Kidney International, Vol. 62 (2002), pp. 1620–1627

HSP-25 and HSP-90 stabilize Na,K-ATPase in cytoskeletalfractions of ischemic rat renal cortex

BETTINA BIDMON, MICHAELA ENDEMANN, THOMAS MULLER, KLAUS ARBEITER, KURT HERKNER,and CHRISTOPH AUFRICHT

Kinderdialyse and Ludwig Boltzmann Institut, Department of Pediatrics, Universitatsklinik f ur Kinder und Jugendheilkunde,AKH Wien, Wien, Austria

HSP-25 and HSP-90 stabilize Na,K-ATPase in cytoskeletal Renal ischemia not only disrupts cellular homeostasisfractions of ischemic rat renal cortex. and causes injury but also induces a cellular stress response

Background. We recently designed an in vitro system based that participates in repair processes and may protect spe-on differential Triton-extractability of Na,K-ATPase from thecific cellular structures against subsequent injury [1, 2].cytoskeletal protein fraction isolated from rat renal cortex after

In the proximal tubule cell, non-lethal ischemic in-renal ischemia. In the present study, we hypothesized that heatshock protein (HSP)-70, HSP-25 and HSP-90 work synergisti- jury induces a transient relocation of Na,K-ATPasecally to stabilize the cytoskeletal anchorage of Na,K-ATPase. into the apical membrane domain [3]. This dislocated

Methods. Cellular proteins were fractionated by differ- Na,K-ATPase remains functional but uses adenosine 5�ential centrifugation into cytoskeletal pellets (I-PEL) obtainedtriphosphate (ATP) to pump sodium back into the urineearly (exhibiting abnormally high Triton extractability of[4]. Several clinical sequelae, such as renal sodium lossNa,K-ATPase) and non-cytoskeletal supernatants (R-SUP)

obtained late (exhibiting high abundance of HSP) after renal and reduced glomerular filtration due to tubuloglo-ischemia. For assessment of the role of HSP-70, HSP-25 and merular feedback, can be explained by this loss of polar-HSP-90 upon in vitro re-compartmentalization, I-PEL was ei-

ity [5].ther incubated in R-SUP with/without HSP antibodies, or inFor Na,K-ATPase to be translocated, it must firstbuffer with/without HSPs at different titers and combinations.

Effects were evaluated by changes of Triton extractability of dissociate from its cytoskeletal anchorage, the mem-Na,K-ATPase after co-incubation. brane-cytoskeletal-complex, which has been defined by

Results. R-SUP was shown to contain significant amounts Triton X-100 extractability [3–5]. The distribution ofof HSP-70, HSP-25 and HSP-90. Incubation of I-PEL in R-SUP

Na,K-ATPase between cytoskeletal pellet and non-cyto-reduced Triton extractability of Na,K-ATPase. Addition ofskeletal supernatant, accompanying reversible changesantibodies against each HSP significantly abolished these ef-

fects of R-SUP. Incubation of I-PEL with purified HSP-70, in cell polarity, represents one marker for the quanti-HSP-25 or HSP-90 each partly reproduced the effects of fication of the ischemic cellular injury of the proximalR-SUP, whereas the combination of all three HSP demon- tubule cell [3]. The increased Triton extractability ofstrated a strong and more than additive effect on the cytoskele-

Na,K-ATPase after renal ischemia is fully reversible dur-tal stabilization of Na,K-ATPase.ing recovery. At the time of the major restoration of theConclusions. The molecular mechanisms responsible for

postischemic re-compartmentalization of Na,K-ATPase in rat cellular distribution of Na,K-ATPase, the transcriptionrenal cortex likely involves interactions between HSP-70, of Na,K-ATPase and of other cytoskeletal elements hasHSP-25 and HSP-90, stress proteins known to be induced in been shown to be reduced even compared to the overallthe ischemic kidney.

decreased transcription rate seen after total cellularischemia [6]. Recent research thus suggests that this res-toration of the membrane-cytoskeletal-complex occursby recycling of cytoskeletal elements, rather than byincreased biosynthesis.

The heat shock proteins (HSP) are ideal candidates forKey words: heat shock proteins, post-ischemic Na,K-ATPase, stress

such post-translational repair mechanisms [7, 8]. Renalprotein, cellular injury, kidney repair, sodium pump, tubuloglomerularfeedback. ischemia, in vivo, has been shown to rapidly induce the

expression of HSP-25, HSP-70 and HSP-90 [9–12]. AllReceived for publication August 22, 2001three classes of HSP showed distinct temporal patternsand in revised form June 4, 2002

Accepted for publication June 6, 2002 of postischemic cellular localization and distribution thatrelate to the recycling of disrupted proteins during recov- 2002 by the International Society of Nephrology

1620

Bidmon et al: HSP-25 and HSP-90 stabilize Na,K-ATPase 1621

ery of the proximal tubule cell from ischemia [9–12].We have designed an in vitro system that is based ondifferential Triton-extractability of Na,K-ATPase fromrat renal cortex [13]. Using this system, we demonstratethat—at least in vitro—HSP-70 is essential for the stabili-zation of the cytoskeletal compartmentalization in ratrenal cortex after renal ischemia.

Our current study tested the hypothesis that HSP-25and HSP-90 are also essential factors for in vitro stabili-zation of Na,K-ATPase in the cytoskeletal fraction of ratrenal cortex obtained after non-lethal in vivo ischemia.

METHODS

Animal procedures

All experiments were performed in male Sprague-Dawley rats weighing 225 to 300 g and anaesthetized by80 mg/kg IP thiobutabarbital sodium. Temperature wasrectally monitored and sustained at normal values on Fig. 1. Scheme of the methodological approach of the incubation pro-

cedures. Cellular proteins were fractionated into cytoskeletal pelletsa warming board throughout the anesthesia period. A(PEL) and non-cytoskeletal supernatants (SUP) by Triton X-100 extrac-polyethylene catheter was secured in an external jugulartion of rat renal cortex obtained after renal ischemia (I) or after recovery

vein and 5 mL 0.9% saline was administered followed from renal ischemia (R). To assess the role of HSP-70, HSP-25 and/orHSP-90 in in vitro repair, aliquots of isolated I-PEL were incubated inby an infusion of 1 mL/h. The abdomen was opened byR-SUP with or without respective anti-HSP antibodies or in buffer witha median incision, kidneys exposed and carefully dis-or without purified HSP at different titers and combinations. A repeat

sected. After anticoagulation with 125 IU heparin, bilat- Triton extraction was performed and differential Triton extractability ofNa,K-ATPase was assessed by Western blots and densitometric analysis.eral renal ischemia was accomplished by selective occlu-

sion of the right renal artery and aorta just proximal tothe left renal artery. After 45 minutes the clamps wereremoved, and reperfusion visually confirmed. After re-

homogenized in the same amount of extraction buffer,flow intervals of 15 minutes the kidneys were rapidlyprotein concentrations were stable and predictable (rangeremoved in one set of animals (I). In another set of6 to 9 �g/�L). PEL were resuspended in extraction bufferanimals (R), the abdominal incision was sutured and theof half the volume of the original homogenate, resultingrats were allowed free access to food and water afterin similar protein concentrations as in SUP (4 to 6 �g/�L).recovery from anesthesia. The next day, after 18 to 20Five hundred �L aliquots of each subsequently describedhours of reflow, anesthesia was repeated and kidneysPEL and 1 mL aliquots of SUP were saved at �70�C.were rapidly harvested. After the protocol, animals were

sacrificed by intracardial KCl injection. Definition of fractions

I-PEL. This fraction, obtained during early reflowCellular protein fractionation(15 min reflow after 45 min ischemia), contains smallKidneys were encapsulated on ice, renal cortex wasamounts of HSP-70, HSP-25 and HSP-90, similar to thatisolated and homogenized in chilled extraction bufferobserved in controls and markedly increased Triton ex-(350 to 450 mg tissue in 15 mL) containing 0.1% Tritontractability of Na,K-ATPase suggestive of severely injuryX-100, 60 mmol/L piperazine-N,N*-bis-(2-ethane-sul-cytoskeletal anchorage [12].fonic acid) (PIPES), 2 mmol/L 1,2-cyclohexanediaminet-

R-SUP. This fraction, obtained at late reflow (18 toetraacedic acid, 1 mmol/L ethylenediamine-tetraacetic acid20 hours reflow after 45 min ischemia), exhibits markedly(EDTA), 1 mmol/L ethylenaglycol-bis(�-amino-ether)-increased amounts of HSP but normal extractability ofN,N�-tetraacetic acid (EGTA), 100 mmol/L sodium chlo-Na,K-ATPase [12]. To minimize interindividual variabil-ride, 0.5 mmol/L phenylmethylsulfonyl fluoride (PMSF)ity, each R-SUP was pooled from three animals (resulting0.75 mg/L leupeptin, and 0.1 mmol/L dithiothreitol (DTT)in uniformly increased HSP levels).using a Potter-Elvehjem homogenizer as previously de-

scribed [10–14]. The homogenate was centrifuged withinIncubation procedures10 to 15 minutes at 35,000 � g for 14 minutes at 4�C to

In the first set of experiments, 100 �L of I-PEL wereseparate the Triton-soluble supernatant protein fractioneither incubated in 100 �L of pure extraction buffer or(SUP � non-cytoskeletal) from the Triton-insoluble pel-

leted fraction (PEL�cytoskeletal). As each kidney was in 100 �L of buffer with 50 �g of anti-HSP-70, with 50 �g

Bidmon et al: HSP-25 and HSP-90 stabilize Na,K-ATPase1622

of anti-HSP-25, with 50 �g of anti-HSP-90 or with acombination of all three anti-HSP antibodies (SPA810;Stressgen Biotechnologies, Victoria, BC, Canada; Fig. 1).After titration experiments (performed in order to deter-mine the dose of exogenous purified HSP correspondingto endogenous HSP present in R-SUP) the second setof experiments consisted of parallel incubation of 100�L of I-PEL either in 100 �L of pure extraction bufferor in 100 �L of buffer with purified HSP-70, HSP-25 orHSP-90, each at titers similar to R-SUP (1�) and at atenth (1/10�) and at a tenfold (10�) concentration (allStressgen Biotechnologies). To assess the effects of indi-vidual HSP, in each experiment an additional aliquot of100 �L of I-PEL was incubated in 100 �L of buffer witheach HSP at 1� and 50 �g of the respective anti-HSPantibody (Stressgen Biotechnologies). As an additionalcontrol, incubation experiments were also performedwith the three HSPs after heat denaturation. These mix-tures were resuspended and incubated for 20 minutes atroom temperature. Differential centrifugation was thenrepeated at 35,000 � g for 15 minutes at 4�C. Dissociated

Fig. 2. Representative Western blot of HSP-70, HSP-25 and HSP-90supernatant fractions were saved for further analysis.in non-cytoskeletal supernatants and cytoskeletal pellets isolated byTriton X-100 extraction of rat renal cortex obtained from controls (C),Western analysis after renal ischemia (I) or after recovery from renal ischemia (R). Non-cytoskeletal supernatants isolated by Triton X-100 extraction from ratProtein concentrations were assessed in duplicates inrenal cortex obtained after recovery from renal ischemia (R-superna-each subfraction by Bradford analysis using BSA as a tants) contained significantly elevated amounts of HSP-70 (�12-fold),

standard. Protein samples were prepared at 1 �g/1 �L HSP-25 (�5-fold) and HSP-90 (�2-fold). N � 6, and all were significantat P 0.05 vs. C-supernatants and I-supernatants.concentration in sodium dodecyl sulfate (SDS) sample

buffer (150 mmol/L Tris-HCl pH 7.5, 14% glycerol,0.0025% bromphenol blue, 30 mmol/L EDTA pH 7.5,10% SDS, 10% mercaptoethanol). A total of 7.5 �g of Data analysisproteins were electrophoresed through 0.1% SDS–7.5%

Differential dissociation of Na,K-ATPase was derived(or 15%) polyacrylamide gel with 4% staking gel andfrom the ratio of specific immunodensitometric signalselectrophoretically transferred to nitrocellulose as pre-at three different exposures in the linear range of theviously described [12]. Non-specific binding sites wereprotein/signal intensity relationship, normalized to anblocked overnight in 0.5% non-fat dry milk, washed, andinternal standard, and were compared between parallelthe membranes were incubated for one hour with theincubations in each experiment.primary antibodies against HSP-25, HSP-70 and HSP-90

(Stressgen Biotechnologies and Sigma Chemical Co.,StatisticsSt. Louis, MO, USA) and against the alpha-subunit of

Wilcoxon signed rank test was used to compare dataNa,K-ATPase (Upstate Biotechnology; Lake Placid, NY,on dissociation of Na,K-ATPase between parallel incu-USA). Detection was accomplished with secondary anti-bations derived from four to eight experiments. Data,bodies (Sigma) and detected with an enhanced chemi-expressed as mean and standard deviation, were consid-luminescence system (ECL) using ECL Western blottingered to be significantly different if P 0.05.analysis system and protocols (Renaissance; NEN-Life

Science Products, Boston, MA, USA).Densitometric analysis was carried out by image analy-

RESULTSsis software (Image Master Analysis Software; Amer-Heat shock protein-70, HSP-25 and HSP-90 weresham Pharmacia Biotech, Uppsala, Sweden). To analyze

strongly expressed after renal ischemia. As shown inthe Western blots, calibration curves were constructedFigure 2, the non-cytoskeletal supernatant R-SUP, iso-to determine the linearity of the ECL detection systemlated by Triton X-100 extraction from rat renal cortexusing serial loading of cellular protein fractions. Lightobtained after recovery from renal ischemia, containsemission was proportional to protein loading in the rangesignificantly increased amounts of these HSP. As shownof 3.75 to 15 �g of a given sample, indicating that fourfold

changes of Na,K-ATPase were within the linear range. in Figure 3, dilution experiments, comparing each spe-

Bidmon et al: HSP-25 and HSP-90 stabilize Na,K-ATPase 1623

Fig. 3. Coomassie stain (left panel) of non-cytoskeletal supernatant ofrat renal cortex obtained after recovery from renal ischemia. Specificsignals of the respective HSP in rat supernatant and in buffer, at thethree titers that were used in the incubation experiments (1/10�, 1�and 10�) are shown in the representative Western blot (right panel).

Fig. 4. Representative Western blot and densitometric analyses dem-onstrating the effects of antibodies against HSP-70, HSP-25 or HSP-90cific HSP band in R-SUP by Western blots against on Triton extractability of Na,K-ATPase. In four to eight experiments,

varying concentrations of purified HSP diluted in pure aliquots of I-PEL were in parallel either incubated in pure extractionbuffer (P/I) or in R-supernatants (R/I) or in R-supernatants with thebuffer, demonstrated similar signal intensity at concen-respective anti-HSP-antibodies (RHSP Ab/I). These mixtures were incu-trations of 113 �g/mL or 2.3% of total cellular protein bated and a repeat Triton extraction performed. Densitometric analysis

for HSP-70, 48 �g/mL or 0.9% for HSP-25, and 305 shows that addition of each anti HSP (RHSP Ab/I) antibodies at leastpartly abolished the effects of R-supernatants (R/I) on cytoskeletal�g/mL or 6.2% for HSP-90 (concentration “1�”).stabilization of Na,K-ATPase. Total Na,K-ATPase remained unchangedTo evaluate for functional effects of these HSP upon with addition of antibodies. Data are shown as mean and standard devia-

the Triton extractability of Na,K-ATPase, we compared tion, P 0.05 vs. R/I.

the amount of soluble Na,K-ATPase that was extractedfrom ischemic cytoskeletal protein fractions I-PEL afterco-incubation in R-SUP with or without addition of anti- ments with each isolated HSP versus a combination of allHSP-antibodies. Figure 4 demonstrates that the addi- three HSP at the same concentrations and ratio as intion of antibodies against HSP-70, HSP-25 or HSP-90, R-SUP. Figure 6 shows that such a cocktail resulted inrespectively, each significantly decreased the in vitro re- much more effective stabilization of Na,K-ATPase thanpair effects of R-SUP on cytoskeletal stabilization of any of the isolated HSPs. The addition of HSP-25 andNa,K-ATPase as demonstrated by stronger signals for HSP-90 to HSP-70, at concentrations in which HSP-25soluble Na,K-ATPase. Total cellular Na,K-ATPase re- demonstrated only weak and HSP-90 no effects, moremained unchanged. than doubled the effect of HSP-70 on the stabilization

As a next step, this stabilizing effect for each purified of Triton extractability of Na,K-ATPase. Heat denatur-HSP in isolation was assessed. As shown in Figure 5, ation of HSP completely abolished these effects. Totalincubation of I-PEL with each of the purified HSPs re- cellular Na,K-ATPase remained unchanged.produced, at least in part, the effects of R-SUP. Thespecificity of these effects was confirmed by co-addition

DISCUSSIONof the respective anti-HSP antibody, which blocked theHSP-mediated cytoskeletal stabilization of Na,K-ATPase. This study is based on a an in vitro assay that measuresThese semiquantitative results are evidence that the ef- the changes of Triton extractability of Na,K-ATPaseficacy of HSPs in the in vitro repair assay was highest after co-incubation of isolated cytoskeletal and non-cyto-for HSP-70, less for HSP-25 and weakest for HSP-90. skeletal cellular protein fractions, obtained from rat re-Antibodies against HSP-70 also reduced the cytoskeletal nal cortex at different reflow times after in vivo renalshift of HSP-70 into the I-PEL. All experiments showed ischemia. Using this approach, we recently demonstratedthat much higher concentrations of isolated HSP than that incubation of cytoskeletal proteins obtained duringwere found in the R-SUP were necessary to achieve a early reflow after renal ischemia (exhibiting severe in-stabilization of Na,K-ATPase. Addition of purified HSP jury of the cytoskeletal anchorage of Na,K-ATPase) andor the respective antibodies had no effect on the total non-cytoskeletal proteins obtained during later reflowcellular Na,K-ATPase (data not shown). (showing high HSP abundance) resulted in transloca-

Finally, in order to investigate for interactions between tion of HSP-70 and dose-dependent stabilization ofHSP-70, HSP-25 and HSP-90, their stabilizing effect on Na,K-ATPase [13]. This stabilization of Na,K-ATPase

was reproducible by addition of purified HSP-70 andNa,K-ATPase was compared between incubation experi-

Bidmon et al: HSP-25 and HSP-90 stabilize Na,K-ATPase1624

Fig. 6. Representative Western-blot and densitometric analyses dem-onstrating the additive effects of purified HSP-70, HSP-25 and HSP-90on stabilization of Triton extractability of Na,K-ATPase in ischemia-injured cytoskeletal protein fractions. In four to eight experiments,aliquots of I-PEL are in parallel either incubated in pure extractionbuffer (P/I) or in R-supernatant (R/I) or in buffer with purified HSP-70, HSP-25 and HSP-90 each and in combination (PHSP/I). These mix-tures were incubated and a repeat Triton extraction was performed.Densitometric analysis shows that, albeit incubation of I-PEL with eachsingle HSP (PHSP/I) reproduced only in part the effects of R-SUP, thecombination of all three HSP was markedly more effective, demonstra-ting more than mere additive effect of these three families of HSP.Heat denaturation (dHSP) completely abolished any effect of this HSPmixture. Total Na,K-ATPase remained unchanged with addition ofpure HSP. Data are shown as mean and standard deviation, *P 0.05vs. P/I.

could be blocked by anti HSP-70 antibody. Therefore,we concluded that this HSP-rich supernatant containedspecific factors, including HSP-70, which were essentialfor the observed in vitro effects.

However, there is increasing evidence that HSPs areless likely to work in isolation, but rather interact duringcellular processes such as protein maturation or repair[7, 15, 16]. HSP families other than HSP-70 thereforealso may have the potential to chaperone disrupted ele-ments of the cytoskeleton and to stabilize the cytoskele-tal re-compartmentalization of the Na,K-ATPase inproximal tubule cells after renal ischemia [17]. As it was

Triton extractability of Na,K-ATPase in ischemia-injured cytoskeletalprotein fractions. In four to eight experiments, aliquots of I-PEL werein parallel either incubated in pure extraction buffer (P/I) or in R-superna-tant (R/I) or in buffer with 3 titers (1/10�, 1�, and 10�) of each purifiedHSP (PHSP/I). In addition, another aliquot of I-PEL was incubated withthe HSP (at 1�) and its specific antibody. These mixtures are incubatedand a repeat Triton extraction was performed. Densitometric analysisshows that incubation of I-PEL with purified HSP (PHSP/I) reproduced,at least in part, the effects of complete R-supernatant. The specificityof the effects of each HSP was confirmed as addition of the specific

Fig. 5. Representative Western-blot and densitometric analysis dem- antibody abolished this cytoskeletal stabilization of Na,K-ATPase. Dataonstrating the effects of HSP-70 (A), HSP-25 (B) or HSP-90 (C ) on are shown as mean and standard deviation; *P 0.05 vs. P/I.

Bidmon et al: HSP-25 and HSP-90 stabilize Na,K-ATPase 1625

not the aim of this study to detect new HSPs in renal Less is known about the role of HSP-90 in the modelof renal ischemia. HSP-90 is one of the most prevalentischemia, we focused on HSP-25 and HSP-90. These two

families of HSP have exhibited in vitro repair abilities cytosolic HSPs, and it is expressed under both normaland stressful conditions [12, 18]. Under control condi-in other systems and are known to be induced in rat

renal cortex after ischemic injury (data that were readily tions, HSP-90 keeps certain key components of the cell’sinternal signaling pathways, including steroid hormoneconfirmed in our “HSP-rich supernatant”) [9–14, 18].

The present study investigated whether in vitro sta- receptors and kinase enzymes, folded in the right confor-mation to respond rapidly to their triggering signals [22].bilization of Na,K-ATPase is dependent on HSP-25

or HSP-90, and whether these HSPs synergize with In the normal kidney, HSP-90 immunoreactivity hasbeen reported to be most pronounced in the distal convo-HSP-70.

Using our in vitro assay, there was a marked reduction luted tubule and in the cortical and medullary collectingduct. After in vivo renal ischemia [12], HSP-90 was rap-of the in vitro effect of the HSP-rich supernatant in the

presence of antibodies against HSP-25 or HSP-90. This idly induced, in a cytoplasmic pattern, in proximal tubulecells. Similarly, HSP-90 immunoreactivity was particu-agrees with our previous study that suggested that

HSP-70 was not the only essential factor for mediating larly strong in the cytoplasm in the later regenerativephase in a model of toxic renal failure, leading the au-translocation of Na,K-ATPase [13]. First, the stabiliza-

tion of Na,K-ATPase by purified HSP-70, even at high thors to conclude that HSP-90 might be involved in thedisposition of damaged proteins and synthesis of newdoses, still appeared less efficient than restoration by

the whole HSP-rich supernatant. Second, effects of the proteins [23–25].The results of the second part of our study demon-HSP-rich supernatant were only weakly influenced by

altering ATP-hydrolysis, whereas the chaperoning ef- strated that HSP-25 and HSP-90 indeed constitute essen-tial factors of the HSP-rich supernatant for the observedfects of HSP-70 are known for their pronounced ATP

dependency [14]. These data were directly confirmed in in vitro effects. Either HSP-25 or HSP-90, when incu-bated with injured cytoskeletal pellet, was able to stabi-the present study, as we re-evaluated the ability of puri-

fied HSP-70 to substitute for the HSP-rich supernatant lize Na,K-ATPase. However, the quantification of thestabilizing effects of each of these HSPs showed that theusing a more quantitative approach. Although increas-

ing titers of HSP-70 resulted in significant and dose- purified HSPs are markedly less potent than the wholeHSP-rich supernatant, even at high concentrations.dependent cytoskeletal re-compartmentalization of the

Na,K-ATPase, this in vitro effect was markedly less than In the final part of our study, we investigated theeffects of combining the three HSPs, mixed at concentra-observed using the whole supernatant, rich in HSP. Even

after a tenfold increase in purified HSP-70, it was less tions and ratios adjusted to those detected in vivo inthe HSP-rich supernatant. If these HSPs demonstratedeffective in stabilizing Na,K-ATPase than the HSP-rich

supernatant. Taken together, these results provide clear interactions or synergy in the in vitro assay, we wouldexpect more than an additive effect on stabilization ofevidence that additional factors other than HSP-70 are

important for repair activity after renal ischemia. Na,K-ATPase. The data obtained after combining HSP-70,HSP-25 and HSP-90 support the concept of synergy be-The “small” stress protein HSP-25 demonstrates se-

quence homology and structural features common to tween these proteins in translocating Na,K-ATPase afteran ischemic insult. At concentrations in which HSP-25the lens protein alpha crystalline, which is involved in

cytoskeletal organization of the lens [19]. The chicken and HSP-90 in isolation demonstrated only minimal, ifany, stabilization of Na,K-ATPase, their addition morehomolog of HSP-25 was originally described as an inhibi-

tor of actin polymerization (IAP), and showed actin- than doubled the effect of HSP-70. Observations thatHSP-70, HSP-25 and HSP-90 reactivate isolated dena-binding (capping) protein properties under control con-

ditions [18]. Overexpression of HSP-25 protected the tured proteins, as well as the interdependency betweenthese HSP, also have been demonstrated in other subcel-actin-based cytoskeleton against disruption [20]. Hence,

increasing evidence supports a major function of HSP-25 lular systems [22, 26–28]. How do these HSP stabilizeNa,K-ATPase in the cytoskeletal fraction of ischemicin the regulation of actin dynamics. As the disruption of

the actin cytoskeleton is an integral component of renal rat renal cortex? We have recently demonstrated thatendogenous as well as exogenously added HSP-70 pref-ischemia, HSP-25 therefore constitutes another attrac-

tive candidate for HSP-mediated repair processes after erentially bound to (unknown) substrates within thisfraction [13, 14]. Upon ATP hydrolysis, HSP-70 releasedrenal ischemia [10, 11, 21]. We recently described in

proximal tubule cells the postischemic induction and re- its substrate and cytoskeletal Na,K-ATPase was stabi-lized [14]. These findings were consistent with “repair”versible redistribution of HSP-25 into the detergent in-

soluble cytoskeletal fraction, a finding consistent with of a disrupted protein involved in the cytoskeletal organi-zation of the proximal tubule cell, mediated by the mo-an increased affinity for disrupted cytoskeletal elements

such as actin in this subfraction [10]. lecular chaperoning mechanisms of HSP-70 [7, 15–17].

Bidmon et al: HSP-25 and HSP-90 stabilize Na,K-ATPase1626

Other recent studies have shown that inducible HSP-25 REFERENCESonly weakly refolds disrupted proteins and does not 1. Siegel NJ, Devarajan P, Van Why SK: Renal cell injury: Meta-directly stimulate the refolding activities of HSP-70 bolic and structural alterations. (Invited Review) Pediatr Res 36:

129–136, 1994[27, 28]. In a manner similar to HSP-25, HSP-90 was2. Thadhani R, Pascual M, Bonventre JV: Acute renal failure. Nshown to prevent the aggregation of various proteins that Engl J Med 334:1448–1460, 1998

have unfolded because of cellular stress and maintained 3. Molitoris BA: Na()-K()-ATPase that redistributes to apicalmembrane during ATP depletion remains functional. Am J Physiolthem in a non-aggregated state, capable of restoring pro-265:F693–F697, 1993teins to their native conformation [24]. HSP-25 and 4. Molitoris BA, Dahl R, Geerdes A: Cytoskeletal disruption and

HSP-90 thus might be effective in our in vitro assay to apical redistribution of proximal tubule Na,K-ATPase during isch-emia. Am J Physiol 263:F488–F495, 1992convert the disrupted cytoskeletal elements to a ‘folding-

5. Molitoris BA: New insights into the cell biology of ischemic acutecompetent’ state, which can subsequently be refolded by renal failure. J Am Soc Nephrol 1:1263–1270, 1991the molecular chaperone HSP-70. Our studies, however, 6. Van Why SK, Mann AS, Ardito T, et al: Expression and molecular

regulation of Na-K-ATPase after renal ischemia. Am J Physiolalso indicate that the HSP-rich supernatant will very267:F75–F85, 1994likely contain additional HSPs, HSP-regulating proteins

7. Morimoto RI, Tissieres A, Georgopoulos C: The Biology of Heat(such as BAG-1) and other factors, which might prove to Shock Proteins and Molecular Chaperons. Cold Spring Harbor,

Cold Spring Harbor Laboratory Press, 1994, pp 1–594be essential for the efficient cellular repair after ischemic8. Minowada G, Welch WJ: Clinical implications of the stress re-injury [29, 30].

sponse. J Clin Invest 95:3–12, 1995Taken together, our results suggest that the molecular 9. Van Why SK, Hildebrandt F, Ardito T, et al: Induction and

intracellular localization of HSP-72 after renal ischemia. Am Jmechanisms responsible for postischemic restoration ofPhysiol 263:F769–F775, 1992the cytoskeletal anchorage of Na,K-ATPase in rat renal

10. Aufricht C, Ardito T, Thulin G, et al: Heat-shock protein 25cortex might involve interactions between all three fami- induction and redistribution during actin reorganization after renallies of HSP that have recently been described to be ischemia. Am J Physiol 274:F215–F222, 1998

11. Schober A, Burger Kentischer A, Muller E, et al: Effect ofinduced in renal tubules after ischemic injury. However,ischemia on localization of heat shock protein 25 in kidney. Kidneythe Triton extractability of Na,K-ATPase is only a surro- Int 54(Suppl 67):S174–S176, 1998

gate assay for the functional state of the cytoskeleton 12. Morita K, Wakui H, Komatsuda A, et al: Induction of heat-shockproteins HSP73 and HSP90 in rat kidneys after ischemia. Renal[3–5]. While re-distribution of Na,K-ATPase to its nativeFail 17:405–419, 1995cytoskeletal compartment is indeed required to re-estab-

13. Bidmon B, Endemann M, Muller T, et al: Heat shock protein-70lish its function, the reduction of its Triton extractability repairs proximal tubule structure after renal ischemia. Kidney Int

58:2400–2407, 2000does not necessarily imply its correct location or restora-14. Aufricht C, Lu E, Thulin G, et al: ATP releases HSP-72 fromtion of its function. These inherent limitations of our in

protein aggregates after renal ischemia. Am J Physiol 274:F268–vitro assay can only be solved by complementary meth- F274, 1998

15. Craig EA, Gambill BD, Nelson RJ: Heat shock proteins: Mo-ods in more complex systems, which also might permitlecular chaperons of protein biogenesis. Microbiol Rev 57:402–an assessment of the overall cellular fate.414, 1993In summary, our current study demonstrates, in an 16. Brown CR, Martin RL, Hansen WJ, et al: The constitutive and

subcellular in vitro system based on differential Triton- stress inducible forms of hsp-70 exhibit functional similarities andinteract with one another in an ATP-dependent fashion. J Cellextractability of cellular proteins isolated from rat renalBiol 120:1101–1112, 1993cortex, that stabilization of Na,K-ATPase in the cy- 17. Liang P, MacRae TH: Molecular chaperones and the cytoskeleton.

toskeletal fraction is rather effected by a certain pattern J Cell Sci 110:1431–1440, 199718. Turman MA, Kahn DA, Rosenfeld SL, et al: Characterizationof ischemia-induced HSP expression than by any given

of human proximal tubular cells after hypoxic preconditioning:isolated HSP. Using such assays for in vitro manipulationconstitutive and hypoxia-induced expression of heat shock proteins

of cellular mechanisms after in vivo renal ischemia, it HSP70 (A, B, and C), HSC70, and HSP90. Biochem Mol Med 60:49–58, 1997may be possible to further define which pattern of stress

19. Arrigo AP, Landry J: Expression and function of the low-molecu-proteins is of particular efficacy. These HSPs might thenlar-weight heat shock proteins, in The Biology of Heat Shock Pro-

become targets for in vivo manipulation to improve the teins and Molecular Chaperons, edited by Morimoto RI, Tissieresoutcome of acute ischemic renal failure. A, Georgopoulos C, Cold Spring Harbor, Cold Spring Harbor

Laboratory Press, 1994, pp 335–37420. Huot J, Houle F, Spitz DR, et al: HSP27 phosphorylation-medi-ACKNOWLEDGMENTS ated resistance against actin fragmentation and cell death induced

by oxidative stress. Cancer Res 56:273–279, 1996This work was supported by 12512-MED from the Austrian Science21. Schober A, Muller E, Thurau K, et al: The response of heatFoundation. Dr. Bidmon was a recipient of a grant from the Erwin

shock proteins 25 and 72 to ischemia in different kidney zones.Schroedinger Foundation. We are grateful to N.J. Siegel and M. Kash-Pflugers Arch 434:292–299, 1997garian for critical and helpful discussions and support. We are grateful

22. Bose S, Weikl T, Bugl H, et al: Chaperone function of Hsp90-to G. Thulin for technical support.associated proteins. Science 274:1715–1717, 1996

23. Johnson BD, Chadli A, Felts SJ, et al: Hsp90 Chaperone activityReprint requests to Prof. Dr. Christoph Aufricht, Kinderdialyse, Uni-requires the full-length protein and interaction among its multipleversitatsklinik fur Kinder und Jugendheilkunde, AKH Wien, Wahringerdomains. J Biol Chem 275:32499–32507, 2000Gurtel 18-20, A-1090 Wien, Osterreich.

E-mail: christoph.aufricht@akh-wien.ac.at 24. Jakob U, Lilie H, Meyer I, et al: Transient interaction of Hsp90

Bidmon et al: HSP-25 and HSP-90 stabilize Na,K-ATPase 1627

with early unfolding intermediates of citrate synthase: Implications 28. Wang K, Spector A: alpha-Crystallin prevents irreversible proteinfor heat shock in vivo. JBC Volume 270:7288–7294, 1995 denaturation and acts cooperatively with other heat-shock proteins

25. Nemoto TK, Ono T, Tanaka K: Substrate-binding characteristics to renature the stabilized partially denatured protein in an ATP-of proteins in the 90kDa heat shock protein family. Biochem J dependent manner. Eur J Biochem 267:4705–4712, 2000354:663–670, 2001 29. Murphy PJ, Kanelakis KC, Galigniana MD, et al: Stoichiometry,

26. Skowyra D, Georgopoulos C, Zylicz M: The E-coli DNA K gene abundance, and functional significance of the hsp90/hsp70-basedproduct, the HSP 70 homolog, can reactivate heat-inactivated RNAmultiprotein chaperone machinery in reticulocyte lysate. J Biolpolymerase in an ATP hydrolysis-dependent manner. Cell 62:939–Chem 276:30092–30098, 2001944, 1990

30. Brehmer D, Rudiger S, Gassler CS, et al: Tuning of chaperone27. Ehrnsperger M, Graber S, Gaestell M, et al: Binding of non-activity of Hsp70 proteins by modulation of nucleotide exchange.native protein to HSP25 during heat shock creates a reservoir of

folding intermediates for reactivation. EMBO J 16:221–229, 1997 Nat Struct Biol 8:427–432, 2001