Angewandte
Chemie
CO2 Fixation
Screening and Engineering the Synthetic Potential of Carboxylating
Reductases from Central Metabolism and Polyketide Biosynthesis
Dominik M. Peter, Lennart Schada von Borzyskowski, Patrick Kiefer, Philipp Christen,
Julia A. Vorholt, and Tobias J. Erb*
Abstract: Carboxylating enoyl-thioester reductases (ECRs)
are a recently discovered class of enzymes. They catalyze the
highly efficient addition of CO2 to the double bond of a,b-
unsaturated CoA-thioesters and serve two biological functions.
In primary metabolism of many bacteria they produce ethyl-
malonyl-CoA during assimilation of the central metabolite
acetyl-CoA. In secondary metabolism they provide distinct a-
carboxyl-acyl-thioesters to vary the backbone of numerous
polyketide natural products. Different ECRs were systemati-
cally assessed with a diverse library of potential substrates. We
identified three active site residues that distinguish ECRs
restricted to C4 and C5-enoyl-CoAs from highly promiscuous
ECRs and successfully engineered a selected ECR as proof-of-
principle. This study defines the molecular basis of ECR
reactivity, allowing for predicting and manipulating a key
reaction in natural product diversification.
longer chain derivatives,[10] as well as isobutylmalonyl-CoA[7b]
and other branched-chain analogues.[11]
These non-traditional extender units are provided by the
reductive carboxylation of a,b-unsaturated acyl-CoA thio-
esters by carboxylating enoyl-thioester reductases (ECR),
a novel class of enzymes that was described only recently. The
prime example in the ECR family is crotonyl-CoA carbox-
ylase/reductase (Ccr) that catalyzes the NADPH-dependent
carboxylation of crotonyl-CoA into ethylmalonyl-CoA and
that is one of the most efficient CO2-fixing enzymes described
to date.[12] As of 2014, the ECR family features more than 900
homologues that can be physiologically divided into two
subfamilies, a large ECR subfamily of so-called primary
metabolism Ccrs (ECR-1) that function in the ethylmalonyl-
CoA pathway, a recently discovered central metabolic path-
way for acetyl-CoA assimilation,[13] and a subfamily of
secondary metabolism ECRs (ECR-2) that are associated
with polyketide biosynthesis.[6]
Given the fact that ECRs are key enzymes to alter the
polyketide backbone, surprisingly little is known on the
structure and catalytic mechanism of these proteins.[14] Yet
such information is indispensable 1) to understand what
factors direct and control the incorporation of non-traditional
extender units into the growing polyketide chain, 2) to assign
the biosynthetic function of ECR homologues for correctly
predicting the polyketide product structure, and 3) to manip-
ulate ECR reactivity to rationally engineer polyketide
biosynthesis.
Motivated by the above, we sought to investigate the
molecular basis for substrate specificity of ECRs in more
detail. We first established a diverse substrate library of
enoyl-CoA thioesters to cover the natural and non-natural
chemical space of polyketide extender units (1–19, Figure 1).
This library was used to test eight phylogenetically diverse
ECRs. From the ECR-1 subfamily that is supposedly specific
for crotonyl-CoA (1) as substrate, we included four homo-
logues: CcrCc from Caulobacter crescentus, CcrSg from
Streptomyces griseus, CcrSb from Streptomyces bottropensis,
and CcrPd from Paracoccus denitrificans. From the ECR-2
subfamily, we included CinF from Streptomyces sp. JS360,
RevT from Streptomyces sp. SN-593, SalG from Salinispora
tropica, and EcrSh from Streptomyces hygroscopicus that
were reported to catalyze the reductive carboxylation of
octenoyl-CoA (CinF),[14a] hexenoyl-CoA (RevT)[11a] and
chlorocrotonyl-CoA, as well as pentenoyl-CoA, repectively
(SalG).[8] Above candidates were heterologously expressed in
Escherichia coli BL21 cells and purified to homogeneity for
individual screens of above substrate library in a HPLC-MS
based in vitro assay (Supporting Information, Figure S1).
T
he biosynthetic potential of nature is impressively reflected
in the diversity of secondary metabolism. More than 325000
natural products have been described to date.[1] All of these
compounds differ strongly with respect to chemical structure
and biological activity,[2] yet the biosynthesis of their struc-
tural backbone is based on simple elongation reactions from
basic building blocks, the so-called extender units.[3]
A
prominent example is the large class of polyketides that are
assembled through subsequent Claisen condensation reac-
tions from a-carboxylacylthioester units.[4] The standard
building blocks in polyketide assembly lines are malonyl-
coenzyme A (CoA) and methylmalonyl-CoA, which are
mainly provided by enzymes of fatty acid metabolism that
a-carboxylate acetyl-CoA and propionyl-CoA, respectively.[5]
However, an increasing number of polyketides show varia-
tions from this common principle, because their side-chains
suggest non-standard extender units to vary the structural
backbone.[6] Additional extender units include ethylmalonyl-
CoA[7] chloroethylmalonyl-CoA,[8] propylmalonyl-CoA[9] and
[*] D. M. Peter, L. Schada von Borzyskowski, Dr. T. J. Erb
Biochemistry and Synthetic Biology of Microbial Metabolism Group
Max-Planck-Institute for terrestrial Microbiology
Karl-von-Frisch-Strasse 10, 35043 Marburg (Germany)
E-mail: toerb@mpi-marburg.mpg.de
D. M. Peter, L. Schada von Borzyskowski, Dr. P. Kiefer, P. Christen,
Prof. J. A. Vorholt, Dr. T. J. Erb
Institute of Microbiology
Eidgençssisch Technische Hochschule (ETH) Zürich
Vladimir-Prelog-Weg 4, 8050 Zürich (Switzerland)
Supporting information and ORCID(s) from the author(s) for this
Angew. Chem. Int. Ed. 2015, 54, 13457 –13461
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
13457