Participants.

A total of 75 healthy adults (Table 1) enrolled in studies during a 3-yr period were identified in our laboratory database according to prespecified age (young: 18–30 yr and middle-aged and older: 50–79 yr) and exercise status (sedentary or exercise trained) criteria for this retrospective analysis. These subjects were previously enrolled in one of the following studies from our laboratory (6, 21, 22, 29) and, as such, subject characteristics data from these participants have been previously published. Twelve sedentary and 13 exercising middle-aged and older adults as well as 9 young sedentary adults in whom maximal O₂ consumption (V̇o_2max) and vascular endothelial function (FMD) had been assessed and who met the prespecified criteria were identified for venous EC analyses. To confirm that group-related changes observed in venous ECs were also present in arterial cells, arterial ECs were studied from a separate cohort from our database consisting of 18 sedentary and 14 exercising middle-aged and older adults as well as 9 young sedentary adults. Because of the more invasive nature of arterial EC collection, sample availability was limited in participants in whom brachial artery FMD and V̇o_2max had been assessed. In addition, a sufficient arterial EC sample was only available for the assessment of the p53/p21 pathway.

Participants were nonobese (body mass index <30 kg/m²), nonsmokers, nondiabetic, and free of other clinical diseases as determined by medical history, physical examination, blood chemistry, and resting and exercise 12-lead ECG. Blood pressure, fasting circulating lipids, and glucose were clinically normal in all participants. Sedentary participants performed no regular aerobic exercise (i.e., <30 min/day, ≤2 days/wk) for at least the preceding 2 yr. Exercise-trained participants performed regular vigorous aerobic exercise including competitive cycling, running, and/or triathlons ≥5 days/wk for ≥45 min/session for at least the preceding 5 yr. All procedures were approved by the Institutional Review Board of the University of Colorado Boulder. The nature, benefits, and risks of the study were explained to the volunteers, and their written informed consent was obtained before participation.

Measurements.

All measurements were performed at the University of Colorado Boulder Clinical Translational Research Center after a 12-h overnight fast and 24-h abstention from alcohol and physical exercise.

Participant characteristics.

Body mass index and total body fat as well as resting arterial systolic and diastolic blood pressures were determined as previously described (7, 21). Leisure time physical activity was determined by the modifiable activity questionnaire (20). V̇o_2max was measured during incremental treadmill exercise testing performed to exhaustion (Balke protocol).

EC protein expression.

EC collection and protein expression measurements were performed as previously described (7, 21). Briefly, J-wires were advanced into an antecubital vein or brachial artery, and cells were recovered by washing and centrifugation. Collected cells were fixed with 3.7% formaldehyde and plated on poly-l-lysine-coated slides (Sigma Chemical, St. Louis, MO) and then frozen at −80°C until analysis. At the time of analysis, cells were rehydrated and incubated with primary antibodies against p53 (1:200, Cell Signaling, Danvers, MA), p21 (1:500, Abcam, Cambridge, MA), and p16 (1:200, Abcam) along with Alexa Fluor 647 fluorescent secondary antibody (Invitrogen, Carlsbad, CA). Cells were stained for vascular-endothelial (VE)-cadherin (1:500, Abcam) for positive identification of the endothelial phenotype and DAPI for nuclear integrity. Slides were viewed with a fluorescence microscope (Eclipse Ni-U, Nikon, Melville, NY), and whole cell fluorescence was assessed using Metamorph Software (Molecular Devices, Sunnyvale, CA). EC protein expression data are reported as ratios to human umbilical vein EC protein expression. All EC sampling occurred during subjects’ initial visits to our laboratory; protein expression measurements were performed on EC slides that had been stored at that time for future use.

Vascular endothelial function.

Brachial artery FMD was determined using duplex ultrasonography (PowerVision 6000, Toshiba, multifrequency linear-array transducer) as previously described (7, 8). FMD is expressed as the percent change in diameter from baseline.

Data analysis.

Statistical analyses were performed with SPSS Statistics (IBM, Chicago, IL). Sample size estimates were based on preliminary analysis (n = 4 subjects/group) of venous and arterial EC p21 expression levels in sedentary young and older adults. These estimates indicated that 9 subjects/group would be adequate to detect group differences with >80% power; we increased the number of older adults to be conservative and to account for additional variance associated with EC protein expression in older adults. Main effects were determined by one-way ANOVA, with least significant differences post hoc analyses used to determine prespecified differences between specific groups. A general linear model with covariates and least significant difference post hoc analysis was used to evaluate the influence of participant characteristics that were significantly different between groups within each cohort of participants (i.e., in the cohort from whom venous ECs were collected and in the cohort from whom arterial ECs were collected). Bivariate relations were assessed by Pearson correlation analysis. Statistical significance was set at α < 0.05. All data are presented as means ± SE.

Participant characteristics.

There were no significant within group differences between participants from whom venous or arterial cells were collected, so participant characteristics data were pooled (Table 1). Physical activity levels were similar between older sedentary and young adults but were higher (P < 0.05) in older exercising adults. V̇o_2max, a measure of maximal aerobic exercise capacity, was lower (P < 0.05) in older sedentary adults but not different between young sedentary and older exercising adults.

Clinically normal values for body weight, body mass index, blood pressure, and circulating factors were observed in all groups (Table 1). However, modest differences were observed with age in the participants from whom venous and arterial ECs were collected, including increased body fat percentage as well as total and LDL-cholesterol (all P < 0.05); diastolic blood pressure and fasting blood glucose levels were also increased with age, but only in the venous EC cohort (P < 0.05). Additionally, in the participants from whom venous and arterial ECs were collected, older exercising adults had lower body mass, body mass index, and body fat percentage than older sedentary adults (all P < 0.05); older exercising adults from whom venous ECs were collected also had lower fasting blood glucose levels than older sedentary adults (P < 0.05).

EC protein expression.

Protein expression of p53, p21, and p16 were 116%, 119%, and 128% greater (P < 0.05), respectively, in venous ECs from older sedentary compared with young adults (Fig. 1). Moreover, p53, p21, and p16 levels were similar between young and older exercising adults. These markers of senescence were 68%, 32%, and 71% lower for p53, p21, and p16, respectively, in venous ECs from older exercising adults compared with older sedentary adults (Fig. 1).

Consistent with the observations in venous ECs, although there were differences in the magnitude of group differences, protein expression of p53 was 26% higher and p21 expression was 23% greater (both P < 0.05) in arterial ECs from older sedentary compared with young adults (Fig. 2). Arterial EC p53 and p21 protein expressions were 35% and 39% lower, respectively, in older exercising compared with sedentary older adults and similar to those of young adults (Fig. 2).

Group differences in p53, p21, and p16 EC protein expression remained significant (all P < 0.05) in venous ECs after covarying for body mass, body mass index, body fat percentage, LDL-cholesterol, total cholesterol, diastolic blood pressure, and fasting blood glucose (Table S1 in the Supplemental Material; Supplemental Material for this article is available at the American Journal of Physiology-Heart and Circulatory Physiology website). Group differences in p53 and p21 EC protein expression remained significant (all P < 0.05) in arterial ECs after covarying for body mass, body mass index, body fat percentage, LDL-cholesterol, and total cholesterol (Table S1). Moreover, there were no within-group differences between male and female subjects (P > 0.05).

Relation of EC expression of markers of senescence with EDD.

EDD, as assessed by brachial artery FMD, was reduced in sedentary older adults compared with young adults (4.1 ± 0.7% vs. 8.9 ± 0.9%, P < 0.05). Brachial artery FMD was higher (P < 0.05) in older exercise-trained adults (9.0 ± 0.6%) compared with older sedentary adults and was not different from young adults (P > 0.05). Protein expression of p53 (r = −0.49, P < 0.05), p21 (r = −0.38, P < 0.05), and p16 (r = −0.58, P < 0.05) were all inversely associated with brachial artery FMD (Fig. 3).

EC senescence with age.

In vitro and preclinical data implicate increased cellular senescence as a driver of age-related vascular endothelial dysfunction (4, 10, 15, 23, 24, 31, 33, 34). To our knowledge, the present study is the first to evaluate markers of cellular senescence in the vascular endothelium in primary (i.e., noncultured) ECs obtained from human subjects. These data extend preclinical observations and provide novel, direct evidence of increased cellular senescence in ECs from healthy older adults. Specifically, our data show increased p53 and p21 expression in biopsied venous and arterial ECs and elevated p16 expression in venous ECs, indicative of increased activation of the p53/p21 and p16 tumor suppressor pathways (19). Collectively, our data indicate that EC senescence increases with healthy aging, even in the absence of overt clinical vascular disease.

EC senescence with age and exercise status.

In contrast to their sedentary peers, older exercising adults did not exhibit age-related increases in p53, p21, or p16 expression. These data build on preclinical data indicating that exercise reduces aortic protein expression of p53 and p16 in young mice (30) and prevents the increase in adipose tissue senescence accompanying high-fat feeding (25). More broadly, these data suggest that habitual aerobic exercise preserves a healthy phenotype in the vascular endothelium. This notion is in agreement with data from our laboratory showing reduced oxidative and inflammatory signaling in ECs obtained from older exercising adults compared with sedentary older adults (21, 29). Collectively, these data may provide evidence for a novel molecular mechanism, reduced EC senescence, underlying the vascular protective effects of aerobic exercise with aging, which may contribute to reduced risk of CVD.

Relation of EC senescence with vascular endothelial function.

The participants in the present study exhibited the expected age-related reduction in EDD with sedentary aging, which was not present in older exercising adults (8, 26). Importantly, the age- and exercise-related changes in EDD were associated with EC senescence. These data suggest that senescence occurs in the vascular endothelium with sedentary aging and impairs vascular endothelial function. Moreover, in the absence of age-associated changes in EC senescence, vascular endothelial function is preserved with healthy aging. Importantly, brachial artery FMD is mediated, at least in part, by NO bioavailability. As such, these data extend in vitro data showing reduced NO production by senescent ECs (24) and support the notion that the link between EC senescence and EDD likely involves NO bioavailability.

Limitations.

We recognize several limitations of our study. First, we had a limited number of ECs from arterial sampling and were therefore unable to assess p16 in arterial ECs. However, protein expression assessed in venous ECs reflects expression observed in arterial ECs (27), and expected age-related changes in protein expression are observed in venous ECs using the same procedures as the present study (7). Second, we assessed p53, a transcription factor associated with senescence but did not assess its nuclear localization. Third, we used p53, its downstream effector p21, and p16 as our indicators of cellular senescence. p21 and p16 are cyclin-dependent kinase inhibitors responsible for the maintenance of the senescent state and are established markers of senescence (19). However, evaluation of additional markers of senescence and the SASP are necessary in future studies to more comprehensively evaluate age- and exercise-related changes in EC senescence. Fourth, mechanistic insight could be provided in future studies by evaluating relations between markers of senescence and oxidative stress and inflammation in ECs, which were not assessed in the present study. Fifth, our general linear model statistical approach may have been subject to model overfitting because of the small sample size. Finally, because of our cross-sectional experimental design, the observations in the present study are associational in nature and do not establish causation.

Conclusions.

The results of the present study provide novel evidence in humans for increased EC senescence with healthy aging and a lack of increase in older adults who perform habitual aerobic exercise. Additionally, our data suggest a link between EC senescence and vascular endothelial dysfunction and indicate that a reduction in endothelial senescence may be one mechanism by which regular aerobic exercise affords protection against increased CVD risk with aging.