In Table 12.1 are given references to four further studies presenting NS EM data on MP/NVoy
(Study 12.7; cf. Chap. 7, Fig. 7.7A, B), MP/non-ionic APol (Study 12.8), and MP/A8-35 (Studies
12.25 and 12.28) complexes. A recent atomic-force microscopy study of MotPS/A8-35 complexes
should also be mentioned (Terahara et al. 2017).
12.3.2.1 Conclusions from Negative-Stain Studies
From this survey of the most informative studies in which TEM was used to examine negatively
stained MP/APol complexes, one can tentatively draw the following conclusions:
• MP/APol (and MP/SMA) complexes are amenable to such studies, and 3D models can be
routinely reconstructed to ~16 to 18-Å resolution (Cvetkov et al. 2011; Kevany et al. 2013;
Tsybovsky et al. 2013; Huynh et al. 2014b) or close to it (Sverzhinsky et al. 2014; Postis et al.
2015), that is to the best resolution that can be achieved using NS.
• Attention should be paid to the fact that removal of free APol may induce the formation of
aggregates, which, in the two studies where they have been examined by TEM/NS (Gohon
et al. 2008; Arunmanee et al. 2014), take the form of filaments, occasionally of sheets
(Arunmanee et al. 2014). Whether these structures form, and at which rate, is likely to be
strongly dependent on the MP, on the nature of the APol, and on conditions such as the
residual concentration of free APol and/or detergent, the ionic strength, and the presence of
divalent cations. Note that the presence of traces of divalent cations degrades the quality of
NMR spectra of MP/A8-35 complexes (Catoire et al. 2009, 2010), certainly due to calciummediated association of some of the complexes via their APol belts (see § 10.3.1.1 in
Chap. 10). It also affects, as shown in Study 12.13, the nature of the aggregates, filaments,
or sheets, formed by APol-depleted OmpF/A8-35 complexes in the presence of lipopolysaccharide (Arunmanee et al. 2014). Adding EDTA to solutions of MP/A8-35, MP/A8-75, or
MP/SMA complexes, all of which are sensitive to the presence of Ca
2+ (Champeil et al. 2000;
Postis et al. 2015), is certainly a good precaution if aggregation is to be avoided.
The formation of superstructures could perhaps be harnessed for the reconstruction of 3D models
by helical reconstruction of filaments or by inducing the formation of 2D crystals. Analysis of the
internal periodicity of the filaments formed by BR (Gohon et al. 2008) or OmpF (Arunmanee et al.
2014) ought to indicate whether protein/protein contacts are involved, which could generate some
order, or only polymer-polymer contacts, as seems to be the case in AcrB/SMA “doublets” (Postis
et al. 2015; see Fig. 12.17), in which case disorder is probable. This point will be further discussed in
§ 12.3.4.3.
• There may be a tendency for the complexes to adhere to the carbon film via the TM region of
the protein (Arunmanee et al. 2014; Postis et al. 2015), which may be a nuisance if a random
distribution of the particles is desired. Ways could be devised to control this phenomenon, as
found in a cryo-EM study of the ryanodine receptor (see below, Study 12.19 in § 12.3.3).
Whereas negatively stained samples are easy to prepare and comparatively easy to study, only
unstained samples trapped in vitreous ice have the potential to yield density maps that, thanks to recent
progress in the collection, correction, and analysis of cryo-EM images, can be interpreted in terms of
atomic structures (see § 12.2.3). As described in the next section, APols are bringing their modest but
useful contribution to such studies by facilitating the preparation and observation of EM samples.
12.3 The Use of Amphipols for Membrane Protein Electron Microscopy Studies
561
(Study 12.7; cf. Chap. 7, Fig. 7.7A, B), MP/non-ionic APol (Study 12.8), and MP/A8-35 (Studies
12.25 and 12.28) complexes. A recent atomic-force microscopy study of MotPS/A8-35 complexes
should also be mentioned (Terahara et al. 2017).
12.3.2.1 Conclusions from Negative-Stain Studies
From this survey of the most informative studies in which TEM was used to examine negatively
stained MP/APol complexes, one can tentatively draw the following conclusions:
• MP/APol (and MP/SMA) complexes are amenable to such studies, and 3D models can be
routinely reconstructed to ~16 to 18-Å resolution (Cvetkov et al. 2011; Kevany et al. 2013;
Tsybovsky et al. 2013; Huynh et al. 2014b) or close to it (Sverzhinsky et al. 2014; Postis et al.
2015), that is to the best resolution that can be achieved using NS.
• Attention should be paid to the fact that removal of free APol may induce the formation of
aggregates, which, in the two studies where they have been examined by TEM/NS (Gohon
et al. 2008; Arunmanee et al. 2014), take the form of filaments, occasionally of sheets
(Arunmanee et al. 2014). Whether these structures form, and at which rate, is likely to be
strongly dependent on the MP, on the nature of the APol, and on conditions such as the
residual concentration of free APol and/or detergent, the ionic strength, and the presence of
divalent cations. Note that the presence of traces of divalent cations degrades the quality of
NMR spectra of MP/A8-35 complexes (Catoire et al. 2009, 2010), certainly due to calciummediated association of some of the complexes via their APol belts (see § 10.3.1.1 in
Chap. 10). It also affects, as shown in Study 12.13, the nature of the aggregates, filaments,
or sheets, formed by APol-depleted OmpF/A8-35 complexes in the presence of lipopolysaccharide (Arunmanee et al. 2014). Adding EDTA to solutions of MP/A8-35, MP/A8-75, or
MP/SMA complexes, all of which are sensitive to the presence of Ca
2+ (Champeil et al. 2000;
Postis et al. 2015), is certainly a good precaution if aggregation is to be avoided.
The formation of superstructures could perhaps be harnessed for the reconstruction of 3D models
by helical reconstruction of filaments or by inducing the formation of 2D crystals. Analysis of the
internal periodicity of the filaments formed by BR (Gohon et al. 2008) or OmpF (Arunmanee et al.
2014) ought to indicate whether protein/protein contacts are involved, which could generate some
order, or only polymer-polymer contacts, as seems to be the case in AcrB/SMA “doublets” (Postis
et al. 2015; see Fig. 12.17), in which case disorder is probable. This point will be further discussed in
§ 12.3.4.3.
• There may be a tendency for the complexes to adhere to the carbon film via the TM region of
the protein (Arunmanee et al. 2014; Postis et al. 2015), which may be a nuisance if a random
distribution of the particles is desired. Ways could be devised to control this phenomenon, as
found in a cryo-EM study of the ryanodine receptor (see below, Study 12.19 in § 12.3.3).
Whereas negatively stained samples are easy to prepare and comparatively easy to study, only
unstained samples trapped in vitreous ice have the potential to yield density maps that, thanks to recent
progress in the collection, correction, and analysis of cryo-EM images, can be interpreted in terms of
atomic structures (see § 12.2.3). As described in the next section, APols are bringing their modest but
useful contribution to such studies by facilitating the preparation and observation of EM samples.
12.3 The Use of Amphipols for Membrane Protein Electron Microscopy Studies
561