INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

TRANSFORMING GROWTH FACTOR (TGF) was a term used initially by Roberts and associates (38) in 1980 to describe a collection of polypeptides that could be extracted from almost any normal tissue of the mouse. These peptides were functionally divided into TGF-, which competed with epidermal growth factor (EGF) for receptor binding and did not require EGF to promote growth of normal rat kidney (NRK) cells in culture, and TGF-, which did not compete with EGF for receptor binding but did require EGF to stimulate growth (7). TGF-s are now known as a family of five multifunctional polypeptides that have complex effects on organ development, cell growth, and differentiation, expression of extracellular matrix proteins, immune responses, angiogenesis, and tissue repair, as reviewed in detail by Roberts and Sporn (39). Mammals produce TGF-1, -2, and -3 but not -4 or -5. The biological effects, which are similar among the TGF-s, depend upon the target cell, the degree of cellular differentiation, and the cellular environment (37). TGF-s are synthesized in latent form and secreted as disulfide-linked homodimers of ~100 kDa. Removal of the amino terminus of latent TGF- in the extracellular space forms a mature, biologically active form of TGF- that has a molecular weight of ~25 kDa (39). How this latent TGF- complex is activated in vivo is unclear, although acidification, protease treatment, and interaction with thrombospondin have also been shown to activate TGF- in vitro (8, 39, 42-44). The complex structure and function of the kidney make it a prime target for the action of a growth factor. TGF-1 is present in glomeruli and all nephron segments of normal rat kidney (6).

It is well accepted that angiotensin II modulates TGF- (3, 20, 25, 28, 49). In apparent contradiction of these findings, however, was the recent report (47) that demonstrated enhanced glomerular expression of TGF- in the Brookhaven strain of salt-sensitive rats, which have a low-renin form of hypertension. Conclusions about the direct role of dietary salt were not entertained in that study, because, at the time of examination, the animals had severe hypertension and renal damage, which provided additional modifying variables. In addition, time-course experiments were omitted, and normotensive strains that did not have renal failure were not included. We hypothesized that dietary salt directly modulated expression of TGF- through a mechanism independent of the renin-angiotensin system. Using immunoassay and Northern hybridization analysis, we document in the present study the effect of dietary salt on the renal expression of TGF- in Sprague-Dawley rats.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Animal preparation. Studies were conducted on 56 male Sprague-Dawley rats obtained at 28 days of age from Charles River Laboratories (Wilmington, MA). Animals were chosen at this age because of our previous experience that showed normal renal function and blood pressure responses to dietary salt over 2 wk of observation (13). The rats were housed under standard conditions and given 0.3% NaCl diet (AIN-76A with 0.3% NaCl; Dyets, Bethlehem, PA) and water ad libitum for 4 days before initiating the experiment. This diet contained sodium chloride in amounts similar to standard rat diet. The animals were then placed in metabolic cages and allowed free access to water and rat diet, which contained either 0.3% (group termed Lo-salt) or 8.0% (AIN-76A with 8.0% NaCl, Dyets) (group termed Hi-salt) sodium chloride (n = 24 for each group). Animals received salt only in the diet. Urine was collected under oil to prevent desiccation. Urine volume and body weight were recorded daily. In other experiments, rats on 8.0% NaCl diet received twice daily intraperitoneal injections of either furosemide (5 mg/kg, denoted Hi-F; Elkins-Sinn, Cherry Hill, NJ) or chlorothiazide (4 mg/kg, denoted Hi-C; Merck, West Point, PA) (n = 4 in each group). Collected urine samples from each rat were filtered and centrifuged at 325 g for 5 min at 4°C. The supernatant was collected and immediately frozen and stored at 80°C until use. Rats were killed on days 1, 4, and 15. Both kidneys were harvested under aseptic conditions for glomerular and tubular cell culture and Northern hybridization analysis, as described below. Blood was collected simultaneously for determination of serum TGF-1 and sodium concentration. Plasma and urine samples were analyzed for sodium and potassium using flame photometry (model IL-943; Instrumentation Laboratories, Lexington, MA).

Glomerular and tubular cell culture. Glomerular and tubular cell preparations were isolated using a graded sieving technique. This protocol has been shown to produce pure and viable glomeruli for study of TGF- (11, 27, 47, 48). Rats were anesthetized with pentobarbital, 50 mg/kg, by intraperitoneal injection. The kidneys were perfused in situ via the aorta with cold isotonic heparinized saline until blanched (50-60 ml saline over 2 min). The renal cortices from each rat were individually dissected and minced to a pastelike consistency. The homogenate was passed successively through a 106-µm metal sieve that excluded blood vessels and a 75-µm nylon sieve that retained the glomeruli and allowed cells and small tubular segments to pass through. Glomeruli and tubules were collected separately and washed three times with ice-cold PBS at 120 g for 2 min, then resuspended at 5 × 10³ glomeruli or 1 × 10⁶ tubular cells/ml of serum-free RPMI 1640 (GIBCO; Life Technologies, Grand Island, NY). All glomerular preparations consisted of more than 95% glomeruli with minimal tubular contamination, as assessed visually at ×40 magnification. Some freshly collected samples of glomeruli were used to obtain total RNA, as described below. In other experiments, samples of glomeruli and tubules in 24-well plates were pretreated with tetraethylammonium chloride (TEA, 3 mM; Sigma, St. Louis, MO) or glybenclamide (10⁵ M, Sigma), in serum-free RPMI media at 37°C. Glybenclamide was dissolved in DMSO; final concentration of DMSO in the medium was 0.002%. TEA in this dose specifically inhibits potassium channels but is relatively nonselective, whereas glybenclamide inhibits K_ATP channels (12, 15, 16, 34, 35); both compounds have been used specifically to examine TGF- production by endothelial cells (34). Following a 30-min incubation period, the medium was removed and replaced with serum-free media that contained the same concentrations of TEA and glybenclamide. All samples were then incubated for 24 h at 37°C. In these same animals, the heart was also dissected and sectioned. Coronal slices of heart tissue were weighed in a petri dish, then minced with a scalpel into small pieces <1 mm in diameter. The sections were rinsed with cold PBS and suspended in serum-free RPMI 1640 at a concentration of 10 mg tissue/ml. After incubation for 24 h, samples of conditioned media were harvested. Phenylmethylsulfonyl fluoride (1 mM, Sigma) was added as a protease inhibitor, and the mixture was centrifuged at 200 g for 5 min at 4°C. The pellet was dissolved in RIPA buffer (50 mM Tris hydrochloride, 150 mM NaCl, 1 mM disodium EDTA, 0.1 mM EGTA, 1.0% Nonidet P-40, 0.1% SDS, 0.5% sodium deoxycholate) for protein assay (Micro BCA protein assay reagent kit; Pierce, Rockford, IL), and the supernatant was collected and stored at 20°C until assay.

TGF-1 assay. Total TGF-1 in rat urine, serum, and conditioned media was determined using an enzyme immunoassay (TGF-1 E_max ImmunoAssay System; Promega, Madison, WI), following the protocol provided by the manufacturer. Briefly, plates containing 96 flat-bottom wells (Falcon; Becton-Dickinson, Oxnard, CA) were coated overnight at 4°C with 100 µl of anti-TGF-1 monoclonal antibody diluted 1:1,000 in buffer that contained 0.025 M sodium bicarbonate and 0.025 M sodium carbonate, pH 9.7. Thereafter, unbound sites in the wells were blocked, using 270 µl of the blocking reagent supplied in the kit for 2 h at room temperature, then 100 µl of the test samples or TGF-1 standard diluted in sample buffer was added to wells. Following incubation for 3 h at room temperature with vigorous shaking, the wells were washed five times with TBST wash buffer (20 mM Tris · HCl, pH 7.6, 150 mM NaCl, and 0.05% Tween-20), then 100 µl of polyclonal anti-TGF-1 diluted in sample buffer was added. The plates were again incubated overnight at 4°C. After five washes with TBST buffer, wells were filled with 100 µl of antibody conjugated with horseradish peroxidase and incubated for 3 h at room temperature with shaking. Following additional washes as in the previous steps, color was developed by adding 100 µl of peroxidase substrate in 3,3',S,S'-tetramethyl benzidine solution. After ~10-min incubation at room temperature, 100 µl of 1 M phosphoric acid were added to stop the color reaction. Optical density was determined at 450 nm using a microplate reader (THERMO_max; Molecular Devices, Menlo Park, CA). Standards were performed in duplicate using TGF-1, 7.8-1,000 pg/ml in sample buffer, and were used to construct standard curves from which the concentrations of the samples were determined. Urine samples from animals on the 0.3% NaCl diet were diluted 1:1 in sample buffer, and urine samples from animals on 8.0% NaCl diet were not diluted. Serum samples were diluted 1:50 in sample buffer for this assay. Because this assay detected only active TGF-1, some samples of media were also acidified to convert latent TGF-1 to the active from. This protocol was provided by the manufacturer and briefly consisted of acidifying 100 µl of sample to pH 3.2 by addition of 2 µl of 1 N HCl for 30 min at room temperature. The transiently acidified samples were then brought to pH 7.4 with 2 µl of 1 N NaOH. Assay then proceeded as described above and allowed determination of total (latent plus active) TGF-1 in the sample.

Northern hybridization. Total RNA was isolated by single-step method of acid guanidinium thiocyanate-phenol chloroform extraction (14). Briefly, kidneys were homogenized in a denaturing solution of 4 M guanidinium isothiocyanate, 0.5% sarcosyl, and 0.1 M -mercaptoethanol in 25 mM sodium citrate (pH 7.0). After phenol/chloroform extraction, the RNA was precipitated twice with isopropanol and washed with 70% ethanol. The concentration and purity of RNA in each sample were determined using optical density at 260 and 280 nm. Thirty micrograms of total RNA from each sample were electrophoresed in 1.5% agarose gels containing 2.2 M formaldehyde and 0.2 M MOPS, pH 7.0, then transferred to a nylon membrane (GeneScreen Plus Hybridization Transfer Membrane; NEN Life Science Products, Boston, MA) by vacuum blotting (model 785; Bio-Rad, Hercules, CA) for 2 h in 10× SSC (1× SSC is 0.15 M NaCl and 0.015 M sodium citrate, pH 7.0). Nucleic acids were cross-linked by ultraviolet irradiation (Stratagene, La Jolla, CA). The membranes were prehybridized for 20 min at 68°C in standard hybridization solution (QuikHyb; Stratagene). They were then hybridized at 68°C for 4 h with cDNA probes for rat TGF-1 and murine TGF-2 and TGF-3 (all kindly provided by Dr. Thomas S. Winokur, University of Alabama at Birmingham). The cDNA probes, which consisted of ~1-kb fragments produced by digestion of the plasmids with Hind III and Xba I and included the entire coding region of the TGF-s (19), were labeled with [-³²P]dCTP by random oligonucleotide priming (Prime-a-Gene Labeling system; Promega). The blots were washed in 2× SSC with 0.1% SDS at room temperature for 30 min and in 0.1× SSC with 0.1% SDS at 60°C for 20-30 min. Membranes were exposed to XAR-5 film (Kodak) at 80°C. The blots were then stripped in solution containing 1 mM Tris · HCl, pH 8.0, 0.1 mM EDTA, and 0.1× Denhardt's solution, at 75°C for 2 h and rehybridized with human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) probe obtained through American Type Culture Collection (Rockville, MD). Autoradiographs were scanned using a densitometry (model 620 Video Densitometer, Bio-Rad). The density of the GAPDH band in the same lane was used to normalize mRNA loading. For quantification, densities of the bands of TGF-s were individually divided by the density of band for GAPDH in the same lane.

Statistical analysis. All data are presented as means ± SE. Significant difference among data sets was determined using either unpaired t-test or one-way analysis of variance with standard post hoc testing (Statview, version 4.5; Abacus Concepts, Berkeley, CA), where appropriate. Simple regression (Statview) was used to determine the relationship between dependent and independent variables. A value of 5% was used to assign statistical significance.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Mean body weights, serum Na⁺ concentration, urinary flow rate, urinary sodium excretion rate, and urinary potassium excretion rate are shown in Table 1. One animal in the Lo-salt group died of unexplained causes. As expected, urinary flow rate and urinary sodium and potassium excretion rates of the Lo-salt groups were less (P < 0.05) than the corresponding Hi-salt groups. Mean body weight of the Lo-salt group on the diet for 4 days was less (P < 0.05) than the corresponding Hi-salt group.

Mean serum concentrations of TGF-1 did not change during salt loading or with addition of furosemide or chlorothiazide (Tables 2 and 3). In marked contrast, urinary excretion of total TGF-1 was much higher than in rats on the 8.0% NaCl diet. The effect of dietary salt occurred by the first day and persisted throughout the 15 days of study. Urinary excretion of TGF-1 correlated (r² = 0.561; P < 0.0001) directly with urinary sodium excretion rates. Both of the diuretics increased urinary flow but did not produce further increases in excretion of TGF-1. In these experiments, urinary excretion of TGF-1 did not correlate with urine flow rate (r² = 0.052; P = 0.3356).

View this table: [in this window] [in a new window]   Table 2.   Effect of dietary salt on relative steady-state levels of mRNA of TGF-1, -2, and -3 and serum levels and urinary excretion of TGF-1

Mean steady-state levels of mRNA of TGF-1, TGF-2, and TGF-3 in the kidney were also greater (P < 0.05) than mRNA levels of these growth factors in animals on 0.3% NaCl diet (Figs. 1 and 2; Tables 2 and 3). Steady-state levels of mRNA correlated (r² = 0.374; P < 0.0001) directly with urinary excretion of total TGF-1. Freshly isolated glomeruli from rats receiving the 8.0% NaCl diet for 4 days also had a higher content of TGF-1 mRNA than rats receiving the 0.3% NaCl diet (Fig. 3). Furosemide produced further increases (P < 0.05) in mRNA of TGF-3, but not TGF-1 or TGF-2, compared with animals on the 8.0% NaCl diet alone (Table 3).

View larger version (48K): [in this window] [in a new window]   Fig. 1.   Northern hybridization study demonstrating the effect of dietary salt on steady-state levels of transforming growth factor (TGF)-1, -2, and -3. An increase in dietary salt produced sustained increases in these growth factors over the 15 days of study. Samples were electrophoresed together on the same gels. Quantified results are in Table 2. GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

View larger version (42K): [in this window] [in a new window]   Fig. 2.   Northern hybridization analysis of effects of addition of twice daily intraperitoneal injections of furosemide (Hi-F, 5 mg/kg/day) or chlorothiazide (Hi-C, 4 mg/kg/day) to the 8.0% NaCl diet. Samples were electrophoresed together on the same gels. Quantified results are in Table 3.

View this table: [in this window] [in a new window]   Table 3.   Effect of CLZ and furosemide on relative steady-state levels of mRNA of TGF-1, -2, and -3 and serum levels and urinary excretion of TGF-1

View larger version (29K): [in this window] [in a new window]   Fig. 3.   Total RNA was obtained from freshly isolated glomeruli from rats on 0.3% NaCl (Lo-salt) and 8.0% NaCl (Hi-salt) diets for 4 days (n = 4 rats in each group). Northern analysis demonstrated a greater relative amount of TGF-1 mRNA in glomeruli of the Hi-salt rats, compared with the Lo-salt rats (0.63 ± 0.04 vs. 0.33 ± 0.03; P = 0.0008).

Production of total (latent plus active), as well as active, TGF-1 was examined in glomerular and cortical tubular cell preparations from rats on the 8.0% NaCl and 0.3% NaCl diets (n = 4 animals in each group). Secretion of total TGF-1 was increased (P < 0.05) in both preparations obtained from animals on the 8.0% NaCl diet, but secretion of active TGF-1 was increased only in glomeruli from the Hi-salt group (Fig. 4). To determine whether potassium channels were involved in TGF- production, we followed the protocol of Ohno and associates (34), who used TEA and glybenclamide to examine TGF-1 production in endothelial cells in culture. Addition of TEA to the medium reduced active and total TGF-1 to levels comparable to those produced by glomeruli obtained from the Lo-salt group (Fig. 4); TEA had no effect on TGF-1 production by tubular cells. Glybenclamide had no effect on TGF-1 production by either preparation. Total TGF-1 production by cardiac tissue from the Hi-salt group was greater (760 ± 46 vs. 563 ± 91 fg · h¹ · mg protein¹; P < 0.05) than that secreted by cardiac tissue from the Lo-salt group. TEA did not alter production of TGF-1 by cardiac tissue from the Hi-salt or Lo-salt groups (Hi-salt, 760 ± 46 vs. 689 ± 37; Lo-salt, 563 ± 91 vs. 650 ± 75 fg · h¹ · mg protein¹).

View larger version (26K): [in this window] [in a new window]   Fig. 4.   A: production of total (latent plus active) TGF-1 by glomeruli and tubules was greater from animals receiving 8.0% NaCl diet for 4 days, compared with those samples from the corresponding Lo-salt group. B: active TGF-1 was greater only from glomeruli from the Hi-salt group. Augmented production was abrogated by addition of tetraethylammonium (TEA) in the Hi-salt group, but was not affected by glybenclamide. * P < 0.05 compared with corresponding Lo-salt group.  P < 0.05 compared with baseline values of the Hi-salt group.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

TGF-s are multifunctional growth factors that play an important role in kidney growth, development, and senescence (2, 9, 10, 24, 33, 39, 41, 48, 50-52). Expressions of TGF-1, -2, and -3 have been identified in both glomeruli and tubules (6, 7, 39, 40, 51, 53, 54). A majority of studies have focused on expression of these growth factors in developmental and pathological states in the kidney, but recent evidence supports a functional role of TGF- in adult animals, including tubular cell hypertrophy and proliferation, stimulation of adenylyl cyclase, and inhibition of cathepsin production (4, 5, 30, 40). In the current studies, we examined the potential role of dietary salt in modifying growth factor expression in the kidney. Physiological adaptation of TGF- expression to dietary salt has not been examined but is an important consideration, particularly in pathological states that produce glomerulosclerosis. Our current findings showed that an increase in dietary salt produced sustained increases in steady-state levels of mRNA of TGF-1, -2, and -3 in the kidney, compared with rats that received a diet that was identical in composition except for the lower content of sodium chloride. While serum concentrations of TGF-1 did not change, the 8.0% NaCl diet increased urinary excretion of TGF-1. Thus the kidney was the source of augmented excretion of TGF-1. Diuretics (furosemide or chlorothiazide), which were added to the 8.0% NaCl diets in part to determine whether increasing urinary flow rates modulated TGF-, did not further augment synthesis of TGF-1. Production of TGF-1 increased in both glomeruli and tubules, although the active form of TGF-1 was secreted in greater amounts only from glomeruli. Enhanced glomerular production of both inactive and active TGF-1 induced by the 8.0% NaCl diet was inhibited by addition of TEA and not glybenclamide. Dietary salt therefore appeared to activate a TEA-sensitive potassium channel that in turn increased glomerular production of TGF-1.

As reviewed in detail by Davies (17), blood coursing through a vessel produces shear stress, a frictional force that directly alters endothelial cell function. An early event in this process is opening of a shear-activated potassium channel (36), which hyperpolarizes the endothelium and increases cytoplasmic calcium (16, 32, 45). Ohno and associates (34) demonstrated shear-induced expression of TGF-1 in endothelial cells in culture. Blockade of the shear-activated potassium channel with TEA, 3 mM, but not glybenclamide, 10⁵ M, produced dramatic reductions in gene transcription and protein activity of TGF-1 (34). Our current studies showed this effect occurs in vivo, specifically in glomeruli and not in tubular cells or in the heart. We propose that an increase in dietary salt, by increasing blood volume (18), produced shear stress in glomeruli and increased expression of TGF-1 specifically in that region of the nephron. Augmented expression of TGF-1 in the tubules and heart was modest and occurred by a different mechanism that was not defined in this study. We suggest that because active TGF-1, but not the latent form, is a low-molecular-weight protein, significant amounts of the active form can appear in glomerular ultrafiltrate and subsequently interact with proximal tubular cells to promote cell growth and hypertrophy (5, 22, 30, 40) and stimulate autocrine production of TGF-, which has been demonstrated to occur in cells in culture (1, 26).

In conclusion, to the extent that our work can be applied in vivo, although angiotensin II stimulates renal production of TGF- (3, 25, 49), dietary salt increases TGF- not through the renin-angiotensin system but instead by a mechanism that involves alteration of shear stress in glomeruli and activation of a TEA-sensitive potassium channel. The role(s) of this growth factor in the renal adaptation to dietary salt is uncertain, although TGF- may alter NO production, which in turn modulates glomerular hemodynamics, glomerular filtration rate, and renin secretion (23, 31, 46). Finally, a pathogenic role of TGF- in glomerulosclerosis has been described (2, 21, 24, 29, 50, 51, 54). One recent report showed increased expression of TGF- in isolated glomeruli from Dahl salt-sensitive rats that had been made hypertensive by a 4-wk treatment with an 8.0% NaCl diet. These animals demonstrated marked nephrosclerosis (47). Although a time course was not performed and control animals that maintained a normal blood pressure and did not manifest nephrosclerosis on the 8.0% NaCl diet were not examined in that study, by directly modifying TGF- expression in glomeruli, the role dietary salt may play in facilitating glomerulosclerosis through a shear stress-related mechanism should be entertained.