Recent Highlights of ATVB Updates on Approaches for Studying Atherosclerosis

Introduction

Atherosclerosis is initiated and propagated over decades with no overt clinical symptoms. The complexity and chronicity of its pathophysiological process have been a major challenge to fully define molecular mechanisms and explore more effective therapeutic targets.^1,2 This Recent Highlights series summarizes ATVB articles published in 2017 and 2018 on the traditional animal model systems and new approaches that have become powerful tools for studying pathogenesis and mechanisms of atherosclerosis.

Animal Models

Many animal models have been used to study atherosclerosis, as summarized in the recent AHA Scientific Statement,³ which provide suggestions on selecting appropriate animal models for a specific atherosclerosis study. Given its value and feasibility as a proof-of-principle model, mouse models have continued to be the most common species to study atherosclerosis.

Two Classic Mouse Models

The two most commonly used mouse models to study atherosclerosis are apolipoprotein E deficient (Apoe^−/−) mouse⁴ and low-density lipoprotein (LDL) receptor deficient (Ldlr^−/−) mouse.⁵ In 61 recent ATVB articles (Figure 1), studies in 41 articles used Apoe^−/− mice,^6–46 13 articles reported Ldlr^−/− mice,^47–59 and 7 articles reported both Apoe^−/− and Ldlr^−/− mice.^60–66 The basis for the more common use of Apoe^−/− mice, instead of Ldlr^−/− mouse model, is unclear. However, it is worth noting that although both Apoe^−/− and Ldlr^−/− mice have profound hypercholesterolemia when fed a saturated fat-enriched diet, the mechanisms of lipoprotein metabolism dysfunction and development of atherosclerosis in these two mouse models may be different.^3,67,68

Figure 1. Summary of 61 publications in ATVB between 2017 and 2018 that have reported Apoe^−/− mice and/or Ldlr^−/− mice.

Among the 61 articles, Apoe^−/− mice were studied in 41 articles, Ldlr^−/− mice were studied in 13 articles, and 7 articles reported both mouse strains. Among articles that studied Apoe^−/− mice, 17 only reported male mice, 8 only reported female mice, and 16 reported both male and female mice. Among articles that studied Ldlr^−/− mice, 5 only reported male mice, 1 only reported female mice, and 7 reported both male and female mice. Among articles that reported both mouse strains, 4 only reported male, 2 reported both male and female mice, and 1 did not provide the sex information.

Recently Developed Mouse Models

Proprotein convertase subtilisin/kexin type 9 (PCSK9) regulates LDL receptor activity.^30,69 A D374Y gain-of-function mutation in human leads to hypercholesterolemia.⁷⁰ Benefitting from the recently advances of using adeno-associated viral vectors (AAV), infection of AAVs carrying this human mutation or its mouse equivalent mutation D377Y in C57BL/6 mice has been used by many researchers to study atherosclerosis.^71–81 Since sex difference is an important influence on development of atherosclerosis,^82,83 some researchers have also studied potential sex difference of this method in inducing hypercholesterolemia and their contributions to atherosclerosis.^80,84 Jarrett and colleagues compared an AAV containing human PCSK9 D374Y mutation (PCSK9D374Y.AAV) to AAV vectors expressing Staphylococcus aureus Cas9 and a guide RNA targeting the Ldlr gene (AAV-CRISPR) in both male and female C57BL/6 mice.⁸⁰ Their study found that the AAV-CRISPR approach to target LDL receptors led to more profound hypercholesterolemia and increases of atherosclerosis than infection with PCSK9 D374Y.AAV in male mice, whereas PCSK9 D374Y.AAV method had more significant effects on plasma cholesterol concentrations and atherosclerosis than the AAV-CRISPR approach for LDL receptor deletion in female mice.⁸⁰ The mechanism for sex dimorphism between these two methods is unclear.

Diet-related Effects

For the mouse models described above, it is common to feed these mice a diet that represents the dietary habits in Western countries.^3,85–88 The currently available Apoe^−/− mice have modest hypercholesterolemia that develop minimal atherosclerosis spontaneously. However, feeding a Western diet profoundly increases plasma cholesterol concentrations and accelerates atherosclerosis. Some laboratories also use a similar fat-enriched diet supplemented with cholesterol (1.25% wt/wt) and cholate, as discussed in the recent AHA statement.³ Some recent studies have provided evidence that gut microbiota influence the development of atherosclerosis,^89–91 although findings from one study do not support that microbiota and choline-supplemented diet affect atherosclerotic lesion size in a hypercholesterolemic mouse model.⁹ Irrespectively, it is important to provide detailed information including mouse housing conditions and whether diets are supplemented with choline when report atherosclerosis studies in animal models.^9,92

Large Animal Models

Despite the much higher expense and difficulty in genetic manipulations, large animal models have potential benefits over mice in providing more relevant pathological and mechanistic insights into the disease in human. Three recent articles in ATVB have reported manipulations of lipoprotein metabolism and the consequences on atherosclerosis in rabbits.^93–95 In one study, an apoA-I-expressing helper-dependent adenoviral vector delivered to carotid arteries led to reduced atherosclerosis in fat diet-fed rabbits.⁹³ In another study, Wang and colleagues generated transgenic rabbits expressing the human endothelial lipase in liver.⁹⁴ These rabbits had lower plasma cholesterol concentrations and less atherosclerotic lesions in both the aorta and coronary arteries, compared to their non-transgenic littermates. These investigators also developed a cholesterol ester transfer protein deficient rabbit model.⁹⁵ These rabbits fed a cholesterol diet for 18 weeks exhibited lower plasma total cholesterol concentrations and atherosclerotic lesions in both the aorta and coronary arteries, compared to their wild type controls.⁹⁵ It was noted that these rabbit models develop atherosclerotic lesions in coronary arteries, which is common in humans, but not commonly observed in mouse models.

Pigs are also an optimal animal model to study atherosclerosis due to their similarities to human anatomy and comparable patterns of atherosclerotic lesions in coronary arteries.³ Recently, a panel of genes that regulate endothelial cell apoptosis induced by shear stress have been identified in aortas of healthy pigs, implicating potential mechanisms in atherosclerosis development.⁹⁶ Burke and colleagues successfully depleted LDL receptor in Yucatan miniature pigs.⁹⁷ A high-fat diet feeding on these pigs led to profound increases of plasma LDL-cholesterol concentrations and large lesions in the aorta and coronary arteries. Bempedoic acid, a drug targeting hepatic ATP-citrate lyase, reduced both plasma cholesterol concentrations and atherosclerosis in this pig model.⁹⁷

Established Genetic Approaches

This section introduces two well-established genetic manipulations in mouse models that have provided remarkable insights into understanding mechanisms of atherosclerosis. One approach is bone marrow transplantation.⁹⁸ The other approach is the LoxP-Cre conditional genetic manipulation first reported by Sauer more than 30 years ago.⁹⁹

Bone Marrow Transplantation

Leukocytes including macrophages, lymphocytes, and neutrophils are present in atherosclerosis, with macrophages being the predominant cell type during development of atherosclerotic lesions.¹⁰⁰ One approach for studying effects of a gene-of-interest in leukocytes is bone marrow transplantation. This approach, since it was first reported by Linton and Fazio in 1995¹⁰¹ and by the Curtiss group¹⁰² in mouse models for atherosclerosis, it has been used widely to study atherosclerosis. In many recent publications of ATVB, with this approach, it was demonstrated that many components involving from transcription to protein functions contribute to either development or prevention of atherosclerosis. Some studies also report mixed chimeric mouse models using this bone marrow transplantation method. For example, mixed chimeric recipient mice were developed by transplanting bone marrow cells from μMT chain-deficient mice (80%) and bone marrow cells from mice deficient in CD40 (20%) into irradiated Ldlr^−/− mice. These chimeric mice had low abundance of CD40 in B cells, which provided evidence that B cell-selective CD40 deficiency reduced atherosclerosis in mice.⁶¹ The recent studies published in ATVB that applied this approach in mouse atherosclerosis models are summarized in Table 1 (Ldlr^−/− mice as recipient) and Table 2 (Apoe^−/− mice as recipient). From these two tables, it is clear that Ldlr^−/− mice are more frequently used as recipient mice for bone marrow transplantation since donor mice do not need to be in an Ldlr^−/− background, as noted previously.^102,103

Table 1.

Bone Marrow Transplantation in Ldlr^−/− Recipient Mice

Manipulation of Gene	Full Name of Gene	Other Manipulations	Effects on Atherosclerosis*	Reference
Depletion of Zhx2	Zinc fingers and homeoboxes 2	Western diet x 18 or 24 weeks	aortic root ↓; en face ↓	47
Depletion of Dab2	Disabled homolog 2	Western diet x 20 weeks	aortic root ↔; en face ↓	50
Depletion of IL-19	Interleukin-19	Western diet x 14 weeks	en face ↑	49
Depletion of ACAT1	Acyl-CoA cholesterol acyltransferase	Western diet x 8 weeks	aortic root ↑	52
Depletion of NLRP3	NACHT, LRR, and PYD domain-containing protein 3		aortic root ↓
Human iNAMPT transgene	Intracellular nicotinamide phosphoribosyltransferase	Western diet x 12 weeks	aortic root ↓	57
human mCAT transgene	Mitochondrial catalase	Western diet x 16 weeks	aortic root ↓	55
Depletion of myeloid EGFR	Epidermal growth factor receptor	High cholesterol diet x 4, 7, or 12 weeks	aortic root ↓	53
Depletion of miR146a	microRNA 146a	High cholesterol diet x 12 weeks	N/A	63
E2-def	Lack of Ly6C− monocytes	High cholesterol diet x 15 weeks	en face ↑	64
Depletion of myeloid MED1	Mediator 1	Western diet x 12 weeks	aortic root ↑; en face ↑	65
Depletion of MHCII (in follicular B cells)	Major histocompatibility II	Western diet x 8 weeks	aortic root ↓	61
Depletion of CD40 (in follicular B cells)	Cluster of differentiation 40	Western diet x 8 weeks	aortic root ↓
Depletion of Blimp-1 (in follicular B cells)	B-lymphocyte-induced maturation protein	Western diet x 8 weeks	aortic root ↓

^*Effects on atherosclerosis, compared to their wild type controls. ↑: increase; ↓: decrease; ↔: no change

Table 2.

Bone Marrow Transplantation in Apoe^−/− Recipient Mice

Manipulation of Gene	Full Name of Gene	Recipient Mouse Strain	Other Manipulations	Effects on Atherosclerosis*	Reference
Human p27kip transgene	Cyclin-dependent kinase inhibitor 1B	Apoe−/−	Unknown	aortic root ↓	12
Depletion of VWF	von Willebrand factor	Apoe−/−	Western diet x 14 weeks	aortic root ↔; en face ↔	19
Depletion of Sema7A	Semaphorin 7A	Apoe−/−	partial carotid artery ligation and Western diet x 2 weeks	carotid artery ↔	24

^*Effects on atherosclerosis, compared to their wild type controls. ↓: decrease; ↔: no change

LoxP-Cre Genetic Manipulations

Cross-breeding mice carrying LoxP franked gene with mice carrying Cre transgene driven by a specific promoter is a well-recognized approach for inducing tissue or cell-specific genetic manipulations.^104,105

LysM-Cre transgenic mice,¹⁰⁶ which generate conditional gene targeting macrophages and granulocytes, have been used frequently to study experimental atherosclerosis. Using mice expressing Cre under the control of the LysM-Cre promoter, attenuation of atherosclerotic development occurred in mice with (1) deletion of myeloid β-catenin, a critical component regulating many cell functions and behaviors differentiation,⁴⁸ (2) depletion of lipin-1, an important enzyme in the glycerolipid synthesis pathway,⁸¹ (3) depletion of ADK (adenosine kinase), an essential enzyme regulating intracellular adenosine levels,⁴⁶ and (4) depletion of EGFR (Epidermal Growth Factor Receptor), a prominent member of the receptor tyrosine kinase family.⁵³ Mice with hypercholesterolemia, such as Apoe^−/− or Ldlr^−/−, were used for these studies. Reduced atherosclerosis through manipulation of any of these genes did not affect plasma cholesterol concentrations, supporting the notion that hypercholesterolemia is crucial in the initiation and development of atherosclerosis, but prevention of atherosclerosis can be achieved even in the presence of profound hypercholesterolemia. This concept was also supported by studies of HMG-CoA (3-hydroxy-3-methylglutaryl-coenzyme A) reductase, the rate-limiting enzyme in cholesterol biosynthesis. Inhibition of HMG-CoA by statins has profound effects on reducing plasma cholesterol concentrations in human, but not in the commonly used hypercholesterolemic mouse models.^1,107–112 Despite its lack of effects on plasma cholesterol concentrations, statins reduce atherosclerosis in mouse models.^111–115 Depletion of HMG-CoA reductase in myeloid cells, as achieved by LysM-Cre activation, attenuates atherosclerosis in Ldlr^−/− mice.⁵¹

Endothelial cells play critical roles in the initiation and development of atherosclerosis.¹¹⁶ Nucleotide P2Y₂ receptor (P2Y₂R) regulates vascular cell adhesion molecule-1 (VCAM-1). P2Y₂ floxed mice were bred with cadherin 5-Cre mice to generate endothelial cell specific P2Y₂ deficiency, which reduced atherosclerosis in Apoe^−/− mice.⁴³ In another study, inducible and endothelial cell-specific alpha5 deficient mice were generated by a cadherin promoter driven Cre-ERT2 with tamoxifen injections. Atherosclerotic lesions were reduced by alpha5 endothelial-specific depletion.¹⁵ It is worth noting that inducible Cre is not only used for cell-specific studies, but also used to induce Cre activity to deplete a gene globally in adult mice, which could avoid potential impact during embryonic development.¹¹⁷ For example, depletion of ApoE was achieved in Apoe floxed mice with Rosa26 Cre-ERT2 at 10–12 weeks of age, with an oral dose of tamoxifen that increased plasma cholesterol concentrations and atherosclerosis.¹⁰

The SM22 promoter has been well-accepted for driving smooth muscle cell-specific genetic manipulation.¹¹⁸ Two recent studies reported smooth muscle cells-specific effects, as achieved by SM22-Cre recombinase activity, on atherosclerosis: Deficiency of IGF1 (insulin-like growth factor) receptor in smooth muscle cells, augmented atherosclerosis;¹⁶ and smooth muscle cell-specific overexpression of USP20 (ubiquitin-specific protease 20, a deubiquitinase) reduced atherosclerosis.⁶²

In addition to the commonly used myeloid, endothelial cell, and smooth muscle cell-specific Cre, several other tissue- or cell-specific Cre recombinase studies were reported to have an effect on mouse atherosclerosis. Brown adipocyte-specific PPARγ (peroxisome proliferator-activated receptor gamma) deficient mice were developed by crossbreeding PPARγ floxed mice with UCP-1 (uncoupling protein 1)-driven Cre mice.¹¹⁹ Deficiency of PPARγ in brown adipocytes led to augmented atherosclerosis.⁸ Adipocyte-specific LRP1 (low-density lipoprotein receptor-related protein 1) deficient mice were generated by mating LRP1 floxed mice with aP2-Cre transgenic mice,¹²⁰ which exhibited accelerated atherosclerosis in carotid arteries.⁵⁶ Epithelial cell-specific androgen receptor deficient mice were generated by K5-Cre recombinase activity in androgen receptor floxed mice, which showed increased atherosclerosis.⁶

Three recent seminal publications have provided comprehensive discussion on adventitial fibroblast,¹²¹ endothelial cell,¹²² and smooth muscle cell-specific¹²³ Cre-recombinase mice. The strengths and weaknesses of multiple Cre-recombinase mouse lines were discussed in these three articles. An important synopsis of these publications is that it is important to validate the location and abundance of the genetic manipulation in Cre transgenic conditional mice.

Emerging Applications

Imaging systems including ultrasound, CT (Computed Tomography), PET (Positron Emission Tomography), and MRI (Magnetic Resonance Imaging) are used for diagnosing, localizing, and characterizing atherosclerotic lesions in humans.^124–129 These imaging systems, along with cytometric analysis and microscopy, have provided insights into understanding mechanisms and pathogenesis of atherosclerosis.¹²⁹ Many of these sophisticated techniques have been applied to animal models to track macrophage biology and behavior because macrophage dynamics are a salient component in the development and progression of atherosclerosis.^{64,79,130–133}

Mass Cytometry

Flow cytometry is traditionally used to quantify a limited number of cell surface markers. Mass cytometry quantifies protein abundance at a single-cell resolution using heavy-metal-conjugated antibodies in tandem with time-of-flight mass spectrometry.¹³⁴ Thomas and colleagues profiled human monocytes with a panel of 36 cell surface markers using this mass cytometry in combination with a computational analysis method,¹³⁵ which defined monocyte classifications by incorporating all cell surface markers simultaneously. Using this method, a small but specific panel of surface markers on human blood monocytes were identified, which improved monocyte subset identification with high accuracy and reproducibility.¹³⁵ This method is expected to facilitate identification of monocyte subsets and enhance consistency in defining monocyte functions not only in atherosclerosis research field but also many other research fields.

Two-photon Microscopy

Microscopy is an essential tool to visualize, characterize, and quantify atherosclerotic lesions. The state-of-the-art intravital two-photon imaging enables visualization of macrophage dynamics in a real-time manner in live animals. In a recent study, Li and colleagues¹³⁶ transplanted aortic arch grafts from Apoe^−/− mice to the right carotid arteries of C57BL/6 CX₃CR1 GFP reporter mice. Recipient monocyte-derived macrophages recruited to atherosclerotic lesions was detectable one month after transplantation, but reduced profoundly 4 months after transplantation, which was visualized using a two-photon microscope on the same animal at two time points.¹³⁶ Two-photon microscopy was also used for intricate characterization of macrophage phenotypes. Williams and colleagues found that in Ldlr^−/− mice fed a fat-enriched diet for 12 – 16 weeks, lipid-laden macrophages accumulated in the intima of carotid artery had different morphology and locations: Large asymmetrical cells were static, but smaller round cells near lesional margins were mobile.⁶⁰ During the progression of atherosclerotic lesions in Apoe^−/− mice, macrophages, as detected by nondegradable phagocytic particles, appeared to move to deeper location within atherosclerotic lesions, whereas during regression of atherosclerotic lesions in Apoe^−/− mice achieved by application of AAV8 vector encoding ApoE, motile macrophages within atherosclerotic lesions were closer to the surface of lesions.⁶⁰ These two recent studies provide insights into understanding complex dynamics of macrophages during the development of atherosclerosis in animal models.

Positron Emission Tomography (PET)

Fluorodeoxyglucose (¹⁸F-FDG) is a marker for cellular uptake of glucose. Since macrophages have high glycolytic capacity, ¹⁸F-FDG-PET has been used to monitor the abundance of macrophage accumulation within atherosclerotic lesions.^137–141 Macrophage behavior and functions are highly heterogeneous.¹³¹ The classic PET using ¹⁸F-FDG alone is not able to identify the heterogeneity of macrophages. To overcome this shortcoming, Tavakoli and colleagues applied PET by combining measurement of glucose (¹⁸F-FDG) and ¹⁴C-glutamine accumulation.¹⁴² This approach identified different patterns of ¹⁸F-FDG and ¹⁴C-glutamine accumulation in macrophages, further demonstrating heterogeneity of macrophages accumulated in atherosclerotic lesions. In addition to ¹⁸F-FDG labeling, a recent study assessed accumulation of monocytes and monocyte-derived macrophages during the regression of atherosclerotic plaques using serial-based scanning with a CCR2 (C-C chemokine receptor 2)-specific imaging probe, which specifically detect CCR2-related responses.¹³⁶ It is envisioned that advancement of different labeling for PET imaging, combined with histological analysis, will enhance accuracy of atherosclerotic lesion characterization.

Near-infrared calcium imaging

Calcification is a critical condition of atherosclerotic diseases.^{33,128,143–149} In addition to histology to visualize calcification in serial sections of atherosclerotic lesions, near-infrared calcium imaging was used for ex vivo molecular imaging and quantification of the calcification process, based on the principle that conjugated bisphosphonate compounds could bind to hydroxyapatite deposited by osteoblast-like cells during the mineralization of calcium.^150,151 Using this ex vivo near-infrared calcium imaging, Ceneri and colleagues found that calcification signal became gradually enhanced in Apoe^−/− mice fed a high fat and high cholesterol diet for 7, 14, or 21 weeks.⁴² The authors also demonstrated that deletion of Rac2, an important signal transducer in bone marrow-derived cells, resulted in increased calcification in atherosclerotic lesions,⁴² confirming versatile effects of macrophages in the development of atherosclerosis.

Molecular Ultrasound and Molecular Tomography

Molecular ultrasound imaging using targeted microbubbles visualizes atherosclerotic components at the molecular level beyond classic ultrasonography.¹⁵² Curaj and colleagues¹⁵³ using JAM-A, a junctional adhesion molecule, targeted microbubbles to detected flow-dependent endothelial dysfunction induced by carotid partial ligation in Apoe^−/− mice. The authors found JAM-A-targeted microbubbles rapidly bound the ligated side, and reached a peak after two weeks of the ligation procedure. This study provides evidence that JAM-A is a useful imaging biomarker to assess endothelial activation and dysfunction at early stages of atherosclerotic plaque formation.¹⁵³

In addition to endothelial dysfunction, neutrophil accumulation is present in early stages of atherosclerosis.^154–156 In Ldlr^−/− mice fed Western diet were imaged at 4, 8, and 12 weeks by fluorescence molecular tomography/x-ray CT after injections of an elastase-targeted fluorescent agent.¹⁵⁷ The fluorescent signal in the area of the aortic arch was highest after 4 weeks of Western diet feeding, but decreased after 8 and 12 weeks of Western diet feeding,¹⁵⁷ implicating elastase-targeted fluorescence molecular tomography is feasible for detecting early development of atherosclerosis.

Summary

The concept that atherosclerosis is a lipid-driven inflammatory disease has been supported by diverse evidence from basic and clinical research.^{131,158–163} Cell types that are involved in atherosclerosis development include leukocytes and resident cell types such as endothelial cells, smooth muscle cells, and cells in the adventitia. Although it is hard to conclude which cell type is more important than the other cell types, or one specific molecule is more important than the other multiple molecules in a single cell type, the recent studies have provided more elaborate insights into many components in the development of atherosclerosis. In addition to animal studies, human studies have also continuously supported the importance of lipoprotein metabolisms and inflammation in the development of atherosclerosis.^{127,148,164–176} We hope that insights from these animal and human studies will enhance our understanding of this complex disease with an ultimate goal to prevent the development of atherosclerosis and improve related adverse outcomes.