Indeed, sequential elimination of three introns in a
gene encoding a protease from the thermophilic fungus
Malbranchea cinnamomea cumulatively reduced production levels in T. reesei (Paloheimo et al. 2016).
Interestingly, the largest reduction was observed
when the intron closest to the transcriptional start
site was removed. Similar results were obtained when
a gene encoding an antifungal protein from Aspergillus
giganteus was expressed in Trichoderma viride (Xu and
Gong 2003). Moreover, the inclusion of an artificial
intron was required for the heterologous production
of GFP and mRFP in the basidiomycete Armillaria
mellea (Ford et al. 2016).
Since it is not straightforward to predict the
impact of retaining introns in a foreign gene on
the final yield of mRNA, it may be useful to try
different gene variants containing no, a few
selected, or all introns, to optimize expression
levels from the GOI.
2. Codon Optimization
Codon usage may vary dramatically between
species (Iriarte et al. 2012) and, consequently,
influence how well a heterologous mRNA is
translated into a protein, reviewed in (Hanson
and Coller 2018; Tanaka et al. 2014). For example, codon composition may change the chromatin structure of the GOI affecting gene
transcription efficiency (Zhou et al. 2016),
whereas rare codons in the beginning of the
transcript may determine the efficiency of
translation initiation (Pop et al. 2014). Moreover, if the mRNA contains abnormally high
levels of rare codons, this may result in premature transcription termination and production
of reduced length mRNA for the ORF (Zhou
et al. 2018). Conversely, rare codons may be
used in the native host as translation pause
sites required for proper and timely folding of
subdomains of the protein structure. If such
codons are replaced by frequently used codons,
overall folding may be compromised (Yu et al.
2015; Zhou et al. 2015).
Different algorithms exist in order to
address the different challenges concerning
optimal codon choices for heterologous protein
production (Gould et al. 2014). Construction of
a codon-optimized gene usually requires de
novo synthesis of the entire gene, which can
be done by fusing a set of overlapping oligonucleotides (Sect. III.E) (Hoover and Lubkowski
2002) or simply be acquired from a commercial
source. Codon optimization may benefit heterologous protein production significantly, especially if the source of the foreign gene and the
new host are distantly related species. However, it is important to stress that predicting
the optimal codon composition for heterologous gene expression is a discipline still in its
infancy. It is therefore often an advantage to
test more than one gene variant during cell
factory construction. This can be exemplified
by the extracellular yield from T. reesei expressing a codon-optimized mammalian monoclonal immunoglobulin (IgG) antibody, which
showed a 40-fold difference, dependent on
which of two companies had performed the
codon optimization (Lin et al. 2006).
B. Selection Markers
Selection markers are mainly used to facilitate
identification of transformants containing the
gene-expression cassette (Dave et al. 2015).
Since targeted integration of the geneexpression cassette can be conducted in a
marker-free manner via CRISPR technology,
the presence of this bio-block in the cassette is
no longer essential. Nevertheless, selective markers will likely remain a common feature in a
gene-expression cassette since strain identification is simplified and more efficient with a
marker and since the marker may also serve
useful post-transformation roles. For example,
a selectable marker can be used to quickly
identify strains, or to promote breeding via
sexual and parasexual cycles, or to maintain
AMA1 plasmid-based expression. Moreover,
to achieve high expression levels, it may be
desirable to obtain a fungus with many copies
of the expression cassette. Such strains can be
selected by employing cassettes containing a
weak selection marker as the strains with a
high copy number will gain a fitness advantage
at heavy selection pressure (Wernars et al.
1985). In some cases, it is necessary to eliminate
the marker after transformation, for example, if
a marker-free strain is desirable or if a marker
234
J. K. H. Rendsvig et al.
gene encoding a protease from the thermophilic fungus
Malbranchea cinnamomea cumulatively reduced production levels in T. reesei (Paloheimo et al. 2016).
Interestingly, the largest reduction was observed
when the intron closest to the transcriptional start
site was removed. Similar results were obtained when
a gene encoding an antifungal protein from Aspergillus
giganteus was expressed in Trichoderma viride (Xu and
Gong 2003). Moreover, the inclusion of an artificial
intron was required for the heterologous production
of GFP and mRFP in the basidiomycete Armillaria
mellea (Ford et al. 2016).
Since it is not straightforward to predict the
impact of retaining introns in a foreign gene on
the final yield of mRNA, it may be useful to try
different gene variants containing no, a few
selected, or all introns, to optimize expression
levels from the GOI.
2. Codon Optimization
Codon usage may vary dramatically between
species (Iriarte et al. 2012) and, consequently,
influence how well a heterologous mRNA is
translated into a protein, reviewed in (Hanson
and Coller 2018; Tanaka et al. 2014). For example, codon composition may change the chromatin structure of the GOI affecting gene
transcription efficiency (Zhou et al. 2016),
whereas rare codons in the beginning of the
transcript may determine the efficiency of
translation initiation (Pop et al. 2014). Moreover, if the mRNA contains abnormally high
levels of rare codons, this may result in premature transcription termination and production
of reduced length mRNA for the ORF (Zhou
et al. 2018). Conversely, rare codons may be
used in the native host as translation pause
sites required for proper and timely folding of
subdomains of the protein structure. If such
codons are replaced by frequently used codons,
overall folding may be compromised (Yu et al.
2015; Zhou et al. 2015).
Different algorithms exist in order to
address the different challenges concerning
optimal codon choices for heterologous protein
production (Gould et al. 2014). Construction of
a codon-optimized gene usually requires de
novo synthesis of the entire gene, which can
be done by fusing a set of overlapping oligonucleotides (Sect. III.E) (Hoover and Lubkowski
2002) or simply be acquired from a commercial
source. Codon optimization may benefit heterologous protein production significantly, especially if the source of the foreign gene and the
new host are distantly related species. However, it is important to stress that predicting
the optimal codon composition for heterologous gene expression is a discipline still in its
infancy. It is therefore often an advantage to
test more than one gene variant during cell
factory construction. This can be exemplified
by the extracellular yield from T. reesei expressing a codon-optimized mammalian monoclonal immunoglobulin (IgG) antibody, which
showed a 40-fold difference, dependent on
which of two companies had performed the
codon optimization (Lin et al. 2006).
B. Selection Markers
Selection markers are mainly used to facilitate
identification of transformants containing the
gene-expression cassette (Dave et al. 2015).
Since targeted integration of the geneexpression cassette can be conducted in a
marker-free manner via CRISPR technology,
the presence of this bio-block in the cassette is
no longer essential. Nevertheless, selective markers will likely remain a common feature in a
gene-expression cassette since strain identification is simplified and more efficient with a
marker and since the marker may also serve
useful post-transformation roles. For example,
a selectable marker can be used to quickly
identify strains, or to promote breeding via
sexual and parasexual cycles, or to maintain
AMA1 plasmid-based expression. Moreover,
to achieve high expression levels, it may be
desirable to obtain a fungus with many copies
of the expression cassette. Such strains can be
selected by employing cassettes containing a
weak selection marker as the strains with a
high copy number will gain a fitness advantage
at heavy selection pressure (Wernars et al.
1985). In some cases, it is necessary to eliminate
the marker after transformation, for example, if
a marker-free strain is desirable or if a marker
234
J. K. H. Rendsvig et al.