totally uncharged NAPols also allow folding in good yield (Bazzacco et al. 2012) seems to indicate that
such an electrostatic mechanism, if present, is not essential.
A related approach, cell-free synthesis of MPs using APols as the accepting medium, is
described in Chap. 7.
A thought-provoking point is why APols seem to provide such a favorable folding medium for
MPs, when neither their chemical structure nor the supramolecular structures they assemble into in
aqueous solutions bear any similarity to lipid bilayers. This puzzling question is briefly discussed in
Box 6.3.
Box 6.3 Why Do Membrane Proteins Adopt Their Native Fold in the Absence of
a Membrane?
To the cell biologist, it may seem strange that such good MP folding yields can be obtained in
APols. MPs are obviously adapted to a membrane environment, viz. the way they expose them to
highly hydrophobic surfaces whose depth matches that of the membrane’s hydrophobic core.
(Re)folding experiments show that an unfolded MP can fold back to its native state if transferred
either to lipid vesicles or to a “membrane-mimetic” environment (§ 6.2 and 6.3). Yet, can APols
really be considered as “membrane-mimetic”, when their chemical structure and the supramolecular
aggregates they form in water (Chap. 4) are so different, chemically and physically, from a lipid
bilayer? Does the high efficiency of MP folding in APols (Table 6.1) tell us something about the
constraints MPs require to fold to their native state?
This question has been debated in a discussion paper to which the reader is referred (Popot
and Engelman 2016). In brief, it is proposed that, because of the conditions in which MPs are
inserted into membranes in vivo, relying, for folding, on the physical constraints provided by a lipid
bilayer would be counterproductive and lead to low folding yields. This is because a folding MP
in vivo is exposed to a highly complex environment that has little to do with a lipid bilayer, crowded
as it is with proteins. Furthermore, physical constraints and chemical interactions differ from one
membrane to the next, so that a given MP will usually be exposed during its synthesis and insertion
to constraints different from those that prevail in the membrane compartment in which it will fulfill
its function. These considerations lead to the proposal that MPs have evolved to be adapted to a
membrane environment, but rely primarily, for their folding, on (i) their TM region being protected
from the aqueous phase, without specific requirements about how this protection is afforded, and
(ii) finding within their own sequence the information required to fold into a functional 3D
structure, without guidance by physical constraints from their environment, such as bilayer
thickness, lateral pressure gradient, etc. This is not to say, of course, that lipids have no influence
on MP structure and function; much to the contrary (cf. Chap. 1). In some cases, the lipid
composition can even modulate the TM topology adopted by a given MP (see e.g. Bogdanov
et al. 2014; Vitrac et al. 2017; and references therein). For a full discussion of this intriguing
question, the reader is referred to Popot and Engelman (2016).
6.3.3
Challenges and Prospects
One of the many open questions is which types of MPs will turn out to be amenable to APol-assisted
folding and which not. MPs with extended extramembrane domains have not been tested yet (OmpA
does feature such a domain, but without a function that could be easily tested to assess its folding). One
may note, however, in this respect, that APols have been observed to protect denatured soluble proteins
from aggregation (Ma et al. 2012; Martin et al. 2014, 2015), which should constitute a favorable factor.
A particularly challenging case is that of oligomeric MPs. There are, however, reasons to expect
that, there as well, the use of APols might provide interesting perspectives:
(i) Several (currently nine) oligomeric MPs, both of the α-helix bundle and of the β-barrel
types, have been folded or expressed in vitro using detergents (Fig. 6.4; reviewed in Popot
2014) – a priori, because of their dissociating properties, much less favorable an environment than APols.
6.3 Amphipol-Assisted Folding of Membrane Proteins
353
such an electrostatic mechanism, if present, is not essential.
A related approach, cell-free synthesis of MPs using APols as the accepting medium, is
described in Chap. 7.
A thought-provoking point is why APols seem to provide such a favorable folding medium for
MPs, when neither their chemical structure nor the supramolecular structures they assemble into in
aqueous solutions bear any similarity to lipid bilayers. This puzzling question is briefly discussed in
Box 6.3.
Box 6.3 Why Do Membrane Proteins Adopt Their Native Fold in the Absence of
a Membrane?
To the cell biologist, it may seem strange that such good MP folding yields can be obtained in
APols. MPs are obviously adapted to a membrane environment, viz. the way they expose them to
highly hydrophobic surfaces whose depth matches that of the membrane’s hydrophobic core.
(Re)folding experiments show that an unfolded MP can fold back to its native state if transferred
either to lipid vesicles or to a “membrane-mimetic” environment (§ 6.2 and 6.3). Yet, can APols
really be considered as “membrane-mimetic”, when their chemical structure and the supramolecular
aggregates they form in water (Chap. 4) are so different, chemically and physically, from a lipid
bilayer? Does the high efficiency of MP folding in APols (Table 6.1) tell us something about the
constraints MPs require to fold to their native state?
This question has been debated in a discussion paper to which the reader is referred (Popot
and Engelman 2016). In brief, it is proposed that, because of the conditions in which MPs are
inserted into membranes in vivo, relying, for folding, on the physical constraints provided by a lipid
bilayer would be counterproductive and lead to low folding yields. This is because a folding MP
in vivo is exposed to a highly complex environment that has little to do with a lipid bilayer, crowded
as it is with proteins. Furthermore, physical constraints and chemical interactions differ from one
membrane to the next, so that a given MP will usually be exposed during its synthesis and insertion
to constraints different from those that prevail in the membrane compartment in which it will fulfill
its function. These considerations lead to the proposal that MPs have evolved to be adapted to a
membrane environment, but rely primarily, for their folding, on (i) their TM region being protected
from the aqueous phase, without specific requirements about how this protection is afforded, and
(ii) finding within their own sequence the information required to fold into a functional 3D
structure, without guidance by physical constraints from their environment, such as bilayer
thickness, lateral pressure gradient, etc. This is not to say, of course, that lipids have no influence
on MP structure and function; much to the contrary (cf. Chap. 1). In some cases, the lipid
composition can even modulate the TM topology adopted by a given MP (see e.g. Bogdanov
et al. 2014; Vitrac et al. 2017; and references therein). For a full discussion of this intriguing
question, the reader is referred to Popot and Engelman (2016).
6.3.3
Challenges and Prospects
One of the many open questions is which types of MPs will turn out to be amenable to APol-assisted
folding and which not. MPs with extended extramembrane domains have not been tested yet (OmpA
does feature such a domain, but without a function that could be easily tested to assess its folding). One
may note, however, in this respect, that APols have been observed to protect denatured soluble proteins
from aggregation (Ma et al. 2012; Martin et al. 2014, 2015), which should constitute a favorable factor.
A particularly challenging case is that of oligomeric MPs. There are, however, reasons to expect
that, there as well, the use of APols might provide interesting perspectives:
(i) Several (currently nine) oligomeric MPs, both of the α-helix bundle and of the β-barrel
types, have been folded or expressed in vitro using detergents (Fig. 6.4; reviewed in Popot
2014) – a priori, because of their dissociating properties, much less favorable an environment than APols.
6.3 Amphipol-Assisted Folding of Membrane Proteins
353