Anti-Inflammatory Properties of Sirtuin 6 in Human Umbilical Vein Endothelial Cells

A prominent feature of inflammatory diseases is endothelial dysfunction. Factors associated with endothelial dysfunction include proinflammatory cytokines, adhesion molecules, and matrix degrading enzymes. At the transcriptional level, they are regulated by the histone deacetylase sirtuin (SIRT) 1 via its actions on the proinflammatory transcription factor nuclear factor-κB (NF-κB). The role of SIRT6, also a histone deacetylase, in regulating inflammation in endothelial cells is not known. The aim of this study was to determine the effect of SIRT6 knockdown on inflammatory markers in human umbilical vein endothelial cells (HUVECs) in the presence of lipopolysaccharide (LPS). LPS decreased expression of SIRT6 in HUVECs. Knockdown of SIRT6 increased the expression of proinflammatory cytokines (IL-1β, IL-6, IL-8), COX-prostaglandin system, ECM remodelling enzymes (MMP-2, MMP-9 and PAI-1), the adhesion molecule ICAM-1, and proangiogenic growth factors VEGF and FGF-2; cell migration; cell adhesion to leukocytes. Loss of SIRT6 increased the expression of NF-κB, whereas overexpression of SIRT6 was associated with decreased NF-κB transcriptional activity. Taken together, these results demonstrate that the loss of SIRT6 in endothelial cells is associated with upregulation of genes involved in inflammation, vascular remodelling, and angiogenesis. SIRT6 may be a potential pharmacological target for inflammatory vascular diseases.


Introduction
Endothelial dysfunction is associated with vasoconstriction, a proinflammatory state, and prothrombotic properties. A number of diseases, including most forms of cardiovascular disease, diabetes, cancer, rheumatoid arthritis, and aging, are associated with endothelial dysfunction [1][2][3]. Endothelial dysfunction is also a characteristic of bacterial sepsis [4,5]. To reduce or prevent the high incidence of morbidity and mortality associated with vascular complications, it is essential to understand further the mechanism by which inflammation regulates endothelial function.
There is now increasing evidence that a number of transcription factors, including the sirtuin (SIRT) family and nuclear factor κB (NF-κB), regulate multiple genes whose products are putatively involved in the regulation of endothelial cell function. The sirtuin family consists of seven enzymes (SIRT1-SIRT7) that share a conserved core catalytic domain but differ in their cellular localisation and tissue distribution [11]. To date, most of the studies have focussed on SIRT1. Numerous genes involved in vascular growth, maturation, and remodelling are dysregulated in SIRT1 deficient endothelial cells [12][13][14][15]. Additionally, administration of resveratrol, a SIRT1 activator, improves endothelial dysfunction [16][17][18][19] and reduces vascular inflammation in mice [20], two central pathophysiological processes involved in the initiation and progression of cardiovascular disease. Importantly, SIRT1 mediates these effects by interacting and inhibiting the proinflammatory transcription factor NF-κB [16][17][18][19][20].
A recent study in mice has suggested a role for SIRT6 in inflammation [21]; however, the role of SIRT6 in endothelial function is not known. In this study, a human umbilical vein endothelial cell (HUVEC) line will be used to determine the effect of the endotoxin lipopolysaccharide (LPS) on SIRT6 expression. In addition, the effect of SIRT6 inhibition on mediators of vascular inflammation will be determined. Specifically, the effect of SIRT6 siRNA on the expression of proinflammatory cytokines (IL-1β and IL-6), COX-prostaglandin system (COX-2 and subsequent prostaglandin release), ECM remodelling enzymes (MMP-2, MMP-9, PAI-1), adhesion molecules (ICAM-1, VCAM-1, P-selectin) and angiogenic growth factors (VEGF and FGF-2) will be assessed. The migration and adhesion properties of SIRT6 deficient HUVECs will also be assessed. In order to elucidate if SIRT6 mediates its effects via NF-κB, luciferase assays will be used to determine the effect of SIRT6 overexpression on NF-κB activity.

Treatment of HUVECs with LPS.
HUVECs (HUV-EC-C) were obtained from ATCC (Manassas, VA, USA). Cells were incubated in 25 cm 2 flask in DMEM/F12 enriched with 10% FCS, 100 U/mL penicillin G, and 100 μg/mL streptomycin until they reached 80% confluence. The cells were the treated with DMEM/F-12 containing 2% heatinactivated FCS in the absence or presence of 10 μg/mL LPS. After 48 h incubation, cells were collected and assessed for SIRT6 expression by Western blotting as detailed below. Six independent experiments were performed.

Gene Silencing of SIRT6 in HUVECs.
HUVECs were incubated in DMEM/F12 enriched with 10% FCS, 100 U/mL penicillin G, and 100 μg/mL streptomycin. Cells were transfected using TransIT-siQUEST reagent according to manufacturer guidelines (Mirus Bio, Madison, WI, USA). Cells were transfected with either 50 nM SIRT6 or nonspecific siRNA from Ambion (Austin, Texas) in DMEM (containing 100 U/mL penicillin G, 100 μg/mL streptomycin) for 72 h. The cells were then treated with either 10 μg/mL LPS (in DMEM/F-12 containing 2% heat-inactivated FCS) for an additional 24 h. Cells were collected and stored at −80 • C until assayed for mRNA expression by qRT-PCR and protein expression by Western blotting as detailed below. Media were collected and stored at −80 • C until assayed for cytokine and prostaglandin release by ELISA as detailed below. Six independent experiments were performed.

Transient Transfection and Luciferase Assay for Detecting
NF-κB Activity. A luciferase assay was utilised to determine possible interactions between SIRT6 and NF-κB. HUVECs were transfected using lipofectamine 2000 as described by the manufacturer (Life Technologies, Carlsbad, CA). Cells were cotransfected with 1 μg reporter (either pNF-κB-luc or negative control; Qiagen) and 1 μg plasmid (empty vector or SIRT6 vector; Origene) for 72 h. The cells were then treated with DMEM/F-12 containing 2% heat-inactivated FCS in the absence or presence of 10 μg/mL LPS for 24 h. The media were collected and assayed for IL-6, IL-8, PGE 2 , and PGF 2α as detailed above. The cells were harvested in lysis buffer and luminescence activity was measured using the Dual-Glo Luciferase Assay kit (Promega) as instructed. Relative luciferase activity was calculated after normalisation with Renilla activity. The results are expressed as a ratio of luciferase activity of the control group. Three independent experiments were performed.

Enzyme
Immunoassays. The release of IL-6 and IL-8 was performed by sandwich ELISA according to the manufacturer's instructions (Invitrogen, Carlsbad, CA). The release of PGE 2 and PGF 2α into the incubation medium was assayed using a commercially available competitive enzyme immunoassay kit according to the manufacturer's specifications (Kookaburra Kits from Sapphire Bioscience, NSW, Australia). The human ICAM-1 ELISA kit was purchased from PeproTech, Inc. (Rocky Hill, NJ).

RNA Extraction and Quantitative RT-PCR (qRT-PCR).
Total RNA was extracted using Tri-Reagent according to manufacturer's instructions (Sigma-Aldrich, Saint Louis Missouri). RNA concentrations were quantified using a spectrophotometer (Smart Spec, Bio-Rad). RNA concentration and purity were measured using a NanoDrop ND1000 spectrophotometer (Thermo Scientific, Pittsburgh, PA). RNA was converted to cDNA using the SuperScript VILO cDNA synthesis kit (Invitrogen, Carlsbad, California, USA) according to the manufacturer's instructions. The cDNA was diluted tenfold, and 0.8 μL of cDNA was used to perform RT-PCR using Sensimix Plus SYBR green (Quantace, Alexandria, NSW, Australia) and 100 nM of prevalidated and designed primers (QuantiTect Primer Assays, Qiagen, Germantown, Maryland, USA). The specificity of the product was assessed from melting curve analysis. RNA without reverse transcriptase during cDNA synthesis as well as PCR reactions using water instead of template showed no amplification. Average gene C T values were normalised to the average GAPDH C T values of the same cDNA sample. Fold differences were determined using the comparative C T method, and the Mediators of Inflammation 3 results are expressed as a ratio of the average of the control group.

Cell Migration Assay.
To assess the migration/invasion ability of SIRT6 deficient cells, a gelatin invasion assay was used. Briefly, transwell inserts (8 μm pore size) were coated on the underside with 0.1% gelatin. HUVECs transfected with NS or SIRT6 siRNA were prepared as detailed above and 5 × 10 5 cells were allowed to migrate across inserts for 5 hours at 37 • C. Unmigrated cells from the upper chamber were removed, and migrated cells in the lower side of the membrane were fixed, stained, and counted. The average number of migrated cells in ten randomly chosen fields of view per insert was taken to quantify the extent of migration. Four independent experiments were performed.

Cell Adhesion
Assays. HUVECs transfected with NS or SIRT6 siRNA were prepared as detailed above and 2 × 10 4 cells in 100 μL per well were added to gelatin precoated 96well plates and allowed to adhere for 1.5 h at 37 • C. Unbound cells were washed, 10 μL of MTT (5 mg/mL stock in PBS) was added per well, and the incubation continued for 4 h. After washing, the bound dye was lysed with 200 μL DMSO and absorbance read at 570 nm. Four independent experiments were performed.

Endothelial-Leukocyte Adhesion Assay.
Polymorphonuclear leukocyte (PMN) cells, isolated from whole blood, were a gift from Dr. Katherine Woods (Ludwig Institute for Cancer Research, Austin Hospital, Heidelberg, Victoria, Australia). For these experiments, 48-well plates were used and HUVECs transfected with NS or SIRT6 siRNA were prepared as detailed above. After 3 washings with serum-free medium, 1 × 10 6 PMN cells per mL were added to each well (final 2.5 × 10 5 PMN cells per well), and incubated at 37 • C for an additional 60 min. After endothelial-leukocyte coculture, nonadherent PMNs cells were removed by washing. The number of adhered PMNs was evaluated by a colorimetric assay using MTT assay as detailed above. Four independent experiments were performed.

Statistical Analysis.
Statistical analyses were performed using a commercially available statistical software package (Statgraphics Plus version 3.1, Statistical Graphics Corp., Rockville, Maryland, USA). Two sample comparisons were analysed by a paired sample comparison. For all other comparisons, analysis was performed using a one-way ANOVA (using Tukey HSD correction to discriminate among the means); homogeneity of data was assessed by Bartlett's test, and when significant, data were logarithmically transformed before further analysis. Statistical difference was indicated by a P value of less than 0.05. Data are expressed as mean ± standard error of the mean (SEM).

Effect of LPS on SIRT6
Expression. The first aim was to determine the effect of LPS on SIRT6 expression in HUVECs. To do this, HUVECs were incubated in the absence or presence of 10 μg/mL LPS for 48 h and protein expression assessed by Western blotting (n = 6 experiments). As shown in Figure 1, LPS significantly decreased SIRT6 protein expression, without affecting β-actin expression.

Efficiency of SIRT6 Knockdown Using siRNA in HUVECs.
Having shown that LPS inhibits SIRT6 expression in HUVECs, the next aim was to determine the effect of SIRT6 knockdown on endothelial function in the presence of LPS. To do this, HUVECs were transfected with and without SIRT6 siRNA and nonspecific (NS) siRNA. After 72 h transfection, cells were incubated in the absence or presence of 10 μg/mL LPS for 24 h. The efficacy of transfection was analysed by qRT-PCR and Western blotting. Compared to NS siRNA transfected HUVECs, transfection with SIRT6 siRNA resulted in a significant decrease in SIRT6 mRNA (Figure 2(a)) and protein expression (Figure 2(b)). There was, however, no effect of SIRT6 siRNA on β-actin protein expression (Figure 2(b)).

Effect of SIRT6 Knockdown on Cytokine and Chemokine
Release and Steady State of Cytokine mRNA. Treatment of HUVECs with LPS significantly increased the release ( Figure  3(a)) and gene expression (Figure 3(b)) of IL-1β, IL-6, and IL-8. In support with the hypothesis that SIRT6 regulates endothelial cell inflammation, siRNA knockdown of SIRT6 is associated with significantly increased gene expression of the proinflammatory cytokine IL-6 and the chemokine IL-8 ( Figure 3(a)). To determine whether inhibition of SIRT6 plays a role in the steady state of cytokine mRNA expression, qRT-PCR analysis was performed after treatment of HUVECs with siRNA. Consistent with the release data, when compared to NS siRNA transfected HUVECs, SIRT6 siRNA induced IL-6 and IL-8 mRNA expression (Figure 3(b)), as well as increasing the mRNA expression of IL-1β.

Effect of SIRT6 Knockdown on COX-Prostaglandin
Pathway. The next aim was to determine the effect of SIRT6 siRNA on the COX-prostaglandin system, specifically COX-1 and COX-2 gene expression and subsequent PGE 2 and PGF 2α release. LPS significantly induced COX-2 mRNA expression (Figure 4(a)) and release of PGE 2 and PGF 2α (Figure 4(b)). As shown in Figure 4, inhibition of SIRT6 significantly increased COX-2 mRNA expression and subsequent prostaglandin release. There was no effect of LPS or SIRT6 knockdown on COX-1 mRNA expression.

Effect of SIRT6 siRNA on ECM Remodelling Enzymes.
We next sought to determine the effect of SIRT6 knockdown on enzymes involved in ECM degradation and ECM remodelling, namely PAI-1, MMP-2, and MMP-9. As shown in Figure 5, LPS induced MMP-9 and PAI-1 gene expression; an effect that was even higher in SIRT6 deficient HUVECs. On the other hand, there was no effect of LPS or SIRT6 inhibition on MMP-2 gene expression. 3.6. Effect of SIRT6 siRNA on Cell Adhesion. To investigate putative promigratory effects of SIRT6 siRNA, endothelial VCAM-1, ICAM-1, and P-selectin mRNA expression was quantified by qRT-PCR. LPS significantly increased the gene expression of the cell adhesion molecule ICAM-1 ( Figure  6(a)). This was also associated with an increase in the secretion of ICAM-1 into the conditioned media ( Figure  6(b)). Additionally, when compared to LPS alone, inhibition of SIRT6 significantly increased ICAM-1 gene expression and release (Figures 6(a), 6(b)). There was, however, no effect of LPS or LPS plus SIRT6 siRNA on VCAM-1 and P-selectin gene expression (Figure 6(a)).
To evaluate the role of SIRT6 in the cell adhesion potential of HUVECs, two different assays were performed. In the first assay, NS or SIRT6 siRNA transfected HUVECs were plated in 1% gelatin-coated 96 well plates for 1.5 h. Thereafter, unattached cells were washed and MTT colorimetric assay was used to determine the number of adherent cells. The optical density of the resulting solution was determined at 570 nm. As depicted in Figure 6 The effect of SIRT6 siRNA transfected HUVECs to bind to leukocytes was also assessed. In these experiments, leukocytes were added to confluent monolayers of HUVECs transfected with NS or SIRT6 siRNA. As shown in Figure  6(d), in NS siRNA transfected HUVECs, stimulation with LPS induced a significant increase in the adhesion of leukocytes. When SIRT6 was knocked out in these cells, the adhesion of leukocytes was further increased.

Effect of SIRT6 siRNA on Cell Migration and Angiogenesis.
Although there was no effect of LPS on the expression of the angiogenic growth factors VEGF and FGF2 ( Figure  7(a)), in HUVECs transfected with SIRT6 siRNA, VEGF and

SIRT6 Regulates the Expression and Activity of NF-κB.
The final aim of this study was to determine if SIRT6 regulates proinflammatory factors in HUVECs via the NF-κB pathway. To assess this, the effect of SIRT6 siRNA on NF-κB transcription was investigated first. As shown in Figure 8(a), knockdown of SIRT6 significantly increased NF-κB p65 mRNA expression. Next, whether SIRT6 regulates the transcriptional activity of NF-κB by luciferase assays using NF-κB-luc reporter was determined. The cells were also transfected with SIRT6 expression vector or control vector pcDNA3.1. As shown in Figure 8(b), transfection with NF-κB luc reporter significantly increased luciferase activity. However, overexpression of SIRT6 inhibited NF-κBmediated luciferase reporter expression in HUVECs. The effect of NF-κB overexpression with and without cotransfection with SIRT6 in HUVECs on a number of NF-κB target genes was also evaluated. As shown in Figure 9, overexpression of NF-κB significantly increased IL-6 and PGF 2α levels in the conditioned media. This increase was inhibited in HUVECs that also overexpressed SIRT6. Similar results were obtained for IL-8 and PGE 2 (data not shown).

Discussion
Epidemiological, clinical, and animal studies demonstrate that bacterial infection can induce endothelial dysfunction cells [6][7][8][9][10]23]. Recent reports highlight an important protective role for SIRT1 in vascular dysfunction. For example, infection decreases SIRT1 expression and activity [22,24] and activation of SIRT1 exerts beneficial effects on endothelial cell dysfunction [18,25]. This is the first study to report that LPS can also decrease SIRT6 expression in HUVECs. This led to the hypothesis that SIRT6 may also play an important role in infection induced endothelial dysfunction. In order to test this theory, SIRT6 was inhibited in HUVECs using siRNA, and then a number of factors known to play an important role in vascular disease were investigated. In SIRT6 deficient cells, in the presence of LPS, there was an increase in the expression of the proinflammatory cytokines IL-1β and IL-6, the chemokine IL-8, ECM degrading enzymes MMP-9 and PAI-1, COX-prostaglandin system, and the angiogenic growth factors FGF-2 and VEGF. The loss of SIRT6 in HUVEC was associated with increased expression of the cell adhesion molecule ICAM-1, as well as increased cell adhesion properties. In addition, loss of SIRT6 increases the expression of the proinflammatory transcription factor NF-κB, whereas overexpression of SIRT6 decreases the transcriptional activity of NF-κB as well as the expression of its target genes. Collectively, this suggests that SIRT6 is a negative regulator of vascular inflammation. Thus, activation of SIRT6 may be beneficial for many inflammatory diseases associated with endothelial dysfunction.
Endothelial cells themselves are a major source of proinflammatory cytokines, which are significantly enhanced during inflammation [9]. These proinflammatory stimuli can induce a phenotypic change of the quiescent endothelium to an activated endothelium. This results in an increase or induction of the endothelial expression of adhesion molecules and chemokines (chemoattractant cytokines), leading to leukocyte recruitment, adhesion, and infiltration into the subendothelium. Activated endothelium also leads to an increase in cytokine expression and production that may act in an autocrine fashion to enhance endothelial cell activation further. In this study, LPS induced the expression and release of the proinflammatory cytokines IL-1β and IL-6, the chemokine IL-8; this increase was even higher in SIRT6 deficient HUVECs. As expected, LPS stimulation induced the adhesion molecule ICAM-1 as well as leukocyte cell adhesion to HUVECs, and this increase was further augmented in SIRT6 deficient cells. Of note, there was no effect of LPS or SIRT6 inhibition on the expression of  Figure 6: Effect of SIRT6 knockdown on cell adhesion. HUVECs were transfected with 50 nM SIRT6 or NS siRNA for 72 h followed by 10 μg/mL LPS for 24 h (n = 6). (a) The gene expression of the adhesion molecules VCAM-1, ICAM-1, and P-selectin was quantified by qRT-PCR. (b) ICAM-1 concentration in the conditioned media was assayed using ELISA. (c) Transfected HUVECs were plated onto gelatincoated 96-well plates for 1.5 h, and cell adhesion assays were performed using MTT. Extracted dye was quantified by absorbance at 570 nm. Each bar represents the mean ± SEM. * P < 0.05 versus NS siRNA + LPS. (d) PMN cells were added to each well and were incubated at 37 • C for 60 min. The number of adhered PMNs was evaluated by a colorimetric assay using MTT assay and quantification by absorbance at 570 nm. Each bar represents the mean ± SEM. * P < 0.05 versus NS siRNA + LPS.
VCAM-1 or P-selection. Although there are no studies on SIRT6 and adhesion molecules, these studies are in support of the few other studies demonstrating that SIRT6 regulates inflammatory cytokines [26,27].
In the process of vascular inflammation, activated endothelial cells can also migrate to surrounding tissues to spur the formation of new blood vessels, or angiogenesis, which plays an important role in a number of pathologies such as atherosclerosis, diabetes, and tumour development [28][29][30]. Indeed, in experimental studies, LPS has been shown to promote human endothelial cell migration. In this study, LPS stimulation not only promoted HUVEC migration but the effect was intensified in SIRT6 deficient HUVECs. VEGF and FGF-2 are two angiogenic growth factors whose activation leads to dysregulated blood vessel formation as seen in various chronic inflammatory conditions such as cancer and rheumatoid arthritis [28,29]. In this study, SIRT6 deficient HUVECs had increased levels of both VEGF and FGF-2.
COX-2 and its proinflammatory products have been implicated in the pathogenesis of several inflammatory diseases including diabetes and atherosclerosis [31]. In addition, COX-2 can also contribute to the pathogenesis of endothelial dysfunction by disrupting vascular homeostasis. It also generates vasoconstrictor substances, including vasoconstrictor prostanoids that have proatherogenic effects [32,33]. HUVECs were cotransfected with 1 μg negative control construct and NF-κB reporter construct, along with 1 μg empty or SIRT6 pCMV6 vector for 72 h. Promoter activity is expressed as a ratio of luciferase activity of the control group, which was set as 1. Each bar represents the mean ± SEM (n = 3 experiments). * P < 0.05 versus cells transfected with NF-κB reporter plus empty pCMV6 vector.
This study demonstrates that in the presence of infection, SIRT6 is a regulator of the COX-prostaglandin pathway in endothelial cells. Specifically, the loss of SIRT6 in HUVECs resulted in a significant increase in COX-2 expression concomitant with an increase in PGE 2 and PGF 2α release. MMPs play a key role in the processes of remodelling and reorganisation of the vasculature, which is a key feature of endothelial dysfunction [34,35]. PAI-1, a member of the serine protease inhibitor family, is a key inhibitor of fibrinolysis; impaired endothelial fibrinolytic potential contributes to the development of thrombotic attacks and atherosclerosis [36,37]. In addition, PAI-1 is also involved degradation of the ECM [36,37]. In this study, siRNA knockdown of SIRT6 in HUVECs increased PAI-1 and MMP-9 expression. This is the first study to demonstrate a role for SIRT6 in the regulation of matrix proteins in endothelial cells; however, a role for SIRT1 has been reported [38].
NF-κB is a master regulatory component driving many inflammatory diseases; it is thought to play an important role in endothelial cell function [1,39,40]. There is now much evidence to show that SIRT1 exerts many of its protective effects in endothelial cells by inhibiting NF-κB activity [41]. In this study, knockdown of SIRT6 was associated with an increase in NF-κB expression. Additionally, using luciferase assays, NF-κB transcriptional activity was significantly reduced in HUVECs that also overexpressed SIRT6. This decrease was also associated with a decrease in the expression of NF-κB target genes, namely, IL-6 and PGF 2α .
Collectively, this suggests that in the presence of SIRT6, HUVECs display hypoactive NF-κB, leading to decreased transcription of proinflammatory genes. Our data is in support of recent studies that report increased expression of NF-κB and its target genes in a number of tissues of SIRT6deficient mice [21].

Conclusions
Functional alterations of the endothelium are hallmark features of many inflammatory diseases, including diabetes, hypertension, atherosclerosis, rheumatoid arthritis, and aging [1,[42][43][44]. Therefore, the inhibition of inflammatory mediators in activated endothelial cells might be of great benefit in the maintenance of endothelial homeostasis and the prevention of vascular diseases. This study highlights the relevance of SIRT6 as an important player in inflammatory reactions of the human endothelium in the context of sepsis. There is now growing literature implicating gut bacteria on the development of insulin resistance [45], cardiovascular disease [46], and atherosclerosis [47]. It is, therefore, tempting to speculate that activators of SIRT6 might be an effective strategy for a number of inflammatory vascular diseases such as diabetes and atherosclerosis.

Conflict of Interests
There is no conflict of interests that could be perceived as prejudicing the impartiality of the research reported.