Lists of other proteins that can be used to form NDs, including other apolipoproteins, saposin A
(see § 3.4.4), α-synuclein (Varkey et al. 2013; Eichmann et al. 2016, 2017), apolipophorins, and
apomyoglobin, are given in Viegas et al. (2016) and Denisov and Sligar (2017). Projects are under way
to develop amphipathic peptides designed to stabilize lipid bilayer patches (see § 3.4.5). In a recent
development, disulfide bond formation protein B (DsbB), a bacterial plasma membrane protein
featuring four TM α-helices, has been expressed in a soluble, functional form in the cytosol of
E. coli after being fused to a truncated version of ApoA-I (Mizrachi et al. 2017). The stabilization of
lipid discs and MPs by styrene-maleic acid copolymers (SMAs) will be discussed in Chaps. 4 and 5.
3.3.3
The Empty Nanodisc
The composition, organization, and dynamics of MP-free NDs have been extensively studied. Among
the major questions investigated were the arrangement of the protein and the degree to which the
behavior of the lipids faithfully reproduces that in protein-free lipid bilayers. The initial controversy
between the “picket fence” model (Phillips et al. 1997), in which short helices have their axis normal to
the plane of the disc (see Fig. 3.6A), and “belt” models (Fig. 3.6B), in which long helices run around its
periphery (Wlodawer et al. 1979; Segrest et al. 1999), has been definitely resolved in favor of the latter
by a host of indirect experimental data, MD simulations (Fig. 3.8), and, finally, direct experimental
determination by NMR (Fig. 3.9) (see e.g. Koppaka et al. 1999; Li et al. 2006; Bibow et al. 2017;
reviewed in Brouillette et al. 2001; Denisov and Sligar 2017). α-Helices are intrinsically rigid due to
the network of hydrogen bonds that stabilizes them. Bends therefore occur preferentially at the weak
Fig. 3.8 Membrane protein-free nanodiscs (NDs) as simulated by molecular dynamics (MD). (A–D) Side
views of NDs after 4.2 ns of all-atom simulation. Each ND is comprised of 160 molecules of DPPC and
two MSP molecules, respectively, MSP1 (A; 200 residues), MSP1 Δ(1–11) (MSP1D1; B; 189 residues),
MSP1 Δ(1–22) (MSPID2; C; 178 residues), and MSP1 Δ(1–22)g (D; same sequence as in C but with the
N-and C-termini of the two MSPs aligned, leaving a gap). MSPs are depicted in tube representation in blue
and red. Prolines are highlighted in sphere representation in yellow and green. Lipid head groups are
shown in orange and acyl chains in gray. (A
0 –D
0 ). Top views. Lipids have been removed to reveal
deviations from perfect circularity in the structure of MSPs, most noticeably the bends induced by the
proline residues (From Shih et al. 2005, # 2005 The Biophysical Society. Published by Elsevier Inc. All
rights reserved. A dashed circle has been fitted around MSPs in Panel B
0 to facilitate comparison with
Fig. 3.13B).
3.3 Nanodiscs
109
(see § 3.4.4), α-synuclein (Varkey et al. 2013; Eichmann et al. 2016, 2017), apolipophorins, and
apomyoglobin, are given in Viegas et al. (2016) and Denisov and Sligar (2017). Projects are under way
to develop amphipathic peptides designed to stabilize lipid bilayer patches (see § 3.4.5). In a recent
development, disulfide bond formation protein B (DsbB), a bacterial plasma membrane protein
featuring four TM α-helices, has been expressed in a soluble, functional form in the cytosol of
E. coli after being fused to a truncated version of ApoA-I (Mizrachi et al. 2017). The stabilization of
lipid discs and MPs by styrene-maleic acid copolymers (SMAs) will be discussed in Chaps. 4 and 5.
3.3.3
The Empty Nanodisc
The composition, organization, and dynamics of MP-free NDs have been extensively studied. Among
the major questions investigated were the arrangement of the protein and the degree to which the
behavior of the lipids faithfully reproduces that in protein-free lipid bilayers. The initial controversy
between the “picket fence” model (Phillips et al. 1997), in which short helices have their axis normal to
the plane of the disc (see Fig. 3.6A), and “belt” models (Fig. 3.6B), in which long helices run around its
periphery (Wlodawer et al. 1979; Segrest et al. 1999), has been definitely resolved in favor of the latter
by a host of indirect experimental data, MD simulations (Fig. 3.8), and, finally, direct experimental
determination by NMR (Fig. 3.9) (see e.g. Koppaka et al. 1999; Li et al. 2006; Bibow et al. 2017;
reviewed in Brouillette et al. 2001; Denisov and Sligar 2017). α-Helices are intrinsically rigid due to
the network of hydrogen bonds that stabilizes them. Bends therefore occur preferentially at the weak
Fig. 3.8 Membrane protein-free nanodiscs (NDs) as simulated by molecular dynamics (MD). (A–D) Side
views of NDs after 4.2 ns of all-atom simulation. Each ND is comprised of 160 molecules of DPPC and
two MSP molecules, respectively, MSP1 (A; 200 residues), MSP1 Δ(1–11) (MSP1D1; B; 189 residues),
MSP1 Δ(1–22) (MSPID2; C; 178 residues), and MSP1 Δ(1–22)g (D; same sequence as in C but with the
N-and C-termini of the two MSPs aligned, leaving a gap). MSPs are depicted in tube representation in blue
and red. Prolines are highlighted in sphere representation in yellow and green. Lipid head groups are
shown in orange and acyl chains in gray. (A
0 –D
0 ). Top views. Lipids have been removed to reveal
deviations from perfect circularity in the structure of MSPs, most noticeably the bends induced by the
proline residues (From Shih et al. 2005, # 2005 The Biophysical Society. Published by Elsevier Inc. All
rights reserved. A dashed circle has been fitted around MSPs in Panel B
0 to facilitate comparison with
Fig. 3.13B).
3.3 Nanodiscs
109