A sandwich enzyme-linked immunosorbent assay for human plasma apolipoprotein AV concentration

Apolipoprotein A-V (apoA-V) is a recently discovered apolipoprotein that appears to have a role in plasma triglyceride (TG) transport. We have developed an ELISA for apoA-V using monoclonal antibodies that has a lower limit of detection of 0.3 ng/ml and linearity up to 20 ng/ml. The ELISA was then used to quantify plasma apoA-V in 196 healthy subjects and 106 patients with insulin-resistant diabetes mellitus. In the healthy subjects, total apoA-V concentration was 179.2 74.8 ng/ml, and it was greater in females than in males ( P 0.005). It was correlated positively with the plasma HDL cholesterol ( r 0.32, P 0.0001), apoA-I ( r 0.27, P 0.0001), and apoE ( r 0.18, P 0.011) concentrations and negatively with plasma TG concentration ( r 0.22, P 0.021). In relation to single nucleotide polymorphism 3 ( 1131C/T) of the apoA-V gene, apoA-V concentration was higher in the T/T type than in the C/C type ( P 0.01). Plasma TG concentration was lower in the T/T type than in the C/C or C/T type ( P 0.05). ApoA-V concentration was lower in the diabetic patients (69.4 44.3 ng/ml; P 0.01) than in the healthy controls. —Ishihara, M., T. Kujiraoka, T. Iwasaki, M. Nagano, M. Takano, J. Ishii, M. Tsuji, H. Ide, I. P. Miller, N. E. Miller, and H. Hattori. A sandwich enzyme-linked immunosorbent assay for human plasma apolipoprotein A-V concentration. J. Lipid Res. 2005. 46: 2015–2022. Supplementary key words triglyceride • diabetes mellitus • immunoassay • single nucleotide polymorphism Plasma triglyceride (TG) levels are influenced by both genetic and environmental factors and are a major independent risk factor for coronary heart disease (1, 2). Plasma TG concentration is influenced by many factors. These include apolipoproteins A-I, A-IV, C-II, and C-III, LPL, LCAT, cholesteryl ester transfer protein, and phospholipid transfer protein (3–11). These factors and their associated gene-environment interactions are of importance in the pathogenesis of coronary heart disease. Apolipoprotein A-V (apoA-V) has recently been identified by comparative sequencing of human and mouse DNA and is located 27 kb distal to the apoA-IV gene in the APOA1/C3/A4 gene cluster on chromosome 11q23 (12). ApoA-V, shown to be expressed mostly in liver and independently named regeneration-associated protein 3, is upregulated after the early phase of liver regeneration after hepatectomy in rat (13). In mice overexpressing the human apoA-V gene, TG concentrations decreased by 50–70%, and in apoA-V gene knockout mice, plasma TG concentrations increased 4-fold (12–14). These results suggest that apoA-V expression may strongly influence, and be negatively associated with, plasma TG concentrations. ApoA-V both enhances lipoprotein lipase-mediated hydrolysis of plasma TG and inhibits hepatic VLDL-TG production (15). ApoA-V also stimulates the efflux of cholesterol from cells by a mechanism independent on the ABCA1 protein, as do other exchangeable apolipoproteins, such as apoA-I and apoA-IV (16). It was recently described that apoA-V mRNA is regulated by peroxisome proliferator-activated receptor agonists (17, 18) and that the liver X receptor ligand T0901317 decreases apoA-V mRNA through the activation of sterol-regulatory element binding protein 1c (SREBP-1c) (19). These results raise the possibility that some TG-lowering agents, such as fenofibrate, may act by altering the expression of apoA-V. In addition, associations have been identified between plasma TG concentrations and several apoA-V polymorphisms, including 1131T/C, 3A/G, S19W, and 1259T/C Abbreviations: apoA-V, apolipoprotein A-V; CBB, Coomassie brilliant blue; MAb, monoclonal antibody; NIDDM, non-insulin-dependent diabetes mellitus; rhapoA-V, recombinant human apolipoprotein A-V; SNP, single nucleotide polymorphism; SREBP-1c, sterol-regulatory element binding protein 1c; TG, triglyceride. 1 To whom correspondence should be addressed. e-mail: hhiro@bml.co.jp Manuscript received 14 April 2005 and in revised form 31 May 2005. Published, JLR Papers in Press, June 16, 2005. DOI 10.1194/jlr.D500018-JLR200 methods by gest, on O cber 9, 2017 w w w .j.org D ow nladed fom

Apo A-V has recently been identified by comparative sequencing of human and mouse DNA, and is located about 27kb distal to the apo A-IV gene in the APOA1/C3/A4 gene cluster on chromosome 11q23 (12). Apo A-V, shown to be expressed mostly in liver and independently named regeneration-associated protein 3 (RAP3), is upregulated after the early phase of liver regeneration after hepatectomy in rat (13). In mice overexpressing the human apo A-V gene, triglyceride concentrations decreased by 50-70%, and in apo A-V gene knockout mice plasma triglyceride concentrations increased approximately fourfold (12 -14). These results suggest that apo A-V expression may strongly influence, and be negatively associated with, plasma triglyceride concentrations. Apo A-V both enhances lipoprotein lipase-mediated hydrolysis of plasma TG and inhibits hepatic very low density lipoprotein (VLDL)-TG 5 by guest, on  www.jlr.org Downloaded from production (15). Apo A-V also stimulates the efflux of cholesterol from cells by a mechanism dependent on the ATP binding cassette transporter A1 (ABC-A1) protein, as do other exchangeable apolipoproteins, such as apo A-I and apo A-IV (16). It has recently been described that apo A-V mRNA is regulated by peroxisome proliferatoractivated receptor-alpha (PPAR-alpha) agonists (17,18), and that the liver X receptor (LXR) ligand T0901317 decreases apo A-V mRNA through the activation of sterol regulatory element-binding protein (SREBP) -1c (19). These results raise the possibility that some triglyceride-lowering agents, such as fenofibrate, may act by altering the expression of apo A-V. In addition, associations have been identified between plasma TG concentrations and several apo A-V polymorphisms, including -1131T/C, -3A/G, S19W and 1259T/C (12,(20)(21)(22)(23)(24)(25)(26). Plasma apo A-V concentrations have recently been measured in Caucasians by using an ELISA procedure that employs polyclonal antibodies against the NH 3 -and COOH-terminus of the protein (27).
We have raised two monoclonal antibodies against human apo A-V, and used them to develop a new sandwich enzyme-linked immunosorbent assay (ELISA). We then used the assay to study plasma total apo A-V concentrations, and the distribution of apo A-V, between high density lipoproteins (HDLs) and other lipoproteins, in healthy subjects.

Subjects
Blood from 196 apparently healthy volunteers (105 males, 91 females) without any medication, who had fasted overnight, was collected at the BML Clinical Reference Laboratory (Saitama, Japan). Blood from 106 non-insulin dependent diabetics (61 males and 45 females) was collected in the outpatient clinic of the Hokkaido Hospital for Social Insurance (Sapporo, Japan) after overnight fasting. EDTA-plasma was isolated immediately by centrifugation at 4ÚC and stored at -80ÚC until use. Subjects were not taking medications. Lipid profiles are shown in Table 1. In healthy subjects, concentrations of total and LDL cholesterol, triglyceride and apo AI were greater, and those of apo A-I and apo E were lower, in males than in females. This study was approved by the ethical committee of Hokkaido Hospital for Social Insurance and BML.
Informed consent was obtained from all subjects.

Cloning of human apo A-V and expression of recombinant apo A-V (rhapo A-V)
7 by guest, on  www.jlr.org

Downloaded from
Human apo A-V cDNA was obtained by RT-PCR from mRNA of HepG2 cells. PCR was carried out using as the sense primer, 5'-GACGGATCCAAAGGCTTCTGGGACTACTTCAGCC-3' and as the antisense primer, 5'-GACGTCGACTCAGGGGTCCCCCAGATGGCTGTGG-3' for the apo A-V cDNA1, and as the sense primer 5'-GACGAATTCAGCAGATAATGGCAAGCATGGCTGC-3' and as the antisense primer, 5'-GACGAATTCTCAGTGATGGTGATGGTGATGGGGGTCCCCCAGATGGCTGTG GCCC-3' for the apo A-V cDNA2. Each apo A-V cDNA was encoding the AA22-363 and AA 1-363 for cDNA1 and cDNA2, respectively, and apo A-V cDNA2 was constructed with a 6xHis tag at the COOH-terminus. The apo A-V cDNA1 was subcloned into the pQE30 plasmid (QIAGEN, CA) to yield the pQE-30/apo A-V1 vector. E. coli. JM109 (Toyobo, Tokyo,Japan) bearing the pQE-30/apo A-V1 plasmid were cultured in TB medium containing 50 mg/L ampicillin at 37ÚC. Expression was induced with 1 mM isopropyl thiogalactopyranoside (IPTG), and after 5h, the cells were blue (CBB) staining, was determined by gel scanning using the Intelligent Quantifier system (BioImage) as previously described (28).
The apo A-V cDNA2 was subcloned into the pEF321 mammalian expression vector (28) to yield the pEF321/apo A-V vector. Chinese hamster ovary (CHO)-K1 cells stably transfected with pEF321/apo A-V vector were cultured in serum-free medium CHO-S-SFM II (Invitrogen, CA) and the culture medium was collected.
Rhapo A-V was partially purified by metal affinity column chromatography using TALON® Metal affinity resin (Clontech, CA). The purity of rhapo A-V was confirmed by SDS-PAGE followed by CBB staining or immunoblotting. For immunoblotting, rhapo A-V was detected with Tetra-His antibody (Qiagen) as the primary antibody and horseradish peroxidase-conjugated anti-mouse IgG (Zymed Laboratories, CA) as the secondary antibody. Bound antibodies were detected with an enhanced chemiluminescence kit (Perkin Elmer Life Sciences, MA).

Preparation of mAbs against apo A-V
Monoclonal antibodies (MAbs) against apo A-V were obtained by the method of DNA-based immunization (29,30). In brief, Balb/c mice were injected subcutaneously with 50 µg plasmids of apo A-V cDNA2 inserted into pcDNA3.1(+) vector (Invitrogen) six times every two weeks. The final immunization was done intraperitoneally with 2.5 µg rhapo A-V from CHO-K1 cells, and spleen cells from the mice were fused with Sp2/0 cells (31). The supernatants of hybridoma cells were screened by ELISA using plates coated with partially purified rhapo A-V (50 ng/well) and by immunoblotting. Positive hybridoma cells were cloned at least three times by limiting dilution and injected intraperitoneally into pristane-primed Balb/c mice. The IgG fraction was isolated from ascitic fluid using protein A-Sepharose FF as previously described (28), dialyzed at 4ÚC against PBS, and stored at -80ÚC. The specificities of mAb B10E and mAb E8E were confirmed by ELISA and immunoblotting against purified HDL and rhapo A-V. MAb isotype was characterized using the IsoStrip mouse monoclonal antibody isotyping kit (Roche Diagnostics, Basel, Switzerland), and was IgG 2a and IgG 1 for mAb B10E and mAb E8E, respectively.

Determination of the apo A-V genotype by Invader® assay
The single nucleotide polymorphism (SNP) 1,131 bp upstream of the transcription start site (T-1131 C; SNP3) of the apo A-V gene was detected by the Invader® assay as previously described (28,32). Primary probes and Invader oligonucleotide for each mutation were designed with Invader® Creator software to have theoretic annealing temperatures of 63ÚC and 77ÚC, respectively, using a nearest-neighbor algorithm on the basis of final probe and target concentrations. The primary probes and Invader oligonucleotides used are shown in Table 2. Genotyping was performed by calculation, using the ratios of net counts with wild primary probe to net counts with mutant primary probe. The accuracy of each genotyping was 100%, determined by comparison with results previously obtained by PCR-RFLP analysis and direct sequencing. Co., Tokyo, Japan) (28). LDL cholesterol concentration was calculated according to Friedewald et al (33). Protein content was determined by the BCA protein assay kit (Pierce, CA) using BSA as a calibrator. SDS-PAGE was performed by the Laemmli method (34) and immunoblotting as described by Towbin et al. (35).

Statistical analysis
Results were expressed as mean ± SD. ANOVA was used for group comparisons.
Correlations were analyzed by the Spearman's rank correlation coefficient. P < 0.05 was considered statistically significant.

Characterization of anti-apo A-V mAbs
The bacterial rhapo A-V purified from the lysate of E. coli showed a major band of about 40 kDa protein (Fig. 1A). This represented more than 95% of total protein after scanning of the gel. Mice were first immunized by DNA injection, followed by partially purified rhapo A-V from CHO culture medium. Two monoclonal antibodies (mAbs) specific for apo A-V were established: mAb B10E and mAb E8E. When rhapo A-V and human plasma were subjected to SDS-PAGE, both mAbs reacted with a single protein (Fig. 1B), the molecular mass of which (40 kDa, Fig. 1B, lane 1 and 2) was similar to that previously reported for human plasma apo A-V (16). The molecular weight of bacterial rhapo A-V (AA22-363) appeared to be greater than those in plasma and culture medium, suggesting that plasma apo A-V may be secreted with more processing. By agarose electrophoresis and Western blotting, apo A-V was detected in the portion corresponding to alpha-lipoproteins, presumably owing to the lower content of apo A-V in the latter (data not shown). There was no evidence of recognition of other plasma proteins. Both mAbs similarly reacted with rhapoA-V from E. coli or CHO cells coated on a microtiterplate (Fig. 2).

Standardization of ELISA for plasma apo A-V concentration
A sandwich ELISA for plasma apo A-V was established using mAb B10E for 14 by guest, on October 14, 2017 www.jlr.org Downloaded from capture and biotinylated mAb E8E for detection. The system showed a dose-dependent response to purified bacterial rhapoA-V, to CHO culture medium expressing rhapo A-V and to plasma, and the reactivity was equal with both bacterial and mammalian rhapo A-V (Fig. 3). For calibration of the ELISA, purified bacterial rhapo A-V was used as the primary calibrator. When subjected to SDS-PAGE and visualized by CBB staining, the purified bacterial rhapo A-V showed a single major 40 kDa band (Fig. 1), which represented > 95% of the total protein in the preparation (as determined by gel scanning using the Intelligent Quantifier system). The protein concentration of this primary rhapo A-V calibrator, assayed using a BCA protein kit with BSA as calibrator, was typically 1.94 mg/mL.
To obtain a calibration curve for the ELISA, dilutions of the primary calibrator were made in PBS containing 5 g/L CHAPS to provide 0.3125 -20.0 ng of rhapo A-V protein per well (15.6 -1000 ng/mL). When the rhapo A-V culture medium, as a secondary calibrator, was diluted in PBS containing 5 g/L CHAPS to cover the apo A-V concentration range 0.3125 -20.0 ng /mL, the curve was identical to that obtained with the primary calibrator (Fig. 4) with plasma were not identical to those obtained with the recombinant proteins (Fig. 3).
Fifty-fold dilution of plasma, in which the diluted aliquot that gave an absorbance between 0.5 and 1.2, was chosen for routine use.
The detergent CHAPS was included in the diluent to avoid any effects of differences between samples in their lipid or apolipoprotein compositions. We Industries). Plasma samples diluted (10-fold) and the rhapo A-V culture medium diluted (20-fold) with PBS gave similar absorbance with each detergent, but most detergents gave higher absorbance in the background (blocking buffer alone), and only two detergents, CHAPS and CHAPSO, showed lower background (absorbance less than 0.1). Therefore we chose PBS containing 5 g/L CHAPS as the sample diluent (data not shown). The day-to-day variation and between-plate-within-day variation in the ELISA were 5.5 ~ 8.8% (n = 5) and 2.2 ~ 3.8% (n = 10), respectively.

Plasma apo A-V concentrations in healthy subjects
Results for apo A-V concentration in healthy men and women are presented in Table 3. The average of plasma apo A-V concentrations was 179.2 ± 74.8 ng/mL, and they were higher in females than in males (P < 0.005). In all subjects combined (both

Plasma apo A-V concentrations in relation to the apo A-V polymorphism
Several single nucleotide polymorphisms (SNPs) of the apo A-V gene are commonly present in humans (12,(21)(22)(23)(24)(25)(26)(27). In the present study, the SNP3 at the position -1131 nt of the apo A-V gene was analyzed in healthy subjects, and the frequency for the T and C alleles was 0.635 and 0.365, respectively. Apo A-V concentration was statistically higher (P < 0.01) in subjects with the T/T type than those with the C/C type but not with the C/T type ( significantly lower in subjects with the T/T type than those with the C/C type (P < 0.01) or the C/T type (P < 0.05).

Plasma apo A-V concentrations in non-insulin dependent diabetes
In both men and women apo A-V concentration was much lower in subjects with non-insulin dependent diabetes mellitus (NIDDM) than in healthy controls (Table   3). This was in spite of the fact that LDL cholesterol and apo B concentrations were higher than in controls, and there were significant differences between the two groups in

Discussion
We have developed a sandwich ELISA for plasma apo A-V concentration, using two monoclonal antibodies against apo A-V produced by DNA injection. The specificity of the antibodies was confirmed by immunoblotting. Both mAb E8E and m Ab B10E reacted with a single protein in human plasma of about 40 kDa molecular mass, which is the same as that previously reported for plasma apo A-V (27). Both mAb E8E and mAb B10E reacted with human VLDLs and HDLs under denaturing conditions (data not shown), and both mAbs reacted similarly with purified rhapo A-V from E. coli and partially purified rhapo A-V from CHO cells coated on to a microtiter plate. The ELISA also measured equally rhapo A-V from bacterial and mammalian cells. Our ELISA can be used to measure up to 1000 ng/mL of plasma apo A-V with linearity.
Plasma apo A-V concentration has been measured by others with a sandwich ELISA using polyclonal antibodies, which were raised against synthetic peptides of NH2-and COOH-terminus of apo A-V (27). Serum apo A-V concentrations observed in 10 subjects (126.5 ± 86.2 ng/mL, mean ± SD, range 24 -258 ng/mL) were somewhat lower on average than those we have observed in healthy Japanese subjects and apo E, our ELISA appears to have greater sensitivity than that of Obrien et al (27), as they used serum samples with 3-fold dilution, whereas our system uses a 50-fold dilution. In our system, PBS containing 5 g/L CHAPS for sample dilution was used to avoid non-specific binding of antigen to a plate, whereas they used PBS containing 1% Triton X-100 for sample dilution, which may affect the reactivity of mAbs to antigen and induce antigen conformational change.
Many studies have described an association of apo A-V SNPs to plasma triglyceride concentration (12,(20)(21)(22)(23)(24)(25)(26). However, there is no information on the relation of plasma apo A-V level to plasma triglycerides. In the present study, the frequency of the C allele at SNP3 was much higher in Japanese than reported in Caucasians (0.37 vs. 0.08), which is consistent with the observation by Nabika et al (36). In addition, the triglyceride level in subjects with the T/T genotype was significantly lower than that in subjects with the T/C or CC genotype. These results suggest that the SNP3 genotype may influence plasma apo A-V concentration (Table 5). This is the first evidence that plasma apo A-V level is inversely associated with plasma triglycerides in humans. This raises the possibility that plasma apo A-V concentration might influence plasma triglyceride transport in vivo. Evidence has been presented by others that apo A-V both enhances the lipolysis of triglyceride-rich lipoproteins and inhibits VLDL triglyceride secretion by liver (15) We also measured plasma apo A-V concentrations in NIDDM patients without drug therapy. Plasma apo A-V concentration in NIDDM was significantly lower than in healthy subjects. In mice with over-or under-expression of plasma apo A-V, plasma apo A-V concentration was reduced and plasma triglycerides were raised (12 -14), suggesting that apo A-V has a crucial role in plasma triglyceride metabolism. It has been described that the APOA5 gene is regulated by peroxisome proliferator-activated receptor-alpha and franesoid X receptor ligands, both nuclear receptors implicated in triglyceride metabolism (37,38). Furthermore, the expression of apo A-V mRNA was decreased by the administration of liver X receptor (LXRs) ligands, which are known to affect triglyceride metabolism in mice and humans (39), through the activation of sterol regulatory element binding protein (SREBP) 1c (40). The expression of SREBP1 c mRNA was increased by insulin treatment in isolated rat hepatocytes (41,42). In insulin resistance and hyperinsulinemia, SREBP-1c levels are elevated (43). These findings suggest that apo A-V expression might be down-regulated in the presence of insulin resistance. Nowak,et al (44) has recently reported that plasma apo A-V level was decreased by the infusion of insulin. Our results in NIDDM subjects are consistent with those in insulin resistance and hyperinsulinemia. However, large epidemiologic studies will be needed to clearly define the relation of hypertriglyceridemia and CHD risk to apo A-V concentration.           (16/12) 152.4 ± 78.7 c, § 1.46 ± 1.14 b Significant differences from the T/T type at position -1131, a P < 0.05 b P < 0.01, c P < 0.005 and from the C/T type at position -1131, § P < 0.01.