Effects of promoter mutations on the in vivo regulation of the cop operon of Enterococcus hirae by copper(I) and copper(II).

Abstract

The cop operon of Enterococcus hirae encodes a repressor, CopY, a copper chaperone, CopZ, and two copper ATPases, CopA and CopB. Regulation of the cop operon is bi-phasic, with copper addition as well as copper chelation leading to induction. Using a plasmid-borne system with a reporter gene, induction of wild-type and mutant cop promoters by high and low copper conditions was investigated. Only mutations that impaired the interaction of CopY with both DNA binding sites had a marked effect on regulation, leading to hyperinduction by copper(I) or copper(II). Chelation of copper(II), but not copper(I), also induced the operon, but induction by copper chelation was not significantly affected by the mutations. E. hirae mutants with reduced extracellular copper reductase activity exhibited the same induction kinetics as wild-type cells. These results show that copper addition and copper chelation induce the cop operon by different routes.

Full text

Page 1

Effects of Promoter Mutations on the in Vivo Regulation
of the cop Operon of Enterococcus hirae
by Copper(I) and Copper(II)
Haibo Wunderli-Ye and Marc Solioz1
Department of Clinical Pharmacology, University of Berne, 3010 Berne, Switzerland
Received May 5, 1999
T he cop operon of E nterococcus hirae encodes a re-
pressor, CopY, a copper chaperone, CopZ, and two
copper AT Pases, CopA and CopB. R egulation of the
cop operon is bi-phasic, with copper addition as well
as copper chelation leading to induction. Using a
plasmid-borne system with a reporter gene, induction
of wild-type and mutant cop promoters by high and
low copper conditions was investigated. Only muta-
tions that impaired the interaction of CopY with both
DNA binding sites had a marked effect on regulation,
leading to hyperinduction by copper(I) or copper(II).
Chelation of copper(II), but not copper(I), also in-
duced the operon, but induction by copper chelation
was not significantly affected by the mutations. E .
hirae mutants with reduced extracellular copper re-
ductase activity exhibited the same induction kinetics
as wild-type cells. T hese results show that copper ad-
dition and copper chelation induce the cop operon by
different routes.
© 1999 Academic Press
Copper is an essential element through its functions
as a cofactor in many redox enzymes. But copper is also
very toxic to cells by damaging biomolecules through
radical formation. Intracellular copper concentrations
therefore need tobe regulated within narrow limits (1).
In the Gram-positive bacterium Enterococcus hirae,
copper homeostasis is accomplished by the cop operon,
which consists of the four genes copY, copZ, copA, and
copB (2). CopY encodes a copper responsive repressor
and copZ a copper chaperone that routes copper intra-
cellularly (3, 4). The two large genes copA and copB
encode CPx-type copper ATPases (5) and earlier data
suggested that CopA serves in copper uptake under
conditions of copper limitation, and CopB in its extru-
sion when intracellular copper reaches toxic levels (6,
7). Expression of the cop operon was shown to be reg-
ulated by ambient copper in a bi-phasic fashion: ex-
pression was minimal in standard growth media; ad-
dition as well as complexation of copper led to the
induction of the operon (6). There was only one pre-
dominant mRNA species apparent under different in-
duction conditions, arguing against the existence of
internal promoter sites of significant strength (D.
Stausak, unpublished).
Expression is apparently controlled by the CopY re-
pressor. By DNaseI footprinting, it had been shown
that CopY protected two regions of 27 and 28 bp fea-
turing an inverted repeat and flanking the transla-
tional start site of the cop operon. Two independent
CopY binding sites were corroborated by DNA band
shift assays, where it was apparent that the CopY-
DNA interaction occurred in two discrete steps by se-
quential binding of CopY to the two sites. The half-
association concentrations of CopY for these sites in
vitro were 5 nM for the operon-distal and 2 nM for the
proximal site (3). CopY requires one zinc(II) ion for the
binding to DNA (4). Copper(I) ions delivered to CopY
with the copper chaperone CopZ in vitro cause the
exchange of the bound zinc by two copper(I) and con-
comitant release of CopY from the promoter (4).
We here analyzed the induction of the cop operon in
vivo, using plasmid-borne wild-type and mutant cop pro-
moters and truncated CopB as a reporter gene. Mutation
of both, but not one, CopY binding site on the promoter
led tohyperinduction. Induction by excess copper or lack
of copper was affected differently, suggesting different
mechanisms ofinduction.An extracellular reductasethat
converts copper(II) to copper(I) was identified and its
influence on induction investigated.
MATERIALS AND METHODS
Materials.
tococcus faecalis or faecium) was obtained from the American Type
Culture Collection. E. coli strains XLmutS and XL1-Blue were ob-
tained from Stratagene Inc. Anti-CopB antibody and the ?copAB
Enterococcus hirae(ATCC9790, formerly called Strep-
1Towhom correspondenceshould beaddressed at Dept. of Clinical
Pharmacology, University of Berne, Murtenstrasse 35, 3010 Berne,
Switzerland. Fax: ?41 31 632 4997. E-mail: marc.solioz@ikp.
unibe.ch.
Biochemical and Biophysical Research Communications 259, 443–449 (1999)
Article ID bbrc.1999.0807, available online at http://www.idealibrary.com on
4430006-291X/99 $30.00
Copyright © 1999 by Academic Press
All rights of reproduction in any form reserved.

Page 2

strain
isothiocyanate-conjugated swine anti-rabbit IgG was obtained from
DAKO (Diagnostics AG, Switzerland). Bicinchoninic acid (BCA) was
bought from Pierce. All other molecular biology reagents and mate-
rials were obtained Roche, Sigma and Merck and were of the highest
grade available.
B1 havebeendescribed previously(8).Fluorescein-
Plasmid construction.
erated as follows: the Klenow-filled EcoRI-HindIII fragment of the
cop promoter region from pOA1 (9) was cloned into the E. coli/E.
hirae shuttle vector pC3 (10), also cut with EcoRI and HindIII and
Klenow-filled. The resultant plasmid pOC1 was cut with BamHI and
SalI and ligated with the 2844 bp BamHI/SalI fragment of the copB
gene to complete it in the resultant plasmid pOB1. From this plas-
mid, pOB2 was obtained by deletion of the 770 bp PstI–PstI frag-
ment, which truncated CopB by 80 amino acids at the C-terminus
(truncated CopB, CopB665). The PAreporter plasmid was generated
similarly from pOM214 (3) that carried the A-61T mutation. To
obtain the PAB (A-61T/A-30T) reporter plasmid pOB361, pOM214
was mutated with the Chameleon mutagenesis kit (Stratagene) us-
ing the mutagenic primers 5?-GTTAAGTTTTCAAATGTAATCG-3?
and the commercial switch-toggle MluI-ScaI primer. From the re-
sultant plasmid pOM30, the 1710 bp SacI/BamHI fragment was
subcloned into pOB2 cut with the same enzymes. Standard molecu-
lar biology methods were conducted by published procedures (11).
The PWT reporter plasmid pOB2 was gen-
Expression assays.
but not de-oxygenated tubes) to an OD546of 0.3–0.5 in N-media (1%
Na2HPO4 ? 2H2O, 1% trypticase peptone, 0.5% yeast extract, 1%
glucose). Following induction with the respective agents for 1h, cell
extracts were prepared by centrifuging the cultures and adding to
the cell pellets 50 ?l of 10 mg/ml lysozyme, 1 mM EDTA, 10 mM
Tris-Cl, pH 8. After 10 min incubation at room temperature, the cells
were taken through a freeze-thaw cycle, 10 ?l of 1 mg/ml DNaseI in
100 mM MgCl2were added, and incubation continued for 5 min at
room temperature. Amounts of these extracts corresponding to the
same number of cells were separated on sodium dodecyl sulfate gels
(12),followed byWestern blotting
phosphate-buffered saline, 10% horse serum, 0.1% Tween 20, for
quenching. Anti-CopB antibodies (8) were used for detection of CopB
and CopB665, and fluoresceinisothiocyanate-conjugated swine anti-
rabbit IgG was used as second antibody. The fluorescence was mea-
sured quantitatively with a Phosphor Imager (Molecular Dynamics).
Cells were grown semi-anaerobically (capped,
as described (13),using
Reductase assays.
determined spectrophotometrically at 37°C as described (14), but
using the following buffer: 200 mM N-morpholinoethane sulfonic
acid, 20 mM Na-citrate, 5 mM MgCl2, 500 ?M CuSO4, 500 ?M BCA,
pH 6. Reactions were started by the addition of 1% glucose. An
extinction coefficient of 4.6 ? 106M?1cm?1at 355 nm was used for the
copper(I)-BCA complex (Cu(I)BCA). Iron reductase was measured
similarly, using 500 ?M bathophenanthroline and 500 ?M FeCl3and
using an extinction coefficient for Fe(II)bathophenanthroline of
2.2 ? 104M?1cm?1at 520 nm (15).
Copper reductase activity in whole cells was
Reductase mutant selection.
washed with 0.1 M NaPi, pH 7, mutagenized for 90 min with 1.6%
ethyl methanesulfonate in the same buffer, followed by washing and
plating on N-media (9). Colonies were transferred to 0.45 ?m nitro-
cellulose membranes and, colonies up, incubated on filter paper
soaked with R-buffer, which was optimized for the plate assay (50
mM Na-citrate, 1 mM BCA, 1 mM CuSO4, pH 6.5) containing 5%
glucose. Within 40 min, colonies with reductase activity turned in-
tensely pink, while reductase deficient mutants stained only faintly.
E. hirae wild-type log cells were
RESULTS
Reporter gene assay for the analysis of the in vivo
regulation of the cop operon.
of the cop operon by copper and the effects of promoter
To study the regulation
mutations, we constructed a plasmid-borne expression
system with the following components: the cop pro-
moter containing the repressor binding sites, the copY
repressor gene, the copZ chaperone gene, and a re-
porter gene encoding a truncated form of CopB,
CopB665 (Fig. 1). Attempts tousethecommon reporter
genes chloramphenicol acetyltransferase or ?-galacto-
sidase in E. hirae had failed for unknown reasons.
CopB665 lacked the C-terminal 80 amino acids and
thus the two C-terminal transmembranous helices. In
wild-type E. hirae cells, CopB functions in copper ex-
trusion and thereby confers copper resistance, allowing
the cells to grow in up to 8 mM CuSO4; cells deleted in
CopB are growth-inhibited by more than 10 ?M CuSO4
(8). When CopB665 was expressed in the ?copAB de-
letion strain B1, the cells remained sensitive to above
10 ?M added copper, confirming that the truncated
CopB665 ATPase could not function in copper extru-
sion.
Induction of wild-type and mutant cop promoters by
copper.In Fig. 2, expression levels of CopB from the
genomic wild-type promoter were compared to expres-
sion levels of the CopB665 reporter protein from
plasmid-borne wild-type and mutant promoters in the
same cell. Both, the genomic promoter and the
plasmid-borne wild-type promoter (PWT) exhibited
closely similar induction kinetics by copper. The one
hour induction time in this and the following experi-
ment had been chosen based on previous experiments
that had shown that steady state levels of induction
were reached one hour after induction, for both, CopB
and CopB665 (16). Expression of genomic CopB was
essentially the same in plasmid bearing and plasmid-
free strains, indicating the absence of significant gene
F IG. 1.
used in this study. WT, genomic cop operon in wild-type E. hirae; B1,
genomic cop operon in ?copAB strain; pOB2, pOB215, pOB361,
plasmid-borne cop operons with CopB665 reporter gene and wild-
type or mutant promoters. The small open boxes represent the two
CopY repressor binding sites of the cop promoter (P). Transcription
is initiated at the arrowhead. The copY, Z, A, and B wild-type genes
areindicated by open boxes and thetruncated copB665 reporter gene
by gray boxes. The erythromycin gene of the cassette used to gener-
ate the ?copAB strain B1 is depicted as a solid bar. Arrows point to
the positions of promoter mutations used. PWT designates the
plasmid-borne wild-type promoter, PA the plasmid-borne A-61T
single mutant promoter, and PAB the A-30T/A-61T double mutant
promoter.
Schematic representation of the strains and plasmids
Vol. 259, No. 2, 1999BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
444

Page 3

dosage effects (16). The promoter mutation A-61T
(start of translation ? 1) had previously been shown to
impede the interaction of CopY with the operon-distal
DNA binding site in vitro. When CopB665 was placed
under the control of the A-61T mutant promoter on a
plasmid (PA), induction kinetics by copper were similar
to those of the wild-type promoter, except for a higher
induction level at 1 mM copper. In this and all further
experiments, the A-61T and the A-30T single muta-
tions exhibited essentially the same properties as the
wild-type promoter and are therefore not shown in
subsequent experiments. In contrast to the single mu-
tations, theplasmid-borneA-61T/A-30T doublemutant
promoter (PAB) was fully induced by only 10 ?M copper,
conditions that did not significantly induce transcrip-
tion from PWTor PA. The reproducible depression of the
induction from PAB at 100 to 300 ?M copper remains
unexplained at present.
Taken together, it appears that one mutated CopY
binding site on the promoter does not strongly disturb
the control of expression, except under very high cop-
per conditions. Mutation of both CopY binding sites in
PABled to hyperinduction by copper. This is in agree-
ment with previous in vitro studies on CopY-DNA in-
teraction by band shift assays (3). In these experi-
ments, the affinity of CopY for PA or PB was only
lowered two- to three-fold, while essentially no CopY
binding to PABcould be observed, suggesting coopera-
tivity in CopY binding tothe twosites on the promoter.
In vivo, the basal level of expression from PABis higher
than in the wild-type, but can still be strongly induced
by copper. Thus, in vivothereis still binding of CopY to
the double mutant promoter and this binding can be
abolished by copper, albeit more easily than binding to
wild-type promoter.
Figure 2C depicts the gradual induction of the
PAB promoter by low concentrations of copper(I) or
copper(II). In this as well as all other experiments
reported here, there was no significant difference be-
tween the effects of copper(I) or copper(II), suggesting
that both ions communicate with the cytoplasm. Also,
the same results were obtained whether copper(I) was
generated by reduction with 1 mM ascorbate or added
as a Cu(I)-acetonitrile complex in the presence of 2%
acetonitrile (not shown, 17).
In wild-typecells, theintracellular copper concentra-
tion is regulated by the copper homeostatic system and
theeffect of externally added copper on thecytoplasmic
concentration is unknown. We therefore employed a
?copAB mutant strain, B1, that is deficient in the
import as well as the export copper ATPase and that is
very sensitive to copper (8). Figure 3 shows that in B1,
induction of the PABdouble mutant promoter was less
responsive to induction by Cu?or Cu2?than in wild-
F IG. 2.
Western blot of wild-type E. hirae expressing CopB from the genomic wild-type promoter and CopB665 from a plasmid-borne wild-type
promoter (PWT), the A-61T mutant plasmid promoter (PA), and the A-61T/A-30T double mutant plasmid promoter (PAB). Cells were induced
for 1 h by the addition of 0 ?M, 10 ?M, 30 ?M, 100 ?M, 300 ?M and 1000 ?M CuSO4(lanes 1–6). (B) Quantification of the data shown in
A. F, CopB from the genomic wild-type promoter; E, CopB665 from plasmid PWT; Œ, CopB665 from plasmid PA; ?, CopB665 from plasmid PAB.
(C) Induction of CopB665 from PABin the low copper concentration range. E, copper(I); F, copper(II). Details of the methods are described
under Materials and Methods.
Copper induction of genomic CopB, and of CopB665 reporter protein from plasmid-borne wild-type and mutant promoters. (A)
Vol. 259, No. 2, 1999BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
445

Page 4

typecells. This suggests that copper import results in a
higher cytoplasmic copper concentration in wild-type
E. hirae compared to B1. No induction from wild-type
or singly mutated promoters could be observed in B1
due to the low copper tolerance of this strain (not
shown).
Induction of thecop promoter by copper(II) chelators.
It had been shown previously that the cop operon is
induced by the chelator o-phenanthroline and that this
induction was a consequence of copper rather than iron
limitation (8). Figure 4A shows that CopB expression
from the genomic wild-type cop promoter was half-
maximally induced at approximately 50 ?M o-phenan-
throline and that there were no significant trans or
gene dosage effects by the presence of reporter plas-
mids. As apparent from Fig. 4B, the o-phenanthroline
induction kinetics of CopB665 from plasmid-borne PWT
and PAB were essentially the same and thus differed
dramatically from induction by copper, which hyperin-
duced PAB. In the ?copAB strain B1, o-phenanthroline
induction of PWT and PAB also behaved similarly, but
half-maximal induction was shifted to 10 ?M. So anal-
ogous to induction by copper, induction by o-phenan-
throline was shifted tolower concentrations in B1 com-
pared to wild-type cells, which again must be ascribed
to the copper homeostatic control in the wild-type. Es-
sentially the same results were obtained using tetra-
thiomolybdate as a copper chelator (not shown). That
wild-type and double mutant promoters responded
similarly to induction by copper(II) chelators suggests
that induction by copper complexation does not work
via release of CopY from the promoter, which would be
expected to lead to hyperinduction of the double mu-
tant promoter. This agrees with in vitro band shift
experiments, where it had been shown that copper, but
not o-phenanthroline, released CopY from the DNA
binding site (3), D. Strausak, unpublished observation).
In contrast to o-phenanthroline and tetrathiomolyb-
date, which complex copper(II), the specific copper(I)
chelator bicinchoninic acid (BCA) did not induce ex-
pression from either wild-type or mutant cop promot-
ers (Fig. 5). The same was observed for the copper(I)
chelator bathocuproin disulfonate (not shown).
Identification and properties of an extracellular cop-
per reductase.The failure of the copper(I) chelators to
induce the cop operon was unexpected. It led to the
hypothesis that an extracellular copper reductase gen-
erates copper(I) in an occluded space, not accessible to
chelators, such as the periplasm. Figure 6 demon-
strates the presence of an extracellular copper reduc-
tase activity in E. hirae. Energy (glucose) dependent
reduction of copper(II) tocopper(I) by whole cells could
be detected in a photometric assay with BCA (18).
When E. hirae was grown in normal media, copper
reductase activity was 0.14 ? 0.03 nmol/min/108cells.
This activity was reduced to approximately 50% if the
cells were grown in the presence of 40 ?M copper(II) or
300 ?M o-phenanthrolineand thus did not appear tobe
F IG. 4.
(A) Induction of genomic CopB in wild-type cells (F) and in wild-type
cells containing alsoa plasmid-borne cop operon (E). (B) Induction of
CopB665 from plasmid PWT (Œ) and PAB (?) in wild-type cells and
from plasmid PWT(F) and PAB(E) in the ?copAB mutant B1. Details
of the procedure were as in Fig. 2.
o-Phenanthroline induction in wild-type and B1 mutant.
F IG. 3.
double mutant B1. E, Induction by copper (I); F, induction by cop-
per(II). Expression levels were determined as in Fig. 2.
Copper induction of CopB665 from PAB in the ?copAB
Vol. 259, No. 2, 1999BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
446

Page 5

strongly regulated. Copper reduction was very sensi-
tive to inhibition by mercury(II), with an I50of 1 ?M
(Fig. 6B). Silver(I) which has a similar ionic radius to
copper(I) inhibited with an I50 of 50 ?M. Nickel(II),
tin(II), cadmium(II) and platinum(I) at 100 ?M had no
significant inhibitory effects. Since platinum(II) is an
inhibitor of Fre1p, a yeast iron and copper reductase
(estimated I50for Pt(II) ? 50 ?M, ref. 14), its effect at
higher concentrations was also investigated. Pt(II) in-
hibited the E. hirae copper reductases activity by 50%
at the maximally attainable concentration of 1 mM,
indicating that the E. hiraeenzyme differs from Fre1p
in its biochemical properties. This E. hirae reductase
activity is, to our knowledge, the first bacterial copper
reductase to be described. An iron reductase activity
that amounted to 70% of the copper reductase activity
could also be detected. Since it was fully inhibited by
100 ?M platinum(II), it was clearly due to a different
catalytic activity (not shown).
Effect of a copper reductase mutant on the induction
of thecop operon. In light of the presence of an extra-
cellular copper reductase activity, it became of interest
to look at the influence of copper reduction on the
regulation of the cop operon. With a colorimetric plate
assay, several mutants with reduced extracellular cop-
per reductase activity were isolated from mutagenized
cells. Mutant C4 and several other mutants that were
identified had a copper reductase activity that was
reduced by 70%. However, no mutants completely de-
ficient in copper reductase could be isolated. In mutant
C4, induction of wild-type and mutant cop promoters
by either copper(I) or copper(II) did not differ signifi-
cantly from that observed in wild-type cells (not
shown).
Taken together, our data show that both, copper(I)
and copper(II) were equally effective in inducing the
cop operon. One intact CopY binding site was sufficient
for effective repression. Mutation of both CopY binding
sites facilitated induction by copper and did so to the
same extent for copper(I) and copper(II). The induction
kinetics by copper(I) or copper(II) were not influenced
by reduced extracellular copper reductase activity. In
contrast to induction by copper, induction by copper
chelators was not influenced by promoter mutations
that impair the interaction of CopY with the promoter,
indicating that this induction occurs by a different
mechanism.
DISCUSSION
The bi-phasic induction of the cop operon of E. hirae
by excess as well as limiting copper reflects its dual role
in copper uptake and copper secretion (9). This co-
regulation of the import and the export copper ATPase
may be a safety mechanism: a sudden change from
copper limiting to copper excess conditions could poi-
son the cell if the import pump is induced, but no
export route is available. The mechanism of this regu-
lation is not fully understood. We thus studied, in vivo,
aspects of wild-type and mutant cop promoter regula-
F IG. 6.
The formation of a copper(I) bicinchoninic acid complex catalyzed by
whole cells was followed spectrophotometrically as described under
Experimental Procedures. Where indicated, 1% glucose or 2 ?M
HgCl2wereadded. (B) Inhibition of copper reductaseactivity by Hg2?
(F), Ag?(E), and Pt2?(Œ).
Copper reductase activity of E. hirae wild-type cells. (A)
F IG. 5.
CopB665 from plasmid-borne PWT (F) and PAB(E) in wild-type cells
by the copper(I) chelator BCA. Details of the procedure were as in
Fig. 2.
Induction of CopB665 by copper chelation. Induction of
Vol. 259, No. 2, 1999BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
447