A Mutation in the Human Apolipoprotein A-I Gene DOMINANT EFFECT ON THE LEVEL AND CHARACTERISTICS OF PLASMA HIGH DENSITY LIPOPROTEINS*

Epidemiologic and genetic data suggest an inverse relationship between plasma high density lipoprotein (HDL) cholesterol and the incidence of premature coronary artery disease. Some of the defects leading to low levels of HDL may be a consequence of mutations in the genes coding for HDL apolipoproteins A-I and A-I1 or for enzymes that modify these particles. A proband with plasma apoA-I and HDL cholesterol that are below 15% of normal levels and with marked bilateral arcus senilis was shown to be heterozygous for a 45-base pair deletion in exon four of the apoA-I gene. This most likely represents a de novo mutation since neither parent carries the mutant allele. The protein product of this allele is predicted to be missing 15 of the 22 amino acids comprising the third amphipathic helical domain. The HDL of the proband and his family were studied. Using anti-A-I and anti-A-I1 immunosorbents we found three popula- tions of HDL particles in the proband. One contained both apoA-I and A-11, Lp(A-I w A-11); one contained apoA-I but no A-11, Lp(A-I w/o A-11); and the third (an unusual one) contained apoA-I1 but no The was in a for followed by quick on After a spin down the condensation, the template-primer was mixed with pl of labeling mix (20 dithiothreitol; 0.6 each of dGTP, dTTP, dCTP; pCi of "'S-dATP (Du Pont-New England Nuclear; specific 1,000 Ci/mmol); and 2 units of Sequenase. The labeling reaction was carried out at room temperature for 2 min. The termi- nation reaction was then done by adding 3.5 p1 of labeling mix to 2.5 pl of each of the four termination mixes (50 mM NaC1; 80 p~ each of dGTP, dATP, dTTP, dCTP; plus 8.0 FM of one of the four dideoxy-ribonucleotides), and incubated at 37" C for 5 min. Three pl of stop solution (95% formamide, 20 mM EDTA, 0.05% bromphenol blue, 0.05% xylene cyanol) was then added to the mixture. The samples were heated to 75" C for 2 min immediately before loading on an 8% polyacrylamide, 7 M urea gel for sequence analysis. At least 200 bases could be read after an overnight exposure.

Epidemiologic and genetic data suggest an inverse relationship between plasma high density lipoprotein (HDL) cholesterol and the incidence of premature coronary artery disease. Some of the defects leading to low levels of HDL may be a consequence of mutations in the genes coding for HDL apolipoproteins A-I and A-I1 or for enzymes that modify these particles.
A proband with plasma apoA-I and HDL cholesterol that are below 15% of normal levels and with marked bilateral arcus senilis was shown to be heterozygous for a 45-base pair deletion in exon four of the apoA-I gene. This most likely represents a de novo mutation since neither parent carries the mutant allele. The protein product of this allele is predicted to be missing 15 of the 22 amino acids comprising the third amphipathic helical domain. The HDL of the proband and his family were studied. Using anti-A-I and anti-A-I1 immunosorbents we found three populations of HDL particles in the proband. One contained both apoA-I and A-11, Lp(A-I w A-11); one contained apoA-I but no A-11, Lp(A-I w/o A-11); and the third (an unusual one) contained apoA-I1 but no A-I. Only Lp(A-I w A-11) and (A-I w/o A-11) were present in the plasma of the proband's parents and brother. Analysis of the HDL particles of the proband by sodium dodecyl sulfate-polyacrylamide gel electrophoresis revealed two protein bands with a molecular mass differing by 6% in the vicinity of 28 kDa whereas the HDL particles of the family members exhibited only a single apoA-I band. The largely dominant effect of this mutant allele (designated apoA-Iseattle) on HDL levels suggests that HDL particles containing any number of mutant apoA-I polypeptides are catabolized rapidly.
Apolipoprotein (apo)' A-I, a polypeptide of 243 amino acids, is synthesized in the liver and small intestine and secreted into plasma in association with lipoprotein particles. ApoA-I is the principal protein of high density lipoprotein (HDL), and its biological functions include activation of lecithincholesterol acyltransferase, the enzyme responsible for cholesterol esterification in plasma, and solubilization of lipids in an aqueous environment (for reviews see Refs. 1 and 2). The other major protein of HDL particles is apoA-11. The entire sequence and structure of the apoA-I gene (located on chromosome 11 q13) have been determined (3)(4)(5). The gene encompasses some 2 kilobases of DNA and is composed of four exons. The fourth exon contains six 66-bp tandemly repeated homologous segments (codons 99-230) that code for six 22-amino repeats believed to form amphipathic helices that mediate the above biological functions (3,6,7).
Considerable interest has been generated in the metabolism of apoA-I and HDL mainly because of results of both epidemiologic and genetic studies which revealed an inverse relationship between low plasma HDL and apoA-I levels and the incidence of premature coronary heart disease (8)(9)(10)(11)(12). These results support the hypothesis that HDL particles promote the transport of cholesterol from peripheral cells to the liver for eventual removal into the biliary tract (13). HDL particles are heterogeneous with respect to size, density, lipid and lipoprotein content. Immunoaffinity chromatography identifies two major classes of HDL particles based on apolipoprotein content: particles containing apoA-I without apoA-I1 and particles containing apoA-I and apoA-I1 (14).
The complete absence of HDL particles has been reported in three kindreds. One of these was shown to be caused by a deletion of the entire apoA-I/C-III/A-IV gene complex (15); another was a result of a rearrangement (inversion) involving the apoA-I and C-I11 genes (10); and the third was caused by the synthesis of a truncated protein, presumably due to premature termination of translation (16).
More than a dozen structural variants of apoA-I have been detected by virtue of altered electrophoretic mobility and later shown to be the result of single amino acid substitutions (1). Only two of these variants, however, were associated with diminished HDL levels. The first, referred to as apoA-IMilano, resulted from the substitution of cysteine for arginine at position 173 (17,18). Individuals heterozygous for this variant had plasma HDL and apoA-I levels approximately 33 and 60% of normal, respectively, but no evidence of atherosclerosis. Low HDL levels were shown to be most likely caused by a more rapid catabolism of the abnormal apoA-I allele. Heterozygotes for the second variant, Pro16s + Arg, were observed in six unrelated families to have significantly lower HDL and apoA-I levels (58 and 62%, respectively) relative to noncarriers, with no associated hypertriglyceridemia (19). In this report we describe a deletion mutation in the apoA-I gene which has a dominant effect on plasma apoA-I and HDL pounds (73 kg). His medical history included tonsillectomy twice in childhood (no abnormal coloration reported), and hay fever-type allergies to a variety of allergens determined by skin testing. He reported smoking four cigarettes and drinking a glass of wine or beer daily. Lipid analyses of blood samples obtained at this and three subsequent visits in 3.5 years showed a gradual increase in plasma cholesterol (163, 181, 216, and 209 mg/dl) and triglyceride (216, 286, 345, and 411 mg/dl) and a slight decrease in HDL cholesterol (7,7,6, and 4 mg/dl). ApoA-I and A-I1 repeatedly measured within the 10-18 mg/dl range.
The family history disclosed no evidence of premature cardiovascular disease in either parent. The father (H. S.), age 67, was treated with diltiazem HC1 for hypertension and allopurinol for hyperuricemia. A radionuclide exercise test had raised the possibility of a prior undetected myocardial infarction but was not definite. The mother (J. S.), age 62, had mild diabetes treated with insulin and hypothyroidism treated with thyroid hormone replacement. She also took lovastatin (Mevacor) and postmenopausal estrogen, which may have increased her HDL. Other medications included the antidepressants nortriptyline (Pamelor) and amitriptyline (Elavil), and Bactrim antibiotic therapy. The brother (B. S.), age 36, had no known clinical disease. Paternal and maternal grandmothers reportedly died of heart failure a t ages 72 and 64, and the grandfathers on the paternal and maternal sides died of trauma and tuberculosis, respectively. There was no known consanguinity.
Preparation and Southern Blot Analysis of DNA-DNA was extracted from peripheral blood leukocytes by the proteinase K-phenol method on an Applied Biosystems (Foster City, CA) model 340A nucleic acid extractor according to the protocol provided by the manufacturer. Conditions for Southern blot analysis were as described (20,21). The apoA-I cDNA probe (pAI-113) (3) was a kind gift of J. Breslow (Rockefeller University, New York).
PCR Amplification-Oligodeoxynucleotide primers used in amplification and sequencing were synthesized on an Applied Biosystems model 380B DNA synthesizer. The sequence and position of these primers are shown in Table I. Three fragments of the apoA-I gene of a normal control and of the proband were PCR amplified (22) using a Perkin-Elmer Cetus Instruments thermal cycler and amplification

+1692
" The number indicates the nucleotide position of the 5' end of the primer on the apoA-I gene, +1 being the first base of exon 1. The numbering system is according to that given in Ref. 3. However, a subsequent publication (53) from the same laboratory designated the transcription initiation site to be 61 bp upstream from that in the original study.
kit. Fragment 1 included 199 bp of the 5' upstream regulatory region, exon 1, and the 5' 17 bp of intron 1 (usingprimers 1 and 2). Fragment 2 encompassed exons 2 and 3 and the intron exon junctions (using primers 3 and 4). Fragment 3 contained exon 4 and the intron 3-exon 4 junction (using primers 5 and 6). Thirty cycles of amplification were employed with 1 min of each of denaturation a t 94" C , annealing a t 58-66" C, and extension a t 72" C.
Direct Sequencing of PCR-amplified DNA-PCR products were purified from the unincorporated dNTPs and nonspecific DNA products by electrophoresis in a low melting point agarose gel (Seaplaque agarose, FMC BioProducts). The isolated gel slice containing the PCR product was then extracted twice with phenol to remove the agarose, and the DNA was then precipitated with ethanol and 0.3 M NaOAc.
Reagents of the Sequenase kit (U. S. Biochemical Corp.) were used for sequencing reactions according to the following procedure. About 1 pmol of the double-stranded PCR product was mixed with 5 pmol of sequencing primer (either one of the PCR primers) in 10 pI of sequencing buffer (40 mM Tris-HC1, pH 7.5, 20 mM MgC1, and 50 mM NaC1). The template-primer mix was heated in a boiling water bath for 5 min followed by quick chilling on ice. After a quick spin to bring down the condensation, the template-primer was mixed with 5 pl of labeling mix (20 mM dithiothreitol; 0.6 p M each of dGTP, dTTP, dCTP; 10 pCi of "'S-dATP (Du Pont-New England Nuclear; specific activity > 1,000 Ci/mmol); and 2 units of Sequenase. The labeling reaction was carried out a t room temperature for 2 min. The termination reaction was then done by adding 3.5 p1 of labeling mix to 2.5 pl of each of the four termination mixes (50 mM NaC1; 80 p~ each of dGTP, dATP, dTTP, dCTP; plus 8.0 FM of one of the four dideoxyribonucleotides), and incubated at 37" C for 5 min. Three pl of stop solution (95% formamide, 20 mM EDTA, 0.05% bromphenol blue, 0.05% xylene cyanol) was then added to the mixture. The samples were heated to 75" C for 2 min immediately before loading on an 8% polyacrylamide, 7 M urea gel for sequence analysis. At least 200 bases could be read after an overnight exposure.
Lipoprotein Fractionation-Fractionation of very low density lipoproteins (VLDL), LDL, and HDL was performed as described (23 Isolation of HDL Particles-HDL particles differing in their apoA-I and A-I1 contents were isolated by a previously established immunoaffinity chromatography procedure (14,24) with the following modifications. Plasma aliquots (12 ml from the proband and 2-4 ml from his parents and brother) were adsorbed sequentially with dextran sulfate-cellulose, anti-A-11-Sepharose CL-4B, and anti-A-I-Sepharose CL-4B to remove, respectively, all apoB-containing lipoproteins (25), apoA-11-containinglipoproteins, and apoA-I-containing lipoproteins that do not have any apoA-11. Nonadsorbed proteins were washed with 0.01 M Tris-HC1 buffer, pH 7.4, containing 0.15 M NaCl, 1 mM EDTA, and 0.05% sodium azide. Lipoproteins bound to dextran sulfate-cellulose and the immunosorbents were separately eluted with 3 M NaSCN in 0.02 M sodium phosphate, pH 7.0, and immediately desalted with Sephadex G-25 (Pharmacia LKB Biotechnology Inc.). In A. S. (but not in his relatives) the lipoproteins eluted from the anti-A-I1 immunosorbent contained twice as much apoA-I1 as A-I. These lipoproteins were adsorbed further with the anti-A-I immunosorbent to separate particles that contained both apoA-I1 and A-I from those that contained apoA-I1 but no apoA-I. Nonadsorbed proteins and adsorbed lipoproteins were concentrated by Micro-Confilt concentrator (Biomolecular Dynamics, Beaverton, OR) for further studies. All isolation and concentration processes were performed a t 4 "C.
Lipid and Protein Analyses-Total and unesterified cholesterol, phospholipid, and triglyceride in plasma and lipoprotein fractions were measured by enzymatic methods (26). The difference between cholesterol and free cholesterol mass was considered as cholesterol ester. ApoA-I, A-11, B, and D, and lecithin-cholesterol acyltransferase mass were quantitated by specific immunoassays (27)(28)(29)(30)(31)(32). The protein composition of the isolated HDL particles was examined by sodium dodecyl sulfate (SDS) 7-20% gradient polyacrylamide gel electropho-resis (PAGE) according to the method of Laemmli (32). HDL particle sizes were determined by nondenaturing gradient PAGE using precast 4-305; gels (Pharmacia) as described (33). Proteins in SDS and nondenaturing gels were visualized with 0.1% Coomassie Blue R-250 and G-250, respectively. Gels were scanned with a laser densitometer and integrated with the LKB 2400 Gelscan software. Chemical cross-linking studies were performed with dimethylsuherimidate according to the method of Swaney and O'Brien (34). The cross-linked products were delipidated and analyzed with SDS-PAGE as described (3.5).

RESULTS
DNA Analysis-Genomic DNA from the proband and a normolipidemic control were first examined for major rearrangements by Southern blot analysis.
DNA fragments (approximately 2.5-14 kilobases in length) generated by digestion with the restriction enzymes BgII, PstI, HindIII, SstI, XmnI, and XbaI did not appear to be different from normal. MspI digests, however, revealed that the proband had both the fragment of expected length (1.08 kilobases) and another that was shorter by approximately 50 bp (Fig. 1). This shorter fragment was not present in either the parents or the brother of the proband.
PCR-amplified DNA fragments encompassing the apoA-I gene of the proband were next subjected to gel electrophoretic analysis. Fragment 1 included 199 bp of 5' upstream sequences, exon 1, and the 5'-17 bp of intron 1. Fragment 2 contained exons 2 and 3 and the intron-exon junctions. Fragment 3 contained exon 4 and the intron 3-exon 4 junction. Two distinct fragments, differing in size by 40-50 bp, were observed when the proband's DNA was used as a template for the oligodeoxynucleotide primers flanking exon 4 (Fig. 1). Only one fragment, corresponding in length to the wild-type allele, was obtained when DNA samples from other members of the nuclear family and several other normolipidemic individuals were used as templates in the amplification reaction. These results are consistent with those of Southern blot analysis after digestion with MspI and furthermore show that the proband is heterozygous for a deletion in exon 4.  (Table I) and electrophoresed on 1% agarose.
All amplified DNA fragments were then purified by electrophoresis on agarose gels and sequenced directly by the dideoxy chain termination method as described in detail under "Materials and Methods." Sequencing of both strands of the shortened exon 4 DNA fragment of the proband revealed the existence of a 45-bp deletion in this exon extending (5' + 3') from either nucleotide 1572 or 1574 (Fig. 2). The discrepancy is caused by the presence of the dinucleotide GC a t both ends of the deletion. In either case this deletion would be expected to result in the synthesis of a protein missing 15 amino acid residues (Gluln6 to Arg,,,) (Fig. 3). Exon 4 of the apoA-I gene contains six homologous tandem repeats (66 bp in length) that code for six 22-amino acid segments (3). The 45-bp deletion described above occurred within repeat number three.

5'
Normal rf Autoradiographs of sequencing gels of the normal and deleted apoA-I alleles. Deleted bases are indicated with a bracket on the autoradiograph and a box on the sequence. Sequencing both strands was primed with primers 7 or 8 ( Table I). The sequence shown is that of the sense strand.
.. The sequence of the normal length exon 4 PCR fragment as well as of the other two fragments of the proband's ApoA-I gene was identical to that of a normal control and to the sequence published previously (3,5). Since neither parent is a carrier of the mutant allele, the apoA-I deletion represents a de nouo mutation. The inheritance of alleles of several polymorphic markers in the family was investigated to test this hypothesis. These included hypervariable regions 3' at the apoB (21,36) and the a-globin gene (37) loci and YNH24 (38) in addition to six dimorphic restriction fragment length polymorphisms at the apoA-I/C-III/A-IV locus (XmnI, MspI (promoter), MspI(intron

XbaI) .
The paternity index (X) and inclusion probability (W) for the proband's father were determined (39) from genotypes at the three hypervariable regions ( Table 11). The calculated cumulative inclusion probability of 0.999 strongly favors the hypothesis of a de nouo mutation.
Lipid and Apolipoprotein Profiles-The plasma cholesterol and triglyceride levels of proband A. S. were 209 and 411 mg/ dl, respectively, at the time this HDL characterization study was performed. His mother (J. S.), who is mildly diabetic and is on insulin treatment, had elevated plasma cholesterol and triglyceride, but the lipid profiles of his father (H. S.) and brother (B. S.) were normal (Table 111). In A. S. and J. S., most of the triglyceride was associated with VLDL although their LDL lipid composition also revealed 7-10% more triglyceride than that of H. S. and B. S. The HDL cholesterol of A. S. was about 10% and below the fifth percentile of that reported for the Lipid Research Clinic Prevalence Study norms (40). His HDL triglyceride and phospholipid, although also low, were disproportionately high for the amount of cholesterol in that fraction. The HDL lipid levels of his parents and brother were normal.
The plasma apoA-I and A-I1 concentrations of A. S. were about 12 and 36% of controls, respectively (Table IV). Thus his plasma A-II/A-I ratio was considerably higher than that in population controls. His apoD and lecithin-cholesterol acyltransferase, two other proteins normally found in HDL,

Paternity index (A) and inclusion probability (W) values
Genotypes a t three highly polymorphic DNA markers were used to calculate the paternity index and the inclusion probability according to standard methods (39,54)  were also reduced. In contrast, the levels of lecithin-cholesterol acyltransferase, apoA-I, A-11, and D were between normal and high normal in the parents and brother. The plasma apoB values of A. S. and his parents were high, above the 95th percentile of control values.
Characterization of HDL Particles-Three populations of HDL particles were found in the plasma of the proband one contained both apoA-I and A-11, Lp(A-I w A-11); one contained apoA-I but no A-11, Lp(A-I w/o A-11); and the third contained apoA-I1 but no A-I, Lp(A-11). In the relatives, Lp(A-I w A-11) and Lp(A-I w/o A-11) were also present, but there was no evidence for the presence of any significant amounts of Lp(A-11). When the HDL particles were examined by SDS-PAGE, one protein band corresponding to the position of purified apoA-I was observed in the HDL particles of the relatives. However, two protein bands at positions around purified apoA-I were observed in the proband's Lp(A-I w A-11) and Lp(A-I w/o A-11) (Fig. 4). The calculated molecular weights of these two protein bands differed by about 6%. Consistent with this observation, the molecular weight of mutant A-I would be predicted to be about 6% lower than  (41,42). Consistent with our earlier observations protein bands with molecular weights similar to the apoCs were seen in the HDL particles of the relatives and were more abundant in Lp(A-I w A-11). However, these protein bands were not detected in the HDL particles of the proband under similar experimental conditions. The lipid and protein compositions of the HDL particles are shown in Table  V. Compared with the HDL of his relatives, the HDL particles of the proband were enriched in triglyceride and depleted of cholesterol ester. Furthermore, the Lp(A-I w A-11) and Lp(A-I w/o A-11) of the proband had relatively lower lipid contents (37 and 34%) than those of his parents and brother (40-52%), suggesting that they were denser HDL particles. His Lp(A-11), however, contained about equal amounts of lipid and protein. The amount of lipid in Lp(A-11) represented nearly half of the total HDL-associated lipids.
The particle size profile of HDL was studied by nondenaturing PAGE. As shown in Fig. 5, the two populations of apoA-I-containing particles in the proband and his family were heterogeneous in size. However, the proband's Lp(A-I w A-11) and Lp(A-I w/o A-11) were significantly enriched with small (7-8-nm hydrated Stokes diameter) and very small (< 7-nm) particles. In contrast, the proband's Lp(A-11) contained a major size species with Stokes diameter of about 8.0 nm  (Fig. 6). Cross-linking of Lp(A-11) with dimethylsuberimidate followed by delipidation and SDS-PAGE showed that this major size species contained proteins with a total molecular weight equivalent to four molecules of apoA-I1 (Fig. 6). (One molecule of apoA-I1 is defined as two identical peptides of 77 amino acids.) Distribution of ApoA-I, A-11, D, and Lecithin-cholesterol Acyltransferase in Plasma-The distribution of these proteins in the various HDL particles, Lp(B) (materials bound to dextran sulfate cellulose), and lipoprotein-deficient plasma (plasma fraction after sequential removal of all apoB, A-11, and A-I by dextran sulfate-cellulose, anti-A-11, and anti-A-I immunosorbents) is shown in Table VI. In the proband, 60% of plasma apoA-I and A-I1 existed in separate HDL particles,  i.e. in Lp(A-I w/o A-11) and Lp(A-11), respectively. This was not the case in his relatives in whom the majority (between 60 and 77%) of plasma apoA-I was located in Lp(A-I w A-11). The molar A-I/A-I1 ratio in the proband's Lp(A-I w A-11) was 0.96. This ratio was lower than the 1.45-1.84 ratio observed in the relatives' Lp(A-I w A-11), and the 1.89 f 0.29 (mean f S.D.) ratio in the Lp(A-I w A-11) of a normolipidemic population (14, 43, 44). The relatively lower A-I/A-I1 ratio in A. S.'s Lp(A-I w A-11) was also evident in the SDS gel (Fig. 4).
In the relatives, no more than 1% of apoA-I and A-I1 was recovered in Lp(B). However, 3% of apoA-I and 8% of apoA-I1 in the plasma of the proband were co-isolated with Lp(B).
ApoD and lecithin-cholesterol acyltransferase were detected in all HDL particles. The distribution of apoD among the HDL particles roughly paralleled the distribution of apoA-I and A-I1 mass. In contrast, HDL-associated lecithin-cholesterol acyltransferase was preferentially located in Lp(A-I w/ o A-11) in all the subjects. However, although the apoD and lecithin-cholesterol acyltransferase in the HDL particles represent a mean of 84 and 80% of total plasma apoD and lecithin-cholesterol acyltransferase, respectively, in the relatives, only 48% of the plasma apoD and 45% of lecithincholesterol acyltransferase in the proband were HDL-associated. Non-HDL apoD was found predominantly in the Lp(B) fraction whereas 84-97% of non-HDL lecithin-cholesterol acyltransferase was located in the lipoprotein-deficient plasma fraction.

DISCUSSION
An individual with highly diminished plasma apoA-I and HDL cholesterol levels (hypoalphalipoproteinemia) was shown by both Southern blot analysis and sequencing of PCRamplified DNA fragments to be heterozygous for a 45-bp deletion in exon 4 of the apoA-I gene. The parents and brother of the proband, none of whom had reduced apoA-I and HDL levels, were shown not to carry the mutant allele. This, together with results of analysis of the inheritance of several polymorphic markers in this family, strongly suggest that this represents a de m u 0 mutation. It is predicted that the protein product of this mutant allele would be missing 15 amino acids (7 of which are charged) from the third 22-residue repeat. Consistent with this, two protein bands in the vicinity of purified apoA-I with molecular weight differing by 6% were observed in an SDS gel. Six homologous 22-residue tandem repeats comprise most of the carboxyl-terminal half of the mature apoA-I polypeptide. These repeats, which also exist in other apolipoproteins (apoA-11, A-IV, C-11, and C-111), represent segments of amphipathic Lu-helices that are important for lipid binding and maintenance of the structural integrity of HDL particles (45).
The presence of this mutant apoA-I missing 15 amino acids had a profound effect on the quantity and characteristics of -- plasma HDL in the proband. In normal individuals, approximately two-thirds of plasma apoA-I and nearly all apoA-I1 are found in some HDL particles. However, in the proband, 60% of plasma apoA-I and A-I1 were found in separate HDL particles. The occurrence of substantial amounts of plasma HDL particles containing apoA-I1 as the major protein component is very unusual. Such quantities of Lp(A-11) have only been reported in Tangier plasma (46,47) and in another HDL-deficient subject we have studied (48). Chemical crosslinking studies showed that most of the Lp(A-11) particles contained proteins with total molecular weight equivalent to four molecules of apoA-11. Interestingly, in vitro complexing of apoA-I1 with HDL lipids also resulted in HDL with four molecules of apoA-II/particle (49).

HS
All of the HDL particles in the proband were rich in triglyceride and poor in cholesterol ester with the proportion of core to surface lipid being normal. This suggests that most of the HDL particles were probably spherical. Despite the relatively low content of cholesterol ester in HDL, the free cholesterol/cholesterol ester ratio in the proband's plasma was normal, suggesting that cholesterol esterification by lecithin-cholesterol acyltransferase was not affected significantly by the low level of apoA-I or by the presence of mutant apoA-I.
We have studied the electrophoretic mobility of the proband's HDL particles on an agarose gel. Because of the small amounts of materials available and the low lipid/protein ratio of some of the particles, we were only able to see, with certainty, a-migrating Lp(A-I w A-11) and Lp(A-11) when the gel was stained for lipid with Sudan black. Staining the gel for protein with Coomassie G-250 revealed both cy-and prep-materials in Lp(A-I w/o A-11).
Since the proband is heterozygous for the mutant allele it is intriguing that his plasma apoA-I and HDL cholesterol levels were less than 15% of population means. A likely explanation of this observation may lie in the multimeric nature of HDL particles and a possible enhanced catabolic rate of mutant apoA-I. We have shown that in normal subjects most Lp(A-I w A-11) contained two molecules of apoA-I whereas Lp(A-I w/o A-11) contained either two, three, or four molecules of apoA-I/particle (35). If mutant apoA-I is catabolized faster than normal A-I, the large fraction of HDL particles which is expected to contain both mutant and normal apoA-I would be catabolized more rapidly than normal HDL. Hypercatabolism has been observed in Tangier patients (50) and carriers of A-Ihlilano (51) and A-IIowa (52). The catabolic rate of apoA-Iseattle remains to be determined.
Two previously described mutations of apoA-I were shown to be associated with decreased plasma apoA-I and HDL cholesterol levels, APOA-IM~~,,,, (Arg,,, + Cys) and Pro165 + Arg. Individuals heterozygous for these alleles had approximately half the normal levels of plasma apoA-I. The trait was inherited in an autosomal co-dominant fashion (17)(18)(19). ApoA-Iseattl, in the heterozygous state has a much more profound effect on both plasma apoA-I and HDL levels than  " _ 10 -6 -46 11 12 32

LP(B)
-" _ plasma apOA-IMilano, presumably because of the greater alteration in primary structure (loss of 15 amino acids) and the suspected higher rate of catabolism. During the past 5 years the proband's plasma triglyceride levels increased to above the 95th percentile of a control population. Whether this is a consequence of the apoA-I deletion mutation remains to be determined.
Although an elevated triglyceride level was observed to be more common among individuals heterozygous for the apOA-IM,lano, no consistent pattern of association was detected in other families with other variants (19).