viously observed in ADH-deficient strains
(Smidt et al. 2012). This elimination of competing products already increased the flux toward
AcCoA, measured as an increase in n-butanol
production
through
a
simultaneously
expressed heterologous pathway (Lian et al.
2014). Schadeweg and Boles (2016b) could
later show that a deletion of ADH isoforms 1–
5, even without a GPD deletion, substantially
increases the production of n-butanol through
a similarly designed pathway. When Ald6/
SeACS
L641P were overexpressed in an ADH-/
GPD-deficient strain, the production of nbutanol unexpectedly decreased due to a strong
accumulation of acetate (Lian et al. 2014). Conversely, overexpression of SeACS
L641P alone
improved the productivity in the same strain
background (Lian et al. 2014; Lian and Zhao
2015). Taken together, the analyses performed
in ADH positive (Shiba et al. 2007) and negative
(Lian et al. 2014; Lian and Zhao 2015) strains
suggest that the balance of ALD and ACS activity is a critical and context-dependent variable
in optimizing the flux through the cytosolic
PDH bypass due to the rate-limiting role of
ACS and the accumulation of toxic acetate
levels.
B. ATP-Independent Pyruvate-to-Acetyl-CoA
Routes
The high energy requirement makes the PDH
bypass a rather inefficient precursor supply
pathway for manufacturing AcCoA-derived
products. Therefore, several alternative ATPindependent AcCoA yielding pathways were
tested in S. cerevisiae for different products.
One possibility to convert acetaldehyde to
AcCoA is via acetylating aldehyde dehydrogenases (A-ALD). Their functionality in yeast
was demonstrated by the ability to complement
the growth defect of an ACS-deficient strain
(Kozak et al. 2014b) and to replace endogenous
acetaldehyde dehydrogenases (Kozak et al.
2016). Schadeweg and Boles (2016b) demonstrated increased n-butanol production via
reverse b-oxidation when they overexpressed
A-ALD from E. coli, mutated to favor the (nonphysiological) reaction direction from acetaldehyde to AcCoA (adhE
A267T/E568K ; MembrilloHernandez et al. 2000) in an ADH-deficient
strain. However, a positive effect of A-ALD
overexpression could only be seen when the
supply of coenzyme A (CoA) was concomitantly increased by overexpression of a heterologous pantothenate kinase (coaA from E. coli)
and pantothenate feeding, demonstrating that
the availability of not only acetyl moieties but
also of the coenzyme is a factor limiting the
synthesis of AcCoA (Schadeweg and Boles
2016b).
This notion likely applies to all AcCoA pathways, as the
overexpression of the pantothenate kinase and pantothenate supplementation was also beneficial in combination with an engineered PDH bypass (Ald6/
SeACS
L641P overexpression) in a strain constructed for
production of the flavonoid naringenin (Liu et al. 2017).
The endogenous pantothenate supply can be improved
by overexpressing the polyamine oxidase Fms1 (Schadeweg and Boles 2016a), which catalyzes the limiting
step of its biosynthesis.
A-ALD pathway is a very promising alternative to the PDH bypass to produce AcCoA
derivatives, owing to its lower energy requirement. However, the theoretically attainable
yield for FA produced by FAS is lower when
A-ALD is used compared to PDH bypass (van
Rossum et al. 2016b) due to cofactor incompatibility (the A-ALD pathway yields NADH, while
FAS requires NADPH). Thus, only if the FAs are
produced via reverse b-oxidation, which can be
engineered to utilize exclusively NADH, the
implementation of the A-ALD pathway appears
feasible. The same is true for other alternative
pathways that convert pyruvate to AcCoA in an
ATP-independent manner and thereby yield
NADH, namely, (1) pyruvate formate lyase
(PFL) combined with formate dehydrogenase,
and (2) cytosolic pyruvate dehydrogenase
complex (cPDH) (van Rossum et al. 2016b).
Although the functionality of PFL in yeast cytosol has been demonstrated (Waks and Silver
2009; Kozak et al. 2014b), it has not been used
for high-level production of FA or other AcCoA
derivatives so far, due to its complex biochemical properties, including strong oxygen sensitivity (Knappe et al. 1969). In contrast, a
significant improvement of n-butanol produc14 Engineering Saccharomyces cerevisiae for Production of Fatty Acids and Their Derivatives
343
(Smidt et al. 2012). This elimination of competing products already increased the flux toward
AcCoA, measured as an increase in n-butanol
production
through
a
simultaneously
expressed heterologous pathway (Lian et al.
2014). Schadeweg and Boles (2016b) could
later show that a deletion of ADH isoforms 1–
5, even without a GPD deletion, substantially
increases the production of n-butanol through
a similarly designed pathway. When Ald6/
SeACS
L641P were overexpressed in an ADH-/
GPD-deficient strain, the production of nbutanol unexpectedly decreased due to a strong
accumulation of acetate (Lian et al. 2014). Conversely, overexpression of SeACS
L641P alone
improved the productivity in the same strain
background (Lian et al. 2014; Lian and Zhao
2015). Taken together, the analyses performed
in ADH positive (Shiba et al. 2007) and negative
(Lian et al. 2014; Lian and Zhao 2015) strains
suggest that the balance of ALD and ACS activity is a critical and context-dependent variable
in optimizing the flux through the cytosolic
PDH bypass due to the rate-limiting role of
ACS and the accumulation of toxic acetate
levels.
B. ATP-Independent Pyruvate-to-Acetyl-CoA
Routes
The high energy requirement makes the PDH
bypass a rather inefficient precursor supply
pathway for manufacturing AcCoA-derived
products. Therefore, several alternative ATPindependent AcCoA yielding pathways were
tested in S. cerevisiae for different products.
One possibility to convert acetaldehyde to
AcCoA is via acetylating aldehyde dehydrogenases (A-ALD). Their functionality in yeast
was demonstrated by the ability to complement
the growth defect of an ACS-deficient strain
(Kozak et al. 2014b) and to replace endogenous
acetaldehyde dehydrogenases (Kozak et al.
2016). Schadeweg and Boles (2016b) demonstrated increased n-butanol production via
reverse b-oxidation when they overexpressed
A-ALD from E. coli, mutated to favor the (nonphysiological) reaction direction from acetaldehyde to AcCoA (adhE
A267T/E568K ; MembrilloHernandez et al. 2000) in an ADH-deficient
strain. However, a positive effect of A-ALD
overexpression could only be seen when the
supply of coenzyme A (CoA) was concomitantly increased by overexpression of a heterologous pantothenate kinase (coaA from E. coli)
and pantothenate feeding, demonstrating that
the availability of not only acetyl moieties but
also of the coenzyme is a factor limiting the
synthesis of AcCoA (Schadeweg and Boles
2016b).
This notion likely applies to all AcCoA pathways, as the
overexpression of the pantothenate kinase and pantothenate supplementation was also beneficial in combination with an engineered PDH bypass (Ald6/
SeACS
L641P overexpression) in a strain constructed for
production of the flavonoid naringenin (Liu et al. 2017).
The endogenous pantothenate supply can be improved
by overexpressing the polyamine oxidase Fms1 (Schadeweg and Boles 2016a), which catalyzes the limiting
step of its biosynthesis.
A-ALD pathway is a very promising alternative to the PDH bypass to produce AcCoA
derivatives, owing to its lower energy requirement. However, the theoretically attainable
yield for FA produced by FAS is lower when
A-ALD is used compared to PDH bypass (van
Rossum et al. 2016b) due to cofactor incompatibility (the A-ALD pathway yields NADH, while
FAS requires NADPH). Thus, only if the FAs are
produced via reverse b-oxidation, which can be
engineered to utilize exclusively NADH, the
implementation of the A-ALD pathway appears
feasible. The same is true for other alternative
pathways that convert pyruvate to AcCoA in an
ATP-independent manner and thereby yield
NADH, namely, (1) pyruvate formate lyase
(PFL) combined with formate dehydrogenase,
and (2) cytosolic pyruvate dehydrogenase
complex (cPDH) (van Rossum et al. 2016b).
Although the functionality of PFL in yeast cytosol has been demonstrated (Waks and Silver
2009; Kozak et al. 2014b), it has not been used
for high-level production of FA or other AcCoA
derivatives so far, due to its complex biochemical properties, including strong oxygen sensitivity (Knappe et al. 1969). In contrast, a
significant improvement of n-butanol produc14 Engineering Saccharomyces cerevisiae for Production of Fatty Acids and Their Derivatives
343