Changing culture conditions to tune expression
Before some of the cloning and engineering strategies previously
discussed, “simpler” approaches may also prove helpful. High-yields
and high-quality are not necessarily linked and higher-yields may, in
fact, produce large quantities of unfolded, non-functional protein. When
this outcome is not intended, for example, if avoiding inclusion body
formation is desired, lowering the culture temperature during protein
induction can lead to higher yields of functional, high-quality material
(Francis & Page, 2010; Li et al., 2001; San-Miguel et al., 2013;
Weickert et al., 1997). This has mostly been explored in E. coli,
however, and little is reported on low temperature cultivation effects
in yeast. The use of synthetic media also facilitates slowed expression
and production compared to nutrient-rich media. Lastly, like strong
promoters, growth conditions can also affect plasmid copy number
(Stueber & Bujard, 1982).
A promising field that has gained much traction over the last decade is
the use of CF systems for high-yield expression. Once used primarily as
a research tool to understand transcription and translation has now been
adapted to produce protein products at scale for synthetic biology
(Garenne et al., 2021; Silverman et al., 2020). CF systems do not suffer
from the production bottlenecks presented by their whole-cell
counterparts, but there are other important considerations. The
intracellular systems and processes that accompany traditional in
vivo expression have effectively been removed, leaving only the
essential protein expression and production machinery. Prokaryotic CF
systems emerged using components isolated from E. coli where
recombinantly expressed, purified, components have been used
successfully – “Protein synthesis Using Recombinant Elements” (PURE
system) (E. D. Carlson et al., 2012; Klammt et al., 2006; Ohashi et al.,
2010). Success has also been met using crude cell lysis extract (Kigawa
et al., 2004). Eukaryotic CF expression has been developed from a
variety of sources ranging from wheat germ to Chinese hamster ovary
(CHO) cells to HeLa cells (Anderson et al., 1989; Brödel et al., 2014;
Harbers, 2014; Weber et al., 1975). There are only a few reports of CF
systems using expression machinery isolated from yeast and most utilize
soluble proteins to demonstrate feasibility of the system (Hodgman &
Jewett, 2013). Extracts from E. coli have been utilized
successfully in CF expression on a multi-liter scale (Figure 3). The
reactions are stable for several hours and can be scaled to produce
gram-level quantities of material (E. D. Carlson et al., 2012). CF yeast
systems have also been utilized successfully to produce active firefly
luciferase at µg-scale (Hodgman & Jewett, 2013). While most of the
large-scale production efforts described utilize soluble proteins,
strategies to adapt this technique to MPs are summarized nicely in
several reviews (Klammt et al., 2006; Sachse et al., 2014). CF In Vitro
Transcription/Translation reaction (IVTT) systems provide a solution to
the potentially limited space available at the plasma membrane that is
necessary to support properly folded, functional MPs. Arguably, it also
leaves proteins less susceptible to the structural and conformational
defects that arise from overcrowding in the membrane during
overexpression. Rather, in the absence of the native cell-membrane, CF
systems rely on artificially supplemented media to support MP
solubility. Mild detergents and other lipids that assume a bilayered
morphological structure are added to the synthetic milieu to keep newly
synthesized hydrophobic MPs soluble and conformationally active (Klammt
et al., 2006). CF systems streamline the production process by omitting
the need for any additional protein refolding protocols to be
established prior to characterization. One important consideration is
the compatibility of detergent and lipid additives not only to the
target protein, but also their effects, if any, on the viability of the
IVTT reaction components. Also, the concentration of detergent and
lipids required to maintain protein solubility and active reaction
components must be determined as well. Detergents must be maintained at
a concentration above their critical micelle concentration, CMC, to
maintain effective solubility properties (Kalipatnapu & Chattopadhyay,
2005; Sachse et al., 2014). Though this is perhaps a slight
simplification of the complex biophysical interactions between MPs and
solubilizing detergents and lipids. At any given time during the course
of the reaction, the concentration of freely available detergent and
associated micelles fluctuates during expression as proteins are
synthesized and occupy micelle aggregates. Further complicating this
process are the changes in by-product accumulation accompanying the
reaction. These are important considerations when optimizing the
synthesis reaction, but advantageously, carrying this out in a CF system
also provides flexibility in the amount of detergent that can be added,
changing the solubilizing capacity of the reaction mixture. Generally,
nonionic detergents such as Brij and Triton X-100 detergents can be
scaled at minimal cost. Other proprietary detergent mixtures such as
Empigen BB®, which is comprised of a heterogeneous
distribution of varying chain length molecules, is also mild and
nondenaturing (Lowthert et al., 1995). These types of detergents are
least likely to cripple the functionality of the components in the
reaction mixture leading to longer reaction times and increased protein
yields. Further, orchestrating a system to remove reaction by-products
as they accumulate and adding substrates can facilitate longer reaction
times and improve yields (Schoborg et al., 2014). Detergent solubilized
proteins can be later reconstituted into bilayered systems for
downstream analysis (Rieth et al., 2020).
Conclusions and Future Directions
A plethora of strategies along with synthetic biology tools for
high-yield recombinant MP production is described. Integral MPs remain
one of the most challenging expression targets due to their extreme
hydrophobicity, which presents challenges to most host organisms when
expressed at high levels. Overexpression often involves recruiting
rigorous gene expression conditions that lead to activation of
stress-related intracellular pathways in yeast that can thwart efforts
to achieve high yields. Engineering native and nonnative hybrid
promoters can help alleviate these problems as well as coupling
activator and repressor sites upstream of the target protein.
Consideration of terminator sequences provides another way to fine-tune
expression and slow the coupled transcription-translation process to
help avoid overwhelming the intracellular translation machinery. Many of
the strategies summarized in this regard relate to the use of expression
plasmids to achieve maximum yields. Indirect approaches have also been
described using metabolic engineering and other genome engineering
techniques to generate KO and KI strains with the aim of redirecting
yeast metabolism and nutrient resources toward producing the target
protein. The use of metabolic/genome engineering to halt non-essential
processes has been demonstrably successful in some cases for soluble
proteins and has yet to be applied to MP expression. Generating KOs to
remove unwanted protease activity can also help increase yields,
although increasing yields is not always coupled with high-quality,
functional proteins. This review is intended to summarize the tools
available to MP researchers looking to produce proteins for
characterization. It describes best approaches that can be applied to
maximize yields, which is often a bottleneck to progress in areas of
structural biology and other types of analyses. It is also intended to
help define areas in need of further development and where future
efforts could be beneficially directed so that more can be learned about
the processes critical to expression. While still somewhat a new
methodology, CF expression for MPs holds tremendous promise as an
alternative to heterologous expression in microbes by removing some of
the critical barriers to overexpression and maximizing yields. Further,
similar strategies used for synthetic biology in microbial systems have
been attempted in CF systems with some success. Genetic components;
transcription/translation, promoters, regulatory elements are beginning
to be explored in CF systems and already make use of well-characterized,
robust systems such as the E. coli -based T7 transcription system
(Garenne et al., 2021). Much of the progress seen with CF systems has
been derived from E. coli . However, the same limitations present
in E. coli expression still plagues its CF counterpart. Namely,
the ability to carry out PTMs. Although, less is known about CF systems
using yeast-derived components much of its progress has been pioneered
by Jewett and co-workers and could potentially complement E.
coli -derived systems for this reason, but without the limitations of
the finite space in the plasma membrane. Along these lines, it may be
advantageous to use CF in tandem with microbial systems by testing an
expression system in vitro prior to transferring the genetic
components to a cell-based system. CF systems are amenable to laboratory
automation and potential scale-up making high-throughput testing
possible, which can help supplement arduous efforts to screen single
expression conditions at one time. CF also makes incorporation of
unnatural amino acids possible for biophysical and other downstream
analysis.