Expression of Arabidopsis MCA1 enhanced mechanosensitive channel activity in the Xenopus laevis oocyte plasma membrane

Abstract

Higher plants sense and respond to osmotic and mechanical stresses such as turgor, touch, flexure and gravity. Mechanosensitive (MS) channels, directly activated by tension in the cell membrane and cytoskeleton, are supposed to be involved in the cell volume regulation under hypotonic conditions and the sensing of these mechanical stresses based on electrophysiological and pharmacological studies. However, limited progress has been achieved in the molecular identification of plant MS channels. Here, we show that MCA1 (mid1-complementing activity 1; a putative mechanosensitive Ca²⁺-permeable channel in Arabidopsis thaliana) increased MS channel activity in the plasma membrane of Xenopus laevis oocytes. The functional and kinetic properties of MCA1 were examined by using a Xenopus laevis oocytes expression system, which showed that MCA1-dependent MS cation currents were activated by hypo-osmotic shock or by membrane stretch produced by pipette suction. Single-channel analyses suggest that MCA1 encodes a possible MS channel with a conductance of 34 pS.

Introduction

Land plants are exposed to a variety of mechanical stresses such as touch, wind, gravity and osmotic stresses, and electrophysiological and pharmacological approaches have been used to provide evidence that MS channels in plants play an important role in sensing those stresses.^1-4 It has been suggested that mechanosensitive inward rectifier channels which permeate potassium act as a primary receptor of osmotic stress,⁵ and stretch-activated (mechanosensitive) channels influence the volume and turgor regulation of guard cells.^2,6 Under hypotonic conditions, bacterial cells also release small osmolytes to reduce intracellular osmolality via the activation of mechanosensitive channels. Recent studies have identified plant homologs of MscS, an MSchannel of small conductance for cell volume regulation in Escherichia coli.⁷ The Arabidopsis MSLs (mscS-like) belong to a gene family with ten genes that are phylogenetically divisible into two groups.⁸ The expression of MSL3 affects the rate of growth of E. coli mutant lacking mscL and mscS (MJF465) after hypoosmotic shock, implying that MSL3 is endowed with cell volume regulation.⁸

The subcellular localization of GFP-labeled MSL3 and the phenotypic feature of knockout mutants suggest that MSL2 and MSL3 are localized in the plastid envelop and regulate the size and shape of plastids.⁸ However, the physiological function of MSL2 and MSL3, and ion selectivity of these channels are still obscure. Haswell et al. reported that MSL9 and MSL10 are expressed in the plasma membrane of root cells, and are responsible for MS anion channel activity.⁹

Repetitive touch stimulation leads to a delay in flowering and an inhibition of inflorescence elongation in Arabidopsis,⁴ and results in the enhancement of expression of touch-induced genes.¹⁰ Externally applied Ca²⁺ and Ca²⁺ channel antagonists have been shown to affect the expression of those genes,^11,12 and a rapid increase in cytoplasmic free Ca²⁺ concentration ([Ca²⁺]_c) occurred in plants subjected to touch, wind and hypotonic stress.^13,14 These observations strongly suggest that the increase in [Ca²⁺]_c through mechanosensitive cation-permeable channels are a key element for mechanosensing in plants. As another group of genes encoding putative mechanosensitive channels in Arabidopsis, MCA1 (mid1-complementing activity 1) and its paralog, MCA2, were isolated via functional complementation of the conditional lethality of a yeast mid1 (mating pheromone-induced death 1) mutant lacking a Ca²⁺ channel component that is activated by membrane stretch when expressed in mammalian cells.^15,16 The subcellular localization of GFP-labeled-MCA1 and -MCA2 in Arabidopsis root cells has shown that both proteins are localized in the plasma membrane.^15,16 A knockout mutant of MCA1 shows a unique phenotype, incapability of penetration from softer to harder agar medium, suggesting that MCA1 is required for sensing the hardness of soil.¹⁵ An aequorin-based analysis of [Ca²⁺]_c in the seedlings of an MCA1 overexpressing Arabidopsis line has revealed that MCA1 participates in Ca²⁺-influx in response to hypotonic stresses,¹⁵ implying that MCA1 is a crucial component (or regulator) of MS cation-permeable channels in Arabidopsis. Most recently, homologous genes in tobacco and rice, NtMCA1, NtMCA2 and OsMCA1, respectively, were identified; they are localized in the plasma membrane, channel Ca²⁺, and may have potential roles in cell proliferation.^16,17 However, direct evidence for mechanical stress-dependent activation of MCA proteins is still lacking.

Here, we report the electrophysiological properties of MCA1 expressed in the plasma membrane of Xenopus laevis oocytes. The magnitude of transmembrane inward currents in response to hypotonic shock or plasma membrane stretching in Xenopus laevis oocytes were dependent on MCA1 expression, suggesting that MCA1 is a possible MS cation-permeable channel activated by such stresses in the cell membrane of Arabidopsis seedlings.

Results

Expression of MCA1 enhances ionic currents induced by hypotonic stress

The complementary (c)RNA generated from MCA1 was microinjected into defolliculated oocytes, and the electrophysiological characteristics of MCA1-expressing oocytes were examined. The membrane fraction from the oocytes was collected 3 d after microinjection, and the expression of MCA1 was examined by immunoblot analysis (Fig. 1G). The MCA1 protein of appropriate molecular mass (48 kDa)¹⁵ was detected in the membrane fraction prepared from cRNA-injected cells, but not from control (water-injected) cells, suggesting that full-length MCA1 was expressed in the membrane of cRNA-injected oocytes.

Figure 1. Expression and analyses of MCA1-dependent membrane currents in Xenopus laevis oocytes. (A) TEVC recordings of the MCA1-expressing oocyte membrane current to test voltages ranging from +80 mV to -120 mV with -20 mV decrements in control (isotonic) solution (40 mM KCl, 2 mM EGTA, and 5 mM HEPES; pH 7.5 adjusted with KOH; 210 mOsmol/L adjusted with D-mannitol). (B) Current records after the bath solution was replaced with a hypoosmotic solution (40 mM KCl, 2 mM EGTA, and 5 mM HEPES; pH 7.5 adjusted with KOH; 110 mOsmol/L adjusted with D-mannitol). (C) Current records after the bath solution was replaced with the K⁺-depleting hypotonic solution (40 mM TEACl, 2 mM EGTA, and 5 mM HEPES; pH 7.5 adjusted with KOH; 110 mOsmol/L adjusted with D-mannitol). Data from (A-C) were from the same oocyte. The amplitude of currents was measured at the time indicated with asterisks. The current-voltage relationships of whole cell currents of MCA1-expressing oocytes were shown in (D) and of water-injected oocytes in (E). Mean values of the current recorded under isotonic (closed circles), hypotonic conditions (open circles) and K⁺-depleting hypotonic conditions (open triangles) are shown. Data represent means ± SEs (n = 9–14). (F) Hypotonic stress-dependent currents were detected in the presence (open circles; n = 14), but was not in the absence (open triangles; n = 3) of K⁺ in the bath solution. The hypotonic stress-dependent currents are obtained by subtracting the current under the isotonic conditions from that under the hypotonic conditions. Note that the ionic contents were not changed in both isotonic and hypotonic conditions. 40 mM KCl were replaced with 40 mM TEACl to maintain the extracellular Cl^- concentration constant in both the isotonic and hypotonic bath solutions. Data represent means ± SEs. (G) Expression of full-length MCA1 in the oocyte membrane. The microsomal fractions corresponding to the indicated amount of oocytes (ranging 0.25 to 1 oocyte) were subjected to SDS-PAGE. The arrowhead indicates the deduced molecular mass of MCA1.

Whole-cell currents were recorded by the 2-electrode voltage clamp (TEVC) technique to test whether hypotonic stress-induced membrane currents were dependent on MCA1 protein expression. The I-V curve of MCA1-expressing oocytes exhibited a characteristic inward rectification in a control (isotonic, 210 mOsmol/L) condition (Fig. 1A). Upon hypotonic (110 mOsmol/L) challenge, the inward current was enhanced in MCA1-expressing oocytes (Fig. 1B, D), while the current was less enhanced in water-injected oocytes (Fig. 1E). The hypotonic stress-sensitive current showed inward rectification and it reversed polarity at ca. 0 mV (Fig. 1F). Under the isotonic control conditions, the I-V relationships of MCA1-expressing oocytes and of water-injected oocytes were almost the same (Fig. 1D and E). These suggest that MCA1 encodes a MS channel sensitive to hypotonic stress.

Potassium ion depletion of the extracellular bath solution largely attenuated the slope of the I-V relationship; MCA1-expressing oocytes showed no hypotonic stimulation-sensitive inward current in a K⁺-depleted solution (Fig. 1C, D and F). Hypotonic conditions cause cell swelling and membrane stretch, resulting in an activation of MS channels.^19,20 Thus the current induced by hypotonic stimulations is thought to be due to an activation of MCA1 expression-dependent MS channels. It should be noticed that the inward currents were also increased slightly under hypotonic conditions in water-injected oocytes probably due to the activation of endogenous MS channels (Fig. 1E, see below).

Patch-clamp analyses of MCA1-dependent MS channel currents

Inward currents were evoked in the MCA1-expressing oocytes in response to negative pressure in the patch pipette at a holding membrane potential of –30 mV. The pressure-dependent inward current in MCA1-expressing oocytes was almost doubled (Fig. 2B) compared with that of the water-injected oocytes using the giant-patch method (Fig. 2A). Similar results were obtained in oocytes expressing MCA2, a homolog of MCA1.² The alteration in the amplitude of the pressure dependent inward current was not obtained in control oocytes expressing an inward-rectifying K⁺ channel KAT1 of A. thaliana (Fig. 2C).²¹

Figure 2. Stretch-activated multichannel currents measured in cell-attached giant patches from water-injected (A) and MCA1-expressing (B) oocytes. The bath and pipette solutions contained 100 mM KCl, 2 mM EGTA and 5 mM HEPES (pH 7.4 adjusted with KOH, 210 mOsmol/L adjusted with D-mannitol). The current increases with increasing in the extent of suction (gray bars, –20 mmHg; black bar, –40 mmHg) at a holding potential of –30 mV. (C) Mean intensities of the negative pressure-induced current in oocytes expressing MCA1 (n = 16), MCA2 (n = 7), KAT1 (n = 5), and water-injected oocytes (n = 11) under a negative pressure of –40 mmHg. Data represent means ± SEs. The letters a and b under the bars denote a significant difference between the two groups (p < 0.05, the two-tailed Student’s t-test).

Single-channel analyses were made in MCA1-expressing oocytes. First, we examined the endogenous MS channel in water-injected oocytes and observed single channel activities with a conductance of 85 ± 7 pS virtually in all the patches examined (12 patches) of the water-injected oocytes (Fig. 3A), and in 50 out of 71 patches of the MCA1-expressing oocytes (Fig. 3B). Two other types of MS channels, which were never found in the water-injected control, were observed in MCA1-expressing oocytes (Fig. 3C and D); they have relatively small conductances, 34 ± 8 pS (Fig. 3C and E) or 15 ± 9 pS (Fig. 3D and E), and found in 16 and 5 patches, respectively, out of 71 patches. These MS channels were not activated by zero pressure in the pipette, but were activated by negative pressures (-15 to -30 mmHg) as shown in multi-channel giant-patch current recordings (Fig. 2A and B). These results suggest that MCA1 encodes MS channels of small conductance, and are activated only in the presence of stress in the membrane.

Figure 3. Single channel currents recorded from a patch membrane of the MCA1-expressing oocyte. Typical traces demonstrate the gating of MS channels of water-injected (A) and of MCA1-expressing (B–D) oocytes activated under negative pressure (–15 to –30 mmHg) and holding potential at –150 mV. (E) The I–V relationships of single-channel currents shown in panels A (closed circle; n = 12), B (open circle; n = 50), C (open triangle; n = 16), and D (open square; n = 5) are shown. The bath and pipette solutions contained 100 mM KCl, 2 mM EGTA and 5 mM HEPES (pH 7.4 adjusted with KOH, 210 mOsmol/L adjusted with D-mannitol). Data represent means ± SEs.

Discussion

The present study indicates that MCA1 is a potential candidate for a plant MS cation-permeable channel based on the following observations. First, the currents induced by hypotonic stimulation were enhanced more in MCA1-expressing oocytes than in water-injected control oocytes. The currents showed inward rectification, and reversed at 0 mV. Second, patch-clamp recordings of channel currents revealed that MCA1-expression-dependent single channel currents were mechanosensitive, and observed at negative membrane potentials. The extrapolated reversal potential of the channels was also around 0 mV (Fig. 3E), which is consistent with those of the hypotonic stimulation-dependent whole-cell currents in MCA1 expressing oocytes. Hypotonic stimulation-induced inward currents in MCA1-expressing oocytes were suppressed when a permeable cation (K⁺) in bath solution was replaced with the impermeable cation TEA. However, TEA may directly block the MS channels as has been reported in a few cases,^22-24 while it does not in other cases e.g., in Arabidopsis root⁹ and mesophyll²⁵ cells. Experiments should be conducted in the future to text whether quaternary ammonium ions including TEA affect the channels activity in MCA1 expressing cells e.g., under the inside-out configuration.²⁴

Our previous studies clearly demonstrated that MCA1 expressed in Arabidopsis and yeast cells permeates Ca²⁺.¹⁵ The whole-cell currents recorded in Arabidopsis mesophyll protoplasts also demonstrated that the trinitrophenol-induced Ca²⁺ current, which is generally thought to be mechanosensitive, was recorded only in MCA1-overexpressing protoplasts, but not in control ones.¹⁵ Taken together, it is highly possible that MCA1 encodes an MS cation (including K⁺ and Ca²⁺)-permeable channel in the plasma membrane of Arabidopsis cells.

Two types of anion channel activities with different conductances were observed in TaALMT1-overexpressing tobacco cultured cells.²⁶ Adapting this interpretation, MCA1 may also form two types of MS channels with different conductances in MCA1-overexpressing oocytes, implying that an unusual oligomer may be assembled. The channel with a conductance of 34 pS is likely to be the main channel encoded by MCA1 in Arabidopsis because it was detected 3 times more frequently than the channel with 15 pS conductance. Further studies should be conducted to test this idea.

Functional analyses of MCA1 in planta have shown that the primary root of mca1-null seedlings fails to penetrate harder agar from softer agar in the two-phase agar method, while that of wild-type seedlings can penetrate both types of agar.¹⁵ In addition, [Ca²⁺]_c of MCA1-overexpressing seedlings increases more than that of wild-type seedlings in response to hypotonic or trinitrophenol stimulations.¹⁵ These observations are consistent with the notion that MCA1 is a component of an MS cation-permeable channel involved in mechanosensing in Arabidopsis. Electrophysiological characterization of MCA1 as an MS channel presented here will facilitate future studies on the molecular and cellular processes of mechanosensing in plants.

Materials and methods

In vitro transcription and cRNA injection

The MCA1, MCA2, and KAT1 cDNA were subcloned into pGEM-HE at the SmaI restriction site, a vector containing 5′ and 3′ untranslated regions from the Xenopus laevis globin gene, which stabilizes transcripts in oocytes²⁷ for efficient expression in oocytes. The plasmid was linearlized using the unique site NheI, and used as a template for the synthesis of capped cRNA using a Message Machine T7 kit (Ambion) as described previously.²⁸

Female Xenopus laevis were purchased from Nasco (Fort Atkinson, WI, USA) and Hamamatsu Seibutsu Kyouzai (Hamamatsu, Japan). Stage V–VI defolliculated oocytes from Xenopus laevis were isolated and stored at 18°C in modified Barth’s saline solution (88 mM NaCl, 1 mM KCl, 2.4 mM NaHCO₃, 0.82 mM MgSO₄, 0.33 mM Ca(NO₃)₂, 0.41 mM CaCl₂, and 20 mM HEPES; pH 7.4 adjusted with Tris, 210 mOsmol/L) supplemented with 50 µg/ml of gentamycin (Sigma). Oocytes were injected with cRNA diluted in 50 nl of RNase-free water (1 µg/µl) using a microinjection system (General Valve Picospritzer III, Parker Hannifin Corp. or NANOJECT II, Drummond Scientific Corp). Control oocytes were injected with the same volume of water. Oocytes were studied 3 d (TEVC recordings) or 4–8 d (patch-clamp analysis) after injection. Oocytes from at least 3 frogs were used for each experiment.

Immunoblot analysis

Total membrane fractions of oocytes were prepared 3 d after injection of cRNA by a method reported previously.^28,29 Protein was separated by sodium dodecyl sulfate-PAGE (SDS-PAGE) and electroblotted onto a polyvinylidene difluoride (PVDF) membrane. The blotted membrane was incubated with anti-MCA1 antiserum¹⁵ at a 1/1000 dilution. The antibody bound to MCA1 was detected with ECL^TM anti-rabbit IgG conjugated with horseradish peroxidase (HRP) (GE healthcare, Buckinghamshire, UK) at a 1/5000 dilution. The washed membrane was developed with the ECL plus immunodetection system (GE healthcare), and the signal intensities were detected with a Bioimaging analyzer (LAS1000; Fuji Film, Tokyo, Japan).

Electrophysiological measurements

Two-electrode voltage clamp (TEVC) recordings were obtained with MEZ-7200 and CEZ-1200 amplifiers equipped with a SET-1201 step pulse generator (Nihon Kohden, Tokyo, Japan). Data were processed with a LabTrax A/D converter and DataTrax2 analyzing software (World Precision Instruments, Sarasota, FL, USA) and stored in a computer. A bath solution (approximately 1 ml in a chamber) was completely replaced within 5 min with a peristaltic pump (2 ml/ min). When the bath solution was changed, recordings were made after 20 min of perfusion.

Patch-clamp recordings were performed with an EPC-10 patch-clamp amplifier (HEKA Electronics, Lambrecht/Pfalz, Germany). The vitelline membrane was mechanically removed with a pair of sharpened forceps before patch-clamp recordings. Giant patch membranes (8–10 MΩof pipette resistance, Fig. 2) and patch membranes of ordinary sizes (50–70 MΩof pipette resistance, Fig. 3) were used for recordings. Data were processed and analyzed using PULSE/PULSEFIT 8.66 (HEKA Electronics) and IGOR 4.22 (WaveMetrics, Lake Oswego, Oregon, USA.). Pressure was applied to the patch pipette interior through the suction port of the pipette holder using a syringe and was monitored with a pressure monitor (PM 015R; World Precision Instruments, Inc., Sarasota, FL USA). Chemicals were obtained from Sigma unless specified otherwise. All the recordings were obtained at room temperature (22–25°C).