(The FASEB Journal. 1999;13:S55-S61.)
© 1999 FASEB
Mechanosensitive channels in bacteria as membrane tension reporters
SERGEI SUKHAREV1
Department of Biology, University of Maryland, College Park, Maryland 20742
1Correspondence: University of Maryland, Department of Biology, Bldg. 144, College Park, MD 20742. E-mail: ss311{at}umail.umd.edu
 |
ABSTRACT
|
|---|
The purpose of this short review is to discuss recent data on the
molecular structure and mechanism of gating of MscL, a mechanosensitive
channel of large conductance from Escherichia coli. MscL is
the first isolated molecule shown to convert mechanical stress of the
membrane into a simple response, the opening of a large aqueous pore.
The functional complex appears to be a stable homo-pentamer of 15-kDa
subunits, the gating transitions in which are driven by stretch forces
conveyed through the lipid bilayer. We have measured the open
probability of MscL and the kinetics of transitions as a function of
membrane tension. The parameters extracted from the single-channel
current recordings and dose-response curves such as the energy
difference between the closed, open, and intermediate conducting
states, and the transition-related changes in protein dimensions
suggest a large conformational rearrangement of the channel complex.
The estimations show that in native conditions MscL openings could be
driven primarily by forces of osmotic nature. The thermodynamic and
spatial parameters reasonably correlate with the available data on the
structure of a single MscL subunit and multimeric organization of the
complex. Combined with the functional analysis of mutations, these data
give grounds to hypotheses on the nature of the channel
mechanosensitivity.—Sukharev, S. Mechanosensitive channels in bacteria
as membranetension reporters.
Key Words: ion channel • membrane stretch • gating • subunit structure
|
INTRODUCTION
|
|---|
IMMEDIATE CELLULAR REACTIONS to mechanical stimuli are
usually manifested as an increase of plasma membrane conductance or
ionic permeability, implying that mechanosensitive ion channels are
likely to be primary receptors for mechanical stress
(1-3)
. Mechanosensitive (MS)2
channels that respond to membrane stretch, deformation, or displacement
with a significant change of open probability
(Po) have been documented in specialized sensory
(1)
and many types of non-sensory cells (2
,
3
). Although some of the MS channels have been
characterized biophysically (2
, 3
) and
pharmacologically (4)
, very few recorded traces have been
attributed to specific molecular entities or genes (5)
. On
the other hand, the cloned channel-like molecules mec-4, mec-6, and
mec-10, which are clearly implicated in normal touch reflexes in
nematode Caenorhabditis elegans (6)
are still
awaiting their electrophysiological examination. Therefore, there are
several reasons why the mechanosensitive channel of large conductance,
MscL, from Escherichia coli attracts our attention.
1) MscL was the first cloned molecule that was shown to
respond to mechanical stimuli by opening a large aqueous pore.
2) MscL is coded by a single 412-bp gene; its functional
complex contains one kind of subunit and comprises a relatively simple
system for exhaustive mutagenesis and structure/function study.
(3)
MscL is relatively stable in detergents; it can be
purified and studied in vitro by functional reconstitution
in liposomes, it responds to membrane tension transmitted via the lipid
bilayer in which it is embedded, and does not require any auxiliary
extramembrane component for gating.
The techniques that allow for patch-clamping of prokaryotes revealed
several types of mechanosensitive channels (Msc's) in the cell
envelopes of gram-negative (7)
as well as gram-positive
bacteria (8)
. Patches from giant E. coli
spheroplasts contain three types of MS channels: a 3-nS MscL
(9)
, a 1-nS MscS (7
, 9
), and
occasionally smaller ~0.3-nS channels called MscM (10)
,
where L, S, and M stand for large, small, and mini, respectively. As
the pressure gradient across the patch membrane increases, MscM
activates first, followed my MscS, whereas MscL, the larger channel,
activates at high pipette pressures often close to the lytic threshold
(10)
.
Extraction of bacterial membranes with octylglucoside followed by gel
filtration and reconstitution into liposomes revealed that MscL and
MscS are functional and represent different molecular entities
(9)
. After several steps of chromatographic fractionation,
reconstitution of individual fractions and multiple patch-clamp assays,
MscL was identified as an approximately 17-kDa protein band associated
with characteristic 3-nS MS channel activities in reconstituted
liposomes (11
, 12
). A typical current
recording of MscL (Fig. 1A
) shows that longer openings are interrupted with
short-lived subconducting states, although longer substates also occur.
View larger version (37K):
[in this window]
[in a new window]
|
Figure 1. Functional characteristics of MscL. A) A typical current
trace of MscL opening events. A membrane patch from giant bacterial
spheroplast was stretched by a pressure gradient of approximately 100
mmHg. The MscS currents are not seen due to desensitization. The
recording buffer contains 0.2 M KCl, 90 mM MgCl2, and 10 mM
CaCl2, 10 mM Hepes, pH 6.0; pipette voltage is +20 mV (see
ref. 7
for technical details). The channel displays incomplete short
openings as well as longer full openings interrupted with short-lived
subconducting states. Long-lived substates occur more frequently at
higher hyperpolarizations (beyond 50 mV). The substates are likely to
reflect the dynamics of a multisubunit complex. B) Images of
a liposome patch at rest (top) and under a 42 mmHg pressure gradient
(bottom). The radius of curvature of the patch is 3.5 µm, which was
used then to calculate the tension. C) The open probability
(Po) of reconstituted MscL as a function of
tension measured on the same patch. The curve was fit by the Boltzmann
distribution.
|
|
|
STRUCTURE
|
|---|
The mscL gene predicts a 136-amino-acid protein with
two highly hydrophobic domains and a hydrophilic carboxy terminus
(11)
. Western blots of separated membrane fractions
indicate the inner membrane localization of MscL in the bacterial cell
(13)
. Using the PhoA-fusion strategy (14)
we
have shown that the two hydrophobic stretches are actually
transmembrane domains. The PhoA activity assayed in strains expressing
seven different MscL-PhoA fusion constructs confirmed that the MscL
polypeptide chain indeed crosses the membrane twice (13)
,
with both termini being cytoplasmic and with the central part of the
protein making a short periplasmic loop (Fig. 2A
). The transmission Fourier transform infrared (FTIR) and
circular dichroism (CD) spectroscopy gave us coherent results that MscL
is approximately 80% 
-helical (15)
. The secondary
structure was found to be highly resistant toward denaturation based on
ellipticity thermal profiles in the range of 25–95°C. Amide
H+/D+ exchange shows that approximately 66% of
the protein is water-accessible, whereas the rest possibly forms a
hydrophobic, water-inaccessible core (in the closed conformation),
partially protected with the lipid bilayer. MscL reconstituted in
oriented lipid bilayers was shown to possess a net transbilayer
orientation using dichroic ratios measured by attenuated total
reflection FTIR (15)
.
View larger version (39K):
[in this window]
[in a new window]
|
Figure 2. The strucuture of MscL and its mode of gating suggested by kinetic and
thermodynamic analyses (21). A) The primary structure and
membrane topology of MscL subunit (from Blount et al., ref. 12
). The
active complex is currently presumed to be a pentamer. B) A
cartoon showing the relative changes of channel dimensions during the
gating. Both the closed and the open conformations are characterized
with the elasticity that is predicted to be higher for the closed
state. When the tension is applied, the channel expands to a certain
extent while it remains closed due to hydrophobic interactions between
the core domains. The subsequent opening leads to a partial relaxation
of internal elastic elements. The energy stored in the initial
deformation drives fast-opening transitions through a series of
subconducting states to the fully open state. The rate-limiting and
most tension-dependent transition is the one between the closed state
and the substate of lowest conductance.
|
|
Evidently, one 15-kDa subunit spanning the membrane twice would be
inadequate to form an aqueous pore of 3 nS in conductance, thus MscL
must be a multimer. Tag-purified functional MscL complexes consist of a
single type of subunit, as judged by sodium dodecyl
sulfate-polyacrylamide gel electrophoresis (SDS-PAGE); then the
question is how many? With the help of molecular tools and antibodies
we have pursued a number of different approaches. We generated
constructs that allowed us to express double and triple tandems of
genetically linked subunits. Patch-clamping of the corresponding
expresser strains revealed functional channels of the same conductance
as wild-type MscL, which were yet much less abundant. These
observations were consistent with the topology of MscL (both termini on
the same side of the membrane) and suggested that the number of
subunits in the functional complex might be even or a multiple of two
and three (13)
. Attempts to cross-link purified MscL
complexes and double tandems with disuccinimidyl suberate (DSS) gave
similar patterns of six bands, therefore the hexameric model was our
first hypothesis (13
, 16
). A recent attempt
of two-dimensional crystallization of tag-purified MscLs reveal
hexagonal lattices of doughnut-shaped particles (17)
,
interpreted as hexamers, but the resolution of unsymmetrized projection
maps achieved in this work seems to be insufficient to draw an
unambiguous conclusion on the number of subunits in the complex.
Recent investigation, however, has indicated that the number of
subunits in the MscL complex is rather five, but the assembly of
functional pentamers is independent of the number of subunit repeats in
the polypeptide precursor, as reported for K+ channels
(18)
. It means that three double tandems may form a forced
hexamer with a functional pentameric core and one extra subunit staying
aside. In mildly solubilized crude membrane extracts native MscL
complexes exist as a homogeneous population of approximately 75-kDa
particles as determined by size-exclusion experiments, which is more
consistent with pentamerization (Sukharev, S. I., Schroeder,
M. J., and McCaslin, D., unpublished results). The products of
MscL covalent cross-linking typically form ladder-like patterns of
bands with incrementally increasing molecular weight on SDS gels
(13)
. A comparison of several cross-linking reagents
allowed us to discern those that make preferentially intra-complex
links from those that tend to bridge subunits from different complexes
as well. DSS, which exhibits a high rate of inter-complex
cross-linking, generates a variety of multimeric products (up to eight
discrete bands plus large-molecular-weight aggregates), with the fifth
band stained most intensely. This ability of DSS to add one or more
subunits "borrowed" from neighbors to a complete MscL complex
biased our initial model (13)
. Other reagents (EDC, DMS,
or DMA) reveal pentamers as the highest degree of cross-linking with no
large intercomplex aggregates. As was mentioned above, the expression
of MscL as genetically engineered double or triple subunit tandems
yields low numbers of functional channels compared with expressed
monomers. When cross-linked in situ, these tandems, as
expected, display a preferential hexameric assembly. The most
convincing bits of information allowing us to discern between 5x and
6x models were gathered from combined tandem expression and
size-exclusion experiments. They showed that mildly solubilized forced
hexamers made from pre-linked dimers elute as a peak of particles
heavier than native MscL complexes, indicating that they contain an
extra subunit and therefore are assembled differently (Sukharev,
S. I., Schroeder, M. J., and McCaslin, D., unpublished
results). A recent report of the crystal structure of a MscL homolog
(see discussion below) confirms the homopentameric structure of MscL.
The observation that MscL complexes are highly uniform suggest that
they must be pre-assembled. This brings another important issue that
MscL assembly is not a dynamic equilibrium of full complexes with
monomers shown to be the case for tension-modulated alamethicin
channels (19)
, but also proposed as the mechanism of
gating for MscL by Häse and colleagues (20)
.
|
ENERGETIC AND SPATIAL PARAMETERS FOR MSCL GATING TRANSITIONS
|
|---|
The patch-clamp technique is highly advantageous because it
permits time-resolved observations of the behavior of a single channel
molecule. When applied to MS channels, however, the technique doesn't
always permit precise control of membrane tension, which is the primary
stimulus for MS channel activation. The unknown parameter in this case
is the patch curvature, which along with the transmembrane pressure
determines the tension acting on the channel. We applied a
high-resolution video-imaging technique first developed by Sokabe and
Sachs (21)
to control for the patch curvature during
electrical recording (22)
. For this purpose, purified
MscLs were reconstituted into phospholipid liposomes that form clean
patches that remain spherical cups and do not change the curvature in
the wide range of pressure gradients (Fig. 1B
). The tension
was calculated according to the law of Laplace from pressure and the
radius of curvature, r. We found that the MscLs'
probability of being open (Po) has a steep
sigmoidal dependence on T with a midpoint of 11.8 dyne/cm
(Fig. 1C
). The maximal slope of the
Po/Pc(T)
dependence was 0.63 dyne/cm per e-fold. We assume that the
mechanosensitivity arises from tension acting on a change of in-plane
area of the channel in the membrane (
A) and therefore the
major tension-dependent term in the free energy is
TA. The experimental curves can be fitted to
the Boltzmann distribution: P =
1/(1+exp(E-TA)/kBT),
which gave the values for the energy difference between the closed and
fully open state in the unstressed membrane E = 18.6
kBT, and for the corresponding channel
expansion A = 6.5 nm2. The internal
cross-section of the MscL pore was independently estimated from channel
conductance as 7 ± 2 nm2 (22)
. The
similarity of changes of internal and external (A)
dimensions indicates that the pore opening directly contributes to the
MscL complex expansion. The numbers for the energy and area change are
large in the molecular scale and their combination makes the MscL
dose-response curve steep. Multiplication of the midpoint tension by
the estimated perimeter of the channel gives a force of 12 µdyne
required to drive the conformational transition. What would be the most
probable physiological source of such a force? Given that the bacterial
cell is a round-ended rod of 2 x 1 µm, the pressure induced by
an osmotic gradient of 20 mM in theory would be sufficient to create
tension strong enough to keep the channel open half of the time. An
open MscL, however, would rapidly dissipate the osmotic gradient in a
small cell because it is permeable to substances as large as a few
kilodaltons (23)
and even small proteins
(24)
. Presumably MscL functions in bacterial cells as a
release valve for osmolytes in the event of a strong hyposmotic shock
(16)
. Cells that happen to be in the thin layers of
liquid, on drying surfaces, or in the foam, and thus may be exposed to
surface forces (surface tension at the air/water interface is 70
dyne/cm) may also deflate themselves via MscL thus reducing the chance
of lysis.
The analysis of current traces by the two-state kinetic model (one
closed and one open state) shows that the rate constant for the forward
(closed-to-open) transition has a 2.5 times steeper tension dependence
than the rate constant for the reverse transition. Thermodynamically it
means that the rate-limiting barrier is positioned closer to the open
state (1)
on the reaction coordinate A,
suggesting that the energetic well for the closed state is wider, and
therefore the mechanical compliance of the channel in the closed
conformation is higher (22)
.
Time-resolved current recordings revealed that MscL, in fact, is
not a binary channel but has at least four conducting states
(22)
. They can be seen as brief shoulders on resolved
raising and falling fronts of current transitions, also as fast
downward deflections from the current level of fully open state (Fig. 1A
). The QuB software (SUNY-Buffalo) allowed us to idealize
these difficult traces and to accomplish a multi-state analysis. It has
been shown that 1) the simplest kinetic scheme
C
S1S2S3O (five-state linear model) has the highest
likelihood compared with any other branched or looped scheme, thus the
channel switches from state to state sequentially; 2) the
transition rates between upper conductance levels are relatively high
and almost independent of tension, whereas the rate-limiting step to
opening is the transition between the closed state and the lowest
conductance substate (CS1). This transition thus involves the
greatest A. When summed over all transitions, the
in-plane area change from closed to fully open was 6 nm2,
agreeing with the value obtained in the two-state analysis. A detailed
thermodynamic analysis of data according to the two- and five-state
models suggested that the channel can be expanded significantly while
remaining fully closed. To reconcile these two conditions, we have to
propose some elastic elements in the channel structure (Fig. 2B
). When the tension deforms the channel to a certain
extent, the internal stress in the complex causes conformational change
(crack) in the core that leads to the first conducting state and then
quickly through the series of intermediate states to the fully open
state. The transitions through intermediate S1, S2, and S3 states are
apparently driven by the energy that is stored in the elastic
deformation and that may explain why they are practically insensitive
to the external tension. However, according to the multistate kinetic
analysis, the last transition (S3O) also has a shallow dependence on
T, thus it must bring a small contribution to the entire
protein expansion (22)
.
|
STRUCTURE-FUNCTION RELATIONSHIPS
|
|---|
The first question of what domains of the protein are involved in
lining the pore is relatively simple. The estimation shows that there
is barely enough protein to line the large MscL pore, thus all
transmembrane domains should be recruited. In the assumption of
pentamerization, 10 parallel transmembrane helices would form a
cylindrical barrel with the internal pore diameter of approximately 2.6
nm. The hexameric assembly (12 parallel helices) would give a
proportionally larger pore of 3.2 nm in diameter (23)
. The
conductance of MscL pore is a linear (non-saturable) function of
specific buffer conductivity (at least up to 2 M KCl), which indicates
a large aqueous pore where the ions behave the same way as in the bulk
solution. Guessing the length of the pore between 3 and 5 nm in
Hille's equation, which links the conductance of a channel with its
geometry and specific conductivity of the electrolyte
(25)
, we obtained pore diameters in the range of 2.9 to
3.5 nm. This makes the pore cross-section slightly larger than
predicted by the 10-helix cylinder model, which means that the pore may
not be exactly cylindrical and/or other parts of the protein such as
the periplasmic loop and terminal domains may contribute to the lining.
It should be mentioned that closed MscL is absolutely non-leaky. This
implies that the closed conformation could be envisioned as a bundle of
tightly packed hydrophobic helixes. If the opening is accompanied by a
6.5-nm2 expansion, then it should involve the filling of
the internal volume with water, whereas the closure should be
associated with the expulsion of up to 33 nm3 of water from
the pore.
Only one mutant out of many studied so far had slightly altered
conductance of the fully open state (26)
and there are no
reports of changed ionic selectivity. But mutations that affect the
channel kinetics have been generated (27)
. These are the
charge reversal of the residue 31 (K31E/D), positioned in the middle of
M1, and a number of mutations of Q56 in the periplasmic loop. Aromatic
substitutions at the position 56 lead to long openings, thereby putting
a higher transition barrier between the open and closed states. But the
role of the loop is probably not limited to setting the duration of
channel openings. Deletion of Q56, which makes the loop slightly
shorter, led to a stretch-insensitive channel phenotype that required
almost twice as much tension to open (27)
. On the other
hand, proteolytic cleavage of the loop leads to channel
hypersensitivity to stretch (28)
. Hypothetically, the
loops may serve as an elastic rim that sets limits to the channel
deformation by external tension.
Soluble proteins or protein complexes are held in well-defined
conformations typically by relatively weak non-covalent bonds and by
interactions with water (29)
. In integral membrane
proteins that expose hydrophobic surfaces to the lipid environment,
leaving slightly more hydrophilic residues in the core
(30)
, interactions with water seem to play a secondary
role. MscL may represent an exception because in the open conformation
the channel lines the surface of a large aqueous pore. Identification
of the type of forces that secure the MscL complex in its closed
conformation and the domains participating in such interactions would
answer the question of what sets the tension sensitivity for MscL
activation. This truly fundamental problem is currently under intensive
investigation. The first diagnostic tests have shown that MscL is
activated at higher pressure in the presence of stabilizing agents
(ammonium sulfate) and opens more easily when a chaotropic salt
(potassium thiocyanate) is added to the bath (31)
. This
suggests that hydrophobic interactions in the complex play a certain
role. Indeed, when the channel opens, some internal hydrophobic
surfaces may become exposed to water and this should bring additional
surface tension that would force the channel to close. Chaotropic or
stabilizing agents would relieve or increase the surface stress,
respectively, thus altering the energy of the open state and therefore
the distribution of channels between the states at a given tension.
This immediately attracts attention to the most hydrophobic parts of
the protein: M1 and M2 domains that may form a hydrophobic lock in the
core of the complex. One may infer that disturbing this part of the
protein by hydrophilic substitutions would increase the channel
sensitivity to tension and/or make the closed channel leaky. As a
consequence, the corresponding mutations may alter the membrane
permeability and therefore become toxic to the host bacterium. In the
recent work by Ou and co-workers (32)
, a random
mutagenesis of MscL combined with screens for toxicity of gene products
revealed a tight cluster of residues especially sensitive to mutations
in the lower portion of the M1 domain. The most severe phenotype
(strongest growth inhibition) arose from hydrophilic or charged
substitutions for G22, V23, G26, and G30. It is important to note that
the named residues and the entire M1 with a conspicuous pattern of
equally spaced glycines is the most conserved region across MscL
homologs from eight distant bacterial species (33)
. All
the four small and generally hydrophobic residues are predicted to be
on the same side of the helix making some kind of a glabrous facet
specially designed to keep helixes in close contact with their
neighbors. The mutations that bring bulky, hydrophilic, or especially
charged side chains to this side of the M1 helix are expected to
destabilize the closed conformation. There was no surprise therefore,
that the electrophysiological examination of these mutants revealed
their significant hypersensitivity to pressure (p), with the leftward
shifts of the Po(p) curves proportional to the
severity of the mutant phenotype as manifested by the inhibition of
growth. Some of the toxic substitutions in M1 clearly change bacterial
cell permeability under varied hyposmotic stress (34)
,
again indicating the presence of channels with inappropriately low
setpoint for pressure activation. Therefore, the analysis of mutations
strongly suggests that hydrophobic and spatial interactions between M1
domains in the core of MscL are the primary factor that stabilizes the
closed conformation of the channel.
While this article was in the process of review, the crystal
structure of the MscL homolog from Mycobacterium
tuberculosis was reported (35)
. The homolog forms a
homopentameric transmembrane barrel. Each subunit consists of two
tilted transmembrane helices (TM1 and TM2) connected with an
extracellular (periplasmic) loop that dips into the pore. The
cytoplasmic carboxy-terminal domains of each subunit are also
-helical and five of them form an additional barrel-like structure
in the cytoplasm. The TM1 helices are in the inside of the channel
barrel forming most of the closed pore lining. The lower portions of
the TM1 helices form a narrow part of the pore (about 2 
wide)
lined mostly with hydrophobic residues (A20, V21, V22), thus forming
the hydrophobic gate or "lock." The channel thus appears to be in
its closed or almost closed conformation. It is peculiar that most of
the aromatic residues on the membrane-facing side of the channel are
close to the cytoplasmic end of the transmembrane barrel. The aromatic
side chains probably provide an efficient anchor to the lipid bilayer
for the channel and assist in transmitting the stretch force to the
gate. The outer radius of the cytoplasmic portion of the channel barrel
[the distance between the channel axis and backbones of the outer
transmembrane helixes (TM2)] is approximately 2 nm, thus the
channels' in-plane area is approximately 12 nm2. Because
the conducting properties of the channel assume no significant ion
interaction with the channel wall (22)
, the fully open
channel must accommodate the pore of 6–8 nm2 in
cross-section, and therefore must expand to the radius of about 2.6 nm.
The structure also predicts that TM1 helix of each subunit interacts
most intensively with the TM2 helix of the adjacent subunit,
topologically permitting a transition from the closed state to a wide
barrel consisting of alternating TM1 and TM2 helices. Because the
equilibrium is strongly biased toward closed conformation, an extensive
molecular modeling would be the most feasible way to predict the
pre-stressed closed and open conformations of MscL.
|
CONCLUDING REMARKS
|
|---|
In spite of the fact that MscL is a prokaryotic channel with as
yet poorly defined function, it comprises one of the most convenient
and tangible systems for studying elemental principles of
mechanosensation. Unlike mechanosensitive channels in specialized
sensory cells that detect low-level signals barely above the thermal
noise (36)
, the MscLs' transition energy is large and the
channel requires significant tension to open it (22)
. This
is consistent with its putative release valve function implying that
the setpoint for activation should be near the lytic tension for the
bilayer. The high transition energy also dictates that at low resting
tension the MscLs' probability of being open is negligibly low
(~10-8), which satisfies the requirement for membrane
integrity for the bacterial energy metabolism.
Accumulation of data about this simple molecular valve has been rapid.
Current directions include comparison of different MscL homologs,
functional evaluation of chimeras, and rigorous quantitative
characterization of the most severe mutations. MscL protein
crystallization reveals the channel in its closed conformation because
the open one appears too unstable. Computational approaches supported
by constraints obtained from functional assays will help to visualize
the open channel.
There are first molecular data on MscS (small mechanosensitive channel
of E. coli) recently identified as the putative 31-kDa
protein YggB (N. Levina and I. Booth, personal communication). It
appears to be structurally unrelated to MscL, porins, or voltage-gated
channels, but has some remote sequence similarity to the family of
two-pore K+ channels, one of which is the mammalian S-like
mechano-gated channel TREK-1 (5)
. It is too early to
discuss molecular mechanisms of their gating but clearly both MscS and
TREK-1 are gated by tension directly transmitted via the lipid bilayer.
Recent identification of an MS channel protein in archaebacteria
(37)
may reveal one more mechanosensitive molecular
design. Combined efforts should bring insights as to what type of
interactions, what domains, structural motifs, and conformations
determine the susceptibility of membrane proteins to physiologically
relevant mechanical stimuli.
|
ACKNOWLEDGMENTS
|
|---|
Parts of this work were supported by NASA Grant NAG5-6526. I
gratefully acknowledge collaboration and helpful discussions with my
colleagues, Drs. C. Kung, P. Blount, B. Martinac, I. Arkin, I. Booth,
W. Sigurdson, and F. Sachs.
|
FOOTNOTES
|
|---|
2 Abbreviations: MS, mechanosensitive;
Po, open probability; Msc's, mechanosensitive
channels; FTIR, Fourier transform infrared spectroscopy; CD, circular
dichroism spectroscopy; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide
gel electrophoresis; DSS, disuccinimidyl suberate.
|
REFERENCES
|
|---|
-
Howard, J., Roberts, W. M., Hudspeth, A. J. (1988) Mechanical transduction by Hair cells. Annu. Rev. Biophys. Chem. 17,99-124[Medline]
-
Sachs, F., Morris, C. (1998) Mechanosensitive ion channels in non specialized cells. Blausetein, M. P. Greger, R. Grunicke, H. Jahn, R. Mendell, L. M. Miyajima, A. Pette, D. Schultz, G. Schweiger, M. eds. Reviews of Physiology and Biochemistry and Pharmacology ,1-78 Springer-Verlag Berlin.
-
Sackin, H. (1995) Mechanosensitive channels. Annu. Rev. Physiol. 57,333-353[Medline]
-
Hamill, O. P., McBride, D. W., Jr (1996) The pharmacology of mechanogated membrane ion channels. Pharmacol. Rev. 48,231-252[Abstract]
-
Patel, A. J., Honore, E., Maingret, F., Lesage, F., Fink, M., Duprat, F., Lazdunski, M. (1998) A mammalian two pore domain mechano-gated S-like K+ channel. EMBO J 17,4283-4290[Medline]
-
Tavernarakis, N., Driscoll, M. (1997) Molecular modelling of mechanotransduction in the nematode Caenorhabditis elegans. Annu. Rev. Physiol. 59,659-689[Medline]
-
Martinac, B., Buechner, M., Delcour, A. H., Adler, J., Kung, C. (1987) Pressure-sensitive ion channel in Escherichia coli. Proc. Natl. Acad. Sci. USA 84,2297-2301[Abstract/Free Full Text]
-
Zoratti, M., Szabo, I. (1991) Stretch-activated composite ion channels in Bacillus subtilis. Biochem. Biophys. Res. Commun. 168,443-450
-
Sukharev, S. I., Martinac, B., Arshavsky, V. Y., Kung, C. (1993) Two types of mechanosensitive channels in the Escherichia coli cell envelope: solubilization and functional reconstitution. Biophys. J. 65,177-183[Abstract/Free Full Text]
-
Berrier, C., Besnard, M., Ajouz, B., Coulombe, A., Ghazi, A. (1996) Multiple mechanosensitive ion channels from Escherichia coli, activated at different thresholds of applied pressure. J. Membr. Biol. 151,175-187[Medline]
-
Sukharev, S. I., Blount, P., Martinac, B., Blattner, F., Kung, C. (1994) A large conductance mechanosensitive channel in E. coli encoded by mscL alone. Nature 368,265-268[Medline]
-
Sukharev, S. I., Martinac, B., Blount, P., Kung, C. (1994) Functional reconstitution as an assay for biochemical isolation of channel proteins: application to the molecular identification of a bacterial mechanosensitive channel. Meth. Enzymol. 6,51-59
-
Blount, P., Sukharev, S. I., Moe, P. C., Schroeder, M. J., Guy, H. R., Kung, C. (1996) Membrane topology and multimeric structure of a mechanosensitive channel protein of Escherichia coli. EMBO J 15,4798-4805[Medline]
-
Boyd, D., Manoil, C., Beckwith, J. (1987) Determinants of membrane protein topology. Proc. Natl. Acad. Sci. USA 84,8525-8529[Abstract/Free Full Text]
-
Arkin, I. T., Sukharev, S., Blount, P., Kung, C., Brunger, A. T. (1998) Helicity, membrane incorporation, orientation and thermal stability of the large conductance mechano-sensitive ion channel from E. coli. Biochim. Biophys. Acta 1369,131-140[Medline]
-
Sukharev, S. I., Blount, P., Martinac, B., Kung, C. (1997) Mechanosensitive channels of Escherichia coli: the MscL gene, protein, and activities. Annu. Rev. Physiol. 59,633-657[Medline]
-
Saint, N., Lacapere, J.-J., Gu, L-Q., Ghazi, A., Martinac, B., Rigaud, J. L. (1998) A hexameric transmembrane pore revealed by two-dimensional crystallization of the large mechanosensitive ion channel (MscL) of Escherichia coli. J. Biol. Chem. 273,14667-14670[Abstract/Free Full Text]
-
McCormack, K., Lin, L., Iverson, L. E., Tanoue, M. A., Sigworth, F. J. (1992) Tandem linkage of Shaker K+ channel subunits does not ensure the stoichiometry of expressed channels. Biophys. J. 63,1406-1411[Abstract/Free Full Text]
-
Opsahl, L. R., Webb, W. W. (1994) Transduction of membrane tension by the ion channel alamethicin. Biophys. J. 66,71-74[Abstract/Free Full Text]
-
Häse, C. C., Minchin, R. F., Kloda, A., Martinac, B. (1997) Cross-linking studies and membrane localization and assembly of radiolabeled large mechanosensitive ion channel (MscL) of Escherichia coli. Biochem. Biophys. Res. Commun. 232,777-782[Medline]
-
Sokabe, M., Sachs, F. (1990) the structure and dynamics of patch-clamped membranes: a study using differential interference contrast light microscopy. J. Cell Biol. 111,599-606[Abstract/Free Full Text]
-
Sukharev, S. I., Sigurdson, W. J., Kung,
C., and Sachs, F. (1999) Energetic and spatial parameters for gating of
the bacterial large conductance mechanosensitive channel,
MscL. J. Gen. Physiol., in press
-
Cruickshank, C., Minchin, R. F., Le Dain, A. C., Martinac, B. (1997) Estimation of the pore size of the large-conductance mechanosensitive ion channel of Escherichia coli. Biophys. J. 73,1925-1931[Abstract/Free Full Text]
-
Ajous, B., Berrier, C., Garrigues, A., Besnard, M., Ghazi, A. (1998) Release of thioredoxin via the mechanosensitive channel MscL during osmotic downshock of Escherichia coli cells. J. Biol. Chem. 273,26670-26674[Abstract/Free Full Text]
-
Hille, B. (1992) Ionic Channels of Excitable Membranes ,296 Sinauer Associates, Inc Sunderland, Mass..
-
Hase, C. C., LeDain, A. C., Martinac, B. (1996) An N-terminal mutant of the large mechanosensitive channel (MscL) with an altered conductance. Biophys. J. 70,A346
-
Blount, P., Sukharev, S. I., Schroeder, M. J., Nagle, S. K., Kung, C. (1996) Single residue substitutions that change the gating properties of a mechanosensitive channel in Escherichia coli. Proc. Natl. Acad. Sci. USA 93,11652-11657[Abstract/Free Full Text]
-
Ajouz, B., Besnard, M., Berrier, C., Martinac, B., Ghazi, A. (1998) Effects of proteolytic treatment on mechanosensitive ion channels of E. coli. Biophys. J. 74,A238
-
Timasheff, S. (1993) The control of protein stability and associacion by weak interactions with water: How do solvents affect these processes. Annu. Rev. Biophys. Biomol. Struct. 22,67-97[Medline]
-
Rees, D. C., DeAntonio, L., Eisenberg, D. (1989) Hydrophobic organization of membrane proteins. Science 245,510-513[Abstract/Free Full Text]
-
Sukharev, S. I., Blount, P., Nagle, S., and
Kung, C. (1995) Experimental evidence that mechanosensitivity of the
large-conductance MS channel (MscL) is determined by hydrophobic
interactions within the channel complex. Biophys. J.68, A132 (abstact)
-
Ou, X., Blount, P., Hoffman, R. J., Kung, C. (1998) One face of a transmembrane helix is crucial in mechanosensitive channel gating. Proc. Natl. Acad. Sci. USA 95,11471-11475[Abstract/Free Full Text]
-
Moe, P. C., Blount, P., Kung, C. (1998) Functional and structural conservation in the mechanosensitive channel MscL implicates elements crucial for mechanosensation. Mol. Microbiol. 28,583-592[Medline]
-
Blount, P., Schroeder, M. J., Kung, C. (1998) Mutations in a bacterial mechanosensitive channel change the cellular response to osmotic stress. J. Biol. Chem. 272,32150-32157[Abstract/Free Full Text]
-
Chang, G., Spencer, R. H., Lee, A. T., Barclay, M. T., Rees, D. C. (1998) Structure of the MscL homolog from Mycobacterium tuberculosis: a gated mechanosensitive ion channel. Science 282,2220-2226[Abstract/Free Full Text]
-
Denk, W., Webb, W. W. (1992) Forward and reverse transition at the limit of sensitivity studied by correlating electrical and mechanical fluctuations in frog saccular hair cells. Hearing Res 60,89-102[Medline]
-
Le Dain, A., Saint, N., Kloda, A., Ghazi, A., Martinac, B. (1998) Mechanosensitive ion channels of the Archaeon Haloferax volcanii. J. Biol. Chem. 273,12116-12119[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
L. V. Nazarenko, I. M. Andreev, A. A. Lyukevich, T. V. Pisareva, and D. A. Los
Calcium release from Synechocystis cells induced by depolarization of the plasma membrane: MscL as an outward Ca2+ channel
Microbiology,
May 1, 2003;
149(5):
1147 - 1153.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
