The role of phosphodiesterase 3 in endotoxin-induced acute kidney injury

Male Sprague Dawley rats weighing between 180 and 230 gm were anesthetized with 80 mg/kg of ketamine and 5 mg/kg of diazepam intraperitoneally. This protocol was approved by the Keimyung University School of Medicine and Massachusetts General Hospital Committee on Animal Research.

Silastic (0.012 inch I.D, 0.025 inch O.D) catheters were placed in the right jugular vein. Endotoxemia was induced by administration of E. coli LPS (1 mg/kg; E. coli 0111:B4, Sigma, St. Luois, MO), single bolus injection through the jugular vein. The LPS dose was selected from the dose used in other experiments to induce mild endotoxemia [17‐22]. Animals were divided into three groups (n = 7/group): 1) Control (0.9% NaCl infusion without LPS); 2) LPS (0.9% NaCl infusion with LPS); 3) Amrinone+LPS (Amrinone infusion with LPS). In the amrinone treated groups, amrinone (Sigma, St. Louis, MO) was continuously infused for 3 h and 15 min at a dose of 40 μg/kg/min (1 ml/kg/h) [23]. LPS was injected 15 min after starting a continuous infusion of either 0.9% NaCl or amrinone. In the control and LPS groups, the same volume of 0.9% NaCl (1 ml/kg/h) was infused continuously. The kidneys were harvested for examination of histopathologic changes (n = 6/group), ultrastructural changes (n = 5/group), and renal tissue cAMP levels (n = 7/group).

Silastic (0.012 inch I.D, 0.025 inch O.D) catheters were placed in the left carotid artery to monitor systemic artery pressure on a Gould recorder (Model R53400, Glen Burnie, MD) and it recorded at 30 minutes intervals.

The expression of PDE3A, PDE3B, and GAPDH was analyzed by reverse transcription-polymerase chain reaction (RT-PCR). Total RNA was isolated by using TRIzole reagent. (Gibco BRL, Grand Island, NY). Five micrograms of total RNA from each specimen were used for complementary DNA synthesis. The PCR reaction was carried out using cDNA prepared using reverse transcriptase. The following protocol was used for the PCR: 95°C for 5 min and 35 cycles of 95°C for 30 s 50°C for 30 s and 72°C for 1 min and 30 s. This was followed by a final extension of 10 min at 65°C. PCR with specific forward and reverse regular or restriction tagged primers was used to amplify PDE transcripts. The PDE3A forward primer was CTG GCC AAC CTT CAG GAA TC and the reverse primer was GCC TCT TGG TTT CCC TTT CTC. The PDE3B forward primer was AAT CTT GGT CTG CCC ATC AGT CC and the reverse primer was TTC AGT GAG GTG GTG CAT TAG CTG [24]. Another PDE3A forward primer was CCG AAT TCC CTT ATC ATA ACA GAA TCC ACG CCA CT and reverse primer was GGG AAT TCG TGT TTC TTC AGG TCA GTA GCC [25]. Forward primer for a 528-bp fragment of the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was ACC ACC ATG GAG AAG GCT GG and reverse primer was CTC AGT GTA GCC CAG GAT GC. RT-PCR results were quantified by densitometry using a Bio RAD imaging densitometer (Model G.S.-690) in conjunction with Molecular Analyst Software, Version 2.1 (Bio Rad laboratories, USA). Optical densities were expressed as arbitrary units.

Blood samples were collected from the inferior vena cava to measure creatinine at the end of the experiment. Blood samples were centrifuged at 4°C for 15 min at 1200 × g in glass tubes, and the serum was stored at -80°C in polystyrene tubes until use. Estimation of BUN and creatinine concentration was performed using a creatinine analyzer (Beckman Instruments, Inc., Fullerton, CA).

Kidney tissue cAMP was measured by a spectrophotometric colorimetric 96 plate ELISA technique. Briefly, the renal cortex was immersed in liquid nitrogen, which was then homogenized in ice-cold 5% trichloroacetic-acid solution and centrifuged. The supernatant was used for the analysis of cAMP using an ELISA kit (R&D Systems, Minneapolis, MN).

The kidneys were fixed in 10% buffered formalin for 24 h before processing and embedding into paraffin. Two sections were cut from each sample and stained with hematoxylin and eosin. The paraffin sections were cut to 5 μm in thickness, mounted on silane-coated glass slides, and stored for 1 h at 60°C. The slides were deparaffinized with xylene, three times, 5 min each, and were rehydrated with graded alcohols (100, 95, 70 and 50%) for 5 min, respectively. After washing with 0.01 M phosphate buffered saline (PBS) for 5 min, the sections were digested with Proteinase K (20 μg/ml) at room temperature for 20 min, and were washed twice with distilled water for 2 min each. The endogenous peroxidase activity was blocked with 3% hydrogen peroxide (H₂O₂) in PBS for 5 min; the slides were rinsed twice with PBS for 5 min. Sections for positive control were treated with 3% H₂O₂ and reacted with DNase I (1–2 U) at 37°C for 30 min, then washed twice with PBS. For negative controls, sections were covered with reaction buffer alone and incubated following same conditions. The sections were incubated 1.5 h with polyclonal antibodies against goat PDE3A and PDE3B (Abcam, Cambridge, UK) at a concentration of 10 μg/ml. For microscopic observation, the sections were counterstained lightly with hematoxylin for one min.

The details for the analysis have been described elsewhere [26]. The cortical tissues were sliced into small pieces, and were fixed in periodate lysine paraformaldehyde (PLP) solution at 4°C overnight. After washing with 0.01 M PBS, the kidney tissues were immersed in graded concentrations of sucrose in PBS (10% for 1 h, 15% for 2 h, 20% for 4 h) at 4°C. Tissue fragments were post-fixed in cacodylate buffered 1% OsO4 for 1 h, and embedded in PolyBed mix resin (Poly/Bed 812, Polysciences Inc., USA). After trimming the PolyBed mix resin blocks, ultrathin sections (100 nm) were cut. Sections were examined using an electron microscope (Hitachi H-7100, Hitachi Co., Japan). Five representative areas of five samples per group were photographed. Two morphologic patterns, intracellular edema and mitochondrial integrity, were graded with a 5-point scale as follows [27]: 0, no abnormality; 1, mild lesions affecting 10% or less of kidney samples; 2, lesions affecting 25% of kidney samples; 3, lesions affecting 50% of kidney samples, and 4, lesions affecting 75% or more of kidney samples. These evaluations were performed in a blinded fashion.

Kidneys from a different group of rats were removed and immediately frozen in liquid nitrogen 3 hours after LPS injection. Samples of isolated kidney were lysed in RIPA buffer (50 μM Tris-HCL [pH 7.5], 150 μM NaCl, 1% Triton X-100, 1% deoxycholate, and 0.1% sodium dodecyl sulfate) that contained 25 μg/ml aprotinin, 2 mM sodium orthovanadate, 25 μg/ml leupeptin, 2 mM phenylmethysulfonyl fluoride, and 50 mM sodium fluoride for 30 minutes on ice. Lysates were clarified by centrifugation at 15,500 × g for 15 minutes at 4°C. The protein concentration of the supernatant was measured with a Bio-Rad Protein Assay Kit. The supernatant was mixed with 2× sample buffer (114 mM Tris, 9% glycerol, 2.7% sodium dodecyl sulfate, 0.02% bromophenol blue, and 4.5% mercaptoehtanol), boiled for 5 minutes, subjected to 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and transferred onto polyvinylidine fluoride (PVDF) membrane (Bio-rad, Hercules, CA). Nonspecific binding sites were blocked by phosphate-buffered saline (PBS)-0.1% Tween 20 supplemented with 5% skim milk for 1 hour at room temperature. The membranes were incubated with anti-iNOS antibody (Santa Cruz Biotechnology, Santa Cruz, CA, 1:200) and anti-PDE3A antibody (Abcam, Cambridge, UK, 1:500) for overnight at 4°C. After extensive washing with three changes of PBS-0.1% Tween 20, the membranes were incubated with horseradish peroxidase-conjugated secondary for one hour. Immunoreactive proteins were visualized with an enhanced chemiluminescense detection system, followed by exposure to Kodak x-ray film (Eastman Kodak, Rochester, NY).

Endotoxin and the PDE inhibitor, both may have an effect on systemic hemodynamics. We monitored mean arterial pressure (MAP) continuously. At baseline (after anesthesia), MAP showed no differences among the groups (Figure 1). MAP did not change in the control and LPS groups throughout the experiment. However, in the amrinone+LPS group, MAP was mildly but significantly lower 30 min after starting of amrinone infusion compared with the two other groups. In the amrinone+LPS group, MAP remained above 80 mm Hg throughout the experiment, while the other 2 groups MAP stayed above 85 mm Hg.

RT-PCR of kidney tissue was carried out to evaluate mRNA expression of PDE3 isoenzymes. RT-PCR yielded the expected size fragment, 300 base pair products, for the primer pair PDE3B. Although we used two differently designed fragments for the primer pairs of PDE3A, RT-PCR did not produce the expected sized fragments of PDE3A. LPS treatment induced PDE3B mRNA expression in kidney tissue; it was 9-fold greater than control (Figure 2A). To confirm no expression of the PDE3A enzyme in the kidney with or without LPS stimulation, we performed immunoblotting of PDE3A. Heart was used for positive control of PDE3A. PDE3A was expressed neither in the control kidney nor LPS treated, while it is expressed in the heart (Figure 2B). This result suggested that PDE3B gene but not PDE3A was stimulated by LPS in the kidney.

To confirm the RT-PCR results and evaluate structural localization of PDE3 protein expression in the kidney tissue, we performed immunohistochemistry of the two PDE3 isoenzymes.

PDE3A was not detectable in both the control (n = 6) and LPS (n = 6) treated kidneys. However, PDE3B was expressed in the distal tubule in the control group in all six kidneys. In the LPS exposed rats, PDE3B is expressed on both the proximal and distal tubule in all six kidneys. PDE3B was not present in glomeruli. A typical staining pattern is demonstrated in Figure 3. These results confirmed that PDE3A was not expressed but PDE3B was expressed in the kidney tissue. The major activating site of PDE3B protein was the proximal tubule.

There was a significant decrease in cAMP in LPS challenged kidney tissue at the end of the experiment (saline 120 ± 9 vs. LPS 79 ± 11 pmol/mg protein; n = 7/group; P < 0.05), a 35% reduction of cAMP level compared with baseline. Amrinone infusion prevented the fall in kidney tissue cAMP level leaving it comparable to control (110 ± 5 pmol/mg protein) (Figure 4). These results indicated that about 75% of the decreased cAMP levels in the LPS treated tissue was preserved by PDE3 inhibition.

Renal function was assessed 3 h after administration of LPS. Renal failure occurred at 3 h as evidenced by a rise in BUN (saline 21 ± 5 vs. LPS 43 ± 8 mg/dl; P < 0.05) and serum creatinine levels (saline 0.52 ± 0.05 vs. LPS 0.73 ± 0.04 mg/dl; n = 7/group; P < 0.01). The PDE3 inhibitor, amrinone, (40 μg/min/kg) significantly attenuated BUN (27 ± 6 mg/dl; P < 0.05) and creatinine levels (0.58 ± 0.04 mg/dl; P < 0.01) (Figure 5). This suggested that PDE3 inhibition attenuated endotoxin-induced renal failure.

Since light microscopic examination of renal tissue revealed normal histology, we investigated ultrastructural changes of the renal proximal tubule. In the control group, there was well preserved integrity of mitochondria without cellular edema. However, there was a substantial change in the mitochondria and in the intracellular edema in the proximal tubules in the LPS group (Figure 6). Amrinone attenuated the change of mitochondrial integrity and cellular edema. The result of the numerical scores and means are presented in Table 1.

Administration of LPS significantly increased iNOS protein in the kidney, while the PDE3 inhibitor significantly attenuated iNOS protein expression (Figure 7). iNOS was detected at 130 kD in the kidney tissue from LPS and to a lesser degree in the amrinone+LPS groups. These results suggested that there may be an association between PDE3 inhibition and iNOS down regulation.