Establishment of a rat model with diet-induced coronary atherosclerosis

Abstract Coronary atherosclerotic disease is a serious disease in humans, but no suitable animal model is available currently for further studies. We used apolipoprotein E gene knockout (ApoE KO) rats to induce hypercholesterolemia through a special high cholesterol/bile salt diet (Paigen diet), then analyzed aortic and coronary atherosclerosis lesions and the myocardial injury in order to establish a novel small animal model of coronary atherosclerosis. Plasma cholesterol of ApoE KO rats increased 7.6-fold compared with wild-type rats after 8 weeks on the Paigen diet. After 10 to 12 weeks of subsisting on the Paigen diet, ApoE KO rats developed mild aortic atherosclerosis with severe coronary atherosclerosis. Hematoxilyn and eosin staining showed that 11 out of 12 ApoE KO male rats had right coronary artery atherosclerosis, 7 of them were>70% occluded. Oil Red O (Lipid Stain), Mac2 immuno-staining and Masson’s trichrome staining demonstrated substantial amounts of lipid, macrophages and collagen fibers in coronary atherosclerosis plaques. In addition, ApoE KO male rats had severe myocardial focal lesions with cholesterol ester as the main component in the lesions. In conclusion, ApoE KO rats developed severe hypercholesterolemia, coronary atherosclerosis and myocardial cholesterol ester deposition after subsisting on the Paigen diet and can be used as a novel animal model for studies on cholesterol metabolism and coronary atherosclerotic disease.


Introduction
Coronary atherosclerotic disease (CAD) is a human disease with a high morbidity and mortality. It is caused by coronary artery atherosclerosis which leads to stenosis or occlusion of the vessels and progresses to myocardial ischemia. Choosing a suitable animal model for coronary heart disease is a necessary approach to study the occurrence, development and outcome of the human disease. Although human atherosclerosis mainly occurs at cardiovascular and cerebrovascular arteries, causing such clinical events as myocardial infarction and stroke, the majority of atherosclerosis lesions mainly occur in the aorta, and rarely, in the coronary arteries in most atherosclerotic animal models [1] . Myocardial ischemia can be artificially induced in mice by coronary artery ligation or injection of isopropyl epinephrine without the pathological lesions of coronary arteries. Therefore, it cannot mimic the natural progression of human CAD. Application of genetic modification techniques including gene deletion of B1 type scavenger receptor (SRB1) [2] and nitric oxide synthase (NOS) [3] has now been developed to generate a number of mouse models of CAD, but extensive genetic manipulations and inefficient long-term breeding are required.
ApoE is one of the most important apolipoprotein in human plasma, and distributed mainly in the low density protein (VLDL), intermediate density lipoprotein (IDL), chylomicrons (CM) and their residues (CMR). ApoE plays its function mainly through the interaction with low density lipoprotein receptors (LDLR), very low density lipoprotein receptors (VLDLR), and LDL receptor related protein (LRP). It can regulate the lipolysis and clearance of plasma lipoprotein, and the synthesis of VLDL. In human plasma, there are 3 ApoE isomers, E2, E3 and E4, with E3 being the wild type isomer. E2 and E4 polymorphism were associated with a higher risk of cardiovascular disease [4] . It has been reported that ApoE null mutation patients suffered from early onset of severe cardiovascular disease, xanthoma and type III hyperlipidemia. The plasma total cholesterol (TC) level can reach as high as 500-700 mg/dL [5][6] .
In 1992, two groups separately generated ApoE knockout mice [7][8][9] . The plasma TC in these mice can reach about 400 mg/dL on a diet comprised of chow. These mice can even develop spontaneous atherosclerosis at 3 months of age. High cholesterol diet can induce severe hypercholesterolemia and aortic atherosclerosis in ApoE KO mice. Currently, ApoE knockout mice have become the most commonly used animal model for the research of hypercholesterolemia and aortic atherosclerosis. Even though simultaneous modification of certain genes at the same time could induce severe coronary atherosclerosis in ApoE KO mice [10][11] , the genetic modification and breeding are difficult.
Compared with mice, rats are large enough to allow surgical operation and imaging examination, thus enabling an easier evaluation of their cardiovascular function. In effect, it is possible to obtain larger biological samples from rats than from mice. Therefore, rats are widely used in the study of cardiovascular physiology and pathology. Most importantly, under some special conditions, certain degrees of coronary atherosclerosis can be induced in wild-type rats, but not mice. If we can promote the development of atherosclerosis through ApoE gene knockout, it is likely to induce more severe coronary atherosclerosis in rats. This type of rat may become a more suitable small animal model for human CAD research.
In this study, ApoE knockout rats (ApoE KO), generated by transcriptional activator-like effector nucleases (TALEN) mediated gene editing method, were used to analyze their plasma lipid contents, aortic and coronary atherosclerotic lesions and myocardial injury produced by chow diet or Paigen diet feeding.

Material and methods
Animals were maintained on a 12-hour light/12-hour dark cycle at 24°C, given water ad libitum and fed a standard laboratory chow diet. The Principles of Laboratory Animal Care (NIH publication No.85Y23, revised 1996) were followed, and the experimental protocol was approved by the Animal Care Committee, Peking University Health Science Center (LA2010-059).

TALEN construction
A pair of TALENs targeting exon 3 of the ApoE gene was created. Each TALEN binds to 20 bp of DNA and the binding sites are separated by a 14-bp spacer region as illustrated in Fig. 1A. TALEN plasmids were linearized and transcribed in vitro using a mMessage mMachine T7 Ultra Kit (Ambion, Austin, TX, USA) according to the manufacturer's instructions. Capped, polyA-tailed mRNAs were cleaned up with a MEGAclear Kit (Ambion). In addition, mRNAs were precipitated, washed, and resuspended at 1 mg/mL in RNase free water, dispensed into aliquots, and stored at 80°C. TALEN mRNAs were subsequently diluted at a final concentration of 10 ng/mL for embryo injection.

Embryo manipulation
Sprague-Dawley (SD) rats were used for generating ApoE knockout rats. Female embryo donors were superovulated and subsequently individually caged with a male SD rat. The following morning, donors were sacrificed and embryos collected from the oviducts. TALEN mRNAs were injected into the cytoplasm using glass injection pipettes. Embryos that survived the injection procedure were surgically transferred to the oviduct of day 0.5 post coitum psuedopregnant recipient SD females that had successfully mated with vasectomized males. DNA sequencing of genome mutated sites Offspring from injected embryos were screened for mutations in the ApoE locus using DNA sequencing. The target sites were amplified by PCR with specific primers (ApoE-F, 5′-GTTGGTCCCATTGCTGACAG-3′, and ApoE-R, 5′-CAGATAGGAGGAACCCCTG-GAT-3′) from genomic DNA. PCR was performed with 35 cycles of a reaction consisting of 45 s of denaturation at 95°C, 45 seconds of annealing at 62°C and 45 seconds of elongation at 72°C. The PCR products were 748 bp. DNA sequencing was performed by SinoGen-oMax Co., Ltd (Beijing, China).

RNA isolation and quantitative real-time PCR
Total RNA was extracted from the liver using Trizol reagent (Invitrogen, Carlsbad, CA, USA) and firststrand cDNA was generated by using an RT kit (Invitrogen). Quantitative real-time PCR was performed using specific primers (r-ApoE-F, 5′-CTGCTGTT-GGTCCCATTGCT-3′, and r-ApoE-R, 5′-CCGAGTC-GGTTGCGTAGATC-3′). Amplifications were performed in 35 cycles using an opticon continuous fluorescence detection system (MJ Research, Foster City, CA, USA) with SYBR green fluorescence (Molecular Probes, Eugene, USA). Each cycle con-sisted of denaturation for 30 seconds at 94°C, annealing for 30 seconds at 56°C, and extension for 30 seconds at 72°C. All samples were quantitated by using the comparative CT method for relative quantitation of gene expression, normalized to GAPDH [12 ].

Blood lipid analysis
Blood was obtained by retro-orbital bleeding. Plasma TC and triglyceride (TG) were determined by using enzymatic methods (Bio Sino, Beijing, China). High density lipoprotein cholesterol (HDL-C) was measured with the TC kit after ApoB-lipoprotein had been precipitated with 20% polyethylene glycol solution. Free cholesterol (FC) was measured using enzymatic methods (Applygen Technologies Inc, Beijing, China). For lipoprotein distribution analysis, pooled plasma samples from 6 rats per group were separated by fast protein liquid chromatography (FPLC) and cholesterol were determined in each fraction. For Western blot analysis, 1 mL of plasma from each sample were loaded and separated by SDS-polyacrylamide gel electrophoresis. The proteins transferred on nitrocellulose membranes were recognized with the required primary antibodies and secondary antibodies conjugated to horseradish peroxidase. The blots were developed by use of the enhanced chemiluminescence detection reagents.

Histological studies
Atherosclerotic lesion area in the aortic root and coronary artery were quantified on cross sections of the aorta as previously described [13] . In brief, rats were sacrificed and flushed with 80 mL 0.01 mol/L phosphate-buffered saline (PBS) through the left ventricle. Tissues were harvested and stored in -80°C or fixed in 4% paraformaldehyde (PFA) for 4 hours, and then transferred to 20% sucrose. The heart was embedded in OCT, snap-frozen in liquid nitrogen and stored at -30°C prior to sectioning. Serial 7 mm sections were obtained working from the apex of the heart toward the origin of the aorta, and sections were mounted from the point where all three aortic valve cusps became clearly visible. Every 20 th section was used for oil red O staining, counterstained with hematoxylin. Atherosclerotic lesion areas were measured using Image J graphic Analysis System and were reported as the average oil red O staining area per section in the first such section for each rat.
Sections were also stained with hematoxylin and eosin (H&E) or Masson's trichrome method for fibrosis analysis. Paraffin-embedded hearts were sectioned at a thickness of 5 mm and stained with H&E or Sirius red for fibrosis analysis. Immuno-detection was performed with Mac2 antibody (Santa Cruz Biotechnology, Dallas, TX) to examine macrophage infiltration.

Heart lipid analysis
Hearts (~50 mg wet weight) were weighed and homogenized in 1 mL PBS. Lipids were extracted as described by Folch et al. [14] and dissolved in 200 mL 3% Triton X-100 for TC and TG analysis using enzymatic methods as described earlier or dissolved in 50 mL choloroform for thin layer chromatography (TLC) to determine cholesterol ester (CE) contains. TLC was performed on silica G-24 plates. The chromatographic developing solution was heptane/diethylether/acetic acid (74:21:4, vol/vol/vol) [15] using lipid from white adipose tissue as TAG control.

Statistical analysis
Quantitative data were given as meanAESEM. Statistical significance was tested using two-tailed Student's t test and one way ANOVA (Turkey posttest) by the computer program Prism (GraphPad Software). A value of P < 0.05 was considered statistically significant.

Generation of ApoE KO rats
Traditional methods of gene targeting are complex with low efficiency and need culture of rat ES cells. In this study, the currently widely used genome editing methods TALEN was used to generate ApoE KO rats. A pair of TALEN plasmids targeting exon 3 of the ApoE gene was created and the target sequence is shown in Fig. 1A. We chose 13 bp deletion mutation founders to generate stability passaged strain. The mutation site and sequences are shown in Fig. 1B. The sequencing peak graph and amino acid changes are shown in Fig. 1C. To verify ApoE knockout efficiency, we examined ApoE mRNA levels in the liver, which possessed high level ApoE expression in wildtype rats. As shown in Fig. 1D, ApoE mRNA expression was undetectable in ApoE KO rats. In addition, we measured ApoE protein levels in plasma and there was no detectable ApoE protein in ApoE KO rats (Fig. 1E). We successfully established ApoE KO rats using TALEN mediated gene editing technique.

Mild aortic atherosclerosis in ApoE KO rats
ApoE KO mice displayed spontaneous aortic atherosclerosis at 3 months of age, and had severe aortic atherosclerosis after high-fat diet feeding [9]. In this study, we can hardly detect atherosclerotic lesions even in 16-month old ApoE KO rats upon chow diet (Data not shown). After 10 to 12 weeks of Paigen diet feeding, oil red O staining in frozen section of aortic roots and the full length of the aorta showed that ApoE KO rats displayed mild aortic atherosclerosis with aortic root lesions (WT: 1.50AE0.47Â10 3 μm 2 ; ApoE KO:254.2AE 33.4Â10 3 μm 2 ) (Fig. 3A and 3B) and full length of aorta lesion (WT: 0.59AE0.15%; ApoE KO:2.58AE0.42%) (Fig. 3C and 3D).

Severe coronary atherosclerosis in ApoE KO rats
Though aortic atherosclerosis was mild in ApoE KO rats, severe coronary atherosclerosis was observed after 10 to 12 weeks of Paigen diet feeding. H&E staining showed 11 of 12 male ApoE KO rats having coronary artery atherosclerosis in the ostium of the right coronary artery, 7 of them with>70% occlusion. Among the females, there were also 4 of 10 ApoE KO rats displaying mild right coronary artery atherosclerosis with < 25% occlusion. We did not detect any coronary atherosclerosis in WT rats (Fig. 4). Oil red O, Mac2 immuno-staining and Masson's trichrome staining showed coronary atherosclerosis plaque in ApoE KO rats containing substantial amounts of lipid, macrophages and collagen fibers (Fig. 4).

Myocardial cholesterol ester deposition in ApoE KO rats
ApoE KO male rats also exhibited severe myocardial focal lesions after Paigen diet feeding. They could be observed by naked eyes on the transverse section of the myocardium (Fig. 5A). Oil red O, Mac2 immunostaining and Sirius red staining showed that the main components of myocardial lesions were neutral lipids, together with macrophages and collagen fibers (Fig.  5B). To detect the lipid composition in the myocardial lesions, we conducted myocardial lipid extraction, and then measured cholesterol and triglyceride contents. It was found that cholesterol levels were dramatically increased (Fig. 5C). We also conducted TLC using myocardial lipid extraction samples. It was shown that the major lipid component in myocardial lesions was CE (Fig. 5D).

Discussion
It was hard to generate gene knockout rats based on the traditional method using ES cell homologous recombination. Nowadays, with the application of novel gene editing techniques (TALEN and CRISPR/ Cas system) knockout rats can be produced quickly and efficiently. There were already reports about ApoE knockout rats [16][17] . The plasma cholesterol levels of ApoE KO rats were 1.5-fold higher than WT controls, but significantly lower than ApoE KO mice, which are consistent with the present data. However, there was no report on the phenotypes upon high-fat diet feeding in these ApoE gene deleted rats. In this study, we analyzed plasma lipid and atherosclerotic features in ApoE KO rats generated by TALEN method. Plasma cholesterol levels were slightly increased in ApoE KO rats upon chow diet. After Paigen diet feeding, ApoE KO rats had severe hypercholesterolemia, coronary atherosclerosis and myocardial cholesterol ester deposition.
There were many differences in phenotypes between ApoE KO rats and mice. 1) Plasma cholesterol levels upon chow diet are mildly elevated in ApoE KO rats but highly elevated in ApoE KO mice (about 170 mg/dL vs. 400 mg/dL). 2) There were more aortic atherosclerotic lesions in ApoE KO rats than in mice either as spontaneous occurrence or by high fat diet induction.
3) There was obvious coronary atherosclerosis in ApoE KO rats, which, however, was nearly completely absent in mice during 3-month Paigen diet induction [18] . 4) ApoE KO rats also displayed severe myocardial CE accumulation after Paigen diet feeding, while there was no report on this phenomenon in ApoE KO mice.
The apparent difference between ApoE KO mice and rats in aortic and coronary atherosclerosis may be related to the difference of vascular diameter and wall structure which would lead to hemodynamic changes. ApoE KO rats readily develop coronary atherosclerosis but fewer lesions in the aorta. Hence, ApoE KO rats can certainly serve as a better model for coronary atherosclerosis, which is the major pathological process in humans. Because ApoE KO rats do not develop spontaneous atherosclerosis in the aorta and the atherosclerotic lesions must be induced by Paigen diet feeding ApoE KO rats are therefore not suitable to serve as a model for the study of spontaneous atherosclerosis. In this regard, ApoE KO mice are superior to the rats.
ApoE KO rats displayed severe myocardial CE accumulation, which probably was a type of cardiac xanthoma, resulting from severe hypercholesterolemia. The reason for massive accumulation of myocardial CE in ApoE KO rats, but not in ApoE KO mice, may be related to undefined difference in metabolic features between the two genera. ApoE KO rats also exhibited skin xanthoma (Supplementary Data), which is similar to certain types of ApoE mutated patients [5][6] . We found that the mRNA expression of oxidized low density lipoprotein (lectin-like) receptor 1 (Lox1) was increased in the heart of ApoE KO rats. Lox1 mediates the uptake of oxidized lipoproteins [19] , and it was reported that lipoproteins from ApoE KO mice were highly oxidized [20] . This may play a certain role in the accumulation of myocardial CE in ApoE KO rats.
According to the results of echocardiography, ApoE KO rats did not display myocardial systolic functional changes after Paigen diet feeding (Supplementary Data). Therefore, although ApoE KO rats had severe coronary atherosclerosis and myocardial CE accumulation, the heart function was still in the compensation phase. Male ApoE KO rats displayed more severe phenotypes than female rats. This may be related to the cardiac protective role of estrogen.
Severe coronary atherosclerosis can be induced in ApoE KO rats by 10 to 12 weeks of Paigen diet. This rat model is easy to operate and handle. Furthermore, the process of diet induced atherosclerosis is similar to the pathological progression of human diseases. ApoE KO rats could serve as a good animal model of coronary atherosclerosis that could be applied to the pathogenic studies and drug screening, and so on.
In the future, we will analyze the impact of different diet compositions and feeding time on the progression and regression of coronary atherosclerosis, cardiac function and the occurrence of acute myocardial infarction in ApoE KO rats. We will also study the pharmacological intervention with lipid-lowering and anti-atherosclerotic agents in these ApoE KO rats. To evaluate the advantages and disadvantages of ApoE KO rats as a novel animal model for coronary atherosclerosis and hypercholesterolemia. In conclusion, ApoE KO rats, generated by the TALEN mediated gene editing method, developed server hypercholesterolemia, coronary atherosclerosis and myocardial CE deposition after subsisting on a Paigen diet. ApoE KO rats can be used as a novel coronary atherosclerosis and hypercholesterolemia animal model.