SIRT3 Modulates Endothelial Mitochondrial Redox State during Insulin Resistance

Emerging evidence indicates that defects in sirtuin signaling contribute to impaired glucose and lipid metabolism, resulting in insulin resistance (IR) and endothelial dysfunction. Here, we examined the effects of palmitic acid (PA) treatment on mitochondrial sirtuins (SIRT2, SIRT3, SIRT4, and SIRT5) and oxidative homeostasis in human endothelial cells (TeloHAEC). Results showed that treatment for 48 h with PA (0.5 mM) impaired cell viability, induced loss of insulin signaling, imbalanced the oxidative status (p < 0.001), and caused negative modulation of sirtuin protein and mRNA expression, with a predominant effect on SIRT3 (p < 0.001). Restoration of SIRT3 levels by mimic transfection (SIRT3+) suppressed the PA-induced autophagy (mimic NC+PA) (p < 0.01), inflammation, and pyroptosis (p < 0.01) mediated by the NLRP3/caspase-1 axis. Moreover, the unbalanced endothelial redox state induced by PA was counteracted by the antioxidant δ-valerobetaine (δVB), which was able to upregulate protein and mRNA expression of sirtuins, reduce reactive oxygen species (ROS) accumulation, and decrease cell death. Overall, results support the central role of SIRT3 in maintaining the endothelial redox homeostasis under IR and unveil the potential of the antioxidant δVB in enhancing the defense against IR-related injuries.


Introduction
Mitochondria are crucial players in the maintenance of endothelial function and regulation of cellular mechanisms responsible for the onset of inflammation and cell death. Mitochondrial dysfunction occurring in the high-fat condition impairs endothelial homeostasis thus constituting a major risk factor for cardiovascular diseases, including diabetes, heart failure, pulmonary hypertension, and insulin resistance (IR) [1]. In this contest, mitochondrial sirtuins (SIRT) contribute to cellular homeostasis due to their pivotal roles in controlling metabolic processes as well as the responses to nutrient availability and oxidative stress [1][2][3]. SIRT, class III histone deacetylases, represent attractive tools for epigenetic intervention due to their crucial role in DNA damage repair and control of cell survival and metabolic homeostasis, including IR [3][4][5].
IR refers to a condition in which insulin-responsive cells undergo a reduced response to insulin, caused by disruption of specific events in the signaling pathways resulting in the onset of diseases such as type 2 diabetes [6]. At the vascular level, insulin signaling begins EC viability was detected using the Cell Counting Kit-8 (CCK-8, Donjindo Molecular Technologies, Inc., Rockville, MD, USA) according to the manufacturer's instructions. Briefly, CCK-8 solution (10 µL) was added to each well and the plate was incubated at 37 • C for 4 h. Cell absorbance was measured at 450 nm with a microplate reader (model 680, Bio-Rad, Hercules, CA, USA) and viability was expressed as the mean of the optical density at 450 nm. All experiments were performed with n = 5 replicates.

SIRT3 Overexpressing
For transient overexpression of SIRT3, EC were seeded in 6-well plates at a density of 1.2 × 10 4 cells/well. The next day, the 70-80% subconfluent cultures were transfected with 20 nM SIRT3 AAV Vector for specific human Sirtuin 3 (438081010110, Applied Biological Materials, Inc. Richmond, BC, Canada) or miRNA mimic Negative Control (MCH00000, Applied Biological Materials, Inc. Richmond, BC, Canada), in serum-and antibiotic-free medium, using Lullaby (LL70500, OZ Biosciences, Marseille, France) as transfection reagent. EC were incubated for 6 h, followed by an additional 12 h of incubation after the addition of FBS, before treatments. In EC overexpressing SIRT3, mimic Negative Control (mimic NC) was used as a control. The transfection efficiency was confirmed by immunoblotting.

Glucose Uptake Analysis
Intracellular glucose content was determined using the Glucose Assay Kit-WST (G264, Dojindo Molecular Technologies, Inc., Rockville, MD, USA), according to the manufacturer's protocol. The absorbance of the WST formazan dye was measured at 450 nm by using a microplate reader (model 680, Bio-Rad, Hercules, CA, USA) and the glucose amount in each sample was calculated by plotting absorbance values with the standard calibration curve. Experiments were performed with n = 4 replicates.

Lactate Dehydrogenase (LDH) Assay
The integrity of the cell membrane related to LDH release was assessed by the LDH Assay Kit-WST (CK12, Dojindo Molecular Technologies, Inc., Rockville, MD, USA), according to the manufacturer's instructions. The working solution (100 µL) was added to the cell suspension (50 µL), followed by a 30 min incubation at room temperature in the dark. The absorbance was measured with a microplate reader (model 680, Bio-Rad, Hercules, CA, USA) at 490 nm, and the LDH content, expressed as a percentage of cytotoxicity, was calculated using the following equation: (Test Substance − Low Control)/(High Control − Low Control) × 100. Experiments were performed with n = 4 replicates.

Lipid Peroxidation Detection
Lipid peroxidation was determined by measuring the cellular content of malondialdehyde (MDA), according to the manufacturer's instructions, with the colorimetric lipid peroxidation assay kit (ab118970, Abcam, Cambridge, UK). Following treatment, EC were harvested, homogenized in 303 µL of MDA lysis solution, and 600 µL of thiobarbituric acid (TBA) reagent was added prior to 1 h of incubation at 95 • C and cooling on ice. The absorbance of the supernatant, containing MDA-TBA adduct, was measured at 532 nm with a microplate reader (model 680, Bio-Rad, Hercules, CA, USA). Total MDA levels, normalized to protein content, were calculated by comparing sample absorbance to a standard curve.

Evaluation of ROS
Intracellular and mitochondrial ROS content was determined with CellROX Green Reagent (C10444, Invitrogen, Waltham, MA, USA) and MitoSOX Red Mitochondrial Superoxide Indicator (M36008, Invitrogen, Waltham, MA, USA), according to manufacturer's instructions. After treatments, EC were stained for 30 min with 5 µM CellROX or MitoSOX fluorogenic probes in complete medium. Cells were imaged on a fluorescence microscope Antioxidants 2022, 11, 1611 4 of 18 using the EVOS FL Cell Imaging System (Thermo Scientific, Rockford, IL, USA) and intracellular and mitochondrial ROS fluorescence intensities were quantified using a BD Accuri C6 cytometer (BD Biosciences, San José, CA, USA). At least 20,000 events were recorded for each sample, and analysis was performed with FLOWJO V10 software (Williamson Way, Ashland, OR, USA). Extracellular ROS content was assessed with the Amplex Red Hydrogen Peroxide/Peroxidase Assay kit (A22188, Invitrogen, Waltham, MA, USA). Amplex red reagent (100 µL), containing 50 µM Amplex Red and 0.1 U HRP/ mL, was added to EC suspension (20 µL), containing 2 × 10 4 live cells in Krebs-Ringer phosphate glucose buffer. After 1 h of incubation at 37 • C, the fluorescence of the oxidized product, 10-acetyl-3,7dihydroxyphenoxazine, was measured with a multiplate reader (model Infinite 2000, Tecan, Männedorf, Switzerland) using a 530 nm excitation wavelength and a 590 nm emission wavelength. H 2 O 2 extracellular content was quantified using a standard curve (0-2 µM concentration range). Experiments were performed with n = 4 replicates. ROS-inducer menadione (50 µM) (M57405, Sigma Aldrich, St. Louis, MO, USA) was incubated for 30 min at 37 • C and used as a positive control.

Programmed Cell Death Mechanisms
Pyroptosis was detected by using the Pyroptosis/Caspase-1 assay (9145, Immuno-Chemistry Technologies, Davis, CA, USA), according to the company's instructions. After treatments, cells were stained with FLICA caspase-1 reagent (FAM-YVAD-FMK) 1 µL/mL medium for 1 h in the dark, before fluorescence microscopy with the EVOS FL Cell Imaging System (Thermo Scientific, Rockford, IL, USA). Nigericin (5 µM) was used as a positive control.
Lysosomal acidification was detected by labeling for 30 min at 37 • C in the dark treated EC with 1 µM LysoTracker Red DND-99 (L7528, Invitrogen, Waltham, MA, USA). The autophagic mechanism was assessed with the autophagy assay kit (ab139484, Abcam, Cambridge, UK), following the manufacturer's instructions. Cells were then washed with PBS and imaged on a fluorescence microscope EVOS FL Cell Imaging System (Thermo Scientific, Rockford, IL, USA) before FACS analysis. Rapamycin (1 µM overnight) was used as a positive control.
Apoptotic mechanisms were investigated with the Annexin V Apoptosis detection kit (556547, BD Pharmigen, Franklin Lakes, NJ, USA), in order to distinguish viable from necrotic and apoptotic cells. After treatments, EC were detached by trypsinization and washed with PBS, and were incubated in 200 µL binding buffer 1× containing 2 µL Annexin V-FITC and 2 µL PI (20 µg/mL) for 30 min.
For all programmed cell mechanisms, flow cytometry analysis was performed using a BD Accuri C6 cytometer (BD Biosciences, San José, CA, USA). For each sample, at least 20,000 events were recorded and the data analyzed by FlowJo V10 software (Williamson Way, Ashland, OR, USA).

RNA Isolation and Quantitative RT-PCR
Total RNA was extracted from EC using a Total RNA purification kit (17200, Norgen-Biotek Corp., Thorold, ON, Canada), according to the company's instructions, and then reverse transcribed into cDNA using a Tetro cDNA Synthesis Kit (BIO-65043, Meridian Bioscience Inc., Cincinnati, OH, USA), following the manufacturer s protocol, on a thermal cycler SureCycler 8800 (Agilent Technologies, Santa Clara, CA, USA). Quantitative real-time assays were carried out using a CFX96 Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA) with a QuantiTect SYBR Green PCR Kit (204,143, Qiagen, Hilden, Germany) and the following primers (0.2 µM): The relative SIRT gene expression levels were determined by comparing the expression of each sirtuin to that of GAPDH using the 2 −∆∆Ct method [∆∆ cycle threshold, Ct = (Ct SIRT-Ct GAPDH) of treated EC-(Ct SIRT-Ct GAPDH) of control], and the data was reported as the mean ± SD of n = 3 independent experiments, with each reaction performed in triplicate.

Statistical Analysis
All experiments were conducted at least three times, and results were reported as mean ± SD. Statistical analysis between two groups was performed using the Student's t test, while the differences among three groups were analyzed by one-way ANOVA followed by Tukey post hoc test. Analyses and graphs were performed using GraphPad Prism version 9.1.2. Differences with p < 0.05 were considered statistically significant.

Effects of PA on IR, Mitochondria Oxidative Status, and Sirtuins
In vitro assays on EC were used as an IR model induced by PA. The results showed the cytotoxic effects induced by PA, with the most relevant effect after treatment with 0.5 mM PA for 48 h (p < 0.001) ( Figure S1A). The IR was confirmed by decreased glucose uptake ( Figure S1B), upregulated phospho-IRS1 protein, reduced phospho-Akt/Akt ratio, and phospho-GSK-3β protein (p < 0.001) (Figures 1 and S6).
The IR assessment was also confirmed by the evaluation of phospho-IRS1, phospho-Akt/Akt ratio, and phospho-GSK-3β protein content under PA exposure with or without insulin treatment (Figures S1 and S13). Results indicated that PA+Insulin was unable to revert the PA-induced phospho-IRS1 increase, as well as the decreased phospho-Akt/Akt and phospho-GSK-3β levels (Figures S1 and S13), thus confirming the IR state. Moreover, exposure to PA resulted in an increased cytokine content (p < 0.001) and mitochondrial ROS production ( Figure 1). Notably, an increased acetylated-SOD2/SOD2 ratio, known to be linked to the suppression of SOD2 activity, was also observed (p < 0.001) (Figures 1 and  S6). Evaluation of protein and mRNA levels of SIRT2, SIRT3, SIRT4, and SIRT5 showed a decreased expression following PA exposure, with the greatest effect on SIRT3 (0.539 ± 0.03 vs. Ctr; ‡ p < 0.001 vs. Ctr. The IR assessment was also confirmed by the evaluation of phospho-IRS1, phospho-Akt/Akt ratio, and phospho-GSK-3β protein content under PA exposure with or without insulin treatment ( Figure S1 and S13). Results indicated that PA+Insulin was unable to revert the PA-induced phospho-IRS1 increase, as well as the decreased phospho-Akt/Akt

SIRT3 + Decreased the PA-Induced Cytotoxicity
The evidence that the cytotoxicity of PA was linked to mitochondrial damage and negative modulation of SIRT3, prompted us to investigate the effect of SIRT3 overexpression (SIRT3 + ) on the in vitro IR conditions ( Figure 2). be linked to the suppression of SOD2 activity, was also observed (p < 0.001) (Figure 1 and Figure S6). Evaluation of protein and mRNA levels of SIRT2, SIRT3, SIRT4, and SIRT5 showed a decreased expression following PA exposure, with the greatest effect on SIRT3 (0.539 ± 0.03 vs. 1.277 ± 0.04 AU in Ctr, p < 0.001) ( Figure 1K,L and Figure S6), suggesting the critical involvement of this sirtuin in the redox state of EC during IR.

SIRT3 + Decreased the PA-Induced Cytotoxicity
The evidence that the cytotoxicity of PA was linked to mitochondrial damage and negative modulation of SIRT3, prompted us to investigate the effect of SIRT3 overexpression (SIRT3 + ) on the in vitro IR conditions ( Figure 2). h Figure 2. SIRT3 + reduced the PA-related cytotoxicity. (A) Immunoblotting analysis of SIRT3 protein levels, (B) LDH assay cytotoxicity, (C) MDA, and (D) extracellular ROS content in EC treated with the empty transfection reagent (Lullaby) or transfected with mimic Negative Control (mimic NC), SIRT3 mimic (SIRT3 + ), mimic Negative Control and then exposed to 0.5 mM PA for 48 h (mimic NC+PA) or SIRT3 mimic before 48 h treatment with PA (SIRT3 + +PA). Control cells (Ctr) were treated with the corresponding volume of HBSS-10 mM Hepes. Western blotting results (n = 5) are expressed as arbitrary units (AU) and represented as boxplots. ‡ p < 0.001 vs. mimic NC; § p < 0.05 vs. mimic NC+PA; # p < 0.01 vs. mimic NC+PA.
The induction of SIRT3 + was confirmed by the evaluation of protein levels (Figure 2A and Figure S7). When the efficacy of SIRT3 + to oppose the PA-induced cytotoxicity was investigated (SIRT3 + +PA), results indicated that SIRT3 + inhibited the LDH release (p < 0.05 vs. mimic NC+PA), as well as the increase in MDA content (p < 0.01 vs. mimic NC+PA) The induction of SIRT3 + was confirmed by the evaluation of protein levels (Figures 2A  and S7). When the efficacy of SIRT3 + to oppose the PA-induced cytotoxicity was investigated (SIRT3 + +PA), results indicated that SIRT3 + inhibited the LDH release (p < 0.05 vs. mimic NC+PA), as well as the increase in MDA content (p < 0.01 vs. mimic NC+PA) and the accumulation of extracellular ROS (p < 0.01 vs. mimic NC+PA) ( Figure 2B,D). Based on the evidence that in our experimental conditions, the mimic negative control (mimic NC) showed no cytotoxic effects compared to the control (Ctr) (Figure 2), the mimic NC was used as a control for further experiments.

SIRT3 + Reverted the PA-Induced IR State and Oxidative Stress
Evaluation of phosphorylated-IRS1, -Akt, and -GSK-3β protein levels showed that SIRT3 overexpression was able to counteract the PA-induced IR state (p < 0.001 vs. mimic NC+PA) (Figures 3A-C and S8).

Effects of SIRT3 + on PA-Induced Inflammation
The efficacy of SIRT3 + to oppose cytokine content induced by PA was then investigated (Figure 4).  The results indicated that the protective effect of SIRT3 + was accompanied by its efficacy in reducing IL-6 and TNF-α levels and protein expression induced by PA (p < 0.01 vs. mimic NC+PA) ( Figure 4A-C and Figure S9), as well as in downregulating the expression levels of NF-κB. Specifically, in SIRT3 + +PA the upregulation of NF-κB protein was lower compared to mimic NC+PA (0.467 ± 0.04 AU vs. 0.657 ± 0.03 AU and 53.78 ± 3.14 AFU vs. 84.7 ± 5.21 AFU, p < 0.05) ( Figure 4D-F and Figure S9). The results indicated that the protective effect of SIRT3 + was accompanied by its efficacy in reducing IL-6 and TNF-α levels and protein expression induced by PA (p < 0.01 vs. mimic NC+PA) ( Figures 4A-C and S9), as well as in downregulating the expression levels of NF-κB. Specifically, in SIRT3 + +PA the upregulation of NF-κB protein was lower compared to mimic NC+PA (0.467 ± 0.04 AU vs. 0.657 ± 0.03 AU and 53.78 ± 3.14 AFU vs. 84.7 ± 5.21 AFU, p < 0.05) ( Figures 4D-F and S9).

SIRT3 + Reduced the Autophagy Induced by PA
Instead, results indicated a reduction of the PA-mediated cytotoxic effects in SIRT3 + +PA treated cells (81.0% ± 3.22 of live cells and 14.8 % ± 1.09 of apoptotic cells, p < 0.05 vs. mimic NC+PA) ( Figure S3C,D).

Effects of δVB on Mitochondrial ROS and Sirtuin Modulation
The antioxidant δVB was tested as a possible modulator of the mitochondrial homeostasis during exposure to PA. Results showed that cell viability was not affected by treatment with δVB up to 72 h of incubation ( Figure S5A). Dose-response indicated that 0.5 mM δVB was the most effective dose in reducing the cytotoxic effects of PA (p < 0.01 vs. PA), acting on the IRS1/Akt pathway ( Figure S5B-D and Figure S14). Pretreatment with 0.5 mM δVB before exposure to PA (δVB+PA) attenuated the mitochondrial ROS cascade (294 ± 17.7 vs. 666 ± 31.4 MFI in PA, p < 0.05) (Figure 7A-C) and upregulated SIRT2, SIRT3, SIRT4, and SIRT5 protein and mRNA, with the highest effect on SIRT3 (p < 0.01) ( Figure   Figure 6. SIRT3 + decreased the PA-induced autophagy. Representative images of fluorescence microscopy and flow cytometry analysis of (A-C) green detection reagent and (F-H) Lysotracker Red, quantified as MFI (scale bars = 100 µm). Western blotting analysis of (D) beclin-1, (E) p62, (I) ATG5, and (J) LC3B II/I in EC transfected with mimic Negative Control (mimic NC), SIRT3 mimic (SIRT3 + ), mimic Negative Control and then exposed to 0.5 mM PA for 48 h (mimic NC+PA) or SIRT3 mimic before 48 h treatment with PA (SIRT3 + +PA). Western blotting results (n = 5) are expressed as arbitrary units (AU) and represented as boxplots. † p < 0.01 vs. mimic NC; ‡ p < 0.001 vs. mimic NC; § p < 0.05 vs. mimic NC+PA; # p < 0.01 vs. mimic NC+PA.

Effects of δVB on Mitochondrial ROS and Sirtuin Modulation
The antioxidant δVB was tested as a possible modulator of the mitochondrial homeostasis during exposure to PA. Results showed that cell viability was not affected by treatment with δVB up to 72 h of incubation ( Figure S5A). Dose-response indicated that 0.5 mM δVB was the most effective dose in reducing the cytotoxic effects of PA (p < 0.01 vs. PA), acting on the IRS1/Akt pathway ( Figures S5B,D and S14). Pretreatment with 0.5 mM δVB before exposure to PA (δVB+PA) attenuated the mitochondrial ROS cascade (294 ± 17.7 vs. 666 ± 31.4 MFI in PA, p < 0.05) ( Figure 7A-C) and upregulated SIRT2, SIRT3, SIRT4, and SIRT5 protein and mRNA, with the highest effect on SIRT3 (p < 0.01) ( Figures 7F,G and S12).

Discussion
Here, we investigated the role of mitochondrial sirtuins in the protection of human endothelial cells against PA-induced IR. The PA-induced mitochondrial dysfunction led to pyroptosis, autophagy, and apoptosis. Transfection with SIRT3 mimics reverted PAinduced ROS accumulation, inflammatory-related pyroptosis, and autophagy. Furthermore, the modulation of SIRT3 levels and the improvement of PA-induced mitochondrial dysfunction have been achieved using the antioxidant δVB.
Among histone deacetylases, mitochondrial SIRT2, SIRT3, SIRT4, and SIRT5 are key metabolic enzymes controlling mitochondrial homeostasis and response to oxidative stress [3,[27][28][29]. Compelling evidence indicates that mitochondrial sirtuins are involved in the regulation of endothelial function given their ability to regulate energy metabolism, ROS production, and cellular redox state [2]. SIRT3, the main mitochondrial deacetylase, regulates multiple metabolic pathways and its expression is reduced in diabetes [2]. Moreover, SIRT3 activation improved IR while its inhibition precipitated IR in adipocytes [30]. Recent findings suggest that endothelial IR induced by SIRT3 deletion could reprogram endothelial cell metabolism, shifting metabolism from oxygen-independent glycolysis to oxygen-dependent oxidative phosphorylation [31,32]. Downregulation of SIRT3 by siRNA reduced insulin response in human umbilical vein endothelial cells (HUVECs), whereas overexpression improved IR induced by palmitate, thus suggesting an association between SIRT3 and insulin sensitivity [32]. Here, overexpression of SIRT3 in endothelial TeloHAEC cells abolished mitochondrial ROS generation caused by PA and counteracted IR injuries through the attenuation of inflammation, which then flowed to the inflammatory form of programmed cell death, pyroptosis. Interestingly, PA induced only a weak apoptotic cell death rate, suggesting that pyroptosis represents the principal cellular phenomenon during PA-induced IR. Calorie restriction and exercise-mediated weight loss decreased mRNA levels of NLRP3 and IL-1β, and improved insulin sensitivity in diabetic subjects [33], demonstrating that pyroptosis contributes to the pathogenesis of IR and diabetes by mediating inflammation and β-cell destruction. The regulation of SIRT2, SIRT3, and SIRT4 expression is known to relate to inflammation and pyroptosis cell death [34][35][36]. To date, the concept of endothelial IR has not been clinically established

Discussion
Here, we investigated the role of mitochondrial sirtuins in the protection of human endothelial cells against PA-induced IR. The PA-induced mitochondrial dysfunction led to pyroptosis, autophagy, and apoptosis. Transfection with SIRT3 mimics reverted PA-induced ROS accumulation, inflammatory-related pyroptosis, and autophagy. Furthermore, the modulation of SIRT3 levels and the improvement of PA-induced mitochondrial dysfunction have been achieved using the antioxidant δVB.
Among histone deacetylases, mitochondrial SIRT2, SIRT3, SIRT4, and SIRT5 are key metabolic enzymes controlling mitochondrial homeostasis and response to oxidative stress [3,[27][28][29]. Compelling evidence indicates that mitochondrial sirtuins are involved in the regulation of endothelial function given their ability to regulate energy metabolism, ROS production, and cellular redox state [2]. SIRT3, the main mitochondrial deacetylase, regulates multiple metabolic pathways and its expression is reduced in diabetes [2]. Moreover, SIRT3 activation improved IR while its inhibition precipitated IR in adipocytes [30]. Recent findings suggest that endothelial IR induced by SIRT3 deletion could reprogram endothelial cell metabolism, shifting metabolism from oxygen-independent glycolysis to oxygen-dependent oxidative phosphorylation [31,32]. Downregulation of SIRT3 by siRNA reduced insulin response in human umbilical vein endothelial cells (HUVECs), whereas overexpression improved IR induced by palmitate, thus suggesting an association between SIRT3 and insulin sensitivity [32]. Here, overexpression of SIRT3 in endothelial TeloHAEC cells abolished mitochondrial ROS generation caused by PA and counteracted IR injuries through the attenuation of inflammation, which then flowed to the inflammatory form of programmed cell death, pyroptosis. Interestingly, PA induced only a weak apoptotic cell death rate, suggesting that pyroptosis represents the principal cellular phenomenon during PA-induced IR. Calorie restriction and exercise-mediated weight loss decreased mRNA levels of NLRP3 and IL-1β, and improved insulin sensitivity in diabetic subjects [33], demonstrating that pyroptosis contributes to the pathogenesis of IR and diabetes by mediating inflammation and β-cell destruction. The regulation of SIRT2, SIRT3, and SIRT4 expression is known to relate to inflammation and pyroptosis cell death [34][35][36]. To date, the concept of endothelial IR has not been clinically established because only a small number of clinical studies have shown the actual response to insulin in endothelial cells of patients with metabolic disorders [37][38][39][40][41][42].
The use of naturally occurring compounds in lifestyle and nutrition improves and/or prevents aging-related diseases such as IR, diabetes, and its vascular complications [19,43]. Recent advances in small metabolites have opened novel avenues in the metabolic route of betaines, metabolites with a quaternary ammonium group involved in the regulation of metabolic homeostasis such as protection against osmotic stress. One such compound, δVB, has been associated with positive health effects (fetal brain development, insulin secretion, and reduced cancer risk) and some negative health outcomes associated with a high-fat diet (cardiovascular disease and fatty liver disease) [44], suggesting that investigations into the metabolic route and role of this novel betaine are still necessary. δVB, at least partially synthesized by gut microbes, has been reported to be present in certain food sources, including milk and meat, and the kinds of marine algae used as foodstuffs [25,44].
In vitro evidence suggests that δVB prevents endothelial damage related to diabetes [19] and displays antineoplastic effects in colon cancer cells and head and neck squamous cell carcinomas [20][21][22]. In the human and mouse heart, δVB affected energy metabolism, showing a similar effect to that of meldonium, a cardioprotective drug used to improve cardiac function after ischemia [16,17]. Clinical studies revealed that treatment of type 2 diabetes with metformin increased the δVB concentration in human serum [45]. Furthermore, increased plasma levels of δVB and other betainized compounds as a consequence of a diet rich in whole grains were correlated with improved insulin resistance and insulin secretion [16,18]. Additionally, δVB was demonstrated to improve glucose tolerance and insulin tolerance in mice [7], suggesting that the higher δVB concentration in plasma might exert some protection against the diabetogenic Western diet. In line with previous data [19], δVB did not show cytotoxicity up to 1 mM concentration and protected EC against PA-induced IR by limiting the inhibitory phosphorylation of IRS1 and restoring phosphorylated Akt expression levels. In particular, δVB reduced oxidative stress and cytokine content, restored the expression levels of SIRT3, and attenuated pyroptosis and autophagy. The beneficial effects mediated by δVB could be related, at least in part, to its ability to positively modulate SIRT3 expression levels, indicating δVB as a possible antioxidant in the prevention of IR. However, it cannot be ruled out that other molecular targets may be involved in this molecular mechanism. Further studies are undoubtedly necessary for a deeper understanding of the molecular pathway through which SIRT3 controls endothelial redox state during IR, in order to develop novel antioxidant-targeted preventive approaches.

Conclusions
Mitochondrial homeostasis is a crucial feature to contrast endothelial injuries related to IR. SIRT3 overexpression improved the endothelial oxidative status and reduced PAinduced programmed cell death, pyroptosis, apoptosis, and autophagy. Furthermore, the antioxidant betaine, δVB, acted as a promising SIRT3 modulator in defining future and innovative preventive strategies for IR.