Endothelial dysfunction and increased arterial stiffness contribute to multiple vascular diseases and are hallmarks of cardiovascular aging. To investigate the effects of aging on shear stress-induced endothelial nitric oxide (NO) signaling and aortic stiffness, we studied young (3–4 mo) and old (22–24 mo) rats in vivo and in vitro. Old rat aorta demonstrated impaired vasorelaxation to acetylcholine and sphingosine 1-phosphate, while responses to sodium nitroprusside were similar to those in young aorta. In a customized flow chamber, aortic sections preincubated with the NO-sensitive dye, 4-amino-5-methylamino-2′,7′-difluorofluorescein diacetate, were subjected to steady-state flow with shear stress increase from 0.4 to 6.4 dyn/cm2. In young aorta, this shear step amplified 4-amino-5-methylamino-2′,7′-difluorofluorescein fluorescence rate by 70.6 ± 13.9%, while the old aorta response was significantly attenuated (23.6 ± 11.3%, P < 0.05). Endothelial NO synthase (eNOS) inhibition, by NG-monomethyl-l-arginine, abolished any fluorescence rate increase. Furthermore, impaired NO production was associated with a significant reduction of the phosphorylated-Akt-to-total-Akt ratio in aged aorta (P < 0.05). Correspondingly, the phosphorylated-to-total-eNOS ratio in aged aortic endothelium was markedly lower than in young endothelium (P < 0.001). Lastly, pulse wave velocity, an in vivo measure of vascular stiffness, in old rats (5.99 ± 0.191 m/s) and in Nω-nitro-l-arginine methyl ester-treated rats (4.96 ± 0.118 m/s) was significantly greater than that in young rats (3.64 ± 0.068 m/s, P < 0.001). Similarly, eNOS-knockout mice demonstrated higher pulse wave velocity than wild-type mice (P < 0.001). Thus impaired Akt-dependent NO synthase activation is a potential mechanism for decreased NO bioavailability and endothelial dysfunction, which likely contributes to age-associated vascular stiffness.
- serine/threonine kinase Akt
- endothelial dysfunction
- nitric oxide synthase
vascular stiffness is emerging as an independent predictor of adverse outcomes in cardiovascular disease (47).1 It is now widely recognized that the endothelium modulates vascular tone acutely (32, 38) and chronically affects vascular remodeling (25, 35) by transducing hemodynamic conditions to the underlying smooth muscle (1, 5). Similarly, the endothelium can regulate the vascular matrix and smooth muscle in response to chemical stimuli. Even though the complicated signaling of endothelial mechanotransduction is still being explored, the “input” and “output” of the process are well documented. The principal input is linear and pulsatile shear stress acting on the endothelial cells (1, 5, 11, 37). The output is a number of vasoactive signaling molecules, the most important of which is the gaseous transmitter nitric oxide (NO), predominantly produced by endothelial NO synthase (eNOS) (1, 17, 44). Endothelial dysfunction, as a result of impaired NO signaling and bioavailability, contributes to the vascular diseases associated with atherosclerosis (25, 27), hypertension (6, 38), and aging (41, 42, 46). This has led to many studies investigating the role of the endothelium in physiological and pathophysiological mechanotransduction, vasorelaxation, and NO signaling.
Paradoxically, previous studies have shown that endothelial dysfunction and reduced NO bioavailability may be associated with an increase in the expression of eNOS, implying that eNOS activity is dependent on other factors (7, 27, 30, 41). Some proposed mechanisms include the following: 1) substrate competitive enzymes (arginase) (2, 46, 50); 2) reactive oxygen species (ROS) (33, 37); 3) endogenous inhibitors of NO synthase (NOS) (asymmetric dimethyl arginine) (21, 39); 4) “uncoupling” of NOS (23, 30); and 5) modulation of upstream activators (kinases) that regulate NOS activity (13, 40, 43). With regard to these kinase modulators, it is now understood that NO synthesis is often associated with amplified eNOS phosphorylation (9, 32). While chemical agonists can cause NOS activation by Ca2+-independent or transient Ca2+-dependent mechanisms (40, 43), mechanical stimuli, such as fluid shear stress and vessel stretch, produce sustained Ca2+-independent increases in NO production, which are exclusively attributed to eNOS phosphorylation (11, 13). Protein kinase B (Akt) is a critical protein modulating cell survival but is also a well-documented activator of eNOS. Both mechanical and chemical stimuli can lead to the activation/phosphorylation of Akt, which, in turn, will phosphorylate eNOS at Ser1177 (9, 11, 32). Although the shear stress-induced phosphorylation cascade has been well elucidated, the stretch-related pathway is just emerging as being unique (24, 34). As aging results in increased vascular stiffness and impaired endothelial NO signaling, we hypothesize that Akt-dependent eNOS phosphorylation is attenuated in aged vessels, leading to the inhibition of eNOS activity and contributing to arterial stiffening.
In this study, we used established and innovative methods to investigate age-related endothelial dysfunction in intact aorta and whole animal models. In isolated aortic tissue, we demonstrated the endothelium-dependent vasorelaxation response to be significantly diminished in aged vessels. We next developed a flow chamber system that allows for real-time fluorescent measurement of NO production rates in the endothelium of intact vessels. With this method, endothelial cell phenotype and paracrine signaling, specifically with regard to shear stress mechanotransduction, are preserved in an in situ environment. We revealed shear stress-derived NO production rates to be significantly impaired in old rat aorta. As a potential molecular regulator of endothelial function, we demonstrated a substantial decrease of the phosphorylated (p)-to-total (t)-Akt ratio in old rat aortic endothelial cells, compared with young. We also observed a proportional reduction in the p/t-eNOS ratio. Finally, we related tissue-level endothelial dysfunction to system-level vascular stiffening through the demonstration of significantly increased pulse wave velocity (PWV) in old rats (16, 18, 45). These complementary techniques yield a comprehensive presentation of physiological and pathophysiological endothelial function and its role in altering vascular stiffness with aging.
MATERIALS AND METHODS
We used young (Y, 3–4 mo) and old (O, 22–24 mo) male Wistar rats obtained from Harlan Laboratories and male eNOS-knockout (KO) mice, with their wild-type background controls (C57BL/6J, 10–12 mo) from Jackson Laboratories. The rats were housed in controlled temperature and light conditions with feeding and watering ad libitum. A group of Y rats designated for PWV study received Nω-nitro-l-arginine methyl ester (l-NAME; 0.7 mg/ml) (Fluka), a nonspecific NOS inhibitor, in their drinking water for 6 days. A treatment of such duration allows for observable PWV changes with limited alterations of vascular wall composition (12, 20). In this group, PWV was measured both before the start and at the end of l-NAME administration. All experimental procedures were approved by the Institutional Animal Care and Use Committee of The Johns Hopkins University School of Medicine.
The animals were heparinized 1 h before death. The rats were euthanized, and a maximum length of aorta, from the distal aortic arch to the femoral bifurcation, was isolated and removed. The dissected vessel was immediately placed in a culture dish with iced Krebs-HEPES buffer [composed of (in mM) 110 NaCl, 4.7 KCl, 25.0 NaHCO3, 1.2 MgSO4, 1.03 KH2PO4, 11.1 d-(+)-glucose, 20.0 HEPES, and 1.87 CaCl2, with a pH of 7.4 at 24°C]. The aorta was then cleaned of additional connective tissue and sectioned into rings. Aortic rings were designated for Western blot analysis, ring/tension experiments, or flow chamber studies. In the vascular tension tests, the endothelium was kept intact (E+) or was physically removed (E−). To accommodate the flow chamber, the rings were cut longitudinally into strips and pinned, endothelial side up, in a Silastic-coated petri dish. The dish was then filled with Krebs-HEPES buffer containing the NO-sensitive fluorescent dye 4-amino-5-methylamino-2′,7′-difluorofluorescein diacetate (DAF-FM DA) (Molecular Probes, Eugene, OR). The vessel strips were incubated with 5 μM DAF-FM DA for 0.5 h at room temperature and in minimal light. These conditions were optimized to achieve endothelium-specific fluorescence in rat aortic tissue (see online supplement). When required, the NOS inhibitor NG-monomethyl-l-arginine (l-NMMA) (Alexis Biochemicals) was added concurrently at 300 μM.
Vascular tension studies.
For analysis of vasocontractility, the aortic rings were placed in 25-ml organ chambers and suspended from a strain gauge, as described previously (4). Briefly, the chambers were filled with oxygenated (95% O2-5% CO2), superfused Krebs buffer [consisting of (in mM) 118 NaCl, 4.7 KCl, 25.0 NaHCO3, 1.16 MgSO4, 1.18 KH2PO4, 11.1 d-(+)-glucose, 3.24 CaCl2], maintained at pH 7.4 and 37°C. The vessels were incrementally stretched to 3 g for optimized contractility and were preconstricted with the previously determined EC50 to phenylephrine (Sigma, St. Louis, MO). Cumulative dose responses to ACh (Sigma) or sodium nitroprusside (SNP) (10−9.5-10−5 M) were obtained to characterize endothelium-dependent and -independent vasorelaxation, respectively. In addition, vasorelaxation responses to sphingosine 1-phosphate (S1P; 10−6-10−5 M; Sigma), a sphingolipid component of HDL known to mediate phosphatidylinositol 3-kinase (PI3K)/Akt-dependent eNOS activation, were acquired. Vasorelaxation is expressed as percent relaxation, as calculated by the percent decrease in tension from the phenylephrine-induced preconstriction.
The fluorescence measurements were accomplished with a variable monochromator-based spectrofluorometer (Photon Technology International, Lawrenceville, NJ) and an inverted fluorescent microscope (Nikon TE2000). The excitation source is a variable monochromator, while a fixed emission filter selects the emission wavelength. The fluorescence data were captured and analyzed with Felix32 (Photon Technologies International). The DAF-FM fluorescence excitation wavelength was set to 485 nm, and the emissions were collected at 510 nm.
Flow chamber and shear stress modeling.
The velocity field in the chamber was modeled through computational finite element analysis using the software program FEMLAB (MathWorks). The velocity field was approximated with the three-dimensional, incompressible, and static Navier-Stokes (Eq. 1), where ρ is fluid density, t is time, u is the velocity field, p is pressure, η is the dynamic viscosity, and F is the volume force. Once the velocity field was obtained, the velocity derivative, δu/δz, was calculated for the midpoint of the channel at the vessel surface. To calculate the shear stress, the velocity gradient value can be used with the viscous shear stress (Eq. 2), where τxz is the shear stress in the x-direction acting on the vessel surface, and viscosity η is approximated as 0.0083 P for Krebs-HEPES buffer at 37°C (22).
We designed our flow chamber to accommodate the maximum output of the tubing pump (Ismatec), while maintaining physiologically relevant experimental shear values (0–10 dyn/cm2). The aorta strip lays on a raised platform and is fixed in place with the “sandwich” plate, creating a continuous channel of uniform geometry. The area of aorta exposed to fluid flow is 6 mm long by 3 mm wide. The chamber is closed with a glass slide and a Series 20 platform (Warner Instruments, Hamden, CT). A Series 20 stage adapter (Warner Instruments) mates with the platform to yield a seamless integration of the flow chamber and the inverted microscope.
Shear stress protocol.
After the 0.5-h loading time, the aorta was placed on its platform, and the flow chamber was assembled. The direction of physiological blood flow was conserved in our flow chamber arrangement. The tubing of the flow apparatus was prefilled with Krebs-HEPES buffer and maintained at 37°C with an inline heater (Warner Instruments). Flow was initiated at ∼1.3 ml/min, or 0.4 dyn/cm2. For 10 min, the flow was not recirculated to wash the vessel of unhydrolyzed DAF-FM DA. Following the stabilization period, flow was recirculated, and fluorescence measurement began. Vessels were continuously excited during the experiments, and the fluorescence was sampled at 0.5-s intervals. Baseline fluorescence was recorded for 0.5 h, after which the shear stress was increased to 6.4 dyn/cm2 for another 0.5 h. At the end of the shear stress response stage, the flow was gradually reduced to its initial value, and the recycled buffer was exchanged for fresh buffer. At this low flow, baseline fluorescence was reestablished, followed by the application of 10 and 100 μM SNP.
The raw fluorescence data were processed with a Savitzky-Golay smoothing function and normalized by initial photon counts (Felix32 software, Photon Technology International). We performed linear fits of selected data to obtain normalized fluorescence rates. The shear response linear fit data did not include the 5 min immediately after a flow change, to allow for response stabilization. For SNP responses, the linear fit region began 2 min after drug application. Fluorescent rate percent change was calculated by finding the percent difference between a fluorescent slope and its respective baseline, low shear stress slope.
eNOS, p-eNOS, Akt, and p-Akt Western blots.
E+ and E− aorta from O and Y rats were homogenized and centrifuged for 30 min at 14,000 g. The protein content of the supernatant was analyzed by the method of Bradford. Protein (100 μg) from the aorta homogenates was electrophoresed. The blots were incubated with their respective monoclonal anti-eNOS (Transduction Laboratories, San Jose, CA), anti-p-eNOS, anti-Akt, and anti-p-Akt (Cell Signaling Technology, Beverly, MA) antibodies, followed by a goat anti-mouse horseradish peroxidase-conjugated secondary antibody. To quantify the resultant blots, individual band intensities were measured (arbitrary units) and the p/t ratios were calculated per lane. The ratios were normalized by the mean Y E+ ratio.
The animals were anesthetized in a closed chamber with isoflurane. Anesthesia was maintained by mask ventilation of 1.5% isoflurane (in 100% O2) with a coupled charcoal scavenging system. Animals were positioned supine on a temperature-controlled printed circuit board (Indus Instruments, Houston, TX) with legs and arms taped to incorporated electrocardiogram electrodes. Body temperature was monitored with a rectal probe (Physitemp, Clifton, NJ) and maintained at 37°C.
Doppler spectrograms of aortic outflow were acquired with a 2-mm-diameter, 10-MHz pulsed Doppler probe (Indus Instruments). Thoracic aortic outflow and abdominal aortic flow profiles were captured. The distance separating the probe locations was also measured. Aortic PWV is calculated as the quotient of the separation distance and the time difference between pulse arrivals, with respect to the R-peak of the ECG. Data analysis was done offline using DSPW software (Indus Instruments).
Data are presented as means ± SE, with sample size (n) being indicated for each reported value. A value of P < 0.05 was considered statistically significant. Vascular response data were analyzed offline with PRISM data analysis software (GraphPad). Dose-response curves were fitted with the sigmoidal dose-response function through nonlinear regression, returning the EC50 and maximal response parameters. Comparisons of these best fit values were accomplished with an F-test. Two-way ANOVA with Bonferroni posttests was employed to compare responses at each concentration dose. Statistical analysis for variance between groups was performed using two-tailed Student t-tests. Paired t-tests were employed for the comparison of data for the same vessel or animal. Unpaired t-tests were used to compare results from different vessels or animals.
Aging and aortic ring vasorelaxation.
The endothelium-dependent agonist ACh induced a dose-dependent relaxation in both Y E+ and O E+ aortic rings (Fig. 1A). However, O E+ aortic rings demonstrate a significant attenuation of the response, as characterized by a lower maximal relaxation compared with that of Y E+ rings (45.4 ± 1.78 vs. 72.4 ± 2.78%, P < 0.001). The ACh dose responses of O E− and Y E− vessels are nearly abolished. Furthermore, the vasorelaxation of O and Y E− rings to the endothelium-independent vasodilator and direct NO donor SNP were similar (maximal response: 99.0 ± 1.94 vs. 100.6 ± 2.10%) (Fig. 1B). These findings support the notion that age-related endothelial dysfunction results from impaired NO bioavailability, which is likely a function of diminished eNOS activity.
DAF-FM fluorescence response in aortic flow chamber.
To verify our ability to detect DAF-FM fluorescence changes in rat aorta strips when mounted in the flow chamber system, we exposed vessels with and without DAF-FM DA incubation to SNP. The fluorescence rate change responses to 100 μM SNP were 1,680 ± 221% for young DAF-FM vessels and 199 ± 2.69% for vessels without DAF-FM (P < 0.05) (Fig. 2A). Moreover, no significant difference exists between the DAF-FM-loaded Y and O aorta in response to SNP (Fig. 2A). Thus DAF-FM-loaded aorta does produce a detectable fluorescence response, independent of vessel age and indicative of NO concentration, with minimal autofluorescence artifacts.
Next, we wished to investigate shear stress-induced NO production at a low baseline shear of 0.4 dyn/cm2 and a high shear of 6.4 dyn/cm2. Both Y and O aorta loaded with DAF-FM demonstrate positive fluorescent rates that increase with elevated shear stress (Fig. 2B). Yet the normalized rates of O aorta are markedly lower than those of Y aorta, particularly at high shear stress. Vessels without DAF-FM exhibit a negligible fluorescent response at both shear stresses. Therefore, shear stress-regulated DAF-FM fluorescence variations are distinguishable with the flow chamber system.
Following validation of shear stress-dependent DAF-FM fluorescence, we tested whether the response could be reduced through NOS inhibition or through aging. In untreated vessels, the shear step resulted in a fluorescent rate increase of 70.6 ± 13.9% (Fig. 2C). In marked contrast, aortic strips incubated with l-NMMA demonstrated a fluorescent rate decrease of −6.53 ± 6.79% (P < 0.01). This subsequent attenuation of fluorescent response supports the hypothesis that shear stress augments NO production through NOS activation. More importantly, when O rat aorta was exposed to identical shear conditions, a significantly attenuated fluorescent rate change was observed (O vs. Y: 23.6 ± 11.3 vs. 70.6 ± 13.9%, P < 0.05) (Fig. 2C). Hence, the diminished fluorescent rate change of the O rat aorta is indicative of decreased NOS responsiveness to shear stress.
The t- and p-eNOS and Akt concentrations.
Given the mechanosensitive role of Akt in activating NOS by phosphorylation, we tested whether eNOS phosphorylation and Akt phosphorylation are attenuated in O rat aortic tissue. As is demonstrated in Fig. 3, A and B, there is a significant decrease in the p/t-Akt ratio in E+ aortic tissue of O rats compared with that of Y rats (0.500 ± 0.078 vs. 1.00 ± 0.131, P < 0.05). By comparing the E+ and E− p/t-Akt ratios, we demonstrate that aged endothelium contains lower p-Akt levels compared with Y (Fig. 3B). Furthermore, this is associated with a significant decrease of the p/t-eNOS ratio in O E+ rat aortic tissue with respect to the Y rat value (0.270 ± 0.653 vs. 1.00 ± 0.110, P < 0.001) (Fig. 3, A and C). In Y and O E− tissue, eNOS is not detected.
Correspondingly, we demonstrate in Fig. 3D that relaxation elicited by S1P, an activator of the PI3K/Akt phosphorylation cascade, is significantly attenuated in O E+ rings compared with Y E+ rings (10−6 M: 3.90 ± 1.39 vs. 11.1 ± 3.64%, P < 0.05, and 10−5 M: 25.6 ± 2.28 vs. 38.0 ± 6.19%, P < 0.05). Sequential addition of ACh (10−5 M) caused further relaxation with magnitudes similar to the dose-response data of Fig. 1A. Taken together, this supports the idea that PI3K/Akt phosphorylation-dependent activation of eNOS is impaired in the endothelium of aging rat aorta.
Given our findings of age-dependent endothelial function and the critical importance of vascular stiffness as an independent predictor of adverse cardiovascular outcome (47), we wished to test the in vivo physiological consequences of impaired NO signaling on aortic stiffness. We utilized PWV measurements as an indicator of vascular distensibility. In Y rats, we measured the PWV to be 3.64 ± 0.068 m/s. The O rats return a significantly greater PWV (5.99 ± 0.191 m/s, P < 0.001) (Fig. 4A), representing a considerable increase in aortic stiffness. A 6-day administration of the NOS inhibitor l-NAME to Y rats produced a significant elevation in PWV (4.96 ± 0.118 m/s, P < 0.001), supporting the idea that NOS plays an active role in modulating vascular stiffness. This idea is further corroborated by Fig. 4B, which demonstrates an increased PWV in eNOS-KO mice compared with wild-type mice (4.44 ± 0.150 vs. 3.47 ± 0.023 m/s, P < 0.001).
Arterial stiffness develops with aging (26) and is observed in several cardiovascular diseases, including atherosclerosis (29), hypertension (6), and diabetes (47, 48). Endothelial dysfunction is closely associated with these conditions and is thought to contribute to the pathophysiological process (44). In general, endothelial NO signaling is controlled by eNOS activation in response to mechanical and chemical stimuli. Although several constituents of NOS regulation have been identified (2, 9, 13, 37, 43), the mechanism contributing to reduced enzyme activity with age has not been fully appreciated. This study demonstrates age-related endothelial dysfunction through diminished vasodilatation and impaired shear stress-induced NO production in intact rat aorta. Physiologically, this manifests as increased vascular stiffness. Furthermore, decreases in p-Akt and p-eNOS levels with aging appear to be critical components of attenuated NOS activity and the resulting increase of arterial stiffness.
Aging affects vascular tone.
Our results of aortic ring vascular tone show diminished vasorelaxation in aged vessels. This is apparent in the ACh dose response, as well as in the responses induced by S1P. Since ACh is a known eNOS activator and the response is abolished by endothelial removal, this strongly suggests that the attenuated relaxation is mediated by NOS. NOS stimulation by ACh is Ca2+ dependent, with the primary mechanism proposed to be Ca2+/CaM binding to eNOS, thus partially activating the enzyme (13, 43). However, Fleming et al. (14) propose additional stimulation arises from phosphorylation cascades that are sensitive to intracellular Ca2+, such as CaMKII phosphorylation of Ser1177. The existence of two additive mechanisms of NOS activation, and the potential inhibition of the latter with aging, can explain the fractional loss of vasorelaxation in aged vessels. The attenuated vasoresponsiveness of O aorta to S1P illustrates this, as S1P modulates NOS activity principally through the PI3K/Akt phosphorylation cascade (8, 28, 32). Nofer et al. (32) reported comparable S1P-induced relaxation magnitudes in isolated aorta of female Wistar rats, and Morales-Ruiz et al. (28) showed enhanced NO production in response to S1P. Moreover, cultured endothelial cells incubated with S1P express elevated Akt and eNOS phosphorylation (8, 28, 32), suggesting a mechanism for the attenuated response in aged aorta. Our control experiments with the NO donor SNP demonstrate that Y and O vessels respond similarly to an extrinsic NO source. Although the arteries of O rats are expected to have diminished compliance (31, 47), Fig. 1B does not suggest that this compromises the NO-induced vasodilatation. Hence the impaired vasorelaxation in O vessels can be attributed to impaired endothelial NO signaling as a consequence of inhibited eNOS activity.
Shear stress-induced NO production.
Our findings presented in Fig. 2 confirm the hypothesis that shear stress regulates NO production through NOS. While untreated aorta produced a considerable increase in NO production, the fluorescent response in aortic strips following NOS inhibition is eliminated. The effect of a similar NOS inhibitor, l-NAME, has been investigated in several cultured cell studies. Qiu et al. (37) demonstrated increasing 4,5-diaminofluorescein (DAF-2) fluorescence correlating with shear stress magnitude. With NOS inhibition by l-NAME, the DAF-2 fluorescence change was abolished. Ye et al. (49) showed an analogous response to shear magnitude and l-NAME in cultured rat endothelial cells. In addition, Pittner et al. (36) showed DAF-2 fluorescence attenuation by l-NAME in vivo.
We also observed that O rat aorta produces less NO than Y rat aorta under equivalent flow conditions. The fluorescent response to a shear stress increase in O aorta was significantly reduced from that of Y aorta, implying diminished shear stress-dependent NO production. The abated NO output is likely a direct consequence of diminished NOS sensitivity to shear stress and may be an indirect consequence of increased ROS production in aged vessels (3, 10). Our p/t-Akt and p/t-eNOS data further support the idea of impaired NOS sensitivity in aging. Nonetheless, the role of ROS as a powerful NO scavenger needs also to be considered (37, 38). Sun et al. (42) demonstrated reduced NO release in response to shear stress in isolated mesenteric arteries of aged rats through fluorometric assessment of nitrites. At 1 dyn/cm2, 24-mo-old rat arteries produced ∼75% of that of the 6-mo-old rats. At 5 and 10 dyn/cm2, the NO production of the O rat arteries was reduced to ∼50% of the Y arteries. These NO production ratios between Y and O vessels are consistent with our results.
eNOS and Akt activity.
Unlike ACh-stimulated NO production, shear stress-induced NO generation is primarily dependent on activation of NOS by phosphorylation (11, 13, 43). Dimmeler et al. (9) have identified Akt as the terminal kinase acting on NOS in this mechanically modulated signal transduction pathway. It was recently discovered by Peng et al. (34) that Akt and NOS phosphorylation are lower in an in vitro model of endothelial cells lining stiff tubes, compared with compliant tubes. These and other studies imply that shear stress and wall stretch initiate distinct upstream pathways that converge at the activation/phosphorylation of Akt by PI3K (19, 24). Hence our shear stress-induced NO production results and the age-associated reduction in vascular wall compliance implicate the PI3K/Akt pathway as a promoter of age-related endothelial dysfunction.
As validation of this hypothesis, we discovered attenuated p/t-Akt expression ratio in O aortic endothelium vs. Y endothelium. Following the accepted mechanotransduction pathway of NOS activation by PI3K/Akt, we demonstrated decreased phosphorylation of Ser1177-eNOS in O rat aorta. Our findings of diminished responsiveness to S1P in aged aorta and its role in inducing Akt/eNOS phosphorylation (8, 28, 32) support this data. Thus, along with age-associated ROS increases (3, 10) and decreased substrate availability (2, 27), reduced p-Akt also appears to impair NOS activity and, thereby, endothelial function. The work of Hambrecht et al. (15) supports this finding through their study of human patients with coronary artery disease. Patients who undertook exercise training were reported to show improved endothelial function in vivo and significant increases in p/t-Akt and p/t-eNOS ratios. As our experiment was performed on rat aortic tissue without intervention, our results are representative of the basal physiological states of Y and O rats. Hence, the results are not directly comparable to the flow chamber measurements, where the stimulus is controlled and equally applied, but they are concordant with the PWV measurements. We demonstrated an increased PWV in O rats, but we did not quantify aorta luminal diameter, which is considered to enlarge with age. Therefore, the mean shear stress acting on the luminal wall is unknown, and the decreased p/t-Akt ratio may be a function of lessened shear stimulation. Further investigation of mean aortic shear stress is necessary to address this potential limitation. Regardless, when considering our results as a whole, it is more likely that the reduced Akt phosphorylation is a consequence of inhibited shear stress mechanotransduction.
Finally, we examined the physiological effects of cardiovascular aging by measurement of PWV. By simple approximation of a distensible pipe (Moens-Korteweg equation), PWV is quadratically proportional to vascular stiffness (16). We observed a significantly greater aortic PWV in O rats, compared with Y rats, indicating an affiliated increase in aortic stiffness. This consequence of aging has been previously confirmed in humans, rats, and mice (16, 18, 45). However, the source of this stiffening has been primarily credited to alterations in passive mechanical properties of aged vessels, such as increased collagen-to-elastin ratio, intimal thickening, and progressive dilation (31, 47). Our finding of increased PWV in eNOS-KO mice suggests that eNOS, and the associated NO signaling, contributes to arterial compliance. We further demonstrated that Y rats developed the aged phenotype of elevated PWV and aortic stiffness following NOS inhibition. Acute l-NAME injections have been shown to produce similar PWV shifts (12), and this effect steadily increases during chronic l-NAME administration (20). The short-term treatment of our study ensures minimal vascular remodeling, thus producing compelling evidence that arterial stiffness is also actively regulated by the endothelium. Similar claims have been made by Jadhav and Kadam (18) and Wang et al. (45) in the context of atherosclerosis. Our findings that old aorta demonstrate attenuated relaxation to ACh and S1P, yet Y and O E− aorta produce similar vasorelaxation in response to equivalent exogenous NO, suggest that NO release, rather than vascular wall composition, has a more prominent effect on vascular tone. In addition, with the novel method of direct, real-time fluorescent measurement of NO in intact aortic tissue, we demonstrated a lower shear stress-induced NO production in aged vessels. We also revealed a possible mechanism for age-related endothelial dysfunction: the impaired Akt-dependent phosphorylation of eNOS. Thus increased vascular stiffness in aging is associated with reduced NO bioavailability and is most likely a direct consequence of attenuated eNOS activation.
The major finding of our study is that of reduced Akt and eNOS phosphorylation in aged rat endothelium. Since Akt is a critical kinase in the shear stress mechanotransduction pathway regulating NOS activity, this supplies a mechanism for diminished NO production and decreased vascular compliance in aged vessels, as observed in the present and previous studies (31, 42, 47, 49). Furthermore, when coupled with the recent findings of stretch-dependent Akt activation (24) and concomitant downstream NOS regulation (34), the results of this study directly implicate age-related arterial stiffening in the progression of endothelial dysfunction. Vascular stiffening, by vessel remodeling and atherosclerotic plaques, will attenuate vessel stretch, thus eliminating a component of Akt activity. Accordingly, less compliant arteries would not allow for full NOS activation, resulting in impaired vasorelaxation and further enhancing the apparent vascular stiffness. Since the majority of endothelial dysfunction studies investigating molecular mechanisms employ traditionally cultured cells, the stretch-regulated component of Akt/NOS activation has been overlooked. Distinguishing the contributions of shear- and stretch-dependent NOS activation may prove critical for characterizing the initiation and propagation of endothelial dysfunction in aging and other disease models.
The development of innovative tissue- and animal-based systems is promoting the combined study of arterial compliance and endothelial function (18, 45). While aging was the focus of this study, many other cardiovascular diseases, such as hypertension, diabetes, and atherosclerosis, are associated with vascular stiffening and endothelial dysfunction. Furthermore, these disorders share various features with aging (3, 31, 47), potentially implicating Akt as a component in these vascular disease processes. In conclusion, the emergence of two distinct NOS-activating mechanotransduction pathways converging at PI3K/Akt will provide attractive targets for therapeutic intervention and will facilitate the enhanced understanding of cardiovascular disease progression.
This work was supported by grants from the National Institute on Aging (RO1 AG-021523) and National Aeronautics and Space Administration (NNH04ZUU005N and through the National Space Biomedical Research Institute, CA00405).
We acknowledge Dr. Derek S. Damron and his laboratory at the Cleveland Clinic for providing us with initial instruction and training on the spectrofluorometer usage.
↵1 The online version of this article contains supplemental data.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society