| ã€Background】Multidrug resistance is usually one of the main causes for failure ofchemotherapy against malignant tumors, including gastric cancer. The MDRmechanisms in gastric cancer cells have been broadly investigated in ourlaboratory and elsewhere, yet they have not been fully elucidated, indicatingthat other unknown molecules or pathways may be involved in thedevelopment of MDR.In mammalian cells, eukaryotic translation initiation factor2α subunit(eIF2α) is phosphorylated by different eIF2α kinases in response to differentstress signals, including anoxia/hypoxia, endoplasmic reticulum stress, aminoacid deprivation, and oxidative stress. This phosphorylation event leads to arapid decrease in global protein biosynthesis concurrent with induction oftranslational expression of genes, including ATF4that function to alleviatecellular damage from stress. Although ATF4may play a pro-apoptotic role under conditions of severe or prolonged stress, ATF4is a potentstress-responsive gene thought to play a protective role by regulating cellularadaptation to adverse circumstances in the integrated stress response (ISR).Recently, overexpression of ATF4was reported to be prominent in a widevariety of tumors and to protect tumor cells against multiple stresses, as wellas a range of cancer therapeutic agents. The potential mechanisms responsiblefor this protection include autophagy induction, promotion of DNA damagerepair, and up-regulation of intracellular glutathione. However, the expressionand function of ATF4in gastric cancer MDR remains unknown.ã€Aims】To investigate the functions and mechanisms of ATF4in gastric cancerMDR.ã€Methods】1. The role of ATF4in gastric cancer MDR1) ATF4levels were detected by Western blot and qPCR in the SGC7901cell line and its MDR variants, SGC7901/VCR and SGC7901/ADR.2)Lentiviral vectors encoding siRNA specific to ATF4and control siRNA weregenerated with the use of PLKO.1-TRC (Addgene) and were designated asLV-siATF4and LV-SCR control, respectively. Lentiviral vector encodinghuman ATF4gene were constructed in FUW-teto (Addgene), designated asLV-ATF4. The empty vector was used as negative control, designated asLV-Vector.3) Stable cell lines were generated by transfection of indicatedlentiviral constructs followed by selection in puromycin or zeocin (Invitrogen),respectively.4) The response of LV-Vector and LV-ATF4stably transfectedSGC7901and AGS, and LV-SCR and LV-siATF4stably transfectedSGC7901/ADR and SGC7901/VCR cells to cisplatin was tested by colony formation assay.5) In vitro drug sensitivity of LV-Vector and LV-ATF4stablytransfected SGC7901cell lines, and LV-SCR and LV-siATF4stablytransfected SGC7901/ADR cells was tested by MTT assay.6) Hoechst33258nuclear staining, DNA fragmentation assay and Western blot for cleaved formcaspase-3were used to investigate the role of ATF4on the sensitivity ofLV-Vector and LV-ATF4stably transfected SGC7901cell lines, LV-SCR andLV-siATF4stably transfected SGC7901/ADR cells and their respectivenontreated counterparts (NC) under cisplatin induced apoptosis.2. The transcription study of the relationship between SIRT1and ATF47) mRNA and protein levels of SIRT1in LV-Vector and LV-ATF4stablytransfected SGC7901cell lines and LV-SCR and LV-siATF4stably transfectedSGC7901/ADR cells were assayed by qPCR and Western blot.8) Cell lysatesfrom LV-Vector and LV-ATF4stably transfected SGC7901cell lines andLV-SCR and LV-siATF4stably transfected SGC7901/ADR cells and theirrespective nontreated counterparts (NC) were blotted with antibodies againstMDR1, MRP, Bcl-2and Bax.9) The5’-flanking sequence of the SIRT1genewas analyzed with bioinformatics softwares (Tfsitescan service, TESS, andGenomatix).10) The1.2kb human SIRT1promoter sequence (-1100to+100bp) was synthesized and cloned into the XhoI and HindIII sites of thepGL3-Basic vector.11) Luciferase reporter assay were used to detect the roleof ATF4in regulating SIRT1promoter activity.12) ChIP assay was used todetect the direct binding of ATF4to the SIRT1promoter.13) Site-directedmutagenesis was used to mutate these two binding sites.14) Luciferasereporter assay were used to investigate the role of the two ATF4binding sitesin regulating SIRT1transactivation.3. The role of SIRT1in the ATF4-induced gastric cancer MDR 15) The response of ATF4stably transfected gastric cancer cells(SGC7901and AGS) after transfection of SIRT1siRNA or scrambled siRNAto cisplatin was tested by colony formation assay.16) In vitro drug sensitivityin ATF4stably transfected SGC7901cells after transfection of SIRT1siRNAor scrambled siRNA were tested by MTT assay.17) Flow cytometry assay,Hoechst33258nuclear staining, and Western blot for cleaved form caspase-3were used to investigate the role of SIRT1on the sensitivity of LV-ATF4stably transfected SGC7901cells under cisplatin induced apoptosis.18)MDR1, MRP, Bcl-2, and Bax expression levels in SGC7901-ATF4cellsfollowing transfection with SIRT1siRNA or scrambled siRNA were examinedby Western blot.19) The basal cytotoxicity of EX-527in LV-Vector andLV-ATF4stably transfected SGC7901cells was examined by MTT assay.20)SIRT1, ATF4, MDR1, MRP, Bcl-2, and Bax expression levels after24hours’incubation with or without the indicated doses of EX-527in LV-Vector andLV-ATF4stably transfected SGC7901cells were detected by Western blot.21)The response of LV-ATF4stably transfected SGC7901cells after preincubatedwith the indicated doses of EX-527for24h to cisplatin or5-fluorouracil wastested by MTT assay.22) The possible synergistic effect of EX-527ondifferent doses of CDDP-and5-FU-mediated inhibition of cell proliferation inSGC7901-ATF4cells was also determined by MTT assay.ã€Results】1. ATF4confers a MDR phenotype to gastric cancer cells and thattargeting ATF4could reverse the resistance of MDR variants in response tochemotherapy.1) Both protein and mRNA levels of ATF4were much higher in theresistant cell lines than in parental cells.2) Lentiviral vectors were successfully established.3) LV-Vector and LV-ATF4stably transfected SGC7901and AGScell lines, and LV-SCR and LV-siATF4stably transfected SGC7901/ADR andSGC7901/VCR cells were successfully established.4) ATF4overexpressionresulted in a nearly3-fold increase in colony numbers compared with emptyvector-expressing cells. While knockdown of ATF4by siRNA in theSGC7901/ADR and SGC7901/VCR cells led to a2-to3-fold reduction in cellnumber when used in combination with CDDP.5) The results of MTT assayindicated that the IC50values of SGC7901-ATF4for ADR, VCR, CDDP, and5-FU were significantly increased as compared to empty vector transfectedcells. In contrast, down-regulation of ATF4significantly reversed the resistanceof SGC7901/ADR cells in response to chemotherapy.6) Treatment ofSGC7901-ATF4and SGC7901/ADR-SCR cells with the indicatedconcentrations of CDDP for36hours did not induce any apoptosis, as assessedby Hoechst nuclear staining and DNA fragmentation assay. In contrast,SGC7901-Vector and SGC7901/ADR-siATF4cells displayed significantapoptosis, with the more frequent appearance of condensed and fragmentednuclei and DNA ladder formation. Moreover, more obvious cleavage ofprocaspase-3was observed after treatment with CDDP in SGC7901-Vector andSGC7901/ADR-siATF4cells as compared to SGC7901-ATF4andSGC7901/ADR-SCR cells, respectively.2. The stress-related SIRT1expression is transactivated by ATF4.7) The overexpression of ATF4was associated with increased SIRT1expression at both the transcriptional and translational levels. In contrast,siRNA knockdown of ATF4in SGC7901/ADR cells resulted in a significantreduction of endogenous SIRT1expression.8) ATF4-proficient cells expressedmore MDR1and an up-regulation of the Bcl-2to Bax ratio as compared to the control cells. Meanwhile, no obvious difference in MRP expression was foundin any of these cell lines.9) Two ATF4putative binding sites were identifiedwithin the-950to-600bp region of the SIRT1promoter with bioinformaticssoftwares (Tfsitescan service, TESS, and Genomatix).10) The1.2kb humanSIRT1promoter sequence (-1100to+100bp) was successfully synthesized andcloned into the XhoI and HindIII sites of the pGL3-Basic vector. The resultingconstruct was confirmed by DNA sequencing.11) The results of luciferasereporter assay showed that the SIRT1promoter activity was markedly activatedby ATF4in a dose-dependent manner.12) ChIP with the ATF4antibody usingSGC7901-ATF4cells showed enrichment of both binding sites within theSIRT1promoter region.13) Site-directed SIRT1promoter reporter mutantswere successfully established.14) The results of luciferase reporter assayshowed that either mutating the binding site1or binding site2reduced theSIRT1promoter activity induced by ATF4. Furthermore, mutation of bothbinding sites abolished the SIRT1promoter activity.3. SIRT1plays a critical role in the ATF4-induced gastric cancer MDR.15) Knockdown of SIRT1by siRNA in SGC7901-ATF4cells led to a>40%reduction in colony number when used in combination with CDDP. Thiseffect was also observed in AGS-ATF4cells.16) The results of MTT assayindicated that knockdown of SIRT1could re-sensitize SGC7901-ATF4cells tochemical drugs, but this did not occur in scrambled siRNA transfected cells.17)After treated with CDDP, the results of Flow cytometry assay and Hoechst33258nuclear staining showed that the apoptotic rate of SIRT1siRNAtransfected SGC7901-ATF4cells was significantly higher than that of thecontrol cells. Moreover, cleavage of procaspase-3was observed as early as12h in SIRT1siRNA transfected SGC7901-ATF4cells, but not in scrambled siRNA treated cells, even after24h of CDDP treatment.18) Down-regulationof MDR1was observed in the SIRT1siRNA-treated cells compared to thecontrol cells. In contrast, no obvious difference of MRP, Bcl-2, and Baxexpression levels were found between the samples.19) The results of MTTassay revealed that EX-527at concentrations up to10μM did not inhibit, butrather slightly increased, the viability of both cell lines.20) Only theexpression of MDR1was down-regulated by EX-527in aconcentration-dependent manner. In contrast, no variation of SIRT1, ATF4,MRP, Bcl-2, or Bax was found after24hours’ incubation with the indicateddoses of EX-527.21) The results of MTT assay showed that EX-527significantly enhanced the cytotoxicity of both cisplatin and5-fluorouracil in adose-dependent manner.22) As expected,10μM EX-527is sufficient topotentiate the cytotoxicity of both drugs.ã€Conclusions】1. ATF4confers a MDR phenotype to gastric cancer cells and thattargeting ATF4provides a method of sensitizing resistant cells to chemicaltreatments.2. ATF4promotes MDR ability of gastric cancer cells through multiplemechanisms.3. SIRT1is a direct transcriptional target of ATF4.4. SIRT1mediates the ATF4-induced MDR effect in gastric cancer cells.5. SIRT1activity also plays a critical role in the ATF4-induced gastriccancer MDR and this role might be mediated partly through MDR1expression. |