Functional Models of the Nickel Pincer Nucleotide Cofactor of Lactate Racemase

Article abstract of DOI:10.1002/anie.201910490

A novel nickel pincer cofactor was recently discovered in lactate racemase. Reported here are three synthetic nickel pincer complexes that are both structural and functional models of the pincer cofactor in lactate racemase. DFT computations suggest the ipso-carbon atom of the pyridinium pincer ligands act as a hydride acceptor for lactate isomerization, whereas an organometallic pathway involving nickel-mediated β-hydride elimination is less favored.

Full text of DOI:10.1002/anie.201910490

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201910490
German Edition: DOI: 10.1002/ange.201910490
Reaction mechanisms
Functional Models of the Nickel Pincer Nucleotide Cofactor of Lactate
Racemase
Renyi Shi, Matthew D. Wodrich, Hui-Jie Pan, Farzaneh Fadaei Tirani, and Xile Hu*
Abstract: A novel nickel pincer cofactor was recently discov-
ered in lactate racemase. Reported here are three synthetic
nickel pincer complexes that are both structural and functional
models of the pincer cofactor in lactate racemase. DFT
computations suggest the ipso-carbon atom of the pyridinium
pincer ligands act as a hydride acceptor for lactate isomer-
ization, whereas an organometallic pathway involving nickel-
mediated b-hydride elimination is less favored.
with thiocarboxylates, we now present models that are not
only more structurally similar to the NPN cofactor, but also
exhibit lactate racemization activity. These species represent
the first functional models of LarA. DFT computations
suggest that isomerization occurs in a pathway similar to path
I (Figure 1b).
Lactate racemase (LarA) converts l-lactic acid into d-lactic
acid, a conversion which is important for the assembly of cell
walls in many microorganisms (Figure 1a).[1] LarA is also the
first enzyme known to contain a pincer metal complex as
a cofactor. This cofactor, called nickel pincer nucleotide
(NPN), is a NiII complex of an SCS pincer ligand based on
a pyridinium-3,5-bisthiocarboxylic acid mononucleotide (Fig-
ure 1a).[2] The NPN cofactor is incorporated into the active
site of LarA through a covalent thioamide bond to Lys186.[3]
While several mechanisms have been proposed for LarA
based on DFT computations,[4] a recent combined experi-
mental and computational study suggested two reaction
pathways.[5] The most probable pathway involves a reversible
proton-coupled hydride transfer (PCHT) to the ipso-carbon
atom of the NPN cofactor (Figure 1b, path I). An alternative
pathway involving dissociation of His200 and reversible
hydride transfer to Ni (path II) seems less favorable but
cannot be excluded. Small-molecule models of enzymes
provide an important tool to probe the mechanism underlying
enzyme reactions. Although metal pincer complexes are
ubiquitous in synthetic chemistry,[6] the structure of the NPN
cofactor is unique. Models of NPN have only begun to be
developed.[7] Recently, we reported a NiII model supported by
a pyridinium-3,5-bisthioamide SCS pincer ligand (1; Fig-
ure 1c).[7] This model mediated dehydrogenation of an
alcohol, but the reaction was irreversible such that no
racemization activity was observed. By replacing thioamides
[*] Dr. R. Shi, Dr. M. D. Wodrich, Dr. H.-J. Pan, Dr. F. F. Tirani,
Prof. X. L. Hu
Laboratory of Inorganic Synthesis and Catalysis, Institute of
Chemical Sciences and Engineering, ꢀcole Polytechnique Fꢁdꢁrale de
Lausanne (EPFL), ISIC-LSCI, BCH 3305,Lausanne 1015 (Switzerland)
E-mail: xile.hu@epfl.ch
Homepage: http://lsci.epfl.ch
Dr. M. D. Wodrich
Laboratory for Computational Molecular Design, Institute of Chem-
ical Science and Engineering, Ecole Polytechnique Fꢁdꢁrale de
Lausanne (EPFL), Lausanne 1015 (Switzerland)
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.201910490.
Figure 1. a) Lactate racemase and its nickel pincer nucleotide (NPN)
cofactor. L has not yet been identified. b) Proposed mechanisms of
lactate racemase. c) Pincer models of LarA.
Angew. Chem. Int. Ed. 2019, 58, 1 – 5
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

Angewandte
Communications
Chemie
Figure 3. Solid-state molecular structures of 5 (left), 6 (middle), and 7
(right).[15] The thermal ellipsoids are displayed at 50% probability. For
5 and 6, only the complex anion is shown (the PPh4 cation is omitted).
For 7, the imidazole molecule is disordered and is modelled by two
partially occupied molecules, only one of which is shown. All hydrogen
atoms are omitted for clarity.
Table S1). Two equivalents of an external base were used to
mimic the basic protein residues involved in the deprotona-
tion of the carboxylic acid and alcohol groups. We found 1,8-
diazabicyclo[5.4.0]undec-7-ene (DBU) to be an appropriate
base. To follow the isomerization by HPLC, the product was
converted into a benzyl ester. The reaction was conducted at
608C (Table 1). Gratifyingly, racemization of l-lactic acid (8)
Figure 2. Synthesis of the pincer proligands and the corresponding Ni
complexes. DMF=N,N-dimethylformamide.
The synthesis of the models is shown in Figure 2. Treat-
ment of pyridine-3,5-dicarboxylic acid (2) with methyl iodide
gave the pyridinium salt 3, which was converted into the
proligand, 1-methyl-3,5-dithiocarboxypyridin-1-ium bromide
(4Br) through consecutive reactions with SOCl2, NaSH, and
HBr (Figure 2a). Reaction of the proligand with Ni-
(OAc)2·4H2O in toluene at 1008C gave the nickel pincer
bromide complex 5, which was isolated as a PPh4 salt by
reaction with PPh4Br. The analogous nickel chloride complex
6 was obtained using HCl and PPh4Cl in the synthesis.
Treatment of 5 with 20 equivalents of imidazole yielded the
neutral imidazole complex 7 (Figure 2b). A similar reaction
of 6 and imidazole, however, yielded a mixture of 6 and 7,
probably as a result of an equilibrium.
Table 1: Summary of the effects of catalysts on the racemization of l-
lactic acid.[a]
Entry
Catalyst
t [h]
ee [%]
1
2
3
4
5
6
5 (10 mol%)
none
6 (10 mol%)
7 (10 mol%)
5 (10 mol%)
NiCl2 (10 mol%)
12
12
12
12
36
12
75
94
76
72
66
93
The UV-vis spectra of the complexes 5–7 exhibit strong
absorptions in the 430–450 nm region with a weaker shoulder
between 550–600 nm (see Figure S2 in the Supporting Infor-
mation). Similar features were observed for the NPN
cofactor[3] and LarA,[5] indicative of a similar electronic
structure for the Ni centers. In the IR spectra of 5–7, a peak at
[a] The ee value was determined as benzyl 2-hydroxypropanoate by HPLC
using an OD-H column (iPrOH/hexane=5:95). The optimized reaction
conditions were l-lactic acid (0.1 mmol), 5 (10 mol%), DBU (2.0 equiv),
and DMA (2.0 mL), 12 h, 608C. After cooling to room temperature, BnBr
(2 equiv) was added to the mixture. The mixture reacted for another 6 h,
508C. Bn=benzyl, DBU=diazabicyclo[5.4.0]undec-7-ene, DMA=N,N-
dimethylacetamide.
1615 cmÀ1 was assigned to the C O vibration (see Figures S5–
=
S7). The solid-state molecular structures of 5–7 were deter-
mined by X-ray crystallography, which confirmed the for-
mation of Ni/SCS pincer complexes (Figure 3). The Ni center
retains the expected square-planar coordination environ-
ment. The Ni–C and Ni–S bond distances were comparable
between all three complexes (see Table S2). Judging from the
was observed using 10 mol% of 5 in the presence of DBU,
giving a mixture of l- and d-lactic acids with 75% ee after
12 hours (entry 1). With either NEt3 or imidazole as the base,
no racemization occurred (see Table S1). Using 5 alone
without base resulted in no racemization (see Table S1). With
DBU alone, only an insignificant amount of isomerization
was observed under the same reaction conditions (Table 1,
entry 2). The chloride complex 6 and the imidazole complex 7
had similar activities to 5 (entries 3 and 4). When the reaction
time was increased to 36 hours, the ee value decreased to 66%
(entry 5). Monitoring the reaction by UV/Vis spectroscopy
(see Figure S1) revealed that the catalyst remained stable
during the reaction. Thus, the incomplete reaction after either
À
À
C O bond length (about 1.22 ꢀ), the C O moiety is best
À
represented by a double bond. Consequently, the C S moiety
is best represented by a single bond, making the S donor
anionic. Compared to the previous model 1,[7] the Ni S bonds
À
in 5–7 are slightly longer (by about 0.05 ꢀ) as a result of
replacing thioamides with thiocarboylates in the pincer
ligands (see Table S2). This elongation might be due to
a trans effect by the two anionic thiolate ligands.
The activities of the models 5–7 in the racemization of
lactic acid, the natural substrate of LarA, was probed (see
2
www.angewandte.org
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2019, 58, 1 – 5
These are not the final page numbers!

Angewandte
Communications
Chemie
12 or 36 hours likely arises from a slow reaction rather than
hypothesize that the anionic thiocarboxylate S donors in 5–7
catalyst decomposition. Assuming the racemization gives an
equal amount of enantiomers, the turnover number is
estimated to be around 3 after 36 hours. When NiCl2 was
used as the catalyst, the ee value was 93%, similar to the
background reaction (Table 1, entries 2 and 6). When 1 was
used as the catalyst, its decomposition was observed upon
addition of the substrates and base (see Figure S1).
confer a higher stability for the nickel complexes compared to
To probe the reaction mechanism leading to lactate
racemization, we turned to DFT computations at the
PBE0[9]-dDSC[10]/TZ2P//M06[11]/def2-SVP theoretical level
in implicit DMA solvent using the COSMO-RS solvation
model[12] as implemented in Gaussian09[13] (M06 computa-
tions) and ADF[14] (PBE0-dDsC computations, see full
computational details in the Supporting Information). We
examined two potential mechanistic pathways (Figure 4)
using computational models of both 5 and 7. The first
pathway (top, Figure 4) involves the ipso-carbon atom of the
pyridinium pincer ligands of the catalysts acting as a hydride
acceptor, while a second pathway (bottom) involves forma-
tion of an alkoxide intermediate followed by b-hydride
elimination to form a nickel hydride species. The free energies
associated with the two pathways for both 5 and 7 are shown
in Figure 5. Clearly, invoking the ipso-carbon atom of the
pyrdinium pincer ligands represents the most favorable
energetic pathway for both 5 and 7, where the rate-determin-
ing transition states lie 9.2 and 15.7 kcalmolÀ1 below the
alkoxide pathway for 5 and 7, respectively. Our computations
also indicate the free-energy profile of 7 to be slightly less
costly than 5, although the overall energetic cost of 30–
35 kcalmolÀ1 would still render the catalysis as being rather
sluggish, consistent with experimental results. We attempted
to prepare C4 by reacting 5 with a strong hydride donor,
NaBH3CN. However, the reaction led to a complex mixture of
unidentified products.
The difference in reactivity between the previous model
1 and current models (5–7; Table 1) might be attributed to
their different stabilities. The complex 1 is unstable in the
racemization medium, whereas 5–7 retain stability. We
Figure 5. Computed free-energy profiles (see full computational details
in the Supporting Informatino for description of theoretical levels) of
potential mechanistic pathways for hydride transfer from lactic acid
involving 5 and 7. The reverse reactions result in lactate racemization.
Figure 4. Potential mechanistic pathways for lactate racemization; only the forward reaction is shown.
Angew. Chem. Int. Ed. 2019, 58, 1 – 5 ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

Angewandte
Communications
Chemie
[6] a) D. Morales-Morales, C. M. Jensen, The Chemistry of Pincer
Compounds, Elsevier, Amsterdam, 2007; b) K. J. Szabꢁ, O. F.
Wendt, Pincer and Pincer-Type Complexes: Applications in
Organic Synthesis and Catalysis, Wiley-VCH, Weinheim, 2014;
c) G. van Koten, D. Milstein, Organometallic Pincer Chemistry,
Springer, Berlin Heidelberg, 2013; d) C. Gunanathan, D. Mil-
stein, Chem. Rev. 2014, 114, 12024 – 12087; e) R. Shi, Z. Zhang,
X. Hu, Acc. Chem. Res. 2019, 52, 1471 – 1483.
the neutral thioamide S donors found in 1. The lower
activities of 5–7 compared to LarA points to the important
catalytic function afforded by an optimized protein environ-
ment.
In summary, Ni pincer complexes that are structurally
similar to the NPN cofactor of LarA have been synthesized
and characterized. These models reproduce, for the first time,
the function of LarA. DFT computations suggest that the
ipso-carbon atom of the pyridinium pincer ligands serve as
a hydride acceptor in these functional models. Overall, our
study provides insights into the structure and mechanism of
LarA.
[7] T. Xu, M. D. Wodrich, R. Scopelliti, C. Corminboeuf, X. Hu,
Proc. Natl. Acad. Sci. USA 2017, 114, 1242 – 1245.
[8] a) I. Buslov, J. Becouse, S. Mazza, M. Montandon-Clerc, X. Hu,
Angew. Chem. Int. Ed. 2015, 54, 14523 – 14526; Angew. Chem.
2015, 127, 14731 – 14734; b) I. Buslov, S. C. Keller, X. Hu, Org.
Lett. 2016, 18, 1928 – 1931; c) J. V. Obligacion, P. J. Chirik, Nat.
Rev. Chem. 2018, 2, 15 – 34.
[9] a) J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1996, 77,
3865 – 3868; b) C. Adamo, V. Barone, J. Chem. Phys. 1999, 110,
6158 – 6170.
Acknowledgements
[10] a) S. N. Steinmann, C. Corminboeuf, J. Chem. Theory Comput.
2010, 6, 1990 – 2001; b) S. N. Steinmann, C. Corminboeuf,
Chimia 2011, 65, 240 – 244; c) S. N. Steinmann, C. Corminboeuf,
J. Chem. Phys. 2011, 134, 044117; d) S. N. Steinmann, C.
Corminboeuf, J. Chem. Theory Comput. 2011, 7, 3567 – 3577.
[11] a) Y. Zhao, D. G. Truhlar, Acc. Chem. Res. 2008, 41, 157 – 167;
b) Y. Zhao, D. G. Truhlar, Theor. Chem. Acc. 2008, 120, 215 –
241.
This work is supported by the Swiss National Science
Foundation. M.D.W. acknowledges C. Corminboeuf (EPFL)
and the Laboratory for Computational Molecular Design
(EPFL) for providing computing resources.
Conflict of interest
[12] A. Klamt, WIREs Comput. Mol. Sci. 2011, 1, 699 – 709.
[13] Gaussian09, RevisionD.01, M. J. Frisch, G. W. Trucks, H. B.
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G.
Scalmani, V. Barone, G. A. Petersson, H. Nakatsuji, X. Li, M.
Caricato, A. Marenich, J. Bloino, B. G. Janesko, R. Gomperts, B.
Mennucci, H. P. Hratchian, J. V. Ortiz, A. F. Izmaylov, J. L.
Sonnenberg, D. Williams-Young, F. Ding, F. Lipparini, F. Egidi, J.
Goings, B. Peng, A. Petrone, T. Henderson, D. Ranasinghe, V. G.
Zakrzewski, J. Gao, N. Rega, G. Zheng, W. Liang, M. Hada, M.
Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T.
Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, K.
Throssell, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M.
Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov,
T. Keith, R. Kobayashi, J. Normand, K. Raghavachari, A.
Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, J. M.
Millam, M. Klene, C. Adamo, R. Cammi, J. W. Ochterski, R. L.
Martin, K. Morokuma, O. Farkas, J. B. Foresman, and D. J. Fox,
Gaussian, Inc., Wallingford CT, 2016.
[14] a) C. Fonseca Guerra, J. G. Snijders, G. te Velde, E. J. Baerends,
Theor. Chem. Acc. 1998, 99, 391 – 403; b) G. te Velde, F. M.
Bickelhaupt, S. J. A. van Gisbergen, C. Fonseca Guerra, E. J.
Baerends, J. G. Snijders, T. Ziegler, J. Comput. Chem. 2001, 22,
931 – 967.
[15] CCDC 1919062 (5), 1919063 (6), and 1919064 (7) contain the
supplementary crystallographic data for this paper. These data
are provided free of charge by The Cambridge Crystallographic
Data Centre
The authors declare no conflict of interest.
Keywords: cofactors · enzymes · nickel · pincer ligands ·
reaction mechanisms
[1] a) E. L. Tatum, W. H. Peterson, E. B. Fred, Biochem. J. 1936, 30,
1892 – 1897; b) J. L. Boer, S. B. Mulrooney, R. P. Hausinger,
Arch. Biochem. Biophys. 2014, 544, 142 – 152.
[2] a) B. Desguin, P. Goffin, E. Viaene, M. Kleerebezem, V. Martin-
Diaconescu, M. J. Maroney, J.-P. Declercq, P. Soumillion, P. Hols,
Nat. Commun. 2014, 5, 3615; b) B. Desguin, T. Zhang, P.
Soumillion, P. Hols, J. Hu, R. P. Hausinger, Science 2015, 349,
66 – 69; c) B. Desguin, P. Soumillion, P. Hols, R. P. Hausinger,
Proc. Natl. Acad. Sci. USA 2016, 113, 5598 – 5603; d) T. Xu, G.
Bauer, X. Hu, ChemBioChem 2016, 17, 31 – 32; e) M. Fellner, B.
Desguin, R. P. Hausinger, J. Hu, Proc. Natl. Acad. Sci. USA 2017,
114, 9074 – 9079; f) R. P. Hausinger, B. Desguin, M. Fellner, J. A.
Rankin, J. Hu, Curr. Opin. Chem. Biol. 2018, 47, 18 – 23.
[3] B. Desguin, M. Fellner, O. Riant, J. Hu, R. P. Hausinger, P. Hols,
P. Soumillion, J. Biol. Chem. 2018, 293, 12303 – 12317.
[4] a) B. Qiu, X. Yang, Chem. Commun. 2017, 53, 11410 – 11413;
b) B. Wang, S. Shaik, Angew. Chem. Int. Ed. 2017, 56, 10098 –
10102; Angew. Chem. 2017, 129, 10232 – 10236; c) M.-J. Yu, S.-L.
Chen, Chem. Eur. J. 2017, 23, 7545 – 7557; d) X. Zhang, L. W.
Chung, Chem. Eur. J. 2017, 23, 3623 – 3630.
[5] J. A. Rankin, R. C. Mauban, M. Fellner, B. Desguin, J.
McCracken, J. Hu, S. A. Varganov, R. P. Hausinger, Biochem-
istry 2018, 57, 3244 – 3251.
Manuscript received: August 16, 2019
Accepted manuscript online: September 19, 2019
Version of record online: && &&, &&&&
4
www.angewandte.org
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2019, 58, 1 – 5
These are not the final page numbers!

Angewandte
Communications
Chemie
Communications
Reaction mechanisms
Pinch-hit: Reported here are the first
functional mimics of the pincer cofactor
of lactate racemase (LarA). The mimics
are both structural and functional models
of the pincer cofactor in lactate racemase.
DFT computations suggest the ipso-
carbon atom of the pyridinium pincer
ligands acts as a hydride acceptor for
lactate isomerization, whereas an organ-
ometallic pathway involving nickel-medi-
ated b-hydride elimination is less favored
R. Shi, M. D. Wodrich, H.-J. Pan,
F. F. Tirani, X. L. Hu*
&&&& — &&&&
Functional Models of the Nickel Pincer
Nucleotide Cofactor of Lactate Racemase
Angew. Chem. Int. Ed. 2019, 58, 1 – 5
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!