Improved Cardiovascular Function in Old Mice After N-Acetyl Cysteine and Glycine Supplemented Diet: Inflammation and Mitochondrial Factors

Abstract

Metabolic, inflammatory, and functional changes occur in cardiovascular aging which may stem from oxidative stress and be remediable with antioxidants. Glutathione, an intracellular antioxidant, declines with aging, and supplementation with glutathione precursors, N-acetyl cysteine (NAC) and glycine (Gly), increases tissue glutathione. Thirty-month old mice were fed diets supplemented with NAC or NAC+Gly and, after 7 weeks, cardiac function and molecular studies were performed. The NAC+Gly supplementation improved diastolic function, increasing peak early filling velocity, and reducing relaxation time, left atrial volume, and left ventricle end diastolic pressure. By contrast, cardiac function did not improve with NAC alone. Both diet supplementations decreased cardiac levels of inflammatory mediators; only NAC+Gly reduced leukocyte infiltration. Several mitochondrial genes reduced with aging were upregulated in hearts by NAC+Gly diet supplementation. These Krebs cycle and oxidative phosphorylation enzymes, suggesting improved mitochondrial function, and permeabilized cardiac fibers from NAC+Gly-fed mice produced ATP from carbohydrate and fatty acid sources, whereas fibers from control old mice were less able to utilize fatty acids. Our data indicate that NAC+Gly supplementation can improve diastolic function in the old mouse and may have potential to prevent important morbidities for older people.

Introduction

The aging process is associated with detrimental cardiovascular functional changes, especially cardiac diastolic dysfunction and large artery stiffening (1). The impaired diastolic function leads to limitation in maximum cardiac output and maximum oxygen consumption (2) and increases the likelihood of developing heart failure (3). The changes in the left ventricle lead to increased reliance on the left atrium for diastolic filling causing increased left atrial volume (LAV) (4,5) and, at least in humans, increased risk of atrial fibrillation (6) and cardio-embolic stroke (7). The large artery stiffening contributes to systolic hypertension leading to left ventricular hypertrophy, and heart failure (8,9). Thus reducing or reversing the age-related changes may have important implications for the aging population.

The pathophysiology underlying these processes is unclear and likely multifactorial, but oxidative damage to the myocardium and arteries increases with age, and so could contribute to the aging changes. Mitochondrial dysfunction occurs with aging, especially manifest as impaired fatty acid oxidation (10–12) and preferential utilization of glucose (13). Additionally mitochondria from old hearts release reactive oxygen species (ROS) (10) which stimulate the inflammatory cascade leading to cardiac fibrosis and further impairment in diastolic function (14). Locally produced pro-inflammatory cytokines, stimulated by ROS, can further modify cardiomyocyte function (15). One of the naturally occurring intracellular defense systems against ROS is glutathione (16–18). Glutathione is present in two forms in cells; oxidized glutathione disulfide (GSSG) and reduced sulfhydryl form glutathione (GSH). GSH is able to neutralize ROS by donating hydrogen, which simultaneously causes its conversion to GSSG. N-acetyl cysteine (NAC) and glycine (Gly) are cell permeant precursors of intracellular glutathione. NAC also possesses a direct reducing property through its thiol-disulfide exchange activity (19). With aging, intracellular glutathione is decreased and results in an oxidative redox potential in both mice and humans (20,21). We have used a N-acetyl cysteine and glycine (NAC+Gly) supplemented diet to successfully modify the antioxidant balance in liver and muscle in aging mice, restore glutathione to the levels observed in young animals and modify metabolism to increase fatty acid metabolism (20).

For this work, we used glutathione precursors (NAC or NAC+Gly) enriched diets to test a potential improvement in the heart function in old mice. The NAC+Gly supplementation improved diastolic function, increasing peak early filling velocity, and reducing relaxation time, LAV, and left ventricle end diastolic pressure. By contrast, cardiac function did not improve with NAC alone. Both NAC and NAC+Gly dietary manipulations exerted anti-inflammatory effects and reduced ROS formation, but only the combination diet reduced leukocyte infiltration into the old hearts. However, with NAC+Gly supplementation, there was a striking increase in expression of critical mitochondrial genes in old hearts. Permeabilized cardiac fibers from old, control mice were unable to utilize fatty acids efficiently, NAC+Gly supplementation resulted in normal fatty acid metabolism, similar to what we showed in liver and skeletal muscle (20). Our data indicate that NAC+Gly, but not NAC alone, improved diastolic function, reduced ROS and inflammation, and improved metabolism, in the old mouse heart.

Methods

Animal Protocols and Study Design

Dietary intervention

Male mice C57BL/6J aged 28–30-month old (n = 35) were obtained from the National Institute of Aging. They were inspected for signs of obvious disease and arrhythmias, housed individually at a Baylor College of Medicine vivarium facility, and divided into groups to be provided ad libitum water and either irradiated chow (control group, n = 12) or supplemented diet: irradiated chow + 1.6 g/kg N-acetyl-L-cysteine (NAC group, n = 6) or irradiated chow + 1.6 g/kg N-acetyl-L-cysteine +1.6 g/kg glycine (NAC+Gly group, n = 17) custom prepared by Harland for our experiments. Mice were fed supplemented diet for a period of 7 weeks. There was no change in body weight over the 7-week interval for any of the groups.

All animal protocols were approved by the Institutional Animal Care and Use Committee of Baylor College of Medicine in accordance with the Guidelines of the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Baseline noninvasive Doppler and Echocardiographic studies were performed at the beginning of the protocol and repeated after 7 weeks of dietary intervention; additional, terminal invasive cardiac function measurements were performed after the noninvasive studies followed by studies of biochemistry and molecular biology.

Noninvasive measurements

All studies were performed on anesthetized mice under 1% isoflurane in oxygen. Mice were first placed in anesthesia chambers until recumbent and placed on a 39°C heated ECG board. The limbs of the mouse were taped to electrodes to assess cardiac rhythm, maintaining the heart rate between 400 and 500 beats per minute while anesthesia was continuously given by nose cone. Body fur was shaved from the anterior thorax and abdominal area and acoustic gel was applied to perform the noninvasive measurements. Cardiac Doppler signals were obtained from the mitral inflow tract, aortic root, thoracic aorta, and abdominal aorta using a 10-MHz probe and analyzed using a real-time signal acquisition and spectrum analyzer system, Doppler Signal Processing Workstation (DSPW, Indus Instruments, Houston, TX) (22). Pulse wave velocity was measured from the thoracic aorta and abdominal aorta signals as previously described (23).

Echocardiographic measurements on 2D and M-mode to assess left atrial anatomy and left ventricular function were obtained using a RMV 710B scanhead; data was analyzed using Vevo 770™ VisualSonics software (Visualsonics, Toronto, Canada) (5).

Invasive measurements

Standard dissection techniques were used to expose the right carotid artery for the invasive measurements made on anesthetized mice under 1% isoflurane in oxygen. Mice were placed on a 39°C heated ECG board for the procedure. Blood pressure was obtained from the carotid artery. Left ventricular end-diastolic pressure (LVEDP), +dP/dt_max, −dP/dt_min, and Tau (time constant of relaxation) using a Millar Mikro-tip catheter transducer (Millar Instruments, Houston, TX) after advancing into the LV.

Cellular, Biochemical, and Molecular Studies

Quantitative PCR (qPCR)

Total RNA was isolated using TRIzol reagent (ThermoFisher Scientific), purified by RNeasy kit (Qiagen, Valencia, CA) and transcribed to cDNA by Verso cDNA Synthesis kit (ThermoFisher Scientific). QPCR was performed on CFX96 Real Time PCR Detection System (Bio-Rad, Hercules, CA) using Sso Advanced Universal SYBR Green mix (Bio-Rad) and gene specific primers. The level of expression of a target mRNA was normalized to an endogenous reference (Hprt or Rpl4) via ΔΔCq method. All primers were evaluated according to MIQE guidelines (24–27). Primer sequences are listed in the Supplementary Table 1.

Protein arrays

Proteins were isolated from snap-frozen hearts using lysis buffer (RayBiotech, Norcross, GA) supplemented with Halt Protease and Phosphatase Inhibitor Cocktail (ThermoFisher Scientific). 500 µg of protein from controls, NAC and NAC+Gly-fed mice hearts was loaded onto mouse cytokine antibody array membranes (RayBiotech). Membranes were processed according to manufacturer’s instructions, and visualized using an Odyssey Imaging System (Licor, Lincoln, NE). Integrated density was assessed by ImageJ software version 1.46r (NIH, Bethesda, MA).

Western blot analysis

50 µg of total heart lysate were separated on 4–15% Tris–HCl gel (Bio-Rad). Proteins were then transferred to a nitrocellulose membrane (Bio-Rad). For immunoblotting, membranes were incubated with the blocking buffer and antibodies diluted according to manufacturer’s protocol. Anti-Sdha antibody (#40489) was purchased from One World Lab (San Diego, CA).

Immunofluorescence staining

Paraffin-embedded zinc-Tris fixed mouse heart sections were stained with the anti-CD45-PE antibody (#103106, BioLegend) and Collagen Type Ia (#600-401-103, Rockland Immunochemicals, Gilbertsville, PA). Nuclei were counterstained with DAPI.

Nonmyocyte isolation and cell culture

Hearts were cut into 1 mm³ pieces, digested with Liberase TM (Roche Diagnostics, Indianapolis, IN) and incubated in a 37°C shaking water bath with regular trituration by pipette to obtain a single cell suspension. Afterwards, cells were centrifuged at 250×g for 5 min. The cell pellet was washed and then suspended in fibroblast growth medium (DMEM/F12 supplemented with 10% FBS and 1% antibiotic-antimycotic), from ThermoFisher Scientific (Waltham, MA).

To synchronize cell cycle, cells were incubated in low glucose (1 g/L) DMEM supplemented with 1% antibiotic-antimycotic. The cell cycle was synchronized within 24 h then the medium was changed to fresh low glucose without serum and supplemented with 5 mM NAC and 5 mM Gly (Sigma, St. Louis, MO) as indicated for additional 2–24 h. Number of mouse hearts are indicated in the figure legend.

Flow cytometry analysis of nonmyocytes

Cells were incubated with antibody to CD45-PECy5 (#103110, Biolegend) or the appropriate IgG control. Calcein (ThermoFisher Scientific) was added to samples to determine cell viability. Cells were analyzed on a Cell Lab Quanta SC flow cytometer (Beckman Coulter, Brea, CA) using the Quanta Analysis software and FlowJo (Tree Star, Ashland, OR).

Detection of intracellular GSH by flow cytometry analysis.

Quiescent fibroblasts were incubated with 5 mM NAC and 5 mM Gly (NAC+Gly), with 5 mM NAC or without supplementation for 4 h. Then cells were detached using TrypLE (ThermoFisher Scientific), spun down and resuspended in GSH Assay buffer (Abcam, Cambridge, MA) supplemented with Thiol Green Indicator (1:200) for 20 min. Events were gated on nucleated cells (Draq5 positive). GSH was detected using FL1 channel and Draq5 using FL3 in the same sample.

Detection of superoxide anion

Quiescent fibroblasts were treated as above. Next cells were incubated with 5 μM MitoSox for 30 min and then analyzed by visualizing changes microscopically or quantified by flow cytometry.

Analysis of mitochondrial respiratory function

High-resolution respirometry (OROBOROS Instruments Corp., Innsbruck, Austria) was used to assess mitochondrial oxygen consumption rate in freshly isolated cardiac tissue procured from the apical area of the heart. Small (10 mg) pieces of heart tissue immediately placed in ice-cold biological preservative solution and separated mechanically by fine forceps. Separated fibrils, permeablized in saponin were loaded into each oxygraph chamber pre-filled with mitochondrial respiration medium for assessing oxygen consumption with 5 mM pyruvate + 2 mM malate or 57.5 µM palmitoyl carnitine and 2 mM ADP added. Data were analyzed by DatLab 4 software (OROBOROS Instruments Corp.) and expressed as oxygen flux (pmol/(s*mg)) normalized by muscle wet weight. The respiratory control ratio (RCR), was estimated by state 3 (ADP-dependent respiration)/state2 (ADP-independent and substrate-based respiration) (28).

Mitochondrial isolation and function

Mitochondria were isolated as previously published and respiration measured at 37°C in Seahorse XF24 Extracellular Flux Analyzer (Agilent, CA) with the Cell Mito Stress Test Kit with 10 mM pyruvate and 2 mM malate and run in triplicate (29).

Data analysis

Comparisons between the baseline measurements and control or NAC or NAC+Gly supplemented chow groups at the end of 7 weeks were analyzed using paired Student’s t-test. Changes between the control and the NAC+Gly supplemented chow groups at the end of the 7-week diet regimen were analyzed using ANOVA for the invasive or terminal studies. Statistical significance was defined as p < .05. All data are shown as mean + SE.

Results

NAC+Gly Improves Cardiac Function in the Old Mouse

The heart function of each mouse was examined noninvasively with echocardiography and Doppler before being placed on the diets (pre) and compared to that 7 weeks later (post). Therefore, each mouse serves as its own control. Heart rates were not different between the groups at baseline (Table 1) and not different between the groups after 7 weeks. However, heart rates were higher for both control and NAC+Gly groups at the post diet evaluation compared to their assessment before the diets started. Systolic function as measured by left ventricular fractional shortening (FS%) (Table 1) or +dP/dt measured invasively (Supplementary Table 2) was not different among the groups and not modified by the diet intervention. Peak aortic flow velocity, which correlates with cardiac output, was increased in both the control and NAC+Gly groups.

Table 1.

Noninvasive Measurements of Systolic Function and Myocardial Performance Index

	Control Pre, N = 9	Control 7 weeks, N = 9	Change (%)	NAC Pre, N = 6	NAC 7 weeks, N = 6	Change (%)	NAC+Gly Pre, N = 13	NAC+Gly 7 weeks, N = 13	Change (%)
Echo
Heart rate (bpm)	408 ± 18	451 ± 27	12 ± 7	469 ± 20	440 ± 14	−5 ± 4	423 ± 11	468 ± 13	12 ± 4*
Volume; s (ul)	63.0 ± 5.0	68.2 ± 8.4	8 ± 11	49.2 ± 4.8	62.1 ± 3.3	32 ± 13	64.1 ± 5.6	58.9 ± 6.3	−6 ± 9
Volume; d (ul)	113 ± 5.0	124 ± 9.5	10 ± 6	93.0 ± 8.5	106.5 ± 3.1	19 ± 9	116.1 ± 6.4	112.0 ± 7.4	−2 ± 6
Fractional shortening (%)	22.3 ± 1.2	23.8 ± 2.0	7 ± 8	23.6 ± 0.7	21.8 ± 1.2	−10 ± 7	23.0 ± 1.5	23.0 ± 1.4	4 ± 8
LVPW; d (mm)	0.64 ± 0.05	0.62 ± 0.05	0 ± 6	0.74 ± 0.04	0.77 ± 0.04	6 ± 10	0.64 ± 0.04	0.59 ± 0.04	−4 ± 8
Doppler
Peak aortic flow velocity (cm/s)	95.6 ± 11.8	110 ± 8.5	22 ± 9	84.3 ± 2.7	77.7 ± 2.9	−7 ± 5	83.2 ± 4.4	105 ± 5.2	28 ± 6**
IVCT (ms)	23.8 ± 2.7	20.7 ± 2.8	−2 ± 16	18.0 ± 1.6	18.9 ± 1.7	24 ± 14	20.3 ± 2.1	13.6 ± 0.6	−23 ± 9*
IVCT/RR	0.16 ± 0.01	0.15 ± 0.01	3 ± 14	0.14 ± 0.01	0.14 ± 0.01	4 ± 10	0.14 ± 0.01	0.11 ± 0.01	−16 ± 10*
Tei Index	0.97 ± 0.07	0.79 ± 0.06	−15 ± 8	0.70 ± 0.03	0.78 ± 0.05	13 ± 7	0.88 ± 0.05	0.66 ± 0.04	−22 ± 6*

* denotes p < .05, ** denotes p < .01.

By contrast, there was dramatic improvement in diastolic function with the NAC+Gly diet as reflected by multiple parameters. There was a 24% increase in the peak early filling velocity (Table 2) with the NAC+Gly diet. A typical response is shown with the E peak velocity increasing from Pre to Post intervention on left ventricular inflow (across the mitral valve) after NAC+Gly diet (Figure 1A). Almost all of the individual mice responded positively to the NAC+Gly supplementation by increasing peak early filling velocity, as shown in Figure 1B. There was no significant change in peak E velocity over the 7-week study in the control diet nor the NAC alone diet (Table 2). Tei index (myocardial performance index) also improved in the NAC+Gly group by 22% compared to no change in the control or NAC groups (Table 1). Heart rate corrected Isovolumic Relaxation Time (IVRT/RR) improved in the NAC+Gly group, and was unaltered in the control or NAC groups (Table 2). Therefore, LV filling was improved by NAC+Gly supplementation.

Table 2.

Noninvasive Measurements of Diastolic Function

	Control Pre, N = 9	Control 7 weeks, N = 9	Change (%)	NAC Pre, N = 6	NAC 7 weeks, N = 6	Change (%)	NAC+Gly Pre, N = 13	NAC+Gly 7 weeks, N = 13	Change (%)
Echo
LA anteroposterior (mm)	2.65 ± 0.1	2.82 ± 0.1	8 ± 7	2.83 ± 0.14	3.03 ± 0.09	8 ± 3	2.72 ± 0.1	2.76 ± 0.1	3 ± 4
LA volume (mm3)	35.7 ± 3.4	34.7 ± 1.7	0 ± 7	29.1 ± 2.1	36.2 ± 1.6	26 ± 5	32.3 ± 3.2	26.8 ± 5.6	−16 ± 16
Doppler
E-peak velocity (cm/s)	67.3 ± 3.7	76.6 ± 5.5	16 ± 9	78.4 ± 3.7	79.6 ± 2.8	2 ± 4	69.8 ± 3.6	84.7 ± 4.9	23. ± 7**
A-peak velocity (cm/s)	45.5 ± 3.7	55.3 ± 4.7	23 ± 8*	58.8 ± 2.1	59.8 ± 2.0	2 ± 2	47.6 ± 3.9	63.5 ± 4.2	43 ± 12*
E-A peak velocity ratio	1.51 ± 0.1	1.43 ± 0.1	−5 ± 3	1.33 ± 0.02	1.33 ± 0.03	0 ± 3	1.61 ± 0.2	1.35 ± 0.1	−9 ± 6
IVRT (ms)	28.6 ± 2.1	20.3 ± 1.7	−21 ± 6*	20.0 ± 1.6	22.3 ± 1.2	17 ± 12	24.0 ± 1.6	17.8 ± 1.1	−24 ± 5**
IVRT/RR	0.19 ± 0.1	0.15 ± 0.1	−15 ± 8	0.16 ± 0.01	0.16 ± 0.01	9 ± 9	0.17 ± 0.1	0.14 ± 0.1	−17 ± 4**

* denotes p < .05, ** denotes p < .01.

Figure 1.

Mice fed NAC+Gly improved diastolic function. (A) Transmitral cardiac Doppler signal from before (left – pre NAC+Gly) and after (right—after 7 weeks of dietary intervention with NAC+Gly). The Early filling (E) wave and Atrial contraction (A) wave are labeled. The ECG is shown below the Doppler signals. (B) The data from individual mice in the NAC+Gly group showing the relatively uniform increase produced by the supplementation on peak Early filling velocity after 7 weeks (p < .05).

With the invasive measures (Supplementary Table 2), the left ventricular end diastolic pressure (LVEDP) was more than 50% lower in the NAC+Gly mice than in the control mice. This is critical because increased peak E filling velocity and shorter IVRT/RR seen with NAC+Gly could have been produced by elevated filling pressures instead of improved diastolic function, called pseudonormalization (5). The peak −dP/dt tended to be better in the NAC+Gly mice and the Tau (time constant of relaxation) also tended to be better in the NAC+Gly-fed mice, though they did not attain statistical significance. The reduced LVEDP, increased peak E filling velocity, and improved relaxation are all consistent with enhanced diastolic function in the NAC+Gly-fed mice.

In normal aging, diastolic dysfunction and persistent increases in left ventricular filling pressures drive much of the age-related enlargement of the left atrium (LA). The LA was larger in the control and NAC-fed mice than those given NAC+Gly. There was a tendency for LA volume to decrease in the NAC+Gly mice, suggesting chronically decreased or unchanged filling pressures in the NAC+Gly mice (Table 2).

Arterial Stiffness

We assessed arterial stiffness noninvasively by measuring foot to foot pulse wave velocity at diastolic pressures. There were no differences in pulse wave velocity between the groups after 7 weeks (control 520.0 ± 42.2 cm/s vs NAC 562 ± 71 cm/s vs NAC+Gly 517.4 ± 39.4 cm/s, p = NS) and the diet had no effect on the pulse wave velocity (PWV) comparing readings before diets to after 7 weeks for any of the groups. Similarly there were no differences in blood pressures in the invasive studies between the control and the NAC+Gly mice (Supplementary Table 2). Therefore, the improved cardiac function was unlikely to be mediated by peripheral hemodynamic factors.

Both NAC and NAC+Gly Diet Supplementations Alter the cardiac Inflammatory State in the Old Heart

Because the production of ROS contributes to a transcriptional activation of several cytokines (30,31), and we demonstrated that there is an elevated presence of several cytokines in the aging heart associated with impaired diastolic function (14), cytokines in the whole heart lysate were analyzed by protein arrays (Table 3). Both NAC and NAC+Gly supplementations caused downregulation of several pro-inflammatory cytokines compared to the control old hearts. Unlike NAC alone, NAC+Gly upregulated the anti-inflammatory cytokine, IL-10, suggesting differing anti-inflammatory shifts in the old NAC+Gly treated mouse compared to NAC alone. QPCR analysis demonstrated significant reductions in MCP-1 and IL-6 mRNA levels for both NAC and NAC+Gly (Figure 2A).

Table 3.

Protein Array Analysis

	Control		NAC			NAC+Gly
	Mean	SEM	Mean	SEM	NAC/control	Mean	SEM	NAC+Gly/con
GCSF	0.213	0.019	0.082	0.009	0.38*	0.025	0.096	0.12*
GMCSF	0.341	0.027	0.158	0.035	0.46*	0.192	0.083	0.56*
IFNγ	0.299	0.135	0.059	0.003	0.20*	0.207	0.096	0.69
IL2	0.286	0.008	0.060	0.005	0.21*	0.259	0.074	0.91
IL3	0.194	0.028	0.081	0.017	0.42*	0.057	0.054	0.30*
IL4	0.598	0.034	0.305	0.061	0.51*	0.497	0.056	0.83
IL5	0.336	0.071	0.072	0.016	0.21*	0.187	0.102	0.56*
IL6	0.260	0.081	0.053	0.019	0.20*	0.079	0.070	0.31*
IL9	0.485	0.031	0.084	0.015	0.17*	0.513	0.097	1.06
IL10	0.170	0.028	0.094	0.023	0.55*	0.270	0.082	1.59*
IL12	0.302	0.020	0.176	0.034	0.58*	0.323	0.113	1.07
IL12p70	0.587	0.050	0.362	0.056	0.62	0.558	0.089	0.95
IL13	0.378	0.067	0.195	0.027	0.52*	0.326	0.065	0.86
IL17	0.221	0.102	0.098	0.008	0.44*	0.172	0.072	0.78
MCP-1	0.533	0.030	0.421	0.058	0.79	0.487	0.033	0.91
MCP-5	0.314	0.019	0.233	0.047	0.74	0.385	0.040	1.23
Rantes	0.426	0.011	0.238	0.069	0.56*	0.355	0.034	0.83
SCF	0.326	0.104	0.055	0.015	0.17*	0.223	0.078	0.68
sTNFRI	0.625	0.026	0.109	0.022	0.17*	0.547	0.086	0.88
Thpo	0.445	0.039	0.340	0.064	0.77	0.261	0.034	0.59*
TNFα	0.407	0.060	0.135	0.031	0.33*	0.351	0.087	0.86
VEGF	0.120	0.081	0.008	0.008	0.07*	0.042	0.022	0.35*

Quantitative analysis of densitometry data acquired by analyzing digitalized membrane images. Results are represented as mean ± SEM. N = 3 for each treatment. Reduced (ratio < 1) or upregulated (ratio > 1) expression of cytokines in hearts isolated from mice subjected to NAC or NAC+Gly supplemented diet for 7 weeks when compared with mice fed control chow.

*p < .05.

Figure 2.

Inflammation and leukocyte infiltration in hearts of NAC or NAC+Gly treated mice. (A) qPCR analysis of mRNA expression for pro-inflammatory cytokines, IL-6 and MCP-1 in total RNA isolated from the whole heart. N = 3, 6, 3 for control, NAC, and NAC+Gly, respectively. (B) Immunofluorescence staining of paraffin heart sections shows the reduced number of CD45⁺ cell (red) in hearts of mice fed NAC+Gly supplemented diet compared to control and NAC alone. Scale bar = 50 µm. Yellow frame depicts enlarged view of CD45⁺ cells. Arrows show minimal red staining in the NAC+Gly heart. In addition, the collagen staining was decreased in the NAC+Gly sections. (C) Flow cytometry analysis of infiltrating leukocytes (CD45^hi) in nonmyocytes isolated from hearts of mice fed control and NAC+Gly supplemented diet. Left panel depicts analysis of total number of leukocytes versus infiltrating leukocytes. Central and right panels show representative histograms. N = 3 for each experimental group. * denotes p < .05.

Staining of the heart sections using anti-CD45 antibody, a marker for all leukocytes, showed dramatically decreased numbers of leukocytes in the hearts from the NAC+Gly treated mice but not from NAC-fed mice compared to controls (Figure 2B). In parallel, the collagen content appeared to be decreased in the NAC+Gly-treated mice Figure 2B.

To quantify and further characterize the CD45⁺ cells, nonmyocytes isolated from control or NAC+Gly hearts were analyzed by flow cytometry. We reported before that gating on low and high CD45 expression discriminates between resident and infiltrating leukocytes (32). A significant reduction of infiltrating leukocytes (CD45^hi) was seen after NAC+Gly supplementation (Figure 2C, left panel). Figure 2C center and right panels depict representative histograms of cardiac leukocytes isolated from control and NAC+Gly-fed mice with a reduction in the infiltrating, CD45^hi population.

Extracellular Matrix Proteins and Modifying Enzymes are Not Affected by NAC or NAC+Gly Enrichment Diet

Since our previous data show a strong correlation between degrees of inflammation, fibrosis, and diastolic dysfunction, we analyzed whole heart homogenates using various markers of matrix proteins and enzymes that affect collagen and matrix deposition (by qPCR, western blot) or their activity (by zymography). No statistically significant changes were detected between control, NAC alone, or NAC+Gly supplemented mice. Supplementary Table 3 lists all examined markers.

Diet Supplementation Alters Cardiac Mitochondria

The decline of mitochondrial function with aging that manifested as reduced mitochondrial membrane potential and diminished expression of gene transcripts for Krebs cycle and oxidative phosphorylation pathway components has been reported in the heart and other tissues (33). We analyzed cardiac transcript levels for genes that are involved in fatty acid metabolism, Krebs cycle, and oxidative phosphorylation, confirming the downregulation in the aging mouse heart: ATP synthase subunit beta (Atp5b), NADP-dependent isocitrate dehydrogenase (Idh2), succinate dehydrogenase complex subunit A (Sdha), succinate dehydrogenase complex subunit C (Sdhc), malate dehydrogenase (Mdh2), and NADH:ubiquinone oxidoreductase subunit B5 (Ndufb5) (Figure 3A). For three of the six genes, Sdha, Mdh2, and Ndufb5, expression levels were doubled in the NAC+Gly supplemented mice compared to controls (Figure 3B). Protein level was also significantly increased for Sdha (Figure 3C). When the diet was supplemented only with NAC, the mRNA, and protein levels for Sdha, Mdh2, and Ndufb5 were unaffected (Figure 3B and C).

Figure 3.

NAC+Gly but not NAC alone upregulates expression of Krebs cycle and oxidative phosphorylation genes. (A) QPCR analysis of mitochondrial genes in young (3-month old) and aged (24-month old) mouse hearts confirming age-related changes. N = 5 for both age group. ATP synthase subunit beta (Atp5b), NADP-dependent isocitrate dehydrogenase (Idh2), succinate dehydrogenase complex subunit A (Sdha), succinate dehydrogenase complex subunit C (Sdhc), malate dehydrogenase (Mdh2), and NADH:ubiquinone oxidoreductase subunit B5 (Ndufb5) (B) Quantification of transcripts in the whole heart. N = 3, 6, 3 for control, NAC and NAC+Gly, respectively. * denotes p < .05. (C) Western blot analysis of Sdha protein (left panel) in whole heart lysates. The right panel shows Ponceau staining of the membrane used for immunoblotting. The lower panel shows densitometry analysis of western blot bands expressed in arbitrary units normalized to the level expressed in control hearts.

The function of isolated mitochondria from control and NAC+Gly mouse hearts was evaluated using carbohydrate, pyruvate, and malate, as substrates using a standard “mito stress” protocol on a Seahorse XF analyzer. To our surprise, oxygen consumption rates were not different between control and NAC+Gly in State 3 respiration using malate and pyruvate and there was no difference in mitochondrial “leak” between groups (Supplementary Figure 1).

We then evaluated saponin permeablized cardiac fibers from control and NAC+Gly hearts for mitochondrial function with the Oroboros oxygraph as shown in Figure 4A. Again, when pyruvate and malate were provided as substrates, there were no significant differences in State 3 respiration nor mitochondrial integrity between control and NAC+Gly hearts. In contrast, when the fatty acid, palmitoyl carnitine, was provided, the cardiac fibers from NAC+Gly mice demonstrated State 3 respiration three times greater than the control fibers. To control for subtle differences in loading or permeability, the ratio of O₂ utilization before and after ADP addition was examined; the NAC+Gly cardiomyocytes had much higher ratio than control (control 1.97 ± 0.03; vs NAC+Gly 4.08 ± 0.42; p < .01) when fatty acid was the provided substrate. There were no differences in the ratio of O₂ utilization rates between the groups (control 3.37 ± 0.33; NAC+Gly 4.18 ± 0.48; p = NS) when carbohydrate, pyruvate, and malate, was the substrate (Figure 4B). While the mitochondria from both groups functioned equivalently with regard to oxidative phosphorylation driven by carbohydrates, the major defect in fatty acid oxidative phosphorylation was found in the control aged animal and was substantially abrogated by treatment with NAC+Gly.

Figure 4.

Oroboros assessment of cardiac fiber oxygen consumption. (A) Oxygen consumption of saponin permeabilized cardiac fibers from a control diet (top row) and a NAC+Gly diet (lower row) fed mouse exposed to either carbohydrate (Malate +Pyruvate) (left column) or fatty acid (Palmitoyl Carnitine – right column) to assess mitochondrial function. Arrow points to line corresponding with O₂ consumption rate. By protocol, fibers are added, then substrate, then ADP which provoked a rapid rise in oxygen consumption. Both old control and NAC+Gly mitochondria used carbohydrate for State 3 respiration however, the myocytes isolated from NAC+Gly-fed mice were able to generate ATP from fatty acids at a much great rate than the old control. (B) The ratio of the State 3 rate to before ADP added was calculated for each substrate. The ADP/PM ratios were similar between groups, while the ADP/PCM (fatty acid) was much greater for NAC+Gly than Control diet old mice. (C) Specific genes involved in fatty acid metabolism were assessed by qPCR. No differences were statistically significant, but tended to be higher in the NAC+Gly then control or NAC hearts. Acad1 = Acyl-CoA dehydrogenase, C-4 to C-12 straight chain; Cpt1 = carnitine palmitoyltransferase I, ECsh1 = Enoyl Coenzyme A hydratase, short chain, 1, mitochondrial, Hadh = hydroxyacyl-coenzyme A dehydrogenase, mitochondrial, Pdk4 = pyruvate dehydrogenase kinase 4.

We then reassessed genes thought critical to fatty acid metabolism and found a trend to increase in genes including the doubling of the rate limiting mitochondrial fatty acid transporter, carnitine palmitoyltransferase 1 (CPT1) (Figure 4C). These data are consistent with greater ability of the NAC+Gly hearts to metabolize fatty acids.

NAC and NAC+Gly Treatment Alter Inflammation and Oxidative State but not Mitochondrial Function in Cardiac Fibroblasts

As we found no differences in cardiomyocyte ROS generation or mitochondrial leak between control and NAC+Gly supplemented mice and our previous reports have demonstrated that cardiac fibroblasts derived from the aging heart are a significant source of secreted pro-inflammatory cytokines (34), we analyzed cardiac fibroblasts in vitro. With quiescent cardiac fibroblasts derived from control, aged mouse hearts, treated with 5 mM NAC or 5 mM NAC+Gly for 24 h, we observed reduced expression of pro-inflammatory cytokines MCP-1 and IL-6 in when compared with nontreated control cells (Figure 5A).

Figure 5.

NAC or NAC+Gly treatment decreases superoxide generation and cytokine expression in quiescent fibroblasts derived from the control diet aged mouse heart. (A) 24 h NAC or NAC+Gly treatment in fibroblasts reduces expression of IL-6 and MCP-1. N = 5. (B) Fibroblasts were treated with NAC or NAC+Gly for 4 h then cells were incubated with Thiol Green Indicator and analyzed by flow cytometry. Events were gated on nucleated cells (Draq5 positive). Upper arrow points to the shift in fluorescence intensity in NAC treated cells, lower arrow denotes a cell population that has higher GSH content in NAC+Gly treated fibroblasts (left panel). N = 4 for each group treatment. Scale bar = 20 μm. Right panel shows quantification of changes of mean fluorescence intensity (MFI) of GSH in fibroblasts treated with NAC or NAC+Gly when compared with control. When both NAC and Gly was provided, the GSH concentration was greatest. Lower (NAC+Gly) and upper (NAC) arrows show the shift in GSH production in treated cells. (C) Immunofluorescent staining of superoxide anion in fibroblasts treated with NAC or NAC+Gly. Quiescent fibroblasts that were treated with 5 mM NAC or NAC+Gly for 2 h were incubated with MitoSox, a fluorogenic dye which, when oxidized by superoxide anion, produces red fluorescence. N = 3. (D) Flow cytometry analysis of fibroblasts treated with of NAC or NAC+Gly for 2 h and incubated with MitoSox. Left panel shows representative histograms and depict shift of fluorescence intensity. Right panel shows quantified mean fluorescent intensity (MFI) in cells treated with NAC or NAC+Gly and compared to its own control (untreated cells). N = 3. All experiments were performed on cardiac fibroblasts derived from 25 to 27-month old mice. * denotes p < .05.

Our recent report described the role of ROS from fibroblasts in the induction of age-dependent cardiac diastolic dysfunction (32). We therefore analyzed the GSH and ROS levels in cultured fibroblasts from control old hearts with and without treatment with 5 mM NAC or 5 mM NAC+Gly for 2 h. Cells were trypsinized and resuspended in GSH assay buffer containing a dye that becomes fluorescent upon reaction with GSH. We found that this brief incubation with NAC or NAC+Gly increased the number of cells that produce higher levels of fluorescence (Figure 5B) by flow cytometry analysis when compared with untreated cells (left panel). The calculated differences in mean fluorescent intensity (MFI) show increased generation of GSH in cells treated with NAC+Gly compared to NAC alone, suggesting that providing both precursors facilitates GSH production (Figure 5B, right panel).

Concurrently with increased GSH levels in fibroblasts treated with NAC or NAC+Gly, ROS production, specifically superoxide anion, was reduced in treated cells (Figure 5C). The flow cytometry analysis confirmed immunofluorescence data; Figure 5D, left panel shows a shift in fluorescence intensity when cells were treated with NAC or NAC+Gly and a 50% reduction in superoxide anion generation was seen (Figure 5D, right panel). The dramatic reduction of ROS levels in cells subjected to either treatment is consistent with the ability of NAC and GSH to decrease ROS in these cells (Figure 5B).

Although cardiac fibroblasts treated with NAC or NAC+Gly for 24 h demonstrated antioxidation effects and reduced inflammation, there were no alterations any mitochondrial gene levels (data not shown), in contrast to what we observed in the whole heart preparation (Figure 3B). This suggests that the observed mitochondrial effects more likely relates to cardiomyocyte mitochondria, but the observed reductions in pro-inflammatory state may be focused in the fibroblasts.

Discussion

The abnormalities seen in aging affect diastolic dysfunction by impacting both active and passive cardiac processes. Seven weeks dietary supplementation with NAC+Gly resulted in improved cardiac diastolic function in old mice without changes in aortic stiffness or systolic function compared to diet supplementation with NAC alone or control diet. These improvements in diastolic function were accompanied by decreased cardiac inflammation, reduced leukocyte infiltration, and cardiac fibrosis. Addition of NAC alone resulted in little change in cardiac function despite the fact that there was a substantial antioxidant and anti-inflammatory effect. In contrast, combining NAC+Gly also resulted in a striking reduction in ventricular stiffness and improvement in diastolic function associated with enhanced mitochondrial fatty acid utilization. The improvements in fatty acid metabolism were similar to those we reported earlier for liver and muscle mitochondria with NAC+Gly supplementation in old mice (20). By contrasting the results with NAC alone to NAC+Gly, we were able to suggest molecular mechanisms that were associated with the improvement in diastolic function.

NAC + Gly Dietary Supplementation Improves Diastolic Function in Old Mice

Diastolic function in older animals and people, as part of normal aging, can be prevented or mitigated. Interventions such as caloric restriction and exercise training have been shown to be associated with improved LV filling. In both older animals and older people who exercise or eat a calorically restricted diet, diastolic function is preserved (1). While as noted above, the dysfunction is multifactorial including impaired calcium handling, energetic dysregulation, and interstitial fibrosis, many of these can be driven by inflammation which is known increased with age (1). In contrast, the experience with antioxidant approaches has not been so positive as described below (35). The supplementation with NAC+Gly clearly improved diastolic function with the enhanced left ventricular filling parameters occurring at reduced filling pressures. The associations between reduced inflammation, reduced fibrosis, improved mitochondrial function, and improved diastolic function are only correlative but not mechanistic.

Reduction of ROS and Inflammation

Both NAC and NAC+Gly diet supplementation reduce expression of pro-inflammatory cytokines in vitro (Figure 5A) and in vivo (Figure 2A; Table 1), which may be directly related to downregulation of ROS levels seen in the fibroblasts (Figure 5C and D). Decreases in pro-inflammatory cytokines, like IL-6, MCP-1, TNF-α, and IFN-γ were noted, but only NAC+Gly increased the anti-inflammatory cytokine Il-10. We demonstrated the old heart expresses higher levels of pro-inflammatory cytokines (34) driven, in part, by ROS and protein damaged by oxidation which can activate multiple components of the inflammatory cascade (36). MCP-1, an inflammatory cytokine, particularly effective in attracting leukocytes into the heart (32), and IL-6 promoting formation of fibroblasts from the leukocytes and propagating fibrosis (34), all contributing to a stiffer ventricle and diastolic dysfunction. The local synthesis of MCP-1 and IL-6 is just one source of these cytokines. The NAC+Gly diet had minimal effects on cardiac MCP-1 protein despite a 50% reduction in mRNA, yet reduced cardiac diastolic dysfunction and leukocyte infiltration. Therefore, it is likely that the chemoattraction in the aging heart is more complicated than just a local MCP-1 effect. Pro-inflammatory cytokines and oxidative damage can also impair mitochondrial function, calcium handling, and worsen the energy dependent aspects of diastolic function (37,38). ROS activate transcription of various pro-inflammatory cytokines, which, in turn, further increase ROS (30,31). Therefore, the potential for both NAC and NAC+Gly to be effective was great.

However, NAC supplementation did not alter cardiac function. One clear difference between the NAC and NAC+Gly hearts was IL-10 protein concentration. In hearts from NAC+Gly supplemented mice. Increases in the anti-inflammatory IL-10 (Table 1) correlated with reduced leukocyte infiltration (Figure 2B). IL-10 plays an active role in preventing leukocyte infiltration in various organs and injury models (39–41). Indeed, IL-10 null mice have diastolic dysfunction, including leukocyte infiltrates, at an early age (42) and our data support a potentially critical role for IL-10 in the old heart.

Changes in Mitochondria

The decline of mitochondrial function with aging has been specifically linked to the reduction of several enzymes that participate in oxidative phosphorylation and the Krebs cycle (33) as well as a slower rate of mitochondrial electron transfer (43). Fatty acid oxidation is impaired in the old heart and glucose becomes the primary energy source available for contraction and relaxation (10–13). The relative inability of the permeablized cardiac fibers from the control diet old mice to generate ATP from palmitoyl carnitine was consistent with this impairment, but both control and NAC+Gly mitochondria functioned similarly when carbohydrate substrate was provided. NAC+Gly (but not NAC alone) dietary supplementation resulted in upregulation of expression of three genes known to decrease with age that participate in oxidative phosphorylation and the Krebs cycle (Mdh2, Ndufb5, and Sdha), which may be consistent with improved mitochondrial metabolism. Both NAC and NAC+Gly diets induced increases in CPT-1, a limiting transported of fatty acid into the mitochondria we implicated in the impaired use of fatty acids in the aging heart (13). These changes in gene expression were observed in whole heart preparations but not in fibroblast cultures. Therefore, the mitochondrial and metabolic effects of NAC+Gly are in the cardiomyocytes.

We found significant differences in the ability of the permeabilized cardiac fibers from NAC+Gly fed old mice to utilize fatty acids, where no diet related differences were seen in the ability to use pyruvate plus malate for ATP production. We hypothesize that both fuel sources are utilized in situ in the young heart. The ATP yield per molecule of oxygen favors carbohydrates as substrates so improved cardiac function in situ correlating with increased fatty acid metabolism was unexpected. In fact, heart failure treatment trials include approaches to suppress cardiac fatty acid metabolism (44). Enhanced fatty acid oxidation has been observed in older people supplemented with a high level of NAC+Gly (45). Our data suggest that utilization of both carbohydrates and fatty acids as fuel sources can be beneficial in the aging heart in the steady state.

The level of ROS is elevated in old cardiomyocyte mitochondria (10). We did not see this modified by the NAC+Gly diet in the isolated mitochondria nor cardiac fibers (20,45). We did see the effect of NAC and NAC+Gly on the ROS produced by the nonmyocytes concomitant with increased intracellular GSH concentrations, and the effects of the NAC+Gly on fibroblast mitochondrial function would be minimized in the mitochondrial assays we used other than the MitoSox assessment. Nevertheless, for the old heart to fully improve its function, both NAC and Gly may need to be supplemented.

Antioxidants in Aging

In general, antioxidant approaches to aging or age-related diseases have had disappointing outcomes (46). Strategies to prevent damage by quenching or preventing free radicals have had modest effects in models where other confounders were controlled (46). Adverse effects, such as preventing gains associated with exercise, underscore that a nonspecific approach may have unintended consequences (47). The importance of intercellular compartmentalization in addressing ROS dictates that antioxidants must be targeted to have desired effects (35,48). For example, mitochondrial targeting of the superoxide scavenger TEMPO or catalase overexpression increased mitochondrial metabolism and cardiac function in old mice (49,50). In our hands, providing the precursors to glutathione allowed targeting to be dictated by the organism, rather than trying to adapt molecules to go one compartment or another.

Glutathione may have other actions in addition to detoxifying ROS. S-glutathionylation is a post-translational modification of a number of proteins, and S-glutathionylation of the cardiac sarcoplasmic reticulum calcium pump (SERCA2a) increased the pumps’ activity (51). Other proteins critical to cardiac function may also undergo this modification (52). Direct glutathione effects may be beneficial in many other milieus (16) and providing both NAC and glycine may have allowed increased glutathione synthesis and S-glutathionylation, independent of the antioxidant effects.

Limitations

In our work, we used NAC alone in some of our studies, but because there was no improvement in cardiac function with NAC, our ultimate outcome, we did not pursue it in all of our efforts. We did not directly assess ROS or GSH levels in the hearts from mice fed NAC or NAC+Gly. We reported increased intracellular GSH in liver and muscle from old mice fed similar diets (20). We did not explore the effects of supplementing the diets with Gly alone, though glycine has been found to have a specific glycine receptor that has protective effects with ischemia and reperfusion (53). Glycine also has anti-inflammatory effects in some models (54) and we recognize this potential confounder. Differences in heart rate for the noninvasive studies can impact the observed function. The heart rate changes from the beginning to end of study are a potential problem, however functional improvements with NAC+Gly were apparent with the invasive studies and with noninvasive measures that are rate independent like the Tei index.

Conclusions

We found that dietary supplementation with NAC+Gly improved diastolic function in old mice associated with a lessened inflammatory state, reduced leukocyte infiltration, and increased expression of genes that could optimize mitochondrial function. With NAC alone, pro-inflammatory mediators were reduced, but the expression of genes regulating metabolism was increased to lesser extend and no improvement in cardiac function occurred. These results show that enhanced diastolic function is achievable in aging, even with a relatively brief intervention and future studies will attempt to clarify the metabolic, inflammatory, and functional interactions in the context of cellular signaling. The potential on this type of intervention to prevent age-related diastolic dysfunction in older people, reducing HFPEF and atrial fibrillation, remains uncertain.

Supplementary Material

Supplementary data is available at The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences online.

Funding

This work was supported by the National Institutes of Health grant (RO1-HL089792), the Huffington Center on Aging, the Medallion Foundation, the JB Katz Foundation, P. Studdert, C. Souki, and the Elaine and Marvy Finger Distinguished Endowed Chair for Translational Research in Metabolic Disorders.

Conflict of interest

K.A.C., R.V.S., A.G., A.R., G.M., C.P.H., M.L.E., D.J.H., S.L., E.R., A.A.G., and A.Z. report no conflicts of interest. G.E.T. reports consulting for Novartis in 2014–15 and is on the Editorial Board and an Associate Editor for the Journal.