Henderson in a landmark review, the latter procedure cannot be applied to biological molecules smaller
than 50–100 kDa (Henderson 1995, 2015; Henderson and McMullan 2013). Alternatively, the contrast
can be increased, and the biological sample made electron-resistant, by embedding it into an electrondense metallic stain, typically a uranium or tungsten salt. This approach can be applied to both
crystalline and single-particle samples. It is most often the first step when embarking on an EM
study and developing protocols for the preparation of samples, because it is quick and easy and yields
within minutes a first view of their nature and quality. What is imaged, in that case, is not the biological
sample itself, but the crust of metal salt that has deposited around it. Due to the grain of the salt, this
approach, known as “negative staining” (NS) because molecules appear as white objects that have
excluded the salt, is limited to resolutions of, at best, ~15 to 18 Å (Ohi et al. 2004; for a recent overview
of “direct imaging methods” relying on the use of stains or metal shadowing and the way they relate to
X-ray and cryo-EM methods, see Miyaguchi 2014).
Except under very special circumstances, which have not been exploited yet and will be briefly
discussed below (§ 12.3.4.3), there is little probability that APols can contribute much to EM studies of
2D crystals. This is because the order in these crystals comes from lateral contacts that, as a rule,
involve mostly the transmembrane (TM) regions of MPs. As discussed in Chap. 11 about Type 1 3D
crystals (see § 11.2.2), APol-coated TM surfaces are unlikely to allow the formation of regular contacts
leading to the formation of well-ordered crystals, unless APols move out of the way and allow direct
protein/protein contacts (§ 12.3.4.3). We will therefore focus on the practice and constraints of singleparticle EM studies.
12.2.2 Single-Particle Electron Microscopy Studies of Biological Molecules:
The Basics
Single-particle EM studies of biological molecules (for general reviews, see e.g. Frank 2006;
Rubinstein 2007; Cheng et al. 2015) require a more moderate purification of the sample than other
structural techniques such as, say, radiation scattering, because averaging of the data over a population
of particles will be done only after the particles to be studied have been selected, based on the
appearance of their projections. Thus, it is perfectly possible, provided that the different forms can
be efficiently distinguished one from another, to work on a mixture of monomeric and dimeric
complexes, whose 3D structure will be reconstructed separately (see e.g. Postis et al. 2015; Study
12.22 below), or on preparations containing a given molecule in several different conformations (see
e.g. Cheng 2015). The particles are spread on an EM grid covered with a carbon film which can be
either continuous or, as is most often the case in cryo-EM, holey, and has been rendered hydrophilic,
usually with a glow discharger. In the case of a continuous carbon film, the particles adhere to it. For
NS, the grid, after rinsing, is exposed to a solution of heavy metal salt, blotted, and air-dried before
being introduced in the electron microscope (for a recent review, see Vahedi-Faridi et al. 2013).
In cryo-EM (for recent reviews, see e.g. Frank 2006; Orlova and Saibil 2011; Lau and Rubinstein
2013; Milne et al. 2013; Bai et al. 2015a; Cheng 2015; Cheng et al. 2015; Nogales and Scheres 2015;
De Zorzi et al. 2016; Vinothkumar and Henderson 2016), the particles are most often suspended in the
aqueous films that form over holes in a carbon film when the grid is blotted (Fig. 12.1A). The grids are
then rapidly plunged (within seconds) into a freezing medium, generally a slurry of liquid ethane
cooled to liquid nitrogen temperature, in which liquid water turns to vitreous ice (Dubochet et al. 1988;
Dubochet 2015). They are dipped in liquid nitrogen and introduced in the microscope in a grid holder
that is kept at either liquid nitrogen or liquid helium temperature. Important features of vitrified
specimens include (i) an amorphous ice layer sufficiently thick to accommodate the particles, but
12.2 Context: Electron Microscopy Studies of Membrane Proteins
535
than 50–100 kDa (Henderson 1995, 2015; Henderson and McMullan 2013). Alternatively, the contrast
can be increased, and the biological sample made electron-resistant, by embedding it into an electrondense metallic stain, typically a uranium or tungsten salt. This approach can be applied to both
crystalline and single-particle samples. It is most often the first step when embarking on an EM
study and developing protocols for the preparation of samples, because it is quick and easy and yields
within minutes a first view of their nature and quality. What is imaged, in that case, is not the biological
sample itself, but the crust of metal salt that has deposited around it. Due to the grain of the salt, this
approach, known as “negative staining” (NS) because molecules appear as white objects that have
excluded the salt, is limited to resolutions of, at best, ~15 to 18 Å (Ohi et al. 2004; for a recent overview
of “direct imaging methods” relying on the use of stains or metal shadowing and the way they relate to
X-ray and cryo-EM methods, see Miyaguchi 2014).
Except under very special circumstances, which have not been exploited yet and will be briefly
discussed below (§ 12.3.4.3), there is little probability that APols can contribute much to EM studies of
2D crystals. This is because the order in these crystals comes from lateral contacts that, as a rule,
involve mostly the transmembrane (TM) regions of MPs. As discussed in Chap. 11 about Type 1 3D
crystals (see § 11.2.2), APol-coated TM surfaces are unlikely to allow the formation of regular contacts
leading to the formation of well-ordered crystals, unless APols move out of the way and allow direct
protein/protein contacts (§ 12.3.4.3). We will therefore focus on the practice and constraints of singleparticle EM studies.
12.2.2 Single-Particle Electron Microscopy Studies of Biological Molecules:
The Basics
Single-particle EM studies of biological molecules (for general reviews, see e.g. Frank 2006;
Rubinstein 2007; Cheng et al. 2015) require a more moderate purification of the sample than other
structural techniques such as, say, radiation scattering, because averaging of the data over a population
of particles will be done only after the particles to be studied have been selected, based on the
appearance of their projections. Thus, it is perfectly possible, provided that the different forms can
be efficiently distinguished one from another, to work on a mixture of monomeric and dimeric
complexes, whose 3D structure will be reconstructed separately (see e.g. Postis et al. 2015; Study
12.22 below), or on preparations containing a given molecule in several different conformations (see
e.g. Cheng 2015). The particles are spread on an EM grid covered with a carbon film which can be
either continuous or, as is most often the case in cryo-EM, holey, and has been rendered hydrophilic,
usually with a glow discharger. In the case of a continuous carbon film, the particles adhere to it. For
NS, the grid, after rinsing, is exposed to a solution of heavy metal salt, blotted, and air-dried before
being introduced in the electron microscope (for a recent review, see Vahedi-Faridi et al. 2013).
In cryo-EM (for recent reviews, see e.g. Frank 2006; Orlova and Saibil 2011; Lau and Rubinstein
2013; Milne et al. 2013; Bai et al. 2015a; Cheng 2015; Cheng et al. 2015; Nogales and Scheres 2015;
De Zorzi et al. 2016; Vinothkumar and Henderson 2016), the particles are most often suspended in the
aqueous films that form over holes in a carbon film when the grid is blotted (Fig. 12.1A). The grids are
then rapidly plunged (within seconds) into a freezing medium, generally a slurry of liquid ethane
cooled to liquid nitrogen temperature, in which liquid water turns to vitreous ice (Dubochet et al. 1988;
Dubochet 2015). They are dipped in liquid nitrogen and introduced in the microscope in a grid holder
that is kept at either liquid nitrogen or liquid helium temperature. Important features of vitrified
specimens include (i) an amorphous ice layer sufficiently thick to accommodate the particles, but
12.2 Context: Electron Microscopy Studies of Membrane Proteins
535