Protective effect of EDTA preadministration on renal ischemia

Male Sprague-Dawley rats weighing about 200 g were used (Charles River Italia, Lecco, Italy) and were allowed water and standard rat chow ad libitum. All the rats were maintained at 22 ± 1°C with a 12/12 hours light/dark cycle.

The rats were anesthetized with an inhaled anesthesia mixture of halothane 2% (Hoechst, Milano, Italy) and oxygen. They were placed on a temperature-regulated table (38°C) (Ugo Basile, Comerio, Lecco, Italy) to maintain body temperature. Kidney ischemia (Isc) was induced by clamping the right renal artery and the right renal vein for 60 minutes with a microsurgical clamp. In the Isc/R group, at the end of the I period, the vascular clamp was removed and reperfusion of 60 minutes was performed. During the surgical procedure the heart rate and the mean arterial blood pressure (MABP) were monitored.

At the end of Isc or of Isc/R, blood samples were obtained by exanguination of rats at the aorta bifurcation level and kidneys were collected and processed for different studies. Blood and kidneys from EDTA-treated-not-ischemized rats were collected 90 minutes after EDTA administration (corresponding to 30 minutes EDTA pre-treatment+60 min Isc).

The right femoral artery was cannulated through a polyethylene catheter and connected to a pressure transducer for the measurement of MABP [15, 16]. The data was collected continuously by means of a computer and were calculated at baseline, at the end of EDTA pre-treatment (e.g 30 minutes after EDTA intravenous injection), at the end of Isc and at the end of postischemic R. In sham-operated rats the values were calculated 90 minutes after EDTA pre-treatment.

EDTA (calcium disodium EDTA) (Collalto, Brescia, Italy) used in human therapy was employed [3], and at the same dosage (e.g. 40 mg/kg body weight). The sterile drug solution of 2 g/10 ml was opportunely diluted in physiological saline and administered by left intrafemoral vein slow infusion.

The inhibitor of NO synthases L-NAME [N(omega)-nitro-L-arginine methyl ester], when required, was injected simultaneously with the EDTA through the intrafemoral vein at the dose of 30 mg/kg body weight, 30 minutes before the induction of Isc or Isc/R.

The rats were randomly allocated to 4 study groups, each composed of 15 rats: group 1, controls; group 2, sham operated: the rats underwent the same surgical procedure, except that the clamp was not applied; group 3, Isc: ischemia was induced for 60 min; group 4, Isc/R: ischemia was induced for 60 min, followed by 60 min reperfusion at room temperature. Other identical 4 groups were studied, in which EDTA treatment was performed. In groups 3 and 4 intrafemoral injection of physiological saline 30 minutes before clamping was performed. The 3 and 4 EDTA-treated groups received a single intravenous injection of EDTA 30 minutes before clamping. In groups 1 and 2 intrafemoral injection of physiological saline or EDTA was performed 90 minutes before kidney removal (= 30 minutes EDTA pre-treatment+60 min Isc).

To take in consideration that EDTA could lead to increase in NO plasmatic levels through increase in eNOS expression, we further performed histological evaluations on two additional groups of rats, to verify whether the eNOS inhibitor L-NAME was able to block the protective effect of EDTA in renal ischemic injury. In such groups the animals were simultaneously treated with EDTA and L-NAME 30 minutes before the induction of Isc (group 5) and 30 minutes before the induction of Isc/R (group 6).

Serum creatinine was measured using a modified Jaffe's reaction, and blood urea nitrogen was measured on the AEROSET system (Abbott Laboratories, Abbott Park, IL) [17].

Kidneys were excised, decapsulated, dissected into 4 pieces along the major ax, fixed by immersion in 4% paraformaldehyde in Dulbecco's PBS (DPBS) overnight at 4°C, cryo-protected in 10% sucrose in DPBS, then embedded in Tissue-Tek medium and frozen in liquid nitrogen. Cryostat-cut four sections/animal (5 μm thick) were submitted to Hematoxylin/Eosin stain; renal damage was evaluated as tubular epithelial cell necrosis, tubular dilation, protein casts and medullary congestion (18). The alterations were semi-quantitatively graded by a pathologist blind to the nature of the experiments. The grading was performed by the following criteria: - =absent, + = barely present, ++ = moderate, +++ = severe. Expression of eNOS, e.g. the endothelial form of the constitutive NO synthase, was assessed on serial sections, with the use of a specific monoclonal antibody (BD Pharmingen, Franklin Lakes, NJ), followed by a Rabbit-anti-Mouse IgG- AlexaFluor488 (Molecular Probes, Eugene, OR). Observations were performed by using an Eclipse 55i microscope (Nikon, Tokyo, Japan), digital images acquired with DS-L1 camera and LUCIA G software (all from Nikon) and mounted using AdobePhotoshop CS software.

The expression of Mac-1 was evaluated by following FACS analysis. Whole blood was incubated with 0.5 μg of FITC-conjugated CD11b monoclonal antibody (clone WT5, isotype mouse IgA, K) (Pharmingen, San Diego, CA) for 20 minutes in ice. After erythrocyte lysis, samples were run on a FACscan (Becton-Dickinson, Mountain View, CA) and gated on PMN parameters. Results are expressed as arbitrary units of mean fluorescence intensity (MFI, a.u.).

The rats were bled off at the aorta bifurcation level. Blood was collected in the presence of 0.065 mM citric acid (Riedel, Hannover, Germany), 0.085 mM sodium citrate (Farmitalia, Milan, Italy) and 2% glucose monohydrate (Riedel) in the blood: anticoagulant ratio of: 7:1. Samples were obtained from rats immediately after the end of each treatment or surgical procedure.

NO release was determined spectrophotometrically [19] by measuring the nitrate/nitrite (NO₂ ^-/NO₃ ^-) concentration in plasma samples from arterial non coagulated blood. Briefly, whole blood was centrifuged and plasma samples were collected, incubated for 30 min at 37°C in the presence of 0.2 U/ml Aspergillus nitrate reductase (Boehringer-Mannheim, Milan, Italy), 50 mM HEPES buffer (pH 7.4), 5 μM flavin adenine dinucleotide (Sigma Aldrich), and 0.1 mM NADPH (Sigma Aldrich). Then, lactate dehydrogenase (Boehringer Mannheim) and sodium pyruvate (Sigma Aldrich) were added to a final concentration of 10 U/ml and 10 mM, respectively, and the samples were incubated for 10 minutes at 37°. The Griess reagent (Sigma Aldrich) was added to the samples (100 μl), and absorbance was measured at 540 nm after 15 minutes incubation at room temperature. Standard curves with increasing concentrations of sodium nitrate and sodium nitrite were run in parallel.

The assay was performed as described [20]. Briefly, the exit of albumin from vessels into the parenchyma of rat kidneys was assayed. The dye solution contained 0.4% albumin (Sigma Aldrich) and 0.5% trypan blue (Sigma Aldrich) in saline. Following laparatomy, animals were perfused with 5 ml dye-solution through the right renal artery for 10 minutes. The perfusate was drawn from the right renal vein. The right kidney was washed with saline in vivo, removed, weighted, suspended and homogenized in buffered phosphate solution at pH 7.4 (1 g tissue dissolved in 3 ml buffer). In treated animals, after halothane anesthesia, EDTA (40 mg/kg) was injected intravenously (through the femoral vein), followed by rat TNFα (R&D System, Abingdon, UK) (0.1 ng/g). TNFα and EDTA, alone or together, were injected 30 minutes before kidney dye perfusion.

Tissue extracts were centrifuged, the supernatants recovered and treated with 10% deoxycholic acid (sodium salt monohydrate, Sigma Aldrich) in saline, to remove lipid interference. Dye was evaluated by spectrophotometer analysis (Pye Unicam SP6-550, Cambridge, United Kingdom) at 540 nm.

The results are expressed as the mean ± SEM of 15 animals in each group. They were analyzed using a two way analysis of variance followed by Bonferroni t-test. The results were considered statistically significant when p < 0.05.

In the rats studied the heart rate did not vary significantly during the experimental procedure (data not shown). To establish whether EDTA could maintain vascular homeostasis, we measured MABP in both untreated and EDTA-treated rats. EDTA treated rats displayed MABP values lower than those of untreated control and sham operated animals. To note, the increase of MABP due to Isc was significantly prevented by EDTA pre-administration (Table 1).

Rats undergoing either renal Isc, obtained by clamping the right renal artery and the right renal vein for 60 minutes, or Isc followed by 60 minutes reperfusion (Isc/R), obtained by removing the clamp, were evaluated for the levels of serum creatinine and blood urea nitrogen (Fig. 1), two parameters routinely used to assess renal function. Both creatinine and urea had a significant increase after the induction of Isc and Isc/R, clearly indicating an impairment of the renal filter function. Interestingly, the administration of EDTA before Isc and Isc/R, maintained both parameters at physiological levels (Fig. 1), thus suggesting a protective role of EDTA toward the renal filter capacity.

To assess whether EDTA, administered before renal Isc or Isc/R induction, protected kidney not only from functional damage but also from structural alterations, we performed histological evaluations, aimed to determine the presence of tubular epithelial cell necrosis, tubular dilation, protein casts and medullary congestion (Fig. 2). For this, kidneys from treated rats were excised and sections were stained with Hematoxylin/Eosin, to compare their architecture with that of control kidneys (Fig. 2a). Picture relative to kidneys from EDTA-treated rats (Fig. 2b) was similar to that of control kidneys (Fig. 2a): indeed, interstitial spaces were maintained and proximal tubule as well as cortical distal segments were preserved. Kidneys from sham-operated rats did not show evidence of important modifications with respect to the controls (data not shown). Kidneys from rats undergoing Isc (Fig. 2c) showed severe renal lesions, mainly tubular, such as dilation and focal engulfment by protein casts. Glomerular and interstitial hemorrhage were also present. Some tubular cells were necrotic, whereas other appeared vacuolized. This picture worsened when kidneys were obtained from rats submitted to R (60 min) after Isc, displaying (Fig. 2e) tubular cast increase and glomerular hypertrophy. Noteworthy, kidneys from animals pre-treated with EDTA before the induction of Isc, (Fig. 2d) failed to show important renal lesions. EDTA pretreatment preserved also the architecture of kidneys submitted to Isc/R (Fig. 2f). No significant differences were evident by comparing panel d and f of Fig. 2. Pictures related to the ascending thick limb in the kidney medulla displayed interstitial hemorrhage at the end of Isc in control kidneys. On the contrary, interstitial hemorrhage was absent in kidneys from EDTA-treated ischemized rats (data not shown). The semiquantitative analysis of renal damage, which represents the mean features for each group of animals, is summarized in Table 2. The use of the eNOS inhibitor L-NAME, simultaneously injected with EDTA before the induction of Isc and Isc/R was able to block the beneficial effects induced by EDTA.

To investigate a putative mechanism of action of EDTA, we considered its effect on PMN, which are largely involved in the damage associated with Isc/R [11, 12].

For this purpose, PMN, isolated from peripheral rat blood, were analyzed for the expression of the pro-adhesive molecule Mac-1 (Fig. 3); the existence of Mac-1 up-regulation is suggestive of PMN activation [21]. Mac-1 expression by PMN obtained from control rats increased significantly after Isc and Isc/R. Following EDTA pretreatment, the increase was significantly impaired in rats submitted to Isc and, at lower extent, to Isc/R.

Being the expression of adhesion molecules, the adhesive and migratory pattern of leukocytes finely regulated by NO, both in physiologic and pathologic conditions [10, 22‐24], we then measured rat NO plasmatic levels. EDTA pre-treatment significantly increased the levels of circulating NO (Fig. 4) both in control and in ischemic rats. Conversely, post-ischemic reperfusion impaired dramatically the production of NO but was not insensitive to EDTA pre-administration: in fact NO production following Isc/R in EDTA-pre-treated rats was similar to that measured in control rats.

As NO in vascular endothelial cells is synthesized primarily by the endothelial form of the constitutive NO-producing enzyme (eNOS), we examined the possibility that a decrease in NO bioavailability might be related to a change in the rate of expression of eNOS. The renal expression of eNOS (Fig. 5), observed in glomerular and interstitial capillaries, was slightly higher and diffuse in animals treated with EDTA (b), as compared to untreated control rats (a). Induction of short time (60 min) Isc, in control rats, produced a loss in the glomerular eNOS and an increase of its interstitial expression (c). When Isc followed EDTA treatment (d), eNOS expression was prevalently assessed inside glomerular capillaries. Kidney sections obtained following Isc/R in controls showed very low expression of eNOS both at glomerular and interstitial levels (e). Kidney sections from rats treated with EDTA before Isc/R (f) displayed fluorescence findings comparable to that of controls (a). The use of L-NAME together with EDTA before the induction of Isc and Isc/R abrogated the increase in eNOS expression due to EDTA treatment alone (data not shown).

It has been recently demonstrated that eNOS has a critical role in regulating the microcirculatory endothelial barrier function in vivo [25]. We investigated whether EDTA influenced the TNFα-induced vascular leakage in kidneys. Vascular leakage values (expressed as μg dye/g fresh kidney and mean ± SEM of 8 rats) are reported in Table 2. A significant increase in dye retention has been shown by kidneys of rats treated with TNFα with respect to kidneys of untreated animals (controls). EDTA treatment alone did not alter the endothelial barrier function. The concomitant administration of EDTA and TNFα resulted in the significant reduction of TNFα-induced leakage, indicating the existence of tights links among EDTA-NO-vascular protection.