Cystamine reduces vascular stiffness in Western diet-fed female mice

Abstract

Consumption of diets high in fat, sugar, and salt (Western diet, WD) is associated with accelerated arterial stiffening, a major independent risk factor for cardiovascular disease (CVD). Women with obesity are more prone to develop arterial stiffening leading to more frequent and severe CVD compared with men. As tissue transglutaminase (TG2) has been implicated in vascular stiffening, our goal herein was to determine the efficacy of cystamine, a nonspecific TG2 inhibitor, at reducing vascular stiffness in female mice chronically fed a WD. Three experimental groups of female mice were created. One was fed regular chow diet (CD) for 43 wk starting at 4 wk of age. The second was fed a WD for the same 43 wk, whereas a third cohort was fed WD, but also received cystamine (216 mg/kg/day) in the drinking water during the last 8 wk on the diet (WD + C). All vascular stiffness parameters assessed, including aortic pulse wave velocity and the incremental modulus of elasticity of isolated femoral and mesenteric arteries, were significantly increased in WD- versus CD-fed mice, and reduced in WD + C versus WD-fed mice. These changes coincided with respectively augmented and diminished vascular wall collagen and F-actin content, with no associated effect in blood pressure. In cultured human vascular smooth muscle cells, cystamine reduced TG2 activity, F-actin:G-actin ratio, collagen compaction capacity, and cellular stiffness. We conclude that cystamine treatment represents an effective approach to reduce vascular stiffness in female mice in the setting of WD consumption, likely because of its TG2 inhibitory capacity.

NEW & NOTEWORTHY This study evaluates the novel role of transglutaminase 2 (TG2) inhibition to directly treat vascular stiffness. Our data demonstrate that cystamine, a nonspecific TG2 inhibitor, improves vascular stiffness induced by a diet rich in fat, fructose, and salt. This research suggests that TG2 inhibition might bear therapeutic potential to reduce the disproportionate burden of cardiovascular disease in females in conditions of chronic overnutrition.

INTRODUCTION

Although stiffening of the vasculature occurs physiologically during aging, it becomes accelerated in pathologic conditions such as overnutrition with a Western diet (WD) high in fat, sugar, and salt, and obesity, hypertension, and type 2 diabetes mellitus (1–3). Increased vascular stiffness, as assessed noninvasively via aortic pulse wave velocity (PWV), has been linked to the progression of atherosclerosis and risk for cardiovascular disease (CVD) (4). In the clinical setting, vascular stiffening is considered a biomarker with independent predictive value for CVD (5). Conversely, it has been shown that reductions in vascular stiffness are associated with improved long-term survival, independently of other risk factors, such as hypertension (6).

The pathophysiology underlying vascular stiffening remains to be fully elucidated. However, it has been shown to involve multiple pathways that result in functional and morphological abnormalities of the endothelium, vascular smooth muscle, and extracellular matrix (7, 8). At the endothelial level, reduced nitric oxide (NO) bioavailability has been consistently linked to increased occurrences of CVD in conjunction with vascular stiffening (4, 5). In vascular smooth muscle, cytoskeletal remodeling has also been shown to contribute to vascular stiffening (9). As for the extracellular matrix, increased collagen deposition and crosslinking have been long recognized as critical players in vascular stiffening (10).

Notably, tissue transglutaminase (TG2) activity is associated with all these vascular wall abnormalities. TG2 is an enzyme ubiquitously expressed in the vasculature that promotes the synthesis and crosslinking of extracellular matrix proteins such as collagen, as well as cytoskeletal remodeling via actin polymerization pathways (11, 12). TG2 is expressed by all cells in the vascular wall, including fibroblasts, smooth muscle, and endothelial cells (13), and when secreted extracellularly, it promotes extracellular matrix protein crosslinking in a calcium-dependent manner (14). Given the prevalent role that vascular stiffening plays in the pathogenesis of CVD, the development of specific therapeutic strategies to reduce, stop, or reverse it are needed. Accordingly, the main goal of this study was to determine the efficacy of cystamine, a nonspecific inhibitor of TG2, at reducing vascular stiffening in WD-fed female mice.

Cystamine is an organic disulfide derived from cystine (15). In vivo, cystamine is partially reduced to cysteamine, which also has TG2 inhibitory effects (15, 16). Cysteamine is currently used clinically for the treatment of cystinosis, a rare autosomal-recessive lysosomal storage disease that results in severe accumulation of cystine in lysosomes (17). Cysteamine also has therapeutic potential for the management of neurodegenerative disorders, such as Huntington’s and Parkinson’s disease, in which there is enhanced TG2 crosslinking activity (18, 19). Furthermore, cysteamine has been studied in conditions of insulin resistance such as nonalcoholic fatty liver disease (20, 21). In addition to this medication, novel compounds with intracellular-specific TG2 inhibitory activity are emerging that might have enhanced pharmacologic potential for CVD (22, 23).

Cumulative research clearly shows that overnutrition resulting from chronic consumption of WD is one of the most prevalent instigators and accelerators of vascular stiffening (24–27). Data also indicate that the resulting obesity and insulin resistance blunt the protective effects of estrogen on the vasculature, leading to more accelerated stiffening in premenopausal females compared with males (28, 29). Therefore, herein we sought to determine the effects of cystamine on vascular stiffening in a female mouse model of chronic exposure to WD. Specifically, we tested the hypothesis that nonspecific TG2 inhibition with cystamine reduces vascular stiffness in female mice fed a WD. As a corollary, we also determined the effects of cystamine on the capacity of human cultured smooth muscle cells to promote collagen remodeling, form actin stress fibers, and soften the cellular cortex.

METHODS

Ethics and Approvals

All animal experiments were approved by the Institutional Animal Care and Use Committee at the University of Missouri and the Subcommittee for Animal Safety at the Harry S. Truman Veterans’ Memorial Hospital. The University of Missouri and the Harry S. Truman Veterans’ Memorial Hospital research unit are accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International. All experiments were randomized and performed in a blinded manner when possible by the investigators. The data reported in this project can be made available upon reasonable request to the corresponding authors.

Animals and Experimental Design

Four-week-old C57BL/6 female mice maintained under the same general conditions were randomly assigned to either ad libitum access to regular chow diet (CD, n = 10) or a high-salt Western diet (WD, n = 20) for 43 wk. The CD contained 3.31 kcal/g of food, 13% kcal from fat, 58% kcal from carbohydrate, and 29% kcal from protein (Laboratory Rodent Diet 5001* Lab Diet) whereas the WD contained 4.65 kcal/g of food, 46% kcal from fat, high-fructose corn syrup 17.5%, sucrose 17.5%, protein 17.6%, and salt 1.6% (TestDiet modified 58Y1; 5APC). In this study, WD was supplemented with additional salt (1.6%), with the aim of enhancing an increased vascular stiffness phenotype. During the last 8 wk on the WD, ten WD-fed mice received the nonspecific TG2-inhibitor, cystamine, in the drinking water at a rate of 40 mg/kg/day (WD + C, n = 10), whereas the other ten received vehicle (WD, n = 10). In vivo cystamine administration in mice has been shown to produce tissular inhibition of transglutaminase activity at doses ranging from 6 to 225 mg/kg/day (30–32). We provided cystamine (0.9 g/L) in drinking water that resulted in a dose of 216 mg/kg/day based on a measured daily water consumption of 7.2 mL/day.

Aortic PWV

In vivo arterial stiffness was assessed via measurements of aortic PWV using Doppler ultrasound (Indus Mouse Doppler System, Webster, TX), as previously described (33). Briefly, mice under 1.5% isoflurane anesthesia were placed in a supine position on a heated platform (42°C) and the transit time between the arrival of a pulse wave detected at the aortic arch with respect to the peak of the ECG-R-wave was determined. A similar transit time was determined for a location in the abdominal aorta. The difference in arrival times of the Doppler pulse waves and the distance between the two locations was assessed. PWV was calculated as the distance between the two locations divided by the time difference and expressed in m/s.

Aortic Stiffness via Atomic Force Microscopy

To evaluate aortic stiffness ex vivo, a 2-mm segment of the thoracic aorta was obtained from each mouse at euthanasia and opened longitudinally. An en face preparation of the segment was prepared by fastening the adventitial surface to a glass coverslip using Cell-Tak. This allowed atomic force microscopy (AFM) probes on a MFP-3D AFM (Asylum Research, Inc., Goleta, CA) mounted on an Olympus IX81 microscope (Olympus, Inc.) to indent the aortic endothelial surface. Stiffness was assessed following a nanoindentation protocol and the subsequent calculation of the Young’s modulus of elasticity from the force curves was generated, as previously described (34–36) and shown in the schematic representation (Fig. 1).

Figure 1.

Cystamine reduces aortic stiffness in Western diet (WD)-fed female mice. Female mice were fed a regular chow diet (CD) or a Western diet (WD) for 43 wk, with half of the WD-fed mice receiving cystamine (216 mg/kg/day) in the drinking water for the last 8 wk of feeding (WD + C). A: aortic pulse wave velocity (PWV), n = 9 or 10 rings (mice)/group. B: schematic representation of atomic force microscopy generation of ex vivo aortic en face force curves [endothelial cell (EC) left] and measured stiffness (right) by treatment group, n = 6–9 rings (mice)/group. C: representative images of adipocytes (scale bar = 50 µm) including zoomed in regions, and plots showing the quantification of perigonadal adipose tissue weight, adipocyte size, and Mac-2 (macrophage)-positive stained area, n = 9 or 10 samples (mice)/group. D: heart, kidneys, and liver weights n = 10 organs (mice)/group. E: body weight, n = 9–10 mice/group. F: blood glucose concentrations at baseline (time = 0), and at 15, 30, 45, 60, and 120 min after an intraperitoneal injection of 1.0 g/kg of dextrose; insets show the area under the curve (AUC) for each treatment group, n = 9 or 10 mice/group. G: heart rate and mean arterial pressure (MAP), n = 10 mice/group. Data are expressed as means ± SE. *P ≤ 0.05 vs. CD, #P ≤ 0.05 vs. WD as determined by two-way ANOVA with repeated measurements and one-way ANOVA followed by Tukey’s range test where appropriate.

Adipocyte Size and Macrophage Infiltration Assessment

Histological assessment of adipocyte size and immune cell (macrophage) infiltration was performed in formalin-fixed visceral (perigonadal) adipose tissue samples processed through paraffin embedding, sectioning at 5 μm, and stained with an antibody versus Mac-2 (1:1,000; Cedarlane, Cat. No. CL8942AP), as previously described (29). Images were acquired with an Olympus IX51 microscope, and adipocyte size and macrophage-positive stained area quantification were done using unbiased MATLAB scripts.

Glucose Tolerance Test

Assessment of glucose tolerance was performed during the week before euthanasia as previously described (35). Briefly, glucose levels in blood drawn from the tail were measured after 5-h of fasting using a handheld glucometer (AlphaTrak 2, Abbott Laboratories, Abbott Park, IL). Blood samples were obtained at baseline and 15, 30, 45, 60, and 120 min following an intraperitoneal injection of 1.0 g/kg dextrose. The glucose concentration area under the curve was calculated using the trapezoidal rule.

Blood Pressure

A CODA tail-cuff system (CODA-HT2; Kent Scientific, Torrington, CT) was used to assess blood pressure and heart rate as previously described (35). Mice were acclimatized to the system for 5 days before the final recording of blood pressure the week before euthanasia.

Ex Vivo Vascular Reactivity

Vasomotor responses were assessed in isolated aortic rings using wire myography, and in isolated mesenteric and femoral arteries using pressure myography, as previously described (24, 35, 37).

In brief, aortic rings (2 mm in length) were collected after euthanasia and cleaned of perivascular adipose tissue in cold (4°C) physiological saline solution (PSS; pH of 7.4). Individual rings were mounted on wire myography chambers (620 M, Danish Myo Technology, Hinnerup, Denmark) containing warm (37°C) PSS gassed with 95% O₂-5% CO₂. The rings were constricted with 80 mM KCl to ensure viability. Viable rings were then preconstricted with the thromboxane mimetic, U-46619 (Cayman Chemical, Cat. No. 16450, 20 nM), and subsequently exposed to cumulative increasing concentrations of acetylcholine (ACh) (from 10⁻⁹ to 10⁻⁵ M in half-log increments) followed by sodium nitroprusside (SNP) (from 10⁻⁹ to 10⁻⁴ M in full-log increments), while maintaining the preconstriction.

Femoral and mesenteric arteries were obtained from each mouse at euthanasia and cannulated in pressure myography chambers (Living Systems Instrumentation, Burlington, VT) containing 3-(N-morpholino) propanesulfonic acid (MOPS)-buffered PSS at 37°C. Pressurized vessels (70 mmHg) without perfusion were exposed abluminally to cumulative increasing concentrations of phenylephrine (from 10⁻⁸ to 10⁻⁴ M in half-log increments) to assess vasoconstriction capacity, and to increasing concentrations of ACh (from 10⁻⁹ to 10⁻⁵ M in half-log increments) followed by SNP (from 10⁻⁸ to 10⁻⁴ M in half-log increments), while maintaining preconstriction with phenylephrine (10 µM). At the end of each experiment, arteries were exposed to Ca²⁺-free PSS containing 2 mM EGTA and 0.1 mM adenosine to obtain maximal passive diameter.

Ex Vivo Mechanical Properties of Mesenteric and Femoral Arteries

After vasomotor responses were accessed, once arteries had reached maximal passive diameter under Ca²⁺-free conditions, intraluminal pressure was changed from 5 to 120 mmHg to obtain passive pressure-diameter curves, as previously described (33, 38). Each intraluminal pressure was maintained for 2 min to achieve a diameter plateau. Wall thickness and pressure-diameter curves were used to calculate the circumferential strain and stress, incremental modulus of elasticity (E_inc), cross-sectional compliance (CSC), and calculated incremental PWV (cPWV_inc) (39). At higher pressures, the main component of stiffness of the vascular wall comes from collagen, quantified by the high modulus of elasticity (E_high), which was calculated by using a linear fitting of the high-pressure region (80–120 mmHg) of the strain-stress curves, as previously reported (33, 34, 38, 40).

Confocal-Multiphoton Microscopy Imaging of Isolated Vessels

The amount of nuclear material, F-actin, and collagen was measured in femoral and mesenteric arteries that were isolated and fixed in 4% paraformaldehyde immediately after euthanasia, as previously described (34, 37). In brief, fixed, permeabilized, blocked, and cannulated vessels were incubated in 300 nM 4′,6-diamidino-2-phenylindole (DAPI, Sigma, Cat. No. D9542) and 200 nM Alexa Fluor 546 phalloidin (Thermo Fisher, Cat. No. A22283) for 1 h at room temperature and subsequently washed with phosphate-buffered saline. Images of nuclei, F-actin, and collagen were obtained using a Leica SP5 confocal-multiphoton microscope (Leica Microsystems, Inc., Morrisville, NC), collagen was imaged using second harmonic generation. Imaris software (Bitplane, Inc., Concord, MA) was used to render three-dimensional reconstruction images. An unbiased MatLab script was used to quantify the volume of the molecules of interest. Volumetric data were normalized by the number of smooth muscle cells (nuclei), when measurements corresponded to the vascular medial layer.

Cell Culture Experimental Design

Human coronary smooth muscle cells were obtained from Thermo Fisher Scientific (Cat. No. C-017-5C, Lot No. 1130140) and cultured as previously described (34). Cells used for immunostaining or for the analysis of collagen compaction were seeded on µ-angiogenesis 15-well ibidi plates (Ibidi, Cat. No. 81506), whereas cells used for Western blot analysis or AFM analyses were seeded on 60-mm cell culture dishes. Cells in half of the wells or dishes were exposed to 1 mM cystamine (Sigma, Cat. No. C121509) or vehicle control before immunostaining, Western blot analysis, AFM, or collagen compaction assessments were performed.

Confocal Microscopy Imaging of Vascular Smooth Muscle Cells in Culture

Confocal microscopy images of human coronary smooth muscle cells were obtained from cells plated in µ-angiogenesis 15-well ibidi plates. TG2 activity was assessed in cells incubated for 24 h at 37°C with 20 µM Alexa Fluor 488 cadaverine (Thermo Fisher Scientific, Cat. No. A30676) in the absence or presence of cystamine (1 mM). At the end of each experiment, cells were fixed in 4% paraformaldehyde, at room temperature for 30 min and permeabilized with 0.5% Triton X-100 for 15 min. This was followed by incubation for 1 h in 1.5 nM DAPI and 200 nM Alexa Fluor 568 phalloidin (Thermo Fisher Scientific, Cat. No. A12380) to stain nuclei and F-actin cellular content, respectively. In an additional set of experiments, cells were stained with 200 nM Alexa Fluor 633 phalloidin (Thermo Fisher Scientific, Cat. No. A22284) and 300 nM Alexa 488 DNAse1 (Thermo Fisher Scientific, Cat. No. D12371) to stain F- and G-actin cellular content, respectively. Images were acquired with a Leica SPE confocal microscope (Leica Microsystems, Inc., Morrisville, NC). Intensity data were quantified using Imaris software and normalized by the total number of cells.

Western Blot Analysis

Specific protein content was assessed in human coronary smooth muscle cell lysates prepared in RIPA buffer (Invitrogen, Cat. No. 89900) supplemented with protease inhibitors (Invitrogen, Cat. No. 1861278) and phosphatase inhibitors (Invitrogen, Cat. No. 1862495). Proteins within samples were separated in Criterion Tris-Glycine eXtended-PAGE precast gels (Bio-Rad) and transferred onto polyvinylidene difluoride membranes. The following antibodies were used to probe for the presence of specific proteins on the membranes, 1:500 phosphorylated-cofilin (Sigma, Cat. No. C8992), 1:750 total-cofilin (Sigma, Cat. No. C8736). Secondary antibody 1:5,000 (Bio-Rad, Cat. No. 1705046) was used to visualize protein bands by chemiluminescence, and densitometry of individual protein bands was quantified using a Bio-Rad ChemiDoc XRS⁺ System (Bio-Rad, Hercules, CA).

Vascular Smooth Muscle Cell Cortical Stiffness

The cortical stiffness of human coronary smooth muscle cells in culture was assessed via AFM as previously described (34). Briefly, AFM cantilever probes were approached onto the cell surface. Stiffness of the cellular cortex was then assessed following a nanoindentation protocol and subsequent calculation of the Young’s modulus of elasticity from the force curves generated, using the Hertz model (34–36).

Vascular Smooth Muscle Cell Collagen Matrix Remodeling

Human coronary smooth muscle cells were seeded onto collagen-coated (1 g/mL, MP Biomedicals, Cat. No. 0216008450) µ-angiogenesis 15-well ibidi plates. Before seeding the cells, five glass beads (100 µm, OMNI International, Cat. No. 19-641) were placed within the collagen matrix in each well, and an image was taken using a Nikon Eclipse TS100 microscope with a ×4 objective and a camera (Amscope MA-1000). After seeding, cells were cultured in the absence or presence of 1 mM cystamine for 24 h, and subsequently the location of the beads was imaged again. Bead displacement was quantified by measuring the distance traveled by each glass bead from its initial position. Briefly, before- and after-incubation images were overlapped in Adobe Illustrator. The after-incubation image was set at the top and its opacity changed to 50% to allow for visualization of both images simultaneously. As the beads were able to travel in any direction, we centered the images using the mark displayed on the bottom of the well. Once the two marks overlapped, we exported the composed image in tiff format to ImageJ. The distance traveled by each of the glass beads was quantified and expressed as mean bead displacement in µm. At the end of each experiment, cells were fixed and stained with DAPI and Alexa Fluor 568 phalloidin as described earlier. Cell numbers were used to assess viability.

Statistical Analyses

Statistical analyses were performed using GraphPad Prism, version 9. The Shapiro–Wilk test was used to determine if data were normally distributed. We used a logarithmic transformation to normalize data that did not have a normal distribution. When transformation did not normalize the data, analyses were performed using the Mann–Whitney U nonparametric test. Two-tailed t tests were used in experiments with only two experimental groups. One-way ANOVA was used in experiments with three groups, and two-way ANOVA with repeated measures was used to analyze data from the time-dependent (i.e., body weight and blood glucose) and vessel myography experiments. Tukey’s multiple range test followed ANOVA for preplanned multiple comparisons, when appropriate. Data are presented as means ± SE. Values of P ≤ 0.05 were considered significant.

RESULTS

Cystamine Reduces Aortic Stiffness in WD-Fed Female Mice

To determine the effect of cystamine on aortic stiffness, female mice fed a WD for 43 wk were treated with either vehicle control (distilled water) or cystamine (216 mg/kg/day) in the drinking water for the last 8 wk of feeding. A regular CD-fed cohort was kept as untreated control. WD increased arterial stiffness as determined by the faster aortic PWV seen in WD- versus CD-fed mice (Fig. 1A), whereas cystamine reduced it as indicated by the slower PWV observed in WD + C versus WD mice (Fig. 1A). These results mirrored our measurements of aortic stiffness, as assessed ex vivo using AFM. Our data showed that WD-fed mice had stiffer aortae compared with those from mice fed CD, whereas WD + C treated mice had softer aortae than those from the WD untreated group (Fig. 1B). To determine whether any effects of cystamine on vascular stiffness occurred in conjunction with changes in body weight, organ weight, glucose tolerance, or blood pressure, we measured these variables in all treatment groups. Results showed that WD-feeding alone did not increase visceral adipose tissue weight or adipocyte size (Fig. 1C) compared with CD-feeding. In comparison, visceral adipose tissue weight and adipocyte size were significantly increased in WD + C versus WD-fed mice (Fig. 1C). Macrophage infiltration of visceral adipose tissue was increased in WD versus CD-fed mice and reduced in WD + C versus WD-fed mice (Fig. 1C). Heart, kidneys, and liver weights were significantly increased in WD versus CD mice (Fig. 1D). Cystamine did not significantly reduce these increases in organ weight (Fig. 1D), but caused an overall increase in body weight compared with WD-fed mice (Fig. 1E). Notably, WD did not increase body weight compared with CD (Fig. 1E). Neither WD-feeding nor cystamine had any effect on glucose tolerance, heart rate, or blood pressure (Fig. 1, F and G).

Cystamine Does Not Improve Vasomotor Function in Female Mice Fed a WD

To determine the effect of in vivo treatment with cystamine on vascular function, isolated aortic rings as well as femoral and mesenteric arteries from all groups were preconstricted with either 20 nM U46619 (aortic rings) or 10 µM phenylephrine (femoral and mesenteric arteries). Preconstricted arteries were then exposed to increasing concentrations of the endothelium-dependent vasodilator, ACh, followed by exposure to the endothelium-independent vasodilator, SNP. WD-feeding affected only aortic ring vasomotor responses by significantly decreasing the vasodilatory effects of both ACh and SNP (Fig. 2A). WD did not cause any significant change in the vasomotor responses of isolated femoral or mesenteric arteries, including those of phenylephrine, which were similar among all the cohorts (data not shown). Notably, cystamine did not significantly affect the vasomotor responses of any isolated vessel (Fig. 2, A–C); however, aortic rings from WD + C had a partial improvement in SNP-induced dilation compared with WD mice. This caused SNP dilations to no longer be statistically different between the WD + C and CD rings (Fig. 2A).

Figure 2.

Cystamine does not improve vasomotor function in female mice fed a Western diet (WD). Aortic rings, as well as femoral and mesenteric arteries were isolated from female mice fed a regular chow diet (CD) or a Western diet (WD) for 43 wk, with half of the WD-fed mice receiving cystamine (216 mg/kg/day) in the drinking water for the last 8 wk of feeding (WD + C). Isolated vessels were preconstricted with 20 nM U46619 (aortic rings) or with 10 µM phenylephrine (femoral and mesenteric arteries) followed by cumulative exposure to increasing concentrations of the vasodilators acetylcholine (ACh) and sodium nitroprusside (SNP) in tandem. A: percent relaxation responses of aortic rings from the CD, WD, and WD + C groups, including the ACh-to-SNP area under the curve (AUC) relaxation ratios, n = 8 rings (mice)/group. B: percent vasodilatory responses of femoral arteries from the CD, WD, and WD + C groups, including the ACh-to-SNP AUC ratios, n = 5–8 arteries (mice)/group. C: percent vasodilatory responses of mesenteric arteries from the CD, WD, and WD + C groups, including the ACh-to-SNP AUC ratios, n = 5 or 6 arteries (mice)/group. Data are expressed as means ± SE. *P ≤ 0.05 vs. CD as determined by two-way ANOVA with repeated measurements and one-way ANOVA followed by Tukey’s range test where appropriate.

Cystamine Reduces Stiffness in Femoral and Mesenteric Arteries in WD-Fed Female Mice

To determine the effect of cystamine on vascular mechanics, pressure-diameter curves were generated in isolated femoral and mesenteric arteries under passive conditions. Structural and mechanical parameters generated from the pressure-diameter curves indicate that WD feeding increased both the femoral and mesenteric artery’s E_inc and cPWV_inc when compared with CD feeding (Fig. 3, A and B). These stiffening effects of WD feeding were diminished by cystamine (WD + C), as determined by its effect of increasing percent initial diameter, shifting the strain-stress curve to the right, and reducing both the E_inc and cPWV_inc. No change in wall dimensions occurred in response to WD feeding or cystamine treatment, as shown by the lack of differences in wall-to-lumen ratios between the groups (Fig. 3, A and B).

Figure 3.

Cystamine softens femoral and mesenteric arteries in Western diet (WD)-fed female mice. Femoral and mesenteric arteries were isolated from female mice fed a regular chow diet (CD) or a Western diet (WD) for 43 wk, with half of the WD-fed mice receiving cystamine (216 mg/kg/day) in the drinking water for the last 8 wk of feeding (WD + C). Pressure-diameter curves were obtained from vessels kept under passive conditions that were mounted onto pressure myography systems. Data from the curves were used to calculate structural and mechanical properties of the vascular wall as pressure changed, which included the following: percent change from initial diameter, wall-to-lumen ratio, circumferential strain-stress, incremental modulus of elasticity (E_inc), high modulus of elasticity (E_high, inset), cross-sectional compliance (CSC), and calculated incremental pulse wave velocity (cPWV_inc). A: structural and mechanical properties of femoral arteries isolated from the CD, WD, and WD + C groups, n = 5–7 arteries (mice)/group. B: structural and mechanical properties of mesenteric arteries isolated from the CD, WD, and WD + C groups, n = 8 arteries (mice)/group. Data are expressed as means ± SE. *P ≤ 0.05 vs. CD, #P ≤ 0.05 vs. WD, as determined by two-way ANOVA with repeated measurements and one-way ANOVA followed by Tukey’s range test where appropriate.

Cystamine Reduces Smooth Muscle F-Actin and Collagen Content in the Femoral and Mesenteric Arteries of Female Mice Fed a WD

To determine the effect of cystamine on vascular smooth muscle F-actin and collagen content, which are two major structural factors contributing to vascular stiffness that are modulated by TG2, femoral and mesenteric arteries from mice belonging to the CD, WD, and WD + C groups were isolated and imaged using a confocal-multiphoton microscope. Images showed that both femoral and mesenteric arteries from WD-fed mice had increased F-actin and collagen content, when compared with CD-fed mice (Fig. 4, A and B). Notably, these increments in F-actin and collagen content were diminished in WD-fed mice treated with cystamine (WD + C, Fig. 4, A and B). No changes in the number of vascular smooth muscle cells within the arterial walls were observed (Fig. 4, A and B).

Figure 4.

Cystamine reduces smooth muscle F-actin and collagen content in the femoral and mesenteric arteries of female mice fed a Western diet (WD). Femoral and mesenteric arteries were isolated from female mice fed a regular chow diet (CD) or a Western diet (WD) for 43 wk, with half of the WD-fed mice receiving cystamine (216 mg/kg/day) in the drinking water for the last 8 wk of feeding (WD + C). The isolated arteries were mounted onto pressure myography systems, pressurized to 70 mmHg while under passive conditions and imaged with a confocal-multiphoton microscope. A: representative images of femoral arteries (scale bar = 30 µm) and quantification of the volume (voxels) of nuclei (4′,6-diamidino-2-phenylindole, DAPI, blue), F-actin (phalloidin, yellow), and collagen (second harmonic generation, SHG, green), n = 5 arteries (mice)/group. B: representative images of mesenteric arteries (scale bar = 30 µm) and quantification of the volume of nuclei, F-actin, and collagen, n = 6 or 7 arteries (mice)/group. SHG images were contrast enhanced equally in all groups to improve visualization. All analyses were performed using the raw images. Data are expressed as means ± SE. *P ≤ 0.05 vs. CD, #P ≤ 0.05 vs. WD, as determined by one-way ANOVA followed by Tukey’s range test where appropriate.

Cystamine Reduces TG2 Activity, F-Actin Content, Cofilin Phosphorylation, and the Ability of Vascular Smooth Muscle Cells to Compact Collagen

To determine the direct effects of cystamine on vascular smooth muscle cells, cultured human coronary smooth muscle cells were exposed to 1 mM cystamine or vehicle control for 24 h. Confocal images showed that cells cultured in the presence of cystamine had reduced cadaverine incorporation and lesser F-actin content compared with controls (Fig. 5A). Further image analyses in conjunction with Western blot analysis indicated that cystamine exposure also reduced F-actin:G-actin ratios and cofilin phosphorylation (Fig. 5, B and C). Assessment of cellular stiffness with AFM indicated that exposure to cystamine softened the cellular cortex (Fig. 5D), whereas culturing cells on collagen embedded with glass microbeads showed that cystamine reduced the capacity of the cells to remodel the collagen substrate and displace the beads within the matrix (Fig. 5E). Notably, although cystamine reduced the capacity of cells to displace the beads and reduce their F-actin content, it did not change the number of cells in the plates (Fig. 5F).

Figure 5.

Cystamine reduces vascular smooth muscle cell’s TG2 activity, F-actin content, cofilin phosphorylation, and ability to compact collagen. Human coronary smooth muscle cells in culture were exposed to 1 mM cystamine or vehicle control for 24 h. A: representative confocal images (scale bar = 50 µm) of tissue transglutaminase (TG2) activity and F-actin content, as assessed by cadaverine incorporation (green) and phalloidin staining (yellow), respectively. The plots represent quantification of the cadaverin and phalloidin fluorescence intensities normalized by the number of nuclei, n = 6 samples (culture dishes)/group. B: representative confocal images (scale bar = 50 µm) of F-actin (phalloidin, red), and G-actin (DNAse1, green), and the quantification of F-actin:G-actin ratio, n = 15 samples (culture dishes)/group. C: representative Western blot images of phosphorylated (P)-cofilin, total (T)-cofilin, and β-actin, and the quantification of P-cofilin:T-cofilin ratios, n = 6 samples (culture dishes)/group. D: schematic drawing of atomic force microscopy (AFM)-generated representative curves and quantification of cell cortical stiffness, n = 8 or 9 samples (culture dishes)/group. E: representative images of collagen compaction as assessed by quantification of glass beads displacement and representative confocal images of human coronary smooth muscle cells (scale bar = 50 µm) and quantification of intensity after cell staining for F-actin (yellow), n = 5 samples (culture dishes)/group. Data are expressed as means ± SE. *P ≤ 0.05 vs. control, as determined by two-tailed t test.

DISCUSSION

The primary finding of this study is that administration of cystamine in the drinking water for 8 wk is capable of reducing arterial stiffness in female mice chronically fed a WD. This is evident by the reduction in aortic PWV observed in cystamine-treated mice (WD + C), when compared with WD-fed mice. Notably, the reduction in aortic PWV, which is the in vivo gold standard for assessing arterial stiffness (41), was not associated with changes in blood pressure, a major physiological parameter affecting PWV (42). This suggests that the reduction in vascular stiffness was not caused by variations in aortic-wall circumferential stress or strain associated with reductions in intravascular pressure. Rather, our results suggest that changes in wall structure were responsible for the reduced vascular stiffness observed in cystamine-treated mice. This is supported by our finding that femoral and mesenteric arteries isolated from WD + C had less F-actin and collagen content than those from WD control mice, which occurred in concert with a rightward shift of the strain-stress curve, reduced E_inc, and slower cPWV_inc. Remarkably, administration of cystamine reduced the increased vascular stiffness caused by WD feeding to levels similar to those of CD-fed mice. This indicates that cystamine is able to reduce vascular stiffness in large and small arteries of female mice fed a WD, and that likely does it via modification of cytoskeletal and extracellular structures within the vascular wall.

We and others have consistently shown that chronic consumption of WDs results in accelerated vascular stiffening (24, 43–45). The mechanisms responsible for this phenomenon remain to be fully uncovered, however there is mounting evidence of a critical role played by endothelial mineralocorticoid receptor activation (25, 26), increased epithelial sodium channel activity (24), oxidative stress, and decreased bioavailable NO, all of which occur in the setting of WD feeding (25, 46). Notably, TG2 is also activated by all of these conditions. Furthermore, additional changes related to vascular stiffening have been demonstrated, including collagen deposition, cytoskeletal abnormalities, and changes in calcium handling in vascular smooth muscle cells (47).

In line with our findings that cystamine reduced arterial stiffness without affecting blood pressure, we and others have previously shown that consumption of a WD associates with enhanced vascular stiffness and endothelial dysfunction before blood pressure increases or overt hypertension (27, 48, 49). In the present study, WD feeding resulted in significant increases in heart, kidneys, and liver weights with no overall increase in body weight. We attribute the overall lack of weight gain to the excess salt content in the WD used, an effect that has been previously reported with dietary salt loading and is mediated by enhanced corticosteroid signaling induced by salt, resulting in ketogenesis, decreased gluconeogenesis, increased fatty acid oxidation, and ultimately loss of muscle mass (50). On the other hand, WD + C-treated mice had greater body weight than WD control mice, which can reflect the influence of TG2 inhibition in promoting adipogenesis, as it has been demonstrated that TG2 transglutaminase activity inhibits adipocyte differentiation by regulating key pathways involving pref-1, β-catenin, Akt, and Rho kinase among others (51). Accordingly, our data support the notion that cystamine induced visceral adipose growth without inflammation, as we found increased visceral adipose tissue weight, and adipocyte size with reduced macrophage infiltration in the WD + C versus WD cohorts. It is notable that cystamine was able to reduce vascular stiffness without reducing body weight or improving glucose tolerance, which indicates that the reduction in vascular stiffness cannot be attributed to cystamine-dependent improvements in metabolic function.

The observation that cystamine did not improve endothelium-dependent dilation, further suggests that cystamine softened the vasculature via its effects on the structural characteristics of arterial walls rather than by enhancing endothelial function. Arguably the partial increased responsiveness to SNP observed in aortic rings isolated from WD + C mice is potentially due to softening of the vascular wall, as relaxation of the vascular smooth muscle contractile machinery would more easily dilate softer vessels after exposure to the same level of NO. Previous studies also indicate that as little as 8–16 wk of WD feeding increases vascular stiffness in female mice (28, 48). Therefore, it is likely that, in our study, after 35 wk of WD consumption (i.e., when cystamine administration was initiated), mice already had increased vascular stiffness. This would suggest that cystamine is not only able to stop the vascular stiffening process but can reverse it. However, PWV was not measured before the administration of cystamine, and cystamine’s ability to soften already stiffened arteries needs to be further explored.

Our observation that aortic ACh-induced relaxation responses were reduced in WD-fed mice in conjunction with their SNP responses further suggests that compensatory mechanisms may occur in the setting of chronic WD consumption that ameliorate endothelial dysfunction. These mechanisms may include switching vasodilatory pathways from NO to other vasodilators such as hydrogen peroxide (52). Although these changes may partially improve endothelial dysfunction, they are likely not able to reduce vascular TG2 activation and arterial wall stiffening. Indeed, our observation that the ratio of ACh to SNP vasodilatory responses did not differ between any of the cohorts suggests that the relationship between endothelial-dependent and independent responses were not significantly affected by cystamine. In comparison, the structural changes induced by cystamine significantly reduced vascular stiffness. Therefore, the reduced ACh responses observed in the WD cohort may be due to reductions in vascular smooth muscle responsiveness associated with stiffening of the cells and vascular wall. The observation that cystamine partially increased responsiveness of the aortic rings to SNP without increasing ACh responses suggests that cystamine ameliorated the impaired responsiveness of vascular smooth muscle to NO without improving endothelial function. Our data further suggest this could be due to the reductions in vascular smooth muscle F-actin content, cellular stiffness, and collagen content effected by cystamine.

Vascular stiffening involves structural changes in the vascular wall that reduce its capacity to be deformed in response to forces generated by physiological processes such as the pulsatile flow of blood. Those structural changes include the production and crosslinking of extracellular matrix proteins with different elastic properties, as well as the polymerization of cytoskeletal proteins that provide vascular cells, in particular smooth muscle, with their capacity to withstand circumferential stress (9, 53, 54). Evidence indicates that in the setting of WD feeding, vascular stiffening is indeed associated with increased amounts of extracellular matrix remodeling, including collagen accumulation, with or without elastin modifications (25, 26). These particular changes in extracellular components have long been associated with vascular stiffening (54). However, relatively recent findings indicate that the stiffness of vascular smooth muscle cells is also an important component of the vascular stiffening process (9, 55). It has been specifically shown that vascular smooth muscle cells present in the wall of stiffened arteries have increased amounts of F-actin in conjunction with an increased phosphorylation and inactivation of cofilin, an enzyme responsible for severing and diminishing F-actin stress fibers (34). Notably, the synthesis and crosslinking of collagen, as well as the phosphorylation of cofilin and polymerization of actin, are processes associated with transglutaminase activity, in particular that of TG2 (53, 56). In addition, we have previously reported that TG2 activity is increased in vessels from WD-fed mice (25), whereas the phosphorylation of cofilin has been shown as augmented in subjects with both hypertension and diabetes, who also have stiffened arteries (34, 57, 58).

In this study, the presence of reduced amounts of collagen and F-actin in the femoral and mesenteric arteries of WD + C mice strongly suggests that cystamine reduced vascular stiffness via its transglutaminase inhibitory properties. Cystamine is the product of cystine decarboxylation with added disulfide bonds between molecules. The biological effects of cystamine include that of transglutaminase inhibition, which results from its capacity to bind the catalytic domain of the enzyme (15, 59). In addition, the reduced form of cystamine serves as a substrate for the transglutaminases and competes for their activity with other substrates (15, 16). Herein, we determined that cystamine indeed diminishes transglutaminase activity in vascular smooth cells, as its presence reduced the incorporation of cadaverine onto cultured cells. Cadaverine is a transglutaminase substrate, commonly used to measure transglutaminase activity, in particular that of TG2 (60). In addition, as vascular smooth muscle cells predominantly synthesize TG2 (56, 61), the observation that cystamine diminished cadaverine incorporation indicates that it worked as a TG2 inhibitor in our in vitro experimental setting.

We also determined that cystamine reduced the collagen compaction capacity of vascular smooth muscle cells in culture. This collagen compaction capacity of vascular smooth muscle has been previously demonstrated to be a feature of TG2 being secreted by the cells (60–62). Thus, cystamine’s ability to reduce the movement of beads embedded within a collagen substrate in our cell culture experiments suggests that it inhibited TG2 and diminished collagen crosslinking. This is also consistent with our observation that collagen density was reduced in femoral and mesenteric arteries from WD + C compared with those from WD-fed mice. Cystamine did not have an effect on the viability of cells in culture, but reduced the amount of F-actin present in each cell as well as the F-actin:G-actin ratio. This indicates that cystamine promoted vascular smooth muscle actin depolymerization, which may be related to the role that TG2 plays as a G protein. In that role, TG2 activity initiates cascades associated with the activation of Rho kinase and its downstream effectors (56, 63, 64). One such cascade involves the activation of LIM kinase, with the subsequent phosphorylation and inactivation of cofilin, which results in actin polymerization, F-actin accumulation, and increased cellular stiffness (34, 65). Indeed, our results showing that F-actin density was reduced in femoral and mesenteric arteries from WD + C versus WD mice further support the concept that cystamine diminished the actin polymerization capacity of TG2, which is consistent with the notion that TG2, acting as a G protein, increases LIM kinase activity, coffin inactivation, F-actin formation, and cellular stiffening (34, 53, 65). The precise mechanisms by which cystamine diminishes the cascade going from TG2 to actin polymerization remains to be experimentally determined. Nonetheless, our results show that reducing smooth muscle F-actin content is one of the features by which cystamine lessens vascular stiffening in WD-fed mice.

Several aspects of this study warrant further consideration. Our data support TG2 inhibition as the primary mechanism by which cystamine reduced vascular stiffness. However, cystamine can contribute to reducing vascular stiffening by additional mechanisms, as it is an antioxidant and a caspase inhibitor (66, 67). Although interrogation of such mechanisms is beyond the scope of the present study, the fact that cystamine did not improve endothelial function or affect cellular viability indicate that its antioxidant and antiapoptotic capabilities may not be major means by which cystamine affects vascular stiffness. Also, our studies were performed in female mice and thus findings cannot be extrapolated to males. Exploring the dimorphic response to TG2 inhibition in males versus females in upcoming studies will provide a more comprehensive understanding of the role played by sex in response to TG2 inhibition and would help to identify subpopulations that would benefit differentially from this strategy.

Unexpectedly, vasodilatory responses to ACh in WD were not different than CD or WD + C in the femoral and mesenteric arteries. Power calculations using previously collected data indicated that an n = 5 per group would be sufficient to detect a difference of 29.7% with a power of 80% and an α = 0.05. Our results, therefore, suggest that either the level of variability increased on the older mice used in this study compared with our previously published data (68) or that compensatory mechanisms alleviating endothelial dysfunction took place that reduced the expected difference between cohorts.

Finally, our study emphasizes the importance of vascular stiffness in the pathogenesis of vascular dysfunction and cardiovascular disease, and is associated with increased risk for CVD, independently of other risk factors (4–6). Of translational importance, this parameter can be readily measured in vivo by noninvasive techniques and therefore can be easily adopted for widespread use in clinical practice. More importantly, to the best of our knowledge, there is a lack of strategies directed specifically at preventing or reversing vascular stiffness.

In summary, the present study adds to current knowledge regarding the role of TG2 inhibition as a potential therapeutic strategy in the management of vascular stiffening. Furthermore, our data suggest that inhibition of TG2 with cysteamine, a medication already used for different conditions, might bear therapeutic potential to reduce vascular stiffness and thus contribute to reduce the disproportionate burden of CVD in females in conditions of chronic overnutrition.

GRANTS

This work was supported by the Veterans Affairs (VA) Merit Grants 5I01BX001981 (to G.L.) and I01BX003391 (to A.W.-C.); National Institutes of Health Grants R01 HL088105 (to L.A.M.-L.), R01HL142770 (to C.M.-A.), K08HL132012 (to G.L.), RO1 NIDDK-DK124329 (to G. Jia), and R01 HL137769 (to J.P.); Fundação de Amparo à Pesquisa do Estado de São Paulo Grant 2018/18854-0 (to L.F.P.); and VA Medical Research Foundation (to V.G.D.).

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

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