The structure of human apolipoprotein C-1 in four different crystal forms
2018; Elsevier BV; Volume: 60; Issue: 2 Linguagem: Inglês
10.1194/jlr.m089441
ISSN1539-7262
AutoresAlexander McPherson, Steven B. Larson,
Tópico(s)Protein Structure and Dynamics
ResumoHuman apolipoprotein C1 (APOC1) is a 57 amino acid long polypeptide that, through its potent inhibition of cholesteryl ester transferase protein, helps regulate the transfer of lipids between lipid particles. We have now determined the structure of APOC1 in four crystal forms by X-ray diffraction. A molecule of APOC1 is a single, slightly bent, α-helix having 13–14 turns and a length of about 80 Å. APOC1 exists as a dimer, but the dimers are not the same in the four crystals. In two monoclinic crystals, two helices closely engage one another in an antiparallel fashion. The interactions between monomers are almost entirely hydrophobic with sparse electrostatic complements. In the third monoclinic crystal, the two monomers spread at one end of the dimer, like a scissor opening, and, by translation along the crystallographic a axis, form a continuous, contiguous sheet through the crystal. In the orthorhombic crystals, two molecules of APOC1 are related by a noncrystallographic 2-fold axis to create an arc of about 120 Å length. This symmetrical dimer utilizes interactions not present in dimers of the monoclinic crystals. Versatility of APOC1 monomer association shown by these crystals is suggestive of physiological function. Human apolipoprotein C1 (APOC1) is a 57 amino acid long polypeptide that, through its potent inhibition of cholesteryl ester transferase protein, helps regulate the transfer of lipids between lipid particles. We have now determined the structure of APOC1 in four crystal forms by X-ray diffraction. A molecule of APOC1 is a single, slightly bent, α-helix having 13–14 turns and a length of about 80 Å. APOC1 exists as a dimer, but the dimers are not the same in the four crystals. In two monoclinic crystals, two helices closely engage one another in an antiparallel fashion. The interactions between monomers are almost entirely hydrophobic with sparse electrostatic complements. In the third monoclinic crystal, the two monomers spread at one end of the dimer, like a scissor opening, and, by translation along the crystallographic a axis, form a continuous, contiguous sheet through the crystal. In the orthorhombic crystals, two molecules of APOC1 are related by a noncrystallographic 2-fold axis to create an arc of about 120 Å length. This symmetrical dimer utilizes interactions not present in dimers of the monoclinic crystals. Versatility of APOC1 monomer association shown by these crystals is suggestive of physiological function. Apolipoprotein C-1 (APOC1) is a 57 amino acid polypeptide of Mr = 6.6 kDa (1Shulman R.S. Herbert P.N. Wehrly K. Fredrickson D.S. The complete amino acid sequence of C–I (apoLp-Ser), an apolipoprotein from human very low density lipoproteins.J. Biol. Chem. 1975; 250: 182-190Abstract Full Text PDF PubMed Google Scholar) that is synthesized predominantly in the liver of mammals. Its gene product includes a 26 amino acid long signal peptide that is proteolytically removed in the rough endoplasmic reticulum upon secretion. In humans, the plasma concentration is about 6 mg/dl. This small protein is remarkable for being the most positively charged protein in the human body, with a lysine content of 17%. The polypeptide contains no histidine, tyrosine, or cysteine, nor is it attached covalently to any carbohydrate (1Shulman R.S. Herbert P.N. Wehrly K. Fredrickson D.S. The complete amino acid sequence of C–I (apoLp-Ser), an apolipoprotein from human very low density lipoproteins.J. Biol. Chem. 1975; 250: 182-190Abstract Full Text PDF PubMed Google Scholar, 2Jong M.C. Hofker M.H. Havekes L.M. Role of ApoCs in lipoprotein metabolism: functional differences between ApoC1, ApoC2, and ApoC3.Arterioscler. Thromb. Vasc. Biol. 1999; 19: 472-484Crossref PubMed Scopus (430) Google Scholar). The protein has been the subject of numerous structural studies by NMR (3Rozek A. Buchko G.W. Cushley R.J. Conformation of two peptides corresponding to human apolipoprotein C-I residues 7–24 and 35–53 in the presence of sodium dodecyl sulfate by CD and NMR spectroscopy.Biochemistry. 1995; 34: 7401-7408Crossref PubMed Scopus (67) Google Scholar, 4Rozek A. Sparrow J.T. Weisgraber K.H. Cushley R.J. Sequence-specific 1H NMR resonance assignments and secondary structure of human apolipoprotein C-I in the presence of sodium dodecyl sulfate.Biochem. Cell Biol. 1998; 76: 267-275Crossref PubMed Scopus (12) Google Scholar, 5Cushley R.J. Okon M. NMR studies of lipoprotein structure.Annu. Rev. Biophys. Biomol. Struct. 2002; 31: 177-206Crossref PubMed Scopus (41) Google Scholar) and other biophysical techniques, but no crystal structure, to our knowledge, has been reported. In human serum, APOC1 is responsible for the activation of esterified lecithin cholesterol. It plays an important role in the exchange of esterified cholesterol between lipoprotein particles and in the removal of cholesterol from tissues. The protein has a high affinity for lipid surfaces, particularly chylomicrons, and both HDL and VLDL particles. In the fasting state, APOC1 is mostly associated with HDL, but in the fed state, it redistributes to the surfaces of chylomicrons and VLDL particles. As a consequence, the presence of APOC1 on a lipoprotein particle may prolong its residence time in the circulation and subsequently facilitate its conversion to LDL (2Jong M.C. Hofker M.H. Havekes L.M. Role of ApoCs in lipoprotein metabolism: functional differences between ApoC1, ApoC2, and ApoC3.Arterioscler. Thromb. Vasc. Biol. 1999; 19: 472-484Crossref PubMed Scopus (430) Google Scholar, 6Dolphin P.J. Lipoprotein metabolism and the role of apolipoproteins as metabolic programmers.Can. J. Biochem. Cell Biol. 1985; 63: 850-869Crossref PubMed Scopus (53) Google Scholar). APOC1 had long been known to be a crucial player in the transfer of lipids between particles and the general clearance of lipids (2Jong M.C. Hofker M.H. Havekes L.M. Role of ApoCs in lipoprotein metabolism: functional differences between ApoC1, ApoC2, and ApoC3.Arterioscler. Thromb. Vasc. Biol. 1999; 19: 472-484Crossref PubMed Scopus (430) Google Scholar), but its mechanism and the manner by which it bound to lipoprotein particles was poorly understood (5Cushley R.J. Okon M. NMR studies of lipoprotein structure.Annu. Rev. Biophys. Biomol. Struct. 2002; 31: 177-206Crossref PubMed Scopus (41) Google Scholar). It was then discovered (7Gautier T. Masson D. de Barros J.P. Human apolipoprotein C-1 accounts for the ability of plasma high density lipoproteins to inhibit the cholesteryl ester transferase protein activity.J. Biol. Chem. 2000; 275: 37504-37509Abstract Full Text Full Text PDF PubMed Scopus (107) Google Scholar, 8de Barros J.P. Boualam A. Gautier T. Dumont L. Vergès B. Masson D. Lagrost L. Apolipoprotein CI is a physiological regulator of cholesteryl ester transfer protein activity in human plasma but not in rabbit plasma.J. Lipid Res. 2009; 50: 1842-1851Abstract Full Text Full Text PDF PubMed Scopus (25) Google Scholar) that APOC1 is a potent inhibitor of cholesteryl ester transferase protein (CETP), which is the crucial enzyme in the transfer of cholesterol between lipid surfaces (tissues) and lipid particles. It is this function, as an inhibitor of CETP, that likely explains its ubiquitous role in lipid metabolism. It is postulated that CETP inhibition by APOC1 may operate through the ability of APOC1 to modify the electrostatic surfaces of lipoprotein particles to which it binds (2Jong M.C. Hofker M.H. Havekes L.M. Role of ApoCs in lipoprotein metabolism: functional differences between ApoC1, ApoC2, and ApoC3.Arterioscler. Thromb. Vasc. Biol. 1999; 19: 472-484Crossref PubMed Scopus (430) Google Scholar, 9Larsson M. Vorrsjo E. Talmud P. Lookene A. Olivecrona G. Apolipoproteins C-I and C-III inhibit lipoprotein lipase activity by displacement of the enzyme from lipid droplets.J. Biol. Chem. 2013; 288: 33997-34008Abstract Full Text Full Text PDF PubMed Scopus (114) Google Scholar). Because of its importance in lipid and cholesterol metabolism, APOC1 naturally plays an important role in medicine (10Kasthuri R.S. McMillan K.R. Flood-Urdangain C. Harvey S.B. Wilson-Grady J.T. Nelsestuen G.L. Correlation of a T45S variant of apolipoprotein C1 with elevated BMI in persons of Indian and Mexican ancestries.Int. J. Obes. (Lond.). 2007; 31: 1334-1336Crossref PubMed Scopus (11) Google Scholar), particularly in atherosclerosis and other cardiovascular conditions (11Hansen J.B. Fernandez J.A. Noto A.T. Deguchi H. Bjorkegren J. Mathiesen E.B. The apolipoprotein C-I content of very-low-density lipoproteins is associated with fasting triglycerides, postprandial lipemia, and carotid atherosclerosis.J. Lipids. 2011; 2011: 271062Crossref PubMed Google Scholar), obesity (12Jong M.C. Voshol P.J. Muurling M. Dahlmans V.E. Romijn J.A. Pijl H. Havekes L.M. Protection from obesity and insulin resistance in mice overexpressing human apolipoprotein C1.Diabetes. 2001; 50: 2779-2785Crossref PubMed Scopus (79) Google Scholar), and diabetes (13Bouillet B. Gautier T. Aho L.S. Duvillard L. Petit J.M. Lagrost L. Vergès B. Plasma apolipoprotein C1 concentration is associated with plasma triglyceride concentration, but not visceral fat, in patients with type 2 diabetes.Diabetes Metab. 2016; 42: 263-266Crossref PubMed Scopus (11) Google Scholar). It would not be appropriate to review the field here, as an extensive literature on lipoproteins in medicine exists. It is, nonetheless, interesting to note that reports have appeared (14Meunier J.C. Russell R.S. Engle R.E. Faulk K.N. Purcell R.H. Emerson S.U. Apolipoprotein c1 association with hepatitis C virus.J. Virol. 2008; 82: 9647-9656Crossref PubMed Scopus (110) Google Scholar, 15Crouchet E. Baumert T.F. Schuster C. Hepatitis C virus-apolipoprotein interactions: molecular mechanisms and clinical impact.Expert Rev. Proteomics. 2017; 14: 593-606Crossref PubMed Scopus (15) Google Scholar, 16Fukuhara T. Ono C. Puig-Basagoiti F. Matsuura Y. Roles of lipoproteins and apolipoproteins in particle formation of hepatitis C virus.Trends Microbiol. 2015; 23: 618-629Abstract Full Text Full Text PDF PubMed Scopus (41) Google Scholar) showing that APOC1 is also associated with the hepatitis C virion. This suggests that its importance in medicine may be even more profound than previously understood. About 25 years ago, we crystallized, for X-ray diffraction analysis, human APOC1 (17Weisgraber K.H. Newhouse Y.M. McPherson A. Crystallization and preliminary X-ray analysis of human plasma apolipoprotein C-I.J. Mol. Biol. 1994; 236: 382-384Crossref PubMed Scopus (14) Google Scholar), using 2-methyl-2,4-pentanediol (MPD) as precipitant along with a nonionic detergent, octyl-β-d-1-thioglucopyranoside in the mother liquor. We ultimately grew crystals in four different crystallographic unit cells. Their properties are presented in Table 1. Two of the crystals, Monocln-1 and Monocln-2, are close variations of one another, the latter a less hydrated form of the former. The crystals, aside from these two, include another distinct monoclinic form, Monocln-3, and one orthorhombic form, designated Orthrhmb. All of the monoclinic crystals diffracted to at least 2.0 Å resolution, whereas the orthorhombic form diffracted to about 3 Å.TABLE 1Crystal and unit cell parametersCrystalSpace Groupa, Åb, Åc, ÅβoVmSolvent, %Resolution, ÅMol/a.u.Monocl-1P2131.0547.6034.1895.741.87341.82Monocl-2P2127.8846.4534.0994.121.64252.02Monocl-3P2135.4549.0637.67105.242.41492.02OrthrhmbP21212134.4653.7071.2090.02.50513.02a.u., asymmetric unit. Open table in a new tab a.u., asymmetric unit. In the years 1994–1995, at least 20 X-ray diffraction datasets were collected from native crystals and putative heavy-atom derivatives. Unfortunately, at that time, none of the heavy-atom derivative trials proved useful in phasing, which is not surprising because the protein contained no histidine, cysteine, or tyrosine. In addition, in spite of our best efforts, we were similarly unsuccessful in obtaining a structure solution using molecular replacement. Thus, the X-ray data were archived, and the project was essentially shelved. Two years ago, we retrieved the X-ray data collected 25 years ago, which remain quite respectable even now (see Table 2), from our archives and resurrected the APOC1 analysis. We benefitted this time, however, by the past 25 years of development of powerful new approaches, computers, and software tools for crystallographic analysis, particularly in the areas of molecular replacement (18Read R.J. Pushing the boundaries of molecular replacement with maximum likelihood.Acta Crystallogr. D Biol. Crystallogr. 2001; 57: 1373-1382Crossref PubMed Scopus (785) Google Scholar, 19Storoni L.C. McCoy A.J. Read R.J. Likelihood-enhanced fast rotation functions.Acta Crystallogr. D Biol. Crystallogr. 2004; 60: 432-438Crossref PubMed Scopus (1094) Google Scholar, 20McCoy A.J. Grosse-Kunstleve R.W. Adams P.D. Winn M.D. Storoni L.C. Read R.J. Phaser crystallographic software.J. Appl. Crystallogr. 2007; 40: 658-674Crossref PubMed Scopus (14432) Google Scholar, 21McCoy A.J. Grosse-Kunstleve R.W. Storoni L.C. Read R.J. Likelihood-enhanced fast translation functions.Acta Crystallogr. D Biol. Crystallogr. 2005; 61: 458-464Crossref PubMed Scopus (1599) Google Scholar) and maximum-likelihood refinement (22Murshudov G.N. Vagin A.A. Dodson E.J. Application of maximum likelihood refinement.In Refinement of Protein Structures. Proceedings of Daresbury Study Weekend. 1996; Google Scholar, 23Murshudov G.N. Vagin A.A. Dodson E.J. Refinement of macromolecular structures by the maximum-likelihood method.Acta Crystallogr. D Biol. Crystallogr. 1997; 53: 240-255Crossref PubMed Scopus (13851) Google Scholar). This time our efforts were rewarded, and we successfully solved the structures of all four of the crystal forms of 25 years ago; we report them here.TABLE 2Data collection, processing, and scaling: Monocl-1, Monocl-2, Monocl-3, and OrthrhmbCrystalMonocl-1Monocl-2Monocl-3OrthrhmbX-ray sourceRigaku RU-200Rigaku RU-200Rigaku RU-200Rigaku RU-200DetectorSDMSSDMSSDMSSDMSMosaicity0.45°0.48°0.45°0.68°S.G. prob.0.8190.8480.6470.442Resolution35.0–1.80 Å34.0–2.0 Å30.0–2.0 Å15.0–3.0 ÅOuter shell1.95–1.802.10–2.002.05–2.03.18–3.0No. of unique Fs7,315 (613)5,912 (406)4,949 (206)2,687 (460)CC 1/20.993 (0.640)0.996 (0.567)0.989 (0.622)0.997 (0.733)Rmerge0.094 (0.457)0.144 (0.656)0.126 (0.583)0.169 (0.491)Rmeas0.114 (0.721)0.152 (0.842)0.146 (0.751)0.180 (0.619)Rpim0.055 (0.315)0.079 (0.417)0.070 (0.350)0.080 (0.369)Completeness85.2 (.760)95.4 (97.3)76.0 (32.5)91.8 (56.3)Multiplicity7.2 (4.5)6.4 (3.7)2.5 (1.5)7.1 (1.7)I/sigma6.7 (3.2)4.5 (1.4)3.7 (1.2)5.5 (1.4)S.G. prob, space group probability. Open table in a new tab Details of the preparation and crystallization of the protein were presented in an earlier paper (17Weisgraber K.H. Newhouse Y.M. McPherson A. Crystallization and preliminary X-ray analysis of human plasma apolipoprotein C-I.J. Mol. Biol. 1994; 236: 382-384Crossref PubMed Scopus (14) Google Scholar). APOC1 was crystallized by sitting-drop vapor diffusion (24McPherson, A., 1982. New York: John Wiley and Sons.Google Scholar) using Cryschem plates (Hampton Research, Aliso Viejo, CA). Reservoirs were 16–18% MPD with 0.10 M sodium acetate and 0.25% octyl-β-d-1-thioglucopyranoside. Crystallization droplets were composed of equal amounts of 8 mg/ml protein in 0.02 M NH4HCO stock solution and reservoir solution. Crystallization was at room temperature. All of the crystal forms reported here appeared from these same crystallization droplets, although every drop contained but one crystal form. Crystals were mounted by conventional means in 0.7–0.8 mm quartz capillaries (24McPherson, A., 1982. New York: John Wiley and Sons.Google Scholar), and X-ray diffraction data were recorded at room temperature using a Rigaku RU-200 generator fitted with a Supper graphite crystal monochromator and operated at 40 kV and 30 mA with twin San Diego Multiwire Systems (SDMS) detectors at a crystal to detector distance of 420 mm. Images were processed with software provided by SDMS (25Howard A.J. Nielsen C. Xuong N.H. Software for a diffractometer with multiwire area detector.Methods Enzymol. 1985; 114: 452-472Crossref PubMed Scopus (280) Google Scholar). Structure amplitudes were obtained by scaling and merging intensities from archived .ARC files using the program AIMLESS (26Evans P. Scaling and assessment of data quality.Acta Crystallogr. D Biol. Crystallogr. 2006; 62: 72-82Crossref PubMed Scopus (3743) Google Scholar, 27Evans P.R. An introduction to data reduction: space-group determination, scaling and intensity statistics.Acta Crystallogr. D Biol. Crystallogr. 2011; 67: 282-292Crossref PubMed Scopus (992) Google Scholar) to yield comprehensive datasets (Table 1). Molecular replacement searches were carried out using the program PHASER (18Read R.J. Pushing the boundaries of molecular replacement with maximum likelihood.Acta Crystallogr. D Biol. Crystallogr. 2001; 57: 1373-1382Crossref PubMed Scopus (785) Google Scholar, 19Storoni L.C. McCoy A.J. Read R.J. Likelihood-enhanced fast rotation functions.Acta Crystallogr. D Biol. Crystallogr. 2004; 60: 432-438Crossref PubMed Scopus (1094) Google Scholar, 20McCoy A.J. Grosse-Kunstleve R.W. Adams P.D. Winn M.D. Storoni L.C. Read R.J. Phaser crystallographic software.J. Appl. Crystallogr. 2007; 40: 658-674Crossref PubMed Scopus (14432) Google Scholar). Rebuilding and most graphics operations relied on the program COOT (28Emsley P. Cowtan K. COOT model - building tools for molecular graphics.Acta Crystallogr. D Biol. Crystallogr. 2004; 60: 2126-2132Crossref PubMed Scopus (23218) Google Scholar), as did quantitative comparisons of models. Refinement of the polypeptide models was accomplished using the program REFMAC (29Murshudov G.N. Skubák P. Lebedev A.A. Pannu N.S. Steiner R.A. Nicholls R.A. Winn M.D. Long F. Vagin A.A. REFMAC5 for the refinement of macromolecular crystal structures.Acta Crystallogr. D Biol. Crystallogr. 2011; 67: 355-367Crossref PubMed Scopus (5948) Google Scholar) from the CCP4 program system (30Collaborative Computational Project The CCP4 Suite: Programs for Protein Crystallography..Didcot, U.K.: Collaborative Computational Project. 1994; : 760-763Google Scholar) based on the maximum-likelihood approach (22Murshudov G.N. Vagin A.A. Dodson E.J. Application of maximum likelihood refinement.In Refinement of Protein Structures. Proceedings of Daresbury Study Weekend. 1996; Google Scholar, 23Murshudov G.N. Vagin A.A. Dodson E.J. Refinement of macromolecular structures by the maximum-likelihood method.Acta Crystallogr. D Biol. Crystallogr. 1997; 53: 240-255Crossref PubMed Scopus (13851) Google Scholar, 29Murshudov G.N. Skubák P. Lebedev A.A. Pannu N.S. Steiner R.A. Nicholls R.A. Winn M.D. Long F. Vagin A.A. REFMAC5 for the refinement of macromolecular crystal structures.Acta Crystallogr. D Biol. Crystallogr. 2011; 67: 355-367Crossref PubMed Scopus (5948) Google Scholar). Shown in Table 1 are the cell parameters and solvent volumes of the four crystal forms with which we worked. The first two crystal forms in Table 1, Monocl-1 and Monocl-2, are very similar in cell dimensions and have the same space group, P21, the latter crystal being a somewhat dehydrated form of the first, having about 13% less volume. Monocl-3 is also of space group P21, but is otherwise unique, and it has significantly greater solvent volume, about 30% more. The last crystal form, Orthrhmb, is clearly different, having an even higher solvent content, about 32%, but it suffers from a limited resolution. In our earlier efforts, we tried to determine a structure for any and all of these crystals, but concentrated primarily on Monocl-1 and Monocl-3. With no progress being made using heavy atoms, we worked to find solutions using molecular replacement. NMR studies of APOC1 indicated that the protein probably consisted of an amino-terminal α-helix and a carboxy-terminal α-helix that, through a hinge sequence near the center, allowed the two helices to form an intramolecular, antiparallel duplex (3Rozek A. Buchko G.W. Cushley R.J. Conformation of two peptides corresponding to human apolipoprotein C-I residues 7–24 and 35–53 in the presence of sodium dodecyl sulfate by CD and NMR spectroscopy.Biochemistry. 1995; 34: 7401-7408Crossref PubMed Scopus (67) Google Scholar, 4Rozek A. Sparrow J.T. Weisgraber K.H. Cushley R.J. Sequence-specific 1H NMR resonance assignments and secondary structure of human apolipoprotein C-I in the presence of sodium dodecyl sulfate.Biochem. Cell Biol. 1998; 76: 267-275Crossref PubMed Scopus (12) Google Scholar). All of our crystal forms contained two molecules of APOC1 as the asymmetric unit, suggesting further that APOC1 might exist as dimers. These considerations, along with some others, implied that the arrangements of antiparallel helices likely existed in solution, and in our crystals, as four-helix bundles, a common motif in proteins. The best available models for such an arrangement, using molecules of comparable size, were the ROP protein (31Banner D.W. Kokkinidis M. Tsernoglou D. Structure of the ColE1 rop protein at 1.7 A resolution.J. Mol. Biol. 1987; 196: 657-675Crossref PubMed Scopus (277) Google Scholar) and later the ROM protein (32Jang S.B. Jeong M.S. Carter R.J. Holbrook E.L. Comolli L.R. Holbrook S.R. Novel crystal form of the ColE1 Rom protein.Acta Crystallogr. D Biol. Crystallogr. 2006; 62: 619-627Crossref PubMed Scopus (6) Google Scholar). Neither, however, had functions similar to APOC1 or were otherwise related. As already noted, molecular-replacement attempts with these probes did not yield unique solutions where side-chain features of APOC1 could be recognized, nor could they be convincingly refined. Without going into unnecessary detail, we recently resumed extensive molecular-replacement efforts using the program PHASER (18Read R.J. Pushing the boundaries of molecular replacement with maximum likelihood.Acta Crystallogr. D Biol. Crystallogr. 2001; 57: 1373-1382Crossref PubMed Scopus (785) Google Scholar, 19Storoni L.C. McCoy A.J. Read R.J. Likelihood-enhanced fast rotation functions.Acta Crystallogr. D Biol. Crystallogr. 2004; 60: 432-438Crossref PubMed Scopus (1094) Google Scholar, 20McCoy A.J. Grosse-Kunstleve R.W. Adams P.D. Winn M.D. Storoni L.C. Read R.J. Phaser crystallographic software.J. Appl. Crystallogr. 2007; 40: 658-674Crossref PubMed Scopus (14432) Google Scholar) from the CCP4 program package, again using four-helix bundle models and as many reasonable variations as we could conceive. All failed. Eventually we deduced, based mainly on intuition combined with observations on difference Fourier maps of our failures, that the APOC1 molecule was not composed of two antiparallel helical segments, but was a single, continuous helix along its entire length. With that realization, when polyalanine α-helices were properly placed, amino acid side chains appropriate to the sequence of APOC1 began to appear, and ultimately the entire APOC1 molecule emerged. Its correctness in Monocl-1 was confirmed not only by the appearance of the amino acid side chains, but by its refined agreement between calculated and observed structure amplitudes and by its use in solving, by molecular replacement, the other three crystal forms. The final refinement statistics for the structures of all crystals are presented in Table 3.TABLE 3Refinement and modelCrystalMonocl-1Monocl-2Monocl-3OrthrhmbNo. of Fs used6,9655,5857,7302,538Rfree test set350 (4.82%)404 (5.11%)532 (5.12%)130 (4.59%)Rworking0.21810.23330.23850.2025Rfree0.27700.28880.30960.2728Mean B factor19.022.925.262.5RMS bond Δ0.01030.01020.00910.0064RMS angle Δ1.22°1.24°1.30°1.03°RMS chrl. vol. Δ0.0650.0700.0740.058Rama outliers1: C terminus000Rotamer outliers0000TwinningNoNoYes: 0.15%NoTLSNoNoNoNoResolution, Å1.752.02.03.0chrl. vol., chiral volume; RMS, root mean square; TLS, translation, libration, screw. Open table in a new tab chrl. vol., chiral volume; RMS, root mean square; TLS, translation, libration, screw. The structure of the asymmetric unit of the crystals designated Monocl-1 is shown in Fig. 1, and some details showing the quality of its electron-density maps are found in Fig. 2. The structure consists of two basically identical long α-helices, closely associated with one another and aligned in an antiparallel fashion. The α-helices are 52–55 amino acids long, with some residues at the termini being apparently disordered in some crystal forms and not visible. The helices are about 75 Å long and have 13–14 helical turns. The helices are not completely straight but are somewhat bent so as to better engage one another and enhance stereochemical complementarity of side chains. The structure is clearly a natural dimer; it is not a consequence of crystal packing. The entire dimer, given the overlap of the monomers, is about 95 Å long. The total buried surface area on the two monomers comprising the dimer is 2,595 Å2, or about 15% of the total surface area (18,355 Å2) of the dimer (program PISA; CCP4).Fig. 2A: A central portion of one of the helices comprising an APOC1 monomer superimposed upon its electron density. The goodness of fit is evident and is typical of that for the entire structure. B: A similar illustration of electron density superimposed upon the model for the structure derived from crystals Monocl-3. C: A detail showing goodness of fit, featuring, at the right, the side chain of Trp41.View Large Image Figure ViewerDownload Hi-res image Download (PPT) On one monomer (A), on the amino-terminal half, there is a near-continuous stack of hydrophobic side chains. These are Leu11-Phe14-Leu18-gap-Leu25-Ileu29, with a gap between Leu18 and Leu25. On the other monomer, closely opposed to the stack on monomer A, is a different stack of hydrophobic side chains from the carboxy-terminal half of molecule B. This stack is Leu53-Val49-Phe46-Trp41-Phe42-Met38-Leu34. The two opposing hydrophobic stacks are not only in direct contact, but are further consolidated by the intercalation of part of the side chain of Trp41 into the gap of the opposite stack between Leu18 and Leu25. It is after Ile29 on monomer A and before Leu34 on monomer B that the two monomers begin to separate somewhat. This is illustrated in Fig. 3A. In addition to the hydrophobic interface, there are also some possible salt bridges and hydrogen bonds between the two monomers, but these must be considered tentative, as they involve flexible side chains. They are Glu13A-Lys52B-Lys21A-Thr45B-Lys50A-Glu19B and Ser35A-Lys30B. A network of hydrogen bonds could exist involving Asp20B-Glu47A-Arg23B-Ser43A. Most of the possible hydrophilic interactions are at the opposite end of the dimer, where the two monomers begin to separate from where the hydrophobic stacks interact and hold the monomers closely together. Electrostatic surface renderings of the dimer, seen in Fig. 4A, reveal an interesting distribution of hydrophobic surface and charged groups. In the orientation at the left in Fig. 4A, it is evident that one-half of the molecule is largely hydrophobic, punctuated here and there by positively charged lysine residues, whereas the other half of the dimer exhibits a cluster of negative charges. Positive charges are, however, also present on that half as well. When the molecule is rotated 180° to show the opposite side, at the right in Fig. 4A, a similar surface is seen. Again, negative charge clustered within one half, the other half almost entirely hydrophobic, with positive charges scattered throughout. The areas of concentrated charge and hydrophobicity switch halves on the two sides of the dimer. This is further apparent in the center orientation in Fig. 4A, where the view is perpendicular. The dimer from crystal Monocl-1, once known, was simply placed in the unit cell of Monocl-2 with the same coordinates; the two helices in the dimer were rigid body-refined and then subjected to restrained refinement (23Murshudov G.N. Vagin A.A. Dodson E.J. Refinement of macromolecular structures by the maximum-likelihood method.Acta Crystallogr. D Biol. Crystallogr. 1997; 53: 240-255Crossref PubMed Scopus (13851) Google Scholar, 33Collaborative Computational Project, Number 4 The CCP4 Suite: programs for protein crystallography.Acta Crystallogr. D. Biol. Crystallogr. 1994; 50: 760-763Crossref PubMed Scopus (19747) Google Scholar). The relevant refinement statistics are presented in Table 3. The individual monomers and the dimer appear virtually unchanged in the Monocl-2 crystals, but the two helices scissor slightly closer to one another by about 2 Å, the pivot being the hydrophobic cluster described above. The number of water molecules and their structure, as might have been expected from the volume change, are slightly different. Residues 5–8 are somewhat less ordered in Monocl-2 on one monomer, but on the other monomer, density is present for Val4, Asp3, and Pro2. These were not visible at all in the Monocl-1 crystal. Although the movement of the two monomers within the dimer toward one another is slight, it significantly increases the interactions between them. The buried surface area is enlarged an additional 7% to 4,050 Å2 and becomes 22% of the entire surface area (19,021 Å2) of the entire dimer. Both the dimers of Monocl-1 and Monocl-2 are predicted by the program PISA (CCP4 System) to be stable dimers in solution. Monocl-2 represents the dimer exhibiting the greatest degree of compression of the constituent monomer helices. As in Monocl-1, the hydrophobic stack interaction at one end of the
Referência(s)