Introduction

Recent research on atherosclerosis has been focusing on mechanisms involved in formation of atherosclerotic plaques. It is now understood that oxidized low-density lipoprotein (LDL) is a trigger for atherosclerotic lesions; it releases several biologically active compounds, such as fragmented oxidized phospholipids, at sites where new plaques form (Caslake and Packard 2003).

As a class, A2 phospholipases (PLA2) regulate phospholipid content and distribution of LDL and high-density lipoprotein (HDL) in plasma. Lipoprotein-associated PLA2 (Lp-PLA2), a serine-dependent enzyme, hydrolyzes sn2-ester bonds on phospholipids when these molecules are fragmented or oxidized (Stremler et al. 1991; Stafforini et al. 1997). Lp-PLA2, encoded by a gene located at 6p12–p21.1, is a 45.4-kDa (441 amino acids) protein expressed in and secreted by monocytes and macrophages (Stafforini et al. 1996a). About 80% of the Lp-PLA2 in plasma is bound to LDL via specific interaction between the C-terminus of apoB-100, the major constitutive protein of the LDL particle, and two domains of Lp-PLA2, one around tyrosine 205 and the other around tryptophan 115–leucine 116. The rest of the plasma Lp-PLA2 is bound to HDL and VLDL (Stafforini et al. 1999; Kujiraoka et al. 2003). The Lp-PLA2 enzyme is also known as platelet-activating factor (PAF) acetylhydrolase (PAF-AH), since one of its known functions is to deactivate PAF (Stafforini et al. 1987a,b). PAF is a potent lipid mediator of inflammatory disease (Snyder 1995).

A view has developed that Lp-PLA2 may have a protective role in atherosclerosis. For instance, this enzyme degrades atherogenic oxidized lipids (Steinbrecher and Pritchard 1989; Stremler et al. 1989; Stafforini et al. 1992), and deficiency of Lp-PLA2 is a predisposing factor for cardiovascular diseases in people of Japanese ethnicity. That predisposition is mainly due to the presence of a common mutant allele, V279F; homozygotes of the mutant allele comprise 0.04 of this population, although V279F has not been reported in Caucasians (Miwa et al. 1988; Stafforini et al. 1996b). Deficiencies of Lp-PLA2 in Japanese are associated with increased risk of stroke (Hiramoto et al. 1997), coronary artery disease (Yamada et al. 1998, 2000), atherosclerotic occlusive disease (Unno et al. 2000), and abdominal aortic aneurysms (Unno et al. 2002).

We describe here two novel mutant alleles of the Lp-PLA2 gene identified in Japanese patients with Lp-PLA2 deficiency and provide experimental evidence that both cause functional defects.

Materials and methods

Patients and clinical profiles

Patients with complete or partial Lp-PLA2 deficiencies were identified from a group of 295 individuals who were followed clinically for hyperlipidemia or diabetes or were receiving health checkups at the outpatient clinic of the Hokkaido Hospital for Social Insurance (Sapporo, Japan). Blood samples were collected after 12–16 h of fasting (Fujita et al. 2003), and all participants provided informed consent for participation in the study. Measurement of serum total cholesterol (TC), HDL, triglycerides (TG), and apolipoproteins was carried out as described previously (Ishii et al. 2002; Takada et al. 2002). LDL cholesterol was calculated according to the Friedewald formula (Friedewald et al. 1972). Plasma Lp-PLA2 (PAF-AH) activity was determined using [3H] PAF, and Lp-PLA2 mass concentration was determined by sandwich enzyme-linked immunosorbent assay (ELISA) as described by Kujiraoka et al. (2003). Genomic DNA was extracted as previously described (Yoshida et al. 2002). Of the 120 subjects exhibiting Lp-PLA2 deficiency, 119 were found to carry a mutant allele of Lp-PLA2 that is common in Japan—V279F (Kujiraoka et al. 2003).

Two subjects displayed discordance between Lp-PLA2 deficiency status and V279F genotype (Table 1). Case 1, a male patient who was 65 years old, 170 cm tall, and 64 kg in weight, had developed type 2 diabetes (NIDDM) and hypertension at age 62; his elder brother had cerebrovascular disease. The second subject (case 2), also male, was 53 years old, 159 cm tall, and 61 kg in weight; he had hyperlipidemia, hypertension, and cardiovascular disease with an occlusive left-coronary lesion indicated by angiography. His mother had died of acute myocardial infarction at age 68 and his father of cancer at 72.

Table 1 Clinical profiles of two patients with plasma Lp-PLA2 deficiency and novel mutations. LDL low-density lipoprotein, HDL high-density lipoprotein
Full size table

Isolation and culture of monocyte-derived macrophages

Fresh blood from both subjects was collected in heparin-containing tubes. After mononuclear cells were isolated by centrifugation with Separate L (Muto Pure Chemicals Co., Osaka, Japan), monocyte-derived macrophages were cultured as previously described (Kosaka et al. 2001). On day 7, the medium was replaced with RPMI supplemented with 10% bovine serum albumin, 100 U/ml penicillin, and 100 μg/ml streptomycin. After culture for 2 days or more, the monocyte-derived macrophages and the culture media were separated and reserved for subsequent experiments.

PCR amplification, DNA sequencing, and PCR-RFLP assay

All 12 exons of the Lp-PLA2 gene were individually amplified by PCR under conditions described previously (Iida et al. 2002) using as primers specific oligonucleotides derived from intronic sequences flanking each exon. Sequences surrounding the variations, as well as primer sequences and experimental conditions, were obtained from published reports (Haga et al. 2002; BLAST Accession number 16173643). Each segment was amplified and sequenced according to procedures described previously (Hattori et al. 2002). V279F and Q281R genotypes of Lp-PLA2 were examined in these patients by Invader assay as previously described (Nagano et al. 2002; Saito et al. 2002) using PCR products of the flanking sequences and probes designed and synthesized by the supplier (Ohnishi et al. 2001).

PCR-based assay for the InsA191 and I317N mutations

PCR-RFLP assays for InsA191 and I317N mutant alleles were carried out by amplifying a 283-bp fragment of exon 3 and a 295-bp fragment of exon 10, respectively, using primers 5′-GAAGGCCAGAAACTATGGGAA-3′ (sense) and 5′-GCGACAGACAATATTAGAGTATCAGG-3′ (antisense) for InsA191 and 5′-CCTCAATGTTGGCTAGTATGTAACC-3′ (sense) and 5′-TCGATTGATATTTACCCAAGCTTCA-3′ (antisense) for I317N, followed by digestion with restriction enzyme Tru9I (New England Biolabs, Beverly, MA, USA) at 37°C for 3 h. Digested samples were electrophoresed on a 7.5% PAGE gel, and the bands were visualized with ethIdium bromide.

Site-directed mutagenesis and transfection of COS-7 cells for V279F, Q281R, and I317N mutant alleles

A cDNA fragment encoding amino acids 42–441 of Lp-PLA2 was obtained by RT-PCR from mRNA expressed by peripheral monocyte-derived macrophages and subcloned into the mammalian expression vector pcDNA3.1(+) (Invitrogen, CA, USA) in the manner described by Kujiraoka et al. (2003). Mutant cDNAs (V279F, Q281R, and I317N) were generated from the pcDNA3.1(+)/Lp-PLA2 vector using QuickChange XL Site-Directed Mutagenesis Kits (Stratagene) according to the manufacturer’s instructions. The integrity of each plasmid was verified by DNA sequencing.

One day before transfection, COS-7 cells (4×105 cells/well) were plated onto 6-well plates (Iwaki Glass, Tokyo, Japan) pre-coated with type I collagen (Nagano et al. 2002). The next day, 1 μg of each plasmid was transfected by means of LIPOFECTAMIN 2000 reagent (Invitrogen) into cultures showing 95% confluence. Cells were collected after 48 h of growth in the same medium, and Lp-PLA2 concentrations in the media were determined by ELISA.

Western-blot analysis

Media from the secretion experiments were incubated for 18 h at 4°C with antihuman Lp-PLA2 monoclonal antibody A7G-coated or 8B1-coated Dynabeads (M-450 Sheep anti-mouse IgG; Dynal) at 1 ml medium/1×107 beads. After incubation, the beads were precipitated by centrifugation and washed three times with 0.1% Tween20 (Nakalai Tesque) containing PBS, resuspended in SDS-PAGE sample buffer, heated to 95°C for 5 min, and then subjected to SDS-PAGE. After electrophoresis, the proteins were transferred to PVDF membranes (Millipore), and Western blotting was performed using a polyclonal antibody that had been raised in rabbits against recombinant human Lp-PLA2 expressed by E. coli (Kujiraoka et al. 2003), and goat HRP antirabbit IgG (Zymed). Detection was achieved by means of Western Lightning Chemiluminescence Reagent (PerkinElmer) according to manufacturer recommendations.

Results

Screening patients for Lp-PLA2 deficiency

We examined Lp-PLA2 concentrations in plasma of 295 individuals at the Hokkaido Hospital for Social Insurance using a highly sensitive sandwich ELISA assay that we had developed recently. We also used the Invader assay to screen these patients for a mutant genotype of the Lp-PLA2 gene, V279F, that is common in the Japanese population. Plasma Lp-PLA2 concentrations among heterozygotes for the V279F mutation (n=36) were about half the amounts for wild-type homozygotes (0.79±0.24 μg/ml versus 1.61±0.44 μg/ml, respectively). Subjects who were homozygous for the V279F mutation (n=2) displayed no Lp-PLA2 mass in plasma. Another previously known but rare mutation, Q281R, was present in some patients with partial Lp-PLA2 deficiency (n=4), i.e., who exhibited concentrations of plasma Lp-PLA2 that were reduced by about half (0.83±0.15 μg/ml). However, one subject (case 1) showed reduced Lp-PLA2 mass (0.58 μg/ml) without any known mutations, and another (case 2) had no detectable Lp-PLA2 mass in plasma but was only heterozygous for the V279F mutation (Table 1).

Detection of novel InsA191 and I317N mutations of Lp-PLA2

Screening the Lp-PLA2 gene for mutations in case 1 identified heterozygosity for a mutation involving insertion of adenine at nucleotide 191 (exon 3); this change had created a premature termination at codon 63 (Fig. 1a) and a consequently truncated protein. We identified case 2 as compound heterozygosity for the common V279F mutation and a novel I317N mutation. The latter had substituted adenine for thymine T→A at nucleotide 950 in exon 10, resulting in substitution of Asn for Ile at codon 317 (I317N) (Fig. 1b). This mutation would create a new N-linked glycosylation site (N-X-S) on the Lp-PLA2 protein. Both mutations were confirmed by PCR-RFLP assays using restriction enzyme Tru9I (data not shown).

Fig. 1a,b
figure 1

Nucleotide sequences of two novel mutations, InsA191 and I317N, of the Lp-PLA2 (PAF-AH) gene in two Japanese patients. a Case 1: insertion of nucleotide A at codon 63 in exon 3, which created a premature termination codon (TAA). b Case 2: T-to-A substitution at codon 317 in exon 10, which replaced a codon for isoleucine (ATC) codon with aspartate (AAC)

Full size image

Impaired secretion of Lp-PLA2 from macrophages carrying InsA191 and I317N mutant alleles

We examined secretion of Lp-PLA2 protein into culture medium from stimulated endogenous monocyte-derived macrophages obtained from each of the two patients. Macrophages of case 1 secreted Lp-PLA2 protein into the medium at half the level of wild type (42.6 ng/mg cell protein versus 82.8 ng/mg cell protein), demonstrating that the truncating mutation, InsA191, indeed caused partial Lp-PLA2 deficiency in that patient (Fig. 2a). No detectable Lp-PLA2 protein was secreted by stimulated macrophages from the compound heterozygote (case 2). However, RT-PCR experiments showed that Lp-PLA2 mRNA was expressed to same extent in those macrophages (Fig. 2b).

Fig. 2a,b
figure 2

Secretion of Lp-PLA2 by cultured monocyte-derived macrophages. a Measurement of Lp-PLA2 mass by ELISA in the culture medium on day 7. Values represent the mean ± SD of triplicate assays from two separate experiments. Wild wild type, HE heterozygote, HO homozygote. Case 1, InsA191 heterozygote; Case 2, compound heterozygote for V279F and I317N mutations. b RT-PCR showing expression of Lp-PLA2 mRNA in macrophages. M size marker, lane 1 wild type, lane 2 V279F heterozygote, lane 3 V279F homozygote, lane 4 case 1, lane 5 case 2

Full size image

Expression of 317N Lp-PLA2 mutant in COS-7 cells

To clarify the molecular mechanism underlying the I317N missense variation, mutant 317N cDNA was transfected into COS-7 cells and expressed in parallel with expression vectors for wild type or 279F and 281R mutant alleles. Values for Lp-PLA2 mass and enzymatic activity, indicated in Fig. 3 by clear or black bars, respectively, show that the previously known common mutant 279F produced neither Lp-PLA2 mass nor activity in COS-7 cells and that the rare 281R allele produced one tenth the mass and activity of the wild-type protein. The novel I317N mutant from case 2 failed to produce any Lp-PLA2 mass or enzymatic activity in transfected cells. Western blotting verified the lack of mutant Lp-PLA2 proteins in lysates of COS-7 cells transfected with the mutant vectors corresponding to I317N or V279F (data not shown).

Fig. 3
figure 3

Lp-PLA2 mass and enzymatic activity in media after COS-7 cultures were transfected separately with expression constructs of different mutant alleles. Lp-PLA2 mass (clear bars) and enzymatic activity (black bars) were measured 48 h after transfection. Values represent the mean ± SD of triplicate assays from two separate experiments

Full size image

Discussion

In the present study, we identified and functionally characterized two novel mutations of the Lp-PLA2 gene, InsA191 and I317N, in Japanese subjects with deficiencies in plasma Lp-PLA2 levels. The InsA191 mutation, which created a premature termination, was present on one allele of the gene in a diabetic (NIDDM) patient with partial Lp-PLA2 deficiency. Monocyte-derived macrophages from this subject secreted only half the wild-type amount of Lp-PLA2 enzyme in vitro.

The I317N mutation, which would create a new N-linked glycosylation site (N317-N-S) in the enzyme’s catalytic domain, was found in a subject with hyperlipidemia and coronary heart disease whose plasma showed complete Lp-PLA2 deficiency due to compound heterozygosity for 317N and 279F. The wild-type amino acid sequence of human Lp-PLA2 (Tjoelker et al. 1995) contains two conserved N-linked glycosylation sites, N423-T-T and N433-S-S, close to the carboxy terminus. Among members of the PLA2 family, glycosylation reduces association of the enzymes with lipoproteins but does not reduce secretion or enzymatic activity (Tselepis et al. 2001). In our experiments, neither monocyte-derived macrophages nor COS-7 cells transfected with a mutant (317N) cDNA expression vector secreted any Lp-PLA2, suggesting that this mutation disturbs processes leading to secretion of mature enzyme. These results verified that both of the novel Lp-PLA2 mutations described here indeed impair proper function of the enzyme.

Lp-PLA2 is also known as PAF-AH, since one of its known functions is to degrade and deactivate PAF and other bioactive phospholipids that mediate inflammatory disease (Stafforini et al. 1997). In some studies, a deficiency of plasma Lp-PLA2 was three times more common in children with severe bronchial asthma than in the general population, suggesting that Lp-PLA2 may play an important role in inflammatory and allergic responses (Stafforini et al. 1999). Several other investigations have revealed that Lp-PLA2 mutations increase susceptibility to inflammatory and allergic diseases (Watson et al. 1995; Tjoelker and Stafforini 2000). The two patients examined in this study did not have histories of allergic diseases, but the data suggest that individuals with deficiencies of plasma Lp-PLA2 are at increased risk of severe responses to specific events that cause allergic or inflammatory overload.

The Lp-PLA2 enzyme is thought to have a protective role in atherosclerosis because numerous reports have associated Lp-PLA2 deficiency with various cardiovascular diseases and because this enzyme catabolizes atherogenic oxidized or fragmented lipids. In fact, deficiency of Lp-PLA2 has been shown to be a predisposing factor for cardiovascular diseases in Japan. On the other hand, recent epidemiological studies carried out in Caucasian populations have suggested that the enzyme might be a positive risk factor for cardiovascular diseases. For example, Packard, Caslake and their colleagues have described a positive association between plasma Lp-PLA2 levels and incidence of coronary heart disease in case-controlled studies among males in Scotland (Caslake et al. 2000; Packard et al. 2000). A similar association between plasma Lp-PLA2 mass and coronary heart disease has been observed in Caucasian females (Blake et al. 2001).

The apparent discrepancy between suggestions of a negative role for Lp-PLA2 in atherosclerosis versus a protective role based on data involving Japanese patients with Lp-PLA2 deficiency appears to reflect the complex nature of the metabolism and activity of this enzyme as well as ethnic specificity of the genetic variations present in Japanese and Caucasian populations. These points need to be carefully examined before a full understanding of the role of this enzyme in vascular and other human diseases can be achieved.