of 2014) have been folded or refolded in vitro using detergents and even more exotic media, which
reproduce essentially none of the constraints exerted by biological membranes (reviewed in Popot
2014; cf. Chap. 6, § 6.2). The apparent paradox vanishes if one considers that folding or refolding is
almost never carried out in pure detergent solutions but, typically, in lipid/detergent mixed micelles, in
lipid vesicles, or in APols, the latter supplemented or not with lipids (see Table 1 in Popot 2014). This
strongly suggests that it is the presence of detergent, more than the absence of a membrane, that
compromises the achievement or the stability of the native fold. On the basis of these observations, it
has been proposed that, for many if not most MPs, the determinants of folding come essentially from
the amino acid sequence, not from physical constraints exerted by the environment. The latter can
obviously modulate the structure and thereby the function of certain MPs (e.g. mechanosensitive
channels), but in many, perhaps most, cases, it plays no critical role in determining the general 3D fold
(discussed in Popot and Engelman 2016). Structural details, however, can be modulated by the
environment and, in particular, by the binding of this or that lipid, which often plays important
functional and/or regulatory roles.
2.5
Solutions to the Instability Problem
In § 2.4, we have described some of the problems associated with handling MPs in detergent solutions,
discussed hypotheses about their origin, and presented some approaches to alleviating their impact, in
particular limiting the volume of the hydrophobic sink and using lipid/detergent mixed micelles rather
than pure detergent. There are, however, other ways, not necessarily exclusive one from the other, to
stabilize MPs in aqueous environments. We can distribute them into three categories:
• A first approach is to optimize the structure of the detergent so as to preserve its ability to
dissolve biological membranes and solubilize MPs while slowing down the rate at which it
inactivates them (§ 2.5.1).
• A second approach is to either choose a more resistant homologue of the protein or to
engineer it so as to adapt it to the detergent (§ 2.5.2).
• A third, somewhat radical approach is to handle the protein in the absence of detergent
(§ 2.5.3).
2.5.1
Making Detergents Less Aggressive
Until the 1970s, biologists had to make do with detergents provided by nature (bile salts, digitonin) or
synthetic detergents developed for industrial purposes (Triton, Tween, Lubrol, Brij, etc.). Even though
some of them are still in use today in membrane biology laboratories (digitonin, because of its
mildness; Triton X-100, because of its efficiency as a solubilizing agent; Tween as an adsorbent
onto hydrophobic surfaces; etc.), each of these compounds presented its set of limitations, such as
heterogeneity, the presence of impurities, the formation of large micelles, UV absorbance, harshness
or, on the contrary, limited solubilizing power, etc. This led to the development of synthetic or
semisynthetic detergents specifically designed for extracting and handling MPs. Among the most
successful of the novel detergents that appeared in the mid-1970s to early 1980s (some of which had
been synthesized years if not decades before but had not yet been discovered by membrane
biochemists), one may cite octyl-β-D-glucoside (OG) (Stubbs et al. 1976; Keana and Roman 1978),
2.5 Solutions to the Instability Problem
81
reproduce essentially none of the constraints exerted by biological membranes (reviewed in Popot
2014; cf. Chap. 6, § 6.2). The apparent paradox vanishes if one considers that folding or refolding is
almost never carried out in pure detergent solutions but, typically, in lipid/detergent mixed micelles, in
lipid vesicles, or in APols, the latter supplemented or not with lipids (see Table 1 in Popot 2014). This
strongly suggests that it is the presence of detergent, more than the absence of a membrane, that
compromises the achievement or the stability of the native fold. On the basis of these observations, it
has been proposed that, for many if not most MPs, the determinants of folding come essentially from
the amino acid sequence, not from physical constraints exerted by the environment. The latter can
obviously modulate the structure and thereby the function of certain MPs (e.g. mechanosensitive
channels), but in many, perhaps most, cases, it plays no critical role in determining the general 3D fold
(discussed in Popot and Engelman 2016). Structural details, however, can be modulated by the
environment and, in particular, by the binding of this or that lipid, which often plays important
functional and/or regulatory roles.
2.5
Solutions to the Instability Problem
In § 2.4, we have described some of the problems associated with handling MPs in detergent solutions,
discussed hypotheses about their origin, and presented some approaches to alleviating their impact, in
particular limiting the volume of the hydrophobic sink and using lipid/detergent mixed micelles rather
than pure detergent. There are, however, other ways, not necessarily exclusive one from the other, to
stabilize MPs in aqueous environments. We can distribute them into three categories:
• A first approach is to optimize the structure of the detergent so as to preserve its ability to
dissolve biological membranes and solubilize MPs while slowing down the rate at which it
inactivates them (§ 2.5.1).
• A second approach is to either choose a more resistant homologue of the protein or to
engineer it so as to adapt it to the detergent (§ 2.5.2).
• A third, somewhat radical approach is to handle the protein in the absence of detergent
(§ 2.5.3).
2.5.1
Making Detergents Less Aggressive
Until the 1970s, biologists had to make do with detergents provided by nature (bile salts, digitonin) or
synthetic detergents developed for industrial purposes (Triton, Tween, Lubrol, Brij, etc.). Even though
some of them are still in use today in membrane biology laboratories (digitonin, because of its
mildness; Triton X-100, because of its efficiency as a solubilizing agent; Tween as an adsorbent
onto hydrophobic surfaces; etc.), each of these compounds presented its set of limitations, such as
heterogeneity, the presence of impurities, the formation of large micelles, UV absorbance, harshness
or, on the contrary, limited solubilizing power, etc. This led to the development of synthetic or
semisynthetic detergents specifically designed for extracting and handling MPs. Among the most
successful of the novel detergents that appeared in the mid-1970s to early 1980s (some of which had
been synthesized years if not decades before but had not yet been discovered by membrane
biochemists), one may cite octyl-β-D-glucoside (OG) (Stubbs et al. 1976; Keana and Roman 1978),
2.5 Solutions to the Instability Problem
81