mRNA EXPRESSION OF TRANSFORMING GROWTH FACTOR-α AND THE EGF RECEPTOR FOLLOWING NEPHROTOXIC RENAL INJURY

Abstract

We studied gene expression for transforming growth factor (TGF)-α, epidermal growth factor (EGF), heparin binding (HB) EGF, and the EGF receptor following acute renal failure induced by mercuric chloride administration to gain insight into potential mechanisms of renal repair. Twenty four hours after HgCl₂, 2 mg/kg, creatinine increased from 0.3 ± 0.01 mg/dl in controls to 2.2 ± 0.26 mg/dl in injured rats (n = 5, p < 0.01). Similar changes were observed after 3 days. Messenger RNA expression for EGF was decreased at 24 hours in HgCl₂ treated rats and remained depressed for at least 3 days. On the other hand steady state mRNA for TGF-α increased nearly 2 fold at day 3 in HgCl₂ treated rats 4 mg/kg. Heparin binding EGF was increased early, by day one in injured kidneys and gene expression for the EGF receptor was increased as well. Immunohistochemistry documented an increase in expression of TGF-α in injured kidneys at distal nephron sites. These studies suggest that TGF-α along with HB EGF may be important ligands for the EGF receptor during repair from renal injury.

• Kidney injury, acute
• Epidermal growth factor
• Transforming growth factor
• Receptor

INTRODUCTION

The mechanisms of repair following acute renal injury remain largely undefined. Recent studies however suggest that the epidermal growth factor system may be importantly involved in this process. They demonstrate that the kidney upregulates the EGF receptor during repair from renal injury [1-3]. In addition, following ischemic injury, the receptor undergoes enhanced tyrosine phosphorylation suggesting activation [[4]]. These observations however are complicated by the finding that EGF itself is decreased in the urine of patients with acute renal injury suggesting that it might not be the ligand [[5]]. Transforming growth factor-α is a 5,600 dalton member of a small family of EGF like peptides. It was originally isolated from murine sarcoma virus transformed cells and it has been demonstrated to be present in a number of tissues including the kidney [6-9]. In normal renal tissue it is confined to distal nephron sites. The growth factor is expressed as a transmembrane protein, which can be cleaved by enzymatic action at the amino and carboxyl terminal ends into a soluble form [[10]]. It is a potent stimulator of the 170,000 dalton EGF receptor which is known to be present in the basolateral membrane of renal tubules [[1]]. When compared to EGF, TGF-α is at least 100 times more potent in stimulating transport in proximal renal tubules [[11]]. We studied gene expression for the growth factor, EGF, heparin binding EGF and the EGF receptor during renal injury induced by HgCl₂ to gain a perspective on their control during renal repair.

METHODS

Animal and Tissue Preparation

Male Sprague-Dawley rats, 175–200 g, Harlan Inc., Indianapolis, IN, were used for all studies. Animals were given a subcutaneous injection of HgCl₂ 2 or 4 mg/kg in sterile saline. Control animals received saline alone. After the injections, animals were allowed access to food and water. At 1 and 3 days in the 2 mg/kg group, or 3 days in the 4 mg/kg group, animals were sacrificed and RNA was prepared by the Trizol method, Life Technologies, Gaithersburg, MD. The right kidney was removed without perfusion, decapsulated, bisected, and exposed to freshly prepared 4% formalin overnight. Four micron sections were prepared from paraffin imbedded tissue. Blood, removed at the time of sacrifice, was centrifuged at 14,000 g and the serum was stored frozen at −70°C. Creatinine and BUN were determined by an autoanalyzer.

Plasmids, Riboprobes, and RNase Protection Assay

The TGF-α riboprobe was prepared from the cDNA provided by Fan et al. [[12]]. The template was a PstI to PstI fragment cloned into pBluesript II KS. After linearization of the plasmid with NotI, T3 was used to transcribe a 334 nt riboprobe that protects a 240 nt portion of the rat TGF-α mRNA. The EGF riboprobe was produced from the cDNA subcloned into pBluescript II KS [[12]]. The plasmid was linearized with XhoI and transcribed with T7 to produce a 669 nt probe that protects a 567 nt fragment. Heparin binding EGF mRNA was studied with a riboprobe produced from the cDNA of a 336 bp RT-PCR product cloned into pGEM3Z [[13]]. The plasmid was linearized with EcoRI and SP6 was used to produce a riboprobe of 375 nt. The rat EGF receptor cDNA subclone consists of a 386 nt Sau3 AI fragment in pBluescript II SK. The plasmid was linearized with AvaII and transcribed with T7.

The 460 nt riboprobe protects a 160 nt fragment of mRNA from the extracellular domain of the full-length EGF receptor [[14]]. We have described the 18S riboprobe in detail in prior work [[15]]. Solution hybridization was performed by the method of Ausbel as previously described [15-16]. Gels were quantitated by exposure to phosphoimaging screens. Data were quantitated using a phosphoimager and ImageQuant software, Molecular Dynamics, Sunnyvale, CA. The 18S riboprobe was used to control for sample loading. Individual lane backgrounds were subtracted from each sample prior to quantitation.

Immunohistochemistry

Deparaffinized tissue sections were hydrated with 100% ethanol and 95% ethanol. Slides were washed with PBS and endogenous peroxidase activity was quenched by incubation with 0.1% H₂O₂ for 30 min. After washing, sections were exposed to 0.05% saponin for 30 sec to expose antigens. After 3 saline rinses, slides were incubated in 3% milk for 30 minutes to suppress non-specific binding. The primary antibody (TGF-α monoclonal, Oncogene Science, Uniondale, NY) was applied in a 1:250 dilution overnight at 4°C. After three washes in saline, sections were incubated with a 1:200 dilution of biotinylated horse antimouse secondary antibody, Vector Laboratories, Burlingame, CA, for 30 min. After 3 PBS washes, streptavidin-horseradish peroxidase (Vector) was applied for 15 minutes as described in the manufacture's directions. The slides were extensive washed and exposed to 1% Triton-X 100 for 30 seconds then covered with diaminobenzidine (Sigma Chemical Co., St. Louis, MO) for 5 minutes. They were then counterstained with hematoxylin. Coverslips were applied after dehydration with 95% then 100% ethanol.

Statistical Analysis

Four-five separate animals were assayed in each group. Data are expressed as means ± S.E. Statistics were performed with STATA, Computer Resource Center, Santa Monica, CA, using an unpaired t-test.

RESULTS

Animals were studied at two concentrations of HgCl₂ to provide a spectrum of injury. Shown in table 1 are the creatinine and BUN values for the various groups. Creatinine was elevated in the rats injected with HgCl₂ (2 mg/kg) at days 1 and 3 but returned to control values by day five; however, BUN values remained elevated throughout. Renal function was quite abnormal in animals injected with HgCl₂, 4 mg/kg, and these animals did not live beyond 4 days. EGF has been studied previously in animals following the induction of acute renal failure [17-18]. In these studies both protein and mRNA diminished rapidly. In order to validate our model we studied the mRNA for EGF (figure 1a). EGF mRNA was reduced in injured kidneys in comparison with controls regardless of the day of study or concentration of HgCl₂ (figure 1b). Heparin binding EGF is another member of the EGF family. It has been shown to be increased in both ischemic and nephrotoxic renal injury [[19]]. As a further validation of the HgCl₂ model we examined mRNA for this growth factor at days 1 and 3. Relative mRNA abundance in control kidneys at day 1 averaged 1.0 ± 0.1 while the relative abundance of mRNA prepared from HgCl₂ treated rats, 2 mg/kg, averaged 4.3 ± 0.3 (n = 4, p < 0.01). Control kidney mRNA averaged 1.0 ± 0.6 (relative mRNA abundance) at day 3 and mRNA values in animals treated with HgCl₂, 2 mg/kg, averaged 9.6 ± 0.8 (n = 4, p < 0.01).

Figure 1A. Solution hybridization assay for in control and mercuric chloride (HgCl₂) EGF injected rats at days 1 and 3. The probe is shown diluted 1:10 and 1:100. Simultaneous hybridization was carried out with 18S shown in the lower panel.

Figure 1B. EGF data corrected for lane background and sample loading using 18S. * p < 0.01.

Table

	HgCl2 (2 mg/kg)		Control
Day	Creatinine (mg/dL)	BUN (mg/dL)	Creatinine (mg/dL)	BUN (mg/dL)
1	2.2 ± 0.26^*	52.8 ± 9.7^* (n = 5)	0.3 ± 0.01^a	12.2 ± 0.7 (n = 5)
3	2.4 ± 0.8^**	112.0 ± 38.2^** (n = 4)	0.6 ± 0.04^b	15.4 ± 0.08 (n = 4)
5	0.8 ± 0.4	38.3 ± 8.9^* (n = 5)	0.6 ± 0.02^b	14.0 ± 1.3 (n = 5)
	HgCl2 (4 mg/kg)		Control
3	4.7 ± 0.32^*	210.1 ± 7.8^* (n = 4)	0.3 ± 0.03^a	19 ± 1.2 (n = 4)

Creatinine and blood urea nitrogen (BUN) values for the various groups. HgCl₂ (mercuric chloride).

^a,bDifferent creatinine standards were used for these groups.

^*p < 0.01

^**p < 0.05, comparisons were made between experimental and control values for that day.

Another group of controls averaged 1.0 ± 0.2, relative mRNA abundance, while kidney mRNA prepared from animals injected with HgCl₂, 4 mg/kg at day 3, averaged 17.3 ± 5.1 (n = 4, p < 0.05). The gel for this latter group is shown in figure 2.

Figure 2. Gel of the probe and protected fragments after solution hybridization for heparin binding EGF at day 3 in controls and mercuric chloride treated rats. 18S is presented in the lower panel.

The main purpose of these studies was to examine TGF-α expression during renal injury. Shown in figure 3aA is the gel from the solution hybridization assay for this growth factor in animals receiving HgCl₂ 4 mg/kg. Figure 3bB depicts the data corrected for sample loading in all groups. There was an increase in TGF-α mRNA in both the low and the high dose HgCl₂ groups at day 3. Changes however were not observed in the day 1 group at 2 mg/kg. Since EGF receptors are known to be increased on the cell surface of renal tubules following injury, we sought the mechanism by an evaluation of mRNA expression using the sensitive solution hybridization assay. Shown in figure 3aA is the gel from these studies. Comparisons are illustrated in figure 3cC. Gene expression for the receptor was increased at day 3 in both the high, 4mg/kg, and the low, 2mg/kg, HgCl₂ groups. Differences were not apparent at day one in the animals receiving the lower dose of mercuric chloride. Immunohistochemistry for TGF-α was performed at day 3 in the 4 mg/kg HgCl₂ animals. Staining for the protein was observed in distal nephrons in control kidneys (figure 4). There was an enhancement of staining at sites of injury in HgCl₂ treated rats particularly in intact distal tubules.

Figure 3A. Gel of the probes and protected fragments for TGF-α and the EGF receptor 3 days following vehicle injection or HgCl₂ 4 mg/kg. A standard curve was prepared from normal rat kidneys; 18S was hybridized as well in these assays.

Figure 3B. Data normalized for background and sample loading for TGF-α mRNA. Four animals were studied in each group. * p < 0.05.

Figure 3C. Corrected data for EGF receptor mRNA in the various groups of control and mercuric chloride treated rats. Four animals were studied in each group. * p < 0.01.

Figure 4. Immunohistochemistry for TGF-α. (A) Injury section stained with control serum alone. (B) control kidney section stained for TGF-α. (C) Cortical section from a rat treated with HgCl₂ 4 mg/kg at day 3. Staining was confined to distal tubules in control kidneys whereas it was more pronounced following the induction of injury. Similar observations were made in multiple other sections from 5 animals.

DISCUSSION

We induced acute renal failure using HgCl₂ because a great deal is known about the pathophysiology of the injury in this model (reviewed in ref 20). Because of their affinity for proteins containing sulfhydryl groups, mercurial compounds interfere with a wide variety of enzymes and structural proteins. They interfere in numerous transport processes and they alter cell permeability. Morphologically HgCl₂ causes damage to S₂ and S³ portions of the proximal tubule. Backleak and obstruction of tubular lumens are likely to be important in the maintenance of the decreased GFR in this model. The purpose of our studies was to further define the components of the EGF system during renal injury. As has been shown in previous work, we found that mRNA for EGF decreased rapidly following the induction of injury [17-18]. Although EGF is diminished in the urine of patients with acute renal failure, some studies show a rebound of EGF in cortical regions of experimental animals and there is evidence for enhanced production of mature EGF from precursors [[5]], [21-22]. Homma et al. demonstrated that heparin binding EGF mRNA was increased in ischemia and nephrotoxic renal injury suggesting that this growth factor might be important in regeneration [[19]]. We observed similar changes in HB EGF mRNA in these studies.

We investigated transforming growth factor- as a potential ligand in the repair process since there is evidence that TGF-α plays a major role in kidney growth and development [[23]]. The growth factor has been shown to accelerate cyst growth in a model of polycystic kidney disease and the renal tubules of mice transgenic for TGF-α demonstrate a potent proliferative response which eventually leads to cyst development [24-25]. These studies indicate that TGF-α is capable of stimulating growth of normal and abnormal tubular epithelium.

Our studies demonstrate an increase in mRNA expression for TGF-α at day 3 in moderately and severely injured kidneys. Of interest was the finding that expression tended to be higher in the more severely injured rats suggesting that the extent of the injury dictates a measured response. Immunohistochemical studies documented enhanced staining for the growth factor in distal nephrons. One possibility is that there is diffusion of important concentrations of the ligand through damaged tissue planes or that there is juxtacrine stimulation of proximal tubule EGF receptors by membrane bound TGF-α. The mechanism of the increase in TGF-α mRNA following injury is unclear. An attractive hypothesis is that growth factor produced from macrophages infiltrating the areas of damage might enhance tubular synthesis of TGF-α as has been observed in other systems [[26]]. The fact that the increase in HB EGF mRNA precedes the increase in TGF-α mRNA suggests that TGF-α might be induced by HB EGF. Our studies also demonstrate that increased EGF receptor mRNA accompanies the increase in the mRNAs of the EGF-like peptides. Although the receptor is known to be increased in the kidney following renal injury, the mechanism of this increase has remained elusive. One possibility is that activation of the receptor induces receptor synthesis as has been demonstrated in-vitro [[27]]. Further studies will be necessary to delineate the exact role of the EGF system in repair of the kidney following acute renal injury; however, these studies suggest that there are contributions by more that one member of the EGF family.

ACKNOWLEDGMENT

These studies were supported by funds from the VA Medical Research Office (MKH).