INTRODUCTION

Top Abstract Introduction Materials and methods Results Discussion References

ONCE INITIATED, renal failure tends to be relentlessly progressive (33), and therapeutic strategies designed to slow or stop this progression have so far been largely ineffective. It is recognized both in animals with experimental renal disease and in humans that declining renal function correlates most strongly with the pathological changes seen in the tubulointerstitium of the kidney (1, 2). These pathological changes are manifest as interstitial fibrosis and scarring together with tubular atrophy (15). The presence of protein, most notably albumin, in the urine of patients with renal disease has conventionally been regarded simply as a marker of the severity of the disease state. Nonetheless, it is recognized that those patients with proteinuria are more likely to develop progressive renal failure than those without proteinuria (6), and recently it has been hypothesized that albuminuria may exert a toxic effect on proximal tubular epithelial cells (PTEC) in its own right, thereby damaging the cells and initiating the process of interstitial fibrosis and scarring (5, 8, 9, 30). Indeed, there is good evidence that proteinuria may result in tubular injury (5), although the mechanisms whereby such injury may occur are unclear. Furthermore, enhanced cellular proliferation has been observed in PTEC cultured in the presence of albumin or nephrotic urine (4). Again, however, the mechanism of this effect has not yet been elucidated.

One current theory suggests that the nephrotoxicity of albuminuric states is not directly determined by the protein molecule itself but that the toxicity resides in other molecules, particularly lipids, carried by albumin. Albumin contains many fatty acid binding sites (14), and it has been shown that in PTEC which are exposed to nephrotic levels of fatty acid-replete albumin, the resting cellular lipid pools suffer major perturbations (29). Furthermore, when proximal tubule segments are exposed to fatty acid-bearing albumin, but not fatty acid-free albumin, they are able to produce a lipid chemoattractant that may have an important role in the development of tubulointerstitial inflammation (17).

Lysophosphatidic acid (LPA) is an intercellular lipid mediator with growth factor-like activities (23, 24). LPA is rapidly produced and released from activated platelets and influences target cells by activating a specific G protein-coupled receptor that is present in numerous cell types (23). As a product of the blood clotting system, LPA is an abundant constituent of serum (but not plasma), where it is present in albumin-bound form. Albumin-bound LPA may account for much of the biological activity of serum (23). Extracellular LPA can also be generated through secretory phospholipase A₂ acting on microvesicles shed from blood cells challenged with inflammatory stimuli (10), suggesting that one of the in vivo functions of LPA is to stimulate proliferative responses at sites of injury and inflammation. Platelet aggregation is commonly observed in the glomerular capillaries in many renal diseases (7), and LPA released by activated platelets is likely to enter the proximal tubule when liberated in the glomerulus either alone or complexed to albumin. Therefore it is possible that LPA exerts receptor-mediated effects on the proximal tubule cells, and it is likely that this effect may be of considerable importance in PTEC pathophysiology.

Using opossum kidney (OK) cells, a kidney proximal tubule epithelial cell line (19) that has many characteristics of the proximal tubule (26), we investigated whether LPA may have a cell signaling function in PTEC by examining its effects on intracellular calcium. In addition, the effects of LPA on cell growth were studied, as renal cell hyperplasia occurs in many renal diseases associated with proteinuria, and it has been suggested that such growth represents a maladaptive response that contributes to the progression of renal failure (34).

	MATERIALS AND METHODS

Top Abstract Introduction Materials and methods Results Discussion References

Materials. OK cells are an immortalized line derived from opossum kidney and were obtained from Dr. J. Caverzasio (Geneva, Switzerland). All tissue culture media and chemicals were obtained from Sigma UK unless otherwise stated.

Cell culture. OK cells were maintained in DMEM-Ham's F-12 mix (DMEM-F12) (GIBCO), supplemented with 10% FCS (GIBCO), 2 mmol/l L-glutamine (GIBCO), 100 U/ml penicillin, and 100 µg/ml streptomycin (Flow Laboratories), at 37°C in a humidified 95% O₂-5% CO₂ atmosphere.

LPA. A stock solution of LPA (oleoyl) was prepared by dissolving it in a 1 mg/ml solution of fatty acid-free bovine serum albumin (FAF-BSA) and distilled water.

Intracellular calcium measurements in suspension. OK cells were grown to confluence in plastic flasks. Cell monolayers were washed twice in cell harvesting solution (0.54 mM sodium EDTA, 154 mM NaCl, and 10 mM HEPES; pH 7.3) and then incubated for 20 min at 37°C in this solution. The cells were gently removed by agitation, washed once in Krebs-Henseleit buffer (in mM: 1 CaCl₂, 118 NaCl, 469 KCl, 1.2 KH₂PO₄, 1.2 MgSO₄ · 7H₂O, 4.2 NaHCO₃, 10 glucose, and 10 HEPES; pH 7.4), then incubated in this solution at 37°C for 30 min prior to loading with the molecular probe. Measurements of intracytoplasmic free Ca²⁺ ([Ca²⁺]_i) were performed with fura 2-AM (Calbiochem).

For measurement of [Ca²⁺]_i, cells were resuspended at 2.5 × 10⁶ cells/ml in Krebs buffer with a 5 µM final concentration of fura 2-AM and incubated at 37°C in a water bath for 45 min (in the dark) with occasional mixing. After incubation the cells were washed once in Krebs buffer and resuspended at 1 × 10⁶ cells/ml in Krebs buffer. The cells were kept in a water bath at 37°C before the experiments were started. Fluorescence of 2 ml of this cellular suspension was monitored with a Perkin-Elmer LS-50B luminescence spectrometer in cuvettes thermostatically controlled at 37°C with constant stirring. Fluorescence of the cellular suspension was first determined using unlabeled cells to correct experimental measurements for autofluorescence. The cell suspension was excited alternately at 340 and 380 nm, and the fluorescence was measured at 510 nm. Ten-nanometer slit widths were used for both excitation and emission. After stabilization of the baseline, agonists were added in 20-µl volumes.

Graphic representations of [Ca²⁺]_i were computed by using the equation (13) where K_d is the dissociation constant of Ca²⁺ for fura 2 (224 nM at 37°C), R_min and R_max are the minimal and maximal fluorescent ratios obtained by perforating the cells with 0.1% Triton X-100 for R_max followed by the addition of an excess of EGTA at 5 mM for R_min. F_min (380 nm) and F_max(380 nm) are the fluorescent intensities after excitation at 380 nm, in the absence and presence of Ca²⁺, respectively.

Single cell intracellular calcium imaging. OK cells were grown for 24 h on sterilized coverslips (22 mm diameter; Chance Proper) in culture dishes (35 × 10 mm) in DMEM-F12 medium (as above) at 37°C. The coverslips were washed twice with Krebs buffer and then incubated in the dark, at room temperature, for 1 h in Krebs buffer supplemented with 5 µM fura 2-AM. The coverslips were then washed twice in Krebs buffer and incubated for a further 30 min to allow for complete de-esterification of the dye before being mounted on the stage of a Nikon Diaphot inverted epifluorescence microscope. Krebs buffer was continuously perfused over the cells at the rate of 5 ml/min. Before application of the stimuli, the buffer was perfused away, and the appropriate concentration of agonist was added to the cells after which the buffer was perfused over the cells once again washing away the stimuli. After subtraction of background fluorescence, images at wavelengths above 510 nm, after excitation at 340 and 380 nm (40 ms at each wavelength), were collected with an intensified charge-coupled device camera (Photonic Science). Experiments were conducted on a Quanticell 700 (Applied Imaging) system. Ratiometric values were converted to approximate [Ca²⁺]_i values using the above equation, in which R_min and R_max are the minimal and maximal fluorescent ratios obtained from a standard curve.

[³H]thymidine proliferation assay. OK cells were plated in 24-well plates and grown to 70-90% confluence. They were then incubated in serum-free and thymidine-free DMEM for 24 h at 37°C. The medium was then replaced with fresh serum-free DMEM alone (control), serum-free DMEM supplemented with 10% FCS, various concentrations of LPA, various concentrations of FAF-BSA, or 10 ng/ml human recombinant epidermal growth factor (EGF) (Calbiochem) as a control for pertussis toxin (PTX) sensitivity. After an additional 24 h, 2 µCi of [³H]thymidine (Amersham Life Science, UK) was added to all wells. In time course experiments, the media containing the agonist was washed out and replaced with serum-free media and incubated at 37°C for the remainder of the 24-h period. After 2-h incubation with [³H]thymidine, the cells were washed three times with DMEM, then incubated with 2 ml of ice-cold 5% trichloroacetic acid (TCA) for 1 h at 4°C. The TCA was removed, and the cells were washed once with fresh ice-cold TCA. Then 2 ml of ice-cold ethanol containing 200 µM potassium acetate was added to each well for 5 min. The cells were then incubated twice in 2 ml of 3:1 mixture of ethanol:ether for 15 min per incubation. After allowing the cell monolayers to air dry, cells were solubilized in 1 ml of 0.1 M sodium hydroxide. [³H]thymidine counts per min (cpm) were measured by adding samples to scintillation fluid (Ultima Gold, Packard) and counted by using a Packard 1900CA Tri-Carb liquid scintillation analyzer.

	RESULTS

Top Abstract Introduction Materials and methods Results Discussion References

Effect of LPA on [Ca²⁺]_i. Basal values of [Ca²⁺]_i averaged 50 ± 15 nM (n = 53) at 1 mM external calcium. As can be seen in Fig. 1A, LPA was found to evoke a typical [Ca²⁺]_i transient in fura 2-labeled OK cells. The calcium signal is initiated within seconds and reaches its peak within 15-20 s of LPA addition and, thereafter, declined to a plateau that was consistently observed to be higher than basal [Ca²⁺]_i. The maximal increase in [Ca²⁺]_i was seen with 100 µM LPA as shown in Fig. 1A. It must be noted at this point that the LPA applied to the cells is bound to FAF-BSA as a carrier; however, the levels of FAF-BSA bound to LPA were found not to have any effect on [Ca²⁺]_i levels in this system. For example, the amount of FAF-BSA present in administering a dose of 100 µM LPA is equivalent to 10 µg/ml FAF-BSA, which was found to have no effect on [Ca²⁺]_i levels (data not shown). The effect of LPA was concentration dependent in the range of 10⁸ to 10³ M (Fig. 1B). The dose response curve calculates an EC₅₀ of 2.6 ×10⁶ M, with 95% confidence intervals 9.7 × 10⁷ to 6.9 × 10⁶ M. 

View larger version (10K): [in this window] [in a new window]   Fig. 1.   A: effect of 100 µM lysophosphatidic acid (LPA) on the intracellular calcium ([Ca2+]i) levels of OK cells. This is a representative experiment of 5 replicate experiments. B: dose-response relationship of LPA-induced [Ca2+]i mobilization as determined from peak values. Data are presented as means ± SE (n = 3).

View larger version (11K): [in this window] [in a new window]   Fig. 2.   A: effect of EGTA pretreatment on LPA-induced [Ca2+]i mobilization in OK cells. Data presented as change in Ca2+-dependent fura-2 fluorescence. This is a representative experiment of 4 replicate experiments. B: effect of thapsigargin pretreatment on LPA-induced [Ca2+]i mobilization in OK cells. This is a representative experiment of 4 replicate experiments

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Effect of pertussis toxin (PTX) pretreatment on LPA-induced [Ca2+]i mobilization in OK cells. This is a representative experiment of 3 replicate experiments.

7

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Comparison of effect of parathyroid hormone (PTH) and LPA on [Ca2+]i mobilization in OK cells. This is a representative experiment of 3 replicate experiments.

View larger version (32K): [in this window] [in a new window]   Fig. 5.   Effect of 10 µM LPA on [Ca2+]i levels of 17 adherent OK cells as measured using the Quanticell-700 cell imaging system. This is a representative experiment of 3 replicate experiments. Arrow, time point at which 10 µM LPA was administered to the cells. Each line shown represents 1 of the 17 cells observed in this experiment. Results show that 14 of 17 cells responded with an increase in [Ca2+]i ranging from ~50 to 150 nM (average = 116 ± 42 nM).

View larger version (26K): [in this window] [in a new window]   Fig. 6.   Effect of LPA and its carrier BSA on proliferation of OK cells. This is a representative experiment of 5 replicate experiments. Results are presented as percentages (±SE, n = 4) of control, which was incubated with serum-free media. * P < 0.05. *** P < 0.001.

View larger version (42K): [in this window] [in a new window]   Fig. 7.   Effect of transient exposure of LPA on proliferation of OK cells. Exposure times of 5, 10, and 60 min and 24 h were used, after which cells were washed and reimmersed in serum-free media for the rest of the 24-h incubation time. This is a representative experiment of 3 replicate experiments. Results are presented as percentages (±SE, n = 4) of control, which was incubated with serum-free media. ** P < 0.01. *** P < 0.001.

View larger version (27K): [in this window] [in a new window]   Fig. 8.   Effect of PTX pretreatment on LPA-, FCS-, and epidermal growth factor (EGF)-induced proliferation in OK cells. Cells were pretreated for 18 h with 50 ng/ml PTX and then stimulated with agonist for 24 h. This is a representative experiment of 3 replicate experiments. Results are presented as percentages (±SE, n = 4) of control, which was incubated with serum-free media. ** P < 0.01 and *** P < 0.001, for comparison between control cells (no PTX treatment) and PTX-pretreated cells.

	DISCUSSION

Top Abstract Introduction Materials and methods Results Discussion References

The pathophysiology of progressive renal scarring in renal disease is poorly understood, but it is likely to be multifactorial. The powerful association between proteinuria, tubulointerstitial scarring, and renal disease progression has led to the hypothesis that either proteinuria per se, or some other unidentified bioactivity in nephrotic glomerular ultrafiltrate, may play an important role in the development of renal scarring and inflammation (3, 28). It is notable therefore that associated with the filtered protein are large quantities of lipid material that enter the proximal tubule in large part complexed with albumin. Lipid molecules can be presented to proximal tubule cells at concentrations far in excess of their maximum solubility by virtue of their capacity to bind to albumin, and indeed intracellular lipid droplets in proximal tubular epithelial cells are a prominent feature of proteinuric states (15).

The potential pathophysiological effects of this filtered lipid material have not been well studied, but it has been demonstrated that incubation of proximal tubular cells with lipidated albumin has profound effects on cell lipid metabolism (29). Furthermore, proximal tubular cells exposed to lipidated albumin are able to produce a monocyte chemoattractant. Exposure of these cells to delipidated albumin, however, does not result in the production of this chemoattractant substance (18). The precise lipid responsible for this effect is unclear, but observations such as these clearly implicate albumin-bound lipids as potential mediators of proximal tubular cell stimulation and toxicity.

In view of the paucity of knowledge regarding lipid signaling in the proximal tubule, the aim of this study was to investigate the effects of potentially important lipid mediators on proximal tubular cell function. LPA was a particularly attractive candidate to examine, since it is likely to be liberated in substantial quantities in inflamed glomeruli and subsequently is very likely to enter the proximal tubule. Specifically, the study set out to investigate whether LPA could stimulate [Ca²⁺]_i changes, as this is a mechanism by which cells regulate many of their activities and responses to extracellular stimuli. In addition, the effect of LPA on proximal tubule cell growth was studied, since derangements of proximal tubule cell growth have been implicated in the progression of renal disease (34).

LPA was found to stimulate a classic [Ca²⁺]_i transient in OK cells that was dose dependent. The EC₅₀ of this response to LPA was higher (2.6 × 10⁶ M) than has been observed in some other cell types. Furthermore, the magnitude of the [Ca²⁺]_i increase was found to be relatively small compared with the [Ca²⁺]_i responses observed in some cell types after LPA (24). The EC₅₀ could be high due to this being a pathophysiological response rather than a physiological process. Nonetheless, the [Ca²⁺]_i responses were found to be comparable to those observed in other investigations in OK cells utilizing established Ca²⁺-mobilizing agonists such as PTH (22). Our results show the [Ca²⁺]_i response to LPA to be biphasic, reflecting an initial phase of Ca²⁺ release from intracellular stores followed by a more prolonged signal that is dependent on Ca²⁺ entry into the cell. In the absence of extracellular Ca²⁺ (i.e., in the presence of 4 mM EGTA), the response was limited to a transient [Ca²⁺]_i rise. This entry of Ca²⁺ into the cell is a universal feature of nonexcitable cells and the most widely accepted model for regulation of Ca²⁺ entry, termed the "capacitative" model, which argues that entry is activated by prior depletion of the inositol 1,4,5-triphosphate-mediated intracellular Ca²⁺ stores (27). Classically, this capacitative Ca²⁺ entry pathway is activated for the duration of store emptying; once triggered, it is independent of the presence or absence of agonist and is inhibited on repletion of the stores (27).

LPA is commonly considered to exert its action via G protein-coupled receptors (32). Results from many studies indicate that the LPA receptor couples to at least three distinct G proteins (24): G_q, which links the receptor to phospholipase C; G_12/13, which mediates Rho activation; and G_i, which triggers Ras-GTP accumulation and inhibition of adenylyl cyclase. As we found the [Ca²⁺]_i response to be immediate upon administration of LPA and insensitive to PTX pretreatment, it would seem that this effect of LPA is receptor mediated and most likely occurs via a PTX-insensitive G protein-linked receptor.

Reports in the literature implicate at least three different G proteins in the transduction of LPA responses in different systems. The obvious question raised by these observations has been whether the responses to LPA stimulation are mediated by unifunctional receptors or whether one type of receptor can mediate its varied responses. A recent report (12) has addressed this issue and has demonstrated that a single LPA receptor can activate multiple LPA-dependent responses in cells from distinct tissue lineages.

One problem of [Ca²⁺]_i measurements in a cell suspension system is that the fluorescence signal obtained is an integration of the signal obtained from many individual cells. Therefore, as LPA stimulated modest [Ca²⁺]_i transients in the suspended OK cells, we hypothesized that this may be due to a fraction of the cells responding with a large [Ca²⁺]_i transient and this signal being diluted due to a proportion of cells showing no response. Consequently, similar experiments were carried out on small populations of adherent cells using the Quanticell-700 cell imaging system. This showed that the great majority (>80%) of the OK cells respond with a [Ca²⁺]_i transient following stimulation with LPA. Also, this system produced results similar to the fluorometer in terms of the magnitude of the [Ca²⁺]_i responses, hence substantiating the initial data.

Both LPA and FAF-BSA were found to exert a mitogenic effect on OK cells, with LPA causing a maximum 5-fold increase in [³H]thymidine incorporation at 100 µM and FAF-BSA causing a 1.5-fold increase at a maximum concentration of 10 µg/ml. LPA was still significantly mitogenic at a concentration of 10 µM, whereas the amount of carrier FAF-BSA present at this dose (1 µg/ml) failed to cause any effect on thymidine incorporation. Therefore, these data suggest that LPA is the predominant stimulator. Although most biological effects of LPA on other cell types are mediated by nanomolar concentrations of the phospholipid, LPA-stimulated proliferation has been found to require micromolar concentrations (16). This was true in our study as well, in which stimulation of proliferation required micromolar concentrations of LPA. The effect of 100 µM LPA in the current studies was found to be equal to that of 10% FCS. The results of the present study are much more marked than the results documented recently with LPA and mouse renal proximal tubule cells (21). These authors demonstrated only a 1.5-fold increase in thymidine incorporation with 100 µM LPA under similar conditions. We have also demonstrated that, in common with many mitogens, a transient 5-min exposure of LPA to the cells is sufficient to produce subsequent proliferation.

The mitogenic effect of LPA was found to be highly PTX sensitive. As discussed above, the LPA receptor couples to at least three distinct G proteins (24). Our data suggest that LPA acts through a PTX-insensitive receptor to cause subsequent [Ca²⁺]_i mobilization; however, the mitogenic action of LPA would seem to occur through a PTX-sensitive receptor system. This pattern has also been reported previously, which suggests that LPA-stimulated mitogenesis (PTX-sensitive) in fibroblasts is independent of the PTX-insensitive G_q-mediated [Ca²⁺]_i mobilization (32).

The role of tyrosine kinases in LPA-mediated proliferation of OK cells in this study is still uncertain. Although we found no inhibition of thymidine incorporation by preincubation of the cells with appropriate concentrations (25) of genistein and herbimycin A, the possibility remains that tyrosine kinases unaffected by these inhibitors are still involved.

A number of functions of proximal tubular cells have been identified which suggest that they can take part in the process of inflammation and scarring. These functions include the production of matrix proteins, proinflammatory cytokines, and chemotactic substances (5). This is not surprising, considering that embryologically they are derived from mesenchymal cells, as are fibroblasts and the cells of the immune system (20). Our evidence would suggest that lipids such as LPA may directly modulate tubular cell function, altering both their growth characteristics and possibly phenotype. It is interesting to speculate that a consequence of LPA signaling within the proximal tubule epithelium may be the production of proinflammatory substances such as growth factors and cytokines (11). This issue is the subject of ongoing research in our laboratory.

In summary, ours is the first report to demonstrate that LPA causes [Ca²⁺]_i signaling within renal proximal tubule cells, and although it has been recently reported that LPA is a mitogenic factor in mouse proximal tubule cells, our evidence suggests that LPA is a considerably more potent mitogenic factor than previously reported (21) and that its action occurs through G_i or a related PTX substrate.