点击显示 收起
【摘要】 Kidney glomeruli are important targets of diabetic nephropathy. We hypothesized a high concentration of glucose could suppress glomerular endothelial nitric oxide synthase (eNOS) by a protein kinase C (PKC) mechanism, as has been found in other tissues. Mouse kidney slices (150-200 µm) were bathed in Hanks' solution with 100 µM L -arginine and exposed to either 5 or 20-30 mM D -glucose. Immunofluorescence identified only eNOS in normal mouse glomeruli. Measurements of glomerular NO concentration with NO-sensitive fluorescent dye (4,5-diaminofluorescein diacetate) using confocal microscopy and NO-sensitive microelectrodes verified that resting glomeruli had active production of NO that was inhibited by N G -nitro- L -arginine methyl ester. High-concentration (20-30 mM) D -glucose inhibited 60-70% of the NO production within 15-30 min; L -glucose at the same concentration did not have any effect. Inhibition of PKC- with 100 nM ruboxistaurin prevented eNOS suppression in high-glucose media. Activation of PKC with 100 nM phorbol ester also suppressed the glomerular NO concentration. We concluded that eNOS in the renal glomerular capillary endothelial cells is suppressed by activity of PKC at high-glucose concentrations comparable to those in diabetic animals and humans. The consequence is a rapid decline in the generation of NO in the glomerular endothelial cells in the presence of a high concentration of glucose.
【关键词】 microelectrode hyperglycemia confocal imaging protein kinase C endothelial nitric oxide synthase
DIABETIC NEPHROPATHY CONTRIBUTES to more than 40% of terminal kidney diseases and is a major diabetic complication ( 43 ). Nephropathy develops in both type I and II diabetic patients and in both groups, one of the first indications of abnormal glomerular function is increased filtration of albumin ( 15 ). This leakage predicts damage of the glomerular membrane allowing unusually large molecules to sieve between the cells. However, what mechanisms mediate diabetic glomerular pathogenesis is unclear. As the first component of the filtration process begins with glomerular endothelial cells and diabetes impairs endothelial nitric oxide synthase (eNOS) function throughout the body, Jover and Mimran ( 25 ) suggested a possible link of impaired endothelial-generated nitric oxide and glomerular permeability function may exist. Although diabetes is associated with chronic eNOS inhibition by a variety of mechanisms ( 22, 26 - 28, 39 - 42 ), it is not clear if hyperglycemia acutely affects eNOS function in glomeruli in normal conditions or very early stages of diabetic mellitus. We are particularly interested in rapid effects of hyperglycemia because most diabetic humans are only occasionally hyperglycemic. In this context, in vivo microcirculatory studies of the brain ( 36 ), skeletal muscle ( 31 ), and small intestine ( 5, 7 ) in normal rats indicate a rapid, within 45 min, suppression of NO formation. Bioassay studies of endothelial-mediated dilation have also found a rapid loss of NO function in the vasculature of normal humans during acute hyperglycemia ( 300 mg/dl) ( 2, 48 ). A common factor in eNOS suppression during acute hyperglycemia is activation of protein kinase C (PKC), and PKC activation is a known factor in diabetic nephropathy ( 12, 14, 20, 24, 30 ). The PKC- II isoform of endothelial cells ( 2, 4, 7 ) appears to be particularly important, as judged by the protection offered by blockade of this PKC isoform by ruboxistaurin (formerly LY-333531, Eli Lilly). As the primary goal of the current study, we tested the general hypothesis that acute exposure to a high concentration of glucose (high glucose) inhibits glomerular NO production through PKC suppression of eNOS.
Histological studies have shown that normal glomerular endothelial cells have high expression of eNOS in humans ( 1, 18 ) and in mice, the species of this study ( 47 ). Functional studies predict that eNOS has an important role in renal protection by maintaining normal glomerular function through inhibition of thrombosis, leukocyte adhesion/activation, apoptosis, and oxidative stress in glomeruli ( 21, 38 ). The consequences of both inhibition of eNOS in normal mice and genetic knockout of eNOS in mice are increased vulnerability to experimentally induced glomerulonephritis ( 21 ). As hyperglycemia has the ability to rapidly and severely compromise eNOS function through a PKC mechanism in other vascular beds, determining if glomerular capillary NO production is impaired rapidly by high glucose was the major goal of this study. The study of glomerular capillaries in terms of their ability to produce NO would be exceedingly difficult in an in vivo environment. Therefore, thin (150-200 µm) slices of mouse kidney were used as an in vitro model. Using this model, we tested three hypotheses. First, that glomerular NO should be primarily produced by eNOS during normal conditions. This hypothesis was evaluated with immunofluorescence and confocal microscopy to verify the normal glomerular cells predominately expressed eNOS but not neuronal NOS (nNOS) and inducible NOS (iNOS). The second hypothesis was glomerular capillary endothelial cells in the tissue slice conditions used were capable of making NO from eNOS and would respond to bradykinin to increase NO production. We tested this hypothesis by direct measurement of the NO concentration with <10-µm outer diameter NO-sensitive microelectrodes and 4,5-diaminofluorescein diacetate (DAF-2 DA) dye to monitor intracellular formation of NO in glomeruli. Third, we hypothesized that hyperglycemia inhibition of eNOS activity would be through PKC mechanism, most likely PKC-. This hypothesis was tested using DAF-2 DA dye with confocal microscopy on renal slices by demonstration that activation of PKC lowered NO formation and that blockade of PKC- before a high-glucose exposure protected NO formation.
MATERIALS AND METHODS
Animals. Male ICR mice of 25-30 g body wt were used for the study (Harlan Industries, Indianapolis, IN). The experimental protocol was approved by the Institutional Animal Care and Use Committee of both the University of North Texas Health Science Center and Indiana University School of Medicine. Mice were anesthetized with an intraperitoneal injection of Inactin, a thiobarbiturate (100 mg/kg body wt). The abdominal aorta was cannulated to allow saline perfusion of the kidney to expel the blood, particularly red blood cells. The inferior vena cava was ligated above and below the renal veins and was cut open to allow egression of perfusion fluids. The kidney was perfused with 15-20 ml of ice-cold isotonic physiological saline (Ringer buffer containing in mM: 130 NaCl, 5 KCl, 2 CaCl 2, 1 MgSO 4, 1 Na 2 HPO 4, 20 HEPES, and 5.5 D -glucose, pH 7.4) for 5 min to remove all the red blood cells in the kidneys. The kidneys were removed and sliced with a vibrotome (World Precision Instruments, Sarasota, FL) into 150- to 200-µm sections. The tissue sections were immersed in ice-cold Ringer buffer until used for either confocal imaging or microelectrode NO measurements. Each renal slice mounted in a perfusion chamber on the microscope stage was superfused with Hanks' buffer (HBSS containing in mM: 2 CaCl 2, 5.37 KCl, 0.44 KH 2 PO 4, 0.49 MgCl 2, 0.406 MgSO 4, 136.89 NaCl, 4.166 NaHCO 3, 0.338 Na 2 HPO 4, 5 D -glucose, pH 7.3-7.4) with 100 µM L -arginine at 37°C. The bathing fluid was bubbled with 95% O 2 -5% CO 2.
Microelectrode measurements of glomerular NO production in a mouse renal slice. NO-sensitive microelectrodes were made and used based on techniques originally developed by Buerk et al. ( 9 ) and Friedemann et al. ( 17 ) and modified to suit our needs ( 6 ). The electrode is a 7- to 8-µm-diameter carbon fiber fully encased in a borosilicate glass (Fredrick Haer, Bowdoinham, ME) microelectrode during the micropipette pulling process. To both stabilize the carbon fiber and provide electrical contact from the fiber to a copper wire in the barrel of the electrode, the electrode barrel was partially filled with conductive epoxy cement (Dylon Industries, Cleveland, OH). The "cement" component of the epoxy flowed by capillary action along the carbon fiber to partially fill the pulled section of the microelectrode and strengthen the fiber-glass wall seal. The sharpened microelectrode tip composed of a carbon fiber with a thin covering of glass had an outer tip diameter of 10-12 µm at most at the end of the sharpened taper. Only the sharpened tip of the microelectrode was NO sensitive because the glass covered the carbon fiber. The electrode tip was coated by immersion in Nafion (Aldrich Chemical, Milwaukee, WI) to make a thin coating (<1 µm) when dried (15 min at 210°C). Nafion formed an electrochemical barrier to exclude negatively charged biological chemicals, such as nitrate, nitrite, and amino acids, from interacting with the electrode. We previously verified the value of this coating to minimize electrode interference by biological compounds ( 6 ). A World Precision Instruments carbon fiber reference electrode was used in the solution of the calibration cell and tissue support device. The microelectrode current was measured with a Keithley model 6517A electrometer (Cleveland, OH) that also supplied the polarization voltage of +0.9V. Microelectrodes were calibrated in a gas tonometer using gas mixtures of 0% NO (pure nitrogen) and 400 and 800 ppm NO in nitrogen for a calibration range of 0-1,200 nM dissolved NO at 37 ± 0.5°C. All microelectrodes had a linear relationship of microelectrode current to NO concentration (). Microelectrodes had a background current of 3-5 pA in nitrogen-equilibrated saline and generated an additional 1-2 pA/1,000 nM NO, which was equal to an increase in output voltage of 100-200 or greater mV/1,000 nM from the electrometer. The overall system can resolve 5- to 10-nM changes in , which was sensitive enough for our experimental purpose.
During tissue measurements, the microelectrode tip is placed 200 µm above the tissue surface to obtain a "0" . However, we could not detect NO in the bath until the microelectrode tip was placed within 50 µm above the tissue surface. In the example record shown in Fig. 1, the microelectrode first touched the glomerulus and then was slightly withdrawn to demonstrate that the NO concentration did change with proximity to the tissue. The vast majority of measurements were made by simply pressing the microelectrode tip against the center of the glomerulus. When the sharpened microelectrode tip touched the tissue surface, there was occasionally a brief mechanical artifact that rapidly dissipated. Thereafter, the measurements were quite stable. After the micropipette was completely withdrawn from the tissue into the bath, the current returned to essentially the baseline 0-nM current equivalent. To account for possible electronic drift of the microelectrode, the pre- and postmeasurement current in the bath was used to predict the virtual 0-nM baseline for any time point.
Fig. 1. Representative data of nitric oxide (NO)-sensitive microelectrode measurements of NO concentration () on renal glomerulus. A : microelectrode contacting the glomerulus followed by a slight withdrawal to reach a steady state. Thereafter, the slice was exposed to 1 mM N G -nitro- L -arginine methyl ester ( L -NAME) and the measured gradually declined toward the baseline current. B : was first measured at normal conditions containing 5 mM D -glucose and then the bath was changed to contain 30 mM D -glucose. The delay in the fall of with glucose represents the slow change of the bathing media followed thereafter by a decline in NO production. A and B : transient spike in the NO current at the end of each record represented withdrawal of the microelectrode to a point 200 µm above the tissue slice.
To minimize disturbances during recordings, various preheated solutions containing drugs were slowly added and simultaneously removed from the test chamber. This avoided thermal effects of the solutions and movement of the tissue slice relative to the micropipette. The total volume of the tissue bath was 15 ml, and the fluid flow was 3-4 ml/min. The microelectrodes were completely insensitive to the very slow motion of fluid flowing through the test chamber because stopping the flow momentarily had no effect on the recorded electrical current.
Immunofluorescent staining. To fix the kidney for immunofluorescent staining, the kidney was first perfused with the PBS to flush out the blood and then freshly prepared 2% paraformaldyhide in PBS for 10 min. The kidney was excised and immersed in the same fixative for 2 h at 4°C. The kidney was then rinsed with PBS and cryoprotected in 30% sucrose in PBS at 4°C for 24 h. Sections of 16-µm thick were cut with a cryostat at -20°C, mounted on glass slides, and immediately processed for immunofluorescent localization of NOS expression according to our previous published methods ( 11 ). Briefly, sections were rinsed 5 min with PBS, incubated in washing buffer (PBS containing 50 mM NH 4 Cl) for 2 x 10 min, and in blocking buffer (washing buffer with 2% BSA and 0.05% Saponin) for 20 min. Sections were incubated with respective primary antibody (5 µg/ml) or preimmune rabbit serum (1:100) in the blocking buffer overnight at 4°C. Primary antibodies used in these experiments are polyclonal antibodies developed in rabbits against eNOS (Sigma, St. Louis, MO; cat. no. N-2643), nNOS (Zymed, San Francisco, CA; cat. no. 61-7000), and iNOS (Transduction Labs, Lexington, KY; cat. no. N32030 ), respectively. After 4 x 5 min washing, sections were incubated in Alexa488 labeled goat anti-rabbit IgG (5 µg/ml; Molecular Probes, Eugene, OR) for 1 h at room temperature followed by 4 x 5 min washing including a 5-min treatment with 2 µg/ml propidium iodide (Molecular Probes) for nucleus staining. Sections were mounted with ProLong antifade medium (Molecular Probes) and imaged with Zeiss LSM510 confocal microscope. Samples were excited with a 488-nm Ar laser and emissions were detected at 505-530 nm (for Alexa488) and 620-680 nm (for propidium iodide) with two separate detectors. Laser power, pinhole size, and detector gain were the same for all samples.
Confocal imaging of fluorescent dye-loaded renal slice. To test the effect of different treatments on glomerular NO production, we preincubated renal slices (150-200 µm) from the same kidney with one of the following solutions for 30 min at 25°C. All solutions were based on the HBSS with 100 µM L -arginine but contained, respectively, 1 ) 5 mM D -glucose + 25 mM L -glucose as osmolar control, 2 ) 30 mM D -glucose, 3 ) 30 mM D -glucose + 100 nM ruboxistaurin for PKC- (both PKC- I and PKC- II) inhibition, 4 ) 5 mM D -glucose + 100 nM PMA for PKC activation, and 5 ) 5 mM D -glucose + 1 mM N G -nitro- L -arginine methyl ester ( L -NAME) for eNOS inhibition. These slices were loaded with 10 µM DAF-2 DA in their respective solutions for 30 min in dark conditions at 25°C, followed by rinsing with the respective solutions three times and kept on ice in dark conditions until imaged in 2 h. During confocal imaging, each slice was put into a chamber with at least 10 ml of media and warmed to 37°C by a heated water jacket on the microscope stage. Samples were excited with a 488-nm argon laser and the emission was detected at 510-550 nm for DAF-2 fluorescence with simultaneous transmitted light imaging by a separate detector. The confocal settings (laser power, pinhole size, detector gain) were set with a normal control sample to avoid saturation of the fluorescent signal and kept identical in the same experiment to get comparable results from samples exposed to different treatments. Calcein/AM (2 µM, non-ion-specific dye, Molecular Probes) was used as a dye-loading control for DAF-2 DA following the same experimental procedure.
Image analysis and statistical methods. Confocal images were analyzed with the Metamorph imaging analysis software (Universal Imaging, Dowingtown, PA). Fluorescence intensity of a glomerulus was obtained after subtraction of the background intensity. Because the background intensity from a glomerulus before dye loading is unpractical in our experiments, we chose a tissue area around the glomerulus under measurement as the background. These peripheral areas around glomeruli were not fluorescent in DAF-2 DA-loaded slices. Fluorescence intensity was expressed as gray scale units/pixel in single glomerulus. Fluorescence intensity of glomeruli in different treatments was normalized to the mean value of the normal control slice (5 mM D -glucose + 25 mM L -glucose) in a given experiment. These normalized values from different experiments were combined according to treatment groups (images of 6-10 glomeruli/treatment in 5 treatment groups/animal and 4 mice tested) for statistical analysis. For tissue slices in which DAF dye was used for comparison between separate slices of tissue, one-way ANOVA and Bonferroni's multiple comparison test were used to compare values obtained from different treatments, and P < 0.05 was used to determine the significance of difference between normal control and each treatment group. For measurements with microelectrodes in which multiple measurements of the same glomerulus were used before and after eNOS blocker treatment in Fig. 2, a two-way analysis of variance (normal vs. eNOS inhibitor and rest vs. bradykinin stimulation) was employed followed by a Tukey honestly significant difference test for individual comparisons. For microelectrode measurements in Fig. 3 of the simple effect of high glucose on the NO concentration in a glomerulus at rest and after acute exposure to high glucose, a simple t -test was used for statistical analysis. For microelectrode studies, P < 0.05 was accepted as a significant effect.
Fig. 2. NO was measured with microelectrodes that were in contact with the glomerulus surface, and then the tissue was exposed to various chemicals to alter the production of NO. One micromolar bradykinin was used to stimulate NO production before and after suppression of endothelial nitric oxide synthase (eNOS) with 0.5 mM L -NAME. L -NAME both decreased the resting concentration of NO and essentially eliminated the increase in NO caused by bradykinin. Data are based on sampling a single glomerulus in each of 7 mice (means ± SE, n = 7). * P < 0.05 vs. control.
Fig. 3. NO-sensitive microelectrode was in contact with the glomerulus surface to obtain a resting and then left in the place for 30 min as 30 mM D -glucose solution entered the fluid chamber. The data set is based on studies of 8 kidney slices from 4 mice (2 slices with a single glomerulus per mouse). As shown in Fig. 1 B, the decline in began within 10 min of exposure to glucose and was at a new steady state in 20-30 min. Data are means ± SE, n = 4. * P < 0.05 vs. control.
RESULTS
NO-sensitive microelectrode measurements of glomerular NO production. Figure 1 shows representative results of the microelectrode experiments during suppression of NO formation both with L -NAME and 30 mM D -glucose. Note that the initial NO recording was higher than the longer term recording after slightly withdrawing the microelectrode. This and many similar observations demonstrated that location in glomerulus influenced the measured. Consequently, in studies of this type, the electrode once placed should be kept at a constant location unless a particularly useful anatomic landmark is available for reference. In general, the use of hydraulic manipulators with no mechanical drift of their position and a vibration-free table was required to obtain highly stable recordings. The addition of 0.5 mM L -NAME (NOS inhibitor) inhibited the NO production from the glomerulus in as short as 10-15 min, but 30 min of exposure was more reliable. Exposure of glomeruli to 30 mM D -glucose, as shown in Fig. 1 B, caused as rapid and as severe depression of the as L -NAME. Although there is a report that high concentrations of D -glucose can directly scavenge NO ( 8 ), we found the remained decreased after the D -glucose was removed by washout from the bath.
As microelectrodes were highly sensitive to changes in bath temperature, the slow changes in during perturbations in Fig. 1 reflected the gradual replacement of the bathing fluid contained pharmacological agents rather than the time-dependent ability of the tissue or microelectrode to respond to a perturbation. We were also concerned that something in the tissue other than NO could activate the electrodes. As has been reported by our previous study using other tissues ( 6 ), we rapidly froze and rewarmed the tissue slices to kill and freeze-fracture cell membranes. After being rewarmed, the electrodes did not detect any signal other than the brief transients associated with the microelectrode tip touching the tissue surface. Therefore, any chemicals present in cells before and after freezing that might leak into the interstitium were unlikely to contribute to the signal recorded during in vitro conditions. The NO-sensitive microelectrodes were unable to respond to D -glucose in our calibration cells, and we assumed this lack of response transferred to the in vitro environment.
In Fig. 2, the was measured at rest and during topical exposure to 1 µM bradykinin, 0.5 mM L -NAME, and 0.5 mM L -NAME + 1 µM bradykinin. At rest, we found the vast majority of glomerular was between 200 and 250 nM, but only if 100 µM L -arginine, the typical rodent plasma concentration, was present. In the absence of L -arginine, is much lower and decreased quite rapidly after the tissues were warmed to 37°C. We did detect low concentrations (<100 nM) of NO from microvessels and some of the renal tubules in the kidney slices, as might be expected because immunocytochemistry revealed eNOS in endothelia in or near these structures. However, the glomerular tissues generated such high by comparison that we concentrated on their NO-regulatory system. Bradykinin (1 µM) topically applied to the kidney slices routinely caused about a 50% increase in that was sustained so long as bradykinin was present. L -NAME was used to suppress any NOS in the preparation. The averaged results of seven animals indicated that 30 min of L -NAME treatment caused the glomerular NO production to decrease 50% ( Fig. 2 ). After exposure to L -NAME, bradykinin was unable to stimulate an increase in the glomerular .
In eight kidney slices from four separate animals, the glomerular was measured at rest with 5 mM D -glucose and 30 min after exposure to 30 mM D -glucose. The microelectrode was placed on a glomerulus and not moved thereafter during the protocol. In every case, there was a progressive decline in the once D -glucose solution was applied over the tissues ( Fig. 3 ) over the following 15-30 min of exposure. The average decline in with D -glucose was actually greater than with L -NAME.
NOS expression in normal glomeruli. These studies using the three NOS antibodies were performed on sections of kidneys from three mice with comparable results in each animal. To identify the sources of NO from NOS expression in normal glomeruli, we performed immunofluorescent staining on normal mouse kidney with specific antibodies to eNOS, nNOS, and iNOS, respectively. Glomeruli were always identified by the eNOS-specific antibody with positive staining in the endothelial cells of the glomerular capillaries ( Fig. 4 A ). When treated with nNOS-specific antibody, the positive staining was only seen in the macula densa in the vascular pole of glomeruli but not inside glomeruli. In comparison, iNOS-specific antibody did not show positive staining either in the glomeruli or in renal tubule epithelial cells.
Fig. 4. Immunofluorescence of NOS isoform expression in normal renal cortex. Cryosections of normal mouse kidney were treated by specific antibodies to eNOS ( A ), neuronal (n)NOS ( B ), inducible (i)NOS ( C ), and preimmune serum ( D ), respectively. The red fluorescence is propidium iodide staining of the cell nucleus. Fluorescence (Alexa 488)-labeled secondary antibody binding to primary antibodies was used to visualize NOS localization (green). Based on 3 mice and multiple kidney slices from each mouse, the consistent finding was eNOS as the only NOS expressed in normal glomeruli ( A ), nNOS was present in only the macula densa (arrow in B ), and iNOS did not express in the normal glomeruli ( C ). Bar = 20 µm; g, glomeruli.
DAF-2 and confocal imaging show that glomeruli have active NO production. To confirm the microelectrode measurements and further differentiate spatial distribution of NO produced by glomerular eNOS and possibly of significant NO production by nNOS of macula densa, we used confocal microscopy imaging on the renal slices loaded with DAF-2 DA. The imaging analysis allowed high spatial resolution as to specific cellular sites of NO formation. In addition, the optical method allowed multiple glomeruli to be studied in the same experiment.
Confocal imaging showed that DAF-2 DA-loaded samples had bright fluorescence from glomeruli and vascular endothelial cells ( Fig. 5 A ) compared with an absence or very low fluorescence from tubular and parenchymal tissues. Figure 5 B demonstrates the fluorescence due to calcein/AM (nonion sensitive) in all cells compared with the brighter fluorescence of DAF-2 DA activated by NO in glomerular cells. Figure 6 A is a confocal image that demonstrated glomeruli were bright due to NO formation, whereas renal tubules were not at normal conditions. We confirmed that NO generation in cells loaded with the DAF-2 dye could be blocked by pharmacological suppression of eNOS with 1 mM L -NAME. As shown in Fig. 6 B, very low fluorescence due to NO formation was found in L -NAME-treated tissue. Then this same tissue slice was exposed to 1 mM sodium nitroprusside to confirm all cells were loaded with DAF-2 and the dye would detect NO from decomposition of nitroprusside. As shown in Fig. 6 C, both the glomerular and tubular cells had bright DAF-2 fluorescence in the presence of exogenous NO. These results indicated that DAF-2 can report cellular NO from either endogenous or exogenous NO sources. The limitation of a DAF-2 signal to glomeruli in Figs. 5 A and 6 A indicated little diffusion of NO from its site of production. As the DAF molecule is reported to respond to NO concentrations as low as 5 nM, diffusion of small concentrations of NO can be evaluated because all cells in the kidney take up the DAF dye as shown in Fig. 6 C.
Fig. 5. Glomeruli were the major structures showing strong NO production visualized by confocal imaging of 4,5-diaminofluorescein diacetate (DAF-2 DA)-loaded kidney slice. A : glomeruli and endothelial cells of arterioles (arrow) were positive for NO. Epithelial cells of renal tubules were generally negative in DAF-2 DA-loaded slices. B : calcein/AM-loaded sample from the same mouse showed homogeneous intracellular green fluorescence in all cells of the slice. Bar = 25 µm.
Fig. 6. Confocal imaging (merged images of fluorescence and transmitted light) of DAF-2 DA-loaded renal slices at rest and with L -NAME blockade of eNOS function. A : typical fluorescence due to NO production. B : tissue slice was exposed to 1 mM L -NAME, which effectively inhibited glomerular NO production. C : same tissue slice inhibited by the L -NAME that became bright after exposed to exogenous NO from 1 mM sodium nitroprusside (SNP). This indicated that L -NAME-treated tissues had loaded DAF-2 DA but the dye was not activated because it lacked biologically generated NO. This experiment was reproduced with similar results in tissue slices from 5 mice. Bar = 20 µm. D -glu, D -glucose.
High glucose inhibits glomerular NO production, and PKC- activity mediates the inhibition. We studied the effect of hyperglycemic D -glucose on glomerular NO production with the imaging method. The microelectrode studies in Figs. 1 and 3 indicated that high glucose suppressed NO formation, and we wished to confirm this observation with an alternative system, the DAF-2 reaction to NO. In addition, ruboxistaurin (previously LY-333531) was available to suppress PKC- (both PKC- I and PKC- II). This drug has been shown to improve peripheral vascular function during diabetes mellitus ( 29 ), including some renal tissues ( 23 ) in chronic studies, and acutely can protect in vivo endothelial cells from hyperglycemia ( 7, 36 ). This allowed us to determine if PKC- might be responsible for suppressed NO production when glomerular endothelial cells are exposed to high glucose. Examples of the images collected are shown in Fig. 7, and the analysis of these images is presented in Fig. 8 in terms of relative intensity measurements of DAF-2 fluorescence. We first tested that L -glucose would not cause a suppression of the NO signal, as shown in Fig. 7 A. Even 25 mM L -glucose with 5 mM D -glucose for metabolic support did not impair generation of NO monitored by DAF-2 fluorescence. However, as shown in Fig. 7 B, 30 mM D -glucose severely suppressed NO formation as registered by DAF-2. On a quantitative basis, the DAF-2 fluorescence intensity during 30 mM D -glucose was 30% of that under normal conditions, as shown in Fig. 8. This relative reduction in NO formation during high glucose with dye is comparable to that recorded with microelectrodes shown in Fig. 3. Figure 7 C presents an image of DAF-2 fluorescence during 30 mM D -glucose following treatment with 100 nM ruboxistaurin. The protected glomerular cells made NO despite the presence of high glucose. The third bar from the left in Fig. 8 presents the relative NO during the 30 mM D -glucose following treatment with 100 nM ruboxistaurin. The relative intensity was equivalent to normal conditions because PKC- suppression protected NO formation during exposure to high glucose. To illustrate that direct PKC activation could suppress NO formation, we used 100 nM PMA at normal D -glucose to activate PKC. The result was significantly decreased glomerular NO production as shown in Figs. 7 D and 8. Cells exposed to PMA did contain DAF-2 because exogenous NO from nitroprusside increased the DAF-2 fluorescence ( Fig. 7 E ). The suppressive effect of PMA on NO generation was so large that it was equivalent to that of slices exposed to 1 mM L -NAME, as shown by the data in Fig. 8. The glomerular NO production was inhibited 60% by PMA based on DAF-2 intensity measurements, and this was equivalent to the decrease in recorded by microelectrodes in Fig. 2 caused by high glucose acting through a PKC mechanism.
Fig. 7. Images of events shown in quantitative form in Fig. 8 using DAF-2 DA dye to monitor NO production. A and B : glomeruli exposed to 5 and 30 mM D -glucose, respectively, to demonstrate change under these conditions as monitored with DAF-2 dye. C : tissue was treated with 100 nM ruboxistaurin before exposure to 30 mM D -glucose and NO formation was present. D : tissue was exposed to 100 nM PMA to activate PKC and NO production was strongly suppressed. E : when tissues exposed to PMA were then exposed to exogenous NO from nitroprusside, the fluorescence of DAF-2 was found. This indicated that PMA did not interfere with DAF-2 loading and sensing intracellular . Bar = 20 µm.
Fig. 8. Quantitative analysis of the DAF-2 fluorescent intensity from studies similar to that shown for Fig. 7. The intensity of the glomeruli exposed to 5 mM D -glucose (5Dg) with 25 mM L -glucose was used as the control value to compare with intensities of glomeruli in sections of the same kidney. Both 30 mM D -glucose (30Dg), 0.5 mM L -NAME, and 100 nM PMA to activate PKC dramatically decreased fluorescence and by analogy, NO production. Ruboxistaurin (rub) treatment prevented 30 mM D -glucose inhibition of glomerular NO production. Quantization of DAF-2 fluorescence in 148 glomeruli (each treatment including glomeruli: 5Dg-41, 30Dg-26, 30Dg+rub-29, 5Dg+ L -name-21, and 5Dg+pma-31) from 4 mice was used for these data (means ± SE, n = 4). ** P < 0.01 vs. control of 5Dg treatment.
DISCUSSION
Correlation of NO measurements with microelectrodes and DAF-2. The first hypothesis tested in this study was to determine if glomeruli of kidney slices could make NO in an in vitro environment and if the regulation of the NO system would be reasonably intact. Direct measurement of the NO concentration with NO-sensitive microelectrodes and DAF-2 DA to monitor intracellular formation of NO in the glomerulus supports this hypothesis. The NO microelectrodes used have been evaluated in a number of in vivo studies and found to report changes in consistent with the current understanding of how eNOS is regulated ( 3, 9 ). As shown in Fig. 2, the NO signal was elevated above the resting concentration by bradykinin, suppressed by L -NAME inhibition of eNOS, and L -NAME eliminated the NO response to topical bradykinin. It is very important to point out that unless 100 µM L -arginine was present constantly in the bathing media, the signal for NO deteriorated rapidly after the tissue slices were warmed to 37°C. The use of 100 µM L -arginine was chosen simply because it is the typical plasma concentration of L -arginine in mammals. We would caution that the recorded may only be valid for the in vitro conditions and could be higher or lower in the in vivo state. Furthermore, the relative changes in NO rather than absolute NO concentrations best exemplify the regulation or suppression of the eNOS mechanism. Although it is not possible to compare directly with the DAF-2 dye measurement to the microelectrode recording of concentration, the relative changes in microelectrode output and dye fluorescence intensity during perturbations involving L -NAME and D -glucose were quite similar. For both techniques, L -NAME reduced the NO signal 50-60%, as shown in Figs. 2 and 8, and high glucose decreased the signal 60-70%, as shown in Figs. 3 and 8. These similar relative responses reinforce that both techniques likely report similar changes in during perturbations.
eNOS is the only NOS expressed in glomeruli of a normal mouse kidney. In the kidney, NO has been recognized an important mediator for renal functions ( 10, 13, 19, 46 ) and as would be expected, eNOS is expressed in human ( 1, 18 ) and mouse ( 47 ) glomerular endothelial cells. iNOS has also been found expressed in mesangial cells of glomeruli, especially under various pathological conditions ( 32, 44, 45 ). Our immunofluorescent results with isoform-specific antibodies to eNOS, nNOS, and iNOS confirmed our hypothesis that NO produced by glomerular cells should be primarily from eNOS ( Fig. 4 ). The nNOS was only expressed in the macula densa, and we could not detect iNOS in the normal glomeruli or renal tubules of the cortex. These results indicated that the in normal glomeruli we found with both NO-sensitive microelectrodes and DAF-2 was solely dependent on eNOS activity.
Inhibition of glomerular eNOS and NO production by PKC at high glucose. With two independent methods, the microelectrode measurements and NO-sensitive fluorescent dye with confocal microscopy, we showed that glomeruli have active in vitro NO production at a physiological D -glucose of 5 mM when sufficient L -arginine is present. As shown in Figs. 2, 6, and 8, eNOS inhibition with L -NAME caused a 50-60% decrease in the NO signal measured both by microelectrodes and DAF-2. In addition, before L -NAME, bradykinin could cause a large increase in that was fully blocked by L -NAME ( Fig. 2 ). NO formation was also very vulnerable to acute 30 mM D -glucose and the NO signal as measured by either microelectrode or DAF-2 was reduced 60-70% ( Figs. 3, 7, and 8 ).
The third hypothesis of the introduction that the suppression of NO by high glucose was associated with PKC activity in the glomerular endothelial cells was verified by two lines of evidence. First, PMA was used to activate PKC and the NO signal monitored with DAF-2 was decreased ( Figs. 7 and 8 ). Second, ruboxistaurin was used to suppress PKC- activity that is associated with hyperglycemia ( 23, 29 ) and this prevented the decline in NO signal measured with DAF-2 ( Figs. 7 and 8 ). These results are quite similar to what has been found by our previous in vivo studies of arteriolar production of NO during hyperglycemia ( 7 ). In those studies in the rat in vivo microvasculature with 15-25 mM D -glucose decreased the by about one-half in 30-45 min and vessels had a depressed dilation to topical acetylcholine or bradykinin. Ruboxistaurin given before hyperglycemia protected the NO formation at rest and during receptor activation with acetylcholine and bradykinin in the in vivo studies. This pattern of response in the in vivo arterioles and that of the in vitro glomerular NO formation during high glucose and PKC blockade ( Fig. 8 ) is quite similar. Therefore, the endothelial NO function in slices of renal tissue appeared to follow a pattern in which hyperglycemia rapidly suppressed eNOS function through a PKC mechanism quite like that in most other vascular beds. The results of our current study are also consistent with other reports indicating that transient hyperglycemia caused an increase of PKC activity in mouse embryos ( 21a ), human platelets ( 1a ), and mesangial cell cultures ( 1b ).
Mechanisms related to PKC-dependent eNOS inhibition have been explored by recent studies. Several reports indicated that PKC mediates phosphorylation of Thr495 (bovine, human, porcine, and rabbit endothelial cells) ( 16, 33, 34, 37 ) or Thr497 (bovine endothelial cells) ( 35 ) in eNOS calmodulin-binding domain correlated to inhibition of eNOS function and decreased NO production. In addition, PKC may inhibit L -arginine transport into endothelial cells causing decreased NO production in endothelial cells ( 49 ). Although a complete picture about mechanisms of PKC activation on eNOS function and NO production is unclear, it is evident from our NO measurements with DAF-2 dye that PKC activation with PMA will cause suppression of eNOS function. Our results further show for the first time that high glucose inhibits glomerular eNOS-mediated NO production through a PKC mechanism in a renal slice of a normal animal.
GRANTS
The study is supported by National Institutes of Health (NIH) Grant DK-064004, Showalter Awards of Indiana University School of Medicine, and the University of North Texas Health Science Center Faculty Research Fund to S. Chu; and NIH Grant HL-25824 to H. G. Bohlen.
ACKNOWLEDGMENTS
Ruboxistaurin was a gift from Eli Lilly to S. Chu and H. G. Bohlen, and we appreciate their generosity.
【参考文献】
Albrecht EW, Stegeman CA, Tiebosch AT, Tegzess AM, and van Goor H. Expression of inducible and endothelial nitric oxide synthases, formation of peroxynitrite and reactive oxygen species in human chronic renal transplant failure. Am J Transplant 2: 448-453, 2002.
Assert R, Scherk G, Bumbure A, Pirags V, Schatz H, and Pfeiffer AF. Regulation of protein kinase C by short term hyperglycaemia in human platelets in vivo and in vitro. Diabetologia 44: 188-195, 2001.
Ayo SH, Radnik R, Garono JA, Troyer DA, and Kreisberg JI. High glucose increases diacylglycerol mass and activates protein kinase C in mesangial cell cultures. Am J Physiol Renal Fluid Electrolyte Physiol 261: F571-F577, 1991.
Beckman JA, Goldfine AB, Gordon MB, Garrett LA, and Creager MA. Inhibition of protein kinase C prevents impaired endothelium-dependent vasodilation caused by hyperglycemia in humans. Circ Res 90: 107-111, 2002.
Bohlen HG. Mechanism of increased vessel wall nitric oxide concentrations during intestinal absorption. Am J Physiol Heart Circ Physiol 275: H542-H550, 1998.
Bohlen HG. Protein kinase II in Zucker obese rats compromises oxygen and flow-mediated regulation of nitric oxide formation. Am J Physiol Heart Circ Physiol 286: H492-H497, 2004.
Bohlen HG and Lash JM. Topical hyperglycemia rapidly suppresses EDRF-mediated vasodilation of normal rat arterioles. Am J Physiol Heart Circ Physiol 265: H219-H225, 1993.
Bohlen HG and Nase GP. Dependence of intestinal arteriolar regulation on flow-mediated nitric oxide formation. Am J Physiol Heart Circ Physiol 279: H2249-H2258, 2000.
Bohlen HG and Nase GP. Arteriolar nitric oxide concentration is decreased during hyperglycemia-induced II PKC activation. Am J Physiol Heart Circ Physiol 280: H621-H627, 2001.
Brodsky SV, Morrishow AM, Dharia N, Gross SS, and Goligorsky MS. Glucose scavenging of nitric oxide. Am J Physiol Renal Physiol 280: F480-F486, 2001.
Buerk DG, Riva CE, and Cranstoun SD. Nitric oxide has a vasodilatory role in cat optic nerve head during flicker stimuli. Microvasc Res 52: 13-26, 1996.
Cachofeiro V, Fortepiani LA, Navarro-Cid J, Lahera V, and Garcia-Estan J. Renal dysfunction after chronic blockade of nitric oxide synthesis. Antioxid Redox Signal 4: 885-891, 2002.
Chu J, Chu S, and Montrose MH. Apical Na + /H + exchange near the base of mouse colonic crypts. Am J Physiol Cell Physiol 283: C358-C372, 2002.
Craven PA, Studer RK, Negrete H, and DeRubertis FR. Protein kinase C in diabetic nephropathy. J Diabetes Complications 9: 241-245, 1995.
D'Amours M, Lebel M, Grose JH, and Lariviere R. Renal and vascular effects of chronic nitric oxide synthase inhibition: involvement of endothelin 1 and angiotensin II. Can J Physiol Pharmacol 77: 8-16, 1999.
DeRubertis FR and Craven PA. Activation of protein kinase C in glomerular cells in diabetes: mechanisms and potential links to the pathogenesis of diabetic glomerulopathy. Diabetes 43: 1-8, 1994.
Fadnes HO and Aukland K. Protein concentration and colloid osmotic pressure of interstitial fluid collected by the wick technique. Microvasc Res 14: 11-25, 1977.
Fleming I, Fisslthaler B, Dimmeler S, Kemp BE, and Busse R. Phosphorylation of Thr(495) regulates Ca 2+ /calmodulin-dependent endothelial nitric oxide synthase activity. Circ Res 88: E68-E75, 2001.
Friedemann MN, Robinson SW, and Gerhardt GA. o -Phenylenediamine-modified carbon fiber electrodes for the detection of nitric oxide. Anal Chem 68: 2621-2628, 1996.
Furusu A, Miyazaki M, Abe K, Tsukasaki S, Shioshita K, Sasaki O, Miyazaki K, Ozono Y, Koji T, Harada T, Sakai H, and Kohno S. Expression of endothelial and inducible nitric oxide synthase in human glomerulonephritis. Kidney Int 53: 1760-1768, 1998.
Galle J and Wanner C. Impact of nitric oxide on renal hemodynamics and glomerular function: modulation by atherogenic lipoproteins? Kidney Blood Press Res 19: 2-15, 1996.
Haneda M, Koya D, and Kikkawa R. Cellular mechanisms in the development and progression of diabetic nephropathy: activation of the DAG-PKC-ERK pathway. Am J Kidney Dis 38: S178-S181, 2001.
Heeringa P, van Goor H, Itoh-Lindstrom Y, Maeda N, Falk RJ, Assmann KJ, Kallenberg CG, and Jennette JC. Lack of endothelial nitric oxide synthase aggravates murine accelerated anti-glomerular basement membrane glomerulonephritis. Am J Pathol 156: 879-888, 2000.
Hiramatsu Y, Sekiguchi N, Hayashi M, Isshiki K, Yokota T, King GL, and Loeken MR. Diacylglycerol production and protein kinase C activity are increased in a mouse model of diabetic embryopathy. Diabetes 51: 2804-2810, 2002.
Hoshiyama M, Li B, Yao J, Harada T, Morioka T, and Oite T. Effect of high glucose on nitric oxide production and endothelial nitric oxide synthase protein expression in human glomerular endothelial cells. Nephron Exp Nephrol 95: e62-e68, 2003.
Ishii H, Jirousek MR, Koya D, Takagi C, Xia P, Clermont A, Bursell SE, Kern TS, Ballas LM, Heath WF, Stramm LE, Feener EP, and King GL. Amelioration of vascular dysfunctions in diabetic rats by an oral PKC B inhibitor. Science 272: 728-731, 1996.
Ishii H, Koya D, and King GL. Protein kinase C activation and its role in the development of vascular complications in diabetes mellitus. J Mol Med 76: 21-31, 1998.
Jover B and Mimran A. Nitric oxide inhibition and renal alterations. J Cardiovasc Pharmacol 38, Suppl 2: S65-S70, 2001.
Kasai N, Sugimoto K, Horiba N, and Suda T. Effect of D -glucose on nitric oxide release from glomerular endothelial cells. Diabetes Metab Res Rev 17: 217-222, 2001.
Keynan S, Hirshberg B, Levin-Iaina N, Wexler ID, Dahan R, Reinhartz E, Ovadia H, Wollman Y, Chernihovskey T, Iaina A, and Raz I. Renal nitric oxide production during the early phase of experimental diabetes mellitus. Kidney Int 58: 740-747, 2000.
Komers R and Anderson S. Paradoxes of nitric oxide in the diabetic kidney. Am J Physiol Renal Physiol 284: F1121-F1137, 2003.
Koya D, Haneda M, Nakagawa H, Isshiki K, Sato H, Maeda S, Sugimoto T, Yasuda H, Kashiwagi A, Ways DK, King GL, and Kikkawa R. Amelioration of accelerated diabetic mesangial expansion by treatment with a PKC inhibitor in diabetic db / db mice, a rodent model for type 2 diabetes. FASEB J 14: 439-447, 2000.
Koya D and King GL. Protein kinase C activation and the development of diabetic complications. Diabetes 47: 859-866, 1998.
Lash JM, Nase GP, and Bohlen HG. Acute hyperglycemia depresses arteriolar NO formation in skeletal muscle. Am J Physiol Heart Circ Physiol 277: H1513-H1520, 1999.
Lee HY, Noh HJ, Gang JG, Xu ZG, Jeong HJ, Kang SW, Choi KH, and Han DS. Inducible nitric oxide synthase (iNOS) expression is increased in lipopolysaccharide (LPS)-stimulated diabetic rat glomeruli: effect of ACE inhibitor and angiotensin II receptor blocker. Yonsei Med J 43: 183-192, 2002.
Lenasi H, Kohlstedt K, Fichtlscherer B, Mulsch A, Busse R, and Fleming I. Amlodipine activates the endothelial nitric oxide synthase by altering phosphorylation on Ser1177 and Thr495. Cardiovasc Res 59: 844-853, 2003.
Li Z, Carter JD, Dailey LA, and Huang YC. Vanadyl sulfate inhibits NO production via threonine phosphorylation of eNOS. Environ Health Perspect 112: 201-206, 2004.
Matsubara M, Hayashi N, Jing T, and Titani K. Regulation of endothelial nitric oxide synthase by protein kinase C. J Biochem (Tokyo) 133: 773-781, 2003.
Mayhan WG and Patel KP. Acute effects of glucose on reactivity of cerebral microcirculation: role of activation of protein kinase C. Am J Physiol Heart Circ Physiol 269: H1297-H1302, 1995.
Michell BJ, Chen Z, Tiganis T, Stapleton D, Katsis F, Power DA, Sim AT, and Kemp BE. Coordinated control of endothelial nitric oxide synthase phosphorylation by protein kinase C and the cAMP-dependent protein kinase. J Biol Chem 276: 17625-17628, 2001.
Mogensen CE. Microalbuminuria and hypertension with focus on type 1 and type 2 diabetes. J Intern Med 254: 45-66, 2003.
Mumtaz FH, Dashwood MR, Khan MA, Thompson CS, Mikhailidis DP, and Morgan RJ. Downregulation of nitric oxide synthase in the diabetic rabbit kidney: potential relevance to the early pathogenesis of diabetic nephropathy. Curr Med Res Opin 20: 1-6, 2004.
Noh H, Ha H, Yu MR, Kang SW, Choi KH, Han DS, and Lee HY. High glucose increases inducible NO production in cultured rat mesangial cells. Possible role in fibronectin production. Nephron 90: 78-85, 2002.
Onozato ML, Tojo A, Goto A, Fujita T, and Wilcox CS. Oxidative stress and nitric oxide synthase in rat diabetic nephropathy: effects of ACEI and ARB. Kidney Int 61: 186-194, 2002.
Prabhakar SS. Tetrahydrobiopterin reverses the inhibition of nitric oxide by high glucose in cultured murine mesangial cells. Am J Physiol Renal Physiol 281: F179-F188, 2001.
Rabkin R. Diabetic nephropathy. Clin Cornerstone 5: 1-11, 2003.
Sade K, Schwartz D, Wolman Y, Schwartz I, Chernichovski T, Blum M, Brazowski E, Keynan S, Raz I, Blantz RC, and Iaina A. Time course of lipopolysaccharide-induced nitric oxide synthase mRNA expression in rat glomeruli. J Lab Clin Med 134: 471-477, 1999.
Saura M, Lopez S, Rodriguez PM, Rodriguez PD, and Lamas S. Regulation of inducible nitric oxide synthase expression in rat mesangial cells and isolated glomeruli. Kidney Int 47: 500-509, 1995.
Shao J, Miyata T, Yamada K, Hanafusa N, Wada T, Gordon KL, Inagi R, Kurokawa K, Fujita T, Johnson RJ, and Nangaku M. Protective role of nitric oxide in a model of thrombotic microangiopathy in rats. J Am Soc Nephrol 12: 2088-2097, 2001.
Teichert AM, Karantzoulis-Fegaras F, Wang Y, Mawji IA, Bei X, Gnanapandithen K, and Marsden PA. Characterization of the murine endothelial nitric oxide synthase promoter. Biochim Biophys Acta 1443: 352-357, 1998.
Williams SB, Goldfine AB, Timimi FK, Ting HH, Roddy MA, Simonson DC, and Creager MA. Acute hyperglycemia attenuates endothelium-dependent vasodilation in humans in vivo. Circulation 97: 1695-1701, 1998.
Zharikov SI, Herrera H, and Block ER. Role of membrane potential in hypoxic inhibition of L -arginine uptake by lung endothelial cells. Am J Physiol Lung Cell Mol Physiol 272: L78-L84, 1997.
作者单位：1 Department of Cell Biology and Genetics, University of North Texas Health Science Center, Fort Worth, Texas 76107; and 2 Department of Cellular and Integrative Physiology, Indiana University School of Medicine, Indianapolis, Indiana 46202