AtMSL9 and AtMSL10

Abstract

Plant cells, like those of animals and bacteria, are able to sense physical deformation of the plasma membrane. Mechanosensitive (MS) channels are proteins that transduce mechanical force into ion flux, providing a mechanism for the perception of mechanical stimuli such as sound, touch and osmotic pressure. We recently identified AtMSL9 and AtMSL10, two mechanosensitive channels in Arabidopsis thaliana, as molecular candidates for mechanosensing in higher plants.¹ AtMSL9 and AtMSL10 are members of a family of proteins in Arabidopsis that are related to the bacterial MS channel MscS, termed MscS-Like (or MSL).² MscS (Mechanosensitive channel of Small conductance) is one of the best-characterized MS channels, first identified as an electrophysiological activity in the plasma membrane (PM) of giant E. coli spheroplasts.^3,4 Activation of MscS is voltage-independent, but responds directly to tension applied to the membrane and does not require other cellular proteins for this regulation.^5,6 MscS family members are widely distributed throughout bacterial and archaeal genomes, are present in all plant genomes yet examined, and are found in selected fungal genomes.^2,7,8 MscS homolgues have not yet been identified in animals.

We previously showed that in wild type protoplasts from the Arabidopsis root, AtMSL9 and AtMSL10 function cooperatively to provide a characteristic WT activity. In this paper, we further investigate the function of AtMSL9 and AtMSL10. We analyze individual protoplasts and argue that in WT cells AtMSL9 and AtMSL10 can function either in cooperation or independently. We also compare the electrophysiological properties of these two channels with that of their bacterial and algal counterparts, and discuss their possible function in planta.

AtMSL9 and AtMSL10 Might Provide Channel Activity Both Cooperatively and Independently in WT Protoplasts

Using the patch clamp technique under the whole-cell configuration, we showed that WT protoplasts are characterized by a mechanosensitive activity elicited by stretching the plasma membrane by application of a positive pressure through the patch pipette (Fig. 1A). The amplitude of each channel-opening event generated in response to pressure was measured, and for the majority of WT protoplasts we observed a homogenous amplitude distribution with a peak value at ∼10 pA (Haswell et al.,¹ Fig. 2C). However, in an analysis of 38 WT protoplasts, we noticed two additional protoplast subsets. In the first subset (4/38 protoplasts) we observed an amplitude distribution centered on 8.1 pA, as shown in Figure 1B. The second subset (4/38 protoplasts) is presented in Figure 1D, and shows an amplitude distribution centered on 20.0 pA. Figure 1C shows the amplitude distribution of a representative protoplast of the predominant population, centered on 10.1 pA (30/38 protoplasts).

Figure 1.

AtMSL9 and AtMSL10 provide MS activity in cooperation or independently in WT protoplasts. (A) Scheme representing the whole-cell configuration used on WT root protoplasts to record MS activity in response to positive pressure. Bath solution: 50 mM CaCl₂, 5 mM MgCl₂, 10 mM MES-Tris and 0.5 mM LaCl₃ (pH 5.6, 450 mOsm). Pipette solution: 150 mM KCl, 2 mM MgCl₂, 5 mM EGTA, 4.2 mM CaCl₂ and 10 mM Tris-HEPES (pH 7.2, 470 mOsm), supplemented with 5 mM MgATP. (B) Representative WT protoplast characterized by AtMSL9 activity. (C) WT protoplast representative of the caracteristic WT activity. (D) WT protoplast representative of AtMSL10 activity. (E) Multimodal Gaussian fit on the entire WT protoplast population. For all figures, the dotted curves represent the multimodal Gaussian fit curve of the entire WT protoplast population. The solid lines represent unimodal Gaussian fit for each representative protoplast.

Figure 2.

Putative role of MSLs as components of mecano- and osmosensing. Physical deformation (A) of the membrane (touch, obstacle, bending) will locally increase the membrane tension, while for a hypoosmotic shock (B) the whole membrane will be affected. The resulting MSLs activation might trigger secondary signals such as membrane depolarization or calcium influx. In the case of osmotic downshock, an electrolyte efflux should counteract the osmotic stress and participate in the osmotic regulation.

These two additional amplitude distributions (Fig. 1B and D) are very similar to those which were observed in msl9-1 and msl10-1 single mutants. The amplitude distribution of the protoplasts in the first subset, with a peak value at 8.1 pA, is close to that of the msl10-1 mutant protoplasts, which exhibit a peak value of 7.3 pA (Haswell et al.,¹ Fig. 3B). The second subset and the msl9-1 mutant (Haswell et al.,¹ Fig. 3A), also have similar amplitude distributions, with peak values at 20.0 and 19.9 pA, respectively (Haswell et al.,¹ Fig. 3A and B). Transient expression of either AtMSL9 or AtMSL10 in msl9-1; msl10-1 double mutant protoplasts (Haswell et al.,¹ Fig. 3C and D) showed that AtMSL9 and AtMSL10 can function independently in this mutant background. Here, observations of individual WT protoplasts suggest that AtMSL9 and AtMSL10 can also function independently in protoplasts derived from a WT plant.

The fit of the amplitude distribution of the population of 38 WT protoplasts with a 3-peak multimodal Gaussian curve (Fig. 1E redrawn from Haswell et al.,¹) reveals a main peak at 9.9 pA and two smaller peaks at 8.1 pA and 18.3 pA respectively. This new fit gives a more general view of the activity of MS channels in root protoplasts. Thus, the majority of protoplasts assayed under our conditions exhibit a WT MS channel activity, likely resulting from AtMSL9 and AtMSL10 functioning in cooperation, while a few protoplasts exhibit an activity due to either AtMSL9 or AtMSL10, each functioning independently. These different populations of protoplasts may reflect the tissue-dependent expression of the MSL9 and MSL10 genes, as they have overlapping but non-identical expression patterns (Haswell et al.,¹ Fig. 1). We have focused our analysis on cortical parenchyma tissue and it will be interesting to analyze protoplasts from other tissues to determine if there is a change in the ratio of AtMSL9 and AtMSL10 activities.

These results could be explained by a model wherein AtMSL9 and AtMSL10 function both separately in homomeric and combinatorially in a heteromultimeric complex. Currently we are producing AtMSL9 and AtMSL10 proteins in acellular⁹ and heterologous systems; we hope that biochemical and electrophysiological studies on these recombinant proteins will allow us to test our hypothesis.

AtMSL9 and AtMSL10 are Similar, but not Identical, to their Bacterial and Algal Homologues

Our initial investigation indicates that some characteristics of the AtMSL9/10 channels differ from previously characterized MscS homologues. The MscS homologue from algae, MSC1, exhibited a conductance of ∼0.4 nS when expressed in E. coli spheroplasts in KCl 200 mM, MgCl₂ 40 mM, CaCl₂ 10 mM symmetric.¹⁰ The conductance of E. coli MscS is generally given as ∼1 nS, though this value was obtained in 200 to 300 mM KCl symmetric.¹¹ In 100 mM KCl, a single-channel conductance of 0.4 nS was reported for MscS.¹² In contrast, AtMSL9/10, under similar ionic conditions, have much smaller conductances (0.04–0.13 nS). This may explain why MSL9 and MSL10 cDNAs do not complement an mscS-, mscL- bacterial mutant (Haswell ES, unpublished results). While both MscS and MSC1 inactivate after a sustained stimulus at certain voltages,^10,13–15 we did not observe a loss in activity from the WT Arabidopsis channel during sustained pressure up to 20 sec (Haswell et al.,¹ Fig. 2C). Experiments in excised patches allowed us to test a large range of pressure (from 0 mm Hg to membrane rupture), an improvement over the whole cell configuration, where we have characterized only the beginning of the activation. The pressure activation threshold of AtMSL10 in excised patches (45–55 mm Hg, Haswell et al.,¹ Fig. S5) is slightly lower than that of either MscS (50–100 mm Hg) or MSC1 (130 mm Hg) in excised patches.^4,10,16

Has AtMSL9/10 Activity Already been Observed in Plant Membranes?

Since 1988, many different MS channels activities have been observed in plant membranes.^17,18 In Arabidopsis, two MS channels activities have been characterized in mesophyll cells.^19,20 The conductances of these activities are in the same range as those we observed with AtMSL9 and AtMSL10, with long opening times (in the range of seconds), strictly mechanosensitive activation (meaning no activity without pressure) and they also appear primarily permeant to chloride. The channel described in detail by Spalding and Goldsmith¹⁹ is voltage-sensitive and activated by low pressure (∼10 mmHg in excised patch configuration) while AtMSL9/10 and the channel characterized by Qi et al.,²⁰ are activated at the same pressure threshold of about 30 mmHg. Neither of these two last channels are permeant to calcium cations, at least not in large quantities. However two characteristics distinguish the channel described by Qi et al., and AtMSL9/10. First, Qi et al., showed that applying repeated pressure desensitizes the channel; and second, they characterized a channel solely sensitive to positive pressure. AtMSL9 and AtMSL10 are activated both by positive and negative pressure and do not exhibit adaptation.

These observations suggest that the channels we have characterized and attributed to the AtMSL9 and AtMSL10 proteins were not already observed in Arabidopsis. However, as the previously described MS channels share several electrophysiological properties with the AtMSL9/10 activity, they may be provided by other members of the MSL family.

Which Function for AtMSL9 and AtMSL10?

At this step of our research the question still remains: in which function are these two MSL channels involved? Though we see a clear phenotype at the cellular level in msl9-1; msl10-1 mutants (the absence of stretch-activated channel activity), the global physiological role of AtMSL9 and AtMSL10 is still unknown. This is different from what has been observed for E. coli MscS, which, together with the functionally related MS channel MscL, prevents cellular lysis after a hypo-osmotic shock.¹² Preliminary osmotic and salt-stress experiments suggest that MscS-Like proteins may have acquired new functions in plants.

As with most MS channels, the activity of the AtMSL9/10 channel found in WT Arabidopsis root protoplasts is directly regulated by the tension applied to the membrane and therefore is only a consequence of the pressure applied through the pipette (Haswell et al.,¹ Fig. S2). In the context of the whole organism, there are two main ways to rapidly change the plasma membrane cell tension (Fig. 2). First, applying pressure to the surface of a cell by either touching or bending a plant organ will slightly change the shape of the cell and then modify the membrane tension. Alternatively, a hypoosmotic shock will result in an influx of water into the cell, thereby increasing membrane tension isotropically.

Both touch and osmotic shock are perceived by plants. The past few years have seen a large amount of data on the increase in cytoplasmic free calcium that occurs within a few seconds after exposure to these environmental signals.^21–23 However, the molecular mechanism underlying this calcium increase remains to be discovered. Toyota et al.,²⁴ using Arabidopsis seedlings expressing the Ca²⁺-sensitive luminescent protein apoaequorin, have suggested that the calcium peak induced by gravistimulation occurs via a mechanosensitive Ca²⁺-permeable channel (MSCC). Also the perception of hypertonic and hypotonic treatments was shown by Hayashi et al.,²⁵ to induce a calcium wave. In these both cases, Gd²⁺, an inhibitor of MS channels, was able to abolish the calcium response.

AtMSL9 and AtMSL10 channels have a relatively high conductance and are not likely to transmit large volume of calcium (Haswell et al.,¹ Fig. S3A). Rather, they may act to rapidly depolarize the membrane, thereby activating dedicated Ca²⁺ channels, or they may directly contribute only small increases in cytosolic Ca²⁺. Alternatively, the activation of AtMSL9/10 is likely to dramatically alter the turgor pressure in plant cells, and they may therefore constitute the origin (or the motor) of the wave of depolarization that has been observed in response to wounding and which has been proposed to be “hydrostatically” propagated.²⁶

Addendum to: Haswell ES, Peyronnet R, Brygoo HB, Meyerowitz EM, Frachisse JM. Two MscS homologs provide mechanosensitive channel activities in the Arabidopsis root. Curr Biol. 2008;18:730–734. doi: 10.1016/j.cub.2008.04.039.