Concentration dependent eVect of GsMTx4 on mechanosensitive channels of small conductance in E. coli spheroplasts
Abstract
The spider peptide GsMTx4, at a saturating concentration of 5 µM, is a potent and selective inhibitor of stretch-activated mechanosensitive (MS) channels found in a wide range of eukaryotic cells. Although the peptide’s structure has been resolved, its mechanism of action is still unclear. Due to its amphipathic nature, the peptide is thought to interact with lipids at the periphery of MS channel proteins. In addition to its action on eukaryotic channels, GsMTx4 also displays antimicrobial properties, inhibiting the growth of various bacterial species at concentrations between 5 and 64 µM. Previous investigations of prokaryotic MS channels, which serve as models for studying MS channel gating, have demonstrated that various amphipathic molecules acting at the protein-lipid interface can modulate channel activity. Based on this, the effects of different concentrations of extracellular GsMTx4 on small-conductance MS channels, MscS and MscK, were analyzed in the cytoplasmic membrane of wild-type E. coli spheroplasts using the patch-clamp technique. The results indicate that GsMTx4 exhibits a biphasic effect, where the peptide’s concentration determines whether it inhibits or potentiates the activity of prokaryotic MS channels.
At low concentrations of 2 and 4 µM, GsMTx4 impairs the gating of prokaryotic MS channels, as evidenced by a decrease in pressure sensitivity. In contrast, at higher concentrations of 12 and 20 µM, the peptide facilitates MS channel opening by increasing pressure sensitivity.
Introduction
Mechanosensitive (MS) channels are found across all domains of life, including animals, plants, fungi, and bacteria. These channels play key roles in regulating osmotic balance and cell volume. In eukaryotic cells, mechanical stretch activates stretch-activated channels (SACs), which have been extensively studied in tissues such as the heart, astrocytes, and skeletal and smooth muscles. Prokaryotic cells also possess stretch-activated channels in their cytoplasmic membranes, where they function as receptors for membrane tension and serve as emergency valves for osmoregulation. Although the structures and sequences of these prokaryotic channels differ from those of eukaryotic channels, they share fundamental mechanosensory functions. The evolutionary conservation of MS channels suggests a common mechanism involving the sensing of membrane tension, curvature, or changes in the trans-bilayer pressure profile, which are converted into cellular responses. One unifying model proposes that changes in the lipid bilayer’s pressure profile, especially at the protein-lipid interface, are critical to MS channel gating.
Over recent years, MS channels in prokaryotes have been utilized as models to investigate the general principles of how biological systems detect mechanical and osmotic stimuli. Electrophysiological studies on E. coli membrane patches have identified four distinct MS channel types: MscL (large conductance), MscS (small conductance), MscK (small conductance, K+-dependent), and MscM (mini-conductance). These channels exhibit different thresholds for activation based on membrane tension, with the largest channels requiring the most tension. The properties of MscS and MscK in E. coli spheroplasts have been previously studied, including their gating thresholds, ion selectivity, conductance, and responses to various modulators.
Amphipathic compounds, which possess both hydrophobic and hydrophilic regions, can induce spontaneous channel gating by inserting into the lipid bilayer and causing local stress, even in the absence of mechanical tension. These molecules likely interact with the membrane in a way that alters local forces, triggering conformational changes in channel proteins. On the other hand, compounds that inhibit MS channel activity are relatively rare, and their mechanisms of action are not fully understood.
The discovery and characterization of the peptide GsMTx4 from the venom of the tarantula Grammostola spatulata marked the identification of a specific pharmacological inhibitor for SACs. Subsequent studies have shown that GsMTx4 is both effective and specific in inhibiting these channels. Structurally, GsMTx4 belongs to the inhibitory cysteine knot peptide family and contains three disulfide bonds in its 34-amino-acid sequence, forming a compact structure with a net positive charge. Its amphipathic nature, with a hydrophobic core and surface, suggests a strong affinity for lipids. The peptide’s mechanism of action does not appear to involve a classic lock-and-key interaction, as its enantiomer is equally active, indicating that the mode of inhibition is likely due to interactions with the lipid environment surrounding the channel proteins.
Studies have shown that GsMTx4 localizes near the protein-lipid interface of both native eukaryotic MS channels and synthetic channels like gramicidin. Its presence affects ionic conductance through the channel pore, consistent with the peptide’s charge being located within the Debye length of conducting ions. GsMTx4 preferentially binds to anionic lipids over zwitterionic ones and can insert into the membrane, penetrating up to nearly 1 nm from the center of the lipid bilayer. Effective modulation of MS channel gating by the peptide requires its close proximity to the boundary lipids of the channel protein, implying that the lipid composition around the channel is critical for its interaction. Computational modeling has further suggested that GsMTx4 can alter the local curvature of the membrane by pulling the lipid leaflets closer together upon insertion.
In addition to its use in studying cellular mechanisms and pathologies, GsMTx4 may provide insights into how mechanical tension in membranes is translated into structural changes in proteins. Until now, the peptide’s effect had not been examined in prokaryotic MS channels. While recent findings have shown that GsMTx4 has antimicrobial effects on bacterial growth, the concentrations required for this activity are significantly higher than those that inhibit eukaryotic MS channels. This raises the possibility that GsMTx4’s antimicrobial action might be due to direct modulation of bacterial MS channels.
To explore this, the peptide was tested on MS channels in E. coli spheroplasts. The experiments were conducted in wild-type bacteria to compare the electrophysiological responses of the channels to previously reported antimicrobial effects. Since bacterial membrane protein and lipid composition can influence channel behavior, it was important to carry out the experiments in native membranes, preserving the natural lipid-protein interactions that might be necessary for GsMTx4’s activity.
In our analysis of MscS and MscK channels, we observed that extracellular application of GsMTx4 at low concentrations resulted in inhibition, as evidenced by the increased tension needed to activate the channels. In contrast, higher concentrations of the peptide enhanced channel activation by reducing the required tension. These results support the hypothesis that GsMTx4 functions as an antimicrobial agent by altering the gating behavior of mechanosensitive channels.
Materials and Methods
E. coli Spheroplasts
Experiments were carried out on giant spheroplasts prepared from wild-type E. coli strain AW737 using established procedures. Recordings were obtained from chromosomally encoded channels preserved in their native lipid environment. Data were collected from several independent preparations to ensure reproducibility.
GsMTx4
The purified spider toxin GsMTx4 was obtained using previously described methods.
Electrophysiology
Channel activity was measured on excised inside-out membrane patches from giant E. coli spheroplasts. These patches were bathed in a standard solution containing either 250 mM KCl or NaCl, supplemented with 90 mM MgCl2 and 5 mM HEPES, adjusted to pH 7.2. The standard pipette solution contained 200 mM of the corresponding chloride salt (K+ or Na+), supplemented with 40 mM MgCl2 and 5 mM HEPES, and was also adjusted to pH 7.2. Solutions without GsMTx4 served as the control condition. The peptide was applied through the patch pipette at final concentrations of 0.5, 2, 4, 12, and 20 µM. Formation of a gigaseal between the pipette and the cytoplasmic membrane was achieved by applying a 30 mV pipette voltage and minimal suction. To ensure gigaseal formation, the tip of the pipette was initially free of the peptide.
Recordings were made while applying suction through the pipette at 30 mV. Pressure was increased in 10 mmHg increments, starting from 0 mmHg, and each step was maintained for at least 45 seconds. In KCl solution, pressure was applied until current saturation was observed across the membrane. The first pressure cycle, lasting about 12 minutes, began 20 minutes after seal formation to allow time for the peptide to diffuse to the membrane. If the seal remained stable, a second and third pressure cycle followed, with each cycle lasting approximately 30 minutes, including a 20-minute interval between cycles. In Na+ solution, pressure was applied until full activation of MscS channels was observed. In K+ solution, where saturation occurred, data from the first cycle were used to determine the activation threshold of MscK channels. Activation threshold was defined as the pressure at which full channel opening first occurred. To evaluate single-channel conductance, current-voltage (I–V) relationships were recorded from -50 to +50 mV under various peptide concentrations.
Data were acquired using an Axonpatch amplifier and Digidata 1440 digitizer with pClamp10 software at a sampling rate of 5 kHz and filtered at 1 kHz. Pipettes were pulled from borosilicate glass capillaries to achieve a bubble number between 3.3 and 3.6, corresponding to a resistance of 2.5 to 2.8 MΩ in standard KCl or NaCl solutions. Data analysis was performed using ClampFit10 and Origin10 software.
Open probability (Po) from recordings in saturated KCl solution was fitted to a two-state Boltzmann distribution. Parameters such as the slope factor, half-activation pressure (p1/2), and free energy difference (ΔG0) were calculated and presented as mean ± standard error. Statistical significance between control and peptide-treated groups was assessed using the Student’s t-test.
Results
Properties of E. coli Mechanosensitive Channels
The impact of GsMTx4 was investigated using inside-out membrane patches derived from wild-type E. coli spheroplasts. To distinguish between the small-conductance mechanosensitive channels MscS and MscK, several characteristics were used, including single-channel conductance, potassium sensitivity, pressure threshold for activation, channel abundance, and behavior during sustained pressure such as deactivation or saturation. These features were consistent with earlier descriptions of MscS and MscK. E. coli also possesses other mechanosensitive channels like MscL and MscM. However, MscL requires significantly higher negative pressure for activation and did not contribute to channel activity observed in this study. MscM, due to its low abundance and small conductance, could be differentiated from MscS and MscK.
In experiments using KCl solution, both MscS and MscK exhibited activity. Identification was based on their known conductance levels and distinct activation thresholds. MscK activity was analyzed in a potassium chloride environment, whereas MscS was studied in sodium chloride solution, as it remains responsive even in the absence of potassium.
Single-Channel Conductance in the Presence of GsMTx4
The conductance of individual MscS and MscK channels was measured at various GsMTx4 concentrations. Analysis across a voltage range from -50 to +50 mV showed that GsMTx4 did not significantly affect the conductance of either channel. Conductance values remained similar to those observed in control conditions and previously published results.
Threshold Pressures for Gating Activity of MscS and MscK
MscS channels exhibit inactivation during increasing pressure, making them unsuitable for Boltzmann-based analysis that requires current saturation. Therefore, activation thresholds were determined by identifying the onset pressure for channel opening. At 2 and 4 µM concentrations of GsMTx4, the activation threshold of MscS increased, suggesting inhibition. Interestingly, at a higher concentration of 20 µM, the threshold decreased, indicating sensitization. Moreover, about 25% of membrane patches treated with 2 and 4 µM peptide exhibited no channel activity, a phenomenon not observed under control conditions or at 20 µM GsMTx4.
For MscK in potassium chloride solution, low concentrations of GsMTx4 (0.5 µM) had no observable effect on activation thresholds. However, 2 and 4 µM shifted the threshold to higher pressures, demonstrating inhibition. At 12 and 20 µM concentrations, the activation threshold shifted to lower values, suggesting sensitization. Spontaneous channel opening occurred under control conditions and at 0.5 and 12 µM GsMTx4, with the effect being more prominent at 20 µM. In contrast, no spontaneous activity was observed at 2 and 4 µM.
Boltzmann Analysis of MscK Gating
MscK channel activity increased with higher pressure stimulation until a maximum open probability was reached, which corresponds to complete channel opening. Data were collected over three sequential pressure cycles for Boltzmann analysis. A consistent pattern emerged, where subsequent pressure cycles led to a shift in the pressure-response curve toward lower pressures. This shift indicates a training effect, in which repeated pressure stimulation facilitates easier channel activation over time. The phenomenon was observed across all concentrations of GsMTx4 tested and appears to reflect the time-dependent action of the peptide, which requires several minutes to exert its full effect.
During the second pressure cycle, approximately 60 minutes after forming a membrane seal, Boltzmann distributions were used to compare the open probability of channels under different peptide conditions. At 0.5 µM GsMTx4, channel activation resembled control conditions. At 2 and 4 µM, the pressure-response curve shifted toward higher activation pressures, indicating inhibitory effects. Conversely, at 12 and 20 µM, the curve shifted toward lower pressures, suggesting sensitization. Recording traces at all tested concentrations confirmed these patterns.
Three parameters derived from the Boltzmann analysis were examined to characterize the gating of MscK: the slope (a) of the fitted curve, the half-activation pressure (p1/2), and the free energy difference (ΔG0) between the open and closed states of the channel. These parameters provide insights into the gating behavior under mechanical stress. ΔG0, in this context, reflects the work required to transition the channel from a closed to an open state in response to lateral membrane tension. The equation ΔG = tΔA – ΔG0 relates the membrane tension (t), the difference in membrane area occupied by the channel (ΔA), and the energy required to maintain the channel in an open state.
The slope parameter (a) increased significantly at 2 µM GsMTx4 and decreased at 12 and 20 µM, with no notable effect at 0.5 µM. At 4 µM, the slope remained unchanged during the first two pressure cycles but decreased during the third, similar to the effects seen at higher peptide concentrations. The half-activation pressure (p1/2) followed a similar pattern. It increased at 2 and 4 µM compared to control and decreased only at 20 µM during the first pressure cycle. In the second and third cycles, 0.5, 12, and 20 µM GsMTx4 did not alter p1/2 compared to control.
Changes in ΔG0 revealed the most significant differences among peptide concentrations, clearly demonstrating a biphasic effect. At 0.5 µM, no effect was observed. At 2 and 4 µM, the free energy increased, indicating that gating was more energetically demanding. This effect was observed in all three cycles at 2 µM, and in the first and second cycles at 4 µM. In contrast, at 12 and 20 µM, the free energy decreased across all pressure cycles, indicating that less energy was required to open the channels. This facilitation of channel gating at high peptide concentrations was also associated with difficulty in forming and maintaining a gigaseal, with an estimated 40% seal loss during the first pressure cycle.
A particularly puzzling observation was the diminishing inhibitory effect of 2 and 4 µM GsMTx4 over time. With continued exposure, the Boltzmann parameters began to resemble those recorded at 12 and 20 µM, suggesting that either the training effect or gradual accumulation of the peptide in the membrane mitigated the inhibition seen at low concentrations.
Discussion
Mechanosensitive Channels in Wild-Type E. coli
This study was conducted on native mechanosensitive channels from giant spheroplasts of wild-type E. coli. There is increasing evidence that both the lipid environment and protein abundance significantly affect channel gating behavior. Prior studies with lipid-altered E. coli strains have shown that replacing phosphatidylethanolamine with monoglucosyldiacylglycerol results in altered MS channel ratios. Preliminary patch-clamp data from these mutants revealed that MscS and MscK channels had reduced activity and pressure responsiveness. These findings, although not yet published in full, highlight the critical role of the lipid environment.
Mechanosensitive channel characteristics also vary depending on the protein-to-lipid ratio in reconstituted liposomes, differing significantly from their behavior in live cells. This sensitivity to lipid composition becomes especially important when evaluating the action of a peptide like GsMTx4, which is thought to interact with the lipid bilayer or at the interface between lipids and proteins. Therefore, this study focused on examining the interaction between GsMTx4 and channels in the native membrane of wild-type E. coli, avoiding potential artifacts introduced by altered lipid compositions or engineered MS channel content. Using the wild-type strain also allowed for comparison with previously reported antimicrobial effects of GsMTx4.
Interaction of GsMTx4 with the Lipid Bilayer
Initial evidence for GsMTx4 binding to lipid surfaces originated from its NMR structure, which revealed its amphipathic nature. Subsequent experimental studies confirmed that the peptide can bind to artificial vesicles. Fluorescence-quenching experiments indicated that GsMTx4 inserts roughly 0.9 nm from the center of the lipid bilayer, a behavior similar to other amphipathic peptides such as hanatoxin-1, which modulates potassium channels.
The affinity of GsMTx4 for lipid membranes depends strongly on lipid composition. The peptide shows a much higher preference for anionic lipids like palmitoyloleoylphosphoglycerol compared to zwitterionic lipids such as palmitoyloleoylphosphocholine. This preference likely results from the peptide’s net positive charge, promoting interactions with negatively charged lipid headgroups, combined with hydrophobic interactions within the bilayer.
Although electrostatic interactions play a role in GsMTx4 membrane association, they do not fully explain its binding behavior. Rather, a combination of hydrophobic and electrostatic forces governs its interaction with the membrane. Molecular modeling suggests two possible binding modes: a deep mode where the peptide penetrates close to the membrane center and induces thinning by drawing the bilayer leaflets together, and a shallow mode where it mainly interacts with the outer leaflet, causing membrane bending.
The balance between these binding modes appears to depend on the lipid headgroups and acyl chain composition. This dual mode of interaction may explain the biphasic effects of GsMTx4 observed in this study, where low concentrations inhibit channel gating, and higher concentrations facilitate it. The extent of membrane interaction likely influences these opposing effects.
Mode of Action of GsMTx4 on Mechanosensitive Channel Gating
It has been demonstrated that GsMTx4, a specific inhibitor of eukaryotic stretch-activated channels, also modulates the gating of prokaryotic small conductance mechanosensitive channels in a concentration-dependent manner. Both MscS and MscK channels show biphasic gating responses. At peptide concentrations of 2–4 micromolar applied to the outer leaflet of the bilayer, channel activity is inhibited, whereas at higher concentrations of 12–20 micromolar, channel activity is potentiated.
Previous studies on eukaryotic channels showed that GsMTx4 has no or only minor effects on single channel conductance. Similarly, no significant change in single channel conductance was observed for prokaryotic mechanosensitive channels at any peptide concentration tested, suggesting that GsMTx4 is not a pore blocker. Instead, the data indicate that the peptide acts as a gating modifier, shifting the dose-response curves to the right or left depending on concentration. The simplest explanation is that the peptide modulates the protein-membrane interface without direct binding to the channel protein itself.
Inhibition at low peptide concentrations may occur due to peptide action near the channel in a ‘shallow’ binding mode. During gating, channels undergo conformational changes that thin the protein as it widens. Boundary lipids also thin under membrane tension, coupling gating to hydrophobic mismatches at the protein-lipid interface. The presence of low peptide concentrations at these sites may interfere with this process. A shift from shallow to deep binding mode at higher concentrations could be driven by energetic costs associated with many peptide molecules bound at the membrane surface, which are reduced by translocation of the peptide deeper into the bilayer.
At high concentrations, the peptide facilitates channel opening by effectively pre-stressing the membrane and channels, lowering the energy required for channel transition and causing more spontaneous openings without applied pressure. This likely corresponds to a ‘deep’ binding mode where peptide insertion into the bilayer facilitates transitions between closed and open states. Hydrophobic mismatches between lipids and channel proteins generate energy costs that can be compensated by local bilayer thickness changes, possibly aided by peptide penetration.
GsMTx4’s ability to influence hydrophobic mismatch has also been demonstrated with gramicidin channels. The peptide increases the lifetime of gramicidin channels of various lengths, attributed to modulation of the hydrophobic mismatch enabling proper subunit alignment. This effect requires peptide proximity to the channel, as conductance is affected.
Another explanation for the biphasic, concentration-dependent effects is GsMTx4’s selective affinity for different lipid species. The cytoplasmic membrane of E. coli is heterogeneous in lipid composition, and channels may reside in distinct lipid environments. The peptide may alter gating by differentially interacting with these lipids or by inducing lipid rearrangement near channels. This lipid sorting could underlie phenomena such as the ‘training effect,’ where channel activity potentiates with repeated peptide exposure.
Similar concentration-dependent dual effects have been observed for other compounds acting on the lipid bilayer. For example, submillimolar concentrations of gadolinium ions block both eukaryotic and prokaryotic mechanosensitive channels, while lower micromolar concentrations stimulate channel activity. Gadolinium is thought to affect membrane mechanics via lipid binding rather than directly interacting with channels. Another example is PIP2, which inhibits and stimulates TREK1 channels depending on concentration and leaflet localization, likely through effects on surrounding phospholipids. GsMTx4 also shows leaflet-specific effects on some channels, highlighting that channel gating mechanisms influence responsiveness to such agents.
Currently, ongoing studies aim to elucidate the role of lipids in mechanosensitive channel kinetics by examining peptide effects on purified channels reconstituted into artificial liposomes with varied lipid content.
Antimicrobial Activity of GsMTx4
Previous studies on bacterial growth revealed antimicrobial activity of GsMTx4 against both Gram-negative and Gram-positive bacteria. The mechanism may involve modulation of mechanosensitive channels or alteration of membrane packing. The minimum effective concentration is significantly lower for Gram-positive bacteria compared to Gram-negative species.
Differences between these bacterial groups likely explain this observation. Gram-negative bacteria possess an additional outer membrane above the peptidoglycan layer, potentially sequestering some peptide and requiring higher concentrations to affect the cytoplasmic membrane. Their cytoplasmic membranes also differ in lipid composition. The outer leaflet of E. coli membranes predominantly contains zwitterionic phosphoethanolamine with some anionic phosphoglycerol, while Gram-positive bacteria membranes are rich in phosphoglycerol lipids, to which GsMTx4 has higher affinity.
Because of these interactions, GsMTx4 may serve as a lead compound for developing new antimicrobial agents.
It remains under discussion whether all mechanosensitive channels gate via bilayer tension, which involves changes in lipid bilayer pressure profiles and curvature caused by hydrophobic mismatch at the protein-lipid interface. Studying the effects of complex compounds like GsMTx4 on prokaryotic mechanosensitive channels can enhance understanding of molecular mechanisms underlying mechanosensory transduction.
Acknowledgments
Gratitude is expressed to Dr. Frederick Sachs for critical manuscript review and constructive comments. Dr. Åke Wieslander and Dr. Daniel Daley are thanked for sharing information on lipid mutant E. coli strains. Appreciation is extended to Paul Rohde for technical assistance. This work was supported by a grant from the National Health and Medical Research Council.