Oxidant Stress and Reduced Antioxidant Enzyme Protection in Polycystic Kidney Disease
Robin L. Maser,
Dianne Vassmer,
Brenda S. Magenheimer and
James P. Calvet
Department of Biochemistry and Molecular Biology, University of Kansas Medical Center, Kansas City, Kansas.
Correspondence to: Dr. Robin L. Maser, Department of Biochemistry and Molecular Biology, University of Kansas Medical Center, 4024 WHE, 3901 Rainbow Boulevard, Kansas City, KS 66160-7421. Phone: 913-588-7425; Fax: 913-588-7440; E-mail: rmaser{at}kumc.edu
ABSTRACT. Oxidative stress has been implicated in the pathogenesisof both acquired and hereditary polycystic kidney disease. Mechanismsof oxidant injury in C57BL/6J-cpk mice and Han:SPRD-Cy ratswith rapidly or slowly progressive polycystic kidney diseasewere explored. Expression of heme oxygenase-1 mRNA, an induciblemarker of oxidative stress, was shown to be increased in cystickidneys of mice and rats in a pattern that reflected diseaseseverity. By contrast, there was a decrease in mRNA expressionof the antioxidant enzymes extracellular glutathione peroxidase,superoxide dismutase, catalase, and glutathione S-transferaseduring disease progression. Renal mRNA levels of these enzymeswere strikingly reduced in rapidly progressive disease in homozygouscystic mice and rats. In slowly progressive disease in heterozygousrats, renal antioxidant mRNA levels were decreased to a greaterextent in cystic males than in the less severely affected females.Protein levels for extracellular glutathione peroxidase werealso reduced in plasma and in cystic kidneys of mice and rats.Plasma extracellular glutathione peroxidase enzymatic activitywas also decreased, whereas the lipid peroxidation productsmalondialdehyde and 4-hydroxy-2(E)-nonenal were increased inkidneys and blood plasma of cystic mice. Reduced antioxidantenzyme protection and increased oxidative damage represent generalmechanisms in the pathogenesis of polycystic kidney disease.
Polycystic kidney disease (PKD) is a common disorder whose pathogenesisis incompletely understood. PKD is primarily characterized bythe progressive development of renal epithelial cysts and inmost instances results in end-stage renal disease. PKD can beinherited as an autosomal recessive (ARPKD) or autosomal dominant(ADPKD) trait with rapidly or slowly progressive clinical courses,respectively. In addition, PKD can be acquired, as seen in themajority of long-term dialysis patients (1). Two genes responsiblefor ADPKD have been identified. They encode proteins that areproposed to function in a macromolecular ion channel-receptorcomplex that transduces signals regulating the state of differentiationof renal epithelial cells (reviewed in [(2,3]). It is not knownhow mutations in either gene result in cyst formation or inloss of renal function.
Evidence has implicated reduced oxidant protection in the pathogenesisof PKD in several rodent models 4). Mice whose bcl-2 gene hasbeen knocked out develop PKD (5), which may be related to theloss of the antioxidant properties of bcl-2 (6). Treatmentsthat reduce renal antioxidant protection or alter renal redoxmetabolism can exacerbate PKD in the Han:SPRD-Cy rat model (7,8).Several antioxidant enzyme genes have also been reported asbeing underexpressed in cystic kidneys of the C57BL/6J-cpk mouse(9).
Antioxidant and detoxicant enzymes play important protectiveroles in the kidney (10). Primarily because of its transportfunctions, the kidney has a very active oxidative metabolismthat results in the production of reactive oxygen species (ROS).If left unchecked, ROS can damage all major cellular componentsand lead to a state of oxidative stress. The primary antioxidantenzymes responsible for protection from ROS are superoxide dismutase,catalase, and glutathione peroxidase (GPx). Detoxicant enzymessuch as glutathione S-transferase (GST) metabolize toxic electrophilesand are considered to be secondary antioxidant enzymes (11).Studies in which rats were fed an antioxidant-deficient, pro-oxidantdiet have demonstrated the significance of an adequate antioxidantdefense in both normal and injured kidneys (12,13). Oxidantinjury is now recognized as playing a key role in the pathophysiologicpathways of a wide variety of progressive clinical and experimentalrenal diseases (14).
Reduced antioxidant enzyme expression in cystic kidneys mightresult in compromised protection against oxidative injury thatcould lead or contribute to renal dysfunction and progressionto renal failure. To test this hypothesis, we assessed markersof oxidative stress and oxidant injury, and we examined themRNA and protein expression of extracellular GPx (EGPx) duringthe development of PKD in the cpk mouse and the Cy rat. Thesemodels, with different modes of transmission and rates of diseaseprogression, have been studied extensively as representativeof human ARPKD and ADPKD (15,16). The study presented here demonstratesthat cystic kidneys of cpk mice and Cy rats have reduced enzymaticprotection against oxidative attack and that they experienceoxidative damage. These results suggest that reduced renal antioxidantprotection and oxidant stress are pathogenetic mechanisms commonto all forms of PKD.
Animals
Colonies of C57BL/6J-cpk mice and Han:SPRD-Cy rats are maintainedat the University of Kansas Medical Center. Animal treatmentwas in accord with the National Institutes of Health Guide forthe Care and Use of Laboratory Animals.
Tissue Collection and Processing
Plasma from heparinized blood was used for EGPx enzyme assaysand Western blot analyses. After perfusion with phosphate-bufferedsaline, kidneys or liver were removed and homogenized in eitherguanidinium thiocyanate buffer (RNA isolation) (17), in 5 timesthe organ weight of 20 mM Tris, pH 7.5, 5 mM ethylenediaminetetraacetate(EDTA), and 0.1 mM phenylmethylsulfonyl fluoride (renal EGPxWestern blot test), or in 50 mM KPO4, pH 7.0, and 1 mM EDTA(liver GPx enzyme assay). Quantitation of protein was by theBio-Rad DC assay kit (Bio-Rad, Hercules, CA) after solubilizationof extracts by addition of sodium dodecyl sulfate to 2%.
Northern Blot Analyses
Total RNA was isolated by the method of Chomczynski and Sacchi(17). RNA samples (5 to 20 µg) were electrophoresed informaldehyde gels and blotted for hybridization with antisenseRNA probes as previously described (18). The mouse EGPx cDNAclone was isolated as described previously (18). GST-Ya, catalase,and MnSOD cDNA were produced by reverse transcriptase–PCRthat used mouse kidney RNA and were confirmed by sequencing.The mouse heme oxygenase-1 (HO-1) cDNA clone was obtained fromS. Sakiyama. As a quantitation control, all blots were rehybridizedwith an oligonucleotide probe for 18S rRNA as described previously(18).
Western Blot Analyses
EGPx immunoanalyses were performed by means of an anti-rat EGPxantibody as described previously (18). Equal amounts of totalkidney protein or equal volumes of blood plasma from individualanimals of a specific age were electrophoresed in duplicatesodium dodecyl sulfate–polyacrylamide gels. One gel wasstained with Coomassie blue to assess protein loading; the othergel was electroblotted, probed with the anti-EGPx antibody,and developed with the chemiluminescent substrate CDP-Star (Tropix,Bedford, MA). Each age group, representing a single litter ofrats or mice, was analyzed separately.
Immunohistochemical Analyses
Paraffin-embedded sections of paraformaldehyde-fixed kidneyswere used. After the paraffin was removed, the sections wereincubated with H2O2 to remove endogenous peroxidase activity,then blocked with 2% dry milk/phosphate-buffered saline or withTerminator Block (BioCare, Walnut Creek, CA; anti-Mn SOD only).Sections were incubated with diluted rabbit anti–HO-1(1:1000; Stressgen, Victoria BC, Canada), anti–GST- (1:1000;Alpha Diagnostics, San Antonio, TX), or anti-MnSOD (1:2000;Stressgen) serum in 3% normal goat serum overnight at 4°C.Sections were incubated with biotinylated secondary antibody(Histostain-SP kit; Zymed, San Francisco, CA) and streptavidin–horseradishperoxidase conjugate, then developed with aminoethyl carbazolefollowed by hematoxylin counterstaining. For controls, the primaryantibody was replaced with an appropriate dilution of normalrabbit serum (all antibodies) or was competed by preincubationwith recombinant HO-1 protein (anti–HO-1 only). Controlswere consistently negative.
GPx Enzymatic Assay
EGPx activity in plasma samples (10 to 15 µl) was measuredat 37°C by following the oxidation of nicotinamide adeninedinucleotide phosphate at 340 nm in an SLM Aminco 3000 Arrayspectrophotometer (Milton Roy) with either H2O2 or tert-butylperoxideas substrate as described by Avissar et al. (19). Cellular GPxactivity was measured in 100,000 x g liver supernatants by useof H2O2 as described by Lawrence and Burk (20). Assay reagentswere from Sigma Chemical Co. (St. Louis, MO).
Lipid Peroxidation Assay
Combined levels of malondialdehyde (MDA) and 4-hydroxy-2(E)-nonenal(HNE) were measured in plasma and kidney extracts from miceby means of a lipid peroxidation assay kit (Calbiochem, La Jolla,CA) according to the manufacturer’s instructions. Bloodwas collected; EDTA was used as the anticoagulant. Plasma wasprepared and used in the assay within 1 h of collection. Wholekidneys from animals perfused with phosphate-buffered salinewere homogenized in 20 mM Tris-HCl, pH 7.4, and supernatantsfrom a 3000 x g spin were used in the assay.
Statistical Analyses
For quantitation, autoradiographic films and stained proteingels were scanned and analyzed by NIH Image software (availableat http://rsb.info.nih.gov/nih-image/). Statistical analysesincluded unpaired t test with F test for equality of variances(enzymatic activity) and two-way (age x phenotype) ANOVA withLevene’s test of error variances (protein measurements).Spearman’s rank correlation was used to study the relationshipbetween renal and plasma EGPx. P < 0.05 was considered significant.
Previous observations suggested that cystic kidneys have analtered redox metabolism and may be in a state of chronic oxidativestress (8,21). To assess oxidative stress, the expression ofHO-1, a widely accepted marker of oxidative stress (22), wasexamined in normal and cystic kidneys of both the C57BL/6J-cpkmouse and the Han:SPRD-Cy rat. Cystic disease in these animalmodels is most likely the result of mutations in different genes(23,24), and disease severity is determined by gene dosage.Homozygous cpk/cpk mice develop cystic kidneys rapidly afterbirth and die of azotemia by 3 to 4 wk of age (15), but heterozygouscpk/+ mice experience no renal abnormalities (25). Homozygouscystic (Cy/Cy) rats also develop a rapidly progressive formof PKD, resulting in death at approximately 3 wk of age; heterozygouscystic (Cy/+) rats develop a more slowly progressive PKD thatis sexually dimorphic (16). Cystic disease in male Cy/+ ratsis more severe, characterized by larger kidneys, earlier developmentof azotemia (approximately 8 wk of age), and death at approximately1 yr of age, in contrast to Cy/+ females, which do not die asa result of renal failure until 2 yr of age (26).
Northern blot analysis showed that HO-1 mRNA levels are increasedin cystic cpk kidney at 14 and 21 d of age, and possibly asearly as 9 d (Figure 1, A and B). HO-1 mRNA levels are alsodramatically increased in Cy/Cy rat kidney at 2 and 3 wk ofage and in male Cy/+ kidney at 8 wk of age (Figure 1, C and D).In contrast, HO-1 mRNA levels are only slightly increasedin female Cy/+ kidney at 8 wk of age. Thus, HO-1 induction indicatesthat oxidant stress occurs in cystic kidneys and serves as ameasure of disease severity in these two models.
Figure 1. Heme oxygenase-1 (HO-1) mRNA levels in normal and cystic kidneys from cpk mice and Cy rats. (A) Northern blot of RNA from kidneys of 9-, 14-, and 21-d-old normal (N) and cystic (C) mice hybridized with a probe for HO-1 mRNA. Normal kidney samples are from both +/+ and cpk/+ mice; cystic kidney samples are from cpk/cpk mice only. Each RNA sample represents 6 to 10 mice. The blot was stripped and rehybridized with an 18S rRNA oligo probe. (B) Renal HO-1 mRNA levels in cystic relative to normal mouse kidneys at each age after normalization to 18S controls. Bars represent the average of two separate Northern blot analyses with different pools of RNA. (C) Northern blot of RNA from kidneys of 2-, 3-, and 8-wk-old normal (+/+) and cystic (Cy/+ and Cy/Cy) male (top) and female (bottom) rats hybridized with a probe for HO-1 mRNA. Each RNA sample represents a single rat. The blot was rehybridized with an 18S oligo probe. (D) Renal HO-1 mRNA levels in cystic relative to normal rat kidneys at each age after normalization to 18S controls. HO-1 levels in male and female +/+ kidneys were set at 1.0 for each age group. Bars represent the average of two Northern blot analyses with different sets of RNA.
A state of oxidative stress can develop when there is an excessivegeneration of ROS or an inadequate antioxidant defense. Previouswork suggested that the expression of numerous protective enzymegenes was reduced in cystic kidneys of cpk mice (9). Therefore,mRNA levels of the secreted GPx (EGPx), the mitochondrial Mnsuperoxide dismutase (MnSOD), the peroxisomal catalase, andthe cytosolic -class GST (GST-Ya) were examined in normal andcystic kidney of cpk mice (Figure 2). Expression of each ofthe mRNAs is significantly reduced in cystic kidney by 14 dof age and is further reduced by 21 d (Figure 2E). To determinewhether the reduced expression of these genes is kidney specific,GST-Ya, MnSOD, and catalase mRNA were examined in liver from3-wk-old normal and cystic mice. No differences in hepatic levelsof these mRNA were detected (data not shown).
Figure 2. Expression of antioxidant and detoxicant enzyme mRNA in cpk mouse kidneys. Northern blot analyses of RNA from kidneys of 7-, 14-, and 21-d-old normal (N) and cystic (C) cpk mice hybridized with probes for (A) extracellular glutathione peroxidase (GPx), (B) Mn superoxide dismutase (SOD), (C) cytosolic -class glutathione-S-transferase (GST), or (D) catalase (Cat) mRNA, then rehybridized with an 18S rRNA oligo probe. Each RNA sample represents 6 to 10 mice. For each mRNA, blots were performed two to five times, with comparable results. The graph below each blot shows relative level of mRNA by age in normal and cystic kidneys after normalization to 18S rRNA. Enzyme mRNA levels increase with age in normal kidneys, but not in cystic kidneys. (E) Level of each mRNA (after normalization to 18S rRNA) in cystic kidney relative to normal kidney at each age in days (da). Both the 4- and 1.5-kb bands of MnSOD mRNA were used in quantitation.
To determine whether reduced expression of these protectiveenzymes is a feature common to both rapidly and slowly progressivePKD, their expression was examined in Cy rat kidney (Figure 3).Cystic kidneys from male and female Cy/Cy rats have dramaticallyreduced levels of EGPx, MnSOD, catalase, and GST-Ya mRNA at3 wk of age. EGPx and GST-Ya mRNA are also decreased at 2 wk(data not shown; MnSOD and catalase were not examined). In thekidneys of Cy/+ rats, reduced expression of all 4 mRNA is detectableas early as 3 wk in males. At 8 wk, all 4 antioxidant mRNA arereduced in kidney from male and female Cy/+ rats; however, thereduction is more evident in males. Thus, in the early stagesof cystic disease in the rat, these protective enzymes are underexpressed,and their degree of underexpression appears to correlate withthe severity of PKD. Furthermore, these results demonstratethat reduced antioxidant enzyme expression is not simply a reflectionof the tubule segment affected because cysts develop in differentsegments in these models—that is, the collecting ductsof the cpk/cpk mouse (15), the proximal tubules of the Cy/+rat, and all tubule segments of the Cy/Cy rat (26).
Figure 3. Expression of antioxidant and detoxicant enzyme mRNA in Cy rat kidneys. Northern blot analyses of RNA from kidneys of 3- and 8-wk-old normal (+/+), heterozygous cystic (Cy/+), and homozygous cystic (Cy/Cy) male (M) and female (F) rats hybridized with probes for (A) extracellular glutathione peroxidase (GPx), (B) cytosolic -class glutathione-S-transferase (GST), (C) Mn superoxide dismutase (SOD), or (D) catalase (Cat) mRNA, then rehybridized with an 18S rRNA oligo probe. (E) Renal antioxidant and detoxicant enzyme mRNA levels in cystic kidneys relative to normal rat kidneys at various ages. After correction to 18S hybridization values, mRNA levels in male and female +/+ kidneys were set at 1.0 for each age group and the level of each mRNA in male and female, Cy/+ and Cy/Cy kidneys was determined relative to +/+ kidneys. Bars represent the average of two separate Northern blot analyses that used different sets of RNA from single rats (except for catalase, which was examined twice with the same set of RNA samples).
To gain insight into where the changes in HO-1 and antioxidantenzyme gene expression are occurring, cystic kidneys from 2-wk-oldcpk mice and from 8-wk-old Cy/+ rats were immunohistochemicallystained for HO-1, GST, or MnSOD proteins (Figure 4). In cpkkidney, HO-1 is expressed in some smaller cortical cysts (mostlikely proximal tubular cysts) and in many nondilated corticaltubules, but is largely absent from the large collecting ductcysts. A similar staining pattern is observed for GST (Figure 4B)and MnSOD (data not shown) in the cpk kidney. Within Cy/+rat kidneys, HO-1 protein is found in the majority of nondilatedcortical tubules and in cysts, but it is reduced or absent incystic epithelial cells overlying a thickened basement membrane(Figure 4C). GST (data not shown) and MnSOD (Figure 4D) stainingis similar to that for HO-1. This immunoanalysis demonstratesthat in cystic kidney of both the cpk mouse and the Cy rat,changes in HO-1 and antioxidant protein expression primarilyinvolve the cystic epithelium with a loss or reduction of stainingwithin many cysts. In addition, a dramatic upregulation of HO-1is observed in normal-appearing tubules.
Figure 4. Localization of heme oxygenase-1 (HO-1), glutathione S-transferase (GST), and Mn superoxide dismutase (MnSOD) protein in cystic kidneys. Immunostaining for HO-1 (A and C), GST (B), and MnSOD (D) proteins in cystic kidneys of 2-wk-old cpk mice (A and B) and 8-wk-old Cy/+ rats (C and D). (A) HO-1 staining is present in some cysts and in many nondilated tubules within the cortex of cpk kidneys. Large collecting duct cysts are negative for HO-1. Normal kidneys showed only occasional HO-1 tubular staining, primarily of macula densa (data not shown). Asterisk indicates cystic tubule with partial staining for HO-1. (B) Nondilated cortical tubules primarily stain for GST. Asterisk indicates cystic tubule with only a few cells expressing GST. (C) HO-1 staining is present in some cysts and completely absent in other cysts within Cy/+ kidneys. Intense staining of HO-1 in nondilated tubules predominates. Arrows indicate area of cyst wall where HO-1 expression is absent in epithelial cells with thickened basement membrane below. (D) MnSOD staining is found in almost all cortical tubules but is absent or reduced in cystic epithelial cells lying on thickened basement membrane (arrows). Original magnification, x100.
To determine whether decreased antioxidant mRNA levels are accompaniedby reductions in protein, we selected EGPx, a secreted enzymethat catalyzes the reduction of hydrogen peroxide, alkyl hydroperoxides,and phospholipid hydroperoxides (reviewed in [27]). EGPx activityis present in many extracellular fluids, including plasma. Studiesin anephric humans and rats have demonstrated that at least70% of the EGPx plasma enzymatic activity is produced by thekidney (28). As such, plasma levels of EGPx should provide anadditional indication of the renal production of this enzyme.The levels of EGPx protein were examined by Western blot analysisin kidney and plasma of normal and cystic mice and rats at variousages. In mice, an overall significant effect of phenotype isobserved for renal (P < 0.0001) and for plasma (P = 0.0029)EGPx protein levels (Figure 5). Plasma EGPx levels reflect renallevels, as shown by a significant positive correlation (Figure 5C;rs = 0.728, P = 0.0036), consistent with a reduced renalproduction in the cystic mice. Renal and plasma EGPx proteinlevels were reduced in cystic rats at 6, 8, and 10 wk of age(Figure 6), and there was an overall significant effect of phenotypefor both (P < 0.01).
Figure 5. Extracellular glutathione peroxidase (EGPx) protein levels in normal and cystic cpk mice. After quantitation of anti-EGPx Western blots, densitometric values were converted to Z scores for each age group to assess the phenotype and age-by-phenotype effects in normal and cystic mice. Bars represent the average Z scores ± SEM for EGPx protein in kidneys from 9-, 11-, 13-, 15-, 17-, 19-, and 21-d-old mice (A) or in plasma from 17-, 19-, and 21-d-old mice (B). Each bar represents two to four mice. (C) Significant positive correlation (rs = 0.728, P = 0.0036) between renal and plasma EGPx levels.
Figure 6. Extracellular glutathione peroxidase (EGPx) protein levels in normal and cystic Cy rats. After quantitation of anti-EGPx Western blots, densitometric values were converted to Z scores for each age group to assess the phenotype and age-by-phenotype effects in normal (+/+) and cystic (Cy/+) rats. Bars represent the average Z scores ± SEM for EGPx protein in kidneys from 4-, 6-, 8-, and 10-wk-old rats (A) and in plasma from 6-, 8-, and 10-wk-old rats (B). Each bar represents four to six rats. There was a positive but NS correlation (rs = 0.233, P = 0.199) between renal and plasma EGPx in the rats (data not shown).
To determine whether the reduced levels of EGPx protein resultin decreased enzymatic activity, EGPx enzyme assays were performedwith plasma from 20-d-old mice (Table 1). Plasma EGPx activityis significantly reduced in cystic mice. GPx family membersare selenium-dependent enzymes, and selenium bioavailabilityhas been shown to affect GPx mRNA, protein, and activity levels(reviewed in [(27]). To rule out defective selenium metabolismas a mechanism for reduced EGPx production, we examined hepaticenzymatic activity and renal and hepatic mRNA levels of theubiquitous, cellular form of GPx (cGPx), a sensitive biomarkerof selenium status. Although cGPx mRNA was reduced in cystickidneys, cGPx mRNA and enzymatic activity levels were not affectedin livers of 19- to 21-d-old cystic mice (data not shown).
Table 1. Plasma extracellular glutathione peroxidase activity in 20-d-old mice
Reduced antioxidant enzyme mRNA, protein and activity levels,and increased HO-1 mRNA levels suggest that there is increasedROS and resulting oxidative damage in cystic kidneys. Membranelipids are a frequent target of ROS and can be oxidized to formlipid hydroperoxides. If not reduced by enzymes (such as GPx),lipid hydroperoxides can undergo further reactions to producevarious aldehydes, including MDA and HNE (29). The combinedMDA + HNE levels were measured in cpk mice (Figure 7) and werefound to be greatly elevated in cystic kidney tissue from miceat 17 and 19 d of age and in plasma from cystic mice at 19 dof age. These results demonstrate that oxidative damage is occurringin cystic mice.
Figure 7. Lipid peroxidation levels in kidney and plasma from normal and cystic cpk mice. (A) Levels of malonaldehyde (MDA) + 4-hydroxy-2(E)-nonenal (HNE) in normal and cystic kidneys from 13-, 17-, and 19-d-old mice. (Left) Total amount of MDA + HNE (nmol/kidney) per whole kidney. (Right) MDA + HNE (nmol/mg protein) per milligram of renal protein (protein was not determined for the 19-d kidney samples). (B) Levels of MDA + HNE (nmol/ml) in plasma from normal and cystic, 13-, 17-, and 19-d-old mice. Each bar represents a pooled sample from two to four mice, except for 17- and 19-d cystic kidney, which are from single mice.
The study presented here establishes that there is reduced antioxidantenzyme protection and increased oxidative stress in two differentmodels of cystic disease and thereby solidifies the view thatoxidant or xenobiotic stress and tissue injury are pathogeneticfactors in the progression of PKD (4,30). Early observationsof the development of cystic disease in long-term dialysis patients(1) and in deconditioned germfree rats treated with nordihydroguaiareticacid (31) led to the idea that environmental insults can resultin renal cyst formation. Later observations in numerous animalmodels suggested that oxidant mechanisms may be at work in cysticdisease (4). Cystic kidneys from Cy/+ rats and bcl-2 knockoutmice have elevated markers of oxidative stress (7,21,32). Treatmentsthat increase renal production of oxidants (33,34) exacerbatePKD in the Cy rat (8,35).
Other studies have made connections between reduced antioxidantprotection and the development (36) or progression (7) of PKDin the rat. Antioxidant promoting treatments (37) have beenshown to attenuate cystic disease in mouse (38) and rat (39,40)models. Finally, connections between human PKD and reduced protectionor oxidant stress exist. ADPKD patients are reported to havereduced plasma EGPx protein and enzyme activity (41) and increasedplasma lipid peroxidation products (42). Cultured ADPKD renalepithelial cells are more sensitive to oxidant treatment thannormal renal epithelial cells (43). In light of the data presentedin this report, it is likely that this increased oxidant sensitivityis the result of decreased antioxidant enzyme expression inthe ADPKD cells.
The study presented here provides direct evidence for oxidativestress, as shown by induction of the HO-1 gene in the cystickidneys of two different animal models of ARPKD and ADPKD andby elevated lipid peroxidation levels in cpk mice. Levels ofHO-1 induction correlate with gene dosage and severity of cysticdisease in that HO-1 induction is higher in homozygous cysticrats and mice with rapidly progressive disease and lower inheterozygous cystic rats with more slowly developing disease.Female Cy/+ rats that develop relatively mild PKD have the lowestHO-1 induction levels. The fact that HO-1 induction is evidentat early stages of PKD suggests that oxidative stress playsa role relatively early in disease progression. As such, theresults presented here, together with the aforementioned studies,strongly suggest that oxidative stress may be a general pathogeneticfeature in all cystic diseases.
How might oxidative stress be involved in the pathogenesis ofPKD? Renal oxidative stress due to decreased oxidant protectionhas been shown to result in increased cell proliferation, increasedextracellular matrix synthesis, increased inflammatory cellinfiltration, and increased apoptosis (44), pathogenic featuresall commonly attributed to cystic kidneys (reviewed in [1,3]).As such, one could envision that the level of oxidative stressand resulting macromolecular damage could determine the rateof progression of PKD and resultant loss of renal function.In extrapolating our observations to human ADPKD, where cystogenesisis proposed to result from a second somatic mutation in oneof the PKD genes (reviewed in [2,45]), we speculate that renaloxidative stress could result in DNA damage, which might increasethe rate of second hits and thereby the severity of PKD.
We have investigated the possibility that reduced renal antioxidantenzyme protection may contribute to oxidative stress in cystickidneys. Renal levels of EGPx, MnSOD, catalase, and GST-Ya mRNAare reduced in cystic mice and rats. The levels of antioxidantenzyme mRNA reduction appeared to reflect the severity of cysticdisease and to coincide with the induction of HO-1 mRNA. EGPxis also reduced at the protein level in the plasma and kidneysof both cystic mice and rats, and at the enzymatic activitylevel in plasma from cystic mice. Primarily thought of as aplasma enzyme, whose major site of production and secretionare renal proximal tubules, recent studies have demonstrateda protective function for EGPx within the kidney. Treatmentsthat reduced renal EGPx protein production were shown to resultin oxidative damage in the kidney (46). Renal EGPx overproductionin transgenic mice protected their kidneys from damage in ischemia/reperfusionexperiments as evidenced by decreased blood urea nitrogen levels,decreased tubular necrosis and apoptosis, and decreased lipidperoxidation (47). In the study presented here, elevated levelsof MDA + HNE are observed in the kidney and plasma of cysticmice. Because lipid hydroperoxides are substrates for EGPx,and are precursors of MDA and HNE, it is quite possible thatthere is a link between reduced EGPx and the increased lipidperoxidation observed in cystic mice.
The results in this report imply that lack of appropriate expressionof protective enzyme genes and oxidative stress are generalpathologic mechanisms of cystic disease progression. Becausereduced antioxidant enzyme expression and increased HO-1 expressionwere observed in two different animal species with cystic diseasecaused by mutations in different genes and affecting differenttubule segments, it appears that these changes are general consequencesof renal cyst formation, possibly as the result of widespreadchanges in renal tubular metabolism and function activated bythe cystogenic process. This idea is supported by the observedloss of antioxidant enzyme expression in cystic tubules andthe observed induction of HO-1 protein in cystic and nondilatedtubules of polycystic kidneys.
What could be responsible for the reduction of antioxidant protectionin cystic kidneys? One important feature common to all formsof PKD is that cystic epithelial cells are abnormally or incompletelydifferentiated (30). A number of antioxidant enzymes are developmentallyexpressed in the rodent kidney, first appearing in fetal kidneytubules that are somewhat differentiated and then increasingin expression during postnatal development (18,48,49). The differentialstaining for antioxidant enzymes within cysts, which was particularlyevident in the previously described, immature cystic epithelialcells overlying thickened basement membrane in the rat kidneys(50), suggests that cellular immaturity may result in loss ofantioxidant enzyme expression. It is possible that the primarygene defect in inherited PKD results in abnormally differentiatedrenal epithelial cells that are incapable of producing or maintainingthe normal complement of protective enzymes. Such a situationcould then lead to renal oxidative stress as the cystic kidneyattempts to meet its metabolic demands, resulting in a viciouscycle of oxidative damage, renal injury, and injury-induceddedifferentiation, eventually leading to apoptosis, loss ofrenal function, and kidney failure.
Acknowledgments
This work was supported by the Polycystic Kidney Disease Foundation(grant 99008 to RLM) and the National Institutes of Health (grantDK57301 to RLM and JPC). We thank Benjamin D. Cowley Jr. forproviding Cy rat Northern blot analyses for initial analysis,S. Sakiyama for the mouse HO-1 cDNA clone, S. Yoshimura forthe chicken anti-rat EGPx antibody, Rosetta Barkley for experthistology work, Jared J. Grantham for critical discussions,and John M. Belmont for statistical analysis and editorial assistance.
Wu G, Somlo S: Molecular genetics and mechanism of autosomal dominant polycystic kidney disease. Mol Genet Metab 69: 1–15, 2000[CrossRef][Medline]
Murcia N, Sweeney WJ, Avner E: New insights into the molecular pathophysiology of polycystic kidney disease. Kidney Int 55: 1187–1197, 1999[CrossRef][Medline]
Torres VE: New insights into polycystic kidney disease and its treatment. Curr Opin Nephrol Hypertens 7: 159–169, 1998[Medline]
Hockenbery DM, Oltvai ZN, Yin XM, Milliman CL, Korsmeyer SJ: Bcl-2 functions in an antioxidant pathway to prevent apoptosis. Cell 75: 241–251, 1993[CrossRef][Medline]
Torres VE, Bengal R, Litwiller R, Wilson D: Aggravation of polycystic kidney disease in Han: PRD rats by buthionine sulfoximine. J Am Soc Nephrol 8: 1283–1291, 1997[Abstract]
Torres VE, Mujwid DK, Wilson DM, Holley KH: Renal cystic disease and ammoniagenesis in Han: SPRD rats. J Am Soc Nephrol 5: 1193–1200, 1994[Abstract]
Maser RL, Calvet JP: Reduced expression of antioxidant and detoxicant enzyme genes in polycystic kidney disease (PKD) [Abstract]. J Am Soc Nephrol 7: 1617, 1996
Ichikawa I, Kiyama S, Yoshioka T: Renal antioxidant enzymes: Their regulation and function. Kidney Int 45: 1–9, 1994[Medline]
Hayes J, Strange R: Potential contribution of the glutathione S-transferase supergene family to resistance to oxidative stress. Free Radic Res 22: 193–207, 1995[Medline]
Nath KA, Salahudeen AK: Induction of renal growth and injury in the intact rat kidney by dietary deficiency of antioxidants. J Clin Invest 86: 1179–1192, 1990
Nath KA, Paller MS: Dietary deficiency of antioxidants exacerbates ischemic injury in the rat kidney. Kidney Int 38: 1109–1117, 1990[Medline]
Haugen E, Nath K: The involvement of oxidative stress in the progression of renal injury. Blood Purif 17: 58–65, 1999[CrossRef][Medline]
Cowley BDJr, Gudapaty S, Kraybill AL, Barash BD, Harding MA, Calvet JP, Gattone VHII: Autosomal-dominant polycystic kidney disease in the rat. Kidney Int 43: 522–534, 1993[Medline]
Chomczynski P, Sacchi N: Single step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162: 156–159, 1987[Medline]
Maser RL, Magenheimer BS, Calvet JP: Mouse plasma glutathione peroxidase: cDNA sequence analysis and renal proximal tubular expression and secretion. J Biol Chem 269: 27066–27073, 1994[Abstract/Free Full Text]
Avissar N, Slemmon J, Palmer I, Cohen H: Partial sequence of human plasma glutathione peroxidase and immunologic identification of milk glutathione peroxidase as the plasma enzyme. J Nutr 121: 1243–1249, 1991
Lawrence R, Burk R: Glutathione peroxidase activity in selenium-deficient rat liver. Biochem Biophys Res Commun 71: 952–958, 1976[CrossRef][Medline]
Torres VE, Bengal RJ, Nickander KK, Grande JP, Low PA: Renal concentration of -tocopherol: Dependence on gender and lack of effect on polycystic kidney disease in Han:SPRD rats. Am J Kidney Dis 31: 687–693, 1998[Medline]
Applegate LA, Luscher P, Tyrrell RM: Induction of heme oxygenase: A general response to oxidant stress in cultured mammalian cells. Cancer Res 51: 974–978, 1991[Abstract/Free Full Text]
Nagao S, Ushijima T, Kasahara M, Yamaguchi T, Kusaka M, Matsuda J, Nagao M, Takahashi H: Closely linked polymorphic markers for determining the autosomal dominant allele (Cy) in rat polycystic kidney disease. Biochem Genet 37: 227–235, 1999[CrossRef][Medline]
Simon E, Cook S, Davisson M, D’Eustachio P, Guay-Woodford L: The mouse congenital polycystic kidney (cpk) locus maps within 1.3 cM of the chromosome 12 marker D12Nyu2. Genomics 21: 415–418, 1994[CrossRef][Medline]
Gattone V, MacNaughton K, Kraybill A: Murine autosomal recessive polycystic kidney disease with multiorgan involvement induced by the cpk gene. Anat Rec 245: 488–499, 1996[CrossRef][Medline]
Schafer K, Gretz N, Bader M, Oberbaumer I, Eckardt K, Kriz W, Bachmann S: Characterization of the Han: SPRD rat model for hereditary polycystic kidney disease. Kidney Int 46: 134–152, 1994[Medline]
Brigelius-Flohe R: Tissue-specific functions of individual glutathione peroxidase. Free Radic Biol Med 27: 951–965, 1999[CrossRef][Medline]
Avissar N, Ornt DB, Yagil Y, Horowitz S, Watkins RH, Kerl EA, Takahashi K, Palmer IS, Cohen HJ: Human kidney proximal tubules are the main source of plasma glutathione peroxidase. Am J Physiol 266: C367–C375, 1994[Abstract/Free Full Text]
Esterbauer H, Schaur R, Zollner H: Chemistry and biochemistry of 4-hydroxynonenal, malonaldehyde and related aldehydes. Free Radic Biol Med 11: 81–128, 1991[CrossRef][Medline]
Calvet JP: Injury and development in polycystic kidney disease. Curr Opin Nephrol Hypertens 3: 340–348, 1994[CrossRef][Medline]
Gardner KD, Evan AP, Reed WP: Accelerated renal cyst development in deconditioned germfree rats. Kidney Int 29: 1116–1123, 1986[Medline]
Hochman A, Liang H, Offen D, Melamed E, Sternin H: Developmental changes in antioxidant enzymes and oxidative damage in kidneys, liver and brain of bcl-2 knockout mice. Cell Mol Biol 46: 41–52, 2000
Nath KA, Croatt AJ, Hostetter TH: Oxygen comsumption and oxidant stress in surviving nephrons. Am J Physiol 258: F1354–F1362, 1990[Abstract/Free Full Text]
Torres VE, Bengal RJ, Chini C, Dousa TP: Effect of extracellular acidification on generation of oxygen free radicals by LLC-PK1 cells [Abstract]. J Am Soc Nephrol 8: 711, 1997
Cowley B, Grantham J, Muessel M, Kraybill A, Gattone V: Modification of disease progression in rats with inherited polycystic kidney disease. Am J Kidney Dis 27: 865–879, 1996[Medline]
Hjelle JT, Guenthner TM, Bell K, Whalen R, Flouret G, Carone FA: Inhibition of catalase and epoxide hydrolase by the renal cystogen 2-amino-4,5-diphenylthiazole and its metabolites. Toxicology 60: 211–222, 1990[CrossRef][Medline]
de Cavanagh E, Inserra F, Ferder L, Fraga C: Enalapril and captopril enhance glutathione-dependent antioxidant defenses in mouse tissues. Am J Physiol Regul Integr Comp Physiol 278: R572–R577, 2000[Abstract/Free Full Text]
Nagao S, Yamaguchi T, Kasahara M, Kusaka M, Matsuda J, Ogiso N, Takahashi H, Grantham JJ: Effect of probucol in a murine model of slowly progressive polycystic kidney disease. Am J Kidney Dis 35: 221–226, 2000[Medline]
Torres VE, Berndt TJ, Okamura M, Nesbit JW, Holley KE, Carone FA, Knox FG, Romero JC: Mechanisms affecting the development of renal cystic disease induced by diphenylthiazole. Kidney Int 33: 1130–1139, 1988[Medline]
Keith DS, Torres VE, Johnson CM, Holley KE: Effect of sodium chloride, enalapril, and losartan on the development of polycystic kidney disease in Han: SPRD rats. Am J Kidney Dis 24: 491–498, 1994[Medline]
Yamamoto Y, Takekoshi Y, Itami N, Honjo T, Kojima H, Yano S, Takahashi H, Saito I, Takahashi K: Enzyme-linked immunosorbent assay for extracellular glutathione peroxidase in serum of normal individuals and patients with renal failure on hemodialysis. Clin Chim Acta 236: 93–99, 1995[CrossRef][Medline]
Merta M, Stipek S, Crkovska J, Platenik J: Lipid peroxidation and activity of superoxide dismutase in autosomal dominant polycystic kidney disease. Contrib Nephrol 115: 109–112, 1995[Medline]
Jonassen J, Cooney R, Scheid C: Sensitivity of autosomal dominant polycystic kidney disease (ADPKD) renal cells to oxidant stress [Abstract]. J Am Soc Nephrol 8: 1728A, 1998
Nath K, Grande J, Croatt A, Haugen J, Kim Y, Rosenberg ME: Redox regulation of renal DNA synthesis, transforming growth factor-beta 1 and collagen gene expression. Kidney Int 53: 367–381, 1998[CrossRef][Medline]
Qian F, Watnick T: Somatic mutation as mechanism for cyst formation in autosomal dominant polycystic kidney disease. Mol Genet Metab 68: 237–242, 1999[CrossRef][Medline]
Dobashi K, Asayama K, Nakane T, Hayashibe H, Kodera K, Uchida N, Nakazawa S: Effect of peroxisome proliferator on extracellular glutathione peroxidase in rat. Free Radic Res 31: 181–190, 1999[Medline]
Ishibashi N, Weisbrot-Lefkowitz M, Reuhl K, Inouye M, Mirochnitchenko O: Modulation of chemokine expression during ischemia/reperfusion in transgenic mice overproducing human glutathione peroxidases. J Immunol 163: 5666–5677, 1999[Abstract/Free Full Text]
Kingsley PD, Whitin JC, Cohen HJ, Palis J: Developmental expression of extracellular glutathione peroxidase suggests antioxidant role in deciduum, visceral yolk sac, and skin. Mol Reprod Dev 49: 343–355, 1998[CrossRef][Medline]
Oberley TD, Sempf JM, Oberley LW: Immunohistochemical localization of antioxidant enzymes during hamster kidney development. Histochemistry J 27: 575–586, 1995
Bachmann S, Ramasubbu K, Schafer K, Uiker S, Gretz N: Tubulointerstitial changes in the Han: SPRD rat model for ADPKD. Contrib Nephrol 115: 113–117, 1995[Medline]
Received for publication March 16, 2001.
Accepted for publication November 1, 2001.
This article has been cited by other articles:
A. E. Courtney, P. T. McNamee, S. Heggarty, D. Middleton, and A. P. Maxwell Association of functional haem oxygenase-1 gene promoter polymorphism with polycystic kidney disease and IgA nephropathy
Nephrol. Dial. Transplant.,
February 1, 2008;
23(2):
608 - 611.
[Abstract][Full Text][PDF]
S. Choudhary, T. Xiao, L. A. Vergara, S. Srivastava, D. Nees, J. Piatigorsky, and N. H. Ansari Role of Aldehyde Dehydrogenase Isozymes in the Defense of Rat Lens and Human Lens Epithelial Cells against Oxidative Stress
Invest. Ophthalmol. Vis. Sci.,
January 1, 2005;
46(1):
259 - 267.
[Abstract][Full Text][PDF]
Y. Sato, K. Harada, K. Kizawa, T. Sanzen, S. Furubo, M. Yasoshima, S. Ozaki, M. Ishibashi, and Y. Nakanuma Activation of the MEK5/ERK5 Cascade Is Responsible for Biliary Dysgenesis in a Rat Model of Caroli's Disease
Am. J. Pathol.,
January 1, 2005;
166(1):
49 - 60.
[Abstract][Full Text][PDF]
L. M. Guay-Woodford Murine models of polycystic kidney disease: molecular and therapeutic insights
Am J Physiol Renal Physiol,
December 1, 2003;
285(6):
F1034 - F1049.
[Abstract][Full Text][PDF]
Top
Abstract
Introduction
(1:1000; Alpha Diagnostics, San Antonio, TX), or anti-MnSOD (1:2000; Stressgen) serum in 3% normal goat serum overnight at 4°C. Sections were incubated with biotinylated secondary antibody (Histostain-SP kit; Zymed, San Francisco, CA) and streptavidin–horseradish peroxidase conjugate, then developed with aminoethyl carbazole followed by hematoxylin counterstaining. For controls, the primary antibody was replaced with an appropriate dilution of normal rabbit serum (all antibodies) or was competed by preincubation with recombinant HO-1 protein (anti–HO-1 only). Controls were consistently negative. 
