THE JOURNAL OF EXPERIMENTAL ZOOLOGY 259~304-315(1991)
Identification of a Stretch-Activated Monovalent Cation Channel From Teleost Urinary Bladder Cells WENHAN CHANG AND CHRISTOPHER A. LORETZ Department of Biological Sciences, State University of New York at Buffalo, Buffalo, New York 14260 ABSTRACT The urinary bladder of euryhaline teleosts is a n important osmoregulatory organ which absorbs N a + , C1-, and water from urine. Using patch clamp technique, single stretchactivated channels, which were permeable to K + and Na+ (PN,/PK= 0.75) and had conductances of 55 and 116 pS, were studied. In excised, inside-out patches which were voltage-clamped in the physiological range of membrane potential, the single-channel open probability (Po)was low (=0.02), and increased to a maximum of 0.9 with applied pipette suction. Single-channel conductance also increased with suction. The channels showed adaptation to applied suction and relaxed to a steady-state activity about 20 seconds after application of suction. The Poincreased u p to 0.9 with strong membrane depolarization (V, = 0 to + 80 mV); however, there was little dependence of Poon membrane potential in the physiological range. The kinetic data suggest t h a t there is one conducting state and at least two non-conducting states of the channel. The open-time constant increased with suction but remained unchanged with membrane potential (V, = - 70 to + 60 mV). The mean closed-time of the channel decreased with suction and membrane depolarization. These results demonstrate the presence of a non-selective monovalent cation channel which may be involved in cell volume regulation in the goby urinary bladder. Additionally, this channel may function as a n enhancer of Na+ influx and K + efflux across the bladder cell as part of transepithelial ion transport if it is located in apical membrane.
Besides functioning as reservoirs for urine, the urinary bladders of some vertebrates function as osmoregulatory organs. From ion flux and electrophysiological studies on fish and amphibians, the urinary bladder has shown its ability to actively absorb Na+, C1-, and H,O from the urine and t o secrete K f into the urine (Dawson and Frizzell, '89; Demarest, '84;Demarest and Machen, '84; Frizzell et al., '79; Loretz and Bern, '80, '81; Nishimura and Imai '82; Renfro, '75, '78; Stokes, '84, '88). These transport processes are controlled by several endogenous regulatory hormones, and they can be altered as well by exogenous drugs (Fisher and Lockard, '88; Handler, '88; Loretz and Bern, '83; Nishimura, '85). Whereas scientists have described the hormonal and pharmacological effects on urinary bladder functions, one potentially important factor has been ignored, namely, the mechanical effects of filling or tissue tension. The rate of urine production determines the fluid load and the volume of the urinary bladder which, in turn, affect the tension of the epithelium lining involved in ion and water transport. Since urine production in aquatic vertebrates varies greatly with environmental salinity (Bentley, '71), bladder volume may regulate bladder ion ab0 1991 WILEY-LISS, INC.
sorption at a local level. An increasing number of reports have shown the involvement of mechanical forces in controlling the biological functions performed by various tissues (Morris, '90; Sachs, '86). Stretch-dependence of membrane ion transport processes has been implicated in systems as diverse as cell volume regulation (Christensen, '87), hair cell mechanotransduction (Howard et al., '88; Ohmori, '88), initiation of smooth muscle contraction (Kirber et al., '88), and mechanotransduction in vascular endothelial cell (Lansman et al., '87). To respond t o the mechanical stimulus of bladder filling, the tissue must have a mechanism t o transduce the stimulus and change functional state. In this report, we describe a stretch-activated channel (SA channel) from single columnar urinary bladder cells; these cells are the location of Na', ClV, and water reabsorption in the goby (Loretz and Bern, '83). In addition t o a description of the effects of membrane tension on channel activity measured using standard patch clamp methodology, we present several biophysical feaReceived J u n e 1, 1990; revision accepted March 18, 1991
STRETCH-ACTIVATED CATION CHANNEL FROM FISH
tures of the channel, i.e., the single-channel conductance, the ionic selectivity and voltagedependence of channels, and some kinetic data. We also compare this SA channel with SA channels from other tissues and finally, we propose possible physiological roles for this SA channel as a mechanical transducer in urinary bladder.
305
bottom of the dish were columnar epithelial cells. Another 10% of the cells were blood cells which could be easily identified by their smaller size. No muscle cells were seen in our cell preparation. Dishes with cells were kept cool and used within 4-6 hours of preparation.
Solutions Table 1shows the basic composition of the solutions used in these experiments. In the text, soluAll experiments were performed on dissociated tions are referred t o by the abbreviations in Table cells from the columnar cell area of the urinary 1. A KC1-rich pipette solution (K-ES) was used bladder in the euryhaline goby, Gillichthys miru- in some experiments but in others, t o facilitate b i l k Gobies weighing 20-50 g were obtained observation and recording of monovalent ion curfrom commercial supplies in California and were rents through SA channels without interference fully adapted in the laboratory to Instant Ocean by currents from frequent K +-selective channels, artificial seawater (Aquarium Systems, Mentor, a pipette solution with a high Na+ concentration OH) at a salinity of 34%0.Fish were maintained (Na electrode solution [Na-ES], 140 mM Na') under a constant 12L:12D photoperiod at 12°C. was used. Unless noted otherwise, we normally Experiments were conducted at room tempera- used Na-ES as pipette solution. With Na-ES in ture (18-20°C). The columnar cell area on the the pipette, the formulation of the G35MES bath mucosal side of the urinary bladder was first solution was such as to yield unique calculated stripped from the opened bladder by using a thin, reversal potentials for the major ionic species, curved piece of polished glass rod, and then disso- specifically: Na' , + 35 mV; K + , a large undefined ciated with collagenase (0.8 mg/ml) in chilled, di- negative potential; C1-, -35 mV; and Ca2+, a valent ion-free Gillichthys bicarbonate-buffered large undefined positive potential. In the experiRinger solution (GBR + 5 mM ethyldiaminete- ments examining the C1 permeability of the chantraacetic acid [EDTAI, composition below) for 35 nel, we first used Na-ES in both pipette and bath minutes. Dissociated cells were collected by gen- medium, and then replaced the Na-ES in the bath tle centrifugation (500g, 15 minutes); the pellet with Na2S0,-BS. In the experiments examining was resuspended in fresh GBR and plated onto the Ca permeability of the channel, again we first 35 mm polystyrene culture dishes. From scanning used Na-ES in both pipette and bath medium, and and transmission electron microscopy studies, we then we added 9 mM CaC1, into the bath t o generfound more than 90% of the cells attaching to the ate a 10-fold Ca chemical gradient across the
MATERIALS AND METHODS Dissociated urinarg bladder cell preparation
TABLE 1 . Composition (mM) of solutions GBR ~
Na * K' Ca2+
Mg2+ c1-
HCO,
so: -
HPO; Isethionate Hepes Mes Sucrose EGTA Glucose PH" ~
~
G35MES ~
161.4 2.5 2.5 1 144.5 5 0.7 20 -
-
5 7.6b
~
35 35 0 35 -
10 35 170 0.1 -
7.6
Na,SO,-BS
.~
~~
~
Na-ES~.
140 1 2 70 10 90 -
140 1 142 10 60 -
-
-
7.6
K-ES ~~
-
140 1 -
142
7.6
apH of all solutions was adjusted by 1 M HC1, 1 M KOH, or 1 M NaOH as approprlate bSolution gassed with 99% 0,-1%CO,
-
10 -
60 -
7.6
W. CHANG AND C.A. LORETZ
306
sure gradient (AP) of -8 cm H,O. A negative pressure gradient is defined as a pressure gradient from cytoplasmic side to extracellular side of the membrane patch. In order to study the dependence of single-channel current and conductance on suction, pipette potential was sequentially Patch clamp apparatus held at - 40, - 50, and - 70 mV, and under each The patch clamp system used in our laboratory pipette potential we increased the syringe suction has been fully described (Loretz and Fourtner, from 0 ml-0.4 ml in 0.1 ml steps and then re'88). The only modification t o our published sys- leased to zero suction. From the above manipulatem was the installation of a 1 ml syringe and a tions, single-channel conductance at a given pimanometer to the suction tubing of the pipette pette suction can be calculated from the 1,-V, holder for the repeated application of known plot. suction. D a t a analysis and statistics Because the application of suction often moved the electrode off the cell membrane, we collected For analysis, recorded data were filtered to only one complete set of on-cell recordings. For yield an effective bandwidth (corner frequency, f,) the on-cell configuration, the membrane potential of 1.5 kHz ( - 3 db, &pole low-pass Bessel filter). of the patch (V,) is equal to the resting mem- Unless specified, the filtered data were digitized brane potential of the cell offset by an applied at 10 kHz and stored on computer hard disk. The voltage (V,) from the clamp circuit. All other digitized data were further analyzed by the membrane patches in this study were excised IPROC-2 computer program (Axon Instruments, inside-out configuration. For these patches, V, Inc., Burlingame, CA) which is an automated was equal t o V,, both reported relative t o the pi- event detection program designed for the analysis pette solution. Inward ( ) current refers to the of single ion channels. By this program, the flow of cations from the extracellular side to the single-channel current (I,) was calculated as the cytoplasmic side of the membrane patch, i.e., from mean current during validated channel openings. the pipette solution into the bath solution. Exci- For membrane patches with only a single chansion of the patch from the cell sometimes resulted nel, the single-channel open probability (Po)was in vesicle formation in the pipette tip; brief expo- calculated from the total current amplitude histosure of the pipette to air generally restored ob- gram with areas under the peaks taken t o be proservable channel activity. In some membrane portional to the time spent in those conductance patches, initial application of suction activated states. For the membrane patch with multiple SA large numbers of channels (more than 5 or 6) channels (as seen in Fig. la), we first.determined which proved too difficult for analysis of single- the number of the channels. By applying suction channel properties. We typically excluded these to the pipette or depolarizing the membrane membrane patches from further analysis. Only patch, either of which would elicit high channel the channel currents which could be activated re- activity, the maximum number of the channels in versibly by suction were recorded and analyzed. the membrane patch can be determined. Under Data collection from patches generally began fol- the assumption that channels in the membrane lowing the appearance of channel activity ac- patch were independent of each other, we calcutivated by brief test application of suction. We lated the Po from the binomial distribution of the recorded spontaneous channel activity, when time spent in the current levels corresponding t o present, at various holding potentials in bath so- different numbers of open channels. Singlelutions for the determination of current-voltage channel conductance (g,) was calculated as the (Ic-Vm)relationships before any other manipula- slope of plots relating I, to V,; regression analysis tions. was applied t o the linear portion of the graph which included the physiological range of memSuction application brane potentials. The plots of I, vs. V, were typiTo study the stretch-dependence of channel ac- cally linear over a wide range (V, = 0 t o -90 tivity, we applied suction t o the membrane patch mV). Since it was difficult to discriminate small simply by retracting the plunger of a 1 ml plastic currents from background current noise, the exsyringe; we calibrated the syringe system by ma- act membrane potential at which no current was nometry. Each 0.1 ml of suction generated a pres- conducted by the channel could not be determembrane patch. Osmotic pressures of all solutions were routinely monitored and never varied from 320 mosmol/l by more than 5 mosmol/l; where necessary, osmotic pressure was adjusted by mannitol or sucrose addition.
+
STRETCH-ACTIVATED CATION CHANNEL FROM FISH
a
307
$1
Suction (AP, cm H 2 0 )
200 msec
1.o
b 0.8
0
0.6
n 0.4 0.2 0.0
0
-1 0
- 20
-30
-40
Suction (AP, cm H20 )
4
- 2 4 crn
4
-32 crn
JI
0 cm
Fig. 1. Suction dependence of channel activity in an excised, inside-out membrane patch and an on-cell membrane patch. a: The current records from an excised, inside-out patch containing 4 SA channels. For this patch, the pipette contained Na-ES and the bath contained GBR; V, was held at - 90 mV. Suction was applied in sequentially increasing steps. The “C” represents the current state when all channels were closed. b The relationship of average Po,in 6 inside-out patches with Na-ES in pipette and GBR in bath, to either
sequentially or non-sequentially increased suctions (pooled data). c: The current records from an on-cell patch. The upward arrows indicate applications of suctions and the downward arrow indicates the release of suction. V, was held at -80 mV; V, was probably near - 130 mV, i.e., the membrane patch was strongly hyperpolarized. I, for the channels in this on-cell recording membrane patch was about 10 PA. The baseline noise was due to the activity of a small spontaneously active channel. (f, = 1.5 kHz)
mined. Therefore, we determined the reversal potential (V,) by the extrapolation of the linear portion of the Ic-Vm plot. The channel showed slow adaptation to applied suction over 20-30 seconds (see Results section for details). We analyzed and reported the first 20-30 seconds of recorded data
directly after the application of suction as the peak response. Adaptation in channel activity was assessed by determining Po for short intervals of time following the application of suction. Powas calculated for single-channel current records 10 seconds in duration, each segment begin-
W.CHANG A N D C.A. LORETZ
308
ning at 2 second intervals throughout the current record. These overlapping sample periods smoothed the data for graphical presentation; data were not combined over transitions in suction. In the study of long-term channel adaptation to applied suction, average Pofor non-overlapping 10 second current records were calculated and plotted. The data were reported as mean i SEM. Statistical significance of differences was assessed using t tests (Dixon and Massey, 1965).
levels of channel conductances. When bathed in GBR, the channels we observed had slope conduc3 pS (n = 4) and 116 i 11 pS tances of 55 (n = 2), with Na-ES in the pipette, and had slope conductances of 70 i 10 pS (n = 2) and 145 i 13 pS (n = 5), with K-ES in the pipette. With G35MES in the bath and Na-ES in the pipette, the channels had slope conductances of 57 5 6 pS (n = 4)and 92 pS (n = 1).Although SA channels exhibited two classes of conductances, they showed no distinguishable difference in suction sensitivity, ion selectivity, voltage-dependence, RESULTS and other biophysical features. Figure 1 illustrates the suction sensitivity of Figure 3 shows the voltage-dependence of SA the channel. From the studies of 17 inside-out ex- channel activity in the absence of membrane cised patches, we found - 8 cm H20 AP was insuf- stretch. Po increased when the membrane was ficient t o induce channel activity. When suction strongly depolarized. Over the physiological equal to or greater than -16 cm H20 AP was range of membrane potential, however, channel applied, channel activity increased with AP (Fig. activity was low in the absence of suction and la,b). Application of - 32 cm H,O AP increased Po independent of membrane potential. In the abup to about 0.9 in some patches. We also observed sence of cytoplasmic-side Ca2+, i.e., the memstretch-dependence of channel activity in one on- brane patch bathed in G35MES, activation of cell patch (Fig. lc). When the cell membrane was channel activity by depolarization was shifted by strongly hyperpolarized, Powas very low and suc- approximately + 20 mV. tion of -24 cm H 2 0 A P slowly induced slight In Figure 3a, when V, was clamped at -50 channel activity. Application of -32 cm H 2 0 A P mV and -80 mV, the channel conducted inward suction activated many channels in the mem- currents, suggesting the possible Na+, C1-, and brane patch after a short delay. After the release Ca2+ permeability of this SA channel. In Figure of suction, channel activity slowly returned t o the 4a, with Na-ES in both pipette and bath, the Vr baseline level. Figure 2 shows that the stretch is approximately zero (0.5 1.2 mV, n = 3). The activation was observed over a physiological replacement of Na-ES in the bath with Na,SO,range of membrane potential, which lies between BS failed t o significantly change the V, (1.8 r - 40 mV and - 70 mV. 1.2, n = 2), suggesting that this channel has very With identical solutions in pipette and bath, low permeability t o C1- ion. A 10-fold increase in individual SA channels expressed either of two Ca2+concentration in the bath also failed to shift the V, (1.0 i 0.8 mV, n = 2). This result suggests low Ca2+ permeability of the SA channel. With 0.5 Na-ES in the pipette and GBR in the bath the V, -40mV (-6.9 ? 2.7 mV, n = 6) was not significantly 0.4 -50 mV different from the theoretical V, for a purely 0 -70 mV Na+-selective channel in GBR ( - 4 mV). The av0.3 erage V, for same inside-out patches bathed in a" the G35MES ( + 14.1 +- 2.2 mV, n = 6) was more 0.2 negative (P< 0.01, two-tailed t test) than the V, for a purely Na+-selective channel in G35MES 0.1 ( + 36 mV). This suggests that the channel is also K + permeable. K + permeability was further con0.0 firmedbyV,(-2.1? 2.1mV,n = 7)ofthechan0 -16 -24 -32 nel current using K-ES in the pipette and GBR in Suction (AP, cm H20) the bath. Using the GHK constant field equation, Fig. 2. Relationship of Poto applied suction for an excised P,,/P, was calculated as about 0.75 for channels inside-out patch when V, was held in the physiological range bathed in the G35MES solution with Na-ES in of membrane potential (-40, -50, and - 7 0 mV). Suction pipette. The calculation of P,,/P, was supported was increased in steps sequentially under each holding potential. by the observation that the conductances of chan_+
_+
~
309
STRETCH-ACTIVATED CATION CHANNEL FROM FISH
a
10
%L
Vm (mV)
I
I
80 msac
-50 5
c
-+-
Na-ES/Na-ES+lOmMCa
--O-
Na-ESINaZS04-BS
- 1 00
0
100
Vm (mV)
C
50
*
lb
I
NaCCESiGBR
+ NaCI-ESIG35MES
10
s v
0.8-
no
I
I
20 I
0 -
0
0.60.4 -
-
1
0
0 100
0
Vm (mV)
0.2-100
- 1 00
-60
-20
20
60
100
Vm (mV) Fig. 3. The channel was voltage gated when the membrane was highly depolarized. a: The current records from an excised, inside-out patch in a pipette filled with Na-ES and bathed in GBR solution. When V, was held a t - 50 and - 80 mV, the channel conducted inward current (upward deflection) with a low single-channel activity. When V, was held at + 30 and + 50 mV, the channel conducted outward current (downward deflection) with a much higher channel activity. The “C” represents the current state when all channels were closed. b: The relationship of average single-channel Po,in 6 excised, inside-out patches with Na-ES in pipette, to V, in GBR and G35MES bath solution. The data points without error bars represent single samples (f, = 1.5 kHz).
nels were greater when K-ES was used in the pipette than the conductances of channels when Na-ES was used. Because of the less defined ion chemical gradients in GBR, we could not calculate a reliable permeability ratio for channels bathed in this solution. Figure 5 presents the effects of suction on I, and g,. In 4 patches, which were held at V, in the
Fig. 4. a: The representative &-V, plots for SA channels in excised, inside-out membrane patches, with either Na-ES o r K-ES in the pipette, and bathed in either GBR (161.4 mM Na+, 2.5 mM K’) or in G35MES (35 mM Na+, 35 mM K + ) . b: The representative Ic-V, plots for SA channel in excised patches, with Na-ES in the pipettes, and bathed in Na-ES, Na,SO,-BS or Na-ES + 10 mM Ca2+.
8
Suction
(AP,cm H20)
*
-
-80
-60
o
Conductance
(ps) 4 1 .O
-16
-40
43.3
-20
Vm (mV) Fig. 5. &-V, plot for a channel in an excised, inside-out patch bathed in G35MES maintained under different levels of suction. Suction was increased sequentially under each holding potential.
310
W.C H A N G A N D C.A. LORETZ
0 range of the physiological membrane potential, both I, and g, in either GBR or G35MES increased Open a with applied suction. Application of - 32 cm H,O AP to some patches increased I, by up to 80%and g, by up t o 40%. In order t o study the kinetics of single SA channels, a set of single-channel data records is v) c needed. Most of the membrane patches we studied C 3 contained more than two channels. Fortunately, we obtained a membrane patch containing only one SA channel and lasting long enough for experimental manipulation. The presence of a single channel in the patch was confirmed by applying - 40 cm H20 AP to the pipette; this level of suction increased Poup t o 0.9. At this high Po, only a single uniform current level for the open 0.0 10.0 20.0 30.0 40.0 50.0 state was observed. Using curve fitting, channel Time (msec) open-time distribution can be fit by one single exponential (Fig. 6a): number of openings = 0 0 e-t’topen, where topen is the open state time constant. 2 The sum of two exponential functions better fit closed-time distributions than a single exponen0 0 tial function (Fig. 6b). These results suggest there 0.3 is one open state and two closed states for this SA channel. The open state time constant, topen, 0 0 increased with applied suction and was indepenc dent of V, (Fig. 7b,e). Because of the infrequent 3 appearance of long closed states of the channel when the membrane was depolarized and when suction was applied, it was difficult t o get reliable closed time constants. Therefore, we have only de0 0 termined the effects of applied suction and V, on fa the mean closed time calculated as the arithmetic mean of those events. The mean closed time de0 creased with the applied suction and with mem0.0 20.0 40.0 60.0 80.0 brane depolarization (Fig. 7c,d). Time (msec) SA channels showed adaptation to applied suction. In Figure 8a, application of sequentiallyFig. 6. a: A sample curve fitting of open-time distribution increasing steps of suction caused increases in by a single exponential function. b: A sample curve fitting of channel activity. For steps in suction t o - 24 and closed-time distribution by the sum of two exponential func-32 cm H,O, increases in channel activity were tions. Both open-time histogram and closed-time histogram abrupt and were followed by slow relaxation to data were sampled from the current record of a single channel in the patch which was voltage clamped a t V, = 60 mV and levels about one-half of those observed immedi- was under no suction condition. ately following application of suction. We also tested the effects of non-sequentially applied steps in suction on channel activity (Fig. 8b). DISCUSSION Channel activity was still dependent on suction when periods of zero-suction were interposed beThe anecdotal observation that the degree of fore and after suction application. The increases urinary bladder filling in the goby depends on the in channel activity were also abrupt and were fol- acclimation salinity (Loretz and Bern ’80) suglowed by slow relaxation. Despite adaptation, the gests that the processing time for the production steady-state level of channel activity was still of the final excreted urine can vary substantially higher than that in the absence of suction (Fig. between freshwater and seawater-adapted fish. In 8c,d). the hypertonic environment, increased urinary
s
Q
STRETCH-ACTIVATED CATION CHANNEL FROM FISH -40mV
0.8
a
d
8
0.4
0.4
0.2
0.0
-50
-70
-40
40
E
0.0
60
0
6-
E
v
4-
-24
-32
2
-4
L
0
-16
-24
-32
4
C
W
a 0
2
c
0
0 -70
0
-16
6
W
?
1
0
b
h
v
-70mV
06
0.2
s
0
08
0.6
8
-50mV
31 1
-50
-40
40
60
I C
E
f
200 150
c
is
100 50
70
-50
-40
40
60
Vrn (rnV)
0
-1 6
-24
-32
Suction (AP, cm H20)
Fig. 7. Panels a, b, and c: Dependence of single-channel kinetics on membrane potential in the absence of suction. Panels d, e, and f: Effects of suction on single-channel kinetics over a physiological range of V,.
bladder filling and residence time in seawateradapted fish promote net NaCl absorption, which subsequently is actively excreted across the gills. Along with NaCl absorption, the water retrieval from the lumen of bladder produces a final urine rich in Mg2+ and SO!- which were passively accumulated across the gut wall as part of the greater drinking activity in seawater-adapted fish. There are multiple modes of control of urinary bladder absorptive transport. In addition to changes in urine delivery and residence time following salinity adaptation, the modification of urine composition is affected by changes in epi-
thelial transport. Hormonal control of urinary bladder ion transport has been documented for the goby (Loretz and Bern, 1983) and in this report, we present our studies on an SA ion channel which may play a role in local regulation of urinary bladder function. Suction-dependence of channel activitu In contrast t o several other types of ion channels seen to be spontaneously active in our patch clamp experiments, we observed one class of channel, which showed activation only with the application of suction. At physiological V,, the spontaneous single-channel open probability (Po)
W.CHANG AND C.A. LORETZ
312
0.2
0.0 20
0
b
0 cml
40
60
I
-24 cm
80
Ocm
I
60
80
100 -32cm
I
120 Ocm
0.4
0
20
40
0.4
100
120
1
100
0
200
300
Time (Sec)
-24 cm
Fig. 8. Adaptation of the goby urinary bladder SA channel to applied suction. a: The effects of sequential applications of increasing suction on Po. b: The effect of non-sequential application of suction on the single-channel Po.c: The effect of non-sequential application of long term suction on the
single-channel Po. The horizontal bar over each panel indicates the suction applied (AP, cm H,O). (Note the change in time axis.) d Partial current records used to generate panel c. Upward arrow indicates the application of suction in cm HZO.
of this channel, in either on-cell or excised, inside-out patches, was low (-0.02) and could be induced by application of pipette suction. When the applied suction was greater than - 16 cm H,O (0.79 x lo4 dyne/cm; Guharay and Sachs, '841, Zn(P,) increased linearly with (AP)' (Fig. 9). This relationship supports the theoretical model proposed by F. Sachs (Guharay and Sachs, '84) that SA channels are activated by energy transduced from membrane tension imposed by the applied pressure difference. The submembrane intracel-
lular cytoskeleton is probably involved in the energy transduction and in modulating the sensitivity of the system. When data from recordings with applied suctions less than -16 cm H,O were included in the analysis, no significant linear relationship between Zn(P,) and (AP)' was observed. This finding is similar t o that observed for SA channels in chick embryonic skeletal muscle. Our interpretation is that the effects of suction are 2-fold: first, application of suction promotes attachment of the membrane patch t o the pipette
313
STRETCH-ACTIVATED CATION CHANNEL FROM FISH
Sachs, '84;Kirber et al., '88; Taniguchi and Guggino, '89). Although this SA channel is not permeable to Ca2+,Ca2+may indeed play a role in channel regulation. In the absence of suction, the mere presence of Ca2+ was insufficient to increase Po in the physiological range of V,; studies of Ca2+ effects on Poin channel under suction will address the possible regulatory role of Ca2+ .
h
0
%
C
Single-channel conductance is dependent on applied suction 0 400 800 1200 In four excised inside-out patches, the singleAP2 (crn' H20) channel current and single-channel conductance Fig. 9. The In(P,) is a linear function of (AP)', when (AP12 increased with applied suction (Fig. 10). We suswas equal to or greater than 256 cm2 H,O (pooled data). pected the increases in current and conductance might be due t o the low cut-off frequency for filglass and, second, further suction increases ten- tering of channel events in our analysis. Theresion on the patch. In studies of 17 membrane fore, we increased f, from 1.5 kHz t o 3.0 kHz and patches, we observed the suction-dependence of increased data sampling frequency from 10 kHz channel activity in a physiological range of mem- to 20 kHz. Despite these alterations, similar debrane potentials. Within this range of membrane pendence of channel current and conductance on potentials the channel activity was low (-0.02) applied suction was observed. We also considered when no suction was applied; while the applica- two other explanations for the increase in conduction of - 32 cm H 2 0 AP induced Poup t o 0.9. This tance with applied suction. One is the ability of result indicates that the SA channel performs its the applied suction to cause a tighter seal bebiological function in the presence of membrane tween membrane patch and pipette glass, which, tension. in turn, decreased the leak current through membrane patch. By reviewing the data record in the Voltage-dependence of channel activitg video tape, no shifting of baseline current suggesThe SA channel reported in this paper showed tive of such an effect occurred when suction was voltage-dependent activation in the absence of applied, ruling out the possibility of decrease in suction only when the membrane was highly de- leak current. Another possibility for the increase polarized (V, between 0 mV and + 80 mV). Simi- in channel conductance is that the applied preslar voltage-dependence was reported in an sure difference on the membrane could create a osteoblast-like cell line (Duncan and Misler, '89) hydraulic pressure, which, in turn, enhanced the and toad smooth muscle (Kirber et al., '88). In the movement of ions and, therefore, current through excitable chick skeletal muscle, the voltagedependence of an SA channel was observed when the membrane was hyperpolarized (Guharay and Sachs, '84). In the physiological range (from 0 mV--90 mV), voltage-dependence was expressed during the application of suction, suggesting that membrane voltage may be a relevant regulator. -4
!
I
The channel is permeable to both K + and N a + ions From the ion substitution experiments, this SA channel shows its selectivity to Naf and K + ,with P,,/PK 7.5, but not to C1- and Ca2+.The general finding for SA channels from other tissues is a relative non-selectivity among cations with K + being more permeable than Na+ in those tissues where selectivity has been tested (Guharay and
0
10
20
30
40
Suction (AP, crn H20) Fig. 10. The dependence of single-channel conductance on applied suction. Average g, was calculated from the channels in 4 patches bathed in GBR with full range of pipette suction steps.
314
W.CHANG AND C.A. LORETZ
the channel (Hille, '84).When we applied suction to the pipette holder, we created a hydrostatic pressure which was oriented from the cytoplasmic side to the extracellular side of membrane patch. If the hydrostatic pressure-induced flow did drag ions through the ion channel, we should have observed a decrease in outward channel current instead of the observed increase. This observation rules out the second possibility. Therefore, we propose that the increases in channel conductance and channel current are due t o the conformational change of channel protein as a result of the change in energy state of cell membrane.
membrane. Although it is too early t o make a conclusion as to the physiological function of this SA channel in the urinary bladder, we do believe the urinary bladder has a mechanical sensor t o monitor and to respond t o the amount of urine production by changing the functional status of the urinary bladder. The discovery of the SA channel in the urinary bladder indicates that the urinary bladder may be more than a urine reservoir or an ion transporting osmoregulatory organ under remote control via hormones, but it may also possess the mechanism for autoregulation of transport.
The channel adapted to the applied suction The SA channels, in the excised, inside-out patches, adapted t o the applied suction in about 20-30 seconds, relaxing to a steady-state level in which the channel activity was still higher than the baseline activity. Considering that the bladder fills slowly, channel activity in situ probably expresses the more modest dependence on stretch suggested by our observations of adaptation. The adaptation of SA channel has also been observed in the yeast plasma membrane (Gustin et al., '88) and in human fibroblasts (Strockbridge and French, '88). The urinary bladder of the goby also undergoes rhythmical contractions, the duration of which are short enough, that they might induce more punctuated increase in channel activation. These results indicate that the responses of the urinary bladder cell t o mechanical forces may involve short term and long term changes in functional status.
ACKNOWLEDGMENTS
Physiological signwcance of the SA channel in urinary bladder Although we can describe the distribution of this channel specifically to the columnar cell of the bladder, we cannot specify the location within the cell. It could be in either the apical membrane or basolateral membrane. If it is in the apical membrane, it can facilitate the Na', C1-, and water reabsorption from the urine by increasing the Na conductance of the apical membrane, which is an intrinsic part of the ion absorptive function of the urinary bladder in goby (Loretz and Bern, '80) and in flounder (Renfro, '75, '78). It might also facilitate K f secretion into urine, a function observed in flounder urinary bladder (Stokes, '88) by increasing K f conductance of the apical membrane. Alternatively, this SA channel may be part of cell volume regulatory mechanism and its cellular location may not be limited to the apical
This study was supported by NSF grant DCB8718633 and BRSG grant 2S07RR0706625. We thank Dr. C.R. Fourtner for valuable discussion.
LITERATURE CITED Bentley, P.J. (1971) Endocrines and Osmoregulation: A Comparative Account of the Regulation of Water and Salt in Vertebrates. Springer-Verlag, New York, pp. 214-218. Christensen, 0. (1987) Mediation of cell volume regulation by Ca+ influx through stretch-activated channels. Nature, 330:66-68. Dawson, D.C., and R.A. Frizzell (1989) Mechanism of active K+ secretion by flounder urinary bladder. Pflugers Arch., 414:393-400. Demarest, J.R. (1984) Ion and water transport by the flounder urinary bladder: Salinity dependence. Am. J . Physiol., 246 (Renal Fluid Electrolyte Physiol., 15):F395-401. Demarest J.R., and T.E. Machen (1984) Passive and active ion transport by the urinary bladder of a euryhaline flounder. Am. J . Physiol., 246 (Renal Fluid Electrolyte Physiol., 15):F402-408. Dixon, W.J., and F.J. Massey (1965) Introduction to statistical analysis. McGraw-Hill, New York, pp. 109-126. Duncan, R., and S. Misler (1989) Voltage-activated and stretch-activated Ba' conducing channels in an osteoblast-like cell line (UMR 106). FEBS Lett. 251:number 1,2, 17-21. Fisher, R.S., and J.W. Lockard (1988) Complex response of epithelial cells to inhibition of Na transport by amiloride. Am. J. Physiol., 254 (Cell Physiol., 23):C297-303. Frizzell, R.A., M. Field, and S.G. Schultz (1979) Sodiumcoupled chloride transport by epithelial tissues. Am. J. Physiol., 236(1) :F 1-8. Guhary F., and F. Sachs (1984) Stretch-activated single ion channel currents in tissue-cultured embryonic chick skeletal muscle. J. Physiol., 352:685-701. Gustin, M.C., X. Zhou, B. Martinac, and C. Kung (1988) A mechanosensitive ion channel in the yeast plasma membrane. Science, 242:762-764. Handler, J.S. (1988) Antidiuretic hormone moves membranes. Am. J . Physiol., 255 (Renal Fluid Electrolyte Physiol., 24):F375-382. Hille, B. (1984) Ion channels of excitable membranes. Sinauer Associates Inc., Massachusetts. Howard, J., M. Robert, and A.J. Hudspeth (1988) Mechanical +
+
STRETCH-ACTIVATED CATION CHANNEL FROM FISH transduction by hair cells. Ann. Rev. Biophys. Chem., 17:99-124. Kirber, M.T., J.V. Walsh, J r . and J.J. Singer (1988) Stretchactivated ion channels in smooth muscle: A mechanism for the initiation of stretch-induced contraction. Pflugers Arch., 412:339-345. Lansman, J.B., T.J. Hallam, and T.J. Rink (1987) Single stretch-activated ion channels in vascular endothelial cells as mechanotransducer? Nature, 325:311-313. Loretz, C.A., and H.A. Bern (1980) Ion transport by the urinary bladder of the gobiid teleost, Gillichthys mirabilis. Am. J. Physiol., 239:R415-423. Loretz, C.A., and H.A. Bern (1981) Stimulation of sodium transport across the teleost urinary bladder by urotensin 11. Gen. Comp. Endocrinol., 43:325-330. Loretz, C.A., and H.A. Bern (1983) Control of ion transport by Gillichthys mirabilis urinary bladder. Am. J. Physiol., 245:R45-52. Loretz, C.A., and Fourtner, C.R. (1988) Functional characterization of a voltage-gated anion channel from teleost fish intestinal epithelium. J. Exp. Biol., 136:383-403. Morris, C.E. (1990) Mechanosensitive ion channels. J . Membr. Biol., 113:93-107. Nishimura, H. (1985) Endocrine control of renal handling of solutes and water invertebrates. Renal Physiol. Biochem., 8:279-300. Nishimura, H., and M. Imai (1982) Control of renal function
315
in freshwater and marine teleosts. Fed. Proc., 41: 2355-2359. Ohmori, H. (1988) Mechano-electrical transduction of the chink hair cell. Prog. Brain Res., 74:ll-20. Renfro, J.L. (1975) Water and ion transport by the urinary bladder of the teleost Pseudopleuronectes americanus. Am. J. Physiol. 228(1):52-61. Renfro, J.L. (1978) Interdependence of the Na+ and C1transport by the isolated urinary bladder of the teleost, Pseudopleurnectes americanus. J . Exp. Zool., 199:181-390. Sachs, F. (1986) Mechanical transduction unification. NIPS, 1:98-100. Stokes, J.B. (1984) Sodium chloride absorption by the urinary bladder of the winter flounder. J. Clin. Invest., 74:7-16. Stokes, J.B. (1988) Passive NaCl transport in the flounder bladder: Predominance of a cellular pathway. Am. J . Physiol., 255 (Renal Fluid Electrolyte Physiol. 24):F229-236. Strockbridge, L.L., and AS. French (1988) Stretch-activated cation channels in human fibroblasts. Biophys. J., 54: 187-190. Taglietti, V., and M. Toselli (1988) A study of stretchactivated channels in the membrane of frog oocytes: Interaction with C a + +ions. J. Physiol., 407:311-328. Taniguchi, J., and W.B. Guggino (1989) Membrane stretch: A physiological stimulator of Ca2' -activated channels in thick ascending limb. Am. J . Physiol. (Renal Fluid Electrolyte Physiol. 26):F347-352.
Identification of a Stretch-Activated Monovalent Cation Channel From Teleost Urinary Bladder Cells WENHAN CHANG AND CHRISTOPHER A. LORETZ Department of Biological Sciences, State University of New York at Buffalo, Buffalo, New York 14260 ABSTRACT The urinary bladder of euryhaline teleosts is a n important osmoregulatory organ which absorbs N a + , C1-, and water from urine. Using patch clamp technique, single stretchactivated channels, which were permeable to K + and Na+ (PN,/PK= 0.75) and had conductances of 55 and 116 pS, were studied. In excised, inside-out patches which were voltage-clamped in the physiological range of membrane potential, the single-channel open probability (Po)was low (=0.02), and increased to a maximum of 0.9 with applied pipette suction. Single-channel conductance also increased with suction. The channels showed adaptation to applied suction and relaxed to a steady-state activity about 20 seconds after application of suction. The Poincreased u p to 0.9 with strong membrane depolarization (V, = 0 to + 80 mV); however, there was little dependence of Poon membrane potential in the physiological range. The kinetic data suggest t h a t there is one conducting state and at least two non-conducting states of the channel. The open-time constant increased with suction but remained unchanged with membrane potential (V, = - 70 to + 60 mV). The mean closed-time of the channel decreased with suction and membrane depolarization. These results demonstrate the presence of a non-selective monovalent cation channel which may be involved in cell volume regulation in the goby urinary bladder. Additionally, this channel may function as a n enhancer of Na+ influx and K + efflux across the bladder cell as part of transepithelial ion transport if it is located in apical membrane.
Besides functioning as reservoirs for urine, the urinary bladders of some vertebrates function as osmoregulatory organs. From ion flux and electrophysiological studies on fish and amphibians, the urinary bladder has shown its ability to actively absorb Na+, C1-, and H,O from the urine and t o secrete K f into the urine (Dawson and Frizzell, '89; Demarest, '84;Demarest and Machen, '84; Frizzell et al., '79; Loretz and Bern, '80, '81; Nishimura and Imai '82; Renfro, '75, '78; Stokes, '84, '88). These transport processes are controlled by several endogenous regulatory hormones, and they can be altered as well by exogenous drugs (Fisher and Lockard, '88; Handler, '88; Loretz and Bern, '83; Nishimura, '85). Whereas scientists have described the hormonal and pharmacological effects on urinary bladder functions, one potentially important factor has been ignored, namely, the mechanical effects of filling or tissue tension. The rate of urine production determines the fluid load and the volume of the urinary bladder which, in turn, affect the tension of the epithelium lining involved in ion and water transport. Since urine production in aquatic vertebrates varies greatly with environmental salinity (Bentley, '71), bladder volume may regulate bladder ion ab0 1991 WILEY-LISS, INC.
sorption at a local level. An increasing number of reports have shown the involvement of mechanical forces in controlling the biological functions performed by various tissues (Morris, '90; Sachs, '86). Stretch-dependence of membrane ion transport processes has been implicated in systems as diverse as cell volume regulation (Christensen, '87), hair cell mechanotransduction (Howard et al., '88; Ohmori, '88), initiation of smooth muscle contraction (Kirber et al., '88), and mechanotransduction in vascular endothelial cell (Lansman et al., '87). To respond t o the mechanical stimulus of bladder filling, the tissue must have a mechanism t o transduce the stimulus and change functional state. In this report, we describe a stretch-activated channel (SA channel) from single columnar urinary bladder cells; these cells are the location of Na', ClV, and water reabsorption in the goby (Loretz and Bern, '83). In addition t o a description of the effects of membrane tension on channel activity measured using standard patch clamp methodology, we present several biophysical feaReceived J u n e 1, 1990; revision accepted March 18, 1991
STRETCH-ACTIVATED CATION CHANNEL FROM FISH
tures of the channel, i.e., the single-channel conductance, the ionic selectivity and voltagedependence of channels, and some kinetic data. We also compare this SA channel with SA channels from other tissues and finally, we propose possible physiological roles for this SA channel as a mechanical transducer in urinary bladder.
305
bottom of the dish were columnar epithelial cells. Another 10% of the cells were blood cells which could be easily identified by their smaller size. No muscle cells were seen in our cell preparation. Dishes with cells were kept cool and used within 4-6 hours of preparation.
Solutions Table 1shows the basic composition of the solutions used in these experiments. In the text, soluAll experiments were performed on dissociated tions are referred t o by the abbreviations in Table cells from the columnar cell area of the urinary 1. A KC1-rich pipette solution (K-ES) was used bladder in the euryhaline goby, Gillichthys miru- in some experiments but in others, t o facilitate b i l k Gobies weighing 20-50 g were obtained observation and recording of monovalent ion curfrom commercial supplies in California and were rents through SA channels without interference fully adapted in the laboratory to Instant Ocean by currents from frequent K +-selective channels, artificial seawater (Aquarium Systems, Mentor, a pipette solution with a high Na+ concentration OH) at a salinity of 34%0.Fish were maintained (Na electrode solution [Na-ES], 140 mM Na') under a constant 12L:12D photoperiod at 12°C. was used. Unless noted otherwise, we normally Experiments were conducted at room tempera- used Na-ES as pipette solution. With Na-ES in ture (18-20°C). The columnar cell area on the the pipette, the formulation of the G35MES bath mucosal side of the urinary bladder was first solution was such as to yield unique calculated stripped from the opened bladder by using a thin, reversal potentials for the major ionic species, curved piece of polished glass rod, and then disso- specifically: Na' , + 35 mV; K + , a large undefined ciated with collagenase (0.8 mg/ml) in chilled, di- negative potential; C1-, -35 mV; and Ca2+, a valent ion-free Gillichthys bicarbonate-buffered large undefined positive potential. In the experiRinger solution (GBR + 5 mM ethyldiaminete- ments examining the C1 permeability of the chantraacetic acid [EDTAI, composition below) for 35 nel, we first used Na-ES in both pipette and bath minutes. Dissociated cells were collected by gen- medium, and then replaced the Na-ES in the bath tle centrifugation (500g, 15 minutes); the pellet with Na2S0,-BS. In the experiments examining was resuspended in fresh GBR and plated onto the Ca permeability of the channel, again we first 35 mm polystyrene culture dishes. From scanning used Na-ES in both pipette and bath medium, and and transmission electron microscopy studies, we then we added 9 mM CaC1, into the bath t o generfound more than 90% of the cells attaching to the ate a 10-fold Ca chemical gradient across the
MATERIALS AND METHODS Dissociated urinarg bladder cell preparation
TABLE 1 . Composition (mM) of solutions GBR ~
Na * K' Ca2+
Mg2+ c1-
HCO,
so: -
HPO; Isethionate Hepes Mes Sucrose EGTA Glucose PH" ~
~
G35MES ~
161.4 2.5 2.5 1 144.5 5 0.7 20 -
-
5 7.6b
~
35 35 0 35 -
10 35 170 0.1 -
7.6
Na,SO,-BS
.~
~~
~
Na-ES~.
140 1 2 70 10 90 -
140 1 142 10 60 -
-
-
7.6
K-ES ~~
-
140 1 -
142
7.6
apH of all solutions was adjusted by 1 M HC1, 1 M KOH, or 1 M NaOH as approprlate bSolution gassed with 99% 0,-1%CO,
-
10 -
60 -
7.6
W. CHANG AND C.A. LORETZ
306
sure gradient (AP) of -8 cm H,O. A negative pressure gradient is defined as a pressure gradient from cytoplasmic side to extracellular side of the membrane patch. In order to study the dependence of single-channel current and conductance on suction, pipette potential was sequentially Patch clamp apparatus held at - 40, - 50, and - 70 mV, and under each The patch clamp system used in our laboratory pipette potential we increased the syringe suction has been fully described (Loretz and Fourtner, from 0 ml-0.4 ml in 0.1 ml steps and then re'88). The only modification t o our published sys- leased to zero suction. From the above manipulatem was the installation of a 1 ml syringe and a tions, single-channel conductance at a given pimanometer to the suction tubing of the pipette pette suction can be calculated from the 1,-V, holder for the repeated application of known plot. suction. D a t a analysis and statistics Because the application of suction often moved the electrode off the cell membrane, we collected For analysis, recorded data were filtered to only one complete set of on-cell recordings. For yield an effective bandwidth (corner frequency, f,) the on-cell configuration, the membrane potential of 1.5 kHz ( - 3 db, &pole low-pass Bessel filter). of the patch (V,) is equal to the resting mem- Unless specified, the filtered data were digitized brane potential of the cell offset by an applied at 10 kHz and stored on computer hard disk. The voltage (V,) from the clamp circuit. All other digitized data were further analyzed by the membrane patches in this study were excised IPROC-2 computer program (Axon Instruments, inside-out configuration. For these patches, V, Inc., Burlingame, CA) which is an automated was equal t o V,, both reported relative t o the pi- event detection program designed for the analysis pette solution. Inward ( ) current refers to the of single ion channels. By this program, the flow of cations from the extracellular side to the single-channel current (I,) was calculated as the cytoplasmic side of the membrane patch, i.e., from mean current during validated channel openings. the pipette solution into the bath solution. Exci- For membrane patches with only a single chansion of the patch from the cell sometimes resulted nel, the single-channel open probability (Po)was in vesicle formation in the pipette tip; brief expo- calculated from the total current amplitude histosure of the pipette to air generally restored ob- gram with areas under the peaks taken t o be proservable channel activity. In some membrane portional to the time spent in those conductance patches, initial application of suction activated states. For the membrane patch with multiple SA large numbers of channels (more than 5 or 6) channels (as seen in Fig. la), we first.determined which proved too difficult for analysis of single- the number of the channels. By applying suction channel properties. We typically excluded these to the pipette or depolarizing the membrane membrane patches from further analysis. Only patch, either of which would elicit high channel the channel currents which could be activated re- activity, the maximum number of the channels in versibly by suction were recorded and analyzed. the membrane patch can be determined. Under Data collection from patches generally began fol- the assumption that channels in the membrane lowing the appearance of channel activity ac- patch were independent of each other, we calcutivated by brief test application of suction. We lated the Po from the binomial distribution of the recorded spontaneous channel activity, when time spent in the current levels corresponding t o present, at various holding potentials in bath so- different numbers of open channels. Singlelutions for the determination of current-voltage channel conductance (g,) was calculated as the (Ic-Vm)relationships before any other manipula- slope of plots relating I, to V,; regression analysis tions. was applied t o the linear portion of the graph which included the physiological range of memSuction application brane potentials. The plots of I, vs. V, were typiTo study the stretch-dependence of channel ac- cally linear over a wide range (V, = 0 t o -90 tivity, we applied suction t o the membrane patch mV). Since it was difficult to discriminate small simply by retracting the plunger of a 1 ml plastic currents from background current noise, the exsyringe; we calibrated the syringe system by ma- act membrane potential at which no current was nometry. Each 0.1 ml of suction generated a pres- conducted by the channel could not be determembrane patch. Osmotic pressures of all solutions were routinely monitored and never varied from 320 mosmol/l by more than 5 mosmol/l; where necessary, osmotic pressure was adjusted by mannitol or sucrose addition.
+
STRETCH-ACTIVATED CATION CHANNEL FROM FISH
a
307
$1
Suction (AP, cm H 2 0 )
200 msec
1.o
b 0.8
0
0.6
n 0.4 0.2 0.0
0
-1 0
- 20
-30
-40
Suction (AP, cm H20 )
4
- 2 4 crn
4
-32 crn
JI
0 cm
Fig. 1. Suction dependence of channel activity in an excised, inside-out membrane patch and an on-cell membrane patch. a: The current records from an excised, inside-out patch containing 4 SA channels. For this patch, the pipette contained Na-ES and the bath contained GBR; V, was held at - 90 mV. Suction was applied in sequentially increasing steps. The “C” represents the current state when all channels were closed. b The relationship of average Po,in 6 inside-out patches with Na-ES in pipette and GBR in bath, to either
sequentially or non-sequentially increased suctions (pooled data). c: The current records from an on-cell patch. The upward arrows indicate applications of suctions and the downward arrow indicates the release of suction. V, was held at -80 mV; V, was probably near - 130 mV, i.e., the membrane patch was strongly hyperpolarized. I, for the channels in this on-cell recording membrane patch was about 10 PA. The baseline noise was due to the activity of a small spontaneously active channel. (f, = 1.5 kHz)
mined. Therefore, we determined the reversal potential (V,) by the extrapolation of the linear portion of the Ic-Vm plot. The channel showed slow adaptation to applied suction over 20-30 seconds (see Results section for details). We analyzed and reported the first 20-30 seconds of recorded data
directly after the application of suction as the peak response. Adaptation in channel activity was assessed by determining Po for short intervals of time following the application of suction. Powas calculated for single-channel current records 10 seconds in duration, each segment begin-
W.CHANG A N D C.A. LORETZ
308
ning at 2 second intervals throughout the current record. These overlapping sample periods smoothed the data for graphical presentation; data were not combined over transitions in suction. In the study of long-term channel adaptation to applied suction, average Pofor non-overlapping 10 second current records were calculated and plotted. The data were reported as mean i SEM. Statistical significance of differences was assessed using t tests (Dixon and Massey, 1965).
levels of channel conductances. When bathed in GBR, the channels we observed had slope conduc3 pS (n = 4) and 116 i 11 pS tances of 55 (n = 2), with Na-ES in the pipette, and had slope conductances of 70 i 10 pS (n = 2) and 145 i 13 pS (n = 5), with K-ES in the pipette. With G35MES in the bath and Na-ES in the pipette, the channels had slope conductances of 57 5 6 pS (n = 4)and 92 pS (n = 1).Although SA channels exhibited two classes of conductances, they showed no distinguishable difference in suction sensitivity, ion selectivity, voltage-dependence, RESULTS and other biophysical features. Figure 1 illustrates the suction sensitivity of Figure 3 shows the voltage-dependence of SA the channel. From the studies of 17 inside-out ex- channel activity in the absence of membrane cised patches, we found - 8 cm H20 AP was insuf- stretch. Po increased when the membrane was ficient t o induce channel activity. When suction strongly depolarized. Over the physiological equal to or greater than -16 cm H20 AP was range of membrane potential, however, channel applied, channel activity increased with AP (Fig. activity was low in the absence of suction and la,b). Application of - 32 cm H,O AP increased Po independent of membrane potential. In the abup to about 0.9 in some patches. We also observed sence of cytoplasmic-side Ca2+, i.e., the memstretch-dependence of channel activity in one on- brane patch bathed in G35MES, activation of cell patch (Fig. lc). When the cell membrane was channel activity by depolarization was shifted by strongly hyperpolarized, Powas very low and suc- approximately + 20 mV. tion of -24 cm H 2 0 A P slowly induced slight In Figure 3a, when V, was clamped at -50 channel activity. Application of -32 cm H 2 0 A P mV and -80 mV, the channel conducted inward suction activated many channels in the mem- currents, suggesting the possible Na+, C1-, and brane patch after a short delay. After the release Ca2+ permeability of this SA channel. In Figure of suction, channel activity slowly returned t o the 4a, with Na-ES in both pipette and bath, the Vr baseline level. Figure 2 shows that the stretch is approximately zero (0.5 1.2 mV, n = 3). The activation was observed over a physiological replacement of Na-ES in the bath with Na,SO,range of membrane potential, which lies between BS failed t o significantly change the V, (1.8 r - 40 mV and - 70 mV. 1.2, n = 2), suggesting that this channel has very With identical solutions in pipette and bath, low permeability t o C1- ion. A 10-fold increase in individual SA channels expressed either of two Ca2+concentration in the bath also failed to shift the V, (1.0 i 0.8 mV, n = 2). This result suggests low Ca2+ permeability of the SA channel. With 0.5 Na-ES in the pipette and GBR in the bath the V, -40mV (-6.9 ? 2.7 mV, n = 6) was not significantly 0.4 -50 mV different from the theoretical V, for a purely 0 -70 mV Na+-selective channel in GBR ( - 4 mV). The av0.3 erage V, for same inside-out patches bathed in a" the G35MES ( + 14.1 +- 2.2 mV, n = 6) was more 0.2 negative (P< 0.01, two-tailed t test) than the V, for a purely Na+-selective channel in G35MES 0.1 ( + 36 mV). This suggests that the channel is also K + permeable. K + permeability was further con0.0 firmedbyV,(-2.1? 2.1mV,n = 7)ofthechan0 -16 -24 -32 nel current using K-ES in the pipette and GBR in Suction (AP, cm H20) the bath. Using the GHK constant field equation, Fig. 2. Relationship of Poto applied suction for an excised P,,/P, was calculated as about 0.75 for channels inside-out patch when V, was held in the physiological range bathed in the G35MES solution with Na-ES in of membrane potential (-40, -50, and - 7 0 mV). Suction pipette. The calculation of P,,/P, was supported was increased in steps sequentially under each holding potential. by the observation that the conductances of chan_+
_+
~
309
STRETCH-ACTIVATED CATION CHANNEL FROM FISH
a
10
%L
Vm (mV)
I
I
80 msac
-50 5
c
-+-
Na-ES/Na-ES+lOmMCa
--O-
Na-ESINaZS04-BS
- 1 00
0
100
Vm (mV)
C
50
*
lb
I
NaCCESiGBR
+ NaCI-ESIG35MES
10
s v
0.8-
no
I
I
20 I
0 -
0
0.60.4 -
-
1
0
0 100
0
Vm (mV)
0.2-100
- 1 00
-60
-20
20
60
100
Vm (mV) Fig. 3. The channel was voltage gated when the membrane was highly depolarized. a: The current records from an excised, inside-out patch in a pipette filled with Na-ES and bathed in GBR solution. When V, was held a t - 50 and - 80 mV, the channel conducted inward current (upward deflection) with a low single-channel activity. When V, was held at + 30 and + 50 mV, the channel conducted outward current (downward deflection) with a much higher channel activity. The “C” represents the current state when all channels were closed. b: The relationship of average single-channel Po,in 6 excised, inside-out patches with Na-ES in pipette, to V, in GBR and G35MES bath solution. The data points without error bars represent single samples (f, = 1.5 kHz).
nels were greater when K-ES was used in the pipette than the conductances of channels when Na-ES was used. Because of the less defined ion chemical gradients in GBR, we could not calculate a reliable permeability ratio for channels bathed in this solution. Figure 5 presents the effects of suction on I, and g,. In 4 patches, which were held at V, in the
Fig. 4. a: The representative &-V, plots for SA channels in excised, inside-out membrane patches, with either Na-ES o r K-ES in the pipette, and bathed in either GBR (161.4 mM Na+, 2.5 mM K’) or in G35MES (35 mM Na+, 35 mM K + ) . b: The representative Ic-V, plots for SA channel in excised patches, with Na-ES in the pipettes, and bathed in Na-ES, Na,SO,-BS or Na-ES + 10 mM Ca2+.
8
Suction
(AP,cm H20)
*
-
-80
-60
o
Conductance
(ps) 4 1 .O
-16
-40
43.3
-20
Vm (mV) Fig. 5. &-V, plot for a channel in an excised, inside-out patch bathed in G35MES maintained under different levels of suction. Suction was increased sequentially under each holding potential.
310
W.C H A N G A N D C.A. LORETZ
0 range of the physiological membrane potential, both I, and g, in either GBR or G35MES increased Open a with applied suction. Application of - 32 cm H,O AP to some patches increased I, by up to 80%and g, by up t o 40%. In order t o study the kinetics of single SA channels, a set of single-channel data records is v) c needed. Most of the membrane patches we studied C 3 contained more than two channels. Fortunately, we obtained a membrane patch containing only one SA channel and lasting long enough for experimental manipulation. The presence of a single channel in the patch was confirmed by applying - 40 cm H20 AP to the pipette; this level of suction increased Poup t o 0.9. At this high Po, only a single uniform current level for the open 0.0 10.0 20.0 30.0 40.0 50.0 state was observed. Using curve fitting, channel Time (msec) open-time distribution can be fit by one single exponential (Fig. 6a): number of openings = 0 0 e-t’topen, where topen is the open state time constant. 2 The sum of two exponential functions better fit closed-time distributions than a single exponen0 0 tial function (Fig. 6b). These results suggest there 0.3 is one open state and two closed states for this SA channel. The open state time constant, topen, 0 0 increased with applied suction and was indepenc dent of V, (Fig. 7b,e). Because of the infrequent 3 appearance of long closed states of the channel when the membrane was depolarized and when suction was applied, it was difficult t o get reliable closed time constants. Therefore, we have only de0 0 termined the effects of applied suction and V, on fa the mean closed time calculated as the arithmetic mean of those events. The mean closed time de0 creased with the applied suction and with mem0.0 20.0 40.0 60.0 80.0 brane depolarization (Fig. 7c,d). Time (msec) SA channels showed adaptation to applied suction. In Figure 8a, application of sequentiallyFig. 6. a: A sample curve fitting of open-time distribution increasing steps of suction caused increases in by a single exponential function. b: A sample curve fitting of channel activity. For steps in suction t o - 24 and closed-time distribution by the sum of two exponential func-32 cm H,O, increases in channel activity were tions. Both open-time histogram and closed-time histogram abrupt and were followed by slow relaxation to data were sampled from the current record of a single channel in the patch which was voltage clamped a t V, = 60 mV and levels about one-half of those observed immedi- was under no suction condition. ately following application of suction. We also tested the effects of non-sequentially applied steps in suction on channel activity (Fig. 8b). DISCUSSION Channel activity was still dependent on suction when periods of zero-suction were interposed beThe anecdotal observation that the degree of fore and after suction application. The increases urinary bladder filling in the goby depends on the in channel activity were also abrupt and were fol- acclimation salinity (Loretz and Bern ’80) suglowed by slow relaxation. Despite adaptation, the gests that the processing time for the production steady-state level of channel activity was still of the final excreted urine can vary substantially higher than that in the absence of suction (Fig. between freshwater and seawater-adapted fish. In 8c,d). the hypertonic environment, increased urinary
s
Q
STRETCH-ACTIVATED CATION CHANNEL FROM FISH -40mV
0.8
a
d
8
0.4
0.4
0.2
0.0
-50
-70
-40
40
E
0.0
60
0
6-
E
v
4-
-24
-32
2
-4
L
0
-16
-24
-32
4
C
W
a 0
2
c
0
0 -70
0
-16
6
W
?
1
0
b
h
v
-70mV
06
0.2
s
0
08
0.6
8
-50mV
31 1
-50
-40
40
60
I C
E
f
200 150
c
is
100 50
70
-50
-40
40
60
Vrn (rnV)
0
-1 6
-24
-32
Suction (AP, cm H20)
Fig. 7. Panels a, b, and c: Dependence of single-channel kinetics on membrane potential in the absence of suction. Panels d, e, and f: Effects of suction on single-channel kinetics over a physiological range of V,.
bladder filling and residence time in seawateradapted fish promote net NaCl absorption, which subsequently is actively excreted across the gills. Along with NaCl absorption, the water retrieval from the lumen of bladder produces a final urine rich in Mg2+ and SO!- which were passively accumulated across the gut wall as part of the greater drinking activity in seawater-adapted fish. There are multiple modes of control of urinary bladder absorptive transport. In addition to changes in urine delivery and residence time following salinity adaptation, the modification of urine composition is affected by changes in epi-
thelial transport. Hormonal control of urinary bladder ion transport has been documented for the goby (Loretz and Bern, 1983) and in this report, we present our studies on an SA ion channel which may play a role in local regulation of urinary bladder function. Suction-dependence of channel activitu In contrast t o several other types of ion channels seen to be spontaneously active in our patch clamp experiments, we observed one class of channel, which showed activation only with the application of suction. At physiological V,, the spontaneous single-channel open probability (Po)
W.CHANG AND C.A. LORETZ
312
0.2
0.0 20
0
b
0 cml
40
60
I
-24 cm
80
Ocm
I
60
80
100 -32cm
I
120 Ocm
0.4
0
20
40
0.4
100
120
1
100
0
200
300
Time (Sec)
-24 cm
Fig. 8. Adaptation of the goby urinary bladder SA channel to applied suction. a: The effects of sequential applications of increasing suction on Po. b: The effect of non-sequential application of suction on the single-channel Po.c: The effect of non-sequential application of long term suction on the
single-channel Po. The horizontal bar over each panel indicates the suction applied (AP, cm H,O). (Note the change in time axis.) d Partial current records used to generate panel c. Upward arrow indicates the application of suction in cm HZO.
of this channel, in either on-cell or excised, inside-out patches, was low (-0.02) and could be induced by application of pipette suction. When the applied suction was greater than - 16 cm H,O (0.79 x lo4 dyne/cm; Guharay and Sachs, '841, Zn(P,) increased linearly with (AP)' (Fig. 9). This relationship supports the theoretical model proposed by F. Sachs (Guharay and Sachs, '84) that SA channels are activated by energy transduced from membrane tension imposed by the applied pressure difference. The submembrane intracel-
lular cytoskeleton is probably involved in the energy transduction and in modulating the sensitivity of the system. When data from recordings with applied suctions less than -16 cm H,O were included in the analysis, no significant linear relationship between Zn(P,) and (AP)' was observed. This finding is similar t o that observed for SA channels in chick embryonic skeletal muscle. Our interpretation is that the effects of suction are 2-fold: first, application of suction promotes attachment of the membrane patch t o the pipette
313
STRETCH-ACTIVATED CATION CHANNEL FROM FISH
Sachs, '84;Kirber et al., '88; Taniguchi and Guggino, '89). Although this SA channel is not permeable to Ca2+,Ca2+may indeed play a role in channel regulation. In the absence of suction, the mere presence of Ca2+ was insufficient to increase Po in the physiological range of V,; studies of Ca2+ effects on Poin channel under suction will address the possible regulatory role of Ca2+ .
h
0
%
C
Single-channel conductance is dependent on applied suction 0 400 800 1200 In four excised inside-out patches, the singleAP2 (crn' H20) channel current and single-channel conductance Fig. 9. The In(P,) is a linear function of (AP)', when (AP12 increased with applied suction (Fig. 10). We suswas equal to or greater than 256 cm2 H,O (pooled data). pected the increases in current and conductance might be due t o the low cut-off frequency for filglass and, second, further suction increases ten- tering of channel events in our analysis. Theresion on the patch. In studies of 17 membrane fore, we increased f, from 1.5 kHz t o 3.0 kHz and patches, we observed the suction-dependence of increased data sampling frequency from 10 kHz channel activity in a physiological range of mem- to 20 kHz. Despite these alterations, similar debrane potentials. Within this range of membrane pendence of channel current and conductance on potentials the channel activity was low (-0.02) applied suction was observed. We also considered when no suction was applied; while the applica- two other explanations for the increase in conduction of - 32 cm H 2 0 AP induced Poup t o 0.9. This tance with applied suction. One is the ability of result indicates that the SA channel performs its the applied suction to cause a tighter seal bebiological function in the presence of membrane tween membrane patch and pipette glass, which, tension. in turn, decreased the leak current through membrane patch. By reviewing the data record in the Voltage-dependence of channel activitg video tape, no shifting of baseline current suggesThe SA channel reported in this paper showed tive of such an effect occurred when suction was voltage-dependent activation in the absence of applied, ruling out the possibility of decrease in suction only when the membrane was highly de- leak current. Another possibility for the increase polarized (V, between 0 mV and + 80 mV). Simi- in channel conductance is that the applied preslar voltage-dependence was reported in an sure difference on the membrane could create a osteoblast-like cell line (Duncan and Misler, '89) hydraulic pressure, which, in turn, enhanced the and toad smooth muscle (Kirber et al., '88). In the movement of ions and, therefore, current through excitable chick skeletal muscle, the voltagedependence of an SA channel was observed when the membrane was hyperpolarized (Guharay and Sachs, '84). In the physiological range (from 0 mV--90 mV), voltage-dependence was expressed during the application of suction, suggesting that membrane voltage may be a relevant regulator. -4
!
I
The channel is permeable to both K + and N a + ions From the ion substitution experiments, this SA channel shows its selectivity to Naf and K + ,with P,,/PK 7.5, but not to C1- and Ca2+.The general finding for SA channels from other tissues is a relative non-selectivity among cations with K + being more permeable than Na+ in those tissues where selectivity has been tested (Guharay and
0
10
20
30
40
Suction (AP, crn H20) Fig. 10. The dependence of single-channel conductance on applied suction. Average g, was calculated from the channels in 4 patches bathed in GBR with full range of pipette suction steps.
314
W.CHANG AND C.A. LORETZ
the channel (Hille, '84).When we applied suction to the pipette holder, we created a hydrostatic pressure which was oriented from the cytoplasmic side to the extracellular side of membrane patch. If the hydrostatic pressure-induced flow did drag ions through the ion channel, we should have observed a decrease in outward channel current instead of the observed increase. This observation rules out the second possibility. Therefore, we propose that the increases in channel conductance and channel current are due t o the conformational change of channel protein as a result of the change in energy state of cell membrane.
membrane. Although it is too early t o make a conclusion as to the physiological function of this SA channel in the urinary bladder, we do believe the urinary bladder has a mechanical sensor t o monitor and to respond t o the amount of urine production by changing the functional status of the urinary bladder. The discovery of the SA channel in the urinary bladder indicates that the urinary bladder may be more than a urine reservoir or an ion transporting osmoregulatory organ under remote control via hormones, but it may also possess the mechanism for autoregulation of transport.
The channel adapted to the applied suction The SA channels, in the excised, inside-out patches, adapted t o the applied suction in about 20-30 seconds, relaxing to a steady-state level in which the channel activity was still higher than the baseline activity. Considering that the bladder fills slowly, channel activity in situ probably expresses the more modest dependence on stretch suggested by our observations of adaptation. The adaptation of SA channel has also been observed in the yeast plasma membrane (Gustin et al., '88) and in human fibroblasts (Strockbridge and French, '88). The urinary bladder of the goby also undergoes rhythmical contractions, the duration of which are short enough, that they might induce more punctuated increase in channel activation. These results indicate that the responses of the urinary bladder cell t o mechanical forces may involve short term and long term changes in functional status.
ACKNOWLEDGMENTS
Physiological signwcance of the SA channel in urinary bladder Although we can describe the distribution of this channel specifically to the columnar cell of the bladder, we cannot specify the location within the cell. It could be in either the apical membrane or basolateral membrane. If it is in the apical membrane, it can facilitate the Na', C1-, and water reabsorption from the urine by increasing the Na conductance of the apical membrane, which is an intrinsic part of the ion absorptive function of the urinary bladder in goby (Loretz and Bern, '80) and in flounder (Renfro, '75, '78). It might also facilitate K f secretion into urine, a function observed in flounder urinary bladder (Stokes, '88) by increasing K f conductance of the apical membrane. Alternatively, this SA channel may be part of cell volume regulatory mechanism and its cellular location may not be limited to the apical
This study was supported by NSF grant DCB8718633 and BRSG grant 2S07RR0706625. We thank Dr. C.R. Fourtner for valuable discussion.
LITERATURE CITED Bentley, P.J. (1971) Endocrines and Osmoregulation: A Comparative Account of the Regulation of Water and Salt in Vertebrates. Springer-Verlag, New York, pp. 214-218. Christensen, 0. (1987) Mediation of cell volume regulation by Ca+ influx through stretch-activated channels. Nature, 330:66-68. Dawson, D.C., and R.A. Frizzell (1989) Mechanism of active K+ secretion by flounder urinary bladder. Pflugers Arch., 414:393-400. Demarest, J.R. (1984) Ion and water transport by the flounder urinary bladder: Salinity dependence. Am. J . Physiol., 246 (Renal Fluid Electrolyte Physiol., 15):F395-401. Demarest J.R., and T.E. Machen (1984) Passive and active ion transport by the urinary bladder of a euryhaline flounder. Am. J . Physiol., 246 (Renal Fluid Electrolyte Physiol., 15):F402-408. Dixon, W.J., and F.J. Massey (1965) Introduction to statistical analysis. McGraw-Hill, New York, pp. 109-126. Duncan, R., and S. Misler (1989) Voltage-activated and stretch-activated Ba' conducing channels in an osteoblast-like cell line (UMR 106). FEBS Lett. 251:number 1,2, 17-21. Fisher, R.S., and J.W. Lockard (1988) Complex response of epithelial cells to inhibition of Na transport by amiloride. Am. J. Physiol., 254 (Cell Physiol., 23):C297-303. Frizzell, R.A., M. Field, and S.G. Schultz (1979) Sodiumcoupled chloride transport by epithelial tissues. Am. J. Physiol., 236(1) :F 1-8. Guhary F., and F. Sachs (1984) Stretch-activated single ion channel currents in tissue-cultured embryonic chick skeletal muscle. J. Physiol., 352:685-701. Gustin, M.C., X. Zhou, B. Martinac, and C. Kung (1988) A mechanosensitive ion channel in the yeast plasma membrane. Science, 242:762-764. Handler, J.S. (1988) Antidiuretic hormone moves membranes. Am. J . Physiol., 255 (Renal Fluid Electrolyte Physiol., 24):F375-382. Hille, B. (1984) Ion channels of excitable membranes. Sinauer Associates Inc., Massachusetts. Howard, J., M. Robert, and A.J. Hudspeth (1988) Mechanical +
+
STRETCH-ACTIVATED CATION CHANNEL FROM FISH transduction by hair cells. Ann. Rev. Biophys. Chem., 17:99-124. Kirber, M.T., J.V. Walsh, J r . and J.J. Singer (1988) Stretchactivated ion channels in smooth muscle: A mechanism for the initiation of stretch-induced contraction. Pflugers Arch., 412:339-345. Lansman, J.B., T.J. Hallam, and T.J. Rink (1987) Single stretch-activated ion channels in vascular endothelial cells as mechanotransducer? Nature, 325:311-313. Loretz, C.A., and H.A. Bern (1980) Ion transport by the urinary bladder of the gobiid teleost, Gillichthys mirabilis. Am. J. Physiol., 239:R415-423. Loretz, C.A., and H.A. Bern (1981) Stimulation of sodium transport across the teleost urinary bladder by urotensin 11. Gen. Comp. Endocrinol., 43:325-330. Loretz, C.A., and H.A. Bern (1983) Control of ion transport by Gillichthys mirabilis urinary bladder. Am. J. Physiol., 245:R45-52. Loretz, C.A., and Fourtner, C.R. (1988) Functional characterization of a voltage-gated anion channel from teleost fish intestinal epithelium. J. Exp. Biol., 136:383-403. Morris, C.E. (1990) Mechanosensitive ion channels. J . Membr. Biol., 113:93-107. Nishimura, H. (1985) Endocrine control of renal handling of solutes and water invertebrates. Renal Physiol. Biochem., 8:279-300. Nishimura, H., and M. Imai (1982) Control of renal function
315
in freshwater and marine teleosts. Fed. Proc., 41: 2355-2359. Ohmori, H. (1988) Mechano-electrical transduction of the chink hair cell. Prog. Brain Res., 74:ll-20. Renfro, J.L. (1975) Water and ion transport by the urinary bladder of the teleost Pseudopleuronectes americanus. Am. J. Physiol. 228(1):52-61. Renfro, J.L. (1978) Interdependence of the Na+ and C1transport by the isolated urinary bladder of the teleost, Pseudopleurnectes americanus. J . Exp. Zool., 199:181-390. Sachs, F. (1986) Mechanical transduction unification. NIPS, 1:98-100. Stokes, J.B. (1984) Sodium chloride absorption by the urinary bladder of the winter flounder. J. Clin. Invest., 74:7-16. Stokes, J.B. (1988) Passive NaCl transport in the flounder bladder: Predominance of a cellular pathway. Am. J . Physiol., 255 (Renal Fluid Electrolyte Physiol. 24):F229-236. Strockbridge, L.L., and AS. French (1988) Stretch-activated cation channels in human fibroblasts. Biophys. J., 54: 187-190. Taglietti, V., and M. Toselli (1988) A study of stretchactivated channels in the membrane of frog oocytes: Interaction with C a + +ions. J. Physiol., 407:311-328. Taniguchi, J., and W.B. Guggino (1989) Membrane stretch: A physiological stimulator of Ca2' -activated channels in thick ascending limb. Am. J . Physiol. (Renal Fluid Electrolyte Physiol. 26):F347-352.
