Basis of recognition between the NarJ chaperone and the
N-terminus of the NarG subunit from Escherichia coli
nitrate reductase
Silva Zakian
1
, Daniel Lafitte
2
, Alexandra Vergnes
1
, Cyril Pimentel
3
, Corinne Sebban-Kreuzer
3
,
Rene
´
Toci
1
, Jean-Baptiste Claude
4
, Franc¸oise Guerlesquin
3
and Axel Magalon
1
1 Laboratoire de Chimie Bacte
´
rienne, Institut de Microbiologie de la Me
´
diterrane
´
e, Centre National de la Recherche Scientifique, Marseille,

than eight metal centres in three distinct subunits
[4–6], and the NarJ chaperone. Dynamic interactions
Keywords
chaperone; metalloproteins; nitrate
reductase; NMR; translocation
Correspondence
A. Magalon, Laboratoire de Chimie
Bacte
´
rienne, Institut de Microbiologie de la
Me
´
diterrane
´
e, Centre National de la
Recherche Scientifique, 31, chemin Joseph
Aiguier 13402 Marseille Cedex 09, France
Fax: +33 491 718 914
Tel: +33 491 164 668
E-mail: [email protected]
(Received 8 December 2009, revised 25
January 2010, accepted 4 February 2010)
doi:10.1111/j.1742-4658.2010.07611.x
A novel class of molecular chaperones co-ordinates the assembly and
targeting of complex metalloproteins by binding to an amino-terminal
peptide of the cognate substrate. We have previously shown that the NarJ
chaperone interacts with the N-terminus of the NarG subunit coming from
the nitrate reductase complex, NarGHI. In the present study, NMR
structural analysis revealed that the NarG(1–15) peptide adopts an a-helical
conformation in solution. Moreover, NarJ recognizes and binds the helical

process strongly resembles the ‘Tat proofreading’ of
periplasmic metalloproteins, of which the best-studied
example relates to E. coli trimethylamine N-oxide
reductase, TorA [8]. The targeting of this enzyme to
the Tat translocase is prevented by the TorD chaper-
one until the molybdenum cofactor has been inserted
[8]. TorD binds the TorA signal peptide, thus shielding
it from the Tat transporter [9,10].
Despite considerable research into chaperone
function, only partial structural information has been
gained on the nature and site of peptide interaction
[9–12]. Biophysical studies have indicated that Tat
signal peptides are unstructured in aqueous solution
and acquire a high degree of secondary structure in
hydrophobic environments, such as those that they may
encounter upon interaction with their partners, either
lipids from the cytoplasmic membrane or proteins such
as chaperones or components of the Tat translocase
[13,14]. Such a situation is encountered in the signal
peptide of Sec substrates, which adopts an a-helical
conformation in the SecA-bound state [15].
In the present study, the interaction between the
NarJ chaperone and the N-terminus of NarG was
studied using a series of biophysical approaches. In
particular, NMR showed that the amphiphilic a-helix
adopted by the N-terminus of NarG within the
NarGHI complex [4] is conserved in NarG(1–15) and
NarG(1–28) peptides. The docking calculation analysis
revealed that NarG(1–15) interacts within a highly
conserved elongated and hydrophobic groove of NarJ.

in agreement with the presence of an a-helix (residues
Ser2-Phe11) in both peptides (Figs 1 and 2). At pH 7,
the observed NH exchange was faster for NarG(1–15)
than for NarG(1–28), indicating the presence of a less-
structured N-terminal helix in the shorter peptide. These
observations were confirmed by structure calculations
of both peptides at pH 4.5 (Fig.3,Table 1). The struc-
ture of NarG(1–28) consisted of an a-helix (residues
2–11) followed by an antiparallel pair of b-strands
(residues 16–19 and 22–25). The N-terminal helix was
similar to that observed in the NarG X-ray structure
(rmsd = 2.84 A
˚
for the backbone) [4]. However, the
orientation of secondary structure elements was rather
different in the solution structure, probably due to the
rearrangement of the N-terminal part of NarG inter-
acting with both NarH and NarI subunits within
the NarGHI complex. Second,
1
H,
15
N-HSQC of
NarG(1–28) at natural abundance showed minor shifts
upon NarJ binding (Fig. S1). These results suggest that
the structural conformation adopted by the peptide in
solution remains unchanged upon complex formation.
Structural properties of the NarJ chaperone
E. coli NarJ is a member of a large family of dedi-
cated chaperones involved in the biogenesis of

NarJT. These observations render it impossible to solve
the structure of both NarJ and NarJT by NMR. 2D
1
H,
15
N-HSQC NMR spectra of both NarJ and NarJT
were found to be very similar (Fig. S2). Moreover,
thermal denaturation analysis of NarJ and NarJT
carried out by DSC entailed a nontwo-state transition
followed by irreversible processes. The temperature
dependence of the partial molar heat capacity of both
proteins was similar (Fig. 4A,B), indicating the exis-
tence of only one structural domain on the protein.
Upon peptide binding, NarJ undergoes a
conformational change
The temperature dependence of the partial molar heat
capacity of free NarJ or NarJT differed considerably
from that of their complexes with NarG(1–15) peptide.
There was a marked increase in thermostability (10 °C)
of both proteins due to peptide binding (Fig. 4A, B).
Moreover, titration of the complex formation between
the NarG(1–15) peptide and
15
N-labelled NarJT was
monitored by 2D
1
H,
15
N-HSQC experiments. Spectrum
analysis showed that most of the NarJ correlation

minimal complex formed between NarJT and the
NarG(1–15) peptide (Table 2). Binding reactions are
often coupled to the absorption or release of protons
by the protein or the ligand. If this is the case, the bind-
ing enthalpy is dependent on the ionization enthalpy of
the buffer in which the reaction takes place. ITC exper-
iments were therefore carried out in Hepes buffer
having a different heat of ionization (20.5 kJÆmol
)1
for
Hepes and 47.4 kJÆmol
)1
for the Tris ⁄ HCl used in the
experiments reported in Table 2) and yielded an identi-
cal biphasic isotherm with unmodified K
d
values. The
enthalpy values obtained for the complex made
between NarJ and any of the NarG peptides were lower
than with Tris ⁄ HCl buffer ()38.8 ± 4 kJÆmol
)1
in
Hepes instead of )69.4 ± 3.8 kJÆmol
)1
for Tris ⁄ HCl
for the first site and )35 ± 3.6 kJÆmol
)1
in Hepes
instead of )62.1 ± 3.1 kJÆmol
)1

loss of water-mediated hydrogen bonds. Interestingly,
this biphasic behaviour disappeared by increasing
the pH, suggesting a protonation event. At pH 8, the
binding isotherm generated a sigmoidal binding curve
that reached saturation with n =1.3±0.2andanapp-
arent K
d
=1±1· 10
)7
m for NarJT ⁄ NarG(1–15)
(Table 2, Fig. 5B). The pKa value of the protonable
A
B
Fig. 3. Ensemble of the backbone traces of the 20 lowest energy
conformers of the solution structure of (A) NarG(1–15) and (B)
NarG(1–28).
A
B
Fig. 2. (A) Sequences of NarG(1–15) (left) and NarG(1–28) (right) and sequential assignments. Collected sequential NOEs are classified into
thick and thin bars according to their relative intensity. (B) NOE distribution versus sequence of NarG(1–15) (left) and NarG(1–28) (right). Intra-
residual NOEs are in white, short NOEs are in light gray, medium-range NOEs are in dark gray, and long-range NOEs are in black.
S. Zakian et al. Structural basis for peptide recognition by NarJ
FEBS Journal 277 (2010) 1886–1895 ª 2010 The Authors Journal compilation ª 2010 FEBS 1889
residue that may be deduced from our data is lower
than 7. Combining the DSC and ITC results, we con-
clude that NarJ does not exhibit two binding sites, but
rather exists as two distinct subpopulations, probably
in rapid exchange in the free state. Each subpopulation
binds the peptide with different affinities, but uses a
similar overall mechanism. Protonation at or near the

NarG(1–15) NarG(1–28)
NMR distance and dihedral constraints
Total NOEs 243 536
Short range (|i)j| £ 1) 190 317
Medium range (1 < |i)j| < 5) 52 141
Long range (|i)j| ‡ 5) 1 78
Average pairwise rmsd
a
(A
˚
)
Heavy 2.24 ± 0.29 1.64 ± 0.19
Backbone 1.39 ± 0.32 0.94 ± 0.16
Ramachandran
Most favoured and
additional allowed (%)
100 96.2
Generously allowed (%) 0 3.8
Disallowed region (%) 0 0
a
Calculated among 20 [NarG(1–15)] and 15 [NarG(1–28)] refined
structures.
AB
C
Fig. 4. Deconvolution of the transition
excess heat capacity of (A) NarJ and
(B) NarJT alone (black traces) or in complex
with NarG(1–15) (red traces). Solid lines,
experimental data; dotted lines,
deconvolution peaks. NarJ 50.9 ± 1 °C;

confirming the existence of a single state of NarJ at this
pH. These results are in full agreement with those
obtained with ITC (Table 2). Interestingly, at pH 7, k
off
varied by nearly a factor of 10 between the two subpop-
ulations, i.e. k
off
= 3.2 ± 1 s
)2
for the minor species of
high affinity and k
off
= 1.9 ± 1 s
)1
for the major
species of lower affinity. Overall, we concluded that
protonation of a specific residue of NarJ modulates the
peptide binding affinity, in particular via the lifespan of
the protein–peptide complex.
Conclusion
One important finding is the structural flexibility of the
NarJ chaperone and its conformational rearrangement
upon NarG binding. Examination of the crystal
structure of several members of this new family of
chaperones [11,16,17] indicates the presence of several
disordered regions. Moreover, ITC data obtained by
others on E. coli TorD [9] and DmsD [12] have sys-
tematically shown a strong decrease in entropy associ-
ated with the complex formation. Overall, structural
flexibility appears to be a common feature of this new

)
TDS
(kJÆmol
)1
)
TDS
corr
a
(kJÆmol
)1
)
NarJ–NarG(1–28) 0.2 ± 0.1 2.3 ± 4 · 10
)9
)69.4 ± 3.8 )22 )20.1 27.3
0.9 ± 0.1 1.7 ± 3.1 · 10
)7
)62.1 ± 3.1 )14.7 )23.4 24
NarJT–NarG(1–28) 0.3 ± 0.1 7.3 ± 3.8 · 10
)9
)57 ± 2.7 )9.6 )10.6 36.8
0.7 ± 0.2 1.9 ± 2.9 · 10
)7
)50.1 ± 3 )2.7 )11.8 35.6
NarJ–NarG(1–15) 0.3 ± 0.2 3.4 ± 4 · 10
)9
)56.4 ± 2.2 )9 )8.1 39.3
0.7 ± 0.1 3.3 ± 3 · 10
)7
)50.8 ± 2.6 )3.4 )13.8 33.6
NarJT–NarG(1–15) 0.3 ± 0.1 3.5 ± 2 · 10

)1
).
S. Zakian et al. Structural basis for peptide recognition by NarJ
FEBS Journal 277 (2010) 1886–1895 ª 2010 The Authors Journal compilation ª 2010 FEBS 1891
metal cofactor insertion processes through additional
contacts with their specific partner [1]. Such structural
flexibility may not only contribute to their high specific-
ity during the binding process, but may also be of para-
mount importance with regard to their multiple
functions during the biogenesis of the partner. An
exception would be the NapD chaperone having a fer-
redoxin-type fold, which undergoes only minor confor-
mational changes upon binding the twin-arginine signal
peptide of NapA [18]. In this case, biogenesis of the
NapA protein is assisted by NapF in charge of cofactor
loading [19,20]. Overall, considering the global confor-
mational change of the chaperone observed upon pep-
tide binding, it is as essential to solve the structure of
the chaperone–peptide complex as to evaluate quantita-
tively the structural flexibility of the chaperone.
An unexpected finding was the discovery of the
pH-dependent modulation of the peptide binding
affinity by changing the lifespan of the chaperone–
peptide complex. Indeed, deprotonation of a yet
unidentified residue of NarJ drastically reduces the pep-
tide binding affinity by 100-fold and the lifespan of the
complex by 10-fold, as judged by k
off
. The physiological
chaperone cycle probably consists of the rapid binding

proteins were produced using M9 minimum media and
15
N-labelled NH
4
Cl.
N-terminal NarG peptides
The NarG(1–15) MSKFLDRFRYFKQKG and NarG(1–28)
MSKFLDRFRYFKQKGETFADGHGQLLNT peptides
used in this study were chemically synthesized and purified
by Synprosis (Marseilles, France). The molecular mass of
each peptide was verified by mass spectrometry.
NMR experiments for NarG peptide structure
calculation
NMR experiments were performed at 293 K, on a 1 mm
peptide sample in 10 mm potassium phosphate buffer at
pH 4.5. Homonuclear NOESY, TOCSY and COSY spectra
and a 24 h
1
H,
15
N-HSQC spectrum at natural abundance
were recorded for each peptide on a Bruker 600 MHz spec-
trometer equipped with a TCN cryoprobe. Spectra were
processed using the topspin 2.1 software (Bruker BioSpin
S.A., Wissembourg Ce
´
dex, France).
C-ter
N-ter
Fig. 6. Interaction surface between NarJT and the N-terminus of

obtain the effective heat of binding [27].
DSC
Heat denaturation measurements were carried out on a
MicroCal VP-DSC instrument (Microcal LLC) at a heating
rate of 1 KÆmin
)1
. The denaturation temperature was deter-
mined as previously described [28]. Because of the irrevers-
ibility of the denaturation process, the excess molar heat
capacity of the protein could not be determined.
BIAcore surface plasmon resonance analysis
All experiments were carried out at 298 K on a BIAcore
3000 apparatus (BIAcore, GE Healthcare Europe GmbH,
Orsay, France). NarJ–His6 was immobilized on a CM5
sensor chip using amine coupling [21]. NarG(1–15) peptide
in 10 mm Tris ⁄ HCl, 150 mm NaCl, 3.4 mm EDTA, 0.005%
surfactant P20 and pH 7 or 8 was then injected over the
test and control (no protein immobilized) surfaces at a flow
rate of 60 lLÆmin
)1
. The sensor surface was regenerated
with an injection of 1 mm NaOH final concentration. The
resulting sensorgrams were evaluated using the biomolecu-
lar interaction analysis evaluation software (BIAcore) to
calculate the kinetic constants of the complex formation.
Molecular docking
A molecular model of NarJT was obtained using modeller
software. Briefly, the NarJ sequence was first used to find
related structures from the Protein Data Bank using the
NCBI server Psi-Blast. To improve the overall quality of

1 Magalon A & Mendel RR (2008) Biosynthesis and
insertion of the molybdenum cofactor. In
EcoSal—Escherichia coli and Salmonella: cellular and
molecular biology (Bo
¨
ck A, Curtiss R III, Kaper JB,
Karp PD, Neidhardt FC, Nystro
¨
m T, Slauch JM &
Squires CL eds). ASM Press, Washington, DC.
2 Sargent F (2007) Constructing the wonders of the
bacterial world: biosynthesis of complex enzymes.
Microbiology, 153, 633–651.
3 Vergnes A, Pommier J, Toci R, Blasco F, Giordano G
& Magalon A (2006) NarJ chaperone binds on two
distinct sites of the aponitrate reductase of Escherichia
coli to coordinate molybdenum cofactor insertion and
assembly. J Biol Chem, 281, 2170–2176.
4 Bertero MG, Rothery RA, Palak M, Hou C, Lim D,
Blasco F, Weiner JH & Strynadka NC (2003) Insights
into the respiratory electron transfer pathway from the
structure of nitrate reductase A. Nat Struct Biol, 10,
681–687.
5 Rothery RA, Blasco F, Magalon A & Weiner JH
(2001) The diheme cytochrome b subunit (Narl) of
S. Zakian et al. Structural basis for peptide recognition by NarJ
FEBS Journal 277 (2010) 1886–1895 ª 2010 The Authors Journal compilation ª 2010 FEBS 1893
Escherichia coli nitrate reductase A (NarGHI):
structure, function, and interaction with quinols. J Mol
Microbiol Biotechnol 3, 273–283.

signal peptide binding to DmsD. Arch Biochem Biophys,
455, 89–97.
13 Kipping M, Lilie H, Lindenstrauss U, Andreesen JR,
Griesinger C, Carlomagno T & Bruser T (2003) Struc-
tural studies on a twin-arginine signal sequence. FEBS
Lett, 550, 18–22.
14 San Miguel M, Marrington R, Rodger PM, Rodger A
& Robinson C (2003) An Escherichia coli twin-arginine
signal peptide switches between helical and unstructured
conformations depending on the hydrophobicity of the
environment. Eur J Biochem , 270, 3345–3352.
15 Gelis I, Bonvin AM, Keramisanou D, Koukaki M,
Gouridis G, Karamanou S, Economou A & Kalodimos
CG (2007) Structural basis for signal-sequence recogni-
tion by the translocase motor SecA as determined by
NMR. Cell, 131, 756–769.
16 Tranier S, Iobbi-Nivol C, Birck C, Ilbert M,
Mortier-Barriere I, Mejean V & Samama JP (2003) A
novel protein fold and extreme domain swapping in the
dimeric TorD chaperone from Shewanella massilia.
Structure, 11, 165–174.
17 Qiu Y, Zhang R, Binkowski TA, Tereshko V,
Joachimiak A & Kossiakoff A (2008) The 1.38 A
crystal structure of DmsD protein from Salmonella
typhimurium, a proofreading chaperone on the Tat
pathway. Proteins, 71, 525–533.
18 Maillard J, Spronk CA, Buchanan G, Lyall V,
Richardson DJ, Palmer T, Vuister GW & Sargent F
(2007) Structural diversity in twin-arginine signal
peptide-binding proteins. Proc Natl Acad Sci USA, 104,

Kaptein R & Thornton JM (1996) AQUA and
PROCHECK-NMR: programs for checking the quality
of protein structures solved by NMR. J Biomol NMR,
8, 477–486.
27 Makarov AA, Tsvetkov PO, Villard C, Esquieu D,
Pourroy B, Fahy J, Braguer D, Peyrot V & Lafitte D
(2007) Vinflunine, a novel microtubule inhibitor,
suppresses calmodulin interaction with the microtubule-
associated protein STOP. Biochemistry 46, 14899–14906.
28 Privalov PL & Potekhin SA (1986) Scanning microcal-
orimetry in studying temperature-induced changes in
proteins. Methods Enzymol, 131, 4–51.
29 Notredame C, Higgins DG & Heringa J (2000)
T-Coffee: a novel method for fast and accurate multiple
sequence alignment. J Mol Biol, 302, 205–217.
Structural basis for peptide recognition by NarJ S. Zakian et al.
1894 FEBS Journal 277 (2010) 1886–1895 ª 2010 The Authors Journal compilation ª 2010 FEBS
30 Sali A & Blundell TL (1993) Comparative protein
modelling by satisfaction of spatial restraints. J Mol
Biol, 234, 779–815.
31 Dominguez C, Boelens R & Bonvin AM (2003)
HADDOCK: a protein-protein docking approach based
on biochemical or biophysical information. J Am Chem
Soc, 125, 1731–1737.
Supporting information
The following supplementary material is available:
Fig. S1. Overlay of
1
H,
15