Enabling microbial syringol conversion through structure-guided protein engineering

Article abstract of DOI:10.1073/pnas.1820001116

Microbial conversion of aromatic compounds is an emerging and promising strategy for valorization of the plant biopolymer lignin. A critical and often rate-limiting reaction in aromatic catabolism is O-aryl-demethylation of the abundant aromatic methoxy groups in lignin to form diols, which enables subsequent oxidative aromatic ring-opening. Recently, a cytochrome P450 system, GcoAB, was discovered to demethylate guaiacol (2-methoxyphenol), which can be produced from coniferyl alcohol-derived lignin, to form catechol. However, native GcoAB has minimal ability to demethylate syringol (2,6-dimethoxyphenol), the analogous compound that can be produced from sinapyl alcohol-derived lignin. Despite the abundance of sinapyl alcohol-based lignin in plants, no pathway for syringol catabolism has been reported to date. Here we used structure-guided protein engineering to enable microbial syringol utilization with GcoAB. Specifically, a phenylalanine residue (GcoA-F169) interferes with the binding of syringol in the active site, and on mutation to smaller amino acids, efficient syringol O-demethylation is achieved. Crystallography indicates that syringol adopts a productive binding pose in the variant, which molecular dynamics simulations trace to the elimination of steric clash between the highly flexible side chain of GcoA-F169 and the additional methoxy group of syringol. Finally, we demonstrate in vivo syringol turnover in Pseudomonas putida KT2440 with the GcoA-F169A variant. Taken together, our findings highlight the significant potential and plasticity of cytochrome P450 aromatic O-demethylases in the biological conversion of lignin-derived aromatic compounds.

Full text of DOI:10.1073/pnas.1820001116

Enabling microbial syringol conversion through
structure-guided protein engineering
Melodie M. Machovinaa,1, Sam J. B. Mallinsonb,1, Brandon C. Knottc,1, Alexander W. Meyersd,1, Marc Garcia-Borràse,1
Lintao Buc, Japheth E. Gadod,f, April Olivera, Graham P. Schmidtc, Daniel J. Hinchenb, Michael F. Crowleyc,
,
Christopher W. Johnsond, Ellen L. Neidleg, Christina M. Paynef, Kendall N. Houke,2, Gregg T. Beckhamd,h,2
John E. McGeehanb,2, and Jennifer L. DuBoisa,2
,
aDepartment of Chemistry and Biochemistry, Montana State University, Bozeman, MT 59717; bCentre for Enzyme Innovation, School of Biological Sciences,
Institute of Biological and Biomedical Sciences, University of Portsmouth, Portsmouth PO1 2UP, United Kingdom; cBiosciences Center, National Renewable
Energy Laboratory, Golden, CO 80401; dNational Bioenergy Center, National Renewable Energy Laboratory, Golden, CO 80401; eDepartment of Chemistry
and Biochemistry, University of California, Los Angeles, CA 90095; fDepartment of Chemical Engineering, University of Kentucky, Lexington, KY 40506;
gDepartment of Microbiology, University of Georgia, Athens, GA 30602; and hCenter for Bioenergy Innovation, Oak Ridge National Laboratory, Oak Ridge,
TN 37830
Contributed by Kendall N. Houk, May 11, 2019 (sent for review November 26, 2018; reviewed by Lindsay D. Eltis, Erika A. Taylor, and Binju Wang)
Microbial conversion of aromatic compounds is an emerging and
promising strategy for valorization of the plant biopolymer lignin.
A critical and often rate-limiting reaction in aromatic catabolism is
O-aryl-demethylation of the abundant aromatic methoxy groups
in lignin to form diols, which enables subsequent oxidative aro-
matic ring-opening. Recently, a cytochrome P450 system, GcoAB,
was discovered to demethylate guaiacol (2-methoxyphenol), which
can be produced from coniferyl alcohol-derived lignin, to form cat-
echol. However, native GcoAB has minimal ability to demethylate
syringol (2,6-dimethoxyphenol), the analogous compound that can
be produced from sinapyl alcohol-derived lignin. Despite the abun-
dance of sinapyl alcohol-based lignin in plants, no pathway for syrin-
gol catabolism has been reported to date. Here we used structure-
guided protein engineering to enable microbial syringol utilization
with GcoAB. Specifically, a phenylalanine residue (GcoA-F169) inter-
feres with the binding of syringol in the active site, and on muta-
tion to smaller amino acids, efficient syringol O-demethylation is
achieved. Crystallography indicates that syringol adopts a produc-
tive binding pose in the variant, which molecular dynamics simula-
tions trace to the elimination of steric clash between the highly
flexible side chain of GcoA-F169 and the additional methoxy group
of syringol. Finally, we demonstrate in vivo syringol turnover in
Pseudomonas putida KT2440 with the GcoA-F169A variant. Taken
together, our findings highlight the significant potential and plastic-
ity of cytochrome P450 aromatic O-demethylases in the biological
conversion of lignin-derived aromatic compounds.
interest (7–12) and essential for economical lignocellulose con-
version (11, 13, 14).
In most plants, lignin comprises primarily coniferyl (G) and
sinapyl (S) alcohol monomers, which have one methoxy group
and two methoxy groups on the aryl ring, respectively. Nearly all
lignin-derived aromatics require O-demethylation of these methoxy
groups as an essential step in their conversion to central interme-
diates. Significant effort has been dedicated to the discovery of
enzymes that can demethylate the methoxy substituents of diverse
aromatic compounds (15–26). Ornston et al. (18, 19) characterized
the O-demethylation of vanillin (4-hydroxy-3-methoxybenzaldehyde)
and vanillate (4-hydroxy-3-methoxybenzoate) analogs by the VanAB
monooxygenase from Acinetobacter baylyi ADP1, which contains a
Rieske nonheme iron center. The three-component LigX
Significance
Lignin is an abundant but underutilized heterogeneous poly-
mer found in terrestrial plants. In current lignocellulosic bio-
refinery paradigms, lignin is primarily slated for incineration,
but for a nonfood plant-based bioeconomy to be successful,
lignin valorization is critical. An emerging concept to valorize
lignin uses aromatic–catabolic pathways and microbes to fun-
nel heterogeneous lignin-derived aromatic compounds to sin-
gle high-value products. For this approach to be viable, the
discovery and engineering of enzymes to conduct key reactions
is critical. In this work, we have engineered a two-component
cytochrome P450 enzyme system to conduct one of the most
important reactions in biological lignin conversion: aromatic O-
demethylation of syringol, the base aromatic unit of S-lignin,
which is highly abundant in hardwoods and grasses.
demethylase P450 lignin biorefinery
|
|
|
ignin is a heterogeneous, recalcitrant biopolymer prevalent in
plant cell walls, where it provides structure, defense against
L
pathogens, and water and nutrient transport through plant tissue
(1). Lignin is synthesized primarily from three aromatic building
blocks (1, 2), making it the only abundant and renewable aro-
matic carbon feedstock available. Due to its recalcitrance, rot
fungi and some bacteria have evolved powerful, oxidative en-
zymes that deconstruct lignin to smaller fragments (3, 4). Once
broken down, the lignin oligomers can be assimilated as a carbon
and energy source through at least four known aromatic-catabolic
pathways (2, 5).
Author contributions: E.L.N., C.M.P., K.N.H., G.T.B., J.E.M., and J.L.D. designed research;
M.M.M., S.J.B.M., B.C.K., A.W.M., M.G.-B., L.B., J.E.G., A.O., G.P.S., D.J.H., M.F.C., C.W.J.,
and K.N.H. performed research; G.T.B., J.E.M., and J.L.D. contributed new reagents/
analytic tools; M.M.M., S.J.B.M., B.C.K., A.W.M., M.G.-B., L.B., J.E.G., M.F.C., C.W.J., C.M.P.,
K.N.H., G.T.B., J.E.M., and J.L.D. analyzed data; and M.M.M., S.J.B.M., K.N.H., G.T.B.,
J.E.M., and J.L.D. wrote the paper.
Reviewers: L.D.E., University of British Columbia; E.A.T., Wesleyan University; and B.W.,
Xiamen University.
The authors declare no conflict of interest.
This open access article is distributed under Creative Commons Attribution License 4.0
(CC BY).
A critical reaction in the aerobic catabolism of lignin-
derived compounds is O-aryl-demethylation, which occurs on
methoxylated lignin-derived compounds to produce aromatic
diols, such as catechol (1,2-dihydroxybenzene), protocatechuate
(3,4-dihydroxybenzoate), and gallate (3,4,5-trihydroxybenzoate).
Next, the aromatic rings are cleaved by intradiol or extradiol
dioxygenases, and the products are funneled into central metab-
olism (6, 7). Harnessing this catabolic capability for transforming
heterogeneous lignin streams into valuable chemicals is of keen
Data deposition: Protein crystal structure coordinates have been deposited in the Protein
Data Bank, https://www.wwpdb.org/ (PDB ID codes 6HQK–6HQT).
1M.M.M., S.J.B.M., B.C.K., A.W.M., and M.G.-B. contributed equally to this work.
2To whom correspondence may be addressed. Email: houk@chem.ucla.edu, gregg.
beckham@nrel.gov, john.mcgeehan@port.ac.uk, or jennifer.dubois1@montana.edu.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
1073/pnas.1820001116/-/DCSupplemental.
www.pnas.org/cgi/doi/10.1073/pnas.1820001116
PNAS Latest Articles
|
1 of 7

monooxygenase system from Sphingobium sp. SYK-6, described
by Masai et al. (22), also contains a Rieske nonheme iron component
that is responsible for O-aryl-demethylation of a model biphenyl
compound that mimics those in lignin. Masai et al. (20, 21) de-
scribed two tetrahydrofolate-dependent enzymes, LigM and
DesA, responsible for O-aryl-demethylation of vanillate and syringate
(4-hydroxy-3,5-dimethoxybenzoate), respectively. Cytochrome P450
systems have also been reported to demethylate aromatic com-
pounds, such as guaiacol, 4-methoxybenzoate, and guaethol
(2-ethoxyphenol) (15, 17, 25, 26); however, the full gene sequences
were either unreported or only recently identified (15, 17, 27), or
the substrate was not of direct interest to lignin conversion (25, 26).
Our recent characterization of a two-component P450 enzyme
system consisting of a reductase, GcoB, and a P450 oxidase, GcoA,
demonstrated that it demethylates diverse aromatic compounds,
including guaiacol (which can be derived from coniferyl alcohol and
represents the aromatic functionality of G-lignin that must undergo
demethylation), guaethol, anisole (methoxybenzene), 2-methylanisole,
and 3-methoxycatechol (3MC) (28) with similar or greater effi-
ciency than other O-aryl-demethylases described in the literature
(22, 24, 29). However, GcoAB shows poor reactivity toward syrin-
gol, which can be derived from sinapyl alcohol via high-temperature
reactions and represents the aromatic functionality of S-lignin that
must undergo O-demethylation for further catabolism via ring-
opening dioxygenases. G- and S-lignin are the major components
of lignin in hardwoods and grasses (1). Due to their abundance, it is
important to find enzymes that can act on the methoxy groups of
both G- and S-lignin subunits. To date, there are no reports de-
scribing syringol O-demethylation or, more broadly, even its ca-
tabolism by microbes. Rather, the best-studied biological reaction
of syringol is its 4–4 dimerization to form cerulignone (30–33).
Although our previous work showed that GcoA was not ef-
fective for syringol O-demethylation, crystallographic studies and
molecular dynamics (MD) simulations indicated that a triad of
active site phenylalanine residues is both highly mobile and im-
portant for positioning the substrate in its catalytically competent
pose. In the present study, we hypothesized that substitution of
GcoA-F169, which has the closest interaction with the bound
substrate, may relax the specificity of the enzyme sufficiently to
permit the O-demethylation of S-lignin type substrates. We
tested that hypothesis using biochemical, structural, computa-
tional, and in vivo approaches. We demonstrate highly efficient
in vitro and in vivo syringol turnover through structure-guided
protein engineering, where the enzyme also retains highly effi-
cient activity toward guaiacol.
Results
The Syringol Binding Mode Can Be Modulated by Active Site
Engineering. Guaiacol assumes a productive orientation in the
active site of GcoA, resulting in a shift in the spin state of the
heme iron from low (S = 1/2) to high (S = 5/2), due to the actions
of amino acid side chains that create a tight-fitting hydrophobic
pocket. The closest contact is with GcoA-F169, which forms a
hydrophobic interaction with the C6 carbon on the aromatic ring
of guaiacol. Previous MD simulations suggest that this residue is
highly mobile, predicting that the productive complex forms dy-
namically (28). Superposition of the cocrystal structures of GcoA
with guaiacol and syringol reveals a shift in the positions of
GcoA-F169 and the reactive syringol methoxy group relative to
the heme (see below). Functionally, these shifts in the GcoA-F169
position permit binding of syringol, although binding in the productive
conformation as measured by the shift in Fe(III) spin state is sub-
stantially diminished relative to binding of guaiacol (Table 1) (28).
We hypothesized that mutation of GcoA-F169 to a smaller
residue (alanine) may relieve the apparent steric clash between it
and the bound ligand in the active site, allowing syringol to adopt a
productive conformation. Thus, we prepared the GcoA-F169A
variant and measured its guaiacol and syringol binding properties
(SI Appendix, Fig. S1). Both ligands were able to stimulate the
Fe(III) spin state conversion at levels close to wild-type (WT)
(Table 1), with KD values in the low micromolar range. Notably,
3MC also bound and induced a spin state change in Fe(III), though
with less affinity than guaiacol or syringol. We concluded that active
site engineering could indeed lead to productive syringol binding
and potentially to turnover by GcoA-F169 variants, and thus sub-
sequently studied their reactivity with guaiacol, syringol, and 3MC.
GcoA-F169A Efficiently Demethylates both Guaiacol and Syringol with
only Limited Uncoupling. Substrate analogs are known to stimulate
the P450 reaction with NADH/O2 without concomitant oxygen-
ation of the organic substrate. This leads to uncoupling of the
NADH and substrate oxidation reactions and to reduction of O2
to either H2O2 or H2O (34). Previous work has shown that
syringol stimulates NADH consumption by WT GcoA (28), al-
though without substantial syringol turnover. To address whether
guaiacol and/or syringol would serve as substrates of GcoA-
F169A, we monitored the disappearance of NADH (UV/vis) and
aromatic substrate (HPLC) over time (Scheme 1). The rates
of organic substrate and NADH consumption were robust and
similar within error regardless of whether guaiacol or syringol
Table 1. Efficacy of GcoA-F169A relative to WT GcoA in binding and demethylating guaiacol
and syringol
WT GcoA
GcoA-F169A
Syringol
Parameter
Guaiacol
Syringol
3MC
Guaiacol
7.1 0.1
3MC
KD, μM*
0.0065 0.002 2.8 0.5 3.7 0.1
1.7 0.07 9.5 0.2
%Fe(III) spin state conversion*
87
1
56
2
62 0.01 72 0.3
76 0.7
11 0.03 5.9 0.01
40 10
290 40 600 90
65 0.6
n/a
n/a
k
cat, s−1†
6.8 0.02
60 10
110 20
—
—
—
n/a‡
n/a
n/a
K
M, mM†
cat/KM, mM−1s−1†
6
1
k
n/a
*KD was measured by titrating 0–60 μM of substrate into a solution containing 2–6 μM WT or F169A GcoA in air-
saturated buffer (25 mM Hepes, 50 mM NaCl, pH 7.5, 25 °C) and recording the ferric spin state change from the low-
spin (417 nm) to high-spin (388 nm) species. The % spin state conversion was calculated by dividing the final high-spin
species (maximum absorbance at 388 nm) by the starting low-spin species (maximum absorbance at 417 nm).
†The Michaelis constants are apparent, as the dioxygen and GcoB concentrations are not known to be saturating.
The conditions used were 0.2 μM GcoAB, 100 μg/mL catalase, 300 μM NADH, 210 μM O2, and 0–300 μM substrate,
25 °C, 25 mM Hepes, 50 mM NaCl, pH 7.5.
‡n/a and dashes: Michaelis constants for 3MC could not be directly measured because of a high level of NADH
uncoupling, indicated by “n/a”; see text. A substantial amount of syringol turnover was not observed for WT
GcoA, again making it impossible to measure Michaelis parameters. This is indicated by dashes.
2 of 7
|
www.pnas.org/cgi/doi/10.1073/pnas.1820001116
Machovina et al.

A broader survey of variants at the GcoA-F169 position
(amino acids S, H, V, I, and L in addition to A) confirmed
that GcoA-F169A exhibits the best catalytic performance in
terms of both specific activity and reaction coupling, although
other small side chains (S and V) also permitted reactivity with
syringol, suggesting that these may permit syringol to assume a
reactive conformation at the heme. Apparent steady-state kinetic
parameters measured in air and at potentially subsaturating
concentrations of GcoB (Table 1 and SI Appendix, Fig. S2 and
Table S2) suggest that GcoA-F169A is a more effective catalyst
toward the first methoxy group of syringol relative to guaiacol,
with kcat/KM[syringol] nearly double kcat/KM[guaiacol]. More-
over, GcoA-F169A has a slightly improved performance with
guaiacol as a substrate relative to the WT enzyme.
Scheme 1. O-demethylation of guaiacol to form catechol and formalde-
hyde (A) and syringol to form pyrogallol and two formaldehydes (B). The
singly demethylated species, 3MC, is expected to form as an intermediate in
reaction (B). See Fig. 1.
was used (SI Appendix, Table S1). This suggests that both
guaiacol and syringol serve as substrates for GcoA-F169A.
Moreover, the oxidative O-demethylation of guaiacol appeared
to be largely coupled to NADH consumption. When NADH and
O2 were present in excess of guaiacol, the measured stoichiometry
of the GcoA-F169A–catalyzed reaction was very close to 1 mole-
cule each of guaiacol and NADH consumed to 1 formaldehyde
and 1 catechol produced (Fig. 1), without overconsumption of
NADH (103 7% coupling efficiency; SI Appendix, Table S1).
Since both methoxy groups of syringol can potentially serve as
substrates, we examined syringol turnover in several ways.
Syringol (100 μM) was first incubated with NADH (200 μM) and
excess dissolved O2 (210 μM), and the reaction with the GcoA-
F169A mutant was allowed to go to completion. As with the
guaiacol reaction, all the NADH and syringol were consumed
(Fig. 1), implying that syringol undergoes two O-demethylations,
producing 3MC and then pyrogallol. However, less formalde-
Structural Analysis Reveals Productive Syringol Reorientation in
GcoA-F169 Variants. Superposition of the structures of GcoA-
ligand complexes indicated a significant rotation and translation
of bound syringol relative to guaiacol (Fig. 2A), and we hy-
pothesized that this could form the basis for the unproductive
syringol uncoupling in the native enzyme. The comparative dis-
tances between the heme and the proximal methoxy carbon of
guaiacol vs. syringol are within 0.4 Å, and even closer between
the heme and methoxy oxygens (within 0.1 Å). In addition, there
is no significant deviation from the plane of the aromatic rings
between these ligands. In contrast, the angle of presentation of
the methoxy to the heme diverges significantly in these com-
plexes. Using the angle between the methoxy oxygen, heme iron,
and terminal methoxy carbon atoms [O-Fe(III)-C] as a conve-
nient readout of relative orientation, there is a 55% increase in
angle from the guaiacol-bound structure (8.3°) compared with
the syringol-bound structure (12.9°).
hyde (170
10 μM) was produced than expected, suggesting
some uncoupling of NADH/O2 consumption from the oxidative
O-demethylation. Consistent with that hypothesis, 50 4 μM of
the singly demethylated intermediate 3MC was observed at the
end of the reaction, even though sufficient NADH/O2 were
present to enable its complete conversion to pyrogallol. Notably,
pyrogallol was not detected under any of the conditions used
here, possibly due to its well-known instability in the presence of
O2 (35).
The stoichiometric analysis was next repeated with NADH
and syringol present in equal concentrations (∼200 μM each;
210 μM O2), conditions expected to permit at most one-half of
the available methoxy groups to react. All of the NADH and
150 6 μM syringol were consumed, and 200 3 μM formal-
dehyde and 120
2 μM 3MC were generated (Fig. 1 and SI
Appendix, Table S2). The accumulation of roughly one-half an
equivalent of 3MC (relative to NADH) under these conditions
suggested that the first O-demethylation of syringol must be
faster than the second, and that the uncoupling reaction is likely
stimulated by 3MC rather than by syringol. Consistent with those
expectations, the rate of 3MC disappearance measured by HPLC
was significantly slower than the disappearance of either guaiacol
or syringol (2.6 0.3 μM 3MC s−1 μmol GcoA-F169A−1 vs. 5.1
0.8 μM syringol s−1 μmol GcoA-F169A−1; SI Appendix, Table S1).
Moreover, the faster consumption of NADH relative to 3MC
suggested diminished reaction coupling (64 10% coupling; SI
Appendix, Tables S1 and S2). In reactions containing 100 μM
3MC and 200 μM NADH, the majority of the initially available
NADH was consumed, and ∼100 μM of formaldehyde was
produced (Fig. 1). We hypothesized that the observed over-
consumption of NADH was due to the uncoupled reaction,
leading to H2O2 production. The production of H2O2 in the
presence of excess NADH/O2 and limiting 3MC was confirmed
using Amplex Red and horseradish peroxidase (HRP) (SI Ap-
pendix, Table S3). Consequently, accurate values for kcat and
KM could not be measured using 3MC as a substrate (Table 1).
Fig. 1. Quantitative analyses of substrate consumption and product gen-
eration indicate nearly complete coupling of NADH/O2 consumption to
substrate O-demethylation for guaiacol and progressively more uncoupling
for syringol and 3MC. NADH (200 μM) and guaiacol, syringol, or 3MC (100 or
200 μM) were incubated in air with 0.2 μM GcoA-F169A and GcoB (each with
25 mM Hepes, 50 mM NaCl, pH 7.5 at 25 °C and 210 μM O2). Reactants and
products were quantified when the UV/vis spectrum ceased changing and
the reaction was deemed complete. The total NADH consumed is compared
with the amounts of formaldehyde and demethylated aromatic compound
produced. Pyrogallol, the O-demethylated product of 3MC, is unstable in air
under the conditions used in the assay and was not detected. Error bars
represent 1 SD from three or more independent measurements. P values
comparing NADH consumption and formaldehyde production were
0.035 for guaiacol, 0.20 for 100 μM syringol, 0.41 for 200 μM syringol, and
0.0035 for 3MC. For NADH consumption and aromatic product production,
these P values were 0.011 for guaiacol, 0.00031 for 100 μM syringol, and
0.0084 for 200 μM syringol.
Machovina et al.
PNAS Latest Articles
|
3 of 7

Fig. 2. Structure-guided active site engineering of GcoA. Superpositions of WT and GcoA-F169A ligand-bound structures of GcoA, the P450 monooxygenase
component of GcoAB, are shown. The heme is colored in bronze stick. (A) The guaiacol (green) and syringol (pink) complexes with WT GcoA are shown with
the position of the GcoA-F169 residue highlighted. The translation and rotation of syringol compared with guaiacol result in a shift of the target methoxy
carbon away from the heme. The Fe(III) to guaiacol methoxy carbon distance is 3.9 Å, and the Fe(III) to proximal syringol methoxy carbon distance is 4.3 Å.
Data from PDB ID: 5NCB (28) and 5OMU (28). (B) The engineered GcoA-F169A-syringol complex (blue) enables positioning of the reactive methoxy group
relative to the heme in a mode consistent with productive guaiacol binding. GcoA-F169 from the guaiacol-bound WT structure is shown in green lines. (C)
Superposition of the WT (green) and GcoA-F169A (yellow) guaiacol-bound complexes reveals that guaiacol sits in an identical position in both crystal
structures. GcoA-F169 from the guaiacol-bound WT structure is shown in green lines.
To investigate this further, we generated multiple high-
resolution GcoA-F169 variant cocrystal structures. A set of
syringol-bound structures (SI Appendix, Table S4 and Fig. S4)
provided direct insight into the minimal reduction in side chain
bulk required to achieve the productive binding mode equivalent
to that of guaiacol. A stepwise trajectory of the bound syringol
toward this optimum orientation with decreasing side chain bulk
was observed in the superposition of four cocrystal structures (SI
Appendix, Fig. S3A). Specifically, GcoA-F169H creates an im-
proved substrate orientation, further improved by GcoA-F169V,
and essentially optimized in both the GcoA-F169S and -F169A
proteins (SI Appendix, Fig. S3A). Indeed, a comparison of the
GcoA-F169A syringol structure with the WT guaiacol structure
revealed an almost perfect alignment relative to the aromatic
rings of each substrate and a methoxy O-Fe-C angle of 8.6°,
within 0.3° of that observed for guaiacol (Fig. 2B).
Each protein variant also crystallized successfully with guaiacol
(SI Appendix, Table S5 and Fig. S5) and the structures
showed that the orientation of the bound ligand remained con-
sistent with that of the WT enzyme (SI Appendix, Fig. S3B). Even
the largest reduction in side chain bulk, represented by the
GcoA-F169A variant, retained the ideal reactive geometry for
the natural substrate (Fig. 1C). Furthermore, comparison of the
surrounding active site architecture confirmed no significant
deviation from the WT. The resolution of these structures (1.66–
2.17 Å) also provided sufficient electron density quality to ex-
plore changes in the hydration of the pocket (SI Appendix, Fig.
S6). While the native enzyme excluded water from the active site
pocket, we were interested to see whether this was maintained
when a new cavity in the pocket was introduced. The syringol-
bound mutants A, S, and V contain an additional ordered water
in the active site (SI Appendix, Fig. S6), which may help maintain
the substrate in a productive binding pose for catalysis. As
expected, the bulkier GcoA-F169H mutant excluded water from
the active site, as with the WT structure.
rather than guaiacol is bound at the WT active site (Fig. 3 A, C,
and E, SI Appendix, Fig. S8, and Movies S1 and S2). This effect is
complemented by the static picture given by the crystal structures
(Fig. 2A), showing that GcoA-F169 is “pushed away” from the
substrate by a distance commensurate with the observed shift of
the substrate.
Opening of the substrate access loop, a larger-scale phenom-
enon closely related to the movement of GcoA-F169, can also
relieve active site crowding. All GcoA crystal structures reported
to date present a closed active site “lid,” but previous MD sim-
ulations demonstrated the ability of the F/G helices (and their
connecting loop) to move away from the active site, thus ex-
posing the active site to solution, particularly in the apo form
(28). Crystal packing may hinder lid opening in the GcoA-F169A
structure; thus, the first two effects (shifting of substrate and
GcoA-F169) are more pronounced in crystal structures. How-
ever, in MD of GcoA in solution, the active site loop is un-
constrained, and the effect of the GcoA-F169 clash with syringol
is observed less at the substrate and more on the enzyme. This
includes the positioning and flexibility of GcoA-F169 (as men-
tioned above) and the substrate access lid (Fig. 3 B, D, and F; SI
Appendix, Fig. S9; and Movies S3 and S4), which is significantly
more prone to open with syringol bound in WT GcoA than with
guaiacol, as well as either substrate bound in MD simulations on
the GcoA-F169A mutant. Open and closed access loops have
been observed in P450 enzyme [P450cam (34) and BM3 (35, 36)]
crystal structures. The open GcoA configurations that we observe
in MD simulations are approximately one-half as open as the
aforementioned open crystal structures and possibly not suffi-
ciently open to allow substrate ingress and egress. However, this
principle observed over 80 ns is likely to be more pronounced over
the course of the full catalytic cycle. We also note that to date,
efforts to crystallize GcoA in the apo state have proven un-
successful; when achieved, this may reveal a more open configu-
ration of the substrate access lid.
Syringol Clashes with both GcoA-F169 and the Substrate Access Lid in
Simulations of WT GcoA. In the WT enzyme with syringol bound,
active site crowding can be relieved in several ways. First, as
already noted, syringol can shift toward the heme (Fig. 2A); this
effect is seen in MD simulations (80 ns) carried out on WT
GcoA with either bound guaiacol or syringol, although the effect
is much subtler than in the crystal structures (SI Appendix, Fig.
S7). A second effect is more pronounced in MD simulations,
which show that GcoA-F169 is significantly more flexible and
perturbed from the crystal structure position when syringol
The above conclusions from 80-ns MD simulations are also
supported by a deeper analysis of three independent 1-μs MD
trajectories of WT GcoA with syringol and guaiacol bound at the
active site, which were originally presented in our previous study
(figure S21 in ref. 28). GcoA-F169 is significantly perturbed from
its crystal structure position and more mobile; this coincides with
an increased propensity to open the substrate access lid, which is
the only region of significant difference in flexibility (SI Appen-
dix, Fig. S10).
4 of 7
|
www.pnas.org/cgi/doi/10.1073/pnas.1820001116
Machovina et al.

conserved positions in the entire protein (Fig. 4 and SI Appendix,
Fig. S14 and Table S7). From a multiple sequence alignment, we
determined that alanine and phenylalanine are the most fre-
quent residues used by CYP255A enzymes at position 169, with
alanine present in the majority of sequences. Thus, the GcoA-
F169A mutant, which showed enhanced turnover on guaiacol
and syringol, is closer to the CYP255A consensus protein than
the WT. It is interesting that although none of the GcoA ho-
mologs in our analyses exhibits a histidine at position 169, the
GcoA-F169H mutant was stable and showed the highest specific
activity on guaiacol. Next to GcoA-F169, A295 and T296 show
the greatest variability of residues in the active site. Besides
these, other residues within 6 Å of the center of mass of the
guaiacol substrate generally show high conservation.
F169A Enables In Vivo Syringol Conversion by GcoA. Finally, we
aimed to demonstrate in vivo conversion of syringol to pyrogallol
using Pseudomonas putida KT2440, which has many native
aromatic-catabolic pathways relevant to lignin conversion (8, 30,
37, 38). To accomplish this, we transformed plasmids expressing
WT GcoA or the GcoA-F169A variant and GcoB into a strain
that constitutively overexpresses PcaHG (SI Appendix, Tables
S8 and S9), a native 3,4-protocatechuate dioxygenase from P.
putida that converts pyrogallol into 2-pyrone-6-carboxylate, a
more stable product (Fig. 5A) (39). 1H NMR analysis of the
culture medium revealed that, when cultured with 20 mM glu-
cose and 1 mM syringol, peaks corresponding to syringol com-
pletely disappeared in cells expressing GcoA-F169A (AM157)
after 6 h, which coincided with the appearance of 3MC, pyro-
gallol, and 2-pyrone 6-carboxylate (Fig. 5B and SI Appendix, Fig.
S15). A standard for 2-pyrone 6-carboxylate was not available;
however, we did verify the presence of 2-pyrone 6-carboxylate
using LC-MS-MS (SI Appendix, Fig. S15). While small amounts
of pyrogallol, 3MC, and 2-pyrone 6-carboxylate were also ob-
served with the WT GcoAB (AM156), nearly 60% of the original
syringol remained after 6 h. Neither pyrogallol nor 3MC was
observed in the strain lacking GcoAB (AM155). These data in-
dicate that the F169A variation enhances in vivo O-demethylation
of syringol to 3MC and pyrogallol by GcoA.
Fig. 3. GcoA-F169 in WT GcoA and the substrate access loop are signifi-
cantly displaced with bound syringol. MD simulations with bound guaiacol
indicate that GcoA-F169 (A) and the substrate access lid (B) are relatively
stable (Movies S1 and S3). Introducing syringol results in increased flexibility
of GcoA-F169 (C) and the substrate access loop (D) (Movies S2 and S4). In A–D,
the position of each of the labeled Phe side chains (or, alternatively, the
substrate access loop) is shown every 4 ns over the course of the 80-ns MD
simulation. Substrate, the Phe side chains, and heme are shown in sticks, and
the Fe atom and the O atom of a reactive heme-oxo intermediate are shown
as spheres. Probability distributions are shown for the rmsd of the six ring
carbons of GcoA-F169 from their crystal structure positions (E) and the re-
action coordinates for opening/closing of the substrate access loop (as de-
fined in SI Appendix; lower values indicate more open configurations) (F).
Discussion and Conclusions
The creation of enzymes that overcome the challenge of lignin
heterogeneity through increased substrate promiscuity is an at-
tractive goal, but this might come at the cost of reduced activity
toward the natural substrate. Unexpectedly, we found that
GcoA-F169A not only binds both guaiacol and syringol in a
We also performed density functional theory (DFT) calcula-
tions on a truncated active site model, demonstrating that the O-
demethylation of syringol proceeds via a similar pathway as
previously described for guaiacol (28) (SI Appendix, Fig. S11 and
Table S6). Optimized transition state geometries and free energy
barriers for the rate-limiting hydrogen atom transfer are likewise
very similar in the two cases. In addition, replica exchange ther-
modynamic integration (RETI) simulations were conducted to
examine relative free energies associated with substrate binding
and mutating GcoA-F169 in the closed state of the enzyme (SI
Appendix, Figs. S12 and S13). These RETI simulations revealed
additional substrate binding modes, made possible by the “softer”
interactions between the substrate and enzyme as the simulations
gradually “turn on” and “turn off” electrostatic and van der Waals
components of intermolecular interactions, and quantify the effect
of this binding flexibility on binding thermodynamics.
Sequence Position 169 in CYP255A Enzymes Is Highly Variable. GcoA
belongs to the CYP255A family of cytochrome P450 enzymes
(27). Conservation analyses of GcoA homologs revealed a no-
tably variable position 169 among active site residues. Moreover,
GcoA-F169 not only is the least conserved of the triad of phe-
nylalanine residues in the active site, it is also among the least
Fig. 4. Bioinformatic analysis of CYP255A sequences indicates variability in
the 169th sequence position. Conservation of residues within 6 Å of guaiacol
in GcoA, determined via an analysis of protein sequences from 482 GcoA
homologs. (Details provided in SI Appendix.) Conservation scores are
reported as percentiles. F169 is less conserved than 73% of the positions in
GcoA. Data from PDB ID: 5NCB (28).
Machovina et al.
PNAS Latest Articles
|
5 of 7

substituted compounds (e.g., vanillin, syringaldehyde). Indeed, in
previous work, we showed that T296 sterically clashes with the
C4 position of vanillin, preventing O-demethylation (28). In-
terestingly, the 296 position is also quite variable according to
the bioinformatics analysis (Fig. 4). The results described here
suggest that we may use a similar structure-guided approach to
investigate the activity of several T296 variants on 4-substituted
compounds. As a large number of lignin degradation products
contain para-substituted R groups that are bulkier than hydrogen,
creating an engineered bacterium in which a minimal number of
genetic insertions lead to maximal lignin bioconversion is of keen
interest for future work. More broadly, the evolutionary trajectory,
substrate specificity, and catalytic efficiency of the CYP255A
family of aromatic O-demethylases will be the subject of future
work aimed at elucidating the relevance of this cytochrome P450
family for microbial lignin conversion.
A
B
Methods
Fig. 5. GcoA-F169A converts syringol in vivo. (A) A pathway for in vivo
syringol O-demethylation to pyrogallol and cleavage to 2-pyrone 6-
carboxylate is proposed. (B) After 6 h, strains were analyzed for their abil-
ity to turn over syringol via 1H NMR spectroscopy. Syringol (green) is com-
pletely converted to pyrogallol (pink), 3MC (blue), or 2-pyrone 6-carboxylate
(purple) in AM157. The WT GcoA enzyme in AM156 shows only small
amounts of conversion. AM155, which does not express GcoA, shows no
conversion.
Protein Expression and Purification. Mutagenesis was performed using pri-
mers listed in SI Appendix, with the Q5 polymerase and KLD enzyme mix
(New England BioLabs) according to the manufacturer’s protocol. Proteins
were expressed as described previously (28).
Crystallization and Structure Determination. Crystallization, diffraction exper-
iments, and structure solution were carried out as described previously (28).
Biochemical Characterization.
Heme quantification of GcoA-F169 mutants. Catalytically active heme bound to
each GcoA mutant was determined as reported previously (28, 40) and de-
scribed in detail in SI Appendix.
Determination of [FAD] and nonheme [Fe] in GcoB. The FAD and 2Fe-2S contents
of GcoB were measured as reported previously (28, 41) and described in
detail in SI Appendix.
Steady-state kinetics of GcoA-F169A. The O-demethylation reactions of guaiacol,
syringol, and 3MC were continuously monitored using the NADH consump-
tion assay as reported previously (28) and described in detail in SI Appendix.
Determination of substrate dissociation constants (KD) with GcoA-F169A. The
equilibrium binding constant, KD, for GcoA-F169A and guaiacol, syringol,
and 3MC was determined as described previously (28) and detailed in
SI Appendix.
Formaldehyde determination. The [formaldehyde] produced on reaction with
GcoA-F169A and substrates was determined using a colorimetric assay with
tryptophan (28, 42); details are provided in SI Appendix.
HPLC for product identification and specific activity measurement. HPLC was used
to verify the O-demethylated product of GcoA-F169A GcoA/GcoB with
guaiacol, syringol, or 3MC. In addition, discontinuous HPLC was used to
determine the specific activity of aromatic product disappearance. Details
are provided in SI Appendix.
Detection of H2O2 via HRP and Amplex Red assay. A colorimetric assay involving
HRP and Amplex Red was used to quantify H2O2 in the reaction between
GcoA-F169A GcoA/GcoB, NADH and guaiacol, syringol, or 3MC, as described
in SI Appendix.
productive orientation analogous to guaiacol in WT GcoA (Fig.
2), but also is more catalytically efficient for O-demethylation
of both guaiacol and syringol relative to WT, where O-
demethylation is well coupled to NADH oxidation (Fig. 1 and
Table 1). Along with this biochemical observation, the bio-
informatics analysis shows that alanine is the most prevalent
residue in the 169th sequence position in the CYP255A family.
Considering these findings together, it is surprising that the WT
GcoA does not possess an alanine at position 169 if guaiacol is
the primary substrate, as was assumed in the original reports of
GcoAB (17, 27, 28). Given that the GcoA-F169A mutation re-
sults in improved turnover of guaiacol, we speculate that either
guaiacol is not the primary substrate or there has been little
evolutionary pressure in Amycolatopsis sp. ATCC 39116 for im-
proved turnover of guaiacol. Another potential explanation
could be that syringol O-demethylation and subsequent ring
cleavage lead to dead-end products that cannot be catabolized
and may be toxic to the microbe; thus, GcoA-F169 could func-
tion to prevent natural syringol catabolism.
As a first step toward enabling syringol catabolism, in vivo
experiments validated the in vitro studies by illustrating efficient
O-demethylation of syringol and 3MC by the GcoA-F169 mutant.
While 2-pyrone 6-carboxylate was detected, pyrogallol is a poor
substrate of PcaHG, as most of the intermediate is lost to oxida-
tion (SI Appendix, Fig. S15), and whether 2-pyrone-6-carboxylate
can be catabolized further is unclear. Future work could focus on
identifying or developing a dioxygenase capable of cleaving py-
rogallol to a product that could be further metabolized. Coupled
with GcoAB, the meta-cleavage pathway of P. putida mt-2 might
enable complete assimilation of syringol if pyrogallol can be effi-
ciently cleaved to 2-hydroxymuconate.
MD, DFT, and Bioinformatics. MD simulations and DFT calculations were
performed following a similar methodology as in our previous work (28). Full
details of the computational methods and references are provided in SI
Appendix. A total of 482 homologous CYP255A sequences were retrieved
from a blastp search against GcoA. After multiple sequence alignment,
conservation was analyzed from relative entropy calculations for each site.
Further details are provided in SI Appendix.
In Vivo Syringol Utilization. Strains used for shake flask experiments were
grown overnight in LB media and resuspended the next day in M9 minimal
media with 20 mM glucose, as described in SI Appendix. Cells were grown
until they reached an OD600 of approximately 1, at which point syringol was
added at a final concentration of 1 mM. 1H NMR spectroscopy was used to
analyze syringol consumption.
Cytochrome P450 systems represent one of the most versatile
classes of enzymes, making them an ideal target for engineering
enhanced activity and substrate promiscuity. Our system is a
prime example. The mutation of a single residue resulted in ef-
ficient turnover of an S-lignin substrate, syringol, in addition to
the native G-subunit substrate, guaiacol, which is not efficiently
achieved in the WT enzyme. The plasticity of the GcoA active
site may be amenable to even more modifications, allowing us to
encompass other lignin monomers as substrates, such as four-
Data Deposition. The atomic coordinates and structure factors have been
deposited in the Protein Data Bank, https://www.wwpdb.org (PDB ID codes
6HQK, 6HQL, 6HQM, 6HQN, 6HQO, 6HQP, 6HQQ, 6HQR, 6HQS, and 6HQT).
6 of 7
|
www.pnas.org/cgi/doi/10.1073/pnas.1820001116
Machovina et al.

ACKNOWLEDGMENTS. We thank Josh Vermaas for his assistance with the
simulation setup and Mark Allen and the staff at Diamond Light Source for
their support and contributions to X-ray crystallography. We also thank
Brenna Black and Bill Michener for their assistance with analytics. This
material is also based on work supported by (while C.M.P. is serving at) the
National Science Foundation (NSF). Any opinions, findings, and conclusions
or recommendations expressed in this material are those of the authors and
do not necessarily reflect the views of the NSF. Support for this work was
provided by the NSF (Grants MCB-1715176, M.M.M. and J.L.D.; CHE-1361104,
to K.N.H.; and CBET-1552355, to C.M.P.) and the Biotechnology and Bi-
ological Sciences Research Council (Grants BB/P011918/1 and BB/L001926/1,
to J.E.M.). This work was partially authored by the Alliance for Sustainable
Energy, LLC, the manager and operator of the National Renewable Energy
Laboratory for the US Department of Energy (DOE) (Contract DE-AC36-
08GO28308). B.C.K., A.W.M., L.B., G.P.S., M.F.C., C.W.J., and G.T.B. received
funding for the MD simulations and in vivo experiments from the US DOE,
Office of Energy Efficiency and Renewable Energy (EERE), Bioenergy Tech-
nologies Office. M.M.M. received funding support from the US DOE, Office
of Science, Office of Workforce Development for Teachers and Scientists,
Office of Science Graduate Student Research program, which is administered
by the Oak Ridge Institute for Science and Education for the DOE under
Contract DE-SC0014664. G.T.B. and J.E.M. received funding for the structural
biology efforts from The Center for Bioenergy Innovation, a US DOE Bioen-
ergy Research Center supported by the Office of Biological and Environmen-
tal Research in the DOE Office of Science. M.G.-B is the recipient of a Ramón
Areces Foundation postdoctoral fellowship. Computer time was provided by
Extreme Science and Engineering Discovery Environment, supported by NSF
Grant ACI-1548562 via allocation MCB-090159 to G.T.B. at the Pittsburgh
Supercomputing Center (Bridges) and the Texas Advanced Computing Center
(Stampede2), by the National Renewable Energy Laboratory Computa-
tional Sciences Center supported by the DOE Office of EERE under Con-
tract DE-AC36-08GO28308, and by the UCLA Institute for Digital Research
and Education. K.N.H. received financial support from the National Insti-
tute for General Medical Sciences (Grant GM-124480).
1. W. Boerjan, J. Ralph, M. Baucher, Lignin biosynthesis. Annu. Rev. Plant Biol. 54, 519–
546 (2003).
2. G. Fuchs, M. Boll, J. Heider, Microbial degradation of aromatic compounds—from one
strategy to four. Nat. Rev. Microbiol. 9, 803–816 (2011).
3. T. D. Bugg, M. Ahmad, E. M. Hardiman, R. Singh, The emerging role for bacteria in
lignin degradation and bio-product formation. Curr. Opin. Biotechnol. 22, 394–400
(2011).
4. D. Floudas et al., The Paleozoic origin of enzymatic lignin decomposition re-
constructed from 31 fungal genomes. Science 336, 1715–1719 (2012).
5. D. Salvachúa et al., Lignin depolymerization by fungal secretomes and a microbial
sink. Green Chem. 18, 6046–6062 (2016).
22. T. Yoshikata et al., Three-component O-demethylase system essential for catabolism
of a lignin-derived biphenyl compound in Sphingobium sp. strain SYK-6. Appl. Envi-
ron. Microbiol. 80, 7142–7153 (2014).
23. A. Harada et al., The crystal structure of a new O-demethylase from Sphingobium sp.
strain SYK-6. FEBS J. 284, 1855–1867 (2017).
24. A. C. Kohler, M. J. L. Mills, P. D. Adams, B. A. Simmons, K. L. Sale, Structure of aryl
O-demethylase offers molecular insight into a catalytic tyrosine-dependent mechanism.
Proc. Natl. Acad. Sci. U.S.A. 114, E3205–E3214 (2017).
25. S. G. Bell et al., The crystal structures of 4-methoxybenzoate bound CYP199A2 and
CYP199A4: Structural changes on substrate binding and the identification of an anion
binding site. Dalton Trans. 41, 8703–8714 (2012).
26. S. G. Bell et al., Investigation of the substrate range of CYP199A4: Modification of the
partition between hydroxylation and desaturation activities by substrate and protein
engineering. Chemistry 18, 16677–16688 (2012).
6. E. Masai, Y. Katayama, M. Fukuda, Genetic and biochemical investigations on bac-
terial catabolic pathways for lignin-derived aromatic compounds. Biosci. Biotechnol.
Biochem. 71, 1–15 (2007).
27. M. Tumen-Velasquez et al., Accelerating pathway evolution by increasing the gene
dosage of chromosomal segments. Proc. Natl. Acad. Sci. U.S.A. 115, 7105–7110 (2018).
28. S. J. B. Mallinson et al., A promiscuous cytochrome P450 aromatic O-demethylase for
lignin bioconversion. Nat. Commun. 9, 2487 (2018).
29. S. C. Farrow, P. J. Facchini, Dioxygenases catalyze O-demethylation and O,O-
demethylenation with widespread roles in benzylisoquinoline alkaloid metabolism
in opium poppy. J. Biol. Chem. 288, 28997–29012 (2013).
30. D. Salvachúa, E. M. Karp, C. T. Nimlos, D. R. Vardon, G. T. Beckham, Towards lignin
consolidated bioprocessing: Simultaneous lignin depolymerization and product
generation by bacteria. Green Chem. 17, 4951–4967 (2015).
31. S. M. Bölicke, W. Ternes, Isolation and identification of oxidation products of syringol
from brines and heated meat matrix. Meat Sci. 118, 108–116 (2016).
32. O. E. Adelakun, T. Kudanga, I. R. Green, M. le Roes-Hill, S. G. Burton, Enzymatic
modification of 2,6-dimethoxyphenol for the synthesis of dimers with high antioxi-
dant capacity. Process Biochem. 47, 1926–1932 (2012).
33. Y. Y. Wan, Y. M. Du, T. S. Miyakoshi, Enzymatic catalysis of 2,6-dimethoxyphenol by
laccases and products characterization in organic solutions. Sci. China B Chem. 51,
669–676 (2008).
34. Y.-T. Lee, R. F. Wilson, I. Rupniewski, D. B. Goodin, P450cam visits an open confor-
mation in the absence of substrate. Biochemistry 49, 3412–3419 (2010).
35. I. F. Sevrioukova, H. Li, H. Zhang, J. A. Peterson, T. L. Poulos, Structure of a cytochrome
P450-redox partner electron-transfer complex. Proc. Natl. Acad. Sci. U.S.A. 96, 1863–
1868 (1999).
36. D. C. Haines, D. R. Tomchick, M. Machius, J. A. Peterson, Pivotal role of water in the
mechanism of P450BM-3. Biochemistry 40, 13456–13465 (2001).
37. D. R. Vardon et al., Adipic acid production from lignin. Energy Environ. Sci. 8, 617–628
(2015).
7. G. T. Beckham, C. W. Johnson, E. M. Karp, D. Salvachúa, D. R. Vardon, Opportunities
and challenges in biological lignin valorization. Curr. Opin. Biotechnol. 42, 40–53
(2016).
8. J. G. Linger et al., Lignin valorization through integrated biological funneling and
chemical catalysis. Proc. Natl. Acad. Sci. U.S.A. 111, 12013–12018 (2014).
9. T. D. Bugg, R. Rahmanpour, Enzymatic conversion of lignin into renewable chemicals.
Curr. Opin. Chem. Biol. 29, 10–17 (2015).
10. O. Y. Abdelaziz et al., Biological valorization of low molecular weight lignin. Bio-
technol. Adv. 34, 1318–1346 (2016).
11. R. Rinaldi et al., Paving the way for lignin valorisation: Recent advances in bio-
engineering, biorefining and catalysis. Angew. Chem. Int. Ed. Engl. 55, 8164–8215
(2016).
12. W. Schutyser et al., Chemicals from lignin: An interplay of lignocellulose fraction-
ation, depolymerisation, and upgrading. Chem. Soc. Rev. 47, 852–908 (2018).
13. A. J. Ragauskas et al., Lignin valorization: Improving lignin processing in the bio-
refinery. Science 344, 1246843 (2014).
14. Z. Sun, B. Fridrich, A. de Santi, S. Elangovan, K. Barta, Bright side of lignin de-
polymerization: Toward new platform chemicals. Chem. Rev. 118, 614–678 (2018).
15. A. Dardas et al., The demethylation of guaiacol by a new bacterial cytochrome P-450.
Arch. Biochem. Biophys. 236, 585–592 (1985).
16. U. Karlson et al., Two independently regulated cytochromes P-450 in a Rhodococcus
rhodochrous strain that degrades 2-ethoxyphenol and 4-methoxybenzoate. J. Bacteriol.
175, 1467–1474 (1993).
17. L. D. Eltis, U. Karlson, K.N. Timmis, Purification and characterization of cytochrome
P450RR1 from Rhodococcus rhodochrous. Eur. J. Biochem. 213, 211–216 (1993).
18. A. Segura, P. V. Bünz, D. A. D’Argenio, L. N. Ornston, Genetic analysis of a chromo-
somal region containing vanA and vanB, genes required for conversion of either
ferulate or vanillate to protocatechuate in Acinetobacter. J. Bacteriol. 181, 3494–3504
(1999).
38. J. Nogales et al., Unravelling the gallic acid degradation pathway in bacteria: The gal
cluster from Pseudomonas putida. Mol. Microbiol. 79, 359–374 (2011).
39. Y. Saeki, M. Nozaki, S. Senoh, Cleavage of pyrogallol by non-heme iron-containing
dioxygenases. J. Biol. Chem. 255, 8465–8471 (1980).
40. F. P. Guengerich, M. V. Martin, C. D. Sohl, Q. Cheng, Measurement of cytochrome
P450 and NADPH-cytochrome P450 reductase. Nat. Protoc. 4, 1245–1251 (2009).
41. S. G. Bell, A. B. Tan, E. O. Johnson, L. L. Wong, Selective oxidative demethylation of
veratric acid to vanillic acid by CYP199A4 from Rhodopseudomonas palustris HaA2.
Mol. Biosyst. 6, 206–214 (2010).
19. B. Morawski, A. Segura, L. N. Ornston, Substrate range and genetic analysis of Aci-
netobacter vanillate demethylase. J. Bacteriol. 182, 1383–1389 (2000).
20. E. Masai et al., A novel tetrahydrofolate-dependent O-demethylase gene is essential
for growth of Sphingomonas paucimobilis SYK-6 with syringate. J. Bacteriol. 186,
2757–2765 (2004).
21. T. Abe, E. Masai, K. Miyauchi, Y. Katayama, M. Fukuda, A tetrahydrofolate-dependent
O-demethylase, LigM, is crucial for catabolism of vanillate and syringate in Sphingo-
monas paucimobilis SYK-6. J. Bacteriol. 187, 2030–2037 (2005).
42. J. Chrastil, J. T. Wilson, A sensitive colorimeter method for formaldehyde. Anal. Bio-
chem. 63, 202–207 (1975).
Machovina et al.
PNAS Latest Articles
|
7 of 7