Asymmetric Synthesis of N-Substituted α-Amino Esters from α-Ketoesters via Imine Reductase-Catalyzed Reductive Amination

Article abstract of DOI:10.1002/anie.202016589

N-Substituted α-amino esters are widely used as chiral intermediates in a range of pharmaceuticals. Here we report the enantioselective biocatalyic synthesis of N-substituted α-amino esters through the direct reductive coupling of α-ketoesters and amines employing sequence diverse metagenomic imine reductases (IREDs). Both enantiomers of N-substituted α-amino esters were obtained with high conversion and excellent enantioselectivity under mild reaction conditions. In addition >20 different preparative scale transformations were performed highlighting the scalability of this system.

Full text of DOI:10.1002/anie.202016589

Angewandte
Communications
Chemie
How to cite: Angew. Chem. Int. Ed. 2021, 60, 8717–8721
International Edition: doi.org/10.1002/anie.202016589
German Edition: doi.org/10.1002/ange.202016589
Biocatalysis
Asymmetric Synthesis of N-Substituted a-Amino Esters from
a-Ketoesters via Imine Reductase-Catalyzed Reductive Amination
Peiyuan Yao+, James R. Marshall+, Zefei Xu, Jesmine Lim, Simon J. Charnock, Dunming Zhu,*
and Nicholas J. Turner*
Abstract: N-Substituted a-amino esters are widely used as
chiral intermediates in a range of pharmaceuticals. Here we
report the enantioselective biocatalyic synthesis of N-substi-
tuted a-amino esters through the direct reductive coupling of a-
ketoesters and amines employing sequence diverse metage-
nomic imine reductases (IREDs). Both enantiomers of N-
substituted a-amino esters were obtained with high conversion
and excellent enantioselectivity under mild reaction conditions.
In addition > 20 different preparative scale transformations
were performed highlighting the scalability of this system.
Figure 1. Representative N-substituted a-amino acid derivatives with
N
-Substituted a-amino acids and their derivatives have
biological activities where the N-substituted a-amino acid scaffolds are
highlighted in blue.
attracted increasing attention in the pharmaceutical and fine
chemical industries in recent years, forming the key scaffolds
in a number of bioactive molecules (Figure 1).[1] For example,
N-methylated analogues of peptides and peptidomimetics can
improve the pharmacokinetic properties of such molecules,
including metabolic stability, membrane permeability, and
oral bioavailability.[2] Although tremendous efforts have been
devoted to the preparation of N-alkyl-a-amino acids,[1a,3]
including bio-inspired asymmetric reductive aminations,[4]
there are a number of limitations to these synthetic methods.
The N-alkylation process often requires genotoxic alkylating
agents, such as alkyl halides,[3a] and can be impractical on large
scale due to difficulty in removing specific protecting
groups.[5] Biomimetic routes also have their constraints due
to the need to preform the imine and the employment of
transition metals such as ruthenium.[4,6]
As a more sustainable and alternative approach to
chemical methods, biocatalysis has become an attractive
option in preparing chiral N-substituted a-amino acids or
their derivatives.[1] Current approaches include N-methyla-
tion of amino acids and peptides by N-methyltransferases,[7]
synthesis of N-arylated aspartic acids catalyzed by ethyl-
enediamine-N,N’-disuccinic acid lyase,[8] and reductive ami-
nation of a-ketoacids by NAD(P)H-dependent oxidoreduc-
tases including opine dehydrogenases (OpDHs), ketimine
reductases and N-methyl amino acid dehydrogenases,[9] and
engineered heme-dependent proteins.[10] However, the cur-
rent enzymatic toolbox for the synthesis of N-substituted a-
amino acids has several limitations including strict stereose-
lectivity for formation of the l-(S)-enantiomer. Many enzyme
families have narrow substrate specificities, with respect to
either the a-ketoacid or amine partner, with high activities
limited to simple primary amines such as methylamine.
Furthermore, with the exception of a single patent reporting
reductive aminations employing OpDHs, limited information
is available regarding stereoselectivities and activities.[11]
Thus, methods for the asymmetric synthesis of N-substituted
a-amino acids, or their derivatives, to access both enantio-
meric series across a broad range of substrates, remain
a significant challenge.
[*] Prof. P. Yao,[+] J. R. Marshall,[+] Prof. N. J. Turner
Department of Chemistry, University of Manchester
Manchester Institute of Biotechnology
131 Princess Street, Manchester, M1 7DN (UK)
E-mail: nicholas.turner@manchester.ac.uk
Prof. P. Yao,[+] Z. Xu, Prof. D. Zhu
National Technology Innovation Center of Synthetic Biology
National Engineering Laboratory for Industrial Enzymes and Tianjin
Engineering Research Center of Biocatalytic Technology
Tianjin Institute of Industrial Biotechnology
Chinese Academy of Sciences
32 Xi Qi Dao, Tianjin Airport Economic Area, Tianjin 300308 (P.R.
China)
E-mail: zhu_dm@tib.cas.cn
Dr. J. Lim, Prof. S. J. Charnock
Prozomix Ltd
Building 4, West End Ind. Estate, Haltwhistle, NE49 9HA (UK)
[+] These authors contributed equally to this work.
Supporting information and the ORCID identification number(s) for
the author(s) of this article can be found under:
https://doi.org/10.1002/anie.202016589.
Imine reductases (IREDs) and reductive aminases
(RedAms) belong to a family of NAD(P)H-dependent
oxidoreductases which have been utilised in the synthesis of
chiral amines employing both reductive amination and cyclic
imine reduction reactions.[12] Reductive amination, employing
wild-type IRED biocatalysts, has helped to define the
substrate scope across both carbonyl acceptors and amine
ꢀ 2021 The Authors. Angewandte Chemie International Edition
published by Wiley-VCH GmbH. This is an open access article under
the terms of the Creative Commons Attribution License, which
permits use, distribution and reproduction in any medium, provided
the original work is properly cited.
Angew. Chem. Int. Ed. 2021, 60, 8717 –8721
ꢀ 2021 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH
8717

Angewandte
Communications
Chemie
partners (primary, and secondary amines).[13] Recently, a large
metagenomic (384 enzymes) IRED panel was generated and
applied to the reductive coupling of b-ketoesters to generate
enantiocomplementary N-substituted b-amino esters on
preparative scales.[14] Furthermore, the synthetic utility of
this enzyme class was recently highlighted on a kilogram scale
using an engineered IRED for the synthesis of a lysine-
specific demethylase-1 (LSD1) inhibitor, GSK2879552.[13a]
We therefore sought to evaluate the wild-type metage-
nomic IRED panel[14] for the reductive amination of aliphatic
and aromatic a-ketoesters, using various amine partners, to
assess both the level of conversion and enantioselectivities, in
order to extend the IRED catalysed reductive amination
scope to this class of compounds.
Table 1: Conversion and enantioselectivity of the top 12 selected IREDs
out of 384 enzymes towards reductive amination between ethyl 2-oxo-4-
phenylbutyrate (1) and propargylamine (a).[a] (S)-1a given in green and
(R)-1a given in blue.
Previous reports had demonstrated the ability of meta-
genomic IREDs and RedAms to accept g- and b-ketoesters
for reductive aminations.[13e,14] However, with a-ketoacids,
specifically pyruvic acid, low activities were observed towards
this substrate using the reductive aminase from Aspergillus
oryzae (AspRedAm).[13e] A genetically engineered Coryne-
bacterium glutamicum strain, expressing an imine reducing
enzyme of a different oxidoreductase family to the metage-
nomic IREDs, belonging to Delta(1)-pyrroline-2-carboxylate/
Delta(1)-piperideine-2-carboxylate reductase (DpkA) was
shown to transform pyruvate to N-methyl-l-alanine.[15]
Hence exploring additional IRED and RedAm sequence
space was of interest to see if we could obtain greater
activities with a-ketoester substrates.
The panel of 384 IREDs (cell-free extracts) was initially
screened for the reductive amination of model substrate ethyl
2-oxo-4-phenylbutyrate (1, 25 mm) with propargylamine (a,
50 mm) in a 100 mL reaction volume. This initial screen
revealed that 99 out of the 384 different IREDs (as shown in
Supplementary Table S3) were found to catalyse the desired
transformation. Further analysis revealed that 78 of these
IREDs were R-selective (of which 35 IREDs exhibited
excellent stereoselectivity with ee > 99%), while 20 were
found to be S-selective where only pIR-338 showed both
excellent conversion and selectivity, and one IRED generated
racemic 1a. The top 5 S-selective and top 7 R-selective
enzymes, with relative activities for the generation of 1a,
are shown in Table 1. These 12 enzymes were subsequently
selected for further reductive amination reactions with
a broader range of a-ketoesters.
Initially the 12 IREDs were evaluated across a variety of
aryl and alkyl a-ketoesters (2–11) ([S] = 50 mm) with prop-
argylamine (a) selected as the amine donor ([S] = 100 mm) on
an analytical scale. All ketoester substrates except 10 and 11
were transformed to the corresponding N-propargyl amino
esters with moderate to high conversion, high stereoselectiv-
ity, and with complementary enantiopreference (Figure 2 and
Supplementary Table S4); larger substituents, such as benzyl
and pentyl groups, were tolerated well. Interestingly, the
stereoselectivities of enzymes such as pIR-117 and pIR-258
were inverted when challenged with different a-ketoesters.
For example, both IREDs generated (R)-2a with a benzyl
substituent at the b-position with > 99% ee. However, when
this substituent was modified to a methyl group, both enzymes
generated (S)-3a with 97% ee for pIR-117 and 94% ee for
[a] Reaction conditions: 25 mm ethyl 2-oxo-4-phenylbutyrate (1), prop-
argylamine (a, 50 mm), 5 mgmLÀ1 lysate of E. coli expressing IRED,
6 UmLÀ1 CDX-901 glucose dehydrogenase (GDH), 0.4 mm NADP+,
62.5 mm glucose, 10% (v/v) DMSO, sodium phosphate buffer (100 mm,
pH 7.5), 100 mL reaction volume, 308C, 200 rpm, 20 h. [b] Enantiomeric
excess (ee) was determined by chiral HPLC. [c] Conversion into product
was determined by GC.
pIR-258. This inversion of stereoselectivity is not uncommon
within this enzyme family and has been previously
observed.[16]
The scope of the selected 12 IREDs was further evaluated
with a variety of amine partners (Table 2) where analytical
scale biotransformations with 1 (50 mm) and amines a, d, e,
and g (100 mm) and amines b, c, f (500 mm) were performed.
Functionalised amines were selected including propargyl-
amine (a), allylamine (d) and 4-methylbenzylamine (g), as
well as linear amines (propylamine (c)) and cyclic amines
(cyclopropylamine (e)). As highlighted in Table 2, pIR-23,
pIR-271, pIR-325, and pIR-338 exhibited excellent stereose-
lectivity towards a–e, whilst again the enantioselectivities of
some IREDs varied depending on the amine partner pre-
sented. The stereoselectivity of pIR-355 was inverted to
afford the R-enantiomers with amines methylamine (b) and
allylamine (d) to generate (R)-1b and (R)-1d respectively.
Only pIR-23 showed activity towards 1 and g with 82%
conversion and 99% ee (R), but no activity was observed for
cyclopentylamine (f).
8718
www.angewandte.org
ꢀ 2021 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH
Angew. Chem. Int. Ed. 2021, 60, 8717 –8721

Angewandte
Communications
Chemie
Table 2: Amine scope of reductive aminations with ethyl 2-Oxo-4-
phenylbutyrate. Conversion and enantioselectivity of selected IREDs
towards Ethyl 2-Oxo-4-phenylbutyrate (1) and different Amines (a–g).
Highlighted in green is the top performing enzyme (both selectivity and
conversion) for the given (S)-product and blue highlights the top
performing enzyme for the given (R)-product.[a]
Figure 2. Reductive amination scope of a-ketoesters with varying
substituents at the b-position. % Conversion into product and enantio-
selectivity (% ee) of 12 selected IREDs towards a-ketoesters (2–9) and
propargylamine (a) where the top IREDs to generate S and R enan-
tiomers are shown with an extended table show in Supplementary
Table S4. No IREDs showed activity towards 10 and 11.
To demonstrate the synthetic practicality of this biocat-
alytic approach, a series of preparative scale reactions was
performed on a 2.5 mmol scale in a total reaction volume of
50 mL (Figure 3). For all of these preparative reactions, good
to excellent conversions were observed (53–99%), and the
products were isolated as their HCl salts in moderate to high
yields (27–80%) with excellent ee values for R-selective
IREDs (98–99%) and good to excellent ee values for S-
selective IREDs (26–99%). With allylamine (d) as the amine
partner, employing metagenomic IRED pIR-271, (R)-1d was
synthesised in 58% isolated yield with excellent enantiose-
lectivity (> 99%), with this enzyme also being employed on
preparative scale to ascertain 13 out of a possible 14 (R)-
products (Figure 3). To highlight the robustness of these
enzymes towards the preparation of 1a, we obtained space
time yields up to 6.6 gLÀ1 dÀ1 with pIR-271. Employing
purified IR-338 for the preparation of 1a we obtained total
turnover numbers (TTN) of 3500 with a turnover frequency
(TOF) of 24 minÀ1. Conversion to 1a was monitored over
24 hours with pIR-271 (Figure S96), where nearly full con-
version was reached after 2.5 hours of reaction time. Owing to
the competing ketone reduction catalysed by endogenous
E. coli ketoreductase activity, lower isolated yields of (S)-1c,
(S)-1d, (R)-2a, and (S)-2a were obtained.[17] This was further
investigated with purified pIR-338 for the preparation of 1a
as shown in Supporting Information Table S5. The origin of
[a] Reaction conditions: 50 mm ethyl 2-oxo-4-phenylbutyrate (1), amine
(100 mm for a, d, e, and g, 500 mm for b, c, and f), 50 mgmLÀ1 E. coli
whole cells expressing IRED, 6 UmLÀ1 CDX-901 GDH, 0.4 mm NADP+,
125 mm glucose, 10% (v/v) DMSO, sodium phosphate buffer (100 mm,
pH 7.5), 500 mL reaction volume, 308C, 200 rpm, 20 h. [b] Conversion
into product was determined by GC. [c] Enantiomeric excess (ee) was
determined by chiral HPLC. [d] Not determined owing to low conversion
for 1 and f, only pIR-23 showed activity towards 1 and g.
this ketoreductase activity appears to be either directly the
CDX-901, or the lysate from which the GDH is formulated,[18]
as purified pIR-338 demonstrated no ketoreductase acitvity.
In summary, we have developed a highly efficient
biocatalytic strategy for the synthesis of N-substituted
amino esters from a-ketoesters and amines catalysed by
IREDs. This approach offers efficient access to various
enantiomerically pure N-substituted amino esters from aryl
and alkyl substituted a-ketoesters with exquisite and com-
plementary enantioselectivities, addressing problems identi-
fied in previous chemocatalytic and biocatalytic approaches
to attain these compounds. The synthetic utility and scal-
ability of this system was highlighted through 28 preparative
scale transformations. This study continues to emphasise the
Angew. Chem. Int. Ed. 2021, 60, 8717 –8721
ꢀ 2021 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH
www.angewandte.org
8719

Angewandte
Communications
Chemie
Conflict of interest
The authors declare no conflict of interest.
Keywords: biocatalysis · chiral amines · imine reductase ·
reductive amination · a-amino acid
[1] a) J. F. Hyslop, S. L. Lovelock, A. J. B. Watson, P. W. Sutton,
G. D. Roiban, J. Biotechnol. 2019, 293, 56 – 65; b) G. Budin, K. S.
Yang, T. Reiner, R. Weissleder, Angew. Chem. Int. Ed. 2011, 50,
9378 – 9381; Angew. Chem. 2011, 123, 9550 – 9553.
[2] a) J. Chatterjee, C. Gilon, A. Hoffman, H. Kessler, Acc. Chem.
Res. 2008, 41, 1331 – 1342; b) J. Chatterjee, F. Rechenmacher, H.
Kessler, Angew. Chem. Int. Ed. 2013, 52, 254 – 269; Angew.
Chem. 2013, 125, 268 – 283; c) M. Luisa Di Gioia, A. Leggio, F.
Malagrinꢀ, E. Romio, C. Siciliano, A. Liguori, Mini-Rev. Med.
Chem. 2016, 16, 683 – 690.
[3] a) L. Aurelio, R. T. C. Brownlee, A. B. Hughes, Chem. Rev. 2004,
104, 5823 – 5846; b) C. Nꢁjera, J. M. Sansano, Chem. Rev. 2007,
107, 4584 – 4671; c) B. Liu, S. Zhu, W. Zhang, C. Chen, Q. Zhou,
J. Am. Chem. Soc. 2007, 129, 5834 – 5835; d) S. Zhu, B. Xu, G.
Wang, Q. Zhou, J. Am. Chem. Soc. 2012, 134, 436 – 442; e) D.
Gillingham, N. Fei, Chem. Soc. Rev. 2013, 42, 4918 – 4931; f) Y.
Zhu, X. Liu, S. Dong, Y. Zhou, W. Li, L. Lin, X. Feng, Angew.
Chem. Int. Ed. 2014, 53, 1636 – 1640; Angew. Chem. 2014, 126,
1662 – 1666; g) V. Arredondo, S. C. Hiew, E. S. Gutman,
I. D. U. A. Premachandra, D. L. Van Vranken, Angew. Chem.
Int. Ed. 2017, 56, 4156 – 4159; Angew. Chem. 2017, 129, 4220 –
4223; h) M. Li, J. Yu, Y. Li, S. Zhu, Q. Zhou, Science 2019, 366,
990 – 994.
[4] H. Kuang, M. L. Brown, R. R. Davies, E. C. Young, M. D.
Distefano, J. Am. Chem. Soc. 1996, 118, 10702 – 10706.
[5] E. Fischer, W. Lipschitz, Ber. Dtsch. Chem. Ges. 1915, 48, 360 –
378.
[6] S. M. So, H. Kim, L. Mui, J. Chin, Eur. J. Org. Chem. 2012, 229 –
241.
[7] a) N. S. van der Velden, N. Kꢂlin, M. J. Helf, J. Piel, M. F.
Freeman, M. Kꢃnzler, Nat. Chem. Biol. 2017, 13, 833 – 835;
b) J. Chatterjee, B. Laufer, H. Kessler, Nat. Protoc. 2012, 7, 432 –
444.
Figure 3. Preparative scale reductive aminations of a-ketoesters with
different amines to generate N-substituted a-amino esters. Each
product isolated as the HCl salt with yields for each given and ee in
brackets. Conversions into product were determined by GC analysis
and ee values were determined by chiral GC analysis and HPLC. Full
figure with conversion data inlcuded is given in Supplementary Fig-
ure S2.
[8] H. G. Fu, A. P. Lujan, L. Bothof, J. L. Zhang, P. G. Tepper, G. J.
Poelarends, ACS Catal. 2019, 9, 7292 – 7299.
[9] a) T. Li, X. Cui, Y. Cui, J. Sun, Y. Chen, T. Zhu, C. Li, R. Li, B.
Wu, ACS Catal. 2020, 10, 7950 – 7957; b) J. F. Hyslop, S. L.
Lovelock, P. W. Sutton, K. K. Brown, A. J. B. Watson, G. D.
Roiban, Angew. Chem. Int. Ed. 2018, 57, 13821 – 13824; Angew.
Chem. 2018, 130, 14017 – 14020; c) T. Dairi, Y. Asano, Appl.
Environ. Microbiol. 1995, 61, 3169 – 3171.
value and applicability of metagenomic imine reductases in
the synthesis of high-value chiral amines.
Acknowledgements
[10] V. Steck, D. M. Carminati, N. R. Johnson, R. Fasan, ACS Catal.
2020, 10, 10967 – 10977.
[11] H. Chen, S. J. Collier, J. Nazor, J. Sukumaran, D. Smith, J. C.
Moore, G. Hughes, J. Janey, G. Huisman, S. Novick, N. Agard, O.
Alvizo, G. Cope, W. L. Yeo, S. NG, (Codexis, Inc.), US Patent
2013/0302859, 2013.
We would like to thank the National Key R&D Program of
China (No. 2019YFA0905100), the Youth Innovation Promo-
tion Association of the Chinese Academy of Sciences (Grant
No. 2016166) and CSC scholarship to support P.Y., and the
Industrial Biotechnology Innovation Centre (IBioIC) and
Biotechnology and Biological Sciences Research Council
(BBSRC) for the awarding the CASE studentship to J.R.M.
from Prozomix Ltd. N.J.T. is grateful to the European
Research Council (ERC) for the award of an Advanced
Grant (Grant number 742987).
[12] a) M. D. Patil, G. Grogan, A. Bommarius, H. Yun, ACS Catal.
2018, 8, 10985 – 11015; b) S. C. Cosgrove, A. Brzezniak, S. P.
France, J. I. Ramsden, J. Mangas-Sanchez, S. L. Montgomery,
R. S. Heath, N. J. Turner in Enzymes in Synthetic Biology,
Vol. 608 (Ed.: N. Scrutton), Elsevier, Amsterdam, 2018, pp. 131 –
149; c) M. Sharma, J. Mangas-Sanchez, N. J. Turner, G. Grogan,
Adv. Synth. Catal. 2017, 359, 2011 – 2025; d) J. Mangas-Sanchez,
S. P. France, S. L. Montgomery, G. A. Aleku, H. Man, M.
Sharma, J. I. Ramsden, G. Grogan, N. J. Turner, Curr. Opin.
Chem. Biol. 2017, 37, 19 – 25; e) M. Lenz, N. Borlinghaus, L.
Weinmann, B. M. Nestl, World J. Microbiol. Biotechnol. 2017, 33,
199; f) G. Grogan, N. J. Turner, Chem. Eur. J. 2016, 22, 1900 –
8720 www.angewandte.org
ꢀ 2021 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH
Angew. Chem. Int. Ed. 2021, 60, 8717 –8721

Angewandte
Communications
Chemie
1907; g) J. H. Schrittwieser, S. Velikogne, W. Kroutil, Adv. Synth.
[15] M. Mindt, J. M. Risse, H. Gruß, N. Sewald, B. J. Eikmanns, V. F.
Catal. 2015, 357, 1655 – 1685.
Wendisch, Sci. Rep. 2018, 8, 12895.
[13] a) M. Schober, C. MacDermaid, A. A. Ollis, S. Chang, D. Khan,
J. Hosford, J. Latham, L. A. F. Ihnken, M. J. B. Brown, D. Fuerst,
M. J. Sanganee, G.-D. Roiban, Nat. Catal. 2019, 2, 909; b) J. I.
Ramsden, R. S. Heath, S. R. Derrington, S. L. Montgomery, J.
Mangas-Sanchez, K. R. Mulholland, N. J. Turner, J. Am. Chem.
Soc. 2019, 141, 1201 – 1206; c) N. Borlinghaus, S. Gergel, B. M.
Nestl, ACS Catal. 2018, 8, 3727 – 3732; d) P. Matzel, M. Gand, M.
Hçhne, Green Chem. 2017, 19, 385 – 389; e) G. A. Aleku, S. P.
France, H. Man, J. Mangas-Sanchez, S. L. Montgomery, M.
Sharma, F. Leipold, S. Hussain, G. Grogan, N. J. Turner, Nat.
Chem. 2017, 9, 961 – 969.
[16] G. A. Aleku, H. Man, S. P. France, F. Leipold, S. Hussain, L.
Toca-Gonzalez, R. Marchington, S. Hart, J. P. Turkenburg, G.
Grogan, N. J. Turner, ACS Catal. 2016, 6, 3880 – 3889.
[17] A. Bornadel, S. Bisagni, A. Pushpanath, S. L. Montgomery, N. J.
Turner, B. Dominguez, Org. Process Res. Dev. 2019, 23, 1262 –
1268.
[18] Y. Li, J. Chen, S. Y. Lun, Appl. Microbiol. Biotechnol. 2001, 57,
451 – 459.
[14] J. R. Marshall, P. Yao, S. L. Montgomery, J. D. Finnigan, T. W.
Thorpe, R. B. Palmer, J. Mangas-Sanchez, R. J. Duncan, R. S.
Heath, K. M. Graham, D. J. Cook, S. J. Charnock, N. J. Turner,
Nat. Chem. 2021, 13, 140—148.
Manuscript received: December 14, 2020
Revised manuscript received: February 1, 2021
Accepted manuscript online: February 8, 2021
Version of record online: March 9, 2021
Angew. Chem. Int. Ed. 2021, 60, 8717 –8721
ꢀ 2021 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH
www.angewandte.org
8721