Much of our knowledge of the function of mechanosensitive channels, including MscL, comes from detailed electrophysiology studies where gating of the channel is monitored by sharp differences in the ion flux through a membrane patch [10,21,49,22,23,34]. A small voltage (∼50mV) is applied across a patch of membrane at the tip of a micropipette. As a function of pressure difference, channel opening events are recorded as stochastic changes in patch current by an ammeter with picoamp (pA) sensitivity. This truly amazing single-molecule spectroscopy technique allows the experimenter to adjust the voltage as well as the pressure difference across the membrane as shown in Fig. 1.14.
The pressure difference across the membrane translates into a lateral membrane tension (via the Laplace-Young relation [65]), responsible for gating the mechanosensitive channel. However, there are two serious problems with this method when probing the mechanisms of mechanosensitive channels.
Arguably, the most serious problem is that oftenpressure difference(J/m3) across the membrane is taken to be the input variable of prime importance, when in facttension(J/m2) is the membrane parameter which governs mechanosensitive gating. Pressure difference is linearly related to tension via the radius of curvature of the membrane, hence in principle the fix is straightforward - image the membrane patch (see Fig. 1.14). While certainly not impossible [21,50], the membrane patch can be difficult to image due to its small size and the fact that it is inside the micropipette. A recent study [50] demonstrated the importance of measuring tension in lieu of pressure difference.
It was shown that using the standard methods for creating “identical” micropipettes, the measured characteristics of a channel varied significantly. However, when the membrane patch was imaged and tension used as the principal input variable, the same data collapsed to within a few percent of each other, as shown in Fig.3.3. In general, if one could perfectly control the size and shape of the micropipette tip used for contacting and sealing the membrane patch, all measurements would be related by a single constant (the radius of curvature). However, variations in micropipette shape and size, as well as variations in how the membrane contacts and adheres to the pipette5 tip all lead to potentially large variations in the perceived gating characteristics of the channel. Additionally, it is difficult to compare the wealth of quantitative data coming from electrophysiology studies to theoretical models when pressure difference, instead of tension, is used as the principal input variable. Tension is routinely measured in micropipette aspiration experiments [26], and in fact, single-channel electrophysiology recordings are possible in such a setup [25] using ion channels
5Glass–bilayer adhesion, specifically in the context of electrophysiology, is discussed in detail in Section3.6.
with conductances lower than MscL. Hence, this technique might provide a useful way to apply known membrane tension to reconstituted MscL channels in well characterized membranes, a topic discussed further in Chapter3.
With tension being used as the variable of prime importance, electrophysiology is poised to put the continuum mechanical view to the test, elucidating the role of lipids in ion channel function.
In particular, the elastic properties of many lipids have been measured [26], enabling a careful examination of the dependence of gating energy on lipid carbon chain length. The simple continuum view we set forth here predicts a quadratic dependence of the lipid thickness deformation energy on hydrophobic mismatch, which is directly linked to carbon chain length. This, of course, has implications for both the function of various transmembrane proteins, and comments meaningfully on the ability of bilayer thickness to segregate proteins in biological membranes.
A second class of intriguing experiments concerns the mechanosensitivity of other ion channels and receptors, generally regarded not to be mechanosensitive [66,67]. This is both interesting from a functional standpoint, in an effort to understand the full physiological effects of these proteins, and as a tool for understanding structural features such as the motions of transmembrane helices.
Performing a similar experiment where lipid carbon chain length is varied around a voltage-gated ion channel (for example) could reveal hidden mechanosensitivity, and energetic analysis from such an experiment could comment on the degree of height and area change during the gating transition.
Indeed by comparing the known electrostatic gating energy of∼15kBT for common voltage-gated ion channels, contributions to the gating energy from membrane deformation can be estimated, and used in two ways [68]. First, one can use such estimates to predict the shift in the gating voltage of these channels as a function of bilayer mechanical attributes like leaflet thickness or tension, as shown in Fig.1.13. Second, the way in which these predicted changes in gating voltage scale with various mechanical properties, can serve as a indicator of which deformation modes are dominant during gating, and hence might indicate relevant structural changes in the protein upon gating.
Such analysis has been performed in detail [68], and the results for a typical voltage-gated ion channel are shown in Fig.1.13.
The second problem facing a complete understanding of the function of mechanosensitive chan- nels is that for many such channels volumetric flow, and not ion flux, is the relevant physiological parameter6. Hence, ion flux is used as a surrogate measurable in place of the true physiological output of the channel. One could argue that ion flux is proportional to volumetric flow, however this assumes that the way ions flow through the channel pore is identical to the way water flows through the pore. Experiments have elucidated the roughly Ohmic nature of mechanosensitive channels [22,75] at low voltage (.80mV), however we know essentially nothing about how a pres- sure gradient across the membrane translates into a volumetric flow. Even the simplest continuum approximation (Hagen-Poiseuille flow) would predict a non-linear function relating the area of the
6The mechanosensitive bacterial channel MscS [12, 69, 18] is another example. Although, there are also mechanosensitive channels that appear to be highly ion selective, such as the bacterial mechanosensitive ion channel MscK [70] and the K2P family of mammalian mechanosensitive channels [71,72,73,74].
−50 0 50 100 150
-4 -2 0 2 4
a)
b)
c)
d)
−100
−50 0 50 100 footprint dila!on
bending
compression
footprint dila!on compression
bending
1.2 1.3 1.4 1.5 1.6 1.7
1.2 1.3 1.4 1.5 1.6 1.7
0.1 0.2 0.3 0.4 0.5
0.1 0.2 0.3 0.4 0.5
-4 -2 0 2 4
Figure 1.13: Mechanosensitivity as a structural reporter. These plots apply to the expected shifts in gating voltage of the Shaker family protein Kv1.2. In all plots the solid gray lines represent thickness deformations, the dotted gray lines represent protein area dilation, and the solid black lines represent midplane bending deformations. a) Shift in gating voltage as a function of tension, assuming the closed channel deforms the membrane. b) Shift in gating voltage as a function of tension, assuming the open channel deforms the membrane. c) Shift in gating voltage as a function of leaflet thickness, assuming the closed channel deforms the membrane. d) Shift in gating voltage as a function of leaflet thickness, assuming the open state deforms the membrane. This figure adapted from [68].
~50mV pA
p = p 1 2
~50mV pA
p2
p1
p2
p1 r
a)
b)
c)
0 200 400 600 800 1000
0 0.2 0.4 0.6 0.8 1
Time (ms) I(t)/Imax
0 400 800 1200 1600 2000 0
0.2 0.4 0.6 0.8 1
Time (ms) I(t)/Imax
p < p 2 1
Figure 1.14: Measurement of tension vs. pressure difference in an electrophysiological experiment.
A channel protein (small blue rectangle) is embedded in a membrane patch (green). A potential of order 50mV is applied across the sealed membrane patch, and channel opening events are measured by an ammeter (circle) with picoamp (pA) sensitivity. a) At low pressure difference, the tension in the patch is low, the mechanosensitive channel is in the closed conformation, and the patch has a large radius of curvature. The plot to the right shows normalized channel current as a function of time for a simulated channel; the open state has low occupation at low tension. b) At high pressure difference, the tension in the patch is high, the mechanosensitive channel will occupy the open state, and the radius of curvature (r) is on the order of microns. The plot to the right shows the open state has high occupation at high tension. c) Optical micrograph of vertically oriented membrane patch at low (top) and high (middle and bottom) pressure differences, illustrating the decrease in the radius of curvature with increase pressure difference (from [50]). The scale bar is 5µm.
channel pore to the volumetric flow, in contrast to the linear relationship between ion flux and channel pore area as predicted by Ohm’s Law [76]. It would be of considerable physical and phys- iological interest to expand our understanding of fluid flow at the molecular level, by measuring the relationship between pressure gradient and volumetric flow through a large-pore channel like MscL. This topic will be discussed in detail in Chapter 3.
1.5.1 Other Experimental Clues
In addition to aforementioned analysis of MscL, other avenues of research have shown interesting links between the function of membrane proteins and the lipid environment. One such avenue is the effects of membrane doping (by toxins, lipids or cholesterol) on channel activity, as schematically
shown in Fig.1.5b. Certain lipid species and other membrane components are clearly required for proper protein function [77,78], but studies using toxins support the idea that the membrane is also a generic mechanical medium with which proteins interact. Rather than having evolved to target a specific channel, some toxins impair the function of multiple membrane proteins, and some small molecules, such as capsaicin [79], and peptide toxins, like those found in spider venom [80], target membrane channels across many species. These broad-ranging effects favor a mechanism that targets a generic property of membrane proteins. It has therefore been proposed that these toxins affect the interactions with the membrane itself. But can these toxins be understood in terms of a coarse-grained membrane model?
As discussed earlier, many studies have shown that bilayer thickness, bending stiffness and monolayer spontaneous curvature can affect the function of embedded proteins [81, 28]. Indeed, although the role of certain proteins (such as mechanosensitive channels) is to respond to membrane mechanical stress, in principle this stress can alter the function of any membrane protein. For ex- ample, the dimerization kinetics of the channel-forming peptide gramicidin A can be controlled by externally applied membrane tension, resulting in membrane thinning and decreasing the hydropho- bic mismatch between the membrane and the gramicidin dimer [25]. Furthermore, using gramicidin A enantiomers as sensors for membrane mechanical properties, the small molecule capsaicin has been shown to indirectly target and trigger the pain receptor TRPV1 by decreasing the bending modulus of lipid bilayers in a concentration-dependent manner, that is, not with a certain fixed stoichiometric relation between toxin and channel, but progressively by altering the membranes bending stiffness [79]. Conversely, voltage-dependent sodium channels are inactivated by capsaicin with no significant change to the conductance properties of the channels, but by an alteration of the gating voltage itself, suggesting that even channels that are not mechanically gated may still be sub- ject to the effects of membrane mechanics through alterations of membrane properties [82,67,83].
In addition, it seems that some peptide toxins target multiple types of stretch–activated cation channels, not by changing membrane properties per se but by changing the effective boundary conditions at or near the protein-lipid interface [80]. This is yet another generic method by which membrane mechanics can couple to protein function, as indicated in Fig. 1.5b. In particular, it seems that either enantiomer of a peptide toxin is localized in the membrane close to the channel and shifts its dose-response curve. Altogether, these experiments show that the entire range of membrane mechanical properties, as well as alteration of protein-lipid boundary conditions, can be utilized to affect channel function. If the changes such membrane dopants cause to bilayer mechanical properties can be carefully quantitated, future experiment may be able to utilize these methods to get a better understanding of the role of lipids in membrane protein function.