Mutation of the 5′-untranslated region stem-loop mRNA structure reduces type I collagen deposition and arterial stiffness in male obese mice

Abstract

Arterial stiffening, a characteristic feature of obesity and type 2 diabetes, contributes to the development and progression of cardiovascular diseases (CVD). Currently, no effective prophylaxis or therapeutics is available to prevent or treat arterial stiffening. A better understanding of the molecular mechanisms underlying arterial stiffening is vital to identify newer targets and strategies to reduce CVD burden. A major contributor to arterial stiffening is increased collagen deposition. In the 5′-untranslated regions of mRNAs encoding for type I collagen, an evolutionally conserved stem-loop (SL) structure plays an essential role in its stability and post-transcriptional regulation. Here, we show that feeding a high-fat/high-sucrose (HFHS) diet for 28 wk increases adiposity, insulin resistance, and blood pressure in male wild-type littermates. Moreover, arterial stiffness, assessed in vivo via aortic pulse wave velocity, and ex vivo using atomic force microscopy in aortic explants or pressure myography in isolated femoral and mesenteric arteries, was also increased in those mice. Notably, all these indices of arterial stiffness, along with collagen type I levels in the vasculature, were reduced in HFHS-fed mice harboring a mutation in the 5′SL structure, relative to wild-type littermates. This protective vascular phenotype in 5′SL-mutant mice did not associate with a reduction in insulin resistance or blood pressure. These findings implicate the 5′SL structure as a putative therapeutic target to prevent or reverse arterial stiffening and CVD associated with obesity and type 2 diabetes.

NEW & NOTEWORTHY In the 5’-untranslated (UTR) regions of mRNAs encoding for type I collagen, an evolutionally conserved SL structure plays an essential role in its stability and posttranscriptional regulation. We demonstrate that a mutation of the SL mRNA structure in the 5’-UTR decreases collagen type I deposition and arterial stiffness in obese mice. Targeting this evolutionarily conserved SL structure may hold promise in the management of arterial stiffening and CVD associated with obesity and type 2 diabetes.

INTRODUCTION

Stiffening of the vasculature, a characteristic feature of obesity and diabetes (1–4), is a causal factor and independent prognosticator of cardiovascular morbidity and mortality (5–11). Indeed, extensive data in humans demonstrate that aortic pulse wave velocity (PWV), the gold standard measurement of arterial stiffness in vivo, independently predicts cardiovascular disease (CVD) risk and mortality (5–8, 12). For example, a meta-analysis revealed that a 1 m/s increase in aortic PWV augments the risk of total cardiovascular events, cardiovascular mortality, and all-cause mortality by 15%, even after adjusting for age, sex, and other common risk factors, including hypertension (13). However, despite the indisputable recognition that arterial stiffening contributes to the pathogenesis of CVD, the molecular mechanisms underlying arterial stiffening in obesity remain largely unknown. Thus, a deeper understanding of such molecular events is critical to identify novel targets to prevent or reverse obesity-associated arterial stiffening and CVD.

Arterial stiffness can be influenced by endothelial function via the modulation of smooth muscle tone and/or by alterations in the integrity of the extracellular matrix (14). Collagen and elastin are the two primary scaffolding proteins in the extracellular matrix that contribute to the structural integrity and elasticity of an arterial wall. Under physiological conditions, a dynamic process of controlled production and degradation typically maintains a stable content of these molecules in the vascular wall. However, in the setting of obesity, chronic low-grade inflammation and oxidative stress disrupt this balance, and result in collagen overproduction (11). This ultimately leads to increased collagen deposition and cross linking, and the development and progression of arterial stiffening (15).

Type I collagen, the predominant collagen subtype contained within the vascular wall, is a heterotrimeric cross-linked protein. Its synthesis is regulated at both transcriptional and post-transcriptional levels. On the latter, an evolutionarily conserved stem-loop (SL) structure in the 5′-untranslated region (UTR) of type I collagen mRNA plays a key role in its stability and post-transcriptional regulation (16–19). This is due to its binding of LARP6 [La Ribonucleoprotein 6, Translational Regulator; also known as Acheron (20)], an RNA-binding protein, thus implicating the 5′SL region as a potential target to modulate arterial stiffening. In fact, it has been previously reported that a mutation in the 5′SL region that disrupts the SL structure and prevents LARP6 binding markedly reduces collagen type I expression in a mouse model of bile duct ligation-induced liver fibrosis, without altering basal expression (21). Accordingly, the purpose of this investigation was to test the hypothesis that a mutation of the SL structure in the 5′-UTR suppresses type I collagen deposition and protects mice against obesity-associated arterial stiffening.

METHODS

Ethics and Approval

All animal procedures were approved by the Institutional Animal Care and Use Committee at the University of Missouri, Columbia. The University of Missouri is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International. Data supporting the findings presented here can be made available upon reasonable request to the corresponding authors.

Animals and Experimental Design

The 5′SL mutant mice on C57BL/6 background used in this study were previously described (21) and were bred in-house by crossing heterozygous 5′SL mutant mice. At 5–6 wk of age, male mice that were genotyped were grouped into 5′SL homozygous mutant and wild-type littermates, and randomly assigned to either remain on standard chow diet (3.35 kcal/g of food consisting of 13.4% kcal from fat, 29.9% kcal from protein, and 56.7% kcal from carbohydrate; Laboratory Rodent Diet 5001*, Lab Diet) or fed a high-fat/high-sucrose diet (HFHS; 5.06 kcal/g of food consisting of 60.1% kcal from fat, 14.9% kcal from protein, and 25.0% kcal from carbohydrate; F1850, Bio Serv) for 28 wk. As such, three experimental groups were created: wild-type chow fed, wild-type HFHS fed, and homozygous 5′SL-mutant HFHS fed. This experimental design allowed us to examine the impact of the 5′SL mutation in the setting of vasculometabolic derangements caused by HFHS feeding. The inclusion of wild-type chow-fed mice was designed to provide evidence that HFHS feeding indeed produced the intended metabolic and vascular phenotype in wild-type mice. Since naive 5′SL mutant mice exhibited no phenotypic abnormalities (21), we did not study chow-fed mutant mice in the present investigation. Moreover, male, but not female, mice were utilized based on our recent report that only male mice develop the intended HFHS diet-induced arterial stiffening (22). For genotype determination, the following primer pair was used: 5′- AAGGGGCCCAGGCCAGTCGTCGGAGC-3′ and 5′- CTTTCCTTATGAATCATCCCGCAGCC-3′, resulting in the detection of the wild-type band at 595 bp and the 5′SL mutant band at around 675 bp. Mice were housed either individually or two to three per cage, based on the number of litters available, in an environmentally controlled facility maintained at thermoneutrality (28°C) on a 12-h light:dark cycle from 0700–1900 with ad libitum access to water and food. Number of mice in a cage did not influence weight gain significantly (data not shown). Food intake in HFHS-fed mice was determined by weighing food in and out over a 5-day period and normalized to the number of mice per cage. This assessment was performed every ∼5 wk throughout the dietary intervention and the data averaged over time. Total energy expenditure (i.e., average of light and dark cycles) was assessed using metabolic cages at 24 wk following initiation of HFHS feeding, as we previously described (23). Food was removed from the cages approximately 3 h before euthanasia. All procedures were performed by investigators that were blinded to the experimental conditions.

Blood Pressure

At 25 wk of the dietary intervention, blood pressure was determined noninvasively using a CODA tail-cuff blood pressure system (CODA-HT2; Kent Scientific, Torrington, CT). Animals were acclimated to the restraints and tail-cuffs for three consecutive days before blood pressure determination. A minimum of eight blood pressure readings were recorded and averaged for each animal (22).

Insulin Tolerance Test

At 26 wk of the dietary intervention, insulin tolerance test (ITT) was performed (24). Briefly, after a 5-h fast, blood was collected from the tail vein and glucose levels were quantified using a glucometer (Alpha Trak 2, Abbott Labs). A baseline measure of blood glucose was taken before giving a sterile solution of regular insulin (1.0 U/mL/kg body wt) via intraperitoneal injection. Glucose measures were taken 15, 30, 45, 60, 90, and 120 min after the insulin injection. Glucose area under the curve (AUC) was calculated using the trapezoidal rule.

Aortic Pulse Wave Velocity

Aortic pulse wave velocity (PWV), the in vivo gold standard measurement of arterial stiffness, was assessed before euthanization using ultrasound imaging (Vevo 2100, VisualSonics, Toronto, ON, Canada) (25). Briefly, mice were anesthetized with isoflurane and placed in the supine position on a heated platform (42°C). Warmed ultrasound gel was applied to the abdominal surface and a 40‐MHz ultrasound transducer (Vevo MS550D, Toronto, ON) was used to collect B‐mode and ECG‐gated kilohertz visualization mode images. Images of the aortic wall and lumen along the longitudinal axis were collected for PWV measurements. PWV was calculated as the ratio of the distance between two locations along the aorta and time delay of the pulse wave between the locations. Data are expressed in meters per second.

Vascular Reactivity and Mechanical Testing in Isolated Arteries

Aortas were harvested and cleaned of perivascular adipose tissue in ice-cold physiological saline solution (PSS; pH 7.4) and cut into 2-mm segments. Aortic rings were then mounted in wire myograph organ bath chambers (620 M, Danish Myo Technology, Hinnerup, Denmark) containing warmed PSS gassed with 95% O₂-5% CO₂ at 37°C, as described previously (22, 26). Aortas were then preconstricted with the prostaglandin H2/thromboxane A2 receptor agonist U-46619 (20 nM) to test vasorelaxation responses to the endothelium-dependent vasodilator acetylcholine (ACh) and the endothelium-independent vasodilator sodium nitroprusside (SNP). The proximal femoral arteries were isolated and cannulated onto glass micropipettes, pressurized at 70 mmHg without flow, and warmed to 37°C in commercial pressure myography chambers (Living Systems Instrumentation, Burlington, VT), as previously described (27). Arteries were preconstricted with phenylephrine (10 µM) and vasodilator responses to ACh, insulin, and SNP were determined. Arteries that failed to respond to the preconstrictor were considered unviable and were excluded from analysis. Pressure myography was also utilized to determine the elastic properties of femoral and mesenteric arteries under passive conditions, as previously described (28–30). To ensure that vessels were under passive conditions, vessels were washed three times in a calcium-free buffer containing 2 mM EGTA and 100 µM adenosine. Vessels were then exposed to consecutive 2-min changes in intraluminal pressure from 5 to 120 mmHg. Throughout the vasoreactivity and mechanical testing protocols, chambers were mounted on inverted microscopes with CCD cameras. Luminal diameter and wall thicknesses were recorded using a video caliper (Living Systems Instrumentation) and a PowerLab data acquisition system (ADInstruments, Inc., Colorado Springs, CO). At the end of the experiment, vessels were fixed with 4% paraformaldehyde for further structural wall composition analysis.

Wall thickness and pressure-diameter curves were used to calculate strain-stress relationships, incremental modulus of elasticity (E_inc, higher values indicate greater stiffness), and incremental pulse wave velocity (cPWV_inc), as we have previously reported (28, 31). The high modulus of elasticity (E_high) was also calculated using a linear fitting from the high-pressure region of the strain-stress curves (30). At high pressures, collagen is considered to be the dominant mechanoelastic element, whereas at lower pressures elastin is dominant (32, 33). Equations used for these analyses were recently summarized in a guidelines article for the measurement of vascular function and structure in isolated arteries (34).

Aortic Stiffness via Atomic Force Microscopy

To evaluate aortic stiffness ex vivo, a 2-mm segment of the mouse thoracic aorta was harvested. The aorta was opened longitudinally (en face), and the adventitial surface of each explant fastened to a plastic coverslip using Cell-Tak. Stiffness was measured by atomic force microscopy (AFM) using a nanoindentation protocol with an MFP-3D AFM (Asylum Research, Inc. Goleta, CA) mounted on an Olympus IX81 microscope (Olympus, Inc.). Analysis was conducted using a tailor-made Python script, as previously described (22, 27, 35).

mRNA Expression in Aortic Tissues

Abdominal aortas were individually homogenized in TRIzol reagent using an Omni Bead Ruptor (Omni International, Kennesaw, GA). Total RNA was isolated using Qiagen’s RNeasy Kit per manufacturer’s instructions. RNA concentration and purity were assessed using a Nanodrop spectrophotometer (Thermo Fisher Scientific, Wilmington, DE). First-strand cDNA was synthesized from total RNA using the 5× iScript RT supermix (Bio-Rad, Hercules, CA). Quantitative real-time PCR was performed as previously described (36, 37) using the Bio-Rad CFX Real-Time PCR System (Bio-Rad). Annealing temperatures were kept at 60°C. GAPDH served as an invariant control. Relative mRNA expression levels were calculated by the 2^−ΔΔCt method. Primer sequences were as follows: COL1A1 sense 5′- CAGAAGATGTAGGAGTCGAG-3′, antisense 5′- TCATAGCCATAGGACATCTG-3′ (GenBank Accession No.: NM_007742); LARP6 sense 5′- AGTCCTTTATCTGAGGTCAC-3′, antisense 5′- GACTCCTAGAGATGACAGAC-3′ (GenBank Accession No.: NM_026235); GAPDH sense 5′- CTCACTCAAGATTGTCAGCA-3′, antisense 5′- GTCTTCTGGGTGGCAGTGAT-3′ (GenBank Accession No.: XM_036165840). Blasting the primers identified no off-targets.

Aortic Picrosirius Red and Verhoeff–Van Gieson Staining

Two-millimeter segments of the thoracic aorta were cleaned of surrounding adipose tissue and fixed in 10% formalin for 48 h followed by 70% ethanol. To retain the structural integrity of the vessel, a 1.5% agar solution was used to house the vessel before paraffin embedding. Blocks were sectioned and stained with picrosirius red (PSR) and Verhoeff–Van Gieson (VVG) by a commercial laboratory (IDEXX Bioanalytics, Columbia, MO).

Images for PSR were taken in the Molecular Cytology Core at the University of Missouri using a Leica DM5500 microscope using polarized light and a ×10 objective. Polarized PSR images show color variations that depend on fiber organizational structure and favor a red staining for collagen type I (38). Red color was extracted from the images and used for analysis of collagen content with a tailor-made MATLAB script. Three regions of interest were selected on empty areas from all images to quantify the background noise, and the averages were subtracted from the images. Pixels with intensities higher than the mean background noise were used for quantification.

Images for VVG were taken using an Olympus IX81 microscope with a ×10 objective. We used a tailor-made MATLAB script to quantify elastin laminae, intima-media thickness, and arterial diameter. Images were analyzed by creating a mask that retains only the intima-media layer for analysis. The area occupied by the elastin laminae, in purple, was quantified after removing the intensity below an established threshold and normalized to the region of interest (i.e., the intima-media layer). The intima-media thickness and arterial diameter were calculated assuming the ring was composed of two concentric circles.

Confocal/Multiphoton Fluorescence Microscopy Imaging of Femoral Arteries

Confocal/multiphoton fluorescence microscopy imaging was used to quantify collagen and elastin content in the vessel wall, as previously described (29, 30, 39). Briefly, before imaging, fixed arteries were rinsed in phosphate-buffered saline (PBS) and in 0.1 M glycine. After cannulation, vessels were flushed with PBS and permeabilized via incubation in 0.5% Triton X-100 for 15 min, followed by incubation for 1 h in 0.2 µM Alexa Fluor 633 Hydrazide (Molecular Probes) to stain elastin. Next, vessels were imaged using a Leica SP5 confocal/multiphoton microscope with a ×63/1.2 numerical aperture water objective. Alexa Fluor 633 was excited with a 633 nm HeNe laser. Collagen was imaged via second-harmonic image generation using a multiphoton laser emitting at 850 nm. Z-stack images were taken at increments of 0.5 µm from outside the vessel wall to mid-diameter. A tailor-made MATLAB scrip was used to quantify the volume of the molecules of interest as previously described (29, 30).

Statistical Analysis

GraphPad Prism (version 8.2.0, GraphPad Prism Software, La Jolla, CA) was used for statistical analysis. The Shapiro–Wilk test was used to confirm that the data were normally distributed. Statistical comparisons were performed using analysis of variance (ANOVA). When ANOVA was significant, the Tukey post hoc test was used for preplanned pairwise comparisons. Differences in aorta diameter and β values for femoral arteries were analyzed using a nonparametric one-way ANOVA (Kruskal–Wallis) test and followed with a Dunn’s post hoc test for multiple comparisons, when appropriate. Values are expressed as means ± SE. A P value less than or equal to 0.05 was considered significant.

RESULTS

HFHS Feeding-Induced Weight Gain, Adiposity, and Insulin Resistance Are Not Altered in Male 5′SL-Mutant Mice

A representative PCR image displaying the successful mutation of the 5′-UTR SL structure is depicted in Fig. 1A. Twenty-eight weeks of HFHS feeding led to marked and similar increases in body weight (Fig. 1B), adiposity (Fig. 1C), and indices of insulin resistance (Fig. 1, D–F) in both 5′SL-mutant and wild-type littermate mice compared with chow-fed wild-type littermates. As expected, the 5′SL mutation did not influence food intake (wild-type = 2.00 ± 0.12 g/day; 5-SL-mutant = 2.08 ± 0.03 g/day; P = 0.416) or total energy expenditure (wild-type = 10.60 ± 0.30 kcal/day, 5′SL-mutant = 10.54 ± 0.36 kcal/day; P = 0.913) in HFHS-fed mice.

Figure 1.

The 5′ stem-loop (SL)-mutant mice are not protected against obesity-associated insulin resistance. A: representative PCR gel image confirming the genotypes (wild-type band at 595 bp and 5′SL-mutant band at ∼675 bp). B: changes in body weight over time; n = 12–21/group. Two-way ANOVA with repeated measurements and Tukey post hoc test. C: percent body fat mass as assessed by echoMRI; n = 10–19/group. One-way ANOVA with Tukey post hoc test. D: blood glucose levels assessed over time in response to an intraperitoneal bolus of insulin injection (i.e., insulin tolerance test, ITT) and glucose area under the curve (AUC); n = 11–18/group. Two-way ANOVA with repeated measurements and Tukey post hoc test; inset: one-way ANOVA with Tukey post hoc test. E: plasma insulin levels; n = 12–19/group. One-way ANOVA with Tukey post hoc test. F: homeostatic model assessment of insulin resistance (HOMA-IR); n = 12–19/group. One-way ANOVA with Tukey post hoc test. All values are expressed as means ± SE. *Significant difference from chow-fed wild-type mice, P ≤ 0.05.

Collagen Type I Deposition in Vasculature is Reduced in 5′SL-Mutant Mice

As displayed in Fig. 2A, there was a trend for aortas from wild-type mice fed HFHS to express increased levels of collagen type I mRNA, and for this effect to be reduced in 5′SL-mutant mice; however, these differences did not reach statistical significance (ANOVA P > 0.05). PSR staining of aortic tissue sections using polarized light fluorescence microscopy did indicate that HFHS-fed wild-type littermates had increased aortic collagen type I deposition (Fig. 2A), which was significantly reduced in 5′SL-mutant mice. Similarly, confocal/multiphoton fluorescence imaging of femoral arteries reveled that HFHS feeding in wild-type littermates caused an increase in collagen fibers in the vessel wall (Fig. 2B), an effect that was significantly attenuated in 5′SL-mutant mice. Aortic and femoral artery elastin content was not significantly affected by diet or genotype (Fig. 2, A and B, ANOVA P > 0.05). The expression of LARP6 that controls stability and post-transcriptional regulation of collagen type I was not significantly altered in aortas across all three groups (Fig. 2A, ANOVA P > 0.05).

Figure 2.

Arterial collagen type I deposition is reduced in 5′ stem-loop (SL)-mutant obese mice. A: mRNA expression of collagen type I and LARP6 (La Ribonucleoprotein 6, Translational Regulator) in abdominal aorta homogenates (n = 8–17/group), as well as determination of red-channel area above threshold of picrosirius red stain (PSR) using polarized light fluorescence microscopy in paraffin-embedded aortic tissue sections (n = 10–14/group). Representative PSR images are displayed to the right (positive stain noted as artificial red). Elastin area was quantified as percent positive Verhoeff–Van Gieson (VVG) stain from the intima-media layer of the artery wall (n = 7–10/group). Representative VVG images are displayed to the right. One-way ANOVA with Tukey post hoc test. B: determination of collagen (using second-harmonic generation, SHG) and elastin (Hydrazide Alexa 633) content in the wall of femoral arteries using confocal/multiphoton fluorescence microscopy imaging (n = 6–9/group). Representative confocal images are displayed to the right (collagen shown in green and elastin in red). One-way ANOVA with Tukey post hoc test. All values are expressed as means ± SE. *Significant difference from wild-type chow, P ≤ 0.05. #Significant difference between wild-type and 5′SL-mutant mice fed high-fat/high-sucrose (HFHS), P ≤ 0.05.

HFHS Feeding-Induced Arterial Stiffening is Blunted in 5′SL-Mutant Mice

As shown in Fig. 3A, male mice fed the HFHS diet exhibited an increase in blood pressure independent of genotype. In wild-type littermates, HFHS feeding also led to an increase in aortic stiffness, as assessed in vivo via PWV (Fig. 3B) and ex vivo via AFM (Fig. 3C), but these stiffness parameters were markedly reduced in 5′SL-mutant mice. Similarly, HFHS feeding increased arterial stiffness in femoral (Fig. 3D) as well as mesenteric (Fig. 3E) arteries from wild-type littermates, but not 5′SL-mutant mice, as determined by their strain-stress relationships, the incremental moduli of elasticity, and calculated PWVs. As illustrated, HFHS feeding caused an increase in incremental moduli of elasticity (i.e., a higher value indicative of a stiffer vessel) in wild-type, but not 5′SL-mutant mice, at high-intraluminal pressures. Changes in distensibility at this high-pressure range suggest alterations in collagen levels (32, 33). Differences at the low-pressure range, which did not occur, would have been suggestive of alterations in elastin (32, 33). Accordingly, these stiffness data are congruent with the structural data presented in Fig. 2. Of note, aortic diameter and intima-media thickness, as assessed in VVG-stained slides, were not significantly influenced by diet or genotype (diameter: wild-type chow = 502.8 ± 14.9 µm, wild-type HFHS = 525.1 ± 15.9 µm, 5′SL-mutant HFHS = 510.3 ± 15.0 µm, ANOVA P = 0.632; intima-media thickness: wild-type chow = 51.8 ± 2.3 µm, wild-type HFHS = 53.6 ± 2.2 µm, 5′SL-mutant HFHS = 49.6 ± 1.8 µm, ANOVA P = 0.443). Femoral and mesenteric artery diameter and wall thickness in pressurized (70 mmHg) vessels were also not significantly affected by diet or genotype (femoral artery diameter: wild-type chow = 306.1 ± 20.1 µm, wild-type HFHS = 288.3 ± 24.2 µm, 5′SL-mutant HFHS = 292.7 ± 16.3 µm, ANOVA P = 0.831; femoral artery wall thickness: wild-type chow = 25.9 ± 2.9 µm, wild-type HFHS = 25.6 ± 3.0 µm, 5′SL-mutant HFHS = 29.8 ± 2.4 µm, ANOVA P = 0.449; mesenteric artery diameter: wild-type chow = 237.8 ± 8.0 µm, wild-type HFHS = 226.1 ± 4.1 µm, 5′SL-mutant HFHS = 235.4 ± 7.3 µm, ANOVA P = 0.480; mesenteric artery wall thickness: wild-type chow = 15.1 ± 1.0 µm, wild-type HFHS = 14.8 ± 0.7 µm, 5′SL-mutant HFHS = 15.7 ± 0.6 µm, ANOVA P = 0.690).

Figure 3.

Obesity-associated arterial stiffening is decreased in 5′ stem-loop (SL)-mutant mice despite no reduction in blood pressure. A: systolic, diastolic, and mean arterial blood pressure (MAP) as determined by tail-cuff plethysmography; n = 5–12/group. One-way ANOVA with Tukey post hoc test. B: aortic pulse wave velocity (PWV) as assessed by VEVO2100 High-Resolution Ultrasound Imaging; n = 5–7/group. One-way ANOVA with Tukey post hoc test. C: stiffness in en face aortic preparations as determined by atomic force microscopy; n = 10–18/group. One-way ANOVA with Tukey post hoc test. D: femoral artery stiffness assessed via strain-stress relationships. β-Values are also included. Femoral artery incremental moduli of elasticity (E_inc) represented as dynes/cm² at increasing pressures. The high moduli of elasticity at high pressures (E_high) highlight the role of collagen and are depicted as a bar graph inset. A calculated incremental pulse wave velocity (cPWV_inc) is also provided. Measurements were made under passive conditions; n = 8–13/group. Two-way ANOVA with repeated measurements and Tukey post hoc test; insets: Kruskal–Wallis with Dunn’s post hoc tests for β-index and one-way ANOVA with Tukey post hoc test for E_high. E: mesenteric arterial stiffness assessed via strain-stress relationships. β-Values are also included. Mesenteric artery E_inc represented as dynes/cm² at increasing pressures. The E_high highlights the role of collagen and is depicted as a bar graph inset. A cPWV_inc is also provided. Measurements were made under passive conditions; n = 9–12/group. Two-way ANOVA with repeated measurements and Tukey post hoc test; insets: one-way ANOVA with Tukey post hoc test. All values are expressed as means ± SE. *Significant difference from wild-type chow, P ≤ 0.05. #Significant difference between wild-type and 5′-SL mutant fed high-fat/high-sucrose (HFHS), P ≤ 0.05.

Reduced Arterial Stiffening in HFHS-Fed 5′SL-Mutant Mice is Accompanied by Increased Vasodilatory Function

As illustrated in Fig. 4, ACh-induced aortic vasorelaxation was increased in HFHS-fed male 5′SL-mutant mice relative to wild-type littermate controls. A slight increase in SNP-induced vasorelaxation was also observed. In the femoral artery, vasodilatory responses to ACh and insulin were also increased in HFHS-fed 5′SL-mutant mice with a small trend for an increase in the SNP response (ANOVA P = 0.091). Phenylephrine-induced vasoconstriction in the femoral artery was not affected by diet or genotype (ANOVA P = 0.259, data not shown).

Figure 4.

The 5′ stem-loop (SL)-mutant mice exhibit enhanced vasodilatory function. A: aortic responses to increasing concentrations of acetylcholine (ACh) and sodium nitroprusside (SNP) following preconstriction with the PGH2/TxA2 receptor agonist U46619; n = 8–12/group. One aortic ring from the HFHS-fed 5′-SL mutant group was removed from analysis as it responded abnormally to ACh with a maximal relaxation response of only 6% (presumably due to endothelial injury resulting from handling) and was considered a statistical outlier. Two-way ANOVA with repeated measurements and Tukey post hoc test. B: femoral artery responses to increasing concentrations of ACh, insulin, and SNP following preconstriction with phenylephrine (Phe); n = 8–15/group. Two-way ANOVA with repeated measurements and Tukey post hoc test. All values are expressed as means ± SE. *Significant difference from wild-type chow, P ≤ 0.05. #Significant difference between wild-type and 5′-SL mutant fed high-fat/high-sucrose (HFHS), P ≤ 0.05.

DISCUSSION

Arterial stiffening is a naturally occurring phenomenon during the aging process (40, 41) that is accelerated by obesity and type 2 diabetes (1–4, 42), likely resulting from adipokine dysregulation, oxidative stress, and inflammation (43–45). Increased arterial stiffness contributes to CVD pathogenesis and represents an independent predictor of adverse cardiovascular events, morbidity, and mortality (5–11). Although increased physical activity and weight loss are typically associated with reduced arterial stiffness (29, 46–50), currently, no effective pharmacological interventions are available to prevent or treat arterial stiffening. Because of the growing prevalence of obesity and type 2 diabetes, as well as increased aging of the population, an understanding of the molecular mechanisms underlying the development and progression of arterial stiffening is imperative for the identification of novel therapeutic targets and strategies to reduce CVD burden. It is well-recognized that a major contributor to arterial stiffening is increased collagen deposition (11). In the 5′-UTR of the mRNA encoding for type I collagen, there is an evolutionally conserved SL structure that is essential in its stability and post-transcriptional regulation (16–19). Here, for the first time, we provide evidence that obese male mice harboring a mutation in the 5′SL structure exhibit reduced type I collagen deposition and decreased arterial stiffness, implicating the 5′SL structure as a potential therapeutic target to ameliorate arterial stiffening.

We report that HFHS feeding for 28 wk results in increased adiposity, insulin resistance, blood pressure, and arterial stiffness in male wild-type littermates, a phenotype that is consistent with previous findings (22, 51). Arterial stiffness was assessed in vivo via aortic PWV, and ex vivo via AFM with nanoindentation in aortic explants and pressure myography in isolated femoral and mesenteric arteries. Notably, relative to wild-type littermates, all of these indices of arterial stiffness increased with HFHS feeding were reduced in the 5′SL-mutant mice. This protective vascular effect occurred without concomitant alterations in insulin resistance or blood pressure, indicating that reduced arterial stiffness in this model is independent of changes in metabolic function or blood pressure. In fact, there is precedence that arterial stiffness can be modulated by pressure-independent mechanisms in hypertension (52). Published reports also indicate that arterial stiffening in obesity precedes the development of hypertension and contributes to its pathogenesis (50). Accordingly, this novel finding that decreased arterial stiffness in obese 5′SL-mutant mice was not accompanied by a reduction in blood pressure should not be interpreted as clinically inconsequential. Actually, data are available indicating that the probability of all-cause survival in patients under antihypertensive therapy is greater in those that exhibit a reduction in arterial stiffness compared with those that show an increase in arterial stiffness, irrespective of the magnitude of blood pressure reduction (53). Thus, arterial stiffness should be considered an important clinical end point and target in patients at risk for developing CVD. Although lifestyle modifications, such as increased physical activity, are recommended for the prevention of CVD, data from human studies also indicate that exercise may not exert aortic destiffening effects once substantial structural changes in the arterial wall and severe stiffening have already developed (54, 55). These reports further underscore the importance of identifying newer therapeutic targets to reduce arterial stiffness.

The present investigation provides the first evidence that repression of type I collagen levels through a mutation of the SL structure in its 5′-UTR effectively blunts arterial stiffening. The downregulation in type I collagen in the vasculature of mutant mice paralleled with a slight decrease in elastin content, particularly in the aorta, though this decrease did not reach statistical significance. Whether this small reduction in arterial elastin in the face of collagen loss serves as a compensatory mechanism to maintain some degree of vascular wall homeostasis is not known. Nevertheless, these findings add to the growing literature indicating that increased collagen deposition and cross linking are important determinants of arterial stiffening (15) and implicate the 5′SL structure in collagen type I as a potential therapeutic target in the setting of increased arterial stiffness. Notably, we show this in a mouse model of diet-induced obesity, insulin resistance, increased blood pressure, and arterial stiffening that resembles the pathophysiology of obesity and type 2 diabetes in humans. Out of several hundred RNA binding proteins, LARP6 is the only one specifically involved in type I collagen regulation (19). LARP6 binds to the 5′SL structure to regulate the stability and translation of type I collagen mRNA (17, 19). Accordingly, based on our novel findings, it is conceivable that blocking LARP6 binding to collagen type I has therapeutic implications in treating obesity-induced arterial stiffening. To that end, future studies should determine whether C9, a small molecule inhibitor that has been reported to block LARP6/collagen type I interaction (56), exerts beneficial therapeutic effects similar to those seen in the 5′SL-mutant mice.

Another noteworthy finding of the present investigation is that reduced arterial stiffness in the 5′SL-mutant mice paralleled enhanced vasodilatory function. Although it is recognized in literature that endothelial dysfunction and the consequent diminished nitric oxide bioavailability contribute to arterial stiffening (57–59), data from the present investigation is supportive of the concept that changes in stiffness may also modulate vasomotor reactivity to dilatory agonists. That is, the observation that SNP-induced dilation, as well as ACh and insulin-induced dilation (i.e., functional markers of nitric oxide-dependent vasodilation), tends to be greater in 5′SL-mutant mice relative to wild-type mice fed HFHS suggests that mutation of the 5′SL structure did not have an impact on nitric oxide production or bioavailability. Rather, it appears that the enhanced vasomotor function phenotype in mutant mice is attributed to reduced collagen deposition and the consequent increase in arterial wall distensibility. However, we cannot rule out the possibility that alterations in collagen expression and deposition might also directly influence the endothelial cell phenotype, including the cytoskeleton, and the release of vasodilatory substances.

Several aspects of this investigation warrant further consideration. First, the present study included only male mice, and thus findings cannot be extrapolated to females. Second, as previously documented (22, 51), our HFHS diet evoked the intended obesity phenotype and vasculo-metabolic derangements. However, we recognize that the use of the standard grain-based chow diet is not the utmost suitable control diet against the administered HFHS, which is a purified ingredient diet. Future studies should consider utilizing a matching low-fat purified ingredient diet as the control to the present HFHS diet. Lastly, it should be noted that although type I collagen provides structural support to multiple tissues under physiological conditions, the 5′SL-mutant mice developed without any gross abnormal phenotype compared with wild-type littermates, an observation that is consistent with that of others using the same model (21) and relevant when considering this mechanism as a potential therapeutic target. Importantly, as noted by Parsons et al. (21), the basal levels of type I collagen produced in the 5′SL-mutant mice appear to be sufficient for normal development, survival, and breeding.

In aggregate, findings from the present study demonstrate that a mutation of the SL mRNA structure in the 5′-UTR, and the presumable loss of LARP6 binding, causes a decrease in collagen type I deposition and arterial stiffness in obese male mice, independent of changes in insulin resistance and blood pressure. Targeting this evolutionarily conserved SL structure may hold promise in the management of arterial stiffening and CVD associated with obesity and type 2 diabetes (Fig. 5).

Figure 5.

Schematic summary of the major finding that obese mice harboring a mutation of the stem-loop (SL) structure in the 5′-untranslated regions of type I collagen mRNA exhibit reduced arterial stiffness. Loss of LARP6 (La Ribonucleoprotein 6, Translational Regulator) binding to the 5′SL structure diminishes translation of type I collagen mRNA and leads to reduced collagen type I deposition and arterial stiffness. This work implicates the 5′SL structure in collagen type I as a potential therapeutic target to ameliorate arterial stiffening and obesity-associated cardiovascular disease (CVD). “+” sign denotes “increase in,” “−” sign denotes “decrease in.”

GRANTS

This project was supported by a grant from the University of Missouri School of Medicine Program Project Planning (to L. A. Martinez-Lemus) and National Heart, Lung, and Blood Institute Grants R01 HL088105 (to L. A. Martinez-Lemus) and R01 HL142770 (to C. Manrique-Acevedo). B. Chandrasekar is a Research Career Scientist (IK6BX004016), and his work is supported by the Veterans Affairs ORD-BLRD Service Award I01-BX004220. This work was also supported by the use of resources and facilities at the Harry S. Truman Memorial Veterans Hospital in Columbia, MO.

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

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