- Email: [email protected]
Rheological behavior of rat mesangial cells during swelling in vitro
~
Biorheology,Vol.M, No.6, pp. 118'7-4011, 199'7 Copyright C 1998 Ebevier Science Ltd Printed in the USA. All rights reserved ()()()6.'55X/97 $17.00 + .00
Pergamon PH SOOO6-355X(98)00023-7
RHEOLOGICAL BEHAVIOR OF RAT MESANGIAL DURING SWELLING IN VITRO
w. CRAELIUSt, CJ. HUANGt,
CELLS
H. GUBER~, AND GE. PALANT~*
tDeparunent of Biomedical Engineering. Rutgers University. Piscataway. NJ; tNephrology Service. Department of Veterans Affairs Hospital. Brooklyn, NY; *University of Nevada School of Medicine, Reno. NV, USA Reprint requests to: William Craelius, Department of Biomedical Engineering, Rutgers University, Piscataway, NJ, 08854 USA; Fax: 732-445-3753; e-mail: [email protected]
ABSTRACT The response of cells to mechanical forces depends on the rheological properties of their membranes and cytoplasm. To characterize those properties, mechanical and electrical responses to swelling were measured in rat mesangial cells (MC) using electrophysiologic and video microscopic techniques. Ion transport rates during hyposmotic exposures were measured with whole-cell recording electrodes. Results showed that cell swelling varied nonlinearly with positive internal pressure, consistent with a viscoelastic cytoplasm. The extrapolated area expansivity modulus for small deformations was estimated to be 450 dyne/cm. Cell swelling, caused either by positive pipet pressure or hyposmotic exposure (40-60 mOsm Kg-1), rapidly induced an outwardly rectifying membrane conductance with an outward magnitude 4-5 times the baseline conductance of 0.9 ± 0.5 nS (p < .01). Swelling-induced (SI) current was weakly selective for K+ over Na", partially reversed upon return to isotonicity, and was antagonized by O.S mM GdCls (p < 0.02; n =6). Isolated cells treated with GdCIs rapidly lysed after hypotonic exposure, in contrast to untreated cells that exhibited regulatory volume decrease (RVD). Our results indicate that volume regulation by Me depends upon a large swelling-induced K+ efflux, and suggest that swelling in MC is a viscoelastic process, with a viscosity dependent on the degree of swelling. © 1998 Elsevier Science Ltd
Introduction Mesangial cells (Me) are components of a reflex that adjusts filtration in response to pressure and volume changes within the glomerulus (for review see Schlondorff, 1996). This mechano-reflex can occur dynamically, through glomerular vasocontrol (Iversen et al., 1992; Blantz et al., 1993; Wang et al.• 1997; Stockand and Sansom, 1997) or chronically, through production of basement membrane (Riser et al., 1992). While details of the reflex are KEYWORDS:
Cell swelling; kidney; membrane stiffness; stretch-activated channels
387
388
Rheology of mesangial cells
Vol. 34, No.6
incomplete, the abundance of mechanically sensitive ion channels in both rat and human MC (Craelius et al., 1989; 1993; Palant, unpublished) and the intimate association of MC with glomerular capillaries and the basement membrane (Sakai and Kriz, 1987) suggest the participation of MC as mechanotransducers. Specifically it can be hypothesized that MC, responding to intraglomerular pressure signals , influence the electrochemical environment and hence their own contractile state, thereby modulating filtration through vasocontroI. To assess the participation of MC in glomerular reflexes, their rheological behavior in response to applied forces requires understanding. The present study describes in particular the mechanoelectrical responses of MC to swelling forces. The mechanotransduction ability of MC may relate not only to glomerular regulation as described above, but also to cellular regulation in the presence of pathological oncotic pressures in the glomerulus Uohnson, 1997).
Materials and Methods Mesangiai cell preparation Mesangial cells were isolated from rat glomeruli using the graded sieving technique. Male Sprague-Dawley rats (Camm Research, Wayne, NJ, USA) weighing 50-75 grams were anesthetized with Metofane, decapitated, bled and placed in 95% ethanol. Kidneys from 8 to 10 rats were excised under sterile conditions using blunt dissection to remove surrounding capsules and serosae. The renal cortices were isolated and purified by sieving through pore sizes of 250, 150 and 75 11m (Thomas Scientific, Swedesboro, NJ, USA). The filtrate, consisting of isolated glomeruli, was then collected atop a sieve with 45 urn pore size and incubated at 37°C for 30 min in 2 ml of collagenase-Type IV, 130 units/mg; clostripain, 0.85 units/mg; and non-specific proteinase, 27 units/mg (Sigma, St. Louis, MO, USA). The specificity of glomerular isolation was > 95%, by microscopic examination. Following enzyme washout, glomeruli were incubated in 75 ml flasks with a 1:1 solution of RPMI + 20% fetal calf serum (Gibeo Labs, Grand Island, New York, USA) and conditioned media prepared from 3T3 fibroblasts (American Type Culture Collection, Rockville, Maryland). The fibroblasts were grown in DMEM + 10% fetal calf serum + 1 % penicillin-streptomycin (Gibco) . The media was refreshed after the first 24 h and was thereafter half-exchanged at two-day intervals until cells became confluent. Identification of spindle-shaped MC was positive under phase contrast microscopy with 2 d of plating. Immunological identification was done by exposing formalin-fixed MC to a-actin antibodies from smooth muscle (Sigma) and antibodies against myosin from smooth muscle and nonsmooth muscle (Biomedical Technologies, Austin, Texas, USA). After staining with immunoperoxidase, MC could be visualized by their strong reaction to anti-a-actin and anti-myosin antibodies. (A separate culture batch of MC was generously provided by Dr. S. Adler of Harbor-UCLA Medical Center.) The morphological and immunological characterization of these cells has been previously published (Ferdows et al., 1989) . For recordings, MC were plated in 30 mm plastic dishes at a density of < 20,000 cells/dish. Cells were studied in the third through the twentieth culture passage.
Electrolyte solutions Cultured MC were bathed at room temperature (22-23°C) for most experiments and at 37°C for the [K+]i measurements. Standard Ringer's composition was (in mM): NaC1150, KCI6. CaCI21.2, MgCI21.2, HEPES In (N-2-hydroxyethypiperazine-N'-2-ethanesulfonicacid) 10, glucose 5.5.
Vol. 34, No.6
389
Rheology of mesangial cells
addition, to block selected currents, some media contained: amiloride 0.1 mM, TEA-el (tetraethylammonium chloride) 16 mM, BaCl2 0.1 mM, TTX 10 J.1M, or GdCIg 0.1-1 mM. Mannitol was used to adjust osmolarities of external solutions to 328 mOsm Kg-I. Table 1 shows the two patch pipet solutions used. In addition, some pipet solutions contained Na2-ATP (l mM) or 8-BromocAMP (0.2 mM). All salts were purchased from Sigma. Osmolarities of electrolyte solutions were measured with a freezing point micro-osmometer (Precision Instruments, Natick, Massachusetts). The osmolarity of the extracellular control Ringer's was 328 mOsm Kg-I, a value within the range of osmolarities (320-350 mOsm Kg-I) of the culture media used for MC in prolonged incubation (see above). Osmolarity of standard electrode solutions was adjusted to be isosmotic with extracellular media (328 mOsm Kg-I). Electrophysiological measurements
Whole-ceIl patch clamp To characterize the conductance activated by hyposmotic stress, the whole-eeIl patch-elamping technique was used (Hamill et al., 1981). Patch-pipets were pulled from 100 J.11 Drummond Microcaps on a Flaming-Brown programmable puller (Sutter Instruments, San Rafael, California, USA), and were back-filled using fine plastic tubing. Cells were placed on the stage of an inverted phase contrast microscope and viewed at 600x magnification (Olympus IMT-2, Lake Success, New York, USA) and the pipet was advanced by a manipulator (Narishige U.S.A., Great Neck, New York, USA). Currents were recorded through patch pipets attached with a giga-ohm seal (Hamill et al., 1981). Tip resistances ranged from 0.8 to 2 MO, and giga-seals with the cell were formed rapidly after slight suction was applied to the pipet. Negative and positive pressures were applied through a side port in the pipet holder and measured with an electronic pressure gage (World Precision Instruments, Sarasota, Florida, USA). Whole-eell recording commenced after rupturing the patch with a suction pulse. Current records were low-pass filtered (-3 dB cutoff 300 Hz) and digitized at 1 ms intervals with 16-bit resolution (VR-ST, Instrutech, Mineola, New York, USA). Holding potential was -40 mV and each I-V sweep consisted of responses to voltage clamps of 200-300 ms applied at 1-2 s intervals beginning at -140 mV and stepping with 10 mV increments to +40 mY. Voltage
Table 1 Patch pipet solution composition Component
El-l (mM)
El-2 (mM)
NaCI
6.0 150.0
10.0
KCI K-Gluconate
135.0 10.0
TEA-CI CaCl2 MgQ 2
1.2
1.2
1.2
1.2
EGTA HEPES
11.0
11.0
10.0
10.0
390
Rheology of 11U!sangial cells
Vol. 34. No.6
errors due to series resistance were minimized using analog compensation. Currents were averaged during the period between 100-200 ms after onset of each voltage step. Currents, voltages, pressure signals, and video images were recorded on VHS tape (Model 420, Vetter, Inc., Reebersburg, PA, USA). Cell conductance was continuously monitored with repeated I-V sweeps after each osmolarity change and during positive pressure applications. Slope conductances for each cell were estimated for both inward and outward currents by fitting lines to both regions. Since reversal potential was about -20 mV, the inward conductance was calculated over the range between -120 and -20 mV, while the outward conductance was calculated over the range between -20 and 40 mV.
Cell volume and surface measurements Plastic dishes (35 mm) containing cultured cells were placed on the stage of an inverted phase contrast microscope and viewed at 600x magnification. Each experiment was monitored with video microscopy and recorded on video tape (Panasonic vHS). Video images were sampled from tape and digitized at 8-bit resolution with a Frame Grabber (Vlab Y/C, Macrosystems, Detroit, MI, USA). Cellular dimensions were quantified with an image processing program written on an Amiga 3000 computer, which has a resolution of 0.25 um/pixel, and measurement error was approximately 0.5 urn. The diameter of each cell was determined as the average of diameters measured at four axes. All cells were approximately spherical. The height of each cell was estimated as the vertical distance between focal planes of its top and bottom surfaces, measured with the focus vernier. Height error was approximately 2 urn. Protocols Four protocols were used to measure rheological responses to swelling in MC. The first protocol measured responses of MC to positive pressures in isosmotic media in whole-cell mode, as depicted in Fig. lAo In these experiments, external osmolarity, I[Co(t)], was equated to internal osmolarity, I[Cj(t)], that was effectively stabilized by dialysis from the pipet. The stretch of the cell during pressurization, Pi, caused water fluxes Qm and Qp. Measurements were made on cells slightly pre-swollen in hyposmolar media to ensure sphericity, and to minimize locally high membrane curvatures that have been described in endothelial cells (Schrnid-Schonbein et al; 1995). Confirmation of sphericity of cells prior to experiments was done by measuring their height-to-diameter ratio, which was 1.1 ± 0.25 in four cells. Positive pressures were applied to the whole-cell pipet starting from 0 and increasing stepwise in 2-5 mm Hg increments to reach swelling < 10%. Cell voltage was held constant at -40 mY. Pressure steps had a rise time of -1 s and cell diameters were measured shortly thereafter (from video tape) for stiffness estimations. Membrane tension, T M was calculated from Laplace law for a sphere as: P·d (1) TM =-4where P = applied pressure from the pipet and d = instantaneous cell diameter. Cell lysis occurred at pressures> 20 mm Hg and was identified as sudden vesiculation and loss of cell contour, usually accompanied by a visible bursting of the membrane, and if recording electrode was attached, sudden increase in leak conductance. Surface areas and volumes were determined for each applied pressure, calculated with 41tr2 and 41tr3/3 (where r is half of measured
Rheology of mesangial cells
Vol. 34, No.6
391
A. Whole Patched Cell
Stretch
t am (Kw) Lei =constant B. Isolated Cell
Stretch
Fig. 1. Experimental Schema. (A) Components of regulatory volume decrease (RVD) in a whole-patched mesangial cell. In the case of wholecell recording mode, the intracellular concentration is dialyzed by pipet solution. establishing a stable cytosolic osmolarity. Addition of hyposmotic solutions results in a stable osmotic gradient (1:Q - l:Co) causing the inflow of water Qrn that is dependent on membrane permeability. K w ' Osmotic solutes can flux through membrane transporters sensitive to membrane stretch, but cannot change roo (B) Components of RVD in isolated mesangial cells. In hyposmotic conditions. osmolarity of external media, rCa, is approximately constant, while the intracellular osmotic concentration, 1:Ci. is time-dependent. Loss of cellular osmotic solutes is mediated through the membrane transporter sensitive to membrane stretch. Restoring of normal cell volume is achieved by water exit, Qrn, determined by water permeability (Kw) and by hydraulic pressures, Pi. derived from increase in cell membrane stretch.
diameter), respectively. The average area expansion modulus, KA of Me was estimated from the tangential slope of the membrane tension vs. cell area curve, according to Evans and Skalak (1979):
392
(2)
Rheology of mesangial cells
Vol. 34, No.6
T KA=- M
6a where 6a is fractional area change. A second protocol involved whole-cell recording during external superfusion with hyposmotic media. In these experiments, water flux, Qm was determined by the total solute concentration gradient, 6C(t) =: ~[Cj(t)]- ~[Co (t)], water permeability, Kw, surface area, and solute effiux induced by stretch caused by swelling (see also Fig. lA). For this protocol, the culture dish was superfused at a rate of 100 ul/s from a perfusion pipet with a tip diameter of 300 11m located 4 mm from the cell. Extracellular osmolarity was reduced by adding aliquots (100-300 ilL) of hyposmotic solution composed of control Ringer's with NaCl omitted. To raise extracellular osmolarity a hypertonic Ringer's (300 mM NaCl) was added. Extracellular osmolarity was changed stepwise from a baseline of 328 mOsm Kg-I to final values within the range 250-360 mOsm Kg-I. A third protocol involved pre-incubation of cells in hyposmotic media for - 20 min prior to whole-eell recording. This protocol provided hyposmotic challenge by intracellular dialysis with a relatively hypertonic solution (Pusch and Neher, 1988). The fourth protocol was visual observation of isolated cells exposed to hyposmotic solutions. In isolated cells, depicted in Fig. IB, water flux, Qm is determined by the total solute concentration gradient, 6C(t) == ~[Cj(t)]- ~ [Co (t)], water permeability, Kw, surface area, internal pressure, Pj, and solute effiux induced by stretch caused by swelling. Results
Electrical response to isosmotic swelling Figure 2 shows typical volume and conductance responses to internally applied pressures. The MC was gradually- pressurized from the whole-eell pipet at a rate of approximately 0.6 mm Hg s-I from 0 (up arrow) to 60 mm Hg in 100 s, Membrane conductance (middle trace) increased with the pressure, reaching a maximum of 4.2 nS at the maximum pressure of 60 mm Hg and rapidly returned to baseline of 1.6 nS upon release of pressure (down arrow). Swelling (top trace) accompanied pressurization, however it transiently declined, suggesting active volume regulation (see Discussion). After release of pressure, volume and conductance immediately returned to baseline, and did the same for repeats in several cells. These results therefore provide direct evidence for swelling induced current.
Mechanical response to isosmotic swelling Swelling of MC was estimated in the above experiments by plotting the instantaneous area expansion against calculated membrane tension (see Methods) as shown in Fig. 3 for 6 cells. Since MC were about five times more expandable than RBC prior to lysis, the viscoelastic properties of the cell could be examined at both small and large deformations. The overall polynomial curve fit of Fig. 3 suggests a bimodal response: at small to moderate deformations there is an apparent loss of stiffness where the curve flattens, but at larger deformations stiffness returns. The extrapolated elastic component of area expansivity at the origin was 450 dyne/em, identical to that previously measured in human RBC. The behavior at larger deformations is consistent with the nonlinear Maxwell model as proposed by Dong et al. (1991) for neutrophils. Specifically, these authors found cytoplasmic viscosity to be
393
Rheowgy of mesangial cells
Vol. 34, No.6
••.,
15 . - - - - - - - - - - - - - - - - - - - - , ~ ~
10
1
,
0 -5
.
• • ... . . . . ..,
]' 5
'
.,. ~
•
,
•
~
•
L-----l_---l._--L._-"-_..L.._..l-_.l-._L..----l
5..-------------------, ~ .Mt"A.4~ ,
.
~----.&.
..,; .&. ...... . ...""
.. . ... .... . . . .. l'\ "
..
I .0
"
. ..... _......t ..· ·• -40
·20
0
20
,
40
60
80
100
120
140
Time(sec)
Fig. 2. Responses to isosmotic swelling . A mesangial cell (MC) , as in Fig. lA, held at -40 mY, was gradually pressurized at a rate of approximately 0.6 mm Hg s-l from 0 (up arrow) to 60 mm Hg in 100 s (bottom trace). Membrane conductance (middle trace) increased with the pressure, reaching a maximum of 4.2 nS at the maximum pressure of 60 mm Hg, Conductance rapidly returned to baseline of 1.6 nS upon release of pressure (down arrow). Swelling (top trace) also accompanied pressurization initially, however it varied as expected, since a regulatory volume decrease (RVD) response is active in MG.
relatively low for small deformations, and that it increased 10-fold at large deformations, due mainly to the involvement of the stiff nucleus. Based on the model of Dong et al., Fig. 3 shows incremental expansion due to viscosity during the pressure steps, with a magnitude inversely proportional to viscosity, as expected. Thus, swelling in Me is a viscoelastic process consistent with a nonlinear Maxwell model.
Responses to hyposmotic swelling In isolated cells, under standard conditions, reducing extracellular osmolarity by 40-60 mOsm Kg-I caused swelling of 10-.20%, that was followed by a complete regulatory volume decrease (RVD) within 1-2 min. A typical example of the normal swelling and RVD response is shown in Fig. 4, wherein swelling of about 15% was followed by return to baseline within about 1 min. In contrast, cells treated with Gd3+ (0.5 mM, n = 5) prior to hyposmotic exposure were unable to regulate and lysed within seconds, as shown in Fig.5A and B. These results suggested the role of a Gd 3+-sensitive solute transport system in RVD, that was further explored using electrophysiological techniques (described below).
394
Vol. 34, No.6
Rheology of mesangial cells
70
60
E ~
50
~ 40
~
c: .Q III c:
30
e 20 10 O-f---.---"""T"""---,-------r-...,.----, 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Fractional Area Expansion (!!.AlA)
Fig. 3. Isosmotic swelling during pressurization. Six mesangial cells (MC), as in Fig. lA, were pressurized in steps. and their membrane tensions were calculated using Eq. (2). Fractional area expansions (M/A) were measured from digitized images at each pressure. Points on graph represent averages of six Me, with bars denoting the standard deviations of both area expansion and tension. The tangential average area expansion modulus at the origin was calculated to be 450 dyne/em. A third order polynomial was fit to the data. 120
-o~
t\ . /
Q)
E :J
(5
110
>
..\/
Q)
o
~ ~ Q) 100 a:
._. • -50
o
.)
\ . '.-. 50
100
Time (sec) Fig. 4. Regulatory volume decrease (RVD) in isolated mesangial cell (MG). Volume response of MC in standard Ringer's exposed to 60 mOsm ~l gradient at time = 0 (-e-). Prior to hyposmotic exposure. volume measurements varied between 100 and 105%. Within 10 s of hyposmotic exposure, volume peaked at 115% and began to sharply decline. with some oscillation. and returned to baseline by 60 s,
Rheology of mesangial cells
Vol. 34, No.6
(a)
Iso-osmotic 3· Gd present
(b)
10 seconds after hypoton ic stress 3+ Gd present Fig. 5. Antagonism of re~lalory volume decrease (RVD) in isolated mesangial cells (MC) by Gd ' - (A) Micrograph shows an Me in isotonic media . ma intaining stable vo lume in the presence of Gd3-+ (0.5 mM L- 1 ) . (B) Same Me about 4 s after exposure to 60 mOsm Kg-I gradient. that caused rapid lysis. RVD was apparently slowed or antagonized by Gd 3-+ .
395
396
Rheology of mesangio1 cells
Vol. 34, No.6
Cells cannot regulate volume normally when attached to a whole-cell electrode, since their internal concentration is clamped by the pipet, but fluxes of ions can be measured, as the cell attempts to regulate. Membrane ionic conductance during hyposmotic exposure was thus used as an indirect measure of solute flux during regulation. In isosmotic conditions, the slope conductance of MC was nearly linear, averaging 0.9 ± 0.5 nS (n = 9 cells) over the range between -120 mV and 40 mV, and was stable within ± 10% during the 10-20 min equilibration period before osmotic challenges. Families of I-V tracings for a typical experiment are shown in Fig. 6 for control (A), 5 min after a reduction of the extracellular osmolarity to 260 mOsm Kg-I (B), and 5 min after addition of GdCl3 (0.5 mM) to the hyposmotic media (C) . The corresponding I-V curves for this experiment (panel D) show that control slope conductance was 0.8 nS, linear, and that after hyposmotic exposure it outwardly rectified, increasing to 1.7 and 2.8 nS in the inward and outward directions, respectively. Inhibition of the conductance increase by GdC l3 (panel C) was consistent with the interpretation of Fig. 5B and will be further discussed below. Swelling induced (SI) current was carried by cations since substitution of internal CI- by gluconater did not shift reversal potential but increased conductance (Huang et al., 1997). The reversal potential of -20 mV indicated a constant permeability ratio, PK+/PNa+, of 3, which did not change during hyposmotic stimuli. Summarized results from six cells are shown in Fig. 7. The SI conductance consisted of a significant increase in the average slope conductance (Iso/Hypo) in the outward direction from 1.2 ± 0.5 to 4.2 ± 3 nS (p < 0.01), as shown in the right panel. Slope conductance in the inward direction also increased, but not significantly. Outward and inward conductances for each cell were estimated by fitting lines to points on either side of the reversal potential. To test the pharmacology of SI current, the conductances of six cells that had been exposed to hypotonic media were measured immediately before adding GdCI,3 (0.5 mM), as shown in the bars marked "Hypo" in the left panel. Within 5 min, SI outward conductance was significantly reduced by GdCI 3 (Hypo + Gd 3 +) from 5.7 ± 1.7 nS to 3.7 ± 2.4 nS (p < 0.02). Cells in hyposmotic media not treated with GdCl3 maintained stable high conductance. Since the concentration of GdCl3 used herein was well above threshold blocking dose for stretch-activated channels (SAC) (Yang et al., 1990), subsequent dose response experiments were done (unpublished), and it was found that 20 J.1M GdCl3 was sufficient to block SI current. Agents that failed to block or prevent SI current included externally applied TEA (16 mM), TTX (10 J.1M), BaCl2 (0.1 mM), or amiloride (0.1 mM). Internal TEA was a common pipet component, and neither ATP nor cAMP influenced SI current. These results provide evidence that SAC are involved in SI current. Hyposmotic challenge of MC was done also by raising their intracellular osmolarity with a relatively hypertonic pipet solution (see Materials and Methods) . This was done by pre-incubating MC in hyposmotic media, 260-270 mOsm Kg-I, for at least 10 min, followed by initiation of recording and dialysis with 328 mOsm Kg-I internal solution. Recordings so obtained on four cells showed that initial conductance was near baseline isosmotic levels of 0.9 nS, but rapidly increased several-fold during the first 5 min. These results suggest that MC equilibrate with hyposmotic media within minutes, as expected if SI current effectively nullifies the osmotic gradient. (See Discussion below.)
~
:-
~
!'" Z ?
HypotonIc
c::n
B
40mV
eontrol
200
50 mS
1100 pA
Hypo
I (pA) 0
mV
D
.1f.
.12' •
-II
·11
i i '
.,
I
•
•
I
o
o
•
o
0
0
•
II. 15
0
0
0
0
II
.
•
• 40
•
·20 it
·40
o 0
i
~
~
0
100
~
H~po+Gd
Iso
I
20
~
iii
aQ
~Q'
-~
0
III
t;"
·100
0
o ·200
Fig. 6. Swelling-induced (SI) current in standard Ringer's, I-V families are shown in control (A), after hyposmotic exposure (B), and with hyposmotic + Gd~+ exposure (C) , I-V families are superimposed currents elicited at voltages from -120 mV to 40 mV in steps of 10 mV, with holding voltage of -40 mV, (D) Shows I-V plots corresponding to control (open squares), hyposmotic (open circles) and hyposmotic with Gd~. 0.5 mM (x), For clarity, lines are not drawn through the points. Pipet composition was FJ-l (Table 1).
~
<0
..:r
398
Vol. 34, No.6
Rheology of mesangial cells (. ) : ~, hyperpolarized regions (+) : 9d, depolarized regions
8
( +)
D
lso
~ Hypo ~ Hypo+Gd 3 + ( +)
7
Vi'
(+)
6
.5.5 CD
u
B 4 u :::J
"g 3 o
o 2
o
(.) (+)
CO lso
-
Hypo
Hypo -
Hypo+Gd
3
+
Fig. 7. Swelling-induced conductance summary. Bars labeled (-) represent conductances in the hyperpolarized region, gh (inward conductance), and bars labeled (+) represent average slope conductance in the depolarized region, gd (outward conductance) , with lines representing standard deviations of gh and gd' The left panel shows the average slope conductance in standard Ringer's before (left) and after (right) being exposed to hyposmotic gradient of 40-60 mOsm Kg""l for about 5 min (p < 0.01, n = 6). The right panel shows the average slope conductance of six cells before (left) and after (right) being exposed to Gd 3+ (0.5 mM) for about 5 min in the presence of an hyposmotic gradient of 40-60 mOsm Kg""l (p < 0.05. n = 6). Note : Data in the two panels are two independent series from different cells.
Discussion Mechanical responses to swelling The rather large scatter in the data of Fig. 3 may be partly caused by video measurement errors and by the transient volume oscillations seen in Fig. 2, however, area measurements were made within a few seconds of pressure steps, to minimize the influence of possible regulatory responses. Large scatter in area expansivity measurements of human red blood cells (RBC) was also reported and attributed to irrecoverable volume changes (Waugh and Evans, 1979). Their study measured a linear elastic modulus for RBC of 450 dyne/em, for area expansions ~ 1.2%. This measurement corresponds almost exactly to the tangent at the origin of Fig. 3, and thus it can be concluded that cultured Me membranes have an area expansivity resembling that of human RBe. The apparent stiffness of MC (Fig. 3) depended on the degree and duration of deformation since membrane tension was considerably damped during low to moderate swelling, but not at larger swelling. The bimodal behavior may reflect separate viscous components in the cytoplasm that can be described as a nonlinear Maxwell model (Dong et al., 1991). Other components of RVD may also contribute to the observed mechanical behavior, including cytoskeletal adjustments during swelling (Funai et al., 1993; Bogen,
Vol. 34, No.6
Rheology of mesangial cells
399
1987) or active contraction (Ausiello et al., 1980). Whether these viscoelastic properties of isolated MC are also exhibited in situ is not known, however, isolated MC exhibit many of their in vivo properties (Ardaillou, 1996).
Electrical response to sweUing Our results describe a solute transport system based on effiux of K+ carried through swelling-activated ionic channels. The solute transport is based on SI current, whose conductance and pharmacology are consistent with those recorded in proximal tubule cells of the kidney (Robson and Hunter, 1994) and with those of the SAC in MC, including activation by hyposmolarity (Craelius et al., 1993). The critical dependence of RVD on K+ conductance increase has been previously shown for cultured astrocytes (Pasantes-Morales et al., 1994), and for renal epitheIoid cells, whose K+ conductance transiently increases 64-fold after hyposmotic challenge (Kersting et al., 1991). Other solute effiux pathways may also playa role in RVD responses of MC, including K+ efflux through furosemide-sensitive pathways (Homma et al., 1990). Based on their estimated density on the cell of 1/~m2 (Craelius et al., 1989) , our maximal response of about 20 nS could be produced by full activation of 4% of available SAC. Thus, the present data can be added to the growing body of direct evidence relating single stretch-activated channel currents with whole-cell currents (Zhang et al., 1994; Ross et al., 1994; Gustin, 1991; Sachs, 1991; Chan and Nelson, 1992; Sackin, 1987; Tseng, 1992; Hagiwara et al., 1992; Chan et al., 1994), in contrast to the one reported failure (Morris, 1991) . Limitations of patch-clamp experiments are the restriction to isolated cells, buffering of [Ca 2+]i and pH, and clamping of voltage, that probably shortcircuit some of the normal components of RVD. In cells exposed to hyposmolarity, [Ca 2+]\ rises dramatically (Tsukahara et al., 1994; McCarty and O'Neil, 1992) and may trigger a cascade of responses to swelling. One likely response is activation of CI- channels, opening a major exit pathway for CI(Hoffmann and Simonsen, 1989; Christensen, 1987) . It is also likely that the internally released Ca2+ would initiate contraction using the actin-myosin network (Ausiello et al., 1980), thereby resisting osmotic swelling. In addition to triggering a cascade of biochemical and mechanical events in the cell, the freed Ca 2+ could also balance a small fraction of charge efflux, while simultaneously reducing the net osmolyte flux (McCarty and O'Neil, 1992) and similar effects could result from RVD-induced acidification, (Hoffmann, 1987). Voltage clamping also prevents changes in membrane stiffness that are produced by voltage gradients (Katnik and Waugh, 1990). None of these conditions of patch clamping, however, apply to our visual observations of rapid hypotonic lysis in cells pre-treated with Gd!l+ (Fig. 2). Thus it can be concluded that K+ effiux occurs with or without Ca 2+ flux and is important to RVD. The degree to which SI current may be transiently electrogenic depends on its kinetics in relationship to counterbalancing anions such as CI- or HCO!!-, and is not answered by our results. Implications The outwardly rectifying current induced by swelling, seen in Fig. 6, represents a large efflux of K+ that could accumulate within certain restricted environments and serve as a depolarizing signal to neighboring cells. Further studies are required to quantify this electrochemical signal arising both within and outside the glomerulus, to determine its significance.
400
Rheology of mesangial cells
Vol. 34, No.6
The bimodal nature of MC viscoelasticity during deformations if exhibited in situ, may be a factor in their long-term regulation of the glomerular environment. An extensive cytoskeleton suggests the ability of MC to internally pressurize during high swelling pressure (Latta et aL, 1960) . The relationship of cytoplasmic viscosity to applied pressure implies that the transference of forces to the cytoskeleton and nucleus becomes more effective as pressure and deformation increase. This phenomenon may explain, in part. the ability of large periodic deformations to stimulate extracellular matrix production by MC through nuclear activation, that is believed to be an adaptive response to mechanical overload (Riser et al., 1992; Akai et al., 1994). The dual viscosities appearing in response to small and large deformations could therefore contribute to the acute vs. chronic adaptations of MC to environmental stress.
Acknowledgments We would like to thank Anthony Dawidowitz for his help with the experiments. These studies were funded by the American Heart Association, New Jersey Affiliate (W. Craelius), and a Merit Review Grant from the Department of Veterans Affairs.
References ARDAILLOU, R (1996) . Biology of glomerular cells in culture. Cell Bioi. Toxicol. 12, 257-261. AKAI, Y., HOMMA, T., BURNS, K.D., YASUDA, T., BADR, K.F., and HARRIS, RC. (1994). Mechanical stretch/relaxation of cultured rat mesangial cells induces protooncogenes and cyclooxygenase. Am. J Physiol. 267 (Cell Physiol. 26): C482-G490. AUSIELLO, D.A., KREIBERG, J.I., ROY, C., and KARNOVSKY, M.J. (1980). Contraction of cultured rat glomerular cells of apparent mesangial origin after stimulation with angiotensin II and arginine vasopressin. J CUn. Invest. 65, 754-760. BLANTZ, RC., GABBAI, F.B., TUCKER, BJ., YAMAMOTO, T., and WILSON, C.B. (1993) . Role of mesangial cell in glomerular response to volume and angiotensin II. Am. J Physiol. 264, F158-F165. BOGEN, D.K. (1987) . Strain energy descriptions of biological swelling I: Single fluid compartment models. J Biomech. Eng. 109, 252-256. CHAN, H-C., and NELSON, D. (1992). Chloride-dependent cation conductance activated during cellular shrinkage. Science 257, 669-671. CHAN, H.C., ru, W.O., CHUNG, Y.W., HUANG, S.J., CHAN, P.S.F., and WONG, P.Y.D. (1994). Swelling-induced anion and cation conductances in human epididymal cells. J Physiol. 478, 449-460. CHRISTENSEN, O. (1987). Mediation of cell volume regulation by Ca 2+ influx through stretch-activated channels. Nature 330, 6tH>8. CRAELIUS, W., EL-SHERIF, N., and PAIANT, C.E. (1989). Stretchactivated ion channels in cultured mesangial cells. Biochem. Biophys. Res. Comm 159, 516-521.
Vol. 34, No.6
Rheology of mesangial cells
401
CRAEUUS. W., ROSS, MJ., HARRIS, D.R. , CHEN, V.K., and PAlANT, C.E. (1993). Membrane currents controlled by physical forces in cultured mesangial cells. Kidney Int. 43, 535-543. DONG, C, SKALAK, R, and SUNG, K.-L.P. (1991). rheology of passive neutrophils. Biorheology 28, 557-567.
Cytoplasmic
EVANS, EA., and SKALAK, R (1979). Mechanics and Thermodynamics oj Biomembranes. CRC Press , Boca Raton. FERDOWS. M., GOLCHINI, K., ADLER S.G., and KURTZ, I. (1989). CI-/base exchange in rat mesangial cells: Regulation of intracellular pH. Kidney Int. 35, 783-789. FUNAI,J., FELDMAN, G., RABBANY, S., HAINES, D., SOMANATH, B., and MOHANlY, S. (1993). Intracellular pressure-volume responses to osmotic and hydrostatic stress: Studies in the giant barnacle muscle cells. FASEB 7, A1628. GUSTIN, M.C. (1991). Single-channel mechanosensitive currents. Technical comments. Science 253, 800. HAGIWARA, N., MASUDA, H., SHODA, R, and IRISAWA, H . (1992) . Stretch-activated anion currents of rabbit cardiac myocytes. J Physiol. 456, 285-302. HAMILL, O.P., MARlY, A., NEHER, E., SAKMANN, B., and SIGWORTH, FJ. (1981) . Improved patch-clamp techniques for highresolution current recording from cells and cell-free membrane patches. Pflugers Arch. 391, 85-100. HOFFMANN, E.K. (1987). Volume regulation in cultured cells. In: Current Topics in Membrane Transport. R Gilles, A Kleinzeller, and L. Boles, Eds., Academic Press. New York, pp. 125-180. (1989) . Membrane HOFFMANN, E.K., and SIMONSEN, L.O. mechanisms in volume and pH regulation in vertebrate cells. Physiol. Rev. 69, 315-382. HOMMA, T., HOOVER, L.R, and HARRIS, RC. (1990). Loop diuretic-sensitive potassium flux pathways of rat glomerular mesangial cells. Am. J Physiol. 258, C862-870. HUANG, C.-]., PALANT, C.E., and CRAELIUS, W. (1997). Low CIinduces a cation conductance in rat mesangial cells. Kidney and Blood Pressure (in press). IVERSEN, B.M., KHAM, F.I., MATRE, K., MORKRID, L., HORVEI, G., BAGCHUS, W., GROND, J., and OFSTAD. J. (1992). Effect of mesangiolysis on autoregulation of renal blood flow and glomerular filtration rate in rats. Am. J Physiol. 262, F361-366. JOHNSON, RJ. (1997). Have we ignored the role of oncotic pressure in the pathogenesis of glomerulosclerosis? Am. J Kidney Dis. 29,147-152.
402
Rheology of mesangial cells
Vol. 34, No.6
KATNIK, C., and WAVGH, R (1990). Alterations of the apparent area expansivity modulus of red blood ceIl membrane by electric fields. Biophys.] 57, 877-882. KERSTING, V., WOJNOWSKI, L, STEIGNER, W., and OBERLEITNER, H. (1991). Hypotonic stress-induced release of KHCOs in fused renal epitheloid (MOCK) cells. Kidney Int. 39, 891-900. LATTA, H., MAVNSBACH, A.B., and MADDEN, S.C. centrolobular region of the glomerulus studied microscopy.] Ultrastruct. Mol. Struct. Res. 4, 455-472.
(1960). The by electron
MCCARlY, N., and O 'NEIL, RG. (1992). Calcium signaling in cell volume regulation. Physiol. Rev. 72, 1037-1061. MORRIS, C.E. (1991). Single-channel mechanosensitive currents. Technical Comments. Science 253,801. PASANTES-MORALES, H., MURRAY, RA., LILJA, L., and MORAN, J. (1994). Regulatory volume decrease in cultured astrocytes. I. Potassium- and chloride-activated permeability. Am. J Physiol. 266, C165-C171. PVSCH, M., and NEHER, E. (1988). Kinetics of loading small cells with various compounds by use of patch pipets. Pflugers Arch. 408, (Suppl. I), 338. RISER B.L., CORTES, P., ZHAO, X., BERNSTEIN, J., DUMLER, F., and NARINS, R.G. (1992). Intraglomerular pressure and mesangial stretching stimulate extraceIlular matrix formation in the rat. J Clin. Invest. 909, 1932-1943. ROBSON, L., and HUNTER, M. (1994). Role of cell volume and protein kinase C in regulation of a CI- conductance in single proximal tubule cells of Rana temporaria.] Physiol. 480, 1-7. ROSS, P.E., GARBER, S.S., and CAHALAN, M.D. (1994). Membrane chloride conductance and capacitance in Jurkat T lymphocytes during osmotic swelling. Biophys.] 66, 169-178. SACHS, F. (1991). Single-channel mechanosensitive currents. Technical comments. Science 253, 800. SACKIN, H. (1987). Stretch-activated potassium channels in renal proximal tubule. Am.]. Physiol. 253, FI253-FI262.
SAKAI, T., and KRIZ, W. (1987). The structural relationship between mesangial cells and basement membrane of the renal glomerulus. Anat. Embryol. 176, 373-386. SCHLONDORFF, D. (1996). Roles of mesangium in glomerular function. Kidney Int. 49, 1583-1585. SCHMID-SCHONBEIN, G.W., KOSAWADA, T., SKALAK, R., and CHIEN, S. (1995). Membrane model of endothelial cells and
Vol. 34, No.6
Rheology of mesangial cells
403
leukocytes. A proposal for the origin of a cortical stress. J Biomech. Eng. 117,171-178. STOCKAND, J.D., and SANSOM, S.C. (1997). Regulation of filtration rate by glomerular mesangial cells in health and diabetic renal disease. Am. J Kidney Dis. 29, 971-981. (1992). Cell swelling increases membrane TSENG, G-N.Y. conductance of canine cardiac cells: Evidence for a volume-sensitive CI channel. Am. J Physiol. 262, C105~1068. TSUKAHARA, H., KRIVENKO, Y., MOORE, L.C., and GOLIGORSKI, M.S. (1994). Decrease in ambient [CI-] stimulates nitric oxide release from cultured rat mesangial cells. Am. J Physiol. 267, F190-F195. WANG, X., AUKLAND, K., BOSTAD, L., and IVERSEN, B.M. (1997). Autoregulation of total and zonal glomerular filtration rate in spontaneously hypertensive rats with mesangiolysis. Kidney and Blood Pressure Res. 20, 11-17. WAUGH, R, and EVANS, E.A. (1979). Thermoelasticity of red blood cell membrane. Biophys. J 26, 115-132. YANG, X-C., and SACHS, F. (1990). Characterization of stretchactivated ion channels in xenopus ooeytes. J Physiol. 431, 103-122. ZHANG, J., RASMUSSON, RL., HALL, S.K., and LIEBERMAN, M. (1994). A chloride current associated with swelling of cultured chick heart cells. J Physiol. 472, 801-82.
Received 28 July 1997; accepted in revised form 19 December 1997.
Biorheology,Vol.M, No.6, pp. 118'7-4011, 199'7 Copyright C 1998 Ebevier Science Ltd Printed in the USA. All rights reserved ()()()6.'55X/97 $17.00 + .00
Pergamon PH SOOO6-355X(98)00023-7
RHEOLOGICAL BEHAVIOR OF RAT MESANGIAL DURING SWELLING IN VITRO
w. CRAELIUSt, CJ. HUANGt,
CELLS
H. GUBER~, AND GE. PALANT~*
tDeparunent of Biomedical Engineering. Rutgers University. Piscataway. NJ; tNephrology Service. Department of Veterans Affairs Hospital. Brooklyn, NY; *University of Nevada School of Medicine, Reno. NV, USA Reprint requests to: William Craelius, Department of Biomedical Engineering, Rutgers University, Piscataway, NJ, 08854 USA; Fax: 732-445-3753; e-mail: [email protected]
ABSTRACT The response of cells to mechanical forces depends on the rheological properties of their membranes and cytoplasm. To characterize those properties, mechanical and electrical responses to swelling were measured in rat mesangial cells (MC) using electrophysiologic and video microscopic techniques. Ion transport rates during hyposmotic exposures were measured with whole-cell recording electrodes. Results showed that cell swelling varied nonlinearly with positive internal pressure, consistent with a viscoelastic cytoplasm. The extrapolated area expansivity modulus for small deformations was estimated to be 450 dyne/cm. Cell swelling, caused either by positive pipet pressure or hyposmotic exposure (40-60 mOsm Kg-1), rapidly induced an outwardly rectifying membrane conductance with an outward magnitude 4-5 times the baseline conductance of 0.9 ± 0.5 nS (p < .01). Swelling-induced (SI) current was weakly selective for K+ over Na", partially reversed upon return to isotonicity, and was antagonized by O.S mM GdCls (p < 0.02; n =6). Isolated cells treated with GdCIs rapidly lysed after hypotonic exposure, in contrast to untreated cells that exhibited regulatory volume decrease (RVD). Our results indicate that volume regulation by Me depends upon a large swelling-induced K+ efflux, and suggest that swelling in MC is a viscoelastic process, with a viscosity dependent on the degree of swelling. © 1998 Elsevier Science Ltd
Introduction Mesangial cells (Me) are components of a reflex that adjusts filtration in response to pressure and volume changes within the glomerulus (for review see Schlondorff, 1996). This mechano-reflex can occur dynamically, through glomerular vasocontrol (Iversen et al., 1992; Blantz et al., 1993; Wang et al.• 1997; Stockand and Sansom, 1997) or chronically, through production of basement membrane (Riser et al., 1992). While details of the reflex are KEYWORDS:
Cell swelling; kidney; membrane stiffness; stretch-activated channels
387
388
Rheology of mesangial cells
Vol. 34, No.6
incomplete, the abundance of mechanically sensitive ion channels in both rat and human MC (Craelius et al., 1989; 1993; Palant, unpublished) and the intimate association of MC with glomerular capillaries and the basement membrane (Sakai and Kriz, 1987) suggest the participation of MC as mechanotransducers. Specifically it can be hypothesized that MC, responding to intraglomerular pressure signals , influence the electrochemical environment and hence their own contractile state, thereby modulating filtration through vasocontroI. To assess the participation of MC in glomerular reflexes, their rheological behavior in response to applied forces requires understanding. The present study describes in particular the mechanoelectrical responses of MC to swelling forces. The mechanotransduction ability of MC may relate not only to glomerular regulation as described above, but also to cellular regulation in the presence of pathological oncotic pressures in the glomerulus Uohnson, 1997).
Materials and Methods Mesangiai cell preparation Mesangial cells were isolated from rat glomeruli using the graded sieving technique. Male Sprague-Dawley rats (Camm Research, Wayne, NJ, USA) weighing 50-75 grams were anesthetized with Metofane, decapitated, bled and placed in 95% ethanol. Kidneys from 8 to 10 rats were excised under sterile conditions using blunt dissection to remove surrounding capsules and serosae. The renal cortices were isolated and purified by sieving through pore sizes of 250, 150 and 75 11m (Thomas Scientific, Swedesboro, NJ, USA). The filtrate, consisting of isolated glomeruli, was then collected atop a sieve with 45 urn pore size and incubated at 37°C for 30 min in 2 ml of collagenase-Type IV, 130 units/mg; clostripain, 0.85 units/mg; and non-specific proteinase, 27 units/mg (Sigma, St. Louis, MO, USA). The specificity of glomerular isolation was > 95%, by microscopic examination. Following enzyme washout, glomeruli were incubated in 75 ml flasks with a 1:1 solution of RPMI + 20% fetal calf serum (Gibeo Labs, Grand Island, New York, USA) and conditioned media prepared from 3T3 fibroblasts (American Type Culture Collection, Rockville, Maryland). The fibroblasts were grown in DMEM + 10% fetal calf serum + 1 % penicillin-streptomycin (Gibco) . The media was refreshed after the first 24 h and was thereafter half-exchanged at two-day intervals until cells became confluent. Identification of spindle-shaped MC was positive under phase contrast microscopy with 2 d of plating. Immunological identification was done by exposing formalin-fixed MC to a-actin antibodies from smooth muscle (Sigma) and antibodies against myosin from smooth muscle and nonsmooth muscle (Biomedical Technologies, Austin, Texas, USA). After staining with immunoperoxidase, MC could be visualized by their strong reaction to anti-a-actin and anti-myosin antibodies. (A separate culture batch of MC was generously provided by Dr. S. Adler of Harbor-UCLA Medical Center.) The morphological and immunological characterization of these cells has been previously published (Ferdows et al., 1989) . For recordings, MC were plated in 30 mm plastic dishes at a density of < 20,000 cells/dish. Cells were studied in the third through the twentieth culture passage.
Electrolyte solutions Cultured MC were bathed at room temperature (22-23°C) for most experiments and at 37°C for the [K+]i measurements. Standard Ringer's composition was (in mM): NaC1150, KCI6. CaCI21.2, MgCI21.2, HEPES In (N-2-hydroxyethypiperazine-N'-2-ethanesulfonicacid) 10, glucose 5.5.
Vol. 34, No.6
389
Rheology of mesangial cells
addition, to block selected currents, some media contained: amiloride 0.1 mM, TEA-el (tetraethylammonium chloride) 16 mM, BaCl2 0.1 mM, TTX 10 J.1M, or GdCIg 0.1-1 mM. Mannitol was used to adjust osmolarities of external solutions to 328 mOsm Kg-I. Table 1 shows the two patch pipet solutions used. In addition, some pipet solutions contained Na2-ATP (l mM) or 8-BromocAMP (0.2 mM). All salts were purchased from Sigma. Osmolarities of electrolyte solutions were measured with a freezing point micro-osmometer (Precision Instruments, Natick, Massachusetts). The osmolarity of the extracellular control Ringer's was 328 mOsm Kg-I, a value within the range of osmolarities (320-350 mOsm Kg-I) of the culture media used for MC in prolonged incubation (see above). Osmolarity of standard electrode solutions was adjusted to be isosmotic with extracellular media (328 mOsm Kg-I). Electrophysiological measurements
Whole-ceIl patch clamp To characterize the conductance activated by hyposmotic stress, the whole-eeIl patch-elamping technique was used (Hamill et al., 1981). Patch-pipets were pulled from 100 J.11 Drummond Microcaps on a Flaming-Brown programmable puller (Sutter Instruments, San Rafael, California, USA), and were back-filled using fine plastic tubing. Cells were placed on the stage of an inverted phase contrast microscope and viewed at 600x magnification (Olympus IMT-2, Lake Success, New York, USA) and the pipet was advanced by a manipulator (Narishige U.S.A., Great Neck, New York, USA). Currents were recorded through patch pipets attached with a giga-ohm seal (Hamill et al., 1981). Tip resistances ranged from 0.8 to 2 MO, and giga-seals with the cell were formed rapidly after slight suction was applied to the pipet. Negative and positive pressures were applied through a side port in the pipet holder and measured with an electronic pressure gage (World Precision Instruments, Sarasota, Florida, USA). Whole-eell recording commenced after rupturing the patch with a suction pulse. Current records were low-pass filtered (-3 dB cutoff 300 Hz) and digitized at 1 ms intervals with 16-bit resolution (VR-ST, Instrutech, Mineola, New York, USA). Holding potential was -40 mV and each I-V sweep consisted of responses to voltage clamps of 200-300 ms applied at 1-2 s intervals beginning at -140 mV and stepping with 10 mV increments to +40 mY. Voltage
Table 1 Patch pipet solution composition Component
El-l (mM)
El-2 (mM)
NaCI
6.0 150.0
10.0
KCI K-Gluconate
135.0 10.0
TEA-CI CaCl2 MgQ 2
1.2
1.2
1.2
1.2
EGTA HEPES
11.0
11.0
10.0
10.0
390
Rheology of 11U!sangial cells
Vol. 34. No.6
errors due to series resistance were minimized using analog compensation. Currents were averaged during the period between 100-200 ms after onset of each voltage step. Currents, voltages, pressure signals, and video images were recorded on VHS tape (Model 420, Vetter, Inc., Reebersburg, PA, USA). Cell conductance was continuously monitored with repeated I-V sweeps after each osmolarity change and during positive pressure applications. Slope conductances for each cell were estimated for both inward and outward currents by fitting lines to both regions. Since reversal potential was about -20 mV, the inward conductance was calculated over the range between -120 and -20 mV, while the outward conductance was calculated over the range between -20 and 40 mV.
Cell volume and surface measurements Plastic dishes (35 mm) containing cultured cells were placed on the stage of an inverted phase contrast microscope and viewed at 600x magnification. Each experiment was monitored with video microscopy and recorded on video tape (Panasonic vHS). Video images were sampled from tape and digitized at 8-bit resolution with a Frame Grabber (Vlab Y/C, Macrosystems, Detroit, MI, USA). Cellular dimensions were quantified with an image processing program written on an Amiga 3000 computer, which has a resolution of 0.25 um/pixel, and measurement error was approximately 0.5 urn. The diameter of each cell was determined as the average of diameters measured at four axes. All cells were approximately spherical. The height of each cell was estimated as the vertical distance between focal planes of its top and bottom surfaces, measured with the focus vernier. Height error was approximately 2 urn. Protocols Four protocols were used to measure rheological responses to swelling in MC. The first protocol measured responses of MC to positive pressures in isosmotic media in whole-cell mode, as depicted in Fig. lAo In these experiments, external osmolarity, I[Co(t)], was equated to internal osmolarity, I[Cj(t)], that was effectively stabilized by dialysis from the pipet. The stretch of the cell during pressurization, Pi, caused water fluxes Qm and Qp. Measurements were made on cells slightly pre-swollen in hyposmolar media to ensure sphericity, and to minimize locally high membrane curvatures that have been described in endothelial cells (Schrnid-Schonbein et al; 1995). Confirmation of sphericity of cells prior to experiments was done by measuring their height-to-diameter ratio, which was 1.1 ± 0.25 in four cells. Positive pressures were applied to the whole-cell pipet starting from 0 and increasing stepwise in 2-5 mm Hg increments to reach swelling < 10%. Cell voltage was held constant at -40 mY. Pressure steps had a rise time of -1 s and cell diameters were measured shortly thereafter (from video tape) for stiffness estimations. Membrane tension, T M was calculated from Laplace law for a sphere as: P·d (1) TM =-4where P = applied pressure from the pipet and d = instantaneous cell diameter. Cell lysis occurred at pressures> 20 mm Hg and was identified as sudden vesiculation and loss of cell contour, usually accompanied by a visible bursting of the membrane, and if recording electrode was attached, sudden increase in leak conductance. Surface areas and volumes were determined for each applied pressure, calculated with 41tr2 and 41tr3/3 (where r is half of measured
Rheology of mesangial cells
Vol. 34, No.6
391
A. Whole Patched Cell
Stretch
t am (Kw) Lei =constant B. Isolated Cell
Stretch
Fig. 1. Experimental Schema. (A) Components of regulatory volume decrease (RVD) in a whole-patched mesangial cell. In the case of wholecell recording mode, the intracellular concentration is dialyzed by pipet solution. establishing a stable cytosolic osmolarity. Addition of hyposmotic solutions results in a stable osmotic gradient (1:Q - l:Co) causing the inflow of water Qrn that is dependent on membrane permeability. K w ' Osmotic solutes can flux through membrane transporters sensitive to membrane stretch, but cannot change roo (B) Components of RVD in isolated mesangial cells. In hyposmotic conditions. osmolarity of external media, rCa, is approximately constant, while the intracellular osmotic concentration, 1:Ci. is time-dependent. Loss of cellular osmotic solutes is mediated through the membrane transporter sensitive to membrane stretch. Restoring of normal cell volume is achieved by water exit, Qrn, determined by water permeability (Kw) and by hydraulic pressures, Pi. derived from increase in cell membrane stretch.
diameter), respectively. The average area expansion modulus, KA of Me was estimated from the tangential slope of the membrane tension vs. cell area curve, according to Evans and Skalak (1979):
392
(2)
Rheology of mesangial cells
Vol. 34, No.6
T KA=- M
6a where 6a is fractional area change. A second protocol involved whole-cell recording during external superfusion with hyposmotic media. In these experiments, water flux, Qm was determined by the total solute concentration gradient, 6C(t) =: ~[Cj(t)]- ~[Co (t)], water permeability, Kw, surface area, and solute effiux induced by stretch caused by swelling (see also Fig. lA). For this protocol, the culture dish was superfused at a rate of 100 ul/s from a perfusion pipet with a tip diameter of 300 11m located 4 mm from the cell. Extracellular osmolarity was reduced by adding aliquots (100-300 ilL) of hyposmotic solution composed of control Ringer's with NaCl omitted. To raise extracellular osmolarity a hypertonic Ringer's (300 mM NaCl) was added. Extracellular osmolarity was changed stepwise from a baseline of 328 mOsm Kg-I to final values within the range 250-360 mOsm Kg-I. A third protocol involved pre-incubation of cells in hyposmotic media for - 20 min prior to whole-eell recording. This protocol provided hyposmotic challenge by intracellular dialysis with a relatively hypertonic solution (Pusch and Neher, 1988). The fourth protocol was visual observation of isolated cells exposed to hyposmotic solutions. In isolated cells, depicted in Fig. IB, water flux, Qm is determined by the total solute concentration gradient, 6C(t) == ~[Cj(t)]- ~ [Co (t)], water permeability, Kw, surface area, internal pressure, Pj, and solute effiux induced by stretch caused by swelling. Results
Electrical response to isosmotic swelling Figure 2 shows typical volume and conductance responses to internally applied pressures. The MC was gradually- pressurized from the whole-eell pipet at a rate of approximately 0.6 mm Hg s-I from 0 (up arrow) to 60 mm Hg in 100 s, Membrane conductance (middle trace) increased with the pressure, reaching a maximum of 4.2 nS at the maximum pressure of 60 mm Hg and rapidly returned to baseline of 1.6 nS upon release of pressure (down arrow). Swelling (top trace) accompanied pressurization, however it transiently declined, suggesting active volume regulation (see Discussion). After release of pressure, volume and conductance immediately returned to baseline, and did the same for repeats in several cells. These results therefore provide direct evidence for swelling induced current.
Mechanical response to isosmotic swelling Swelling of MC was estimated in the above experiments by plotting the instantaneous area expansion against calculated membrane tension (see Methods) as shown in Fig. 3 for 6 cells. Since MC were about five times more expandable than RBC prior to lysis, the viscoelastic properties of the cell could be examined at both small and large deformations. The overall polynomial curve fit of Fig. 3 suggests a bimodal response: at small to moderate deformations there is an apparent loss of stiffness where the curve flattens, but at larger deformations stiffness returns. The extrapolated elastic component of area expansivity at the origin was 450 dyne/em, identical to that previously measured in human RBC. The behavior at larger deformations is consistent with the nonlinear Maxwell model as proposed by Dong et al. (1991) for neutrophils. Specifically, these authors found cytoplasmic viscosity to be
393
Rheowgy of mesangial cells
Vol. 34, No.6
••.,
15 . - - - - - - - - - - - - - - - - - - - - , ~ ~
10
1
,
0 -5
.
• • ... . . . . ..,
]' 5
'
.,. ~
•
,
•
~
•
L-----l_---l._--L._-"-_..L.._..l-_.l-._L..----l
5..-------------------, ~ .Mt"A.4~ ,
.
~----.&.
..,; .&. ...... . ...""
.. . ... .... . . . .. l'\ "
..
I .0
"
. ..... _......t ..· ·• -40
·20
0
20
,
40
60
80
100
120
140
Time(sec)
Fig. 2. Responses to isosmotic swelling . A mesangial cell (MC) , as in Fig. lA, held at -40 mY, was gradually pressurized at a rate of approximately 0.6 mm Hg s-l from 0 (up arrow) to 60 mm Hg in 100 s (bottom trace). Membrane conductance (middle trace) increased with the pressure, reaching a maximum of 4.2 nS at the maximum pressure of 60 mm Hg, Conductance rapidly returned to baseline of 1.6 nS upon release of pressure (down arrow). Swelling (top trace) also accompanied pressurization initially, however it varied as expected, since a regulatory volume decrease (RVD) response is active in MG.
relatively low for small deformations, and that it increased 10-fold at large deformations, due mainly to the involvement of the stiff nucleus. Based on the model of Dong et al., Fig. 3 shows incremental expansion due to viscosity during the pressure steps, with a magnitude inversely proportional to viscosity, as expected. Thus, swelling in Me is a viscoelastic process consistent with a nonlinear Maxwell model.
Responses to hyposmotic swelling In isolated cells, under standard conditions, reducing extracellular osmolarity by 40-60 mOsm Kg-I caused swelling of 10-.20%, that was followed by a complete regulatory volume decrease (RVD) within 1-2 min. A typical example of the normal swelling and RVD response is shown in Fig. 4, wherein swelling of about 15% was followed by return to baseline within about 1 min. In contrast, cells treated with Gd3+ (0.5 mM, n = 5) prior to hyposmotic exposure were unable to regulate and lysed within seconds, as shown in Fig.5A and B. These results suggested the role of a Gd 3+-sensitive solute transport system in RVD, that was further explored using electrophysiological techniques (described below).
394
Vol. 34, No.6
Rheology of mesangial cells
70
60
E ~
50
~ 40
~
c: .Q III c:
30
e 20 10 O-f---.---"""T"""---,-------r-...,.----, 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Fractional Area Expansion (!!.AlA)
Fig. 3. Isosmotic swelling during pressurization. Six mesangial cells (MC), as in Fig. lA, were pressurized in steps. and their membrane tensions were calculated using Eq. (2). Fractional area expansions (M/A) were measured from digitized images at each pressure. Points on graph represent averages of six Me, with bars denoting the standard deviations of both area expansion and tension. The tangential average area expansion modulus at the origin was calculated to be 450 dyne/em. A third order polynomial was fit to the data. 120
-o~
t\ . /
Q)
E :J
(5
110
>
..\/
Q)
o
~ ~ Q) 100 a:
._. • -50
o
.)
\ . '.-. 50
100
Time (sec) Fig. 4. Regulatory volume decrease (RVD) in isolated mesangial cell (MG). Volume response of MC in standard Ringer's exposed to 60 mOsm ~l gradient at time = 0 (-e-). Prior to hyposmotic exposure. volume measurements varied between 100 and 105%. Within 10 s of hyposmotic exposure, volume peaked at 115% and began to sharply decline. with some oscillation. and returned to baseline by 60 s,
Rheology of mesangial cells
Vol. 34, No.6
(a)
Iso-osmotic 3· Gd present
(b)
10 seconds after hypoton ic stress 3+ Gd present Fig. 5. Antagonism of re~lalory volume decrease (RVD) in isolated mesangial cells (MC) by Gd ' - (A) Micrograph shows an Me in isotonic media . ma intaining stable vo lume in the presence of Gd3-+ (0.5 mM L- 1 ) . (B) Same Me about 4 s after exposure to 60 mOsm Kg-I gradient. that caused rapid lysis. RVD was apparently slowed or antagonized by Gd 3-+ .
395
396
Rheology of mesangio1 cells
Vol. 34, No.6
Cells cannot regulate volume normally when attached to a whole-cell electrode, since their internal concentration is clamped by the pipet, but fluxes of ions can be measured, as the cell attempts to regulate. Membrane ionic conductance during hyposmotic exposure was thus used as an indirect measure of solute flux during regulation. In isosmotic conditions, the slope conductance of MC was nearly linear, averaging 0.9 ± 0.5 nS (n = 9 cells) over the range between -120 mV and 40 mV, and was stable within ± 10% during the 10-20 min equilibration period before osmotic challenges. Families of I-V tracings for a typical experiment are shown in Fig. 6 for control (A), 5 min after a reduction of the extracellular osmolarity to 260 mOsm Kg-I (B), and 5 min after addition of GdCl3 (0.5 mM) to the hyposmotic media (C) . The corresponding I-V curves for this experiment (panel D) show that control slope conductance was 0.8 nS, linear, and that after hyposmotic exposure it outwardly rectified, increasing to 1.7 and 2.8 nS in the inward and outward directions, respectively. Inhibition of the conductance increase by GdC l3 (panel C) was consistent with the interpretation of Fig. 5B and will be further discussed below. Swelling induced (SI) current was carried by cations since substitution of internal CI- by gluconater did not shift reversal potential but increased conductance (Huang et al., 1997). The reversal potential of -20 mV indicated a constant permeability ratio, PK+/PNa+, of 3, which did not change during hyposmotic stimuli. Summarized results from six cells are shown in Fig. 7. The SI conductance consisted of a significant increase in the average slope conductance (Iso/Hypo) in the outward direction from 1.2 ± 0.5 to 4.2 ± 3 nS (p < 0.01), as shown in the right panel. Slope conductance in the inward direction also increased, but not significantly. Outward and inward conductances for each cell were estimated by fitting lines to points on either side of the reversal potential. To test the pharmacology of SI current, the conductances of six cells that had been exposed to hypotonic media were measured immediately before adding GdCI,3 (0.5 mM), as shown in the bars marked "Hypo" in the left panel. Within 5 min, SI outward conductance was significantly reduced by GdCI 3 (Hypo + Gd 3 +) from 5.7 ± 1.7 nS to 3.7 ± 2.4 nS (p < 0.02). Cells in hyposmotic media not treated with GdCl3 maintained stable high conductance. Since the concentration of GdCl3 used herein was well above threshold blocking dose for stretch-activated channels (SAC) (Yang et al., 1990), subsequent dose response experiments were done (unpublished), and it was found that 20 J.1M GdCl3 was sufficient to block SI current. Agents that failed to block or prevent SI current included externally applied TEA (16 mM), TTX (10 J.1M), BaCl2 (0.1 mM), or amiloride (0.1 mM). Internal TEA was a common pipet component, and neither ATP nor cAMP influenced SI current. These results provide evidence that SAC are involved in SI current. Hyposmotic challenge of MC was done also by raising their intracellular osmolarity with a relatively hypertonic pipet solution (see Materials and Methods) . This was done by pre-incubating MC in hyposmotic media, 260-270 mOsm Kg-I, for at least 10 min, followed by initiation of recording and dialysis with 328 mOsm Kg-I internal solution. Recordings so obtained on four cells showed that initial conductance was near baseline isosmotic levels of 0.9 nS, but rapidly increased several-fold during the first 5 min. These results suggest that MC equilibrate with hyposmotic media within minutes, as expected if SI current effectively nullifies the osmotic gradient. (See Discussion below.)
~
:-
~
!'" Z ?
HypotonIc
c::n
B
40mV
eontrol
200
50 mS
1100 pA
Hypo
I (pA) 0
mV
D
.1f.
.12' •
-II
·11
i i '
.,
I
•
•
I
o
o
•
o
0
0
•
II. 15
0
0
0
0
II
.
•
• 40
•
·20 it
·40
o 0
i
~
~
0
100
~
H~po+Gd
Iso
I
20
~
iii
aQ
~Q'
-~
0
III
t;"
·100
0
o ·200
Fig. 6. Swelling-induced (SI) current in standard Ringer's, I-V families are shown in control (A), after hyposmotic exposure (B), and with hyposmotic + Gd~+ exposure (C) , I-V families are superimposed currents elicited at voltages from -120 mV to 40 mV in steps of 10 mV, with holding voltage of -40 mV, (D) Shows I-V plots corresponding to control (open squares), hyposmotic (open circles) and hyposmotic with Gd~. 0.5 mM (x), For clarity, lines are not drawn through the points. Pipet composition was FJ-l (Table 1).
~
<0
..:r
398
Vol. 34, No.6
Rheology of mesangial cells (. ) : ~, hyperpolarized regions (+) : 9d, depolarized regions
8
( +)
D
lso
~ Hypo ~ Hypo+Gd 3 + ( +)
7
Vi'
(+)
6
.5.5 CD
u
B 4 u :::J
"g 3 o
o 2
o
(.) (+)
CO lso
-
Hypo
Hypo -
Hypo+Gd
3
+
Fig. 7. Swelling-induced conductance summary. Bars labeled (-) represent conductances in the hyperpolarized region, gh (inward conductance), and bars labeled (+) represent average slope conductance in the depolarized region, gd (outward conductance) , with lines representing standard deviations of gh and gd' The left panel shows the average slope conductance in standard Ringer's before (left) and after (right) being exposed to hyposmotic gradient of 40-60 mOsm Kg""l for about 5 min (p < 0.01, n = 6). The right panel shows the average slope conductance of six cells before (left) and after (right) being exposed to Gd 3+ (0.5 mM) for about 5 min in the presence of an hyposmotic gradient of 40-60 mOsm Kg""l (p < 0.05. n = 6). Note : Data in the two panels are two independent series from different cells.
Discussion Mechanical responses to swelling The rather large scatter in the data of Fig. 3 may be partly caused by video measurement errors and by the transient volume oscillations seen in Fig. 2, however, area measurements were made within a few seconds of pressure steps, to minimize the influence of possible regulatory responses. Large scatter in area expansivity measurements of human red blood cells (RBC) was also reported and attributed to irrecoverable volume changes (Waugh and Evans, 1979). Their study measured a linear elastic modulus for RBC of 450 dyne/em, for area expansions ~ 1.2%. This measurement corresponds almost exactly to the tangent at the origin of Fig. 3, and thus it can be concluded that cultured Me membranes have an area expansivity resembling that of human RBe. The apparent stiffness of MC (Fig. 3) depended on the degree and duration of deformation since membrane tension was considerably damped during low to moderate swelling, but not at larger swelling. The bimodal behavior may reflect separate viscous components in the cytoplasm that can be described as a nonlinear Maxwell model (Dong et al., 1991). Other components of RVD may also contribute to the observed mechanical behavior, including cytoskeletal adjustments during swelling (Funai et al., 1993; Bogen,
Vol. 34, No.6
Rheology of mesangial cells
399
1987) or active contraction (Ausiello et al., 1980). Whether these viscoelastic properties of isolated MC are also exhibited in situ is not known, however, isolated MC exhibit many of their in vivo properties (Ardaillou, 1996).
Electrical response to sweUing Our results describe a solute transport system based on effiux of K+ carried through swelling-activated ionic channels. The solute transport is based on SI current, whose conductance and pharmacology are consistent with those recorded in proximal tubule cells of the kidney (Robson and Hunter, 1994) and with those of the SAC in MC, including activation by hyposmolarity (Craelius et al., 1993). The critical dependence of RVD on K+ conductance increase has been previously shown for cultured astrocytes (Pasantes-Morales et al., 1994), and for renal epitheIoid cells, whose K+ conductance transiently increases 64-fold after hyposmotic challenge (Kersting et al., 1991). Other solute effiux pathways may also playa role in RVD responses of MC, including K+ efflux through furosemide-sensitive pathways (Homma et al., 1990). Based on their estimated density on the cell of 1/~m2 (Craelius et al., 1989) , our maximal response of about 20 nS could be produced by full activation of 4% of available SAC. Thus, the present data can be added to the growing body of direct evidence relating single stretch-activated channel currents with whole-cell currents (Zhang et al., 1994; Ross et al., 1994; Gustin, 1991; Sachs, 1991; Chan and Nelson, 1992; Sackin, 1987; Tseng, 1992; Hagiwara et al., 1992; Chan et al., 1994), in contrast to the one reported failure (Morris, 1991) . Limitations of patch-clamp experiments are the restriction to isolated cells, buffering of [Ca 2+]i and pH, and clamping of voltage, that probably shortcircuit some of the normal components of RVD. In cells exposed to hyposmolarity, [Ca 2+]\ rises dramatically (Tsukahara et al., 1994; McCarty and O'Neil, 1992) and may trigger a cascade of responses to swelling. One likely response is activation of CI- channels, opening a major exit pathway for CI(Hoffmann and Simonsen, 1989; Christensen, 1987) . It is also likely that the internally released Ca2+ would initiate contraction using the actin-myosin network (Ausiello et al., 1980), thereby resisting osmotic swelling. In addition to triggering a cascade of biochemical and mechanical events in the cell, the freed Ca 2+ could also balance a small fraction of charge efflux, while simultaneously reducing the net osmolyte flux (McCarty and O'Neil, 1992) and similar effects could result from RVD-induced acidification, (Hoffmann, 1987). Voltage clamping also prevents changes in membrane stiffness that are produced by voltage gradients (Katnik and Waugh, 1990). None of these conditions of patch clamping, however, apply to our visual observations of rapid hypotonic lysis in cells pre-treated with Gd!l+ (Fig. 2). Thus it can be concluded that K+ effiux occurs with or without Ca 2+ flux and is important to RVD. The degree to which SI current may be transiently electrogenic depends on its kinetics in relationship to counterbalancing anions such as CI- or HCO!!-, and is not answered by our results. Implications The outwardly rectifying current induced by swelling, seen in Fig. 6, represents a large efflux of K+ that could accumulate within certain restricted environments and serve as a depolarizing signal to neighboring cells. Further studies are required to quantify this electrochemical signal arising both within and outside the glomerulus, to determine its significance.
400
Rheology of mesangial cells
Vol. 34, No.6
The bimodal nature of MC viscoelasticity during deformations if exhibited in situ, may be a factor in their long-term regulation of the glomerular environment. An extensive cytoskeleton suggests the ability of MC to internally pressurize during high swelling pressure (Latta et aL, 1960) . The relationship of cytoplasmic viscosity to applied pressure implies that the transference of forces to the cytoskeleton and nucleus becomes more effective as pressure and deformation increase. This phenomenon may explain, in part. the ability of large periodic deformations to stimulate extracellular matrix production by MC through nuclear activation, that is believed to be an adaptive response to mechanical overload (Riser et al., 1992; Akai et al., 1994). The dual viscosities appearing in response to small and large deformations could therefore contribute to the acute vs. chronic adaptations of MC to environmental stress.
Acknowledgments We would like to thank Anthony Dawidowitz for his help with the experiments. These studies were funded by the American Heart Association, New Jersey Affiliate (W. Craelius), and a Merit Review Grant from the Department of Veterans Affairs.
References ARDAILLOU, R (1996) . Biology of glomerular cells in culture. Cell Bioi. Toxicol. 12, 257-261. AKAI, Y., HOMMA, T., BURNS, K.D., YASUDA, T., BADR, K.F., and HARRIS, RC. (1994). Mechanical stretch/relaxation of cultured rat mesangial cells induces protooncogenes and cyclooxygenase. Am. J Physiol. 267 (Cell Physiol. 26): C482-G490. AUSIELLO, D.A., KREIBERG, J.I., ROY, C., and KARNOVSKY, M.J. (1980). Contraction of cultured rat glomerular cells of apparent mesangial origin after stimulation with angiotensin II and arginine vasopressin. J CUn. Invest. 65, 754-760. BLANTZ, RC., GABBAI, F.B., TUCKER, BJ., YAMAMOTO, T., and WILSON, C.B. (1993) . Role of mesangial cell in glomerular response to volume and angiotensin II. Am. J Physiol. 264, F158-F165. BOGEN, D.K. (1987) . Strain energy descriptions of biological swelling I: Single fluid compartment models. J Biomech. Eng. 109, 252-256. CHAN, H-C., and NELSON, D. (1992). Chloride-dependent cation conductance activated during cellular shrinkage. Science 257, 669-671. CHAN, H.C., ru, W.O., CHUNG, Y.W., HUANG, S.J., CHAN, P.S.F., and WONG, P.Y.D. (1994). Swelling-induced anion and cation conductances in human epididymal cells. J Physiol. 478, 449-460. CHRISTENSEN, O. (1987). Mediation of cell volume regulation by Ca 2+ influx through stretch-activated channels. Nature 330, 6tH>8. CRAELIUS, W., EL-SHERIF, N., and PAIANT, C.E. (1989). Stretchactivated ion channels in cultured mesangial cells. Biochem. Biophys. Res. Comm 159, 516-521.
Vol. 34, No.6
Rheology of mesangial cells
401
CRAEUUS. W., ROSS, MJ., HARRIS, D.R. , CHEN, V.K., and PAlANT, C.E. (1993). Membrane currents controlled by physical forces in cultured mesangial cells. Kidney Int. 43, 535-543. DONG, C, SKALAK, R, and SUNG, K.-L.P. (1991). rheology of passive neutrophils. Biorheology 28, 557-567.
Cytoplasmic
EVANS, EA., and SKALAK, R (1979). Mechanics and Thermodynamics oj Biomembranes. CRC Press , Boca Raton. FERDOWS. M., GOLCHINI, K., ADLER S.G., and KURTZ, I. (1989). CI-/base exchange in rat mesangial cells: Regulation of intracellular pH. Kidney Int. 35, 783-789. FUNAI,J., FELDMAN, G., RABBANY, S., HAINES, D., SOMANATH, B., and MOHANlY, S. (1993). Intracellular pressure-volume responses to osmotic and hydrostatic stress: Studies in the giant barnacle muscle cells. FASEB 7, A1628. GUSTIN, M.C. (1991). Single-channel mechanosensitive currents. Technical comments. Science 253, 800. HAGIWARA, N., MASUDA, H., SHODA, R, and IRISAWA, H . (1992) . Stretch-activated anion currents of rabbit cardiac myocytes. J Physiol. 456, 285-302. HAMILL, O.P., MARlY, A., NEHER, E., SAKMANN, B., and SIGWORTH, FJ. (1981) . Improved patch-clamp techniques for highresolution current recording from cells and cell-free membrane patches. Pflugers Arch. 391, 85-100. HOFFMANN, E.K. (1987). Volume regulation in cultured cells. In: Current Topics in Membrane Transport. R Gilles, A Kleinzeller, and L. Boles, Eds., Academic Press. New York, pp. 125-180. (1989) . Membrane HOFFMANN, E.K., and SIMONSEN, L.O. mechanisms in volume and pH regulation in vertebrate cells. Physiol. Rev. 69, 315-382. HOMMA, T., HOOVER, L.R, and HARRIS, RC. (1990). Loop diuretic-sensitive potassium flux pathways of rat glomerular mesangial cells. Am. J Physiol. 258, C862-870. HUANG, C.-]., PALANT, C.E., and CRAELIUS, W. (1997). Low CIinduces a cation conductance in rat mesangial cells. Kidney and Blood Pressure (in press). IVERSEN, B.M., KHAM, F.I., MATRE, K., MORKRID, L., HORVEI, G., BAGCHUS, W., GROND, J., and OFSTAD. J. (1992). Effect of mesangiolysis on autoregulation of renal blood flow and glomerular filtration rate in rats. Am. J Physiol. 262, F361-366. JOHNSON, RJ. (1997). Have we ignored the role of oncotic pressure in the pathogenesis of glomerulosclerosis? Am. J Kidney Dis. 29,147-152.
402
Rheology of mesangial cells
Vol. 34, No.6
KATNIK, C., and WAVGH, R (1990). Alterations of the apparent area expansivity modulus of red blood ceIl membrane by electric fields. Biophys.] 57, 877-882. KERSTING, V., WOJNOWSKI, L, STEIGNER, W., and OBERLEITNER, H. (1991). Hypotonic stress-induced release of KHCOs in fused renal epitheloid (MOCK) cells. Kidney Int. 39, 891-900. LATTA, H., MAVNSBACH, A.B., and MADDEN, S.C. centrolobular region of the glomerulus studied microscopy.] Ultrastruct. Mol. Struct. Res. 4, 455-472.
(1960). The by electron
MCCARlY, N., and O 'NEIL, RG. (1992). Calcium signaling in cell volume regulation. Physiol. Rev. 72, 1037-1061. MORRIS, C.E. (1991). Single-channel mechanosensitive currents. Technical Comments. Science 253,801. PASANTES-MORALES, H., MURRAY, RA., LILJA, L., and MORAN, J. (1994). Regulatory volume decrease in cultured astrocytes. I. Potassium- and chloride-activated permeability. Am. J Physiol. 266, C165-C171. PVSCH, M., and NEHER, E. (1988). Kinetics of loading small cells with various compounds by use of patch pipets. Pflugers Arch. 408, (Suppl. I), 338. RISER B.L., CORTES, P., ZHAO, X., BERNSTEIN, J., DUMLER, F., and NARINS, R.G. (1992). Intraglomerular pressure and mesangial stretching stimulate extraceIlular matrix formation in the rat. J Clin. Invest. 909, 1932-1943. ROBSON, L., and HUNTER, M. (1994). Role of cell volume and protein kinase C in regulation of a CI- conductance in single proximal tubule cells of Rana temporaria.] Physiol. 480, 1-7. ROSS, P.E., GARBER, S.S., and CAHALAN, M.D. (1994). Membrane chloride conductance and capacitance in Jurkat T lymphocytes during osmotic swelling. Biophys.] 66, 169-178. SACHS, F. (1991). Single-channel mechanosensitive currents. Technical comments. Science 253, 800. SACKIN, H. (1987). Stretch-activated potassium channels in renal proximal tubule. Am.]. Physiol. 253, FI253-FI262.
SAKAI, T., and KRIZ, W. (1987). The structural relationship between mesangial cells and basement membrane of the renal glomerulus. Anat. Embryol. 176, 373-386. SCHLONDORFF, D. (1996). Roles of mesangium in glomerular function. Kidney Int. 49, 1583-1585. SCHMID-SCHONBEIN, G.W., KOSAWADA, T., SKALAK, R., and CHIEN, S. (1995). Membrane model of endothelial cells and
Vol. 34, No.6
Rheology of mesangial cells
403
leukocytes. A proposal for the origin of a cortical stress. J Biomech. Eng. 117,171-178. STOCKAND, J.D., and SANSOM, S.C. (1997). Regulation of filtration rate by glomerular mesangial cells in health and diabetic renal disease. Am. J Kidney Dis. 29, 971-981. (1992). Cell swelling increases membrane TSENG, G-N.Y. conductance of canine cardiac cells: Evidence for a volume-sensitive CI channel. Am. J Physiol. 262, C105~1068. TSUKAHARA, H., KRIVENKO, Y., MOORE, L.C., and GOLIGORSKI, M.S. (1994). Decrease in ambient [CI-] stimulates nitric oxide release from cultured rat mesangial cells. Am. J Physiol. 267, F190-F195. WANG, X., AUKLAND, K., BOSTAD, L., and IVERSEN, B.M. (1997). Autoregulation of total and zonal glomerular filtration rate in spontaneously hypertensive rats with mesangiolysis. Kidney and Blood Pressure Res. 20, 11-17. WAUGH, R, and EVANS, E.A. (1979). Thermoelasticity of red blood cell membrane. Biophys. J 26, 115-132. YANG, X-C., and SACHS, F. (1990). Characterization of stretchactivated ion channels in xenopus ooeytes. J Physiol. 431, 103-122. ZHANG, J., RASMUSSON, RL., HALL, S.K., and LIEBERMAN, M. (1994). A chloride current associated with swelling of cultured chick heart cells. J Physiol. 472, 801-82.
Received 28 July 1997; accepted in revised form 19 December 1997.