Endothelial dysfunction in renal arcuate arteries of obese Zucker rats: The roles of nitric oxide, endothelium-derived hyperpolarizing factors, and calcium-activated K+ channels

The roles of nitric oxide (NO), endothelium-derived hyperpolarizing factors (EDHF), and calcium-activated K+ (KCa) channels in diabetes-associated endothelial dysfunction of small renal arteries are not clear. The present study investigated acetylcholine (ACh)-induced vasorelaxation of renal arcuate arteries from obese Zucker (OZ) rats at different diabetes durations, and the relative contribution of NO, EDHF, and KCa channels to the endothelial dysfunction. OZ rats of 7 weeks (prediabetic stage), 12 weeks (early diabetic stage), and 20 weeks (late diabetic stage), and time-matched lean control rats, were studied. Segments of arcuate arteries (130 to 180 μm) were isolated, cannulated and pressurized. Vascular endothelial functions were tested using ACh-induced vasodilation. Our experiments demonstrated: (1) ACh-elicited vasodilation was impaired in OZ rats of 20 weeks, but not in rats of 7 and 12 weeks; (2) inhibition of NO or EDHF (contributed by epoxyeicosatrienoic acids [EETs]) production significantly decreased ACh-induced vasodilation in both lean and OZ rats of 20 weeks. The reduction of ACh-induced vasodilation by inhibition of NO or EDHF formation was less in OZ rats, as compared to lean rats; and (3) inhibition of KCa channels markedly reduced ACh-induced vasodilation in lean control rats, but not in OZ rats of 20 weeks. Our observations indicated that endothelium-dependent vasodilation in renal arcuate arteries is impaired in diabetes mellitus; NO and EDHF, mainly EETs, dominate the ACh-induced vasodilation in renal arcuate arteries; the contribution of NO and EETs is impaired in diabetic rats; KCa channels are involved in ACh-induced vasodilation; and the activity of KCa channels is downregulated in diabetes mellitus.


Introduction
Endothelial dysfunction plays a key role in the pathogenesis of diabetic vascular complications, which account for most of the morbidity and mortality [1]. Substantial clinical and experimental a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 diameter from 130 to 180 μm were chosen. Arterial segments were cannulated at both ends onto glass micropipettes and secured, and the lumen of the vessel was filled with MOPS-buffered solution containing 1% albumin. The transmural pressure was set at 60 mmHg and continuously monitored. Neither transluminal flow nor oxygenation was applied to the cannulated vessels. The internal diameter of the vessels was recorded by a computerized diameter tracking system (Diamtrak, Montech Pty Ltd., Australia).

Experimental procedures
After cannulation, the vessel diameter was measured at 60 mm Hg (maximum diameter [Dmax]). Then, the organ bath was heated from room temperature (~26˚C) to 37.5˚C and superfused continuously with a MOPS-buffered physiological saline solution (37.5˚C, PH 7. To observe ACh-induced vasodilation after 30-minute equilibration in the bath, cannulated arcuate arteries were exposed to 1 μmol/l phenylephrine (PE) to produce a 50-60% contraction of maximum diameter; if not, PE was increased to 5μmol/l. When the vessel diameter is at a stable value it is considered to be in the steady PE-preconstriction state. After that, increasing doses of ACh were applied to incubation bath to elicit vasorelaxation. The vessel internal diameter was recorded once the vessel reached a steady state.

Data collection and statistical analysis
All data were presented as mean ± SD. Vessel diameter changes were presented as percentage (%) of dilation of the preconstriction, calculated as follows: % of vasodilation = [(Dagonist-Dbase)/ (Dmax-Dbase)]x100, where Dmax was the diameter of the vessel at room temperature when the vessels were exposed to the pressure of 60 mm Hg, Dbase was the vessel diameter at steady constriction state induced by PE before ACh stimulation, and Dagonist was the diameter of the vessel after ACh stimulation. The maximum dilation was represented as 100%, and baseline diameter was 0%. Comparisons were made with the use of paired Student's t-test or ANOVA with a post hoc Bonferroni test, as appropriate. The acceptable level of significance was defined as P<0.05.

Body weights and blood glucose levels
Body weight and blood glucose data for each age-matched lean and OZ rats are presented in Fig 1. Body weight was increased as the age of rats increased, but body weight of each group of OZ rats was greater than that of the lean rats (484.8±38.62 g compared to 325.56±13.00 g at 20 weeks) ( Fig 1A). As expected, blood glucose levels were significantly increased in OZ rats at 12 weeks (440.6±73.04 mg/dl in OZ rats; 154.6±16.50 mg/dl in lean rats) and 20 weeks (574.2 ±50.63 mg/dl in OZ rats; 168.6±19.74 mg/dl in lean rats) ( Fig 1B).

Vascular responses to ACh in renal arcuate arteries of OZ rats at different diabetic stages
ACh induced a concentration-dependent vasorelaxation in both lean and OZ rats of 7 weeks, 12 weeks, and 20 weeks (Fig 2). Removal of the endothelium by passing 1 ml air bubble through the vessel lumen abolished the vascular dilative responses (data not shown), which supports the assertion that ACh-induced vasorelaxation is endothelium-dependent in rat arcuate arteries. There was no significant difference of ACh-induced vasodilation between lean control and OZ rats at the ages of 7 (Fig 2A) and 12 weeks (Fig 2B). But the dilation was significantly reduced in OZ rats at 20 weeks of age (late diabetic stage) ( Fig 2C). The dilation induced by 10 μmol/l ACh was only 28.42±6.53% in OZ rats compared to 66.37±11.44% in lean rats of 20 weeks. Since SNP is an endothelium-independent vasodilator, we then used SNP to test smooth muscle function. SNP elicited a similar vasodilation in both OZ and lean rats of 20 weeks (Fig 2D). Based on the impaired vasodilation results seen here, all the following experiments were conducted on renal arteries isolated from rats at the age of 20 weeks.
The contribution of NO and EDHF to ACh-elicited vasodilation in renal arcuate arteries of late diabetic OZ rats.
The contribution of NO and COX-derived prostaglandins was assessed by L-NNA and indomethacin, respectively. Incubation of vessels with L-NNA had a significant inhibitory effect on vasodilation induced by ACh, in both lean and OZ rats ( Fig 3A). However, vasodilation was much less reduced by L-NNA in OZ rats than in lean rats. The reduction induced by 10 μmol/l ACh after L-NNA was 13.21±4.38% in OZ rats compared to 41.71±18.06% in lean rats ( Fig  3B). Treatment with indomethacin did not cause an obvious inhibition of ACh-induced vasodilation (data not shown). The contribution of EDHF was assessed by the combined inhibition of NOS and COX. Combined application of L-NNA and indomethacin did not further enhance the inhibitory effect of L-NNA. The reduction elicited by L-NNA plus indomethacin ( Fig 3D) was similar to that induced by L-NNA alone ( Fig 3B). However, after combined incubation with L-NNA and indomethacin, ACh-induced vasodilation in OZ rats was less than that in lean control rats. The vasodilation induced by 10 μmol/l ACh was 17.40±7.21% in OZ rats compared to 28.38±12.55% in lean rats after application of L-NNA plus indomethacin ( Fig 3C).

The contribution of EETs/EDHFs to ACh-induced vasodilation in renal arcuate arteries of late diabetic OZ rats
Since CYP epoxygenase is highly expressed in renal microvessels, and CYP epoxygenase metabolites, EETs, are identified as EDHFs in regulating renal vascular function, PPOH and 17-ODYA were used to inhibit epoxygenase activity. 10 μmol/L PPOH significantly decreased the ACh-induced vasodilation in renal arteries in both OZ and lean rats (Fig 4 A). The reduction evoked by PPOH in OZ rats was significantly less than that in lean rats. The reduction induced by 10 μmol/l ACh after PPOH in OZ rats was 12.16±8.05% compared to 27.88 ±12.33% in lean rats (Fig 4B). Incubation of 10 μmol/l 17-ODYA had a similar inhibitory effect (Fig 4C and 4D). The reduced vasodilation by 10 μmol/l ACh after 17-ODYA was 9.34±6.77% in OZ rats compared to 29.19±15.69% in lean rats (Fig 4D).
K Ca channel responses to ACh in renal arcuate arteries of late diabetic OZ rats EDHFs, in particular EETs, hyperpolarize smooth muscle through activation of K Ca channels. We used charybdotoxin and apamin to block K Ca channels. Incubation of vessels with charybdotoxin and apamin had an inhibitory effect on the vasodilation induced by ACh in lean rats (Fig 5A), but not in OZ rats (Fig 5B). A combined incubation with L-NNA, indomethacin, charybdotoxin and apamin almost abolished the ACh-induced vasodilation in both lean ( Fig  5A) and OZ rats (Fig 5B). Blockade of K ATP or K ir channels by application of 30 μmol/l glibenclamide or 30 μmol/l barium chloride did not have any effect on the vasodilation induced by ACh (data not shown). To further test the activity of K Ca channels, NS 1619 was used to activate them. NS 1619 elicited a dose-dependent vasodilation. 100 μmol/l NS 1619 almost evoked a full vasodilation in both lean and OZ rats. But the vasodilation induced by NS 1619 at low concentrations (less than 50 μmol/l) was significantly reduced in OZ rats than in lean rats (Fig 5C).

Discussion
The present study investigated the influence of diabetes on the different stages and components of endothelium-dependent vasodilation in the renal arcuate arteries of OZ rats in vitro. The major findings are: (1) ACh-elicited vasodilation was impaired in late diabetic OZ rats at 20 weeks, but not in 7 week prediabetic or 12 week early diabetic rats; (2) Both NO and EDHF (contributed by EETs) mediated the ACh-elicited vasodilation in renal arcuate arteries of late diabetic OZ rats, and their contributions were reduced; and (3) K Ca channels were involved in the ACh-elicited vasodilation and their function was impaired in late diabetic OZ rats.
Diabetes mellitus is a group of chronic metabolic disorders characterized not only by hyperglycemia, protein and lipid metabolism disorder, but also by macro-and micro-angiopathic vascular complications resulting from endothelial dysfunction, inflammation, apoptosis, and so on [1,26]. Endothelial dysfunction is partially described as the impairment of endotheliumdependent relaxation. Few studies have examined the temporal nature of the onset of impaired endothelial-dependent dilation in renal vessels from animal models of type 2 diabetes. Gealekman et al. [17] found ACh-elicited vasodilation of renal interlobar arteries was significantly attenuated in 22-week-old OZ rats, but unimpaired in 8-week-old OZ rats. In our study, the ACh-induced vasodilation in renal arcuate arteries was investigated in OZ rats at different diabetic stages. ACh elicited a concentration-dependent vasodilation in pressurized renal arcuate arteries in vitro. Compared to the age-matched lean control rats, no difference was demonstrated in the prediabetic (7 weeks) and early diabetic (12 weeks) stages. But at the age of 20 weeks, ACh-induced vasodilation was significantly reduced in OZ rats. Our current observation, in agreement with the above report, revealed the importance of disease duration or the severity of diabetes in the development of vascular endothelial dysfunction. Reduced eNOS activity and NO production, or decreased NO bioavailability, are some of the most important mechanisms of endothelial dysfunction in diabetes mellitus. But there are different observations in renal vasculature. Enhanced [19], normal [27], and impaired [18] NO production and/or activity have been demonstrated in rodent models of type 2 diabetes. In the current study, NO appeared to dominate the vasodilatory responses to ACh in renal resistance arteries in both OZ diabetic and lean rats at 20 weeks of age. Inhibition of NOS markedly decreased ACh-induced vasodilation in these animals. The reduction was less in OZ rats than in lean rats, confirming impaired NO production or activity in the renal vessels of diabetic rats.
EETs have been identified as EDHFs in the kidney [13,15,21]. However, few studies have investigated the role of EDHF in renal arteries of diabetes mellitus, especially in those of type 2 diabetes. Zhao et al. [15] reported that endothelium-dependent EDHF-evoked dilation was impaired in the renal preglomerular vasculature of 20-week-old OZ rats. In our preparation, a portion of endothelium-dependent vasodilation induced by ACh was still observed in both lean rats and OZ rats after the combined inhibition of NOS and COX, but vasodilation was decreased in OZ rats after blocking NOS and COX. Consistent with the previous study, our results indicated that EDHF played an important role in mediating ACh-induced vasodilation in renal arcuate arteries, and the contribution of EDHF was impaired under diabetic conditions.
To further investigate the EDHF identity, PPOH, a specific CYP epoxygenase inhibitor, was tested in our preparation. Treatment with PPOH reduced ACh-elicited vasodilation in both OZ rats and lean control rats at 20 weeks of age. The reduction by PPOH was less in OZ diabetic rats. The results were confirmed by use of another irreversible CYP epoxygenase inhibitor, 17-ODYA. Our observations indicated that CYP epoxygenase metabolites, namely EETs, contributed to ACh-induced vasodilation in renal arteries, and this contribution was reduced in diabetic rats.
The opening of K Ca channels on vascular smooth muscle cells, and the resulting hyperpolarization, is the major feature of EDHF-mediated vasorelaxation. Experiments using electrophysiological and pharmacological approaches have revealed functional K Ca channels in renal microvessels [28,29]. Blockage of K Ca channels diminished the vasodilatory response to 11, 12-EET in renal microvessels [21]. However, the role of K Ca channels in renal arteries in diabetes mellitus has not been thoroughly studied. It is known that diabetic endothelial status is associated with increased production of reactive oxygen species (ROS) [30]. Also, ROS, especially hydrogen peroxide (H 2 O 2 ) and peroxynitrite (ONOO -), effectively inhibited K Ca channel function [31,32]. In the present study, inhibition of K Ca channels with charybdotoxin and apamin decreased ACh-induced vasodilation in lean rats, but not in OZ diabetic rats at 20 weeks of age, which indicated that K Ca channel activity was impaired in diabetic animals. This conclusion was further confirmed by the application of NS 1619, a specific K Ca channel agonist. NS 1619 elicited-vasodilation was reduced in diabetic OZ rats compared to controls. Our current observation, in agreement with the findings reported in retinal vascular smooth muscle [33], supported the idea that the activity of K Ca channels was downregulated in diabetes mellitus. In our preparation, inhibition of K ATP and K ir channels by glibenclamide and low concentrations of barium chloride did not elicit marked alteration of ACh-induced vasodilation, suggesting that K ATP and K ir channels play little or no role in ACh-induced vasodilation in renal arcuate arteries.
In conclusion, the present study provided evidence supporting the hypothesis that NO, EDHF (mainly EETs), and K Ca channels dominate ACh-induced vasodilation in renal arcuate arteries; furthermore, the contribution of NO and EDHF was impaired, and the activity of K Ca channels was downregulated, in renal arcuate arteries of late diabetic OZ rats (Fig 6). This study provides a specific basis for evaluating the appropriate timing of pharmacological intervention to prevent endothelial dysfunction in renal microvessels under diabetic conditions. It also provides mechanistic basis for developing novel drugs to ameliorate endothelial dysfunction in renal small arteries in diabetes mellitus.

Author Contributions
Conceptualization: Dandan Yin, Ying Li.  NO is derived from endothelial NO synthase (eNOS), while prostacyclin (PGI 2 ) and epoxyeicosatrienoic acids (EETs, as EDHFs) are generated from arachidonic acids by cyclooxygenase (COX), and cytochrome P-450 (CYP) epoxygenase, respectively. EETs facilitate the hyperpolarization of smooth muscle cells and vascular relaxation through K + efflux mediated by the opening of K Ca channels either in endothelial cells or in smooth muscle cells. The K + released from K Ca channels of endothelial cells into the subendothelial space (potential connective tissue space beneath the endothelium) subsequently actives inward rectifying K + (K ir ) channels on smooth muscle cells, while producing K + efflux as well. In the present study, the contribution of NO and EDHF to endothelium vasodilation was impaired and the activity of K Ca channels was downregulated in renal arcuate arteries of obese Zucker (OZ) rats at 20 weeks of age, but PGI 2 had no effect on ACh-elicited vasodilation. Barium chloride, inhibitor of the K ir channel; Charybdotoxin (Cha)+Apamin, inhibitor of the K Ca channel; Glibenclamide, inhibitor of the K ATP channel; Indomethacin, inhibitor of COX; L-NNA, inhibitor of eNOS; NS1619, agonist of the K Ca channel; and PPOH or 17-ODYA, inhibitors of CYP epoxygenase. https://doi.org/10.1371/journal.pone.0183124.g006 The roles of NO, EDHF and K Ca channels in endothelial dysfunction of renal arcuate arteries