Differences in Immunolocalization of Kim-1,RPA-1,and RPA-2 in Kidneys of Gentamicin-,Cisplatin-,and Valproic Acid-Treated Rats: Potential Role of iNOS and Nitrotyrosine

Animals

Male ten-, twenty-five-, forty-, and eighty-day-old SD rats (Harlan, Indianapolis, IN) were used in this study. Rats were maintained in plastic cages in a controlled environment at 21°C–23°C with a twelve-hour light-dark cycle, fed Purina rodent laboratory chow (Purina Mills, St. Louis, MO), and provided water ad libitum. The protocol was approved by the Institutional Animal Care and Use committee of the FDA White Oak Animal Program (Silver Spring, MD). Animal care procedures for the study conformed to the 1996 ILAR (Institute of Laboratory Animal Resources) Guide for the Care and Use of Laboratory Animals (NRC 1996).

Experimental Procedures

Statistics

The nonparametric Kruskal-Wallis rank test was used to assess difference in renal lesion scores. A p value < .05 was considered statistically significant.

Histopathology

In control rats in the Gen study, no renal injury was found in any proximal tubular segment (Figures 1A and 1B). In Gen-treated rats, epithelial cell necrosis (Figure 1C) and apoptosis (Figure 1D) were located predominantly in the S1/S2 segments of proximal tubules and to a lesser extent in the S3 segment. In control rats in the Cis study, neither tubular epithelial cell necrosis nor apoptosis was found in medullary CD or LH (Figures 1E and 1F). In contrast, in forty-day-old rats treated with 3 mg/kg of Cis, epithelial cell necrosis and apoptosis primarily occurred in the S3 segments of proximal straight tubules (figure not shown), CD (figure not shown), and LH in the medulla (Figures 1G and 1H). Although the epithelial cell injury (degeneration, regeneration, necrosis, and apoptosis) induced by Gen and Cis was similar, the injury was located in the S1/S2 segments (proximal convoluted tubules) and the S3 segments of proximal straight tubules for Gen nephrotoxicity and in the medullary CD or medullary LH for Cis nephrotoxicity (Figures 1C, 1D, 1G, and 1H).

Renal injury was more severe in Gen- and Cis-treated adult rats (forty or eighty days old) than in immature rats (ten days old), and renal injury was greater in rats treated with Gen for fourteen days compared to those treated for six days (Table 2). In the Cis study, renal injury in aged rats treated with 6 mg/kg of Cis (average lesion score = 3.1) was more severe than renal injury in aged rats treated with 3 mg/kg of Cis (average lesion score = 2.8) or 1 mg/kg of Cis (average lesion score = 0.7).

Immunolocalization of iNOS and Nitrotyrosine

In saline-treated rats, expression of iNOS (Figure 2A) and nitrotyrosine (Figure 2C) was not found in either the renal cortex or the medulla. In contrast, in Gen-treated rats, iNOS (Figure 2B) and nitrotyrosine (Figure 2D) were expressed in the cytoplasm of epithelial cells in the proximal tubules. Moreover, iNOS and nitrotyrosine were also expressed in the flat epithelial cells in the dilated tubules (Figures 2B and 2D), the morphology of which was identical to that detected in RPA-1–stained CD in Gen-treated rats (Figures 4B–4D). In saline-treated control rats in the Cis study, iNOS (Figure 2E) and nitrotyrosine (Figure 2G) were not apparent in the epithelial cells of CD or LH in the medulla. In contrast, in Cis-treated rats, iNOS was expressed predominantly in the epithelial cells of LH of the medulla, although slight expression of iNOS was also discernible in the epithelial cells of CD in the medulla (Figure 2F). Nitrotyrosine expression was found only in the epithelial cells of LH of the medulla (Figure 2H).

Immunolocalization of Kim-1 in Gen and Cis Nephrotoxicity

Saline-treated, ten-day-old rats did not exhibit positive immunoperoxidase staining for Kim-1 in any segment of the nephron in any renal tissue sections examined (Figure 3A). In contrast, in Gen-treated, ten-day-old rats, Kim-1 expression was located predominantly in the cortex and slightly in the outer stripe (Figure 3B). In saline-treated, ten-day-old rats, Kim-1 expression was not detected in the S1/S2 or S3 segments (Figure 3C). However, in Gen-treated, ten-day-old rats, Kim-1 expression was detected primarily in the S1/S2 segments (Figure 3D). In eighty-day-old rats treated with 100 mg/kg Gen for fourteen days, Kim-1 was expressed in the proximal convoluted tubular cells of the S1/S2 segments (Figure 3E) and in the proximal straight tubular cells of the S3 segments (Figure 3F). In Cis nephrotoxicity, Kim-1 induction was predominantly localized to epithelial cells of the S3 segments, proximal straight tubules (Figures 3G and 3H) in rats treated with either 3 mg/kg (Figure 3G) or 6 mg/kg Cis (Figure 3H). In both Gen and Cis nephrotoxicity, the extent of Kim-1 expression in the S1/S2 or S3 segments increased with time for Gen-treated rats and dose for Cis-treated rats. In addition, the immunostaining intensity correlated with the severity of tubular epithelial cell injury (Table 2).

Immunolocalization of RPA-1 in Gen and Cis Nephrotoxicity

Positive staining with mAb RPA-1 (isotype IgG) and negative control staining with the nonimmune mouse IgG₁ were performed concurrently. Using the negative control antiserum (mouse IgG₁)_, RPA-1–like expression was not found in CD in the cortex and medulla in rats treated with saline (figure not shown). However, weak nonspecific binding of the RPA-1 antibody was identified in necrotic proximal tubular cells in Gen-treated rats but not in saline control rats (figure not shown). The results were similar to those in a previous study of Gen nephrotoxicity in rats (Zhang, Shaw et al. 2008).

Using the mAb RPA-1 in ten-day-old, saline-treated control rats in the Gen study, constitutive (basal) levels of RPA-1 expression were found mainly in the inner medullary CD epithelial cells and sporadically in the outer medullary and the cortical CD epithelial cells (Figure 4A). The lumens of normal CDs were very small and narrow (Figure 4A). In contrast, in ten-day-old rats treated with 100 mg/kg Gen, RPA-1 was expressed in the epithelial cells of CD in the cortex (Figures 4B-4D). The lumens of CD stained with RPA-1 appeared irregular in shape and larger. Strong, positive-staining amorphous materials were found in some of the lumens (Figure 4B). It was evident that nonspecific binding of RPA-1 occurred in the epithelial cells of proximal tubules (Figure 4B). No significant alterations of RPA-1 expression were found in the medullae of ten-day-old rats treated with 100 mg/kg Gen. In contrast to the intensive staining of cortical CD in the Gen study, RPA-1 expression was comparable between saline-treated rats (Figure 4E) and Cis-treated rats (Figure 4F). In eighty-day-old, saline-treated control rats in the Cis study, constitutive levels of RPA-1 expression were detected in a few CD epithelial cells in the medulla (Figure 4G). In contrast, in eighty-day-old rats treated with 3 mg/kg Cis, RPA-1 expression was increased in the medullary CD, particularly in the proliferative tubular epithelial cells of the dilated CD (Figure 4H). In Cis-treated rats, RPA-1 expression in the epithelial cells of the medullary CD increased with rat age (Table 2).

Immunolocalization of RPA-2 in Gen and Cis Nephrotoxicity

The negative control antiserum (mouse IgM) for RPA-2 showed no staining in the cortical and medullary LH (Figures 5A and 5B) in saline-treated control rats, but nonspecific binding of the anti RPA-2 was detected in the necrotic proximal epithelial cells in Gen-treated rats (figure not shown), which is similar to the findings of our previous Gen study (Zhang, Shaw et al. 2008). In Gen-treated rats, RPA-2 was expressed in the epithelial cells in the thick ascending segment of LH, but the staining intensity was weak and blurred (figures not shown). In addition, nonspecific binding of RPA-2 was also located in the cytoplasm of necrotic proximal cells or their debris in rats treated with 100 mg/kg Gen for six days.

In control saline-treated, eighty-day-old rats in the Cis study, constitutive levels of RPA-2, with weak and blurry staining in most of the LH and distinct staining in some of the LH, were found in the medulla (Figure 5C). In contrast, RPA-2 expression was increased in the tubular epithelial cells of LH in the medulla in eighty-day-old rats treated with 1 mg/kg Cis (Figure 5D), 3 mg/kg Cis (Figures 5E and 5F), and 6 mg/kg Cis (Figures 5G and 5H). The staining intensity increased with the increase in dose (1, 3, or 6 mg/kg). Expression of RPA-2 in Cis-injured tubular epithelial cells of the medullary LH also increased with rat age (Table 2).

Histopathology and Immunolocalization of Kim-1, RPA-1, and RPA-2 in VPA Study

The hepatotoxicant VPA induced hepatocyte degeneration (cytoplasmic vacuolization) and necrosis and hemorrhage (Espandiari et al. 2008), but it did not induce renal lesions in any segment of the cortex (Figure 6A) or the medulla (Figure 6B). In twenty-five-day-old rats treated with 650 mg/kg VPA, Kim-1 was not expressed in the epithelial cells of proximal tubules (Figure 6C) and in the outer stripe (Figure 6D). Constitutive levels of RPA-1 were expressed in the tubular epithelial cells of CD in the corticomedullary junction (Figure 6E) and in the cortex (Figure 6F). Very low constitutive levels of RPA-2 were expressed in the tubular epithelial cells of LH in the cortex (Figure 6G) and in the corticomedullary junction (Figure 6H). Expression of iNOS and nitrotyrosine was not found in saline-and VPA-treated rats (figure not shown).

Immunohistochemical Patterns of iNOS, Nitrotyrosine, Kim-, RPA-1, and RPA-2 Expression in the S1/S2 and S3 Segments Associated with Gen Nephrotoxicity

Microscopic examination demonstrated that Gen-induced AKI was predominantly localized to the S1/S2 segments, with minor involvement of the S3 segment. The IHC data showed that induction of iNOS and nitrotyrosine, in association with expression of Kim-1, was greater in the S1/S2 segments of the cortex than in the S3 segments in the corticomedullary rays. Nonspecific binding of anti-RPA-1 and anti-RPA-2 was identified in the necrotic proximal cells of the same segments. These spatial distributions of AKI biomarkers and oxidative and nitrosative stress markers lend support to our hypothesis that NO production was greater in the S1/S2 segments. However, lack of evidence obtained from specially designed studies (such as quantification of iNOS and nitrotyrosine activity in micro-dissected segments or use of NOS knockout animals) leaves this hypothesis inconclusive at present. In addition, the specialized histological features and functional properties of various nephron segments may also contribute to the topographic distribution of renal injury. The morphologically distinct segments (S1, S2, and S3) of rat kidney proximal tubule cells have been investigated in detail by Maunsback (1966a, 1966b, 1966c). In the rat kidney, the S1 segment has a tall brush border, a prominent endocytic system, numerous mitochondria, and extensive invaginations of the basolateral plasma membrane; the S2 segment has a less prominent brush border than the S1 segment and large secondary lysosomes; and the S3 segment has a tall brush border, less prominent endocytic system than in the S1 and S2 segments, sparse basolateral invaginations, and scattered smaller mitochondria (Madsen and Tisher 2004; Maunsback 1966a, 1966b, 1966c). The prominent endocytic activity and abundant mitochondria in S1-segment epithelia and abundant secondary lysosomes in the S2 segment epithelia with lower endocytic activity and fewer mitochondria may explain why there is general predilection for injury in the S1/S2 segments than in the S3 segment in Gen nephrotoxicity, since it is generally accepted that Gen is taken into the cell mainly by endocytosis (Greaves 1990). It is postulated that damage to other segments may occur in Gen nephrotoxicity, in terms of physiological function of downstream nephron effects resulting from hyperfiltration and functional nephron loss. Other biochemical and molecular mechanisms also play a role in Gen nephrotoxicity; however, a detailed discussion is beyond the scope of this article.

Immunohistochemical Patterns of iNOS, Kim-1, RPA-1, and RPA-2 Expression in the S3 Segment, CD, and LH associated with Cis Nephrotoxicity

Cis-induced AKI was limited almost exclusively to the medulla, specifically, the S3 segments in the outer medulla, and the LH and CD in the outer and the inner medulla. At the sites of Cis-induced AKI, the induction of iNOS and nitro-tyrosine was associated with increased expression of Kim-1, RPA-1, and RPA-2. The mechanisms by which Cis induced iNOS and nitrotyrosine; increased the expression of Kim-1, RPA-, and RPA-2; and caused renal injury are not yet fully elucidated.

Immediately after Cis administration, the proximal and distal tubular reabsorption rates of sodium and water are significantly decreased (Daugaard et al. 1987; Daugaard 1990). The reduction of reabsorption capacities in the S1/S2 segments and the S3 segments occur concomitantly with reduced renal blood flow rates, increased renal vascular resistance, and increased urinary levels of sodium and potassium. Therefore, it is possible that the CD and LH, which are permeable to water and salts, will be overburdened by increased urinary sodium, potassium, and water (Daugaard et al. 1987). As a result, the competing burdens of high oxygen demand and insufficient oxygen supply in the LH and CD segments may make these segments more susceptible to hypoxic injury.

A Common Final Pathway Leading to Nephrotoxicity via Nitrotyrosine Generation

Activation of iNOS has been implicated in the mechanisms of various forms of renal injury induced by both nephrotoxicants (e.g., Gen, Cis, chromate) and other forms of renal insult (e.g., ischemia/reperfusion) (Table 1). There is growing evidence to suggest that most of the cytotoxicity attributed to NO is rather a result of peroxynitrite, which is formed by the reaction between superoxide anion and nitric oxide (NO^•) (Pacher et al. 2007). Therefore, attenuation of the toxic levels of nitrotyrosine and selective inhibition of iNOS have become pharmacological strategies against nitrosative stress-mediated toxicity in vivo (Denicola and Radi 2005; Pacher et al. 2007). Recent IHC studies have demonstrated that the hydroxyl radical scavenger S-allylmercaptocysteine (a garlic-derived compound) decreased expression of nitrotyrosine, dinitrophenol, and 4-hydroxy-2-nonenal in kidney tissue sections (Pedraza-Chaverri et al. 2004). Concurrent attenuation of increases in serum creatinine, blood urea nitrogen (BUN), and urinary N-acetyl-β-D-glucosaminidase (NAG), as well as reduced kidney damage (proximal tubular cell necrosis) in rats treated with Gen was also observed. These findings suggest that oxidative and nitrosative stress are implicated in Gen nephrotoxicity (Pedraza-Chaverri et al. 2004). In addition, the peroxynitrite decomposition catalyst 5,10,15,20-tetrakis (4-sulfonatophenyl) porphyrinato iron (III) (a soluble complex to catalyze the destruction of peroxynitrite) decreased expression of nitrotyrosine in the kidney tissue sections and attenuated levels of serum creatinine, BUN, NAG, and total proteins in Cis-treated rats. These findings demonstrated that Cis nephrotoxicity was peroxynitrite mediated (Chirino et al. 2004). Recently, a selective and irreversible iNOS inhibitor, N-[3-(aminomethyl)benzyl] accetamine (1400W), was shown to suppress oxidative stress (decreased renal cortex malondialdehyde content and decreased expression of 4-hydroxy-2-nonenal in the kidney tissue sections) and nitrosative stress (decreased expression of nitrotyrosine in the tissue sections). These alterations were accompanied by a reduction in kidney damage (increased proteinuria and decreased creatinine clearance) and improvement of renal function). These results suggest that NO^•-derived iNOS contributes to Cis nephrotoxicity (Chirino et al. 2008).

In contrast to previous studies, the present study demonstrated increased expression of iNOS and nitrotyrosine in the epithelial cells of proximal tubules, CD, or LH in Gen- and Cis-treated rats (Figures 2). Therefore, these findings support the hypothesis of a common final pathway leading to Gen or Cis nephrotoxicity via nitrotyrosine generation (the production of NO and superoxide); however, further studies of decreased formation of toxic levels of peroxynitrite are required to reach specific conclusions.

Kim-1, RPA-1, and RPA-2 as Site-specific Biomarkers for Gen and Cis Nephrotoxicity

The present study demonstrated that the expression of Kim-1, RPA-1, and RPA-2 was significantly increased in Gen- and Cis-induced renal tubular epithelial cell injury as compared with their expression in rats treated with saline or VPA, a non-nephrotoxicant. The present study also demonstrated that the immunolocalizations of Kim-1, RPA-1, and RPA-2 were distinctly different.

Kim-1 was mainly expressed in the S1/S2 segments and less in the S3 segments of the cortex in Gen-induced renal injury, whereas it was expressed mainly in the S3 segment in Cis nephrotoxicity. The immunolocalization of Kim-1 is clearly consistent with histopathology for Gen versus Cis nephrotoxicity. These findings demonstrate that Kim-1 is a specific and sensitive biomarker for proximal tubular injury.

The anti-RPA-1 monoclonal antibody recognizes an antigen in the epithelial cells of CD (Falkenberg et al. 1996; Hildebrand et al. 1999). A recent study (Harpur et al. 2008) has shown urinary RPA-1 to be a sensitive and specific biomarker of renal papillary CD injury after treatment with N-phenylanthranillic acid (NPAA), a drug that targets papillary CD. However, the present study revealed that RPA-1 expression was increased in the epithelial cells of CD in the cortex in Gen nephrotoxicity (Figures 4B-4D), in association with morphological evidence of cellular injury, whereas RPA-1 expression was increased in the epithelial cells of CD in the medulla in Cis nephrotoxi-city (Figure 4H). These findings may explain our earlier study of Gen nephrotoxicity in rats (Shaw et al. 2007), in which the urinary levels of RPA-1 were significantly elevated four- to twelve-fold higher (dose-dependent) than controls at seventy-two hours after treatment. Therefore, RPA-1 has the potential to serve as a novel biomarker for cortical CD injury induced by Gen, as well as for papillary CD injury induced by NPAA.

The monoclonal anti-RPA-2 recognizes an antigen in the epithelial cells of LH (Falkenberg et al. 1996; Hildebrand et al. 1999). Scant information has been published on RPA-2 distribution and induction in drug-induced nephrotoxicity. The present study demonstrated that RPA-2 expression was increased in the epithelial cells of LH in the medulla (Figures 5C-5H) in association with morphological evidence of cellular injury in Cis nephrotoxicity. No urinary RPA-2 data are available at this time. Although the information regarding RPA-2 in this study is incomplete, the up-regulation of RPA-2 was clearly visible, and RPA-2 has the potential to serve as a novel biomarker for Cis nephrotoxicity.