positions with respect to the protein, but they would nevertheless indicate where the APol belt
lies. This approach, which has been validated for the couple biotinylated A8-35/monomeric
streptavidin, using the mitochondrial cytochrome bc 1 complex and a bacterial secretion
protein, CsgG, as test MPs, will be described in an upcoming article (Perry et al. 2018).
The thiomorpholine-tagged version of A8-35 (see Chap. 4) could be used in the same manner,
in combination with gold nanoparticles (cf. Suárez-Suárez et al. 2013).
In cryo-EM, one could conceivably play on the nature of the counterions. A belt of A8-35
can be expected to carry about 0.88 carboxylates per nm
2 of APol/water interface, or, in
projection, for a belt ~4 nm high and 2 nm thick, ~4.4 carboxylates per nm
2 . AUC
experiments have shown that MP/A8-35 complexes can be handled in the presence of any
alkaline cation from Li
+ to Cs
+ (Gohon 2002). Rough calculations indicate that, were two
otherwise identical samples of MP/A8-35 complexes to be imaged in the presence of
50–100 mM of either LiCl or CsCl, the increase in contrast of the APol belt over the
surrounding solution would be of ~2.3Â when substituting Cs
+ to Na
+
, ~3Â when comparing
Li
+ and Cs
+
, which ought to be easily detected.
• Inducing the complexes to adopt preferred orientations. Tagged APols could possibly be used
to bias the orientation of MP/APol complexes. For instance, a carbon film onto which would
have been adsorbed a layer of Ni/NTA-carrying lipids (Kelly et al. 2008) could possibly be
used to induce MPs trapped with a poly-histidine-tagged APol to lie with their TM region in
contact with the carbon film. We have speculated elsewhere (Le Bon et al. 2014a) on the
possibility to use DNA scaffolds to organize in space MPs trapped with an oligonucleotidetagged APol (cf. Selmi et al. 2011).
• In the same order of ideas, it would be interesting to push somewhat further the study of the
filaments and, if applicable, the 2D sheets that some MP/APol complexes tend to form upon
depletion of free APol (see above, §§ 12.3.1 and 12.3.2). It would be important, in particular, to
establish whether the formation of filaments is mediated by APol/APol or by protein/protein
contacts, which could be done by studying the periodicity of the protein along the filaments. In
the first case, not much order can be expected from the proteins. In the second case, on the
contrary, a regular helix could form, which could be amenable to helical reconstruction. This
would be especially valuable for MPs whose small size renders them difficult to orient for
single-particle image reconstruction and that refuse to yield 2D or 3D crystals.
It is not clear why the side-by-side association into filaments does not proceed further to
form 2D sheets, as one might have expected. One possibility is that electrostatic repulsion
between APol belts opposes it, favoring the addition of new molecules at the extremities
rather than at the side of filaments. It would be of great interest to examine the effect of factors
such as the charge density of the APol (SAPols ¼ A8-75 > A8-35 > NAPols; see Chap. 4), the
ionic strength, or the addition of small amounts of neutral detergent. In the case of E. coli
OmpF, one factor has already been identified: addition of LPS in the presence of divalent
cations favors the transition from a 1D organization, filaments, to a 2D one, sheets or crystals
(Arunmanee et al. 2014; see above, § 12.3.2). As suggested in the latter work, it would be
intriguing to carry such studies further and examine whether APols could not be used as
modulators in controlling the formation of MP 2D crystals.
• Finally, another intriguing possibility, which might perhaps deserve examination, would be to
trap MPs with an APol carrying free-radical scavenger groups, whose synthesis has been
developed (F. Giusti and M. Zoonens, unpublished data), in the hope of slowing down
radiation damage during illumination.
578
12 The Use of Amphipols for Electron Microscopy
lies. This approach, which has been validated for the couple biotinylated A8-35/monomeric
streptavidin, using the mitochondrial cytochrome bc 1 complex and a bacterial secretion
protein, CsgG, as test MPs, will be described in an upcoming article (Perry et al. 2018).
The thiomorpholine-tagged version of A8-35 (see Chap. 4) could be used in the same manner,
in combination with gold nanoparticles (cf. Suárez-Suárez et al. 2013).
In cryo-EM, one could conceivably play on the nature of the counterions. A belt of A8-35
can be expected to carry about 0.88 carboxylates per nm
2 of APol/water interface, or, in
projection, for a belt ~4 nm high and 2 nm thick, ~4.4 carboxylates per nm
2 . AUC
experiments have shown that MP/A8-35 complexes can be handled in the presence of any
alkaline cation from Li
+ to Cs
+ (Gohon 2002). Rough calculations indicate that, were two
otherwise identical samples of MP/A8-35 complexes to be imaged in the presence of
50–100 mM of either LiCl or CsCl, the increase in contrast of the APol belt over the
surrounding solution would be of ~2.3Â when substituting Cs
+ to Na
+
, ~3Â when comparing
Li
+ and Cs
+
, which ought to be easily detected.
• Inducing the complexes to adopt preferred orientations. Tagged APols could possibly be used
to bias the orientation of MP/APol complexes. For instance, a carbon film onto which would
have been adsorbed a layer of Ni/NTA-carrying lipids (Kelly et al. 2008) could possibly be
used to induce MPs trapped with a poly-histidine-tagged APol to lie with their TM region in
contact with the carbon film. We have speculated elsewhere (Le Bon et al. 2014a) on the
possibility to use DNA scaffolds to organize in space MPs trapped with an oligonucleotidetagged APol (cf. Selmi et al. 2011).
• In the same order of ideas, it would be interesting to push somewhat further the study of the
filaments and, if applicable, the 2D sheets that some MP/APol complexes tend to form upon
depletion of free APol (see above, §§ 12.3.1 and 12.3.2). It would be important, in particular, to
establish whether the formation of filaments is mediated by APol/APol or by protein/protein
contacts, which could be done by studying the periodicity of the protein along the filaments. In
the first case, not much order can be expected from the proteins. In the second case, on the
contrary, a regular helix could form, which could be amenable to helical reconstruction. This
would be especially valuable for MPs whose small size renders them difficult to orient for
single-particle image reconstruction and that refuse to yield 2D or 3D crystals.
It is not clear why the side-by-side association into filaments does not proceed further to
form 2D sheets, as one might have expected. One possibility is that electrostatic repulsion
between APol belts opposes it, favoring the addition of new molecules at the extremities
rather than at the side of filaments. It would be of great interest to examine the effect of factors
such as the charge density of the APol (SAPols ¼ A8-75 > A8-35 > NAPols; see Chap. 4), the
ionic strength, or the addition of small amounts of neutral detergent. In the case of E. coli
OmpF, one factor has already been identified: addition of LPS in the presence of divalent
cations favors the transition from a 1D organization, filaments, to a 2D one, sheets or crystals
(Arunmanee et al. 2014; see above, § 12.3.2). As suggested in the latter work, it would be
intriguing to carry such studies further and examine whether APols could not be used as
modulators in controlling the formation of MP 2D crystals.
• Finally, another intriguing possibility, which might perhaps deserve examination, would be to
trap MPs with an APol carrying free-radical scavenger groups, whose synthesis has been
developed (F. Giusti and M. Zoonens, unpublished data), in the hope of slowing down
radiation damage during illumination.
578
12 The Use of Amphipols for Electron Microscopy