Molecular Microbiology (2011) 81(6), 1577–1592 䊏
x
The Mycobacterium tuberculosis Ag85A is a novel
diacylglycerol acyltransferase involved in lipid
body formation mmi_7792 1577..1592
Ayssar A. Elamin,1 Matthias Stehr,1 Ralf Spallek,3
Manfred Rohde2 and Mahavir Singh1,3*
Departments of 1Gene Regulation and Differentiation
and 2Medical Microbiology, Helmholtz Centre for
Infection Research, 38124 Braunschweig, Germany.
3 Introduction
LIONEX Diagnostics and Therapeutics GmbH, D-38126
Braunschweig, Germany.
Summary
Mycobacterium tuberculosis accumulates large
amounts of triacylglycerol (TAG) which acts as
storage compounds for energy and carbon. The
mycobacterial triacylglycerols stored in the form of
intracellular lipid droplets are essential for long-term
survival of M. tuberculosis during a dormant state. We
report here that when the M. tuberculosis myco-
lytransferase Ag85A is overexpressed in Mycobacte-
rium smegmatis mc2155, cell morphology was
changed and the cells became grossly enlarged. A
massive formation of lipid bodies and a change in lipid
pattern was observed simultaneously. We suspected
a possible role of Ag85A in the acyl lipid metabolism
and discovered that the enzyme possesses acyl-
CoA:diacylglycerol acyltransferase (DGAT) activity in
addition to its well-known function as mycolyltrans-
ferase. Ag85A mediates the transesterification of dia-
cylglycerol using long-chain acyl-CoA as acyl donors.
The Km and Kcat values for palmitoleoyl-coenzyme A
were 390 mM and 55.54 min-1 respectively. A docking
model suggests that palmitoleoyl-coenzyme A and
1,2-dipalmitin occupy the same active site as treha-
lose 6,6⬘-dimycolate and trehalose 6⬘-monomycolate.
The site-directed Ser126Ala mutation of the active site
proved that this residue is involved in the catalytic
activity of this enzyme. Although not proven conclu-
sively for dormant stage of M. tuberculosis, our novel
finding about the synthesis of TAGs by Ag85A
strongly suggests that Ag85A may play a significant
Accepted 27 July, 2011. *For correspondence. E-mail msi@
helmholtz-hzi.de; Tel. (+49) 531 6181 5320/5309; Fax (+49) 531
6181 5399.
© 2011 Blackwell Publishing Ltd
doi:10.1111/j.1365-2958.2011.07792.
First published online 23 August 2011
role in the formation of lipid storage bodies and thus
also in the establishment and maintenance of a per-
sistent tuberculosis infection.
More than a third of the world’s population is infected with
Mycobacterium tuberculosis, the bacterium that causes
tuberculosis (TB). M. tuberculosis is the leading cause of
death from a single bacterial species (Coker, 2004). TB
disease kills approximately 1.7 million people worldwide
annually (Russell et al., 2010). The emergence of new
strains of M. tuberculosis which have become resistant to
standard antibiotic therapy has given urgency to the
search for a better vaccine and new drugs against this
pathogen (Bass et al., 1994; de Jong et al., 2004).
Mycobacterium tuberculosis shows a remarkable prop-
erty of existing in different states of invasion (infection),
colonization and persistence. It also has outstanding
mechanisms to escape from elimination and immune or
therapeutic eradication (Casadevall and Pirofski, 2000).
The major obstacle for host defence mechanisms (Jarlier
and Nikaido, 1994) and therapeutic intervention is the
unusual robust cell wall of mycobacterium. The cell wall
structure of M. tuberculosis is unique among prokaryotes,
and it is a major determinant of virulence for the bacterium
(Belisle et al., 1997; Daffe and Draper, 1998). The inner
peptidoglycan sacculus of the mycobacterial cell enve-
lope is wrapped by a polysaccharide layer of mycolyl-
arabinogalactan. The outer envelope consists of trehalose
6,6′-dimycolate (TDM; cord factor) where the mycolic
acids of TDM interact with the mycolyl-residues from the
layer beneath. The mycolic acid-containing layers have
width of ~10 nm and limit the penetration of hydrophilic
substances, and water-soluble antibiotics, whereas the
inner saccharide layer inhibits the penetration of such
substances. Due to the extremely low permeability of the
cell wall mycobacteria show a high degree of intrinsic
resistance to most antibiotics and chemotherapeutic
agents (Ducati et al., 2006). Thus the biosynthesis of
mycolic acids and TDM have become a major target of
drug research (Brennan and Nikaido, 1995). TDM is
exclusively synthesized by the enzymes of the antigen 85
1578 A. A. Elamin et al. 䊏
(Ag85) complex Ag85A, B and C, which are acyltrans-
ferases and convert TMM (trehalose 6′-monomycolate)
to TDM (Belisle et al., 1997). Antigen 85C synthesizes
around 50% of the cell wall linked mycolic acid (Jackson
et al., 1999) and antigen 85A has been shown to be crucial
for the growth in nutrient-poor media and macrophage-like
cell line (Armitige et al., 2000).
Therapeutic eradication of TB is affected by the ability of
the bacterium to survive up to decades in a dormant state,
primarily in hypoxic granulomas in the lung (Murphy and
Brown, 2007). Unfortunately, very little is known about the
biology of the dormant phase and no specific persister
genes have been identified yet (Cardona, 2007; Lewis,
2007; Rao and Li, 2009). This lack of well-defined targets
specific to the dormancy phase has been a major obstacle
for the identification and development of novel leads
(Murphy and Brown, 2007). One important factor for long-
term survival of the bacillus seems to be the ability to store
and metabolize fatty acids, which are thought to be the
major energy source in the dormant phase (Daniel et al.,
2004). Genes that encode enzymes required to metabolize
fatty acids as the major carbon source, such as isocitrate
lyase, were found to be essential for persistence (McKin-
ney et al., 2000; Meena and Rajni, 2010). Actinomycetes
synthesize and accumulate large amounts of triacylglyc-
erol (TAG) which act as storage compounds for energy and
carbon (Olukoshi and Packter, 1994; Alvarez et al., 2000;
Alvarez and Steinbüchel, 2002). Mycobacteria store TAGs
in the form of intracellular lipid droplets during hypoxia-
induced non-replicating persistence (Garton et al., 2002;
Daniel et al., 2004). These bacteria are phenotypically
drug-resistant and probably the cause for prolonged TB
treatment required for therapy (Low et al., 2010). Lipid
body positive acid fast bacilli have been suggested to be
a biomarker for non-replicating M. tuberculosis cells in
sputum (Garton et al., 2008; Low et al., 2010). Out of 15
putative TAG synthases (Tgs) (diacylglycerol acyltrans-
ferase) genes known, only for tgs1 the involvement in
TAG accumulation has been proven experimentally
(Kalscheuer and Steinbüchel, 2003; Daniel et al., 2004;
Sirakova et al., 2006).
During our overexpression studies with the M. tubercu-
losis Ag85A in M. smegmatis mc2155 we observed a
massive formation of lipid bodies and a change of the lipid
pattern, showing the involvement of Ag85A in TAG synthe-
sis. Prompted by this finding we carried out detailed enzy-
matic characterization of Ag85A and report for the first
time that M. tuberculosis Ag85A possess, in addition to its
mycolyltransferase activity, also diacylglycerol acyltrans-
ferase (DGAT) activity. This finding strongly suggests that
Ag85A is a TAG synthase and might be a key player in both
cell wall stability and the biosynthesis of storage com-
pounds for the survival of M. tuberculosis in the dormant
state.
© 2011 Blackwell Publishing Ltd, Molecular Microbiology, 81, 1577–1592
Results
Overexpression of Ag85A changes cell shape, size and
results in formation of giant cells
The recombinant M. smegmatis-Rv3804c showed rapid
overexpression of Ag85A from the mycobacterial heat
shock promoter Phsp60 even before induction (Fig. 1) as
observed previously (Stover et al., 1991). After 18 h no
further increase of expression was observed (Fig. 1I) and
the recombinant M. smegmatis-Rv3804c showed pro-
minent changes in cell size, morphology and surface
structure. The increase in cell length correlated to the
extended expression of Ag85A. A majority of the wild-type
M. smegmatis mc2155 cells were ~2 mm in length
(Fig. 1A) whereas M. smegmatis-Rv3804c cells increased
in length up to 4–5 mm (Fig. 1B and C). These results
indicated that expression of Ag85A interfered with cell
division and led to filamentation. Some of the elongated
cells showed buds and bulb-like structures (Fig. 1D and
E). M. smegmatis-Rv3804c cells cultured in liquid media
(7H9) containing Tween-80 showed significant clumping
and aggregation specifically in the late stationary growth
phase (Fig. 1–2 and H), while the control M. smegmatis
mc2155 did not show any tendency to clump (Fig. 1F1 and
G). To investigate the possibility that the observed aggre-
gation was not due to the presence of kanamycin, the
M. smegmatis-pMV261 cells containing the empty vector
were cultured in the same condition. No aggregation phe-
notypes or any changes in cell size were observed (data
not shown).
Immunoelectron microscopy revealed that Ag85A is
mostly localized in the inner side of cell membrane or cell
wall and at the surface of TAG inclusions (Fig. 2) as has
been found earlier in mycobacteria residing in phago-
somes (Beatty and Russell, 2000). Gold labelling of
Ag85A showed that the enzyme is also present in the
cytoplasm.
Overexpression of Ag85A induces lipid body formation
and thickening of the cell wall in recombinant
M. smegmatis
Mycobacterium smegmatis mc2155 and recombinant
M. smegmatis-Rv3804c were cultured under similar
conditions as described in Experimental procedures
and samples were collected after induction. Transmis-
sion electron microscopic (TEM) on ultrathin sections of
the cells revealed a significant change only in the in
morphology in M. smegmatis-Rv3804c cells and concur-
rently the appearance of neutral lipid vesicles (Fig. 3A
and B). Sections of M. smegmatis-Rv3804c showed an
abundance of electron-transparent vesicles within the
cytoplasm that had no internal structure. These discrete
structures appeared to be bordered by a thin membrane,
 Diacylglycerol acyltransferase activity of Ag85A 1579

Fig. 1. Overexpression of Mycobacterium tuberculosis Ag85A altered the morphology of M. smegmatis cells.
A–H. FESEM analysis revealed enlargement and branching of M. smegmatis-Rv3804c (B and C) with changing of surface shape, also after
induction the cell exhibits tubular like structures (D and E). Wild-type M. smegmatis mc2155 was used as control (A and G). (F) M. smegmatis
mc2155 [1] M. smegmatis-Rv3804c [2] were cultured in 7H9 liquid media with ADC, 0.2% glycerol and 0.05% Tween 80 until an OD600 = 1.8
then the aggregation was tested by allowing the cells to settle for 15 min. The aggregation was clearly visible in FESEM images of
M. smegmatis-Rv3804c (H) while M. smegmatis mc2155 does not show any clumping (G).
I. Kinetics of Ag85A overexpression associated with induced changes in cell morphology by using SDS/PAGE on 12% (w/v) polyacrylamide
gels stained with Coomassie Brilliant Blue R-250. Molecular mass marker proteins were from Fermentas (L). Samples were collected from the
main culture at 12 (lane 2), 14 (lane 3), 16 (lane 4) and 18 h (lane 5) and also after 2 h induction (lane 6) at 45°C with 0.002% H2O2, wild-type
M. smegmatis mc2155 (lane 1) used as control.

were variable in sizes and occupied a substantial portion
of the total cell volume (approximately 30–70%).
After induction, all intermediate stages of matured lipid
bodies (MLBs), lipid prebodies (PBs) and released preb-
odies (RPBs) (Alvarez and Steinbüchel, 2002) were
observed (Fig. 3D right panel and Fig. 3E). The PBs were
visible as flat to spherical, relatively electron-dense struc-
tures close to the cytoplasmic membrane as described
previously (Daniel et al., 2004). PBs did not show a clearly
visible oleogenous layer as observed by Alvarez and
Steinbüchel (2002). The PBs enlarged and dissociated
into the cytoplasm forming RPBs. These RPBs enlarged
further to form MLBs, which were clearly separated from
the cytoplasm with an implied boundary layer. MLBs had
no visible contact to the cytoplasm membrane (Fig. 3D).
The mean thickness of cell wall of M. smegmatis-Rv3804c
was thicker (13 nm) compared with that of the control
M. smegmatis mc2155 (8 nm) (Fig. 3C).

© 2011 Blackwell Publishing Ltd, Molecular Microbiology, 81, 1577–1592

1580 A. A. Elamin et al. 䊏

Fig. 2. Localization of Mycobacterium tuberculosis Ag85A in M. smegmatis mc2155. Localization was observed by electron microscopic
sections using a monoclonal antibody against Ag85A and post-embedding immunogold labelling (black dots). Ag85A was observed at inner
side of cell membrane, cell wall, in the cytoplasm and on the surface of TAG inclusions. PB, lipid prebody; RPB, released prebody; MLB,
matured lipid body; CW, cell wall; CM, cell membrane.

Lipid body development an approximate twofold increase of TDM and TMM in

M. smegmatis-Rv3804c compared with M. smegmatis
Ag85A overexpression led to an increased amount of lipid
mc2155 (Fig. 6A). This result is in agreement with the TLC
droplets in the cell. Using Bodipy 493/503 staining, the
results which show an increase of the glycolipids TMM and
application of flow cytometry allows the monitoring of lipid
TDM for M. smegmatis-Rv3804c (Fig. 6C). Expression of
droplet development. Bodipy 493/503 selectively stains
Ag85A modulates also to the composition and concen-
neutral lipids and is more specific for cellular lipid droplets
tration of mycolic acids. Upon HPLC analysis of their
than staining with Nile red (Gocze and Freeman, 1994).
mycolates, both strains have two slightly distinct
M. smegmatis mc2155 and M. smegmatis-Rv3804c cells
HPLC-double-cluster patterns (Fig. 6B). In comparison
were collected from the main culture at 4, 6, 8, 12, 16 h
with M. smegmatis mc2155 the double-cluster pattern from
and after induction (Figs 4 and 5). M. smegmatis-Rv3804c
M. smegmatis-Rv3804c shows a longer retention time on a
showed a higher fluorescence intensity (101–104) com-
C18 reverse-phase column, indicating the presence of
pared with M. smegmatis mc2155 (100–102), indicating
mycolic acids with longer carbon chains (Fig. 6B) (Wong
the presence of neutral lipid droplets in M. smegmatis-
et al., 1979; Butler et al., 1986; Cage, 1992). TLC results
Rv3804c (Raschke and Knorr, 2009). Fluorescence
indicate as well a slight increase of the mycolic acid con-
intensity of M. smegmatis-Rv3804c increased after 12 h
centration in the M. smegmatis-Rv3804c. The relative
(Fig. 4D–F) showing the increasing amount of neutral
amounts of a′-mycolic acids, a-mycolic acids and epoxy
lipids in the cells. M. smegmatis-mRv3804c harbouring the
mycolic acids increased (Fig. 6C).
plasmid pMV261::85A* (mutated Ag85A) was collected
Thus, Ag85A induces a quantitative and qualitative
at same time points and stained with Bodipy 493/503.
modulation of glycolipids and mycolic acids of the cell
M. smegmatis-mRv3804c strain showed very few lipid
wall. In addition we also observed an increase of storage
bodies in the cytoplasm in all time points (Fig. S2). Obvi-
lipids in the cell. M. smegmatis-Rv3804c shows a higher
ously M. smegmatis-mRv3804c showed changes in the
amount of TAGs in the TLC analysis (Fig. 6D).
cell size and morphology as M. smegmatis-Rv3804c but
without any effect on the lipid bodies formation.
Ag85A has DGAT activity
Ag85A modulates the synthesis of glycolipids, mycolic
The increase of TAGs in M. smegmatis-Rv3804c
acids and TAGs in M. smegmatis-Rv3804c
prompted us to ask the question whether Ag85A has an
In order to confirm that the altered morphology of acyltransferase activity required for TAG synthesis. TAGs
M. smegmatis-Rv3804c phenotype was also the cause of are synthesized by DGAT. DGAT enzymes mediate TAG
changes in cell envelope lipid composition, we analysed formations from long-chain acyl-coenzyme A (CoA) as
the lipid content of wild-type M. smegmatis and acyl donor and long-chain fatty alcohols or diacylglycerols
M. smegmatis-Rv3804c cells. We performed the total lipid as acyl acceptors (Wältermann et al., 2007).
analysis using our newly reported mycolyltransferase In order to test for DGAT activity, purified recombinant
assay and the equations (Elamin et al., 2009). The expres- M. tuberculosis Ag85A from Escherichia coli was assayed
sion of Ag85A led to an increase of concentration of with palmitoleoyl-CoA as acyl donor and 1,2-dipalmitoyl-
the glycolipids TDM and TMM in the cell. We detected sn-glycerol (1,2-dipalmitin) as the acyl acceptor. The

© 2011 Blackwell Publishing Ltd, Molecular Microbiology, 81, 1577–1592

 Diacylglycerol acyltransferase activity of Ag85A 1581

Fig. 3. Ag85A overexpression leads to lipid bodies formation and

changes the cell wall thickness.
A–D. Giant cells overexpressing Ag85A (right panel) from the heat
shock promoter phsp60 exhibit to numerous lipid bodies and a
thicker cell wall than wild type. The wild-type (M. smegmatis
mc2155) cells (left panel) displays normal shape, size with no lipid
bodies. Fig. 4. Monitoring of lipid body development. Flow cytometry
E. Schematic display of the stages of lipid body formation in using Bodipy 493/503 staining showed an increase of lipid bodies
M. smegmatis-Rv3804c (D, right panel). PB, lipid prebody; RPB, during the incubation time and induction for M. smegmatis-
released prebody; MLB, matured lipid body; CW, cell wall; CM, cell Rv3804c, while M. smegmatis mc2155 showed no change in the
membrane. fluorescence histogram. Cells were collected for FACS analysis at
(A) 4, (B) 6, (C) 8, (D) 12, (E) 16 h and (F) after induction.
© 2011 Blackwell Publishing Ltd, Molecular Microbiology, 81, 1577–1592
1582 A. A. Elamin et al. 䊏

Fig. 5. Bodipy 493/503 staining giant cells overexpressing Ag85A from the heat shock promoter. Fluorescence images of
M. smegmatis-Rv3804c cells stained with Bodipy 493/503 and were collected at (A) 4, (B) 6, (C) 8, (D) 12, (E) 16 h and (F), (H) and (I) after
induction, M. smegmatis mc2155 used as control (G). The white arrows (A) indicated the starting of lipid accumulation and the arrows in (B)
indicated the lipid prebodies (PBs).

formation of the product TAG was verified by TLC (Fig. 7D
and E). The TLC results show clearly the production of a
band at the same Rf as the tripalmitin standard. This
product band was analysed by MALDI-TOF-MS and was
verified as tripalmitin. When using the Ser126Ala mutant of
Ag85A, no band corresponding to tripalmitin could be
detected, confirming Ser126 as the catalytic residue for
DGAT activity (data not shown). To determine the kinetic
parameters for Ag85A, the enzyme was assayed using an
enzyme assay which quantifies the thiol groups of released
CoA-SH (Arabolaza et al., 2008) (see Experimental
procedures). Palmitoleoyl-CoA was used as acyl donor
in concentrations ranging from 0–2.5 mM and with 1,2-
dipalmitoyl-sn-glycerol (1,2-dipalmitin) as the acyl accep-
tor at a fixed concentration of 2 mM. The reaction curve
indicated that the DGAT reaction followed Michaelis–
Menten kinetics (Fig. 7A). For palmitoleoyl-CoA an
apparent Km of 390 mM was determined. Vmax was
3166 nmol mg-1 min-1 and Kcat of 55.54 min-1. The
Ser126Ala mutant did not show any detectable activity
(Fig. 7A).
The Km values reported here for Ag85A are quite similar
to those reported for the M. tuberculosis DGAT TGS1
(Sirakova et al., 2006), an enzyme having preference for
C26:0-CoA over C18:1-CoA with Km values 306 and
540 mM respectively. Another reported kinetic studies for
DGAT of M. tuberculosis TGS1–4 showed similar range of
values by using diolein and oleoyl-CoA as substrates
(Wältermann et al., 2007).
The molecular binding model (see below) predicted that
1,2-dipalmitin and palmitoleoyl-CoA can compete for
binding to the active site and elevated concentrations of

© 2011 Blackwell Publishing Ltd, Molecular Microbiology, 81, 1577–1592

 Diacylglycerol acyltransferase activity of Ag85A 1583

Fig. 6. Modulation of glycolipids, mycolic acids and TAGs in M. smegmatis-Rv3804c.

A. Quantification of TMM and TDM by the mycoltransferase assay.
B. HPLC trace of mycolic acid p-bromophenacyl esters from Mycobacterium smegmatis mc2155 and M. smegmatis-Rv3804c, showing a
double-cluster pattern. The pattern of M. smegmatis-Rv3804c has a longer retention time.
C and D. Total lipid analysis of different strains by analytical TLC. (C) Petroleum ether:diethyl ether:acetone (80:20:2, by vol.) was used as
mobile phase. M. smegmatis-Rv3804c shows stronger signal for glycolipids (TMM, TDM) and mycolic acids. Lane 1, M. smegmatis mc2155.
Lane 2, M. smegmatis-Rv3804c. (D) n-hexane:diethyl ether (9:1, by vol.) was used as mobile phase. M. smegmatis-Rv3804c has a stronger
signal for TAGs. Staining was performed with Coomassie blue as described (Elamin et al., 2009). Lane 1, M. smegmatis mc2155. Lane 2,
M. smegmatis-Rv3804c.

© 2011 Blackwell Publishing Ltd, Molecular Microbiology, 81, 1577–1592

1584 A. A. Elamin et al. 䊏

Fig. 7. DGAT activity of recombinant M. tuberculosis Ag85A and the Ser126Ala mutant.
A. Michaelis–Menten kinetics of Ag85A for palmitoleoyl-coenzyme A as substrate. The different substrate concentrations tested follow the
Michaelis–Menten equation (䊉). The Km and Vmax of palmitoleoyl-CoA were determined as 390 mM and 3166 nmol mg-1 min-1 respectively. The
mutant Ag85A (Ser126Ala) did not have any detectable activity ( ).
B and C. The apparent Km and Kcat of 1,2-dipalmitin for M. tuberculosis Ag85A were calculated at two concentrations of palmitoleoyl-coenzyme
A, 100 mM (B) and 700 mM (C). The Km and Kcat values for 1,2-dipalmitin were 9.09 mM and 38.57 min-1 respectively at palmitoleoyl-coenzyme
A concentration of 100 mM. At 700 mM of palmitoleoyl-coenzyme A the Km and Kcat values for 1,2-dipalmitin were 4.34 mM and 124.3 min-1
respectively.
D. TLC analysis of the extracted lipids from the reactions mixture. 1: Dipalmitin as reference, 2: Tripalmitin as reference, 3: Ag85A with 700 mM
palmitoleoyl-coenzyme without dipalmitin A, 4: Ag85A with 700 mM palmitoleoyl-coenzyme A and 10 mM dipalmitin, 5: Ag85A with 700 mM
palmitoleoyl-coenzyme A and 20 mM dipalmitin, 6: Ag85A with 700 mM palmitoleoyl-coenzyme A and 30 mM dipalmitin and 7: Ag85A with
700 mM palmitoleoyl-coenzyme A and 40 mM dipalmitin.
E. MALDI-TOF-MS analysis of the TLC spot obtained from Ag85A reaction, which was scratched from the TLC plates and extracted by
chloroform. The mass spectrum showed the formation of tripalmitin (arrow).
F. Model for inhibition of activity by dipalmitin at low palmitoleoyl-coenzyme A concentration. At high concentration of the acyl donor no
inhibitory effect is observed. Only at low palmitoleoyl-coenzyme A concentration dipalmitin may bind to the active site and block the entry of
the acyl donor.

1,2-dipalmitin should result in substrate inhibition. To test Substrate recognition

the model, we assayed enzyme activity at concentration
lower (100 mM) and at nearly double concentration of the
Km of palmitoleoyl-CoA (700 mM). The inhibitory effect of
1,2-dipalmitin at fixed palmitoleoyl-CoA concentration of
100 mM is clear (Ki = 84 mM) (Fig. 7B). At 700 mM almost
no inhibition of 1,2-dipalmitin was observed (Ki = 6.8 mM).
The Km for 1,2-dipalmitin was determined as 9.09 mM and
4.34 mM at palmitoleoyl-CoA concentrations of 100 and
700 mM respectively (Fig. 7B and C).
In summary, the experiments show clearly that Ag85A of
M. tuberculosis exhibits DGAT activity and is an acyltrans-
ferase involved in TAG biosynthesis.
Since the palmitoleoyl-CoA and 1,2-dipalmitin were
shown to be acyl donor and acyl acceptor, respectively,
we investigated the possible binding site of both sub-
strates. The crystal structure of the homologous protein
Ag85C has been determined (Ronning et al., 2000).
Similar to Ag85C, Ag85A contains a catalytic triad formed
by residues Ser126, His262 and Glu230 and also pos-
sesses the highly conserved serine 126 in a deep
trehalose binding groove as reported for Ag85B and C
(Ronning et al., 2000; Anderson et al., 2001). The acyl
donor palmitoleoyl-CoA was docked to the trehalose

© 2011 Blackwell Publishing Ltd, Molecular Microbiology, 81, 1577–1592

 Diacylglycerol acyltransferase activity of Ag85A 1585

Fig. 8. Proposed catalytic mechanism of acyl transfer and hypothetical binding mode of acyl substrates.
Upper panel. Schematic representation of the proposed catalytic mechanism of acyl transfer. The reaction comprises three-steps: (1) Ser126
allows a nucleophilic attack on the carboxylate carbon of the acyl donor palmitoleoyl-coenzyme A, leading to an acyl-enzyme adduct. (2) The
hydroxyl group of the acyl acceptor 1,2-dipalmitoyl-sn-glycerol performs a nucleophilic attack on the carboxylate carbon of the acyl-enzyme
adduct. (3) The triacylglycerol product is released. The transferred palmitoleoyl moiety is in red.
Lower panel. Proposed binding mode for the acyl substrates. (A) Docked palmitoleoyl-coenzyme A in the binding pocket of antigen 85A. The
thioester carbonyl group is oriented in the vicinity of Ser126 (arrow). Carbon atoms of the docked palmitoleoyl-coenzyme A molecule are
green. (B) Proposed intermediate with the palmitoleic acid covalently attached to Ser126. (C) Complex of the Ag85-palmitoleoyl adduct with
1,2-dipalmitoyl-sn-glycerol. Carbon atoms of the 1,2-dipalmitoyl-sn-glycerol are in grey. The figure shows that the palmitoyl chains are
positioned on top of the palmitoleoyl-Coenzyme A molecule and may block the entry of the acyl donor at high 1,2-dipalmitin concentration. The
position of the active-site Ser126 is indicated with an asterisk. The figure was prepared with Pymol (http://www.pymol.org/).

binding pocket with affinities of -9.3 kcal mol-1. The
thioester carbonyl group is located in close proximity to
the serine 126 and allows a nucleophilic attack by the
serine hydroxyl group leading to the proposed acyl-
enzyme adduct (Ronning et al., 2000). The long palmito-
leoyl chain occupies the deep hydrophobic groove on
the protein as observed for palmitoyl in thioesterase 1
(Fig. 8A) (Bellizzi et al., 2000). The acyl acceptor 1,2-
dipalmitin was docked with an affinity of 4.7 kcal mol-1
to the assumed acyl-enzyme adduct (Fig. 8B). 1,2-
Dipalmitin accommodates the same space in the binding
pocket as trehalose in Ag85B (Anderson et al., 2001). The
palmitoyl chains cover the binding site of palmitoleoyl-
CoA and may block the entry of the acyl donor at high
1,2-dipalmitin concentration. The pantoic acid component
of CoA and an ester group of 1,2-dipalmitin adapts the
glucose ring structure of trehalose bound to Ag85B (PDB
1F0P) and the head region of the palmitoleoyl chain occu-
pies approximately the same position as MPD bound to
Ag85B (Anderson et al., 2001) (data not shown).
Discussion
The enzymes of the antigen 85 complex (Ag85A, B and
C) synthesize the most abundant glycolipid of the myco-
bacterial cell wall, the cord factor. The cord factor (TDM)
is essential for the integrity of the mycobacterial cell wall
and pathogenesis (Nguyen et al., 2005). The enzymes
of the antigen 85 complex may possess as well the
capability to confuse and evade the host immune
system (Armitige et al., 2000; Harth et al., 2002;
Ronning et al., 2004).
In order to study the quantitative effect of Ag85A on the
of mycobacterial cell wall synthesis, we expressed the
M. tuberculosis Ag85A in the non-pathogenic M. smegma-
tis mc2155 which was used as a model organism because
it is fast growing and shares many features with pathogenic
mycobacteria (He and De Buck, 2010). The recombin-
ant M. smegmatis-Rv3804c expressing M. tuberculosis
Ag85A displayed thickening of the cell wall and major
changes in morphology. Unexpectedly, in addition to the
morphological changes, an extensive formation of lipid
bodies was observed. The presence of neutral and
saturated TAGs which are a major components of lipid
inclusions (Daniel et al., 2004; Sirakova et al., 2006; Wäl-
termann et al., 2007) were confirmed by Bodipy 493/503
staining. TAGs are the major storage molecules for eukary-
otic organisms (Yen et al., 2005; 2008; Saha et al., 2006)

© 2011 Blackwell Publishing Ltd, Molecular Microbiology, 81, 1577–1592

1586 A. A. Elamin et al. 䊏
but these are not common storage compounds of bacteria
(Alvarez et al., 2000). Only bacteria belonging to the
genera Acinetobacter (Kalscheuer and Steinbüchel,
2003), Streptomyces (Olukoshi and Packter, 1994; Arabo-
laza et al., 2008), Rhodococcus (Kalscheuer and Stein-
büchel, 2003) and Mycobacterium (Daniel et al., 2004;
Sirakova et al., 2006; Low et al., 2010) accumulate TAGs.
We found that the recombinant Ag85A was partly
membrane-bound and also localized in the cytoplasm or on
surface of lipid inclusions in the recombinant M. smegma-
tis strain as observed for other lipid-synthesizing enzymes
such as the 14 kDa phasin from Rhodococcus ruber
and phasin PhaP1 from Ralstonia eutropha (Pieper-Fürst
et al., 1995).
Flow cytometry revealed a major difference in stain-
ing pattern between M. smegmatis mc2155 cells and
M. smegmatis-Rv3804c which contained very large lipid
bodies. An increase in lipid droplet size led to an
increase in fluorescence intensity allowing a qualitative
statement about the lipid droplet development. The con-
stantly observed low fluorescence in wild type could
result from the fact that the mycobacterial cells contain
different types of neutral lipids. Comparing FACS infor-
mation with the cell images provided by fluorescence
microscope, we conclude that the increase of fluores-
cence is caused by the increased size of neutral lipid
droplets. The TEM and fluorescence staining experi-
ments showed that lipid biosynthesis begins at periph-
eral lipid domains close to the cytoplasm membrane
followed by the subsequent formation of PBs. The
spherical PBs enlarge and are released from the
membrane forming prebodies followed by full release
into the cytoplasm to form mature lipid bodies
(Wältermann et al., 2005). It is difficult to distinguish
between MLBs from RPBs but the former were larger
and clearly separated from the cytoplasm. The control
(parental) M. smegmatis mc2155 strain (Fig. S1) and
M. smegmatis-mRv3804c, the later contains the mutated
Ag85A, showed no or very rarely few lipid bodies in the
cytoplasm. Interestingly, M. smegmatis-mRv3804c dis-
played the same large cell phenotype which was
observed in M. smegmatis-Rv3804c, but without induc-
tion of the lipid bodies (Fig. S2) supporting the possibility
that the Ag85A may had DGAT activity.
Acyl-CoA:DGAT catalyse the biosynthesis of TAGs
(Kalscheuer and Steinbüchel, 2003). The accumulation of
neutral lipids by Ag85A transformants of M. smegmatis
mc2155 prompted us to examine Ag85A for Acyl-
CoA:DGAT activity. We found that Ag85A mediates TAG
formation from long-chain acyl-CoA as acyl donor and
1,2-dipalmitoyl-sn-glycerol (1,2-dipalmitin) as the acyl
acceptor (Fig. 7) as is known for DGAT enzymes from other
organisms (Wältermann et al., 2007). The DGAT activity of
Ag85A is in the same range as is reported for fungal DGAT2
© 2011 Blackwell Publishing Ltd, Molecular Microbiology, 81, 1577–1592
(isolated from lipid bodies) (Lardizabal et al., 2001) and
His-Sco0958 DGAT from Streptomyces coelicolor. (Arabo-
laza et al., 2008). The enzyme exhibits a high Km of 390 mM
as reported previously for M. tuberculosis TGS1 (Sirakova
et al., 2006).
The known Acyl-CoA:DGAT such as DGAT1 or
DGAT2 belong to two distinct protein families, have dif-
ferent physiological roles and exhibit no sequence
homology to each other (Kalscheuer and Steinbüchel,
2003; Yen et al., 2005; 2008). Ag85A does not belong to
either of the two known DGAT families. It does not
contain putative DGAT motifs, which have been pro-
posed by Liu et al. (2010) or Kalscheuer and Stein-
büchel (2003). The DGAT activity of Ag85A includes two
consecutive reactions: fatty acyl-CoA hydrolysis (thio-
esterification), and transfer of the acyl group to the dia-
cylglycerol (transesterification). Ag85A shares these two
catalytic activities with other serine hydrolases such as
thioesterases (Bellizzi et al., 2000; Cao et al., 2009;
Dias et al., 2010) and carboxylesterases (Satoh and
Hosokawa, 1998; Bencharit et al., 2006). Thioesterases
exhibit only acyl-CoA hydrolase activity, while some
eukaryotic carboxylesterases also exhibit additional acyl-
transferase (ACAT) activity like human carboxylesterase
1 (Satoh and Hosokawa, 1998; Bencharit et al., 2006).
Thioesterases and carboxylesterases share a catalytic
mechanism where a conserved serine hydrolyses the
substrate by a nucleophilic attack. The Ag85A homo-
logue Ag85C is proposed to act as a serine hydrolase
exhibiting mycolyl esterase/transferase activity. Ag85A
contains the same catalytic triad as Ag85C formed by
residues Ser126, His262 and Glu230 and possesses a
deep substrate binding groove near the active-site
serine as reported for Ag85B and C (Ronning et al.,
2000; Anderson et al., 2001). The positions of the sub-
strates in the docking model suggest a similar mecha-
nism for Ag85A as proposed for Ag85C (Ronning et al.,
2000). The active-site Ser126, verified by the analysis of
a Ser126Ala mutant, allows a nucleophilic attack by the
serine hydroxyl group on the carboxylate carbon of the
fatty acyl-coenzyme, forming an acyl-enzyme intermedi-
ate. Finally the short distance of the hydroxyl group of
the acyl acceptor diacylglycerol would permit a nucleo-
philic attack on the carboxylate carbon of the acyl-
enzyme adduct. These findings, establish for the first
time a link between cell wall and TAG biosynthesis,
which will shed more light and expand our understand-
ing of the role of Ag85A in M. tuberculosis pathogenesis
and persistence. Since the current data do not comprise
final proof of Ag85A involvement in TAG synthesis under
dormant conditions in M. tuberculosis we have initiated
further work on construction of suitable k.o. mutants and
their detailed characterization under various growth con-
ditions including dormancy.

Experimental procedures
Gene cloning and transformation into M. smegmatis
mc2155
The gene coding for Ag85A (FbpA; Rv3804c) was amplified by
PCR using genomic DNA from M. tuberculosis H37Rv as
template. The sequence of PCR sense primer for Rv3804c
was 5′-ggt acc gga tcc atg cag ctt gtt gac agg gt-3′ and that of
anti-sense primer was 5′-ggc tgt atc gat cta gtg atg gtg atg gtg
atg ggc gcc ctg ggg cgc ggg cc-3′. The PCR products were
purified by agarose gel-electrophoresis followed by extraction
using the QIAquick kit (Qiagen, Germany). After cleavage with
restriction endonucleases FastDigest BamHI and FastDigest
ClaI (Fermentas) fragments were purified and ligated in
pMV261 (Stover et al., 1991), digested with the same restric-
tion endonucleases to generate plasmid pMV261::85A. The
ligation mixture was used to transform electrocompetent cells
of E. coli Top10 F⬘ (Invitrogen). The cells were grown in
Luria–Bertani (LB) broth and LB agar supplemented with
kanamycin (20 mg ml-1) and incubated at 37°C. pMV261::85A
was isolated from transformed colonies and plasmid integrity
was confirmed by DNA sequencing (Department of Genom-
analytik, HZI Braunschweig). For transformation into
M. smegmatis mc2155 pMV261::85A was isolated using the
QIAprep Spin Miniprep Kit (Qiagen, Germany). Preparation
and transformation of M. smegmatis mc2155 was carried out
as described before (Pelicic et al., 1996) with minor modifica-
tions. The resulting suspensions from electroporation were
incubated at 37°C for 4 h and then plated on kanamycin
(20 mg ml-1)-containing plates (Middlebrook 7H10 agar
supplemented with ADC, 0.2% glycerol and 0.05% Tween 80.
Transformed colonies were confirmed by PCR amplification
and DNA sequencing. The M. smegmatis strain mc2155, har-
bouring the plasmid pMV261::85A was named M. smegmatis-
Rv3804c.
Gene cloning, transformation into E. coli BL21 and
production of Ag85A
For kinetic analysis, Ag85A was produced in large amounts
as recombinant 6¥His-tagged protein. The gene encoding the
secreted form of antigen 85A (FbpA; Rv3804c) was amplified
by PCR. The sequence of PCR sense primer for Rv3804c
was 5′-ggc tgt cat atg gca ttt tcc cgg ccg ggc tt-3′ and that of
anti-sense primer was 5′-ggt acc gga tcc cta ggc gcc ctg ggg
cgc gg-3′. The PCR product was digested with FastDigest
NdeI and FastDigest BamHI (Fermentas). The agarose-gel-
purified DNA was ligated with T4 DNA ligase (Roche) in
double-digested pET-28a (+) (Novagen). The final plasmid
pET28::Ag85A contained the DNA sequence coding for
antigen 85A with an N-terminal 6¥histidine tag. The expres-
sion plasmid pAg85A was transformed into BL21 (DE3) cells
(Novagen). The correct orientation and integrity of inserts
were confirmed by PCR amplification and DNA sequence
analysis. BL21 (DE3) cells containing pET28::Ag85A were
cultivated and the recombinant protein was purified as
reported before (Elamin et al., 2009).
Site-directed mutagenesis
Site-directed mutagenesis was performed by use of a
QuikChange site-directed mutagenesis kit (Stratagene). The
© 2011 Blackwell Publishing Ltd, Molecular Microbiology, 81, 1577–1592
Diacylglycerol acyltransferase activity of Ag85A 1587
introduction of single specific mutations into Ag85A to
mutated serine 126 to alanine (tcg to gcg) was achieved by
the use of pET28::Ag85A and pMV261::85A as templates
DNA together with the following pair of mutagenic primers:
forward 5′-gtc gtc ggt ctt gcg atg gct gct tc-3′ and reverse
5′-gaa gca gcc atc gca aga ccg acg ac-3′ (the nucleotide
changes giving the appropriate mutations are shown in
bold). After initial denaturation at 95°C for 0.5 min, the
cycling parameters were 0.5 min at 95°C followed by 1 min
at 55°C and 7 min at 68°C (16 cycles). The reaction mix-
tures were placed on ice for 2 min, and then the parental,
supercoiled double-stranded DNA was digested with 1 ml of
DpnI at 37°C for 1 h before being transformed into compe-
tent E. coli Top10 F⬘. Mutation was verified by DNA
sequencing. The expression and purification of pET28::85A*
(mutated 85A) was done by using BL21 (DE3) cells
(Novagen) as described (Elamin et al., 2009). The isolation
and transformation of pMV261::85A* (mutated Ag85A) into
M. smegmatis mc2155 was done as described above. The
M. smegmatis strain mc2155, harbouring the plasmid
pMV261::85A* was named M. smegmatis-mRv3804c.
Expression of Ag85A in M. smegmatis-Rv3804c and
M. smegmatis-mRv3804c
Mycobacterium smegmatis mc2155 was grown in 7H9
medium (Difco) supplemented with ADC, 0.2% glycerol and
0.05% Tween 80. Induction of the recombinant gene was
done by shifting the incubation temperature from 37 to 45
degrees for 2 h if not otherwise stated.
Preparation for field emission scanning electron
microscopy (FESEM)
Mycobacterium smegmatis mc2155, M. smegmatis-Rv3804c
and M. smegmatis mc2155-pMV261 were grown at 37°C for
48 h in 75 ml flasks containing 25 ml of Middlebrook 7H9
broth, supplemented with ADC, 0.2% glycerol and 0.05%
Tween 80 and kanamycin (20 mg ml-1) (except for M. smeg-
matis mc2155) and diluted to obtain 103–104 cells ml-1. The
diluted suspensions were added to 100 ml flasks containing
Middlebrook 7H9 broth until an OD600 of 0.8–1 was reached,
and then the cultures were incubated at 45°C for additional
2 h. Samples of 5 ml were fixed with a solution containing 2%
glutaraldehyde and 5% formaldehyde in cacodylate buffer
(0.1 M cacodylate, 0.09 M sucrose, 0.01 M MgCl2, 0.01 M
CaCl2, pH 6.9) for 1 h on ice and washed with TE buffer
(20 mM TRIS, 2 mM EDTA, pH 6.9). After washing with TE
buffer samples placed onto poly-l-lysine-coated slides and
then dehydrated with 10%, 30%, 50%, 70%, 90%, 100%
acetone, each step for 10 min on ice. The last 100% acetone
step was performed at room temperature. Samples were then
critical-point-dried with liquid CO2 (CPD 030, Bal-Tec, Liecht-
enstein) mounted onto aluminium stubs and sputter coated
with gold (SCD 500, Bal-Tec, Liechtenstein) before examina-
tion in a Zeiss field emission scanning electron microscope
Gemini DSM852 (Zeiss, Oberkochen, Germany) at an accel-
eration voltage of 5 kV using the Everhart-Thornley SE detec-
tor and the inlens SE detector in a 50:50 ratio. Images are
stored onto a 230 MB MO disk.
1588 A. A. Elamin et al. 䊏
Transmission electron microscopy
The same collected samples as described above were fixed
in a fixation solution containing 5% formaldehyde and 2%
glutaraldyhde in cacodylate buffer (0.1 M cacodylate, 0.01 M
CaCl2, 0.01 M MgCl2, 0.09 M sucrose, pH 6.9) and washed
with cacodylate buffer. Samples were then osmificated with
1% aqueous osmium for 1 h at room temperature, washed
and pellets of the samples were embedded in 2% water
agar and cut into small cubes. Dehydration was achieved
with a graded series of acetone (10%, 20% and 50%) for
30 min on ice followed by contrasting with 2% uranyl acetate
in 70% acetone for overnight at 4°C and further dehydrated
with 90% and 100% acetone. Samples in the 100% acetone
step were allowed to reach room temperature and were infil-
trated with the epoxy resin according to Spurr’s formula for
a medium resin: 1 part 100% acetone/1 part resin for over-
night, 1 part 100% acetone/2 parts resin for 8 h, pure resin
for overnight and several changes the following 2 days.
Samples were then transferred to resin filled gelatine cap-
sules and polymerized for 8 h at 75°C. Ultrathin sections
were cut with a diamond knife, picked up with formvar-
coated copper grids (300 mesh) and counterstained with 2%
aqueous uranyl acetate and lead citrate. After air-drying
samples were examined in a Zeiss transmission electron
microscope TEM910 at an acceleration voltage of 80 kV.
Images were recorded digitally with a Slow-Scan CCD-
Camera (ProScan, 1024 ¥ 1024, Scheuring, Germany) with
ITEM Software (Olympus Soft Imaging Solutions, Münster,
Germany).
Immunoelectron microscopic labelling
Mycobacterium smegmatis mc2155 and M. smegmatis-
Rv3804c cells which treated as described were fixed with 1%
formaldehyde directly in the growth medium for 2 h at 4°C,
washed twice with cacodylate buffer containing 10 mM
glycine. Samples were dehydrated with a graded series of
ethanol (10%, 30%, 50%, 70%, 90% and 100%) and embed-
ded in LRWhite resin. Ultrathin sections were cut with a
diamond knife and collected onto formvar coated nickel grids
(300 mesh). Sections were incubated with mouse anti-85A
mAb IgG (Lionex, Germany) for 12 h at 4°C (1:5 dilution).
After washing with PBS bound antibodies were made visible
with protein A gold complexes (15 nm in diameter, 1:200
dilution of the stock solution, incubation for 1 h at room
temperature). After washing in PBS and distilled water sec-
tions were counterstained with 4% aqueous uranyl acetate
for 2 min before examination in a Zeiss transmission electron
microscope EM910 at an acceleration voltage of 80 kV and
calibrated magnifications. Images were recorded digitally
with a Slow-Scan CCD-Camera (ProScan, 1024 ¥ 1024,
Scheuring, Germany) with ITEM-Software (Olympus Soft
Imaging Solutions, Münster, Germany).
Bodipy labelling
Bodipy dyes have potential applications as stains for neutral
lipids in eukarotic as well as in prokaryotic cells for staining
intracellular lipid droplets (Gocze and Freeman, 1994). One
© 2011 Blackwell Publishing Ltd, Molecular Microbiology, 81, 1577–1592
millilitre of M. smegmatis mc2155, M. smegmatis-Rv3804c
and M. smegmatis-mRv3804c cultures which were grown as
described were centrifuged, washed with PBS and resus-
pended in 90 ml of PBS. Ten microlitres of a 1 mg ml-1 solution
of Bodipy 493/503 (Invitrogen) in ethanol was added and the
bacteria were incubated for 10 min in the dark at room
temperature. After washing twice with PBS 2 ml of the Bodipy-
labelled bacteria were put onto a glass slide, covered with a
coverslip and examined using a Zeiss photomicroscope with
an attached Hrc camera and Axiovision software 4.7. For time
collection experiments the samples were collected from the
main culture at 4, 6, 8, 12 and 16 h and also after 2 h induc-
tion at 45°C.
Flow cytometry
Depending on given cell density, the cell number used for
measurement was adjusted leading to 700–1500 events per
second. Cell suspension volume used was 100 ml and the
flow rate was adjusted. The PI (Invitrogen) staining was
used to determine the vitality. PI was added to gain a final
concentration of 15 mM. The sample was incubated for
10 min and directly measured. The lipid bodies staining
were performed using Bodipy 493/503 (Invitrogen) as
described previously. Flow cytometry was conducted using
a BD FACS Canto II (BD Biosciences) equipped with an
air-cooled 15 mW 500 nm Argon-Laser. For both PI and
Bodipy 493/503 staining the same settings were used. Flow
cytometry data were analysed using the BD FACSDiva Soft-
ware (BD Biosciences).
Mycolic acids extraction and derivatization procedures
Mycobacterial cultures (0.5 ml) (M. smegmatis mc2155 and
M. smegmatis-Rv3804c) were harvested, washed and resus-
pended in 2 ml of saponification reagent (25% KOH in 50%
ethanol) per gram whole cells followed by saponification by
autoclaving for 1 h at 121°C. Saponified cells were cooled to
room temperature and acidified to pH 1 with 6 N HCI. The
mycolic acids were extracted by adding 2 ml of chloroform
per gram cells, mixed in a separation funnel to assist sepa-
ration of the chloroform layer. The extracted mycolic acids
were evaporated to dryness and resuspended in chloroform
for derivatization. The mycolic acids were derivatized to
p-bromophenacyl esters by the method of Durst et al. (1975)
by p-bromophenacyl-8 reagent (Pierce Chemical). Derivati-
zation was carried out for 45 min at 85°C in glass tubes. After
derivatization, the p-bromophenacyl esters of the mycolic
acids were filtered through 0.45-mm-pore nylon 66 membrane
filters, evaporated to dryness and resuspended in chloroform
for analysis.
HPLC conditions
HPLC of the derivatized mycolic acids was performed with an
HPLC model 450 data system controller (KNAUER’s Smart-
line, Germany). The UV light absorbing p-bromophenacyl
esters were detected with a variable-wavelength detector
(Smartline UV Detector 2500) set at 254 nm. A reverse-phase
C-18 cartridge column (22 cm by 2.1 mm; Pierce) was used

to separate the mycolic acid esters. A solvent gradient system
of chloroform and methanol was controlled by two Smartline
Pump 1000 solvent delivery pumps (KNAUER) adjusted
to a constant flow rate of 0.6 ml min-1. The column was
equilibrated to 10% chloroform-90% methanol over a 13 min
period. After the injection of 5 ml of samples, the solvent
concentration was changed linearly over a 1 min period to
25% chloroform-75% methanol; over the next 30 min period,
the solvent concentration was changed linearly to 70%
chloroform-30% methanol.
The HPLC fractions and total lipids were analysed by
thin-layer chromatography (TLC) on silica plates (Macherey-
Nagel, Germany) with the different solvents system. For ana-
lytical purposes, the spots were visualized with Coomassie
blue. Each developed TLC plate was air dried and immersed
in a staining solution consisting of 0.04% Brilliant blue R
(Sigma) in 25% methanol. After 30 min, the plate was
removed from the staining solution, immersed in 25% metha-
nol (destaining solution), and allowed to stand for 2–5 min
with occasional agitation. Then the plate was removed from
the destaining solution and excess liquid on the surface of the
plate was soaked up with filter paper by gentle pressing. The
box density was measured by Quantity one program (version
4.6.2) (Bio-Rad Laboratories).
Total lipids extraction and quantification of TMM
and TDM
Bacterial cultures (0.5 l) from M. smegmatis mc2155 and
M. smegmatis-Rv3804c were used for total lipids extraction
as described previously (Elamin et al., 2009). Quantification
of TMM and TDM from the total lipid of mycobacterium cells
was performed by the mycolyltransferase assay as described
earlier (Elamin et al., 2009). The final TMM and TDM concen-
tration was obtained by subtraction of the glucose concen-
tration of the negative control without antigen 85A from the
glucose concentration obtained by the reaction.
Determination of DGAT activity with purified
recombinant antigen 85A
The DGAT activity was assayed for purified recombinant pro-
teins from BL21 (DE3) (Ag85A and mutated Ag85A) by using
5,5-dithiobis (2-nitrobenzoic acid) (DTNB), which specifically
reacts with thiol groups of released CoA-SH (Arabolaza et al.,
2008). For the determination of kinetic constants of Ag85A
with palmitoleoyl-CoA as substrate the final concentration of
palmitoleoyl-CoA was varied between 25 and 2000 mM while
the concentration of 1,2-dipalmitin was kept constant at
2 mM. Stock solution of palmitoleoyl-CoA was dissolved in
CHAPS (5 mM). 1,2-Dipalmitin was suspended in CHAPS
(40 mM) and solubilized by adding drops of chloroform/
methanol (1:1 by vol.).
The reaction was carried out in a buffer containing 100 mM
potassium phosphate buffer (pH 7.0), 5% (v/v) ethanol and
10 mM MgCl2. The assay was performed in a total volume of
260 ml at 37°C. The reaction was initiated by adding pure
Ag85A (57 mg) to the mixture (6 mM in final concentration).
After 1 min an aliquot of 60 ml was taken and the reaction was
stopped by the addition of 60 ml of 1% (w/v) trichloroacetic
© 2011 Blackwell Publishing Ltd, Molecular Microbiology, 81, 1577–1592
Diacylglycerol acyltransferase activity of Ag85A 1589
acid. 100 ml of the supernatant was taken after centrifugation
(17 000 g for 10 min), and added to 100 ml of 2 mM DTNB
dissolved in Tris-HCl (pH 7.8) in a F-bottom 96-well plates
(Greiner Bio-One). The product formation was measured by
reading the absorbance at 405 nm with a microplate reader
MRX Revelation (Dynex Technologies GmbH, Germany).
The rate of product formation (nmol mg-1 min-1) was calcu-
lated using a standard curve of acetylcysteine from 0 to
100 mM. Kinetic data were calculated with GraphPad Prism
software using the Michaelis–Menten model (GraphPad Soft-
ware, La Jolla, CA, USA).
The apparent Km 1,2-dipalmitin for M. tuberculosis Ag85A
was determined at two concentrations of palmitoleoyl-CoA,
100 mM and 700 mM. The product rate was calculated as
mentioned above. In order to verify the acylation of diacylg-
lycerol by Ag85A to TAG, the assay was performed with
the same conditions as mentioned before for Ag85A (the
protein final concentration increased to 12 mM) with 700 mM
palmitoleoyl-CoA and different concentration of 1,2-
dipalmitin (0–2000 mM). The reaction mixture was incubated
at 37°C for 2 h, and stopped and extracted by the addition
of 1 ml of chloroform with gentle agitation for 6–12 h. The
mixture was allowed to stand for 15 min followed by sepa-
ration of the chloroform layer. The extracted lipids were
dried and resuspended in chloroform. The lipids were analy-
sed by TLC with n-hexane: Acetic acid (9:1, by vol.) as
mobile phase. Dipalmitin and tripalmitin (Sigma-Aldrich)
were used as references. Staining was performed with Coo-
massie blue in methanol as described before. The TLC spot
corresponding to tripalmitin was scratched from the TLC
plates and extracted by chloroform and analysed by MALDI-
TOF-MS.
Docking
Co-ordinates for the composition of the substrates were
obtained by the HIC up server (http://xray.bmc.uu.se/hicup/).
Palmitoleic acid was extracted from the PDB entry 1FK3.
Dipalmitoyl-sn-glycero-3-phosphate was extracted from the
PDB entry 1HG4. Acetyl coenzyme was extracted from 3F5O
(Cao et al., 2009). The substrates palmitoleoyl-CoA and
1,2-dipalmitin were constructed using the program Coot
(Emsley and Cowtan, 2004). Energy minimization, addition
of full charges and polar hydrogens to the substrate atoms
were performed by the Dundee PRODRG2 Server (http://
davapc1.bioch.dundee.ac.uk/prodrg/) (Schüttelkopf and
van Aalten, 2004). The co-ordinates of Ag85A were obtained
from the PDB entry 1SFR. AutoDock Tools (ADT) (http://
autodock.scripps.edu/resources/adt) (Sanner, 1999) was
used for the preparation of the pdbqt files (addition of hydro-
gens to the macromolecule and definition of rotatable torsion
angles for the ligand). The ligand palmitoleoyl-CoA was
docked within a search area defined by a grid box of size
x = 32, y = 24 and z = 16 with a spacing of 1.0 Å and centred
at x = 41.38, y = 62.42, z = 3.10, encompassing the trehalose
binding pocket.
Vina (Trott and Olson, 2010) was used for molecular
docking. The hypothetical palmitoleoyl-Ag85A adduct was
constructed with Coot by moving the molecule slightly so that
one carbonyl-oxygen was near to the oxygen of serine126 to
allow binding. 1,2-Dipalmitin was docked to the palmitoleoyl-
1590 A. A. Elamin et al. 䊏
Ag85A complex by Vina with the following parameters: grid
area x = 16, y = 26 and z = 16 with a spacing of 1.0 Å,
centred at x = 45.40, y = 68.016, z = 3.48. The program was
run with an exhaustiveness of 20.
Acknowledgements
We thank our colleagues, Wulf Oehlmann, Manfred Nimtz and
Ina Schleicher, for their help and technical support. Ayssar
Elamin is indebted to the Deutscher Akademischer Austaus-
chdienst (DAAD) for the award of doctoral scholarship.
References
Alvarez, H.M., and Steinbüchel, A. (2002) Triacylglycerols in
prokaryotic microorganisms. Appl Microbiol Biotechnol 60:
367–376.
Alvarez, H.M., Kalscheuer, R., and Steinbüchel, A. (2000)
Accumulation and mobilization of storage lipids by Rhodo-
coccus opacus PD630 and Rhodococcus ruber NCIMB
40126. Appl Microbiol Biotechnol 54: 218–223.
Anderson, D.H., Harth, G., Horwitz, M.A., and Eisenberg, D.
(2001) An interfacial mechanism and a class of inhibitors
inferred from two crystal structures of the Mycobacterium
tuberculosis 30 kDa major secretory protein (Antigen 85B),
a mycolyl transferase. J Mol Biol 307: 671–681.
Arabolaza, A., Rodriguez, E., Altabe, S., Alvarez, H., and
Gramajo, H. (2008) Multiple pathways for triacylglycerol
biosynthesis in Streptomyces coelicolor. Appl Environ
Microbiol 74: 2573–2582.
Armitige, L.Y., Jagannath, C., Wanger, A.R., and Norris, S.J.
(2000) Disruption of the genes encoding antigen 85A and
antigen 85B of Mycobacterium tuberculosis H37Rv: effect
on growth in culture and in macrophages. Infect Immun 68:
767–778.
Bass, J.B., Jr, Farer, L.S., Hopewell, P.C., O’Brien, R.,
Jacobs, R.F., Ruben, F., et al. (1994) Treatment of tuber-
culosis and tuberculosis infection in adults and children.
American Thoracic Society and The Centers for Disease
Control and Prevention. Am J Respir Crit Care Med 149:
1359–1374.
Beatty, W.L., and Russell, D.G. (2000) Identification of myco-
bacterial surface proteins released into subcellular com-
partments of infected macrophages. Infect Immun 68:
6997–7002.
Belisle, J.T., Vissa, V.D., Sievert, T., Takayama, K., Brennan,
P.J., and Besra, G.S. (1997) Role of the major antigen
of Mycobacterium tuberculosis in cell wall biogenesis.
Science 276: 1420–1422.
Bellizzi, J.J., 3rd, Widom, J., Kemp, C., Lu, J.Y., Das, A.K.,
Hofmann, S.L., and Clardy, J. (2000) The crystal structure
of palmitoyl protein thioesterase 1 and the molecular basis
of infantile neuronal ceroid lipofuscinosis. Proc Natl Acad
Sci USA 97: 4573–4578.
Bencharit, S., Edwards, C.C., Morton, C.L., Howard-Williams,
E.L., Kuhn, P., Potter, P.M., and Redinbo, M.R. (2006)
Multisite promiscuity in the processing of endogenous
substrates by human carboxylesterase 1. J Mol Biol 363:
201–214.
Brennan, P.J., and Nikaido, H. (1995) The envelope of
mycobacteria. Annu Rev Biochem 64: 29–63.
© 2011 Blackwell Publishing Ltd, Molecular Microbiology, 81, 1577–1592
Butler, W.R., Ahearn, D.G., and Kilburn, J.O. (1986) High–
performance liquid chromatography of mycolic acids as a
tool in the identification of Corynebacterium, Nocardia,
Rhodococcus, and Mycobacterium species. J Clin Micro-
biol 23: 182–185.
Cage, G.D. (1992) High-performance liquid chromatography
patterns of Mycobacterium gordonae mycolic acids. J Clin
Microbiol 30: 2402–2407.
Cao, J., Xu, H., Zhao, H., Gong, W., and Dunaway-Mariano,
D. (2009) The mechanisms of human hotdog-fold
thioesterase 2 (hTHEM2) substrate recognition and cataly-
sis illuminated by a structure and function based analysis.
Biochemistry 48: 1293–1304.
Cardona, P.J. (2007) New insights on the nature of latent
tuberculosis infection and its treatment. Inflamm Allergy
Drug Targets 6: 27–39.
Casadevall, A., and Pirofski, L.A. (2000) Host-pathogen inter-
actions: basic concepts of microbial commensalism, coloni-
zation, infection, and disease. Infect Immun 68: 6511–6518.
Coker, R.J. (2004) Review: multidrug-resistant tuberculosis:
public health challenges. Trop Med Int Health 9: 25–40.
Daffe, M., and Draper, P. (1998) The envelope layers of
mycobacteria with reference to their pathogenicity. Adv
Microb Physiol 39: 131–203.
Daniel, J., Deb, C., Dubey, V.S., Sirakova, T.D., Abomoelak,
B., Morbidoni, H.R., and Kolattukudy, P.E. (2004) Induction
of a novel class of diacylglycerol acyltransferases and tria-
cylglycerol accumulation in Mycobacterium tuberculosis as
it goes into a dormancy-like state in culture. J Bacteriol
186: 5017–5030.
Dias, M.V., Huang, F., Chirgadze, D.Y., Tosin, M., Spiteller,
D., Dry, E.F., et al. (2010) Structural basis for the activity
and substrate specificity of fluoroacetyl-CoA thioesterase
FlK. J Biol Chem 285: 22495–22504.
Ducati, R.G., Ruffino-Netto, A., Basso, L.A., and Santos, D.S.
(2006) The resumption of consumption – a review on
tuberculosis. Mem Inst Oswaldo Cruz 101: 697–714.
Durst, H.D., Milano, M., Kikta, E.J., Jr, Connelly, S.A., and
Grushka, E. (1975) Phenacyl esters of fatty acids via crown
ether catalysts for enhanced ultraviolet detection in liquid
chromatography. Anal Chem 47: 1797–1801.
Elamin, A.A., Stehr, M., Oehlmann, W., and Singh, M. (2009)
The mycolyltransferase 85A, a putative drug target of
Mycobacterium tuberculosis: development of a novel
assay and quantification of glycolipid-status of the myco-
bacterial cell wall. J Microbiol Methods 79: 358–363.
Emsley, P., and Cowtan, K. (2004) Coot: model-building tools
for molecular graphics. Acta Crystallogr D Biol Crystallogr
60: 2126–2132.
Garton, N.J., Christensen, H., Minnikin, D.E., Adegbola, R.A.,
and Barer, M.R. (2002) Intracellular lipophilic inclusions of
mycobacteria in vitro and in sputum. Microbiology 148:
2951–2958.
Garton, N.J., Waddell, S.J., Sherratt, A.L., Lee, S.M., Smith,
R.J., Senner, C., et al. (2008) Cytological and transcript
analyses reveal fat and lazy persister-like bacilli in tuber-
culous sputum. PLoS Med 5: e75.
Gocze, P.M., and Freeman, D.A. (1994) Factors underlying
the variability of lipid droplet fluorescence in MA-10 Leydig
tumor cells. Cytometry 17: 151–158.
Harth, G., Horwitz, M.A., Tabatadze, D., and Zamecnik, P.C.

(2002) Targeting the Mycobacterium tuberculosis 30/32-
kDa mycolyl transferase complex as a therapeutic strategy
against tuberculosis: proof of principle by using antisense
technology. Proc Natl Acad Sci USA 99: 15614–15619.
Epub 12002 Nov 15611.
He, Z., and De Buck, J. (2010) Cell wall proteome analysis of
Mycobacterium smegmatis strain MC2 155. BMC Microbiol
10: 121.
Jackson, M., Raynaud, C., Laneelle, M.A., Guilhot, C.,
Laurent-Winter, C., Ensergueix, D., et al. (1999) Inactiva-
tion of the antigen 85C gene profoundly affects the mycolate
content and alters the permeability of the Mycobacterium
tuberculosis cell envelope. Mol Microbiol 31: 1573–
1587.
Jarlier, V., and Nikaido, H. (1994) Mycobacterial cell wall:
structure and role in natural resistance to antibiotics. FEMS
Microbiol Lett 123: 11–18.
de Jong, B.C., Israelski, D.M., Corbett, E.L., and Small, P.M.
(2004) Clinical management of tuberculosis in the context
of HIV infection. Annu Rev Med 55: 283–301.
Kalscheuer, R., and Steinbüchel, A. (2003) A novel bifunc-
tional wax ester synthase/acyl-CoA:diacylglycerol acyl-
transferase mediates wax ester and triacylglycerol
biosynthesis in Acinetobacter calcoaceticus ADP1. J Biol
Chem 278: 8075–8082.
Lardizabal, K.D., Mai, J.T., Wagner, N.W., Wyrick, A.,
Voelker, T., and Hawkins, D.J. (2001) DGAT2 is a new
diacylglycerol acyltransferase gene family: purification,
cloning, and expression in insect cells of two polypeptides
from Mortierella ramanniana with diacylglycerol acyltrans-
ferase activity. J Biol Chem 276: 38862–38869.
Lewis, K. (2007) Persister cells, dormancy and infectious
disease. Nat Rev Microbiol 5: 48–56.
Liu, Q., Siloto, R.M., and Weselake, R.J. (2010) Role
of cysteine residues in thiol modification of acyl-
CoA:diacylglycerol acyltransferase 2 from yeast. Biochem-
istry 49: 3237–3245.
Low, K.L., Shui, G., Natter, K., Yeo, W.K., Kohlwein, S.D.,
Dick, T., et al. (2010) Lipid droplet-associated proteins are
involved in the biosynthesis and hydrolysis of triacylgly-
cerol in Mycobacterium bovis bacillus Calmette-Guerin.
J Biol Chem 285: 21662–21670.
McKinney, J.D., Honer zu Bentrup, K., Munoz-Elias, E.J.,
Miczak, A., Chen, B., Chan, W.T., et al. (2000) Persistence
of Mycobacterium tuberculosis in macrophages and mice
requires the glyoxylate shunt enzyme isocitrate lyase.
Nature 406: 735–738.
Meena, L.S., and Rajni (2010) Survival mechanisms of
pathogenic Mycobacterium tuberculosis H37Rv. FEBS J
277: 2416–2427.
Murphy, D.J., and Brown, J.R. (2007) Identification of gene
targets against dormant phase Mycobacterium tuberculo-
sis infections. BMC Infect Dis 7: 84.
Nguyen, L., Chinnapapagari, S., and Thompson, C.J. (2005)
FbpA-dependent biosynthesis of trehalose dimycolate is
required for the intrinsic multidrug resistance, cell wall
structure, and colonial morphology of Mycobacterium
smegmatis. J Bacteriol 187: 6603–6611.
Olukoshi, E.R., and Packter, N.M. (1994) Importance of
stored triacylglycerols in Streptomyces: possible carbon
source for antibiotics. Microbiology 140: 931–943.
© 2011 Blackwell Publishing Ltd, Molecular Microbiology, 81, 1577–1592
Diacylglycerol acyltransferase activity of Ag85A 1591
Pelicic, V., Reyrat, J.M., and Gicquel, B. (1996) Generation
of unmarked directed mutations in mycobacteria, using
sucrose counter-selectable suicide vectors. Mol Microbiol
20: 919–925.
Pieper-Fürst, U., Madkour, M.H., Mayer, F., and Steinbüchel,
A. (1995) Identification of the region of a 14-kilodalton
protein of Rhodococcus ruber that is responsible for the
binding of this phasin to polyhydroxyalkanoic acid granules.
J Bacteriol 177: 2513–2523.
Rao, P.K., and Li, Q. (2009) Protein turnover in mycobacterial
proteomics. Molecules 14: 3237–3258.
Raschke, D., and Knorr, D. (2009) Rapid monitoring of cell
size, vitality and lipid droplet development in the oleagi-
nous yeast Waltomyces lipofer. J Microbiol Methods 79:
178–183.
Ronning, D.R., Klabunde, T., Besra, G.S., Vissa, V.D.,
Belisle, J.T., and Sacchettini, J.C. (2000) Crystal structure
of the secreted form of antigen 85C reveals potential
targets for mycobacterial drugs and vaccines. Nat Struct
Biol 7: 141–146.
Ronning, D.R., Vissa, V., Besra, G.S., Belisle, J.T., and Sac-
chettini, J.C. (2004) Mycobacterium tuberculosis antigen
85A and 85C structures confirm binding orientation and
conserved substrate specificity. J Biol Chem 279: 36771–
36777. Epub 32004 Jun 36710.
Russell, D.G., Barry, C.E., 3rd, and Flynn, J.L. (2010) Tuber-
culosis: what we don’t know can, and does, hurt us. Science
328: 852–856.
Saha, S., Enugutti, B., Rajakumari, S., and Rajasekharan, R.
(2006) Cytosolic triacylglycerol biosynthetic pathway in
oilseeds. Molecular cloning and expression of peanut
cytosolic diacylglycerol acyltransferase. Plant Physiol 141:
1533–1543.
Sanner, M.F. (1999) Python: a programming language for
software integration and development. J Mol Graph Model
17: 57–61.
Satoh, T., and Hosokawa, M. (1998) The mammalian car-
boxylesterases: from molecules to functions. Annu Rev
Pharmacol Toxicol 38: 257–288.
Schüttelkopf, A.W., and van Aalten, D.M. (2004) PRODRG: a
tool for high-throughput crystallography of protein-ligand
complexes. Acta Crystallogr D Biol Crystallogr 60: 1355–
1363.
Sirakova, T.D., Dubey, V.S., Deb, C., Daniel, J., Korotkova,
T.A., Abomoelak, B., and Kolattukudy, P.E. (2006) Identifi-
cation of a diacylglycerol acyltransferase gene involved in
accumulation of triacylglycerol in Mycobacterium tubercu-
losis under stress. Microbiology 152: 2717–2725.
Stover, C.K., de la Cruz, V.F., Fuerst, T.R., Burlein, J.E.,
Benson, L.A., Bennett, L.T., et al. (1991) New use of BCG
for recombinant vaccines. Nature 351: 456–460.
Trott, O., and Olson, A.J. (2010) AutoDock Vina: improving
the speed and accuracy of docking with a new scoring
function, efficient optimization, and multithreading.
J Comput Chem 31: 455–461.
Wältermann, M., Hinz, A., Robenek, H., Troyer, D., Reichelt,
R., Malkus, U., et al. (2005) Mechanism of lipid-body for-
mation in prokaryotes: how bacteria fatten up. Mol Micro-
biol 55: 750–763.
Wältermann, M., Stöveken, T., and Steinbüchel, A. (2007) Key
enzymes for biosynthesis of neutral lipid storage com-
1592 A. A. Elamin et al. 䊏

pounds in prokaryotes: properties, function and occurrence
of wax ester synthases/acyl-CoA:diacylglycerol acyltrans-
ferases. Biochimie 89: 230–242.
Wong, M.Y., Steck, P.A., and Gray, G.R. (1979) The major
mycolic acids of Mycobacterium smegmatis. Characteriza-
tion of their homologous series. J Biol Chem 254: 5734–
5740.
Yen, C.L., Monetti, M., Burri, B.J., and Farese, R.V., Jr (2005)
The triacylglycerol synthesis enzyme DGAT1 also cata-
lyzes the synthesis of diacylglycerols, waxes, and retinyl
esters. J Lipid Res 46: 1502–1511.
Yen, C.L., Stone, S.J., Koliwad, S., Harris, C., and Farese,
R.V., Jr (2008) Thematic review series: glycerolipids. DGAT
enzymes and triacylglycerol biosynthesis. J Lipid Res 49:
2283–2301.
Supporting information
Additional supporting information may be found in the online
version of this article.
Please note: Wiley-Blackwell are not responsible for the
content or functionality of any supporting materials supplied
by the authors. Any queries (other than missing material)
should be directed to the corresponding author for the
article.

© 2011 Blackwell Publishing Ltd, Molecular Microbiology, 81, 1577–1592