gp91ds-tat | Colforsin

Intermittent local periodontal inﬂammation causes endothelial dysfunction of the systemic artery via increased levels of hydrogen peroxide concomitantly with overexpression of superoxide dismutase

a b s t r a c t
Background: The present study was designed to examine whether the intermittent local periodontal inﬂamma- tion induces endothelial dysfunction of the systemic artery caused by oxidative stress and if increased levels of hydrogen peroxide coexisted with overexpression of superoxide dismutase (SOD) as well as NADPH oxidase con- tribute to the oxidative stress.
Methods: The rats in lipopolysaccharides (LPS) group received 1500 μg LPS injection to bilateral gingiva of the lower jaw a week interval from eight- to eleven-week-old. Isolated mandibles or aortas were subjected to the evaluation of histopathological changes, isometric force recordings, reactive oxygen species using 2′,7′- dichloroﬂuorescin diacetate (10−5 mol/L) and protein expression of NADPH oxidase subunits and SOD, respec- tively.
Results: Mandible sections demonstrated the periodontal inﬂammation only in the LPS group at three days, but not seven days, after the LSP injection. Acetylcholine (10−9 to 10−5 mol/L)-induced relaxation was reduced only in aortas from the LPS group. Gp91ds-tat and PEG-catalase restored the impaired dilation in arteries from the LPS group. Levels of reactive oxygen species were enhanced in aortas from the LPS group, whereas the incre- ment was abolished by the treatment with gp91-ds-tat or PEG-catalase. Expression of a NADPH oxidase subunit p47phox and CuZn-SOD increased in the LPS group.
Conclusions: The intermittent local periodontal inﬂammation induces systemic endothelial dysfunction caused by overproduction of reactive oxygen species in the systemic artery of rats and that overexpression of CuZn-SOD as well as a NADPH oxidase cytosolic subunit contributes to increased levels of hydrogen peroxide in blood vessels of this animal model.

1.Introduction
Periodontal disease is one of the potential causes of subclinical ath- erosclerotic cardiovascular diseases in humans whereas the mecha- nisms are unclear [1]. The removal of plaque, at least, every 24 h requires to the prevention of gingival inﬂammation even in subjects without periodontal diseases, indicating that the avoidance of mild or intermittent periodontal inﬂammation plays a key role in dental health [2]. Whether such intermittent, as well as mild, periodontal inﬂamma- tion affects the systemic vascular dysfunction is, however, has been un- known since previous studies with the periodontal disease animal models intended to reproduce the continuous as well as severe peri- odontal inﬂammation with the alveolar bone loss [3].The prolonged endotoxemia caused by a significant amount of LPS application enhances the production of superoxide resulting from the activation of NADPH oxidase [4]. Chronic systemic inﬂammation increases levels of hydrogen peroxide in cardiovascular tissues [5]. Su- peroxide, which undergoes dismutation by superoxide dismutase (SOD), results in the hydrogen peroxide production [6]. These findings draw the idea that increased levels of hydrogen peroxide coexisting with the overexpression of SOD as well as NADPH oxidase may also con- tribute to the oxidative stress in the chronic local inﬂammation.The present study has produced a rat model of local inﬂammation caused by the mild periodontal disease with the intermittent cure within the several day’s intervals. It aimed at evaluation whether this pathological condition impairs the acetylcholine-induced endothelium-dependent relaxation in the aorta, whether inhibitors of NADPH oxidase and hydrogen peroxide restore the relaxation, and whether the disease state causes arterial oxidative stress including in- creased levels of hydrogen peroxide coexisting with overexpression of SOD as well as NADPH oxidase.

2.Methods
Mean blood pressure and heart rate in rats at eight- and twelve-week-old in each group.Eight-week-old male Wistar rats (Japan SLC, Inc., Hamamatsu, Japan) in the LPS group received LPS (Sigma Aldrich Inc., St. Louis, MO, USA) total 1500 μg (750 μg for each side) injection to bilateral gingiva of the lower jaw at the first molar with a week interval from eight- to eleven-week-old (Fig. 1). Rats in the LPS and Sham groups similarly had the intraperitoneal injection of LPS (150 μg) at eight-week-old to add immunological sensitization [7]. Rats in the Normal group received normal saline with the same volume for the LPS group. The hemody- namics including systemic blood pressure and heart rate were noninva- sively evaluated using the tail-cuff method (BP-98E; Softron, Tokyo, Japan) with a week interval from eight- to twelve-week-old. The injected dose of LPS in the current study was non-lethal as the previous studies reported [8] and did not induce hemodynamic changes as shown in Table 1. Above procedures were performed under inhalation of 2 L/min air with 3% isoﬂurane in the incubator heated to 38 °C. The concentration of isoﬂurane was selected to obtain surgical anesthetic condition considering the minimum alveolar anesthetic concentration of isoﬂurane that produces immobility in 50% of subjects given a nox- ious stimulation in rats has been reported as about 1.3% [9]. Immediately after hemodynamic measurements at twelve-week- old, rats were euthanatized by exsanguination under inhalation of 2 L/ min air with 3% isoﬂurane. The thoracic aorta and mandible were quickly isolated and were used for the histopathological analysis, organ chamber experiments, measurements of in situ reactive oxygen species production and Western immunoblotting analysis. The rats in LPS (lipopolysaccharide) group received LPS total 1500 μg injection to bilateral gingiva of the lower jaw at the first molar four times with a week interval from eight- to eleven-week-old. Rats in the LPS and Sham groups similarly had the intraperitoneal injec- tion of LPS (150 μg) at eight-week-old. Rats in the Normal group received normal saline with the same volume for LPS group. Data are shown as the means ± SD, n = 7 each.

Isolated mandibles, which were fixed in 10% neutral-buffered for- malin for 24 h at 4 °C, were subsequently put in the 10% solution of formic acid for six weeks at 4 °C. After the decalcification, the specimens were trimmed, dehydrated in a sequential ethanol series using an auto- mated processor and embedded in paraffin wax. Serial 6-μm-thick sec- tions through the centers of the first molar on the buccal-lingual plane were obtained, stained with hematoxylin and eosin.The periodontal inﬂammation was histopathologically evaluated using the light microscopy. A blinded examiner, who did not know the group to which the rat belonged, examined the degree of inﬂammatory cell infiltration in a randomly chosen section of the interproximal area within 1 × 1 mm square using the inﬂammatory score as follows: 0, no inﬂammatory cells; 1, slight inﬂammation (a few inﬂammatory cells in the gingival connective tissue); 2, moderate inﬂammation (sev- eral inﬂammatory cells in the gingival connective tissue); and 3, severe inﬂammation (more than 1/3 of the cells in the gingival connective tis- sue were inﬂammatory cells) [10] The upper and lateral views of the LPS injection site in the current study. LPS total 1500 μg (750 μg each) was injected to bilateral gingival of the lower jaw at the first molar indicated with black arrows.
Aortic rings of 2.5 mm in length were connected to an isometric force transducer and suspended in an organ chamber filled with 10 mL of Krebs-Ringer bicarbonate solution (the control Krebs-Ringer solution, 37 °C, pH 7.4) bubbled with 95% O2–5% CO2 gas mixture of the following composition (mmol/L): NaCl 119, KCl 4.7, CaCl2 2.5, MgSO4 1.17, KH2PO4 1.18, NaHCO3 25, and glucose 5.5. Each ring was gradually stretched to the optimal point of its length-tension curve as determined by the contraction to the α adrenoceptor agonist phenyl- ephrine (3 × 10−7 mol/L). The optimal resting force was achieved at ap- proximately 1.5 g. Some rings were incubated with a non-selective nitric oxide synthase inhibitor Nω-nitro-L-arginine methyl ester hydro- chloride (L-NAME, 3 × 10−4 mol/L), a superoxide inhibitor Tiron (10 mmol/L), a selective NADPH oxidase inhibitor gp91ds-tat (10−6 mol/L) or a selective hydrogen peroxide inhibitor catalase- polyethylene glycol (PEG-catalase, 200 U/mL) for 15 min. 5 min after completion of the incubation, phenylephrine (3 × 10−7 mol/L) was added to the organ chambers to contract arterial rings. During submax- imal contraction, which reached plateau 10 to 15 min after the addition of contractile agents, concentration-response curves to acetylcholine (10−9–10−5 mol/L) or a nitric oxide donor 1-Hydroxy-2-oxo-3-(N- methyl-3-aminopropyl)-3-methyl-1-triazene (NOC-7, 10−9– 10−5 mol/L) were obtained in a cumulative fashion with three to 6 min interval. The relaxation was expressed as a percentage of the maximal relaxation in response to papaverine (3 × 10−4 mol/L), which was added at the end of experiments to produce the maximal re- laxation (100%) of arterial rings.

Each isolated aorta with endothelium was immediately frozen at
−80 °C. The twenty-micrometer unfixed sections were cut on a cryostat and mounted onto microscope slides. They had further incubation in the control Krebs-Ringer solution with or without PEG-catalase (200 U/mL) or gp91ds-tat (10−6 mol/L). The phosphate-buffered saline (pH 7.4) containing 2′,7′-dichloroﬂuorescin diacetate (DCF-DA, 10−5 mol/L) as well as Hoechst 33258 (1 μg/mL) was applied to each section in a light-protected chamber at 37 °C for 20 min. Images of cellular ﬂuores- cence were acquired using a microscope fitted with BZ-II analyzer soft- ware (Model BZ-9000 Generation II, Keyence, Osaka, Japan). Settings were identical for the acquisition of images from all of the arterial slices. The relative ratio according to the control ﬂuorescence of aortic rings treated with PEG-catalase (200 U/mL) in the Normal group was used to standardize the DCF-DA ﬂuorescence in each specimen.The isolated aortas with endothelium were quickly frozen at −80 °C. Cytosolic and membranous fractions were prepared and used for West- ern immunoblotting analysis [11]. Blood vessels, which were stored at −80 °C until use, were powdered under liquid nitrogen and solved in 300 μL cell permeabilization buffer containing 1% protease inhibitor. The lysate was centrifuged at 500 ×g for 15 min at 4 °C, and the super- natant ﬂuid was used for the measurement of total protein levels. A por- tion of the supernatant ﬂuid was centrifuged at 100,000 ×g for 60 min at 4 °C and used as a cytosolic fraction, and the pellet was used as a mem- brane fraction. The protein concentrations in the cytosolic, as well as membrane fractions, were estimated by the BCA protein assay (Thermo Fisher Scientific Inc., Rockford, IL, USA). The same amount of protein was separated by SDS–PAGE and transferred to PVDF membranes (Immobilon-P, EMD Millipore Corp, Billerica, MA, USA). These mem- branes were assessed with antibodies against p47-phox (1:500 dilu- tion; Millipore Corp., Billerica, MA, USA), Rac1 (1:500 dilution; Cell Signaling Technology, Inc., Danvers, MA, USA), NOX2 (1:500 dilution; Cell Signaling Technology, Inc., Danvers, MA, USA), CuZn-SOD (1:500 dilution; Millipore Corp., Billerica, MA, USA) and Na+/K+-ATPase (1:500 dilution; Cell Signaling Technology, Inc., Danvers, MA, USA) for the membrane fraction, and CuZn-SOD (1:500 dilution; Millipore Corp., Billerica, MA, USA) and glyceraldehyde-3-phosphate dehydroge- nase (GAPDH, 1:500 dilution; GeneTex, Inc., Irvine, CA, USA) for the cy- tosolic fraction. The results were quantified based on the expression level of Na+/K+-ATPase or GAPDH using the Image Analyzer System (ImageQuant LAS4000, GE Healthcare UK Ltd., Pollards Wood, UK).

The following pharmacological agents were used: acetylcholine, catalase-polyethylene glycol (PEG-catalase), dimethyl sulfoxide, Nω- Nitro-L-arginine methyl ester hydrochloride (L-NAME), papaverine, phenylephrine and Tiron (Sigma Aldrich Inc., St. Louis, MO, USA), Hoechst 33258 (Nacalai Tesque, Kyoto, Japan), 1-Hydroxy-2-oxo-3- (N-methyl-3-aminopropyl)-3-methyl-1-triazene (NOC-7, Dojindo Mo- lecular Technologies, Inc., Kumamoto, Japan), formic acid (Wako Pure Chemical Industries, Ltd., Osaka, Japan), 2′,7′-dichloroﬂuorescin diacetate (DCF-DA, Goryo Chemical, Inc., Sapporo, Japan) and gp91ds- tat (Genemed Synthesis Inc., San Antonio, TX, USA). Drugs were dis- solved in distilled water such that volumes of b 60 mL are added to the perfusion system. The stock solution of Tiron was prepared in di- methyl sulfoxide, and the highest concentration of dimethyl sulfoxide was 1.74 × 10−6 mol/L. Our preliminary confirmed that this vehicle does not affect vasomotor function in our experimental condition [12]. The concentrations of drugs are expressed as the final molar concentration.Statistical analysis was performed using PASW Statistics 18™ (IBM Japan Inc., Tokyo, Japan). The continuous variables were expressed as the means ± SD and the data were analyzed by one-way ANOVA or two-way ANOVA with Scheffe’s test. The ordinal variables were expressed as the median (interquartile range), and the data were ana- lyzed Mann-Whitney U test. Differences were considered to be statisti- cally significant when P is b 0.05.

3. Results
Mean blood pressure and heart rate in rats at eight- and twelve- week-old did not differ among the Normal, Sham, and LPS groups (Table 1).The histopathological analysis using mandible sections stained with hematoxylin and eosin verified the existence of intermittent periodon- tal inﬂammation in the LPS group. Sections through the centers of the first molar on the buccal-lingual plane demonstrate inﬂammatory cells infiltration only in the LPS group at three days, but not seven days, after the last LPS injection (Fig. 2). The mandible sections applied to the inﬂammatory cell infiltration score also demonstrated the signif- icant inﬂammatory cell infiltration only in the LPS group at three days after the last LPS injection (P b 0.01, Table 2). The sections after the first LPS injection showed the same results as above findings (n = 5, data not shown).Acetylcholine caused the concentration-dependent relaxation in aortas from all groups, whereas L-NAME abolished it (P b 0.05 or 0.01, Fig. 3A). Acetylcholine-induced relaxation was reduced only in aortas. The representative pictures of mandible sections stained with hematoxylin and eosin in the Normal and LPS groups (n = 6 each). The sections through the centers of the first molar on the buccal-lingual plane demonstrated inﬂammatory cells infiltration only in the LPS group at three days (within the yellow square), but not seven days, after the last LPS injection. The yellow bar indicates 500 μm. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) from the LPS group (P b 0.05 or 0.01, Fig. 3A). The NOC-7-induced, endothelium-independent relaxation did not differ between the LPS and Normal groups (Fig. 3B). Gp91ds-tat and PEG-catalase, but not Tiron, restored the impaired dilation in aortas from the LPS group (P b 0.05, Fig. 4A&B).Levels of reactive oxygen species were enhanced in aortas from the LPS group (P b 0.01), whereas the increment was abolished by the treat- ment with gp91ds-tat or PEG-catalase (Fig. 5A&B).
Expression of a NADPH oxidase subunit p47phox and CuZn-SOD in- creased in the membrane protein fraction of isolated aortas in the LPS group (P b 0.01) whereas that of NADPH oxidase subunits rac1 and NOX2 in the membrane fraction and CuZn-SOD in the cytosolic fraction did not differ between the Normal and LPS groups (Fig. 6A&B).

4.Discussion
The histopathological analysis including the inﬂammatory cell infil- tration score evaluation showed the existence of the periodontal inﬂammation at three days, but not seven days, after the last as well as the first LPS injection. The oral pathology verifies that the intermit- tent periodontal inﬂammation developed in rats of the LPS group. More- over, the inﬂammatory cell infiltration did not coexist with the periodontal pocket formation as well as the alveolar bone resorption, in- dicating that the inﬂammatory periodontal model in the current study is not severe enough to diagnose as periodontitis [3,10,13,14]. It is cru- cial to note that the mild intermittent periodontal inﬂammation did not cause changes in the blood pressure and heart rate as previous ani- mal models of experimental periodontitis showed them, indicating the negligible role of hemodynamic changes derived from the gingival LPS application in the study results [3,15]. Acetylcholine-induced relaxation was reduced only in aortas from the LPS group whereas L-NAME abolished it in aortas from all groups, and NOC-7-induced relaxation did not differ between the LPS and Nor- mal groups. L-NAME or NOC-7 is the selective inhibitor of nitric oxide synthase and nitric oxide donor, respectively [16,17]. The intermittent periodontal inﬂammation resulted from the gingival LPS application ap- pears to impair the systemic vasodilation mediated by endogenous ni- tric oxide derived from the synthase. These results are consistent with the previous studies in humans and animals regarding periodontitis, demonstrating the endothelial dysfunction in the systemic arteries [3, 15,18,19].

The intermittent periodontal inﬂammation enhanced levels of reac- tive oxygen species in aortas, which was abolished by the treatment with gp91ds-tat or PEG-catalase. Gp91ds-tat is a competitive inhibitor of the NADPH oxidase cytosolic subunit p47phox assembly to the mem- branous subunits [20], and the PEG-catalase is a selective hydrogen per- oxide scavenger [21]. These results indicate that the endothelial dysfunction of the current study is due to hydrogen peroxide via activa- tion of NADPH oxidase. This conclusion is further supported by the re- sults that the protein expression of a NADPH oxidase subunit p47phox increased in the membrane fraction of aortas in the LPS group. Why the experimental condition of this study enhanced hydrogen peroxide production is critical to note since the NADPH oxidase is an oxidoreduc- tase to produce superoxide as the primary product [22]. Superoxide un- dergoes the dismutation by SOD, resulting in the increased levels. The endothelium-dependent relaxation in response to acetylcholine (Left, 10−9–10−5 mol/L) or endothelium-independent relaxation to NOC-7 (Right, 10−9–10−5 mol/L) in aortas from the Normal, Sham, and LPS groups, respectively. L-NAME (3 × 10−4 mol/L) was applied to some aortic rings. Differences between the control rings in the Normal group and those in the LPS or Sham group, the control rings in each and those treated with L-NAME are statistically significant (#: P b 0.05, *: P b 0.01). The data were expressed as means ± SD; n refers to the number of rats from which the artery was taken, and they were expressed as percent of maximal relaxation induced by papaverine (3 × 10−4 mol/L)hydrogen peroxide [6]. The protein expression of CuZn-SOD enhanced in the aortas from the LPS group, indicating that the intermittent peri- odontal inﬂammation induces the compensatory overexpression of the enzyme enough to dismutate the superoxide derived from the NADPH oxidase. These findings are in agreement with the previous studies that increased hydrogen peroxide coexisted with the overex- pression of SOD as well as NADPH oxidase in cardiovascular tissues con- tributes to the oxidative stress in the chronic inﬂammation [5](Left) The endothelium-dependent relaxation in response to acetylcholine (10−9–10−5 mol/L) in aortas from the Normal and LPS groups in the presence or absence of Tiron (10 mmol/L) or gp91ds-tat (10−6 mol/L). #Differences between the control rings in the Normal group and control rings in the LPS group or rings treated with Tiron or gp91ds-tat in the LPS group are statistically significant (P b 0.05). (Right)

The relaxation in response to acetylcholine in aortas from the Normal and LPS groups in the presence or absence of PEG- catalase (200 U/mL). #Difference between the control rings in the Normal group and those in the LPS group is statistically significant (P b 0.05). The data were expressed as means ± SD; n refers to the number of rats from which the artery was taken, and they were expressed as percent of maximal relaxation induced by papaverine (3 × 10−4 mol/L)(Left) Representative images of in situ reactive oxygen species, which were labeled with the green ﬂuorescence using 2′,7′-dichloroﬂuorescin diacetate (DCF-DA, 10−5 mol/L) and those of without the ﬂuorescence labeling (the differential interference contrast image), in the aortic rings with 20 μm thick. (Right) Relative levels of reactive oxygen species in the aortic rings treated with or without PEG-catalase (200 U/mL) or gp91ds-tat (10−6 mol/L) in the Normal and LPS groups. The relative ratio according to the control ﬂuorescence of aortic rings treated with PEG-catalase (200 U/mL) in the Normal group was used to standardize the DCF-DA ﬂuorescence in each specimen. *Differences between the control rings in the Normal group and control rings or rings treated with Tiron or gp91ds-tat in the LPS group are statistically significant (P b 0.01). The data were expressed as means ± SD; n refers to the number of subjects from which the artery was taken.(Upper) Representative images of the protein expression of p47phox, rac1, NOX2, Na+/K+ ATPase and CuZn-SOD in the membrane fraction, and CuZn-SOD and GAPDH in the cytosolic fraction from the aortas in the Normal and LPS groups. (Lower) In the bar graph, the band density in the membrane and cytosolic fractions was expressed as means ± SD.Difference between the rings in the Normal and those in the LPS group is statistically significant (P b 0.01). The results were quantified based on the expression level of Na+/K+- ATPase or GAPDH; n refers to the number of rats from which the artery was taken.Moreover, hydrogen peroxide production correlated well with the plasma renin activity, suggesting the causal and critical roles in the chronic cardiovascular diseases in humans [23]. The oxidative stress in response to the intracellular redox state, however, seems to be depen- dent on the seriousness of inﬂammation since the complete experimen- tal periodontitis induces increased levels of superoxide, but not hydrogen peroxide, in the systemic arterial walls, which is less likely in the cardiovascular tissues under chronic inﬂammation of humans [3,5,23].The relationship between cardiovascular diseases and chronic in- ﬂammation including periodontitis has long been suggested although mechanisms of the interlink remain unclear in detail [24]. None of the pre-existing biomarkers, including cytokines, proteins, and enzymes, provides the definite answer to the above question [25,26]. To find the new inﬂammatory biomarker that alone or in the combination of some is, therefore, currently necessary to explain the cardiovascular pa- thology related to chronic inﬂammation. Indeed, previous studies on periodontitis in humans and animals demonstrated the increased levels of systemic C-reactive protein, matrix metalloproteinases, and cyto- kines, indicating the involvement of several inﬂammatory biomarkers [18,19,27]. Especially, roles of cytokines including interleukin-8, which is associated with adverse outcome of chronic heart failure may have to be studied in the future [28].

5.Conclusion
The present study is the first to demonstrate that the intermittent local periodontal inﬂammation induces the systemic arterial endothelial dysfunction caused by overproduction of reactive oxygen species in the rats and that overexpression of CuZn-SOD as well as a NADPH oxidase contributes to increased levels of hydrogen peroxide in blood vessels of this animal model. These results verified that the local periodontal disease is one of the therapeutic targets to avoid oxidative stress toward systemic blood vessels, and suggest that the oral gp91ds-tat care to prevent the in- termittent and mild periodontal inﬂammation is increasing important to protect cardiovascular tissues from the oxidative stress.