A second obstacle is more cultural, if not societal, in nature. Biologists aim at understanding
how a live organism functions. For this reason, they strive to provide the proteins they work on with as
natural-like an environment as possible. Synthetic polymers do not look very natural, and the fact that
they nevertheless manage to preserve to a large degree one of the key features of the membrane
environment – lipid/protein interactions – is underestimated, as is, in my view, overestimated the
importance of the physical forces exerted by lipid bilayers. There is an illusion in believing that any
in vitro system, even the most seductive of them, that of protein-bound nanodiscs, provides a
solubilized membrane protein with a perfectly natural environment. One will always deal with a
more or less satisfying approximation thereof. This fear of straying too far away from nature is
compounded by the legitimate but paralyzing concern that, even if one has carefully weighed the
pros and cons of the various experimental systems and chosen that providing the best compromise
between biological relevance and experimental expediency, some systems will make publication or
grant approval easier than others. Journal referees and granting agencies tend to look more favorably to
works where one’s protein is expressed under a functional form in eukaryotic cells rather than folded
from inclusion bodies produced in bacteria, and where it is studied in a supposedly native lipid bilayer
rather than an “artificial” environment. All this is understandable, but it tends to push experimenters to
use very costly and complex procedures where cheaper, simpler, and more resolutive ones would have
led them more rapidly to the same conclusions, at least at a first level of description. Yet, detailed
information – e.g. an atomic model – gathered thanks to the simple and convenient but more “artificial”
system provided by amphipols can be checked against data, possibly less accurate or more fragmentary, obtained in more sophisticated and work-intensive in vitro systems, such as nanodiscs, and/or
spot measurements carried out in vivo. Fortunately, there is a growing realization that each experimental system has its advantages and limitations, that none of them is systematically the best, and that
the risk of falling victim to distortions and perturbations induced by one particular medium can be
reduced by confronting results obtained in different media. The fact that identical results have been
obtained, e.g. in electron cryomicroscopy, using amphipols and nanodiscs ought to make the former
more widely accepted as a medium in which to gather biologically relevant data.
A third obstacle, which is of a very practical nature, is due to the difficulties in providing the
community of biologists with a supply of diverse and well-controlled molecules. Polymer chemistry is
not always easy, and the fact that the final product is necessarily a mixture does not simplify the
standardization of the batches. Commercially speaking, amphipols are a “niche” market in the sense
that, as long as no large-scale investigations rely on their use, the amounts sold will always remain
relatively minute. This does not encourage private companies to invest in scaling up synthesis routes
and developing quality control procedures. It explains why, whereas a large number of amphipol
structures have been synthesized by organic chemists and validated by biochemists and biophysicists,
only a handful of them are commercially available. Having to rely on the generosity of academic
colleagues can be uncomfortable. For one thing, it entails the risk that a change in scientific priorities,
the loss of a key coworker, or the drying up of funding cause production to be discontinued before a
given research project is completed. For another, it can be compromised by unreasonable demands in
terms of publication signatures. The simplest route out of this predicament is probably that networks of
laboratories interested in exploiting a given type of amphipol or amphipol derivative combine their
resources to have sufficiently large batches synthesized and distributed, making it profitable for a
company to develop the production. Once the activation barrier of setting up the procedures has been
overcome and the initial investment has paid off, maintaining a steady production is relatively
straightforward. Among the many important roles that organic chemists still have to play in this
field, one will be to develop molecules whose synthesis is increasingly simpler, cheaper, and more
robust, so as to facilitate their dissemination.
Final Comments
685
how a live organism functions. For this reason, they strive to provide the proteins they work on with as
natural-like an environment as possible. Synthetic polymers do not look very natural, and the fact that
they nevertheless manage to preserve to a large degree one of the key features of the membrane
environment – lipid/protein interactions – is underestimated, as is, in my view, overestimated the
importance of the physical forces exerted by lipid bilayers. There is an illusion in believing that any
in vitro system, even the most seductive of them, that of protein-bound nanodiscs, provides a
solubilized membrane protein with a perfectly natural environment. One will always deal with a
more or less satisfying approximation thereof. This fear of straying too far away from nature is
compounded by the legitimate but paralyzing concern that, even if one has carefully weighed the
pros and cons of the various experimental systems and chosen that providing the best compromise
between biological relevance and experimental expediency, some systems will make publication or
grant approval easier than others. Journal referees and granting agencies tend to look more favorably to
works where one’s protein is expressed under a functional form in eukaryotic cells rather than folded
from inclusion bodies produced in bacteria, and where it is studied in a supposedly native lipid bilayer
rather than an “artificial” environment. All this is understandable, but it tends to push experimenters to
use very costly and complex procedures where cheaper, simpler, and more resolutive ones would have
led them more rapidly to the same conclusions, at least at a first level of description. Yet, detailed
information – e.g. an atomic model – gathered thanks to the simple and convenient but more “artificial”
system provided by amphipols can be checked against data, possibly less accurate or more fragmentary, obtained in more sophisticated and work-intensive in vitro systems, such as nanodiscs, and/or
spot measurements carried out in vivo. Fortunately, there is a growing realization that each experimental system has its advantages and limitations, that none of them is systematically the best, and that
the risk of falling victim to distortions and perturbations induced by one particular medium can be
reduced by confronting results obtained in different media. The fact that identical results have been
obtained, e.g. in electron cryomicroscopy, using amphipols and nanodiscs ought to make the former
more widely accepted as a medium in which to gather biologically relevant data.
A third obstacle, which is of a very practical nature, is due to the difficulties in providing the
community of biologists with a supply of diverse and well-controlled molecules. Polymer chemistry is
not always easy, and the fact that the final product is necessarily a mixture does not simplify the
standardization of the batches. Commercially speaking, amphipols are a “niche” market in the sense
that, as long as no large-scale investigations rely on their use, the amounts sold will always remain
relatively minute. This does not encourage private companies to invest in scaling up synthesis routes
and developing quality control procedures. It explains why, whereas a large number of amphipol
structures have been synthesized by organic chemists and validated by biochemists and biophysicists,
only a handful of them are commercially available. Having to rely on the generosity of academic
colleagues can be uncomfortable. For one thing, it entails the risk that a change in scientific priorities,
the loss of a key coworker, or the drying up of funding cause production to be discontinued before a
given research project is completed. For another, it can be compromised by unreasonable demands in
terms of publication signatures. The simplest route out of this predicament is probably that networks of
laboratories interested in exploiting a given type of amphipol or amphipol derivative combine their
resources to have sufficiently large batches synthesized and distributed, making it profitable for a
company to develop the production. Once the activation barrier of setting up the procedures has been
overcome and the initial investment has paid off, maintaining a steady production is relatively
straightforward. Among the many important roles that organic chemists still have to play in this
field, one will be to develop molecules whose synthesis is increasingly simpler, cheaper, and more
robust, so as to facilitate their dissemination.
Final Comments
685