Low Expression of the Apolipoprotein B mRNA–Editing Transgene in Mice Reduces LDL Levels but Does Not Cause Liver Dysplasia or Tumors

From the Gladstone Institute of Cardiovascular Disease (X.Q., M.E.B., J.B., I.L., T.L.I.), the Cardiovascular Research Institute (X.Q., J.B., T.L.I.), and the Department of Pathology (T.L.I.), University of California, San Francisco; and the Department of Pharmacology (S.Y.), Osaka City University Medical School, Osaka, Japan.

Correspondence to Xiaobing Qian, MD, PhD, Gladstone Institute of Cardiovascular Disease, 365 Vermont St, San Francisco, CA 94103. E-mail xiaobing_qian{at}quickmail.ucsf.edu

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
Abstract—Hepatic expression of apolipoprotein (apo) B mRNA–editing enzyme catalytic polypeptide 1 (APOBEC-1) has been proposed as a gene therapy approach for lowering plasma low density lipoprotein (LDL) levels. However, high-level expression of APOBEC-1 in transgenic mouse and rabbit livers causes liver dysplasia and hepatocellular carcinoma. To determine the physiological and pathological effects of low-level hepatic expression of APOBEC-1, we used a 52-kb rat APOBEC-1 genomic clone (RE4) to generate transgenic mice expressing low levels of APOBEC-1 (2 to 5 times those in nontransgenic mice). Liver function, liver histology, editing of apoB mRNA at the normal editing site (C⁶⁶⁶⁶), and abnormal editing at multiple sites (hyperediting) in these mice were compared with those in transgenic mice expressing intermediate (I-20) or high (I-28) levels of APOBEC-1 in the liver. Hyperediting of mRNA coding for the novel APOBEC-1 target 1 (NAT1) was also examined. In the high-expressing I-28 line, 50% of the mice had palpable tumors at 15 weeks of age, whereas in the intermediate-expressing I-20 line, 50% of the mice had evidence of liver tumors after 1 year. In contrast, low-expressing RE4 mice had normal liver function and histology and did not develop liver tumors when examined at 3 to 17 months of age. Moreover, hyperediting of apoB and NAT1 mRNA in the liver was robust in the I-20 mice but barely detectable in the RE4 mice. The low-level expression resulted in sufficient APOBEC-1 to edit essentially all apoB mRNA at the normal editing site, virtually eliminating apoB-100 and LDL in the plasma of RE4 mice. When RE4 mice were crossed with human apoB transgenic mice, which possess high plasma LDL concentrations, plasma LDL levels in the offspring were reduced to very low levels. These results indicates that long-term hepatic expression of APOBEC-1 at low levels sufficient to eliminate LDL does not cause apparent liver damage or liver tumors in transgenic mice. RE4 APOBEC-1 transgenic mice should prove valuable for studying the roles of apoB-containing lipoproteins in lipid metabolism and atherosclerosis.

Key Words: APOBEC-1 • cholesterol • NAT1 • RNA editing • tumorigenesis

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
Two forms of apoB are involved in normal lipoprotein metabolism.¹ ApoB-100, the full-length protein, is synthesized in the liver and is essential for the assembly and secretion of VLDLs,^{2 3} the precursors of LDLs. LDLs transport about two thirds of the cholesterol in the plasma and are the major atherogenic lipoproteins in humans. ApoB-100 is the major protein component of LDL and mediates their uptake by the LDL receptor. ApoB-100 is also a component of other atherogenic lipoproteins, such as Lp(a) and ß-VLDL. ApoB-48, the amino-terminal 48% of apoB-100, is synthesized in the small intestine of all mammalian species and is essential for the assembly and secretion of chylomicrons.³ ApoB-48 is formed by apoB mRNA editing of the same nuclear mRNA that encodes apoB-100.⁴ The apoB mRNA–editing catalytic polypeptide 1 (APOBEC-1) deaminates the cytidine in codon 2153 (C⁶⁶⁶⁶AA) to form a uridine. The resulting in-frame stop codon (U⁶⁶⁶⁶AA) causes premature translation termination and formation of apoB-48.⁵

ApoB mRNA editing has significant functional consequences on lipoprotein metabolism. ApoB-48, the product of this editing, lacks the LDL receptor–binding domain of apoB, which is located at the carboxyl-terminal half of apoB-100, and therefore cannot bind to the LDL receptor. However, unlike those containing apoB-100, lipoproteins containing apoB-48 possess multiple copies of apoE, which serves as a ligand to mediate the clearance of the particles through both the LDL receptor and the LDL receptor–related protein.⁶ Even though single copies of apoB-100 and apoE bind to the LDL receptor with similar affinities, lipoproteins with multiple copies of apoE bind to multiple LDL receptors, increasing the affinity of the lipoprotein-receptor interaction.⁷ As a result, apoB-48–containing particles are cleared from the circulation more rapidly (in a matter of minutes) than are LDLs (2 to 3 days), which contain only 1 apoB-100 molecule.⁸ There is also evidence that apoB-100–containing particles can be converted to LDL, whereas apoB-48–containing particles cannot.⁹ Because apoB-48 does not attach to apo(a), it is not involved in the formation of the atherogenic Lp(a).¹

Unlike humans and rabbits, which express APOBEC-1 in the small intestine only, rats and mice express APOBEC-1 in multiple tissues, including the liver.³ As a result of the hepatic expression of APOBEC-1, these species secrete both apoB-100- and apoB-48–containing lipoproteins from the liver.^{9 10 11} Because apoB-48–containing lipoproteins are cleared rapidly from the circulation, are not converted to LDL, and cannot form Lp(a),^{1 12} mice and rats have lower apoB-100 and LDL levels and are more resistant to diet-induced atherosclerosis than are most other species. This evidence suggests several hypotheses. First, increased hepatic editing activity in species that already partially edit apoB mRNA, such as mice and rats, would further lower LDL concentrations. Second, introducing APOBEC-1 into the liver of species that normally lack the enzyme, such as rabbits and humans, would lower plasma LDL concentrations. Third, hepatic expression of APOBEC-1 could be used therapeutically to reduce the levels of atherogenic lipoproteins, LDL, and Lp(a) and to prevent atherosclerosis.

Two approaches have been employed to determine whether the introduction of APOBEC-1 into the livers of experimental animals will reduce plasma levels of LDL and Lp(a). One method has been to use adenoviral vectors to produce short-term expression of exogenous APOBEC-1 in the livers of mice and rabbits. This approach has been described in 4 studies. In the first, expression of APOBEC-1 markedly but transiently decreased the concentrations of mouse apoB-100 and LDL in mice.¹³ In the second, APOBEC-1 reduced plasma cholesterol levels by 30% in the LDL receptor–deficient Watanabe heritable hyperlipidimic rabbit.¹⁴ In the third, APOBEC-1 expression decreased the plasma levels of human apoB-100 and Lp(a) in transgenic mice expressing human Lp(a).¹⁵ In the fourth, APOBEC-1 lowered plasma LDL in LDL receptor–knockout mice.¹⁶ These studies support the hypothesis that gene therapy with APOBEC-1 can lower plasma apoB-100 and LDL levels. However, because of the brief duration of transgene expression, the long-term effects of increased APOBEC-1 expression in the liver could not be assessed in these studies, and atherosclerosis studies could not be performed in these animal models.

Adopting another approach, we expressed high levels of rabbit APOBEC-1 cDNA in the livers of transgenic mice and rabbits.¹⁷ In these animals, apo-B mRNA was extensively edited in the livers, and the plasma levels of apo-B100 and LDL were reduced. However, massive overexpression of APOBEC-1 in the liver caused liver dysplasia and hepatocellular carcinomas. We hypothesized that the overexpression of APOBEC-1 caused the aberrant hepatic editing of other mRNAs, resulting in serious pathological changes. This hypothesis is supported by the discovery of apoB mRNA editing at multiple sites downstream of C⁶⁶⁶⁶ (hyperediting) and the hyperediting of a novel mRNA designated NAT1 (novel APOBEC-1 target 1) in the livers of animals overexpressing APOBEC-1.^{18 19} Functional studies have indicated that NAT1 is most likely a translation repressor,¹⁹ and aberrant editing in the livers of these transgenic animals caused missense mutation of NAT-1, greatly reducing the levels of this protein. Although it remains to be determined whether editing of NAT-1 caused the liver dysplasia and tumorigenesis in these animals, it is clear that high-level expression of APOBEC-1 in the liver can have detrimental effects.

To study regulation of the rat APOBEC-1 gene, we generated transgenic mice (RE4) expressing low levels of APOBEC-1 by using a rat APOBEC-1 genomic clone.²⁰ The tissue distribution and relative levels of APOBEC-1 transgene expression in various tissues in these mice closely resembled those of the rat and mouse APOBEC-1 gene.²⁰ In this study, we used these mice to determine whether long-term, low-level expression of APOBEC-1 would lower LDL levels without causing liver damage. Using these mice and the mice expressing intermediate (I-20) and high (I-28) levels of APOBEC-1,¹⁷ we investigated the correlation between the level of APOBEC-1 editing activity and pathogenesis in the liver and hyperediting of apoB and NAT1 mRNAs. We also investigated the effects of low-level APOBEC-1 expression on lipid and lipoprotein metabolism in the RE4 mice and in double-transgenic mice expressing both APOBEC-1 and human apoB-100.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Mice
Rat APOBEC-1 transgenic mice (RE4) were generated with a 52-kb genomic clone containing 4 kb of the 5' and 30 kb of the 3' flanking sequence of the rat APOBEC-1 gene.²⁰ RNase protection assays demonstrated that expression of the rat APOBEC-1 transgene closely mimicked the patterns of endogenous mouse and rat APOBEC-1 expression.²⁰ Transgenic mice overexpressing rabbit APOBEC-1 (I-20 and I-28) were generated with a full-length cDNA fragment of rabbit APOBEC-1 as described previously.¹⁷ Expression of the rabbit transgene was controlled by the promoter and hepatic control region of the human apoE gene; therefore, the transgenes were specifically expressed in the liver.¹⁷ All APOBEC-1 transgenic mice used in this study were hybrids of genetic strains C57BL/6 (25%), SJL (25%), and ICR (50%). Human apoB transgenic mice were generated with an insert from a P1 plasmid (p158) (J.B. et al, unpublished data). These mice were hybrids of genetic strains C57BL/6 and SJL. APOBEC-1/apoB double-transgenic mice were generated by mating RE4 mice with human apoB mice. F1 mice from this mating were used in this study. All animal experiments were carried out in compliance with guidelines for the humane treatment of animals from the National Institutes of Health and the University of California, San Francisco.

Determination of ApoB mRNA–Editing Activity in Mouse Liver S100 Extracts
Liver S100 extracts were prepared from age- and sex-matched nontransgenic mice, transgenic mice expressing the rat APOBEC-1 gene (RE4),²⁰ or transgenic mice expressing the rabbit APOBEC-1 cDNA (I-20 or I-28)¹⁷ as follows. Liver segments were homogenized in S100 buffer A (10 mmol/L HEPES, pH 7.9, 1.5 mmol/L MgCl₂, 10 mmol/L KCl, 0.2 mmol/L EGTA, 1 mmol/L EDTA, 1 mmol/L DTT, 0.15 mmol/L spermine, and 0.5 mmol/L spermidine). The homogenates were cleared of tissue debris by centrifugation at 4000g at 4°C for 10 minutes. A 0.11 volume of S100 buffer B (0.3 mol/L HEPES, pH 7.9, 30 mmol/L MgCl₂, and 1.4 mol/L KCl) was added to the cleared supernatant, and the mixture was centrifuged at 100 000g for 50 minutes at 4°C. The S100 extract was then dialyzed against buffer D (20 mmol/L HEPES, pH 7.9, 125 mmol/L KCl, 1 mmol/L EDTA, 0.2 mmol/L EGTA, 20% glycerol, and 1 mmol/L DTT) and the following protease inhibitors: 1 mmol/L PMSF, 1 mmol/L benzamidine, 10 µmol/L leupeptin, 10 µmol/L pepstatin, 10 U/mL aprotinin, and 20 µg/mL soybean trypsin inhibitor. S100 extract (100 µg) was incubated with 10 ng of synthetic substrate apoB mRNA at 30°C for 2 hours. The editing of the substrate apoB mRNA (354 nucleotides of rabbit apoB) was determined by primer extension as described previously.²¹

Histopathological Analysis
Mouse livers were fixed with phosphate-buffered 10% formalin (Fisher Scientific), embedded in paraffin, and sectioned by standard techniques. Sections were stained with hematoxylin/eosin and examined by light microscopy. All histological sections were graded blindly. Levels of liver enzymes in mouse plasma were determined by standard techniques in the Clinical Laboratories, San Francisco General Hospital Medical Center.

Analysis of ApoB and NAT1 mRNA Editing
Total RNA isolated from transgenic and control nontransgenic mouse livers was treated with DNase and reverse transcribed with a first-strand cDNA synthesis kit (Life Technologies). Selected fragments of apoB and NAT1 cDNA were amplified by PCR. Primers used to amplify mouse apoB were mouse M49 (5'-CTGAATG CATCTGACTGGGAGAGACAAGTAGCTG-3') and mouse M50 (5'-CGGATATGATCTGTTCGTCAAGC-3'). Primers used to amplify human apoB were human M49 (5'-CTGAATTCATTCAATT GGGAGAGACAAGTTTCAC-3') and human M50 (5'-CG GATATGATAGTGCTCATCAAGAC-3'). Primers used to detect the editing of multiple cytidines in apoB mRNA were M51 (5'-ATCATAACTACTTTTAATATACTG-3') for editing of C⁶⁶⁶⁶, M52 (5'-TTCATCAAGAATTTTTAACTTTTC-3') for editing of C⁶⁷³⁸ and C⁶⁷⁴³, and M54 (5'-TTTTTAAGTCATGTGGATCATAAT-3') for editing of C⁶⁶⁷⁵. Primers used to amplify NAT1 were D45620–32UP (5'-GGTGTGAACAAATGGTGAGAAT-3') and D45620–146LP (5'-TTTCAAGTATCACAATGTTTATTG-3'). PCR products were analyzed by primer extension with primer PE83 (5'-ACAAGTATATAAAATCAGGGCATG-3') to detect the C-to-U editing at C³⁷²⁵ and C³⁷³⁰ as described previously.^{18 19} Five mice from each group were analyzed. The editing of apoB and NAT-1 mRNAs was quantified with the Scananalytics Ambis Radioisotopic Imaging System. Results are shown as mean±SD.

Analysis of Plasma Lipids and Lipoproteins
Lipid levels were measured in fresh plasma from age- and sex-matched nontransgenic mice, transgenic mice expressing the rat APOBEC-1 gene (RE4), and double-transgenic mice expressing both rat APOBEC-1 and human apoB-100 (n=3 in each group). All mice were fed a chow diet and were fasted for 6 hours before blood samples were drawn by cardiac puncture. EDTA at a final concentration of 2.5 mmol/L was used as an anticoagulant. Plasma was obtained by centrifugation at 16 000g for 10 minutes at 4°C, and the samples were stored with 1 mmol/L PMSF.

To compare the apoB-100 and apoB-48 levels in the mice of different groups, plasma samples (or lipoprotein fractions isolated by ultracentrifugation) were analyzed on SDS-polyacrylamide gels followed by immunoblotting. An affinity-purified polyclonal antibody against rat LDL was used to detect mouse apoB, and monoclonal antibody 1D1 was used to detect human apoB.²²

Lipoproteins in 200 µL of plasma were fractionated by Superose 6 chromatography as described previously.²³ Major lipoprotein classes eluted from the Superose 6 column were pooled [fractions 17 to 19, VLDL; 20 to 23, IDL; 24 to 29, LDL and HDL₁; and 30 to 36, HDL] and concentrated to a volume of 50 µL with Centricon filters. Aliquots (3 µL) of concentrated lipoprotein fractions were separated on 1% agarose gels at 90 V for 45 minutes. The gels were stained for neutral lipids with fat red 7B (Sigma Chemical Co).

Cholesterol and triglycerides in 5-µL plasma sample or in a 100-µL sample from each 500-µL Superose 6 fraction were measured with a Spectrum lipid analyzer (Abbott Laboratories). Protein concentrations were determined by the method of Lowry et al.²⁴

	Results

Top Abstract Introduction Methods Results Discussion References

 
Correlation of Tumorigenesis and APOBEC-1 Levels in the Livers of Transgenic Mice
To compare the expression levels of APOBEC-1 in RE4, I-20, and I-28 mice, we measured the enzymatic activities of APOBEC-1 in mouse liver S-100 extracts with an in vitro editing assay. Hepatic APOBEC-1–mediated editing activity was 8 and 12 times higher in I-20 and I-28 mice, respectively, but only 3 times higher in RE4 mice than in nontransgenic mice (Figure 1A).

View larger version (20K): [in this window] [in a new window]   Figure 1. Correlation of hepatic apoB mRNA–editing levels and pathogenic effects. A, Editing activity in mouse liver S100 extracts. Liver S100 extracts from age- and sex-matched nontransgenic (Non-tg) mice, transgenic mice expressing rat APOBEC-1 (RE4), and lines of mice expressing rabbit APOBEC-1 (I-20 and I-28) (n=5 per group) were analyzed for apoB mRNA–editing activity by an in vitro editing assay and primer extension analysis. Data are mean±SD. B, Kaplan-Meier disease-free survival curves of transgenic mice expressing APOBEC-1. Internal organs of transgenic mice were examined at the indicated time points, and all grossly abnormal tissues were excised and examined histopathologically.

In the I-28 mice, which had the highest level of APOBEC-1 expression, 50% of the mice had visible liver tumors after only 15 weeks, and >80% had tumors within 6 months (Figure 1B). In the I-20 mice, hepatic dysplasia was apparent shortly after birth, and slightly >50% of the mice had evidence of liver tumors after 1 year. The livers of all RE4 mice appeared normal after 65 weeks (12 livers examined between 12 to 20 weeks, 30 livers examined between 22 to 24 weeks, and 11 livers examined between 53 to 65 weeks) and were similar in size and appearance to those of nontransgenic controls examined at 3 to 17 months of age. Histopathological examination of liver sections showed almost no abnormalities in RE4 mice (n=18) expressing low levels of APOBEC-1, although in the older mice (n=3 each group, 65 weeks), we observed some fatty changes in the RE4 mice and the nontransgenic mice (n=3 each group, 65 weeks). In contrast, the livers of I-20 mice (n=12) overexpressing APOBEC-1 showed fatty deposits and hepatic dysplasia (Figure 2), similar to the results previously described for the livers of these transgenic mice.¹⁷ Thus, the incidence of hepatic dysplasia and tumors was correlated with the APOBEC-1 expression levels.

View larger version (148K): [in this window] [in a new window]   Figure 2. Liver morphology of RE4 and I-20 mice. Representative photomicrographs of hematoxylin/eosin–stained liver sections of a 6-month-old nontransgenic mouse (A, 125x), a 6-month-old RE4 transgenic mouse (B, 125x), and a 3-month-old I-20 transgenic mouse (C, 125x) and (D, 250x). Nontransgenic and RE4 mice show no morphological changes, whereas the I-20 mice show fatty deposits and hepatic dysplasia.

To evaluate the effect of APOBEC-1 expression on liver function, we measured the alanine aminotransferase and aspartate aminotransferase levels in transgenic mouse plasma (Figure 3). I-20 mice had elevated liver enzymes, indicating liver dysfunction (Figure 3). In contrast, RE4 mice had normal liver enzyme levels. Thus, long-term, low-level APOBEC-1 expression did not cause liver dysfunction or tumors in transgenic mice.

View larger version (14K): [in this window] [in a new window]   Figure 3. Liver function in RE4 and I-20 mice. Plasma levels of liver enzymes alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were measured in 6-month-old nontransgenic (Non-tg) and RE4 mice and in 3-month-old I-20 mice. Data are mean±SD, n=3 per group.

Effect of APOBEC-1 Expression on Hyperediting of ApoB and NAT1 mRNA
Site-specific editing of apoB mRNA by APOBEC-1 requires additional unidentified auxiliary protein factors.¹¹ Some of these auxiliary proteins may determine the specificity of editing through the binding to apoB mRNA and recruiting of APOBEC-1 to the specific site of editing, C⁶⁶⁶⁶. When APOBEC-1 is overexpressed in either animals^{17 18} or tissue culture cells,²⁵ the specificity of apo-B editing is lost, resulting in the hyperediting of other cytidines in apoB mRNA. These results suggest that the relative levels of APOBEC-1 and other auxiliary factors are important for site-specific editing and that the most likely mechanism for the liver dysplasia and tumors in animals overexpressing APOBEC-1 is the hyperediting of mRNAs that control cell proliferation. We have identified one mRNA, NAT1, that is hyperedited in the liver tumors in transgenic mice but is not edited in nontransgenic mice.¹⁹ Therefore, the degree of hyperediting of apo-B and NAT1 may represent the loss of balanced stoichiometry in the editing complex and consequently, the loss of specificity. To test the effect of APOBEC-1 expression on the specificity of editing, we used RT-PCR and primer extension to assess hyperediting at several sites on apoB and NAT1 mRNAs from transgenic mouse livers. As shown in Figure 4A, C⁶⁷³⁸, C⁶⁷⁴³, and C⁶⁶⁷⁵ in apoB mRNA, which were not edited in nontransgenic control mice, were hyperedited in the livers of I-20 mice (37.4±2.6%, 10.3±0.8%, and 7.3±0.6%, respectively). In RE4 mice, however, these cytidines were either not edited (C⁶⁷³⁸) or only minimally edited (C⁶⁷⁴³, 1±0.5%; C⁶⁶⁷⁵, 0.6±0.2%). Similarly, 2 cytidines in NAT1 mRNA that were hyperedited in I-20 mice (C³⁷²⁵, 7.4±0.6%; C³⁷³⁰, 18.3±0.9%) were either not edited (C³⁷²⁵) or only minimally edited (C³⁷³⁰, 1.3±0.3%) in RE4 mice (Figure 4B). These results indicate a dose-dependent effect of APOBEC-1 expression on hyperediting of both apoB and NAT1 mRNAs.

View larger version (14K): [in this window] [in a new window]   Figure 4. Hyperediting of apoB and NAT-1 in APOBEC-1 transgenic mice. A, Hyperediting of apoB mRNA at C6738, C6743, and C6675 and B, hyperediting of NAT-1 mRNA at C3725 and C3730 in the livers of nontransgenic (Non-tg) and APOBEC-1 transgenic mice (RE4 and I-20) was determined by primer extension after RT-PCR. Data are mean±SD (n=5 per group).

Effect of APOBEC-1 Expression on ApoB-100 and LDL Levels
Although low-level APOBEC-1 expression resulted in only minimal hyperediting of apoB and NAT1 mRNAs and no liver tumors, it was sufficient to edit almost all apoB mRNA at C⁶⁶⁶⁶ and dramatically reduce plasma apo-B100 and LDL levels. Using RT-PCR and primer extension analysis, we determined the editing of apoB mRNA at C⁶⁶⁶⁶ in the livers of control nontransgenic, RE4 transgenic, and I-20 transgenic mice. An average of 69% of apoB mRNA in the nontransgenic mouse liver was edited. In I-20 transgenic mice, which had 8 times the editing activity, >95% of apoB mRNA was edited. Surprisingly, in the low-expressing RE4 mice, an average of 93% of apoB mRNA was edited (Figure 5A). Western blot analysis of apoB in mouse plasma lipoproteins isolated by ultracentrifugation (d<1.210 g/mL) was unable to detect apoB-100 in both RE4 and I-20 mice (Figure 5B). These results indicate that even low levels of APOBEC-1 expression dramatically reduce apoB-100 concentrations in transgenic mice.

View larger version (16K): [in this window] [in a new window]   Figure 5. Low-level APOBEC-1 expression in transgenic liver dramatically reduces plasma apoB-100 levels. A, Hepatic editing of apoB mRNA at C6666 in nontransgenic mice (Non-tg), RE4, and I-20 transgenic mice, as determined by primer extension after RT-PCR. (Non-tg 69.3±2.2%; RE4, 93.0±1.6; and I-20, 95.7±2.8%; n=5 per group.) B, Western blot analysis demonstrated the elimination of apoB-100 from mouse plasma lipoproteins (d<1.210 g/mL) in transgenic RE4 and I-20 mice.

To examine the effect of low-level APOBEC-1 expression on LDL levels in transgenic mice, we analyzed plasma lipoproteins in nontransgenic and RE4 mice (Figure 6). Superose 6 chromatographic column fractions representing VLDL, IDL, LDL, and HDL were pooled, concentrated, and subjected to agarose gel electrophoresis (Figure 7). Compared with nontransgenic mice, RE4 mice had similar total cholesterol levels (the Table) but lower LDL cholesterol levels (Figures 6 and 7).

View larger version (32K): [in this window] [in a new window]   Figure 6. Low-level APOBEC-1 expression in transgenic livers dramatically reduces LDL levels. Representative Superose 6 profiles of (A) nontransgenic mice, (B) APOBEC-1 transgenic (RE4) mice, (C) human apoB transgenic mice, and (D) APOBEC-1 and human apoB double-transgenic mice. Two separate experiments were performed on age- and sex-matched mice. For each experiment, plasma was pooled from 3 animals. Data are shown for 1 Superose 6 fractionation. TC indicates total cholesterol; TG, triglycerides.

View larger version (55K): [in this window] [in a new window]   Figure 7. Agarose gel of Superose 6–separated lipoproteins from (A) nontransgenic and (B) APOBEC-1 (RE4), (C) human apoB, and (D) APOBEC-1/human apoB transgenic mice. V indicates VLDL; I, IDL; L, LDL and LDL/HDL1; and H, HDL.

Mice have intrinsically low plasma concentrations of apoB-100 and LDL. To assess further the efficacy of low APOBEC-1 expression in reducing plasma apoB-100 and LDL levels in an animal model with a lipoprotein profile resembling that of humans, we bred RE4 mice with human apoB transgenic mice, which have elevated levels of total cholesterol and LDL cholesterol (the Table and Figure 6C). Low-level APOBEC-1 expression significantly decreased plasma total cholesterol, triglycerides, and LDL cholesterol concentrations in double-transgenic mice compared with human apoB–only transgenic mice (the Table and Figures 6C, 6D, 7C, and 7D) and reduced plasma apoB-100 to undetectable levels (Figure 8A). In human apoB transgenic mice, mouse and human apoB mRNAs were edited at comparable levels (75% and 77%, respectively) (Figure 8B). In 2 APOBEC-1/human apoB double-transgenic mice, editing at C⁶⁶⁶⁶ increased to 95% and 92% for mouse and human apoB mRNAs, respectively. These results indicate that even low-level hepatic expression of APOBEC-1 efficiently reduces elevated apoB-100 and LDL levels, making the mice resistant to hypercholesterolemia induced by overproduction of apoB.

View larger version (39K): [in this window] [in a new window]   Figure 8. Editing of human apoB mRNA and elimination of human apoB-100 from APOBEC-1/human apoB double-transgenic (Tg) mice. A, Western blot analysis of human apoB in total plasma of human apoB transgenic mice and APOBEC-1 (RE4)/human apoB double-transgenic mice. B, Editing of apoB mRNA at C6666 in the livers of 2 human apoB transgenic mice (ApoB Tg) and 2 APOBEC-1 (RE4)/human apoB double-transgenic mice (APOBEC-1 x ApoB Tg) was determined by primer extension after RT-PCR.

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
In the current study we used transgenic mice expressing low and high levels of APOBEC-1 in the liver to investigate the efficacy and safety of hepatic APOBEC-1 expression. The results clearly showed that the liver dysplasia and hepatocellular carcinoma depend on the level of APOBEC-1 expression. Low-expressing RE4 mice had histologically normal livers and no evidence of altered liver function for up to 65 weeks. High-expressing I-28 mice developed tumors by 10 to 20 weeks of age, and intermediate-expressing I-20 mice had evidence of liver dysplasia after birth but on average did not develop evidence of tumors until almost 1 year of age (Figure 1B). Therefore, for APOBEC-1 to be useful as a therapeutic agent, it is critical to maintain hepatic APOBEC-1 expression at a level that permits efficient normal editing but does not alter the function and proliferation of hepatocytes. As demonstrated by this study, such a goal is achievable, at least in mice.

It has been shown that AU-rich mRNAs bind APOBEC-1, regardless of their intrinsic ability to undergo editing.²⁶ However, the remarkable specificity of apoB mRNA editing under physiological conditions is thought to be determined by auxiliary proteins in the editing complex. We and others have proposed that hyperediting represents a loss of specificity of the apo B mRNA–editing machinery caused by a shift of stoichiometry between APOBEC-1 and the auxiliary proteins.^{18 25} The correlation between expression levels of APOBEC-1 and the degree of hyperediting of both NAT1 and apoB mRNAs observed here supports this hypothesis.

The low-expressing transgenic mice were generated by the use of a rat transgene, whereas the transgenic mice expressing medium and high levels of APOBEC-1 were generated by the use of a rabbit gene. Because the rat gene is more similar to the mouse gene, the difference in toxicity may have been in part due to species difference. However, it is unlikely that this is the main cause of the difference in toxicity for the following two reasons. First, overexpression of rabbit APOBEC-1 in transgenic rabbit caused liver dysplasia¹⁷ ; second, overexpression of rat APOBEC-1 in rat hepatoma cell line McArdle cells resulted in hyperediting.²⁵ Thus, homologous biological systems displayed the same phenotype when APOBEC-1 was overexpressed.

The current study demonstrates that a low level of hepatic apoB mRNA editing activity virtually eliminates LDL from the plasma. This finding supports the hypothesis that increased hepatic apoB mRNA editing can lower apoB-100 and LDL cholesterol levels. Indeed, low-level expression of APOBEC-1 even reduced plasma apoB-100 and LDL levels in the human apoB transgenic mice, which have greatly elevated plasma apoB-100 and LDL cholesterol. Similarly, adenovirus-mediated hepatic expression of APOBEC-1 dramatically decreased plasma apoB-100 and LDL levels in Watanabe heritable hyperlipidemic rabbits and LDL receptor–null mice, which also have greatly elevated plasma LDL cholesterol levels.^{14 16} These results indicate that increasing hepatic APOBEC-1 activity by either gene therapy or pharmaceutical treatment could reduce or prevent hypercholesterolemia in mice and rabbits and probably in humans as well.

Despite these encouraging results for the potential therapeutic use of hepatic APOBEC-1 expression, it is important to note that more rigorous control of the specificity of APOBEC-1 editing may be required in species that do not normally express APOBEC-1 in the liver, such as humans and rabbits. Human and rabbit livers possess auxiliary proteins that supplement APOBEC-1 and permit editing,^{21 27} but deficiencies in their amount or composition may result in hyperediting at a lower level of APOBEC-1 expression than in mouse liver.

The RE4 mice expressing low levels of APOBEC-1 should prove useful for studying the roles of apoB-containing lipoproteins in lipid metabolism and atherosclerosis. Although our findings demonstrate that increased hepatic editing effectively reduces apoB-100 and prevents hypercholesterolemia, it is not known whether lipoproteins containing apoB-48 are as atherogenic as those containing apoB-100. Intestinally derived apoB-48-containing lipoproteins are cleared rapidly after their appearance in the circulation.⁸ The low LDL cholesterol levels and relative resistance to diet-induced hypercholesterolemia in rodents further indicate that apoB-48–containing lipoproteins do not contribute to a high plasma cholesterol level. Studies of hypobetacholesterolemic patients indicate that expression of other truncated forms of apoB can also lower LDL levels.²⁸ In heterozygous subjects with hypobetacholesterolemia, who are resistant to atherosclerosis, LDL cholesterol levels are one-third to one-half of normal. These observations support the appealing hypothesis that inducing hepatic apoB-48 expression in humans might be used to lower LDL cholesterol levels and reduce atherogenic events. However, other studies suggest that apoB-48–containing lipoproteins may be atherogenic. Type III hypercholesterolemia is characterized clinically by the elevation of plasma cholesterol and triglyceride levels, the accumulation of cholesterol-rich ß-VLDL particles (apoB-48–containing ß-VLDL and apoB-100–containing VLDL remnants) in the plasma, and the formation of lipid deposits in the skin and artery wall.^{6 29} In vitro experiments using mouse peritoneal macrophages have shown that the chylomicron remnants are effective in causing cholesteryl ester accumulation in these cells.³⁰ Recently, transgenic apoB-100– and apoB-48–only mice on the apoE-null background were used to study the atherogenicity of apoB proteins.³¹ This study indicated that in the absence of apoE, apoB-48 is at least as atherogenic as apoB-100. Thus, susceptibility to atherosclerosis in these animal models depends on total cholesterol levels. However, because apoE is an important component of apoB-48–containing lipoproteins, it is not known whether the apoE-depleted apoB-48 particles have the same lipid composition or whether the apoB-48 has the same conformation on these particles as on those containing apoE. Therefore, the atherogenic properties of apoB-48 in the presence of apoE remain to be elucidated.

The RE4 mice characterized in this study expressed low levels of APOBEC-1 in the liver and showed no signs of pathological effects from the transgene. When these mice were crossed with human apoB transgenic mice, no apoB-100 was detectable in the plasma of their double-transgenic offspring. Transgenic mice overexpressing human apoB develop severe atherosclerosis when they are fed a high-fat, high-cholesterol diet.³² Thus, the APOBEC-1/human apoB double-transgenic mouse can be used to determine whether conversion of all apoB-100 in the apoB-overexpressing mice to apoB-48 will prevent or promote the development of atherosclerosis. Furthermore, genetic crossing of this mouse model with other transgenic or knockout mouse models harboring other changes in lipoprotein metabolism and the vasculature will allow us to generate new transgenic mouse lines with an apoB-48–only phenotype on these altered genetic backgrounds. These mouse models can be used to dissect the interrelationship between apoB and the other risk or beneficial factors of atherosclerosis.

	Selected Abbreviations and Acronyms

 

apo = apolipoprotein APOBEC-1 = apoB mRNA–editing catalytic polypeptide 1 Lp(a) = lipoprotein(a) NAT1 = novel APOBEC-1 target 1 PCR = polymerase chain reaction RT = reverse transcribed

	Acknowledgments

 
This research was funded in part by the National Heart, Lung, and Blood Institute program project grant HL47660 (to T.L.I.), a Molecular and Cellular Basis of Cardiovascular Disease training grant (to X.Q.), and a postdoctoral fellowship from the American Heart Association, Western States Affiliate (to X.Q.). We thank Drs Stanley Rall, Bob Pitas, and Yadong Huang for their critical comments on this manuscript; Amy Corder, John Carroll, and Stephen Gonzales for graphics; Susannah Patarroyo and September Plumlee for manuscript preparation; and Gary Howard and Stephen Ordway for editorial support.

Received December 2, 1997; accepted February 6, 1998.

	References

Top Abstract Introduction Methods Results Discussion References

 

1. Chan L. Apolipoprotein B, the major protein component of triglyceride-rich and low density lipoproteins. J Biol Chem. 1992;267:25621–25624.[Free Full Text]
2. Innerarity TL. Plasma lipoproteins. In: Dulbecco R, eds. Encyclopedia of Human Biology. Vol 6. San Diego, Calif: Academic Press; 1991:23–35.
3. Innerarity TL, Borén J, Yamanaka S, Olofsson S-O. Biosynthesis of apolipoprotein B48-containing lipoproteins: regulation by novel post-transcriptional mechanisms. J Biol Chem. 1996;271:2353–2356.[Free Full Text]
4. Powell LM, Wallis SC, Pease RJ, Edwards YH, Knott TJ, Scott J. A novel form of tissue-specific RNA processing produces apolipoprotein-B48 in intestine. Cell. 1987;50:831–840.[Medline] [Order article via Infotrieve]
5. Johnson DF, Poksay KS, Innerarity TL. The mechanism for apo-B mRNA editing is deamination. Biochem Biophys Res Commun. 1993;195:1204–1210.[Medline] [Order article via Infotrieve]
6. Mahley RW, Rall SC Jr. Type III hyperlipoproteinemia (dysbetalipoproteinemia): the role of apolipoprotein E in normal and abnormal lipoprotein metabolism. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The Metabolic and Molecular Bases of Inherited Disease. 7th ed. New York, NY: McGraw-Hill; 1995:1953–1980.
7. Pitas RE, Innerarity TL, Mahley RW. Cell surface receptor binding of phospholipid · protein complexes containing different ratios of receptor-active and inactive E apoprotein. J Biol Chem. 1980;255:5454–5460.[Free Full Text]
8. Innerarity TL, Mahley RW. Enhanced binding by cultured human fibroblasts of apo-E-containing lipoproteins as compared with low density lipoproteins. Biochemistry. 1978;17:1440–1447.[Medline] [Order article via Infotrieve]
9. Elovson J, Huang YO, Baker N, Kannan R. Apolipoprotein B is structurally and metabolically heterogeneous in the rat. Proc Natl Acad Sci U S A. 1981;78:157–161.[Abstract/Free Full Text]
10. Greeve J, Altkemper I, Dieterich J-H, Greten H, Windler E. Apolipoprotein B mRNA editing in 12 different mammalian species: hepatic expression is reflected in low concentrations of apoB-containing plasma lipoproteins. J Lipid Res. 1993;34:1367–1383.[Abstract]
11. Teng B, Burant CF, Davidson NO. Molecular cloning of an apolipoprotein B messenger RNA editing protein. Science. 1993;260:1816–1819.[Abstract/Free Full Text]
12. Young SG. Recent progress in understanding apolipoprotein B. Circulation. 1990;82:1574–1594.[Abstract/Free Full Text]
13. Teng B, Blumenthal S, Forte T, Navaratnam N, Scott J, Gotto AM Jr, Chan L. Adenovirus-mediated gene transfer of rat apolipoprotein B mRNA-editing protein in mice virtually eliminates apolipoprotein B-100 and normal low density lipoprotein production. J Biol Chem. 1994;269:29395–29404.[Abstract/Free Full Text]
14. Kozarsky KF, Bonen DK, Giannoni F, Funahashi T, Wilson JM, Davidson NO. Hepatic expression of the catalytic subunit of the apolipoprotein B mRNA editing enzyme (apobec- 1) ameliorates hypercholesterolemia in LDL receptor-deficient rabbits. Hum Gene Ther.. 1996;7:943–957.[Medline] [Order article via Infotrieve]
15. Hughes SD, Rouy D, Navaratnam N, Scott J, Rubin EM. Gene transfer of cytidine deaminase apoBEC-1 lowers lipoprotein(a) in transgenic mice and induces apolipoprotein B editing in rabbits. Hum Gene Ther. 1996;7:39–49.[Medline] [Order article via Infotrieve]
16. Teng B, Ishida B, Forte TM, Blumenthal S, Song L-Z, Gotto AM Jr, Chan L. Effective lowering of plasma, LDL, and esterified cholesterol in LDL receptor–knockout mice by adenovirus-mediated gene delivery of apoB mRNA editing enzyme (Apobec1). Arterioscler Thromb Vasc Biol. 1997;17:889–897.[Abstract/Free Full Text]
17. Yamanaka S, Balestra ME, Ferrell LD, Fan J, Arnold KS, Taylor S, Taylor JM, Innerarity TL. Apolipoprotein B mRNA-editing protein induces hepatocellular carcinoma and dysplasia in transgenic animals. Proc Natl Acad Sci U S A. 1995;92:8483–8487.[Abstract/Free Full Text]
18. Yamanaka S, Poksay KS, Driscoll DM, Innerarity TL. Hyperediting of multiple cytidines of apolipoprotein B mRNA by APOBEC-1 requires auxiliary protein(s) but not a mooring sequence motif. J Biol Chem. 1996;271:11506–11510.[Abstract/Free Full Text]
19. Yamanaka S, Poksay KS, Arnold KS, Innerarity TL. A novel translational repressor mRNA is edited extensively in livers containing tumors caused by the transgene expression of the apoB mRNA-editing enzyme. Gene Dev. 1997;11:321–333.[Abstract/Free Full Text]
20. Qian X, Balestra ME, Innerarity TL. Two distinct TATA-less promoters direct tissue-specific expression of the rat apo-B editing catalytic polypeptide 1 gene. J Biol Chem. 1997;272:18060–18070.[Abstract/Free Full Text]
21. Yamanaka S, Poksay KS, Balestra ME, Zeng G-Q, Innerarity TL. Cloning and mutagenesis of the rabbit apoB mRNA editing protein: a zinc motif is essential for catalytic activity, and noncatalytic auxiliary factor(s) of the editing complex are widely distributed. J Biol Chem. 1994;269:21725–21734.[Abstract/Free Full Text]
22. Pease RJ, Milne RW, Jessup WK, Law A, Provost P, Fruchart J-C, Dean RT, Marcel YL, Scott J. Use of bacterial expression cloning to localize the epitopes for a series of monoclonal antibodies against apolipoprotein B100. J Biol Chem. 1990;265:553–568.[Abstract/Free Full Text]
23. Horie Y, Fazio S, Westerlund JR, Weisgraber KH, Rall SC Jr. The functional characteristics of a human apolipoprotein E variant (cysteine at residue 142) may explain its association with dominant expression of type III hyperlipoproteinemia. J Biol Chem. 1992;267:1962–1968.[Abstract/Free Full Text]
24. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem. 1951;193:265–275.[Free Full Text]
25. Sowden M, Hamm JK, Smith HC. Overexpression of APOBEC-1 results in mooring sequence-dependent promiscuous RNA editing. J Biol Chem. 1996;271:3011–3017.[Abstract/Free Full Text]
26. Anant S, MacGinnitie AJ, Davidson NO. apobec-1, the catalytic subunit of the mammalian apolipoprotein B mRNA editing enzyme, is a novel RNA-binding protein. J Biol Chem. 1995;270:14762–14767.[Abstract/Free Full Text]
27. Giannoni F, Bonen DK, Funahashi T, Hadjiagapiou C, Burant CF, Davidson NO. Complementation of apolipoprotein B mRNA editing by human liver accompanied by secretion of apolipoprotein B48. J Biol Chem. 1994;269:5932–5936.[Abstract/Free Full Text]
28. Linton MF, Farese RV Jr, Young SG. Familial hypobetalipoproteinemia. J Lipid Res. 1993;34:521–541.[Medline] [Order article via Infotrieve]
29. Kane JP, Chen GC, Hamilton RL, Hardman DA, Malloy MJ, Havel RJ. Remnants of lipoproteins of intestinal and hepatic origin in familial dysbetalipoproteinemia. Arteriosclerosis. 1983;3:47–56.[Abstract/Free Full Text]
30. Fainaru M, Mahley RW, Hamilton RL, Innerarity TL. Structural and metabolic heterogeneity of ß-very low density lipoproteins from cholesterol-fed dogs and from humans with type III hyperlipoproteinemia. J Lipid Res. 1982;23:702–714.[Abstract]
31. Véniant MM, Pierotti V, Newland D, Cham CM, Sanan DA, Walzem RL, Young SG. Susceptibility to atherosclerosis in mice expressing exclusively apolipoprotein B48 or apolipoprotein B100. J Clin Invest. 1997;100:180–188.[Medline] [Order article via Infotrieve]
32. Purcell-Huynh DA, Farese RV Jr, Johnson DF, Flynn LM, Pierotti V, Newland DL, Linton MF, Sanan DA, Young SG. Transgenic mice expressing high levels of human apolipoprotein B develop severe atherosclerotic lesions in response to a high-fat diet. J Clin Invest. 1995;95:2246–2257.