Membrane currents controlled by physical forces in cultured mesangial cells

Kidney international, Vol. 43 (1993), PP. 535—543

Membrane currents controlled by physical forces in cultured mesangial cells WILLIAM CRAELIUS, MARGERY J. Ross, DENNIS R. HARRIS, VICTOR K. CHEN,

and CAiuos E. PALANT Department of Biomedical Engineering, Rutgers University, Piscataway, New Jersey; Nephrology Service and Research Service, Department of Veterans Affairs Medical Center, Brooklyn, Department of Medicine State University of New York Health Sciences Center at Brooklyn, New York, USA

Membrane currents controlled by physical forces in cultured mesangial

cells. Mechanically-activated ion channels (MACs) of cultured rat mesangial cells were stimulated by applying suction to patch pipets or by exposing cells to hypoosmotic media. MAC density was estimated as 1.5 0.4 per t2. In the absence of any stimulus, MAC open probabil-

ities (N * P) were < 0.0001 increasing as a function of stretch or extracellular hypoosmolarity. Single channel mean open time during stretch increased with patch depolarization whereas hyperpolarization

of the membrane delayed MAC inactivation. Ionic conductance of MACs, based on average slope conductances at hyperpolarized potentials, was 76 pS in high external K (N = 5) and 40 pS in high external Na (N = 8). PKtPNa* was estimated to be 4.7. MACs did not permeate C1, at least outwardly. Whole cell currents in response to voltage steps applied to resting cells in control conditions were approximately ohmic

active responses has been mounting for a wide variety of cells in physiological systems [5]. In particular, cultured mesangial cells change morphologically, alter their growth characteristics

and metabolic activity during periods of cyclic stretchingrelaxation [6]. While the intermediate steps involved in response to stretch are not known, a likely candidate for the transduction process is the mechanically-activated ion channel (MAC) that carries a mixed current of K and Nat, with which the mesangial cell is richly endowed [7].

In addition to their possible role as mechanotransducers, MACs have been implicated in the adaptational response to between —120 mV and 40 mY and were linearly and reversibly hypoosmotic conditions [8, 9]. Mesangial cells possess a budependent on extracellular osmolarity. Our results demonstrate that: (1) MACs can be activated by both negative hydrostatic pressures applied

to the pipet and by osmotic gradients; (2) MAC kinetic behavior is sensitive to membrane potential; (3) MACs may participate in cellular responses to physical forces.

metanide-sensitive electroneutral ion carrier [10] that may be of

importance in regulation of cell volume during an osmotic challenge. At present, little additional information on the role of

ion channels in mesangial cell responses to osmotic stress is

available. In the present study we have examined the activation mechanisms of MACs, their ionic permeability and voltage-depenWithin the glomerulus, mesangial cells are closely apposed to dent behavior as well as their susceptibility to a variety of ion both the basement membrane and capillary endothelium. Fine channel blockers. To determine whether MACs are triggered by microtendinous mesangial processes extend into the glomerular osmotic swelling, single channel responses were measured in basement membrane and contact capillary lumena across endo- hypoosmotic media and compared to MACs elicited by pipet thelial fenestrations [1, 2]. At the opposite extreme of the suction. In order to determine whether MACs influence whole glomerular capillary, extraglomerular mesangial cells or "lacis" cell behavior the whole-cell patch clamp technique was used to cells sit across a narrow fluid cleft that separates them from the measure membrane conductance of cells exposed to anisotonic

macula densa epithelium, and come in close proximity to conditions. The results described herein provide evidence for glomerular arterioles, partially surrounding the efferent arte- single ion channel responses as well as macroscopic ionic riole in its intraglomerular portion [3]. These mesangial mi- currents resulting from reductions in extracellular osmolarity. croenvironments are very hydrodynamic, created by hemody-

namic pressures within glomerular capillaries and osmotic pressure arising as plasma mixes with tubular fluids [4]. As a result, dynamic wall tension is an integral part of the mesangial cell environment, and is likely to play an important role in its function. Evidence that membrane wall tensions can directly evoke

Received for publication January 27, 1992 and in revised form October 29, 1992 Accepted for publication October 29, 1992

© 1993 by the International Society of Nephrology

Methods

Cell culture

Rat mesangial cells were provided by Dr. Sharon G. Adler (Harbor-UCLA Medical Center, Los Angeles, California, USA) and obtained from male, 250 gram Sprague-Dawley rat glomeruli using the graded sieving technique. Details of the procedures used for culture and functional as well as immunological characterization of these cells have been previously published [11]. In the present studies mesangial cells were studied in the third through the sixtieth culture passage.

535

536

Craelius et al: Mechano-activated channels in mesangial cells

Table 1. Electrode solutions Sot. Nagluconate Kgluconate KC1 CsCI

CholineCl CaCl2 BaC12

EGTA HEPESb pH Titration

El-la

El-2

El-5

E1-6

135

143

El-3 —

El-4

—

—

—

—

—

—

135

—

—

—

—

—

—

—

150

15

1.5

1.5

— —

— — — — —

— — — 40 — —

20

20

7.4 CsOH

7.4 CsOH

El-l

— — — — 10

— — — — 10

— — — — 10

— — 1

— 11 10

150

— — —

100

USA), and were back-filled using fine plastic tubing. Tip resistances ranged from 5 to 8 M1l. Gigaseal formation occurred

rapidly upon pipet contact with the cell surface, often in the absence of suction, but slight suction (approximately 5 mm Hg) was usually applied to facilitate sealing. For sealing and mem-

brane stretching, negative pressures were applied to the pipet from a side port in the pipet holder through a syringe in series with a Hg manometer. During each experiment, suction was applied, as gradual pulses, ranging from 7 to 25 mm Hg. Seal resistance was monitored by periodic current pulses and was

not affected by episodes of suction. Estimates of channel

number in the patch were made by counting the maximum open level that was achieved, which usually occurred during maximum suction. All concentrations are in m. For whole cell recordings pipets Currents and holding potentials were amplified either with a contained E1-4. In some experiments E1-4 was modified by the addition List EPC-5 (Darmstadt, Germany) or a Warner PC-SO! (Warner of CaCl2 in the following concentrations: 0.1, 2, 4 and 8 mM. 7.4 NaOH

20

7.4 7.4 7.2 7.4 NaOH KOH KOH CsOH

a Ethyleneglycol-bis-(/3-aminoethylether)-N-N'-tetracetic acid b N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid

Instruments, Hamden, Connecticut, USA) amplifiers and recorded on FM tape, with bandwith from DC-18 KHz (Instrutech VR-lOO Digital Recorder, Mineola, New York, USA). Current records were played back from tape through low pass

filters (Bessel, -3dB cutoff 300 to 500 Hz, Warner Inst.) and Electrolyte solutions digitized at 0.1 msec intervals with 16-bit accuracy (Instrutech Cells were bathed at room temperature in a standard Ringer's VR-ST). Longer sampling intervals were sometimes used for solution composed of (in mM): NaCl 150, KC1 6, CaC12 1.2, purposes of compressed data display. Values for single channel MgC12 1.2, glucose 5.5, N-2-hydroxyethypiperazine-N'-2- mean open times and cumulative open probabilities (N * P) ethanesulfonic acid (HEPES) 20, pH 7.4. Patch electrode were obtained using a threshold analysis program [13]. Mean solutions are given in Table 1. open times were determined from single channel openings, and Osmolarities of electrolyte solutions were measured on a open probabilities (N * F) were computed from the sum of freezing point osmometer (Advanced Instruments, Needham (channel open times * level number) total elapsed time [13]. Heights, Massachusetts, USA). Osmolarity of our control Ring- A minimum of 1 second period was used to compute channel er's solution was 328 mOsm/liter, a value within the range of kinetics and probabilities. Holding potentials in the cell-atculture media used for mesangial cells in prolonged incubation tached configuration are expressed as, Vp, representing the (high glucose Dulbecco's modified Eagle's medium and RPMI potential within the pipet relative to ground. from the Gibco Corp., Grand Island, New York, USA). This The permeability ratio, PK+/PNa+, was estimated from the osmolarity value was used to attenuate osmotic stress upon average values of reversal potentials from a number of experireplacement of culture media by Ringer's solution. To study the ments, using the following equation.

effects of whole cell swelling on ion currents, extracellular osmolarity was changed by adding to the extracellular solution, 100 .d aliquots of a hypotonic Ringer's (75 mt NaCl), to lower osmolarity, or a hypertonic Ringer's (300 ms'i NaC1), to raise

Erev = Erev,K+

—

Erev,Na*

= Rt/zF In (PK*{K]o/PNa÷[Na]o) [14]

(1)

osmolarity. The most extreme change in [Na]0 used in the protocol was 21 mrvi. Since Na permeates through MACs, this

where ErevK+ and ErevNa* represent the MAC reversal poten-

change in ionic strength would produce a shift in reversal tial in high K (nominally Na free) and high Na (nominally potential that may have biased the results. Based on the K free) solutions, respectively, obtained from the intercept of permeability ratio PK*/PNa* of 4.7 (Results), the maximum the mean slope conductance of the linear portion of the currentexpected shift in reversal potential (Erev) is < 1 mV using the voltage relationships (at hyperpolarized potentials), PK* and Goldman-Hodgkin-Katz equation. This value is smaller than PNa* are the permeability for K and Na, and [K]0 and the voltage steps used in our whole-cell recording protocols and [Na]0 are the electrode concentrations of K and Nat, respectively. Since equation 1 assumes constant resting memhence should not have measurably affected our results. brane potential, the error in Erev is affected by variations from Electrophysiological measurements cell to cell. Our previous study [101 measured resting potentials Cells were placed on the stage of an inverted phase contrast in 18 mesangial cells at 370, finding a value of —43 12 mY microscope (Olympus IMT-2; Lake Success, New York, USA) (mean SD). Whole-cell recording was done after rupturing the patch with and the patch electrode was advanced by a hydraulic manipua large (> 25 mm Hg) suction pulse. Holding potential was —40 lator (Narishige USA, Great Neck, New York, USA). Currents were recorded through patch pipets attached with giga-ohm seal mV and 10 or 20 mY voltage steps of 100 msec were applied at using methods described by Hamill et al [121. Electrodes were 2 to 4 second intervals under computer control. Currents were pulled from 100 d Drummond Microcaps (Drummond Scien- measured under steady state conditions 100 msec after onset of tific, Broomall, Pennsylvania, USA) on a Flaming-Brown pro- each voltage step. Voltage errors due to series resistance were grammable puller (Sutter Instruments, San Rafael, California, minimized using analog compensation.

Craelius et a!: Mechano-activated channels in mesangial cells

537

0.

500 msec

0— 3—

Fig. 1. MAC response to suction. Channels opened upon application of suction, 7 mm Hg (ON), and closed upon release (OFF). Three channels were evident during the suction (numbers on left). Holding voltage, Vi,, was

*1111.1 I. n..r I. i—.--.-

Vp=-llOmV

i._,riryT

- _______

ON

+

—100 mY, and pipet contained (in mM): Nagluconate 135; CsCI 15; HEPES 10; pH 7.4 (El-i).

OFF

7mm Hg

1.0

All chemicals were purchased from Sigma Chemical Co. (St. Louis, Missouri, USA). Results are expressed as mean SD or SE. Data were fitted to a linear regression using proprietary software (Cricket Graph, Cricket Software, Malvern, Pennsylvania, USA).

0.9

0.8 0.7

Results

0.6 .0

Single channel activation and kinetics 2 0.5 0. MAC responses were elicited in approximately 70% of the a) 150 cell-attached patches examined. Channel number averaged 0. O 0.3 14.6 4 per patch in 7 cells from culture passages 3 to 5. Based on an estimated surface area of 10 ,u2 in each patch MAC 0.2 density was 1.5 0.4 ji2 in these cells. 0.1 In the absence of suction, N * P for "spontaneous" events was < 0.0001, since each 100 second period, contained fewer 0.0 than 10 openings, assuming an average open time of 10 milli30 25 20 15 10 5 0 seconds. The typical experimental protocol consisted of appliPressure, mm Hg cation of several suction pulses in order to elicit MACs repeatedly in the same patch. MAC responses could be elicited in the Fig. 2. Open probability (N * P) of cell-attached MACs as a function same patch for more than 10 suction pulses at various holding of applied suction. Vp = 90 mV, and pipet contained (in mM): potentials. Figure 1 illustrates typical MAC responses in a Na-gluconate 135; CsCI 15; HEPES 10; pH 7.4 (El-l). cell-attached patch exposed to Na-gluconate (El-l, Table 1) and depolarized by 110 mV (—Vp). Immediately upon application of there was an increasing delay in closing as the membrane was 7 mm Hg suction to the pipet (ON), MACs opened in the hyperpolarized. For this experiment a Na-gluconate electrode outward direction (downward), and remained open until suction (El-i) was used. Prolonged delays of inactivation, sometimes was released (OFF). The open level fluctuated rapidly from 1 to exceeding 25 seconds, were present in patches hyperpolarized 3 during the stimulus, reflecting the opening and closing of at more than 60 mV. The highly nonlinear relationship between the inactivation delays and membrane voltage is shown in least 3 channels. Open probabilities (N * P) were a direct function of applied Figure 5, which represents data from 15 cell-attached patches. In addition to suction, hypoosmotic stimuli were used to suction as shown in Figure 2, for a cell-attached patch, with a Na-gluconate electrode (El-l), held at hyperpolarized potential. activate MACs, as shown in Figure 6. All records presented At the maximum suction applied (25 mm Hg), open probability were obtained with a KC1 electrode (El-3, Table 1). In the top trace, suction of 7 mm Hg was applied to a cell-attached patch (N * P) peaked at 0.55, with 5 channels in the patch. Kinetics of MACs were dependent on holding potentials as held at a Vp = 50 mV, in control Ringer's solution (328 mOsm). illustrated in Figures 3 to 5. Figure 3A illustrates an experiment The single channel open time and mean current for these with a KC1 electrode (El-3, Table 1), where a patch was held at channels were 8.0 miliseconds and 4.0 pA 2 respectively (N constant tension, and Vp was changed stepwise from —100 to = 236 events). The lower trace shows records from the same 80 mV. Duration of channel openings decreased from > 1 patch after a five minute exposure to an hypoosmotic Ringer's second at Vp of —100 mV to approximately 100 milliseconds at (278 mOsm), without suction. Note that MACs did not open in hyperpolarized membrane voltages. This shift was independent the absence of suction in control Ringer's solution. The single of the direction of voltage change, with no hysteresis observed. channel open time and mean current for the hypoosmotically1.5, The nearly monotonic increase in mean open time for this patch activated channels were 9.3 miliseconds and 4.9 pA is illustrated in Figure 3B. respectively (N = 223 events). The histograms at right show Another type of voltage-dependence is shown in Figure 4. In similar hi-modal distribution of open times for both types of each trace, suction was applied between the arrows at the stimuli. A total of six cells were studied under similar condiholding potential indicated. Note that at depolarized potentials, tions and showed MAC openings with pipet suction and hychannels closed immediately upon release of suction, however poosmolarity as illustrated in Figure 6. Open probability

538

Craelius et al: Mechano-activated channels in mesangial cells

A

Level

10.

B

—100

-.

—40 —30

-

.

- ---- —

.—i

--0

-.

0.1

A A

—10 —-——''-'-- —0 1

>

.

20

50

AA

1.

—70

.JL_IL JL

.

A

AA A 0.01

iii _1 —100

—50

50

0

100

150

-Vp, mV

80

5 pA 50 ms Fig. 3. A. Influence of holding potential on mean open time of MACs. Single channel currents were recorded from patch held at constant tension while V, was changed in steps from —100 mV to 80 mV. Pipet contained (in mM): KC1 150; CaCI2 1; EGTA 11; HEPES 10; pH 7.2 (El-3). B. Relationship of mean open time to holding potential. Total number of events represented is approximately 5000 from same cell shown in Figure 3A. Note that the abscissa denotes —Vi,, such that depolarizing voltages are at right (see Methods).

(N * P) in the absence of an osmotic challenge was <0.0001 and rose to values as high as 0.2 on exposure to hypoosmolarity. However, there was considerable variability in the relation between N * P and the degree of hypoosmolarity for individual cell-attached patches, and this precluded its quantitative analysis. Permeability of MA Cs

using Equation 1 (Methods) was 4.7. Given that cell membrane potentials were not simultaneously measured the value of 4.7

represents only an estimate of the true K'Na of MACs. These results indicate that mesangial cell MACs are moderately

selective for K over Na and that currents carried by these cations are not significantly affected by the accompanying anion. Further permeability data are provided below. Neither inward nor outward MAC currents were observed in

A series of MAC I-V relationships were obtained for cell- electrode solutions containing either Ba2 or Ca2 as the attached patches having either Na or K as the principal principal cation (El-5 and El-6, Table 1; N = 10 patches), cation as depicted in Figure 7. Lines of regression were suggesting that divalent cations did not permeate MACs and determined separately for the hyperpolarized and depolarized potentials, and error bars represent standard deviations. Since MACs inwardly rectified in all solutions examined, conductances were estimated from the linear slope of the hyperpolarized region. With Na as the principal cation (El-l and El-la,

Fig. 7A), slope conductance in the linear (hyperpolarized) portion of the plot was 40 p5 (r = 0.94) and reversal potential 0

mY (N = 8 patches) with inward rectification present. In a

single cell, reduction in external Cl from 15 m to 1.5 ma'i did not significantly alter the I-V relationship (data not shown). With K as the principal cation (El-2, Table 1), represented in Figure 7B conductance averaged 76 pS (r = 0.98) and reversal potential was 39 my (N = 5) with inward rectification again evident. Replacement of K-gluconate (El-2) by KC1 (El-3) did

block outwardly-directed cation currents. The lack of effect of

Ba2 and Cl on MACs is also exemplified by additional records shown in Figure 8. In this experiment a mesangial cell was repeatedly studied in the cell-attached mode under control conditions with Na-gluconate electrodes (El-l), as well as BaCl2 and choline-Cl electrodes. Following the first response with El- 1 (record A) the electrode solution was replaced first by a an electrode where the BaC12 concentration of El-S was reduced to

50 m (Fig. 8, record B) and then by a choline-Cl electrode

(El-4; Fig. 8, record C). Constant suction of 25 mm Hg

(pressure value not shown) failed to elicit a MAC response in BaCI2 or in choline-Cl as evidenced by the virtual absence of channel openings. In the recovery period a vigorous MAC response was again elicited with El-i (Fig. 8, record D). Figure not alter the I-V relationship. Furthermore, substitution of 8 also illustrates an additional example of delayed inactivation cations with choline-Cl (El-4) resulted in loss of inward currents of MACs in hyperpolarized patches (top compressed record of (data not shown). The permeability ratio PK*/PNa+ calculated A, Fig. 8 and record D). All records were obtained at a Vp value

539

Craelius et a!: Mechano-activated channels in mesangial cells

90mV ION

OFF

80 mV

40 35

!

30 25 C

20 TON

OFF

E

>

IlpA lsec

7OmV

C

15 0

10

ON OFF

C

C

DC COD C

5

ON OFF

C

0

—150 —100

C

C C DC

—50

C

DDD

C

0

50

100

150

-Vp, mV -80 mV

ON OFF

ON OFF

Fig. 5. Influence of holding potential on inactivation of MACs. Results

presented are pooled from measurements in 15 cells.

-100 mV

measured at the voltage steps indicated after a five minute exposure to an extracellular osmolarity of 316 mOsm, from an each trace, suction was applied between the arrows at the holding initial osmolarity of 328 mOsm. The lower set of superimposed potential indicated on the left. As the membrane is hyperpolarized currents was recorded from the same cell, approximately five (from bottom to top) there is increasing delay in closing following release of suction. Pipet composition was (in mM): Na-gluconate 135; minutes after the osmolarity was increased to 360 mOsm. Note Fig. 4. Slow inactivation of MACs at hyperpolarized potentials. In

that the envelope of currents is substantially reduced after the change to hypertonic conditions. Whole-cell I-V relationships for four cells at three different of 50 mY. Taken together these results support the conclusion extracellular osmolarities, are represented in Figure 10. The that mesangial cell MACs did not permeate divalent cations and slope of each regression line represents the average conducreinforced the notion presented above that C1 was not per- tance at the osmolarities indicated at right. The error bars meant in an outward direction. represent the standard error in conductance across cells. As can Ca2 concentrations from 0.1 m to 8 mM (N = 21 patches) be seen, conductances were ohmic in all osmolarities. Erev foT did not noticeably affect MACs with K as the principal cation the currents at all osmolarities tested averaged —41 mY 2 mV in the electrode (El-3). MAC responses remained when the (mean 1.9 SD). Conductance initially averaged 3.8 nS following compounds were included in the electrode solution: SD) during baseline conditions of 328 mOsm, it (mean lanthanum chloride 33 M (N = 4 patches); amiloride 10 M (N increased to 6.5 nS 1 (mean SD) during exposure to 316 = 6 patches); 4,4'-diisothiocyanostilbene-2-2'-disulfonic acid, mOsm, and declined to 2.8 nS 0.7 (mean SD) during DIDS, 1 M (N = 4 patches); quinidine 500 M (N = 3 patches); exposure to 360 mOsm. Thus there was an inverse relationship verapamil 50 M/ml (N = 7 patches) and gadolinium chloride 10 between whole cell conductance and external osmolarity, hav/LM (N = 4 patches). ing a slope 228 pS/mOsm decrease in osmolarity (P <0.01). The nature of the ionic conductance induced by the fall in Whole-cell currents extracellular osmolarity could not be determined without speFollowing some cell attached recordings of MACs with cific pharmacological blockers, however, the data were consisK-filled pipets (El-3), the patch was ruptured with a large tent with a current carried by MACs. The expected Erev for suction pulse, permitting measurement of whole-cell currents. MACs, with a PK*/PNa+ of 4.7 (see above) under whole-cell Electrode composition was designed to mimic intracellular recording conditions (El-3, Table 1) was calculated from the conditions using EGTA to buffer Ca2 to < l0— M. To compare Goldman-Hodgkin-Katz equation to be —35 mY. The reasonthe effects of membrane tension on whole cell-currents the able agreement between the predicted and observed value, —41 extracellular osmolarity was changed stepwise from a baseline mY, suggest that MACs participated in the whole-cell reof 328 mOsm to selected values within the range 316 mOsm to sponses seen in Figure 10. 360 mOsm. During these steps cell conductance was measured Since osmolarity was changed by adjusting the NaC1 concenwith voltage steps between —120 mY and 40 mY after each tration either up or down (Methods), some of the effects observed may have been due to changes in ionic strength, osmolarity change. Records from a representative experiment are shown in particularly with respect to Nat A worst-case estimate of the Figure 9. The series of superimposed currents at top were maximum expected shift in Erev due to the most extreme change CsC1 15; HEPES 10; pH 7.4 (El-l).

540

Craelius et a!: Mechano-activated channels in mesangial cells

A Isoosmotic + suction

100 80 60

JJJLftLJJJLrL

C

40

0

20

C)

tmean =9.3 msec t1

=2msec

t2

=72 msec

0

—3 —2 —1 0 1 Log time, seconds

B Hypoosmotic

100

tmean = 7.95 msec

80 60

t1

' 0

II

J 4pA

=3msec t2 = 43 msec

—3—2—1 0 Log time, seconds

Vp=50 mV

200 msec Fig. 6. MACs are triggered by an hypoosmotic medium. In the top trace, MACs were recorded in the cell attached patch configuration, in response

to suction (7 mm Hg) in control Ringer's solution (328 mOsm). In the bottom trace MACs were recorded after a 5 minute exposure to a relatively hypoosmotic Ringer's (278 mOsm). The population histograms on the right show similar bimodal distribution for both type of stimuli. Results are representative of measurements in 6 cells.

A

—150

B

10

—100

—50

—150

50

100

10

—100

—50

150

50

100

150

-Vp, mV

—Vp, mV

—10

—10

—20

—20

Fig. 7. Current-voltage relationships of MA Cs channels. Constant suction was applied while V, was changed in 10 mY steps. Conductances (g) and reversal potentials (Erev) for each electrode were calculated as the mean regression line for currents at hyperpolarized potentials (g) and the intercept of the mean (E). A. I-V relationships for cells studied with Na-gluconate electrodes (El-l or El-la, Table I). Values presented are mean SD for 8 different patches. Mean conductance was 40 pS (r = 0.94) and Erev was 0 mY (N = 8 patches). B. I-V relationships for cells studied with K electrode (EI-2, Table 1). Mean conductance, g, was 76 pS (r = 0.98) and Erev was 39 mY (N = 5 patches).

in extracellular Na concentration (21 mM) can be made again using the Goldman-Hodgkin-Katz equation. It was estimated

consistent with the minor changes observed in Erev. It was therefore concluded that the change in Na concentration did

that this value was less than 1 mY, and is thus entirely not play a significant role relative to the osmotic stimuli applied.

541

Craelius et al: Mechano-activated channels in mesangial cells

2 sec.

+ OFF

/F

F

I II I

I

S

S S S

Na-gluc I.j.._..__ —"--

A 18 sec. gap

BaCI2

14mm Hg

B

.. _______________ - - ___-_ — _r •jj -

Choline-Cl

C I

ri

11-1

Na-gluc

II + ON

10 sec.

gap

'. ____

OFF ci 7 mm Hg 500 msec Fig. 8. Comparison of MAC currents recorded in the same cell with high extracellular Na pipets versus BaCl2 and choline-Cl pipets. Trace A was obtained from a cell-attached patch with an electrode containing (in mM): Na-gluconate 135; CsCI 15; HEPES 10; pH 7.4 (El-i) showing typical

MAC response (ON-OFF). Trace was shortened for convenience and this is shown as an 18 second gap in the record. B. A segment from a continuous record obtained with an electrode similar to El-5 (BaCI2 100 mM; HEPES 20; pH 7.4, Table 1) in which BaCl2 concentration was reduced to 50 m. Suction was maintained at 25 mm Hg throughout the record (suction level not shown). No MACs were observed. C. Another segment of a record obtained with a pipet containing (in mM): choline-Cl 150; HEPES 20; pH 7.4, (El-4, Table 1) during constant suction of 25 mm Hg showing only a rare channel opening. D. Recovery of MAC response when Na-gluconate electrode (El-i, Table 1) is again used. Vp = 50 mY for all records.

sensors of intraglomerular physical forces. They are activated Rat mesangial cells possess a MAC that permeates both K by stretch applied to small patches of membrane or by whole and Na, but not C1 or Ca2 and has voltage-dependent cell swelling, and carry a relatively large current composed kinetics. Mesangial cell MACs thereby share similarities with primarily of K. While it is possible that MACs serve diverse those described on smooth [15] and skeletal muscle [16], in functions in mesangial cells, an osmoregulatory role seems terms of relatively high permeability to both K and Nat, likely. The current responses shown in Figures 9 and 10 in voltage dependence of mean open time, and insensitivity to a particular suggest the osmosensitivity of mesangial cells. It is variety of ion channel blockers [reviewed in 5]. The apparent suggested that transmembrane water fluxes could be regulated ubiquity of MACs in both prokaryotic and eukaryotic cells [8, 9, by movement of K through MACs. While MACs do permeate 15—22] suggests a diversity of functions. To this date, MACs Na to a significant degree, they appear to be sufficiently have been implicated in electromechanical coupling where they permeable to K to provide reasonable volume regulation. The physiological significance of mesangial cell MACs are may mediate a depolarizing current that triggers stretch-induced contraction of frog gastric smooth muscle [15], vascular supported by separate lines of evidence. Firstly, it is very likely smooth muscle [23] and rat myocardium [24]. Recent data that mesangial cells are subjected to physical stretching during obtained in the fungus Uromyces demonstrates that MACs act changes in distal nephron hydrodynamics. For example, in as surface mechanosensors that facilitate infectivity [25]. A role rabbits and rats it has been observed that intercellular spaces of in volume regulatory decrease has been demonstrated in epi- macula densa epithelium are dilated by volume expansion with thelia were Ca2 or K fluxes through MACs are presumably either isotonic or hypotonic electrolyte solutions [26]. Due to the virtual absence of a limiting basement membrane between coupled to water effiux [8, 9, 18, 22]. Our results suggest that MACs could function as membrane the macula densa and the extraglomerular mesangial cell, it is Discussion

542

Craelius et al: Mechano-activated channels in ,nesangial cells

600

A 316 mOsm

316 mOsm

40 mV r—-----—-------—---—--—---—-----------400

P328 mOsm

360 mOsm

—100 —80 —60

V, mV —200

—400

—120 mV

Fig. 10. Whole-cell I-V plots. Data from four cells are summarized. The slope of each regression line indicates the average conductance at 100 pA 50 mS

B 360 mOsm 40 mV

the osmolarities indicated. Error bars represent standard error in conductance. See Results section for average conductance values. Ringer's solution was composed of (in mM): NaCI 150; KCI 6; CaC12 1.2; MgC12 1.2; glucose 5.5, N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES) 20, pH 7.4. Osmolarity was changed by adding to the extracellular solution, 100 sl aliquots of a hypotonic Ringer's (75

mM NaC1), to lower osmolarity, or a hypertonic Ringer's (300 mM NaCI), to raise osmolarity. Electrode composition was (in mM): KCI 150; CaC12 1; ethyleneglycol-bis-(13--aminoethylether)-N-N'-tetracetic acid (EGTA) 11; HEPES 10, pH 7.2, (E1-3).

-120 mV

kidney cells [9] are also activated by both mechanisms. The question remains as to whether ion channels other than MACs

Fig. 9. Whole-cell currents as a function of osmolarity, The upper trace is a family of superimposed whole-cell currents elicited at an

participated in the whole-cell current responses in hypoosmotic media presented in Figures 9 and 10. Recently, two ion chan-

osmolarity of 316 mOsm. The lower set of current traces was recorded from the same cell after increasing osmolarity to 360 mOsm, which reduced conductance about threefold. All traces were recorded with (in mM): KC1 150, EGTA 11, CaCl2 1, HEPES 10, pH 7.2, (El-3, Table 1).

nels, a 40 pS K channel and a 25 pS non-selective cation channel were reported in cell-attached and mostly isolated patches of mesangial cells. These channels were markedly sensitive to activation by 10—6 M Ca2 [31]. Our whole-cell pipets contained high concentrations of EGTA and a pH of 7.2

which presumably maintained cytosolic Ca2 in the iO M thus apparent that the latter will be physically stressed under range [32]. This low intracellular Ca2 activity should have conditions of volume expansion. Secondly, as osmosensors, inhibited, albeit not suppressed, Ca2-activated ion channels. It MACs could have a significant role in a system whereby is therefore possible that their activity contributed to whole cell glomerular filtration rate is adjusted in response to distal tubular currents. An effective blocker of MACs is critically needed to fluid flow rate and/or tonicity (tubuloglomerular feedback mech- further assess the relative contribution of various ion channels anism or TGF). The physicochemical properties of the fluid to osmoregulation in mesangial cells. transported across the macula densa are of critical importance MACs could be activated by in situ mechanical stress other in the transmission of TGF signals by mesangial cells [27, 28]. than cell swelling. As mentioned in the Introduction mesangial There is strong evidence that perimesangial hypertonicity and cells project microtendinous attachments that insert in the Cl— concentration [27, 29] may trigger TGF, and that distal glomerular basement membrane [1, 2] at sites where glomerular tubular fluid hypotonicity can act similarly [30]. The generation pressures and wall tension may be augmented as a result of a of mixed ionic currents in response to alterations in extracellu- variety of pathophysiological conditions. Mechanical stress, lar osmolarity may be a useful mode of transmission of elect- imposed by abnormally high glomerular pressures and flows rotonic signals from the mesangial syncytium to the adjacent could increase the open probability of MACs, altering regulamicrovasculature. Whether MACs participate in TGF elicited tion of mesangial cell by-products through increased ion fluxes. by volume infusion into the distal nephron is a possibility that In support of this hypothesis is the demonstration that stretchdeserves further consideration and study. dependent Na influx, presumably through MACs, serves as a The similarity of MAC properties during both suction and stimulus for protein synthesis in mammalian myocardium [33]. In summary, rat mesangial cells possess a MAC that permehypoosmotic stimuli (Fig. 6) strongly support the concept that MACs can be activated by either mechanism. Other MACs, ates K and Na but not Cl— or Ca2 and is activated by both such as those of Necturus proximal tubule [8] and opossum suction applied to the patch pipet or by an osmotic gradient in

543

Craelius et a!: Mechano-activated channels in inesangial cells

the absence of suction. Opening of MACs in response to an hypoosmotic challenge constitutes evidence of their cellular role. MACs could act in concert with ion carriers and other ion channels to confer mesangial cells osmosensitive properties and

as basis for mechanosensation of intraglomerular physical forces. Acknowledgments This work was supported by a Merit Review Nephrology Board grant from the Department of Veterans Affairs to C.E.P. and a Grant-in-Aid from the American Heart Association, New Jersey affiliate to W.C. We recognize the support and encouragement of Dr. Geoffrey M. Berlyne. We also wish to thank Dr. Eli A. Friedman for financial support.

Reprint requests to Carlos E. Palant (ill F), 800 Poly Place, Brooklyn, New York 11209, USA.

References 1. SAKAI 1, Kiuz W: The structural relationship between mesangial cells and basement membrane of the renal glomerulus. Anat Embryol 176:373—386, 1987 2. Kiuz W, ELGER M, LEMLEY K, SAIi T: Structure of the glomer-

14. HILLE B: Selective permeability: Independence, in Ionic Channels

of Excitable Membranes, edited by HILLE B, Massachusetts, Sinauer Associates, 1984, pp. 226—271

15. KIRBER MT, WALSH JV JR, SINGER JJ: Stretch-activated ion

channels in smooth muscle: a mechanism for the initiation of stretch-induced contraction. Pflugers Arch 412:339—345, 1988 16. GUHARAY F, SACHS F: Stretch-activated single ion channel currents in tissue-cultured embryonic chick skeletal muscle. J Physiol 352:685—701, 1984 17. MARTINAC B, BUECHNER M, DELCOUR AH, ADLER J, KUNG C:

Pressure-sensitive ion channel in Escherichia coli. Proc NatI Acad Sci USA 8:2297—2301, 1987

18. CHRISTENSEN 0: Mediation of cell volume regulation by Ca2 influx through stretch-activated channels. Nature 330:66—68, 1987 19. GUSTIN MC, ZHOU X-L, MARTINAC B, KUNG C: A mechanosensitive ion channel in the yeast plasma membrane. Science 242:762— 765, 1988

20. MCCANN JD, Li M, WELSH MJ: Identification and regulation of whole-cell chloride currents in airway epithelium. J Gen Physiol 94: 1015—1036, 1989

21. YANG X-C, SAcH5 F: Characterization of stretch-activated ion channels in Xenopus oocytes. J Physiol 431:103—122, 1990 22. HURST AM, HUNTER M: Stretch-activated channels in single early distal tubule cells of the frog. J Physiol 430:13—24, 1990

ular mesangium: A biomechanical interpretation. Kidney mt 23. DAvis MJ, DoNovITz JA, HOOD JD: Stretch-activated single 38(Suppl)30:S2—S9, 1990 3. ELGER M, SAKAI T, WINKLER D, Katz W: Structure of the outflow

channel and whole-cell currents in vascular smooth muscle cells. Am J Physiol 262 (Cell Physiol 31):C1083—C1088, 1992

segment of the efferent arterioles in rat superficial glomeruli. 24. CiELIuS W, CHEN V: Mechanical activation of cardiac myocytes. Contrib Nephrol 95: 1—12, 1991

4. TAUGNER R, HACKENTHAL E: The Juxtaglomerular Apparatus, edited by TAUGNER R, HACKENTI-SAL E, Berlin, Springer-Verlag, 1989, pp. 45—67 5. FRENCH AS: Mechanotransduction. Ann Rev Physiol 54:135—152,

1992

6. HAlutis RC, HARALSON MA, BADR KF: Continuous stretch-relax-

ation in culture alters mesangial cell morphology, growth characteristics, and metabolic activity. Lab Invest 66:548—554, 1992 7. CRAELIUS W, EL-SHERIF N, PALANT CE: Stretch-activated ion channels in cultured mesangial cells. Biochim Biophys Res Cornmun 159:516—521, 1989

8. SACKIN H: A stretch-activated K channel sensitive to cell volume. Proc Nat! Acad Sci USA 86:1731—1735, 1989 9. UBL J, MURER H, KOLB HA: Ion channels activated by osmotic and mechanical stress in membranes of Opposum kidney cells. J Memb Biol 104:223—232, 1988

10. MALLIS L, GUBER H, ADLER SG, PALANT CE: Intracellular chloride activity in cultured mesangial cells. Renal Physiol Biochem 14:12—14, 1991

11. FERDOWS M, GOLCHINI K, ADLER SG, KURT? I: Cl/base ex-

change in rat mesangial cells: Regulation of intracellular pH. Kidney mt 35:783—789, 1989 12. HAMILL OP, MARTY A, NEHER E, SAKMANN B, SIGWORTH FJ:

J Mol Cell Cardiol 22(Suppl l):S33, 1990 25. ZHOU X-L, STUMPF MA, HOCH HC, KUNG C: A mechanosensitive

channel in whole cells and in membrane patches of the fungus Uromyces. Science 253:1415—1417, 1991

26. KAISSLING B, Katz W: Variability of intercellular spaces between macula densa cells: A transmission electron microscopic study in rabbits and rats. Kidney Int 22(Suppl 12):S9—517, 1982 27. PERSSON B-E, SAKAI T, MARSH DJ: Juxtaglomerular interstitial

hypertonicity in Amphiuma: Tubular origin-TGF signal. Am J Physiol 254 (Renal Fluid Electrol Physiol 23):F445—F449, 1988 28. BLANTZ RC, THOMSON SC, PETERSON OW, GABBAI FG: Physio-

logic adaptations of the tubuloglomerular feedback system. Kidney mt 38:577—583, 1990 29. SCHNERMANN J, PLOTH DW, HERMLE M: Activation of tubuloglo-

merular feedback by chloride transport. Pflugers Arch 362:229—240, 1976 30. NAVAR LG, INSCHO EW, IBARROLA M, CARMINES PK: Communi-

cation between the macula densa and the afferent arteriole. Kidney Int 39(Suppl 32):S78—S82, 1991 31. MATSUNAGA H, YAMASHITA N, MIYAJIMA Y, OKUDA T, CFIANG H, OGATA E, KUROKAWA K: Ion channel activities of cultured rat

mesangial cells. Am J Physiol 261

(Renal

Fluid Electrol Physiol

30):F808—F814, 1991

Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflugers Arch

32. T5IEN RY, RINK TJ: Neutral carrier ion selective microelectrodes for measurement of intracellular free calcium. Biochem Biophys

391:85—100, 1981

Acta 599:623—638, 1980 33. KENT RL, HOOBER K, COOPER G: Load responsiveness of protein synthesis in adult mammalian myocardium: Role of cardiac deformation linked to sodium influx. Circ Res 64:74—85, 1989

13. SIGw0RTH FJ: An example of analysis, in Single Channel Recording, edited by SAKMANN B, NEHER E, New York, Plenum Press, 1983,

pp. 301—321