Kidney International, Vol. 56 (1999), pp. 1700–1709

Polycations induce calcium signaling in glomerular podocytes

FRANK RUDIGER, RAINER GREGER, ROLAND NITSCHKE, ANNA HENGER, PETER MUNDEL, andHERMANN PAVENSTADT

Division of Nephrology, Department of Medicine, and Department of Physiology, University of Freiburg, Freiburg, Germany,and Division of Nephrology, Albert Einstein College of Medicine, Bronx, NY, USA

Polycations induce calcium signaling in glomerular podocytes. rejection of macromolecules increases as the molecularBackground. The neutralization of the polyanionic surface charge becomes more negative, supporting the idea that

of the podocyte by perfusion of kidneys with polycations, such negative charges of the filtration barrier, which are foundas protamine sulfate, leads to a retraction of podocyte footin all layers, hinder the passage of negatively chargedprocesses and proteinuria. This study investigates the effects

of protamine sulfate or anionic, neutral, or cationic dextrans molecules [1–3]. The basement membrane and the podo-on the cytosolic calcium activity ([Ca21]i) in podocytes. cytes, but not the open fenestrated endothelial cells,

Methods. [Ca21]i was measured in single cultured differenti- serve as an effective barrier to plasma proteins [1]. Itated mouse podocytes with the fluorescence dye fura-2/AM.has been suggested that the basement membrane sufficesResults. Protamine sulfate caused a concentration-depen-to explain the charge and size-selective properties ofdent and partially reversible increase of [Ca21]i (EC50 approxi-

mately 1.5 mmol/liter). Pretreatment of the cells with heparin the glomerulus to macromolecules [4, 5]. However, big(100 U/liter) inhibited the protamine sulfate-mediated increase molecules such as ferritin are able to permeate the glo-of [Ca21]i. Like protamine sulfate, diethylaminoethyl dextran

merular basement membrane at sites of podocytes de-(DEAE-dextran) concentration dependently increased [Ca21]itachment, and a selective injury of podocytes leads toin podocytes (EC50 approximately 20 nmol/liter), whereas dex-

tran sulfate or uncharged dextran (both 10 mmol/liter) did albuminuria [6, 7]. Therefore, it is believed that podo-not influence [Ca21]i. A reduction of the extracellular Ca21 cytes play a crucial role as a filter restricting the passageconcentration (from 1 mmol/liter to 1 mmol/liter) partially in- of macromolecules.hibited the protamine sulfate and the DEAE-dextran–induced

The basement membrane of podocytes is covered with[Ca21]i response. Flufenamate (100 mmol/liter) or Gd31 (10sulfated glycosaminoglycans, mainly heparan sulfate andmmol/liter), which are known to inhibit nonselective ion chan-

nels, did not influence the [Ca21]i increase induced by prot- the slit membrane, and the foot processes of podocytesamine sulfate. In the presence of thapsigargin (50 nmol/liter), are decorated with a negatively charged glycocalyx [re-an inhibitor of the endoplasmic reticulum Ca21-ATPase, both viewed in 8]. Neutralization of the polyanionic surfaceprotamine sulfate and DEAE-dextran increased [Ca21]i.

by perfusion of kidneys with the polycation protamineConclusions. The data indicate that polycations increase po-docyte [Ca21]i. The increase of [Ca21]i may be an early event sulfate (PS) retracts podocyte foot processes and inducesin the pathogenesis of protamine sulfate-mediated retraction a displacement of the slit membranes [9]. The PS-inducedof podocyte foot processes. injury of podocytes, which is morphologically very simi-

lar to that observed in proteinuric states, is preventedby neutral or anionic macromolecules [9]. Therefore, it

The glomerular filtration barrier prevents all but the is believed that the anionic glycocalyx on the podocytesmallest plasma proteins from entering the urinary space. surface is essential for the maintenance of foot processIt interacts with proteins by virtue of their size and charge structure and the characteristics of the filtration barriercharacteristics. Neutral dextrans and large nonmetabo- [8, 9]. It has been speculated that charge modulationslized organic molecules are excluded from the filtration activate intracellular signaling pathways, which regulateon the basis of size and shape but not of charge. The extracellular matrix proteins of podocytes, but the pre-

cise interrelations between alterations of podocytecharge and cellular functions are not well understoodKey words: podocyte charge, intracellular calcium, protamine sulfate,

dextran, cell signaling. [10]. Cytosolic Ca21 is an important second messengerthat regulates cellular functions in literally all cells. In

Received for publication August 18, 1998this study, we investigated whether charge alterationsand in revised form April 13, 1999

Accepted for publication June 7, 1999 induced by PS or dextrans influence [Ca21]i in podocytesand examined the possible mechanisms involved. 1999 by the International Society of Nephrology

1700

Rudiger et al: Polycations in glomerular podocytes 1701

METHODS measured noise and autofluorescence of the cells did notexceed 1% of the fura-2 fluorescence and were thereforeCell culturenot subtracted. The photomultiplier was used in the pho-

This study used conditionally immortalized mouse po- ton counting mode, and the data were digitized with 12-docyte clone cells, harboring a temperature-sensitive mu- bit resolution and were recorded continuously on thetant of immortalized SV-40 large T-antigen coupled with hard disc of an AT computer. The photon counts fora promoter whose activity is enhanced by g-interferon the three excitation wavelengths were calculated for each[11]. Podocytes were maintained in RPMI 1640 medium turn of the wheel. In 15 experiments, the calibration of(Life Technologies, Eggenstein, Germany) supplemented the fura-2 fluorescence signal could be performed at thewith 100 g/liter fetal calf serum (FCS; Boehringer Mann- end of the protocol using the Ca21 ionophore ionomycinheim, Mannheim, Germany), 100 kU/liter penicillin, and (5 mmol/liter) and low and high Ca21 buffers. To vary100 mg/liter streptomycin (Life Technologies) in a humid the free Ca21 activity, the solutions were prepared ac-atmosphere with 5% CO2. To propagate podocytes, the cording to established techniques with egtazic acidculture medium was supplemented with 10 kU/liter (EGTA) and nitnilotriacetic acid (NTA) as Ca21 buffers.mouse recombinant g-interferon (Sigma, Deisenhofen, The Ca21 activity was calculated from a standard equa-Germany) to enhance the expression of T-antigen, and tion and was determined with a Ca21-selective electrodecells were cultivated at 338C (permissive conditions). To (Radiometer, Copenhagen, Denmark) [13]. [Ca21]i wasinduce differentiation, podocytes were cultured on type-I calculated from the fluorescence ratio according to thecollagen (Biochrom, Berlin, Germany) at 378C without equation described by Grynkiewicz, Poenie, and Tsieng-interferon (nonpermissive conditions) and in the pres- [14]. A KD for the fura-2–Ca21 complex of 224 nmol/ence of 10 g/liter FCS for two to three weeks. Differenti- liter (378C) was assumed. The given concentrations for

the [Ca21]i peak refer to the highest value of the fluores-ated podocytes stained positively for the Wilm’s tumorcence ratio. The [Ca21]i plateaus refer to the fluorescenceantigene and synaptopodin, a characteristic marker forratio 60 seconds after the [Ca21]i peak.podocyte foot processes in vivo [11]. Podocytes between

passages 6 and 20 were used in all experiments.ChemicalsMurine Balb3T3 fibroblasts (ATCC, Rockville, MD,

The following agents were used: Fura-2/AM, PS,USA) were grown in RPMI 1640 medium supplementedTriton-X 100, Gd31, flufenamate, staurosporine, heparinwith 100 g/liter FCS, 100 kU/liter penicillin, and 100 mg/(Sigma), diethylaminoethyl dextran (DEAE-dextran;liter streptomycin in a humid atmosphere with 5% CO2.positively charged dextran), dextran-sulfate (negativelycharged dextran), dextran-T 500 (neutral dextran; Phar-Measurements of the intracellularmacia Biotech, Uppsala, Sweden), herbimycin A, iono-calcium activity [Ca21]i

mycin, calyculin A, KN 62, lysozyme (Calbiochem, BadMeasurements of [Ca21]i were performed in single po-Soden, Germany), and thapsigargin (Molecular Probes,docytes or fibroblasts with an inverted fluorescence mi-Eugene, OR, USA).croscope recently described (bath temperature, 378C,

bath solution changes within one second) [12]. In short, Statisticspodocytes were incubated with the Ca21-sensitive dye

The data are given as mean values 6 sem (N), wherefura-2/AM (5 mmol/liter; Sigma) for 30 minutes at 378C.N refers to the number of experiments. The average

They were mounted in a bath chamber on the stage of fluorescence ratio from cells that were rinsed with aan inverted microscope and perfused with a Ringer-like Ringer-like solution before and after application of thesolution containing (in mmol/liter): NaCl 145, K2HPO4 agonist was taken as a control value. In the case of an1.6, KH2PO4 0.4, CaCl2 1.3, MgCl2 1.03, d-glucose 5, pH irreversible effect of the agonist, only the basal fluores-7.4. The light from a 75 W xenon lamp was directed cence ratio before the addition of the agonist was takenthrough an infrared light filter (Tempax; Schott, Mainz, as control. Paired or unpaired t-test was used to compareGermany) to avoid thermal damage of the three excita- mean values within one experimental series. A P valuetion filters mounted in a motor-driven filter wheel (10 of , 0.05 was accepted to indicate statistical significance.cycles/second). The excitation filters were bandpass fil-ters with transmission maxima at 340, 360, and 380 nm

RESULTS(Delta Light & Optics, Lyngby, Denmark). A dichroicProtamine sulfate, DEAE-dextran, and lysozymemirror (FT 425; Zeiss, Oberkochen, Germany) and aincrease the cytosolic calcium activity ([Ca21]i)bandpass filter (500 to 530 nm; Delta Light & Optics)in podocyteswere used in the emission light pass. The field of view

of the photomultiplier could be chosen by means of an Figure 1A shows an original fluorescence recording ofadjustable rectangular diaphragm. Experiments were the effect of 1 mmol/liter PS on [Ca21]i in a single podo-

cyte. The addition of PS caused a slow and sustainedperformed with a window including one single cell. The

Rudiger et al: Polycations in glomerular podocytes1702

Fig. 1. Protamine sulfate (PS) in-creases the cytosolic calcium activity([Ca21]i) in podocytes (d) and fibro-blasts (m). A typical fluorescence re-cording of the effect of PS (1 mmol/liter) on [Ca21]i in a podocyte (A)and a fibroblast (B). The fluorescenceratio with 340/380 nm excitation isshown as a function of time. (B) Con-centration response curve of the ef-fect of PS on [Ca21]i in podocytes andfibroblasts. In this and subsequentfigures, the values were taken at theagonist-induced [Ca21]i peak. The as-terisk indicates significant differencesbetween control and peak values.

[Ca21]i increase to a peak, followed by a decline to a nmol/liter DEAE-dextran. Figure 2B shows the concen-tration-response curve of the effect of DEAE-dextranplateau (N 5 122). The [Ca21]i peak was reached afteron [Ca21]i in podocytes (N 5 56). One hundred nmol/267 6 17 and 229 6 46 seconds after the addition of 1liter DEAE-dextran increased [Ca21]i from 59 6 8 toand 10 mmol/liter PS, respectively (N 5 15 to 49). After373 6 58 nmol/liter (N 5 13). The effect of DEAE-the removal of PS and rinsing podocytes with Ringer-dextran was concentration dependent, with an EC50 oflike solution, [Ca21]i returned to the baseline value inabout 20 nmol/liter.60% of the experiments. The [Ca21]i response to PS was

Figure 3 shows the effect of 10 mmol/liter of lysozymeconcentration dependent with an EC50 of approximatelyon [Ca21]i of a podocyte. Lysozyme increased [Ca21]i in1.5 mmol/liter for PS (Fig. 1C). One and 10 mmol/literpodocytes (N 5 5), but in comparison to PS and DEAE-PS increased [Ca21]i from 31 6 4 to a peak of 249 6 35dextran, the effect of lysozyme was much weaker. Innmol/liter or from 23 6 6 to a peak of 325 6 104 nmol/contrast to PS and DEAE-dextran, in the presence ofliter, respectively (N 5 15 to 49). After three minuteslysozyme, [Ca21]i declined to baseline [Ca21]i values.in the presence of 1 and 10 mmol/liter PS, [Ca21]i declined

to a plateau of 68 6 5 and 129 6 31 nmol/liter, respec- Influence of heparin on protamine sulfate-inducedtively (N 5 15 to 49). [Ca21]i increase in podocytes

Like podocytes, mouse fibroblasts responded to PS Pretreatment of the cells with negatively charged hep-with a [Ca21]i increase (N 5 26). The effect of PS on arin (100 U/liter) for approximately five minutes attenu-[Ca21]i of fibroblasts was concentration dependent with ated the [Ca21]i increase induced by 0.5 mmol/liter PSan EC50 of approximately 1.5 mmol/liter for PS. Figure by 93 6 4% (N 5 5). Conversely, the addition of heparin1B shows an original fluorescence recording of the effect during the application of 0.5 mmol/liter PS attenuatedof 1 mmol/liter PS on [Ca21]i in a fibroblast. Figure 1C the [Ca21]i increase significantly by 38 6 16% (N 5 5).shows the concentration-response curve. Figure 4 shows original recordings and the summary of

Figure 2A shows an original recording of the effect of the effect of heparin on PS-mediated [Ca21]i increase.100 nmol/liter DEAE-dextran on [Ca21]i of a podocyte.

Effect of a low extracellular Ca21 on the [Ca21]iDEAE-dextran induced an increase of [Ca21]i in podo-response elicited by protamine sulfate andcytes, whereas negatively charged dextran sulfate or un-DEAE-dextrancharged dextran did not influence [Ca21]i (both 10 mmol/

liter, N 5 5, data not shown). The [Ca21]i peak was Pretreatment of podocytes with a solution with a re-duced extracellular Ca21 concentration (from 1mmol/reached after 143 6 11 seconds after the addition of 100

Rudiger et al: Polycations in glomerular podocytes 1703

Fig. 3. Lysozyme increases [Ca21]i in podocytes. (A) Fluorescencerecording of the effect of 10 mmol/liter lysozyme on [Ca21]i (displayedFig. 2. Diethylaminoethyl dextran (DEAE-dextran) increases the cy-as fluorescence ratio) in podocytes. (B) Summary of five experimentstosolic calcium activity ([Ca21]i) in podocytes. (A) Typical fluorescence(N 5 5). The asterisk indicates significant differences between controlrecording of the effect of 100 nmol/liter DEAE-dextran on [Ca21]i and peak values.(displayed as fluorescence ratio) in podocytes. (B) Concentration re-

sponse curve of DEAE-dextran. The asterisk indicates significant differ-ences between control and peak values.

were performed. Preincubation of podocytes with 100mmol/liter flufenamate or 10 mmol/liter Gd31 for three

liter to 1 mmol/liter) partially prevented the effect of 10 to five minutes did not significantly alter the effect ofmmol/liter PS and 100 nmol/liter DEAE by 74 6 6 and 0.5 mmol/liter PS on [Ca21]i (N 5 8 for both). Figure 623 6 4%, respectively (N 5 5 to 10; Fig. 5). Moreover, summarizes the data.the effect of lower concentrations of PS (1 mmol/liter)on [Ca21]i was completely prevented under these condi- Effect of thapsigargin on protamine sulfate- ortions (N 5 7, data not shown). Even after the removal of DEAE-dextran–induced [Ca21]i increasePS and the subsequent readdition of extracellular Ca21,

Figure 7A shows the effect of thapsigargin in the pres-[Ca21]i increased to a value comparable to the rise ofence and absence of PS or extracellular adenosine 5’-[Ca21]i observed in the presence of PS.triphosphate (ATP). Thapsigargin (50 nmol/liter), an in-hibitor of the endoplasmic reticulum Ca21-ATPase [17],Flufenamate and Gd31 do not influence protamine

sulfate-mediated [Ca21]i increase increased [Ca21]i from 70 6 6 to a peak value of 195 6

15 nmol/liter (N 5 44). In the presence of thapsigargin,The data mentioned earlier in this article indicate thatthe addition of PS (10 mmol/liter) still increased [Ca21]ithe PS-induced increase of [Ca21]i was at least in partfrom 28 6 11 to 203 6 51 nmol/liter (N 5 10, not signifi-caused by an influx of Ca21 from the extracellular space.cantly different from the effect of PS under control con-To characterize the PS-mediated Ca21 influx mecha-ditions). In the presence of thapsigargin, extracellularnisms, experiments with flufenamate and Gd31, which

are known to inhibit nonselective cation channels [15, 16], ATP (100 mmol/liter), a known Ca21-mobilizing agonist

Rudiger et al: Polycations in glomerular podocytes1704

Fig. 4. The protamine sulfate-mediated [Ca21]i

increase is inhibited by heparin. (A, left panel)Typical fura-2 fluorescence recording of the effectof a pretreatment of podocytes with heparin (100U/liter) on the [Ca21]i increase (displayed as fluo-rescence ratio) induced by 0.5 mmol/liter PS.(Right panel) Summary of the experiments (N 55). (B, left panel) Typical fura-2 fluorescence re-cording of the effect of heparin (100 U/liter) dur-ing the PS-induced [Ca21]i increase. (B, rightpanel) Summary of the experiments (N 5 5). Theasterisk indicates significant differences betweenthe effect of PS in the presence or absence ofheparin.

Fig. 5. Effect of a low extracellular Ca21 con-centration (Ca21↓, 1 mmol/liter) on the [Ca21]i

response elicited by (A) protamine sulfate (10mmol/liter) or (B) by diethylaminoethyl dextran(DEAE, 100 nmol/liter). (Left panels) Originalrecordings. (Right panels) Summaries of the ex-periments (N 5 5 to 10). Note that in the pres-ence of a low extracellular Ca21 concentration,DEAE-dextran and PS still increased [Ca21]i.The asterisk indicates significant differences be-tween low extracellular Ca21 and PS or DEAE-dextran in the presence of low extracellularCa21.

caused a small but significant decrease of the [Ca21]i but significant increase of [Ca21]i from 12 6 6 to 53 69 nmol/liter (N 5 9).plateau from 97 6 12 to 73 6 10 nmol/liter (N 5 22).

Figure 7B shows the effect of ATP or PS in the pres- Figure 8 shows similar experiments for DEAE-dex-tran. In the presence of thapsigargin, DEAE-dextranence of thapsigargin and the effect of reduced extracellu-

lar Ca21. A reduction of the extracellular Ca21 (from 1 (100 nmol/liter) still increased [Ca21]i from 29 6 11 to229 6 83 nmol/liter (N 5 8, not significantly differentmmol/liter to 1 mmol/liter) completely and significantly

attenuated the thapsigargin-induced [Ca21]i plateau from the effect of DEAE-dextran under control condi-tions). Figure 8B shows the effect of DEAE-dextran in(N 5 8). Under these experimental conditions, that is,

in the presence of thapsigargin and a low extracellular the presence of thapsigargin and a reduced extracellularCa21. A reduction of the extracellular Ca21 (from 1Ca21 concentration, extracellular ATP (100 mmol/liter)

did not induce a significant [Ca21]i change in podocytes mmol/liter to 1 mmol/liter) completely and significantlyinhibited the thapsigargin-induced [Ca21]i increase. Un-(N 5 8), whereas 10 mmol/liter PS again induced a small

Rudiger et al: Polycations in glomerular podocytes 1705

been reported as a model of IgA nephropathy [18]. Inthis model of experimental nephropathy, all dextransinduce glomerular injury, but the pattern and the severityof glomerular lesions are charge dependent; that is, cat-ionic dextrans cause more severe glomerular alterations[18]. Pasi et al reported that after injection of DEAE-dextran, rats developed a proteinuria in a nephroticrange, depositions of dextrans in the glomerular basalmembrane, and a damage of podocytes, but no deposi-tion of immune complexes [19].

Very recently, we have observed that incubation ofpodocytes with PS for more than two hours altered theirmorphology; that is, podocyte cell bodies rounded up,

Fig. 6. Flufenamate and Gd31 do not significantly influence the PSand a retraction of their foot processes occurred. The(0.5 mmol/liter)-mediated [Ca21]i increase. Summary of the effect of a

preincubation of 100 mmol/liter flufenamate or 10 mmol/liter Gd31 for effects of PS on podocyte morphology were preventedthree to five minutes on PS-mediated [Ca21]i increase shown here as in the presence of heparin (Mundel et al, unpublishedthe change in fluorescence ratio.

observations). This suggests that PS alters the morphol-ogy of cultured podocytes in a similar manner as in podo-cytes in vivo [9], indicating that the in vitro model usedin this study might be a useful tool in investigating theder these experimental conditions, that is, in the presenceinteractions between podocytes charge, their morpho-of thapsigargin and at a low extracellular Ca21 concentra-logical alterations, and cellular signaling events.tion, DEAE-dextran still significantly increased [Ca21]i

In this study, we investigated whether polycations mod-from 19 6 4 to 129 6 30 nmol/liter (N 5 7).ulate [Ca21]i in podocytes. The study was performed withdifferentiated podocytes, which exhibited an arborizedHerbimycin A, staurosporine, KN-62, and calyculin Amorphology and expressed podocyte-specific markers [11].do not influence a PS-mediated [Ca21]i increase

Positively charged DEAE-dextrans or PS increasedin podocytes[Ca21]i in podocytes. The effect of PS and DEAE-dex-It has been recently suggested that tyrosine kinase istran on [Ca21]i was partly but not completely reversible.activated in PS-induced injury of podocytes [10]. To clar-This suggests that PS and DEAE-dextran may have in-ify whether tyrosine kinase or other signaling pathwaysduced an impairment of the [Ca21]i regulation or that

are involved in PS-induced [Ca21]i increase, the [Ca21]i some cells were irreversibly damaged by the cations.response to PS (0.5 mmol/liter) was investigated after However, short-term incubation of podocytes with PSpreincubation with herbimycin A, staurosporine, KN-62 or DEAE-dextran for up to one hour did not increaseand calyculin A, inhibitors of tyrosine kinase, protein lactate dehydrogenase release, and PS induced a partiallykinase C, Ca21/calmodulin-dependent protein kinase II, reversible depolarization of podocytes; however, theand protein phosphatase, respectively, for about five membrane voltage did not collapse in the presence of PSminutes (N 5 5 to 8). None of the substances significantly (Pavenstadt et al, unpublished observations). Therefore,altered PS-induced [Ca21]i increase in podocytes. Figure acute administration of PS or DEAE dextran probably9 summarizes the data. does not induce irreversible toxicity in podocytes. A

partial irreversible PS effect has been also reported inglomerular epithelial cells with a cobblestone phenotype.DISCUSSIONIn these cells, the removal of PS failed to protect cells

A retraction of foot processes is a major reaction pat- from albumin leakage [20].tern of podocytes that occurs in many glomerular dis- Like in podocytes, PS increased [Ca21]i in fibroblasts,eases. As a consequence, podocytes become simplified, indicating that the [Ca21]i response to PS is a more gener-and the capillary wall is covered by flat epithelial cell- alized phenomena. Podocytes are a very specialized celllike podocytes. It has been suggested that a reduction type, and the [Ca21]i transients shown here, even afterof the surface charge of podocytes leads to a change a single injection of PS, caused a retraction of the footof podocyte shape, but the cellular mechanisms of the processes, which was not seen in fibroblasts. Further-interaction of podocyte charge and shape are not clearly more, the PS-induced retraction of foot processes has aunderstood [8]. severe functional consequence, namely proteinuria.

In vivo studies show that perfusion of isolated kidneys Although we cannot completely exclude some nonspe-with PS results in a retraction of podocytes foot pro- cific cellular effects of the used agonists, there is goodcesses, probably by neutralizing the podocyte anionic evidence that the effects of PS and DEAE-dextran were

due to their positive charge: (a) Negatively charged orsurface charge [9]. An injection of dextrans into rats has

Rudiger et al: Polycations in glomerular podocytes1706

Fig. 7. Influence of thapsigargin (Thaps) on the protamine sulfate (PS)-induced [Ca21]i increase. (A, Left panel) Typical recording of the [Ca21]i

(shown as fluorescence ratio) response to PS (10 mmol/liter) and ATP (100 mmol/liter) in the presence of thapsigargin (50 nmol/liter) and anextracellular Ca21 concentration of 1.3 mmol/liter. (A, right panel) Summary of the experiments (N 5 10). The asterisk indicates significantdifferences. In the presence of thapsigargin, ATP induced a significant decrease of [Ca21]i, whereas PS significantly increased [Ca21]i. Thapsigargindid not significantly influence the magnitude of the [Ca21]i increase induced by PS (Fig. 2). (B, left panel) Typical recording of the effect of PS(10 mmol/liter) or ATP (100 mmol/liter) on [Ca21]i in the presence of thapsigargin and a low extracellular Ca21 concentration (Ca21↓, 1 mmol/liter).(B, right panel) Summary of the experiments (N 5 6). Note that in the presence of thapsigargin and with low extracellular Ca21 concentration,ATP did not induce a significant [Ca21]i change in podocytes, whereas PS induced a small but significant increase of [Ca21]i.

neutral dextrans did not influence [Ca21]i in podocytes. Lysozyme, a cationic protein with bactericidal activitythat can be released by neutrophils, has been suggested(b) Neutralization of positive charges by preincubation

of podocytes with heparin prevented the effect of PS, to bind to the glomerular basement membrane duringglomerular injury and to induce proteinuria by chargeand the addition of heparin during the PS-induced [Ca21]i

increase reduced the [Ca21]i response to PS. Similar pro- neutralization [21]. Like PS and DEAE-dextran, lyso-zyme increased [Ca21]i in podocytes. However, in com-tective effects of heparin have been also observed in situ,

where heparin restored PS-mediated damage of podo- parison to PS and DEAE-dextran, lysozyme was muchless potent in increasing [Ca21]i, and in contrast to PScytes foot processes [9]. (c) In comparison to PS, DEAE-

dextran was approximately 75 times more potent in in- and DEAE-dextran, [Ca21]i returned to baseline valueseven in the continued presence of lysozyme. The differ-creasing [Ca21]i, and the Hill coefficient for PS and

DEAE-dextran was approximately 0.9 and 2.0, respec- ent effects of lysozyme on [Ca21]i may be explained byits fewer positive charges (17 positive charges per mole).tively. These differences may be, at least in part, ex-

plained by the different amounts of positive charges of In addition, by looking at the structure of lysozyme, itis obvious that in comparison to PS and DEAE-dextran,the agents: PS possesses 21 and DEAE-dextran 1140

positive charges per mole. Therefore, there is a correla- lysozyme possesses also many negative charged aminoacids and that within the protein, the positive chargedtion between the amount of positive charges and the

efficiency of these agents to increase [Ca21]i. amino acids are widely distributed and not densely con-

Rudiger et al: Polycations in glomerular podocytes 1707

Fig. 8. Effect of thapsigargin (Thaps) on diethylaminoethyl dextran (DEAE-dextran)-induced [Ca21]i increase. (A, left panel) Typical recordingof the [Ca21]i response (displayed as fluorescence ratio) to DEAE-dextran (100 nmol/liter) in the presence of thapsigargin (50 nmol/liter) and anextracellular Ca21 concentration of 1.3 mmol/liter. (A, right panel) summary of the experiments (N 5 8). (B, left panel) Typical recording of theeffect of DEAE-dextran (100 nmol/liter) in the presence of thapsigargin and low extracellular Ca21 concentration (Ca21↓, 1 mmol/liter). (B, rightpanel) Summary of the experiments (N 5 7). The asterisk indicates significant differences.

centrated [22]. Subsequent studies will have to clarify Similar to the effect of PS, the DEAE-dextran–whether the distribution and the density of positive mediated [Ca21]i increase was only partially inhibited incharges within the protein may influence the pattern and a low extracellular Ca21 concentration, indicating thatthe magnitude of the [Ca21]i response. DEAE-dextran released Ca21 from intracellular stores

It has been reported that perfusion of kidneys with and induced Ca21 influx.PS in a solution with a reduced extracellular Ca21 con- To characterize the ion channel, which may be respon-centration ameliorated the PS-mediated podocyte foot sible for the PS-mediated Ca21 efflux, experiments withprocess injury by approximately 50%, indicating that an ion channel blockers have been performed. In the pres-influx from Ca21 is crucial for the PS-mediated damage ence of flufenamate or Gd31, which are known to inhibitof podocyte foot processes [9]. The [Ca21]i increase in- nonselective cation channels [15, 16], a PS-induced in-duced by 10 mmol/liter PS in this study was partially

crease of Ca21 was not inhibited, indicating that PS didprevented in the presence of a low extracellular Ca21

not open flufenamate- or Gd31-sensitive nonselectiveconcentration, and the effect of 1 mmol/liter PS was com-cation channels. Recently, we have shown that mousepletely abolished under these experimental conditions.podocytes do not possess functionally relevant L-typeThis may indicate that low concentrations of PS onlyCa21 channels, and therefore, studies with L-type Ca21

induce a Ca21 influx from the extracellular space or thatchannel blockers were not performed [23].an undectable Ca21 release from intracellular stores oc-

To characterize further the mechanisms whereby po-curs and that higher concentrations of PS mediate alycations increase [Ca21]i in podocytes, [Ca21]i measure-detectable Ca21 release from intracellular stores in addi-

tion to the influx of Ca21. ments were performed in the presence of thapsigargin,

Rudiger et al: Polycations in glomerular podocytes1708

inhibit the effect of PS, suggesting that it may not bemediated by protein kinase C, Ca21/calmodulin-depen-dent protein kinase II, or protein phosphatases.

In conclusion, our data indicate that positively chargedagents increased [Ca21]i in podocytes by activating theinflux of Ca21 and mobilization of Ca21 from thapsigar-gin-insensitive intracellular Ca21 stores. The effects ofthe cations are probably mediated by their positivecharge. The [Ca21]i increase may be an early cellularresponse that is activated by neutralization of the resi-dent anionic charges. There is a great body of evidenceshowing that proteinuria is associated with charge alter-ations of the glomerular filtration barrier and podocytesFig. 9. Herbimycin A (Herbi, 1 mmol/liter), staurosporine (Stauro, 100specifically in proteinuric diseases, such as membranousnmol/liter), KN-62 (10 mmol/liter), and calyculin A (10 nmol/liter) do

not significantly influence the protamine sulfate (PS; 0.5 mmol/liter)- and diabetic nephropathy [28, 29]. Our study on themediated [Ca21]i increase in podocytes. Summary of the experiments interaction between charge and podocyte function ex-(N 5 5 to 8). Cells were preincubated with the agents for approximately

pands the knowledge base about the mechanisms of pro-five minutes, after which PS was added.teinuria.

ACKNOWLEDGMENTSan inhibitor of the endoplasmic reticulum Ca21-ATPase

This work was supported by the Deutsche Forschungsgemeinschaft[17]. After emptying InsP3-sensitive Ca21 stores by thap- (DFG), Pa 483/1-5 and Fi 691/1-1. We thank Mrs. C. Hupfer and T.

Kilic for their excellent technical assistance.sigargin, PS and DEAE-dextran still increased [Ca21]i inpodocytes. In contrast, the vasoactive agent ATP, which

Reprint requests to Dr. H. Pavenstadt, Medizinische Universitats-is known to increase [Ca21]i in human and mouse podo- klinik, Abt. IV, Labor C4, Hugstetter Str. 55, D-79106 Freiburg, Ger-cytes by a release of Ca21 from intracellular stores and many.

by an influx of Ca21 from the extracellular space (unpub-lished observations) [24], did not induce an increase, but REFERENCESa small decrease of [Ca21]i. This suggests that PS and 1. Comper WD, Glasgow EF: Charge selectivity in kidney ultrafil-DEAE-dextran increase [Ca21]i by a release of Ca21 from tration. Kidney Int 47:1242–1251, 1996

2. Bohrer MP, Baylis C, Humes HD, Glassock RJ, Robertson CR,thapsigargin insensitive stores. This was further sup-Brenner BM: Permselectivity of the glomerular capillary wall:ported by experiments in the presence of thapsigargin Facilitated filtration of circulating polycations. J Clin Invest 61:72–

and with a reduced extracellular Ca21 concentration, 78, 19783. Caulfield JP, Farquhar MG: Distribution of anionic sites inwhere PS and DEAE-dextran still increased [Ca21]i. The

glomerular basement membranes: Their possible role in filtrationexistence of thapsigargin-insensitive stores has been re- and attachment. Proc Natl Acad Sci USA 73:1646–1650, 1976cently proposed for other cell types [25]. In contrast, 4. Rennke HG, Cotran RS, Venkatachalam MA: Role of molecular

charge in glomerular permeability: Trace studies with cationizedthe effect of extracellular ATP was also inhibited underferritins. J Cell Biol 67:638–646, 1975these experimental conditions, indicating that extracellu-

5. Kanwar YS, Farquhar MG: Presence of heparan sulfate in thelar ATP was not able to mobilize Ca21 from thapsigargin- glomerular basement membrane. Proc Natl Acad Sci USA 76:1303–

1307, 1979insensitive intracellular Ca21 stores. The nature of the6. Kanwar YS, Rosenzweig LJ: Altered glomerular permeability asCa21 stores and the intracellular mechanisms involved in

a result of focal detachment of the visceral epithelium. Kidney Intthe Ca21 effects of PS and DEAE-dextran remain open. 21:565–574, 1982

7. Laurens W, Battaglia C, Foglieni C, De Vos R, MalanchiniIt has been observed the PS-mediated podocyte injuryB, Van Damme B, Vanrenterghem Y, Remuzzi G, Remuzzi A:is associated with an increased tyrosine phosphorylationDirect podocyte damage in the single nephron leads to albuminuria

at the newly formed tight junctions and the basal mem- in vivo. Kidney Int 47:1078–1086, 19958. Kerjaschki D: Dysfunction of cell biological mechanisms of vis-brane of the foot processes [10]. Tyrosine kinase-regu-

ceral epithelial cells (podocytes) in glomerular diseases. Kidneylated [Ca21]i signaling has been reported in different cellInt 45:300–313, 1994

types, such as endothelia or smooth muscle cells [26, 27]. 9. Kerjaschki D: Polycation-induced dislocation of slit diaphragmsand formation of cell junctions in rat kidney glomeruli: The effectsTo study whether an inhibition of tyrosine kinase mightof low temperature, divalent cations, colchicine, and cytochalasininfluence the [Ca21]i response to PS, experiments withB. Lab Invest 39:430–440, 1978

herbimycin A, an inhibitor of tyrosine kinase, were per- 10. Kurihara H, Anderson JM, Farquhar MG: Increased Tyr phos-phorylation of ZO-1 during modification of tight junctions betweenformed. Herbimycin A did not inhibit the [Ca21]i increaseglomerular foot processes. Am J Physiol 268:F514–F524, 1995mediated by PS, indicating that the PS-mediated Ca21

11. Mundel P, Reiser J, Borja AZ, Pavenstadt H, Davidson GR,influx is not controlled by tyrosine kinase. Like herbi- Kriz W, Zeller R: Rearrangements of the cytoskeleton and cell

contacts induce process formation during differentiation of condi-mycin A, staurosporine, KN-62, and calyculin A did not

Rudiger et al: Polycations in glomerular podocytes 1709

tionally immortalized mouse podocyte cell lines. Exp Cell Res glomerular epithelial cell monolayers to albumin. J Lab Clin Med123:437–446, 1994236:248–258, 1997

21. Camussi G, Tetta C, Vercellone A: Platelets and platelet-derived12. Nitschke R, Leipziger J, Greger R: Intracellular Ca21 transientscationic proteins in human and experimental glomerular pathology.in HT29 cells induced by hypotonic cell swelling. Pflugers ArchContrib Nephrol 69:27–36, 1989423:274–279, 1993

22. Blake CCF, Koenig DF, Mair GA, North ACT, Phillips DC,13. Gogelein H, Greger R, Schlatter E: Potassium channels in the Sarma VR: Structure of hen egg-white lysozyme. Nature 206:757–basolateral membrane of the rectal gland of Squalus acanthias: 761, 1965Regulation and inhibitors. Pflugers Arch 401:107–113, 1987 23. Huber TB, Gloy J, Henger A, Schollmeyer P, Greger R, Mun-

14. Grynkiewicz G, Poenie M, Tsien RY: A new generation of Ca21del P, Pavenstadt H: Catecholamines modulate podocyte func-

indicators with greatly improved fluorescence properties. J Biol tion. J Am Soc Nephrol 9:335–345, 1998Chem 260:3340–3350, 1985 24. Pavenstadt H, Spath M, Schlunck G, Nauck M, Fischer R,

Wanner C, Schollmeyer P: Effect of nucleotides on the cytosolic15. Gogelein H, Dahlem D, Englert HC, Lang HJ: Flufenamic acid,calcium activity and inositol phosphate formation in human glo-mefenamic acid and niflumic acid inhibit single nonselective cationmerular epithelial cells. Br J Pharmacol 107:189–195, 1992channels in the rat exocrine pancreas. FEBS Lett 268:79–82, 1990

25. Baumgarten LB, Lee HC, Villereal ML: Multiple intracellular16. Schumann S, Greger R, Leipziger J: Flufenamate and Gd31 inhibitCa21 pools exist in human foreskin fibroblast cells: The effect ofstimulated Ca21 influx in the epithelial cell line CFPAC-1. Pflugers BK on release and filling of the non-cytosolic Ca21 pools. Cell

Arch 428:583–589, 1994 Calcium 17:41–52, 199517. Thastrup O, Cullen PJ, Drobak BK, Hanley MR, Dawson AP: 26. Kruse HJ, Negrescu EV, Weber PC, Siess W: Thrombin-induced

Thapsigargin, a tumor promoter, discharges intracellular Ca21Ca21 influx and protein tyrosine phosphorylation in endothelial

stores by specific inhibition of the endoplasmatic reticulum Ca21- cells is inhibited by herbimycin A. Biochem Biophys Res CommunATPase. Proc Natl Acad Sci USA 87:2466–2470, 1990 202:1651–1656, 1994

27. Himpens B, De Smedt H, Bollen M: Modulation of nucleocyto-18. Isaacs KL, Miller F: Role of antigen size and charge in immunesolic [Ca21] gradient in smooth muscle by protein phosphorylation.complex glomerulonephritis. I. Active induction of disease withFASEB J 8:879–883, 1994dextran and derivates. Lab Invest 47:198–205, 1982

28. Guasch A, Deen WM, Myers BD: Charge selectivity of the glo-19. Pasi A, Dendorfer U, Holthofer H, Nelson PJ, Tazzari S,merular filtration barrier in healthy and nephrotic humans. J ClinArmelloni S, Fornasieri A, D’Amico G, Schlondorff D: Charac- Invest 92:2274–2282, 1993

terization of nephropathy induced by immunization with high mo- 29. Ina K, Kitamura H, Nakamura M, Ono J, Takaki R: Loss oflecular weight dextran. Nephrol Dial Transplant 12:1849–1855, sulfated carbohydrate from the glomerular podocyte as a cause of1997 albuminuria in experimental diabetic rats: Ultrastructural histo-

chemical study. J Diabetes Complications 5:173–175, 199120. Hammes M, Singh A: Effect of polycations on permeability of