Molecular Medicine

Altered Metabolism of LDL in the Arterial Wall Precedes Atherosclerosis RegressionNovelty and Significance

Emil D. Bartels, Christina Christoffersen, Marie W. Lindholm, Lars B. Nielsen
https://doi.org/10.1161/CIRCRESAHA.115.307182
Circulation Research. 2015;117:933-942
Originally published September 10, 2015

Jump to

Abstract

Rationale: Plasma cholesterol lowering is beneficial in patients with atherosclerosis. However, it is unknown how it affects entry and degradation of low-density lipoprotein (LDL) particles in the lesioned arterial wall.

Objective: We studied the effect of lipid-lowering therapy on LDL permeability and degradation of LDL particles in atherosclerotic aortas of mice by measuring the accumulation of iodinated LDL particles in the arterial wall.

Methods and Results: Cholesterol-fed, LDL receptor–deficient mice were treated with either an anti-Apob antisense oligonucleotide or a mismatch control antisense oligonucleotide once a week for 1 or 4 weeks before injection with preparations of iodinated LDL particles. The anti-Apob antisense oligonucleotide reduced plasma cholesterol by ≈90%. The aortic LDL permeability and degradation rates of newly entered LDL particles were reduced by ≈50% and ≈85% already after 1 week of treatment despite an unchanged pool size of aortic iodinated LDL particles. In contrast, the size, foam cell content, and aortic pool size of iodinated LDL particles of aortic atherosclerotic plaques were not reduced until after 4 weeks of treatment with the anti-Apob antisense oligonucleotide.

Conclusions: Improved endothelial barrier function toward the entry of plasma LDL particles and diminished aortic degradation of the newly entered LDL particles precede plaque regression.

Introduction

Atherosclerosis is a leading cause of disability and death.1 Plasma low-density lipoprotein (LDL) particles promote the development of atherosclerosis when they cross the arterial endothelium. The LDL particles can be trapped in the intima-inner media and cause inflammation.2,3 Intimal inflammation attracts monocytes from the blood stream. They transform into macrophages that engulf and degrade the LDL particles, resulting in foam cell formation, that is, a pathological hallmark of atherosclerosis.2 Although plasma cholesterol lowering is beneficial in patients with atherosclerosis, it is unknown how it may affect the barrier function of the endothelium and subsequent metabolism of LDL particles in atherosclerotic lesions.

In This Issue, see p 905

Editorial, see p 909

Previous studies of the kinetics of LDL particles in the arterial intima of humans and experimental animal models have shown that although the endothelial LDL permeability is increased in atherosclerotic lesions,4–6 most LDL particles that enter the arterial wall leave (efflux) without contributing to the growth of the atherosclerotic lesion. However, the residence time of the newly entered particles is increased in lesioned versus nonlesioned arterial intima.4 This contributes to an increased pool size of LDL particles in the arterial intima and likely to increased susceptibility of the arterial LDL particles to aggregate and undergo other modifications within the arterial wall,7–9 thus promoting the growth of the lesion. The latter observation is one important foundation for the response to retention hypothesis of atherosclerosis.10

The rates of entry and degradation of LDL particles in the arterial wall have previously been studied in the aorta of rabbits. Both aortic LDL permeability and degradation of LDL particles are increased in atherosclerotic lesions.3,11–14 Hence, atherogenesis is accompanied by local changes in the arterial wall, which favors accumulation and degradation of plasma-derived LDL particles. However, it is not known whether these changes are irreversible and may react to, for example, cholesterol-lowering therapy.

Statin treatment reduces the size and changes the composition of established atherosclerotic lesions in humans.15 The reduction in the size of the atherosclerotic lesions is only modest although cardiovascular morbidity is markedly reduced.16 Morphological evaluations of atherosclerotic lesions, for example, in cholesterol-fed rabbits, suggest that lowering of plasma cholesterol (ie, after withholding cholesterol feeding) leads to plaque remodeling with formation of more collagen-rich atherosclerotic lesions after 6 to 12 months, suggesting that the long-term effect of plasma cholesterol lowering is a more stable plaque.17–20 This is also supported by data in mice, where genetic ablation of hepatic MTTP expression in LDL receptor–deficient mice (LDLr−/−) reduced plasma very low-density lipoprotein (VLDL) and LDL and led to more stable, less inflamed plaques.21 More recent studies where aortic segments from atherosclerotic donor mice with high plasma VLDL/LDL levels were transplanted into healthy recipient mice with low plasma VLDL/LDL levels suggest that atherosclerotic plaques are highly dynamic. Hence, transplantation of lesioned aortas into mice with nonatherogenic plasma lipid profiles reduces macrophage content in the transplanted aorta content within few days.22 Also, the phenotype of the remaining plaque macrophages shifts toward a less inflammatory M2 phenotype.23–26 Such cellular changes could possibly affect the rate of cellular degradation of LDL particles in the vessel wall and as such the regression of the atherosclerotic plaques. A deeper understanding of the mechanisms of atherosclerosis regression is an increasingly important area with the introduction and regulatory approval of potent lipid-lowering drugs, for example, proprotein convertase subtilisin kexin-9 inhibitors.27

We recently used an anti-Apob antisense oligonucleotide (ASO), which effectively reduces hepatic Apob mRNA expression by >90%.28 This compound allows elimination of the plasma VLDL/LDL exposure of a diseased arterial wall without the need for transplantation of aortic segments or change of diet. The aim of this study was to investigate the effect of lowering VLDL/LDL cholesterol on the aortic LDL permeability and the degradation rates of newly entered LDL particles in the arterial wall of mice with pre-existing atherosclerotic lesions.

Methods

An expanded Methods section is available in the Online Data Supplement.

Animals

Ldlr−/− mice were kept on a high-cholesterol diet for 16 to 19 weeks before treated weekly with anti-Apob ASO targeting Apob mRNA or control ASO (5 mg/kg per week, Santaris Pharma29) for 1 or 4 weeks. The studies are outlined in Online Table I. All animal experiments were approved by the Danish Animal Experiments Inspectorate (Dyreforsoegstilsynet).

Isolation and Radioactive Labeling of LDL Particles Using Iodogen

Isolation and labeling of LDL particles with 125I or 131I were done essentially as previously described.30,31

LDL Permeability, Degradation, and Plasma Contamination

For estimation of arterial wall LDL permeability, mice (n=34) were injected with 125I-LDL in a tail vein 90 minutes before removal of the aorta. The same mice were also injected with 131I-LDL 47½ minutes before removal of the aorta (n=17; to assess whether the accumulation of LDL particles coupled to radiolabeled iodine [*I-LDL] increased linearly with time) or 5 minutes before removal of the aorta (n=17; to assess plasma contamination of the arterial tissue). Blood samples were drawn before injection and after 5, 45, 60, and 90 minutes.

For estimation of arterial wall LDL degradation, mice (n=31) were injected in a tail vein with 125I and 131I-TC/125I-LDL 23 to 25 hours before removal of aorta. Blood samples were drawn before injection and after 5 and 180 minutes and 23 to 25 hours.

Aliquots (10 µL) of plasma were precipitated with 15% trichloroacetic acid to remove free I and counted in a γ-counter, which allows the correction for 131I spillover into the 125I spectrum (1470 Automatic Counter, PerkinElmer Danmark A/S, Skovlunde, Denmark) for 10 minutes. The tissues were minced before precipitated with trichloroacetic acid and centrifuged. After centrifugation, the aortic radioactivity in the supernatant and precipitate was counted for 60 minutes.

Calculations

The calculations of the aortic LDL permeability and aortic degradation of radiolabeled LDL particles were corrected for the contamination of aortic tissue with plasma radioactivity.13,32

Tissue Samples

Before the last blood sample, the mice were anesthetized with an intraperitoneal injection of hypnorm/midazolam. Subsequently, tissue samples were obtained. The aorta was removed down to diaphragm, carefully cleaned from perivascular fat, and opened en face. The aortas were divided in an upper part (from the aortic root to the sixth rib) and a lower part (from the sixth rib to the diaphragm). Plaque surface areas were analyzed from photos taken with a Leica DFC290 digital camera mounted on a surgical microscope. The aortic intima-inner media was separated from the outer media and adventitia tissue using microforceps and a surgical microscope.

Biochemistry

Plasma cholesterol and triglycerides were determined with enzymatic methods (Chod-PAP; Roche, Denmark and GPO-Trinder; Sigma-Aldrich, Brondby, Denmark).

Size exclusion chromatography of mouse plasma pools (200 µL) was performed with a superose 6 column.33 The apo AI and apoB levels in pooled gel filtration fractions corresponding to VLDL, LDL, high-density lipoprotein (HDL), or protein were determined with Western blot.30

Tissue Lipid Analysis

Aortic free and esterified cholesterol content and triglyceride content were measured with quantitative thin layer chromatography after extraction with chloroform/methanol.34

Histological Analyses

Hearts were immersion-fixed in paraformaldehyde (4%) and embedded in Tissue-Tek OCT (Sakura Finetek, Denmark). The proximal part of the heart containing the aortic root was sectioned in 10-µm sections and mounted on Superfrost Plus microscope slides (Thermo Scientific, Hvidovre, Denmark). Sections from the aortic root were stained with oil red O (ORO) to visualize lipids, collagen (Massons trichrome), macrophages (with anti-CD68 [catalog number: ab53444; Abcam, United Kingdom] or anti-CD163 antibodies [a kind gift from Dr Søren Mostrup; University of Aarhus]), scavenger receptors (anti-MSR [MSR1/scavenger receptor class A; catalog number: NBP-1-00092; Novus Biologicals], anti-CD36 [catalog number: NB400-144; Novus Biologicals), and anti-SRBI [NB400-104; Novus Biologicals]), and nitrotyrosine (antinitrotyrosine; catalog number: ab53444; Abcam, United Kingdom).

mRNA Expression

mRNA expression in total thoracic aorta and in laser-microdissected macrophages was determined with real time polymerase chain reaction using the ABI 7900 HT sequence detection system (Applied Biosystems, Foster City, CA).

Statistical Analysis

GraphPad Prism 4.0 (GraphPad Software Inc, San Diego) was used for statistical analyses. Two-group comparisons were done with Student t test or Mann–Whitney as stated in table and figure legends. P≤0.05 was considered significant.

Results

Anti-Apob ASO Effectively Reduces Plasma Cholesterol in Ldlr−/− Mice

To induce atherosclerosis in aorta, Ldlr−/− mice (n=98) were fed a cholesterol-enriched diet for 16 weeks. Subsequently, while continuing the diet, the mice were randomly allocated to treatment with anti-Apob ASO or control ASO (5 mg/kg IP once weekly) for 1 or 4 weeks (Online Table I). Sixteen additional mice were fed the cholesterol-enriched diet for 19 weeks before treated with ASO for 1 week, and 6 mice were fed the diet for 20 weeks without receiving any ASO treatment (Online Table I).

Anti-Apob ASO treatment lead to a sustained ≈90% reduction of total plasma cholesterol in cholesterol-fed Ldlr−/− mice when compared with control ASO treatment (Table 1). The reduction of plasma cholesterol was achieved within 2 to 3 days (Online Figure I). Plasma triglycerides were also reduced by the anti-Apob ASO (Table 1) consistent with the reduction in plasma VLDL. Accordingly, on size exclusion chromatography, plasma VLDL and LDL cholesterol and apoB were essentially eliminated in the cholesterol-fed anti-Apob ASO–treated Ldlr−/− mice both after 1 and 4 weeks of treatment (Figure 1A and 1D). The HDL cholesterol appeared slightly reduced by 35% after 1 week of anti-Apob ASO treatment (Online Figure IIA), whereas no changes in HDL cholesterol peak height was seen after 4 weeks of treatment (Online Figure IIB). Thus, the HDL/LDL ratio was markedly increased in anti-Apob ASO compared with control ASO–treated mice after both 1 and 4 weeks of treatment. Body weight was similar in anti-Apob ASO and control ASO–treated mice (Table 1).

View this table:
Table 1.

Plasma Lipids and Body Weight After Treatment With Anti-Apob or Control ASO

Figure 1.
Figure 1.

Effect of anti-Apob antisense oligonucleotide (ASO) on plasma lipoproteins. Lipoprotein size distribution was determined with fast protein liquid chromatography (FPLC) after 1 week (A) and 4 weeks (C) of treatment with anti-Apob ASO (dotted line) or control ASO (full line). Western blots of pooled FPLC fractions for apoB48/100 and apo AI after 1 week (B) or 4 (D) weeks of ASO treatment (A and C). HDL indicates high-density lipoprotein; LDL, low-density lipoprotein; and VLDL, very low-density lipoprotein.

Effect of Anti-Apob ASO on Plaque Size and Composition

After 1 week of treatment, the size of the atherosclerotic lesions in aorta, as judged either by en face analysis of the aorta or in histological cross section of the proximal aorta, was not affected by the anti-Apob ASO (Figure 2A and 2B). Also, there was no effect of anti-Apob ASO on plaque lipid content (as judged by ORO staining and chemical measurements of aortic cholesterol; Online Figure IIIA), collagen content (as judged by trichrome staining; Figure 2C and 2D), or the size of the necrotic core (Online Figure IIIC). In contrast, immunohistochemical staining of sections from the aortic root showed a significant reduction in the nitrotyrosine-stained area (an indirect marker of modification by reactive nitrogen species) after 1 week of treatment with the anti-Apob ASO (Figure 3G). Nevertheless, the amount of subendothelial foam cells (Figure 3E and 3F), the mRNA expression of inflammatory genes in the luminal subendothelial foam cells (Online Table III), and the expression of the scavenger receptors MSR, SR-BI, and CD36 were unchanged after 1 week of the cholesterol-lowering treatment (Online Figure IVA–IVC).

Figure 2.
Figure 2.

Effect of anti-Apob antisense oligonucleotide (ASO) on atherosclerotic plaque size and composition after 1 week (A–D) or 4 weeks (E–H) of treatment. Plaque size was quantified both en face as surface coverage of the aorta (A and E) and as cross-sectional lesion area in the aortic root (B and F). C and G, Lipid staining was done with Oil red O. D and H, Collagen visualized with Masons trichrome. Open boxes: control ASO–treated mice. Full boxes: anti-Apob ASO–treated mice. Data are presented as median with individual data points. Two-group comparisons were done with Mann–Whitney test.

Figure 3.
Figure 3.

Effect of anti-Apob antisense oligonucleotide (ASO) treatment on macrophage foam cells in the aortic root. CD68-positive foam cells indicated with arrows in control ASO–treated mice (A) and anti-Apob ASO–treated mice (B) after 4 weeks of treatment. Trichrome staining of the aortic root after 4 weeks of treatment in control ASO–treated mice (C) and anti-Apob ASO–treated mice (D). The subendothelial foam cells are indicated with arrows. The cross-sectional area of foam cells was quantified in the mice after 1 week (E) and 4 (F) weeks of ASO treatment. Nitrotyrosine positive area in the aortic root after 1 week (G) and 4 weeks (H) of ASO treatment. Open boxes: control ASO–treated mice. Full boxes: anti-Apob ASO–treated mice. Data are presented as median with individual data points. Two-group comparisons were done with Mann–Whitney test.

After 4 weeks of treatment, the morphology of the lesions was clearly affected. Hence, the size of the lesions and the ORO-stained plaque lipids were reduced, and the plaques contained relatively more collagen in the anti-Apob ASO–treated mice when compared with control ASO–treated mice (Figure 2E–2H). Remarkably, the anti-Apob ASO–treated mice lacked, or only had few, CD68-positive cells in the luminal parts of the atherosclerotic lesions, whereas these cells were abundant in the control ASO–treated mice (Figure 3A and 3B). This morphological difference also appeared in trichrome-stained sections where subendothelial foam cells were scarce in the anti-Apob ASO–treated mice (Figure 3C and 3D). On morphometry, the amount of subendothelial foam cells was reduced by ≈70% (P<0.0001) in anti-Apob ASO compared with control ASO–treated mice (Figure 3E and 3F). The disappearance of the subluminal foam cells was accompanied by an 85% to 95% reduction in the mRNA expression of inflammatory genes in the aortic arch after 4 weeks of treatment (Table 2). Prototypical markers of M1 macrophages (Il-6, Inos, and Cd11c) and M2 macrophages (Cd206, Tgf-β, and Cd163) were reduced to the same extent. In contrast, the mRNA expression of Ccr7 and Tnf-α was unaffected by the anti-Apob ASO treatment (Table 2). Despite the reduction of foam cells, the aortic total cholesterol content (Online Figure IIB) and the size of the necrotic core (Online Figure IIID) were not significantly reduced after 4 weeks of treatment. Immunohistochemical staining of sections from the aortic root showed that the reduction in nitrotyrosine-stained area observed after 1 week was sustained after 4 weeks of treatment with the anti-Apob ASO (Figure 3H), whereas trends toward changes in the expression of the scavenger receptors MSR (P=0.10), CD36 (P=0.06), and SR-BI (P=0.06) did not reach statistical significance (Online Figure IVD–IVF).

View this table:
Table 2.

Expression of Inflammatory Genes in the Thoracic Aorta After Treatment With Anti-Apob or Control ASO for 4 Weeks

Anti-Apob ASO Reduces Aortic LDL Permeability

The effect of plasma cholesterol reduction on the aortic LDL permeability was assessed by measuring the entry of LDL particle–bound *I in the intima-inner media of the aortic arch 45 and 90 minutes on an intravenous injection of *I-LDL before the removal of the aorta (studies 1 and 3).

The entry of *I in aortic intima-inner media increased from 45 minutes to 90 minutes after both 1 and 4 weeks of ASO treatment (Figure 4A and 4B). This indicates that the *I-LDL entry divided by the exposure time for ≤90 minutes primarily reflects the unidirectional flux of *I-LDL into the arterial wall and hence the aortic endothelial LDL permeability (in nL cm−2 h−1, where nL cm−2 denotes nanoliters of plasma equivalents per square centimeter aortic surface area).32 The apparent larger entry of I-LDL from 45 to 90 minutes than from 0 to 45 minutes could reflect a lag in entry of I-LDL particularly in the anti-Apob ASO–treated mice. Notably, there was no difference in the rate of disappearance of *I-LDL from the plasma between anti-Apob ASO and control ASO–treated mice (Online Figure VA and VB). Using the entry of *I-LDL during 90 minutes to calculate LDL permeability, the anti-Apob ASO caused a ≈50% reduction of the aortic LDL permeability from 24±4 to 12±2 nL cm−2 h−1 (P=0.03) after 1 week of treatment and from 34±6 to 18±3 nL cm−2 h−1 (P=0.01) after 4 weeks of treatment. The permeability of the vessel wall toward plasma proteins was also investigated by intravenous injection of Evans blue. After 1 week of treatment with the anti-Apob ASO, there was no difference in Evans blue accumulation in aorta between anti-Apob ASO and control ASO–treated mice (data not shown). Thus, the data suggest that plasma cholesterol lowering rapidly improves the barrier function of the aortic endothelium toward the entry of plasma LDL particles but have no or less effect on the entry of albumin-sized particles in the aortic vessel wall.

Figure 4.
Figure 4.

Entry of labeled low-density lipoprotein (LDL) in aortic intima-inner media after 1 week (A) or 4 weeks (B) of treatment. Accumulation of *I-LDL in aorta was measured 45 minutes (n=3–4 mice per group) and 90 minutes (n=7–9 mice per group) after injection of *I-LDL. Open boxes: control antisense oligonucleotide (ASO)–treated mice. Full boxes: anti-Apob ASO–treated mice. Data are presented as medians with interquartile ranges. *I indicates radiolabeled iodine.

Effect of Anti-Apob ASO on Aortic Pool Size of Undegraded LDL and LDL Degradation

The effect of the anti-Apob ASO on aortic LDL accumulation and LDL degradation was assessed by measuring the accumulation of *I-tyramine cellobiose (*I-TC) and *I in aortic intima-inner media during 24 hours after intravenous injection of *I-TC/*I-LDL (studies 4 and 5). The plasma decay of the labeled LDL particles during 24 hours was similar in anti-Apob ASO and control ASO–treated mice (Online Figure VC and VD).

The content of undegraded *I-LDL in the intima-inner media divided by the plasma concentration of *I-LDL at 24 hours provides an estimate of the pool size of undegraded LDL particles in the aortic intima-inner media.35 The pool size of undegraded LDL particles did not differ between anti-Apob ASO and control ASO–treated mice after 1 week (Figure 5A) but was reduced by ≈55% (P=0.03) in the anti-Apob ASO–treated mice after 4 weeks of treatment (Figure 5B).

Figure 5.
Figure 5.

Effect of anti-Apob antisense oligonucleotide (ASO) on pool size and degradation of *I/*I-TC-LDL in aortic intima-inner media after 1 week (A and C) or 4 weeks (B and D) of treatment. Pool size (A and B) and degradation of *I/*I-TC-LDL (C and D) were measured after 24 hours of exposure to double-labeled LDL assuming full equilibrium of labeled LDL between plasma and aorta at this time point. Open boxes: control ASO–treated mice. Full boxes: anti-Apob ASO–treated mice. Data are presented as median with individual data points. Two-group comparisons were done with Mann–Whitney test. *I indicates radiolabeled iodine; LDL, low-density lipoprotein; and TC, tyramine cellobiose.

The aortic LDL degradation was reduced from 1.7±0.21 nL cm−2 h−1 in control ASO–treated mice to 0.24±0.15 nL cm−2 h−1 in the anti-Apob ASO–treated mice (P=0.0006) already after 1 week of treatment. After 4 weeks of treatment with the anti-Apob ASO, LDL degradation was essentially eliminated from 0.73±0.07 nL cm−2 h−1 in control ASO–treated mice to 0.02±0.01 nL cm−2 h−1 in anti-Apob ASO–treated mice (P=0.0003; Figure 5C and 5D).

Effect of Anti-Apob ASO on Pinocytic Uptake of LDL Particles

To assess whether the sudden changes in LDL degradation on acute lipid-lowering treatment could be attributed to changes in macrophage pinocytosis, the nonreceptor-mediated uptake of LDL-sized particles by arterial wall macrophages was assessed by intravenous injection of pegylated fluorescent nanoparticles. After 1 week of anti-Apob ASO or control ASO treatment, there was no difference in the plaque accumulation of pegylated nanoparticles (Online Figure VI).

Discussion

Plasma LDL cholesterol fuels atherogenesis, and reduction of plasma LDL cholesterol improves the outcome of cardiovascular disease.36 However, little is known about the metabolism of LDL in the arterial wall during cholesterol-lowering treatment and atherosclerosis regression. The main results of this study are that lowering of plasma cholesterol leads to a pronounced reduction in the LDL entry in the arterial wall and in the propensity of the newly entered LDL particle to be degraded in the arterial wall, which precedes changes in the cellular composition and size of atherosclerotic lesions.

Using an anti-Apob ASO to lower plasma cholesterol, we achieved an essential elimination of plasma VLDL and LDL in Ldlr−/− mice with pre-existing atherosclerosis. Notably, this change occurred without changing the cholesterol-enriched diet. Within 4 weeks, atherosclerotic lesions become smaller and more collagen rich, suggesting plaque stabilization, which is in accordance with previous results from studies in cholesterol-fed rabbits.17–20 The sizes of the necrotic core and the aortic total cholesterol content were unchanged, possibly reflecting that the major pool of extracellular cholesterol is relatively resistant to efflux in the 4-week timespan studied, although the foam cell content was highly dynamic and disappeared already after 4 weeks of treatment. The discordance between the unchanged cholesterol levels in the thoracic aorta measured by thin layer chromatography and the reduced ORO staining in the aortic root of anti-Apob ASO–treated mice could, at least in part, be caused by changes in the hydrolysis/re-esterification cycle of cholesteryl esters in macrophages37 because unesterified cholesterol is not detected with ORO. The mass influx of LDL cholesterol is the product of the plasma LDL cholesterol concentration and the endothelial LDL permeability.38 Thus, any reduction of plasma VLDL and LDL is paralleled by a reduction in mass influx of cholesterol into the arterial wall. This study suggests that a reduction of the endothelial LDL permeability also contributes to an even further reduced mass influx of proatherogenic lipoproteins after cholesterol lowering. Hence, the aortic LDL permeability was reduced already 1 week after commencing anti-Apob ASO treatment, apparently without any changes in the general permeability toward albumin-sized plasma proteins. This finding suggests that the endothelial LDL permeability is specifically improved before the lesion size is reduced. Notably, the LDL permeability of the atherosclerotic murine aorta was comparable with what has previously been observed in rabbits, man, and pigs,5,6,11,39 supporting the validity of the present results in mice.

In addition to reducing the influx of *I-LDL, cholesterol lowering also lead to an 86% decline in fraction of newly entered *I-LDL that was degraded in the intima-inner media 1 week after commencement of anti-Apob ASO treatment. Notably, the reduction in the fraction of newly entered *I-LDL that was degraded within the arterial intima-inner media was more substantial than the reduction in influx of *I-LDL and occurred in the setting of an unchanged size of the pool of intact *I-LDL within the arterial wall. Because the macrophage is the primary cell type responsible for uptake and degradation of LDL particles,40 this result implies that the likelihood of a newly entered LDL particle to be taken up and degraded by arterial macrophage foam cells is dramatically reduced on lowering of plasma cholesterol. Remarkably, this was not because of the absence of macrophage foam cells in the arterial wall after 1 week. Instead, the result could reflect that cholesterol lowering lessens modifications of proteins by reactive nitrogen species within the arterial wall as suggested by decreased nitrotyrosine immune staining. Although the present immunohistochemical data on nitrosylation do not ascertain modifications of apoB per se, a decrease in LDL modifications within the arterial wall—and hence a decrease in cellular uptake—would be a plausible mechanism for our finding of lower degradation rates of newly entered LDL particles after anti-Apob ASO. Our results do not support the other major mechanistic possibility, namely, that the phenotype of the macrophages might have changed toward a cell type with less propensity to perform receptor-dependent or receptor-independent uptake of arterial LDL particles because we did not detect changes in the scavenger receptor protein expression, pinocytosis of nanoparticles, or even macrophage mRNA expression profiles.

After 4 weeks of treatment with anti-Apob ASO, the aortic degradation of *I-LDL was almost abolished, despite the fact that the reduction of the size of the arterial pool of *I-LDL was only 65%. This decline in degradation of newly entered LDL particles was paralleled by a massive reduction of subendothelial CD68-positive foam cells, which were abundant in the plaques of the control mice. The loss of macrophages was paralleled by a pronounced reduction in the mRNA content of macrophage-denominating inflammatory genes. Previous reports of gene expression in CD68-positive cells dissected from atherosclerotic plaques undergoing regression have suggested a prominent role for CCR7 in the process.24 We did not detect any difference in total aortic mRNA expression of CCR7 between mice treated 4 weeks with the anti-Apob and the control ASO. But because the number of foam cells was markedly reduced at this time point, we suspect that the per cell expression of CCR-7 mRNA likely is increased on cholesterol lowering.

Despite similar changes in plaque composition, the changes in total aortic gene expression after 4 weeks of ASO treatment and the lack of changes in gene expression in laser-microdissected foam cell areas after 1 week of ASO treatment are in some contrast with previous studies of murine atherosclerosis regression. Hence, a selective induction of M2 markers and a downregulation of M1 markers have been seen already a few days after lipid-lowering interventions.26 Notably, in our model, we do not see an increase in plasma HDL accompanying the LDL-lowering treatment. Increases of HDL as seen in previous regression studies41 could affect both speed of regression, unloading of cellular cholesterol and the inflammatory phenotype of plaque macrophages.

In this study, the loss of foam cells was not accompanied by increased size of the necrotic core, implying that the foam cells may have left the arterial wall. Trogan et al42 demonstrated that foam cells indeed leave a transplanted arterial wall at an increased rate on exposure to a low plasma cholesterol environment. Hence, the disappearance of the foam cells as seen in this study might reflect egress to the blood.

The findings of an unchanged size of the necrotic core and similar levels of total aortic cholesterol content in anti-Apob ASO compared with control ASO–treated mice could imply that the majority of the cholesterol in atherosclerotic lesions is not easily mobilized in the 4-week timespan of this study despite marked changes in plasma lipid levels. This is in line with previous studies of atheroma regression in rhesus monkeys with established atherosclerotic disease, where the cholesterol content in coronary artery was significantly elevated despite regression of plaque size after >3 years on a low cholesterol regression diet.43

In conclusion, this study provides new insight into fundamental aspects of atherosclerosis regression. Hence, both improved endothelial barrier function toward plasma LDL particles and decreased propensity of newly entered plasma LDL particles to become degraded precedes morphological lesion regression with the loss of foam cells and markedly reduced lesion inflammation.

Acknowledgments

We thank Lis Nielsen, Maria Kristensen, Karen Rasmussen, and Birgitte E.S. Nielsen for excellent technical assistance and Connie Bundgaard for language editing. We also acknowledge the Core Facility for Integrated Microscopy, Faculty of Health and Medical Sciences, University of Copenhagen, for providing equipment and supervision.

Sources of Funding

This study received grant support from the Danish Research Council (C. Christoffersen), Novo Nordisk Foundation (E.D. Bartels and C. Christoffersen), the A.P Møller Foundation for the Advancement of Medical Science, Copenhagen, Denmark (E.D. Bartels and L.B. Nielsen), and Rigshospitalet (L.B. Nielsen).

Disclosures

M.W. Lindholm is an employee of Roche Innovation Center Copenhagen, Hoersholm, Denmark. The other authors report no conflicts.

Footnotes

  • In August 2015, the average time from submission to first decision for all original research papers submitted to Circulation Research was 14.31 days

  • The online-only Data Supplement is available with this article at http://circres.ahajournals.org/lookup/suppl/doi:10.1161/CIRCRESAHA.115.307182/-/DC1.

  • Nonstandard Abbreviations and Acronyms
    ASO
    antisense oligonucleotide
    HDL
    high-density lipoprotein
    *I
    radiolabeled iodine
    *I-TC
    *I-tyramine cellobiose
    LDL
    low-density lipoprotein
    ORO
    oil red O
    VLDL
    very low-density lipoprotein

  • Received July 9, 2015.
  • Revision received September 7, 2015.
  • Accepted September 10, 2015.

References

  1. 1.↵
    1. Nguyen C,
    2. Cheng-Lai A
    . The polypill: a proposed global solution to cardiovascular disease. Cardiol Rev. 2012;21:49–54. doi: 10.1097/CRD.0b013e3182755429.
    OpenUrl
  2. 2.↵
    1. Tabas I,
    2. Williams KJ,
    3. Borén J
    . Subendothelial lipoprotein retention as the initiating process in atherosclerosis: update and therapeutic implications. Circulation. 2007;116:1832–1844. doi: 10.1161/CIRCULATIONAHA.106.676890.
    OpenUrlAbstract/FREE Full Text
  3. 3.↵
    1. Schwenke DC,
    2. Carew TE
    . Initiation of atherosclerotic lesions in cholesterol-fed rabbits. I. Focal increases in arterial LDL concentration precede development of fatty streak lesions. Arteriosclerosis. 1989;9:895–907.
    OpenUrlAbstract/FREE Full Text
  4. 4.↵
    1. Nielsen LB,
    2. Stender S,
    3. Jauhiainen M,
    4. Nordestgaard BG
    . Preferential influx and decreased fractional loss of lipoprotein(a) in atherosclerotic compared with nonlesioned rabbit aorta. J Clin Invest. 1996;98:563–571. doi: 10.1172/JCI118824.
    OpenUrlCrossRefPubMed
  5. 5.↵
    1. Schwenke DC,
    2. Carew TE
    . Initiation of atherosclerotic lesions in cholesterol-fed rabbits. II. Selective retention of LDL vs. selective increases in LDL permeability in susceptible sites of arteries. Arteriosclerosis. 1989;9:908–918.
    OpenUrlAbstract/FREE Full Text
  6. 6.↵
    1. Nielsen LB,
    2. Grønholdt ML,
    3. Schroeder TV,
    4. Stender S,
    5. Nordestgaard BG
    . In vivo transfer of lipoprotein(a) into human atherosclerotic carotid arterial intima. Arterioscler Thromb Vasc Biol. 1997;17:905–911.
    OpenUrlAbstract/FREE Full Text
  7. 7.↵
    1. Devlin CM,
    2. Leventhal AR,
    3. Kuriakose G,
    4. Schuchman EH,
    5. Williams KJ,
    6. Tabas I
    . Acid sphingomyelinase promotes lipoprotein retention within early atheromata and accelerates lesion progression. Arterioscler Thromb Vasc Biol. 2008;28:1723–1730. doi: 10.1161/ATVBAHA.108.173344.
    OpenUrlAbstract/FREE Full Text
  8. 8.↵
    1. Bostrom MA,
    2. Boyanovsky BB,
    3. Jordan CT,
    4. Wadsworth MP,
    5. Taatjes DJ,
    6. de Beer FC,
    7. Webb NR
    . Group v secretory phospholipase A2 promotes atherosclerosis: evidence from genetically altered mice. Arterioscler Thromb Vasc Biol. 2007;27:600–606. doi: 10.1161/01.ATV.0000257133.60884.44.
    OpenUrlAbstract/FREE Full Text
  9. 9.↵
    1. Gustafsson M,
    2. Levin M,
    3. Skålén K,
    4. Perman J,
    5. Fridén V,
    6. Jirholt P,
    7. Olofsson SO,
    8. Fazio S,
    9. Linton MF,
    10. Semenkovich CF,
    11. Olivecrona G,
    12. Borén J
    . Retention of low-density lipoprotein in atherosclerotic lesions of the mouse: evidence for a role of lipoprotein lipase. Circ Res. 2007;101:777–783. doi: 10.1161/CIRCRESAHA.107.149666.
    OpenUrlAbstract/FREE Full Text
  10. 10.↵
    1. Williams KJ,
    2. Tabas I
    . The response-to-retention hypothesis of early atherogenesis. Arterioscler Thromb Vasc Biol. 1995;15:551–561.
    OpenUrlFREE Full Text
  11. 11.↵
    1. Nielsen LB,
    2. Nordestgaard BG,
    3. Stender S,
    4. Kjeldsen K
    . Aortic permeability to LDL as a predictor of aortic cholesterol accumulation in cholesterol-fed rabbits. Arterioscler Thromb. 1992;12:1402–1409.
    OpenUrlAbstract/FREE Full Text
  12. 12.↵
    1. Nordestgaard BG,
    2. Nielsen LB
    . Atherosclerosis and arterial influx of lipoproteins. Curr Opin Lipidol. 1994;5:252–257.
    OpenUrlCrossRefPubMed
  13. 13.↵
    1. Juul K,
    2. Nielsen LB,
    3. Munkholm K,
    4. Stender S,
    5. Nordestgaard BG
    . Oxidation of plasma low-density lipoprotein accelerates its accumulation and degradation in the arterial wall in vivo. Circulation. 1996;94:1698–1704.
    OpenUrlAbstract/FREE Full Text
  14. 14.↵
    1. Nielsen LB
    . Transfer of low density lipoprotein into the arterial wall and risk of atherosclerosis. Atherosclerosis. 1996;123:1–15.
    OpenUrlCrossRefPubMed
  15. 15.↵
    1. Kaneda H,
    2. Terashima M,
    3. Yamaguchi H
    . The role of intravascular ultrasound in the determination of progression and regression of coronary artery disease. Curr Atheroscler Rep. 2012;14:175–185. doi: 10.1007/s11883-012-0234-3.
    OpenUrlCrossRefPubMed
  16. 16.↵
    1. Brown BG,
    2. Zhao XQ,
    3. Sacco DE,
    4. Albers JJ
    . Lipid lowering and plaque regression. New insights into prevention of plaque disruption and clinical events in coronary disease. Circulation. 1993;87:1781–1791.
    OpenUrlAbstract/FREE Full Text
  17. 17.↵
    1. Weber G,
    2. Fabbrini P,
    3. Capaccioli E,
    4. Resi L
    . Repair of early cholesterol-induced aortic lesions in rabbits after withdrawal from short-term atherogenic diet. Scanning electron-microscopical (SEM) and transmission electron-microscopical (TEM) observations. Atherosclerosis. 1975;22:565–572.
    OpenUrlCrossRefPubMed
  18. 18.↵
    1. Wilson RB,
    2. Miller RA,
    3. Middleton CC,
    4. Kinden D
    . Atherosclerosis in rabbits fed a low cholesterol diet for five years. Arteriosclerosis. 1982;2:228–241.
    OpenUrlAbstract/FREE Full Text
  19. 19.↵
    1. Daugherty A,
    2. Schonfeld G,
    3. Sobel BE,
    4. Lange LG
    . Metabolism of very low density lipoproteins after cessation of cholesterol feeding in rabbits. A factor potentially contributing to the slow regression of atheromatous plaques. J Clin Invest. 1986;77:1108–1115. doi: 10.1172/JCI112409.
    OpenUrlCrossRefPubMed
  20. 20.↵
    1. Kockx MM,
    2. De Meyer GR,
    3. Buyssens N,
    4. Knaapen MW,
    5. Bult H,
    6. Herman AG
    . Cell composition, replication, and apoptosis in atherosclerotic plaques after 6 months of cholesterol withdrawal. Circ Res. 1998;83:378–387.
    OpenUrlAbstract/FREE Full Text
  21. 21.↵
    1. Feig JE,
    2. Parathath S,
    3. Rong JX,
    4. Mick SL,
    5. Vengrenyuk Y,
    6. Grauer L,
    7. Young SG,
    8. Fisher EA
    . Reversal of hyperlipidemia with a genetic switch favorably affects the content and inflammatory state of macrophages in atherosclerotic plaques. Circulation. 2011;123:989–998. doi: 10.1161/CIRCULATIONAHA.110.984146.
    OpenUrlAbstract/FREE Full Text
  22. 22.↵
    1. Reis ED,
    2. Li J,
    3. Fayad ZA,
    4. Rong JX,
    5. Hansoty D,
    6. Aguinaldo JG,
    7. Fallon JT,
    8. Fisher EA
    . Dramatic remodeling of advanced atherosclerotic plaques of the apolipoprotein E-deficient mouse in a novel transplantation model. J Vasc Surg. 2001;34:541–547. doi: 10.1067/mva.2001.115963.
    OpenUrlCrossRefPubMed
  23. 23.↵
    1. Feig JE,
    2. Shang Y,
    3. Rotllan N,
    4. Vengrenyuk Y,
    5. Wu C,
    6. Shamir R,
    7. Torra IP,
    8. Fernandez-Hernando C,
    9. Fisher EA,
    10. Garabedian MJ
    . Statins promote the regression of atherosclerosis via activation of the CCR7-dependent emigration pathway in macrophages. PLoS One. 2011;6:e28534. doi: 10.1371/journal.pone.0028534.
    OpenUrlCrossRefPubMed
  24. 24.↵
    1. Trogan E,
    2. Feig JE,
    3. Dogan S,
    4. Rothblat GH,
    5. Angeli V,
    6. Tacke F,
    7. Randolph GJ,
    8. Fisher EA
    . Gene expression changes in foam cells and the role of chemokine receptor CCR7 during atherosclerosis regression in ApoE-deficient mice. Proc Natl Acad Sci U S A. 2006;103:3781–3786. doi: 10.1073/pnas.0511043103.
    OpenUrlAbstract/FREE Full Text
  25. 25.↵
    1. Llodrá J,
    2. Angeli V,
    3. Liu J,
    4. Trogan E,
    5. Fisher EA,
    6. Randolph GJ
    . Emigration of monocyte-derived cells from atherosclerotic lesions characterizes regressive, but not progressive, plaques. Proc Natl Acad Sci U S A. 2004;101:11779–11784. doi: 10.1073/pnas.0403259101.
    OpenUrlAbstract/FREE Full Text
  26. 26.↵
    1. Feig JE,
    2. Vengrenyuk Y,
    3. Reiser V,
    4. Wu C,
    5. Statnikov A,
    6. Aliferis CF,
    7. Garabedian MJ,
    8. Fisher EA,
    9. Puig O
    . Regression of atherosclerosis is characterized by broad changes in the plaque macrophage transcriptome. PLoS One. 2012;7:e39790. doi: 10.1371/journal.pone.0039790.
    OpenUrlCrossRefPubMed
  27. 27.↵
    1. Cannon CP,
    2. Cariou B,
    3. Blom D,
    4. McKenney JM,
    5. Lorenzato C,
    6. Pordy R,
    7. Chaudhari U,
    8. Colhoun HM
    ; ODYSSEY COMBO II Investigators. Efficacy and safety of alirocumab in high cardiovascular risk patients with inadequately controlled hypercholesterolaemia on maximally tolerated doses of statins: the ODYSSEY COMBO II randomized controlled trial. Eur Heart J. 2015;36:1186–1194. doi: 10.1093/eurheartj/ehv028.
    OpenUrlAbstract/FREE Full Text
  28. 28.↵
    1. Krzystanek M,
    2. Pedersen TX,
    3. Bartels ED,
    4. Kjaehr J,
    5. Straarup EM,
    6. Nielsen LB
    . Expression of apolipoprotein B in the kidney attenuates renal lipid accumulation. J Biol Chem. 2010;285:10583–10590. doi: 10.1074/jbc.M109.078006.
    OpenUrlAbstract/FREE Full Text
  29. 29.↵
    1. Straarup EM,
    2. Fisker N,
    3. Hedtjärn M,
    4. Lindholm MW,
    5. Rosenbohm C,
    6. Aarup V,
    7. Hansen HF,
    8. Ørum H,
    9. Hansen JB,
    10. Koch T
    . Short locked nucleic acid antisense oligonucleotides potently reduce apolipoprotein B mRNA and serum cholesterol in mice and non-human primates. Nucleic Acids Res. 2010;38:7100–7111. doi: 10.1093/nar/gkq457.
    OpenUrlAbstract/FREE Full Text
  30. 30.↵
    1. Christoffersen C,
    2. Pedersen TX,
    3. Gordts PL,
    4. Roebroek AJ,
    5. Dahlbäck B,
    6. Nielsen LB
    . Opposing effects of apolipoprotein m on catabolism of apolipoprotein B-containing lipoproteins and atherosclerosis. Circ Res. 2010;106:1624–1634. doi: 10.1161/CIRCRESAHA.109.211086.
    OpenUrlAbstract/FREE Full Text
  31. 31.↵
    1. Christoffersen C,
    2. Ahnström J,
    3. Axler O,
    4. Christensen EI,
    5. Dahlbäck B,
    6. Nielsen LB
    . The signal peptide anchors apolipoprotein M in plasma lipoproteins and prevents rapid clearance of apolipoprotein M from plasma. J Biol Chem. 2008;283:18765–18772. doi: 10.1074/jbc.M800695200.
    OpenUrlAbstract/FREE Full Text
  32. 32.↵
    1. Nielsen LB,
    2. Nordestgaard BG,
    3. Stender S,
    4. Niendorf A,
    5. Kjeldsen K
    . Transfer of lipoprotein(a) and LDL into aortic intima in normal and in cholesterol-fed rabbits. Arterioscler Thromb Vasc Biol. 1995;15:1492–1502.
    OpenUrlAbstract/FREE Full Text
  33. 33.↵
    1. Christoffersen C,
    2. Jauhiainen M,
    3. Moser M,
    4. Porse B,
    5. Ehnholm C,
    6. Boesl M,
    7. Dahlbäck B,
    8. Nielsen LB
    . Effect of apolipoprotein M on high density lipoprotein metabolism and atherosclerosis in low density lipoprotein receptor knock-out mice. J Biol Chem. 2008;283:1839–1847. doi: 10.1074/jbc.M704576200.
    OpenUrlAbstract/FREE Full Text
  34. 34.↵
    1. Pedersen TX,
    2. Bro S,
    3. Andersen MH,
    4. Etzerodt M,
    5. Jauhiainen M,
    6. Moestrup S,
    7. Nielsen LB
    . Effect of treatment with human apolipoprotein A-I on atherosclerosis in uremic apolipoprotein-E deficient mice. Atherosclerosis. 2009;202:372–381. doi: 10.1016/j.atherosclerosis.2008.04.041.
    OpenUrlCrossRefPubMed
  35. 35.↵
    1. Nielsen LB,
    2. Stender S,
    3. Kjeldsen K,
    4. Nordestgaard BG
    . Specific accumulation of lipoprotein(a) in balloon-injured rabbit aorta in vivo. Circ Res. 1996;78:615–626.
    OpenUrlAbstract/FREE Full Text
  36. 36.↵
    1. Gutierrez J,
    2. Ramirez G,
    3. Rundek T,
    4. Sacco RL
    . Statin therapy in the prevention of recurrent cardiovascular events: a sex-based meta-analysis. Arch Intern Med. 2012;172:909–919. doi: 10.1001/archinternmed.2012.2145.
    OpenUrlCrossRefPubMed
  37. 37.↵
    1. Brown MS,
    2. Ho YK,
    3. Goldstein JL
    . The cholesteryl ester cycle in macrophage foam cells. Continual hydrolysis and re-esterification of cytoplasmic cholesteryl esters. J Biol Chem. 1980;255:9344–9352.
    OpenUrlFREE Full Text
  38. 38.↵
    1. Nordestgaard BG,
    2. Tybjaerg-Hansen A,
    3. Lewis B
    . Influx in vivo of low density, intermediate density, and very low density lipoproteins into aortic intimas of genetically hyperlipidemic rabbits. Roles of plasma concentrations, extent of aortic lesion, and lipoprotein particle size as determinants. Arterioscler Thromb. 1992;12:6–18.
    OpenUrlAbstract/FREE Full Text
  39. 39.↵
    1. Nordestgaard BG,
    2. Hjelms E,
    3. Stender S,
    4. Kjeldsen K
    . Different efflux pathways for high and low density lipoproteins from porcine aortic intima. Arteriosclerosis. 1990;10:477–485.
    OpenUrlAbstract/FREE Full Text
  40. 40.↵
    1. Steinberg D,
    2. Pittman RC,
    3. Carew TE
    . Mechanisms involved in the uptake and degradation of low density lipoprotein by the artery wall in vivo. Ann N Y Acad Sci. 1985;454:195–206.
    OpenUrlCrossRefPubMed
  41. 41.↵
    1. Feig JE,
    2. Rong JX,
    3. Shamir R,
    4. Sanson M,
    5. Vengrenyuk Y,
    6. Liu J,
    7. Rayner K,
    8. Moore K,
    9. Garabedian M,
    10. Fisher EA
    . HDL promotes rapid atherosclerosis regression in mice and alters inflammatory properties of plaque monocyte-derived cells. Proc Natl Acad Sci U S A. 2011;108:7166–7171. doi: 10.1073/pnas.1016086108.
    OpenUrlAbstract/FREE Full Text
  42. 42.↵
    1. Trogan E,
    2. Fayad ZA,
    3. Itskovich VV,
    4. Aguinaldo JG,
    5. Mani V,
    6. Fallon JT,
    7. Chereshnev I,
    8. Fisher EA
    . Serial studies of mouse atherosclerosis by in vivo magnetic resonance imaging detect lesion regression after correction of dyslipidemia. Arterioscler Thromb Vasc Biol. 2004;24:1714–1719. doi: 10.1161/01.ATV.0000139313.69015.1c.
    OpenUrlAbstract/FREE Full Text
  43. 43.↵
    1. Armstrong ML,
    2. Megan MB
    . Lipid depletion in atheromatous coronary arteries in rhesus monkeys after regression diets. Circ Res. 1972;30:675–680.
    OpenUrlAbstract/FREE Full Text

Novelty and Significance

What Is Known?

  • Cholesterol lowering reduces the size of pre-existing atherosclerotic lesions only modestly.

  • Nevertheless, cardiovascular morbidity is markedly reduced.

What New Information Does This Article Contribute?

  • Cholesterol lowering in mice rapidly reduced the entry into the aorta wall and subsequent cellular degradation of low-density lipoprotein (LDL) particles in mice with pre-existing atherosclerosis.

  • Hence, reduced intra-aortic metabolism of LDL particles precedes morphological regression of pre-existing atherosclerosis on lowering of plasma cholesterol.

Atherosclerosis is a common cause of cardiovascular disease. Plasma LDL particles promote atherosclerosis by entering the arterial wall where the LDL particles can be modified and cause inflammation and foam cell formation. Although it is well understood how lowering of plasma cholesterol prevents development of atherosclerotic lesions, much less is known about molecular, cellular, or physiological effects on pre-existing atherosclerotic lesions. This study showed that aggressive plasma cholesterol–lowering treatment immediately improved the barrier function of the aortic endothelium toward plasma LDL particles in mice. Moreover, the degradation rate of newly entered LDL particles by cells in aorta was dramatically reduced. This profound change in the metabolism of a newly entered LDL particle by the arterial wall occurred already after 1 week of treatment and before any morphological signs of atherosclerosis regression. The data may imply that within arteries with pre-existing atherosclerosis, either modifications of LDL particles or the ability of foam cells to take up and degrade extracellular LDL particles are reduced when plasma LDL particles are lowered. Further delineation of the molecular and cellular mechanisms may point at new targets for treating pre-existing atherosclerosis at the level of the arterial wall.

View Abstract

This Issue

Jump to

Article Tools

Share this Article

  • Email
  • Altered Metabolism of LDL in the Arterial Wall Precedes Atherosclerosis RegressionNovelty and Significance
    Emil D. Bartels, Christina Christoffersen, Marie W. Lindholm and Lars B. Nielsen
    Circulation Research. 2015;117:933-942, originally published September 10, 2015
    https://doi.org/10.1161/CIRCRESAHA.115.307182
    del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Related Articles

Cited By...

Subjects