（ROC）curve,Kaplan-Meier survival curve and concordance index（C-index）were used to evaluate and compare the four PR-DE-FRGs’value in predicting the prognosis of OSCC patients.Univariate Cox regression was used to further determine the influence of 4 PR-DE-FRGs on the prognosis of OSCC patients from the TCGA cohort.After comprehensive comparison,we got the gene（AURKA）with the highest predictive value in the prognosis of OSCC patients.External cohorts（GSE41613GSE65858）were used to further validate the best prognostic value of AURKA in OSCC.3.In the TCGA cohort,the importance scores of the 4 PR-DE-FRGs were ranked using the random forest algorithm while its diagnostic value in OSCC was evaluated separately using the diagnostic ROC curve.External cohort（GSE30784GSE37991）was used to further validate their diagnostic value.The gene（AURKA）with the highest diagnostic value was screened out.4.The expression differences of AURKA between 18 cancers and adjacent tissues in the TCGA database were further analyzed.Multivariate Cox regression was used to determine whether AURKA could independently affect the OSCC patients’prognosis.5.Gene set enrichment analysis was used to explore the possible biological functions of AURKA.6.Used TCGA’s OSCC RNA sequencing data to compare the differences in AURKA expression among different clinical grades,stages and survival status subgroups.We also analyzed the correlation between the expression of AURKA and cell adhesion-related genes.7.We also further analyzed the correlation between AURKA and ferroptosis marker gene expression.8.We also further analyzed the correlation between AURKA and multidrug resistance gene expression.Part II:1.Collected OSCC tissue samples,adjacent normal tissue samples and corresponding clinical information.The RNA and protein expression of AURKA in tissues were detected by quantificational real time polymerase chain reaction（QRT-PCR）and western blot（WB）.We compared the expression difference of AURKA between OSCC and normal oral tissues.In addition,the differences of the relative AURKA’s RNA levels in subgroups of patients with different clinicopathological features were also compared.SCC9 and HSC-3 were infected with lentivirus.The infected cells with the highest infection efficiency were selected and screened with puromycin to obtain cells with stable AURKA’s overexpression.QRT-PCR and WB verified the overexpression of AURKA in the SCC9/HSC-3.3.CCK-8 detected the survival rate of empty,empty virus and AURKA overexpressed SCC9/HSC-3 cells at 24,48 and 72 hours.Then we compared the difference in survival rate among the three groups of cells.4.We also compared the differences in the migration ability assessed by the scratch test among the three groups of cells.5.We also compared the differences in the invasion ability assessed by Transwell assay among the three groups of cells.Part III:1.Compared the differences among the reactive oxygen species and Fe²⁺levels in SCC9/HSC-3 cells with empty,empty virus and AURKA overexpression.2.We further compared the differences among the RNA levels of ferroptosis markers and NRF2-KEAP1 pathway genes（GPX4,SLC7A11,FTH1,NRF2 and KEAP1）detected by QRT-PCR in the three groups of SCC9/HSC-3.Part IV:1.CCK-8 detected the cell viability of CAL27 and CAL27/CDDP treated with different concentrations of CDDP.Next,we draw the growth curves of the two groups of cells,calculated the corresponding IC50,and compared the difference of IC50 between the two groups of cells.2.QRT-PCR and WB were used to detect the RNA and protein levels of AURKA in these two groups of cells,respectively.We compared their differences between the two groups.3.The lentivirus was used to infect CAL27/CDDP again.According to the intensity of fluorescent protein expression,the infected cells with the highest infection efficiency were selected and screened with puromycin to obtain the CAL27/CDDP with stable AURKA’s overexpression.QRT-PCR and WB were again used to verify the overexpression of AURKA in the CAL27/CDDP.4.CCK-8 was used to detect the cell viability of AURKA overexpressed HSC-3 and CAL27/CDDP under different concentrations of cisplatin treatment.We plotted the growth curves of the three groups of cells,calculated the corresponding IC50,and compared the differences of IC50 among the three groups of cells.Results:Part I:1.Six DE-FRGs that were highly correlated with both OSCC and OSCC cisplatin resistance were screened.Kaplan-Meier survival analysis preliminarily identified that AURKA,SLC2A1,RRM2 and DUOX1 had an influence on the prognosis of OSCC patients.AURKA had the highest C-index and highest AUC value among the 4 PR-DE-FRGs.In univariate Cox regression analysis,only AURKA showed a significant effect on prognosis.Similar results were also obtained in the external cohort.2.In the diagnostic ROC curves based on internal and external cohorts,AURKA showed the highest AUC value.3.AURKA was upregulated in OSCC tissues and OSCC cisplatin-resistant cells and was generally upregulated in 16 cancers.After we adjusted relevant clinical factors,AURKA was still an influencing factor for the prognosis of OSCC patients.4.Enrichment analysis observed that the high AURKA expression was related to the functions and pathways related to tumor malignant progression.Low AURKA expression correlated with ferroptosis related function.5.AURKA expression was higher in samples with higher clinical grade.ICAM1/EGFR/CD44 was significantly positively correlated with the AURKA expression.6.GPX4/SLC7A11/FTH1 was significantly positively correlated with the AURKA expression.7.There was a significant positive correlation between AURKA and MRP1（ABCC1）/PGP expression.Part II:1.AURKA was upregulated in the OSCC tissues.The RNA level of AURKA was significantly higher in the tissues of OSCC patients with higher clinical grade.2.The infection efficiency of SCC9/HSC-3was the highest with virus in MOI=20/80.AURKA was significantly and stably overexpressed in SCC9/HSC-3.3.The viability of SCC9/HSC-3 with stable AURKA’s overexpression was significantly higher at 24,48 and 72 hours.4.The migration rate of SCC9/HSC-3 with stable AURKA’s overexpression was significantly faster after scratching.5.The number of SCC9/HSC-3 with stable AURKA’s overexpression across the chamber was significantly higher.Part III:1.The fluorescence intensity reflecting the reactive oxygen species and Fe²⁺levels in SCC9/HSC-3 with stable AURKA’s overexpression was lower.2.The RNA expressions of GPX4,SLC7A11,FTH1 and NRF2 were higher in SCC9/HSC-3 with stable AURKA’s overexpression when the RNA expression of KEAP1 were lower.Part IV:1.Under different concentrations of cisplatin treatment,the CAL27/CDDP viability was significantly higher than that of CAL27.The IC50 of CAL27/CDDP was significantly higher than that of CAL27.3.The RNA and protein expression levels of AURKA in CAL27/CDDP were higher than those in CAL27.4.The infection efficiency of CAL27/CDDP was the highest with virus in MOI=40.AURKA was significantly overexpressed in the CAL27/CDDP with stable AURKA’s overexpression.5.AURKA overexpressed HSC-3 and CAL27/CDDP viability was significantly higher under different concentrations of cisplatin treatment.The corresponding IC50 was also significantly higher.Conclusions:Part I:1.Bioinformatics analysis identified and validated that AURKA had the best diagnostic and prognostic value.This gene was highly associated with OSCC and OSCC cisplatin resistance.2.AURKA was up-regulated in 17 cancers including OSCC and OSCC cisplatin-resistant cells,and could independently and negatively affect OSCC patiednts’prognosis.3.The increased expression of AURKA may promote the malignant progression,inhibit ferroptosis,and promote cisplatin resistance in OSCC.Part II:1.The expression of AURKA was up-regulated in OSCC tissues.The increased expression of AURKA may promote the malignant progression of OSCC.2.The increased expression of AURKA promoted the proliferation of SCC9/HSC-3.3.The increased expression of AURKA promoted the migration of SCC9/HSC-3.4.The increased expression of AURKA promoted the invasion of SCC9/HSC-3.Part III:The increased expression of AURKA inhibited ferroptosis in OSCC,which may be achieved through the NRF2-KEAP1 pathway.Part IV:AURKA was upregulated in CAL27/CDDP.The increased expression of AURKA reduced the sensitivity of OSCC cells to cisplatin and promotes their cisplatin resistance.