Repurposing an Aldolase for the Chemoenzymatic Synthesis of Substituted Quinolines

Article abstract of DOI:10.1021/acscatal.1c01398

Quinoline derivatives are important natural products and pharmaceuticals, but their synthesis can be challenging due to poor yields, harsh reaction conditions, and instability of starting materials. Here we report the chemoenzymatic synthesis of quinaldic acids under mild conditions using an aldolase, trans-o-hydroxybenzylidenepyruvate hydratase-aldolase (NahE, or HBPA). A series of 2-aminobenzaldehydes derived from reduction of the corresponding nitro analogue were reacted with pyruvate in the presence of NahE to give substituted quinolines in up to 93% isolated yield. This reaction differs from the aldol condensation catalyzed by NahE in vivo, instead resembling the heterocycle formation catalyzed by its homologue, dihydrodipicolinate synthase.

Full text of DOI:10.1021/acscatal.1c01398

pubs.acs.org/acscatalysis
Letter
Repurposing an Aldolase for the Chemoenzymatic Synthesis of
Substituted Quinolines
Douglas J. Fansher, Richard Granger, Satinderpal Kaur, and David R. J. Palmer*
Cite This: ACS Catal. 2021, 11, 6939−6943
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Quinoline derivatives are important natural products and
pharmaceuticals, but their synthesis can be challenging due to poor yields,
harsh reaction conditions, and instability of starting materials. Here we
report the chemoenzymatic synthesis of quinaldic acids under mild
conditions using an aldolase, trans-o-hydroxybenzylidenepyruvate hydra-
tase-aldolase (NahE, or HBPA). A series of 2-aminobenzaldehydes derived
from reduction of the corresponding nitro analogue were reacted with
pyruvate in the presence of NahE to give substituted quinolines in up to
93% isolated yield. This reaction differs from the aldol condensation
catalyzed by NahE in vivo, instead resembling the heterocycle formation
catalyzed by its homologue, dihydrodipicolinate synthase.
KEYWORDS: aldolase, NahE, quinoline, quinaldic acid, biocatalysis, chemoenzymatic synthesis
recently been reviewed.12 These reactions typically rely upon
harsh or toxic reagents, high temperatures, and/or long
reaction times, often accompanied by poor yields.11 There
have been some recent syntheses of quinaldic acids, typically as
the corresponding esters. Huang et al. showed that copper
triflate could catalyze a three-component reaction of aniline,
ethyl glyoxalate, and arylalkyne at room temperature (Figure
2a),13 and McNulty later showed the reaction could also be
carried out with silver triflate.14 The scope of the reaction was
limited to arylalkynes and to anilines bearing an electron-
donating group in the para position. Gabrielli et al. reported a
one-pot aza-Michael−Henry reaction using a polystyrene-
bound phosphazene base (Figure 2b), although the yields did
not exceed 64%.15 Most recently, Wang et al. reported a
photocatalytic approach to quinaldic esters in moderate to
good yields (Figure 2c).16 Again, the scope was limited to
aniline derivatives bearing an electron-donating group in the
para position reacting with styrene and closely related
compounds. Here we report that an aldolase can be used to
catalyze the synthesis of 17 different quinaldic acids in
moderate to excellent yields under mild conditions.
he quinoline ring system is found widely in both natural
Tproducts and pharmaceuticals, most famously in quinine
and related drugs.1,2 The quinoline-2-carboxylic acid (or
quinaldic acid) motif is found in the cyclic oligopeptide
antibiotic thiostrepton3 produced by some species of
streptomycetes, in the HIV-1 protease inhibitor palinavir,4 in
an inhibitor of the nucleic acid demethylase AlkB,5 and in an
anticancer ruthenium complex,6 as shown in Figure 1. Gonec
et al. have reported antibacterial and cytotoxic properties of an
array of quinoline-2-carboxamides.7 Quinolines are usually
synthesized by classical methods, including the Doebner-
8−11
̈
Miller, Friedlander, and Pfitzinger reactions,
as have
There is strong interest in making chemical processes,
including quinoline synthesis,17 more “green” by reducing toxic
catalysts and solvents: for example, by the adaptation of
Received: March 26, 2021
Revised: May 7, 2021
Published: June 1, 2021
Figure 1. HIV-1 protease inhibitor palinavir (top), an anticancer
ruthenium complex (bottom left), and a nucleic acid demethylase
inhibitor (bottom right). The quinaldic acid derived portion of each
molecule is shown in blue.
https://doi.org/10.1021/acscatal.1c01398
© 2021 American Chemical Society
ACS Catal. 2021, 11, 6939−6943
6939

ACS Catalysis
pubs.acs.org/acscatalysis
Letter
replaced by 3-fluoropyruvate, the resulting aldol product does
not dehydrate, resulting in the generation of two stereo-
centers.25 Very recently Moreno et al. observed that non-
aromatic aldehydes can act as substrates which also form
products that do not undergo dehydration.28 No authors have
observed formation of the cis isomer of the product by NahE.
Although NahE is reported to tolerate substitution on the
aromatic ring of benzaldehyde, 2-aminobenzaldehyde (2-AB)
is notably absent from these reports, presumably due to its low
stability.24,25,27
We envisioned using NahE from P. putida to generate cyclic
products as a novel route to substituted quinolines using 2-AB
and related compounds as starting materials, as shown in
Figure 2d. Such a reaction would likely require the enzyme to
catalyze the transimination reaction in the active site, so that
the cyclic product would form rather than the “normal” 2-
aminobenzylidenepyruvate product. 2-AB derivatives readily
undergo self-condensation and thus, if commercially available,
tend to be expensive. Indeed, we found that 2-AB polymerizes
to an insoluble yellow-brown gum when it is added directly to
aqueous buffer. We avoided this drawback by working with 2-
nitrobenzaldehyde, which we reduced with Fe0 in aqueous
ethanol and catalytic HCl.29 Although TLC indicated complete
conversion, isolation of the resulting aminoaldehyde resulted in
considerable degradation. We therefore eliminated this
isolation by adding DMSO after the workup but before
removal of the solvent and using the solution of 2-AB in
DMSO directly in the subsequent enzymatic reaction as
described in the Supporting Information.
NahE bearing an N-terminal hexahistidine tag was purified
from Escherichia coli cells transformed with an expression
vector bearing the P. putida nahE gene as described in the
Supporting Information. We found that stirring 2-hydrox-
ybenzaldehyde with a 3-fold excess of pyruvate and NahE in
phosphate buffer at pH 6.5 overnight at room temperature
gave the expected trans-o-hydroxybenzylidenepyruvate in 98%
isolated yield. We therefore adapted these conditions to our
reaction of interest. Briefly, 2-AB in DMSO was added
dropwise over 5−10 min to a gently stirred solution of NahE
and pyruvate in potassium phosphate buffer (pH 6.5), so that
the final conditions had a 3-fold excess of pyruvate, 0.02 mol %
of NahE, and ≤10% v/v DMSO. After gentle shaking for 16 h
at room temperature, the reaction mixture was acidified and
extracted with ethyl acetate to give quinaldic acid in 90% yield.
Reactions performed without NahE or with heat-inactivated
NahE resulted in no detectable formation of quinoline.
Short reaction times resulted in the presence of a second
product in addition to quinaldic acid. We had anticipated that
NahE might produce different products, as depicted in Figure
4. The aldol product (A) might be released from the enzyme
by hydrolysis and then cyclize and dehydrate nonenzymatically
to form the quinaldic acid, while the “normal” aldol
condensation product, trans-2-aminobenzylidenepyruvate (B),
might form reversibly. Monitoring the reaction progress using
NMR and HPLC with UV−visible detection showed that only
one additional product could be observed, which could be
identified as B. The 1H NMR spectrum of this product shows a
pair of doublets at 6.77 and 7.85 ppm that share a large
coupling constant (J = 16 Hz), consistent with a trans-alkene
(Figure S2). HPLC chromatograms show that this transient
product has a strong UV absorbance with a maximum at 296
nm, consistent with extended conjugation; trans-2-hydrox-
ybenzylidenepyruvate shows a very similar spectrum (Figures
Figure 2. Recent quinaldic acid forming reactions: (a) tandem
synthesis of quinaldic esters;13 (b) aza-Michael−Henry reaction;15 (c)
aerobic oxidative dehydrogenative coupling;16 (d) the NahE-catalyzed
reaction reported here.
enzymes to catalyze reactions under mild aqueous conditions
on a practical scale. There are many enzymes that catalyze the
formation of N-heterocycles. Our laboratory has studied a type
1 (Schiff base dependent) aldolase, dihydrodipicolinate
synthase (DHDPS), that catalyzes the condensation of
pyruvate with a β-aminoaldehyde, (S)-aspartate-β-semialde-
hyde.18,19 The conventional aldol carbon−carbon bond
forming step is followed by a transimination in the active
site to release a heterocyclic product (Figure 3a and Figure
Figure 3. Reactions of the homologous aldolases DHDPS (a) and
NahE (b). The NahE-catalyzed reaction is shown in the aldol
direction, although its metabolic function is the retro-aldol reaction.
S1). DHDPS is a member of the N-acetylneuraminate lyase
(NAL) subfamily of aldolases, which includes at least 13
functionally distinct members.20,21 Several members of this
family have been characterized and share a similar protein fold
and active site architecture. DHDPS is the only member of this
family known to have evolved to catalyze heterocycle
formation.
One member of the NAL family that has received attention
recently is trans-o-hydroxybenzylidenepyruvate hydratase-
aldolase, also called NahE.22−27 NahE catalyzes the reversible
hydration (i.e., oxa-Michael addition of water) and retro-aldol
reaction of trans-o-hydroxybenzylidenepyruvate (see Figure
3b), a step in bacterial naphthalene degradation. Like many
aldolases, it is selective for the first substrate in the bond-
forming direction, pyruvate, and much less so for the second
substrate, 2-hydroxybenzaldehyde. The high-resolution struc-
ture of NahE from Pseudomonas putida has recently been
reported,26 and Howard et al. have shown that, if pyruvate is
6940
https://doi.org/10.1021/acscatal.1c01398
ACS Catal. 2021, 11, 6939−6943

ACS Catalysis
pubs.acs.org/acscatalysis
Letter
Figure 4. Proposed reaction pathway of NahE-catalyzed formation of
quinaldic acid, showing the covalent enzyme-bound aldol inter-
mediate. The aldol product A is not observed by NMR nor HPLC,
whereas the appearance and disappearance of B are observed.
S3 and S4). By quenching the reaction after 5 s, we could
observe in the NMR spectrum that quinaldic acid formation
was already apparent (Figure S5), and within minutes both
quinaldic acid and B increased as 2-AB was consumed, but
quinaldic acid concentration increased much more quickly,
indicating that this is the preferred pathway. The proportion of
B in solution relative to 2-AB and quinaldic acid did not
exceed 16% (Figure S6), and B was gradually converted to
quinaldic acid, consistent with the reversible formation of B
and essentially irreversible accumulation of quinaldic acid. No
evidence of aldol product A was observed. These results are
consistent with cyclization taking place in the active site, as in
the DHDPS mechanism. The presence of a trans double bond
precludes cyclization; therefore, transimination must take place
prior to dehydration (Figure S1).
The substrate scope of the reaction was investigated using a
series of substituted 2-AB derivatives, as shown in Figure 5.
Several conclusions can be drawn from the results. First, the
reaction tolerates a substituent in any position on the ring,
allowing synthesis of 5-, 6-, 7-, and 8-substituted quinaldic
acids, as shown for the chloro-substituted series 1a−d, which
are produced in comparable yield. The reaction also
accommodates fluorine and bromine substitution (1e−h). 2-
Amino-4-cyanobenzaldehyde was converted to the 7-cyano-
quinaldic acid (1i) in excellent yield, but substitution with a
methoxycarbonyl group in the same position (1j) resulted in a
notably lower yield. Both of these substituents are electron-
withdrawing, suggesting that steric effects may account for this
difference. Examination of the effects of electron-donating
substituents (1k−n) suggests that lower yields are obtained
when such groups are para to the aldehyde of the substrate
(1k,n). Although the 6-hydroxyquinaldic acid could be
obtained in 70% yield, the reduction of the starting material
was difficult to control, making the process less reliable. The
presence of more than one additional substituent (1o,p)
resulted in poor yields. We could not isolate any product using
the much bulkier 2-amino-2-formylchromone (1q).
Figure 5. Substrate scope of the chemoenzymatic synthesis of
substituted quinaldic acids. Average isolated yields of two or more
reactions starting from the corresponding 2-nitrobenzaldehyde are
shown, except for (a) yield of the isolated methyl ester and (b) yield
of the reaction from the 2-aminobenzaldehyde. (c) n.d. = no product
detected.
Using 1.8 g (11 mmol) of 5-fluoro-2-nitrobenzaldehyde
resulted in the isolation of 1.35 g of 1e (68%).
In summary, we have shown that the aldolase NahE can be
used to generate the quinoline ring system to form substituted
quinaldic acids under mild conditions and in high yields. To
our knowledge this is the first reported use of biocatalysis to
form quinolines.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acscatal.1c01398.
■
sı
Experimental methods, NahE sequence and accession
number, supporting figures, and chemical character-
ization (PDF)
AUTHOR INFORMATION
Corresponding Author
■
NahE is very robust under these conditions. Extending the
reaction time from 16 to 48 h improved the yield from poorer
substrates such as 1p from 11% to 26% and from 1j from 57%
to 89%. The process can also be carried out on a gram scale.
David R. J. Palmer − Department of Chemistry, University of
Saskatchewan, Saskatoon, Saskatchewan, Canada S7N 5C9;
orcid.org/0000-0001-9300-7460; Email: dave.palmer@
usask.ca
6941
https://doi.org/10.1021/acscatal.1c01398
ACS Catal. 2021, 11, 6939−6943

ACS Catalysis
pubs.acs.org/acscatalysis
Letter
(12) Weyesa, A.; Mulugeta, E. Recent advances in the synthesis of
biologically and pharmaceutically active quinoline and its analogues: a
review. RSC Adv. 2020, 10 (35), 20784−20793.
(13) Huang, H.; Jiang, H.; Chen, K.; Liu, H. A Simple and
Convenient Copper-Catalyzed Tandem Synthesis of Quinoline-2-
carboxylates at Room Temperature. J. Org. Chem. 2009, 74 (15),
5476−5480.
(14) McNulty, J.; Vemula, R.; Bordón, C.; Yolken, R.; Jones-Brando,
L. Synthesis and anti-toxoplasmosis activity of 4-arylquinoline-2-
carboxylate derivatives. Org. Biomol. Chem. 2014, 12 (2), 255−260.
(15) Gabrielli, S.; Giardinieri, A.; Sampaolesi, S.; Ballini, R.; Palmieri,
A. A New One-Pot Synthesis of Quinoline-2-carboxylates under
Heterogeneous Conditions. Molecules 2016, 21 (6), 776.
(16) Wang, J.; Li, L.; Guo, Y.; Li, S.; Wang, S.; Li, Y.; Zhang, Y.
Visible-light-enabled aerobic oxidative Csp3-H functionalization of
glycine derivatives using an organic photocatalyst: access to
substituted quinoline-2-carboxylates. Org. Biomol. Chem. 2020, 18
(40), 8179−8185.
(17) Nainwal, L. M.; Tasneem, S.; Akhtar, W.; Verma, G.; Khan, M.
F.; Parvez, S.; Shaquiquzzaman, M.; Akhter, M.; Alam, M. M. Green
recipes to quinoline: A review. Eur. J. Med. Chem. 2019, 164, 121−
170.
(18) Phenix, C. R.; Nienaber, K.; Tam, P. H.; Delbaere, L. T. J.;
Palmer, D. R. J. Structural, functional and calorimetric investigation of
MosA, a dihydrodipicolinate synthase from Sinorhizobium meliloti
L5−30, does not support involvement in rhizopine biosynthesis.
ChemBioChem 2008, 9 (10), 1591−1602.
(19) Skovpen, Y. V.; Palmer, D. R. J. Dihydrodipicolinate Synthase
from Campylobacter jejuni: Kinetic Mechanism of Cooperative
Allosteric Inhibition and Inhibitor-Induced Substrate Cooperativity.
Biochemistry 2013, 52 (32), 5454−5462.
(20) Watanabe, S.; Watanabe, Y.; Nobuchi, R.; Ono, A. Biochemical
and Structural Characterization of l-2-Keto-3-deoxyarabinonate
Dehydratase: A Unique Catalytic Mechanism in the Class I Aldolase
Protein Superfamily. Biochemistry 2020, 59 (32), 2962−2973.
(21) Lawrence, M. C.; Barbosa, J. A. R. G.; Smith, B. J.; Hall, N. E.;
Pilling, P. A.; Ooi, H. C.; Marcuccio, S. M. Structure and mechanism
of a sub-family of enzymes related to N-acetylneuraminate lyase. J.
Mol. Biol. 1997, 266 (2), 381−399.
Authors
Douglas J. Fansher − Department of Chemistry, University of
Saskatchewan, Saskatoon, Saskatchewan, Canada S7N 5C9
Richard Granger − Department of Chemistry, University of
Saskatchewan, Saskatoon, Saskatchewan, Canada S7N 5C9
Satinderpal Kaur − Department of Chemistry, University of
Saskatchewan, Saskatoon, Saskatchewan, Canada S7N 5C9
Complete contact information is available at:
https://pubs.acs.org/10.1021/acscatal.1c01398
Funding
This work was funded by a NSERC Discovery Grant to
D.R.J.P. and a University of Saskatchewan Dean’s Scholarship
to D.J.F.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors thank K. Fransishyn, P. Zhu, and other staff of the
Saskatchewan Structural Sciences Centre for their expertise
and D. Sanders for helpful discussions and shared equipment.
REFERENCES
■
(1) Shang, X.-F.; Morris-Natschke, S. L.; Liu, Y.-Q.; Guo, X.; Xu, X.-
S.; Goto, M.; Li, J.-C.; Yang, G.-Z.; Lee, K.-H. Biologically active
quinoline and quinazoline alkaloids part I. Med. Res. Rev. 2018, 38 (3),
775−828.
(2) Shang, X.-F.; Morris-Natschke, S. L.; Yang, G.-Z.; Liu, Y.-Q.;
Guo, X.; Xu, X.-S.; Goto, M.; Li, J.-C.; Zhang, J.-Y.; Lee, K.-H.
Biologically active quinoline and quinazoline alkaloids part II. Med.
Res. Rev. 2018, 38 (5), 1614−1660.
(3) Priestley, N. D.; Smith, T. M.; Shipley, P. R.; Floss, H. G. Studies
on the biosynthesis of thiostrepton: 4-(1-hydroxyethyl)quinoline-2-
carboxylate as a free intermediate on the pathway to the quinaldic acid
moiety. Bioorg. Med. Chem. 1996, 4 (7), 1135−1147.
(4) Lamarre, D.; Croteau, G.; Wardrop, E.; Bourgon, L.; Thibeault,
D.; Clouette, C.; Vaillancourt, M.; Cohen, E.; Pargellis, C.; Yoakim,
C.; Anderson, P. C. Antiviral properties of palinavir, a potent inhibitor
of the human immunodeficiency virus type 1 protease. Antimicrob.
Agents Chemother. 1997, 41 (5), 965.
(5) Woon, E. C. Y.; Demetriades, M.; Bagg, E. A. L.; Aik, W.;
Krylova, S. M.; Ma, J. H. Y.; Chan, M.; Walport, L. J.; Wegman, D.
W.; Dack, K. N.; McDonough, M. A.; Krylov, S. N.; Schofield, C. J.
Dynamic Combinatorial Mass Spectrometry Leads to Inhibitors of a
2-Oxoglutarate-Dependent Nucleic Acid Demethylase. J. Med. Chem.
2012, 55 (5), 2173−2184.
(6) Camm, K. D.; El-Sokkary, A.; Gott, A. L.; Stockley, P. G.;
Belyaeva, T.; McGowan, P. C. Synthesis, molecular structure and
evaluation of new organometallic ruthenium anticancer agents. Dalton
Transactions 2009, No. 48, 10914−10925.
(7) Gonec, T.; Bobal, P.; Sujan, J.; Pesko, M.; Guo, J.; Kralova, K.;
Pavlacka, L.; Vesely, L.; Kreckova, E.; Kos, J.; Coffey, A.; Kollar, P.;
Imramovsky, A.; Placek, L.; Jampilek, J. Investigating the Spectrum of
Biological Activity of Substituted Quinoline-2-Carboxamides and
Their Isosteres. Molecules 2012, 17, 613−644.
(8) Marco-Contelles, J.; Pérez-Mayoral, E.; Samadi, A.; Carreiras, M.
d. C.; Soriano, E. Recent Advances in the Friedlander Reaction. Chem.
Rev. 2009, 109 (6), 2652−2671.
(22) Eaton, R. W. trans-o-Hydroxybenzylidenepyruvate Hydratase-
Aldolase as a Biocatalyst. Appl. Environ. Microbiol. 2000, 66 (6), 2668.
(23) Ferrara, S.; Mapelli, E.; Sello, G.; Di Gennaro, P. Character-
ization of the aldol condensation activity of the trans-o-hydrox-
ybenzylidenepyruvate hydratase-aldolase (tHBP-HA) cloned from
Pseudomonas fluorescens N3. Biochim. Biophys. Acta, Proteins
Proteomics 2011, 1814 (5), 622−629.
(24) Sello, G.; Di Gennaro, P. Aldol Reactions of the trans-o-
Hydroxybenzylidenepyruvate Hydratase-Aldolase (tHBP-HA) from
Pseudomonas fluorescens N3. Appl. Biochem. Biotechnol. 2013, 170
(7), 1702−1712.
(25) Howard, J. K.; Muller, M.; Berry, A.; Nelson, A. An Enantio-
̈
and Diastereoselective Chemoenzymatic Synthesis of α-Fluoro β-
Hydroxy Carboxylic Esters. Angew. Chem., Int. Ed. 2016, 55 (23),
6767−6770.
(26) LeVieux, J. A.; Medellin, B.; Johnson, W. H.; Erwin, K.; Li, W.;
Johnson, I. A.; Zhang, Y. J.; Whitman, C. P. Structural Character-
ization of the Hydratase-Aldolases, NahE and PhdJ: Implications for
the Specificity, Catalysis, and N-Acetylneuraminate Lyase Subgroup
of the Aldolase Superfamily. Biochemistry 2018, 57 (25), 3524−3536.
(27) Suzuki, T.; Takizawa, N. Purification and enzymatic character-
ization of trans-o-hydroxybenzylidenepyruvate hydratase-aldolase
from Rhodococcus opacus and enzymatic formation of α, β-
unsaturated ketones. Biosci., Biotechnol., Biochem. 2019, 83 (10),
1884−1888.
̈
(9) Sloop, J. C. Quinoline formation via a modified Combes
reaction: examination of kinetics, substituent effects, and mechanistic
pathways. J. Phys. Org. Chem. 2009, 22 (2), 110−117.
(10) Shvekhgeimer, M. G. A. The Pfitzinger Reaction. (Review).
Chem. Heterocycl. Compd. 2004, 40 (3), 257−294.
(11) Denmark, S. E.; Venkatraman, S. On the Mechanism of the
Skraup-Doebner-Von Miller Quinoline Synthesis. J. Org. Chem. 2006,
71 (4), 1668−1676.
(28) Moreno, C. J.; Hernández, K.; Charnok, S. J.; Gittings, S.;
Bolte, M.; Joglar, J.; Bujons, J.; Parella, T.; Clapés, P. Synthesis of γ-
Hydroxy-α-amino Acid Derivatives by Enzymatic Tandem Aldol
Addition-Transamination Reactions. ACS Catal. 2021, 11 (8), 4660−
4669.
6942
https://doi.org/10.1021/acscatal.1c01398
ACS Catal. 2021, 11, 6939−6943

ACS Catalysis
pubs.acs.org/acscatalysis
Letter
(29) Diedrich, C. L.; Haase, D.; Saak, W.; Christoffers, J.
̈
Regioselectivity of Friedlander Quinoline Syntheses. Eur. J. Org.
Chem. 2008, 2008 (10), 1811−1816.
6943
https://doi.org/10.1021/acscatal.1c01398
ACS Catal. 2021, 11, 6939−6943