Background

The degree of oxidative stress and the severity of subsequent tissue injury may depend on an imbalance between the excessive production of reactive oxygen species and the antioxidant defense. The antioxidants include the enzymes superoxide dismutase (SOD), catalase, and glutathione peroxidase (GPX), which detoxify reactive oxygen species. Catalase (E.C.1.11.1.6) is a major enzyme that catalyzes the decomposition of hydrogen peroxide (H2O2) and plays a role in cellular antioxidant defense mechanisms [1]. The main reaction of catalase is the catalytic reaction (2H2O2 → O2+ 2H2O), which is essential for the removal of excessive H2O2 and for regulating the H2O2 concentration [2]. Catalase limits the accumulation of H2O2 generated by various oxidases in tissue, and serves as a substrate for the Fenton reaction to produce the highly injurious hydroxyl radicals. Genetic defects of catalase were first documented by Takahara [3] in Japanese patients who exhibited a deficiency of blood catalase enzyme activity (acatalasemia) [4, 5]. Subsequently, an acatalasemic mouse strain (Csb) was established by Feinstein, Suter, and Jaroslow [6] from the progeny of x-ray-irradiated mice.

Focal segmental glomerulosclerosis (FSGS) is a common cause of nephrotic syndrome in both children and adults [7, 8]. The clinicopathological syndrome may be classified as primary, secondary or familial. The primary defect in FSGS lies in the filtration barrier of the glomeruli. Disruption of the filtration barrier results in the loss of permselectivity, and macromolecules such as albumin are allowed to enter the urine. Chen et al. [9] reported that BALB/c mice were susceptible to renal toxicity arising from the administration of the anthracycline antibiotic, adriamycin (ADR), with selective injury to podocytes resulting in severe proteinuria and progressive renal failure [9, 10]. This was described as the first experimental model of FSGS in mice. The activities of antioxidant enzymes including catalase, GPX and Mn-SOD and the glutathione concentration in renal cortex were decreased by ADR nephropathy in BALB/c mice [11]. The level of nitric oxide in kidney homogenates [12], and the urinary levels of nitrite/nitrate [10] were also increased in the ADR nephropathy model. The administration of the soluble receptor for advanced glycation endproducts (AGEs) suppressed AGE generation and reactive oxygen species in the ADR nephropathy mice [13].

In the present study, we hypothesized that a defect in the antioxidant system in the form of catalase deficiency would enhance proteinuria, glomerular sclerosis, and eventually lead to the loss of renal function. This hypothesis was tested using the ADR nephropathy model, which is a well-established model of progressive FSGS, in acatalasemic mice.

Methods

Experimental animal protocol

Male wild-type mice (C3H/AnLCsaCsa) and male homozygous acatalasemic mutant mice (C3H/AnLCsbCsb) were used at the age of 8 - 10 weeks. Animals were housed in cages and fed standard chow and water ad libitum. ADR (10, 15 and 20 mg/kg BW) dissolved in saline was intravenously administered to both acatalasemic mutant mice and control wild-type mice [14]. In the C3H/AnL strain, the dose of 10 mg/kg BW did not induce significant albuminuria or increase mortality (Additional file 1: Figure S1A). The 20 mg/kg BW dose resulted in a rapid increase in mortality in both groups of mice (Additional file 1: Figure S1C). The ADR-induced cardiotoxicity might have led to the high mortality resulting from the higher dose of ADR [15]. A total of 55.3% of the acatalasemic mice and 59.2% of the wild-type mice survived 8 weeks after the injection of 15 mg/kg BW ADR (Additional file 1: Figure S1B) with a substantial amount of albuminuria. Therefore, 15 mg/kg BW ADR was considered to be a suitable concentration to investigate the effects of a functional catalase deficiency in the ADR nephropathy model. In the control mice, the same volume of saline was injected intravenously.

Mice were divided into subgroups (n = 6-15/group). Their body weight was measured at 0, 4 and 8 weeks. The mice were sacrificed at 0, 4 and 8 weeks after ADR administration, then their kidneys and hearts were harvested, washed with saline, blotted dry on gauze, and weighed as described previously [16, 17]. The whole kidney weight and heart weight were expressed as a percentage of the body weight determined at the time the mice were sacrificed. Twenty-four hour urine samples were collected in metabolic cages every 4 weeks. Immediately before death, blood samples were drawn. The serum creatinine, blood urea nitrogen (BUN), and urinary albumin excretion (UAE) levels were measured, and the creatinine clearance (Ccr) was calculated as described previously [16]. The experimental protocol was approved by the Ethics Review Committees for Animal Experimentation of Okayama University Graduate School (OKU-2009226).

Reagents and antibodies

Chemicals and reagents of analytical grade were purchased from Sigma Co. Ltd. (St. Louis, Missouri) or Wako Pure Chemical Ind. (Osaka, Japan) unless stated otherwise. The mouse monoclonal antibody to 4-hydroxy-2-nonenal (4-HNE) was obtained from Nof Life Science (Tokyo, Japan). The N-histofine MOUSESTAIN KIT was obtained from NICHIREI BIOSCIENCES INC. (Tokyo, Japan). The catalase assay kit was obtained from Cayman (Ann Arbor, Michigan).

Light and electron microscopic studies

The kidneys were removed, fixed in 10% buffered formalin, and embedded in paraffin. Paraffin sections (3-μm thick) were stained with periodic acid-Schiff (PAS) and Masson's trichrome stain. Each tissue section was evaluated under an Olympus light microscope (Olympus, Tokyo, Japan) with a high-resolution digital camera system (Penguin 600CL; Pixera Co., Los Gatos, CA). The area of glomeruli was measured using a Microanalyzer software program (version 1.1, Japan Poladigital Co., Tokyo, Japan) [18]. Glomerulosclerosis was quantified using the percentage increase in the relative mesangial matrix area (the PAS-positive area within the glomerulus divided by the glomerular capillary area; high magnification). All mean values were calculated from 10 glomeruli. Electron microscopy was performed for the mouse kidney specimens as described previously [16, 17]. A quantitative analysis was performed to count the number of podocyte foot processes per 10 μm of glomerular basement membrane in each glomerulus by electron microscopy. The mean number of podocyte foot processes was defined as the effacement score. The area of interstitial fibrosis in the cortex was evaluated with Masson's trichrome as described previously [19], with some modifications. Under low magnification, the number of casts in five randomly selected non-overlapping fields from the cortical region was counted and averaged as the cast score.

Immunohistochemistry

To detect lipid oxidation products, paraffin sections (3-μm thick) were stained using an N-histofine MOUSESTAIN KIT as described previously [20]. A mouse monoclonal antibody to 4-HNE was used as the primary antibody (1:100 dilution). Under low magnification, five randomly selected non-overlapping fields from the cortical region were analyzed. The 4-HNE positive areas that were stained in brown were picked up on digital images, and the percentage of the 4-HNE positive area relative to the whole area of the field was calculated (% area).

Renal catalase activity

Kidney samples were stored in a -80°C freezer until being assayed. The catalase activity in each kidney was determined by an ELISA using a catalase assay kit [21]. All procedures were performed according to the manufacturer's instructions.

Toll-like receptor-4 gene (Tlr4) mutation analysis in exon 3 in wild-type (C3H/AnLCsaCsa) and acatalasemic mice (C3H/AnLCsbCsb)

The detection of the Tlr4 mutation was performed using nested-PCR as described previously [22]. The nested PCR for detection of the missense mutation, a C to A transversion (Pro712His), which was reported in the C3H/HeJ mouse strain [23], was performed using the primers shown in Additional file 2: Table S1. The primers TLR4-ex3-F and TLR4-ex3-R were used for direct sequencing of DNA extracted from the mouse kidneys.

Statistical analyses

The data were shown as the means ± SE. The normal distribution and homogeneity of variance were checked, and logarithmic transformations were made for the variables if needed. Multiple comparisons between groups were made by Scheffe's test or the Steel-Dwass test. A Kaplan-Meier analysis and the Log-Rank statistic were used to explore the effects of ADR in both groups of mice. The statistical analysis was performed using the Excel add-in software Statcel 2 program (OMS, Tokyo, Japan) or the JMP 9 software program (SAS Institute Inc., Cary, North Carolina). P < 0.05 denoted the presence of a statistically significant difference.

Results

Changes in body weight, kidney weight, and heart weight in the mouse ADR nephropathy model

The ADR nephropathy model has been extensively studied in animals of the BALB/c background. Therefore, we first tried to characterize this model in our mouse strains. The body weight, relative kidney weight and relative heart weight were similar between the groups at the start of the experiment (Table 1). Both groups of mice lost a similar amount of body weight by 8 weeks after ADR administration. However, the relative kidney weight in acatalasemic mice was lower at 8 weeks than that in wild-type mice after ADR administration. The relative heart weights were similar for both groups during the experimental period.

Table 1 The metabolic and laboratory data of the mice
Full size table

The increase in albuminuria is significantly accelerated in acatalasemic mice after ADR administration

To evaluate the effect of acatalasemia on ADR-induced alterations of biological functions, we measured the parameters of renal function (Table 1) and UAE (Figure 1). The basal levels of BUN and Ccr were not significantly different between the wild-type and acatalasemic mice. In both groups, the renal function did not significantly change during the experimental period. However, UAE was elevated significantly after ADR administration at weeks 4 and 8 in both groups. The UAE in the acatalasemic mice was higher than that in wild-type mice after ADR administration at week 4.

Figure 1
figure 1

The changes in urinary albumin excretion (UAE) in wild-type (open bars) and acatalasemic mice (closed bars). The UAE significantly increased in both groups at 4 and 8 weeks after Adriamycin (ADR) administration. The elevation of UAE in acatalasemic mice was higher at 4 weeks after ADR administration compared to that in wild-type mice. Each column consists of the means ± SE. N = 9 to 39 animals/group. **: P < 0.01 vs. wild-type ADR mice at the same time point. #: p < 0.05 vs. week 0 in the same group. ##: p < 0.01 vs. week 0 in the same group.

Full size image

Acatalasemia accelerates glomerulosclerosis and tubulointerstitial fibrosis

At the start of the experiment, no histological abnormalities of the glomeruli were observed at the light microscopic level in either group (Figure 2A and 2B). The glomeruli in the acatalasemic mice were significantly hypertrophied after ADR administration at week 4 (Figure 2D and 2G). In the acatalasemic mice, glomerulosclerosis developed earlier than in wild-type mice, and it was significantly increased at week 4 compared with that in wild-type mice after ADR administration (Figure 2C through F and 2H). A decrease in the effacement score of the podocyte foot processes was observed in glomeruli of both groups of mice (Additional file 3: Figure S2A through F), however, there were no significant changes in the effacement scores in either group at week 4 and 8 after ADR administration (Additional file 3: Figure S2G). Tubulointerstitial fibrosis developed significantly in the kidneys of acatalasemic mice group (Figure 3B, D, F and 3G) at 4 and 8 weeks after ADR administration, with significant cast formation noted at week 4 (Figure 3H). The tubulointerstitial changes in the kidneys of wild-type mice after ADR administration were not statistically significant compared to that of acatalasemic mice (Figure 3A, C, E, G and 3H).

Figure 2
figure 2

The renal histology in wild-type and acatalasemic mice. Light micrographs of a glomerulus of wild-type (A, C and E) and acatalasemic (B, D and F) glomeruli at 0, 4 and 8 weeks after adriamycin (ADR) administration are shown. The week 0 glomeruli were microscopically normal. Note that the segmental sclerosis is significant in both acatalasemic and wild-type glomeruli at the later time points. The glomerular area (G) and sclerosis index (H) of wild-type (open bars) or acatalasemic (closed bars) mice are also shown. A through F: periodic acid-Schiff stain. Scale bars: 20 μm (A through F). G and H: Each column shows the means ± SE. N = 6 to 15 animals/group. *: P < 0.05 vs. wild-type ADR mice at the same time point. #: p < 0.05 vs. week 0 in the same group. ##: p < 0.01 vs. week 0 in the same group.

Full size image
Figure 3
figure 3

Light micrographs of the wild-type ( A, C and E ) and acatalasemic ( B, D and F ) kidney cortex at 0, 4 and 8 weeks after adriamycin (ADR) administration are shown. Note that interstitial fibrosis significantly developed in the kidneys of the acatalasemic mice. The fibrosis scores (G) and cast scores (H) of wild-type (open bars) and acatalasemic (closed bars) mice are also shown. A through F: Masson's trichrome stain. Scale bars: 200 μm (A through F). G and H: Each column includes the means ± SE. N = 6 to 15 animals/group. #: p < 0.05 vs. week 0 in the same group. ##: p < 0.01 vs. week 0 in the same group.

Full size image

Acatalasemia enhances the accumulation of lipid peroxides in the kidneys of the ADR nephropathy model mice

Modification of proteins and lipids by oxidative stress is believed to play a central role in a variety of biological activities, such as apoptosis and extracellular matrix expansion [18, 24, 25]. The major content of the cell membrane is lipids, and thus, lipid peroxidation may cause renal injury. We next examined whether acatalasemia influenced the lipid peroxidation products in the tubulointerstitial compartment of the kidneys in ADR nephropathy. At the start of the experiment, there was almost no accumulation of 4-HNE in the kidneys of either group (Figure 4A and 4B). There was an increase in the 4-HNE antibody labeling of various patches in the tubules of kidneys after ADR administration in both groups (Figure 4C through 4F). The degree of 4-HNE accumulation in the cortex of the kidneys in acatalasemic mice was significantly higher than that in wild-type mice 8 weeks after ADR administration (Figure 4G).

Figure 4
figure 4

Immunohistochemical staining of 4-hydroxy-2-nonenal (4-HNE). Wild-type (A, C and E) and acatalasemic (B, D and F) cortex samples at 0, 4 and 8 weeks after adriamycin (ADR) administration are shown. The 4-HNE positive areas (G) of wild-type (open bars) or acatalasemic (closed bars) mice are also shown. Scale bars: 50 μm (A through F). G: Each column shows the means ± SE. N = 6 to 8 animals/group. **: P < 0.01 vs. wild-type ADR mice at the same time point. ##: p < 0.01 vs. week 0 in the same group.

Full size image

The catalase activity in acatalasemic ADR nephropathy kidneys did not significantly change

The maintenance of tissue homeostasis requires an appropriate balance between oxidants and antioxidants. Two major antioxidant enzymes, catalase and GPX, are physiologically involved in the detoxification of H2O2, and thus protect tissues from oxidant-mediated injury. Therefore, we next examined the catalase activity in acatalasemic ADR kidneys. Although the renal catalase activity of the wild-type mice was increased significantly at week 4 after ADR administration, that from the acatalasemic mice was decreased compared with wild-type mice at the start of the experiment, and remained low during the entire experimental period (Figure 5). Since high catalase levels are found in erythrocytes, we compared the catalase activity between saline-perfused and unperfused kidneys after the removal of residual blood. The catalase activity of the perfused kidney was 6.0 ± 0.2% lower than that of the unperfused kidney [17]. Since the catalase activity was enhanced in wild-type mice 4 weeks after ADR administration and was thought to contribute to reactive oxygen elimination, we investigated the levels of other renal antioxidant enzymes, such as GPX and SOD, in wild-type or acatalasemic mice 4 weeks after ADR administration. The activity levels of GPX and SOD did not show any remarkable changes in either of the groups of mice (data not shown). Since SOD catalyzes the conversion of the superoxide anion radical to H2O2 and O2, it has been suggested that the level of production of ROS mediated by SOD may be similar in the ADR-nephropathy kidneys in both groups.

Figure 5
figure 5

The renal catalase activities in wild-type (open bars) and acatalasemic mice (closed bars). Each column consists of the means ± SE. N = 5 to 10 animals/group. *: P < 0.05 vs. wild-type adriamycin (ADR) mice at the same time point. **: P < 0.01 vs. wild-type ADR mice at the same time point. ##: p < 0.01 vs. week 0 in the same group.

Full size image

Toll-like receptor-4 mutation analysis

To examine the background of the mouse strains, we performed a mutation analysis of Tlr4 exon 3. Tlr4 did not show a C to A transversion due to a missense mutation in the C3H/AnL mouse strain, which was reported to be present in the C3H/HeJ strain (Additional file 4: Figure S3) [23].

Discussion

In the present study, acatalasemic mouse strains deficient in catalase activity were used as an animal model. These mice were found to be more susceptible to functional and morphological alterations in the kidneys induced by adriamycin than wild-type mice. The level of albuminuria and glomerulosclerosis in the acatalasemic mice after adriamycin injection was significantly higher than that in the wild-type mice. The renal catalase activity in these mice remained low, without compensatory upregulation of GPX or SOD. Collectively, these data suggest that the increased ROS, particularly the hydroxyl radical, resulting from the reduction of catalase activity, may be involved in the acceleration of glomerulosclerosis found under acatalasemic disease conditions.

Although some rat strains show complete susceptibility to ADR, most mouse strains do not. Zheng et al. showed that AKR/J, C3H/HeJ, CBA/J, C57BL/10J, LP/J, SWR/J, SJL/J, and 129S6/SvEvTac mice were resistant to ADR nephropathy, whereas 129S1/SvImJ and BALB/cByJ mice were susceptible [26, 27]. They also showed that the susceptible allele for the adriamycin was present in the DOXNPH locus [28]. We did not confirm whether this allele is present in the C3H/AnL mouse strain. Instead, we performed a mutation analysis of the TLR4 gene, because it has been thought that TLR4 is involved in progressive renal fibrosis [29]. TLR4 is considered to be the critical component of the LPS receptor complex. In 1998, a point mutation in the TLR4 gene was found to be the molecular basis of the LPS hyporesponsiveness in C3H/HeJ mice [23]. The C3H/AnL mice used in the present study did not have this mutation. Therefore, ADR induced mild renal fibrosis in both groups, although the degree of interstitial fibrosis was only significant in the acatalasemic mice.

ADR induced severe albuminuria, thus leading to values ranging from 23 to 226 mg/mgCr in BALB/cj mice and a value of 0.04 mg/gCr in B6/D2 mice [30]. In our study, ADR induced relatively mild albuminuria, with 0.6 mg/mgCr in wild-type and 1.5 mg/mgCr in acatalasemic mice. The C3H/AnL mouse strain is considered to have mild sensitivity to ADR. A single dose of ADR (9.5 mg/kg BW) brought about 40% segmental glomerulosclerosis in the BALB/c mouse strain in a previous study [31]. The wild-type mice which we used in this study had only 15% segmental glomerulosclerosis at week 8 after administration, while the acatalasemic mice had 20% glomerulosclerosis at week 8. This indicates that when the catalase activity is decreased, the resistance to ADR is diminished, and segmental glomerulosclerosis is induced. The number of foot processes of the podocytes did not differ between the mouse groups, indicating that the foot process effacement of the podocytes was affected to a similar degree in both mouse groups. The discrepancy between the level of albuminuria and the degree of foot process effacement may be due to the fact that this evaluation procedure compares only the surviving podocytes in the unsclerosed glomerulus.

Noiri et al. previously investigated the percentage of cortical interstitial volume in the ADR model [19], and Turnberg et al. evaluated tubulointerstitial injury, including cast formation [32]. We evaluated the presence of tubulointerstitial injury in similar analyses, and found significant tubulointerstitial injury only in the acatalasemic mice. The fact that acatalasemia exacerbates pulmonary fibrosis in bleomycin-induced lung injury [33] and chlorhexidine gluconate-induced peritoneal fibrosis [34] was already demonstrated. In this study, we showed that acatalasemia exacerbates renal fibrosis in a murine FSGS model.

The overexpression of catalase prevented albuminuria and interstitial fibrosis in the angiotensinogen transgenic mice [35]. In our study, the deficiency of catalase accelerated the albuminuria and glomerular sclerosis in the ADR nephropathy model. These data suggest that reactive oxygen species may contribute to progressive renal injury. Injac et al. reported that the activity of catalase, which detoxifies hydrogen peroxide to H2O, was increased after ADR administration in rats [36]. We measured the catalase activity in the whole kidneys of wild-type mice, and found that its level increased after ADR administration. Therefore, the differences in the albuminuria and the degree of glomerulosclerosis between wild-type and acatalasemic mice may be due to the significant difference of catalase activity in kidneys in these mice and the inability of the acatalasemic mice to increase catalase activity under oxidative stress.

Sheerin et al. reported that ADR-injected mice that died in the complement protein C3+/+ group at 6 weeks had a high serum urea level prior to death [37]. In the current study, all mice survived for at least 4 weeks after the injection of 15 mg/kg BW ADR; however 44.7% of the acatalasemic mice and 40.8% of the wild-type mice died 8 weeks after the administration of ADR. These mortality rates were not significantly different between the two groups (Additional file 1: Figure S1B). Since the incidence of albuminuria was analyzed in all of the mice that had survived at 8 weeks, but histological studies were not performed in more than half of the mice that survived at 8 weeks, bias might have been introduced into the analysis. We did not take blood samples or evaluate the serum urea levels in ADR-treated mice that died just before the end of the 8 weeks in this study. In addition, the administration of ADR can cause cardiotoxicity [15] and gastrointestinal toxicity in mice [38], however we could not confirm these toxicities in our experiments.

Acatalasemia is a rare human disease [3, 39]. It is unknown whether albuminuria and glomerulosclerosis are related to low catalase activity in humans, although the total antioxidant capacity is correlated with albuminemia, and inversely correlated with proteinuria and anti-DNA antibodies in subjects with lupus nephritis [40]. In addition, apoptosis, which is thought to be related to oxidative stress, correlated with the immunoserological activities of lupus nephritis [41] and idiopathic early FSGS [42]. However, the mechanism(s) by which oxidative stress influences human renal disease is largely unknown. Further studies are needed to elucidate whether and how a low catalase activity in humans influences either albuminuria or glomerulosclerosis associated with kidney diseases.

Conclusions

We herein demonstrated that ADR induced more albuminuria and glomerulosclerosis in catalase-deficient mice. Treatment with catalase supplementation may contribute to the suppression of progressive renal injury with proteinuria and glomerulosclerosis.