Biochemical Studies of Mycobacterial Fatty Acid Methyltransferase: A Catalyst for the Enzymatic Production of Biodiesel

Article abstract of DOI:10.1016/j.chembiol.2015.09.011

Summary Transesterification of fatty acids yields the essential component of biodiesel, but current processes are cost-prohibitive and generate waste. Recent efforts make use of biocatalysts that are effective in diverting products from primary metabolism to yield fatty acid methyl esters in bacteria. These biotransformations require the fatty acid O-methyltransferase (FAMT) from Mycobacterium marinum (MmFAMT). Although this activity was first reported in the literature in 1970, the FAMTs have yet to be biochemically characterized. Here, we describe several crystal structures of MmFAMT, which highlight an unexpected structural conservation with methyltransferases that are involved in plant natural product metabolism. The determinants for ligand recognition are analyzed by kinetic analysis of structure-based active-site variants. These studies reveal how an architectural fold employed in plant natural product biosynthesis is used in bacterial fatty acid O-methylation. Mycobacterial fatty acid methyltransferases are employed as biocatalysts for the production of biodiesel. Petronikolou and Nair describe structural and biochemical characterization of a mycobacterial fatty acid methyltransferase, reveal an unexpected homology to enzymes involved in plant primary metabolism, and provide insights into substrate preference.

Full text of DOI:10.1016/j.chembiol.2015.09.011

Article
Biochemical Studies of Mycobacterial Fatty Acid
Methyltransferase: A Catalyst for the Enzymatic
Production of Biodiesel
Graphical Abstract
Authors
Nektaria Petronikolou, Satish K. Nair
Correspondence
snair@illinois.edu
In Brief
Mycobacterial fatty acid
methyltransferases are employed as
biocatalysts for the production of
biodiesel. Petronikolou and Nair describe
structural and biochemical
characterization of a mycobacterial fatty
acid methyltransferase, reveal an
unexpected homology to enzymes
involved in plant primary metabolism, and
provide insights into substrate
preference.
Highlights
d
d
d
M. marinum FAMT catalyzes transesterification of fatty acids
to produce biodiesel
Structure of FAMT reveals similarity to plant primary
metabolism methyltransferases
Kinetic characterization of active-site mutants shows basis
for substrate specificity
Petronikolou & Nair, 2015, Chemistry & Biology 22, 1–11
November 19, 2015 ª2015 Elsevier Ltd All rights reserved
http://dx.doi.org/10.1016/j.chembiol.2015.09.011

Please cite this article in press as: Petronikolou and Nair, Biochemical Studies of Mycobacterial Fatty Acid Methyltransferase: A Catalyst for the
Enzymatic Production of Biodiesel, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.09.011
Chemistry & Biology
Article
Biochemical Studies of Mycobacterial Fatty Acid
Methyltransferase: A Catalyst for the Enzymatic
Production of Biodiesel
Nektaria Petronikolou1 and Satish K. Nair1,2,3,
,3
*
1
Department of Biochemistry, University of Illinois at Urbana Champaign, 600 South Mathews Avenue, Urbana, IL 61801, USA
2Center for Biophysics and Computational Biology and University of Illinois at Urbana Champaign, 600 South Mathews Avenue,
Roger Adams Lab Room 430, Urbana, IL 61801, USA
3Institute for Genomic Biology, University of Illinois at Urbana Champaign, 600 South Mathews Avenue, Urbana, IL 61801, USA
*Correspondence: snair@illinois.edu
http://dx.doi.org/10.1016/j.chembiol.2015.09.011
SUMMARY
has properties similar to those of petro-diesel, and can be
used as a motor fuel without major engine modifications (Demi-
Transesterification of fatty acids yields the essential rbaxs, 2002). It is a mixture of long-chain fatty acid alkyl esters
component of biodiesel, but current processes are (FAAEs), produced mainly by the transesterification of fatty acids
derived from vegetable oils and animal fats (Atadashi et al.,
2010). Many biodiesel precursors are derived from triglycerides,
cost-prohibitive and generate waste. Recent efforts
make use of biocatalysts that are effective in divert-
ing products from primary metabolism to yield fatty
acid methyl esters in bacteria. These biotransforma-
tions require the fatty acid O-methyltransferase
which are esters of three equivalents of fatty acids with one
equivalent of glycerol. Triglycerides can undergo transesterifica-
tion with an alcohol in the presence of a catalyst to yield FAAEs
and glycerol as final products. However, this process is expen-
sive and generates waste, negating the financial and environ-
mental benefits of using biodiesel (Leung et al., 2010).
(
FAMT) from Mycobacterium marinum (MmFAMT).
Although this activity was first reported in the
literature in 1970, the FAMTs have yet to be biochem-
With the goal of formulating less energy-intensive and more
ically characterized. Here, we describe several environment-friendly methods for biodiesel production, several
crystal structures of MmFAMT, which highlight an laboratories have engineered microorganisms for the overpro-
unexpected structural conservation with methyl- duction of fatty acid precursors (Kung et al., 2012; Steen et al.,
2
010; Li et al., 2008; Nawabi et al., 2011). More recently, Escher-
transferases that are involved in plant natural
product metabolism. The determinants for ligand
recognition are analyzed by kinetic analysis of struc-
ture-based active-site variants. These studies reveal
how an architectural fold employed in plant natural
product biosynthesis is used in bacterial fatty acid
O-methylation.
ichia coli has been engineered to directly produce fatty acid
methyl esters (FAMEs) and 3-hydroxy fatty acid methyl esters
(
(
3-OH FAMEs) by using a fatty acid O-methyltransferase
FAMT) from Mycobacterium marinum (MmFAMT) for transester-
ification in situ (Figure 1A) (Nawabi et al., 2011). Although the
yield of FAMEs produced by this pathway was lower than the
maximal FAAE yields reported (Steen et al., 2010; Nawabi
et al., 2011), the use of FAMT holds promise for in vivo biodiesel
production.
INTRODUCTION
The fatty acid O-methyltransferases are S-adenosylmethionine
SAM)-dependent enzymes that catalyze the transfer of a methyl
(
The increasing demand for energy, coupled with the need for en- group to the carboxyl group of a fatty acid acceptor (Struck
ergy independence and security, have been the main forces et al., 2012). Canonical SAM-dependent methyltransferases
N
driving the growing interest in the use of biofuels as a potential catalyze alkylation on heteroatoms via an S 2-type nucleophilic
sustainable energy resource. The Energy Independence and Se- substitution mechanism (Blumenthal et al., 1999). Mycobacteria
curity Act of 2007 mandates increased production of biofuels specifically possess a large number of SAM-dependent methyl-
with the aim of reducing the national dependence on unstable transferases that catalyze the modification of mycolic acids, the
foreign suppliers, as well as reducing pollutants produced by very long fatty acids found in mycobacterial cell walls (Barry
the consumption of fossil fuels (US Government, 2007). Although et al., 1998). These enzymes can be broadly classified with regard
biofuels hold great promise, the current methodologies used for to the identity of the methyl acceptor, and include the cyclopro-
their production require large quantities of water, fertilizers, and pane fatty acid synthases (which catalyze C-methylation of a cis
pesticides, rendering biofuels less environmentally friendly and double bond) (Huang et al., 2002), and the O-methyltransferases
sustainable than desired (Naik et al., 2010).
that catalyze methylation on the hydroxyl moiety of phthiocerol,
In recent years, biodiesel has emerged as a viable resource for phenolphthiocerol, or ketomycolate (Yuan et al., 1998; Yuan and
utilization in green energy production. In addition to favorable Barry, 1996). There are three cyclopropane fatty acid synthases
properties such as biodegradability and low toxicity, biodiesel with different regiospecificities; CmaA1 catalyzes the installation
Chemistry & Biology 22, 1–11, November 19, 2015 ª2015 Elsevier Ltd All rights reserved
1

Please cite this article in press as: Petronikolou and Nair, Biochemical Studies of Mycobacterial Fatty Acid Methyltransferase: A Catalyst for the
Enzymatic Production of Biodiesel, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.09.011
1991; Safayhi et al., 1991; Sastry et al., 1994). Although they
were first discovered in mycobacteria in 1970 (Akamatsu and
Law, 1970), there are very few published studies about this class
of methyltransferases (Nawabi et al., 2011; Akamatsu and Law,
1970;Safayhietal., 1991), andtheirphysiologicalroleinmycobac-
teria has yet to be elucidated.
The identification of the mycobacterial FAMTs was facilitated
by the sequence analysis of mycobacterial proteins, and
sequence-based structural alignments suggested that they
contain a Rossmann-like fold found in class I methyltransferases
(Liscombe et al., 2012). The primary sequences of the mycobac-
terial enzymesalso suggestweak, butnotable, similaritiesto plant
natural product methyltransferases, such as dimethylxanthine
methyltransferase (PDB: 2EFJ) (McCarthy and McCarthy, 2007)
and salicylate methyltransferase (PDB: 1M6E) (Zubieta et al.,
2
003), as well as other members of a recently characterized
protein family classified as the SABATH methyltransferase
Pfam03492 [D’Auria etal., 2003];so named forthefirstcharacter-
(
ized members of this fold class) (Figure 1A). However, these plant
enzymes are not known to catalyze methylation of fatty acid sub-
strates, and the sequence identity between the mycobacterial
enzymes and the plant O-methyltransferases is less than 25%.
To understand the basis for this biotechnologically relevant
enzyme, we carried out biochemical and biophysical character-
ization of the MmFAMT that has been previously utilized for the
in vivo production of biodiesel (Nawabi et al., 2011). Cocrystal
structures of the enzyme with a variety of bound substrates allow
for the identification of features that bestow specificity to fatty
acid substrates. Kinetic characterization of active-site variants,
as deduced from the structural data, facilitates the assignment
of roles for various residues and the proposal for a reaction
mechanism consistent with the data. These studies should pro-
vide a framework for future engineering experiments aimed at
adapting FAMTs for in vivo high-yield production of biodiesel
from fatty acid precursors.
RESULTS AND DISCUSSION
Figure 1. Small Molecule Methyltransferase Reactions and
Michaelis-Menten Kinetics
Kinetic Characterization of M. marinum FAMT
(A) Reactions catalyzed by the fatty acid O-methyltransferase (FAMT) and two
The kinetic parameters for wild-type MmFAMT for a panel of
fatty acids and 3-hydroxy fatty acids were determined using a
coupled assay (Dorgan et al., 2006; Wooderchak et al., 2008)
that monitored SAM turnover. In this assay, S-adenosylhomo-
cysteine (SAH), the product of SAM, is hydrolyzed by an SAH
nucleosidase (the Pfs SAH nucleosidase has a catalytic effi-
of its structural homologs from the SABATH family (7-MXMT, 7-methylxan-
thine N-methyltransferase; SAMT, salicylic acid O-methyltransferase). Free
fatty acids can be released by the action of a thioesterase (TES) on acyl-ACP
(
(
(
acyl carrier protein).
B) Michaelis-Menten kinetic plots for the wild-type MmFAMT for octanoate
C8), decanoate (C10), 3-hydroxyoctanoate (3-OH-C8), and 3-hydrox-
ydecanoate (3-OH-C10). All measurements were conducted in triplicate and
the mean average value for each are shown. Error bars indicate the standard
deviations of each data point.
6
ꢀ1 ꢀ1
ciency of 11.6 3 10
M s [Choi-Rhee and Cronan, 2005],
which is greater than that of MmFAMT) to S-ribosylhomocys-
teine and adenine. The latter is subsequently deaminated by
an adenine deaminase, resulting in a decrease in absorbance
ꢀ
1
ꢀ1
of the distal cyclopropane ring on a-mycolic acid when overex- at 265 nm (Dε z 6,700 M cm ), and allowing for continuous
pressed (Yuan et al., 1995; Glickman, 2003), PcaA installs that monitoring of the reaction (Dorgan et al., 2006). The catalytic
proximal cyclopropane ring in a-meroacid (Glickman et al., efficiency of MmFAMT for free fatty acids was found to in-
2
000), and CmaA2 catalyzes the cis-cyclopropane synthesis in crease with chain length (the kcat/K
M
value for C8 octanoic
, and for C10 decanoic acid is
) (Table 1 and Figure 1B). The 3-hydroxy com-
an adjacent hydroxyl group, during the formation of both keto- pounds were turned over less efficiently, with kcat/K values that
3
ꢀ1 ꢀ1
methoxymycolates (Yuan and Barry, 1996). Additional examples acid is 1.49 3 10
M
s
4
ꢀ1 ꢀ1
include MmaA4 that introduces a methyl branch, together with 2.97 3 10 M
s
M
and methoxymycolates (Boissier et al., 2006). In contrast to these are roughly an order of magnitude lower than for the correspond-
other SAM-dependent enzymes, little is known about FAMTs (Na- ing fatty acid. The kinetic parameters for MmFAMT reported here
wabi et al., 2011; Akamatsu and Law, 1970; Orpiszewski et al., differ from values determined previously via a discontinuous
2
Chemistry & Biology 22, 1–11, November 19, 2015 ª2015 Elsevier Ltd All rights reserved

Please cite this article in press as: Petronikolou and Nair, Biochemical Studies of Mycobacterial Fatty Acid Methyltransferase: A Catalyst for the
Enzymatic Production of Biodiesel, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.09.011
related plant SABATH family of methyltransferases, which share
limited sequence similarities (25% sequence identity), and all of
Table 1. Steady-State Kinetic Parameters for MmFAMT Wild-
Type and Mutants Against Different Substrates
ꢀ
2
ꢀ1
ꢀ1 ꢀ1
s )
which function on small-molecule natural products (Liscombe
et al., 2012). To understand this substrate specificity and scope,
we determined several binary and ternary complex structures of
˚
MmFAMT in complex with SAH (1.7 A resolution), SAH and octa-
˚
noate (1.6 A resolution), and SAH and 3-hydroxydecanoate
˚
(1.9 A resolution). We also determined the structure of an
MmFAMT active-site mutant (Q154A) in complex with SAH and
˚
3-hydroxydecanoate (1.85 A resolution). Initial crystallographic
MmFAMT
Wild-Type
C8
K
M
(mM)
k
cat (10
s
)
k
cat/K
M
(M
308 ± 9
46.1 ± 0.5
24.4 ± 0.4
48.8 ± 1.4
59.9 ± 1.1
59.8 ± 3.2
1,497
C10
8.2 ± 0.5
2,039 ± 172
201 ± 8
29,756
239
3-OH-C8
3-OH-C10
2,975
23,450
SAM
Q31A
C8
25.5 ± 4.0
phases were determined by single-wavelength anomalous
diffraction using selenomethionine (SeMet)-labeled protein
a
a
a
a
a
a
(SAH and decanoic acid), and phases for subsequent structures
3-OH-C10
were determined by molecular replacement using the resultant
model as a search probe. All crystallographic data collection
and refinement statistics are provided in Table 2.
Q154A
C8
5,238 ± 427
1,028 ± 43
14.7 ± 0.4
4.2 ± 0.0
28.1
40.9
3-OH-C10
The overall structure of MmFAMT consists of a di-domain
architecture composed of a Rossmann-like a/b fold that is com-
mon among diverse class I methyltransferases, which is further
elaborated with an all a-helical domain that caps the SAM-bind-
ing site (Figure 2A). The Rossmann-like fold core is formed from
discontinuous regions of the polypeptide chain and space resi-
dues Ser11 through Asp217, Pro255 through Ala287, and
Pro354 through the carboxy terminus. This core consists of a
seven-stranded b-sheet core that is sandwiched between sets
W155F
C8
a
a
a
a
1.37 ± 0.03
3.05 ± 0.16
3-OH-C10
Y24F
a
a
SAM
a
Could not be determined.
824 ± 19
assay that employed radiolabeled SAM as a donor (Nawabi et al., of helical regions. The helical capping domain is similarly
011). These differences likely arise from variations in the two composed of disjointed segments and forms part of the active
2
methods used for characterization. However, in the context of site, in a fashion similar to that found in plant natural product
this work, our kinetic data provide an internal standard for methyltransferases (McCarthy and McCarthy, 2007; Zubieta
analyzing the effects of specific amino acid mutations on the cat- et al., 2003; Zhao et al., 2008).
alytic activity of MmFAMT, as reported below.
Despite a low conservation in primary sequence, a search of
the PDB (Bernstein et al., 1977) using the Dali server (Holm and
Rosenstrom, 2010) shows that MmFAMT is structurally homolo-
Binding Affinity of MmFAMT for SAM and SAH
The product SAH is known to inhibit some SAM-dependent gous to the SABATH class of plant natural product methyltrans-
methyltransferases. To characterize whether MmFAMT is simi- ferases, which themselves are only distantly related to other
larly subject to product inhibition, we carried out isothermal titra- well-characterized small-molecule methyltransferases (D’Auria
tion calorimetric (ITC) analysis to measure the binding affinity to et al., 2003). Representative structural homologs include the sal-
SAM and SAH (Table S1 and Figure S1). MmFAMT binds to SAM icylic acid (SA) O-methyltransferase from Clarkia breweri (PDB:
˚
with a dissociation constant (K
SAH occurs with a K
0.70 mM). For both ligands, binding is driven largely by the en- methylxanthine N-methyltransferase from Coffea canephora
D
) of 33.6 mM, while binding of 1M6E; root-mean-square deviation [RMSD] of 2.5 A over 330
D
two orders of magnitude smaller Ca atoms; 23% sequence identity) (Zubieta et al., 2003), the di-
(
˚
thalpic component. These data are consistent with the inhibition (PDB: 2EFJ; RMSD of 3.2 A over 328 Ca atoms; 22% sequence
of MmFAMT by the product SAH. It should be noted that all mea- identity) (McCarthy and McCarthy, 2007), and indole-3-acetic
surements using SAM as the ligand yielded a number of binding acid O-methyltransferase from Arabidopsis thaliana (PDB:
˚
sites of two (n = 2.06 ± 0.01), while measurements with SAH gave 3B5I; RMSD of 2.6 A over 312 Ca atoms; 23% sequence identity)
a number of sites of one (n = 0.93 ± 0.01). The basis for this (Zhao et al., 2008) (Figure S2). The unexpected structural similar-
discrepancy is not clear, as the structural data do not reveal ev- ities between MmFAMT and the plant SABATH O-methyltrans-
idence for multiple binding sites. Nonetheless, it is evident from ferases reflect the similar substrate repertoires of these
the binding isotherms that MmFAMT binds tightly to SAH. These enzymes, specifically that they both catalyze methylation of car-
results are in agreement with product inhibition observed in boxylic acid on an otherwise hydrophobic substrate. However,
several members of this family (Kung et al., 2015; Obianyo and the nature of the hydrophobic groups differs as the plant en-
Thompson, 2012; Lee et al., 2005). Hence, both in vivo and zymes utilize larger, bulkier scaffolds that are typically aromatic,
in vitro efficacy of enzymatic methylation of free fatty acids using while the FAMTs function on fatty acids containing a long hydro-
FAMTs may be improved if the reactions are coupled to an SAH carbon acyl chain (Figure 1A).
nucleosidase to hydrolyze the SAH product.
The plant SABATH methyltransferases described above are
all homodimers in solution and the corresponding crystal struc-
tures demonstrate a two-fold symmetric arrangement. Likewise,
Crystal Structures of MmFAMT
The unique specificity of bacterial FAMTs for fatty acid sub- MmFAMT is homodimeric both in solution and in the crystal lattice
strates cannot be understood in the context of the distantly (Figure S3). In plant O-methyltransferases that are not members of
Chemistry & Biology 22, 1–11, November 19, 2015 ª2015 Elsevier Ltd All rights reserved
3

Please cite this article in press as: Petronikolou and Nair, Biochemical Studies of Mycobacterial Fatty Acid Methyltransferase: A Catalyst for the
Enzymatic Production of Biodiesel, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.09.011
Table 2. Data Collection, Phasing, and Refinement Statistics
Native
SeMet
C8
3-OH-C10
Q154A-3-OH-C10
Data Collection
Space group
P2
1
P2
1
P2
1
P2
1
P2
1
˚
a, b, c (A), b ( )
ꢂ
63.1, 66.1, 98.4, 107.5 62.7, 65.7, 97.5, 107.8 63.1, 66.1, 98.1, 107.4 62.8, 65.9, 98.2, 107.1 62.8, 65.6, 98.2, 107.3
˚
Resolution (A)
a
50–1.7 (1.73–1.7)
5.4 (52.1)
50–1.95 (1.98–1.95)
5.5 (58.8)
50–1.6 (1.63–1.6)
6.0 (57.3)
50–1.9 (1.93–1.9)
6.5 (54.1)
50–1.85 (1.88–1.85)
7.0 (81.4)
b
R
sym (%)
I/s(I)
25.6 (2.6)
15.9 (2.2)
20.5 (1.9)
18.2 (2.1)
14.5 (1.9)
Completeness (%) 100 (99.9)
100 (100)
100 (99.7)
4.2 (3.5)
99.5 (100)
4.1 (4.1)
100 (100)
Redundancy
Phasing
5.0 (4.8)
4.1 (4.1)
4.7 (4.7)
c
FOM
0.382/0.157
Refinement
˚
Resolution (A)
25.0–1.7
80,265
25.0–1.6
95,148
25.0–1.9
55,801
25.0–1.85
61,413
No. of reflections
d
R
work/Rfree
19.5/22.2
19.5/21.8
19.3/23.1
19.3/22.4
No. of atoms
Protein
Fatty acid
SAH
5,504
-
5,533
20
5,525
26
5,525
44
52
52
52
52
Water
793
802
563
570
B factors
Protein
Fatty acid
SAH
16.7
-
16.9
16.2
8.1
24.1
17.2
12.8
30.9
23.9
22.2
12.6
31.2
7.9
27.9
Water
28.4
RMSD
˚
Bond lengths (A)
0.000
1.07
0.005
1.07
0.006
1.15
0.006
1.11
ꢂ
Bond angles ( )
a
Highest-resolution shell is shown in parentheses.
b
c
R
sym = Sj(I
i
ꢀ <I
i
>)jSI
i i i
, where I is the intensity of the ith reflection and <I > is the mean intensity.
Mean figure of merit (acentric/centric).
d
R factor = S(jFobsj ꢀ kjFcalcj)/S jFobsj, and Rfree is the R value for a test set of reflections consisting of a random 5% of the diffraction data not used in
refinement.
the SABATH class, notably chalcone O-methyltransferase and plane of the b strands. In the MmFAMT-SAH structure, the py-
caffeic acid O-methyltransferase, the dimeric arrangement is rimidine ring of SAH adenine is sandwiched between the side
functional, with residues from one monomer contributing to the chain of Phe134, via a p-stacking interaction on one side and
active site of the other subunit (Zubieta et al., 2001, 2002). Val99 on the other (Figure 2B). Extensive hydrogen bonding in-
Although the SABATH enzymes are also dimeric, oligomerization teractions further stabilize the bound SAH, including the interac-
is not needed for activity. Oligomerization of MmFAMT is similarly tion between Ser133 and the adenine amine, between Asp98
likely not functional, as the monomeric entity contains all of the and both hydroxyls of the ribose, and between Tyr24 and
necessary residues involved in catalysis. Likewise, dimerization Ser150 and the a-carboxylate of SAH. Additional contacts are
2
˚
only buries ꢁ1,314 A of solvent-accessible surface area (corre- mediated through van der Waals interactions with non-polar res-
sponding to 8% of the total surface area of each monomer), which idues that contribute to form the SAM/SAH binding site. An anal-
is far less than for the functionally homodimeric enzymes. The ogous network of interacting residues stabilizes bound SAM/
dimerization interface of MmFAMT is similar in organization to SAH in structures of the plant natural product O-methyltrans-
that of the SA O-methyltransferase (Zubieta et al., 2003) and ferases. The extensive set of interactions and the hydrophobicity
indole-3-acetic acid O-methyltransferase (Zhao et al., 2008), of the binding site rationalize why uncharged SAH is a strong
both of which also likely function as a monomer.
competitive inhibitor of MmFAMT.
The cocrystal structure of MmFAMT with SAH and octanoate
shows that the substrate-binding site is largely localized within
SAM/SAH and Substrate-Binding Pockets
Similar to other class I methyltransferases, the SAM/SAH binding the a-helical domain that caps the Rossmann-like fold
site is situated in the Rossmann-like fold domain, and the ligand domain (Figure S2A). This capping domain contains a binding
is bound in an extended manner, roughly perpendicular to the cavity that is sufficiently contoured to accommodate fatty acid
4
Chemistry & Biology 22, 1–11, November 19, 2015 ª2015 Elsevier Ltd All rights reserved

Please cite this article in press as: Petronikolou and Nair, Biochemical Studies of Mycobacterial Fatty Acid Methyltransferase: A Catalyst for the
Enzymatic Production of Biodiesel, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.09.011
Figure 2. Structure of Fatty Acid Methyl-
transferase and Binding Pocket of SAM/
SAH
(
A) Ribbon diagram of the MmFAMT/SAH crystal
structure, with one monomer colored in gray and
the other colored in purple (Rossmann-like fold
domain) and cyan (capping domain).
(
(
B) Simulated annealing difference Fourier map
-F ) contoured to 2.5s (blue) and 10s (red)
F
o
c
showing the SAM-binding site of MmFAMT. The
coordinates for bound SAH were omitted during
map calculations. The bound ligand is colored
yellow, and residues that interact with the ligand
are colored gray.
substrates with acyl chain lengths up to C12. Numerous aliphatic Lastly, Gln154 is positioned to hydrogen bond with the carbox-
and aromatic residues encircle the acyl chain of the fatty acid ylate of the substrate in the MmFAMT/SAH/octanoate structure.
substrate, including Trp155, Met214, Phe222, Tyr224, Val256, However, a modest rotation of the plane of the substrate, neces-
Phe311, and Leu316 (Figure 3A). MmFAMT residues Tyr174, sary to maximize interactions of the 3-hydroxyl group with
Met 227, and Asn228 enclose the upper side of the binding cav- Gln154, results in a slight movement of this residue away from
ity, and multiple conformers are observed for the latter two the substrate carboxylate in the MmFAMT/SAH/3-hydrodeca-
residues. The planar, extended conformation of the fatty acid noate structure.
substrate is set by the side chains of Met19, Ala315, and
In the cocrystal structure of MmFAMT with 3-hydroxydeca-
Leu316 on one side of the substrate and by Trp151 and Tyr224 noate, only the S enantiomer was bound in the active site,
on the opposite side. In the SA O-methyltransferase cocrystal although an enantiomeric mixture of 50% R and 50% S was
structure, Met150 and Met308 similarly sandwich the substrate used in the crystallization condition. This enantiomeric selectivity
for proper positioning at the SAM methyl donor (Zubieta et al., is due to the presence of hydrophobic residues (Phe311 and
2
003). Lastly, hydrogen bonding with Gln31 and Trp155 engage Trp151) on one side of the active site in close proximity to the
and orient the carboxylate moiety of the fatty acid substrate for carboxyl group, and the presence of Gln154 on the other side
methyl transfer, similar to the Gln25/Trp151 pair utilized by SA in a position favorable for hydrogen bonding with the 3-hydroxyl
O-methyltransferase to orient the carboxylate of its substrate. group of the S enantiomer. Although the physiological role of
The relative disposition of the a-helical capping and Ross- MmFAMT is yet to be elucidated, this enantiomeric specificity
mann-like fold domains is slightly different than that observed may be of physiological relevance. 3-Hydroxylated fatty acids
in structures of plant O-methyltransferases with a similar bilobed are found in mycobacterial phospholipids (Alugupalli et al.,
structure, which may also contribute to different substrate scope 1994, 1995), enter b-oxidation as (S)-3-hydroxyacyl coenzyme
of the FAMTs (Figure S2).
A (Shimakata et al., 1979; Williams et al., 2011), and are part of
A comparison of the MmFAMT/SAH cocrystal structures with the mycolic acid biosynthesis as (R)-3-hydroxy-acyl-ACP (acyl
bound octanoate and 3-hydroxydecanoate suggests a rationale carrier protein) (Sacco et al., 2007; Marrakchi et al., 2014).
for understanding why MmFAMT shows a preference for the Methylation of the carboxyl group of the (S)-3-hydroxy fatty acids
M
larger substrate (K values of C10 substrates lower by a factor may be part of any of these processes by regulating the fate of
of ꢁ10–40 relative to the C8 counterparts; Table 1). While the (S)-3-hydroxy fatty acids. This hypothesis, however, is yet
Met227 and Asn228 are observed as multiple conformers in to be investigated.
the cocrystal structure with octanoate (Figure 3B), both of these
residues exist in single conformations oriented away from the Molecular Basis for Fatty Acid Substrate Specificity
substrate-binding cavity to accommodate the larger acyl chain As noted, the structure of MmFAMT shares several features with
in the SAH/3-hydroxydecanoate structure (Figure 3D). This that of SA O-methyltransferase (Zubieta et al., 2003), including
movement is accompanied by a compensatory reorganization the presence of a hydrophobic a-helical domain that caps the
of a loop spanning residues Leu166 through Gln171, located SAM-binding site. While both enzymes utilize a similar constella-
near the omega carbon of the fatty acid, resulting in tighter pack- tion of active-site residues to orient the substrate carboxylate
ing against C9 and C10 of the acyl chain (Figure 3D). Movement (Gln31/Trp155 in MmFAMT and Gln25/Trp151 in SA O-methyl-
of these side chains fills the substrate-binding cavity but only transferase), the capping domain establishes specificity for the
with acyl chains of length C10–C12, and likely explains why hydrophobic part of the substrate (Figures 4A and 4B). Specif-
these longer-chain fatty acids are better substrates than octa- ically, the contours of each cavity are optimized for its cognate
noate (Figures 3B and 3D). There are limited interactions in the substrate. In SA O-methyltransferase the active site is border-
active site with the 3-hydroxyl moiety, as Gln154 is located within lined by Ile 225, Trp226, Tyr255, and Phe347, and in MmFAMT
hydrogen bonding distance and the thioether of Met214 is more by Tyr174, Phe222, Tyr224, Met227, and Asn228. The replace-
˚
than 3.2 A away. Placement of the polar 3-hydroxyalcohol in a ment of Ile225 and Trp226 (in SA O-methyltransferase) by
hydrophobic binding pocket without sufficient compensatory in- Met227 and Asn228 (in MmFAMT) creates a larger cavity to
M
teractions may explain the higher K values for 3-hydroxy-con- accommodate the longer acyl chain of the fatty acid (Figures 3,
taining substrates relative to their acyl counterparts (Table 1). 4A, and 4B).
Chemistry & Biology 22, 1–11, November 19, 2015 ª2015 Elsevier Ltd All rights reserved
5

Please cite this article in press as: Petronikolou and Nair, Biochemical Studies of Mycobacterial Fatty Acid Methyltransferase: A Catalyst for the
Enzymatic Production of Biodiesel, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.09.011
To further probe this function, we generated site-specific muta-
tions at several of these residues and determined kinetic param-
eters for the variant enzymes (Table 1). The Gln31/Ala mutation
had an immense effect on the enzyme activity, as no product for-
mation could be detected even when 20-fold more enzyme and
much greater concentrations (up to 5 mM) of substrate were
used. Similarly, the Trp155/Phe mutation resulted in a near
M
1,000-fold decrease in catalytic efficiency (kcat/K ). As Gln31
and Trp155 are situated near the carboxylate of the substrate,
the Gln31/Ala and Trp155/Phe mutations likely compromise
binding and orientation of the substrate, consistent with the loss
of activity in each of these mutants.
The Gln154/Ala mutation had a large effect on the K
MmFAMT with octanoate as a substrate (17-fold increase), and
a smaller effect on the K for 3-hydroxydecanoate (5-fold in-
M
of
M
crease). This was an unexpected result, as Gln154 is within
hydrogen bonding distance to the 3-hydoxyl group, and a
mutation at this residue would be expected to cause a much
M
greater increase in the K for the 3-hydroxy fatty acid. To
provide a rationale for this observation, we determined the
˚
1.85-A resolution cocrystal structure of MmFAMT Gln154/Ala
in complex with SAH and 3-hydroxydecanoate. The structure
shows that the 3-hydroxyl group rotates toward the sulfur
of Met214 to compensate for the loss of interaction with
Gln154 (Figure S4), explaining why this mutation results in only
a modest increase in the K
M
for 3-hydroxydecanoate. The
Figure 3. FAMT Active Site with Bound Substrates
A and C) Difference Fourier maps (F -F ) contoured to 2.5s (blue) showing the
bound fatty acid substrates (colored green).
B and D) Surface cut-away diagram showing that binding of the larger sub-
strate results in a smaller cavity due to movement of Leu166 through Gln171.
Gln154/Ala mutation also results in the carboxylate shifting
away to result in a less optimal orientation for methyl transfer,
consistent with a 10-fold lower kcat for the mutant relative to
wild-type MmFAMT.
(
o
c
(
Although the constrained smaller cavity of SA O-methyltrans- Conclusions
ferase may not accommodate a fatty acid, the larger cavity of Our collective structural and biochemical analysis demonstrates
MmFAMT raised the question as to whether MmFAMT can utilize how MmFAMT uses a SABATH plant natural product methyl-
SA as a substrate despite the lack of residues that would anchor transferase architecture to catalyze the methylation of fatty
the aromatic ring in an optimal orientation. To further investigate acid substrates. This adaptation is a result of minor alterations
ꢂ
this, we carried out reactions (2 hr incubation at 37 C) of in secondary structural elements, and of the changes in the
MmFAMT with SA in triplicate, and analyzed the products by disposition of the a-helical capping domain, which engages
gas chromatography-mass spectrometry (GC-MS). Surprisingly, the substrate, relative to the Rossmann-like fold domain that
MmFAMT is able to catalyze methylation on SA (Table S2). A harbors the methyl donor. Nonetheless, the identification of
closer examination of an active-site superposition of the cocrys- MmFAMT as a structural homolog of the SABATH class of meth-
tal structures of MmFAMT/SAH/octanoate and SA O-methyl- yltransferases extends the function of Pfam03492 (D’Auria et al.,
transferase/SAH/SA reveals that SA can be accommodated 2003) to beyond plant metabolism.
into the MmFAMT active site. Although the positioning of Prior studies on the plant SABATH members suggest that
Phe311 (Cys307 in SA O-methyltransferase) would seemingly these enzymes do not require a general base to deprotonate
cause some steric clashes, the residue at the other side of the ar- the substrate methyl acceptor, as its carboxylate is likely
a
omatic ring (Tyr255 in SA O-methyltransferase) is a smaller ionized at physiological pH due to its low pK (Zubieta et al.,
Val256 in MmFAMT, which would allow for positioning of SA as 2003). Likewise, there are no residues in the active site of
a substrate. Equivalent, but distinct, residues in the MmFAMT MmFAMT that can abstract the proton from the fatty acid
active site could orient SA. However, in SA O-methyltransferase, carboxylate prior to methyl transfer, suggesting that the
Ile225 and Trp226, which are replaced by the aforementioned enzyme active site simply serves to facilitate proximity and
flexible Met 227-Asn228 in MmFAMT, constrict the upper end orientation of the reactive groups. The side-chain amide of
of the active site to accommodate the smaller SA substrate (Fig- Gln31 and the indole nitrogen of Trp155 are within hydrogen
ures 4A and 4B). Consequently, it is likely that SA may not bind bonding distance from the two carboxylate oxygens, and
very well in the MmFAMT active site.
structure-guided engineering studies (Zubieta et al., 2003)
demonstrate that replacement of the equivalent Gln in plant
O-methyltransferases establishes selectivity against different
Kinetic Analysis of Wild-Type and Mutant MmFAMTs
The cocrystal structures of MmFAMT identified a number of res- methyl acceptors. The replacement of both of these residues
idues that may play a role in either substrate binding or catalysis. in the MmFAMT homolog from Mycobacterium smegmatis
6
Chemistry & Biology 22, 1–11, November 19, 2015 ª2015 Elsevier Ltd All rights reserved

Please cite this article in press as: Petronikolou and Nair, Biochemical Studies of Mycobacterial Fatty Acid Methyltransferase: A Catalyst for the
Enzymatic Production of Biodiesel, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.09.011
Figure 4. Methyltransferase Active Site Comparisons
(
A and B) Comparison of active sites in (A) MmFAMT (gray) with bound SAH (pink), octanoic acid (green) and (B) the SABATH family enzyme SA O-methyl-
transferase (gray) with bound SAH (pink) and salicylic acid (cyan).
C) Structure-based alignment of the primary sequences of MmFAMT, MsXMT, and SA O-methyltransferase. The green triangles demarcate residues involved in
(
interactions with the carboxylic acid, the brown circles demarcate residues that form the fatty acid binding pocket, and the magenta diamonds indicate residues
that interact with SAH/SAM.
(52% sequence identity, UniProt: A0R0D4) (MsXMT in Fig- mologs for the in vivo production of various FAMEs of desired
ure 4C) may account for the lack of activity against fatty acid length.
substrates in the latter (Nawabi et al., 2011).
Prior attempts to utilize MmFAMT for production of biodiesel in
Studies of plant SABATH enzymes have determined that small E. coli (Nawabi et al., 2011) resulted in yields that were consider-
changes in the primary amino acid sequence of these enzymes ably lower than for other reported fatty acid ethyl esterification
can establish methyltransferase activity on structural scaffolds processes (Steen et al., 2010). From the crystal structure of
with diverse chemical structures. Similarly, small changes in MmFAMT, as well as the previous in vivo studies (Nawabi
the primary sequence of mycobacterial methyltransferases ho- et al., 2011), it is evident that MmFAMT can methylate me-
mologous to MmFAMT (Figures 5 and S6) provide a pool of en- dium-chain fatty acids (up to 12–14 carbons long), which could
zymes that have been engineered by nature to utilize fatty acids account for the low observed yields. Specifically, low compati-
of different chain lengths. The identification of the active site res- bility between MmFAMT and the fatty acid thioesterases
idues of MmFAMT that recognize the carboxyl of the substrate used to generate the fatty acid substrate would result in
as well as the residues that line the hydrophobic pocket that har- methylation of only a fraction of the available fatty acids. The uti-
bors the fatty acid tail can guide the selection of MmFAMT ho- lization of complementary combinations of thioesterases with
Chemistry & Biology 22, 1–11, November 19, 2015 ª2015 Elsevier Ltd All rights reserved
7

Please cite this article in press as: Petronikolou and Nair, Biochemical Studies of Mycobacterial Fatty Acid Methyltransferase: A Catalyst for the
Enzymatic Production of Biodiesel, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.09.011
BAA-535) and primers designed based on the published sequence. The
gene was cloned into the NdeI/BamHI sites of a pET28b vector, and this
plasmid was used as template for the generation of the site-specific mutants
by PCR (Table S3). The integrity of all recombinant plasmids was confirmed
by sequencing (ACGT).
Protein Expression and Purification
Expression vectors bearing wild-type or mutant MmFAMT were transformed
into E. coli Rosetta 2(DE3) cells for heterologous protein production. A 5-ml
starter culture was inoculated in 2 L of Luria-Bertani growth medium supple-
mented with 50 mg/ml kanamycin and 25 mg/ml chloramphenicol. The culture
ꢂ
was grown at 37 C until the absorbance at 600 nm reached 0.6–0.8, at which
point protein production was induced by addition of 0.3 mM isopropyl-b-D-1-
ꢂ
thiogalactopyranoside. The culture was then cooled to 18 C and grown for an
additional 18 hr. Cells were collected by centrifugation, resuspended in 20 mM
Tris-HCl (pH 8.0), 500 mM NaCl, and 10% glycerol buffer, and lysed by multiple
passages through a C5 Emulsiflex (Avestin) cell homogenizer. Following
centrifugation of the lysate, the supernatant was applied to a 5-ml His-Trap
(
GE Biosciences) column that was previously equilibrated with 20 mM Tris-
HCl (pH 8.0), 1 M NaCl, and 30 mM imidazole. The column was extensively
washed with the same buffer, and elution of specifically bound protein was
carried out using a gradient of increasing imidazole concentration. Fractions
containing protein of the highest purity (as determined by SDS-PAGE) were
pooled and the polyhistidine affinity tag was removed by overnight incubation
ꢂ
with thrombin at 4 C. Samples were further purified using size-exclusion chro-
matography (Superdex HiloadTM 200 16/60) in a buffer of 20 mM HEPES
Figure 5. Sequence Similarity Network
The network (Gerlt et al., 2015; Shannon et al., 2003) contains 838 nodes and
(
pH 7.5) and 300 mM KCl. Samples were concentrated using a 10,000-Da mo-
lecular weight cutoff Amicon centrifugal filter. SeMet MmFAMT was produced
by repression of methionine biosynthesis in defined media prior to protein in-
duction (Doublie, 1997), and was purified as described above. All proteins
242,549 edges. Each node represents amino acid sequences that are 95% or
more identical, and each edge connects a pair of sequences at an E value of
ꢀ
30
better than 1 3 10 . This E value is the one generated by the EFI-EST (Gerlt
ꢂ
were flash-frozen in liquid nitrogen and stored at ꢀ80 C. Prior to freezing,
et al., 2015), and is not identical to E values generated by BLAST (Gerlt et al.,
2
.5% glycerol (final concentration) was added to samples of wild-type and
2015; Altschul et al., 1990; Altschul, 1993). Teal, plants; indigo, actinobacteria;
mutant proteins used for kinetic analysis, but not to samples used for crystal-
lographic studies.
fuchsia, proteobacteria; green, cyanobacteria; red, fungi; plum, other.
mycobacterial methyltransferases that accept different fatty acid
substrate lengths can result in better yields. Based on our re-
sults, additional metabolic engineering, for example, overex-
pression of an efficient SAH nucleosidase to avert product
inhibition by SAH, can further increase the methyl ester produc-
tion. These data provide a starting point for engineering efforts
directed at exploiting both MmFAMT and homologs for the in vivo
production of various FAMEs of desired length.
Crystallization
Initial crystallization conditions were determined by the sparse matrix sam-
pling method using commercial screens. Crystals of the MmFAMT/SAH com-
plex were grown using the hanging-drop vapor diffusion method. In brief, 0.9 ml
of protein at 4 mg/ml concentration was incubated with 2 mM SAM for 1 hr on
ice, and was subsequently mixed with 0.9 ml of precipitant solution (0.1 M
2
MgCl , 0.1 M N-(2-acetamido)iminodiacetic acid [pH 6.5], 12% polyethylene
glycol 6000) and 0.2 ml of 7% (v/v) 1-butanol, and equilibrated over a well
containing the same precipitant solution at room temperature. Although the
crystallization media used SAM, the resultant structure revealed density corre-
sponding to SAH, due to either hydrolysis or the presence of trace amounts of
SAH in the SAM preparation. Crystals grew within 2 days, and were transiently
soaked in precipitant solution supplemented with 30% glycerol prior to vitrifi-
cation by direct immersion in liquid nitrogen. Ligand bound crystals were
grown similarly using protein that was incubated with 2 mM SAH, 2 mM (3-hy-
droxy) fatty acid, and 1 mM DTT for 30 min prior to crystallization.
SIGNIFICANCE
Methylation of free fatty acids affords a viable route toward
the production of biodiesel. The structure of the mycobacte-
rial FAMT reveals an architectural conservation with en-
zymes involved in plant natural product biosynthesis. Kinetic
analysis of structure-based variants provides a rationale for
substrate specificity. These data provide the framework for
further engineering experiments aimed at expanding the
substrate scope of this biotechnologically relevant catalyst.
Data Collection, Phasing, and Structure Determination
X-Ray diffraction data were collected at Life Sciences Collaborative Access
Team (LS-CAT), Sector 21, Argonne National Laboratory. All data were in-
dexed, integrated, and scaled using either HKL2000 or XDS. Initial crystallo-
graphic phases were determined by single-wavelength anomalous diffraction
from the SeMet-labeled protein crystals of the MmFAMT-SAH complex using
data collected near the SE absorption edge (l = 0.97872 A˚ ). Due to the low
EXPERIMENTAL PROCEDURES
symmetry of the SeMet crystals (monoclinic setting), care was taken during
Chemicals and Reagents
crystal alignment to maximize the simultaneous collection of Bijvoet pairs.
A four-fold redundant dataset was collected to 1.95 A˚ resolution (Rmerge of
All chemical reagents were purchased from Sigma-Aldrich unless otherwise
noted. All of the materials used for protein production and purification were
purchased from GE Healthcare.
5.5%, I/s(I) of 2.2 in the highest-resolution shell) using a MAR CCD detector.
The heavy-atom substructure was determined using SHELX (Sheldrick,
2
010). The heavy-atom positions were imported into SHARP (Bricogne et al.,
Cloning and Site-Specific Mutagenesis
2003) for maximum-likelihood refinement resulting in preliminary phases with
a mean figure of merit of 0.382. Density modification using solvent flattening
and non-crystallographic symmetry averaging yielded a map of exceptional
MmFAMT (GenBank: NC010612; gene MMAR3356) was amplified using PCR
with template genomic DNA of Mycobacterium marinum strain M (ATCC
8
Chemistry & Biology 22, 1–11, November 19, 2015 ª2015 Elsevier Ltd All rights reserved

Please cite this article in press as: Petronikolou and Nair, Biochemical Studies of Mycobacterial Fatty Acid Methyltransferase: A Catalyst for the
Enzymatic Production of Biodiesel, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.09.011
quality, which allowed for nearly all of the main chain and roughly half of the
side chains to be automatically built using Buccaneer (Cowtan, 2006, 2008)
as implemented in the CCP4 suite of programs (Winn et al., 2011). Additional
cycles of manual rebuilding interspersed with crystallographic refinement us-
ing REFMAC5 (Murshudov et al., 2011) resulted in the final model. The final
cycles of model building and refinement were carried out against high-resolu-
tion data collected on native crystals of MmFAMT-SAH.
chased as enantiomeric mixtures. However, it is clear from the crystal struc-
ture of the enzyme complexed with SAH and 3-OH-C10 that the enzyme
utilizes only the S enantiomer (Figures 3C and S4). Subsequently, the enantio-
meric ratio was determined by Mosher ester analysis (Hoye et al., 2007), and
was found to be 1:1.
End-point Activity Assay
Cocrystal structures with bound ligands were all determined by molecular
replacement as implemented in the Phenix program suite (Adams et al.,
The activity of the wild-type enzyme for SA was investigated by GC-MS. 100-ml
reactions containing 50 mM wild-type MmFAMT, 500 mM SAH nucleosidase,
ꢂ
2
010), using the refined coordinates of SeMet MmFAMT as search model.
5 mM SAM, and 10 mM substrate were incubated at 37 C for 2 hr. The samples
The resultant solutions were subsequently used as starting models for
several rounds of automated model building using the ARP/wARP web
server (Winn et al., 2011; Murshudov et al., 2011; Langer et al., 2008), fol-
lowed by rounds of manual rebuilding using Coot (Emsley et al., 2010),
combined with crystallographic refinement using REFMAC5 (Murshudov
et al., 2011). Ligands were built in Coot, and water molecules were added
using the ARP/wARP solvent building software of the CCP4 suite (Lamzin
and Wilson, 1993), and confirmed by manual inspection. In all cases, the
quality of the in-progress model was routinely monitored using both
the free R factor (Read et al., 2011) and MolProbity (Chen et al., 2010) for
quality assurance.
were cleaned from the enzyme with a 10,000-Da molecular weight cutoff
Amicon spin filter, and analyzed by the Roy J. Carver Biotechnology Center
(University of Illinois at Urbana Champaign). All reactions were performed in
the same buffer used for kinetic analysis supplemented with 5% methanol
(final concentration). To confirm that methanol did not deactivate the enzyme
and that possible observed inactivity was not due to its addition, reactions
containing wild-type MmFAMT and C8 were included. Control reactions
without addition of the enzyme were also analyzed. Results are means ±
SEM of triplicate experiments.
Isothermal Titration Calorimetry
The binding affinity of wild-type MmFAMT for SAM and SAH was measured at
ꢂ
2
5 C using a VP-ITC microcalorimeter (Microcal). Protein and ligands were in
Determination of Kinetic Parameters
the same buffer used for kinetic analysis. For binding of SAM, 1.1–1.15 mM
SAM was injected into the reaction cell containing 30–35 mM protein in 28 suc-
cessive aliquots at 300-s intervals and 20.5-s duration, with a reference power
of 2 mcal/s. For binding of SAH, 0.52–0.54 mM SAH was injected into the reac-
tion cell containing 40–42 mM protein in 28 successive aliquots at 240-s inter-
vals and 20.5-s duration, with a reference power of 6 mcal/s. All injections were
For all experiments, SAM was further purified by high-performance liquid
chromatography (HPLC) using a C18 column (Vydac; 5-mm particle size,
4.6 3 250 mm) and monitoring absorbance at 260 nm. The column was
washed for 30 min with solvent B (methanol with 0.1% trifluoroacetic acid),
and equilibrated for 15 min with solvent A (water with 0.1% trifluoroacetic
acid). SAM was injected into the column and a gradient elution was applied
as follows: wash with 5 ml of solvent A, elute with a linear gradient to a final
1
0 ml in volume, except for the first injection which was 4 ml and was excluded
from data analysis. The protein-ligand buffer was used in the reference cell,
and a titration of the ligand into just the buffer was subtracted from the mea-
surements. Non-linear regression with a single-site fitting model (MicroCal
Origin) was applied for data analysis, and the thermodynamic parameters
were calculated using the Gibbs free energy equation (DG = DH ꢀ TDS) and
2
0% of solvent B, and wash with 5 ml of solvent B. The flow rate was 1 ml/min
throughout the procedure. The fraction of SAM collected was subsequently
ꢂ
lyophilized and stored at ꢀ20 C. Fresh solution of SAM was prepared before
each experiment.
The kinetic parameters of the wild-type and mutant proteins were deter-
mined using a photospectrometric assay that monitors the production of
SAH (Dorgan et al., 2006; Wooderchak et al., 2008). All enzyme reactions
the relationship DG = ꢀRTlnK
a
. Results are means ± SEM of duplicate exper-
iments. For all binding experiments, freshly purified protein was used, and
commercial preparations of SAM were purified by HPLC as described above.
ꢂ
were performed in 100 mM HEPES (pH 7.8) and 300 mM KCl at 37 C, and
monitored at 265 nm for up to 20 min. Control reactions without addition of
the substrate were also included to take into account any background hydro-
lysis of SAM over time. For determination of the Michaelis-Menten parameters
of the wild-type enzyme and mutants for the fatty acids and 3-hydroxy fatty
acids, a 150-ml reaction volume contained the following components: 0.1–
Sequence Similarity Network
A sequence similarity network was generated by using the Enzyme Function
Initiative Enzyme Similarity Tool (EFI-EST: http://efi.igb.illinois.edu/efi-est/)
(
Gerlt et al., 2015) with the sequence of MmFAMT as the query for a BLASTP
Altschul et al., 1990; Altschul, 1993) search of the UniProtKB database (http://
(
2.0 mM MmFAMT, 1 mM SAH nucleosidase, 0.2 mM adenine deaminase,
mM MnSO , 80 mM SAM, and various concentrations of fatty acids and 3-hy-
www.uniprot.org/) (UniProt Consortium, 2015). Only sequences of 300–450
1
4
amino acids were included for the subsequent generation of a network with
droxy fatty acids. For determination of the kinetic parameters for SAM, a 150-ml
reaction volume contained 0.2 mM MmFAMT, 1 mM SAH nucleosidase, 0.2 mM
ꢀ
30
E-values equal to or lower than 1 3 10 . This network was visualized using
Cytoscape 3.2.1 (Figure 5) (Shannon et al., 2003).
adenine deaminase, 1 mM MnSO
4
, 3 mM C8, and various concentrations of
and Vmax values
SAM. Based on the initial reaction rates, the apparent K
M
were determined using the Michaelis-Menten function of Origin (OriginLab).
Results are means ± SEM of triplicate experiments.
SUPPLEMENTAL INFORMATION
For mutants with very increased K
the enzyme could not be achieved due to limited solubility of the substrates in
the reaction buffer. In these cases, the apparent K and Vmax values could not
be determined, but the kcat/ values were obtained by plotting the observed
M
values for C8 or 3-OH-C10, saturation of
Supplemental Information includes Supporting Methods, six figures, and four
tables and can be found with this article online at http://dx.doi.org/10.1016/j.
chembiol.2015.09.011.
M
K
M
rates (kobs) at four different substrate concentrations (Figure S5). For the Y24F
ACKNOWLEDGMENTS
mutant, the kinetic parameters for SAM, and consequently for C8 and 3-OH-
C10, could not be determined, as the K
M
value increased such that saturation
We would like to thank the staff at Life Sciences Collaborative Access Team
(LS-CAT, Argonne National Laboratory, Argonne, IL) for facilitating data
collection, and Dr. Alexander Ulanov from the Metabolomics Center (Roy J.
Carver Biotechnology Center, University of Illinois at Urbana-Champaign) for
the GC-MS analysis. We would also like to thank Drs. Spencer Peck and Bijoy
Desai for helpful discussions about the kinetic studies.
of the enzyme could not be achieved due to limitations of the assay (concen-
tration of SAM used should be kept below 250 mM to remain in the linear range
of the spectrophotometer [Dorgan et al., 2006]).
To attest that the coupling enzymes were not rate limiting, the initial rates of
0.1, 0.2, and 0.4 mM wild-type enzyme were determined by addition of 3 mM
C8. The means of the observed rates ± SEM were found to be the same (Fig-
ure S5A), indicating that the coupling enzymes are not rate limiting. Conse-
quently, the measured rate corresponds to the rate of MmFAMT.
Received: May 13, 2015
Revised: September 4, 2015
Accepted: September 24, 2015
Published: October 29, 2015
All substrates were purchased from Sigma-Aldrich, except for the 3-OH-C8,
which was purchased from Matreya. The 3-hydroxy fatty acids were pur-
Chemistry & Biology 22, 1–11, November 19, 2015 ª2015 Elsevier Ltd All rights reserved
9

Please cite this article in press as: Petronikolou and Nair, Biochemical Studies of Mycobacterial Fatty Acid Methyltransferase: A Catalyst for the
Enzymatic Production of Biodiesel, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.09.011
REFERENCES
Gerlt, J.A., Bouvier, J.T., Davidson, D.B., Imker, H.J., Sadkhin, B., Slater, D.R.,
and Whalen, K.L. (2015). Enzyme Function Initiative-Enzyme Similarity Tool
(EFI-EST): a web tool for generating protein sequence similarity networks.
Biochim. Biophys. Acta 1854, 1019–1037.
Adams, P.D., Afonine, P.V., Bunkoczi, G., Chen, V.B., Davis, I.W., Echols, N.,
Headd, J.J., Hung, L.-W., Kapral, G.J., Grosse-Kunstleve, R.W., et al. (2010).
PHENIX: a comprehensive Python-based system for macromolecular struc-
ture solution. Acta Crystallogr. D Biol. Crystallogr. 66, 213–221.
Glickman, M.S. (2003). The mmaA2 gene of Mycobacterium tuberculosis en-
codes the distal cyclopropane synthase of the alpha-mycolic acid. J. Biol.
Chem. 278, 7844–7849.
Akamatsu, Y., and Law, J.H. (1970). The enzymatic synthesis of fatty acid
methyl esters by carboxyl group alkylation. J. Biol. Chem. 245, 709–713.
Glickman, M.S., Cox, J.S., and Jacobs, W.R., Jr. (2000). A novel mycolic acid
cyclopropane synthetase is required for cording, persistence, and virulence of
Mycobacterium tuberculosis. Mol. Cell 5, 717–727.
Altschul, S.F. (1993). A protein alignment scoring system sensitive at all evolu-
tionary distances. J. Mol. Evol. 36, 290–300.
Altschul, S.F., Gish, W., Miller, W., Myers, E.W., and Lipman, D.J. (1990). Basic
Holm, L., and Rosenstrom, P. (2010). Dali server: conservation mapping in 3D.
local alignment search tool. J. Mol. Evol. 215, 403–410.
Nucleic Acids Res. 38, W545–W549.
Alugupalli, S., Portaels, F., and Larsson, L. (1994). Systematic study of the
Hoye, T.R., Jeffrey, C.S., and Shao, F. (2007). Mosher ester analysis for the
determination of absolute configuration of stereogenic (chiral) carbinol car-
bons. Nat. Protoc. 2, 2451–2458.
3
2
-hydroxy fatty acid composition of mycobacteria. J. Bacteriol. 176, 2962–
969.
Alugupalli, S., Lan e´ elle, M.A., Larsson, L., and Daff e´ , M. (1995). Chemical char-
acterization of the ester-linked 3-hydroxy fatty acyl-containing lipids in
Mycobacterium tuberculosis. J. Bacteriol. 177, 4566–4570.
Huang, C.C., Smith, C.V., Glickman, M.S., Jacobs, W.R., Jr., and Sacchettini,
J.C. (2002). Crystal structures of mycolic acid cyclopropane synthases from
Mycobacterium tuberculosis. J. Biol. Chem. 277, 11559–11569.
Atadashi, I.M., Aroua, M.K., and Aziz, A.A. (2010). High quality biodiesel and its
Kung, Y., Runguphan, W., and Keasling, J.D. (2012). From fields to fuels:
diesel engine application: a review. Renew. Sustainable Energy Rev. 14, 1999–
recent advances in the microbial production of biofuels. ACS Synth. Biol. 1,
2008.
4
98–513.
Barry, C.E., III, Lee, R.E., Mdluli, K., Sampson, A.E., Schroeder, B.G., Slayden,
R.A., and Yuan, Y. (1998). Mycolic acids: structure, biosynthesis and physio-
logical functions. Prog. Lipid Res. 37, 143–179.
Kung, P.-P., Huang, B., Zehnder, L., Tatlock, J., Bingham, P., Krivacic, C.,
Gajiwala, K., Diehl, W., Yu, X., and Maegley, K.A. (2015). SAH derived potent
and selective EZH2 inhibitors. Bioorg. Med. Chem. Lett. 25, 1532–1537.
Bernstein, F.C., Koetzle, T.F., Williams, G.J., Meyer, E.F., Jr., Brice, M.D.,
Rodgers, J.R., Kennard, O., Shimanouchi, T., and Tasumi, M. (1977). The
Protein Data Bank: a computer-based archival file for macromolecular struc-
tures. J. Mol. Biol. 112, 535–542.
Lamzin, V.S., and Wilson, K.S. (1993). Automated refinement of protein
models. Acta Crystallogr. D Biol. Crystallogr. 49, 129–147.
Langer, G., Cohen, S.X., Lamzin, V.S., and Perrakis, A. (2008). Automated
macromolecular model building for X-ray crystallography using ARP/wARP
version 7. Nat. Protoc. 3, 1171–1179.
Blumenthal, R.M., Cheng, X., and Fauman, E.B. (1999). Structure and
Evolution of AdoMet-dependent Methyltransferases, First Edition (World
Scientific), pp. 1–38.
Lee, W.J., Shim, J.Y., and Zhu, B.T. (2005). Mechanisms for the inhibition
of DNA methyltransferases by tea catechins and bioflavonoids. Mol.
Pharmacol. 68, 1018–1030.
Boissier, F., Bardou, F., Guillet, V., Uttenweiler-Joseph, S., Daffe, M.,
Quemard, A., and Mourey, L. (2006). Further insight into S-adenosylmethio-
nine-dependent methyltransferases: structural characterization of Hma, an
enzyme essential for the biosynthesis of oxygenated mycolic acids in
Mycobacterium tuberculosis. J. Biol. Chem. 281, 4434–4445.
Leung, D.Y.C., Wu, X., and Leung, M.K.H. (2010). A review on biodiesel pro-
duction using catalyzed transesterification. Appl. Energy 87, 1083–1095.
Li, Q., Du, W., and Liu, D. (2008). Perspectives of microbial oils for biodiesel
Bricogne, G., Vonrhein, C., Flensburg, C., Schiltz, M., and Paciorek, W. (2003).
Generation, representation and flow of phase information in structure determi-
nation: recent developments in and around SHARP 2.0. Acta Crystallogr.
D Biol. Crystallogr. 59, 2023–2030.
production. Appl. Microbiol. Biotechnol. 80, 749–756.
Liscombe, D.K., Louie, G.V., and Noel, J.P. (2012). Architectures, mechanisms
and molecular evolution of natural product methyltransferases. Nat. Prod.
Rep. 29, 1238–1250.
Chen, V.B., Arendall, W.B., Headd, J.J., Keedy, D.A., Immormino, R.M.,
Kapral, G.J., Murray, L.W., Richardson, J.S., and Richardson, D.C. (2010).
MolProbity: all-atom structure validation for macromolecular crystallography.
Acta Crystallogr. D Biol. Crystallogr. 66, 12–21.
Marrakchi, H., Lan e´ elle, M.-A., and Daff e´ , M. (2014). Mycolic acids: structures,
biosynthesis, and beyond. Chem. Biol. 21, 67–85.
McCarthy, A.A., and McCarthy, J.G. (2007). The structure of two N-methyl-
transferases from the caffeine biosynthetic pathway. Plant Physiol. 144,
Choi-Rhee, E., and Cronan, J.E. (2005). A nucleosidase required for in vivo
function of the S-adenosyl-L-methionine radical enzyme, biotin synthase.
Chem. Biol. 12, 589–593.
8
79–889.
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, 355–367.
Cowtan, K. (2006). The Buccaneer software for automated model building. 1.
Tracing protein chains. Acta Crystallogr. D Biol. Crystallogr. 62, 1002–1011.
Cowtan, K. (2008). Fitting molecular fragments into electron density. Acta
Crystallogr. D Biol. Crystallogr. 64, 83–89.
Naik, S.N., Goud, V.V., Rout, P.K., and Dalai, A.K. (2010). Production of
first and second generation biofuels: a comprehensive review. Renew.
Sustainable Energy Rev. 14, 578–597.
D’Auria, J.C., Chen, F., and Pichersky, E. (2003). The SABATH family of MTS in
Arabidopsis thaliana and other plant species. In Recent Advances in
Phytochemistry, T.R. John, ed. (Elsevier), pp. 253–283.
Nawabi, P., Bauer, S., Kyrpides, N., and Lykidis, A. (2011). Engineering
Escherichia coli for biodiesel production utilizing a bacterial fatty acid methyl-
transferase. Appl. Environ. Microbiol. 77, 8052–8061.
Demirbasx, A. (2002). Diesel fuel from vegetable oil via transesterification and
soap pyrolysis. Energy Sourc. 24, 835–841.
Obianyo, O., and Thompson, P.R. (2012). Kinetic mechanism of protein argi-
Dorgan, K.M., Wooderchak, W.L., Wynn, D.P., Karschner, E.L., Alfaro, J.F.,
Cui, Y., Zhou, Z.S., and Hevel, J.M. (2006). An enzyme-coupled continuous
spectrophotometric assay for S-adenosylmethionine-dependent methyltrans-
ferases. Anal. Biochem. 350, 249–255.
nine methyltransferase 6 (PRMT6). J. Biol. Chem. 287, 6062–6071.
Orpiszewski, J., Hebda, C., Szyku1a, J., Powls, R., Clasper, S., and Rees, H.H.
(1991). Multiple forms of O-methyltransferase involved in the microbial conver-
sion of abietic acid into methyl abietate by Mycobacterium sp. FEMS
Microbiol. Lett. 82, 233–236.
Doublie, S. (1997). Preparation of selenomethionyl proteins for phase determi-
nation. Methods Enzymol. 276, 523–530.
Emsley, P., Lohkamp, B., Scott, W.G., and Cowtan, K. (2010). Features and
Read, R.J., Adams, P.D., Arendall, W.B., 3rd, Brunger, A.T., Emsley, P.,
Joosten, R.P., Kleywegt, G.J., Krissinel, E.B., Lutteke, T., Otwinowski, Z.,
development of Coot. Acta Crystallogr. D Biol. Crystallogr. 66, 486–501.
10 Chemistry & Biology 22, 1–11, November 19, 2015 ª2015 Elsevier Ltd All rights reserved

Please cite this article in press as: Petronikolou and Nair, Biochemical Studies of Mycobacterial Fatty Acid Methyltransferase: A Catalyst for the
Enzymatic Production of Biodiesel, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.09.011
et al. (2011). A new generation of crystallographic validation tools for the pro-
Williams, K.J., Boshoff, H.I., Krishnan, N., Gonzales, J., Schnappinger, D., and
Robertson, B.D. (2011). The Mycobacterium tuberculosis b-oxidation genes
echA5 and fadB3 are dispensable for growth in vitro and in vivo.
Tuberculosis (Edinb.) 91, 549–555.
tein data bank. Structure 19, 1395–1412.
Sacco, E., Covarrubias, A.S., O’Hare, H.M., Carroll, P., Eynard, N., Jones, T.A.,
Parish, T., Daffe, M., Backbro, K., and Quemard, A. (2007). The missing piece
of the type II fatty acid synthase system from Mycobacterium tuberculosis.
Proc. Natl. Acad. Sci. USA 104, 14628–14633.
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., et al. (2011).
Overview of the CCP4 suite and current developments. Acta Crystallogr.
D Biol. Crystallogr. 67, 235–242.
Safayhi, H., Anazodo, M.I., and Ammon, H.P. (1991). Calmodulin- and Ca2(+)-
insensitive fatty acid methyltransferase from RINm5F cells. Inhibition by trifluo-
perazine and W7. Int. J. Biochem. 23, 769–772.
Wooderchak, W.L., Zhou, Z.S., and Hevel, J. (2008). Assays for S-adenosyl-
methionine (AdoMet/SAM)-dependent methyltransferases. Curr. Protoc.
Toxicol., Chapter 4, Unit4.26.
Sastry, B.V., Vidaver, P.S., Janson, V.E., and Franks, J.J. (1994). S-adenosyl-
L-methionine-mediated enzymatic methylations in the rat retinal membranes.
J. Ocul. Pharmacol. 10, 253–263.
Yuan, Y., and Barry, C.E. (1996). A common mechanism for the biosynthesis of
methoxy and cyclopropyl mycolic acids in Mycobacterium tuberculosis. Proc.
Natl. Acad. Sci. USA 93, 12828–12833.
Shannon, P., Markiel, A., Ozier, O., Baliga, N.S., Wang, J.T., Ramage, D., Amin,
N., Schwikowski, B., and Ideker, T. (2003). Cytoscape: a software environment
for integrated models of biomolecular interaction networks. Genome Res. 13,
Yuan, Y., Lee, R.E., Besra, G.S., Belisle, J.T., and Barry, C.E., 3rd (1995).
Identification of a gene involved in the biosynthesis of cyclopropanated my-
colic acids in Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. USA 92,
2498–2504.
Sheldrick, G. (2010). Experimental phasing with SHELXC/D/E: combining
chain tracing with density modification. Acta Crystallogr. D Biol. Crystallogr.
6
630–6634.
Yuan, Y., Zhu, Y., Crane, D.D., and Barry, C.E., 3rd (1998). The effect of
oxygenated mycolic acid composition on cell wall function and macrophage
growth in Mycobacterium tuberculosis. Mol. Microbiol. 29, 1449–1458.
6
6, 479–485.
Shimakata, T., Fujita, Y., and Kusaka, T. (1979). Purification and characteriza-
tion of 3-hydroxyacyl-CoA dehydrogenase of Mycobacterium smegmatis.
J. Biochem. 86, 1191–1198.
Zhao, N., Ferrer, J.L., Ross, J., Guan, J., Yang, Y., Pichersky, E., Noel, J.P.,
and Chen, F. (2008). Structural, biochemical, and phylogenetic analyses sug-
gest that indole-3-acetic acid methyltransferase is an evolutionarily ancient
member of the SABATH family. Plant Physiol. 146, 455–467.
Steen, E.J., Kang, Y., Bokinsky, G., Hu, Z., Schirmer, A., McClure, A., Del
Cardayre, S.B., and Keasling, J.D. (2010). Microbial production of fatty-acid-
derived fuels and chemicals from plant biomass. Nature 463, 559–562.
Zubieta, C., He, X.Z., Dixon, R.A., and Noel, J.P. (2001). Structures of two nat-
ural product methyltransferases reveal the basis for substrate specificity in
plant O-methyltransferases. Nat. Struct. Biol. 8, 271–279.
Struck, A.W., Thompson, M.L., Wong, L.S., and Micklefield, J. (2012). S-ad-
enosyl-methionine-dependent methyltransferases: highly versatile enzymes
in biocatalysis, biosynthesis and other biotechnological applications.
Chembiochem 13, 2642–2655.
Zubieta, C., Kota, P., Ferrer, J.L., Dixon, R.A., and Noel, J.P. (2002). Structural
basis for the modulation of lignin monomer methylation by caffeic acid/5-hy-
droxyferulic acid 3/5-O-methyltransferase. Plant Cell 14, 1265–1277.
UniProt Consortium. (2015). UniProt: a hub for protein information. Nucleic
Acids Res. 43, D204–D212.
Zubieta, C., Ross, J.R., Koscheski, P., Yang, Y., Pichersky, E., and Noel, J.P.
(2003). Structural basis for substrate recognition in the salicylic acid carboxyl
methyltransferase family. Plant Cell 15, 1704–1716.
US Government. (2007). Energy Independence and Security Act of 2007
(EISA). Public Law 110-140, 1492–1801.
Chemistry & Biology 22, 1–11, November 19, 2015 ª2015 Elsevier Ltd All rights reserved 11