Attenuation of Interstitial Fibrosis and Tubular Apoptosis in db/db Transgenic Mice Overexpressing Catalase in Renal Proximal Tubular Cells

Rabbit polyclonal antibodies against bovine CAT and monoclonal antibodies against β-actin were purchased from Sigma-Aldrich Canada (Oakville, ON, Canada). Anti-cleaved caspase-3 polyclonal antibody, anti-Bax polyclonal antibody, and monoclonal anti-collagen type IV antibody were obtained from New England Biolabs (Pickering, ON, Canada), Santa Cruz Biotechnologies (Santa Cruz, CA), and Chemicon International (Temecula, CA), respectively. rCAT cDNA was a gift from Dr. Paul E. Epstein (University of Louisville, Louisville, KY). pKAP2 plasmid containing the kidney-specific androgen-regulated protein (KAP) promoter responsive to testosterone stimulation was obtained from Dr. Curt Sigmund (University of Iowa, Iowa, IA) and has been described elsewhere (30). Placebo pellets or pellets containing 5 mg testosterone with a 91-day release schedule (Cat. #NA-151) were purchased from Innovative Research of America (Sarasota, FL). Oligonucleotides were synthesized by Invitrogen (Burlington, ON, Canada). Restriction and modifying enzymes were from either Invitrogen, La Roche Biochemicals (Dorval, QC, Canada), or Amersham-Pharmacia Biotech (Baie d’Urfé, QC, Canada).

Tg mice (C57Bl/6 background) overexpressing rCAT in RPTCs (line #688) were created in our lab (by J.S.D.C) and have been described previously (24). Homozygous rCAT-Tg mice were then crossed with heterozygous db/m⁺ mice (B6.Cg m^+/+ Lepr^db in C57BL/6J background) (Jackson Laboratories, Bar Harbor, ME). Breeding was continued until homozygous db/db rCAT-Tg mice were obtained. db/db mice (B6.Cg + Lepr^db/+ Lepr^db/+ on a C57BL/6J background) were also purchased from Jackson Laboratories with a specific request for animals with persistent hyperglycemia. The blood glucose levels on arrival were >20 and >25 mmol/l glucose for female and male db/db mice, respectively. These levels were maintained throughout the experimental periods until 20 weeks of age. Homozygous db/db rCAT-Tg mice were identified by the PCR of genomic DNA for the rCAT transgene (24) and the mutated leptin receptor gene (31). The rCAT sense primer 5′-AAGCGGCCGCATGGCGGACAGCCGGGACCC-3′ and the HA (a tag sequence encoding amino acid residues 98–106 [YPYDVPDYA] of human influenza virus hemagglutinin) anti-sense primer 5′-GGCGTAGTCAGGCACGTCGT-3′ (32), the mouse leptin receptor gene forward primer 5′-AGAACGGACACTCTTTGAAGTCTC-3′ and the reverse primer 5′-CATTCAAACCATAGTTTAGGTTTGTGT-3′ of the mouse leptin receptor gene (31), respectively, were used for PCR. The DNA fragment (135 bp) of mouse leptin receptor gene was then digested with the restriction enzyme RsaI for 1 h at 37°C. If the allele was mutated, the 135-bp DNA fragment was cut into two smaller DNA fragments, appearing as 106- and 29-bp DNA fragments on 3% agarose gel electrophoresis.

Six to eight mice at 20 weeks of age were studied per group. Non-Tg age- and sex-matched db/m⁺ littermates served as controls. Females were implanted at week 8 with testosterone pellets (5 mg testosterone under 91-day release) or placebo pellets. All animals received standard mouse diet and water ad libitum. Animal care met the standards set forth by the Canadian Council on Animal Care, and all procedures were approved by the CHUM Animal Care Committee.

Blood pressure was monitored with a BP-2000 tail-cuff pressure instrument (Visitech Systems, Apex, NC) every 2 weeks for a period of 12 weeks, starting at age 8 weeks, measured three to four times per week and averaged. Each mouse was trained for at least for 5 consecutive days (30–60 min of blood pressure measurements) before the first study blood pressure measurements. Body weight was recorded, and blood glucose level was quantified weekly with an ACCU-CHEK Compact Plus glucose meter (Roche Diagnostics, Laval, QC, Canada). The mice were housed in metabolic cages to obtain 24-h urine samples for the assessment of albuminuria by ELISA (Albuwell and Creatinine companion; Exocell, Philadelphia, PA) (24,32). At the indicated times, the animals were killed by CO₂, and their kidneys were removed immediately. The left kidney was used for histology and immunohistochemistry. The right kidney was used for proximal tubule isolation by Percoll gradient (24,33). Aliquots of freshly isolated mouse renal proximal tubules (mRPTs) were immediately used for protein extraction or total RNA isolation.

Western blot analysis was performed as described previously (24,33). The membrane was first blotted with rabbit anti-CAT and then re-blotted with anti–β-actin monoclonal antibodies and chemiluminescent developing reagent (La Roche Biochemicals). The relative densities of CAT and β-actin bands were quantified by computerized laser densitometry (ImageQuant software, version 5.1; Molecular Dynamics).

Aliquots of freshly isolated proximal tubules were immediately used for ROS measurement by the lucigenin method (24,33). ROS production in mRPTs was normalized with protein concentration and expressed as relative light units per microgram of protein.

Kidneys were collected in Tissue-Tek cassettes (VWR Canlab, Montreal, QC, Canada), dipped immediately in ice-cold formaldehyde (10% in PBS) and fixed for 24 h at 4°C. The cassettes were then processed by the CHUM Pathology Department. Tissue sections (four to five specimens per group) were stained with periodic acid schiff or Masson's trichrome and analyzed visually under a light microscope by an observer unaware of the treatments. The collected images were analyzed and quantified using NIH ImageJ software.

The mean glomerular volume (V_G) was determined by the method of Weibel (34) with the aid of an image analysis software system (Motics Images Plus 2.0; Motic, Richmond, BC, Canada). The V_G was estimated by the mean glomerular tuft area (A_T) derived from the light microscopic measurement of 30 random sectional profiles of glomeruli from each group (n = 6 animals per group) using the formula: V_g = β/k × A_T^1.5, where β = 1.382 (shape coefficient for spheres) and k = 1.1 (size distribution coefficient).

Formaldehyde-fixed, paraffin-embedded kidney sections (5-μm thick) were deparaffinized in xylene and rehydrated. Apoptosis was assessed with a transferase-mediated dUTP nick-end labeling (TUNEL) kit (La Roche Biochemicals) according to the supplier's instructions (24). Immunohistochemical examination was performed by the standard avidin-biotin-peroxidase complex method (ABC Staining System; Santa Cruz Biotechnologies) (24,32) using primary anti-CAT polyclonal antibody (1:500 dilution), anti-cleaved caspase-3 polyclonal antibody (1:50 dilution), anti-collagen type IV monoclonal antibody (1:50 dilution), or anti-Bax polyclonal antibody (1:50 dilution). Targeted proteins were visualized by color development with 3,3′-diaminobenzidine tetrahydrochloride and counterstained with hematoxylin.

The percentage of RPTCs that stained positive on TUNEL assay and active caspase-3 and Bax immunostaining was estimated quantitatively as previously described (24). Briefly, five to six fields (∼500–600 cells/field) were randomly selected for each group of animals. The number of RPTCs containing TUNEL-positive active caspase-3 or Bax-staining was divided by the total number of RPTCs counted and multiplied by 100 to calculate the percentage of positively stained RPTCs.

Total RNA was used in real-time quantitative PCR to quantify the amount of Ang, p53, and Bax mRNA expressed in mRPTs as described previously (24). The forward and reverse primers corresponding to ANG (NM_007428, forward: 5′-ACAGACACCGAGATGCTGTT-3′ and reverse: 5′-CCACGCTCTCTGGATTTATC-3′), p53 (AB020317, forward: 5′ TGATGGAGA GTATTTCACCC-3′ and reverse: 5′-GGGCATCCTTTAACTCTAAG-3′), Bax (L22472, forward: 5′-CTGATCAGAACCATCATGGG-3′ and reverse: 5′-ACAAAGATGGTCACTG TCTG-3′), and β-actin (NM 031144, forward: 5′-ATGCCATCCTGCGTCTGGACCTGGC-3′ and reverse: 5′-AGCATTTGCGGTGCACGATGGAGGG-3′) were used in real-time quantitative PCR assays.

Data were expressed as means ± SD. The data were analyzed by one-way ANOVA using Bonferroni correction. P < 0.05 was considered statistically significant.

Figure 1A presents data from the specific PCR analysis of the rCAT-HA transgene and mutated leptin-receptor gene in offspring of the rCAT-Tg line 688 crossbred with heterozygous db/m⁺ mice. Animals displaying rCAT-HA transgene and the mutated leptin-receptor gene were used in subsequent experiments.

CAT protein expression in the RPTs of db/db rCAT-Tg mice was increased compared with heterozygous db/m⁺, db/m⁺ rCAT-Tg and db/db mice, respectively (Fig. 1B). Enhanced rCAT expression in RPTCs of db/m⁺ rCAT-Tg and db/db rCAT-Tg was documented by immunostaining of CAT in mouse kidneys (Fig. 1C). Furthermore, ROS generation was significantly augmented in freshly isolated RPTs from db/db mice compared with db/m⁺ or db/m⁺ rCAT-Tg mice (Fig. 1D). In contrast, the ROS level was significantly attenuated in RPTs of db/db rCAT-Tg mice compared with db/db mice. Taken together, these results confirm that KAP2 directs rCAT transgene expression in RPTCs of db/db rCAT-Tg mice.

Body weights (Fig. 2A) and blood glucose levels (Fig. 2B) of male and female db/db and db/db rCAT-Tg mice were significantly higher than in db/m⁺ and db/m⁺ rCAT-Tg mice, respectively. Body weight and blood glucose level did not differ significantly in nondiabetic db/m⁺ and db/m⁺ rCAT-Tg mice or in db/db and db/db rCAT-Tg mice. Blood glucose in female db/db rCAT-Tg mice appeared to be somewhat lower than in female db/db mice, but the difference did not reach statistical significance. These results indicate that rCAT overexpression in RPTCs alone is not effective in preventing body weight gain and hyperglycemia in db/db mice.

Average systolic blood pressure was increased in both male and female db/db mice, from week 10, compared with db/m⁺ mice, but the differences did not become statistically significant until weeks 12 and 14 in male and female db/db mice, respectively (Fig. 3A). In contrast, blood pressure remained unchanged in male and female db/db rCAT-Tg mice and in db/m⁺ and db/m⁺ rCAT-Tg mice (Fig. 3A).

Slight increases in the 24-h urinary albumin-to-creatinine ratio were detectable in both male and female db/db and db/db rCAT-Tg mice after week 16 compared with db/m⁺ and db/m⁺ rCAT-Tg mice, respectively (Fig. 3B). However, the albumin-to-creatinine ratio at week 16 did not differ significantly in both male and female db/db and db/db rCAT-Tg. In sharp contrast, the 24-h urinary albumin-to-creatinine ratio was significantly decreased (P ≤ 0.05) at 18 weeks in both male and female db/db rCAT-Tg mice compared with db/db mice. These results indicate that rCAT overexpression in RPTCs effectively prevented hypertension development and albuminuria progression in db/db mice.

As expected, male db/m⁺ (Fig. 4A, a and e) and db/m⁺ rCAT-Tg mice (Fig. 4A, b and f) had normal kidney morphology. However, db/db mouse kidneys (Fig. 4A, c and g) displayed cellular edema, reabsorption of droplets, and hypertrophy of glomerular and proximal tubular cell compared with db/m⁺ mice. Massive inflammatory cell infiltration was also seen in the Bowman's capsule of glomeruli in db/db mice. On the other hand, db/db rCAT-Tg mice (Fig. 4A, d and h) exhibited glomerular hypertrophy and a milder inflammatory cell infiltration. Moreover, extracellular matrix (ECM) protein expression was increased in db/db mouse kidneys compared with those of db/m⁺ and db/m⁺ rCAT-Tg and was normalized in db/db rCAT-Tg mice. V_G were significantly increased in db/db mice compared with db/m⁺ or db/m⁺ rCAT-Tg mice. The V_G values in db/db rCAT-Tg mice appeared to be slightly larger than in db/db mice but did not reach statistical significance (Fig. 4B).

Likewise, Masson's trichrome staining and immunostaining revealed, respectively, enhanced expression of collagenous components (Fig. 5A) and immunoreactive collagen type IV (Fig. 5B) in db/db mouse kidneys compared with db/m⁺ and db/m⁺ rCAT-Tg. Once again, staining for collagenous components and immunostaining for collagen type IV were normalized in db/db rCAT-Tg mouse kidneys, indicating that rCAT overexpression in RPTCs effectively prevented interstitial fibrosis in db/db mice.

Next, we investigated whether CAT overexpression could prevent apoptosis induced by hyperglycemia in mouse RPTCs. The TUNEL assay disclosed positively stained nuclei in RPTCs of db/db mice (Fig. 6A, c and g) but not in RPTCs of db/m⁺ mice (Fig. 6A, a and e), db/m⁺ rCAT-Tg (Fig. 6A, b and f) or db/db rCAT-Tg mice (Fig. 6A, e and h). The percentage of the TUNEL-positive RPTCs was significantly higher in db/db compared with db/m⁺ and db/m⁺ rCAT-Tg mice (Fig. 6B). In contrast, the number of TUNEL-positive RPTCs was significantly attenuated in db/db rCAT-Tg mice compared with db/db mice.

Markedly elevated active caspase-3 expression was detected in RPTCs from db/db mice (Fig. 7A, c) but not in db/m⁺, db/m⁺ rCAT-Tg, or db/db rCAT-Tg mice (Fig. 7A, a, b, and d). Bax expression was also enhanced in RPTCs from db/db mice (Fig. 7B, g) but not in db/m⁺, db/m⁺ rCAT-Tg, or db/db rCAT-Tg mice (Fig. 7B, e, f, and h). These data demonstrate that RPTC apoptosis in diabetic db/db mice can be attenuated by rCAT overexpression. These observations were confirmed by quantitation of the immunostaining of active caspase-3 and Bax (Fig. 7C). Importantly, active caspase-3 and Bax immunostaining were significantly attenuated in db/db rCAT-Tg mice. These data demonstrate that rCAT overexpression effectively attenuated RPTC apoptosis in diabetic db/db mice.

Expression of ANG mRNA (Fig. 8A), Bax mRNA (Fig. 8B), and p53 mRNA (Fig. 8C) was significantly elevated in RPTs from db/db mice compared with db/m⁺, db/m⁺ rCAT-Tg, and db/db rCAT-Tg mice. These increases were prevented in db/db rCAT-Tg mice.

We report here that rCAT overexpression in RPTCs of db/db mice effectively attenuates hypertension, albuminuria, interstitial fibrosis, tubular apoptosis, and proapoptotic gene expression, demonstrating an important role for ROS in tubular injury in diabetes.

Enhanced formation of ROS induces apoptosis of podocytes and mesangial and tubular cells (6,9,35). To investigate whether CAT can attenuate kidney injury in type 2 diabetic db/db mice, we created db/db Tg mice overexpressing rCAT specifically in their RPTs. The CAT transgene and the mutated insulin-2 gene were expressed in kidneys of homozygous db/db CAT-Tg mice. CAT protein expression (assessed by Western blotting) was at least twofold higher in RPTs from db/db CAT-Tg mice compared with db/m⁺, db/m⁺ CAT-Tg, or db/db mice. Likewise, immunostaining for CAT and ROS generation were higher in RPTCs of db/db CAT-Tg mice compared with those of db/m⁺, db/m⁺ CAT-Tg, and db/db mice. These findings confirm that KAP directs transgene expression in RPTCs (30,32).

The db/db mouse (Lepr^db/db) is a useful animal model to investigate the pathogenesis of nephropathy in type 2 diabetes because the renal lesions in db/db mice (BKS.Cg-m^+/+ background (36,37) and B6.Cg-m^+/+ background (38), as well as those without background specified (39,40), closely resemble those in human type 2 diabetes. These include albuminuria, mesangial matrix expansion, glomerulosclerosis, glomerular and tubular hypertrophy and apoptosis, and ECM gene expression, as well as macrophage accumulation and activation. In the present study, we detected increases in body weight and blood glucose level in both male and female db/db and db/db rCAT-Tg mice in comparison with db/m⁺ and db/m⁺ rCAT-Tg mice, respectively. These features are characteristic of db/db mice of both backgrounds (36–40).

Longitudinal experiments revealed that baseline blood pressure in both male and female db/db mice was significantly higher than in db/m⁺ and db/m⁺ rCAT mice. Between weeks 14 and 20, systolic blood pressure of db/db mice increased on average by 20–30 mmHg (P < 0.05) compared with db/m⁺ and db/m⁺ rCAT-Tg mice (average systolic blood pressure in db/m⁺ and db/m⁺ rCAT-Tg mice: 100 to 110 mmHg). At the present, we have no explanation for the lack of blood pressure changes in db/db mice in another study (36). A possible explanation might be that we have performed longitudinal blood pressure measurement (starting from week 8 to week 20) in the same animals (Fig. 3A), as opposed to cross-sectional blood pressure measurements in previous studies.

Interestingly, systolic blood pressure was normalized in db/db rCAT-Tg mice as compared with db/db mice, indicating that attenuation of ROS generation by CAT can prevent development of hypertension in db/db mice. At present, little is known about the mechanisms that lead to elevated blood pressure in db/db. One possibility is that enhanced intrarenal RAS activation may underlie the increases in blood pressure. Indeed, we have previously reported that Tg mice overexpressing ANG in their RPTCs do develop hypertension (32). Furthermore, our present data show that ANG mRNA expression is threefold higher in RPTs from db/db mice than in db/m⁺ mice and is normalized in db/db rCAT-Tg mice.

Since microalbuminuria is an important clinical marker for the early detection of hypertension- or diabetes-induced nephropathy, we monitored urinary albuminuria using the albumin-to-creatinine ratio. Albuminuria was detectable only after week 16 in db/db and db/m⁺ rCAT-Tg mice. In male db/db mice, albuminuria increased further at week 18 to levels comparable with those in female db/db mice at week 16. rCAT overexpression in db/db rCAT-Tg mice significantly (P ≤ 0.05) reduced albuminuria in both male and female mice at weeks 18 and 20 compared with db/db mice. To our best knowledge, this is the first report that rCAT overexpression in RPTCs can effectively attenuate the development of hypertension and progression of albuminuria in both male and female db/db mice. Overexpression of rCAT in female db/db mice was, however, more effective in lowering the albumin-to-creatinine ratio than in male db/db mice. This may be explained by differences in the effect of exogenous and endogenous testosterone on the kidney-specific androgen-regulated protein (KAP2) promoter, which controls the expression of the rCAT-HA transgene in Tg mice and is regulated by blood testosterone levels. In female Tg mice, testosterone pellets were implanted to stimulate the rCAT-HA expression in the kidney chronically. Since male Tg mice were implanted with placebo pellets, the rCAT-HA expression is regulated by endogenous blood testosterone level that shows interindividual variation. Hence, the response in testosterone-treated female Tg mice would be more marked as compared with male Tg mice. Taken together, these observations imply a link between renal ROS generation, hypertension, and albuminuria. However, it remains to be investigated whether albuminuria is secondary to elevated blood pressure, whether hypertension and albuminuria are unrelated events, and/or whether enhanced intrarenal ROS generation alone is capable of inducing albuminuria independently of systemic hypertension.

Histological examinations confirmed the presence of characteristic features of renal injury in the kidneys of db/db mice (36). In contrast, db/db rCAT-Tg mice exhibited glomerular hypertrophy, a milder inflammatory cell infiltration, and renal histology similar to that of db/m⁺ rCAT-Tg mice. Furthermore, ECM protein and collagen expression were attenuated in the kidneys of db/db rCAT-Tg mice. By using immunostaining, we also detected enhanced collagen type IV expression in glomeruli and tubulointerstitial spaces in db/db mice. Surprisingly, collagen type IV expression was completely normalized in db/db rCAT-Tg compared with db/m⁺ and db/m⁺ rCAT-Tg mice. These data suggest a critical role for tubular ROS generation in the development of interstitial fibrosis in diabetic db/db mice.

High glucose is a potent inducer of apoptosis in RPTCs via ROS generation (6,9). Consistently, we detected significantly higher number of apoptotic cells in RPTs of diabetic db/db mice than in RPTCs of db/m⁺ or db/m⁺ rCAT-Tg mice, whereas apoptosis was significantly reduced in db/db rCAT-Tg mice. These observations indicate that ROS generation in RPTCs of db/db mice actually induces tubular apoptosis and that CAT overexpression can at least partially prevent it. Indeed, active caspase-3 and Bax expression was normalized in db/db rCAT-Tg mice. Furthermore, RPTs from db/db mice exhibited markedly elevated expression of the proapoptotic genes p53 and Bax. Our observations that Bax and p53 mRNA expression was considerably lower in db/db rCAT-Tg mice than in db/db mice lend further support to the concept that rCAT overexpression attenuates RPTC apoptosis in db/db mice.

The precise mechanism(s) by which ROS cause hypertension and tubular injury (albuminuria, interstitial fibrosis, and RPTC apoptosis) in db/db mice remains unclear. One possibility is that ROS stimulate ANG production and activate the intrarenal RAS as RPTCs express all RAS components (41). Ang II promotes sodium reabsorption via stimulation of sodium/hydrogen exchanger activity in RPTCs (42), ultimately leading to development of hypertension. Furthermore, Ang II is capable of stimulating transforming growth factor-β 1 (TGF-β1) and subsequently enhancing ECM protein, collagen type IV, and proapoptotic genes in RPTCs, leading to tubular injury (interstitial fibrosis and cellular apoptosis) (43). Indeed, neutralization of TGF-β alleviated fibrosis in diabetic animal models, including db/db mice (44). Moreover, TGF-β1 was reported to promote apoptosis after renal injury in both in vivo and in vitro models (45,46). We have previously found that intrarenal ANG gene expression is essential for TGF-β1 gene expression in rat RPTCs exposed to high glucose in vitro (47) and that Tg mice overexpressing ANG in RPTCs exhibit hypertension and albuminuria (32). Our present data showed at least threefold higher ANG mRNA expression in RPTs of db/db mice and normalization of ANG mRNA expression in db/db rCAT-Tg mice. Taken together, these findings support a role for intrarenal ROS and RAS in the induction of interstitial fibrosis and RPTC apoptosis.

The present results may have clinical implications to type 2 diabetes. Since tubular apoptosis is detectable in human diabetic kidneys with albuminuria (27,28), and tubular atrophy appears to be a better indicator of disease progression than glomerular pathology (13–18), we suggest that RPTC apoptosis may be an initial mechanism for tubular atrophy in diabetes. Our data point toward ROS as one of the key mediators of this process. The source(s) of ROS production in RPTs remains, however, to be identified. Among these sources, mitochondrial (29) and membrane-bound NADPH oxidase-derived (48) ROS production have already been reported. Interestingly, diabetes is associated with enhanced expression of NADPH oxidase subunits, Nox2, Nox4, and p47^phox mRNA and p47^phox in mouse glomeruli (49), indicating that these NADPH oxidase subunits might be involved in ROS production in RPTs. Clearly, more experiments along these lines are warranted.

In summary, the present study suggest a critical role for CAT in attenuating hypertension, albuminuria, interstitial fibrosis, and RPTC apoptosis in db/db mice in vivo. Our studies raise the possibility that selective activation of this enzyme may provide a novel approach in preventing or reversing the pathophysiological manifestations of diabetic nephropathy including tubular atrophy in type 2 diabetes.

This work was supported by a grant from the Canadian Institutes of Health Research (MOP 62920 to J.S.D.C.), the Kidney Foundation of Canada, and the U.S. National Institutes of Health (HL-48455 to J.R.I.).

The authors thank Dr. Paul Goodyer (Montreal Children's Hospital, MUHC, Montreal, QC, Canada) for advice on the quantitation of the glomerular volume and Mr. Ovid M. Da Silva, Editor, Research Support Office, Research Centre, CHUM, for editing this manuscript.