Rapid regulation of apolipoprotein A-I secretion in HepG2 cells by a factor associated with bovine high-density lipoproteins

The effect of fetal bovine serum (FBS) on the secretion of apolipoprotein A-I (apo A-I) by HepG2 cells was studied. The cells incubated with FBS always secreted more apo A-I than the cells incubated with serum-free medium. The changes in the rate of apo A-I secretion were observed within 1 h after addition or depletion of serum. The high-density lipoproteins (HDL) or the lipoprotein-deficient serum (LPDS) obtained from FBS also stimulated apo A-I secretion rapidly to the same level as obtained with FBS. Addition of low-density lipoproteins did not have any effect. The rate of general protein synthesis was not affected by short-term incubations with or without serum or HDL. The rate of apolipoprotein E secretion by these cells did not change significantly, parallel to the changes in apo A-I secretion in the presence or absence of FBS. It is concluded that serum may have a factor that plays a specific role in the regulation of apo A-I secretion by the liver cells and this factor is associated with the HDL fraction.     

Remodeling of human plasma lipoproteins by detergent perturbation

Detergent perturbation, the treatment of total human plasma lipoproteins (TLP) with sodium cholate and its subsequent removal, has been used to study lipoprotein dynamics and stability. At physiological TLP concentrations, detergent perturbation converts low-density lipoproteins (LDL) and high-density lipoproteins (HDL) to higher-particle weight species with the concomitant release of apo A-I but not apo A-II as a lipid-poor species. Detergent perturbation of isolated HDL also releases lipid-poor apo A-I and forms larger HDL species, whereas detergent perturbation of an isolated LDL has no effect on its size. A model is presented in which detergent perturbation induces transfer of PC from metastable HDL and LDL to mixed micelles with sodium cholate. The remaining LDL and HDL are unstable because of the loss of their surface components, phospholipid and/or apo A-I, and fuse to give larger LDL and HDL particles. These effects on HDL, i.e., PC transfer, apo A-I dissociation, and particle fusion, emulate the activity of human plasma phospholipid transfer protein. Thus, detergent perturbation is a new and potentially powerful method for determining lipoprotein stability, studying the mechanisms for remodeling of plasma lipoproteins, and preparing new forms of HDL and LDL with unique interactions with lipoprotein transporters and receptors.     

Evidence in vitro that hepatic lipase reduces the concentration of apolipoprotein A-I in rabbit high-density lipoproteins

Incubation of rabbit plasma in vitro with hepatic lipase resulted in the hydrolysis of triacylglycerol in high-density lipoproteins (HDL) and a reduction in HDL particle size. These changes were accompanied by a decrease in the concentration of apolipoprotein A-I (apo A-I) in the HDL. The loss of apo A-I was demonstrated independently by ultracentrifugation, size exclusion chromatography and gradient gel-immunoblot analysis. It was unrelated to hydrolysis of HDL phospholipids but did correlate with the reduction in HDL particle size. These studies suggest that the concentration of apo A-I in HDL may be influenced by factors which regulate the metabolism of HDL core lipid constituents.     

Characterization of Na+-dependent phosphate uptake in cultured kidney cells (JTC-12) from monkey.

O-(4-Diazo-3-[125I]iodobenzoyl)sucrose ([125I]DIBS), a novel labelling compound specifically designed to study the catabolic sites of serum proteins [De Jong, Bouma, & Gruber (1981) Biochem. J. 198, 45-51], was applied to study the tissue sites of degradation of serum lipoproteins. [125I]DIBS-labelled apolipoproteins (apo) E and A-I, added in tracer amounts to rat serum, associate with high-density lipoproteins (HDL) just like conventionally iodinated apo E and A-I. No difference is observed between the serum decays of chromatographically isolated [125I]DIBS-labelled and conventionally iodinated HDL labelled specifically in either apo E or apo A-I. When these specifically labelled HDLs are injected into fasted rats, a substantial [125I]DIBS-dependent 125I accumulation occurs in the kidneys and in the liver. No [125I]DIBS-dependent accumulation is observed in the kidneys after injection of labelled asialofetuin or human low-density lipoprotein. It is concluded that the kidneys and the liver are important sites of catabolism of rat HDL apo E and A-I.     

Apolipoprotein A-I and apolipoprotein SAA half-lives during acute inflammation and amyloidogenesis

The half-lives of two apolipoproteins of mouse high-density lipoproteins, apo A-I and apo SAA, were determined in normal animals and compared with those having an inflammatory condition, or inflammation leading to AA amyloid deposition. The apo A-I half-life was considerably shorter in animals with inflammation and in those that were preamyloidotic, than in controls (t1/2 of 3-3.5 h vs. 10-12 h). The average loss of apo A-I in controls over the first 10 h was 31.1 micrograms/ml per h, while that in inflamed animals was 58.7 micrograms/ml per h a 2-fold increase in apo A-I clearance. The apo SAA half-life was similar in all groups of animals and was of the order of 1.5 h. The concentration of apo SAA during inflammation is however considerably higher (500-1000-fold) than in controls, which implies a much greater clearance rate during inflammation and involving a process which is apparently not saturable. In addition to hypertriglyceridemia and Tangier's disease, ordinary acute inflammation can now be added to those pathological conditions which lead to a significant decrease in apo A-I half-life.     

Mechanism of dissociation of human apolipoprotein A-I from complexes with dimyristoylphosphatidylcholine as studied by guanidine hydrochloride denaturation

The reversibility of the binding of human apolipoprotein A-I (apo A-I) to phospholipid has been monitored through the influence of guanidine hydrochloride (Gdn-HCl) on the isothermal denaturation and renaturation of apo A-1/dimyristoylphosphatidylcholine (DMPC) complexes at 24 degree C. Denaturation was studied by incubating discoidal 1:100 and vesicular 1:500 mol/mol apo A-I/DMPC complexes with up to 7 M Gdn-HCl for up to 72 h. Unfolding of apo A-I molecules was observed from circular dichroism spectra while the distribution of protein between free and lipid-associated states was monitored by density gradient ultracentrifugation. The ability of apo A-I to combine with DMPC in the presence of Gdn-HCl at 24 degrees C was also investigated by similar procedures. In both the denaturation and renaturation of 1:100 and 1:500 complexes, the final values of the molar ellipticity and the ratio of free to bound apo A-I at various concentrations of Gdn-HCl are dependent on the initial state of the lipid and protein; apo A-I is more resistant to denaturation when Gdn-HCl is added to existing complexes than to a mixture of apo A-I and DMPC. There is an intermediate state in the denaturation pathway of apo A-I/DMPC complexes which is not present in the renaturation; the intermediate comprises partially unfold apo A-I molecules still associated with the complex by some of their apolar residues. Complete unfolding of the alpha helix and subsequent desorption of the apo A-I molecules from the lipid/water interface involve cooperative exposure of these apolar residues to the aqueous phase. The energy barrier associated with this desorption step makes the binding of apo A-I to DMPC a thermodynamically irreversible process. Consequently, binding constants of apo A-I and PC cannot be calculated simply from equilibrium thermodynamic treatments of the partitioning of protein between free and bound states. Apo A-I molecules do not exchange freely between the lipid-free and lipid-bound states, and extra work is required to drive protein molecules off the surface. The required increased in surface pressure can be achieved by a net mass transfer of protein to the surface; in vivo, increases in the surface pressure of lipoproteins by lipolysis can cause protein desorption.     

Binding of human apolipoprotein A-IV to human hepatocellular plasma membranes

We have investigated the binding of human apolipoprotein A-IV (apo A-IV) to human hepatocellular plasma membranes. Addition of increasing concentrations of radiolabeled apo A-IV to hepatic plasma membranes, in the presence and absence of a 25-fold excess of unlabeled apo A-IV, revealed saturation binding to the membranes with a KD of 154 nM and a binding maximum of 1.6 ng/microgram of membrane protein. The binding was temperature-insensitive, partially calcium-dependent, abolished when apo A-IV was denatured by guanidine hydrochloride or when the membranes were treated with Pronase and decreased when apo A-IV was incorporated into phospholipid/cholesterol proteoliposomes. In displacement studies using purified apolipoproteins and isolated lipoproteins, only unlabeled apo A-IV, apo A-I and high-density lipoproteins effectively competed with radiolabeled apo A-IV for membrane binding sites. We conclude that human apo A-IV exhibits high-affinity binding to isolated human hepatocellular plasma membranes which is saturable, reversible and specific.     

Characterization of human high-density lipoprotein subclasses LP A-I and LP A-I/A-II and binding to HepG2 cells

Plasma HDL can be classified according to their apolipoprotein content into at least two types of lipoprotein particles: lipoproteins containing both apo A-I and apo A-II (LP A-I/A-II) and lipoproteins with apo A-I but without apo A-II (LP A-I). LP A-I and LP A-I/A-II were isolated by immuno-affinity chromatography. LP A-I has a higher cholesterol content and less protein compared to LP A-I/A-II. The average particle mass of LP A-I is higher (379 kDa) than the average particle weight of LP A-I/A-II (269 kDa). The binding of 125I-LP A-I to HepG2 cells at 4 degrees C, as well as the uptake of [3H]cholesteryl ether-labelled LP A-I by HepG2 cells at 37 degrees C, was significantly higher than the binding and uptake of LP A-I/A-II. It is likely that both binding and uptake are mediated by apo A-I. Our results do not provide evidence in favor of a specific role for apo A-II in the binding and uptake of HDL by HepG2 cells.     

The conformation of apolipoprotein A-I in high-density lipoproteins is influenced by core lipid composition and particle size: a surface plasmon resonance study

Plasma high-density lipoproteins (HDL) are a heterogeneous group of particles that vary in size as well as lipid and apoprotein composition. The effect of HDL core lipid composition and particle size on apolipoprotein (apo) A-I structure was studied using surface plasmon resonance (SPR) analysis of the binding of epitope-defined monoclonal antibodies. The association and dissociation rate constants of 12 unique apo A-I-specific monoclonal antibodies for isolated plasma HDL were calculated. In addition, the association rate constants of the antibodies were determined for homogeneous preparations of spherical reconstituted HDL (rHDL) that contained apo A-I as the sole apolipoprotein and differed either in their size or in their core lipid composition. This analysis showed that lipoprotein size affected the conformation of domains dispersed throughout the apo A-I molecule, but the conformation of the central domain between residues 121 and 165 was most consistently modified. In contrast, replacement of core cholesteryl esters with triglyceride in small HDL modified almost the entire molecule, with only two key N-terminal domains of apo A-I being unaffected. This finding suggested that the central and C-terminal domains of apo A-I are in direct contact with rHDL core lipids. This immunochemical analysis has provided valuable insight into how core lipid composition and particle size affect the structure of specific domains of apo A-I on HDL.     

Speciation of human plasma high-density lipoprotein (HDL): HDL stability and apolipoprotein A-I partitioning

The distribution of apolipoprotein (apo) A-I between human high-density lipoproteins (HDL) and water is an important component of reverse cholesterol transport and the atheroprotective effects of HDL. Chaotropic perturbation (CP) with guanidinium chloride (Gdm-Cl) reveals HDL instability by inducing the unfolding and transfer of apo A-I but not apo A-II into the aqueous phase while forming larger apo A-I deficient HDL-like particles and small amounts of cholesteryl ester-rich microemulsions (CERMs). Our kinetic and hydrodynamic studies of the CP of HDL species separated according to size and density show that (1) CP mediated an increase in HDL size, which involves quasi-fusion of surface and core lipids, and release of lipid-free apo A-I (these processes correlate linearly), (2) >94% of the HDL lipids remain with an apo A-I deficient particle, (3) apo A-II remains associated with a very stable HDL-like particle even at high levels of Gdm-Cl, and (4) apo A-I unfolding and transfer from HDL to water vary among HDL subfractions with the larger and more buoyant species exhibiting greater stability. Our data indicate that apo A-I's on small HDL (HDL-S) are highly dynamic and, relative to apo A-I on the larger more mature HDL, partition more readily into the aqueous phase, where they initiate the formation of new HDL species. Our data suggest that the greater instability of HDL-S generates free apo A-I and an apo A-I deficient HDL-S that readily fuses with the more stable HDL-L. Thus, the presence of HDL-L drives the CP remodeling of HDL to an equilibrium with even larger HDL-L and more lipid-free apo A-I than with either HDL-L or HDL-S alone. Moreover, according to dilution studies of HDL in 3 M Gdm-Cl, CP of HDL fits a model of apo A-I partitioning between HDL phospholipids and water that is controlled by the principal of opposing forces. These findings suggest that the size and relative amount of HDL lipid determine the HDL stability and the fraction of apo A-I that partitions into the aqueous phase where it is destined for interaction with ABCA1 transporters, thereby initiating reverse cholesterol transport or, alternatively, renal clearance.     

Putative mechanisms of action of probucol on high-density lipoprotein apolipoprotein A-I and its isoproteins kinetics in rabbits

Probucol is a widely prescribed lipid-lowering agent, the major effects of which are to lower cholesterol in both low- and high-density lipoproteins (LDL and HDL, respectively). The mechanism of action of probucol on HDL apolipoprotein (apo) A-I kinetics was investigated in rabbits, with or without cholesterol feeding. 125I-labeled HDL was injected intravenously, and blood samples were taken periodically for 6 days. Kinetic parameters were calculated from the apo A-I-specific radioactivity decay curves. Fractional catabolic rate (FCR) and synthetic rate (SR) of apo A-I in rabbits fed a normal chow and normal chow with 1% probucol were similar. Apo A-I FCR of the rabbits fed 0.5% cholesterol was significantly increased but there were no changes in SR, compared to findings in the normal chow-fed group. Apo A-I FCR of the rabbits fed 1% probucol with 0.5% cholesterol (both 1 month and 2 months) was significantly increased compared to findings in rabbits fed the normal chow as well as 0.5% cholesterol diet group, while SR of apo A-I was significantly reduced in the former groups. Kinetics at 1 month after discontinuation of 1% probucol (under cholesterol feeding) showed a similar FCR of HDL-apo A-I to that of the rabbits fed 0.5% cholesterol, but the SR of apo A-I remained lower. Apo A-I isoproteins kinetics assessed by autoradiography of isoelectric focusing slab gels showed that the synthesis of proapo A-I was significantly reduced in the 1% probucol with 0.5% cholesterol administered, compared to the 0.5% cholesterol group. Thus, the action of probucol on HDL apo A-I kinetics was only prominent in case of higher serum cholesterol levels. The decreased HDL or apo A-I seen with probucol was apparently the result of an increase in FCR and a decrease in SR of HDL-apo A-I. A decreased synthesis of apo A-I remained evident even 1 month after discontinuing probucol. The action of probucol on the intracellular synthetic processes of apo A-I was revealed by the reduced synthesis of proapo A-I.     

Molecular dynamics simulations on discoidal HDL particles suggest a mechanism for rotation in the apo A-I belt model

Apolipoprotein A-I (apo A-I) is the major protein component of high-density lipoprotein (HDL) particles. Elevated levels of HDL in the bloodstream have been shown to correlate strongly with a reduced risk factor for atherosclerosis. Molecular dynamics simulations have been carried out on three separate model discoidal high-density lipoprotein particles (HDL) containing two monomers of apo A-I and 160 molecules of palmitoyloleoylphosphatidylcholine (POPC), to a time-scale of 1ns. The starting structures were on the basis of previously published molecular belt models of HDL consisting of the lipid-binding C-terminal domain (residues 44-243) wrapped around the circumference of a discoidal HDL particle. Subtle changes between two of the starting structures resulted in significantly different behavior during the course of the simulation. The results provide support for the hypothesis of Segrest et al. that helical registration in the molecular belt model of apo A-I is modulated by intermolecular salt bridges. In addition, we propose an explanation for the presence of proline punctuation in the molecular belt model, and for the presence of two 11-mer helical repeats interrupting the otherwise regular pattern of 22-mer helical repeats in the lipid-binding domain of apo A-I.     

Sterol efflux to apolipoprotein A-I originates from caveolin-rich microdomains and potentiates PDGF-dependent protein kinase activity

The kinetics of sterol efflux from human aortic smooth muscle cells equilibrated with a [(3)H]benzophenone-modified photoactivable free cholesterol analogue ((3)H-FCBP) did not differ significantly from those labeled with free cholesterol ((3)H-FC). Trypsin digestion of caveolin cross-linked by photoactivation of FCBP led to association of radiolabel in a single low molecular weight fraction, indicating relative structural homogeneity of caveolin-bound sterol. These findings were used to investigate the organization of sterols in caveolae and the ability of these domains to transfer sterols to apolipoprotein A-I (apo A-I), the major protein of human plasma high-density lipoproteins (HDL). During long-term (4-5 h) incubation with apo A-I, caveolin-associated (3)H-FC and (3)H-FCBP decreased, in parallel with an increase in apo A-I-associated sterol. Assay of caveolin-associated labeled sterols indicated that caveolae were a major source of sterol lost from the cells during HDL formation. Short-term changes of sterol distribution in caveolae were assayed using platelet-derived growth factor (PDGF). PDGF was without effect on FC efflux in the absence of apo A-I, but when apo A-I was present, PDGF increased FC efflux approximately 3-fold beyond the efflux rate catalyzed by apo A-I alone. At the same time, caveolin-associated FC decreased, and PDGF-dependent protein kinase activity was stimulated. Parallel results were obtained with (3)H-FCBP-equilibrated cells, in which apo A-I potentiated a PDGF-mediated reduction of radiolabel cross-linked to caveolin following photoactivation. These results suggest that sterols within caveolae are mobile and selectively transferred to apo A-I. They also suggest a novel role for sterol efflux in amplifying PDGF-mediated signal transduction.     

A Model of Lipid-Free Apolipoprotein A-I Revealed by Iterative Molecular Dynamics Simulation

Apolipoprotein A-I (apo A-I), the major protein component of high-density lipoprotein, has been proven inversely correlated to cardiovascular risk in past decades. The lipid-free state of apo A-I is the initial stage which binds to lipids forming high-density lipoprotein. Molecular models of lipid-free apo A-I have been reported by methods like X-ray crystallography and chemical cross-linking/mass spectrometry (CCL/MS). Through structural analysis we found that those current models had limited consistency with other experimental results, such as those from hydrogen exchange with mass spectrometry. Through molecular dynamics simulations, we also found those models could not reach a stable equilibrium state. Therefore, by integrating various experimental results, we proposed a new structural model for lipid-free apo A-I, which contains a bundled four-helix N-terminal domain (1–192) that forms a variable hydrophobic groove and a mobile short hairpin C-terminal domain (193–243). This model exhibits an equilibrium state through molecular dynamics simulation and is consistent with most of the experimental results known from CCL/MS on lysine pairs, fluorescence resonance energy transfer and hydrogen exchange. This solution-state lipid-free apo A-I model may elucidate the possible conformational transitions of apo A-I binding with lipids in high-density lipoprotein formation.     

Human neutrophils employ the myeloperoxidase/hydrogen peroxide/chloride system to oxidatively damage apolipoprotein A-I.

The structural integrity of apolipoprotein A-I (apo A-I) is critical to the physiological function of high-density lipoprotein (HDL). Oxidized lipoproteins are thought to be of central importance in atherogenesis, and oxidation products characteristic of myeloperoxidase, a heme protein secreted by activated phagocytes, have been detected in human atherosclerotic tissue. At plasma concentrations of halide ion, hypochlorous acid is a major product of the myeloperoxidase-hydrogen peroxide-chloride system. We therefore investigated the effects of activated human neutrophils, a potent source of myeloperoxidase and hydrogen peroxide, on the protein and lipid components of HDL. Both free and HDL-associated apo A-I exposed to activated human neutrophils underwent extensive degradation as monitored by RP-HPLC and Western blotting with a polyclonal antibody to apo A-I. Replacement of the neutrophils with reagent HOCl resulted in comparable damage (at molar oxidant : HDL subclass 3 ratio = 100) as observed in the presence of activated phagocytes. Apo A-I degradation by activated neutrophils was partially inhibited by the HOCl scavenger methionine, by the heme inhibitor azide, by chloride-free conditions, by the peroxide scavenger catalase, and by a combination of superoxide dismutase (SOD)/catalase, implicating HOCl in the cell-mediated reaction. The addition of a protease inhibitor (3,4-dichloroisocoumarin) further reduced the extent of apo A-I damage. In contrast to the protein moiety, there was little evidence for oxidation of unsaturated fatty acids or cholesterol in HDL3 exposed to activated neutrophils, suggesting that HOCl was selectively damaging apo A-I. Our observations indicate that HOCl generated by myeloperoxidase represents one pathway for protein degradation in HDL3 exposed to activated phagocytes.     

Interaction of apolipoprotein A-II with recombinant HDL containing egg phosphatidylcholine, unesterified cholesterol and apolipoprotein A-I

The preparation of discoidal, recombinant HDL (r-HDL) containing various phospholipids, apolipoproteins and a range of concentrations of unesterified cholesterol has been reported by several investigators. The present study describes the preparation of r-HDL containing both apolipoprotein (apo) A-I and apo A-II. r-HDL with 100:1 (mol:mol) egg PC.apo A-I and 0 (Series I), 5 (Series II) or 10 (Series III) mol% unesterified cholesterol were prepared by the cholate dialysis method. The resulting complexes had a Stokes' radius of 4.7 nm and contained two molecules of apo A-I per particle. When the r-HDL (2.0 mg apo A-I) were supplemented with 1.0 mg of apo A-II, one of the apo A-I molecules was replaced by two molecules of apo A-II. This modification was not accompanied by a loss of phospholipid, nor by major change in particle size. The addition of 2.5 or 4.0 mg of apo A-II resulted in the displacement of both apo A-I molecules from a proportion of the r-HDL and the formation of smaller particles (Stokes' radius 3.9 nm), which contained half the original number of egg PC molecules and three molecules of apo A-II. The amount of apo A-I displaced was dependent on the concentration of unesterified cholesterol in the r-HDL: when 2.5 mg of apo A-II was added to the Series I, II and III r-HDL, 44, 60 and 70%, respectively, of the apo A-I was displaced. Addition of 4.0 mg of apo A-II did not promote further displacement of apo A-I from any of the r-HDL. By contrast, the association of apo A-II with r-HDL was independent of the concentration of unesterified cholesterol and was a linear function of the amount of apo A-II which had been added. It is concluded that (1), the structural integrity of egg PC.unesterified cholesterol.apo A-I r-HDL, which contain two molecules of apo A-I, is not affected when one of the apo A-I molecules is replaced by two molecules of apo A-II; (2), when both apo A-I molecules are replaced by apo A-II, small particles which contain three molecules of apo A-II are formed; and (3), the displacement of apo A-I from r-HDL is facilitated by the presence of unesterified cholesterol in the particles.     

Studies on the interaction between apolipoprotein A-II-enriched HDL3 and cultured bovine aortic endothelial (BAE) cells

The specific binding of high-density lipoproteins (HDL) to a number of cell and membrane types has been reported. The aim of this study was to investigate the ligand specificity of HDL binding sites on bovine aortic endothelial (BAE) cells and in particular to investigate the role of apo A-II in the interaction. In order to do this we prepared AII-HDL3 particles by incubating HDL3 with apo HDL. These particles were enriched in apo A-II, contained virtually no apo A-I, and were similar to HDL3 in terms of size and lipid composition. As these particles resemble the native HDL3 structure we believe they are probably a more suitable model for investigation of ligand specificity than artificial recombinants. AII-HDL3 particles were shown to bind to cells with similar affinity and capacity as HDL3. Further experiments indicated that HDL3 and AII-HDL3 competed with each other for binding and displayed similar affinities for a common binding site(s). The results suggest that apo A-II, as well as apo A-I, play an important role in the process of HDL recognition by putative HDL receptors on endothelial cells.     

Apolipoproteins and their electrophoretic pattern throughout the reproductive cycle in the green frog Rana esculenta

Lipoprotein fractions in Rana esculenta were separated using the same salt intervals currently applied for human lipoproteins. Very low density lipoproteins (VLDL), low density lipoproteins (LDL) and high density lipoproteins (HDL) were analyzed with reference to the electrophoretic pattern. The lipoprotein electrophoretic pattern in males and females throughout the reproductive cycle showed minor differences. In general, each fraction was characterized by a specific apolipoprotein content. VLDL and LDL fractions were dominated by a high molecular weight (MW) band, most likely the counterpart of human Apolipoprotein B (apo B). The apo B in R. esculenta cross reacted, although weakly, with antibodies raised against chicken apo B. The HDL fraction showed a band with an apparent MW of 29 kDa. The electrophoretic mobility of the protein moiety of HDL was similar to human apolipoprotein A-I (apo A-I). However, HDL apolipoprotein of R. esculenta did not cross react with antibodies against chicken apo A-I under either denaturing or native conditions. The HDL apolipoprotein of R. esculenta was purified by DEAE-Sephacel chromatography followed by HPLC. Its amino acid composition showed a moderate correlation with trout, salmon, chicken and human apo A-I.