Apolipoprotein E deficiency accelerates atherosclerosis development in miniature pigs

ABSTRACT Miniature pigs have advantages over rodents in modeling atherosclerosis because their cardiovascular system and physiology are similar to that of humans. Apolipoprotein E (ApoE) deficiency has long been implicated in cardiovascular disease in humans. To establish an improved large animal model of familial hypercholesterolemia and atherosclerosis, the clustered regularly interspaced short palindromic repeats (CRISPR)-associated protein 9 system (CRISPR/Cas9) was used to disrupt the ApoE gene in Bama miniature pigs. Biallelic-modified ApoE pigs with in-frame mutations (ApoEm/m) and frameshift mutations (ApoE−/−) were simultaneously produced. ApoE−/− pigs exhibited moderately increased plasma cholesterol levels when fed with a regular chow diet, but displayed severe hypercholesterolemia and spontaneously developed human-like atherosclerotic lesions in the aorta and coronary arteries after feeding on a high-fat and high-cholesterol (HFHC) diet for 6 months. Thus, these ApoE−/− pigs could be valuable large animal models for providing further insight into translational studies of atherosclerosis.


INTRODUCTION
Atherosclerosis is a chronic inflammatory disease characterized by the accumulation of proliferative smooth muscle cells, macrophages, lipids, cholesterol, calcium deposits and cellular debris in arterial walls (Libby et al., 2011;Mastenbroek et al., 2015). The most common clinical manifestations of this disease are myocardial infarction, ischemic stroke and ischemic damage to kidneys and intestines, due to occlusive thrombosis and severe arterial stenosis (Libby et al., 2013;Murray et al., 2012). Although the pathogenesis of atherosclerosis remains elusive, plasma lipid levels, particularly those of low-density lipoprotein (LDL) cholesterol, correlate with the risk of atherogenesis.
Apolipoprotein E (ApoE) is a major component of very-lowdensity lipoprotein (VLDL) and chylomicrons, playing a critical role in the metabolism of cholesterol and clearance of lipoprotein remnants (Mahley, 1988). ApoE was initially recognized for its importance in lipoprotein metabolism and cardiovascular disease because dysfunction of ApoE protein results in familial dysbetalipoproteinemia (type III hyperlipoproteinemia), which often leads to atherosclerosis (Eichner et al., 2002;Georgiadou et al., 2011;Ghiselli et al., 1981). Thus, ApoE knockout (KO) rodent models have been favored in atherosclerosis studies and have provided valuable insights into mechanisms underlying this disease (Chistiakov et al., 2017;Iqbal et al., 2016;Niimi et al., 2016;Piedrahita et al., 1992;Plump et al., 1992). However, rodent models have limitations because many important human atherosclerosis features, including plaque rupture and thrombosis, rarely occur in rodents (Al-Mashhadi et al., 2013;Bentzon and Falk, 2010). In addition, rodents are dissimilar to humans in vascular size and lipoprotein profiles and metabolism, which have limited studies on intravascular devices and hampered the translation of experimental findings into clinic applications (Al-Mashhadi et al., 2013;Langheinrich et al., 2007). Hence, suitable large animal models that can adequately mimic human atherosclerosis are needed.
Currently, pigs are broadly used as large animal models in biomedical research because they share many similarities with humans with respect to physiology, pathology, immunology and the cardiovascular system (Aigner et al., 2010;Reyes et al., 2014;Xi et al., 2004). Moreover, compared to other large animals such as non-human primates, pigs are more ethically and economically acceptable, making them preferable models for cardiovascular disease research. However, it is very difficult to develop advanced atherosclerotic lesions in normal pigs, even with a high-fat and high-cholesterol (HFHC)-diet induction. These considerations suggest the need for the development of genetically modified miniature pigs.
The clustered regularly interspaced short palindromic repeats (CRISPR)-associated protein 9 (CRISPR/Cas9) system has emerged as a powerful and efficient tool for gene editing (Hsu et al., 2014;Mali et al., 2013;Niu et al., 2014), and has significantly facilitated the generation of genetically modified pig models (Chen et al., 2015;Estrada et al., 2015;Hai et al., 2014;Zhang et al., 2017). Recently, there was a report of ApoE/LDL receptor (LDLR) double-KO minipigs produced by CRISPR/Cas9-mediated gene targeting (Huang et al., 2017). Although slightly increased plasma lipid levels were observed, an atherosclerosis phenotype was not ever described in those pigs, possibly because in-frame (IF) mutations in both the ApoE and LDLR alleles may give rise to truncated ApoE and LDLR proteins with partially retained function (Huang et al., 2017).
In the present study, we used the CRISPR/Cas9 system to disrupt the ApoE gene, and established ApoE KO Bama miniature pigs by somatic cell nuclear transfer (SCNT) technology. We have found that ApoE KO pigs fed an HFHC diet had severe hypercholesterolemia and spontaneously developed human-like atherosclerotic lesions in the aorta and coronary arteries. These results indicated that the ApoE KO pigs are an ideal model for the study of human atherosclerosis.

RESULTS
sgRNA-directed site-specific indels in primary pig fetal fibroblasts (PFFs) To disrupt the ApoE gene in Bama miniature pigs, two single-guide RNAs (sgRNAs) were designed based on exon 2 and 3 sequences determined by PCR sequencing. The precise target sites are shown in Fig. 1A. To test their cleavage efficiency, these sgRNAs were cloned into a Cas9 targeting vector, pX330 (#42230, Addgene, Cambridge, MA, USA) and introduced into PFFs. At 48 h post-transfection, primary PFFs were harvested for genomic DNA extraction. PCR amplicons spanning the ApoE target region were subjected to a T7 endonuclease 1 digestion assay followed by gel electrophoresis. The cleavage fragments shown in Fig. 1B indicated that both sgRNAs could induce insertions/deletions (indels) in the target region of the ApoE gene, although sgRNA1 (19.2%) had higher cleavage efficiency than sgRNA2 (3.1%). To establish ApoE KO cell lines, the Cas9-sgRNA1 targeting vector was co-transfected with a neomycin expression plasmid into early passages of PFFs derived from a 35-day-old male Bama mini fetus. After 10 days of selection with G418, resistant cell colonies were harvested. PCR amplification was performed on each colony and the resultant fragments were subjected to TA cloning for Sanger sequencing. Of the 44 colonies analyzed, 27 were identified as having biallelic modifications, 14 of which were homozygous, with 6 indel types ranging from 1 to 39 base pairs (bp) in the ApoE gene target region (Fig. 1C).

Production of ApoE KO pigs by SCNT
ApoE biallelic-modified colonies (A7 and A22) were pooled and used as nuclear donors for SCNT to produce ApoE KO piglets. Morula-and blastula-stage reconstructed embryos were transferred to 7 surrogates, yielding 3 full-term pregnancies, from which 17 live-born male piglets were delivered ( Fig. 2A). Seven of them died before weaning (Table 1). Genomic DNA from the ear tissues of the 10 remaining piglets (designated M1 to M10) was isolated, and mutations were detected via PCR and Sanger sequencing. Piglets M1, M2, M3, M6, M7 and M10 carried a 39 bp deletion corresponding to A22 donor cells; piglets M4, M5 and M8 harbored a 5 bp deletion mutation derived from A7 donor cells; and piglet M9 bore a 4 bp deletion in the ApoE target region that did not originate from the A7 and A22 colonies, indicating that a small portion of cells with a 4 bp deletion in the ApoE region was in the donor-cell mixture (Fig. 2B).
Because the undesired off-target effects of the CRISPR/Cas9 system are still a concern, we computationally predicted off-target sites by screening the pig genome using an online tool (http://www. rgenome.net/cas-offinder/) based on sequence homology to the ApoE sgRNA1 sequence, allowing for un-gapped alignments with up to 4 mismatches. In this way, a total of 81 potential off-target sites (OTSs) were identified. To verify whether off-target cleavage occurred in these genetically modified piglets, 25 OTSs (Table S1) were randomly selected for PCR amplification and subjected to Sanger sequencing. The primers for amplifying the off-target sites are listed in Table S2. The results indicated that none of the sequencing reads exhibited indel mutations, suggesting that there were no off-target effects at these sites. Elevated serum cholesterol levels in ApoE −/− pigs Body weights and appetites of the ApoE targeted piglets, and their age-matched wild-type (WT) Bama minipigs, were comparable at birth and after weaning (data not shown). Plasma ApoE in all piglets was determined by western blot analysis. The ApoE protein was absent in mutant piglets harboring a 4 or 5 bp deletion (M4, M5, M8 and M9), whereas the mutant piglets bearing a 39 bp deletion showed obvious ApoE expression (Fig. 2C). We speculated that a 4 or 5 bp deletion in the ApoE gene region would give rise to a frameshift (FS) mutation, whereas a 39 bp deletion would lead to a 13 amino acids (aa) IF deletion in the ApoE protein. To assess the potential effects of different ApoE gene mutations on serum lipids, these 10 mutant piglets were categorized into ApoE m/m (13 aa deletion) and ApoE −/− (FS mutation) groups. Plasma cholesterol and triglyceride levels were measured in mutant and WT animals 4 weeks after weaning. No significant difference in fasting serum triglyceride concentrations was observed among WT, ApoE m/m and ApoE −/− piglets. The ApoE m/m founders had comparable total cholesterol level as the controls. However, a significantly increased serum cholesterol was observed in ApoE −/− piglets [2.57±0.55 mM (mean±s.d.)] compared to WT (1.35±0.27 mM) and ApoE m/m pigs (1.45±0.21 mM), which was mostly contributed by LDL cholesterol (Fig. 3).

Severe hypercholesterolemia during HFHC feeding
Three-month-old ApoE targeted piglets and age-matched WT pigs were challenged with a HFHC diet containing 30% saturated fat and 1.5% cholesterol for 3 consecutive months. WT pigs fed with chow diet were used as controls. Fasting blood was collected monthly for the assessment of plasma cholesterol and triglyceride levels. During this period, two ApoE m/m piglets (M7 and M10) died because of infection. Unlike WT and ApoE m/m piglets, ApoE −/− piglets maintained on a HFHC diet exhibited chylemia, as indicated by the turbid appearance of their sera (Fig. 4A). Compared to WT pigs on normal chow, increased total cholesterol levels, mostly contributed by high-density lipoprotein (HDL) cholesterol, were observed in WT pigs fed an HFHC diet. However, triglyceride levels did not significantly differ between WT pigs fed an HFHC or normal diet (Fig. 4B). Among all groups, the ApoE −/− group had the highest levels of triglycerides and total cholesterol after 3 months of HFHC feeding, with mean levels of 1.47 and 42.86 mM, respectively (Fig. 4B). ApoE m/m piglets showed less severe hypercholesterolemia, as their total cholesterol levels were about 2-fold that of WT groups on HFHC feeding (14.96 vs 6.76 mM), and their plasma triglyceride levels were comparable to that of WT controls (0.52 vs 0.40 mM) (Fig. 4B). LDL but not HDL cholesterol levels were significantly increased in HFHC-fed ApoE m/m pigs compared to HFHC-fed WT pigs. Interestingly, both LDL and HDL cholesterol levels were dramatically elevated in the HFHC-fed ApoE −/− group compared to the other groups.

Progressive atherosclerosis in ApoE targeted pigs
To facilitate development of atherosclerosis plaques in these pigs, HFHC diet feeding was prolonged to 6 months. Serum ApoE protein in the WT, ApoE m/m and ApoE −/− groups was examined by western blot. As expected, ApoE protein was absent in ApoE −/− piglets. An upregulated expression of ApoE was observed in ApoE m/m piglets compared to WT controls (Fig. 4C). Moreover, the molecular mass of ApoE protein in ApoE m/m piglets was smaller than that in WT animals, due to the deletion of 13 aa in the ApoE protein or degradation of the truncated ApoE protein. To test whether disruption of ApoE accelerated atherosclerosis development, 2 pigs from each group (WT, ApoE m/m and ApoE −/− ) were euthanized after 6 months of HFHC feeding. ApoE expression in these sacrificed pig livers was also examined by western blot with β-tubulin as an internal control. Consistent with serum ApoE levels, ApoE expression was elevated in the ApoE m/m group and absent in the ApoE −/− group compared to WT pigs (Fig. 4D). The ApoE −/− pigs  displayed a significantly higher percentage of aortic surface area covered with lesions than ApoE m/m pigs, namely a 2.1-fold increase compared with the ApoE m/m pigs (Fig. 5). However, atherosclerotic lesions were barely seen in the aorta and coronary arteries of WT pigs even when fed a HFHC diet. Interestingly, coronary area covered with atherosclerotic lesion was comparable between ApoE m/m and ApoE −/− pigs fed a HFHC diet (Fig. 5). The aorta from WT, ApoE m/m and ApoE −/− pigs were further stained to assess the presence of atherosclerosis. WT pigs on HFHC feeding exhibited no significant pathological alterations, whereas ApoE m/m pigs had appreciable aorta areas of fatty streaks, intimal thickening as well as apparent foam cell accumulation as indicated by CD68 staining. The ApoE −/− pigs showed progressive atherosclerotic lesions of fibrous plaque, which was symbolic of fibrosis as well as hyaline degeneration in aorta (Fig. 6).

DISCUSSION
Pigs have advantages over rodent models in recapitulating the human pathophysiology of atherosclerosis because they are similar to humans with regard to vascular size and physiology. Modeling atherosclerosis in minipigs has been hampered by a low propensity to atherosclerosis even with long-term feeding with atherogenic diets. Genetically modified pig models have been considered promising in solving the problem. However, utilization of genetically modified pigs as atherosclerosis models is still in its infancy, with the exception of PSCK9 transgenic pigs created by transposon integration and LDLR KO pigs generated by homologydirected recombination (HDR)-based methods (Al-Mashhadi et al., 2013;Davis et al., 2014;Li et al., 2016). Here, we successfully established novel porcine models of accelerated atherosclerosis in Bama miniature pigs by disrupting the endogenous ApoE gene using the CRISPR/Cas9 system. These pigs exhibited moderate, yet consistently, elevated serum cholesterol levels when fed a standard chow diet, and showed severe hypercholesterolemia and developed aortic and coronary atherosclerotic lesions when fed a HFHC diet. With high mutagenesis efficiency, CRISPR/Cas9 technology has been extensively used for gene targeting in pigs (Chen et al., 2015;Estrada et al., 2015;Zhang et al., 2017). Consistent with previous studies, high gene-targeting efficiency of the CRISPR/Cas9 system was achieved in PFFs: 11 and 28 of 48 colonies were identified as monoallelic and biallelic ApoE modified, respectively, which was much higher than that garnered with HDR-based ApoE targeting in rodents (Piedrahita et al., 1992;Plump et al., 1992). A total of 17 homozygous ApoE KO piglets were produced by a single round of SCNT and no off-target mutagenesis was detected in them, supporting that the high efficiency and relative simplicity of the CRISPR/Cas9 technology can greatly facilitate the efficient production of gene KO pigs. Due to its preference for the errorprone non-homologous end joining (NHEJ) pathway, the CRISPR/ Cas9 system has the advantage of being able to induce multiple gene modifications. Founder pigs with −4 or −5 bp deletions in the ApoE targeting site were generated and showed identical serum lipid profiles. Thus, we suspected that epigenetic status had little impact on the phenotype of the cloned founder pigs.
Hypercholesterolemia has long been implicated in atherosclerosis. Similar to PCSK9 transgenic pigs and LDLR KO pigs, the ApoE −/− pigs showed increased plasma cholesterol levels on a low-fat diet, resembling the human familial hypercholesterolemia phenotype. Interestingly, when fed a HFHC diet for 3 months, total plasma cholesterol levels plateaued to ∼40 mM in ApoE −/− pigs, but only reached ∼20-25 mM in PCSK9 transgenic and LDLR KO pigs. In addition, HDL cholesterol levels were dramatically increased in  ApoE −/− pigs, whereas the PCSK9 transgene and LDLR KO did not exert statistically significant effects on HDL cholesterol levels as compared to WT controls. Contrary to relatively low triglyceride levels in PCSK9 and LDLR KO pigs, hypertriglyceridemia was observed in ApoE −/− pigs by HFHC feeding. Thus, the ApoE −/− pigs could be beneficial for delineating the underlying pathophysiology of hypercholesterolemia and atherosclerosis because their serum lipid profiles differ from existing pig models, which cannot sufficiently recapitulate human disease. More recently, a line of ApoE/LDLR KO pigs generated by CRISPR/Cas9-mediated gene targeting has been reported (Huang et al., 2017). Surprisingly, cholesterol levels in those ApoE/LDLR KO pigs were lower than those of LDLR KO pigs reported previously (Davis et al., 2014;Li et al., 2016) or ApoE KO pigs in this study. Moreover, ApoE protein was still present in the ApoE/LDLR KO pigs and no atherosclerosis phenotype was reported. Considering that those pigs harbored a −18 bp deletion in the ApoE gene and +1/−15 bp indels in the LDLR gene, we speculated that IF-mutated ApoE and LDLR proteins did not completely lose their functions. Thus, the ApoE −/− pigs established in this study are the first ApoE KO pig model with severe atherosclerosis phenotypes.
Although the results presented here are encouraging, there were some limitations to this study. First, we did not investigate the advanced stages of atherosclerosis, such as plaque rupture or thrombosis. Thus, a study with a longer duration is needed to establish the full spectrum of atherosclerotic plaque progression from onset to rupture. Second, this study also did not take into account the effects of gender and puberty on phenotype because only male ApoE −/− pigs were investigated and development of coronary atherosclerosis was observed in the ApoE −/− pigs on HFHC feeding for 6 months, when they have reached puberty. Finally, in addition to its implication in lipoprotein metabolism and cardiovascular disease, ApoE exerts pleiotropic roles in Alzheimer's disease (Shi et al., 2017;Yu et al., 2014), immunoregulation (Braesch-Andersen et al., 2013) and cognition (Davies et al., 2014). However, the developmental consequences of ApoE depletion in other tissues/organs were not evaluated, being beyond the scope of this study.
In conclusion, ApoE −/− pigs fed a HFHC diet showed severe hypercholesterolemia and developed progressive atherosclerotic lesions. These pigs could serve as ideal large animal models for the elucidation of ApoE gene functions and translational studies of atherosclerosis.

Ethics statement
WT mature Landrace gilts and WT Bama minipigs were purchased from Chia Tai Co., Ltd (Huaiyin, China) and housed in a large animal facility affiliated with Nanjing Medical University, Nanjing, China. Standard chow diets were purchased from Chia Tai Co., Ltd and the HFHC diet was purchased from Xietong Organism Co., Ltd (Nanjing, China). Standard pig husbandry procedures were applied to all animals. Embryo transplantation and sacrifice procedures were conducted under isoflurane anesthesia to minimize animal suffering. All animal experiments were carried out in accordance with the guidelines approved by the Institutional Animal Care and Use Committee (IACUC) of the Nanjing Medical University.

T7E1 cleavage assay
PFFs were isolated from Chinese Landrace fetuses at day 35 of gestation using 200 U/ml collagenase (Invitrogen) and 25 kU/ml DNaseI (Invitrogen), and cultured as previously described (Chen et al., 2015). PFFs transfected with or without Cas9-sgRNA targeting plasmids (as described above) were cultured for 48 h. Genomic DNA was extracted using a DNA extraction kit (TianGen, Beijing, China) and the genomic regions surrounding the CRISPR/Cas9 target site were PCR amplified. PCR primers used are as follows: 5′-CCTCAGGTGGTCTAGGTTGG-3′ and 5′-TTTTGCAATGGAGGAGTGGC-3′ for sgRNA1; 5′-CGCCTC-TCTCTGTTCATTGC-3′ and 5′-TTTTGCAATGGAGGAGTGGC-3′ for sgRNA2. The PCR conditions were as follows: 95°C for 5 min, followed by 30 cycles of 95°C for 10 s, 60°C for 30 s, and 72°C for 40 s, and a final 72°C for 7 min. A total of 250 ng of the purified PCR product was mixed with NEB Buffer 2, denatured, and annealed to allow formation of heteroduplexes using the following conditions: 95°C for 5 min, 95°C to 85°C ramping at −2°C/s, 85°C to 25°C at −0.1°C/s, and 4°C hold. After reannealing, the products were digested with 1 μl of T7 endonuclease I (NEB, Beverly, MA, USA) at 37°C for 15 min and then run on a 2% agarose gel stained with ethidium bromide.

PFF transfection and selection
Approximately 5 μg of the ApoE Cas9-sgRNA targeting plasmid was co-transfected with 1 μg of the neomycin expression plasmid (pCMV-tdTomato) into 1×10 6 early passage of PFFs using the Basic Fibroblast Nucleofection Kit (Amaxa Biosystems/Lonza, Cologne, Germany), according to the manufacturer's instructions. A total of 800 μg/ml of G418 (Gibco, Grand Island, NY, USA) was applied to the post-transfection cells 48 h later and maintained for 10-14 days thereafter. The G418-resistant colonies were seeded in 24-well plates and then passaged to 12-well plates. Approximately 1/5 of the cells of a single colony were lysed with NP-40 (55°C for 30 min, 95°C for 10 min) for PCR screening and 4/5 of the remaining cells were used for SCNT. The primers used in amplifying the target regions were as follows: forward: 5′-CCTCAGGTGGTCTAGGTTGG-3′, reverse: 5′-TTTTGCAATGGAGGA-GTGGC-3′. The PCR conditions were as follows: 95°C for 5 min, followed by 35 cycles of 95°C for 10 s, 60°C for 30 s, and 72°C for 50 s, and a final 72°C for 4 min. The PCR products were subcloned into a pMD18-T vector (Takara Clontech, Tokyo, Japan) according to the manufacturer's instructions. Fifteen to 20 individual clones were picked and sequenced.

SCNT and embryo transfer
The oocytes for SCNT were collected from ovaries purchased from a local slaughterhouse and cultured for 42-44 h in vitro. The mature oocytes were enucleated as described elsewhere . A single targeted PFF cell was transferred into the perivitelline space of the enucleated oocyte. After electrofusion of membranes between the donor cell and recipient cytoplast, the reconstructed embryos were cultured in embryo development medium at 38.5°C for 24 h. About 250 reconstructed oocytes were transferred into the uterus of an estrus-synchronized recipient gilt. Pregnancy was confirmed by ultrasound 30 days after transplantation and monitored until the perinatal period. All cloned piglets were born by normal delivery.

Serum lipid chemistry
Whole blood was drawn from the jugular veins of overnight-fasted pigs at the indicated time points into serum-separating tubes (BD Diagnostics, Franklin Lakes, NJ, USA). Sera were prepared by centrifugation at 500 rcf for 15 min at room temperature. The levels of total cholesterol, triglycerides, HDL cholesterol and LDL cholesterol were determined by Dian Diagnostics (Nanjing, China).

Histology analysis
The aorta and coronary arteries were isolated from overnight-fasted pigs and fixed in 10% buffered formalin. Sudan IV was used to stain the atherosclerotic lesions. Images were collected on an Olympus FSX100 microscope (Olympus, Tokyo, Japan). The stained area was quantified using ImageJ software (NIH, Bethesda, MD, USA) and expressed as a percentage of the total image area. The aorta was divided into 1-cm segments taken perpendicular to the direction of blood flow and embedded in paraffin. H&E, Masson's trichrome and immunohistochemistry staining were conducted on 5 μm sections to confirm lesion details as described previously (Chen et al., 2018). CD68 monoclonal antibody (KP1; 1:100; Thermo Fisher Scientific) and HRP-conjugated goat anti-mouse secondary antibody (1:500; Proteintech, Rosemont, IL, USA) were used for immunostaining.

Statistical methods
All data are expressed as the means±s.d. P-values were determined using a 2-tailed Student's t-test or 1-way analysis of variance (ANOVA) with a Tukey's post hoc correction for multiple group comparisons. A P-value of less than 0.05 was considered statistically significant.