Sirt1 during childhood is associated with microvascular function later in life

Abstract

Microvascular dysfunction often precedes other age-related macrovascular conditions and predicts future cardiovascular risk. Sirtuin 1 (Sirt1) has recently emerged as a protein that protects the vasculature and reduces the risk of cardiovascular diseases. We tested the hypothesis that lower Sirt1 during childhood is associated with a reduced microvascular function during adulthood. Thirty-four adults (34 ± 3 yr) from the Augusta Heart Study returned to participate in the present clinical observational study. Sirt1 was assessed in samples collected during both adulthood and participants’ childhood (16 ± 3 yr), and data were divided based on childhood Sirt1 concentrations: <3 ng/dL (LowCS; n = 16) and ≥3 ng/dL (HighCS; n = 18). MVF was evaluated in all of the adults using laser-Doppler flowmetry coupled with three vascular reactivity tests: 1) local thermal hyperemia (LTH), 2) post-occlusive reactive hyperemia (PORH), and 3) iontophoresis of acetylcholine (ACh). The hyperemic response to LTH was significantly (P ≤ 0.044) lower in the LowCS than in the HighCS group. Similarly, the LowCS also exhibited an ameliorated (P ≤ 0.045) response to the PORH test and lower (P ≤ 0.008) vasodilation in response to iontophoresis of ACh when compared with the HighCS. Positive relationships were identified between childhood Sirt1 and all MVF reactivity tests (r≥0.367, P ≤ 0.004). Novel observations suggest that lower Sirt1 during childhood is associated with premature microvascular dysfunction in adulthood. These findings provide evidence that Sirt1 may play a critical role in microvascular function and have therapeutic potential for the prevention of age-associated vascular dysfunction in humans.

NEW & NOTEWORTHY With a longitudinal cohort, novel observations from the present study demonstrate that individuals who had lower Sirt1 early in life exhibit premature microvascular dysfunction during adulthood and may be at higher risk to develop CVD. These results provide experimental evidence that Sirt1 may play an important role in microvascular function with age and represent a potential therapeutic target to prevent premature vascular dysfunction.

INTRODUCTION

Microvascular dysfunction is considered a crucial mechanism in the development and progression of cardiovascular disease (CVD). Indeed, a dysfunctional microvasculature often precedes other macrovascular pathological processes (27, 29) and is associated with increased cardiovascular mortality (36). The cutaneous microcirculation represents an accessible model of generalized microvascular function through the assessment of endothelial-dependent vasodilatory mechanisms (23, 27). Numerous studies have identified cutaneous microvascular dysfunction in a variety of cardiovascular pathologies (29, 59, 70) that are accompanied with reduced nitric oxide (NO) bioavailability (23) and conduit vessel endothelial dysfunction (9, 24, 28).

Sirtuin 1 (Sirt1) belongs to a class of highly conserved histone deacetylases that regulate critical biological functions involved in aging (8, 38, 67). Sirt1 prevents premature senescence through a large number of mechanisms. Indeed, low expression and/or the dysregulation of this protein are associated with vascular dysfunction and overt CVD (2, 69). Conversely, greater Sirt1 improves endothelial function, reduces oxidative stress, ameliorates systemic inflammation, and, overall, exerts cardioprotective effects (44, 74). Sirt1 mediates vascular protection through regulating different processes, including the deacetylation of the endothelial nitric oxide synthase (eNOS) and the associated increase in NO synthesis (13). Indirectly, Sirt1 reduces oxidative stress, preserving the degradation of NO and increasing endothelial-dependent vasodilation (77). Thus, targeting an increase in Sirt1 is emerging as a potential therapeutic approach for different cardiovascular (CV)-related diseases (48, 51, 56, 65).

To date, there have been very few studies that have assessed the impact that Sirt1 has on microvascular function in humans. Thus, the purpose of the present study was to examine whether Sirt1 during childhood could be associated with microvascular function later in life. We hypothesized that lower concentrations of Sirt1 during childhood would be associated with microvascular dysfunction during early adulthood.

MATERIALS AND METHODS

Experimental design.

All participants presented to the Laboratory of Integrative Vascular and Exercise Physiology (LIVEP) on two separate occasions, a preliminary day and an experimental day. The preliminary day consisted of the assessment of body composition, blood pressure, blood oxygen saturation, and standard clinical laboratory values. For the experimental visit, participants reported to the LIVEP at 8:00 am following an overnight fast and having abstained from tobacco, caffeine, and vigorous physical activity for 24 h and from vitamin supplementation for 72 h. A standard venipuncture blood draw was performed, and assessments of microvascular function and arterial stiffness were conducted. Data were analyzed based on concentrations of Sirt1 during childhood between those individuals who had concentrations <3 ng/mL (LowCS) and those who exhibited concentrations >3 ng/mL (HighCS). The cutoff point was established after a percentile assessment of available published data that analyzed Sirt1 using similar methodology (median/Q2 = 3.12 ng/mL) (1, 15, 30, 39, 40) and our present data (median/Q2 = 2.96 ng/mL). Investigators were blinded to Sirt1 childhood concentrations during data collection and analysis.

Participants.

Thirty-four adults (34 ± 3 yr) that had been participating in the Augusta Heart Study (25), a longitudinal investigation evaluating cardiovascular health, were recruited to participate in the present study. Participants were excluded from the present investigation if they 1) had a body mass index (BMI) >40 kg/m² (class III obesity), 2) were postmenopausal or pregnant, 3) were diagnosed with any cardiovascular, pulmonary, renal, hepatic, cerebral, or metabolic disease, 4) were prescribed any vasoactive medications (i.e., nitrates, β-blockers, angiotensin-converting enzyme inhibitors, PDE-5 inhibitors, etc.), or 5) had symptoms of uncontrolled hypertension. All participants were informed of the objectives and possible risks of the investigation before written consent/assent for participation was obtained. The study followed the principles of the Declaration of Helsinki and was approved by the Institutional Review Board at Augusta University (No. 611895).

Demographic characteristics and clinical laboratory values.

Standard anthropometric assessments of height, weight, calculated body mass index (BMI), and body composition using dual-energy X-ray absorptiometry (DXA; QDR-4500W; Hologic, Inc.) to evaluate the percentage of body fat were performed. Resting systolic and diastolic blood pressures were evaluated in triplicate using established protocols (26), and the average of three assessments was used to calculate mean arterial pressure.

Following an overnight fast, venous blood samples (∼30 ml) were collected from the antecubital vein into EDTA Vacutainer systems (BD, Franklin Lakes, NJ). All blood samples were centrifuged at 3,000 rpm at 4°C for 10 min to separate plasma. Plasma samples were stored at −80°C for future analysis. Fasting concentrations of standard biochemical values for lipids [total cholesterol (TC), high-density lipoproteins (HDL), low-density lipoproteins (LDL), and triglycerides] and glucose concentrations were obtained using the Cholestech LDX point of care analyzer (Alere, Providence, RI). Hemoglobin and hematocrit values were obtained using the HemoPoint H2 analyzer (Stanbio Laboratory, Boerne, TX). Concentrations of high-sensitivity C-reactive protein (hsCRP), hemoglobin A_1c (HbA_1c), estradiol, progesterone, and testosterone were obtained from standard core laboratory techniques (Laboratory Corporation of America Holdings, Burlington, NC).

Concentrations of Sirt1.

Plasma samples collected at two different points were used to assess Sirt1 concentrations. Samples were collected as part of the Augusta Heart Study, when participants were in their childhood (16 ± 3 yr), and samples were collected during the actual experimental visit when participants were adults (34 ± 3 yr). Samples were stored at −80°C until the assessment. Circulating concentrations of Sirt1 were measured using a highly sensitive enzyme-linked immunosorbent assay Kit (MyBiosource) with a detection range from 0.78 ng/mL to 50 ng/mL, an intra-assay coefficient of variation of ≤8%, and an interassay coefficient of ≤12%.

Cutaneous microvascular function.

Changes in cutaneous microvascular blood flux were quantified using laser-Doppler flowmetry (MoorVMS-LDF; Moor Instruments). Briefly, the right arm of each participant was extended laterally at ∼80° of shoulder abduction, and the distal forearm was secured in a vacuum-packed pillow (Vacpac). A forearm cuff was placed immediately distal to the medial epicondyle, and an iontophoresis gel electrode was placed on the dorsal part of the wrist. A heater probe and an iontophoresis chamber were placed on the ventral surface of the forearm using double-sided adhesive tape. The placement was carefully selected, avoiding any area with hair, broken skin, areas of skin pigmentation (or tattoos), and visible veins.

After a 20-min acclimation period and in a temperature-controlled room (22 ± 2°C) to achieve a hemodynamic steady state, microvascular function was determined using three endothelial-dependent vasodilation reactivity tests: 1) local thermal hyperemia (LTH), 2) post-occlusive reactive hyperemia (PORH), and 3) iontophoresis of acetylcholine. Baseline (BL) flux was determined by calculating a 30-s average before the pertinent protocol was started, and we assured that flux after the PORH protocol returned to baseline flux before LTH or iontophoresis was started. A biological zero (B0), to control for the Brownian movement of macromolecules in cutaneous interstitial space, was determined while the forearm cuff was inflated during the PORH protocol and subtracted from both baseline and peak responses. Cutaneous blood flow was indexed as red blood cell flux (RBF) in perfusion units (PU) and as cutaneous vascular conductance (CVC): RBF divided by mean arterial blood pressure. Results are presented as 1) the maximal hyperemic response (peak), 2) the area under the curve (area), 3) the relative change in flux expressed as a percentage of maximal dilation (CVC_max), and 4) the time to peak (TTP) that represents the time from the start of the stimuli to the maximal hyperemic response. Skin resistance (SR) was also calculated for every individual using Ohm’s law.

Local thermal hyperemia.

Local thermal hyperemia (LTH) was determined using a heater probe (SH02-SHP1H; Moor Instruments) placed proximal to the occlusion cuff on the ventral surface of the forearm. The LTH chamber was heated at >0.1°C/s until 44°C for 25 min to elicit maximal microvascular dilation (10). No discomfort was reported from the participants during the heating process, as has been previously described with same temperature (61). The maximal response was quantified as the peak response and expressed as CVC_max. All the other variables related to the heating protocol are defined with the subscript LTH.

Post-occlusive reactive hyperemia.

Endothelial microvascular function was determined by recording the flux on the forearm before and after a 5-min occlusion period to provoke shear stress-mediated reactive hyperemia (10). Occlusion was induced by a forearm cuff placed immediately distal to the medial epicondyle that was inflated to 250 mmHg for 5 min. All variables related to this protocol are defined with the subscript PORH.

Iontophoresis of acetylcholine.

Endothelial-dependent dilation was also evaluated through iontophoresis of acetylcholine (ACh) by filling a ventral chamber (MIC-ION1R-P1; Moor Instruments) distal to the occlusion cuff with 0.2 ml of 2% ACh solution (purity >99% TLC; Sigma-Aldrich). Electrode (cathode) and chamber (anode) were connected to a battery-powered iontophoresis controller (MIC2; Moor Instruments). After a baseline recording, ACh was delivered using an anodal current of 100 μA for 20 s and repeated for a total of seven times with 60-s intervals in between doses. Total electric charge (mC) was calculated by the relation between electric current (μA) and time (s). All variables related to this protocol are defined with the subscript ACh.

Arterial stiffness.

Arterial stiffness was determined noninvasively using the SphygmoCor XCEL system (AtCor Medical, Sydney, Australia). Briefly, with applanation tonometry, augmentation index (AIx) was determined in duplicate from the left brachial artery and normalized for a heart rate of 75 beats/min (AIx75). In addition, carotid-femoral pulse wave velocity (cfPWV) was determined by simultaneously recording electrocardiographic-gated carotid and femoral artery waveforms. Collectively, these assessments of arterial stiffness were used as a marker of macrovascular health.

Endothelial function.

With a 12-MHz linear transducer, simultaneous B-mode and blood velocity profiles (duplex mode) of the brachial artery were obtained (Logiq 7; GE Medical Systems, Milwaukee, WI). A forearm occlusion cuff (D. E. Hokanson, Bellevue, WA), placed immediately distal to the medial epicondyle, was rapidly inflated to 250 mmHg for 5 minu (E-20 rapid cuff inflator; D.E. Hokanson) to induce arterial occlusion and subsequent reactive hyperemia of the brachial artery (60). R-wave gaiting (Accusync 72; Accusync Medical Research Corporation, Milford, CT) was used to capture end-diastolic arterial diameters for automated offline analysis of brachial artery vasodilation (Medical Imaging Applications, Coralville, IA). Peak diameter was determined by the highest 5-s average following cuff release. Shear rate was calculated as follows: shear rate = mean blood velocity × 8/vessel diameter (60). Cumulative shear rate [area under the curve (AUC)] was also calculated based on the trapezoidal rule, every 4 s for the first 20 s following cuff release, and for the rest of the data collection period, every 5 s (20). FMD is expressed as the percent increase in peak diameter from baseline diameter and also normalized by shear and expressed as FMD/shear (20).

Statistical analysis.

The data were analyzed using SPSS version 25 (SPSS Inc., Chicago, IL) and expressed as means ± SE unless otherwise noted. For all statistical analyses, significance was set at P < 0.05. The Shapiro-Wilk test was used to analyze the normality of the measurement distribution. Group differences between participants (LowCS vs. HighCS) were determined by independent group t-tests when normality was met or Mann-Whitney U-tests when normality was not met. The response to the incremental charge of ACh through iontophoresis was analyzed by a two-way ANOVA (group by time). Results were also controlled by potential cofounding variables, including blood pressure, lipid panel, body mass index during both adulthood and childhood, and HbA_1c. Pearson’s correlation coefficients were used to examine the relationships between Sirt1 and microvascular function and between Sirt1 concentrations during childhood and adulthood. Effect size calculations using Cohen’s d were reported for primary outcomes to represent small (Cohen’s d = 0.2), medium (Cohen’s d = 0.5), and large (Cohen’s d = 0.8) effect sizes (6, 32).

RESULTS

Participant characteristics and clinical laboratory values.

Demographic characteristics and laboratory values for participants in the LowCS and HighCS groups are presented in Table 1. No differences in subject demographics and body composition were observed between participants from both groups. Similarly, no differences were identified in the clinical laboratory results between both groups, with the exception of HDL concentrations, which were significantly (P = 0.044) higher in the LowCS group. Glucose, liver panel, hemoglobin, and hematocrit were also similar between groups and all within normal limits. In addition, no differences (P ≥ 0.296) were observed in the hemodynamic variables with similar heart rate and blood pressure between individuals from both groups.

Table 1.

Participant characteristics and laboratory values

Variable	Low CS	High CS	P Value
n	16	18
Sex, men/women	7/9	12/6	0.190
Race, C/AA	7/9	10/8	0.507
Childhood
Age, yr	15 ± 2	16 ± 3	0.528
Height, cm	164 ± 12	169 ± 8	0.358
Weight, kg	67 ± 20	70 ± 16	0.755
BMI, kg/m2	24.8 ± 4.7	24.5 ± 4.5	0.966
Adulthood
Age, yr	33 ± 2	3 ± 3	0.228
Height, cm	170 ± 12	175 ± 6	0.387
Weight, kg	81 ± 22	87 ± 19	0.389
BMI, kg/m2	27.9 ± 5.9	28.5 ± 5.9	0.564
Waist-to-hip ratio	0.9 ± 0.1	0.9 ± 0.2	0.456
Body fat, %	35 ± 10	33 ± 11	0.669
Heart rate, beats/min	65 ± 14	66 ± 7	0.841
SBP, mmHg	118 ± 7	120 ± 8	0.422
DBP, mmHg	74 ± 6	76 ± 6	0.296
MAP, mmHg	89 ± 6	91 ± 6	0.299
PP, mmHg	44 ± 4	45 ± 8	0.407
TC, mg/dL	178 ± 23	174 ± 35	0.721
HDL, mg/dL	59 ± 16	48 ± 13	0.029
LDL, mg/dL	100 ± 29	104 ± 32	0.681
TRIG, mg/dL	99 ± 40	112 ± 69	0.557
TC-to-HDL ratio	3.3 ± 0.9	3.9 ± 1.0	0.104
GLU, mg/dL	91 ± 6	98 ± 18	0.206
HbA1c, %	5.4 ± 0.3	5.6 ± 0.3	0.166
hsCRP, mg/L	1.9 ± 1.5	1.8 ± 1.7	0.931
Hb, g/dL	14.3 ± 1.9	15.4 ± 2.6	0.215
Hct, %	41 ± 8	45 ± 7	0.052
AST, IU/L	19 ± 4	22 ± 8	0.250
ALT, IU/L	17 ± 9	24 ± 14	0.166

Values are means ± SD; boldfaced value indicates statistical significance. AA, African American; AST, aspartate transaminase; ALT, alanine transaminase; BMI, body mass index; C, Caucasian; DBP, diastolic blood pressure; GLU, glucose; Hb, hemoglobin; HbA_1c, hemoglobin A_1c; Hct, hematocrit; HDL, high-density lipoproteins; HighCS, high childhood sirtuin 1 (Sirt1); hs-CRP, high-sensitive C-reactive protein; LDL, low-density lipoproteins; LowCS, low childhood Sirt1; MAP, mean arterial pressure; PP, pulse pressure; SBP, systolic blood pressure; TC, total cholesterol; TRIG, triglycerides.

Concentrations of Sirt1.

A strong, significant (P < 0.001) correlation was identified between concentrations of Sirt1 during childhood and adulthood (r = 0.793; Fig. 1), inferring that the storage of prior samples did not affect the concentration of Sirt1. Notably, childhood concentrations of Sirt1 were significantly (P = 0.001, Cohen’s d = 1.21) lower in the LowCS compared with the HighCS group (2.4 ± 0.1 vs. 4.8 ± 0.5 ng/mL, respectively). In addition, both LowCS and HighCS exhibited a similar reduction in Sirt1 when comparing values from childhood to adulthood (ΔSirt1 = 0.3 ± 0.2 vs. 0.5 ± 0.4 ng/mL, respectively, P = 0.612). Furthermore, concentrations of Sirt1, both at early ages and later in life, tended to be negatively associated with the inflammatory marker hsCRP (childhood: r = −0.313, P = 0.071; adulthood: r = −0.319, P = 0.065).

Fig. 1.

Circulating concentrations of sirtuin 1 (Sirt1) in all of the participants during their childhood (16 ± 3 yr) and adulthood (34 ± 3 yr). Differences were identified using Pearson’s correlations; n = 34. LowCS, low childhood Sirt1 (○); HighCS, high childhood Sirt1 (gray circles).

Cutaneous microvascular function.

Red blood cell flux and CVC data for all three microvascular reactivity tests are presented in Table 2. Notably, no differences (P = 0.725) in skin resistance were identified between groups (LowCS: 160.349 ± 62.031 Ω vs. HighCS: 147.586 ± 68.553 Ω).

Table 2.

Microvascular function during the 3 reactive protocols

Variable	LowCS	HighCS	P Value
LTH
Baseline, PU	10 ± 1	9 ± 1	0.479
Peak, PU	152 ± 9	186 ± 15	0.069
Area, PU/s	181,489 ± 14,931	224,820 ± 13,341	0.037
Baseline, PU/mmHg	0.10 ± 0.02	0.14 ± 0.02	0.206
Peak, PU/mmHg	1.75 ± 0.12	2.14 ± 0.16	0.044
TTP, s	1381 ± 20	1338 ± 19	0.138
PORH
Baseline, PU	8 ± 1	8 ± 1	0.458
Peak, PU	37 ± 3	49 ± 6	0.097
Area, PU/s	3,470 ± 298	4,797 ± 415	0.012
Baseline, PU/mmHg	0.10 ± 0.01	0.10 ± 0.01	0.578
Peak, PU/mmHg	0.52 ± 0.05	0.74 ± 0.09	0.045
TTP, s	29 ± 3	29 ± 4	0.950
Iontophoresis with ACh
Baseline, PU	9 ± 2	10 ± 1	0.487
Peak, PU	54 ± 7	75 ± 8	0.066
Area, PU/s	46,570 ± 5,390	59,758 ± 3,578	0.044
Baseline, PU/mmHg	0.10 ± 0.02	0.12 ± 0.01	0.462
Peak, PU/mmHg	0.54 ± 0.08	0.87 ± 0.09	0.003
TTP, s	15 ± 2	16 ± 1	0.730

Values are means ± SE. Boldfaced values indicate statistical significance. ACh, acetylcholine; AUC, area under the curve; HighCS, high childhood sirtuin 1; LowCS, low childhood sirtuin 1; LTH, local thermal hyperemia; PORH, post-occlusive reactive hyperemia; PU, perfusion units; TTP, time to peak.

Local thermal hyperemia.

Baseline RBF and CVC were similar (P ≥ 0.206) between groups for the LTH reactivity test. Likewise, both groups show a comparable initial peak in response to local heating (RBF, P = 0.627; CVC, P = 0.610). However, during the plateau phase, the LowCS group exhibited a significantly (P = 0.037, Cohen’s d = 0.66) attenuated RBF response and a blunted (P = 0.044, Cohen’s d = 1.14) maximal response (CVC; Fig. 2A) to the thermal provocation when compared with the HighCS group. Similar results (P ≤ 0.044) of maximal response during the thermal protocol were observed after controlling for potential factors that may impact microvascular function, including blood pressure, lipids, HbA_1c, and BMI.

Fig. 2.

Microvascular function assessment in low childhood sirtuin 1 (Sirt1) (LowCS; ○) and high childhood Sirt1 (HighCS; gray circles) during 3 reactivity tests. A: maximal (max) response during local thermal hyperemia (LTH). B: maximal response during post-occlusive reactive hyperemia (PORH). C: cumulative charge response to iontophoresis with acetylcholine (ACh). D: maximal response during iontophoresis with ACh. Group differences were determined by independent group t-tests or Mann-Whitney U-tests, and the response to the incremental charge of ACh through iontophoresis was analyzed by a 2-way ANOVA (group by time); n = 34. CVC, cutaneous vascular conductance; mC, miliCoulombs; PU, perfusion units. *P < 0.05 vs. LowCS.

Post-occlusive reactive hyperemia.

Baseline RBF and CVC were similar between groups (P ≥ 0.458) during the PORH test. However, the HighCS group showed an overall (P = 0.012; Cohen’s d = 1.17) greater hyperemic response to shear stress compared with the LowCS group. A similar response was also observed in the peak CVC (P = 0.045, Cohen’s d = 0.92). In addition, Fig. 2B illustrates the significantly lower (P = 0.023, Cohen’s d = 1.01) maximal CVC hyperemic response (%CVC_max) in the LowCS when compared with the HighCS (31 ± 3 vs. 43 ± 4% CVC_max, respectively). Results were similar (P ≤ 0.044) after controlling for biological covariate factors.

Iontophoresis of acetylcholine.

Consistent with the PORH data, no changes (P ≥ 0.462) in baseline RBF or CVC were observed during iontophoresis with ACh. However, individuals from the LowCS group exhibited an overall (P = 0.044, Cohen’s d = 0.63) diminished response to ACh that was also evident after controlling for changes in blood pressure (P = 0.003, Cohen’s d = 1.08). Figure 2C illustrates the response to the iontophoresis test with ACh plotted against the cumulative current. The drug delivery resulted in a progressive increase in perfusion in individuals from both groups. However, as illustrated in Fig. 2D, a significantly (P ≥ 0.044) greater response to ACh was identified in the HighCS group when compared with the LowCS group beginning at low cumulative current (≥4 mC). Similarly, a significantly (P = 0.008, Cohen’s d = 0.92) lower maximal dilation (%CVC_max) in response to ACh was observed in the LowCS compared with the HighCS group (30 ± 4 vs. 47 ± 4%CVC_max, respectively). Similar results (P ≤ 0.021) were observed after controlling for potential factors impacting microvascular dilation.

Arterial stiffness.

No statistical differences (P ≥ 0.428) in the arterial tonometry assessments for either AIx (18 ± 3 vs. 21 ± 3%, respectively) or adjusted AIx75 (16 ± 3 vs.17 ± 3, respectively) were observed between LowCS and HighCS. Similarly, no differences (P = 0.195) were identified in cfPWV (5.3 ± 0.6 vs. 6.2 ± 0.4, respectively) assessment between LowCS and HighCS groups.

Endothelial function.

No differences (P ≥ 0.857) were observed in the diameter of the brachial artery at baseline between the LowCS and the HighCS groups (3.31 ± 0.33 vs. 3.38 ± 0.22 cm, respectively). Although not statistically significant (P = 0.071), FMD tended to be higher in the HighCS group compared with the LowCS (8.3 ± 0.8 vs. 6.2 ± 0.7%), even when FMD was normalized for shear stress (0.17 ± 0.02 vs. 0.13 ± 0.02%/s, AUC respectively, P = 0.063).

Relationships between Sirt1 and microvascular function.

Significant correlations were identified between all microvascular function reactivity tests and Sirt1 concentrations. Specifically, childhood concentrations of Sirt1 were positively associated with maximal hyperemic responses to LTH (r = 0.596, P < 0.001; Fig. 3A), PORH (r = 0.367, P = 0.004; Fig. 3B), and iontophoresis with ACh (r = 0.450, P = 0.008; Fig. 3C). Similarly, concentrations of Sirt1 during adulthood were also positively associated with the responses to LTH (r = 0.458, P = 0.006), PORH (r = 0.347, P = 0.044), and iontophoresis with ACh (r = 0.412, P = 0.016).

Fig. 3.

Relationship between concentrations of sirtuin 1 (Sirt1) during childhood and microvascular function assessed during adulthood through maximal response to local thermal hyperemia (LTH; A), maximal response to post-occlusive reactive hyperemia (B), and maximal response to iontophoresis with acetylcholine (ACh; C). Differences were identified using Pearson’s correlations coefficients; n = 34. LowCS, low childhood Sirt1 (○); HighCS, high childhood Sirt1 (gray circles). CVC, cutaneous vascular conductance; PU, perfusion units.

DISCUSSION

With a longitudinal cohort, novel observations from the present study demonstrate that concentrations of Sirt1 during childhood are positively associated with microvascular function later in life. Specifically, individuals that had lower Sirt1 early in life exhibit premature microvascular dysfunction and may be at higher risk to develop CVD. These results provide experimental evidence that Sirt1 may play an important role in microvascular function with age and represent a potential therapeutic target to prevent premature vascular dysfunction.

Sirt1 and cardiovascular protection.

Sirt1 has emerged as a regulator of critical biological functions that delay the aging phenotype (8, 38, 67) and protect the development of multiple disorders (63). Indeed, dysregulation of Sirt1 has been associated with many age-related diseases, including CVD (11, 54), diabetes (49), and neurodegenerative syndromes (17). Thus, the evaluation of this protein throughout the lifespan and its relationship with cardiovascular health seems to be a crucial area to be investigated. However, minimal information is available in humans, with the majority of the data completed either in vitro (64, 73) or using animal (37, 47, 55) studies.

Results from our longitudinal cohort reveal that concentrations of Sirt1 during childhood are significantly associated with concentrations of Sirt1 later in life. Sirt1 is known to be degraded by multiple environmental factors and lifestyle choices, including poor diet, physical inactivity, environmental pollutants, or psychological stress, among others (18). Although we cannot rule out the specific contribution that these factors exert in our participants during the 17 yr between samples, our results support that Sirt1 followed a progressive decrease with the aging process in both groups. Given the reliance between Sirt1 and protection against the development of chronic diseases, these results have high clinical relevance, since the preservation of Sirt1 during early years may protect the individual from the development of premature age-related disorders (74). Although not statistically significant, our data also suggest a potential association between lower Sirt1 during both childhood and adulthood and greater expression of a clinical marker of inflammation (hsCRP). Following traditional criteria of stratification (19), 63% of individuals from the LowCS group were at moderate (hsCRP = 1–3 mg/L) or high (hsCRP > 3 mg/L) risk to develop cardiovascular diseases versus the 39% of the participants from the HighCS group. Although these results may support the cardioprotective role of Sirt1, they should be interpreted with caution. Further exploration will be needed to draw more concrete conclusions.

Sirt1 and microvascular function.

The cutaneous microcirculation represents an accessible tissue that serves as a model of generalized microvascular function. As in other vascular beds, microvasculature endothelial dysfunction is one of the earliest indicators of senescence and/or disease and is predictive of future cardiovascular events (27). Abundant literature has also described that the dysregulation of Sirt1 is associated with endothelial dysfunction and subsequent premature vascular dysfunction (4, 13, 75). Results from the present study provide further understanding of previous observations with novel information about the association between Sirt1 during childhood and microvascular function later in life. Notably, our findings suggest that those individuals with lower Sirt1 at early ages exhibit a premature microvascular dysfunction during young adulthood, which may increase the risk to develop CVD. It is important to note that senescence and disease impact both macro- and microvasculature; however, microvascular health has been proposed to be a stronger predictor of long-term outcomes and cardiovascular events compared with the health of larger vessels (43, 68). Our results support this notion since our apparently healthy cohort already exhibits a reduced microvascular dysfunction and a trend toward lower conduit vascular function. Thus, results from the present study may have high clinical relevance and identify a relationship between Sirt1 during childhood and cardiovascular health later in life.

As is true with larger vascular beds, a reduction in endothelial-derived dilators, including NO, negatively affects the functionality of the microvessels (27). In the present study, we evaluated endothelial-dependent vasodilation in the microvasculature using three different reactivity tests. Importantly, the findings of each of these tests were consistent with each other and suggest that lower Sirt1 during childhood is associated with a reduced endothelial-dependent microvascular dilation during adulthood. Indeed, different mechanisms regulate the different responses to the reactivity provocation (44, 45). For example, NO has a predominant role in the maximal dilatory response to thermal provocation (3, 50). The blood flow response to an increase in shear stress is also associated with NO synthesis and with an increase in Sirt1 (5). In addition, iontophoresis of ACh triggers the synthesis and the release of NO, prostaglandins, and endothelium-derived hyperpolarizing factors, causing relaxation of the surrounding smooth muscle (46, 52). Considering that Sirt1 promotes NO synthesis (44, 53) and activates the AMP kinase pathway promoting vasodilatation by releasing other vasodilators (14, 78), it is not surprising that the responses to the three reactivity tests were all diminished in those individuals who exhibited lower Sirt1 during childhood.

Finally, it is worth noting that individuals from both groups exhibited a similar time response to elicit vasodilation during the different reactivity tests. These findings may suggest that the neuromuscular and metabolic mechanisms involved in the microvascular response are still preserved in both groups. These results are not surprising since our cohort comprises young and seemingly healthy individuals that have not yet developed structural changes (22, 57, 66). This hypothesis is also confirmed by the arterial stiffness values that were not only similar between groups but also all within healthy ranges (12, 58). Our results are in line with previous observations that emphasize the predictive role that microvascular function has as an early marker of cardiovascular disease (27). In addition, present findings support the use of Sirt1 early in life as a potential therapeutic target to not only prevent microvascular deterioration but also reduce CVD risk later in life.

Experimental considerations.

Sirt1 has recently emerged as a promising mediator associated with multiple health benefits. Although the data from the present study also support the cardioprotective role of Sirt1, there are some experimental considerations that should be noted when interpreting these results. Sirt1 was originally described as a nuclear protein. However, it shuttles between the nucleus and cytoplasm in response to divergent extracellular stimuli (71). Interestingly, Sirt1 has been measured in circulation (30, 39, 41, 72), although its precise origin and its relationship to tissue content have yet to be completely elucidated. A possibility is that Sirt1 is secreted as part of extracellular vesicles, which is not surprising considering that proteins packaged into these vesicles are part of the mechanisms of biogenesis, and Sirt1 has a critical role in exosome secretion and lysosome function (33–35, 45). This hypothesis has been supported by previous experiments that demonstrated the presence of another sirtuin in exosomes from oligodendroglial cells (16). Importantly, this theory will support prevailing results in the literature that describe reduced circulating Sirt1 in different physiological and disease conditions [i.e., aging (76), frailty (31), obesity (42), Alzheimer’s disease (30), or COPD (7)], since exosome secretion is also decreased during pathological processes (21, 62). Unfortunately, the lack of enough samples from our participants does not allow us to fully explore the analysis of extracellular vesicles in our cohort.

Conclusion.

In conclusion, the findings of the present study provide new insight into the relationship among Sirt1, microvascular function, and cardiovascular health in humans. Our results suggest that individuals with reduced Sirt1 during childhood may exhibit a premature microvascular dysfunction during early adulthood. The impairments in microvascular function and reductions in endothelial-dependent mechanisms observed in the present study expand previous findings that support the role of Sirt1 in the pathophysiology of cardiovascular disease. These results provide evidence that Sirt1 may play a critical role in microvascular protection and may have therapeutic potential for the treatment and prevention of CVD later in life.

GRANTS

This work was supported in part by American Heart Association Grants 16POST31080031 and 19CDA34660318 (to P.R.-M.) and National Heart, Lung, and Blood Institute Grant P01-HL-06999 (to J.S.P., G.H., and R.A.H.).

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

P.R.-M., G.H., J.S.P., and R.A.H. conceived and designed research; P.R.-M. and J.L. performed experiments; P.R.-M. and J.T. analyzed data; P.R.-M. and R.A.H. interpreted results of experiments; P.R.-M. prepared figures; P.R.-M. and R.A.H. drafted manuscript; P.R.-M., J.L., J.T., G.H., J.S.P., and R.A.H. edited and revised manuscript; P.R.-M., J.L., J.T., G.H., J.S.P., and R.A.H. approved final version of manuscript.