at a level of !1 mg per mL of lysate, an amount comparable to that which can be obtained in vivo in
1 L of Escherichia coli culture. Lysates are most often prepared from E. coli, but extracts from
archaebacteria, protozoans, yeasts, wheat germs and other higher plant cells, insect cells, rabbit
reticulocytes, or cultured human cell lines can also be employed with the view of facilitating either
folding or posttranslational modifications of the target proteins. This comes, however, at the expense of
simplicity, yield, and/or cost control (reviewed and critically compared in Zemella et al. 2015).
Cell-free expression offers many attractive features, among which are to do away with the
toxicity issue and to allow labeling using limited amounts of isotopically labeled or non-natural amino
acids for X-ray crystallography or for solution or solid-state NMR structural investigations, etc. (see
e.g. Kigawa et al. 1999; Kigawa 2010; Maslennikov et al. 2010; Reckel et al. 2011; Abdine et al.
2012). In addition, it is possible to supply the lysate with lipids, detergents, or other surfactants,
thereby providing the overexpressed proteins with an amphipathic environment destined to keeping it
soluble and facilitating its folding. This makes CFE an extremely interesting approach for the
production of MPs.
7.2
Context: Cell-Free Expression of Membrane Proteins
The putative advantages of CFE of MPs over in vivo expression are schematized in Fig. 7.2. On the left
side of the figure are indicated the various steps required by in vivo approaches and the associated
potential bottlenecks. Emphasized on the right side are the relative simplicity of in vitro expression and
the many compounds that can be added to the lysate in order to stabilize the target protein, favor its
folding, keep it soluble, or label it.
Application of CFE to MPs started in the mid-1990s (see e.g. Sonar et al. 1993; Van Gelder et al.
1994; Huppa and Ploegh 1997; Bogdanov and Dowhan 1998; for relatively recent reviews, the reader
is referred, for example, to Rajesh et al. 2011; Ge and Xu 2012; Kai et al. 2012, 2015; Maeda and
Schertler 2013; Hein et al. 2014; Rues et al. 2014, 2016; Sachse et al. 2014; Henrich et al. 2015; as well
as to the list of MPs expressed by CFE compiled in Popot 2014). The most commonly used lysates are
derived either from E. coli or from wheat germ, supplemented or not with surfactants (for a recent
discussion of the pros and cons of prokaryotic vs. eukaryotic lysates, see Zemella et al. 2015). MP
production by CFE has long remained marginal, but it picked up steam after, roughly, 2003 (Figs. 7.3
and 7.4A).
MPs obtained in vitro using CFE can be synthesized either in the presence of a surfactant – lipid
vesicles, nanodiscs, bicelles, detergent or detergent-lipid mixed micelles, APols, etc. – in which case
they probably insert and fold in the course of biosynthesis, or they can be left to precipitate from a
surfactant-free lysate. The precipitates thus obtained are usually much easier to dissolve than inclusion
bodies and can often be solubilized using a non-denaturing detergent (see e.g. Klammt et al. 2004,
2005, 2006; Keller et al. 2008; Junge et al. 2010; Hein et al. 2014; and references therein). In some
cases, however, they are dissolved using either SDS or urea (see e.g. Focke et al. 2016), which makes
it necessary to fold the target MP from the denatured state in which it is recovered (for an overview,
see Chap. 6, § 6.2, or Popot 2014). E. coli lysates have yielded by far the largest amount of properly
folded MPs (28 of them by the end of 2013), but since 2007, the use of wheat germ lysates has
been progressing (Fig. 7.4A). Other lysates are being tested, e.g. the reticulocyte lysate supplemented
with dog pancreas microsomes that has been used to express, fold, and assemble a T cell receptor-CD3
complex (Huppa and Ploegh 1997).
CFE has been much more often used to express α-helical than β-barrel MPs (32 vs. 4 at the end
of 2013; Fig. 7.4B). However, this situation may have a circumstantial rather than a real technical
basis. Indeed, β-barrel MPs appear particularly easy to fold from inclusion bodies (see Buchanan et al.
7.2 Context: Cell-Free Expression of Membrane Proteins
363
1 L of Escherichia coli culture. Lysates are most often prepared from E. coli, but extracts from
archaebacteria, protozoans, yeasts, wheat germs and other higher plant cells, insect cells, rabbit
reticulocytes, or cultured human cell lines can also be employed with the view of facilitating either
folding or posttranslational modifications of the target proteins. This comes, however, at the expense of
simplicity, yield, and/or cost control (reviewed and critically compared in Zemella et al. 2015).
Cell-free expression offers many attractive features, among which are to do away with the
toxicity issue and to allow labeling using limited amounts of isotopically labeled or non-natural amino
acids for X-ray crystallography or for solution or solid-state NMR structural investigations, etc. (see
e.g. Kigawa et al. 1999; Kigawa 2010; Maslennikov et al. 2010; Reckel et al. 2011; Abdine et al.
2012). In addition, it is possible to supply the lysate with lipids, detergents, or other surfactants,
thereby providing the overexpressed proteins with an amphipathic environment destined to keeping it
soluble and facilitating its folding. This makes CFE an extremely interesting approach for the
production of MPs.
7.2
Context: Cell-Free Expression of Membrane Proteins
The putative advantages of CFE of MPs over in vivo expression are schematized in Fig. 7.2. On the left
side of the figure are indicated the various steps required by in vivo approaches and the associated
potential bottlenecks. Emphasized on the right side are the relative simplicity of in vitro expression and
the many compounds that can be added to the lysate in order to stabilize the target protein, favor its
folding, keep it soluble, or label it.
Application of CFE to MPs started in the mid-1990s (see e.g. Sonar et al. 1993; Van Gelder et al.
1994; Huppa and Ploegh 1997; Bogdanov and Dowhan 1998; for relatively recent reviews, the reader
is referred, for example, to Rajesh et al. 2011; Ge and Xu 2012; Kai et al. 2012, 2015; Maeda and
Schertler 2013; Hein et al. 2014; Rues et al. 2014, 2016; Sachse et al. 2014; Henrich et al. 2015; as well
as to the list of MPs expressed by CFE compiled in Popot 2014). The most commonly used lysates are
derived either from E. coli or from wheat germ, supplemented or not with surfactants (for a recent
discussion of the pros and cons of prokaryotic vs. eukaryotic lysates, see Zemella et al. 2015). MP
production by CFE has long remained marginal, but it picked up steam after, roughly, 2003 (Figs. 7.3
and 7.4A).
MPs obtained in vitro using CFE can be synthesized either in the presence of a surfactant – lipid
vesicles, nanodiscs, bicelles, detergent or detergent-lipid mixed micelles, APols, etc. – in which case
they probably insert and fold in the course of biosynthesis, or they can be left to precipitate from a
surfactant-free lysate. The precipitates thus obtained are usually much easier to dissolve than inclusion
bodies and can often be solubilized using a non-denaturing detergent (see e.g. Klammt et al. 2004,
2005, 2006; Keller et al. 2008; Junge et al. 2010; Hein et al. 2014; and references therein). In some
cases, however, they are dissolved using either SDS or urea (see e.g. Focke et al. 2016), which makes
it necessary to fold the target MP from the denatured state in which it is recovered (for an overview,
see Chap. 6, § 6.2, or Popot 2014). E. coli lysates have yielded by far the largest amount of properly
folded MPs (28 of them by the end of 2013), but since 2007, the use of wheat germ lysates has
been progressing (Fig. 7.4A). Other lysates are being tested, e.g. the reticulocyte lysate supplemented
with dog pancreas microsomes that has been used to express, fold, and assemble a T cell receptor-CD3
complex (Huppa and Ploegh 1997).
CFE has been much more often used to express α-helical than β-barrel MPs (32 vs. 4 at the end
of 2013; Fig. 7.4B). However, this situation may have a circumstantial rather than a real technical
basis. Indeed, β-barrel MPs appear particularly easy to fold from inclusion bodies (see Buchanan et al.
7.2 Context: Cell-Free Expression of Membrane Proteins
363