Structural basis of hypoxic gene regulation by the Rv0081 transcription factor of Mycobacterium tuberculosis

Abstract

The transcription factor Rv0081 of Mycobacterium tuberculosis controls
hypoxic gene expression and acts as a regulatory hub in the latent phase of
tuberculosis (TB) infection. We report here the crystal structure of Rv0081 at
2.9 �A resolution revealing that it belongs to the well-known ArsR/SmtB family
proteins. However, unlike other members in this family, Rv0081 has neither
a metal-binding domain nor does it possess Cys residues, suggesting an
alternate mechanism of gene regulation. Our structural and biochemical analyses
suggest the molecular basis for the recognition of self-regulatory DNA
sequences and a plausible mechanism of regulation of Rv0081 in the latent
phase of TB infection.

Full text

Structural basis of hypoxic gene regulation by the Rv0081
transcription factor of Mycobacterium tuberculosis
Ashwani Kumar
1
, Swastik Phulera
1,†
, Arshad Rizvi
2
, Parshuram J. Sonawane
1
,
Hemendra S. Panwar
1
, Sharmistha Banerjee
2
, Arvind Sahu
1
and Shekhar C. Mande
1,‡
1 National Centre for Cell Science, SP Pune University Campus, Pune, India
2 Department of Biochemistry, University of Hyderabad, Hyderabad, India
Correspondence
S. C. Mande, National Centre for Cell
Science, NCCS Complex, SP Pune
University Campus, Ganeshkhind, Pune 411
007, India
Tel: +91 20 2570 8235
E-mail: shekhar@nccs.res.in
†
Present address
Vollum Institute, Oregon Health and Science
University, Portland, OR, USA
‡
Present address
Council of Scientific and Industrial Research
(CSIR), New Delhi, India
(Received 6 January 2019, revised 26
March 2019, accepted 28 March 2019,
available online 26 April 2019)
doi:10.1002/1873-3468.13375
Edited by Stuart Ferguson
The transcription factor Rv0081 of Mycobacterium tuberculosis controls
hypoxic gene expression and acts as a regulatory hub in the latent phase of
tuberculosis (TB) infection. We report here the crystal structure of Rv0081 at
2.9

A resolution revealing that it belongs to the well-known ArsR/SmtB fam-
ily proteins. However, unlike other members in this family, Rv0081 has nei-
ther a metal-binding domain nor does it possess Cys residues, suggesting an
alternate mechanism of gene regulation. Our structural and biochemical anal-
yses suggest the molecular basis for the recognition of self-regulatory DNA
sequences and a plausible mechanism of regulation of Rv0081 in the latent
phase of TB infection.
Database
Structural data are available in the Protein Data Bank under the accession
number –6JMI
Keywords: ArsR/SmtB family; FHL regulator; hypoxia protein;
phosphoserine; Rv0081
A significant aspect of the pathogenesis of Mycobac-
terium tuberculosis (Mtb), the tuberculosis (TB) causing
bacterium, is the latent or dormant stage, where the
non-replicating bacteria persist in an infected person
asymptomatically and successfully evade host immunity
leading to survival inside the hostile macrophage envi-
ronment [1,2]. Apart from slow replication, these
mycobacteria are reported to develop resistance to
anti-mycobacterial agents. While almost all the avail-
able drugs act on the actively dividing bacteria, the
prolonged treatment is followed to eliminate the non-
replicating subpopulations that continue to thrive and
lead to reduced susceptibility to standard TB drugs.
Reactivation of the dormant bacteria contributes to an
active infection and consequent spread in the human
population, making the control and eradication of TB
highly challenging. Understanding the processes of
Mtb latency and reactivation are therefore crucial for
controlling the spread of TB [3]. The latency and reac-
tivation of Mtb has been experimentally demonstrated
to be intimately related to oxygen tension and based on
several experimental observations, laboratory models
that mimic clinical Mtb latency have been established
[4–7]. It has been observed in cell culture-based models
that progressively increasing hypoxic condition, where
mycobacteria are grown as sealed, standing cultures
Abbreviations
Mtb, Mycobacterium tuberculosis; SPR, surface-plasmon resonance; MSA, multiple sequence alignment; TB, tuberculosis; FHL, formate
hydrogen lyase; PTMs, post-translational modifications.
982 FEBS Letters 593 (2019) 982–995 ª2019 Federation of European Biochemical Societies

(the Wayne model) induced similar physiological, tran-
scriptional and proteome changes as exhibited by dor-
mant mycobacteria [7,8]. Subsequently, re-aeration, in
which mycobacteria grown under hypoxic conditions
are transferred to aerated, shaking cultures, could
mimic reactivation in vitro [6,9,10]. Using such models,
the hypoxia sensing mechanism of Mtb has been delin-
eated through genetic manipulations and comparative
transcriptomics [11–14] Hypoxia activated DosRS/
DosT (DevRS-Rv2027c) and MprAB two-component
signal transduction regulatory systems were found to
regulate nearly 48 genes implicated in dormancy and
granuloma-like conditions [15–17]. Later, the DosR-
dependent regulon was observed to be induced in
response to nitric oxide, carbon monoxide, low pH in
macrophages and in both early and late mouse infec-
tions, establishing the significance of this regulatory
operon in dormancy [18–21]. Besides DosR-mediated
initial hypoxic response, sustained hypoxia spanning to
4–7 days showed induction of the set of 230 genes,
referred to as Enduring Hypoxic Response (EHR), that
facilitated the shift to a persistent, metabolically inac-
tive, but viable state [6,22]. A systematic high through-
put study under hypoxia and re-aeration conditions
identified several transcriptional regulators outside that
of the DosR or EHR response [23]. This study identi-
fied Rv0081, a part of the DosR regulon, as a major
regulatory hub for multiple hypoxia induced pathways.
This supported our earlier simulation studies where we
had predicted DosR and DosS, along with Rv0081 as
important regulators of latency [24]. Thus, Rv0081 is
likely to be a major factor mediating transcriptional
changes during the transition between normoxic and
hypoxic conditions.
Rv0081, the first gene of the rv0081-rv0088 operon,
belongs to the ArsR/SmtB family of prokaryotic met-
alloregulatory transcriptional repressors [25]. This fam-
ily of repressors is known to regulate the expression of
genes linked to heavy metal ion stress including those
involved in metal uptake, efflux or detoxification [26].
Rv0081 was established as a repressor where Rv0081
self regulates its expression by binding to an inverted
repeat element upstream to the locus rv0081-rv0088
[25]. Additionally, Rv0081 was shown to co-regulate
the genes of rv0081-rv0088 operon, predicted to encode
formate hydrogenlyase (FHL) complex. FHL is
involved in the degradation of formic acid, which is
formed fermentatively under anaerobic conditions, into
carbon dioxide and hydrogen [27,28]. As mycobacteria
would require shifting the metabolism to anaerobic
conditions inside granuloma during latency, Rv0081
might act as one of the key regulators deciding the fate
of mycobacterial latency. By virtue of controlling the
expression of genes on the rv0081-rv0088 operon,
Rv0081 might be required to sense oxygen levels in the
cells, and thereby trigger the gene expression under
low oxygen levels. Detailed characterization of Rv0081
would therefore be essential to understand its role in
sensing varying levels of oxygen and switching genes
between dormancy and actively dividing forms.
ArsR/SmtB family of regulators are known to
respond to toxic levels of metals by virtue of the con-
served metal-binding motif - ELCV(C/G)D, termed as
the ‘metal-binding box’ in a-3 and a-5 helices [26,29].
It was previously reported that Rv0081 does not have
the conserved metal-binding residues [25], and we
observed the same in this study. This suggests that
DNA-binding properties of Rv0081 might be metal
independent. In Mtb, metal-independent DNA binding
is regulated by either thiol switch or post-transcrip-
tional modifications, such as phosphorylation [30–33],
acetylation [34,35] or methylation [36,37] and we can
postulate one or combination of such mechanisms reg-
ulates the activity of Rv0081. With the emerging sig-
nificance of Rv0081 in dormancy and our own
observation of its indirect dependence on GroEL [38],
we have determined the crystal structure of Rv0081
and identified the possible molecular basis for binding
of Rv0081 to its cognate DNA element. A plausible
mechanism of regulation by Rv0081 is proposed,
which might be due to post-translational modifications
(PTMs) of Rv0081 during control of gene expression.
Materials and methods
Purification, crystallization and structure
determination of Rv0081
The (His)
6
-tagged Rv0081 protein was purified to homogene-
ity (Data S1) and protein concentration of 1.5–2.0 mgmL
1
was used for crystallization trials. Diffractable crystals were
obtained in a condition with 0.1 MCH
3
COONa, 0.2–0.4 M
(NH
4
)
2
SO
4
and 25% w/v PEG 4000 in 2 lL drops set at
room temperature. The crystals were stabilized in crystalliza-
tion condition with 20% glycerol before flash-freezing in liq-
uid-nitrogen for data collection at 100K.
To determine the structure of Rv0081 using molecular
replacement (MR), six different models from Protein Data
Bank (4GGG, 2KJB, 3PQK, 1R1U, 2KKO and 3GW2)
were used (described in detail in Supporting Information).
A dimer of a chainsaw or poly-Ala model of these of PDB
structures individually was used as search models for MR
performed by Phaser [39]. Initial refinement was done in
CCP4i in REFMAC5 [40] module and visualized in COOT
[41]. Further final manual model building and refinement
was done in COOT and PHENIX [42,43] refinement mod-
ules respectively.
983FEBS Letters 593 (2019) 982–995 ª2019 Federation of European Biochemical Societies
A. Kumar et al. Structure of Hypoxia protein Rv0081

Site-directed mutagenesis for Rv0081 mutants
DNA-binding residues of the Rv0081, namely S48, S49,
S52 and Q53, identified by the superimposition of Rv0081
structure with NolR-DNA complex (PDB: 4OMY) were
selected for mutagenesis. Following two mutants were gen-
erated where these residues were either changed to alanine
or aspartic acid using Site-directed mutagenesis (Data S1);
(a) A mutant of Rv0081 with S48A, S49A S52A and Q53A
named as Rv0081 S/Q?A henceforth; (b) A mutant of
Rv0081 with S48D, S49D and S52D named as Rv0081 S?
D henceforth. The Rv0081 S/Q?A and Rv0081 S?D
mutants were purified in the same way as that of wild-type
Rv0081 as described before.
Electrophoretic mobility shift assay
DNA probes
Biotin-labelled and unlabelled oligonucleotides used for
performing Electrophoretic mobility shift assay (EMSA)
(detailed in sequence Table S2) were synthesized and
desalted or PAGE purified (IDT, Coralvillie, IA, USA).
EMSA with Rv0081 protein and Wt-DNA probes as well
as mutant DNA probes (M1, M2 and M3) were performed
using LightShift
TM
Chemiluminescent EMSA Kit (Thermo
Fisher Scientific, Waltham, MA, USA) as per the manufac-
turer’s recommendations (mentioned in details in Data S1).
Further to test the effect of specific mutations of Rv0081
on DNA binding, EMSA were performed with Rv0081 S/
Q?A or Rv0081 S?D with ds-Wt-DNA. Additionally,
competitive EMSA were also performed for Rv0081 or
Rv0081 S/Q?A with labelled ds-Wt-DNA and unlabelled
ds-Wt-DNA or mutated oligos. All the above experiments
were performed in duplicates.
Surface-plasmon resonance
DNA and protein binding studies
To validate the interaction of Rv0081 proteins (Rv0081-
wild-type or Rv0081 S/Q?A or Rv0081 S?D) with ds
DNA oligos (biotin-labelled Wt, M1 and M2), surface-plas-
mon resonance analyses were carried out using the Biacore
2000 system (Biacore AB, Uppsala, Sweden; as described in
the in Data S1).
Kinetics studies
Kinetic measurements were performed by flowing various
concentrations of Rv0081 and Rv0081 S/Q?A proteins
over the SA-chip immobilized with the Wt and M2-biotiny-
lated oligos. The data collected were fitted with suitable
Table 1. Crystallography data collection and refinement statistics.
Rv0081 Rv0081
PDB ID 6JMI
Beamline XRD1, Elettra
Synchrotron
Detector Pilatus 2M
Wavelength (

A) 0.9686
Resolution range (

A) 45.57–2.896 (3–2.896)
a
Space group P 4
1
2
1
2 (92)
Unit cell a=63.317

A, b=63.317

A,
c=262.604

Aa=b=c=90°
Total reflections 168 683 (15 717)
Unique reflections 12 678 (1198)
Multiplicity 13.3 (13.1)
Completeness (%) 99.54 (97.15)
Mean I/sigma (I) 12.37 (1.04)
Wilson B-factor (

A)
2
78.52
R
merge
0.1841 (2.443)
R
meas
0.1915 (2.539)
R
pim
0.05187 (0.6803)
CC
1/2
0.999 (0.509)
CC*1 (0.822)
Matthew’s coefficient for four
monomers in asymmetric unit
2.89
Refinement statistics
Reflections used in refinement 12 678 (1191)
Reflections used for R-free 1264 (8124)
R
work
0.2610 (0.3515)
R
free
0.3013 (0.3988)
CC (work) 0.958 (0.639)
CC (free) 0.950 (0.598)
Number of non-hydrogen atoms 2788
Macromolecules 2768
Ligands (SO2
4ion) 4
Protein residues 375
rmsd bond length (

A) 0.002
rmsd bond angles (°) 0.48
Ramachandran favored (%) 96.62
Ramachandran allowed (%) 3.10
Ramachandran outliers (%) 0.28
Rotamer outliers (%) 2.97
Clash-score (%) 3.90
Average B-factor (

A
2
) 75.17
Macromolecules (%) 74.85
Reflections used in refinement 118.99
R
merge
=∑
hkl
∑
i
∣I
i
(hkl)

I(hkl) ∣/∑(hkl) ∑
i

I(hkl) with irunning over
the number of independent observations of reflection hkl, where I
i
,

Iare intensity and average intensity of reflection hkl respectively.
R
work
=∑
hkl
‖F
o
(hkl) ∣∣F
c
(hkl) ‖/∑
hkl
∣F
o
(hkl).
R
free
=∑
hklεT
‖F
o
(hkl)∣F
c
(hkl)‖/∑
hklεT
F
o
(hkl), where Tis a test
data set randomly selected from the observed reflections prior to
refinement. Test data set was not used throughout refinement and
10% of the total unique reflections. F
o
(hkl) and F
c
(hkl) are the
observed and calculated structure factor amplitudes.
CC*=[2 CC
1/2
/(1 +CC
1/2
)]
½
, where CC
1/2
Pearson correlation coef-
ficient.
a
Statistics for the highest resolution shell (3.0–2.90)

A of X-ray
diffraction data are shown in parentheses.
984 FEBS Letters 593 (2019) 982–995 ª2019 Federation of European Biochemical Societies
Structure of Hypoxia protein Rv0081 A. Kumar et al.

kinetic models, that is either with Bi-valent model for Wt
oligo sequences or 1 : 1 Langmuir model for the mutant
oligo sequences using the BIAEVALUATION software version
3.2 (Biacore).
Results
Overall structural features of Rv0081
The structure determined by MR showed four chains
in the asymmetric unit. The best MR solution was
obtained in the space group P4
1
2
1
2 with a poly-Ala
model of 1R1U. Moreover, this solution did not yield
any short contacts between symmetry-related mole-
cules thereby suggesting the solution to be correct. The
structure was refined to reasonable geometrical param-
eters (Table 1). One cell length being very long could
be explained by the presence of two dimers in the crys-
tal asymmetric unit arranged in a side-to-side manner.
Earlier it has been reported that the molecule is likely
to be a monomer; however, recently it has been
reported that Rv0081 exist in a dimeric form [44] and
also forms higher oligomers at higher concentration
[45]. Therefore, we tested the presence of dimeric asso-
ciation through multi-angle light scattering coupled
with size-exclusion chromatography (SEC-MALS).
These results confirmed that in vitro purified protein is
indeed dimeric in tested buffer conditions (Fig. S1A–
C) and supports earlier finding [44].
Furthermore, to confirm the SEC-MALS results of
the biological assembly, we calculated the area buried
among the four different Chains. The Chains A and
D; and B and C possess the interface area of 1499.5
and 1465.9

A
2
respectively. On the other hand, the
interface area between chains AD and chain BC is
609

A
2
. Chain A and Chain D thereby appear to form
one biological dimer whereas Chain B and Chain C
form the other biological dimer. The four chains in the
asymmetric unit together therefore constitute two bio-
logical dimers (Fig. 1).
In each of the chains, a number of residues at the
N- and C-termini were not visible in the electron den-
sity, and hence were excluded from the final model.
Moreover, the region of the structure that is likely to
be involved in DNA binding is highly disordered in all
the four chains. The final model therefore constitutes
of residues Ser 3 to Ala 102 in Chain A; Glu 4 to Ser
41, Asp 43 to Leu 46, Ser 49 to Ala 102 in Chain B;
Ser 3 to Val 44, Ser 49 to Ala 102 in Chain C; and
Glu 4 to Leu 39, Asp 43 to Leu 46, Asn 50 to Ala 65
and Tyr 75 to Val 101 in Chain D. Presumably
because of such large segments of structural disorder,
the average temperature factor is high, leading to poor
diffraction quality of the crystals.
The overall structure is archetypical winged helix-
turn-helix motif DNA binding proteins and is similar to
other proteins in the ArsR/SmtB family (Table S2).
Structural superposition with apo-CzrA (PDB: 1R1U)
and apo-SmtB (PDB: 1R1T) yielded overall a rmsd of
1.8 and 1.9

A (Fig. S2A,B) respectively. Maximum
deviation is seen in the region where DNA-binding site
is located. The disordered DNA binding region of the
Fig. 1. Structure of the Rv0081 dimer. The structure of Rv0081 is typical of the ArsR/SmtB family of metal-dependent transcriptional
repressors. (A) Crystal asymmetric unit contains four identical chains as shown in four different colours. The extremities of the polypeptide
chains at either of the N- and C- termini make up the interface between two dimers. At this stage it is not clear if the tetrameric assembly
is biologically relevant or not. (B) Two monomers possess an extensive monomer-monomer interface, clearly suggesting that the biological
assembly is that of a dimer. The interface is principally made up of the 1st and the last helices. Only in one chain, that is, in Chain A, the
entire molecule could be fitted in the electron density. In all the other chains, there are disordered regions which could not be fitted in
electron density (see text for details).
985FEBS Letters 593 (2019) 982–995 ª2019 Federation of European Biochemical Societies
A. Kumar et al. Structure of Hypoxia protein Rv0081

structure is suggestive of induced fit binding to the
DNA [46].
Rv0081 lacks metal-binding motif but has
conserved DNA binding residues
The members of the ArsR/SmtB family transcriptional
repressors are known to bind DNA in a metal-regu-
lated manner. There are three major classes of metal-
binding regulators in this family, those which bind
metal defined by ‘a3 motif’ (e.g. ArsR), ‘a5 motif’ (e.g.
CzrA, NmtR, and CmtR) and ‘a3 and a5’together (e.g.
SmtB, CadC and ZiaR). We compared the structural
regions of Rv0081 with the corresponding region of
metal bound CzrA-Zn
2
(PDB: 1R1V) with overall rmsd
of 1.5

A and SmtB-Zn
2
(PDB: 1R22) with overall rmsd
of 1.8

A. Surprisingly, we found that the residues
presumably involved in binding metals are mutated in
Rv0081 (Fig. 2A). For example, in the a5 family, in
the structure, CzrA-Zn
2
, the residues involved are Asp
84, His 86, His 97 and His 100 whereas in a3-a5 family
structure, SmtB-Zn
2
, the residues are Asp 104, His 106,
His 117 and Glu 120. The superposed residues in
Rv0081 are Ala 79, Asp 81, Val 92 and Arg 95 respec-
tively, with both a5 and a3-a5 family. The superposi-
tion of Rv0081 with a3 family proteins was not
performed owing to unavailability of the crystal struc-
ture. Multiple sequence alignment (MSA) of Rv0081
with ArsR (a3), CzrA (a5) and SmtB (a3-a5) proteins
also showed that Rv0081 did not have metal-binding
residues (Fig. 2B). Further, we examined the Rv0081
structure for potential metal-binding sites in the
PAR3D program [47] and found that there are no
potential metal-binding sites in Rv0081. It therefore
Fig. 2. Superposition of Rv0081 structure with the SmtB-Zn
2
structure bound to metals (PDB: 1R22) and MSA of representative of a3, a5
and a3-a5 regions of ArsR/SmtB family with Rv0081. (A) In the a3-a5 family, SmtB-Zn
2
complex (purple ribbon structure), Zn
2+
ion only
seen at a5 region, not at a3 region. Residues of SmtB, which are coordinated with Zn
2+
are Asp104, His 106, His’117 and Glu’120 at a5
region, where the latter two residues are contributed by another monomer in the dimer. Structural superposition shows that these residues
in the corresponding region of the Rv0081 structure (green ribbon structure) are Ala 79, Asp 81, Val’ 92 and Arg’ 95. Therefore, the metal
coordinating ability of Rv0081 appears to have been lost during evolution. Similar comparisons with the regions of a3 and a5 also reinforce
the same conclusions. (B) MSA of ArsR (a3), CzrA (a5) and SmtB (a3-a5) protein sequence with Rv0081 sequence, the region highlighted
with green and red represents a3 and a5 regions respectively. Superpositions and MSA of a3, a5 and a3-a5 regions of ArsR/SmtB family
with Rv0081 showed that Rv0081 might have lost its ability to bind metals.
986 FEBS Letters 593 (2019) 982–995 ª2019 Federation of European Biochemical Societies
Structure of Hypoxia protein Rv0081 A. Kumar et al.

appears that the activity of Rv0081 binding to DNA
might not be regulated by metals.
By comparing the Rv0081 structure with other
ArsR/SmtB transcription factors, we could delineate
the regions likely to play important roles in DNA
recognition (Fig. 3A). From the structural superposi-
tion of NolR-DNA complex of Sinorhizobium fredii
(PDB: 4OMY), there are two kinds of residues that
are seen to interact with the DNA, the residues
involved in contacting the DNA bases and the residues
interacting with phosphate backbone. In the structure
of NolR-DNA complex, the base recognizing residues
are Gln 56, Ser 57, Ser 60 and Gln 61 [48]. Superposi-
tion of the two structures suggests that corresponding
residues in Rv0081 would be Ser 48, Ser 49, Ser 52
and Gln 53. Similarly, the phosphate binding residues
Fig. 3. Superposition of Rv0081 structure with the NolR-DNA structure bound to DNA. (A) The superposition of Rv0081 with NolR-DNA
complex (PDB: 4OMY) clearly indicates two different kinds of binding modes- (a) residues which are involved in base-recognition (red dotted
lines) and (b) those involved in contacting the phosphate backbone (black dotted lines). DNA binding domains of NolR are highlighted in box.
NolR residues interacting with bases are highlighted in purple and those with phosphate in light brown colour. Bases of DNA are highlighted
in white and gray and phosphate backbone in cyan colour (Box i and iii). The superposed residues of Rv0081 that might potentially interact
with bases are highlighted in yellow and those with phosphate in red colour (Box ii and iv). Thus, the superposition suggests the sequence
of the DNA involved in DNA-Rv0081 interactions. (B) Surface diagram of dimer of (chain A, green and chain D, cyan) Rv0081 Structure
shows that the base-contacting residues (yellow) lie at the extremities of the structure, whereas the phosphate contacting residues (red) fill
in the intervening surface of the structure. (C) Clustering of basic residues in Rv0081, that is, Lys 15, His 19 and Arg 22, is suggestive of
anion binding site. Fo-Fc map contoured at 5rlevel shows a strong density in this region. We have modelled SO2
4ion in this density (from
the crystallization reagent). See text for details.
987FEBS Letters 593 (2019) 982–995 ª2019 Federation of European Biochemical Societies
A. Kumar et al. Structure of Hypoxia protein Rv0081

in Rv0081 would be Lys 15, His 19, Arg 22, Gln 54,
Asn 71 and Tyr 75. In the surface diagram representa-
tion of dimer of Rv0081 in Fig. 3B, base recognizing
residues denoted as yellow lie at the extremity of the
dimer and phosphate binding residues in red colour
span the entire surface. Thus, residues involved in
DNA recognition could be predicted from compar-
isons with other structures that are available in com-
plex with cognate DNA.
In the Rv0081 structure, Lys15, His19 and Arg22
form a cluster of basic residues, suggesting their possi-
bility of anion binding. Indeed, Fo-Fc maps show a
strong peak at 5rabove mean electron density level in
three of the four chains (Fig. 3C). Although our
resolution is not very high (2.9

A), and thereby it does
not permit us the correct identity of ligand in this
strong density, we have modelled sulfate ion in this
density. Moreover, such a clustering of basic residues
is reminiscent of anion binding sites such as sulfates
[49,50] and phosphates [51,52], or acetate or formate
[53,54].
DNA binding by Rv0081
Interaction studies on Rv0081 protein with
autoregulatory elements
DNA binding of Rv0081 was confirmed for a 33 bp
oligonucleotide with its sequence derived from an
upstream sequence (67 to 35) of the Rv0081- Rv0088
operon. This region encompasses the R1 and R2
regions, where the protein is believed to bind first to the
R2 region, and then cooperatively to the R1 region [25].
From the structure, we identified five bases which are
likely to be contacted by the protein. These analyses
suggested that Rv0081 might bind to the sequence
TATCT in the R1 region and TCTTC in the R2 region
(Fig. 4A). We mutated these five bases in one or both of
the R1 and R2 regions into GGGGG and tested bind-
ing by the protein (Fig. 4A). EMSA analyses with the
wild-type (Wt) or mutated oligos showed that the pro-
tein binds to the native sequence (Fig. 4B). This binding
is sequence-specific since bound biotinylated Wt-oligos
competed with non-biotinylated Wt-oligos but did not
Fig. 4. EMSA and SPR of the wild-type Rv0081 with different oligo sequences. EMSA and SPR were carried out as described in the text with
a 33 base pair (from 67 to 35 relative to the Rv0081 translational start site) native region (Wt) and mutated oligos M1, M2 and M3. (A) Two
DNA motifs interacting with Rv0081 are labelled as R1 and R2. Five potential bases both in R1 (TATCT) and R2 (TCTTC), interacting with
Rv0081 in (Wt) oligo were mutated to GGGGG in either R1 (M2) or R2 (M1) as well as both (M3). (B) EMSA was performed with the Rv0081
protein (1 lM) and biotinylated DNA probes (25 nM) Wt, M1, M2 and M3. Shifted bands are indicated by ‘*’ and the bar plot below shows the
intensity of shifted bands. (C) SPR analysis of the Rv0081 protein (100 nM) with immobilized DNA probes of either Wt (red), M1 (green) or M2
(blue) mutants performed using Biacore 2000. Average response units for the Rv0081 protein with ds-Wt as well as M1 and M2 oligos
are shown using a bar graph inset (Data are mean SD of three experiments). Weak binding to M2 (R1 mutant) and no binding to M1
(R2 mutant) oligo sequences suggest that the Rv0081 protein might initially bind the R2 region, and then bind cooperatively to the R1 region.
988 FEBS Letters 593 (2019) 982–995 ª2019 Federation of European Biochemical Societies
Structure of Hypoxia protein Rv0081 A. Kumar et al.

compete with scrambled-oligos RS1, RS2 and RS3
(Fig. S3A). EMSA results clearly show that Rv0081
binding efficiency diminishes in the R1-mutated oligo,
whereas it is almost completely abolished in the
R2-mutated oligo (Fig. 4B). Corroboratively, surface-
plasmon resonance (SPR) with purified Rv0081 and
immobilized oligos [Wt, M1 (R2-mutated region) and
M2 (R1-mutated region)] shows reduced binding for the
R1-mutated region and almost completely abolished
binding for R2-mutated region (Fig. 4C). Thus, these
results indicate that the Rv0081 protein binds to the R2
region more efficiently than the R1 region, confirming
the earlier findings [25].
Mutational analysis of Rv0081 suggest modulation of
DNA binding
Based on the superimposition of Rv0081 structure with
NolR-DNA complex, we identified four key residues
of Rv0081 for interaction with DNA bases. These resi-
dues, S48, S49, S52 and Q53, were selected for the
mutational analysis. We generated two mutants of
Rv0081 proteins one where all these four residues were
mutated to Ala (Rv0081 S/Q?A), and the other where
Ser were mutated to Asp (Rv0081 S?D; Fig. 5A) and
EMSA as well as SPR experiments were performed to
compare the interaction of Rv0081 and its mutants
(Rv0081 S/Q?A, Rv0081 S?D). Binding of Rv0081
and Rv0081 S/Q?A mutant were in concentration-
dependent manner with ds-Wt DNA (EMSA). The
binding was modestly reduced in Rv0081 S/Q?A
mutant whereas it was almost completely abolished in
Rv0081 S?D mutant (Fig. 5B). The similar pattern of
interactions was observed with both the mutant pro-
teins using the SPR. Rv0081 S/Q?A mutant showed a
moderate binding response, whereas Rv0081 S?D dis-
played substantially reduced binding response
(Table 2). It is important to point out here that
Rv0081 S/Q?A failed to show equivalent binding
response to that of Rv0081 even at a concentration
that is 32-times (3200 nM) higher than that of Rv0081
(Figs 5C, 6C and S4A) whereas Rv0081 S?D showed
barely detectable binding response 6400 nMconcentra-
tion (Figs 5C, S4B).
Fig. 5. EMSA and SPR of the wild-type and mutants (S/Q?A, S?D) of Rv0081 with wild-type (Wt) oligonucleotide sequences. (A)
Schematic representation of the amino acid residues of wild-type Rv0081 protein (Rv0081) and its mutants Rv0081 S/Q?A and Rv0081 S?
D. (B) EMSA was performed with increasing concentration (0.5, 1 and 5 lM) of either Rv0081, Rv0081 S/Q?A or Rv0081 S?D proteins
with Wt biotinylated DNA probe and shifted bands are indicated by ‘*’. (C) Binding interaction of the Rv0081 or its mutant proteins with the
Wt oligonucleotide sequence. The biotin-labelled Wt oligo (test flow cell) or the scrambled oligo (control flow cell) was immobilized on a
streptavidin chip. The specific binding response (shown as sensogram overlay plots) of Rv0081 (100 nM), Rv0081 S/Q?A mutant (3200 nM)
and Rv0081 S?D mutant (6400 nM) to the Wt oligonucleotide is shown in red, blue and green, respectively. The inset bar graph shows the
average binding of all the three proteins to ds-Wt oligos. (Data are mean SD of three experiments).
Table 2. Summary of interactions of Rv0081, its Ala and Asp
mutants with Wt-type oligo R1, R 2 and both R1 and R2 mutants.
Auto regulating element
and its mutants Rv0081 and its Ala and Asp mutants
Oligos Rv0081 Rv0081 S/Q?A Rv0081 S?D
Wt +++ ++ 
M1 (R2 mutant)  
M2 (R1 mutant) +
M3 (R1 +R2 mutant)  
989FEBS Letters 593 (2019) 982–995 ª2019 Federation of European Biochemical Societies
A. Kumar et al. Structure of Hypoxia protein Rv0081

Kinetics of Rv0081-DNA binding
To calculate affinities of the Rv0081 and its mutants to
the wild-type (Wt) or mutant oligos, different concen-
tration of proteins were flowed over the oligo-immobi-
lized chip. The SPR results clearly showed that the
native oligonucleotide sequence (ds-Wt DNA) binds
with greater affinity to the native Rv0081 protein.
In order to obtain the affinity for this interaction,
the data were fitted with a Bivalent model (BIA
evaluation) (v
2
=2.74). The analyses indicated
K
D1
=1.04 910
7
M and K
D2
=5.2 910
2
M
(Fig. 6A, Table 3), suggesting that one of the binding
sites has higher affinity. This is consistent with our
EMSA and SPR analysis data (Fig. 4B,C), confirming
that the R2 region has the higher affinity for Rv0081.
The R2 mutated oligonucleotide (M1 oligo) failed
almost completely to bind (Fig. 4C), on the other
hand, the R1 mutated oligonucleotide (M2 oligo)
bound to Rv0081 with 100-fold lower affinity (K
D
1.56 910
5
M; Fig. 6B, Table 3). Effect of mutation
on low affinity R1 region also suggests the possibility of
cooperative binding of both R1 and R2 regions with
Rv0081.
Discussion
In this study, we report the structure of Rv0081 at
2.9

A resolution (Fig. 1). The dimeric assembly of
Rv0081 is canonical of transcriptional repressors of
ArsR/SmtB family. In the crystal structure, the DNA-
binding regions are ordered only in one of the four
chains. However, they are juxtaposed symmetrically in
the biological dimer (Fig. 1B) suggesting that the
mode of DNA binding is similar to that of the most
bacterial transcriptional repressors of this family [26].
The DNA binding by Rv0081 can be imagined as
being spread through one surface of the molecule
(Fig. 3B), with the base recognizing residues (Ser 48,
Ser 49, Ser 52 and Gln 53) lie at the extremities of the
dimer, whereas the entire length of the intervening sur-
face being dotted with basic amino acids which are
likely to interact with the phosphate backbone of the
DNA. The phosphate-recognition residues from
the two monomers therefore cover the entire surface of
the dimer between the two base-recognition sites.
These residues potentially involved in DNA binding
appear to be conserved in the homologs of Rv0081
from different genera of bacteria and different species
of mycobacteria (Fig. S5).
It has been suggested that Rv0081 fails to bind
metals [25]. In the other structures of the ArsR/SmtB
family of proteins, one or two metal-binding sites are
known, which are designated as either a3, a5 or both
a3-a5[26,29]. Superposition of the Rv0081 structure
with other family members of the ArsR/SmtB
revealed that in both these putative metal-binding
regions the residues in Rv0081 are either non-polar,
or mutated as to be devoid of metal-binding. Further,
MSAs and phylogenetic analysis of different classes
(a3, a5 and a3-a5) of ArsR/SmtB family with Rv0081
amino acid sequence also supported our observation
(Figs S6, S7 and S8). We therefore agree with the
reported study that Rv0081 is unlikely to bind cations
[25]. As other members of this family-sense metal-
dependent redox changes, the mechanism of Rv0081
to sense the hypoxic conditions clearly appears to be
Fig. 6. Binding analysis of Rv0081 wild-type or the mutant protein
(Rv0081 S/Q?A) to the immobilized oligos (Wt/M2 oligos). The
sensogram overlay plots of Rv0081 (wild-type protein) flowed at
the indicated concentrations over the Wt (A) or the mutant oligo
M2 (B) is shown. Similarly the sensogram overlay plots of the
mutant protein (Rv0081 S/Q?A) is shown in (C). The solid lines
represent the fitting of the data either by bivalent model (A, C) or
with the 1 : 1 Langmuir binding model. The details of kinetics are
described in Table 3.
990 FEBS Letters 593 (2019) 982–995 ª2019 Federation of European Biochemical Societies
Structure of Hypoxia protein Rv0081 A. Kumar et al.

distinctly different and metal-independent. Wide rang-
ing oxygen sensing mechanisms in prokaryotic cells
have been described earlier, which include those medi-
ated by metals [55], iron-sulfur clusters [56]orby
means of Cysteine-disulfide exchange [57,58]. In the
absence of reversible metal-binding, oxygen sensing is
mediated by thiol redox switch as revealed by crystal
structures of many members of ArsR/SmtB family
[59–64]. Rv0081 does not possess any such canonical
site for sensing oxygen levels. However, it interacted
with haeme non-covalently (Fig. S9).
In the absence of cation binding and thiol redox
switch, the mode of transcriptional regulation of
Rv0081 in response to hypoxic conditions therefore
remains enigmatic. Based on previous studies we can
hypothesize that under hypoxic conditions, the Rv0081
might get post-translationally modified to alter its
DNA-binding ability by phosphorylation [30–33],
acetylation [34,35], methylation [36,37] or modulation
by formylation [65–67]. First, to check the possibility of
phosphorylation of Rv0081, three Serines which are
postulated to be involved in base-recognition become
excellent candidates to test this hypothesis (Fig. 5). We
have shown that two mutants (Rv0081 S/Q?A Rv0081
S?D) of Rv0081 show significant differences in DNA
binding ability (Fig. 5), where the former binds DNA
with lower affinity, the latter almost failed to bind to
DNA (Figs 5C and 6C, Table 3). The Serine to Aspar-
tate change being a mimic of a phospho-Serine, this
result is suggestive of Ser-phosphorylation in Mtb
Rv0081. Our analysis with Mtb cultures grown under
normoxic and hypoxic conditions suggest the possibility
of differential phosphorylation of the Rv0081 protein,
with elevated levels of phosphorylation under the
hypoxic conditions (Fig. S10). However, such a novel
mechanism needs to be further validated with various
laboratory models of hypoxia. Secondly, it might be
possible that Rv0081 also senses formate levels, instead
of oxygen levels, and thereby influences the expression
of the rv0081-rv0088 operon. This possible mechanism
further needs to be investigated if it is similar to the
mechanism of fhlA transcription factor of Escherichia
coli which regulates the FHL genes by sensing the for-
mate levels [68,69]. Additionally, as Rv0081 is the first
gene and repressor of the FHL operon it might be regu-
lated by formate ions. Lastly, the correct modelling of
the strong density for anions in Fo-Fc maps of Rv0081
(Fig. 3C) with correct potential ligand(s) would be
important for the understanding of the regulation of
Rv0081 by various PTMs.
Finally, the structure of Rv0081 and accompanying
solutions studies reveal that different levels of DNA-
binding regulation by Rv0081 might exist in Mtb.
Galagan et al. [23] have reported Rv0081 to be a regu-
latory hub of many genes in Mtb, and it is likely that
such multiple mechanisms of DNA binding offer
redundancy to switch-off and -on different genes via
different mechanisms. The crystal structure of Rv0081,
therefore, reveals mechanisms of DNA binding which
are distinct from other members of the ArsR/SmtB
family members.
Acknowledgements
The funding was provided by the Department of
Biotechnology, Ministry of Science and Technology
grants BT/PR3260/BRB/10/967/2011 and BT/PR1545
0/COE/34/46/2016. We thank Dr. Thomas C. Zahrt,
Medical College of Wisconsin, USA for providing
Rv0081 plasmid. We thank Dr. Saikrishanan (IISER
Pune) and beamline staff at the Elettra synchrotron
Trieste (XRD1) for their assistance during data collec-
tion. We thank Ms. Tanuja N Bankar, Dr. Krish-
nasastry and Dr. Ramanmurthy for generation of
poly-Ab of Rv0081 in NCCS animal house facility,
Dr. C.M Santosh Kumar for his suggestion in Rv0081
purification and Ms. Kriti Chopra for help in prepar-
ing phylogenetic trees. The authors acknowledge finan-
cial assistance from ICMR (AK), CSIR (SP, AR and
HSP) and SERB-NPDF (PJS).
Table 3. Kinetic and affinity data for interactions of Rv0081 with its auto-regulating element. k
a
, association rate constant; k
d
, dissociation
rate constant; K
D
, equilibrium dissociation constant. Data shown here are representative of three independent experiments.
Oligo Protein k
d1
(1/s)/k
a1
(1/Ms) K
D1
(M)
k
d2
(1/s)/k
a2
(1/Ms) K
D2
(M) Rmax Chi
2
Fitting model
Wt Rv0081 3.41 910
2
/3.28 910
5
1.04 910
7
26.101/4.97 5.25 910
2
161 2.74 Bivalent
Wt Rv0081
S/Q?A mutant
5.54 910
2
/2.82 910
3
1.96 910
5
0.261/4.38 2.17 910
4
101 0.549 Bivalent
k
d
(1/s)/k
a
(1/Ms) K
D
(M)
M2
(R1 mutant)
Rv0081 9.92 910
2
/6.34 910
4
1.56 910
6
–784 13.5 Langmuir (1 : 1)
991FEBS Letters 593 (2019) 982–995 ª2019 Federation of European Biochemical Societies
A. Kumar et al. Structure of Hypoxia protein Rv0081

Author contributions
AK, SCM designed the study; AK, SP, SCM solved
the structure; AK, AR, SB performed the Mtb H37Rv
experiments; AK, PJS performed EMSA; HSP and
AKS contributed in SPR experiments; AK, SB, PJS,
SCM wrote the manuscript. SCM obtained the fund-
ing and supervised the project. All the authors have
read the manuscript.
References
1 Cadena AM, Fortune SM and Flynn JL (2017)
Heterogeneity in tuberculosis. Nat Rev Immunol 17,
691–702.
2 Abel L, Fellay J, Haas DW, Schurr E, Srikrishna G,
Urbanowski M, Chaturvedi N, Srinivasan S, Johnson DH
and Bishai WR (2017) Genetics of human susceptibility to
active and latent tuberculosis: present knowledge and
future perspectives. Lancet Infect Dis 18, e64–e75.
3 Arora A, Chandra NR, Das A, Gopal B, Mande SC,
Prakash B, Ramachandran R, Sankaranarayanan R,
Sekar K, Suguna K et al. (2011) Structural biology of
Mycobacterium tuberculosis proteins: the Indian efforts.
Tuberculosis 91, 456–468.
4 Dutta NK and Karakousis PC (2014) Latent
tuberculosis infection: myths, models, and molecular
mechanisms. Microbiol Mol Biol Rev 78, 343–371.
5 Kupz A, Zedler U, St€
aber M and Kaufmann SHE
(2016) A mouse model of latent tuberculosis infection
to study intervention strategies to prevent reactivation.
PLoS ONE 11, e0158849.
6 Rustad TR, Harrell MI, Liao R and Sherman DR
(2008) The enduring hypoxic response of
Mycobacterium tuberculosis.PLoS ONE 3, e1502.
7 Wayne LG and Hayes LG (1996) An in vitro model for
sequential study of shiftdown of Mycobacterium
tuberculosis through two stages of nonreplicating
persistence. Infect Immun 64, 2062–2069.
8 Alnimr AM and Alnimr AM (2015) Dormancy models
for Mycobacterium tuberculosis: a minireview. Braz J
Microbiol 46, 641–647.
9 Gengenbacher M and Kaufmann SHE (2012)
Mycobacterium tuberculosis: success through dormancy.
FEMS Microbiol Rev 36, 514–532.
10 Kondratieva T, Azhikina T, Nikonenko B, Kaprelyants
A and Apt A (2014) Latent tuberculosis infection: what
we know about its genetic control? Tuberc Edinb Scotl
94, 462–468.
11 Gerasimova A, Kazakov AE, Arkin AP, Dubchak I
and Gelfand MS (2011) Comparative Genomics of the
Dormancy Regulons in Mycobacteria. J Bacteriol 193,
3446–3452.
12 Gopinath V, Raghunandanan S, Gomez RL, Jose L,
Surendran A, Ramachandran R, Pushparajan AR,
Mundayoor S, Jaleel A and Kumar RA (2015) Profiling
the Proteome of Mycobacterium tuberculosis during
dormancy and reactivation. Mol Cell Proteomics 14,
2160–2176.
13 Rosenkrands I, Slayden RA, Crawford J, Aagaard C,
Barry CE and Andersen P (2002) Hypoxic response of
Mycobacterium tuberculosis studied by metabolic
labeling and proteome analysis of cellular and
extracellular proteins. J Bacteriol 184, 3485–3491.
14 Starck J, K€
allenius G, Marklund B-I, Andersson DI and

Akerlund T (2004) Comparative proteome analysis of
Mycobacterium tuberculosis grown under aerobic and
anaerobic conditions. Microbiology 150, 3821–3829.
15 Bagchi G, Chauhan S, Sharma D and Tyagi JS (2005)
Transcription and autoregulation of the Rv3134c-devR-
devS operon of Mycobacterium tuberculosis.Microbiol
Read Engl 151, 4045–4053.
16 Chauhan S, Sharma D, Singh A, Surolia A and Tyagi
JS (2011) Comprehensive insights into Mycobacterium
tuberculosis DevR (DosR) regulon activation switch.
Nucleic Acids Res 39, 7400–7414.
17 Park H-D, Guinn KM, Harrell MI, Liao R, Voskuil
MI, Tompa M, Schoolnik GK and Sherman DR (2003)
Rv3133c/dosR is a transcription factor that mediates
the hypoxic response of Mycobacterium tuberculosis.
Mol Microbiol 48, 833–843.
18 Voskuil MI, Schnappinger D, Visconti KC, Harrell MI,
Dolganov GM, Sherman DR and Schoolnik GK (2003)
Inhibition of respiration by nitric oxide induces a
Mycobacterium tuberculosis dormancy program. J Exp
Med 198, 705–713.
19 Rustad TR, Sherrid AM, Minch KJ and Sherman DR
(2009) Hypoxia: a window into Mycobacterium
tuberculosis latency. Cell Microbiol 11, 1151–1159.
20 Kumar A, Deshane JS, Crossman DK, Bolisetty S, Yan
B-S, Kramnik I, Agarwal A and Steyn AJC (2008)
Heme oxygenase-1-derived carbon monoxide induces
the Mycobacterium tuberculosis dormancy regulon.
J Biol Chem 283, 18032–18039.
21 Rohde K, Yates RM, Purdy GE and Russell DG
(2007) Mycobacterium tuberculosis and the environment
within the phagosome. Immunol Rev 219,37–54.
22 Gupta S and Chatterji D (2005) Stress responses in
mycobacteria. IUBMB Life 57, 149–159.
23 Galagan JE, Minch K, Peterson M, Lyubetskaya A,
Azizi E, Sweet L, Gomes A, Rustad T, Dolganov G,
Glotova I et al. (2013) The Mycobacterium tuberculosis
regulatory network and hypoxia. Nature 499, 178–183.
24 Hegde SR, Rajasingh H, Das C, Mande SS and Mande
SC (2012) Understanding communication signals during
mycobacterial latency through predicted genome-wide
protein interactions and boolean modeling. PLoS ONE
7, e33893.
25 He H, Bretl DJ, Penoske RM, Anderson DM and
Zahrt TC (2011) Components of the Rv0081-Rv0088
992 FEBS Letters 593 (2019) 982–995 ª2019 Federation of European Biochemical Societies
Structure of Hypoxia protein Rv0081 A. Kumar et al.

locus, which encodes a predicted formate hydrogenlyase
complex, are coregulated by Rv0081, MprA, and
DosR in Mycobacterium tuberculosis.J Bacteriol 193,
5105–5118.
26 Busenlehner LS, Pennella MA and Giedroc DP (2003)
The SmtB/ArsR family of metalloregulatory
transcriptional repressors: structural insights into
prokaryotic metal resistance. FEMS Microbiol Rev 27,
131–143.
27 McDowall JS, Murphy BJ, Haumann M, Palmer T,
Armstrong FA and Sargent F (2014) Bacterial formate
hydrogenlyase complex. Proc Natl Acad Sci USA 111,
E3948–E3956.
28 Pinske C and Sawers RG (2016) Anaerobic formate
and hydrogen metabolism. EcoSal Plus 7,1–32.
29 Eicken C, Pennella MA, Chen X, Koshlap KM,
VanZile ML, Sacchettini JC and Giedroc DP (2003) A
metal-ligand-mediated intersubunit allosteric switch in
related SmtB/ArsR zinc sensor proteins. J Mol Biol
333, 683–695.
30 Singh A, Gupta R, Vishwakarma RA, Narayanan PR,
Paramasivan CN, Ramanathan VD and Tyagi AK
(2005) Requirement of the mymA operon for
appropriate cell wall ultrastructure and persistence of
Mycobacterium tuberculosis in the Spleens of Guinea
Pigs. J Bacteriol 187, 4173–4186.
31 Kumar P, Kumar D, Parikh A, Rananaware D, Gupta
M, Singh Y and Nandicoori VK (2009) The
Mycobacterium tuberculosis protein kinase K modulates
activation of transcription from the promoter of
mycobacterial monooxygenase operon through
phosphorylation of the transcriptional regulator VirS.
J Biol Chem 284, 11090–11099.
32 Kalantari A, Derouiche A, Shi L and Mijakovic I
(2015) Serine/threonine/tyrosine phosphorylation
regulates DNA binding of bacterial transcriptional
regulators. Microbiol Read Engl 161, 1720–1729.
33 Cohen-Gonsaud M, Barthe P, Canova MJ, Stagier-
Simon C, Kremer L, Roumestand C and Molle
V (2009) The Mycobacterium tuberculosis Ser/Thr
Kinase substrate Rv2175c is a DNA-binding protein
regulated by phosphorylation. J Biol Chem 284,
19290–19300.
34 Ghosh S, Padmanabhan B, Anand C and Nagaraja V
(2016) Lysine acetylation of the Mycobacterium
tuberculosis HU protein modulates its DNA binding
and genome organization. Mol Microbiol 100,
577–588.
35 Bi J, Gou Z, Zhou F, Chen Y, Gan J, Liu J, Wang H
and Zhang X (2018) Acetylation of lysine 182 inhibits
the ability of Mycobacterium tuberculosis DosR to bind
DNA and regulate gene expression during hypoxia.
Emerg Microbes Infect 7, 108.
36 Shell SS, Prestwich EG, Baek S-H, Shah RR, Sassetti
CM, Dedon PC and Fortune SM (2013) DNA
Methylation impacts gene expression and ensures
hypoxic survival of Mycobacterium tuberculosis.PLoS
Pathog 9, e1003419.
37 Phelan J, de Sessions PF, Tientcheu L, Perdigao J,
Machado D, Hasan R, Hasan Z, Bergval IL, Anthony
R, McNerney R et al. (2018) Methylation in
Mycobacterium tuberculosis is lineage specific with
associated mutations present globally. Sci Rep 8, 160.
38 Sharma A, Rustad T, Mahajan G, Kumar A, Rao
KVS, Banerjee S, Sherman DR and Mande SC (2016)
Towards understanding the biological function of the
unusual chaperonin Cpn60.1 (GroEL1) of
Mycobacterium tuberculosis.Tuberc Edinb Scotl 97,
137–146.
39 McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn
MD, Storoni LC and Read RJ (2007) Phaser
crystallographic software. J Appl Crystallogr 40, 658–
674.
40 Murshudov GN, Skub
ak P, Lebedev AA, Pannu NS,
Steiner RA, Nicholls RA, Winn MD, Long F and
Vagin AA (2011) REFMAC5 for the refinement of
macromolecular crystal structures. Acta Crystallogr D
Biol Crystallogr 67, 355–367.
41 Emsley P and Cowtan K (2004) Coot: model-building
tools for molecular graphics. Acta Crystallogr D Biol
Crystallogr 60, 2126–2132.
42 Adams PD, Grosse-Kunstleve RW, Hung L-W, Ioerger
TR, McCoy AJ, Moriarty NW, Read RJ, Sacchettini
JC, Sauter NK and Terwilliger TC (2002) PHENIX:
building new software for automated crystallographic
structure determination. Acta Crystallogr D Biol
Crystallogr 58, 1948–1954.
43 Echols N, Grosse-Kunstleve RW, Afonine PV,
Bunkoczi G, Chen VB, Headd JJ, McCoy AJ, Moriarty
NW, Read RJ, Richardson DC et al. (2012) Graphical
tools for macromolecular crystallography in PHENIX.
J Appl Crystallogr 45, 581–586.
44 Dong S, Ding Z, Wang Y, Yang Y, Mao Y and Wang
Y (2017) Transcription factor Rv0081 from
Mycobacterium tuberculosis: purification, crystallization
and initial crystallographic analysis. Acta Crystallogr F
Struct Biol Commun 73, 281–285.
45 Sun X, Zhang L, Jiang J, Ng M, Cui Z, Mai J, Ahn
SK, Liu J, Zhang J, Liu J et al. (2018) Transcription
factors Rv0081 and Rv3334 connect the early and the
enduring hypoxic response of Mycobacterium
tuberculosis.Virulence 9, 1468–1482.
46 Longo A, Guanga GP and Rose RB (2007) Structural
basis for induced fit mechanisms in DNA recognition
by the Pdx1 homeodomain. Biochemistry 46,
2948–2957.
47 Goyal K and Mande SC (2008) Exploiting 3D
structural templates for detection of metal-binding sites
in protein structures. Proteins Struct Funct Bioinforma
70, 1206–1218.
993FEBS Letters 593 (2019) 982–995 ª2019 Federation of European Biochemical Societies
A. Kumar et al. Structure of Hypoxia protein Rv0081

48 Lee SG, Krishnan HB and Jez JM (2014) Structural
basis for regulation of rhizobial nodulation and
symbiosis gene expression by the regulatory protein
NolR. Proc Natl Acad Sci USA 111, 6509–6514.
49 Pereira CT, Roesler C, Faria JN, Fessel MR and Balan
A (2017) Sulfate-Binding Protein (Sbp) from
Xanthomonas citri: structure and functional insights.
Mol Plant-Microbe Interact 30, 578–588.
50 Copley RR and Barton GJ (1994) A structural analysis
of phosphate and sulphate binding sites in proteins.
Estimation of propensities for binding and conservation
of phosphate binding sites. J Mol Biol 242, 321–329.
51 Hirsch AKH, Fischer FR and Diederich F (2007)
Phosphate recognition in structural biology. Angew
Chem Int Ed Engl 46, 338–352.
52 Schenk G, Gahan LR, Carrington LE, Mitic N,
Valizadeh M, Hamilton SE, de Jersey J and Guddat
LW (2005) Phosphate forms an unusual tripodal
complex with the Fe-Mn center of sweet potato purple
acid phosphatase. Proc Natl Acad Sci USA 102,
273–278.
53 James TL and Cohn M (1974) The role of the lysyl
residue at the active site of creatine kinase. Nuclear
Overhauser effect studies. J Biol Chem 249,
2599–2604.
54 K€
uhnel K, Blankenfeldt W, Terner J and Schlichting I
(2006) Crystal structures of chloroperoxidase with its
bound substrates and complexed with formate, acetate,
and nitrate. J Biol Chem 281, 23990–23998.
55 Waldron KJ, Rutherford JC, Ford D and Robinson NJ
(2009) Metalloproteins and metal sensing. Nature 460,
823–830.
56 Crack JC, Green J, Thomson AJ and Le Brun NE
(2014) Iron-sulfur clusters as biological sensors: the
chemistry of reactions with molecular oxygen and nitric
oxide. Acc Chem Res 47, 3196–3205.
57 Green J and Paget MS (2004) Bacterial redox sensors.
Nat Rev Microbiol 2, 954–966.
58 Paulsen CE and Carroll KS (2013) Cysteine-mediated
redox signaling: chemistry, biology, and tools for
discovery. Chem Rev 113, 4633–4679.
59 Cook WJ, Kar SR, Taylor KB and Hall LM (1998)
Crystal structure of the cyanobacterial metallothionein
repressor SmtB: a model for metalloregulatory proteins.
J Mol Biol 275, 337–346.
60 Guimar~
aes BG, Barbosa RL, Soprano AS, Campos
BM, de Souza TA, Tonoli CCC, Leme AFP,
Murakami MT and Benedetti CE (2011) Plant
pathogenic bacteria utilize biofilm growth-associated
repressor (BigR), a novel winged-helix redox switch, to
control hydrogen sulfide detoxification under hypoxia.
J Biol Chem 286, 26148–26157.
61 Liu T, Nakashima S, Hirose K, Shibasaka M,
Katsuhara M, Ezaki B, Giedroc DP and Kasamo K
(2004) A novel cyanobacterial SmtB/ArsR family
repressor regulates the expression of a CPx-ATPase and
a metallothionein in response to both Cu(I)/Ag(I) and
Zn(II)/Cd(II). J Biol Chem 279, 17810–17818.
62 Moinier D, Slyemi D, Byrne D, Lignon S, Lebrun R,
Talla E and Bonnefoy V (2014) An ArsR/SmtB family
member is involved in the regulation by arsenic of the
arsenite oxidase operon in Thiomonas arsenitoxydans.
Appl Environ Microbiol 80, 6413–6426.
63 Osman D and Cavet JS (2010) Bacterial metal-sensing
proteins exemplified by ArsR-SmtB family repressors.
Nat Prod Rep 27, 668–680.
64 Ye J, Kandegedara A, Martin P and Rosen BP (2005)
Crystal structure of the Staphylococcus aureus pI258
CadC Cd(II)/Pb(II)/Zn(II)-responsive repressor.
J Bacteriol 187, 4214–4221.
65 Lee S (2013) Post-translational modification of proteins
in toxicological research: focus on lysine acylation.
Toxicol Res 29,81–86.
66 Kim J-M, Seok O-H, Ju S, Heo J-E, Yeom J, Kim D-
S, Yoo J-Y, Varshavsky A, Lee C and Hwang C-S
(2018) Formyl-methionine as an N-degron of a
eukaryotic N-end rule pathway. Science 362, eaat0174
67 Wi
sniewski JR, Zougman A and Mann M (2008) Ne-
Formylation of lysine is a widespread post-translational
modification of nuclear proteins occurring at residues
involved in regulation of chromatin function. Nucleic
Acids Res 36, 570–577.
68 Maupin JA and Shanmugam KT (1990) Genetic
regulation of formate hydrogenlyase of Escherichia coli:
role of the fhlA gene product as a transcriptional
activator for a new regulatory gene, fhlB. J Bacteriol
172, 4798–4806.
69 Schlensog V and B€
ock A (1990) Identification and
sequence analysis of the gene encoding the
transcriptional activator of the formate hydrogenlyase
system of Escherichia coli.Mol Microbiol 4, 1319–1327.
Supporting information
Additional supporting information may be found
online in the Supporting Information section at the end
of the article.
Data S1. Supplementary methods.
Table S1. Oligonucleotides used for EMSA and SPR
experiments.
Table S2. List of selected proteins of different metal-
binding motif of ArsR/SmtB family.
Fig. S1. Purification and characterization Rv0081,
Rv0081 S/Q?A and Rv0081 S?D A. Equal amount of
purified proteins were loaded on 15% SDS/PAGE. B.
Size-exclusion chromatography profile (SEC) of
Rv0081, Rv0081 S/Q?A and Rv0081 S?D loaded on
Superdex-200 10/300-GL with flow rate of
0.3 mLmin
1
. C. SEC-MALS analysis of Rv0081.
994 FEBS Letters 593 (2019) 982–995 ª2019 Federation of European Biochemical Societies
Structure of Hypoxia protein Rv0081 A. Kumar et al.

Purified Rv0081 protein was separated on Superdex-200
10/300-GL size-exclusion chromatography column cou-
pled with multi-angle light scattering (SEC-MALS;
Wyatt Tech.). Analyses of SEC-MALS results suggest
the molecule to be dimeric in solution. D. Circular
dichroism of the all three protein Rv0081 (wild-type,
red), Rv0081 S/Q?A (Blue) and Rv0081 S?D (green)
shows that the helical contents of all three proteins are
similar.
Fig. S2. A: Superposition of structure of Rv0081 (green)
with apo-CzrA (red) (a5 family) and B: Superposition of
structure of Rv0081 (green) with apo-SmtB (pink) 3a-5a
family.
Fig. S3. Competition EMSA: EMSA were performed for
Rv0081 (panel A) or Rv0081 S/Q?A (panel B) proteins
with Wt-labelled oligo and competing them with differ-
ent unlabelled scrambled oligos RS1, RS2 and RS3.
Rv0081 (1lM) protein (Panel A) or Rv0081 S/Q?A
(1lM) protein (Panel B) were incubated with biotin
labelled ds-Wt oligo (25 nM) in presence of 50-folds
excess of different scramble oligos. After incubation, the
complexes were resolved on 6.5% non-denaturing-
PAGE. Signal was detected using the Lightshift EMSA
assay kit as per manufacturer’s instructions and imaged
using Imager600 (GE).
Fig. S4. SPR analysis of binding of Rv0081 S/Q?A and
Rv0081 S?D mutant to the Wt oligo.
Fig. S5. A: MSA - Rv0081 homologous proteins. B: Phy-
logenetic tree - Rv0081 homologous proteins.
Fig. S6. MSA and phylogenetic tree of a3 motif members
of ArsR/SmtB family members. A: MSA of a3 motif, B:
Phylogenetic tree a3 motif.
Fig. S7. MSA and phylogenetic tree of a5 motif members
of ArsR/SmtB family members.
Fig. S8. MSA and phylogenetic tree of a3-a5 motif mem-
bers of ArsR/SmtB family members.
Fig. S9.UV–vis absorption spectra of Hemin (3 lM)
bound to 3 lMBSA (A) or Rv0081 (B). (a) absorption
spectra of BSA or Rv0081, respectively; (b) absorption
spectrum of Hemin; (c) absorption spectra of [(BSA or
Rv0081) +Hemin] and (d) absorption spectra of [(BSA
or Rv0081) +Hemin]-[Hemin]. Hemin showed A
kmax
at
390 nM, however, protein bound Hemin showed a red
shift. For [BSA+Hemin] A
kmax
was at 394–396 nMand
for [Rv0081 +Hemin] A
kmax
was at 396–400 nM.(C)
UV–vis absorption spectra of hemin (0–10 lM) bound to
3lMRv0081 (left panel). As the concentration of Hemin
increased the spetra shifted toleft (blue shift). Increasing
concentration of Hemin alone did not show the blue
shift (right panel). The blue shift in the left panel could
be a result of increasing concentration of free Hemin (for
example see [12]). Hemin is ferric chloride heme.Data
shown here are representative of one of three indepen-
dent experiments. A
kmax
represent the absorbence max-
ima at particular wavelength.
Fig. S10. Phosphorylation of Rv0081 under Hypoxia.
995FEBS Letters 593 (2019) 982–995 ª2019 Federation of European Biochemical Societies
A. Kumar et al. Structure of Hypoxia protein Rv0081