Identification and characterization of bifunctional proline racemase/hydroxyproline epimerase from archaea: Discrimination of substrates and molecular evolution

Article abstract of DOI:10.1371/journal.pone.0120349

Proline racemase (ProR) is a member of the pyridoxal 5′-phosphate-independent racemase family, and is involved in the Stickland reaction (fermentation) in certain clostridia as well as the mechanisms underlying the escape of parasites fromhost immunity in eukaryotic Trypanosoma. Hydroxyproline epimerase (HypE), which is in the same protein family as ProR, catalyzes the first step of the trans-4-hydroxy-L-proline metabolism of bacteria. Their substrate specificities were previously considered to be very strict, in spite of similarities in their structures and catalytic mechanisms, and no racemase/epimerase fromthe ProR superfamily has been found in archaea.We here characterized the ProR-like protein (OCC-00372) from the hyperthermophilic archaeon, Thermococcus litoralis (TlProR). This protein could reversibly catalyze not only the racemization of proline, but also the epimerization of 4-hydroxyproline and 3-hydroxyproline with similar kinetic constants. Among the four (putative) ligand binding sites, one amino acid substitution was detected between TlProR (tryptophan at the position of 241) and natural ProR (phenylalanine). TheW241F mutant showed a significant preference for proline over hydroxyproline, suggesting that this (hydrophobic and bulky) tryptophan residue played an importance role in the recognition of hydroxyproline (more hydrophilic and bulky than proline), and substrate specificity for hydroxyproline was evolutionarily acquired separately between natural HypE and ProR. A phylogenetic analysis indicated that such unique broad substrate specificity was derived from an ancestral enzyme of this superfamily.

Full text of DOI:10.1371/journal.pone.0120349

RESEARCH ARTICLE
Identification and Characterization of
Bifunctional Proline Racemase/
Hydroxyproline Epimerase from Archaea:
Discrimination of Substrates and Molecular
Evolution
Seiya Watanabe*, Yoshiaki Tanimoto, Hisashi Nishiwaki, Yasuo Watanabe
Faculty of Agriculture, Ehime University, Matsuyama, Ehime, 790–8566, Japan
* irab@agr.ehime-u.ac.jp
Abstract
Proline racemase (ProR) is a member of the pyridoxal 5’-phosphate-independent racemase
family, and is involved in the Stickland reaction (fermentation) in certain clostridia as well as
the mechanisms underlying the escape of parasites from host immunity in eukaryotic Trypa-
nosoma. Hydroxyproline epimerase (HypE), which is in the same protein family as ProR, cat-
alyzes the first step of the trans-4-hydroxy-L-proline metabolism of bacteria. Their substrate
specificities were previously considered to be very strict, in spite of similarities in their struc-
tures and catalytic mechanisms, and no racemase/epimerase from the ProR superfamily has
been found in archaea. We here characterized the ProR-like protein (OCC_00372) from the
hyperthermophilic archaeon, Thermococcus litoralis (TlProR). This protein could reversibly
catalyze not only the racemization of proline, but also the epimerization of 4-hydroxyproline
and 3-hydroxyproline with similar kinetic constants. Among the four (putative) ligand binding
sites, one amino acid substitution was detected between TlProR (tryptophan at the position of
241) and natural ProR (phenylalanine). The W241F mutant showed a significant preference
for proline over hydroxyproline, suggesting that this (hydrophobic and bulky) tryptophan resi-
due played an importance role in the recognition of hydroxyproline (more hydrophilic and
bulky than proline), and substrate specificity for hydroxyproline was evolutionarily acquired
separately between natural HypE and ProR. A phylogenetic analysis indicated that such
unique broad substrate specificity was derived from an ancestral enzyme of this superfamily.
OPEN ACCESS
Citation: Watanabe S, Tanimoto Y, Nishiwaki H,
Watanabe Y (2015) Identification and
Characterization of Bifunctional Proline Racemase/
Hydroxyproline Epimerase from Archaea:
Discrimination of Substrates and Molecular Evolution.
PLoS ONE 10(3): e0120349. doi:10.1371/journal.
pone.0120349
Academic Editor: Paul Taylor, University of
Edinburgh, UNITED KINGDOM
Received: October 31, 2014
Accepted: January 21, 2015
Published: March 18, 2015
Copyright: © 2015 Watanabe et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any
medium, provided the original author and source are
credited.
Data Availability Statement: All relevant data are
within the paper and its Supporting Information files.
Introduction
Funding: The authors have no support or funding to
report.
L-Proline can serve as a complete source of carbon and energy or of nitrogen for organisms.
The metabolism of L-proline is generally initiated through its oxidation by L-proline dehydro-
genase (EC 1.5.99.8; L-ProDH) to form Δ1-pyrroline-5-carboxylate (Pyr5C) [1] (Fig. 1A).
Competing Interests: The authors have declared
that no competing interests exist.
PLOS ONE | DOI:10.1371/journal.pone.0120349 March 18, 2015
1 / 17

Bifunctional Proline Racemase/Hydroxyproline Epimerase
Fig 1. Putative metabolic network of proline and hydroxyproline in archaea. (A) Metabolic networks of L-proline, D-proline, and T3LHyp. (B) Schematic
gene clusters related to the metabolism of proline and/or T3LHyp by archaea. Homologous genes are indicated in the same color and correspond to Fig. 1A.
Putative genes in the box were purified and characterized in this study (see Fig. 1C). No clustering was found around the FaProR gene. C-C and C-T indicate
a pair of catalytic amino acid residues of ProR superfamily enzymes. Gray putative genes are sequentially similar to other (amino acid) transporters. (C)
Purification of recombinant His6 tag proteins. Five micrograms each of the purified protein were applied to a 12% (w/v) gel. (D) Western blot analysis. One
microgram each of the purified protein were applied to a 12% (w/v) gel.
doi:10.1371/journal.pone.0120349.g001
PLOS ONE | DOI:10.1371/journal.pone.0120349 March 18, 2015
2 / 17

Bifunctional Proline Racemase/Hydroxyproline Epimerase
Fig 2. Library of proline derivatives. TlProR can enable the reversible racemization and epimerization of all proline derivatives (arrows).
doi:10.1371/journal.pone.0120349.g002
Following its spontaneous hydrolysis, the L-glutamate γ-semialdehyde produced is then oxi-
dized to L-glutamate by Pyr5C dehydrogenase (EC 1.2.1.88). L-Proline is alternatively metabo-
lized to 5-aminopentanoate through two enzymatic steps. L-Proline racemase (EC 5.1.1.4;
ProR) first catalyzes the racemization of L-proline to D-proline, followed by reductive cleavage
to yield 5-aminopentanoate by D-proline reductase (EC 1.21.4.1). This pathway is only opera-
tive in certain clostridia including Clostridium sticklandii [2] and Clostridium difficile [3], clini-
cally significant nosocomial pathogens, and Trypanosoma species including Trypanosoma
cruzi [4, 5] and Trypanosoma vivax [6], the causative agents of Chagas disease and animal try-
panosomiasis, respectively. In the former, the pathway is involved in the so-called “Stickland
reaction (fermentation)”, whereas ProR in the latter has been implicated in the mechanisms
underlying the escape of parasites from host immunity as a B-cell mitogen.
Hydroxyproline has been detected in certain proteins, especially collagen, as well as some
peptide antibiotics. In mammalian systems [1], the L-proline residue in procollagen is post-
translationally hydroxylated to trans-4-hydroxy-L-proline (T4LHyp) or trans-3-hydroxy-L-
proline (T3LHyp). Among the several stereoisomers of hydroxyproline including T4LHyp and
T3LHyp (Fig. 2), T4LHyp is the most common in nature. In contrast to mammalians [1], some
bacteria have been shown to metabolize T4LHyp to α-ketoglutarate via three intermediates
through four enzymatic steps [7, 8]. Of these, hydroxyproline 2-epimerase (HypE; EC 5.1.1.8)
first catalyzes the isomerization of T4LHyp to cis-4-hydroxy-D-proline (C4DHyp), which is
then converted to Δ1-pyrroline-4-hydroxy-2-carboxylate (Pyr4H2C) by C4DHyp dehydroge-
nase (D-HypDH). In the metabolism of T3LHyp (Fig. 1A), T3LHyp dehydratase (EC 4.2.1.77;
T3LHypD) initially catalyzes the dehydration of T3LHyp to Δ1-pyrroline-2-carboxylate
(Pyr2C) via a putative Δ2-pyrroline-2-carboxylate intermediate [9]. The Pyr2C is then con-
verted by NAD(P)H-dependent Pyr2C reductase (EC 1.5.1.1) to yield L-proline, which is me-
tabolized through the general degradation of L-proline described above. This pathway is most
commonly found in mammals [9], bacteria [10] and archaea [11].
Racemase enzymes catalyze the deprotonation/reprotonation of the chiral carbon (Cα) of
both amino acid enantiomers, resulting in the stereoinversion of chiral centers. ProR is a mem-
ber of the pyridoxal 5’-phosphate (PLP)-independent racemase family along with HypE and
T3LHypD (although the reaction involving the latter is not epimerization) [9, 12, 13]. These
enzymes have been classified into three types based on the two specific residues at the active
sites [9]: the Cys-Cys type (ProR and HypE), Cys-Thr type (T3LHypD), and Ser-Cys or Ser-
Thr type (function unknown). In the cases of ProR and HypE, catalysis is based on the same
1,1-proton transfer mechanism using two general acidic/basic cysteine residues located on
PLOS ONE | DOI:10.1371/journal.pone.0120349 March 18, 2015
3 / 17

Bifunctional Proline Racemase/Hydroxyproline Epimerase
opposite faces of the active site [12]. On the other hand, their substrate specificities were previ-
ously considered to be very strict: proline for ProR [13] and T4LHyp and cis-4-hydroxy-L-pro-
line (C4LHyp) for HypE [13, 14]. Although the catalytic mechanism of T3LHypD currently
remains unclear, dehydration may start by the abstraction of a proton from the Cα of the sub-
strate by an active site cysteine residue as a general basic catalyst [9]. This enzyme can also only
utilize T3LHyp as a substrate.
As described above, no racemase/epimerase from the ProR superfamily had been found in
archaea. We here identified and characterized a ProR-like protein with a pair of Cys-Cys as
(putative) active sites from the hyperthermophilic marine archaeon, Thermococcus litoralis
[15], which differed from the previously isolated T3LHypD enzyme [11]. This protein could re-
versibly catalyze not only the racemization of proline, but also the epimerization of 4-hydroxy-
proline and 3-hydroxyproline with similar kinetic constants. Site-directed mutagenesis
revealed several important amino acid residue(s) responsible for discriminating between pro-
line and hydroxyproline. A phylogenetic analysis indicated that such unique broad substrate
specificity was derived from an ancestral enzyme of this superfamily. To the best of our knowl-
edge, this is the first study on the bifunctional enzyme of ProR and HypE from archaea.
Materials and Methods
Materials
T3LHyp was purchased from Kanto Chemical (Tokyo, Japan). C4LHyp, L-pipecolate, and D-
pipecolate were obtained from Tokyo Chemical Industry (Tokyo, Japan). T4DHyp and
C3LHyp were from Sigma Aldrich (USA). T4LHyp, C4DHyp, L-proline, and D-proline were
from Wako Pure Chemical Industries (Osaka, Japan).
General procedures
Basic recombinant DNA techniques were performed as described by Sambrook et al. [16]. Ar-
chaeal genomic DNA was prepared using a DNeasy Tissue Kit (Qiagen). PCR was carried out
using a GeneAmp PCR System 2700 (Applied Biosystems) for 30 cycles in 50 μl of reaction
mixture containing 1 U of KOD FX DNA polymerase (TOYOBO), appropriate primers (15
pmol), and template DNA under the following conditions: denaturation at 98°C for 10 s, an-
nealing at 50°C for 30 s, and extension at 68°C for time periods calculated at an extension rate
of 1 kbp∙min−1. DNA sequencing was carried out using the BigDye Cycle Sequencing Kit
ver.3.1 (Applied Biosystems) and appropriate primers with the Genetic Analyzer 3130 (Ap-
plied Biosystems). High-pressure liquid chromatography (HPLC) was performed using an Agi-
lent 1120 Compact LC system (TOSOH). Protein concentrations were determined by the
method of Lowry et al. [17] with bovine serum albumin as the standard. SDS-PAGE was per-
formed as previously described by Laemmli [18]. Western blot analysis was carried out using
an ECL Western Blotting Analysis System (GE Healthcare) and RGS∙His HRP antibody (horse-
radish peroxidase-fused mouse monoclonal antibody against Arg-Gly-Ser-His6 in the N-termi-
nal additional peptide of the expressed recombinant proteins (Qiagen))
Plasmid construction for expression of recombinant proteins
The primer sequences used in this study are shown in S1 Table. In this study, the prefixes Tl
(T. litoralis), Fa (Ferroplasma acidiphilum), Hj (Haloarcula japonica), Cd (C. difficile), and Ab
(Azospirillum brasilense) were added to gene symbols or protein designations where necessary
for clarity. The TlProR (GenBank accession number OCC_00372), FaProR (FACI_I-
FERC00001G0786), HjProR (C444_02196), CdProR (CD630_32370), and AbHypE genes
PLOS ONE | DOI:10.1371/journal.pone.0120349 March 18, 2015
4 / 17

Bifunctional Proline Racemase/Hydroxyproline Epimerase
(BAN78527) were amplified by PCR using primers containing appropriate restriction enzyme
sites at the 5’- and 3’-ends and the genomic DNA of T. litoralis DSM 5473, F. acidiphilum DSM
12658, H. japonica DSM 6131, C. difficile 630, or A. brasilense ATCC 29145 as a template, re-
spectively. Each amplified DNA fragment was introduced into BamHI-SalI sites (for the
TlProR, FaProR, and CdProR genes), BamHI-PstI (for the HjProR gene), or BamHI-HindIII
(for the AbHypE gene) in pETDuet-1 (Novagen), a plasmid vector used to confer an N-termi-
nal His6 tag on expressed proteins, to obtain pET/TlProRWT, pET/FaProRWT, pET/HjProR,
pET/CdProR, and pET/AbHypEWT, respectively.
In the present study, we used F. acidiphilum DSM 12658 instead of Ferroplasma acidarma-
nus fer1 as a target microorganism, for which there was 99% identity in the 16S rRNA se-
quence. We successfully amplified the FaProR gene by genomic PCR using oligonucleotide
primers designed from the F. acidarmanus fer1 genome sequence. Unless otherwise noted, Fer-
roplasma sp. hereafter indicates strain DSM 12658.
Expression and purification of His6-tagged recombinant proteins
Escherichia coli BL21-CodonPlus(DE3)-RIL (Novagen) harboring the constructed pETDuet-1
plasmid was grown at 37°C to a turbidity of 0.6 at 600 nm in Super broth medium (pH 7.0, 12
g tryptone, 24 g yeast extract, 5 ml glycerol, 3.81 g KH2PO4, and 12.5 g K2HPO4 per liter) con-
taining 50 mg/liter ampicillin. After the addition of 1 mM isopropyl-β-D-thiogalactopyrano-
side (IPTG), the culture was grown for a further 6 h to induce the expression of the His6-
tagged protein. The grown cells were harvested by centrifugation at 30,000 × g for 20 min, sus-
pended in Buffer A (50 mM sodium phosphate buffer (pH 8.0) containing 300 mM NaCl and
10 mM imidazole), disrupted by sonication for 20 min at appropriate intervals on ice using
Ultra Sonic Disruptor UD-211 (TOMY SEIKO Co., Ltd, Tokyo, Japan), and then centrifuged
at 108,000 × g for 20 min at 4°C. The supernatant was loaded onto a Ni-NTA Superflow col-
umn (Qiagen) equilibrated with Buffer A linked to the BioAssist eZ system (TOSOH). The col-
umn was washed with Buffer B (50 mM sodium phosphate buffer (pH 8.0) containing 300 mM
NaCl, 10% (v/v) glycerol, and 50 mM imidazole). The enzymes were then eluted with Buffer C
(pH 8.0, Buffer B containing 250 mM imidazole instead of 50 mM imidazole), concentrated by
ultrafiltration with Centriplus YM-30 (Millipore), dialyzed against 50 mM Tris-HCl buffer
(pH 8.0) containing 50% (v/v) glycerol, and stored at −35°C until use.
The native molecular mass of recombinant proteins was estimated by gel filtration, which
was carried out using a HPLC system at a flow rate of 1 ml/min. The purified enzyme was load-
ed onto a TSKgel G3000SWXL column (TOSOH) equilibrated with 50 mM Tris-HCl buffer
(pH 8.0). A high molecular weight gel filtration calibration kit (GE Healthcare) was used for
molecular markers.
Enzyme assay by HPLC
Purified protein was added to 50 mM Tris-HCl buffer (pH 8.0) (1 ml) containing 10 mM sub-
strate, and incubated at 50°C (for TlProR and HjProR) or 30°C (FaProR, CdProR, and
AbHypE). After varying the incubation times, the enzyme reaction was stopped by rapidly
incubating at −80°C. To this solution, 25 μl each of 10 mM 4-fluoro-7-nitrobenzofurazan
(NBD-F) in ethanol and 100 mM borate buffer (pH 8.0) was then added and incubated at 60°C
for 1 min. A total of 1.15 ml of 5 mM HCl was finally added to the reaction mixture, followed
by filtration through a 0.22-μm filter (Millipore). HPLC analysis was carried out using an
Ultron ES-CD chiral separation column (150 × 2.0 mm, Shinwa Chemical Industries). The
flow rate was 1 ml∙min−1, and isocratic elution with 20 mM potassium phosphate (pH 6.0)
containing 10% (v/v) methanol was monitored by absorbance at 530 nm. Non-enzymatic
PLOS ONE | DOI:10.1371/journal.pone.0120349 March 18, 2015
5 / 17

Bifunctional Proline Racemase/Hydroxyproline Epimerase
conversion of L$D isomers was observed even after being incubated for 12 h, and the differ-
ence in background turnover between 30°C and 50°C was negligible. This method was used to
estimate the effects of substrate specificity and inhibition by pyrrole-2-carboxylate (PYC)
on activity.
Spectrophotometric enzyme assay
Racemase/epimerase activity for proline or hydroxyproline(s) was spectrophotometrically as-
sayed at 50°C by monitoring the reduction rate of an artificial electron acceptor in the coupling
system with dehydrogenase for proline or hydroxyproline, using a Shimadzu UV-1800 spectro-
photometer (Shimadzu GLC Ltd., Tokyo, Japan).
Activity toward L-proline and T3LHyp. The standard reaction mixture contained 0.05 mM
2,6-dichloroindophenol (Cl2Ind) and purified D-ProDH from Pyrobaculum islandicum (20 μg)
[19] in 50 mM Tris-HCl buffer (pH 8.0).
Activity toward D-proline. The standard reaction mixture contained 0.05 mM Cl2Ind and
purified L-ProDH from Aeropyrum pernix (20 μg) [20] in 50 mM Tris-HCl buffer (pH 8.0).
Activity toward T4LHyp. The standard reaction mixture contained 0.25 mM p-iodonitrote-
trazolium violet (INT), 0.06 mM phenazine methosulfate (PMS), and purified D-HypDH from
A. brasilense (20 μg) [21] in 50 mM Tris-HCl buffer (pH 8.0).
All reactions were initiated by the addition of 100 mM substrate (100 μl) with a final reac-
tion volume of 1 ml. The millimolar absorption coefficients (ε) for Cl2Ind and INT were
19.1 mM−1∙cm−1 at 600 nm and 15.0 mM−1∙cm−1 at 490 nm, respectively. The detectable limit
of specific activity was *0.001 unit/mg protein, due to non-enzymatic absorbance change by
Cl2Ind and PMS/INT. This method was used to determine the kinetic parameters, Km and kcat,
which were calculated by a Lineweaver-Burk plot.
Identification of reaction products
To remove glycerol in the stored buffer, purified TlProR (10 μg) was dialyzed at 4°C overnight in
50 mM potassium phosphate buffer (pH 7.0), lyophilized, and solved in D2O (600 μl) containing
20 mM T4LHyp or T3LHyp. NMR spectra were recorded at 25°C on a JEOL JNM-EC400 NMR
spectrometer (JEOL Ltd., Tokyo, Japan) operating at 400 MHz. 2,2-Dimethyl-2-silapentane-5-
sulfonate was used as an internal standard. 1H NMR spectra of T4LHyp and T3LHyp (400 MHz,
D2O) were as follows: T4LHyp, 4.41 (1H, t, J = 4 Hz), 4.33 (1H, dd, J1 = 10 Hz, J2 = 8 Hz), 3.47
(1H, dd, J1 = 12 Hz, J2 = 4 Hz), 3.35 (1H, ddd, J1 = 12 Hz, J2 = J3 = 2 Hz), 2.42 (1H, dddd, J1 =
14 Hz, J2 = 8 Hz, J3 = J4 = 2 Hz), and 2.15 (1H, ddd, J1 = 14 Hz, J2 = 10 Hz, J3 = 4 Hz); C4DHyp,
4.57 (1H, m), 4.21 (1H, dd, J1 = 11 Hz, J2 = 4 Hz), 3.36 (1H, dd, J1 = 12 Hz, J2 = 4 Hz), 2.50
(1H, ddd, J1 = 14 Hz, J2 = 10 Hz, J3 = 4 Hz), and 2.25 (1H, m); T3LHyp, 4.68 (1H, m), 4.07
(1H, s), 3.60 (1H, m), 3.50 (1H, m), and 2.04 (2H, m); C3DHyp, 4.71 (1H, m), 4.14 (1H, d, J =
4 Hz), 3.56 (1H, m), 3.48 (1H, m), and 2.18 (2H, m).
Site-directed mutagenesis
A mutation was introduced into the TlProR, FaProR, or AbHypE genes by sequential steps of
PCR [22] using sense and antisense primers (S1 Table) and pET/TlProRWT (for TlProRF62S
(CTG!AGG), TlProRF62V (CTG!GTT), TlProRL221H (CTG!GAG), TlProRF62S/L221H
,
TlProRW241F (GTG!GAC), TlProRW241C (CTG!TGT), and TlProRW241Y mutants
(CTG!TAT)), pET/FaProRWT (for FaProRF240W mutant (GCC!AAG)), or pET/AbHy-
pEWT (for AbHypEC226F mutant (GTG!TTC)) as a template (the codons used for each mu-
tation were in parentheses). The coding region of the mutated genes was confirmed by
PLOS ONE | DOI:10.1371/journal.pone.0120349 March 18, 2015
6 / 17

Bifunctional Proline Racemase/Hydroxyproline Epimerase
subsequent sequencing in both directions. Mutant proteins were expressed and purified by
the same procedures as those for the wild-type enzyme.
Amino acid sequence alignment and phylogenetic analysis
Protein sequences were analyzed using the Protein-BLAST and Clustal W program distributed
by DDBJ (DNA Data Bank of Japan) (www.ddbj.nig.ac.jp). The phylogenetic tree was pro-
duced using the TreeView 1.6.1. program.
Results
Putative ProR or HypE, but not the T3LHypD gene from archaea
We previously reported that OCC_00387 and OCC_00362(7) from T. litoralis encoded
T3LHypD and Pyr2C reductase are involved in (putative) T3LHyp metabolism, respectively
[11] (Fig. 1A). These genes were clustered on the genome together with several other hypothet-
ical genes (Fig. 1B), among which OCC_00357 (α-subunit) and OCC_00352 (β) (genes) were
homologous to α4β4-type L-ProDH from Pyrococcus horikishii (PH1363 and PH1364 with 69
and 80% identities, respectively) [23], while OCC_00347 (α), OCC_00342 (γ), OCC_00337 (δ),
and OCC_00332 (β) (genes) were similar to αβγδ-type L-ProDH from Thermococcus profun-
dus (pdhA, pdhF, pdhX, and pdhB with 81, 72, 72, and 64% identities, respectively) [24]. Since
T3Lhyp is metabolized to L-proline via a Pyr2C intermediate [9–11], the gene cluster may be
related to both the (general) metabolism of L-proline and T3LHyp. Another ProR-like gene
(OCC_00372) was also detected in this gene cluster, and the putative amino acid sequence con-
tained two cysteine residues at the active sites (Cys88-Cys251) specific for ProR or HypE, but
not T3LHypD (see below). Furthermore, by a homology search using the Protein-BLAST pro-
gram, some archaea including F. acidarmanus (F. acidiphilum) and H. japonica were found to
possess the homologous gene (see below in detail). Therefore, in the present study, we selected
three ProR-like genes from T. litoralis, F. acidiphilum, and H. japonica as target genes (referred
to as TlProR, FaProR, and HjProR genes, respectively). Furthermore, ProR from C. difficile
[13] and HypE from A. brasilense [21] were used as controls (referred to as CdProR and
AbHypE, respectively).
Preparation of recombinant His6-tag proteins
After cloning all target genes into the vector pETDuet-1, the recombinant enzymes were suc-
cessfully expressed in E. coli cells as His6-tagged enzymes, and purified to homogeneity using a
nickel-chelating affinity column (Fig. 1C). Additional amino acid residues including the His6-
tag at the N-terminal was confirmed by western blot analysis with an anti-His6-tag antibody
(Fig. 1D). The apparent molecular masses of TlProR, FaProR, HjProR, CdProR, and AbHypE,
as estimated by SDS-PAGE, were 40 (38.1), 40 (38.5), 42 (35.8), 40 (37.7) and 37 (34.1) kDa
(values in parentheses indicate the calculated molecular mass of the enzyme with His6-tag),
while those estimated by analytic gel filtration were *80 kDa, respectively (data not shown).
Therefore, all these enzymes appeared to be dimeric.
Substrate specificity of TlProR
Stereoisomer identification and quantification by potential enzyme reactions were determined
by chiral separation column chromatography, in which a sample was derivatized with NBD-F.
The typical retention times (min) of the tested substrates (Fig. 2) were follows: L-proline, 12.9;
D-proline, 11.6; T4LHyp, 8.2; C4DHyp, 8.7; C4LHyp, 9.5; T4DHyp, 7.1; T3LHyp, 8.0;
PLOS ONE | DOI:10.1371/journal.pone.0120349 March 18, 2015
7 / 17

Bifunctional Proline Racemase/Hydroxyproline Epimerase
Fig 3. Substrate specificities of TlProR, FaProR, HjProR, CdProR, and AbHypE. The reaction mixture (1 ml) consisted of 10 mM of substrate and each
purified enzyme as follows in Tris-HCl buffer (pH 8.0): TlProR, 1.3 μg; FaProR, 100 μg; HjProR, 570 μg; CdProR, 3.6 μg; AbHypE, 4.9 μg. L-A2C is L-
azetidine-2-carboxylate. Samples at the indicated times were analyzed by HPLC (means S.D., n = 3). Values on the bars of TlProR indicate specific activity
(unit/mg protein), which are similar to those in Tables 1 and 2.
doi:10.1371/journal.pone.0120349.g003
(putative) C3DHyp, 8.5; C3LHyp, 9.8; (putative) T3DHyp, 7.3; L-pipecolate, 15.4; D-pipeco-
late, 14.7; L-azetidine-2-carboxylate, 9.1; (putative) D-azetidine-2-carboxylate, 8.6.
TlProR could catalyze the conversion of not only L-proline, but also 4-hydroxy-L-proline
(T4LHyp (63) and C4LHyp (69)) and 3-hydroxy-L-proline (T3LHyp (73) and C3LHyp (47))
(values in parentheses indicate specific activity as a percentage of L-proline) (Fig. 3). The reac-
tion was completely bi-directional, and the reverse reactions were also similar: D-proline (130),
C4DHyp (160), and T4DHyp (97) (values in parentheses indicate specific activity as a
PLOS ONE | DOI:10.1371/journal.pone.0120349 March 18, 2015
8 / 17

Bifunctional Proline Racemase/Hydroxyproline Epimerase
Fig 4. Unique inhibition of archaeal ProR-like enzymes by pyrrole-2-carboxylate (PYC; inset). Reactions were performed for 30 min with the same
conditions as those in Fig. 3, except for the presence of several concentration of PYC. L-Proline (for TlProR, FaProR, HjProR, and CdProR) or T4LHyp (for
AbHypE) was used as a substrate. Relative specific activity values were expressed as percentages of the values obtained in the absence of PYC (means
S.D., n = 3). Data for the ProR of T. cruzi (TcProR) are from Berneman et al. [25]. IC50 values were calculated by curve fitting using ImageJ software (http://
rsb.info.nih.gov/ij/).
doi:10.1371/journal.pone.0120349.g004
percentage of L-enantiomer). Furthermore, using large amounts of the purified enzyme, we
could also estimate specific activities for L-pipecolate (0.7), D-pipecolate (3.6), and L-azeti-
dine-2-carboxylate (0.16), which are the six- and four-membered ring analogs of
proline, respectively.
Another previously reported difference in the properties of ProR and HypE is the profile of
inhibition by pyrrole-2-carboxylate (PYC), a transition state analogue of proline [6, 13, 25]:
ProR enzymes from trypanosomatids and clostridia were inhibited at even 0.01 and 1 mM, re-
spectively, whereas HypE reactions were only affected by high amounts of PYC (*10 mM), as
described in CdProR and AbHypE (Fig. 4). In spite of the significant HypE activity, the IC50
value of TlProR (0.217 mM) was *28-fold higher than that of AbHypE (6.00 mM), confirming
dual activity toward proline and 4-hydroxyproline.
Kinetic analysis of TlProR
A more sensitive spectrophotometric assay method using proline or hydroxyproline dehydro-
genase as a coupling enzyme was developed to estimate kinetic properties. The kinetic parame-
ters of proline and hydroxyproline determined from the Lineweaver-Burk plot are shown in
Tables 1 (L!D) and 2 (D!L). The catalytic efficiency (kcat/Km) values for L-proline and
T4LHyp were similar, whereas a preference for D-proline over C4DHyp (45-fold) was identi-
fied and was caused by a 75-fold lower Km for D-proline. This may have been partially due to
differences in the Km values of L-ProDH (0.28 mM) [20] and D-ProDH (4.2 mM) [19] as the
coupling enzyme used. TlProR was assayed at 50°C, which was far from the optimum tempera-
ture (90*100°C; data not shown). Although the kcat/Km value for T4LHyp of TlProR was 27-
fold lower than that for AbHypE, the specific activity of the former at 100°C, estimated by
HPLC analysis, was similar to that of the latter: 29.1 and 58.6 unit/mg protein, respectively.
Zhao et al. [26] recently characterized 51 ProR-like proteins from bacteria. Of these, the
kcat/Km value for T4LHyp of HypE from Pseudomonas putida was 192 and 3430-fold higher
than those for T3LHyp and L-proline, respectively, and was caused by a markedly higher kcat
value. Similar results were also obtained for other (putative) HypE proteins. In contrast, the ki-
netic constant for T3LHyp of TlProR was similar to those of L-proline and T4LHyp. Thus, to
PLOS ONE | DOI:10.1371/journal.pone.0120349 March 18, 2015
9 / 17

Bifunctional Proline Racemase/Hydroxyproline Epimerase
the best of our knowledge, this is the first example of a ProR-like protein being capable of uti-
lizing not only proline, but also 4-hydroxyproline and 3-hydroxyproline as a substrate.
Identification of reaction products by TlProR
T4LHyp was incubated with TlProR in D2O at various incubation times. 1H NMR spectra
showed the progressive loss of H1 peaks, and additionally contained resonances associated with
C4DHyp, a potential product: an exchange of the α-proton with solvent deuterium (Fig. 5A).
Similar phenomena were also observed when T3LHyp was used as a substrate instead of
T4LHyp (Fig. 5B). These results were expected for a 1,1-proton transfer reaction that equili-
brates the configurations at Cα of not only 4-hydroxyproline, but also 3-hydroxyproline (novel
activity), as proposed previously [12].
Phylogenetic analysis of TlProR
As expected from the preliminary annotation, TlProR belongs to the ProR superfamily, which
consists of four subfamilies: the archetype ProR (blue in Fig. 6), HypE (light green), T3LHypD
(orange), and function unknown protein (yellow). Only two putative proteins from the
Table 1. Kinetic parameters for L-proline, T4LHyp, and T3LHyp.
Specific activity a (units/mg protein)
Km (mM)
k
cat (min− 1
)
k
cat/Km (min− 1ꢀmM− 1
)
ProR
HypE
Enzymes
Substrates
TlProRWT
L-Proline
T4LHyp
T3LHyp
L-Proline
T4LHyp
T3LHyp
L-Proline
T4LHyp
L-Proline
T4LHyp
L-Proline
T4LHyp
L-Proline
T4LHyp
L-Proline
T4LHyp
L-Proline
T4LHyp
L-Proline
T4LHyp
L-Proline
T4LHyp
L-Proline
T4LHyp
3.45
9.69
1.21 0.12
6.08 1.76
0.502 0.021
1.29 0.14
4.28 0.84
16.4 3.2
20.5 2.5
6.50 0.39
154 13
167 10
138 4
‒
728 146
216 4
122 13
1.1 b
0.32
‒
3.67
430 10
TlProRW241F
5.19
190 14
147 5
2.13
92.0 10.5
17.5 2.5
55.4 6.1
39.0 1.7
12.0 1.1
2.35 0.17
5.51 0.99
7.84 0.49
N.D.
21.7 1.7
1.07 0.07
2.71 0.05
6.01 0.11
0.0774 0.0010
0.379 0.012
2.88 0.15
8.67 0.18
N.D.
6.8
137
0.45
0.160
0.488
0.670
0.0195
0.0384
0.0691
0.0972
0.001>
0.001>
0.29
TlProRL221H
TlProRF62S
TlProRF62V
TlProRF62S/L221H
FaProRWT
0.20
0.33
‒
6.22 0.62
1.93 0.44
0.905 0.075
N.D. c
N.D.
N.D.
N.D.
31.6 5.3
6.12 0.30
53.0 8.0
2.17 0.13
8.53 0.47
N.D.
35.8 5.6
2.37 0.10
25.0 3.4
1.13 0.01
0.387 0.002
0.473 0.008
0.331 0.005
99.6 57
N.D.
2.9
1.4
‒
0.0209
0.117
0.00822
14.2
FaProRF240W
CdProR
0.717 0.032
849 27
N.D.
0.0975
0.124
58.6
AbHypEWT
AbHypEC226F
N.D.
N.D.
N.D.
‒
1.39 0.15
N.D.
4400 340
N.D.
3180 100
N.D.
0.001>
0.001>
‒
N.D.
N.D.
N.D.
aUnder standard assay conditions in the “Materials and Methods”.
bRatio of L-proline to T4LHyp or T3LHyp in kcat/Km.
cNot determined due to trace activity.
doi:10.1371/journal.pone.0120349.t001
PLOS ONE | DOI:10.1371/journal.pone.0120349 March 18, 2015
10 / 17

Bifunctional Proline Racemase/Hydroxyproline Epimerase
Table 2. Kinetic parameters for D-proline and C4DHyp.
Specific activity a (units/mg protein)
K
m (mM)
k
cat (min− 1
)
k
cat/Km (min− 1ꢀmM− 1
)
ProR
HypE
Enzymes
Substrates
TlProRWT
D-Proline
C4DHyp
D-Proline
C4DHyp
D-Proline
C4DHyp
D-Proline
C4DHyp
D-Proline
C4DHyp
D-Proline
C4DHyp
D-Proline
C4DHyp
D-Proline
C4DHyp
D-Proline
C4DHyp
3.45
3.48
0.0919 0.006
6.65 0.69
0.194 0.020
6.70 0.68
2.54 0.46
4.18 0.81
4.00 0.17
3.14 0.15
0.156 0.004
N.D.
133 2
1570 90
34.8 1.8
990 82
6.25 0.24
6.84 0.19
1.59 0.00
0.387 0.008
0.274 0.010
9.04 0.23
N.D.
45.1 b
230 12
191 5
TlProRW241F
TlProRL221H
TlProRF62S
TlProRF62V
TlProRF62S/L221H
FaProRWT
4.87
158
4.3
1.4
‒
0.681
41.8 2.7
17.3 2.7
6.66 1.31
1.55 0.09
0.299
0.111
0.0283
0.0186
0.0309
0.00661
>0.001
>0.001
0.164
0.859 0.012
1.41 0.03
N.D.
N.D.
N.D.
N.D.
‒
N.D.
N.D.
N.D.
1.17 0.09
N.D. c
7.09 0.12
N.D.
6.09 0.35
N.D.
‒
0.00234
0.141
FaProRF240W
CdProR
3.12 0.08
N.D.
7.21 0.06
N.D.
2.31 0.04
N.D.
‒
0.00230
3.92
0.612 0.158
N.D.
149 34
N.D.
245 8
‒
>0.001
N.D.
aUnder standard assay conditions in the “Materials and Methods”.
bRatio of D-proline to C4DHyp in kcat/Km.
cNot determined due to trace activity.
doi:10.1371/journal.pone.0120349.t002
hyperthermophilic archaeon Thermococcus sibiricus (TSIB_0633) and Thermococcus sp. ES1
(TES1_1056) showed >85% sequential identity (referred to as TlProR subfamily; red), and the
corresponding genes formed a similar gene cluster to T. litoralis, together with (putative) L-
ProDH and Pyr2C reductase genes (Fig. 1B). Although TlProR (subfamily) is not strongly re-
lated to any of the subclasses of the other members (30*40% sequence identity), a Protein-
BLAST analysis revealed that TlProR was a close, but distinct subfamily of the ProR subfamily;
the (putative) ligand binding sites were similar to those of ProR, but not HypE or T3LHypD, as
described below. Furthermore, we found that two putative proteins (Cys-Cys type) from the
hyperacidophilic archaeon F. acidarmanus (FaProR) and Ferroplasma sp. type II were located
between the TlProR and ProR subfamilies (purple in Fig. 6). Several halophilic archaea includ-
ing H. japonica (HjProR) (pink) also possessed similar types of proteins (genes), but formed a
distinct subfamily to any other subfamily, and were often clustered with putative bacterial type
(but not archaeal) L-ProDH (PutA) (Fig. 1B).
Characterization of FaProR and HjProR
To further estimate the evolutionary insight of archaeal ProR-like protein(s), FaProR and
HjProR were characterized enzymatically using purified recombinant proteins (Fig. 1C). The
racemization activity of proline was found to be significant in FaProR, although specific activity
was *10-fold lower than that of TlProR (Fig. 3). On the other hand, the kcat/Km value for
T4LHyp was *3-fold lower than that for L-proline, which was attributed to a 17-fold lower
kcat value, and we could not determine the kinetic parameters of the epimerization of C4DHyp
PLOS ONE | DOI:10.1371/journal.pone.0120349 March 18, 2015
11 / 17

Bifunctional Proline Racemase/Hydroxyproline Epimerase
Fig 5. Representative 1H NMR spectra for epimerase activity toward T4LHyp (A) and T3LHyp (B) by
TlProR. Asterisks are peaks derived from an internal standard. Left panels show the assignments of protons
in D2O. The dashed line box indicates the progressive loss of H1 peaks.
doi:10.1371/journal.pone.0120349.g005
to T4DHyp due to this low activity (Tables 1 and 2). Although similar results were obtained
from the HPLC analysis (Fig. 3), L- or D-hydroxyproline(s) was significantly converted to an-
other enantiomer(s) after being incubated for 12 h. These results suggested that the substrate
specificity of FaProR was more similar to the known ProR enzymes than that of TlProR, con-
firming a phylogenetic relationship, as described above.
Although HjProR could also catalyze proline racemization, (low) specific activity was un-
measurable possibly due to the partial proteolysis (Fig. 1CD) and/or the enzyme from extreme
halophilus: activation by salts is frequency necessary for full activity recovery [27]. Therefore,
substrate specificity was preliminarily estimated by HPLC analysis using samples that had been
incubated for 12 h (Fig. 3), and the enzyme could utilize both proline and hydroxyproline(s) as
a substrate, similar to TlProR.
PLOS ONE | DOI:10.1371/journal.pone.0120349 March 18, 2015
12 / 17

Bifunctional Proline Racemase/Hydroxyproline Epimerase
Fig 6. Phylogenetic tree of the ProR superfamily. The number on each branch indicates the bootstrap value. The circles, squares, and triangles at the end
of each branch are enzymes from bacteria, archaea, and eukaryotes, respectively. Proteins with asterisks were used for Fig. 7. Proteins in circles were
functionally characterized.
doi:10.1371/journal.pone.0120349.g006
Identification of amino acid residue(s) for discriminating between proline
and hydroxyproline
In the crystal structure of the enzyme-T4LHyp complex of HypE from Pseudomonas protegens
(PDB ID 4J9X) (33% identity with TlProR) [26], seven amino acid residues were detected close
to the ligands, except for two catalytic cysteine residues (referred to as sites 1*7) (Fig. 7).
Among them, sites 3, 6, and 7 were previously reported to be contained in two conserved mo-
tifs for the ProR superfamily (Met87-Cys-Gly-His90 and Asp247-Arg-Ser-Pro-Cys-Gly-Thr-
Gly254; the numbers for TlProR), and formed hydrogen bonds with the carboxyl group or pyr-
rolidine nitrogen atom (shaded in green) [13].
TlProR possessed a tryptophan residue (Trp241) at site 5 (shaded in pink), which was substi-
tuted to phenylalanine or cysteine/tyrosine in all known ProR (and also FaProR and HjProR)
and HypE enzymes, respectively. Therefore, we constructed TlProRW241F, TlProRW241C
,
PLOS ONE | DOI:10.1371/journal.pone.0120349 March 18, 2015
13 / 17

Bifunctional Proline Racemase/Hydroxyproline Epimerase
Fig 7. Partial multiple sequence alignment of deduced amino acid sequences of TlProR. A, B, and E are enzymes from archaea, bacteria, and
eukaryotes, respectively. Consensus segments of the ProR superfamily are shown as a line on the sequence. Catalytic cysteine and/or threonine residues
are shaded in red. Gray-shaded letters indicate highly conserved amino acid residues. Seven substrate binding sites are shaded in yellow and green, and the
former interact with the carboxyl group or pyrrolidine nitrogen atom. The tryptophan residue, which is important for the discrimination of substrates (Typ241 in
TlProR), is shaded in pink. Gray-shaded letters indicate highly conserved amino acid residues. The secondary structures of ProR from T. cruzi (PDB ID
1W61), α-helix (rectangles) and β-sheets (arrows), are shown under the sequence.
doi:10.1371/journal.pone.0120349.g007
TlProRW241Y, FaProRF241W, and AbHypEC226F mutants at an equivalent position. Of these,
only the TlProRW241F, FaProRF240W, and AbHypEC226F mutants successfully expressed in E.
coli cells, as well as the wild-type (WT) enzyme (Fig. 1C). TlProRW241F showed 5.3, 430, and
5.8-fold lower kcat/Km values for T4LHyp, T3LHyp, and C4DHyp respectively, mainly due to a
marked decrease in kcat values, whereas no significant effects were found in the kinetic con-
stants of proline (Tables 1 and 2). FaProRF241W showed similar kinetic constants for proline
and hydroxyproline to the WT enzyme. The AbHypEC226F mutant could utilize neither proline
nor hydroxyproline(s) as a substrate. These results clearly indicated that tryptophan in the
TlProR residue played important role(s) in both substrate specificity and structural folding,
and favorably interacted with both the 4 and 3-hydroxy groups of hydroxyproline.
Phenylalanine at site 1 for ligand binding may only exist in ProR(-like) enzymes including
TlProR, while histidine at site 4 could be specific for HypE. The three constructed TlProR mu-
tants, F62S, F62V, and L221H, which mimic the natural HypE enzyme, moderately shifted sub-
strate specificity toward 4-hydroxyproline (in particular, direction of L!D), and this was
mainly attributed to a marked decrease in kcat values; however, the prominent loss of both
ProR and HypE catalysis was also observed, and there was no synergistic effect in a double mu-
tant, F62S/L221H (Tables 1 and 2).
Discussion
The most interesting finding of the present study was the manner by which TlProR recognized
both proline and 4-hydroxyproline (and 3-hydroxyproline) as substrates. Goytia et al. [13] pro-
posed that several amino acid residues were responsible for this discrimination, in which two
aromatic phenylalanine residues of ProR (at sites 1 and 5 in Fig. 7) may play the most impor-
tant role(s) through hydrophobic interaction(s) between the enzyme and pyrrolidine ring of
proline. If this hypothesis is correct, a mutation at the corresponding sites in ProR or HypE
could significantly change each substrate specificity. However, when these amino acid residues
are exchanged between ProR and HypE, the resultant mutant(s) can generally utilize neither
proline nor hydroxyproline as a substrate: V60F for HypE from Pseudomonas aeruginosa [13],
F62S and F62V for TlProR (this study), and C226F for AbHypE (this study). On the other
hand, there is no doubt that TlProR (directly) evolved from an ancestor with phenylalanine at
site 5 such as FaProR (Fig. 6). In other words, the W241F mutation (but not W241C and
W241Y) mimicked the natural evolutionary process, through which no radical loss of activity
occurred.
PLOS ONE | DOI:10.1371/journal.pone.0120349 March 18, 2015
14 / 17

Bifunctional Proline Racemase/Hydroxyproline Epimerase
FaProR possessed the same ligand binding sites as natural ProR (Fig. 7). Therefore, it is un-
expected that the (hydrophobic and bulky) tryptophan at site 5 of TlProR played an important
role in the activity of hydroxyproline, which is more hydrophilic and bulkier than proline. No
equivalent tryptophan was detected in the other members of the ProR superfamily, and many
HypE enzymes including Sinorhizobium meliloti [28] possessed (the same aromatic and hydro-
philic) tyrosine residue at that position, strongly suggesting that substrate specificity had been
convergently acquired between TlProR and (natural) HypE. On the other hand, the environ-
ment in the active site pocket of TlProR seems to be suitable for recognizing hydroxyproline as
a substrate independently of the amino acid residue at site 5, because FaProR, which possesses
phenylalanine, but not tryptophan, at this position, as well as TlProR, shows similar IC50 value
for inhibition by PYC to TlProR (Fig. 4). Several bacterial ProR-like proteins of the Ser-Thr
type at the (putative) active sites (see in “Introduction”) have recently been functionally anno-
tated as T3LHypD (Type II in Fig. 6) [26]. The ligand binding sites (1, 2, 4, and 5) were
completely different from those of known T3LHypD (Type I) (Fig. 7). These results indicate
that convergent evolution frequently occurred in the ProR superfamily, and also that there was
more than one pattern for favorable binding of the same substrate. A previous study proposed
that the last universal common ancestor (well denoted “LUCA” or “LCA”) was a hyperthermo-
philic organism [29] such as T. litoralis (98°C of maximum temperature for the growth) [15],
and also that they possessed enzymes with significant broad substrate specificity [30]; there-
fore, the unique properties of TlProR may be (partially) derived from an ancient protein of the
ProR superfamily.
The physiological role of TlProR has not yet been established. Since T. litoralis possesses no
homologous gene to D-proline reductase, the physiological function of TlProR may be to epi-
merize hydroxyproline, but not racemize proline, confirming the evolutionary scenario pro-
posed above. T. litoralis significantly accumulated T4LHyp as an intercellular organic solute,
while other Thermococcus species including T. celer, T. stetteri, and T. zilligii did not [31]. Fur-
thermore, archaeal D-ProDH was previously shown to be capable of converting C4DHyp to
Pyr4H2C [19]; therefore, the homologous pathway involved in the metabolism of T4LHyp by
bacteria may also be operative in T. litoralis [7, 8]. In another hypothesis, Pyr5C was suggested
to be spontaneously taumerized to L-glutamate γ-semialdehyde via a “5-hydroxy-L-proline”
intermediate (Fig. 1A). If TlProR is also able to epimerize 5-hydroxy-L-proline to the D-enan-
tiomer due to broad substrate specificity, D-glutamate is finally produced via D-glutamate γ-
semialdehyde. In bacteria, D-glutamate was reported to be produced from L-glutamate by glu-
tamate racemase (EC 5.1.1.3), and used as a building block of cell-wall peptidoglycan [32]. Al-
though D-glutamate has also been detected in several archaea, the homolog from archaea
functions as an aspartate racemase (EC 5.1.1.13) [33]. Therefore, TlProR may be involved in an
alternative pathway of D-glutamate biosynthesis. The development of the gene disruption of
T. litoralis will be useful for understanding the physiological role(s) of the TlProR gene in more
detail.
Supporting Information
S1 Table. Primers used in this study. aLower case letters indicate additional bases for intro-
ducing the digestion sites of restriction enzymes in parentheses. bOnly sense primers are
shown. Underlining indicates mutated regions.
(PDF)
PLOS ONE | DOI:10.1371/journal.pone.0120349 March 18, 2015
15 / 17

Bifunctional Proline Racemase/Hydroxyproline Epimerase
Acknowledgments
We thank Dr. Takenori Satomura (University of Fukui, Japan) and Dr. Hiroshi Sekiya (Matsu-
yama University, Japan) for the gift of L-ProDH from A. pernix and D-ProDH from P. islandi-
cum, and genomic DNA of C. difficile 630, respectively. Our thanks extend especially to Mr.
Masashi Yoshida (Shinwa Chemical Industries, Kyoto, Japan) for his help in the HPLC analysis
of chiral proline derivatives.
Author Contributions
Conceived and designed the experiments: SW. Performed the experiments: SW YT. Analyzed
the data: SW HN. Contributed reagents/materials/analysis tools: SW YW. Wrote the paper:
SW.
References
1. Wu G, Bazer FW, Burghardt RC, Johnson GA, Kim SW, Knabe DA, et al. Proline and hydroxyproline
metabolism: implications for animal and human nutrition. Amino Acids. 2011; 40: 1053–1063. doi: 10.
1007/s00726-010-0715-z PMID: 20697752
2. Bouillaut L, Self WT, Sonenshein AL. Proline-dependent regulation of Clostridium difficile Stickland
metabolism. J Bacteriol. 2013; 195: 844–854. doi: 10.1128/JB.01492-12 PMID: 23222730
3. Jackson S, Calos M, Myers A, Self WT. Analysis of proline reduction in the nosocomial pathogen Clos-
tridium difficile. J Bacteriol. 2006; 188: 8487–8495. PMID: 17041035
4. Chamond N, Grégoire C, Coatnoan N, Rougeot C, Freitas-Junior LH, da Silveira JF, et al. Biochemical
characterization of proline racemases from the human protozoan parasite Trypanosoma cruzi and defi-
nition of putative protein signatures. J Biol Chem. 2003; 278: 15484–15494. PMID: 12735293
5. Reina-San-Martín B, Degrave W, Rougeot C, Cosson A, Chamond N, Cordeiro-Da-Silva A, et al. A B-
cell mitogen from a pathogenic trypanosome is a eukaryotic proline racemase. Nat Med. 2000; 6: 890–
897. PMID: 10932226
6. Chamond N, Cosson A, Coatnoan N, Minoprio P. Proline racemases are conserved mitogens: charac-
terization of a Trypanosoma vivax proline racemase. 2009;Mol Biochem Parasitol. 165: 170–179. doi:
10.1016/j.molbiopara.2009.02.002 PMID: 19428664
7. Watanabe S, Yamada M, Ohtsu I, Makino K. α-Ketoglutaric semialdehyde dehydrogenase isozymes in-
volved in metabolic pathways of D-glucarate, D-galactarate and hydroxy-L-proline: molecular and met-
abolic convergent evolution. J Biol Chem. 2007; 282: 6685–6695. PMID: 17202142
8. Watanabe S, Morimoto D, Fukumori F, Shinomiya H, Nishiwaki H, Kawano-Kawada M, et al. Identifica-
tion and characterization of D-hydroxyproline dehydrogenase and Δ1-pyrroline-4-hydroxy-2-carboxyl-
ate deaminase involved in novel L-hydroxyproline metabolism of bacteria: metabolic convergent
evolution. J Biol Chem. 2012; 287: 32674–32688. PMID: 22833679
9. Visser WF, Verhoeven-Duif NM, de Koning TJ. Identification of a human trans-3-hydroxy-L-proline
dehydratase, the first characterized member of a novel family of proline racemase-like enzymes. J Biol
Chem. 2012; 287: 21654–21662. doi: 10.1074/jbc.M112.363218 PMID: 22528483
10. Watanabe S, Tanimoto Y, Yamauchi S, Tozawa Y, Sawayama S, Watanabe Y. Identification and char-
acterization of trans-3-hydroxy-L-proline dehydratase and Δ1-pyrroline-2-carboxylate reductase in-
volved in trans-3-hydroxy-L-proline metabolism of bacteria. FEBS Open Bio. 2014; 4: 240–250. doi: 10.
1016/j.fob.2014.02.010 PMID: 24649405
11. Watanabe S, Tozawa Y, Watanabe Y. Ornithine cyclodeaminase/μ-crystallin homolog from hyperther-
mophilic archaeon Thermococcus litoralis functions as a novel Δ1-pyrroline-2-carboxylate reductase in-
volved in putative trans-3-hydroxy-L-proline metabolism. FEBS Open Bio. 2014; 4: 617–626. doi: 10.
1016/j.fob.2014.07.005 PMID: 25161870
12. Buschiazzo A, Goytia M, Schaeffer F, Degrave W, Shepard W, Grégoire C, et al. Crystal structure, cata-
lytic mechanism, and mitogenic properties of Trypanosoma cruzi proline racemase. Proc Natl Acad Sci
U S A. 2006; 103: 1705–1710. PMID: 16446443
13. Goytia M, Chamond N, Cosson A, Coatnoan N, Hermant D, Berneman A, et al. Molecular and structural
discrimination of proline racemase and hydroxyproline-2-epimerase from nosocomial and bacterial
pathogens. PLoS One. 2007; 2: e885. PMID: 17849014
14. Finlay TH, Adams E. Kinetic and structural studies of hydroxyproline 2-epimerase. J Biol Chem. 1970;
245: 5248–5260. PMID: 5469165
PLOS ONE | DOI:10.1371/journal.pone.0120349 March 18, 2015
16 / 17

Bifunctional Proline Racemase/Hydroxyproline Epimerase
15. Neuner A, Jannasch HW, Belkin S, Stetter KO. Thermococcus litoralis sp. nov.: a new species of ex-
tremely thermophilic marine archaebacteria Arch Microbiol. 1990; 153: 205–207.
16. Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning: a Laboratory Manual, 3rd ed. Cold Spring Har-
bor, NY: Cold Spring Harbor Laboratory; 2001.
17. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the folin phenol reagent. J
Biol Chem. 1951; 193: 265–275. PMID: 14907713
18. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Na-
ture. 1970; 227: 680–685. PMID: 5432063
19. Satomura T, Kawakami R, Sakuraba H, Ohshima T. Dye-linked D-proline dehydrogenase from hyper-
thermophilic archaeon Pyrobaculum islandicum is a novel FAD-dependent amino acid dehydrogenase.
J Biol Chem. 2002; 277: 12861–12867. PMID: 11823469
20. Sakuraba H, Satomura T, Kawakami R, Kim K, Hara Y, Yoneda K, et al. Crystal structure of novel dye-
linked L-proline dehydrogenase from hyperthermophilic archaeon Aeropyrum pernix. J Biol Chem.
2012; 287: 20070–20080. doi: 10.1074/jbc.M111.319038 PMID: 22511758
21. Watanabe S, Hiraoka Y, Endo S, Tanimoto Y, Tozawa Y, Watanabe Y. An enzymatic method to esti-
mate the content of L-hydroxyproline. J Biotechnol. 2015 doi: 10.1016/j.jbiotec.2015.01.026
22. Penning TM, Jez JM. Enzyme redesign. Chem Rev. 2001; 101: 3027–3046. PMID: 11710061
23. Kawakami R, Sakuraba H, Tsuge H, Goda S, Katunuma N, Ohshima T. A second novel dye-linked L-
proline dehydrogenase complex is present in the hyperthermophilic archaeon Pyrococcus horikoshii
OT-3. FEBS J. 2005; 272: 4044–4054. PMID: 16098188
24. Kawakami R, Sakuraba H, Ohshima T. Gene and primary structures of dye-linked L-proline dehydroge-
nase from the hyperthermophilic archaeon Thermococcus profundus show the presence of a novel het-
erotetrameric amino acid dehydrogenase complex. Extremophiles. 2004; 8: 99–108. PMID: 15064976
25. Berneman A, Alves-Ferreira M, Coatnoan N, Chamond N, Minoprio P. Medium/high throughput D-
amino acid oxidase colorimetric method for determination of D-amino acids. application for amino acid
racemases. J Microbial Biochem Technol. 2010; 2: 139–146.
26. Zhao S, Sakai A, Zhang X, Vetting MW, Kumar R, Hillerich B, et al. Prediction and characterization of
enzymatic activities guided by sequence similarity and genome neighborhood networks. Elife. 2014; 3:
e03275.
27. Madern D, Ebel C, Zaccai G. Halophilic adaptation of enzymes. Extremophiles. 2000; 4: 91–98. PMID:
10805563
28. White CE, Gavina JM, Morton R, Britz-McKibbin P, Finan TM. Control of hydroxyproline catabolism in
Sinorhizobium meliloti. Mol Microbiol. 2012; 85: 1133–1147. doi: 10.1111/j.1365-2958.2012.08164.x
PMID: 22804907
29. Akanuma S, Nakajima Y, Yokobori S, Kimura M, Nemoto N, Mase T, et al. Experimental evidence for
the thermophilicity of ancestral life. Proc Natl Acad Sci U S A. 2013; 110: 11067–11072. doi: 10.1073/
pnas.1308215110 PMID: 23776221
30. Jensen RA. Enzyme recruitment in evolution of new function. Annu Rev Microbiol. 1976; 30: 409–425.
PMID: 791073
31. Lamosa P, Martins LO, Da Costa MS, Santos H. Effects of temperature, salinity, and medium composi-
tion on compatible solute accumulation by Thermococcus spp. Appl Environ Microbiol. 1998; 64:
3591–3598. PMID: 9758772
32. Fisher SL. Glutamate racemase as a target for drug discovery. Microb Biotechnol. 2008; 1: 345–360.
doi: 10.1111/j.1751-7915.2008.00031.x PMID: 21261855
33. Long Z, Lee JA, Okamoto T, Sekine M, Nimura N, Imai K, et al. Occurrence of D-amino acids and a pyri-
doxal 5’-phosphate-dependent aspartate racemase in the acidothermophilic archaeon, Thermoplasma
acidophilum. Biochem Biophys Res Commun. 2001; 281: 317–321. PMID: 11181048
PLOS ONE | DOI:10.1371/journal.pone.0120349 March 18, 2015
17 / 17