structures, lest none of them be thoroughly explored. As a redeeming feature, once a basic structure has
been carefully validated, it is (usually) a relatively simple matter to identify and fine-tune the chemistry
required to develop an array of derivatives, each carrying one or another label or tag (§ 4.4). Most such
modifications will not significantly affect the physical-chemical behavior of the polymer (see § 4.4.2).
Therefore, the grinding work that has been invested upstream to validate the basic molecule and
understand its advantages and drawbacks will usually remain exploitable.
In the following survey, we will successively consider (i) the original polyacrylate-based APols
(§ 4.2.1), (ii) other ionic APols (§ 4.2.2), and (iii) non-ionic APols (§ 4.2.3). A selection of polymers
that have been used to keep MPs soluble in aqueous solutions, whether they qualify as APols or not, is
provided in Table 4.1, along with some key references.
4.2.1
Polyacrylate-Based Amphipols with Carboxylates as Their Hydrophilic
Moieties
The concept of APols emerged from discussions between two polymerists, Roland Audebert and
Christophe Tribet, and a membrane biologist, myself. R. Audebert’s laboratory had been studying the
properties of long poly(acrylic acid) (PAA) molecules sparsely modified with relatively long alkyl
chains (C 12 or C 18 ). By themselves, these polymers disperse in water as individual molecules and form
fluid solutions. However, in the presence of detergent, the alkyl chains partition into detergent
micelles, creating nodes, and, thereby, a meshwork of polymers, turning the solution into a gel (see
e.g. Sarrazin-Cartalas et al. 1994). R. Audebert was initially interested in forming nodes with MPs
instead of micelles. I was only moderately enthused by the idea of using my hard-to-purify and fragile
MPs to make gels, but much more so to devise ways to make them more stable in aqueous solutions.
This led me to suggest making much shorter polymers, so that each protein would surround itself with
several of them and cross-linking would be limited or avoided. The resulting complexes might be
water-soluble, and, hopefully, eliminating the detergent would stabilize the protein (cf. Chap. 2).
Because long hydrophilic loops extending into the solution would be a hindrance for many biophysical
applications, such as crystallization or NMR, the alkyl chains had to be spaced much more closely than
in Roland’s original polymers, which meant that, in order to achieve a high enough aqueous solubility,
they had to be short, which led us to settle for octyl chains. A highly flexible main chain, such as
provided by PAA, was also a desirable feature, because it would facilitate a close coverage of the
corrugated surface of the proteins and the formation of small particles. Because we were entering a
totally unchartered territory, we decided to play both on the length of the polymers, which would
determine how many of them would bind to a given protein and modulate the risks of cross-linking
(cf. Yamamoto et al. 2000b), and on the charge density, which would affect the persistence length
(rigidity) of the polymers and, perhaps, the stability of the protein. This led to the design of the four
polymers, A8-35, A8-75, A34-35, and A34-75, described in the princeps publication (Tribet et al.
1996).
A8-35 (Fig. 4.1a) was obtained by grafting a commercial ~5-kDa PAA preparation first with
octylamine, yielding A8-75, and then with isopropylamine (Fig. 4.2). The resulting amphipathic
polymer comprises ~25 octyl chains per 100 monomers, ~40 isopropyl groups, and ~35 free
carboxylates. Whereas carboxylates and octyl chains confer amphipathy to A8-35, the isopropyl
groups limit its charge density, which, if it is too high, seems to negatively affect MP stability (see
Chap. 5). Isopropyls do not contribute much to the global hydrophobicity at low or ambient temperature (Takei et al. 1993), but they do interact with the hydrophobic TM surface of MPs (see Chap. 5).
The average mass per octyl chain of the sodium salt of A8-35 is close to 500 Da, similar to that of many
detergents, but higher than that of those carrying octyl chains, such as octylglucoside or
4.2 Amphipol Chemical Structure and Synthesis
153
been carefully validated, it is (usually) a relatively simple matter to identify and fine-tune the chemistry
required to develop an array of derivatives, each carrying one or another label or tag (§ 4.4). Most such
modifications will not significantly affect the physical-chemical behavior of the polymer (see § 4.4.2).
Therefore, the grinding work that has been invested upstream to validate the basic molecule and
understand its advantages and drawbacks will usually remain exploitable.
In the following survey, we will successively consider (i) the original polyacrylate-based APols
(§ 4.2.1), (ii) other ionic APols (§ 4.2.2), and (iii) non-ionic APols (§ 4.2.3). A selection of polymers
that have been used to keep MPs soluble in aqueous solutions, whether they qualify as APols or not, is
provided in Table 4.1, along with some key references.
4.2.1
Polyacrylate-Based Amphipols with Carboxylates as Their Hydrophilic
Moieties
The concept of APols emerged from discussions between two polymerists, Roland Audebert and
Christophe Tribet, and a membrane biologist, myself. R. Audebert’s laboratory had been studying the
properties of long poly(acrylic acid) (PAA) molecules sparsely modified with relatively long alkyl
chains (C 12 or C 18 ). By themselves, these polymers disperse in water as individual molecules and form
fluid solutions. However, in the presence of detergent, the alkyl chains partition into detergent
micelles, creating nodes, and, thereby, a meshwork of polymers, turning the solution into a gel (see
e.g. Sarrazin-Cartalas et al. 1994). R. Audebert was initially interested in forming nodes with MPs
instead of micelles. I was only moderately enthused by the idea of using my hard-to-purify and fragile
MPs to make gels, but much more so to devise ways to make them more stable in aqueous solutions.
This led me to suggest making much shorter polymers, so that each protein would surround itself with
several of them and cross-linking would be limited or avoided. The resulting complexes might be
water-soluble, and, hopefully, eliminating the detergent would stabilize the protein (cf. Chap. 2).
Because long hydrophilic loops extending into the solution would be a hindrance for many biophysical
applications, such as crystallization or NMR, the alkyl chains had to be spaced much more closely than
in Roland’s original polymers, which meant that, in order to achieve a high enough aqueous solubility,
they had to be short, which led us to settle for octyl chains. A highly flexible main chain, such as
provided by PAA, was also a desirable feature, because it would facilitate a close coverage of the
corrugated surface of the proteins and the formation of small particles. Because we were entering a
totally unchartered territory, we decided to play both on the length of the polymers, which would
determine how many of them would bind to a given protein and modulate the risks of cross-linking
(cf. Yamamoto et al. 2000b), and on the charge density, which would affect the persistence length
(rigidity) of the polymers and, perhaps, the stability of the protein. This led to the design of the four
polymers, A8-35, A8-75, A34-35, and A34-75, described in the princeps publication (Tribet et al.
1996).
A8-35 (Fig. 4.1a) was obtained by grafting a commercial ~5-kDa PAA preparation first with
octylamine, yielding A8-75, and then with isopropylamine (Fig. 4.2). The resulting amphipathic
polymer comprises ~25 octyl chains per 100 monomers, ~40 isopropyl groups, and ~35 free
carboxylates. Whereas carboxylates and octyl chains confer amphipathy to A8-35, the isopropyl
groups limit its charge density, which, if it is too high, seems to negatively affect MP stability (see
Chap. 5). Isopropyls do not contribute much to the global hydrophobicity at low or ambient temperature (Takei et al. 1993), but they do interact with the hydrophobic TM surface of MPs (see Chap. 5).
The average mass per octyl chain of the sodium salt of A8-35 is close to 500 Da, similar to that of many
detergents, but higher than that of those carrying octyl chains, such as octylglucoside or
4.2 Amphipol Chemical Structure and Synthesis
153