The first nonradioactive fluorescence assay for phosphatidylglycerol: prolipoprotein diacylglyceryl transferase that initiates bacterial lipoprotein biosynthesis

Article abstract of DOI:10.1016/j.ab.2012.01.018

The unique and physiologically vital bacterial enzyme, prolipoprotein diacylglyceryl transferase (Lgt), which catalyzes the committed first step in the posttranslational transfer of diacylglyceryl group from phosphatidylglycerol to the prospective N-terminal cysteine of prolipoproteins, remains to be characterized for want of a simpler but equally sensitive nonradioactive assay. We, for the first time, report a coupled enzymatic fluorescence assay for Lgt using the de novo synthetic peptide substrate MKATKSAVGSTLAGCSSHHHHHH. The assay is based on the conversion of the by-product, glycerol-1-phosphate, to dihydroxyacetone using an alkaline phosphatase-glycerol dehydrogenase combination and estimating the fluorescence of the coupled reduction of resazurin to resorufin. The minimum amount of glycerol-1-phosphate, and hence the modified peptide, detected by this method is approximately 20 pmol, thereby making this assay a promising alternative to the radioactive assays. The assay is rapid, more convenient, less laborious, and suitable for purification and characterization of Lgt.

Full text of DOI:10.1016/j.ab.2012.01.018

Analytical Biochemistry 423 (2012) 163–170
Contents lists available at SciVerse ScienceDirect
Analytical Biochemistry
journal homepage: www.elsevier.com/locate/yabio
The first nonradioactive fluorescence assay for phosphatidylglycerol:prolipoprotein
diacylglyceryl transferase that initiates bacterial lipoprotein biosynthesis
⇑
Srividhya Sundaram, Sanchari Banerjee, Krishnan Sankaran
Centre for Biotechnology, Anna University, Chennai 600 025, India
a r t i c l e i n f o
a b s t r a c t
Article history:
The unique and physiologically vital bacterial enzyme, prolipoprotein diacylglyceryl transferase (Lgt),
which catalyzes the committed first step in the posttranslational transfer of diacylglyceryl group from
phosphatidylglycerol to the prospective N-terminal cysteine of prolipoproteins, remains to be character-
ized for want of a simpler but equally sensitive nonradioactive assay. We, for the first time, report a
coupled enzymatic fluorescence assay for Lgt using the de novo synthetic peptide substrate MKATK
SAVGSTLAGCSSHHHHHH. The assay is based on the conversion of the by-product, glycerol-1-phosphate,
to dihydroxyacetone using an alkaline phosphatase–glycerol dehydrogenase combination and estimating
the fluorescence of the coupled reduction of resazurin to resorufin. The minimum amount of glycerol-1-
phosphate, and hence the modified peptide, detected by this method is approximately 20 pmol, thereby
making this assay a promising alternative to the radioactive assays. The assay is rapid, more convenient,
less laborious, and suitable for purification and characterization of Lgt.
Received 8 December 2011
Received in revised form 17 January 2012
Accepted 19 January 2012
Available online 27 January 2012
Keywords:
Bacterial lipid modification
Prolipoprotein diacylglyceryl transferase
Peptide substrate
Coupled enzyme assay
Enzyme characterization
Ó 2012 Elsevier Inc. All rights reserved.
More than 15,000 posttranslationally modified lipoproteins
have been either identified or predicted in both gram-positive
and gram-negative bacteria. These lipoproteins perform essential
functions in structural integrity, nutrient uptake, environmental
signal transduction, cell division, sporulation, conjugation, protein
secretion, adhesion, and antibiotic resistance [1]. They are synthe-
sized as precursors with tripartite signal sequence, with a charac-
teristic C-terminal ‘‘lipobox’’ (consensus sequence [LVI] [ASTVI]
[GAS] [C]) [2]. During maturation, the invariant Cys sequentially
undergoes diacylglyceryl modification catalyzed by phosphatidyl-
glycerol:prolipoprotein diacylglyceryl transferase (Lgt),1 signal
sequence cleavage at the modification site catalyzed by lipopro-
tein-specific signal peptidase (Lsp II), and N-acylation of the partially
lipid-modified Cys catalyzed by apolipoprotein N-acyl transferase
(Lnt) [3]. The resultant N-acyl-S-diacylglyceryl-Cys anchors bacterial
lipoproteins to the cell membrane at the N-terminal end so that bulk
of the protein efficiently functions in the aqueous compartment.
Recently, the role of Sec (secretory) and TAT (twin arginine
translocase) pathways in differentially recognizing the slow- and
fast-folding lipoproteins, respectively, for lipid modification has
been identified [4]. This unique adaptation at the hydrophobic–
aqueous interface by bacteria has potential in protein engineering
applications, including enzyme-linked immunosorbent assay
(ELISA), liposomal integration, targeted drug delivery, and biosen-
sors [5].
Of the three enzymes, Lgt is most significant because it
catalyzes the first committed step of the unique and ubiquitous
pathway seen only in bacteria. This enzyme, associated on the
cytosolic side of the inner membrane, catalyzes the characteristic
transfer of diacylglyceryl moiety from phosphatidylglycerol (PG)
to the thiol group of lipobox Cys. The 31.6-kDa [6] enzyme is essen-
tial for the growth and viability of gram-negative organisms [7].
Lgt mutants of gram-positive bacteria were viable but showed
physiological deficiencies and attenuated virulence [8–11]. In-
depth study of the unique activity of Lgt has been hindered due
to its low abundance, difficulty to overexpress by recombinant
methods, poor stability, and lack of simpler nonradioactive assays.
The original radioactive assay developed for Lgt using a synthetic
peptide corresponding to the N-terminal 24 amino acids of the pro-
totype prolipoprotein, the precursor of the most abundant murein
lipoprotein, led to the discovery of Lgt, its preference for negatively
charged phospholipid substrate, particularly PG, and the actual
pathway [3]. The fate of glycerol-1-phosphate released in the Lgt
reaction is not yet known, but homologues of glycerol-1-phosphate
dehydrogenase (G1PDH) genes in gram-positive Bacillus subtilis
predict the possible use of glycerol phosphate in the synthesis of
glycerophospholipid backbone [12]. Comparison of primary
⇑
Corresponding author. Fax: +91 44 2235 0299.
E-mail address: ksankran@yahoo.com (K. Sankaran).
Abbreviations used: Lgt, prolipoprotein diacylglyceryl transferase; Sec, secretory;
1
TAT, twin arginine translocase; PG, phosphatidylglycerol; G1PDH, glycerol-1-phos-
phate dehydrogenase; IMV, inverted membrane vesicle; OG, b-octyl glucopyranoside;
EDTA, ethylenediaminetetraacetic acid; DTT, dithiothreitol; GDH, glycerol dehydro-
genase; LB, Luria–Bertani; RFU, relative fluorescence units; WE, water extract; CPM,
counts per minute; SDS–PAGE, sodium dodecyl sulfate–polyacrylamide gel
electrophoresis.
0003-2697/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved.
doi:10.1016/j.ab.2012.01.018

164
Coupled fluorescence assay for Lgt / S. Sundaram et al. / Anal. Biochem. 423 (2012) 163–170
sequences of Lgt from phylogenetically distinct species of
bacteria—Escherichia coli, Salmonella typhimurium, Haemophilus
influenza, and Staphylococcus aureus—revealed a conserved 103-
HGGLIG-108 motif, possibly involved in catalysis [6]. Site-directed
mutagenesis of all the histidine residues and seven of the tyrosine
residues and chemical modification of Lgt with diethyl pyrocar-
bonate (DEPC) implicated His103 and Tyr235 in Lgt activity [13].
The enzyme is very basic (deduced pI of 10.4), with an abundance
of Arg residues in gram-negative organisms and an equal abun-
dance of both Lys and Arg residues in gram-positive organisms.
The inability to precipitate hydrophilic peptide substrates and
their diacylglyceryl-modified product by 90% acetone saturated
with ammonium sulfate in the original peptide assay led to the
development of a generic radioactive assay for Lgt employing a
hydrophilic peptide synthesized de novo [14] and the differential
paper-electrophoretic separation of PG (negatively charged and
radioactive), unmodified substrate (positively charged and nonra-
dioactive), and lipid-modified product (positively charged and
radioactive) owing to their opposite charges. This assay led to the
discovery that E. coli Lgt is superficially associated with the inner
leaflet of inner membrane by hydrophobic interactions. Although
low abundance of the enzyme and the inability to overexpress
were major disadvantages in its isolation, water extraction of the
inverted membrane vesicle (IMV) provided a convenient enriched
source of the enzyme [14]. Very recently, it has been shown using
engineered lipo-EGFP (enhanced green fluorescent protein) that
Sec- or TAT-bound prolipoproteins could be the actual substrates
of Lgt in vivo [4].
dissolved in 100 mM sodium carbonate–bicarbonate buffer. All of
the enzyme reagents were stored at À20 °C. NAD+, NADH, glyc-
erol-1-phosphate, resazurin, and resorufin were dissolved in water
and stored at À20 °C for approximately a month. PG was also ex-
tracted from S. aureus.
Design of fluorometric assay
The new fluorescence assay was designed based on the sensitive
enzymatic estimation of glycerol-1-phosphate released in Lgt-cat-
alyzed reaction. The sequential steps (Fig. 1) in the assay are as
follows:
(i) Lgt catalyzes the transfer of diacylglyceryl moiety from PG to
the thiol group of the invariable cysteine in the lipobox of
the protein/peptide substrate, releasing the modified pro-
lipoprotein/peptide and glycerol-1-phosphate;
(ii) glycerol-1-phosphate is cleaved in the presence of alkaline
phosphatase into glycerol and inorganic phosphate (Pi);
(iii) the released glycerol undergoes dehydrogenation in the
presence of NAD+ and GDH, forming dihydroxyacetone and
NADH+H+; and
(iv) the NADH is then coupled to diaphorase-catalyzed reduction
of resazurin to resorufin (kex = 550 nm and kem = 590 nm), a
highly fluorescent product.
The resulting fluorescence will be directly proportional to the
amount of glycerol-1-phosphate formed in the reaction catalyzed
by Lgt and, hence, will reflect the Lgt activity.
Both the radioactive assays indeed helped to unravel important
characteristics of Lgt, but the serious disadvantages associated
with radioactive assays have hindered the kinetic characterization
and purification of this interesting enzyme. Therefore, we have
developed a new, enzymatically coupled fluorescence assay for
Lgt by estimating the by-product of the reaction, glycerol-1-phos-
phate. The newly developed nonradioactive assay is shown to be as
sensitive and reliable as the radioactive assays.
Assay components
Partial purification of GDH from E. coli DH5
a
The extraction and partial purification of GDH was carried out
as described previously [15] but with alterations in the buffers
used. E. coli DH5a was grown overnight in 1.5 L of Luria–Bertani
Materials and methods
(LB) medium, harvested by centrifugation, and washed twice with
0.9% NaCl. The pellet was resuspended in 0.1 M Tris–HCl (pH 7.5)
and sonicated at 15,000 psi. The turbid supernatant was centri-
fuged at 12,000g for 45 min to obtain the cytosolic fraction, which
was heated at 70 °C for 90 min to get enriched heat-stable protein
preparation. After removing the denatured protein by centrifuga-
tion at 12,000g for 20 min, the enzyme present in the supernatant
(because the enzyme is heat stable) was precipitated with 90%
ammonium sulfate. The pellet was dissolved in 50 mM Tris–HCl
(pH 7.5) to a final protein concentration of 10 mg/ml and dialyzed
against the same buffer for 16 h at 4 °C to remove residual ammo-
nium sulfate. The enzyme extract was then stored in aliquots at the
final protein concentration of 5 mg/ml at À20 °C.
All fluorescence measurements were made with the BioTek
Synergy HT multimode microplate reader at kex = 550 12.5 nm
and kem = 590 15 nm.
Reagents
The following chemicals were used in the study. The synthetic
peptide MKATKSAVGSTLAGCSSHHHHHH was dissolved in 0.1 N
HCl at 5 mg/ml concentration and stored in aliquots at À20 °C.
PG, glycerol-1-phosphate, b-octyl glucopyranoside (OG), and res-
orufin were obtained from Sigma Chemical. NAD+, NADH, glycerol,
resazurin, diaphorase of Clostridium sp. (500 U), and alkaline phos-
phatase of calf intestinal mucosa (200 U) were purchased from Sis-
co Research Laboratories. [9,10] Palmitate (50 Ci/mmol) was
purchased from GE Healthcare.
Buffers used in the study were as follows: TED buffer (20 mM
Tris–HCl, pH 8.0, containing 5 mM ethylenediaminetetraacetic acid
[EDTA] and 4 mM dithiothreitol [DTT]), TBS buffer (20 mM Tris–
HCl, pH 8.0, containing 0.9% [w/v] NaCl), TED–NaCl buffer
(20 mM Tris–HCl, pH 8.0, containing 5 mM EDTA, 4 mM DTT, and
0.5 M NaCl), 100 mM Tris–HCl buffer (pH 8.0), 100 mM sodium
carbonate–bicarbonate buffer (pH 10.0), and 0.1 M sodium acetate
buffer (pH 4.5).
Kinetic characterization of GDH preparation
The activity of GDH was determined by measuring the absor-
bance of NADH at 340 nm during the enzymatic reaction (15 min
at 37 °C) with 10 mM glycerol and 1 mM NAD+ in 100 mM sodium
carbonate–bicarbonate buffer (pH 10.0). The KM of GDH for glyc-
erol was measured by varying it from 25 to 500 mM with 1 mM
NAD+, and the KM for NAD+ was measured by varying it from 0.5
to 50 mM with 500 mM glycerol. Using a Lineweaver–Burk plot
of the kinetic data, the KM values of E. coli GDH for glycerol and
NAD+ were found to be 95 and 2.3 mM, respectively. The best prep-
arations had a specific activity of 74 nmol/min/mg protein. The
preparation was stable for approximately a week when stored at
À20 °C.
Phosphatidylglycerol was solubilized in TED buffer containing
1% OG. Glycerol dehydrogenase (GDH) and diaphorase were
dissolved in 100 mM Tris–HCl, and alkaline phosphatase was

Coupled fluorescence assay for Lgt / S. Sundaram et al. / Anal. Biochem. 423 (2012) 163–170
165
Fig.1. Schematic representation of steps involved in the new fluorometric assay for Lgt.
Estimation of glycerol-1-phosphate using GDH preparation: partial
reaction in Lgt assay
The high KM of GDH for glycerol demanded high enzyme con-
centration in the assay mix to estimate glycerol-1-phosphate in
(WE) was prepared by suspending the IMV pellet thoroughly in
double distilled water at a protein concentration of 5 mg/ml and
incubating the suspension on ice for 1 h, which was then centri-
fuged at 1,00,000g for 1 h. The supernatant was used as the
water-soluble fraction of IMV, and the pellet, suspended in the
same volume of TED buffer, served as the water-insoluble fraction
of IMV.
the range of 20 to 500 pmol. Here, 150 lg of partially purified
GDH along with 0.1 U of alkaline phosphatase, 10 mM NAD+,
0.2 U of diaphorase, and 1 mM resazurin in TED buffer was found
to be optimal for the estimation of glycerol-1-phosphate. After
15 min of incubation at 37 °C in the dark, fluorescence was mea-
sured in a fluorescence microplate reader, where 200 relative fluo-
rescence units (RFU) corresponded to 1 pmol of resorufin.
Preparation of radioactive PG from S. aureus
S. aureus, which has approximately 70% PG in its total phospho-
lipid content [16], was inoculated in nutrient broth and allowed to
grow at 37 °C. At A600nm = 0.3, 10 lCi/ml [9,10] palmitate (specific
activity = 50 Ci/mmol) was added and grown for another 3 h at
37 °C. The radiolabeled cells were then harvested and suspended
in 0.3 ml of TE buffer (pH 8.0) and extracted with 2:1 chloro-
form:methanol according to the Bligh and Dyer method [17]. The
chloroform phase containing PG was aspirated into a fresh tube;
radioactivity was counted in a liquid scintillation counter, and then
the phase was dried in aliquots. Whenever required, the dried res-
idue was suspended in TED buffer to the final concentration of
Preparation of IMV and its extraction with water as Lgt source
The preparation of IMV and water extract (WE) of Lgt was based
on the method described by Selvan and Sankaran [14]. E. coli
BL21(DE3) cells were used as host for hyperexpressing Lgt in the
plasmid pT7GT under the T7 promoter system. Host cells contain-
ing T7GT and the vector control plasmids were grown in 1.5 L of LB
medium overnight, with 100 lg/ml ampicillin selection. After har-
vesting the cells, the pellet was washed twice with 150 ml of TBS
buffer, suspended in 20 ml of TE buffer (pH 8.0) containing 0.5 M
NaCl, and passed twice through French press at 15,000 psi. After
removing the debris by low-speed centrifugation at 3000g for
5 min, the turbid supernatant was centrifuged at 1,25,000g for
1 h. The IMV pellet obtained in this manner was suspended in
TED buffer to a final protein concentration of 5 mg/ml and stored
100,000 counts per minute (CPM)/ll [14].
Coupled fluorescence assay
Diacylglyceryl modification of peptide by Lgt
The assay for Lgt was performed in a total volume of 100 ll con-
as 50-ll aliquots at À80 °C. The water-soluble fraction of Lgt
taining 20 ll of 5x TED buffer, 40 lM peptide, 200 lM PG, 0.1% (w/

166
Coupled fluorescence assay for Lgt / S. Sundaram et al. / Anal. Biochem. 423 (2012) 163–170
v) OG, and IMV (50
l
g of protein) or WE (10
l
g of protein), GDH
concentrations varying from 10 to 50
tion volume and from 2 to 10 g for WE in a 100-
ume. For both fluorescence and radioactive assays, to determine
the KM and Vmax for the peptide, it was varied from 10 to 80
at a constant PG concentration of 200 M. To determine the KM
for PG, it was varied from 25 to 250 M at a constant peptide con-
centration of 40 M.
l
g for IMV in a 100-
ll reac-
(150
l
g of protein), 10 mM NAD+, 0.2 U of diaphorase, and 1 mM
l
l
l reaction vol-
resazurin. IMV or WE from the vector control served as negative
control. After 20 min of incubation at 37 °C in the dark, fluores-
cence was measured in a fluorescence microplate reader. The activ-
ity of Lgt enzyme was converted from relative fluorescence units to
reaction velocity (v) (i.e., picomoles of peptide converted per min-
ute) using the resorufin standard graph.
l
M
l
l
l
Thermal and pH inactivation studies
Tricine SDS–PAGE of the modified and unmodified peptide
The effect of temperature on Lgt activity was assessed by per-
forming the assay with IMV and WE fractions preincubated at 50
and 80 °C for 15 min. In addition, to validate the effect of pH, the
assay was performed with 0.1 M sodium acetate buffer (pH 4.5)
and 0.1 M sodium carbonate–bicarbonate buffer (pH 10.0). Both
fluorometric and radioactive assays were performed in parallel.
Tricine sodium dodecyl sulfate–polyacrylamide gel electropho-
resis (SDS–PAGE) [18] of the reaction mix demonstrates slower
mobility of lipid-modified peptide (product) over unmodified pep-
tide (substrate). The gel was prepared with the separating portion
with 16.5% acrylamide with 6% crosslinking and the stacking por-
tion with 5% acrylamide and 3% crosslinking. Lgt assay mix con-
taining 200
dilutions of the WE–Lgt as enzyme source was mixed with sample
solubilizing buffer after 20 min of the reaction time and 80 l of
the assay mix were loaded in the tricine gel. The electrophoresis
was performed at a constant voltage of 100 V until the dye front
touched the bottom of the gel. Gels were stained by the silver
staining method [19].
lM PG, 40 lM peptide, and TED buffer with different
Results
l
The radioactive assays of Lgt based on the direct estimation of
the amount of tritiated palmitate incorporated into the synthetic
peptide substrate after reaction with the enzyme preparation and
palmitoyl-labeled PG were not convenient for the kinetic charac-
terization and purification of this unique enzyme and its interest-
ing enzyme activity. Hence, an enzymatic fluorescence assay based
on the estimation of the by-product, glycerol-1-phosphate, after
converting it to glycerol and oxidizing it to dihydroxyacetone in
the presence of NAD+ with concomitant reduction of resazurin to
highly fluorescent resorufin had been developed. It has been
shown below that the assay was comparable to the radioactive as-
say in terms of sensitivity but better suited for higher throughput
needed for kinetic characterization and purification.
Radioactive assay for diacylglyceryl modification of peptide
The radioactive assay and paper-electrophoretic separation of
assay components were performed as described by Selvan and
Sankaran [14]. The assay mix contained 40
sponding to 200,000 counts, 0.1% OG, and IMV or WE of Lgt corre-
sponding to 50 or 10 g, respectively. After incubation at 37 °C for
20 min, 2.5 l of this mix (corresponding to 20,000 counts) was ap-
lM peptide, PG corre-
l
l
plied at the center of 19 Â 0.6-cm Whatman No. 2 paper strips wet-
ted with 100 mM Tris borate buffer (pH 8.2) and electrophoresed at
500 V for 1 h. The strips were then dried at room temperature.
From the air-dried strip, a 1-cm-long piece was cut at 1.5 cm away
in the cathodic direction from the center (which corresponds to
modified peptide) and counted in a liquid scintillation counter.
The fluorescence measurement of resorufin could determine 1 pmol of
the final product
Because resorufin is the ultimate product of the coupled enzy-
matic resazurin reduction assay, the linearity and limit of sensitiv-
ity of the assay were determined as a function of the amount of
resorufin expected under the assay conditions. As shown in
Fig. 2, the fluorescence of resorufin was linear in the broad range
of 1 to 1000 pmol and permitted the measurement of a minimum
of 1 pmol of the substance. Although this was the lowest practical
Functional characterization of Lgt
Determination of KM and Vmax
To determine the linearity between enzyme amount and
activity, fluorescence change was measured at different enzyme
Fig.2. Standard graph of resorufin. The fluorescence response due to resorufin was linear from 1 to 1000 pmol, indicating its utility in developing a sensitive optical assay for
Lgt. The inset shows the useful range of 1 to 100 pmol under the assay conditions for Lgt.

Coupled fluorescence assay for Lgt / S. Sundaram et al. / Anal. Biochem. 423 (2012) 163–170
167
limit, suitable for the measurement of the amount of glycerol-1-
phosphate, expected from the Lgt reaction of 4000 pmol of the sub-
strate, the actual sensitivity was actually dependent on the cou-
pling of alkaline phosphatase, GDH, and diaphorase activities in
measuring glycerol-1-phosphate released.
The detection level of glycerol-1-phosphate was as low as 20 pmol
under the assay conditions used for measuring Lgt activity
The sensitivity of the new assay, therefore, was determined
using glycerol-1-phosphate standard and measuring the fluores-
cence after the enzyme-coupled dye reduction under the condi-
tions used in the Lgt assay. The increase in fluorescence was
linear in the range of 20 to 500 pmol of glycerol-1-phosphate
(Fig. 3). For 4000 pmol of peptide, this would correspond to the
conversion of 0.5 to 12.5% of the substrate, the initial velocity con-
ditions. In other words, the amount of Lgt required to convert 1% of
the substrate in 15 min can be measured under the standardized
assay conditions. Under the same assay conditions, 1 pmol of glyc-
erol and 1 pmol of NADH could be estimated (data not shown),
showing that these do not limit the overall sensitivity.
Fig.3. Standard graph of glycerol-1-phosphate using the new method. The mini-
mum amount of glycerol-1-phosphate that could be estimated with the new assay
is approximately 20 pmol, and the range covers the initial velocity conditions for
Lgt.
Fig.4. Effects of peptide substrate concentration (A), PG substrate concentration (B), amount of enzyme (C), and time (D) on the reaction rate of Lgt in both IMV and WE
preparations. The assay conditions are as described in Materials and Methods. The Michaelis–Menten kinetics with respect to the two substrates is apparent and reproducible
in terms of KM already reported. The rate of the reaction increased linearly as a function of enzyme concentration, with the WE showing 5 times higher specific activity
compared with IMV. The Lgt reaction with respect to time (panel D) was linear for 30 min, unlike the abrupt cessation of reaction at 20 min as reported in the radioactive
assay. The rate of the reaction (panel D) was similar for both IMV (5 pmol/min) and WE (6 pmol/min).

168
Coupled fluorescence assay for Lgt / S. Sundaram et al. / Anal. Biochem. 423 (2012) 163–170
Basic kinetic characteristics matched with what has been established
by radioactive methods
The new assay was tested against the known kinetic behavior of
the enzyme as established by the radioactive methods. The activity
of Lgt with respect to both lipid and peptide substrate showed the
typical Michaelis–Menten behavior, with the activity linear up to
40 lM peptide and 200 lM PG (Figs. 4A and B). The rates increased
linearly in the range, where the primary reaction catalyzed by Lgt
remained rate limiting for the coupled system. Vmax of WE–Lgt was
823 pmol/min/mg protein, whereas Vmax of IMV–Lgt was
143 pmol/min/mg protein, indicating that the specific activity of
the WE–Lgt was approximately 5 times that of IMV–Lgt (Fig. 4C).
The Lgt reaction was seen to progress linearly beyond 20 min, un-
like what was reported in the radioactive method, but began to
stop by 30 min (Fig. 4D).
The lipid-modified peptide showed a clear shift in the movement on a
tricine gel
Fig.5. Tricine SDS–PAGE analysis of the Lgt reaction. In this standard analysis, the
lipid-modified peptide appeared as the retarded band, and its intensity after silver
staining shows a progressive increase with a concomitant decrease in the intensity
of the peptide substrate as
a function of enzyme concentration. Evidently,
Because the fluorescence enzymatic assay is an indirect mea-
sure of Lgt, a direct demonstration of the correlation needed to
be established. The tricine gel analysis of the modified and unmod-
ified peptides provided the direct evidence. As can be seen from
Fig. 5, the silver staining revealed the product, modified peptide,
as a retarded band appearing as a result of Lgt activity. The inten-
sity of the retarded band increased as the radioactive and fluores-
cence intensities increased with the product formation. The
sensitivities of the fluorescence and radioactivity were compara-
ble, with the latter being slightly more sensitive.
fluorescence and radioactivity associated with lipid modification in respective
assays are proportional to enzyme concentration (lanes 3–6). The sensitivities of
radioactive and fluorescence assays are comparable, with the former being slightly
more sensitive than the latter. Both radioactive and fluorescence assays are more
sensitive than silver staining.
same behavior was observed for both membrane-bound and WE
fractions in the radioactive assay reported.
Discussion
Comparable kinetic constants derived from both fluorescence and
radioactive assays for Lgt
The three major turnover events of PG in gram-negative bacte-
ria are its conversion to cardiolipin, the transfer of glycerol-1-phos-
phate to the membrane-derived oligosaccharides, and the transfer
of diacylglyceryl group to N-terminal cysteine of prolipoproteins
[20] by Lgt that initiates the bacterial lipid modification pathway.
None of the enzymes of PG turnover is essential for cell division
and growth except Lgt [21]. Lgt is the sole member of transferases
known to transfer diacylglyceryl moiety of PG, and the novelty of
Lgt action, although intriguing, has remained only partially inves-
tigated for its substrate specificity, membrane localization, and
basic kinetics using difficult radioactive assays [3,14]. It is to be
noted that, due to the difficulty of working with this low-abundant
enzyme and not-easy-to-assay conditions, the biochemical charac-
terization of Lgt has not been reported elsewhere. The commercial
potential in comprehensive lipid modification in a single enzy-
matic step led us to develop the first laboratory bioreactor (proto-
type) containing immobilized E. coli Lgt for in vitro lipid
modification of proteins and peptides. For deeper understanding
of this interesting enzyme and its exploitation for commercial
applications, we needed a simpler optical assay, but one as sensi-
tive as the radioactive assay, to save on the expensive peptide/PG
substrates.
To establish the applicability of the newly developed nonradio-
active assay for the kinetic characterization and purification of Lgt,
both the fluorescence and radioactive assays were performed in
parallel and the results compared. The kinetic constants derived
from rate of reaction versus substrate concentration (peptide or
PG), as shown in Table 1, revealed a good match between the
two assays employing crude IMV and enriched or partially purified
WE. As can be seen from Table 1, the values for basic kinetic
parameters (KM and Vmax) of Lgt for the peptide as well as PG
matched those obtained by the radioactive assay performed in par-
allel. In addition, both fractions of the enzyme, membrane bound
and WE, showed identical kinetic behavior (Table 1), indicating
nearly 100% recovery, a fact readily verifiable with the radioactive
assay results. In our previous study using the radioactive assay, we
had reported nearly identical affinity of Lgt for the peptide and PG
substrates [14], unlike the results obtained in the current study.
The KM for PG was higher for both of the Lgt preparations, IMV
and WE, as confirmed by both fluorescence and radioactive assays
performed simultaneously.
Matching results on thermal and pH stabilities from radioactive and
fluorescence assays
The fluorescence assay is comparable to radioactive assay in terms of
sensitivity and reliability
In another comparative test between the radioactive and fluo-
rescence methods to establish the general applicability of the new-
ly developed assay, heat and pH inactivation of the two enzyme
preparations, IMV and WE, were performed; the results are shown
in Table 2. By both radioactive and fluorescence assays, IMV and
WE preparations of Lgt exhibited maximum activity at 37 °C and
pH 8.0 [14]. As seen from the values in Table 2, the enzyme lost
most of its activity when pretreated at 50 °C and pH 10.0. In addi-
tion, complete abolition of activity was observed at pH 4.5. The
A number of fluorescence assay methods with reagents such as
iodoacetamido fluorescein, dansyl aziridine, monobromobimane,
dansyl chloride, and 4-amino fluorescein have been attempted
(unpublished results) for either the direct measurement (modified
peptide) or indirect measurement (unmodified peptide) of Lgt
activity. But all of these methods were not suitable owing to prob-
lems of low sensitivity or poor stability of the fluorogenic reagents
or interference from by-products. Because coupled enzymatic

Coupled fluorescence assay for Lgt / S. Sundaram et al. / Anal. Biochem. 423 (2012) 163–170
169
Table 1
Kinetic constants of Lgt for peptide and PG substrates: Comparison of fluorescence and radioactive assays.
Peptide
IMV
PG
WE
IMV
WE
KM
(lM)
Vmax (pmol/min)
KM
(lM)
Vmax (pmol/min)
KM
(lM)
Vmax (pmol/min)
KM(l
M)
Vmax(pmol/min)
Fluorescence
Radioactive
32
13
12
12
24
10
13
13
69
59
7
15
68
52
8
16
Table 2
Fluorescence and radioactivity associated with heat-inactivated, pH-inactivated enzyme reactions and reaction blanks.
Fluorescence (RFU)
Clone
Radioactive (CPM)
Vector control
IMV
Clone
IMV
Vector control
IMV
IMV
WE
WE
WE
WE
Control (37 °C, pH 8.0)
982
112
26
21
165
95
1302
125
3
26
142
88
87
21
36
15
62
34
101
26
102
42
25
47
93
61
87
48
1256
42
78
85
112
52
2989
56
34
73
142
59
58
32
28
51
67
46
69
49
96
35
55
47
47
46
34
44
D
D
50 °C
80 °C
pH 4.5
pH 10.0
Peptide blank
PG blank
Enzyme blank
118
48
124
65
64
52
87
51
methods have been reported and are commercially available for
glycerol estimation, we attempted to extend such a method to
the measurement of glycerol-1-phosphate, which is the other
product of Lgt reaction. To match the sensitivity of the radioactive
assay with minimal use of the peptide substrate, we needed to im-
prove the detection limit of glycerol-1-phosphate by coupling the
reaction to a fluorogenic dye with high molar extinction coefficient
and emission peak in a region where there is minimal background
signal or autofluorescence [22]. Resazurin is a redox-sensitive and
stable water-soluble dye whose highly fluorescent reduced form,
resorufin, has been used effectively to measure cell death [22]
and viability and drug resistance [23] as well as to perform analysis
of triglycerides in serum [24], diagnostic procedures of tuberculo-
sis (TB) specimens [25], and high-throughput analysis of lipases
and esterases [26]. By employing three sequential enzymatic reac-
tions catalyzed by alkaline phosphatase, GDH, and diaphorase, sub-
sequent to Lgt action, we were able to develop the reported
fluorescence assay for Lgt. The KM and Vmax of each enzyme reagent
was measured (results not shown) to determine the optimal cou-
pling conditions in facilitating the detection of Lgt activity under
initial velocity conditions. In terms of sensitivity, even 0.5% conver-
sion of the 4 nmol of the peptide normally used, which is as low as
20 pmol of the modified peptide, could be detected.
temperature of commercial Cellulomonas GDH at which Lgt is inac-
tivated, cessation was observed after 20 min, supporting the infer-
ence of product inhibition by glycerol-1-phosphate (data not
shown). The initial lag until 5 min was found to be due to the accu-
mulation of substrate required for the detectable activity of alka-
line phosphatase given that the assay of alkaline phosphatase as
a function of time using commercially available glycerol-1-phos-
phate showed the same lag during the initial 5 min (data not
shown).
With respect to the kinetics of Lgt, comparable KM values of IMV
(32
fractions for peptide substrate. These values did not differ much
from the reported KM values for the hydrophilic peptide (6.1 M)
[14] (radioactive assay data in this study) and the hydrophobic
peptide (30 M) [13] determined by radioactive assays. For PG
substrate, however, higher KM values of 69 and 68 M were ob-
lM) and WE (24 lM) fractions indicated similar affinity of both
l
l
l
served for IMV–Lgt and WE–Lgt, respectively. These values are
approximately 5-fold higher than the previously reported values
of 10 lM [13] and 9.6 lM [14]. However, the kinetic constants de-
rived from radioactive assays performed in parallel closely match
the current data. Direct evidence for the reliability of the fluores-
cence assay, however, was provided by comparing the results of
the fluorescence and radioactive assays with the gel analysis of
the reaction products for peptide modification as a function of en-
zyme concentration.
The coupled enzymatic fluorescence assay points to product inhibition
by glycerol-1-phosphate
The easier and more convenient fluorescence assay will now aid the
detailed study of Lgt, a unique enzyme of great interest to
enzymologists and protein engineers
The fluorescence assay exhibited two features not seen in radio-
active assay when the Lgt reaction was measured as a function of
time. The initial lag for 5 min and the extension of the linearity be-
yond 20 min with gradual cessation after 30 min appear to be the
consequence of coupled enzymatic reactions. However, on per-
forming the assay in two stages (i.e., by adding the coupling en-
zymes along with the dye at stipulated time intervals during Lgt
reaction), the cessation could be observed at 20 min, indicating
product inhibition by glycerol-1-phosphate. In other words, the
use of glycerol-1-phosphate by alkaline phosphatase in the cou-
pled single-step assay may be providing some relief, resulting in
the reaction extending and gradually stopping by 30 min. When
the assay was performed in two steps, Lgt reaction at 37 °C
followed by enzymatic coupling reaction at 50 °C, the optimal
Developing a cost-effective yet sensitive assay was the major
aim of this study, and thus the assay was optimized from that an-
gle. One of the assay components, GDH, is commercially available
from Cellulomonas sp. (specific activity = 100 U/mg protein), where
0.01 U/assay will suffice. However, the assay needed to be per-
formed in two steps because the optimal temperature of Cellulo-
monas GDH is 50 °C, whereas for Lgt it is 37 °C. Hence, we
resorted to using E. coli GDH. Owing to its high thermal stability,
a simple enrichment step by heating the E. coli lysate for 90 min
at 70 °C provided the second enzymatic reagent. Despite the sim-
ple preparation, the problems encountered with E. coli GDH are

170
Coupled fluorescence assay for Lgt / S. Sundaram et al. / Anal. Biochem. 423 (2012) 163–170
[3] K. Sankaran, H.C. Wu, Lipid modification of bacterial prolipoprotein: transfer of
diacylglyceryl moiety from phosphatidylglycerol, J. Biol. Chem. 269 (1994)
19701–19706.
[4] H. Shruthi, P. Anand, V. Murugan, K. Sankaran, Twin arginine translocase
pathway and fast-folding lipoprotein biosynthesis in E. coli: interesting
implications and applications, Mol. Biosyst. 6 (2010) 999–1007.
[5] S. Kamalakannan, V. Murugan, M.V. Jagannadham, R. Nagaraj, K. Sankaran,
Bacterial lipid modification of proteins for novel protein engineering
applications, Protein Eng. Des. Sel. 17 (2004) 721–729.
[6] H.Y. Qi, K. Sankaran, K. Gan, H.C. Wu, Structure–function relationship of
bacterial prolipoprotein diacylglyceryl transferase: functionally significant
conserved regions, J. Bacteriol. 177 (1995) 6820–6824.
[7] K. Gan, K. Sankaran, M.G. Williams, M. Aldea, K.E. Rudd, S.R. Kushner, H.C. Wu,
The umpA gene of Escherichia coli encodes phosphatidylglycerol:
prolipoprotein diacylglyceryl transferase (lgt) and regulates thymidylate
synthase levels through translational coupling, J. Bacteriol. 177 (1995) 1879–
1882.
[8] M. Baumgartner, U. Karst, B. Gerstel, M. Loessner, J. Wehland, L. Jansch,
Inactivation of Lgt allows systematic characterization of lipoproteins from
Listeria monocytogenes, J. Bacteriol. 189 (2007) 313–324.
[9] S. Leskelä, E. Wahlstrom, V.P. Kontinen, M. Sarvas, Lipid modification of
prelipoproteins is dispensable for growth but essential for efficient protein
secretion in Bacillus subtilis: characterization of the lgt gene, Mol. Microbiol. 31
(1999) 1075–1085.
[10] C.M. Petit, J.R. Brown, K. Ingraham, A.P. Bryant, D.J. Holmes, Lipid modification
of prelipoproteins is dispensable for growth in vitro but essential for virulence
in Streptococcus pneumoniae, FEMS Microbiol. Lett. 200 (2001) 229–233.
[11] H. Stoll, J. Dengjel, C. Nerz, F. Gotz, Staphylococcus aureus deficient in lipidation
of prelipoproteins is attenuated in growth and immune activation, Infect.
Immun. 73 (2005) 2411–2423.
[12] H. Guldan, R. Sterner, P. Babinger, Identification and characterization of a
bacterial glycerol-1-phosphate dehydrogenase: Ni2+-dependent AraM from
Bacillus subtilis, Biochemistry 47 (2008) 7376–7384.
its shorter shelf life (50% activity loss after 1 week in À20 °C) and
the need to use a greater amount (150 g of protein) of approxi-
l
mately 0.07 U for each reaction. In the final analysis, however,
the easier extraction procedure of GDH from E. coli and the avail-
ability of cost-effective components such as NAD+, diaphorase,
alkaline phosphatase, and resazurin served as major advantages
of this assay, especially in replacing the difficult and cumbersome
radioactive methods.
In our recent study, we reported that, in vivo, the translocation
of lipoproteins could be either Sec dependent (slow-folding lipo-
proteins) or TAT dependent (fast-folding lipoproteins), and it was
postulated that Sec-bound (unfolded proteins) or Tat-bound
(folded proteins) prolipoproteins are in fact the substrates for lipid
modification by Lgt [4]. However, in vitro, Lgt is shown to modify
free peptide/prolipoprotein substrates. The mechanistic reason
for this behavior of Lgt has not yet been investigated. This property,
however, may be attributable to the physiological necessity of the
low-abundant Lgt to efficiently modify during translocation
approximately 100 different lipoproteins per cell, with some of
them such as murein lipoprotein being present in very high num-
bers and some others being present in very low abundance. The
comparison of in vitro conversion of peptide substrates containing
both Sec and TAT signals with lipobox would be an interesting
study using this simple assay.
The availability of this generic and rapid assay will aid not only
such mechanistic studies but also the purification of the enzyme,
screening for its potent inhibitors as possible antibacterials, and
immobilization for protein engineering applications. The method
is readily adaptable for high-throughput screening of Lgt clones
in a 96-well format and is now being used by our group for bio-
chemical characterization and purification of Lgt.
[13] K. Sankaran, K. Gan, B. Rash, H.Y. Qi, H.C. Wu, P.D. Rick, Roles of histidine-103
and tyrosine-235 in the function of the prolipoprotein diacylglyceryl
transferase of Escherichia coli, J. Bacteriol. 179 (1997) 2944–2948.
[14] A.T. Selvan, K. Sankaran, Localization and characterization of prolipoprotein
diacylglyceryl transferase (Lgt) critical in bacterial lipoprotein biosynthesis,
Biochimie 90 (2008) 1647–1655.
[15] R.E. Asnis, A.F. Brodie, A glycerol dehydrogenase from Escherichia coli, J. Biol.
Chem. 203 (1953) 153–159.
[16] S.A.
Short,
D.C.
White,
Metabolism
of
phosphatidylglycerol,
lysylphosphatidylglycerol, and cardiolipin of Staphylococcus aureus, J.
Bacteriol. 108 (1971) 219–226.
Conclusion
[17] E.G. Bligh, W.J. Dyer, A rapid method of total lipid extraction and purification,
Can. J. Biochem. Physiol. 37 (1959) 911–917.
[18] H. Schägger, G. Von Jagow, Tricine–sodium dodecyl sulfate–polyacrylamide gel
electrophoresis for the separation of proteins in the range from 1 to 100 kDa,
Anal. Biochem. 166 (1987) 368–379.
[19] L. Kerenyi, F. Gallyas, A highly sensitive method for demonstrating proteins in
electrophoretic, immunoelectrophoretic, and immunodiffusion preparations,
Clin. Chim. Acta 38 (1972) 465–467.
[20] K. Yokota, M. Kito, Transfer of the phosphatidyl moiety of phosphatidylglycerol
to phosphatidylethanolamine in Escherichia coli, J. Bacteriol. 151 (1982) 952–
961.
We have developed a new fluorescence-based assay for Lgt
based on the estimation of glycerol-1-phosphate in which direct
measurement of resorufin, in a series of enzymatic reactions,
would be related to its activity. The assay is an ideal method for
further characterization of this significant enzyme and analysis of
inhibitors.
Acknowledgments
[21] C.R.H. Raetz, Genetic control of phospholipid bilayer assembly, New Comp.
Biochem. 4 (1982) 435–477.
The Department of Biotechnology is acknowledged for a fellow-
ship to Srividhya Sundaram from the Centre for Excellence pro-
gramme-mode support to P. Kaliraj. We thank the Lady Tata
Memorial Trust and the Council of Scientific and Industrial Re-
search (CSIR) for funding assistance to Sanchari Banerjee. Tamil
Selvan is acknowledged for his useful suggestions in the design
of the assay.
[22] R.H. Batchelor, M. Zhou, Use of cellular glucose-6-phosphate dehydrogenase
for cell quantitation: applications in cytotoxicity and apoptosis assays, Anal.
Biochem. 329 (2004) 35–42.
[23] J.C. Palomino, A. Martin, M. Camacho, H. Guerra, J. Swings, F. Portaels,
Resazurin microtiter assay plate: Simple and inexpensive method for
detection of drug resistance in Mycobacterium tuberculosis, Antimicrob.
Agents Chemother. 46 (2002) 2720–2722.
[24] H. Winartasaputra, V.N. Mallet, S.S. Kuan, G.G. Guilbault, Fluorometric and
colorimetric enzymic determination of triglycerides (triacylglycerols) in
serum, Clin. Chem. 26 (1980) 613–617.
[25] P. Farnia, F. Mohammadi, M. Mirsaedi, A.J. Zarife, J. Tabatabee, K. Bahadori, M.
Bahadori, M.R. Masjedi, A.A. Velayati, Application of oxidation-reduction assay
for monitoring treatment of patients with pulmonary tuberculosis, J. Clin.
Microbiol. 42 (2004) 3324–3325.
[26] M. Schmidt, U.T. Bornscheuer, High-throughput assays for lipases and
esterases, Biomol. Eng. 22 (2005) 51–56.
References
[1] I.C. Sutcliffe, R.R. Russell, Lipoproteins of gram-positive bacteria, J. Bacteriol.
177 (1995) 1123–1128.
[2] M.M. Babu, M.L. Priya, A.T. Selvan, M. Madera, J. Gough, L. Aravind, K. Sankaran,
A database of bacterial lipoproteins (DOLOP) with functional assignments to
predicted lipoproteins, J. Bacteriol. 188 (2006) 2761–2773.