Biochemical and Biophysical Research Communications 560 (2021) 27e31

Contents lists available at ScienceDirect

Biochemical and Biophysical Research Communications

journal homepage: www.elsevier.com/locate/ybbrc

HU-AB simulacrum: Fusion of HU-B and HU-A into HU-B-A, a

functional analog of the Escherichia coli HU-AB heterodimer
Kanika Arora, Bhishem Thakur, Archit Gupta, Purnananda Guptasarma*
Centre for Protein Science, Design and Engineering (CPSDE), Department of Biological Sciences, Indian Institute of Science Education and Research (IISER)
Mohali, Knowledge City, Sector-81, SAS Nagar, Punjab, 140306, India

a r t i c l e i n f o a b s t r a c t

Article history: In enteric bacteria such as Escherichia coli, there are two homologs of the DNA-binding nucleoid asso-
Received 16 April 2021 ciated protein (NAP) known as HU. The two homologs are known as HU-A and HU-B, and exist either in
Accepted 26 April 2021 the form of homodimers (HU-AA, or HU-BB) or as heterodimers (HU-AB), with different propensities to
Available online 5 May 2021
form higher-order oligomers. The three different dimeric forms dominate different stages of bacterial
growth, with the HU-AB heterodimer dominating cultures in the stationary phase. Due to similarities in
Keywords:
their properties, and the facile equilibrium that exists between the dimeric forms, the dimers are difficult
Nucleoid associated proteins
to purify away from each other. Although HU-AA and HU-BB can be purified through extensive ion-
Histone-like protein HU
Heterodimer formation
exchange chromatography, reestablishment of equilibrium interferes with the purification of the HU-
Construct of a heterodimer simulacrum AB heterodimer (which constitutes ~90% of any population with equal numbers of HU-B and HU-A
DNA-binding chains). Here, we report the creation of a functional analog of HU-AB that does not appear to partition
Protein Engineering to generate any minority populations of HU-AA or HU-BB. The analog was constructed through genetic
fusion of the HU-B and HU-A chains into a single polypeptide (HUeB-A) with a glycine/serine-rich linker
of 11 amino acids separating HU-B from HU-A, and a histidine tag at the N-terminus of HU-B. HU-B-A
folds to bind 4-way junction DNA, and displays a significant tendency to form dimers (i.e., analogs of HU
tetramers), and a higher thermodynamic stability than HU-BB or HU-AA, thus explaining why it domi-
nates mixtures of HU-B and HU-A chains.
© 2021 Elsevier Inc. All rights reserved.

1. Introduction contamination by the wild-type (untagged) forms of HU-A and HU-

The protein HU plays a role in the compaction of bacterial nu-
cleoids [1]. Enteric bacteria such as E. coli have two homologs of HU,
known as HU-A and HU-B [2,3]. These homologs associate into HU-
A homodimers (HU-AA), HU-B homodimers (HU-BB) as well as
heterodimers of HU-A and HU-B (HU-AB). It is known that homo-
dimers of wild-type HU-AA and HU-BB interact spontaneously to
give rise to a ~90% population of HU-AB heterodimers, when equal
amounts of HU-AA and HU-BB homodimers are mixed together
[4e6]. Further, it is also known that such spontaneous trans-
formation of HU-AA and HU-BB homodimers into the HU-AB het-
erodimer appears not to occur if the chains constituting the HU-AA
and HU-BB homodimers possess N-terminal 6xHis histidine tags,
with this allowing purification of N-terminally histidine-tagged
HU-AA, or HU-BB, from overexpressing cells, without
* Corresponding author.
E-mail address: guptasarma@iisermohali.ac.in (P. Guptasarma).
B [7]. Our own (unpublished) observations suggest that the ex-
change of subunits between HU-AA and HU-BB tends to be
discouraged when HU chains possess histidine tags, or any other
protein (e.g., another protein domain in genetic fusion) at the N-
terminus, because the N-terminal extension frustrates HU hetero-
dimer formation by (a) strengthening inter-subunit contacts, and
(b) preventing formation of a higher-order oligomer that acts as a
subunit exchange-facilitating intermediate. Notably, a hetero-
tetrameric intermediate has been proposed to be involved in the
exchange of subunits between HU-AA and HU-BB [8].
In the context of this paper, the non-occurrence of subunit ex-
change between N-terminally histidine-tagged HU-AA and HU-BB
allows facile generation and purification of homodimers of HU-
AA, and HU-BB, respectively, from cells overexpressing HU-A and
HU-B chains. As these histidine-tagged HU chains do not appear to
participate in any subunit exchange with wild-type HU, to generate
any heterodimers of the histidine-tagged form on one HU homolog,
and the wild-type form of the other homolog, only pure histidine-
tagged homodimers are obtained from HU-A or HU-B

https://doi.org/10.1016/j.bbrc.2021.04.107
0006-291X/© 2021 Elsevier Inc. All rights reserved.
K. Arora, B. Thakur, A. Gupta et al.
overexpression [7]. However, for the exact same reason(s), such
chains cannot then be used to generate HU-AB heterodimers,
through mixing of homodimers of histidine-tagged HU-AA and HU-
BB. Therefore, ion-exchange chromatography followed by crystal-
lization of wild-type HU from stationary phase cultures appears to
be the only method available for generation and purification of HU-
AB heterodimers [9,10]. This is because it is difficult to purify wild-
type HU-AB heterodimers from any population of HU chains
through ion-exchange chromatography alone, as such a population
would reestablish the equilibrium that causes HU-AB to constitute
only ~90% of the HU population, with the remaining population
becoming reorganized to generate ~10% of HU-AA and HU-BB
homodimers. Thus, the three-dimensional structure of HU-AB has
been determined [10], because an enriched population of HU-AB
proved to suffice for its crystallization (which is an acknowledged
method for the purification of any protein). Not much else is known
about HU-AB.
In this paper, we describe a novel approach for creating and
studying pure HU-AB, by generating a simulacrum of the HU-AB
heterodimer, instead of the heterodimer itself. Our approach in-
volves genetic fusion of HU-B and HU-A chains into a single poly-
peptide chain, with a linker separating HU-B and HU-A, and an N-
terminal histidine tag to potentially prevent subunit exchange, as
shown in Fig. 1A. The rationale for the creation of this simulacrum is
that the likelihood of the HU-B sequence preferentially encoun-
tering the HU-A sequence, to form heterodimeric contacts with it,
may be considered to be greatly increased if the HU-B and HU-A
sequences happen to lie upon the very same polypeptide chain.
Thus, we created this simulacrum of HU-AB by using a glycine/
serine-rich linker of 11 residues, consisting of the sequence, N-
SGGGGSGGGGS-C, to separate the HU-B sequence from the HU-A
sequence. In an extended form, this linker is anticipated to be
~38.5 Å (or longer) in length, assuming a length of 3.5 Å per amino
acid residue. The linker could thus be anticipated to quite easily
span the distance of ~28.7 Å separating the C-terminus of the HU-B
polypeptide and the N-terminus of the HU-A polypeptide in the
structure of a crystallized HU-AB heterodimer. It is easy to see that
even if the trajectory of the linker in the simulacrum of the HU-AB
structure takes it around the circumference of the HU-AB hetero-
dimer, the linker would still be long enough to span the distance
from the C-terminus of the HU-B chain to the N-termimus of the
HU-A chain, as shown by the schematic diagram in Fig. 1A.
In the remainder of this paper, we show that the polypeptide
incorporating the fusion of the HU-B and HU-A sequences forms a
well-folded structure that binds to 4-way junction DNA, displays
Fig. 1. Panel A shows the structure of the difficult-to-maintain HU-AB heterodimer (PDB ID: 2O97). The electron density for the DNA binding loops at the top of the figure is missing
in the crystal structure. The two chains are schematically shown to be linked by a glycine-serine-rich linker peptide. Panel B. The synthesis of the gene encoding the 6xHis-HU-B-
linker-HU-A fusion through splicing by overlap extension PCR (SOE-PCR) is shown schematically, along with a schematic representing the sequence of fusions at the bottom of the
panel.
28
Biochemical and Biophysical Research Communications 560 (2021) 27e31
oligomerization, and also shows a thermodynamic stability that is
higher than HU-AA or HU-BB, explaining why it forms spontane-
ously when there is no kinetic barrier to its formation.
2. Materials and methods
Cloning, expression and purification of HU-BA (the HU-AB
simulacrum), and HU-AA and HU-BB. HU-AA and HU-BB were
purified from clones already in our possession, as described earlier
[11,12]. To obtain the fusion product, HU-B-A, HU-B was fused to N-
terminal of HU-A. An 11 residues-long glycine-serine linker
(SGGGGSGGGGS) was introduced between the two proteins to
allow flexibility of folding of chain segments. The schematic of the
construction is shown in Fig. 1B. At first, the linker was added to the
C-terminal of HU-B by a splicing PCR reaction (PCR1) using the
primers, HU-B Forward: 50 - AGCTACTGGATCCATGAA-
TAAATCTCAATT GATCG-30 and Linker Reverse: 50 -
GCTGCCACCTCCGCCTGAACCTCCTCCACCTGAGTT TACCGCGTCTTTC-
3’. Separately, the linker was also added to the N-terminus of HU-A
by a splicing PCR reactions (PCR2) using the primers, HU-A Reverse:
50 -GAATACTCCCGGGTTACTTA ACTGCGTCTTTCAATG-30 and Linker
Forward: 50 -AGCGGTGGAGGAGGATCAGGCGG AGGTGGCAGCAT-
GAAC AAGACTCAACTG-3'. The two PCR products were then joined
together through a third splicing PCR reaction (PCR3), using HU-B
Forward and HU-A Reverse primers. The spliced product was
then digested and ligated into a pQE-30 vector between the Bam HI
and Hind III restriction enzyme sites, and transformed into E. coli
XL1-Blue cells. The successful transformation was confirmed
through sequencing of the plasmid purified from these cells. The
protein was also expressed in these cells. Purification of the protein
was performed using standard Ni-NTA imidazole metal affinity
chromatography (IMAC) using Ni-NTA agarose resin (Qiagen), with
extensive washing with 1 M NaCl to remove bound DNA (and any
other proteins bound to such DNA) using a method identical to that
described by us earlier [11,12].
Analytical chromatography and electrophoresis, including
glutaraldehyde cross-linking. Analytical gel filtration chromatog-
raphy was performed using a 24 ml Superdex-75 (10/300) Increase
(GE) column run on an AKTA Purifier 10 workstation (GE). The
column was equilibrated with 20 mM, pH 7.4, PBS buffer made with
sodium phosphate, containing 150 mM NaCl. SDS-PAGE experi-
ments were carried out using the standard Laemmli procedures. For
checking the oligomeric status of the fusion protein, 0.5 mg/ml of
protein was incubated with 0.01, 0.05 and 1% glutaraldehyde, and
incubated for 5 min at room temperature, prior to loading of control
K. Arora, B. Thakur, A. Gupta et al.
and glutaraldehyde-treated protein on 15% SDS-PAGE gels.
Circular dichroism (CD) studies. CD spectra were collected on a
Chirascan Spectrophotometer (Applied Photophysics, UK) using a
quartz cuvette of 1 mm path length and a protein concentration of
~0.5 mg/ml in 20 mM phosphate buffered saline of pH 7.4, con-
taining 150 mM NaCl. Sample and buffer spectra were collected
between 200 and 250 nm, and spectra obtained were corrected for
background. CD monitoring of thermal denaturation was per-
formed by heating protein samples from 20  C to 90  C using the
spectropolarimeter's peltier block arrangement, with data being
collected at 2  C intervals. For chemical denaturation experiments,
protein (0.5 mg/ml) was pre-incubated over-night in solutions of
different urea concentrations, prior to collection of CD spectra.
Gibb's free energy calculations. Gibb's free energy calculations
were performed for the individual homodimers, HU-AA and HU-BB,
and for the fusion protein, HU-AB, by monitoring equilibrium
chemical denaturation in the presence of different urea concen-
trations, after overnight incubation of protein samples in urea. The
fraction of protein unfolded (fU), or folded (fF) were calculated as
follows: fU¼(yF-y)/(yF-yU); fF ¼ 1-fU. The equilibrium constant, Keq,
for all molar concentrations of urea was calculated using the
following equation: Keq ¼ fU/fF. The Gibbs free energy associated
with each unfolding transition at every molar concentration of urea
was calculated using the equation: DG ¼ -RTlnKeq, where the values
of R (gas constant) and K (Kelvin) were taken to be 8.314 J/K.mol
and 298 K. The value of DG obtained for unfolding transitions at
each urea concentration, for all three proteins, was plotted using
Origin Pro 5 software (Microcal) against the respective urea con-
centrations, followed by linear fitting using the equation y¼mx þ c.
The intercept c represents a range of DG values for the protein to
undergo a transition from partially-unfolded state (at 1.5. M urea
concentration) into a nearly-completely-unfolded state.
3. Results
HU-B-A (6xHis-HU-B-linker-HU-A fusion) is folded and adopts
helical structure. Fig. 2A shows the far-UV circular dichroism (CD)
spectrum of the purified HU-AB simulacrum, 6xHis-HU-B-linker-
HU-A, or HU-B-A. Measurement of raw ellipticity (rather than mean
residue ellipticity) was used to plot the spectrum, due to issues
with accurate measurement of protein concentration (since neither
HU-A nor HU-B chains possess tryptophan or tyrosine residues).
The far-UV-CD spectrum was found to be well-folded and entirely
similar to the spectra of HU-AA and HU-BB homodimers published
by our laboratory [11]. Thus, the fusion construct, consisting of HU-
B and HU-A chains fused into a single chain (HUeB-A) constituting
a simulacrum of HU-AB was found to have naturally folded to adopt
a structure apparently simulating the HU-AB heterodimer, through
intra-molecular association of HU-B and HU-A chain sequences.
The sequence of the HU-B-A construct is shown in Supplementary
Fig. 1.
HU-B-A is functional and binds to DNA. Fig. 2B establishes that
the HU-B-A (6xHis-HU-B-linker-HU-A) construct binds to Holliday
(4-way) junction DNA (4WJ DNA), in an assay similar to the stan-
dardized test used to examine such binding [11e13]. As the figure
shows, this simulacrum of the HU-AB heterodimer gives rise to a
significant shift in the mobility of 4WJ DNA, establishing that it
binds to DNA with a protein (HUeB-A) concentration-dependent
reduction of intensity of the 4WJ DNA band, with concomitant in-
crease in intensity of the DNA smear(s) resulting from binding of
HU-B-A to DNA, in a manner in which the protein associates and
dissociates with 4WJ DNA on the timescale of electrophoresis
(resulting in a smear instead of a sharp band), entirely akin to what
is seen with HU-BB, or HU-BA, or other proteins derived from HU
[11,12]. At HU-B-A concentrations between 20 and 45 mM, the
29
Biochemical and Biophysical Research Communications 560 (2021) 27e31
sample of 4WJ DNA (1 mM) is seen to be also visible in the wells of
the gel, due to the creation of a large population with excessively-
restricted mobility.
HU-B-A is a mixture of monomers (simulating HU-AB dimers)
and dimers (simulating HU-AB tetramers). Fig. 2C shows the
chromatographic profile of the folded HU-B-A (6xHis-HU-B-linker-
HU-A) protein construct subjected to gel filtration chromatography
on a ~24 ml Superdex-75 (GE) column. The elution profile, as well
as the elutions observed at ~10 ml, and ~12 ml, suggest that 6xHis-
HU-B-linker-HU-A construct consists of species with molecular
weights of ~21 kDa, corresponding to the molecular weight of the
HU-AB heterodimer (~20.7 kDa), i.e., a monomer of 6xHis-HU-B-
linker-HU-A, as well as another species of twice this molecular
weight (a tetramer of ~42 kDa, or a dimer of 6xHis-HU-B-linker-
HU-A). This suggests that the 6xHis-HU-B-linker-HU-A protein
construct has undergone both intramolecular association between
its HU-B and HU-A sequences (to form a simulacrum of the HU-AB
heterodimer), and also intermolecular association between two
such folded HU-B-A chains (to form a dimer simulating an HU-AB
tetramer). Notably, when the two elutions are collected and chro-
matographed again, the ~10 ml elution splits again into ~10 and
~12 ml elutions, whereas the ~12 ml elution does not do so, sug-
gesting that the dissociation of the dimer of HU-B-A into mono-
meric HU-B-A is faster than the association of HU-B-A into the
dimer of HU-B-A. The data supporting this is shown in
Supplementary Fig. 2. Glutaraldehyde cross-linking was performed
to confirm the 6xHis-HU-B-linker-HU-A fusion's tendency to
display higher-order oligomers. In Fig. 2D, lane 1 of the SDS-PAGE
corresponds to the control 6xHis-HU-B-HU-A polypeptide, corre-
sponding to a ~22 kDa band for HU-B-A. In lane 2, upon cross-
linking with 0.1% glutaraldehyde, the ~22 kDa band shows a
higher mobility corresponding to a molecular weight of ~20 kDa. It
is known that such an increase in mobility arises from the extensive
intramolecular cross-linking of molecules by glutaraldehyde, which
then does not allow molecules to unfold fully in the presence of
SDS, and leads to their faster migration and apparent (anomalous)
interpretation of their molecular weight, with a reduction in
staining by the Coomassie dye [14]. In lane 2, as well as in lanes
corresponding to higher percentages of glutaraldehyde (lanes 3 and
4), there is an additional band at ~40 kDa, corresponding to a dimer
of the said ~20 kDa band. The ~40 kDa band further establishes,
over and above the gel filtration data, that HU-B-A is a mixture of
monomers (corresponding to an HU-AB dimer), and dimers (cor-
responding to an HU-AB tetramer).
HU-B-A undergoes a reversible, partial thermal unfolding
(biphasic) transition. Fig. 3A shows the CD spectra of the folded
6xHis-HU-B-linker-HU-A polypeptide construct at 20  C (black
curve), 90  C (red curve), and after cooling to 20  C from heated
state (blue curve). Approximately 50% unfolding is observed at
90  C, with near-complete refolding of this unfolded structure
observed, upon cooling. The likely interpretation of this data is that
the helical N-terminal domain of both HU-A and HU-B chain sec-
tions undergoes unfolding, whereas the C-terminal domains of HU-
A and HU-B (largely hydrophobically-stabilized through inter-
subunit inter-domain contacts) remain folded. A plot of the
change observed in ellipticity, as a function of temperature, is
shown in Fig. 3B. This plot suggests that the two N-terminal domain
sections (i.e., derived from the N-terminal domains of the HU-A and
HU-B sequences) probably do not cooperate such unfolding. Upon
heating at a constant heating rate of 2  C/minute, fitting of the data
obtained to a double sigmoidal curve yields two transitions with
apparent melting temperatures of 43.69  C and 71.55  C,
respectively.
HU-B-A is more stable to urea unfolding that either HU-AA or
HU-BB. To assess the stability of HU-B-A, 6xHis-HU-B-linker-HU-A
K. Arora, B. Thakur, A. Gupta et al. Biochemical and Biophysical Research Communications 560 (2021) 27e31

Fig. 2. Folding, DNA-binding and dimerization of HU-B-A. Panel A. Circular dichroism (CD) spectrum of HU-B-A at 20  C. Panel B. Electrophoretic mobility shift assay (EMSA) gel
showing shift in the mobility of 4WJ DNA upon binding to HU-B-A in a concentration-dependent manner. Panel C. Gel filtration chromatographic profile of HU-B-A on a Superdex-75
Increase (GE) column. Panel D. SDS-PAGE examination of HU-B-A (lane 1) as a function of glutaraldehyde concentration varying from 0% (lane 1), through 0.1% (lane 2), and 0.5%
(lane 3), to 1.0% (lane 4). Protein molecular weight markers are also shown (lane 5) with sizes of markers in kDa marked alongside.

Fig. 3. Thermal and chemical destabilization of the structure of HU-B-A. Panel A. CD spectra of HU-B-A at 20  C (black curve), 90  C (red curve) and after cooling back to 20  C
(blue curve), Panel B. Reduction in the raw ellipticity of HU-B-A at 222 nm in CD data collected as a function of temperature. Panel C. Reduction in the raw ellipticity of HU-B-A, HU-
BB and HU-AA, at 222 nm in CD data collected as a function of urea concentration, following overnight incubation, at 20  C. Panel D. Plots of variations in Gibbs free energy of HU-B-
A, HU-BB, and HU-AA, as a function of urea concentration. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this
article.)

30
K. Arora, B. Thakur, A. Gupta et al.
Table 1
Table showing Cm and DG values for urea-based unfolding of homo and hetero-
dimeric HU species.
HU dimer Cm [M] DGurea (KJ/mol)
HU-AA 1.77 1.8
HU-BB 2.14 5.8
HU-B-A 2.91 8.1
protein was incubated overnight in different concentrations of urea.
The loss in structure was assessed by monitoring changes in the CD
signal at 222 nm, as shown in Fig. 3C. The ellipticity (mdeg) values
for different urea concentrations were plotted and fitted to a
sigmoidal equation to calculate Cm values. The same data was also
used to calculate the change in free energy (DG) associated with
chemical unfolding, as shown in Table 1. The same approach was
used with HU-AA and HU-BB, to calculate the DG of unfolding of the
two homodimeric HU forms, for comparison. A comparison of pa-
rameters for HU-AA, HU-BB and HU-B-A is shown in Table 1. From
the table, and from Fig. 3C and D, it is evident that HU-B-A is more
thermodynamically stable than HU-BB or HU-AA.
4. Discussion
The folding, DNA-binding behavior, and thermodynamic sta-
bility of an HU-B-A construct (created as a simulacrum of the HU-
AB heterodimer found in E. coli) has been definitively described
in this paper. The findings establish that the simulacrum was suc-
cessfully constructed through fusion of the genes encoding HU-B
and HU-A with a flexible linker between them, and a histidine
tag at the N-terminal end of the fusion. HU-B-A could be shown to
undergo folding, associate into a dimer, and bind to DNA, demon-
strating that the association between HU-B and HU-A that ordi-
narily occurs in inter-molecular fashion can also be caused to occur
in an intramolecular fashion. The protein is not fully unfolded by
heat, but substantially unfolded by urea. Comparison of the Cm and
DG for HU-BA with HU-AA, and HU-BB establishes that the heter-
odimer displays the highest Cm and DG, causing it to be the most
chemically- and thermodynamically-stable of the three dimeric
forms of HU, and explaining why wild-type HU-AA and HU-BB
transform spontaneously into a ~90% population of HU-AB. We
know from our unpublished results that the inter dimer interface is
very stable in HU-AA and HU-BB homodimers which favour het-
erodimer formation despite this, as discussed earlier [4e7,15],
probably through the formation of a hetero-tetrameric intermedi-
ate between wild-type HU-AA and HU-BB homodimers (disallowed
by the presence of N-terminal histidine tags). We propose, there-
fore, that the higher thermodynamic stability of the HU-AB heter-
odimer (suggested by the behavior of its HU-B-A simulacrum) is the
reason that HU-A and HU-B chains prefer to form heterodimers and
remain heterodimeric.
Declaration of competing interest
The authors declare no conflicts of interest for the paper titled
31
Biochemical and Biophysical Research Communications 560 (2021) 27e31
“HU-AB simulacrum: Fusion of HU-B and HU-A into HU-B-A, a
functional analog of the Escherichia coli HU-AB heterodimer”.
Acknowledgements
KA thanks the University grant commission, Government of
India, for a doctoral fellowship. BT thanks the Council of Scientific
and Industrial Research (CSIR), New Delhi, for a doctoral fellowship.
AG thanks the Department of biotechnology, Government of India,
for a doctoral fellowship. PG thanks the Ministry of Human
Resource Development (MHRD), Government of India, for a Centre
of Excellence Grant (MHRD-14-0064) in Protein Science, Design
and Engineering.
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.bbrc.2021.04.107.
References
[1] R.T. Dame, The role of nucleoid-associated proteins in the organization and
compaction of bacterial chromatin, Mol. Microbiol. 56 (2005) 858e870.
[2] Y. Kano, K. Osato, M. Wada, F. Imamoto, Cloning and sequencing of the HU-2
gene of Escherichia coli, Mol. Gen. Genet. 209 (1987) 408e410.
[3] Y. Kano, S. Yoshino, M. Wada, K. Yokoyama, M. Nobuhara, F. Imamoto, Mo-
lecular cloning and nucleotide sequence of the HU-1 gene of Escherichia coli,
Mol. Gen. Genet. 201 (1985) 360e362.
[4] T. Ali Azam, A. Iwata, A. Nishimura, S. Ueda, A. Ishihama, Growth phase-
dependent variation in protein composition of the Escherichia coli nucleoid,
J. Bacteriol. 181 (1999) 6361e6370.
[5] J. Rouviere-Yaniv, N.O. Kjeldgaard, Native Escherichia coli HU protein is a
heterotypic dimer, FEBS Lett. 106 (1979) 297e300.
[6] L. Claret, J. Rouviere-Yaniv, Variation in HU composition during growth of
Escherichia coli: the heterodimer is required for long term survival, J. Mol.
Biol. 273 (1997) 93e104.
[7] O. Pellegrini, J. Oberto, V. Pinson, M. Wery, J. Rouviere-Yaniv, Overproduction
and improved strategies to purify the threenative forms of nuclease-free HU
protein, Biochimie 82 (2000) 693e704.
[8] N. Garnier, K. Loth, F. Coste, R. Augustyniak, V. Nadan, C. Damblon, B. Castaing,
An alternative flexible conformation of the E. coli HUbeta(2) protein: struc-
tural, dynamics, and functional aspects, Eur. Biophys. J. 40 (2011) 117e129.
[9] T. Aki, S. Adhya, Repressor induced site-specific binding of HU for transcrip-
tional regulation, EMBO J. 16 (1997) 3666e3674.
[10] F. Guo, S. Adhya, Spiral structure of Escherichia coli HUalphabeta provides
foundation for DNA supercoiling, Proc. Natl. Acad. Sci. U.S.A. 104 (2007)
4309e4314.
[11] B. Thakur, K. Arora, A. Gupta, P. Guptasarma, The DNA-binding protein HU is a
molecular glue that attaches bacteria to extracellular DNA in biofilms, J. Biol.
Chem. (2021) 100532.
[12] B. Thakur, A. Gupta, P. Guptasarma, A novel protein-engineered dsDNA-
binding protein (HU-Simulacrum) inspired by HU, a nucleoid-associated
DNABII protein, Biochem. Biophys. Res. Commun. 534 (2021) 47e52.
[13] S. Ghosh, A. Grove, Histone-like protein HU from Deinococcus radiodurans
binds preferentially to four-way DNA junctions, J. Mol. Biol. 337 (2004)
561e571.
[14] P. Tiwari, P. Kaila, P. Guptasarma, Understanding anomalous mobility of
proteins on SDS-PAGE with special reference to the highly acidic extracellular
domains of human E- and N-cadherins, Electrophoresis 40 (2019) 1273e1281.
[15] J. Ramstein, N. Hervouet, F. Coste, C. Zelwer, J. Oberto, B. Castaing, Evidence of
a thermal unfolding dimeric intermediate for the Escherichia coli histone-like
HU proteins: thermodynamics and structure, J. Mol. Biol. 331 (2003) 101e121.