Systematic search for zinc-binding proteins in
Escherichia coli
Akira Katayama
1,2
, Atsuko Tsujii
2
, Akira Wada
3
, Takeshi Nishino
2
and Akira Ishihama
1
National Institute of Genetics, Department of Molecular Genetics, Mishima, Shizuoka, Japan;
2
Nippon Medical School, Department
of Biochemistry and Molecular Biology, Bunkyo-ku, Tokyo, Japan;
3
Osaka Medical School, Department of Physics, Takatsuki,
Osaka, Japan
AsystematicsearchforEscherichia coli proteins with the
zinc-binding activity was performed using the assay of
radioactive Zn(II) binding to total E. coli proteins fraction-
ated by two methods of two-dimensional gel electrophoresis.
A total of 30–40 radioactive spots were identified, of which
14 have been assigned from N-terminal sequencing. In
addition to five known zinc-binding proteins, nine zinc-
binding proteins were newly identified including: acetate
kinase (AckA), DnaK, serine hydroxymethyltransferase
(GlyA), transketolase isozymes (TktA/TktB), translation
elongation factor Ts (Tsf), ribosomal proteins L2 (RplB),
L13 (RplM) and one of S15 (RpsO), S16 (RpsP) or S17

of the zinc-binding motif have been identified, including
those tetrahedrally co-ordinated to imidazole nitrogen
atoms from His and thiol groups from Cys [2]. The
functions of protein-bound zinc are beginning to catch up
with the increasing number of zinc-containing proteins. Up
to the present time, the structural and functional roles of
zinc have been analysed in detail with zinc-containing
proteins from higher eukaryotes, but little is known about
the zinc-binding proteins in prokaryotes. Ros homologues
that exit in plant-associated agrobacteria are the only
bacterial proteins with the typical C2H2-type zinc-finger
motif [4]. Various types of zinc-containing protein with the
zinc finger or zinc cluster exist in yeast, which, however,
lacks proteins with the hormone receptor-type zinc-binding
motif. This type of zinc-binding proteins appears in
Caenorhabditis elegans [5] and the number of this type of
zinc-binding proteins increases in higher animals.
The aim of this study is to identify as many zinc-binding
proteins as possible in the model prokaryote, Escherichia
coli. For a systematic and experimental detection of zinc-
binding proteins, we employed a conventional method of
radioactive zinc blotting with whole cell extracts fraction-
ated by two methods of two-dimensional gel electrophor-
esis, the widely used O’Farrell system [6] and the newly
developed radical-free and highly reducing (RFHR) method
[7]. Results indicate that most of the newly identified
bacterial zinc-binding proteins do not contain the known
zinc-binding motifs, most of which have been identified in
higher eukaryotes. This report also describes the affinity and
specificity of zinc binding for some of the E. coli zinc-

2
and DNase I (Sigma)
were added to the final concentration of 2% (v/v), 10 m
M
and 16.7 UÆmL
)1
, respectively. After removal of insoluble
materials (unlysed cells and cell wall aggregates) by centrif-
ugation at 10 000 g for 10 min at 4 °C, the supernatant
fraction was used as the whole cell extract. The supernatant
fraction was prepared after removal of ribosomes by
centrifugation for 1 h at 100 000 g at 4 °C. Protein
concentration was determined with a Bio-Rad protein assay
kit (Bio-Rad).
Total acid-soluble proteins were prepared from the whole
cell lysate by adding 2 vol. of acetic acid at 4 °C [10]. One
hour after addition of acetic acid, the acid-treated extract
was centrifuged for 10 min at 10 000 g at 4 °Ctoremove
acid-precipitated proteins.
Two dimensional gel electrophoresis
Two methods of two-dimensional gel electrophoresis system
were employed throughout this study. In the O’Farrell
method [6], proteins were separated according to isoelectric
point by isoelectric focusing in the first dimension and then
according to molecular size after denaturation with SDS in
the second dimension. In this series of experiments, we
employed the Pharmacia system and apparatus. In brief, the
18-cm Dry Strips, pH 4–7 (Pharmacia), were rehydrated for
12 h with 0.5 mL of rehydration solution [8
M

Blotting of radioactive zinc
Proteins in the second dimension gels were blotted onto
poly(vinylidene difluoride) (PVDF) membranes (Nippon
Genetics, Japan) using a semidry blotting apparatus (Bio-
Rad). The protein-blotted membranes were subjected to
binding assay of radioactive zinc essentially according to the
method of Mazen et al. [12]. In brief, the protein-blotted
membranes were soaked with buffer A (10 m
M
Tris/HCl,
pH 7.5 at 4 °C) for 1 h at room temperature, and then
incubated in buffer B (10 m
M
Tris/HCl, pH 7.5 at 4 °C,
0.1
M
KCl) containing 0.2 lCiÆmL
)165
ZnCl
2
(specific activ-
ity, 1.84 mCiÆmg
)1
; NEN Life Science Products, Inc.) for
15 min. After washing for 15 min in buffer B, the membranes
were exposed to imaging plates and the plates were analyzed
with a BAS-2000 imaging analyzer (Fuji Film Co., Japan).
Construction of expression plasmids for zinc-binding
proteins
The genes coding for zinc-binding proteins were amplified

lysozyme, 0.2 m
M
phenylmethanesulfonyl fluoride and 1 m
M
dithiothreitol,
followed by sonication. The cell lysates were centrifuged for
2 h at 100 000 g at 4 °C, and the supernatant was directly
subjected to affinity chromatograpy on Ni
2+
-nitrilotriacetic
acid agarose columns (Qiagen) previously equilibrated with
lysis buffer. After washing with lysis buffer containing 3 m
M
imidazole, column-bound proteins were eluted with elution
buffer containing 200 m
M
imidazole. Peak fractions of each
zinc-binding protein were pooled and after dialysis against
storage buffer [10 m
M
Tris/HCl, pH 8.0 at 4 °C, 10 m
M
MgCl
2
,0.1m
M
EDTA, 1 m
M
dithiothreitol, 0.2
M

soluble fraction after removal of the group of abundant
proteins such as membrane proteins, ribosomal proteins
and nucleoid-associated DNA-binding proteins (or nucleoid
proteins). The overall pattern of Zn(II) binding was
essentially the same with that obtained using whole cell
extracts (compare Fig. 1B,D). Some Zn(II)-binding spots
detected with the whole cell extracts disappeared, and
instead some new radioactive spots were detected because of
the increase in relative levels for those proteins in the soluble
protein fraction. Because the intensity of Zn(II) binding and
the staining intensity with CBB do not correlate and because
the pattern of radioactive Zn(II) binding differs between
two different preparations of the cell extract, we concluded
that the filter binding assay herein employed allows the
detection of at least a group of proteins with Zn(II)-binding
activity. It should be noted, however, that the intensity of
Zn(II) radioactivity thus detected reflects both the affinity to
Zn(II) and the protein concentration.
To increase in the resolution of basic proteins, we also
employed the RFHR method of two-dimensional gel
electrophoresis [7]. In addition, artefacts arising from
oxidation of proteins during electrophoresis could also be
avoided by using the RFHR method. Figure 2 shows one
example of the Zn(II)-binding assay for acid-soluble basic
proteins separated by the RFHR method. Some of the basic
proteins were newly identified as having Zn(II)-binding
activity. After sequence analysis, these basic proteins were
identified as specific ribosomal proteins (see below). The
ribosomal proteins detected in the region of gel electro-
Fig. 1. Radioactive Zn(II) binding assay of E. coli total proteins fractionated by the O’Farrell method of two-dimensional gel electrophoresis. Awhole

gel electrophoresis. In the case of AhpC/Ppa spot, Ppa was
identified to be the Zn(II)-binding component, because
the isolated inorganic pyrophosphatase (Ppa) exhibited the
Zn(II)-binding activity (see below), in agreement with the
previous observations [14–16]. Lipoamide dehydrogenase
(LpdA) [17], fructose-1,6-bisphosphatase aldolase (Fba)
[18,19], phosphotransacetylase (Pta) [20] and RNA poly-
merase a subunit (RpoA) [21] have all been shown to
contain or bind zinc in its isolated state. All these
observations indicate that the method employed in this
study is useful for the identification of yet unidentified
Zn(II)-binding proteins. Because the intracellular concen-
trations of Zn(II)-binding proteins thus identified are,
however, different in the cell extracts analysed, the different
intensity of radioactivity for these spots does not necessarily
represent the relative affinity of Zn(II)-binding.
Fig. 2. Radioactive Zn(II) binding assay of E. coli acid-soluble proteins fractionated by the RFHR method of two-dimensional gel electrophoresis.
Acid-soluble proteins (380 mg), prepared from the whole cell lysate of exponentially growing E. coli W3110 type-A, were fractionated by the
RFHR method [7] of two-dimensional PAGE. After electrophoresis, proteins were blotted onto PVDF membranes followed by staining with
Commassie Brilliant Blue R250 (A). The locations of low-molecular-mass ribosomal proteins are indicated. The protein-blotted membrane was also
subjected to the binding assay of radioactive Zn(II) as described in Materials and methods (B), followed by exposure to an imaging plate, which was
then analysed with a BAS-2000 image analyser (Fuji, Japan). The Zn(II)-binding proteins indicated by arrows were analysed for the N-terminal
sequences. The spot numbers correspond to the gene numbers listed in Table 1.
Table 1. Zinc binding proteins. Protein spots with zinc-binding activity
were cut out from PVDF membranes and immediately subjected to
N-terminal microsequencing. From the sequence results, the genes
coding for these proteins were identified. Protein spots, AhpC/Ppa,
AtpA/LpdA, RpsO/RpsP/RpsQ and TktA/Tkt, were found to contain
multiple proteins, among which Ppa and LpdA were identified to have
the zinc-binding activity. Asterisks indicate the previouslyidentified

formed E. coli cells, and affinity-purified the His
6
-tagged
proteins to apparent homogeneity. Equal amounts of the
purified Zn(II)-binding proteins were subjected to the
Zn(II)-binding assay. Figure 3 shows the
65
Zn(II)-binding
activity for two proteins, RplM (ribosomal protein L13) and
RpoA (RNA polymerase a subunit), in parallel with four
known heavy metal-binding proteins, Zn(II)-binding Def
(peptide deformylase) [22,23], Hg(II)-binding MerR (mer-
cury export regulation protein) with Zn(II)-binding activity
[24,25] and Fe(II)-binding Zn(II)-containing Fur (ferric
uptake regulation protein) [26,27] and Zn(II)-binding BSA
(bovine serum albumin) [28], and one control protein,
RpoD (RNA polymerase r subunit) with no known activity
of Zn(II)-binding.
65
Zn(II) binding was detected even at low
protein concentrations for Def, MerR, Fur and RpoA
(Fig. 3C), and in addition, for RplM and BSA at high
protein concentrations (Fig. 3D). In addition to RplM and
RpoA, Zn(II)-binding activity was confirmed for other
three purified proteins, AckA (acetate kinase), Ppa (inor-
ganic pyrophosphatase) and DnaJ (see below). Here we
detected a low level of Zn(II) binding for RpoD, the RNA
polymerase r
70
subunit (see also below). At present,

Zn(II) as described in Materials and methods (C and D), followed by exposure to an imaging plate which was analysed with a BAS-2000 image
analyser. CH represents the hexahistidine (His) tag added at the C-terminus of each protein. RPase core indicates the RNA polymerase core enzyme
with the subunit composition a
2
bb¢.
Ó FEBS 2002 Escherichia coli zinc-binding proteins (Eur. J. Biochem. 269) 2407
Affinity of Zn(II) binding by the purified Zn(II)-binding
proteins
Using the purified proteins, we then measured the Zn(II)-
binding affinity. For this purpose, the level of protein-bound
65
Zn(II) was measured using the same amounts of the newly
identified Zn(II)-binding proteins, AckA and Ppa, and the
known metal-binding proteins, Def, MerR and Fur, in the
presence of a fixed amount of radioactive Zn(II) and
increasing concentrations of nonradioactive Zn(II). The
level of Zn(II) binding to all the test proteins increased
concomitantly with the increase of Zn(II) concentration
[note the decrease in specific radioactivity of
65
Zn(II)]
(Fig. 4). The affinity of Zn(II) binding by both AckA and
Ppa was estimated to be within the same range of those of
the reference proteins, Def, MerR and Fur.
Specificity of Zn(II) binding by the whole cell extracts
To examine the specificity of Zn(II) binding for the known
and novel E. coli Zn(II)-binding proteins, we performed the
competitive inhibition assay of
65
Zn(II) binding with other

. The amount of radioactive Zn(II)
binding decreased by the addition of nonradioactive Zn(II)
(Fig. 6B) and the reduction level was apparently the same
between the test proteins. In contrast, the Zn(II) binding
activity by these proteins was not affected by the addition of
Mg(II) (Fig. 6C) and Fe(III) (Fig. 6D). Although the Fur
protein has a high affinity to Fe(III), the binding of
radioactive Zn(II) to Fur was not interfered with the
addition of Fe(III) (Fig. 6D), being consistent with the
finding that the site of zinc binding is different from the iron-
binding site [27].
The Zn(II) binding activity of AckA, Def, DnaJ, Fur,
MerR, Ppa, and RplM was also tested in the presence of
Ca(II), Cu(II) and Cd(II) (Fig. 7), but none of these metals
affected the binding of radioactive Zn(II) to the test
proteins. Taken the results of all these competition assays
together we concluded that the observed Zn(II)-binding
activity by the test proteins represents the specific binding of
Zn(II).
Fig. 4. Dose-dependent binding of Zn(II) to purified E. coli zinc-binding proteins. The newly identified E. coli Zn(II)-binding proteins, AckA and
Ppa, were analysed, together with the known Zn(II)-binding proteins, Def, MerR, Fur and Sp1, for radioactive Zn(II) binding in the presence of
indicated concentrations of nonradioactive ZnCl
2
. Proteins analysed were: lane 1, AckA(CH) + Ppa(CH) + MerR(CH) + Sp1; and lane 2,
Def(CH) + Fur(CH). After electrophoresis, the proteins were blotted onto PVDF membranes and the protein-blotted membranes were subjected
to the binding assay of
65
ZnCl
2
(0.2 lCiÆmL

M
ZnCl
2
(C), MgCl
2
(D), NiCl
2
(E) or CdCl
2
(F). The filters were treated as described in Fig. 4.
Fig. 6. Competition of radioactive Zn(II) binding by nonradioactive metals, Zn(II), Mg(II) and Fe(III), to purified E. coli zinc-binding proteins.
Mixtures of the purified Zn(II)-binding proteins were subjected to radioactive Zn(II)-binding assay in the absence (A) or presence of 100 l
M
nonradioactive Mg(II) (B), Zn(II) (C) or Fe(III) (D). Proteins analysed were: Lane 1, AckA; lane 2, AckA(CH) + Ppa(CH) + Sp1; lane 3, Ppa;
lane 4, RpoA(CH) + RplM(CH); lane 5, RpoA + MerR(CH); lane 6, Fur(CH); lane 7, Def(CH); lane 8, DnaJ(CH). After SDS/PAGE, gels were
treated as described in Fig. 4.
Ó FEBS 2002 Escherichia coli zinc-binding proteins (Eur. J. Biochem. 269) 2409
details for zinc-containing metalloproteins from higher
animals (reviewed in [3]). In contrast, relatively little is
known about the roles of zinc in function and structure of
proteins from prokaryotes. The total number of known
zinc-containing E. coli protein species that have been
experimentally examined to date is not more than 20
(Table 2). Here we performed, for the first time in E. coli
molecular genetics, a systematic search for E. coli proteins
Fig. 7. Competition of radioactive Zn(II) binding to purified E. coli zinc-binding proteins by nonradioactive metals, Ca(II), Cu(II) and Cd(II). Mixtures
of the purified Zn(II)-binding proteins (100 pmol each) were subjected to the Zn(II)-binding assay (
65
ZnCl
2

formation of native conformations or the expression of their
intrinsic functions.
The success of detection of specific Zn(II)-binding
proteins agrees with the prediction that the formation of
Zn(II)-binding site includes short stretches of the protein
sequence that can be easily refolded after denaturation
during electrophoretic separation of proteins. The E. coli
proteins detected in this study must belong to a group of
proteins with similar affinity to Zn(II). In fact, the binding
affinity of Zn(II) as measured by competition assay (see
Fig. 4) was similar between this group of proteins. Conse-
quently It can not be excluded that Zn(II)-binding motifs,
which require longer stretches of the protein and can not be
refolded after one cycle of denaturation-renaturation treat-
ment, could not be detected by the method herein employed.
Under laboratory culture conditions, E. coli expresses at
most 1000 genes out of 4000 ORFs on the genome, as
detected by two-dimensional gel electrophoresis. In this
study, we identified 20–30 Zn(II)-binding proteins among
300–400 E. coli proteins (or 5–10% of total open reading
frames). If the relative amount of Zn(II)-binding proteins in
terms of total protein species is the same for other E. coli
proteins, the total number of zinc-binding proteins in E. coli
could be about 200–400 (or 5–10% of a total of 4000 open
reading frames on the E. coli genome). Most of the Zn(II)-
binding proteins herein detected do not contain any of
the known zinc-binding motifs, which were identified in
zinc-containing proteins from higher eukaryotes. After a
computational search for Zn(II)-binding E. coli proteins,
however, we identified only a total of about 30 proteins

employed (Fig. 2), however, L36 migrated outside the gel
space that was subjected to Zn(II)-binding assay.
ACKNOWLEDGEMENTS
We thank Katsunori Yata (National Institute of Genetics, RI Centre)
for expression and purification of Fur and MerR proteins, and Yukio
Sugiura (Kyoto University) for the gift of Sp1. This work was
supported by Grants-in-Aid from the Ministry of Education, Science,
Culture and Sports of Japan, and by CREST fund from the Japan
Science Corporation (JSP).
REFERENCES
1. Vallee, B.L. & Auld, D.S. (1990) Zinc coordination, function, and
structure of zinc enzymes and other proteins. Biochemistry 29,
5647–5659.
2. Coleman, J.E. (1992) Zinc proteins: enzymes, storage proteins,
transcription factors, and replication proteins. Annu. Rev. Bio-
chem. 61, 897–946.
3. Vallee, B.L. & Falchuk, K.H. (1993) The biochemical basis of zinc
physiology. Physiol. Rev. 73, 79–118.
4. Bouhouche, N., Syvanen, M. & Kado, C.I. (2000) The origin
of prokaryotic C2H2 zinc finger regulators. Trends Microbiol. 8,
77–81.
Table 2. Zinc-binding proteins in Escherichia coli.
Protein Function Reference
Ada Transcription factor for ada operon [34]
Cdd Cytidine deaminase [35]
Def Peptide deformylase (PDF) [22]
DnaG Primase for DNA replication [36]
DnaJ Molecular chaperone [30]
Fba Fructose 1,6-bisphosphatase [18]
Fpg Formamidepyrimidine-DNA glycosylase [37]

10. Hardy, S.J., Kurland, C.G., Voynow, P. & Mora, G. (1969) The
ribosomal proteins of Escherichia coli, I. Purification of the 30S
ribosomal proteins. Biochemistry 8, 2897–2905.
11. Wada, A., Yamazaki, Y., Fujita, N. & Ishihama, A. (1990)
Structure and probable genetic location of a Ôribosome modu-
lation factorÕ associated with 100S ribosomes in the stationary-
phase Escherichia coli cells. Proc.NatlAcad.Sci.USA87,
2657–2661.
12. Mazen, A., Grandwohl, G. & de Murcia, G. (1988) Zinc-binding
proteins detected by protein labeling. Anal. Biochem. 172, 39–42.
13. Katayama, A., Fujita, N. & Ishihama, A. (2000) Mapping of
subunit-subunit contact surfaces on the b¢ subunit of Escherichia
coli RNA polymerase. J. Biol. Chem. 275, 3583–3592.
14. Avaeva, S.M., Ignatov, P., Kurilova, S., Nazarova, T., Rodina, E.,
Vorobyeva, N., Oganessyan, V. & Harutyunyan, E. (1996)
Escherichia coli inorganic pyrophosphatase: site-directed muta-
genesis of the metal binding sites. FEBS Lett. 399, 99–102.
15. Efimova, I.S., Salminen, A., Pohjanjoki, P., Lapinniemi, J.,
Magretova,N.N.,Cooperman,B.S.,Goldman,A.,Lahti,R.&
Baykov, A.A. (1999) Directed mutagenesis studies of the metal
binding site at the subunit interface of Escherichia coli inorganic
pyrophosphatase. J. Biol. Chem. 274, 3294–3299.
16. Kankare, J., Salminen, T., Lahti, R., Cooperman, B.S., Baykov,
A.A. & Goldman, A. (1996) Crystallographic identification of
metal-binding sites in Escherichia coli inorganic pyrophosphatase.
Biochemistry 35, 4670–4677.
17. Olsson, J.M., Xia, L., Eriksson, L.C. & Bjornstedt, M. (1999)
Ubiquinone is reduced by liposamide dehydrogenase and
this reaction is potently stimutaled by zinc. FEBS Lett. 448,
190–192.

resistance proteins and their use in biosensers for the detection of
bioavailable heavy metals. J. Inorg. Biochem. 79, 225–229.
26. Althaus,E.W.,Outten,C.E.,Olson,K.E.,Cao,H.&Oı
´
Halloran,
T.V. (1999) The ferric uptake regulation (Fur) repressor is a zinc
metalloprotein. Biochemistry 38, 6559–6569.
27. Gonzalez de Peredo, A., Saint-Pierre, C., Adrait, A., Lacquamet,
L., Latour, J.M., Michaud-Soret, I. & Forest, E. (1999) Identifi-
cation of the two zinc-bound cysteines in the ferric uptake reg-
ulation protein from Escherichia coli: chemical modification and
mass spectrometry analysis. Biochemistry 38, 8582–8589.
28. Ohyoshi, K., Hamada, Y., Nakata, K. & Kohata, S. (1999) The
interaction between human and bovine serum albumin and zinc
studied by a competitive spectrophotometry. Inorg. Biochem. 75,
213–218.
29. Singh-Wissmann, K., Ingram-Smith, C., Miles, R.D. & Ferry,
J.G. (1998) Identification of essential glutamates in the acetate
kinase from Methanosarcina thermophila. J. Bacteriol. 180, 1129–
1134.
30. Banecki,B.,Liberek,K.,Wall,D.,Wawrzynow,A.,Georgopo-
ulos,C.,Bertoli,E.,Tanfani,F.&Zylicz,M.(1996)Structure–
function analysis of the zinc finger region of the DnaJ molecular
chaperone. J. Biol. Chem. 271, 14840–14848.
31. Martinez-Yamout,M.,Legge,G.B.,Zhang,O.,Wright,P.E.&
Dyson, H.J. (2000) Solution structure of the cysteine-rich domain
of the Escherichia coli chaperone protein DnaJ. J. Mol. Biol. 300,
805–818.
32. Wada, A. & Sako, T. (1987) Primary structures of and genes for
new ribosomal proteins A and B in Escherichia coli. J. Biochem.

Shoolingin-Jordan. (1999) X-ray structure of 5-aminolevuinic
acid dehydratase from Escherichia coli complexed with the inhi-
bitor levulinic acid at 2.0 A
˚
resolution. Biochemistry 38,
4266–4276.
40. Gonzalez,J.C.,Peariso,K.,Penner-Hahn,J.E.&Matthews,R.G.
(1996) Cobalamin-independent methionine synthase from
Escherichia coli: a zinc metalloenzyme. Biochemistry 35, 12228–
12234.
2412 A. Katayama et al. (Eur. J. Biochem. 269) Ó FEBS 2002
41. Goulding, C.W. & Matthews, R.G. (1997) Cobalamin-
dependent methionine synthase from Escherichia coli: involve-
ment of zinc in homocsteine synthesis. Biochemistry 36,
15749–15757.
42. Sowadski, J.M., Handschumacher, M.D., Murthy, H.M., Foster,
B.A. & Wyckoff, H.W. (1985) Refined structure of alkaline
phosphatase from Escherichia coli at 2.8 A
˚
resolution. J. Mol. Biol.
186, 417–433.
43. Markov, D., Naryshkina, T., Mustaev, A. & Severinov, K. (1999)
A zinc-binding site in the largest subunit of DNA-dependent RNA
polymerase is invovled in enzyme assembly. Genes Dev. 13, 2439–
2448.
44. Chatterji, D., Wu, C.W. & Wu, F.Y. (1986) Spatial relationship
between the intrinsic metal in the beta subunit and cysteine-132
in the sigma subunit of Escherichia coli RNA polymerase: a
resonance energy transfer. Arch. Biochem. Biophys. 244, 218–225.
45. Navaratnam, S., Myles, G.M., Strange, R.W. & Sancar, A. (1989)