https://orcid.org/0000-0003-1472-7975
Abstract
Arterial thrombosis as a result of plaque rupture or erosion is a key event in acute cardiovascular events. Sirtuin 5 (SIRT5) belongs to the lifespan-regulating sirtuin superfamily and has been implicated in acute ischaemic stroke and cardiac hypertrophy. This project aims at investigating the role of SIRT5 in arterial thrombus formation.
Sirt5 transgenic (Sirt5Tg/0) and knock-out (Sirt5−/−) mice underwent photochemically induced carotid endothelial injury to trigger arterial thrombosis. Primary human aortic endothelial cells (HAECs) were treated with SIRT5 silencing-RNA (si-SIRT5) as well as peripheral blood mononuclear cells from acute coronary syndrome (ACS) patients and non-ACS controls (case–control study, total n = 171) were used to increase the translational relevance of our data. Compared to wild-type controls, Sirt5Tg/0 mice displayed accelerated arterial thrombus formation following endothelial-specific damage. Conversely, in Sirt5−/− mice, arterial thrombosis was blunted. Platelet function was unaltered, as assessed by ex vivo collagen-induced aggregometry. Similarly, activation of the coagulation cascade as assessed by vascular and plasma tissue factor (TF) and TF pathway inhibitor expression was unaltered. Increased thrombus embolization episodes and circulating D-dimer levels suggested augmented activation of the fibrinolytic system in Sirt5−/− mice. Accordingly, Sirt5−/− mice showed reduced plasma and vascular expression of the fibrinolysis inhibitor plasminogen activator inhibitor (PAI)-1. In HAECs, SIRT5-silencing inhibited PAI-1 gene and protein expression in response to TNF-α. This effect was mediated by increased AMPK activation and reduced phosphorylation of the MAP kinase ERK 1/2, but not JNK and p38 as shown both in vivo and in vitro. Lastly, both PAI-1 and SIRT5 gene expressions are increased in ACS patients compared to non-ACS controls after adjustment for cardiovascular risk factors, while PAI-1 expression increased across tertiles of SIRT5.
SIRT5 promotes arterial thrombosis by modulating fibrinolysis through endothelial PAI-1 expression. Hence, SIRT5 may be an interesting therapeutic target in the context of atherothrombotic events.
1. Introduction
The formation of a blood clot in the arterial tree (i.e. arterial thrombosis) mainly results from rupture or erosion of an atherosclerotic plaque.1 Upon exposure of the plaque content, glycoprotein (GP) VI and integrin α2β1 mediate platelet adhesion and aggregation in response to collagenous plaque components and initiate arterial thrombosis.2 Then, activation of the coagulation cascade by exposure of vascular tissue factor (TF) to circulating factor VII (FVII) leads to thrombus formation and stabilization via thrombin-mediated deposition of fibrin.3 The sudden reduction in oxygen and energy supply causes the death of the parenchyma downstream of the arterial obstruction with irreversible organ damage as seen in most cardiovascular (CV) diseases such as myocardial infarction, ischaemic stroke, and peripheral artery disease. As CV diseases constantly rate among the top causes of death worldwide,4 greater efforts are needed in order to understand the pathophysiology of these disorders to describe new molecular targets potentially improving their outcome.5
Ageing is a major risk factor for all CV complications.6 Intriguingly, genes regulating lifespan (i.e. ageing and longevity genes) also play pivotal roles in the development of age-dependent vascular dysfunction and CV diseases.7–10 Specifically, the NAD+-dependent superfamily of proteins called sirtuins (SIRTs) are enzymes with functions in post-translational protein modification involved in several physiological and pathological processes, including ageing, energy control, and stress resistance (i.e. hypoxia and inflammation).11–13 Initially described as a mitochondrial sirtuin, Sirtuin 5 (SIRT5) was reported to regulate different metabolic pathways including Krebs cycle, glycolysis, urea cycle, and fatty acid oxidation.14 However, this protein was recently found to locate also in the cytoplasm where it regulates extra-mitochondrial processes.15,16 Differently from other sirtuins, SIRT5 holds a very weak deacetylase activity, being more efficient in removal of succinyl, malonyl, and glutaryl groups.17 The role of this sirtuin in pathophysiology of age-related CV diseases remains poorly characterized; we previously reported SIRT5 to mediate experimental cerebral ischaemia/reperfusion damage by increasing the blood–brain barrier permeability through occludin degradation.12 Similarly, SIRT5 appears to play a deleterious role also in the context of cardiac hypertrophy where Sirt5−/− animals show reduced left ventricular dilation and cardiac dysfunction after transverse aortic constriction.18 Lastly, SIRT5 seems to have protective effects in an experimental model of myocardial infarction, as animals lacking SIRT5 show increased infarct size after transient left anterior descending coronary ligation.19 Yet, whether SIRT5 plays a role in the formation of an arterial thrombus which—contrary to the clinical situation—does not occur in experimental myocardial infarction remains to be explored.
Hence, the present study aims at investigating the role of SIRT5 in arterial thrombosis. To this end, we assessed the effect of SIRT5 overexpression and knockout in an in vivo murine model of photochemical-induced arterial thrombosis. Importantly, this model is based on an endothelial-specific injury which is thought to mirror the endothelial damage underlying formation of an arterial thrombus in humans. Then, to increase the translational relevance of this work and further explore the molecular mechanism, we confirm the animal findings in SIRT5-depleted primary human aortic endothelial cells (HAECs). Finally, employing peripheral blood mononuclear cell (PBMC) as a surrogate cell, we evaluated the expression of SIRT5 in humans with and without acute coronary syndrome (ACS).
2. Methods
2.1 Animals
About 12-week-old SIRT5 knockout (Sirt5−/−) or transgenic (Sirt5Tg/0) male mice together with corresponding wild-type (WT) littermate controls all on C57BL/6 background were used for all experiments. Sirt5−/−and Sirt5Tg/0 mice were generated as described previously.12,20,21 All animals were kept in a temperature-controlled animal facility under normal light/dark cycle with free access to food and water. All procedures were approved by the Cantonal Veterinary Authority, Switzerland (ZH293/14). Animal experiments were performed conforming to the Directive 2010/63/EU of the European Parliament and of the Council of 22 September 2010 on the protection of animals used for scientific purposes.
2.2 In vivo carotid artery thrombosis model
Wild-type, Sirt5−/−, and Sirt5Tg/0 male mice underwent photochemical injury of the common carotid artery (CCA) as previously described.22–24 Briefly, mice were anaesthetized by i.p. injection of 87 mg/kg sodium pentobarbital (Butler, Columbus, OH, USA). The depth of anaesthesia was confirmed by the absence of twitch reflex. Rose Bengal (Fischer Scientific, Fair Lawn, NJ, USA) was diluted to 12 mg/mL in phosphate-buffered saline and then injected through the tail vein at a concentration of 63 mg/kg. Mice were placed in supine position under a dissecting microscope and the right CCA was exposed by a midline cervical incision. A Doppler flow probe (Model 0.5 VB, Transonic Systems, Ithaca, NY, USA) was applied and connected to a flowmeter (Model T106, Transonic Systems, Ithaca, NY, USA). About 5–10 min after Rose Bengal injection, a 1.5 mW green light laser (540 nm; Melles Griot, Carlsbad, CA, USA) was applied to the site of injury at a distance of 6 cm from the artery for 60 min or until thrombosis occurred. From the onset of injury, carotid blood flow and heart rate were continuously monitored up to 120 min. Occlusion was defined as blood flow <0.1 mL/min for at least 1 min. Cyclic flow variations were recorded and thrombus embolization defined as an increase of blood flow to >0.1 mL/min after previous decrease below said level lasting <1 min. At the end of the experiment, animals were euthanized by overdose of carbon dioxide.
2.3 Blood cell counts
Total blood cell count was performed on a ScilVet ABCplus (Horiba, Kyoto, Japan) using EDTA-anticoagulated blood.
2.4 Collagen-induced platelet aggregation in mice
For platelet aggregation studies, washed platelets were obtained from 3.8% citrate anticoagulated blood drawn terminally by cardiac puncture in animals deeply anesthetized with isoflurane (5%), as previously described.25 Thereafter, washed platelets were re-suspended in Thyrode’s buffer (134 mM NaCl; 0.34 mM Na2HPO4; 2.9 mM KCl; 12 mM NaHCO3; 20 mM Hepes; 5 mM glucose; 0.35% (w/v) bovine serum albumin; pH 7.0) and platelet counts were normalized to 200 000/µL. Platelets were activated with collagen (final concentration 5 µg/mL) and maximal aggregation (%), lag phase (s), and slope of aggregation (%/min) were assessed using light transmission aggregometry (APACT 4004 aggregometer, Haemochrom Diagnostica GmbH, Essen, Germany).
2.5 Coagulation assays
Plasma from citrated blood (3.2% citrate, 1/10) was extracted by 15 min centrifugation (2500 g at 4°C) and stored immediately at −80°C until analysis. Prothrombin time (PT) and activated partial thromboplastin time (aPTT) were assessed by an analyzer (Start4; Diagnostica Stago, Asnieres, France) using the appropriate reagents (Roche Diagnostics, Basel, Switzerland).
2.6 Tissue factor activity assay
TF activity was determined as previously described by colorimetric ACTICHROME® TF assay according to the manufacturer’s recommendation (Cat No. 846, American Diagnostica, Stamford, CT, USA).26,27 Aortas were lysed (50 mmol/L Tris–HCl, 100 mmol/L NaCl, 0.1% Triton X-100, pH 7.4), and total protein concentration was determined by Bradford protein assay according to the manufacturer’s recommendations (VWR Life Science AMRESCO, Solon, OH, USA). Arterial lysates were mixed with factor VIIa and X following the manufacturer’s instructions which leads to the conversion of factor X to Xa; factor Xa subsequently cleaves the chromogenic substrate SPECTROZYME FXa. Optical density of cleaved SPECTROZYME FXa was determined at 490 nm by Nanodrop 2000 Spectrophotometer (Thermo Scientific, Waltham, MA, USA) and subtracted from absorbance at 405 nm. Finally, TF (pM) content was calculated, according to a standard curve. Finally, TF concentration as detected by the colorimetric assay was normalized to the total protein content of the sample and expressed as pmol/g of total protein.
2.7 Plasma sampling for TF, TF pathway inhibitor, plasminogen activator inhibitor-1, and D-dimer
Blood was collected via intracardiac puncture and immediately mixed with EDTA. The EDTA-blood solution was then centrifuged for 15 min at 3000 g as previously described.28 Plasma was collected and snap-frozen in liquid nitrogen. Protein content was measured employing colorimetric enzyme-linked immunosorbent assays (ELISA) species-specific following the manufacturer instruction. Specifically, DY3178-05 was used for TF (R&D systems, Minneapolis, MN, USA), ab217776 was used for TF pathway inhibitor (TFPI; Abcam, Cambridge, UK), DY3828-05 was used for plasminogen activator inhibitor-1 (PAI-1) (R&D systems), and abx258705 was used for D-dimer (Abbexa, Cambridge, UK) following the manufacturers’ instructions.
2.8 Cell culture experiments
Primary HAECs (Lonza, Basel, Switzerland) were cultured as described.29 Briefly, adhering HAECs were grown to confluence in fibronectin-coated 75 cm2 flasks in endothelial growth medium (Cell Applications Inc., San Diego, CA, USA) supplemented with 10% foetal bovine serum (FBS). Cells were detached by using Tripsin/EDTA and reseeded in 12-multi-well plates (120 000/well). Cells were grown to 80% confluence before being transfected with SIRT5 small interfering RNA (siSIRT5; Santa Cruz, Dallas, TX, USA) or scrambled siRNA (siSCR; Microsynth, Balgach, Switzerland) for 6 h using the Lipofectamine® 3000 transfection kit according to manufacturer’s recommendations (Invitrogen, Carlsbad, CA, USA) as previously described.30 Next, cells were rendered quiescent for 24 h in medium containing 0.5% FBS before stimulation with tumour necrosis factor α (TNF-α; R&D systems) for 6 h. Cytotoxicity was assessed by a colorimetric assay to detect lactate dehydrogenase (LDH; Roche, Basel, Switzerland).
2.9 Western blotting
Protein expression was determined by western-blot analysis as previously described.31,32 Murine aortas and endothelial cells were lysed (Tris 50 mM, NaCl 150 mM, EDTA 1 mM, NaF 1 mM, DTT 1 mM, aprotinin 10 mg/mL, leupeptin 10 mg/mL, Na3VO4 0.1 mM, phenylmethylsulfonyl fluoride 1 mM, and NP-40 0.5%) and total protein concentration was determined according to the manufacturer’s recommendations (Bio-Rad Laboratories AG, Fribourg, Switzerland); 20–30 µg of total protein lysates were separated on an 8% or 10% SDS–PAGE before being transferred to a polyvinylidene fluoride membrane by wet transfer (Bio-Rad Laboratores AG, Fribourg, Switzerland). Membranes were incubated with primary antibodies against SIRT5 (No. 8779S; 1:1000; Cell Signalling), TFPI (No. ADG72; 1:2000; American Diagnostica, Pfungstadt, Germany), PAI-1 (No. sc5297, 1:1000, Santa Cruz), phospho-AMPKα (No. 2535S; 1:1000, Cell Signaling), AMPKα (No. 5832 T; 1:1000, Cell Signaling), phospho-ERK (No. 4060S; 1:1000, Cell Signaling), ERK (No. 9272S; 1:1000, Cell Signaling), phospho-p38 (No. 9211S, 1:1000, Cell Signaling), p38 (No. 9212S, 1:1000, Cell Signaling), phospho-Janus kinases (JNK; No. 9251S, 1:1000, Cell Signaling), JNK (No. 9252S, 1:1000, Cell Signaling), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (No. MAB374; 1:40 000; Merck Millipore, Billerica, MA, USA) over-night at 4°C on a shaker. Secondary antibodies anti-mouse (No. 1031-05; Southern Biotech) and anti-rabbit (No. 4050-05; Southern Biotech) were obtained from Southern Biotechnology (Birmingham, AL, USA) and applied for 1 h at room temperature. Densitometric analyses were performed (Amersham Imager 600, General Electric; Healthcare Europe GmbH, Glattbrugg, Switzerland) and protein expression was normalized to GAPDH (for SIRT5, TFPI, and PAI-1) or total protein (for phosphorylated proteins).
2.10 Study population
The study cohort consisted of two subgroups of consecutively recruited subjects over the same time period: (i) a community-based subgroup of 120 individuals, recruited among the general population of the city of Athens, Greece with no acute or chronic clinically overt CV disease (non-ACS group, controls) and (ii) a subgroup of 51 patients with ACS recruited from the cardiac care unit of the Alexandra Hospital, Athens, Greece within 12 h of symptom onset. Exclusion criteria for both groups included autoimmune disease, cancer, acute renal failure, acute stroke, chronic inflammatory disease, or active infection.
Baseline characteristics of the cohort and an extensive CV history were obtained (Table 1). History of CV disease was confirmed by a previous coronary angiography or definite positive functional stress test for myocardial ischaemia or history of hospital admission for previous ACS. Current ACS was confirmed by ECG criteria, levels of cardiac enzymes, and coronary angiography at the acute phase (Supplementary material online, Table S1). Arterial hypertension was defined as blood pressure (BP) >140/90 mmHg or history of medical treatment with antihypertensive medications. Blood pressure was measured twice at resting position. Diabetes mellitus was defined according to the latest criteria as fasting plasma glucose (FPG) ≥126 mg/dL (excluding hospital admission period) or glycated haemoglobin ≥6.5% or treatment with anti-diabetic drugs. Hyperlipidaemia was defined according to the lipid profile or history of hypolipidemic treatment in patients without ACS. All individuals provided written informed consent, and the local ethics committee approved the protocol. This investigation conforms to the principles outlined in the Declaration of Helsinki.
Descriptive characteristics of the study cohort
| Study cohort | Non-ACS | ACS | P-value* | |
|---|---|---|---|---|
| (n = 171) | (n = 120) | (n = 51) | ||
| Age, years (mean ± SD) | 58.1 ± 13.2 | 57.7 ± 13.4 | 59.9 ± 11.6 | 0.311 |
| Male gender, n (%) | 97 (56.7) | 51 (42.5) | 46 (90.2) | <0.001 |
| Smoking | 81 (47.4) | 44 (36.7) | 37 (72.5) | <0.001 |
| Type 2 diabetes mellitus | 22 (12.9) | 15 (12.5) | 7 (13.7) | 0.807 |
| Dyslipidaemia | 84 (49.1) | 51 (42.5) | 33 (64.7) | 0.012 |
| Arterial hypertension | 73 (42.7) | 45 (37.5) | 28 (54.9) | 0.043 |
| Creatinine, mg/dL [median (IQR)] | 0.80 (0.36) | 0.75 (0.40) | 0.87 (0.27) | 0.070 |
| CRP, mg/L [median (IQR)] | 1.6 (3.8) | 1.0 (2.3) | 4.8 (7.9) | <0.001 |
| Medication at admission | ||||
| Aspirin | 16 (9.4) | 3 (2.5) | 13 (25.5) | <0.001 |
| Statins | 51 (29.8) | 37 (30.8) | 14 (27.5) | 0.718 |
| RAAS-inhibitors | 42 (24.6) | 29 (24.2) | 13 (25.5) | 0.848 |
| Beta-blockers | 28 (16.4) | 17 (14.2) | 11 (21.6) | 0.261 |
| CCBs | 24 (14.0) | 22 (18.3) | 2 (3.9) | 0.015 |
| Nitrates | 1 (0.6) | 0 (0) | 1 (1.96) | 0.298 |
| Diuretics | 16 (9.4) | 15 (12.5) | 1 (1.96) | 0.041 |
| Study cohort | Non-ACS | ACS | P-value* | |
|---|---|---|---|---|
| (n = 171) | (n = 120) | (n = 51) | ||
| Age, years (mean ± SD) | 58.1 ± 13.2 | 57.7 ± 13.4 | 59.9 ± 11.6 | 0.311 |
| Male gender, n (%) | 97 (56.7) | 51 (42.5) | 46 (90.2) | <0.001 |
| Smoking | 81 (47.4) | 44 (36.7) | 37 (72.5) | <0.001 |
| Type 2 diabetes mellitus | 22 (12.9) | 15 (12.5) | 7 (13.7) | 0.807 |
| Dyslipidaemia | 84 (49.1) | 51 (42.5) | 33 (64.7) | 0.012 |
| Arterial hypertension | 73 (42.7) | 45 (37.5) | 28 (54.9) | 0.043 |
| Creatinine, mg/dL [median (IQR)] | 0.80 (0.36) | 0.75 (0.40) | 0.87 (0.27) | 0.070 |
| CRP, mg/L [median (IQR)] | 1.6 (3.8) | 1.0 (2.3) | 4.8 (7.9) | <0.001 |
| Medication at admission | ||||
| Aspirin | 16 (9.4) | 3 (2.5) | 13 (25.5) | <0.001 |
| Statins | 51 (29.8) | 37 (30.8) | 14 (27.5) | 0.718 |
| RAAS-inhibitors | 42 (24.6) | 29 (24.2) | 13 (25.5) | 0.848 |
| Beta-blockers | 28 (16.4) | 17 (14.2) | 11 (21.6) | 0.261 |
| CCBs | 24 (14.0) | 22 (18.3) | 2 (3.9) | 0.015 |
| Nitrates | 1 (0.6) | 0 (0) | 1 (1.96) | 0.298 |
| Diuretics | 16 (9.4) | 15 (12.5) | 1 (1.96) | 0.041 |
Continuous variables were compared using independent samples t-test or Mann–Whitney U test based on distribution. Categorical variables were compared using Fisher’s exact χ2 test. Bold font indicates statistical significance.
ACS, acute coronary syndrome; SD, standard deviation; IQR, inter-quartile range; CRP, C-reactive protein; RAAS, renin–angiotensin–aldosterone system; CCBs, calcium-channel blockers.
P-value for comparison between non-ACS individuals and ACS patients. Continuous variables are presented as mean±SD or median (IQR) when normal distribution did not apply. Categorical variables are presented as absolute count (percentage).
Descriptive characteristics of the study cohort
| Study cohort | Non-ACS | ACS | P-value* | |
|---|---|---|---|---|
| (n = 171) | (n = 120) | (n = 51) | ||
| Age, years (mean ± SD) | 58.1 ± 13.2 | 57.7 ± 13.4 | 59.9 ± 11.6 | 0.311 |
| Male gender, n (%) | 97 (56.7) | 51 (42.5) | 46 (90.2) | <0.001 |
| Smoking | 81 (47.4) | 44 (36.7) | 37 (72.5) | <0.001 |
| Type 2 diabetes mellitus | 22 (12.9) | 15 (12.5) | 7 (13.7) | 0.807 |
| Dyslipidaemia | 84 (49.1) | 51 (42.5) | 33 (64.7) | 0.012 |
| Arterial hypertension | 73 (42.7) | 45 (37.5) | 28 (54.9) | 0.043 |
| Creatinine, mg/dL [median (IQR)] | 0.80 (0.36) | 0.75 (0.40) | 0.87 (0.27) | 0.070 |
| CRP, mg/L [median (IQR)] | 1.6 (3.8) | 1.0 (2.3) | 4.8 (7.9) | <0.001 |
| Medication at admission | ||||
| Aspirin | 16 (9.4) | 3 (2.5) | 13 (25.5) | <0.001 |
| Statins | 51 (29.8) | 37 (30.8) | 14 (27.5) | 0.718 |
| RAAS-inhibitors | 42 (24.6) | 29 (24.2) | 13 (25.5) | 0.848 |
| Beta-blockers | 28 (16.4) | 17 (14.2) | 11 (21.6) | 0.261 |
| CCBs | 24 (14.0) | 22 (18.3) | 2 (3.9) | 0.015 |
| Nitrates | 1 (0.6) | 0 (0) | 1 (1.96) | 0.298 |
| Diuretics | 16 (9.4) | 15 (12.5) | 1 (1.96) | 0.041 |
| Study cohort | Non-ACS | ACS | P-value* | |
|---|---|---|---|---|
| (n = 171) | (n = 120) | (n = 51) | ||
| Age, years (mean ± SD) | 58.1 ± 13.2 | 57.7 ± 13.4 | 59.9 ± 11.6 | 0.311 |
| Male gender, n (%) | 97 (56.7) | 51 (42.5) | 46 (90.2) | <0.001 |
| Smoking | 81 (47.4) | 44 (36.7) | 37 (72.5) | <0.001 |
| Type 2 diabetes mellitus | 22 (12.9) | 15 (12.5) | 7 (13.7) | 0.807 |
| Dyslipidaemia | 84 (49.1) | 51 (42.5) | 33 (64.7) | 0.012 |
| Arterial hypertension | 73 (42.7) | 45 (37.5) | 28 (54.9) | 0.043 |
| Creatinine, mg/dL [median (IQR)] | 0.80 (0.36) | 0.75 (0.40) | 0.87 (0.27) | 0.070 |
| CRP, mg/L [median (IQR)] | 1.6 (3.8) | 1.0 (2.3) | 4.8 (7.9) | <0.001 |
| Medication at admission | ||||
| Aspirin | 16 (9.4) | 3 (2.5) | 13 (25.5) | <0.001 |
| Statins | 51 (29.8) | 37 (30.8) | 14 (27.5) | 0.718 |
| RAAS-inhibitors | 42 (24.6) | 29 (24.2) | 13 (25.5) | 0.848 |
| Beta-blockers | 28 (16.4) | 17 (14.2) | 11 (21.6) | 0.261 |
| CCBs | 24 (14.0) | 22 (18.3) | 2 (3.9) | 0.015 |
| Nitrates | 1 (0.6) | 0 (0) | 1 (1.96) | 0.298 |
| Diuretics | 16 (9.4) | 15 (12.5) | 1 (1.96) | 0.041 |
Continuous variables were compared using independent samples t-test or Mann–Whitney U test based on distribution. Categorical variables were compared using Fisher’s exact χ2 test. Bold font indicates statistical significance.
ACS, acute coronary syndrome; SD, standard deviation; IQR, inter-quartile range; CRP, C-reactive protein; RAAS, renin–angiotensin–aldosterone system; CCBs, calcium-channel blockers.
P-value for comparison between non-ACS individuals and ACS patients. Continuous variables are presented as mean±SD or median (IQR) when normal distribution did not apply. Categorical variables are presented as absolute count (percentage).
2.11 Isolation of PBMCs
Peripheral blood from consenting individuals was collected in EDTA-tubes (BD Vacutainer) and PBMCs were isolated by ficoll density gradient centrifugation. Isolated PBMCs were washed twice with phosphate-buffered saline and subsequently lysed in Trizol and stored at −80°C until further processing, as previously described.33
2.12 Quantitative real-time polymerase chain reaction
Total RNA was extracted from patient samples with the Direct-zol RNA MiniPrep kit (Zymo Research) including a DNase digestion step according to manufacturer’s instructions. One microgram (1 μg) of total RNA was reverse transcribed into cDNA with the use of MLV reverse transcriptase kit (Invitrogen, Thermo Fisher Scientific) according to manufacturer’s instructions. For the quantification of PAI-1 (SERPINE) expression a SYBR-green-based assay was used (PowerUp SYBR Green Master Mix, Applied Biosystems) with the following set of primers: forward: 5′-GGGCCATGGAACAAGGATGA-3′; reverse: 5′-CTCCTTTCCCAAGCAAGTTG-3′. RPLP0 and ACTB served as the housekeeping genes (RPLP0: forward: 5′-TCGACAATGGCAGCATCTAC-3′; reverse: 5′-ATCCGTCTCCACAGACAAGG-3′; ACTB: forward: 5′-GCA CAG AGC CTC GCC TT-3′ and reverse: 5′-GTT GTC GAC GAC GAG CG-3′). A melting curve analysis was performed after amplification to verify the accuracy of the amplicon. For the quantification of SIRT5 mRNA levels, we used a predesigned TaqMan Gene Expression Assay (Hs00978328_m1, Applied Biosystems), while HPRT1 (Hs02800695_m1) was used as the housekeeping gene. RT–PCR was performed in a QuantStudio 7 Flex RT–PCR cycler (Applied Biosystems, Foster City, CA, USA) according to the manufacturer’s instructions. The relative mRNA expression levels were determined using the formula 2−ΔCt, where ΔCt = Ct(gene)−Ct(housekeeping gene) and Graph Pad Prism 6 software (GraphPad Software, Inc, La Jolla, CA, USA).
2.13 Statistical analysis
Statistical analysis was performed using SPSS v.24 and GraphPad Prism 6 software. Experimental data were analysed by using one-way analysis of variance (ANOVA) with Bonferroni’s post hoc test for multiple comparisons or unpaired two-tailed Student’s t-test where appropriate. For the clinical study, the distribution of variables was assessed by Kolmogorov–Smirnov and Shapiro–Wilk tests of normality. To compare continuous variables between two groups, we used independent samples t-test or Mann–Whitney U test depending on data distribution, while χ2 test was used to compare categorical variables. We used ANCOVA to control for the effect of traditional CV risk factors (i.e. sex, age, arterial hypertension, hyperlipidaemia, type 2 diabetes mellitus, and smoking) on associations observed with SIRT5 and PAI-1. Statistical significance was set at probability value P < 0.05 for all experiments.
3. Results
3.1 Sirt5 overexpression facilitates thrombus formation and increases thrombus firmness in vivo
To assess the effects of Sirt5 overexpression on thrombosis, time to arterial occlusion was analysed in Sirt5Tg/0 mice and WT littermates following photochemical-induced carotid endothelial injury. Time to stable thrombotic occlusion was significantly accelerated in Sirt5Tg/0 animals as compared to WT littermates (P < 0.001, Figure 1A). A representative trace of mean blood flows from start of the experiments to carotid occlusion is reported for the two groups (Figure 1B). The irregular pattern of the traces reflects the dissolving action of the fibrinolytic system on the forming thrombus. Embolization of a formed thrombus and blood flow restoration >0.1 mL/min occurred less often in Sirt5Tg/0 mice than in controls (P < 0.05, Figure 1C). Importantly, no significant differences in initial blood flow, heart rate, and body weight were observed between the two groups (Figure 1D–F).
Impact of Sirt5 overexpression on photochemical-induced arterial thrombosis in vivo in mice. (A) Sirt5Tg/0 mice showed reduced time to formation of an occlusive thrombus in their carotid arteries after endothelial-specific damage as compared to Sirt5+/+ littermates. (B) Representative trace of mean blood flow until occlusion (mean flow ≤ 0.1 mL for 1 min) in the two study groups. (C) Sirt5 overexpression reduced the number of episodes of thrombus embolization. (D–F) No difference in terms of initial heart rate, blood flow, and weight were reported among Sirt5Tg/0 and wild-type littermates. n = 7–8 different mice per group. (A and C–F): unpaired two-tailed Student’s t-test. *P < 0.05; ***P < 0.001.
3.2 Loss of Sirt5 delays time to arterial thrombotic occlusion in vivo
To investigate the potential role of Sirt5 as a therapeutic target in the context of arterial thrombosis, thrombus formation was analysed in carotid arteries of Sirt5−/− mice and WT littermates. Compared to WT controls, the formation of a stable occluding thrombus in Sirt5−/− animals was delayed by 2.14-fold (P < 0.001, Figure 2A). Opposite to what was reported for Sirt5Tg/0 animals, the analysis of cyclic blood flow variation (Figure 2B) showed that thrombus embolization occurred more frequently in Sirt5−/− mice than in controls (P < 0.05, Figure 2C). Again, no significant differences in initial blood flow, heart rate, and body weight were observed between the two groups (Figure 2D–F).
Effects of Sirt5 deletion on carotid arterial thrombosis in vivo in mice. (A) Sirt5−/− mice showed delayed time to formation of an occlusive thrombus in their carotid arteries after endothelial-specific damage as compared to Sirt5+/+ littermates. (B) Representative trace of mean blood flow until occlusion (mean flow ≤ 0.1 mL for 1 min) in the two study groups. (C) Sirt5 deletion increased the number of episodes of thrombus embolization. (D–F) No difference in terms of initial heart rate, blood flow, and weight were reported among Sirt5−/− and wild-type littermates. n = 8 different mice per group. (A and C–F): unpaired two-tailed Student’s t-test. *P < 0.05; ***P < 0.001.
3.3 Sirt5 deficiency does not affect platelet count, volume, and reactivity to collagen ex vivo
In order to investigate the potential mechanisms underlying the protective effect of Sirt5 deletion in arterial thrombosis, we investigated whether the deficiency of this protein affects platelet count and function. Complete blood count analysis revealed that platelet numbers and volumes were unchanged in Sirt5−/− and WT mice (Figure 3A and B, respectively). In addition, reactivity of washed platelets to collagen was assessed by light transmission aggregometry. Sirt5 deficiency did not affect collagen-induced platelet reactivity as demonstrated by similar maximal aggregation, rate (slope) of aggregation and lag phase in Sirt5−/− animals, and WT littermates (Figure 3C–E, respectively).
Impact of Sirt5 deletion on platelet count, volume, and ex vivo aggregation. (A and B) Sirt5−/− animals did not show any difference in terms of platelet count or mean platelet volume as compared to Sirt5+/+ animals. (C–E) Ex vivo collagen-induced platelet aggregation as assessed by light transmission aggregometry did not show any difference in terms of maximal aggregation, rate (slope) of aggregation and lag phase in Sirt5−/− animals, and WT littermates. n = 5 different mice per group. (A–E): unpaired two-tailed Student’s t-test.
3.4 Loss of Sirt5 does not affect coagulation times, vascular, and plasma TF expression and activity
Together with platelets, activation of the coagulation cascade by TF triggers the formation of an arterial thrombus after endothelial damage. Thus, we evaluated the function of extrinsic and intrinsic coagulation pathways by measuring aPTT and PT, respectively. Sirt5−/− and WT mice did not show any difference in term of aPTT and PT (Figure 4A and B, respectively).
Effects of Sirt5 deletion on coagulation pathways and fibrinolytic system. (A) aPTT was similar in plasma from Sirt5−/− and Sirt5+/+ mice. (B) No difference in term of PT were noted between Sirt5−/− and Sirt5+/+ mice. (C) Plasma levels of TF did not differ between Sirt5−/− and Sirt5+/+ mice. (D) Similarly, also arterial TF activity did not differ among the two study groups. (E and F) No difference in TFPI plasma and vascular levels as detected between Sirt5−/− mice and wild-type littermates. (G) Plasma D-dimer concentration was significantly increased in Sirt5−/− animals as compared to Sirt5+/+ littermates (H) Sirt5−/− animals exhibited reduced plasma PAI-1 levels as compared to wild-type littermates. (I) Similarly, also arterial PAI-1 protein expression was blunted in animal lacking Sirt5. (J) Phosphorylation levels of AMPK subunity α were lower in Sirt5−/− animals than in wild-type controls. (K) Phosphorylation levels of the MAP kinases ERK1/2 were higher in Sirt5−/− animals than in wild-type controls. n = 5–6 different mice per group. (A–K): unpaired two-tailed Student’s t-test. *P < 0.05. aPTT, activated partial thromboplastin time; AU, arbitrary unit; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; ERK, extracellular signal-regulated kinases; PAI-1, plasminogen activator inhibitor; PT, prothrombin time; TF, tissue factor; TFPI, tissue facto pathway inhibitor.
Accordingly, Sirt5−/− and WT animals displayed similar levels of circulating TF (Figure 4C). Similarly, also TF activity in aortic lysates of animals lacking Sirt5 remained unchanged as compared to WT littermates (Figure 4D). TFPI is the major physiological inhibitor of TF and thus of the extrinsic coagulation cascade.34 Here, we analysed both plasma and vascular pools of this protein and found no differences between Sirt5−/− and WT animals (Figure 4E and F, respectively).
3.5 Sirt5 deletion associates with increased fibrinolysis via reduced PAI-1 expression
Fibrin is the end product of the coagulation cascade, its lysis into fibrin degradation products (such as D-dimer) has the fundamental homeostatic function to reduce thrombus stability and facilitate its dissolution after healing. According to the reduced thrombus firmness highlighted by the higher number of thrombus embolization episodes in Sirt5−/− animals, those animals also showed increased activation of the fibrinolytic system as showed by higher plasma levels of D-dimer as compared to control littermates (Figure 4G).
Plasmin is the major fibrinolysin and is enzymatically activated from plasminogen by tissue or urokinase-type plasminogen activators (tPA and uPA, respectively). Both tPA and uPA are rapidly inhibited by PAI-1, which is mainly produced by vascular endothelial cells and released into the circulation. Here, we found that Sirt5−/− animals have blunted plasma circulating levels of this pro-coagulant factor as compared to WT mice (P < 0.05, Figure 4H). Accordingly, we also found that the arterial tissue PAI-1 expression was reduced in mice lacking Sirt5 as compared to controls (P < 0.05, Figure 4I). On the opposite, animal overexpressing this sirtuins showed increase arterial PAI-1 levels (P < 0.05, see Supplementary material online, Figure S1). AMPK is a ubiquitous protein kinase that mediates cellular response to nutritional and environmental stress by regulating several key genes among which PAI-1. WB analysis of arterial tissue displayed increased phosphorylation of AMPK sub-unit α denoting an increased activation of this metabolic pathway in Sirt5−/− animals compared to Sirt5+/+ mice (P < 0.05, Figure 4J). Further, endothelial PAI-1 expression can be induced by different stimuli including different cytokines and hypoxia through activation of different mitogen-activated protein (MAP) kinases. Accordingly, Sirt5−/− animals showed blunted phosphorylation levels of the MAP kinase ERK1/2 than WT littermates (P < 0.05, Figure 4K).
3.6 Silencing of SIRT5 reduces PAI-1 expression in stimulated primary HAECs
To test the translational relevance of our in vivo findings, PAI-1 expression was assessed in SIRT5-silenced primary HAECs under basal conditions and after stimulation with TNF-α. SIRT5 silencing was achieved in HAECs by transfection with SIRT5 siRNA (siSIRT5) and compared to siSCR-treated cells (P < 0.01, Figure 5A). As expected, stimulation with TNF-α increased the expression of PAI-1 mRNA in endothelial cells treated with the control siRNA (P < 0.05, Figure 5B). Interestingly, treatment with TNF-α did not induce PAI-1 mRNA expression in SIRT5-silenced HAECs thus, transcript levels remained similar to those of unstimulated cells and significantly lower than those of stimulated control HAECs (P < 0.05 for all, Figure 5B). Likewise, treatment with TNF-α increased PAI-1 protein expression in control HAECs, while this effect was lost in SIRT5-silenced cells (P < 0.05, Figure 5C). As a result and similarly to what was observed after thrombosis in vivo, endothelial PAI-1 protein levels after TNF-α stimulation were significantly lower in siSIRT5-treated as compared to siSCR cells (P < 0.05, Figure 5C).
Effects of SIRT5 silencing on TNF-α-mediated PAI-1 expression in primary human aortic endothelial cells. (A) In HAECs, SIRT5 protein expression is significantly reduced after transfection with SIRT5 small interfering RNA (siSIRT5) as compared to control siRNA (siSCR). (B) PAI-1 mRNA levels are induced in siSCR HAECS by treatment with 10 ng/mL TNF-α, this effect is lost upon SIRT5 silencing. (C) Accordingly, PAI-1 protein levels were increased in stimulated siSCR-treated cells while remaining similar to baseline in stimulated SIRT5-silenced HAECs. (D) Phosphorylation levels of the AMPK activator sub-unity α were higher in stimulated SIRT5-silenced HAECs than in control cells. (E–G) MAP kinases phosphorylation was induced by TNF-α in siSCR-treated control cells, in SIRT5-silenced HAECs this effect was specifically lost for ERK1/2 while not for p38 and JNK. n = 8–10 independent experiments. (A) unpaired two-tailed Student’s t-test; (B–G): one-way analysis of variance (ANOVA) with Bonferroni post hoc test. *P < 0.05, **P < 0.01. ERK1/2, extracellular signal-regulated kinases; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; HAECs, human aortic endothelial cells; JNK, c-Jun N-terminal kinases; MAP, mitogen-activated protein; PAI-1, plasminogen activator inhibitor 1; TNF-α, tumour necrosis factor α.
Mirroring the animal findings, stimulated siSIRT5-treated cells showed increased activation of AMPK pathways as underlined by augmented phosphorylation of its α sub-unit as compared to HAECs treated with siSCR (P < 0.05 for siSCR vs. siSIRT5 TNF-α and P < 0.01 for siSCR TNF-α vs. siSIRT5 TNF-α; Figure 5D). In endothelial cells, stimulation with TNF-α triggers the activation of different MAPK pathways.35 Accordingly, TNF-α increased the phosphorylation of ERK1/2, p38, and JNK in siSCR-treated control cells (P < 0.05 for all, Figure 5E–G, respectively). On the opposite, after stimulation with TNF-α, p-ERK1/2 levels remained unaltered in SIRT5-silenced HAECs as compared to unstimulated ones (Figure 5E). Interestingly, this effect was specific for the ERK1/2 pathway, as it was not observed for p38 and JNK (Figure 5F and G).
3.7 Sirt5 expression in PBMCs from patients with ACS is increased and correlates with PAI-1
To increase the translational relevance of our data, we analysed SIRT5 and PAI-1 mRNA levels in PBMCs of patients suffering from an ACS and compared to those of a control group. In light of the fact that endothelial cells could not be easily obtained, we chose PBMCs as a surrogate cell type. A total of 171 patients were included in the analysis, 120 non-ACS control subjects and 51 ACS patients. As expected, rates of traditional CV risk factors such as smoking habits, arterial hypertension, and hyperlipidaemia were higher in ACS patients as compared to non-ACS controls (Table 1). SIRT5 expression was significantly increased in PBMCs from ACS patients as compared to non-ACS controls (P = 0.03). After adjusting for CV risk factors, SIRT5 remained significantly higher in the ACS group compared to non-ACS individuals (adjusted P = 0.009, Figure 6A). Although PAI-1 levels were not significantly increased in ACS in the univariate model, after adjusting for traditional CV risk factors PAI-1 expression was significantly associated with presence of ACS (adjusted P = 0.004, Figure 6B). Of importance, PAI-1 levels significantly increased across tertiles of SIRT5 expression in the whole cohort demonstrating the biological relevance of SIRT5–PAI-1 regulatory axis in humans (Figure 6C).
SIRT5 and PAI-1 gene expression in patients with ACS. (A) SIRT5 mRNA expression is increased in PBMCs from ACS patients as compared to non-ACS controls. (B) Similarly, also PAI-1 expression is increased in patients with ACS as compared to non-ACS. (C) PAI-1 transcript levels were found to be increased across tertiles of SIRT5 in the whole cohort. n = 120 patients for non-ACS and n = 51 patients for ACS. (A and B): Analysis of covariance (ANCOVA); (C): P-value adjusted by the Dunn’s correction for multiple comparisons. ACS, acute coronary syndrome; PAI-1, plasminogen activator inhibitor 1; PBMCs, peripheral blood mononuclear cells.
4. Discussion
Here, we demonstrate for the first time that SIRT5 plays a deleterious role in the pathophysiology of arterial thrombosis and that it holds potential as a novel therapeutic target in this context. Translational experimental data obtained in rodents and different human cells, substantiate our conclusions: (i) carotid arterial thrombosis following endothelial-specific damage is enhanced in transgenic mice overexpressing Sirt5 as compared to WT littermates; (ii) conversely, in animals lacking Sirt5 expression the formation of a stable occluding thrombus is delayed and occurs only after several thrombus formation and lysis cycles; (iii) loss of Sirt5 specifically increases fibrinolytic activity without affecting the activation of platelets to collagen or the coagulation cascade; this effect is mediated by reduced plasma and tissue levels of the fibrinolysis inhibitor PAI-1 through modulation of AMPK and ERK1/2 activation; (iv) SIRT5 silencing impairs TNFα-induced PAI-1 expression in primary HAECs at the transcriptional level by increasing AMPK activation and specifically blunting ERK1/2 phosphorylation, but not that of p38 and JNK; (v) SIRT5 gene expression is increased in PBMCs from patients with ACS, while PAI-1 mRNA expression increases across tertiles of SIRT5.
Arterial thrombosis underlies most of the acute ischaemic complications of age-dependent CV diseases whose incidence is set to increase in the coming decades.5 The identification of novel therapeutic targets remains of utmost importance and we have previously shown that different genes regulating life span are also involved in the pathophysiology of age-dependent CV afflictions and can act as effective molecular targets for their prevention and treatment.36,37 In the present study, we investigated SIRT5, a gene belonging to the lifespan-regulating sirtuin superfamily whose role in arterial thrombosis remains undetermined. As for other sirtuins, also SIRT5 has been previously linked to lifespan regulation as well as to different age-dependent disease.6 Here, we report a deleterious role for Sirt5 in the pathophysiology of arterial thrombosis after endothelial-specific damage by using two different genetically modified animal models. Specifically, we reported reduced time to thrombotic carotid occlusion for Sirt5Tg/0 mice and delayed thrombosis in those lacking Sirt5.
Arterial thrombosis occurs as a result of the interplay between components of the vascular wall, platelets, and coagulation cascade. Early after endothelial damage, platelets recognize collagenous plaque components via GP VI and integrin α2β1, get activated, and aggregate at the site of injury thus initiating the thrombus formation.38 To investigate the mechanism underlying the different thrombosis dynamics in Sirt5−/− animals, we then characterized collagen-induced platelet aggregation. Sirt5−/− platelets did not differ from WT ones in terms of count and mean volume. In addition, ex vivo collagen-induced aggregation of Sirt5−/− and Sirt5+/+ platelets showed very similar dynamics as underlined by similar maximal aggregation, slope of aggregation, and lag phase in light transmission aggregometry. Although a role for other sirtuins (Sirt1, 2, and 3) in regulation of platelet function was previously reported,39,40 to the best of our knowledge this is the first study specifically investigating Sirt5. Soon after platelet aggregation, the thrombus is then stabilized by the formation of thrombin and fibrin as the result of activation of the extrinsic coagulation cascade by TF.38 Accordingly, in this investigation, we sought to address the different thrombotic responses by examining the extrinsic and intrinsic coagulation pathways. Unexpectedly, we did not observe any difference between the two groups. Fibrinolysis constitutes the third haemostatic pathway. The degradation of fibrin polymers due to the plasmin enzymatic action contributes to clot removal and blood flow restoration with production of fibrin degradation products such as D-dimer.41 As such, the fibrinolytic system physiologically antagonizes stable thrombus formation.42 In this work, Sirt5Tg/0 and Sirt5−/− animals showed, respectively, reduced and increased frequency of thrombus embolization during photochemically induced carotid thrombosis thus pointing towards fibrinolysis as a possible mediator of the different thrombosis times recorded. Accordingly, sampling of Sirt5−/− plasma after thrombosis revealed increased fibrinolytic activity as underlined by higher levels of D-dimer. The fibrinolytic system is tightly regulated by PAI-1 which inhibits plasmin activators thus playing a pro-coagulant role. Coherently to the delayed time to carotid occlusion, deletion of Sirt5 associated to reduced vascular and plasma levels of the fibrinolysis inhibitor PAI-1.
Endogenous fibrinolysis represents an expanding research field for CV risk stratification and is more and more seen as a potential pharmacological target to improve CV outcome.43 Mirroring the in vivo data, SIRT5-silencing reduced PAI-1 expression also in vitro in stimulated primary HAECs by modulating its gene expression. With respect to the mechanisms underlying the observed effects, both in vivo and in vitro data consistently point towards AMPK and ERK1/2 as possible mediators. Within cells, Sirt5 resides primarily in the mitochondria where it regulates several metabolic pathways by post-translational protein modification.17 Loss of SIRT5 was previously shown to increase mitochondrial NADH, suppress ATP production, and increase AMP/ATP ratio in human embryonic kidney cells and mouse hearts under stress conditions.18 Under Sirt5-mediated ATP depletion, cytoplasmic AMPK—among the better described intracellular energetic sensors—undergoes allosteric transition and is phosphorylated at Thr172 by upstream kinases leading to its activation.18,44 Accordingly, we could demonstrate an increased phosphorylation of AMPKα indicating an activation of AMPK pathway in vivo in Sirt5−/− arterial lysates as well as in vitro in stimulated SIRT5-silenced HAECs. Recently, different investigations reported AMPK activation to directly regulate PAI-1 expression by inducing small heterodimer partner and repressing NF-κΒ activation in hepatocytes and endothelial cells.45,46 Furthermore, AMPK was shown to reduce the activation of ERK1/2 pathway, a known regulator of PAI-1 transcription in response to different stimuli.47–49 In the present study, in line with reduced PAI-1 levels, we reported blunted ERK1/2 phosphorylation both in arteries of Sirt5−/− animals as well as in SIRT5-silenced HAECs stimulated with TNF-α. Of interest, the decreased phosphorylation was specific for ERK1/2 as both MAP kinases p38 and JNK remained unaltered.
Lastly, in an exploratory human investigation, PBMCs from whole blood were selected as a surrogate cell type to investigate the relationship between SIRT5 and PAI-1. Interestingly, SIRT5 mRNA expression levels were found to be increased in ACS patients as compared to non-ACS controls. This finding was reinforced by adjustment for traditional CV risk factors such as male sex, age, arterial hypertension, hyperlipidaemia, type 2 diabetes mellitus, and smoking. Although few DNA sequence variants (DSVs) within the SIRT5 regulatory regions were recently identified in patients suffering from acute myocardial infarction which may alter the expression of this gene,50 to the best of our knowledge this is the first research study showing a significant up-regulation of such a gene in patients with an ACS. Alongside SIRT5, also PAI-1 levels were significantly associated with ACS incidence after adjustment for the traditional CV risk factors. This finding is in line with previous literature showing plasma fibrinolytic system to be markedly down-regulated in patients with coronary artery disease and to predict ACS occurrence and outcome.51 Specifically, PAI-1 levels are increased in patients with acute myocardial infarction52–54 and a recent meta-analysis pooling different studies showed an association between PAI-1 and major adverse CV events in patients with coronary artery disease.55 In the present study, we reported that patients in the highest tertile of SIRT5 also had the highest PAI-1 expression demonstrating the biological relevance of SIRT5—PAI-1 regulatory axis in humans. These data are in line with in vivo and in vitro findings herein reported and support the concept of SIRT5 being a negative regulator of the protective fibrinolytic system activity.
This work presents some limitations that should be considered when interpreting the herein reported findings. First, although the use of systemic genetic models for experimental investigations is widely accepted, it does not allow to specifically investigate which cell type primarily concurs to determine the observed phenotype. Yet, our results suggested that alterations in platelet—major regulators of arterial thrombosis together with endothelial cells—did not account for the thrombosis alterations. Second, the exact molecular pathway linking loss of SIRT5 with blunted PAI-1 expression remains to be carefully delineated. However, both direct and indirect (through ERK1/2) effects of AMPK over activation on PAI-1 gene expression may be critically involved as previously discussed. Finally, PBMCs were chosen as surrogate cells for studying the relationship between SIRT5 and PAI-1 in a clinical cohort of humans with and without ACS. Clearly, results from this proof-of-principle experiment remain purely associative and will require further observations, including measurement of PAI-1 and SIRT5 protein levels and AMPK phosphorylation in ACS patients, to assess the cause and effect relationship and support our conclusions regarding their possible clinical applications.
In conclusion, this study demonstrates that SIRT5 mediates arterial thrombosis by modulating fibrinolysis via PAI-1. The translational nature of our data suggests SIRT5 as a potential mediator of acute atherothrombotic events and sets the stage for further investigations to validate its clinical relevance.
Supplementary material
Supplementary material is available at Cardiovascular Research online.
Authors’ contributions
L.L., A.A., M.X.M., and G.G.C. conceived the project and the experimental set-up. L.L., A.A., N.R.B., V.N., Y.M.P., and S.C. performed animal and in vitro experiments. N.I.V., K.Sta., and K.Ste. coordinated the enrolment of patients and analysed clinical data. L.L., N.I.V., K.Sta., K.Ste., and G.G.C. wrote the article. All authors provided critical input, critically reviewed the article, and contributed to discussion and interpretation of results.
Conflict of interest: LL and GGC are co-inventors on a provisional patent application that was filed in May 2019. The patent relates to the use of antibodies which specifically bind interleukin (IL)-1α to reduce various sequelae of ischemia-reperfusion injury to the central nervous system. All other authors report no conflict of interest.
Funding
The present work was supported by the Swiss Heart Foundation, the Swiss National Science Foundation [G.G.C. (310030_175546); T.F.L. and G.G.C. (310030_166576)], the Alfred and Annemarie von Sick Grants for Translational and Clinical Research Cardiology and Oncology to G.G.C. and a donation by Hans-Peter Wild to the Foundation for Cardiovascular Research–Zurich Heart House. G.G.C. and F.P. are the recipients the recipients of a H.H. Sheikh Khalifa bin Hamad Al Thani Foundation Assistant Professorship at the Faculty of Medicine, University of Zurich. The clinical part of the project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Grant Agreement No. 759248; MODVASC) and the German Research Foundation (DFG SFB834 Project Number 75732319) to K.Ste. as well as institutional funding to K.Sta.
Data availability
The data underlying this article will be shared on reasonable request to the corresponding author.
References
This study illustrates a novel role for Sirtuin 5 in arterial thrombosis by regulating fibrinolysis through plasminogen activator inhibitor 1 (PAI-1). These results shed new light onto the pathophysiology of arterial thrombus formation which underlies most of the acute atherosclerotic complications. In addition, they further affirm the intrinsic relationship between lifespan-regulating genes, vascular dysfunction, and age-related cardiovascular disease, thus indicating these genes as potential targets for cardiovascular prevention and therapy. Further studies will be needed to assess the predictive ability of SIRT5 in patients with acute cardiovascular or cerebrovascular events. In addition, the design of specific SIRT5 inhibitors will allow trials aiming at investigating the efficacy of SIRT5 blockage in the clinical setting.
