Advertisement

Cytokine signature of inflammation mediated by autoreactive Th-cells, in calf muscle of claudicating patients with Fontaine stage II peripheral artery disease

Published:August 21, 2020DOI:https://doi.org/10.1016/j.trsl.2020.08.008
      Peripheral artery disease (PAD), a severe atherosclerotic condition primarily of the elderly, afflicts 200 million individuals, worldwide, and is associated with lower extremity myopathy. Circulating markers of inflammation have been linked to risk and severity of PAD but the contribution of local inflammation to myopathy remains unknown. We evaluated, by ELISA, calf muscle of PAD patients (N = 23) and control subjects (N = 18) for local expression of inflammatory cytokines including Granulocyte/Monocyte Colony-Stimulating Factor (GM-CSF), Interleukin 17A (IL-17A), Interferon ϒ (IFN-ϒ), tumor necrosis factor α (TNF-α), and Interleukin 6 (IL-6). One or more of these cytokines were expressed in nineteen patients and 2 controls and coordinated expression of GM-CSF, IL-17A, IFN-ϒ, and TNF-α, a signature of activated, MHC Class II dependent autoreactive Th-cells, was unique to 11 patients. GM-CSF is the central driver of tissue-damaging myeloid macrophages. Patients with this cytokine signature had a shorter (P= 0.017) Claudication Onset Distance (17 m) compared with patients lacking the signature (102 m). Transforming Growth Factor β1 (TGFβ1) and Chemokine Ligand 5 (CCL5) were expressed coordinately in all PAD and control muscles, independently of GM-CSF, IL-17A, IFN-ϒ, TNF-α, or IL-6. TGFβ1 and CCL5 and their gene transcripts were increased in PAD muscle, consistent with increased age-associated inflammation in these patients. Serum cytokines were not informative of muscle cytokine expression. We have identified a cytokine profile of autoimmune inflammation in calf muscles of a significant proportion of claudicating PAD patients, in association with decreased limb function, and a second independent profile consistent with increased “inflammaging” in all PAD patients.

      Abbreviations:

      COD (Claudication Onset Distance), dNTPs (Deoxynucleotide Triphosphates), DTT (Dithiothreitol), PWD (Peak Walking Distance), QoL (Quality of Life), TRIS (Tris(hydroxymethyl)aminomethane), vSMC (Vascular Smooth Muscle Cells), WIQ (Walking Impairment Questionnaire)
      AT A GLANCE COMMENTARY
      George P. Casale, et al

      Background

      Circulating markers of inflammation have been identified as predictors of risk and severity of peripheral artery disease (PAD). However, the contribution of local inflammation to lower extremity myopathy remains unknown.

      Translational Significance

      We identified a cytokine signature of MHC Class II dependent autoimmune inflammation in calf muscles of half of claudicating PAD patients, associated with decreased limb function. We found, in addition, a second and independent cytokine profile consistent with increased age-associated inflammation in all PAD patients. Our findings raise the possibility of targeted intervention with clinically effective anti-inflammatory agents.

      INTRODUCTION

      Peripheral artery disease and calf muscle damage

      Peripheral Artery Disease (PAD) is a severe atherosclerotic condition primarily of the elderly, with a prevalence of more than 200 million individuals, worldwide. The disease is diagnosed by an Ankle-Brachial Index (ABI) of less than 0.90 with limitation of lower leg function and carries an increased risk of coronary artery disease (CAD), carotid artery disease, and cardiovascular mortality.
      • Brevetti G.
      • Giugliano G.
      • Brevetti L.
      • Hiatt W.R.
      Inflammation in peripheral artery disease.
      During exercise, atherosclerotic plaques in the arteries supplying the lower legs, restrict blood flow causing a deficit in oxygen and nutrient delivery in relation to demand.
      • Hiatt W.R.
      • Armstrong E.J.
      • Larson C.J.
      • Brass E.P.
      Pathogenesis of the limb manifestations and exercise limitations in peripheral artery disease.
      Chronic intermittent ischemia of the lower leg is associated with pathology of the lower leg muscles. Ischemia, per se, and myopathy primarily of the calf muscle, produce limb dysfunction,
      • Koutakis P.
      • Miserlis D.
      • Myers S.A.
      • et al.
      Abnormal accumulation of desmin in gastrocnemius myofibers of patients with peripheral artery disease: associations with altered myofiber morphology and density, mitochondrial dysfunction and impaired limb function.
      • Koutakis P.
      • Myers S.A.
      • Cluff K.
      • et al.
      Abnormal myofiber morphology and limb dysfunction in claudication.
      • Schieber M.N.
      • Hasenkamp R.M.
      • Pipinos I.I.
      • et al.
      Muscle strength and control characteristics are altered by peripheral artery disease.
      and limit physical activity. Ischemia is regarded as the ultimate cause of pathological changes to skeletal muscle myofibers, nerves, and blood vessels of the lower legs of PAD patients.
      • Pipinos II
      • Judge A.R.
      • Selsby J.T.
      • et al.
      The myopathy of peripheral arterial occlusive disease: part 1. Functional and histomorphological changes and evidence for mitochondrial dysfunction.
      Reactive oxygen species (ROS) generated in response to ischemia produce oxidative damage to the myofibers, which increases with advancing disease and correlates with reduced myofiber cross-sectional area, reduced ABI,
      • Weiss D.J.
      • Casale G.P.
      • Koutakis P.
      • et al.
      Oxidative damage and myofiber degeneration in the gastrocnemius of patients with peripheral arterial disease.
      and selective loss of Type II myofibers.
      • Koutakis P.
      • Weiss D.J.
      • Miserlis D.
      • et al.
      Oxidative damage in the gastrocnemius of patients with peripheral artery disease is myofiber type selective.
      In addition to impairing cellular function by damaging macromolecules, ROS may induce an inflammatory response that has myopathic effects.
      • Brevetti G.
      • Giugliano G.
      • Brevetti L.
      • Hiatt W.R.
      Inflammation in peripheral artery disease.
      ,
      • Gillani S.
      • Cao J.
      • Suzuki T.
      • Hak D.J.
      The effect of ischemia reperfusion injury on skeletal muscle.
      Reduced calf muscle cross-sectional area and increased calf muscle fat content have been associated with increasing quartile of serum/plasma C-reactive protein (CRP).
      • McDermott M.M.
      • Ferrucci L.
      • Guralnik J.M.
      • et al.
      Elevated levels of inflammation, d-dimer, and homocysteine are associated with adverse calf muscle characteristics and reduced calf strength in peripheral arterial disease.

      Inflammatory biomarkers and severity of atherosclerosis in patients With PAD

      The severity of atherosclerosis in PAD patients is represented by (1) the higher rate of 1-year atherothrombotic events among PAD compared to CAD and cerebrovascular disease patients and (2) the high prevalence (up to 90%) of CAD among PAD patients compared to less than 25% PAD among CAD patients.
      • Brevetti G.
      • Giugliano G.
      • Brevetti L.
      • Hiatt W.R.
      Inflammation in peripheral artery disease.
      These observations together with the positive association of the prevalence of hypoechoic (inflamed) plaques in the carotid arteries with hypoechoic plaques in the femoral arterial bed, among PAD patients, suggested to some investigators that inflammation in the large atherosclerotic arterial beds of the affected PAD limbs is a principal contributor to systemic disease in PAD patients. Studies of inflammatory markers in the blood of PAD and CAD patients are consistent with this hypothesis. Relative to the aorta, blood from the femoral vein of the affected limb of PAD-plus-CAD patients contained increased numbers of activated neutrophils, in contrast to CAD-only patients who displayed no transfemoral increase.
      • Brevetti G.
      • Piscione F.
      • Schiano V.
      • Galasso G.
      • Scopacasa F.
      • Chiariello M.
      Concomitant coronary and peripheral arterial disease: relationship between the inflammatory status of the affected limb and the severity of coronary artery disease.
      PAD-plus-CAD patients whose transfemoral increases were greater than the median had a greater number of coronary stenoses, and an increased prevalence of 3-vessel CAD and myocardial infarct compared to PAD-plus-CAD patients whose transfemoral increases were less than the median. These findings are consistent with those of an earlier study
      • Violi F.
      • Criqui M.
      • Longoni A.
      • Castiglioni C.
      Relation between risk factors and cardiovascular complications in patients with peripheral vascular disease. Results from the A.D.E.P. study.
      that documented an increased incidence of cardiac, cerebral and peripheral events, combined, in PAD patients with elevated leukocyte count. Moreover, the plasma concentration of soluble vascular cell adhesion molecule-1 was found to be a strong and independent predictor of risk of cardiovascular events, in PAD patients with intermittent claudication.
      • Silvestro A.
      • Brevetti G.
      • Schiano V.
      • Scopacasa F.
      • Chiariello M.
      Adhesion molecules and cardiovascular risk in peripheral arterial disease. Soluble vascular cell adhesion molecule-1 improves risk stratification.

      Inflammatory biomarkers and risk of developing hemodynamic limitation and PAD

      In addition to their association with cardiovascular events, inflammatory markers have been linked to risk of PAD and hemodynamic limitation in the lower legs of PAD patients. CRP was found to be a strong and independent predictor of the development of PAD among healthy men who participated in the Physicians’ Health Study.
      • Ridker P.M.
      • Stampfer M.J.
      • Rifai N.
      Novel risk factors for systemic atherosclerosis: a comparison of C-reactive protein, fibrinogen, homocysteine, lipoprotein(a), and standard cholesterol screening as predictors of peripheral arterial disease.
      Baseline plasma CRP was higher (P < 0.006) among men who developed PAD (0.14 mg/dL) in the 9-year follow-up compared to men who did not develop PAD (0.10 mg/dL). The relative risk for developing PAD was 2.5 times higher among men in the highest quartile of baseline CRP compared to men in the lowest quartile. Within the Rotterdam Study of the incidence and risk factors of disabling diseases among men and women aged 55 years and older, plasma markers of inflammation were investigated as potential predictors of ABI.
      • van der Meer I.M.
      • de Maat M.P.M.
      • Bots M.L.
      • et al.
      Inflammatory mediators and cell adhesion molecules as indicators of severity of atherosclerosis.
      Both CRP and Interleukin-6 (IL-6) were positively associated with hemodynamic limitation in the lower leg as indicated by their correlation with decreasing ABI. As part of the Edinburgh Artery Study,
      • Tzoulaki I.
      • Murray G.D.
      • Lee A.J.
      • Rumley A.
      • Lowe G.D.
      • Fowkes F.G.
      C-reactive protein, interleukin-6, and soluble adhesion molecules as predictors of progressive peripheral atherosclerosis in the general population: Edinburgh Artery Study.
      CRP in addition to IL-6 and soluble Intercellular Adhesion Molecule-1 were shown to be independent predictors of the decline of ABI, over 12 years of follow-up. In the same study, soluble vascular cell adhesion molecule-1 and E-selectin were not associated with changes in ABI.

      Inflammatory biomarkers and limb dysfunction in PAD patients

      Inflammation has been linked to limb dysfunction in PAD patients. A study
      • Nylaende M.
      • Kroese A.
      • Stranden E.
      • et al.
      Markers of vascular inflammation are associated with the extent of atherosclerosis assessed as angiographic score and treadmill walking distances in patients with peripheral arterial occlusive disease.
      of 127 patients with PAD verified by arteriography, found that plasma IL-6 exhibited an inverse correlation with maximum treadmill walking distance. In contrast, tumor necrosis factor-α (TNF-α), monocyte chemoattractant protein-1, and CD40 ligand were not associated with treadmill walking distance. A more extensive study of the association of inflammation with limb function in PAD patients was completed by McDermott et al
      • McDermott M.M.
      • Greenland P.
      • Green D.
      • et al.
      D-dimer, inflammatory markers, and lower extremity functioning in patients with and without peripheral arterial disease.
      who determined the associations of serum CRP, fibrinogen, and amyloid A with 6-minute walking distance, 4-m walking velocity, and a summary performance score. CRP but not fibrinogen or amyloid A was associated with 6-minute walking distance and summary performance score. Higher CRP concentrations were associated independently with poorer 6-minute walking distance and lower summary performance score. In a prospective study of PAD patients,
      • McDermott M.M.
      • Liu K.
      • Guralnik J.M.
      • et al.
      Functional decline in patients with and without peripheral arterial disease: predictive value of annual changes in levels of C-reactive protein and D-dimer.
      greater annual increases in serum CRP were found to be associated with greater annual decline in 6-minute walking distance, over a period of 3 years

      Local inflammation in calf muscle of PAD patients

      The association of inflammatory markers in circulating blood with risk and severity of PAD is well established. This association, however, is based on circulating markers that likely originate from active atherosclerotic plaques in these patients and does not address local inflammation in calf muscle or other lower leg muscles. Myofiber pathology in the lower legs of PAD patients is characterized by mitochondrial dysfunction,
      • Pipinos II
      • Judge A.R.
      • Selsby J.T.
      • et al.
      The myopathy of peripheral arterial occlusive disease: part 1. Functional and histomorphological changes and evidence for mitochondrial dysfunction.
      ,
      • Brass E.P.
      • Hiatt W.R.
      Acquired skeletal muscle metabolic myopathy in atherosclerotic peripheral arterial disease.
      oxidative damage,
      • Weiss D.J.
      • Casale G.P.
      • Koutakis P.
      • et al.
      Oxidative damage and myofiber degeneration in the gastrocnemius of patients with peripheral arterial disease.
      ,
      • Koutakis P.
      • Weiss D.J.
      • Miserlis D.
      • et al.
      Oxidative damage in the gastrocnemius of patients with peripheral artery disease is myofiber type selective.
      intracellular accumulation of damaged proteins,
      • Koutakis P.
      • Miserlis D.
      • Myers S.A.
      • et al.
      Abnormal accumulation of desmin in gastrocnemius myofibers of patients with peripheral artery disease: associations with altered myofiber morphology and density, mitochondrial dysfunction and impaired limb function.
      ,
      • Weiss D.J.
      • Casale G.P.
      • Koutakis P.
      • et al.
      Oxidative damage and myofiber degeneration in the gastrocnemius of patients with peripheral arterial disease.
      abnormal morphology,
      • Koutakis P.
      • Myers S.A.
      • Cluff K.
      • et al.
      Abnormal myofiber morphology and limb dysfunction in claudication.
      ,
      • Cluff K.
      • Miserlis D.
      • Naganathan G.K.
      • et al.
      Morphometric analysis of gastrocnemius muscle biopsies from patients with peripheral arterial disease: objective grading of muscle degeneration.
      degeneration, and replacement with fibrotic tissue.
      • Ha D.M.
      • Carpenter L.C.
      • Koutakis P.
      • et al.
      Transforming growth factor-beta 1 produced by vascular smooth muscle cells predicts fibrosis in the gastrocnemius of patients with peripheral artery disease.
      These features are consistent with myofiber necrosis, a process that releases both damaged and intact molecules normally sequestered in the cell and, consequently, may be recognized as danger-associated molecular patterns (DAMPs),
      • Rock K.L.
      • Kono H
      The inflammatory response to cell death.
      autoantigens, or neoantigens. Recognition of DAMPs by pattern recognition receptors on macrophages and dendritic cells increases transcription of genes that code for inflammatory cytokines and chemokines.
      • Takeuchi O.
      • Akira S.
      Pattern recognition receptors and inflammation.
      In contrast, sequestered self-proteins released during necrosis may be immunogenic and in association with MHC class II molecules on antigen presenting cells activate autoreactive Th-cells,
      • Fernando M.M.
      • Stevens C.R.
      • Walsh E.C.
      • et al.
      Defining the role of the MHC in autoimmunity: a review and pooled analysis.
      thereby driving chronic autoimmune inflammation.
      • Becher B.
      • Tugues S.
      • Greter M.
      GM-CSF: from growth factor to central mediator of tissue inflammation.
      On the other hand, lipid peroxidation generates oxidatively damaged proteins that are highly immunogenic neoantigens.
      • Kurien B.T.
      • Scofield R.H.
      Autoimmunity and oxidatively modified autoantigens.
      In association with MHC class I alleles
      • Calis J.J.
      • Maybeno M.
      • Greenbaum J.A.
      • et al.
      Properties of MHC class I presented peptides that enhance immunogenicity.
      these neoantigens drive chronic T-cell activation and inflammation. We evaluated local expression of inflammatory cytokines including GM-CSF, IL-17A, IFN-ϒ, TNF-α, and IL-6, as well as TGFβ1
      • Shen W.
      • Li Y.
      • Zhu J.
      • Schwendener R.
      • Huard J.
      Interaction between macrophages, TGF-beta1, and the COX-2 pathway during the inflammatory phase of skeletal muscle healing after injury.
      and CCL5
      • Locati M.
      • Deuschle U.
      • Massardi M.L.
      • et al.
      Analysis of the gene expression profile activated by the CC chemokine ligand 5/RANTES and by lipopolysaccharide in human monocytes.
      and now present data supporting inflammation mediated by autoreactive Th-cells,
      • Komuczki J.
      • Tuzlak S.
      • Friebel E.
      • et al.
      Fate-mapping of GM-CSF expression identifies a discrete subset of inflammation-driving T helper cells regulated by cytokines IL-23 and IL-1beta.
      in calf muscle of claudicating patients with PAD, as well as increased age-associated inflammation in all PAD patients.

      METHODS

      Human subjects

      This study conforms to the ethical principles for medical research involving human subjects, as set forth by the World Medical Association (Declaration of Helsinki). The Institutional Review Boards of the VA Nebraska Western Iowa Medical Center and University of Nebraska Medical Center approved the experimental protocol and all subjects gave informed consent. We recruited 23 consecutive patients scheduled for lower extremity operations for the treatment of symptomatic PAD (Table I). Diagnosis of PAD was based on medical history, physical examination, decreased ABI (<0.9), and computerized or standard arteriography showing stenosis and/or occlusions in the arteries supplying the legs. We selected patients with Fontaine Stage II disease characterized by intermittent claudication in the absence of rest pain or tissue loss. Anatomic pattern of disease and planned treatment for each PAD patient enrolled are presented in Table II. We recruited 18 control patients scheduled for lower extremity operations for indications other than PAD (Table I). These patients had no history of PAD symptoms, normal blood flow to their legs and normal pulses in their lower extremities. Medical condition and planned treatment for each control subject enrolled are presented in Table III. All controls had normal ABIs at rest and after stress and all led sedentary lifestyles.
      Table IDemographics of claudicating patients with peripheral artery disease and control subjects
      CTRLPAD-IIP value
      Number of subjects1823_
      Median age (years)63.8 ± 1.763.8 ± 7.6
      Randomization t test was used for continuous variables.
      0.991
      Height (m)1.76 ± 0.021.79 ± 0.01
      Randomization t test was used for continuous variables.
      0.422
      Weight (kg)90.9 ± 3.790.6 ± 4.2
      Randomization t test was used for continuous variables.
      0.958
      Body mass index (BMI)29.9 ± 1.228.6 ± 1.2
      Randomization t test was used for continuous variables.
      0.438
      Ankle brachial index1.08 ± 0.010.57 ± 0.04
      Randomization t test was used for continuous variables.
      0.0001
      Gender % (male/female)97.6/2.4100/0
      Yates’ Continuity Corrected Chi-Square test was used for categorical variables.
      0.901
      Smoking % (N)
      Never33.3 (6)4.3 (1)
      Yates’ Continuity Corrected Chi-Square test was used for categorical variables.
      0.042
      Former33.3 (6)47.8 (11)
      Yates’ Continuity Corrected Chi-Square test was used for categorical variables.
      0.538
      Current33.3 (6)47.8 (11)
      Yates’ Continuity Corrected Chi-Square test was used for categorical variables.
      0.538
      Coronary artery disease %27.8 (5)34.8 (8)
      Yates’ Continuity Corrected Chi-Square test was used for categorical variables.
      0.889
      Obesity %27.8 (5)26.1 (6)
      Yates’ Continuity Corrected Chi-Square test was used for categorical variables.
      1.000
      Dyslipidemia %77.8 (14)82.6 (19)
      Yates’ Continuity Corrected Chi-Square test was used for categorical variables.
      1.000
      Diabetes mellitus %22.2 (4)30.4 (7)
      Yates’ Continuity Corrected Chi-Square test was used for categorical variables.
      0.815
      COPD %22.2 (4)26.1 (6)
      Yates’ Continuity Corrected Chi-Square test was used for categorical variables.
      1.000
      Renal insufficiency %5.6 (1)17.4 (4)
      Yates’ Continuity Corrected Chi-Square test was used for categorical variables.
      0.504
      Statins %77.8 (14)91.3 (21)
      Yates’ Continuity Corrected Chi-Square test was used for categorical variables.
      0.441
      Family history % cardiovascular diseases27.8 (5)26.1 (6)
      Yates’ Continuity Corrected Chi-Square test was used for categorical variables.
      1.000
      Hypertension %72.2 (13)78.3 (18)
      Yates’ Continuity Corrected Chi-Square test was used for categorical variables.
      0.936
      Abbreviations: CTRL, control subjects; PAD-II, claudicating patients with Fontaine Stage IIPeripheral Artery Disease; COPD, Chronic Obstructive Pulmonary Disease.
      Obesity: Body mass index >30.
      Renal Insufficiency: Creatinine Clearance <60 mL/min/1.73m2.
      Continuous variables are presented as mean ± S.E.M.
      low asterisk Randomization t test was used for continuous variables.
      Yates’ Continuity Corrected Chi-Square test was used for categorical variables.
      Table IIPatients with peripheral artery disease: anatomic pattern of disease and planned treatment
      Segment of arterial tree involvedTreatment
      Treatment Abbreviations: AF2, aorto-bifemoral bypass grafting; TF2, thoracobifemoral bypass grafting; CFA-BKPop (L), left common femoral artery to below-Knee popliteal artery bypass graft; SFA-BKPop (L), left superficial femoral artery to below-knee popliteal artery bypass graft; CFA-AKPop (L), left common femoral artery to above-knee popliteal artery bypass graft; SFA-TP (L), left superficial femoral artery to tibioperoneal trunk bypass graft; SET, supervised treadmill exercise therapy; HB-Exer, home-based exercise; Stent-AI(BI), stent-aortoIliac bilateral
      Ankle-brachial index
      Patient numberAorto-IliacFemoral-poplitealTibialPlannedLeftRight
      P1PositivePositiveNegativeAF20.640.10
      P2NegativePositiveNegativeCFA-BKPop (L)0.661.05
      P3NegativePositiveNegativeSFA-BKPop (L)0.541.16
      P4NegativePositiveNegativeCFA-AKPop (L)0.660.78
      P5PositivePositiveNegativeAF20.160.10
      P6PositivePositiveNegativeTF20.100.10
      P7NegativePositiveNegativeCFA-BKPop (L)0.551.03
      P8NegativePositiveNegativeSFA-TP (L)0.430.74
      P9PositiveNegativeNegativeSET0.660.68
      P10PositivePositiveNegativeAF20.620.28
      P11PositivePositiveNegativeHB-Exer0.400.43
      P12NegativePositiveNegativeSET0.560.68
      P13PositivePositiveNegativeSET0.680.66
      P14PositiveNegativeNegativeHB-Exer0.960.58
      P15PositiveNegativeNegativeHB-Exer1.030.66
      P16PositiveNegativeNegativeHB-Exer0.801.00
      P17NegativePositiveNegativeHB-Exer0.560.70
      P18NegativePositiveNegativeHB-Exer0.080.87
      P19PositivePositiveNegativeHB-Exer0.570.51
      P20PositivePositiveNegativeHB-Exer0.480.66
      P21PositiveNegativeNegativeStent-AI(Bi)0.840.64
      P22NegativePositiveNegativeHB-Exer0.340.76
      P23NegativePositiveNegativeHB-Exer0.580.88
      Treatment Abbreviations: AF2, aorto-bifemoral bypass grafting; TF2, thoracobifemoral bypass grafting; CFA-BKPop (L), left common femoral artery to below-Knee popliteal artery bypass graft; SFA-BKPop (L), left superficial femoral artery to below-knee popliteal artery bypass graft; CFA-AKPop (L), left common femoral artery to above-knee popliteal artery bypass graft; SFA-TP (L), left superficial femoral artery to tibioperoneal trunk bypass graft; SET, supervised treadmill exercise therapy; HB-Exer, home-based exercise; Stent-AI(BI), stent-aortoIliac bilateral
      Table IIIControl subjects: medical condition and treatment
      Medical conditionAnkle-brachial index
      Control numberDiagnosisTreatment
      Treatment Abbreviations: CEA, carotid endarterectomy; EVAR, endovascular abdominal aortic aneurysm repair; vein stripping for varicose veins; OAR, open abdominal aortic aneurysm repair. Each Biopsy was collected during the surgical procedure.
      LeftRight
      C1Carotid artery diseaseCEA1.021.02
      C2Abdominal aortic aneurysmEVAR1.101.10
      C3Carotid artery diseaseCEA1.131.14
      C4Iliac artery aneurysmEVAR1.101.10
      C5Varicose veinsVein stripping1.101.10
      C6Abdominal aortic aneurysmOAR1.220.95
      C7Carotid artery diseaseCEA1.061.06
      C8Popliteal aneurysmSFA-Pop RSVG1.101.10
      C9Abdominal aortic aneurysmEVAR1.111.03
      C10Varicose veinsVein stripping1.121.10
      C11Abdominal aortic aneurysmEVAR1.091.05
      C12Varicose veinsVein stripping1.101.10
      C13Carotid artery diseaseCEA1.031.01
      C14Varicose veinsVein stripping1.171.14
      C15Abdominal aortic aneurysmEVAR0.971.01
      C16Carotid artery diseaseCEA1.081.08
      C17Carotid artery diseaseCEA1.101.10
      C18Abdominal aortic aneurysmEVAR1.011.11
      Treatment Abbreviations: CEA, carotid endarterectomy; EVAR, endovascular abdominal aortic aneurysm repair; vein stripping for varicose veins; OAR, open abdominal aortic aneurysm repair.Each Biopsy was collected during the surgical procedure.

      Biopsy of the calf muscle

      Calf muscle biopsies weighing approximately 250 mg were obtained with a 6 mm Bergstrom needle, from the anteromedial aspect of the muscle belly, 10 cm distal to the tibial tuberosity. A biopsy was taken from the more diseased leg (lower ABI) of each PAD patient. Each biopsy was divided and immediately processed. Some pieces were frozen for ELISA and biochemical analyses and some were fixed in cold methacarn for future microscopic study.

      Quantification of cytokines in calf muscles of control subjects and patients with PAD, by sandwich ELISA

      We analyzed expression of 7 inflammatory cytokines in the calf muscle of both PAD patients and control subjects. These included Tumor Necrosis Factor alpha (TNF-α), Interferon gamma (IFN-γ), Interleukin 17A (IL-17A), Interleuklin 6 (IL-6), Granulocyte-Macrophage Colony-Stimulating Factor (GM-CSF), Chemokine Ligand 5 (CCL5), and Transforming Growth Factor beta 1 (TGFβ1). The lower limit of detection for these cytokines was 21.4, 12.6, 28.4, 14.0, 48.3, 8.0, and 19.7 pg/mL, respectively.

      Calf muscle homogenates

      One day prior to analysis, an 80 mg piece of calf muscle (stored in liquid nitrogen vapor) was washed with ice-cold Tris-buffered saline (50 mM Tris and 150 mM NaCl at pH 7.5) containing 30% v/v protease inhibitor mix (Sigma P-8340), blotted and then placed in 660 μL of the same buffer with protease inhibitor. The tissue was homogenized, on ice, with a Dounce homogenizer and then centrifuged at 14,000 x g for 20 minutes, at 4°C. Supernatant (570 μL), beneath the fat layer, was collected and stored at −80°C.

      ELISA procedure

      On the day of assay, the samples were thawed and diluted 1:1 (v:v) in sample dilution buffer supplied with the Multi-Analyte ELISA kit (Qiagen, Germantown, MD; product # 336111). Diluted samples and standards were added to an ELISA plate consisting of 8-well columns. Wells in each column were coated with capture antibody specific for 1 cytokine. After the appropriate incubations and washes according to the Qiagen protocol, biotinylated detection antibody wash added to the wells. This was followed by washes and then incubation for 30 minutes with streptavidin-HRP. Finally, the plate was washed, incubated for 15 minutes with development solution and then treated with stop solution. The plate was read with a microplate reader, at 450 nm and at 570 nm. The 450 absorbances were corrected for any plate imperfections by subtracting the corresponding 570 absorbances. Absorbances of samples and standards, in addition, were corrected for negative controls (sample dilution buffer instead of sample or standard per well). Cytokine concentrations (picograms/mL) of the sample dilutions added to the wells were computed from sample absorbance and absorbance of the corresponding cytokine standard. Cytokine concentrations in the original muscle specimen were computed from the sample dilutions, with the original addition of 80-mg muscle to 660 μL tris-buffered saline considered to be a dilution of 80 μL in a total of 740 μL (ie, a dilution of 4/37).

      Quantification of cytokine message in calf muscles of PAD patients and controls subjects

      Total RNA was extracted from frozen muscle (∼20 mg), with Trizol Reagent (Invitrogen, Austin, TX) per the manufacturer's instructions. RNA extracts were treated with DNAse and then purity and quantity of RNA were determined by Agilent Bioanalyzer (Agilent Technologies, Inc., Santa Clara, CA). About 200 ng per sample were used for the synthesis of cDNA in a 20 μL reaction volume containing RNA, random hexamer (Promega, Madison, WI), 0.5 mM dNTPs, 1 unit of RNAsin (Promega, Madison, WI), and 1 unit of Improm-II Reverse Transcriptase (Promega, Madison, WI) in a buffer containing 50 mM Tris-HCl pH 8.3, 75 mM KCl, 3 mM MgCl2, and 2 mM DTT. Reverse Transcriptase (RT) reactions were incubated at 42°C for 1 hour, followed by inactivation at 75°C for 15 minutes. After briefly chilling on ice, RNase H (Promega, Madison, WI) was added to remove RNA strand from RT reactions, which subsequently were diluted 5-fold with nuclease-free water. The synthesized cDNA's were aliquoted and stored at −84°C, and subsequently amplified by qPCR (CFX Connect, Bio-Rad Laboratories, Hercules, CA). For qPCR, 10% of each DNA aliquot was used with 200 pmol of each primer pair (Table IV) in DyNAmo SYBR Green qPCR mix (Fisher Scientific, Pittsburgh PA) according to the manufacturer's instructions. qPCR was implemented with a CFX96 connect (Bio-Rad, Hercules, CA). Cycling times were once at 95°C for 10 minutes followed by 40 cycles at 95°C for 15s and 60°C for 1 minute. Melting curves were analyzed for specificity.
      Table IVPrimers for quantification of cytokine mRNAs by the quantitative polymerase chain reaction
      mRNA5’———————————–3’ForwardReverseGenBank
      Interleukin 6 (IL 6)ACAAATGCCAGCCTGCTGACGAAG

      TCTGACCAGAAGAAGGAATGCCCA
      Forward

      Reverse
      NM_000600
      Interleukin 17A (IL 7A)CCGGCTGGAGAAGATACTGGTG

      GAACTCTAATGAGTTTAGTCCG
      Forward

      Reverse
      NM_002190
      Interferon-γ (IFN-γ)ACTAGGCAGCCAAC TAAGCAAGA

      CATCAGGGTCACCTGACACATTCA
      Forward

      Reverse
      NM_000619
      Tumor Necrosis Factor-α (TNF-α)CTGACAAGCTGCCAGGCAGGTTCTCTT

      ATCAGGAAGGAGAAGAGGCTGAGGAAC
      Forward

      Reverse
      NM_000594
      Transforming Growth Factor-β1 (TGF-β1)CCGCAAAGACTTTTCCCCAGACC

      ACCTAGATGGGCGCGATCTGGTA
      Forward

      Reverse
      NM_000660
      Chemokine Ligand 5 (CCL5)AGTGTGTGCCAACCCAGAGAAGAA

      TGCCTGTTTCTGCTTGCTCTTGTCCT
      Forward

      Reverse
      NM_002985
      Human Myosin Heavy ChainTCGAAAATCAGAAAAGGAGCGGAT

      GAGGGTTCATGGGGAAGACTTGGT
      Forward

      Reverse
      NM_005963
      Human Myosin Heavy Chain: Reference mRNA for quantification of cytokine message.

      Graded treadmill test and 6-minute walking distance

      Walking performance of each PAD patient was evaluated as Peak Walking Distance (PWD) and Claudication Onset Distance (COD) with the Gardner-Skinner treadmill test
      • Gardner A.W.
      • Skinner J.S.
      • Cantwell B.W.
      • Smith L.K.
      Progressive vs single-stage treadmill tests for evaluation of claudication.
      and 6-minute walking distance
      • Montgomery P.S.
      • Gardner A.W.
      The clinical utility of a six-minute walk test in peripheral arterial occlusive disease patients.
      administered indoors in a 20-m hallway.

      Quality of life assessments

      Physical function and health-related quality-of-life (QoL) were assessed with the Walking Impairment Questionnaire and Short Form 36 Health Survey,
      • Myers S.A.
      • Johanning J.M.
      • Stergiou N.
      • Lynch T.G.
      • Longo G.M.
      • Pipinos II
      Claudication distances and the Walking Impairment Questionnaire best describe the ambulatory limitations in patients with symptomatic peripheral arterial disease.
      on the same day that walking performance was assessed.

      Statistics

      Baseline characteristics for PAD and control subjects were compared with the Randomization t test for continuous variables and Yate's Continuity Corrected Chi Squared Test for categorical variables. Differences in the prevalence of detectable cytokine in calf muscles of PAD patients and Control subjects were determined with the Yates’ Continuity Corrected Chi Square and differences in cytokine concentrations were determined with the Randomization t test. Correlations and their significance among cytokine measurements were determined as the Spearman Rank or Kendall's Tau correlation coefficient. Differences between PAD patients with and those without the signature cytokine profile of autoreactive Th cells were determined with the Randomization t test for continuous variables and Fisher's Exact Test for categorical variables. All analyses were completed with NCSS statistical software version 19.0.1 (NCSS, LLC, Kaysville, UT) Statistical significance was held at 0.05.

      RESULTS

      Inflammatory cytokines in the calf muscles of patients with PAD

      Calf muscle homogenates and serum samples from PAD patients (N = 23) and control subjects (N = 18) were assayed by ELISA, for the 5 inflammatory cytokines IL-6, IL-17A, IFN-γ, TNF-α, and GM-CSF which were detected at picogram to nanogram per mL in calf muscle of PAD patients (Table V). Cytokine was detected in biopsy specimens from 19 PAD patients and specimens from 15 of these patients co-expressed 2 or more of these inflammatory cytokines. IL-17A, IFN-γ, TNF-α, and GM-CSF were co-expressed in the calf muscles of 11 PAD patients and muscles of 7 of these patients expressed, in addition, IL-6. Co-expressed inflammatory cytokines were strongly correlated (Table VI), establishing coordinated expression of these cytokines in the affected muscles of 65% of PAD patients. IL-17A, IFN-γ, and GM-CSF were not detected in the calf muscles of control subjects, whereas IL-6 was detected in the calf muscle of one control subject and TNF-α in that of another (Table V). Neither PAD nor control serum contained detectable IL-6, IFN-γ, TNF-α, or GM-CSF while 5 PAD sera and 4 control sera contained detectable IL-17A (Table VII). IL-17A in PAD serum was not correlated with PAD muscle IL-17A. Cytokines in the serum of PAD patients did not reflect the inflammatory environment within calf muscle. These data establish local inflammatory activity in the affected muscles of a substantial proportion of PAD patients.
      Table VCytokine concentrations in calf muscle biopsies of PAD patients and control subjects, determined by enzyme-linked immunosorbent assay (ELISA)
      PAD MuscleControl Muscle
      CytokinePositive (total) patientsValue for single positive (pg/mL)Median (pg/mL)Inter quartile range (pg/mL)Positive (total) controlsValue for single positive (pg/mL)Median (pg/mL)Inter quartile range (pg/mL)P value (PAD vs control)
      IL-614 (23)NA169424841 (18)2184NANA
      Yates’ Continuity Corrected Chi-Square.
      0.0009
      IFN-γ14 (23)NA153362260 (18)NANANA
      Yates’ Continuity Corrected Chi-Square.
      0.0002
      TNF-α13 (23)NA996215,3681 (18)4092NANA
      Yates’ Continuity Corrected Chi-Square.
      0.0020
      IL-17A11 (23)NA317552760 (18)NANANA
      Yates’ Continuity Corrected Chi-Square.
      0.0021
      GM-CSF11 (23)NA244427160 (18)NANANA
      Yates’ Continuity Corrected Chi-Square.
      0.0021
      TGFβ123 (23)NA35,01953,23418 (18)NA888212,239
      Randomization t test.
      0.0011
      CCL523 (23)NA21,26740,96818 (18)NA35602538
      Randomization t test.
      0.0001
      Abbreviations: IL-6, interleukin 6; IFN-ϒ, interferon gamma; TNF-α, tumor necrosis factor alpha; IL-17A, interleukin 17A; GM-CSF, Granulocyte Monocyte Colony Stimulating Factor; TGFβ1, transforming growth factor beta1; CCL5, chemokine ligand 5 also known as RANTES.
      low asterisk Yates’ Continuity Corrected Chi-Square.
      Randomization t test.
      Table VICorrelations among cytokines detected by ELISA in calf muscle of PAD Patients.
      CYTOKINEIL-6IL-17AIFN-γTNF-αGM-CSFTGFβ1CCL5
      IL-61

      ---
      Spearman Rho correlation coefficient.
      0.929

      p-value.
      0.0025
      0.800

      0.0096
      0.857

      0.0065
      0.929

      0.0025
      0.371

      0.1910
      0.147

      0.6154
      N=14N=7N=9N=8N=7N=14N=14
      IL-17A

      ---
      1

      ---
      0.900

      0.0002
      0.855

      0.0008
      0.946

      0.0001
      0.082

      0.8110
      0.064

      0.8525
      N=11N=11N=11N=11N=11N=11
      IFN-γ

      -----
      1

      ---
      0.879

      0.0001
      0.955

      0.0001
      0.073

      0.8054
      0.121

      0.6806
      N=14N=13N=11N=14N=14
      TNF-α

      ----
      1

      ---
      0.891

      0.0002
      -0.220

      0.4706
      -0.220

      0.4706
      N=13N=11N=13N=13
      GM-CSF

      -----
      1

      ---
      0.046

      0.8944
      0.046

      0.8944
      N=11N=11N=11
      TGFβ1

      -----
      1

      ---
      0.959

      0.0001
      N=23N=23
      N: The number of patients positive for the cytokine or cytokine pair.
      low asterisk Spearman Rho correlation coefficient.
      p-value.
      Table VIICytokine concentrations in sera of PAD patients and control subjects, determined by ELISA
      PAD SERUMControl SERUM
      CytokinePositive (total) patientsMedian (pg/mL)Inter quartile range (pg/mL)Positive (Total) controlsMedian (pg/mL)Inter quartile range (pg/mL)P value (PAD vs control)
      IL-60 (23)NANA0 (18)NANAND
      IFN-γ0 (23)NANA0 (18)NANAND
      TNF-α0 (23)NANA0 (18)NANAND
      IL-17A5 (23)7637124 (18)14962071
      Randomization t test.
      0.1970
      GM-CSF0 (23)NANA0 (18)NANAND
      TGFβ123 (23)17365615518 (18)144813428
      Randomization t test.
      0.0006
      CCL523 (23)8247433018 (18)8920468
      Randomization t test.
      0.1750
      low asterisk Randomization t test.

      Walking performance and QoL of PAD of patients who co-expressed IL-17A, IFN-γ, TNF-α, and GM-CSF in their calf muscle and PAD patients who did not co-express these cytokines

      We identified 11 PAD patients who co-expressed IL-17A, IFN-γ, TNF-α, and GM-CSF in calf muscle. Walking performance and QoL were compared (Table VIII) with the remaining 12 PAD patients of whom 2 co-expressed IFN-γ and TNF-α, 1 expressed TNF-α, only, and the remainder did not express any of the 4 inflammatory cytokines. These 2 groups did not differ in any of the demographic parameters identified in Table I or in anatomic pattern of disease Table IX. If there were a strong and direct association of local inflammation with leg function parameters, we would expect reduced walking performance and/or decreased QoL in those patients who co-expressed the 4 inflammatory cytokines. We found a significantly (P= 0.017) lower median COD (17 vs 102 m) in patients who co-expressed all 4 cytokines, and, although median PWD was substantially lower (55 vs 147 m) in these patients, it did not reach statistical significance (P= 0.169). Six-minute walking distance was nearly the same in both groups of patients with a median of 324 m for the 11 patients who co-expressed all 4 cytokines and 305 m for the remaining 12 patients. None of the QoL measures were significantly different between these 2 groups of patients.
      Table VIIIWalking performance and quality-of-life of patients co-expressing IL-17A, IFN-γ, TNF-α, and GM-CSF compared to patients who did not co-express these cytokines
      Group 1: Patients co-expressing IL-17A, IFN-γ, TNF-α, and GM-CSFGroup 2: Patients not co-expressing IL-17A, IFN-γ, TNF-α, and GM-CSF
      Walking performance / Quality-of-LifeMedian (N)LCL -> UCLMedian (N)LCL -> UCL
      Randomization t test.
      P value
      Six-minute walk distance (meters)324 (11)210 -> 353305 (11)204 -> 3260.498
      Claudication onset distance (meters)17 (7)5 -> 99102 (8)21 -> 1170.017
      Peak walking distance (meters)55 (10)23 -> 215147 (11)55 -> 1900.169
      WIQ-150 (11)25 -> 5050 (11)25 -> 631.000
      WIQ-268 (11)29 -> 8350 (11)25 -> 640.164
      WIQ-367 (11)38 -> 7544 (11)17 -> 810.486
      WIQ-450 (11)8 -> 7567 (11)33 -> 750.187
      SF-physical function50 (11)22 -> 6545 (11)20 -> 700.814
      SF-pain45 (11)23 -> 6045 (11)23 -> 550.511
      SF-general health60 (11)35 -> 6950 (11)30 -> 700.714
      WIQ, Walking Impairment Questionnaire; SF, short form 36.
      LCL, lower confidence limit; UCL, upper confidence limit.
      Of the patients in Group 2, 3 expressed IFN-γ, or IFN-γ plus TNF-α, and the remainder expressed none of the 4 inflammatory cytokines.
      low asterisk Randomization t test.
      Table IXThe relationship between the level of arterial tree blockages of claudicating patients with peripheral artery disease and the autoimmune inflammatory cytokine profile in their calf muscle
      Segment of the arterial tree with hemodynamically significant blockages.
      Anatomic disease pattern
      PAD patients with the autoimmune cytokine profile (N = 11)PAD patients without the autoimmune cytokine profile (N = 12)
      Fisher's Exact Test: The test is for the null hypothesis that the distribution of blockages is independent of the autoimmune cytokine profile. The null hypothesis is rejected at a P value of 0.05 or less.
      P value
      Aorto-Iliac involvement with or without a concurrent Femoro-Popliteal involvement.
      Aorto-Iliac
      8 yes / 3 no5 yes / 7 no0.2138
      Femoro-Popliteal involvement with or without an Aorto-Iliac involvement.
      Femoro-Popliteal
      6 yes / 5 no11 yes / 1 no0.0686
      Aorto-Iliac and Femoro-Popliteal4 yes / 7 no4 yes / 8 no1.0000
      Aorto-Iliac only4 yes / 7 no2 yes / 10 no0.3707
      Femoro-Popliteal only3 yes / 8 no7 yes / 5 no0.2138
      low asterisk Segment of the arterial tree with hemodynamically significant blockages.
      Aorto-Iliac involvement with or without a concurrent Femoro-Popliteal involvement.
      Femoro-Popliteal involvement with or without an Aorto-Iliac involvement.
      Fisher's Exact Test: The test is for the null hypothesis that the distribution of blockages is independent of the autoimmune cytokine profile. The null hypothesis is rejected at a P value of 0.05 or less.

      TGFβ1and CCL5 in calf muscle biopsies and serum samples of PAD patients and control subjects

      Calf muscle biopsies and serum samples collected from PAD patients (N = 23) and control subjects (N = 18) were evaluated for 2 more cytokines with inflammatory properties: TGFβ1, a pro-fibrotic inflammatory cytokine,
      • Han G.
      • Li F.
      • Singh T.P.
      • Wolf P.
      • Wang X.J.
      The pro-inflammatory role of TGFbeta1: a paradox?.
      and CCL5, a chemokine with monocyte chemotactic activity. TGFβ1 and CCL5 were detected at high picogram per mL of calf muscle of all PAD patients and control subjects, however, median concentrations were 3.94 and 5.67 times greater, respectively, in PAD compared to control muscle (Table V). The greater concentrations of TGF β1 (P < 0.002) and CCL5 (P < 0.001) in PAD muscle are consistent with increased gene expression in the affected calf muscles of PAD patients. It is well known that megakaryocytes synthesize these and other growth factors, cytokines, and chemokines that are sequestered in their fragmentation products, that is, blood platelets.
      • Meyer A.
      • Wang W.
      • Qu J.
      • et al.
      Platelet TGF-beta1 contributions to plasma TGF-beta1, cardiac fibrosis, and systolic dysfunction in a mouse model of pressure overload.
      ,
      • Fava R.A.
      • Casey T.T.
      • Wilcox J.
      • Pelton R.W.
      • Moses H.L.
      • Nanney L.B.
      Synthesis of transforming growth factor-beta 1 by megakaryocytes and its localization to megakaryocyte and platelet alpha-granules.
      Since platelets are expected to be present in muscle biopsies and their contents released in muscle homogenates, we extended our analysis of TGFβ1 and CCL5 to evaluate potential contributions from contaminating platelets. Median concentrations of TGFβ1 and CCL5 in PAD and control sera (Table VII) were similar with PAD/control ratios of 1.20 and 0.92, respectively. If platelets were a principal source of these cytokines in muscle, we would expect these ratios to be close to 1.0 reflecting the platelet cytokine complement released during serum preparation. As noted, PAD/control ratios for muscle were 3.94 and 5.67 for TGFβ1 and CCL5, respectively. More conclusively, these cytokines exhibited strong correlations across PAD muscles (N = 23; Kendall's Tau = 0.834; P< 0.0001) and across control muscles (N = 18; Kendall's Tau = 0.779; P< 0.0001; Fig 1, A and B) but were not correlated across PAD or control sera (Fig 1, C and D). If contaminating platelets accounted for TGFβ1 and CCL5 in muscle, we would expect the correlations in muscle to occur in serum. We conclude that coordinated expression of these 2 inflammatory cytokines in muscle biopsy specimens is a local event in both PAD and control muscle.
      Fig 1
      Fig 1TGFβ1 and CCL5 proteins were correlated in calf muscle across PAD patients (Panel A) and control subjects (Panel B) but were not correlated in serum across PAD patients (Panel C) or control subjects (Panel D). Muscle homogenates were prepared from biopsies of calf muscle of the more affected leg (lower ABI) of 23 patients and from calf muscle of 18 controls. Serum was collected at the time of biopsy. Cytokines were quantified by sandwich ELISA (Qiagen Kit; Germantown, MD) and their concentrations expressed as picograms per mL (pg/mL) muscle. Correlations are expressed as Kendall's Tau (R) with P values.

      TGFβ1 and CCL5 message in calf muscles of PAD patients and control subjects

      In another test for local gene expression, we quantified TGFβ1 and CCL5 transcripts in calf muscles of PAD patients and control subjects, by RT qPCR. Total RNA was extracted from frozen muscle biopsies of a subset of the PAD patients (N = 12) and control subjects (N = 12) and cDNAs were prepared by reverse transcriptase reactions. Subsequently, cDNAs for TGFβ1 and the reference protein myosin heavy chain (MyHC) were amplified by qPCR. The median transcript ratio for TGFβ1 (TGFβ1/MyHC) was nearly 3.5 times greater (P= 0.0260) in PAD muscle (ratio, LCL -> UCL: 0.00893, 0.00020 –> 0.02714) compared to control muscle (ratio, LCL -> UCL: 0.00259, 0.00055 –> 0.00694), suggesting greater inflammatory activity in the affected leg muscles of PAD patients. The median transcript ratio for CCL5 (CCL5/MyHC) was 43 times greater (P= 0.0005) in PAD muscle (0.00729, 0.00014 –> 0.01828) compared to control muscle (0.00017, 0.00011 –> 0.00037). The greater concentrations of TGFβ1 and CCL5 transcripts in PAD compared to control muscle support increased gene expression in the affected muscle of PAD patients. This interpretation was strengthened by analysis of the correlations between RNA transcripts for TGFβ1 and CCL5 and the corresponding proteins determined by ELISA. Transcripts for both TGFβ1 and CCL5 and their corresponding proteins in PAD muscle, were strongly correlated (Kendall's Tau = 0.833, P< 0.0001 and 0.778, P< 0.0001, respectively) (Fig 2, A and B), however, cytokine transcripts and protein in control muscle were not correlated (Fig 2, C and D). The absence of correlation in control muscle is consistent with low-level, steady-state gene expression.
      Fig 2
      Fig 2TGFβ1 and CCL5 proteins correlated with their corresponding gene transcripts in calf muscle of PAD patients (Panels A and B) but did not correlate in calf muscle of control subjects (Panels C and D). Muscle homogenates were prepared from biopsies of 12 PAD patients and 12 control subjects. Cytokines were quantified by sandwich ELISA (Qiagen; Germantown, MD) and their concentrations expressed as picograms per mL (pg/mL) muscle. Gene transcripts were quantified by RT qPCR and abundances expressed as a ratio to transcripts of the human myosin heavy chain gene. Correlations are expressed as Kendall's Tau (R) with P values.

      DISCUSSION

      Studies by many investigators have documented associations between circulating inflammatory cytokines and limb function, walking performance, and QoL as well as risk for PAD.
      • Brevetti G.
      • Giugliano G.
      • Brevetti L.
      • Hiatt W.R.
      Inflammation in peripheral artery disease.
      ,
      • McDermott M.M.
      • Ferrucci L.
      • Guralnik J.M.
      • et al.
      Elevated levels of inflammation, d-dimer, and homocysteine are associated with adverse calf muscle characteristics and reduced calf strength in peripheral arterial disease.
      ,
      • Brevetti G.
      • Piscione F.
      • Schiano V.
      • Galasso G.
      • Scopacasa F.
      • Chiariello M.
      Concomitant coronary and peripheral arterial disease: relationship between the inflammatory status of the affected limb and the severity of coronary artery disease.
      ,
      • Ridker P.M.
      • Stampfer M.J.
      • Rifai N.
      Novel risk factors for systemic atherosclerosis: a comparison of C-reactive protein, fibrinogen, homocysteine, lipoprotein(a), and standard cholesterol screening as predictors of peripheral arterial disease.
      ,
      • Tzoulaki I.
      • Murray G.D.
      • Lee A.J.
      • Rumley A.
      • Lowe G.D.
      • Fowkes F.G.
      C-reactive protein, interleukin-6, and soluble adhesion molecules as predictors of progressive peripheral atherosclerosis in the general population: Edinburgh Artery Study.
      • Nylaende M.
      • Kroese A.
      • Stranden E.
      • et al.
      Markers of vascular inflammation are associated with the extent of atherosclerosis assessed as angiographic score and treadmill walking distances in patients with peripheral arterial occlusive disease.
      • McDermott M.M.
      • Greenland P.
      • Green D.
      • et al.
      D-dimer, inflammatory markers, and lower extremity functioning in patients with and without peripheral arterial disease.
      However, none of these studies evaluated local inflammation in lower extremity muscles, and the outcomes of all these studies are readily interpreted as representative of the burden of inflamed atherosclerotic plaques in these patients. Atherosclerotic disease in PAD patients is severe
      • Brevetti G.
      • Giugliano G.
      • Brevetti L.
      • Hiatt W.R.
      Inflammation in peripheral artery disease.
      and the burden of plaques is expected to be directly related to both limb dysfunction and serum concentrations of inflammatory cytokines and acute-phase proteins. We identified 2 groups of PAD patients at the extremes of cytokine expression in the calf muscle. One group of 11 patients co-expressed 4 inflammatory cytokines (IL-17A, IFN-γ, TNF-α, and GM-CSF) and another group of 12 patients did not co-expressed these 4 cytokines. If local inflammation has a strong, direct effect on walking performance or QoL, we would expect significant differences between the 2 groups. Mean COD was significantly lower (83% lower; P= 0.017) in co-expressors (17 m) compared to patients who did not co-express these 4 cytokines (102 m). PWD also was reduced in co-expressors (55 m vs 147 m) but did not reach significance (P= 0.169). Given the small population size, these differences suggest a strong effect of autoimmune inflammation on walking performance.
      Detection of GM-CSF in calf muscle of PAD patients but not in calf muscle of control subjects is a significant finding of the present study. Until recently, GM-CSF was considered primarily a hemopoietic growth factor, but now it is clear that the cytokine is for the most part not required for development of macrophages and granulocytes.
      • Becher B.
      • Tugues S.
      • Greter M.
      GM-CSF: from growth factor to central mediator of tissue inflammation.
      GM-CSF is a central and essential mediator of tissue inflammation.
      • Komuczki J.
      • Tuzlak S.
      • Friebel E.
      • et al.
      Fate-mapping of GM-CSF expression identifies a discrete subset of inflammation-driving T helper cells regulated by cytokines IL-23 and IL-1beta.
      The cytokine is absent from the circulation and is produced and active locally at sites of tissue inflammation.
      • Becher B.
      • Tugues S.
      • Greter M.
      GM-CSF: from growth factor to central mediator of tissue inflammation.
      ,
      • Hamilton J.A.
      • Anderson G.P.
      GM-CSF Biology.
      GM-CSF is increased at diverse sites of inflammation and is essential for disease progression in models of arthritis and multiple sclerosis (MS).
      • Becher B.
      • Tugues S.
      • Greter M.
      GM-CSF: from growth factor to central mediator of tissue inflammation.
      ,
      • Komuczki J.
      • Tuzlak S.
      • Friebel E.
      • et al.
      Fate-mapping of GM-CSF expression identifies a discrete subset of inflammation-driving T helper cells regulated by cytokines IL-23 and IL-1beta.
      The central mechanism of chronic inflammation across diverse inflammatory diseases including rheumatoid arthritis (RA), psoriasis (PS), MS, and diabetes is activation of tissue damaging myeloid (vs embryonically derived resident) macrophages by GM-CSF secreted from activated autoreactive Th-cells.
      • Komuczki J.
      • Tuzlak S.
      • Friebel E.
      • et al.
      Fate-mapping of GM-CSF expression identifies a discrete subset of inflammation-driving T helper cells regulated by cytokines IL-23 and IL-1beta.
      The ability of these Th-cells to initiate tissue inflammation requires their production of GM-CSF under the influence of IL-23 which is essential for development of many inflammatory diseases.
      • Becher B.
      • Durell B.G.
      • Noelle R.J.
      Experimental autoimmune encephalitis and inflammation in the absence of interleukin-12.
      ,
      • Shiomi A.
      • Usui T.
      Pivotal roles of GM-CSF in autoimmunity and inflammation.
      Clinical treatment of PS patients with blocking antibody to IL-23 has produced improvement in these patients and targeting GM-CSF with blocking antibody has produced improvement in patients with RA.
      • Becher B.
      • Tugues S.
      • Greter M.
      GM-CSF: from growth factor to central mediator of tissue inflammation.
      ,
      • Wicks I.P.
      • Roberts A.W.
      Targeting GM-CSF in inflammatory diseases.
      GM-CSF+ Th-cells are part of a functionally integrated Th-cell subset that secretes GM-CSF, IFN-ϒ, TNF-α and IL-17A, a multi-cytokine signature.
      • Komuczki J.
      • Tuzlak S.
      • Friebel E.
      • et al.
      Fate-mapping of GM-CSF expression identifies a discrete subset of inflammation-driving T helper cells regulated by cytokines IL-23 and IL-1beta.
      Though not part of this Th-cell subset signature, IL-6 expression is upregulated along with Toll-Like receptors in GM-CSF stimulated myeloid macrophages.
      • Shiomi A.
      • Usui T.
      Pivotal roles of GM-CSF in autoimmunity and inflammation.
      The presence of GM-CSF together with strong correlations (Spearman coefficients of 0.855–0.955) among these 4 inflammatory cytokines in the calf muscles of 11/23 claudicating PAD patients' supports a local autoimmune inflammation in their calf muscles. Since activation of autoreactive Th-cells is MHC Class II restricted,
      • Fernando M.M.
      • Stevens C.R.
      • Walsh E.C.
      • et al.
      Defining the role of the MHC in autoimmunity: a review and pooled analysis.
      ,
      • Komuczki J.
      • Tuzlak S.
      • Friebel E.
      • et al.
      Fate-mapping of GM-CSF expression identifies a discrete subset of inflammation-driving T helper cells regulated by cytokines IL-23 and IL-1beta.
      ,
      • Goverman J.
      Autoimmune T cell responses in the central nervous system.
      we suggest that MHC Class II haplotype is a significant risk factor for autoimmune inflammation in calf muscle of PAD patients. Moreover, the availability of therapeutic, blocking antibodies to GM-CSF and IL-23 and their documented safety and efficacy in clinical trials, offer the opportunity for specific intervention in PAD.
      • Becher B.
      • Tugues S.
      • Greter M.
      GM-CSF: from growth factor to central mediator of tissue inflammation.
      ,
      • Komuczki J.
      • Tuzlak S.
      • Friebel E.
      • et al.
      Fate-mapping of GM-CSF expression identifies a discrete subset of inflammation-driving T helper cells regulated by cytokines IL-23 and IL-1beta.
      ,
      • Wicks I.P.
      • Roberts A.W.
      Targeting GM-CSF in inflammatory diseases.
      Tissue inflammation has been reported as profibrotic and injurious to tissue, often leading to organ failure, and, conversely, as essential for tissue repair and maintenance of organ function. Addressing this problem, Lavine et al.
      • Lavine K.J.
      • Epelman S.
      • Uchida K.
      • et al.
      Distinct macrophage lineages contribute to disparate patterns of cardiac recovery and remodeling in the neonatal and adult heart.
      studied macrophage populations in a model of cardiomyocyte injury in neonatal and adult mice. In neonatal mice, cardiac recovery occurred in association with cardiomyocyte proliferation, angiogenesis, and minimal inflammation. In adult mice, cardiac repair was limited and was associated with a robust inflammatory response and tissue damage. Cardiac recovery in neonatal mice was linked to the presence of resident (embryonically derived) macrophages in the injured heart, while inflammation and tissue damage in the injured adult heart was linked to replacement of resident macrophages with monocyte-derived (myeloid) macrophages. In the present study, we documented expression of GM-CSF, a central communicator between activated autoreactive Th-cells and tissue damaging myeloid macrophages,
      • Becher B.
      • Tugues S.
      • Greter M.
      GM-CSF: from growth factor to central mediator of tissue inflammation.
      as part of an inflammatory cytokine signature of autoreactive Th-cells
      • Komuczki J.
      • Tuzlak S.
      • Friebel E.
      • et al.
      Fate-mapping of GM-CSF expression identifies a discrete subset of inflammation-driving T helper cells regulated by cytokines IL-23 and IL-1beta.
      in calf muscle of a significant proportion (11/23) of claudicating patients with PAD. Consequently, we propose that myeloid macrophages predominate in the inflamed calf muscles of these PAD patients and are the immediate effectors of muscle damage and fibrosis. Our future studies will address this proposal.
      In this study, we found that TGFβ1 and CCL5 were co-expressed in all calf muscle specimens but were present at higher concentrations in the calf muscles of PAD patients compared to control subjects. These cytokines exhibited strong correlations (R = 0.957 and 0.913, respectively) across PAD muscles and across control muscles, suggesting their origin from a common process in each group. For both TGFβ1 and CCL5, the strong correlation between mRNA and protein in PAD muscle indicated upregulated gene expression. In contrast, the absence of detectable correlation in control muscle indicated low-level, steady-state gene expression, consistent with the observed lower concentrations of these cytokines in control muscle. Our findings with control specimens are consistent with chronic, low-grade inflammation that develops with age, a process that has been termed “inflammaging.”
      • Franceschi C.
      • Garagnani P.
      • Vitale G.
      • Capri M.
      • Salvioli S
      Inflammaging and 'Garb-aging'.
      Several mechanisms have been proposed for inflammaging, including a heightened acute-phase response, reduced capacity to resolve immune responses, increased oxidative stress, and innate and/or adaptive immune responses to cellular self-debris including both damaged proteins and intact immunoprivileged molecules.
      • Franceschi C.
      • Garagnani P.
      • Vitale G.
      • Capri M.
      • Salvioli S
      Inflammaging and 'Garb-aging'.
      ,
      • Woods J.A.
      • Wilund K.R.
      • Martin S.A.
      • Kistler B.M.
      Exercise, inflammation and aging.
      Currently, increased accumulation of cellular debris, misfolded proteins, and exposed immunoprivileged molecules are considered the predominant stimuli for inflammaging, which is consistent with an imbalance of cell damage and repair and the decline of autophagy with age.
      • Franceschi C.
      • Garagnani P.
      • Vitale G.
      • Capri M.
      • Salvioli S
      Inflammaging and 'Garb-aging'.
      Oxidatively modified proteins accumulate with advancing age and should be included among these stimuli.
      • Kurien B.T.
      • Scofield R.H.
      Autoimmunity and oxidatively modified autoantigens.
      ,
      • Woods J.A.
      • Wilund K.R.
      • Martin S.A.
      • Kistler B.M.
      Exercise, inflammation and aging.
      ,
      • Kregel K.C.
      • Zhang H.J.
      An integrated view of oxidative stress in aging: basic mechanisms, functional effects, and pathological considerations.
      The precise mechanism of inflammaging is not known but it is possible that the condition is mediated by activation of pattern recognition receptors on macrophages, dendritic cells and/or neutrophils, by DAMPs
      • Takeuchi O.
      • Akira S.
      Pattern recognition receptors and inflammation.
      that may be expressed on oxidatively damaged proteins,
      • Gill R.
      • Tsung A.
      • Billiar T.
      Linking oxidative stress to inflammation: toll-like receptors.
      nonoxidatively damaged proteins
      • Franceschi C.
      • Garagnani P.
      • Vitale G.
      • Capri M.
      • Salvioli S
      Inflammaging and 'Garb-aging'.
      ,
      • Franceschi C.
      • Campisi J.
      Chronic inflammation (inflammaging) and its potential contribution to age-associated diseases.
      and unaltered immunoprivileged proteins, and other molecules
      • Rock K.L.
      • Kono H
      The inflammatory response to cell death.
      ,
      • Bianchi M.E.
      DAMPs, PAMPs and alarmins: all we need to know about danger.
      released from damaged or dying cells. Alternatively, inflammaging may be mediated by Th-cells that are not autoreactive but react with neoantigens generated by oxidative damage to self-proteins.
      • Kurien B.T.
      • Scofield R.H.
      Autoimmunity and oxidatively modified autoantigens.
      ,
      • Calis J.J.
      • Maybeno M.
      • Greenbaum J.A.
      • et al.
      Properties of MHC class I presented peptides that enhance immunogenicity.
      We propose that upregulated inflammaging and Th-cell mediated autoimmune inflammation occur concurrently in damaged muscle of the lower legs of claudicating patients with PAD.

      Limitations

      A limitation of the present study is the relatively small number of PAD patients and control subjects included in the study. Despite this limitation, we observed a statistically significant (P= 0.017) and markedly lower COD for patients who expressed the cytokine signature of activated autoreactive Th-cells (17 m) compared to those patients who did not express this signature in their calf muscles (102 m). Moreover, PWD was markedly lower in patients who expressed this cytokine signature (55 m) compared to those who did not (147 m) but did not reach statistical significance (P= 0.169).
      Serum cytokine measurements were not informative of muscle cytokine concentrations and did not serve as markers of local inflammation. Our assay was sensitive, specific, and reproducible. Adjusting for serum dilution, lower limits of detection for our ELISA were 28.0, 56.8, 19.6, 42.8, and 96.6 pg/mL for IL-6, IL-17A, IFN-γ, TNF-α, and GM-CSF, respectively. It is possible that these cytokines are present at lower concentrations and may be informative of muscle cytokines. However, it has been reported that blood values of inflammatory cytokines are not informative of local inflammation.
      • Franceschi C.
      • Garagnani P.
      • Vitale G.
      • Capri M.
      • Salvioli S
      Inflammaging and 'Garb-aging'.
      Another limitation of the manuscript is our inability to determine precisely the contribution of locally produced TGFβ1 and CCL5 to our measurements of calf muscle, because of the potential contribution of these cytokines from possible platelet contamination as well as TGFβ1 production by hyperplastic vSMC in small veins and arteries of PAD muscle.
      • Ha D.M.
      • Carpenter L.C.
      • Koutakis P.
      • et al.
      Transforming growth factor-beta 1 produced by vascular smooth muscle cells predicts fibrosis in the gastrocnemius of patients with peripheral artery disease.
      Strong correlations of these cytokines in PAD and control muscle and the absence of correlation in serum, however, support the conclusion that our cytokine measurements were largely unaffected by platelets or vSMC. Instead, our measurements likely represent low-grade inflammation in calf muscles of PAD patients and control subjects.

      CONCLUSION

      The principal finding of the present study was the presence of a signature cytokine profile of inflammation mediated by autoreactive Th-cells, in calf muscle of nearly 50% of claudicating patients with PAD. This signature of autoimmune inflammation comprised highly coordinated expression of 4 inflammatory cytokines (GM-CSF, IL-17A, IFN-ϒ, and TNF-α), was associated with decreased leg function and was not found in control subjects. Patient serum was not informative of this signature in muscle. GM-CSF is the central and essential driver of tissue damage mediated by myeloid macrophages. Clinically tested antibodies against GM-CSF are relatively safe therapeutic agents that have been effective in treating RA, an autoimmune inflammatory disease, and offer the potential for effective therapeutic intervention for ca.50% of claudicating patients with PAD. Since Th-cell mediated autoimmune inflammation is MHC Class II restricted, it may be possible to identify PAD patients at high risk of autoimmune inflammation based on their MHC Class II haplotype, thus, substituting an invasive biopsy-based diagnostic procedure with a much less invasive procedure based on a blood sample. More research is needed to assess (1) MHC Class II haplotype as a biomarker of autoimmune inflammation in claudicating patients with PAD and (2) the effectiveness of anti-GM-CSF therapy for PAD patients with inflamed myopathy.

      Acknowledgments

      Conflicts of Interest: All authors have read the journal's policy on disclosure. There are no conflicts of interest to disclose.
      NIH Grant R01AG034995 , R01AG049868 , and R01HD090333 and the Charles and Mary Heider Fund for Excellence in Vascular Surgery
      All authors have read the journal's authorship agreement
      All authors have reviewed and approved this manuscript.

      References

        • Brevetti G.
        • Giugliano G.
        • Brevetti L.
        • Hiatt W.R.
        Inflammation in peripheral artery disease.
        Circulation. 2010; 122: 1862-1875
        • Hiatt W.R.
        • Armstrong E.J.
        • Larson C.J.
        • Brass E.P.
        Pathogenesis of the limb manifestations and exercise limitations in peripheral artery disease.
        Circ Res. 2015; 116: 1527-1539
        • Koutakis P.
        • Miserlis D.
        • Myers S.A.
        • et al.
        Abnormal accumulation of desmin in gastrocnemius myofibers of patients with peripheral artery disease: associations with altered myofiber morphology and density, mitochondrial dysfunction and impaired limb function.
        J Histochem Cytochem. 2015; 63: 256-269
        • Koutakis P.
        • Myers S.A.
        • Cluff K.
        • et al.
        Abnormal myofiber morphology and limb dysfunction in claudication.
        J Surg Res. 2015; 196: 172-179
        • Schieber M.N.
        • Hasenkamp R.M.
        • Pipinos I.I.
        • et al.
        Muscle strength and control characteristics are altered by peripheral artery disease.
        J Vasc Surg. 2017; 66 (e12): 178-186
        • Pipinos II
        • Judge A.R.
        • Selsby J.T.
        • et al.
        The myopathy of peripheral arterial occlusive disease: part 1. Functional and histomorphological changes and evidence for mitochondrial dysfunction.
        Vasc Endovascular Surg. 2007; 41: 481-489
        • Weiss D.J.
        • Casale G.P.
        • Koutakis P.
        • et al.
        Oxidative damage and myofiber degeneration in the gastrocnemius of patients with peripheral arterial disease.
        J Transl Med. 2013; 11: 230
        • Koutakis P.
        • Weiss D.J.
        • Miserlis D.
        • et al.
        Oxidative damage in the gastrocnemius of patients with peripheral artery disease is myofiber type selective.
        Redox Biol. 2014; 2: 921-928
        • Gillani S.
        • Cao J.
        • Suzuki T.
        • Hak D.J.
        The effect of ischemia reperfusion injury on skeletal muscle.
        Injury. 2012; 43: 670-675
        • McDermott M.M.
        • Ferrucci L.
        • Guralnik J.M.
        • et al.
        Elevated levels of inflammation, d-dimer, and homocysteine are associated with adverse calf muscle characteristics and reduced calf strength in peripheral arterial disease.
        J Am Coll Cardiol. 2007; 50: 897-905
        • Brevetti G.
        • Piscione F.
        • Schiano V.
        • Galasso G.
        • Scopacasa F.
        • Chiariello M.
        Concomitant coronary and peripheral arterial disease: relationship between the inflammatory status of the affected limb and the severity of coronary artery disease.
        J Vasc Surg. 2009; 49: 1465-1471
        • Violi F.
        • Criqui M.
        • Longoni A.
        • Castiglioni C.
        Relation between risk factors and cardiovascular complications in patients with peripheral vascular disease. Results from the A.D.E.P. study.
        Atherosclerosis. 1996; 120: 25-35
        • Silvestro A.
        • Brevetti G.
        • Schiano V.
        • Scopacasa F.
        • Chiariello M.
        Adhesion molecules and cardiovascular risk in peripheral arterial disease. Soluble vascular cell adhesion molecule-1 improves risk stratification.
        Thromb Haemost. 2005; 93: 559-563
        • Ridker P.M.
        • Stampfer M.J.
        • Rifai N.
        Novel risk factors for systemic atherosclerosis: a comparison of C-reactive protein, fibrinogen, homocysteine, lipoprotein(a), and standard cholesterol screening as predictors of peripheral arterial disease.
        JAMA. 2001; 285: 2481-2485
        • van der Meer I.M.
        • de Maat M.P.M.
        • Bots M.L.
        • et al.
        Inflammatory mediators and cell adhesion molecules as indicators of severity of atherosclerosis.
        Arterioscleros Thrombos Vasc Biol. 2002; 22: 838-842
        • Tzoulaki I.
        • Murray G.D.
        • Lee A.J.
        • Rumley A.
        • Lowe G.D.
        • Fowkes F.G.
        C-reactive protein, interleukin-6, and soluble adhesion molecules as predictors of progressive peripheral atherosclerosis in the general population: Edinburgh Artery Study.
        Circulation. 2005; 112: 976-983
        • Nylaende M.
        • Kroese A.
        • Stranden E.
        • et al.
        Markers of vascular inflammation are associated with the extent of atherosclerosis assessed as angiographic score and treadmill walking distances in patients with peripheral arterial occlusive disease.
        Vasc Med. 2006; 11: 21-28
        • McDermott M.M.
        • Greenland P.
        • Green D.
        • et al.
        D-dimer, inflammatory markers, and lower extremity functioning in patients with and without peripheral arterial disease.
        Circulation. 2003; 107: 3191-3198
        • McDermott M.M.
        • Liu K.
        • Guralnik J.M.
        • et al.
        Functional decline in patients with and without peripheral arterial disease: predictive value of annual changes in levels of C-reactive protein and D-dimer.
        J Gerontol A Biol Sci Med Sci. 2006; 61: 374-379
        • Brass E.P.
        • Hiatt W.R.
        Acquired skeletal muscle metabolic myopathy in atherosclerotic peripheral arterial disease.
        Vasc Med. 2000; 5: 55-59
        • Cluff K.
        • Miserlis D.
        • Naganathan G.K.
        • et al.
        Morphometric analysis of gastrocnemius muscle biopsies from patients with peripheral arterial disease: objective grading of muscle degeneration.
        Am J Physiol Regul Integr Comp Physiol. 2013; 305: R291-R299
        • Ha D.M.
        • Carpenter L.C.
        • Koutakis P.
        • et al.
        Transforming growth factor-beta 1 produced by vascular smooth muscle cells predicts fibrosis in the gastrocnemius of patients with peripheral artery disease.
        J Translat Med. 2016; 14: 39-51
        • Rock K.L.
        • Kono H
        The inflammatory response to cell death.
        Annu Rev Pathol. 2008; 3: 99-126
        • Takeuchi O.
        • Akira S.
        Pattern recognition receptors and inflammation.
        Cell. 2010; 140: 805-820
        • Fernando M.M.
        • Stevens C.R.
        • Walsh E.C.
        • et al.
        Defining the role of the MHC in autoimmunity: a review and pooled analysis.
        PLoS Genet. 2008; 4e1000024
        • Becher B.
        • Tugues S.
        • Greter M.
        GM-CSF: from growth factor to central mediator of tissue inflammation.
        Immunity. 2016; 45: 963-973
        • Kurien B.T.
        • Scofield R.H.
        Autoimmunity and oxidatively modified autoantigens.
        Autoimmun Rev. 2008; 7: 567-573
        • Calis J.J.
        • Maybeno M.
        • Greenbaum J.A.
        • et al.
        Properties of MHC class I presented peptides that enhance immunogenicity.
        PLoS Comput Biol. 2013; 9e1003266
        • Shen W.
        • Li Y.
        • Zhu J.
        • Schwendener R.
        • Huard J.
        Interaction between macrophages, TGF-beta1, and the COX-2 pathway during the inflammatory phase of skeletal muscle healing after injury.
        J Cell Physiol. 2008; 214: 405-412
        • Locati M.
        • Deuschle U.
        • Massardi M.L.
        • et al.
        Analysis of the gene expression profile activated by the CC chemokine ligand 5/RANTES and by lipopolysaccharide in human monocytes.
        J Immunol. 2002; 168: 3557-3562
        • Komuczki J.
        • Tuzlak S.
        • Friebel E.
        • et al.
        Fate-mapping of GM-CSF expression identifies a discrete subset of inflammation-driving T helper cells regulated by cytokines IL-23 and IL-1beta.
        Immunity. 2019; 50 (e6): 1289-1304
        • Gardner A.W.
        • Skinner J.S.
        • Cantwell B.W.
        • Smith L.K.
        Progressive vs single-stage treadmill tests for evaluation of claudication.
        Med Sci Sports Exerc. 1991; 23: 402-408
        • Montgomery P.S.
        • Gardner A.W.
        The clinical utility of a six-minute walk test in peripheral arterial occlusive disease patients.
        J Am Geriatr Soc. 1998; 46: 706-711
        • Myers S.A.
        • Johanning J.M.
        • Stergiou N.
        • Lynch T.G.
        • Longo G.M.
        • Pipinos II
        Claudication distances and the Walking Impairment Questionnaire best describe the ambulatory limitations in patients with symptomatic peripheral arterial disease.
        J Vasc Surg. 2008; 47: 550-555
        • Han G.
        • Li F.
        • Singh T.P.
        • Wolf P.
        • Wang X.J.
        The pro-inflammatory role of TGFbeta1: a paradox?.
        Int J Biol Sci. 2012; 8: 228-235
        • Meyer A.
        • Wang W.
        • Qu J.
        • et al.
        Platelet TGF-beta1 contributions to plasma TGF-beta1, cardiac fibrosis, and systolic dysfunction in a mouse model of pressure overload.
        Blood. 2012; 119: 1064-1074
        • Fava R.A.
        • Casey T.T.
        • Wilcox J.
        • Pelton R.W.
        • Moses H.L.
        • Nanney L.B.
        Synthesis of transforming growth factor-beta 1 by megakaryocytes and its localization to megakaryocyte and platelet alpha-granules.
        Blood. 1990; 76: 1946-1955
        • Hamilton J.A.
        • Anderson G.P.
        GM-CSF Biology.
        Growth Fact. 2004; 22: 225-231
        • Becher B.
        • Durell B.G.
        • Noelle R.J.
        Experimental autoimmune encephalitis and inflammation in the absence of interleukin-12.
        J Clin Invest. 2002; 110: 493-497
        • Shiomi A.
        • Usui T.
        Pivotal roles of GM-CSF in autoimmunity and inflammation.
        Mediators Inflamm. 2015; 2015568543
        • Wicks I.P.
        • Roberts A.W.
        Targeting GM-CSF in inflammatory diseases.
        Nat Rev Rheumatol. 2016; 12: 37-48
        • Goverman J.
        Autoimmune T cell responses in the central nervous system.
        Nat Rev Immunol. 2009; 9: 393-407
        • Lavine K.J.
        • Epelman S.
        • Uchida K.
        • et al.
        Distinct macrophage lineages contribute to disparate patterns of cardiac recovery and remodeling in the neonatal and adult heart.
        Proceed Natl Acad Sci. 2014; 111: 16029-16034
        • Franceschi C.
        • Garagnani P.
        • Vitale G.
        • Capri M.
        • Salvioli S
        Inflammaging and 'Garb-aging'.
        Trends Endocrinol Metab. 2017; 28: 199-212
        • Woods J.A.
        • Wilund K.R.
        • Martin S.A.
        • Kistler B.M.
        Exercise, inflammation and aging.
        Aging Dis. 2012; 3: 130-140
        • Kregel K.C.
        • Zhang H.J.
        An integrated view of oxidative stress in aging: basic mechanisms, functional effects, and pathological considerations.
        Am J Physiol Regul Integr Comp Physiol. 2007; 292: R18-R36
        • Gill R.
        • Tsung A.
        • Billiar T.
        Linking oxidative stress to inflammation: toll-like receptors.
        Free Radic Biol Med. 2010; 48: 1121-1132
        • Franceschi C.
        • Campisi J.
        Chronic inflammation (inflammaging) and its potential contribution to age-associated diseases.
        J Gerontol A Biol Sci Med Sci. 2014; 69: S4-S9
        • Bianchi M.E.
        DAMPs, PAMPs and alarmins: all we need to know about danger.
        J Leukoc Biol. 2007; 81: 1-5