R ES E A RC H | R E PO R TS

6. L. A. van der Veen, P. C. J. Kamer, P. W. N. M. van Leeuwen, METALLOPROTEINS

Angew. Chem. Int. Ed. 38, 336–338 (1999).
7. S. Pereira, M. Srebnik, J. Am. Chem. Soc. 118, 909–910
(1996).
8. N. Sakai, K. Nozaki, H. Takaya, J. Chem. Soc. Chem. Commun.
(4): 395–396 (1994).
A tethered niacin-derived pincer
9. T. Gadzikwa, R. Bellini, H. L. Dekker, J. N. H. Reek, J. Am.
Chem. Soc. 134, 2860–2863 (2012).
10. T. C. Nugent, Chiral Amine Synthesis: Methods, Developments
complex with a nickel-carbon bond in
and Applications (Wiley VCH, Weinheim, Germany, 2010).
11. U. T. Bornscheuer, R. J. Kazlauskas, Hydrolases in Organic
Synthesis (Wiley VCH, Weinheim, Germany, 2006).
lactate racemase
12. M. T. Robak, M. A. Herbage, J. A. Ellman, Chem. Rev. 110,
3600–3740 (2010). Benoît Desguin,1 Tuo Zhang,2 Patrice Soumillion,3 Pascal Hols,3
13. T. E. Müller, K. C. Hultzsch, M. Yus, F. Foubelo, M. Tada, Chem.
Rev. 108, 3795–3892 (2008).
Jian Hu,2,4* Robert P. Hausinger1,2*
14. P. W. Roesky, T. E. Müller, Angew. Chem. Int. Ed. 42, 2708–2710
(2003).
Lactic acid racemization is involved in lactate metabolism and cell wall assembly of
15. M. R. Gagné et al., Organometallics 11, 2003–2005
(1992). many microorganisms. Lactate racemase (Lar) requires nickel, but the nickel-binding
16. S. Hong, T. J. Marks, Acc. Chem. Res. 37, 673–686 (2004). site and the role of three accessory proteins required for its activation remain enigmatic.
17. M. Kawatsura, J. F. Hartwig, J. Am. Chem. Soc. 122, We combined mass spectrometry and x-ray crystallography to show that Lar from
9546–9547 (2000).
Lactobacillus plantarum possesses an organometallic nickel-containing prosthetic group.
18. Z. Zhang, S. D. Lee, R. A. Widenhoefer, J. Am. Chem. Soc. 131,

Downloaded from www.sciencemag.org on July 2, 2015

5372–5373 (2009). A nicotinic acid mononucleotide derivative is tethered to Lys184 and forms a tridentate
19. A. L. Reznichenko, H. N. Nguyen, K. C. Hultzsch, Angew. Chem. pincer complex that coordinates nickel through one metal-carbon and two metal-sulfur
Int. Ed. 49, 8984–8987 (2010). bonds, with His200 as another ligand. Although similar complexes have been previously
20. R. Dorta, P. Egli, F. Zürcher, A. Togni, J. Am. Chem. Soc. 119,
synthesized, there was no prior evidence for the existence of pincer cofactors in enzymes.
10857–10858 (1997).
21. J. S. Zhou, J. F. Hartwig, J. Am. Chem. Soc. 130, 12220–12221 The wide distribution of the accessory proteins without Lar suggests that it may play a
(2008). role in other enzymes.
22. S. Zhu, N. Niljianskul, S. L. Buchwald, J. Am. Chem. Soc. 135,
15746–15749 (2013).

L
23. S. Zhu, S. L. Buchwald, J. Am. Chem. Soc. 136, 15913–15916
(2014). actate dehydrogenases are nearly ubiqui- racemase (LarA) was initially stymied by the
24. S.-L. Shi, S. L. Buchwald, Nat. Chem. 7, 38–44 (2015). tous enzymes that reduce pyruvate to form loss of activity and Ni content in the first few
25. N. Niljianskul, S. Zhu, S. L. Buchwald, Angew. Chem. Int. Ed. 54, either D- or L-lactic acid. The D-isomer is hours after LarA protein purification from re-
1638–1641 (2015).
26. Y. Mikki, K. Hirano, T. Satoh, M. Miura, Angew. Chem. Int. Ed.
specifically incorporated into the cell wall combinant Lactococcus lactis cells (2). To cir-
52, 10830–10834 (2013). of certain bacteria, where it can contribute cumvent this problem, we included sulfite in
27. R. P. Rucker, A. M. Whittaker, H. Dang, G. Lalic, J. Am. Chem. to vancomycin resistance (1). Lactate racemase the purification. This compound was identified
Soc. 134, 6571–6574 (2012). (Lar), which interconverts the D- and L-isomers as a potent mixed inhibitor (fig. S1) and was
28. A. M. Berman, J. S. Johnson, J. Am. Chem. Soc. 126, 5680–5681
(2004).
of lactic acid, was recently shown to use nickel shown both to stabilize LarA activity and delay
29. D. Lee, J. Yun, Tetrahedron Lett. 45, 5415–5417 (2004). as a cofactor (2), bringing the number of known Ni loss (fig. S2, A and B). Sulfite most probably
30. D. H. Apella, Y. Moritani, R. Shintani, E. M. Ferreira, S. L. Buchwald, Ni-dependent enzymes to nine (3). The structure protects LarA from oxidation because incuba-
J. Am. Chem. Soc. 121, 9473–9474 (1999). of the recombinant apoprotein from Thermo- tion of the enzyme in microaerobic environments
31. C. Deutsch, N. Krause, Chem. Rev. 108, 2916–2927 anaerobacterium thermosaccharolyticum (LarATt) had an equivalent effect (fig. S2C), whereas
(2008).
32. B. H. Lipshutz, K. Noson, W. Chrisman, A. Lower, J. Am. Chem.
lacks the cofactor, leaving open questions of the other reducing agents [dithiothreitol and tris
Soc. 125, 8779–8789 (2003). metallocenter composition and coordinating lig- (2-carboxyethyl)phosphine] did not protect Lar
33. G. M. Whitesides, E. R. Stedronsky, C. P. Casey, J. San Filippo Jr., ands. Lar activity requires the products of the activity. LarA purified in the presence of sulfite
J. Am. Chem. Soc. 92, 1426–1427 (1970). structural gene (larA) and three accessory genes showed greater Ni content and higher specific
34. K. R. Fandrick et al., J. Am. Chem. Soc. 133, 10332–10335 (2011).
35. A. de Meijere, F. Diederich, Metal-Catalyzed Cross-Coupling
(larB, larC, and larE). LarE is a putative adenosine activity (table S1), and its absorbance spectrum
Reactions (Wiley VCH, Weinheim, 2004). 5´-triphosphate (ATP)–using enzyme of the PP- revealed electronic transitions at 375, 440, and
36. T. G. Gant, J. Med. Chem. 57, 3595–3611 (2014). loop superfamily that, when isolated from cells 550 nm (fig. S2D). These features were absent
37. J. Won, D. Noh, J. Yun, J. Y. Lee, J. Phys. Chem. A 114, also producing LarB and LarC, is able to activate in LarA apoprotein obtained from cells not ex-
12112–12115 (2010).
Lar apoprotein in vitro (2). The requirement for pressing the accessory proteins (fig. S2D), sug-
38. Y. Zhao, T. M. Swager, J. Am. Chem. Soc. 137, 3221–3224 (2015).
these accessory proteins and the absence of an gesting that the chromophore is associated with
ACKN OW LEDG MEN TS identified Ni-binding site in the reported LarATt the metallocenter.
We are grateful to the National Institutes of Health (grant GM structure were compatible with the presence of Protein molecular mass differences observed
58160 to S.L.B.) and the University of Pittsburgh (P.L.) for financial a Ni-binding cofactor in Lar. by using LC-ESI-MS further demonstrate the
support. The content is solely the responsibility of the authors To interrogate the nature of Ni-binding in presence of an additional compound in LarA
and does not necessarily represent the official views of the
National Institutes of Health. We acknowledge Y. Zhao (Swager Lar, we used liquid chromatography–electrospray holoprotein. Whereas the mass of LarA apo-
group) for helpful discussions. We thank Y.-M. Wang and ionization–mass spectrometry (LC-ESI-MS) to protein corresponded to its predicted 47,393
M. T. Pirnot for assistance with the preparation of the manuscript. identify a prosthetic group on the enzyme. We daltons (for protein missing its N-terminal Met)
Calculations were performed at the Center for Simulation and also determined the x-ray structure of LarA (Fig. 1A), the LarA holoprotein purified in the
Modeling at the University of Pittsburgh and the Extreme Science
and Engineering Discovery Environment (XSEDE) supported by containing the cofactor. Biochemical charac- presence of 50 mM sulfite was 451 T 1 daltons
the National Science Foundation. terization of Lactobacillus plantarum lactate greater (Fig. 1B). Because of the LC purifica-
1
tion prior to injection, the protein only retained
Department of Microbiology and Molecular Genetics,
SUPPLEMENTARY MATERIALS
Michigan State University, East Lansing, MI 48824, USA.
strongly attached compounds, and hence, sul-
www.sciencemag.org/content/349/6243/62/suppl/DC1 2
Department of Biochemistry and Molecular Biology, Michigan fite or other loosely bound molecules would
Materials and Methods State University, East Lansing, MI 48824, USA. 3Institute of Life not account for this mass increase. When in-
Supplementary Text Sciences, Université Catholique de Louvain, B-1348 Louvain-la- cubated for 1 day at room temperature in buffer
Figs. S1 to S6 Neuve, Belgium. 4Department of Chemistry, Michigan State
University, East Lansing, MI 48824, USA.
containing only 0.1 mM sulfite, a substantial
17 April 2015; accepted 19 May 2015 *Corresponding author. E-mail: [email protected] (JH); portion of LarA holoprotein decreased in mass
10.1126/science.aab3753 [email protected] (RPH) by 30 T 1 and 56 T 1 daltons (Fig. 1C), indicating

66 3 JULY 2015 • VOL 349 ISSUE 6243 sciencemag.org SCIENCE

 RE S EAR CH | R E P O R T S

a degradation of the compound that is accom- Fig. 1. Posttranslational modification of LarA

panied by loss of Lar activity (fig. S2A). by a nicotinic acid derivative. (A to E) The mass
To further investigate whether the additional of (A) LarA apoprotein, (B) LarA holoprotein, (C)
mass in LarA is associated with the enzyme ac- LarA holoprotein after incubating 1 day in buffer
tivity and Ni content, we used mutagenesis to containing only 0.1 mM of sulfite, (D) K184A variant
substitute eight of the most conserved and pre- of LarA, and (E) LarA holoprotein purified from
sumed active site residues (table S1). Four larA cells grown in the presence of d4-nicotinic acid.
mutants (encoding the H108A, H172A, H200A, The ordinate represents the abundance relative to
and K298A LarA variants) essentially abolished the highest peak in percent. In parentheses are the
Lar activity in cell extracts, whereas another four mass differences compared with the theoretical
mutants (encoding D72A, R75A, K184A, and Q295A mass of the LarA apoprotein lacking its N-terminal
LarA variants) showed residual activity in vivo, with methionine.
the purified R75A and Q295A variants retaining
in vitro activity (table S1). (Single-letter abbrevia-
tions for the amino acid residues are as follows:
A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H,
His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro;
Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and
Y, Tyr. In the variants, other amino acids were
substituted at certain locations; for example,
H108A indicates that histidine at position 108
was replaced by alanine.) All variants that
could be purified (R75A, H172A, K184A, and
Q295A) exhibited reduced Ni content, but only
K184A LarA completely lacked both Ni and
activity—which is equivalent to apoprotein (table
S1). Moreover, the mass of K184A LarA corre-
sponded to that of the variant apoprotein (Fig.
1D). This result implies that K184 is involved in
the binding of the unknown compound that seems
to be required for Lar activity and Ni coordina-
tion. The reduced Lar activity and Ni content in
the other variants may be explained by the de-
stabilization of the active site leading to rapid
inactivation and Ni loss.

Fig. 2. LC-ESI-MS analysis of modified tryptic peptides of

LarA in positive ionization mode and isotopic model of the
1610-dalton species. (A) Doubly positively charged tryptic pep-
tides of LarA were analyzed by means of LC-ESI-MS with samples
derived from cells grown in the presence of (top) nicotinic acid
and (bottom) d4-nicotinic acid. (B) Isotopic model of the doubly
negatively charged 1610-dalton species (left) calculated with the
Micromass MassLynx (Waters, Milford, MA) software package
and (right) as observed in negative ionization mode.

SCIENCE sciencemag.org 3 JULY 2015 • VOL 349 ISSUE 6243 67

R ES E A RC H | R E PO R TS

Table 1. Mass spectrometric analysis of nicotinic acid–containing tryptic fragments of LarA.

Observed mass (OM)* OM – peptide (–1161.676)† OM – peptide – NAMN + H2O (–1478.706)† D‡ Theoretical mass of D

1610.570 T 0.014 448.894 T 0.014 131.864 T 0.014 CS Ni –2H

2
2+ +
131.864
............................................................................................................................................................................................................................................................................................................................................
1586.637 T 0.020 424.961 T 0.020 107.931 T 0.020 CO2S2 107.934
............................................................................................................................................................................................................................................................................................................................................
1570.646 T 0.009 408.970 T 0.009 91.940 T 0.009 COS2 91.939
............................................................................................................................................................................................................................................................................................................................................
1554.658 T 0.010 392.982 T 0.010 75.952 T 0.010 CS2/CO2S 75.944/75.962
............................................................................................................................................................................................................................................................................................................................................

*Observed mass (OM, in daltons) is the average of masses observed by using positive and negative ionization modes (Fig. 2A and fig. S4). The “plus/minus” represents
the 95% confidence interval (normal distribution, n = 4 m/z peaks). †Mass of the tryptic peptide (residues 184 to 194) or tryptic peptide plus NAMN subtracted
from OM (in daltons). ‡Additional atoms that are likely to be present.

To investigate the nature of the cofactor, we nucleotide (NAMN) derivative in all species represents a snapshot of the active LarA. As shown
grew recombinant Lc. lactis cells expressing (fig. S6 and supplementary text). This result in the Fo-Fc omit map (Fig. 3B), Ni is coordinated
the lar operon in chemically defined medium also suggests that the NAMN derivative is at- with the side-chain of His200 and a NAMN de-
(4) that was selectively depleted of particular tached to the lysine through its carboxylic acid rivative that is covalently attached to Lys184 and
vitamins. Niacin (nicotinic acid or nicotinamide) rather than by the sugar phosphate. After sub- interacts with several highly conserved residues.
was necessary for generating Lar activity (fig. tracting the mass of the NAMN, the additional The electron density profile supports a linkage of
S3), suggesting that niacin is a precursor of the mass (131.864 T 0.014 daltons) corresponds to C3 of the pyridinium ring to Lys184 with strong
LarA cofactor. To directly test this hypothesis, the addition of CS2Ni2+ (131.864 daltons) because electron density, which is comparable with the
we grew Lc. lactis cells in the presence of tetra- the closest alternative formula (CO2SNi2+: 131.882 density of the S of the spatially close Met224, at
deuterated nicotinic acid and analyzed the pu- daltons) falls outside the 95% confidence interval the position expected for O of an amide bond (fig.
rified LarA by means of MS. The resulting mass of the measurement (Table 1). The masses of the S10), strongly suggesting S occupies this position.
of LarA showed a 2 T 0.5–dalton mass increase observed fragments also support this formula The electron density on C5 of the pyridinium ring
compared with the holoenzyme isolated from (fig. S5 and supplementary text). In addition, the has a shape and density (comparable with that of
control cultures (Fig. 1, B and E), indicating that isotopic distribution of this species is in agreement the S of Met224) (fig. S10), suggesting a thiocarbox-
the nicotinic acid was incorporated into the pro- with the theoretical model. In particular, the ylic acid where the S faces Ni. Collectively, the
tein and that two deuterium atoms had under- lower abundance of the M+1 species versus the crystallographic data suggest a cofactor with
gone substitution. M and M+2 species is a fingerprint of Ni presence CS2Ni added to the NAMN (Fig. 3B), which is
To better define the structure of this nicotinic (Fig. 2B) because it reflects the isotopic distribu- consistent with the MS data (Table 1). As a result,
acid derivative, we digested LarA containing nor- tion of Ni (58Ni, 68%; 59Ni, 0%; and 60Ni, 26% Ni is coordinated by the S atoms of a thioamide
mal or deuterated cofactors with trypsin and abundance). and a thiocarboxylic acid along with His200 and
analyzed the peptides by means of LC-ESI-MS. Assuming the 1610-dalton species is the native C4 of the pyridine ring, forming a distorted square
We detected four peptides eluting with similar Ni-containing cofactor, the 1554-dalton species planar complex. A previous study suggested Ni
retention times in both positive and negative could directly arise from it by loss of Ni2+ (Table 1), coordination by two His (2), but this cofactor
ionization mode, all of which contained a 2.010 T which is consistent with the loss of 56 T 1 daltons was not included when modeling those data and
0.003–dalton mass increase when d4-nicotinic from the holoprotein depleted of sulfite (Fig. 1C), could substitute for one of the His side chains.
acid containing protein was digested, demonstrat- whereas the 1586- and 1570-dalton species could The additional thiocarboxylic acid and the Ni-C
ing the presence of a covalently bound nicotinic arise from gain of O2 or half O2, respectively, by the bond explain the loss of two deuterium atoms
acid derivative (Fig. 2A, Table 1, and fig. S4A). 1554-dalton species. An alternative 1554-dalton from the d4-nicotinic acid in the final cofactor
The molecular mass of the 1610-dalton species species could arise from loss of S or O from the (Figs. 1E and 2A).
corresponds to the addition of 448.894 T 0.014 1586- or 1570-dalton species, respectively (Table 1, The LarA metallocenter resembles pincer com-
daltons to the peptide (R)K184SVLPGIASYK(T) fig. S7, and supplementary text). plexes that are widely applied in organic syn-
(1162 daltons), which is 2 daltons smaller than To establish the structure of the NAMN deriva- thesis or organometallic catalysis (5, 6), but such
the 451 T 1 dalton compound found in the holo- tive and examine how Ni binds at the catalytic site, a complex has not been observed in enzymes.
protein (Table 1 and Fig. 1B), which suggests that we purified LarA in the presence of sulfite and Following the chemical nomenclature, LarA has
the same compound is present, probably in a more crystallized it within 3 days in order to minimize a (SCS)Ni(II) pincer complex that we name a
oxidized state. Moreover, the 1610-dalton species activity loss. We solved the structure at 3.0 Å reso- pyridinium-3-thioamide-5-thiocarboxylic acid mono-
was 10 times more abundant after digestion in lution by means of molecular replacement using nucleotide Ni pincer. Although organometallic
microaerobic conditions (fig. S4B). Together, these the structure of LarATt as a template (table S2) (2). (SCS)Ni(II) pincer complexes have been synthe-
findings strongly suggest that the 1610-dalton spe- The overall fold of LarA is similar to LarATt, but sized (7, 8), the chemistry of these complexes
cies contains the original cofactor and that the the two LarA molecules in one asymmetric unit remains underexplored (9). Pincer complexes are
other species represent degradation products aris- adopt distinct conformations (figs. S8A and S9): usually very stable (5); however, Ni is lost from
ing, in part, from its oxidation. The 1610-dalton LarA with an open catalytic site is nearly super- the (SCS)Ni cofactor of LarA in aerobic environ-
species and the two other major derivatized pep- imposable to LarATt, whereas LarA in the closed ments in the absence of sulfite (fig. S2, B and C),
tide species showed MS/MS fragment ions in posi- conformation has a much narrower cleft at the correlating with the oxidation of the cofactor
tive ionization mode that matched fragments of active site with the N- and C-domains approach- by molecular oxygen (Table 1). The basis of this
the residue 184-197 peptide with modified Lys184 ing each other, leading to a sealed catalytic cen- oxidation remains unclear, but the S moieties of
(fig. S5 and supplementary text), which is in agree- ter (fig. S8B). The strong anomalous signal of Ni (SCS)Ni(II) complexes are presumably prone to
ment with our mutagenesis results. All four deriv- indicated that only the molecule in the closed con- oxidative decomposition (9). Oxidation of a sul-
atized peptide species showed fragmentation formation has a Ni atom in the active site (Fig. 3, fur in LarA (SCS)Ni might lead to the loss of co-
patterns in negative ionization mode consistent A and B, and fig. S8C). Because Ni is absolutely ordination of Ni (fig. S11), accounting for the
with the loss of ribose phosphate (212 daltons), required for LarA activity, this observation sug- observed Ni loss and explaining the instability of
indicating the presence of a nicotinic acid mono- gested that the closed conformation most likely enzyme activity.

68 3 JULY 2015 • VOL 349 ISSUE 6243 sciencemag.org SCIENCE

 RE S EAR CH | R E P O R T S

Fig. 3. Structure of LarA holoprotein and pu-

tative catalytic mechanism. (A) Structure of
LarA in the active conformation. The protein is
illustrated in cartoon mode, and the (SCS)Ni
complex, His200, and Lys184 are shown as a stick
model, with the cofactor C atoms in bright yellow,
O in red, N in blue, S in dark yellow, and P in orange.
Ni is depicted as a green sphere. (B) An expanded
view of the LarA catalytic site as seen from the
perspective of the arrow in (A). The color code for
atoms is the same as in (A).The Fo-Fc omit map of
the (SCS)Ni cofactor (s = 4.0) is shown in blue, and
the anomalous difference map of nickel (s = 5.0) is
shown in pink.The coordination bonds of nickel are
shown as light blue dashed lines (length in
angstroms). (C) Hypothetical catalytic mechanism
of LarA. Straight arrows are coordination; curved
arrows indicate the movement of electrons; dotted
lines indicate the hydrogen bonds. An alternative
site for hydride transfer is directly onto the Ni.

The presence of a metal-C bond is unusual in pyruvate intermediate (Fig. 3C). Thus, the (SCS) 12. T. Spatzal et al., Science 334, 940 (2011).
enzyme cofactors, but some well-known exam- Ni pincer complex may act to reversibly cap- 13. S. W. Ragsdale, H. G. Wood, W. E. Antholine, Proc. Natl. Acad.
Sci. U.S.A. 82, 6811–6814 (1985).
ples are adenosyl cobalamin (10), [Fe]-hydrogenase ture a hydride from the substrate, which is an-
14. T. Hiyama, S. Fukui, K. Kitahara, J. Biochem. 64, 99–107 (1968).
(11), and nitrogenase (12), and they are found as alogous to the role of nicotinamide adenine 15. H. Katagiri, T. Sugimori, K. Imai, Agric. Biol. Chem. 25, 281–289
catalytic intermediates as in CO dehydrogenase dinucleotide (NAD) in traditional hydride tran- (1961).
(13). The reason for the presence of a Ni-C bond sfer enzymes (18), but by remaining tethered to 16. D. Dennis, N. O. Kaplan, Biochem. Z. 338, 485–495 (1963).
in Lar is not clear, but its presence suggests a the protein, it facilitates substrate racemization. 17. A. Cantwell, D. Dennis, Biochemistry 13, 287–291 (1974).
18. P. Belenky, K. L. Bogan, C. Brenner, Trends Biochem. Sci. 32,
role in catalysis. Past reports suggested a hydride Of additional interest, synthesis of the (SCS)Ni 12–19 (2007).
transfer mechanism for lactate racemization complex requires LarB, LarC, and LarE, which
(14–17). The participation of the (SCS)Ni cofactor are widely distributed in microorganisms that AC KNOWLED GME NTS
in LarA catalysis would require that the substrate lack LarA supporting a cofactor role in other We thank E. L. Hegg, A. D. Jones, and O. Riant for fruitful
binds nearby. In the crystal structure, bound sul- enzymes (2). discussions; L. Chen at the Michigan State University Mass
Spectrometry and Metabolomics Core for instructions and
fate forms hydrogen bonds with six conserved help with data collection; and J. Brunzelle and Z. Wawrzak
residues at the presumed catalytic site (fig. S12), RE FERENCES AND NOTES at the Life Sciences Collaborative Access Team of Advanced
which strongly suggests a similar binding mode 1. C. T. Walsh, Science 261, 308–309 (1993).
Proton Source for assistance with data collection. The LarA
for lactate. If bound at this site, the C2 of lactate structure has been deposited with the Protein Data Bank
2. B. Desguin et al., Nat. Commun. 5, 3615 (2014).
(accession code 4YNS). The work of P.H. and P.S. was
would be located directly on top of the C4 of the 3. J. L. Boer, S. B. Mulrooney, R. P. Hausinger, Arch. Biochem.
supported by grants from the Belgian National Fund for
pyridinium ring and adjacent to the Ni atom, Biophys. 544, 142–152 (2014).
Scientific Research (FNRS) and the Université Catholique de
4. G. Zhang, D. A. Mills, D. E. Block, Appl. Environ. Microbiol. 75,
where hydride transfer is expected to take place 1080–1087 (2009).
Louvain–Fonds Spéciaux de Recherche. R.P.H. acknowledges
(Fig. 3, B and C). Furthermore, two conserved support from Michigan State University.
5. J. T. Singleton, Tetrahedron 59, 1837–1857 (2003).
histidine residues, His108 and His174, are situated 6. G. van Koten, D. Milstein, Eds., Organometallic Pincer
on each side of the substrate and may participate Chemistry (Springer, Germany, 2013), vol. 40. SUPPLEMENTARY MATERIALS
7. S. M. Peterson, M. L. Helm, A. M. Appel, Dalton Trans. 44,
in the racemization mechanism by acting as bases 747–752 (2015).
www.sciencemag.org/content/349/6243/66/suppl/DC1
for deprotonating the hydroxyl group of one or Materials and Methods
8. C. A. Kruithof et al., Organometallics 27, 4928–4937 (2008).
Supplementary Text
the other lactate enantiomer and as acids for re- 9. D. Zargarian, A. Castonguay, D. M. Spasyuk, in Organometallic
Figs. S1 to S12
protonating the pyruvate intermediate (Fig. 3, Pincer Chemistry, G. van Koten, D. Milstein, Eds.
Tables S1 to S3
(Springer-Verlag, Berlin, 2013), vol. 40, pp. 131–174.
B and C). Arg75 forms hydrogen bonds with the 10. L. Randaccio, S. Geremia, N. Demitri, J. Wuerges, Molecules 15,
References (19–30)
(SCS)Ni cofactor and additionally could be in- 3228–3259 (2010). 30 March 2015; accepted 28 May 2015
volved in the stabilization of the hypothesized 11. S. Shima et al., Dalton Trans. 41, 767–771 (2012). 10.1126/science.aab2272

SCIENCE sciencemag.org 3 JULY 2015 • VOL 349 ISSUE 6243 69

 A tethered niacin-derived pincer complex with a nickel-carbon bond
in lactate racemase
Benoît Desguin et al.
Science 349, 66 (2015);
DOI: 10.1126/science.aab2272

This copy is for your personal, non-commercial use only.

If you wish to distribute this article to others, you can order high-quality copies for your
colleagues, clients, or customers by clicking here.
Permission to republish or repurpose articles or portions of articles can be obtained by
following the guidelines here.

The following resources related to this article are available online at

www.sciencemag.org (this information is current as of July 2, 2015 ):

Downloaded from www.sciencemag.org on July 2, 2015

Updated information and services, including high-resolution figures, can be found in the online
version of this article at:
http://www.sciencemag.org/content/349/6243/66.full.html
Supporting Online Material can be found at:
http://www.sciencemag.org/content/suppl/2015/07/01/349.6243.66.DC1.html
A list of selected additional articles on the Science Web sites related to this article can be
found at:
http://www.sciencemag.org/content/349/6243/66.full.html#related
This article cites 27 articles, 8 of which can be accessed free:
http://www.sciencemag.org/content/349/6243/66.full.html#ref-list-1
This article has been cited by 1 articles hosted by HighWire Press; see:
http://www.sciencemag.org/content/349/6243/66.full.html#related-urls
This article appears in the following subject collections:
Biochemistry
http://www.sciencemag.org/cgi/collection/biochem

Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the
American Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. Copyright
2015 by the American Association for the Advancement of Science; all rights reserved. The title Science is a
registered trademark of AAAS.