The gravistimulation-induced very slow Ca2+ increase in Arabidopsis seedlings requires MCA1, a Ca2+-permeable mechanosensitive channel

Gravity is a critical environmental factor affecting the morphology and function of plants on Earth. Gravistimulation triggered by changes in the gravity vector induces an increase in the cytoplasmic free calcium ion concentration ([Ca2+]c) as an early process of gravity sensing; however, its role and molecular mechanism are still unclear. When seedlings of Arabidopsis thaliana expressing apoaequorin were rotated from the upright position to the upside-down position, a biphasic [Ca2+]c-increase composed of a fast-transient [Ca2+]c-increase followed by a slow [Ca2+]c-increase was observed. We find here a novel type [Ca2+]c-increase, designated a very slow [Ca2+]c-increase that is observed when the seedlings were rotated back to the upright position from the upside-down position. The very slow [Ca2+]c-increase was strongly attenuated in knockout seedlings defective in MCA1, a mechanosensitive Ca2+-permeable channel (MSCC), and was partially restored in MCA1-complemented seedlings. The mechanosensitive ion channel blocker, gadolinium, blocked the very slow [Ca2+]c-increase. This is the first report suggesting the possible involvement of MCA1 in an early event related to gravity sensing in Arabidopsis seedlings.

The [Ca 2+ ] c -increases induced by gravistimulation have been studied in more detail in Arabidopsis (Arabidopsis thaliana) seedlings with aequorin, a luminous Ca 2+ -reporting protein [12][13][14] . When gravistimulated by turning from an upright to upside-down position, the seedlings showed a biphasic [Ca 2+ ] c -increase in their hypocotyls and petioles 14 imaged with an ultrasensitive photon counting camera; a fast-transient [Ca 2+ ] c -increase and a subsequent slow [Ca 2+ ] c -increase. The fast-transient [Ca2 + ] c -increase depended on the rotational velocity 13 but not on the rotational angle, whereas the slow [Ca 2+ ] c -increase depended on the rotational angle but not the rotational velocity 14 .
Recently, the slow [Ca 2+ ] c -increase has been demonstrated as a gravistimulation-specific Ca 2+ -response using μg conditions produced by parabolic flights (PF) 13 ; the μg conditions allowed the rotation of seedlings without gravistimulation, the fast-transient [Ca 2+ ] c -increase was induced by rotation under the μg condition, and the slow [Ca 2+ ] c -increase was initiated by transition from μg to ca. 1.5 g when the µg condition was terminated 12,14 , confirming the idea that the fast-transient is principally induced by rotational stimulation, and the slow increase is genuinely induced by gravistimulation.
The slow [Ca 2+ ] c -increase was significantly attenuated by the potential MSCC inhibitors Gd 3+ and La 3+ , and the Ca 2+ chelator 1,2-bis (2-aminophenoxy) ethane-N,N,N′,N′-tetraacetic acid (BAPTA), suggesting that it depends on the Ca 2+ influx via MSCCs in the plasma membrane. The slow transient [Ca 2+ ] c -increase was also attenuated by the actin-disrupting drugs cytochalasin B and latrunculin B 14 . These observations agree with the "starch-statolith model" 6,15,16 , which includes MSCCs and actin filament networks; sedimentation of "starchstatolith" increases the stress within the actin filament network and activates MSCCs 16 . However, the molecular entities of plasma membrane MSCCs responsible for sensing and responding to the gravistimulation have not been elucidated.
The functional properties of MCA1 have been examined in a variety of studies; Ca 2+ uptake was increased in Arabidopsis seedlings and yeast cells by overexpressing MCA1 in the plasma membrane 19 . Seedlings of mca1knockout lines lacked the ability to penetrate their roots into a harder agar 19 . Membrane stretching elevates [Ca 2+ ] c in CHO cells expressing MCA1 19 , and expression of Arabidopsis MCA1 leads to enhanced mechanosensitive cation channel activity (34 pS) in the Xenopus laevis oocyte plasma membrane 25 . Elongation growth of hypocotyls was suppressed under the hypergravity condition in wild-type Arabidopsis seedlings, whereas the extent of the suppression was reduced in mca1-knockout seedlings, but was augmented in MCA1-overexpressing seedlings 28,29 . These findings suggest that MCA1 is a plasma membrane Ca 2+ -permeable MS cation channel that is potentially involved in gravity sensing and the subsequent morphological changes in Arabidopsis seedlings.
In this study, [Ca 2+ ] c -increases triggered by rotating Arabidopsis seedlings under 1-5 g conditions were investigated. With a backward rotation from an upside-down to an upright position, a novel, "very slow" [Ca 2+ ] c -increase following the biphasic [Ca 2+ ] c -increase was found. In addition, mca1-knockout seedlings showed a reduced amplitude of the very slow [Ca 2+ ] c -increase, and the reduction was partially restored in MCA1-complemented seedlings. This is the first report that characterizes the very slow [Ca 2+ ] c -increase by using 1-5 g gravitational acceleration, and the role of MCA1 in the gravistimulation-induced [Ca 2+ ] c -increase.

Results
A novel very slow [Ca 2+ ] c -increase induced by backward rotation from the upside-down to the upright position. Arabidopsis seedlings containing aequorin were rotated to the upside-down position and then rotated back to the upright position under 1-5 g conditions (Fig. 1A), and gravistimulation-induced [Ca 2+ ] c -changes were monitored. As described previously 14 Table S1), then backward rotation to the upright position was given at 400 s after the first rotation where the slow [Ca 2+ ] c -increase almost disappeared (Fig. 1B). As described previously 14 , the backward rotation also promoted the biphasic Ca 2+ -signal with similar kinetics to that in the first rotation, but the amplitude of the second slow [Ca 2+ ] c -increase was apparently attenuated, and often not observed (black single arrowhead with amplitude A 1 ′ in Figs ] c -increase, the amplitude of the response was almost saturated at 3 g acceleration (nearly the same peak amplitudes of [Ca 2+ ] c -increase were detected in 3-5 g, Fig. 3B), whereas the amplitude of the slow [Ca 2+ ] c -increase was augmented, depending on the g magnitude; the saturation was not seen in the range from 1 to 5 g (Fig. 3A). Thus, the g dependency of the slow and the very slow [Ca 2+ ] c -increases is not the same.  Table S1).
The [Ca 2+ ] c -increase in mca1-knockout seedlings induced by the same stimulation protocol showed that the amplitudes of the very slow [Ca 2+ ] c -increases were profoundly decreased (Figs. 2B, 3, 4D,F). The [Ca 2+ ] c -increase in MCA1-complemented mca1-knockout seedlings induced by the same stimulation protocol showed that the   Fig. 3B inset). The kinetic parameters (T 1 , T 2 , τ 1 , τ 2 ) and the amplitudes of the fast transients (A 0 , A 0 ′) were not changed apparently in mca1-knockout seedlings as shown in Supplementary Table S1. Note that the slow [Ca 2+ ] c -increase was also affected to a lesser extent under 1-4 g, but the difference between WT and mca1 mutants   Fig. 4G). These data suggest that MCA1, a mechanosensitive cation channel in the plasma membrane, is involved in the gravistimulation-induced [Ca 2+ ] c -increases.  (Fig. 4D).

Kinetic parameters of the very slow [Ca 2+
] c -increase were not affected by a sudden decrease in gravitational acceleration from 3 to 1 g. The PF experiment revealed that when gravistimulation triggered the response, a sudden gravitational decrease from 2 g to µg did not attenuate the slow [Ca 2+ ] c -increase; i.e., the response was not affected by the sudden g decrease. This indicates that once the gravity sensing process is activated, the slow [Ca 2+ ] c -response proceeds irrespective of environmental gravity changes. The amplitude of the centrifugation was reduced from 3 to 1 g and the time course of the very slow [Ca 2+ ] c -increase of WT plants was examined (Fig. 5A); the time course was not changed apparently by the 3-1 g sudden decline, as shown in Fig. 5B, suggesting that once the gravity sensing process is activated, the very slow [Ca 2+ ] c -response proceeds irrespective of environmental gravity changes.

Discussion
In the present study, the [Ca 2+ ] c changes in response to changes in the gravity vector under hypergravity conditions up to 5 g were examined. The backward rotation from the upside-down to upright positions induced a very slow [Ca 2+ ] c -increase. The very slow [Ca 2+ ] c -increase was profoundly diminished in mca1-knockout  ] c -increase was not changed when the gravitational acceleration declined from 3 to 1 g at the peak of the response, which agrees with the hypothesis that gravistimulation produced by the backward rotation leads to the activation of the plasma membrane MSCCs (e.g., MCA1), followed by a molecular interaction and signaling process that were not affected by gravity changes.
Our previous pharmacological study using Ruthenium Red suggested the involvement of endomembrane Ca 2+ -permeable channels in the slow [Ca 2+ ] c -increase. The intracellular inositol 1,4,5-trisphosphate (InsP 3 ) level is known to increase within 15 s of gravistimulation in oat shoot pulvini, suggesting the possible involvement of phospholipase C (PLC)-dependent signaling. Thus, once the gravity sensing process is triggered, it can be followed by a signaling process including the activation of PLC, production of InsP 3 , and InsP 3 -induced Ca 2+ -release (IICR) from the ER (or the vacuole), which may be less sensitive to changes in the gravity environment. This explains why the kinetic parameters of the slow and very slow [Ca 2+ ] c -increases were not affected by the different magnitudes of gravitational acceleration (Supplementary Table S1); i.e., the slow kinetics may be predominantly governed by the Ca 2+ release processes from cellular organelles 13,31 .
Enhanced gravitropism by hypergravity (hyper-gravitropism) is reported in Arabidopsis roots 32 , stems 33 and hypocotyls 34 . The analysis of the plant growth under the sequential hypergravity conditions (30-500 g for 1 day) shows that hypergravity suppressed elongation growth of hypocotyls, but this effect was reduced in hypocotyls of mca-null mutants compared with the wild type 28 , suggesting that MCAs are involved in the sensing of gravity signals in plants. Hyper-gravitropism was not examined in this study, since it may need very high g condition and long observation period according on the above study.
The fast-transient [Ca 2+ ] c -increase was presumably a genuine response to rotation, because it was induced by rotatory stimulation even under µg, and its amplitude increases in an angular acceleration-dependent manner. Recently, it was demonstrated that the amplitude of the fast-transient [Ca 2+ ] c -increase (A 0 and A 0 ′) is augmented by the gravitational acceleration from μg to 2 g. The present study confirmed these results under a wider range of gravitational acceleration, from 1 to 5 g. In contrast, the amplitude of the slow [Ca 2+ ] c -increase depended on the rotational angle (change in the direction of gravity) 12,14 , not on the rotational velocity, and it is a genuine g-change response. These support the idea that the backward rotation induces fast transient, slow and very slow [Ca 2+ ] c -increases as illustrated in Fig. 1, however, the possibility is not excluded that further investigation reveals additional complexity in the [Ca 2+ ] c -increase.
Hyper-gravistimulation, achieved by reorienting the specimens 90° under the hypergravity condition (e.g., 5 g), enhances gravitropic curvature in Arabidopsis hypocotyls and roots 32 , whereas gravistimulation less than 1 g (e.g., 0.39-0.93 g), created by centrifugation in space, reduced gravitropic curvature in lentil roots 35 . These observations are consistent with the gravity-dependency of the slow and very slow [Ca 2+ ] c -increases in 1-5 g conditions. Thus, plants that have evolved on earth (1 g) may be capable of transducing a wide range of gravitational changes into the gravitropic response by potentiating the amplitude of Ca 2+ -signaling.
Our results suggest that plants use MCA1 for detecting the gravity vector from an upside-down to upright position, i.e., changes from an unfavorable position to their accustomed position. On the other hand, plants may use multiple molecular machineries for sensing gravity vector changes (e.g., upright to upside-down position). This may be the reason why the amplitude of the slow [Ca 2+ ] c -increase was less affected in mca1-knockout seedlings, i.e., multiple MSCCs species may be responsible for the slow [Ca 2+ ] c -increase, and mca1-knockout may only partially affect the slow [Ca 2+ ] c -increase (e.g., 3 g condition).
The characteristics of the [Ca 2+ ] c -increase are consistent with the starch-statolith hypothesis: sedimentation of the high-density plastid, and amyloplast in response to the gravity vector, increase stress in the actin filaments, which may activate mechanosensitive Ca 2+ -permeable channels (e.g., MCA1) 16 . This idea is supported by the experimental result that mechanical stress in the actin cytoskeleton can activate MS channels 36 . A centrifuge microscope revealed sedimentary movements of amyloplasts under hypergravity conditions 33 . In this study backward rotation specifically activated MCA1, leading to the very slow [Ca 2+ ] c -increase, implying that the plant distinguishes the forward and backward rotation. At present the cellular mechanism that distinguishes the forward and backward rotation is not known, however, it might be worth to propose a hypothetical cellular mechanism behind this. The forward rotation will cause the sedimentation of the high-density plastid, which will increase stress in the certain set of actin filaments that is newly encountered with the plastid, and the stress increase in the actin filaments will activate multiple mechanosensitive Ca 2+ -permeable channels including MCA1. On the other hand, the backward rotation will also cause the sedimentation, but this may increase stress in the "different" set of actin filaments that was used to be associated with the plastid under the upright position before gravistimulation and are primarily connected with MCA1, and the stress increase causes the very slow [Ca 2+ ] c -increase.
MCA2 is the only paralog of MCA1 in Arabidopsis. Ca 2+ uptake activity was lower in the roots of mca2null plants than those of wild-type plants, and overlapping function of MCA1 and MCA2 in gravitropism is reported 20 . However, the gravistimulation induced Ca 2+ response in MCA2 knockdown mutant was not examined in this study, because the [Ca 2+ ] c increases are recorded in hypocotyls and petioles of seedlings but not in roots in our experimental condition (coelenterazine-CP does not penetrate the gel substrate).
The physiological role of the very slow [Ca 2+ ] c -increase has not been elucidated at present. The Ca 2+ signals were collected from hypocotyls and petioles as mentioned in our preceding report 14 [37][38][39] . These ideas should be examined in future studies.

Materials and methods
Plant materials, growth conditions, and reconstitution of aequorin. The materials and methods of this study are basically the same as those used in our previous study 14  ] c -dependent aequorin luminescence was monitored and processed using a photon counter (model SUC-100, SCIENTEX Co.) at 1-s intervals. This configuration enabled the detection of gravistimulation-induced aequorin luminescence generated from approximately 50 seedlings. Four sets of this configuration were used for the experiment; each set was mounted in a light-tight dark box and aequorin luminescence was recorded and stored on a computer. This box was vertically rotated 180° by a computer-controlled DC motor system (model BX5120AM-100S, ORIENTAL MOTOR Co.) at an angular velocity of 6 rpm and an angular acceleration of 2.5 rad s −2 , because it induced nearly the maximum [Ca 2+ ] c -increase compared with that by 90° rotation 14 . The plates were mounted near the rotation axis, producing a centrifugal acceleration of ca. 10 -3 g during uniform circular motion; the tangential acceleration at the start of rotation was ca. 10 -2 g. These devices were mounted on a custom-made aluminum rack that was fixed on the bucket of a swing type 5 g centrifuge machine (W3 experimental unit, Japan Aerospace Exploration Agency, Japan, Supplementary Fig. S1). Measurements were made when the experimental setup reached a constant acceleration (2-5 g). The amplitude and kinetic parameters of the fast-transient, slow, and very slow [Ca 2+ ] c -increases were analyzed, as described below. Gravitational acceleration applied to the specimen was monitored using an accelerometer (model CXL04LP3, Crossbow Technology, Inc.) and stored on a computer. The complementation test was conducted as a separate experiment following the WT and mca1 mutant experiments.
Treatment with chemical agents. GdCl 3 (10 mM stock) was added to a final concentration of 100 µM to the aequorin-reconstitution medium in a petri dish 1 h prior to the removal of the medium. Plant growth medium containing coelenterazine-CP and GdCl 3 was removed from the dish 30 min before the experiment 14 to retain the inhibitory action of Gd 3+ during the experiment.
Data analyses. In this paper, the luminescence ratio, rather than the calibrated [Ca 2+ ] c , was used for all analyses, because it was not possible to discharge the remaining aequorin and monitor the signal in the 5 g centrifuge machine. The luminescence ratio was calculated by dividing the aequorin luminescence intensity by the basal luminescence intensity; the average aequorin luminescence intensity before the first gravistimulation at the gravitational acceleration for each experiment (i.e., − 200 to − 50 s in Fig. 1A). The Ca 2+ response showed statistical nature; e.g., the amplitude of the fast transient, the slow and very slow Ca 2+ increases varied from response to response. Data were analyzed using two-tailed Student's t test or one-way ANOVA statistical analysis with the software Origin version 9, which was also used to estimate the decay time constant. The number of observations (n) denotes the number of experiments made with independent samples/plates containing approximately 50 seedlings. All experiments were repeated more than three times, and all the data obtained were analyzed.