Induction of Atherosclerosis in Mice and Hamsters Without Germline Genetic EngineeringNovelty and Significance
Rationale: Atherosclerosis can be achieved in animals by germline genetic engineering, leading to hypercholesterolemia, but such models are constrained to few species and strains, and they are difficult to combine with other powerful techniques involving genetic manipulation or variation.
Objective: To develop a method for induction of atherosclerosis without germline genetic engineering.
Methods and Results: Recombinant adeno-associated viral vectors were engineered to encode gain-of-function proprotein convertase subtilisin/kexin type 9 mutants, and mice were given a single intravenous vector injection followed by high-fat diet feeding. Plasma proprotein convertase subtilisin/kexin type 9 and total cholesterol increased rapidly and were maintained at high levels, and after 12 weeks, mice had atherosclerotic lesions in the aorta. Histology of the aortic root showed progression of lesions to the fibroatheromatous stage. To demonstrate the applicability of this method for rapid analysis of the atherosclerosis susceptibility of a mouse strain and for providing temporal control over disease induction, we demonstrated the accelerated atherosclerosis of mature diabetic Akita mice. Furthermore, the versatility of this approach for creating atherosclerosis models also in nonmurine species was demonstrated by inducing hypercholesterolemia and early atherosclerosis in Golden Syrian hamsters.
Conclusions: Single injections of proprotein convertase subtilisin/kexin type 9–encoding recombinant adeno-associated viral vectors are a rapid and versatile method to induce atherosclerosis in animals. This method should prove useful for experiments that are high-throughput or involve genetic techniques, strains, or species that do not combine well with current genetically engineered models.
Germline genetically engineered animals, mainly Apoe−/− or Ldlr−/− mice, have become the preferred models for studying atherosclerosis.1,2 However, several powerful genetic techniques are impractical to combine with these models because of the extensive amount of breeding required. These include conditional and inducible knockouts of candidate disease genes, high-throughput gene targeting,3 mapping of susceptibility genes in complex mouse crosses,4,5 and analysis of the atherosclerosis-modulating effect of polygenic traits, such as accelerated aging or type 2 diabetes mellitus.6,7 Furthermore, some aspects of human physiology may be better modeled in other small animals than mice.
Editorial, see p 1672
In This Issue, see p 1671
Moderate hypercholesterolemia and small foam cell lesions can be achieved in some strains of mice by feeding them a cholate-supplemented diet, but cholate has pleiotropic effects that diminish its usefulness even when the restriction to foam cells is acceptable.8 The present study describes an alternative and more effective method for inducing atherosclerosis that circumvents the need for germline-encoded models. Proprotein convertase subtilisin/kexin type 9 (PCSK9) binds hepatic low-density lipoprotein (LDL) receptors directing them for degradation in lysosomes, and PCSK9 hyperactivity increases LDL levels in humans and mice.9 Here, we show that a single injection of recombinant adeno-associated virus (rAAV) encoding gain-of-function mutant forms of PCSK9 is sufficient to induce atherosclerosis in mice and hamsters.
An expanded Methods section is available in the Online Data Supplement.
Materials and Animals
rAAV plasmids were constructed to encode gain-of-function forms of murine or human PCSK9 (pAAV/D377Y-mPCSK9 or pAAV/D374Y-hPCSK9; Online Figure I). For both vectors, expression of PCSK9 was driven by an efficient liver-specific promoter. Viral vectors in serotype 8 capsids (rAAV8-D377Y-mPCSK9 and rAAV8-D374Y-hPCSK9) were produced by the UNC Vector Core (Chapel Hill, NC). Female C57BL/6NTac mice were from Taconic, female Ldlr−/− mice (B6.129S7-Ldlrtm1Her/J) from Jackson Laboratories, and male Golden Syrian Hamsters from Harlan Laboratories. Breeding pairs of C57BL/6-Ins2Akita/J and C57BL/6J mice were from Jackson Laboratories, and male offspring were used for experiments. The Danish Animal Experiments Inspectorate approved all experiments.
Viral vectors were delivered via a single tail vein injection in mice and a single intraperitoneal injection in hamsters, and animals were challenged with either Western-type diet containing 21% fat and 0.21% cholesterol or cholate-supplemented Paigen diet containing 16% fat, 1.25% cholesterol, and 0.5% sodium cholate.
Aortic arch atherosclerotic lesion area (down to the supreme intercostal artery) was quantified en face after Oil Red O staining. Plaque morphology was examined in aortic root cross sections.
Hypercholesterolemia Induced by PCSK9 Gene Transfer
Hydrodynamic injections of naked D377Y-mPCSK9– and D374Y-hPCSK9–encoding plasmids, a nonviral gene transfer approach, caused transient increase of plasma mPCSK9 and hPCSK9, respectively, in C57BL/6NTac mice (n=4 in each group; Online Figure IIA and IIB). This resulted in a temporally shifted phase of hypercholesterolemia (Online Figure IIC and IID), which, however, was not maintained long enough to cause atherosclerosis.
Tissue tropism for AAV serotype 8 was visualized in C57BL/6NTac mice by bioluminescence using the luciferase reporter gene (rAAV8-CMV-Luc, 5.0×1011 vector genomes). High levels of expression were seen in the upper abdomen consistent with efficient liver transduction (Figure 1A). rAAV8-CMV-Luc transduction did not increase plasma cholesterol levels, alanine aminotransferase activity, or serum amyloid A compared with saline-injected control mice (n=5 in each group; Online Figure IIIA–IIIC).
To achieve long-term hypercholesterolemia, C57BL/6NTac mice, 8 weeks of age, were given rAAV8-D377Y-mPCSK9 or rAAV8-D374Y-hPCSK9 at 3 different dosages (2.0×1010, 1.0×1011, or 5.0×1011 vector genomes) and fed either Western-type or Paigen diet. Ldlr−/− and saline-injected control mice were treated similarly. All mice thrived well (Online Table I).
D377Y-mPCSK9 increased rapidly, peaking at day 56, and was stably maintained at high levels (Figure 1B). This resulted in strong reduction of hepatic LDL receptor levels (Figure 1C) and significant dose-dependent hypercholesterolemia on both diets (Figure 1D and 1E). As expected, cholesterol levels were higher on Paigen diet. Results with the rAAV8-D374Y-hPCSK9 vector were similar (Online Figure IV).
Size-exclusion chromatography showed that low-dose rAAV8-D374Y-PCSK9 vector injection caused an isolated increase in LDL-sized lipoproteins on Western-type diet, whereas more very-low-density lipoprotein (VLDL)-sized lipoproteins were seen at higher D374Y-hPCSK9 levels and in Ldlr−/− mice (Online Figure V). On Paigen diet, most plasma cholesterol was carried in VLDL-sized lipoproteins in all groups.
Plasma alanine aminotranferase activity and serum amyloid A were increased similarly in vector-injected and Ldlr−/− groups compared with controls (Online Figure VI), indicating no hepatocellular injury or systemic inflammation caused by rAAV8 transduction alone.
Atherosclerosis Induced by PCSK9 Gene Transfer
Mice were euthanized 12 weeks after rAAV8 vector injection to quantify atherosclerosis (Figure 2A). Controls had no or minimal lesions on either diet, whereas all rAAV8-injected mice developed atherosclerosis in a dose-dependent manner (Figure 2B). Similar results were seen with the rAAV8-D374Y-hPCSK9 vector (Online Figure VII).
Histological analysis of aortic root lesions showed advanced plaque development with foam cells, smooth muscle cells, and fibrous tissue (Figure 2C–2F). All, except 1 rAAV8-D377Y-mPCSK9–injected (high dose) and 1 Ldlr−/− mouse, exhibited necrotic core formation.
Applications of the Method
To show how the method can be used to examine the atherosclerosis susceptibility of a mouse strain without intercrossing with Apoe−/− or Ldlr−/− mice, we injected rAAV8-D377Y-mPCSK9 (1.0×1011 vector genomes) in 26-week-old diabetic Akita mice with the Ins2Akita mutation and analyzed atherosclerosis after 10 weeks on Western-type diet. Consistent with previous studies in the Ldlr−/− background,10 we found augmented hypercholesterolemia and atherosclerosis compared with wild-type littermates (Figure 3). In this particular example, the effect of diabetes mellitus may have been underestimated because the level of plasma mPCSK9 was found to be lower in diabetic mice compared with controls (21 300 versus 56 000 ng/mL; n=4 both groups; P<0.03; Mann–Whitney). This will probably rarely be the case but serves to illustrate that analyzing for potential differences in the PCSK9 level obtained in groups is important when applying the atherosclerosis-inducing vector.
To provide an example of how the method can be used to create atherosclerosis models in nonmurine species, we injected Golden Syrian hamsters with a single intraperitoneal dose of the rAAV8-D374Y-hPCSK9 vector at 10 weeks of age (1.2×1012 vector genomes). After 10 days, D374Y-hPCSK9 was efficiently expressed in plasma and the level stayed elevated (>2000 ng/mL) throughout the study (Online Figure VIII). Marked hypercholesterolemia developed on Western-type diet (Figure 4A) with most cholesterol carried in VLDL- and LDL-sized lipoproteins (Figure 4B). After 12 weeks, all rAAV8-injected hamsters had scattered lesions in the aortic arch covering 3% to 6% of the surface area, whereas saline-injected controls showed no or minimal lesion development (Figure 4C and 4D). Histology of the aortic arch revealed foam cell lesions (Figure 4E).
Our experiments show that a single injection of PCSK9-encoding rAAV8 vectors is sufficient to induce atherosclerosis in mice and hamsters when combined with a cholate-free Western-type diet. This method should prove useful for high-throughput experiments or applications that involve genetic techniques, strains, or species that do not combine well with germline genetically engineered models. By circumventing the need for extensive breeding programs, it may also contribute to reducing the number of mice used for atherosclerosis research.
The efficiency of the technique stems from the noncell autonomous mechanism of action of PCSK9. PCSK9 is secreted into plasma and exerts its main function in the extracellular space.11 As a result, ubiquitous reduction of hepatic LDL receptor levels can be achieved even if only a subgroup of hepatocytes is transduced with the rAAV8 vector.
The extent of atherosclerosis induced by the viral vectors was less than that of Ldlr−/− mice on Western-type diet, but more important for most research applications, lesions progressed to the fibroatheromatous stage and the coefficient of variation did not seem to be increased. If more extensive atherosclerosis is required, previous studies have shown that hepatic expression of rAAV8-encoded genes can be maintained far beyond the 12-week study period of the present study.12
Because AAV vectors have broad species tropism and the PCSK9 and LDLR genes are conserved across mammals,13 we hypothesized that the technique would not be constrained to the mouse, and we were able to induce hypercholesterolemia and early aortic atherosclerosis in Golden Syrian hamsters.
In conclusion, gain-of-function PCSK9-encoding rAAV8 vectors provide a versatile tool for experimental atherosclerosis research that overcomes some of the limitations of germline genetically engineered models and can be applied in mice and hamsters and potentially in other species.
We thank Dorte Wilhardt Qualmann, Lisa Maria Røge, Zahra Partovi Nasr, Mie Aarup, Charlotte Wandel, Marie Bek Lund, and Rasmus Otkjær Bak for technical assistance.
Sources of Funding
This work was supported by The Danish Heart Foundation, Aase og Ejnar Danielsens Fond, Fonden til Lægevidenskabens Fremme, and Snedkermester Sophus Jacobsen & Hustru Astrid Jacobsens Fond.
In February 2014, the average time from submission to first decision for all original research papers submitted to Circulation Research was 13.8 days.
The online-only Data Supplement is available with this article at http://circres.ahajournals.org/lookup/suppl/doi:10.1161/CIRCRESAHA.114.302937/-/DC1.
- Nonstandard Abbreviations and Acronyms
- low-density lipoprotein
- proprotein convertase subtilisin/kexin type 9
- recombinant adeno-associated virus
- Received October 27, 2013.
- Revision received March 26, 2014.
- Accepted March 27, 2014.
- © 2014 American Heart Association, Inc.
- Getz GS,
- Reardon CA
- Stylianou IM,
- Bauer RC,
- Reilly MP,
- Rader DJ
- Novella S,
- Dantas AP,
- Segarra G,
- Vidal-Gómez X,
- Mompeón A,
- Garabito M,
- Hermenegildo C,
- Medina P
- Kim JH,
- Stewart TP,
- Soltani-Bejnood M,
- Wang L,
- Fortuna JM,
- Mostafa OA,
- Moustaid-Moussa N,
- Shoieb AM,
- McEntee MF,
- Wang Y,
- Bechtel L,
- Naggert JK
- Getz GS,
- Reardon CA
- Maxwell KN,
- Breslow JL
- Zhou C,
- Pridgen B,
- King N,
- Xu J,
- Breslow JL
Novelty and Significance
What Is Known?
Gene-modified models, such as Apoe- or Ldlr-deficient mice, are widely used for studying hypercholesterolemia and atherosclerosis.
Hyperactivity of proprotein convertase subtilisin/kexin type 9, a key regulator of low-density lipoprotein receptor protein levels, causes hypercholesterolemia in mice and humans.
Recombinant adeno-associated virus vectors are efficient tools for liver-directed gene delivery and provide long-term gene expression.
What New Information Does This Article Contribute?
A single injection of recombinant adeno-associated virus encoding a proprotein convertase subtilisin/kexin type 9 gain-of-function mutant induces persistent hypercholesterolemia and atherosclerosis in mice and hamsters.
This technique provides an alternative method for the induction of atherosclerosis that circumvents the need for breeding with Apoe- or Ldlr-deficient mice.
Several genetic techniques are available for studying the pathophysiology of atherosclerosis in mice, including complex strain crosses and conditional knockout techniques. However, ready application of these methods with current atherosclerosis models is limited by the extensive breeding required to generate the mice. Here, we report that injection of recombinant adeno-associated viruses encoding a gain-of-function proprotein convertase subtilisin/kexin type 9 mutant induces persistent hypercholesterolemia and atherosclerosis in mice and hamsters. This method could be useful for high-throughput experiments or for studies of atherosclerosis in complex mouse strains. It could also be used to generate nonmurine models.