A high-throughput screening fluorescence polarization assay for fatty acid adenylating enzymes in Mycobacterium tuberculosis

Article abstract of DOI:10.1016/j.ab.2011.06.037

Mycobacterium tuberculosis, the etiological agent of tuberculosis (TB), encodes for an astonishing 34 fatty acid adenylating enzymes (FadDs), which play key roles in lipid metabolism. FadDs involved in lipid biosynthesis are functionally nonredundant and serve to link fatty acid and polyketide synthesis to produce some of the most architecturally complex natural lipids including the essential mycolic acids as well as the virulence-conferring phthiocerol dimycocerosates, phenolic glycolipids, and mycobactins. Here we describe the systematic development and optimization of a fluorescence polarization assay to identify small molecule inhibitors as potential antitubercular agents. We fluorescently labeled a bisubstrate inhibitor to generate a fluorescent probe/tracer, which bound with a K_D of 245 nM to FadD28. Next, we evaluated assay performance by competitive binding experiments with a series of known ligands and assessed the impact of control parameters including incubation time, stability of the signal, temperature, and DMSO concentration. As a final level of validation the LOPAC1280 library was screened in a 384-well plate format and the assay performed with a Z-factor of 0.75, demonstrating its readiness for high-throughput screening.

Full text of DOI:10.1016/j.ab.2011.06.037

Analytical Biochemistry 417 (2011) 264–273
Contents lists available at ScienceDirect
Analytical Biochemistry
journal homepage: www.elsevier.com/locate/yabio
A high-throughput screening fluorescence polarization assay for fatty acid
adenylating enzymes in Mycobacterium tuberculosis
⇑
Kimberly D. Grimes, Courtney C. Aldrich
Center for Drug Design, University of Minnesota, 516 Delaware St. SE, Minneapolis, MN, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 21 May 2011
Received in revised form 24 June 2011
Accepted 28 June 2011
Available online 2 July 2011
Mycobacterium tuberculosis, the etiological agent of tuberculosis (TB), encodes for an astonishing 34 fatty
acid adenylating enzymes (FadDs), which play key roles in lipid metabolism. FadDs involved in lipid bio-
synthesis are functionally nonredundant and serve to link fatty acid and polyketide synthesis to produce
some of the most architecturally complex natural lipids including the essential mycolic acids as well as
the virulence-conferring phthiocerol dimycocerosates, phenolic glycolipids, and mycobactins. Here we
describe the systematic development and optimization of a fluorescence polarization assay to identify
small molecule inhibitors as potential antitubercular agents. We fluorescently labeled a bisubstrate
inhibitor to generate a fluorescent probe/tracer, which bound with a KD of 245 nM to FadD28. Next,
we evaluated assay performance by competitive binding experiments with a series of known ligands
and assessed the impact of control parameters including incubation time, stability of the signal, temper-
ature, and DMSO concentration. As a final level of validation the LOPAC1280 library was screened in a
384-well plate format and the assay performed with a Z-factor of 0.75, demonstrating its readiness for
high-throughput screening.
Keywords:
Mycobacterium tuberculosis
Adenylation
Adenylate-forming enzymes
Fluorescence polarization
High throughput screening
Ó 2011 Elsevier Inc. All rights reserved.
Tuberculosis (TB)1 is an infectious disease caused by the acid-fast
gram-positive bacillus Mycobacterium tuberculosis (Mtb) that is read-
ily spread by aerosolized droplets from coughing or sneezing of an
actively infected individual [1]. Improvements in public health and
ultimately the discovery of effective chemotherapeutic agents in
the middle of last century led to a drastic reduction of TB mortality
in the industrialized world. However, TB has continued to rage in the
developing world and now infects approximately 2 billion individu-
als resulting in nearly two million deaths annually, making TB the
second leading cause of infectious disease mortality superseded only
by HIV/AIDS [1]. The emergence of multidrug resistant (MDR-TB)
strains, and more recently, extensively drug resistant (XDR-TB)
strains that are virtually untreatable, demands the development of
new TB drugs.
Mycobacteria contain a unique cell envelope, composed of many
novel lipids, which provides a permeability barrier that shields the
bacterium from environmental stress, provides intrinsic resistance
to chemotherapeutic agents, and plays a key role in persistence [2–
5]. Lipids in the mycobacterial envelope include the covalently at-
tached mycolic acids that are essential as well as several noncova-
lently associated lipids such as the phthiocerol dimycocerosates
(PDIMs), phenolic glycolipids (PGLs), sulfolipids (SLs), and myco-
bactins (MBTs) that are critical for virulence (Fig. 1) [6–9]. Fatty acid
degradation in mycobacteria is also believed to be important and
several lines of evidence support the hypothesis that mycobacteria
are primarily lipolytic in vivo, deriving their nutrients by break-
down of host fatty acids [10,11].
The FadD enzymes are found in both lipid biosynthetic as well
as catabolic pathways [12]. M. tuberculosis encodes for an astonish-
ing 34 fadDs, by comparison, Escherichia coli encodes a single fadD
[13]. The large number of fadDs found in Mtb highlights the impor-
tance of lipid metabolism in this organism. FadDs catalyze a two-
step reaction. In the first step, the FadD binds a substrate alkanoic
acid and ATP, and then catalyzes their condensation to form an
intermediate acyl-adenylate with the release of pyrophosphate.
In a second step, the FadD binds an acceptor molecule, which can
be either a molecule of coenzyme A (CoA) or an acyl carrier protein
⇑
Corresponding author. Fax: +1 612 626 8154.
E-mail address: aldri015@umn.edu (C.C. Aldrich).
Abbreviations used: ACP, acyl carrier protein; Boc, tert-butoxycarbonyl; DME, 1,2-
1
dimethoxyethane; DMF, N,N-dimethylformamide; FAAL, fatty acyl-AMP ligase; FACL,
fatty acyl CoA ligase; FadD, fatty acid adenylating enzyme; FP, fluorescence
polarization; HTS, high-throughput screening; ITC, isothermal titration calorimetry;
MBT, mycobactin; MDR-TB, multidrug resistant tuberculosis; MesG, 7-methylthiogu-
anosine; Mtb, Mycobacterium tuberculosis; NHS, N-hydroxylsuccimidyl; PDIM,
phthiocerol dimycocerosate; PGL, phenolic glycolipid; PKS, polyketide synthase;
pTsOH, p-toluenesulfonic acid; SL, sulfolipid; TAMRA, 5-carboxytetramethylrhod-
amine; TB, tuberculosis; TCEP, tris(2-carboxyethyl)phosphine; TFA, trifluoroacetic
acid; XDR-TB, extensively drug resistant tuberculosis.
(ACP) domain of
a polyketide synthase (PKS) enzyme, then
catalyzes the transfer of the acyl moiety of the acyl-adenylate onto
0003-2697/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.ab.2011.06.037

Fluorescence polarization assay for FadD28 / K.D. Grimes, C.C. Aldrich / Anal. Biochem. 417 (2011) 264–273
265
Phthiocerol dimycocerosate A
A
B
2
Phenolic Glycolipid (PGL, )
1
(PDIM A,
)
O
18
O
O
OMe
O
O
O
O
OMe
HO
O
O
O
O
O
O
HO
O
O
OH
O
O
O
O
C
O
N
OH
O
OH
H
N
OH
O
N
O
N
N
H
O
O
3
Mycobactins (MBT,
)
O
O
OH
D
O
HO
HO
O3SO
O
O
O
OH
O
O
OH
O
O
Sulfolipid-1 (SL-1, 4)
O
E
H
H
OH
H
H
O
O
α-Mycolic acid (α-MA, 5)
Fig.1. Unique lipids found in the Mycobacterium tuberculosis cell envelope. The FadDs are involved in the biosynthesis of the lipid portions in each of these molecules as
highlighted in blue. (A) Phthiocerol dimycocerosate A (PDIM A, 1) biosynthesis requires FadD26 and FadD28; (B) phenolic glycolipids (PGLs, 2) require FadD22, FadD28, and
FadD29; (C) the mycobactins (MBTs, 3) utilize FadD33; (D) sulfolipids, represented by SL-1 4, require FadD23; and (E) the mycolic acids, represented
5), employ FadD32.
a-mycolic acid (a-MA,
O
R
AMP
Lipid Catabolism
SCoA
NH2
N
CoASH
ATP
OH
8
N
N
O
O
O O
P
FACL
R
R
N
O
O
(b)
O
ACP
(a)
PPi
10
6
7
FAAL
HO
OH
ACP
S
Lipid Biosynthesis
O
AMP
SH
9
R
ACP
SH
ACP
O
O P
O-
O
O
O
SH
N
H
N
H
OH
Fig.2. Mechanism of fatty acid adenylating enzymes (FadDs) in Mycobacterium tuberculosis. FadDs catalyze a two-step reaction: (a) the first half-reactions of both FACL and
FAAL classes of FadD are identical and involve the formation of an acyl-adenylate intermediate 7 from a fatty acid 6 and ATP with the release of pyrophosphate; (b) the second
half-reaction for the FACL and FAAL classes of FadD diverges and involves the transfer of the acyl group of the acyl-adenylate 7 onto the nucleophilic thiol of CoA for FACLs, the
first step in lipid catabolism or the thiolation domain of an acyl carrier protein (ACP), which is the first step in fatty acid biosynthesis.
the nucleophilic sulfur atom of the acceptor molecule resulting in a
CoA thioester or an acylated-ACP product (Fig. 2) [12,14].
As a result of this functional dichotomy, the FadDs in Mtb are
grouped into two classes: fatty acyl-CoA ligases (FACLs) involved
in fatty acid catabolism and long chain fatty acyl-AMP ligases
(FAALs) involved in fatty acid biosynthesis [12]. The precise
biochemical roles of the 20 annotated FACLs are largely unknown:
FadD6, FadD13, FadD15, FadD17, and FadD19 have been biochem-
ically characterized as CoA ligases, but the native substrates for
these enzymes have not been identified [12,15,16]. Transposon

266
Fluorescence polarization assay for FadD28 / K.D. Grimes, C.C. Aldrich / Anal. Biochem. 417 (2011) 264–273
mutagenesis suggested the FACLs were nonessential, which may be
due to functional redundancy [17]. Indeed FadD6, FadD15, and
FadD19 were shown to possess a remarkably broad substrate spec-
ificity [15]. By contrast, the FAAL class of FadDs appears to be func-
tionally nonredundant and serves to link fatty acid and polyketide
synthesis in mycobacteria [12,18]. FadD32, for example, is required
for mycolic acid biosynthesis and targeted genetic disruption con-
firmed its essentiality [19–21]. FadD33 is responsible for attach-
ment of the lipid moiety onto the mycobactins [22–24]. FadD26
and FadD28 are required for synthesis of the phthiocerol and
mycocerosatic acids in the PDIMs [25–27]. FadD22 in conjunction
with FadD29 is required for assembly of the phenolphthiocerol li-
pid in the PGLs [27,28]. The sulfolipids use FadD23 for biosynthesis
of the phthioceranic acid and two hydroxyphthioceranic acid
groups [29].
The identification of specific small molecule inhibitors against
each class of FadDs or selective inhibitors of an individual FadD
is expected to help decipher the functional role that the FadDs play
in lipid metabolism and could additionally lead to the develop-
ment of a new class of antitubercular agents. Simple bisubstrate
inhibitors of the FadDs have been described that serve as useful
tool compounds, but these inhibitors possess only modest potency,
display little selectivity, and do not represent useful drug-like leads
[14,21]. High-throughput screening represents an alternate meth-
od for identifying potential lead compounds with more chemically
tractable scaffolds.
We recently reported the development of a coupled steady-
state kinetic assay for FadDs employing hydroxylamine as a surro-
gate acceptor molecule, which led to hydroxamate products [30].
The pyrophosphate generated in the first half-reaction catalyzed
by the FadD was measured using pyrophosphatase and purine
nucleoside phosphorylase in conjunction with the chromogenic
product 7-methylthioguanosine (MesG). While this assay repre-
sents an excellent functional assay for secondary screening, it is
not suitable for HTS due to the requirement for two coupling en-
zymes, potential interference caused by the low wavelength of
detection, and low activity of most FadDs. Instead, we elected to
design a fluorescent polarization (FP) displacement assay to iden-
tify active-site-directed FadD inhibitors.
assay parameters including incubation time, temperature, and
DMSO concentration. The performance of the assay was then eval-
uated in a 384-well plate format against the LOPAC1280 library.
Materials and methods
General methods, reagents, procedures
All commercial reagents (Sigma–Aldrich, Fisher, Fluka) were
used as provided. 5-Carboxytetramethylrhodamine N-hydroxy-
succinimide ester was obtained from Berry & Associates (Dexter,
MI). Dichlorobis(triphenylphospine)palladium (II) and copper io-
dide were purchased from Strem Chemicals (Newburyport, MA).
Sulfamoyl chloride was prepared by the method of Heacock with-
out recrystallization [35]. 2-Iodoadenosine 12 was prepared in
three steps from guanosine using the method of Matsuda et al.
[36]. tert-Butyl (2-(2-(2-hydroxyethoxy)ethoxy)ethyl)carbamate
was synthesized in two steps from commercially available 2-[2-
(2-chloroethoxy)ethoxy]ethanol [37,38]. An anhydrous solvent
dispensing system (J.C. Meyer) using two packed columns of neu-
tral alumina was used for drying N,N-dimethylformamide (DMF)
and the solvent was dispensed under Argon. Anhydrous dime-
thoxyethane (DME) and acetonitrile (MeCN) were used as pro-
vided. All reactions were performed under an inert atmosphere
of dry argon in oven-dried (150 °C) glassware. Flash chromatogra-
phy was performed on an ISCO Combiflash Companion purification
system with prepacked silica gel cartridges and the indicated sol-
vent system. 1H and 13C NMR experiments were recorded on a Var-
ian 600 MHz spectrometer. Proton chemical shifts are reported in
ppm from an internal standard of residual methanol (3.31 ppm)
or chloroform (7.21 ppm). Carbon chemical shifts are reported
using an internal standard of residual methanol (49.1 ppm) or
chloroform (77.2 ppm). Proton chemical data are reported as
follows: chemical shift, multiplicity (s = singlet, d = doublet,
t = triplet, q = quartet, p = pentet, m = multiplet, br = broad, ovlp =
overlapping), coupling constant, and integration. 13C NMR data
for probe 11 were obtained by HSQC and HMBC experiments.
High-resolution mass spectra were acquired on an Agilent TOF II
TOF/MS instrument equipped with an ESI or APCI interface. Semi-
Fluorescence polarization assays have been widely used for
high-throughput screening due to their operational simplicity
and robust performance [31–33]. FP displacement assays involve
displacement of a fluorescently labeled ligand, also referred to as
a tracer, by a small molecule from a macromolecular receptor,
which is usually a protein. The degree of polarization of the fluo-
rescent ligand is related to its rotational correlation time and hence
molecular mass. When a fluorescent ligand is bound to a high
molecular weight receptor such as a protein, it tumbles slowly
and the emitted light remains largely polarized. However, when
a fluorescent ligand is displaced into solution, by a competitive
ligand, the fluorescent ligand tumbles rapidly and the emitted light
is depolarized. For analysis of ligand dissociation constants, anisot-
ropy is more convenient to use than polarization values; conse-
quently we will use anisotropy for data analysis rather than
polarization in the FP assay [33].
Herein, we describe the development of a fluorescence polariza-
tion displacement assay for the discovery of inhibitors of the FAAL
class of FadDs. We selected FadD28 as a model FadD since this
enzyme is required for biosynthesis of the virulence conferring
PDIM and DIM lipids, FadD28 can be readily overexpressed and
purified from E. coli, and FadD28 is the only FadD from Mtb that
has been structurally characterized [14,34]. We first prepared a
fluorescently labeled active-site-directed FP probe that exhibited
nanomolar affinity for FadD28. We then performed competitive
displacement experiments with a series of known ligands of vary-
ing potency to validate the assay. Next, we examined the impact of
preparative HPLC was performed on a Phenomenex Gemini 10 lm
C18 110A (250 ꢀ 10.0 mm) column operating at 4.0 mL/min with
detection at 254 nm.
Cloning, expression, and purification of FadD28
FadD28 was cloned, overproduced, and purified as previously
reported to afford approximately 90 mg/L of culture [30].
Synthesis of fluorescent probe 11
tert-Butyl 2-{2-[2-(prop-2-yn-1-yloxy)ethoxy]ethoxy}ethylcar-
bamate (13): to
a solution of tert-butyl 2-[2-(2-hydroxyeth-
oxy)ethoxy]ethylcarbamate (920 mg, 3.69 mmol, 1.0 eq) in THF
(40 mL) was added sodium hydride (60% dispersion in oil,
221 mg, 5.54 mmol, 1.5 eq) in one portion at 0 °C. After 0.5 h, prop-
argyl bromide (80% in toluene, 411 lL, 3.69 mmol, 1.0 eq) was
added and the reaction mixture was warmed to 23 °C and stirred
for 16 h. The reaction was quenched with saturated aqueous NH4Cl
(10 mL) and extracted with EtOAc (3 ꢀ 25 mL). The combined ex-
tracts were washed with saturated aqueous NaCl (50 mL), dried
(MgSO4), and concentrated. Purification by flash chromatography
(0–50% hexanes/EtOAc) afforded the title compound (525 mg,
50%) as a colorless oil: Rf 0.62 (3:2 EtOAc/hexane); 1H NMR
(600 MHz, CDCl3): d 1.45 (s, 9H), 2.44 (t, J = 2.4 Hz, 1H), 3.32–
3.33 (m, 2H), 3.55 (t, J = 4.8 Hz, 2H), 3.63–3.64 (m, 2H), 3.65–3.67

Fluorescence polarization assay for FadD28 / K.D. Grimes, C.C. Aldrich / Anal. Biochem. 417 (2011) 264–273
267
(m, 2H), 3.69–3.70 (m, 2H), 3.72–3.73 (m, 2H), 4.23 (d, J = 2.4 Hz,
2H), 5.03 (br s, 1H); 13C NMR (150 MHz, CDCl3): d 28.3, 40.3,
58.3, 68.9, 70.08, 70.13, 70.3, 70.4, 74.7, 78.9, 79.5, 155.9; HRMS
(APCI+) calcd for C14H26NO5 [M+H]+ 288.1805, found 288.1788 (er-
ror 5.9 ppm).
115.8, 120.3, 142.8, 146.3, 150.3, 156.8, 158.5; HRMS (ESIꢁ) calcd
for
C
27H40N7O11
S
[MꢁH]ꢁ 670.2512, found 670.2533 (error
3.1 ppm).
2-[3-(2-{2-[2-(tert-Butoxycarbonylamino)ethoxy]ethoxy}ethoxy)-
prop-1-ynyl]-20,30-O-isopropylidene-50-O-[N-(n-tetradecanoyl)sul-
famoyl]adenosine triethylammonium salt (18). To a solution of 16
(40 mg, 0.06 mmol, 1.0 eq) in DMF (5 mL) at 0 °C was added N-
hydroxysuccinimidyl tetradecanoate 17 (29 mg, 0.09 mmol,
1.5 eq) and Cs2CO3 (58 mg, 0.18 mmol, 3.0 eq). The reaction was
stirred for 16 h at 23 °C and then filtered to remove salts and the fil-
trate was concentrated. Purification by flash chromatography
(90:10:0.5 EtOAc/MeOH/NEt3) afforded the title compound
(45 mg, 77%) as a yellow oil: Rf 0.4 (9:1 EtOAc/MeOH); 1H NMR
(600 MHz, CD3OD) d 0.89 (t, J = 6.6 Hz, 3H), 1.25–1.31 (m, 29H),
1.39 (s, 3H), 1.42 (s, 9H), 1.56 (p, J = 7.2 Hz, 2H), 1.62 (s, 3H), 2.17
(t, J = 7.2 Hz, 2H), 3.18–3.23 (m, 8H), 3.51 (t, J = 5.4 Hz, 2H), 3.61–
3.62 (m, 2H), 3.66–3.67 (m, 2H), 3.71–3.72 (m, 2H), 3.79–3.81 (m,
2H), 4.23–4.28 (m, 2H), 4.46 (s, 2H), 4.53 (br s, 1H), 5.10 (dd,
J = 5.7, 1.8 Hz, 1H), 5.30 (dd, J = 6.0, 3.0 Hz, 1H), 6.22 (d, J = 3.0 Hz,
1H), 8.50 (s, 1H); 13C NMR (150 MHz, CD3OD) d 9.3, 14.6, 23.8,
25.8, 27.5, 27.7, 28.9, 30.4, 30.58, 30.63, 30.69, 30.74, 30.81,
30.87, 30.88, 33.2, 40.3, 41.4, 47.9, 59.6, 69.9, 70.6, 71.2, 71.4,
71.5, 71.6, 80.1, 82.7, 83.3, 85.8, 85.9, 86.5, 91.6, 115.5, 119.8,
142.3, 147.3, 150.8, 157.2, 158.5, 182.9; HRMS (ESIꢁ) calcd for
2-[3-(2-{2-[2-(tert-Butoxycarbonylamino)ethoxy]ethoxy}eth-
oxy)prop-1-ynyl]adenosine (14): compound 13 (431 mg,
1.5 mmol, 3.0 eq) was added dropwise via a syringe pump over
0.5 h to a solution of 2-iodoadenosine 12 (197 mg, 0.5 mmol,
1.0 eq), dichlorobis(triphenylphospine)palladium (II) (17.5 mg,
0.025 mmol, 0.05 eq), and copper iodide (7.1 mg, 0.038 mmol,
0.076 eq) in CH3CN/Et3N (1:1 v/v, 10 mL). The reaction was stir-
red 3 h at 23 °C, and then concentrated in vacuo. Purification by
flash chromatography (0–10% EtOAc/MeOH) afforded the title
compound (260 mg, 94%) as a yellow oil: Rf 0.3 (8:2 EtOAc/
MeOH); 1H NMR (600 MHz, CD3OD) d 1.42 (s, 9H), 3.21–3.23
(m, 2H), 3.51 (t, J = 6.0 Hz, 2H), 3.61–3.63 (m, 2H), 3.65–3.67
(m, 2H) 3.70–3.72 (m, 2H), 3.75 (dd, J = 12.0, 2.4 Hz, 1H), 3.79–
3.80 (m, 2H), 3.90 (dd, J = 12.6, 2.4 Hz, 1H), 4.17 (q, J = 2.4 Hz,
1H), 4.33 (dd, J = 5.4, 3.0 Hz, 1H), 4.45 (s, 2H), 4.71 (t,
J = 6.0 Hz, 1H), 5.94 (d, J = 6.0 Hz, 1H), 6.58 (br s, 1H), 8.34 (s,
1H); 13C NMR (150 MHz, CD3OD) d 28.9, 41.4, 58.5, 63.5, 70.5,
71.2, 71.3, 71.5, 71.6, 72.6, 75.6, 80.2, 83.2, 85.9, 88.1, 91.3,
120.7, 142.9, 146.8, 150.2, 157.4, 158.5; HRMS (ESIꢁ) calcd for
C
24H35N6O9 [MꢁH]ꢁ 551.2471, found 551.2480 (error 1.6 ppm).
2-[3-(2-{2-[2-(tert-Butoxycarbonylamino)ethoxy]ethoxy}eth-
C
41H66N7O12S [MꢁH]ꢁ 880.4496, found 880.4496 (error 0 ppm).
2-[3-(2-{2-[2-({4-Carboxy-3-[6-(dimethylamino)-3-(dimethy-
oxy)prop-1-ynyl]-20,30-O-isopropylideneadenosine (15): to a solu-
tion of 14 (125 mg, 0.23 mmol, 1.0 eq) in acetone (5 mL) was
liminio)-3H-xanthen-9-yl]benzoyl}amino)ethoxy]ethoxy}ethoxy)
prop-1-ynyl]-50-O-[N-(n-tetradecanoyl)sulfamoyl]adenosine trie-
thylammonium salt (11): to 18 (10 mg, 0.01 mmol, 1.0 eq) was
added 80% TFA (1.5 mL). The reaction was stirred at 23 °C for 4 h,
and then concentrated in vacuo to remove all traces of TFA. The
added dimethoxypropane (139 lL, 1.13 mmol, 4.9 eq) and p-
toluenesulfonic acid (88 mg, 0.25 mmol, 1.09 eq). The reaction
was stirred at 23 °C for 16 h. To the resulting suspension was
added solid NaHCO3 to quench the reaction and then the acetone
was removed in vacuo. The residue was partitioned between EtOAc
(20 mL) and saturated aqueous NaHCO3. The aqueous layer was
further extracted with EtOAc (3 ꢀ 20 mL). The combined extracts
were washed with saturated aqueous NaCl (50 mL), dried (MgSO4),
and concentrated. Purification by flash chromatography (0–10%
CH2Cl2/MeOH) afforded the title compound (120 mg, 88%) as a yel-
low oil: Rf 0.5 (9:1 CH2Cl2/MeOH); 1H NMR (600 MHz, CD3OD) d
1.39 (s, 3H), 1.42 (s, 9H), 1.62 (s, 3H), 3.22 (t, J = 5.4 Hz, 2H), 3.51
(t, J = 6.0 Hz, 2H), 3.61–3.63 (m, 2H), 3.65–3.67 (m, 2H) 3.70–3.74
(m, 3H), 3.79–3.81 (m, 3H), 4.37 (q, J = 3.6 Hz, 1H), 4.46 (s, 2H),
5.04 (dd, J = 6.0, 3.0 Hz, 1H), 5.25 (dd, J = 6.0, 3.0 Hz, 1H), 6.14 (d,
J = 3.6 Hz, 1H), 8.36 (s, 1H); 13C NMR (150 MHz, CD3OD) d 25.8,
27.8, 28.9, 41.5, 59.6, 63.8, 70.6, 71.2, 71.4, 71.6, 71.7, 80.1, 83.0,
83.1, 85.5, 86.2, 88.3, 92.9, 115.4, 120.3, 142.8, 147.2, 150.3,
residue was dissolved in DMF (350 lL) and triethylamine (6.0 lL,
0.045 mmol, 4.5 eq) and 5-carboxytetramethylrhodamine N-
hydroxysuccinimide ester 19 (TAMRA, 8.0 mg, 0.015 mmol,
1.5 eq) were added. The flask was covered with foil and the mixture
was stirred at 23 °C for 5 h. The mixture was directly purified by
semipreparative HPLC (injected 5 ꢀ 70
lL) using a Phenomenex
m C18 110A (250 ꢀ 10.0 mm) column and a linear gra-
Gemini 10
l
dient of 20–100% CH3CN/50 mM triethylammonium bicarbonate
(pH 7.5) over 15 min followed by 100% CH3CN for 10 min. The
retention time of the product was 19.0 min and the appropriate
fractions were pooled and lyophilized to afford the title compound
(6.0 mg, 48%) as a magenta solid: 1H NMR (600 MHz, CD3OD) d 0.89
(t, J = 7.2 Hz, 3H), 1.25–1.31 (m, 29H), 1.60 (p, J = 7.2 Hz, 2H), 2.20 (t,
J = 7.8 Hz, 2H), 3.19 (q, J = 7.8 Hz, 6H), 3.25 (s, 6H), 3.27 (s, 6H),
3.67–3.68 (m, 4H), 3.69–3.70 (m, 6H), 3.74 (t, J = 5.4 Hz, 2H),
4.16–4.17 (m, 1H), 4.25–4.34 (ovlp m, 3H), 4.30 (ovlp s, 2H), 4.43
(t, J = 4.8 Hz, 1H), 5.84 (d, J = 5.4 Hz, 1H), 6.74 (br s, 2H), 6.87 (d,
J = 9.0 Hz, 1H), 6.92 (d, J = 9.0 Hz, 1H), 7.16 (d, J = 9.6 Hz, 1H), 7.19
(d, J = 9.0 Hz, 1H), 7.69 (d, J = 7.2 Hz, 1H), 8.12 (d, J = 7.8 Hz, 1H),
8.49 (s, 1H), 8.54 (s, 1H); 13C NMR (150 MHz, CD3OD) d 9.5, 14.6,
23.9, 27.7, 30.3 (3C), 30.6, 30.8 (2C), 30.9 (2C), 33.2, 40.5, 41.0
(2C), 41.1 (2C), 41.2, 48.0, 59.6, 68.8, 70.5, 70.8, 71.7 (2C), 71.9,
72.0, 76.4, 79.6, 83.7, 84.4, 89.3, 96.9 (2C), 114.5 (2C), 114.9 (2C),
119.6, 128.9, 129.8, 131.5, 132.4 (2C), 136.9, 138.8, 142.1, 147.6,
155.5, 158.6, 162.1, 162.3 (2C), 166.4, 169.2, 172.4, 180.1 (missing
C2 and C6 of the adenine ring, for complete assignment by COSY,
HMQC, and HMBC, see Supplementary data); HRMS (ESIꢁ) calcd
for C58H74N9O14S [MꢁH]ꢁ 1152.5081, found 1152.5014 (error
5.8 ppm).
157.3, 158.5; HRMS (ESIꢁ) calcd for
C
27H39N6O9 [MꢁH]ꢁ
591.2784, found 591.2787 (error 0.5 ppm).
2-[3-(2-{2-[2-(tert-Butoxycarbonylamino)ethoxy]ethoxy}eth-
oxy)prop-1-ynyl]-20,30-O-isopropylidene-50-O-(sulfamoyl)adeno-
sine (16): to a solution of 15 (70 mg, 0.12 mmol, 1.0 eq) in DME
(10 mL) was added sodium hydride (60% dispersion in mineral
oil, 13.0 mg, 0.33 mmol, 2.75 eq) at 0 °C. After 0.5 h, solid sulfa-
moyl chloride (20.4 mg, 0.18 mmol) was added and the reaction
was stirred for 1 h at 23 °C. The mixture was quenched with MeOH
(5 mL), and then concentrated onto Celite. Purified by flash chro-
matography (0–20% EtOAc/MeOH) afforded the title compound
(50 mg, 62%) as a yellow oil: Rf 0.75 (8:2 EtOAc/MeOH); 1H NMR
(600 MHz, CD3OD) d 1.40 (s, 3H), 1.42 (s, 9H), 1.62 (s, 3H), 3.22
(t, J = 5.4 Hz, 2H), 3.52 (t, J = 5.4 Hz, 2H), 3.61–3.62 (m, 2H),
3.65–3.66 (m, 2H) 3.71–3.72 (m, 2H) 3.79–3.80 (m, 2H), 4.28
(dd, J = 10.8, 4.8 Hz, 1H), 4.38 (dd, J = 10.5, 4.2 Hz, 1H), 4.48 (s,
2H), 4.53 (q, J = 4.8 Hz, 1H), 5.14 (dd, J = 5.4, 3.0 Hz, 1H), 5.39
(dd, J = 6.0, 3.0 Hz, 1H), 6.14 (d, J = 2.4 Hz, 1H), 8.23 (s, 1H); 13C
NMR (150 MHz, CD3OD) d 25.8, 27.6, 28.9, 41.4, 59.6, 70.1, 70.7,
71.2, 71.4, 71.5, 71.7, 80.2, 83.2, 83.8, 85.6, 85.7, 86.1, 92.0,
Equilibrium dissociation constant of probe 11
Direct binding experiments were performed to determine the
equilibrium dissociation constant (KD) between FP probe 11 and
FadD28. Threefold serial dilutions of a 3ꢀ stock solution of FadD28

268
Fluorescence polarization assay for FadD28 / K.D. Grimes, C.C. Aldrich / Anal. Biochem. 417 (2011) 264–273
(10
l
L, 75
l
M) in 1ꢀ FP buffer (30 mM TrizmaꢂHCl, pH 7.5, 1 mM
Isothermal titration calorimetry
MgCl2, 1 mM TCEP, 0.0025% Igepal CA-630) were added to 20 lL
of a 1.5X stock solution of 11 (150 nM) in 1ꢀ FP buffer in a 384-
well black plate (Corning No. 3575). The plate was shaken for 5 s
and read (kex = 530, kem = 590, emission cutoff = 590 nm, G fac-
tor = 1.104) after a 10 min incubation period at 23 °C on a Molecu-
lar Devices Spectramax M5e plate reader. The KD was determined
by fitting the experimentally observed anisotropies (AOBS) to
Eqs. (1) and (2) by nonlinear regression analysis using Mathemat-
ica 7 (Wolfram Research Inc.):
ITC experiments were carried out on a Microcal VP-ITC micro-
calorimeter (Microcal, Inc.). All measurements were taken at
20 °C in 20 mM TrizmaꢂHCl, pH 7.5, 1 mM MgCl2, and 0.0025% Ige-
pal. FadD28 was dialyzed (2 L) against the buffer described above
for 16 h at 4 °C, and all ligand solutions were prepared in the final
dialysate. In individual titrations, ligands were injected into the en-
zyme solution. Ligand and protein concentrations were 370
ligand 24 and 24 M FadD28. Titrations were carried with a stir-
ring speed of 307 rpm and 300 s interval between 10 L injections.
lM for
l
l
QFSBAB þ ð1 ꢁ FSBÞAF
The first injection was excluded from data fitting. Titrations were
run past the point of enzyme saturation to determine and correct
for heats of dilution. The experimental data were fit to a theoretical
titration curve using the Origin software package (version 7.0) pro-
vided with the instrument to afford values of KA (the association
constant in Mꢁ1), n (the number of binding sites per monomer),
AOBS
¼
ð1Þ
1 ꢁ ð1 ꢁ QÞFSB
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
KD1 þ LST þ RT ꢁ ðKD1 þ LST þ RT Þ ꢁ 4LSTRT
¼
2LST
FSB
ð2Þ
Here AOBS is the observed anisotropy, Q is the ratio of the fluorescent
intensities of the bound and free probe, FSB is the fraction of the
bound probe, AF and AB are the anisotropies of the free and bound
probe, KD1 is the probe’s dissociation constant, LST is the total probe
concentration, and RT is the concentration of FadD28.
and
DH (the binding enthalpy change in kilocalories per mole).
The affinity of the ligand for the protein is given as the dissociation
constant (KD = 1/KA). ITC experiments were performed in triplicate
and analyzed independently, and the thermodynamic values ob-
tained were averaged.
Determination of dissociation rate constant of 11ꢂFadD28
Determination of enzyme inhibition (Kiapp
)
The dissociation rate constant of 11 and FadD28 was deter-
mined by incubation of 100 nM 11 with 535 nM FadD28 in FP as-
Apparent Ki values were determined using a coupled continu-
ous assay employing FadD28, pyrophosphatase, and nucleoside
phosphorylase under initial velocity conditions as described [30].
say buffer (1ꢀ) in
a
total volume of 29
lL. Next, 1 lL of
tetradecanoyl-AMS 24 (100
lM final concentration) was added
and FP measurements were immediately acquired in the kinetic
mode and taken every 15 s until a plateau was reached. Data were
fit to the one-phase exponential decay equation.
Reactions contained 4.5 lM FadD28 in a buffer of 50 mM
TrizmaꢂHCl, pH 8.0, 2.5 mM ATP, 5 mM MgCl2, 0.5 mM DTT,
150 mM hydroxylamine, pH 7, 0.1 U nucleoside phosphorylase,
0.04 U pyrophosphatase, 0.2 mM 7-methylthioguanosine (MesG),
off t
AOBS ¼ ðAU ꢁ AFÞ10ꢁk þ AF;
ð3Þ
and 33 lM tetradecanoic acid (KM = 5.3 lM). Reactions were run
in 96-well half-area UV Star plates (Greiner) and the cleavage of
MesG was monitored at A360 on a Molecular Devices Spectramax
M5e plate reader. Kai pp values were determined by fitting the
concentration–response plots to the Morrison equation since
the inhibitors exhibited tight-binding behavior (Kai pp 6 100 ꢀ [E])
[39].
where AU is the observed anisotropy of 11 bound to 535 nM FadD28,
AF is the anisotropy of the free probe, AOBS is the observed anisot-
ropy at time t, and koff is the dissociation rate constant. All experi-
ments were performed in duplicate.
Competitive displacement experiments
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
ð½EꢃT þ ½IꢃT þ KappÞ ꢁ ð½EꢃT þ ½IꢃT þ KappÞ ꢁ 4½EꢃT ½IꢃT
For each ligand, a threefold serial dilution series was prepared
in DMSO to provide a final concentration ranging from 100 to
mi
m0
i
i
¼ 1 ꢁ
ð5Þ
2½EꢃT
0.005
l
M. The ligand solutions were added (1
lL in DMSO) to a
FP assay mixture (29
l
L) containing 11 (100 nM final concentra-
Determination of Z0 and S/N for parameter optimization
tion), FadD28 (535 nM final concentration) in FP buffer (1ꢀ) in a
384-well black plate (Corning No. 3575). The plate was shaken
for 5 s and read after a 10 min incubation period at 23 °C as de-
scribed for the direct binding experiment. The KD of each com-
pound was determined by fitting the observed anisotropies (AOBS
to Eqs. (1) and (4) by nonlinear regression analysis using Mathem-
FP assay mixture (29
centration), 11 (100 nM final concentration) in FP Buffer (1ꢀ) was
added to 48 wells (384-well black plate) containing 1.0 L (100
final volume) of tetradecanoyl-AMS 24 in DMSO to serve as a posi-
tive control and 48 wells containing 1.0 L DMSO as the negative
lL) containing FadD28 (535 nM final con-
l
lM
)
l
atica 7:
control. For FP signal stability studies, the plate was read at 10,
30, 60, 120, 180, and 1080 min at 23 °C. For thermal stability stud-
ies, the plate was incubated at 23, 25, and 28 °C for 0.5 h and then
read. The S/N and Z0 values were determined from
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
ða2 ꢁ 3bÞ cosðh=3Þ ꢁ a
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
FSB
¼
with
ð4Þ
3KD1 þ 2 ða2 ꢁ 3bÞ cosðh=3Þ ꢁ a
a ¼ KD1 þ KD2 þ LST þ LT ꢁ RT
AU ꢁ AF
b ¼ ðLT ꢁ RT ÞKD1 þ ðLST ꢁ RT ÞKD2 þ KD1KD2
c ¼ ꢁKD1KD2RT
S=N ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ð6Þ
ð7Þ
rU2
þ
rF2
2
3
ð3rU þ 3rF
AU ꢁ AF
Þ
Z0 ¼ 1 ꢁ
ꢁ2a3 þ 9ab ꢁ 27c
2
6
4
7
5
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
h ¼ arccos
3
ða2 ꢁ 3bÞ
where AU is the observed anisotropies in the absence of competitive
ligand 24, AF is the observed anisotropies in the presence of 100
Here KD2 is the compounds equilibrium dissociation constant, LT is
the respective concentration, and all other parameters are equiva-
lently defined as for Eq. (2).
lM
competitive ligand 24, and rU and rF are the respective standard
deviations.

Fluorescence polarization assay for FadD28 / K.D. Grimes, C.C. Aldrich / Anal. Biochem. 417 (2011) 264–273
269
High-throughput screening
activity – an important consideration since the mycobacterial FadD
enzymes possess low activity [31–33]. As illustrated in Fig. 3, when
the fluorescent probe 11 binds to FadD28 and is excited with
plane-polarized light the macromolecule complex tumbles slowly
in solution and the emitted light remains largely polarized. How-
ever, when the fluorescent probe 11 is displaced by a small mole-
cule ligand into solution, the fluorescent probe is able to rotate
rapidly and consequently the emitted light is depolarized relative
to the bound probe.
The LOPAC1280 (Library of Pharmacologically Active Com-
pounds, Sigma) library was screened for assay validation. The
library was formatted into four 384-well black flat bottom plates
and diluted to 1.25 mM in DMSO using a Biomek 3000 automated
workstation. These plates were used to prepare screening plates in
duplicate that contained 320 LOPAC compounds, 16 DMSO nega-
tive control wells (1
lL), and 16 wells positive control wells con-
taining 24 (1 L each, 100
l
l
M final volume) per plate. The FP
The most important factor in the development of an FP assay is
the design of the fluorescent probe, which consists of three compo-
nents: the ligand (typically active-site directed for enzyme-based
assays), linker (separates the ligand from the fluorophore), and
fluorophore (Fig. 4) [40]. 50-O-[N-(Tetradecanoyl)sulfamoyl]adeno-
sine (tetradecanoyl-AMS, 24), a mimic of the intermediate acyl-
adenylate, was selected as the ligand based on potency (vide infra)
and excellent solubility (i.e., longer chain acyl groups increase po-
tency at the expense of solubility). Ideally, the linker should be of
sufficient length to enable attachment of the fluorophore without
disrupting binding of the probe and also not have too many de-
grees of freedom, which can cause depolarization referred to as
the ‘‘propeller effect’’ [41]. A triethyleneglycol linker was expected
to provide sufficient length to access the solvent-exposed surface
of the protein. Attachment of the linker at the C-20 hydroxyl was
initially explored, but this modification substantially reduced bind-
ing affinity (not shown). We next, attempted to install the linker at
the C-2 position of the purine and found this was tolerated as de-
scribed below. When selecting an appropriate fluorophore, it is
advantageous to have a fluorophore that has absorption in a range
>500 nm to avoid interference by autofluorescent molecules that
assay mixture was prepared directly before the assay and con-
tained FP buffer (30 mM TrizmaꢂHCl, pH 7.5, 1 mM MgCl2, 1 mM
TCEP, and 0.0025% Igepal CA-630), 535 nM FadD28, and 100 nM
11. Next, 29 lL of the FP assay mixture was added to each well
and the plates were shaken for 5 s and read after a 30 min incuba-
tion period at 23 °C as described for the direct binding experiment.
The robustness of the assay was determined by calculating the Z0
value for each plate using Eq. (7). The KD’s of hit compounds (Ta-
ble 2) were determined against FadD28 using the competitive
binding assay described above.
Results and discussion
Assay design and rationale
Our goal was to develop an assay amenable to high-throughput
screening for the discovery of small molecule inhibitors of FadD28
from M. tuberculosis. The fluorescence polarization (FP) format was
selected since it is homogeneous, allows the identification of
active-site-directed inhibitors, and does not require enzyme
Fig.3. Fluorescence polarization assay. General schematic for FP assay: FadD28 is represented by the gray crescent shape while the fluorescent ligand 11 is depicted as a black
oval connected via a short linker to a pink fluorophore. (A) When the fluorescent probe 11 is bound to FadD28 and excited with plane-polarized light, the macromolecule
complex tumbles slowly and the emitted light remains largely polarized. (B) When the fluorescent probe 11 is displaced into the bulk solution, it rotates rapidly and the
emitted light is depolarized relative to the bound probe.
NH2
N
O
N
O
O
S
O
N
(
)
12
N
N
O
O
O
O
O
N
H
O H
OH
COO
Ligand
Linker
N
O
N
11
Fluorophore
Fig.4. Scaffold for fluorescent probe 11.

270
Fluorescence polarization assay for FadD28 / K.D. Grimes, C.C. Aldrich / Anal. Biochem. 417 (2011) 264–273
NH2
N
NH2
N
N
N
13
N
MeO
OMe
O
N
HO
N
NHBoc
HO
N
I
3
O
O
O
NHBoc
p-TsOH, (CH3)2CO
88%
3
Et3N, CuI, PdCl2(PPh3)2,
MeCN 94%
OH
HO
OH
HO
12
14
O
NH2
N
NH2
N
O
N
N
N
O
O
( )12
O
N
S
N
N
HO
H2N
O
N
17
NH2SO2Cl
O
O
O
O
O
NHBoc
NHBoc
DME
62%
3
3
Cs2CO3, DMF
77%
O
O
O
O
15
16
NH2
N
NH2
N
N
N
N
O
O
O
O
O O
S
S
O
N
N
( )12
( )12
N
N
O
N
O
1. 80% TFA/H2O
O
O
O
O
N
NHBoc
3
2. NEt3, DMF
(48% over two steps)
3
Et3NH
H
Et3NH
O H
OH
O
O
COO
O
O
18
11
19
N
O
N
N
O
COO
O
N
O
N
Scheme 1.
Fig.5. Direct binding of fluorescent probe 11 (100 nM) to FadD28 as measured by
fluorescence polarization. Experimental points are shown with error bars and the
line is the result of fitting data to Eqs. (1) and (2).
Fig.6. Competitive displacement of fluorescent probe 11 (100 nM) from FadD28
(535 nM) with 24. Experimental points are shown with error bars and the line is the
result of fitting data to Eqs. (1) and (3).
are abundant in a high throughput libraries [42–44]. Consequently,
we chose to use red shifted 5-carboxytetramethylrhodamine
(TAMRA). Based on these aforementioned considerations, we de-
signed ligand 11 (Fig. 4).
N-hydroxysuccimidyl (NHS) ester of tetradecanoic acid 17 medi-
ated by Cs2CO3 furnished acylsulfamate 18. Deprotection of the
acetonide and Boc groups of 18 was performed with an 80% aque-
ous TFA solution and the crude product was directly coupled with
the NHS ester of 5-carboxytetramethylrhodamine 19 to provide
fluorescent probe 11 as the triethylammonium salt in 19% yield
over six steps.
Probe synthesis
The synthesis of the fluorescent probe 11 began with the Sono-
gashira coupling of 2-iodoadenosine 12 [36] and alkyne linker 13
[37,38,45] to afford 14 [46] (Scheme 1). Protection of the 20- and
30-hydroxyls of 14 as the acetonide provided nucleoside 15, which
was sulfamoylated on the 50 position to yield 16. Acylation with the
Determination of probe KD
Direct binding experiments were performed to determine the
dissociation constant (KD) between 11 and FadD28. Titration of

Fluorescence polarization assay for FadD28 / K.D. Grimes, C.C. Aldrich / Anal. Biochem. 417 (2011) 264–273
271
Table 1
KD’s and Kapp’s of FadD28 inhibitors.
i
Kai pp
(l
M)b
Ligand
KD (l
M)a
20
21
22
23
24
>100
>100
ntc
NH2
N
N
N
O
O
S
H2N
N
O
O
HO
OH
>100
>100
NH2
N
N
N
O
O
S
O
N
N
O
O
HO
OH
1.65 0.49
0.23 0.04
0.19 0.004
NH2
N
N
N
O
O
O
O
S
N
N
O
HO
OH
7.4 0.6
NH2
N
N
N
O
O
O
S
N
N
O
O
HO
O
OH
2.5 0.2
NH2
N
N
N
O
O
S
N
N
O
O
HO
OH
a
b
c
KD values were determined using the fluorescence polarization assay with 100 nM probe 11 and 535 nM FadD28.
Kai pp values were determined using the FadD28 hydroxamate–MesG coupled assay employing 33
lM tetradecanoic acid and 4.5
lM FadD28.
Not tested.
100 nM probe 11 with FadD28 in FP assay buffer yielded a KD of
245 nM (Fig. 5) with a Q value of 0.91 [47]. Igepal CA-630, a non-
ionic detergent, was used in the assay mixture to prevent the
pronounced variation in the curvature of the meniscus in the
microplate wells observed for various protein concentrations that
affected fluorescent readings [48]. Additionally, non-ionic deter-
gents reduce the potential for false positives of aggregate-based
inhibitors from library screening [49–51]. The anisotropy of the
unbound probe was 0.06, whereas the anisotropy of the probe
bound to FadD28 was 0.25. The large anisotropy difference
between the bound and free probe provides a robust signal.
yielded a KD of 365 219 nM,
D
H of ꢁ6.9 0.5 kcal/mol, and n value
of 1.1 0.2, indicating one substrate binding site per monomer. We
also determined the apparent Ki value (Kapp) of 2.5
l
M for 24 with
i
FadD28 using a coupled kinetic assay employing saturating concen-
trations of tetradecanoic acid (Table 1). The close agreement be-
tween the KD’s of ligand 24 obtained by FP and ITC provides
confidence in the assay. We additionally evaluated a series of
ligands 20–23 (Table 1 and Fig. S1 in Supplementary data) using
the competitive displacement FP assay and the coupled kinetic as-
say [30]. The results between both assays showed excellent correla-
tion and KD and Ki values both decreased with increasing chain
length of the acyl moiety. The FP assay was unable to discriminate
the potency of ligands 23 and 24, since the range of resolvable
inhibitor potency is limited by the affinity of the fluorescent probe
[40].
Determination of KD using FP competitive binding assay
Competitive displacement assays were performed with several
ligands as a means to provide an initial level of assay validation.
Roehrl et al. suggest that the fraction of the bound probe (FSB
)
Impact of control parameters
should be greater than or equal to 0.5 to ensure sufficient signal
[47]. Consequently, we adjusted the concentration of the fluores-
cent probe 11 and FadD28 to 100 and 535 nM, respectively, to pro-
vide a FSB = 0.50. The first ligand we examined was the parent ligand
tetradecanoyl-AMS 24, from which probe 11 was designed [30].
Titration of tetradecanoyl-AMS 24 into a solution of 100 nM 11
and 535 nM FadD28 provided a displacement curve (Fig. 6), which
was analyzed as described under ‘‘Materials and Methods’’ to pro-
vide a KD of 186 4 nM. For direct comparison, we performed iso-
thermal titration calorimetry (ITC) of ligand 24 with FadD28 that
The optimization of several key factors, such as incubation time,
stability of FP signal, temperature, and DMSO concentration was
next assessed. In order to determine the necessary incubation time
required for the assay, the dissociation rate of the probe 11 with
FadD28 was measured by following the monoexponential time-
dependent decrease in anisotropy following addition of a large ex-
cess of tight-binding ligand 24 (100 lM) to prebound probe 11
(100 nM) and FadD28 (535 nM). The dissociation rate-constant koff
is 0.0148 0.0005 sꢁ1 that corresponds to a half-life of 46.9 s,

272
Fluorescence polarization assay for FadD28 / K.D. Grimes, C.C. Aldrich / Anal. Biochem. 417 (2011) 264–273
Table 2
KD values of LOPAC hits.a
Ligand
KD
2.4 0.5
(l
M)a
25
NH2
N
[Na+]3
N
N
O
P
O
-O
O
O
P
O
N
S
O
O-
O-
O H
OH
26
1.3 0.2
NH2
[Na+]4
N
N
N
O
P
O
P
O
P
-O
O
O
N
S
O
O-
O-
O-
O H
OH
27
28
6.1 2.0
7.6 2.4
O
S
OH
NH2
OH
OH
a
KD values were determined using the fluorescence polarization assay with
100 nM probe 11 and 535 nM FadD28.
25 and 2-methylthio ATP tetrasodium 26 showed full displacement
yielding KD’s of 2.4 0.5 and 1.3 0.16 lM, respectively (Table 2).
The long chain fatty acid derivatives tetradecylthioacetic acid 27
and DL-sphinganine 28 showed only partial displacement, but the
observed anisotropies could still be fit to Eqs. (1) and (4) to provide
Fig.7. Effects of incubation time (A) and temperature (B) on assay performance.
Experimental points are shown with error bars.
KD’s of 6.1 2.0 and 7.6 2.4 lM, respectively.
suggesting that a 10 min incubation time is adequate. Next, the
robustness of the assay was determined by evaluating the Z0 value
over time. Fluorescent probe 11 and FadD28 in FP buffer were
incubated with 24 as a positive control or with DMSO as a negative
control. The observed anisotropies were recorded at 10, 30, 60, 120,
240 and 1080 min and fit to Eq. (7) to obtain a Z0 value [52]. The Z0
value remained greater than 0.7 over the 24 h time period
(Fig. 7A); thus the assay demonstrated excellent stability. We also
showed that the signal stability as measured by Z0 did not vary sig-
nificantly from 23 to 28 °C (Fig. 7B). In high throughput libraries,
compounds are generally dissolved in DMSO; therefore, we deter-
mined the maximum concentration of DMSO tolerated in this as-
say. Fluorescent probe 11 and FadD28 in FP buffer were
incubated with varying concentrations of DMSO for 10 min and
the observed anisotropies were recorded. From this analysis con-
centrations of 5% DMSO were acceptable, but the anisotropy
greatly diminished at higher DMSO concentrations (data not
shown); consequently, we elected to use 3% DMSO in our final
assay conditions.
Conclusion
In this body of work, we have described the development of a
fluorescence polarization displacement assay amenable to high-
throughput screening for the discovery of active-site-directed
inhibitors of FadD28 from M. tuberculosis. We designed and synthe-
sized a fluorescent probe/tracer 11 that tightly binds FadD28 based
on the known inhibitor 24 by attachment of the 5-carboxytetram-
ethylrhodamine fluorophore on the C-2 position of the nucleoside
moiety of 24 via a triethyleneglycol spacer. Competitive displace-
ment experiments were performed with a series of known ligands
demonstrating that the assay was robust and well-behaved. The
assay was relatively insensitive to incubation time and tempera-
ture, but displayed a sharp dependence on DMSO concentration.
As a final level of validation the LOPAC library was successfully
screened and four ligands were identified that resulted in a greater
than 25% displacement of the probe/tracer including two nucleo-
tides and two fatty acids. These results are wholly consistent with
the fatty acid–nucleoside-based FP probe/tracer 11 and demon-
strate the assays readiness for high-throughput screening. We
expect that the described assay can be adapted to study other FadD
enzymes by suitable modification of the acyl chain to modulate
binding affinity.
High-throughput screening
In order to further validate the FP assay for high throughput
screening, the LOPAC1280 library was screened in duplicate
against FadD28 in a 30
lL 384-well plate format. The assay per-
Acknowledgments
formed well with a calculated Z0 factor P0.75 for each plate and
a signal-to-noise >14 with four compounds exhibiting greater than
25% displacement. Z0 values between 0.5 and 1.0 are considered
acceptable; thus the assay was deemed suitable for high-through
screening [52]. The four hit compounds were then retested using
the FP competitive displacement assay to determine their respec-
tive dissociation constants (Table 2 and Fig. S2 in Supplementary
data). The four compounds discovered can be grouped into two
categories: ATP derivatives and long chain fatty acid derivatives.
Of the ATP derivatives, 2-methylthio-ADP trisodium salt hydrate
We thank Daniel Wilson for testing inhibitors 21–24 in the
coupled kinetic assay. This research was supported by a Grant from
the National Institutes of Health (NS066415).
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.ab.2011.06.037.

Fluorescence polarization assay for FadD28 / K.D. Grimes, C.C. Aldrich / Anal. Biochem. 417 (2011) 264–273
273
[26] O.A. Trivedi, P. Arora, A. Vats, M.Z. Ansari, R. Tickoo, V. Sridharan, D. Mohanty,
R.S. Gokhale, Dissecting the mechanism and assembly of a complex virulence
mycobacterial lipid, Mol. Cell 17 (2005) 631–643.
[27] R. Simeone, M. Leger, P. Constant, W. Malaga, H. Marrakchi, M. Daffe, C.
Guilhot, C. Chalut, Delineation of the roles of FadD22, FadD26 and FadD29 in
the biosynthesis of phthiocerol dimycocerosates and related compounds in
Mycobacterium tuberculosis, FEBS J. 277 (2010) 2715–2725.
[28] J.A. Ferreras, K.L. Stirrett, X. Lu, J.S. Ryu, C.E. Soll, D.S. Tan, L.E. Quadri,
Mycobacterial phenolic glycolipid virulence factor biosynthesis: mechanism
and small-molecule inhibition of polyketide chain initiation, Chem. Biol. 15
(2008) 51–61.
[29] J. Lynett, R.W. Stokes, Selection of transposon mutants of Mycobacterium
tuberculosis with increased macrophage infectivity identifies fadD23 to be
involved in sulfolipid production and association with macrophages,
Microbiology 153 (2007) 3133–3140.
[30] D.J. Wilson, C.C. Aldrich, A continuous kinetic assay for adenylation enzyme
activity and inhibition, Anal. Biochem. 404 (2010) 56–63.
[31] T.J. Burke, K.R. Loniello, J.A. Beebe, K.M. Ervin, Development and application of
fluorescence polarization assays in drug discovery, Comb. Chem. High
Throughput Screen. 6 (2003) 183–194.
[32] M. Hafner, E. Vianini, B. Albertoni, L. Marchetti, I. Grune, C. Gloeckner, M.
Famulok, Displacement of protein-bound aptamers with small molecules
screened by fluorescence polarization, Nat. Protoc. 3 (2008) 579–587.
[33] M.H. Roehrl, J.Y. Wang, G. Wagner, A general framework for development and
data analysis of competitive high-throughput screens for small-molecule
inhibitors of protein–protein interactions by fluorescence polarization,
Biochemistry 43 (2004) 16056–16066.
[34] A. Goyal, M. Yousuf, E. Rajakumara, P. Arora, R.S. Gokhale, R.
Sankaranarayanan, Crystallization and preliminary X-ray crystallographic
studies of the N-terminal domain of FadD28, a fatty-acyl AMP ligase from
Mycobacterium tuberculosis, Acta Crystallogr. Sect. F: Struct. Biol. Cryst.
Commun. 62 (2006) 350–352.
References
[1] World Health Organization, Fact Sheet on Tuberculosis, >http://www.who.int/
tb/publications/2010/factsheet_tb_2010_rev21feb11.pdf>
16.05.11).
[2] K. Takayama, C. Wang, G.S. Besra, Pathway to synthesis and processing of mycolic
acids in Mycobacterium tuberculosis, Clin. Microbiol. Rev. 18 (2005) 81–101.
[3] P.J. Brennan, H. Nikaido, The envelope of mycobacteria, Annu. Rev. Biochem. 64
(1995) 29–63.
[4] V. Jarlier, H. Nikaido, Mycobacterial cell wall: structure and role in natural
resistance to antibiotics, FEMS Microbiol. Lett. 123 (1994) 11–18.
[5] C.E. Barry 3rd, K. Mdluli, Drug sensitivity and environmental adaptation of
mycobacterial cell wall components, Trends Microbiol. 4 (1996) 275–281.
[6] K.C. Onwueme, C.J. Vos, J. Zurita, J.A. Ferreras, L.E. Quadri, The dimycocerosate ester
polyketide virulence factors of mycobacteria, Prog. Lipid Res. 44 (2005) 259–302.
[7] M.B. Reed, P. Domenech, C. Manca, H. Su, A.K. Barczak, B.N. Kreiswirth, G.
Kaplan, C.E. Barry 3rd, A glycolipid of hypervirulent tuberculosis strains that
inhibits the innate immune response, Nature 431 (2004) 84–87.
(accessed
on
[8] M.W. Schelle, C.R. Bertozzi, Sulfate metabolism in mycobacteria,
ChemBioChem 7 (2006) 1516–1524.
[9] J.J. De Voss, K. Rutter, B.G. Schroeder, H. Su, Y. Zhu, C.E. Barry 3rd, The salicylate-
derived mycobactin siderophores of Mycobacterium tuberculosis are essential for
growth in macrophages, Proc. Natl. Acad. Sci. USA 97 (2000) 1252–1257.
[10] J.D. McKinney, K. Honer zu Bentrup, E.J. Munoz-Elias, A. Miczak, B. Chen, W.T.
Chan, D. Swenson, J.C. Sacchettini, W.R. Jacobs Jr., D.G. Russell, Persistence of
Mycobacterium tuberculosis in macrophages and mice requires the glyoxylate
shunt enzyme isocitrate lyase, Nature 406 (2000) 735–738.
[11] J. Timm, F.A. Post, L.G. Bekker, G.B. Walther, H.C. Wainwright, R. Manganelli,
W.T. Chan, L. Tsenova, B. Gold, I. Smith, G. Kaplan, J.D. McKinney, Differential
expression of iron-, carbon-, and oxygen-responsive mycobacterial genes in
the lungs of chronically infected mice and tuberculosis patients, Proc. Natl.
Acad. Sci. USA 100 (2003) 14321–14326.
[35] D. Heacock, C.J. Forsyth, K. Shiba, K. Musier-Forsyth, Synthesis and aminoacyl-
tRNA synthetase inhibitory activity of prolyl adenylate analogs, Bioorg. Chem.
24 (1996) 273–289.
[36] A. Matsuda, M. Shinozaki, T. Yamaguchi, H. Homma, R. Nomoto, T. Miyasaka, Y.
Watanabe, T. Abiru, Nucleosides and nucleotides. 103. 2-Alkynyladenosines: a
novel class of selective adenosine A2 receptor agonists with potent
antihypertensive effects, J. Med. Chem. 35 (1992) 241–252.
[12] O.A. Trivedi, P. Arora, V. Sridharan, R. Tickoo, D. Mohanty, R.S. Gokhale,
Enzymic activation and transfer of fatty acids as acyl-adenylates in
mycobacteria, Nature 428 (2004) 441–445.
[13] S.T. Cole, R. Brosch, J. Parkhill, T. Garnier, C. Churcher, D. Harris, S.V. Gordon, K.
Eiglmeier, S. Gas, C.E. Barry 3rd, F. Tekaia, K. Badcock, D. Basham, D. Brown, T.
Chillingworth, R. Connor, R. Davies, K. Devlin, T. Feltwell, S. Gentles, N. Hamlin,
S. Holroyd, T. Hornsby, K. Jagels, A. Krogh, J. McLean, S. Moule, L. Murphy, K.
Oliver, J. Osborne, M.A. Quail, M.A. Rajandream, J. Rogers, S. Rutter, K. Seeger, J.
Skelton, R. Squares, S. Squares, J.E. Sulston, K. Taylor, S. Whitehead, B.G. Barrell,
Deciphering the biology of Mycobacterium tuberculosis from the complete
genome sequence, Nature 393 (1998) 537–544.
[14] P. Arora, A. Goyal, V.T. Natarajan, E. Rajakumara, P. Verma, R. Gupta, M. Yousuf,
O.A. Trivedi, D. Mohanty, A. Tyagi, R. Sankaranarayanan, R.S. Gokhale,
Mechanistic and functional insights into fatty acid activation in
Mycobacterium tuberculosis, Nat. Chem. Biol. 5 (2009) 166–173.
[37] E. Fernandez-Megia, J. Correa, I. Rodriguez-Meizoso, R. Riguera,
A click
approach to unprotected glycodendrimers, Macromolecules 39 (2006) 2113–
2120.
[38] L. Lebeau, P. Oudet, C. Mioskowski, Synthesis of new phospholipids linked to
steroid-hormone derivatives designed for two-dimensional crystallization of
proteins, Helv. Chim. Acta 74 (1991) 1697–1706.
[39] R.A. Copeland, Tight Binding Inhibitors. Evaluation of Enzyme Inhibitors in
Drug Discovery, Wiley, Hoboken, NJ, 2005. pp. 185–187.
[40] X. Huang, Fluorescence polarization competition assay: the range of resolvable
inhibitor potency is limited by the affinity of the fluorescent ligand, J. Biomol.
Screen. 8 (2003) 34–38.
[15] P. Arora, A. Vats, P. Saxena, D. Mohanty, R.S. Gokhale, Promiscuous fatty acyl
CoA ligases produce acyl-CoA and acyl-SNAC precursors for polyketide
biosynthesis, J. Am. Chem. Soc. 127 (2005) 9388–9389.
[41] B.A. Lynch, K.A. Loiacono, C.L. Tiong, S.E. Adams, I.A. MacNeil, A fluorescence
polarization based Src-SH2 binding assay, Anal. Biochem. 247 (1997) 77–82.
[42] K.L. Vedvik, H.C. Eliason, R.L. Hoffman, J.R. Gibson, K.R. Kupcho, R.L. Somberg,
K.W. Vogel, Overcoming compound interference in fluorescence polarization-
based kinase assays using far-red tracers, Assay Drug. Dev. Technol. 2 (2004)
193–203.
[43] G.R. Nakayama, P. Bingham, D. Tan, K.A. Maegley, A fluorescence polarization
assay for screening inhibitors against the ribonuclease H activity of HIV-1
reverse transcriptase, Anal. Biochem. 351 (2006) 260–265.
[44] A. Vaasa, I. Viil, E. Enkvist, K. Viht, G. Raidaru, D. Lavogina, A. Uri, High-affinity
bisubstrate probe for fluorescence anisotropy binding/displacement assays
with protein kinases PKA and ROCK, Anal. Biochem. 385 (2009) 85–93.
[45] E.W. van Tilburg, M. Gremmen, J. von Frijtag Drabbe Kunzel, M. de Groote, A.P.
IJzerman, 2, 8-Disubstituted adenosine derivatives as partial agonists for the
adenosine A2A receptor, Bioorg. Med. Chem. 11 (2003) 2183–2192.
[46] F. Guo, S. Das, B.M. Mueller, C.F. Barbas 3rd, R.A. Lerner, S.C. Sinha, Breaking the
one antibody-one target axiom, Proc. Natl. Acad. Sci. USA 103 (2006) 11009–
11014.
[47] M.H. Roehrl, J.Y. Wang, G. Wagner, Discovery of small-molecule inhibitors of
the NFAT–calcineurin interaction by competitive high-throughput
fluorescence polarization screening, Biochemistry 43 (2004) 16067–16075.
[48] J. Neres, D.J. Wilson, L. Celia, B.J. Beck, C.C. Aldrich, Aryl acid adenylating
enzymes involved in siderophore biosynthesis: fluorescence polarization
assay, ligand specificity, and discovery of non-nucleoside inhibitors via high-
throughput screening, Biochemistry 47 (2008) 11735–11749.
[16] G. Khare, V. Gupta, R.K. Gupta, R. Gupta, R. Bhat, A.K. Tyagi, Dissecting the role
of critical residues and substrate preference of a fatty acyl-CoA synthetase
(FadD13) of Mycobacterium tuberculosis, PLoS One 4 (2009) e8387.
[17] C.M. Sassetti, D.H. Boyd, E.J. Rubin, Genes required for mycobacterial growth
defined by high density mutagenesis, Mol. Microbiol. 48 (2003) 77–84.
[18] R.S. Gokhale, P. Saxena, T. Chopra, D. Mohanty, Versatile polyketide enzymatic
machinery for the biosynthesis of complex mycobacterial lipids, Nat. Prod.
Rep. 24 (2007) 267–277.
[19] D. Portevin, C. de Sousa-D’Auria, H. Montrozier, C. Houssin, A. Stella, M.A.
Laneelle, F. Bardou, C. Guilhot, M. Daffe, The acyl-AMP ligase FadD32 and
AccD4-containing acyl-CoA carboxylase are required for the synthesis of
mycolic acids and essential for mycobacterial growth: identification of the
carboxylation product and determination of the acyl-CoA carboxylase
components, J. Biol. Chem. 280 (2005) 8862–8874.
[20] S. Gavalda, M. Leger, B. van der Rest, A. Stella, F. Bardou, H. Montrozier, C.
Chalut, O. Burlet-Schiltz, H. Marrakchi, M. Daffe, A. Quemard, The Pks13/
FadD32 crosstalk for the biosynthesis of mycolic acids in Mycobacterium
tuberculosis, J. Biol. Chem. 284 (2009) 19255–19264.
[21] M. Leger, S. Gavalda, V. Guillet, B. van der Rest, N. Slama, H. Montrozier, L.
Mourey, A. Quemard, M. Daffe, H. Marrakchi, The dual function of the
Mycobacterium tuberculosis FadD32 required for mycolic acid biosynthesis,
Chem. Biol. 16 (2009) 510–519.
[22] L. Rindi, L. Fattorini, D. Bonanni, E. Iona, G. Freer, D. Tan, G. Deho, G. Orefici, C.
Garzelli, Involvement of the fadD33 gene in the growth of Mycobacterium
tuberculosis in the liver of BALB/c mice, Microbiology 148 (2002) 3873–3880.
[23] L. Rindi, D. Bonanni, N. Lari, C. Garzelli, Requirement of gene fadD33 for the
growth of Mycobacterium tuberculosis in a hepatocyte cell line, New Microbiol.
27 (2004) 125–131.
[24] R. Krithika, U. Marathe, P. Saxena, M.Z. Ansari, D. Mohanty, R.S. Gokhale, A
genetic locus required for iron acquisition in Mycobacterium tuberculosis, Proc.
Natl. Acad. Sci. USA 103 (2006) 2069–2074.
[25] T.D. Sirakova, A.M. Fitzmaurice, P. Kolattukudy, Regulation of expression of
mas and fadD28, two genes involved in production of dimycocerosyl
phthiocerol, a virulence factor of Mycobacterium tuberculosis, J. Bacteriol. 184
(2002) 6796–6802.
[49] S.L. McGovern, B.T. Helfand, B. Feng, B.K. Shoichet, A specific mechanism of
nonspecific inhibition, J. Med. Chem. 46 (2003) 4265–4272.
[50] B.Y. Feng, B.K. Shoichet,
A detergent-based assay for the detection of
promiscuous inhibitors, Nat. Protoc. 1 (2006) 550–553.
[51] B.Y. Feng, A. Simeonov, A. Jadhav, K. Babaoglu, J. Inglese, B.K. Shoichet, C.P.
Austin, A high-throughput screen for aggregation-based inhibition in a large
compound library, J. Med. Chem. 50 (2007) 2385–2390.
[52] J.H. Zhang, T.D. Chung, K.R. Oldenburg, A simple statistical parameter for use in
evaluation and validation of high throughput screening assays, J. Biomol.
Screen. 4 (1999) 67–73.