Metabolic engineering, defined as the rational engineering of microorganisms towards production goals, has greatly evolved since its conception over three decades ago. Once
applied to improve organisms to produce existing chemicals through endogenous
metabolism, it is now a promising approach also for the biosynthesis of non-natural
compounds through the expression of designed synthetic metabolic pathways. Improved
over billions of years by evolution, enzymes are however less adapted to new catalytic
functions as required by synthetic metabolism. The present work was aimed at the
construction of artificial routes for the biosynthesis of small molecules through the
application of concepts of enzyme engineering.
(L)-2,4-dihydroxybutyrate (DHB) is a non-natural compound of industrial interest for the
synthesis of methionine analogues, whose biological production was previously
demonstrated by expression of a two-step pathway via homoserine in Escherichia coli.
The pathway sequentially employs homoserine (HMS) transaminase and 2-keto-4-
hydroxybutyrate (OHB) reductase activities. Rational enzyme design was used to
improve the last catalytic step of the pathway. Simultaneous expression of the evolved
OHB reductase Ec-Mdh-5Q and HMS transaminase Ec-AlaC A142P:Y275D variant in
an engineered homoserine-overproducing E. coli strain resulted in the production of 89.0
mM DHB from glucose, the highest titer reported to date.
Of industrial interest is also the synthesis of 1,3-propanediol (PDO), a metabolite
generated from glycerol catabolism in various Clostridia species. Expanding the susbtrate
range to sugars would render PDO production more flexible. Therefore, a six-step
synthetic pathway yielding PDO from glucose via malate was conceived. While the three
first reaction steps were previously demonstrated, the remaining DHB dehydrogenase,
OHB decarboxylase and PDO oxidoreductase activities were identified in candidate
enzymes acting on sterically cognate substrates. Improved enzyme activities were
obtained by sequence- and structure-based protein design. The feasibility of the PDO
pathway was validated though expression of all required enyme activities in a single E.
coli strain, while further improvements were achieved through co-cultivation of two E.
coli strain expressing partial segments of the pathway (up to 3.8 mM PDO).
During the design and construction of the PDO pathway, OHB decarboxylases which
release 3-hydroxypropanal as product of OHB decarboxylation, were found to be
catalytically low efficient. To this end, a transcription factor-based metabolite sensor
towards high-throughput detection of aldehydes in E. coli was developed. After
Optimization tasks of the metabolite sensor through 5’-UTR engineering rendered the
sensor more sensitive to target compounds. In a proof of concept, simultaneous
expression of the metabolite sensor and a segment of the PDO pathway in E. coli afforded
the the discrimination of two OHB decarboxylases with distinct kinetic properties.
Therefore, the metabolite sensor can be implemented in directed evolution campaigns
aiming at OHB reductase development |