genes encoding transcription factors and transporters (Legras et al. 2010). The role of the
identified transporters Tpo1 and Pdr12 in octanoic and decanoic acid efflux was further analyzed in growth tests with knockout mutants
(Legras et al. 2010). An overexpression of such
efflux-involved transporters is one approach
that could increase product yields—in the
case, that secretion is a bottleneck. For example, overexpression of Pdr12 was shown to
increase the secretion of short branched-chain
FA at early time points of production (Yu et al.
2016). However, transporter overexpression
can also have unwanted side effects, such as
slower growth, depleting ATP from other cellular processes or altering the plasma membrane
composition, as was hypothesized based on
other Pdr12 overexpression studies (Nyga ˚rd
et al. 2014). Due to a lack of yeast efflux
pumps known to specifically transport FA and
their derivatives, another viable approach is the
screening of heterologous transporters. In a
recent study, 12 human or Arabidopsis thaliana
(putative) transporters were screened, and one
of them, human FATP1, improved overall cell
fitness and fatty alcohol (Hu et al. 2018) as well
as 1-alkene (Zhou et al. 2018) production and
secretion in S. cerevisiae. For alkane resistance
of S. cerevisiae, endogenous efflux pumps,
namely, Snq2 and Pdr5, as well as heterologous
transporters, namely, Y. lipolytica ABC2 and
ABC3, were shown to have a positive effect
(Chen et al. 2013a; Ling et al. 2013).
2. Methods for Increasing Yeast Robustness
To avoid the re-entrance of the products, the
robustness of the plasma membrane needs to
be improved. Octanoic acid, for instance, disrupts the plasma membrane composition, leading to membrane leakage and cell death (Legras
et al. 2010; Liu et al. 2013). This effect was
reduced by rationally engineering plasma
membrane composition through increasing
the oleic acid content, either by external supply
(Liu et al. 2013) or by expression of a mutated
AcCoA carboxylase, Acc1
S1157A
(BesadaLombana et al. 2017). The increase in the average chain length of membrane FA, as well as
higher cis-monounsaturated FA levels, was
shown to provide higher tolerance to toxic FA
(Liu et al. 2013; Besada-Lombana et al. 2017).
As tolerance is usually a complex phenotype, which cannot solely be improved by rational engineering of single genes, adaptive
laboratory evolution (ALE) is an interesting
alternative. In an ALE experiment, a strain is
grown over many generations, with increasing
concentrations of the toxic compound, for
selection of cells with an enhanced tolerance
phenotype. Causal mutations in the final strain
are identified by whole genome sequencing
(Dragosits and Mattanovich 2013; Mans et al.
2018). When performed with E. coli, an evolved
strain not only showed higher tolerance to octanoic acid but also produced higher titers, which
was attributed to changes in membrane composition and fluidity (Royce et al. 2015). A similar
approach would be possible for S. cerevisiae, for
which ALE has been performed successfully, for
example, for increased resistance to high temperatures (Caspeta et al. 2014) and acetic acid
(Gonza ´lez-Ramos et al. 2016) or alcohol tolerance (Gonza ´lez-Ramos et al. 2013; Davis Lo ´pez
et al. 2018), but not for short- or medium-chain
FA tolerance.
A method that can lead to similar results as
ALE was termed global transcription machinery engineering, gTME (Alper et al. 2006). It
can induce a remodeling of the transcriptome
and therefore target polygenic traits, such as
tolerance. gTME relies on the random mutagenesis of a transcription factor that regulates
the transcription of several genes and was successfully applied to enhance S. cerevisiae tolerance to ethanol (Alper et al. 2006), as well as to
adapt to growth on lignocellulosic hydrolysates
through improved xylose utilization (Liu et al.
2011). When using this method in combination
with an appropriate selection for mutants with
increased growth in the presence of the toxic
FA, it could be a valuable tool for future tolerance engineering but has so far not been
applied to S. cerevisiae for FA tolerance.
14 Engineering Saccharomyces cerevisiae for Production of Fatty Acids and Their Derivatives
353
identified transporters Tpo1 and Pdr12 in octanoic and decanoic acid efflux was further analyzed in growth tests with knockout mutants
(Legras et al. 2010). An overexpression of such
efflux-involved transporters is one approach
that could increase product yields—in the
case, that secretion is a bottleneck. For example, overexpression of Pdr12 was shown to
increase the secretion of short branched-chain
FA at early time points of production (Yu et al.
2016). However, transporter overexpression
can also have unwanted side effects, such as
slower growth, depleting ATP from other cellular processes or altering the plasma membrane
composition, as was hypothesized based on
other Pdr12 overexpression studies (Nyga ˚rd
et al. 2014). Due to a lack of yeast efflux
pumps known to specifically transport FA and
their derivatives, another viable approach is the
screening of heterologous transporters. In a
recent study, 12 human or Arabidopsis thaliana
(putative) transporters were screened, and one
of them, human FATP1, improved overall cell
fitness and fatty alcohol (Hu et al. 2018) as well
as 1-alkene (Zhou et al. 2018) production and
secretion in S. cerevisiae. For alkane resistance
of S. cerevisiae, endogenous efflux pumps,
namely, Snq2 and Pdr5, as well as heterologous
transporters, namely, Y. lipolytica ABC2 and
ABC3, were shown to have a positive effect
(Chen et al. 2013a; Ling et al. 2013).
2. Methods for Increasing Yeast Robustness
To avoid the re-entrance of the products, the
robustness of the plasma membrane needs to
be improved. Octanoic acid, for instance, disrupts the plasma membrane composition, leading to membrane leakage and cell death (Legras
et al. 2010; Liu et al. 2013). This effect was
reduced by rationally engineering plasma
membrane composition through increasing
the oleic acid content, either by external supply
(Liu et al. 2013) or by expression of a mutated
AcCoA carboxylase, Acc1
S1157A
(BesadaLombana et al. 2017). The increase in the average chain length of membrane FA, as well as
higher cis-monounsaturated FA levels, was
shown to provide higher tolerance to toxic FA
(Liu et al. 2013; Besada-Lombana et al. 2017).
As tolerance is usually a complex phenotype, which cannot solely be improved by rational engineering of single genes, adaptive
laboratory evolution (ALE) is an interesting
alternative. In an ALE experiment, a strain is
grown over many generations, with increasing
concentrations of the toxic compound, for
selection of cells with an enhanced tolerance
phenotype. Causal mutations in the final strain
are identified by whole genome sequencing
(Dragosits and Mattanovich 2013; Mans et al.
2018). When performed with E. coli, an evolved
strain not only showed higher tolerance to octanoic acid but also produced higher titers, which
was attributed to changes in membrane composition and fluidity (Royce et al. 2015). A similar
approach would be possible for S. cerevisiae, for
which ALE has been performed successfully, for
example, for increased resistance to high temperatures (Caspeta et al. 2014) and acetic acid
(Gonza ´lez-Ramos et al. 2016) or alcohol tolerance (Gonza ´lez-Ramos et al. 2013; Davis Lo ´pez
et al. 2018), but not for short- or medium-chain
FA tolerance.
A method that can lead to similar results as
ALE was termed global transcription machinery engineering, gTME (Alper et al. 2006). It
can induce a remodeling of the transcriptome
and therefore target polygenic traits, such as
tolerance. gTME relies on the random mutagenesis of a transcription factor that regulates
the transcription of several genes and was successfully applied to enhance S. cerevisiae tolerance to ethanol (Alper et al. 2006), as well as to
adapt to growth on lignocellulosic hydrolysates
through improved xylose utilization (Liu et al.
2011). When using this method in combination
with an appropriate selection for mutants with
increased growth in the presence of the toxic
FA, it could be a valuable tool for future tolerance engineering but has so far not been
applied to S. cerevisiae for FA tolerance.
14 Engineering Saccharomyces cerevisiae for Production of Fatty Acids and Their Derivatives
353