Bi-enzymatic Conversion of Cinnamic Acids to 2-Arylethylamines

Article abstract of DOI:10.1002/cctc.201902128

The conversion of carboxylic acids, such as acrylic acids, to amines is a transformation that remains challenging in synthetic organic chemistry. Despite the ubiquity of similar moieties in natural metabolic pathways, biocatalytic routes seem to have been overlooked for this purpose. Herein we present the conception and optimisation of a two-enzyme system, allowing the synthesis of β-phenylethylamine derivatives from readily-available ring-substituted cinnamic acids. After characterisation of both parts of the reaction in a two-step approach, a set of conditions allowing the one-pot biotransformation was optimised. This combination of a reversible deaminating and irreversible decarboxylating enzyme, both specific for the amino acid intermediate in tandem, represents a general method by which new strategies for the conversion of carboxylic acids to amines could be designed.

Full text of DOI:10.1002/cctc.201902128

Accepted Article
Title: Bi-enzymatic Conversion of Cinnamic Acids to 2-Arylethylamines
Authors: Nicholas Weise, Prasansa Thapa, Syed Ahmed, Rachel
Heath, Fabio Parmeggiani, Nicholas Turner, and Sabine
Flitsch
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201902128
Link to VoR: http://dx.doi.org/10.1002/cctc.201902128
A Journal of
www.chemcatchem.org

10.1002/cctc.201902128
ChemCatChem
COMMUNICATION
Bi-enzymatic Conversion of Cinnamic Acids to 2-Arylethylamines
Nicholas J. Weise,[a] Prasansa Thapa,[a] Syed T. Ahmed,[a] Rachel S. Heath,[a] Fabio Parmeggiani,[a]
Nicholas J. Turner,[a] and Sabine L. Flitsch*[a]
Abstract: The conversion of carboxylic acids, such as acrylic acids,
to amines is a transformation that remains challenging in synthetic
organic chemistry. Despite the ubiquity of similar moieties in natural
metabolic pathways, biocatalytic routes seem to have been
overlooked for this purpose. Herein we present the conception and
optimisation of a two-enzyme system, allowing the synthesis of β-
phenylethylamine derivatives from readily-available ring-substituted
cinnamic acids. After characterisation of both parts of the reaction in
a two-step approach, a set of conditions allowing the one-pot
biotransformation was optimised. This combination of a reversible
deaminating and irreversible decarboxylating enzyme, both specific
for the amino acid intermediate in tandem, represents a general
method by which new strategies for the conversion of carboxylic
acids to amines could be designed.
acid as the first step in the process, followed by an irreversible
release of CO2. Cinnamic acid derivatives were chosen due to
the additional alkene moiety conjugated to the carboxyl group to
be converted, thus allowing additional and orthogonal
transformations in a stepwise fashion. It was reasoned that,
despite the potentially unfavourable equilibrium of the initial
enzymatic step, the subsequent decarboxylation would serve to
remove the amination product and thus drive overall conversion,
much in the same way as has been shown with the use of
aldolases and oxidases to produce a variety of amino alcohols.[5]
To this end, a combination of phenylalanine ammonia lyase
(PAL) and an aromatic amino acid decarboxylase (AADC) was
considered (Scheme 1). This was due to the widely reported use
of ammonia lyase enzymes for synthetic purposes[6–10] and the
decarboxylation chemistry of AADCs uniquely enabled by the α-
amino moiety,[5,11] avoiding possible siphoning of the PAL
carboxylic acid substrate into a decarboxylated by-product.
Owing to the use of truly catalytic chemical groups by both
enzymes - 4-methylideneimidazole-5-one (MIO) for PAL[8,9,12]
Research into methods allowing the interconversion of simple
functional groups has been
a focus in synthetic organic
chemistry over the last century. Despite sustained efforts in this
area, there are still a number of such transformations which
feature unstable intermediates, lengthy multi-step syntheses and
multiple purification steps, resulting in elevated energy
consumption, waste generation and solvent use. Examples of
this are heteroatom migrations (including the Curtius
rearrangement and modifications of Hofmann, Lossen and
Schmidt degradations), which remain some of the most facile
methods to produce primary amines from a carboxylic acid
starting material (Scheme 1).[1] In spite of the utility of such
approaches, chemical routes to these high value compounds
require multi-step synthesis and several isolation steps. The
harsh conditions required for many of the latter steps may prove
incompatible with other moieties in the starting material and the
and pyridoxal-5’-phosphate (PLP) for AADC[5,11]
- it was
predicted that no additional supplementations or recycling
systems would be required, with each holoenzyme being
regenerated during their respective catalytic cycles.
need
for
chromatographic
purification
steps
using
environmentally-damaging solvents is usually undesirable.
Furthermore, the use of potentially explosive and highly toxic
azide derivatives (in Curtius-type procedures) is of particular
concern, mainly in an industrial setting.
Biocatalytic strategies have gained much attention in the
areas of synthetic chemistry and pharmaceutical science, due to
the ever increasing demand for concise, sustainable and mild
approaches to chemical synthesis. The ability to combine
renewably-sourced and highly selective biological catalysts
under ambient and compatible conditions, without the need for
specialised work up of intermediates, presents great advantages
over current synthetic methods.[2–4] The range of enzymes
available in nature and near infinite combination that could be
imagined from these poses the possibility of designing new
strategies for classical chemical transformations, complementary
to established synthetic methods and even to other enzymatic
routes.
Scheme 1. Methods allowing the conversion of 2-arylacrylic acids to 2-
arylethylamines.
Substituted β-arylethylamines 3 are key building blocks of
pharmaceuticals and bioactive molecules, and the parent
compound β-phenylethylamine 3a is known to function as a
neurotransmitter and antidepressant.[11] There is evidence in the
literature of the conversion of various cinnamic acids 1 to the
corresponding amines 3, by a two-step procedure based on
hydrogenation followed by Curtius or Hofmann rearrangements,
e.g. in the synthesis of peptidomimetic probes and small
molecule pharmaceutical candidates.[13–16] For example,
By performing a retrosynthetic analysis, a two-enzyme
pathway was envisaged to obtain β-phenylethylamine
derivatives from easily accessible cinnamic acids. This involved
placing the amination of the candidate unsaturated carboxylic
conversion of 3-bromocinnamic acid required
a five step
This article is protected by copyright. All rights reserved.

10.1002/cctc.201902128
ChemCatChem
COMMUNICATION
synthesis with an overall yield of 24%, a method that would still
require a de-protection step to afford the free amine.[14] To
complement existing strategies, a biocatalytic route requiring
mild aqueous conditions was sought, using two enzymes: the
cyanobacterial phenylalanine ammonia lyase AvPAL from
substrates with ortho-groups larger than a fluorine atom gave
extremely low conversions (<10%) with EfPheDC. Increasing the
size of substituents at the 4-position had a less dramatic but still
noticeable effect, with the para-bromo-, para-methyl- and para-
chloro- giving moderate conversions. For the nitro substituents,
low conversions were seen all around, but with a preference for
the 3-substituents, as seen with chloro- and bromoarene
isomers. To add evidence to the assertion that the
Anabaena
variabilis[8,9,17] and
a
decarboxylase
from
Enterococcus faecium, referred to from hereon as EfPheDC.[11]
Production of both enzymes as either wet (EfPheDC) or dry
(AvPAL) whole cell biocatalysts allowed simple investigation of
their pH optima. The enzymes were shown to operate under
very different conditions in this respect, as consistent with
previous studies of PAL enzymes[18,19] and the potential
involvement of AADCs with bacterial acid stress responses.[20]
As such, initial experiments focussed on the development of a
sequential reaction whereby cells, water and ammonium
carbamate (required for PAL-mediated amination) were removed
via centrifugation and evaporation, leaving a crude isolate[21] for
conversion by EfPheDC under separate conditions. Following
initial test with temperatures ranging from 20 to 60°C, the
optimum incubation was found to be at 30°C (see Supporting
Information). Using this method, a panel of 15 cinnamic acids
1a-o known to be good PAL substrates was tested (Table 1).
decarboxylase displayed arene substitution preference,
a
smaller panel of various phenylalanine derivatives was tested
with this enzyme, revealing low conversions for D-phenylalanine,
rac-2-bromo- and L-4-nitro- compounds (see Supporting
Information). However, investigations of the enantioselectivity of
AvPAL showed low enantiomeric excess values only for the
nitrophenylalanines, as observed previously.[8,9] This revealed
that, although EfPheDC displayed a clear preference for L-amino
acids, production of racemic reaction intermediates by AvPAL
was unlikely to be the reason for low overall conversion in most
cases (see Supporting Information).
Having demonstrated the effectiveness of AvPAL-EfPheDC
tandem reactions for the synthesis of various 2-arylethylamines,
the compatibility of the two steps for construction of
a
biocatalytic cascade was investigated. It was hypothesised that
the catalytic activity of the PLP-dependent decarboxylase may
be affected detrimentally by the high concentrations of ammonia
associated with the initial amination step. Biotransformations of
L-phenylalanine in ammonium chloride buffer revealed this to be
the case, with conversions of ~20% only observed at ammonia
concentrations 8-fold to 20-fold lower than commonly required
for PAL reactions (see Supporting Information). Interestingly, the
use of the diluted reaction buffer in the first cascade attempt
gave just 14% conversion by AvPAL and no traces of β-
phenylethylamine, possibly due to the short-lived activity of
EfPheDC and slow amination rate under these conditions. In an
attempt to rebalance the ratio of cinnamate to ammonia as the
primary driving force for the initial amination reaction, various
substrate concentrations were tested, and an increasing amount
of β-phenylethylamine could be observed with decreasing
substrate concentration (see Supporting Information).
In spite of the improvements seen (conversion up to 34% of
2 mM starting material after 22 h) by increasing the rate of
amination, the focus was shifted towards further optimisations
designed to increase the efficiency of the decarboxylation step.
It was reasoned that this would increase removal rate of the
intermediate amino acid and increase the rate of the PAL
reaction to maintain the equilibrium. Initially, changing the
ammonia concentration for the cascade reaction to 1000 mM (to
favour the first step) and 250 mM (to favour the second step)
both resulted in lower levels of the amine (Table 2, entries 1-3).
As such, 500 mM was used as a compromise with the pH either
increased or decreased to favour each step in turn (Table 2,
entries 4-6). This resulted in a slight improvement at the more
acidic pH and extremely poor yield of 3a at the more basic pH.
As this reinforced the idea that the activity of the decarboxylase
was the bottleneck of the cascade, the loading of EfPheDC-
containing whole cells was varied. As expected, there was a
positive correlation between the amount of catalyst and the
effective siphoning of the PAL-product to amine (Table 2, entries
7-10). This also resulted in greater overall conversion with
Table 1. Telescopic amination-decarboxylation of various cinnamic acids to
yield the corresponding 2-arylethylamines as catalysed by AvPAL (step 1) and
EfPheDC (step 2).[a]
______________________________________________________________
Conv.
(1 to 2)[b]
[%]
Conv.
(2 to 3)[b]
[%]
Overall
conv.[b]
[%]
Subs.
1a
R
H
99
93
92
1b
1c
1d
1e
1f
o-Br
m-Br
p-Br
o-Cl
m-Cl
p-Cl
o-F
98
94
94
97
93
95
>99
92
96
2
2
81
37
7
75
34
7
96
48
98
87
89
86
45
98
78
84
1g
1h
1i
m-F
p-F
1j
1k
1l
o-Me
85
73
99
98
91
2
<1
35
<1
21
2
p-Me
51
<1
22
2
1m
1n
1o
o-NO2
m-NO2
p-NO2
[a] Reaction conditions: step 1. 32 mM 1. 10 mg mL-1 AvPAL E. coli dry
cells, 4 M ammonium carbamate, pH 9.9, 30°C, 250 rpm, 22h; step 2. 10
mg mL-1 EfPheDC E. coli wet cells, 100 mM sodium citrate pH 4.4, 30°C,
250 rpm, 22 h. [b] Determined by reverse phase HPLC analysis on a non-
chiral phase.
The decarboxylation step was found to give variable
conversions, with the unsubstituted, fluoro- and 3-chloro/bromo-
compounds giving the highest values. Interestingly, any
This article is protected by copyright. All rights reserved.

10.1002/cctc.201902128
ChemCatChem
COMMUNICATION
increasing decarboxylase loading, as high as 88% with 40 mg
mL–1 cells, but with the relative ratio of 1a to 2a remaining
constant. This proves that, as the thermodynamic equilibrium of
the amination reaction remains unaltered by the removal of
substrate, the method can serve to increase the overall
productivity of the system and approach higher conversions to
amine.
of substrates, such as to allow formal biocatalytic NH3/CO2
exchange with a variety of acrylic acids.
Acknowledgements
SLF acknowledges RCUK (BB/L013762/1; BB/M027791/1;
BB/M02903411; BB/M028836/1) and ERC (788231-ProgrES-
ERC-2017-ADG). This work was funded by the European
Union’s 7th Framework program FP7/2007-2013 under grant
agreement nos. 289646 (KYROBIO) and 115360 (CHEM21)
with the latter including EFPIA companies’ in-kind contribution
for the Innovative Medicine Initiative.
Table 2. Effect of varying ammonia concentration, buffer pH and EfPheDC
whole cell catalyst loading on the product distribution in the AvPAL-EfPheDC
cascade.[a]
______________________________________________________________
Ammonium
carbamate
[mM]
EfPheDC
E. coli wet cells
[mg mL-1]
1a
2a
3a
pH
[%][b]
[%][b]
[%][b]
Keywords: arylethylamines • biocatalysis • enzyme cascades •
1000
500
250
500
500
500
500
500
500
500
7.7
7.7
7.7
8.8
7.7
6.6
6.6
6.6
6.6
6.6
20
20
20
20
20
20
0
56
53
16
72
53
19
28
20
19
12
29
13
56
23
13
43
72
47
43
27
15
34
28
5
ammonia lyases • decarboxylases
References
34
39
0
[1]
[2]
N. Ohtsuka, M. Okuno, Y. Hoshino, K. Honda, Org. Biomol. Chem.
2016, 9046–9054.
U. T. Bornscheuer, G. W. Huisman, R. J. Kazlauskas, S. Lutz, J. C.
Moore, K. Robins, Nature 2012, 485, 185–194.
M. T. Reetz, J. Am. Chem. Soc. 2013, 135, 12480–12496.
H. S. Toogood, N. S. Scrutton, Catal. Sci. Technol. 2013, 3, 2182-
2194.
10
20
40
33
39
61
[3]
[4]
[a] Reaction conditions: 2 mM 1a, 10 mg mL-1 AvPAL E. coli dry cells,
30°C, 250 rpm, 22 h. [b] Determined by reverse phase HPLC analysis on
a non-chiral phase.
[5]
a) J. Steinreiber, M. Schuermann, M. Wolberg, F. Van Assema, C.
Reisinger, K. Fesko, D. Mink, H. Griengl, Angew. Chem. Int. Ed.
2007, 46, 1624–1626. b) D. Baud, O. Peruch, P.-L. Saaidi, A.
Fossey, A. Mariage, J.-L. Petit, M. Salanoubat, C. Vergne-Vaxelaire,
V. de Berardinis, A. Zaparucha, Adv. Synth. Catal. 2017, 359,
1563–1569.
This work has demonstrated the biocatalysed conversion of
easily accessible arylacrylic acids to β-arylethylamines using a
combination of PAL and AADC enzymes. The integration of the
well-characterised PAL biocatalyst with an AADC not widely
used for biotransformations prompted a substrate screen of the
latter enzyme, showing it to have a strict enantiopreference for L-
amino acids, broad substrate scope (particularly meta-
substituted) and possibly mild inhibition of activity in the
presence of D-enantiomers. The two biocatalytic steps could be
easily run sequentially in the same vessel, without
chromatographic purifications or complex separation steps in
between. This was achieved through simple removal of whole
cells and sublimation of the volatile reaction buffer following the
first step, allowing up to 98% conversion overall at ~30 mM
substrate concentration. The different pH optima and ammonia
tolerance exhibited by each enzyme warranted optimisation of a
cascade approach for the one-pot conversion of cinnamic acid to
its corresponding amine. By focussing on the requirements of
the second, irreversible decarboxylation step (i.e., low pH, low
ammonia concentration and increased cell loading) it was found
that conversion up to 61% to β-phenylethylamine could be
achieved, with the AADC acting to remove the initial amination
product thus enabling further consumption of starting material to
redress the equilibrium of the PAL reaction. This strategy,
combining an amino acid forming enzyme with an orthogonal
decarboxylase, could potentially be applied to different classes
[6]
[7]
[8]
[9]
Y. Zhou, S. Wu, Z. Li, Angew. Chem. Int. Ed. 2016, 55, 11647–
11650.
E. Busto, R. C. Simon, W. Kroutil, Angew. Chem. Int. Ed. 2015, 54,
10899–10902.
F. Parmeggiani, N. J. Weise, S. T. Ahmed, N. J. Turner, Chem. Rev.
2018, 118, 73–118.
a) S. L. Lovelock, R. C. Lloyd, N. J. Turner, Angew. Chem. Int. Ed.
2014, 53, 4652–4656. b) F. Parmeggiani, S. L. Lovelock, N. J.
Weise, S. T. Ahmed, N. J. Turner, Angew. Chem. Int. Ed. 2015, 54,
4608–4611.
[10]
[11]
B. de Lange, D. J. Hyett, P. J. D. Maas, D. Mink, F. B. J. van
Assema, N. Sereinig, A. H. M. de Vries, J. G. de Vries,
ChemCatChem 2011, 3, 289–292.
a) A. Marcobal, B. De Las Rivas, R. Muñoz, FEMS Microbiol. Lett.
2006, 258, 144–149. b) H. C. Hailes, Y. Wang, N. Tappertzhofen, D.
Méndez-Sánchez, M. Bawn, B. Lyu, J. Ward, Angew. Chem. Int. Ed.
2019, 15, 10120-10125
[12]
S. D. Christenson, W. Liu, M. D. Toney, B. Shen, J. Am. Chem. Soc.
2003, 125, 6062–6063.
[13]
[14]
C. L. Nesloney, J. W. Kelly, J. Org. Chem. 1996, 61, 3127–3137.
C. Karlsson, M. Blom, M. Johansson (neé Varedian), A. M. Jansson,
E. Scifo, A. Karlén, T. Govender, A. Gogoll, Org. Biomol. Chem.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201902128
ChemCatChem
COMMUNICATION
2015, 13, 2612–2621.
[15]
[16]
D. Boger, D. Yohannes, J. Org. Chem. 1990, 55, 6000–6017.
L. M. Smith, M. J. Orwat, Z. Hu, W. Han, C. Wang, K. A. Rossi, P. J.
Gilligan, K. B. Pabbisetty, H. Osuna, J. R. Corte, A. R Rendina, J. M
Luettgen, P. C Wong, R. Narayanan, T. W. Harper, J. M. Bozarth, E.
J Crain, A. Wie, V. Ramamurthy, P. E. Morin, B. Xin, J. Zheng, D. A.
Seiffert, M. L Quan, P. Y. S. Lam, R. R. Wexler, D. J. P. Pinto,
Bioorg. Med. Chem. Lett. 2016, 26, 472–478.
[17]
[18]
[19]
[20]
[21]
M. C. Moffitt, G. V Louie, M. E. Bowman, J. Pence, P. Joseph, B. S.
Moore, Biochemistry 2008, 46, 1004–1012.
L. Zhu, L. Zhou, W. Cui, Z. Liu, Z. Zhou, Biotechnol. Rep. 2014, 3,
21–26.
H. Z. Jaliani, S. Farajnia, S. A. Mohammadi, A. Barzegar, S. Talebi,
Appl. Biochem. Biotechnol. 2013, 171, 1805–1818.
J. Lin, M. P. Smith, K. C. Chapin, H. S. Baik, G. N. Bennett, J. W.
Foster, Appl. Env. Microbiol. 1996, 62, 3094–3100.
N. J. Weise, S. T. Ahmed, F. Parmeggiani, E. Siirola, A. Pushpanath,
U. Schell, N. J. Turner, Catal. Sci. Technol. 2016, 6, 4086–4089.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201902128
ChemCatChem
COMMUNICATION
Entry for the Table of Contents (Please choose one layout)
Layout 2:
COMMUNICATION
H
AMMONIA LYASE
+
H
Nicholas J. Weise, Prasansa Thapa,
Syed T. Ahmed, Rachel S. Heath, Fabio
Parmeggiani, Nicholas J. Turner, and
Sabine L. Flitsch*
H
H
H
R
R
DECARBOXYLASE
CO2H
NH2
H
NH3
CO2
up to 98% conv.
Page No. – Page No.
Biocatalysis: A combination of phenylalanine ammonia lyase and decarboxylase
enzymes allows the conversion of aromatic acrylic acids to the corresponding
ethylamines. The reaction can be performed either as a tandem reaction with facile
work up in between or as a biocatalytic cascade in a common reaction medium. A
variety of ring-substituted cinnamic acid derivatives can serve as starting materials
with varying conversion.
Bi-enzymatic Conversion of Cinnamic
Acids to 2-Arylethylamines
This article is protected by copyright. All rights reserved.