Download MicroReview Metal ion homeostasis and intracellular

Molecular Microbiology (1998) 28(3), 403–412
MicroReview
Metal ion homeostasis and intracellular parasitism
Daniel D. Agranoff and Sanjeev Krishna*
Division of Infectious Diseases, Department of Cellular
and Molecular Sciences, St George’s Hospital Medical
School, Cranmer Terrace, London SW17 0RE, UK.
Summary
Bacteria possess multiple mechanisms for the transport of metal ions. While many of these systems may
have evolved in the first instance to resist the detrimental effects of toxic environmental heavy metals,
they have since become adapted to a variety of important homeostatic functions. The ‘P’-type ATPases play
a key role in metal ion transport in bacteria. A Cu þATPase from the intracellular bacterium Listeria monocytogenes is implicated in pathogenesis, and similar
pumps in Mycobacterium tuberculosis and M . leprae
may play a comparable role. Intracellular bacteria
require transition metal cations for the synthesis of
superoxide dismutases and catalases, which constitute an important line of defence against macrophage-killing mechanisms. The macrophage protein
Nramp1 , which confers resistance to a variety of intracellular pathogens, has also been shown recently to
be a divalent amphoteric cation transporter. Mycobacterial homologues have recently been identified
by genomic analysis. These findings suggest a model
in which competition for divalent cations plays a pivotal role in the interaction between host and parasite.
Introduction
The dynamic relationship between pathogenic bacteria and
their hosts is an area of profound medical and scientific
interest. The long-recognized phenomenon of ‘nutritional
immunity’, in which sequestration of iron and possibly
other metals occurs as a non-specific host response to
infection (Beisel, 1977), hints in general terms at the possibility of keen competition between host and parasite for
essential metal ions. However, much of the extensive literature on metal ion homeostasis in bacteria has addressed
Received 22 November, 1997; revised 20 January, 1998; accepted
23 January, 1998. *For correspondence. E-mail s.krishna@sghms.
ac.uk; Tel. (0181) 725 5836; Fax (0181) 725 3487.
Q 1998 Blackwell Science Ltd
itself to mechanisms of environmental heavy metal resistance (Silver and Phung, 1996), and it is not clear how, if
at all, this relates to pathogenicity.
The antiquity of major metal transporter families is evident from their high degree of sequence, and presumably
functional, conservation between bacteria and complex
eukaryotes. Bacterial resistances to a wide variety of
toxic heavy metal ions, including Agþ, AsO2– , AsO43– ,
Cd 2þ, Co 2þ, CrO42– , Cu 2þ, Hg2þ, Ni 2þ, Pb 2þ, Sb 3þ,
TeO32– , Tl þ and Zn 2þ, are well recognized, and their
genetic determinants are frequently plasmid borne (Silver
and Phung, 1996). Four broad strategies for minimizing
toxicity from metal ions are discernible in bacteria. These
involve: (i) the ATP-driven efflux pumps comprising the
‘P’-type ATPases and ATP-binding cassette (ABC) superfamilies (examples are the copB Cu þ efflux and the cadA
Cd 2þ /Zn 2þ efflux ‘P’-type ATPases of Enterococcus hirae
and Staphylococcus aureus , respectively, and the ABCbased arsenic efflux pumps in a number of Gram-positive
and Gram-negative bacteria); (ii) chemiosmotic systems
(metal ion/proton antiporters), in which the energy for
the secondarily active transport of the metal ion is provided
by proton gradients, e.g. the Cd 2þ, Zn 2þ, Co2þ (Czc) threecomponent systems found in the soil bacteria Alcaligenes
spp.; (iii) redox systems, of which the merA mercuric reductase of Pseudomonas aeruginosa is the prototype; and (iv)
metallothioneins – small metal-binding proteins previously
believed to be confined to animal cells but recently identified in the cyanobacterium Synechococcus (Silver and
Phung, 1996).
Over the course of evolution, these systems have been
selected to perform physiological functions extending
beyond those in bacteria. Thus, for example, the proteins
associated with human copper metabolism, whose genes
are mutated in the copper transport disorders Menkes
and Wilson’s diseases, are copper-transporting ‘P’-type
ATPases that bear a closer resemblance to bacterial
heavy metal ‘P’-type ATPases than they do to other mammalian ATPases (Bull and Cox, 1994).
In recent years, it has become apparent that the regulation of metal ions plays a key role in a much broader sphere
of bacterial homeostasis. These areas include the generation of electrochemical ionic gradients across membranes,
which, in turn, provide the driving force for the secondarily
active transport of metabolic precursors; the regulation of
404 D. D. Agranoff and S. Krishna
cell volume, which depends on the control of intracellular
K þ in response to changes in the osmolality of the external
milieu; and the regulation of trace metal ions such as
copper, manganese, zinc and iron, which are essential
components of metalloenzymes necessary for the survival of both prokaryotic and eukaryotic cells. Copper, for
example, is an essential cofactor in a variety of redox
enzyme systems, including lysyl oxidase, cytochrome c
oxidase and several bacterial superoxide dismutases
(SODs), as are iron, zinc and manganese (Sehn and
Meier, 1994). Iron is located at the active centre of catalase,
an enzyme important for the disposal of toxic peroxide
anions. These redox systems protect the organism against
reactive oxygen intermediates, which may be generated
either endogenously during aerobic respiration or exogenously by neutrophils or macrophages, the preferred host
cell targets of a number of medically important human
intracellular pathogens, such as Mycobacterium tuberculosis and M . leprae , Legionella pneumophila, Listeria
monocytogenes , Mycoplasma spp. and Chlamydia spp.
Exposure to such metal ions, however, is a double-edged
sword, for it is precisely those redox properties harnessed
in the context of SODs that, in other circumstances, render
them highly toxic through interference with the functioning
of intracellular macromolecules and the generation of toxic
free radicals. Thus, mechanisms for metal ion homeostasis or, more specifically, metal ion transport may constitute
major adaptations to intracellular survival and replication
among pathogenic bacteria.
These concepts are reinforced by new findings (Gunshin
et al., 1997) which indicate that Nramp1 (natural resistanceassociated macrophage protein), which is associated with
resistance to a variety of unrelated bacterial and protozoan intracellular pathogens (Vidal et al ., 1995), is probably itself a heavy amphoteric metal ion transporter. The
‘P’-type ATPases, a ubiquitous and highly conserved
superfamily of cation transporters, are also prime candidates for a central role in intracellular survival in addition
to their housekeeping functions. Indeed, recent evidence
for the role of a putative copper-importing ‘P’-type ATPase
as a virulence factor in L . monocytogenes (Francis and
Thomas, 1997a,b) points to their potential importance for
intracellular pathogens. Knockout mutants exhibited no
phenotypic impairment when grown in competition with
wild types in liquid media or in J774 or HeLa cell monolayers but were cleared much more rapidly from the tissues
of infected mice. Thus, it now becomes possible to bridge
the conceptual gap between the hitherto ill-defined importance of metal ions in host resistance and bacterial pathogenicity.
‘P’-type ATPases
The ‘P’-type ATPases are fundamental components of
the cell’s homeostatic apparatus. They constitute a large,
ubiquitous and diverse superfamily of integral membrane
proteins, containing 8–10 hydrophobic membrane-spanning helices (Moller et al., 1996). Their core structure is
conserved throughout the three major domains of living
organisms: the Archaea, Prokaryota and Eukaryota. Most
are involved in the transport of cations across cellular and
subcellular membrane systems. Identified substrate specificities encompass Naþ, K þ, Hþ, Ca 2þ, Mg 2þ, Cu þ, Cd 2þ
and Zn 2þ as well as aminophospholipids (Tang et al .,
1996). Their common biochemical characteristics include
the utilization of energy derived from hydrolysis of the terminal pyrophosphate bond of ATP for the vectorial transport
of cations across the cell membrane and a dual-energy
state catalytic cycle involving a phosphorylated acyl-intermediate (hence ‘P’ type). The basic mechanism of ion
translocation has been worked out for the Naþ /K þ and sarcoplasmic reticulum Ca 2þ ATPases and probably applies
in general terms to all ‘P’-type ATPases. ATP-dependent
phosphorylation at a well-defined site on the cytoplasmic
side of the protein is coupled with binding and occlusion
of one or more intracellular cations. This generates a highenergy intermediate and induces conformational changes
causing translocation of the cation across the membrane.
Binding of an extracellular counterion to the now lowenergy-state protein is associated with dephosphorylation
and transport of this ion into the cell. The orientation of the
ATPase in the membrane is, therefore, defined by the fact
that ATP is only likely to be available intracellularly. Consequently, as ATP-dependent phosphorylation appears to
be a common feature of ‘P’-type ATPases, those pumps
whose function appears to be the uptake of a physiologically important ion (see below) must presumably export
another intracellular cation. In many cases, this secondary
ion has yet to be identified.
Once thought to be poorly represented in bacteria, recent
genome-sequencing initiatives have to date identified more
than 50 bacterial ATPase sequences, including representatives of all the major subgroups. Although highly divergent
in many regions, they display several extraordinarily wellconserved amino acid motifs at the phosphorylation and
phosphatase sites and at the ATP-binding and cation
transduction domains. These sequences can be regarded
as signature motifs essential for classification (Fig. 1).
‘P’-type ATPases are subclassified on the basis of cation
specificity and membrane topology as proposed by Lutsenko and Kaplan (1995). The subgroups encompass the
putative heavy metal transporters such as E . hirae copA
(designated type I ATPases), a second subgroup, designated type II ATPases [the Naþ /K þ and Hþ /K þ ATPases
of animals, the sarco(endo)plasmic reticulum calcium
ATPases (SERCA), plasma membrane calcium ATPases
(PMCA), the proton ATPases of fungi and lower plants
(HA) and aminophospholipid translocases], and the
Q 1998 Blackwell Science Ltd, Molecular Microbiology, 28, 403–412
Bacterial ion transport 405
Fig. 1. Summary of alternative
nomenclatures for the ‘P’-type ATPases with
schematic representations of the three
subclasses of bacterial ATPases showing
their proposed membrane topologies,
configuration of putative membrane-spanning
helices and conserved motifs. Note that the
type I ATPases possess an additional pair of
transmembrane helices in their N-terminal
region, while the type II proteins have a much
extended C-terminus, accommodating 4–6
transmembrane helices. The several
designations for each subclass have been
proposed by various authors on the basis of
different criteria, i.e. membrane topology,
function, motifs etc., but are essentially
overlapping. The motifs conserved among all
subclasses are the ‘TGE’ or phosphatase site,
the proline residue in the M4 transmembrane
helix flanked by subclass-specific residues
(Cpx, PEGL, PTTI) believed to form part of
the ion channel, the phosphorylation site motif
DKTGTXT, and the hinge region motif,
GDGXND, forming part of the ATP-binding
domain. GMXCXXC is the heavy metal
binding motif occurring in many but not all
type I ATPases. M1–M6, transmembrane
helices. SERCA, sarco(endo)plasmic
reticulum calcium ATPase; PMCA, plasma
membrane calcium ATPase; HA, proton
ATPase of fungi and lower plants.
high-affinity K þ uptake ATPase exemplified by the kdpB
protein found only in bacteria (type III ATPases).
Among bacterial sequences, cation specificities for Cd 2þ,
Zn 2þ, Cu þ/2þ, Agþ and Mg 2þ have been demonstrated,
while presumptive Hþ and Ca 2þ transporters have been
recognized on the basis of sequence homology. Many
remain functionally uncharacterized. The majority of bacterial sequences are monosubunit members of the type I
or heavy metal subclass and are probably involved in tolerance to heavy metal ions or in the scavenging of essential trace metals. Whether this apparent predominance of
type I pumps is a true reflection of the distribution of ‘P’type ATPases in bacteria or whether it is an artifact of
selective approaches to cloning will be clarified by the
systematic sequencing of bacterial genomes currently in
progress.
Functional studies on bacterial ‘P’-type ATPases
In this section, we review those ‘P’-type ATPases that have
been investigated experimentally. Table 1 summarizes
the properties of all ‘P’-type ATPases identified in the 10
prokaryotic genomes sequenced to date, as well as those
of other functionally characterized ATPases discussed
below.
The E. coli KdpB ATPase
The inducible K þ uptake system of E . coli has been well
Q 1998 Blackwell Science Ltd, Molecular Microbiology, 28, 403–412
characterized (Altendorf et al ., 1992) and consists of a
complex of three proteins, kdpABC , which are encoded
as a single operon. kdpA contains a putative periplasmic
K þ binding site, while kdpC may be analogous to the beta
subunit of the mammalian heterodimeric Naþ /K þ ATPase.
The catalytic subunit, kdpB , is a ‘P’-type ATPase. K þ transport by kdpB is electrogenic (Fendler et al ., 1996). Homologues in numerous other Gram-negative bacteria, one
Gram-positive bacterium, Bacillus acidocaldarius (Treuner-Lange et al ., 1997), and in M . tuberculosis (accession
no. Z92539) have been identified. kdpB differs sufficiently
from the type I and type II ‘P’-type ATPases by virtue of its
membrane topology (see Fig. 1) to merit its inclusion in a
subclass of its own (the type III ‘P’-type ATPases).
The kdpABC system is a high-affinity system (reflected
by a K m for K þ of 2 mM) for the uptake of K þ at low external
K þ concentrations, in contrast to the constitutively active
TrK and Kup systems with only a modest affinity for K þ
(K m 1.5 mM). In this respect, it resembles several of the
other bacterial ‘P’-type ATPase-based ion transport
mechanisms in providing a recruitable and tightly regulated system for reinforcing existing constitutive transporters under extreme or rapidly changing environmental
conditions. It is likely that the kdp system plays a role in
the regulation of cell volume, as the transcriptional regulation of the kdp operon is stimulated by low, and inhibited by
high, turgor pressures (Epstein, 1992). This regulatory
system appears to be mediated by two proteins, kdpD
and KdpE , which are encoded as an adjacent operon
copA
hpcopA
copA
cadA
pacL
mgtA
cop1
cop2
cadA1
cadA2
pacL1
pacL2
pacL3
pma1
kdpB
Haemophilus influenzae
Helicobacter pylori
Mycoplasma genitalium
Mycoplasma pneumoniae
Synechocystis sp.
ykvW
yvgW
yvgX
yloB
Bacillus subtilis
ATZN
HRA-1
HRA-2
ATCU
mgtA
kdpB
ORF
Methanococcus jannaschii
Escherichia coli
pacS
copB
ATPase
name
Archaeoglobus fulgidus
A: Completed genomes
Organism
D90904
D90915
D64005
D64005
D64005
D90911
D90905
D90910
D90910
MG071
P47317
HP1072
HP1503
L46864
HI0290
U00039
U16658
U16659
U58330
U14003
K02670
BG13325
BG14105
BG14106
BG13384
U67563
AF0473
AF0152
Acc. no.
Table 1. Properties of bacterial ‘P’-type ATPases.
745
780
642
721
945
953
972
905
690
872
874
746
789
687
723
733
732
722
835
899
683
637
702
803
890
805
805
691
No. amino
acids
79.9
82.6
68.7
76.7
102.5
103.6
106.1
97.6
73
94.9
96.3
81.8
86.5
74.9
78.1
76.8
78.4
78.5
87.8
99.5
72.1
68.6
75.2
85.8
97.1
89.6
86.4
75.3
MW (kDa)
I
I
I
I
II
II
II
II
III
II
II
I
I
I
I
I
I
I
I
II
III
I
I
I
II
II
I
I
Class
GMTCQSC
?Cu þ/ 2þ
?Cu
?Cu 2þ
?Cd 2þ
?Cd 2þ
?Ca 2þ
?Ca 2þ
?Ca 2þ
?H þ
?K þ
2þ
?Mg
GMRCAAC
GMKCAGC
–
GMDCTSC
–
–
–
–
–
–
–
?Ca 2þ
2þ
GMTCTAC
CSHC
NLDCPDC
Cu
efflux?
?Cuþ /2þ
?Cd 2þ efflux
þ /2þ
GMDCAAC
(MXHX/C/GX)
Histidine-rich N-terminus
GLSCGHC, GMSCASC
–
–
?Zn 2þ
Unknown heavy metal
Unknown heavy metal
?Cu 2þ
?Mg 2þ
K þ uptake
–
GLDCSNC
(GMTCAAC) × 2
–
–
?H þ
Unknown heavy metal
Unknown heavy metal
Unknown heavy metal
?Ca 2þ
GMTCAMC
Histidine-rich N-terminus
Heavy metal binding
motif
Unknown heavy metal
Cu 2þ
Transport activity
CPC
CPC
SPC
CPC
PEGL
PEGL
PEGL
PEGL
PTTI
PEGL
PEGL
CPC
CPC
CPC
CPC
CPC
CPH
CPH
CPC
PEML
PTTI
SPC
CPC
CPC
PEGL
PAVL
CPC
CPH
M4 motif
SEHPLA
TRHPLA
SEHPIG
STHPIA
–
–
–
–
–
–
–
SEHVIA
LAHSAI
STHPIA
ANHPIA
ATHPLA
SSHPIA
STHPLA
SSHPLA
–
–
SSHPLA
SQHPIA
SEHPLG
–
–
SEHPIA
SEHPIA
HP motif
et
et
et
et
et
et
et
et
et
et
al .
al .
al .
al .
al .
al .
(1997)
(1997)
(1997)
(1997)
(1997)
(1997)
al. (1997)
al . (1997)
al . (1997)
al . (1997)
Kaneko
Kaneko
Kaneko
Kaneko
Kaneko
Kaneko
Kaneko
Kaneko
Kaneko
et
et
et
et
et
et
et
et
et
al .
al .
al .
al .
al .
al .
al .
al .
al .
(1996)
(1996)
(1996)
(1996)
(1996)
(1996)
(1996)
(1996)
(1996)
Himmelreich et al . (1996)
Fraser et al . (1995)
Tomb et al. (1997)
Tomb et al. (1997)
Tomb et al. (1997)
Fleischmann et al. (1995)
Blattner
Blattner
Blattner
Blattner
Blattner
Blattner
Kunst
Kunst
Kunst
Kunst
Bult et al . (1996)
Klenk et al . (1997)
Klenk et al. 1997)
References
406 D. D. Agranoff and S. Krishna
Q 1998 Blackwell Science Ltd, Molecular Microbiology, 28, 403–412
Q 1998 Blackwell Science Ltd, Molecular Microbiology, 28, 403–412
ctpA
ctpB
cadA
copA
copB
cadA
ctpA
fixI
mgtA
mgtB
cadA(p)
cadA(c)
pacS
pacL
ctaA
ctpA
ctpB
yhho
iod7
y71
2097 *
830
y39
y21
kdpB
781
751
723
727
745
651
654
757
903
909
727
805
748
927
791
856
797
709
Z74025
Z95210
Z92539
Z46257
Z46257
M90750
L13292
L13292
U15554
U15554
M36068
U07843
M57715
J04551
P4738
D16437
D16436
U04356
762
752
772
789
719
657
Z74410
Y07638
Z74025
Z79700
Z92771
AL021184
82.3
78.2
78.2
78.4
81.5
69.2
69.5
79.6
99.8
100.5
78.8
86.9
79.7
99.7
83.7
89.9
85
74.6
79.02
77.5
79.5
80.1
76.5
67.9
I
I
I
I
I
I
I
I
II
II
I
I
I
II
I
I
I
I
I
I
I
I
II
II
III
heavy
heavy
heavy
heavy
heavy
heavy
heavy
metal
metal
metal
metal
metal
metal
metal
Unknown heavy metal
Unknown heavy metal
Cd 2þ efflux?
Cu þ /Ag þ uptake
Cu þ /Ag þ efflux
Cd 2þ efflux
Cu 2þ
Unknown heavy metal
Mg 2þ
Mg 2þ
Cd 2þ /Zn 2þ efflux
Cd 2þ efflux
Cu þ / 2þ efflux?
?Mg 2þ
Cu þ / 2þ uptake?
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
?Ca 2þ
?H þ
?K þ
GMSCSCC
GMLCAAC
GFTCANC
GMTCANC
(MXHXXMSGMXHSH) × 3
GFTCANC
–
NAYCGTC?
–
–
–
(GFSCANC) × 2
GMGCAAC
–
GMKCAGK
–
–
–
GMSCSAC
GMSCAAC
–
–
–
CPC
CPC
CPC
CPC
CPH
CPC
CPC
CPC
PEML
PEML
CPC
CPC
CPC
PEGL
CPC
PEGL
PEGL
PTTI
CPC
CPC
APC
CPC
CPC
SPC
SEHAVA
SEHSVA
SQHPLA
SEHPLG
ANHPLA
SQHPLA
SEHPLA
SRHPIA
–
–
SQHPLA
SQHPLA
SEHPLA
–
SRHPLA
SEHSVA
SEHAMA
SEHPLA
SEHPIG
SRHPLA
SEHPLG
SEHPLG
–
–
–
Fsihi and Cole (1995)
Fsihi and Cole (1995)
Ivey et al . (1992)
Odermatt et al. (1993)
Odermatt et al . (1993)
Lebrun et al . (1994)
Francis and Thomas (1997)
Khan et al . (1989)
Snavely et al . (1991)
Snavely et al . (1991)
Nucifora et al . (1989)
Witte et al . (1986)
Kanamaru et al. (1993)
Kanamaru et al . (1993)
Phung et al . (1994)
–
–
–
–
–
–
–
–
–
–
Section A lists, by organism, all ‘P’-type ATPases identified in the 10 prokaryotic genomes completely sequenced at the time of writing, permitting comparison of ATPase number and diversity
between genomes. Sequences were identified initially through the relevant genome sequencing project’s own open reading frame annotations and checked for completeness through our own
similarity searches using the TBLASTN algorithm, which searches translations of all six reading frames against a query protein sequence. Section B summarizes additional sequences discussed
in the text.
Acc. no: EMBL/GenBank Accession number. No. amino acids, MW, calculated number of amino acids and molecular weight. Class refers to the classificatory scheme based on that of Lutsenko
and Kaplan (1995) discussed in the text. The transport activity quoted for most of the proteins derived from the completed genome data is speculative and based solely on sequence similarity to
proteins in which the cation specificity has been experimentally verified. The ‘heavy metal binding motif’, GMXCXXC (where X ¼ any amino acid residue), is characteristic of the N-terminal region of
most type I ‘P’-type ATPases, although in several type I proteins this is replaced by a histidine-rich repeating motif, while in others no obvious heavy-metal binding region has so far been recognized. The M4 motif refers to the invariant proline residue (and subclass-defined adjacent residues) found in the transmembrane helix (M4) preceding the phosphorylation site motif, DKTGTXT. The
‘HP’ motif is another characteristic type I ATPase motif, located 34–43 residues downstream of the phosphorylated aspartyl residue.
P, plasmid encoded; C, chromosomally encoded. *This sequence had only just appeared in the public domain at the time of writing and had not yet appeared on the EMBL/GenBank databases.
Since then a further sequence (not shown) has been identified.
B: Additional sequences
Mycobacterium leprae
Mycobacterium leprae
Bacillus firmus
Enterococcus hirae
Enterococcus hirae
Listeria monocytogenes
Listeria monocytogenes
Rhizobium melliloti
Salmonella typhimurium
Salmonella typhimurium
Staphylococcus aureus
Staphylococcus aureus
Synechococcus sp. 7942
Synechococcus sp. 7943
Synechococcus sp. 7942
Mycobacterium tuberculosis
Bacterial ion transport 407
408 D. D. Agranoff and S. Krishna
and whose sequences place them in the sensor–effector
family of bacterial regulators.
Copper-transporting ‘P’-type ATPases
There is genetic and biochemical evidence for coppertransporting ‘P’-type ATPases in Enterococcus hirae ,
Helicobacter pylori , Listeria monocytogenes and Synechococcus spp. In E . hirae , copA and copB form part of
a single inducible operon that also incorporates two transcriptional regulators, copY and copZ , located upstream
of the ATPase sequences (Odermatt et al ., 1993). copA
exhibits 35% sequence identity to copB and, remarkably,
43% identity with the human Menkes disease gene, underscoring the probable role of both encoded proteins in
copper transport. The N-termini copA and copB contain the heavy metal binding motifs GMXCXXC and
MXHXXMSGXHS respectively.
copA mediates Cu þ uptake, as demonstrated by
increased Cu þ dependence in copA knockout mutants,
whereas copB mediates Cu þ extrusion as demonstrated
by enhanced sensitivity to the toxic effects of Cu þ in
copB and copAB knockouts, increased Cu þ uptake by
copB and copAB knockouts compared with wild-type
cells and copA knockouts and the need for a functioning
copB gene for the extrusion of 110Agþ (which substitutes
for Cu þ ) from previously loaded cells (Odermatt et al .,
1993). Cu þ accumulation by everted membrane vesicles
engineered to overexpress copB has been shown to be
dependent on ATP hydrolysis and to be inhibitable by
vanadate ions, confirming that copB is functionally a ‘P’type ATPase (Solioz and Odermatt, 1995).
Knockout studies also show that copY is a transcriptional repressor and copZ a transcriptional activator of
the copAB operon. Interestingly, both copY and copZ contain heavy metal binding motifs, reflecting their role in
mediating the effect of ambient copper ions on gene
expression.
Helicobacter pylori and Listeria monocytogenes
Cu-ATPases
The complete genomic sequence of H . pylori reveals two
putative Cu-ATPases, hpcopA and copA (see Table 1).
Only hpcopA has been functionally characterized (Ge
and Taylor, 1996) and forms part of a single operon with
a short, downstream-encoded polypeptide, hpCopP , containing a heavy metal binding motif. Knockout mutants of
hpcopA display reduced environmental copper tolerance,
supporting a role for copA in copper extrusion.
The ctpA gene of the intracellular pathogen L . monocytogenes also appears to be involved in copper homeostasis (Francis and Thomas, 1997a). Unusually, its product is
shorter than other bacterial type I ‘P’-type ATPases, with
a truncated N-terminus lacking a heavy metal-binding
motif. This is reminiscent of the initial report concerning
the H . pylori hpcopA gene, which was also thought to
lack a portion of the N-terminal region (Ge et al ., 1995).
However, no upstream sequence containing the heavy
metal motif has so far been identified. Significantly, knockout mutants were cleared more rapidly from the livers of
infected mice than wild-type organisms and were rapidly
out-competed by wild types in mixed infections (Francis
and Thomas, 1997b). This is the first direct evidence for
the role of a ‘P’-type ATPase in virulence.
Cadmium-transporting ‘P’-type ATPases
The Cd 2þ efflux ATPases represent one of several bacterial mechanisms for Cd 2þ resistance. They occur widely in
Gram-positive bacteria including Staphylococcus aureus ,
L . monocytogenes , H . pylori and Bacillus subtilis (the
last two recognizable on the basis of homology only). S .
aureus possesses a plasmid-borne cadA gene expressed
as part of a single operon with a short upstream-encoded
open reading frame, cadC , which may represent a metal
ion-responsive transcriptional repressor, and a chromosomally encoded ‘P’-type ATPase, which lacks a cadC homologue (Silver et al ., 1993). cadA transports and confers
resistance to both Cd 2þ and Zn 2þ, while the chromosomal
gene product transports only Cd 2þ. There is direct biochemical evidence for ATP-driven cadmium transport by
cadA in everted membrane vesicles of B . subtilis expressing cadA (Tsai et al ., 1992). The cadC,A operon is capable of conferring high levels of Cd 2þ resistance on B .
subtilis (a cadmium-sensitive organism more readily manipulated than S . aureus ). cadC,A expression is inducible by
Cd 2þ, Zn 2þ, Pb 2þ and Bi 3þ. Cd 2þ is its most powerful
inducer.
The amino acid sequences of the chromosomally
encoded B . firmus cadA and the plasmid-encoded L .
monocytogenes cadA are highly similar to those of S .
aureus , and both are associated with cadC homologues.
In L . monocytogenes , experimental evidence supports
the role of cadA in Cd 2þ resistance as well as its induction
by Cd 2þ, although it differs from its S . aureus homologue
in not transporting Zn 2þ (Lebrun et al ., 1994).
Magnesium- and calcium-transporting ‘P’-type ATPases
Mg 2þ is the most abundant divalent cation in biological
systems. It exhibits the largest hydrated ionic radius of
all biological cations, while being the most charge dense
in its unhydrated state. In this respect, in differs from the
otherwise chemically closely related Ca 2þ.
Salmonella typhimurium possesses two Mg 2þ-transporting ‘P’-type ATPases, mgtA and mgtB , both of which
mediate Mg 2þ uptake, confer tolerance to Mg 2þ-limiting
Q 1998 Blackwell Science Ltd, Molecular Microbiology, 28, 403–412
Bacterial ion transport 409
environments and are tightly regulated by the external
Mg 2þ concentration (Roof, 1994). Both bear a far closer
resemblance to the eukaryotic SERCA and PMCA
ATPases than to the bacterial heavy metal pumps and
can clearly be classified as type II ‘P’-type ATPases
(Fig. 1). They exhibit homologies to the eukaryotic
SERCA ATPases in regions surrounding at least five of
the six amino acid residues suggested by Clarke et al .
(1989) as being critical for calcium binding and ion transduction in the rabbit SERCA pump. Differences with
respect to certain key residues may reflect the different
physicochemical properties of Mg 2þ and Ca 2þ.
Although both appear to perform the same function,
they differ significantly in terms of amino acid sequence
(only 50% identity), kinetic parameters and regulation
(both are inducible but to greatly differing extents). mgtA
displays 91% identity with a putative Mg 2þ transporter in
E . coli (mgtA ) (Burland et al ., 1995), suggesting that the
two S . typhimurium genes are not the result of a recent
gene duplication. Moreover, the mgtB locus occurs within
a bicistronic operon, mgtCB , incorporating a smaller gene,
mgtC , which encodes a protein essential for the full
expression of the Mg 2þ-scavenging phenotype (Snavely
et al ., 1991). A further feature of these transporters is the
fact that, unlike all other biochemically characterized ‘P’type ATPases, the direction of cation transport is down
rather than against the electrochemical gradient. It is
unclear why an energy-dependent system should be
required for this process. An important clue as to their
function is provided by the observation that mgtB transcription is greatly increased during phagocytosis of a
pathogenic S . typhimurium strain into the vacuole of an
epithelial cell line (Portillo et al ., 1992). More recently, it
has been shown that the mgtCB operon forms part of a
pathogenicity island in S . typhimurium and that mgtC
expression is essential both for intramacrophage survival
and for virulence in a mouse model (Blanc-Potard and
Groisman, 1997). The intramacrophage survival defect
exhibited by mgtC knockout mutants can be rescued by
exposure to Mg 2þ, indicating a central role for this ion in
the interaction between macrophage and parasite in this
instance.
Six putative bacterial Ca 2þ ATPases have been identified, although this is an inference so far based on sequence homologies alone (see Table 1).
Mycobacterial ‘P’-type ATPases
The pathogenicity of the important human bacterial pathogens Mycobacterium tuberculosis and Mycobacterium
leprae is at least in part related to their ability to grow
and multiply within host macrophages. The intraphagosomal microcellular environment differs markedly from
the extracellular milieu in terms of osmolality, metal ion
Q 1998 Blackwell Science Ltd, Molecular Microbiology, 28, 403–412
concentrations, pH and nutrient availability, demanding
the modulation of the parasite’s homeostatic responses.
Thirteen mycobacterial ‘P’-type ATPases have appeared
on the databases: 11 in M . tuberculosis and two in M .
leprae. Assignation of function is so far based on sequence
homology alone, although we are attempting the functional
characterization of some of these proteins. The apparently
restricted number of ‘P’-type ATPases in M . leprae may
be a reflection of the proteomic streamlining characteristic
of obligate parasites, in which functional redundancy is
kept to a minimum. This characteristic is seen throughout
the M . leprae genome in which many members of paralogous gene families have become pseudogenes.
Both of the M . leprae ATPases (ctpA and ctpB ) and 7 of
the 11 M . tuberculosis proteins (ctpA, ctpB , ioD7 , yhho ,
y71 , 2097 and 830) are putative ‘heavy metal’ pumps.
Sequence alignments place ctpA , ctpB and ioD7 in the
Cu 2þ-transporting group, while yhho has some affinities
with the Cd 2þ transporters. However, these inferences
are highly tentative in the absence of functional data. M .
tuberculosis also possesses a putative Ca2þ pump ( y39 ),
a putative proton pump ( y21 ) and a kdpB ATPase (see
Table 1).
Nramp and Mramp – a new class of metal ion
transporter
The energy expended by one or more ‘P’-type ATPases in
maintaining transmembrane electrochemical gradients is
also likely to contribute to the functioning of secondarily
active transporters of metal ions. The Nramp (natural
resistance-associated macrophage protein) family are
novel members of the latter class of transporter and constitute a group of homologous hydrophobic membrane
proteins associated with resistance to infection in mice
by such phylogenetically and immunologically distinct
intracellular pathogens as Mycobacterium bovis (BCG),
Salmonella typhimurium and Leishmania donovani . The
mechanism is independent of immunological resistance.
First identified in inbred populations of mice (Vidal et al .,
1995), homologues are now known from a wide variety
of organisms, including humans, insects, plants, yeast
(Cellier et al ., 1995; 1996) and, most intriguingly, the intracellular pathogens M . leprae and M . tuberculosis (accession nos. U15184 and Z95212 respectively). The latter,
which we have designated ‘Mramp ’, was recently identified through the M . tuberculosis genome-sequencing
initiative undertaken by the Sanger Centre, UK. Interestingly, no homologues have been identified in the completely sequenced genomes of free-living bacteria with
the exception of E. coli . In mammals, two isotypes have
been identified – Nramp1 and Nramp2 – the former
restricted to cells of the macrophage/monocyte lineage,
the latter expressed ubiquitously (Cellier et al ., 1996).
410 D. D. Agranoff and S. Krishna
Fig. 2. Model (modified from Supek et al .,
1997; Gunshin et al ., 1997) illustrating the
proposed competition between macrophage
and intraphagosomal mycobacteria for
divalent cations. It is envisaged that Nramp1 ,
located in the membrane of the early
phagosome, functions as a metal ion/proton
symporter, which deprives the mycobacteria
of divalent cations by export from
phagosomes. The mycobacteria may counter
this process through the actions of Mramp
and the type 1 ‘P’-type ATPases, located in
the bacterial plasma membrane, that pump
metal ions from the phagosome into the
bacterium. The associated accumulation of
protons by the mycobacteria could be
countered by a putative bacterial proton efflux
pump. Divalent cations are required by
mycobacteria for the proper functioning of
superoxide dismutase (SOD) and catalase
(CAT), which serve as a defence against
reactive oxygen intermediates such as the
superoxide anion (O2¹ ) and hydroxyl radical
(.OH) generated by the macrophage. X 2þ,
divalent cation, probably Fe 2þ /Cu 2þ /Mn 2þ /
Zn 2þ. Myc, Mycobacterium.
Accumulating evidence points to the function of the
Nramp family as divalent cation transporters of broad specificity (Gunshin et al ., 1997). Thus, the 10–12 hydrophobic
membrane-spanning domains common to all members are
highly conserved, and the amphipathic character of several
of these domains is very suggestive of a transmembrane
hydrophilic pathway. Secondly, a consensus transport
motif common to a variety of bacterial periplasmic transporters (the prokaryotic binding protein-dependent transport signature) is located in the cytoplasmic loop after
transmembrane segment 8 (Fig. 2). Thirdly, the yeast
Nramp homologue, SMF1 , has been shown to be associated with high-affinity Mn 2þ uptake (Supek et al ., 1997).
Most recently, a ubiquitously expressed rat homologue
of Nramp2 (designated DCT1 ) has been shown through
formal electrophysiological studies in Xenopus oocytes
expressing DCT1 mRNA to function as a divalent cation/
proton symporter of broad specificity (Gunshin et al .,
1997). Fe 2þ uptake may be its main physiological role,
although significant uptake of the other predominantly
transition element cations, Zn 2þ, Mn 2þ, Cu 2þ, Co 2þ,
Ni 2þ, Cd 2þ and Pb 2þ, was also demonstrated. Maximal
uptake was observed at an extra-oocytic pH of 5.5.
Competition for heavy metal cations may be related
to intracellular survival
The discovery that Nramp proteins are divalent cation
transporters and the observation that many macrophage-inhabiting intracellular pathogens possess active
cation-transporting mechanisms of their own permits a
new conceptual synthesis. While the Nramp cation transport specificity is fairly broad, it is nevertheless restricted
predominantly to the divalent cations of transition elements.
Many of these have several stable oxidation states enabling
them to take part in redox reactions. As discussed earlier,
several of these metals are precisely those required for the
functioning of the redox enzymes, superoxide dismutase
and catalase. Parasite-encoded versions of these enzymes
represent a major line of defence against the reactive
oxygen intermediates generated by the macrophage during the ‘respiratory burst’, one of the major bactericidal
mechanisms used by the host.
Mycobacteria reside within the phagolysosome compartment of the macrophage, the arena in which many of
the host’s killing mechanisms operate. This is a dynamic
microenvironment, which alters as the phagosome progresses along a well-defined maturation pathway. M .
tuberculosis arrests this maturation at the early phagosome stage, preventing the acidification of the phagosome
below a pH of about 6.3. Other parasites, such as Leishmania spp., reside in a pH below 5.0 (Russell, 1995).
Immunolocalization studies have demonstrated that
Nramp1 is recruited to the early phagosome membrane
(Gruenheid et al ., 1997). This strongly suggests that
Nramp1 is involved in the modulation of the divalent cation
Q 1998 Blackwell Science Ltd, Molecular Microbiology, 28, 403–412
Bacterial ion transport 411
composition of the early phagosome, rendering it hostile to
the parasite. It is therefore tempting to speculate that the
mechanism of action of Nramp1 is to deplete the phagosome of the essential heavy metal cations required by
the parasite for its own defences (Gunshin et al ., 1997).
The parasite, in response, may compete against this
deprivation using its own armoury of metal ion transporters, which, in the case of M . tuberculosis , may include
Mramp and the several heavy metal ‘P’-type ATPases.
The delineation of the relative contributions of Mramp
and ‘P’-type ATPases to cation homeostasis in M . tuberculosis awaits experimental verification. The proposed
macrophage–parasite interaction is illustrated in Fig. 2.
Important questions requiring experimental elucidation
include the functional characterization of Mramp (and its
pH dependence), its localization (probably the mycobacterial plasma membrane), the relative orientations of Nramp ,
Mramp and the mycobacterial type I ‘P’-type ATPases in
their respective membranes and the manner of their
expression during infection. If Mramp and Nramp1 behave
as proton/divalent cation symporters like DCT1 , then the
relatively acidic intraphagosomal environment (even in M .
tuberculosis -infected cells) would favour ion fluxes consistent with our model.
Concluding remarks
The recent identification of Nramp as a divalent cation
transporter coupled with the discovery of complementary
heavy metal transporters in important intracellular pathogens suggests that metal ion dynamics may prove to play
a central role in the preimmunological interaction between
macrophage and parasite. The experimental dissection of
the details of these processes is an important objective
that promises to provide fascinating insights into a fundamental area of biology and that may eventually lead to
the identification of new therapeutic strategies.
Acknowledgements
D.A. is an MRC (UK) Clinical Training Fellow and S.K. is a
Wellcome Trust Senior Research Fellow in Clinical Science.
References
Altendorf, K., Siebers, A., and Epstein, W. (1992) The kdp
ATPase of Escherichia coli . Ann NY Acad Sci 671: 228–
243.
Beisel, W.R. (1977) Magnitude of the host nutritional
responses to infection. Am J Clin Nutr 30: 1236–1247.
Blanc-Potard, A.-B., and Groisman, E.A. (1997) The Salmonella selC locus contains a pathogenicity island mediating
intramacrophage survival. EMBO J 16: 5376–5385.
Blattner, F.R., Plunkett, III, G., Bloch, C.A., Perna, N.T.,
Burland, V., et al . (1997) The complete genome sequence
of Escherichia coli K-12. Science 277: 1453–1474.
Q 1998 Blackwell Science Ltd, Molecular Microbiology, 28, 403–412
Bull, P.C., and Cox, D.W. (1994) Wilson disease and Menkes
disease: new handles on heavy metal transport. Trends
Biochem Sci 10: 246–252.
Bult, C.J., White, O., Olsen, G.J., Zhou, L., and Fleischmann,
R.D. (1996) Complete genome sequence of the methanogenic archaeon, Methanococcus jannaschii . Science 273:
1058–1073.
Burland, V.D., Plunkett, III, G., Sofia, H.J., Daniels, D.L.,
and Blattner, F.R. (1995) Analysis of the Escherichia coli
genome VI: DNA sequence of the region from 92.8 through
100 min. Nucleic Acids Res 23: 2105–2119.
Cellier, M., Prive, G., Belouchi, A., Kwan, T., Rodrigues,
V., Chia, W., et al . (1995) Nramp defines a family of
membrane proteins. Proc Natl Acad Sci USA 92: 10089–
10093.
Cellier, M., Belouchi, A., and Gros, P. (1996) Resistance to
intracellular infections: comparative genomic analysis of
Nramp . Trends Genet 12: 201–204.
Clarke, D., Loo, T., Inesi, G., and MacLennan, D. (1989)
Location of high affinity Ca-binding sites within the predicted transmembrane domain of the sarcoplasmic reticulum Ca ATPase. Nature 339: 476–478.
Epstein, W. (1992) kdp - a bacterial P-type ATPase whose
expression and activity are regulated by turgor pressure.
Acta Physiol Scand 607 (suppl.): 193–199.
Fendler, K., Drose, S., Altendorf, K., and Bamberg, E. (1996)
Electrogenic K þ transport by the kdp -ATPase of Escherichia coli . Biochemistry 35: 8009–8017.
Fleischmann, R.D., Adams, M.D., White, O., Clayton, R.A.,
Kirkness, E., et al . (1995) Whole-genome random sequencing and assembly of Haemophilus influenzae Rd. Science
269: 496–512.
Francis, M.S., and Thomas, C.J. (1997a) The Listeria monocytogenes gene ctpA encodes a putative P-type ATPase
involved in copper transport. Mol Gen Genet 253: 484–
491.
Francis, M.S., and Thomas, C.J. (1997b) Mutants in the ctpA
copper transporting P-type ATPase reduce virulence of
Listeria monocytogenes . Microb Pathog 22: 67–78.
Fraser, C.M., Gocayne, J.D., White, O., Adams, M.D.,
Clayton, R.A., et al . (1995) The minimal gene complement
of Mycoplasma genitalium . Science 270: 397–403.
Fsihi, H., and Cole, S.T. (1995) The Mycobacterium leprae
genome: systemic sequence analysis identifies key catabolic enzymes, ATP-dependent transport systems and a
novel polA locus associated with genomic variability. Mol
Microbiol 16: 909–919.
Ge, Z., and Taylor, D.E. (1996) Helicobacter pylori genes
hpcopA and hpcopP constitute a cop operon involved in
copper export. FEMS Microbiol Lett 145: 181–188.
Ge, Z., Hiratsuka, K., and Taylor, D.E. (1995) Nucleotide
sequence and mutational analysis indicate that two Helicobacter pylori genes encode a P-type ATPase and a cationbinding protein associated with copper transport. Mol
Microbiol 15: 97–106.
Gruenheid, S., Pinner, E., Desjardins, M., and Gros, P.
(1997) Natural resistance to infection with intracellular
pathogens: the Nramp1 protein is recruited to the membrane of the phagosome. J Exp Med 185: 717–730.
Gunshin, H., Mackenzie, B., Berger, U.V., Gunshin, Y.,
Romero, M.F., Boron, W.F., et al . (1997) Cloning and
412 D. D. Agranoff and S. Krishna
characterisation of a mammalian proton-coupled metal-ion
transporter. Nature 388: 482–488.
Himmelreich, R., Hilbert, H., Plagens, H., Pirkl, E., Li, B.C.,
and Herrmann, R. (1996) Complete sequence analysis of
the genome of the bacterium. Mycoplasma pneumoniae
24: 4420–4449
Ivey, D., Guffanti, A., Shen, Z., Kudyan, N., and Krulwich, T.
(1992) The cadC gene product of alkaphilic Bacillus firmus
OF4 partially restores Na resistance to an Escherichia coli
strain lacking an Na/H Antiporter (NhaA). J Bacteriol 174:
4878–4884.
Kahn, D., David, M., Domergue, O., Daveran, M.-L., Ghal, J.,
Hirsch, P.R., and Batut, J. (1989) Rhizobium meliloti fixGHI
sequence predicts involvement of a specific cation pump in
symbiotic nitrogen fixation. J Bacteriol 171: 929–939.
Kanamura, K., Kashiwagi, S., and Mizuno, T. (1993) The
cyanobacterium, Synechococcus sp PCC7942, possesses
two distinct genes encoding cation-transporting P-type
ATPases. FEBS Lett 330: 99–104.
Kaneko, T., Sato, S., Kotani, H., Tanaka, A., Asamizu, E., et
al . (1996) Sequence analysis of the genome of the unicellular cyanobacterium Synechocystis sp. strain PCC6802.
II. Sequence determination of the entire genome and
assignment of potential protein-coding regions. DNA Res
3: 109–136.
Klenk, H.P., Clayton, R.A., Tomb, J.F., White, O., Nelson,
K.E., et al . (19997) The complete genome sequence
of the hyperthermophilic, sulphate-reducing archaeon
Archaeoglobus fulgidus . Nature 390: 364–370.
Kunst, F., Ogasawara, N., Moszer, I., Albertini, A.M., Alloni,
G., et al . (1997) The complete genome sequence of the
Gram-positive bacterium Bacillus subtilis . Nature 390:
249–256.
Lebrun, M., Audurier, A., and Cossart, P. (1994) Plasmidborne cadmium resistance genes in Listeria monocytogenes are similar to cadA of Staphylococcus aureus and
are induced by cadmium. J Bacteriol 176: 3040–3048.
Lutsenko, S., and Kaplan, J.H. (1995) Organization of P-type
ATPases: significance of structural diversity. Biochemistry
34: 15607–15613.
Moller, J., Juul, B., and le Maire, M. (1996) Structural organization, ion transport and energy transduction of P-type
ATPases. Biochim Biophys Acta 1286: 1–51.
Nucifora, G., Chu, L., Mistra, T.K., and Silver, S. (1989) Cadmium resistance from Staphylococcus aureus plasmid
pI258 cad A gene results from a cadmium-efflux ATPase.
Proc Natl Acad Sci USA 86: 3544–3545.
Odermatt, A., Suter, H., Krapf, R., and Solioz, M. (1993)
Primary structure of two P-type ATPases involved in copper homeostasis in Enterococcus hirae . J Biol Chem 268:
12775–12779.
Phung, L.T., Ajlani, G., and Haselkorn, R. (1994) P-type
ATPase from the cyanobacterium Synechococcus 7942
related to the human Menkes and Wilson disease gene
product. Proc Natl Acad Sci USA 91: 9651–9654.
Portillo, G., Foster, J.W., Maguire, M.E., and Finlay, B.B.
(1992) Characterization of the microenvironment of Salmonella typhimurium -containing vacuoles within MDCK
epithelial cells. Mol Microbiol 6: 3289–3297.
Roof, S.M., and Maguire, M.E. (1994) Magnesium transport
systems: genetics and protein structure (a review). J Am
Coll Nutr 13: 424–428.
Russell, D. (1995) Mycobacterium and Leishmania: stowaways in the endosomal network. Trends Cell Biol 5:
125–128.
Sehn, A.P., and Meier, B. (1994) Regulation of an in vivo
metal-exchangeable superoxide dismutase from Propionibacterium shermanii exhibiting activity with different metal
cofactors. Biochem J 304: 803–808.
Silver, S., and Phung, L.T. (1996) Bacterial heavy metal
resistance: new surprises. Annu Rev Microbiol 50: 753–
789.
Silver, S., Nucifora, G., and Phung, L.T. (1993) Human
Menkes X-chromosome disease and the staphylococcal
cadmium-resistance ATPase: a remarkable similarity in
protein sequence. Mol Microbiol 10: 7–12.
Snavely, M.D., Miller, C.G., and Maguire, M.E. (1991) The
mgtB Mg 2þ transport locus of Salmonella typhimurium
encodes a P-type ATPase. J Biol Chem 266: 815–823.
Solioz, M., and Odermatt, A. (1995) Copper and silver transport by copB -ATPase in membrane vesicles of Enterococcus hirae . J Biol Chem 270: 9217–9221.
Supek, F., Supekova, L., Nelson, H., and Nelson, N. (1997)
Function of metal ion homeostasis in the cell division
cycle, mitochondrial protein processing, sensitivity to
mycobacterial infection and brain function. J Exp Biol
200: 321–330.
Tang, X., Halleck, S., Schlegel, R., and Williamson, P. (1996)
A subfamily of P-type ATPases with aminophospholipid
transporting activity. Science 272: 1495–1497.
Tomb, J.F., White, O., Kerlavage, A.R., Clayton, R.A., Sutton, G.G., et al . (1997) The complete genome sequence
of the gastric pathogen Helicobacter pylori . Nature 388:
539–547.
Treuner-Lange, A., Kuhn, A., and Durre, P. (1997) The kdp
System of Clostridium acetobutylicum : cloning, sequencing and transcriptional regulation in response to potassium concentration. J Bacteriol 179: 4501–4512.
Tsai, K.J., Yoon, K.P., and Lynn, A.R. (1992) ATP-dependent
cadmium transport by the cadA cadmium resistance determinant in everted membrane vesicles of Bacillus subtilis .
J Bacteriol 174: 116–121.
Vidal, S., Gros, P., and Skamene, E. (1995) Natural resistance to infection with intracellular parasites: molecular
genetics identifies Nramp1 as the Bcg/Ity/Lsh locus. J
Leukoc Biol 58: 382–390.
Witte, W., Green, L., Misra, T.K., and Silver, S. (1986) Resistance to mercury and to cadmium in chromosomally resistant Staphylococcus aureus. Antimicrob Agents Chemother
29: 663–669.
Q 1998 Blackwell Science Ltd, Molecular Microbiology, 28, 403–412