Separate Sets of Mutations Enhance Activity and Substrate Scope of Amine Dehydrogenase

Article abstract of DOI:10.1002/cctc.201902364

Mutations were introduced into the leucine amine dehydrogenase (L-AmDH) derived from G. stearothermophilus leucine dehydrogenase (LeuDH) with the goals of increased activity and expanded substrate acceptance. A triple variant (L-AmDH-TV) including D32A, F101S, and C290V showed an average of 2.5-fold higher activity toward aliphatic ketones and an 8.0 °C increase in melting temperature. L-AmDH-TV did not show significant changes in relative activity for different substrates. In contrast, L39A, L39G, A112G, and T133G in varied combinations added to L-AmDH-TV changed the shape of the substrate binding pocket. L-AmDH-TV was not active on ketones larger than 2-hexanone. L39A and L39G enabled activity for straight-chain ketones as large as 2-decanone and in combination with A112G enabled activity toward longer branched ketones including 5-methyl-2-octanone.

Full text of DOI:10.1002/cctc.201902364

Accepted Article
Title: Separate sets of mutations enhance activity and substrate scope
of amine dehydrogenase
Authors: Robert D Franklin, Conner J Mount, Bettina R Bommarius,
and Andreas Sebastian Bommarius
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.201902364
Link to VoR: http://dx.doi.org/10.1002/cctc.201902364
A Journal of
www.chemcatchem.org

10.1002/cctc.201902364
ChemCatChem
COMMUNICATION
Separate sets of mutations enhance activity and substrate scope
of amine dehydrogenase
Robert D. Franklin,‡[b] Conner J. Mount,‡[c] Bettina R. Bommarius,[b] and Andreas S. Bommarius*[a]
[a]
Prof. A. S. Bommarius
School of Chemical and Biomolecular Engineering
School of Chemistry and Biochemistry
Parker H. Petit Institute for Bioengineering and Bioscience
Engineered Biosystems Building, 950 Atlantic Drive, Atlanta, GA 30332-2000, USA
E-mail: andreas.bommarius@chbe.gatech.edu
R. D. Franklin, Dr. B. R. Bommarius
School of Chemical and Biomolecular Engineering
Parker H. Petit Institute for Bioengineering and Bioscience
Georgia Institute of Technology
[b]
[c]
Engineered Biosystems Building, 950 Atlantic Drive, Atlanta, GA 30332-2000, USA
C. J. Mount
School of Chemistry and Biochemistry
Parker H. Petit Institute for Bioengineering and Bioscience
Georgia Institute of Technology
Engineered Biosystems Building, 950 Atlantic Drive, Atlanta, GA 30332-2000, USA
‡
These authors contributed equally to the reported work
Supporting information for this article is given via a link at the end of the document.
Abstract: Mutations were introduced into the leucine amine
dehydrogenase (L-AmDH) derived from G. stearothermophilus
leucine dehydrogenase (LeuDH) with the goals of increased activity
and expanded substrate acceptance. A triple variant (L-AmDH-TV)
including D32A, F101S, and C290V showed an average of 2.5-fold
higher activity toward aliphatic ketones and an 8.0 °C increase in
melting temperature. L-AmDH-TV did not show significant changes in
relative activity for different substrates. In contrast, L39A, L39G,
A112G, and T133G in varied combinations added to L-AmDH-TV
changed the shape of the substrate binding pocket. L-AmDH-TV was
not active on ketones larger than 2-hexanone. L39A and L39G
enabled activity for straight-chain ketones as large as 2-decanone and
in combination with A112G enabled activity toward longer branched
ketones including 5-methyl-2-octanone.
acid dehydrogenase scaffolds[6] or have been identified from
natural sources.[7] Others have introduced new mutations to
increase activity for new substrates by altering the size of the
ketone binding pocket.[6h] Still more work has been performed on
enzyme immobilization,[6e, 8] rate law determination,[9] whole-cell
biocatalysis,[6d, 10] and multi-enzyme cascades[8b, 11] to enable the
use of amine dehydrogenases in the large-scale synthesis of
chiral amines.
For decades, chiral amine compounds have proven to be key
intermediates for blockbuster drugs. Many of the top selling small
molecule drugs today contain chiral amine groups.[1] Recently,
there has been a growing interest in utilizing biocatalytic reductive
amination of prochiral ketones to produce chiral amines.
Biocatalytic production of chiral amines offers distinct advantages
over classical heterogeneous catalysis. Key enzyme families in
the field include ω-transaminases,[2] reductive aminases,[3] and
amine dehydrogenases.[4] Amine dehydrogenases (AmDHs), first
developed in 2012[5], catalyze the reductive amination of prochiral
ketones to form chiral primary amines with the addition of
aqueous ammonia. The reaction is dependent on a hydride
transfer from NADH to form NAD+ [Scheme 1]. To-date, all the
AmDHs engineered from amino acid dehydrogenases produce
(R)-amines. The first AmDH, called leucine amine dehydrogenase
(L-AmDH) was developed through targeted libraries of mutations
in the active site of the leucine dehydrogenase (LeuDH) from
Geobacillus stearothermophilus. After multiple rounds of
mutations, activity toward keto acids was removed, and new
activity toward ketones was obtained. Since 2012, multiple groups
have produced similar amine dehydrogenases from other amino
Scheme 1. Reductive amination of prochiral ketones to form chiral amines with
an amine dehydrogenase
In the present work, we sought to improve the published L-AmDH
(referred to here as the base case) through two separate sets of
mutations. The first group was selected to increase activity and
stability without necessarily impacting substrate specificity.
Position V291 was shown to be important for substrate binding in
LeuDH.[12] In the base case enzyme, this residue was mutated
from valine to cysteine. It was speculated that mutating back to
valine [Figure 1A] could positively impact activity. Two residues
1
This article is protected by copyright. All rights reserved.

10.1002/cctc.201902364
ChemCatChem
COMMUNICATION
farther away from the active site were also shown to be promising
candidates in a 2004 patent on mutations in LeuDHs.[13] Based on
the reported results, F101S and D32A were incorporated into the
L-AmDH base case.
In the second group of mutations, three residues close to the
substrate side chain were identified as potential targets for
altering the L-AmDH substrate scope. Kataoka and Tanizawa
found that for G. stearothermophilus LeuDH, L39K and A112G
mutations both caused major changes in the relative specific
activities for different amino acid and keto acid substrates.[12]
Recently, Chen et al. showed mutations homologous to A112G
and T133G enhanced activity toward larger and bulkier ketones
in multiple engineered L-AmDHs.[6h] Replacing the residues at
positions L39, A112, and T133 with smaller residues was
hypothesized to enable activity toward larger ketones in our L-
AmDH as well. The positions of the proposed mutations relative
to the substrate binding pocket can be seen in Figure 1B.
Combinations of the six proposed mutations were introduced into
L-AmDH and the resulting variants were screened for activity on
a variety of straight and branched methyl ketones of different
sizes. Other mutations were tested but showed poor results (see
Supplementary Information).
Figure 1. A homology model (constructed with SWISS-MODEL[14]) of the engineered L-AmDH sequence mapped onto a B. stearothermophilus LeuDH cryo-EM
structure (PDB ID: 6ACF).[15] Because this structure is in the apo form, the holo-structure the PheDH from Rhodococcus sp. (PDB ID: 1C1D)[16] was aligned to the
homology model, the phenylalanine substrate and NAD+ cofactor position from this alignment can give a good idea of how the substrate and cofactor would be
positioned in L-AmDH. Subfigure A shows the positions of D32, F101, and C290 relative to the active site. Subfigure B shows the positions of L39, A112, and T133
in the active site.
The L-AmDH mutations D32A, F101S, and C290V were
generated sequentially using overlap extension PCR, followed by
cloning and protein expression. Specific activity toward various
aliphatic ketones was measured in 1 mL batch reactions by
following the decrease in concentration of NADH with UV-visible
spectrophotometry and is recorded in milliunits per milligram of
enzyme, with one unit defined as the amount of enzyme required
to catalyze the conversion of one micromole of substrate in one
minute. The results of these assays are recorded in Table 1.
Across the six substrates for which activity could be quantified,
the combined D32A/F101S/C290V mutant (L-AmDH-TV) showed
an average of a 2.5-fold increase in activity compared to the base
case enzyme. The largest improvement was seen for 4-methyl-2-
pentanone where L-AmDH-TV had 3.4-fold higher activity than
the base case. Interestingly, the relative activity of the enzyme
toward different substrates did not change due to the mutations,
but rather the mutations increased activity for all substrates in
roughly equal proportions. Additionally, differential scanning
fluorimetry (DSF) experiments showed the melting point of L-
AmDH-TV to be 73.6 °C, an increase of 8.0 °C compared to L-
AmDH at 65.6 °C.
Table 1. Activity enhancement of L-AmDH toward aliphatic ketones and the
generation of L-AmDH-TV
Specific Activity (mU/mg)
D32A/F101S/
C290V (“TV”)
Substrate
L-AmDH
D32A
D32A/F101S
1
2
3
4
5
6
n.d.[a]
87.9
430.5
144.8
n.d.
n.d.
n.d.
n.d.
61.4
888.3
172.2
n.d.
181.7
1363.9
278.6
n.d.
225.5
1303.6
266.4
n.d.
n.d.
n.d.
n.d.
n.d.
2
This article is protected by copyright. All rights reserved.

10.1002/cctc.201902364
ChemCatChem
COMMUNICATION
Based on results reported for a different set of engineered L-
AmDHs,[6h] A112G and T133G were investigated as additional
sites for binding pocket expansion. The TV/A112G variant shows
new activity for 2-heptanone, and a new optimal substrate length
of six carbons rather than five, as was found for TV [Table 2]. The
further addition of T133G increased activity for 2-pentanone
through 2-heptanone, while enabling new activity for 2-octanone
and 2-nonanone. Finally, the addition of L39A to TV/A112G and
TV/A112G/T133G shifted the optimal substrate length to 7
carbons and enabled activity toward 2-decanone. The addition of
T133G to TV/L39A/A112G did not have the same synergistic
effects as were found when TV/A112G, suggesting that the
benefits of L39A and T133G are not additive.
7
n.d.
n.d.
n.d.
n.d.
8
n.d.
n.d.
n.d.
n.d.
9
453.4
531.1
549.6
945.3
1406.5
1083.3
959.2
1661.4
1145.1
1146.5
1808.5
1056.2
10
11
Reaction conditions: 0.4-1.0 µM enzyme, 20 mM substrate, 200 µM NADH, 4M
NH4Cl/NH4OH, pH 9.6, 25 °C, 1 mL reaction volume. [a] n.d. denotes an activity
level below the limit of detection.
Sequence alignments of the active sites of multiple amino acid
dehydrogenases suggested L-AmDH position L39 as a potential
target for expansion of the substrate binding pocket to
accommodate larger substrates.[12] Leucine is conserved at this
position for leucine dehydrogenases, but a lysine at the
homologous position in glutamate dehydrogenases is likely
responsible for binding the acid group on the glutamate substrate
sidechain. The residue is positioned with its side chain pointing in
toward the substrate binding site. The L-AmDH-TV scaffold
presented in the previous section formed the basis for further
mutations. L-AmDH-TV/L39A and L39G mutants were produced.
Figure 2 shows a comparison of specific activities toward straight-
chain ketones of varying lengths. For L-AmDH-TV, 2-pentanone
shows the highest activity at 1.31 U/mg, while zero activity was
observed for ketones longer than 2-hexanone. When the leucine
at position L39 is replaced with the smaller alanine, activity toward
2-butanone and 2-pentanone are greatly reduced, while activity
was more than doubled for 2-hexanone. More interestingly,
TV/L39A showed activity for ketones as large as 2-decanone, with
the highest activity found for 2-heptanone at 644 mU/mg. When
the residue was further mutated to glycine, the n-ketone with the
highest activity shifted to 2-nonanone. While TV/L49G showed
decreased activity for 2-butanone through 2-heptanone, activity
was increased compared to TV/L39A for 2-octanone through 2-
decanone.
Table 2. Effect of A112G and T133G on activity toward n-ketones
Specific Activity (mU/mg)
TV/
L39A/
A112G/
T133G
TV/
A112G/
T133G
TV/
L39A/
A112G
TV/
A112G
Substrate
TV
1
2
3
4
5
6
7
8
n.d.[a]
225.5
1303.6
266.4
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
80.5
166.6
31.9
n.d.
565.1
777.3
596.9
159.2
145.9
n.d.
35.1
19.0
308.9
446.1
268.6
268.8
140.5
177.7
431.5
237.4
171.2
92.9
n.d.
n.d.
n.d.
n.d.
n.d.
Reaction conditions: 0.4-1.0 µM enzyme, 20 mM substrate, 200 µM NADH, 4M
NH4Cl/NH4OH, pH 9.6, 25 °C, 1 mL reaction volume.
The reported variants were also tested for their ability to convert
longer branched ketones, 12 and 13 [Table 3]. The combination
of L39A and A112G allowed for the conversion of these bulky
ketones, while each mutation on its own was insufficient. TV/L39G
also showed low but measurable activity toward 13 at 39.4 mU/mg.
As seen with the straight-chain ketones, T133G does not increase
activity for larger substrates when in combination with L39A.
Table 3. Activity of L-AmDH variants toward long branched ketones
Activity Toward 12
(mU/mg)
Activity Toward 13
(mU/mg)
Enzyme Variant
TV
n.d.[a]
n.d.
n.d.
TV/L39A
n.d.
TV/L39G
n.d.
39.4
n.d.
TV/A112G
n.d.
Figure 2. Expansion of the L-AmDH binding pocket to accommodate ketones
of larger sizes due to mutations at position L39. Reaction conditions: 0.4-1.0 µM
enzyme, 20 mM substrate, 200 µM NADH, 4M NH4Cl/NH4OH, pH 9.6, 25 °C, 1
mL reaction volume.
TV/A112G/T133G
TV/L39A/A112G
TV/L39A/A112G/T133G
n.d.
n.d.
258.8
177.9
186.4
161.5
3
This article is protected by copyright. All rights reserved.

10.1002/cctc.201902364
ChemCatChem
COMMUNICATION
[1]
[2]
[3]
D. Ghislieri, N. J. Turner, Topics in Catalysis 2014, 57,
284-300.
M. D. Patil, G. Grogan, A. Bommarius, H. Yun, Catalysts
2018, 8, 254.
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.
M. D. Patil, G. Grogan, A. Bommarius, H. Yun, ACS Catal.
2018, 8, 10985-11015.
M. J. Abrahamson, E. Vazquez-Figueroa, N. B. Woodall,
J. C. Moore, A. S. Bommarius, Angew. Chem., Int. Ed.
2012, 51, 3969-3972.
a) M. J. Abrahamson, J. W. Wong, A. S. Bommarius, Adv.
Synth. Catal. 2013, 355, 1780-1786; b) B. R. Bommarius,
M. Schurmann, A. S. Bommarius, Chem. Commun.
(Cambridge, U. K.) 2014, 50, 14953-14955; c) L. J. Ye, H.
H. Toh, Y. Yang, J. P. Adams, R. Snajdrova, Z. Li, ACS
Catal. 2015, 5, 1119-1122; d) A. Pushpanath, E. Siirola, A.
Bornadel, D. Woodlock, U. Schell, ACS Catal. 2017, 7,
3204-3209; e) J. Lowe, A. A. Ingram, H. Groger, Bioorg.
Med. Chem. 2018, 26, 1387-1392; f) V. Tseliou, T. Knaus,
M. F. Masman, M. L. Corrado, F. G. Mutti, Nat. Commun.
2019, 10, 3717; g) S. K. Au, B. R. Bommarius, A. S.
Bommarius, ACS Catal. 2014, 4, 4021-4026; h) F. F.
Chen, G. W. Zheng, L. Liu, H. Li, Q. Chen, F. L. Li, C. X.
Li, J. H. Xu, ACS Catal. 2018, 8, 2622-2628.
Reaction conditions: 0.4-1.0 µM enzyme, 20 mM substrate, 200 µM NADH, 4M
NH4Cl/NH4OH, pH 9.6, 25 °C, 1 mL reaction volume.
To further demonstrate the improvements to the applicability of L-
AmDH to convert larger ketones, selected substrates were
converted at a 50 mL preparative scale for 24 hours with both L-
AmDH and TV/L39A. As shown in Table 4, relative conversion,
(measured after derivatization with benzoyl chloride[7a]) between
the two enzymes was in line with their relative specific activities.
Additionally, the already exquisite enantioselectivity of the L-
AmDH[5] was not impacted by the mutations, as measured after
diastereomeric derivatization[7b] [Figures S4, S5, S8]. Aggregation
occurred in all samples occurred over the course of the reaction
and likely limited overall conversion. In the future, this could be
mitigated through immobilization.
[4]
[5]
[6]
Table 4. Conversion values for preparative-scale reactions
Conversion after 24 hours
[7]
a) O. Mayol, K. Bastard, L. Beloti, A. Frese, J. P.
Turkenburg, J. L. Petit, A. Mariage, A. Debard, V. Pellouin,
A. Perret, V. de Berardinis, A. Zaparucha, G. Grogan, C.
Vergne-Vaxelaire, Nat. Catal. 2019, 2, 324-333; b) O.
Mayol, S. David, E. Darii, A. Debard, A. Mariage, V.
Pellouin, J. L. Petit, M. Salanoubat, V. de Berardinis, A.
Zaparucha, C. Vergne-Vaxelaire, Catal. Sci. Technol.
2016, 6, 7421-7428.
Substrate
Concentration
L-AmDH
TV/L39A
3
5
7
10 mM
10 mM
2 mM
65.4 ± 2.2%
1.7 ± 0.1%
n.d.
48.5 ± 1.6%
56.3 ± 4.4%
51.9 ± 0.5%
[8]
[9]
a) A. A. Caparco, A. S. Bommarius, J. A. Champion,
AIChE J. 2018, 64, 2934-2946; b) W. Bohmer, T. Knaus,
F. G. Mutti, ChemCatChem 2018, 10, 731-735; c) H. Ren,
Y. Zhang, J. Su, P. Lin, B. Wang, B. Fang, S. Wang, J.
Biotechnol. 2017, 241, 33-41; d) J. Liu, B. Q. W. Pang, J.
P. Adams, R. Snajdrova, Z. Li, ChemCatChem 2017, 9,
425-431.
R. D. Franklin, J. A. Whitley, J. M. Robbins, A. S.
Bommarius, Chem. Eng. J. 2019, 369, 634-640.
J. Liu, Z. Li, Biotechnol. Bioeng. 2019, 116, 536-542.
a) S. Yoon, M. D. Patil, S. Sarak, H. Jeon, G. H. Kim, T. P.
Khobragade, S. Sung, H. Yun, ChemCatChem 2019, 11,
1898-1902; b) F. G. Mutti, T. Knaus, N. S. Scrutton, M.
Breuer, N. J. Turner, Science 2015, 349, 1525-1529; c) H.
Wang, Y.-C. Zheng, F.-F. Chen, J.-H. Xu, H.-L. Yu,
ChemCatChem 2020, DOI 10.1002/cctc.201902253; d) F.
F. Chen, Y. H. Zhang, Z. J. Zhang, L. Liu, J. P. Wu, J. H.
Xu, G. W. Zheng, J. Org. Chem. 2019, 84, 14987-14993.
K. Kataoka, K. Tanizawa, J. Mol. Catal. B: Enzym. 2003,
23, 299-309.
Reaction conditions: 2 mg each of AmDH and cbFDH, 1 mM NAD+, ketone
concentration as listed, 2M NH4COOH/NH4OH, pH 8.5, 21 °C, 50 mL reaction
volume, rotating at 20 rpm for 24 hours.
In conclusion, we found two sets of mutations to improve L-AmDH
which result in either increased specific activity and
thermodynamic stability, or in an altered substrate specificity
towards longer or branched methylketones. While the first set of
mutations acts synergistically to increase L-AmDH activity and
stability, the second set of three mutations for change substrate
specificity does not. Instead, that second set enables picking a
desired specificity trait, such as a long side chain of a
methylketone or branched alkylmethylketone, with a specific
mutation. Thus, the current work is an important step towards a
differentiated family of sufficiently stable amine dehydrogenases.
[10]
[11]
[12]
[13]
[14]
J. Rozzell, L. Hua, M. Mayhew, S. Novick (Biocatalytics
Inc), US20040115691A1, 2004.
a) S. Bienert, A. Waterhouse, T. A. de Beer, G. Tauriello,
G. Studer, L. Bordoli, T. Schwede, Nucleic Acids Res,
2017, 45, D313-D319; b) A. Waterhouse, M. Bertoni, S.
Bienert, G. Studer, G. Tauriello, R. Gumienny, F. T. Heer,
T. A. P. de Beer, C. Rempfer, L. Bordoli, R. Lepore, T.
Schwede, Nucleic Acids Res. 2018, 46, W296-W303.
H. Yamaguchi, A. Kamegawa, K. Nakata, T. Kashiwagi, T.
Mizukoshi, Y. Fujiyoshi, K. Tani, J. Struct. Biol. 2019, 205,
11-21.
Acknowledgements
Funding for the reported work by the National Science Foundation
through NSF CBET grant 1512848 and NSF I/UCRC grant
1540017 to the Center for Pharmaceutical Development is
gratefully acknowledged. We also wish to acknowledge the
Biopolymer Characterization (BPC) core facility at the Parker H.
Petit Institute for Bioengineering and Bioscience at the Georgia
Institute of Technology for the use of its shared equipment,
services and expertise.
[15]
[16]
N. M. Brunhuber, J. B. Thoden, J. S. Blanchard, J. L.
Vanhooke, Biochemistry 2000, 39, 9174-9187.
Keywords: amine dehydrogenases • biocatalysis • enzyme
catalysis • protein engineering • reductive amination
4
This article is protected by copyright. All rights reserved.

10.1002/cctc.201902364
ChemCatChem
COMMUNICATION
Entry for the Table of Contents
Mutations at six locations in an engineered leucine amine dehydrogenase (L-AmDH) greatly improved activity, stability, and substrate
scope. A triple variant enzyme showed 1.8 to 3.4-fold increase in specific activity compared to parent and an 8.0 °C increase in
melting temperature. A second set of mutations reshaped the binding pocket to enable new activity for straight-chain ketones up to
ten carbons long, compared to six for the parent.
5
This article is protected by copyright. All rights reserved.