Article reSeArcH
We investigated whether Na+ ions bind to GIPCs using isother-
mal titration calorimetry (ITC). We produced lipid vesicles from the
mass-spectrometry-analysed GIPC–IPC mixtures extracted from wild-
type and moca1 seedlings (Fig. 5a–d). The mixtures contained more
than 90% GIPCs and more than 90% IPCs for wild-type and moca1
seedlings, respectively (Fig. 5d). Thus, the properties of Na+ binding
to these lipids using wild-type or moca1 mixtures approximate the
properties of Na+ binding to GIPCs or IPCs, respectively. ITC showed
that Na+ binds to GIPCs and IPCs with a dissociation constant (Kd)
of 0.315 ± 0.083 mM and 0.286 ± 0.063 mM, respectively (Fig. 6a, b;
Extended Data Fig. 9a–f; P > 0.05). The number of apparent binding
sites for GIPCs and IPCs was 1.09 ± 0.02 and 0.77 ± 0.09, respectively
(Extended Data Fig. 9c; P < 0.01), in agreement with the presence of
negative charges on PO 4 − and GlcA− in GIPCs and only PO 4 − in IPCs.
Other thermal properties of GIPCs and IPCs were similar (Extended
Data Fig. 9d–f). We also analysed the binding ability of K+ and Li+ to
GIPCs and IPCs (Extended Data Fig. 9g–j). For all three ions, there
were consistently more binding sites in GIPCs than in IPCs, whereas
for a given cation the binding affinities to GIPCs and IPCs were similar.
Given that lipid micro-domains in the plant plasma membrane con-
tain a large amount of GIPCs^27 ,^30 , these results indicate that disruption
of GIPC biosynthesis might reduce GIPC content in moca1, leading to
the reduction in Na+-binding sites and thereby preventing the subse-
quent cell-surface potential depolarization that gates Ca^2 + channels
(Fig. 6c).Discussion
We have tackled the long-standing issue of whether salt-induced Ca^2 +
signalling serves as a salt-sensing mechanism in plants. We identi-
fied the first molecular component required for salt-induced [Ca^2 +]i
elevation, MOCA1, and revealed a biochemical function of GIPCs as
monovalent-cation sensors. Notably, considering that the GT8 family
members in non-plant organisms do not have multiple transmembrane
domains, the mutation in TM6 in moca1 may suggest that in plants,
unique multiple transmembrane domains in the GT8 family have
evolved as an adaptation to salt stress environments (Extended Data
Fig. 1 0). Furthermore, identification of GIPCs as sensors in the salt
CBL–CIPK pathway will shed light on the molecular mechanisms that
underlie the detection of other nutrients via CBL–CIPK pathways^7 ,^8 ,^34 –^36.
We propose a working model for plant salt cation sensing (Fig. 6c):
Na+ ions bind to GIPCs, and depolarize the cell-surface potential to
gate Ca^2 + influx channels. The functioning of ion channels and recep-
tors in the membrane depends critically on how their transmembrane
segments are embedded in the membrane^29 ,^31 ,^32 , and the regulation
of ion channels by cell-surface potentials was recorded more than
40 years ago in animals^33 , although the exact molecular mechanisms
of this process remain unknown. Sphingolipids are structural com-
ponents of membranes found in lipid micro-domains, and also act as
intracellular second messengers in animals^30 ,^32 –^34. However, not much
is known about the binding of sphingolipids to ion channels to gate
them via cell-surface charges. On the other hand, phosphatidylinositol
binds ion channels in the cytoplasmic leaflet and regulates ion channel
function^37 –^40. Evidently, GIPC-mediated salt sensing does not resemble
any known sensory system found in other organisms. Our findings
allow us to propose that rather than a sole salt sensor, cation-sensing****
WTmoca1–25
Zeta potential (mV)–10–15–20Protoplast–30WTmoca140802060Survival rate (%)(^100) Protoplast
0
[NaCl] (mM)
105 15
0
100
80
60
40
20
0
moca1 922.85 (IPC)
1,260.99 (GIPC)
800 900 1,000 1,100 1,200 1,300 1,400
Mass (m/z)
Intensity (%)
100
80
60
40
20
0
WT 1,260.99 (GIPC)
922.85 (IPC)
80
60
100
Relative content (%)
40
20
WT moca1
0
IPC
IPC
GIPC
GIPC
WT
moca1
Content (r.u.)
2
3
4
5
1
(^0) IPC GIPC
f
a
b
c
d
e
Intensity (%)
Fig. 5 | MOCA1-related GIPCs are responsible for NaCl-induced
changes in cell-surface potentials. a, b, Matrix-assisted laser desorption/
ionization-mass spectrometry (MALDI-MS) analysis of GIPCs extracted
from wild-type (a) or moca1 leaves (b). IPC and GIPC series A differ
by their number of saccharide units, from zero (IPC) to two (GIPC
series A). c, d, Relative contents of IPC and GIPC series A in leaves from
experiments as in a and b were quantified with wild-type GIPC content
normalized as 1 (r.u., relative unit; c), or as a relative percentage between
IPC and GIPC with the total of IPC and GIPC as 100%
(d). Data are presented as mean ± s.d. (n = 3; Student’s t-test,
*P < 0.001, P < 0.01). e, ζ potentials of mesophyll cell protoplasts
plotted as a function of applied [NaCl]. Date are from four representative
experiments (mean ± s.d.; n = 20 protoplasts per data point; two-way
ANOVA, P < 0.001). f, Survival rates of protoplasts plotted as a function
of applied [NaCl]. Data are presented as mean ± s.d. (n = 5 repeats (15
protoplasts); two-way ANOVA, P < 0.001).
Na+
Na+ Na+
Ca2+
Out
In
InOuta b0.40Time (min)
10 20 3000.81.21.6ITC (μcal s–1)400 123 4
Na+-to-lipid (outside) ratio01020304000.040.080 123 4Injectant (kcal mol–1)0.12cTime (min)Na+-to-lipid (outside) ratio
GIPCGIPC
–––––––GIPCNa+ GIPC Na+Fig. 6 | Na+ binds to GIPCs and gates Ca^2 + influx channels. a, b, ITC
analysis of Na+ binding to GIPCs from wild-type GIPC–IPC mixture with
90% GIPCs. ITC data (a) and plots of injected heat for NaCl injections
into the sample cell are shown (b). Six independent experiments were
performed, and similar results were obtained. c, Model of how GIPCs
sense salt and gate Ca^2 + influx channels.
15 AUGUSt 2019 | VOl 572 | NAtUre | 345