Use of techniques derived from graph theory to compare secondary structure motifs in proteins

A substructure matching algorithm is described that can be used for the automatic identification of secondary structural motifs in three-dimensional protein structures from the Protein Data Bank. The proteins and motifs are stored for searching as labelled graphs, with the nodes of a graph corresponding to linear representations of helices and strands and the edges to the inter-line angles and distances. A modification of Ullman's subgraph isomorphism algorithm is described that can be used to search these graph representations. Tests with patterns from the protein structure literature demonstrate both the efficiency and the effectiveness of the search procedure, which has been implemented in FORTRAN 77 on a MicroVAX-II system, coupled to the molecular fitting program FRODO on an Evans and Sutherland PS300 graphics system.     

The primary structure of apolipoprotein A-I from rabbit high-density lipoprotein

The amino acid sequence of rabbit apolipoprotein A-I (apo A-I) has been determined by degradation and alignment of two overlapping sets of peptides obtained from tryptic and staphylococcal digestions. All of the peptides of rabbit apo A-I resulting from digestion by staphylococcal protease were isolated and sequenced except residues 33-37. A digestion with trypsin was employed to find overlapping and missing peptides. The N-terminus of rabbit apo A-I was confirmed by sequencing the intact protein up to 20 residues while the C-terminus was identified through its homology with human apo A-I. The protein contains 241 residues in its single chain. Its primary structure is highly homologous to the reported canine apo A-I (80%) and human apo A-I (78%), but exhibits less similarity with rat apo A-I (60%). Like human apo A-I, rabbit apo A-I contains very little histidine (2) and methionine (1); it does however have two residues of isoleucine. Based on a comparison of the hydrophobic-hydrophilic character of apo A-I residues with that of the two synthetic peptides that activated lecithin: cholesterol acyltransferase (Pownall et al. and Yokoyama et al.), we found that the five segments with the highest corresponding homologies on the protein are located within the N-terminal half. This suggests that the N-terminal half of apo A-I contains the major portion of regions activating lecithin: cholesterol acyltransferase.     

Characterization of apolipoprotein A-I structure using a cysteine-specific fluorescence probe

Two new Cys mutants of proapolipoprotein A-I, D9C and A232C, were created and expressed in Escherichia coli systems. Specific labeling with the thiol-reactive fluorescence probe, 6-acryloyl-2-dimethylaminonaphthalene (acrylodan), was used to study the structural organization and dynamic properties of the extreme regions of human apolipoprotein A-I (apoA-I) in lipid-free and lipid-bound states. Spectroscopic approaches, including circular dichroism and various fluorescence methods, were used to examine the properties of the mutant proteins and of their covalent adducts with the fluorescence probe. The mutations themselves had no effect on the structure and stability of apoA-I in the lipid-free state and in reconstituted HDL (rHDL) complexes. Furthermore, covalent modification with acrylodan did not alter the properties of the apoA-I variants in the lipid-bound state nor in the lipid-free A232C mutant, but it affected the structure and local stability of the lipid-free protein in the D9C mutant. Fluorescence results using the acrylodan probe confirmed a well-organized structure in the N-terminal region of apoA-I. Also, they suggested a three-dimensional structure in the C-terminal region, stabilized by protein-protein contacts. When Trp residues and acrylodan were used as donor-acceptor pairs for fluorescence resonance energy transfer (FRET), average distances could be measured. Both intensity and lifetime changes of the Trp emission indicated a protein folding in solution that brings the C-terminus of the protein near the Trp residues in the N-terminal half of the sequence. Also, the N- and C-terminal domains of apoA-I appeared to be near each other in rHDL having two apoA-I per particle.     

Structural analysis of human apolipoprotein A-I variants. Amino acid substitutions are nonrandomly distributed throughout the apolipoprotein A-I primary structure

In the course of an electrophoretic mutation screening program of 32,000 dried blood samples from newborns, 17 genetic variants of apolipoprotein A-I (apoA-I) were found and structurally analyzed. The following defects were identified by the combined use of high performance liquid chromatography, time-of-flight secondary ion mass spectrometry, and sequence analysis: Pro3----Arg (1 x), Pro4----Arg (1 x), Asp89----Glu (1 x), Lys107----0 (4 x), Lys107----Met (2 x), Glu139----Gly (2 x), Glu147----Val (1 x), Pro165----Arg (4 x), and Glu198----Lys (1 x). The distribution of point mutations in the apoA-I gene leading to these 9 and 11 other variants of apoA-I reported previously was statistically analyzed. Substitutions are overrepresented in the 10 amino-terminal amino acids (p less than 0.001, chi 2-test) and in residues 103-177 (p less than 0.025, chi 2-test) or residues 103-198 (p less than 0.05, chi 2-test), respectively. We further noted the following. (i) Prolines were substituted by arginine or histidine residues at a frequency much higher than expected on the basis of random nucleotide substitutions (5 out of 18 "electrically non-neutral" amino acid substitutions, p less than 0.001, chi 2-test). These substitutions are the result of transversions of cytosines contained within stretches of at least 5 consecutive cytosines in the apoA-I gene. The observed hypervariability of the apoA-I amino terminus, therefore, might be caused by a hot spot for mutation formed by the 7 subsequent cytosines in codons 3, 4, and 5. (ii) CpG dinucleotides were overrepresentatively affected by C----T transitions (5 out of 18 electrically nonneutral amino acid substitution, p less than 0.001, chi 2-test). The hypervariability of the apoA-I alpha-helical domain might therefore be caused by CpG dinucleotides predominantly occurring in codons 120-208 of apoA-I (82 out of 125). (iii) Comparison of mutation sites in the human apoA-I gene with sites of nonsynonymous substitutions revealed that amino acid substitutions found in human apoA-I were predominantly localized in areas that were little conserved during mammalian evolution. These regions may therefore represent areas of less structural constraint for the function of apoA-I.     

Cloning and structure analysis of the rat apolipoprotein A-I cDNA

Apolipoprotein A-I, the major protein in mammalian high-density lipoprotein, acts as a cofactor for lecithin-cholesterol acyltransferase during the formation of cholesterol ester and as such, is thought to promote cholesterol efflux from peripheral cells to the liver. In this paper, we report the partial purification of rat liver apolipoprotein A-I mRNA by a polysome immunoadsorption technique, and its cDNA cloning. Isolation of two overlapping cDNA clones enabled us to derive the whole rat apolipoprotein A-I cDNA coding sequence. Comparison of the deduced protein sequence with its human counterpart reveals a striking homology between the prepropeptide precursors. Both mature protein amino-terminal regions are very homologous, suggesting that this particular domain could be involved in lipid/protein binding or lecithin-cholesterol acyltransferase activation.     

The secondary structure of peptides derived from caseins: a circular dichroism study

Three peptides have been formed by proteolytic digestion of individual casein proteins and their secondary structures characterised by far-UV circular dichroism (CD). Peptide alpha s1(1-23), residues 1-23 of alpha s1-casein, was generated by treatment of the parent protein with chymosin. Peptides beta(1-28) and beta(1-52), residues 1-28 and 1-52 of beta-casein, were plasmin- and chymotrypsin-generated fragments, respectively. Analysis of the CD spectra revealed that in aqueous solution all three peptides have secondary structures composed exclusively of beta-sheet and random coil. A limited amount of alpha-helix was formed in two of the three peptides upon treatment with high concentrations (greater than 40% (v/v] of 2,2,2-trifluoroethanol. Partial dephosphorylation (60%) of beta(1-28) and beta(1-52) by treatment with alkaline phosphatase resulted in homogeneous preparations, as judged by polyacrylamide gel electrophoresis, which exhibited increased hydrophobicity. This reduction in the level of phosphorylation of serine residues 15, 17, 18 and 19 led to increased propensity for helix formation in the peptides in the presence of 2,2,2-trifluoroethanol, but no alpha-helical structures were detected in the dephosphorylated peptides in the absence of 2,2,2-trifluoroethanol.     

Functional similarities of human and chicken apolipoprotein A-I: dependence on secondary and tertiary rather than primary structure

To investigate the sequence requirements for apolipoprotein (apo) AI functions, comparisons of human and chicken apoAI were performed. In lipid binding assays, chicken apoAI was capable of transforming phospholipid vesicles into discoidal bilayer structures, similar in both size and apolipoprotein content to those produced with human apoAI under the same conditions. Human and chicken apoAI were indistinguishable in their relative abilities to prevent phospholipase C-induced aggregation of human low density lipoprotein. This activity, which is dependent upon formation of a stable interaction with the modified lipoprotein, represents a sensitive measure of apolipoprotein association with spherical lipoprotein particles. The ability of chicken versus human apoAI to mobilize the regulatory pool of cholesterol available for esterification by acyl-CoA:cholesterol acyltransferase by human fibroblasts was also assessed. Lipid-free chicken and human apoAI were equivalent in their ability to deplete cholesterol from this pool, as were intact chicken high density lipoprotein (HDL) and human HDL(3). Based on the overall sequence identity of chicken and human apoAI (48%), and comparison of regions thought to be responsible for key apoAI functions, these data indicate that amphipathic alpha-helical structure, rather than specific amino acid sequence, is the major determinant of apoAI lipid binding and ability to mobilize the regulatory pool of cellular cholesterol.     

Affinity of nonhomologous amphiphilic peptides toward a monoclonal antibody raised against apolipoprotein A-I

Monoclonal antibodies against human apolipoprotein A-I (apoA-I) were generated by the hybridoma technique. Clone G-10 was selected on the basis of its highest titer. The affinity of this antibody toward a series of synthetic peptides differing in length, amino acid composition, and amphiphilicity was tested by using both the indirect and the competitive enzyme-linked immunosorbent techniques (ELISA). From these measurements we calculated dissociation constants of the complexes of the antibody with apoA-I bound to the surface of the microtiter plate, apoA-I in solution, and any of the several peptides in solution. The dissociation constant (Kd) of the immobilized apoA-I/anti-apoA-I-complex, Kd = 2 x 10(-9) M, was significantly lower than that of the complex resulting from the interaction between anti-apoA-I and either apoA-I in solution or any of the several amphiphilic helical peptides in solution. Peptides devoid of amphiphilic secondary structure were inert. These data are consistent with the proposal that monoclonal G-10 recognizes in antigenic peptides an alpha-helical secondary structure of defined hydrophilic-lipophilic balance and comparatively less the specific amino acid side chains. We propose that the highest contribution to the free energy of binding (8 Kcal/mole) is derived from the docking of the helix to the antibody. It follows that in probing the specificity of a monoclonal antibody the conformation and the physical environment of the interacting antigen must be taken into account.     

Cysteine substitutions in apolipoprotein A-I primary structure modulate paraoxonase activity

Paraoxonase (PON) is transported primarily on apolipoprotein A-I (apoA-I) -containing high-density lipoprotein (HDL) and is thought to protect against early atherogenic events including low-density lipoprotein (LDL) oxidation and monocyte migration. It has been proposed that apoA-I may be necessary for PON's association with plasma HDL. On the basis of this, we examined the effect of apoA-I on PON's enzymatic activity and its ability to associate with HDL. Additionally, we examined whether changes in apoA-I primary structure (cysteine substitution mutations) could modulate these effects. Chinese hamster ovary cells stably transfected with human PON1A cDNA were incubated in the presence and absence of recombinant wild-type apoA-I (apoA-I(WT)) and specific Cys substitution mutations. Extracellular accumulation of PON activity in the presence of apoA-I(WT) was 0.095 +/- 0.013 unit/mg of cell protein (n = 7) compared to 0.034 +/- 0.010 unit/mg of cell protein in the absence of apoA-I (n = 7), a 2.79-fold increase in activity when apoA-I was incubated with the cells. Lipid-free apoA-I did not increase PON activity, while preformed nascent HDL increased PON activity only 30%, suggesting that maximal PON activity is lipid-dependent and requires coassembly of PON and apoA-I on nascent HDL. The cysteine mutations R10C, R27C, and R61C significantly increased (p < 0.01) PON activity 32.6% +/- 14.7%, 31.6% +/- 18.9%, and 27.4% +/- 20%, respectively, over that of wild type (WT). No changes in PON activity were observed with apoA-I cysteine substitution mutations in the C-terminal portion of the protein. The data suggest that, for optimal PON activity, coassembly of the enzyme onto nascent HDL is required and that the N-terminal region of apoA-I may be important in the assembly process.     

Probing the lipid-free structure and stability of apolipoprotein A-I by mutation

To probe the secondary structure of the C-terminus (residues 165-243) of lipid-free human apolipoprotein A-I (apoA-I) and its role in protein stability, recombinant wild-type and seven site-specific mutants have been produced in C127 cells, purified, and studied by circular dichroism and fluorescence spectroscopy. A double substitution (G185P, G186P) increases the protein stability without altering the secondary structure, suggesting that G185 and G186 are located in a loop/disordered region. A triple substitution (L222K, F225K, F229K) leads to a small increase in the alpha-helical content and stability, indicating that L222, F225, and F229 are not involved in stabilizing hydrophobic core contacts. The C-terminal truncation Delta(209-243) does not change the alpha-helical content but reduces the protein stability. Truncation of a larger segment, Delta(185-243), does not affect the secondary structure or stability. In contrast, an intermediate truncation, Delta(198-243), leads to a significant reduction in the alpha-helical content, stability, and unfolding cooperativity. The internal 11-mer deletion Delta(187-197) has no significant effect on the conformation or stability, whereas another internal 11-mer deletion, Delta(165-175), dramatically disrupts and destabilizes the protein conformation, suggesting that the presence of residues 165-175 is crucial for proper apoA-I folding. Overall, the findings suggest the presence of stable helical structure in the C-terminal region 165-243 of lipid-free apoA-I and the involvement of segment 209-243 in stabilizing interactions in the molecule. The effect of the substitution (G185P, G186P) on the exposure of tryptophans located in the N-terminal half suggests an apoA-I tertiary conformation with the C-terminus located close to the N-terminus.     

Denaturation of apolipoprotein A-I and the monomer form of apolipoprotein A-I(Milano).

In this study the thermal and denaturant induced unfolding of apolipoprotein A-I (apo A-I) and the monomer form of apolipoprotein A-I(Milano) (apo A-I(M)) was followed. Dimer apo A-I(M) was reduced with dithiothreitol, which was present in the protein solutions in all experiments. Thermal denaturation is followed by differential scanning calorimetry (DSC) and far-UV and near-UV CD. Both apo A-I and monomer apo A-IM have a broad asymmetric DSC peak that could be deconvoluted into three non two-state transitions, apo A-I being more stable than the monomer apo A-IM. Estimation of melting of tertiary structure by near-UV CD is lower than that for secondary structure determined from far-UV. This together with the non two-state unfolding of the proteins observed with DSC is indicative of unfolding via a molten globular-like state. Apo A-I and monomer apo A-I(M) are equally susceptible to guanidinum chloride, half-unfolded at 1.2 M denaturant. The presence of 0.5 and 1.0 M denaturant, lower and equalize the denaturation temperatures of the proteins, respectively.     

The covalent structure of apolipoprotein A-I from canine high density lipoproteins

The complete amino acid sequence of apolipoprotein A-I (apo-A-I) from canine serum high density lipoproteins (HLD) has been determined by automated Edman degradation of the intact protein and proteolytic fragments derived therefrom. The major strategy involved analysis of overlapping sets of peptides generated by cleavage at lysyl residues with Myxobacter protease and by tryptic hydrolysis at arginines in the citraconylated protein derivative. Canine apo-A-I has 232 residues in its single polypeptide chain and its covalent structure is highly homologous to one of the two reported sequences for human apo-A-I. As in the case for the human apoprotein, predictive analysis of the canine apo-A-I sequence suggests that it comprises a series of amphiphilic alpha helices punctuated by a periodic array of prolyl residues. Human HDL contains a second major protein component, apolipoprotein A-II (apo-A-II) that is lacking in HDL from dog serum. The absence of apo-A-II in canine HDL raised the possibility that the apo-A-I from this source might contain within its primary structure sequences related to apo-A-II and thus perform the dual function of both proteins in one. Our analysis proves that canine apo-A-I has all of the structural features of human apo-A-I and that it is not an A-I: A-II hybrid molecule.     

Preferential binding of apolipoprotein E derived peptides with oxidized phospholipid

The physiological function of apolipoprotein E (apoE) includes transport and metabolism of lipids and its C-terminal domain harbors high affinity lipid-binding sites. Although the binding of apoE with non-oxidized phospholipid containing membranes has been characterized earlier, the interaction of apoE or its fragments with oxidized phospholipid containing membrane has never been studied. In this study we have compared the interaction of amphipathic helical peptide sequences derived from the C-terminal domain of apoE with membrane vesicles containing oxidized phospholipid, 1-palmitoyl-2-azelaoyl-sn-glycero-3-phosphocholine (PazePC), with membrane vesicles without PazePC. The interaction was studied by monitoring (a) fluorescence emission maxima of the peptides, (b) acrylamide quenching of the peptides tryptophan residues and (c) by measuring the equilibrium binding constants by resonance energy transfer (RET) analysis. Our result shows that peptide sequence 202-223, 245-266 and 268-289 of apoE has higher affinity towards membrane containing PazePC, compared to membrane without PazePC. Presence of 1mM divalent cation or 50 mM NaCl in the buffer decreased the binding of peptides to PazePC containing membrane vesicles suggesting possible involvement of the electrostatic interaction in the binding. These observations suggest that the preferential binding of apoE to oxidized phospholipid containing membrane may play a role in the anti-oxidative properties of apoE.     

Protein heterogeneity of lipoprotein particles containing apolipoprotein A-I without apolipoprotein A-II and apolipoprotein A-I with apolipoprotein A-II isolated from human plasma

The protein heterogeneity of fractions isolated by immunoaffinity chromatography on anti-apolipoprotein A-I and anti-apolipoprotein A-II affinity columns was analyzed by high resolution two-dimensional gel electrophoresis. The two-dimensional gel electrophoresis profiles of the fractions were analyzed and automatically compared by the computer system MELANIE. Fractions containing apolipoproteins A-I + A-II and only A-I as the major protein components have been isolated from plasma and from high density lipoproteins prepared by ultracentrifugation. Similarities between the profiles of the fractions, as indicated by two-dimensional gel electrophoresis, suggested that those derived from plasma were equivalent to those from high density lipoproteins (HDL), which are particulate in nature. The established apolipoproteins (A-I, A-II, A-IV, C, D, and E) were visible and enriched in fractions from both plasma and HDL. However, plasma-derived fractions showed a much greater degree of protein heterogeneity due largely to enrichment in bands corresponding to six additional proteins. They were present in trace amounts in fractions isolated from HDL and certain of the proteins were visible in two-dimensional gel electrophoresis profiles of the plasma. These proteins are considered to be specifically associated with the immunoaffinity-isolated particles. They have been characterized in terms of Mr and pI. Computer-assisted measurements of protein spot-staining intensities suggest an asymmetric distribution of the proteins (as well as the established apolipoproteins), with four showing greater prominence in particles containing apolipoprotein A-I but no apolipoprotein A-II.     

Lipid dependant disorder-to-order conformational transitions in apolipoprotein CI derived peptides

In contrast to the notion established for many years that protein function depends on rigid 3D structures, nowadays there is important evidence suggesting that non-structured segments of proteins play important roles in protein function. Therefore, disorder-to-order dynamic conformational transitions have been proposed as an attractive mechanism involved in protein-protein recognition. Our laboratory using Langmuir monolayers of apolipoproteins has previously shown that upon lateral compression at the air/water and phospholipid/water interfaces, there is an important movement of the C-terminal segment of apolipoprotein CI towards the air, considered the hydrophobic region of the monolayer and the acyl-chain region of the interface when phospholipids are used. Here, in an attempt to define secondary structure changes that might occur within this C-terminal segment of apoCI while moving from the monolayer interface back and forth its hydrophobic region, employing three peptides derived from apoCI we studied by circular dichroism and dynamic light scattering their conformational properties when associated to a series of amphipathic lipids and lipid-like molecules. Our results show that a series of lysophospholipids present the ability to modulate the formation of an α helix at the C-terminal peptide of apoCI through a disorder-to-order transition while forming small lipid/peptide aggregates below 10 nm in diameter.     

Influence of tertiary structure domain properties on the functionality of apolipoprotein A-I

The tertiary structure of apolipoprotein (apo) A-I and the contributions of structural domains to the properties of the protein molecule are not well defined. We used a series of engineered human and mouse apoA-I molecules in a range of physical-biochemical measurements to address this issue. Circular dichroism measurements of alpha-helix thermal unfolding and fluorescence spectroscopy measurements of 8-anilino-1-napthalenesulfonic acid binding indicate that removal of the C-terminal 54 amino acid residues from human and mouse apoA-I has similar effects; the molecules are only slightly destabilized, and there is a decrease in hydrophobic surface exposure. These results are consistent with both human and mouse apoA-I adopting a two-domain tertiary structure, comprising an N-terminal antiparallel helix bundle domain and a separate less ordered C-terminal domain. Mouse apoA-I is significantly less resistant than human apoA-I to thermal and chemical denaturation; the midpoint of thermal unfolding of mouse apoA-I at 45 degrees C is 15 degrees C lower and the midpoint of guanidine hydrochloride denaturation (D1/2) occurs at 0.5 M as compared to 1.0 M for human apoA-I. These differences reflect the overall greater stability of the helix bundle formed by residues 1-189 in human apoA-I. Measurements of the heats of binding to egg phosphatidylcholine (PC) small unilamellar vesicles and the kinetics of solubilization of dimyristoyl PC multilamellar vesicles indicate that the more stable human helix bundle interacts poorly with lipids as compared to the equivalent mouse N-terminal domain. The C-terminal domain of human apoA-I is much more hydrophobic than that of mouse apoA-I; in the lipid-free state the human C-terminal domain (residues 190-243) is partially alpha-helical and undergoes cooperative unfolding (D1/2 = 0.3 M) whereas the isolated mouse C-terminal domain (residues 187-240) is disordered in dilute solution. The human C-terminal domain binds to lipid surfaces much more avidly than the equivalent mouse domain. Human and mouse apoA-I have very different tertiary structure domain contributions for achieving functionality. It is clear that the stability of the N-terminal helix bundle, and the hydrophobicity and alpha-helix content of the C-terminal domain, are critical factors in determining the overall properties of the apoA-I molecule.     

Predicting the structure of apolipoprotein A-I in reconstituted high-density lipoprotein disks.

In reconstituted high-density lipoproteins, apolipoprotein A-I and phosphatidylcholines combine to form disks in which the amphipathic alpha-helices of apolipoprotein A-1 bind to the edge of a lipid bilayer core, shielding the hydrophic lipid tails from the aqueous environment. We have employed experimental data, sequence analysis, and molecular modeling to construct an atomic model of such a reconstituted high-density lipoprotein disk consisting of two apolipoprotein A-I proteins and 160 palmitoyloleoylphosphatidylcholine lipids. The initial globular domain (1-47) of apolipoprotein A-I was excluded from the model, which was hydrated with an 8-A shell of water molecules. Molecular dynamics and simulated annealing were used to test the stability of the model. Both head-to-tail and head-to-head forms of a reconstituted high-density lipoprotein were simulated. In our simulations the protein contained and adhered to the lipid bilayer while providing good coverage of the lipid tails.     

Apolipoprotein A-I(Milano) and apolipoprotein A-I(Paris) exhibit an antioxidant activity distinct from that of wild-type apolipoprotein A-I

Apolipoprotein A-I(Milano) (apoA-I(Milano)) and apoA-I(Paris) are rare cysteine variants of apoA-I that produce a HDL deficiency in the absence of cardiovascular disease in humans. This paradox provides the basis for the hypothesis that the cysteine variants possess a beneficial activity not associated with wild-type apoA-I (apoA-I(WT)). In this study, a unique antioxidant activity of apoA-I(Milano) and apoA-I(Paris) is described. ApoA-I(Milano) was twice as effective as apoA-I(Paris) in preventing lipoxygenase-mediated oxidation of phospholipids, whereas apoA-I(WT) was poorly active. Antioxidant activity was observed using the monomeric form of the variants and was equally effective before and after initiation of oxidative events. ApoA-I(Milano) protected phospholipid from reactive oxygen species (ROS) generated via xanthine/xanthine oxidase (X/Xo) but failed to inhibit X/Xo-induced reduction of cytochrome c. These results indicate that apoA-I(Milano) was unable to directly quench ROS in the aqueous phase. There were no differences between lipid-free apoA-I(Milano,) apoA-I(Paris), and apoA-I(WT) in mediating the efflux of cholesterol from macrophages, indicating that the cysteine variants interacted normally with the ABCA1 efflux pathway. The results indicate that incorporation of a free thiol within an amphipathic alpha helix of apoA-I confers an antioxidant activity distinct from that of apoA-I(WT). These studies are the first to relate gain of function to rare cysteine mutations in the apoA-I primary sequence.