- Email: [email protected]
THE ROLE OF NITRIC OXIDE IN OBSTRUCTIVE NEPHROPATHY
0022-5347/00/1634-1276/0 THE JOURNAL OF UROLOGY® Copyright © 2000 by AMERICAN UROLOGICAL ASSOCIATION, INC.®
Vol. 163, 1276 –1281, April 2000 Printed in U.S.A.
THE ROLE OF NITRIC OXIDE IN OBSTRUCTIVE NEPHROPATHY ANDREW HUANG, LANE S. PALMER, DAVID HOM, ELSA VALDERRAMA AND HOWARD TRACHTMAN From the Divisions of Pediatric Urology, Pathology and Nephrology, Schneider Children’s Hospital, Long Island Jewish Medical Center, New Hyde Park, New York
ABSTRACT
Purpose: Ureteral obstruction leads to tubulointerstitial fibrosis and loss of renal function. Nitric oxide production ameliorates fibrosis due to obstructive uropathy. However, nitric oxide is produced by 3 isoforms of the enzyme, nitric oxide synthase. We evaluated the role of inducible nitric oxide synthase in obstructive uropathy using nitric oxide synthase knockout mice, and determined whether the administration of L-arginine to promote nitric oxide synthesis by alternative nitric oxide synthase isoforms modulates renal fibrosis in these animals. Materials and Methods: Complete unilateral ureteral obstruction was created in wild-type C57 and inducible nitric oxide synthase knockout mice. Control animals of each strain underwent sham surgery. Throughout the experiment mice had free access to untreated tap water or water supplemented with 10 gm./l. L-arginine. Animals were sacrificed 1 and 2 weeks, respectively, after creation of unilateral ureteral obstruction. We obtained serum as well as bladder and obstructed renal pelvic urine, and determined the nitrite level in each fluid. Renal cortical thickness was measured in the normal and obstructed kidneys. The degree of tubulointerstitial fibrosis was evaluated by trichrome staining and type I collagen deposition in kidney tissue specimens. Results: Nitrite was significantly decreased in the serum, bladder and renal pelvic urine of inducible nitric oxide synthase knockout mice with unilateral ureteral obstruction compared with that in wild-type C57 mice at 1 and 2 weeks (p ⬍0.05). In knockout mice with unilateral ureteral obstruction 1 week in duration that drank tap or L-arginine supplemented water nitrite in serum and each urine sample was higher than in sham operated knockout controls. The level returned to baseline after 2 weeks of obstruction (p ⬍0.05). After 2 weeks of obstruction there was significantly greater cortical thinning in knockout than in C57 mice (p ⬍0.05). Moreover, knockout mice given L-arginine supplemented water for 2 weeks had even greater cortical thinning than after 1 week or than mice given tap water for 1 to 2 weeks (p ⬍0.05). Decreased renal cortical thickness in knockout mice after 2 weeks of obstruction was associated with less intense trichrome staining and a virtual absence of type I collagen deposition compared with findings in the wild-type C57 strain. Conclusions: Inducible nitric oxide synthase knockout mice with unilateral ureteral obstruction have significantly lower nitrite in serum and urine than wild-type C57 mice. Knockout mice also have more severe renal cortical thinning than C57 animals after creation of unilateral ureteral obstruction. Providing L-arginine supplemented water to inducible nitric oxide synthase knockout mice exacerbates the loss of cortical thickness. Alterations in cortical thinning that we observed in knockout mice were associated with decreased tubulointerstitial fibrosis and a decreased net renal extracellular matrix accumulation. These data indicate that endothelial or neuronal nitric oxide synthase may be more important than inducible nitric oxide synthase for modulating renal fibrosis in obstructive uropathy. KEY WORDS: kidney, nitric oxide, ureteral obstruction
Obstructive uropathy in children is usually the result of a congenital structural or neurological lesion of the urinary system. In severe cases this condition leads to progressive glomerulosclerosis, tubulointerstitial fibrosis and loss of renal function. Complete unilateral ureteral obstruction is a well described experimental model of this renal disease.1 Numerous factors have been implicated in the chronic renal injury that develops in this disorder, including eicosanoids, angiotensin II, transforming growth factor- and nitric oxide.2– 4 Inhibition of nitric oxide synthesis exacerbates and enhanced production of nitric oxide ameliorates fibrosis in exAccepted for publication November 12, 1999. Supported by a grant from the Schneider family.
perimental obstructive uropathy.5, 6 Nitric oxide is synthesized from the guanidino nitrogen of L-arginine by nitric oxide synthase.5 However, there are 3 distinct isoforms of nitric oxide synthase in mammals.7 Endothelial and neuronal nitric oxide synthase are the 2 forms of the enzyme expressed constitutively in the basal state that are present in endothelial and neuronal cells, respectively. The other form, inducible nitric oxide synthase, is present in various cells, including glomerular mesangial and tubular epithelial cells.8, 9 While activation of endothelial or neuronal nitric oxide synthase leads to slight transient increases in nitric oxide production, stimulation of inducible nitric oxide synthase leads to marked and sustained elevations in nitric oxide synthesis. Because of the importance of nitric oxide in pro-
1276
1277
ROLE OF NITRIC OXIDE IN OBSTRUCTIVE NEPHROPATHY
gressive renal disease and the complex nature of nitric oxide production, we performed experiments with unilateral ureteral obstruction in wild-type C57 and inducible nitric oxide synthase knockout mice to determine the specific role of inducible nitric oxide synthase in modulating fibrosis due to obstructive uropathy. In addition, we assessed the impact of endothelial nitric oxide synthase and neuronal nitric oxide synthase in the fibrotic process by providing supplemental L-arginine, the precursor to nitric oxide, to these 2 strains of mouse. MATERIALS AND METHODS
Control wild-type C57BL/6 mice were provided and inducible nitric oxide synthase knockout mice were bred from homozygous animals for the study. All mice were fed standard rodent chow. At age 9 weeks they were anesthetized with an intraperitoneal injection of 50 mg./kg. pentobarbital. The right colon was reflected medially. The proximal right ureter was dissected free and ligated with 4-zero silk ties. Control mice underwent sham surgery without ureteral ligation. No mice died as a result of operative complications at the time of right ureteral obstruction creation or sham surgery. Mice were given free access to untreated tap water or water supplemented with 10 gm./l. L-arginine. L-arginine drinking water was started 24 hours preoperatively and continued until sacrifice. There were 5 to 12 mice in each experimental group. Animals were anesthetized with an intraperitoneal injection of 50 mg./kg. pentobarbital 1 and 2 weeks after the onset of unilateral ureteral obstruction or sham surgery. A blood sample was obtained, and urine was collected from the bladder and obstructed renal pelvis. The kidneys were perfused with cold Hanks balanced salt solution, removed and incised coronally. Sections were fixed by immersion in 10% formalin. To determine the nitrite level blood samples were centrifuged at 1,000 rpm for 5 minutes and the serum was removed. Plasma and urine samples were tested for nitrite using the Greiss reaction.10 Briefly, serum samples were deproteinized by adding a 1/20 volume of zinc sulfate and then diluted 1:1 with distilled water. Urine samples were diluted 1:6 with distilled water. Samples were incubated for 10 minutes with an equal volume of Greiss reagent (1% sulfanilamide, 0.1% naphthylethylene diamine dihydrochloride and 2.5% phosphoric acid). Absorbance was measured in a spectrophotometer at 550 nm. A standard curve with known nitrite concentrations of 1 to 50 m. was included in each assay. Serum and urine samples were assayed for urea nitrogen and creatinine, respectively, using an automated analyzer. Urinary nitrite was expressed as nmol./mg. creatinine. For histological evaluation tissues were embedded in paraffin, and 4 . sections were stained with hematoxylin and eosin, and Masson’s trichrome. Cortical thickness was measured in the region of the kidney with the best preservation of renal parenchymal structures. Specimens were examined histologically for interstitial fibrosis using the trichrome stain and graded on a scale of 0 —foci of fibrosis in less than 10% of the specimen, 1—10% to 25%, 2—26% to 50%, 3—51% to 75%, 4 — greater than 75%. For immunohistochemical study primary antibodies to types I (No. AB765) and III (No. AB747) collagen were used in a 1:50 dilution and incubated with the renal sections for 32 minutes. After exposure to 1:10 dilution of a solution to unmask tissue antigens for 40 minutes at 40C in a water bath, slides were rinsed with water and stained using an automated immunoperoxidase stainer. Sections were counterstained with hematoxylin and 1.5% bluing slide ammonia in 70% ethanol. They were then dehydrated and a coverglass was applied. A single observer (E. V.) blinded to treatment groups evaluated all histopathological and immunohisto-
chemical slides. Data are shown as the mean plus or minus standard deviation. We used analysis of variance and the t test to compare findings in each group at the various time points for the various fluids administered and findings in the 2 mouse strains. RESULTS
There was no acute mortality during the surgical procedure. Fewer than 5% of the mice of either strain died more than 24 hours after the onset of unilateral ureteral obstruction. Blood urea nitrogen ranged from 29 to 59 mg./dl. As expected, because ureteral obstruction was unilateral, there were no differences in either group in blood urea nitrogen regardless of the type of drinking water provided, or whether mice were sacrificed 1 or 2 weeks postoperatively (data not shown). In sham operated C57 and inducible nitric oxide synthase knockout mice there were no differences in plasma or urine nitrite with time regardless of the type of drinking water consumed (data not shown). Therefore, we combined the data on the 2 drinking solutions for each mouse strain. Plasma and urinary nitrite was markedly higher in C57 than in knockout mice drinking tap or L-arginine supplemented water (p ⬍0.05, tables 1 and 2). Plasma nitrite did not differ in C57 mice with unilateral ureteral obstruction for 1 or 2 weeks and unobstructed mice regardless of the type of drinking water. Moreover, plasma nitrite was significantly lower in knockout mice with obstruction than in their C57 counterparts (fig. 1). Plasma nitrite increased in knockout mice 1 week after the onset of unilateral ureteral obstruction and then returned to baseline after 2 weeks of obstruction. This pattern was noted regardless of whether tap or L-arginine supplemented water was consumed (p ⫽ 0.002 and 0.0008, respectively, fig. 1 and table 1). Bladder urine nitrite in C57 animals with unilateral ureteral obstruction did not vary with time compared with that in sham operated animals for the type of water consumed (p ⬎0.05, table 2). Bladder urine nitrite was significantly lower in knockout than in C57 mice at each time point and for each type of drinking water (p ⬍0.05, table 2). However, in knockout mice bladder urine nitrite significantly increased after 1 week of obstruction compared with that in sham operated animals. The concentration decreased to baseline after 2 weeks of obstruction. We noted this pattern whether knockout mice drank untreated or L-arginine supplemented water (p ⫽ 0.002 and 0.009, fig. 2 and table 2). Pelvic urine nitrite in C57 mice that drank tap water was lower than bladder urine nitrite in sham operated and obstructed animals (p ⬍0.05, table 2). To our knowledge there is no evident explanation for the decrease in pelvic and bladder urine nitrite in C57 mice given L-arginine drinking water for 1 week, especially in view of the increased concentrations after 2 weeks of L-arginine supplementation. Knockout mice with obstruction 2 weeks in duration that were given tap
TABLE 1. Plasma nitrite concentration Water C57 mice: Tap L-arginine Tap L-arginine Tap L-arginine Knockout mice: Tap L-arginine Tap L-arginine Tap L-arginine
Obstruction (wks.)
Mean Nitrite ⫾ SD (M.)
0 0 1 1 2 2
72.4 ⫾ 14.7 62.3 ⫾ 22.8 68.1 ⫾ 31.1 48.2 ⫾ 26.0 77.6 ⫾ 8.8 77.1 ⫾ 24.0
0 0 1 1 2 2
5.5 ⫾ 3.7 5.6 ⫾ 2.9 15.9 ⫾ 7.1 17.7 ⫾ 6.2 4.6 ⫾ 4.2 5.7 ⫾ 3.7
1278
ROLE OF NITRIC OXIDE IN OBSTRUCTIVE NEPHROPATHY
TABLE 2. Urine nitrite from the bladder of sham operated, unobstructed animals, and the bladder and kidneys of obstructed animals Water
Obstruction (wks.)
Mean Nitrite ⫾ SD (nmol./mg. creatinine)
Sham operated bladder urine C57 mice: Tap L-arginine Knockout mice: Tap L-arginine
0 0
7,911.9 ⫾ 3,758.9 5,797.5 ⫾ 3,924.3
0 0
670.8 ⫾ 221.2 560.8 ⫾ 281.9
Obstructed bladder and kidney urine C57 mice: Tap L-arginine Tap L-arginine Knockout mice: Tap L-arginine Tap L-arginine
1 1 2 2
9,219.5 ⫾ 2,634.0 3,382.0 ⫾ 668.0 6,028.0 ⫾ 1,204.1 4,935.2 ⫾ 2,433.0
1 1 2 2
1,395.6 ⫾ 478.5 1,545.0 ⫾ 834.3 324.8 ⫾ 382.4 430.6 ⫾ 197.3
1,277.5 ⫾ 1,295.0 314.0 ⫾ 150.0 1,647.0 ⫾ 1,513.1 1,263.0 ⫾ 577.8 860.7 ⫾ 241.5 811.8 ⫾ 593.7 268.9 ⫾ 102.9 471.4 ⫾ 289.5
FIG. 1. Serum nitrite concentration in M. From left to right bars represent C57 and knockout mice given tap and L-arginine supplemented drinking water, respectively. Asterisk indicates p ⬍0.005 versus sham operated (SHAM) knockout mice and those with unilateral ureteral obstruction (UUO) 2 weeks in duration.
water had significantly lower pelvic and bladder urine nitrite than sham operated mice or knockout mice with obstruction 1 week in duration (p ⫽ 0.04, fig. 3, table 2). Overall serial changes in bladder and pelvic urine nitrite in knockout mice with obstruction paralleled the alterations in plasma nitrite, that is an increase at 1 week followed by a return to baseline after 2 weeks. These alterations were most prominent in plasma and bladder urine samples. Histological analysis revealed no structural abnormalities of the right kidney in sham operated C57 or inducible nitric oxide synthase knockout mice. Trichrome staining did not show any renal fibrosis in the renal interstitium in these animals. In each mouse strain unilateral ureteral obstruction caused considerable attenuation of the renal parenchyma and widespread tubular dilatation. However, normal glomeruli and tubular structures were still evident in the cortex of C57 and knockout mice. Cortical thickness did not differ in C57 mice regardless of whether the duration of unilateral ureteral obstruction was 1 or 2 weeks. Cortical thickness was also unaffected by the type of drinking water. After 1 week of unilateral ureteral obstruction there were no differences in cortical thickness in C57 and knockout mice
FIG. 2. Bladder urine nitrite. Hatched bars represent knockout mice given tap water. Crosshatched bars represent knockout mice given L-arginine supplemented drinking water. Asterisk indicates p ⬍0.01 versus sham operated (SHAM) knockout mice and those with unilateral ureteral obstruction (UUO) 2 weeks in duration.
FIG. 3. Pelvic urine nitrite of obstructed right kidney. From left to right bars represent C57 and knockout mice given tap and L-arginine supplemented drinking water, respectively. Asterisk indicates p ⬍0.05 versus knockout mice with unilateral ureteral obstruction (UUO) 2 weeks in duration.
for either type of drinking water. However, after 2 weeks of complete ureteral obstruction the renal cortex was thinner in knockout mice that drank tap and supplemented water (p ⫽ 0.006 and 0.004, respectively, fig. 4 and table 3). Moreover, while knockout mice given tap water did not have progressive changes in cortical thickness from 1 to 2 weeks, those that drank L-arginine supplemented water had significantly more renal cortical thinning after 2 than after 1 week of obstruction (p ⫽ 0.0006, fig. 4, table 3). Differences in renal cortical thickness in the 2 strains of mice were paralleled by the intensity of trichrome staining in renal tissue specimens after 2 weeks of unilateral ureteral obstruction. Thus, in C57 versus knockout mice the mean staining score was 1.5 ⫾ 0.2 versus 0.4 ⫾ 0.1 (p ⫽ 0.002). Administering L-arginine supplemented drinking water to wild-type mice with unilateral ureteral obstruction for 2 weeks had almost no affect on trichrome stain intensity. In contrast, there was a virtual absence of fibrosis in the L-arginine treated knockout mice. Figure 5 shows the difference in renal fibrosis in C57 and knockout mice after 2 weeks of obstruction. Immunohistochemical study revealed that types I and III collagen were not present in the unobstructed left kidney of either strain. In addition, type III collagen was not detected
ROLE OF NITRIC OXIDE IN OBSTRUCTIVE NEPHROPATHY
FIG. 4. Renal cortical thickness in mm. of obstructed right kidney. From left to right bars represent C57 and knockout mice given tap and L-arginine supplemented drinking water, respectively. Asterisk indicates p ⬍0.005 versus C57 mice, 2 weeks of unilateral ureteral obstruction (UUO), and tap and L-arginine supplemented drinking water. Double asterisks indicate p ⬍0.001 versus knockout mice with unilateral ureteral obstruction 2 weeks in duration and tap water.
TABLE 3. Histological evaluation of the cortical thickness of the obstructed kidneys Water C57 mice: Tap L-arginine Tap L-arginine Knockout mice: Tap L-arginine Tap L-arginine
Obstruction (wks.)
Mean Cortical Thickness ⫾ SD (mm.)
1 1 2 2
1.69 ⫾ 0.47 1.96 ⫾ 0.23 1.75 ⫾ 0.27 1.69 ⫾ 0.46
1 1 2 2
1.43 ⫾ 0.31 1.71 ⫾ 0.29 1.15 ⫾ 0.38 0.76 ⫾ 0.24
p Value
0.0056
0.0006 0.0056 0.0037, 0.0006
in the obstructed kidney removed from wild-type or knockout mice. In contrast, after 2 weeks of obstruction type I collagen was present in all 6 and in 1 of the 5 obstructed kidneys in C57 and knockout mice, respectively (p ⫽ 0.015). This pattern of type I collagen staining was observed in animals given tap or L-arginine supplemented drinking water. Figure 6 shows immunohistochemical staining for type I collagen in a C57 and a knockout mouse. DISCUSSION
Chronic urinary tract obstruction is a common cause of end stage renal disease in childhood, accounting for 30% to 50% of all pediatric cases.11 Unilateral ureteral obstruction is a well described experimental model of tubulointerstitial fibrosis.1 Factors contributing to the severity of this process include eicosanoids, angiotensin II and transforming growth factor.3, 12, 13 Nitric oxide is involved in normal renal physiology by regulating local arteriolar tone, tubular sodium handling and mesangial cell proliferation14 as well as by modulating production of extracellular matrix proteins in vitro. In studies using cultured rat mesangial cells increased nitric oxide production led to the decreased production of collagen and fibronectin, and the increased synthesis of laminin.14 In addition, nitric oxide increases the activity of a 72 kDa. neutral metalloproteinase.15 These coordinated changes of decreased synthesis and increased degradation of matrix proteins sug-
1279
gest that nitric oxide may have an anti-fibrotic role in the kidney in health and disease. Nitric oxide has a role in various experimental renal diseases. In vivo models provide conflicting data on the effect of nitric oxide on renal injury. In acute mesangioproliferative glomerulonephritis induced by infusion of antithymocyte antibody administration of a nitric oxide synthase inhibitor decreased proteinuria, improved glomerular filtration and decreased glomerular mesangial expansion.16 In contrast, in a seven-eighths nephrectomy model of chronic renal failure in rats administration of L-arginine enhanced nitric oxide production and ameliorated kidney injury.17 Nitric oxide has been specifically evaluated in experimental models of obstructive uropathy. Administering L-arginine before the onset of unilateral ureteral obstruction resulted in nearly complete restoration of renal blood flow and glomerular filtration after obstruction resolved. Infusion of a nitric oxide synthase inhibitor before obstruction resolved resulted in complete loss of function in the affected kidney.18 In addition, L-arginine markedly decreased infiltration of macrophages into the interstitium of the obstructed kidney and improved the glomerular filtration rate.19 Morrissey et al studied the effect of enalapril and L-arginine or enalapril and n-nitro-L-arginine methyl ester, an inhibitor of nitric oxide synthase, in rats with unilateral ureteral obstruction 5 days in duration. 6 They evaluated numerous parameters, including interstitial volume, macrophage infiltration, collagen IV and smooth muscle actin expression as well as messenger RNA levels of transforming growth factor-, type IV collagen and tissue inhibitor of metalloproteinase-1. Animals treated with enalapril or L-arginine had higher urinary nitric oxide and significantly decreased interstitial volume. These data suggest that increased nitric oxide production has a beneficial effect by decreasing tubulointerstitial fibrosis after the onset of unilateral ureteral obstruction. We demonstrated that renal cortical thinning after the onset of unilateral ureteral obstruction was more severe in inducible nitric oxide synthase knockout than in C57 mice. Less extensive cortical thinning in C57 mice was associated with an increased intensity of trichrome staining and type I collagen deposition in the tubulointerstitium. These findings indicate that greater cortical thinning in knockout mice reflects decreased renal fibrosis and implies that attenuated renal scarring was not the consequence of nitric oxide synthesized by inducible nitric oxide synthase. The increases in serum, and bladder and obstructed renal pelvic urine nitrite after 1 week of unilateral ureteral obstruction in knockout mice must have been due to endothelial or neuronal nitric oxide synthase. The increments in serum and urine nitrite levels in knockout mice with unilateral ureteral obstruction were transient and returned to baseline by 2 weeks. However, we speculate that after 1 week increased nitric oxide produced by endothelial or neuronal nitric oxide synthase caused a decrease in the net renal accumulation of extracellular matrix materials and the worsening of cortical thinning after 2 weeks of obstruction in knockout mice. Exaggerated cortical thinning and decreased fibrosis after 2 weeks of obstruction in knockout mice given L-arginine supplemented drinking water are also consistent with the hypothesis that noninducible nitric oxide synthase derived nitric oxide is more important for down regulating renal fibrosis during prolonged ureteral obstruction. A preliminary report describes the effect of unilateral ureteral obstruction in inducible nitric oxide synthase knockout mice. In apparent contradiction to our findings, Hochberg et al noted increased interstitial volume expressed as a percent of parenchymal volume in knockout mice with unilateral ureteral obstruction.20 However, it is conceivable that absolute cortical and medullary volume was decreased compared with that in wild-type mice. In addition, differences in age may explain the seemingly discrepant response to unilateral
1280
ROLE OF NITRIC OXIDE IN OBSTRUCTIVE NEPHROPATHY
FIG. 5. Light micrography of mice with unilateral ureteral obstruction 2 weeks in duration. Trichrome stain, reduced from ⫻225. a, in C57 mouse note area of trichrome stain and interstitial fibrosis (arrow). b, knockout mouse.
FIG. 6. Light micrography of type I collagen in mice with unilateral ureteral obstruction 2 weeks in duration. Immunohistochemical staining, reduced from ⫻225. a, in C57 mouse note area of type I collagen deposition in interstitium (arrow). b, knockout mouse.
ureteral obstruction in our study and that of Hochberg et al. Finally, Hochberg et al did not assess the impact of L-arginine supplementation on the severity of renal fibrosis in inducible nitric oxide synthase knockout and wild-type mice with unilateral ureteral obstruction. The genetic deletion of inducible nitric oxide synthase in knockout mice may have affected the expression and activity of the other 2 isoforms of the enzyme. There is evidence of an inverse correlation in inducible and endothelial nitric oxide synthase activity. In a rat model of septic shock Schwartz et al noted that high levels of nitric oxide production due to cytokine induced stimulation of inducible nitric oxide synthase were associated with decreased endothelial nitric oxide synthase activity and selective inhibition of endothelial nitric oxide synthase activity restored inducible nitric oxide synthase.21 In our study of inducible nitric oxide synthase knockout mice complete absence of inducible nitric oxide synthase derived nitric oxide synthesis may have led to a compensatory increase in endothelial or neuronal nitric oxide synthase activity, which in turn had a protective and anti-fibrotic effect on kidneys in the obstructed model. This response may have been amplified when supplemental L-arginine was given to knockout mice. In contrast, L-arginine had no discernible effect on renal fibrosis in C57
mice. The continued usefulness of inducible nitric oxide synthase knockout mice is confirmed by the recent study of Ling et al, which documented that these mice are less susceptible to ischemia reperfusion renal injury.22 Further study to evaluate the specific role of endothelial nitric oxide synthase in unilateral ureteral obstruction is currently under way at our laboratory using endothelial nitric oxide synthase knockout mice. CONCLUSIONS
The homozygous deficiency of inducible nitric oxide synthase leads to significantly lower levels of nitrite in the blood and urine of knockout mice. After 2 weeks of complete unilateral ureteral obstruction mice lacking inducible nitric oxide synthase had a significantly greater degree of renal cortical thinning, less tubulointerstitial fibrosis and virtual absence of type I collagen deposition in the kidney compared with findings in wild-type C57 mice. These histopathological changes were preceded by transient elevation in nitrite in serum and urine that was detected after 1 week of unilateral ureteral obstruction but returned to baseline after 2 weeks. The administration of L-arginine exaggerated the loss of renal parenchyma in knockout mice after the onset of unilat-
ROLE OF NITRIC OXIDE IN OBSTRUCTIVE NEPHROPATHY
eral ureteral obstruction. Based on these data we speculate that endothelial or neuronal nitric oxide synthase may have a more significant role than inducible nitric oxide synthase for modulating renal fibrosis resulting from obstructive nephropathy. Dr. John Mudgett provided the original control wild-type mice C57BL/6 breeding pair and Albert Tarectecan performed the immunohistochemical studies. REFERENCES
1. Klahr, S.: New insights into the consequences and mechanisms of renal impairment in obstructive nephropathy. Am J Kidney Dis, 18: 689, 1991 2. Klahr, S., Harris, K. and Purkerson, M. L.: Effects of obstruction on renal functions. Pediatr Nephrol, 2: 34, 1988 3. Ishidoya, S., Morrissey, J., McCracken, R. et al: Angiotensin II receptor antagonist ameliorates renal tubulointerstitial fibrosis caused by unilateral ureteral obstruction. Kidney Int, 47: 1285, 1995 4. Wright, E. J., McCaffrey, T. A., Robertson, A. P. et al: Chronic unilateral ureteral obstruction is associated with interstitial fibrosis and tubular expression of transforming growth factorbeta. Lab Invest, 74: 528, 1996 5. Klahr, S. and Morrissey, J.: Editorial: renal disease: the two faces of nitric oxide. Lab Invest, 72: 1, 1995 6. Morrissey, J. J., Ishidoya, S., McCracken, R. et al: Nitric oxide generation ameliorates the tubulointerstitial fibrosis of obstructive nephropathy. J Am Soc Nephrol, 7: 2202, 1996 7. Nathan, C.: Nitric oxide as a secretory product of mammalian cells. FASEB J, 6: 3051, 1992 8. Shultz, P. J., Archer, S. L. and Rosenberg, M. E.: Inducible nitric oxide synthase mRNA and activity in glomerular mesangial cells. Kidney Int, 46: 683, 1994 9. Markewitz, B. A., Michael, J. R. and Kohan, D. E.: Cytokineinduced expression of nitric oxide synthase in rat renal tubule cells. J Clin Invest, 91: 2138, 1993 10. Levine, J. J., Pettei, M. J., Valderrama, E. et al: Nitric oxide and inflammatory bowel disease: evidence for local intestinal production in children with active colonic disease. J Pediatr Gastroenterol Nutr, 26: 34, 1998
1281
11. Foreman, J. W. and Chan, J. C.: Chronic renal failure in infants and children. J Pediatr, 113: 793, 1988 12. Kaneto, H., Morrissey, J. and Klahr, S.: Increased expression of TGF-beta1 mRNA in the obstructed kidney of rats with unilateral ureteral ligation. Kidney Int, 44: 313, 1993 13. Kaneto, H., Morrissey, J., McCracken, R. et al: Enalapril reduces collagen type IV synthesis and expansion of the interstitium in the obstructed rat kidney. Kidney Int, 45: 1637, 1994 14. Trachtman, H., Futterweit, S. and Singhal, P.: Nitric oxide modulates the synthesis of extracellular matrix proteins in cultured rat mesangial cells. Biochem Biophys Res Commun, 207: 120, 1995 15. Trachtman, H., Futterweit, S., Garg, P. et al: Nitric oxide stimulates the activity of a 72-kDa neutral matrix metalloproteinase in cultured rat mesangial cells. Biochem Biophys Res Commun, 218: 704, 1996 16. Narita, I., Border, W. A., Ketteler, M. et al: Nitric oxide mediates immunologic injury to kidney mesangium in experimental glomerulonephritis. Lab Invest, 72: 17, 1995 17. Reyes, A. A., Purkerson, M. L., Karl, I. et al: Dietary supplementation with L-arginine ameliorates the progression of renal disease in rats with subtotal nephrectomy. Am J Kidney Dis, 20: 168, 1992 18. Reyes, A. A., Martin, D., Settle, S. et al: EDRF role in renal function and blood pressure of normal rats and rats with obstructive uropathy. Kidney Int, 41: 403, 1992 19. Reyes, A. A., Porras, B. H., Chasalow, F. I. et al: L-arginine decreased the infiltration of the kidney by macrophages in obstructive nephropathy and puromycin-induced nephrosis. Kidney Int, 45: 1346, 1994 20. Hochberg, D. A., Chen, J., Vaughn, E. D., Jr. et al: Interstitial fibrosis is exacerbated in unilateral ureteral obstruction (UUO) in mice lacking the gene for inducible nitric oxide synthase (iNOS). J Urol, suppl., 68: 68, abstract 257, 1998 21. Schwartz, D., Mendonca, M., Schwartz, I. et al: Inhibition of constitutive nitric oxide synthase (NOS) by nitric oxide generated by inducible NOS after lipopolysaccharide administration provokes renal dysfunction in rats. J Clin Invest, 100: 439, 1997 22. Ling, H., Edelstein, C., Gengaro, P. et al: Attenuation of renal ischemia-reperfusuion injury in inducible nitric oxide synthase knockout mice. Amer J Physiol, suppl., 277: F383, 1999
Vol. 163, 1276 –1281, April 2000 Printed in U.S.A.
THE ROLE OF NITRIC OXIDE IN OBSTRUCTIVE NEPHROPATHY ANDREW HUANG, LANE S. PALMER, DAVID HOM, ELSA VALDERRAMA AND HOWARD TRACHTMAN From the Divisions of Pediatric Urology, Pathology and Nephrology, Schneider Children’s Hospital, Long Island Jewish Medical Center, New Hyde Park, New York
ABSTRACT
Purpose: Ureteral obstruction leads to tubulointerstitial fibrosis and loss of renal function. Nitric oxide production ameliorates fibrosis due to obstructive uropathy. However, nitric oxide is produced by 3 isoforms of the enzyme, nitric oxide synthase. We evaluated the role of inducible nitric oxide synthase in obstructive uropathy using nitric oxide synthase knockout mice, and determined whether the administration of L-arginine to promote nitric oxide synthesis by alternative nitric oxide synthase isoforms modulates renal fibrosis in these animals. Materials and Methods: Complete unilateral ureteral obstruction was created in wild-type C57 and inducible nitric oxide synthase knockout mice. Control animals of each strain underwent sham surgery. Throughout the experiment mice had free access to untreated tap water or water supplemented with 10 gm./l. L-arginine. Animals were sacrificed 1 and 2 weeks, respectively, after creation of unilateral ureteral obstruction. We obtained serum as well as bladder and obstructed renal pelvic urine, and determined the nitrite level in each fluid. Renal cortical thickness was measured in the normal and obstructed kidneys. The degree of tubulointerstitial fibrosis was evaluated by trichrome staining and type I collagen deposition in kidney tissue specimens. Results: Nitrite was significantly decreased in the serum, bladder and renal pelvic urine of inducible nitric oxide synthase knockout mice with unilateral ureteral obstruction compared with that in wild-type C57 mice at 1 and 2 weeks (p ⬍0.05). In knockout mice with unilateral ureteral obstruction 1 week in duration that drank tap or L-arginine supplemented water nitrite in serum and each urine sample was higher than in sham operated knockout controls. The level returned to baseline after 2 weeks of obstruction (p ⬍0.05). After 2 weeks of obstruction there was significantly greater cortical thinning in knockout than in C57 mice (p ⬍0.05). Moreover, knockout mice given L-arginine supplemented water for 2 weeks had even greater cortical thinning than after 1 week or than mice given tap water for 1 to 2 weeks (p ⬍0.05). Decreased renal cortical thickness in knockout mice after 2 weeks of obstruction was associated with less intense trichrome staining and a virtual absence of type I collagen deposition compared with findings in the wild-type C57 strain. Conclusions: Inducible nitric oxide synthase knockout mice with unilateral ureteral obstruction have significantly lower nitrite in serum and urine than wild-type C57 mice. Knockout mice also have more severe renal cortical thinning than C57 animals after creation of unilateral ureteral obstruction. Providing L-arginine supplemented water to inducible nitric oxide synthase knockout mice exacerbates the loss of cortical thickness. Alterations in cortical thinning that we observed in knockout mice were associated with decreased tubulointerstitial fibrosis and a decreased net renal extracellular matrix accumulation. These data indicate that endothelial or neuronal nitric oxide synthase may be more important than inducible nitric oxide synthase for modulating renal fibrosis in obstructive uropathy. KEY WORDS: kidney, nitric oxide, ureteral obstruction
Obstructive uropathy in children is usually the result of a congenital structural or neurological lesion of the urinary system. In severe cases this condition leads to progressive glomerulosclerosis, tubulointerstitial fibrosis and loss of renal function. Complete unilateral ureteral obstruction is a well described experimental model of this renal disease.1 Numerous factors have been implicated in the chronic renal injury that develops in this disorder, including eicosanoids, angiotensin II, transforming growth factor- and nitric oxide.2– 4 Inhibition of nitric oxide synthesis exacerbates and enhanced production of nitric oxide ameliorates fibrosis in exAccepted for publication November 12, 1999. Supported by a grant from the Schneider family.
perimental obstructive uropathy.5, 6 Nitric oxide is synthesized from the guanidino nitrogen of L-arginine by nitric oxide synthase.5 However, there are 3 distinct isoforms of nitric oxide synthase in mammals.7 Endothelial and neuronal nitric oxide synthase are the 2 forms of the enzyme expressed constitutively in the basal state that are present in endothelial and neuronal cells, respectively. The other form, inducible nitric oxide synthase, is present in various cells, including glomerular mesangial and tubular epithelial cells.8, 9 While activation of endothelial or neuronal nitric oxide synthase leads to slight transient increases in nitric oxide production, stimulation of inducible nitric oxide synthase leads to marked and sustained elevations in nitric oxide synthesis. Because of the importance of nitric oxide in pro-
1276
1277
ROLE OF NITRIC OXIDE IN OBSTRUCTIVE NEPHROPATHY
gressive renal disease and the complex nature of nitric oxide production, we performed experiments with unilateral ureteral obstruction in wild-type C57 and inducible nitric oxide synthase knockout mice to determine the specific role of inducible nitric oxide synthase in modulating fibrosis due to obstructive uropathy. In addition, we assessed the impact of endothelial nitric oxide synthase and neuronal nitric oxide synthase in the fibrotic process by providing supplemental L-arginine, the precursor to nitric oxide, to these 2 strains of mouse. MATERIALS AND METHODS
Control wild-type C57BL/6 mice were provided and inducible nitric oxide synthase knockout mice were bred from homozygous animals for the study. All mice were fed standard rodent chow. At age 9 weeks they were anesthetized with an intraperitoneal injection of 50 mg./kg. pentobarbital. The right colon was reflected medially. The proximal right ureter was dissected free and ligated with 4-zero silk ties. Control mice underwent sham surgery without ureteral ligation. No mice died as a result of operative complications at the time of right ureteral obstruction creation or sham surgery. Mice were given free access to untreated tap water or water supplemented with 10 gm./l. L-arginine. L-arginine drinking water was started 24 hours preoperatively and continued until sacrifice. There were 5 to 12 mice in each experimental group. Animals were anesthetized with an intraperitoneal injection of 50 mg./kg. pentobarbital 1 and 2 weeks after the onset of unilateral ureteral obstruction or sham surgery. A blood sample was obtained, and urine was collected from the bladder and obstructed renal pelvis. The kidneys were perfused with cold Hanks balanced salt solution, removed and incised coronally. Sections were fixed by immersion in 10% formalin. To determine the nitrite level blood samples were centrifuged at 1,000 rpm for 5 minutes and the serum was removed. Plasma and urine samples were tested for nitrite using the Greiss reaction.10 Briefly, serum samples were deproteinized by adding a 1/20 volume of zinc sulfate and then diluted 1:1 with distilled water. Urine samples were diluted 1:6 with distilled water. Samples were incubated for 10 minutes with an equal volume of Greiss reagent (1% sulfanilamide, 0.1% naphthylethylene diamine dihydrochloride and 2.5% phosphoric acid). Absorbance was measured in a spectrophotometer at 550 nm. A standard curve with known nitrite concentrations of 1 to 50 m. was included in each assay. Serum and urine samples were assayed for urea nitrogen and creatinine, respectively, using an automated analyzer. Urinary nitrite was expressed as nmol./mg. creatinine. For histological evaluation tissues were embedded in paraffin, and 4 . sections were stained with hematoxylin and eosin, and Masson’s trichrome. Cortical thickness was measured in the region of the kidney with the best preservation of renal parenchymal structures. Specimens were examined histologically for interstitial fibrosis using the trichrome stain and graded on a scale of 0 —foci of fibrosis in less than 10% of the specimen, 1—10% to 25%, 2—26% to 50%, 3—51% to 75%, 4 — greater than 75%. For immunohistochemical study primary antibodies to types I (No. AB765) and III (No. AB747) collagen were used in a 1:50 dilution and incubated with the renal sections for 32 minutes. After exposure to 1:10 dilution of a solution to unmask tissue antigens for 40 minutes at 40C in a water bath, slides were rinsed with water and stained using an automated immunoperoxidase stainer. Sections were counterstained with hematoxylin and 1.5% bluing slide ammonia in 70% ethanol. They were then dehydrated and a coverglass was applied. A single observer (E. V.) blinded to treatment groups evaluated all histopathological and immunohisto-
chemical slides. Data are shown as the mean plus or minus standard deviation. We used analysis of variance and the t test to compare findings in each group at the various time points for the various fluids administered and findings in the 2 mouse strains. RESULTS
There was no acute mortality during the surgical procedure. Fewer than 5% of the mice of either strain died more than 24 hours after the onset of unilateral ureteral obstruction. Blood urea nitrogen ranged from 29 to 59 mg./dl. As expected, because ureteral obstruction was unilateral, there were no differences in either group in blood urea nitrogen regardless of the type of drinking water provided, or whether mice were sacrificed 1 or 2 weeks postoperatively (data not shown). In sham operated C57 and inducible nitric oxide synthase knockout mice there were no differences in plasma or urine nitrite with time regardless of the type of drinking water consumed (data not shown). Therefore, we combined the data on the 2 drinking solutions for each mouse strain. Plasma and urinary nitrite was markedly higher in C57 than in knockout mice drinking tap or L-arginine supplemented water (p ⬍0.05, tables 1 and 2). Plasma nitrite did not differ in C57 mice with unilateral ureteral obstruction for 1 or 2 weeks and unobstructed mice regardless of the type of drinking water. Moreover, plasma nitrite was significantly lower in knockout mice with obstruction than in their C57 counterparts (fig. 1). Plasma nitrite increased in knockout mice 1 week after the onset of unilateral ureteral obstruction and then returned to baseline after 2 weeks of obstruction. This pattern was noted regardless of whether tap or L-arginine supplemented water was consumed (p ⫽ 0.002 and 0.0008, respectively, fig. 1 and table 1). Bladder urine nitrite in C57 animals with unilateral ureteral obstruction did not vary with time compared with that in sham operated animals for the type of water consumed (p ⬎0.05, table 2). Bladder urine nitrite was significantly lower in knockout than in C57 mice at each time point and for each type of drinking water (p ⬍0.05, table 2). However, in knockout mice bladder urine nitrite significantly increased after 1 week of obstruction compared with that in sham operated animals. The concentration decreased to baseline after 2 weeks of obstruction. We noted this pattern whether knockout mice drank untreated or L-arginine supplemented water (p ⫽ 0.002 and 0.009, fig. 2 and table 2). Pelvic urine nitrite in C57 mice that drank tap water was lower than bladder urine nitrite in sham operated and obstructed animals (p ⬍0.05, table 2). To our knowledge there is no evident explanation for the decrease in pelvic and bladder urine nitrite in C57 mice given L-arginine drinking water for 1 week, especially in view of the increased concentrations after 2 weeks of L-arginine supplementation. Knockout mice with obstruction 2 weeks in duration that were given tap
TABLE 1. Plasma nitrite concentration Water C57 mice: Tap L-arginine Tap L-arginine Tap L-arginine Knockout mice: Tap L-arginine Tap L-arginine Tap L-arginine
Obstruction (wks.)
Mean Nitrite ⫾ SD (M.)
0 0 1 1 2 2
72.4 ⫾ 14.7 62.3 ⫾ 22.8 68.1 ⫾ 31.1 48.2 ⫾ 26.0 77.6 ⫾ 8.8 77.1 ⫾ 24.0
0 0 1 1 2 2
5.5 ⫾ 3.7 5.6 ⫾ 2.9 15.9 ⫾ 7.1 17.7 ⫾ 6.2 4.6 ⫾ 4.2 5.7 ⫾ 3.7
1278
ROLE OF NITRIC OXIDE IN OBSTRUCTIVE NEPHROPATHY
TABLE 2. Urine nitrite from the bladder of sham operated, unobstructed animals, and the bladder and kidneys of obstructed animals Water
Obstruction (wks.)
Mean Nitrite ⫾ SD (nmol./mg. creatinine)
Sham operated bladder urine C57 mice: Tap L-arginine Knockout mice: Tap L-arginine
0 0
7,911.9 ⫾ 3,758.9 5,797.5 ⫾ 3,924.3
0 0
670.8 ⫾ 221.2 560.8 ⫾ 281.9
Obstructed bladder and kidney urine C57 mice: Tap L-arginine Tap L-arginine Knockout mice: Tap L-arginine Tap L-arginine
1 1 2 2
9,219.5 ⫾ 2,634.0 3,382.0 ⫾ 668.0 6,028.0 ⫾ 1,204.1 4,935.2 ⫾ 2,433.0
1 1 2 2
1,395.6 ⫾ 478.5 1,545.0 ⫾ 834.3 324.8 ⫾ 382.4 430.6 ⫾ 197.3
1,277.5 ⫾ 1,295.0 314.0 ⫾ 150.0 1,647.0 ⫾ 1,513.1 1,263.0 ⫾ 577.8 860.7 ⫾ 241.5 811.8 ⫾ 593.7 268.9 ⫾ 102.9 471.4 ⫾ 289.5
FIG. 1. Serum nitrite concentration in M. From left to right bars represent C57 and knockout mice given tap and L-arginine supplemented drinking water, respectively. Asterisk indicates p ⬍0.005 versus sham operated (SHAM) knockout mice and those with unilateral ureteral obstruction (UUO) 2 weeks in duration.
water had significantly lower pelvic and bladder urine nitrite than sham operated mice or knockout mice with obstruction 1 week in duration (p ⫽ 0.04, fig. 3, table 2). Overall serial changes in bladder and pelvic urine nitrite in knockout mice with obstruction paralleled the alterations in plasma nitrite, that is an increase at 1 week followed by a return to baseline after 2 weeks. These alterations were most prominent in plasma and bladder urine samples. Histological analysis revealed no structural abnormalities of the right kidney in sham operated C57 or inducible nitric oxide synthase knockout mice. Trichrome staining did not show any renal fibrosis in the renal interstitium in these animals. In each mouse strain unilateral ureteral obstruction caused considerable attenuation of the renal parenchyma and widespread tubular dilatation. However, normal glomeruli and tubular structures were still evident in the cortex of C57 and knockout mice. Cortical thickness did not differ in C57 mice regardless of whether the duration of unilateral ureteral obstruction was 1 or 2 weeks. Cortical thickness was also unaffected by the type of drinking water. After 1 week of unilateral ureteral obstruction there were no differences in cortical thickness in C57 and knockout mice
FIG. 2. Bladder urine nitrite. Hatched bars represent knockout mice given tap water. Crosshatched bars represent knockout mice given L-arginine supplemented drinking water. Asterisk indicates p ⬍0.01 versus sham operated (SHAM) knockout mice and those with unilateral ureteral obstruction (UUO) 2 weeks in duration.
FIG. 3. Pelvic urine nitrite of obstructed right kidney. From left to right bars represent C57 and knockout mice given tap and L-arginine supplemented drinking water, respectively. Asterisk indicates p ⬍0.05 versus knockout mice with unilateral ureteral obstruction (UUO) 2 weeks in duration.
for either type of drinking water. However, after 2 weeks of complete ureteral obstruction the renal cortex was thinner in knockout mice that drank tap and supplemented water (p ⫽ 0.006 and 0.004, respectively, fig. 4 and table 3). Moreover, while knockout mice given tap water did not have progressive changes in cortical thickness from 1 to 2 weeks, those that drank L-arginine supplemented water had significantly more renal cortical thinning after 2 than after 1 week of obstruction (p ⫽ 0.0006, fig. 4, table 3). Differences in renal cortical thickness in the 2 strains of mice were paralleled by the intensity of trichrome staining in renal tissue specimens after 2 weeks of unilateral ureteral obstruction. Thus, in C57 versus knockout mice the mean staining score was 1.5 ⫾ 0.2 versus 0.4 ⫾ 0.1 (p ⫽ 0.002). Administering L-arginine supplemented drinking water to wild-type mice with unilateral ureteral obstruction for 2 weeks had almost no affect on trichrome stain intensity. In contrast, there was a virtual absence of fibrosis in the L-arginine treated knockout mice. Figure 5 shows the difference in renal fibrosis in C57 and knockout mice after 2 weeks of obstruction. Immunohistochemical study revealed that types I and III collagen were not present in the unobstructed left kidney of either strain. In addition, type III collagen was not detected
ROLE OF NITRIC OXIDE IN OBSTRUCTIVE NEPHROPATHY
FIG. 4. Renal cortical thickness in mm. of obstructed right kidney. From left to right bars represent C57 and knockout mice given tap and L-arginine supplemented drinking water, respectively. Asterisk indicates p ⬍0.005 versus C57 mice, 2 weeks of unilateral ureteral obstruction (UUO), and tap and L-arginine supplemented drinking water. Double asterisks indicate p ⬍0.001 versus knockout mice with unilateral ureteral obstruction 2 weeks in duration and tap water.
TABLE 3. Histological evaluation of the cortical thickness of the obstructed kidneys Water C57 mice: Tap L-arginine Tap L-arginine Knockout mice: Tap L-arginine Tap L-arginine
Obstruction (wks.)
Mean Cortical Thickness ⫾ SD (mm.)
1 1 2 2
1.69 ⫾ 0.47 1.96 ⫾ 0.23 1.75 ⫾ 0.27 1.69 ⫾ 0.46
1 1 2 2
1.43 ⫾ 0.31 1.71 ⫾ 0.29 1.15 ⫾ 0.38 0.76 ⫾ 0.24
p Value
0.0056
0.0006 0.0056 0.0037, 0.0006
in the obstructed kidney removed from wild-type or knockout mice. In contrast, after 2 weeks of obstruction type I collagen was present in all 6 and in 1 of the 5 obstructed kidneys in C57 and knockout mice, respectively (p ⫽ 0.015). This pattern of type I collagen staining was observed in animals given tap or L-arginine supplemented drinking water. Figure 6 shows immunohistochemical staining for type I collagen in a C57 and a knockout mouse. DISCUSSION
Chronic urinary tract obstruction is a common cause of end stage renal disease in childhood, accounting for 30% to 50% of all pediatric cases.11 Unilateral ureteral obstruction is a well described experimental model of tubulointerstitial fibrosis.1 Factors contributing to the severity of this process include eicosanoids, angiotensin II and transforming growth factor.3, 12, 13 Nitric oxide is involved in normal renal physiology by regulating local arteriolar tone, tubular sodium handling and mesangial cell proliferation14 as well as by modulating production of extracellular matrix proteins in vitro. In studies using cultured rat mesangial cells increased nitric oxide production led to the decreased production of collagen and fibronectin, and the increased synthesis of laminin.14 In addition, nitric oxide increases the activity of a 72 kDa. neutral metalloproteinase.15 These coordinated changes of decreased synthesis and increased degradation of matrix proteins sug-
1279
gest that nitric oxide may have an anti-fibrotic role in the kidney in health and disease. Nitric oxide has a role in various experimental renal diseases. In vivo models provide conflicting data on the effect of nitric oxide on renal injury. In acute mesangioproliferative glomerulonephritis induced by infusion of antithymocyte antibody administration of a nitric oxide synthase inhibitor decreased proteinuria, improved glomerular filtration and decreased glomerular mesangial expansion.16 In contrast, in a seven-eighths nephrectomy model of chronic renal failure in rats administration of L-arginine enhanced nitric oxide production and ameliorated kidney injury.17 Nitric oxide has been specifically evaluated in experimental models of obstructive uropathy. Administering L-arginine before the onset of unilateral ureteral obstruction resulted in nearly complete restoration of renal blood flow and glomerular filtration after obstruction resolved. Infusion of a nitric oxide synthase inhibitor before obstruction resolved resulted in complete loss of function in the affected kidney.18 In addition, L-arginine markedly decreased infiltration of macrophages into the interstitium of the obstructed kidney and improved the glomerular filtration rate.19 Morrissey et al studied the effect of enalapril and L-arginine or enalapril and n-nitro-L-arginine methyl ester, an inhibitor of nitric oxide synthase, in rats with unilateral ureteral obstruction 5 days in duration. 6 They evaluated numerous parameters, including interstitial volume, macrophage infiltration, collagen IV and smooth muscle actin expression as well as messenger RNA levels of transforming growth factor-, type IV collagen and tissue inhibitor of metalloproteinase-1. Animals treated with enalapril or L-arginine had higher urinary nitric oxide and significantly decreased interstitial volume. These data suggest that increased nitric oxide production has a beneficial effect by decreasing tubulointerstitial fibrosis after the onset of unilateral ureteral obstruction. We demonstrated that renal cortical thinning after the onset of unilateral ureteral obstruction was more severe in inducible nitric oxide synthase knockout than in C57 mice. Less extensive cortical thinning in C57 mice was associated with an increased intensity of trichrome staining and type I collagen deposition in the tubulointerstitium. These findings indicate that greater cortical thinning in knockout mice reflects decreased renal fibrosis and implies that attenuated renal scarring was not the consequence of nitric oxide synthesized by inducible nitric oxide synthase. The increases in serum, and bladder and obstructed renal pelvic urine nitrite after 1 week of unilateral ureteral obstruction in knockout mice must have been due to endothelial or neuronal nitric oxide synthase. The increments in serum and urine nitrite levels in knockout mice with unilateral ureteral obstruction were transient and returned to baseline by 2 weeks. However, we speculate that after 1 week increased nitric oxide produced by endothelial or neuronal nitric oxide synthase caused a decrease in the net renal accumulation of extracellular matrix materials and the worsening of cortical thinning after 2 weeks of obstruction in knockout mice. Exaggerated cortical thinning and decreased fibrosis after 2 weeks of obstruction in knockout mice given L-arginine supplemented drinking water are also consistent with the hypothesis that noninducible nitric oxide synthase derived nitric oxide is more important for down regulating renal fibrosis during prolonged ureteral obstruction. A preliminary report describes the effect of unilateral ureteral obstruction in inducible nitric oxide synthase knockout mice. In apparent contradiction to our findings, Hochberg et al noted increased interstitial volume expressed as a percent of parenchymal volume in knockout mice with unilateral ureteral obstruction.20 However, it is conceivable that absolute cortical and medullary volume was decreased compared with that in wild-type mice. In addition, differences in age may explain the seemingly discrepant response to unilateral
1280
ROLE OF NITRIC OXIDE IN OBSTRUCTIVE NEPHROPATHY
FIG. 5. Light micrography of mice with unilateral ureteral obstruction 2 weeks in duration. Trichrome stain, reduced from ⫻225. a, in C57 mouse note area of trichrome stain and interstitial fibrosis (arrow). b, knockout mouse.
FIG. 6. Light micrography of type I collagen in mice with unilateral ureteral obstruction 2 weeks in duration. Immunohistochemical staining, reduced from ⫻225. a, in C57 mouse note area of type I collagen deposition in interstitium (arrow). b, knockout mouse.
ureteral obstruction in our study and that of Hochberg et al. Finally, Hochberg et al did not assess the impact of L-arginine supplementation on the severity of renal fibrosis in inducible nitric oxide synthase knockout and wild-type mice with unilateral ureteral obstruction. The genetic deletion of inducible nitric oxide synthase in knockout mice may have affected the expression and activity of the other 2 isoforms of the enzyme. There is evidence of an inverse correlation in inducible and endothelial nitric oxide synthase activity. In a rat model of septic shock Schwartz et al noted that high levels of nitric oxide production due to cytokine induced stimulation of inducible nitric oxide synthase were associated with decreased endothelial nitric oxide synthase activity and selective inhibition of endothelial nitric oxide synthase activity restored inducible nitric oxide synthase.21 In our study of inducible nitric oxide synthase knockout mice complete absence of inducible nitric oxide synthase derived nitric oxide synthesis may have led to a compensatory increase in endothelial or neuronal nitric oxide synthase activity, which in turn had a protective and anti-fibrotic effect on kidneys in the obstructed model. This response may have been amplified when supplemental L-arginine was given to knockout mice. In contrast, L-arginine had no discernible effect on renal fibrosis in C57
mice. The continued usefulness of inducible nitric oxide synthase knockout mice is confirmed by the recent study of Ling et al, which documented that these mice are less susceptible to ischemia reperfusion renal injury.22 Further study to evaluate the specific role of endothelial nitric oxide synthase in unilateral ureteral obstruction is currently under way at our laboratory using endothelial nitric oxide synthase knockout mice. CONCLUSIONS
The homozygous deficiency of inducible nitric oxide synthase leads to significantly lower levels of nitrite in the blood and urine of knockout mice. After 2 weeks of complete unilateral ureteral obstruction mice lacking inducible nitric oxide synthase had a significantly greater degree of renal cortical thinning, less tubulointerstitial fibrosis and virtual absence of type I collagen deposition in the kidney compared with findings in wild-type C57 mice. These histopathological changes were preceded by transient elevation in nitrite in serum and urine that was detected after 1 week of unilateral ureteral obstruction but returned to baseline after 2 weeks. The administration of L-arginine exaggerated the loss of renal parenchyma in knockout mice after the onset of unilat-
ROLE OF NITRIC OXIDE IN OBSTRUCTIVE NEPHROPATHY
eral ureteral obstruction. Based on these data we speculate that endothelial or neuronal nitric oxide synthase may have a more significant role than inducible nitric oxide synthase for modulating renal fibrosis resulting from obstructive nephropathy. Dr. John Mudgett provided the original control wild-type mice C57BL/6 breeding pair and Albert Tarectecan performed the immunohistochemical studies. REFERENCES
1. Klahr, S.: New insights into the consequences and mechanisms of renal impairment in obstructive nephropathy. Am J Kidney Dis, 18: 689, 1991 2. Klahr, S., Harris, K. and Purkerson, M. L.: Effects of obstruction on renal functions. Pediatr Nephrol, 2: 34, 1988 3. Ishidoya, S., Morrissey, J., McCracken, R. et al: Angiotensin II receptor antagonist ameliorates renal tubulointerstitial fibrosis caused by unilateral ureteral obstruction. Kidney Int, 47: 1285, 1995 4. Wright, E. J., McCaffrey, T. A., Robertson, A. P. et al: Chronic unilateral ureteral obstruction is associated with interstitial fibrosis and tubular expression of transforming growth factorbeta. Lab Invest, 74: 528, 1996 5. Klahr, S. and Morrissey, J.: Editorial: renal disease: the two faces of nitric oxide. Lab Invest, 72: 1, 1995 6. Morrissey, J. J., Ishidoya, S., McCracken, R. et al: Nitric oxide generation ameliorates the tubulointerstitial fibrosis of obstructive nephropathy. J Am Soc Nephrol, 7: 2202, 1996 7. Nathan, C.: Nitric oxide as a secretory product of mammalian cells. FASEB J, 6: 3051, 1992 8. Shultz, P. J., Archer, S. L. and Rosenberg, M. E.: Inducible nitric oxide synthase mRNA and activity in glomerular mesangial cells. Kidney Int, 46: 683, 1994 9. Markewitz, B. A., Michael, J. R. and Kohan, D. E.: Cytokineinduced expression of nitric oxide synthase in rat renal tubule cells. J Clin Invest, 91: 2138, 1993 10. Levine, J. J., Pettei, M. J., Valderrama, E. et al: Nitric oxide and inflammatory bowel disease: evidence for local intestinal production in children with active colonic disease. J Pediatr Gastroenterol Nutr, 26: 34, 1998
1281
11. Foreman, J. W. and Chan, J. C.: Chronic renal failure in infants and children. J Pediatr, 113: 793, 1988 12. Kaneto, H., Morrissey, J. and Klahr, S.: Increased expression of TGF-beta1 mRNA in the obstructed kidney of rats with unilateral ureteral ligation. Kidney Int, 44: 313, 1993 13. Kaneto, H., Morrissey, J., McCracken, R. et al: Enalapril reduces collagen type IV synthesis and expansion of the interstitium in the obstructed rat kidney. Kidney Int, 45: 1637, 1994 14. Trachtman, H., Futterweit, S. and Singhal, P.: Nitric oxide modulates the synthesis of extracellular matrix proteins in cultured rat mesangial cells. Biochem Biophys Res Commun, 207: 120, 1995 15. Trachtman, H., Futterweit, S., Garg, P. et al: Nitric oxide stimulates the activity of a 72-kDa neutral matrix metalloproteinase in cultured rat mesangial cells. Biochem Biophys Res Commun, 218: 704, 1996 16. Narita, I., Border, W. A., Ketteler, M. et al: Nitric oxide mediates immunologic injury to kidney mesangium in experimental glomerulonephritis. Lab Invest, 72: 17, 1995 17. Reyes, A. A., Purkerson, M. L., Karl, I. et al: Dietary supplementation with L-arginine ameliorates the progression of renal disease in rats with subtotal nephrectomy. Am J Kidney Dis, 20: 168, 1992 18. Reyes, A. A., Martin, D., Settle, S. et al: EDRF role in renal function and blood pressure of normal rats and rats with obstructive uropathy. Kidney Int, 41: 403, 1992 19. Reyes, A. A., Porras, B. H., Chasalow, F. I. et al: L-arginine decreased the infiltration of the kidney by macrophages in obstructive nephropathy and puromycin-induced nephrosis. Kidney Int, 45: 1346, 1994 20. Hochberg, D. A., Chen, J., Vaughn, E. D., Jr. et al: Interstitial fibrosis is exacerbated in unilateral ureteral obstruction (UUO) in mice lacking the gene for inducible nitric oxide synthase (iNOS). J Urol, suppl., 68: 68, abstract 257, 1998 21. Schwartz, D., Mendonca, M., Schwartz, I. et al: Inhibition of constitutive nitric oxide synthase (NOS) by nitric oxide generated by inducible NOS after lipopolysaccharide administration provokes renal dysfunction in rats. J Clin Invest, 100: 439, 1997 22. Ling, H., Edelstein, C., Gengaro, P. et al: Attenuation of renal ischemia-reperfusuion injury in inducible nitric oxide synthase knockout mice. Amer J Physiol, suppl., 277: F383, 1999