One-pot chemoenzymatic synthesis of trolline and tetrahydroisoquinoline analogues

Article abstract of DOI:10.1039/c7cc08024g

Chemoenzymatic reaction cascades can provide access to chiral compounds from low-cost starting materials in one pot. Here we describe one-pot asymmetric routes to tetrahydroisoquinoline alkaloids (THIAs) using the Pictet-Spenglerase norcoclaurine synthase (NCS) followed by a cyclisation, to give alkaloids with two new heterocyclic rings. These reactions operated with a high atom economy to generate THIAs in high yields.

Full text of DOI:10.1039/c7cc08024g

ChemComm
View Article Online
View Journal
COMMUNICATION
One-pot chemoenzymatic synthesis of trolline
and tetrahydroisoquinoline analogues†
Cite this:DOI: 10.1039/c7cc08024g
b
Jianxiong Zhao,a Benjamin R. Lichman, ‡b John M. Ward
and
a
Received 26th October 2017,
Accepted 30th November 2017
Helen C. Hailes
*
DOI: 10.1039/c7cc08024g
rsc.li/chemcomm
Chemoenzymatic reaction cascades can provide access to chiral have incorporated an amine oxidase or transaminases to
compounds from low-cost starting materials in one pot. Here we generate aldehydes:13,14 the latter strategy used a second PSR
describe one-pot asymmetric routes to tetrahydroisoquinoline to produce a tetrahydroprotoberberine.14 Engineered microbial
alkaloids (THIAs) using the Pictet–Spenglerase norcoclaurine synthase pathways to natural opioids incorporating NCS have also been
(NCS) followed by a cyclisation, to give alkaloids with two new described.15,16 The reported TfNCS X-ray crystal structure and
heterocyclic rings. These reactions operated with a high atom mechanistic studies,17–19 together with the aldehyde promiscuity
economy to generate THIAs in high yields.
and recent ketone acceptance,20 have highlighted the significant
potential of using NCSs more widely.
The use of biocatalytic strategies in synthetic applications
The alkaloid trolline 2, isolated from the flowers of Trollius
continues to offer many advantages compared to traditional chinensis, has been reported to be effective against Staphylococcus
chemical approaches due to their sustainability, the use of mild aureus and possess antiviral activity.21 There are few reported
reaction conditions and the high levels of stereocontrol that syntheses, however 3 was used in a Bischler–Napieralski strategy
can be achieved.1 Biocatalysts that enable C–C bond formation to give 2 in 44% yield (Scheme 1).22 Another 3-step approach
are particularly useful and norcoclaurine synthases (NCSs) are gave 2 in 73% yield, however the amine starting material was not
of significant interest for the synthesis of tetrahydroisoquino-
line alkaloids (THIAs).2 THIAs are a large group of secondary
metabolites with a range of pharmacological activities includ-
ing the analgesic morphine, anti-hypertensive magnoflorine,
and anti-mycobacterial leucoxine.3–5
In plant biosynthetic pathways, NCS catalyses the Pictet–Spengler
reaction (PSR)6 between dopamine 1 and 4-hydroxyphenyl
acetaldehyde (4-HPAA) to generate (S)-norcoclaurine.7 Recently,
it has been established that a range of aldehydes, particularly
substituted phenylacetaldehydes, and some heteroaromatic
and aliphatic substrates, can be accepted by recombinant wild-
type (WT) Thalictrum flavum (TfNCS) and Coptis japonica NCS
(CjNCS) to generate single-isomer THIAs.8–10 NCS has been used
in chemoenzymatic cascades, with hypochlorite to produce the
aldehyde component.11,12 In addition, NCS enzyme cascades
a Department of Chemistry, University College London, Christopher Ingold Building,
20 Gordon Street, London, WC1H 0AJ, UK. E-mail: h.c.hailes@ucl.ac.uk
b Department of Biochemical Engineering, University College London, Gower Street,
London, WC1E 6BT, UK
† Electronic supplementary information (ESI) available: Experimental details.
Preparation of NCS enzymes, reaction details, compound characterization and
analytical details. See DOI: 10.1039/c7cc08024g
Scheme 1 Previous syntheses of trolline (a and b) and the approach used
‡ Current address: John Innes Centre, Norwich Research Park, Norwich, NR4
in this work (c and d).
7UH, UK.
This journal is ©The Royal Society of Chemistry 2017
Chem. Commun.

View Article Online
Communication
ChemComm
commercially available.23 Multistep asymmetric strategies have yields, 89%, of the lactam. A one-pot procedure was then
been described. One used Jacobsen’s catalyst and HCN to established with 1 and 5 using first step a and then, via the
establish the C-1 stereochemistry,24 while another started with addition of sodium carbonate and adjustment of pH, step b, to
4 and used the (R,Sa)-N-pinap chiral catalyst to give (S)-2 in give 2 in 97% yield (para : ortho 18 : 1) (Table 1). An extraction
41% yield (5 steps);25 strong bases and toxic solvents were also and purification protocol was developed12,20 to avoid the need
required.25
for chromatographic purification, and rac-2 was isolated in
To develop more sustainable rapid routes to these important 81% yield (para : ortho 34 : 1 by HPLC). The one-pot protocol
THIAs, and inspired by the acceptance of non-aromatic was then applied to aldehyde 7 (n = 2). Notably, more linear
aldehydes by NCS, here we describe efficient chemocatalytic THIA 8 was generated in the first step while less cyclized
cascades to 2 and its analogues in up to 96% yield and 499% product 9 was formed (and some ortho-product). The formation
ee. One-pot chemoenzymatic cascades offer many advantages, of lactam 9 was readily achieved upon the addition of sodium
such as avoiding the need for intermediate isolation.26–28 carbonate, and isolated (69% yield, para : ortho 50 : 1). When
Two strategies to trolline 2 were adopted: first a biomimetic aldehyde 10 (n = 3) was used, the linear THIA 11 (72% yield,
phosphate-mediated PSR29 and subsequent cyclisation to para : ortho 7 : 1) was formed. No cyclisation to give 12 was
establish the one-pot reaction conditions. Secondly, a one-pot observed and 11 was isolated by preparative HPLC.
NCS-mediated PSR and then cyclisation to give (S)-trolline 2
The stereoselective approach using NCS was then investi-
and its analogues. Initial reactions were carried out using 1 and gated. Initially, the reaction conditions for the NCS reaction
commercially available 5 under aqueous potassium phosphate with the novel aldehyde substrates were explored. Using
(KPi) conditions.29 The PSR readily occurred at pH 6 to give a WT-D29TfNCS (see the ESI†) and 1 and 5, the use of acetonitrile
mixture of the linear THIA 6 and, due to spontaneous ring and DMSO as co-solvents was investigated as recent work has
cyclization, 2. A ratio of amine : aldehyde of 1 : 1.5 gave higher highlighted several-fold higher product formation with DMSO,
yields. A reaction temperature of 60 1C was optimal; above this, perhaps due to enzyme stabilisation effects.20 Here DMSO also
some product decomposition occurred. In addition, as dopamine gave a significantly higher yield (B2-fold increase) compared to
is oxidatively sensitive, ascorbate was added to avoid side-product acetonitrile after a 3 h reaction. The extension of the reaction
formation.20 These conditions gave a mixture of linear and time from 3 h to 6 h, and the use of 1.5 equiv. of aldehyde,
cyclised products in a quantitative yield (Table 1). It should be enhanced the combined yield of THIAs to 86%: no ortho-
noted that the major para-isomer 6 was formed, with some minor products were formed. The yield of 2 formed in step a using
ortho-isomer (ratio 15 : 1).12 The reaction conditions for the cycli- NCS was lower than that using KPi, reflecting the lower reaction
sation of linear intermediate 6 were then investigated. Acidic temperatures used. Furthermore, the amount of DMSO co-solvent
and basic conditions were explored using 6 ( para : ortho 15 : 1) could be reduced to 1% with little effect on the reaction conversion,
(ESI;† Fig. S1), and sodium carbonate at pH 7.5 gave the highest and this enhanced the ease of product isolation.
Table 1 One-pot routes to rac-2, (S)-trolline 2 and analogues via step a and steps a + b
Step aa or f
Steps a + bb
Linear product
Cyclised product
Product yieldc
( para : ortho)
Isolated product yield
( para : ortho)
Aldehyde
Solvent
yieldc ( para : ortho)
yieldc ( para : ortho)
Product eeh
KPi step aa
5
7
10
50% CH3CN/KPi
50% CH3CN/KPi
50% CH3CN/KPi
6 32% (15 : 1)
8 75% (8 : 1)
11 72% (7 : 1)
2 68% (22 : 1)
9 14% (13 : 1)
Only 11
2 97% (18 : 1)
9 89% (9 : 1)
11 72% (7 : 1)
2 81% (34 : 1)d
9 69% (50 : 1)e
11 26% (7 : 1)e
na
na
na
Enzymatic step a f
5g
5g
5
10% CH3CN/HEPES
6 35%
6 61%
6 71%
6 67%
8 82%
11 92%
2 3%
2 7%
2 15%
2 19%
9 15%
only 11
na
na
na
2 75%
9 96%
11 92%
na
na
na
nd
nd
nd
95%
96%
499%
10% DMSO/HEPES
10% DMSO/HEPES
1% DMSO/HEPES
1% DMSO/HEPES
1% DMSO/HEPES
5
7
10
2 74%d
9 87%d
11 63%e
a
b
1 and aldehyde (1 : 1.5) in KPi buffer (0.3 M)/CH3CN (1 : 1), 18 h, under Ar, 60 1C, pH 6, and ascorbic acid (1 equiv.). Na2CO3 (1 M), pH 7.5, and 4 h.
c
e
d
HPLC yields: calculated by analytical HPLC. Isolated in high purity by a basic and then acidic extraction procedure using EtOAc and then MeOCO2Me.
Isolated by preparative HPLC. 1 and aldehyde (ratio 1 : 1.5) in co-solvent/HEPES buffer (pH 7.5 and 0.1 M), 37 1C, WT-TfNCS (0.1 mg mLÀ1), sodium
f
g
h
ascorbate (1 equiv.), and 6 h. As for f but 1:5 in a ratio of 1.5 : 1, and 3 h reaction. The ees determined by chiral HPLC: for trolline the absolute
stereochemistry was confirmed by the optical rotation. na, not applicable. nd, not determined.
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2017

View Article Online
ChemComm
Communication
Following the reaction conditions established for step b, non-productive one (Fig. 1F). Such alternative favourable bind-
(S)-2 was then formed in the one-pot 2-step chemoenzymatic ing arrangements of shorter, less sterically demanding aliphatic
cascade in 75% yield, and 95% ee by chiral HPLC. In a aldehydes could lower the enzyme efficiency leading to reduced
preparative scale reaction, (S)-trolline 2 was readily isolated in yields. This has implications in other synthetic applications
74% yield without the use of chromatographic methods using using less sterically challenging aldehydes, and provides insights
the extraction protocol developed. Remarkably, although alde- for future NCS enzyme engineering.
hyde 5 is structurally quite different from the natural substrate
The CjNCS and three variants of D29TfNCS (A79I, A79F, and
4-hydroxyphenylacetaldehyde, it was still readily accepted. The F80L) with modified residues at the active site entrance
extension to aldehyde 7 (n = 2) similarly gave lactam 9 in 96% region9,18,20 were also screened against 1 and 5, 7, and 10 (ESI;†
yield (87% isolated yield on a preparative scale) and 96% ee. Fig. S3 and S4). Similar reaction profiles were observed with
When using aldehyde 10 (n = 3), the linear THIA 11 was formed WT-D29TfNCS, although A79F gave slightly higher conversions with
in 92% yield and in very high optical purity (499% ee) (Table 1). 7: a scale-up reaction (40 mL) generated lactam 9 in 93% isolated
The NCS reaction profiles using WT-TfNCS and aldehydes 5, yield and 499% ee. The higher ee may reflect steric effects of a
7, and 10 revealed that the bioconversion proceeded more more bulky residue on the orientation of intermediates.
rapidly with the longer chain aldehyde (ESI;† Fig. S2). Compu-
The one-pot, 2-step reaction sequence was then extended to
tational docking, using the NCS X-ray structure (5NON),19 of the dopamine analogues 13–15 using aldehyde 5. In the KPi-mediated
imine reaction intermediates when using 1 and 5, 7, and 10 was reactions, THIAs were readily formed in up to quantitative yields,
carried out to try and understand this (ESI;† Table S4). Two again containing predominantly para-isomers rac-16–18, with
‘dopamine-first’ binding modes were observed, both involving some ortho-product: when using 14, THIAs were formed as a
the key interaction between the catechol 3-OH and Lys-122 mixture of isomers (see the ESI†). For WT-TfNCS reactions with
residues.9,18,19 In productive binding modes (Fig. 1A–C), the 13, (S)-16 was formed in 51% yield as the only regioisomer and
intermediate occupies a conformation conducive to the subse- 93% ee (Fig. 2). The bulkier amine metaraminol 14 reacted with 5
quent cyclisation step. There were also non-productive binding in 76% yield, and gave (1S)-17 (at C-1) as the major isomer (ratio
modes (Fig. 1D–F), with the ester substituent extended into a 10: 1) where the minor isomers were a mixture of (1R)-17 (at C-1)
water channel. The docking experiments suggested that as the and the ortho-product (ratio B1 : 1). The stereochemistry at C-1
aldehyde chain length increases, the productive pose becomes was confirmed by 2D 1H NMR spectroscopy. While higher
more favourable relative to the non-productive one. For example, enzymatic stereoselectivities were observed with metaraminol
docking calculations with 5 (n = 1) identified the non-productive 14 and phenylacetaldehyde,28 here a non-aromatic aldehyde,
pose (Fig. 1D) to be more favourable than the productive pose which can give rise to alternative binding arrangements, led to
(Fig. 1A). By comparison, docking with 10 (n = 3) indicated lower selectivities. When using 15, (S)-18 was formed in 57%
the productive pose (Fig. 1C) as more favourable than the yield (87% ee). This is the first time that a fluorinated dopamine
Fig. 1 Reaction imine intermediates docked into the NCS X-ray structure (5NON).19 Binding modes depicted are the most favoured productive binding
modes enabling cyclisation (A–C) or extended non-productive modes (D and E). (A) Productive imine intermediate with aldehyde 5. (B) Productive imine
intermediate with aldehyde 7. (C) Productive imine intermediate with aldehyde 10. (D) Non-productive iminium intermediate with aldehyde 5. (E) Non-
productive iminium intermediate with aldehyde 7. (F) Non-productive iminium intermediate with aldehyde 10.
This journal is ©The Royal Society of Chemistry 2017
Chem. Commun.

View Article Online
Communication
ChemComm
Nature, 2012, 485, 185; (c) J. H. Schrittwieser and V. Resch, RSC Adv.,
2013, 3, 17602.
2 For reviews see: (a) A. Bonamore, M. Barba, B. Botta, A. Boffi and
A. Macone, Molecules, 2010, 15, 2070; (b) F. Ghirga, D. Quaglio, P. Ghirga,
S. Berardozzi, G. Zappia, B. Botta, M. Mori and I. D’Acquarica, Chirality,
2016, 28, 169.
´
3 E. Sverrisdottir, T. M. Lund, A. E. Olesen, A. M. Drewes, L. L.
Christrup and M. Kreilgaard, Eur. J. Pharm. Sci., 2015, 74, 45.
4 M. A. Rashid, K. R. Gustafson, Y. Kashman, J. H. Cardellina, J. B.
McMahon and M. R. Boyd, Nat. Prod. Lett., 1995, 6, 153.
5 J. D. Guzman, T. Pesnot, D. A. Barrera, H. A. Davies, E. McMahon,
D. Evangelopoulos, P. N. Mortazavi, T. Munshi, A. Maitra, E. D.
Lamming, R. Angell, M. C. Gershater, J. M. Redmond, D. Needham,
J. M. Ward, L. E. Cuca, H. C. Hailes and S. Bhakta, J. Antimicrob.
Chemother., 2015, 70, 1691.
¨
6 J. Stockigt, A. P. Antonchick, F. Wu and H. Waldmann, Angew.
Chem., Int. Ed., 2011, 50, 8538.
7 N. Samanani, D. K. Liscombe and P. J. Facchini, Plant J., 2004,
40, 302.
¨
8 B. M. Ruff, S. Brase and S. E. O’Connor, Tetrahedron Lett., 2012,
53, 1071.
Fig. 2 THIAs 16–18 prepared using dopamine analogues and the one-pot
9 T. Pesnot, M. C. Gershater, J. M. Ward and H. C. Hailes, Adv. Synth.
Catal., 2012, 354, 2997.
2-step cascade.
10 M. Nishihachijo, Y. Hirai, S. Kawano, A. Nishiyama, H. Minami,
T. Katayama, Y. Yasohara, F. Sato and H. Kumagai, Biosci., Biotechnol.,
Biochem., 2014, 78, 701.
11 A. Bonamore, I. Rovardi, F. Gasparrini, P. Baiocco, M. Barba,
C. Molinaro, B. Botta, A. Boffi and A. Macone, Green Chem., 2010,
12, 1623.
12 J. J. Maresh, S. O. Crowe, A. A. Ralko, M. D. Aparece, C. M. Murphy,
M. Krzeszowiec and M. W. Mullowney, Tetrahedron Lett., 2014,
55, 5047.
13 A. Bonamore, L. Calisti, A. Calcaterra, O. H. Ismail, M. Gargano,
I. D’Acquarica, B. Botta, A. Boffi and A. Macone, ChemistrySelect,
2016, 1, 1525.
14 B. R. Lichman, E. D. Lamming, T. Pesnot, J. M. Smith, H. C. Hailes
and J. M. Ward, Green Chem., 2015, 17, 852.
15 W. C. DeLoache, Z. N. Russ, L. Narcross, A. M. Gonzales, V. J. Martin
and J. E. Dueber, Nat. Chem. Biol., 2015, 11, 465.
16 S. Galanie, K. Thodey, I. J. Trenchard, M. F. Interrante and
C. D. Smolke, Science, 2015, 349, 1095.
17 A. Ilari, S. Franceschini, A. Bonamore, F. Arenghi, B. Botta, A. Macone,
A. Pasquo, L. Bellucci and A. Boffi, J. Biol. Chem., 2009, 284, 879.
18 B. R. Lichman, M. C. Gershater, E. D. Lamming, T. Pesnot, A. Sula,
N. H. Keep, H. C. Hailes and J. M. Ward, FEBS J., 2015, 282, 1137.
19 B. R. Lichman, A. Sula, T. Pesnot, H. C. Hailes, J. M. Ward and
N. H. Keep, Biochemistry, 2017, 56, 5274.
analogue has successfully been used with NCS to our knowledge.
Small-scale reactions were performed to obtain THIAs 16–18
via the one-pot reactions and the products were isolated by
preparative HPLC.
Computational modelling of the aminol intermediate with
15 confirmed a similar fit into the active site compared to
dopamine (ESI;† Fig. S5). Note that ees were determined by
chiral HPLC and the absolute stereochemistry of 16 and 18 was
assigned following the established stereoselectivity of the NCS
enzymes at C-1 and the correlation to the known stereochemistry
of trolline 2.
Overall, this work highlights the potential of using efficient
one-pot chemoenzymatic cascades, not requiring the purifica-
tion of intermediates, for the rapid stereoselective synthesis
of natural and novel alkaloids. It also illustrates the wide
substrate tolerance of the NCS enzyme. Furthermore, it demon-
strates the application of the sustainable asymmetric catalyst
NCS in aqueous media for the synthesis of tricyclic alkaloids.
We gratefully acknowledge UCL (Dean’s Prize) and the China
Scholarship Council-UCL Joint Research Scholarship for funding
to J. Z. and the Wellcome Trust for studentship funding to
B. R. L. We also thank K. Karu (UCL Mass Spectrometry Facility)
and A. E. Aliev (UCL NMR Facility) of the Department of Chemistry.
20 B. R. Lichman, J. Zhao, H. C. Hailes and J. M. Ward, Nat. Commun.,
2017, 8, 14883.
21 R. F. Wang, X. W. Yang, C. M. Ma, S. Q. Cai, J. N. Li and Y. Shoyama,
Heterocycles, 2004, 63, 1443.
´
´
22 L. Moreno, J. Parraga, A. Galan, N. Cabedo, J. Primo and D. Cortes,
Bioorg. Med. Chem., 2012, 20, 6589.
23 J. Selvakumar, R. S. Rao, V. Srinivasaptiyan, S. Marutheeswaran and
C. R. Ramanathan, Eur. J. Org. Chem., 2015, 2175.
24 T. Kanemitsu, Y. Yamashita, K. Nagata and T. Itoh, Heterocycles,
2007, 74, 199.
25 W. Lin and S. Ma, Org. Chem. Front., 2017, 4, 958.
26 E. Garcia-Junceda, I. Lavendera, D. Rother and J. H. Schrittwieser,
J. Mol. Catal. B: Enzym., 2015, 114, 1.
27 H. C. Hailes, P. A. Dalby and J. M. Woodley, J. Chem. Technol.
Biotechnol., 2007, 82, 1063.
Conflicts of interest
There are no conflicts to declare.
28 V. Erdmann, B. R. Lichmann, J. Zhao, R. C. Simon, W. Kroutil,
J. M. Ward, H. C. Hailes and D. Rother, Angew. Chem., Int. Ed., 2017,
56, 12503.
29 T. Pesnot, M. C. Gershater, J. M. Ward and H. C. Hailes, Chem.
Commun., 2011, 47, 3242.
Notes and references
1 For reviews see: (a) B. M. Nestl, S. C. Hammer, B. A. Nebel and
B. Hauer, Angew. Chem., Int. Ed., 2014, 53, 3070; (b) U. T. Bornscheuer,
G. W. Huisman, R. J. Kazlauskas, S. Lutz, J. C. Moore and K. Robins,
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2017