An Acyl Transfer Reaction Catalyzed by an Epimerase MarH

Article abstract of DOI:10.1021/acscatal.5b02198

MarH, a small protein (129 amino acids) belonging to the cupin superfamily, was previously characterized as an epimerase involved in the (2S,3S)-β-methyltryptophan formation in the maremycin biosynthesis. Here, MarH was discovered to act as an acyltransferase that can catalyze the 3-O-acylation of chloramphenicol. Furthermore, MarH can catalyze N-acylation of deacylated chloramphenicol analogue thereby activating them for 3-O-acylation. By systematic site-directed mutagenesis, H64 was revealed as a potential catalytic base that deprotonates the acyl acceptor substrate. Nucleophilic attack at the carbonyl carbon of the acyl donor then gives the acylation product.

Full text of DOI:10.1021/acscatal.5b02198

Letter
pubs.acs.org/acscatalysis
An Acyl Transfer Reaction Catalyzed by an Epimerase MarH
Mo Han,†,‡ Haixing Yin,§,‡ Yi Zou,† Nelson L. Brock,† Tingting Huang,† Zixin Deng,† Yiwen Chu,*,§
and Shuangjun Lin*,†
†State Key Laboratory of Microbial Metabolism, Joint International Laboratory on Metabolic & Developmental Sciences, School of
Life Sciences & Biotechnology, Shanghai Jiao Tong University, 800 Dongchuan Road, Shanghai 200240, China
§Sichuan Industrial Institute of Antibiotics, Chengdu University, 168 Huaguan Road, Chengdu 610052, China
S
* Supporting Information
ABSTRACT: MarH, a small protein (129 amino acids)
belonging to the cupin superfamily, was previously charac-
terized as an epimerase involved in the (2S,3S)-β-methyl-
tryptophan formation in the maremycin biosynthesis. Here,
MarH was discovered to act as an acyltransferase that can
catalyze the 3-O-acylation of chloramphenicol. Furthermore,
MarH can catalyze N-acylation of deacylated chloramphenicol analogue thereby activating them for 3-O-acylation. By systematic
site-directed mutagenesis, H64 was revealed as a potential catalytic base that deprotonates the acyl acceptor substrate.
Nucleophilic attack at the carbonyl carbon of the acyl donor then gives the acylation product.
KEYWORDS: cupin protein, MarH, multifunctional enzyme, acyltransferase, chloramphenicol acetyltransferase
expression plasmid pET28a-marH (hereafter named as strain-I)
ost enzymes are special proteins produced by organisms
Mthat can catalyze diverse chemical reactions in life and the control plasmid pET28a (named as strain-II) were,
respectively, cultured in LB medium containing both
kanamycin and Cml (Supporting Information, SI). After
fermentation, the supernatants were analyzed by high-perform-
ance liquid chromatography−mass spectrometry (HPLC-MS).
Three peaks showed the same UV absorption as Cml, but their
relative intensities differed in the strains. In particular, 4 was the
dominant product in strain-I while 2 was the major product in
strain-II (Figure 1A). High-resolution mass (HR-MS) analyses
suggested that they were the acetylation products of 1 (Table
S2 and Figure S14). The three compounds were purified, and
comprehensive NMR analysis confirmed 2−4 to be 3-O-acetyl
Cml, 1-O-acetyl Cml and 1,3-O-diacetyl Cml, respectively
(Figure S15−20). Thus, we presumed that MarH might exhibit
the CAT activity in addition to its epimerase activity.
process. Due to their versatility, chemo-, regio-, enantioselec-
tivity, and mild reaction conditions,1 they are explored as
catalysts in many chemical industries.2 The selectivity and
specificity of enzymes are evidenced by the induced-fit toward
their substrates: the binding of substrates induces a conforma-
tional change of an enzyme. The conformational change
operates like a molecular switch to recognize correct
substrates.3 A growing number of enzymes with substrate
ambiguity4 and multifunctional enzymes with different
domains, such as fatty acid synthetases,5 polyketide synthases,6
nonribosomal peptide synthetases,7 and many others8 have
been reported. In addition, single-domain enzymes with
catalytic promiscuity are also known,9 but most of them
catalyze similar or sequential reactions.10
To verify this speculation, MarH was overexpressed in E. coli
BL21(DE3) instead of E. coli BL21(DE3)pLysS to avoid
contamination of CATs. The in vitro CAT activity assays (SI)
proved that the purified MarH efficiently catalyzed the acetyl
transfer from acetyl coenzyme A (Ac-CoA, 5) to Cml,
generating both monoacetylated Cml 2 and 3, and diacetylated
Cml 4 (Figure 1B).
The time-dependent assays showed that initially only 2 was
produced, followed by production of 3, and that 4 appeared last
(Figure S1). This result implicated that 3 was nonenzymatically
generated from 2, which was confirmed by incubation of 2 in
Tris-HCl buffer at 30 °C (Figure S2A). The nonenzymatic
transformation may take place via a six-membered cyclic
MarH (AHJ60977) belongs to the cupin superfamily that is
widely found in prokaryotic and eukaryotic organisms.11 This
versatile superfamily includes metal-independent epimerases12
and metal-dependent enzymes such as dioxygenases,13
decarboxylases,14 hydroxylases,15 and more.16 Most cupin
proteins catalyze a single reaction. An exception is the barley
germin that can act as both oxalate oxidase and manganese
superoxide dismutase.17 Previously, we have shown that MarH
catalyzes the epimerization of (R)-β-methylindolepyruvate in
the biosynthesis of (2S,3S)-β-methyltryptophan from L-
tryptophan.18 Here, we report that MarH can also catalyze
the acylation of chloramphenicol (Cml, 1) even though it
shows no homology to Cml acetyltransferases (CATs).
When MarH was overexpressed in E. coli BL21(DE3) pLysS
in Luria−Bertani (LB) medium supplemented with Cml, we
accidently discovered that MarH could catalyze the acetylation
of Cml. The strain E. coli BL21(DE3)pLysS harboring the
Received: September 30, 2015
Revised: November 26, 2015
Published: December 30, 2015
© 2015 American Chemical Society
788
DOI: 10.1021/acscatal.5b02198
ACS Catal. 2016, 6, 788−792

ACS Catalysis
Letter
Figure 1. Activity and substrate specificity of MarH for acyl-donors. (A) HPLC profiles of fermentation extracts of strain-I (i) and strain-II (ii). (B)
HPLC profiles of acylation of Cml catalyzed by MarH with different acyl-CoAs. (i) Cml with boiled MarH; (ii) 5 incubated with MarH for 15 min,
and (iii) 5, (iv) 6, (v) 7 as donors incubated with MarH for 2 h. (C) Proposed acylation of Cml catalyzed by MarH.
Figure 2. (A) Catalytic efficiency of MarH and its mutants. (B) Proposed mechanism of MarH-catalyzed Cml acetylation.
transition state (Figure S2B).19 These results augured that
MarH could not directly transfer the acetyl group to the 1-
hydroxy group of Cml and also indicated that the diacetylated
product 4 was derived from the acetylation of 3, not 2 (Figure
1C). To confirm this hypothesis, we conducted two parallel
reactions with either purified 2, or the mixture of 2 and 3 at
equilibrium composition. In the reaction with 2, 4 was barely
formed until the transformation of 2 to 3 reached the
equilibrium at 20 min. After this time point, 4 steadily
increased, along with the gradual decrease of 2, while 3
remained unchanged (Figure S3A). In contrast, in the reaction
with the mixture of 2 and 3, 4 was produced from the
beginning of the reaction and steadily increased along with the
decrease of 3, but 2 almost remained unchanged until the
reaction proceeded for 30 min (Figure S3B). These results
established regiospecificity of MarH for the 3-O-acetylation of
Cml. Overall, the acetylation of Cml catalyzed by MarH is
similar to that of CATs (Figure 1C).20
residues H62, H64, and E68 of the motif 1 (HxHxxxE) as well
as P102, G104, and H107 of the motif 2 (PxGxxH) that are
characteristics of the cupin superfamily proteins (Figure S4).21
Five additional amino acid residues F31, F38, P75, C120, and
G122 were shared by MarH and canonical CATs (Figure S4).
In order to locate the key residues for acyl transfer activity of
MarH, site-directed mutagenesis of these amino acid residues
was performed, and the resulting mutant proteins of MarH
were overexpressed and purified (Figure 2A and S5). All
mutant proteins were confirmed to retain the overall protein
structure of MarH by circular dichroism (CD) spectra (Figure
S6).
The enzymatic efficiencies of MarH and the mutants were
determined (SI). The activities of the F31A and F38A mutants
increased in about 50%. It was speculated that the replacement
of phenylalanine by alanine provided more space for Cml
binding. The mutant H64A exhibited significantly reduced
CAT activity, whereas other mutants retained comparative
activities to the wild type MarH (Figure 2A). Taken together,
H64 might act as a catalytic base in the acetylation of Cml
catalyzed by MarH, which is similar to CATs and other
acetyltransferase.22 The 3-hydroxy group of Cml might be
deprotonated by the imidazole of the H64 residue. Nucleophilic
CATs catalyze the acetylation of Cml using a conserved
histidine residue as the catalytic base to activate the acetyl-
acceptor. MarH belongs to the cupin superfamily without any
homology to CATs. Sequence alignment of MarH and several
cupin superfamily proteins revealed the conserved amino acid
789
DOI: 10.1021/acscatal.5b02198
ACS Catal. 2016, 6, 788−792

ACS Catalysis
Letter
attack at the carbonyl carbon of acetyl-CoA completes the
acetylation of Cml (Figure 2B). To confirm this speculation,
CAT activity of MarH was tested at different pH values (SI).
MarH showed relatively higher CAT activity under alkaline
conditions that are beneficial for deprotonation (Figure S7). In
addition, four acidic residues (D36, E47, D87, and E92), which
might deprotonate H64, were also mutated to Ala, but the
resultant mutation had no effect on CAT activity (Figure 2A).
Intriguingly, H107 from motif 2 had been confirmed as a
catalytic base for the epimerase activity,18 whereas H64 from
motif 1 was demonstrated now to act as the catalytic base for
the CAT activity. Structural studies to reveal why MarH can
catalyze both the epimerization and acylation are in progress.
In our previous work, three homologues of MarH including
StnK3 (AFW04574), ACPL_6197 (AEV87082), and
Cwoe_4835 (ADB53248) could catalyze the same epimeriza-
tion with similar efficiency as MarH.18 All three proteins were
tested whether they could also catalyze the acetylation of Cml.
Both the monoacetylated and diacetylated products of Cml (2,
3, and 4) were detected in the typical in vitro assays (Figure
S8). Furthermore, the steady-state kinetic parameters for these
four proteins were acquired with varying Ac-CoA concen-
trations (SI and Figure S9) and summarized in Table 1. MarH
and StnK3 showed similar catalytic efficiency to the canonical
CATs.20,23
Acyltransferases are widely distributed in nature. They not
only play important roles in life processes24 but also participate
in the biosynthesis of numerous secondary metabolites.25
Moreover, several acyltransferases have been successfully
applied for the synthesis of pharmaceuticals.26 For medicinal
applications, acylated derivatives often have improved biological
activities, stabilities, or availabilities, compared to the original
compounds.27 Because no acyltransferase has ever been
reported from the cupin superfamily, we were curious about
whether MarH can recognize other acyl donor substrates.
Butyryl-CoA (6), isobutyryl-CoA (7), and six other acyl CoAs
(8−13) were tested as putative acyl donors in typical assays
(Figure S10). HPLC and HR-MS analysis showed that only
monobutyrylated and monoisobutyrylated products of Cml
(14−17) were formed in the reactions with 6 and 7 as acyl
donors (Figure 1B and Table S2). The structure of 14 was
confirmed by NMR analysis (Figure S21−22). The position of
acylation is easily deducible from the chemical shift of the 3-H
that was found at δH = 3.74−3.60 ppm for Cml28 compared to
δH = 4.32−4.14 ppm for the corresponding butyrylated
compound (Table S4). 16 was identified only by HR-MS due
to limited amounts of substrate 7 (Table S2 and Figure S14).
Comparative kinetic parameters were obtained for 6 as a donor
substrate (Km = 185.4
29.50 μM and kcat = 1.56
0.08
min−1). It is notable that no dibutyrylated or diisobutyrylated
Cml could be detected even by increasing the enzyme
concentration and extending the reaction time, probably due
to the increased steric hindrance of the butyryl and isobutyryl
groups. No product was observable in the reactions with other
putative acyl donors, implying that the variability of acyl donors
recognized by MarH is limited. However, compared to
canonical CATs,23 MarH still presents a relatively flexible
substrate specificity for acyl donors.
Table 1. Kinetic Parameters for MarH and Its Homologues
enzyme
MarH
Km (μM)
kcat (min−1
)
kcat/Km (mM−1 min−1
)
88.39 5.99
88.72 2.68
72.91 8.81
104.9 20.40
1.26 0.04
2.00 0.01
0.67 0.02
0.46 0.02
15.12
22.54
9.19
StnK3
ACPL_6197
Cwoe_4835
4.39
Figure 3. Acceptor substrate specificity of MarH. (A) Structures of the putative acceptors; (B) HPLC profiles of acetylation catalyzed by MarH with
Ac-CoA and putative acceptors. (i) 18 and (iii) 19 incubated with boiled MarH for 3 h; (ii) 18 and (iv) 19 incubated with MarH for 3 h. (C)
Proposed acetylation of 18 catalyzed by MarH.
790
DOI: 10.1021/acscatal.5b02198
ACS Catal. 2016, 6, 788−792

ACS Catalysis
Letter
To gain insights into the acceptor substrate specificity,
chemo- and stereoselectivity of MarH, compounds with similar
structures as Cml (18−21) were tested as putative acetyl
acceptors in the in vitro assays (Figure 3A). HPLC and HR-MS
analysis revealed that only 18 was acetylated by MarH,
generating four acetylated products (Figure 3B and Table
S3). To elucidate the structures of these products, 23, 24, and
25 were purified from a biotransformation of 18 using E. coli
BL21(DE3) containing pET28a-marH. NMR analysis (SI)
revealed that 23 and 24 were indeed diacetylated products
while 25 was the triacetylated product of 18 (Figure S23−S37).
As before, the acylation position was assigned on the basis of
chemical shift variations of the adjacent hydrogen atoms and
furthermore confirmed by HMBC data. Notably, the amino
group of all three products was acetylated. However, only trace
amounts of 22 could be obtained that were insufficient for
NMR analysis. Because the mono-O-acetylated products of
Cml were always observed as an interconvertible mixture and
no further peaks with the same HR-MS as 22 were found, we
concluded that 22 might be the N-acetylated product of 18.
The low amounts of 22 accumulated in the biotransformation
of 18 by MarH implicated that MarH first catalyzed the
acetylation of the amino group of 18, further initiating the 3-O-
acetylation of 18 that proceeded with a faster reaction rate. To
verify this assumption, we carried out time-dependent
enzymatic assays of MarH with 18 and Ac-CoA as substrates
(Figure S11). This time-course analysis not only confirmed this
hypothesis but also showed that 23 could spontaneously react
to 24 that was further acetylated to 25 (Figure 3C and S12).
MarH could catalyze the acetyl transfer to 18, but not to 19,
exhibiting the stringent (1R,2R)-selectivity. Because 20 and 21
could not be acetylated by MarH, we speculated that the nitro
group is essential for substrate binding to MarH. Taken
together, MarH was proven to be a robust enzyme that not
only functions as an epimerase18 but also acts as an O- and N-
acyltransferase.
Finally, to test whether marH could function as a Cml
resistance gene in vivo, the anti-Cml bioassay was carried out
using E. coli BL21(DE3) strains containing the expression
plasmid pET28a-marH and the control plasmid pET28a as
indicator strains, respectively. The strain E. coli BL21(DE3)/
pET28a cannot grow on LB-agar medium containing Cml
while the strain E. coli BL21(DE3)/pET28a-marH showed an
anti-Cml effect (Figure S13). These findings indicated that
MarH can be used not only as a robust acyltransferase but also
as a Cml resistance gene in vivo.
In summary, we have discovered a group of small cupin
proteins (129−132 aa residues), which can catalyze not only
epimerization of β-methylindolepyruvate but also esterification
and amide bond formation of Cml by acting as a dual
acyltransferase. Studies of the reaction kinetics indicated that
these small cupin proteins had comparative catalytic efficiency
to canonical CATs. By systematic point mutation study, H64
was discovered to be the key active site and a similar catalytic
model was proposed to that of CATs. Overall, MarH showed
multifunctional properties, flexible substrate specificity, and
strict regio- and enantioselectivity, providing a potential to be
explored as an efficient biocatalyst used for organic synthesis.
Experimental details, NMR, Q-TOF, and other supple-
mental data (PDF)
AUTHOR INFORMATION
■
Corresponding Authors
*E-mail: linsj@sjtu.edu.cn.
*E-mail: chuyiwen@cdu.edu.cn.
Author Contributions
‡These authors contributed equally (M.H. and H.Y.).
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Professor Wen Liu at Shanghai Institute of Organic
Chemistry CAS for gifting acyl CoAs. This work was financially
supported by the National Natural Science Foundation of
China (31425001 and 21372154 for S.L.; 31121064 for Z.D.)
and the grants from MOE of China and the Leopoldina
Fellowship Program (German National Academy of Sciences
Leopoldina, LPDS 2013-12 for N.L.B.).
REFERENCES
■
(1) Barbayianni, E.; Kokotos, G. ChemCatChem 2012, 4, 592−608.
Kelleher, N. L.; Belshaw, P. J.; Walsh, C. T. J. Am. Chem. Soc. 1998,
120, 9716−9717.
(2) Thomas, S. M.; DiCosimo, R.; Nagarajan, V. Trends Biotechnol.
2002, 20, 238−242.
(3) Hatzakis, N. S. Biophys. Chem. 2014, 186, 46−54. Johnson, K. A.
J. Biol. Chem. 2008, 283, 26297−26301.
(4) Krebs, G.; Hugonet, L.; Sutherland, J. D. Angew. Chem., Int. Ed.
2006, 45, 301−305.
(5) Maier, T.; Leibundgut, M.; Ban, N. Science 2008, 321, 1315−
1322.
(6) Dutta, S.; Whicher, J. R.; Hansen, D. A.; Hale, W. A.; Chemler, J.
A.; Congdon, G. R.; Narayan, A. R.; Hakansson, K.; Sherman, D. H.;
Smith, J. L.; Skiniotis, G. Nature 2014, 510, 512−517.
(7) Tanovic, A.; Samel, S. A.; Essen, L. O.; Marahiel, M. A. Science
2008, 321, 659−663.
(8) Winzer, T.; Kern, M.; King, A. J.; Larson, T. R.; Teodor, R. I.;
Donninger, S. L.; Li, Y.; Dowle, A. A.; Cartwright, J.; Bates, R.;
Ashford, D.; Thomas, J.; Walker, C.; Bowser, T. A.; Graham, I. A.
Science 2015, 349, 309−312. Farrow, S. C.; Hagel, J. M.; Beaudoin, G.
A.; Burns, D. C.; Facchini, P. J. Nat. Chem. Biol. 2015, 11, 728−732.
Drake, E. J.; Nicolai, D. A.; Gulick, A. M. Chem. Biol. 2006, 13, 409−
419. Campbell, M. G.; Smith, B. C.; Potter, C. S.; Carragher, B.;
Marletta, M. A. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, E3614−E3623.
(9) Babtie, A. C.; Bandyopadhyay, S.; Olguin, L. F.; Hollfelder, F.
Angew. Chem., Int. Ed. 2009, 48, 3692−3694. Bornscheuer, U. T.;
Kazlauskas, R. J. Angew. Chem., Int. Ed. 2004, 43, 6032−6040.
(10) DeLoache, W. C.; Russ, Z. N.; Narcross, L.; Gonzales, A. M.;
Martin, V. J.; Dueber, J. E. Nat. Chem. Biol. 2015, 11, 465−471.
Rodionov, R. N.; Jarzebska, N.; Weiss, N.; Lentz, S. R. Trends
Pharmacol. Sci. 2014, 35, 575−582. Charlier, D.; Kholti, A.; Huysveld,
N.; Gigot, D.; Maes, D.; Thia-Toong, T. L.; Glansdorff, N. J. Mol. Biol.
2000, 302, 409−424.
(11) Dunwell, J. M.; Purvis, A.; Khuri, S. Phytochemistry 2004, 65, 7−
17.
(12) Jakimowicz, P.; Tello, M.; Meyers, C. L.; Walsh, C. T.; Buttner,
M. J.; Field, R. A.; Lawson, D. M. Proteins: Struct., Funct., Genet. 2006,
63, 261−265.
(13) Fetzner, S. Appl. Environ. Microbiol. 2012, 78, 2505−2514.
Stipanuk, M. H.; Simmons, C. R.; Karplus, P. A.; Dominy, J. E., Jr.
Amino Acids 2011, 41, 91−102. Hangasky, J. A.; Taabazuing, C. Y.;
Valliere, M. A.; Knapp, M. J. Metallomics 2013, 5, 287−301.
(14) Anand, R.; Dorrestein, P. C.; Kinsland, C.; Begley, T. P.; Ealick,
S. E. Biochemistry 2002, 41, 7659−7669.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acscatal.5b02198.
■
S
791
DOI: 10.1021/acscatal.5b02198
ACS Catal. 2016, 6, 788−792

ACS Catalysis
Letter
(15) Hewitson, K. S.; McNeill, L. A.; Riordan, M. V.; Tian, Y. M.;
Bullock, A. N.; Welford, R. W.; Elkins, J. M.; Oldham, N. J.;
Bhattacharya, S.; Gleadle, J. M.; Ratcliffe, P. J.; Pugh, C. W.; Schofield,
C. J. J. Biol. Chem. 2002, 277, 26351−26355.
(16) Dunwell, J. M.; Culham, A.; Carter, C. E.; Sosa-Aguirre, C. R.;
Goodenough, P. W. Trends Biochem. Sci. 2001, 26, 740−746. Uberto,
R.; Moomaw, E. W. PLoS One 2013, 8, e74477.
(17) Woo, E. J.; Dunwell, J. M.; Goodenough, P. W.; Marvier, A. C.;
Pickersgill, R. W. Nat. Struct. Biol. 2000, 7, 1036−1040.
(18) Zou, Y.; Fang, Q.; Yin, H.; Liang, Z.; Kong, D.; Bai, L.; Deng, Z.;
Lin, S. Angew. Chem., Int. Ed. 2013, 52, 12951−12955.
(19) Lin, S.; Van Lanen, S. G.; Shen, B. Proc. Natl. Acad. Sci. U. S. A.
2009, 106, 4183−4188.
(20) Nakagawa, Y.; Nitahara, Y.; Miyamura, S. Antimicrob. Agents
Chemother. 1979, 16, 719−723.
(21) Shutov, A. D.; Kakhovskaya, I. A. Mol. Biol. (Moscow, Russ. Fed.,
Engl. Ed.) 2011, 45, 529−535.
(22) Leslie, A. G.; Moody, P. C.; Shaw, W. V. Proc. Natl. Acad. Sci. U.
S. A. 1988, 85, 4133−4137. Ma, X.; Koepke, J.; Panjikar, S.; Fritzsch,
G.; Stockigt, J. J. Biol. Chem. 2005, 280, 13576−13583.
(23) Shaw, W. V. Methods Enzymol. 1975, 43, 737−755.
(24) Grunstein, M. Nature 1997, 389, 349−352.
(25) Demain, A. L. Appl. Microbiol. Biotechnol. 1999, 52, 455−463.
Bennett, R. N.; Wallsgrove, R. M. New Phytol. 1994, 127, 617−633.
(26) Gonzalez-Sabin, J.; Moran-Ramallal, R.; Rebolledo, F. Chem. Soc.
Rev. 2011, 40, 5321−5335.
(27) Xie, X.; Tang, Y. Appl. Environ. Microbiol. 2007, 73, 2054−2060.
Nevarez, D. M.; Mengistu, Y. A.; Nawarathne, I. N.; Walker, K. D. J.
Am. Chem. Soc. 2009, 131, 5994−6002. Yan, Y.; Chen, J.; Zhang, L.;
Zheng, Q.; Han, Y.; Zhang, H.; Zhang, D.; Awakawa, T.; Abe, I.; Liu,
W. Angew. Chem., Int. Ed. 2013, 52, 12308−12312.
(28) Groß, F.; Lewis, E. A.; Piraee, M.; van Pee
White, R. L. Bioorg. Med. Chem. Lett. 2002, 12, 283−286.
́
, K.-H.; Vining, L. C.;
792
DOI: 10.1021/acscatal.5b02198
ACS Catal. 2016, 6, 788−792