Nucleotide Sequence Encoding the Carboxyl-Terminal Half of Apolipoprotein B from Spontaneously Hypercholesterolemic Pigs
Total Page:16
File Type:pdf, Size:1020Kb
Nucleotide sequence encoding the carboxyl-terminal half of apolipoprotein 6 from spontaneously hypercholesterolemic pigs Charles Purtell,’ Nobuyo Maeda, t David L. Ebert,’ Mary Kaiser,’ Sissel Lund-Katz; * Stephen L. Sturley,’ Voula Kodoyianni,tt Kurt Grunwald; David N. Nevin,’ Robert J. Aiello,’ and Alan D. Attie’.* Departments of Biochemistry and Comparative Biosciences: University of Wisconsin-Madison, Madison, WI 53706; Department of Pathology,t University of North Carolina, Chapel Hill, NC 27599; Department of Physiology and Biochemistry: * Medical College of Pennsylvania, Philadelphia, PA 19129; and Department of Biochemistry and Molecular Biology Laboratory, tt University of Wisconsin-Madison, Madison, WI 53706 Abstract Previous studies from this laboratory characterized population (1). Approximately two-thirds of plasma the hypercholesterolemia of pigs with a mutant allele of cholesterol is transported on low density lipoprotein apolipoprotein B (apoB), designated Lpb5. This apoB allele is as- sociated with low density lipoprotein (LDL) particles deficient in (LDL) particles. The best-characterized mutations as- binding to the LDL receptor. To identify potential causative mu- sociated with hypercholesterolemia are those affecting the tations in Lpb5 DNA, 10.6 kb of genomic DNA, encoding the function and expression of the LDL receptor, which is carboxyl-terminal 58% of apoB were sequenced from the Lpb5 primarily responsible for LDL catabolism (2). However, allele and from an allele encoding phenotypically normal apoB. since LDL receptor mutations account for only about 1% Comparison of the two DNA sequences revealed 33 polymor- phisms, 13 of which resulted in amino acid poIymorphisms. To of individuals with hypercholesterolemia, mutations in determine whether any of the amino acids at the polymorphic other genes undoubtedly contribute to the high preva- positions in Lpb5-encoded apoB were unique to that isoform, lence of hypercholesterolemia. those positions were sequenced in four other pig apoB alleles en- Apolipoprotein B (apoB) is the major protein compo- coding phenotypically normal apoB. None of the amino acids nent of LDL and a ligand for the LDL receptor. Im- were by themselves uniquely encoded by the Lpb5 allele. However, a unique haplotype consisting of Asp,,, in conjunc- munogenetic studies in pigs first showed that variation in tion with Ala3s17 distinguished the Lpb5-encoded apoB from all apoB structure might be associated with hyper- other allelic isoforms sequenced in this region. To gain insight cholesterolemia (3). In these studies, pigs were cross- into changes in the tertiary structure of the mutant apoB, ‘3C- immunized with pig serum and the resulting allo-antisera NMR analysis of LDL reductively methylated with [I3C]- were used to follow the inheritance pattern of epitopes, formaldehyde was performed. LDL has lysine residues that ti- trate at pH 10.5 and others that titrate at pH 8.9. The latter some of which were subsequently shown to reside on apoB residues are thought to include those involved in the interaction (4). This approach enabled the discrimination of eight pig of LDL with the LDL receptor. LDL from Lpb5 pigs possessed apoB alleles, designated Lpb’-8 (3). a smaller proportion of lysine residues titrating at pH 8.9 than One pig apoB allele, designated Lpb5, occurs in pigs did LDL from non-Lpb5 pigs, suggesting that the Lpb5-encoded that have an elevation in plasma LDL (4). Cultured apoB is altered in a manner affecting the microenvironment of particular lysine residues.-Purtell, C., N. Maeda, D. L. fibroblasts from these pigs express normal LDL receptor Ebert, M. Kaiser, S. Lund-Katz, S. L. Sturley, V. activity (4). The LDL elevation is due to an 85% lower Kodoyianni, K. Grunwald, D. N. Nevin, R. J. Aiello, and affinity of this apoB variant for the LDL receptor (5) and A. D. Attie. Nucleotide sequence encoding the carboxyl- a pronounced overproduction of a buoyant LDL subspe- terminal half of apolipoprotein B from spontaneously hyper- cies (6, 7). cholesterolemic pigs. J. Lipid Res. 1993. 34: 1323-1335. Supplementary key words low density lipoproteins DNA LDL receptor-related protein Abhreviations: apo, apolipoprotein; LDL, low density lipoprotein; PCR, polymerase chain reaction; CD, circular dichroism; NMR, Hypercholesterolemia is a major risk factor for coro- nuclear magnetic resonance; LRP, LDL receptor-related protein; PBS, nary heart disease. Genetic factors contribute about 50% phosphate-buffered saline. to the variance in plasma cholesterol levels in the human IT0 whom correspondence should be addressed. Journal of Lipid Research Volume 34, 1993 1323 This is an Open Access article under the CC BY license. The cross-immunization experiments that identified residues were consolidated into one chemical shift, sug- the allelic apoB isoforms in pigs constituted a non- gesting that the more complex pattern was a reflection of prejudiced screen for polymorphic proteins in plasma. A the unique environments of two distinct pools of apoB ly- similar cross-immunization occurs in patients undergoing sine within the context of an intact LDL particle (12). multiple blood transfusions. Analysis of sera from such Lund-Katz et al. (13) showed a significant decline in the patients led to an immunological classification of human proportion of lysines titrating at the lower pH in apoB apoB isoforms, the ag system (8). Because the most anti- possessing the Arg3500+ Gln mutation. genic protein identified in these analyses was apoB, it was We recently cloned and sequenced the portion of a nor- inferred that apoB is an unusually polymorphic protein. mal pig apoB allele, Lpb7, encoding a carboxyl-terminal More recently, studies of a highly repetitive region down- segment of apoB (14). This region is suspected to contain stream of the human apoB gene indicate that this locus is domains necessary for binding to the LDL receptor. In also unusually polymorphic in non-coding regions of the addition, we previously sequenced a short (1.1 kb) seg- gene (9). ment from the Lpb5 allele (15). Although several sequence Most of the human apoB polymorphisms do not appear differences between Lpb5 and Lpb7 were detected, none of to be correlated with a clinical phenotype. However, a the individual sequence variations were found to be relatively common Arg + Gln substitution at amino acid unique to the Lpb5 allele when three other apoB alleles residue 3500 is associated with an apoB molecule essen- from non-hypercholesterolemic pigs were sequenced wi- tially devoid of receptor binding activity (10). Since apoB thin that segment. We concluded that mutations responsi- truncations that delete the region surrounding Arg3500do ble for the defective receptor recognition by the Lpb5- not reduce the binding affinity of apoB for the LDL encoded apoB might lie outside the putative receptor receptor (ll), it would appear that the Arg + Gln substi- binding domain encoded by the 1.1-kb segment (15). Al- tution prohibits receptor binding by altering the structure ternatively, the causative alteration is more complex than of a distant segment of apoB. a single missense mutation. This interpretation is supported by structural analysis In the present study, we have sequenced 10,686 and of intact mutant apoB. Lund-Katz et al. (12) developed an 10,403 bp at the 3' ends of pig apoB alleles Lpb5 and Lpb7, NMR technique for assessing the microenvironment of respectively (the difference in length relates to a unique lysine residues on apoB. By labeling the exposed lysine 283-bp insertion in intron 28 in the Lpb5 allele). The se- residues via reductive methylation with [ Wlformalde- quence described here encodes 58% of the apoB protein. hyde, they were able to discern two distinct lysine chemi- This enabled the identification of all positions where the cal shifts, corresponding to distinct pK,s for the E- two alleles differ within this segment. In addition, the I3C- amines. Upon denaturation in 7 M urea, the lysine NMR technique was applied to the comparison of normal Poly A site 848 editing site I KH E KX I P 1 35 6 7 8 9 1011 13 14 15 1618 20 -+ +-+3 4-4-44-34- 3 34-3 4- 34- 3 34 24 12 17 19 DEA 154 (0.3) DEP 900 E 1.3-5 DEC 11 DN1-1 HX5 BX3 m -I = -- DEA 154-6- DEC 11-4 VX200 SH2 HX7 I-- --CPC 49 VX 1.3-8x3 BH 2 - ---VX 600 7 DN 1-3 DEA 96 (A 96) - NllO B VRA 5 - Fig. 1. DNA sequencing strategy. The shaded areas indicate introns. Letters indicate restriction endonuclease sites used to generate subclones: B, BumH1; S, Sac I; E, Eco R1; P, Psl I; X, Xba I; K, Kpn I; Bg, Bgl I, H, Hind 111. Heavy bars indicate segments subcloned into plasmids. Numbers indicate the locations and orientations of the synthetic oligonucleotide primers. 1324 Journal of Lipid Research Volume 34, 1993 versus mutant pig LDL particles to assess the relative con- Sequence analysis formations of lysine residues in particles with varying More than 90% of the reported sequence, including all affinities for the LDL receptor. exon sequence, was obtained three or more times. The re- maining 10% was performed twice. In more than 50% of the reported sequence, both strands were sequenced. Au- MATERIALS AND METHODS toradiograms were read at least twice by different people. The sequence was entered by typing or via digitizer into Subcloning of genomic DNA fragments the University of Wisconsin Genetics Computer Group Genomic clones (7 .O-10.4 kb) from non-hypercholes- (UWGCG) program “SeqEd.” Subsequent analysis was terolemic (Lpb7) and hypercholesterolemic (Lpb5)pigs (14) done employing other GCG programs (21) and DNA were digested with appropriate restriction enzymes; frag- Strider (22). Homology searches in the Genbank database ments were separated by agarose gel electrophoresis, and were carried out using the FASTA program (23) accessed isolated from agarose gel slices (16). The gel-purified frag- through the EMBL Network file server (24, 25).