Experimental Studies On The Effects Of SIRT3 In Anti-cancer Drug-induced Cardiotoxicity And Mechanisms

BackgroundAlong with the rapid advancements of effective modern anti-cancer chemotherapies,the undesired cardiotoxicity of anti-cancer chemotherapeutic agents has also gained increasing research attention.Clinically frequent cardiovascular adverse effects of anti-cancer drugs include elevated blood pressure,left ventricular ejection fraction（LVEF）decline,arrhythmia,among others.These adverse effects may often lead to loss of quality of life in chemotherapy-receiving cancer patients,forced discontinuation of chemotherapy in some patients,and may even cause some cancer survivors to die of cardiovascular diseases,posing a tremendous hindrance to the progression of anti-cancer treatment.Despite this,the physiological,pathological and molecular mechanisms of anti-cancer drug-induced cardiotoxicity is vastly unclear.Understanding of the cellular and molecular mechanisms of anti-cancer drug-induced cardiotoxicity and exploration of preventive and therapeutic strategies are of critical significance for providing more comprehensive and long-term medical service for cancer patients.The tyrosine kinase inhibitors（TKIs）are a type of widely-used anti-cancer chemotherapeutic agents,which produce their anti-tumor effects by inhibiting the tyrosine kinases downstream of vascular endothelial growth factors,suppressing tumor angiogenesis and in turn weakening the microvascular blood supply.Unfortunately,because the heart is,similar to tumors,also highly metabolic and heavily dependent on blood supply,the normal physiological functioning of the heart is also profoundly sensitive to TKI-induced inhibition of the microvasculature.Previously existing research have reported that TKIs can induce a series of cardiovascular adverse effects,such as elevated blood pressure and left ventricular dysfunction.Sunitinib is a frequently-used TKI agent that in China is used for the first-line treatment of cancers including renal cell carcinoma.Sunitinib has been clinically reported to induce cardiovascular adverse effects,one of the most common being left ventricular dysfunction.Recent research has demonstrated that injury and loss of cardiac microvascular pericytes and the consequent myocardial microvascular dysfunction is the key pathological alteration in sunitinib-induced left ventricular dysfunction.Pericytes are a group of cells wrapped around the basal membrane of capillaries throughout the body,which wield important regulatory roles in the structure and functions of the microvascular system.Pericytes in the heart are strict regulators of the architecture,diameter,blood flow,and wall permeability of coronary microvessels,and have been shown to participate in a range of cardiovascular conditions.Through potent antagonism of platelet derived growth factor receptor β（PDGFRβ）,sunitinib induces cardiac microvascular pericyte coverage loss,which leads to reduced tortuosity,increased wall permeability of cardiac capillaries and damaged microvascular blood supply,and eventually left ventricular dysfunction.Sirtuin 3（SIRT3）is an essential deacetylase located in the mitochondrial matrix.Targeting a wide range of protein substrates and coordinating a complex signaling network,SIRT3 has been shown to play roles in numerous cardiovascular diseases and tumors.Recent literature reported that SIRT3 inhibited autophagy in human hepatocytes augmenting sensitivity to exogenous lipotoxic agents.Another study also reported that human SIRT3 inhibited the glutathione S-transferase pi 1（GSTP1）/c-Jun N-terminal kinase（JNK）pathway,boosting hepatoma cell sensitivity to TKI agent sorafenib.Furthermore,it was also reported that the GSTP1/JNK pathway contributed to autophagy in human osteosarcoma cells.Nevertheless,there has been no study to date that investigated whether SIRT3 could potentially affect cardiac pericyte autophagy,if so,whether this effect of SIRT3 on autophagy was mediated by the GSTP1/JNK pathway,and whether the SIRT3/GSTP1/JNK/autophagy axis in cardiac pericytes could influence sunitinib-induced left ventricular dysfunction.Thus,the exact impact of SIRT3 on sunitinib-induced left ventricular dysfunction and its underlying mechanisms remain to be elaborated.Based on the above research background,in the present study,using sunitinib-induced left ventricular dysfunction mouse models and sunitinib-induced toxicity pericyte models,we investigated on in vivo and in vitro levels the mechanisms of sunitinib-induced cardiotoxicity as well as the role of SIRT3 in sunitinib-induced left ventricular dysfunction and its cellular and molecular mechanisms.Results from this study may provide experimental evidence for establishing SIRT3 as a novel potential molecular target for prevention and treatment strategies of sunitinib-induced cardiotoxicity.Aims1.To investigate the impact of SIRT3 on sunitinib-induced left ventricular dysfunction;2.To investigate whether impact of SIRT3 on sunitinib-induced left ventricular dysfunction is mediated by autophagy;3.To investigate whether SIRT3 regulation of autophagy is mediated by GSTP1/JNK pathway.Materials and methods1.AnimalsAll procedures performed involving animals complied with the Animal Management Rules of the Chinese Ministry of Health（Document No.55,2001）and the ethical standards of the Animal Care and Use Committee of Shandong University.Eight-week-old male 129 wild type（WT）mice were purchased from Charles River Laboratories,the United States.Eight-week-old male and female SIRT3-knockout（KO）mice were purchased from Jackson Laboratory,the United States,with genetic background 129S6/SvEvTac.Mice were bred in-house after purchase.SIRT3-overexpressing mice were constructed via caudal vein injection of SIRT3-overexpressing lentivirus（LV）,and mice injected with lentiviral vehicle served as controls.2.Establishment of sunitinib-induced left ventricular dysfunction mouse modelA total of 64 mice were weighed,and treated with sunitinib of 40 mg/kg/d or placebo（olive oil）of corresponding volume via oral gavage for 4 weeks.Mice were fed with laboratory standard chow and water during the experiment.3.Blood pressure measurementsSystolic blood pressure（systolic blood pressure,SBP）,mean blood pressure（mean blood pressure,MBP）and diastolic blood pressure（diastolic blood pressure,DBP）of mice were measured using Softron non-invasive tail-cuff system.Measurements were repeated 3 times for each mouse and mean value of each parameter was used.4.EchocardiographyThe VisualSonics Vevo 770 echocardiography machine was used to obtain M-mode,two-dimensional mode,and pulse wave Doppler（PWD）mode imaging at least 3 times per mouse,and LVEF,fractional shortening（FS）and E/A ratio were calculated.5.Immunohistochemical stainingMice were euthanized,myocardial tissue slices were produced,and hematoxylin and eosin（H&E）staining,Masson trichrome staining,wheat germ agglutinin（WGA）fluorescent staining,isolectin I-B4（IB4）immunofluorescent staining,neural/glial antigen 2（NG2）and IB4 double-labeling immunofluorescent staining,and SIRT3 and NG2 double-labeling immunofluorescent staining were performed to assess histological changes,collagen storage,cardiomyocyte hypertrophy,microvessel density,microvascular pericyte coverage and pericyte SIRT3 expression level,respectively.Zeiss confocal laser scanning microscopy was employed to obtain images,and the Image Pro Plus 6.0 software was employed for analysis.6.Extraction and culture of primary cardiac pericytesMice 3 weeks old were euthanized,the myocardium were obtained in sterile condition,minced in Hank’s Balanced Salt Solution（HBSS）,digested in a protease solution,filtrated using a copper net into singular cells,and pericyte constituent was separated using the Percoll density gradient centrifugation.The pericytes were cultured in endothelial cell medium for the first 3 passages,and the medium was switched to pericyte medium upon the third passage.Immunofluorescent staining of NG2 and PDGFRB was used for identification of pericytes.pericytes were extracted using SIRT3-KO mice.LV-SIRT3 pericytes were established by transfection of LV-SIRT3 into WT pericytes.GSTP1-silenced pericytes were established via transfection of siRNA-GSTP1 into WT pericytes.7.Extraction and culture of primary cardiomyocytesNeonatal rats were euthanized,the myocardium were obtained in sterile condition,minced,digested with collagenase B repeatedly,and cardiomyocytes were collected using the differential attachment method,and cultured in the Dulbecco’s Modified Eagle Medium（DMEM）.Immunofluorescent staining of.cardiac troponin I（cTn1）was used for identification of cardiomyocytes.8.Cell viability assayCell Counting Kit-8（CCK-8）was used to determine cell viability in vitro.9.Monitor of autophagyCells of robust growth were transfected with mCherry-green fluorescent protein（GFP）-microtubule-associated protein 1A/1B-light chain 3（LC3）adenovirus and observed under confocal microscopy.Number of autophagosomes of at least 5 cells from each group were analyzed10.ImmunoblottingTotal protein were extracted from cardiac tissue and cells,separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis（SDS-PAGE）,transferred onto polyvinylidene fluoride（PVDF）membranes,blocked,incubated in primary antibodies at 4℃ overnight,then in secondary antibodies at room temperature for 2 h,and protein bands were visualized through electrochemiluminescence（ECL）,images were obtained using LAS-4000 chemiluminescence reader,and analyzed using ImageJ 1.46r software.11.Co-immunoprecipitationTotal proteins were incubated with primary antibodies and protein A-Sepharose at 4℃ overnight,beads were washed with wash buffer for 5 times,and boiled and eluted for 10 min.Precipitated samples were SDS-PAGE electrophoresed,and immunoblotted using primary antibodies against the co-immunoprecipitated protein to assess level of co-immunoprecipitation.12.Statistical analysisAll data are obtained from at least 3 independently-repeated experiments,processed using the SPSS 20.0 software,and expressed as the mean ± standard deviation（SD）.The Student’s two-tailed t-test was used to evaluate the significance of the differences between 2 groups.Differences with p<0.05 were considered statistically significant.Results1.Establishment of sunitinib-induced left ventricular dysfunction mouse modelsWT mice were administrated with sunitinib（40 mg/kg/d）or vehicle for 28 days.Blood pressure measurement results show that sunitinib-treated mice had SBP of 136.8±7.7 mmHg,while vehicle-treated mice had SBP of 111.0±5.2 mmHg.Echocardiography showed that sunitinib-treated mice had LVEF of 0.75±0.13 and FS of 0.48±0.03,while vehicle-treated mice had LVEF of 0.93±0.05 and FS of 0.67±0.11.These results showed that compared with vehicle-treated mice,sunitinib-treat mice exhibited significantly elevated blood pressures and impaired left ventricular systolic functions,indicating that sunitinib-induced left ventricular dysfunction mouse models were correctly established.2.Effect of sunitinib on mouse cardiac pericytesH&E staining,Masson trichrome staining,WGA staining and IB4 immunofluorescence of the cardiac sections of sunitinib-treated and vehicle-treated mice showed that,compared with vehicle-treated mice,sunitinib treatment did not induce inflammatory cell infiltration,fibrosis,cardiomyocyte hypertrophy,or microvascular density decrease,suggesting that sunitinib-induced left ventricular dysfunction is not mediated by these pathological processes.NG2/IB4 double-labeling immunofluorescent staining on cardiac sections and immunoblotting analysis of NG2 of cardiac protein of sunitinib-treated and vehicle-treated mice showed that compared with vehicle-treated mice,sunitinib-treated mice showed significantly lower pericyte coverage in the cardiac microvessels.We further treated cultured primary mouse cardiac pericytes with a concentration ladder of sunitinib（0,2,4,6,8,10 μM）,and CCK-8 assay showed that sunitinib dose-dependently downregulated viability of pericytes.These results are consistent with previous studies showing that cardiac pericyte loss is the main pathological alteration of sunitinib-induced left ventricular dysfunction.3.Effect of sunitinib on mouse cardiac pericyte SIRT3 levelsImmunoblotting analysis of cellular protein of sunitinib-or vehicle-treated primary mouse pericytes and primary rat cardiomyocytes showed that sunitinib dose-dependently upregulated SIRT3 levels in pericytes,but did not affect SIRT3 levels in cardiomyocytes.SIRT3/NG2 double-labeling immunofluorescent staining on mouse cardiac sections showed that,compared with vehicle-treated mice,SIRT3 levels in cardiac pericytes were upregulated in sunitinib-treated mice.These results indicate that sunitinib upregulates cardiac pericyte SIRT3 levels in vitro and in vivo,which suggests that cardiac pericyte SIRT3 may play a critical role in sunitinib-induced left ventricular dysfunction.4.Effect of SIRT3 on mouse sensitivity to sunitinib-induced cardiotoxicityWe treated WT,SIRT3-KO,LV-vehicle and LV-SIRT3 mice with sunitinib or vehicle.Results showed that,compared with WT mice,the blood pressure rise and left ventricular dysfunction were significantly suppressed in SIRT3-KO mice,while vehicle-treated WT and SIRT3-KO mice did not exhibit significant differences in cardiac functions;compared with LV-vehicle mice,the blood pressure increase and left ventricular dysfunction were worsened in LV-SIRT3 mice,while vehicle treatment did not induce differences in cardiac functions in LV-vehicle and LV-SIRT3 mice.These results validate that SIRT3 boosted sensitivity of mice to sunitinib-induced cardiotoxicity.5.Effect of SIRT3 on pericyte sensitivity to sunitinib-induced injuryWe performed NG2/IB4 double-labeling immunofluorescent staining on mouse cardiac sections,as well as immunoblotting analysis of cardiac proteins of sunitinib or vehicle-treated WT,SIRT3-KO,LV-vehicle and LV-SIRT3 mice.It was shown that compared with WT mice,cardiac pericyte loss in SIRT3-KO mice was relieved,while compared with LV-vehicle mice,cardiac pericyte loss in LV-SIRT3 mice was aggravated.These results demonstrate that SIRT3 sensitized pericytes to sunitinib-induced damage in vivo.We then performed CCK-8 assays on sunitinib or vehicle-treated WT,SIRT3-KO,LV-vehicle and LV-SIRT3 cultured pericytes.We found that SIRT3-KO pericytes suffered lighter sunitinib-induced viability decline compared with WT pericytes,and LV-SIRT3 pericytes suffered graver sunitinib-induced viability decline compared with LV-vehicle pericytes.These results demonstrate that SIRT3 sensitized pericytes to sunitinib-induced damage in vitro.Taken together,it can be inferred that SIRT3 sensitizes pericytes to sunitinib-induced injury both in vivo and in vitro.6.SIRT3 contributes to sunitinib-induced autophagic inhibition in pericytesWe performed immunoblotting analysis of apoptotic markers in sunitinib-or vehicle-treated cardiac pericytes,and results suggested that sunitinib did not alter apoptotic markers,indicating that sunitinib-induced pericyte injury was not mediated by apoptosis.We next assessed autophagic flux in sunitinib-and vehicle-treated pericytes using immunoblotting and mCherry-GFP-LC3 adenovirus transfection,and found that,compared with vehicle-treated pericytes,sunitinib-treated pericytes exhibited significant increases of cellular LC3-I-to-LC3-II conversion and dramatic cytoplasmic autophagic vesicle accumulation,indicating that sunitinib inhibited late-stage autophagy in cardiac pericytes.Next we pretreated pericytes with late-stage autophagy inhibitor bafilomycin A1 or autophagic activator rapamycin,and found that compared with control pericytes,sunitinib-induced pericyte viability decline was alleviated by rapamycin,and worsened by bafilomycin A1,which indicates that sunitinib-induced pericyte injury was at least partially mediated by autophagic inhibition.We next asked the role of SIRT3 in cardiac pericyte autophagy.Immunoblotting and adenovirus transfection showed that SIRT3-KO suppressed sunitinib-induced increase in LC3-I-to-LC3-II conversion and autophagic vesicle accumulation,while LV-SIRT3 aggravated these effects.These results indicate that SIRT3 augmented sunitinib-induced autophagic inhibition in cardiac pericytes.7.SIRT3 sensitizes pericytes to sunitinib-induced autophagic inhibition viaGSTP1/JNK pathwayWe first examined protein levels and phosphorylations of the 5’ adenosine monophosphate-activated protein kinase（AMPK）/mechanistic target of rapamycin（mTOR）pathway,and found that there was no significant differences of protein levels and phosphorylations of this pathway in sunitinib-or vehicle-treated pericytes,indicating that this pathway was not involved in sunitinib-induced pericyte injury.Next,we performed immunoblotting and siRNA transfection experiments on sunitinib-or vehicle-treated pericytes to assess protein levels and phosphorylations of the GSTP1/JNK pathway,as well as performed co-immunoprecipitations to assess direct interactions between SIRT3 and GSTP1,finding that cardiac pericyte SIRT3 directly binds with GSTP1,and that SIRT3 downregulates protein levels of GSTP1,promoting phosphorylation and activation of downstream protein JNK,inhibiting autophagy and therefore sensitizing pericytes to the injury of sunitinib.Conclusions1.SIRT3 promotes sensitivity of mice to sunitinib-induced left ventricular dysfunction;2.SIRT3 promotes sensitivity of mice to sunitinib injury via inhibiting cardiac pericyte autophagy;3.SIRT3 inhibits cardiac pericyte autophagy via inhibition of GSTP1/JNK pathway.Background Pericytes are a group of regulatory and multipotent periendothelial cells wrapped around the outer basal membrane of microvessels throughout the microvascular system.First discovered by Rouget in 1873,pericytes were for a long time assumed to be merely supportive and physiologically unimportant cells involved in vasoconstriction.However,in the past 50 years,basic and clinical studies regarding physiological and pathological functions of pericytes from various systems and organs strikingly expanded our knowledge of pericytes.Emerging evidence in various fields has shown that pericytes are essential regulators of physiological processes such as angiogenesis,vascular permeability and substance exchange,blood flow,extracellular matrix organization and signaling and the blood-brain barrier,and has demonstrated robust healing potentials in reparative fibrosis,scar formation,growth and metastasis of tumors and arterial collateralization.Due to the unique disposition of pericytes as interface between the microvasculature and the tissue compartment,pericytes are a first-hand player in the vascular/tissue substance exchange.This makes structural and functional alterations of pericytes particularly influential to highly-metabolic tissues that heavily rely on microcirculatory blood supply and substance exchange,such as tumors and the heart.Studies regarding drug metabolism in tissues have shown that pericytes are direct participants in modulation of tissue sensitivity and resistance towards vascularly-delivered anti-cancer chemotherapeutic drugs: pericytes in thyroid cancer act as a shield and are the main cell group to produce resistance against anticancer chemotherapeutic agents sorafenib and vemurafenib;cardiac pericytes,on the other hand,have been identified as the cellular target of tyrosine kinase inhibitor sunitinibinduced cardiotoxicity.Nevertheless,more aspects of the roles of cardiac pericytes in cardiac drug metabolism and underlying mechanisms remain unknown.Further investigations of effects of cardiac pericytes in cardiac drug sensitivity and resistance as well as cellular molecular mechanisms would contribute to developing safer and more potent cardiovascular pharmaceutics.Sirtuin 3（SIRT3）is a mitochondrion-located nicotinamide adenine dinudeotide（NAD）-dependent deacetylase.Recently,SIRT3 has been reported to regulate sensitivity and resistance of different types of cells towards anti-cancer chemotherapeutic drugs.For instance,SIRT3 augmented breast cancer cell resistance towards tamoxifen and cisplatin,but enhanced sensitivity of hepatoma cells towards sorafenib,doxorubicin and epirubicin.In our previous studies,it was demonstrated that mouse cardiac pericyte SIRT3 promoted pericyte sensitivity to sunitinib-induced cardiotoxicity via inhibition of glutathione S-transferase pi 1（GSTP1）.Apart from this,no other study has investigated effects of SIRT3 in cardiac pericyte drug metabolism.A more profound and comprehensive understanding of roles of Sirt3 in cardiac pericyte drug metabolism remains to be established.In this study,we performed tandem mass tag（TMT）-labeling quantitative proteomics and bioinformatic analyses on in vitro cultured primary mouse cardiac pericytes,and reveal that mouse cardiac pericyte SIRT3 deletion predominantly alters the G^P450-associated drug metabolism pathway.We further validate through in vitro experiments that SIRT3 deletion effectively suppresses pericyte sensitivity towards toxicity of anti-cancer chemotherapeutic drugs sorafenib and vemurafenib,and regulation of sensitivity against sorafenib is mediated through upregulation of glutathione S-transferase alpha 4（GSTA4）.These findings broaden the understanding of the impact of pericytes on drug metabolism,proposing SIRT3 as key regulator of cardiac pericyte CYP450-associated drug metabolism pathway,which may provide cellular and molecular avenue for establishing more efficient cardiovascular systemassociated drug delivery and toxicity protection. Aims1.To investigate influence of cardiac pericyte SIRT3 on CYP450-associated drug metabolsim proteome;2.To investigate cardiac pericyte SIRT3 regulation of GSTA4 protein levels;3.To investigate cardiac pericyte SIRT3 modulation of sensitivity to anti-cancer drugs through regulation of GSTA4.Materials and methods1.Extraction and culture of primary mouse cardiac pericytes All animal experiments involved in this study complied with the Animal Management Rules of the Chinese Ministry of Health（Document No.55,2001）and the ethical standards of the Animal Care and Use Committee of Shandong University.Eight-week-old male 129 wild type（WT）mice were purchased from Charles River Laboratories,the United States.Eight-week-old male and female SIRT3-knockout（KO）mice were purchased from Jackson Laboratory,the United States.After purchase,mice were bred in-house.For extraction of primary mouse cardiac pericytes,40 3-week-old male mice（20 WT mice and 20 Sirt3-KO mice）were used.Ten mice were used for extraction of each pooled sample,and 2 pooled samples were averaged in each group in the proteomic analysis.Three-week-old mice were euthanized,the myocardium excised in sterile condition,mechanically minced in Hank’s Balanced Salt Solution(HBSS^ digested in a protease cocktail,dispersed into single cells by repeated filtration through a copper net,and the pericyte constituent was isolated using the Percoll density gradient centrifugation.Pericytes were resuspended in endothelial cell medium,cultured,passaged,and then upon the third passage,the medium was changed to pericyte medium,immunofluorescent staining of neural/glial antigen 2（NG2）and platelet derived growth factor receptors 3（PDGFRp）was used for confirmation of pericyte identity.2.Protein sample preparation Pericytes converging at optimal growth condition were used,the medium was discarded,and cells were rinsed with phosphate buffer saline（PBS）.Cells were then gently detached in PBS,collected into centrifugation tubes and centrifuged at 12000 rpm for 5 min,removing supernatants.Samples were lysed with SDT lysis buffer at100°C for 15 min,and then centrifuged at 14000 g for 15 min.Supernatants were quantified for protein concentration using the bicinchoninic acid（BCA）assay,assessed using sodium dodecyl sulfate-polyacrylamide gel electrophoresis（SDS-PAGE）according to protein concentration,and gels were stained with Coomassie brilliant blue staining.3.Filter aided sample preparation（FASP）and TMTIOplex? labeling FASP was performed on samples.Dithiothreotol（DTT）was used to reduced disulfide bonds,iodoacetamide was used to oxidize disulfide bonds and cysteine residues.Samples were digested using 4 g/L trypsin in ammonium bicarbonate,centrifuged,filtrates were collected,and peptides were desalinated using C18 Cartridge.Peptides were lyophilized,redissolved with formic acid,and peptide contents were estimated by ultraviolet spectrophotometry.Subsequently,samples were labeled with TMTIOplex? Isobaric Label Reagent Set.4.Liquid chromatography-mass spectrometry（LC-MS/MS）analysis and data processing Samples were separated using the Easy nLC system.Q Exactive Plus mass spectrometer was used to perform LC-MS/MS analysis.The generated Raw files were converted to.mgf files using Proteome Discoverer 2.1.Files were submitted to MASCOT2.6 engine and searched in the Uniprot_MusMusculus₁6998一20180905database,and.dat files were obtained.5.Bioinformatic analysis Bioinformatic analyses were performed on differentially-expressed proteins:Gene Ontology（GO）functional annotations were analyzed using Blast2GO;The Kyoto Encyclopedia of Genes and Genomes（KEGG）KEGG Orthology And Links Annotation（KOALA）database was used for pathway annotation analysis;the Python library matplotlib was used to perform protein clustering analysis and generate hierarchical clustering thermal maps;Ingenuity Pathway Analysis（IPA?）was performed on differential proteins in significant pathways,and functional interaction networks were generated.6.Immunoblotting Cardiac pericyte protein were extracted,separated using SDS-PAGE,transferred onto polyvinylidene fluoride（PVDF）membranes,blocked,incubated in primary antibodies at 4°C overnight,in secondary antibodies at room temperature for 2 h,and protein bands were visualized using electrochemiluminescence（ECL）,images were captured with LAS-4000 chemiluminescence reader,and analyzed using ImageJ1.46 r software.7-Cell treatment and viability assay Cultured cardiac pericytes were treated with sorafenib（0,2,4,6,8,10 \xM）for48 h or vemurafenib(0,100,200,300,400,500 nMol} for 72 h.Cell Counting Kit-8（CCK-8）was used to determine cell viability in vitro.8.siRNA transfection Cells were transfected with 50 nMol of siRNA silencing GSTA4（5;-3’）GCUGCCAAGUACAACUUGUTTACAAGUUGUACUUGGCAGCTT.9.Statistical analysis All data are generated from at least 3 independently-repeated experiments,analyzed using the SPSS 20.0 software,and represented as the mean ± standard deviation（SD）.The Student’s two-tailed t-test was used to determine significance of the differences between 2 groups.Differences with p < 0.05 were considered statistically significant.Results1.Quantification of cardiac pericyte proteome We extracted and cultured primary mouse cardiac pericytes and confirmed their identity using immunofluorescence against pericyte markers NG2 and PDGFRp.TMT quantitative proteomic analysis results show that a total of 38383 peptides and 5110 proteins were identified from the protein samples.Proteins were matched in the Uniprot—MusMusculuS-16998₂0180905 database.2.Effect of cardiac pericyte SIRT3 ablation on CYP450-associated drug metabolism pathway Among the 5110 identified proteins,415 proteins were differentially expressed to statistical significance in between WT and SIRT3-KO samples.159 proteins were significantly upregulated in SIRT3-KO pericytes compared with WT pericytes,and 256 proteins significantly downregulated.For the differential proteins,277 protein pathway maps were identified in KE6 G pathway annotation analysis;differential protein KEGG pathway enrichment analysis further showed that CYP450-associated drug and xenobiotic metabolism pathway is the second most significantly altered pathways,involving 16 differentially expressed proteins.On the other hand,GO functional annotation analysis recognized total of 3711 GO annotations.On level one annotations,2541（68.47%）annotations of biological processes,683（18.40%）of molecular functions,and 487（13.12%）of cellular compartments were identified,indicating that biological processes was the most significantly altered functional annotation in Sirt3-KO pericytes compared with WT pericytes.Within the biological process annotation category,level 2 subcategories cellular process（15.96%）and metabolic processes（15.96%）were identified with the most protein annotation hits.GO enrichment analysis of differential proteins revealed that xenobiotic metabolic process of biological processes was significantly altered with 9 differential proteins.Collectively,GO functional annotation analysis results support that CYP450-associated drug metabolism pathway was significantly affected in SIRT3-KO pericytes compared with WT pericytes.3.Effect of cardiac pericyte SIRT3 ablation on differential proteins in CYP450-associated drug metabolism pathway KEGG pathway annotation analysis revealed that among the 16 differential proteins of CYP450-associated drug metabolism pathway,10 were upregulated in SIRT3-KO pericytes compared with WT pericytes,including CYP450 1B1（CYP1B1）,3glutathione S-transferases（GSTs）6STA4,glutathione S-transferase alpha 3（GSTA3）and glutathione S-transferase mu 1（GSTM1）,and 2 uridine diphosphate glucuronosyltransferases（UDPGTs）（UDPGT1-7C and UDPGT1-6）,while 6 were downregulated,including 3 Flavin containing monooxygenases（FMOs）（FMOl,FM02,FM03）.IPA? reveals that these differential proteins are connected in a functional interaction network by molecular functions or interactions.4.SIRT3 ablation desensitizes pericytes to anti-cancer drug-induced injury To corroborate the bioinformatic finding that Sirt3 was involved with CYP450-associated drug metabolism,we treated cultured cardiac pericytes with anti-cancer drugs sorafenib（0,2,4,6,8,10 pM）for 48 h or vemurafenib（0,100,200,300,400,500 nMol）for 72 h.CCK-9 results show that both sorafenib and vemurafenib induced WT cardiac pericyte viability declines in a concentration-dependent manner;however,compared with WT pericytes,the sorafenib-and vemurafenib-induced viability declines in Sirt3-KO pericytes were significantly blunted.These results confirm that cardiac pericyte Sirt3 ablation alleviates toxicities induced by anticancer chemotherapeutic agents sorafenib and vemurafenib.5.SIRT3 ablation in pericytes upregulates GSTA4 Among the 3 upregulated GSTs,GSTA4 was upregulated to the highest statistical significance.We further validated via immunoblotting that Sirt3 deletion induced significant upregulation of GSTA4 protein expression.6.SIRT3 ablation desensitizes pericytes to sorafenib-induced injury via upregulation ofGSTA4 We silenced GSTA4 expression in SIRT3-K0 pericytes using siRNA transfection,and then treated WT pericytes,SIRT3-K0 pericytes and SIRT3-KO pericytes transfected with siRNA-GSTA4 with sorafenib or vemurafenib.Cell viability assays show that sorafenib and vemurafenib induced cell viability declines in WT pericytes;compared with WT pericytes,SIRT3-KO pericytes exhibited significantly mitigated sorafenib* and vemurafenib-induced cell viability declines;compared with SIRT3-KO pericytes,GSTA4-silenced SIRT3-K0 pericytes showed significantly aggravated sorafenib-induced viability decline,but the vemurafenib-induced viability decline was not altered.These results demonstrate that SIRT3 ablation alleviates sorafenibinduced toxicity through GSTA4 upregulation,but effect of SIRT3 ablation on vemurafenib-induced toxicity is not mediated by GSTA4.Conclusions1.Mouse cardiac pericyte SIRT3 ablation significantly alters CYP450-associated drug metabolsim proteome;2.Mouse cardiac pericyte SIRT3 ablation suppresses cellular sensitivity towards sorafenib-and vemurafenib-induced injuries;3.Mouse cardiac pericyte SIRT3 ablation upregulates GSTA4 levels;4.Mouse cardiac periycte SIRT3 ablation suppresses cellular sensitivity towards sorafenib-induced injury via upregulation of GSTA4,but effect on vemurafenibinduced injury is not mediated by GSTA4.