Hypertriglyceridaemia and low plasma HDL in a patient with apolipoprotein A‐V deficiency due to a novel mutation in the APOA5 gene

APOA5 encodes a novel apolipoprotein (apo A‐V) which appears to be a modulator of plasma triglyceride (TG). In apoA5 knock out mice plasma TG level increases almost fourfold, whereas in human APOA5 transgenic mice it decreases by 70%. Some SNPs in the APOA5 gene have been associated with variations in plasma TG in humans. In addition, hypertriglyceridaemic (HTG) patients have been identified who carried rare nonsense mutations in the APOA5 gene (Q139X and Q148X), predicted to result in apo A‐V deficiency. In this study we report a 17‐year‐old male with high TG and low high density lipoprotein cholesterol (HDL‐C), who at the age of two had been found to have severe HTG and eruptive xanthomas suggesting a chylomicronaemia syndrome. Plasma postheparin LPL activity, however, was normal and no mutations were found in LPL and APOC2 genes. The sequence of APOA5 gene revealed that the patient was homozygous for a point mutation (c.289 C>T) in exon 4, converting glutamine codon at position 97 into a termination codon (Q97X). Apo A‐V was not detected in patient’s plasma, indicating that he had complete apo A‐V deficiency. The administration of a low‐fat and low‐oligosaccharide diet, either alone or supplemented with ω‐3 fatty acids, started early in life, reduced plasma TG to a great extent but had a negligible effect on plasma HDL‐C. Loss of function mutations of APOA5 gene may be the cause of severe HTG in patients without mutations in LPL and APOC2 genes.


Introduction
Hypertriglyceridaemia (HTG) usually refers to fasting plasma triglyceride (TG) level above the 95th percentile for age and sex; it can be divided into primary and secondary type. Primary HTG is assumed to have a strong genetic basis; however, the molecular defect underlying primary HTG is known in <10% of cases [1].
Hyperlipidaemia type I and hyperlipidaemia type V are lipoprotein disorders characterized by the pathological presence of chylomicrons in the fasting state. Their clinical features include eruptive xanthomas, lipaemia retinalis, hepato-splenomegaly, recurrent abdominal pain and increased risk of pancreatitis [2]. The accumulation of chylomicrons (type I) or chylomicrons and VLDL (Type V) is caused by defects in the lipolytic cascade due to loss of function mutations in the genes encoding lipoprotein lipase (LPL) or its activator apo C-II (APOC2) [2].
Familial HTG (hyperlipoproteinaemia type IV) is characterized by an isolated elevation of VLDL. This disorder has a population prevalence of 5-10% but its molecular basis is still unknown; it is likely to be polygenic and strongly affected by environmental factors and metabolic stress [1].
In familial combined hyperlipidaemia (FCHL) the elevation of plasma TG is due to increased level of VLDL, which is associated with increased level of LDL and hepatic overproduction of apolipoprotein B [3]. The molecular basis underlying FCHL is largely unknown; genetic variants in APOA1 ⁄ C3 ⁄ A4 ⁄ A5 cluster and in the recently identified USF1 gene appear to contribute to the phenotype [4,5].
The increased TG levels found in dysbetalipoproteinaemia (hyperlipoproteinaemia type III) is due to the accumulation of TG-rich lipoprotein remnants (e.g. b-VLDL). Patients with this disorder are typically homozygous for the rare apo E-2 isoform, which shows a defective binding to LDL-receptor when compared with the most common apo E-3 isoform. The phenotypic expression of dysbetalipoproteinaemia, however, requires some additional factors such as obesity, type 2 diabetes or hypothyroidism [3].
The recently identified APOA5 gene, which encodes apolipoprotein A-V (apo A-V), appears to be a key modulator of plasma TG homeostasis in rodents as well as in humans. When the human APOA5 gene was expressed in transgenic mice, the plasma TG level decreased by 70%, whereas when mouse apoa5 gene was inactivated, the plasma TG increased fourfold, compared with control mice [6]. These results suggested an inverse correlation between apo A-V concentration and plasma TG level in mice. Single nucleotide polymorphisms in the APOA5 gene (e.g. 1131T ⁄ C in the 5¢-flanking region and c.56 C ⁄ G in the coding region) have been associated with differences in plasma TG levels in humans [6][7][8][9][10]. Carriers of the rare alleles )1131C or c.56 G have higher plasma TG than carriers of the corresponding common alleles [7][8][9][10]. Interestingly, human studies do not support the inverse correlation between plasma apo A-V and TG concentration that has emerged from studies in mice. In fact, plasma apo A-V level was found to be positively correlated with TG level in subjects from the general population, as well as in patients with HTG and in patients with type 2 diabetes [11][12][13][14][15][16].
Rare loss of function mutations in APOA5 gene have been reported in humans [17]. In 2005 we identified a 9-year-old boy with severe HTG, who was homozygous for a nonsense mutation predicted to result in truncated apo A-V (Q148X) [18]. He was regarded to have complete apo A-V deficiency as no apo A-V was detected in his plasma [17]. Another patient with type V hyperlipidaemia and recurrent episodes of pancreatitis was found to be homozygous for a different apo A-V mutation (Q139X) predicted to cause complete apo A-V deficiency [19]. So it looks as though apo A-V deficiency in humans recapitulates, to some extent, the plasma lipid abnormalities observed in apoa5 knock-out mice, suggesting that the absence of apo A-V causes a substantial disruption of the metabolism of TG-rich lipoproteins.
In the present study we report a 17-year-old patient with primary HTG detected early in life, who was found to be homozygous for a novel nonsense mutation of APOA5 gene. No apo A-V was detectable in his plasma, thus suggesting that he has complete apo A-V deficiency.

Subject
Proband T.L. is a 17-year-old Caucasian male, who at the age of two was found to have eruptive cutaneous xanthomas and severe HTG in the fasting state, suggesting hyperchylomicronaemia syndrome. He never had abdominal pain, hepatosplenomegaly nor pancreatitis and developed normally. Routine laboratory tests, with the exception of plasma lipids, were within the reference range. The proband has been treated with low-fat and low-oligosaccharide diet since infancy; the good compliance to this diet was associated with a substantial reduction of plasma TG level (Table 1) and a progressive disappearance of eruptive xanthomas. At the age of 12 he started treatment with x-3 fatty acids (3 g day )1 ) which, however, was discontinued after 2 years because of poor compliance. Medium Chain Triglycerides (MCT) supplementation or fibrate therapy were never prescribed.
At age 16 the proband was referred to the Lipid Clinic of the University Hospital with the clinical diagnosis of primary severe HTG (type V). Alcohol assumption was denied. He was a nonsmoker and practised regular physical activity (basketball player). On physical examination he was found healthy and lean (BMI 19.5 kg m )2 , waist circumference 71 cm). He had normal blood pressure and heart rate; no skin, tendons, thoracic, abdominal or vascular abnormalities were observed. Tonsils, eyes and fundus oculi examinations were normal. Abdominal ultrasound scan confirmed the absence of visceral abnormalities. Routine laboratory tests for hepato-biliary, renal and thyroid function were normal, as well as blood glucose and blood cell count. ECG was normal.
On a low-fat diet, the level of total cholesterol (T-C), TG and HDL cholesterol (HDL-C) were 4.08, 5.90 and 0.75 mmol L )1 , respectively ( Table 1). The addition of x-3 fatty acids (3 g day )1 ) lowered triglycerides to 2.98 mmol L )1 but had little effect on plasma HDL-C (Table 1).
Proband's parents were apparently unrelated; they were in good health and had no family history of dyslipidaemia or premature coronary heart disease. The proband's 7-year-old brother was also in good health. Routine laboratory tests of these family members were within the normal values (Table 2).

Plasma lipid analysis and LPL assay
T-C, HDL-C, TG, apo A-I and apo B levels were determined by standard methods and plasma lipoprotein were separated by density gradient ultracentrifugation in a SW41 Beckman rotor at 170 000 · g at 15°C [20]. The measurement of plasma apo A-V and apo C-III was performed by ELISA assay; immunoblot analysis of plasma apo A-V was conducted as described [11]. Lipoprotein and hepatic lipase activities in postheparin plasma were measured as previously reported [18].

Analysis of candidate genes
Genomic DNA was extracted from peripheral blood leucocytes by a standard procedure. All exons and the promoter region of LPL and APOC2 genes were amplified by polymerase chain reaction from genomic Plasma level of TG, T-C, HDL-C and LDL-C are given in mmol L )1 ; plasma level of apo B and apo A-I are given in mg dL )1 .
Diet + x-3 fatty acids 3 g day )1 (oral ethyl-ester of polyunsaturated fatty acids EPA and DHA 85%, EPA ⁄ DHA = 0.9-1.5). n.d., not determined. Plasma level of TG, T-C, HDL-C and LDL-C are given in mmol L )1 ; plasma level of apo B and apo A-I are given in mg dL )1 .
DNA and sequenced as described [18]. The APOA5 gene was amplified and sequenced as reported [18]. The mutation of APOA5 gene found in the proband, which eliminates a PstI restriction site in exon 4, was confirmed by restriction fragment analysis of PCRamplified 5¢ region of exon 4, using the following primers: 5¢-CGG CCT GGA TAT CTG TCC CC-3¢ (forward primer) and 5¢-CCC AGT GCC TGC AAA GGC TC-3¢ (reverse primer). The amplification conditions were 96°C for 5 min, 96°C for 1 min per 61.1°C for 1 min per 72°C for 1 min for 35 cycles, followed by 72°C for 5 min. The amplification product (662 bp) was digested with PstI and the fragments were separated by 2% agarose gel electrophoresis. This procedure was used to screen the mutation in 100 randomly selected normolipidaemic subjects.
Nomenclature of the mutation was according to the recommendations of the Human Genome Variation Society (the numerical series of codons includes the sequence of the signal peptide, http://www.hgvs.org/ mutnomen).
The proband and family members were also genotyped for the following polymorphisms: )482 C ⁄ T in APOC3 gene; 2, 3 and 4 in APOE gene [19]. The polymorphisms of APOA5 gene were assessed by direct sequencing.

Ethical aspects
Informed consent was obtained from the patient and his family and the study protocol was approved by the local ethical committee of the University Hospital of Modena.

Plasma lipid and lipoproteins
The plasma lipid levels of the proband, recorded since 2 years of age, are shown in Table 1. Plasma TG level shows large fluctuations over time (especially in early teens), mostly in relation to the compliance of the patient to the low-fat and low-oligosaccharide diet and treatment with x-3 fatty acids. Plasma HDL-C levels showed minor changes and were less affected by dietary treatment. Plasma lipoprotein profile performed when the patient's TG level was 5.90 mmol L )1 showed a marked increase of the peak corresponding to lipoproteins of density <1.030 g mL )1 (Chylomicrons, VLDL + IDL) and a reduction of LDL and HDL peaks (data not shown). Plasma postheparin LPL activity (9.5 lmol F-FA mL )1 ) was within the control range (8-15 lmol FFA mL )1 ). The APOE genotype of the proband was 34. The parents and the brother of the proband were found to have normal lipid levels ( Table 2).

Sequence of APOA5 gene
Since the sequence analysis of LPL and APOC2 genes in the proband revealed no abnormalitites, we sequenced his APOA5 gene. This analysis revealed that the proband was homozygous for a single nucleotide substitution in exon 4 (c.289 C>T), which converts the glutamine codon at position 97 into a termination codon (Q97X) (Fig. 1). The presence of this mutation was confirmed by restriction fragment analysis after digestion of exon 4 with the restriction enzyme Pst1, one site of which is abolished by the mutation (Fig. 2).
The parents and the brother of the proband were found to be heterozygous for the mutation (Figs 1  and 2). APOA5 gene polymorphisms in the proband's family are shown in Table 3. No carrier of the mutation was found amongst 100 randomly selected normolipidaemic individuals, nor in 30 patients with primary severe HTG negative for mutations in LPL or APOC2 genes. The proband was homozygous for the common allele in the promoter of APOC3 gene ()482 C ⁄ C), whilst his parents were heterozygous ()482 C ⁄ T).

Plasma apo A-V and apo C-III levels
Apo A-V was undetectable in proband's plasma; the level of apo A-V in proband's father (26.9 lg dL )1 ) was within the control range (25.8 ± 14.6 lg dL )1 ). Immunoblot analysis of proband's plasma failed to reveal the presence of either wild type or truncated apo A-V (Fig. 3). Plasma apo C-III level in the proband was 18.4 mg dL )1 and 10.8 mg dL )1 in his father (in fasting normolipidaemic subjects 10.6 ± 3.6 mg dL )1 ).

Discussion
In this study we report a novel mutation in APOA5 gene which causes a complete apo A-V deficiency. This mutation was discovered during a survey of 48 patients with severe primary HTG, suspected to have genetic defects in the lipolytic cascade. Amongst the HTG patients found to be negative for mutations in the LPL and APOC2 genes (30 subjects) we identified three patients who carried pathogenic mutations of APOA5 gene. The clinical and biochemical features of the first two patients were described previously [18,21]. The patient investigated in the present study was found to have eruptive xanthomas and milky plasma in the fasting state at the age of 2 years, suggestive of primary hyperchylomicronaemia syndrome. He was put on a restricted dietary regimen (low-fat and low-oligosaccharide diet) which was effective in reducing plasma TG and eliminating eruptive xanthomas. Additional benefits were obtained with supplementation of x-3 fatty acids. During the 15 years of follow-up plasma TG exceeded the critical level of 10 mmol L )1 (regarded to be the threshold for the risk of pancreatitis) only on one occasion. It is most likely that the strict dietary regimen, the regular physical activity during adolescence and the maintenance of an ideal BMI have contributed to prevent excessive elevations of plasma TG over the years. The normal postheparin LPL activity (performed at the age of 12) cast doubt on the presence of defects in the lipolytic cascade and the patient was referred to the laboratory for molecular diagnosis at the age of 17. After the exclusion of LPL and APOC2 genes, the analysis of APOA5 gene showed that proband was homozygous for a point mutation in APOA5 gene, predicted to result in a short truncated protein (Q97X). Neither wild type nor truncated apo A-V was detected in proband's plasma, thus confirming that the patient had a complete apo A-V deficiency. The truncated protein Q97X, if synthesized at all, is expected to be completely devoid of function, as it lacks some domains involved in lipid binding (residues 192-238 and 301-343), LPL activation (residues 192-238) and heparin binding (residues 186-227) [22][23][24].
Phenotypically proband T.L. resembles the patient with complete apo A-V deficiency who was homozygous for the truncated apo A-V Q148X we reported previously [18]. Both these young patients had severe HTG since infancy, which was transmitted as a recessive trait in the family. In both patients the administration of a low-fat diet, especially if supplemented with x-3 fatty acids, induced a substantial reduction of plasma TG level (Table 4).
Although the mechanism by which apo A-V affects plasma TG level is not fully understood, several hypotheses have been proposed including: (i) an intracellular effect on hepatic VLDL assembly and secretion (apo A-V would down-regulate the assembly and secretion of VLDL) and (ii) an effect on plasma removal of TG-rich lipoprotein, either by a direct modulation of the activity of lipoprotein lipase or by an effect on the uptake of these lipoproteins (or their remnants) by hepatocytes [24].
Some studies indicate that apo A-V facilitates the action of LPL. In view of its high affinity for heparin, apo A-V would increase the binding of TG-rich lipoproteins to glycosaminoglycans present on the surface of endothelial cells, thus rendering these lipoproteins more accessible to LPL [25]. Recently, Nilsson et al. [26] provided evidence that apo A-V binds to two endocytotic receptors of the LDLR family, LRP and SorLA, and mediates the binding of chylomicrons to these receptors. They also put forward the hypothesis that the main action of apo A-V is exerted at the sinusoidal surface of the hepatocytes rather than in the plasma compartment. Apo A-V (which is synthesized exclusively in the liver)  secreted by the hepatocytes would be entrapped by heparan-sulphate proteglycans and endocytotic receptors present at the cell surface [27]. In this way apo A-V would facilitate lipoprotein uptake (i.e. chylomicron remnants removal). These results are in line with the observation that the clearance of chylomicron remnants is reduced in apoa5 knock-out mice [28]. In this context one might assume that in patients with apo A-V deficiency the LPL-mediated TG hydrolysis of chylomicrons and VLDL is less efficient and the clearance of lipoprotein remnants by the liver is delayed. The magnitude of the accumula-tion in plasma of TG-rich lipoproteins in these patients is probably related to the amount of these lipoproteins entering the plasma compartment. Excessive secretion of chylomicrons due to a high fat intake, or excessive secretion of VLDL by the liver induced by various metabolic stimuli, may overload the less efficient LPL-mediated hydrolysis of TG-rich lipoproteins and impair the capacity of the liver to capture lipoprotein remnants, in case of apo A-V deficiency. This may explain why in our two patients with apo A-V deficiency (Q97X and Q148X) dietary interventions, aimed at reducing chylomicron formation (reduced oral fat load) and ⁄ or VLDL synthesis and secretion (administration of x-3 fatty acids) [29], have such a profound TG-lowering effect.
The other remarkable finding in proband T.L. is the persistently low level of plasma HDL-C, which was only marginally affected by dietary treatment. The hypothesis that the absence of apo A-V, which is a minor protein component of HDL, renders HDL more unstable and more easily removed from the circulation, is unlikely in view of the minute amounts of apo A-V present in human plasma. It is more likely that the delayed TG hydrolysis, due apo A-V deficiency per se or due to the increased level of apo C-III (an inhibitor of LPL) [30], reduces the availability of surface components of TG-rich lipoproteins, which contribute to HDL formation. It was recently shown that adenovirus mediated delivery of apoA-V into HTG APOC3 transgenic mice not only reduced triglyceride and VLDL-C level but also increased HDL-C level with an increase of both the number and size of HDL particles [31]. These data suggest that in addition to its TG-lowering effects, apoA-V plays a significant role in modulating HDL maturation and cholesterol metabolism.
The three heterozygotes for Q97X mutation investigated in this study (Table 2) have normal plasma TG, suggesting that partial apo A-V deficiency is not sufficient per se to induce HTG and other predisposing factors are required. We previously reported that only five out of ten carriers of Q148X mutation had moderate elevation of plasma TG (>2 mmol L )1 ; the highest level observed was 4.65 mmol L )1 ). Carriers with elevated plasma TG tended to be older, to have higher body mass index and higher fasting plasma glucose level than normotriglyceridaemic carriers [18]. Amongst the carriers of Q139X belonging to one of the families reported by Marcais et al. [19], two had severe HTG and four had plasma TG levels of 2 mmol L )1 or less. The two individuals with severe HTG had diabetes and low LPL activity.
In conclusion, we report the third mutation in APOA5 gene that results in apo A-V deficiency in a patient with primary HTG. We also confirmed that strict dietary treatment and supplementation with x-3 fatty acids induce a substantial reduction of plasma TG level.
We predict that the systematic analysis of the major candidate genes controlling the lipolytic cascade in patients with primary HTG will lead to the discovery of other loss of function mutations of APOA5 gene.

Conflict of interest
No conflict of interest was declared. Deceased at the age of 60 from a cause unrelated to dyslipidaemia [19]