Biochemical and structural basis of triclosan resistance in a novel enoyl-acyl carrier protein reductase

Article abstract of DOI:10.1128/AAC.00648-18

Enoyl-acyl carrier protein reductases (ENR), such as FabI, FabL, FabK, and FabV, catalyze the last reduction step in bacterial type II fatty acid biosynthesis. Previously, we reported metagenome-derived ENR homologs resistant to triclosan (TCL) and highly similar to 7- hydroxysteroid dehydrogenase (7-AHSDH). These homologs are commonly found in Epsilonproteobacteria, a class that contains several human-pathogenic bacteria, including the genera Helicobacter and Campylobacter. Here we report the biochemical and predicted structural basis of TCL resistance in a novel 7-AHSDH-like ENR. The purified protein exhibited NADPH-dependent ENR activity but no 7-AHSDH activity, despite its high homology with 7-AHSDH (69% to 96%). Because this ENR was similar to FabL (41%), we propose that this metagenome-derived ENR be referred to as FabL2. Homology modeling, molecular docking, and molecular dynamic simulation analyses revealed the presence of an extrapolated six-amino-acid loop specific to FabL2 ENR, which prevented the entry of TCL into the active site of FabL2 and was likely responsible for TCL resistance. Elimination of this extrapolated loop via site-directed mutagenesis resulted in the complete loss of TCL resistance but not enzyme activity. Phylogenetic analysis suggested that FabL, FabL2, and 7-AHSDH diverged from a common short-chain dehydrogenase reductase family. This study is the first to report the role of the extrapolated loop of FabL2-type ENRs in conferring TCL resistance. Thus, the FabL2 ENR represents a new drug target specific for pathogenic Epsilonproteobacteria.

Full text of DOI:10.1128/AAC.00648-18

AAC Accepted Manuscript Posted Online 11 June 2018
Antimicrob. Agents Chemother. doi:10.1128/AAC.00648-18
Copyright © 2018 American Society for Microbiology. All Rights Reserved.
1
2
Biochemical and Structural Basis of Triclosan Resistance in a Novel Enoyl-Acyl
Carrier Protein Reductase
3
4
5
Running title: Novel triclosan resistant enoyl ACP reductase
6
7
8
Raees KhanaϮ, Amir ZebbϮ, Nazish Roya, Roniya Thapa Magara, Hyo Jeong Kima, Keun Woo
Leeb#, Seon-Woo Leea#
9
10
11
aDepartment of Applied Bioscience, Dong-A University, Busan 49315, Republic of Korea
12
13
14
bDivision of Applied Life Science (BK21 Plus Program), Plant Molecular Biology and
Biotechnology Research Center (PMBBRC), Research Institute of Natural Science (RINS),
Gyeongsang National University, Jinju 52828, Republic of Korea
15
16
17
18
19
20
#Address correspondence to Seon-Woo Lee, seonlee@dau.ac.kr and to Keun Woo Lee
kwlee@gnu.ac.kr
ϮThese authors contributed equally to this work.
21
22
23
24
25
1

26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
ABSTRACT
Enoyl-acyl carrier protein reductase (ENR), such as FabI, FabL, FabK and FabV, catalyzes the
last reduction step in bacterial type II fatty acid biosynthesis. Previously, we reported
metagenome-derived ENR homologs resistant to triclosan (TCL) and highly similar to 7-α
hydroxysteroid dehydrogenase (7-AHSDH). These homologs are commonly found in
Epsilonproteobacteria, a class that contains several human pathogenic bacteria, including the
genera Helicobacter and Campylobacter. Herein, we report the biochemical and predicted
structural basis of TCL resistance in a novel 7-AHSDH-like ENR. The purified protein exhibited
NADPH-dependent ENR activity but no 7-AHSDH activity, despite its high homology with 7-
AHSDH (69%-96%). Because this ENR was similar to FabL (41%), we propose that this
metagenome-derived ENR is referred to as FabL2. Homology modeling, molecular docking, and
molecular dynamic simulation analyses revealed the presence of an extrapolated six-amino acid
loop specific to FabL2 ENR, which prevented the entry of TCL into the active site of FabL2 and
was likely responsible for TCL resistance. Elimination of this extrapolated loop via site-directed
mutagenesis resulted in the complete loss of TCL resistance but not enzyme activity.
Phylogenetic analysis suggested that FabL, FabL2, and 7-AHSDH diverged from a common
short-chain dehydrogenase reductase family. This study is the first to report the role of the
extrapolated loop of FabL2-type ENRs in conferring TCL resistance. Thus, the FabL2 ENR
represents a new drug target specific for pathogenic bacteria of Epsilonproteobacteria.
2

46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
INTRODUCTION
Enoyl-acyl carrier protein (ACP) reductase (ENR) catalyzes the last step of the bacterial type II
fatty acid synthesis (FASII) cycle to reduce the enoyl-ACP to fully saturated acyl-ACP (Fig. 1A
and 1B). NADH, NADPH, or reduced flavin mononucleotide (FMNH2) function as coenzymes
in this reduction reaction (1). A majority of the enzymes involved in the FASII cycle are
relatively conserved among bacteria, except ENR (2). Till date, four prototypic bacterial ENR
isozymes have been reported, including FabI (3), FabL (4), FabV (5), and FabK (6). Except for
FabK, which is an FMN-containing protein, all ENR isozymes belong to the short-chain
dehydrogenase reductase (SDR) superfamily (7). These ENRs share low sequence similarity
(15%–30%), although their active sites and specific sequence motifs required for coenzyme
binding are highly conserved (8–10).
Triclosan [5-chloro-2-(2,4-dichlorophenoxy)phenol] (TCL) is a broad-spectrum
antimicrobial that targets ENR of various organisms, thereby blocking FASII and ultimately
preventing microbial growth (11). For years, TCL has been incorporated in a variety of consumer
and personal care products worldwide (12–14) because of its potential antimicrobial activity.
However, TCL resistance is prevalent among bacteria, and various mechanisms have been
proposed as the basis for this resistance, including high ENR expression (15), mutant ENR
versions tolerant to TCL (16), cell membrane modifications (17), various efflux pumps (15, 18),
TCL-degrading enzymes (19), novel ENRs, and other unknown TCL resistance determinants
(20). Additionally, TCL has been known to impose selective pressure on bacterial pathogens,
thereby inducing co- or cross-resistance to other antibiotics (16, 20–25). Furthermore, there are
increasing concerns regarding the excessive use of TCL and its negative effects on the
environment and public health (12, 13).
3

69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
FabI is known as the only effective ENR target for TCL, although substitutions of various
key amino acid residues in FabI result in significant TCL resistance (20, 26–28). FabK has been
reported to confer either low (20) or high resistance to TCL (6), whereas other ENRs such as
FabL confer low resistance, and FabV (5), 7-α hydroxysteroid dehydrogenase (7-AHSDH)-like
ENR homologs, and FabG-like ENR homologs confer high resistance to TCL (20). Being pivotal
for bacterial survival and growth, ENRs have been used as potential targets for various
antimicrobials for decades, and various synthetic ENR inhibitors have either been marketed or
are currently being developed or in trial (29).
Previously, we identified a novel, TCL-tolerant, 7-AHSDH-like protein homolog
(KT982367.1; AOR51268.1) from soil metagenome, which complements ENR activity in a
conditional Escherichia coli mutant JP1111 (fabIts) and renders the bacteria completely tolerant
to TCL (20). The gene encoding this 7-AHSDH-like protein is of special interest, as it is present
in Epsilonproteobacteria, a class that contains many bacterial pathogens of humans, including
Helicobacter pylori and Campylobacter jejuni (20). The gene encoding the 7-AHSDH-like
homolog from H. pylori and C. jejuni has been shown to confer TCL resistance in E. coli (20). In
the present study, we aimed to determine whether the 7-AHSDH-like protein possesses dual
enzymatic activity of ENR and 7-AHSDH and its role in TCL resistance. Based on phylogenetic
analysis, biochemical characterization, and molecular simulation, the 7-AHSDH-like enzyme
showed divergent evolution from the SDR family, with a unique TCL resistance mechanism. We
propose that this 7-AHSDH-like enzyme be referred to as FabL2. Because the FabL2-type ENRs
are commonly found along with FabI type ENRs (30) in the pathogenic bacterial group of
Epsilonproteobacteria, FabL2 has significant implications in healthcare and drug discovery.
4

92
93
MATERIALS AND METHODS
Bacterial strains, plasmids, culture condition, and general DNA manipulation
The E. coli strains DH5α, EPI300, and BL21 (DE3) were grown at 37°C in Luria–Bertani (LB)
broth or on LB agar media containing appropriate antibiotics: TCL (1–600 µg/ml; Sigma–
Aldrich Co., St. Louis, MO, USA), chloramphenicol (50 µg/ml), ampicillin (100 µg/ml), or
kanamycin (50 µg/ml). Recombinant DNA manipulation was performed as previously described
(31). Oligonucleotide synthesis and DNA sequencing were conducted at the DNA sequencing
facility of MacroGen (Seoul, Korea). Nucleotide and amino acid sequences were compared using
the online version of BLAST and ORF finder, publicly available at the National Center for
Biotechnology Information portal (NCBI; http://blast.ncbi.nlm.nih.gov). Multiple sequence
alignments were performed using BioEdit v7.2.5 and GeneDoc v2.7 software.
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
Phylogenetic analysis
Phylogenetic analysis was performed as previously described (20) for metagenomic FabL2
ENR using amino acid sequences of FabL2 and its homologs, prototypic FabL, FabI, FabV,
FabK ENRs, and prototypic 7-AHSDH from Comamonas testosteroni and its homologs
retrieved from the UniRef50 database (updated on September 19, 2017). Top 10 entries were
selected from each homology search. All identified sequences compiled together with the
closely related prototypic ENRs and metagenomic FabL2, and redundant sequences were
removed using the online Decrease Redundancy program (32). Sequence alignment and
phylogenetic tree construction were performed with MEGA 6 (33) using the MUSCLE
algorithm (34). To analyze the alignment output in MEGA 6, the maximum likelihood method
was used in combination with the nearest-neighbor-interchange strategy, resulting in the
5

115
116
117
118
119
deletion of gaps present in less than 50% of the sequences and generating 500 bootstrapped
replicates resampling data sets to evaluate the confidence.
Expression and purification of FabL2 ENR
A gene encoding FabL2 ENR was PCR amplified from pBF1-4 (20) using gene-specific forward
120
121
primer (5′-ATTCAAGGATCCTAGAGACATGACAAATATGAAAGGCAA-3′) and reverse
primer (5′-TTATCATCTTTTAACATATAATAGATGGTCGACTTTCAA-3′) containing
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
BamHI and SalI restriction sites, respectively. The amplified PCR product was digested with
BamHI and SalI restriction endonucleases and cloned into pET-30b(+) expression vector to
generate the recombinant vector, pEBF1-4.
To express the FabL2 protein, pEBF1-4 was transformed into E. coli BL21 (DE3) cells,
and recombinant cells were selected on LB agar medium containing kanamycin. E. coli cells
carrying pEBF1-4 were grown in 200 ml of LB supplemented with kanamycin at 37°C until
reaching an optical density of 0.5 at 600 nm (OD600). To induce protein expression, isopropyl β-
D-1-thiogalactopyranoside (IPTG) (1 mM) was added to the bacterial culture during the late
exponential phase. For protein purification, E. coli cells were harvested, re-suspended in 5 ml of
binding buffer [20 mM Tris-Cl, 0.5 M NaCl, 40 mM Imidazole, (pH 8.0)], and subjected to
sonication (Sonic Dismembrator Model 500; Fisher Scientific) for 2 min (pulse ON: 5 s, pulse
OFF: 10 s). This mixture was then centrifuged at 3,500 × g for 6 min at 25°C. The supernatant
was collected and re-centrifuged at 17,000 × g for 10 min at 25°C and filtered using a 0.45-µm
membrane filter. The fusion protein was purified using AKTA prime liquid chromatography
system (GE Healthcare, Buckinghamshire, UK) with His Trap™ HP affinity column (1 ml bed
6

137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
volume; GE Healthcare). The identity of the purified fusion protein was confirmed by denaturing
sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE).
Analysis of FabL2 enzymatic activity
While the deduced amino acid sequence of FabL2 showed high sequence similarity to that of 7-
AHSDH, fabL2 has previously shown to complement the ENR mutant E. coli (20). To test if
FabL2 possesses dual enzymatic activity, enzyme assays were performed using the purified
fusion protein. Biochemical characterization of the purified fusion protein was performed to
determine the optimum reaction conditions and Michaelis–Menten kinetics for ENR activity. All
enzyme assays were performed as previously described (35) with slight modifications. Briefly,
ENR activity was measured in a 100-µl volume containing NADH/NADPH cofactors (250 µM),
crotonyl-coenzyme A (CoA) substrate (200 µM), FabL2 protein (450 nM), and sodium
phosphate buffer (100 mM; pH 7.0) at 25°C. Crotonyl-CoA, NADH, and NADPH were
purchased from Sigma–Aldrich. Enzymatic reactions were monitored using the UV/Vis
Spectrophotometer (DU730 Life Science; Beckman Coulter Inc., Fullerton, CA, USA) at 340 nm
and 30-s intervals for a total of 3 min. Because the protein did not exhibit ENR activity with
NADH and preferred NADPH as a cofactor, the latter was used in all subsequent enzyme assays.
To determine the value of the Michaelis–Menten constant (Km) of the protein, 100 nM of
purified protein was added to 100 µl of the reaction mixture containing 200 μM NADPH and
varying concentrations of crotonyl-CoA (3, 6, 12, 24, 36, and 48 μM). To determine the Km
value of NADPH, 100 nM of protein was added to 60 μM of crotonyl-CoA and varying
concentrations of NADPH (5, 10, 15, 20, 30, 50, and 75 μM). The oxidation of NADPH cofactor
was spectrophotometrically measured at 340 nm. The reaction mixtures were incubated at 25°C
7

160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
for 10 min. To determine the optimal buffer composition and pH for enzymatic activity of the
protein, reactions were conducted using different buffers over a wide pH range: 100 mM sodium
citrate buffer (pH 3.2, 4.2, 5.2, and 6.2) and 100 mM sodium phosphate buffer (pH 6.5, 7, 7.5,
and 8.0). To determine whether FabL2 possessed 7-AHSDH activity, enzyme assays were
performed using cholic acid (Sigma–Aldrich) as the substrate and NADH and NADPH as
cofactors, as previously described (36). All kinetic reactions were performed in triplicates. The
initial velocity of the reaction was calculated from the linear phase of the progress curves. To
calculate the Km values both for the substrate and cofactor, data were fitted to the standard
Michaelis–Menten equation.
Site-directed mutagenesis
Sequence comparison, homology modeling, and docking analysis revealed that FabL2-type ENR
carries an extrapolated highly flexible loop comprising six amino acid residues (Y96–V101); this
was specific to FabL2 ENR only, as other known ENRs lack this loop. To test if this extrapolated
flexible loop was involved in TCL resistance and enzymatic activity, overlap extension PCR was
used to delete the 18-bp loop (Fig. S1, region b). Briefly, two PCRs were performed to amplify
the overlapping fragments A and C of the FabL2 gene (Fig. S1) from pBF1-4 using the primer
177
178
179
180
pairs 7A-1/A-2 (7A-1: 5′-GCCAAAGCGTTGTCAGGTG-3′ and 7A-2: 5′-
AATCATCGCATTGCTTACGAAG-3′) and 7A-3/A-4 (7A-3: 5′-
TCGTAAGCAATGCGATGATTGGCGGATACGGTAAATTTAT-3′ and 7A-4: 5′-
CCCGTCATATTACTCGTTCCCA-3′), respectively. The PCR conditions were as follows:
8

181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
initial denaturation at 95°C for 3 min, 30 cycles of denaturation at 95°C for 30 s, annealing at
variable temperatures (55°C or 63°C for the amplifying fragment A or C, respectively) for 30 s,
and extension at 72°C for 30 s, followed by a final extension at 72°C for 5 min. The amplified
PCR products A and C were gel-purified. This was followed by a fusion PCR using fragments A
and C as templates in equimolar concentrations without any primers and the following conditions:
initial denaturation at 95°C for 3 min, 10 cycles of denaturation at 95°C for 1 min, annealing at
50°C for 30 s, and extension at 72°C for 1 min 35 s, followed by a final extension at 72°C for 5
min. The fusion product was subsequently amplified using 7A-1 and 7A-4 primers (Fig. S1) and
the following conditions: initial denaturation at 95°C for 3 min, 20 cycles of denaturation at
95°C for 1 min, annealing at 63°C for 30 s, and extension at 72°C for 1 min 35 s, followed by a
final extension at 72°C for 5 min. The purified fusion product was cloned into the pGEM-T Easy
vector and transformed into E. coli DH5α to confirm the deletion of the loop and test for TCL
resistance as previously described (20). The mutated version of FabL2 was designated as
mFabL2.
Complementation
To investigate the ENR activity of mFabL2, complementation studies were performed. The
recombinant pGEM-T Easy plasmid carrying mFabL2 was transformed into the conditional
temperature-sensitive fabI mutant of E. coli, JP1111, which is unable to grow at a high
temperature of 42°C (37). E. coli JP1111 containing the mFabL2 vector were grown in triplicates
on LB agar medium supplemented with ampicillin (100 µg/ml) and IPTG at 30°C and 42°C. The
growth of E. coli JP1111 at 42°C for 48ꢀh indicated complementation of FabI ENR activity.
9

204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
TCL resistance test
To determine and compare the growth and TCL resistance of E. coli DH5α expressing either
metagenomic FabL2 or mFabL2, growth assays were performed in LB broth supplemented
with ampicillin and various concentrations of TCL (0–600 µg/ml). E. coli DH5α expressing
Bacillus velezensis FabL homolog (WP_003155478.1) was used as a positive control at
similar TCL concentrations, and E. coli DH5α carrying empty pGEM-T Easy vector was used
as a negative control. Bacterial growth was monitored using UV/Vis Spectrophotometer
(DU730 Life Science; Beckman Coulter Inc., Fullerton, CA, USA) by measuring OD600 over 96
h.
Homology modeling of FabL2
Homology modeling is the construction of an atomic model of the target protein utilizing
experimentally determined structures of evolutionarily-related proteins (38). First, the FabL2
protein sequence (target) was analyzed against the Protein Data Bank (PDB) using the BLASTP
tool in NCBI to identify the suitable protein structure (template)
(https://blast.ncbi.nlm.nih.gov/Blast.cgi?PROGRAM=blastp&PAGE_TYPE=BlastSearch).
Subsequently, the target–template alignment was subjected to MODELLER program
implemented in Discovery Studio v4.5. Ten iterative models of FabL2 were generated, and the
best model was selected based on the lowest probability density function (PDF) total energy and
discrete optimized protein energy (DOPE) score for further analysis.
It is noteworthy to mention that the homology model does not reflect the conformation or
orientation of amino acids comprising side chains in their physiological state. To obtain the
native conformation of FabL2, an unrestrained molecular dynamic (MD) simulation was
10

227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
performed with CHARMm36 force field in GROMACS v5.0.7 (39). Briefly, the system was
solvated in an octahedral box of transferable intermolecular potential three position (TIP3P)
water model. Counter-ions (Na+) were added to neutralize the system. The steepest descent
minimization with a maximum tolerance of 10 kJ/mol/nm was employed to avoid any
unfavorable interactions. The system was equilibrated in two phases. In the first phase, NVT
equilibration was conducted for 100 ps at 300 K. The temperature was maintained with a V-
rescale thermostat. In the second phase, heavy atoms were restrained, and solvent molecules with
counter-ions were allowed to move during the 100-ps simulation at 300 K and 1.0 bar pressure
using the Parrinello–Rahman barostat. The final production step was conducted for 10 ns under
periodic boundary conditions with NPT ensemble and bond constraint algorithm, linear
constraint solver (LINCS). The representative structure of FabL2 was extracted from the last 6-
ns trajectory using the clustering method. The stereochemical quality of the MD-refined model
of FabL2 was verified using PROCHECK implemented in SAVES web server
(http://services.mbi.ucla.edu/PROCHECK/). The MD-refined model of FabL2 was also validated
by ProSA-web (https://prosa.services.came.sbg.ac.at/prosa.php) for its accuracy of potential
errors. The Z-score of ProSA measures the deviation of the total energy of the structure with
respect to an energy distribution derived from random conformations.
Molecular docking simulation of TCL into FabL2
Molecular docking is a computational technique used to predict the binding affinity and
orientation of a ligand in the binding site of a protein. The two-dimensional (2D) structure of
TCL was drawn in Accelrys Draw v4.2 and converted into three-dimensional (3D) structure in
Discovery Studio v4.5. The MD-refined model of FabL2 and TCL were used as input data in
11

250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
Genetic Optimization of Ligand Docking (GOLD) v5.2.2 program. The binding site of FabL2
was traced from its catalytic residues using Define and Edit Binding Site tools implemented in
Discovery Studio v4.5. Docking results were analyzed with the GOLD fitness score that includes
hydrogen bond (H-bond) energy, van der Waals energy, and ligand torsion strains. The best
docking pose was selected based on the GOLD fitness score and H-bonding with catalytic
residues.
Accession number(s)
The nucleotide sequence of pBF1 harboring the FabL2 gene has been deposited in the GenBank
database under the accession number KT982367. The FabL2 ENR protein sequence has been
deposited in the GenBank database under the accession number AOR51268.1.
RESULTS AND DISCUSSION
Phylogenetic analysis of FabL2, 7-AHSDH-like protein, and prototypic ENRs
ENR catalyzes the final reduction step in the bacterial FASII cycle and is indispensable for
establishing and maintaining the rate of fatty acid biosynthesis (40, 41). Amino acid sequence
analysis revealed that FabL2 shared significant homology (69%-96%) with the 7-AHSDH
homologs of Epsilonproteobacteria and relatively less similarity with FabL (41%), FabI (27%),
and prototypic 7-AHSDH (34%) (20). Moreover, FabL2 ENR shared similar structural features
such as highly conserved tyrosine and lysine residues of the active site with prototypic FabI and
FabL ENRs. Similarly, key residues of the enzyme such as Ser146, Lys163, Thr193 and RINA
like sequences were strictly conserved among the FabL2 ENR and prototypic 7-α-HSDH (20).
12

272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
Additionally, the FabL2 protein conferred complete TCL tolerance when expressed in E. coli
and complemented the ENR activity in the E. coli mutant, JPP1111, carrying the fabIts mutation
(20). Phylogenetic analysis of the FabL2 protein with other prototypic ENRs and 7-AHSDH
proteins and their homologs revealed that FabL2-type ENRs clustered as a separate clade (Fig. 2),
suggesting that FabL2 diverged from either closely related FabL and FabI ENRs or from 7-
AHSDH during evolution. Therefore, we designated this enzyme as FabL2. Consequently,
distinguishing these types of ENRs based only on sequence comparison/annotation is not always
ideal (20).
NADPH-dependent ENR activity of the FabL2 protein
A fusion protein of FabL2 was purified and confirmed to be of expected size (33.56 kDa) using
SDS-PAGE (Fig. S2A, S2B). The conversion of NADH/NADPH cofactors to NAD/NADP at
340 nm was monitored to assess the ENR activity. Because the purified protein exhibited
maximum activity in 100 mM sodium phosphate buffer (pH 7.0) (Fig. S2C), all subsequent
assays were performed using this buffer. No enzymatic activity was observed in 100 mM sodium
citrate buffer regardless of the pH (data not shown). The purified protein catalyzed the turnover
of NADPH (Km = 27.64 µM) (Fig. 3A) into NADP in the presence of crotonyl-CoA (Km =
9.627 µM) as a substrate (Fig. 3B). These Km values for the metagenomic ENR were equivalent
to those reported for Chlamydia trachomatis and E. coli ENRs (4, 35), although these values
were slightly lower than those reported by Ward et al. and Basso et al. (42, 43). The purified
protein did not utilize NADH as a cofactor, thus, exhibiting ENR activity only with NADPH.
Among other prototypic ENRs, FabL from B. subtilis showed high similarity to FabL2
ENR (41%), which uses NADPH as a cofactor (4), whereas FabV using NADPH as a cofactor (1)
13

295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
did not show any similarity to FabL2. The kcat values for FabL2 with NADPH (1.09 µM/min)
and crotonyl-ACP (0.18283 µM/min) (Table S1) were within the range reported previously (44)
but were lower than those for the ENR from C. trachomatis (35). These variations might be due
to the use of different substrates. Overall, biochemical analyses confirmed that FabL2 enzyme
possesses an NADPH-dependent ENR activity. However, despite its high similarity with 7-
AHSDH from Epsilonproteobacteria, FabL2 did not exhibit 7-AHSDH activity when tested with
cholic acid substrate and NADH and NADPH cofactors (data not shown). These data suggest
that the enzyme FabL2 from soil metagenome is a bona fide ENR similar to FabL.
Predicted structure of FabL2
Sequence analysis of metagenomic TCL-resistant FabL2 showed 41.0% sequence similarity with
FabL from B. subtilis. Sequence alignment revealed Gly102, Tyr160, Lys167, and Phe204 as the
catalytic residues of FabL2 (Fig. 4A) and an extrapolated loop to be tested for TCL resistance
(Fig. 4B) and ENR function (Fig. 4C). Therefore, the structure of FabL from B. subtilis (PDB ID:
3OID; chain A) was considered as template for homology modeling of FabL2. Among ten
predicted models, the best model of FabL2 was selected based on the lowest molecular
probability density function (MOLPDF) score of 1,414.23 and DOPE score of −26,020.47. The
representative structure of FabL2 was extracted after MD simulation refinement. The
stereochemical quality of the refined FabL2 structure revealed that 88.6% of the residues
occupied the most favored region of the Ramachandran plot (45) (Fig. S3A). These results
suggest that phi (ɸ) and psi (ψ) backbone dihedral angles in the modeled structure are reasonably
accurate. Analysis of FabL2 with ProSA-web (46) revealed a Z-score of −7.31, which was within
the range of Z-scores of experimentally determined structures (Fig. S3B).
14

318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
The MD-refined structure of FabL2 has an architecture similar to that reported for FabL
proteins (47). Briefly, the overall structure of FabL2 comprises a central 7-stranded parallel β-
sheet (β1–β7) sandwich-like structure flanked on both sides by three α-helices, forming an
NADPH-binding Rossman-like fold (48) (Fig. 5A). Our modeled FabL2 structure also exhibited
the same folding pattern in the substrate-binding region (α8 and α9) located near the carboxyl
end of β6 and β7 as previously described for different ENRs and other members of the SDR
family (4, 49, 50). Despite the high similarity of FabL2 with FabL from B. subtilis, FabL2
contained an extrapolated region extending between Tyr96 and Val101 residues (Fig. 4A, Fig.
5A and 5B). The structural superimposition of FabL2 and FabL affirmed that the extrapolated six
amino acid residues formed a loop (Fig. 5A, orange color). The role of this loop in substrate
specificity and TCL resistance of FabL2 was subsequently validated.
Interaction of TCL with FabL2
The best docking pose of TCL with FabL2 revealed a GOLD fitness score of 53.00. Despite the
high docking score, TCL was flipped away (~5.7 Å) from the catalytic site of FabL2 and was
bound at the rim region of the tunnel leading to the substrate-binding site (Fig. 5B). Furthermore,
despite the phenol moiety, the phenoxy group of TCL was oriented toward the catalytic pocket of
FabL2. Molecular interactions between FabL2 and TCL implied that the Arg98 residue of the
extrapolated loop formed two H-bonds with the backbone oxygen and phenolic oxygen of TCL
(Fig. 5C). Moreover, other non-polar interactions confirmed the binding of TCL at the rim region
of the tunnel leading to the catalytic site of FabL2 (Fig. 5C). Our rational approach concluded
that TCL could not reach the catalytic site of FabL2 due to H-bonding with Arg98 and other non-
polar interactions (Fig. 5A and 5B). This flexible loop may determine the shape and size of the
15

341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
tunnel. We further speculate that the extrapolated loop is highly flexible and plays a key role in
TCL resistance of FabL2.
Based on our results, we hypothesized that removing this extrapolated flexible loop
(Tyr96–Val101) from FabL2 renders it sensitive to TCL. To test this hypothesis, we created
mutant FabL2 (mFabL2) lacking the extrapolated loop and predicted its structure as previously
described for FabL2 (Fig. 4B). The best model had the lowest PDF and DOPE scores of 1,353.46
and −24,680.74, respectively. PROCHECK analysis of mFabL2 revealed that 92.6% of amino
acid residues occupied the most favored region of the Ramachandran plot (Fig. S3C). The Z-
score of mFabL2 model was −6.74, which followed the Z-score pattern of wild-type FabL2 (Fig.
S3D). Both FabL2 and mFabL2 showed a similar overall topology and model quality, except for
the tunnel leading to catalytic cavity (Fig. 5D). Based on these data, we speculate that the native
function of mFabL2 was unaffected by the deletion of the flexible loop.
Although the docking score of mFabL2 was the same as that of the wild-type FabL2
(53.00), TCL occupied the catalytic active site of mFabL2. The binding of TCL with mFabL2
was similar to that observed with other ENR family members with respect to the orientation of
the phenol moiety toward the substrate-binding site (Fig. 5E). Moreover, the H-bonding between
TCL and Tyr154 of mFabL2 stabilized its orientation in the active site of mFabL2 (Fig. 5F).
Further, the H-bonding between amino acid residues of mFabL2 and NADPH generate stable
enzyme complex of mFabL2 (Fig. 5G). This binding pattern of TCL is conserved across all ENR
family members (28, 47, 51, 52).
MD simulation of the mFabL2-TCL complex
16

363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
Detailed analysis of the binding mode of TCL with the active site of mFabL2 was conducted via
20-ns MD simulation. The root mean square deviation (RMSD) of the Cα atoms (Fig. 5H), the
simulation of the mFabL2-TCL complex (Fig. 5I) and analysis of molecular interactions (Fig. 5J,
Fig. S4A and S4B) indicated that the system remained stable during the entire simulation period.
This mechanism of targeting the catalytic Tyr by TCL is well documented among other ENR
family members (28, 47, 52). Our analysis also revealed an additional interaction between
NADPH and TCL via H-bonding (Fig. S4A and S4B), which may strengthen the binding of TCL
with mFabL2. Moreover, other molecular interactions, including π-π, alkyl-alkyl, π-alkyl, and
van der Waals interactions were observed between the catalytic site residues of mFabL2 and
TCL (Table S2, Fig. 5G). Taken together, these data validated our hypothesis and revealed that
the deletion of the extrapolated residues did not disturb the native folding of mFabL2 and restore
its sensitivity to TCL. Experimental validation confirmed the native function of mFabL2 as well
as its inhibition by TCL.
The extrapolated loop of FabL2 is involved in TCL resistance
Bioinformatics analysis revealed that the FabL2 has an extra six-residue loop (Tyr96–Val101),
which was specific to and supposed to be involved in TCL tolerance. These six residues extend
the loop dramatically, which pushes the Gly102 (conserved catalytic residue) away, creating a
new topology of the TCL-binding site of FabL2. We speculate that Arg98, which is sequestered
between Gly102 and TCL, interferes with their binding. Moreover, docking analysis of mFabL2
revealed that TCL is able to access the active site of mFabL2. Therefore, we conclude that the
Tyr96–Val101 loop is responsible for the observed TCL tolerance of FabL2; its removal may
result in the loss or reduction of TCL resistance. As expected, deletion of the Tyr96–Val101 loop
17

386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
via site-directed mutagenesis resulted in the loss of TCL resistance in mFabL2 (minimum
inhibitory concentration, 2.5 µg/ml), whereas the wild-type FabL2 was capable of conferring
resistance to TCL even at concentrations as high as 600 µg/ml (Fig. S5A and S5B). Moreover,
complementation analysis revealed that mFabL2 retained its ENR activity (Fig. 4 C). This result
indicates that the Tyr96–Val101 loop is involved in TCL tolerance but not in ENR activity. The
strict amino acid conservation of this extrapolated loop (Fig. S6A) suggests that the loop was
recently introduced into FabL2 of Epsilonproteobacteria.
The extrapolated loop is highly unique and is present only in FabL2-type ENR and its
homologs in Epsilonproteobacteria; it is absent in the closely related prototypic FabL-type ENRs
and prototypic 7-AHSDH homologs (Fig. S6A, Fig. S6B, Table S3). It is unclear how and why
these enzymes have evolved to contain this extrapolated loop. However, this extrapolated loop
(Tyr96 – Gly102) in FabL2 is involved in the topology of tunnel leading to the enzyme active
site. The residues of the target loop are not considered as catalytic moieties of ENR, since
removal of the loop did not affect ENR activity. Although we have not tested specific point
mutation of the extrapolated loop, we speculate that specific point mutation of the loop would
not alter the overall structure of the loop. Our rational approach suggested that extrapolated loop
is highly flexible and fluctuates back and forth to open and close the opening of the tunnel
leading to the active site of FabL2 (Fig. 5A). However, this prediction still awaits further
biochemical investigation.
Our docking analysis of TCL into the active site of FabL2 suggested that TCL is flipped
away from docking site by ~5.7 Å, and hence TCL could not reach the catalytic site. If we delete
this loop, the main entrance of the tunnel will remain opened and TCL would be able to reach the
active site of FabL2. In fact, deletion of extrapolated loop could successfully abolish TCL
18

409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
resistance in mFabL2. Finally, we perceived that extrapolated loop of FabL2 serves as a
checkpoint to selectively allow substrate(s) to reach the active site, which needs to be further
verified by structural characterization. Taken together, this study showed that minor changes in
the structure of bacterial proteins due to small-scale structural variations in the coding sequence
can render the bacteria resistant to antibiotics.
CONCLUSIONS
We conclude that FabL2 ENR confers complete TCL tolerance via a unique extrapolated loop in
its protein structure. This study is the first to report TCL tolerance conferred by residues other
than those directly interacting with the substrate or cofactor. Furthermore, the presence of TCL-
resistant FabL2 ENR homologs among the human pathogenic bacteria of the
Epsilonproteobacteria class indicates that these bacteria may be unaffected by TCL treatment.
Additionally, although the amino acid sequence of FabL2 was highly similar to that of 7-
AHSDH, the lack of 7-AHSDH activity in FabL2 indicates that sequence alignments alone are
not sufficient for determining protein function.
Acknowledgements
This research was supported by a grant from the Next-Generation BioGreen 21 Program to S.-W.
L. (Project No. PJ01313101) through Rural Development Administration and the Pioneer
Research Center Program to K.W. L. (NRF-2015M3C1A3023028) through the National
Research Foundation of Korea funded by the Ministry of Science, ICT and Future Planning,
Republic of Korea.
19

432
433
434
435
436
437
438
439
440
441
442
Transparency declaration
None to declare.
Author contributions
S. W. L., R. K., K.W.L. and A.Z developed the framework for evaluating TCS resistance of the
novel FabL2 type ENR. S.W.L. and K.W.L conceived, organized, and supervised the project. R.
K., A.Z., K.W.L. and S. W. L. interpreted the results and prepared the manuscript. R. K., N.R.,
R.T.M. and H.J.K. performed experiments and analyzed the data. All authors contributed to the
final version of the manuscript. The guarantor is S.W. L.
References
443
444
1. Massengo-Tiassé RP, Cronan JE. 2009. Diversity in enoyl-acyl carrier protein reductases.
Cell. Mol. Life Sci. 66:1507–1517.
445
446
2. Jackowski S, Murphy CM, Cronan JE, Rock CO. 1989. Acetoacetyl-acyl carrier protein
synthase. A target for the antibiotic thiolactomycin. J. Biol. Chem. 264: 7624–7629.
447
448
449
3. Bergler H, Wallner P, Ebeling A, Leitinger B, Fuchsbichler S, Aschauer H, Kollenz G,
Högenauer G, Turnowsky F. 1994. Protein EnvM is the NADH-dependent enoyl-ACP
reductase (FabI) of Escherichia coli. J. Biol. Chem. 269:5493–5496.
450
451
4. Heath RJ, Su N, Murphy CK, Rock CO. 2000. The enoyl-[acyl-carrier-protein] reductases
FabI and FabL from Bacillus subtilis. J. Biol. Chem. 275:40128–40133.
452
453
5. Massengo-Tiassé RP, Cronan JE. 2008. Vibrio cholerae FabV defines a new class of enoyl-
acyl carrier protein reductase. J. Biol. Chem. 283:1308–1316.
20

454
455
6. Heath RJ, Rock CO. 2000b. Microbiology: A triclosan-resistant bacterial enzyme. Nature
406:145–146.
456
457
7. White SW, Zheng J, Zhang YM, Rock CO. 2005. The structural biology of type II fatty acid
biosynthesis. Annu. Rev. Biochem. 74:791–831.
458
459
8. Joernvall H, Persson B, Krook M, Atrian S, Gonzalez-Duarte R, Jeffery J, Ghosh D. 1995.
Short-chain dehydrogenases/reductases (SDR). Biochemistry 34:6003–6013.
460
461
9. Kallberg Y, Oppermann U, Jörnvall H, Persson B. 2002. Short‐chain
dehydrogenases/reductases (SDRs). Eur. J. Biochem. 269:4409–4417.
462
463
464
10. Persson B, Kallberg Y, Oppermann U, Jörnvall H. 2003. Coenzyme-based functional
assignments of short-chain dehydrogenases/reductases (SDRs). Chem. Biol. Interact.
143:271–278.
465
466
11. Heath RJ, Rubin JR, Holland DR, Zhang E, Snow ME, Rock CO. 1999. Mechanism of
triclosan inhibition of bacterial fatty acid synthesis. J. Biol. Chem. 274:11110–11114.
467
468
12. Halden RU. 2014. On the need and speed of regulating triclosan and triclocarban in the
United States. Environ. Sci. Technol. 48:3603–3611.
469
470
471
472
13. Yueh MF, Tukey RH. 2016. Triclosan: a widespread environmental toxicant with many
biological effects. Annu. Rev. Pharmacol. Toxicol. 56:251–272.
14. Saleh S, Haddadin RN, Baillie S, Collier PJ. 2011. Triclosan–an update. Lett. Appl.
Microbiol. 52:87–95.
21

473
474
475
15. Yazdankhah SP, Scheie AA, Høiby EA, Lunestad BT, Heir E, Fotland TØ, Naterstad K,
Kruse H. 2006. Triclosan and antimicrobial resistance in bacteria: an overview. Microb.
Drug Resist. 12:83–90.
476
477
478
16. McMurry LM, McDermott PF, Levy SB. 1999. Genetic evidence that InhA of
Mycobacterium smegmatis is a target for triclosan. Antimicrob. Agents Chemother. 43:711–
713.
479
480
481
482
17. Russell AD. 2004. Whither triclosan? J. Antimicrob. Chemother. 53:693–695.
18. McMurry LM, Oethinger M, Levy SB. 1998b. Overexpression of marA, soxS, or acrAB
produces resistance to triclosan in laboratory and clinical strains of Escherichia coli. FEMS
Microbiol. Lett. 166:305–309.
483
484
19. Lee DG, Zhao F, Rezenom YH, Russell DH, Chu KH. 2012. Biodegradation of triclosan by a
wastewater microorganism. Water research 6:4226-4234.
485
486
487
20. Khan R, Kong HG, Jung YH, Choi J, Baek KY, Hwang EC, Lee SW. 2016. Triclosan
resistome from metagenome reveals diverse enoyl acyl carrier protein reductases and
selective enrichment of triclosan resistance genes. Sci Rep. 6:1–13.
488
489
490
21. Fernando DM, Xu W, Loewen PC, Zhane GG, Kumar A. 2014. Triclosan can select for an
AdeIJK-overexpressing mutant of Acinetobacter baumannii ATCC 17978 that displays
reduced susceptibility to multiple antibiotics. Antimicrob. Agents Chemother. 58:6424–6431.
491
492
22. Schweizer HP. 2001. Triclosan: a widely used biocide and its link to antibiotics. FEMS
Microbiol. Lett. 202:1–7.
22

493
494
495
496
497
23. Chuanchuen R, Beinlich K, Hoang TT, Becher A, Karkhoff-Schweizer RR, Schweizer HP.
2001. Cross-resistance between triclosan and antibiotics in Pseudomonas aeruginosa is
mediated by multidrug efflux pumps: exposure of a susceptible mutant strain to triclosan
selects nfxB mutants overexpressing MexCD-OprJ. Antimicrob. Agents Chemother. 45:428–
432.
498
499
500
24. Bailey AM, Paulsen IT, Piddock LJ. 2008. RamA confers multidrug resistance in Salmonella
enterica via increased expression of acrB, which is inhibited by chlorpromazine. Antimicrob.
Agents Chemother. 52:3604–3611.
501
502
25. Yu BJ, Kim JA, Pan JG. 2010. Signature gene expression profile of triclosan-resistant
Escherichia coli. J. Antimicrob. Chemother. 65:1171–1177.
503
504
26. McMurry LM, Oethinger M, Levy SB. 1998a. Triclosan targets lipid synthesis. Nature
394:531–532.
505
506
507
508
27. Fan F, Yan K, Wallis NG, Reed S, Moore TD, Rittenhouse SF, DeWolf WE, Huang J,
McDevitt D, Miller WH, Seefeld MA. 2002. Defining and combating the mechanisms of
triclosan resistance in clinical isolates of Staphylococcus aureus. Antimicrob. Agents
Chemother. 46:3343–3347.
509
510
28. Stewart MJ, Parikh S, Xiao G, Tonge PJ, Kisker C. 1999. Structural basis and mechanism
of enoyl reductase inhibition by triclosan. J. Mol. Biol. 290:859–865.
511
512
29. Campbell JW, Cronan Jr JE. 2001. Bacterial fatty acid biosynthesis: targets for antibacterial
drug discovery. Annu. Rev. Microbiol. 55:305–332.
23

513
514
515
516
517
30. Khan R, Roy N, Choi K, Lee SW. 2018. Distribution of triclosan-resistant genes in major
pathogenic microorganisms revealed by metagenome and genome-wide analysis. PLoS ONE.
13:e0192277.
31. Sambrook J, Fritsch E, Maniatis T. 1989. Molecular Cloning: a Laboratory Manual, 4th ed,
vol 1. Cold Spring Harbor Laboratory Press, New York.
518
519
520
521
522
523
524
525
526
527
528
529
530
32. Sikic K, Carug O. 2010. Protein sequence redundancy reduction: comparison of various
method. Bioinformation 5:234–239.
33. Tamura K, Stecher G, Peterson D, Filipski A, Kumar S. 2013. MEGA6: molecular
evolutionary genetics analysis version 6.0. Mol. Biol. Evol. 30:2725–2729.
34. Edgar RC. 2004. MUSCLE: Multiple sequence alignment with high accuracy and high
throughput. Nucleic. Acids. Res. 32:1792–1797.
35. Yao J, Abdelrahman YM, Robertson RM, Cox JV, Belland RJ, White SW, Rock CO. 2014.
Type II fatty acid synthesis is essential for the replication of Chlamydia trachomatis. J. Biol.
Chem. 289:22365–22376.
36. Yoshimoto T, Higashi H, Kanatani A, Lin XS, Nagai H, Oyama H, Kurazono K, Tsuru D.
1991. Cloning and sequencing of the 7 alpha-hydroxysteroid dehydrogenase gene from
Escherichia coli HB101 and characterization of the expressed enzyme. J. Bacteriol.
173:2173–2179.
531
532
37. Egan AF, Russell RR. 1973. Conditional mutations affecting the cell envelope of Escherichia
coli K-12. Genet Res. 21:139–152.
533
534
38. Chothia C, Lesk AM. 1986. The relation between the divergence of sequence and structure in
proteins. EMBO J. 5:823–826.
24

535
536
537
39. Huang J, Rauscher S, Nawrocki G, Ran T, Feig M, de Groot, BL, Grubmuller H, MacKerell
AD Jr. 2017. CHARMM36m: an improved force field for folded and intrinsically disordered
proteins. Nat. Methods 14:71–73.
538
539
540
40. Heath RJ, Rock CO. 1995. Enoyl-acyl carrier protein reductase (fabI) plays a determinant
role in completing cycles of fatty acid elongation in Escherichia coli. J. Biol. Chem.
270:26538–26542.
541
542
41. Heath RJ, Rock CO. 1996. Regulation of fatty acid elongation and initiation by acyl-acyl
carrier protein in Escherichia coli. J. Biol. Chem. 271:1833–1836.
543
544
545
546
42. Ward WH, Holdgate GA, Rowsell S, McLean EG, Pauptit RA, Clayton E, Nichols WW,
Colls JG, Minshull CA, Jude DA, Mistry A. 1999. Kinetic and structural characteristics of
the inhibition of enoyl (acyl carrier protein) reductase by triclosan. Biochemistry 38:12514–
12525.
547
548
549
550
43. Basso LA, Zheng R, Musser JM, Jacobs WR Jr, Blanchard JS. 1998. Mechanisms of
isoniazid resistance in Mycobacterium tuberculosis: enzymatic characterization of enoyl
reductase mutants identified in isoniazid-resistant clinical isolates. J. Infect. Dis. 178:769–
775.
551
552
553
44. Xu H, Sullivan TJ, Sekiguchi JI, Kirikae T, Ojima I, Stratton CF, Mao W, Rock FL, Alley
MR, Johnson F, Walker SG. 2008. Mechanism and inhibition of saFabI, the enoyl reductase
from Staphylococcus aureus. Biochemistry 47:4228–42236.
554
555
45. Ramachandran GN, Ramakrishnan C, Sasisekharan V. 1963. Stereochemistry of polypeptide
chain configurations. J. Mol. Biol. 7: 95–99.
25

556
557
46. Wiederstein M, Sippl MJ. 2007. ProSA-web: interactive web service for the recognition of
errors in three-dimensional structures of proteins. Nucleic Acids Res. 35:407–410.
558
559
47. Kim KH, Ha BH, Kim SJ, Hong SK, Hwang KY, Kim EE. 2011. Crystal structures of enoyl-
ACP reductases I (FabI) and III (FabL) from B. subtilis. J. Mol. Biol. 406:403–415.
560
561
48. Rossmann MG, Moras D, Olsen KW. 1974. Chemical and biological evolution of a
nucleotide-binding protein. Nature 250:194–199.
562
563
564
565
566
567
568
569
49. Seefeld MA, Miller WH, Newlander KA, Burgess WJ, DeWolf WE, Elkins PA, Head MS,
Jakas DR, Janson CA, Keller PM, Manley PJ. 2003. Indole naphthyridinones as inhibitors of
bacterial enoyl-ACP reductases FabI and FabK. J. Med. Chem. 46:1627–1635.
50. Kavanagh KL, Jörnvall H, Persson B, Oppermann U. 2008. Medium-and short-chain
dehydrogenase/reductase gene and protein families. Cell Mol. Life Sci. 65:3895.
51. Sivaraman S, Sullivan TJ, Johnson F, Novichenok P, Cui G, Simmerling C, Tonge PJ. 2004.
Inhibition of the bacterial enoyl reductase FabI by triclosan: a structure− reactivity analysis
of FabI inhibition by triclosan analogues. J. Med. Chem. 47:509–518.
570
571
572
52. Qiu X, Abdel‐Meguid SS, Janson CA, Court RI, Smyth MG, Payne DJ. 1999. Molecular
basis for triclosan activity involves a flipping loop in the active site. Protein Sci. 8:2529–
2532.
573
574
26

575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
Figure Legends
Fig. 1. Bacterial type II fatty acid synthesis (FASII) cycle. (A) Generalized FASII cycle; Fatty
acid biosynthesis is initiated by 3-oxoacyl-acyl carrier protein (ACP) synthase (FabH), which
links malonyl-ACP with either acetyl-CoA or 2-methylbutyryl-CoA. The resulting β-ketoacyl-
ACP is further reduced by 3-oxoacyl-ACP reductase (FabG), resulting in β-hydroxybutyryl-ACP,
which is further dehydrated by 3-hydroxyacyl-ACP dehydratase (FabZ) to produce crotonyl-
ACP. The last step in the cycle is the conversion of crotonyl-ACP to acyl-ACP by enoyl-ACP
reductase (ENR), which is a target for triclosan activity. (B) ENRs catalyze the reduction
reaction.
Fig. 2. Phylogenetic analysis of FabL2 ENR and its homologs. Maximum likelihood analysis
was performed with well-characterized 7-AHSDH, FabL, FabV, FabI, and FabK (in bold) and
their homologs, with sequence identity >50% using the Uniref50 database. Bootstrap values are
shown for each node with >50% support in a bootstrap analysis of 500 replicates. Scale bar
represents 0.2 estimated amino acid substitutions per residue.
Fig. 3. Biochemical analysis of the metagenomic ENR FabL2. Initial velocities were
determined in triplicates as a function of (A) NADPH concentration and (B) crotonyl-CoA
concentration. Data were fitted to the Michaelis–Menten nonlinear regression equation using
GraphPad Prism version 7. The fitted line and Km values are shown.
Fig. 4. Sequence alignment of FabL2, mutant FabL2 (mFabL2), and template structure.
(A) Sequence alignment of FabL2 and template structure (PDB ID: 3OID), which is the
27

598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
crystallographic structure of Bacillus subtilis FabL. The extrapolated mismatched six amino
acids comprising the highly flexible loop of FabL2 are shown in red boxes, and the highly
conserved catalytic active site residues are shown in magenta boxes. (B) Sequence alignment of
mFabL2 and template structure. (C) Complementation analysis of m-FabL2 ENR. Each plate has
been divided into three sections: 1, JP111 with pGEM-T Easy only; 2, JP1111 carrying E. coli
FabI in pGEM-T Easy; 3, JP1111 carrying m-FabL2 in pGEM-T Easy. Plates were incubated at
30°C and 42°C for 48 hours.
Fig. 5. Homology model and docking of FabL2, mFabL2 and triclosan (TCL). (A)
Molecular dynamic optimized model of FabL2. The α-helices and β-strands are labeled and the
extrapolated loop of 7-AHSDH is shown in orange. NDAPH is shown as a stick model. (B)
FabL2-TCL complex after docking. TCL is shown as a stick model in magenta; it is bound at the
rim region of the tunnel leading to the catalytic site. The size of the tunnel opening is determined
by the loop (orange). The putative catalytic site of FabL2 is shown as a sphere (light blue). (C)
Two-dimensional (2D) representation of molecular interactions between FabL2 and TCL. Two
H-bonds between TCL and the Arg98 residue of FabL2 are shown. (D) Molecular dynamic
optimized model of mFabL2. NDAPH is shown as a stick model. (E) mFabL2-TCL complex
after docking. TCL is shown as a stick model (magenta), and the putative catalytic site is shown
as a sphere (light blue). (F) Two-dimensional (2D) representation of molecular interactions
between mFabL2 and TCL after docking. The absence of the extrapolated loop in mFabL2
significantly widened the opening of the tunnel leading to the catalytic site of mFabL2. The
hydrogen bond is shown as green dashed lines, while the corresponding residue is depicted as
green closed circles. (G) 2D representation of mFabL2 and NADPH interactions. All interactions
28

621
622
623
624
625
626
627
628
are indicated with dashed lines: H-bonds, black; salt bridge interactions, orange; and other
hydrophobic interactions, light magenta. Amino acid residues connected via H-bonds are
depicted as green closed circles. All amino acid residues are labeled with their 3-letter code,
followed by chain ID in protein structure and their respective amino acid number. (H) Root mean
square deviation of Cα-atoms of mFabL2 representing its stability during the simulation. (I)
Potential energy of the system, indicating the stability of the mFabL2-TCL complex. (J) Total
number of H-bonds between NADPH bound mFabL2 and TCL during the entire simulation
period.
29