generally not be positioned identically with respect to one another from one complex to the next,
creating heterogeneity. Only poorly diffracting crystals of BR have been obtained to date (unpublished
observations cited in Denisov and Sligar (2017)). When the MP sports huge extramembrane domains,
crystallization may perhaps be achieved under conditions where all contacts form between those and
the rotational disorder of the MSPs does not prevent the growth of well-ordered crystals. Another
opportunity that does not seem to have been discussed yet would be to turn to advantage the
inconvenience of trapping a MP in too narrow a ND: if protein/protein contacts between the guest
MP and MSPs cannot be avoided, then perhaps the proteins may lock in the register permitting the
most favorable contacts, generating homogeneous complexes. This might be an incentive to conduct
crystallization trials using the smallest NDs that will accommodate the target MP. Appropriate
engineering, like that of hydrogen bonds or salt bridges, could possibly help stabilizing a unique
arrangement. Finally, a promising approach is to transfer ND-trapped MPs to lipidic mesophases.
When BR/APol complexes are mixed with a mesophase, the protein and the APol part ways, diffuse
separately, and highly ordered crystals of APol-free BR form (Polovinkin et al. 2014), providing a way
to crystallize MPs that have been folded or stabilized using APols (cf. Chap. 11, § 11.2.2.2). The same
process has recently been extended to ND-trapped ones (Nikolaev et al. 2017).
It is possible to fold denatured MPs to their native state while incorporating them into NDs. BR
from Halobacterium salinarum has been refolded starting from a mixture of MSP1D1, DMPC,
bacterio-opsin, and retinal dissolved in SDS and removing SDS with Bio-Beads (see the Methods
section in Etzkorn et al. 2013). A similar procedure has been applied to BR from Exiguobacterium
sibiricum and to the homotetrameric K
+ channel KcsA from Streptomyces lividans (Shenkarev et al.
2013). Extending this approach to other MPs, particularly complex ones, might not always be
straightforward, given the many parasitic reactions that can take place and lead to aggregation and
the large number of parameters to be optimized, but the procedure can obviously be made to work and
is worth keeping in mind.
3.4
Amphipathic Peptides
Amphipathic peptides that act to destabilize membranes, such as antimicrobial peptides, bee venom
melittin, or numerous bacterial toxins, are widespread in nature (for a review, see e.g. Peters et al.
2010). Their use as detergent substitutes to stabilize MPs in aqueous solutions was pioneered in the
early 1990s (Dempsey and Sternberg 1991; Schafmeister et al. 1993) and has since branched in many
different directions. Even though none of the approaches that have been investigated has yet become
widely used, a rapid survey is in order. The structure of the peptides and the organization of the
complexes they form with MPs and/or lipids are extremely variable (Table 3.2). One can categorize
them as follows:
• Relatively long peptides (typically 20 to 25 residues long) that are designed to fold into
laterally amphipathic α-helices whose length is comparable to the thickness of the hydrophobic core of a lipid bilayer and whose hydrophobic face is expected to adsorb onto the
hydrophobic transmembrane (TM) surface of MPs. They can be plain oligopeptides
(“peptitergents”; Schafmeister et al. 1993; Table 3.2, line 1; § 3.4.1) or endowed with
bound fatty acyl chains meant to provide a “soft” interface between the peptide and the MP
(“lipopeptide detergents”; McGregor et al. 2003; Table 3.2, line 2; § 3.4.2).
118
3 Alternatives to Detergents for Handling Membrane Proteins in Aqueous Solutions
creating heterogeneity. Only poorly diffracting crystals of BR have been obtained to date (unpublished
observations cited in Denisov and Sligar (2017)). When the MP sports huge extramembrane domains,
crystallization may perhaps be achieved under conditions where all contacts form between those and
the rotational disorder of the MSPs does not prevent the growth of well-ordered crystals. Another
opportunity that does not seem to have been discussed yet would be to turn to advantage the
inconvenience of trapping a MP in too narrow a ND: if protein/protein contacts between the guest
MP and MSPs cannot be avoided, then perhaps the proteins may lock in the register permitting the
most favorable contacts, generating homogeneous complexes. This might be an incentive to conduct
crystallization trials using the smallest NDs that will accommodate the target MP. Appropriate
engineering, like that of hydrogen bonds or salt bridges, could possibly help stabilizing a unique
arrangement. Finally, a promising approach is to transfer ND-trapped MPs to lipidic mesophases.
When BR/APol complexes are mixed with a mesophase, the protein and the APol part ways, diffuse
separately, and highly ordered crystals of APol-free BR form (Polovinkin et al. 2014), providing a way
to crystallize MPs that have been folded or stabilized using APols (cf. Chap. 11, § 11.2.2.2). The same
process has recently been extended to ND-trapped ones (Nikolaev et al. 2017).
It is possible to fold denatured MPs to their native state while incorporating them into NDs. BR
from Halobacterium salinarum has been refolded starting from a mixture of MSP1D1, DMPC,
bacterio-opsin, and retinal dissolved in SDS and removing SDS with Bio-Beads (see the Methods
section in Etzkorn et al. 2013). A similar procedure has been applied to BR from Exiguobacterium
sibiricum and to the homotetrameric K
+ channel KcsA from Streptomyces lividans (Shenkarev et al.
2013). Extending this approach to other MPs, particularly complex ones, might not always be
straightforward, given the many parasitic reactions that can take place and lead to aggregation and
the large number of parameters to be optimized, but the procedure can obviously be made to work and
is worth keeping in mind.
3.4
Amphipathic Peptides
Amphipathic peptides that act to destabilize membranes, such as antimicrobial peptides, bee venom
melittin, or numerous bacterial toxins, are widespread in nature (for a review, see e.g. Peters et al.
2010). Their use as detergent substitutes to stabilize MPs in aqueous solutions was pioneered in the
early 1990s (Dempsey and Sternberg 1991; Schafmeister et al. 1993) and has since branched in many
different directions. Even though none of the approaches that have been investigated has yet become
widely used, a rapid survey is in order. The structure of the peptides and the organization of the
complexes they form with MPs and/or lipids are extremely variable (Table 3.2). One can categorize
them as follows:
• Relatively long peptides (typically 20 to 25 residues long) that are designed to fold into
laterally amphipathic α-helices whose length is comparable to the thickness of the hydrophobic core of a lipid bilayer and whose hydrophobic face is expected to adsorb onto the
hydrophobic transmembrane (TM) surface of MPs. They can be plain oligopeptides
(“peptitergents”; Schafmeister et al. 1993; Table 3.2, line 1; § 3.4.1) or endowed with
bound fatty acyl chains meant to provide a “soft” interface between the peptide and the MP
(“lipopeptide detergents”; McGregor et al. 2003; Table 3.2, line 2; § 3.4.2).
118
3 Alternatives to Detergents for Handling Membrane Proteins in Aqueous Solutions