| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Lipid Research, Vol. 48, 1754-1762, August 2007
Copyright © 2007 by American Society for Biochemistry and Molecular Biology
Department of Laboratory Medicine, Division of Clinical Chemistry, Lund University, University Hospital, SE-20502 Malmö, Sweden
Published, JLR Papers in Press, May 24, 2007.
1 To whom correspondence should be addressed. e-mail: bjorn.dahlback{at}med.lu.se
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Supplementary key words retinol • retinoic acid • intrinsic fluorescence • high density lipoprotein • signal peptide • cholesterol
Abbreviations: 1,8-ANS, 1-anilinonaphthalene-8-sulfonic acid; apoM, apolipoprotein M; Kd, dissociation constant; RBP, retinol binding protein
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
A noteworthy feature of the apoM amino acid sequence is the lack of a signal peptidase cleavage site, explaining why circulating apoM contains its signal peptide (1). This unusual property is found in two other HDL-associated proteins: paraoxonase-1 and the haptoglobin-related protein (5). The retained signal peptide of apoM serves as a hydrophobic anchor, binding apoM to the phospholipid layer of the lipoproteins (O. Axler et al., unpublished data).
Structural analysis including homology modeling has predicted apoM to belong to the lipocalin protein family (6). Members of the lipocalin protein family are typically small secreted proteins, which are functionally diverse and have weak sequence similarity but high similarity at the tertiary structural level (7, 8). The majority of lipocalins are responsible for the storage and transport of compounds that have low solubility or are chemically sensitive, such as vitamins, steroids, and metabolic products (7, 8). The lipocalins share a structurally conserved ß-barrel as a central motif, which consists of an eight-stranded antiparallel ß-sheet closed back on itself and a short 
-helix. The ß-barrel encloses a hydrophobic pocket, which in many lipocalins serves as a ligand binding site (7, 8). We previously constructed an apoM model based partly on the structure of retinol binding protein (RBP) and mouse major urinary protein (Fig. 1
) (6). RBP is a 21 kDa lipocalin present in plasma. Its main function is to transport vitamin A (all-trans-retinol). The three-dimensional structure of RBP has been experimentally determined, and the structural details of the retinol binding have been elucidated (9, 10).
|
The aim of this study was to investigate the binding of several small lipophilic ligands to apoM. The predicted lipocalin domain of human apoM was produced in a prokaryotic expression system as well as in stably transfected HEK 293 cells, and intrinsic fluorescence quenching was used to test the binding of a panel of hydrophobic substances to apoM. ApoM was found to bind retinol and retinoic acid, whereas other hydrophobic vitamins, steroid hormones, and fatty acid derivatives did not associate with apoM. These results suggest that apoM is indeed a lipocalin, but the physiological importance of retinol binding remains unknown.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
-tocopherol) were purchased from MP Biomedicals. 1-Anilinonaphthalene-8-sulfonic acid (1,8-ANS) was from Fluka. Acetyl-CoA, vitamin D3 (cholecalciferol), and 3-hydroxy butyric acid were from Sigma-Aldrich. Vitamin K was from Roche. Testosterone was kindly provided by Dr. Pirkko Härkönen (University Hospital Malmö). Human purified RBP was a kind gift from Dr. Bo Åkerström (Lund University). Polyclonal and monoclonal antibodies were generated in house using recombinant human apoM as antigen. Polyclonal antibodies were raised in rabbits and monoclonal antibodies (mab23, mab42, and mab58) were raised in mice using standard techniques (11).
Vector construction for expression of the human apoM lipocalin domain in Escherichia coli
The cDNA encoding the lipocalin domain of apoM (amino acid residues 22–188) was introduced into a pET15b bacterial expression vector (Novagen) using the BamHI and NdeI cleavage sites. The pET15b vector contains the coding sequence for an N-terminal histidine tag followed by a thrombin cleavage site. After the thrombin cleavage site, the vector construct encodes five amino acids, GSHMN, before the start of the apoM sequence. The thrombin cleavage site was thus placed a few amino acids before the start of the apoM sequence, the reason being that we had experienced poor cleavage in other constructs in which the apoM sequence followed directly upon the cleavage site. This is presumably attributable to steric hindrance, as the cysteine at position 23 is involved in a disulfide bridge with cysteine 167 (6). The five extra amino acid residues did not seem to affect the folding, as demonstrated in Results. The BigDye Sequencing Kit (Perkin-Elmer) was used to confirm that the inserted apoM sequence was correct.
Mutagenesis of apoM
ApoM (accession number NM 019101) contains two tryptophans, Trp47 and Trp100, which is why we denote the wild-type variant apoMWW (Fig. 1). Three mutants, Trp47Phe (apoMFW), Trp100Phe (apoMWF), and Trp47Phe, Trp100Phe (apoMFF), were constructed using the Quickchange Site-Directed Mutagenesis Kit (Stratagene). Two complementary oligonucleotides for each mutation were used for this purpose. The sequences for the sense oligonucleotides were 5'-GAGGTCCACTTGGGCCAGTTCTACTTTATCGCAGGG-3' (FW and FF) and 5'-CTCTGTGTGCCCCGGAAATTCATCTACCACCTGACTG-3' (WF and FF). Mutated nucleotides are highlighted by underlined letters. The oligonucleotides were ordered from MWG Biotech. The mutations were verified by DNA sequencing using a BigDye Sequencing Kit (Perkin-Elmer).
Expression and purification of recombinant apoM in E. coli
The pET15b vector containing the apoM sequence was transfected into E. coli strain BL21 (DE3). Cultures were grown under agitation at 37°C in 2 liters of Luria-Bertani medium containing 100 µg/ml ampicillin. Gene expression was induced when the absorbance at 600 nm was 0.6–0.8 by the addition of isopropyl-1-thio-ß-D-galactopyranoside to a final concentration of 1 mM. Expression was allowed for 4 h, and then the cells were harvested by centrifugation and resuspended in ice-cold 20 mM Tris-HCl and 0.15 M NaCl, pH 7.4. Lysozyme (Amersham Biosciences; 100 µg/ml final concentration) and benzamidine (0.5 M) were added to the bacteria, followed by incubation for 45 min at room temperature. The bacteria were sonicated in intermittent 10 s bursts for 30 min until the solution was clear and nonviscous and then centrifuged again. The pellet containing the inclusion bodies was suspended in 30 ml of 6 M guanidine-HCl and 20 mM Tris-HCl, pH 8.0, containing 10 mM reduced glutathione overnight at room temperature and then centrifuged again. The supernatant was collected and applied to a nickel column (20 ml) (nickel resin Superflow; Qiagen) equilibrated with the same buffer. After washing with the equilibration buffer, the apoM was eluted with a linear 0–0.5 M imidazole gradient (260 ml). Fractions containing recombinant apoM were identified by dot-blot analysis using a polyclonal anti-human apoM antiserum.
The fractions with apoM were pooled and refolded by dialysis at 4°C. The pool (diluted to 100 µg/ml in the dialysis buffer) was first dialyzed against 3 M guanidine HCl, 20 mM Tris-HCl (pH 8.0), 4 mM reduced glutathione, 0.4 mM oxidized glutathione, and 10% glycerol for 2–3 days and then against 20 mM Tris-HCl (pH 8.0), 4 mM reduced glutathione, 0.4 mM oxidized glutathione, and 10% glycerol for 1 day. Iodacetamide (5 mM) was added to block free sulfhydryl groups before final dialysis against 20 mM Tris-HCl, pH 8.0, containing 10% glycerol for 1 day. The recombinant apoM was further purified through anion-exchange chromatography on a Q-Sepharose column (Amersham Biosciences). The column (100 ml) was equilibrated in 20 mM Tris-HCl (pH 8.0) and 10% glycerol, and after application of the refolded apoM and washing with the same buffer, the column was eluted with a 500 ml linear NaCl gradient (0–0.3 M). Fractions were analyzed by unreduced 15% PAGE in the presence of SDS, and those containing monomeric apoM were pooled and dialyzed against 20 mM Tris-HCl (pH 8.0) and 10% glycerol overnight at 4°C. The histidine tag was successfully cleaved off by thrombin. After thrombin cleavage, the apoM was repurified on another Q-Sepharose column (25 ml; 400 ml 0–0.3 M NaCl gradient). Finally, the pooled fractions from the second Q-Sepharose purification were applied on a cation-exchange chromatography column (SP-Sepharose; Amersham Biosciences) (2 ml) equilibrated with 20 mM Tris-HCl (pH 8.0) and 10% glycerol to remove the remaining thrombin; thrombin bound to the column, whereas apoM passed through. The apoM concentration was determined by absorption at 280 nm using an extinction coefficient (E1%, 1 cm) of 13.7, which was determined as described (12). The purified apoM was analyzed by 15% SDS-PAGE and by nondenaturing gel electrophoresis (13).
Vector construction for the expression of the human apoM lipocalin domain in a eukaryotic expression system
To express soluble apoM, a signal peptidase cleavage site was generated by mutagenesis of glutamine 22 to alanine (apoMQ22A) in full-length apoM cDNA in pcDNA3 (Invitrogen) (6) using the Quikchange Site-Directed Mutagenesis Kit (Stratagene). The sequence for the sense oligonucleotide (MWG Biotech) was 5'-CCTTAACTCCATCTACGCGTGCCCTGAGCACAGTC-3', with the mutated oligonucleotides being underlined. The presence of the mutation was verified by DNA sequencing using the BigDye Sequencing Kit (Perkin-Elmer).
Expression and purification of apoM produced in HEK 293 cells
The apoMQ22A cDNA was used to transfect HEK 293 cells using Lipofectin (Invitrogen), and stable transfectants were selected using G-418 and tested with the apoM ELISA. Clones secreting high levels of apoM were expanded, and serum-free Optimem medium was collected and stored at –20°C. To purify apoM, ammonium sulfate (70%) was added to the medium (1 liter) and the precipitate was collected by centrifugation at 5,000 rpm for 25 min. The pellet was dissolved in 20 mM Tris, pH 8.0, dialyzed against the same buffer overnight at 4°C, and loaded onto a 5 ml mab23 (9 mg/ml) HiTrap column (Amersham Biosciences) (3). The column was washed with TBS, pH 7.4, and apoM was eluted with 0.1 M glycine, pH 2.2, the pH being immediately neutralized with unbuffered Tris. The purity of the apoM was assessed by 15% SDS-PAGE.
ApoM sandwich ELISA
A sandwich ELISA for apoM based on two monoclonal antibodies, mab58 and mab42, was used to quantify apoM, as described previously (3, 11). In brief, a 96-well Costar plate (Corning, Inc.) was coated with mab58 and quenched according to standard techniques. Serial dilutions of wild-type apoM (apoMWW and apoMQ22A) and apoM mutants and Li-heparin plasma were made in TBS (50 mM Tris-HCl and 0.15 M NaCl, pH 7.4) containing 1% BSA (Sigma-Aldrich) and 1% Triton X-100 (MP Biomedicals) and added to the wells. ApoM bound to mab58 was detected by biotinylated mab42, streptavidin-avidin-horseradish peroxidase (DAKOCytomation A/S), and 1,2-phenylenediamine dihydrocloride tablets (DAKOCytomation A/S), according to the manufacturer's instructions.
Gel filtration of apoM
Recombinant apoM (0.5 ml of 20 µg/ml) in TBS containing 1% BSA (Sigma-Aldrich) was applied to a gel filtration column, Superose 12 10/300 GL (Amersham Biosciences), and equilibrated in TBS. Fractions of 0.27 ml were collected and analyzed using the apoM ELISA. The total protein concentration in the fractions was measured with the Bio-Rad Protein Assay according to the manufacturer's instructions.
Intrinsic fluorescence quenching
The ligand binding experiments were performed at 25 ± 0.1°C with a Fluoromax-3 spectrofluorometer (Jabin Yvon Horiba Group) with slits set at 5 nm bandwidth. The binding studies were performed using excitation at 295 nm. In initial experiments, increasing concentrations of the proteins (in TBS) were tested until the signals for intrinsic fluorescence were high enough. The emission maximum for each protein was determined from the emission spectra received during these titrations. Ligands were added stepwise to a 1 ml apoM protein solution in 0.5–5 µl aliquots from a 1 mM stock solution of the ligand, the maximum volume of ligand added not exceeding 20 µl. The slight volume increase in the course of the titration was neglected. After each added aliquot, the protein sample was mixed and incubated in the dark for 1 min before the fluorescence measurement. In each of the measurements, a blank consisting of the ligand in TBS was subtracted from the actual fluorescence. For each ligand, three to five independent titrations were performed. Binding was evaluated by following the quenching of protein fluorescence attributable to energy transfer. The ligands tested were the following: bilirubin, cholesterol, ß-estradiol, 3-hydroxy butyric acid, oleic acid, palmitic acid, platelet-activating factor, progesterone, prostaglandin E1, all-trans-retinol, all-trans-retinoic acid, 9-cis-retinoic acid, vitamin D3 that had been dissolved in DMSO, arachidonic acid, testosterone, vitamin K that had been dissolved in 99.5% ethanol, vitamin E that had been dissolved in acetone, and acetyl-CoA that had been dissolved in distilled water.
Data analysis of binding
The binding affinity between apoM and the different ligands was calculated by fitting to two different equations, 1 and 2, according to the theory of bimolecular complex formation.
 | (Eq. 1) |
 | (Eq. 2) |
Kd is the apparent dissociation constant for the binding between apoM and the ligand. Kd and IPL were fitted as free parameters. See the supplementary material for the derivation of the equation.
1,8-ANS competition assay
1,8-ANS is a hydrophobic probe that is known to interact with hydrophobic regions in proteins (14–17). The affinity of other ligands for the 1,8-ANS binding site can be assessed by observing the decrease of 1,8-ANS-apoM fluorescence in the presence of increasing concentrations of competitor ligand (16). 1,8-ANS also shows a reversible shift of its own fluorescence emission maximum from 512 to 470 nm when excited at 370 nm after binding to hydrophobic surfaces. 1,8-ANS-bound fluorescence measurements were performed with the ligands binding to apoM in the intrinsic fluorescence measurements. A precomplex was formed by adding 2 µl of a 5 mM stock solution of 1,8-ANS in DMSO to 1 ml of 1.6 µM apoMWW. Fluorescence intensity was measured at 500 nm after excitation at 370 nm. Just as in the intrinsic fluorescence experiments, ligands were added to the protein in 0.5–5 µl aliquots of 1 mM stock solutions to a total of 20 µl. After each added aliquot of ligand, the protein sample was mixed and incubated in the dark for 1 min before the fluorescence measurement. Slit widths of 5 nm were used for both excitation and emission. For each ligand, five independent titrations were performed.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
2 region (data not shown). On gel filtration chromatography, the elution volume of apoMWW was similar to that of RBP, suggesting apoM to be monomeric. ApoMWW was stable upon storage in a refrigerator for weeks. It was highly soluble and could have been concentrated to 30 mg/ml without any visible precipitation.
ApoMQ22A was expressed in HEK 293 cells and used to compare with the prokaryotically expressed apoM. The Q22A mutation created a signal peptidase cleavage site, and stably transfected HEK 293 cells expressed soluble apoM (residues 22–188) lacking the signal peptide at a concentration of 
3 mg/l. The yield of purified apoMQ22A was 200 µg/l Optimem medium. The purified apoMQ22A migrated as two bands on SDS-PAGE under both reducing and nonreducing conditions, which corresponded to glycosylated (20 kDa) and nonglycosylated (15 kDa) apoM (Fig. 2A, B). On gel filtration chromatography, the elution volume of apoMQ22A was similar to that of apoMWW.
Reactivity of apoM variants in a monoclonal antibody-based sandwich ELISA
A sandwich ELISA for apoM based on two monoclonal antibodies (mab42 and mab58) was used to compare the immunoreactivities of the recombinant apoM variants with that of apoM in plasma. On Western blotting, the two monoclonal antibodies only reacted with unreduced apoM (Fig. 2C, D), showing that their epitopes were conformation-dependent. Similar dose-response curves were obtained with the five recombinant apoM variants as with apoM from plasma (Fig. 3
). The ELISA was used to determine the concentration of the mutants, using the apoMWW values to create the standard curve. The reason for this approach was that the 280 nm absorbance is dependent on the number of tryptophans in the molecule. A semiquantitative Western blot using polyclonal antibodies against apoM confirmed that the concentration determination made with the ELISA was correct (data not shown).
|
|
0.1 µM, which corresponds well with previously reported results (21, 22). ApoMWW expressed in E. coli bound all-trans-retinol and its two metabolites in the same manner as apoMQ22A expressed in HEK 293 cells (Fig. 6, Table 1). Also, the two apoM mutants apoMFW and apoMWF bound all-trans-retinol and its two metabolites with the same Kd values as did apoMWW (Table 1). The large variance observed in the experiment using all-trans-retinoic acid was caused by high background fluorescence from the substance itself. This also explains the negative values that were seen in some experiments at high concentrations of all-trans-retinoic acid after subtraction of the background fluorescence from the fluorescence of the apoM-retinoic acid complex (Fig. 6B).
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Among the lipocalins, a tryptophan located at the bottom of the hydrophobic pocket is highly conserved (9, 23). According to homology modeling, the tryptophan at position 47 in apoM corresponds to the conserved tryptophan. In addition, apoM has a tryptophan at position 100 (i.e., at the top of the binding pocket) (Fig. 1) (6). These two tryptophans made it possible to use intrinsic fluorescence quenching to investigate the binding of hydrophobic ligands. Intrinsic fluorescence quenching studies have been used successfully for ligand binding studies for RBP, apoD, and several other lipocalins (21, 22, 24, 25). In this study, RBP was used as a positive control. All-trans-retinol and all-trans-retinoic acid bound to RBP with a Kd of 0.1 µM, which agrees well with previously reported results (21, 22).
Of the many different hydrophobic substances tested, only vitamin A (all-trans-retinol) and its two derivatives, all-trans-retinoic acid and 9-cis-retinoic acid, showed significant binding to apoM, with Kd values of 2–3 µM using this method. The three substances bound to apoMQ22A, apoMWW, and the two apoM mutants, apoMFW and apoMWF, with similar Kd values (Table 1). Together with the observation that the mutations did not affect binding to the monoclonal antibodies, this indicates that the apoM mutants were correctly folded. The binding of the three substances quenched the intrinsic fluorescence of apoMFW and apoMWF to the same extent as was observed for apoMWW. This suggests that both tryptophans are involved in the quenching of the fluorescence in apoM and that the polarity of the environment around both tryptophans is affected by the binding of retinol and retinoic acid. In vivo, apoM is linked to the lipoproteins with the retained signal peptide, and at present we cannot exclude the possibility that the signal peptide affects the ligand binding properties of apoM, even though the lipocalin domain of apoM is an independently folded unit.
In an attempt to evaluate whether retinol and retinoic acid bound in the hydrophobic pocket of apoM, an ANS competition assay was used. 1,8-ANS is a much used "hydrophobic probe" for proteins (e.g., apoD), membranes, etc. It is noteworthy that ANS bound to apoM with much lower affinity than did retinol and retinoic acid. Moreover, in the competition assay, both retinoic acids were only able to partially displace bound 1,8-ANS from apoM. The displacement of 1,8-ANS could be the result of a conformational change in apoM upon binding of retinoic acid or direct competition between the substances. 1,8-ANS is conventionally considered to bind to hydrophobic surfaces through its nonpolar anilinonaphthalene group. However, Matulis et al. (17) showed that 1,8-ANS binding to a number of proteins can occur through electrostatic forces, making it possible for several 1,8-ANS molecules to bind to the same apoM molecule. In conclusion, the experiments using ANS suggested that ANS and the retinoids possibly bind to the same site, the affinity for the retinoids being higher than that for ANS. It is possible that the relatively bulky ANS molecule did not fit in the hydrophobic pocket of apoM, whereas the smaller retinoids could fit better.
Retinoids (vitamin A and its derivatives) exhibit multiple and diverse biological functions and are important for vision, the regulation of gene expression affecting reproduction, hematopoiesis, and growth (10, 26, 27). All-trans-retinol represents the biological nonactive transport form and serves as the precursor of active intracellular retinoids as all-trans-retinoic acid and 9-cis-retinoic acid (28). With the exception of vision, the actions of retinoids are mediated by these two retinoic acid isomers. Several human lipocalins have been shown to bind retinoids (21). ApoD binds all-trans-retinol tightly but also many other substances, including cholesterol, fatty acids, and pheromones (21, 24, 25, 29). RBP is the main transporter of retinol in blood, but there are alternative ways to transport retinoids (30). Dietary retinoids are incorporated into chylomicrons as retinyl esters along with the other dietary lipids in the intestine. The majority of dietary retinoids are cleared by and stored within the liver. The liver secretes retinol into the circulation bound to RBP to meet the vitamin A requirements of tissues. Compared with other tissues, the eye displays a strong preference for retinol uptake when retinol is delivered bound to RBP (28, 31, 32). RBP knockout mice are viable and fertile even though they have a phenotype with several different features. The RBP knockout mice are able to use an alternative pathway to acquire retinol to the eye (28, 30, 31). Whether or not apoM is involved in the transport of retinol as a backup system for RBP remains to be determined.
In conclusion, the binding of retinol and retinoic acid to recombinant apoM supports the hypothesis that apoM is a lipocalin, although the physiological relevance of this binding has yet to be elucidated.
|
ACKNOWLEDGMENTS |
|---|
Manuscript received February 28, 2007 and in revised form May 1, 2007 and in re-revised form May 23, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  J. Ahnstrom, O. Axler, M. Jauhiainen, V. Salomaa, A. S. Havulinna, C. Ehnholm, R. Frikke-Schmidt, A. Tybjaerg-Hansen, and B. Dahlback Levels of apolipoprotein M are not associated with the risk of coronary heart disease in two independent case-control studies J. Lipid Res., September 1, 2008; 49(9): 1912 - 1917. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Christoffersen, J. Ahnstrom, O. Axler, E. I. Christensen, B. Dahlback, and L. B. Nielsen The Signal Peptide Anchors Apolipoprotein M in Plasma Lipoproteins and Prevents Rapid Clearance of Apolipoprotein M from Plasma J. Biol. Chem., July 4, 2008; 283(27): 18765 - 18772. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Christoffersen, M. Jauhiainen, M. Moser, B. Porse, C. Ehnholm, M. Boesl, B. Dahlback, and L. B. Nielsen Effect of Apolipoprotein M on High Density Lipoprotein Metabolism and Atherosclerosis in Low Density Lipoprotein Receptor Knock-out Mice J. Biol. Chem., January 25, 2008; 283(4): 1839 - 1847. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. Axler, J. Ahnstrom, and B. Dahlback An ELISA for apolipoprotein M reveals a strong correlation to total cholesterol in human plasma J. Lipid Res., August 1, 2007; 48(8): 1772 - 1780. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Journal of Biological Chemistry |
| Molecular and Cellular Proteomics | ASBMB Today |
