Upregulation of Ciliary Neurotrophic Factor (CNTF) and CNTF Receptor ␣ in Rat Kidney with Ischemia-Reperfusion Injury

Home , Tumor necrosis factor

J Am Soc Nephrol 12: 749–757, 2001 Upregulation of Ciliary Neurotrophic Factor (CNTF) and CNTF Receptor ␣ in Rat Kidney with Ischemia-Reperfusion Injury

CHUL WOO YANG,* SUN WOO LIM,† KI WHAN HAN,† HEE JONG AHN,* JUNG HEE PARK,* YOUNG HEE KIM,† MATTHIAS KIRSH,‡ JUNG HO CHA,† JOO HYUN PARK,* YONG SOO KIM,* JIN KIM,† and BYUNG KEE BANG* *Division of Nephrology, Department of Internal Medicine, Kangnam St. Mary’s Hospital, and †Department of Anatomy, The College of Medicine, The Catholic University of Korea, Seoul, Korea; and ‡Institute of Anatomy, University of Freiburg, Freiburg, Germany.

Abstract. Ciliary neurotrophic factor (CNTF) is presumed to weak and limited to the descending thin limb of the loop of play a role as a survival factor in neuronal cells, but little is Henle. With I/R injury, CNTF mRNA and protein expressions known about its role in the kidney. To investigate this, the were strikingly increased as compared with the sham-operated expression of CNTF and CNTF receptor ␣ (CNTFR ␣) was rat kidney, and the immunoreactivity of CNTF was mainly analyzed in the ischemic rat kidney. An ischemia/reperfusion observed in the regenerating proximal tubules. The expression (I/R) injury was induced by clamping both renal arteries for 45 of CNTFR ␣ mRNA was also increased after I/R injury, and its min. Animals were killed at 1, 2, 3, 5, 7, 14, and 28 d after location and expression patterns were similar to the expression ischemia. The expression of CNTF and CNTFR ␣ was moni- of CNTF. These findings suggest a possible role of CNTF as a tored by reverse transcription-PCR, in situ hybridization, im- growth factor during renal tubular repair processes after I/R munoblotting, immunohistochemistry, and electron micros- injury and an autocrine or paracrine function of CNTF acting copy. In sham-operated rat kidneys, CNTF expression was against CNTFR ␣.

The kidney has a remarkable capacity for restoring its structure Ciliary neurotrophic factor (CNTF) was originally identified and function after ischemic injury (1–3). Recovery is depen- as a trophic molecule for the survival of embryonic chicken dent on the ability of the remaining tubular cells to dediffer- ciliary neurons in vitro (9). Subsequent studies have shown that entiate, enter the cell cycle, proliferate, reline the damaged CNTF is a neuronal growth factor that has a regulatory role in areas along the nephron, and then redifferentiate. This replica- local neuronal healing and regeneration (10,11). The CNTF tive repair process of renal tubule cells is mediated predomi- mRNA and protein are widely expressed in the brain, heart, nantly by the paracrine or autocrine release of growth factors, lung, liver, kidney, and testis of the rat, in addition to prefer- and these factors are expressed transiently in a coordinated ential expression in the sciatic nerve (12,13). Apart from neu- manner during renal regeneration (4). ronal tissue, CNTF is most abundant in the kidney, but its role Several growth factors have been identified in the kidney, and precise localization are yet to be determined. and these growth factors act in a coordinated manner. Of these, The effects of CNTF are mediated by a CNTF-specific epidermal growth factor, insulin-like growth factor-1, and he- ligand-binding ␣ subunit (CNTFR ␣) (14,15). Functionally, patocyte growth factor are well known (5). In addition, hepa- CNTF receptors are critical for the developing nervous system rin-binding epidermal growth factor-like growth factor (6), and are needed to sustain life. Whereas disruption of the CNTF fibroblast growth factor (7), and transforming growth factor-␤1 (8) are associated with the repair process. Additional growth gene results in only modest changes within the central nervous ␣ factors are believed to be involved in the repair process after system (16), mice with the CNTFR null mutation (“knock- ischemic injury because of the complicated nature of the outs”) die during the perinatal period and display profound ␣ process. neuron damage (17). The expression of CNTFR mRNA is restricted primarily to nervous system and skeletal muscle, and a small amount of CNTFR ␣ mRNA is present in the kidney Received March 16, 2000. Accepted September 13, 2000. (18) Correspondence to Dr. Jung Ho Cha, Associate Professor, Department of From the data that have accumulated on the role of CNTF as Anatomy, The Catholic University of Korea, 505 Banpo-Dong, Seocho-Ku, a growth factor in nervous tissue, we hypothesized that CNTF 137-701, Seoul, Korea. Phone: 82-2-590-1163; Fax: 82-2-536-3110; E-mail: may play a role as a growth factor in the kidney. To test this, [email protected] the response of CNTF and CNTFR ␣ to ischemia/reperfusion 1046-6673/1204-0749 Journal of the American Society of Nephrology (I/R) injury was evaluated during regeneration of renal tubular Copyright © 2001 by the American Society of Nephrology cells after I/R injury. Our findings support the hypothesis. 750 Journal of the American Society of Nephrology J Am Soc Nephrol 12: 749–757, 2001

Table 1. Oligonucleotide primer pairs for RT-PCR

Gene Gene Size Oligonucleotide Primers Position Size Bank

CNTF 21 S GAT GGC TTT CGC AGA GCA AAC 77 550 X17457 22 AS GCA CAG TAA AGA AGA GTA GTC T 627 CNTFR␣ 20 S GCT GTG ACC TGG AGG GTA AA 240 425 S54212 20 AS GTG ATA GCC GTG GTG TTG TG 664 GAPDH 20 S AAT GCA TCC TGC ACC ACC AA 469 515 M17701 21 AS GTA GCC ATA TTC ATT GTC ATA 984

Materials and Methods animals (6 sham and 6 I/R injury) were included at each time point. Animal Preparation and Induction of Acute Renal Presence of acute renal failure was determined by measuring the Failure concentration of creatinine in serum at each experimental time point. The serum creatinine levels were measured with an autoanalyzer Male Sprague-Dawley rats, weighing approximately 250 to 300 g, (Roche Diagnostics, Nutley, NJ). The serum creatinine levels (mg/L) were housed under a 12-h light/dark cycle, and food and water were were above 1.0 mg/dl on day 1 after I/R injury. The sham-operated available ad libitum. The experimental protocol used in this study was rats showed no significant difference at each time point, and the approved by the Animal Ethics Review Committees of our Institution. average serum creatinine concentrations were approximately 0.5 Animals were anesthetized with an intraperitoneal injection of ket- mg/L. amine (2 g/kg body wt). After the abdomen was opened through a midline incision, both renal pedicles were exposed and cleaned by blunt dissection. Microvascular clamps were placed on both renal Preservation of Kidneys arteries to block renal blood flow. Core body temperature was main- The kidneys were perfused briefly through the abdominal aorta tained by placing the animals on a homeothermic table. After 45 min, with phosphate-buffered saline (PBS) to rinse out the blood and the clamps were removed and blood flow was returned to the kidneys. subsequently were fixed by in vivo perfusion with a periodate-lysine- Animals were killed at 1, 2, 3, 5, 7, 14, and 28 d after reperfusion. The paraformaldehyde (PLP) solution for 4 min. They were cut into sham operation was performed in a similar manner, except for the sagittal slices and then immersed in PLP overnight at 4°C. After being clamping of the renal vessels. Total number of animals was 84; 12 rinsed in PBS, tissues were dehydrated in a graded series of ethanol

Figure 1. Immunoblot (A and B) and immunohistochemical detection (C and D) of proliferating cell nuclear antigen (PCNA) in rat kidney after sham operation and ischemia/reperfusion (I/R) injury. Representative immunoblots for PCNA and relative optical density in cortex (A) and medulla (B). *, P Ͻ 0.05 versus sham. In the sham-operated rat kidney (C), there were few PCNA-positive cells. After ischemia, numerous PCNA-positive cells were observed in the renal tubular cells (D). Optical densities represent the mean Ϯ SEM for six animals. Magnification, ϫ200. J Am Soc Nephrol 12: 749–757, 2001 Upregulation of CNTF and CNTFR ␣ in Ischemic Rat Kidney 751

Figure 3. Immunoblotting for CNTF protein in rat kidneys after sham Figure 2. Semiquantitative reverse transcription-PCR (RT-PCR) for operation and I/R injury. (A) Representative immunoblot for CNTF ciliary neurotrophic factor (CNTF) mRNA using glyceraldehyde-3- protein in the cortex. A band of 22 kD corresponds to the molecular phosphate dehydrogenase (GAPDH). A band of 550 bp corresponds to weight of CNTF protein. (B) Representative immunoblot for CNTF CNTF. Differences in the CNTF/GAPDH ratios indicate the relative protein in the medulla. *, P Ͻ 0.05 versus sham. Optical densities differences in CNTF mRNA levels between the samples. (A) CNTF represent the mean Ϯ SEM for six animals. mRNA expression in the cortex. (B) CNTF mRNA expression in the medulla. *, P Ͻ 0.05 versus sham. Optical densities represent the mean Ϯ SEM for six animals. 0.01. After the sections were washed with 0.05 M Tris buffer, they were dehydrated in a graded series of ethanol and embedded in Epon-812. The embedded 50-␮m-thick sections were examined, and 1 ␮m of semi-thin and embedded in wax (polyethylene glyco 400 disterate; Poly- sections were cut and photographed on Olympus Photomicroscope sciences, Inc., Warrington, PA). For immunohistochemistry using a (Tokyo, Japan) equipped with differential-interference contrast. pre-embedding method, PLP-fixed tissues were cut on a vibratome For the immunoelectron microscopy, some of the immunostained (Lancer Vibratomes Series 10 00; Technical Products International, vibratome sections were postfixed with 1% osmium tetroxide and St. Louis, MO) to a thickness of 50 ␮m. embedded in Epon-812. Ultrathin sections were stained with lead citrate and observed with a transmission electron microscope Immunohistochemistry for Proliferating Cell Nuclear (1200EX, JOEL, Tokyo, Japan). Antigen and CNTF Wax sections were dewaxed and hydrated through the ethanol Semiquantitative Reverse Transcription-PCR for CNTF series and were immunostained with the primary antibody to prolif- mRNA and CNTFR ␣ mRNA erating cell nuclear antigen (PCNA; PC-10, DAKO, Glostrup, Den- Total RNA was isolated from Tri Reagent (MRC, Cincinnati, OH) mark) using Vectastain ABC kit (Vector Laboratories, Burlington, according to the manufacturer’s instructions. First-strand cDNA was CA) according to the manufacturer’s instructions. Diaminobenzidine reverse-transcribed from the RNA using random hexanucleotide prim- was used as chromogen. After immunostaining, sections were coun- ers (Life Technologies, Gaithersburg, MD) and Moloney murine terstained with hematoxylin. leukemia virus reverse transcriptase (Life Technologies) as previously Immunohistochemistry using a pre-embedded method was per- prescribed (20). The cDNA was then amplified by PCR for 25 to 30 formed as described previously (19). The 50-␮m-thick vibratome cycles. Ten ␮l of each reaction cup was run on a 1.5% agarose gel, ␮ sections were washed with 50 mM NH 4Cl in PBS. Before incubation which contained 1 g/ml ethidium bromide. Reverse transcription- with primary antibody, the tissue sections were incubated for 3 h with PCR (RT-PCR) products were quantified by densitometry of a pho- PBS containing 1% bovine serum albumin, 0.05% saponin, and 0.2% tograph of ethidium bromide–stained agarose. The number of PCR gelatin (solution A). Tissue sections were then incubated overnight at cycles was optimized to measure the amount of mRNA in the linear 4°C with monoclonal antibody against CNTF (diluted to 1:10; Chemi- range, and an analysis of cycle sequencing revealed that the sequence con International Inc., Temecula, CA) in PBS containing 1% bovine was identical to position 77-627 in rat CNTF cDNA and 240-664 in serum albumin (solution B). After washes with solution A, the tissue rat CNTFR ␣ cDNA. In the semiquantitative measurement, CNTF and sections were incubated for 2 h with peroxidase-conjugated goat CNTFR ␣ primers were coamplified with glyceraldehyde-3-phos- anti-mouse IgG Fab fragment (Jackson ImmunoResearch Laborato- phate dehydrogenase (GAPDH) primers, and the CNTF/GAPDH or ries, West Grove, PA), diluted 1:50 in solution B. The tissues were CNTFR ␣/GAPDH product ratios were calculated and considered an then rinsed, first in solution A and subsequently in 0.05 M Tris buffer index of CNTF or CNTFR ␣ mRNA expressions. Six animals were (pH 7.6). For the detection of horseradish peroxidase, the sections used for PCR in each time point, and three measurements were used were incubated in 0.1% 3,3'-diaminobenzidine in 0.05 M Tris buffer per animal. The primers used for CNTF, CNTFR ␣, and GAPDH

for 5 min, after which H 2O 2 was added to a final concentration of amplification are listed in Table 1. 752 Journal of the American Society of Nephrology J Am Soc Nephrol 12: 749–757, 2001

Figure 4. Light micrographs of a 50-␮m-thick vibratome section (A, C through F) and a 1-␮m-thick section (B) illustrating immunostaining for CNTF in kidneys from sham-operated rats (A and B) and from rats with I/R injury on days 1 (C), 3 (D), 7 (E), and 28 (F). (A and B) Strong CNTF immunostaining in the inner stripe of the outer medulla (ISOM) is evident. High magnification micrographs of the transition between the outer stripe of the outer medulla (OSOM) and ISOM (B); descending thin limb (*) showed strong CNTF immunostaining, but proximal tubules (PT) were negative. (C through F). From 1 d after ischemic injury, induced CNTF immunostaining started to appear between OSOM and the cortex (Cx). CNTF immunostaining in OSOM and Cx remained strong to 7 d after ischemic injury and decreased gradually. Magnifications: ϫ15 in A; ϫ480 in B; ϫ18 in C through F.

Immunoblotting of CNTF and PCNA Proteins were separated by sodium dodecyl sulfate-polyacrylamide Tissues were homogenized in a buffer containing 10 mmol/L Tris gel electrophoresis in 15% polyacrylamide gels and were electroblot- Cl (pH 7.6), 150 mmol/L NaCl, 1% (wt/vol) sodium deoxycholate, ted onto Bio-Blot nitrocellulose (Costar, Cambridge, MA). Nonspe- 1% (vol/vol) Triton X-100, 0.1% (wt/vol) sodium dodecyl sulfate, 1% cific binding was blocked by incubating the blots for1hin5%

(vol/vol) aprotinin, 2 mmol/L Na 3VO 4, and freshly added leupeptin (wt/vol) nonfat milk. CNTF protein was detected with a monoclonal (1 ␮g/ml), pepstatin (1 ␮g/ml]), and phenylmethylsulfonyl fluoride (1 anti-CNTF antibody (Chemicon International Inc.), and PCNA pro- mmol/L). Homogenates were centrifuged at 16,000 ϫ g for 15 min at tein was detected with anti-PCNA antibody (PC10, Dako, Copenha- 4°C, and protein concentrations were determined on supernatants gen, Denmark) by incubation overnight. Primary antibody incubation using the Bradford method protein microassay (Bio-Rad, Hercules, was followed by six washes with Tris-buffered saline containing CA). Homogenates were boiled for 5 min in Laemmli sample buffer. 0.005% Tween 20 (TBS-T). The blots were then incubated with J Am Soc Nephrol 12: 749–757, 2001 Upregulation of CNTF and CNTFR ␣ in Ischemic Rat Kidney 753

Figure 5. Light (A and B) and electron (C) micrographs illustrating immunostaining for CNTF in OSOM of rat kidneys rendered ischemic after 1d(A)and7d(BandC)before rats were killed. (A and B). CNTF immunostaining was demonstrated in some cells (arrowheads) of proximal tubules (PT) of the OSOM. Exfoliated cells (arrows in A) showed no immunostaining. (C) Note the three types of cells: negatively (cell I), moderately (cell II) and densely immunostained (cell III) cells. Cell III is almost devoid of microvilli. Cell II has short microvilli compared with cell I. A prominent nucleolus (*) is shown in the nucleus of cell III, and heterochromatin is restricted to the nuclear membrane. Note the tight junctions (arrows) and many mitochondria (arrowheads) and free ribosomes in CNTF-immunostained cells (cells II and III). Bar in C represents 1 ␮m. Magnifications: ϫ240 in A; ϫ480 in B; ϫ5500 in C.

peroxidase-conjugated antibody, goat anti-mouse IgG for CNTF and prehybridization buffer was composed of 50% formamide, 4 ϫ SSC, mouse anti-rat IgG for PCNA for 30 min. Antibody-reactive protein 10% dextran sulfate, 1 ϫ Denhardt’s solution, and 1 ␮g/␮l salmon was detected using an enhanced chemiluminescence kit (Amersham sperm DNA. The hybridization buffer was identical with the prehy- Life Science, Buckinghamshire, UK). Densitometric analysis was bridization buffer except that salmon sperm DNA was substituted by performed using Imagemaster VDS software (Pharmacia Biotech, 150 ng/␮l CNTFR ␣ riboprobe. After posthybridization washing, Piscataway, NJ). Six animals were used for PCR in each time point, sections were incubated with antidigoxigenin antiserum conjugated and three measurements were used per animal. with alkaline phosphatase (Boehringer Mannheim, Mannheim, Ger- many), and histochemical detection was then performed using the In Situ Hybridization for CNTFR ␣ 4-Nitroblue tatrazolium chloride/5-Bromo-4-chloro-3-indolyl-phosphate The riboprobe for CNTFR ␣ used in the present study was kindly mixture (Boehringer Mannheim). provided by Dr. H. D. Hofmann (21). After dewaxing, the sections were treated in 0.2 N HCl for 20 min and incubated in 20 ␮g/ml Statistical Analyses pepsin (0.1 N HCl) for 20 min at room temperature, which was Data reported are mean Ϯ SEM, and all statistical analyses were followed by three washes with PBS. Prehybridization and hybridiza- calculated with SYSTAT for Macintosh v. 5.2 (SYSTAT Inc., Chi- tion steps were carried out at 53°C for 1 h and 15 h, respectively. The cago, IL). The sham-operated rats showed no significant variations at 754 Journal of the American Society of Nephrology J Am Soc Nephrol 12: 749–757, 2001

a maximum on day 7. The relative optical densities in each lane, taking day 1 as a 100% reference point, were as follows: day 1, 100%; day 2, 791%; day 3, 622%; day 5, 791%; day 7, 1600%; day 14, 278%; and day 28, 239%. The pattern of CNTF expression in the medulla was similar to that of the cortex (Figure 3B). The relative optical densities of CNTF protein were as follows: day 1, 99%; day 2, 305%; day 3, 250%; day 5, 265%; day 7, 433%; day 14, 272%; and day 28, 295%.

Localization of CNTF Protein in Rat Kidneys with Sham Operation and I/R Injury In rat kidneys with sham operations, CNTF immunoreactivity was observed only in the descending thin limb of the loop of Henle (Figure 4, A and B). With I/R injury, CNTF expression began to appear in the outer part of the outer stripe of the outer medulla (OSOM) (Figure 4C) and progressively expanded into the cortex and the OSOM up to 7 d (Figure 4, D and E). Thereafter, CNTF expression in the cortex and OSOM decreased gradually (Figure 4F). The CNTF-positive cells in the OSOM were predominantly proximal tubule cells, which were squamous or cuboidal in shape and had less developed Figure 6. Semiquantitative RT-PCR for ciliary neurotrophic factor apical microvilli (Figure 5, A and B). In contrast, the exfoliated receptor ␣ (CNTFR ␣) mRNA using GAPDH. A band of 550 bp cells showed no CNTF immunoreactivity (Figure 5A). corresponds to CNTFR ␣ mRNA. Differences in the CNTFR ␣/GAPDH ratio indicate the relative differences in the CNTFR ␣ Immunoelectron microscopy showed densely immuno- mRNA levels between the samples. *, P Ͻ 0. 05 versus sham. Optical stained cells with no or poorly developed apical microvilli densities represent the mean Ϯ SEM in six animals. (Figure 5C). These cells seemed to be regenerating large and bright nuclei with prominent nucleoli, many mitochondria and free ribosomes, and intact intercellular junctions and basal any time point, so the data at day 1 were used as a control reference. lamina. Comparisons between these data and those for I/R rats were analyzed using analyzed using the t test, and the level assumed for statistical Expression of CNTFR ␣ mRNA in Rat Kidneys after significance was P Ͻ 0.05. Sham Operation and I/R Injury The expression pattern of CNTFR ␣ mRNA was similar to Results that of the CNTF protein. The expression of CNTFR ␣ mRNA Proliferative Activity Assessed by Immunoblotting and in the medulla (Figure 6) was increased on day 1 (0.24 Ϯ 0.05 Immunohistochemistry for PCNA versus 0.55 Ϯ 0.10; P Ͻ 0.05) and on day 2 (versus 0.65 Ϯ Immunoblotting to detect the PCNA protein demonstrated a 0.11; P Ͻ 0.05) as compared with the sham-operated rat. This single band of 36 kD (Figure 1). The expression pattern of increase of CNTFR ␣ mRNA was observed up to day 7 (versus PCNA protein was similar in the cortex and medulla. A strik- 0.50 Ϯ 0.11; P Ͻ 0.05), but thereafter its levels were progres- ing increase was observed on day 2 (cortex, 48.1-fold; medulla, sively decreased (day 14, versus 0.38 Ϯ 0.05; day, 28; versus 23-fold as compared with sham-operated controls), and a small 0.24 Ϯ 0.05). increase was observed on day 7 (cortex, 31.9-fold; medulla, 11.5-fold as compared with the sham-operated controls). Localization of CNTFR ␣ mRNA in Rat Kidneys with PCNA-positive cells were rarely found in the sham-operated Sham Operation and I/R Injury rat kidney (Figure 1C) but were increased in numbers in In agreement with results from RT-PCR, the hybridization kidneys with I/R injury (Figure 1D). signal of CNTFR ␣ mRNA was detectable in sham-operated rat kidneys (Figure 7, A through C). With I/R injury, CNTFR Expression of CNTF mRNA and Protein in Rat Kidneys ␣ mRNA expression progressively increased in both the renal with Sham Operation and I/R Injury cortex and the medulla up to day 7(Figure 7, D through F). The CNTF mRNA expression in the cortex and the medulla S3 segment of the proximal tubules in the OSOM showed the began to increase on days 1 to 2 and was maximal on days 5 most prominent increase in signal (Figure 8B). In the cortex, to 7. It then gradually decreased (Figure 2). Expression of the initial part of the proximal tubules, continuous with the CNTF protein (Figure 3) was similar to CNTF mRNA expres- glomerular urinary pole, and the S2 segment of the proximal sion. In the cortex (Figure 3A), a very faint band for CNTF tubules in the medullary ray showed strong signal intensity protein was detected in sham-operated rat kidney. With I/R (Figures 7F and 8A). Thereafter, in situ signals of CNTFR ␣ injury, CNTF protein began to increase on day 1 and achieved mRNA decreased gradually, but a moderate level of CNTFR ␣ J Am Soc Nephrol 12: 749–757, 2001 Upregulation of CNTF and CNTFR ␣ in Ischemic Rat Kidney 755

Figure 7. Expression of CNTFR ␣ mRNA in kidneys from sham-operated rats (A through C) and I/R injury on days 1 (D), 3 (E), 7 (F), and 28 (G). In low magnification (A), no clear hybridization signal was observed. At high magnification (B and C), weak signals were detected in the distal tubules (DT) and the descending thin limb of Henle’s loop (DTL). There were no signals in the proximal tubules (PT). Induction of in situ signals of CNTFR ␣ mRNA was observed in the outer medulla (OM) on day 1 (D), and its signals were increased and expanded into the cortex on day 7 (E and F). Thereafter, in situ signals of CNTFR ␣ mRNA decreased gradually. Note persistent moderate levels of CNTFR ␣ mRNA on day 28 (G). The asterisk indicates proximal tubules showing strong hybridization signal in a medullary ray. Magnifications: ϫ15 in A; ϫ300 in B and C; ϫ18 in D through G. mRNA was observed in the thick ascending limb of the inner occurred in parallel with the recovery of normal renal struc- stripe of the outer medulla (ISOM) on day 28 (Figures 7G and tures. This finding suggests that CNTF may have a potential 8C). role as a growth factor in the repair process of renal tubular cells after ischemic injury. To the best of our knowledge, the Discussion present study shows for the first time the induction of CNTF in Our study demonstrates striking changes of CNTF expres- rat kidney after I/R injury. sion in rat kidneys after I/R injury. Low levels of CNTF in rat The presence of CNTF and CNTFR ␣ in rat kidney is still kidneys were significantly increased in the regenerating prox- controversial. Sto¨ckli et al. (22) reported that CNTF mRNA is imal renal tubular cells after I/R injury, and this increase undetectable in the kidney, but Ohta et al. (12,13) reported that 756 Journal of the American Society of Nephrology J Am Soc Nephrol 12: 749–757, 2001

Figure 8. In situ hybridization localization of CNTFR ␣ in kidneys 3 (A and B) and 28 d (C) after ischemic injury. (A) Part of the renal cortex. Note the strongly stained initial part of the proximal tubule (PTi) and the unstained part of the proximal tubule (PT) which is just distal of PTi. (B) S3 segments showing strong hybridization signal. Note that the cells with the protruding shapes (arrows) and extruded cells found in the lumen showed no hybridization signals. (C) Part of the outer medulla. Moderate hybridization signal intensity was observed in the thick ascending limbs of Henle’s loop. Magnification, ϫ300.

CNTF mRNA and protein are present. For CNTFR ␣, Davis et the control group). This finding suggests that CNTF regulates al. (23) first reported that it is expressed exclusively within the processes that occur later, such as redifferentiation of kidney nervous system and skeletal muscle. These discrepant results epithelia. seem to be related to the low level of CNTF and CNTFR ␣ in Distribution of the CNTF receptor is also important as the kidney. In the present study, the presence of both CNTF CNTF signal depends on CNTF binding to CNTFR ␣, and this and CNTFR ␣ was confirmed with RT-PCR and/or immuno- process is essential for the trophic effect of CNTF to occur blotting, and their specific locations were established with the (27). Our study shows the close relationship between CNTF use of immunohistochemistry or in situ hybridization. and CNTFR ␣ in location as well as expression pattern after The anatomic location of CNTF-producing cells is important I/R injury. Both CNTF and CNTFR ␣ were strongly expressed if we are to understand the possible role of CNTF in rat kidney on the regenerating proximal tubular cells or were expressed on with I/R. In general, ischemic renal injury in rat induces growth tubules located close by (CNTF in the descending thin limb of factors at the distal (S3) segment of the proximal tubule (1,24). Henle’s loop; CNTFR ␣ in the thick ascending limb). In This site is the most vulnerable to I/R injury (25). Our study addition, a similar expression pattern of CNTF and CNTFR ␣ showed a significant increase in CNTF immunoreactivity in the was observed during the regeneration period. This suggests S3 segment of the proximal tubule, a site at which CNTF that interaction between CNTF and CNTFR ␣ is mediated by normally is not expressed. Further study of proximal tubules the autocrine or paracrine mechanism during the course of using electron microscopy revealed that CNTF-positive cells regeneration after injury. were morphologically similar to regenerating tubular cells. Our study demonstrates the striking increase of CNTF in the Conversely, there was no immunoreactivity of CNTF in the rat kidney after I/R injury, but the mechanism that regulates detached or exfoliated tubular cells. These immunohistochem- CNTF expression is unknown. In the absence of injury, neu- ical and electron microscopic studies provide evidence that the rotrophic protein cannot be secreted, because there is no signal primary source of induced CNTF is regenerating proximal sequence that could allow its secretion according to the clas- tubular cells and that necrotic cells do not produce CNTF in rat sical vesicular mechanism (22). Moreover, several studies kidney with I/R injury. failed to demonstrate the release of significant amounts of the The proliferative activity of tubular cells is maximal at 2 to protein by cultured CNTF-expressing cells (28,29). This has 3 d and on day 7 in rat kidneys after I/R injury (26). This fostered the belief that CNTF can be released only from dam- biphasic pattern was confirmed in this study using PCNA aged cells. Recently, evidence was presented for release of immunoblotting (Figure 1). During this period, several growth CNTF through other secretory pathways that do not require factors seem to be working in a coordinated manner. The conventional signal sequences. Cultured astrocytes could be expressions of epidermal growth factor (day 1) and hepatocyte stimulated by exogenous cytokines, for example, by interleu- growth factor (6 to 12 h) are maximal during the early phase, kin-1 and tumor necrosis factor-␣, to secrete CNTF (30). and insulin-like growth factor-1 is detectable at 3 to 7 d (5). In Interestingly, these two cytokines are implicated in the patho- the present study, CNTF expression was maximal on day 7 genesis of I/R injury and upregulated in renal tubular cells with (16-fold in the cortex, 4.3-fold in the medulla as compared with I/R injury (31,32). Therefore, we speculate that inflammatory J Am Soc Nephrol 12: 749–757, 2001 Upregulation of CNTF and CNTFR ␣ in Ischemic Rat Kidney 757 cytokines produced after I/R injury may be responsible for the factor-binding protein as a ciliary neurotrophic factor receptor CNTF production in regenerating renal tubular cells. In vitro component. J Biol Chem 268: 7628–7631, 1993 study will be needed to define the roles of these cytokines in 16. Masu Y, Wolf E, Holtmann B, Sendtner M, Brem G, Thoenen H: regulating CNTF expression in normal and injured conditions. Disruption of the CNTF gene results in motor neuron degener- In summary, our study demonstrates the upregulation of ation. Nature 365: 27–32, 1993 CNTF and CNTFR ␣ at regenerating renal tubular cells after 17. DeChiara TM, Vejsada R, Poueymirou WT, Acheson A, Suri C, Conover JC, Friedman B, McClain J, Pan L, Stahl N, et al.: Mice I/R injury. This finding suggests a role for CNTF as a growth lacking the CNTF receptor, unlike mice lacking CNTF, exhibit factor in renal tissue repair. profound motor neuron deficits at birth. Cell 83: 313–322, 1995 18. Ip NY, McClain J, Barrezueta NX, Aldrich TH, Pan L, Li Y, Acknowledgments Wiegand SJ, Firedman B, Davis S, Yancopoulos GD: The ␣ The authors acknowledge the financial support of the Korea Re- component of the CNTF receptor is required for signaling and search Foundation (KRF 1999-041-F00160). Part of this study was defines potential CNTF targets in the adults and during devel- presented at the annual meeting of American Society of Nephrology, opment. Neuron 10: 89–102, 1993 Miami, November 1999. 19. Kim J, Kim YH, Cha JH, Tisher CC, Madsen KM: Intercalated cell subtypes in connecting tubule and cortical collecting duct of References rat and mouse. J Am Soc Nephrol 10: 1–12, 1999 1. Toback FG: Regeneration after acute tubular necrosis. Kidney Int 20. Rathbun RK, Faulkner GR, Ostroski MH, Christianson TA, 41: 226–246, 1992 Hughes G, Jones G, Cahn R, Maziarz AR, Royle G, Keeble W, 2. Bacallao R, Fine LG: Molecular events in the organization of et al.: Inactivation of the Fanconi anemia group C gene augments renal tubular epithelium: From nephrogenesis to regeneration. interferon-r-induced apoptotic response in hematopoietic cells. Am J Physiol 257: F913–F924, 1989 Blood 90: 974–985, 1997 3. Bonventre JV: Mechanisms of ischemic renal failure. Kidney Int 21. Lee MY, Naumann T, Kirsch M, Frotscher M, Hofmann HD: 43: 1160–1178, 1993 Transient up-regulation of ciliary neurotrophic factor receptor- 4. Humes HD, Lake EW, Liu S: Renal tubule cell repair following alpha mRNA in axotomized rat septal neurons. Eur J Neurosci 9: acute renal injury. Miner Electrolyte Metab 21: 353–365, 1995 622–626, 1997 5. Hammerman MR: Growth factors and apoptosis in acute renal 22. Sto¨ckli KA, Lottspeich F, Sendtner M, Masiakowski P, Carroll P, injury. Curr Opin Nephrol Hypertens 7: 419–424, 1998 Go¨tz R, Lindholm D, Thoenen H: Molecular cloning, expression 6. Homma T, Sakai M, Cheng HF, Yasuda T, Coffey RJ Jr, Harris and regional distribution of rat ciliary neurotrophic factor. Nature RC: Induction of heparin-binding epidermal growth factor-like 342: 920–923, 1989 growth factor mRNA in rat kidney after acute injury. J Clin 23. Davis S, Aldrich TH, Valenzuela DM, Wong VV, Furth ME, Invest 96: 1018–1025, 1995 Squinto SP, Yancopoulos GD: The receptor for ciliary neurotro- 7. Ichimura T, Maier JAM, Madag T, Zhang G, Stevens JL: FGF-1 phic factor. Science 253: 59–63, 1991 in normal and regenerating kidney: Expression in mononuclear, 24. Safirstein R: Gene expression in nephrotoxic and ischemic acute interstitial, and regenerating epithelial cells. Am J Physiol 269: renal failure. J Am Soc Nephrol 4: 1387–1395, 1994 F653–F662, 1995 25. Venkatachalam MA, Bernard DB, Donohoe JF, Levinsky NG: 8. Basile DP, Hammerman MR: TGF-beta in renal development Ischemic damage and repair in the proximal tubule: Differences and renal growth. Miner Electrolyte Metab 24: 144–148, 1998 among the S1, S2, and S3 segments. Kidney Int 14: 31–49, 1978 9. Adler R, Landa KB, Manthorpe M, Varon S: Cholinergic neu- 26. Shimizu A, Yamanaka N: Apoptosis and cell desquamation in rotrophic factors: Intraocular distribution of trophic activity for repair process of ischemic tubular necrosis. Virchows Arch B ciliary neurons. Science 204: 1434–1436, 1979 10. Ip NY, Nye SH, Boulton TG, Davis S, Taga T, Li Y, Birren SJ, Cell Pathol Incl Mol Pathol 64: 171–180, 1993 Yasukawa K, Kishimoto T, Anderson DJ, et al.: CNTF and LIF 27. Adler R: Ciliary neurotrophic factor as an injury factor. Curr act on neuronal cells via shared signaling pathways that involve Opin Neurobiol 3: 785–789, 1993 the IL-6 signal transducing receptor component gp 130. Cell 69: 28. Lillien LE, Sendtner M, Rohrer H, Hughes SM, Raff MC: 1121–1132, 1992 Type-2 astrocyte development in rat brain cultures is initiated by 11. Ikeda K, Iwasaki Y, Shiojima T, Kinoshita M: Neuroprotective a CNTF-like protein produced by type-1 astrocytes. Neuron 1: effect of various cytokines on developing spinal motoneurons 485–494, 1988 following axotomy. J Neurol Sci 135: 109–113, 1996 29. Lin LFH, Mismer D, Lile JD, Armes LG, Butler ET, Vannice JL, 12. Ohta K, Hara H, Hayashi K, Itoh N, Ohi T, Ohta M: Tissue Collins F : Purification, cloning, and expression of ciliary neu- expression of rat ciliary neurotrophic factor (CNTF) mRNA and rotrophic factor (CNTF). Science 246: 1023–1025, 1989 production of the recombinant CNTF. Biochem Mol Biol Int 35: 30. Kamiguchi H, Yoshida K, Sagoh M, Sasaki H, Inaba M, Waka- 283–290, 1995 moto H, Otani M, Toya S: Release of ciliary neurotrophic factor 13. Ohta M, Ohi T, Nishimura M, Itoh N, Hayashi K, Ohta K: Distri- from cultured astrocytes and its modulation by cytokines. Neu- bution and age-related changes in ciliary neurotrophic factor protein rochem Res 20: 1187–1193, 1995 in rat tissues. Biochem Mol Biol Int 40: 671–678, 1996 31. Donnahoo KK, Shames BD, Harken AH, Meldrum DR: Review 14. Davis S, Aldrich TH, Stahl N, Pan L, Taga T, Kishimoto T, Ip article: The role of tumor necrosis factor in renal ischemia- NY, Yancopoulos GD: LIFR␤ and gp130 as heterodimerizing reperfusion injury. J Urol 162: 196–203, 1999 signal transducers of the tripartite CNTF receptor. Science 260: 32. Azuma H, Nadeau K, Takada M, Mackenzie HS, Tinley NL: 1805–1808, 1993 Cellular and molecular predictors of chronic renal dysfunction 15. Stahl N, Davis S, Wong V, Taga T, Kishimoto T, Ip NY, after initial ischemia/reperfusion injury of a single kidney. Trans- Yancopoulos GD: Cross-linking identifies leukemia inhibitory plantation 64: 190–197, 1997