Variation in lipoprotein(a) concentration associated with different apolipoprotein(a) alleles.

Plasma lipoprotein(a) (Lp(a)) concentrations vary considerably between individuals. To examine the variation for products of the same and different apolipoprotein(a) (apo(a)) alleles, conditions were established whereby phenotyping immunoblots could be used to estimate the concentration of Lp(a) associated with the constituent apo(a) isoforms. In these studies 28 distinct isoforms were identified, each differing by a single kringle IV unit. Tracking the isoforms through 10 families showed that there could be up to 200-fold difference in the Lp(a) concentration associated with the same-sized isoform produced from different alleles. In contrast there was typically < 2.5-fold variation in the Lp(a) concentration associated with the same allele. However, there were four occasions where the concentration associated with a particular allele was reduced below the typical range from one generation to the next. A nonlinear, inverse trend with isoform size was apparently superimposed upon the other factors that determine Lp(a) concentration. Inheritance of familial hypercholesterolemia or familial-defective apoB100 had little consistent effect upon Lp(a) concentration. In both the families and in other unrelated individuals the distribution of isoforms and their associated concentrations provided evidence for the presence of at least two and possibly more subpopulations of apo(a) alleles with different sizes and expression.

Relationship between apolipoprotein(a) phenotype, lipoprotein(a) concentration in plasma, and low density lipoprotein receptor function in a large kindred with familial hypercholesterolemia due to the pro664----leu mutation in the LDL receptor gene.

In a large kindred of 66 individuals, 22 were identified as heterozygous and 3 as homozygous for a mutation (pro664----leu) in the LDL-receptor gene that gives rise to familial hypercholesterolaemia (FH). All the heterozygotes had a raised level of plasma total cholesterol and low density lipoprotein cholesterol, but were remarkably free from premature coronary disease. Determination of apolipoprotein(a) (apo(a)) phenotype and lipoprotein(a) (Lp(a)) concentration in plasma revealed that in many instances, involving individuals with various apo(a) phenotypes, there was no difference in plasma Lp(a) concentration between an FH heterozygote and an unaffected sibling with the same apo(a) phenotype. No significant difference in Lp(a) concentration was observed between groups of FH and non-FH of the same apo(a) phenotype, although in each case the mean value for the FH group was greater than that for the non-FH group. There was also evidence for an inherited trait that markedly increased Lp(a) concentration, which did not segregate with apo(a) phenotype or the defective LDL-receptor allele. The data provide no evidence for a strong multiplicative interaction between the gene loci for apo(a) and the LDL receptor.

Characterization of a novel cellular defect in patients with phenotypic homozygous familial hypercholesterolemia.

Familial hypercholesterolemia (FH) is characterized by a raised concentration of LDL in plasma that results in a significantly increased risk of premature atherosclerosis. In FH, impaired removal of LDL from the circulation results from inherited mutations in the LDL receptor gene or, more rarely, in the gene for apo B, the ligand for the LDL receptor. We have identified two unrelated clinically homozygous FH patients whose cells exhibit no measurable degradation of LDL in culture. Extensive analysis of DNA and mRNA revealed no defect in the LDL receptor, and alleles of the LDL receptor or apo B genes do not cosegregate with hypercholesterolemia in these families. FACS((R)) analysis of binding and uptake of fluorescent LDL or anti-LDL receptor antibodies showed that LDL receptors are on the cell surface and bind LDL normally, but fail to be internalized, suggesting that some component of endocytosis through clathrin-coated pits is defective. Internalization of the transferrin receptor occurs normally, suggesting that the defective gene product may interact specifically with the LDL receptor internalization signal. Identification of the defective gene will aid genetic diagnosis of other hypercholesterolemic patients and elucidate the mechanism by which LDL receptors are internalized.

Deficiency of PPARalpha disturbs the response of lipogenic flux and of lipogenic and cholesterogenic gene expression to dietary cholesterol in mouse white adipose tissue.

PPARalpha-deficiency in mice fed a high-carbohydrate, low-cholesterol diet was associated with a decreased weight of epididymal adipose tissue and an increased concentration of adipose tissue cholesterol. Consumption of a high (2% w/w) cholesterol diet resulted in a further increase in the concentration of cholesterol and a further decrease in epididymal fat pad weight in PPARalpha-null mice, but had no effect in the wild-type. These reductions in fat pad weight were associated with an increase in hepatic triacylglycerol content, indicating that both PPARalpha-deficiency and cholesterol altered the distribution of triacylglycerol in the body. Adipose tissue de novo lipogenesis was increased in PPARalpha-null mice and was further enhanced when they were fed a cholesterol-rich diet; no such effect was observed in the wild-type mice. The increased lipogenesis in the chow-fed PPARalpha-null mice was accompanied paradoxically by lower mRNA expression of SREBP-1c and its target genes, acetyl-CoA carboxylase and fatty acid synthase. Consumption of a high-cholesterol diet increased the mRNA expression of these genes in the PPARalpha-deficient mice but not in the wild-type. De novo cholesterol synthesis was not detectable in the adipose tissue of either genotype despite a relatively high expression of the mRNA's encoding SREBP-2 and 3-hydroxy-3-methylglutaryl Coenzyme A reductase. The mRNA expression of these genes and of the LDL-receptor in adipose tissue of the PPARalpha-deficient mice was lower than that of the wild-type and was not downregulated by cholesterol feeding. The results suggest that PPARalpha plays a role in adipose tissue cholesterol and triacylglycerol homeostasis and prevents cholesterol-mediated changes in de novo lipogenesis.

Protein kinases and their substrates in brown adipose tissue from newborn rats.

The 10000 X g supernatant fraction of brown fat from newborn rats catalyzed the cyclic AMP-dependent phosphorylation of both histone and a preparation of proteins from the same subcellular fraction (endogenous proteins). The apparent affinity for ATP was lower for the phosphorylation of the endogenous proteins than for the phosphorylation of histone. In order to discover whether the phosphorylation of histone and the endogenous proteins were catalyzed by different enzymes, the 100000 X g supernatant was fractionated by ion-exchange and adsorption chromatography. Three different cyclic AMP-dependent protein kinases and one cyclic AMP-independent protein kinase were separated and partially purified. Each of these enzymes catalyzed the phosphorylation of both substrates, and the difference in apparent Km for ATP remained. Neither affinity chromatography on histone-Sepharose, nor electrophoresis on polyacrylamide gels resulted in the separation of the phosphorylation of histone from that of the endogenous proteins of any of the partially purified kinases. Moreover, experiments in which the phosphorylated substrates were separated by differential precipitation with trichloroacetic acid showed that the endogenous proteins competitively inhibited the phosphorylation of lysine-rich histone. It is concluded that each of the partially purified kinase preparations contains protein kinase, which catalyzes the phosphorylation of both substrates. The difference in apparent Km for ATP was found to be due to the presence in the endogenous protein preparation of a low molecular weight compound which competes with ATP. This was not ATP nor the modulator protein. The ratio of the phosphorylation of endogenous proteins to that of histone was much higher for the cyclic AMP-independent kinase preparation than for the other enzymes. Electrophoresis of the endogenous substrates in the presence of sodium dodecyl sulphate showed that the enzyme phosphorylated a greater number of proteins than did the cyclic AMP-dependent kinases. The phosphorylation of endogenous proteins relative to that of histone was significantly lower for one of the cyclic AMP-dependent kinases than for the other two. This difference was not reflected in a different pattern of phosphorylation of the individual proteins of the endogenous mixture.

The separation, properties and possible subunit composition of adenosine 3',5'-monophosphate-dependent protein kinases in brown adipose tissue.

Two 8.5-S protein kinases (ATP : protein phosphotransferase EC 2.7.1.37) and one 6.6-S protein kinase were purified 500--1000-fold from the acid-soluble fraction of brown adipose tissue. The catalytic properties of the kinases were similar. Each kinase was activated by cyclic AMP and had two components of cyclic AMP binding. In the presence of 200 nM cyclic AMP, undissociated kinase activity sedimented at 7.7 or 5.5 S. Free catalytic activity (3.2 S) could be detected but was unstable. Free regulatory units could not be detected. The 8.5-S protein kinase was dissociated by freezing and thawing to a 7.7-S variety with loss of the higher affinity component of binding. The 7.7-S kinase was sedimented through linear gradients of sucrose containing different concentrations of cyclic AMP. At each concentration, kinase activity lost from the holoenzyme peak (% of original) was identical with the amount of cyclic AMP bound at equilibrium (% oof maximum). Similar experiments on the 8.5-S kinase showed that the binding component with higher affinity was not associated with the release of catalytic activity. The results were consistent with the propostal that the kinases isolated contained one more cyclic AMP binding subunit than catalytic subunit (3 : 2 for 8.5 S and 2 : 1 for 6.6 S) and that this extra subunit was released to give an equal number of subunits of each type before catalytic activity was liberated.

Protein kinases in brown adipose tissue of developing rats. I. GTP as nucleotide substrate for histone phosphorylation.

100 000 times g soluble extracts from interscapular brown adipose tissue catalyzed the transfer of the terminal phosphoryl group from GTP to histone. Maximal velocity was achieved only with both cyclic AMP and ATP present. The cyclic AMP dose-response curve was the same as for the ATP-utilizing enzyme, with maximum stimulation at 0.5 muM. ATP (1--100muM) increased the rate of histone phosphorylation with GTP as the radioactive substrate. Higher concentrations had a dilution effect similar to that of GTP on the ATP-utilizing enzyme. Similar effects were observed with ADP and AMP. The apparent Km values for histone were the same with both GTP and ATP as nucleotide substrates. The effects of pH, purified beef muscle kinase inhibitor and of NaCl were also the same. Maximum velocities of histone phosphorylation from ATP and those from GTP were almost the same in brown fat of all age groups testes, Separated on histone-Sepharose, the GTP-utilizing activity was absolutely dependent on the re-addition of the ATP-utilizing enzyme (a linear relationship with a slope of approx. 0.95). An extremely active nucleotide phosphotransferase activity was found in the same subcellular fraction. The rate of equilibration of the gamma-32-P between GTP and ATP could account for all the histone phosphorylation with [gamma-32-P] GTP. It is concluded that, in spite of the presence of nucleotide phosphotransferase and ATP-protein kinase activities, a direct transfer from GTP to a protein substrate cannot be excluded. Also, histone may not be the natural protein acceptor for GTP-linked phosphorylation.

Protein kinases in brown adipose tissue of developing rats. II. two soluble kinase activities and their affinities for nucleotide and protein substrates.

A heat-stable, soluble component of brown adipose tissue from newborn rats was found to be readily phosphorylated by protein kinase of the same subcellular fraction. The concentration of this component in brown fat decreased with the age of the animals. A boiled crude microsomal preparation from rat liver was also phosphorylated by brown fat protein kinase. The GTP-linked phosphorylation of the endogenous heat-stable protein was not stimulated by ATP (in contrast to phosphorylation of histone). The maximum velocity of phosphorylation achieved with GTP was about 2.5 times higher than that with ATP as nucleotide substrate. This difference was not due to ATPase activity in the assay. With histone as the protein acceptor both activities were the same. The affinity of protein kinase(s) for ATP was lower with the endogenous heat-stable brown-fat protein and with boiled microsomes (Km of 0.21 mM and 0.17 mM, respectively) than with histone (Km of 0.05 M). No detectable ATPase activity was present in either acceptor protein. It is concluded that the 100 000 times g supernatant fraction from brown fat of infant rats contains two protein kinase activities. One preferentially uses ATP and histone as substrates and the other uses endogenous heat-stable soluble proteins and either ATP or GTP.

Catabolism of low-density lipoprotein by fibroblasts cultured in medium supplemented with saturated or unsaturated free fatty acids.

The increase in the degree of unsaturation of phospholipid fatty acids that occurred when fibroblasts were grown in medium supplemented with linoleic acid was associated with an increase in the degradation of LDL by both receptor-mediated and receptor-independent pathways. There was not a corresponding increase in the number of surface receptors for LDL.

Effect of insulin and prednisolone on cyclic nucleotides and phosphoenolpyruvate carboxykinase activity in brown fat and liver of developing rats.

The concentrations of cyclic AMP and cyclic GMP in brown fat and liver of both suckling and adult rats at fixed times after injection of insulin (2.5 U/100 g body weight) or prednisolone (2.5 mg/100 g body weight) were compared with the activity of phosphoenolpyruvate carboxykinase assayed 24 h after the injections. A stimulus that produced an increase in cyclic AMP content also produced an increase in the enzyme activity. If the content of cyclic GMP was also increased there was no rise in phosphoenolpyruvate carboxykinase activity. A rise in the content of cyclic GMP alone was associated with a reduction in the activity of the enzyme. These preliminary results indicate that cyclic AMP could be involved in the induction of phosphoenolpyruvate carboxykinase and that cyclic GMP may somehow be related to its repression. The known differences in the response of phosphoenolpyruvate carboxykinase activity to insulin and prednisolone in different tissues and at different stages of ontogenic development may thus be linked to differences in the responsiveness of enzymes concerned with the metabolism of cyclic nucleotides.

Abnormal structure and co-operative binding of low-density lipoprotein receptors containing the Glu-80-->Lys mutation.

The properties of low-density lipoprotein (LDL) receptors containing a Glu to Lys substitution at position 80 have been studied in fibroblasts from a homozygous familial hypercholesterolaemic subject (MB) and in monkey COS cells transfected with the mutant cDNA. Receptors containing the Glu-80-->Lys mutation were processed more slowly than the normal protein and only approx. 50% reached the surface as the mature form. Both cell types exhibited a normal concentration binding curve for beta-very-low-density lipoproteins (beta-VLDL) but an atypical, sigmoidal curve for LDL. The mature mutant receptor protein migrated abnormally slowly on SDS-PAGE under non-reducing conditions but normally under reducing conditions or after treatment with neuraminidase. It also showed an unusual ability to form dimers that were stable in detergents. Transfected normal and mutant receptors were apparently cleaved on the surface of the cells to give a product lacking the NH2-terminal portion of the protein, which was resistant to further proteolytic digestion. The results suggest that the Glu-80-->Lys substitution produces a change in the conformation of the protein, stabilized by polysaccharide chains, which results in a strong self-association of receptor molecules that affects their ability to bind LDL.

Determinants of variable response to statin treatment in patients with refractory familial hypercholesterolemia.

Interindividual variability in low density lipoprotein (LDL) cholesterol (LDL-C) response during treatment with statins is well documented but poorly understood. To investigate potential metabolic and genetic determinants of statin responsiveness, 19 patients with refractory heterozygous familial hypercholesterolemia were sequentially treated with placebo, atorvastatin (10 mg/d), bile acid sequestrant, and the 2 combined, each for 4 weeks. Levels of LDL-C, mevalonic acid (MVA), 7-alpha-OH-4-cholesten-3-one, and leukocyte LDL receptor and hydroxymethylglutaryl coenzyme A reductase mRNA were determined after each treatment period. Atorvastatin (10 mg/d) reduced LDL-C by an overall mean of 32.5%. Above-average responders (LDL-C -39.5%) had higher basal MVA levels (34.4+/-6.1 micromol/L) than did below-average responders (LDL-C -23.6%, P<0.02; basal MVA 26.3+/-6.1 micromol/L, P<0.01). Fewer good responders compared with the poor responders had an apolipoprotein E4 allele (3 of 11 versus 6 of 8, respectively; P<0.05). There were no baseline differences between them in 7-alpha-OH-4-cholesten-3-one, hydroxymethylglutaryl coenzyme A reductase mRNA, or LDL receptor mRNA, but the latter increased in the good responders on combination therapy (P<0.05). Severe mutations were not more common in poor than in good responders. We conclude that poor responders to statins have a low basal rate of cholesterol synthesis that may be secondary to a genetically determined increase in cholesterol absorption, possibly mediated by apolipoprotein E4. If so, statin responsiveness could be enhanced by reducing dietary cholesterol intake or inhibiting absorption.

Variation at position 162 of peroxisome proliferator-activated receptor alpha does not influence the effect of fibrates on cholesterol or triacylglycerol concentrations in hyperlipidaemic subjects.

The fibrate group of drugs are widely used to lower plasma lipid concentrations, especially in diabetic subjects. They exert their effects by activating peroxisome proliferator-activated receptor alpha (PPARalpha), which regulates the transcription of a number of genes involved in hepatic lipid metabolism. The discovery of polymorphisms in the PPARalpha gene raises the possibility that different variants could be associated with different responses to fibrate therapy. We have examined this in a random sample of 96 lipid clinic subjects who showed a wide range of response to fibrates. Of the known polymorphisms in PPARalpha, the only difference detected in this sample was the Leu/Val change at position 162. However, this change did not influence baseline plasma cholesterol or triacylglycerol concentrations, and was not associated with any difference in the effectiveness of fibrate treatment. Thus, there is no evidence that variation in the PPARalpha gene at position 162 is responsible for the differential response to fibrates in non-diabetic hyperlipidaemic subjects.

Non-saturable degradation of LDL by monocyte-derived macrophages leads to a reduction in HMG-CoA reductase activity with little synthesis of cholesteryl esters.

In monocyte-derived macrophages from both normal and familial hypercholesterolaemic (FH) subjects, degradation of low density lipoprotein (LDL) through non-saturable pathways produced the same fall in 3-hydroxy-3-methylglutaryl-CoA reductase activity as receptor-mediated degradation of acetylated LDL, yet did not lead to as great an increase in incorporation of [14C]oleate into cholesteryl esters. Studies using FH cells showed that the simultaneous addition of LDL did not reduce oleate incorporation resulting from degradation of acetylated LDL, and that there was a similar relationship for both lipoproteins between the increase in net oleate incorporation and the increase in the cholesteryl ester content of the cells. FH cells maintained in serum-free medium accumulated more free cholesterol than cells in lipoprotein-deficient serum when incubated with LDL but not when incubated with acetylated LDL. The results suggest that the cholesterol released from non-saturable degradation of LDL is more easily removed from the cells by acceptors in the medium than cholesterol released from receptor-mediated uptake of acetylated LDL, and is not readily available for esterification.

Binding and degradation of heavy and light subfractions of low density lipoprotein by cultured fibroblasts and macrophages.

The heavy and light subfractions of low density lipoprotein (LDL) were bound to the same extent and with the same affinity by the LDL receptors of cultured human fibroblasts, both when assayed at 4 degrees C and when assayed at 37 degrees C. They were also degraded similarly by the low affinity, LDL-receptor-mediated pathway exhibited by normal human monocyte-derived macrophages maintained in medium containing whole serum. Neither of the subfractions was taken up by the 'scavenger' pathway in mouse peritoneal or human monocyte-derived macrophages. Assuming that the LDL particles were not altered during isolation, the results provide no evidence to suggest that the higher fractional catabolic rate of light LDL observed in vivo can be explained by any preferential catabolism through LDL-receptor-mediated pathways.

The effect of nicotinic acid and acipimox on lipoprotein(a) concentration and turnover.

This study examines the effect of nicotinic acid (1 g t.d.s.) on serum Lp(a) concentration in a group of patients with type II hyperlipidaemia selected on the basis of a plasma Lp(a) concentration greater than 30 mg/dl. Reductions in total cholesterol, triglyceride, LDL-cholesterol and Lp(a) were 16.3%, 25.5%, 23.7% and 36.4%, respectively, with an increase in HDL cholesterol of 37.3%. The reduction in Lp(a) concentration did not correlate with any other lipoprotein changes. In order to establish the mechanism of the fall in Lp(a) concentration, in vivo turnover of autologous Lp(a) was studied in three subjects before and whilst taking nicotinic acid. The fractional catabolic rate in Lp(a) was unaltered in the subjects on therapy, indicating that nicotinic acid did not increase catabolism of Lp(a) but decreased the synthetic rate. Since nicotinic acid was poorly tolerated we examined the effect of acipimox, an analogue of nicotinic acid on lipoproteins using a placebo controlled double-blind crossover design in a group of hyperlipidaemic patients again selected with plasma Lp(a) concentration greater than 30 mg/dl. Acipimox was better tolerated than nicotinic acid but the percentage changes in lipoprotein concentrations were smaller.

Catabolism of lipoprotein(a) in familial hypercholesterolaemic subjects.

The in vivo turnover of autologous lipoprotein(a) (Lp(a)) was studied in four heterozygous familial hypercholesterolaemic (FH) subjects and four subjects who were hyperlipidaemic but not FH. Each of the FH subjects exhibited a much lower fractional catabolic rate (FCR) for LDL than each of the non-FH subjects. Lp(a) was purified by sequential density gradient centrifugations and was radio-iodinated. The labelled Lp(a) ran as a single band on electrophoresis in gradient polyacrylamide gels. Less than 5% of the label was in lipid, with about 40% of the remainder on apolipoprotein B (apo B) and 60% on apo(a). Labelled and unlabelled Lp(a) competed equally poorly with LDL for binding to LDL receptors on cultured fibroblasts. The FCR of Lp(a), calculated from the decay of the specific radioactivity of the Lp(a) isolated from the daily blood samples, was the same in FH subjects as in non-FH subjects. There was no consistent relationship between Lp(a) FCR and the plasma Lp(a) concentration or between FCR and the Lp(a) phenotype, at least within this sample of subjects. There was a strong association between Lp(a) concentration and production rate, with values for non-FH and FH subjects falling on the same line. The rate of decline of radioactivity in whole plasma was consistently slower than the fall in specific radioactivity of the isolated Lp(a). This difference was more marked in FH subjects than in non-FH subjects and resulted from the accumulation of radioactivity derived from the injected Lp(a) at a lower density than Lp(a), in the fractions containing LDL. The amount of radioactivity in this fraction increased for the first few days after injection and then fell, the fall being more rapid in non-FH than in FH subjects. These results provide no evidence for the involvement of LDL receptors in the catabolism of Lp(a) itself but suggest that they could be responsible for some of the clearance of the lipid and apo B components after removal of apo(a) in the circulation.

Lipoprotein(a) in subjects with familial defective apolipoprotein B100.

The plasma lipoprotein(a) (Lp(a)) concentration and apolipoprotein(a) (apo(a)) phenotype were determined in the members of two families affected with familial defective apo B100 (FDB), resulting from the Arg3500----Gln mutation in apo B that disrupts binding to LDL receptors. Eleven different phenotypic species of apo A were identified, five of which were present in both families. Although there was a general increase in Lp(a) concentration as the size of the predominant apo(a) component decreased, there was considerable variability and in three clear instances the concentration of an inherited phenotypic species was atypically low. In five cases where a direct comparison could be made, the plasma Lp(a) concentration was significantly higher in heterozygous FDB subjects than in their non-FDB siblings or close relatives with the same phenotype. However, in vitro competition studies using purified Lp(a) that had been reduced with dithiothreitol to remove the apo(a) component, indicated that the Lp(a) from FDB heterozygotes contained a smaller proportion of defective particles than their LDL. Lp(a) particles containing normal and binding-defective apo B were present at approximately the same concentration, suggesting that the increase in Lp(a) concentration observed in FDB subjects could not be explained by the inability of the particles containing the defective apo B100 to be cleared through LDL-receptor mediated processes.

Influence of genotype at the low density lipoprotein (LDL) receptor gene locus on the clinical phenotype and response to lipid-lowering drug therapy in heterozygous familial hypercholesterolaemia. The Familial Hypercholesterolaemia Regression Study Group.

The relationship between molecular defect and clinical phenotype has been examined in 42 patients with heterozygous familial hypercholesterolaemia (FH) and premature coronary heart disease. The defined defects included mutations in the low density lipoprotein (LDL)-receptor gene (23/42) or the apolipoprotein B Arg3500Gln mutation (5/42). Mean LDL-cholesterol was higher, both before and during treatment with simvastatin and bile acid sequestrants, in patients predicted as having a 'severe' mutation than in those with a 'mild' mutation (8.72 +/- 2.02 mmol/l vs 6.63 +/- 1.8, P = 0.05 before and 4.51 +/- 0.90 mmol/l vs 3.19 +/- 0.58, P = 0.05 during treatment). Maximum inducible LDL-receptor activity in cultured lymphoblasts was inversely correlated with LDL-cholesterol before (r2 = 0.499, P = 0.002) and during (r2 = 0.478, P = 0.004) treatment in patients with a defined mutation in the LDL-receptor gene, but not in the 14 patients with no detectable molecular defect. LDL-cholesterol concentrations before and during treatment were significantly correlated in patients with a defined LDL-receptor gene mutation (r2 = 0.548, P = 0.0001), but not in those with no detectable genetic defect. All these correlations were weak, however and there were no differences in the response to treatment in terms of either relative reduction or absolute decrease in LDL-cholesterol concentration between patients with different LDL-receptor defects. We conclude that only part of the variable phenotype of heterozygous FH patients is explained by different LDL-receptor defects and that other factors determine the severity of their hypercholesterolaemia and the onset of coronary disease.

Detection and quantitation of apolipoprotein(a) mRNA in human liver and its relationship with plasma lipoprotein(a) concentration.

Northern blotting and hybridisation with specific probes was used to detect and quantitate apolipoprotein(a) (apo(a)) mRNA in total RNA isolated from 25 human liver samples. A total of 14 different transcripts were identified suggesting that there are at least 15 different alleles at the apo(a) locus including a probable null allele. Apo(a) mRNA sizes were linearly correlated with the electrophoretic mobility of plasma apo(a) glycoprotein isoforms, and differed, in many cases, by the equivalent of one Kringle 4 unit. To investigate the relationship between apo(a) mRNA size and its concentration in the liver, and between hepatic apo(a) mRNA concentration and plasma lipoprotein(a) (Lp(a)) levels, apo(a) mRNA was quantified by densitometric scanning of autoradiograms of Northern blots. Overall, there was a significant inverse correlation between apo(a) mRNA size and its concentration in the liver, despite a marked interindividual variability in the relative amounts of similar-sized transcripts. In each heterozygous individual, the difference in concentration between the two mRNA species was determined by the difference in size. However, there was not a significant relationship between hepatic apo(a) mRNA concentration and plasma Lp(a) levels in this group. These findings emphasise the importance of mechanisms other than the rate of transcription of the apo(a) gene in the regulation of Lp(a) synthesis.