High-density lipoprotein (HDL) particles are cholesterol and lipid transport containers. Mature HDL particles destined for the liver develop through the formation of intermediate discoidal HDL particles, which are the primary acceptors for cholesterol. Here we present the three-dimensional structure of reconstituted discoidal HDL (rdHDL) particles, using a shortened construct of human apolipoprotein A-I, determined from a combination of nuclear magnetic resonance (NMR), electron paramagnetic resonance (EPR) and transmission electron microscopy (TEM) data. The rdHDL particles feature a protein double belt surrounding a lipid bilayer patch in an antiparallel fashion. The integrity of this structure is maintained by up to 28 salt bridges and a zipper-like pattern of cation-π interactions between helices 4 and 6. To accommodate a hydrophobic interior, a gross 'right-to-right' rotation of the helices after lipidation is necessary. The structure reflects the complexity required for a shuttling container to hold a fluid lipid or cholesterol interior at a protein:lipid ratio of 1:50.
Subscribe to Journal
Get full journal access for 1 year
only $8.25 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Tax calculation will be finalised during checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
Zannis, V.I. et al. HDL biogenesis, remodeling, and catabolism. Handb. Exp. Pharmacol. 224, 53–111 (2015).
Kingwell, B.A. & Chapman, M.J. Future of high-density lipoprotein infusion therapies: potential for clinical management of vascular disease. Circulation 128, 1112–1121 (2013).
Navab, M., Reddy, S.T., Van Lenten, B.J. & Fogelman, A.M. HDL and cardiovascular disease: atherogenic and atheroprotective mechanisms. Nat. Rev. Cardiol. 8, 222–232 (2011).
Kingwell, B.A., Chapman, M.J., Kontush, A. & Miller, N.E. HDL-targeted therapies: progress, failures and future. Nat. Rev. Drug Discov. 13, 445–464 (2014).
Kontush, A. et al. Structure of HDL: particle subclasses and molecular components. Handb. Exp. Pharmacol. 224, 3–51 (2015).
Rogers, D.P. et al. Truncation of the amino terminus of human apolipoprotein A-I substantially alters only the lipid-free conformation. Biochemistry 36, 288–300 (1997).
Segrest, J.P., Garber, D.W., Brouillette, C.G., Harvey, S.C. & Anantharamaiah, G.M. The amphipathic alpha helix: a multifunctional structural motif in plasma apolipoproteins. Adv. Protein Chem. 45, 303–369 (1994).
Segrest, J.P. et al. The amphipathic helix in the exchangeable apolipoproteins: a review of secondary structure and function. J. Lipid Res. 33, 141–166 (1992).
Gursky, O. Crystal structure of Δ(185-243)ApoA-I suggests a mechanistic framework for the protein adaptation to the changing lipid load in good cholesterol: from flatland to sphereland via double belt, belt buckle, double hairpin and trefoil/tetrafoil. J. Mol. Biol. 425, 1–16 (2013).
Denisov, I.G., Grinkova, Y.V., Lazarides, A.A. & Sligar, S.G. Directed self-assembly of monodisperse phospholipid bilayer Nanodiscs with controlled size. J. Am. Chem. Soc. 126, 3477–3487 (2004).
Silva, R.A.G.D. et al. Structure of apolipoprotein A-I in spherical high density lipoproteins of different sizes. Proc. Natl. Acad. Sci. USA 105, 12176–12181 (2008).
Asztalos, B.F. et al. Role of LCAT in HDL remodeling: investigation of LCAT deficiency states. J. Lipid Res. 48, 592–599 (2007).
Zannis, V.I., Chroni, A. & Krieger, M. Role of apoA-I, ABCA1, LCAT, and SR-BI in the biogenesis of HDL. J. Mol. Med. 84, 276–294 (2006).
Gauthamadasa, K. et al. Speciated human high-density lipoprotein protein proximity profiles. Biochemistry 49, 10656–10665 (2010).
Bruce, C., Chouinard, R.A. Jr. & Tall, A.R. Plasma lipid transfer proteins, high-density lipoproteins, and reverse cholesterol transport. Annu. Rev. Nutr. 18, 297–330 (1998).
Qiu, X. et al. Crystal structure of cholesteryl ester transfer protein reveals a long tunnel and four bound lipid molecules. Nat. Struct. Mol. Biol. 14, 106–113 (2007).
Huang, R. et al. Apolipoprotein A-I structural organization in high-density lipoproteins isolated from human plasma. Nat. Struct. Mol. Biol. 18, 416–422 (2011).
Hagn, F., Etzkorn, M., Raschle, T. & Wagner, G. Optimized phospholipid bilayer nanodiscs facilitate high-resolution structure determination of membrane proteins. J. Am. Chem. Soc. 135, 1919–1925 (2013).
Bibow, S. et al. Measuring membrane protein bond orientations in nanodiscs via residual dipolar couplings. Protein Sci. 23, 851–856 (2014).
Frey, L., Lakomek, N.A., Riek, R. & Bibow, S. Micelles, bicelles, and nanodiscs: comparing the impact of membrane mimetics on membrane protein backbone dynamics. Angew. Chem. Int. Ed. Engl. http://dx.doi.org/10.1002/anie.201608246 (2016).
Salzmann, M., Wider, G., Pervushin, K., Senn, H. & Wuthrich, K. TROSY-type triple-resonance experiments for sequential NMR assignments of large proteins. J. Am. Chem. Soc. 121, 844–848 (1999).
Bellstedt, P. et al. Resonance assignment for a particularly challenging protein based on systematic unlabeling of amino acids to complement incomplete NMR data sets. J. Biomol. NMR 57, 65–72 (2013).
Spera, S. & Bax, A. Empirical correlation between protein backbone conformation and Cα and Cβ13C nuclear magnetic resonance chemical shifts. J. Am. Chem. Soc. 113, 5490–5492 (1991).
Wishart, D.S. & Sykes, B.D. Chemical shifts as a tool for structure determination. Methods Enzymol. 239, 363–392 (1994).
Oda, M.N. et al. The secondary structure of apolipoprotein A-I on 9.6-nm reconstituted high-density lipoprotein determined by EPR spectroscopy. FEBS J. 280, 3416–3424 (2013).
Zhou, N.E., Zhu, B.Y., Sykes, B.D. & Hodges, R.S. Relationship between amide proton chemical-shifts and hydrogen-bonding in amphipathic α-helical peptides. J. Am. Chem. Soc. 114, 4320–4326 (1992).
Blanco, F.J. et al. NMR chemical shifts: a tool to characterize distortions of peptide and protein helices. J. Am. Chem. Soc. 114, 9676–9677 (1992).
Bracken, C., Carr, P.A., Cavanagh, J. & Palmer, A.G. III. Temperature dependence of intramolecular dynamics of the basic leucine zipper of GCN4: implications for the entropy of association with DNA. J. Mol. Biol. 285, 2133–2146 (1999).
Kuntz, I., Kosen, P. & Craig, E. Amide chemical shifts in many helices in peptides and proteins are periodic. J. Am. Chem. Soc. 113, 1406–1408 (1991).
Lakomek, N.A., Ying, J. & Bax, A. Measurement of 15N relaxation rates in perdeuterated proteins by TROSY-based methods. J. Biomol. NMR 53, 209–221 (2012).
Lakomek, N.A. et al. Internal dynamics of the homotrimeric HIV-1 viral coat protein gp41 on multiple time scales. Angew. Chem. Int. Ed. Engl. 52, 3911–3915 (2013).
Segrest, J.P. et al. A detailed molecular belt model for apolipoprotein A-I in discoidal high density lipoprotein. J. Biol. Chem. 274, 31755–31758 (1999).
Phillips, M.C. New insights into the determination of HDL structure by apolipoproteins. J. Lipid Res. 54, 2034–2048 (2013).
Thomas, M.J., Bhat, S. & Sorci-Thomas, M.G. Three-dimensional models of HDL apoA-I: implications for its assembly and function. J. Lipid Res. 49, 1875–1883 (2008).
Borhani, D.W., Rogers, D.P., Engler, J.A. & Brouillette, C.G. Crystal structure of truncated human apolipoprotein A-I suggests a lipid-bound conformation. Proc. Natl. Acad. Sci. USA 94, 12291–12296 (1997).
Mason, J.M. & Arndt, K.M. Coiled coil domains: stability, specificity, and biological implications. ChemBioChem 5, 170–176 (2004).
Nolte, R.T. & Atkinson, D. Conformational analysis of apolipoprotein A-I and E-3 based on primary sequence and circular dichroism. Biophys. J. 63, 1221–1239 (1992).
Mei, X. & Atkinson, D. Crystal structure of C-terminal truncated apolipoprotein A-I reveals the assembly of high density lipoprotein (HDL) by dimerization. J. Biol. Chem. 286, 38570–38582 (2011).
Bashtovyy, D., Jones, M.K., Anantharamaiah, G.M. & Segrest, J.P. Sequence conservation of apolipoprotein A-I affords novel insights into HDL structure-function. J. Lipid Res. 52, 435–450 (2011).
Sorci-Thomas, M.G., Bhat, S. & Thomas, M.J. Activation of lecithin: cholesterol acyltransferase by HDL ApoA-I central helices. Clin. Lipidol. 4, 113–124 (2009).
Glukhova, A. et al. Structure and function of lysosomal phospholipase A2 and lecithin: cholesterol acyltransferase. Nat. Commun. 6, 6250 (2015).
Sorci-Thomas, M.G. & Thomas, M.J. The effects of altered apolipoprotein A-I structure on plasma HDL concentration. Trends Cardiovasc. Med. 12, 121–128 (2002).
Clore, G.M. & Garrett, D.S. R-factor, free R, and complete cross-validation for dipolar coupling refinement of NMR structures. J. Am. Chem. Soc. 121, 9008–9012 (1999).
Marley, J., Lu, M. & Bracken, C. A method for efficient isotopic labeling of recombinant proteins. J. Biomol. NMR 20, 71–75 (2001).
Gans, P. et al. Stereospecific isotopic labeling of methyl groups for NMR spectroscopic studies of high-molecular-weight proteins. Angew. Chem. Int. Ed. Engl. 49, 1958–1962 (2010).
Lee, D., Hilty, C., Wider, G. & Wüthrich, K. Effective rotational correlation times of proteins from NMR relaxation interference. J. Magn. Reson. 178, 72–76 (2006).
Shen, Y. & Bax, A. Protein backbone and sidechain torsion angles predicted from NMR chemical shifts using artificial neural networks. J. Biomol. NMR 56, 227–241 (2013).
Battiste, J.L. & Wagner, G. Utilization of site-directed spin labeling and high-resolution heteronuclear nuclear magnetic resonance for global fold determination of large proteins with limited nuclear overhauser effect data. Biochemistry 39, 5355–5365 (2000).
Permi, P., Rosevear, P.R. & Annila, A. A set of HNCO-based experiments for measurement of residual dipolar couplings in 15N,13C,(2H)-labeled proteins. J. Biomol. NMR 17, 43–54 (2000).
Pannier, M., Veit, S., Godt, A., Jeschke, G. & Spiess, H.W. Dead-time free measurement of dipole-dipole interactions between electron spins. 2000. J. Magn. Reson. 213, 316–325 (2011).
Polyhach, Y. et al. High sensitivity and versatility of the DEER experiment on nitroxide radical pairs at Q-band frequencies. Phys. Chem. Chem. Phys. 14, 10762–10773 (2012).
Jeschke, G. et al. DeerAnalysis2006—a comprehensive software package for analyzing pulsed ELDOR data. Appl. Magn. Reson. 30, 473–498 (2006).
Güntert, P. Automated NMR structure calculation with CYANA. Methods Mol. Biol. 278, 353–378 (2004).
Duss, O., Yulikov, M., Jeschke, G. & Allain, F.H. EPR-aided approach for solution structure determination of large RNAs or protein-RNA complexes. Nat. Commun. 5, 3669 (2014).
Polyhach, Y., Bordignon, E. & Jeschke, G. Rotamer libraries of spin labelled cysteines for protein studies. Phys. Chem. Chem. Phys. 13, 2356–2366 (2011).
Jeschke, G. Conformational dynamics and distribution of nitroxide spin labels. Prog. Nucl. Magn. Reson. Spectrosc. 72, 42–60 (2013).
Gallivan, J.P. & Dougherty, D.A. Cation-π interactions in structural biology. Proc. Natl. Acad. Sci. USA 96, 9459–9464 (1999).
Tang, G. et al. EMAN2: an extensible image processing suite for electron microscopy. J. Struct. Biol. 157, 38–46 (2007).
Scheres, S.H. RELION: implementation of a Bayesian approach to cryo-EM structure determination. J. Struct. Biol. 180, 519–530 (2012).
We thank N.A. Lakomek and S.A. Müller for insightful discussions and critical reading of the manuscript. This work was supported by the ETH fellowship program, financed by the ETH Zurich and the EU COFUND program under award number FEL-2012-1 (to S.B.); the Swedish research council (Ventenskap Rådet VR-2011-13 to C.N.C.); Wenner-Gren Stiftelsern WG-13 (to C.N.C.); the NCCR TransCure (H.S.); the SNF (grant 205320_166164 to H.S., grant 200020_157034 to G.J., and grant 144444 to R.R.); and Eurostars grant E! 9803 funded by the State Secretariat for Education, Research and Innovation of the Swiss Confederation (to P.G.). Plasmid pET-28a was a generous gift from the Zerbe lab (University of Zurich, Zurich, Switzerland) given with permission from G. Wagner (Harvard University, Cambridge, Massachusetts, USA).
The authors declare no competing financial interests.
Integrated supplementary information
(a) Overlay of two TROSY-HNCO spectra of ApoA-I with helix 5 (red spectrum) and without helix 5 (black spectrum). Virtually identical peak shifts are found for the majority of the residues indicating the same arrangement of both ApoA-I helices irrespective of helix 5. (b) Chemical shift differences of H-N moieties were calculated according to the formula in the inset. The threshold was set to 20 Hz accounting for the increased molecular weight (ca. 158 kDa for ApoA-I containing helix 5 vs. 108 kDa for ApoA-I without helix 5) and the corresponding difference in helix bending due to a larger circumference (see panel (c)). As a comparison, the chemical shift difference found for the K106-K107 peak in the upper panel of (a) is 25 Hz. Significant chemical shift differences are found N-terminal of helix 5, at the beginning of helix 6, as it would be expected, since the ‘insertion’ of helix 5 produces a different spin environment. Signals from residues C-terminal of helix 5 are missing or could not be assigned. Furthermore, chemical shift changes for the linkers between helices 6 and 7 as well as helices 7 and 8 are found. (c) The amide secondary chemical shifts for ApoA-I with helix 5 (red dots) and without helix 5 (black dots) are shown along the amino acid sequence. The wave pattern is attributed to the curvature of the helices (see extended data figure 8 and supplementary information for more details). The pronounced peak overlap indicates the same orientation of the helices. However, the expected reduced bending of helices due to a larger circumference of the larger ApoA-I construct (red dots) is visible and most pronounced for helix 6. The red dots are on average closer to zero, which is a clear indication that hydrogen bond lengths are shorter for the convex solvent exposed side of the helix and longer for the concave lipid-facing side of the helix. This would also explain the on average higher chemical shift differences found for helix 6. Both constructs are in the same RR orientation, since a change from a solvent exposed position towards the protein-protein interface would produce very pronounced chemical shift changes for all affected residues along the entire sequence. Similarly, a registry change upon helix 5 incorporation would produce a different environment for all residues located at the dimer interface. Since we do not observe this pattern of large chemical shifts throughout the sequence, but only localized at the linkers, the rotation and registry of both ApoA-I helices is independent of the presence of helix 5. Hence, HDL particles assembled using MSPdH5 are representative for the full-length ApoA-I. (d) Cα secondary chemical shifts are presented for ApoA-I without helix 5(black bars) and ApoA-I containing helix 5 (red dots). The difference of both shifts is shown as blue squares. (e) Intensity ratio of rdHDL particles without (I0) and with Gadolinium tagged PE-DTPA (IPE-DTPA) for ApoA-I without helix 5 (black bars) and ApoA-I containing helix 5 (red dots). The color code in the background is according to the coloring in Figure 4e in the main manuscript. (f) Size exclusion profile (upper panel) using a S200 column and SDS gel (lower left panel) and native PAGE (lower right panel). As indicated only the main peak was used for NMR experiments.
rdHDLs (white) were visible as single rings, double rings or disc-on-disc arrangement, the latter having been artificially induced during the negative stain grid preparation process. Scale bar is 50 nm.
Supplementary Figure 3 2D planes of triple-resonance NMR spectroscopy experiments to show the spectral resolution of rdHDL.
(a) 2D plane with a 15N frequency of 120.2 ppm of a 3D tr-HNCA of uniformly 13C,15N,2H-labelled rdHDL. (b) The corresponding 2D plane of a 3D tr-HNCO spectra of uniformly 13C,15N,2H-labelled rdHDL. The cross peaks are labelled with the corresponding sequential assignment.
(a) Sequence-specific 13Cα secondary chemical shifts (δΔ13Cα) along the amino acid sequence of MSPΔH5 (black bars) overlaid with the TALOS-N secondary structure prediction probability (pink points) based on 1HN, 15N, 13CO and 13Cα chemical shifts. Both entities suggest a helical secondary structure throughout the protein. δΔ13Cα is the difference between measured chemical shift and the corresponding random coil chemical shift of a perdeuterated protein at a temperature of 42°C and a pH of 7.4 determined using the database from the University of Copenhagen (see Materials and Methods). Positive δΔ13Cα values indicate an α-helical conformation. (b) [1H]15N-hetero-nuclear NOE data (hetNOE) along the amino acid sequence (black bars). Negative or near zero values indicate highly flexible 15N-1H moieties with motions faster than ca. 1 ns. The error of the measurement is shown as red markers, whereas the back-calculated values from Fast ModelFree are shown as blue markers. Values close to 1 indicate the absence of such motions. (c) Longitudinal 15N R1 relaxation rates (black bars) versus the amino acid sequence. 15N-1H moieties with larger R1 rates indicate the presence of higher flexibility in the ps-ns time range. In particular, the termini show some mobility. The error of the measurement is shown as red markers, whereas the back-calculated values from Fast ModelFree are shown as blue markers. (d) Transverse 15N R2 (R1ρ) relaxation rates versus the amino acid sequence (black bars) recorded with a spin-lock field of 2 kHz, which effectively removed exchange contributions slower than 80 μs. The error of the measurement is shown as red markers, whereas the back-calculated values from Fast ModelFree are shown as blue markers. (e) Representation of the calculated order parameter S2 (left axis) and internal correlation time τe (right axis). Both values were calculated using relax (green markers for the order parameter and green bars for the internal correlation time) and Fast ModelFree (blue markers for the order parameter and blue bars for the internal correlation time). The back-calculated hetNOE, R1 and R1ρ values from Fast Modelfree are used for panels (b)-(e). (f) Transverse relaxation 15N R2β relaxation rates versus the amino acid sequence measured on the slowly relaxing β-component at 900 (black circles) and 700 (red circles) MHz 1H frequency. A few 15N-1H moieties (such as at the N-terminus) have faster relaxation rates at 900 MHz than at 700 MHz, indicating the presence of local chemical exchange. (g) Transverse cross-correlated relaxation, ηxy, was determined as described previously and converted into the rotational correlation time in a similar way as with the TRACT experiment. An average correlation time τc of 34 ns was determined, identical to previous TRACT measurements. The homogeneity of τc along the sequence indicates uniform tumbling as one entity. (h) Presaturation experiments were performed on a 900 MHz spectrometer (black bars) and a 600 MHz spectrometer (red dots) using a presaturation time of 1 sec. Two different protein batches were used to test the reproducibility. Error bars in panel b, c, d and f correspond to one standard deviation from 30 single-exponential fits of the experimental relaxation rates.
(a) The amide secondary chemical shifts (δΔ1HN), i.e. the experimental chemical shifts of rdHDL minus the corresponding ‘random coil’ shifts, are shown along the amino acid sequence. The black connected dots are experimental data, while the red lines symb›olize the wave-like pattern of the helices. In addition, Pro residues with their sequential neighbor i+1 as well as aromatic residues are specifically labeled, since aromatic ring-currents and prolines can change amide proton shifts. The dashed lines indicate residues with the assigned heptad letter ‘a’ and ‘d’. The wave pattern is attributed to the curvature of the helices, allowing the solvent and lipid-facing side of the helices to be determined. In the proposed modified heptad sequence (abc[c’]defg), letters ‘a’ and ‘d’ are pointing towards a hydrophobic interface (i.e., the lipid bilayer). (b) the amino acid sequence of MSPΔH5 with the highlighted heptad sequence (abc[c’]defg), with ‘a’ and ‘d’ underlined. (c) The amide secondary chemical shifts (δΔ1HN) of rdHDL in DMPC (black) and POPC (red) overlap almost perfectly, indicating a similarity of the structure, regardless of the used lipid.
Supplementary Figure 6 Paramagnetic relaxation enhancement (PRE) broadening profiles for the collection of intra- and intermolecular distance restraints.
(a-h) Paramagnetic relaxation enhancement (PRE) broadening profiles versus the amino acid sequence are shown for various MTSL-Cys variants of rdHDL as indicated. Intensity ratios were calculated by comparing peak intensities from a sample tagged with paramagnetic MTSL and the corresponding peak intensities from a sample tagged with a diamagnetic MTSL analog (control). (a) The measurements of intra-molecular PRE broadening were established using a sample with 15N-2H-MTSL-Cys67-MSPΔH5 assembled with 14N-1H-wildtype-MSPΔH5 in a ratio of 1:4. (b)-(h) Intermolecular PRE broadenings to probe the intermolecular arrangement of two MSPΔH5 copies in an rdHDL particle were established by assembling 14N-1H-MTSL-Cys-variant-MSPΔH5 and 15N-2H-wildtype-MSPΔH5 in a ratio of 3:1. (i) Cartoon representation of the rotamer cloud of MTSL. The center of the rotamer cloud was calculated as the population-weighted mean coordinate of the nitroxide moiety that is represented by a dummy Cα atom whose distance is fixed to the protein backbone by 4 restraints (N, C’, Cα and O of the respective cysteine-mutated residue). Intra- and Intermolecular distance restraints were calculated between amide protons and the Cα dummy atom representing the N-O nitroxide moiety. Error bars were calculated with Gaussian error propagation using the signal-to-noise ratios from the dia- and paramagnetic samples.
EPR measurements were used to obtain very long-range distance restraints. These are imperative for the definition of the discoidal shape and the confirmation of a fixed arrangement of the MSPΔH5 molecules within the rdHDL assembly. (a)-(f) Left panels: Intensity-normalized, phase-corrected primary DEER data (i.e. data before background correction). Experimental time traces (shown as a black line in the left panel) were acquired using the Q-band DEER experiment. Middle panels: corresponding DEER distance distributions for paramagnetic MTSL attached to the indicated residue. Experimental distance distributions (black line in the corresponding middle panel) were computed using Tikhonov regularization, as implemented in DeerAnalysis, assuming a homogeneous spatial distribution of the observer spins (3D background model). The light blue lines in the middle panels represent the ensemble averaged simulated distance distribution using the 10 structures with the lowest CYANA target function. The dotted red line in the middle panels represent the ensemble averaged distances found within the 10 conformers of our structure between the Cα dummy atoms representing the N-O nitroxide moiety determined by the rotamer approach. Right panels: error estimation for the experimental distance distributions performed with the Validation tool of DeerAnalysis (here possible distance distributions are confined between the dark blue and the red lines. Black lines represent corresponding generalized normalized distributions. Grey areas indicate an extent of possible distance distribution models seen during the error computation. The light blue lines in the right panels represent simulated distance distributions for each of the 10 structures with the lowest CYANA target function. Simulations were done with the rotamer library approach as implemented in the MMM package. The rotamer library R1A_298K_UFF_216_CASD.mat was used. (g)-(i) The Q-band DEER traces were acquired on the mutant with MTSL at position 213 with full (black) and strongly reduced (blue) inversion efficiency of the pump pulse to determine whether more than two MSP monomers assemble to rdHDL particles (see Supplementary Information for more details). (g) Normalized and phase corrected primary DEER traces. (h) DEER form factors. (i) Overlay of two form factors after normalization to the same modulation depth. They correspond perfectly. Experimental and simulated distance distribution show a good match. The broadest distance distribution is obtained between the more flexible sites 67. For sites 166 an apparent long-distance shift is fully within the experimental uncertainty. A long-distance shift between the experimental and simulated distributions for sites 213 can partially be explained by overweighted simulated rotamers, which extend fully into the solvent phase. Simulation of the distribution of rotamers at the solvent-lipid interface is complicated and has, to our knowledge, not been done yet. In the rotamer library, no distinction is made between spin labels that experience a homogeneous environment (solvent or lipid) or inhomogeneous environment (e.g. located at the lipid-solvent interface), resulting in a larger uncertainty of the predicted rotamer distribution for such a case compared to the case of a homogeneous environment.
(a) [1H,13C] HMQC of the methyl region showing the proS methyls of Leucine and Valines (note that the sequence does not contain Isoleucines). Selective labeling was achieved following the approach from the Boisbouvier group (see Materials and Methods section for more details). (b) 2D plane of a 3D CHH-NOESY-HMQC using the methyl-selective labeled sample to determine intra- and intermolecular methyl-methyl NOEs. (c) Excerpt from 2D [1H,15N]-TROSY spectra which were measured to determine intermolecular PRE effects. In order to do so, 14N-1H-MTSL-Cys-variant-MSPΔH5 (here in this case 213,192,148) and 15N-2H-wildtype-MSPΔH5 were assembled. Since only the 15N-2H-wildtype-MSPΔH5 is NMR-visible which has no MTSL tag, signal broadening must stem from the other NMR-invisible but MTSL-tagged molecule and therefore intermolecular. As it can be seen, several residues experience intermolecular PRE broadening. (d) Long range distance restraints (blue lines) collected from PRE data and NOEs are mapped onto the 3D structure of rdHDL.
Supplementary Figure 9 Structural reorganization of apoA-I from its soluble to its lipid-bound state after rdHDL formation using LL and RR color-coding.
(a) The amphipathic helices of ApoA-I can also be represented by an 11-mer repeat (instead of a heptad repeat) with prolines marking the beginning of each helix, which are therefore assigned the helical wheel position 1. Residues ‘2’, ‘5’ and ‘9’ are the ‘left’ residues, whereas residues ‘4’, ‘7’ and ‘11’ are designated the ‘right’ residues. The amphipathic helices of ApoA-I rotate upon rdHDL formation either via a ‘left to left’ (LL) or ‘right to right’ (RR) rotation of the helices to expose their hydrophobic residues (in this case residues ‘1’ and ‘8’) to the lipid bilayer. Residues that are designated as ‘left’ (forming the dimer interface upon a LL rotation) are colored in magenta whereas residues designated as ’right’ (forming the dimer interface upon a RR rotation) are colored in green. (b) Structure of the lipid-containing rdHDL particle. The two helices rotate RR (green dimer interface), exposing the hydrophobic residues to the lipid bilayer and the ‘left’ residues (magenta) to the solvent. (c) Rigid-body docking using an unrestricted grid search results in the RR rotation already when using only the 6 EPR long-range restraints. For docking, we used the apoA-I monomer structure derived from our CYANA structure calculation and assumed it to be fully rigid. No NMR restraints between the two apoA-I protomers were included and no test for clashes between the protomers was made, demonstrating that the coarse-grained arrangement (RR or LL) can be deduced from a small fraction of the restraints used in the full CYANA structure calculation (see supplementary note 1 for further information). (d) Crystal structure of lipid-free dimeric ApoA-I (pdb-code 3R2P). The upper illustration shows the proposed solvent-oriented side of the structure, whereas the lower illustration shows the proposed lipid-facing side of the structure. In this lipid-free assembly hydrophobic residues occupy the dimer interface, preventing a LL or RR rotation. (e) Crystal structure of lipid-free tetrameric ApoA-I (pdb-code 1AV1). The formation of a dimer of dimers results in a LL rotation for the two dimers (strip of residues colored in magenta), whereas the dimerization of both dimers is realized via a RR rotation (strip of residues colored in green). This assures that the hydrophobic residues of all four ApoA-I molecules are buried inside the tetramer and are not solvent exposed.
Supplementary Figure 10 Amide secondary chemical shifts of rdHDL and structural clustering of functionally relevant familial mutations of apoA-I.
(a) Residues with a positive amide secondary chemical shift Δδ1HN (see Fig 4 in main text) are colored yellow on top of a surface representation of the 3D structure of rdHDL, which is cut in half to show the interior. The surface representation of the conformer with the lowest CYANA target function is shown. A positive amide secondary chemical shift Δδ1HN is indicative of a shorter hydrogen bond, which is expected to be located in the inner side of the MSPΔH5 molecules of the rdHDL. (b) A surface representation of the 3D structure of the rdHDL is shown. Known apoA-I mutations that do not affect HDL plasma concentrations are mapped onto the upper apoA-I molecule in blue, while mutations affecting HDL plasma concentrations are mapped in red that cluster structurally.
About this article
Cite this article
Bibow, S., Polyhach, Y., Eichmann, C. et al. Solution structure of discoidal high-density lipoprotein particles with a shortened apolipoprotein A-I. Nat Struct Mol Biol 24, 187–193 (2017). https://doi.org/10.1038/nsmb.3345
Assembly of phospholipid nanodiscs of controlled size for structural studies of membrane proteins by NMR
Nature Protocols (2018)
Journal of Molecular Medicine (2018)