Hypertension and decreased aortic compliance due to reduced elastin amounts do not increase atherosclerotic plaque accumulation in Ldlr−/− mice

Abstract

Background and Aims

High blood pressure and reduced aortic compliance are associated with increased atherosclerotic plaque accumulation in humans. Animal studies support these associations, but additional factors, such as fragmented elastic fibers, are present in most previous animal studies. Elastin heterozygous (Eln+/−) mice have high blood pressure and reduced aortic compliance, with no evidence of elastic fiber fragmentation and represent an appropriate model to directly investigate the effects of these factors on atherosclerosis

Methods and Results

Eln+/− and Eln+/+ mice were crossed with low density lipoprotein receptor knockout (Ldlr−/−) and wild-type (Ldlr+/+) mice and fed normal or Western diet (WD) for 16 weeks. We hypothesized that on WD, Eln+/−Ldlr−/− mice with high blood pressure and reduced aortic compliance would have increased atherosclerotic plaque accumulation compared to Eln+/+Ldlr−/− mice. We measured serum cholesterol and cytokine levels, blood pressure, aortic compliance, and plaque accumulation. Contrary to our hypothesis, we found that on WD, Eln+/−Ldlr−/− mice do not have increased plaque accumulation compared to Eln+/+Ldlr−/− mice. At the aortic root, there are no significant differences in plaque area between Eln+/−Ldlr−/− and Eln+/+Ldlr−/− mice on WD (p = .89), while in the ascending aorta, Eln+/−Ldlr−/− mice on WD have 29% less normalized plaque area than Eln+/+Ldlr−/− mice on WD (p = 0.009).

Conclusion

Using an atherogenic mouse model, we conclude that increased blood pressure and reduced aortic compliance are not direct causes of increased aortic plaque accumulation. We propose that additional insults, such as fragmentation of elastic fibers, are necessary to alter plaque accumulation.

Introduction

Atherosclerosis is an inflammatory disease with both environmental and genetic components^{1, 2}. High systolic blood pressure^3–6 and decreased aortic compliance^{7, 8} are associated with increased atherosclerotic plaque accumulation, although the relationships are complicated and often contradictory. Better understanding of the risk factors in atherosclerosis would help inform preventative and palliative care. Although they have limitations, mouse models of atherosclerosis, including apolipoprotein-E (Apoe−/−)^{9, 10} and low density lipoprotein receptor (Ldlr−/−)¹¹ deficient animals, have become invaluable tools for studying mechanisms of atherosclerosis¹². Apoe−/− and Ldlr−/− mice fed a Western diet (WD, 20% fat, 0.15% cholesterol) develop hyperlipidemia and atherosclerosis.

Previously, Van Herck et al.¹³ investigated the relationship between atherosclerosis and aortic compliance by crossing Apoe−/− mice with mice haploinsufficient for a mutation in the fibrillin-1 gene (C1039G+/−). The authors found that on WD C1039G+/−Apoe−/− mice had decreased aortic compliance, which led to increased accumulation of atherosclerotic plaque, and promoted plaque instability compared to C1039G+/+Apoe−/− mice. Fibrillin-1 is a major component of microfibrils in the extracellular matrix that associate with developing elastic fibers. Mutations in fibrillin-1 cause Marfan Syndrome, an autosomal dominant disorder with skeletal, ocular, and cardiovascular manifestations. Cardiovascular symptoms include dilation and dissection of the ascending aorta¹⁴, and have been associated with dysregulated transforming growth factor beta (TGF-β) signaling¹⁵. C1039G−/− mice die soon after birth due to failure of the arterial wall, while C1039G+/− mice show defects in elastic fiber structure and local elastolysis, but have a normal lifespan^{16, 17}. Elastic fiber fragmentation and injection of elastin-derived peptides have been shown to potentiate atherosclerosis¹⁸. It is possible that TGF-β signaling and elastic fiber fragmentation contribute to the increased plaque accumulation and instability in C1039G+/−Apoe−/− mice, and that decreased aortic compliance is not a major factor.

During elastic fiber formation, microfibrils interact with tropoelastin, which is then crosslinked into insoluble elastin. Elastin haploinsufficient mice (Eln+/−) mice have about 60% of wild-type elastin levels, stable hypertension, and decreased aortic compliance¹⁹. Elastin haploinsufficiency in humans causes Supravalvular Aortic Stenosis, which can occur as an isolated disease or as a component of Williams-Beuren Syndrome²⁰. Mice and humans with elastin haploinsufficiency have thinner, more numerous elastic lamellae across the aortic wall, but they do not appear fragmented²¹. Hence, Eln+/− mice represent an appropriate model to isolate the effects of aortic compliance and hypertension on atherosclerosis progression, without complications of elastic fiber fragmentation and dysregulated TGF-β signaling. We bred Eln+/− mice to Ldlr−/− mice and fed them WD. We measured serum lipid and cytokine levels, blood pressure, aortic compliance, and atherosclerotic plaque accumulation and composition to determine if increased blood pressure and decreased aortic compliance in Eln+/−Ldlr−/− mice leads to increased plaque accumulation compared to Eln+/+Ldlr−/− mice.

Materials and Methods

Mice

Female B6.129S7-Ldlr^tm1Her/J −/− (Ldlr−/−) (Jackson Laboratory, stock #002207) were bred with male Eln+/− mice²². Males at the F3 – F6 generation were used because previous data on blood pressure and aortic stiffness in Eln+/− mice were obtained for males¹⁹. Genotypes included in the study are: Eln+/+Ldlr−/−, Eln+/−Ldlr−/−, Eln+/+Ldlr+/+, Eln+/−Ldlr+/+. After weaning at three weeks of age, mice were provided with normal diet (ND) or WD (AIN-76A, Purina Test Labs) for 16 weeks. All protocols were approved by the Institutional Animal Care and Use Committee.

Blood pressure, serum chemistry, and tissue collection

Mice were anesthetized with 2% isoflurane and intra-aortic blood pressure was measured with a 1.2F solid-state catheter (Transonic). Whole blood was collected from a subset of mice via cardiac puncture and serum was separated. Lipid levels were quantified by Advanced Veterinary Laboratory using an automated chemistry analyzer. Additional serum samples for Eln+/+Ldlr−/− and Eln+/−Ldlr−/− mice on ND and WD were analyzed for inflammatory cytokines and TGF-β1 using electrochemiluminescence immunoassays from Mesoscale Discovery and read on a MESO Quickplex SQ 120. Only analytes with levels higher than 10 pg/ml on the mouse inflammatory cytokine multiplex assay are presented. These include interleukin 6 (IL6), interleukin 10 (IL10), chemokine (C-X-C motif) ligand 1 (CXCL1), and tumor necrosis factor (TNF).

The heart was removed and the proximal region was frozen at −80°C in Tissue Tek OCT for sectioning of the aortic root. In about half of the mice, the ascending aorta was removed for mechanical testing and a small piece of the left carotid artery was fixed for wall structure analysis. In the rest of the mice, the entire aorta from the root to the iliac bifurcation was removed for en face plaque analysis.

Mechanical testing

The ascending aorta was mounted at the approximate in vivo length in a pressure myograph (110P, Danish Myotechnology) in physiologic saline solution at 37°C, as described previously²³. Arteries were inflated from 0 – 175 mmHg in steps of 25 mmHg (12 sec/step) while pressure, outer diameter, and axial force were recorded at 1 Hz. The diameter compliance was calculated as the change in diameter for each pressure step, and is an inverse measure of aortic stiffness.

Plaque quantification and characterization

Aortas for en face preparation were fixed in 10% neutral buffered formalin overnight, cut longitudinally, pinned to black dissection wax, stained with Oil Red O in propylene glycol, and imaged²⁴. Outlines of the ascending, thoracic, and descending regions of the aorta were defined and positive Oil Red O pixels were traced manually in Image J (NIH). Plaque area for each region was normalized to the aortic surface area. For aortic valve analysis, 5 μm frozen sections of the aortic root were cut with a cryostat. Slides were stained with Oil Red O in 60% isopropanol²⁵, but were not counterstained. Images were taken of slides at ~10 μm intervals. Three images for each mouse where clear sections of the aortic valves could be seen were analyzed for total Oil Red O positive pixels using Matlab software (Mathworks) and averaged.

Adjacent sections of the aortic valves for a subset of Eln+/+Ldlr−/− and Eln+/−Ldlr−/− mice on WD were stained with F4/80 antibody (ab16911, Abcam) followed by DAB and imaged to visualize macrophage content and localization. Adjacent sections of the aortic valves for the same mice were also examined by fluorescence microscopy. Sections were labeled for elastin, collagen, smooth muscle cells (SMCs), and cell nuclei. Alexa Fluor 633 Hydrazide (Life Technologies) was used for elastin^{26, 27}. CNA35 (kindly provided by Magnus Hook, Texas A&M) labeled with Oregon Green 488 (Life Technologies) was used for collagen²⁸. Alpha smooth muscle actin (αSMA) primary antibody (A5228, Sigma) followed by Alexa Fluor 555 goat anti-mouse secondary antibody (Life Technologies) was used for SMCs. The cell nuclei were stained with Hoechst 34580 (Life Technologies). The percentage of the plaque area staining positive for F4/80, αSMA, and collagen was calculated from thresholded images of at least two aortic valves/mouse using Image J software.

Arterial wall structure

Sections of the left common carotid artery were examined by histology for seven Eln+/+Ldlr−/− and six Eln+/−Ldlr−/− mice on WD. Two – three mm long pieces of the artery were fixed in 10% neutral buffered formalin overnight, dehydrated in a graded series of ethanol, embedded in paraffin, sectioned, stained with H&E, Verhoeff Van Gieson (VVG) or picrosirius red (PSR) and imaged.

Statistics

Three-way ANOVA (SPSS) was used to evaluate the effects of diet, Ldlr genotype, Eln genotype and the interactions between each independent variable on the measured dependent variables. Cytokine levels were only measured in Ldlr−/− mice, so two-way ANOVA was used to evaluate the effects of diet and Eln genotype. Additionally, two-way student's t-test with unequal variance was used to evaluate differences in plaque amounts and composition between Eln+/−Ldlr−/− and Eln+/+Ldlr−/− mice on WD. P < 0.05 was considered significant.

Results

Cholesterol increases with WD and Ldlr−/− genotype, but is not affected by Eln genotype

Total cholesterol, triglyceride, and low density lipoprotein (LDL) levels all increase with WD (200 – 400%), Ldlr−/− genotype (400 – 4000%), and show a significant interaction between diet and Ldlr genotype (Table 1). Eln genotype does not affect total cholesterol, triglyceride, or LDL levels. High density lipoprotein (HDL) levels are increased about 40% with WD, but are not affected by Ldlr or Eln genotype (Table 1).

Table 1.

Serum cholesterol levels and blood pressures in Eln+/+ and Eln+/− mice crossed with Ldlr+/+ and Ldlr−/− mice and fed ND or WD for 16 weeks after weaning. Total cholesterol, triglyceride (tri), low density lipoprotein (LDL) and high density lipoprotein (HDL) serum concentrations were measured. Diet and Ldlr genotype significantly affect most measures of serum cholesterol, while Eln genotype does not. Systolic (sys) and pulse pressures measured with a solid-state catheter are significantly affected by diet, Ldlr genotype, and Eln genotype, while diastolic (dias) pressures are affected only by diet. Heart rate (hr) is not affected by any of the independent variables. Significance was determined by three-way ANOVA for diet, Ldlr genotype, Eln genotype and all two- and three-way interactions.

			serum cholesterol (mg/dl)				blood pressure (mmHg)
Diet	Ldlr	Eln	total	tri	LDL	HDL	sys	dias	pulse	hr (bpm)
ND	Ldlr+/+	Eln+/+	138 ± 32	62 ± 13	6 ± 3	69 ± 11	109 ± 12	74 ± 10	35 ± 4	497 ± 62
		Eln+/−	169 ± 39	55 ± 7	7 ± 3	83 ± 13	122 ± 19	79 ± 12	43 ± 11	514 ± 57
	Ldlr−/−	Eln+/+	441 ± 50	117 ± 35	123 ± 16	80 ± 14	114 ± 9	76 ± 7	37 ± 4	514 ± 68
		Eln+/−	432 ± 114	266 ± 181	170 ± 83	87 ± 13	122 ± 17	77 ± 11	44 ± 8	510 ± 82
WD	Ldlr+/+	Eln+/+	238 ± 77	66 ± 18	15 ± 5	120 ± 29	117 ± 12	78 ± 7	39 ± 6	484 ± 89
		Eln+/−	209 ± 77	58 ± 12	12 ± 3	114 ± 20	122 ± 12	81 ± 8	40 ± 7	506 ± 85
	Ldlr−/−	Eln+/+	1040 ± 143	740 ± 351	698 ± 229	99 ± 17	123 ± 16	82 ± 10	41 ± 8	482 ± 58
		Eln+/−	1001 ± 85	808 ± 215	596 ± 216	112 ± 10	141 ± 17	85 ± 9	56 ± 10	480 ± 58
P value		diet	<0.001	<0.001	<0.001	<0.001	0.002	0.011	0.006	0.154
		Ldlr	<0.001	<0.001	<0.001	0.632	0.014	0.299	0.001	0.786
		Eln	0.596	0.215	0.628	0.122	<0.001	0.115	<0.001	0.565
	interactions		diet × Ldlr (<0.001)	diet × Ldlr (<0.001)	diet × Ldlr (<0.001)	-	-	-	diet × Ldlr (0.027), diet × Ldlr × Eln (0.018)	-
N/group			5 – 7	5 – 7	5 – 7	5 – 7	9 – 17	9 – 17	9 – 17	9 – 17

Italics indicate P < 0.05. Data are presented as mean ± SD.

Systolic and pulse pressure increase with WD, Ldlr−/− genotype and Eln+/− genotype

Systolic pressure is increased 8% with WD, 6% with Ldlr−/− genotype, and 10% with Eln+/− genotype, with no interactions between the independent variables (Table 1). Diastolic blood pressure is increased 7% with WD, but is not affected by Ldlr or Eln genotype (Table 1). Pulse pressure is increased 11% with WD, 13% with Ldlr−/− genotype, and 20% with Eln+/− genotype, with interactions between diet and Ldlr genotype and all three independent variables (Table 1). Heart rate is not significantly affected by diet, Ldlr or Eln genotype (Table 1).

Aortic compliance is reduced in Eln+/− mice at systolic pressures, but is not affected by diet or Ldlr genotype

Average pressure-diameter curves are shown in Fig. 1A. Ldlr genotype and diet do not have significant effects on the diameter for any applied pressure. Diameter is decreased 7 – 11% at pressures between 0 – 50 and 150 – 175 mmHg in Eln+/− mice (p = <.001 – .013). There are significant interactions between Eln genotype and diet at 0 – 50 mmHg (p = .007 – .042) and between all three independent variables at pressures above 0 mmHg (p = .001 – .045). Average pressure-compliance curves are shown in Fig. 1B. Ldlr genotype and diet do not have significant effects on the compliance for any applied pressure. The compliance curves for Eln+/− mice are shifted so that the peak occurs at lower pressures than Eln+/+, with a steep drop in compliance at higher pressures. At pressures of 50 and 75 mmHg, compliance is 21 – 43% higher in Eln+/− aorta compared to Eln+/+ (p = <.001 and .004). At pressures between 100 – 175 mmHg (near and above systolic pressure values), compliance is 19 – 53% lower in Eln+/− aorta compared to Eln+/+ (p = <.001 – .004). There are significant interactions between Eln and Ldlr genotype at pressures between 0 – 75 mmHg (p = .002 – .017).

Figure 1.

Ascending aortic diameter is reduced in Eln+/− mice compared to Eln+/+ (A, B), as determined from in vitro mechanical tests. Ascending aortic compliance, calculated as the local slope of the pressure-diameter curve, is reduced in Eln+/− mice compared to Eln+/+ at pressures near and above systolic values (C, D). Diet and Ldlr genotype have no significant effects on the diameter or compliance at any pressure. N = 4 – 8/group. * = P < .05 for Eln genotype by three-way ANOVA.

Atherosclerotic plaque is not increased in Eln+/− mice

Cross-sections of the aortic root were stained with Oil Red O to estimate atherosclerotic plaque accumulation. Total plaque at the aortic root is increased with WD (p < 0.001) and Ldlr−/− genotype (p < 0.001), but is not affected by Eln genotype (p = 0.991). There are significant interactions between diet and Ldlr genotype (p < 0.001). There are no significant differences for the plaque area at the aortic root between Eln+/+Ldlr−/− and Eln+/−Ldlr−/− mice on WD (p = 0.89) (Fig. 2A). Representative images of plaque at the aortic root are shown in Figs. 2B and C. En face preparations of the aorta were stained with Oil Red O as an additional estimate of plaque accumulation. As most of the plaque accumulation occurs in the ascending aorta in Ldlr−/− mice after 16 weeks of WD, we focused our analyses on that region. The fraction of ascending aortic lumen covered in plaque is increased with WD (p < 0.001) and Ldlr−/− genotype (p < 0.001), and is decreased in Eln+/− mice (p = 0.010). There are significant interactions between all two-way combinations and the three-way combination of independent variables (p < 0.001 – 0.011). Eln+/−Ldlr−/− mice on WD have 29% less normalized plaque area in the ascending aorta than Eln+/+Ldlr−/− mice on WD (p = 0.009) (Fig. 2D). Representative images of plaque in the ascending aorta are shown in Figs. 2E and F. In the thoracic and abdominal regions, there are no significant differences in plaque accumulation between Eln+/+Ldlr−/− and Eln+/−Ldlr−/− mice on WD, but the plaque amounts in these regions are highly variable (Supplemental Fig. I). When total normalized plaque area of the entire aorta is calculated, there are no significant differences in plaque amounts between Eln+/+Ldlr−/− and Eln+/−Ldlr−/− mice on WD (Supplemental Fig. I).

Figure 2.

Atherosclerotic plaque amounts are similar or reduced in Eln+/−Ldlr−/− mice on WD compared to Eln+/+Ldlr−/− mice on WD, despite high blood pressure and reduced aortic compliance in Eln+/−Ldlr−/− mice. Total plaque area at the level of the aortic valve is similar in Eln+/+Ldlr−/− and Eln+/−Ldlr−/− mice on WD (A), as quantified by Oil Red O staining of aortic root sections (B, C). N = 10 – 16/group. Normalized (norm) plaque area in the ascending aorta is reduced in Eln+/−Ldlr−/− mice on WD compared to Eln+/+Ldlr−/− mice on WD (D), as quantified by Oil Red O staining of en face preparations of the ascending aorta (E, F). N = 6 – 9/group. Significance between Eln+/+Ldlr−/− and Eln+/−Ldlr−/− mice on WD determined by student's t-test.

Cross-sections of the aortic root for Eln+/+Ldlr−/− and Eln+/−Ldlr−/− mice on WD were analyzed for plaque composition. Representative F4/80 images for macrophage amount and localization are shown in Figs. 3A and B. The plaques have a layer of macrophages on top of a necrotic core, with no differences in the amount of F4/80 positive cells (Fig. 3E). Fluorescence staining for elastin, collagen, SMCs, and cell nuclei for the same plaques are shown in Figs. 3C and D. A layer of SMCs is located on top of the necrotic core, where there are few cell nuclei. There are no differences in the amount of cells staining positively for αSMA (Fig. 3F). Collagen stains brightly in the valves and faintly under the SMCs. There is a significant increase in area of the plaque staining positively for collagen in Eln+/−Ldlr−/− mice on WD compared to Eln+/+Ldlr−/− mice on WD (p = 0.02) (Fig. 3G).

Figure 3.

Macrophage and SMC amounts are similar, but collagen amount is increased for aortic root plaques in Eln+/−Ldlr−/− mice on WD compared to Eln+/+Ldlr−/− mice on WD. F4/80 staining shows a layer of macrophages on top of a necrotic core (A, B). Fluorescent staining of the same plaque shows αSMA positive cells (red) on top of the necrotic core containing few cell nuclei (blue). Collagen (green) stains brightly in the valve tissue and shows patchy staining (arrows) below the layer of αSMA positive cells. Elastic fibers (magenta) are visible in the aortic wall. Quantification of the percent area of the plaque staining positively for F4/80 (E), αSMA (F), and collagen (G) shows an increase in collagen amounts for Eln+/−Ldlr−/− mice. N = 3 – 6/group.

Circulating cytokine levels are not affected by Eln genotype

Immunoassays were used to measure serum levels of inflammatory cytokines and TGF-β1 in Eln+/+Ldlr−/− and Eln+/−Ldlr−/− mice on ND and WD. IL6, IL10, CXCL1, and TNF are increased 140 – 600% with WD, but are not affected by Eln genotype (Table 2). TGF-β1 levels are not affected by diet or Eln genotype (Table 2).

Table 2.

Levels of inflammatory cytokines and TGF-β1 were measured using electrochemiluminescence immunoassays on serum from Eln+/+Ldlr−/− and Eln+/−Ldlr−/− mice and fed ND or WD for 16 weeks after weaning. IL6, IL10, CXCL1, and TNF levels are significantly affected by diet, but not Eln genotype. TGF-β1 levels are not affected by diet or Eln genotype. Significance was determined by two-way ANOVA for diet, Eln genotype and the interaction between diet and Eln genotype.

			inflammatory cytokines (pg/ml)				TGF-β1 (ng/ml)
Diet	Ldlr	Eln	IL6	IL10	CXCL1	TNF	TGF-β1
ND	Ldlr−/−	Eln+/+	12 ± 5	10 ± 2	29 ± 10	8 ± 2	39 ± 10
		Eln+/−	13 ± 3	10 ± 3	26 ± 8	7 ± 1	52 ± 12
WD	Ldlr−/−	Eln+/+	89 ± 28	29 ± 20	185 ± 160	23 ± 7	28 ± 30
		Eln+/−	86 ± 73	20 ± 6	202 ± 144	29 ± 15	37 ± 26
P value		Diet	0.021	0.005	0.019	0.007	0.332
		Eln	0.964	0.231	0.913	0.697	0.412
	Diet × Eln		0.956	0.274	0.873	0.579	0.879
N/group			2 – 6	2 – 6	2 – 6	2 – 6	2 – 6

Italics indicate P < 0.05. Data are presented as mean ± SD.

Elastic fibers are not fragmented in Eln+/− arteries

Cross-sections of the left common carotid artery were processed for histology to visualize the wall structure away from the plaque area of Eln+/+Ldlr−/− and Eln+/−Ldlr−/− mice on WD (Fig. 4). H&E staining shows similar cell density and orientation. VVG staining shows that Eln+/−Ldlr−/− arteries have thinner, more numerous elastic lamellae than Eln+/+Ldlr−/− arteries, with no evidence of increased elastic fiber fragmentation. PSR staining shows that collagen organization appears similar in Eln+/+Ldlr−/− and Eln+/−Ldlr−/− arteries.

Figure 4.

Representative histology images of carotid artery sections stained with H&E for cell nuclei (purple), Verhoeff Van Gieson (VVG) for elastic fibers (black) and picrosirius red (PSR) for collagen (red). Eln+/−Ldlr−/− mice on WD have thinner, more numerous elastic lamellae across the arterial wall than Eln+/+Ldlr−/− mice on WD, but the lamellae are not fragmented. N = 6 – 7/group. There is also no evidence of elastic fiber fragmentation at the level of the aortic valve under the plaque (Fig. 3).

Discussion

Eln+/+ and Eln+/− mice were bred with Ldlr+/+ and Ldlr−/− mice and fed ND or WD for 16 weeks after weaning to determine if hypertension and decreased aortic compliance in Eln+/−Ldlr−/− mice increases atherosclerotic plaque accumulation compared to Eln+/+Ldlr−/− mice. WD increases total cholesterol, triglyceride, LDL and HDL levels. Ldlr−/− genotype increases total cholesterol, triglycerides, and LDL levels. Total cholesterol and triglyceride changes are consistent with previous studies of Ldlr mice on WD^29–31. LDL and HDL levels are also consistent with previous results for Ldlr mice on WD³². The serum cholesterol measurements confirm that the combination of diet and Ldlr genotype has the expected results on Eln mice.

WD, Ldlr−/− genotype, and Eln+/− genotype increase systolic blood pressure and pulse pressure, with Eln+/− genotype having the most significant effect. WD has been shown to increase blood pressure in wild-type mice^{33, 34}. To our knowledge, Ldlr−/− genotype has not previously been associated with increases in systolic or pulse pressure. Eln+/− mice are known to have increased systolic and pulse pressure^{19, 35}. Our blood pressure results confirm previous studies and show that Eln+/−Ldlr−/− mice on WD have the highest systolic and pulse pressure of all groups. The increase in pulse pressure for Eln+/− mice is similar to the increase in pulse pressure for aging humans³⁶. In clinical studies, there are significant associations between high blood pressure and atherosclerotic plaque accumulation^3–6, but animal studies of relationships between hypertension and atherosclerosis have been contradictory³⁷.

One complication in animal studies is that modulation of the renin-angiotensin system (RAS) is often used to alter blood pressure. Ang II is the active end product of the RAS and has a range of effects on cells including encouraging proliferation and apoptosis and activating inflammatory and oxidative stress pathways³⁸. Infusing Ang II in Apoe−/− mice accelerates atherosclerosis³⁹, while inhibiting Ang II in Apoe−/− mice through pharmaceutical⁴⁰ or genetic⁴¹ approaches reduces plaque accumulation, independent of blood pressure lowering effects. When norepinephrine or Ang II are infused in Apoe−/− mice to obtain equivalent increases in blood pressure, there are much more modest increases in plaque amounts with norepinephrine than Ang II compared to control. However, norepinephrine still causes increased blood pressure and increased plaque deposition compared to control⁴². Previous studies showed that renin activity is high in Eln+/− mice, suggesting increased levels of Ang II, but aldosterone levels are normal, suggesting that downstream effects of Ang II may be mediated differentially in Eln+/− mice¹⁹. Increased blood pressure and increased renin activity in Eln+/− mice would be expected to increase atherosclerotic plaque accumulation in a hyperlipidemic model, but our results show that Eln+/−Ldlr−/− mice on WD have similar or reduced plaque amounts compared to controls.

Two recent studies investigate links between hypertension and atherosclerosis in animals without RAS activation. Vitamin D deficiency coupled with WD in Ldlr−/− and Apoe−/− mice increases blood pressure and increases atherosclerotic plaque accumulation⁴³. However, vitamin D deficiency also activates macrophage ER stress and accelerates atherosclerosis by inducing phenotypic changes in macrophages. Plaque accumulation in vitamin D deficiency is decreased by a chemical chaperone that decreases ER stress, without any effects on blood pressure, indicating that macrophage ER stress is the driving factor for plaque accumulation, not blood pressure. Compared to ND, a low sodium diet given to Apoe−/− mice results in normal blood pressure, increased inflammatory markers, and increased atherosclerotic plaque accumulation⁴⁴. Contrarily, high sodium diet given to Apoe−/− mice results in high blood pressure, decreased inflammatory markers and decreased atherosclerotic plaque accumulation, indicating that inflammatory markers are associated with atherosclerosis, but high blood pressure is not. The data from the vitamin D and sodium studies highlight the role of inflammation over hypertension as a factor in atherosclerosis susceptibility.

The results of the vitamin D and sodium studies are consistent with our findings that Eln+/−Ldlr−/− mice with hypertension, but no indication of abnormal inflammation in the aortic wall, do not have increased plaque accumulation compared to Eln+/+Ldlr−/− mice on WD. Serum levels of inflammatory cytokines, IL6, IL10, CXCL1, and TNF, which all play important roles in atherogenesis⁴⁵, are increased in Ldlr−/− mice on WD compared to ND, but are not different in Eln+/− mice compared to Eln+/+. Macrophage amounts and localization are not different in aortic root plaques in Eln+/+Ldlr−/− and Eln+/−Ldlr−/− mice on WD, indicating that the infiltration of inflammatory cells in the plaque is not affected by elastin haploinsufficiency and the resulting hypertension and reduced aortic compliance. Although SMC amounts are similar, collagen amounts are higher in aortic root plaques for Eln+/−Ldlr−/− mice on WD compared to Eln+/+Ldlr−/− mice on WD, indicating increased plaque stability².

Aortic compliance is significantly lower in Eln+/− mice than Eln+/+ at physiologic pressures, consistent with previous studies^{19, 23}. Diet and Ldlr genotype have no significant effects on aortic compliance. In B6D2F1 mice, WD decreases aortic compliance as measured by pulse wave velocity⁴⁶ and in vitro mechanical tests⁴⁷. Previous studies have found reduced aortic compliance in Ldlr−/− mice⁴⁸. Differences in the effects of diet and Ldlr genotype on aortic compliance may be due to differences in mouse strain, diet protocol, and measurement methods. Despite significantly decreased aortic compliance, Eln+/−Ldlr−/− mice on WD do not show increased atherosclerotic plaque accumulation or decreased plaque stability.

C1039G+/−Apoe−/− mice have decreased aortic compliance and increased plaque accumulation and decreased plaque stability, with no changes in blood pressure compared to control mice¹³. C1039G+/− mice have fragmented elastic fibers^{16, 17} and dysregulated TGF-β signaling¹⁵. Injecting elastin-derived peptides increases plaque accumulation in Apoe−/− mice on ND and Ldlr−/− mice on WD¹⁸. The effect is blocked in the absence of immune PI3Kγ, indicating that fragmentation of elastic fibers, release of elastin-derived peptides, and the subsequent immune response may contribute to increased plaque accumulation in C1039G+/−Apoe−/− mice. TGF-β has complex effects on atherosclerosis and immune reactivity in mouse and man^{49, 50}, but it is reasonable to believe that altered TGF-β regulation may also affect atherosclerosis accumulation in C1039G+/−Apoe−/− mice, independent of reductions in aortic compliance. C1039G+/− mice have higher levels of circulating TGF-β1 that increase with age and correlate with aortic root dilation⁵¹. We found no significant differences in circulating TGF-β1 levels between Eln+/+Ldlr−/− and Eln+/−Ldlr−/− mice on ND or WD. TGF-β signaling could be further investigated through quantification of gene and protein levels of TGF-β pathway members within the aortic wall at different time points, but our current results show no evidence of altered TGF-β signaling in Eln+/− mice.

In human studies, a 3 – 21% reduction in arterial compliance is observed in individuals with atherosclerotic plaques, compared to individuals with no detectable plaques^{7, 8}. The reduction in compliance is consistent with that observed in Eln+/− mice at systolic blood pressure. However, human studies generally include older individuals (mean age > 65)^{7, 8}. Elastin degrades and fragments with age, leading to decreased aortic compliance^{52, 53}. Because elastin degradation and decreased aortic compliance are intimately related, it is difficult to separate them in human studies with aging individuals or in mouse models with fragmented elastic fibers. Eln+/− mice are unique in that they have reduced aortic compliance, but no indication of elastic fiber fragmentation. Our data show that Eln+/−Ldlr−/− mice on WD do not have increased atherosclerotic plaque accumulation or instability compared to Eln+/+Ldlr−/− mice on WD. Our results suggest that additional insults, besides high blood pressure and low aortic compliance, such as activation of inflammatory pathways through elastic fiber fragmentation or TGF-β dysregulation, are necessary to alter atherosclerosis progression.

Limitations

Atherosclerosis in mice is different than in humans. However, mice allow us to investigate specific contributions to atherosclerosis in a controlled manner that cannot be done in human studies. We used Ldlr−/− mice as our atherogenic model and may have observed different results in another model. We measured plaque accumulation at one specific time point, so we may have missed differences in the time course of plaque progression that depend on blood pressure or aortic compliance. We examined plaque composition and cytokine levels in a subset of Eln+/+Ldlr−/− and Eln+/−Ldlr−/− mice on WD. More comprehensive analyses of plaque composition and progression and cytokine signaling are needed to determine if Eln+/− genotype affects characteristics of atherosclerotic disease other than plaque amounts. We measured anesthetized blood pressure, which will be different than ambulatory blood pressure. Although absolute pressure values vary between anesthetized and awake animals, the differences between groups are likely to remain. We suggest that additional insults, such as elastic fiber fragmentation, are necessary to alter atherosclerosis progression. Alternate mouse models, such as fibulin-5 knockout mice (Fbln5−/−), that have reduced aortic compliance, as well as fragmented elastic fibers^{54, 55} could be used to further investigate this idea.

Highlights.

• -Elastin insufficient (Eln+/−) mice have hypertension and reduced aortic compliance.
• -Eln+/−Ldlr−/− mice do not have more atherosclerotic plaque than Eln+/+Ldlr−/− mice.
• -Hypertension and reduced aortic compliance may not directly affect atherosclerosis.