Hypomorphic Apolipoprotein E Mice

In creating an allelic variant of mouseApoe designed to resemble human apolipoprotein E4 (apoE4), we generated hypomorphic apoE (hypoE) mice that express only ∼5% of normal apoE mRNA levels in all tissues. Insertion of aneo cassette flanked by loxP sites in the third intron of Apoe reduced expression of the Arg-61 allelic variant in hypoE mice and resulted in plasma apoE levels that were ∼2–5% of normal. Unlike other mouse models with low levels of circulating apoE, hypoE mice had a nearly normal lipoprotein cholesterol profile when fed a chow diet. Further reduction of apoE expression in hypoE/Apoe −/− heterozygous mice led to an increase in remnant lipoprotein-associated cholesterol levels, demonstrating that hypoE mice express close to the threshold level of Arg-61 apoE required for a normal lipoprotein profile. Unlike wild type mice, hypoE mice were susceptible to diet-induced hypercholesterolemia, which was fully reversed within 3 weeks after resumption of a chow diet. In Mx1-Cre transgenic hypoE mice, plasma apoE levels returned to normal within 10 days after gene repair and removal of the neo cassette following induction of Cre recombinase. HypoE mice provide the opportunity for conditional gene repair by crossing with inducible or lineage/cell type-specific Cre transgenic mice, generating new models to dissect the roles of apoE in atherosclerosis regression, immunoregulation, and neurodegeneration.

In creating an allelic variant of mouse Apoe designed to resemble human apolipoprotein E4 (apoE4), we generated hypomorphic apoE (hypoE) mice that express only ϳ5% of normal apoE mRNA levels in all tissues. Insertion of a neo cassette flanked by loxP sites in the third intron of Apoe reduced expression of the Arg-61 allelic variant in hypoE mice and resulted in plasma apoE levels that were ϳ2-5% of normal. Unlike other mouse models with low levels of circulating apoE, hypoE mice had a nearly normal lipoprotein cholesterol profile when fed a chow diet. Further reduction of apoE expression in hypoE/Apoe ؊/؊ heterozygous mice led to an increase in remnant lipoprotein-associated cholesterol levels, demonstrating that hypoE mice express close to the threshold level of Arg-61 apoE required for a normal lipoprotein profile. Unlike wild type mice, hypoE mice were susceptible to diet-induced hypercholesterolemia, which was fully reversed within 3 weeks after resumption of a chow diet. In Mx1-Cre transgenic hypoE mice, plasma apoE levels returned to normal within 10 days after gene repair and removal of the neo cassette following induction of Cre recombinase. HypoE mice provide the opportunity for conditional gene repair by crossing with inducible or lineage/cell type-specific Cre transgenic mice, generating new models to dissect the roles of apoE in atherosclerosis regression, immunoregulation, and neurodegeneration.
Apolipoprotein E (apoE) 1 is an important structural and functional protein component of lipoproteins that plays a prominent role in lipid metabolism in plasma and in the central nervous system (1,2). As a high affinity ligand for the low density lipoprotein (LDL) receptor, the LDL receptor-related protein, and heparan sulfate proteoglycans, apoE mediates the uptake of plasma remnant lipoproteins by the liver (3,4). In addition, apoE participates in diverse biological processes, such as intracellular cholesterol utilization (5), cell growth (6), immunoregulation (7)(8)(9), and neuronal growth and repair (2).
Tissue-specific control elements in the Apoe gene restrict its expression to hepatocytes (10), astrocytes (11), skin fibroblasts (12), adipocytes, and macrophages (13). Hepatocyte-derived apoE, the major source of plasma apoE (14), is responsible for receptor-mediated uptake of remnant lipoproteins in the liver by the secretion-capture pathway (15,16). ApoE secreted by hepatocytes into the space of Disse associates with incoming remnant lipoproteins and with heparan sulfate proteoglycans bound to hepatic sinusoidal surfaces. This local enrichment in apoE facilitates remnant clearance through receptor-mediated processes. In the brain, astrocytes are the major source of apoE, which serves in lipid homeostasis in the central nervous system (17)(18)(19). Macrophage-derived apoE promotes remnant lipoprotein uptake and retards the development of atherosclerosis in Apoe Ϫ/Ϫ mice (20 -22). ApoE also participates in the regression of atherosclerosis (23,24), contributes to the production of very low density lipoprotein (VLDL) triglycerides (25,26), impairs VLDL-triglyceride lipolysis (27), and enhances the production of VLDL-apoB (28). In addition, apoE has been suggested to participate in the regulation of inflammatory immune responses that protect against bacterial infection (29) and to act as an antioxidant to protect against atherosclerosis (30).
The transplantation of wild type (WT) bone marrow into Apoe Ϫ/Ϫ mice as a source of non-liver-derived apoE demonstrated that levels of plasma apoE equivalent to 10% of normal are sufficient to reduce plasma cholesterol levels to a normal range (20,21,31). In transgenic Apoe Ϫ/Ϫ mice expressing WT apoE in the adrenal gland, 3% but not 1% of normal plasma apoE levels substantially reduced plasma cholesterol levels (32). However, in the bone marrow transplantation model, low levels of apoE failed to restore a normal plasma lipoprotein profile. Unlike WT mice, which transport ϳ75-80% of their plasma cholesterol in high density lipoproteins (HDL), Apoe Ϫ/Ϫ mice transplanted with WT bone marrow and expressing 2-5% of plasma apoE transport ϳ30%-40% of their plasma cholesterol in HDL (31).
Recently, we generated an allelic variant of murine apoE, Arg-61, by gene targeting (33). The targeting vector included a floxed neomycin (neo) cassette in the third intron to follow the mutation. Removal of the neo cassette by Cre-mediated recombination resulted in normal apoE expression levels. However, its retention resulted in hypomorphic apoE (hypoE) mice. Hy-poE mice express reduced levels of apoE mRNA (ϳ5% of normal) in all tissues examined, giving rise to ϳ2-5% of normal apoE levels in plasma. Other examples of hypomorphic genes created in mice by inserting a neo cassette into an intron have been described (34 -38). , and a fellowship from the Heart and Stroke Foundation of Canada (to R. L. R). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Here we report that the hypoE mice have a nearly normal lipoprotein profile when fed a chow diet, but they are very susceptible to diet-induced hypercholesterolemia. The hypercholesterolemia can be reversed in Mx1-Cre transgenic hypoE mice by removing the neo cassette following induction of Cre recombinase with polyinosinic-polycytidylic ribonucleic acid (pIpC). This induction results in restoration of normal levels of plasma apoE.
Thus, hypoE mice are a new model of reduced apoE expression that will provide additional insight into the physiological roles of apoE. Moreover, hypoE mice represent a unique opportunity to study the role of tissue-specific expression of apoE by using Cre-loxP technology.

EXPERIMENTAL PROCEDURES
Generation of a Hypomorphic Allele of Apoe-A sequence replacement gene-targeting strategy was previously used to substitute arginine for the mouse equivalent of human Thr-61 as described ( Fig. 1A) (33). Chimeric mice harboring a mutant Apoe allele, Apoe neoϩ , in which intron 3 contained a neo cassette flanked by loxP sites, were crossed with C57BL/6 female mice to generate Apoe neoϩ/WT mice. These heterozygous mice were intercrossed to generate Apoe neoϩ/neoϩ mice. The mice were weaned at 21 days of age and housed in a barrier facility with a 12-h light/12-h dark cycle. Unless otherwise noted, they were fed a chow diet containing 4.5% fat (Ralston Purina, St. Louis, MO).
Northern Blot Analysis of Total RNA-After extraction from several tissues and organs with Triazol reagent (Invitrogen), total RNA (ϳ20 g) was electrophoresed in a 1% agarose gel containing 20% formaldehyde, transferred to Hybond membrane (Amersham Biosciences), and hybridized to a mouse apoE cDNA probe labeled with [ 32 P]dCTP in Quickhyb solution (Stratagene, La Jolla, CA) at 65°C overnight. The blot was washed in 0.3ϫ standard sodium citrate (150 mM NaCl, 15 mM sodium citrate) and 0.1% SDS at 55°C for 1 h and exposed to x-ray film overnight. A second blot of identical samples run on the same gel was hybridized with a mouse ␤-actin probe. Signals were quantified with a phosphor imager and quantification software (Bio-Rad QUANTITY ONE).
Lipid and Lipoprotein Determination-Lipids and lipoproteins were measured in 8 -15-week-old male mice that had been fasted for 4 h, anesthetized, and bled by retro-orbital puncture. Lipoproteins were fractionated by fast performance liquid chromatography (FPLC) on a Superose 6 column (Amersham Biosciences), and plasma was examined by agarose gel electrophoresis (Universal Gel/8, Helena Laboratories, Beaumont, TX). Cholesterol and triglyceride levels in plasma and FPLC fractions were determined with colorimetric assays (Spectrum (Abbott) and Triglycerides (Roche Molecular Biochemicals), respectively). Sta-tistical analysis was performed with the nonparametric Mann-Whitney test.
ApoE and ApoB Quantitation-Fasted mouse plasma was subjected to SDS-PAGE with 10 -20% or 4 -15% gels and transferred to nitrocellulose. Western blotting was performed with rabbit antisera against mouse apoE (33) and apoB. Polyclonal antisera against mouse apoB100 and apoB48 were raised using mouse LDL (d ϭ 1.006 -1.063 g/ml) isolated from Ldlr Ϫ/Ϫ mouse plasma by sequential density ultracentrifugation. New Zealand White rabbits were immunized with 100 g of purified mouse LDL emulsified in complete Freund's adjuvant. Rabbits were boosted twice with antigen emulsified in incomplete Freund's adjuvant.
Western blots were incubated with primary antibodies at a dilution of 1:5000, and bound primary antibody was detected by a horseradish peroxidase-conjugated anti-rabbit antibody (Invitrogen). Signals were generated by incubating membranes with chemiluminescent reagent (Amersham Biosciences) and exposing them to x-ray film (Eastman Kodak Co.). Signals were quantified with a phosphor imager and quantification software (Bio-Rad QUANTITY ONE).

RESULTS
Generation of a Hypomorphic Apoe Allele-The hypoE mice expressing reduced levels of apoE were generated by homologous recombination in embryonic stem cells. A neo cassette flanked by loxP sites was inserted into Apoe intron 3 to help follow the replacement of the human equivalent of Thr-61 by an arginine (Fig. 1A) (33). Correctly targeted embryonic stem cell clones were injected into blastocysts, and chimeric mice were crossed with C57BL/6 mice to generate mice that were heterozygous for the neo cassette (Apoe WT/neoϩ ). Heterozygous mice were intercrossed to produce homozygous hypoE mice (Apoe neoϩ/neoϩ ).
Characterization of HypoE Mice-The apoE mRNA levels in the liver, brain, and spleen in hypoE mice were ϳ5% of those in WT mice, suggesting a common mechanism for the reduced expression of the targeted allele (Fig. 1B). Other organs and tissues that normally express low levels of apoE gave barely detectable signals. The plasma apoE levels in male hypoE mice were ϳ2-5% of those in WT mice (Fig. 1C). Female mice expressed similar levels, and male hypoE mice were used to characterize the lipoprotein phenotype. In chow-fed mice, the total plasma cholesterol and triglyceride levels were slightly higher in hypoE mice than in WT mice (98 Ϯ 14 versus 65 Ϯ 5 mg/dl, n ϭ 7, p ϭ 0.003; 49 Ϯ 14 versus 26 Ϯ 8 mg/dl, n ϭ 7, p ϭ FIG. 1. Generation of hypoE mice. A, targeting strategy. Homologous recombination of the targeting vector with the Apoe locus results in inserting a neo cassette into intron 3 and converts the equivalent of human Thr-61 to an arginine as previously described (33). B, Northern blot demonstration of reduced expression of the Apoe hypomorphic allele. Total RNA from the liver, brain, and spleen was isolated from WT mice and hypoE mice and subjected to Northern blot analysis with apoE cDNA probe or a control mouse ␤-actin probe. C, Western blot of mouse plasma; comparison of hypoE mouse plasma (1 l) with serially diluted WT mouse plasma. 0.007, respectively). However, the lipoprotein cholesterol profiles were similar in hypoE, WT, and Arg-61 Cre-deleted mice (Fig. 2). HypoE mice had slightly more cholesterol in the VLDL, intermediate density lipoproteins, and LDL lipoprotein fractions than the WT mice. Most of the plasma cholesterol in the hypoE mice was associated with HDL, as in WT mice (65-70% versus 75-80% in WT mice) (Table I). This finding is in contrast to Apoe Ϫ/Ϫ mice engineered to express levels of apoE ϳ2-5% of WT (31). In these mice, a significant portion of plasma cholesterol is associated with VLDL and LDL lipoproteins, and only ϳ30 -40% of plasma cholesterol is associated with HDL (31). Agarose gel electrophoresis confirmed the nearly normal lipoprotein profile in hypoE mice indicated by FPLC (Fig. 2,  inset).
As shown by SDS-PAGE Western blot analysis, hypoE mice had lower levels of apoB100 and higher levels of apoB48 in plasma than WT mice (Fig. 3). The hypoE mice and Apoe Ϫ/Ϫ mice had very similar levels of apoB100, approximately 75% lower than in WT mice as judged by densitometry. In contrast, hypoE mice had 8-fold more apoB48 than WT mice.
The effect of a further reduction in apoE expression on lipoprotein metabolism was examined by crossing hypoE and Apoe Ϫ/Ϫ mice. Heterozygous Apoe neoϩ/Ϫ mice did not have sig-nificantly higher plasma cholesterol and triglyceride levels than hypoE mice (114 Ϯ 32 versus 98 Ϯ 14 mg/dl, n ϭ 7, p ϭ 0.164, and 58 Ϯ 16 versus 49 Ϯ 14 mg/dl, n ϭ 7, p ϭ 0.47). However, Apoe neoϩ/Ϫ mice carried more plasma cholesterol as remnant lipoproteins than hypoE mice (Table I). Taken together, these results suggest that hypoE mice express close to the lower limit of apoE required to maintain a nearly normal lipoprotein profile.
Diet-induced Hypercholesterolemia in HypoE Mice-Next, the susceptibility of hypoE mice to diet-induced hypercholesterolemia was determined. On a high fat Western diet (21% fat, 0.12% cholesterol), hypoE mice had higher plasma levels of cholesterol (238 Ϯ 63 versus 133 Ϯ 25 mg/dl, n ϭ 6, p ϭ 0.003) and triglyceride (93 Ϯ 23 versus 22 Ϯ 13 mg/dl, n ϭ 6, p Ͻ 0.001) than WT mice. FPLC analysis revealed an increase of cholesterol, mainly in the HDL and VLDL fractions, in hypoE mice (Fig. 4). In contrast, the Paigen diet (16% fat, 1.25% cholesterol, 0.5% cholic acid) markedly increased the accumulation of all classes of remnant lipoproteins in hypoE mice, and their plasma cholesterol and triglyceride levels were much higher than those of WT mice (1146 Ϯ 141 versus 227 Ϯ 31 mg/dl, n ϭ 6, p Ͻ 0.001 and 61 Ϯ 33 versus 18 Ϯ 7 mg/dl, n ϭ 6, p ϭ 0.01, respectively). Similar responses to these two diets have been observed in other mouse models of low level apoE expression (20,21). The hypercholesterolemia in hypoE mice was fully reversed 3 weeks after resumption of a chow diet (Fig.  5). These results demonstrate that hypoE mice are far more susceptible to diet-induced hypercholesterolemia than WT mice and that very different lipoprotein profiles and plasma lipid levels can be obtained using different diet formulations.
The Effects of Gene Repair on Plasma Lipoprotein Metabolism in HypoE Mice-We have previously demonstrated that removal of the neo cassette from the targeted Apoe allele by crossing hypoE mice with Cre-deleted transgenic mice resulted in organ-wide reversal of the hypomorphic effect (33). Levels of apoE mRNA in Cre-deleted mice are identical to WT in the liver, brain, and spleen, and plasma lipid levels and lipoprotein   profiles are similar (33). To test the effects of conditional gene repair on plasma lipid metabolism, we crossed hypoE mice with inducible Mx1-Cre transgenic mice. Induction of these mice has been demonstrated to lead to Cre-mediated recombination in the liver and bone marrow (39,40). Uninduced Mx1-Cre transgenic hypoE mice had plasma apoE levels identical to those of nontransgenic hypoE mice, and they were equally susceptible to diet-induced hypercholesterolemia (not shown). A single intraperitoneal injection of 250 g of pIpC in Mx1-Cre transgenic hypoE mice increased plasma apoE levels within 2 days, and normal plasma apoE levels were restored within 10 days (Fig.  6). Restoration of plasma apoE levels completely reversed dietinduced hypercholesterolemia, resulting in a plasma cholesterol level of 65 mg/dl and a WT lipoprotein profile in two separate mice (not shown). DISCUSSION This study shows that introducing a neo cassette flanked by loxP sites into Apoe intron 3 to create a human apoE4-like allelic variant (Arg-61 apoE) results in mice with reduced apoE mRNA expression in all tissues. The generation of a nonproductive splice variant from a major portion of primary RNA transcripts has been proposed as an explanation for the hypomorphic effect in other models (35)(36)(37). Despite having plasma apoE levels of ϳ2-5% of normal, hypoE mice have a lipoprotein cholesterol profile similar to that of WT mice. Significantly, most of the plasma cholesterol (65-70%) is associated with HDL, as in WT mice. Unlike WT mice, however, hypoE mice are highly susceptible to diet-induced hypercholesterolemia, which is rapidly reversed when the mice are fed a normal chow diet. The reversal of remnant accumulation is consistent with our conclusion that 2-5% of normal plasma apoE levels can support effective remnant clearance.
HypoE mice differ from previous mouse models with reduced plasma apoE levels. For example, several lines of transgenic Apoe Ϫ/Ϫ mice expressing WT mouse apoE in the adrenal gland have reduced plasma cholesterol levels with 3% but not with 1% of WT plasma apoE levels (32). The lipoprotein profile of the transgenic Apoe Ϫ/Ϫ mice expressing 3% of WT apoE was not reported in the study and therefore cannot be compared with hypoE mice. Apoe Ϫ/Ϫ mice engineered to express apoE from nonhepatic sources at ϳ2-5% of normal levels by bone marrow transplantation have reduced plasma cholesterol levels; however, only ϳ30 -40% of their plasma cholesterol is associated with HDL (31) versus 65-70% in hypoE mice and 75-80% in WT mice. We speculate that the presence of hepatocyte-derived apoE in the hypoE mice may explain the difference in remnant lipoprotein metabolism between hypoE mice and mice expressing low levels of plasma apoE from nonhepatic sources. The hypoE mouse model demonstrates that ϳ2-5% of normal plasma apoE is close to the threshold level of apoE required for normal lipoprotein metabolism in mice fed a chow diet. A potential contributing factor to this observation may be that the Arg-61 apoE is more effective in remnant clearance than WT apoE. However, we have demonstrated that plasma lipid levels and lipoprotein profiles are similar in WT and genetargeted mice expressing normal levels of Arg-61 apoE, suggesting that, if there are differences, they may be small (33). To evaluate the potential differences between the Arg-61 and WT apoE isoforms, we are currently generating WT hypoE mice by inserting a neo cassette flanked by loxP sites into intron 3 of the WT Apoe gene.
The susceptibility of hypoE mice to hypercholesterolemia induced by the two diets used was predictable, given the welldocumented accumulation of plasma cholesterol in Apoe Ϫ/Ϫ mice and in Apoe Ϫ/Ϫ mice expressing low levels of mouse or human apoE (20 -22, 31, 32, 41, 42). However, the hypoE mouse model provides the opportunity to produce a spectrum of lipoprotein and plasma cholesterol levels by dietary manipula- tion. Plasma cholesterol levels in hypoE mice increased by 2.5-fold on the high fat Western diet and by 8 -12-fold on the Paigen diet. As a result, plasma lipid and lipoprotein levels can be more tightly modulated by diet composition in hypoE mice than in WT mice.
Another interesting feature of hypoE mice was the skewing of plasma apoB100 and apoB48 levels. HypoE mice had lower levels of apoB100 and higher levels of apoB48 than WT mice. In hypoE mice, apoB100 lipoproteins are probably cleared more rapidly from the circulation by the LDL receptor, whereas the clearance of apoB48 lipoproteins, which depend on apoE, may be delayed (3,43). The LDL receptor-related protein also may fail to clear apoB48 remnants effectively due to limiting amounts of circulating apoE, resulting in increased levels of apoB48. Alternatively, the reduced levels of apoB100 in hypoE mice may result from decreased apoB100 secretion by the liver. Indeed, hepatic apoE expression has been reported to promote apoB secretion (28). To define better the mechanisms that lead to reduced levels of plasma apoB100 in hypoE mice, it will be necessary to determine the production rate of apoB in vivo and in primary cultures of hepatocytes from hypoE mice.
The hypoE mouse model also offers the possibility of taking advantage of emerging Cre-loxP technology. We demonstrate here that in Mx1-Cre transgenic hypoE mice, removal of the neo cassette by Cre-mediated recombination restored normal plasma apoE levels following induction of Cre recombinase. By crossing hypoE mice with tissue-specific Cre transgenic mice, it will be possible to restore normal levels of apoE expression in selected tissue. For example, crossing hypoE mice with Mac1-Cre transgenic mice obtained from Dr. G. Kollias (available at www.fleming.gr) will produce mice in which apoE expression is fully restored only in macrophages. These mice could help to elucidate the role of macrophage-derived apoE in plasma lipoprotein metabolism and in the prevention of atherosclerosis. Restoring normal plasma apoE levels in Mx1-Cre and in tissueand cell lineage-specific Cre transgenic hypoE mice, such as astrocyte-, oligodendrocyte-, and neuron-specific Cre, represents a unique way to study the biological roles of apoE in the brain. Achieving physiological and permanently sustained levels of apoE expression after Cre-mediated excision of the neo cassette represents an approach to study the contribution of apoE to atherosclerosis regression not available in current models. Moreover, when WT hypoE mice become available, isoform-specific differences between Arg-61 apoE and WT apoE will provide the means to determine the contribution of apoE4 domain interaction as a distinguishing feature between apoE4 and apoE3 in atherosclerosis regression.
In summary, we report the development of mice with reduced apoE expression or hypoE mice. The reversibility of the hypomorphic effect in hypoE mice provides the opportunity for an expanded study of the role of tissue-derived apoE in existing tissue-specific and inducible Cre transgenic mice (44) and in new lines as they become available.