Structural insights into the function of the nicotinate mononucleotide:phenol/p-cresol phosphoribosyltransferase (ArsAB) enzyme from Sporomusa ovata

Article abstract of DOI:10.1021/bi301142h

Cobamides (Cbas) are cobalt (Co) containing tetrapyrrole-derivatives involved in enzyme-catalyzed carbon skeleton rearrangements, methyl-group transfers, and reductive dehalogenation. The biosynthesis of cobamides is complex and is only performed by some

Full text of DOI:10.1021/bi301142h

Article
pubs.acs.org/biochemistry
Structural Insights into the Function of the Nicotinate
Mononucleotide:phenol/p‑cresol Phosphoribosyltransferase (ArsAB)
Enzyme from Sporomusa ovata
Sean A. Newmister,†,§ Chi Ho Chan,‡,§ Jorge C. Escalante-Semerena,*,‡,# and Ivan Rayment*,†
‡
†Departments of Biochemistry and Bacteriology, University of Wisconsin, Madison, Wisconsin 53706, United States
#Department of Microbiology, University of Georgia, Athens, Georgia 30602, United States
S
* Supporting Information
ABSTRACT: Cobamides (Cbas) are cobalt (Co) containing
tetrapyrrole-derivatives involved in enzyme-catalyzed carbon
skeleton rearrangements, methyl-group transfers, and reductive
dehalogenation. The biosynthesis of cobamides is complex and
is only performed by some bacteria and achaea. Cobamides
have an upper (Coβ) ligand (5′-deoxyadenosyl or methyl) and
a lower (Coα) ligand base that contribute to the axial Co
coordinations. The identity of the lower Coα ligand varies
depending on the organism synthesizing the Cbas. The
homoacetogenic bacterium Sporomusa ovata synthesizes two
unique phenolic cobamides (i.e., Coα-(phenolyl/p-cresolyl)-
cobamide), which are used in the catabolism of methanol and 3,4-dimethoxybenzoate by this bacterium. The S. ovata ArsAB
enzyme activates a phenolic lower ligand prior to its incorporation into the cobamide. ArsAB consists of two subunits, both of
which are homologous (∼35% identity) to the well-characterized Salmonella enterica CobT enzyme, which transfers nitrogenous
bases such as 5,6-dimethylbenzimidazole (DMB) and adenine, but cannot utilize phenolics. Here we report the three-
dimensional structure of ArsAB, which shows that the enzyme forms a pseudosymmetric heterodimer, provide evidence that only
the ArsA subunit has base:phosphoribosyl-transferase activity, and propose a mechanism by which phenolic transfer is facilitated
by an activated water molecule.
obamides are cobalt-containing modified cyclic tetrapyr-
role-derivatives, which are members of a broad family that
the lower ligand (Figure 1C). Phenolic cobamides were first
described in the homoacetogenic bacterium Sporomusa ovata.8,9
Two features of the S. ovata cobamides are of note. First, the
phenolic compound is covalently attached to the ribosyl group
via an O-glycosidic bond, rather than the N-glycosidic bond
found in all other known cobamides; second, unlike any other
cobamides, phenolic cobamides cannot exist in the ‘base-on’
conformation because neither phenol nor p-cresol contains an
atom that can establish a coordination bond with the Co ion of
the corrin ring. Enzymes that require cobamides in the base-on
conformation for catalysis (e.g., glycerol dehydratase, diol
dehydratase, ethanolamine ammonia-lyase) cannot use or are
inhibited by phenolic cobamides.10,11
Recently, the S. ovata arsAB genes encoding the two subunits
of the enzyme responsible for the conversion of phenol/p-
cresol to the corresponding α-O-glycosidic riboside mono-
phosphate were identified. The arsAB-encoded enzyme was
isolated to homogeneity, and initial analyses of its activity were
performed in vivo and in vitro.12 The enzyme functions as a
heterodimer; i.e., neither subunit is active by itself. The two
C
include all forms of heme (iron), chlorophylls (magnesium),
and coenzyme F430 (nickel).1 Several features distinguish
cobamides from other members of this family of molecules.
As depicted in Figure 1, a cobamide has an upper (Coβ) ligand
covalently bound to the cobalt ion of the ring, and a lower
(Coα) ligand interacting with the Co ion via a coordination
bond.
Vitamin B12 is a cobamide (Cba) that contains a cyano (CN)
group as the Coβ ligand (Figure 1). When DMB is the Coα
ligand, it is known as cyanocobalamin. Cobamides are in their
coenzymic form when the Coβ ligand is 5′-deoxyadenosine
(AdoCba). AdoCba participates in radical mediated molecular
rearrangement such as in diol dehydratase and methylmalonyl-
CoA mutase.2,3 Cobamides also serve as transient methyl-group
carriers in Co(I)Cba-dependent methyltransferases.4 In nature,
the nucleoside base of cobamides varies, depending on the
microorganism synthesizing it.5,6 Purines and purine analogues
linked to the ribosyl group via an N-glycosidic bond can form
coordination bonds with the cobalt ion, while bases linked to
the ribosyl group via an O-glycosidic bond cannot (Figure
1B,C).
Received: August 23, 2012
Revised: October 5, 2012
Published: October 5, 2012
Only some bacteria and archaea synthesize cobamides,7 and
unique among cobamides are those with phenol or p-cresol as
© 2012 American Chemical Society
8571
dx.doi.org/10.1021/bi301142h | Biochemistry 2012, 51, 8571−8582

Biochemistry
Article
Figure 1. Chemical structures of cobamides, vitamin B12, and phenolic cobamides. The upper Coβ ligand (R) varies. When R = a cyano (CN) group,
the cobamide is in its vitamin form; when R = 5′deoxyadenosine, the cobamide is in its coenzymic form. The chemical nature of the lower (Coα)
ligand base (B*) is diverse.5
subunits are evolutionarily related and share 37% and 57%
amino acid identity and similarity, respectively. Furthermore, in
addition to phenol and p-cresol, the enzyme converts DMB to
α-DMB riboside monophosphate (α-ribazole-5′-phosphate or
α-RP). The latter feature of ArsAB is noteworthy because, to
date, none of the enzymes known to transfer DMB or any other
nitrogenous base can utilize phenol or p-cresol.13 This feature
of ArsAB is even more intriguing in light of the fact that both
ArsA and ArsB proteins are homologues of the well-
characterized nicotinate mononucleotide (NaMN):DMB phos-
phoribosyltransferase (CobT) enzyme of Salmonella enterica
(SeCobT).12 ArsA and ArsB share 37% and 36% amino acid
sequence identity with SeCobT, but prior to this study it was
unknown whether the phosphoribosyltransferase activity of
ArsAB was provided by one or both subunits of the inferred
heterodimer. The general scheme for the activation of the base
is shown below in eq 1.
subtle but profound evolutionary changes required for ArsAB
to phosphoribosylate phenolic bases. Remarkably, only the
active site in ArsA was occupied by substrate. Analysis of the
ArsB active site identified an arginyl side chain that may block
access to the site. The apparent inactivity of the ArsB subunit
suggests a strictly structural role for ArsB in the phosphoribosyl
transferase activity of this enzyme, though this does not
preclude some other unidentified activity for that active site.
EXPERIMENTAL PROCEDURES
■
Construction of Plasmids Encoding ArsAB and
Variants. The arsA and arsB coding sequences were amplified
from plasmid pARSAB712 using primers listed in Table 1. The
fragment was cut with SmaI and NheI enzymes and was ligated
into the StuI and NheI sites of plasmid pH6T, a pTEV5
derivative,14 yielding plasmid pARSAB22. Plasmid pARSAB22
directed the synthesis of H6−ArsA and tag-less ArsB. Variants
of ArsAB were constructed using the QuikChange method
(Stratagene) using pARSAB22 as template. The plasmids
constructed and primers used are listed in Table 1.
base + nicotinate mononucleotide (NaMN) + enzyme
(1)
→ α‐riboside‐5′‐P + nicotinate
ArsAB Enzyme Isolation. ArsAB was overproduced using
plasmid pARSAB22 in strain JE13607, a derivative of E. coli
BL21(λ DE3) that lacks the cobT gene encoding the
NaMN:5,6-dimethylbenzimidazole (DMB) phosphoribosyl-
transferase (CobT) enzyme of this bacterium. Strain JE13607
was used to ensure that native E. coli CobT protein did not
contaminate the ArsAB heterodimer during subsequent kinetic
analyses or interfere with protein folding.12 Strain JE13607
harboring pARSAB22 was grown in six 2-L super broth (SB)
medium cultures (tryptone 35 g, yeast extract 20 g, NaCl 5 g,
NaOH 2N, 2.5 mL per liter of deionized H2O). Expression of
Crystallographic analyses of SeCobT (a homodimer) in
complex with diverse substrates13 provided structural explan-
ations for how the enzyme phosphoribosylates most of the
nitrogenous bases found in cobamides, with the exception of
phenol and p-cresol.
Here we report the three-dimensional crystal structure of S.
ovata ArsAB in its substrate-free form (2.1 Å), and in complex
with DMB (1.5 Å), p-cresol (2.1 Å), phloroglucinol (1.6 Å),
and phenol (2.4 Å). Comparisons of the ArsAB complexes with
the corresponding SeCobT complexes shed light onto the
8572
dx.doi.org/10.1021/bi301142h | Biochemistry 2012, 51, 8571−8582

Biochemistry
Article
Preparation of ArsAB Proteins for Crystallography.
For crystallographic studies, the concentration of NaCl and
HEPES in ArsAB solutions was gradually reduced by dialysis (5
steps), from 50 mM HEPES pH 8.0 and 300 mM NaCl to 10
mM HEPES pH 7.5; the resulting solution was concentrated to
10 mg/mL before drop-freezing in liquid nitrogen. The
concentration of ArsAB was determined using the combined
calculated extinction coefficients of ArsA and ArsB at 280 nm
(21150 cm−1 M−1) utilizing a NanoDrop 1000 spectropho-
tometer (Thermo).
Table 1. List of Plasmids and Primers Used in This Study
plasmid
pH6T
genotype and description
overexpression vector that fuses the N-terminus of the protein
of interest to a H6 tag, which can be removed by rTEV
protease, bla +
pARSAB22 So arsAB+ in pH6T, bla+
pARSAB31 So arsAB+ (ArsA M87Q) in pH6T, bla+
pARSAB32 So arsAB+ (ArsA I321S) in pH6T, bla+
pARSAB33 So arsAB+ (ArsA M87Q; I321S) in pH6T, bla+
primer name
ArsA F
primer sequence
Crystallization of ArsAB in the Substrate-Free State.
ArsAB heterodimers were screened for initial crystallization
conditions by vapor diffusion at 25 and 4 °C with a 144-
condition sparse matrix screen developed in the Rayment
laboratory. Single, diffraction-quality crystals were grown by
hanging drop vapor diffusion by mixing 2 μL of 10 mg/mL
ArsAB heterodimers in 10 mM HEPES pH 7.6 with 2 μL of
reservoir solution containing 100 mM 3-(N-morpholino)-
propanesulfonic acid (MOPS pH 7.1), 12.5% methyl ether
polyethylene glycol 5000 (MEPEG5K), 20 mM NaCl, 12%
ethylene glycol, and 10 mM phloroglucinol. Hanging droplets
were immediately nucleated from an earlier spontaneous
crystallization event with a cat’s whisker. Crystals grew to
approximate dimensions of 200 × 200 × 400 μm within 7 days.
Phloroglucinol was identified from an additive screen as a
compound that dramatically improved the size, stability, and
diffraction properties of the crystals. Subsequent structural
studies showed that phloroglucinol binds in the active site of
ArsA, which in hindsight is not surprising since it shares
structural similarity to phenolic substrates. Additionally,
phloroglucinol was found at low-occupancy in a different
orientation in the corresponding ArsB site (Figure S1,
Supporting Information). In order to prepare crystals of the
substrate-free protein, the crystals were soaked in a solution
containing 90 mM MOPS pH 7.1, 11.25% MEPEG5K, 25 mM
NaCl, 12% ethylene glycol, and 25 mM imidazole for 1 week.
The crystals of ArsAB were unstable in a synthetic mother
liquor that lacked an aromatic base, however imidazole was not
observed in the crystal lattice. The soaked crystals were
transferred in two steps into a cryoprotectant solution which
contained 100 mM MOPS pH 7.1, 12.5% MEPEG5K, 25 mM
NaCl, 17% ethylene glycol, and 25 mM imidazole and rapidly
plunged into liquid nitrogen. Substrate-free ArsAB crystallized
in the space group P212121 with unit cell dimensions of a = 51.3
Å, b = 78.0 Å, and c = 151.8 Å where there was a single ArsAB
heterodimer in the asymmetric unit.
Crystallization of ArsAB Complexed With DMB.
Crystals were grown by mixing 2 μL of ArsAB at 10 mg/mL
containing 5 mM DMB, and 10 mM HEPES pH 7.6 with 2 μL
of reservoir solution containing 100 mM MOPS pH 7.1, 12.5%
MEPEG5K, 20 mM NaCl, 12% ethylene glycol and 5 mM
diethylenetriamine. Hanging droplets were nucleated immedi-
ately by streak-seeding where after the crystals grew to
approximate dimensions of 200 × 200 × 400 μm within 7
days. The crystals were transferred stepwise into a cryoprotect-
ing solution containing 100 mM MOPS 7.1, 12.5% MEPEG5K,
20 mM NaCl, 15% ethylene glycol, and 5 mM diethylenetri-
amine and rapidly plunged into liquid nitrogen. ArsAB
complexed with DMB crystallized in the space group P21
with unit cell dimensions of a = 52.7 Å, b = 77.4 Å, c =
152.3 Å, α = 90.0°, β = 90.2°, and γ = 90.0° where there were
two heterodimers of ArsAB in the asymmetric unit.
AGC TCG CCC GGG G ATG AGT TTA CTG CAA
GCA ACA GTA GCG
ArsB R
GCA GCT AGC GCT TGC TAA TCT CTA ACA TCC
TTG C
ArsA M87Q F CCG ATA GAA ACA ACA ATT CAT CAG ACA GCT
AAT TAT CTT ATC TC
ArsA M87Q R GAG ATA AGA TAA TTA GCT GTC TGA TGA ATT
GTT GTT TCT ATC GG
ArsA I321S F
CGG CTG GGA GAG GGT AGC GGG GCT TCT ATG
GTT G
ArsA I321S R
CAA CCA TAG AAG CCC CGC TAC CCT CTC CCA
GCC G
arsAB was induced by the addition of isopropyl-β-D-1-
thiogalactopyranoside (IPTG, 0.5 mM) when cultures reached
log-phase (∼0.5 OD600). After induction, cultures were shifted
to 20 °C for 16 h. Cells were harvested by centrifugation
(Beckman/Coulter Avanti J20-XPI refrigerated centrifuge,
equipped with a JLA-8.1000 rotor; 15 min at 4 °C; 6000g),
and frozen at −80 °C until used. Frozen pellets were
resuspended in buffer A [50 mM 4-(2-hydroxyethyl)-1-
piperazine-N′-(2-ethanesulfonic acid) (HEPES), pH 8), con-
taining 300 mM NaCl, and 10 mM imidazole]; 1 mg/mL
lysozyme and DNase were added to the cell suspension prior to
disruption by sonication (60 s, 28% duty, 2 s pulses, setting 9)
using a 500 Sonic Dismembrator (Fisher Scientific). Cell debris
was removed by centrifugation (Beckman/Coulter Avanti J25-I
refrigerated centrifuge, equipped with a JA-25.50 rotor; 45 min
̈
at 4 °C; 43000g). Clarified extract was applied onto an AKTA
FPLC Purifier system (Amersham Biosciences) equipped with
a 5-mL Ni-charged HisTrap fast flow (FF) column (GE
Healthcare) equilibrated at a flow rate of 2 mL/min. After
loading, the column was washed with 50 mL of buffer A before
application of a 50-mL linear gradient of imidazole from 10 to
300 mM. Fractions containing ArsAB were pooled and His7-
tagged recombinant tobacco etch virus (His7-rTEV) protease
15 was added at 1:100 His7-rTEV:ArsAB ratio to cleave the His6
tag fused to ArsA.16 Tag-less ArsAB was reapplied onto the Ni-
charged column to separate His7-rTEV and other contami-
nants; tag-less ArsAB was found in the flow through. The
ArsAB enzyme was dialyzed into 50 mM HEPES, pH 8.0, 100
mM NaCl, and 15% v/v glycerol then flash frozen in liquid
nitrogen and stored at −80 °C until used.
Preparation of Variant Proteins. ArsAB variants encoded
in pARSAB31, pARSAB32 and pARSAB33 were overexpressed
in strain JE13607 in lysogenic broth (LB) as described above in
1-L scale. Clarified extracts in buffer A were applied onto a 1-
mL bed volume of Ni-NTA Superflow resin (Qiagen)
equilibrated with buffer A and washed with 10 column volumes
of buffer A containing 20 mM imidazole. The variant enzymes
were eluted with 10 column volumes of buffer A containing 300
mM imidazole. The His6 tag fused to ArsA was cleaved as
described above. Variant proteins were stored at −80 °C in the
same buffer as wild type ArsAB until used.
8573
dx.doi.org/10.1021/bi301142h | Biochemistry 2012, 51, 8571−8582

Biochemistry
Article
Table 2. X-ray Data Collection and Refinement Statistics
complex
pdb ID
phloroglucinol
phenol
p-cresol
P212121
substrate-free
P212121
DMB
space group
P212121
P212121
P21
wavelength (Å)
resolution range
reflections: measured
reflections: unique
redundancy
0.979
1.54
0.979 Å
50−1.95 (1.98−1.95)
1052638
46511
0.979 Å
50−2.24 (2.28−2.24)
649606
30319
0.979 Å
a
50−0.1.50 (1.55−1.50)
1708377
100980
9.9 (3.4)
99.7 (95.3)
47.0 (3.3)
5.2 (40.0)
18.9
50−2.4 (2.45−2.4)
223958
24812
50−1.45 (1.48−1.45)
1678842
216492
8.94 (5.7)
99.5 (96.8)
26.8 (8.7)
19.2 (52.8)
20.5
11.2 (5.7)
99.4 (87.8)
49.8 (5.6)
5.4 (38.2)
19.4
11.7 (5.4)
99.4 (88.8)
47.0 (3.8)
5.9 (54.8)
20.5
4.8 (3.3)
97.6 (95.1)
36.9 (6.3)
4.4 (22.9)
18.6
completeness (%)
average I/σ
b
Rmerge (%)
c
Rwork
c
Rfree
22.2
24.1
25.6
28.3
21.9
protein atoms
ligand atoms
4839
9571
4794
4818
9892
23
14
12
0
78
water molecules
average B factors (Å2)
Ramachandran (%)
most favored
546
204
186
105
1320
23.373
22.343
42.122
57.298
23.151
96.63
2.24
1.12
96.76
1.85
1.39
96.89
1.71
1.40
96.46
2.15
1.38
97.02
1.85
1.13
allowed
disallowed
rms deviations
bond lengths (Å)
bond angles (deg)
0.025
2.538
0.013
0.019
2.042
0.014
1.718
0.024
2.45
1.808
a
b
Values in parentheses are for the highest-resolution shell. Rmerge = (Σ|I(hkl) − I| × 100.)/ Σ |I(hkl)| where the average intensity I is taken over all
c
symmetry equivalent measurements and I(hkl) is the measured intensity for a given observation. Rfactor= (Σ|F(obs) − F(calc)| × 100.)/ Σ |F(obs)|, where
Rwork refers to the Rfactor for the data utilized in the refinement and Rfree refers to the Rfactor for 5% of the data that were excluded from the refinement.
Crystallization of ArsAB Complexed With p-Cresol.
Crystals of ArsAB in complex with p-cresol were prepared in an
similar fashion to those required to obtain ArsAB in the
substrate-free state with the exception that 20 mM p-cresol
replaced the 25 mM imidazole in the soaking and
cryoprotecting solutions. Crystals of ArsAB complexed with
p-cresol belong to the space group P212121 with unit cell
dimensions of a = 52.5 Å, b = 77.9 Å, and c = 152.2 Å and
contained a single ArsAB heterodimer in the asymmetric unit.
Crystallization of ArsAB Complexed With Phenol.
Crystals of ArsAB in complex with phenol were grown by
combining 2 μL of 10 mg/mL ArsAB heterodimers in 10 mM
HEPES pH 7.6 with 2 μL of reservoir solution containing 100
mM MOPS pH 7.1, 13.0% MEPEG5K, 20 mM NaCl, 12%
ethylene glycol, and 25 mM phenol. Crystals were transferred
stepwise into a cryoprotecting solution composed of 100 mM
MOPS pH 7.1, 13.2% MEPEG5K, 17% ethylene glycol, and 25
mM phenol and frozen by rapidly plunging into liquid nitrogen.
The crystals of ArsAB complexed with phenol belong to the
space group P212121 with unit cell dimensions of a = 52.7 Å, b
= 77.4 Å, c = 152.3 Å, α = 90.0°, β = 90.0°, and γ = 90.0° and
two ArsAB heterodimers in the asymmetric unit.
Data Collection and Structure Determination for
ArsAB Substrate-Free, DMB, p-cresol, and Phlorogluci-
nol Complexes. X-ray data for substrate-free ArsAB and
ArsAB complexed with DMB, p-cresol, and phloroglucinol were
collected at 100 K on the Structural Biology Center beamline
19BM at the Advanced Photon Source in Argonne, IL.
Diffraction data were integrated and scaled with HKL3000.17
Data collection statistics are given in Table 2. The ArsAB
heterodimer structure was determined using the structure of
CobT from Salmonella enterica (PDB entry 1L4B) as a
molecular replacement search model in the program Molrep.18
Final models were generated with alternate cycles of manual
model building and least-squares refinement using the
programs Coot19 and Refmac.20 Refinement statistics are
presented in Table 2.
Data Collection and Structure Determination for
ArsAB Complexed with Phenol. X-ray data for the ArsAB
complexed with phenol were collected at 100 K with a Bruker
AXS Platinum 135 CCD detector equipped with Montel optics
and controlled by the Proteum software suite (Bruker AXS
Inc.). All data sets were integrated with SAINT version 7.06A
and internally scaled with SADABS version 2005/1. The
structure was determined by molecular replacement using
PHASER21 in which ArsAB from the DMB complex structure
was used as a search model. Model refinement was performed
by alternate cycles of manual building with Coot19 and
restrained refinement with Refmac5.22 Refinement statistics
are presented in Table 2.
ArsAB Kinetic Assays. Pseudo-first-order kinetics of the
ArsAB-catalyzed reaction was performed using saturating
concentrations of NaMN (5 mM), varying amounts of DMB
or phenol (0.010−2 mM), and ArsAB protein at 0.25 μM.
Stock solutions of DMB and phenol substrates were made in
30% (v/v) ethanol. All reaction mixtures contained 100 mM
glycine pH 9 and 3% (v/v) ethanol at 37 °C. Triplicate assays
were performed and analyzed as described elsewhere.12
ArsAB Activity Assays. Activities of ArsAB and variants on
DMB and p-cresol were measured with 3 mM of substrates
(DMB/p-cresol and NaMN) and 0.2 μg/μL enzyme in 100
mM glycine, pH 9.0. The amount of enzyme used in this assay
was normalized to the least amount of variant protein needed
to give detectable activity. Twenty-five microliters of each
reaction was taken at 1, 6, and 13 min and boiled to stop the
reaction. An equal volume of 20 mM ammonium acetate buffer
8574
dx.doi.org/10.1021/bi301142h | Biochemistry 2012, 51, 8571−8582

Biochemistry
Article
Figure 2. Stereo representation of ArsAB heterodimer. ArsA is colored in blue, ArsB is colored in green. DMB is shown in yellow.
pH 4.5 (buffer B) was added to each reaction, and the protein
was filtered using a Spin-X column (Costar) before separation
on the HPLC. The product of each reaction was separated on a
Shimadzu Prominence UFLC equipped with a Kinetex 2.6 μm
C18 100 Å 100 × 4.60 mm column (Phenomenex) at a flow
rate of 0.8 mL/min. After injection, the column was developed
for 0.6 min in 97% B and 3% acetonitrile (ACN) before a linear
gradient to 60% B and 40% ACN over 6.4 min. Next, the
column was developed for 3.2 min to 100% ACN and held at
100% ACN for 3.2 min. Retention times (minutes) were as
follows: DMB, 9.7; p-cresol 11.2; α-DMB-riboside mono-
phosphate, 7.4 and α-p-cresolyl-riboside monophosphate 7.6.
Reactions were quantified against a standard curve derived from
0.1 - 30 nmol substrates.
MEVI residues (274 vs 24, respectively). Finally, the resulting
protein could be concentrated readily to more than 10 mg/mL.
Open reading frame (ORF) prediction software Gene-
Mark.hmm also predicted the downstream translational start
codon as the likely ORF for the gene.23 The plasmid
pARSAB22 was used to prepare ArsAB enzyme for all
subsequent crystallographic and kinetic studies.
Structure of the Substrate-Free ArsAB Heterodimer.
Crystals of substrate-free ArsAB were obtained by soaking out
the crystallization additive phloroglucinol, which bound in the
ArsA active site. The final substrate-free-ArsAB model contains
334 of 348 expected amino acids for the ArsA subunit which
extend from Ser2-Asn339. For the ArsB subunit the model
contains 328 of 350 expected amino acid residues, which extend
from Leu2-Ala333. The electron density is continuous for ArsA
except for a break between Ser208 and Leu212 in ArsA. This
disordered region consists of a loop that folds over the putative
binding site for the substrate NaMN. In the ArsA subunit this
loop is displaced by the N-terminus of a symmetry-related ArsB
molecule in what is presumably a crystal packing interaction.
The length of the break in this region varies among the
substrate complexes determined. There is also one break in the
electron density for ArsB between Ala70 and Met78. This break
occurs in a surface loop that spans the active site cavity. The
structure of substrate-free-ArsAB is shown in Figure 2.
Despite sharing only 37% sequence identity, ArsA and ArsB
exhibit highly similar tertiary structures with a root-mean-
square deviation (RMSD) of 1.3 Å for 311 structurally
equivalent α-carbon atoms. As expected, ArsA and ArsB have
a very similar fold to that described for SeCobT24 with an
overall RMSD of 1.25 Å with respect to either ArsA or ArsB
relative to PDB coordinates 1D0S. The overall fold contains
two domains where the large domain consists of a six-stranded
parallel β-sheet surrounded by α-helices in the β6, β5, β4, β1,
β2, β3 arrangement characteristic of the classic Rossmann
dinucleotide binding motif. The small domain is defined by a
three-helix bundle, which is built from two N-terminal helices
and a longer C-terminal helix that spans both domains.
RESULTS AND DISCUSSION
■
Identification of the Initiating Methionine of ArsA.
The initial characterization of the S. ovata arsAB genes
identified two potential translational initiation sites in the
arsA-coding region.12 The difference between the two putative
ArsA proteins was four amino acids (i.e., MEVI). In the
previous report on ArsAB,12 an overexpression vector
(pARSAB7), was used to isolate an N-terminally tagged protein
that included a hexahistidine (His6) tag and a rTEV protease
site followed by the MEVI residues mentioned above. The
same vector also directed the synthesis of ArsB without a tag.
After isolation of His6-ArsA, the His6-tag was removed with
rTEV protease,15 leaving two additional glycines prior to the
MEVI residues. ArsB copurified with ArsA, but the over-
production of ArsAB from pARSAB7 was poor under the
conditions tested, and the enzyme could not be concentrated to
more than 3 mg/mL (data not shown). This led to the question
of whether the true start codon was the alternative methionine.
Expression of arsAB from plasmid pARSAB22 eliminated the
MEVI residues of ArsA but retained the His6 affinity-tag and
the rTEV enzyme recognition site. Protein expression driven by
this construct yielded ArsAB enzyme that was readily purified
under multiple conditions. Furthermore, the specific activity
(nmol of product/min/mg of protein) of this material was 10-
fold higher than that of the enzyme containing the additional
The ArsAB heterodimer exhibits the same quaternary
arrangement as the SeCobT homodimer, with numerous
8575
dx.doi.org/10.1021/bi301142h | Biochemistry 2012, 51, 8571−8582

Biochemistry
Article
Figure 3. Structural alignment of ArsA with ArsB in stereo representation. ArsA (blue) with p-cresol bound (yellow) was superposed onto ArsB
(green). The original positions of ArsA and p-cresol are depicted in light gray. Arrows indicate areas of greatest divergence:1 the loop that folds over
the second substrate [ArsA: Gly206-Thr218, ArsB: Gly189-Gly197],2 the loop that projects into the active site on the adjacent subunit [ArsA: Val33-
Gly41, ArsB: Leu27-His35], and3 additional C-terminal helix in ArsB.
contacts along the dimer interface. The total buried surface area
of the dimer is about 2300 Å2/subunit as computed with the
program AREAIMOL in the ccp4 package.25 In SeCobT and
ArsAB the active site is located at the dimer interface and is
built from both subunits of the dimer. The C-terminus of one
Veillonellaceae family, suggesting that that the evolution of
heterodimeric SeCobT homologues may be rooted in these
bacteria.
Structure of ArsAB Heterodimer in Complex With
DMB. The ArsAB heterodimer in complex with DMB
crystallized in the space group P21 with two heterodimers in
the asymmetric unit. No significant differences in tertiary
structure were observed between the ArsAB heterodimer in
either crystal system (RMSD: 0.45 Å), and inspection of the
unit cell revealed no substantial change in crystal packing
interactions. It is unclear whether the shift in space group was
due to the presence of DMB, or by a change in crystallization
conditions which included the crystallization additive dieth-
ylenetriamine.
There is unambiguous electron density for DMB in the ArsA
active sites for both heterodimers in the asymmetric unit
(Figure 4A). The overall RMSD between the two heterodimers
in the asymmetric unit is 0.24 Å. Given the similarity between
the two heterodimers, the discussion and figures are based on
chains A and B in the asymmetric unit. The substrate binding
site lies in a cavity at the interface of the ArsA and ArsB
subunits, near the periphery of the dimer in a manner that is
similar to that seen for SeCobT.24 The cavity is defined by
helices 5 and 6 and their connecting loop in ArsA, the loop
leading into helix 16 in ArsA, and β4 and its adjoining loop in
ArsA. Additionally, the loop between helices 2 and 3 of the
ArsB subunit forms part of the cavity (Figure 4B). The cavity in
which DMB binds is hydrophobic. DMB sits atop Met87 and
Met177, Ile179, and Leu317 in ArsA and Pro30 in ArsB. An
imidazole nitrogen on DMB forms a hydrogen bond with the
proposed active site residue Glu319 (2.9 Å) (Glu317 in
SeCobT).26
24
subunit and a small N-terminal loop of the second subunit
form the site and result in two structurally equivalent active
sites per dimer. Due to the heterodimeric nature of ArsAB, the
two active sites in the dimer are structurally distinct.
There are notable structural differences in the surface loops
that surround the active site cavities of ArsA and ArsB. The
loop between helices 2 and 3 in ArsB (Leu27-Leu34) that
forms part of the active site for ArsA curls away at Pro30 in
ArsB in a fashion similar to that of SeCobT and provides
sufficient room for substrate binding in the ArsA active site.
Conversely, in ArsB the corresponding loop from ArsA
penetrates further into the analogous site in ArsB (Figure 3).
ArsB also contains an additional C-terminal helix spanning
Phe327-Glu331 that projects from the three helix bundle of the
small domain and is directed away from the ArsA active site.
There is no observable electron density for the C-terminal 17
amino acids of ArsB. These differences, along with others
discussed later, indicate that DMB phosphoribosyltransferase
activity has been lost from the ArsB active site.
The observation that ArsAB heterodimers crystallized
supports the earlier suggestion that neither ArsA nor ArsB
can assemble independently to form an active enzyme.12 ArsAB
is likely the result of a gene duplication event onto which
selective forces resulted in the coevolution of ArsA and ArsB
into functional a heterodimer. Why functional homodimers
were not favored remains unclear. At present, information
available from genome databases shows that the tandem
organization of arsAB is only found in some members of
8576
dx.doi.org/10.1021/bi301142h | Biochemistry 2012, 51, 8571−8582

Biochemistry
Article
Figure 4. (A) Stereoview of the electron density for DMB in the ArsA active site. The electron density (3.5σ) was calculated from coefficients of the
form Fo − Fc where DMB was omitted from the phase calculation and refinement. ArsB has been omitted for clarity. (B) The DMB binding site is
formed by contributions from both subunits: ArsA is shown in blue and ArsB is shown in green.
Figure 5. Alignment of SeCobT and ArsAB in complex with DMB. Coordinates for SeCobT·DMB were derived from PDB ID: 1D0S.24 DMB from
SeCobT is colored violet. DMB from ArsAB is colored yellow.
DMB Binding Is Conserved. Alignment of DMB-bound
structures for ArsAB and SeCobT reveals that DMB binds in a
nearly identical orientation in both enzymes (Figure 5). This is
consistent with the sequence and structural conservation of the
two enzymes. The character of most of the side chains in the
DMB binding site are conserved between ArsA and SeCobT,
many of which are hydrophobic. Notable exceptions are the
substitution of both Gln88 and Leu175 in SeCobT to
methionine in ArsA. These methionine residues form a large
part of the hydrophobic binding surface for DMB in ArsA. The
most apparent difference in first shell interactions between
SeCobT and ArsAB is the replacement of Ser80 in SeCobT with
Tyr79 in ArsA. This substitution likely compensates for the
absence of hydrophobic surface contributed by the C-terminus
of SeCobT. There is no evidence that the largely disordered C-
terminus of ArsB makes contacts with the active site of ArsA.
Tyr79 of ArsA may play a role in the activation of phenolics as
discussed later. A key conserved interaction in the active site of
8577
dx.doi.org/10.1021/bi301142h | Biochemistry 2012, 51, 8571−8582

Biochemistry
Article
Figure 6. The active site of ArsB is inconsistent with NaMN:DMB phosphoribosyltransferase activity. This shows the alignment of SeCobT products
complex with the ArsB active site. Coordinates for SeCobT in complex with its reaction products came from the PDB (accession number 1D0V).24
α-Ribazole-5′-phosphate (α-RP) and nicotinate (NA) are colored violet. ArsB is colored green. SeCobT is colored gray. The loop that folds over the
NaMN binding site in SeCobT is highlighted in dark gray. This loop is shortened by 11 residues in ArsB as can be seen in the far right of the image.
Figure 7. Stereoview for the electron density of (A) p-cresol and (B) phenol bound in the active site of ArsA. The electron density (3.0σ p-cresol,
2.0σ phenol) was calculated from coefficients of the form Fo − Fc where p-cresol and phenol were omitted from the phase calculations and
refinements. ArsB has been omitted for clarity.
ArsA is the hydrogen bond between Glu319 (Glu317 in
SeCobT) and one of the imidazole nitrogen atoms of DMB. In
addition to its role in coordinating and orienting the substrate,
this residue has been proposed to act as an active site base24 in
SeCobT. This constellation of interactions is also observed in
ArsAB providing strong evidence that ArsAB enzyme may
utilize the same catalytic mechanism as SeCobT to catalyze
phosphoribosyl transfer.
The second substrate, NaMN, was not captured in complex
with ArsAB. Indeed, the portion of the NaMN binding site in
SeCobT that lies atop the phosphate moiety is disordered in the
structurally equivalent region in ArsAB. This disordered loop
8578
dx.doi.org/10.1021/bi301142h | Biochemistry 2012, 51, 8571−8582

Biochemistry
Article
Figure 8. (A) Alignment of SeCobT and ArsAB in complex with p-cresol. Coordinates for SeCobT·p-cresol and for SeCobT·NAMN came from PDB
accession numbers 1JHU and 1L4M respectively.13 p-Cresol is positioned by hydrogen bonds to an active site water molecule and the side chain of
Tyr79. p-cresol (magenta) is bound to SeCobT, and p-cresol (yellow) is bound to ArsA. (B) A surface representation of ArsA and bound p-cresol
shows that Tyr79 plays a role in positioning of the phenolic substrate. (C) Proposed mechanism for phenolic activation. ArsA is depicted in light
blue whereas SeCobT is colored in white.
spans helices 9 and 10 in ArsA, and may undergo a disorder/
order transition upon binding the second substrate. Alter-
natively, crystal-packing interactions may have displaced these
residues as there are numerous contacts between the N-
terminus of a symmetry related molecule of ArsB (Glu3 and
Glu4) and the NaMN binding pocket in ArsA (Lys215).
Despite the lack of structural information in this region, many
key interactions, such as those between side chain of SeGlu174
and the amide nitrogen of SeGly176 with the ribose moiety are
observed to be in structurally equivalent positions in ArsA
(Glu179 and Gly 178). The cavity in which the nicotinate
moiety binds is also maintained, however a slight rearrange-
ment is observed around Ser293 in ArsA as compared to
SeCobT, where SeSer291 interacts with an oxygen of the
carboxylate group of nicotinate. In ArsA, the peptide backbone
of the Ser293 has rotated nearly 180° leaving the side chain
directed toward the solvent rather than the nicotinate binding
cavity.
DMB Phosphoribosyl Transferase Activity Is not
Maintained in ArsB. While there is a strong conservation of
active site architecture and substrate binding between SeCobT
and ArsA, the equivalent site in ArsB shows considerable
divergence (Figure 6). The most striking difference is the
position of the side chain of Arg83, which lies directly within
the DMB binding site. Additionally, the hydrophobic loop that
is contributed by the adjacent ArsA subunit lies much deeper in
the cavity and is not curled away from the site by Pro30 (ArsB)
as is observed in the ArsA DMB binding site. Considered
together, these structural features of ArsAB likely preclude the
binding of DMB in a conformation that is catalytically
competent. Interestingly, the proposed catalytic residue
Glu303 is maintained in a structurally equivalent position in
ArsB. The NaMN-binding site has also undergone several
structural rearrangements as depicted in Figure 6. The loop that
folds over the phosphate moiety in NaMN in SeCobT is
shortened in ArsB and tracks along the outer surface of the
cavity rather than toward the interior. A sequence alignment of
SeCobT and ArsB shows an 11 residue gap in this region in
ArsB. Lastly, the loop that includes Tyr79 in ArsA and spans β1
and helix 4 is disordered in ArsB but is directed inward toward
the active site cavity before becoming disordered. The
structural divergence of the active site in ArsB could be the
result of selective pressures leading to heterodimer formation,
or to the ability of ArsA to transfer phenolic bases. At this point,
it is unknown whether the ArsB site has any catalytic activity
whatsoever. It is conceivable that the ArsB active site has
evolved to perform another function. In this respect, although
the ArsB active site does not have base:phosphosylribosyl
transferase activity, the presence of phloroglucinol (Figure S1,
Supporting Information) in the ArsB active site suggests that
the capacity to bind phenolic bases has been maintained.
Structure of ArsAB Heterodimer in Complex with
Phenol and p-Cresol. To gain insight into the mechanism by
which ArsAB generates phenolic α-riboside monophosphates,
crystal structures were determined for ArsAB in complex with
both p-cresol and phenol. There is clear electron density in the
ArsA active site for both substrates (Figure 7). The phenolic
substrates bind in the same location in the active site cavity at
8579
dx.doi.org/10.1021/bi301142h | Biochemistry 2012, 51, 8571−8582

Biochemistry
Article
the interface between the ArsA and ArsB subunits. The
constellation of side chains surrounding the bound substrate
remains unchanged relative to the complex with DMB with the
notable exception of the hydrogen bonding pattern about
Glu319 and the loop between helices 4 and 5, most notably at
Tyr79. While Glu319 participates in a direct hydrogen bond
with the imidazole nitrogen in complex DMB (2.9 Å), it has
been replaced with a well-ordered water molecule with phenolic
substrates bound. In the case of p-cresol, this water molecule
(W1) is 2.8 Å from Oε2 of Glu319 and 2.4 Å from the hydroxyl
moiety on p-cresol (Figure 8). The presence of W1 results in a
rotation of the planar substrate away from Glu319 and toward
the NaMN binding site. This position is additionally
maintained by the rotation of Tyr79 3.6 Å inward from that
of the DMB complex, resulting in a close interaction with p-
cresol (Figure 8). Taken together, these interactions likely
explain how p-cresol is positioned for a nucleophilic attack at
the C1 carbon of ribose following deprotonation, which is
facilitated by the close hydrogen bond with W1, as it is ideally
positioned to transfer a proton to Glu319.
Comparison of SeCobT with ArsAB Bound to Phenolic
Substrates. SeCobT is unable to catalyze phosphoribosyl
transfer to phenolic substrates, whereas this is readily
accomplished by ArsAB. The presence of a water molecule in
the active site of the ArsAB:phenolic complexes compared to
the absence of a water molecule in the phenolic complexes with
SeCobT suggests that a water molecule plays a key role in the
positioning and deprotonation of p-cresol for attack at the C1
carbon of ribose. In SeCobT, p-cresol hydrogen bonds directly
(2.6 Å) with Oε2 of Glu317. This close hydrogen bond
between p-cresol and Glu317, places the phenolic hydroxyl
group hydroxyl group in a position that is most likely too far
from the C1 carbon of ribose (>4.0 Å) for the reaction to
occur. The reason that SeCobT does not allow inclusion of a
water molecule in the equivalent position of W1 in ArsAB is
not immediately obvious, but careful examination of the
comparison of the active sites in ArsAB and SeCobT suggests
several possibilities.
Superposition of the active sites (Figure 9) shows that in
order for ArsA to accommodate a water molecule the catalytic
glutamate Glu319 (Glu317 in SeCobT) must move away from
its position in SeCobT. This would appear to be facilitated by
the replacement of Gln88 in SeCobT by Met87 in ArsA. This
change removes a hydrogen bonding interaction which is
predicted to stabilize the position of the catalytic glutamate in
the active site. There is also a difference in the position and
conformation of the backbone atoms for the polypeptide chain
that surrounds the catalytic base. This change appears to be
coupled to a number of changes in this region. In particular,
replacement of Ser319 in SeCobT by Ile321 in ArsA requires a
rearrangement in the backbone to accommodate the loss of a
hydrogen bond.
Figure 9. Inclusion of a water molecule in the active site of ArsA.
Structural overlay of ArsAB and SeCobT complexed with p-cresol
shows that subtle changes in the position of the catalytic glutamate
allow the incorporation of a water molecule into the active site of
ArsA. It is hypothesized that these differences are facilitated by
removal of a hydrogen bond in the active site (replacement of a
glutamine by a methionine) and conformational changes in the
backbone associated with Ile321. The coordinates of SeCobT
complexed with p-cresol were taken from PDB accession number
1JHU.13
than one residue of the base-binding pocket. A full under-
standing of the structural basis for the utilization of phenolic
substrates by ArsAB must await further structural and kinetic
studies of variants.
Efficiency of the ArsAB Enzyme as a Function of its
Base Substrate. Earlier work by Stupperich et al. showed that
p-cresolyl-Cba and phenolyl-Cba comprised >90% of the
cobamides synthesized by S. ovata,8,9 yet data shown in Table
4 indicate that the ArsAB enzyme efficiently uses DMB as its
base substrate. However, the KM of ArsAB for DMB reported
herein (415 μM) is much higher than other CobT homologues,
namely S. enterica (KM < 10 μM) and Pseudomonas denitrificans
(16 nM).27,28 The dramatic increase in KM for DMB in ArsAB
may be a consequence of the evolutionary changes required for
this enzyme to phosphoribosylate phenolic bases. From this
perspective the higher KM for DMB observed with ArsAB
suggests that increased flexibility needed to accommodate a
water molecule might be a compromise to maintain both
activities. The crystallographic snapshots of ArsAB and SeCobT
show DMB bound in the same position and orientation in both
enzyme active sites so there is no easy structural explanation for
the observed difference in KM between ArsAB and SeCobT.
Interestingly, the interaction between the C-terminal residues
contributed by the neighboring subunit in SeCobT and DMB
are apparently lacking in ArsAB. This could affect on-rates and
off-rates of substrate and thereby KM.
The hypothesis that Met87 and Ile321 play a role in enabling
ArsAB to catalyze phosphribosyl transfer to phenolic substrates
was tested by preparing the reverse mutations M87Q and I321S
(Table 3). The effect of the M87Q mutation is most
pronounced with respect to phenolic specificity, as the specific
activity with p-cresol drops by 2.6 fold while a slight increase
(1.2 fold) in activity with DMB was observed. Introduction of
serine at Ile321 was deleterious to both phenolic and DMB
activation, suggesting that the role of this amino acid residue in
ArsA is not well understood. From these experiments it is clear
that specificity for phenolic substrates is influenced by more
The catalytic efficiency of the activation of DMB and phenol
for ArsAB differ only by a factor of 2 (Table 4), suggesting that,
8580
dx.doi.org/10.1021/bi301142h | Biochemistry 2012, 51, 8571−8582

Biochemistry
Article
a
Table 3. Specific Activities for Variants of ArsA towards DMB and p-Cresol
substrate
wild type
M87Q
I321S
M87Q/I321S
p-cresol
5,6-dimethylbenzimidazole
37.6 0.8
14.2 1.6
8.3 1.7
0.7 0.1
2.4 1.0
160.7 7.9
192.6 12.2
16.7 1.9
a
Specific activities are represented in nmol product formed/min/mg enzyme, from 1 to 13 min, three time points taken to ensure linearity, averages
and standard deviations taken from three replicate reactions.
Accession Codes
Table 4. Kinetic Parameters of the ArsAB-Catalyzed
Reaction
a
X-ray coordinates for the substrate-free form and in complex
with DMB, p-cresol, phloroglucinol, and phenol have been
deposited in the Research Collaboratory for Structural
Bioinformatics, Rutgers University, New Brunswick, N. J.
(Protein Data Bank entries 4HDN, 4HDR, 4HDM, 4HDK, and
4HDS respectively)
substrate
KM (μM)
kcat (s−1
)
kcat/KM (s−1 M−1
)
5,6-dimethylbenzimidazole
phenol
415 56
50 11
1.6 0.08
0.4 0.05
3.9 × 103
8 × 103
a
Reaction mixtures contained either 1.9 mM phenol + [14C−U]-
phenol (0.1 mM, 60 mCi mmol−1), or 1.98 mM 5,6-dimethylbenzi-
midazole (DMB) + [14C-2]-DMB (0.02 mM, 43 mCi mmol−1). Other
components present in the reaction mixture and the protocol for the
assessment of product formation are described under Experimental
Procedures.
AUTHOR INFORMATION
Corresponding Author
■
*(I.R.) Department of Biochemistry, University of Wisconsin,
433 Babcock Dr., Madison, WI 53706; phone, (608) 262-0437;
fax, (608) 262-1319; e-mail, ivan_rayment@biochem.wisc.edu.
(J.C.E.-S.) Department of Microbiology, University of Georgia,
527 Biological Sciences Building, 120 Cedar Street, Athens, GA
30602, phone, (706) 542 2651; fax, 706 542 2815; e-mail
jcescala@uga.edu.
under some conditions, this bacterium may synthesize
cobalamin (Coβ-5,6-dimethylbenzimidazolyl-Cba) or other
cobamides containing purines or purine analogues.5,12 Notably,
while the apparent KM for phenol was 8-fold lower than the
apparent KM for DMB, the turnover number when phenol was
the substrate was 4-fold slower than the one calculated for the
reaction when DMB was the substrate (Table 4). It is
noteworthy that phenolyl-Cba comprises only 20% of the
pool of phenolic-Cbas synthesized by S. ovata so that the
kinetic parameters for phenol could differ from those for p-
cresol. The kinetic parameters for p-cresol were not measured
because 14C-p-cresol necessary for the kinetic assay developed
and used in this study is not commercially available.
Author Contributions
§These authors contributed equally to this work.
Funding
This work was supported in part by NIH Grants GM083987
and GM086351 to I.R. and R37 GM40313 to J.C.E.-S. Use of
the SBC 19 μB beamline at the Argonne National Laboratory
Advanced Photon Source was supported by the U.S. Depart-
ment of Energy, Office of Energy Research, under Contract No.
W-31-109-ENG-38.
Notes
The authors declare no competing financial interest.
CONCLUSIONS
■
The structural studies described here confirm that the ArsAB
heterodimer from Sporomusa ovata contains one functional
active site for nicotinate mononucleotide:phenol/p-cresol
phosphoribosyltransfer, which is formed mostly by amino
acid residues from the ArsA subunit. The role of the ArsB
subunit is unclear; though it does retain the ability to weakly
bind phenolic bases, which allows the possibility that this
component of the protein has evolved a new but unknown
function. The complexes of ArsAB with phenol and p-cresol
suggest that ArsAB has evolved to accommodate phosphor-
ibosyl transfer to phenolic substrates by allowing a water
molecule to enter into the active site. This water molecule
serves to reposition phenolic substrates such that the hydroxyl
moiety is closer to the putative location of the C1 carbon on
ribose which is predicted to facilitate nucleophilic attack.
Comparison of the structure of ArsAB with the NaMN:DMB
phosphoribosyltransferase (CobT) enzyme of Salmonella
enterica suggests that comparatively few changes are required
to facilitate phosphoribosyl transfer to phenolic substrates.
ABBREVIATIONS USED
■
Ado, adenosyl; cobamide, Cba; 5,6-dimethylbenzimidazole,
DMB; nicotinate mononucleotide, NaMN; 4-(2-hydroxyeth-
yl)-1-piperazine-N′-(2-ethanesulfonic acid, HEPES; methyl
ether polyethylene glycol 5000, MEPEG5K; NaMN, DMB
phosphoribosyltransferase (CobT) enzyme of Salmonella
enterica, (SeCobT); 3-(N-morpholino)propanesulfonic acid,
MOPS; root-mean-square deviation, RMSD
REFERENCES
■
(1) Battersby, A. R. (2000) Tetrapyrroles: the pigments of life. Nat.
Prod. Rep. 17, 507−526.
(2) Banerjee, R. (2003) Radical carbon skeleton rearrangements:
catalysis by coenzyme B12-dependent mutases. Chem. Rev. 103, 2083−
2094.
(3) Yamanishi, M., Yunoki, M., Tobimatsu, T., Sato, H., Matsui, J.,
Dokiya, A., Iuchi, Y., Oe, K., Suto, K., Shibata, N., Morimoto, Y.,
Yasuoka, N., and Toraya, T. (2002) The crystal structure of coenzyme
B12-dependent glycerol dehydratase in complex with cobalamin and
propane-1,2-diol. Eur. J. Biochem. 269, 4484−4494.
(4) Matthews, R. G., Koutmos, M., and Datta, S. (2008) Cobalamin-
dependent and cobamide-dependent methyltransferases. Curr. Opin.
Struct. Biol. 18, 658−666.
(5) Renz, P. (1999) Biosynthesis of the 5,6-dimethylbenzimidazole
moiety of cobalamin and of other bases found in natural corrinoids, in
Chemistry and Biochemistry of B12 (Banerjee, R., Ed.) pp 557−575,
John Wiley & Sons, Inc., New York.
ASSOCIATED CONTENT
* Supporting Information
A stereoview of the electron density for phloroglucinol (pho)
bound to ArsB in the complex of ArsAB with phloroglucinol.
This material is available free of charge via the Internet at
http://pubs.acs.org.
■
S
8581
dx.doi.org/10.1021/bi301142h | Biochemistry 2012, 51, 8571−8582

Biochemistry
Article
(6) Yi, S., Seth, E. C., Men, Y. J., Stabler, S. P., Allen, R. H., Alvarez-
Cohen, L., and Taga, M. E. (2012) Versatility in Corrinoid Salvaging
and Remodeling Pathways Supports the Corrinoid-Dependent
Metabolism of Dehalococcoides mccartyi. Appl. Environ. Microbiol. 78,
7745−7752.
tide:5,6- dimethylbenzimidazole phosphoribosyltransferase (CobT)
from Salmonella typhimurium complexed with 5,6-dimethybenzimida-
zole and its reaction products determined to 1.9 Å resolution.
Biochemistry 38, 16125−16135.
(25) Winn, M. D., Ballard, C. C., Cowtan, K. D., Dodson, E. J.,
Emsley, P., Evans, P. R., Keegan, R. M., Krissinel, E. B., Leslie, A. G.,
McCoy, A., McNicholas, S. J., Murshudov, G. N., Pannu, N. S.,
Potterton, E. A., Powell, H. R., Read, R. J., Vagin, A., and Wilson, K. S.
(2011) Overview of the CCP4 suite and current developments. Acta
Crystallogr., Sect. D 67, 235−242.
(26) Cheong, C. G., Escalante-Semerena, J. C., and Rayment, I.
(2002) Capture of a labile substrate by expulsion of water molecules
from the active site of nicotinate mononucleotide:5,6-dimethylbenzi-
midazole phosphoribosyltransferase (CobT) from Salmonella enterica.
J. Biol. Chem. 277, 41120−41127.
(27) Trzebiatowski, J. R., and Escalante-Semerena, J. C. (1997)
Purification and characterization of CobT, the nicotinate-mononucleo-
tide:5,6-dimethylbenzimidazole phosphoribosyltransferase enzyme
from Salmonella typhimurium LT2. J. Biol. Chem. 272, 17662−17667.
(28) Cameron, B., Blanche, F., Rouyez, M. C., Bisch, D., Famechon,
A., Couder, M., Cauchois, L., Thibaut, D., Debussche, L., and Crouzet,
J. (1991) Genetic analysis, nucleotide sequence, and products of two
Pseudomonas denitrificans cob genes encoding nicotinate-nucleotide:
dimethylbenzimidazole phosphoribosyltransferase and cobalamin (5′-
phosphate) synthase. J. Bacteriol. 173, 6066−6073.
(7) Escalante-Semerena, J. C., Warren, M. J. (2008) Biosynthesis and
Use of Cobalamin (B12), in EcoSal - Escherichia coli and Salmonella:
Cellular and Molecular Biology (Bock, A., Curtiss III, R., Kaper, J. B.,
̈
Karp, P. D., Neidhardt, F. C., Nystrom, T., Slauch, J. M., Squires, C. L.,
̈
Eds.) ASM Press, Washington, D. C.
(8) Stupperich, E., Eisinger, H. J., and Krautler, B. (1989)
̈
Identification of phenolyl cobamide from the homoacetogenic
bacterium Sporomusa ovata. Eur. J. Biochem. 186, 657−661.
(9) Stupperich, E., Eisinger, H. J., and Krautler, B. (1988) Diversity of
̈
corrinoids in acetogenic bacteria. p-Cresolylcobamide from Sporomusa
ovata, 5-methoxy-6-methylbenzimidazolylcobamide from Clostridium
formicoaceticum and vitamin B12 from Acetobacterium woodii. Eur. J.
Biochem. 172, 459−464.
(10) Abend, A., Bandarian, V., Nitsche, R., Stupperich, E., Retey, J.,
and Reed, G. H. (1999) Ethanolamine ammonia-lyase has a “base-on”
binding mode for coenzyme B12. Arch. Biochem. Biophys. 370, 138−
141.
́
(11) Poppe, L., Stupperich, E., Hull, W. E., Buckel, T., and Retey, J.
(1997) A base-off analogue of coenzyme-B12 with a modified
nucleotide loop–1H-NMR structure analysis and kinetic studies with
(R)-methylmalonyl-CoA mutase, glycerol dehydratase, and diol
dehydratase. Eur. J. Biochem. 250, 303−307.
(12) Chan, C. H., and Escalante-Semerena, J. C. (2011) ArsAB, a
novel enzyme from Sporomusa ovata activates phenolic bases for
adenosylcobamide biosynthesis. Mol. Microbiol. 81, 952−967.
(13) Cheong, C. G., Escalante-Semerena, J. C., and Rayment, I.
(2001) Structural investigation of the biosynthesis of alternative lower
ligands for cobamides by nicotinate mononucleotide: 5,6-dimethyl-
benzimidazole phosphoribosyltransferase from Salmonella enterica. J.
Biol. Chem. 276, 37612−37620.
(14) Rocco, C. J., Dennison, K. L., Klenchin, V. A., Rayment, I., and
Escalante-Semerena, J. C. (2008) Construction and use of new cloning
vectors for the rapid isolation of recombinant proteins from
Escherichia coli. Plasmid 59, 231−237.
(15) Blommel, P. G., and Fox, B. G. (2007) A combined approach to
improving large-scale production of tobacco etch virus protease.
Protein Expr. Purif. 55, 53−68.
(16) Blommel, P. G., Becker, K. J., Duvnjak, P., and Fox, B. G. (2007)
Enhanced bacterial protein expression during auto-induction obtained
by alteration of lac repressor dosage and medium composition.
Biotechnol. Prog. 23, 585−598.
(17) Otwinowski, Z., and Minor, W. (1997) Processing of X-ray
diffraction data collected in oscillation mode. Methods Enzymol. 276,
307−326.
(18) Vagin, A., and Teplyakov, A. (2000) An approach to multi-copy
search in molecular replacement. Acta Crystallogr. D Biol. Crystallogr.
56, 1622−1624.
(19) Emsley, P., and Cowtan, K. (2004) Coot: model-building tools
for molecular graphics. Acta Crystallogr. D Biol. Crystallogr. 60, 2126−
2132.
(20) Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997)
Refinement of macromolecular structures by the Maximum-Likelihood
Method. Acta Crystallogr. D Biol. Crystallogr. 53, 240−255.
(21) McCoy, A. J., Grosse-Kunstleve, 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.
(22) Murshudov, G. N., Skubak, P., Lebedev, A. A., Pannu, N. S.,
Steiner, R. A., Nicholls, R. A., Winn, M. D., Long, F., and Vagin, A. A.
(2011) REFMAC5 for the refinement of macromolecular crystal
structures. Acta Crystallogr. D Biol. Crystallogr. 67, 3553−3567.
(23) Lukashin, A. V., and Borodovsky, M. (1998) GeneMark.hmm:
new solutions for gene finding. Nucleic Acids Res. 26, 1107−1115.
(24) Cheong, C. G., Escalante-Semerena, J. C., and Rayment, I.
(1999) The three-dimensional structures of nicotinate mononucleo-
8582
dx.doi.org/10.1021/bi301142h | Biochemistry 2012, 51, 8571−8582