Endothelium-dependent relaxations in the aorta from K_2p6.1 knockout mice

Abstract

K_2P6.1 or TWIK-2, a two-pore domain K channel, is an important regulator of cardiovascular function. K_2P6.1 is highly expressed in vascular smooth muscle and endothelium. Mice (8–12 wk) lacking functional K_2P6.1 (K_2P6.1^−/−) are hypertensive and have enhanced vascular contractility. It is not known whether the lack of functional K_2P6.1 in endothelium has a role in the vascular dysfunction in K_2P6.1^−/− mice. We tested the hypothesis: K_2P6.1^−/− mice have impaired endothelium-dependent relaxations. K_2P6.1^−/− mice were ∼35 mmHg more hypertensive than WT mice at both 8–12 wk (young adult) and 20–24 wk (mature mice, P < 0.01; n = 8–10). Endothelium-dependent relaxations of the thoracic aorta were evaluated by isometric myography after contraction with phenylephrine (10⁻⁶ M). Maximal ACh-dependent relaxations were increased from 65 ± 1% to 73 ± 1% in the aorta from young adult (P < 0.01; n = 6) and from 45 ± 1% to 74 ± 1% in the aorta from mature (P < 0.001; n = 5) K_2P6.1^−/− mice compared with K_2P6.1^+/+ littermates. However, in the aorta from young adult and mature K_2P6.1^+/+ mice, 10⁻⁵ M indomethacin, a cyclooxygenase inhibitor, increased maximal ACh relaxations to knockout levels. Enhanced relaxation was also seen with ATP, a P_2Y purinergic agonist, and A23187, a nonreceptor-based agonist in mature K_2P6.1^−/− mice. Mature adult aorta from K_2P6.1^−/− showed an attenuated ACh-mediated contraction in the presence of nitro-l-arginine methyl ester (l-NAME) and without precontraction of 0.97 mN vs. 7.5 mN in K_2P6.1^−/− and K_2P6.1^+/+ (P < 0.001; n = 5). In summary, K_2P6.1^−/− mice, which are hypertensive, have enhanced endothelium-dependent relaxations in the aorta due to the suppression of an indomethacin-sensitive constrictor component.

two-pore domain potassium channels (K_2P) are a recently discovered family of K channels that are ubiquitously expressed. Of the 15 gene members in the K_2P family (gene name KCNK), more than half of the members are expressed in the vasculature, with K_2P2.1 (TREK-1), K_2P3.1 (TASK-1), and K_2P6.1 (TWIK-2) being highly expressed (18). The function of these K_2P members in vascular smooth muscle is not fully understood, although recent studies have attempted to clarify their role in regulating vascular tone (4, 16, 23, 26, 36). We recently showed that gene deletion of K_2P6.1 resulted in vascular smooth muscle cell depolarization, altered aortic function, increased peripheral vascular resistance, and hypertension in 8–12-wk-old male mice (23). These initial studies in the K_2P6.1-deficient mice (K_2P6.1^−/−) focused on vascular smooth muscle. However, K_2P6.1 is expressed not only in the vascular smooth muscle but also in the endothelium (23). It is not known whether the lack of K_2P6.1 in the endothelium is involved with altered aortic function in K_2P6.1^−/− mice. This issue is especially important since attenuations in endothelium-dependent relaxations [primarily through nitric oxide (NO)] are closely associated with hypertension (12, 13, 37).

The purpose of this study was to determine whether endothelium-dependent relaxations in the aorta contribute to vascular dysfunction in young adult (8–12 wk) and mature (20–24 wk) K_2P6.1^−/− mice. If the endothelium is involved with the hypertension, we predict that endothelium-dependent relaxations (primarily NO) would be attenuated.

MATERIALS AND METHODS

All studies were approved by the Institutional Animal Care and Use Committee at the Baylor College of Medicine. K_2P6.1^−/− animals were created and maintained on a C57BL6-SV129 hybrid background, as previously described (23). Male F1 hybrid littermate controls were used for this study. Studies were conducted on young adult (8–12 wk) and mature (20–24 wk) mice.

Blood pressure was measured using either tail cuff plethysmography or telemetry after implantation of a catheter in the carotid artery. For tail cuff plethysmography, mice were trained to enter a restraining tube on five successive days. On the fifth day, several plethysmograph readings were obtained and averaged.

For other mice, blood pressure was measured using a telemeter system (PA-C10; Data Science International). Two hours prior to the surgery, the mice received ketoprofen (10 mg/kg sc) as an analgesic. The mice were anesthetized with 2% isoflurane in O₂, and a blood pressure catheter was inserted into the right carotid artery. The telemeter device was placed subcutaneously in the abdomen. Once the surgery was complete, the mice were allowed to recover for 7 days. On day 8, blood pressure was continuously recorded for a 24-h period.

Changes in force generated by aortic segments were measured using wire myography, as previously described (23). Segments were allowed to equilibrate in Krebs buffer (gassed with 5% CO₂/21% O₂/balance N₂) for 1 h at 37°C, and then adjusted to the resting tension of 1.2 mN. Each segment was contracted three times with 40 mM hypertonic KCl followed by one 60 mM hypertonic KCl contraction prior to each experiment. Some segments were precontracted with 10⁻⁶ M phenylephrine (EC₈₀) prior to relaxation. Aortic contractions to PE were similar between K_2P6.1^+/+ and K_2P6.1^−/− mice, as shown in previous studies (23). Contraction values for phenylephrine were 9.9 ± 1.4 and 8.4 ± 1.6 mN for WT and KO (P = 0.48; n = 5) in young adult and 13.3 ± 2.5 and 11.3 ± 0.7 mN for WT and KO (P = 0.46; n = 4) in mature K_2P6.1 mice, respectively. Furthermore, the effects of PE with the addition of the blockers l-NAME, indomethacin, and tiron were no different between genotype. The EC₅₀ and the maximum effect were calculated using Sigma Plot 10 with the following equation: y = min + [(max − min)/1 + (x/EC₅₀)^−Hillslope].

For histological analysis, aortic segments were fixed in 10% neutral buffered formalin, paraffin-embedded, sectioned at 5 μm, and stained with the Movat method, as previously described (23). A Zeiss Axioplan II upright microscope was used to capture bright-field images.

Values are expressed as means ± SE and the two-tailed unpaired t-test. The standard error of the least square means (SELSM) replaced the SE when repeated-measures (RM) two-way ANOVA with the post hoc Holm-Sidak test for multiple comparisons was used. The level of significance was set at P < 0.05.

RESULTS

Telemetry measurements of mean arterial blood pressure (MAP), systolic blood pressure (SBP), and heart rate (HR) in young adult (8–12 wk) wild-type and K_2P6.1^−/− mice over a 24-h period are shown in Fig. 1. MAP and SBP were significantly increased (P < 0.05 RM ANOVA) at all times over 24-h in K_2P6.1^−/− mice (Fig. 1, A and B). HR between the K_2P6.1^+/+ and K_2P6.1^−/− mice was similar (Fig. 1C). Diastolic blood pressure was also significantly increased in K_2P6.1^−/− compared with K_2P6.1^+/+, while pulse pressure and cage activity were similar in the two genotypes (data not shown).

Fig. 1.

Twenty-four-hour blood pressure telemetry in nonanesthetized young adult (8–12 wk) K_2P6.1 mice. Mean arterial pressure in mmHg (A), systolic blood pressure in mmHg (B), and heart rate in beats per minute (C) over 24 h (^#P < 0.05 repeated-measures ANOVA. *P < 0.05 **P < 0.01 using the Holm-Sidak test for multiple comparisons, n = 6). The white bar indicates the inactive phase, and the dark bar represents the active phase of the animal in the 24-h clock. Values are expressed as means ± SE of the least square means (LSM).

Next, we determined whether the hypertension progressed as the animals aged from young adult to maturity. We chose tail cuff plethysmography to measure consecutive SBP and HR over 12 wk in individual mice. Figure 2 shows significantly increased SBP in the young adult (+29 mmHg) and mature (+40 mmHg) K_2P6.1^−/− mice compared with age-matched K_2P6.1^+/+ mice (Fig. 2, A and B, respectively) (P < 0.01; n = 10). This indicates sustained systolic hypertension from 8 to 24 wk in the K_2P6.1^−/− mice. Interestingly, there was no significant change in SBP between the young adult and mature K_2P6.1^−/− mice, indicating lack of overt hypertension progression with maturation.

Fig. 2.

Consecutive tail cuff plethysmography on K_2P6.1 mice from young adult (8–12 wk) to maturity (20–24 wk). Systolic blood pressures (A) and heart rates (B) are shown (**P < 0.01 unpaired t-test, n = 8–10). Values are expressed as means ± SE.

Since background strain affects blood pressure in various animal models (17, 19, 31), we tested blood pressure by tail cuff plethysmography in mature (20–24 wk) mice that were backcrossed on a C57 background for 10 generations. K_2P6.1 deficiency on a C57 background showed increased SBP (+29 mmHg) with no change in heart rate (P < 0.001; n = 7, data not shown). This result indicates that the blood pressure effects of K_2P6.1 deficiency were not dependent on the mixed background strain of the mice.

We then tested endothelium-dependent relaxations in aortic segments from K_2P6.1^+/+ and K_2P6.1^−/− mice using wire myography. Young adult K_2P6.1^−/− mice showed an increased maximal endothelium-dependent relaxation from 47 ± 1% to 77 ± 1% (P < 0.001; n = 8) to carbachol, an ACh analog (Fig. 3A), with a 21% leftward shift in the concentration response curve EC₅₀, from 2.3 ± 0.1 μM in WT to 1.8 ± 0.1 μM in KO. Relaxations to carbachol were abolished in the presence of the nitric oxide synthase (NOS) inhibitor nitro-l-arginine methyl ester (l-NAME). Because enhanced relaxations can also result from increased sensitivity of the smooth muscle to NO itself (14), we performed a dose-response experiment to the NO donor, MAHMA NONOate, with and without l-NAME pretreatment. The aortic segments showed no difference in relaxation to MAHMA NONOate between genotypes, indicating that the enhanced endothelium-dependent relaxations in K_2P6.1^−/− mice were not due to increased sensitivity of the smooth muscle to NO (Fig. 3B).

Fig. 3.

Vascular relaxation of young adult (8–12 wk) K_2P6.1 thoracic aortas. Vehicle or nitro-l-arginine methyl ester (l-NAME) (10⁻⁴ M) was given to aortic segments 60 min prior to the start of the experiment. Aortas were precontracted with phenylephrine (10⁻⁶ M) prior to relaxation (^##P < 0.01 RM ANOVA; *P < 0.05 Holm-Sidak test for multiple comparisons; n = 8). Values are expressed as means ± SELSM.

We then characterized the enhanced endothelium-dependent relaxations of young adult aortic segments using ACh. Maximal relaxation to ACh was increased from 65 ± 1% to 73 ± 1% in the aorta from young adult (P < 0.01; n = 6) K_2P6.1^−/− mice compared with K_2P6.1^+/+ littermates (Fig. 4A) with a 73% leftward shift in the concentration response curve EC₅₀ from 83 ± 6 nM in WT to 22 ± 2 nM in KO. Preincubation of young adult aortic segments with 10⁻⁵ M indomethacin, a cyclooxygenase inhibitor, did not affect the relaxation to ACh in the K_2P6.1^−/−; however, it did enhance maximum relaxations in the K_2P6.1^+/+ to K_2P6.1^−/− levels (Fig. 4A). Preincubation of the aortic segments with 10⁻³ M tiron, a superoxide anion scavenger, did not significantly affect the relaxation to ACh in the K_2P6.1^−/− but resulted in a leftward shift in the EC₅₀ by 90% in the K_2P6.1^+/+ (P < 0.05; n = 7) (Fig. 4B). Combined tiron and indomethacin did not produce a greater effect in either genotype than indomethacin alone (data not shown). The relaxations to ACh in the aorta from either K_2P6.1^−/− or K_2P6.1^+/+ mice were not affected by catalase, an enzyme that rapidly degrades H₂O₂ (data not shown), indicating that H₂O₂ is not involved. This reinforces the proposal that a robust, endothelium-dependent contractile mechanism that is present in wild-type mice is largely suppressed in aortas from K_2P6.1^−/− mice.

Fig. 4.

Characteristics of ACh-dependent relaxation in young adult (8–12 wk) K_2P6.1 mice. 10⁻⁵ M indomethacin (indo) (A) and 10⁻³ M tiron (B) were preincubated with aortic segments for 60 min prior to contraction with phenylephrine (10⁻⁶ M) and subsequent relaxation. The same vehicle data (dotted line) are shown in both A and B (#P < 0.05 repeated measures ANOVA. $P < 0.05 repeated-measures ANOVA interaction. *P < 0.05 Holm-Sidak test for multiple comparisons; n = 7 or 8). Values are expressed as means ± SELSM.

Because the production of prostanoid-constricting factors and superoxide anion levels increase with age (13), we tested endothelium-dependent relaxations in the aorta of mature (20–24 wk) K_2P6.1 mice. Maximum ACh-dependent relaxations were increased from 45 ± 1% to 74 ± 1% in K_2P6.1^−/− mice compared with K_2P6.1^+/+ littermates with a leftward shift in the EC₅₀ by 52% (Fig. 5A, P < 0.001; n = 5). Preincubation with indomethacin enhanced the ACh-dependent relaxation in K_2P6.1^+/+ mice, making it similar to that in the K_2P6.1^−/− mice (Fig. 5A). Analysis of the change in relaxation with age revealed that the K_2P6.1^+/+ showed a significant (P < 0.05; n = 5) decrease from 63 ± 1% to 45 ± 1% in ACh-dependent relaxations from young adult to maturity (Fig. 5B), while the K_2P6.1^−/− mice had no change in ACh-dependent relaxations with age (Fig. 5C). The relaxations in aortic segments from young adult and mature K_2P6.1^−/− mice were similar in the presence or absence of indomethacin. In addition, the ACh-dependent relaxations of K_2P6.1^−/− were similar to K_2P6.1^+/+ with indomethacin.

Fig. 5.

ACh-dependent relaxations of the thoracic aorta from young adult and mature K_2P6.1 mice. Vehicle or 10⁻⁵ M indomethacin (indo) was given to aortic segments 60 min prior to the start of the experiment. Aortas were contracted with phenylephrine (10⁻⁶ M) prior to relaxation. Mature K_2P6.1 vehicle (A) or indo-treated (B) and (C) overlay of vehicle data from Fig. 4A (dotted line) with vehicle data from Fig. 5A (dotted line) (#P < 0.05 and ##P < 0.01 RM ANOVA. *P < 0.05 and **P < 0.01, Holm-Sidak test for multiple comparisons; n = 5). Values are expressed as means ± SELSM.

We tested other agents that relax aortic segments through an endothelium-dependent mechanism. Relaxations to ATP, a P_2Y receptor agonist (Fig. 6A) were increased at 10⁻⁴ M from 27 ± 4% to 54 ± 6% (P < 0.05; n = 4), while relaxations to A23187, a Ca²⁺ ionophore (Fig. 6C), were increased at 10⁻³ M from 12 ± 8% to 57 ± 10% (P < 0.05; n = 4) in aortic segments obtained from 20–24-wk-old K_2P6.1^+/+ and K_2P6.1^−/− mice, respectively. Indomethacin had little to no effect on the relaxations in segments from K_2P6.1^−/− mice but enhanced the relaxations in the K_2P6.1^+/+ to K_2P6.1^−/− levels (Fig. 6, B and D).

Fig. 6.

Vascular relaxation of the thoracic aorta from mature (20–24 wk) K_2P6.1 mice. Vehicle or 10⁻⁵ M indomethacin (indo) was given to aortic segments 60 min prior to the start of the experiment. Aortas were contracted with phenylephrine (10⁻⁶ M) prior to relaxation. ATP (A and B), A23187 (C and D). A and C show vehicle-treated, while B and D show indo-treated (#P < 0.05 repeated-measures ANOVA; *P < 0.05 Holm-Sidak test for multiple comparisons; n = 5). Values are expressed as means ± SELSM.

To further confirm that the aortas of mature adult K_2P6.1^−/− produced less indomethacin-sensitive, endothelium-dependent contraction than K_2P6.1^+/+, we preincubated the aortic segments with 10⁻⁵ M l-NAME and performed a dose-response experiment to ACh in the absence of PE contraction. This method has been used to exclusively study endothelium-dependent contracting factors without the inhibitory effects of NO on the contraction and in the suppression of NOS activity (38). Mature adult aorta from K_2P6.1^−/− showed an attenuated ACh-mediated contraction of 0.97 mN vs. 7.5 mN in K_2P6.1^−/− and K_2P6.1^+/+ (Fig. 7A, P < 0.01 RM ANOVA, P < 0.001 RM ANOVA interaction; n = 5). This indicates that the K_2P6.1^−/− had a largely suppressed endothelium-dependent contraction that was present in the K_2P6.1^+/+. Furthermore, indomethacin abolished the contractions in both genotypes (Fig. 7B).

Fig. 7.

Endothelium-dependent contractions from the aorta of mature (20–24 wk) K_2P6.1 mice. l-NAME (10⁻⁵ M) alone or in combination with indomethacin (indo) (10⁻⁵ M) were given to aortic segments 60 min prior to the start of the experiment (^##P < 0.01 repeated-measures ANOVA. ^$$$P < 0.001 repeated-measures ANOVA interaction. **P < 0.01 and ***P < 0.001 Holm-Sidak test for multiple comparisons; n = 5). Values are expressed as means ± SELSM.

Finally we performed a histological analysis on the aortic segments to assess for differences in morphology using Movat staining. We found no differences in medial thickness and luminal diameter in the mature K_2P6.1^−/− vs. K_2P6.1^+/+ mice (Fig. 8). Furthermore, on visual inspection, there was no obvious myocyte disarray or alterations in elastin or collagen organization.

Fig. 8.

Movat staining of the thoracic aorta from K_2P6.1^+/+ (A and B) and K_2P6.1^−/− (C and D) mice. B and D are magnified to show detail of the vascular wall (n = 4).

DISCUSSION

The purpose of this study was to determine whether K_2P6.1^−/− mice have impaired endothelium-dependent relaxations that would indicate endothelial dysfunction. Because endothelial dysfunction is associated with hypertension and often precedes its onset (13, 37), it was reasonable to hypothesize that endothelial dysfunction (decreased endothelium-dependent relaxations) would occur in K_2P6.1^−/− mice. Surprisingly, not only was there no attenuation of relaxations, but relaxations of the aorta by endothelium-dependent mechanisms were actually enhanced in K_2P6.1^−/− mice (Figs. 3–6). It follows that endothelium dysfunction cannot account for the hypertension that we report in K_2P6.1^−/− mice.

The enhanced relaxations in aortas from young adult K_2P6.1^−/− mice resulted from the suppression of an indomethacin-sensitive/ROS constrictor component, which was present in aortas from K_2P6.1^+/+ mice. In the young adult mice, the addition of the ROS scavenger, tiron, had no effect on relaxations in the K_2P6.1^−/− aorta; however, it substantially enhanced relaxation in the K_2P6.1^+/+ (Fig. 4) at the doses of ACh < 10⁻⁷ M. As the normal mice (i.e., K_2P6.1^+/+) mature (20–24 wk), relaxations of the aorta to ACh decline (Fig. 5) through increased indomethacin-sensitive vasoactive prostanoids and ROS generated from cyclooxygenase reactions (1, 32, 37). Interestingly, endothelium-dependent relaxations were not altered with maturation in the aorta from K_2P6.1^−/− mice. In the aorta from young and mature adult mice, indomethacin was able to enhance relaxation in K_2P6.1^+/+ to that observed in K_2P6.1^−/− mice. Our data would indicate that the difference in relaxations to ACh in K_2P6.1^+/+ and K_2P6.1^−/− mice involves an indomethacin-sensitive/ROS constrictor component (Fig. 7). It is noteworthy that our previous studies (23) in K_2P6.1^−/− aortas showed greater, not less sensitivity, to the thromboxane prostanoid receptor agonist U46619 than K_2P6.1^+/+, which rules out the possibility that the K_2P6.1^−/− aortas are insensitive to at least some vasoactive prostanoids.

Hypertension is strongly associated with endothelial dysfunction in humans and animal models (5, 9, 24, 25, 34). The extent of endothelial dysfunction appears related to the elevated blood pressures, as reversal of hypertension causes normalization of endothelial function, except in the presence of combined pathologies (39, 41). Rescue of endothelial function with blood pressure control has been extensively documented in the human (3, 8, 20, 21, 35, 40), rat (10, 11, 30), and mouse (22, 33). Our results demonstrate that attenuation of endothelium-dependent relaxations is not necessary for hypertension to occur. Furthermore, attenuation of endothelium-dependent relaxations does not necessarily result from hypertension. Thus, the mechanism of hypertension in the K_2P6.1^−/− mice may be altogether independent of the endothelium.

Our results can be compared and contrasted to endothelial function in both the spontaneously hypertensive rat (SHR) and the borderline hypertensive rat (BHR). The SHR and the BHR showed a similar stage of indomethacin-insensitive enhanced endothelium-dependent relaxations to ACh (2, 6, 7, 15, 27, 28) and A23187 (27). Enhanced endothelium-dependent relaxations occurred in young adult (6 wk) rats in the developing stages of hypertension only. By maturity (16–20 wk) in the SHR, when significant hypertension occurred, there was marked indomethacin-sensitive endothelial dysfunction. In K_2P6.1^−/− mice systolic blood pressure did not significantly progress from young adult to maturity (Fig. 2), as with the SHR model. It is possible that the maintenance of enhanced endothelial relaxations in K_2P6.1^−/− mice limited the progression of hypertension. It is also possible that the decreased indomethacin-sensitive/ROS constrictor component allowed the K_2P6.1^−/− mice to, thereby, “escape” the deleterious feedback loop between hypertension and endothelial dysfunction.

Although there are numerous possibilities as to why the aorta from K_2P6.1^−/− mice do not have an indomethacin-sensitive/ROS constrictor component, one point of speculation involves Ca²⁺ regulation in the endothelium. The membrane potential is a determinant of the driving force for extracellular Ca²⁺ to move into the cytoplasm of endothelium. The more hyperpolarized the membrane becomes, the greater the driving force for Ca²⁺ to enter the endothelium. If the absence of K_2P6.1 results in a depolarization of the endothelium [as it does in vascular smooth muscle (23)], then endothelial Ca²⁺ may not reach sufficient concentrations for the liberation of arachidonic acid, the major substrate and limiting factor for the synthesis of constrictor prostanoids through the cyclooxygenase pathway. Ca²⁺ acts to anchor phospholipase A₂ (PLA₂), the enzyme responsible for liberating arachidonic acid, onto phospholipid membranes. Of note, the driving force for Ca²⁺ into the endothelium in the absence of functional K_2P6.1 could result in the inability to achieve sufficient endothelial Ca²⁺ for arachidonic acid liberation by PLA₂. Alternatively, Ca²⁺ increases could be sufficient to activate endothelial NOS, an enzyme that is activated by much lower concentrations of Ca²⁺ than PLA₂ (see discussion in Ref. 42). Because NO is a potent and long-term inhibitor of indomethacin-sensitive prostanoids/ROS constrictor components, this could result in decreased endothelium-dependent and indomethacin-sensitive contraction in the K_2P6.1^−/− (1, 25, 29, 38). Further study of endothelial membrane potential and calcium concentrations, as well as the direct measurements of NO and ROS constrictor activity, may help clarify the vascular phenotype in the K_2P6.1 animals.

Perspectives and Significance

We show that K_2P6.1^−/− mice have enhanced endothelium-dependent relaxations in young adult and mature mice concomitant with significant systolic hypertension. The enhanced relaxations result from decreased synthesis of indomethacin-sensitive/ROS constrictor components, which oppose NO. Studies on humans do not indicate that hypertension is associated with any period of overt adaptive or compensatory increases in endothelium-dependent relaxations. This study is the first to describe a model in which sustained hypertension is linked not to endothelial dysfunction but rather to endothelial “hyper”-function. It may be the endothelium itself is producing a compensatory response to hypertension, which limits its progression.

GRANTS

This study was supported by National Institutes of Health Grants 5R21HL-098921-02, NIH R01-HL088435, and R21-NS077413.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Author contributions: E.E.L. and R.M.B. conception and design of research; E.E.L. and R.F.C. performed experiments; E.E.L., S.P.M., and R.M.B. analyzed data; E.E.L., L.M.P., S.P.M., and R.M.B. interpreted results of experiments; E.E.L. and R.M.B. prepared figures; E.E.L. and R.M.B. drafted manuscript; E.E.L., L.M.P., R.F.C., S.P.M., and R.M.B. edited and revised manuscript; E.E.L., L.M.P., R.F.C., S.P.M., and R.M.B. approved final version of manuscript.