Somatic CelI and Molecular Genetics, Vol. 18, No. 5, 1992, pp. 443-450
Molecular Cloning and Nucleotide Sequence of cDNA Encoding a Functional Murine Low-Density-Lipoprotein Receptor 1 William J. Poivino, 2,3 David A. Dichek, 2 James Mason, 2 and W. French Anderson 2 2Molecular Hematology Branch, National Heart Lung and Blood Institute, Bethesda, Maryland 20892; and 3Clinical Pharmacology, Merck Sharp & Dohme Research Laboratories, Rahway, New Jersey 07065 Received 12 August 1992
A b s t r a c t - - I h e low-density-lipoprotein (LDL ) receptor is an important mediator of mammalian
cholesterol metabolism; its congenital absence in humans is' characterized by hypercholesterolemia, atherosclerosis, and coronary artery disease. We report here the identification and cloning of a cDNA encoding the murine LDL receptor. The cDNA insert is 4467 base pairs" in length and the deduced amino acid sequence bears" 78.2% homology with the reported human sequence. This murine cDNA was subcloned into a retroviral-based expression vectol; LmLSN1, and transfected into 3T3 cells. The production of functional LDL receptor was confirmed by tigand binding qf DiI-LDL cholesterol
INTRODUCTION
The low-density-lipoprotein (LDL) receptor is a 160-kDa transmembrane protein that functions in the uptake and metabolism of LDL. The congenital absence of functional LDL receptors in homozygous familial hypercholesterolemia (FH) leads to markedly elevated serum levels of LDL, atherosclerosis, and premature coronary artery disease with myocardial infarction and death often in the first decade of life. The heterozygous state, which afflicts approximately one in 500 in the United States, contributes to the prevalence of clinically evident coronary, artery disease (1). Significant advances have been made in our understanding of the molecular basis of FH. The nucleic acid sequence and derived amino acid sequence for the receptor protein
has been determined, in whole or in part, in the rabbit (2), cow (3, 4), rat (5), hamster (6), and in man (7-9). In addition, the specific mutations responsible for the FH phenotype have been determined in several human kindred (10-12) and in the Watanabe heritable hs,percholesterolemic (WHHL) rabbit (2). An animal model for the study of hypercholesterolemia is important, given the societal costs of the morbidity and mortality associated with atheroscterosis and coronary artery disease. Such a model would help both to elucidate the pathophysiology of this disorder and to evaluate the efficacy of potential therapeutic interventions. Thus an ideal animal model would provide large numbers of animals with a defined genetic background and a well-characterized phenotype similar to human atherosclerosis.
1Sequence data from this article has been deposited with the EMBL/GenBank Libraries under Accession No. X64414. 443 0740-7750/92/0700-0443506.50/0 © 1992 Plenum PublishingCorporation
444
At present, the animal models available for the study of LDL receptor deficiency are the W H H L rabbit and a recently described kindred of hypercholesterolemic rhesus monkeys (13). The W H H L rabbit model, while useful, has been limited by variability in the cholesterol levels of individual rabbits, which has been problematic in some studies (14) but not in others (15, 16). In addition the W H H L rabbit model is also limited by problems with fertility (C. Hansen, personal communication). Identification of the LDL receptor gene in the mouse would provide the means to attempt targeted gene replacement (17, 18) in mouse ES cells with potential generation of an LDL-receptor-deficient mouse. This is an attractive theoretical model for the study of hypercholesterolemia and/or atherosclerosis for the following reasons: (1) the specific mutation causing the receptor deficiency could be engineered as desired, (2) the genetic background of laboratory mice is well defined, and (3) the short generation time and ease of management of laboratory mice would allow investigations to be conducted in larger numbers and in a shorter time. We report here the identification and cloning of a cDNA encoding a functional murine LDL receptor. MATERIALS AND M E T H O D S
Materials. The mouse liver cDNA library in )t-ZAP was purchased from Stratagene (La Jolla, California). Oligonucleotides were a kind gift from Genetic Therapy, Inc. (Gaithersburg, Maryland) and were 32p end-labeled using T4-polynucleotide kinase. DH10B competent E. coli were obtained from Bethesda Research Laboratories (Bethesda, Maryland). Sanger sequencing was performed using a Sequenase kit (U.S. Biochemicals, Cleveland, Ohio). The plasmid pLXSN was kindly provided by Dr. A. Dusty Miller (Fred Hutchinson Cancer Center, Seattle, Washington). Lipoprotein-
Polvino et ai.
deficient serum and 1,1 '-dioctadecyl-3,3,3',3'tetramethyl-indocarbocyanine perchloratelabeled LDL (DiI-LDL) were purchased from Biomedical Technologies, Inc. (Stoughton, Massachusetts). Screening of cDNA Library. Radiolabeled oligonucleotides were used to screen the cDNA library. Oligonucleotides 20 or 21 nucleotides in length were synthesized complementary to the 5' and 3' ends of the recently described rat cDNA coding region. The cDNA library was plated at a density of 50,000 PFU/150-mm plate on a lawn of BB4 E. coli. The plaques were lifted onto nitrocellulose filters in duplicate, denatured in situ, baked onto the filters, and hybridized according to previously reported procedures (19). Plaques that hybridized to both oligonucleotides employed as probes were identified and extracted from the original plates. Two additional rounds of plaque purification were required to purify the isolates to clonality. The cDNA insert was subcloned into the pBluescript plasmid according to the manufacturer's in vivo excision protocol (Stratagene). The resultant plasmid was transformed into DH10B competent E. coli and the plasmid purified on a CsC1 gradient (19). Sequencing. The pBluescript plasmid containing the desired cDNA insert was sequenced utilizing a Sequenase kit according to the manufacturer's instructions. A series of oligonucleotides were synthesized to allow sequencing of approximately 300 nucleotides from each individual primer. Both strands of the insert were sequenced in their entirety. Subcloning and Transfection. Restriction endonuclease digestions and blunt-end ligations were performed according to previously reported procedures (19). Plasmids were CaPO4 coprecipitated (19) into subconfluent cocultures of PA317 (20) and GP+E-86 (21) retroviral packaging cell lines. The transfected cells were grown for seven days in DMEM supplemented with 10% fetal bovine
Mouse LDL Receptor Sequence and Expression
serum (37°C, 5% CO2) and then selected in 0.8 mg/ml G418 for an additional four days. Staining for LDL Receptor. Prior to analyzing for LDL binding, cells were grown for 24 h in DMEM supplemented with 10% bovine lipoprotein-deficient serum. LDL receptor activity was identified by the addition of DiI-labeled LDL to the culture medium at 10 p.g/ml tbr 5 h at 37°C. Unbound ligand was removed by washing with phosphate-buffered saline. The cells were examined using fluorescense microscopy with standard rhodamine filters. Sequence Analysis. The nucleotide sequence was analyzed with MacVector Version 3.0. The derived amino acid sequence was compared to the amino acid sequences of other mammalian LDL receptors utilizing homology matrix analysis and maximal alignment functions. RESULTS
Isolation of cDNA and Sequence Analysis. A total of 5 x 10s plaques were screened to yield one plaque that hybridized to probes complementary to the 5' and 3' ends of the rat LDL receptor cDNA coding region. The resultant clone contained a 4.5-kb cDNA insert, which included a start and stop codon separated by 2586 bp of open reading frame (Fig. 1). There were flanking regions of 121 bp 5' to the initiator ATG codon, and 1760 bp 3' to the stop codon. This extensive 3' tail is seen also in the human LDL receptor cDNA, which contains a tail 2.5 kb in length
(7). The deduced mouse amino acid sequence was 862 amino acids in length and was organized into the five structural domains that have been described for other species (2, 3, 5, 7). The mouse receptor protein was composed of an N-terminal sequence of 21 hydrophobic amino acids analogous to the leader sequence of the human precursor molecule. This sequence was followed by a 293-amino acid ligand-
445
binding domain, rich in cysteine residues, comprised of seven direct repeats of approximately 40 amino acids each with a nineamino acid "spacer" between repeats 4 and 5. The next region of approximately 400 amino acids comprised the region of homol-~ ogy with epidermal growth factor precursor, including the consensus sequence of three direct repeats previously reported in the rabbit (2) and human (7) amino acid sequences. This region was followed by a 60-amino acid sequence containing 20 serine and threonine residues, which corresponds to the region of O-linked glycosylation in the human protein. The transmembrane domain was composed of 22 hydrophobic amino acids flanked by arginine residues. The terminal 50 amino acids constituted the cytosolic domain of the protein and was highly conserved across species. Homology matrix analysis of the derived amino acid sequence for the mouse LDL receptor revealed the direct repeat organization in the ligand-binding and EGF homology domains (Fig. 2a). The degree of homology among mouse, rat, human, bovine, rabbit, and hamster amino acid sequences was striking, as represented by the comparison of the mouse amino acid sequence to that of the human protein (Fig. 2b). The one clear area of divergence was in the O-linked sugar domain. Not surprisingly, the mouse amino acid sequence was most similar to the rat sequence in this region. Closer inspection of these O-linked sugar domains revealed the presence of an 18-amino acid insert in the rat sequence that was not present in any other species. Of note, this insert contained two of the three consecutive six-residue direct repeats that occurred within a 24-residue stretch of the rat protein (5). With the exclusion of this 18-amino acid insert from the rat amino acid sequence, the remaining rat O-linked sugar domain had 73.2% homology with the corresponding mouse sequence. It is possible that this structure may have,
446
Polvino et al.
~Ja' gg~ gP
a a~xe
~ = ¢ C ~ G C T C A G C T G C C T ~ CT gC C C ~ ~ g T T ~ ¢ eCAT C T T C A T T T C C T T ~ C
~ T A ~ G T
T T T A T A T A ~ T A T T GT C T ~ G = A C ~ ~
=
~
~TG~eAC
~AC
TG G C T g T g ~ T ~ T ~ C ~
T C ~ C A C T ~ C ~ G ~ W ~ m ~ C C
T¢ TG C C T g ~ T C T
2~8
~ G G ~ C C C T G ~ A ~
3098
Fig. 1. Nucleic acid and derived amino acid sequences of the m o u s e L D L receptor c D N A insert. T h e sequences
corresponding to signal sequence and transmembrane domains are underlined. Nucleotide and amino acid sequence numbers are indicated. arisen in the rat via duplication of genetic material during the evolutionary process (22). Analysis of homology based on the h u m a n exon-intron organization (Fig. 3)
revealed a high degree of conservation across species with the notable exceptions of exon 1 and exons 15 and 16, which correspond to the leader sequence and to the regions of O-linked glycosylation, respectively. The de-
Mouse LDL Receptor Sequence and Expression ~ " .'t'-"- "'." loo-
Molecular Cloning and Nucleotide Sequence of cDNA Encoding a Functional Murine Low-Density-Lipoprotein Receptor 1 William J. Poivino, 2,3 David A. Dichek, 2 James Mason, 2 and W. French Anderson 2 2Molecular Hematology Branch, National Heart Lung and Blood Institute, Bethesda, Maryland 20892; and 3Clinical Pharmacology, Merck Sharp & Dohme Research Laboratories, Rahway, New Jersey 07065 Received 12 August 1992
A b s t r a c t - - I h e low-density-lipoprotein (LDL ) receptor is an important mediator of mammalian
cholesterol metabolism; its congenital absence in humans is' characterized by hypercholesterolemia, atherosclerosis, and coronary artery disease. We report here the identification and cloning of a cDNA encoding the murine LDL receptor. The cDNA insert is 4467 base pairs" in length and the deduced amino acid sequence bears" 78.2% homology with the reported human sequence. This murine cDNA was subcloned into a retroviral-based expression vectol; LmLSN1, and transfected into 3T3 cells. The production of functional LDL receptor was confirmed by tigand binding qf DiI-LDL cholesterol
INTRODUCTION
The low-density-lipoprotein (LDL) receptor is a 160-kDa transmembrane protein that functions in the uptake and metabolism of LDL. The congenital absence of functional LDL receptors in homozygous familial hypercholesterolemia (FH) leads to markedly elevated serum levels of LDL, atherosclerosis, and premature coronary artery disease with myocardial infarction and death often in the first decade of life. The heterozygous state, which afflicts approximately one in 500 in the United States, contributes to the prevalence of clinically evident coronary, artery disease (1). Significant advances have been made in our understanding of the molecular basis of FH. The nucleic acid sequence and derived amino acid sequence for the receptor protein
has been determined, in whole or in part, in the rabbit (2), cow (3, 4), rat (5), hamster (6), and in man (7-9). In addition, the specific mutations responsible for the FH phenotype have been determined in several human kindred (10-12) and in the Watanabe heritable hs,percholesterolemic (WHHL) rabbit (2). An animal model for the study of hypercholesterolemia is important, given the societal costs of the morbidity and mortality associated with atheroscterosis and coronary artery disease. Such a model would help both to elucidate the pathophysiology of this disorder and to evaluate the efficacy of potential therapeutic interventions. Thus an ideal animal model would provide large numbers of animals with a defined genetic background and a well-characterized phenotype similar to human atherosclerosis.
1Sequence data from this article has been deposited with the EMBL/GenBank Libraries under Accession No. X64414. 443 0740-7750/92/0700-0443506.50/0 © 1992 Plenum PublishingCorporation
444
At present, the animal models available for the study of LDL receptor deficiency are the W H H L rabbit and a recently described kindred of hypercholesterolemic rhesus monkeys (13). The W H H L rabbit model, while useful, has been limited by variability in the cholesterol levels of individual rabbits, which has been problematic in some studies (14) but not in others (15, 16). In addition the W H H L rabbit model is also limited by problems with fertility (C. Hansen, personal communication). Identification of the LDL receptor gene in the mouse would provide the means to attempt targeted gene replacement (17, 18) in mouse ES cells with potential generation of an LDL-receptor-deficient mouse. This is an attractive theoretical model for the study of hypercholesterolemia and/or atherosclerosis for the following reasons: (1) the specific mutation causing the receptor deficiency could be engineered as desired, (2) the genetic background of laboratory mice is well defined, and (3) the short generation time and ease of management of laboratory mice would allow investigations to be conducted in larger numbers and in a shorter time. We report here the identification and cloning of a cDNA encoding a functional murine LDL receptor. MATERIALS AND M E T H O D S
Materials. The mouse liver cDNA library in )t-ZAP was purchased from Stratagene (La Jolla, California). Oligonucleotides were a kind gift from Genetic Therapy, Inc. (Gaithersburg, Maryland) and were 32p end-labeled using T4-polynucleotide kinase. DH10B competent E. coli were obtained from Bethesda Research Laboratories (Bethesda, Maryland). Sanger sequencing was performed using a Sequenase kit (U.S. Biochemicals, Cleveland, Ohio). The plasmid pLXSN was kindly provided by Dr. A. Dusty Miller (Fred Hutchinson Cancer Center, Seattle, Washington). Lipoprotein-
Polvino et ai.
deficient serum and 1,1 '-dioctadecyl-3,3,3',3'tetramethyl-indocarbocyanine perchloratelabeled LDL (DiI-LDL) were purchased from Biomedical Technologies, Inc. (Stoughton, Massachusetts). Screening of cDNA Library. Radiolabeled oligonucleotides were used to screen the cDNA library. Oligonucleotides 20 or 21 nucleotides in length were synthesized complementary to the 5' and 3' ends of the recently described rat cDNA coding region. The cDNA library was plated at a density of 50,000 PFU/150-mm plate on a lawn of BB4 E. coli. The plaques were lifted onto nitrocellulose filters in duplicate, denatured in situ, baked onto the filters, and hybridized according to previously reported procedures (19). Plaques that hybridized to both oligonucleotides employed as probes were identified and extracted from the original plates. Two additional rounds of plaque purification were required to purify the isolates to clonality. The cDNA insert was subcloned into the pBluescript plasmid according to the manufacturer's in vivo excision protocol (Stratagene). The resultant plasmid was transformed into DH10B competent E. coli and the plasmid purified on a CsC1 gradient (19). Sequencing. The pBluescript plasmid containing the desired cDNA insert was sequenced utilizing a Sequenase kit according to the manufacturer's instructions. A series of oligonucleotides were synthesized to allow sequencing of approximately 300 nucleotides from each individual primer. Both strands of the insert were sequenced in their entirety. Subcloning and Transfection. Restriction endonuclease digestions and blunt-end ligations were performed according to previously reported procedures (19). Plasmids were CaPO4 coprecipitated (19) into subconfluent cocultures of PA317 (20) and GP+E-86 (21) retroviral packaging cell lines. The transfected cells were grown for seven days in DMEM supplemented with 10% fetal bovine
Mouse LDL Receptor Sequence and Expression
serum (37°C, 5% CO2) and then selected in 0.8 mg/ml G418 for an additional four days. Staining for LDL Receptor. Prior to analyzing for LDL binding, cells were grown for 24 h in DMEM supplemented with 10% bovine lipoprotein-deficient serum. LDL receptor activity was identified by the addition of DiI-labeled LDL to the culture medium at 10 p.g/ml tbr 5 h at 37°C. Unbound ligand was removed by washing with phosphate-buffered saline. The cells were examined using fluorescense microscopy with standard rhodamine filters. Sequence Analysis. The nucleotide sequence was analyzed with MacVector Version 3.0. The derived amino acid sequence was compared to the amino acid sequences of other mammalian LDL receptors utilizing homology matrix analysis and maximal alignment functions. RESULTS
Isolation of cDNA and Sequence Analysis. A total of 5 x 10s plaques were screened to yield one plaque that hybridized to probes complementary to the 5' and 3' ends of the rat LDL receptor cDNA coding region. The resultant clone contained a 4.5-kb cDNA insert, which included a start and stop codon separated by 2586 bp of open reading frame (Fig. 1). There were flanking regions of 121 bp 5' to the initiator ATG codon, and 1760 bp 3' to the stop codon. This extensive 3' tail is seen also in the human LDL receptor cDNA, which contains a tail 2.5 kb in length
(7). The deduced mouse amino acid sequence was 862 amino acids in length and was organized into the five structural domains that have been described for other species (2, 3, 5, 7). The mouse receptor protein was composed of an N-terminal sequence of 21 hydrophobic amino acids analogous to the leader sequence of the human precursor molecule. This sequence was followed by a 293-amino acid ligand-
445
binding domain, rich in cysteine residues, comprised of seven direct repeats of approximately 40 amino acids each with a nineamino acid "spacer" between repeats 4 and 5. The next region of approximately 400 amino acids comprised the region of homol-~ ogy with epidermal growth factor precursor, including the consensus sequence of three direct repeats previously reported in the rabbit (2) and human (7) amino acid sequences. This region was followed by a 60-amino acid sequence containing 20 serine and threonine residues, which corresponds to the region of O-linked glycosylation in the human protein. The transmembrane domain was composed of 22 hydrophobic amino acids flanked by arginine residues. The terminal 50 amino acids constituted the cytosolic domain of the protein and was highly conserved across species. Homology matrix analysis of the derived amino acid sequence for the mouse LDL receptor revealed the direct repeat organization in the ligand-binding and EGF homology domains (Fig. 2a). The degree of homology among mouse, rat, human, bovine, rabbit, and hamster amino acid sequences was striking, as represented by the comparison of the mouse amino acid sequence to that of the human protein (Fig. 2b). The one clear area of divergence was in the O-linked sugar domain. Not surprisingly, the mouse amino acid sequence was most similar to the rat sequence in this region. Closer inspection of these O-linked sugar domains revealed the presence of an 18-amino acid insert in the rat sequence that was not present in any other species. Of note, this insert contained two of the three consecutive six-residue direct repeats that occurred within a 24-residue stretch of the rat protein (5). With the exclusion of this 18-amino acid insert from the rat amino acid sequence, the remaining rat O-linked sugar domain had 73.2% homology with the corresponding mouse sequence. It is possible that this structure may have,
446
Polvino et al.
~Ja' gg~ gP
a a~xe
~ = ¢ C ~ G C T C A G C T G C C T ~ CT gC C C ~ ~ g T T ~ ¢ eCAT C T T C A T T T C C T T ~ C
~ T A ~ G T
T T T A T A T A ~ T A T T GT C T ~ G = A C ~ ~
=
~
~TG~eAC
~AC
TG G C T g T g ~ T ~ T ~ C ~
T C ~ C A C T ~ C ~ G ~ W ~ m ~ C C
T¢ TG C C T g ~ T C T
2~8
~ G G ~ C C C T G ~ A ~
3098
Fig. 1. Nucleic acid and derived amino acid sequences of the m o u s e L D L receptor c D N A insert. T h e sequences
corresponding to signal sequence and transmembrane domains are underlined. Nucleotide and amino acid sequence numbers are indicated. arisen in the rat via duplication of genetic material during the evolutionary process (22). Analysis of homology based on the h u m a n exon-intron organization (Fig. 3)
revealed a high degree of conservation across species with the notable exceptions of exon 1 and exons 15 and 16, which correspond to the leader sequence and to the regions of O-linked glycosylation, respectively. The de-
Mouse LDL Receptor Sequence and Expression ~ " .'t'-"- "'." loo-
