established, for the first time, that APol-trapped MPs could be studied by both NS and cryo-EM
approaches. These observations, however, seem to have remained largely unnoticed outside of the
ATP-synthase community.
A cryo-EM study with reconstruction of a 3D model of a MP complex was first presented in
Study 12.3, taking the mitochondrial Complex I as a test MP (Flötenmeyer et al. 2007). Complex I
(NADH:ubiquinone oxidoreductase) was purified from the fungus Neurospora crassa in the presence
of the detergent Triton X-100 and transferred to A8-35 using Bio-Beads SM-2 and ultrafiltration. It is
likely that the latter procedure removed most if not all of the free APol, a significant point in view of
some aspects of the cryo-EM observations (see below). A comparison of cryo-EM grids prepared
either with the Triton X-100-solubilized or with the A8-35-trapped preparation is shown in Fig. 12.9.
Panels a, c, and e show samples in Triton X-100, which were blotted on perforated carbon films and
fast-frozen in liquid ethane. The resulting ice films show a pronounced thickness gradient across the
holes (Fig. 12.9E), and the protein tends to clump and aggregate at the edges (Fig. 12.9C). With
samples of A8-35-trapped Complex I (Panels b, d, f), an ice film much more uniform in thickness is
obtained (Panel f), in which individual Complex I particles are better dispersed and can be more easily
identified (Panel d) and picked for image processing. This phenomenon considerably facilitated the
collection of a sufficient number of images of isolated particles. Its cause and possible ways to control
it will be discussed below (§ 12.3.4.2).
Amphipols and electron cryomicroscopy of membrane proteins
(# 2018 Francis Haraux)
A total of 1200 particles were selected, normalized, and subjected to multivariate analysis (van
Heel and Frank 1981). The top 16 class averages are shown in Fig. 12.10A. The class average
projections were then cross-correlated with a set of angular projections of a 3D model taken from
the reconstruction of negatively stained Complex I (Guénebaut et al. 1998). The matching projections
of the model are shown in Fig. 12.10B. There is good agreement between the class average projections
of the cryo-EM images and the corresponding projections of the negative-stain model, making it
possible to use the latter as a reference for the angular reconstitution method (Serysheva et al. 1995).
550
12 The Use of Amphipols for Electron Microscopy
approaches. These observations, however, seem to have remained largely unnoticed outside of the
ATP-synthase community.
A cryo-EM study with reconstruction of a 3D model of a MP complex was first presented in
Study 12.3, taking the mitochondrial Complex I as a test MP (Flötenmeyer et al. 2007). Complex I
(NADH:ubiquinone oxidoreductase) was purified from the fungus Neurospora crassa in the presence
of the detergent Triton X-100 and transferred to A8-35 using Bio-Beads SM-2 and ultrafiltration. It is
likely that the latter procedure removed most if not all of the free APol, a significant point in view of
some aspects of the cryo-EM observations (see below). A comparison of cryo-EM grids prepared
either with the Triton X-100-solubilized or with the A8-35-trapped preparation is shown in Fig. 12.9.
Panels a, c, and e show samples in Triton X-100, which were blotted on perforated carbon films and
fast-frozen in liquid ethane. The resulting ice films show a pronounced thickness gradient across the
holes (Fig. 12.9E), and the protein tends to clump and aggregate at the edges (Fig. 12.9C). With
samples of A8-35-trapped Complex I (Panels b, d, f), an ice film much more uniform in thickness is
obtained (Panel f), in which individual Complex I particles are better dispersed and can be more easily
identified (Panel d) and picked for image processing. This phenomenon considerably facilitated the
collection of a sufficient number of images of isolated particles. Its cause and possible ways to control
it will be discussed below (§ 12.3.4.2).
Amphipols and electron cryomicroscopy of membrane proteins
(# 2018 Francis Haraux)
A total of 1200 particles were selected, normalized, and subjected to multivariate analysis (van
Heel and Frank 1981). The top 16 class averages are shown in Fig. 12.10A. The class average
projections were then cross-correlated with a set of angular projections of a 3D model taken from
the reconstruction of negatively stained Complex I (Guénebaut et al. 1998). The matching projections
of the model are shown in Fig. 12.10B. There is good agreement between the class average projections
of the cryo-EM images and the corresponding projections of the negative-stain model, making it
possible to use the latter as a reference for the angular reconstitution method (Serysheva et al. 1995).
550
12 The Use of Amphipols for Electron Microscopy