Nobuyoshi Takahashi md, Department of Urology, Hirosaki University School of Medicine, 5 Zaifu-cho, Hirosaki, Aomori 036-8563, Japan. Email: email@example.com
Background: Recent studies have shown that nitric oxide (NO) synthases, particularly inducible nitric oxide synthase (i-NOS), are induced in acute rejection episodes following heart, liver, pancreas and kidney allotransplantation. Furthermore, tissue and cellular injury has been demonstrated to be mediated by peroxynitrite (ONOO-), a metabolite of NO as well as a potent oxidant. However, a detailed relationship between NO, i-NOS and graft injury in transplantation remains elusive.
Methods: The present study used the following models of renal transplantation in rats: allografts (n = 5, Brown–Norway to Lewis [LEW] rats), isografts (n = 5, LEW to LEW) and allografts treated with aminoguanidine (AG), an i-NOS inhibitor (n = 5). Blood urea nitrogen (BUN), serum creatinine (SCr) and urinary and serum nitrosocompounds (NOx) were measured on days 2, 4 and 7 post-transplant. Western blot analysis of i-NOS protein expression and measurement of i-NOS activity were carried out in grafts harvested on Day 7, along with immunohistochemical and histopathological examinations.
Results: In the allograft group, both BUN and SCr levels increased markedly on Day 7, in parallel with a sharp increase in NOx. A band stained by anti-i-NOS antibody was detected at approximately 130 kDa, along with high levels of i-NOS activity and diffusely distributed i-NOS-positive cells (macrophages). Histologically, an acute rejection episode was confirmed (Grade 3 according to Banff classifications). In the AG group, reduced renal function and graft injury were significantly less severe than in the allograft group.
Conclusions: In rat renal allograft acute rejection, markedly increased levels of serum NOx were observed, along with enhanced tissue i-NOS activity, together resulting in graft injury. AG administration suppressed the increase of serum NOx levels, with concomitant mitigation of tissue injury and renal function impairment.
During episodes of allograft rejection, activation of alloantigen by antigen-presenting cells results in infiltration of macrophages and lymphocytes (including T cells) into the graft.1 Augmented production of cytokines such as TNF-α and IFN-γ induces massive expression of inducible nitric oxide synthase (i-NOS), greatly increasing the production of nitric oxide (NO). Recent studies have demonstrated marked induction of i-NOS expression in experimental models of cardiac, hepatic, pancreatic and renal allograft rejection.2–4
The present study used a rat renal allograft rejection model to investigate changes in NO metabolic products (NOx), expression of i-NOS protein and i-NOS activity in graft tissues, and examined the relationship between NO, i-NOS and graft injury. Results were compared with a renal isograft model, and with a rat renal allograft model treated with aminoguanidine (AG), an i-NOS inhibitor.
Materials and methods
Male Brown–Norway (BN) rats (225–250 g; 10–12 weeks old) and Lewis (LEW) rats (225–250 g; 10–12 weeks old) were used in the present study. All rats were obtained from a single source (CLEA Japan, Tokyo, Japan).
Rat renal transplant models
To produce an allograft model (allograft group; n= 5), kidneys from BN donors were transplanted into LEW recipients. For an isograft model (isograft group; n= 5), kidneys from LEW donors were transplanted into LEW recipients. In the allograft group treated with AG (AG group; n= 5), an allograft model underwent intraperitoneal injection of AG, an i-NOS inhibitor (Sigma Corp, St Louis, MO, USA); intraperitoneal injections commenced from Day 1 at a daily dose of 400 mg/kg of bodyweight.
The left kidney from the donor was transplanted into the right abdominal cavity of the recipient. Donor blood vessels were severed, together with the abdominal aorta and inferior vena cava, and the ureter was detached near the bladder and separated. The extracted graft was immediately irrigated from the aorta with heparin-containing physiological saline at 4°C. The aorta and inferior vena cava of the graft were then anastomosed to the corresponding recipient vessels in an end-to-side fashion. Bilateral native nephrectomy was performed at the same time. Blood flow to the graft was restored within 60 min.
Measurements of nitrosocompounds, blood urea nitrogen and creatinine
After transplantation, conditions had stabilized and recipients were kept in a metabolic cage from day 2 post-transplantation onwards. From days 2–6, urine was collected every 24 h and analyzed. After centrifugation, supernatants were collected. On days 2, 4 and 7 200 mL of blood was collected from the caudal vein, and sera were collected after centrifugation. Blood urea nitrogen (BUN) and serum creatinine (SCr) levels were measured using enzymatic methods. Urinary and serum concentrations of nitrosocompound (NOx = NO2 + NO3) were measured using Griess Reagent. (NO2/NO3 Assay Kit-C, Griess Reagent Kit, Wako Pure Chemicals, Tokyo, Japan.)
Western blot analysis of i-NOS protein expression
Graft harvested on Day 7 was homogenized in 1× SDS buffer. After adding 100 µM phenylmethylsulfonylfloride (PMSF), the homogenate was centrifuged for 15 min at 14 000 ×g. Protein content of the supernatant was measured according to the Bradford method using a U-3200 absorptiometer (Hitachi, Japan). The sample was subjected to 8% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), then transferred onto a nitrocellulose membrane. Immunological staining was performed on the nitrocellulose membrane using antirat i-NOS antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA) to investigate expression of i-NOS protein.
Measurement of i-NOS activity
A small sample of renal tissue was homogenized in Tris-EDTA buffer containing 0.5% Triton-X-100, and PMSF was added. After the sample was kept on ice for 30 min, homogenate was centrifuged at 2°C, 14000 ×g for 20 min. Protein content of the supernatant was measured using the Bradford method, and supernatants were diluted to obtain samples of fixed protein content. Induced NOS activity was measured using an i-NOS Activity Measurement Kit (Cerep, Rueil-Malmaison, France).
Harvested graft was sectioned and fixed in 10% buffered formaldehyde, embedded in paraffin, and stained with hematoxylin and eosin (HE), periodic acid Schiff (PAS), periodic acid-methenamine-silver (PAM) and phoshotungstic acid hematoxylin (PTAH). Sections were evaluated for evidence of rejection and tubulointerstitial injury.
Immunohistochemistry was performed on thin frozen sections using the antirat i-NOS antibody to determine expression and distribution of i-NOS protein in harvested graft tissues.
The Mann–Whitney U-test was used to analyze data. Values of P < 0.05 were considered statistically significant.
Changes in renal graft function
Changes in graft function are shown in Figure 1. In the isograft group, concentrations of both BUN and SCr peaked on Day 2 post-transplant, declining thereafter and normalizing by Day 7. In the allograft group, BUN and SCr concentrations also peaked on Day 2 and demonstrated decreases on Day 4. However, these changes were followed by a sharp increase to levels diagnostic of severe impairment on Day 7, indicating an acute rejection episode (P < 0.01). In the AG group, both BUN and SCr levels declined continuously after peaking on Day 2, compared to the allograft group (P < 0.05).
Changes in serum and urinary NOx levels
Changes in serum and urinary concentrations of NOx are shown in Figure 2. In the isograft group, plasma NOx levels were slightly increased from Day 2, remaining at 40–50 µmol/L until Day 7. In the allograft group, plasma NOx levels were mildly increased on Day 2, followed by a sharp increase to approximately 400 µmol/L on Day 4 and a further increase to approximately 500 µmol/L on Day 7 (P < 0.01). Changes in the AG group resembled those in the isograft group. Urinary NOx levels in all three groups showed no marked variations, but levels were reduced on Day 7 compared to preoperative values.
Western blot analysis of i-NOS protein expression in tissue extracts
In all rats in the allograft and AG groups, a band stained by anti-i-NOS antibody was detected with equal intensity at approximately 130 kDa. No such band was observed in the isograft group (data not shown).
Measurement of graft tissue i-NOS activity
I-NOS activity levels in graft tissues are shown in Figure 3. The allograft and AG groups showed significantly higher i-NOS activity levels than the isograft group (P = 0.049). No difference was observed between the allograft and AG group.
In the allograft group, striking infiltration of mononuclear cells was found in the stroma around blood vessels and tubules, and inside some glomeruli. Complete obstruction of arterioles, caused by thrombosis, was also observed.
In tubules, necrosis of epithelial cells, vacuolization and luminal obstruction were observed. These findings indicated an acute rejection episode (Grade 3, according to Banff classifications 1997). In the isograft group, few rats displayed tubular necrosis and focal infiltration of inflammatory cells, with no involvement of arterioles or glomeruli. In the AG group, moderately necrotic and swollen tubules predominated, with obvious perivascular mononuclear cell infiltration in edematous areas (Grade 1 to 2). Injury was significantly less severe than in the allograft group. Spaces between tubules were probably due to edema.
Numerous small reddish-purple PAS-positive particles were observed in tubular epithelial cells in the allograft group only. No such particles were seen in either the AG or isograft group.
In the allograft group, purplish, reticulate PTAH-positive materials indicating the presence of thrombus were observed in arterioles and glomerular capillaries. No such reticulate structures were observed in either the AG or isograft group.
No significant differences were observed between the three groups (data not shown).
Immunohistochemical findings of graft tissues are shown in Figure 4 (lower panels). In the allograft and AG groups, i-NOS reactivity was observed in mononuclear inflammatory cells with abundant cytoplasm, which presumably represented macrophages. In the allograft group, i-NOS-positive cells were diffusely distributed throughout the perivascular and peritubular stroma, even infiltrating the interior of some glomeruli. In the AG group, however, i-NOS-positive cells were not localized in glomeruli, but in perivascular and peritubular areas of the medulla. No i-NOS immunoreactivity was identified in the isograft group.
Recent studies have demonstrated that NO synthases, especially i-NOS, are induced in acute rejection episodes following allotransplantation of heart, liver, pancreas and kidney.2–4 Furthermore, the localization of i-NOS has been reported in infiltrating leukocytes.4 Conversely, the actions of NO are not limited to vasodilatation, as initially proposed, and cellular injury has also been demonstrated to be mediated by peroxynitrite (ONOO-).5–7 However, a detailed relationship between NO, i-NOS and tissue injury in transplantation has not been examined.
In the present study, a rat renal transplantation model was used to investigate whether NO production in acute rejection episodes is associated with graft injury, and whether an i-NOS inhibitor mitigates renal impairment.
In the isograft model, graft function was transiently lowered on Day 2 after grafting, but improved gradually to display normal function by Day 7. These results reflect acute lowering of renal function due to ischemic-reperfusion injury, followed by recovery. Histopathological findings demonstrated no mononuclear cell infiltration, and graft injury was limited to necrosis in small numbers of tubules.
In the allograft model, transient ischemic-reperfusion injury and initial recovery were observed as per the isograft model, but graft function showed further marked deterioration after Day 4. HE staining showed findings indicative of acute rejection, such as marked mononuclear cell infiltration, necrosis of vascular and tubular cells, destruction of vascular endothelium and tunica media, and tubular obstruction.
Although serum NOx levels were mildly increased in the isograft model, NOx production was drastically increased in the allograft model, because of acute rejection. NOx was measured instead of NO, because NO displays an extremely short life of several seconds in vivo, and measurement of accumulated metabolites (NO2 and NO3) is substantially easier.8 Because local NO production is reportedly correlated with whole blood NOx, increased serum NOx levels in the allograft model would seem likely to indicate augmented NO production in the graft.9 Indeed, Western blot analysis of i-NOS protein expression in graft tissue demonstrated an i-NOS-positive band at approximately 130 kDa in the allograft model, but not in the isograft model. Furthermore, i-NOS activity level in graft tissue was significantly higher in the allograft model than in the isograft model. A mild increase of serum NOx levels was detected in both the isograft and allograft models on Day 2. In a previous study using a rat renal ischemic-reperfusion injury model, increased NOS activity and NO production were observed and attributed to increased endothelial NOS activity derived from vascular endothelium.10 In the present experiment, circulation was interrupted for 50–90 min, potentially accounting for the early mild increase in i-NOS.
In the allograft group, i-NOS immunoreactivity was observed in infiltrating mononuclear cells with abundant cytoplasm, and the morphology of these cells was strongly suggestive of macrophages. The i-NOS expressed in those cells can be expected to produce large quantities of NO. The drastic increase in NOx production after Day 4 in the allograft group probably signifies initiation of acute rejection and increased expression of i-NOS from around this time. In an experiment on allografting using sponge-matrix, inflammatory cell infiltration began on Day 3 after grafting, along with the production of TNF-α and IFN-γ.11 Following expression of IFN-γ, concentrations of cytokines such as IL-2 and IL-6 also increase in cells infiltrating the allograft.12 Acute rejection occurring from 3 to 4 days post-transplant might, therefore, induce expression of i-NOS inside infiltrating macrophages, thereby leading to increased NO production.
In the AG group, the deterioration of graft function, which had been observed in the allograft group, was suppressed after Day 4. Serum NOx level was only slightly increased after grafting. Histopathologically, although infiltration by numerous mononuclear cells was observed, necrosis and swelling of tubular epithelial cells were mild compared to the allograft group. Immunohistochemistry also showed infiltration of i-NOS-positive mononuclear cells with abundant cytoplasm, as seen in the allograft group. In addition, PAS-positive granules were found in tubular cells in the allograft group, but not in the AG group.13 PAS-positive substances presumably represent protein particles, and are found in conditions of glomerular injury such as nephrotic syndrome. PTAH staining was then conducted to determine if microthrombi contribute to glomerular injury. The results showed numerous microthrombi inside glomeruli in the allograft group, but none in the AG group. Thrombus formation is well known to result from vascular endothelial damage leading to platelet adhesion and aggregation in addition to activation of coagulation factors. In the allograft group, injury of endothelial cells in glomerular capillaries might cause thrombus formation that lowers glomerular function, resulting in increased protein excretion and the appearance of PAS-positive protein particles in tubules.
Although inhibition of NO production was found to correlate with amelioration of tissue injury, NO per se does not posses strong chemical reactivity.14 Instead, in the presence of inflammatory cells such as macrophages, coexistence of NO and superoxide anion (O2–) generates ONOO-, which is associated with various pathological conditions.5,6 Through processes such as the oxidation of sulfhydryl (SH) base, lipid peroxidation and nitrification of aromatic amines,15 ONOO- destroys DNA by activating poly ADP ribose polymerase (PARP)16 causing tissue and cellular injuries. In the allograft model, O2– production is probably also augmented by inflammatory cell infiltration, and might bind with the increased NO, resulting in generation of ONOO- that causes graft injury.
Nitric oxide generated by macrophages enters the blood and reacts with oxyhemoglobin to form NO3–, whereas ONOO- is unstable and isomerizes to form NO3–.17 NO3– does not affect renal function, and is excreted in urine together with NO2–. Urinary NOx excretion was found to be lowered in all three groups compared to baseline, probably reflecting a delay in recovery of renal function after transplantation, even in the isograft group.
With regard to i-NOS activity in graft tissues, significant differences might be expected between the isograft, allograft and AG groups. However, no significant difference in i-NOS activity was detected between the allograft and AG group. We believe that this discrepancy might be due to the methods used for measuring in vivo i-NOS activity. Presumably AG showed no competitive inhibitory effect on L-arginine, the precursor in NO production, because of the reversible, weak binding to i-NOS protein; therefore, in the process of extraction, after adding an excess of L-arginine following tissue homogenization, AG was expected to be released and devitalized, leading to reactivation of remaining i-NOS protein in tissues.18 Consequently, in the AG group, i-NOS activity in tissues per se was suspected to be low enough to suppress NO production, with a decreased plasma NOx level and amelioration of graft injury.
Although i-NOS protein expression was detected in the AG group, AG inhibited I-NOS activity, suppressing NO production and ONOO- formation and thereby reducing cellular impairment. This drug action might account for the differences in renal function and histopathological findings observed between the allograft and AG group. In the AG group, O2– would remain in tissues, because of decreased production of NO. The injurious effects of O2– and the direct effects of cytokines must therefore also be considered. Edema in graft tissue was more pronounced in the AG group than in the allograft group. Because a high frequency of edema has been observed under conditions of abundant O2– production, such as bronchial asthma,19 the edema observed in the present study does not represent an entirely unexpected finding.
In future studies, co-administration of i-NOS inhibitor with an agent capable of removing O2–, such as superoxide dismutase (SOD), or inhibitors of xanthine oxidase (an O2– producing agent), such as allopurinol, might more effectively reduce the tissue or cellular injury. In a mouse model with influenza virus-induced pneumonia, co-administration of the NOS inhibitor, NG-monomethyl-L-arginine, with SOD and allopurinol resulted in improvement of disease parameters and survival rate.20,21
Recently, the development of new immunosuppressants such as cyclosporine A and FK506 have markedly improved graft survival and substantially decreased the frequency of acute rejection episodes.22 However, poor absorption of immunosuppressants might result in low blood concentrations, allowing the development of acute rejection and graft injury. In these cases, although symptoms and abnormal laboratory results permit diagnosis to a certain extent, early diagnostic markers have remained unavailable. Initiation of treatment is therefore often delayed, during which time tissue injury progresses. Accelerated acute rejection caused by low blood concentration of immunosuppressants due to poor absorption occurs most frequently on days 4–7 post-transplant. During this latent period, prophylactic administration of i-NOS inhibitor, SOD or allopurinol seems likely to prevent tissue injury and benefit therapy.
When acute rejection occurs, an increased NO production is thought to accompany inflammatory cell infiltration into the graft tissue. Takahashi et al. reported the usefulness of measuring serum NOx level as an early indicator of acute rejection in human recipients of renal allograft.23 However, in order to ensure its efficacy, a further multistudy is needed in the future.