# paper.pdf ## Metadata - PDFFormatVersion=1.7 - Language=de-DE - IsLinearized=true - IsAcroFormPresent=false - IsXFAPresent=false - IsCollectionPresent=false - IsSignaturesPresent=false - Author=Kai Huang - CreationDate=D:20260513155741+05'30' - Creator=Adobe InDesign 15.1 (Windows) - ModDate=D:20260514172010+05'30' - Producer=Adobe PDF Library 15.0 - Subject=DOI: 10.1161/CIRCRESAHA.125.326768 Circulation Research, Volume. 138, Issue. 11, BACKGROUND: Presently, we investigated hypothesized roles and mechanisms of cell type–specific, selective activation of different vascular NOX (NADPH oxidase) isoforms in obesity and metabolic syndrome. Methods: Wild-type, NOX1/2/4 global knockout mice, endothelial/VSMC-specific NOX1 knockout mice, or vascular endothelial-speicifc NOX1 knockin mice were exposed to high-fat diet feeding prior to phenotypical analyses of obesity and metabolic syndrome, as well as of exercise capacity, skeletal muscle mitochondrial function, and novel genetic signatures. Results: Expression of NOX1 was significantly upregulated in wild-type mice fed a high-fat diet. Global knockout of NOX1 (NOX1−/y), rather than of NOX2/NOX4, markedly abrogated high-fat feeding–induced body weight/fat mass gain, preadipocyte differentiation, fatty liver, glucose intolerance, and insulin/leptin resistance. Intriguingly, endothelial-specific NOX1 knockout (Cdh5cre [cadherin 5 (vascular endothelial cadherin) promoter–driven Cre recombinase (endothelial-specific Cre expression)]-NOX1CKO), rather than vascular smooth muscle-specific NOX1 knockout (Myh11cre [myosin heavy chain 11 promoter–driven Cre recombinase (smooth muscle–specific Cre expression)]-NOX1CKO), substantially alleviated obesity and metabolic syndrome. Consistently, endothelial-specific NOX1 knockin mice (Cdh5cre-NOX1CKI) fed a high-fat diet displayed exaggerated metabolic disorders. Endothelial cell–specific knockout/knockin of NOX1 was confirmed using endothelial cell washout experiments. Food/water intakes were not different from corresponding controls in high-fat fed NOX1−/y, Cdh5cre-NOX1CKO, or Cdh5cre-NOX1CKI mice, indicating no difference in energy intake. Instead, spontaneous activity, exercise capacity, mitochondrial oxygen consumption/ATP production, skeletal muscle mitochondrial function (ROS production and swelling activity), and mitochondrial cristae structure were all substantially improved in NOX1−/y or Cdh5cre-NOX1CKO mice, indicating augmented energy expenditure attributed to preserved skeletal muscle mitochondrial function. Supportively, Cdh5cre-NOX1CKI mice displayed deteriorated exercise capacity and skeletal muscle mitochondrial dysfunction. Endothelium-dependent vasorelaxation was restored in high-fat fed NOX1−/y or Cdh5cre-NOX1CKO mice, confirming improved endothelial function. RNA-sequencing identified 4 genes (Cntnap4 [contactin-associated protein-like 4], Sgsm1, Tll2, and Syt9) and 7 genes (Odf3l2, Col9a1 [collagen type IX alpha 1 chain], Cldn23, Atp5g2, Nkx6-3, Ntsr2, and Zfp69) significantly downregulated/upregulated in high-fat fed Cdh5cre-NOX1CKO mice, among which Cntnap4 and Col9a1 linked to muscular disorders. Importantly, we observed marked upregulation of NOX1 in isolated coronary arteries from human patients with obesity. CONCLUSIONS: Taken together, our data for the first time establish a novel and paradigm-shifting concept that endothelial NOX1 drives systematic metabolic phenotypes, via impairment in skeletal muscle mitochondrial dysfunction with novel genetic signatures. Tissue-specific targeting of endothelial NOX1 and novel candidate genes may prove to be robustly effective in treating obesity and metabolic syndrome. - Title=Endothelial NOX1 Drives Obesity via Skeletal Muscle Mitochondrial Dysfunction - Trapped=False - dc:format=application/pdf - dc:title=Endothelial NOX1 Drives Obesity via Skeletal Muscle Mitochondrial Dysfunction - dc:creator=Kai Huang,Yuanli Huang,Yuhan Zhang,Yixuan Zhang,Nicholas W. Hatch,Julie K. Freed,Hua Cai - dc:subject=metabolic syndrome ◼,mitochondrial diseases ◼,muscle skeletal ◼,NADPH oxidases ◼,obesity ◼,reactive oxygen species - dc:description=DOI: 10.1161/CIRCRESAHA.125.326768 Circulation Research, Volume. 138, Issue. 11, BACKGROUND: Presently, we investigated hypothesized roles and mechanisms of cell type–specific, selective activation of different vascular NOX (NADPH oxidase) isoforms in obesity and metabolic syndrome. Methods: Wild-type, NOX1/2/4 global knockout mice, endothelial/VSMC-specific NOX1 knockout mice, or vascular endothelial-speicifc NOX1 knockin mice were exposed to high-fat diet feeding prior to phenotypical analyses of obesity and metabolic syndrome, as well as of exercise capacity, skeletal muscle mitochondrial function, and novel genetic signatures. Results: Expression of NOX1 was significantly upregulated in wild-type mice fed a high-fat diet. Global knockout of NOX1 (NOX1−/y), rather than of NOX2/NOX4, markedly abrogated high-fat feeding–induced body weight/fat mass gain, preadipocyte differentiation, fatty liver, glucose intolerance, and insulin/leptin resistance. Intriguingly, endothelial-specific NOX1 knockout (Cdh5cre [cadherin 5 (vascular endothelial cadherin) promoter–driven Cre recombinase (endothelial-specific Cre expression)]-NOX1CKO), rather than vascular smooth muscle-specific NOX1 knockout (Myh11cre [myosin heavy chain 11 promoter–driven Cre recombinase (smooth muscle–specific Cre expression)]-NOX1CKO), substantially alleviated obesity and metabolic syndrome. Consistently, endothelial-specific NOX1 knockin mice (Cdh5cre-NOX1CKI) fed a high-fat diet displayed exaggerated metabolic disorders. Endothelial cell–specific knockout/knockin of NOX1 was confirmed using endothelial cell washout experiments. Food/water intakes were not different from corresponding controls in high-fat fed NOX1−/y, Cdh5cre-NOX1CKO, or Cdh5cre-NOX1CKI mice, indicating no difference in energy intake. Instead, spontaneous activity, exercise capacity, mitochondrial oxygen consumption/ATP production, skeletal muscle mitochondrial function (ROS production and swelling activity), and mitochondrial cristae structure were all substantially improved in NOX1−/y or Cdh5cre-NOX1CKO mice, indicating augmented energy expenditure attributed to preserved skeletal muscle mitochondrial function. Supportively, Cdh5cre-NOX1CKI mice displayed deteriorated exercise capacity and skeletal muscle mitochondrial dysfunction. Endothelium-dependent vasorelaxation was restored in high-fat fed NOX1−/y or Cdh5cre-NOX1CKO mice, confirming improved endothelial function. RNA-sequencing identified 4 genes (Cntnap4 [contactin-associated protein-like 4], Sgsm1, Tll2, and Syt9) and 7 genes (Odf3l2, Col9a1 [collagen type IX alpha 1 chain], Cldn23, Atp5g2, Nkx6-3, Ntsr2, and Zfp69) significantly downregulated/upregulated in high-fat fed Cdh5cre-NOX1CKO mice, among which Cntnap4 and Col9a1 linked to muscular disorders. Importantly, we observed marked upregulation of NOX1 in isolated coronary arteries from human patients with obesity. CONCLUSIONS: Taken together, our data for the first time establish a novel and paradigm-shifting concept that endothelial NOX1 drives systematic metabolic phenotypes, via impairment in skeletal muscle mitochondrial dysfunction with novel genetic signatures. Tissue-specific targeting of endothelial NOX1 and novel candidate genes may prove to be robustly effective in treating obesity and metabolic syndrome. - dc:rights=© 2026 The Authors. - xmp:createdate=2026-05-13T15:57:41+05:30 - xmp:creatortool=Adobe InDesign 15.1 (Windows) - xmp:modifydate=2026-05-14T17:20:10+05:30 - xmp:metadatadate=2026-05-14T17:20:10+05:30 - xmpmm:instanceid=uuid:32a80ca7-e496-4513-ac03-d3e4e9c867ba - xmpmm:originaldocumentid=xmp.did:DE4D607E8A24681180838231FB1F3433 - xmpmm:documentid=xmp.id:2531485f-b85b-9f4f-8a37-eceded2394f5 - xmpmm:renditionclass=proof:pdf - xmpmm:versionid=1 - xmpmm:derivedfrom=xmp.iid:d15dbe4b-14c0-5a47-b995-414d8d9be121xmp.did:134860d9-7a4e-044e-92db-6bf078d5762edefault - xmpmm:history=convertedfrom application/x-indesign to application/pdfAdobe InDesign 15.1 (Windows)2026-05-13T15:57:41+05:30converteduuid:ce785028-76bf-4c41-8e60-d288f0f85d15converted to PDF/A-3aPreflight2026-05-14T17:20:09+05:30 - xmprights:marked=True - pdf:producer=Adobe PDF Library 15.0 - pdf:trapped=False - pdfuaid:part=1 - pdfaid:part=3 - pdfaid:conformance=A - pdfaextension:schemas=http://ns.adobe.com/pdf/1.3/pdfAdobe PDF SchemainternalA name object indicating whether the document has been modified to include trapping informationTrappedTexthttp://ns.adobe.com/xap/1.0/mm/xmpMMXMP Media Management SchemainternalUUID based identifier for specific incarnation of a documentInstanceIDURIinternalThe common identifier for all versions and renditions of a document.OriginalDocumentIDURIhttp://ns.adobe.com/pdfx/1.3/pdfxAdobe Document Info PDF eXtension SchemainternalID of PDF/X standardGTS_PDFXVersionTextinternalConformance level of PDF/X standardGTS_PDFXConformanceTextinternalCompany creating the PDFCompanyTextinternalDate when document was last modifiedSourceModifiedTexthttp://www.aiim.org/pdfa/ns/id/pdfaidPDF/A ID SchemainternalPart of PDF/A standardpartIntegerinternalAmendment of PDF/A standardamdTextinternalConformance level of PDF/A standardconformanceTexthttp://www.aiim.org/pdfua/ns/id/pdfuaidPDF/UA ID SchemainternalPart of PDF/UA standardpartInteger ## Contents ### Page 1 ORIGINAL RESEARCH # Endothelial NOX1 Drives Obesity via SkeletalMuscle Mitochondrial Dysfunction Kai Huang,* Yuanli Huang,* Yuhan Zhang[ ![Equation 1]() ](https://orcid.org/0009-0004-3328-2400) , Yixuan Zhang, Nicholas W. Hatch, Julie K. Freed[ ![Equation 1]() ](https://orcid.org/0000-0002-9367-2281) , Hua Cai[ ![Equation 1]() ](https://orcid.org/0000-0001-8747-6156) Correspondence to: Hua Cai, MD, PhD, Department of Anesthesiology and Perioperative Medicine, Department of Medicine, David Geffen School of Medicine,University of California, 650 Charles E Young Dr, Los Angeles, Los Angeles, CA 90095. Email *K. Huang and Y. Huang contributed equally. Supplemental Material is available at. © 2026 The Authors. Circulation Research is published on behalf of the American Heart Association, Inc., by Wolters Kluwer Health, Inc. This is an open access articleunder the terms of the License, which permits use, distribution, and reproduction in any medium, providedthat the original work is properly cited, the use is noncommercial, and no modifications or adaptations are made. BACKGROUND: Presently, we investigated hypothesized roles and mechanisms of cell type–specific, selective activation ofdifferent vascular NOX (NADPH oxidase) isoforms in obesity and metabolic syndrome. METHODS: Wild-type, NOX1/2/4 global knockout mice, endothelial/VSMC-specific NOX1 knockout mice, or vascularendothelial-speicifc NOX1 knockin mice were exposed to high-fat diet feeding prior to phenotypical analyses of obesity andmetabolic syndrome, as well as of exercise capacity, skeletal muscle mitochondrial function, and novel genetic signatures. RESULTS: Expression of NOX1 was significantly upregulated in wild-type mice fed a high-fat diet. Global knockout ofNOX1 (NOX1−/y), rather than of NOX2/NOX4, markedly abrogated high-fat feeding–induced body weight/fat mass gain,preadipocyte differentiation, fatty liver, glucose intolerance, and insulin/leptin resistance. Intriguingly, endothelial-specificNOX1 knockout (Cdh5cre [cadherin 5 (vascular endothelial cadherin) promoter–driven Cre recombinase (endothelial-specific Cre expression)]-NOX1CKO), rather than vascular smooth muscle-specific NOX1 knockout (Myh11cre [myosinheavy chain 11 promoter–driven Cre recombinase (smooth muscle–specific Cre expression)]-NOX1CKO), substantiallyalleviated obesity and metabolic syndrome. Consistently, endothelial-specific NOX1 knockin mice (Cdh5cre-NOX1CKI)fed a high-fat diet displayed exaggerated metabolic disorders. Endothelial cell–specific knockout/knockin of NOX1 wasconfirmed using endothelial cell washout experiments. Food/water intakes were not different from corresponding controls inhigh-fat fed NOX1−/y, Cdh5cre-NOX1CKO, or Cdh5cre-NOX1CKI mice, indicating no difference in energy intake. Instead,spontaneous activity, exercise capacity, mitochondrial oxygen consumption/ATP production, skeletal muscle mitochondrialfunction (ROS production and swelling activity), and mitochondrial cristae structure were all substantially improved in NOX1−/yor Cdh5cre-NOX1CKO mice, indicating augmented energy expenditure attributed to preserved skeletal muscle mitochondrialfunction. Supportively, Cdh5cre-NOX1CKI mice displayed deteriorated exercise capacity and skeletal muscle mitochondrialdysfunction. Endothelium-dependent vasorelaxation was restored in high-fat fed NOX1−/y or Cdh5cre-NOX1CKO mice,confirming improved endothelial function. RNA-sequencing identified 4 genes (Cntnap4 [contactin-associated protein-like4], Sgsm1, Tll2, and Syt9) and 7 genes (Odf3l2, Col9a1 [collagen type IX alpha 1 chain], Cldn23, Atp5g2, Nkx6-3, Ntsr2,and Zfp69) significantly downregulated/upregulated in high-fat fed Cdh5cre-NOX1CKO mice, among which Cntnap4 andCol9a1 linked to muscular disorders. Importantly, we observed marked upregulation of NOX1 in isolated coronary arteriesfrom human patients with obesity. CONCLUSIONS: Taken together, our data for the first time establish a novel and paradigm-shifting concept that endothelialNOX1 drives systematic metabolic phenotypes, via impairment in skeletal muscle mitochondrial dysfunction with novelgenetic signatures. Tissue-specific targeting of endothelial NOX1 and novel candidate genes may prove to be robustlyeffective in treating obesity and metabolic syndrome. GRAPHIC ABSTRACT: A is available for this article. Key Words: metabolic syndrome ◼ mitochondrial diseases ◼ muscle, skeletal ◼ NADPH oxidases ◼ obesity ◼ reactive oxygen species [Meet the First Author, see p e000756](https://circres.ahajournals.org/content/138/11/1615)|[Editorial, see Article by Craige](https://circres.ahajournals.org/content/138/11/1589) ### Page 2 ## Novelty and Significance ### What Is Known? - • Oxidative stress is involved in the development of obe-sity and metabolic syndrome. - • Activation of NOX (NADPH oxidase) has been impli-cated in obesity and metabolic syndrome. ### What New Information Does This ArticleContribute? - • NOX1 (NOX isoform 1), but not NOX2 (NOX isoform2) or NOX4 (NOX isoform 4), mediates the develop-ment of obesity and metabolic syndrome. - • Endothelial-specific activation of NOX1 drives obesityand metabolic syndrome attributed to exercise inca-pacity and reduced energy expenditure. - • Endothelial-specific NOX1 activation mediates skeletalmuscle mitochondrial dysfunction to result in deficien-cies in exercise capacity and energy expenditure. - • Novel genetic signatures of skeletal Cntnap4( contactin-associated protein-like 4) and Col9a1 (col-lagen type IX alpha 1 chain) have been identified aspotential new targets for the treatment of obesity andmetabolic syndrome. - • Our findings for the first time mechanistically demon-strate a critical driving role of endothelial NOX1 activationin the development of obesity and metabolic syndrome. Oxidative stress attributed to activation of NOXhas been implicated in obesity and metabolic syn-drome. However, this study reveals a novel role ofendothelial-specific activation of NOX1 in driving theseconditions by impairing exercise capacity to reduceenergy expenditure, while NOX2 and NOX4 are dis-pensable. Mechanistically, endothelial NOX1 activationinduces skeletal muscle mitochondrial dysfunction,leading to metabolic deficiencies. In addition, novelgenetic signatures, Cntnap4 and Col9a1, were iden-tified as potential therapeutic targets. These findingsprovide the first mechanistic evidence that endothelialNOX1 is a key driver of obesity and metabolic syn-drome, offering new insights for targeted interventions. ## Nonstandard Abbreviations and Acronyms BMI body mass index Cdh5cre cadherin 5 (vascular endothelial cad-herin) promoter–driven Cre recombinase(endothelial-specific Cre expression) Cntnap4 contactin-associated protein-like 4 Col9a1 collagen type IX alpha 1 chain DHFR dihydrofolate reductase EC endothelial cell eNOS endothelial nitric oxide synthase IPA Ingenuity Pathway Analysis Myh11cre myosin heavy chain 11 promoter–driven Cre recombinase (smoothmuscle–specific Cre expression) NOX NADPH oxidase NOX1 NOX isoform 1 NOX1CKI cre-inducible NOX1flox/flox knockin/floxed mice NOX1CKO cre-inducible NOX1flox/flox knockout/floxed mice NOX2 NOX isoform 2 NOX3 NOX isoform 3 NOX4 NOX isoform 4 PCR polymerase chain reaction RCR respiratory control ratio ROS reactive oxygen species UCLA University of California, Los Angeles WT wild type Obesity has proven to be a global epidemic with, how-ever, unclear molecular mechanisms for its develop-ment and progression.1 It is estimated that 39% to49% of the population worldwide (2.8–3.5 billion people)is overweight or obese.1,2 Obesity is a chronic, complex,heterogeneous disease that impacts all organ systems,resulting in severe health consequences and a huge eco-nomic burden for patients and society.3,4 WHO estimatesthat at least 2.8 million people die each year as a resultof being obese or overweight.5 Although pharmacologi-cal, surgical, and device interventions have been used totreat obesity, these practices have not been consistentlyeffective in managing obesity and its complications at thepopulation level.4,6,7 Currently, there are 6 major Food andDrug Administration–approved antiobesity medications:phentermine, orlistat, phentermine/topiramate extendedrelease, lorcaserin, naltrexone sustained release/bupro-pion sustained release, and liraglutide.3 Nonetheless, allof these drugs have varying degrees of adverse effectsor contraindications and are not optimal for long-termtreatment of obesity.3 Therefore, identification of detailedmolecular mechanisms of obesity is in urgent need toenable development of novel and effective therapeuticsfor the treatment of obesity and metabolic syndrome. Oxidative stress has been implicated in the devel-opment of obesity and metabolic syndrome.8,9 We havepreviously established an innovative, paradigm-shiftingconcept that vascular oxidative stress is a driving forcefor the development of systematic obesity and metabolicsyndrome, rather than being a mere consequence.8 Micewith overproduction of reactive oxygen species (ROS) inthe vasculature displayed markedly exaggerated obesity ### Page 3 and metabolic syndrome in response to high-fat feeding.8p22phox (also known as cytochrome b-245 light chain),as the only membrane-bound subunit for NOX (NADPHoxidase), is required for catalytic activation of NOX1(NOX isoform 1), NOX2 (NOX isoform 2), NOX3 (NOXisoform 3), and NOX4 (NOX isoform 4),10,11 among whichNOX1, NOX2, and NOX4 are expressed in the vascularcells.12,13 Of note, NOX family enzymes have emergedas the predominant sources of ROS production dur-ing the pathogenesis of cardiovascular diseases.11,14–18We and others have identified differential roles of dif-ferent NOX isoforms in mediating the development ofdifferent pathological conditions of hypertension, aorticaneurysms, diabetic vascular dysfunction under eithertype 1 or type 2 conditions, ischemia-reperfusion injuryof the heart, and cardiac arrhythmia.11,19–30 Therefore, wesought to identify whether there is a direct, causal roleof NOX in driving obesity and metabolic syndrome inresponse to high-fat feeding, the specific isoform(s) ofNOX that is(are) involved, potential tissue/cell-specificroles of NOX isoforms, and the underlying molecularmechanisms. In this study, our data clearly demonstrate that globalknockout of NOX1, rather than of NOX2 or NOX4,protects against obesity and metabolic syndromeunder high-fat feeding. NOX1 expression was mark-edly upregulated in wild-type (WT) mice fed a high-fatdiet. Endothelial cell (EC) NOX1, rather than vascularsmooth muscle NOX1, is specifically involved in thisprocess. While energy intake was not different, high-fat feeding–induced impairment in energy expenditure,spontaneous activity, exercise capacity, mitochondrialoxygen consumption/ATP production, skeletal mus-cle mitochondrial function, and mitochondrial cristaestructure were all substantially improved in global orendothelial-specific NOX1 knockout mice, whereasthese phenotypes and mechanistic alternations wereexaggerated in endothelial-specific NOX1 knockinmice. The endothelial-specific knockout/knockin ofNOX1 was validated using EC washout experiments.Therefore, enhanced energy expenditure attributed toimproved skeletal muscle mitochondrial function resultsin restored spontaneous activity and exercise capacity,which underlies protection against obesity and meta-bolic syndrome by global or endothelial-specific deletionof NOX1. Endothelium-dependent vasorelaxation wasrestored in high-fat–fed NOX1−/y or Cdh5cre (cadherin5 [vascular endothelial cadherin] promoter–driven Crerecombinase [endothelial-specific Cre expression])-cre-inducible NOX1flox/flox knockout (Cdh5cre-NOX1CKO),confirming improved endothelial function. Novel targetsof Cntnap4 (contactin-associated protein-like 4) andCol9a1 (collagen type IX alpha 1 chain) were identi-fied by RNA-sequencing and validation analyses inCdh5cre-NOX1CKO mice to account for the mechanis-tic regulations of skeletal muscle function. Importantly,we observed marked upregulation of NOX1 in the coro-nary arteries of human patients with obesity. For the firsttime, these findings establish a novel, significant, spe-cific, and selective role of endothelial NOX1 in drivingobesity and metabolic syndrome via impairment in skel-etal muscle mitochondrial function/energy expenditurewith novel genetic signatures, targeting of which may berobustly effective in treating human metabolic disorders. ## METHODS ### Data Availability The original data supporting the main findings of this studyare available from the corresponding author upon reasonablerequest. Except for the most critical methods paragraphs pre-sented below, other detailed methods have been included inthe. ### Generation of Endothelial Specific NOX1CKOand NOX1 CKI mice and Experimental Design The NOX1 floxed mice were generated by the TransgenicMouse Facility at the University of California, Irvine, using theEuropean Conditional Mouse Mutagenesis Program–targetedmale embryonic stem cells cell line. In brief, a loxP (locus ofX-over P1 sites recognized by Cre recombinase) NOX1 tar-geting construct with simultaneous insertion of an orphan loxPsite and an FRT-neo-FRT (Flippase recognition target [FRT]-flanked neomycin resistance cassette)-loxP resistance cas-sette was inserted into the NOX1 genomic locus to target exon3 to exon 8 (E3–E8) of NOX1 by Cre-mediated excision. Forgenotyping of NOX1 floxed mice the polymerase chain reaction(PCR) primers used were 5′- GTACTAAACAGAGAAAGGCTCTCCC-3′, 5′-TACTGGGACAGCTTCCTGCATCCC-3′, and 5′-CAACGGGTTCTTCTGTTAGTCC-3′. The PCR amplificationwas performed using the conditions of 95 °C for 3 minutes,followed by 35 cycles of 92 °C for 30 s, 61 °C for 30 s, and72 °C for 30 s. PCR products were separated on 1.8% aga-rose gel, with a 407-base pair band indicating the presenceof the mutant allele. A 527-base pair product from PCR indi-cates the WT allele. The mice were backcrossed in-house tothe C57BL/6 background for >10 generations. The floxed mice used to generate endothelial-specificNOX1 knockin mice were created in-house. A floxed stopcodon was inserted before the genomic locus of the NOX1gene that can be excised once Cre is induced by tamoxi-fen. For the NOX1CKI genotyping, the PCR primers usedwere 5′-CTCAAGCTTCGAATATGGGAAACTG-3′ and5′-GATCCAATGGCTTTCTCAGTGTACG-3′. The PCR amplifi-cation was performed using the conditions of 95 °C for 5 min-utes, followed by 29 cycles of 92 °C for 30 s, 56 °C for 30 s,and 72 °C for 30 s. PCR products were separated on 1.8%agarose gel, with a 290-base pair band indicating the presenceof a transgenic band. The mice were backcrossed in-house tothe C57BL/6 background for >10 generations. The Cdh5 (cadherin 5; PAC [P1 artificial chromosome])-CreERT2 (Cre recombinase–estrogen receptor T2 fusionprotein [tamoxifen-inducible Cre recombinase]) founder mice(endothelial Cre mice)31 were generously provided by Dr RalfAdams from the Vascular Development Laboratory, Cancer ### Page 4 Research UK London Research Institute, London, UnitedKingdom. For Cdh5 (PAC)-CreERT2 genotyping, PCR primersused were 5′-GCCTGCATTACCGGTCGATGCAACGAGTG-3′and 5′-CTGGCAATTTCGGCTATACGTAACAGGGTG-3′. ThePCR amplification was performed using the conditions of 95 °Cfor 5 minutes, followed by 29 cycles of 92 °C for 30 s, 60 °Cfor 30 s, and 72 °C for 30 s. PCR products were separated on1.8% agarose gel, with a 345-base pair band indicating thepresence of a transgenic band. The tg(Myh11-cre/ERT2)1Soff(vascular smooth muscle cells [VSMC] Cre) was originally pur-chased from Jackson Laboratory (Stock No: 019079). ForMyh11-cre/ERT2 genotyping, PCR primers used were 5′-TGACCC CAT CTC TTC ACT CC-3′ and 5′-AGT CCC TCACAT CCTCAGGTT-3′. The PCR amplification was performed using theconditions of 94 °C for 2 minutes, followed by 10 cycles of94 °C for 20 s, 65 °C for 15 s, and 68 °C for 15 s, and then fol-lowed by 28 cycles of 94 °C for 15 s, 60 °C for 15 s, and 72 °Cfor 15 s. PCR products were separated on 1.8% agarose gel,with a 287-base pair band indicating the presence of a trans-genic band. The NOX1 floxed mice were crossed with Cdh5-creor Myh11-cre to generate Cdh5cre-NOX1CKO (EC-specificNOX1 knockout mice) or Myh11cre (myosin heavy chain 11promoter–driven Cre recombinase [smooth muscle–specificCre expression])-NOX1CKO (VSMC-specific NOX1 knock-out mice) for experiments. The inducible NOX1flox/flox knockin/floxed mice were crossed with Cdh5-cre mice to generateCdh5cre-NOX1CKI (EC-specific NOX1 knockin/overexpress-ing) mice for experiments. All of the mice have been bred andmaintained at the University of California, Los Angeles (UCLA).Tamoxifen was dissolved in a sunflower seed oil/ethanol (10:1)mixture at 10 mg/mL, and 5-week-old Cdh5cre-NOX1CKO/WT, Myh11cre-NOX1CKO/WT, and Cdh5cre-NOX1CKI/WTmice were injected with 2 mg of tamoxifen intraperitoneally for5 consecutive days.8,32,33 ### Animal Studies Using WT and Various Knockoutand Knockin Mice All animals mentioned above and experimental procedureswere approved by the institutional animal care and use com-mittee at UCLA. Before experimentation, animals were kept inventilated cages with free access to water and standard chowand cared for by Division of Laboratory Animal Medicine staffuntil 8 weeks old. Mice were randomly divided into 2 dietarygroups and fed either regular chow (5053, 13.2% fat; LabDiet,St. Louis, MO) or a high-fat diet (with 42% calories from fat,TD.88137, Envigo, New Jersey, NY) for 8 weeks. ### Measurements of Food and Water Intake, andDetermination of Oxygen Consumption inIndividual Animals Using Metabolic Cages Diet and water intake of mice were monitored weekly for 8consecutive weeks until the end of the protocol. Oxygen con-sumption was measured using an Oxymax ComprehensiveLab Animal Monitoring System (Oxymax-CLAMS; ColumbusInstruments, Columbus, OH). Eight weeks after the mice werefed a regular diet or a high-fat diet, the animals were subjectedto Metabolic Cage analyses. Animals were individually housedin an open-circuit chamber for 3 days with ad libitum accessto food and water. Oxygen consumption was measured by theOxymax system for mice housed individually, and the data wereanalyzed using the CLAMS Examination Tool (CLAX; ColumbusInstruments), version 2.1.0.34 ### Assessment of Spontaneous Activity After 8 weeks of dietary intervention (regular chow versus high-fat diet), mouse spontaneous activity in the 12-hour dark cyclewas monitored by video-recording using infrared webcams, andthe real-time motions of the individual animals were analyzedusing motion-detection software as we previously published.8 ### Assessment of Exercise Capacity As mentioned above, mice were randomly assigned to either aregular chow or a high-fat diet group and fed accordingly for 8weeks before subsequent experiments. Mouse exercise capac-ity was examined using a specialized mouse Treadmill (Exer3/6 model, Columbus Instruments) as previously describedwith modifications.35 Before measurement, mice were trainedfor 2 days at a constant speed of 6 m/min for 30 minutes.For assessment of exercise capacity, mice were subjected toa 5-minute warm-up period at 6 m/min, after which tread-mill speed was incrementally increased to 30 m/min, and themice were run for 40 minutes with speed escalating at 1.2 m/min every 2 minutes. An electric shock at 1.0-mA current wasapplied at the back of the treadmill to motivate mice to run.Exhaustion was defined as the point when mice remained onshockers for >10 s and were not able to jump back on thetrack. Exercise capacity was evaluated by time, total distancetraveled, and speed at exhaustion point. ### Mitochondrial Isolation and Assessment ofRespirational Capacity Upon harvesting of tissues, mitochondria were freshly isolatedby differential centrifugation. In brief, freshly harvested skeletalmuscles were homogenized using a DWK Life Sciences KimbleKontes Duall Tissue Grinders (size 23, Kimble, Thermo-FisherScientific, Grand Island, NY) in homogenization buffer con-taining 250-mmol/L sucrose, 1-mmol/L ethylene glycol-bis(β-aminoethyl ether)-N, N, N′, N′-tetraacetic acid, 10-mmol/LHEPES, and 10-mmol/L Tris-HCl, pH 7.4. The homogenateswere centrifuged at 800g for 7 minutes, and then, the superna-tants were transferred to new tubes and centrifuged at 4000gfor 15 minutes. The pellet was rinsed with homogenizationbuffer without ethylene glycol-bis (β-aminoethyl ether)-N, N,N′, N′-tetraacetic acid and then centrifuged at 4000g for 15minutes. Rinsed pellets were resuspended in homogenizationbuffer without ethylene glycol-bis (β-aminoethyl ether)-N, N, N′,N′-tetraacetic acid. Respirational capacity of fresh mitochondriawas examined using a Seahorse XF 24-3 analyzer (SeahorseBioscience, UCLA Biosciences Cellular Bioenergetics Core)according to the manufacturer’s instructions. The respira-tory control ratio (RCR) was calculated as State III/State IVo,induced by the addition of Oligomycin. ### Mitochondrial Swelling Assay Freshly isolated mitochondria were incubated with swellingbuffer containing 250-mmol/L sucrose, 10-mmol/L Tris (pH7.4), and 5-mmol/L succinate for 1 minute at room temperature ### Page 5 before swelling was initiated by the addition of 250-μmol/LCaCl2. Mitochondrial swelling was measured by monitoringthe decrease in absorbance at 540 nm using a Synergy PlateReader (Synergy H1 Hybrid Reader, Biotek/Agilent, SantaClara, CA) as we previously described.8,27 ### Examination of Mitochondrial Cristae Structuresby Electron Microscopy Freshly harvested soleus muscle was immediately immersed inelectron microscopy grade fixing solution (0.1M PB with 2.5%glutaraldehyde and 4% paraformaldehyde, pH 7.4). Specimenprocessing and imaging were performed at the UCLA BrainResearch Institute Electron Microscopy Core Laboratory.Muscular mitochondria were imaged by JEOL 100CX trans-mission electron microscope (JEOL Ltd) at magnifications of×2900 and ×7200, respectively. ### RNA-Sequencing/RNA-Sequencing Analysis RNAs isolated from soleus muscles of experimental mice (3each group from 4 experimental groups of Cdh5cre-WT miceunder regular diet, Cdh5cre-WT mice under high-fat diet,Cdh5cre-NOX1CKO mice under regular diet, and Cdh5cre-NOX1CKO mice under high-fat diet) were subjected to RNA-sequencing at the UCLA Technology Center for Genomics &Bioinformatics. KAPA Stranded RNA-SeqKit was used for pre-paring the libraries for RNA-sequencing. Sequencing was per-formed on an Illumina HiSeq 3000 for a single read, 50 runs.Data quality checks were done on Illumina SAV. Demultiplexingwas performed by using Illumina Bcl2fastq2, version 2.17, pro-gram. Reads were first mapped to the latest UCSC transcriptset with Bowtie2, version 2.1.0, and gene expression levels wereestimated by RSEM, version 1.2.15. Gene expression was nor-malized by the trimmed mean of M-values. The edgeR programwas used to identify the differentially expressed genes. RNAspecies that increase >2-fold or decrease >2-fold in the soleusmuscle of WT mice under high-fat feeding (group 3) comparedwith WT mice with a regular diet (group 1) were screened out.Then, protein-coding RNAs were chosen from these RNAspecies. The protein-coding RNAs decreased by >1.5-fold orincreased by >1.5-fold in Cdh5Cre-NOX1CKO mice underhigh-fat feeding (group 4) compared with WT littermates underhigh-fat feeding (group 3) were selected. The significantlyupregulated or downregulated protein-coding RNAs (P<0.05)were further analyzed by Ingenuity Pathway Analysis (IPA; seebelow) to explore their relationship to skeletal muscle function. ### IPA Analysis The differentially regulated protein-coding RNAs identifiedby RNA-sequencing analysis were analyzed by QIAGEN IPA(content version: 68752261,[-](https://www.qiagenbioinformatics.com/products/ingenuity-pathway-analysis)) to reveal thoseimplicated in skeletal and muscular disorders for subsequentanalyses. ### Analysis of NOX1 Expression in HumanPatients Coronary artery samples of a total of 15 human subjects (bodymass index [BMI] <25, n=6; BMI >25, n=9) were used forthe determination of NOX1 expression. All protocols wereapproved by the institutional review board at the MedicalCollege of Wisconsin to collect fresh, otherwise discarded, ex-planted, donated hearts from the Versiti Institute. Specimensfrom individuals were categorized based on BMI. Coronaryarteries were dissected and immediately transferred to ice-coldPBS solution to preserve tissue integrity. Connective tissue wascarefully removed, and the arteries were divided into sections.Sections of the coronary arteries were lysated and subjected toWB analysis for NOX1 expression as described above. ### Statistical Analysis Statistical analyses were performed using GraphPad Prism10.6.0 (GraphPad Software, San Diego, CA). Normality wasassessed using the Shapiro-Wilk test for data sets with n≥6;normally distributed data were analyzed using parametric tests,while nonnormally distributed data or data sets with n<6 wereanalyzed using nonparametric tests. For 2-group comparisons, the Student t test was used fornormally distributed data with equal variances (determined bythe F-test), the Welch t test for normally distributed data withunequal variances, and the Mann-Whitney U test for nonpara-metric data. For comparisons involving ≥3 groups with n≥6,variance homogeneity was assessed using the Brown-Forsythetest, and 1-way ANOVA or 2-way ANOVA (for 2 independentfactors) with the Tukey or Newman-Keuls multiple comparisontest was used or nonparametric data with the Kruskal-Wallistest followed by the Dunn post hoc test. For comparisonsinvolving 4 groups with n<6, aligned rank transform ANOVAfollowed by 2-tailed Welch t tests was used (eg, for data inFigure 1F, when a 2-group comparison between WT and NOX1gKO mice under a high-fat diet is the data to be analyzed). For experiments with repeated measurements with n≥6,overall effects were assessed by 2-way repeated-measuresANOVA, and differences at individual time points were evalu-ated using 2-way ANOVA with the Tukey or Newman-Keulspost hoc test. For experiments with repeated measurementswith n<6, a mixed-effects model (restricted maximum likeli-hood [REML]) with the Dunnett or Holm-Šídák multiple com-parison tests was used. Nonlinear regression (nonlinear curvefitting) was applied to analyze relaxation responses. Post hocpairwise comparisons were adjusted for multiple testing, andstatistical significance was defined as 2-sided P<0.05. ## RESULTS ### Global NOX1 Knockout Selectively AttenuatesHigh-Fat Feeding–Induced Obesity andMetabolic Syndrome We have previously demonstrated that vascular-drivenROS production mediates high-fat-diet–induced obe-sity and metabolic syndrome, establishing a paradigm-shifting concept that vascular ROS function servesas a causal factor, rather than a consequence, for/ofthe development of obesity and metabolic syndrome.Overexpression of p22phox in the vascular cells resultsin activation of NOX and consequent skeletal mito-chondrial dysfunction to result in impaired spontaneous ### Page 6 ![Figure 1]() activity and energy expenditure.8 Of note, all of theNOX isoforms, except for NOX5 that is not presentin rodents, require p22phox for enzymatic activation.11NOX3 is expressed primarily in the inner ear.11 There-fore, we next examined direct causal role(s) of NOX iso-forms (NOX1, NOX2, and NOX4) in mediating high-fat ### Page 7 feeding–induced obesity and metabolic syndrome,employing high-fat–fed global knockout mice with dele-tion in NOX1, NOX2, or NOX4. As shown in Figure 1,in NOX1 global knockout mice, verified by genotyping(Figure 1A), high-fat-diet–induced body weight gainwas substantially attenuated (Figure 1B and 1C). How-ever, body weight gain was not affected by NOX2 globalknockout (), while NOX4 global knockoutincreased body weight by contrast in response to high-fat feeding (). Consistent with decreasedbody weight gain, adipose mass gain was also reducedalmost to regular diet levels in high-fat–fed NOX1global knockout mice (Figure 1D), while lean mass wasnot different between high-fat–fed NOX1 global knock-out and WT littermate groups (Figure 1E). High-fat-diet–induced increase in liver weight was also markedlyattenuated in NOX1 global knockout mice (Figure 1F).Of note, organ weights of lung, spleen, and kidney werenot affected by high-fat feeding or regulated in NOX1global knockout mice, while NOX1 global knockout alsohad no effects on elevated heart weights in response tohigh-fat feeding (). In addition,primary preadipocytes were isolated from high-fat–fedWT, NOX1 global knockout, and tgsm/p22phox mice. Cul-tured preadipocytes were induced to differentiate usinga differentiation cocktail, and adipocyte differentiationactivity was analyzed as described in the Methods sec-tion. As shown in Figure 1G, NOX1 deletion signifi-cantly attenuated preadipocyte differentiation, which isconsistent with reduced fat mass gain in vivo. On thecontrary, preadipocytes from high-fat–fed tgsm/p22phoxmice had increased adipogenesis (Figure 1G). We next examined whether global NOX1 deletionis able to modify metabolic parameters under high-fatfeeding. We examined glucose clearance rate in high-fat–fed animals. As shown in Figure 1H, intraperitonealglucose challenge revealed a marked delay in glucoseclearance or increase in glucose intolerance in high-fat–fed WT mice compared with the regular diet group, whileNOX1 global knockout substantially attenuated thisresponse. However, high-fat–fed NOX2 global knockoutmice () and NOX4 global knockout mice ([-](https://www.ahajournals.org/doi/suppl/10.1161/CIRCRESAHA.125.326768)) displayed no difference from high-fat–fed WTmice in glucose intolerance. Of note, circulating insulinand leptin levels were significantly upregulated in high-fat–fed WT mice compared with the regular diet group(Figure 1I and 1J). NOX1 global deletion in mice, how-ever, markedly attenuated plasma insulin levels at 6 and8 weeks (Figure 1I) and leptin levels at 2, 4, 6, and 8weeks (Figure 1J). Importantly, NOX1 protein expressionwas indeed markedly upregulated in high-fat–fed WTmice (Figure 1K). Taken together, these results clearly demonstrate thatNOX1, instead of NOX2 or NOX4, selectively mediateshigh-fat feeding–induced obesity and metabolic syn-drome in mice. ### Endothelial-Specific Deletion of NOX1Abrogates High-Fat Feeding–Induced Obesityand Metabolic Syndrome We have shown that NOX-derived vascular ROS playa causal role in the development of obesity and meta-bolic syndrome,8 and that NOX1 deletion robustly alle-viates high-fat feeding–induced phenotypes of obesityand metabolic syndrome as described above. NOX1is expressed in ECs, smooth muscle cells, and cardio-myocytes in the cardiovascular system.11 Based on ourprevious findings that NOX activation impairs skeletalmuscle mitochondrial function, hence impairing spon-taneous activity,8 we hypothesize that vascular NOX1activation drives initial ROS production under high-fatfeeding and consequent ROS-dependent triggering ofmitochondrial dysfunction in nearby skeletal muscle. Todissect the specific cell/tissue type in the vasculaturethat is specifically involved in NOX1-driven develop-ment of obesity and metabolic syndrome, we generatedendothelial or VSMC-specific NOX1 knockout mice forthe first time in-house and subjected the animals tohigh-fat feeding. To induce cell type–specific knockout of NOX1, wefirst created a Cre-inducible NOX1 knockout gene con-struct by inserting an orphan loxP site and an FRT-neo-FRT-loxP resistance cassette inserted into the NOX1genomic locus to target exon 3 to exon 8 (E3–E8) ofNOX1 (Figure 2A). The floxed NOX1 founders wereverified by genotyping (Figure 2B) and backcrossedin-house to the C57BL/6 background for >10 genera-tions. Then, the mice were crossed with Cdh5Cre mice,31which were also verified by genotyping (Figure 2B) togenerate endothelial-specific NOX1 knockout mice(Cdh5Cre-NOX1CKO mice). Using EC washout experi-ments as we previously published to examine EC-spe-cific gene expression, we found that NOX1 expressionwas successfully knocked out in Cdh5Cre-NOX1CKOmice compared with the littermate controls (Figure 2C).NOX1 expression was only blunted in the EC fractionbut not in the EC-denuded aortic fraction (Figure 2C).eNOS (endothelial nitric oxide synthase) was used asan internal control for successful isolation of ECs fromthe aortas, with it being present only in the EC fraction(Figure 2C). Eight-week-old Cdh5Cre-NOX1CKO mice and lit-termate controls were subjected to a high-fat diet or aregular diet for 8 weeks. As shown in Figure 2D and 2E,compared with Cdh5Cre-WT littermates, high-fat feed-ing–induced body weight gain was completely attenu-ated in Cdh5Cre-NOX1CKO mice, to the baseline asin regular diet–fed mice. Consistent with decreasedbody weight gain, adipose mass gain was also substan-tially lower in high-fat–fed Cdh5Cre-NOX1CKO mice(Figure 2F and 2G), while lean mass was not differentbetween the Cdh5Cre-NOX1CKO and Cdh5Cre-WT ### Page 8 ![Figure 1]() ### Page 9 littermate groups after high-fat feeding (Figure 2H and2I). In addition to decreased fat mass gain, we alsoobserved markedly attenuated liver weight in high-fat–fed Cdh5Cre-NOX1CKO mice compared with theirCdh5Cre-WT littermates (Figure 2J), whereas organweights of heart, lung, spleen, and kidney were notaffected by endothelial-specific NOX1 knockout (). Endothelial-specific NOX1 deletionalso significantly attenuated preadipocyte differentiation(Figure 2K), which is consistent with reduced fat massgain in vivo. We further examined metabolic parameters in high-fat–fed endothelial-specific NOX1 knockout mice. Asshown in Figure 2L, intraperitoneal glucose challengerevealed that endothelial-specific NOX1 deletion com-pletely attenuated high-fat feeding–induced glucoseintolerance to the levels in regular diet–fed animals. Inaddition, endothelial-specific NOX1 deletion in miceabrogated elevations in plasma insulin levels at 8 weeks(Figure 2M) and that in plasma leptin levels at 6 and 8weeks (Figure 2N). Taken together, these results for the first time dem-onstrate that NOX1 situated in the ECs plays an impor-tant role in mediating high-fat-diet–induced obesity andmetabolic syndrome in mice. ### Smooth Muscle–Specific Deletion of NOX1 HasNo Effect on High-Fat Feeding–Induced Obesityand Metabolic Syndrome NOX1 is expressed in ECs and vascular smooth musclecells in the vasculature.11 NOX1 in either cell type canbe activated first to potentially transduce ROS signalsto provoke mitochondrial dysfunction in adjacent skeletalmuscle. Therefore, we also examined whether specificdeletion of NOX1 in smooth muscle cells has any effectson high-fat feeding–induced obesity and metabolicsyndrome. NOX1 floxed mice were crossed with Myh-11Cre mice to generate smooth muscle–specific NOX1knockout mice (Myh11Cre-NOX1CKO mice) for the firsttime, which were verified by genotyping (). Asshown in, body weight gain andfat mass gain in Myh11Cre-NOX1CKO and Myh11Cre-NOX1WT mice were significantly and equally increasedin response to high-fat feeding compared with chew diet–fed Myh11Cre-NOX1CKO and Myh11Cre-NOX1WTmice, respectively. Notably, lean mass was increased inMyh11Cre-NOX1CKO mice compared with Myh11Cre-WT littermates after high-fat feeding for 4 to 8 weeks(), but there was no significantchange when normalized to the tibial length. We also examined whether smooth muscle–spe-cific NOX1 deletion was able to modify metabolicparameters under high-fat feeding. As shown in[-](https://www.ahajournals.org/doi/suppl/10.1161/CIRCRESAHA.125.326768), intraperitoneal glucose challenge revealedequally impaired glucose tolerance in high-fat–fedMyh11Cre-NOX1CKO and Myh11Cre-NOX1WT micecompared with regular diet–fed Myh11Cre-NOX1CKOand Myh11Cre-NOX1WT mice, respectively. In addition,high-fat feeding resulted in similar phenotypes of insulinand leptin resistance in both Myh11Cre-NOX1CKO miceand Myh11Cre-WT mice (). Of note,no evidence is available on NOX1 regulation in vascularsmooth muscle cells of human obese subjects, indicat-ing potential irrelevance of its regulation. Taken together,these data indicate that smooth muscle–specific NOX1ablation in mice has no effect on phenotypes of obesityand metabolic syndrome. Therefore, NOX1 activation inECs, rather than in smooth muscle cells, plays a specificand selective causal role in mediating high-fat-diet–induced obesity and metabolic syndrome in mice. ### Endothelial-Specific NOX1 TransgenesisInduces Exaggerated Obesity and MetabolicSyndrome in Response to High-Fat Feeding To further evaluate the intermediate role of endothelialNOX1 in high-fat feeding–induced obesity and metabolicsyndrome, cre-inducible NOX1 knockin/floxed mice weregenerated for the first time in-house, as described in Fig-ure 3A, and backcrossed to the C57BL/6 backgroundfor >10 generations. Then, the mice were crossed withCdh5Cre mice to generate endothelial-specific NOX1knockin (Cdh5Cre-NOX1CKI mice) mice, which wereverified by genotyping (Figure 3B). Using EC washoutexperiments as we previously published to examine EC-specific gene expression, we found that NOX1 expres-sion was successfully knocked in in Cdh5Cre-NOX1CKOmice compared with the littermate controls (Figure 3C).NOX1 expression was only enhanced in the EC fractionbut not in the EC-denuded aortic fraction (Figure 3C).eNOS was used as an internal control for successful ### Page 10 ![Figure 2]() ### Page 11 isolation of ECs from the aortas, with it being presentonly in the EC fraction (Figure 3C). Eight-week-old Cdh5Cre-NOX1CKI and littermatecontrol mice were subjected to a high-fat diet or a regu-lar diet for 8 weeks. As shown in Figure 4D and 4E, bodyweight gain of Cdh5Cre-NOX1CKI mice was signifi-cantly further increased compared with that of Cdh5Cre-WT littermates on a high-fat diet. Consistently, adiposemass gain was also significantly further increased inhigh-fat–fed Cdh5Cre-NOX1CKI mice (Figure 3Fand 3G), while lean mass was not different betweenCdh5Cre-NOX1CKI and Cdh5Cre-WT littermate groups(Figure 3H and 3I). Of note, liver weight in high-fat–fedCdh5Cre-NOX1CKI mice was also significantly higherthan that of high-fat–fed Cdh5Cre-WT littermates (Fig-ure 3J), whereas organ weights of heart, lung, spleen,and kidney were not different in endothelial-specificNOX1 knockin mice compared with control littermates(). Furthermore, preadipocytesisolated from high-fat–fed Cdh5Cre-NOX1CKI mice hada significantly higher differentiation rate compared withthat of Cdh5Cre-WT littermate mice fed a high-fat diet(Figure 3K), which is consistent with further increasedfat mass gain in vivo. In addition, high-fat–fed Cdh5Cre-NOX1CKI micedisplayed more severely delayed glucose clearancecompared with Cdh5Cre-WT littermates (Figure 3L).High-fat–fed Cdh5Cre-NOX1CKI mice also had furtherincreased plasma insulin levels at 2, 4, 6, and 8 weeks(Figure 3M) and further elevated leptin levels at 4, 6, and8 weeks (Figure 3N). Taken together, these results additionally confirm thatendothelial-specific activation of NOX1 plays a vital rolein mediating high-fat feeding–induced obesity and meta-bolic syndrome. ### NOX1 Global Knockout Abrogates Obesityand Metabolic Syndrome via Preservation ofSkeletal Function to Maintain SpontaneousActivity and Exercise Capacity Because NOX1 global knockout and WT mice hadsimilar energy and water intake during high-fat feed-ing (Figure 4A and 4B), we examined whether energyexpenditure was altered in high-fat–fed NOX1 globalknockout mice by measuring spontaneous activity andexercise capacity. We assessed the spontaneous activityof regular diet–fed or high-fat-diet–fed animals using aninfrared video monitoring system as we previously pub-lished.8 Of note, spontaneous activity was increased inNOX1 global knockout mice at baseline when fed a reg-ular diet (Figure 4C). High-fat feeding markedly reducedspontaneous activity in WT littermates, which was com-pletely restored to more than baseline in NOX1 globalknockout mice (Figure 4C), indicating robust protectiveeffects of NOX1 deletion in preserving spontaneousactivity to maintain normal energy expenditure duringhigh-fat feeding. To further evaluate exercise capacity inhigh-fat–fed NOX1 global knockouts versus WT litter-mates, we used a mouse treadmill to measure exercisecapacity, with parameters recorded for exhaustion time,exhaustion travel distance, and exhaustion speed foreach animal. Compared with high-fat–fed WT littermates,NOX1 global knockout mice fed a high-fat diet displayedsignificantly longer exhaustion time, longer traveled dis-tance, and higher exhaustion speed (Figure 4D through4F, respectively). Importantly, the completely preservedspontaneous activity and much improved exercise capac-ity ensure sufficient energy expenditure to alleviate thedevelopment of obesity. It is known that oxygen consumption is positively cor-related with physical activity. We next examined real-timeoxygen consumption in mice using metabolic cages(Oxymax-CLAMS; Columbus Instruments). We foundthat oxygen consumption was enhanced in high-fat–fedNOX1 global knockout mice compared with WT litter-mates (Figure 4G and 4H). The VCO2 showed the sametrend as for VO2 (Figure 4I and 4J). These data indicatethat NOX1 global knockout restores energy expenditurein high-fat–fed animals. The increased energy expendi-ture is attributed to completely preserved spontaneousactivity and augmented exercise capacity as describedabove. In addition, we examined ATP production by assess-ing skeletal muscle mitochondrial respiratory control ratio(RCR) using Seahorse. Skeletal muscle mitochondrialRCR was significantly decreased in high-fat–fed WT lit-termates, which was, however, completely recovered tobaseline by NOX1 global knockout (Figure 4K). Giventhat RCR is an index of the tightness of coupling betweenrespiration and phosphorylation, a higher value of RCRin NOX1 global knockout mice indicates a tighter linkbetween the amount of O2 consumed and the formation ### Page 12 ![Figure 3]() ### Page 13 of ATP, confirming that the mitochondria are functioningmore efficiently in NOX1 global knockout mice. Likewise,calcium-induced mitochondrial swelling, as a readout ofmitochondrial dysfunction, was significantly increased inhigh-fat–fed WT littermates, which was, however, com-pletely attenuated in NOX1 global knockout mice (Fig-ure 4L), indicating preserved mitochondrial integrity. Inaddition, we examined the cristae structure of the mito-chondria in the soleus muscle by transmission electronmicroscopy. Notably, swollen mitochondria with perturba-tion of cristae structure were observed in tgsm/p22phox micefed with a high-fat diet (Figure 4M). It has been reportedthat the oxidative phosphorylation system is organized infolded cristae.36 Upon perturbation of cristae, the elec-tron transport chain is less efficient, and mitochondrialperformance will decrease. Importantly, the density andstructure of mitochondrial cristae were fully restoredin high-fat–fed NOX1 global knockout mice comparedwith that of WT littermates (Figure 4M), further indicat-ing preserved mitochondrial respiratory performance. Wealso measured mitochondrial ROS production as one ofthe indicators of changes in mitochondrial function. Asshown in Figure 4, skeletal muscle mitochondrial pro-duction of superoxide was significantly decreased inhigh-fat–fed NOX1 global knockout mice compared withWT littermates. Taken together, these data indicate thatNOX1 ablation preserved spontaneous activity and exer-cise capacity via maintenance of skeletal muscle func-tion attributed to preserved mitochondrial function, whichwas reflected by increased oxygen consumption andATP production, reduced mitochondrial swelling activityand superoxide production, and restored mitochondrialcristae structure. ### Endothelial-Specific NOX1 Deletion AbrogatesObesity and Metabolic Syndrome viaPreservation of Skeletal Function to MaintainSpontaneous Activity and Exercise Capacity Because global NOX1 ablation protects mice fromhigh-fat feeding–induced obesity via preservation ofmitochondrial function to maintain spontaneous activityand exercise capacity as described above, we furtherexamined the specific role of endothelial NOX1 in thisresponse. Because Cdh5Cre-NOX1CKO and Cdh5Cre-WT mice had similar food and water intake under high-fatfeeding (Figure 5A and 5B), we compared spontaneousactivity and exercise capacity between the 2 groups.High-fat feeding markedly reduced spontaneous activityin WT littermates, which was restored to near baseline inCdh5Cre-NOX1CKO mice (Figure 5C), indicating robustprotective effects of endothelial NOX1 deletion in pre-serving spontaneous activity to maintain normal energyexpenditure during high-fat feeding. Data indicate thatcompared with high-fat–fed WT littermates, high-fat–fedCdh5Cre-NOX1CKO mice also had preserved exercisecapacity, reflected by significantly longer exhaustiontime, longer traveled distance, and higher exhaustionspeed (Figure 5D through 5F, respectively). Besides,oxygen consumption (Figure 5G and 5H) and carbondioxide production (Figure 5I and 5J) were both signifi-cantly enhanced in high-fat–fed Cdh5Cre-NOX1CKOmice compared with Cdh5Cre-WT littermates. Thesedata indicate that endothelial NOX1 knockout restoresenergy expenditure via preserved spontaneous activityand exercise capacity to protect from high-fat feeding–induced obesity and metabolic syndrome. Skeletal mitochondrial functions were examined nextto validate the elucidated mechanisms, if similar in NOX1global knockout mice. Transmission electron microscopyresults revealed that the density and structure of mito-chondrial cristae in soleus muscle Cdh5Cre-NOX1CKOwere fully restored compared with high-fat–fedCdh5Cre-WT littermates (Figure 5K). Besides, the signifi-cant increase in calcium-induced mitochondrial swellinginduced by high-fat feeding in Cdh5Cre-WT littermateswas completely attenuated in Cdh5Cre-NOX1CKO mice(Figure 5L), indicating protected mitochondrial integrity byendothelial-specific NOX1 deletion. In addition, increasedmitochondrial superoxide production in skeletal muscle ofhigh-fat–fed Cdh5Cre-WT littermates was significantlyalleviated in high-fat–fed Cdh5Cre-NOX1CKO mice(Figure 5M). Therefore, endothelial-specific NOX1 dele-tion protects against high-fat feeding–induced obesity viapreservation of skeletal muscle mitochondrial function tomaintain efficient energy expenditure. ### Smooth Muscle–Specific NOX1 Deletion HasNo Effect on Skeletal Function and ExerciseCapacity in Response to High-Fat Feeding In addition to phenotype analyses for obesity and meta-bolic syndrome in high-fat–fed Myh11Cre-NOX1CKOmice as described above, we also examined changesin skeletal function and exercise capacity in these ani-mals under high-fat feeding. As shown in, data from treadmill exercise indicate thatMyh11Cre-NOX1CKO and Myh11Cre-WT littermateshad similar exhaustion time, distance traveled, andexhaustion speed under high-fat feeding. In addition, ### Page 14 ![Figure 4]() ### Page 15 calcium-induced mitochondrial swelling and mitochon-drial superoxide production were not affected in high-fat–fed Myh11Cre-NOX1CKO mice compared withhigh-fat–fed Myh11Cre-WT littermates (). Therefore, endothelial NOX1, rather than smoothmuscle NOX1, plays a critical intermediate role in high-fat-diet–induced obesity and metabolic syndrome,knockout of which protected against the phenotypes viapreservation of skeletal mitochondrial function to main-tain spontaneous activity and exercise capacity for suf-ficient energy expenditure. ### Endothelial-Specific NOX1 TransgenesisWorsens Obesity via Deterioration of SkeletalDysfunction to Impair Exercise Capacity In further experiments, we also examined exercise capacityand skeletal mitochondrial function in endothelial-specificNOX1 knockin mice. As shown in Figure 6A and 6B,Cdh5Cre-NOX1CKI and Cdh5Cre-WT mice had simi-lar food and water intake. With treadmill training, high-fat–fed Cdh5Cre-NOX1CKI mice displayed significantlyshorter exhaustion time, shorter traveled distance, andlower exhaustion speed compared with high-fat–fedCdh5Cre-WT littermates (Figure 6C through 6E, respec-tively). In addition, mitochondrial swelling in the skeletalmuscle of Cdh5Cre-NOX1CKI mice was deterioratedcompared with WT littermates after high-fat feeding(Figure 6F), indicating worsened mitochondrial integrityby overexpression of endothelial NOX1. Transmissionelectron microscopy results showed lower density anddisrupted structure of mitochondrial cristae in Cdh5Cre-NOX1CKI compared with Cdh5Cre-WT mice under high-fat feeding (Figure 6G). Skeletal muscle mitochondrialsuperoxide production was also significantly increasedin high-fat–fed Cdh5Cre-NOX1CKI mice comparedwith high-fat–fed Cdh5Cre-WT littermates (Figure 6H).Therefore, overexpression of NOX1 specifically in theendothelium deteriorated exercise incapacity and skeletalmitochondrial dysfunction to further worsen phenotypesof obesity and metabolic syndrome. ### Restoration of Endothelium-DependentVasorelaxation in High-Fat Fed Global andEndothelial-Specific NOX1 Knockout Mice Endothelium-dependent vasorelaxation was assessedas we previously reported.34,37 Importantly, the impairedendothelium-dependent vasorelaxation in high-fat–fedWT mice was completely restored in both global NOX1knockout mice () and endothelial-specificNOX1 knockout mice (), confirming improve-ment in endothelial function in these animals underhigh-fat diet feeding. This supports the hypothesis thatendothelial-specific knockdown of NOX1 exerts protec-tive effects on obesity and metabolic syndrome via ini-tial preservation of endothelial function and subsequentprotection of adjacent skeletal muscle mitochondrialfunction, resulting in restored spontaneous and exerciseactivities to facilitate energy expenditure. ### Novel Genetic Signatures in High-Fat FedEndothelial-Specific NOX1 Knockout MiceRelevant to Skeletal Muscle Function: Cntnap4and Col9a1 An RNA-sequencing experiment was performed to elu-cidate potential targets of endothelial-specific activationof NOX1 in modulating skeletal muscle function to resultin obesity and metabolic syndrome. The data indicatethat 1912 RNA species were upregulated > 2-fold inthe soleus muscle of WT mice under high-fat feeding(group 3) compared with WT mice with a regular diet(group 1), among which 667 were protein-coding RNAs(Figure 7A). Among the 667 differentially expressedgenes, 408 were downregulated >1.5-fold in Cdh5Cre-NOX1CKO mice under high-fat feeding (group 4) com-pared with WT littermates under high-fat feeding (group3; Figure 7B), and 23 protein-coding RNAs were fur-ther screened out by significant differences (P<0.05)between group 3 and group 1 (Figure 7B and 7C). Inaddition, 4 protein-coding RNAs were significantly down-regulated (P<0.05) in group 4 compared with group 3among the 23 protein-coding RNAs (Figure 7B and 7D). ### Page 16 ![Figure 2]() ### Page 17 ![Figure 3]() ### Page 18 On the other hand, 1998 RNA species were down-regulated >2-folds in group 3 compared with group1, and 665 were protein-coding RNAs among them(Figure 7E). Out of these 665 differentially expressedgenes, 336 protein-coding RNAs were increased >1.5-fold in group 4 compared with group 3 (Figure 7F), and32 protein-coding RNAs were filtered out according tosignificant differences (P<0.05) between group 3 andgroup 1 (Figure 7F and 7G). Furthermore, 7 protein-coding RNAs were significantly upregulated (P<0.05) ingroup 4 compared with group 3 among the 32 protein-coding RNAs (Figure 7F and 7H). Next, QIAGEN IPA was used to explore the relation-ship between the modulated genes in the soleus muscleof Cdh5Cre-NOX1CKO mice and the function of skele-tal and muscular disorders. Among the 4 downregulatedgenes (group 4 versus group 3) and 7 upregulated genes(group 4 versus group 3) screened out by 4 consecutivesteps of comparison as described above, only Cntnap4and Col9a1 were found to connect to the function ofskeletal and muscular disorders (Figure 7I). Results ofWestern blotting analyses confirmed that Cntnap4 wasincreased by high-fat feeding while decreased by endo-thelial NOX1 knockout, and Col9a1 was upregulatedby high-fat feeding while downregulated by endothelialNOX1 knockout (Figure 7J). Therefore, Cntnap4 and Col9a1 in the soleus mus-cle may represent novel genetic signatures underlyingendothelial-specific NOX1 activation-driven develop-ment of obesity and metabolic syndrome, via impairmentin spontaneous activity and exercise capacity under ahigh-fat diet. ### Upregulation of NOX1 in Coronary Arteries ofHuman Patients With Obesity To further validate the translational relevance of our find-ings, we examined the expression of NOX1 in humancoronary artery samples isolated from subjects with obe-sity and normal controls. Coronary artery samples of atotal of 15 human subjects (BMI <25, n=6; BMI >25,n=9) were used for the determination of NOX1 expres-sion. The normal weight group consisted of 6 subjects(age: 35.5±9.17 years) with BMI<25, at 18.8±1.33,while the obese group of 9 patients had BMI>25 at32.5±1.79 (age: 54.4±4.17 years). The authoriza-tion and collection of the samples are described in theMethods section. Freshly isolated coronary arteries weredissected, removed of connective tissue carefully, sec-tioned for separate storage, and snap-frozen. Sectionsof the coronary arteries were lysated and subjected toWB analysis for NOX1 expression as described in theMethods section. Importantly, we found that NOX1 pro-tein abundance was significantly increased in the coro-nary arteries of human patients with obesity. Shown arerepresentative Western blots and grouped quantitativedata in Figure 8. One limitation is that we have not beenable to isolate EC fractions from these tissue samples toexamine NOX1 expression selectively in ECs, which war-rants further investigation when additional samples arecollected from future cohorts. ![Figure 4]() ## DISCUSSION The most significant findings of this study are given asfollows: (1) global knockout of NOX1, rather than ofNOX2/NOX4, substantially abrogated features of obe-sity and metabolic syndrome, including increased bodyweight, fat mass, liver weight, adipocyte differentiation,glucose intolerance, plasma insulin, and leptin levels. (2)endothelial-specific NOX1 deletion, rather than vascularsmooth muscle–specific NOX1 deletion, substantiallyalleviated features of obesity and metabolic syndromein response to high-fat feeding, establishing a novel,selective, and tissue-specific role of endothelial NOX1 ### Page 19 in driving systematic metabolic responses. This is furtherconfirmed by data collected using endothelial-specificNOX1 knockin mice, in which high-fat feeding resulted inexaggerated phenotypes of obesity and metabolic syn-drome. (3) Global and EC-specific deletion of NOX1 inmice led to preserved energy expenditure under high-fatfeeding via restoration of skeletal muscle mitochondrialfunction to result in preserved spontaneous activity andexercise capacity, therefore alleviating the developmentof obesity and metabolic syndrome. This is further vali-dated by data from EC-specific knockin mice in whicha high-fat-diet–induced worsened skeletal muscle mito-chondrial dysfunction and exercise incapacity. (4) RNA-sequencing analyses identified novel genetic signaturesof Cntnap4 and Col9a1 in soleus muscles of endothelial-specific NOX1 knockout mice under a high-fat diet, whichare linked to the modulation of skeletal muscle function.(5) NOX1 expression was significantly upregulated inhuman patients with obesity, indicating robust transla-tional relevance of our findings. These observations areparadigm-shifting in establishing a novel, significant,specific, and selective role of endothelial-specific activa-tion of NOX1 in driving obesity and metabolic syndromeunder high-fat feeding, via deficiencies in skeletal musclemitochondrial function to result in impaired spontaneousactivity, exercise capacity, and energy expenditure. Accumulating evidence has demonstrated that theNOX family of enzymes plays a major role in the patho-genesis of various vascular diseases, such as hyperten-sion, aortic aneurysms, atherosclerosis, and diabeticvascular complications, as well as for the development ofcardiac disorders, including ischemia-reperfusion injury,myocardial infarction, heart failure, and cardiac arrhyth-mias.11 NOX1, NOX2, or NOX4 lies upstream of DHFR(dihydrofolate reductase) deficiency to induce abdominalaortic aneurysm formation.21,23 NOX4-induced ROS pro-duction mediates development of thoracic aortic aneu-rysm and abdominal aortic aneurysm in Marfan syndromemice.22 NOX4 activation, subsequent ROS production, andredox-sensitive CaMKII (calcium/calmodulin-dependentprotein kinase II) activation underlie the pathogenesisof cardiac arrhythmia.30 Either NOX1 or NOX2, or both,has been shown to be activated in the processes of ath-erosclerosis,38 while other studies implicated a protectiverole of NOX4 in atherosclerosis utilizing different modelsystems.39–41 As the only membrane-bound subunit forNOX, p22phox is required for catalytic activation of NOX1to NOX4,10,11 and excessive vascular ROS producedby p22phox plays a causal role in obesity by promotinginflammation, adipogenesis, and exercise intolerance.8However, the differential roles and related molecularmechanisms of NOX isoforms in the modulation of obe-sity and metabolic syndrome had remained completelyunknown until our present study. High-fat-diet–inducedbody weight gain, fat mass gain, liver weight gain, preadi-pocyte differentiation, glucose intolerance, insulin/leptinresistance, and hypercholesterolemia were all substan-tially alleviated in NOX1 global knockout mice but unaf-fected in NOX2 or NOX4 global knockout mice. Of note,this protective effect was not derived from changes inenergy intake, as WT and NOX1 global knockout micehad similar intake of food and water. On the other hand,high-fat feeding–induced impairment in spontaneousactivity, exercise capacity, respiratory performance, andskeletal muscle mitochondrial function were all mark-edly abrogated in NOX1 global knockout mice to resultin improved ATP production and energy expenditure. Inaddition, endothelium-dependent vasorelaxation was alsofully restored in global and endothelial-specific NOX1knockout mice, confirming improvement in endothelialfunction as the initiating factor of the protective signaling.Taken together with our earlier observations that vascularp22phox-dependent NOX drives development of obesityand metabolic syndrome, our data at first verified thatvascular NOX1 rather than NOX2 or NOX4 is the sourceof vascular oxidative stress that is acting as a driving forceto induce phenotypes of obesity and metabolic syndromeunder high-fat feeding. These findings further establishthe paradigm-shifting concept that vascular-derived oxi-dative stress from activated NOX1 is a cause of obesityand metabolic syndrome rather than a consequence. Oxidative stress has been shown to impair skel-etal muscle energy metabolism in mice and humanpatients.35,42–46 Angiotensin II–induced oxidative stresscauses deficiency in mitochondrial respiration in skeletalmuscle of C57BL/6 mice.44 Systemic oxidative stresshas been found to be associated with impaired skeletalmuscle energy metabolism in patients with metabolicsyndrome or heart failure.42,43 Our previous work alsorevealed skeletal muscle mitochondrial dysfunction inmice with vascular ROS overproduction.8 However, theexact source of ROS and how it impacts skeletal musclefunction have remained unclear until our present study. Inthis work, we elucidated that vascular endothelial NOX1,rather than smooth muscle cell NOX1, produced ROS todamage skeletal muscle mitochondrial function, result-ing in impaired spontaneous activity and exercise capac-ity under high-fat diet. Consistently, endothelial-specificoverexpression of NOX1 in mice led to deterioratedskeletal muscle mitochondrial dysfunction in response tohigh-fat feeding. Besides, mitochondrial ROS productionwas decreased by endothelial-specific NOX1 knockoutwhile increased by endothelial-specific NOX1 knockin. Ofnote, blood vessels are closely intertwined with skeletalmuscle tissues lying between the fascicles and bundlesof muscle fibers. Each muscle is supplied by many capil-laries, which contain lots of ECs rather than smooth mus-cle cells. The increase in bulk muscle blood flow observedduring exercise is accompanied by a substantial elevationin capillary recruitment within the muscle that expands thesurface area.47 This close association reduces the diffu-sion distances, allowing for efficient exchange of oxygen ### Page 20 and nutrients required for contraction and rapid removalof injurious metabolic wastes. Therefore, vascular ROS iscapable of diffusing to adjacent skeletal muscle cells toinduce mitochondrial dysfunction and mitochondrial ROSproduction in these cells. Exercise intolerance and mitochondrial complex Iand II deficiencies have been reported in high-fat–fedC57/BL6 mice.35 Alongside these changes, the respira-tory control ratio (RCR), as the single most useful gen-eral measure of function in isolated mitochondria,48 wasimpaired in high-fat–fed WT mice in our present study butsubstantially restored by global and endothelial-specificknockout of NOX1. Excessive mitochondrial swelling,representative of impaired mitochondrial integrity thatoccurs following increased opening of mitochondrialpermeability transition pores, is a central player that caninduce cell death through apoptosis or necrosis, depend-ing on the availability of ATP.49 Mitochondrial swelling hasbeen reported in skeletal muscle of mice fed a high-fatdiet due to increased ROS production in the muscle.50In this study, we found markedly increased mitochondrialswelling by high-fat feeding in WT mice, while global andendothelial-specific deletion of NOX1 completely attenu-ated this process. However, endothelial-specific NOX1knockin further increased mitochondrial swelling underhigh-fat feeding. We also found that high-fat feeding–induced marked increase in mitochondrial ROS produc-tion in skeletal muscle was substantially alleviated byglobal or endothelial-specific knockout of NOX1. By con-trast, NOX1-specific overexpression in ECs exaggeratedmitochondrial ROS production in skeletal muscle. Takentogether, these data indicate that endothelial-specificNOX1 activation induces skeletal muscle mitochondrialdysfunction that is characterized by reduced respira-tory performance, increased mitochondrial swelling,and increased mitochondrial ROS production. The mito-chondrial electron transport chain is located withinmitochondrial cristae where it generates the electrochem-ical potential across the invaginated inner mitochondrialmembranes, which is used for ATP synthesis.51 Cristaeorientation, curvature, and density per surface area arenonrandom, exhibiting directionality and orientation thatmay enable cristae membranes of adjacent mitochondriato become physically and energetically coordinated.51In this work, we revealed that disorganized mitochon-drial cristae in WT mice fed a high-fat diet, which wascompletely abrogated by global or endothelial-specificdeletion of NOX1. Of note, we found much lower den-sity and more disrupted structure of mitochondrial cris-tae in endothelial-specific NOX1 knockin mice. Theseresults showed that endothelial-specific NOX1 activationdestroys mitochondrial cristae structure and ATP produc-tion. Therefore, endothelial-specific activation of NOX1impairs skeletal muscle mitochondrial function through adiversity of processes, resulting in impairment in sponta-neous activity and exercise capacity to suppress energyexpenditure. This conclusion is supported by data fromboth endothelial-specific knockout and knockin mice. Global NOX1 knockout mice have been previouslystudied for their roles in diabetic vascular dysfunctionand atherogenesis, as reported by others and our ownwork.8,24,52,53 Besides being implicated to mediate dia-betic endothelial dysfunction and accelerated atherogen-esis,8,24,52–54 it was reported that NOX1 is involved in thedysfunction of liver sinusoids in nonalcoholic fatty liverdisease.55 These findings seem to share similarity with ourprevious work establishing a critical role of p22phox in medi-ating obesity and metabolic syndrome as a catalytic sub-unit capable of activating NOX1,8 which was highlightedby an editorial commentary emphasizing a paradigm-shifting novel concept of vascular ROS driving systematicresponses of obesity and metabolic syndrome.56 It wouldalso be critical to further examine detailed mechanisticpathways by which endothelial NOX1 is triggered to initi-ate the processes of obesity and metabolic dysfunction. Importantly, the only 2 novel genes of Cntnap4and Col9a1 were specifically screened out via RNA-sequencing and IPA analysis in this study, which areregulated in the skeletal muscle to be responsible forthe changes in function induced by high-fat feeding thatwas reversed by endothelial-specific deletion of NOX1.Existing studies indicate that Cntnap4 is a member ofthe neurexin superfamily of transmembrane moleculesthat have critical functions in neuronal cell communica-tion.57 It was shown that Cntnap4 plays a critical role inosteogenesis, which is the first functional annotation forCntnap4 in the musculoskeletal system.57 Interestingly,recent studies indicate that microRNA-34 uniquely reg-ulates Cntnap4 expression, which may be regulated inobesity to contribute to Cntnap4-dependent modulationof skeletal muscle function.58 The Col9a1 gene has beenimplicated in the initial stages of skeletal muscle cell dif-ferentiation and found associated with the longissimusdorsi muscle in the pig.59 Some of the genetic variantsof Col9a1 have been identified.60 It would be interestingto explore whether these variants are present in humanpatients with obesity. In this research, we found thatCntnap4 was upregulated, while Col9a1 was downreg-ulated in the skeletal muscle of high-fat–fed WT mice,and endothelial-specific NOX1 ablation reversed theseresponses, indicating that the regulation of these novelgenes likely underlie high-fat feeding–induced skeletalmuscle phenotypes that are correctable by deletion ofendothelial NOX1. Detailed molecular pathways control-ling the regulation of Cntnap4 and Col9a1 in skeletalmuscle warrant further investigations. Importantly, utilizing coronary arteries isolated fromhuman patients with obesity and control subjects, wefound that NOX1 protein expression was markedlyincreased in obese subjects (BMI >25) compared withnormal weight controls (BMI <25). This indicates thatthe sophisticated mechanistic insights we identified for ### Page 21 vascular NOX1-driven development of obesity and met-abolic syndrome, are relevant to human subjects withgreat translational potential. Of note, this regulation weobserved is consistent with previous work by Matsumotoat al,55 reporting upregulated NOX1 mRNA expression inhuman nonalcoholic steatohepatitis patients. Of note, our studies were performed primarily usingmale mice. It would be important to also examine thenew pathways and mechanisms in female mice. It wasshown that female mitochondria have higher functionalcapacity and greater resistance to oxidative damage thanmale mitochondria.61 Whereas mitochondria might be lessinjured in obese females, preservation of mitochondrialfunction might have more sensitive outcomes in reducingobesity and metabolic syndrome. It was proposed that sexsteroid hormones and gut microbiota diversity, chromo-somal and genetic variables, and behavioral and sociocul-tural variables all influence obesity development in menand women. It is, therefore, a complex question to addresshow the endothelial-specific NOX1 activation drives obe-sity in females differentially, which warrants immediatefurther studies.62 Many novel mediators in adipose tissueshave been recently revealed,63,64 and it is also important tounderstand how they differ between males and femalesin regulating adipogenesis and obesity. In conclusion, observations in our present study firmlyestablish a direct causal role of endothelial-specific activa-tion of NOX1 in driving the development of systematic dis-orders of obesity and metabolic syndrome. ROS producedfrom activated NOX1 in the ECs diffuses to adjacent skel-etal muscle to induce mitochondrial dysfunction, resultingin impaired spontaneous activity and exercise capacity,and ultimately a severe deficiency in energy expenditureto allow development of obesity and metabolic syndrome.Novel genes of Cntnap4 and Col9a1 in skeletal musclewere identified being relevant to these processes. NOX1protein expression was indeed upregulated in the coro-nary arteries of human patients with obesity. Therefore,specific and selective targeting of endothelial NOX1, orCntmap4/Col9a1 in the skeletal muscle, may prove to beinnovatively and robustly effective for the treatment andprevention of obesity and metabolic syndrome. ## ARTICLE INFORMATION Received May 09, 2025; revision received March 25, 2026; accepted March 30,2026. ### Affiliations Division of Molecular Medicine, Department of Anesthesiology and PerioperativeMedicine, Division of Cardiology, Department of Medicine, David Geffen Schoolof Medicine, University of California, Los Angeles (K.H., Y.H., Yuhan Zhang, YixuanZhang, N.W.H., H.C.). Department of Anesthesiology, Medical College of Wiscon-sin, Milwaukee (J.K.F.). ### Sources of Funding This work was supported by the National Institutes of Health, National Heart,Lung, and Blood Institute awards HL142951 (to H. Cai), HL154754 (to H. Cai),and HL162407 (H. Cai). ### Disclosures None. ### Supplemental Material Supplemental Materials and Methods Figures S1–S6 Major Resources Table Uncropped Western Blots References 8, 27, 31, 34 ## REFERENCES 1. Powell-Wiley TM, Poirier P, Burke LE, Despres JP, Gordon-Larsen P, Lavie CJ,Lear SA, Ndumele CE, Neeland IJ, Sanders P, et al; American Heart Asso-ciation Council on Lifestyle and Cardiometabolic Health; Council on Car-diovascular and Stroke Nursing; Council on Clinical Cardiology; Council onEpidemiology and Prevention; and Stroke Council. Obesity and cardiovascu-lar disease: a scientific statement from the American Heart Association. Cir-culation. 2021;143:e984–e1010. doi: 10.1161/CIR.0000000000000973 2. Maffetone PB, Rivera-Dominguez I, Laursen PB. Overfat and underfat: newterms and definitions long overdue. Front Public Health. 2016;4:279. doi:10.3389/fpubh.2016.00279 3. Srivastava G, Apovian CM. Current pharmacotherapy for obesity. Nat RevEndocrinol. 2018;14:12–24. doi: 10.1038/nrendo.2017.122 4. Blüher M. Obesity: global epidemiology and pathogenesis. Nat Rev Endocri-nol. 2019;15(5):288–298. doi: 10.1038/s41574-019-0176-8 5. The Lancet Gastroenterology Hepatology. Obesity: another ongo-ing pandemic. Lancet Gastroenterol Hepatol. 2021;6:411. doi:10.1016/S2468-1253(21)00143-6 6. Cardel MI, Atkinson MA, Taveras EM, Holm JC, Kelly AS. Obesity treatmentamong adolescents: a review of current evidence and future directions. JAMAPediatr. 2020;174:609–617. doi: 10.1001/jamapediatrics.2020.0085 7. Cignarella A, Busetto L, Vettor R. Pharmacotherapy of obesity: an update.Pharmacol Res. 2021;169:105649. doi: 10.1016/j.phrs.2021.105649 8. Youn JY, Siu KL, Lob HE, Itani H, Harrison DG, Cai H. Role of vascular oxi-dative stress in obesity and metabolic syndrome. Diabetes. 2014;63:2344–2355. doi: 10.2337/db13-0719 9. Forrester SJ, Kikuchi DS, Hernandes MS, Xu Q, Griendling KK. Reac-tive oxygen species in metabolic and inflammatory signaling. Circ Res.2018;122:877–902. doi: 10.1161/CIRCRESAHA.117.311401 10. Santillo M, Colantuoni A, Mondola P, Guida B, Damiano S. NOX signalingin molecular cardiovascular mechanisms involved in the blood pressurehomeostasis. Front Physiol. 2015;6:194. doi: 10.3389/fphys.2015.00194 11. Zhang Y, Murugesan P, Huang K, Cai H. NADPH oxidases and oxidasecrosstalk in cardiovascular diseases: novel therapeutic targets. Nat Rev Car-diol. 2020;17:170–194. doi: 10.1038/s41569-019-0260-8 12. Brandes RP, Schroder K. Differential vascular functions of Nox fam-ily NADPH oxidases. Curr Opin Lipidol. 2008;19:513–518. doi:10.1097/MOL.0b013e32830c91e3 13. Ray R, Shah AM. NADPH oxidase and endothelial cell function. Clin Sci(Lond). 2005;109:217–226. doi: 10.1042/CS20050067 14. Cai H, Harrison DG. Endothelial dysfunction in cardiovascular dis-eases: the role of oxidant stress. Circ Res. 2000;87:840–844. doi:10.1161/01.res.87.10.840 15. Cai H, Griendling KK, Harrison DG. The vascular NAD(P)H oxidases astherapeutic targets in cardiovascular diseases. Trends Pharmacol Sci.2003;24:471–478. doi: 10.1016/S0165-6147(03)00233-5 16. Cai H. Hydrogen peroxide regulation of endothelial function: origins,mechanisms, and consequences. Cardiovasc Res. 2005;68:26–36. doi:10.1016/j.cardiores.2005.06.021 17. Cai H. NAD(P)H oxidase-dependent self-propagation of hydro-gen peroxide and vascular disease. Circ Res. 2005;96:818–822. doi:10.1161/01.RES.0000163631.07205.fb 18. Lassegue B, San Martin A, Griendling KK. Biochemistry, physiology, andpathophysiology of NADPH oxidases in the cardiovascular system. Circ Res.2012;110:1364–1390. doi: 10.1161/CIRCRESAHA.111.243972 19. Matsuno K, Yamada H, Iwata K, Jin D, Katsuyama M, Matsuki M, Takai S,Yamanishi K, Miyazaki M, Matsubara H, et al. Nox1 is involved in angio-tensin II-mediated hypertension: a study in Nox1-deficient mice. Circulation.2005;112:2677–2685. doi: 10.1161/CIRCULATIONAHA.105.573709 20. Bendall JK, Rinze R, Adlam D, Tatham AL, de Bono J, Wilson N,Volpi E, Channon KM. Endothelial Nox2 overexpression potentiates vas-cular oxidative stress and hemodynamic response to angiotensin II: studies ### Page 22 in endothelial-targeted Nox2 transgenic mice. Circ Res. 2007;100:1016–1025. doi: 10.1161/01.RES.0000263381.83835.7b 21. Siu KL, Li Q, Zhang Y, Guo J, Youn JY, Du J, Cai H. NOX isoforms in thedevelopment of abdominal aortic aneurysm. Redox Biol. 2017;11:118–125.doi: 10.1016/j.redox.2016.11.002 22. Huang K, Wang Y, Siu KL, Zhang Y, Cai H. Targeting feed-forward sig-naling of TGFβ/NOX4/DHFR/eNOS uncoupling/TGFβ axis withanti-TGFβ and folic acid attenuates formation of aortic aneurysms:novel mechanisms and therapeutics. Redox Biol. 2021;38:101757. doi:10.1016/j.redox.2020.101757 23. Huang K, Narumi T, Zhang Y, Li Q, Murugesan P, Wu Y, Liu NM, Cai H.Targeting MicroRNA-192-5p, a downstream effector of NOXs (NADPHoxidases), reverses endothelial DHFR (dihydrofolate reductase) defi-ciency to attenuate abdominal aortic aneurysm formation. Hypertension.2021;78:282–293. doi: 10.1161/HYPERTENSIONAHA.120.15070 24. Youn JY, Gao L, Cai H. The p47phox- and NADPH oxidase organiser 1(NOXO1)-dependent activation of NADPH oxidase 1 (NOX1) mediatesendothelial nitric oxide synthase (eNOS) uncoupling and endothelial dys-function in a streptozotocin-induced murine model of diabetes. Diabetologia.2012;55:2069–2079. doi: 10.1007/s00125-012-2557-6 25. Oak JH, Cai H. Attenuation of angiotensin II signaling recouples eNOSand inhibits nonendothelial NOX activity in diabetic mice. Diabetes.2007;56:118–126. doi: 10.2337/db06-0288 26. Youn JY, Zhou J, Cai H. Bone Morphogenic protein 4 mediates NOX1-dependent eNOS uncoupling, endothelial dysfunction, and COX2 induc-tion in type 2 diabetes mellitus. Mol Endocrinol. 2015;29:1123–1133. doi:10.1210/ME.2014-1313 27. Siu KL, Lotz C, Ping P, Cai H. Netrin-1 abrogates ischemia/reperfusion-induced cardiac mitochondrial dysfunction via nitric oxide-dependentattenuation of NOX4 activation and recoupling of NOS. J Mol Cell Cardiol.2015;78:174–185. doi: 10.1016/j.yjmcc.2014.07.005 28. Matsushima S, Kuroda J, Ago T, Zhai P, Ikeda Y, Oka S,Fong GH, Tian R, Sadoshima J. Broad suppression of NADPH oxidaseactivity exacerbates ischemia/reperfusion injury through inadvertent down-regulation of hypoxia-inducible factor-1α and upregulation of peroxisomeproliferator-activated receptor-α. Circ Res. 2013;112:1135–1149. doi:10.1161/CIRCRESAHA.111.300171 29. Zhang J, Youn JY, Kim AY, Ramirez RJ, Gao L, Ngo D, Chen P, Scovotti J,Mahajan A, Cai H. NOX4-dependent hydrogen peroxide overproduction inhuman atrial fibrillation and HL-1 atrial cells: relationship to hypertension.Front Physiol. 2012;3:140. doi: 10.3389/fphys.2012.00140 30. Zhang Y, Shimizu H, Siu KL, Mahajan A, Chen JN, Cai H. NADPH oxi-dase 4 induces cardiac arrhythmic phenotype in zebrafish. J Biol Chem.2014;289:23200–23208. doi: 10.1074/jbc.M114.587196 31. Wang Y, Nakayama M, Pitulescu ME, Schmidt TS, Bochenek ML, Sakakibara A,Adams S, Davy A, Deutsch U, Luthi U, et al. Ephrin-B2 controls VEGF-induced angiogenesis and lymphangiogenesis. Nature. 2010;465:483–486. doi: 10.1038/nature09002 32. Monvoisin A, Alva JA, Hofmann JJ, Zovein AC, Lane TF,Iruela-Arispe ML. VE-cadherin-CreERT2 transgenic mouse: a model forinducible recombination in the endothelium. Dev Dyn. 2006;235:3413–3422. doi: 10.1002/dvdy.20982 33. Wirth A, Benyo Z, Lukasova M, Leutgeb B, Wettschureck N, Gorbey S,Orsy P, Horvath B, Maser-Gluth C, Greiner E, et al. G12-G13-LARG-mediated signaling in vascular smooth muscle is required for salt-inducedhypertension. Nat Med. 2008;14:64–68. doi: 10.1038/nm1666 34. Guo Z, Zhang Y, Liu C, Youn JY, Cai HL. Toll-like receptor 2 (TLR2) knock-out abrogates diabetic and obese phenotypes while restoring endothe-lial function via inhibition of NOX1. Diabetes. 2021;70:2107–2119. doi:10.2337/db20-0591 35. Yokota T, Kinugawa S, Hirabayashi K, Matsushima S, Inoue N, Ohta Y,Hamaguchi S, Sobirin MA, Ono T, Suga T, et al. Oxidative stress in skeletalmuscle impairs mitochondrial respiration and limits exercise capacity in type2 diabetic mice. Am J Physiol Heart Circ Physiol. 2009;297:H1069–H1077.doi: 10.1152/ajpheart.00267.2009 36. Cogliati S, Enriquez JA, Scorrano L. Mitochondrial cristae: where beautymeets functionality. Trends Biochem Sci. 2016;41:261–273. doi:10.1016/j.tibs.2016.01.001 37. Youn JY, Wang T, Blair J, Laude KM, Oak JH, McCann LA, Harrison DG,Cai H. Endothelium-specific sepiapterin reductase deficiency in DOCA-salthypertension. Am J Physiol Heart Circ Physiol. 2012;302:H2243–H2249.doi: 10.1152/ajpheart.00835.2011 38. Barry-Lane PA, Patterson C, van der Merwe M, Hu Z, Holland SM, Yeh ET,Runge MS. p47phox is required for atherosclerotic lesion progression inApoE(-/-) mice. J Clin Invest. 2001;108:1513–1522. doi: 10.1172/JCI11927 39. Schurmann C, Rezende F, Kruse C, Yasar Y, Lowe O, Fork C,van de Sluis B, Bremer R, Weissmann N, Shah AM, et al. The NADPH oxi-dase Nox4 has anti-atherosclerotic functions. Eur Heart J. 2015;36:3447–3456. doi: 10.1093/eurheartj/ehv460 40. Langbein H, Brunssen C, Hofmann A, Cimalla P, Brux M,Bornstein SR, Deussen A, Koch E, Morawietz H. NADPH oxidase 4 pro-tects against development of endothelial dysfunction and atherosclerosisin LDL receptor deficient mice. Eur Heart J. 2016;37:1753–1761. doi:10.1093/eurheartj/ehv564 41. Gray SP, Di Marco E, Kennedy K, Chew P, Okabe J, El-Osta A, Calkin AC,Biessen EA, Touyz RM, Cooper ME, et al. Reactive oxygen species can pro-vide atheroprotection via NOX4-dependent inhibition of inflammation andvascular remodeling. Arterioscler Thromb Vasc Biol. 2016;36:295–307. doi:10.1161/ATVBAHA.115.307012 42. Yokota T, Kinugawa S, Yamato M, Hirabayashi K, Suga T, Takada S,Harada K, Morita N, Oyama-Manabe N, Kikuchi Y, et al. Systemic oxida-tive stress is associated with lower aerobic capacity and impaired skeletalmuscle energy metabolism in patients with metabolic syndrome. DiabetesCare. 2013;36:1341–1346. doi: 10.2337/dc12-1161 43. Yokota T, Kinugawa S, Hirabayashi K, Yamato M, Takada S, Suga T,Nakano I, Fukushima A, Matsushima S, Okita K, et al. Systemic oxidativestress is associated with lower aerobic capacity and impaired skeletal mus-cle energy metabolism in heart failure patients. Sci Rep. 2021;11:2272. doi:10.1038/s41598-021-81736-0 44. Inoue N, Kinugawa S, Suga T, Yokota T, Hirabayashi K, Kuroda S, Okita K,Tsutsui H. Angiotensin II-induced reduction in exercise capacity is associatedwith increased oxidative stress in skeletal muscle. Am J Physiol Heart CircPhysiol. 2012;302:H1202–H1210. doi: 10.1152/ajpheart.00534.2011 45. Kinugawa S, Wang Z, Kaminski PM, Wolin MS, Edwards JG,Kaley G, Hintze TH. Limited exercise capacity in heterozygous man-ganese superoxide dismutase gene-knockout mice: roles of super-oxide anion and nitric oxide. Circulation. 2005;111:1480–1486. doi:10.1161/01.CIR.0000159261.11520.63 46. Kuwahara H, Horie T, Ishikawa S, Tsuda C, Kawakami S, Noda Y, Kaneko T,Tahara S, Tachibana T, Okabe M, et al. Oxidative stress in skeletal muscle causessevere disturbance of exercise activity without muscle atrophy. Free Radic BiolMed. 2010;48:1252–1262. doi: 10.1016/j.freeradbiomed.2010.02.011 47. Sylow L, Kleinert M, Richter EA, Jensen TE. Exercise-stimulated glucoseuptake - regulation and implications for glycaemic control. Nat Rev Endocri-nol. 2017;13:133–148. doi: 10.1038/nrendo.2016.162 48. Brand MD, Nicholls DG. Assessing mitochondrial dysfunction in cells. Bio-chem J. 2011;435:297–312. doi: 10.1042/BJ20110162 49. Javadov S, Chapa-Dubocq X, Makarov V. Different approaches to model-ing analysis of mitochondrial swelling. Mitochondrion. 2018;38:58–70. doi:10.1016/j.mito.2017.08.004 50. Bonnard C, Durand A, Peyrol S, Chanseaume E, Chauvin MA, Morio B,Vidal H, Rieusset J. Mitochondrial dysfunction results from oxidative stressin the skeletal muscle of diet-induced insulin-resistant mice. J Clin Invest.2008;118:789–800. doi: 10.1172/JCI32601 51. Picard M, McManus MJ, Csordas G, Varnai P, Dorn GW 2nd,Williams D, Hajnoczky G, Wallace DC. Trans-mitochondrial coordination ofcristae at regulated membrane junctions. Nat Commun. 2015;6:6259. doi:10.1038/ncomms7259 52. Sobey CG, Judkins CP, Rivera J, Lewis CV, Diep H, Lee HW,Kemp-Harper BK, Broughton BR, Selemidis S, Gaspari TA, et al. NOX1deficiency in apolipoprotein E-knockout mice is associated with elevatedplasma lipids and enhanced atherosclerosis. Free Radic Res. 2015;49:186–198. doi: 10.3109/10715762.2014.992893 53. Zhang Y, Youn JY, Huang K, Zhang Y, Cai H. Alleviation of accelerateddiabetic atherogenesis in STZ-treated apoE/NOX1 DKO mice, apoE-/-/tg-EC-DHFR mice, and by folic acid. Redox Biol. 2025;82:103570. doi:10.1016/j.redox.2025.103570 54. Sheehan AL, Carrell S, Johnson B, Stanic B, Banfi B, Miller FJ Jr. Role forNox1 NADPH oxidase in atherosclerosis. Atherosclerosis. 2011;216:321–326. doi: 10.1016/j.atherosclerosis.2011.02.028 55. Matsumoto M, Zhang J, Zhang X, Liu J, Jiang JX,Yamaguchi K, Taruno A, Katsuyama M, Iwata K, Ibi M, et al. The NOX1 iso-form of NADPH oxidase is involved in dysfunction of liver sinusoids in non-alcoholic fatty liver disease. Free Radic Biol Med. 2018;115:412–420. doi:10.1016/j.freeradbiomed.2017.12.019 56. Aroor AR, DeMarco VG. Oxidative stress and obesity: the chicken or theegg? Diabetes. 2014;63:2216–2218. doi: 10.2337/db14-0424 57. Li C, Zheng Z, Ha P, Chen X, Jiang W, Sun S, Chen F, Asatrian G,Berthiaume EA, Kim JK, et al. Neurexin superfamily cell membrane receptorcontactin-associated protein like-4 (Cntnap4) is involved in neural EGFL-like ### Page 23 1 (Nell-1)-responsive osteogenesis. J Bone Miner Res. 2018;33:1813–1825. doi: 10.1002/jbmr.3524 58. McNeill EM, Warinner C, Alkins S, Taylor A, Heggeness H, DeLuca TF,Fulga TA, Wall DP, Griffith LC, Van Vactor D. The conserved microRNAmiR-34 regulates synaptogenesis via coordination of distinct mechanismsin presynaptic and postsynaptic cells. Nat Commun. 2020;11:1092. doi:10.1038/s41467-020-14761-8 59. Fan B, Onteru SK, Nikkila MT, Stalder KJ, Rothschild MF. The COL9A1 geneis associated with longissimus dorsi muscle area in the pig. Anim Genet.2009;40:788. doi: 10.1111/j.1365-2052.2009.01885.x 60. Hofrichter MAH, Doll J, Habibi H, Enayati S, Vahidi Mehrjardi MY, Müller T,Dittrich M, Haaf T, Vona B. Exome-wide copy number variation analysis iden-tifies a COL9A1 in frame deletion that is associated with hearing loss. EurJ Med Genet. 2019;62:103724. doi: 10.1016/j.ejmg.2019.103724 61. Mauvais-Jarvis F. Sex differences in energy metabolism: natural selection,mechanisms and consequences. Nat Rev Nephrol. 2024;20:56–69. doi:10.1038/s41581-023-00781-2 62. Koceva A, Herman R, Janez A, Rakusa M, Jensterle M. Sex- and gender-related differences in obesity: from pathophysiological mechanisms to clini-cal implications. Int J Mol Sci. 2024;25:7342. doi: 10.3390/ijms25137342 63. Cheng Y, Yuan Q, Vergnes L, Rong X, Youn JY, Li J, Yu Y,Liu W, Cai H, Lin JD, et al. KDM4B protects against obesity and meta-bolic dysfunction. Proc Natl Acad Sci USA. 2018;115:E5566–E5575. doi:10.1073/pnas.1721814115 64. Félix-Soriano E, Sáinz N, Gil-Iturbe E, Collantes M, Fernández-Galilea M,Castilla-Madrigal R, Ly L, Dalli J, Moreno-Aliaga MJ. Changes in brown adiposetissue lipid mediator signatures with aging, obesity, and DHA supplementationin female mice. FASEB J. 2021;35:e21592. doi: 10.1096/fj.202002531R