Studies on structural and functional divergence among
seven WhiB proteins of Mycobacterium tuberculosis
H37Rv
Md. Suhail Alam, Saurabh K. Garg* and Pushpa Agrawal
Institute of Microbial Technology, CSIR, Chandigarh, India
Mycobacterium tuberculosis has a remarkable ability to
survive under hostile conditions it encounters during
infection [1]. Despite extensive research directed
towards understanding the physiology of M. tuberculo-
sis and its molecular pathogenesis [1–3], many funda-
mental questions about the mechanisms of survival
during early infection and persistence remain poorly
understood. Among several intriguing questions, are:
(a) what are the bacterial determinants necessary for
early infection, (b) how does the bacterium counteract
or evade its host’s defenses to survive the vigorous host-
immune response, (c) what regulates the transition from
initial growth to persistence and back to active growth,
(d) are the bacteria present in a non-replicating ‘spore-
like’ state or do they replicate at all during latency, and
(e) how does the bacterium adapt to survive under the
anaerobic and nutritionally altered environment within
the granuloma? The answers to these questions are
likely to provide insight into the mechanisms by which
M. tuberculosis establishes infection and persists within
Keywords
iron–sulfur cluster; Mycobacterium
tuberculosis; protein disulfide reductase;
redox system; WhiB
Correspondence
P. Agrawal, Institute of Microbial

fide of insulin, a model substrate. However, the reduction efficiency varied
significantly. Surprisingly, WhiB2 did not reduce the insulin disulfide, even
though its basic properties were similar to those of others. The structural
and functional divergence among WhiB proteins indicated that each WhiB
protein is a distinguished member of the same family and together they
may represent a novel redox system for M. tuberculosis.
Abbreviations
ANS, 8-anilinonapthalene-1-sulfonate; GSH, reduced glutathione; GSSG, oxidized glutathione; IAA, iodoacetamide; ThT, thioflavin T;
Trx, thioredoxin.
76 FEBS Journal 276 (2009) 76–93 ª 2008 The Authors Journal compilation ª 2008 FEBS
the host and the means to eliminate latent infection, a
phase of the disease that poses the most significant
obstacle to the eradication of tuberculosis. To survive
and establish successful infection, M. tuberculosis
appears to have acquired a strong network of genes to
sense and respond to stress conditions; the properties of
many of these are poorly understood.
A family of genes, whiB, has received attention
because of their involvement in cell division (whiB2),
fatty acid metabolism and pathogenesis (whiB3), antibi-
otic resistance (whiB7) and in sensing a variety of stress
conditions [4–9]. Seven genes, whiB1 ⁄ Rv3219, whiB2 ⁄
Rv3260c, whiB3 ⁄ Rv3416, whiB4 ⁄ Rv3681c, whiB5 ⁄
Rv0022c, whiB6 ⁄ Rv3862c and whiB7 ⁄ Rv3197A, have
been identified in M. tuberculosis [10,11] as orthologs of
the whiB gene of Streptomyces coelicolor A3(2), which
has been shown to be involved in sporulation [12].
Although, WhiB proteins are annotated as putative
transcription factors [12], to date it has not been
shown directly that these proteins work as transcrip-

activity [17], redox sensing [18] and the coordination of
metal cofactors [19]. The functional importance of the
conserved cysteine residues in iron–sulfur cluster coor-
dination and protein disulfide reductase has been dem-
onstrated in WhiB4 [15]. Recently, cysteines of WhiB3
have also been shown to act as a ligand for the O
2
- and
NO-responsive [4Fe–4S] cluster [9].
The presence of four conserved cysteines and a
CXXC motif in WhiB proteins from M. tuberculosis
raises several questions: are all WhiB proteins coordi-
nated with an iron–sulfur cluster? If yes, then what are
their basic properties? Are the iron–sulfur clusters
equally oxidation labile? Does removal of the iron–
sulfur cluster lead to disulfide bond formation? Are the
structural features of mycobacterial WhiB proteins
similar? Do all WhiB proteins behave like protein
disulfide reductase? The objective of this study is to
answer several of the questions raised above.
This is the first study to report the biochemical and
biophysical properties of WhiB2, WhiB5, WhiB6 and
WhiB7 of M. tuberculosis and also compare the prop-
erties of all seven WhiB proteins. We show that, simi-
lar to WhiB3 and WhiB4, other freshly purified WhiB
proteins also coordinate a [2Fe–2S] cluster which
respond differently to the oxidizing environment.
Except WhiB2, apo WhiB5, WhiB6 and WhiB7 also
reduce insulin in vitro, but the efficiency of the reduc-
tion varies. An extensive biophysical study suggested

spectra of the purified proteins were recorded in the
Md. S. Alam et al. Molecular properties of M. tuberculosis WhiB proteins
FEBS Journal 276 (2009) 76–93 ª 2008 The Authors Journal compilation ª 2008 FEBS 77
range 200–700 nm. In addition to a peak at 280 nm,
two additional peaks at  333–340 and  420–
424 nm, along with two broad shoulders at  460 and
 560–580 nm were observed (Fig. 1). The peaks were
characteristic of an [2Fe–2S] cluster [20], therefore, it
was assumed that freshly purified WhiB1, WhiB2,
WhiB5, WhiB6 and WhiB7 also coordinated the [2Fe–
2S] cluster. The absorption spectra of different WhiB
proteins were largely indistinguishable, however, in
WhiB6 and WhiB7, the shoulder at  460 nm was
more prominent than in others. This subtle change in
the peak pattern may be because of their differential
electronic environment. The nature and type of amino
acids and their side-chain orientations around iron–
sulfur cluster coordination sites are the likely cause of
minor variations in the electronic properties, which
were reflected in their absorption spectra.
The brownish appearance of the protein purified in
the presence of 8 m urea indicated that the iron–sulfur
cluster of WhiB proteins had survived treatment by a
denaturant, a feature very similar to WhiB3 [14] and
WhiB4 [15]. Unlike proteins purified from the soluble
fraction, which had a spectral feature typical of the
[2Fe–2S] cluster, proteins in 8 m urea showed a single
peak at  400–415 nm (Fig. S3). The differential peak
features may be due to the solvent-induced confor-
mational change, which is possibly because of changes

IscS ⁄ Rv3025c, a cysteine desulfurase [9] of M. tuber-
culosis was cloned, expressed in E. coli and purified by
metal-affinity chromatography (data not shown). The
WhiB proteins were incubated in the reaction mixture
along with FeCl
3
, IscS and
35
S-cysteine. We observed
an IscS-dependent mobilization of sulfur from l-cyste-
ine to the iron–sulfur cluster of WhiB proteins
(Fig. 2A). In the control reactions, where IscS was
excluded or the iron concentration was limited (10-fold
less), we did not observe any signal (Fig. 2A). Further
characterization of the iron–sulfur cluster of the recon-
stituted samples could not be carried out because none
of the samples gave an EPR signal at 120K using
liquid nitrogen (data not shown). It is possible that a
further decrease in temperature (using liquid helium)
would be required in order to detect the EPR signal.
Nevertheless, the absorption spectra of the reconsti-
tuted proteins showed a single peak at  420 nm indi-
cating the presence of a [4Fe–4S] cluster (Fig. 2B). The
presence of a similar cluster has been reported in
WhiB3 and WhiB4.
The iron content of proteins purified from the solu-
ble fraction, from inclusion bodies, under denaturing
conditions and after refolding was similar (Table 1).
The data clearly suggested that the protein fold
responsible for holding the iron–sulfur cluster was

0.030
0.045
0.060
0.075
0.090
~560–580 nm
420 nm
340 nm
WhiB1, 50 µM
WhiB1, 50 µM, alkylated
WhiB2, 50 µM
WhiB2, 50 µM, alkylated
WhiB3, 50 µM
WhiB3, 50 µM, alkylated
WhiB4, 50 µM
WhiB4, 50 µM, alkylated
WhiB5, 50 µM
WhiB5, 50 µM, alkylated
WhiB6, 50 µM
WhiB6, 50 µM, alkylated
WhiB7, 50 µM
WhiB7, 50 µM, alkylated
Absorbance
λ
(nm)
λ
(nm)
λ
(nm)
λ

Absorbance
337 nm
300 350 400 450 500 550 600 650 700
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
~560– 580 nm
420 nm
337 nm
Absorbance
300 350 400 450 500 550 600 650 700
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
0.18
~560–580 nm
424 nm
333 nm

the purified proteins (50 l
M) with 20 mM IAA for 1 h at 25 °C in the dark and the spectra were recorded (thin line) after the baseline correc-
tion. The spectra for WhiB3 and WhiB4 are taken from Alam & Agrawal [14] and Alam et al. [15], respectievely.
Md. S. Alam et al. Molecular properties of M. tuberculosis WhiB proteins
FEBS Journal 276 (2009) 76–93 ª 2008 The Authors Journal compilation ª 2008 FEBS 79
clusters were most stable against air oxidation in
WhiB6 and WhiB7, and most labile in WhiB1.
To study the sensitivity towards oxidized glutathione
(GSSG), reduced glutathione (GSH) and dithiothreitol,
proteins were incubated with 10 mm of each agent and
the absorbance at 424 nm (A
424
) was recorded at dif-
ferent time intervals up to 42 h. All WhiB proteins
showed differential sensitivity towards oxidation by
GSSG, and similar to air oxidation, the iron–sulfur
clusters of WhiB6 and WhiB7 were comparatively
more stable (Fig. 3B). A reducing environment (in the
presence of GSH or dithiothreitol) significantly low-
ered the rate of disintegration of the iron–sulfur cluster
in each of the WhiB proteins (Fig. 3C,D). Therefore,
disassembly of the iron–sulfur cluster under oxidizing
conditions and its stability under reducing conditions
suggested that the iron–sulfur clusters of M. tuberculo-
sis WhiB proteins are redox sensitive. We assume that
the iron–sulfur clusters of different WhiB proteins
would respond differently to the oxidative stress
encountered by M. tuberculosis in vivo.
Iron–sulfur clusters of WhiB proteins are
differentially exposed to the external environment

M urea) 0.198 ± 0.050
WhiB6 (refolded) 0.208 ± 0.072
WhiB6 (alkylated) 0.007 ± 0.003
WhiB7 (native) 0.182 ± 0.035
WhiB7 (in 8
M urea) 0.189 ± 0.066
WhiB7 (refolded) 0.175 ± 0.020
WhiB7 (alkylated) 0.007 ± 0.005
+ + – + + +
10X 10X 10X 10X 10X 0.1X
WhiB
35
S-Cys
FeCl
3
IscS
– + + + + +
+ – + + + +
WhiB1
WhiB2
WhiB3
WhiB4
WhiB5
WhiB6
WhiB7
Auto radiogram
Ponceau S
stained
300 350 400 450 500 550 600
0.0

with methanol for 5 s and washed thoroughly. The membrane was
equilibrated with a buffer containing 50 m
M Tris ⁄ HCl, pH 9.0,
150 m
M NaCl and 10 mM dithiothreitol for 5 min. In total, 30 lL
(10 lL at one time) of the indicated samples were spotted, air dried
and developed using Phosphorimager (Bio-Rad, Hercules, CA,
USA). (B) Absorption spectra of a representative in vitro reconsti-
tuted WhiB protein. All seven WhiB proteins showed similar
features.
Molecular properties of M. tuberculosis WhiB proteins Md. S. Alam et al.
80 FEBS Journal 276 (2009) 76–93 ª 2008 The Authors Journal compilation ª 2008 FEBS
order to study the surface accessibility of the iron–
sulfur cluster of WhiB proteins, freshly purified proteins
were incubated with 20 mm EDTA and the degree of
iron chelation was monitored by recording the A
424
at
different intervals up to 20 h. Immediate chelation of
iron was not observed in any WhiB protein. However,
as time increased, the degree of iron chelation
increased and the extent of chelation varied (Table 2).
Almost 30% of the iron was chelated within 2 h in
WhiB1 and WhiB2, whereas in WhiB6 and WhiB7 it
was negligible over the same period. Even after 20 h of
incubation, iron chelation was  15–20% (minimum)
in WhiB6 and WhiB7, whereas it was  60% (maxi-
mum) in WhiB1 and WhiB2; in the other proteins it
varied between 20% and 60% (Table 2). From the
data, it appears that the surface accessibility of the

120
Effect of air
0 h
6 h
12 h
24 h
48 h
WhiB7
WhiB6
WhiB5
WhiB4
WhiB3
WhiB2
WhiB1
% Change in A
420
0
15
0303
45
0606
7575
0909
105
021021
Effect of GSSG (10 mM)
0 h
2 h
6 h
20 h

WhiB3
WhiB2
WhiB1
% Change in A
420
15
0
0303
45
0606
7575
0909
105
021021
Effect of dithiothreitol (10 mM)
0 h
2 h
6 h
20 h
30 h
42 h
WhiB7
WhiB6
WhiB5
WhiB4
WhiB3
WhiB2
WhiB1
% Change in
A

WhiB6 100 96 ± 3 89 ± 2 84 ± 2
WhiB7 100 98 ± 2 96 ± 4 80 ± 4
Md. S. Alam et al. Molecular properties of M. tuberculosis WhiB proteins
FEBS Journal 276 (2009) 76–93 ª 2008 The Authors Journal compilation ª 2008 FEBS 81
total increase in the mass after reduction of the disul-
fide bond reflects the total number of cysteines present
in the thiol and disulfide forms. In the oxidized state,
both WhiB2 and WhiB5 showed a major peak corre-
sponding to the theoretical molecular mass of the
recombinant protein. However, the reduced proteins
had increased molecular masses, representing alkyl-
ation of four cysteine residues in each case (Fig. 5).
Although, WhiB5 and WhiB6 did not show any mobil-
ity differences under reducing conditions, a similar
increase in mass was found after reduction (Fig. 5).
The difference in mass between the oxidized and
reduced forms suggested the presence of four cysteine-
thiols in the reduced apo WhiB proteins. Because none
of the cysteines was present in a thiol form in the oxi-
dized protein (except for one in WhiB6 which has five
cysteines), it was concluded that the apo form of all
WhiB proteins contained two intramolecular disulfide
bonds.
All apo WhiB proteins, except WhiB2, reduce the
insulin disulfide
Previously, we reported that apo WhiB1 [13] WhiB3
[14] and WhiB4 [15] are protein disulfide reductases.
The enzymatic activity of WhiB4 was shown to be gov-
erned by the CXXC motif [15]. Because the CXXC
motif is present in all WhiB family members of

Therefore, we assume that in WhiB proteins, one disul-
fide bond is formed between the two cysteines of
the CXXC motif (CXXXC in the case of WhiB5) and
the other between the remaining two cysteines. The
assumption is supported by our earlier data, in which
a similar arrangement of intramolecular disulfide
bonds in WhiB4 was established [15]. In WhiB6, one
of the intramolecular disulfide bonds appeared to have
formed between Cys53 and Cys56 but the involvement
of cysteines for the second bond is little hard to pre-
dict, as it contains five cysteines (Cys12, Cys34, Cys53,
Cys56 and Cys62).
Divergence in the secondary structure
composition of WhiB proteins of M. tuberculosis
The multiple sequence alignment of WhiB proteins of
M. tuberculosis showed 49–66% sequence homology
and 31–50% identity with respect to each other
(Table S1). However, because of the variation in amino
acid composition, it is possible that structural variations
may be an important determinant of their functional
properties in vivo. Therefore, the structural organization
of each M. tuberculosis WhiB protein was studied using
biophysical tools. The secondary structure of each
WhiB5
26.9
20.0
36.5
WhiB6
18.4
14.4

because their molar ellipticities varied significantly
(Fig. 6). The spectra showed a-helix, b-strand and
random coil features. However, the proportion of each
feature varied among WhiB proteins, as evident from
the difference in negative molar ellipticity at specific
wavelengths, i.e. 208 and 222 nm (a helix signature),
218 nm (b strand signature), 202–204 nm (random coil
signature). In WhiB5 and WhiB6, the structure was
dominated by a helices and b strands and the propor-
tion of these structural elements was higher in WhiB6.
WhiB1, WhiB2 and WhiB4 showed relatively increased
molar ellipticity at 202–204 nm, indicating the presence
of a significant proportion of random coils (Fig. 6). The
Fig. 5. MALDI-TOF spectroscopic analysis of oxidized and reduced form of apo WhiB proteins. ‘Oxd’ represents the ‘oxidized and alkylated’
protein, whereas ‘Red’ represents ‘reduced and alkylated’ protein.
Table 3. Protein disulfide reductase activity of apo WhiB proteins.
The data for each protein sample (3 l
M) are expressed as
means ± SD (three independent protein preparations).
Samples
Reductase activity
(· 10
)3
DA
650
nmÆmin
)2
)
WhiB1 4.78 ± 0.25
WhiB2 0.56 ± 0.16

–5000
–4000
–3000
–2000
–1000
0
1000
2000
WhiB5, Oxd
WhiB5, Red
λ
(nm)
190 200 210 220 230 240 250
λ
(nm)
190 200 210 220 230 240 250
λ
(nm)
190 200 210 220 230 240 250
λ
(nm)
190 200 210 220 230 240 250
λ
(nm)
λ
(nm)
190 200 210180 220 230
240 250
–6000
–5000

–10 000
–8000
–6000
–4000
–2000
0
2000
[
θ]
MRW
(deg·cm
2
·
dmol
–1
)
[
θ]
MRW
(deg·cm
2
·
dmol
–1
)
[
θ]
MRW
(deg·cm
2

(deg·cm
2
·
dmol
–1
)
[
θ]
MRW
(deg·cm
2
·
dmol
–1
)
[
θ]
MRW
(deg·cm
2
·
dmol
–1
)
WhiB1, Oxd
WhiB1, Red
Fig. 6. Secondary structure analyses of apo WhiB proteins. Far-UV
CD spectra of apo proteins (0.2 mgÆmL
)1
) were recorded at 25 °C.

recognition of a helices and less reliable for b strands,
thus the data obtained from CD spectroscopy are an
approximate assessment. To estimate the level of
b-structures in different WhiB proteins, a thioflavin T
(ThT)-binding assay was performed. ThT shows strong
fluorescence in the presence of crossed b-sheet struc-
tures [27,28]. The binding of ThT to each apo WhiB
protein was measured by fluorescence spectroscopy.
20 30 40 50 60 70 80 90 100
–8000
–6000
–4000
–2000
0
Temp (°C)
222 (deg·cm
2
·dmol
–1
)

WhiB1, Oxd
WhiB1, Red
–2400
–1600
–800
0
WhiB3, Oxd
WhiB3, Red
222 (deg·cm

WhiB4, Oxd
WhiB4, Red
222 (deg·cm
2
·dmol
–1
)

20 30 40 50 60 70 80 90 100
Temp (°C)
–4000
–3000
–2000
–1000
0
WhiB5, Oxd
WhiB5, Red
222 (deg·cm
2
·dmol
–1
)

20 30 40 50 60 70 80 90 100
Temp (°C)
–5000
–4000
–3000
–2000
–1000

with increasing temperature from 25 to 90 °C. The thermal denaturation kinetics of WhiB1 (0.2 mgÆmL
)1
) was studied as described in Garg
et al. [13], whereas the data of WhiB3 and WhiB4 were taken from Alam & Agrawal [14] and Alam et al. [15] respectively.
Md. S. Alam et al. Molecular properties of M. tuberculosis WhiB proteins
FEBS Journal 276 (2009) 76–93 ª 2008 The Authors Journal compilation ª 2008 FEBS 85
The fluorescence intensities of each protein varied with
respect to each other (Fig. 8). In WhiB5 and WhiB6 it
was several fold higher than in the others, whereas for
the rest it was within a similar range (Fig. 8). From
the data, it appears that b-sheet structures are a major
contributors in WhiB5 and WhiB6, but the proportion
is relatively low and similar in other WhiB proteins. It
should be noted that WhiB5 aligned only with WhiB3
and WhiB4, whereas the WhiB6 sequence aligned only
with WhiB4 (NCBI; />blast/bl2seq/wblast2.cgi) (Table S1), suggesting that at
the amino acid sequence level both WhiB5 and WhiB6
differ from the others, and the difference is clearly
reflected in their secondary structure. The other impor-
tant point is that although the WhiB proteins were
predicted to attain a a-helical structure [16], WhiB5
and WhiB6 appeared to have significant amounts of
b strand. In WhiB5 and WhiB6 the reduction of intra-
molecular disulfide bonds resulted in a slight decrease
in b-sheet structure (decreased fluorescence intensity),
whereas this was reversed in the other proteins.
Structural diversification at the tertiary structure
level among different WhiB proteins
Both CD spectroscopy and the ThT-binding assay pro-
vide information only about the secondary structure.

proteins. Polyclonal antibodies against all WhiB pro-
teins were raised in rabbit and their cross-reactivity
was tested by ELISA. Dilutions of primary antibodies
which gave A
450
= 0.5 ± 0.05 were selected for the
cross-reactivity assay and were as follows: 1 : 35 000
for WhiB1; 1 : 25 000 for WhiB2; 1 : 20 000 for
WhiB3, WhiB4 and WhiB5; 1 : 15 000 for WhiB6 and
1 : 50 000 for WhiB7. The activity of the positive con-
trol (antibody raised against that particular antigen)
was taken to be 100% and cross-reactivity with other
WhiB proteins was represented relative to the positive
0
20
40
60
150
200
250
300
WhiB7
WhiB6
WhiB5
WhiB4
WhiB3
WhiB2
WhiB1
Fluorescence itensity (at 482 nm)
Oxidized

tivity suggested that the antigenic sites (mostly confor-
mational epitopes) on different WhiB proteins are
heterogeneous. We assume that the conformational dif-
ference among various WhiB proteins is the possible
cause of this heterogeneity. Because WhiB proteins
share significant sequence homology minor cross-reac-
tivity may be due to the presence of antibodies which
are reacting to the linear epitopes or to conserved con-
formational epitopes. The data showed that there are
significant structural differences among WhiB proteins
of M. tuberculosis.
Discussion
The purified recombinant WhiB proteins had spectral
resemblance to proteins coordinating the [2Fe–2S] clus-
ter. Earlier studies also showed that recombinant
WhiB3 [9] and WhiB4 [15] of M. tuberculosis, purified
under normal conditions, coordinate a [2Fe–2S] clus-
ter. However, based upon the in vitro reconstitution
assay, both were found to be [4Fe–4S] cluster coordi-
nating proteins. Iron–sulfur clusters are one of the
most ancient and versatile cofactors of several impor-
tant class of proteins and have been implicated in vari-
ety of functions [29,30]. Iron and sulfur can be
combined in several ways to produce different cluster
types, e.g. [2Fe–2S], [3Fe–4S], [4Fe–4S] and more com-
plex structures [31]. The susceptibility of iron–sulfur
clusters to oxidation makes them a good sensor of
redox conditions within a cell [32,33]. It has been com-
monly observed that [4Fe–4S] clusters are highly oxi-
dation labile and rapidly transform into the relatively

120
AntiWhiB4 Ab
0
20
40
60
80
100
100
120
Anti-WhiB5 Ab
0
20
40
60
80
100
100
120
Anti-WhiB6 Ab
0
20
40
60
80
80
100
120
Anti-WhiB7 Ab
0

Protein ligands for the canonical clusters are typi-
cally sulfide ions of cysteines. The involvement of four
conserved cysteines in the coordination of the [4Fe–4S]
cluster has been confirmed in WhiB3 [9] and WhiB4
[15]. The brown color of the purified protein and the
iron–sulfur cluster-specific peaks also disappeared
when proteins were treated with IAA (Figs 1 and S2),
supporting the notion that the cysteines are the proba-
ble iron–sulfur cluster coordination sites in WhiB
proteins of M. tuberculosis. Because WhiB6 has five
cysteine residues, a detailed investigation is needed to
determine the exact coordination sites. Although the
arrangement of cysteines in the primary structure is
not identical, the ‘C-X
2
-C-X
5
-C’ arrangement (except
for WhiB5 where it is ‘C-X
3
-C-X
7
-C’) is conserved and
the position of the N-terminal cysteine appears to be
flexible. Therefore, in the primary structure, although
the cysteines lie far apart with respect to each other, in
a 3D form they are likely to come closer to hold the
iron–sulfur cluster. The formation of two intramolecu-
lar disulfide bonds upon cluster removal strengthened
the argument that in the 3D form, the cysteines are in

dependent detoxification system. Instead, it contains
mycothiol [43] and a mycothiol reductase [44]. In the
absence of functional oxyR, fnr, etc. [10,45] it appears
that M. tuberculosis has evolved several accessory
signaling molecules and a protective network which
remain unexplored. The presence of the WhiB family
in the form of disulfide reductase is likely to be an
important components of this unexplored system.
Although insulin is a non-natural substrate, the dif-
ference in activity indicates that the efficiency of reduc-
tion of target disulfides by WhiB proteins may vary
in vivo. Moreover, the involvement of a particular
WhiB protein in vivo would depend upon their struc-
tural features and redox potential. The redox potential
of disulfide oxidoreductases has been shown to be
largely determined by the amino acids present in and
around the active site (CXXC motif) [46,47]. WhiB
proteins of M. tuberculosis did not show any conserva-
tion in the dipeptide present between the two cysteines
of CXXC motif (Fig. S6). Therefore, we argue that
each protein is likely to have a different redox poten-
tial. The crystal structures of several thioredoxins have
shown that the active site is located at the N-terminus
of a-2 helix of thioredoxin fold and is separated by a
kink from rest of the helix due to presence of a proline
[25,48–50]. The kink provides proper positioning of
the active site to facilitate electron flow during the
thiol–disulfide exchange reactions [51]. Interestingly, in
all WhiB proteins except WhiB2, the amino acid imme-
diately after the C-terminal cysteine of the CXXC

than a DNA–protein interaction. To date, there has
not been a single report to show that the WhiB pro-
teins work as a DNA-binding protein. It is possible
that they act as an essential structural component of a
multiprotein complex and ⁄ or are directly or indirectly
involved in mediating the assembly of transcription
factors that regulate the expression of downstream
gene(s).
Iron–sulfur clusters are known to modulate the
structural and functional properties of several pro-
teins [29,56]. Because the cysteine-thiols would be
ligated to iron–sulfur cluster in holo WhiB proteins,
they are not free for electron flow and disulfide
exchange. Therefore, there is a high possibility that
removal of the cluster is essential for the reductase
activity of WhiB proteins. Indeed, the iron–sulfur
cluster of M. tuberculosis WhiB4 has a regulatory
role [15]. Removal of the iron–sulfur cluster under
oxidizing conditions is associated with conforma-
tional change followed by a gain of disulfide reduc-
tase activity.
A fundamental question is that if one thioredoxin-
related protein ‘can do it all’, why was evolution direc-
ted to choose multiple Trx-like proteins. We do not
have clear answer to this question, but in general, mul-
tigene families are thought to have evolved to address
questions of substrate specificity. In a scenario in
which there is a difference in structural organization
among WhiB proteins and because different whiB
genes are involved in different cellular processes, the

purged with nitrogen before storage or transfer. Standard
recombinant DNA techniques were followed as described
elsewhere [58].
Protein overexpression and purification
Gene-specific primers were used to amplify the complete
ORFs (without stop codon) of whiB2 ⁄ Rv3260c,
whiB5 ⁄ Rv0022c, whiB6 ⁄ Rv3862c, whiB7 ⁄ Rv3197A and
iscS ⁄ Rv3025c using genomic DNA of M. tuberculosis
H37Rv as a template. The nucleotide sequence of primers
is listed in Table S2. The authenticity of PCR products
was determined by sequencing both DNA strands in an
ABI Prism automated DNA Sequencer 310. PCR-ampli-
fied ORFs were cloned at EcoRI ⁄ XhoI(whiB 2 and
whiB6), EcoRI ⁄ SalI(whiB5 and whiB7) or SacI ⁄ HindIII
(iscS) sites of pET-29a to get pET-O (O defines the
respective ORFs) under isopropyl thio-b-d-galactoside-
inducible promoter. The pET-O transformed E. coli BL21
(DE3) cells were grown at 37 °C in Luria–Bertani broth
containing 30 lgÆmL
)1
kanamycin until D
600
was  0.5–
0.6 and expression was induced by the addition of
0.3 mm isopropyl thio-b-d-galactoside for 18–20 h at
16 °C (for WhiB proteins) or for 3 h at 30 °C (for IscS).
Cells from 1000 mL culture were re-suspended in 10 mL
buffer A (50 mm NaH
2
PO

air oxidation, readings from control samples were sub-
stracted from the test samples and the values thus obtained
are represented. The initial reading was set to 100% and
the change in A
420
(residual) is expressed relative to the
reading at t = 0. Suitable baseline corrections were made
before recording the absorbance.
In vitro assembly of iron–sulfur cluster by IscS
using radiolabeled cysteine
The purified proteins were dialyzed against buffer B
(50 mm Tris ⁄ HCl, pH 9.0, 150 mm NaCl, 10 mm dith-
iothreitol). Purified IscS from M. tuberculosis was used as
a cysteine desulfurase. The reaction (500 lL) was set in
buffer B and contained 30 lm WhiB protein, 2 lm IscS,
300 lm FeCl
3
, 300 lml-cysteine and 15 lCi
35
S-labeled
l-cysteine (BARC, Mumbai, India). The reaction was car-
ried out at 10 °C for 16–18 h in darkness. After incubation,
the samples were extensively dialyzed against buffer B and
absorption spectra were recorded. In order to study the
incorporation of
35
S into the iron–sulfur cluster and there-
fore in the WhiB proteins, 30 lL samples were spotted onto
a poly(vinylidene difluoride) membrane, air dried and visu-
alized by phosphorimaging.

)1
.
Data were recorded at every 0.5 °C, with an 8 s averaging
time and at a 1.5 nm bandwidth. For all spectral measure-
ments, a path length of 1 mm was used. Ten spectra were
averaged for each sample.
Fluorescence spectroscopy
All fluorescence measurements were recorded at 25 °Cina
Perkin–Elmer Luminescence Spectrometer LS50B (Wal-
tham, MA, USA). To measure the surface hydrophobicities,
10 lm of oxidized apo proteins in buffer C was mixed with
ANS (250 lm) and incubated for 2 min at 25 °C. There-
after, the samples were excited at 388 nm and fluorescence
emission spectra were recorded in the range 400–600 nm.
The surface hydrophobicity (SH) was calculated using a
formula:
SH(AU) ¼
Maximum emission fluorescence intensity (AU)
X
where X ¼
total number of hydrophobic amino acids
total number of amino acids in the protein
:
To study the binding of ThT to different WhiB
proteins, 50 lm ThT was added to 5 lm apo proteins in
buffer C. The excitation was set at 430 nm and the emis-
sion was recorded between 420 and 600 nm. All spectra
were recorded at a scan speed of 50 nmÆmin
)1
. The

given and blood was collected after 5 days. The serum was
isolated and the titer was determined by ELISA. Use of
animals for the generation of polyclonal antibodies was
carried out with prior approval from the Institutional Ani-
mal Ethics Committee (institutional registration number
55 ⁄ 1999 ⁄ CPCSEA, dated 11 November 1999). Animal
handling and experimental design was performed in accor-
dance with the approved guidelines.
Each well of a flat-bottom microtitre ELISA plate
was coated with 0.5 lg of purified WhiB protein (100
llÆwell
)1
) in 0.05 m bicarbonate buffer, pH 9.6, and incu-
bated for overnight at 4 °C. After blocking, different dilu-
tions of test sera (1 : 5000 to 1 : 80 000) prepared in
NaCl ⁄ P
i
containing 0.1% skimmed milk were added
(100 lL) to each well and incubated at 37 °C for 3 h. Goat
anti-rabbit IgG (100 lL) conjugated to horseradish peroxi-
dase were diluted 1 : 5000 in NaC ⁄ P
i
containing 0.1%
skimmed milk, added to wells and incubated at 37 °C for
1 h. For the activity assay, 100 lL of peroxidase substrate,
3,3¢,5,5¢-tetramethylbenzidine (Bangalore Genei, Bangalore,
India) (diluted 1 : 20 in water), was added to the wells and
incubated at 37 °C for 30 min. The reaction was stopped
by adding 25 lLof1m H
2

Æcm
)1
for WhiB1, 13 140 m
)1
Æcm
)1
for
WhiB2, 18 830 m
)1
Æcm
)1
for WhiB3 and WhiB4,
16 980 m
)1
Æcm
)1
for WhiB5, 27 200 m
)1
Æcm
)1
for WhiB6 and
17 550 m
)1
Æcm
)1
for WhiB7. Throughout, the study protein
concentrations were calculated using A
280
reading.
Acknowledgements

terium tuberculosis. Proc Natl Acad Sci USA 102 ,
12200–12205.
7 Raghunand TR & Bishai WR (2006) Mapping essential
domains of Mycobacterium smegmatis WhmD: insights
into WhiB structure and function. J Bacteriol 188,
6966–6976.
8 Geiman DE, Raghunand TR, Agarwal N & Bishai WR
(2006) Differential gene expression in response to expo-
sure to antimycobacterial agents and other stress condi-
tions among seven Mycobacterium tuberculosis whiB-like
genes. Antimicrob Agents Chemother 50, 2836–2841.
9 Singh A, Guidry L, Narasimhulu KV, Mai D, Tromb-
ley J, Redding KE, Giles GI, Lancaster JR Jr & Steyn
AJ (2007) Mycobacterium tuberculosis WhiB3 responds
Md. S. Alam et al. Molecular properties of M. tuberculosis WhiB proteins
FEBS Journal 276 (2009) 76–93 ª 2008 The Authors Journal compilation ª 2008 FEBS 91
to O
2
and nitric oxide via its [4Fe–4S] cluster and is
essential for nutrient starvation survival. Proc Natl
Acad Sci USA 104, 11562–11567.
10 Cole ST, Brosch R, Parkhill J, Garnier T, Churcher C,
Harris D, Gordon SV, Eiglmeier K, Gas B, Barry CE
3rd et al. (1998) Deciphering the biology of Mycobac-
terium tuberculosis from the complete genome sequence.
Nature 393, 537–544.
11 Camus JC, Pryor MJ, Medigue C & Cole ST (2002)
Re-annotation of the genome sequence of Mycobacte-
rium tuberculosis H37Rv. Microbiology 148, 2967–
2973.

teins (Spiro TG, ed.), pp. 67–96. Wiley-Interscience,
New York, NY.
21 Paraskevopoulou C, Fairhurst SA, Lowe DJ, Brick P &
Onesti S (2006) The Elongator subunit Elp3 contains a
Fe4S4 cluster and binds S-adenosylmethionine. Mol
Microbiol 59, 795–806.
22 Bertini I, Cowan JA, Luchinat C, Natarajan K &
Piccioli M (1997) Characterization of a partially
unfolded high potential iron protein. Biochemistry 36,
9332–9339.
23 Loferer H, Wunderlich M, Hennecke H & Glockshuber
R (1995) A bacterial thioredoxin-like protein that is
exposed to the periplasm has redox properties compara-
ble with those of cytoplasmic thioredoxins. J Biol Chem
270, 26178–26183.
24 Holmgren A (1979) Thioredoxin catalyzes the reduction
of insulin disulfides by dithiothreitol and dihydrolipoa-
mide. J Biol Chem 254, 9627–9632.
25 Shi Y, Tang W, Hao SF & Wang CC (2005) Contribu-
tions of cysteine residues in Zn
2
to zinc fingers and
thiol-disulfide oxidoreductase activities of chaperone
DnaJ. Biochemistry 44, 1683–1689.
26 Holmgren A (1985) Thioredoxin. Annu Rev Biochem 54,
237–271.
27 LeVine H III (1993) Thioflavine T interaction with syn-
thetic Alzheimer’s disease beta-amyloid peptides: detec-
tion of amyloid aggregation in solution. Protein Sci 2,
404–410.

of Mycobacterium tuberculosis persistence by gene and
protein expression profiling. Mol Microbiol 43, 717–731.
37 Ramakrishnan L, Federspiel NA & Falkow S (2000)
Granuloma-specific expression of Mycobacterium viru-
lence proteins from the glycine-rich PE-PGRS family.
Science 288, 1436–1439.
38 Fernando MR, Nanri H, Yoshitake S, Nagata-Kuno K
& Minakami S (1992) Thioredoxin regenerates proteins
inactivated by oxidative stress in endothelial cells. Eur J
Biochem 209, 917–922.
Molecular properties of M. tuberculosis WhiB proteins Md. S. Alam et al.
92 FEBS Journal 276 (2009) 76–93 ª 2008 The Authors Journal compilation ª 2008 FEBS
39 Nakamura H, Nakamura K & Yodoi J (1997) Redox
regulation of cellular activation. Annu Rev Immunol 15,
351–369.
40 Nishinaka Y, Masutani H, Nakamura H & Yodoi J
(2001) Regulatory roles of thioredoxin in oxidative
stress-induced cellular responses. Redox Rep 6, 289–295.
41 Holmgren A (1989) Thioredoxin and glutaredoxin sys-
tems. J Biol Chem 264, 13963–13966.
42 Arner ES & Holmgren A (2000) Physiological functions
of thioredoxin and thioredoxin reductase. Eur J Bio-
chem 267, 6102–6109.
43 Newton GL, Arnold K, Price MS, Sherrill C, Delcard-
ayre SB, Aharonowitz Y, Cohen G, Davies J, Fahey
RC & Davis C (1996) Distribution of thiols in microor-
ganisms: mycothiol is a major thiol in most actinomyce-
tes. J Bacteriol 178, 1990–1995.
44 Patel MP & Blanchard JS (1999) Expression, purifica-
tion, and characterization of Mycobacterium tuber-

53 Balmer Y, Koller A, del Val G, Manieri W, Schurmann
P & Buchanan BB (2003) Proteomics gives insight into
the regulatory function of chloroplast thioredoxins.
Proc Natl Acad Sci USA 100, 370–375.
54 Hayashi T, Ueno Y & Okamoto T (1993) Oxidoreduc-
tive regulation of nuclear factor kappa B. Involvement
of a cellular reducing catalyst thioredoxin. J Biol Chem
268, 11380–11388.
55 Schenk H, Klein M, Erdbrugger W, Droge W & Schu-
lze-Osthoff K (1994) Distinct effects of thioredoxin and
antioxidants on the activation of transcription factors
NF-kappa B and AP-1. Proc Natl Acad Sci USA 91,
1672–1676.
56 Kiley PJ & Beinert H (2003) The role of Fe–S proteins
in sensing and regulation in bacteria. Curr Opin Micro-
biol 6, 181–185.
57 Raghava GP, Solanki RJ, Soni V & Agrawal P (2000)
Fingerprinting method for phylogenetic classification
and identification of microorganisms based on variation
in 16S rRNA gene sequences. Biotechniques 29, 108–
112.
58 Sambrook J & Russell DW (2001) Molecular Cloning:
A Laboratory Manual, 3rd edn. Cold Spring Harbor
Laboratory Press, Cold Spring Harbor, NY.
59 Plank DW, Kennedy MC, Beinert H & Howard JB
(1989) Cysteine labeling studies of beef heart aconitase
containing a 4Fe, a cubane 3Fe, or a linear 3Fe cluster.
J Biol Chem 264, 20385–20393.
60 Carroll KS, Gao H, Chen H, Leary JA & Bertozzi CR
(2005) Investigation of the iron–sulfur cluster in Myco-