Metabolic and genome engineering efforts to increase
membrane protein yield
In recent years, whole genome methods to increase recombinant protein
production have emerged. This has been mostly reported for E.
coli . However, S. cerevisiae still remains one of the more
popular eukaryotic expression platforms. Both model microbe systems are
well-characterized and contain reasonably-sized genomes that enable
genetic manipulation to be feasible. As our understanding of the
interplay between multiple intracellular processes continues to evolve,
this has shed light on additional indirect factors that influence the
outcome of protein expression. This requires a top-down systems level
analysis of simultaneous cell processes that take place under specific
environmental conditions. Metabolic engineering aims to address these
challenges. To look at how this approach can be applied to MP expression
requires a recognition and understanding of the processes not only
directly involved in but also those affected downstream by their
synthesis, folding, PTM, and translocation. Transcriptomics and
metabolomics analyses are two ways that the effects of growth conditions
can be assessed by looking at changes in transcript levels, mRNA, of
genes involved in certain metabolic pathways (R. Carlson et al., 2002;
Chae et al., 2017; Dromms & Styczynski, 2012; Guan et al., 2018; Park
et al., 2005; Trethewey, 2004; Yuan et al., 2018). Similarly,
metabolomics is used to assess signature metabolites and their relative
quantities, which are by-products from these pathways and interconnected
processes. Information can be gleaned from these studies about the
pathways that are most affected by expression conditions. Metabolic
engineering aims to divert resources that support identified systems
away from non-essential pathways and redirecting them toward recombinant
expression. The rationale is that this results in prolonged expression
leading toward increased production levels. Employing a rational
approach to achieve this requires a systems level understanding of the
host metabolism including those involved in transcription, translation
and flux of metabolites, which also includes proteins. An analysis of
the host expression platform helps in deciding which pathways are
critically effected during protein expression. As a result of these
analyses one strategy is to engineer KO yeast strains that are lacking
pathways least critical for survival of the organism. Also, targeting
genes responsible for the translation bottleneck created when strong
promoters generate higher rates of mRNA synthesis helps alleviate this
stress and associated degradation of mRNA. This was successfully
demonstrated in E. coli when 36 genes responsible for
non-essential functions were selectively knocked out (Sharma et al.,
2020; L. Zhou et al., 2022). The results showed that protein expression
could be increased by 1.5-fold when carried out at lower temperatures,
25 °C.
Alternatively, knockin (KI) strains also show promise as a means of
increasing production. In this way, strains are engineered to carry out
important processes that lead to proper protein function. This is a
common strategy employed in the latest gene editing technologies such as
CRISPR/Cas9. This allows for targeted gene insertion to coax the host
into producing gene products that are either non-native to the host
organism or to restore gene function that has been lost (Giuliano et
al., 2019; Ye et al., 2021). Recently, CRISPR/Cas9 technology was used
successfully to introduce the T7 RNA polymerase gene, an established
robust expression system, into a strain of E. coli naturally
lacking this system (Ye et al., 2021). While some genome-wide strategies
have been used successfully in E. coli , an extensive exploration
of these methods in S. cerevisiae has not yet been reported. The
addition of folding chaperones such as Hsp150 and PDI1 has seen some
success, mostly for the recombinant expression of soluble proteins (Kim
et al., 2014). Yeast strains deficient in lysosomal compartment
proteases along with other organellar proteases is another strategy that
has been implemented to circumvent protein degradation, truncation and
increase yields by at least 10-fold (Tomimoto et al., 2013). Processes
such as glycosylation also have significant effects on protein yield and
quality control. Mutagenesis efforts have been directed toward
humanizing yeast strains to produce glycosylation patterns on
recombinantly expressed proteins similar to those produced in humans
(Hamilton & Gerngross, 2007) (Figure 2). As proof of concept, humans,
for example, lack the ability to produce sialic
acid, N -glycolylneuraminic acid, while other mammals can produce
it. This leads to aberrant sialylation. Knocking out its production in
mammalian and other cell lines can potentially alleviate this problem.
This is an important consideration in the design and production of
therapeutic proteins where posttranslational differences in the
resulting protein can have severe immunological effects. For MPs along
with other non-therapeutic proteins, glycosylation affects folding,
stability, half-life and function (Helenius & Aebi, 2001; Mitra et al.,
2006). All of which are important considerations for high-yield,
high-quality recombinant expression.