Heat Shock Protein 60 Is Released in Immune-Mediated Glomerulonephritis and Aggravates Disease: In Vivo Evidence for an Immunologic Danger Signal
Andreas Lang*,
Dirk Benke,
Frank Eitner*,
Daniel Engel,
Svenja Ehrlich,
Minka Breloer,
Emma Hamilton-Williams,
Sabine Specht,
Achim Hoerauf,
Jürgen Floege*,
Arne von Bonin and
Christian Kurts
* Division of Nephrology and Clinical Immunology, University of Aachen, Aachen; Institute of Molecular Medicine and Experimental Immunology, University Clinic of Bonn, Bonn; Bernhard-Nocht-Institute for Tropical Medicine, Hamburg; and Institute for Medical Parasitology, University Clinic of Bonn, Bonn, Germany
Address correspondence to: Dr. Christian Kurts, Institute of Molecular Medicine and Experimental Immunology, Friedrich-Wilhelms-Universität, 53105 Bonn, Germany. Phone: +49-228-287-1031; Fax: +49-228-287-1052; E-mail: ckurts{at}web.de
Heat shock proteins (Hsp) are ubiquitous intracellular proteinsthat can be released in various forms of cellular stress. SomeHsp, such as Hsp60, have been shown to stimulate directly Tcell-mediated immune responses in vitro. Here, it is demonstratedthat Hsp60 is released from the kidneys and excreted into theurine of mice with nephrotoxic nephritis (NTN), a model of rapidlyprogressive glomerulonephritis. For examining the functionalrelevance of Hsp60 release, this protein was injected into micewith subnephritogenic NTN, in which only transient proteinuriaand minimal organ damage occur that do not progress to terminalkidney failure. Injection of Hsp60 strikingly aggravated disease,as evidenced by global glomerular necrosis, tubulointerstitialdamage, and complete anuria after 10 to 12 d. This effect wasmediated neither by endotoxin contaminations of Hsp60 nor byautologous antibodies. It was strictly T cell dependent butnot associated with a systemic Th1/Th2 shift. Thus, Hsp60 isan endogenous mediator stimulating immune effector mechanismsthat contribute to the progression of NTN. These findings demonstratein vivo that Hsp60 fulfills criteria of immunologic danger signalsand suggest that such signals may be involved in immune-mediatedkidney disease.
The induction of immunity usually requires antigen and non-antigen-specificstimulatory signals (1), such as adjuvants or pathogen-associatedmolecular patterns (PAMP), e.g., bacterial LPS or hypomethylatedDNA (1,2). The danger hypothesis (3) postulates endogenous immunestimulators that are released during infections by intrinsiccells of the body and that alert the immune system on the needto induce immunogenic responses. This may be important in infectionswith pathogens that do not express PAMP (3). Danger signalsare suspected to play important roles also in transplant andtumor rejection and in autoimmunity (3). Recently, the dangerhypothesis has gained substantial support by the molecular identificationof danger signals: Purified uric acid was shown to activateantigen-presenting cells (APC) and displayed adjuvant activityin vaccination experiments (4). In that study, a high molecularweight danger signal was also detected, suggesting the existenceof additional danger signals.
Heat shock proteins (Hsp) have long been suspected to act asdanger signals (3,5–7). Initially described as a familyof intracellular proteins essential for various vital cell functions(8,9), Hsp have later been demonstrated to stimulate potentlyimmune responses. Although some biologic effects observed arenow suspected to have been due to endotoxin contaminations ofrecombinant Hsp preparations, there is evidence that Hsp canbind to PAMP receptors of APC and stimulate T cell activation(5,6,9–13). Hsp are attractive as danger signals becausethey are evolutionarily highly conserved, constitutively presentin virtually all body cells, and upregulated and released inresponse to cellular stress. In infections, this release couldsupport induction of immunity. Hsp release, however, is notspecific to infections but also occurs in response to physicalstimuli, such as heat or irradiation, and in noninfectious inflammatorydiseases (9). Under these conditions, Hsp release from injuredtissue may aggravate autoimmune responses. In support of thishypothesis, injection of Hsp70 was shown recently to convertT cell tolerance into immune-mediated diabetes (7). In vitro,Hsp60 has been shown to be particularly efficient at T cellstimulation, presumably through a PAMP receptor (14–16),and thus has been proposed to represent a danger signal (5,6,10,11).
The effect of danger signals on models of kidney disease invivo is unclear. Therefore, we studied the role of Hsp60 innephrotoxic nephritis (NTN), a murine model of necrotic glomerulonephritis(GN) induced by injection of sheep antiserum against glomerularantigens nephrotoxic sheep serum (NSS). Sheep Ig is depositedin the kidney and is presented by intrinsic renal cells to CD4T cells, which are essential disease mediators (17,18). Kidneydamage is characterized by glomerular leukocyte infiltration,mesangial damage, and crescent formation. Depending on the extentof the initial damage, either glomerular repair ensues or glomerularnecrosis, tubulointerstitial damage, and progression to terminalkidney failure. Kidney disease models related to NTN were shownto be aggravated by PAMP such LPS (19) or CpG-DNA (20). As Hsp60is also thought to stimulate PAMP receptors, we investigatedits role in NTN.
Mice, Reagents, NTN Model, and Overload Proteinuria Model
Female 6- to 8-wk-old C57/BL6 mice in specific pathogen-freecondition were obtained from Bomholtgard (Ny, Denmark). Allexperiments were performed in accordance with local animal ethicsprocedures. Reagents were from Sigma-Aldrich (Steinheim, Germany),if not specified otherwise. Recombinant human Hsp60 with lowendotoxin content was from Stressgen Biomol (Hamburg, Germany)or Loke Diagnostics Aps (Arhus, Denmark). Only preparationsthat contained <1 EU endotoxin/mg Hsp60 as determined bylimulus amebocyte lysate assay were used (21). NTN was inducedby injection of NSS as described (22–24). A dose-responsecurve identified the amount of NSS required for induction ofproteinuria and terminal kidney failure (Table 1). Injectionof 4 µl/g body wt NSS led to proteinuria in most of therecipients but never to terminal kidney failure; 6 µlled to terminal kidney failure in half, and 8 µl led toterminal kidney failure in all recipients (23–25) (datanot shown). In the present study, 4 µl/g body wt NSS wasused for determining aggravation of proteinuria, and 5 µlwas used for aggravation of histologic damage. In control mice,irrelevant sheep Ig was injected. T cell depletion was performedby injection of anti-CD90 antibody (clone T24) purified fromculture supernatant. Overload proteinuria was induced by dailyintravenous injection of increasing amounts of BSA (2, 4, 6,8, and 10 mg) as described (26). All animal studies were approvedby Institutional and Government Review Boards.
Determination of Hsp60 Concentration by Western Blot
Tissue incubation supernatants, urine samples, and recombinantHsp60 were separated by electrophoresis on a polyacrylamidegel and blotted onto nitrocellulose membrane. Hsp60 was revealedusing an anti-Hsp60 monoclonal antibody (clone LK-1, isotypeIgG1; Stressgen Biomol, Hamburg, Germany). A band at 60 kD wasanalyzed by densitometry on a Gel Doc system (Bio Rad, Munich,Germany) using Quantity One 4.2 software. Hsp60 concentrationwas determined by linear regression of logarithmic density values.Spike recovery assays demonstrated that this technique alloweddetection of Hsp60 concentrations >5 µg/ml. A commercialELISA for Hsp60 (Stressgen Biomol) allowed detection of Hsp60neither in murine urine nor in serum. According to the manufacturer,the latter was noted also by other groups and was interpretedas due to soluble factors inhibiting the test.
Flow Cytometry
After treatment with FcBlock (Pharmingen, Heidelberg, Germany),cells were stained for 15 min on ice with fluorochrome-labeledmAb specific for CD3 (clone 2c11), CD4 (clone GK1.5), CD8 (clone53.6.7), B220 (clone RA36B2), NKH1–1 cells (clone PK136),CD11b (clone Mac-1), and CD11c (clone HL3; Pharmingen). Deadcells were excluded with propidium iodide. Flow cytometry wasperformed on a Becton Dickinson FACScalibur (Becton Dickinson,San Diego, CA).
Ig Levels
Serum levels of Ig specific for sheep Ig levels were assessedby ELISA (coating with 10 µg/ml sheep Ig [Serotec, Düsseldorf,Germany] overnight; blocking with 1% BSA; and serum diluted1:100 in PBS, biotinylated rat anti-murine IgG1 and IgG2a [Pharmingen],streptavidin peroxidase [Boehringer Mannheim, Mannheim, Germany],and substrate 3,5,3',5'-tetramethylbenzidine [Roth, Karlsruhe,Germany] in DMSO). Total serum IgG1 and IgG2a levels againstsheep Ig were determined using commercial ELISA kits (Pharmingen).
Cytokine Production by Splenocytes
Spleens were removed from mice, and single-cell suspensionswere obtained. Splenocytes (2 x 105 cells/well) were culturedin 100 µl of DMEM (10% FCS) for 72 h in the presence of10 µg/ml sheep globulin or on plates that were coatedwith 10 µg/ml anti-CD3 (clone 2c-11). Concentrations ofIFN-, IL-4, and IL5 in culture supernatants were measured bycommercially available ELISA kits (Pharmingen). Flow cytometryensured that the numbers of the major leukocyte populationsin culture were comparable between experimental groups duringculture (Table 2 and data not shown).
Miscellaneous Determinations
Proteinuria was determined by Bradford protein assay (Bio Rad),urine albumin excretion was determined by Albuwell M ELISA specificfor murine albumin (Exocell, Philadelphia, PA), and leukocyturiawas determined using Combur test strips (Roche, Mannheim, Germany).Determination of serum creatinine as well as the generationof paraffin sections of Methyl Carnoy’s fixed tissue andperiodic acid-Schiff (PAS) staining were performed as describedpreviously (25). For fibrinogen staining, paraffin sectionswere incubated with rabbit anti-human fibrinogen antiserum (cross-reactivewith mouse; DAKO, Glostrup, Denmark), followed by biotinylatedgoat anti-rabbit antibody (Vector, Burlingame, CA), ABC-Elitereagent (Vector), and 3,3'-diaminobenzidine with nickel chlorideenhancement. Renal Ig deposits were detected using biotinylatedhorse anti-mouse IgG (Vector). Slides were counterstained withmethyl green. Fibrinogen deposits were semiquantitatively scoredby a blinded observer who assessed the stained area in at least30 consecutive glomeruli per animal, using the following system:0 = <5%, 1 = 5 to 25%, 2 = 26 to 50%, 3 = 51 to 75%, 4 =>76% of glomerular area staining positively as described(24,27). Statistical significances of scored values were analyzedby the Mann-Whitney U test. The other determinations were analyzedusing the unpaired t test.
Release of Hsp60 in Experimental Kidney Disease
To investigate whether Hsp60 release occurred in NTN, we determinedHsp60 concentrations in the urine of mice 8 d after injectionof 5 µl/g body wt NSS (for titration, see Table 1). Whereaslow levels were found in the urine of healthy control mice (7.1± 2.4 µg/ml; n = 4), those in mice with NTN weremuch higher (80.4 ± 18.0 µg/ml; n = 5; P < 0.001;Figure 1A). Theoretically, this excretion could result fromlocal production in the kidney or could simply be due to glomerularfiltration of serum Hsp60, because NTN is characterized by highproteinuria (Figure 1B). The latter possibility was addressedby measuring excretion of murine albumin in the urine. Thisprotein has a molecular weight comparable to that of Hsp60 andshould be handled similarly during glomerular filtration. Asexpected, it was detectable in the urine of mice with NTN, whereashealthy mice showed no detectable albumin excretion (Figure 1C),demonstrating that serum proteins of approximately 60 kDare excreted in NTN. The albumin serum concentration is approximately20 mg/ml (28). Western blot analysis established that the serumconcentration of Hsp60 was <10 µg/ml (data not shown).Thus, although proteinuria as a result of NTN theoreticallyallowed filtration of Hsp60, there was little of this proteinavailable in the serum for such filtration as compared to albumin,arguing against a significant contribution of filtration toHsp60 urinary excretion. This conclusion was supported by observationsin overload proteinuria, a noninflammatory model of proteinuria(Figure 1B) (26). In this model, excretion of murine albuminwas significantly higher than in healthy mice (5.9 ±1.9 versus 0.2 ± 0.1 g/L; n = 4; P < 0.001), albeitnot as high as in NTN (Figure 1C). In contrast, Hsp60 excretionin overload proteinuria was not significantly increased (13.3± 7.8 versus 7.1 ± 2.4 µg/ml), as opposedto the impressive increase in NTN (Figure 1A). Thus, 60-kD proteinswith high serum concentrations such as albumin are excretedinto the urine in both noninflammatory and inflammatory renaldamage. In contrast, Hsp60 is present in low concentrationsin the serum and is selectively excreted into the urine in inflammatoryrenal disease. These findings suggest that the high urinaryexcretion of Hsp60 in NTN was mostly due to local release fromthe kidney.
Figure 1. Release of heat shock protein 60 (Hsp60) in nephrotoxic nephritis (NTN). (A through C) Urine samples from groups of healthy mice () 8 d after induction of NTN () or protein overload (
) were assayed for concentrations of Hsp60 (A), total protein (B), and murine albumin (C). (D through E) Kidneys and livers from mice before () or 8 d after () NTN induction (D) and kidneys 2 d after ligation of the left ureter () and the contralateral kidney (; E) were explanted, decapsulated, and incubated as whole organs in 2 ml of Hanks buffer for 2 h at 4°C to allow extracellular Hsp60 to diffuse from the organ into the medium. Hsp60 concentration in culture supernatant and urine was determined by Western blot analysis. Shown are means and SD of groups of four to five animals. Similar results were seen in a separate experiment in groups of three to four mice.
To substantiate this conclusion, we designed an assay to quantifyrelease of Hsp60 into the extracellular compartment of the diseasedkidney. This was not possible by immunohistochemistry becauseof the high constitutive intracellular content of Hsp60 in mostkidney cells, in particular in tubular epithelial cells (29–31).Instead, we cultured whole murine kidneys 8 d after injectionof NSS for 2 h at 4°C. This approach was chosen becausehomogenization of the kidney would lead to cellular destructionand to release of intracellular Hsp60. We speculated that theincubation of an explanted kidney for a short time would allowextracellular Hsp60 to diffuse into the culture supernatant.This incubation was performed at 4°C, under the assumptionthat this would reduce metabolic processes causing unwantedHsp60 release, for example as a result of energy or oxygen deprivationof the explanted organ. Indeed, Hsp60 release from kidneys withNTN was significantly increased as compared with that from normalkidneys (34.1 ± 14.6 versus 6.4 ± 5.2 µg/mlper g organ wt; n = 4; P = 0.01; Figure 1D). No significantincrease was seen in the supernatants of the livers (NTN 3.2± 0.5 versus control 2.7 ± 1.7 µg/ml perg organ wt; n = 3; Figure 1D) of the same animals, demonstratingorgan specificity of Hsp60 release in NTN. Release of Hsp60was detected also in a non-immune-mediated yet inflammatorykidney disease model, obstructive uropathy (obstructed kidney10.7 ± 1.1 versus contralateral kidney 4.73 ±1.1 µg/ml per g organ wt; n = 5; P = 0.06; Figure 1E).However, this release was smaller and statistically less significantas compared with NTN. Thus, Hsp60 release occurred also in non-immune-mediatedrenal disease.
Hsp60 Severely Aggravates NTN
To investigate the consequences of Hsp60 release in NTN, wedecided to co-inject recombinant Hsp60 in this disease model.To be able to detect aggravation of NTN, we used a low doseof NSS (4 µl/g body wt), which caused minor proteinuriaand leukocyturia (Table 1). A single co-injection of Hsp60 6h after NSS aggravated proteinuria and leukocyturia, first detectable4 to 5 d after disease induction and very marked on day 12 (Figure 2, A and B),when kidneys were taken for histologic examination.On that day, histologic signs of nephritis were manifest onlyin mice that received a co-injection of NSS and Hsp60 (datanot shown). Quantification of glomerular fibrinogen depositionby immunohistochemistry has been used as a parameter for damagein NTN (27). However, significant increase of this parameterwas not seen under these conditions (Figure 2C). To render histologicchanges more evident, we performed a series of experiments usinga slightly higher dose of NSS (5 µl/g body wt). This causedproteinuria in most animals from day 3 on, which was increasedin the group that received co-injections of Hsp60 on day 7 (Figure 2D).From then on, animals that received a co-injection beganto show signs of wasting. On day 12, some animals had succumbedto disease. The surviving mice showed elevated serum creatininelevels (Figure 2E) and extremely high proteinuria or were anuric.Histology revealed pronounced signs of kidney damage in theseanimals: Severe global glomerular necrosis, tubulointerstitialdamage, multiple PAS-positive protein casts in atrophic, anddilated tubuli were apparent (Figure 3). In contrast, animalsthat received a injection of NSS alone showed only segmentalglomerular necrosis, focal PAS-positive protein casts, tubularatrophy and dilation, and no extracapillary proliferation (Figure 3).Sections of mice that received an injection of Hsp60 alonewere indistinguishable from those of control animals. The glomeruliof mice that received a co-injection showed significantly higherfibrinogen deposits than those of mice that were given NSS alone(Figures 2F and 3), further illustrating aggravation of diseaseby Hsp60.
Figure 2. Hsp60 severely aggravates NTN. (A through C) Groups of four mice received an intravenous injection of 4 µl/g body wt nephrotoxic sheep serum (NSS) or sheep Ig, followed by 50 µg of Hsp60 or PBS after 6 h. On day 12, proteinuria (A), leukocyturia (B), and glomerular fibrinogen deposits (C) were determined. (D through F) Groups of four mice received intravenous injections of 5 µl/g body wt NSS or sheep Ig, followed after 6 h by 50 µg of Hsp60, 50 µg of Hsp60 boiled for 2 h, or PBS. Proteinuria for day 7 is given (D) because one of the four animals in the NSS+Hsp60 group died on day 8 and two had become anuric by day 12, when serum was taken for creatinine determination (E) and kidneys were taken for immunohistologic assessment of fibrinogen (F) deposits. Similar results were obtained in a separate experiment in groups of three mice.
Figure 3. Kidneys of mice from experiments using 5 µl of NSS (Figure 2, C through F) were taken for representative histology. Paraffin sections were stained for periodic acid-Schiff (rows 1 and 2), fibrinogen (row 3), and IgG deposition (row 4). Arrows indicate a segmental mesangial necrosis (row 2) or segmental fibrinogen deposits (row 3). Sections from a previous study (24) were used as a positive control and showed clear Ig deposits (data not shown).
Mechanisms Involved in Hsp60-Mediated Aggravation of NTN
Some biologic effects of Hsp60 have been suggested to resultfrom endotoxin contaminations in certain commercial preparations(32,33). Furthermore, a rat NTN model has been reported to beaffected by LPS (19). To exclude such effects in our systems,we determined the endotoxin content of all Hsp60 preparationsby limulus amebocyte lysate assay and used only preparationswith <1 EU endotoxin/mg Hsp60 (equivalent to 1 ng of theLPS used in the present study). These preparations were shownrecently to improve T cell activation and effector functionsindependent of endotoxin contaminations (21). To control foran in vivo influence of such small endotoxin contamination,we took advantage of the heat stability of endotoxins and includedin the experiments using 5 µl/g body wt NSS an experimentalgroup that received co-injections of boiled Hsp60. Boiling denaturesproteins and has been used to incapacitate various Hsp (7,11,13,34).In mice that received a co-injection of such boiled Hsp60, proteinuria(Figure 2D), creatinine levels (Figure 2E), leukocyturia (datanot shown), fibrinogen deposits (Figure 2F), and histologicdamage (Figure 3) were similar to those observed in mice thathad received NSS alone, demonstrating that the aggravation ofNTN was mediated by a heat-sensitive component of Hsp60. Toinvestigate further the influence of potential endotoxin contaminations,we co-injected 1 ng of LPS with NSS. This dose was 20-fold higherthan the LPS contamination detected in the 50-µg Hsp60that was used in the co-injection experiments in this study.This dose of LPS did not aggravate proteinuria (Figure 4, A and B),serum creatinine (NSS: 20.5 ± 1.3 µM, NSS+LPS:20.7 ± 1.8 µM; n = 4), glomerular fibrinogen deposits(Figure 4C), or histologic damage (data not shown). In summary,these controls indicate that endotoxin was not responsible forthe Hsp60-mediated aggravation of NTN.
Figure 4. T cells are essential for Hsp60-mediated aggravation of NTN. Mice were depleted of T cells by intraperitoneally injection of T24 antibody or rat Ig as a control. On the next day, 5 µl/g body wt NSS or sheep Ig was injected, followed by 50 µg of Hsp60, 10 ng of LPS, or PBS after 6 h. Proteinuria was given on day 7 (1) and day 12 (B). Fibrinogen deposition in kidney sections (B) were determined on day 12. In the NSS+Hsp60 control group, three of four mice became anuric between day 9 and day 11, and two succumbed to disease on day 11. Therefore, data points () were given instead of SD in this group. Similar results were obtained in a separate experiment in groups of two to three mice.
Previous studies have demonstrated that T cells are essentialin NTN (17). Hsp60 has been shown to increase T cell proliferationand cytokine production in vitro (6,11). To evaluate whetherthe Hsp60-mediated aggravation of NTN was T cell dependent,we depleted T cells in mice using anti-Thy1-specific mAb andinduced NTN on the following day. On that day, the proportionof T cells in the blood was reduced from 35 to 40% of totalCD3+ lymphocytes to <1% (data not shown). This depletiondid not significantly change proteinuria in mice that had receivedNSS without Hsp60 (Figure 3D). In contrast, the aggravationof NTN after co-injection of Hsp60 was completely abrogatedin the absence of T cells, as determined by histologic damage(Figure 4C and data not shown) and proteinuria (Figure 4, A and B).T cell-depleted mice showed only sporadic glomerulardamage and neither fibrinogen deposits (Figure 4C) nor tubulointerstitialdamage (data not shown), regardless of whether Hsp60 was co-injectedor not. The functionality of the Hsp60 preparation used in thisexperiment was demonstrated by aggravation of NTN in the positivecontrol group that had received NSS, Hsp60, and rat Ig as acontrol for the anti-Thy1 antibody. Only in this group was proteinuriaincreased by day 7 (Figure 4A), and on day 12, two mice haddied, one mouse was anuric, and one showed severe proteinuria(Figure 3D). Both surviving mice showed glomerular necrosis,tubular atrophy, mononuclear tubulointerstitial infiltration,and fibrinogen deposits (Figure 3C). Thus, T cells were essentialfor Hsp60-mediated aggravation of NTN. On the day of analysis,i.e., day 12, CD3+ T cell numbers in the blood had risen toapproximately half of that in control animals, indicating thatT cells were required in the early induction phase of diseaseto mediate Hsp60-induced aggravation of disease.
A shift of the T cell response toward Th1 has been shown tobe associated with crescentic GN (18,35). Hsp60 caused a Th1shift in vitro (6,36). To analyze whether Hsp60 aggravated NTNin our model via a Th1 shift, we examined the ability of spleencells to produce cytokines in response to sheep Ig in vitro.Significant changes were observed in the production neitherof the Th1 cytokine IFN- nor of the Th2 cytokines IL-4 and IL-5(Table 2). This could not be explained by changes in the cellularcomposition of splenic leukocyte populations, as the total cellnumbers and the proportions of T cells (Table 2) and of B cells,T cells, NK cells, macrophages, and CD11b+ and CD11b–dendritic cells (data not shown) were not different betweenexperimental groups. A Th1-type response can also be demonstratedby a shift of from the serum Ig subclass IgG1 to IgG2a. We didnot detect significant changes in total IgG1 concentrationsbetween mice that received an injection of Hsp60 or not (Table 2).It is interesting that nephritic animals showed lower totalIgG2a levels as compared with controls that received an injectionof control sheep serum (Table 2). When we examined Ig titersspecific for sheep Ig, a decrease was seen for both subclasses(Table 2), arguing against an important role of the autologousphase in our model. This was supported by immunohistochemicalstaining of kidney sections for deposits of murine Ig. No suchdeposits were detected in the glomeruli and in the tubulointerstitium(Figure 3). Taken together, these findings do not support aTh1 shift or an aggravation of the autologous phase as underlyingcauses for the T cell-mediated aggravation of NTN induced byHsp60.
Recent studies have demonstrated that uric acid (4) and Hsp70(7) can activate APC and boost resulting T cell responses inanimal models. These findings are considered to represent thefirst in vivo identifications of immunologic danger signals.These signals have been defined as endogenous mediators releasedin harmful conditions, such as infections, that stimulate APCto induce immunogenic responses (3). Our present study providesin vivo evidence that also Hsp60 fulfills criteria of a dangersignal. We demonstrated that Hsp60 was released from the kidneyin NTN and aggravated this disease model. Experimental conditionsthat normally cause minor kidney damage and glomerular repairled to complete organ failure in the presence of Hsp60. Theseresults demonstrate that this molecule can shift the balancebetween tolerance and immunity toward immunity. They are consistentwith recent findings in a transgenic model of diabetes, in whichHsp70 converted T cell tolerance into autoimmunity (7). Ourstudy extends this previous report to a nontransgenic situationin a different organ and by demonstrating release of the dangersignal from the affected organ. Upregulation of various Hsp,including Hsp60, has been demonstrated by immunohistochemistryin kidney disease (29–31), but there has been little directevidence for their release. Our findings do not prove that Hsp60release from the kidney can directly aggravate NTN. This directproof would require the ability to specifically block Hsp60function in the kidney. Nevertheless, our findings support theidea of a positive feedback loop consisting of kidney disease-mediatedrelease of Hsp60, immune-mediated aggravation of kidney disease,further release of Hsp60, and so on. Such vicious circles mayrepresent general mechanisms relevant in the perpetuation ofimmune-mediated diseases.
Hsp60-mediated aggravation of NTN involved immune effector mechanisms,because it was strictly T cell dependent. CD4 T cells have beenshown to be essential in the progression of crescentic GN toterminal kidney failure (17). Recent studies suggest APC aslikely intermediates between Hsp60 and T cells, as they seemedto express an Hsp60 receptor whose stimulation led to APC activation(10,14–16). In vitro, APC activated under the influenceof Hsp60 were shown to induce CD4 T cells to produce increasedamounts of inflammatory cytokines such as IFN- and IL-12 (5,6,11),which have been shown to aggravate crescentic GN (18). An involvementof APC in our findings, however, remains to be shown. A rolefor CD8 T cells is less likely, as these cells were irrelevantin crescentic GN (35). Furthermore, Hsp60, unlike Hsp70 andgp96, does not seem to introduce chaperoned peptides into thecross-presentation pathway, which could recruit CD8 T cellsas effectors. T cell responses against Hsp70 have been shownto be important in CdCl2-induced interstitial nephritis (37),and T cells specific for Hsp60 have been shown to contributeto colitis in immunocompromised mice (38). However, if thisplayed a role in the present study, then it would be difficultto explain why pathology was confined to the kidney and whysystemic administration of Hsp60 alone did not cause any pathology.An increase in the autologous Ig response was not observed,consistent with previous observations in crescentic GN (35).This response was even diminished, suggesting that humoral immuneresponses were not accountable for the Hsp60-mediated effects.Likewise, we did not detect an Hsp60-induced shift of the immuneresponse against sheep Ig toward Th1, which has been shown toaggravate crescentic GN (18,35) and which is a well-documentedeffect of Hsp60 (6,36). It is interesting that we noted in nephriticanimals, regardless of Hsp60, a decrease in total Ig2a serumlevels (Table 2) and an increase in splenic production of IL-4and IL-5 induced by anti-CD3 antibody (data not shown), whichstimulates all T cells, including those not specific for thenephritogenic sheep Ig antigen. These findings argue againsta systemic Th1 shift and suggest that the NSS used in the presentstudy (22) induced immune reactions different from those instudies on crescentic GN (35). In summary, these results donot support a systemic or sheep Ig-specific Th1 shift or anaggravation of the autologous phase as crucial mechanisms inHsp60-mediated aggravation of NTN.
The target cell stimulated by Hsp60 in our system remains tobe identified. Addressing this question would be possible bytargeting the elusive Hsp60 receptor. As candidate receptors,the Toll-like receptors TLR 2 (14,16), TLR 4 (14,15), and CD14(10) have been proposed. These molecules have also been shownto mediate APC activation by pathogen-derived endotoxins (32).Therefore, we performed extensive controls to exclude that endotoxincontaminations in the recombinant Hsp60 preparations used wereresponsible for our observations. These controls do not strictlyrule out the notion that Hsp60 may bind small amounts of endotoxinsand increase their ability to stimulate PAMP receptors (32),but even if Hsp60 were not sufficient on its own, it would stillbe an essential component of a danger signal. Strong evidencefor endotoxin-independent effects are transgenic mice expressingthe Hsp gp96, in which ubiquitous APC activation and a lupus-likephenotype were observed (39). In such transgenic animals, endotoxincontaminations could be excluded. Finally, direct stimulationof T cells by Hsp60, as recently demonstrated in vitro (16),is unlikely to explain the present findings, because such stimulationwas reported to attenuate the resulting T cell response.
Our findings may have diagnostic and therapeutic implications.The presence of Hsp60 in the urine of animals with NTN raisesthe possibility of using such molecules as noninvasive markersfor immune-mediated kidney disease. Blocking of Hsp60 functionis not likely to represent a feasible therapeutic approach,because of the numerous vital housekeeping functions of Hsp(8). Even if specific blockade of Hsp60 or its receptor werepossible, then other Hsp, such as Hsp70 or gp96, may possessoverlapping functionality, so incapacitation of only one suchprotein may be compensated by others. However, components ofcommon Hsp signaling pathways may represent promising targetsfor therapeutic strategies to slow progression of immune-mediatedkidney disease.
Acknowledgments
This work was funded by a project research grant of the Germanstate of Nordrhein-Westfalen. C.K. was supported by a Heisenbergfellowship from the Deutsche Forschungsgemeinschaft (grant Ku1063/2-1).A.H. acknowledges support from the Deutsche Forschungsgemeinschaft(grant Ho2009/1-4). M.B. is funded by the Fritz Bender Foundation.
We thank Birgit Gerstner and Steffi Schweistal for animal husbandryand Monika Wirtz for technical assistance. Dr. K. Assmann, Universityof Nijmegen, The Netherlands, kindly provided nephrotoxic sheepserum.
Footnotes
Published online ahead of print. Publication date availableat www.jasn.org.
Janeway CA Jr, Medzhitov R: Innate immune recognition.
Annu Rev Immunol 20
: 197
–216, 2002[CrossRef][Medline]
Schnare M, Barton GM, Holt AC, Takeda K, Akira S, Medzhitov R: Toll-like receptors control activation of adaptive immune responses.
Nat Immunol 2
: 947
–950, 2001[CrossRef][Medline]
Gallucci S, Matzinger P: Danger signals: SOS to the immune system.
Curr Opin Immunol 13
: 114
–119, 2001[CrossRef][Medline]
Shi Y, Evans JE, Rock KL: Molecular identification of a danger signal that alerts the immune system to dying cells.
Nature 425
: 516
–521, 2003[CrossRef][Medline]
Chen W, Syldath U, Bellmann K, Burkart V, Kolb H: Human 60-kDa heat-shock protein: A danger signal to the innate immune system.
J Immunol 162
: 3212
–3219, 1999[Abstract/Free Full Text]
Breloer M, Dorner B, More SH, Roderian T, Fleischer B, von Bonin A: Heat shock proteins as "danger signals": Eukaryotic Hsp60 enhances and accelerates antigen-specific IFN-gamma production in T cells.
Eur J Immunol 31
: 2051
–2059, 2001[CrossRef][Medline]
Millar DG, Garza KM, Odermatt B, Elford AR, Ono N, Li Z, Ohashi PS: Hsp70 promotes antigen-presenting cell function and converts T-cell tolerance to autoimmunity in vivo.
Nat Med 9
: 1469
–1476, 2003[CrossRef][Medline]
Hartl FU: Molecular chaperones in cellular protein folding.
Nature 381
: 571
–579, 1996[CrossRef][Medline]
Srivastava PK, Menoret A, Basu S, Binder RJ, McQuade KL: Heat shock proteins come of age: Primitive functions acquire new roles in an adaptive world.
Immunity 8
: 657
–665, 1998[CrossRef][Medline]
Kol A, Lichtman AH, Finberg RW, Libby P, Kurt-Jones EA: Cutting edge: Heat shock protein (HSP) 60 activates the innate immune response: CD14 is an essential receptor for HSP60 activation of mononuclear cells.
J Immunol 164
: 13
–17, 2000[Abstract/Free Full Text]
Bethke K, Staib F, Distler M, Schmitt U, Jonuleit H, Enk AH, Galle PR, Heike M: Different efficiency of heat shock proteins (HSP) to activate human monocytes and dendritic cells: Superiority of HSP60.
J Immunol 169
: 6141
–6148, 2002[Abstract/Free Full Text]
Habich C, Baumgart K, Kolb H, Burkart V: The receptor for heat shock protein 60 on macrophages is saturable, specific, and distinct from receptors for other heat shock proteins.
J Immunol 168
: 569
–576, 2002[Abstract/Free Full Text]
Kol A, Bourcier T, Lichtman AH, Libby P: Chlamydial and human heat shock protein 60s activate human vascular endothelium, smooth muscle cells, and macrophages.
J Clin Invest 103
: 571
–577, 1999[Medline]
Vabulas RM, Ahmad-Nejad P, da Costa C, Miethke T, Kirschning CJ, Hacker H, Wagner H: Endocytosed HSP60s use toll-like receptor 2 (TLR2) and TLR4 to activate the toll/interleukin-1 receptor signaling pathway in innate immune cells.
J Biol Chem 276
: 31332
–31339, 2001[Abstract/Free Full Text]
Ohashi K, Burkart V, Flohe S, Kolb H: Cutting edge: Heat shock protein 60 is a putative endogenous ligand of the toll-like receptor-4 complex.
J Immunol 164
: 558
–561, 2000[Abstract/Free Full Text]
Zanin-Zhorov A, Nussbaum G, Franitza S, Cohen IR, Lider O: T cells respond to heat shock protein 60 via TLR2: Activation of adhesion and inhibition of chemokine receptors.
FASEB J 17
: 1567
–1569, 2003[Abstract/Free Full Text]
Li S, Kurts C, Kontgen F, Holdsworth SR, Tipping PG: Major histocompatibility complex class II expression by intrinsic renal cells is required for crescentic glomerulonephritis.
J Exp Med 188
: 597
–602, 1998[Abstract/Free Full Text]
Timoshanko JR, Holdsworth SR, Kitching AR, Tipping PG: IFN-gamma production by intrinsic renal cells and bone marrow-derived cells is required for full expression of crescentic glomerulonephritis in mice.
J Immunol 168
: 4135
–4141, 2002[Abstract/Free Full Text]
Karkar AM, Rees AJ: Influence of endotoxin contamination on anti-GBM antibody induced glomerular injury in rats.
Kidney Int 52
: 1579
–1583, 1997[Medline]
Anders HJ, Banas B, Linde Y, Weller L, Cohen CD, Kretzler M, Martin S, Vielhauer V, Schlondorff D, Grone HJ: Bacterial CpG-DNA aggravates immune complex glomerulonephritis: Role of TLR9-mediated expression of chemokines and chemokine receptors.
J Am Soc Nephrol 14
: 317
–326, 2003[Abstract/Free Full Text]
Osterloh A, Meier-Stiegen F, Veit A, Fleischer B, von Bonin A, Breloer M: Lipopolysaccharide-free heat shock protein 60 activates T cells.
J Biol Chem 279
: 47906
–47911, 2004[Abstract/Free Full Text]
Assmann KJ, Tangelder MM, Lange WP, Schrijver G, Koene RA: Anti-GBM nephritis in the mouse: Severe proteinuria in the heterologous phase.
Virchows Arch A Pathol Anat Histopathol 406
: 285
–299, 1985[CrossRef][Medline]
Neugarten J, Feith GW, Assmann KJ, Shan Z, Stanley ER, Schlondorff D: Role of macrophages and colony-stimulating factor-1 in murine antiglomerular basement membrane glomerulonephritis.
J Am Soc Nephrol 5
: 1903
–1909, 1995[Abstract/Free Full Text]
Janssen U, Ostendorf T, Gaertner S, Eitner F, Hedrich HJ, Assmann KJ, Floege J: Improved survival and amelioration of nephrotoxic nephritis in intercellular adhesion molecule-1 knockout mice.
J Am Soc Nephrol 9
: 1805
–1814, 1998[Abstract]
Krüger T, Benke D, Eitner F, Lang A, Wirtz M, Hamilton-Williams EE, Engel D, Giese B, Müller-Newen G, Floege J, Kurts C: Identification and functional characterization of dendritic cells in the healthy murine kidney and in experimental glomerulonephritis.
J Am Soc Nephrol 15
: 613
–621, 2004[Abstract/Free Full Text]
Eddy AA: Interstitial nephritis induced by protein-overload proteinuria.
Am J Pathol 135
: 719
–733, 1989[Abstract]
Heeringa P, van Goor H, Itoh-Lindstrom Y, Maeda N, Falk RJ, Assmann KJ, Kallenberg CG, Jennette JC: Lack of endothelial nitric oxide synthase aggravates murine accelerated anti-glomerular basement membrane glomerulonephritis.
Am J Pathol 156
: 879
–888, 2000[Abstract/Free Full Text]
Brenner DA, Buck M, Feitelberg SP, Chojkier M: Tumor necrosis factor-alpha inhibits albumin gene expression in a murine model of cachexia.
J Clin Invest 85
: 248
–255, 1990
Somji S, Todd JH, Sens MA, Garrett SH, Sens DA: Expression of heat shock protein 60 in human proximal tubule cells exposed to heat, sodium arsenite and CdCl(2).
Toxicol Lett 115
: 127
–136, 2000[CrossRef][Medline]
Birk OS, Gur SL, Elias D, Margalit R, Mor F, Carmi P, Bockova J, Altmann DM, Cohen IR: The 60-kDa heat shock protein modulates allograft rejection.
Proc Natl Acad Sci U S A 96
: 5159
–5163, 1999[Abstract/Free Full Text]
Trieb K, Dirnhofer S, Krumbock N, Blahovec H, Sgonc R, Margreiter R, Feichtinger H: Heat shock protein expression in the transplanted human kidney.
Transpl Int 14
: 281
–286, 2001[CrossRef][Medline]
Wallin RP, Lundqvist A, More SH, von Bonin A, Kiessling R, Ljunggren HG: Heat-shock proteins as activators of the innate immune system.
Trends Immunol 23
: 130
–135, 2002[CrossRef][Medline]
Gao B, Tsan MF: Recombinant human heat shock protein 60 does not induce the release of tumor necrosis factor alpha from murine macrophages.
J Biol Chem 278
: 22523
–22529, 2003[Abstract/Free Full Text]
Arnold D, Faath S, Rammensee H, Schild H: Cross-priming of minor histocompatibility antigen-specific cytotoxic T cells upon immunization with the heat shock protein gp96.
J Exp Med 182
: 885
–889, 1995[Abstract/Free Full Text]
Tipping PG, Holdsworth SR: T cells in glomerulonephritis.
Springer Semin Immunopathol 24
: 377
–393, 2003[CrossRef][Medline]
Flohe SB, Bruggemann J, Lendemans S, Nikulina M, Meierhoff G, Flohe S, Kolb H: Human heat shock protein 60 induces maturation of dendritic cells versus a Th1-promoting phenotype.
J Immunol 170
: 2340
–2348, 2003[Abstract/Free Full Text]
Weiss RA, Madaio MP, Tomaszewski JE, Kelly CJ: T cells reactive to an inducible heat shock protein induce disease in toxin-induced interstitial nephritis.
J Exp Med 180
: 2239
–2250, 1994[Abstract/Free Full Text]
Kuckelkorn U, Ruppert T, Strehl B, Jungblut PR, Zimny-Arndt U, Lamer S, Prinz I, Drung I, Kloetzel PM, Kaufmann SH, Steinhoff U: Link between organ-specific antigen processing by 20S proteasomes and CD8(+) T cell-mediated autoimmunity.
J Exp Med 195
: 983
–990, 2002[Abstract/Free Full Text]
Liu B, Dai J, Zheng H, Stoilova D, Sun S, Li Z: Cell surface expression of an endoplasmic reticulum resident heat shock protein gp96 triggers MyD88-dependent systemic autoimmune diseases.
Proc Natl Acad Sci U S A 100
: 15824
–15829, 2003[Abstract/Free Full Text]
Received for publication April 6, 2004.
Accepted for publication November 10, 2004.
This article has been cited by other articles:
H. Bendz, B.-C. Marincek, F. Momburg, J. W. Ellwart, R. D. Issels, P. J. Nelson, and E. Noessner Calcium Signaling in Dendritic Cells by Human or Mycobacterial Hsp70 Is Caused by Contamination and Is Not Required for Hsp70-mediated Enhancement of Cross-presentation
J. Biol. Chem.,
September 26, 2008;
283(39):
26477 - 26483.
[Abstract][Full Text][PDF]
A. Osterloh, A. Veit, A. Gessner, B. Fleischer, and M. Breloer Hsp60-mediated T cell stimulation is independent of TLR4 and IL-12
Int. Immunol.,
March 1, 2008;
20(3):
433 - 443.
[Abstract][Full Text][PDF]
J. Scholz, V. Lukacs-Kornek, D. R. Engel, S. Specht, E. Kiss, F. Eitner, J. Floege, H.-J. Groene, and C. Kurts Renal Dendritic Cells Stimulate IL-10 Production and Attenuate Nephrotoxic Nephritis
J. Am. Soc. Nephrol.,
March 1, 2008;
19(3):
527 - 537.
[Abstract][Full Text][PDF]
H. Bendz, S. C. Ruhland, M. J. Pandya, O. Hainzl, S. Riegelsberger, C. Brauchle, M. P. Mayer, J. Buchner, R. D. Issels, and E. Noessner Human Heat Shock Protein 70 Enhances Tumor Antigen Presentation through Complex Formation and Intracellular Antigen Delivery without Innate Immune Signaling
J. Biol. Chem.,
October 26, 2007;
282(43):
31688 - 31702.
[Abstract][Full Text][PDF]
A. Osterloh, U. Kalinke, S. Weiss, B. Fleischer, and M. Breloer Synergistic and Differential Modulation of Immune Responses by Hsp60 and Lipopolysaccharide
J. Biol. Chem.,
February 16, 2007;
282(7):
4669 - 4680.
[Abstract][Full Text][PDF]
Top
Abstract
Introduction
, IL-4, and IL5 in culture supernatants were measured by commercially available ELISA kits (Pharmingen). Flow cytometry ensured that the numbers of the major leukocyte populations in culture were comparable between experimental groups during culture (Table 2 and data not shown). 
) 8 d after induction of NTN (
) or protein overload (
) were assayed for concentrations of Hsp60 (A), total protein (B), and murine albumin (C). (D through E) Kidneys and livers from mice before (
) were given instead of SD in this group. Similar results were obtained in a separate experiment in groups of two to three mice.