Báo cáo khoa học: Molecular dissection of the biosynthetic relationship between phthiocerol and phthiodiolone dimycocerosates and their critical role in the virulence and permeability of Mycobacterium tuberculosis doc

Molecular dissection of the biosynthetic relationship
between phthiocerol and phthiodiolone dimycocerosates
and their critical role in the virulence and permeability of
Mycobacterium tuberculosis
Roxane Sime
´
one, Patricia Constant, Wladimir Malaga, Christophe Guilhot, Mamadou Daffe
´
and
Christian Chalut
De
´
partement Me
´
canismes Mole
´
culaires des Infections Mycobacte
´
riennes, Institut de Pharmacologie et de Biologie Structurale, Toulouse,
France
Mycobacteria are the agents of several important
human diseases, including tuberculosis and leprosy,
and remain an important cause of mortality and
morbidity worldwide. According to the World
Health Organization ( />Mycobacterium tuberculosis, the causative agent of
tuberculosis in humans, is responsible for more than
8 million new cases and kills 2 million people every
year. Little is known about the molecular mechanisms
of mycobacterial pathogenicity, but accumulated data
Keywords
ketoreductase; Mycobacterium tuberculosis;

phenolic glycolipids; phthiocerol
dimycocerosates; tuberculosis
Correspondence
C. Chalut, Institut de Pharmacologie et de
Biologie Structurale, 205 route de Narbonne,
31077 Toulouse Cedex, France
Fax: +33 5 6117 5994
Tel: +33 5 6117 5473
E-mail:
(Received 12 January 2007, revised 9
February 2007, accepted 14 February 2007)
doi:10.1111/j.1742-4658.2007.05740.x
Phthiocerol dimycocerosates and related compounds are important mole-
cules in the biology of Mycobacterium tuberculosis, playing a key role in
the permeability barrier and in pathogenicity. Both phthiocerol dimyco-
cerosates, the major compounds, and phthiodiolone dimycocerosates, the
minor constituents, are found in the cell envelope of M. tuberculosis, but
their specific roles in the biology of the tubercle bacillus have not been
established yet. According to the current model of their biosynthesis,
phthiocerol is produced from phthiodiolone through a two-step process in
which the keto group is first reduced and then methylated. We have previ-
ously identified the methyltransferase enzyme that is involved in this pro-
cess, encoded by the gene Rv2952 in M. tuberculosis. In this study, we
report the construction and biochemical analyses of an M. tuberculosis
strain mutated in gene Rv2951c. This mutation prevents the formation of
phthiocerol and phenolphthiocerol derivatives, but leads to the accumula-
tion of phthiodiolone dimycocerosates and glycosylated phenolphthiodio-
lone dimycocerosates. These results provide the formal evidence that
Rv2951c encodes the ketoreductase catalyzing the reduction of phthiodio-
lone and phenolphthiodiolone to yield phthiotriol and phenolphthiotriol,

which are the substrates of the methyltransferase encoded by gene Rv2952.
We also compared the resistance to SDS and replication in mice of the
Rv2951c mutant, deficient in synthesis of phthiocerol dimycocerosates but
producing phthiodiolone dimycocerosates, with those of a wild-type strain
and a mutant without phthiocerol and phthiodiolone dimycocerosates. The
results established the functional redundancy between phthiocerol and
phthiodiolone dimycocerosates in both the protection of the mycobacterial
cell and the pathogenicity of M. tuberculosis in mice.
Abbreviations
CFU, colony-forming unit; CI, competition index; DIM, diester of phthiocerol or phthiodiolone; DIM A, phthiocerol dimycocerosates; DIM B,
phthiodiolone dimycocerosates; Hyg, hygromycin; Km, kanamycin; PGL, phenolglycolipid; PGL-tb, phenolglycolipid from M. tuberculosis.
FEBS Journal 274 (2007) 1957–1969 ª 2007 The Authors Journal compilation ª 2007 FEBS 1957
strongly suggest that the mycobacterial cell envelope
plays a key role in both the pathogenesis and resist-
ance to the various hostile environments encountered
by pathogenic mycobacteria during the infection pro-
cess. The mycobacterial cell envelope is a complex and
unusual structure [1]. A key feature of this structure is
its very high lipid content, consisting of up to 60% of
the dry weight of the bacteria. Among these lipids, the
diesters of phthiocerol and phenolic glycolipids (PGLs)
have attracted attention for years. They are specifically
found in slow-growing species that include the patho-
genic species Mycobacterium leprae, Mycobacterium
ulcerans, Mycobacterium marinum and members of the
M. tuberculosis complex [2]. In M. tuberculosis, the
diesters of phthiocerol and phthiodiolone, called DIMs,
are composed of a mixture of long chain b-diols that
are esterified by multimethyl-branched fatty acids
named mycocerosic acids [2] (Fig. 1). The chemical

structures of PGLs are very similar to those of
DIMs, except that they harbor a phthiocerol chain
Fig. 1. Proposed biosynthetic pathway leading to DIM A and PGL-tb from DIM B and glycosylated phenolphthiodiolone dimycocerosates,
respectively. The keto, hydroxyl and methoxyl groups are boxed in rectangles. p, p¢ ¼ 2–4; n, n¢ ¼ 16–18; m2 ¼ 15–17; m1 ¼ 20–22;
R ¼ C
2
H
5
or CH
3
.
Phthiocerol dimycocerosates in M. tuberculosis R. Sime
´
one et al.
1958 FEBS Journal 274 (2007) 1957–1969 ª 2007 The Authors Journal compilation ª 2007 FEBS
x-terminated by an aromatic nucleus, the so-called
phenolphthiocerol, which in turn is glycosylated [2]
(Fig. 1). In M. tuberculosis the major constituents,
phthiocerol dimycocerosates (DIM A) and glycosylated
phenolphthiocerol dimycocerosates (PGL-tb), are usu-
ally accompanied by minor structural variants, called
phthiodiolone dimycocerosates (DIM B) and phenol-
phthiodiolone dimycocerosates, which contain a keto
group in place of the methoxy group at the terminus
of the b-diols (Fig. 1).
Several laboratories have shown that both DIMs
and PGLs contribute to the permeability barrier
formed by the cell envelope of M. tuberculosis and to
virulence [3–6]. Nevertheless, their precise molecular
mechanisms of action are still unknown, and the speci-

fic roles of the various members of the DIM family,
e.g. DIM A and DIM B, in these functions have never
been investigated. The recent developments in our
understanding of the biosynthetic pathway of DIMs
have provided the means to investigate the contribu-
tion of these compounds to the biology of M. tubercu-
losis. Considerable efforts, started more than 40 years
ago, have been devoted to deciphering the complex
biosynthetic pathways of DIMs and PGLs. These
efforts have led to the identification of more than 20
proteins required either for the formation or for the
translocation of these compounds [7]. Remarkably,
most of these genes are clustered in a 70 kb region of
the M. tuberculosis chromosome [3,4,8]. Several lines
of evidence indicate that DIM A is formed from
DIM B in a two-step process [9]: the reduction of the
keto function carried by DIM B into a hydroxyl
group, to form phthiotriol dimycocerosates, followed
by the methylation of the hydroxyl group, catalyzed
by a methyltransferase, to yield DIM A (Fig. 1). This
model is supported by the characterization of the two
enzymes responsible for these enzymatic reactions.
Indeed, we previously showed that the protein encoded
by Rv2952 catalyzes the methylation of phthiotriol and
phenolphthiotriol to form phthiocerol and phenol-
phthiocerol, respectively [10]. Subsequently, Onwueme
et al. [11] identified Rv2951c as being the gene enco-
ding the phthiocerol ketoreductase that catalyzes
the reduction of the keto moiety in phthiodiolone
during DIM biosynthesis. However, the role of

Rv2951c has been investigated in M. ulcerans and
Mycobacterium kansasii by gene complementation
studies, and no formal evidence that Rv2951c has the
same function in M. tuberculosis has been reported to
date.
In this article, we describe the contruction of an
Rv2951c M. tuberculosis mutant. Biochemical analyses
of this strain demonstrated that Rv2951c catalyzes
the reduction of the keto moiety of both DIM B and
phenolphthiodiolone dimycocerosates in M. tuber-
culosis. Comparison of the phenotypes of this Rv2951c
mutant with those of the wild-type strain and a mutant
deficient in DIM production demonstrates that DIM A
and DIM B fulfil redundant functions regarding both
the resistance of M. tuberculosis to SDS and its viru-
lence in the mouse model.
Results
Disruption of the Rv2951c gene in M. tuberculosis
H37Rv and biochemical characterization of the
Rv2951c gene-disrupted mutant
It has recently been proposed that Rv2951c encodes an
oxydoreductase involved in the reduction of the keto
group of DIM B to yield phthiotriol dimycocerosates,
which in turn would be methylated to give DIM A
[11]. This proposal was based on the observation that
some M. ulcerans and M. kansasii strains naturally
produce DIM B but not DIM A, and that M. ulcerans
harbored a mutation within the Rv2951c ortholog.
Gene complementation studies with a multicopy plas-
mid carrying a functional Rv2951c ortholog from

M. marinum partially restored DIM A synthesis in the
recombinant strains [11]. Although these studies
strongly suggested that Rv2951c catalyzes the reduc-
tion of the keto group of DIM B, the role of Rv2951c
remains to be demonstrated.
To formally establish the function of this enzyme
in the synthesis of DIM in M. tuberculosis, we construc-
ted an Rv2951c knockout M. tuberculosis H37Rv
mutant strain, named PMM74, by replacing the
wild-type allele of Rv2951c with a kanamycin (km)-dis-
rupted allele using the temperature-sensitive ⁄ sacB proce-
dure [12] (supplementary Fig. S1A). Lipids were then
extracted from the PMM74 mutant and analyzed by
TLC. As shown in Fig. 2A, the disruption of Rv2951c
in M. tuberculosis selectively abolished the production
of DIM A but not that of DIM B. When equivalent
amounts of lipids were loaded on TLC plates, we
observed that the mutant cells accumulated more
DIM B than did the wild-type cells. To further quantify
the lipids produced by the various strains, cells were
labeled with [1-
14
C]propionate, a precursor known to be
incorporated into methyl-branched fatty acyl-contain-
ing lipids, including DIMs. Analysis of the labeling
(Fig. 2B) confirmed that there were no traces of DIM A
in the DRv2951c::km mutant, and revealed that the
amounts of DIM B accumulated by the DRv2951c::km
mutant (46% of the labeled lipids) corresponded to
those of DIM A + DIM B produced by the wild-type

R. Sime
´
one et al. Phthiocerol dimycocerosates in M. tuberculosis
FEBS Journal 274 (2007) 1957–1969 ª 2007 The Authors Journal compilation ª 2007 FEBS 1959
cells (47% of the labeled lipids). Complementation of
the Rv2951c mutation by the introduction of a wild-type
allele Rv2951c in the PMM74 mutant fully restored the
production of DIM A (Fig. 2A), indicating that the
phenotypic differences observed between the mutated
and the wild-type strains relied solely on the disruption
of the Rv2951c gene. Together, these results esta-
blished that Rv2951c is involved in the biosynthesis of
DIM A and that DIM B is a precursor of DIM A in
M. tuberculosis.
To further characterize the structure of the DIM-
like substances produced by the PMM74 mutant,
lipids exhibiting R
f
values similar to those of DIM A
and DIM B were purified by preparative TLC and
analyzed by MALDI-TOF MS. The mass spectrum of
the purified lipids exhibiting a TLC mobility similar
to that of DIM B showed a series of pseudomolecular
ion (M +Na
+
) peaks at m ⁄ z 1346, 1360, 1374, 1388,
1402, 1416, 1430, 1444, 1458, 1472, and 1486
(Fig. 3A). These values were identical to those
observed in the mass spectrum of DIM B purified
from the wild-type strain (data not shown), confirm-

ing that PMM74 still produced DIM B. In sharp con-
trast, no pseudomolecular ion peaks that would
correspond to DIM A were seen in the mass spectrum
of purified lipids from PMM74 whose R
f
value was
similar to that of DIM A (Fig. 3B). Indeed, the
expected pseudomolecular ion (M +Na
+
) peaks were
observed in the mass spectrum of the DIM A from
the wild-type strain at m ⁄ z 1362, 1376, 1390, 1404,
1418, 1432, 1446, 1460, 1474, 1488, and 1502
(Fig. 3C). The m ⁄ z value for each peak in this series
is 16 mass units higher than that of DIM B, due to
the reduction of the keto function of DIM B, fol-
lowed by the methylation of the resulting hydroxyl
group. Thus, inactivation of the Rv2951c gene in
M. tuberculosis abolishes the production of DIM A
but does not affect that of DIM B.
Because the biosynthetic pathways of DIMs and
PGLs are known to be closely related, we next focused
on the role of Rv2951c in PGL-tb biosynthesis. Most
of the M. tuberculosis strains, such as H37Rv, are
devoid of PGL-tb, due to a frameshift mutation in the
pks15 ⁄ 1 gene, and production of PGL-tb can be
restored by introducing a functional pks15 ⁄ 1 gene into
these strains [13]. Accordingly, both the PMM74 and
wild-type strains were transformed with plasmid
pPET1 carrying a functional M. bovis BCG pks15 ⁄ 1

gene, and the structure of the expected PGL-tb was
determined after lipid extraction. The PMM74:pPET1
strain produced a major glycoconjugate exhibiting
slightly higher TLC mobility than PGL-tb pro-
duced from H37Rv:pPET1 (Fig. 2C). The glycolipids
B
C
A
D
Fig. 2. TLC analyses of lipids extracted from the M. tuberculosis
H37Rv DRv2951c::km and DppsE::km mutant strains. (A) TLC ana-
lysis of DIMs from M. tuberculosis H37Rv, the PMM74
(DRv2951c::km) mutant, and the PMM74:pRS01-complemented
strains. Lipids extracts were dissolved in CHCl
3
, loaded onto the
TLC plate, and run in petroleum ether ⁄ diethylether (90 : 10, v ⁄ v).
DIMs were visualized by spraying the plate with 10% phosphomo-
lybdic acid in ethanol, followed by heating. The positions of DIM A
(arrows) and DIM B (arrowheads) are indicated. (B) TLC analysis of
radiolabeled DIMs from M. tuberculosis H37Rv and the PMM74
mutant strains. Lipids were visualized by using a PhophorImager
system (Molecular Dynamics, Sunnyvale, CA, USA). The positions
of DIM A (arrow) and DIM B (arrowheads) are indicated. (C) TLC
analysis of glycolipids extracted from M. tuberculosis H37Rv,
H37Rv:pPET1, and PMM74:pPET1. Lipids were dissolved in CHCl
3
and run in CHCl
3
⁄ CH

3
OH (95 : 5, v ⁄ v). Glycoconjugates were visu-
alized by spraying the TLC plate with 0.2% anthrone (w ⁄ v) in con-
centrated H
2
SO
4
, followed by heating. The position of PGL-tb
(arrow) is indicated. (D) TLC analysis of radiolabeled DIMs extracted
from M. tuberculosis H37Rv and the PMM56 (DppsE::km) mutant
strains. Lipids were visualized by using a PhophorImager system.
The positions of DIM A and DIM B are indicated.
Phthiocerol dimycocerosates in M. tuberculosis R. Sime
´
one et al.
1960 FEBS Journal 274 (2007) 1957–1969 ª 2007 The Authors Journal compilation ª 2007 FEBS
produced by the PMM74:pPET1 and H37Rv:pPET1
strains were also analyzed by MALDI-TOF MS
following purification. The mass spectrum of the PGL-
like compound from the PMM74:pPET1 strain showed
a series of pseudomolecular ion (M +Na
+
) peaks
16 mass units lower than for PGL-tb from the
wild-type strain (series of M +Na
+
peaks at m ⁄ z 1864,
1878, 1892, 1906, 1920, 1934, 1948, 1962, 1976, 1990,
2004, and 2018) (Fig. 4). Because the Rv2951c gene
was shown to be involved in the modification of the

keto group of DIM B to yield DIM A, we speculated
that the glycolipid produced by the PMM74:pPET1
mutant might be a glycosylated phenolphthiodiolone
dimycocerosate. To confirm this hypothesis, this glyco-
lipid was further analyzed by
1
H-NMR (Fig. 5). All
the proton signal resonances typical of PGL-tb were
detected, with the notable exception of those corres-
ponding to the methoxyl group of the phenolphthio-
cerol dimycocerosate portion of PGL-tb expected at
3.32 p.p.m. (singlet, 3H) [14,15]. This observation was
consistent with the absence of the proton resonance at
2.85 p.p.m. corresponding to that of the methine pro-
ton of the carbon bearing the methoxyl group [14,15].
Furthermore, the proton resonances of the methyl
group b (Fig. 5) was observed at 1.05 p.p.m., instead
of 0.87 p.p.m. for the corresponding methyl resonances
in phthiocerol and phenolphthiocerol dimycocerosates
[16]. Together, these results clearly demonstrated that
the PGL produced by the PMM74 mutant strain is a
triglycosylated phenolphthiodiolone dimycocerosate,
most likely a tri-O-methyl-fucosyl-(a1–3)-rhamnosyl-
(a1–3)-2-O-methyl-rhamnosyl-a-phenolphthiodiolone
dimycocerosate, and therefore that Rv2951c is also
implicated in the biosynthesis of PGL-tb in M. tuber-
culosis by catalyzing the reduction of the keto group
of phenolphthiodiolone.
Construction of a DIM-less mutant of
M. tuberculosis H37Rv by disruption of the

ppsE gene
The construction of a M. tuberculosis mutant unable
to synthesize DIM A but proficient at DIM B produc-
tion, namely the PMM74 mutant strain, prompted us
to address the question of the specific role of DIM A
and DIM B in the biology of M. tuberculosis.We
chose to compare the phenotypes of the DRv2951c::km
mutant (DIM A
–
, DIM B
+
) with those of the
wild-type strain (DIM A
+
, DIM B
+
) and those of a
DIM-less mutant. The last of these was constructed by
insertion ⁄ deletion within the ppsE gene which encodes
a polyketide synthase required for the formation of the
b-diol chain [17,18] (supplementary Fig. S1B). Bio-
chemical analyses of the resulting mutant, named
PMM56, confirmed that it was unable to synthesize
either DIM A or DIM B (Fig. 2D).
A
B
C
Fig. 3. MALDI-TOF mass spectra of purified lipids exhibiting TLC
mobilities similar to those of DIM B (A) and DIM A (B) from
M. tuberculosis PMM74 and of DIM A (C) from M. tuberculosis

H37Rv (wild type).
R. Sime
´
one et al. Phthiocerol dimycocerosates in M. tuberculosis
FEBS Journal 274 (2007) 1957–1969 ª 2007 The Authors Journal compilation ª 2007 FEBS 1961
Effect of the absence of DIM A on the
susceptibility of M. tuberculosis to SDS
In a previous study, we demonstrated that DIMs are
involved in the resistance of the tubercle bacillus to
detergent, a feature related to the cell envelope per-
meability [4]. To determine the contribution of DIM A
and DIM B to this resistance, we compared the sensi-
tivity to SDS of three M. tuberculosis strains: the
wild type (DIM A
+
, DIM B
+
), the PMM74 mutant
(DIM A
–
, DIM B
+
), and the PMM56 mutant
(DIM A
–
, DIM B
–
). The three strains were incubated
with 0.1% SDS for 1, 4 and 8 days, and their survival
was then evaluated (Fig. 6). The DIM-less mutant

(PMM56) was much more sensitive to SDS than the
wild-type strain: after 1 day of exposure to the deter-
gent, the number of colony-forming units (CFUs) was
20-fold lower for PMM56 than for the wild-type
strain, confirming our previous observation [4]. In
sharp contrast, no difference in CFU number was
detected between the wild-type strain and the DIM A
mutant (PMM74) that was still able to synthesize
DIM B. The survival curves of the wild type and
PMM74 were almost superimposable for the first two
time points, and the number of CFUs was even higher
for PMM74 than for the wild type after 8 days of
incubation with SDS. These data indicate that the lack
of DIM A in PMM74 did not induce an important
structural cell wall modification.
Effect of the absence of DIM A on the virulence
of M. tuberculosis in mice
DIM deficiency has been shown to be associated with
virulence attenuation of M. tuberculosis in mice [3,4,8].
To investigate the role of DIM A and DIM B in this
phenotypic modification, we performed in vivo compe-
tition assays in mice to compare the virulence of
PMM56 (DIM-less) and PMM74 (DIM A-less) mutant
strains to that of the wild-type strain. Mice were
infected intranasally either with a mixture of
strains H37Rv:pMV361H and PMM74 or with a mix-
ture of strains H37Rv:pMV361H and PMM56. The
infectious dose used for each mouse was around
5 · 10
3

CFU, in a ratio (CFUs of mutant inocula-
ted ⁄ CFUs of wild type inoculated) of 0.722 for the
PMM74 ⁄ H37Rv:pMV361H inoculum and 0.944 for
the PMM56 ⁄ H37Rv:pMV361H inoculum, as deter-
mined by growth on 7H11 plates containing either
kanamycin (Km) or hygromycin (Hyg). Mice were
killed 1 or 21 days postinfection, and the mutant and
wild-type loads in lungs and spleens were determined
on the basis of growth on selective media.
As expected, 21 days postinfection, we observed a
marked difference between the number of CFUs of
the wild-type strain and that of the PMM56 mutant
strain in lungs and in the spleen (Fig. 7A). Indeed,
both strains multiplied in lungs, but whereas an aver-
age of 5.92 · 10
6
CFU was recovered from lungs for
the wild-type strain, only 3.91 · 10
4
CFU were recov-
ered for the PMM56 mutant strain. We also observed
a growth defect for the DIM-less mutant in the
spleen: on day 21, an average of 8.57 · 10
3
CFU
was recovered for the wild-type strain against
1.24 · 10
2
CFU for the PMM56 strain. A competition
index (CI) was determined by calculating the ratio of

mutant to wild-type bacteria after correcting for the
ratio of these strains in the inoculum. It appeared
that, 21 days after infection, the ratio of PMM56 to
wild-type bacteria was diminished more than 100-fold
(CI ¼ 9.28 · 10
)3
) in lungs and 18-fold (CI ¼
5.39 · 10
)2
) in the spleen, relative to the initial infect-
ing ratio (Fig. 7B). In contrast, mycobacteria deficient
in DIM A but synthesizing DIM B (the DRv2951c::km
mutant) did not show significantly attenuated growth
in comparison to the wild-type strain in mice. Indeed,
21 days postinfection, averages of 1.41 · 10
7
CFU and
A
B
Fig. 4. MALDI-TOF mass spectra of the purified glycolipid from
M. tuberculosis PMM74:pPET1 (A) and of PGL-tb from M. tubercu-
losis H37Rv:pPET1 (B).
Phthiocerol dimycocerosates in M. tuberculosis R. Sime
´
one et al.
1962 FEBS Journal 274 (2007) 1957–1969 ª 2007 The Authors Journal compilation ª 2007 FEBS
of 1.50 · 10
4
CFU, respectively, were recovered from
the lungs and the spleen for the wild-type strain,

against 1.10 · 10
7
CFU and 1.50 · 10
4
CFU for
the PMM74 mutant strain. These CFU counts esta-
blished that mice infected with PMM74 had similar
ratios of mutant to wild-type bacteria in both lungs
and spleen after 21 days as compared to the ratio of
bacteria in the initial inoculum (CI ¼ 1.04 and 1.44)
(Fig. 7B).
These mixed-infection experiments confirmed that
DIMs are major determinants for the pathogenicity of
M. tuberculosis. Moreover, the ability of the PMM74
mutant strain to replicate and persist in the lungs and
spleens of mice clearly indicates that DIM A is not
required for full virulence of M. tuberculosis when
DIM B is produced in the mycobacterial cell at the
same level as DIM A in the wild-type cell.
Discussion
The main objectives of this study were: (a) to fur-
ther characterize the biosynthetic pathways of two
Fig. 5. The
1
H-NMR spectrum of the purified glycolipid from PMM74:pPET1. The structure of the analyzed compound is shown above the
spectrum (p, p¢ ¼ 2–4; n, n¢ ¼ 16–18; m ¼ 15–17; R ¼ C
2
H
5
). The two doublets at 6.97 and 7.08 p.p.m. (g, h) are assigned to phenolic

proton resonances. Three anomeric proton resonances are seen at 5.50 p.p.m. (1H, i) and 5.15 p.p.m. (2H, i¢). The four singlets at
3.5–3.6 p.p.m. (j) are assigned to sugar-linked methoxyl proton resonances. The multiplet centered at 4.83 p.p.m. (a) is attributed to methine
resonances of esterified b-diol. The doublet at 1.15 p.p.m. (f) corresponds to the resonance of a methyl group in the a position of the fatty
acyl residues. The signals that correspond to the resonance of the methine proton of the a carbon (d) and that of the carbon bearing the
methyl group near the keto group (d¢) are observed at 2.55 p.p.m. Signals of several terminal methyl proton resonances are seen at
0.8–1.0 p.p.m. (e), consistent with the presence of multimethyl branched fatty acyl residues. The resonance of the protons of the methyl
group adjacent to the keto group is seen as a signal at 1.05 p.p.m. (b). The two arrows show the proton signal resonance positions corres-
ponding to the methoxyl group (expected at 3.32 p.p.m.) and the methine proton of the carbon bearing the methoxyl group (expected at
2.85 p.p.m.) of the phenolphthiocerol dimycocerosate portion in PGL-tb.
R. Sime
´
one et al. Phthiocerol dimycocerosates in M. tuberculosis
FEBS Journal 274 (2007) 1957–1969 ª 2007 The Authors Journal compilation ª 2007 FEBS 1963
important virulence factors for M. tuberculosis, i.e.
DIMs and PGL-tb; and (b) to address the question of
the function of the two related molecules DIM A and
DIM B. According to the accepted model of DIM and
PGL biosynthesis [17], DIM A and PGL-tb are expec-
ted to be derived from DIM B and glycosylated phe-
nolphthiodiolone dimycoserosates, respectively, after
two enzymatic steps: the reduction of the keto group
to give phthiotriol and glycosylated phenolphthiotriol
dimycocerosates, and the methylation of the hydroxyl
group by the previously identified methyltransferase
encoded by Rv2952 [10]. Onwueme et al. [11] have
demonstrated, in a recent study, that the lack of a
functional Rv2951c ortholog in some M. ulcerans and
M. kansasii strains was responsible for diacyl phthioc-
erol deficiency in these strains. In addition, they have
shown that complementation of M. ulcerans and

M. kansasii with a functional Rv2951c gene from
M. marinum leads to the accumulation of diacyl
phthiotriols [11], indicating that Rv2951c encodes the
phthiodiolone ketoreductase that catalyzes the forma-
tion of phthiotriol, the substrate of the Rv2952 methyl-
transferase in these strains.
In the present study, we constructed and biochemi-
cally characterized an M. tuberculosis strain harboring
a mutation within the Rv2951c gene. Upon transfer of
Fig. 6. Susceptibility to SDS of the M. tuberculosis wild-type
(DIM A
+
, DIM B
+
) (circle), PMM74 (DIM A
–
, DIM B
+
) (triangle) and
PMM56 (DIM A
–
, DIM B
–
) (square) mutant strains. Strains were
inoculated in 7H9 supplemented with ADC to which 0.1% SDS
was added. The number of viable bacteria was evaluated by plating
serial dilutions of the different cultures onto 7H11 supplemented
with OADC and incubation at 37 °C. The values shown are the
means ± standard deviations for three independent experiments.
B

A
*
Fig. 7. Competition between mutant and wild-type strains in infected mice. (A) Numbers of CFUs recovered from the lungs and the spleen
of mice infected with a mixture of the wild-type H37Rv:pMV361H and DIM-less PMM56 strains (left panel) or with a mixture of the wild-
type H37Rv:pMV361H and DIM A-less PMM74 strains (right panel), at 1 day (J1) or 21 days (J21) after infection. The CFU numbers were
determined by plating dilutions of homogenized tissues on 7H11 media containing either Km (for CFU counts of the mutant strains) or Hyg
(for CFU counts of the wild-type strain). White, black and gray bars represent the numbers of CFUs corresponding to the H37Rv:pMV361H,
the PMM56 and the PMM74 strains, respectively. Values are means ± standard deviations (error bars) of CFU counts for five infected mice.
*When spleen homogenates from the five mice infected with the H37Rv::pMV361H ⁄ PMM56 mixture were plated on 7H11 plates contain-
ing Km, no colonies were obtained for four mice, indicating that fewer than 50 bacteria of the PMM56 mutant strain were present in the
spleen of these mice. For the determination of the number of PMM56 bacteria present in the spleen of these four mice, we thus chose a
value of 50 CFU per spleen. The average number of CFUs recovered from the spleen of mice infected with the H37Rv::pMV361 ⁄ PMM56
mixture is therefore overestimated for the PMM56 mutant. (B) CI for the DIM-less PMM56 (white bars) and the DIM A-less PMM74 (gray
bars) mutant strains in the lungs and spleen of the infected mice after 21 days of infection. CI is defined as the ratio of mutant to wild-type
CFUs in the organ divided by the mutant to wild-type bacterial ratio in the inoculum.
Phthiocerol dimycocerosates in M. tuberculosis R. Sime
´
one et al.
1964 FEBS Journal 274 (2007) 1957–1969 ª 2007 The Authors Journal compilation ª 2007 FEBS
plasmid pPET1, this strain accumulated both DIM B
and glycosylated phenolphthiodiolone dimycoserosates,
but was unable to synthesize DIM A and PGL-tb.
This mutant strain did not produce the intermediates
phthiotriol and glycosylated phenophthiotriol dimyco-
cerosates found in the DRv2952::km mutant strain,
demonstrating that the products of Rv2952 and
Rv2951c are not involved in the same enzymatic step
of the DIM A and PGL-tb pathway. This phenotype
was not due to a polar effect on a downstream gene,
as the transfer of a functional Rv2951c gene carried on

a mycobacterial plasmid fully reversed this biochemical
phenotype. Thus, our data extend the results reported
by Onwueme et al.toM. tuberculosis, and further
demonstrate that the ketoreductase encoded by
Rv2951c is involved in the biosynthetic pathway of
PGL-tb by catalyzing the formation of phenolphthiot-
riol dimycoserosates.
Previous reports have established that DIMs are
important virulence factors of M. tuberculosis and
contribute to the cell envelope permeability barrier
[3–5,19]. However, these reports were based on the
phenotypic analysis of mutants deficient in both
DIM A and DIM B biosynthesis, and did not allow
the precise definition of the role of each of the mole-
cules. With our DIM-less and DIM A-less mutants
derived from the same parental strain, we first
addressed the question of the specific role of DIM A
and DIM B in the resistance of cells to SDS. We
demonstrated that the occurrence of comparable
amounts of DIM B can complement the absence of
DIM A with regard to the SDS resistance, suggesting
that the two molecules fulfill redundant functions
regarding the protection of the mycobacterial cell
against environmental attack. These results also sug-
gest that the methoxyl group located at the terminal
end of the phthiocerol chain does not significantly
contribute to the structural organization of the myco-
bacterial cell envelope. We next used the mouse infec-
tion model to investigate the effect of the lack of
DIM A production on virulence attenuation of the

resulting bacteria. We found that, unlike the DIM-
less mutant, the mutant deficient in DIM A replicated
as well as the wild-type strain in mice, indicating that:
(a) DIM A is not required per se for full virulence of
M. tuberculosis; and (b) DIM B can contribute to the
same extent as DIM A in pathogenicity when this
compound is overproduced in the mycobacterial cell.
Interestingly in this experiment, the DIM-less mutant
(PMM56) was clearly defective for growth in both
lung and spleen, which is consistent with previous
findings by Rousseau et al. [19]. These data are in
contrast with those of Cox et al. [3], who found that
DIMs are required for optimal growth in the lungs
but not in the spleen of mice. The discrepancy
between these results might be explained by the dif-
ferent genetic backgrounds of the M. tuberculosis
strains used in these studies or ⁄ and by different
experimental conditions. Indeed, Cox et al. [3] infec-
ted mice by intravenous injection with a high infec-
tious dose (10
6
CFU), whereas Rousseau et al. used,
like us, the intranasal route of infection and a low
infectious dose (10
4
CFU).
The precise molecular mechanisms by which DIMs
act in the course of infection are still unclear. It is
possible that DIMs contribute to pathogenicity pas-
sively, by protecting the tubercle bacillus against the

antimicrobial responses of the host during infection.
Indeed, several lines of evidence, including our SDS
experimental data, indicate that these molecules con-
tribute to the structural organization of the myco-
bacterial cell envelope and are involved in the cell
wall permeability barrier of mycobacteria [4]. The
growth defect observed for the DIM-less mutant in
mice could therefore result from altered cell wall per-
meability. In contrast, the PMM74 mutant strain
may behave like the wild-type strain in mice because
this mutant has an unaffected cell envelope organiza-
tion. Alternatively, the external localization of DIMs
in the cell envelope raises the possibility that these
compounds act in vivo by interacting with some com-
ponent of the host cell and by modulating the host
immune response to contain the infection. This
hypothesis is supported by recent data suggesting
that DIMs modulate the host immune response in
the very early steps of infection [19]. In that situ-
ation, it can be inferred from our results that
DIM B is able to fulfill the same function as DIM A
in this immunomodulation activity, suggesting that
the methoxyl group carried by DIM A is not
involved in this process.
Our study provides the first structure–function ana-
lysis of DIMs in the pathogenicity of the tubercle
bacillus. Nevertheless, more experiments are required
to clarify the precise roles played by DIMs in the cell
envelope architecture and in virulence. The generation
of M. tuberculosis mutants producing DIM derivatives,

such as the PMM74 mutant, could be very useful for
addressing this issue. Indeed, the biochemical charac-
terization of these mutants and the analysis of their
growth characteristics in various cellular and animal
models may lead to the identification of the structural
motif(s) of DIMs involved in pathogenesis, and
thereby provide clues to decipher the mechanism by
which these compounds contribute to the pathogenesis
of tuberculosis.
R. Sime
´
one et al. Phthiocerol dimycocerosates in M. tuberculosis
FEBS Journal 274 (2007) 1957–1969 ª 2007 The Authors Journal compilation ª 2007 FEBS 1965
Experimental procedures
Bacterial strains, growth media and culture
conditions
Plasmids were propagated at 37 °CinEscherichia coli DH5a
in LB broth or LB agar (Invitrogen, Cergy Pontoise, France)
supplemented with either Km (40 lgÆmL
)1
) or Hyg
(200 lgÆmL
)1
). M. tuberculosis H37Rv, PMM56 and
PMM74 strains were grown at 37 °C in Middlebrook 7H9
broth (Invitrogen) containing ADC (0.2% dextrose, 0.5%
BSA fraction V, 0.0003% beef catalase) and 0.05% Tween-
80 when necessary, and on solid Middlebrook 7H11 broth
containing ADC and 0.005% oleic acid (OADC). For bio-
chemical analyses, mycobacterial strains were grown as sur-

face pellicles on Sauton’s medium. When required, Km and
Hyg were used at concentrations of 40 lgÆmL
)1
and
50 lgÆmL
)1
, respectively. Sucrose 2% (w ⁄ v) was used to sup-
plement 7H11 for the construction of the PMM74 mutant.
General DNA techniques
Molecular cloning experiments were performed using stand-
ard procedures. The cloning vectors used were pGEM-T
(Promega, Lyon, France) and pPR27 [20]. Mycobacterial
genomic DNA was extracted from 5 mL of saturated cul-
tures as previously described [21]. PCR experiments for
plasmid constructions or genomic analysis were performed
with standard conditions on a GeneAmp PCR system 2700
thermocycler (Applied Biosystems, Courtaboeuf, France).
PCR was performed in a final volume of 50 lL containing
2.5 units of Pfu DNA polymerase (Promega).
Construction of M. tuberculosis H37Rv
gene-disrupted mutants
A 2749 bp DNA fragment containing the Rv2951c gene
of M. tuberculosis H37Rv flanked by 760 and 843 bp at the
5¢- and 3¢-end, respectively, was amplified by PCR from
genomic DNA using oligonucleotides 2951A and 2951B
(Table 1), and cloned into pGEM-T to give pCG163. An
internal Rv2951c fragment of 801 bp was removed by a ClaI
digestion and substituted by a km resistance cassette [22] to
yield pCG167. This plasmid was subsequently digested by
PmeI, and the 4542 bp fragment that contained the disrup-

ted Rv2951c gene and its flanking regions was purified and
inserted at the XbaI site of pPR27, a mycobacterial thermo-
sensitive suicide plasmid harboring the counterselectable
marker sacB, to give plasmid pCG175. This vector was elec-
trotransformed in M. tuberculosis H37Rv, and transform-
ants were selected on 7H11 supplemented with OADC and
Km at 32 °C [12]. Two clones were selected and grown in
5 mL of 7H9 medium containing Tween-80 and Km at
32 °C for 3 weeks. Several dilutions of these cultures were
then plated onto 7H11 agar plates containing OADC, Km
and 2% sucrose, and incubated at 39 °C. PCR screening for
disruption of Rv2951c was performed with a set of specific
primers (2951C; 2951D; 2951E; res1; res2) (Table 1) after
extraction of the genomic DNA from several Km- and
sucrose-resistant colonies. One clone giving the correspond-
ing pattern for disruption of Rv2951c was selected for fur-
ther analyses and named PMM74 (supplementary Fig. S1A).
To construct a DIM-less mutant of H37Rv, we chose to
inactivate the ppsE gene, one of the genes shown to be
involved in the formation of the phthiocerol backbone
[17,18]. We used the strategy described by Bardarov et al.
[23]. A 2660 bp fragment of the ppsE gene was amplified
using M. tuberculosis genomic DNA and primers ppsE1 and
ppsE2 (Table 1) in a final volume of 50 lL containing
2.5 units of Taq DNA polymerase (Roche Molecular Bio-
chemicals, Meylan, France). The PCR fragment was inserted
within the vector pGEM-T to give pWM39. The km resist-
ance cassette was then inserted between the KpnI and BglII
sites of the ppsE gene fragment generating a 523 bp deletion
to yield pWM40. The PmeI fragment from pWM40 was then

cloned within the cosmid vector pYUB854 [23]. The resulting
Table 1. Oligonucleotides used in this study.
Gene Oligonucleotide Sequence (5¢-to3¢)
Rv2951c 2951A GCTCTAGAGTTTAAACGATCTCATTGTTGGGGCGC
2951B GCTCTAGAGTTTAAACATAGTCAATGAACTTGTACGC
2951C AGGAAGGCCGGCAAATGGC
2951D TTCACGTGAGATAAGCTCCC
2951E ACGGTTTCGGTGAAGCCAG
2951L ACAATTAATTAACAGTATGTACGAGCGATGCG
2951M ACAAAAGCTTGGCGCAAATCATAGCTTCTTG
ppsE ppsE1 GACTAGTTTAAACGGATCGACGAGTTCGACGC
ppsE2 GACTAGTTTAAACGAGGCACTGTGACCAGATGC
ppsE3 CGTTCTGGAGCAACCTTCG
ppsE4 GGTCGAGGAAGTACGTGAC
res res1 GCTCTAGAGCAACCGTCCGAAATATTATAAA
res2 GCTCTAGATCTCATAAAAATGTATCCTAAATCAAATATC
Phthiocerol dimycocerosates in M. tuberculosis R. Sime
´
one et al.
1966 FEBS Journal 274 (2007) 1957–1969 ª 2007 The Authors Journal compilation ª 2007 FEBS
cosmid, named pWM41, was then cut with PacI and ligated
with the mycobacteriophage phAE87. The ligation products
were encapsidated in vitro using the Gigapack III XL kit
(Stratagene, La Jolla, CA, USA), and the mix was used to
infect E. coli HB101, following the manufacturer’s recom-
mendations. Transfectants were selected on LB plates con-
taining Km. A recombinant phagemid containing the
disrupted gene construct was selected and named phWM01.
This phagemid was transferred by electroporation in
M. smegmatis, and phage particules were prepared as des-

cribed previously [23]. These particles were then used to
infect M. tuberculosis H37Rv, and M. tuberculosis allelic
exchange mutants were selected on 7H11 agar plates supple-
mented with OADC and Km. Five clones were then analyzed
using primers res1, res2, ppsE3, and ppsE4 (Table 1). One
clone, named PMM56, gave the pattern corresponding to all-
elic exchange and was retained for further analysis (supple-
mentary Fig. S1B).
Complementation of the M. tuberculosis H37Rv
Rv2951c gene-disrupted mutant
A region covering the Rv2951c gene and 173 bp upstream of
the start codon was PCR-amplified from M. tuberculosis
H37Rv genomic DNA using oligonucleotides 2951L and
2951M (Table 1). The PCR product was purified, digested
with PacI and HindIII endonuclease restriction enzymes,
and cloned between the PacI and HindIII sites of pMV361-
eHyg, a pMV361 derivative lacking the pHsp60 promoter
and carrying a hyg resistance marker [24]. The resulting
plasmid, named pRS01, was transferred into PMM74 cells
by electrotransformation, and tranformants were selected on
7H11 agar plates supplemented with OADC, Km and Hyg.
Extraction and purification of DIMs and PGLs
For each strain, bacterial cells were harvested from 200 mL
cultures and killed by incubation in CHCl
3
⁄ CH
3
OH (1 : 2,
v ⁄ v) for 1 day at room temperature. Lipids were extracted
twice with CHCl

3
⁄ CH
3
OH (1 : 1, v ⁄ v), washed twice with
water, and dried before analysis. Extracted mycobacterial
lipids were analyzed by TLC after resuspension in CHCl
3
at a final concentration of 20 mgÆmL
)1
. Equivalent
amounts of lipids from each extraction were spotted onto
silica gel G60 plates (20 · 20 cm, Merck, Paris, France)
and separated with either petroleum ether ⁄ diethylether
(90 : 10, v ⁄ v) for DIM analysis or CHCl
3
⁄ CH
3
OH (95 : 5,
v ⁄ v) for PGL analysis. DIMs and PGLs were visualized by
spraying the plates with 10% phosphomolybdic acid in eth-
anol and with a 0.2% anthrone solution (w ⁄ v) in concen-
trated H
2
SO
4
, followed by heating, respectively. DIMs and
PGLs were purified by preparative TLC as described above,
and recovered by scraping silica gel from the plates.
DIM A and DIM B were quantified by radiolabeling
newly synthesized lipids of M. tuberculosis. Briefly, 20 lCi

(7.4 · 10
5
Bq) of sodium [1-
14
C]propionate (specific activ-
ity, 2.03 · 10
12
BqÆmol
)1
; MP Biomedicals, Illkirch, France)
were added to log-phase cultures of the wild-type M. tuber-
culosis H37Rv, and the PMM74 and the PMM56 mutant
strains, and incubated for 20 h with continuous shaking.
Lipids were then extracted, and separated by TLC as des-
cribed before, and labeled compounds were quantified using
a PhosphorImager (Amersham Pharmacia Biosciences,
Little Chalfont, UK).
MALDI-TOF MS
MALDI-TOF MS was performed using a voyager DE-STR
MALDI-TOF instrument (PerSeptive Biosystems, Framing-
ham, MA, USA) equipped with a pulse nitrogen laser emit-
ting at 337 nm, as previously described [25]. Samples were
analyzed in the reflector mode using an extraction delay time
set at 100 ns and an accelerating voltage operating in positive
ion mode of 20 kV. The mass spectra were mass assigned by
external calibration. Samples (1 lLofa1mgÆmL
)1
solution
in CHCl
3

) were directly applied onto the sample plate. The
matrix solution [0.5 lL of 2,5-dihydroxybenzoic acid at
10 mgÆmL
)1
in CHCl
3
⁄ CH
3
OH (1 : 1, v ⁄ v)] was added. The
sample was then allowed to crystallize at room temperature.
NMR spectroscopy
The
1
H-NMR spectra were recorded at 295 K on a Bruker
(Wissembourg, France) DMX 500 apparatus using a 5 mm
broadband inverse (BBI) probe. Samples were dissolved in
100% CDCl
3
. The chemical shifts obtained by NMR
spectroscopy were assigned using chloroform as the refer-
ence for protons (7.23 p.p.m).
SDS resistance assays
Cultures of strains H37Rv, PMM56 and PMM74 of
M. tuberculosis were grown to mid-logarithmic phase in
7H9 supplemented with ADC and Km when necessary, and
centrifuged at 10 000 g for 10 min at room temperature
using a Jouan CR412 centrifuge with T4 swing out rotor
(Jouan, Saint-Herblain, France). The cell concentrations
were adjusted to allow the inoculation of 10 mL cultures at
a final D

600
of 0.02 with 100 lL of bacterial suspension.
SDS was added to a final concentration of 0.1%, and cul-
tures were incubated for 8 days at 37 °C. Aliquots were col-
lected after 0, 1, 4, and 8 days of growth, and the number
of viable bacteria was evaluated by plating serial dilutions
on 7H11 medium.
Competition assays in mice
For the mixed-infection experiments in mice, the M. tuber-
culosis H37Rv wild-type strain was previously transformed
R. Sime
´
one et al. Phthiocerol dimycocerosates in M. tuberculosis
FEBS Journal 274 (2007) 1957–1969 ª 2007 The Authors Journal compilation ª 2007 FEBS 1967
with pMV361H, a mycobacterial plasmid derived
from pMV361 [24] by replacement of the km cassette
by a hyg resistance marker. Inocula were prepared by
mixing H37Rv:pMV361H cells with either PMM74
or PMM56 mutant cells at a final concentration of
1.25 · 10
5
CFUÆmL
)1
for each strain in a final volume
of 500 lL. The actual CFUs of each strain in both inocula
were determined by plating serial dilutions on 7H11 plates
containing either Km or Hyg before infection.
Ten BALB ⁄ c mice were infected intranasally [26] with
20 lL(5· 10
3

CFU) of the H37Rv:pMV361H ⁄ PMM56
mixture, and 10 mice were infected with 20 lL(5· 10
3
CFU) of the H37Rv:pMV361H ⁄ PMM74 mixture. For each
infection group, five mice were killed 1 day after infection
to determine the number of CFUs that seeded in lungs, and
the remaining five mice were killed 21 days after infection
to count the number of bacteria present in lungs and
spleen. Bacteria were recovered from the spleen and lungs
by homogenizing tissues in 5 mL of NaCl ⁄ P
i
containing
0.05% Tween-80, and the number of viable bacteria in the
organs of infected mice were determined by plating serial
dilutions of homogenates on 7H11 plates containing either
Km or Hyg.
All investigations with mice were carried out according
to the CNRS guidelines for animal experimentation, and
were approved by the ‘Comite
´
re
´
gional d’e
´
thique pour
l’expe
´
rimentation animale’.
Acknowledgements
We are grateful to Drs Franc¸ oise Laval and Anne

Lemassu (IPBS, Toulouse) for their valuable assistance
with MS and NMR spectroscopy, respectively. This
work was supported by the ‘Agence Nationale de la
Recherche’ (grant ANR-06-MIME-037-01).
References
1 Daffe
´
M & Draper P (1998) The envelope layers of
mycobacteria with reference to their pathogenicity. Adv
Microb Physiol 39 , 131–203.
2 Daffe
´
M & Lane
´
elle MA (1988) Distribution of phthio-
cerol diester, phenolic mycosides and related compounds
in mycobacteria. J Gen Microbiol 134, 2049–2055.
3 Cox JS, Chen B, McNeil M & Jacobs WR Jr (1999)
Complex lipid determines tissue-specific replication of
Mycobacterium tuberculosis in mice. Nature 402,
79–83.
4 Camacho LR, Constant P, Raynaud C, Lane
´
elle MA,
Triccas JA, Gicquel B, Daffe
´
M & Guilhot C (2001)
Analysis of the phthiocerol dimycocerosate locus of
Mycobacterium tuberculosis. Evidence that this lipid is
involved in the cell wall permeability barrier. J Biol

Chem 276, 19845–19854.
5 Goren MB, Brokl O & Schaefer WB (1974) Lipids of
putative relevance to virulence in Mycobacterium tuber-
culosis: correlation of virulence with elaboration of sul-
fatides and strongly acidic lipids. Infect Immun 9,
142–149.
6 Reed MB, Domenech P, Manca C, Su H, Barczak AK,
Kreiswirth BN, Kaplan G, Barry CE 3rd (2004) A gly-
colipid of hypervirulent tuberculosis strains that inhibits
the innate immune response. Nature 431, 84–87.
7 Onwueme KC, Vos CJ, Zurita J, Ferreras JA & Quadri
LE (2005) The dimycocerosate ester polyketide virulence
factors of mycobacteria. Prog Lipid Res 44, 259–302.
8 Camacho LR, Ensergueix D, Perez E, Gicquel B &
Guilhot C (1999) Identification of a virulence gene
cluster of Mycobacterium tuberculosis by signature-
tagged transposon mutagenesis. Mol Microbiol 34,
257–267.
9 Gastambide-Odier M, Delaume
´
ny JM & Lederer E
(1963) Biosynthesis of phthiocerol: incorporation
of methionine and propionic acid. Chem Ind 31,
1285–1286.
10 Perez E, Constant P, Laval F, Lemassu A, Laneelle
MA, Daffe
´
M & Guilhot C (2004) Molecular dissection
of the role of two methyltransferases in the biosynthesis
of phenolglycolipids and phthiocerol dimycoserosate in

the Mycobacterium tuberculosis complex. J Biol Chem
279, 42584–42592.
11 Onwueme KC, Vos CJ, Zurita J, Soll CE & Quadri LE
(2005) Identification of phthiodiolone ketoreductase, an
enzyme required for production of mycobacterial diacyl
phthiocerol virulence factors. J Bacteriol 187,
4760–4766.
12 Pelicic V, Reyrat JM & Gicquel B (1996) Generation of
unmarked directed mutations in mycobacteria, using
sucrose counter-selectable suicide vectors. Mol Microbiol
20, 919–925.
13 Constant P, Perez E, Malaga W, Lane
´
elle MA, Saurel
O, Daffe
´
M & Guilhot C (2002) Role of the pks15 ⁄ 1
gene in the biosynthesis of phenolglycolipids in the
Mycobacterium tuberculosis complex. Evidence that all
strains synthesize glycosylated p-hydroxybenzoic methyl
esters and that strains devoid of phenolglycolipids har-
bor a frameshift mutation in the pks15 ⁄ 1 gene. J Biol
Chem 277, 38148–38158.
14 Daffe
´
M, Lacave C, Lane
´
elle MA & Lane
´
elle G (1987)

Structure of the major triglycosyl phenol-phthiocerol of
Mycobacterium tuberculosis (strain Canetti). Eur J Bio-
chem 167, 155–160.
15 Daffe
´
M, Lane
´
elle MA, Lacave C & Lane
´
elle G (1988)
Monoglycosyldiacylphenol-phthiocerol of Mycobac-
terium tuberculosis and Mycobacterium bovis . Biochem
Biophy Acta 958, 443–449.
16 Daffe
´
M & Servin P (1989) Scalar, dipolar-correlated
and J-resolved 2D-NMR spectroscopy of the specific
Phthiocerol dimycocerosates in M. tuberculosis R. Sime
´
one et al.
1968 FEBS Journal 274 (2007) 1957–1969 ª 2007 The Authors Journal compilation ª 2007 FEBS
phenolic mycosides of Mycobacterium tuberculosis. Eur
J Biochem 185, 157–162.
17 Kolattukudy PE, Fernandes ND, Azad AK,
Fitzmaurice AM & Sirakova TD (1997) Biochemistry
and molecular genetics of cell-wall lipid biosynthesis in
mycobacteria. Mol Microbiol 24, 263–270.
18 Trivedi OA, Arora P, Vats AM, Ansari MZ, Tickoo R,
Sridharan V, Mohanty D & Gokhale RS (2005) Dissect-
ing the mechanism and assembly of a complex virulence

mycobacterial lipid. Mol Cell 17, 631–643.
19 Rousseau C, Winter N, Pivert E, Bordat Y, Neyrolles
O, Ave P, Huerre M, Gicquel B & Jackson M (2004)
Production of phthiocerol dimycocerosates protects
Mycobacterium tuberculosis from the cidal activity of
reactive nitrogen intermediates produced by macro-
phages and modulates the early immune response to
infection. Cell Microbiol 6, 277–287.
20 Pelicic V, Jackson M, Reyrat JM, Jacobs WR Jr,
Gicquel B & Guilhot C (1997) Efficient allelic exchange
and transposon mutagenesis in Mycobacterium
tuberculosis. Proc Natl Acad Sci USA 94, 10955–
10960.
21 Belisle JT & Sonnenberg MG (1998) Isolation of geno-
mic DNA from mycobacteria. Methods Mol Biol 101,
31–44.
22 Malaga W, Perez E & Guilhot C (2003) Production
of unmarked mutations in mycobacteria using site-
specific recombination. FEMS Microbiol Lett 219,
261–268.
23 Bardarov S, Bardarov S Jr, Pavelka MS Jr, Sambanda-
murthy V, Larsen M, Tufariello J, Chan J, Hatfull G &
Jacobs WR Jr (2002) Specialized transduction: an
efficient method for generating marked and unmarked
targeted gene disruptions in Mycobacterium tuberculosis,
M. bovis BCG and M. smegmatis. Microbiology 148,
3007–3017.
24 Stover CK, de la Cruz VF, Fuerst TR, Burlein JE,
Benson LA, Bennett LT, Bansal GP, Young JF, Lee
MH, Hatfull GF et al. (1991) New use of BCG for

recombinant vaccines. Nature 351, 456–460.
25 Laval F, Lane
´
elle MA, De
´
on C, Monsarrat B & Daffe
´
M (2001) Accurate molecular mass determination of
mycolic acids by MALDI-TOF mass spectrometry. Anal
Chem 73, 4537–4544.
26 Pethe K, Alonso S, Biet F, Delogu G, Brennan MJ,
Locht C & Menozzi FD (2001) The heparin-binding
haemagglutinin of M. tuberculosis is required for extra-
pulmonary dissemination. Nature 412, 190–194.
Supplementary material
The following supplementary material is available
online:
Fig. S1. Construction and characterization of M. tuber-
culosis H37Rv nRv2951c::km and nppsE::km mutant
strains.
This material is available as part of the online article
from
Please note: Blackwell Publishing is not responsible
for the content or functionality of any supplementary
materials supplied by the authors. Any queries (other
than missing material) should be directed to the corres-
ponding author for the article.
R. Sime
´
one et al. Phthiocerol dimycocerosates in M. tuberculosis

FEBS Journal 274 (2007) 1957–1969 ª 2007 The Authors Journal compilation ª 2007 FEBS 1969