INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

PROSTAGLANDINS (PGs) are potent autocoids formed from arachidonic acid via the action of cyclooxygenase under both physiological and pathological conditions (9). The actions of prostanoids are quite diverse, exerting cellular (1) and systemic (25) effects dependent on pharmacological concentration (36) and/or the presence of distinct PG receptor subtypes (12). Prostaglandin E₂ (PGE₂) is a major metabolite of arachidonic acid in mammalian kidney (34), where it modulates renal hemodynamics, metabolism, water and ion transport, and sympathetic nerve activity (18, 27). PGs also protect a variety of tissues, including the kidney (30), from a diverse array of toxicants, and this property has been termed "cytoprotection" (16, 32). Because PGE₂-mediated cytoprotection is observed in cultured cells, a cellular mechanism of action has been proposed (30, 33), although the cellular and molecular mechanisms associated with this response remain unknown.

The pharmacological effects of PGE₂ are primarily receptor mediated, and currently four PGE₂ receptor [E-prostanoid (EP)] subtypes have been cloned, namely the EP₁, EP₂, EP₃, and EP₄ receptors (for a review, see Refs. 11, 12, 17, 31 and references therein). EP₁ receptors are coupled to inositol phospholipid (IP)-related signal transduction, whereas cyclic nucleotide metabolism is regulated positively by EP₂/EP₄ receptors and negatively by EP₃ receptors (31). EP₃ alternative splice variants have been identified that differentially couple to cyclic nucleotide- or IP-related signal transduction (17). Modulation of adenosine 3',5'-cyclic monophosphate (cAMP) metabolism and IP turnover, in turn, regulates the activity of a cAMP-dependent protein kinase A (PKA) and PKC, respectively (22).

PKC represents a multigene family of protein kinases similar in size, structure, and function that transduce signals from a wide variety of stimuli, including growth factors, hormones, and neurotransmitters (for a review, see Ref. 28). The role of PKC in IP-related signaling has been firmly established. In this pathway, IPs are hydrolyzed by phospholipase C to the ultimate second messengers, inositol-1,4,5-trisphosphate (IP₃) and diacylglycerol (DAG) (28). IP₃ releases calcium from intracellular stores, whereas DAG is the endogenous activator of PKC (28). Endogenous DAG and exogenously added cell-permeant DAG analogs are rapidly metabolized, limiting their application in studying PKC-related signaling (28). For this reason, phorbol esters, metabolically stable tumor promoters that are potent activators of PKC, have been used as surrogates of DAG (4).

PKC plays an important role in the transduction of multiple signals into the nucleus (29), including activation of the activator protein-1 (AP-1) transcriptional complex, which is considered a "nuclear third messenger" in this pathway (23). The AP-1 transcription factor is a heterodimeric complex composed of c-jun (c-Jun, Jun B, Jun D) and c-fos (c-Fos, Fos B, Fra-1) protooncogene family members, as either a Jun/Jun homodimer or Jun/Fos heterodimer (23, 29). On formation, this complex specifically binds to a target DNA sequence (TGAC/GTCA) referred to as the 12-O-tetradecanoylphorbol-13-acetate (TPA) responsive element (TRE) (23). TPA-induced binding of AP-1 to the TRE is independent of protein synthesis, suggesting this response is regulated by posttranslational modification of existing proteins (2, 3).

Conjugation of hydroquinone with glutathione (GSH) in the liver results in the formation of several isomeric nephrotoxic GSH conjugates (20). Histochemical analysis of kidney sections following administration of the most potent metabolite, 2,3,5-(trisglutathion-S-yl)-hydroquinone [2,3,5-(trisglutathion-S-yl)-HQ], shows cytotoxicity initially localized to proximal tubule epithelial cells in the S3 segment (24). The present studies were conducted to determine whether PGE₂ affords cytoprotection against 2,3,5-(trisglutathion-S-yl)-HQ-mediated cytotoxicity in a renal proximal tubule epithelial cell line (LLC-PK₁) and to determine the cellular and molecular nature of the cytoprotective response. We report that PGE₂ offers cytoprotection against chemical-induced cytotoxicity in LLC-PK₁ cells and that this event is mediated by a PKC-coupled receptor, which is pharmacologically distinct from currently classified EP receptor subtypes.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Chemicals. 2,3,5-(Trisglutathion-S-yl)-HQ was synthesized as previously described (24) and was >99% pure as determined by high-performance liquid chromatography. PGE₂, PGE₂ methyl ester, 17-phenyltrinor PGE₂ (PT-PGE₂), 11-deoxy-16,16-dimethyl PGE₂ (DDM-PGE₂), sulprostone, PGE₁, 11-deoxy PGE₁, and PGA₂ were obtained from Cayman Chemical (Ann Arbor, MI). N-(2{[3-(4-bromophenyl)-2-propenyl]-amino}-ethyl)-5-isoquinolinesulfonamide (H-89) and calphostin C were purchased from Calbiochem (La Jolla, CA). Formaldehyde, glacial acetic acid, glycerol, and ethanol were purchased from Fisher Scientific (Houston, TX). TRE and AP2 consensus sequences were purchased from Promega (Madison, WI). [-³²P]ATP (3,000 Ci/mmol) was obtained from NEN (Beverly, MA). Poly[d(I-C)] was purchased from Boehringer-Mannheim (Indianapolis, IN). 4-bromo-A-23187 (4-Br-A-23187) was a product of Molecular Probes (Eugene, OR). All other chemicals were from Sigma Chemical (St. Louis, MO).

Cell culture. LLC-PK₁ cells were obtained from the American Type Culture Collection (CL101) at passage 181. Cells were maintained in Dulbecco's modified Eagle's medium (DMEM) (JRH Biosciences, Lenexa, KS) supplemented with 4 g/l D-glucose (Sigma) and 10% fetal bovine serum (FBS) (Atlanta Biologicals, Norcross, GA) in 5% CO₂-95% air at 37°C. Cells were subcultured by trypsinization, and all experiments were conducted with 5-day postconfluent cultures at passage levels 187-200.

Pretreatment of LLC-PK₁ cells with PGs. LLC-PK₁ cells were seeded in 24-well plates and maintained in 10% FBS-DMEM until 5 days postconfluency, and media were replaced every 2 days. Postconfluent cultures were rinsed with phosphate-buffered saline (PBS) and exposed to PGs in 10% FBS-DMEM for specified periods of time. Prior to chemical challenge, media were aspirated, and cell monolayers were rinsed three times with PBS to remove residual PGs.

Neutral red assay. The neutral red assay employs lysosomal membrane integrity as an index of cell viability. This assay affords the most sensitive detection of chemical challenge in LLC-PK₁ cells compared with other cytotoxicity assays, such as those that measure plasma membrane integrity (lactate dehydrogenase leakage) or mitochondrial function [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay] (26). Vehicle or PG-pretreated cells were rinsed three times with PBS and exposed to 2,3,5-(trisglutathion-S-yl)HQ (300 µM) in 0.1% FBS-DMEM supplemented with 25 mM N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES) (pH 7.4) for 2 h in a final volume of 0.5 ml. HEPES was required to maintain physiological pH following addition of 2,3,5-(trisglutathion-S-yl)-HQ due to the presence of small amounts of acetic acid in the purified product. After chemical challenge, cells were washed three times with PBS and exposed to 50 µg/ml neutral red in 0.1% FBS-DMEM supplemented with 25 mM HEPES (pH 7.4) for 1 h. To quantify neutral red uptake, monolayers were washed once with 1 ml of a 1% formaldehyde-1% calcium chloride solution followed by aspiration. Neutral red was extracted from the cells with 1 ml of a 1% glacial acetic acid-50% ethanol solution for 15 min at room temperature while protected from light. The extracted dye was quantified at 540 nm using a Shimadzu UV-160 spectrophotometer, and results are expressed as percent of control. For time-course studies, values are expressed as percent of protection rather than percent of control, using the following equation to calculate the value: [(neutral red absorbance_A)  (neutral red absorbance_B)]/[(neutral red absorbance_C)  (neutral red absorbance_B)], where neutral red absorbance_A is LLC-PK₁ cells pretreated with DDM-PGE₂ and subsequently exposed to 2,3,5-(trisglutathion-S-yl)-HQ; neutral red absorbance_B is treatment with 2,3,5-(trisglutathion-S-yl)-HQ alone; and neutral red absorbance_C is absorbance for control groups.

Electrophoretic mobility shift assay (EMSA). EMSAs were carried out as described previously (6). LLC-PK₁ cells were collected and lysed in buffer A [25 mM HEPES, pH 7.6, 1.5 mM EDTA, 10% glycerol, 1 mM dithiothreitol (DTT), and 0.1 mg/ml phenylmethylsulfonyl fluoride] using 20 strokes with a Dounce homogenizer. The homogenate was centrifuged at 12,000 revolutions/min (rpm) in an Eppendorf microcentrifuge at 4°C for 5 min, and the supernatant was discarded. The remaining pellet was centrifuged for 10 s, and the residual supernatant was aspirated. The pellet was extracted with 40 µl of buffer A supplemented with 0.5 M KCl for 1 h on ice. Extracted pellets were centrifuged at 14,000 rpm for 20 min at 4°C, and the supernatant was designated the nuclear extract. Protein concentrations were determined by the method of Bradford (7) using bovine serum albumin as standard. For EMSAs, 10 µg of nuclear extract were incubated in a reaction mixture consisting of 18.8 mM HEPES, 40 mM KCl, 1.1 mM EDTA, 7.5% glycerol, 0.75 mM DTT, and 62.5 ng/µl poly[d(I-C)] for 15 min at 20°C to reduce interference by nonspecific DNA binding proteins. Addition of 3.5 nM TRE probe labeled with [-³²P]ATP was added for 15 min to determine AP-1 binding activity. Bound TRE was separated on a 5% polyacrylamide nondenaturing gel for 2 h at 120 V, dried, and, exposed to Hyperfilm-MP (Amersham) for autoradiography.

PKC assay kit. Measurements of PKC activity were obtained using a PKC assay kit (GIBCO-BRL, Gaithersburg, MD), as described previously (37).

Statistics. Individual comparisons were made using the Student's t-test or analysis of variance with a post hoc Student-Newman-Keuls test, as appropriate. P < 0.05 was accepted as significant.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

The neutral red assay affords the most sensitive detection of quinone-thioether-mediated cytotoxicity in LLC-PK₁ cells (26) and was therefore used in the present studies. Pretreatment of LLC-PK₁ cells with 0.01-40 µM PGE₂ for 24 h protected against 2,3,5-(trisglutathion-S-yl)-HQ-mediated cytotoxicity in a concentration-dependent manner (Fig. 1, control values are 100%). In contrast, cotreatment of LLC-PK₁ cells with PGE₂ (0.1-40 µM) did not protect against 2,3,5-(trisglutathion-S-yl)-HQ-mediated cytotoxicity (data not shown). PG-treated groups alone were not different from respective controls (data not shown).

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Prostaglandin E2 (PGE2)-mediated cytoprotection in LLC-PK1 cells. LLC-PK1 cells were pretreated with 0.01-40 µM PGE2 () in 10% fetal bovine serum-Dulbecco's modified Eagle's medium (FBS-DMEM) for 24 h and subsequently exposed to 300 µM 2,3,5-(trisglutathion-S-yl)-hydroquinone [2,3,5-(trisglutathion-S-yl)-HQ] for 2 h. Cell viability measurements were obtained using a neutral red assay, as described in MATERIALS AND METHODS. Values represent means ± SE (n = 3). * Significantly different from control,  significantly different from the 2,3,5-(trisglutathion-S-yl)-HQ-treated group; P < 0.05. Similar results were observed in 2 separate experiments.

In our initial studies with PGE₂, the cytoprotective dose-response curve was variable, and we suspected this trend was secondary to chemical stability. It is well known that PGE₂ undergoes -elimination in basic solutions (pH > 7.4) to form the corresponding A- and B-series PGs (15, 19), and minor changes toward alkaline pH occur during routine handling of cells in culture. Thus LLC-PK₁ cells were treated with PGE₂ in media buffered to pH 7.4 or 7.8 to assess the pH and structural dependence of the cytoprotective response. Pretreatment of cells with 10 µM PGE₂ for 24 h at pH 7.8 eliminates its cytoprotective properties, whereas pretreatment at pH 7.4 retains the cytoprotective properties (Fig. 2). Increasing the pH from 7.4 to 7.8 did not affect cell viability or 2,3,5-(trisglutathion-S-yl)-HQ-mediated cytotoxicity. These data suggest a structural requirement for cytoprotection and demonstrate that minor changes in pH can dramatically effect the cytoprotective response to PGE₂.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Effect of pH on PGE2-induced cytoprotection. LLC-PK1 cells were pretreated with 10 µM PGE2 in 10% FBS-DMEM + HEPES buffered to pH 7.4 (open bars) or pH 7.8 (solid bars) for 24 h and subsequently challenged with 300 µM 2,3,5-(trisglutathion-S-yl)-HQ in 0.1% FBS-DMEM for 2 h. Cell viability measurements were obtained using a neutral red assay as described in MATERIALS AND METHODS. Values represent means ± SE (n = 3). * Significantly different from respective control,  significantly different from respective 2,3,5-(trisglutathion-S-yl)-HQ-treated group; P < 0.05. Similar results were observed in 2 separate experiments.

In subsequent studies, cytoprotection against 2,3,5-(trisglutathion-S-yl)-HQ-mediated cytotoxicity was evaluated in cells pretreated with EP receptor agonists (PT-PGE₂, 11-deoxy PGE₁, sulprostone), a synthetic stable PGE₂ analog (DDM-PGE₂), PGE₁, or PGA₂. Because synthetic analogs exhibit EP receptor subtype specificity and/or enhanced stability in basic solutions, they offer advantages over PGE₂ for evaluating biological activity in vitro. PG pretreatment alone did not affect cell viability, relative to controls (DDM-PGE₂, 109 ± 1.7%; PT-PGE₂, 102 ± 0.6%; 11-deoxy PGE₁, 100 ± 2.8%; sulprostone, 104 ± 1.7%; PGE₁, 100 ± 3.1%; PGA₂, 91 ± 5.2%). Cytoprotection against 2,3,5-(trisglutathion-S-yl)-HQ (300 µM; 2 h)-mediated cytotoxicity was only observed in cells pretreated with DDM-PGE₂ but not in cells pretreated with agonists for the EP_1-4 receptor subtypes (PT-PGE₂, 11-deoxy PGE₁, sulprostone; all concentrations were 20 µM). In addition, neither 20 µM PGA₂ nor 40 µM PGE₁ induced cytoprotection against 2,3,5-(trisglutathion-S-yl)-HQ-mediated cytotoxicity (Fig. 3).

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Prostaglandin-mediated cytoprotection is structurally limited. LLC-PK1 cells were pretreated with 17-phenyltrinor PGE2 (PT-PGE2, open bar), 11-deoxy PGE1 (solid bar), sulprostone (shaded bar), 11-deoxy-16,16-dimethyl PGE2 (DDM-PGE2, latticed bar), PGE1 (left hatched bar), or PGA2 (right hatched bar) in 10% FBS-DMEM for 24 h (all concentrations were 20 µM, except PGE1, which was 40 µM) and subsequently challenged with 300 µM 2,3,5-(trisglutathion-S-yl)-HQ in 0.1% FBS-DMEM for 2 h. Cell viability measurements were obtained using a neutral red assay as described in MATERIALS AND METHODS. Values represent means ± SE (n = 3).  Significantly different from the 2,3,5-(trisglutathion-S-yl)-HQ-treated group, P < 0.05. Similar results were observed in at least 2 separate experiments.

Thus only DDM-PGE₂ reproduced PGE₂-mediated cytoprotection in LLC-PK₁ cells. This stable analog was therefore used to further investigate the mechanism of PGE₂-mediated cytoprotection. The induction of cytoprotection by DDM-PGE₂ exhibited time dependence (Fig. 4). Cytoprotection against 2,3,5-(trisglutathion-S-yl)-HQ-mediated cytotoxicity was first detected after an 8 h exposure to 1 µM DDM-PGE₂, with the maximal cytoprotective response occurring between 20 and 24 h.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Time-dependent induction of cytoprotection by DDM-PGE2. LLC-PK1 cells were pretreated with 1.0 µM DDM-PGE2 in 10% FBS-DMEM for 0.5-24 h and subsequently challenged with 300 µM 2,3,5-(trisglutathion-S-yl)-HQ in 0.1% FBS-DMEM for 2 h. Cell viability measurements were obtained using a neutral red assay as described in MATERIALS AND METHODS. Values represent means ± SE (n = 3). Significantly different from the 2,3,5-(trisglutathion-S-yl)-HQ-treated group, P < 0.05. Similar results were observed in 2 separate experiments.

Pretreatment of LLC-PK₁ cells with 0.001-40 µM DDM-PGE₂ for 24 h resulted in a concentration-related induction of cytoprotection against 2,3,5-(trisglutathion-S-yl)-HQ-mediated cytotoxicity (Fig. 5). Cytoprotection was maximal between 0.1-1.0 µM DDM-PGE₂ in LLC-PK₁ cells and, therefore, all subsequent pretreatment protocols employed 1 µM DDM-PGE₂ for 24 h.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Concentration-dependent induction of cytoprotection by DDM-PGE2. LLC-PK1 cells were pretreated with 0.001-40 µM DDM-PGE2 in 10% FBS-DMEM for 24 h and subsequently challenged with 300 µM 2,3,5-(trisglutathion-S-yl)-HQ in 0.1% FBS-DMEM for 2 h. Cell viability measurements were obtained using a neutral red assay as described in MATERIALS AND METHODS. Values represent means ± SE (n = 3).  Significantly different from the 2,3,5-(trisglutathion-S-yl)-HQ-treated group, P < 0.05. Similar results were observed in 2 separate experiments.

PGE₂-mediated responses are associated with cyclic nucleotide- or IP-related signal-transduction pathways in mammalian cells (17). To determine whether these pathways are associated with the cytoprotective response, LLC-PK₁ cells were pretreated with agents modulating cAMP- [forskolin, dibutyryl cAMP (DBcAMP)], calcium (4-Br-A-23187)-, or PKC (TPA)-related signaling. Pretreatment of cells with forskolin (0.1-30 µM), DBcAMP (0.0001-1.0 mM), or 4-Br-A-23187 (0.001-1.0 µM) for 24 h did not induce cytoprotection against 2,3,5-(trisglutathion-S-yl)-HQ-mediated cytotoxicity (data not shown), consistent with the negative effects of the EP receptor agonists (see Fig. 3). The ability of forskolin to increase cAMP levels in LLC-PK₁ cells was verified using a cAMP radioimmunoassay. Calcium-related signaling induced by 4-Br-A-23187 was verified by measuring the induction of calcium-responsive genes, including gadd 153 (data not shown). These data indicate that neither cAMP- or calcium-related signal-transduction pathways are involved in the cytoprotective response of LLC-PK₁ cells to PGE₂. In contrast, pretreatment of cells with TPA (10 ng/ml, 24 h), a potent PKC activator, protected against 2,3,5-(trisglutathion-S-yl)-HQ-mediated cytotoxicity (Fig. 6A).

View larger version (37K): [in this window] [in a new window]   Fig. 6.   Modulation of DDM-PGE2- and TPA-induced cytoprotection by the protein kinase inhibitor N-(2{[3-(4-bromophenyl)-2-propenyl]-amino}-ethyl)-5-isoquinolinesulfonamide (H-89). LLC-PK1 cells were pretreated with 2.0-60 µM H-89 for 1 h prior to treatment with 10 ng/ml 12-O-tetradecanoylphorbol-13-acetate (TPA, A) or 1 µM DDM-PGE2 (B) for 24 h and subsequently challenged with 300 µM 2,3,5-(trisglutathion-S-yl)-HQ for 2 h. Cell viability measurements were obtained using a neutral red assay as described in MATERIALS AND METHODS. Values represent means ± SE (n = 3). * Significantly different from control, P < 0.05.  Significantly different from the 2,3,5-(trisglutathion-S-yl)-HQ [2,3,5-(trisGSyl)HQ]-treated group, P < 0.05. f Significantly different from groups pretreated with DDM-PGE2 or TPA and subsequently challenged with 2,3,5-(trisglutathion-S-yl)-HQ, P < 0.05. Similar results were observed in 2 separate experiments.

To further differentiate between PKC and PKA signaling in the cytoprotective response to DDM-PGE₂, LLC-PK₁ cells were pretreated with the protein kinase inhibitor H-89, which inhibits PKA in the nanomolar range and PKC in the micromolar range (8). Inhibition of TPA-mediated cytoprotection by H-89 was observed as the concentration approached the inhibitory constant (K_i, 31.7 µM) for PKC but not at a concentration (2 µM) 40-fold higher than the K_i (48 nM) for PKA (Fig. 6A). The ability of H-89 to inhibit TPA-mediated cytoprotection in LLC-PK₁ cells indicates that concentrations of H-89 sufficient for the inhibition of PKC are reached in this system. Similar studies were conducted with the DDM-PGE₂ analog to functionally implicate PKC activity in the cytoprotective response. H-89 (20-60 µM) also inhibited DDM-PGE₂-mediated cytoprotection, as the concentration approached the K_i for PKC (Fig. 6B). Inhibition of TPA- and DDM-PGE₂-mediated cytoprotection by H-89 was not secondary to H-89-mediated cytotoxicity (control, 100 ± 2.9%; 60 µM H-89, 101 ± 3.5%). Although the use of a more selective PKC inhibitor would have been desirable in these experiments, calphostin C was extremely cytotoxic to LLC-PK₁ cells, causing cell lysis within the first 4 h of treatment (data not shown).

DDM-PGE₂-mediated cytoprotection exhibited a latency suggestive of the involvement of transcriptional activity or activities (Fig. 4). PKC regulates the activity of the AP-1 transcriptional complex (23), and, therefore, we examined the induction of TRE binding activity as preliminary evidence for a transcriptional component in the cytoprotective response and as a molecular marker of PKC activation (29). Peak TRE binding activity occurred 2 h after treatment of LLC-PK₁ cells with 1 µM DDM-PGE₂ (Fig. 7A) or 1 h after treatment with 10 ng/ml TPA (Fig. 7B). The induction of TRE binding activity in nuclear extracts from cells treated with TPA (0.1-100 ng/ml) for 1 h or DDM-PGE₂ (0.05-10 µM) for 2 h exhibited concentration dependence (Fig. 8). Specificity of the binding response was confirmed for control, 1.0 µM DDM-PGE₂, and 10 ng/ml TPA groups by addition of unlabeled TRE, which competitively eliminated the inducible band, or unrelated, DNA (AP2 responsive element), which was without effect. H-89 (0.2-60 µM) pretreatment (1 h) fully inhibited DDM-PGE₂-mediated TRE binding activity as the H-89 concentration approached the K_i for PKC (Fig. 9A). Similarly, 20 µM H-89 dramatically reduced peak TRE binding activity induced by 10 ng/ml TPA (Fig. 9B). The induction of TRE binding activity by DDM-PGE₂ is consistent with a role for transcriptional activities in the cytoprotective response.

View larger version (55K): [in this window] [in a new window]   Fig. 7.   TPA responsive element (TRE) binding activity in nuclear extracts from LLC-PK1 cells treated with DDM-PGE2 (A) or TPA (B). Nuclear extracts from LLC-PK1 cells treated with 1 µM DDM-PGE2 or 10 ng/ml TPA for 0.5-7 h were incubated with a 32P-labeled TRE in an electrophoretic mobility shift assay (EMSA) as described in MATERIALS AND METHODS. Protein-DNA complexes were separated on a 5% native polyacrylamide gel and visualized by autoradiography. Similar results were observed in 2 separate experiments. AP-1, activator protein-1.

View larger version (46K): [in this window] [in a new window]   Fig. 8.   Concentration-related induction of peak TRE binding activity in nuclear extracts from LLC-PK1 cells treated with DDM-PGE2 or TPA. Nuclear extracts from LLC-PK1 cells treated with 0.05-10 µM DDM-PGE2 or 0.1-100 ng/ml TPA were incubated with a 32P-labeled TRE in an EMSA as described in MATERIALS AND METHODS. Protein-DNA complexes were separated on a 5% native polyacrylamide gel and visualized by autoradiography. Specificity for the binding reaction was confirmed by addition of excess unlabeled TRE or excess nontarget DNA (AP2 consensus sequence). Similar results were observed in 2 separate experiments. Top arrowhead, AP-1:TRE; bottom arrowhead, free TRE.

View larger version (49K): [in this window] [in a new window]   Fig. 9.   Inhibition of DDM-PGE2 (A)- or TPA (B)-inducible TRE binding activity by H-89. LLC-PK1 cells were pretreated with H-89 prior to challenge with DDM-PGE2 or TPA, and nuclear extracts were prepared and incubated with a 32P-labeled TRE in an EMSA as described in MATERIALS AND METHODS. A: effects of 0.2-60 µM H-89 on inducible TRE binding activity in nuclear extracts from LLC-PK1 cells treated with 1µM DDM-PGE2. B: effect of 20 µM H-89 on inducible TRE binding activity in nuclear extracts from LLC-PK1 cells treated with TPA. Protein-DNA complexes were separated on a 5% native polyacrylamide gel and visualized by autoradiography. Similar results were observed in 2 separate experiments. Top arrowhead, AP-1:TRE; bottom arrowhead, free TRE.

Structure-function relationship studies were extended to measurements of TRE binding activity to further evaluate the presence of PKC-coupled EP receptors. Induction of TRE binding activity is directly associated with receptor-mediated activation of PKC and thus would represent a more sensitive indication of the presence of these receptors, since it is unlikely that chemical stability issues would significantly impact this response. TRE binding activity was observed in LLC-PK₁ cells treated for 2 h with DDM-PGE₂ and PGE₂ but not PT-PGE₂, sulprostone, or PGE₁ (all concentrations were 20 µM, Fig. 10), consistent with the negative effects of the EP receptor agonists. Specificity for the binding reaction was confirmed by incubation of nuclear extracts with unlabeled TRE or nontarget DNA (AP2 responsive element).

View larger version (79K): [in this window] [in a new window]   Fig. 10.   TRE binding activity in nuclear extracts from LLC-PK1 cells treated with PT-PGE2, DDM-PGE2, sulprostone, PGE1, or PGE2. Nuclear extracts from prostaglandin (20 µM)-treated LLC-PK1 cells were incubated with a 32P-labeled TRE in an EMSA as described in MATERIALS AND METHODS. Protein-DNA complexes were separated on a 5% continuous polyacrylamide gel and visualized by autoradiography. Specificity for the binding reaction was confirmed by addition of excess unlabeled TRE or nontarget DNA (AP2 consensus sequence). Similar results were observed in 2 separate experiments. DMSO, dimethyl sulfoxide. Top arrowhead, AP-1:TRE; bottom arrowhead, free TRE.

To rule out the possibility that DDM-PGE₂-mediated TRE binding activity occurs via the direct activation of PKC by this analog, in a manner similar to phorbol ester, PKC activity was measured in isolated cell homogenates treated with TPA or DDM-PGE₂. In contrast to TPA, DDM-PGE₂ did not increase PKC activity in isolated LLC-PK₁ cell homogenates (data not shown).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

We have shown that PGE₂ induces cytoprotection against quinone-thioether-mediated cytotoxicity in LLC-PK₁ cells (Fig. 1). Moreover, structure-activity relationships do not support a role for any of the presently classified EP receptor subtypes in the cytoprotective response. In contrast, DDM-PGE₂, a stable analog of PGE₂, acts as a potent agonist for the putative receptor-mediated effects of PGE₂ (Fig. 3). Evidence supporting the presence of a DDM-PGE₂ receptor include 1) the induction of cytoprotection by PGE₂ is not observed under pH-restrictive conditions (pH 7.8), suggesting a structural requirement for cytoprotection, thus implicating a receptor; 2) PGE₂ and DDM-PGE₂ induce the binding of nuclear proteins to a -³²P-labeled TRE, suggesting this response is mediated by a common receptor; 3) DDM-PGE₂ does not directly activate PKC in isolated LLC-PK₁ cell homogenates (data not shown), in accordance with a mechanism of receptor-mediated TRE binding activity; and 4) cytoprotection by DDM-PGE₂ occurs at concentrations comparable to those eliciting functional EP receptor-mediated responses, including cytoprotection in cortical neurons (1), sodium and water transport in renal cortical collecting ducts (18), and muscle relaxation in piglet saphenous veins (10).

PGE₂ undergoes -elimination to form A and B series prostaglandins at pH > 7.4 (15, 19). PGA₂, but not PGB₂, PGD₂, PGE₂, or PGF₂, increases the biosynthesis of -glutamylcysteine synthetase, which subsequently results in increased glutathione levels in L-1210 and NIH/3T3 cells (21). Upregulation of antioxidant defense mechanisms would significantly impact quinone-thioether-mediated cytotoxicity, which exhibits a prominent oxidative stress-related component (26, 31). However, under conditions favoring PGE₂ degradation (pH 7.8), cytoprotection is not observed (Fig. 2), consistent with the observation that exogenous PGA₂ does not induce cytoprotection (Fig. 3). Although B-series prostaglandins were not tested in these studies, the fact that PGE₂ degradation is associated with B-series formation, but not cytoprotection, argues against a role for B-series PGs in the cytoprotective response. In addition, the induction of cytoprotection by DDM-PGE₂, an analog which is stable at pH  9.0, is pH insensitive (data not shown), suggesting that changes in pH per se cannot account for the loss of cytoprotection in PGE₂-pretreated cells under pH-restrictive (pH 7.8) conditions. Collectively, these data implicate PGE₂ in the cytoprotective response, and this prostanoid is known to elicit its effects through its interactions with EP receptors (12).

In contrast to PGE₂, EP receptor subtype-specific agonists failed to induce cytoprotection against 2,3,5-(trisglutathion-S-yl)-HQ-mediated cytotoxicity (Fig. 3). However, PGE₁ was also inactive in this system (Fig. 3), questioning a role for EP receptors in the cytoprotective response to PGE₂. PGE₂ and PGE₁ share a cyclopentane ring structure with keto and hydroxyl groups in positions 9 and 11, respectively. PGE₂ and PGE₁ are equipotent at displacing [³H]PGE₂ from EP₂ (5) and EP₃ (35) receptors in ligand binding studies, with PGE₁ slightly less potent using EP₁ receptor preparations (14). The biological activities of PGE₁ and PGE₂ are nearly identical, and both induce cytoprotection against N-methyl-D-aspartate receptor-mediated glutamate cytotoxicity via EP₂ receptors in cultured cortical neurons (1). Thus PGE₁ would be expected to induce cytoprotection if the response is mediated by an EP receptor. Additional evidence supporting the differential effects of PGE₂ and PGE₁ include the observation that TRE binding activity is induced by PGE₂ but not PGE₁ (Fig. 10).

EP receptors have only recently been cloned, and alternative splice variants have been identified, raising the possibility that novel subtypes exist and may mediate the present cytoprotective response (12). EP₁/EP₃ receptors are coupled to IP-related signal transduction, whereas cyclic nucleotide metabolism is regulated positively by EP₂/EP₄ receptors and negatively by EP₃ receptors (17). To facilitate identification of the PGE₂/DDM-PGE₂ receptor, we activated second messenger systems to define pathways capable of inducing cytoprotection. Although many actions of PGs involve cyclic nucleotide-related signaling, cytoprotection is not induced by agents that activate this pathway (DBcAMP, forskolin; data not shown). PG treatment is also associated with elevations in intracellular calcium (17), but pretreatment with a calcium ionophore (4-Br-A-23187) alone was ineffective. In contrast, IP turnover is associated with activation of PKC, and pretreatment of cells with TPA, a potent activator of PKC (28), induces cytoprotection against 2,3,5-(trisglutathion-S-yl)-HQ-mediated cytotoxicity (Fig. 6).

To further dissociate these pathways and provide additional support for protein kinase activity in the cytoprotective response to DDM-PGE₂, the protein kinase inhibitor H-89 was employed. This inhibitor has the advantage of differentially inhibiting PKA and PKC activities as a function of concentration (K_{i PKA}, 48 nM; K_{i PKC}, 31.7 µM) (8). The cytoprotective responses to DDM-PGE₂ and TPA are sensitive to H-89 at concentrations approaching the K_i for PKC but not at concentrations 40-fold higher than the K_i for PKA (Fig. 6), consistent with a PKC-dependent mechanism. H-89 alone did not modulate cell viability, suggesting that inhibition of cytoprotection and TRE binding activity is not secondary to overt cytotoxicity, as observed with the selective PKC inhibitor calphostin C (data not shown). The highest concentration of calphostin C, which was not cytotoxic (10 nM), marginally inhibited the cytoprotective response to DDM-PGE₂ (10-15%, data not shown), as would be predicted (half-maximal inhibitory concentration, 50 nM). The differential cytotoxicity of H-89 and calphostin C in LLC-PK₁ cells suggests the presence of a critical PKC isoform(s) that regulates cell viability and suggest that H-89 may be a useful probe to identify this isoform(s).

PKC regulates the activity of the AP-1 transcriptional complex, which is considered a nuclear third messenger in this cascade. On activation, AP-1 specifically binds to a target DNA sequence referred to as the TRE (23). DDM-PGE₂ induces binding of nuclear proteins to a ³²P-labeled TRE (Figs. 7-10), providing molecular evidence for the involvement of PKC in this response. Furthermore, the induction of TRE binding activity is observed in cells treated with cytoprotective agents (DDM-PGE₂ and PGE₂) but not by prostanoids known to interact with PKC-coupled EP receptors (PT-PGE₂, sulprostone, PGE₁) (Fig. 10). Consistent with a PKC-dependent mechanism, H-89 inhibits inducible TRE binding activity as the concentration approaches the K_i for PKC (Fig. 9).

The mechanism of PG-mediated cytoprotection in vivo may involve modulation of renal hemodynamics (13), direct cellular actions (1), or perhaps a combination of both. The latency of the cytoprotective response to DDM-PGE₂ in LLC-PK₁ cells (Fig. 4) suggests cytoprotection occurs at the cellular level, consistent with reports elsewhere (1, 30, 33). Interestingly, inhibition of DDM-PGE₂-mediated cytoprotection (Fig. 6) and inducible TRE binding activity (Fig. 9) by H-89 were observed, and this correlation raises the possibility that cytoprotection is transcriptionally regulated. This suggestion is consistent with the delayed onset of this effect (Fig. 4) and with the observation that PGE₂ as a cotreatment does not induce cytoprotection (data not shown). Alternatively, the cytoprotective response was first detected 8 h after treatment of cells with DDM-PGE₂ (Fig. 4), a time point where inducible TRE binding activity was returning to control (Fig. 7). This correlation suggests the induction of cytoprotection is secondary to PKC downregulation. Studies are ongoing to delineate the mechanism of PG-mediated cytoprotection and the role of PKC in this response.

In conclusion, evidence is presented suggesting that cytoprotection by PGE₂ and DDM-PGE₂ in LLC-PK₁ cells is receptor mediated. The increased potency of DDM-PGE₂, relative to PGE₂, in the cytoprotective response may be related to the enhanced stability of this analog. Alternatively, modification of PGE₂ to the DDM-PGE₂ analog may increase its affinity for the putative receptor that mediates the cytoprotective response. The latter suggestion is consistent with the general decreased potency of PGE₂ with respect to the induction of cytoprotection and TRE binding activity. The observation that known EP receptor agonists and PGE₁ are inactive in this system suggest the involvement of a receptor unrelated to presently known EP receptor subtypes. The PGE₂/DDM-PGE₂ receptor is coupled to PKC as evidenced by 1) the induction of TRE binding activity by PGE₂ and DDM-PGE₂, which does not result from the direct activation of PKC; 2) the inhibition of DDM-PGE₂-mediated cytoprotection and TRE binding activity by H-89; and 3) the induction of cytoprotection by phorbol ester.