Directory UMM :Data Elmu:jurnal:P:PlantScience:Plant Science_BioMedNet:101-120:
477
Plant ion channels: from molecular structures to
physiological functions
Sabine Zimmermann* and Hervé Sentenac†
Progress in identification of plant ion channels and
development of electrophysiological analyses in heterologous
expression systems and in planta, in combination with reverse
genetic approaches, are providing the possibility of
associating molecular entities with physiological functions.
Recently, the first attempts to determine in vivo functions
using knockout mutants demonstrated the roles of root ion
channels. The search for proteins interacting with such
channels leads to an even more complex view of the
concerted action in protein networks.
Addresses
*Max-Planck-Institut für Molekulare Pflanzenphysiologie, D-14424
Potsdam, Germany; e-mail: [email protected]
†Biochimie et Physiologie Moléculaire des Plantes, INRA/ENSAM/CNRS URA 2133, Place Viala, F-34 060 Montpellier cedex 1,
France; e-mail [email protected]
Current Opinion in Plant Biology 1999, 2:477–482
1369-5266/99/$ — see front matter © 1999 Elsevier Science Ltd.
All rights reserved.
Abbreviations
ABA
abscisic acid
AKT1
Arabidopsis K+ transporter 1
GFP
green fluorescent protein
IRK
inward-rectifying K+
KAT1
K+ Arabidopsis thaliana channel 1
KCO
K+ channel family related to the Ca2+-activated, outwardly
rectifying K+ channel
KST1
K+ channel of Solanum tuberosum
P domain pore domain
SKOR
Stellar outward-rectifying K+ channel
TWIK
tandem of P domains in a weak inward-rectifying K+
Introduction
Channels, when open, form selective pores through
which transported ions can move without inducing a
general conformational change of the protein. The maximum velocity at which ions can move through channels
(107 ions per second) is therefore several orders of magnitude higher than the velocity at which ions can be
transported by carriers. Ion channels are involved in the
control of membrane potential and signal transduction in
plants [1], as in animals. In plants, they are also involved
in sustained transport, such as uptake of ions from the
soil solution or secretion of ions into the xylem sap [2].
The first channels identified in plants were two K+ channels from Arabidopsis, AKT1 (Arabidopsis K+ transporter 1)
[3] and KAT1 (K+ Arabidopsis thaliana channel 1 ) [4], both
cloned in 1992 by functional complementation of yeast
mutant strains defective for K+ transport. Various molecular approaches and electronic cloning have thereafter
revealed a large multigene family of channels related to
AKT1 and KAT1, sharing homologies with animal Shakertype channels and grouped in the so-called plant Shaker
family, two other families of K+ channels related to the
animal TWIK (tandem of P-domains in a weak inwardrectifying K+) and IRK (inward-rectifying K+) channels
[5•], and four other gene families likely to encode cation or
anion channels from their homologies with animal channels. Given that ∼ 70% of the open reading frames in the
Arabidopsis genome have now been fully or partially
sequenced, it may be assumed that most plant ion channel
families with counterparts in animals have been identified.
This advancement has shifted research priorities from
cloning to characterization of the gene products and identification of their roles in planta. Although work in this field
is in its infancy, molecular approaches, aimed at broadening the view by identifying regulatory mechanisms and
interacting networks, have commenced.
The first part of this review is focused on the work of the
past two years on plant Shaker channels, which has provided both highly significant results at the biological level and
a paradigm in this field of research by closely associating
tools from molecular biology, reverse genetics and electrophysiology. The second and third part of the review
present exciting progress in characterization of other ion
channel families and analysis of channel regulation.
Plant K+ channels of the Shaker family
At present, all K+ channels cloned, from prokaryotes to
eukaryotes, are classified into four groups according to
their structural features [6] (Figure 1). They all contain a
characteristic domain, named H5 or the pore (P) domain,
which forms part of the aqueous pore of the channel and
controls permeation and ionic selectivity [7••]. They differ in the number of transmembrane segments (two,
four, six or eight) and P domains (one or two) present
per polypeptide. A milestone in elucidating the structure–function relationship of K+ channels was the
determination of the three-dimensional structure of a
bacterial K+ channel from the two transmembrane segments/one pore family [7••].
Structure–function relationship
Regions similar to the six transmembrane segments (called
S1–S6) and the P domain of animal Shaker channels are
present in the first channels identified in Arabidopsis,
AKT1, KAT1 and in their relatives (Figure 1). The postulated topology derived from animal Shaker channels
(Figure 2) has been confirmed for KAT1 [8]. Experimental
evidence has been obtained indicating that the S4 segment
and the P domain in the plant Shaker-like channels play
the same roles as their animal counterparts [6]. Namely,
the S4 domain senses the transmembrane voltage and
478
Cell biology
Figure 1
Figure 2
(a)
Shaker-type
S1
Voltage sensor
S6
P
+
+
+
+
+
(b)
S1
Exterior
S2 S4
Membrane
S3
S2
S6
S3
S4 S6
S5
S5
Cytosol
S1
C
N
P
IRK-type
P
P
cNMP
Aqueous pore
TWIK-type
P1
Amino
terminus
Anky
KHA
P2
Carboxyl-terminus
Current Opinion in Plant Biology
TOK-type
P1
P2
Current Opinion in Plant Biology
Secondary structure of the four K+ channel types. Channels belonging
to the Shaker family share a typical hydrophobic structure consisting of
six transmembrane segments (S1–S6), the P domain being present
between the fifth and the sixth segments. The second family, named IRK
in animal cells, includes channels displaying two transmembrane
segments, the P domain being present between them. The third and
fourth groups, named TWIK in animal cells and TOK (two-pore
outwardly rectifying K+ channel) in yeast, correspond to polypeptides
with two P domains and either four or eight putative transmembrane
segments. In the former case, the hydropathy profile suggests a union
of two IRK-like structures. In the latter case, the hydropathy profile is
reminiscent of a Shaker-like structure attached to an IRK-like structure.
The first three families have members in both animal [6] and plant cells
[5•]. Only in yeast has a channel of the fourth family been identified [5•].
controls channel activation, while the P domain forms the
aqueous pore and controls permeation.
In this field of research, KAT1 and its counterpart KST1
(K+ channel of Solanum tuberosum) in potato [9] have been
used as models, at least in part because these channels can
be easily characterized in Xenopus oocytes. Furthermore,
within the Shaker family, KAT1 was the first channel known
to be endowed with inward rectification (i.e. mediating
K+ influx), as all the animal channels cloned at that time
were characterized as outward rectifiers (i.e. mediating
K+ efflux). This discovery gave rise to a question debated
within the electrophysiologist community (plant and animal), namely whether hyperpolarization-induced KAT1
inward currents were related to the channel activation [10]
Structure of plant Shaker-like channels. The proposed structure is
derived from sequence and structural homologies with animal Shaker
channels, direct analyses of the transmembrane topology, mutagenesis
of pore mutants and biochemical characterization (see Figure 1 and
text). (a) The P domain between the fifth and sixth transmembrane
region forms part of the aqueous pore [12,13••]. The voltage sensor in
the fourth transmembrane span [16,18] is surrounded by the other
transmembrane spans to isolate positive charges. Functional domains
are present in the cytosolic carboxyl terminus: a putative cyclic
nucleotide-binding domain (cNMP), an interaction domain (KHA), and in
the AKT-like subfamily only, ankyrin repeats (Anky). (b) Tetramerization
of four α-subunits results in the formation of functional channels.
or to recovery from inactivation [11]. The former hypothesis has recently received further support [12,13••].
Furthermore, recent identification of an outwardly rectifying K+ channel SKOR (Steller outward-rectifying
K+ channel) [14••] and of a weakly rectifying K+ channel
AKT3 [15•] among the plant Shaker family clearly indicates
that similar structures can be endowed with different rectification mechanisms. The plant Shaker family is therefore
providing an exciting model for electrophysiologists interested in structure–function relationship analyses.
Site-directed mutagenesis of the KAT1 S4 domain coupled
to electrophysiological characterization of mutated
channnels in Xenopus oocytes has recently identified positively-charged amino acids as voltage sensing residues
[12,13••]. A similar strategy has identified mutations in the
sequence encoding the P domain of KAT1 that alter channel permeation properties and confer increased sensitivity
to blocking by Ca2+ [16] or stimulation by H+ [17•].
Although the AKT1 channel cannot be characterized in
Xenopus oocytes (for reasons that have not yet been elucidated), random mutagenesis coupled to characterization of
mutated channels by functional expression in yeast has provided evidence that the P domain controls permeation in
Plant ion channels Zimmermann and Sentenac
Figure 3
Leaf
Knockout mutants highlight the role of Shaker channels
in mineral nutrition
Expression of the AKT2 gene (also called AKT3) in
phloem tissues has been reported recently, suggesting a
role for the encoded K+ channel in long distance K+ transport via the phloem vasculature [15•]. Expression studies
and electrophysiological approaches support the hypothesis that KAT1 [24,25] and its counterpart KST1 in potato
[9] mediate K+ influx in guard cells leading to stomatal
opening, probably in both the low (
1 mM) K+ concentration range [26]. It should be noted,
however, that differences have been found between the
functional properties of the major guard cell inward channel and those of KAT1 expressed in Xenopus oocytes [27•].
Such differences could result from the fact that characterization in heterologous systems might provide a distorted
view, because of artefactual interactions and/or lack of
control by plant proteins [28•].
Using knockout mutants, clear-cut evidence has been
obtained for the roles of two Arabidopsis K+ channels AKT1
and SKOR (Figure 3). AKT1 was characterized as an
Root
(b)
Phloem
Stem
(a)
In plant Shaker-like K+ channels, a putative cyclic
nucleotide binding site and a domain rich in hydrophobic
and acidic residues, called KHA, are present downstream
from the hydrophobic core, in the cytoplasmic carboxyterminal domain (Figure 1). In some channels (the AKT
subfamily), several ankyrin repeats are present between
these two domains and could play a role in protein–protein
interactions [3]. Expression of AKT1 or KAT1 cDNAs in
insect cells results in the appearance of membrane proteins
with an apparent molecular weight that is four times that of
the corresponding predicted polypeptide, indicating that
AKT1 and KAT1 are tetrameric channels ([22]; S Urbach,
I Cherel, H Sentenac, F Gaymard, unpublished data), like
their animal counterparts. When expressed in insect cells,
the cytoplasmic carboxyterminal region of AKT1 forms
highly stable homotetrameric structures, suggesting a role
for this region in channel tetramerization [22]. Expression
of a fusion protein between KST1 and green fluorescent
protein (GFP) has revealed that the KHA domain is not
essential for the expression of functional channels in insect
cells but plays a role in channel clustering [23].
(c)
(d)
Xylem
this channel also [18]. In KST1, two histidine residues present at the external face of the channel, one within the P
domain and the other within the S3–S4 linker, have been
shown to act as pH sensing elements, upregulating the
channel activity upon acidification of the medium [19]. A
similar upregulation has been shown for KAT1, but involves
distinct molecular mechanisms [17•]. Comparison of KAT1
currents at different K+/Rb+ ratios has led to the conclusion
that KAT1 conduction requires several K+ ions to be present
simultaneously within the pore [20]. This so-called multiion pore behavior is also supported by the demonstration
that the impermeant ion methylammonium blocks K+ and
NH4+ currents through KAT1 differently [21].
479
Current Opinion in Plant Biology
K+ transport within Arabidopsis thaliana. Ion channels have been
identified by (a) employing insertion mutants to mediate K+ uptake in
the root (AKT1 [31••]) and (b) K+ secretion into the xylem for long
distance transport (SKOR [14••]). (c) AKT2/3 [15•] is assumed to
contribute to K+ distribution via the phloem. (d) KAT1 was shown to be
responsible for K+ uptake into guard cells during stomatal opening [25].
inwardly rectifying K+ channel [29] preferentially
expressed in the root epiderm and cortex [30], suggesting a
role for AKT1 in K+ uptake from the soil solution. This has
been definitively confirmed by phenotype characterization
of a knockout mutant [31••]. Growth of the mutant is
reduced under limiting K+ concentrations (
Plant ion channels: from molecular structures to
physiological functions
Sabine Zimmermann* and Hervé Sentenac†
Progress in identification of plant ion channels and
development of electrophysiological analyses in heterologous
expression systems and in planta, in combination with reverse
genetic approaches, are providing the possibility of
associating molecular entities with physiological functions.
Recently, the first attempts to determine in vivo functions
using knockout mutants demonstrated the roles of root ion
channels. The search for proteins interacting with such
channels leads to an even more complex view of the
concerted action in protein networks.
Addresses
*Max-Planck-Institut für Molekulare Pflanzenphysiologie, D-14424
Potsdam, Germany; e-mail: [email protected]
†Biochimie et Physiologie Moléculaire des Plantes, INRA/ENSAM/CNRS URA 2133, Place Viala, F-34 060 Montpellier cedex 1,
France; e-mail [email protected]
Current Opinion in Plant Biology 1999, 2:477–482
1369-5266/99/$ — see front matter © 1999 Elsevier Science Ltd.
All rights reserved.
Abbreviations
ABA
abscisic acid
AKT1
Arabidopsis K+ transporter 1
GFP
green fluorescent protein
IRK
inward-rectifying K+
KAT1
K+ Arabidopsis thaliana channel 1
KCO
K+ channel family related to the Ca2+-activated, outwardly
rectifying K+ channel
KST1
K+ channel of Solanum tuberosum
P domain pore domain
SKOR
Stellar outward-rectifying K+ channel
TWIK
tandem of P domains in a weak inward-rectifying K+
Introduction
Channels, when open, form selective pores through
which transported ions can move without inducing a
general conformational change of the protein. The maximum velocity at which ions can move through channels
(107 ions per second) is therefore several orders of magnitude higher than the velocity at which ions can be
transported by carriers. Ion channels are involved in the
control of membrane potential and signal transduction in
plants [1], as in animals. In plants, they are also involved
in sustained transport, such as uptake of ions from the
soil solution or secretion of ions into the xylem sap [2].
The first channels identified in plants were two K+ channels from Arabidopsis, AKT1 (Arabidopsis K+ transporter 1)
[3] and KAT1 (K+ Arabidopsis thaliana channel 1 ) [4], both
cloned in 1992 by functional complementation of yeast
mutant strains defective for K+ transport. Various molecular approaches and electronic cloning have thereafter
revealed a large multigene family of channels related to
AKT1 and KAT1, sharing homologies with animal Shakertype channels and grouped in the so-called plant Shaker
family, two other families of K+ channels related to the
animal TWIK (tandem of P-domains in a weak inwardrectifying K+) and IRK (inward-rectifying K+) channels
[5•], and four other gene families likely to encode cation or
anion channels from their homologies with animal channels. Given that ∼ 70% of the open reading frames in the
Arabidopsis genome have now been fully or partially
sequenced, it may be assumed that most plant ion channel
families with counterparts in animals have been identified.
This advancement has shifted research priorities from
cloning to characterization of the gene products and identification of their roles in planta. Although work in this field
is in its infancy, molecular approaches, aimed at broadening the view by identifying regulatory mechanisms and
interacting networks, have commenced.
The first part of this review is focused on the work of the
past two years on plant Shaker channels, which has provided both highly significant results at the biological level and
a paradigm in this field of research by closely associating
tools from molecular biology, reverse genetics and electrophysiology. The second and third part of the review
present exciting progress in characterization of other ion
channel families and analysis of channel regulation.
Plant K+ channels of the Shaker family
At present, all K+ channels cloned, from prokaryotes to
eukaryotes, are classified into four groups according to
their structural features [6] (Figure 1). They all contain a
characteristic domain, named H5 or the pore (P) domain,
which forms part of the aqueous pore of the channel and
controls permeation and ionic selectivity [7••]. They differ in the number of transmembrane segments (two,
four, six or eight) and P domains (one or two) present
per polypeptide. A milestone in elucidating the structure–function relationship of K+ channels was the
determination of the three-dimensional structure of a
bacterial K+ channel from the two transmembrane segments/one pore family [7••].
Structure–function relationship
Regions similar to the six transmembrane segments (called
S1–S6) and the P domain of animal Shaker channels are
present in the first channels identified in Arabidopsis,
AKT1, KAT1 and in their relatives (Figure 1). The postulated topology derived from animal Shaker channels
(Figure 2) has been confirmed for KAT1 [8]. Experimental
evidence has been obtained indicating that the S4 segment
and the P domain in the plant Shaker-like channels play
the same roles as their animal counterparts [6]. Namely,
the S4 domain senses the transmembrane voltage and
478
Cell biology
Figure 1
Figure 2
(a)
Shaker-type
S1
Voltage sensor
S6
P
+
+
+
+
+
(b)
S1
Exterior
S2 S4
Membrane
S3
S2
S6
S3
S4 S6
S5
S5
Cytosol
S1
C
N
P
IRK-type
P
P
cNMP
Aqueous pore
TWIK-type
P1
Amino
terminus
Anky
KHA
P2
Carboxyl-terminus
Current Opinion in Plant Biology
TOK-type
P1
P2
Current Opinion in Plant Biology
Secondary structure of the four K+ channel types. Channels belonging
to the Shaker family share a typical hydrophobic structure consisting of
six transmembrane segments (S1–S6), the P domain being present
between the fifth and the sixth segments. The second family, named IRK
in animal cells, includes channels displaying two transmembrane
segments, the P domain being present between them. The third and
fourth groups, named TWIK in animal cells and TOK (two-pore
outwardly rectifying K+ channel) in yeast, correspond to polypeptides
with two P domains and either four or eight putative transmembrane
segments. In the former case, the hydropathy profile suggests a union
of two IRK-like structures. In the latter case, the hydropathy profile is
reminiscent of a Shaker-like structure attached to an IRK-like structure.
The first three families have members in both animal [6] and plant cells
[5•]. Only in yeast has a channel of the fourth family been identified [5•].
controls channel activation, while the P domain forms the
aqueous pore and controls permeation.
In this field of research, KAT1 and its counterpart KST1
(K+ channel of Solanum tuberosum) in potato [9] have been
used as models, at least in part because these channels can
be easily characterized in Xenopus oocytes. Furthermore,
within the Shaker family, KAT1 was the first channel known
to be endowed with inward rectification (i.e. mediating
K+ influx), as all the animal channels cloned at that time
were characterized as outward rectifiers (i.e. mediating
K+ efflux). This discovery gave rise to a question debated
within the electrophysiologist community (plant and animal), namely whether hyperpolarization-induced KAT1
inward currents were related to the channel activation [10]
Structure of plant Shaker-like channels. The proposed structure is
derived from sequence and structural homologies with animal Shaker
channels, direct analyses of the transmembrane topology, mutagenesis
of pore mutants and biochemical characterization (see Figure 1 and
text). (a) The P domain between the fifth and sixth transmembrane
region forms part of the aqueous pore [12,13••]. The voltage sensor in
the fourth transmembrane span [16,18] is surrounded by the other
transmembrane spans to isolate positive charges. Functional domains
are present in the cytosolic carboxyl terminus: a putative cyclic
nucleotide-binding domain (cNMP), an interaction domain (KHA), and in
the AKT-like subfamily only, ankyrin repeats (Anky). (b) Tetramerization
of four α-subunits results in the formation of functional channels.
or to recovery from inactivation [11]. The former hypothesis has recently received further support [12,13••].
Furthermore, recent identification of an outwardly rectifying K+ channel SKOR (Steller outward-rectifying
K+ channel) [14••] and of a weakly rectifying K+ channel
AKT3 [15•] among the plant Shaker family clearly indicates
that similar structures can be endowed with different rectification mechanisms. The plant Shaker family is therefore
providing an exciting model for electrophysiologists interested in structure–function relationship analyses.
Site-directed mutagenesis of the KAT1 S4 domain coupled
to electrophysiological characterization of mutated
channnels in Xenopus oocytes has recently identified positively-charged amino acids as voltage sensing residues
[12,13••]. A similar strategy has identified mutations in the
sequence encoding the P domain of KAT1 that alter channel permeation properties and confer increased sensitivity
to blocking by Ca2+ [16] or stimulation by H+ [17•].
Although the AKT1 channel cannot be characterized in
Xenopus oocytes (for reasons that have not yet been elucidated), random mutagenesis coupled to characterization of
mutated channels by functional expression in yeast has provided evidence that the P domain controls permeation in
Plant ion channels Zimmermann and Sentenac
Figure 3
Leaf
Knockout mutants highlight the role of Shaker channels
in mineral nutrition
Expression of the AKT2 gene (also called AKT3) in
phloem tissues has been reported recently, suggesting a
role for the encoded K+ channel in long distance K+ transport via the phloem vasculature [15•]. Expression studies
and electrophysiological approaches support the hypothesis that KAT1 [24,25] and its counterpart KST1 in potato
[9] mediate K+ influx in guard cells leading to stomatal
opening, probably in both the low (
1 mM) K+ concentration range [26]. It should be noted,
however, that differences have been found between the
functional properties of the major guard cell inward channel and those of KAT1 expressed in Xenopus oocytes [27•].
Such differences could result from the fact that characterization in heterologous systems might provide a distorted
view, because of artefactual interactions and/or lack of
control by plant proteins [28•].
Using knockout mutants, clear-cut evidence has been
obtained for the roles of two Arabidopsis K+ channels AKT1
and SKOR (Figure 3). AKT1 was characterized as an
Root
(b)
Phloem
Stem
(a)
In plant Shaker-like K+ channels, a putative cyclic
nucleotide binding site and a domain rich in hydrophobic
and acidic residues, called KHA, are present downstream
from the hydrophobic core, in the cytoplasmic carboxyterminal domain (Figure 1). In some channels (the AKT
subfamily), several ankyrin repeats are present between
these two domains and could play a role in protein–protein
interactions [3]. Expression of AKT1 or KAT1 cDNAs in
insect cells results in the appearance of membrane proteins
with an apparent molecular weight that is four times that of
the corresponding predicted polypeptide, indicating that
AKT1 and KAT1 are tetrameric channels ([22]; S Urbach,
I Cherel, H Sentenac, F Gaymard, unpublished data), like
their animal counterparts. When expressed in insect cells,
the cytoplasmic carboxyterminal region of AKT1 forms
highly stable homotetrameric structures, suggesting a role
for this region in channel tetramerization [22]. Expression
of a fusion protein between KST1 and green fluorescent
protein (GFP) has revealed that the KHA domain is not
essential for the expression of functional channels in insect
cells but plays a role in channel clustering [23].
(c)
(d)
Xylem
this channel also [18]. In KST1, two histidine residues present at the external face of the channel, one within the P
domain and the other within the S3–S4 linker, have been
shown to act as pH sensing elements, upregulating the
channel activity upon acidification of the medium [19]. A
similar upregulation has been shown for KAT1, but involves
distinct molecular mechanisms [17•]. Comparison of KAT1
currents at different K+/Rb+ ratios has led to the conclusion
that KAT1 conduction requires several K+ ions to be present
simultaneously within the pore [20]. This so-called multiion pore behavior is also supported by the demonstration
that the impermeant ion methylammonium blocks K+ and
NH4+ currents through KAT1 differently [21].
479
Current Opinion in Plant Biology
K+ transport within Arabidopsis thaliana. Ion channels have been
identified by (a) employing insertion mutants to mediate K+ uptake in
the root (AKT1 [31••]) and (b) K+ secretion into the xylem for long
distance transport (SKOR [14••]). (c) AKT2/3 [15•] is assumed to
contribute to K+ distribution via the phloem. (d) KAT1 was shown to be
responsible for K+ uptake into guard cells during stomatal opening [25].
inwardly rectifying K+ channel [29] preferentially
expressed in the root epiderm and cortex [30], suggesting a
role for AKT1 in K+ uptake from the soil solution. This has
been definitively confirmed by phenotype characterization
of a knockout mutant [31••]. Growth of the mutant is
reduced under limiting K+ concentrations (