| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Correspondence to: Guido Franceschini, at the Center E. Grossi Paoletti, Department of Pharmacological Sciences, Via Balzaretti 9 20133, Milano, Italy., Guido.Franceschini{at}unimi.it (E-mail)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The apolipoprotein A-IMilano (apoA-IM) is a molecular variant of apoA-I characterized by the Arg173
Cys substitution, leading to the formation of homodimers A-IM/A-IM. Upon interaction with palmitoyloleoylphosphatidylcholine, A-IM/A-IM forms only two species of reconstituted HDL (rHDL) particles, with diameters of 7.8 and 12.5 nm. We used limited proteolysis to analyze the conformation of A-IM/A-IM in the two rHDL particles, in comparison with that of apoA-I in rHDL of similar size. ApoA-I in the small, 7.8-nm rHDL is degraded to a greater extent (50% after 6 h) than in the large rHDL (<10% degraded after 6 h). The protease susceptibility of A-IM/A-IM in small and large rHDL is instead remarkably the same, with A-IM/A-IM being much more sensitive to proteolytic digestion (50% degraded after 10 min) than apoA-I. The identification of the proteolytic fragments by immunoblotting, N-terminal sequencing, and molecular mass determination, shows that the N-terminus of both proteins is resistant to proteolysis, with six cleavage sites located in the central and carboxy-terminal portions of the molecules. Cleavage in the middle of apoA-I occurs at distinct sites in 7.8-nm (Lys118) and 12.7-nm (Arg123) rHDL, indicating a different conformation in small and large rHDL particles. The A-IM/A-IM instead adopts a unique and identical conformation in small and large rHDL, with the carboxy-terminal portion of the molecule being remarkably more accessible to the proteases than in apoA-I.
This suggests the presence of a novel carboxy-terminal domain in A-IM/A-IM, not organized in a compact structure and not shared by wild-type apoA-I, which may account for the unique functional properties of A-IM/A-IM.—Calabresi, L., G. Tedeschi, C. Treu, S. Ronchi, D. Galbiati, S. Airoldi, C. R. Sirtori, Y. Marcel, and G. Franceschini. Limited proteolysis of a disulfide-linked apoA-I dimer in reconstituted HDL. J. Lipid Res. 2001. 42: 935–942.
Supplementary key words: limited proteolysis, protein conformation, apolipoprotein A-IMilano
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Apolipoprotein A-I (apoA-I) is the major protein constituent of human HDL, and is believed to be responsible for the antiatherogenic and antithrombogenic properties of HDL (1). Indeed, apoA-I overexpression in transgenic mice and rabbits increases the number of circulating HDL particles and protects the animals from the development of diet- or gene-induced atherosclerosis (2) (3) (4). Moreover, adenovirus-mediated transfer of the apoA-I gene into apoE-deficient mice remarkably inhibits neointima formation after endothelial denudation (5), suggesting a direct protective effect of apoA-I on the arterial wall. A number of mechanisms, including facilitation of reverse cholesterol transport (6), prostacyclin stabilization (7), activation of fibrinolysis (8), and modulation of cell-cell interactions (9), have been proposed to explain this protective effect. The structural requirements for these different functions in atherosclerosis protection are mostly unknown.
ApoA-I is synthesized both in the liver and in the intestine as pre-pro-apoA-I, and is then processed to the mature form of 243 residues (10). ApoA-I is predicted to consist of a series of eight 22 amino acids, proline punctuated, amphipathic 
-helices, and two 11 amino acids tandem repeats (11) (12) (13). The recent crystallization of a large fragment of apoA-I lacking the N-terminal 43 residues provided the first direct information on the three-dimensional structure of the lipid-free molecule (14). According to these data, the protein adopts a continuously curved conformation, with a horseshoe-like shape due to the presence of kinks at the site of proline residues (14).
Upon interaction with phospholipids, apoA-I generates a number of discretely sized discoidal particles resembling nascent HDL, also known as reconstituted HDL (rHDL) (15). These rHDL mimic most of the physiological properties of plasma HDL and proved to be a valuable tool in identifying structural requirements for some HDL functions (9) (16) (17) (18). Well-defined rHDL made with apoA-I and synthetic phosphatidylcholines, and with specific ratios between the components, have very reproducible sizes and compositions (15). The rHDL particles can be grouped into classes according to the number of apoA-I molecules per particle, and within each class into particles of different sizes. These subclasses are thought to arise from different apoA-I conformations, reflecting the number of amphipathic helices in contact with lipids on the edge of the disc (19). Two models have been proposed for the structure of apoA-I around the periphery of the rHDL discs. The "belt" model describes apoA-I organization in terms of the X-ray structure, with two molecules of apoA-I forming a pair of continuous amphipathic helices parallel to the plane of the disc (14) (20) (21). In the "picket-fence" model, the amphipathic helices of apoA-I arrange into tightly packed antiparallel segments, oriented perpendicular to the plane of the disc (22) (23).
The apoA-IMilano (apoA-IM) has been the first described molecular variant of human apolipoproteins (24). The apoA-IM differs from wild-type apoA-I for an Arg173Cys substitution, leading to the formation of disulfide-linked dimers (25). Although the structure of the lipid-free and lipid-bound apoA-IM monomer differs only slightly from wild-type apoA-I (26), the introduction of an interchain disulfide bridge in the apoA-IM homodimer (A-IM/A-IM) remarkably alters the physico-chemical properties of apoA-I (27). When compared with apoA-I, the lipid-free A-IM/A-IM displays a higher -helical content and a more folded tertiary structure (27). Upon interaction with phospholipids, A-IM/A-IM forms only two species of rHDL with a diameter of 7.8 and 12.5 nm, which contain one or two A-IM/A-IM molecules per particle, respectively (28). Under the same experimental conditions, apoA-I forms predominantly rHDL particles with a diameter of 9.6 nm, from which a 7.8-nm rHDL is derived (28), and other rHDL subspecies, including a 12.7-nm rHDL (28). When incorporated in these two rHDL, the A-IM/A-IM proved more effective than apoA-I in promoting cell cholesterol efflux (17) and less capable of activating the lecithin:cholesterol acyltransferase enzyme (16). In the present studies, we used limited proteolysis as a tool for investigating the conformation of A-IM/A-IM and apoA-I in the same small and large rHDL particles.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Materials
A-IM/A-IM was expressed in the E. Coli and purified by conventional chromatographic procedures (27). The A-IM/A-IM batch used in the present study contained 
98% of A-IM/A-IM, as determined by SDS-PAGE (not shown). Normal apoA-I was purified from human blood plasma, as previously described (26). Before use, lyophilized A-IM/A-IM and apoA-I were dissolved in 20 mM phosphate buffer, pH 7.4, containing 6 M Gdn-HCl and extensively dialyzed against 10 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1 mM NaN3, and 0.01% EDTA (reconstitution buffer). Protein concentration of the stock solutions was assayed by amino acid analysis, performed on a Beckman 6300 amino acid analyzer after acidic hydrolysis of samples in 6 M HCl for 45 min at 155°C. L--palmitoyloleoylphosphatidylcholine (POPC), sodium cholate, elastase, and chymotrypsin were purchased from Sigma (St. Louis, MO). Trypsin was purchased from Boehringer (Indianapolis, IN).
Preparation and characterization of rHDL
Discoidal rHDL containing A-IM/A-IM or apoA-I and POPC, with a diameter of 7.8 or 12.5;–12.7 nm, were prepared by the cholate dialysis technique (29), as previously described (28). The size of the particles was estimated by nondenaturing gradient gel electrophoresis (GGE) (28), using the Pharmacia Phast System (Pharmacia Biotech). The number of apolipoprotein molecules per rHDL particle was determined by cross-linking with dimethylsuberimidate (DMS) (27). Cross-linked samples were analyzed by SDS-PAGE on 4;–10% acrylamide gradient slab gels, using the Tris-tricine buffer system of Schagger & Van Jagou (30) in a Mini Protean slab minigel apparatus (BioRad). Phospholipid content of rHDL was determined by an enzymatic method (31). Proteins were measured by the method of Lowry et al. (32), using bovine serum albumin as standard.
Circular dichroism (CD) spectra were recorded with a Jasco J500A spectropolarimeter at the constant temperature of 25°C. Molar mean residue ellipticity (
) was expressed in degrees · cm2 · dmol-1, and calculated as:
where obs is the observed ellipticity in degrees, 115 is the mean residue molecular weight of the proteins, l is the optical path length in centimeters, and c is the protein concentration in grams/milliliter. All the spectra were baseline corrected. The -helical content was calculated by the method of Chang, Wu, and Yang (33).
Limited proteolysis experiments
Reconstituted HDL containing A-IM/A-IM or apoA-I (about 0.1 mg/ml) were mixed with trypsin at the ratio of 1:50 (w/w) and incubated at 37°C. At various time intervals, aliquots were removed, SDS was added to obtain a final concentration of 4%, and samples were boiled for 2 min to terminate the reaction. The digestion products were stored at - 20°C until analyzed by electrophoresis.
Analysis of proteolytic products
The digestion products were analyzed by SDS-PAGE on 10;–16% acrylamide gradient slab gels, using the Tris-tricine buffer system (30). After electrophoresis, the peptide bands in the gels were either visualized with Coomassie Blue or electrophoretically transferred to polyvinylidene fluoride membranes. The membranes were then stained with Coomassie Blue, and the bands of interest were carefully cut from the membrane and subjected to automated sequence analysis on a pulsed-liquid sequencer (Applied Biosystems, Foster City, CA) equipped with a 120A Applied Biosystems PTH-analyzer.
The molecular weight of the trypsin fragments was determined by flight-mass spectrometry (MALDI/TOF-MS) using a Vestec Lasertec MALDI-TOF instrument (Perspective, Freburg, Germany) operating in a linear mode. Ions formed by a pulse ultraviolet laser beam (nitrogen laser, 
= 337 nm) were accelerated through 28 kV. The instrument was calibrated with bovine heart cytochrome c (MW = 12,327) and flagellin from Bacillus subtilis (MW = 32,626). The calibration proteins and the samples were dissolved in a solution of ferulic acid in 50% acetonitrile:50% water with trifluoroacetic acid 0.1%, to a final concentration of 50 pmoles/µl.
Western blotting
Some monoclonal antibodies (mAbs) used in these studies were a generous gift of Dr. G. R. Castro (A05, A17, A03, and A44), and Drs. S. Marcovina and A. L. Catapano (A-I-11 and A-I-57); all mAbs have been previously described and characterized (34) (35). After trypsin digestion, the proteolytic fragments were separated by SDS-PAGE on 10;–16% acrylamide gradient slab gels, and then electrophoretically transferred to membranes. After electroblotting, the membrane was incubated with albumin to prevent nonspecific binding. After blocking, the membrane was incubated with the primary antibody (mouse mAb) at 37°C for 1 h, washed, and incubated with the secondary antibody (goat anti-mouse IgG coupled to horseradish peroxidase). The immunocomplexes were detected using an enhanced chemiluminescence kit.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Upon interaction with POPC, A-IM/A-IM forms only two species of rHDL particles, with diameters of 7.8 and 12.5 nm (28). In the present studies we investigated the conformation of A-IM/A-IM in these two particles, in comparison with that of apoA-I in rHDL of comparable sizes. The rHDL preparations used here have the same characteristics of those used in the previous functional studies (16) (17). In particular, the 7.8- and 12.5-nm A-IM/A-IM rHDL contain one and two molecules of the dimer per particle, and display the same -helical content of 62%. The 7.8- and 12.7-nm apoA-I rHDL contain two and three apoA-I molecules per particle, with an -helical content of 61% and 71%.
Limited proteolysis of A-IM/A-IM and apoA-I rHDL
Limited proteolysis was used to investigate the conformation of A-IM/A-IM and apoA-I in rHDL particles. In initial experiments, each of the four rHDL was treated with proteolytic enzymes of differing specificities, chymotrypsin, trypsin, and elastase. The three proteases gave similar cleavage patterns (data not shown), i.e., consistent with previous findings (36) (37). Trypsin was used in further experiments.
The rHDL were subjected to tryptic digestion for various lengths of time, and the degree of protein degradation was estimated from the densitometry of SDS-PAGE gels ( Fig 1). Since equal amounts of protein were analyzed, the decrease in the intensity of the band corresponding to the intact protein truly reflects degradation of this protein. The extent of hydrolysis of apoA-I varies between small and large rHDL, such that apoA-I in the 7.8-nm rHDL is digested to a greater extent than apoA-I in the 12.7-nm rHDL ( Fig 2). ApoA-I in the small rHDL is degraded by 50% after 6 h of proteolysis, with digestion being detectable at the earliest time points (Fig 2). ApoA-I in the large rHDL is instead resistant to proteolytic degradation, with 90% of the protein remaining intact after prolonged digestion (Fig 2).
|
|
Different from apoA-I, the protease susceptibility of A-IM/A-IM in 7.8- and 12.5-nm rHDL particles is remarkably similar (Fig 1), suggesting that A-IM/A-IM adopts a similar conformation in small and large rHDL. This result agrees with our previous observations that A-IM/A-IM has the same secondary and tertiary structure when incorporated into small or large rHDL (28). It is noteworthy that A-IM/A-IM is definitely more susceptible to proteolytic digestion than apoA-I, as indicated by its pronounced degradation at shorter times. Fifty percent of A-IM/A-IM is already degraded after 10 min of proteolysis, with only 10% of the protein remaining intact after 6 h of incubation (Fig 2). By contrast, the protease susceptibility of the monomeric A-IM in rHDL is remarkably similar to that of apoA-I (not shown).
Particle integrity over the time course of the proteolysis experiment was assessed by nondenaturing gradient gel electrophoresis. No changes in rHDL particle size are detected in control incubations without trypsin. Both apoA-I rHDL particles are stable over the 6 h of proteolysis. A-IM/A-IM rHDL do not change in size during the first 30 min of incubation, despite a significant protein degradation (40% of A-IM/A-IM remaining intact); the particles become more heterogeneous at 2 h, and some protein dissociates from rHDL after 6 h of proteolysis.
Identification of cleavage sites
At the same time intervals that protein degradation was estimated, the cleavage products generated by trypsin digestion were separated by SDS-PAGE (Fig 1). The identification of the cleavage sites in apoA-I and of A-IM/A-IM was achieved by i) immunoblotting of SDS-PAGE gels with a panel of mAbs directed against epitopes distributed all along the apoA-I sequence ( Fig 3 and Fig 4), ii) N-terminal sequencing of proteolytic fragments isolated by SDS-PAGE (Table 1 and Table 2), and iii) determination of the molecular mass of the digestion products by mass spectrometry (Table 1 and Table 2).
|
|
|
|
Treatment of 7.8-nm apoA-I rHDL with trypsin yields six additional bands by SDS-PAGE, corresponding to polypeptides with molecular masses ranging between 24.9 and 13.2 kDa (Fig 1, Table 2). All these products are recognized by mAbs 4H1 and AO5 (Fig 4), which bind to the apoA-I sequences Glu2-Trp8 and Ser25-Leu82 (34) (Fig 3), indicating that all peptides contain the apoA-I N-terminus, as proved by N-terminal sequencing (Table 1). Therefore, no cleavage occurs in the amino-terminal region of apoA-I, when the protein is bound to POPC in rHDL. Bands 2 to 4 on SDS-PAGE react with all the tested mAbs (Fig 4), suggesting that they correspond to peptides lacking different portions of the C-terminus, which is not recognized by any antibody. The molecular masses of these large fragments, and of the small complementary peptides, are consistent with cleavage at residues Arg215, Lys195, and Arg188 (Table 1). Band 5 reacts with mAbs mapping to the amino-terminal and central regions of apoA-I, and it corresponds to the peptide cleaved at Arg160 (Table 1). Band 6 reacts with mAbs recognizing the N-terminus of apoA-I (4H1, A05 and A-I-11) (Fig 4). Molecular mass and sequence analysis indicate this band indeed includes two distinct peptides, of 14.1 and 13.9 kDa, with N-terminal sequences starting at Val119 and Asp1, respectively (Table 1). Therefore, band 6 contains the complementary peptides generated by cleavage of apoA-I at Lys118, which cannot be solved owing to insufficient resolution by SDS-PAGE. The Val119-Gln243 fragment may be generated early during digestion, being further degraded toward its carboxy-terminal end, with formation of smaller fragments. Indeed, immunoreactivity of band 6 with mAbs mapped to the central portion of apoA-I (A44 and A-I-57) is observed at the earliest (not shown) but not latest (Fig 4) time points after the addition of protease.
All together, immunoblotting, N-terminal sequencing, and mass spectrometry show that apoA-I in the 7.8-nm rHDL is cleaved by trypsin at Lys94, Lys118, Arg160, Arg188, Lys195, and Arg215 (Table 1, Fig 5). The Asp1-Arg215, Asp1-Lys195, and Asp1-Lys118 peptides (bands 2, 3, and 6, respectively, in Fig 1) are formed immediately after the digestion starts; the three other fragments, Asp1-Arg188, Asp1-Arg160, and Asp1-Lys94 (bands 4, 5, and 7, respectively, in Fig 1), appear later during digestion, indicating a lower exposure of the corresponding cleavage sites to the protease or a formation secondary to first cleavages. Although the proteolysis experiments were not designed to evaluate the kinetics of the degradation process, the identification of earlier and later cleavage sites is consistent with the apparently multiphases decay of the intact protein (Fig 2).
|
As indicated above, apoA-I in the 12.7-nm rHDL is much more resistant to trypsin digestion than in the small rHDL. A clear-cut visualization of the proteolytic products can be achieved only after prolonged digestion (24 h, not shown), when seven bands can be identified by SDS-PAGE. The proteolytic products have been characterized by N-terminal sequencing and mass spectrometry. Five of the generated peptides (bands 2;–5 and 7) are identical to those obtained after digestion of 7.8-nm apoA-I rHDL. By contrast, band 6 includes two peptides that are different from those found in degraded 7.8-nm apoA-I rHDL; these proteolytic products, with masses of 14.5 and 13.6 kDa, correspond to the Asp1-Arg123 and the complementary Ala124-Gln243 peptides (Fig 5). Therefore, cleavage in the central portion of apoA-I occurs at distinct sites in small (Lys118) and large (Arg123) rHDL particles.
Trypsin cleavage of A-IM/A-IM in 7.8- and 12.5-nm rHDL particles gives the same eight additional bands by SDS-PAGE, corresponding to proteolytic fragments ranging in molecular mass from 49.7 to 11.0 kDa (Fig 1, Table 2). The band with an approximate molecular mass of 19 kDa (band 7) indeed includes two distinct products, with masses of 18.8 and 18.5 kDa, as determined by N-terminal sequencing (Table 2). Therefore, nine large peptides are generated by digestion of A-IM/A-IM with trypsin. Bands 2;–5 on SDS-PAGE react with all the mAbs (Fig 4), suggesting that they correspond to peptides lacking different portions of the carboxy-terminal region of the protein. Three of these peptides, of 49.7 kDa (band 2), 45.3 kDa (band 3), and 43.7 kDa (band 4), contain an intact N-terminus and the interchain disulfide bridge at Cys173; the molecular masses of these fragments, and of the small complementary peptides, are consistent with cleavage at residues Arg215, Lys195 and Arg188 (Table 2). The 43.7 homodimer (band 4) is an intermediate, and is further cleaved at Lys94 of one chain to generate the 32.8-kDa heterodimer (band 5) and the complementary Asp1-Lys94 peptide (band 9) (Table 2). Band 6 corresponds to a disulfide-linked peptide with molecular mass of 27.3 kDa and the Ala124-Gln243 sequence (Table 2); this peptide, therefore, should not react with mAbs recognizing the N-terminus (4H1 and AO5). Some immunoreactivity is, however, found on immunoblotting (Fig 4), possibly due to the co-migrating A-IMilano monomer, which is a contaminant (<2%) of the A-IM/A-IM preparation; some intact N-terminus is indeed detected by sequencing. Band 7 is recognized by all the mAbs and, indeed, consists of two peptides with N-terminal sequences starting at Asp1 and Thr161 (Table 2). Therefore, cleavage in the central portion of A-IM/A-IM, at Arg123 and Arg160, gives two monomeric peptides of 18.8 kDa (band 7) and 14.5 kDa (band 8), containing the intact N-terminus, together with the complementary homodimers of 27.3 kDa (band 6) and 18.5 kDa (band 7). Five primary cleavage sites, located at Arg123, Arg160, Arg188, Lys195, and Arg215, and one secondary cleavage site at Lys94, are thus present in the lipid-bound A-IM/A-IM (Fig 5). As observed for apoA-I, the presence of earlier and later cleavage sites is consistent with the multiphases kinetics of intact protein degradation (Fig 2).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our previous studies (28) have shown that, upon interaction with phospholipids, A-IM/A-IM forms only two species of rHDL, with diameters of 7.8 and 12.5 nm, containing one or two A-IM/A-IM molecules per particle, respectively, whereas apoA-I forms rHDL particles of increasing size with increasing lipid:protein ratio in the incubation mixture. We hypothesized that the introduction of an interchain disulfide bridge in the A-IM/A-IM protein restricts rHDL particle size heterogeneity by restraining the intrinsic structural adaptability of the wild-type monomeric apoA-I. In the present studies, we have relied on limited proteolysis to probe the conformation of A-IM/A-IM on the surface of the two rHDL particles, in comparison with that of wild-type apoA-I in rHDL of comparable size. Limited proteolysis has been widely used to study the domain structure of proteins (38), due to the correspondence between accessibility to the protease and the lack of an organized secondary structure. Proteolytic digestion of a globular protein is expected to occur at surface loops and random segments, or at flexible hinges between protein domains, rather than inside rigid elements of secondary structure, such as helices or pleated sheets (38). The present work demonstrates that (i) apoA-I in small rHDL is remarkably more sensitive to proteolytic digestion than in large rHDL; (ii) A-IM/A-IM in small and large rHDL shows the same susceptibility to proteolysis, being much more sensitive to digestion than apoA-I; and (iii) the amino-terminal region of both apoA-I and A-IM/A-IM is resistant to proteolysis, with the cleavage sites located in the central and carboxy-terminal portions of the molecules.
ApoA-I in small rHDL was definitely more susceptible to proteolytic digestion than in large rHDL. The observation of a distinct susceptibility to proteases of apoA-I in rHDL of different size is in agreement with previous results of Dalton and Swaney (36), who observed an inverse relationship between the size of rHDL and the extent of apoA-I degradation, and with the more rapid denaturation of apoA-I in small than large particles (39). Furthermore, Kunitake et al. (40) showed that the apoA-I of plasma-derived pre-ß-migrating HDL, somewhat comparable to the present 7.8-nm rHDL particles in terms of size and surface charge (indeed, all the rHDL used in this study migrate in pre-ß position), displays an enhanced sensitivity to proteolysis than apoA-I of -migrating, larger HDL. All together, these findings fully support the concept, originally developed from spectroscopic (22) and epitope mapping (19) studies, of a distinct conformation of apoA-I in rHDL of different size that contain the same number of apoA-I molecules.
The change in conformation caused by variations in apoA-I rHDL size seems to involve a hinged domain, that is, a mobile portion of the molecule constituted by a pair of consecutive amphipathic helices, which can exist either bound to lipids in the large rHDL or looping out of the small rHDL into the aqueous milieu (19). This concept, originally developed on the basis of the picket-fence model of apoA-I structural organization in rHDL (19), may also adapt to the alternative belt model (20) (21). It is reasonable that, once dissociated from the lipid core of the small rHDL, this hinged domain loses its helical structure, as indicated by the reduced -helical content of small than large rHDL particles (22) (28). Two regions, encompassing the helices 4 and 5 (residues 99;–143), or 5 and 6 (residues 121;–165) of apoA-I (11), have been proposed to form the hinge (19) (41), but its exact position has never been clearly demonstrated. The present work has identified distinct cleavage sites at Lys118 and Arg123 in small and large apoA-I rHDL, respectively, indicating a different conformation of this portion of apoA-I in the two rHDL particles, therefore supporting the concept that is indeed the 99;–143 region that forms the hinge.
Although 7.8- and 12.7-nm apoA-I rHDL display a distinct susceptibility to proteolytic digestion, A-IM/A-IM shows the same sensitivity to proteases when bound to small or large rHDL particles. This is in agreement with the almost identical spectral properties of A-IM/A-IM in the same small and large rHDL (28), and indicates that A-IM/A-IM has a limited structural plasticity compared with apoA-I. The generation of apoA-I rHDL particles of different sizes and containing a constant number of apoA-I molecules per particle has been explained by the conformational flexibility of the hinged domain in the central portion of apoA-I (19). In the case of both small and large A-IM/A-IM rHDL, cleavage in the central portion of the A-IM/A-IM molecule occurs at the same position as in the 12.7-nm apoA-I rHDL, where the hinged domain is in the lipid-bound conformation (19). This indicates that also in the two A-IM/A-IM rHDL, the hinged domain is in contact with the lipid bilayer and assumes helical structure. Because the two A-IM/A-IM rHDL display an -helical content similar to the 7.8-nm apoA-I rHDL and lower than the 12.7-nm apoA-I rHDL (28), a portion of A-IM/A-IM, distinct from the hinged domain, is not in contact with lipids and loses the -helical structure.
The present results can be adapted to both models proposed for the structure of apoA-I around the periphery of the rHDL. The belt model for rHDL discs with a diameter of 10;–10.5 nm and containing two apoA-I molecules per particle, describes apoA-I organization in terms of the X-ray structure (14), with apoA-I forming a pair of continuous amphipathic helices parallel to the plane of the disc (21). In the picket-fence model for rHDL containing two or three apoA-I molecules per particle, the amphipathic apoA-I helices arrange into tightly packed antiparallel segments, oriented perpendicular to the plane of the disc (22) (23). Both picket-fence and belt models for A-IM/A-IM rHDL have been recently proposed (1) (42). The present experiments were not designed to distinguish between the two models, and the results do not favor either hypothesis. It is noteworthy that, although the picket-fence model (1) is easily adaptable to the small and large rHDL containing one and two A-IM/A-IM molecules per particle, the proposed belt model of a 9;–10-nm A-IM/A-IM rHDL (42) appears incompatible with the experimentally generated rHDL sizes (28). Whether it is organized in the picket-fence or belt manner, the A-IM/A-IM adopts a conformation that is different from that of apoA-I, with the C-terminal portion being more susceptible to proteolytic degradation. It seems reasonable to speculate that the disulfide bridge in A-IM/A-IM sets a structural constraint that forcedly displaces out of the discs the C-terminal residues of A-IM/A-IM, starting approximately at Ala187 (11). This would form a distinctly folded domain, not detectable in apoA-I, that, not involved in lipid-binding, would rapidly lose its helical structure, i.e., consistent with the circular dichroism data (28).
Cleavage sites in apoA-I and A-IM/A-IM were located in the central and carboxy-terminal portions of the protein with no cleavage occurring in the amino-terminal domain. These results are consistent with those of previous studies on apoA-I proteolysis in homogeneous, POPC-containing rHDL (36) (37), but differ from those of Lins et al. (43), who showed cleavage by trypsin and other proteases at residues 43;–48 of apoA-I in DMPC-containing complexes. Roberts et al. (44) recently reported that the oxidative state of apoA-I is a major determinant of protein conformation and protease susceptibility: oxidation of two methionines at residues 112 and 148 in the primary sequence switches the protease cleavage sites from the N-terminus to the central and carboxy-terminal regions of the protein. In the present study, the lack of cleavage in the amino-terminal region of both apoA-I and A-IM/A-IM is not due to protein oxidation, as demonstrated by determination of protein masses by mass spectrometry (apoA-I: 28,078 Da; A-IM/A-IM: 56,050 Da). These somewhat conflicting results might be due to differences in the size and composition of rHDL, in the rHDL purification procedure used, and/or in the proteolytic conditions.
In conclusion, the introduction of an interchain disulfide bridge in apoA-I remarkably affects the conformation of the protein when bound to lipids, with generation of a unique C-terminal domain. The specific conformation of the lipid-bound A-IM/A-IM would account for the delayed in vivo catabolism (45), and for the peculiar properties of this protein in lipoprotein metabolism (16) (17) (46) and atherosclerosis protection (1).
|
ACKNOWLEDGMENTS |
|---|
We are most grateful to Dr. H. Ageland (Pharmacia, Sweden) for providing the recombinant A-IM/A-IM. We thank Drs. G. R. Castro, S. Marcovina and A. L. Catapano for providing the monoclonal antibodies.
Manuscript received July 31, 2000; and in revised form January 3, 2001
Abbreviations: ApoA-I, apolipoprotein A-I; rHDL, reconstituted HDL; apoA-IM, apolipoprotein A-IMilano; A-IM/A-IM, A-IM homodimer; POPC, palmitoyloleoylphosphatidylcholine; GGE, gradient gel electrophoresis; DMS, dimethylsuberimidate; CD, circular dichroism; mAbs, monoclonal antibodies
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  E. Favari, M. Gomaraschi, I. Zanotti, F. Bernini, M. Lee-Rueckert, P. T. Kovanen, C. R. Sirtori, G. Franceschini, and L. Calabresi A Unique Protease-sensitive High Density Lipoprotein Particle Containing the Apolipoprotein A-IMilano Dimer Effectively Promotes ATP-binding Cassette A1-mediated Cell Cholesterol Efflux J. Biol. Chem., February 23, 2007; 282(8): 5125 - 5132. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. G. Rocco, L. Mollica, E. Gianazza, L. Calabresi, G. Franceschini, C. R. Sirtori, and I. Eberini A Model Structure for the Heterodimer apoA-IMilano-apoA-II Supports Its Peculiar Susceptibility to Proteolysis Biophys. J., October 15, 2006; 91(8): 3043 - 3049. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Wang, B. G. Sharifi, T. Pan, L. Song, A. Yukht, and P. K. Shah Bone Marrow Transplantation Shows Superior Atheroprotective Effects of Gene Therapy With Apolipoprotein A-I Milano Compared With Wild-Type Apolipoprotein A-I in Hyperlipidemic Mice J. Am. Coll. Cardiol., October 3, 2006; 48(7): 1459 - 1468. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Lee, P. T. Kovanen, G. Tedeschi, E. Oungre, G. Franceschini, and L. Calabresi Apolipoprotein composition and particle size affect HDL degradation by chymase: effect on cellular cholesterol efflux J. Lipid Res., March 1, 2003; 44(3): 539 - 546. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. G. Cole, W. L. Nowatzke, C. L. Bisgaier, and B. R. Krause Method-dependent Changes in ""HDL-Cholesterol"" with Recombinant Apolipoprotein A-IMilano Infusion in Healthy Volunteers Clin. Chem., April 1, 2002; 48(4): 680 - 681. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE |
