Snapshots of the Catalytic Cycle of an O2, Pyridoxal Phosphate-Dependent Hydroxylase

Article abstract of DOI:10.1021/acschembio.8b00039

Enzymes that catalyze hydroxylation of unactivated carbons normally contain heme and nonheme iron cofactors. By contrast, how a pyridoxal phosphate (PLP)-dependent enzyme could catalyze such a hydroxylation was unknown. Here, we investigate RohP, a PLP-dependent enzyme that converts l-arginine to (S)-4-hydroxy-2-ketoarginine. We determine that the RohP reaction consumes oxygen with stoichiometric release of H₂O₂. To understand this unusual chemistry, we obtain ～1.5 ? resolution structures that capture intermediates along the catalytic cycle. Our data suggest that RohP carries out a four-electron oxidation and a stereospecific alkene hydration to give the (S)-configured product. Together with our earlier studies on an O₂, PLP-dependent l-arginine oxidase, our work suggests that there is a shared pathway leading to both oxidized and hydroxylated products from l-arginine.

Full text of DOI:10.1021/acschembio.8b00039

Subscriber access provided by MT ROYAL COLLEGE
Article
Snapshots of the catalytic cycle of an O2,
pyridoxal phosphate-dependent hydroxylase
Jason B. Hedges, Eugene Kuatsjah, Yi-Ling Du, Lindsay D. Eltis, and Katherine S Ryan
ACS Chem. Biol., Just Accepted Manuscript • DOI: 10.1021/acschembio.8b00039 • Publication Date (Web): 21 Feb 2018
Downloaded from http://pubs.acs.org on February 23, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Chemical Biology is published by the American Chemical Society. 1155 Sixteenth
Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 33
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
Snapshots of the catalytic cycle of an O , pyridoxal phosphate-
2
dependent hydroxylase
Authors
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
4
3
Jason B. Hedges, Eugene Kuatsjah, YiꢀLing Du, Lindsay D. Eltis, Katherine S.
1*
Ryan
Author Affiliations
1
2
3
Department of Chemistry, Genome Science and Technology Program, and Department
of Microbiology and Immunology, The University of British Columbia, Vancouver,
British Columbia, Canada.
4
Institute of Pharmaceutical Biotechnology, School of Medicine, Zhejiang University,
Hangzhou, China
*
Correspondence: ksryan@chem.ubc.ca
ABSTRACT
Enzymes that catalyze hydroxylation of unactivated carbons normally contain hemeꢀ and
nonꢀheme iron cofactors. By contrast, how a pyridoxal phosphate (PLP)ꢀdependent
enzyme could catalyze such a hydroxylation was unknown. Here, we investigate RohP, a
PLPꢀdependent enzyme that converts
Lꢀarginine to (S)ꢀ4ꢀhydroxyꢀ2ꢀketoarginine. We
determine that the RohP reaction consumes oxygen, with stoichiometric release of H O .
2
2
To understand this unusual chemistry, we obtain ~1.5 Å resolution structures that capture
intermediates along the catalytic cycle. Our data suggest that RohP carries out a fourꢀ
electron oxidation and a stereospecific alkene hydration to give the (S)ꢀconfigured
product. Together with our earlier studies on an O , PLPꢀdependent
Lꢀarginine oxidase,
2
our work suggests that there is a shared pathway leading to both oxidized and
hydroxylated products from
Lꢀarginine.
1
ACS Paragon Plus Environment

ACS Chemical Biology
Page 2 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
INTRODUCTION
3
Activation of carbonꢀhydrogen (CꢀH) bonds on sp ꢀhybridized carbons is an enduring
1
,2
challenge in chemical synthesis. Metalloenzymes, nature’s solution to this challenge,
have evolved the incredible capacity to catalyze the stereospecific functionalization of Cꢀ
H bonds under mild conditions. Common cofactors for such enzymeꢀcatalyzed reactions
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
are heme and nonꢀheme irons. For example, in the case of hydroxylation of
Lꢀarginine,
the Fe(II)ꢀ, αꢀketoglutarateꢀdependent enzymes VioC, YcfD, and OrfP catalyze
3
3–5
hydroxylations of the sp ꢀhybridized carbons of the Lꢀarginine side chain. By contrast,
organic cofactors such as pyridoxal phosphate (PLP) are not normally expected to
3
catalyze installation of hydroxyl groups onto unactivated sp ꢀhybridized carbons.
PLP is one of the most widely used cofactors for enzymatic manipulation of
6
amino acid substrates. In the course of catalysis, many PLPꢀdependent enzymes form a
carbanionic intermediate, called a quinonoid, which can result from deprotonation or
7
decarboxylation of an amino acidꢀPLP adduct. Normally, this quinonoid intermediate is
sequestered in the active site of the enzyme and thus protected from reaction with O2.
However, in several cases this quinonoid intermediate can be subject to offꢀpathway
8
reactions with electrophiles, including O2. The O ꢀdependent oxidative deamination
2
9
catalyzed by DOPA decarboxylase and the O ꢀdependent oxidative decarboxylation
2
10
catalyzed by ornithine decarboxylase are two wellꢀcharacterized examples of such
paracatalytic enzymatic activities. More recently, the diversity of PLPꢀdependent
enzymes has expanded with the discovery of enzymes that use O as a coꢀsubstrate to
2
perform their primary enzymatic activities. For example, both petunia
phenylacetaldehyde synthase and the celesticetin biosynthetic enzyme CcbF catalyze O2ꢀ,
1
1,12
PLPꢀdependent decarboxylationꢀcoupled oxidative deaminations,
and Cap15 is an O2ꢀ
,
PLPꢀdependent monooxygenaseꢀdecarboxylase that generates uridineꢀ5'ꢀcarboxamide
13
on the pathway to capuramycin.
Intriguingly, the scope of such O ꢀ, PLPꢀdependent reactions is not limited to
2
oxidative decarboxylations but also includes functionalization of unactivated CꢀH bonds.
In 1977, Eguchi and coworkers purified an active fraction from Streptomyces eurocidicus
SF 506 that could convert Lꢀarginine to 2ꢀketoarginine and 4ꢀhydroxyꢀ2ꢀketoarginine
1
4
using PLP and O . This work tantalizingly suggested that a PLPꢀdependent enzyme
2
2
ACS Paragon Plus Environment

Page 3 of 33
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
could use O to catalyze a reaction that resulted in hydroxylation of an unactivated carbon
2
center. However, the identity of the enzyme, the configuration of its product, and its
likely mechanism involved remained unknown. More recently, Silvaggi and coworkers
revealed that the foldꢀtype I PLPꢀdependent enzyme MppP is responsible for production
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
of 2ꢀketoarginine (2) and 4ꢀhydroxyꢀ2ꢀketoarginine (4, configuration unknown) from
Lꢀ
arginine on the pathway to the nonꢀproteinogenic amino acid ꢀenduracididine, a
L
1
5
precursor to the antibiotic mannopeptimycin. They demonstrated that MppP consumes
O , suggesting that this enzyme has evolved the ability to activate O for a challenging
2
2
hydroxylation reaction. They also solved two crystal structures of MppP. One of these
structures shows PLP bound as an internal aldimine to Lys221, and the second structure
has
Dꢀarginine, which is not a substrate, bound in an external aldimine with PLP.
However, the Nꢀterminus is disordered and the active site is exposed to solvent in all but
one chain of the holoꢀenzyme, where it forms a small αꢀhelix that points away from the
active site and into the solvent. The lack of a closed active site and the lack of structures
with native substrates bound in the active site raise questions about how catalysis might
proceed past formation of the external aldimine. Furthermore, the specific role of oxygen
in this reaction remains unknown.
In our group, we discovered a predicted foldꢀtype I PLPꢀdependent enzyme Ind4
16,17
from the indolmycin biosynthetic pathway.
This enzyme, like MppP, catalyzes the
twoꢀelectron oxidation of ʟꢀarginine to give 2ꢀketoarginine (2) (Figure 1). But, unlike
MppP, Ind4 also catalyzes the fourꢀelectron oxidation of ꢀarginine to give
L
didehydroarginine (2ꢀaminoꢀ5ꢀguanidinopentaꢀ2,4ꢀdienoate, 8a) (Figure S1). The imine
tautomer (8b) of this product can be intercepted by the downstream enzyme Ind5 for a
stereospecific NADHꢀdependent reduction to give Dꢀdehydroarginine (9) (Figure S1). In
the absence of Ind5, the imine tautomer is hydrolyzed to compound 6, which reacts with
H O released from the reaction to give compound 7 (Figure 1). As we unraveled this
2
2
enzymology, we were curious whether enzymes similar to Ind4 are more widely
distributed in natural product pathways, and we identified a putative biosynthetic gene
cluster encoding an Ind4 homolog. Here, we report our discovery and characterization of
the new Ind4 homolog, which we name RohP, and we demonstrate that it catalyzes an
MppPꢀlike reaction. To provide a mechanistic framework for understanding how RohP
3
ACS Paragon Plus Environment

ACS Chemical Biology
Page 4 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
catalyzes a hydroxylation – instead of an Ind4ꢀlike oxidation – we carry out extensive
mass spectral, kinetic, stoichiometric, and Xꢀray crystallographic experiments. Our work
suggests that both the RohP hydroxylase and the Ind4 oxidase catalyze challenging
reactions of
crystallographic and stoichiometric studies support a mechanism where both the oxidase
and hydroxylase carry out an O ꢀdependent fourꢀelectron oxidation on ꢀarginine, giving
Lꢀarginine through similar oxidative pathways. In particular, our
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
L
2
a PLPꢀtethered didehydroarginine. Whereas Ind4 then releases didehydroarginine as a
product, RohP additionally catalyzes a stereospecific alkene hydration on the PLPꢀ
tethered didehydroarginine, using water to give the hydroxylated product.
RESULTS AND DISCUSSION
RohP is an
L-arginine hydroxylase In our previous work, we discovered the O , PLPꢀ
2ꢀ
dependent oxidase Ind4, which generates the conjugated didehydroarginine (8a) from
Lꢀ
arginine (1) (Figure S1). We were curious whether similar enzymes are found in other
bacteria. Using BLASTꢀP to search for similar enzymes in other bacteria, we identified
an enzyme (accession number: AEW92768.1) from Streptomyces cattleya NRRL 8057
1
7
with 43% sequence identity to Ind4, 40% sequence identity to the Ind4ꢀlike enzyme
17
15
Pel4, 32% sequence identity to MppP, and 31% sequence identity to the MppPꢀlike
18
enzyme EndP (Figure S2). This newly identified enzyme is encoded in a putative
biosynthetic gene cluster unrelated to the indolmycin, mannopeptimycin, or other
characterized gene clusters, and the function corresponding to this cluster is unknown
(
Figure S3). This close sequence identity of the encoded enzyme to both Ind4 and MppP
suggests that it could have either Ind4ꢀlike or MppPꢀlike activity, or perhaps catalyze a
different reaction on ꢀarginine. As a first step in elucidating the product of this gene
L
cluster, we determined the function of this enzyme. Based on the sequencing data, this
enzyme is annotated with ~20 amino acid truncation in the Nꢀterminus when compared to
the other, characterized enzymes (Figure S2). We purified the recombinant enzyme from
E. coli and tested whether it is also active on
enzyme was inactive on ꢀarginine as purified or with exogenous PLP (Figure S4A).
After determining that the enzyme was inactive with ꢀarginine, we reꢀsequenced the
Lꢀarginine. However, we found that the
L
L
upstream region of DNA from S. cattleya, and we observed that one cytosine was missing
4
ACS Paragon Plus Environment

Page 5 of 33
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
from the deposited sequence. With the correct number of cytosines, an alternate
annotation of the gene would have a different start site and encode an enzyme 25 amino
acids longer that aligns along the full length of MppP, Ind4, and the other characterized
enzymes (Figure S2).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
We recloned the gene including the new start site into pET28a, purified the
enzyme from E. coli, and assayed the enzyme with
we derivatized the reaction with dansyl chloride and analyzed it by HPLC. This
experiment revealed that the ꢀarginine was consumed in the overnight assay (Figure
S4B), suggesting that the enzyme – like Ind4, Pel4, and MppP – reacts directly with
Lꢀarginine. After 16 h of incubation,
L
Lꢀ
arginine in aerobic conditions, without the need for an exogenously provided ketoꢀacid
for regeneration of the PLP cofactor. To determine the product(s) of this enzyme, we first
+
employed ESIꢀMS analysis to reveal that four new ions were generated: [M+H] 146,
+
+
+
[M+H] 162, [M+H] 174, and [M+H] 190 (Figure 2). Previously, we had observed both
+
+
the [M+H] 146 and the [M+H] 174 ions in the reaction of Lꢀarginine with Ind4. We
+
previously assigned the ion at [M+H] 174 as 2ꢀketoarginine (2), and we showed that the
+
ion at [M+H] 146 was 4ꢀguanidinobutyric acid (3), arising from the nonꢀenzymatic
1
7,19,20
+
decarboxylation of 2 with H O produced in the course of the reaction.
The [M+H]
2
2
ion of 190 that we observe in this reaction could correspond to the MppP product 4ꢀ
14,15
hydroxyꢀ2ꢀketoarginine (4, configuration unknown),
which could decarboxylate in
the presence of H O to give 3ꢀhydroxyꢀ4ꢀguanidinobutyric acid (5, configuration
2
2
+
+
unknown), having an [M+H] ion of [M+H] 162.
To determine whether these ions correspond to the previously identified products,
we added catalase to the reaction mixture to consume any H O produced. With catalase
2
2
+
+
present, we only observed the ions at [M+H] 174 and [M+H] 190 (Figure 2), supporting
that 2 and 4 are the initial products of the enzymatic reaction that decarboxylate in the
presence of H O to give 3 and 5, respectively. Furthermore, using high resolution ESIꢀ
2
2
MS to analyze all four products, we were able to confirm their elemental compositions,
which correspond to the formulas for the proposed products (Table S1). To further
validate the structures of these molecules, the reaction of Lꢀarginine with the S. cattleya
enzyme was scaled up to provide sufficient material for NMR analysis. This reaction was
carried out without catalase, and the 6 h incubation time allowed complete conversion of
5
ACS Paragon Plus Environment

ACS Chemical Biology
Page 6 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
the initial products to the corresponding decarboxylated molecules. A combination of Hꢀ
13
1
1
1
13
1
13
,
Cꢀ, Hꢀ H COSY, Hꢀ C HSQC, and Hꢀ C HMBC NMR were employed to
characterize the products. The analysis identified 3 and 5 in the reaction mixture (Figure
S5), suggesting that the products of the reaction are 2 and 4, which nonꢀenzymatically
convert to 3 and 5, respectively, in the presence of H O . Despite our NMR and mass
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
spectral analysis, the configuration of the hydroxyl in 5 (and by, extension, in 4)
remained unknown. Because this enzyme we identified catalyzes hydroxylation on an
arginineꢀderived molecule, we named it RohP to indicate its substrate (Lꢀarginine = R),
its activity (hydroxylation = OH), and its related activity to the “P” enzymes EndP and
MppP.
Kinetics and stoichiometry of the RohP-catalyzed reaction We hypothesized that
RohP requires O for activity the same way Ind4 and MppP do. Accordingly, we carried
2
out the RohP reaction with
L
ꢀarginine in a series of buffers and monitored the O2
consumption using an oxygraph (Figure S6A). We observed an O consumption rate of
2
ꢀ1
ꢀ1
1
5 µM min in 40 mM MOPS pH 7.2, a rate that decreased to 8 ꢀM min in the presence
of catalase (Figure S6B). This approximate halving of the rate of O consumption in the
2
presence of catalase – an enzyme that converts two molecules of H O to one molecule of
2
2
O – suggests that the RohPꢀcatalyzed reaction consumes O with stoichiometric
2
2
conversion of O to H O . To further confirm the stoichiometry of the reaction, we
2
2
2
ꢀ1
incubated 2.5 ꢀM RohP with 300 ꢀM
Horseradish Peroxidase (Typeꢀ1) and 1 mM ABTS and quenched the reaction when 100
M of O was consumed. In this condition, we observed the production of 96 ± 6 ꢀM of
Lꢀarginine in the presence of 0.1 mg mL
ꢀ
2
H O , indicating stoichiometric conversion of O to H O . We then determined that 77 ±
2
2
2
2
2
4
ꢀM of ꢀarginine was consumed when 100 ꢀM of O was consumed. This result
L
2
suggests there are two reaction pathways, as we observed for Ind4. In one pathway,
arginine undergoes a twoꢀelectron oxidation, with consumption of one equivalent of O2.
In the second pathway, arginine undergoes a fourꢀelectron oxidation, with consumption
of two equivalents of O . The observed O :Lꢀarginine stoichiometry can be rationalized if
2
2
the first, less oxidizing pathway consumes twice as much
Lꢀarginine as the more
oxidizing pathway. That is, for every 100 µM of O consumed in total, 50 µM is
2
6
ACS Paragon Plus Environment

Page 7 of 33
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
consumed by the first pathway to convert 50 µM of
by the more oxidizing pathway to convert 25 µM of
L
ꢀarginine, and 50 µM is consumed
Lꢀarginine. The first pathway leads to
2
, as observed for Ind4, whereas the second pathway should give a more oxidized
product, such as 4.
We then determined steadyꢀstate kinetic parameters for the RohPꢀcatalyzed
reaction. We determined that in the presence of airꢀsaturated buffer, RohP has a K = 40
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
m
ꢀ1
±
10 µM and a kcat = 7.8 ± 0.6 min for
Lꢀarginine (Figure S6C). These values are
consistent with those previously reported for the related enzymes Ind4 (K = 69 ± 6 µM,
m
ꢀ1
ꢀ1
kcat = 3.5 ± 0.1 min ) and MppP (K = 50 ± 8 µM and k = 13.2 ± 0.6 min ).
m
cat
Additionally, we determined that in the presence of saturating amounts of
Lꢀarginine,
ꢀ1
RohP has a K = 78 ± 6 µM and a k = 10.2 ± 0.2 min for O (Figure S6D), which can
m
cat
2
be compared to the values previously obtained for Ind4 (K = 90 ± 10 µM, k = 4.6 ±
m
cat
ꢀ1
0.1 min ).
The hydroxyl group in the RohP product derives from water, not oxygen or H O
2
2
Our work demonstrates stoichiometric conversion of O to H O , meaning there are no
2
2
2
remaining oxygen atoms that could be incorporated into the product from either O or
2
H O . Thus, these stoichiometry results suggest that the hydroxyl oxygen atom in 4ꢀ
2
2
hydroxyꢀ2ꢀketoarginine derives from H O. To further interrogate the source of this
2
1
8
oxygen atom, we carried out the RohP reaction in 50% H O in the absence of catalase
2
(Figure 2) and analyzed the results by LCꢀMS. Our results show ten unique ions. Four of
+
these ions, those at [M+H] 146, 162, 174, and 190, are also present in the reaction in
unlabeled water, and correspond to unlabeled compounds 3, 5, 2, and 4, respectively.
+
Additionally, four ions, those at [M+H] 148, 164, 176, and 192, are two mass units
1
8
higher, and should arise from hydrolysis of the likely imine products by H O, giving the
2
+
corresponding labelled carbonyl oxygens. Finally, the remaining two ions, [M+H] 166
+
+
and 194, are 4 Da heavier when compared to [M+H] 162 and [M+H] 190, meaning two
labeled oxygens are present. One label should result from hydrolysis of the imine product
1
8
by H O, giving a labeled carbonyl, and the second label should result from
2
18
incorporation of H O into the hydroxyl group. Altogether, our stoichiometric and
2
labeling studies support that the hydroxyl oxygen derives from water, not O or H O .
2
2
2
7
ACS Paragon Plus Environment

ACS Chemical Biology
Page 8 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
UV-Visible spectrum of RohP To further investigate the RohPꢀcatalyzed reaction, we
analyzed the UVꢀVisible spectra of RohP. Initially, the spectrum of RohP (30 µM)
contains a single peak with a λmax at 422 nm, typical of an enzymeꢀPLP internal aldimine.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
With the addition of excess
Lꢀarginine (300 µM) we observed a decrease in the
absorbance at 422 nm (Figure S7). At the same time, a peak at 515 nm, which is
diagnostic of a quinonoid, increased over the 10 min incubation. Finally, a peak at 563
nm increased initially and then decreased after 5 min over the course of the experiment.
Previously, we assigned a peak at 567 nm from the Ind4 reaction as a more conjugated
quinonoid intermediate, and Silvaggi and coworkers assigned a similar peak at 560 nm
from the MppP reaction as a more conjugated quinonoid intermediate. Our work here
suggests that RohP, like Ind4 and MppP, may transition through a more conjugated
quinonoid intermediate with a λmax at 563 nm.
Crystal structure of holo-RohP Our stoichiometric, mass spectral, and spectroscopic
results suggest that RohP consumes two molecules of O to carry out a fourꢀelectron
2
oxidation and incorporates water to yield 4ꢀhydroxyꢀ2ꢀketoarginine. However, several
questions remained unanswered. First, what is the stereochemistry of the 4ꢀhydroxyꢀ2ꢀ
ketoarginine product? Second, what intermediates arise in such a reaction scheme? To
address these questions, we turned to an Xꢀray crystallographic investigation of RohP.
We obtained four Xꢀray crystallographic structures of RohP at resolutions from
1
.50 to 1.55 Å (Tables S1 and S2). Each of these structures crystallized as homodimers in
space group C2. First, we determined the initial structure of holoꢀRohP by soaking RohP
crystals with 5 mM PLP for 4 h. We used MppP (PDB: 5DJ1) as a model for molecular
replacement. HoloꢀRohP exhibits an overall structure that is typical of other foldꢀtype I
PLP enzymes and is very similar to MppP (RMSD of 1.215 Å for Cα pairs, Figure S8A).
The RohP monomer consists of a large domain (residues 25ꢀ265) and a small domain
(residues 266ꢀ393) both of which exhibit an αꢀβꢀα motif. Active sites are sandwiched
between the large and small domains in individual monomers and exposed to solvent.
Unlike other PLP enzymes of foldꢀtype I and similar to MppP, there are only minor
contributions from one monomer to the active site of the other monomer. As was the case
8
ACS Paragon Plus Environment

Page 9 of 33
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
for MppP, the Nꢀterminal of the RohP holoenzyme is incomplete and missing Nꢀterminal
residues 1ꢀ25.
The RohP internal aldimine is formed between Lys235 and PLP, as was observed
for Lys221 and PLP in MppP. Furthermore, all of the same residues that have stabilizing
interactions with the internal aldimine in MppP are also observed in RohP (Figure S8B;
Table S3). The conserved residue Lys243 (Lys229 in MppP) provides positive charge to
stabilize the PLP phosphate. The position of Ser95 (Ser91 in MppP) also causes the
phosphate to rotate away from the plane of the pyridine ring, as observed in MppP.
Additional interactions serve to stabilize the pyridine ring of PLP. Asp198, conserved in
foldꢀtype I PLPꢀenzymes is located near the pyridinyl nitrogen and stabilizes positive
charges on the pyridinium cation through a hydrogen bond (Figure S9A). Asn167
hydrogen bonds with the hydroxyl group of the PLP pyridine ring. As is the case with
many PLP enzymes, an aromatic residue, Phe119, is located ~3.5 Å above the pyridine
ring. Finally, the phosphate also forms a waterꢀmediated hydrogenꢀbonding network with
Asp232 (Asp218 in MppP) and Tyr92 (Tyr88 in MppP) of the other monomer of the
homodimer. This structure of RohP provided a good initial model and was used to build
structures produced from subsequent experiments.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Trapping RohP the first quinonoid The slow rate of RohP catalyzed oxidation of
arginine afforded the opportunity to capture intermediates in the catalytic cycle by
aerobic soaking of ꢀarginine. A RohP quinonoid intermediate structure was produced by
soaking a crystal with 22 mM ꢀarginine for 90 s. In the resulting structure, there is no
evidence of a Schiff base linkage between Lys235 and PLP. Instead, positive F ꢀF
Lꢀ
L
L
o
c
electron density in the shape of arginine was found to extend from the C4' aldehyde of
PLP. Both the external aldimine (EA1) and quinonoid (Q1) intermediate were modelled
into the electron density (Figure S10). Although both EA1 and Q1 fit well to the density,
Q1 was modelled into the final structure (Figure 3A) because of its fit to the available
density and our observation that a quinonoid intermediate with an absorbance at 515 nm
accumulates in the reaction of RohP with Lꢀarginine (Figure S7), suggesting that Q1 is
more stable than EA1. Despite the formation of a quinonoid, the overall structure is very
similar to that of holoꢀRohP, with the two structures superposing with an RMSD of 0.106
9
ACS Paragon Plus Environment

ACS Chemical Biology
Page 10 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Å across Cα pairs. In the active site, only Asn121 exhibits a minor change in
conformation, rotating towards the quinonoid (Figure S9B). Furthermore, when the
quinonoid is present, the guanidium of Arg367 forms a salt bridge with the carboxylic
acid of the arginine substrate. Additionally, Leu266 of the second monomer, which is
unchanged in position relative to holoꢀRohP, now forms part of the periphery of the
active site and helps to orient the guanidium end of the quinonoid. Because the Nꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
terminal amino acids are disordered in this structure, the sidechain of the Lꢀarginine
substrate is pointed into solvent. As such, Cγ is 3.9 Å from His34 (Figure 4B), which is a
residue conserved among RohP, MppP, and Ind4 (Figure S2) but not observed in other
foldꢀtype I aminotransferases.
Trapping of a more conjugated quinonoid orders the N-terminus of RohP When we
soaked a RohP crystal with 4 mM PLP overnight and 10 mM Lꢀarginine for 5 min, we
observed the crystal change from yellow to red (Figure S11). We cryoprotected and flashꢀ
froze this red crystal. Solving the structure of this red crystal again revealed unknown
positive F ꢀF electron density in the shape of arginine in the active site. Initially, several
o
c
different intermediates were built using Phenix eLBOW and then modelled into the
available F ꢀF density (Figures S12AꢀD). However, none of these intermediates gave a
o
c
satisfactory fit to the available density. Thus, we employed an alternate approach,
2
1
utilizing ARP/wARP to build a ligand structure based upon the unknown density. The
initial ARP/wARP output was adjusted in COOT, refined and found to match available
omit density well (Figure S12E). This modelled ligand has bond lengths that closely
match what would be expected for the more conjugated quinonoid (Q2) intermediate,
containing a double bond between Cβ and Cγ positions of the arginine (Figure S12F). To
interrogate whether modeling Q2 was reasonable for such a red crystal, we carried out
spectroscopic analysis on other frozen red crystals using the microspectrophotometer
outfitted on beamline 9ꢀ2 at the Stanford Synchrotron Radiation Lightsource. This work
revealed that such red crystals have peaks with a λmax of 515 and 563 nm (Figure S13A),
matching the peaks observed by UVꢀVisible spectroscopy in solution (Figure S7). We
solved the structure of one of these crystals, which diffracts to lower, 2.0 Å resolution,
revealing that it too contains a PLP adduct (Figures S13B, C). As we observe a more
10
ACS Paragon Plus Environment

Page 11 of 33
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
conjugated quinonoid intermediate in the UVꢀvisible spectra both in solution and in
crystallo (Figures S7 and S13A), it is likely that Q2 is a more stable intermediate than a
more conjugated external aldimine (EA2). Therefore, we built Q2 into the final structure
based on the ARP/wARP coordinates (Figures 3B and S12E). However, the ~1.5 Å
resolution of this structure does not allow us to unambiguously assign the pattern of
double bonds. Furthermore, a mixture of intermediates may contribute to the density we
observed.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In this structure of the Q2 intermediate, electron density for residues 13ꢀ25 of the
Nꢀterminus is present, forming a small αꢀhelix that closes off the active site, isolating the
active site from bulk solvent (Figure 3B). With the ordering of the Nꢀterminus, the amide
nitrogen of Leu16 forms a hydrogen bond with Asp120. Additionally, the hydroxyl
sidechain of Thr17 forms a hydrogen bond with the guanidium group of Q2 and pushes
the guanidium deeper into the active site where it forms additional hydrogen bonds with
Ser95 and the carbonyl oxygen of Val264. The conserved residue Glu20 becomes
ordered near the carboxylic acid and Cβ of Q2. Phe119 also exhibits a dual conformation,
with the additional conformer rotating from its previous position above the pyridine ring
to a position ~3.5 Å above the carboxylic acid of Q2. Asn121 also moves ~180° from its
previous position to partially occupy the area that was vacated by the rotation of Phe119.
Collectively, these changes push Q2 deeper into the active site, into a position where
Glu20 is 4.0 Å away from Cβ and His34 is 3.2 Å and 3.5 Å away from Cβ and Cδ,
respectively, of the arginine substrate (Figure 4C).
Structure of RohP-product complex We obtained a final snapshot of the RohP catalytic
cycle by soaking RohP crystals with 10 mM
Lꢀarginine overnight. During this
experiment, the crystal changed from yellow to red and then back to yellow. We
cryoprotected and flashꢀfroze the resulting yellow crystal. The structure we obtained has
Nꢀterminal residues 14ꢀ25 present, though the density of these residues in chain B is
weaker. Therefore, these residues were modelled with an occupancy of ~0.7ꢀ0.8 in chain
B, compared to full occupancy in chain A. The active site of the enzyme also remains
largely unchanged compared to the Q2 structure, with Asn121 remaining flipped relative
to its initial position, and Phe119 exhibiting a dual conformation away from the pyridine
11
ACS Paragon Plus Environment

ACS Chemical Biology
Page 12 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ring of PLP. However, in this structure, PLP has again formed an internal aldimine with
Lys235 (Figure S9D). Strong F ꢀF density was present above the internal aldimine,
o
c
displaying a shape consistent with a product containing a 4ꢀhydroxy group. Both the
enamine and hydrolysis product of the imine tautomer were modelled into the available
density to determine their fit (Figure S14); the larger negative density over the enamine
double bond led us to model 4ꢀhydroxyꢀ2ꢀketoarginine into the final structure. We
suspect that the dynamic nature of the Nꢀterminus should allow for hydrolysis of the
enamine product over the extended incubation period. As the stereochemistry of the 4ꢀ
hydroxyl was still undetermined, both the R and S enantiomers of the molecule were
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
modelled into the positive F ꢀF density present in the active site (Figure 5). Refinement
o
c
of the Rꢀenantiomer produced strong negative F ꢀF density around the hydroxyl group,
o
c
while the same refinement of the Sꢀenantiomer produced no negative F ꢀF density
o
c
around the hydroxyl group. Identical results were also observed for both enantiomers of
the enamine product (Figure S14). These results support that RohP catalyzes the
production of (S)ꢀ4ꢀhydroxyꢀ2ꢀketoarginine (4) from Lꢀarginine.
Characterization of the RohP-His34Ala variant Based upon our crystallographic
results, His34 appears likely to be involved in the installation of the 4ꢀhydroxyl group,
due to its position relative to the quinonoid intermediates (Figure 4). To probe the
function of His34 we created a His34Ala variant of RohP with siteꢀdirected mutagenesis.
ESIꢀMS analysis was again employed to determine the product(s) of the His34Ala
variant. The production of 4 was abolished, however, the variant was still able to produce
2
. The corresponding decarboxylation product 3 was also detected (Figure S15). That this
variant was only able to perform one oxidation, and no other intermediates were detected,
suggests that His34 may also be involved in catalyzing the second oxidation of the
arginine substrate and possibly the final hydration reaction.
Discussion
In this work, we investigate RohP, an enzyme that uses pyridoxal phosphate to catalyze
the transformation of
Lꢀarginine and O to two products: 2ꢀketoarginine (2) and (S)ꢀ4ꢀ
2
hydroxyꢀ2ꢀketoarginine (4). While previous studies by Eguchi and Silvaggi highlighted
12
ACS Paragon Plus Environment

Page 13 of 33
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
this enzymatic reaction, the key question of how an enzyme uses PLP and O to
2
3
hydroxylate an unactivated, sp ꢀhybridized carbon remained unaddressed. Here, we use
detailed mass spectral, kinetic, stoichiometric, and Xꢀray crystallographic analysis to
build a firm mechanistic framework for understanding this group of PLPꢀ, O ꢀdependent
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
hydroxylases. Our work demonstrates that RohP and the oxidase Ind4 share many key
features: both stoichiometrically convert O to H O , both generate quinonoid and
2
2
2
conjugated quinonoid intermediates, and both produce the less oxidized product, 2ꢀ
ketoarginine, which results from hydrolysis of the corresponding imine. The enzymes
differ only in whether they produce didehydroarginine or (S)ꢀ4ꢀhydroxyꢀ2ꢀketoarginine.
From our Xꢀray crystal structures containing trapped quinonoid and conjugated
quinonoid intermediates, along with a structure with product bound, we unveil a shared
mechanism whereby both enzymes can catalyze a fourꢀelectron oxidation of Lꢀarginine.
However, only RohP can utilize water to carry out a stereospecific alkene hydration.
Our crystallographic work shows that RohP forms an external aldimine at the αꢀ
amino group of arginine, exactly as would be expected for a PLPꢀdependent
aminotransferase. In our mechanistic proposal (Figure 6), we suggest that the Cα proton
of the external aldimine (EA1) is abstracted by Lys235, with the resulting anionic species
stabilized by formation of a quinonoid intermediate (Q1). Q1 is a typical intermediate in
PLPꢀdependent aminotransferases, one that we also observe accumulate by UVꢀVisible
spectroscopy. We then propose that Q1, as for Ind4, reacts directly with O , oxidizing the
2
bound amino acid and releasing H O . The exact mechanism of how O interacts with Q1
2
2
2
is currently unknown, as is the mechanism for release of H O . After oxidation, the
2
2
resulting external aldimine intermediate (EA2) forms. This intermediate can undergo one
of two fates: it can either be attacked by Lys235, which will reform the internal aldimine
and release an enamine product, which tautomerizes to the imine and is then hydrolyzed
to give 2. Alternatively, the oxidized intermediate can remain in the active site. If it
remains in the active site, the Cγ proton can undergo rapid deprotonation by the adjacent
His34, shuttling electron density into the cofactor to give the more conjugated quinonoid
(
Q2) intermediate. Then, a second molecule of O could react with Q2, again oxidize the
2
substrate, and release a second molecule of H O . Now, the PLPꢀtethered
2
2
didehydroarginine could undergo hydration. One possible scenario is that the double
13
ACS Paragon Plus Environment

ACS Chemical Biology
Page 14 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
bond between CγꢀCδ is first protonated at the Cδ position. The resulting carbocation at
Cγ could be stabilized through resonance with density from the PLP pyridine ring. Now
His34 deprotonates an adjacent water to insert the hydroxyl group at Cγ and give the
resulting PLPꢀtethered final product. The stereospecificity of the hydration appears to be
promoted by the positioning of His34, which is positioned to the si face of the alkene. At
the same time, the phosphate of PLP occludes solvent access to the re face of the alkene.
To complete the catalytic cycle, the 4ꢀhydroxy product is then released when Lys235
attacks to release the enamine with the concomitant formation of the RohPꢀPLP internal
aldimine. The imine tautomer is then hydrolyzed by water to produce the final product 4.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Our work raises exciting questions about O , PLPꢀdependent oxidases. First, what
2
are the structural features that distinguish arginine oxidases like Ind4 from arginine
hydroxylases like RohP and MppP? We suggest that the remarkable feat of RohP –
installation of a hydroxyl group – could be the stereospecific hydration of a PLPꢀtethered
didehydroarginine. If this possibility is true, what causes the hydroxylase RohP to
catalyze the hydration, whereas the oxidase Ind4 releases didehydroarginine as a product?
The two types of enzymes have all residues conserved in their active sites – including
His34 – suggesting that other residues will be key to determining the product outcome
(
Table S3). An Ind4 structure will be essential for pinpointing which residues are critical
to determining product outcome. Second, in both RohPꢀlike and Ind4ꢀlike enzymes, 2 is
produced. Is the production of 2 an unavoidable waste product for such ꢀarginine, PLPꢀ,
L
O ꢀdependent enzymes, or is there a purpose for production of 2? Elucidation of the full
2
biosynthetic pathway will begin to address this issue. Finally, an unresolved question is
why a select group of PLPꢀdependent enzymes are able to use O to catalyze oxidation
2
8
,11,12
reactions and how O is activated during catalysis.
Answers to these questions await
2
further study.
METHODS
General methods Primers were purchased from Integrated DNA Technologies. DNA
sequencing was carried out by NAPS Unit DNA Sequencing Facility (The University of
British Columbia). Reagents were purchased from Anatrace, Bio Basic Inc., Gold
14
ACS Paragon Plus Environment

Page 15 of 33
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
Biotechnology, Hampton Research, New England Biolabs (NEB), Thermo Fisher
Scientific Canada, and VWR International.
Cloning, expression, and purification The gene rohP, which encoded RohP as originally
annotated, was amplified from genomic DNA of S. cattleya (NRRL 8057, DSM 46488)
by the polymerase chain reaction using the primers RohPꢀF 5′ꢀ
AGCAGCCATATGAAGTACAACCTCGCCGACGCCꢀ3′ (NdeI site underlined) and
RohPꢀR 5′ꢀAGTAGTCTCGAGTCAGCGGCCATGGCGGTCꢀ3′ (XhoI site underlined).
The reꢀannotated rohP, which coded for the active form of RohP with an intact Nꢀ
terminus, was similarly amplified using the primers RohPꢀFꢀ2 5′ꢀ
AGCAGCCATATGCACCCGCAAGꢀCGACCꢀ3′ and RohPꢀR. The products were
digested with NdeI and XhoI, and ligated into similarly digested plasmid pET28a (EMD
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Millipore) to produce a Nꢀterminal His ꢀtagged protein. Siteꢀdirected mutagenesis of
6
RohP was performed using the Q5 SiteꢀDirected Mutagenesis Kit (New England
Biolabs), using the primers H34AꢀF 5′ꢀCGCCGACGCCGCCACCCACCAGꢀ3′ and
H34AꢀR
5′ꢀGAGGTTGTACTTCATGGTCAꢀGCGCCTGGATCTCGTGꢀ3′.
The
nucleotide sequence of the cloned rohP was confirmed, and the plasmid was transformed
into E. coli BL21 (DE3) cells for protein production.
E. coli cell cultures were grown at 37 °C in LuriaꢀBertani (LB) medium
containing 50 ꢁg/mL kanamycin to an OD600 of 0.8–1.0, and then cooled to 16 °C.
Protein expression was induced with 0.1 mM βꢀDꢀ1ꢀthiogalactopyranoside (IPTG), and
the cultures were grown for an additional 16 h at 16 °C. Cells were harvested by
centrifugation and frozen at –20 °C until protein purification.
For purification, the cells were thawed and reꢀsuspended in 20 mM HEPES, 50
mM NaCl (pH 7.5) buffer and sonicated to lyse the cells. The lysate was then centrifuged
at 40,500 g for 45 min to remove insoluble material. The lysate was then applied to a
column containing ~1 mL Chelating Sepharose™ Fast Flow resin (GE Lifesciences)
charged with NiSO ·6H O. The lysate was gravity filtered through the resin, and the resin
4
2
was washed with 20 mL of 20 mM HEPES, 50 mM NaCl (pH 7.5) buffer. The column
was then washed with 5 mL portions of 20 mM HEPES, 50 mM NaCl (pH 7.5)
containing 5, 10, 20, 50, 100, 200, 300, and 500 mM imidazole in a stepwise manner to
elute the protein. The fractions containing RohP could be identified by the yellow color,
15
ACS Paragon Plus Environment

ACS Chemical Biology
Page 16 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
and most of the protein eluted in the fractions that contained 200 and 300 mM imidazole.
The RohP containing fractions were combined and concentrated to a volume of ~5 mL
using an Amicon Ultra Centrifugal filter (10,000 molecular weight cutꢀoff, EMDꢀ
TM
Millipore). The concentrated fraction was loaded into a HiLoad
Superdex 16/600
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Superdex column (GE Amersham Biosciences) preꢀequilibrated with 20 mM HEPES, 50
ꢀ1
mM NaCl (pH 7.5) buffer. The protein was eluted using a flow rate of 1 mL min . The
fractions containing purified RohP were pooled to a final concentration of ~50 µM and
dialyzed against HEPES buffer containing 250 µM PLP for 16 h. Excess PLP was
removed by further dialysis with 20 mM HEPES, 50 mM NaCl (pH 7.5). Finally, RohP
ꢀ1
was concentrated to ~20 mg mL by centrifugation with an ultra centrifugal filter (10,000
molecular weight cutꢀoff, EMDꢀMillipore). At this concentration, the purified RohP
remained stable at 4 °C and was stored in the dark.
In vitro biochemical assays and product analysis Initial in vitro reactions (100 µL)
contained 30 µM RohP and 1 mM Lꢀarginine, in 20 mM Tris, 50 mM NaCl (pH 7.5)
buffer and proceeded for 16 h at room temperature. HPLC analysis of the reaction was
carried out after preꢀcolumn derivatization with dansylꢀchloride (DNSꢀCl). A reaction
mixture of 50 µL was treated with 80 mM Li CO , 70 µL of CH CN, and 30 µL of 5 mM
2
3
3
DNSꢀCl dissolved in CH CN. The reaction was carried out at room temperature for 1 h
3
and then 40 µL of 2% ethylamine was added to the mixture to react with excess DNSꢀCl.
The mixture was centrifuged, and 20 µL of the supernatant was subjected to HPLC
analysis. HPLC analysis was carried out on a 1260 HPLC apparatus (Agilent), using a
Luna C18(2), 5 µm, 4.6 mm ID × 250 mm column (Phenomenex). Elution was performed
ꢀ1
at 0.5 mL min using a mobileꢀphase consisting of a linear gradient of water and
acetonitrile ((v/v): 95:5 to 50:50, 0 to 15 min; 0:100, 15 to 22 min), with both solvents
containing 0.05% (v/v) trifluoroacetic acid. DNSꢀArg was detected at a wavelength of
3
30 nm.
In vitro assays for ESIꢀMS analysis contained 10 µM RohP and 1 mM
Lꢀarginine,
in 20 mM HEPES, 50 mM NaCl (pH 7.5) buffer in a 100 µL reaction mixture. These
reactions were carried out for 4 h at room temperature and then quenched with an equal
volume of methanol. Precipitated protein was removed by centrifugation and 10 µL of
16
ACS Paragon Plus Environment

Page 17 of 33
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
the supernatant was subjected to ESIꢀMS analysis. MS analysis was performed with a
6120 Quadrupole LC/MS system (Agilent) operated in positive ion mode.
NMR analysis of RohP reaction products The reaction mixture (5 mL) contained 20 µM
RohP, and 10 mM Lꢀarginine in 20 mM sodium phosphate buffer (pH 7.2). The mixture
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
was incubated in an unsealed 50 mL vial at 25 °C and shaken at 120 rpm for 6 h. The
solvent was then evaporated overnight using a SpeedVac plus vacuum concentrator. The
dried solids were reꢀsuspended in 600 µL D O and centrifuged to remove any residual
2
undissolved solids prior to NMR analysis. All spectra were acquired utilizing a Bruker
Avance 600 MHz spectrophotometer.
Steady-state kinetics for the RohP reaction Kinetic assays were performed by
monitoring the consumption of O using a Clarkꢀtype polarographic O electrode
2
2
1
7
(
Hansatech, Pentney, UK) similar to the method described previously. The electrode
was calibrated daily using airꢀsaturated water and sodium hydrosulfite according to the
manufacturer’s instructions. The standard assay was performed in 1 mL of airꢀsaturated
4
0 mM MOPS (I = 0.1 M, pH 7.2) at 25 °C containing 500 ꢁM Lꢀarginine. The reaction
was initiated with the addition of RohP to 1 ꢁM. The observed rates were corrected with
background O consumption prior to reaction initiation. The effect of pH on the rate of
2
the RohPꢀcatalyzed reaction was evaluated using airꢀsaturated 20 mM buffers (I = 0.1 M)
of MES (pH 6.0), PIPES (pH 6.7), MOPS (pH 7.2), HEPES (pH 7.4), HEPPS (pH 8.0),
and TAPS (pH 8.5).
The steadyꢀstate kinetic parameters of RohP with respect to
determined at ambient oxygen levels by varying the concentration of ꢀarginine (8 – 500
µM) using 1 µM and 5 µM of RohP, respectively. The steadyꢀstate kinetic parameters of
RohP with respect to oxygen were measured using 17 – 685 µM O at an ꢀarginine
Lꢀarginine were
L
L
2
2
concentration of 500 mM. O concentrations were established by bubbling mixtures of O2
and N into the reaction chamber prior to reaction initiation. The final oxygen
2
concentrations were standardized to the level of airꢀsaturated buffer before the gas
bubbling. Kinetic parameters were determined either by leastꢀsquares fitting of the
MichaelisꢀMenten equation to the data using LEONORA or by Hill equation using Origin
8
.1 (OriginLab corp., Northampton, MA).
17
ACS Paragon Plus Environment

ACS Chemical Biology
Page 18 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Stoichiometry of the RohP reaction The production of H O was evaluated by comparing
2
2
the reaction rates catalyzed by 2.5 µM RohP and 0.3 mM ʟꢀarginine in the presence or
absence of ~4000 U of catalase. H O production was also measured colorimetrically by
2
2
ꢀ1
supplementing the reaction with 0.1 mg mL horseradish peroxidase (HRP) (Typeꢀ1) and
mM 2,2'ꢀazinoꢀbis(3ꢀethylbenzothiazolineꢀ6ꢀsulphonic acid) (ABTS). Upon the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
consumption of 100 µM oxygen, the reaction was quenched with 1 volume of 10% TCA.
The experimental values were compared to a calibration curve of H O .
2
2
For quantification of the
Lꢀarginine:O stoichiometry, reaction mixtures (1 mL)
2
containing 12 µM RohP and 250 µM
Lꢀarginine in 40 mM MOPS (I = 0.1 M, pH 7.2)
buffer were incubated at 25 °C. The mixture was equilibrated to ambient O levels prior
2
to the addition of RohP. Upon addition of RohP the O concentration was monitored by
2
Oxygraph and the reaction was quenched with one volume of MeOH after 100 µM of O2
was consumed. The
Lꢀarginine remaining after the reaction with RohP by was quantified
by HPLC after
preꢀcolumn derivatization with oꢀphthaldialdehyde
(OPA)/mercaptopropionic acid (MPA). The quenched solution was incubated with 1
ꢀ
volume of 0.4 M borate buffer (pH 10.2) and 1 volume of OPA/MPA solution (1 mg mL
1
OPA, 0.1% (v/v) MPA in 0.1 M borate buffer (pH 10.2)) for 3 min before HPLC
analysis. HPLC analysis was carried out on a 1260 HPLC apparatus (Agilent), using a
Poroshell 120, ECꢀC18, 2.7 µm, 4.6 mm ID × 50 mm column (Agilent). Elution was
ꢀ1
performed at 1.0 mL min using a mobileꢀphase consisting of a linearꢀgradient of water
and acetonitrile ((v/v): 98:2, 0 to 1 min; 80:20, 1 to 5 min; 50:50, 5 to 6.5 min; 0:100, 6.5
to 8 min; 98:2, 8 to 10 min), with both solvents containing 0.05% (v/v) trifluoroacetic
acid. Derivatized
Lꢀarginine was detected at a wavelength of 330 nm. The experimental
values were compared to a calibration curve of
Lꢀarginine that was done using the same
conditions. The data are reported as mean values ± s.d., with n = 5.
Spectroscopic analysis of the RohP reaction Aerobic reaction mixtures (1 mL) were
prepared in a quartz cuvette with a path length of 1 cm and contained 30 µM of enzyme
and 300 µM of substrate in 20 mM HEPES, 50 mM NaCl, pH 7.5. All solutions were
room temperature and air saturated with oxygen. The reaction was initiated upon addition
of substrate. Spectra were recorded using a Varian Cary 100 Bio UVꢀVis
spectrophotometer (Agilent).
18
ACS Paragon Plus Environment

Page 19 of 33
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
ꢀ1
Crystallization Initial crystallization conditions were identified by screening 4 mg mL
RohP against the Index HT Screen (Hampton Research) and Top96 Crystal Screen
Anatrace). Optimization of the initial crystallization conditions was carried out at room
temperature using hanging drop vapor diffusion. Diffractionꢀquality crystals were
(
ꢀ1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
obtained by mixing 1.5 ꢁL of 8 mg mL protein and an equal volume of a crystallization
solution composed of 0.2 M sodium malonate (pH 7.0), and 16–20% (w/v) PEG 3350,
using hanging drop vapor diffusion over a 500 µL reservoir of the crystallization
solution. Large, yellow crystals appeared after approximately 7 d, though some crystals
took up to 4 weeks to form under these conditions.
The structures of RohP in complex with intermediates and product were obtained
by adding solutions of PLP to the ~3 ꢁL drops containing crystals of RohP, soaking
overnight, and then adding ʟꢀarginine solutions for timed soaks. The structure of the
holoꢀRohP resulted from adding 1 ꢁL of 20 mM PLP to a drop containing crystals for an
overnight soak, giving a final concentration of 5 mM PLP. The structure of the first RohP
quinonoid intermediate (Q1) resulted from adding 0.5 ꢁL 20 mM PLP for an overnight
soak and then adding 1 ꢁL 100 mM
mM PLP and 22 mM ꢀarginine. The structure of the second RohP quinonoid
intermediate (Q2) resulted from adding 1.0 ꢁL 20 mM PLP for an overnight soak and
then adding 1 ꢁL 50 mM ꢀarginine for 5 min, giving final concentrations of 4 mM PLP
and 10 mM ꢀarginine. The RohPꢀ4ꢀhydroxyꢀ2ꢀketoarginine structure resulted from
adding 1.0 ꢁL 20 mM PLP and 1 ꢁL 50 mM ꢀarginine for an overnight soak, giving
final concentrations of 4 mM PLP and 10 mM ꢀarginine. After soaking as described, the
Lꢀarginine for 90 s, giving final concentrations of 2.2
L
L
L
L
L
crystals were cryoprotected using solution composed of 0.2 M sodium malonate (pH 7.0),
15% (w/v) PEG 3350, and 20% (v/v) ethylene glycol and flashꢀfrozen in liquid nitrogen
prior to Xꢀray data collection.
Data collection, structure determination and model refinement Xꢀray diffraction data
were collected at the Canadian Light Source (Saskatoon, Canada), beamline 08IDꢀ1, at a
wavelength of 0.97949 Å using MX300ꢀHE and Pilatus 6M detectors. UVꢀVisible
spectroscopic data of RohP crystals were collected using the microspectrophotometer at
beamline BL9ꢀ2 at the Stanford Synchrotron Radiation Light Source (Menlo Park, United
2
2
23
States). All data sets were integrated using iMOSFLM, and scaled using AIMLESS.
19
ACS Paragon Plus Environment

ACS Chemical Biology
Page 20 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
RohP crystallized with two units in the asymmetric unit, forming a homodimer in the
space group C . The crystal structures of each RohP complex was phased by PHASERꢀ
2
24
MR in the Phenix software package, using the
Lꢀarginine, γꢀhydroxylase SwMppP as a
model (PDB: 5DJ1 chain D, 32% identity across 393 amino acid residues). The initial
2
5
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
results from molecular replacement were input into Phenix Autobuild for additional
model building. The Autobuild output was subjected to several rounds of manual
26
27
inspection and building in COOT, and refinement in phenix.refine using translationꢀ
liberationꢀscrew (TLS) refinement. Alternate conformations of side chains and molecules
from the crystallization solution were added to the model where appropriate. Solvent
molecules were added automatically in phenix.refine and examined manually in COOT.
Refinement statistics are listed in Table 1. Nonꢀstandard ligand restraints were generated
2
8
using Phenix eLBOW. The second quinonoid intermediate (Q2) was initially fit to the
2
1
F ꢀF omit density using ARP/wARP version 7.6. The coordinates generated were used
o
c
to produce a restraint file containing the ARP/wARP optimized geometry, and the entire
structure was subjected to additional refinement in phenix.refine.
The holoꢀRohP structure has the Nꢀterminus through residue 26 disordered in both
chains. The first RohP quinonoid intermediate (Q1) structure has the Nꢀterminus through
residue 25 disordered in both chains. The second RohP quinonoid intermediate (Q2)
structure has the Nꢀterminus through residue 13 disordered in both chains. The structure
of the RohP with (S)ꢀ4ꢀhydroxyꢀ2ꢀketoarginine in the active site has the Nꢀterminus
through residue 13 disordered in Chain A and through residue 14 disordered in Chain B.
In all structures, C-terminal residues 391–393 are disordered in each monomer, except in
Chain B of both quinonoid structures, where residues 392–393 are disordered.
ASSOCIATED CONTENT
Supporting Information
Supporting Information is available free of charge on the ACS Publications website at
DOI:
High resolution ESIꢀMS data, Xꢀray collection data, Supporting Figures S1ꢀS15
Data Deposition
20
ACS Paragon Plus Environment

Page 21 of 33
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
The atomic coordinates and structure factors for the crystal structures reported have been
deposited in the Protein Data Bank. PDB ID Codes 6C3A (RohPꢀ4ꢀhydroxyꢀ2ꢀ
ketoarginine), 6C3B (RohPꢀholoenzyme), 6C3C (RohPꢀquinonoid I) and 6C3D (RohPꢀ
quinonoid II).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Author Information
Corresponding Author
*
ksryan@chem.ubc.ca
ORCID
Jason B. Hedges: 0000ꢀ0003ꢀ2922ꢀ0085
Eugene Kuatsjah: 0000ꢀ0002ꢀ8793ꢀ2331
YiꢀLing Du: 0000ꢀ0001ꢀ7072ꢀ4803
Lindsay D. Eltis: 0000ꢀ0002ꢀ6774ꢀ8158
Katherine S. Ryan: 0000ꢀ0003ꢀ1643ꢀ1940
Declaration of Interests
The authors declare no competing interests.
Acknowledgements We thank M. Higgins for assistance with the initial Xꢀray
crystallographic work. Our work was supported by funding from the Natural Sciences
and Engineering Research Council of Canada (RGPINꢀ2016ꢀ03778 to K.S.R., and
1
71359ꢀ13 to L.D.E.), the Alfred P. Sloan Foundation (FGꢀ20166503 to K.S.R.), and the
Canadian Institutes of Health Research (FDNꢀ148381 and 201312MSHꢀ322191ꢀ209186
to K.S.R.). L.D.E. holds a Tier 1 Canada Research Chair. Research described in this
paper was performed using beamline 08IDꢀ1 at the Canadian Light Source and beamline
9
ꢀ2 at the Stanford Synchrotron Radiation Light Source.
References
1) Labinger, J. A., and Bercaw, J. E. (2002) Understanding and exploiting CꢀH bond
(
activation. Nature 417, 507–514.
(2) Roudesly, F., Oble, J., and Poli, G. (2017) Metalꢀcatalyzed CꢀH
activation/functionalization: The fundamentals. J. Mol. Catal. Chem. 426, 275–296.
21
ACS Paragon Plus Environment

ACS Chemical Biology
Page 22 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
3) Helmetag, V., Samel, S. A., Thomas, M. G., Marahiel, M. A., and Essen, L.ꢀO. (2009)
Structural basis for the erythroꢀstereospecificity of the
viomycin biosynthesis. FEBS J. 276, 3669–3682.
4) Ge, W., Wolf, A., Feng, T., Ho, C., Sekirnik, R., Zayer, A., Granatino, N., Cockman,
Lꢀarginine oxygenase VioC in
(
M. E., Loenarz, C., Loik, N. D., Hardy, A. P., Claridge, T. D. W., Hamed, R. B.,
Chowdhury, R., Gong, L., Robinson, C. V., Trudgian, D. C., Jiang, M., Mackeen, M. M.,
McCullagh, J. S., Gordiyenko, Y., Thalhammer, A., Yamamoto, A., Yang, M., LiuꢀYi,
P., Zhang, Z., SchmidtꢀZachmann, M., Kessler, B. M., Ratcliffe, P. J., Preston, G. M.,
Coleman, M. L., and Schofield, C. J. (2012) Oxygenaseꢀcatalyzed ribosome
hydroxylation occurs in prokaryotes and humans. Nat. Chem. Biol. 8, 960–962.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(5) Chang, C.ꢀY., Lyu, S.ꢀY., Liu, Y.ꢀC., Hsu, N.ꢀS., Wu, C.ꢀC., Tang, C.ꢀF., Lin, K.ꢀH.,
Ho, J.ꢀY., Wu, C.ꢀJ., Tsai, M.ꢀD., and Li, T.ꢀL. (2014) Biosynthesis of streptolidine
involved two unexpected intermediates produced by a dihydroxylase and a cyclase
through unusual mechanisms. Angew. Chem. Int. Ed. Engl. 53, 1943–1948.
(
6) Phillips, R. S. (2015) Chemistry and diversity of pyridoxalꢀ5′ꢀphosphate dependent
enzymes. Biochim. Biophys. Acta BBA - Proteins Proteomics 1854, 1167–1174.
7) Toney, M. D. (2011) Controlling reaction specificity in pyridoxal phosphate enzymes.
(
Biochim. Biophys. Acta BBA - Proteins Proteomics 1814, 1407–1418.
(8) Bunik, V. I., Schloss, J. V., Pinto, J. T., Dudareva, N., and Cooper, A. J. L. (2011) A
survey of oxidative paracatalytic reactions catalyzed by enzymes that generate
carbanionic intermediates: implications for ROS production, cancer etiology, and
neurodegenerative diseases. Adv. Enzymol. Relat. Areas Mol. Biol. 77, 307–360.
(9) Bertoldi, M., Cellini, B., Montioli, R., and Borri Voltattorni, C. (2008) Insights into
the mechanism of oxidative deamination catalyzed by DOPA decarboxylase.
Biochemistry 47, 7187–7195.
(10) Sakai, K., Miyasako, Y., Nagatomo, H., Watanabe, H., Wakayama, M., and
Moriguchi, M. (1997)
ornithine oxidase activity. J. Biochem. 122, 961–968.
11) Kaminaga, Y., Schnepp, J., Peel, G., Kish, C. M., BenꢀNissan, G., Weiss, D.,
LꢀOrnithine decarboxylase from Hafnia alvei has a novel Lꢀ
(
Orlova, I., Lavie, O., Rhodes, D., Wood, K., Porterfield, D. M., Cooper, A. J. L., Schloss,
J. V., Pichersky, E., Vainstein, A., and Dudareva, N. (2006) Plant phenylacetaldehyde
synthase is a bifunctional homotetrameric enzyme that catalyzes phenylalanine
decarboxylation and oxidation. J. Biol. Chem. 281, 23357–23366.
(12) Wang, M., Zhao, Q., Zhang, Q., and Liu, W. (2016) Differences in PLPꢀdependent
cysteinyl processing lead to diverse Sꢀfunctionalization of lincosamide antibiotics. J. Am.
Chem. Soc. 138, 6348–6351.
(13) Huang, Y., Liu, X., Cui, Z., Wiegmann, D., Niro, G., Ducho, C., Song, Y., Yang, Z.,
and Van Lanen, S. G. (2018) Pyridoxalꢀ5′ꢀphosphate as an oxygenase cofactor: discovery
of a carboxamideꢀforming, αꢀamino acid monooxygenaseꢀdecarboxylase. Proc. Natl.
Acad. Sci. U. S. A. 115, 974–979.
(14) Nakane, A., Nakamura, T., and Eguchi, Y. (1977) A novel metabolic fate of arginine
in Streptomyces eurocidicus. Partial resolution of the pathway and identification of an
intermediate. J. Biol. Chem. 252, 5267–5273.
(15) Han, L., Schwabacher, A. W., Moran, G. R., and Silvaggi, N. R. (2015)
Streptomyces wadayamensis MppP is a pyridoxal 5′ꢀphosphateꢀdependent Lꢀarginine αꢀ
22
ACS Paragon Plus Environment

Page 23 of 33
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
deaminase, γꢀhydroxylase in the enduracididine biosynthetic pathway. Biochemistry 54,
7
029–7040.
(16) Du, Y.ꢀL., Alkhalaf, L. M., and Ryan, K. S. (2015) In vitro reconstitution of
indolmycin biosynthesis reveals the molecular basis of oxazolinone assembly. Proc. Natl.
Acad. Sci. U. S. A. 112, 2717–2722.
(17) Du, Y.ꢀL., Singh, R., Alkhalaf, L. M., Kuatsjah, E., He, H.ꢀY., Eltis, L. D., and
Ryan, K. S. (2016) A pyridoxal phosphateꢀdependent enzyme that oxidizes an
unactivated carbonꢀcarbon bond. Nat. Chem. Biol. 12, 194–199.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
18) Yin, X., and Zabriskie, T. M. (2006) The enduracidin biosynthetic gene cluster from
Streptomyces fungicidicus. Microbiol. 152, 2969–2983.
19) Li, C., and Lu, C.ꢀD. (2009) Arginine racemization by coupled catabolic and
anabolic dehydrogenases. Proc. Natl. Acad. Sci. U. S. A. 106, 906–911.
20) Vlessis, A. A., Bartos, D., and Trunkey, D. (1990) Importance of spontaneous alphaꢀ
(
(
ketoacid decarboxylation in experiments involving peroxide. Biochem. Biophys. Res.
Commun. 170, 1281–1287.
(21) Langer, G., Cohen, S. X., Lamzin, V. S., and Perrakis, A. (2008) Automated
macromolecular model building for Xꢀray crystallography using ARP/wARP version 7.
Nat. Protoc. 3, 1171–1179.
(22) Battye, T. G. G., Kontogiannis, L., Johnson, O., Powell, H. R., and Leslie, A. G. W.
(
2011) iMOSFLMꢂ: a new graphical interface for diffractionꢀimage processing with
MOSFLM. Acta Crystallogr. D Biol. Crystallogr. 67, 271–281.
23) Evans, P. R., and Murshudov, G. N. (2013) How good are my data and what is the
resolution? Acta Crystallogr. D Biol. Crystallogr. 69, 1204–1214.
24) McCoy, A. J., GrosseꢀKunstleve, R. W., Adams, P. D., Winn, M. D., Storoni, L. C.,
(
(
and Read, R. J. (2007) Phaser crystallographic software. J. Appl. Crystallogr. 40, 658–
674.
(25) Terwilliger, T. C., GrosseꢀKunstleve, R. W., Afonine, P. V., Moriarty, N. W., Zwart,
P. H., Hung, L.ꢀW., Read, R. J., and Adams, P. D. (2008) Iterative model building,
structure refinement and density modification with the PHENIX AutoBuild wizard. Acta
Crystallogr. D Biol. Crystallogr. 64, 61–69.
(
26) Emsley, P., and Cowtan, K. (2004) Cootꢂ: modelꢀbuilding tools for molecular
graphics. Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132.
27) Afonine, P. V., Mustyakimov, M., GrosseꢀKunstleve, R. W., Moriarty, N. W.,
(
Langan, P., and Adams, P. D. (2010) Joint Xꢀray and neutron refinement with
phenix.refine. Acta Crystallogr. D Biol. Crystallogr. 66, 1153–1163.
(28) Moriarty, N. W., GrosseꢀKunstleve, R. W., and Adams, P. D. (2009) electronic
Ligand Builder and Optimization Workbench (eLBOW): a tool for ligand coordinate and
restraint generation. Acta Crystallogr. D Biol. Crystallogr. 65, 1074–1080.
2
3
ACS Paragon Plus Environment

ACS Chemical Biology
Page 24 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure Legends:
+
Figure 1. Ind4ꢀ and RohPꢀcatalyzed reactions. [M+H] ions observed from both RohP
and Ind4ꢀcatalyzed reactions are shown in purple, from only the RohPꢀcatalyzed
reactions are in red, and from only the Ind4ꢀcatalyzed reaction are in blue. Molecules in
gray arise from nonꢀenzymatic reactions with H O .
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
Figure 2. RohPꢀcatalyzed oxidation of
L
ꢀarginine. Liquid chromatographyꢀmass
spectrometry analysis of the products of the RohP reaction with
L
ꢀarginine ( ꢀArg). The
L
reaction conditions are indicated above each spectrum, and representative integrated total
ion chromatograms for each reaction are shown.
Figure 3. Modelling of RohPꢀquinonoid intermediates. Left Panel: (a) The first
quinonoid intermediate (Q1) modelled into the density present in the active site of the
first RohP intermediate structure. (b) The second quinonoid intermediate (Q2) with a
double bond between arginine Cβ and Cγ modelled into the density present in the active
site of the second intermediate structure. Both F ꢀF omit maps are displayed in green and
o
c
contoured at 3.0 σ. Middle panel: Depicts the entire protein structure as modelled in each
structure. The two monomers are indicated with light and dark coloring. The Nꢀterminal
helix, absent in holoꢀRohP, is colored magenta here. Right panel: View of active sites,
with nearby conserved residues surrounding the quinonoid intermediates depicted as
sticks.
Figure 4. Distances between modelled quinonoid intermediates and conserved residues
Glu20 and His34 in Chain A of the RohP homodimer. (a) The RohP holoenzyme has only
a water molecule (red sphere) positioned 2.4 Å from His34. (b) The Nꢀterminus and
Glu20 are absent in the RohPꢀquinonoid I. His34 is positioned 3.9 Å from the Cγ atom of
the bound substrate and 2.5 Å from a water molecule. (c) In the RohPꢀquinonoid II, the
Nꢀterminus is ordered, including Glu20, which is now positioned 4.0 Å from the Cβ of
the bound substrate. Furthermore, the new positioning of the amino acid substrate places
its Cβ and Cδ atoms 3.2 Å and 3.5 Å, respectively, from His34. No ordered water was
24
ACS Paragon Plus Environment

Page 25 of 33
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
present near His34. (d) The product (S)ꢀ4ꢀhydroxyꢀ2ꢀketoarginine has its Cβ 3.3 Å from
Glu20 and its Cδ 3.6 Å from His34. The 4' hydroxyl group is positioned 2.7 Å from
His34.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. Possible stereoisomers of 4ꢀhydroxyꢀ2ꢀketoarginine. (a) The Rꢀenantiomer
modelled into F ꢀF omit density present in the active site. (b) The density present around
o
c
the Rꢀenantiomer after refinement using Refmac. (c) The Sꢀenantiomer modelled into Foꢀ
Fc omit density present in the active site. (d) The density present around the Sꢀenantiomer
after refinement using Refmac. The F ꢀF maps are contoured at 3.0 σ with positive and
o
c
negative density indicated as green and red, respectively. The 2F ꢀF maps are contoured
o
c
at 1.0 σ in gray. Ligands were built using eLBOW in the Phenix software suite.
Figure 6. Proposed mechanism of the RohP catalyzed reaction to give 4.
25
ACS Paragon Plus Environment

ACS Chemical Biology
Page 26 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 1. RohP XꢀRay Refinement Statistics
HoloꢀRohP
6C3B]
Quinonoid I
Quinonoid II
Int. Aldimine –
[
(Q1) [6C3C]
(Q2) [6C3D]
Product [6C3A]
Refinement
a
Rwork
0.1522 (0.2332)
0.1624 (0.2409) 0.1599 (0.2326) 0.1528 (0.2251)
0.1866 (0.2601) 0.1856 (0.2454) 0.1729 (0.2505)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Rfree
0.1725 (0.2476)
No. nonꢀhydrogen atoms
Protein
6859
5951
828
7079
5942
1025
112
7079
6123
847
7121
6140
904
Solvent
Ligands
80
109
77
RMSD Bonds (Å)
RMSD Angles (°)
0.006
0.87
0.006
0.83
98.62
1.38
0
0.006
0.82
98.27
1.73
0
0.006
0.92
98.66
1.34
0
Ramachandran favored (%) 98.75
Ramachandran allowed (%) 1.25
Ramachandran outliers (%)
0
2
Average B factor (Å )
21.82
20.05
34.22
33.36
20
18.33
16.39
29.33
20.39
ꢀ
24.84
23.11
36.41
31.83
17
23.51
21.60
35.48
35.48
17
Protein
Solvent
Ligands
No. TLS groups
a
Data from the highestꢀresolution shell is indicated in parentheses.
2
6
ACS Paragon Plus Environment

Page 27 of 33
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
+
Figure 1. Ind4- and RohP-catalyzed reactions. [M+H] ions observed from both RohP and Ind4-catalyzed
reactions are shown in purple, from only the RohP-catalyzed reactions are in red, and from only the Ind4-
catalyzed reaction are in blue. Molecules in gray arise from non-enzymatic reactions with H
2 2
O .
82x38mm (600 x 600 DPI)
ACS Paragon Plus Environment

ACS Chemical Biology
Page 28 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. RohP-catalyzed oxidation of L-arginine. Liquid chromatography-mass spectrometry analysis of the
products of the RohP reaction with L-arginine (L-Arg). The reaction conditions are indicated above each
spectrum, and representative integrated total ion chromatograms for each reaction are shown.
1
40x171mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 29 of 33
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. Modelling of RohPꢀquinonoid intermediates. Left Panel: (a) The first quinonoid intermediate (Q1)
modelled into the density present in the active site of the first RohP intermediate structure. (b) The second
quinonoid intermediate (Q2) with a double bond between arginine Cβ and Cγ modelled into the density
o c
present in the active site of the second intermediate structure. Both F ꢀF omit maps are displayed in green
and contoured at 3.0 σ. Middle panel: Depicts the entire protein structure as modelled in each structure. The
two monomers are indicated with light and dark coloring. The Nꢀterminal helix, absent in holoꢀRohP, is
colored magenta here. Right panel: View of active sites, with nearby conserved residues surrounding the
quinonoid intermediates depicted as sticks.
140x77mm (300 x 300 DPI)
ACS Paragon Plus Environment