INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

IN THE KIDNEY, tubular injury resulting from ischemic or nephrotoxic insult elicits a regenerative response, restoring the structural and functional integrity of tubular epithelium (7, 9). Nephrogenic repair occurring after renal tissue damage involves a rise of mitotic activity, as well as a transient dedifferentiation in the epithelium lining renal tubules. Proximal tubule segments, which are particularly sensitive to ischemia and nephrotoxins, are also those where regeneration is the most frequently observed. Regenerative hyperplasia in injured renal tubules may be accompanied by hyperplastic foci in kidney interstitium, especially in cases of severe tubular necrosis (15).

In an early study focusing on the mechanism of renal tubular regeneration (24), we used a combined in vitro/in vivo approach to show that the proliferative response consecutive to gentamicin-induced renal tissue injury could occur in absence of serum factors. This observation led us to propose that nephrogenic repair is regulated by endogenous factors acting in paracrine or autocrine fashion, with little if any involvement of endocrine mediators (15). At that time, polypeptide growth factors already appeared as likely candidates for the control of tubular regeneration.

As a result of the growing interest in the field during the last decade, various experimental approaches have been followed to identify growth factors involved in the control of kidney tubular regeneration. In a number of studies, the authors have examined the enhancing effect of growth factor administration on the process of regenerative hyperplasia and the recovery of renal function after kidney ischemia or nephrotoxin-induced acute tubular necrosis (3, 8, 19, 20, 21). Other studies, including ours, have concentrated on the modifications undergone by growth factors or their cognate receptors during a regenerative response in renal tissue (4, 13, 16, 22, 25, 26, 28-30). Finally, the role of growth factors in nephrogenic repair has been also examined by observing changes in the expression of growth factor or growth factor receptor genes after acute tubular necrosis (4, 12, 13, 31). These different lines of investigation have led to the identification of several growth factors, such as epidermal growth factor (EGF), insulin-like growth factor (IGF)-I, and hepatocyte growth factor (HGF), as possible mediators in nephrogenic repair, although some data are still a matter of debate and must be interpreted with caution. Indeed, the finding that a given growth factor promotes renal tissue repair when administered after acute tubular injury does not necessarily imply that the same factor is produced in response to injury and that it plays a role in the regenerative process. Besides, measurement of growth factors or of the expression of the corresponding genes gives little indication concerning their biological effect on renal cells. Thus we decided to gain further insight into nephrogenic repair by utilizing an alternative method that would allow us to detect in renal tissue those soluble compounds endowed with mitogenic properties.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Treatment of animals and sampling of renal tissue. In vivo studies were carried out on female Wistar rats (160-180 g body wt) supplied by a commercial breeding farm (Iffa Credo, L'Arbresle, France). The animals were taken care of and treated in compliance with the guidelines specified by the Belgian Ministry of Middle Class and Agriculture (license no. LA1500021). Upon their arrival, the rats were transferred to an animal facility, submitted to a regular 12:12-h light-dark cycle, and allowed 1 wk of acclimation before starting gentamicin treatment. During this period and until termination, standard rat chow and tap water were provided ad libitum.

Rats were treated with the gentamicin sulfate formulation meant for parenteral administration in clinics (Géomycine; Schering-Plough, Brussels, Belgium). The original preparation was diluted in 0.9% NaCl to administer the drug in a volume of ~0.5 ml/animal. The actual amount of gentamicin solution injected intraperitoneally was adjusted according to the body weight of each individual, which was determined immediately before dosing. A daily dose of 50 mg/kg or 100 mg/kg was delivered twice daily for four consecutive days. Control (sham treated) animals received on the same schedule an equivalent volume of intraperitoneal 0.9% NaCl. Rats were killed either 16 h (referred to as day 1) or 88 h (referred to as day 4) after last gentamicin injection.

The experimental animals were terminated by decapitation and exsanguination under CO₂ narcosis. Both kidneys were exposed by a midline incision of the abdominal wall, removed, and bisected. Renal necropsies containing cortex and outer stripe of outer medulla tissue were excised by sharp dissection and snap frozen in dry ice. These samples were kept frozen at 20°C until further processing. For morphological study, one-quarter of a kidney was fixed by immersion in Duboscq-Brazil fluid. In some experiments, blood recovered during exsanguination was allowed to clot at room temperature, and the serum was recovered for bioassay.

Processing of renal tissue for morphological analysis. Fixed tissue specimens were dehydrated in graded ethanol solutions, finished in butanol, and embedded in paraffin wax (Paraplast Plus) following routine procedure. Paraffin sections of 4-5 µm thickness were cut on a Reichert Autocut 2040 microtome and mounted on glass slides. Histopathological examination of renal tissue was performed after staining with periodic acid-Schiff (PAS), hemalun, and luxol fast blue.

For the immunocytochemical demonstration of proliferating cell nuclear antigen (PCNA), dewaxed tissue sections were rehydrated in distilled water and were pretreated with 10% HCl (60 min at 60°C) to unmask hidden epitopes. Endogenous peroxidase activity in tissue was quenched by a 5-min exposure to 0.5% H₂O₂. Thereafter, the sections were rinsed in phosphate-buffered saline (PBS; 0.04 M Na₂HPO₄, 0.01 M KH₂PO₄, 0.12 M NaCl, pH 7.2) and incubated in PBS containing 5% normal goat serum (NGS) to prevent nonspecific adsorption during immunostaining. The primary antibody applied to the sections was a monoclonal mouse anti-PCNA (clone PC10, 1:40 dilution in PBS containing 5% NGS). After a rinse in PBS, this was followed by exposure to biotinylated goat anti-mouse immunoglobulin G (IgG) antibodies (1:50 dilution in PBS, 5% NGS). Immunocomplexes were revealed by incubation for 30 min in presence of biotinylated peroxidase-avidin (ABC) complexes. Bound peroxidase activity was visualized by incubation in a solution of 0.02% 3,3'-diaminobenzidine, 0.01% H₂O₂ in PBS. Finally, sections were counterstained with PAS, hemalun, and luxol fast blue. The specificity of immunolabeling was checked in preliminary experiments by omitting the first layer of antibody. No immunostaining was observed in this case.

Assessment of cell proliferation. Renal tissue sections (1 section/experimental animal) immunostained for PCNA were observed at ×400 magnification to estimate the number of nuclei displaying PCNA immunoreactivity. The evaluation procedure was standardized by fitting into one of the microscope eyepieces an additional lens engraved with a 1-cm² square grid that covered an actual area of 0.081 mm² in the microscopic field. Each tissue section was scanned until a total of 20 fields (i.e., a total surface of 1.62 mm²) was attained. The number of PCNA-positive nuclei was expressed per unit surface of tissue section.

Preparation of renal tissue extracts. Renal tissue necropsies (0.8-1.4 g wet wt) were thawed and homogenized in 9 vol of ice-cold distilled water using a Potter-Elvehjem tissue grinder equipped with a motor-driven Teflon pestle. All subsequent steps of the procedure were run at 0-4°C. After a 30-min incubation, the homogenate was centrifuged for 30 min at 100,000 g. The resulting supernatant was recovered, and the sediment was suspended in the same volume of distilled water as that used for homogenization. The preparation was spun in the same conditions as above, and the supernatant was added to that of the first centrifugation. Pooled supernatants were lyophilized. The dry residue was suspended in 6 vol of 0.2 M HCl and centrifuged for 30 min at 100,000 g. The supernatant was neutralized with 1 M NaOH and lyophilized again. The residue was suspended in a volume of Dulbecco's modified Eagle's medium (DMEM) equivalent to four times the wet weight of renal tissue and spun for 30 min at 100,000 g. The final supernatant (referred to as renal extract) was sterilized by serial passages through 0.8-µm and 0.22-µm filters and stored at 20°C. The protein content of renal extracts was determined before bioassays on cell cultures.

EGF radioimmunoassay. The level of soluble EGF immunoreactivity in renal tissue extracts was determined by a double-antibody nonequilibrium radioimmunoassay (RIA), as described in a previous publication (16).

Cell cultures. The biological activity of growth factors and renal tissue extracts was assayed on the established cell line NRK-52E (normal rat kidney, ATCC no. 1571). Culture studies were performed at 37°C in a normal atmosphere, humidified cell incubator. For routine maintenance, cells were cultured in 75-cm² flasks containing DMEM supplemented with 10% fetal calf serum (FCS), 20 mM N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid, 100 U/ml penicillin, 100 µg/ml streptomycin, and 0.25 µg/ml amphotericin B. Cells were passed once a week, with a renewal of the culture medium 2 days after seeding. For passages and measurement of growth, cells were dislodged from the vessel bottom by treatment with trypsin/EDTA and counted in a model Z1 cell counter (Coulter Electronics, Luton, UK).

Assay of mitogenic activity. Cells in DMEM + 10% FCS were plated at an initial density of 2 × 10⁴ cells/cm² in 12-well plates (day 0). At day 2, the seeding medium was replaced by fresh DMEM of the same composition. At day 9, serum-containing DMEM was changed for serum-free DMEM-Ham's F-12 medium supplemented with 30 mM NaSeO₃ and 10 µg/ml transferrin. The mitogenic activity of growth factors and renal tissue extracts was tested at day 12, by exposing for 26 h the growth-arrested cells to these additives and measuring the amount of [³H]thymidine (1 µCi/ml medium) incorporated into cell DNA during a period of 4-8 h preceding the end of exposure. The effect of anti-EGF or anti-IGF-I on the mitogenic activity of growth factors and tissue extracts was examined by preincubating the latter for 72 h in presence of these antisera.

To evaluate the rate of [³H]thymidine incorporation, the cell layer was rinsed with Dulbecco's PBS after aspiration of the culture medium and treated for 10 min with 2 ml of ice-cold 10% trichloroacetic acid (TCA). The supernatant was discarded, and residual acid soluble radioactivity was washed out by two rinses with the same volume of 10% TCA. The residue containing acid-insoluble radioactivity was solubilized by adding 2 ml of 2.5% sodium dodecyl sulfate in 0.2 M NaOH. The cell lysate was recovered after a 30-min incubation at room temperature and assayed for protein content and radioactivity. Radioactivity of cell lysates was monitored in 0.5-ml aliquots by scintillation spectrometry in a Tri-Carb 1600CA liquid scintillation counter (Canberra Packard, Meriden CT), after mixing with 10 ml of scintillation cocktail (Instagel Plus, Canberra Packard).

Binding studies. The analysis of EGF binding to NRK-52E cells was carried out using ¹²⁵I-labeled mouse EGF. At day 0, cells were seeded in 12-well plates at a concentration of 4 × 10⁴ cells/cm² in DMEM + 10% FCS. The medium was replaced by fresh medium of the same composition at day 2. At day 6, serum-free DMEM containing 0.1% bovine serum albumin was substituted for serum-containing medium. At day 7, the medium was changed for fresh medium containing ¹²⁵I-EGF at concentrations varying from 0.001 to 0.9 nM, and the cell culture dishes were transferred to 4°C. After an incubation of 48 h, cell-bound and soluble radioactivities were measured in cell lysates and the culture medium, respectively. Specific EGF binding was calculated by subtracting values of cell-associated radioactivity measured in presence of a 100-fold excess of unlabeled EGF. A similar approach was followed for the analysis of IGF-I binding, except that the cells were exposed for 16 h to ligand concentrations varying from 0.1 to 52 µM.

Protein assay. Protein concentration was measured by the method of Lowry et al. (17), using bovine serum albumin as a reference. No NaOH was included in reagent A for the solubilization of proteins.

Immunochemicals, radiolabeled compounds, and other reagents. Polyclonal anti-rat EGF antiserum was raised in rabbits by immunization with rat submaxillary gland EGF purified according to the method of Schaudies and Savage (27). Priming of the animals was achieved by subcutaneously injecting 100 µg of EGF in complete Freund's adjuvant. Primary injection was followed by several boosts of 100 µg EGF in incomplete Freund's adjuvant given at 1-mo intervals. Blood samples were collected 14 days after immunogen administration. Immunization and blood collection were carried out in the Laboratory of Hormonology (CER, Marloie, Belgium). Immunoglobulins were purified from rabbit serum when its antibody titer, as evaluated by indirect enzyme-linked immunosorbent assay, proved to be satisfactory. Purification was performed by volume exclusion using the E-Z-SEP kit of Middlesex Sciences (Pharmacia Biotech Benelux, BJ Roosendaal, The Netherlands). Rat IGF-I and rabbit anti-human IGF-I antiserum cross-reacting with rat IGF-I were purchased from GroPep (Adelaide, Australia). Monoclonal mouse anti-PCNA (clone PC10), biotinylated goat anti-mouse IgG antibodies, and ABC complex kit came from Dako (Glostrup, Denmark). NGS was obtained from the Laboratory of Hormonology (CER, Marloie, Belgium). [³H]thymidine (5 Ci/mmol), ¹²⁵I-EGF (~66 mCi/mg), and ¹²⁵I-IGF-I (~260 mCi/mg) were supplied by Amersham Life Science (Ghent, Belgium). DMEM was purchased from ICN/Flow Laboratories (Asse-Relegem, Belgium). DMEM-Ham's F-12, as well as FCS and other culture supplements, was obtained from Life Technologies (Merelbeke, Belgium). Tyrphostin A46 [-cyano-(3,4-dihydroxy)-cinnamide] came from Calbiochem/Euro Biochem (Bierges, Belgium). Other reagents were obtained from commercial sources and were of analytical grade, except for those used in routine histological procedures.

Statistical analysis. When no more than two groups were considered for comparison, statistical analysis was performed using unpaired two-tailed Student's t-test. Comparison of more than two groups was achieved by analysis of variance (ANOVA), followed by Tukey's post hoc test. Significance level was set at 0.05.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Animal studies. Light microscopy examination of renal cortex tissue after histological staining (Fig. 1, A, C, and E) confirmed the development of nephrotoxic insult in the kidneys of rats exposed to gentamicin. Morphological abnormalities and lesions were mostly observed in proximal tubules, with no or little involvement of distal tubules, collecting ducts, and glomeruli. At the gentamicin dosage of 50 mg/kg, sections of proximal tubules exhibited evidence of cell desquamation and the presence of cell debris in the lumen (Fig. 1C). In less affected proximal tubules, epithelial cells appeared swollen and contained conspicuous PAS-positive inclusions (Fig. 1C). After administration of 100 mg/kg gentamicin, proximal tubule damage became more extensive, as indicated by clusters of necrotic tubule sections with a more or less denuded basement membrane (Fig. 1E). Necrotic injury induced by 100 mg/kg gentamicin was occasionally accompanied by peritubular infiltration.

View larger version (163K): [in this window] [in a new window]   Fig. 1.   Tubular damage and regeneration following gentamicin administration. A, C, and E: histological staining with periodic acid-Schiff, hemalun, and luxol fast blue (magnification, ×190). A: light microscopy appearance of renal cortex in a sham-treated rat. C: tubular injury induced by 50 mg/kg gentamicin (killed at day 1). Cell necrosis and desquamation can be observed in several proximal tubule sections. E: tubular injury after 100 mg/kg gentamicin (killed at day 1). Tubule sections exhibit an almost completely denuded basement membrane. Cell debris appears in lumina. Note that the glomerulus and distal tubules remain unremarkable. B, D, and F: immunostaining of proliferating cell nuclear antigen (PCNA) and counterstaining as above (magnification, ×480). B: renal cortex in a sham-treated rat. A few PCNA-labeled nuclei are seen in proximal tubule sections and a collecting duct. D: proximal tubule showing evidence of regenerative hyperplasia. Epithelial cells are flattened and display an ill-defined brush border, and several cells in the section exhibit PCNA immunolabeling (50 mg/kg gentamicin, killed at day 4). F: section of a regenerating tubule with a dedifferentiated epithelium and prominent PCNA immunolabeling (100 mg/kg gentamicin, killed at day 4).

Increase of cell proliferation subsequent to gentamicin-induced tubular necrosis was revealed by the immunostaining of PCNA, a DNA polymerase -accessory protein whose concentration increases in proliferating cells. As shown in Fig. 2, gentamicin administration resulted in a marked elevation of nuclei staining with anti-PCNA antibody. The increase of proliferative activity started as soon as 1 day after the cessation of gentamicin administration. At day 4, many tubular profiles in renal cortex were lined with a flattened, dedifferentiated epithelium where cells showed no or an ill-defined brush border. These profiles invariably contained more PCNA-positive nuclei compared with intact proximal tubules in controls (Fig. 1, B and D). Tubular profiles with morphological signs of regeneration and numerous PCNA-positive nuclei were particularly prominent in kidneys of rats treated with 100 mg/kg gentamicin (Fig. 1F).

View larger version (25K): [in this window] [in a new window]   Fig. 2.   Regenerative hyperplasia consecutive to gentamicin-induced tubular injury. PCNA-positive cells were counted and expressed per unit surface, as detailed in METHODS. Columns show mean values (error lines are ±SD) of treated and control groups (n = 5-8 animals). * Values significantly higher than mean control value (ANOVA and Tukey's post hoc test).

In a rat model of nephrogenic repair after renal ischemia, we have previously shown that tubular regeneration is associated with a decrease of tissue-bound EGF in renal tissue and a peak of low-molecular-weight EGF in kidney soluble fractions (26). In the current study, soluble immunoreactive EGF was determined by RIA in renal tissue extracts prepared from gentamicin-treated rats and control animals. As illustrated by Fig. 3, the level of EGF immunoreactivity was higher in kidney extracts obtained from animals exposed to gentamicin, irrespective of the dosage. Thus renal tubular injury/regeneration was accompanied by an elevation of soluble renal EGF.

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Levels of soluble epidermal growth factor (EGF) in renal tissue extracts. Immunoreactive EGF (irEGF) was determined by radioimmunoassay in renal tissue extracts prepared from animals belonging to each group (n = 5-9). Each value is the mean of 3 determinations.

Cell culture studies. A typical growth curve, as well as the kinetic behavior of NRK-52E cells grown in our experimental conditions, is illustrated in Fig. 4. As shown by the time course measurement of cell number and proteins in cultures (Fig. 4, A and B), cell populations reached a plateau ~9-10 days after plating. As could be expected, the slowing down of cell growth occurred in parallel with a considerable reduction of [³H]thymidine incorporation into cell DNA (Fig. 4C). Thus the rate of DNA synthesis at 10 days was only 5% of that evaluated during exponential growth. For the measurement of mitogenic activity associated with growth factors and renal tissue extracts, serum-free medium with selenite and transferrin was substituted for serum-containing medium at day 9, i.e., 4 days before the measurement of DNA synthesis.

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Growth curves and proliferative activity of normal rat kidney NRK-52E cells. Cells were plated in complete medium at an initial density of 2 × 104 cells/cm2. A and B: growth curves determined by cell counting and protein assay, respectively. C: proliferative activity assessed by [3H]thymidine incorporation into cell DNA. Radioactivity was corrected according to the amount of cell protein, since cell number increased during the period of observation. Arrow indicates time when serum-free medium with selenite and transferrin was substituted for serum-containing medium. Each value is the mean of 4 determinations (error lines are ±SD).

Figure 5 shows the effect of purified growth factors on [³H]thymidine incorporation into DNA of growth-arrested NRK-52E cells. Both EGF and IGF-I exerted an enhancing effect on DNA synthesis. In this regard, EGF appeared more potent than IGF-I, since a half-maximum effect was observed at EGF concentrations around 1.7 ng/ml medium, vs. ~5 ng/ml in the case of IGF-I. Similar data were obtained when cell numbers were evaluated instead of DNA synthesis (not shown). This shows that growth-arrested NRK-52E cells resume their proliferation upon the addition of EGF or IGF-I to serum-free medium.

View larger version (13K): [in this window] [in a new window]   Fig. 5.   Response of growth-arrested NRK-52E cells to growth factors. Effect of EGF (A) and insulin-like growth factor I (IGF-I, B) on [3H]thymidine incorporation was tested on cells made quiescent by serum deprivation, as indicated in METHODS. Radioactivity was corrected according to the amount of cell protein. Each value is the mean of 4 determinations (error lines are ±SD).

The specificity of the mitogenic effect exhibited by EGF and IGF-I was confirmed by studying the binding of these growth factors to NRK-52E cells. Radiolabeled EGF bound in a saturable fashion to the cell surface. Scatchard analysis of the saturation curve generated a curvilinear plot suggesting the presence of two binding sites with different K_d and B_max values (Fig. 6A). We estimated a K_d of 9.3 pM and a B_max of 23,565 receptors/cell for the high-affinity site and a K_d of 43.4 pM and a B_max of 50,635 receptors/cell for the low-affinity site. Kinetic analysis of IGF-I binding (Fig. 6B) revealed one class of binding sites with K_d and B_max values of 11.4 nM and 190,455 receptors/cell, respectively. Thus the lower mitogenic potency of IGF-I was matched by a lower affinity toward the cognate receptor.

View larger version (11K): [in this window] [in a new window]   Fig. 6.   Scatchard plots of EGF (A) and IGF-I (B) binding to NRK-52E cells.

The biological activity of renal tissue extracts was tested on cultured cells in conditions where growth factors had been observed to promote cell proliferation. Compared with control cultures, i.e., cells growing in serum-free medium with no renal tissue extract, a higher proliferative activity was seen in cultures containing tissue extract from intact kidneys (Fig. 7). However, tissue extracts prepared from kidneys of rats treated with gentamicin were endowed with more pronounced mitogenic properties, compared with control extracts, especially when their effect on [³H]thymidine incorporation was expressed relative to their protein content (Fig. 8). The presence of mitogenic substance(s) related to tubular injury/regeneration was specific to renal tissue, as revealed by the testing of rat serum in the same experimental conditions. Although the addition of homologous serum increased cell proliferation, compared with cells grown in serum-free medium, no difference was found in this respect between gentamicin-treated rats and controls (Fig. 7).

View larger version (27K): [in this window] [in a new window]   Fig. 7.   Effect of renal tissue extracts and rat sera on NRK-52E cell proliferation. Animals treated with 100 mg/kg gentamicin were killed at day 1. Proliferative response of growth-arrested NRK-52E cells was assessed by measurement of [3H]thymidine incorporation. Since some cultures contained an excess of protein resulting from the addition of tissue extracts or sera, radioactivity was expressed per well. Each value is the mean of 4 determinations (error lines are ±SD). Results were analyzed by ANOVA and Tukey's post hoc test. * P < 0.05 compared with control cultures, with no extract or serum.  P < 0.05 compared with tissue extract from sham-treated animals.

View larger version (32K): [in this window] [in a new window]   Fig. 8.   Mitogenic activity of renal tissue extracts as a function of gentamicin dosage and time of death. Variation of kidney protein content due to nephrotoxic damage was taken into account by expressing the mitogenic effect (rate of [3H]thymidine incorporation) according to protein concentration in tissue extracts. Each value is the mean of 4 determinations (error lines are ±SD). Results were analyzed by ANOVA and Tukey's post hoc test. * P < 0.05 compared with renal tissue extract from sham-treated rats.  P < 0.05 compared with rats exposed to 50 mg/kg gentamicin and killed at day 1.  P < 0.05 compared with rats exposed to 100 mg/kg gentamicin and killed at day 1.

The possible existence of a link between the mitogenic activity of tissue extracts from kidneys undergoing nephrogenic repair and the response of NRK-52E cells to polypeptide growth factors was examined by using tyrphostin A46, an inhibitor of tyrosine kinase activity. Results are presented in Fig. 9. Addition of tyrphostin A46 to cell cultures whose growth was stimulated by EGF largely abolished the enhancing effect of this growth factor on cell proliferation, since DNA synthesis in tyrphostin-containing cultures was reduced by 80% compared with that measured in cultures without inhibitor. The proliferative activity of cells exposed to extracts from intact kidneys showed a decrease of 45% in presence of tyrphostin A46. The reduction of DNA synthesis induced by the tyrosine kinase inhibitor reached ~70% in cultured cells growing in presence of tissue extracts obtained from gentamicin-treated rats.

View larger version (39K): [in this window] [in a new window]   Fig. 9.   Effect of tyrphostin on proliferative response of NRK-52E cells to EGF and renal tissue extracts. Control cultures with no tyrphostin contained an equivalent amount of tyrphostin diluent (dimethyl sulfoxide). In EGF-containing cultures, growth factor was present at a concentration of 10 ng/ml. Each value is the mean of 4 determinations (error lines are ±SD). Tyrphostin-containing cultures were compared with corresponding controls by Student's t-test. * P < 0.05.

The actual nature of the mitogenic substance(s) present in renal tissue extracts from gentamicin-treated rats was further explored by utilizing polyclonal antibodies raised against selected polypeptide growth factors. As depicted in Fig. 10A, anti-EGF antiserum did not substantially modify DNA synthesis in cultured cells exposed to renal tissue extracts from control and treated rats, even though the antiserum caused a 65% decrease in the stimulation of cell proliferation induced by purified EGF. Anti-IGF-I did not modify DNA synthesis in cells exposed to extracts from control kidneys but significantly reduced by 25% the cell DNA synthesis in presence of renal tissue extracts obtained from treated rats (Fig. 10B). Similarly, anti-IGF-I induced a 40% decrease of DNA synthesis in growth-arrested cells stimulated with purified IGF-I (Fig. 10B).

View larger version (35K): [in this window] [in a new window]   Fig. 10.   Effect of anti-growth factor antisera on NRK-52E cell response to growth factors and renal tissue extracts. Extracts were prepared from kidneys of animals treated with 100 mg/kg gentamicin and killed at day 1. Growth factors or tissue extracts were preincubated with antisera as indicated in METHODS. Control cultures consisting of growth-arrested cells without additive in culture medium are shown for comparison. Each value is the mean of 4 determinations (error lines are ±SD). Statistical comparisons were made with corresponding controls, where growth factors or tissue extracts had not been preincubated with antisera. * P < 0.05, Student's t-test.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Nephrogenic repair is the major process underlying the reversibility of acute renal failure in human clinics, as well as in experimental animals (7, 9). There is now a general consensus that regenerative hyperplasia occurring in the kidney after acute tubular necrosis is submitted to a positive, and possibly a negative, control exerted by endogenous polypeptide growth factors (10). The current study concentrated on the presence of soluble mitogenic factors in regenerating kidneys.

Acute tubular necrosis was induced in the experimental animals by the administration of high doses of gentamicin. This compound was selected because of its moderate cytotoxic potential, compared with other nephrotoxins such as HgCl₂, CdCl₂, or cisplatin. In preliminary experiments, we observed that gentamicin had no effect on cell growth below a concentration of 0.1 mM in the culture medium. Thus we concluded that the amount of residual gentamicin in kidney tissue extracts of treated animals was not likely to interfere with our bioassay system.

Our cell culture experiments were conducted on NRK-52E cells. Since these cells were originally derived from the rat kidney, this made it possible to work in a homologous system. NRK-52E are epithelial-like cells, and the fact that they exhibit the EGF receptor, as do proximal tubule cells, suggests that they originate from the proximal segment of the nephron. Our data demonstrate that growth-arrested NRK-52E cells are responsive to EGF and IGF-I, two growth factors probably involved in nephrogenic repair.

Intracellular signaling downstream of growth factor receptors depends on the tyrosine kinase activity of these receptors. In our study, we observed that the presence of tyrphostin partly abrogated the proliferative response of growth-arrested NRK-52E cells to renal tissue extracts. This finding allowed us to attribute the mitogenic effect of tissue extracts to the activity of growth factors rather than to the presence of nutrients that might have been extracted from kidney tissue. Strikingly, tyrphostin modified the mitogenic effect of extracts from intact kidneys, albeit to a lesser extent than that of regenerating kidneys. Yet, the rate of cell turnover in normal kidneys is very low. It is conceivable that the kidney produces, even in the absence of lesion, a significant amount of growth factor required for the regulation of renal function or renal cell metabolism but with no effect on renal cell proliferation.

Both EGF and IGF-I have been shown to alleviate renal dysfunction and stimulate nephrogenic repair after renal ischemia or nephrotoxin-induced tubular injury (3, 8, 20, 21). These factors are synthesized in renal tissue, and their receptors have been found on proximal tubule cells (9). They could therefore play a role in the endogenous control of regenerative hyperplasia. However, several studies indicate that physiologically inactive EGF and IGF-I are converted into active forms by different mechanisms.

In the intact kidney, EGF is synthesized by distal tubule cells as a large molecular weight precursor (proEGF) which mostly remains associated with the plasma membrane (23). Acute tubular injury is associated with a transient increase of soluble EGF and a decrease of proEGF expression (16, 23, 25, 26) in renal tissue. Recent in vitro studies on isolated kidney membranes have shown that proEGF could be converted into soluble EGF by the proteolytic activity of kallikrein-like enzymes (14).

Although there is still some uncertainty concerning the site of IGF-I production in the kidney, solution hybridization ribonuclease (RNase) protection assay and immunocytochemistry have shown the presence of both IGF-I mRNA and IGF-I in collecting ducts (1). More recently, IGF-I expression and immunostainable IGF-I have also been found in Henle's loops and in glomerular tufts (2, 18). Like EGF receptors (30), IGF-I receptors are present on proximal tubule cells (5, 6). During nephrogenic repair subsequent to renal ischemia, IGF-I RIA has disclosed a decrease of this growth factor in kidney tissue (4, 31). A parallel decline of IGF-I mRNA has been found by solution hybridization RNase protection assay (4, 31). On the other hand, tubular regeneration also seems to be accompanied by a change in the topology of IGF-I synthesis, since IGF-I mRNA and immunoreactive IGF-I have been demonstrated in regenerating proximal tubules, whereas they are absent in intact proximal tubules (18). Although nephrogenic repair did not affect the overall expression of IGF-I receptor, it resulted in an apparent increase in receptor density probably due to posttranscriptional upregulation (31). When the augmentation of IGF-I receptors was visualized by ¹²⁵I-IGF-I binding and histoautoradiography, it was found to occur in the regenerating zone of postischemic kidneys (18). Of note, nephrogenic repair was also shown to entail a decrease in the expression of insulin-like growth factor binding proteins (31).

Taken altogether, data from the examination of EGF and IGF-I during nephrogenic repair suggest that 1) membrane-anchored proEGF present in distal tubules undergoes enzymatic processing and releases soluble EGF capable of interacting in a paracrine fashion with EGF receptors on proximal tubules (26); and 2) the bioavailability of renal IGF-I increases (due to the reduction of insulin-like growth factor binding proteins), and proximal tubule cells become more responsive to this growth factor (due to receptor upregulation), so that IGF-I might also promote regenerative hyperplasia despite the fact that its synthesis diminishes in the postnecrotic kidney (31). In addition, tubular regeneration might also be stimulated by an autocrine action of IGF-I (18).

The mitogenic activity of tissue extracts from regenerating kidneys but not that of intact kidneys diminished when the extracts were incubated in presence of anti-IGF-I antiserum, thus pointing to IGF-I as a major component responsible for that mitogenic activity. This suggests the presence in regenerating kidneys of increased amounts of biologically active IGF-I involved in the autocrine and/or the paracrine stimulation of regenerative hyperplasia. This could appear surprising in view of the decrease of IGF-I expression during nephrogenic repair. However, it is conceivable that the parallel decline of IGF-I binding protein expression described in previous studies (4, 31) might augment the level of biologically active IGF-I.

The lack of anti-EGF effect on the proliferative response to treated kidney extracts came out as an unexpected observation, more especially as the same antiserum partly suppressed the effect of purified EGF on growth-arrested cells. Yet, the current study confirmed previous observations (25, 26) by showing an increase of soluble immunoreactive EGF in regenerating kidneys. Two hypotheses can be put forward to explain the lack of effect of the anti-EGF antiserum. A first explanation would be that extracts from regenerating kidneys contained such an excess of EGF that it could not be neutralized by the amount of anti-EGF antibodies added to the system. We think it is unlikely inasmuch as our anti-EGF had an inhibitory effect on the cell proliferative response to 10 ng/ml purified EGF in culture medium, a concentration higher than the EGF level that could have been attained in cultures by the addition of tissue extracts from regenerating kidneys. An alternative possibility would be the presence within renal tissue extracts of EGF-related products, distinct from mature EGF, which could not be efficiently neutralized in vitro by anti-EGF antiserum. Indeed, recent work on isolated renal membranes (14) has shown that the proteolytic cleavage of membrane-bound proEGF releases, in addition to 6-kDa EGF, two other EGF-related entities of higher molecular weight (~45 kDa). These larger forms of soluble EGF exhibit a mitogenic activity toward cultured cells but do not cross react well with anti-EGF antiserum raised against mature 6-kDa EGF.

HGF is another growth factor that promotes tubular regeneration and alleviates kidney dysfunction when administered to experimental animals after induction of ischemic or toxic renal insult (19). Thus it emerges as another growth factor possibly involved in nephrogenic repair. An increase of HGF mRNA in renal interstitial cells and an elevation of HGF activity in plasma and kidney tissue have been observed following renal ischemia or HgCl₂ administration (12). Postischemic kidneys also exhibited an enhanced expression of c-met, the HGF receptor. (13). Although HGF behaves as a potent mitogen when tested on rabbit tubular epithelial cells in secondary culture (11), we do not know whether it is active on NRK-52E cells. However, it is noteworthy that HGF seems to act on nephrogenic repair in an endocrine as well as in a paracrine fashion (12), whereas in our experimental model there is no increase of mitogenic activity in the serum of rats undergoing renal tubular regeneration.

To conclude, the present in vivo/in vitro study unequivocally demonstrates the existence of a mitogenic activity in tissue extracts from regenerating kidneys and shows that it is related to IGF-I. To our knowledge, this is the first demonstration of an increase of biologically active IGF-I in kidneys undergoing nephrogenic repair. The approach that we describe here can be modified at different levels and used to explore further the intricate mechanism of nephrogenic repair.