Ephx2‑gene deletion affects acetylcholine‑induced relaxation in angiotensin‑II infused mice: role of nitric oxide and CYP‑epoxygenases

Abstract
Previously, we showed that adenosine A2A receptor induces relaxation independent of NO in soluble epoxide hydrolase- null mice (Nayeem et al. in Am J Physiol Regul Integr Comp Physiol 304:R23–R32, 2013). Currently, we hypothesize that Ephx2-gene deletion affects acetylcholine (Ach)-induced relaxation which is independent of A2AAR but dependent on NO and CYP-epoxygenases. Ephx2−/− aortas showed a lack of sEH (97.1%, P < 0.05) but an increase in microsomal epoxide hydrolase (mEH, 37%, P < 0.05) proteins compared to C57Bl/6 mice, and no change in CYP2C29 and CYP2J protein (P > 0.05). Ach-induced response was tested with nitro-L-arginine methyl ester (L-NAME) NO-inhibitor; 10−4 M), N-(methylsulfonyl)-2-(2-propynyloxy)-benzenehexanamide (MS-PPOH) (CYP-epoxygenase inhibitor; 10−5 M), 14,15-epox- yeicosa-5(Z)-enoic acid (14,15-EEZE, an epoxyeicosatrienoic acid-antagonist; 10−5 M), SCH-58261 (A2AAR-antagonist; 10−6 M), and angiotensin-II (Ang-II, 10−6 M). In Ephx2−/− mice, Ach-induced relaxation was not different from C57Bl/6 mice except at 10−5 M (92.75 ± 2.41 vs. 76.12 ± 3.34, P < 0.05). However, Ach-induced relaxation was inhibited with L-NAME (Ephx2−/−: 23.74 ± 3.76% and C57Bl/6: 11.61 ± 2.82%), MS-PPOH (Ephx2−/−: 48.16 ± 6.53% and C57Bl/6: 52.27 ± 7.47%), and 14,15-EEZE (Ephx2−/−: 44.29 ± 8.33% and C57Bl/6: 39.27 ± 7.47%) vs. non-treated (P < 0.05). But, it did not block with SCH-58261 (Ephx2−/−: 68.75 ± 11.41% and C57Bl/6: 66.26 ± 9.43%, P > 0.05) vs. non-treated (P > 0.05). Interestingly, Ang-II attenuates less relaxation in Ehx2−/− vs. C57Bl/6 mice (58.80 ± 7.81% vs. 45.92 ± 7.76, P < 0.05). Our data suggest that Ach-induced relaxation in Ephx2−/− mice depends on NO and CYP-epoxygenases but not on A2A AR, and Ephx2-gene deletion attenuates less Ach-induced relaxation in Ang-II-infused mice.

Introduction
The nitric oxide (NO) and endothelium-derived hyperpolar- izing factor (EDHF) are released from vascular endothe- lium in human arteries [1, 2]. Epoxyeicosatrienoic acids (EETs) act as EDHFs, which are cytochrome-P450 epox- ygenase-derived metabolites (from arachidonic acid, AA) [3, 4]. EETs can cause hyperpolarization in canine coro- nary arteries [5–7]. The enzyme soluble epoxide hydrolase (sEH) metabolizes EETs to dihydroxyeicosatrienoic acids (DHETs), and its sEH inhibition or deletion could increase EET activity [8–13].Cytochrome P-450-epoxygenases, including CYP2C (CYP2C29) and CYP2J (CYP2J2, CYP2J5), have beendetected on endothelium [4, 11, 14–21]. Yang et al. [22] and Hanif et al. [11] showed upregulation of CYP2J2 protectsvascular endothelium against hypoxia-reoxygenation injury/ ischemia reperfusion injury and enhanced coronary reac- tive hyperemic response [11, 22]. Endothelial CYP2C and CYP2J-epoxygenases are the main source of epoxide gen- eration in the vascular system [4, 14, 23]. EETs present in small-resistance vessels, coronary arteries, and the aorta [11–13, 17, 24, 25]. In addition, EETs play an important role in the relaxation of isolated coronary arteries at a low concentration and are involved in increased coronary reac- tive hyperemic response [11–13, 24, 25]. EETs-induced responses have been seen in both bovine coronary arteries and mouse aortas which is inhibited by the 14,15-epoxyei- cosa-5(Z)-enoic acid (EET antagonist) [18, 19, 25, 26].The sEH enzyme (Ephx2 gene) is commonly present in various mammalian tissues, including the kidney, intestine, liver, heart, coronary arteries, and aortas [9–13, 25, 27–29]. sEH metabolizes four forms of EET regioisomers to their corresponding diols [27, 30].

In the presence of CYP-epox- ygenases, conversion of arachidonic acids into epoxides then into diols by sEH, diminishes the beneficial effects on car- diovascular functions. Inhibition or deletion of sEH enables beneficial epoxides to accumulate and stay for longer periods after they are generated from arachidonic acids [12, 13, 24, 25]. Several reports indicate that sEH inhibition could be a promising approach for controlling of hypertension and vascular diseases [31–34]. However, Hercule et al. indi- cated that there is a possibility CYP450-eicosanoids activate eNOS and release NO in sEH−/− and WT mouse mesenteric arteries [35, 36]. NO and EETs maintain a stable arterial wall viscosity during blood flow increase, and this endothe- lial adaptive regulation is lost during essential hypertension [37]. The sEH inhibitor improved endothelium-dependent relaxations to acetylcholine, assessed by myography in iso- lated coronary arteries. This improvement was related to a restoration of EETs and NO pathways, as shown by the increased inhibitory effects of the NOS and cytochrome P-450 epoxygenase inhibitors l-NA and MS-PPOH on these relaxations [38].

In contrast, we demonstrated that adenosineA2A receptor-induced relaxation response was NO-independ- ent in sEH−/− mice, where CYP450-epoxygenases-derivedEETs may play an important role in mouse aortic relaxa- tion [25]. In addition, sEH inhibition reduced the blood pressure in Ang-II-induced hypertension [34, 39] and sEH deletion restricted the reduction of coronary reactive hyper- emic response in Ang-II-infused sEH−/− vs. WT mice (un- published data from our lab). Taking all of these findings into consideration, we hypothesized that Ephx2-gene dele- tion affects Ach-induced relaxation which is independent of A2AAR but dependent on NO and CYP-epoxygenases.Sinal et al. [40] described the generation of Ephx2−/− mice. Ephx2−/− and C57Bl/6 mice were provided by Dr. Dar- ryl Zeldin, NIEHS/NIH. West Virginia University Insti- tutional Animal Care and Use Committee approved all animal care and experimentation protocols and were in accordance with the principles and guidelines of the NIH’s Guide for the Care and Use of Laboratory Animals. We used both male and female mice (14–18 wk. old) equally in our study.Phenylephrine and ACh (Sigma Chemicals, St. Louis, MO) and Angiotensin-II (Human), BACHEM (Bubendorf, Switzerland) dissolved in distilled water. 14,15-EEZE and MS-PPOH (Cayman Chemicals), and SCH 58261 and L-NAME (Sigma Chemicals, St. Louis, MO) were dis- solved in DMSO [25]. CYP2J2 and Cyp2c29 (Dr. Zeldin, NIEHS/NIH), sEH (Santa Cruz Biotechnology), mEH (Institut für Pharmakologie undToxikologie, Zürich, Swit- zerland), and β-actin (Santa Cruz Biotechnology) antibod- ies were obtained and used for Western blot experiments [25].Mice were euthanized with pentobarbital sodium (100 mg/ kg ip).

According to our previously described protocol [17–20, 41–48], after thoracotomy, the aortas were gently removed and cleaned of fat and connective tissue [25]. In brief, aortas from both Ephx2−/− and C57Bl/6 mice were treated with 1 ml of lysis buffer for protein extraction [25]. Gel electrophoresis and Western blot analysis were done according to the protocol described by our proto- col [17–20, 41–48]. Following blocking with nonfat dry milk, the nitrocellulose membranes were incubated with polyclonal primary antibodies raised against for CYP2J2 and CYP2C29 (Dr. Zeldin, NIEHS/NIH), monoclonal sEH (Santa Cruz Biotechnology) microsomal epoxide hydro- lase (mEH, Institut für Pharmakologie und Toxikologie, Zürich, Switzerland), and β-actin (Santa Cruz Biotechnol- ogy) [25]. We used the secondary antibody, horseradish peroxidase-conjugated anti-rabbit IgG [25]. The mem- branes were developed using enhanced chemiluminescence(Amersham Biosciences) and exposed to X-ray film for appropriate time [25]. Mouse does not have CYP2J2, but antibody likely to recognize some of the mouse CYP2Js.Ephx2−/− and C57Bl/6 mice were euthanized with pento- barbital sodium (100 mg/kg intraperitoneally) [25].

The aorta was removed after thoracotomy and cut into rings of 3–4 mm in length [25]. The rings were suspended between the two wire hooks [25]. Two rings were suspended in organ baths containing 10 ml of modified Krebs–Hense- leit buffer [17–21, 25, 42, 44, 49]. After the equilibration period (60 min), tissues were contracted with KCl (50 mM) to assess the viability of the tissue [25]. Rings were then constricted with phenylephrine (PE; 10−7 M), and tension was monitored continuously with a fixed range precision force transducer (125C; BIOPAC Systems) connected to amplifiers (Data Acquisition system 100B; BIOPAC Systems) [25]. Data were recorded using MP100 WSW, BIOPAC digital acquisition system and analyzed using Acknowledge 3.5.7 software (BIOPAC Systems) [25]. Ach- dose (10−12–10−5 M)-dependent response experiments were conducted as previously described by our laboratory with adenosine agonist (NECA), adenosine A1 receptor agonist (CCPA) and adenosine A2A receptor agonist (CGS 21680) [17–21, 25, 42, 44, 49]. The aortic rings were washed several times with Krebs–Henseleit solution and allowed to equili- brate for 30 min before the experimental protocol began. Results are expressed as percentages downward or upward on PE-induced precontraction [25].Effect of nitric oxide inhibitor (L-NAME) on ACh- induced response in Ephx2−/− and C57Bl/6 mice. Concen- tration–response curves (CRCs) were obtained by cumula- tive addition of drugs in 1-log increments as described by us [17–21, 25, 42, 44, 49]. A single CRC was constructed for each ring in parallel in pairs of rings from either Ephx2−/− and C57Bl/6 mice in the same organ bath [25]. L-NAME (10 µM & 100 µM) was added 30 min before the PE contraction and was present throughout the ACh CRC.Effects of N-(methylsulfonyl)-2-(2-propynyloxy)-ben- zenehexanamide (MS-PPOH) (CYP-epoxygenase inhibitor on Ach-induced CRC in Ephx2−/− and C57Bl/6 mice.

Ach- CRC was obtained with and without MS-PPOH (10 µM), and MS-PPOH was added 30 min before PE contraction and was present throughout the ACh CRC [25], as similarly described earlier by us [11, 12, 17, 18, 21, 44, 49].Effects of epoxyeicosatrienoic acids (EETs) receptor antagonist 14,15-EEZE on Ach-induced CRC in Ephx2−/− and C57Bl/6 mice. Ach-CRC was obtained with and without 14,15-EEZE (10 µM), which was added 30 min before the PEcontraction and was present throughout the ACh CRC [25], as similarly described earlier by us [17–19, 25].Effect of A2A AR antagonists (SCH 58261) on Ach- induced CRC in Ephx2−/− and C57Bl/6 mice. Ach-inducedCRC obtained as described above [25]. SCH 58261 (1 µM) was added 30 min before PE contraction and was present throughout the experiment as described earlier by us [17, 18, 25].Effects of Angiotensin-II (Ang-II, 1 µM) on Ach-induced CRC in Ephx2−/− and C57Bl/6 mice. Ach-CRC was obtained with and without Ang-II (1 µM), and Ang-II was added 30 min before the PE contraction and was present through- out the ACh CRC [25], as similarly described earlier by us [20, 50].Statistical analysisData were reported as mean ± SE. One-way ANOVA was used to compare difference among groups, and two-way ANOVA was used for repeated measures, followed by Tukey post hoc test to compare the Ach-induced vascular responses of non-treated vs. inhibitor/antagonist (L-NAME, MS-PPOH, 14,15-EEZE, SCH 58261 and Ang-II)-treatedEphx2−/− and C57Bl/6 mice [25]. Differences (Ephx2−/− vs. C57Bl/6) between non-treated vs. (L-NAME, MS-PPOH, 14,15-EEZE, SCH 58261 and Ang-II)-treated Ephx2−/− and C57Bl/6 mice, and L-NAME-treated vs. (MS-PPOH, 14,15- EEZE, SCH 58261 and Ang-II)-treated Ephx2−/− and C57Bl/6 mice were considered significant when P < 0.05 [25]. Further, densitometry of Western blot analysis (mEH, CYP2J2, Cyp2c29 and sEH) expressed as mean ± SE in arbi- trary units [25]. All the statistical analyses were performed using GraphPad Prism statistical package [25].

Results
Western blot analysis showed 97% less sEH(~ 60 kDa) pro- tein in Ephx2−/− vs. C57Bl/6 mice (P < 0.05, Fig. 1a) but showed 37% more mEH (~ 50 kDa) protein in Ephx2−/− vs. C57Bl/6 mice (P < 0.05, Fig. 1B). There was no significant difference observed between Ephx2−/− vs. C57Bl/6 mice (P > 0.05, Fig. 2a, b) concerning CYP2C29 (~ 60 kDa) and CYP2J (~ 60 kDa) proteins.Concentration response curve (CRC) for AChand the effect of nitric oxide inhibitor in Ephx2−/−and C57Bl/6 miceACh caused a concentration (10−7–10−5 M)-dependent relaxation in both Ephx2−/− and C57Bl/6 mice, but theresponse was not significantly different (Ach-10−7 and 10−6 M, P > 0.05) except (Ach-10−5 M, P < 0.05) between Ephx2−/− and C57Bl/6 mice (92.75 ± 2.41 vs. 76.12 ± 3.34, Fig. 3). In addition, L-NAME (10 µM and 100 µM) had altered vascular response significantly (P < 0.05) in bothC57Bl/6 (10 µM L-NAME at Ach 10−5 M, 36.25 ± 2.64vs non-treated 76.12 ± 3.34, P < 0.05, Fig. 4a) andEphx2−/− (10 µM L-NAME at Ach 10−5 M, 48.79 ± 8.54vs non-treated 92.75 ± 2.41, P < 0.05, Fig. 4b). Simi- larly, with 100 µM L-NAME-treated C57Bl/6 (at Achobserved between (100 µM L-NAME)-treated Ephx2−/− vs. (100 µM L-NAME)-treated C57Bl/6 mice (P < 0.05, Fig. 5). At Ach 10−5 M, there is a significant difference between 100 µM L-NAME-treated C57Bl/6 (11.61 ± 2.82) vs.100 µM L-NAME-treated Ephx2−/− (23.74 ± 3.76, P < 0.05,Fig. 5).At 10−5 M Ach, MS-PPOH reduced the Ach-induced dose-dependent relaxation in C57Bl/6 (52.27 ± 7.47 vs. non-treated 76.12 ± 3.34, P < 0.05, Fig. 6a) as well as in Ephx2−/− mice (48.16 ± 6.53 vs. non-treated 92.75 ± 2.41, P < 0.05, Fig. 6b). No significant change was observed between MS-PPOH-treated C57Bl/6 vs. MS-PPOH- treated Ephx2−/− mice (P > 0.05, Fig. 7).

However, there was a significant difference in the reduction of Ach-induced relaxation observed between MS-PPOH (10−5 M)-treated vs. L-NAME (10−4 M)-treated C57Bl/6 mice (at 10−5 M Ach, MS-PPOH: 52.27 ± 7.47 vs. L-NAME: 11.61 ± 2.82,P < 0.05, Fig. 7). Similarly, there was a significant differ- ence in the reduction of Ach-induced relaxation observed between MS-PPOH (10−5 M)-treated vs. L-NAME (10−4 M)-treated Ephx2−/− mice (at 10−5 M Ach, MS- PPOH: 48.16 ± 6.53 vs. L-NAME: 23.74 ± 3.76, P < 0.05,Fig. 7). In addition, there was a significant difference in the reduction of Ach-induced dose (10−7 to 10−5 Mlar response in aortic rings of Ephx2−/− mice. *P < 0.05 between non-treated Ephx2−/− and 10−5 M L-NAME-treated Ephx2−/− mice.+P < 0.05 between non-treated Ephx2−/− and 10−4 M L-NAME- treated Ephx2−/− mice. Values expressed as mean ± SE, n = 8 per groupCRC for Ach with or without 14,15‑EEZE (EETs receptor antagonist, 10−5 M)‑treated C57Bl/6 and Ephx2−/− mice−5At 10 M Ach, 14,15-EEZE reduced the Ach-induceddose-dependent relaxation in C57Bl/6 (39.27 ± 7.47 vs. non-treated 76.12 ± 3.34, P < 0.05, Fig. 8a) as well as in Ephx2−/− mice (44.29 ± 8.33 vs. non-treated 92.75 ± 2.41, P < 0.05, Fig. 8b). There was no significant change observed between 14,15-EEZE-treated C57Bl/6 vs. 14,15-EEZE- treated Ephx2−/− mice (P > 0.05, Fig. 9). However, there was a significant difference in the reduction of Ach-induced relaxation between 14,15-EEZE-treated vs. L-NAME- treated C57Bl/6 mice (at 10−5 M Ach, 14,15-EEZE: 39.27 ± 7.47 vs. L-NAME: 11.61 ± 2.82, P < 0.05, Fig. 9).

Similarly, there was a significant difference in the reduc- tion of Ach-induced relaxation observed between 14,15- EEZE (10−5 M)-treated vs. L-NAME (10−4 M)-treated Ephx2−/− mice (at 10−5 M Ach, 14,15-EEZE: 44.29 ± 8.33C57Bl/6 mice. @P < 0.05 between non-treated Ephx2−/− and 10−4 M L-NAME-treated Ephx2−/− mice. Also, the effect of L-NAME (10−4 M) on Ach-induced dose-dependent vascular response between C57Bl/6 and Ephx2−/− mice. *P < 0.05 between non-treated Ephx2−/− and 10−5 M L-NAME-treated Ephx2−/− mice. *@P < 0.05 between the 10−4 M L-NAME-treated *C57Bl/6 and 10−4 M L-NAME-treated @Ephx2−/− mice. Values expressed as mean ± SE, n = 8 per groupAch)-dependent relaxation observed between MS-PPOH- treated vs. L-NAME-treated C57Bl/6 and Ephx2−/− mice (P < 0.05, Fig. 7).vs. L-NAME: 23.74 ± 3.76, P < 0.05, Fig. 9). In addition,there was a significant difference in the reduction of Ach- induced dose (10−7 to 10−5 M Ach)-dependent relaxation observed between 14,15-EEZE-treated vs. L-NAME-treated C57Bl/6 and Ephx2−/− mice (P < 0.05, Fig. 9).CRC for Ach with and without SCH 58261 (A2A AR‑antagonist) in Ephx2−/− and C57Bl/6 miceAch produced a concentration-dependent relaxation in both Ephx2−/− and C57Bl/6 mice. A2A AR-antagonist SCH 58261 produced no significant change in Ach-inducedMS-PPOH-treated C57Bl/6 mice. *P < 0.05 between non-treated Ephx2−/− and 10−5 M MS-PPOH-treated Ephx2−/− mice [25]. Values expressed as mean ± SE, n = 8 per group11.61 ± 2.82, P < 0.05, Fig. 11). Similarly, there was a signif- icant difference in the reduction of Ach-induced relaxation observed between SCH-58261-treated vs. L-NAME-treatedvs. L-NAME: 23.74 ± 3.76, P < 0.05, Fig. 11). In addition,there was a significant difference was observed between SCH 58261-treated vs. L-NAME-treated C57Bl/6 and−/−dose-dependent relaxation (P > 0.05, Fig. 10a, b).

However, there was a significant difference in the reduction of Ach- induced relaxation observed between SCH 58261 (10−6 M)- treated vs. L-NAME (10−4 M)-treated C57Bl/6 mice (at 10−5 M Ach, SCH 58261: 66.26 ± 9.43 vs. L-NAME:Ach produced a concentration-dependent relaxation in both Ephx2−/− and C57Bl/6 mice. Ang-II (1 µM) attenuates Ach- induced dose-dependent relaxation in both C57Bl/6 (at Ach 10−7–10−5 M; 9.06 ± 5.35, 32.71 ± 9.26, 45.92 ± 7.76vs. non-treated 37.24 ± 4.21, 59.29 ± 4.19, 76.12 ± 3.34,P < 0.05, Fig. 12a) and Ephx2−/− (at Ach 10−7–10−5 M;21.34 ± 4.23, 50.92 ± 5.95, 58.80 ± 4.81 vs. non-treated51.70 ± 3.91, 65.12 ± 9.95, 92.75 ± 2.41, P < 0.05, Fig. 12b)mice. Interestingly, Ang-II treatment attenuates more Ach- induced dose-dependent relaxation in C57Bl/6 (at Ach 10−7–10−5 M; 9.06 ± 5.35, 32.71 ± 9.26, 45.92 ± 7.76,Fig. 13a) vs. Ephx2−/− mice (at Ach 10−7–10−5 M; 21.34 ± 4.23, 50.92 ± 5.95, 58.80 ± 4.81 P < 0.05, Fig. 13a).Nevertheless, there was a significant difference in the reduc- tion of Ach-induced relaxation observed between Ang-II (10−6 M)-treated vs. L-NAME (10−4 M)-treated C57Bl/6 mice (at 10−5 M Ach, Ang-II: 45.92 ± 7.76 vs. L-NAME:11.61 ± 2.82, P < 0.05, Fig. 13b). Similarly, there was a sig- nificant difference in the reduction of Ach-induced relaxa- tion observed between Ang-II (10−6 M)-treated vs. L-NAME(10−4 M)-treated Ephx2−/− mice (at 10−5 M Ach, Ang-II: 58.80 ± 4.81 vs. L-NAME: 23.74 ± 3.76, P < 0.05, Fig. 13b).In addition, there was a significant difference in the reduc- tion of Ach-induced dose (10−7 to 10−5 M Ach)-dependent relaxation observed between Ang-II-treated vs. L-NAME- treated C57Bl/6 and Ephx2−/− mice (P < 0.05, Fig. 13b).

Discussion
The Ephx2−/− mice show no sEH protein expression com- pared to C57Bl/6 mice, whereas mEH (microsomal epox- ide hydrolase, Ephx1) was significantly upregulated in Ephx2−/− compared to C57Bl/6 mice (Fig. 1a, b). mEH is often expressed at lower levels than sEH and displays much lower rates of EET hydrolysis in vitro; however, it is anchored to ER membranes, interacts with P450 epoxyge- nases, and may participate in a coupled reaction to hydrolyze EETs during low, basal rates of EET formation [51–53]. Importantly, in sEH-deficient mice or tissues, mEH func- tionally compensates to hydrolyze the majority of EETs produced [53, 54]. The effect of mEH on EET is biologi- cally relevant, as mEH/sEH-null hearts have significantly improved post-ischemic heart function compared to sEH-null hearts [53]. Thus, the compensatory upregulation of mEH−/−mean ± SE, n = 8 per groupobserved in our Ephx2 mice may significantly affect EETshydrolysis in the current study. Therefore, mEH is upregu-lated in Ephx2−/− compared to C57Bl/6 mice, and there is no change in CYP2C29 and CYP2J protein expression in Ephx2−/− compared to C57Bl/6 mice (Fig. 2a, b). This is the first study to investigate the Ach-induced dose-dependentvascular (aortic) response interaction between CYP-epox- ygenases, soluble epoxide hydrolase, and nitric oxide in Ephx2−/− and C57Bl/6 mice. We are reporting some novel findings due to this interaction. First, L-NAME attenuates most of the Ach-induced dose-dependent relaxation in both Ephx2−/− and C57Bl/6 mice. Thus, it is clear that Ach- induced relaxation is almost all dependent on NO. However, attenuation of Ach-induced dose-dependent relaxation due to L-NAME in Ephx2−/− is less than C57Bl/6 mice.

Therefore,sEH nullification in Ephx2−/− mice plays a role in resisting the reduction of Ach-induced dose-dependent relaxation due to L-NAME in Ephx2−/− compared to C57Bl/6 mice. (1) CYP-epoxygenase inhibitor (MS-PPOH) was able to par- tially attenuate Ach-induced dose-dependent relaxation in both Ephx2−/− and C57Bl/6 mice, thus, demonstrating that CYP-epoxygenases are partially involved in Ach-induced dose-dependent relaxation in both Ephx2−/− and C57Bl/6 mice. (2) The EETs antagonist (14,15-EEZE) was also able to partially attenuate Ach-induced dose-dependent relaxation in both Ephx2−/− and C57Bl/6 mice. Therefore, EETs seem to be involved in Ach-induced dose-dependent relaxation in both Ephx2−/− and C57Bl/6 mice. (3) The A2A AR-antago- nist (SCH 58261) did not alter Ach-induced dose-dependent relaxation in both Ephx2−/− and C57Bl/6 mice. Hence, it appears that Ach-induced dose-dependent relaxation is inde- pendent of A2A AR but dependent on NO. (4) Ang-II also attenuates less Ach-induced dose-dependent relaxation in Ephx2−/− compared to C57Bl/6 mice. According to our data, sEH plays an important role in the in Ach-induced dose- dependent regulation of vascular response (Fig. 14).Genes related to Cytochrome P450-epoxygenases poly- morphism have been identified in various communities regarding its role in cardiovascular function and in blood pressure regulation [55–62]. Also, individuals with varia- tions in EPHX2 and ω-hdroxylase genes have an increased risk of coronary heart disease, ischemic stroke, restenosis, diabetes heart, heart failure, ischemic stroke in Caucasians, Chinese, and in the black Americans with increase in blood pressure [25, 55, 62–66].

Seubert et al. reported that dif- ferential renal sEH gene expression in pre-hypertensive, hypertensive, and spontaneously hypertensive rats (SHR)Ephx2 deletion on Ach-induced dose-dependent vascular response in aortic rings of C57Bl/6 and Ephx2−/− mice. *P < 0.05 between C57Bl/6 vs. Ephx2−/− mice only at 10−5 M of Ach. Also, effect of L-NAME and Ang-II on Ach-induced dose-dependent vascular response in aortic rings of C57Bl/6 and Ephx2−/− mice. #P < 0.05 between non-treated or Ang-II treated C57Bl/6 and L-NAME-treated C57Bl/6 mice. @P < 0.05 between non-treated or Ang-II-treated Ephx2−/− and L-NAME-treated Ephx2−/− mice. $@P < 0.05 between Ang-II-treated Ephx2−/− and L-NAME-treated Ephx2−/− mice.%#P < 0.05 between Ang-II-treated C57Bl/6 and L-NAME-treated C57Bl/6 mice. Values expressed as mean ± SE, n = 8 per group[25, 67], and mice lacking or inhibiting sEH have enhanced coronary reactive hyperemic response in Ephx2−/− and t-AUCB-treated mice [12, 13, 25]. However, Hercule et al. indicated that CYP450-eicosanoids activate endothelial NO synthase (eNOS) and NO release in Ephx2−/− and WT mouse mesenteric arteries [35, 36]. In contrast, we demon- strated that adenosine A2A receptor-induced relaxation was NO-independent in sEH-null mice, where CYP450-epoxy- genases-derived EETs play an important role in aortic relax- ation [25]. In addition, sEH inhibition reduced the blood pressure in Ang-II-induced hypertension [34, 39] and sEH deletion restricted the reduction of coronary reactive hyper- emic response in Ang-II-infused sEH−/− vs. WT mice (un- published data from our lab).

Therefore, we hypothesized that Ephx2 gene deletion attenuates Ach-induced relaxation which is independent of A2AAR but dependent on NO and CYP-epoxygenases.Role of sEH in Ach‑induced relaxationAch-induced dose-dependent aortic relaxation in Ephx2−/− was not different (P < 0.05) from C57Bl/6 mice except at 10−5 M Ach (P < 0.05, Fig. 3). This is very different from adeno- sine (5′-N-ethylcarboxamidoadenosine, adenosine ago- nist, NECA) and adenosine A2A receptor (A2A AR agonist, CGS 21680)-induced dose-dependent aortic relaxation in sEH−/− compared to C57Bl/6 (P < 0.05) [25], and this A2A AR- induced vascular response was independent of NO and COX [17, 18, 21, 25]. ACh data between Ephx2−/− and C57Bl/6 mice suggest that there is no relationship between ACh and the presence or absence of sEH except at 10−5 M of Ach, which is similar to as we reported earlier from our lab [25].Role of NO in Ach-induced relaxation. L-NAME a nitric oxide synthase inhibitor (100 µM) had almost com- pletely blocked ACh-induced vascular response in both Ephx2−/− and C57Bl/6 mice (P < 0.05, Fig. 4a, b). However, there is a slightly significant difference observed at 10−4 M of L-NAME in Ach-induced dose-dependent vascular response between Ephx2−/− and C57Bl/6 mice (P < 0.05, Fig. 5). This indicates that the deletion of the Ephx2 gene confers resist- ance to the reduction of Ach-induced dose-dependent aortic relaxation caused by L-NAME treatment in Ephx2−/− com- pared to C57Bl/6 mice. Due to deletion of Ephx2 gene, sEH enzyme activity is aborted and the conversion of EETs into dihydroxyeicosatrienoic acids (DHETs) may be stopped, caus- ing epoxides to accumulate and stay for longer periods after they are generated from arachidonic acids [13, 25]. This is a compelling approach for controlling hypertension and vascu- lar diseases [12, 13, 24, 25, 31–34].

Also, EETs have been demonstrated to increase the release of NO in cultured bovine aortic endothelial cells (BEACs), and EETs/DHETs can induce endothelial NO release to modulate vascular tone [36, 68]. Interestingly, our observation shows deletion of Ephx2gene results in a resistance to the reduction of Ach-induced dose-dependent aortic relaxation causing by L-NAME treat- ment in Ephx2−/− compared to C57Bl/6 mice, and may be due to a higher availability of EETs in Ephx2−/− compared to C57Bl/6 mice. Similarly, we have seen in NECA (adenosine) and CGS 21680 (A2AAR agonist)-induced dose-dependent aortic relaxation in Ephx2−/− compared to C57Bl/6 mice [25], as well as enhanced coronary reactive hyperemic response in Ephx2−/− vs. C57Bl/6 mice [13]. In addition to that, NECA and CGS 21680 significantly enhanced aortic relaxation in eNOS−/− vs. C57Bl/6 mice and this relaxation was dependent on A2A AR through CYP-epoxygenases-EETs pathway but not eNOS [21]. Therefore, according to the current study, CYP- epoxygenases-EETs pathway is both A2A AR-dependent but NO-independent and NO-dependent but A2A AR-independent.MS-PPOH (CYP-epoxygenases inhibitor) was able to par- tially block Ach-induced dose-dependent aortic relaxation in both Ephx2−/− and C57Bl/6 mice compared to non- treated mice (Fig. 6a, b). No difference was found between Ephx2−/− and C57Bl/6 mice. MS-PPOH has significantly less blocking action than L-NAME-treated Ephx2−/− and C57Bl/6 mice (Fig. 7). These findings indicate that CYP-epoxyge- nases are also involved in Ach-induced dose-dependent aortic relaxation, as similarly reported in mouse mesenteric arteries [36]. In contrast, we have observed NECA (adenosine) and CGS 21680 (A2AAR agonist)-induced dose-dependent aortic relaxation independent of NO and COX in sEH−/− compared to C57Bl/6 mice [25], as well as enhanced coronary reac- tive hyperemic response in Ephx2−/− vs.

C57Bl/6 mice [13]. However, CYP-epoxygenases are involved in Ach-induced dose-dependent aortic relaxation, as similarly reported in mouse mesenteric arteries [36].Role of EETs in Ach-induced relaxation. 14,15-EEZE (EETs antagonist) was able to block partially Ach-induced dose-dependent aortic relaxation in both Ephx2−/− and C57Bl/6 mice compared to non-treated mice (Fig. 8a, b). There was no difference between 14,15-EEZE-treated Ephx2−/− and 14,15-EEZE-treated C57Bl/6 mice. 14,15-EEZE has signifi- cantly less reduction in blocking action than L-NAME-treated Ephx2−/− and C57Bl/6 mice (Fig. 9). That means EETs are involved in Ach-induced dose-dependent aortic relaxation, as similarly reported in mouse mesenteric arteries [36].SCH-58261 (1 µM, A2AAR-antagonist) was not able to block Ach-induced dose-dependent aortic relaxation in both Ephx2−/− and C57Bl/6 mice (Fig. 10a, b). In addition,there is a significant difference between SCH-58261-treated Ephx2−/− and SCH-58261-treated C57Bl/6 mouse Ach- induced dose-dependent aortic vascular responses (Fig. 11). These data are in agreement with our earlier reports that NECA (adenosine) and CGS 21680 (A2AAR agonist)- induced dose-dependent aortic relaxation are independ- ent of NO and COX in sEH−/− and C57Bl/6 mice [17, 18, 21, 25], as well as enhanced coronary reactive hyperemic response in sEH−/− vs. C57Bl/6 mice [13]. In addition, the difference between the blocking action of L-NAME vs. non- treated Ephx2−/−/C57Bl/6 mice and SCH-58261-treated Ephx2−/−/C57Bl/6 mice are the same (Fig. 11).

Effect of Ang‑II in Ach‑induced relaxation Ang-II (1 µM) was able to partially block Ach-induced dose- dependent aortic relaxation in both Ephx2−/− and C57Bl/6 mice compared to non-treated mice (Fig. 12a, b), and there is a significant difference was found between the blocking action of Ang-II in Ephx2−/− compared to C57Bl/6 mice (Fig. 13a). This shows that the absence of sEH activity due to lack of Ephx2 gene in Ephx2−/− compared to C57Bl/6 mice has an advantage in reducing the blocking action of Ang-II in Ach-induced dose-dependent aortic relaxa- tion. As mentioned earlier, the inhibition or deletion of sEH makes beneficial epoxides accumulate and reside for longer periods after they are generated from arachidonic acids, and sEH inhibition may be a compelling approach for controlling hypertension and vascular diseases [12, 13, 24, 25, 31–34]. Also, Ang-II upregulated sEH [69] and sEH inhibition blocked Ang-II-induced increase in blood pres- sure in rats [70]. Further, Ang-II blocking action on Ach- induced dose-dependent relaxation in both Ephx2−/− and C57Bl/6 mice has significantly lesser reduction in blocking action than L-NAME-treated Ephx2−/− and C57Bl/6 mice (Fig. 13b). Both Ang-II and L-NAME-blocked Ach-induced dose-dependent relaxation in Ephx2−/− and C57Bl/6 mice. However, in the absence of sEH in Ephx2−/− mice, Ang-II and L-NAME-blocking action is significantly less than the blocking action in C57Bl/6 mice. Therefore, we conclude that inhibition or deletion of sEH in mice may reduce the toxic action of both Ang-II and L-NAME in Ach-induced dose-dependent relaxation in Ephx2−/− compared to C57Bl/6 mice.

Conclusion
In conclusion, inhibition or absence of sEH through deletion of the Ephx2 gene appears to be critical for acetylcholine- induced dose-dependent relaxation while challenged by Angiotensin-II or L-NAME. When the sEH enzyme was absent in Ephx2−/− mice, Ach-induced dose-dependent relaxation was similar to C57Bl/6 mice except at 10−5 M of Ach. However, when the Ach-induced dose-dependent relax- ation was challenged by Ang-II or L-NAME, absence of sEH made a significant difference in the resistance to block in Ach-induced dose-dependent relaxation compared to the presence of sEH in C57Bl/6 mice. Therefore, we conclude that inhibition or deletion of sEH in mice may reduce the toxic action of both Ang-II and L-NAME in Ach-induced dose-dependent relaxation in Ephx2−/− compared to C57Bl/6 mice.

Summary
Deletion of sEH in Ephx2−/− results in mEH (Ephx1) upreg- ulation. Therefore, the absence of sEH in the Ephx2−/− mice results in mEH having a small role in hydrolyzing EETs into DHETs in the current study as mentioned by others [51, 52, 54]. L-NAME attenuates most of the Ach-induced dose- dependent relaxation in both Ephx2−/− and C57Bl/6 mice. Thus, it is clear that Ach-induced relaxation is almost solely dependent on NO. However, attenuation of Ach-induced dose-dependent relaxation due to L-NAME in Ephx2−/− is less than C57Bl/6 mice. CYP-epoxygenase inhibitor (MS- PPOH) was able to partially attenuates Ach-induced dose- dependent relaxation in both Ephx2−/− and C57Bl/6 mice. Thus, it demonstrates that CYP-epoxygenases are partially involved in Ach-induced dose-dependent relaxation in both Ephx2−/− and C57Bl/6 mice. EETs antagonist (14,15- EEZE) was also able to partially attenuate Ach-induced dose-dependent relaxation in both Ephx2−/− and C57Bl/6 mice. Therefore, EETs seem to be involved in Ach-induced dose-dependent relaxation in both Ephx2−/− and C57Bl/6 mice. However, A2A AR-antagonist (SCH 58261) did not alter Ach-induced dose-dependent relaxation in either Ephx2−/− and C57Bl/6 mice. Hence, it appears that Ach- induced dose-dependent relaxation is independent of A2A AR but dependent on NO. Ang-II also attenuates less Ach- induced dose-dependent relaxation in Ephx2−/− compared to C57Bl/6 mice. According to our data, sEH plays an impor- tant role in the in Ach-induced dose-dependent regulation of vascular response. Therefore, we conclude that SCH58261 inhibition or deletion of sEH in mice may reduce the toxic action of both Ang-II and L-NAME in Ach-induced dose-dependent relaxation in Ephx2−/− compared to C57Bl/6 mice.