Liver-specific Inactivation of the Abetalipoproteinemia Gene Completely Abrogates Very Low Density Lipoprotein/Low Density Lipoprotein Production in a Viable Conditional Knockout Mouse

Conventional knockout of the microsomal triglyceride transfer protein large subunit (ℓMTP) gene is embryonic lethal in the homozygous state in mice. We have produced a conditional ℓMTP knockout mouse by inserting loxP sequences flanking exons 5 and 6 by gene targeting. Homozygous floxed mice were born live with normal plasma lipids. Intravenous injection of an adenovirus harboring Cre recombinase (AdCre1) produced deletion of exons 5 and 6 and disappearance of ℓMTP mRNA and immunoreactive protein in a liver-specific manner. There was also disappearance of plasma apolipoprotein (apo) B-100 and marked reduction in apoB-48 levels. Wild-type mice showed no response, and heterozygous mice, an intermediate response, to AdCre1. Wild-type mice doubled their plasma cholesterol level following a high cholesterol diet. This hypercholesterolemia was abolished in AdCre1-treated ℓMTP−/− mice, the result of a complete absence of very low/intermediate/low density lipoproteins and a slight reduction in high density lipoprotein. Heterozygous mice showed an intermediate lipoprotein phenotype. The rate of accumulation of plasma triglyceride following Triton WR1339 treatment in ℓMTP−/− mice was <10% that in wild-type animals, indicating a failure of triglyceride-rich lipoprotein production. Pulse-chase experiments using hepatocytes isolated from wild-type and ℓMTP−/− mice revealed a failure of apoB secretion in ℓMTP−/−animals. Therefore, the liver-specific inactivation of the ℓMTP gene completely abrogates apoB-100 and very low/intermediate/low density lipoprotein production. These conditional knockout mice are a usefulin vivo model for studying the role of MTP in apoB biosynthesis and the biogenesis of apoB-containing lipoproteins. Conventional knockout of the microsomal triglyceride transfer protein large subunit (ℓMTP) gene is embryonic lethal in the homozygous state in mice. We have produced a conditional ℓMTP knockout mouse by inserting loxP sequences flanking exons 5 and 6 by gene targeting. Homozygous floxed mice were born live with normal plasma lipids. Intravenous injection of an adenovirus harboring Cre recombinase (AdCre1) produced deletion of exons 5 and 6 and disappearance of ℓMTP mRNA and immunoreactive protein in a liver-specific manner. There was also disappearance of plasma apolipoprotein (apo) B-100 and marked reduction in apoB-48 levels. Wild-type mice showed no response, and heterozygous mice, an intermediate response, to AdCre1. Wild-type mice doubled their plasma cholesterol level following a high cholesterol diet. This hypercholesterolemia was abolished in AdCre1-treated ℓMTP−/− mice, the result of a complete absence of very low/intermediate/low density lipoproteins and a slight reduction in high density lipoprotein. Heterozygous mice showed an intermediate lipoprotein phenotype. The rate of accumulation of plasma triglyceride following Triton WR1339 treatment in ℓMTP−/− mice was <10% that in wild-type animals, indicating a failure of triglyceride-rich lipoprotein production. Pulse-chase experiments using hepatocytes isolated from wild-type and ℓMTP−/− mice revealed a failure of apoB secretion in ℓMTP−/−animals. Therefore, the liver-specific inactivation of the ℓMTP gene completely abrogates apoB-100 and very low/intermediate/low density lipoprotein production. These conditional knockout mice are a usefulin vivo model for studying the role of MTP in apoB biosynthesis and the biogenesis of apoB-containing lipoproteins. Abetalipoproteinemia is an autosomal recessive disorder characterized by the almost complete absence of circulating apolipoprotein (apo) 1The abbreviations used are: apo, apolipoprotein; MTP, microsomal triglyceride transfer protein; ℓMTP, MTP large subunit; VLDL, very low density lipoproteins; IDL, intermediate density lipoproteins; LDL, low density lipoproteins; HDL, high density lipoproteins; kb, kilobase(s); PAGE, polyacrylamide gel electrophoresis; FPLC, fast protein liquid chromatography; FIAU, 1-(2′-deoxy-2′-fluoro-β-d-arabinofuranosyl)-5-iodouracil. B-containing lipoproteins (1Kane J.P. Havel R.J. Scriver C.R. Beaudet A.L. Sly W.S. Valle D. The Metabolic and Molecular Bases of Inherited Disease. McGraw-Hill, Inc., New York1995: 1853-1885Google Scholar). In addition to the lipoprotein abnormalities, patients with abetalipoproteinemia also suffer from severe anemia with acanthocytosis, fat malabsorption, and progressive neurodegenerative syndromes. Abetalipoproteinemia is caused by mutations in the gene for the large subunit of the microsomal triglyceride transfer protein (MTP) (2Wetterau J.R. Aggerbeck L.P. Bouma M.-E. Eisenberg C. Munck A. Hermier M. Schmitz J. Gay G. Rader D.J. Gregg R.E. Science. 1992; 258: 999-1001Crossref PubMed Scopus (644) Google Scholar, 3Sharp D. Blinderman L. Combs K.A. Kienzie B. Ricci B. Wager-Smith K. Gil C.M. Turck C.W. Bouma M.-E. Rader D.J. Aggerbeck L.P. Gregg R.E. Gordon D.A. Wetterau J.R. Nature. 1993; 365: 65-69Crossref PubMed Scopus (408) Google Scholar, 4Shoulders C.C. Brett D.J. Bayliss J.D. Narcisi T.M.E. Jarmuz A. Grantham T.T. Leoni P.R.D. Bhattacharya S. Pease R.J. Cullen P.M. Levi S. Byfield P.G.H. Purkiss P. Scott J. Hum. Mol. Genet. 1993; 2: 2109-2116Crossref PubMed Scopus (228) Google Scholar, 5Narcisi T.M.E. Shoulders C.C. Chester S.A. Read J. Brett D.J. Harrison G.B. Grantham T.T. Fox M.F. Povey S. de Bruin T.W.A. Erkelens D.W. Muller D.P.R. Lloyd J.K. Scott J. Am. J. Hum. Genet. 1995; 57: 1298-1310PubMed Google Scholar). MTP is an integral protein in the endoplasmic reticulum. It was originally described as a protein that accelerates the transport of lipids (triglycerides, cholesteryl ester, and phosphatidylcholine) between synthetic membranes (6Wetterau J.R. Zilversmitt D.B. J. Biol. Chem. 1984; 259: 10863-10866Abstract Full Text PDF PubMed Google Scholar). MTP is a heterodimer composed of a 97-kDa subunit (designated ℓMTP) and a 58-kDa subunit, which turned out to be protein-disulfide isomerase (7Wetterau J.R. Combs K.A. Spinner S.N. Joiner B.J. J. Biol. Chem. 1990; 265: 9800-9807Abstract Full Text PDF PubMed Google Scholar). ApoB-100 and apoB-48 secretion from the cell requires the presence of functional MTP. In its absence, all the newly synthesized intracellular apoB is degraded, and little of it is secreted (8Leiper J.M. Bayliss J.D. Pease R.J. Brett D.J. Scott J. Shoulders C.C. J. Biol. Chem. 1994; 269: 21951-21954Abstract Full Text PDF PubMed Google Scholar, 9Gordon D.A. Jamil H. Sharp D. Mullaney D. Yao Z. Greeg R.E. Wetterau J. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 7628-7632Crossref PubMed Scopus (189) Google Scholar). It is believed that MTP facilitates the stabilization of newly synthesized apoB-100 (10Rustaeus S. Stillemark P. Lindberg K. Gordon D. Olofsson S.-O. J. Biol. Chem. 1998; 273: 5196-5203Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar) and apoB-48 (11Gordon D.A. Jamil H. Gregg R.E. Olofsson S.-O. Borén J. J. Biol. Chem. 1996; 271: 33047-33053Abstract Full Text Full Text PDF PubMed Scopus (143) Google Scholar) in a narrow “window” shortly after completion of apoB translation but before the addition of the major amount of lipids to the lipoprotein particle (reviewed in Refs. 12Gordon D.A. Curr. Opin. Lipidol. 1997; 8: 131-137Crossref PubMed Scopus (81) Google Scholar, 13Yao Z. Tran K. McLeod R.S. J. Lipid Res. 1997; 38: 1937-1953Abstract Full Text PDF PubMed Google Scholar, 14Wetterau J.R. Lin M.C.M. Jamil H. Biochim. Biophys. Acta. 1997; 1345: 136-150Crossref PubMed Scopus (286) Google Scholar). The pathophysiological basis of abetalipoproteinemia is postulated to be an almost complete failure of apoB secretion from the liver and small intestine because of the absence of functional MTP in affected patients. An animal model would be valuable in studying the molecular basis of the lipoprotein abnormalities and elucidating the pathophysiological basis of the other non-lipoprotein complications found in abetalipoproteinemia. Unfortunately, the genetic inactivation of the ℓMTP locus in mice is embryonic lethal in the homozygous state ((15) and see below). In this communication, we report the production and characterization of a viable mouse model of abetalipoproteinemia using a conditional knockout strategy. A mouse 129 strain λ genomic library was purchased (Stratagene) and screened with mouse ℓMTP cDNA (16Nakamuta M. Chang B.H.J. Hoogeveen R. Li W.-H. Chan L. Genomics. 1996; 33: 313-316Crossref PubMed Scopus (21) Google Scholar). Two overlapping clones encompassing exons 3 to 8 were used to construct two types of targeting vectors. For straight replacement-type targeting construct, a neo cassette was inserted into exon 7 of the ℓMTP gene between a SmaI and aXhoI site (data available from L. C. upon request). The conditional targeting construct was designed by inserting a neo-loxP cassette in the XbaI site of intron 6 and a loxP fragment in the BamHI site of intron 4 of the ℓMTP (Fig.1 A). A thymidine kinase cassette was ligated to the 5′ end of the construct. An R1 ES cell line was obtained from Dr. Andras Nagy at the University of Toronto. The cells were expanded to passage 14 and used to generate knockout mice as described previously (17Nakamuta M. Chang B.H.J. Zsigmond E. Kobayashi K. Lei H. Ishida B.Y. Oka K. Li E. Chan L. J. Biol. Chem. 1996; 271: 25981-25988Abstract Full Text Full Text PDF PubMed Scopus (82) Google Scholar). Three positive ES cell clones were injected into blastocysts of C57BL/6J, and chimeric mice were obtained. They were mated with C57BL/6J mice, and germ line transmission was confirmed by Southern blot analysis. Most experiments were conducted on siblings of F2 or F3 mice. The mice were weaned at 21 days and fed either a chow diet (Teklad 7001) or a high fat cholesterol diet (ICN 960393) containing 1.23% cholesterol and 17.84% fat. Total RNA from the liver and small intestine were isolated using TriZOL (Life Technologies, Inc.). A polymerase chain reaction product containing the first 350 base pairs of the mouse ℓMTP cDNA were cloned into pBluescript KS vector and used as a probe for RNase protection assay. The authenticity of the clone was verified by sequencing, and the antisense strand RNA was transcribed by using MAXIscript (Ambion). The assay was done using the RPAII kit (Ambion) with 5 μg of total RNA following the instructions of the vendor's manual. A β-actin probe was used as an internal control in the assay. Liver and small intestine were removed, and total proteins were extracted by a Wheaton No. 6 hand-held homogenizer in buffer B (10 mm Hepes pH 7.4, 2.5 mm sodium phosphate monobasic, 250 mm sucrose, 5 mm EDTA, and 0.5 mm phenylmethylsulfonyl fluoride). Microsomal fractions were isolated from extracts of liver and small intestine by ultracentrifugation at 100,000 ×g for 1 h. The pellet was resuspended in buffer B, and the protein concentration was determined by a DC Protein Assay Kit from Bio-Rad. Ten μg of microsomal protein was loaded onto a 4–15% gradient polyacrylamide gel, and a rabbit anti-bovine ℓMTP antiserum (a gift from Dr. David Gordon of Bristol-Myers Squibb Pharmaceutical Research Institute) was used for Western blot analysis. One hundred μg of microsomal protein was used to measure the MTP-mediated 14Ctriglyceride transfer activity as described in Jamilet al. (18Jamil H. Dickson Jr., J.K. Chu C.-H. Lago M.W. Rinehart J.K. Biller S.A. Gregg R.E. Wetterau J.R. J. Biol. Chem. 1995; 270: 6549-6554Abstract Full Text Full Text PDF PubMed Scopus (143) Google Scholar). Blood was collected after a 4–5-h fast, and total plasma cholesterol and triglyceride concentrations were measured by enzymatic kits (Sigma Diagnostics). Two hundred μl of pooled plasma from 3–4 animals was loaded on a FPLC system with 2 Superose 6 columns connected in series (Pharmacia FPLC System, Amersham Pharmacia Biotech). 0.5-ml fractions were collected using an elution buffer (1 mm EDTA, 154 mmNaCl, and 0.02% NaN3, pH 8.2) (19Cole T. Kitchens R.T. Daugherty A. Schonfeld G. FLPC BioCommunique. 1988; 4: 4-6Google Scholar). Lipid contents in individual fractions were determined with enzymatic assay kits (Sigma Diagnostics). Very low density (VLDL), intermediate density (IDL), low density (LDL), and high density (HDL) lipoproteins are well separated by this technique (19Cole T. Kitchens R.T. Daugherty A. Schonfeld G. FLPC BioCommunique. 1988; 4: 4-6Google Scholar). A replication-defective adenovirus containing recombinant Cre recombinase, AdCre1, was a gift from Dr. Frank Graham (McMaster University, Hamilton, Ontario, Canada) (20Anton M. Graham F.L. J. Virol. 1995; 69: 4600-4606Crossref PubMed Google Scholar). It was amplified in 293 cells and purified as described previously (21Kobayashi K. Oka K. Forte T. Ishida B.Y. Teng B.-B. Ishimura-Oka K. Nakamuta M. Chan L. J. Biol. Chem. 1996; 271: 6851-6860Google Scholar). Eight-week-old male mice were injected with 3 × 109 plaque-forming units of AdCre1 through a jugular vein. The AdCre1-treated mice were fed either a normal chow or a high cholesterol diet before and after injection. At day 10–21 after adenovirus administration, the mice were sacrificed and studied. Hepatocytes were isolated from AdCre1-treated mice by White's method (22White A. Clin. Genet. 1997; 52: 326-337Crossref PubMed Scopus (11) Google Scholar) except that the perfusion was done on an anesthetized animal through the portal vein instead of on excised liver. A pulse-chase experiment using 35Smethionine was done to determine apoB degradation and secretion in these cells (23Yeung S.J. Chen S.-H. Chan L. Biochemistry. 1996; 35: 13843-13848Crossref PubMed Scopus (170) Google Scholar). Triglyceride secretion in vivo was quantified by the intravenous administration of Triton WR1339 (24Li X. Catalina F. Grundy S.M. Patel S. J. Lipid Res. 1996; 37: 210-220Abstract Full Text PDF PubMed Google Scholar). Plasma triglycerides were measured at 1, 2, 3, and 4 h after treatment; the triglyceride accumulation remained linear during this time. As reported by Raabe et al. (15Raabe M. Flynn L.M. Zlot C.H. Wong J.S. Véniant M.M. Hamilton R.L. Young S.G. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 8686-8691Crossref PubMed Scopus (223) Google Scholar), we produced ℓMTP knockout mice by gene targeting in ES cells and found that inactivation of the ℓMTP locus in mice is embryonic lethal in the homozygous state. 2B. H.-J. Chang, M. Nakamuta, and L. Chan, unpublished observations. We have therefore produced a conditional knockout construct shown in Fig.1 A. It would insert two loxP sequences encompassing exons 5 and 6. Deletion of these exons would be predicted to inactivate the ℓMTP protein because it causes a shift in the translation frame of the mRNA if the remaining exons are correctly into a was verified by of genomic ES cell with BamHI and Southern blot using a probe the targeting vector (Fig.1 A). The presence of a fragment instead of the wild-type fragment of the vector by 14 of clones that were and a with 3 of 14 ES cell clones were injected into C57BL/6J 8 chimeric mice with indicating of ES were with C57BL/6J and germ line transmission was in of the by Southern blot of (Fig.1 Heterozygous and homozygous floxed ℓMTP−/− mice were obtained by These animals were and produced The and of and ℓMTP−/− mice were The plasma cholesterol and triglyceride are in the types of animals were on a chow diet Therefore, of the loxP sequences and the neo cassette to the of ℓMTP Two after these animals were on a high cholesterol was an of the plasma cholesterol but little in plasma triglyceride the were the of animals lipids in wild-type and floxed MTP knockout AdCre1 administration AdCre1 administration cholesterol diet for 2 after AdCre1 administration cholesterol between and between and of mice before AdCre1 administration were 6 from and and 4 from after AdCre1 treatment were 4 for between and between and in a The of mice before AdCre1 administration were 6 from and and 4 from after AdCre1 treatment were 4 for We the of the of Cre recombinase on the ℓMTP gene and its in wild-type and knockout of Cre recombinase was by the intravenous administration of a purified recombinant adenovirus the recombinase (AdCre1) (20Anton M. Graham F.L. J. Virol. 1995; 69: 4600-4606Crossref PubMed Google Scholar). We the liver and small intestine and extracted and from these after AdCre1 Southern blot we found that AdCre1 treatment produced a deletion of the floxed exons in the ℓMTP gene in the liver. using the probe shown produced a in the floxed ℓMTP gene deletion of the floxed of the it produced a with the of the between the loxP which an site in exon we the blot from extracted from the small it was that the floxed ℓMTP locus remained The liver is because are by liver cells L. Curr. Opin. Lipidol. 1995; PubMed Scopus Google Scholar). For heterozygous ℓMTP knockout mice, deletion of the floxed ℓMTP was also in the liver but the small intestine (data We liver and small intestine ℓMTP mRNA following AdCre1 We used RNase protection assay with an ℓMTP cDNA fragment and a β-actin cDNA fragment as internal As shown in 3 the ℓMTP mRNA is in the liver of ℓMTP−/− mice, the ℓMTP mRNA level in mouse liver was In the concentration of ℓMTP mRNA in the small intestine of mice of all and in the liver of wild-type mice was by AdCre1 Therefore, was between ℓMTP mRNA and the presence of an ℓMTP and deletion of exons 5 and 6 to the absence of ℓMTP the antisense probe used to a of the ℓMTP gene 5′ to the these that in the AdCre1-treated mouse if the ℓMTP gene were transcribed at the RNA was and its concentration low that it was by RNase these at the and mRNA are at the protein we isolated from the liver and small intestine and using an ℓMTP As shown in 3 B, a ℓMTP immunoreactive was in wild-type mouse liver and small in the AdCre1 treatment was a reduction in of the immunoreactive in the but small of the heterozygous floxed mice. In the homozygous floxed ℓMTP−/− mice, the Cre gene transfer to the liver completely the ℓMTP to be a in the of the ℓMTP in the small intestine following AdCre1 ℓMTP a days in cells M.C.M. Gordon D. Wetterau J.R. J. Lipid Res. 1995; Full Text PDF PubMed Google These that in mice, days of of ℓMTP gene is no immunoreactive ℓMTP in the liver We determined the MTP activity in the isolated from AdCre1-treated floxed heterozygous and floxed homozygous ℓMTP−/− The shown in 3 that the MTP triglyceride transfer activity of isolated from the liver of heterozygous floxed mice is to that of the wild-type and the activity of the homozygous floxed ℓMTP−/− mice is almost to level In the MTP triglyceride transfer activity in all of animals is very and for and homozygous knockout Therefore, in the liver and small intestine is between ℓMTP protein and MTP functional which that ℓMTP, and protein-disulfide is these The plasma of the types of mice before and after AdCre1 treatment are shown in adenovirus administration, the mice were on a high cholesterol and their plasma cholesterol and triglyceride were in the of AdCre1 was a reduction in the plasma cholesterol concentration in the homozygous ℓMTP−/− mice, which was that in wild-type The cholesterol level in the heterozygous animals was intermediate between of wild-type and homozygous knockout The plasma triglyceride level was between the of animals, it to be in the homozygous mice. functional ℓMTP is for apoB biogenesis D.A. Curr. Opin. Lipidol. 1997; 8: 131-137Crossref PubMed Scopus (81) Google Scholar, 14Wetterau J.R. Lin M.C.M. Jamil H. Biochim. Biophys. Acta. 1997; 1345: 136-150Crossref PubMed Scopus (286) Google Scholar), we the plasma for apoB-100 and apoB-48 by analysis. We plasma from these mice 2 following AdCre1 In animals that were on chow 3 was no in wild-type and heterozygous knockout but In homozygous knockout animals, apoB-100 was and apoB-48 was and In wild-type animals that were fed a high cholesterol diet 3 plasma apoB-100 and apoB-48 were on the with apoB-48 a the apoB-100 In the AdCre1-treated heterozygous floxed mice apoB-48 and apoB-100 we the plasma from AdCre1-treated homozygous floxed ℓMTP−/− mice, we found a marked reduction in the of the apoB-48 and an almost complete disappearance of the apoB-100 Therefore, inactivation of the ℓMTP locus in the liver to an complete absence of circulating apoB-100 and a marked in plasma These would be with the of apoB-100 and apoB-48 production by the liver with the of apoB-48 production in the small intestine in the AdCre1-treated homozygous floxed ℓMTP−/− mice. We the plasma lipoproteins in the of mice by FPLC Plasma was obtained from heterozygous floxed and homozygous floxed to 21 days following AdCre1 In animals, in with the ℓMTP−/− mice a complete absence of the and 4 and a In heterozygous animals, were in the and with no in the with wild-type animals that were fed a high cholesterol wild-type mice a a and a In the homozygous floxed ℓMTP−/− mice, was a complete absence of lipoproteins in the and There was a slight reduction in the Therefore, the apoB-100 production from the almost complete of MTP activity in the ℓMTP−/− mice was to a to high cholesterol diet The heterozygous mice, which intermediate ℓMTP and MTP activity an intermediate lipoprotein 4 The plasma lipoprotein triglyceride was in the in wild-type It was abolished in the of the marked in the ℓMTP−/− mice, we measured the rate of triglyceride production in these The intravenous administration of Triton WR1339 the of triglyceride-rich lipoproteins. The accumulation of plasma triglyceride following Triton treatment the triglyceride secretion rate from the liver and the animals were on a diet during this the secretion from the liver. As shown in B, the triglyceride secretion rate in ℓMTP−/− mice was to that in the wild-type The rate in heterozygous knockout mice was intermediate the absence of in the knockout animals was a result of failure of Pulse-chase experiments on hepatocytes isolated from wild-type and homozygous knockout animals revealed that was complete failure of secretion of apoB in ℓMTP−/− animals, which would for the absence of production in these production was normal 4 In we have produced a viable abetalipoproteinemia gene knockout mouse model using a strategy. We found that the liver-specific of the ℓMTP gene was to completely the plasma to a high cholesterol diet. knockout of the ℓMTP gene is embryonic lethal in the homozygous the conditional knockout mice be a valuable model for studying the and of abetalipoproteinemia as well as the role of MTP in apoB biosynthesis and the biogenesis of apoB-containing lipoproteins in We Dr. Frank Graham (McMaster University, Hamilton, Canada) for AdCre1, David Gordon and Jamil Squibb Pharmaceutical Research Institute) for to MTP and Dr. Oka and for with AdCre1 Dr. for and for