Abstract
F2-Isoprostanes are isomers of the prostaglandin PGF2α. At least one compound of this group, 8-epi-PGF2α, exhibits biological activity, and therefore special interest is focused on the mechanism of isoprostane formation: enzyme catalyzed or radical mediated. We analyzed the formation of isoprostanes in vitro and in vivo. In both systems, purified cyclooxygenase isoenzymes and cell models specific for the cyclooxygenase isoenzymes, 8-epi-PGF2α formation could be totally suppressed by cyclooxygenase inhibitors. Indomethacin inhibited concentration-dependent 8-epi-PGF2α formation in platelets stimulated with calcium ionophore, arachidonic acid or thrombin. Nordihydroguaiaretic acid, an antioxidant, blocked isoprostane formation with a similar IC50 value as thromboxane B2 synthesis, pointing toward cyclooxygenase as the primary target of inhibition. Based on the turnover number, cyclooxygenase-2 formed higher levels of 8-epi-PGF2α than cyclooxygenase-1. Endogenous 8-epi-PGF2α production in rat mesangial cells correlated well with the mRNA and protein expression of cyclooxygenase-2 during interleukin-1 induction. However, in contrast to human platelets, which produced different forms of isoprostanes, rat mesangial cells appeared to form only 8-epi-PGF2α. Further, this indicates that mesangial cells may represent a cellular origin for renal 8-epi-PGF2αformation. Next, we analyzed the formation of isoprostanes in humans. A direct correlation was observed between indomethacin treatment and the decrease in 8-epi-PGF2α and isoprostane levels, but compared with other prostanoids the inhibition was less pronounced. In summary, based on the in vitro studies, a clear cyclooxygenase-dependent formation of isoprostanes, especially 8-epi-PGF2α, was observed. However, in vivoadditional formation via cyclooxygenase enzyme-independent mechanisms is likely.
Isoprostanes are recently discovered PG-like products formed from arachidonic acid (Morrow et al., 1990). The isoprostanes consist of stereoisomers and regioisomers of the common PGs and were first reported for PGF2. Later, isoprostanes of the E2 and D2 series were discovered (Morrow et al., 1994). The best characterized isoprostane is 8-epi-PGF2α, which has been found to modulate platelet aggregation (Morrow et al., 1992a); later potent vasoconstriction and bronchoconstriction were described in isolated lungs of rabbit (Banerjee et al., 1992) and rat (Kang et al., 1993). In rats, the intrarenal administration of 8-epi-PGF2α exerted reduction in the glomerular filtration rate and renal plasma flow (Takahashi et al., 1992). Interestingly, these effects can be fully prevented by antagonists of the thromboxane receptor (Banerjeeet al., 1992; Takahashi et al., 1992), indicating an interaction with this receptor subtype. Recently, a separate receptor for 8-epi-PGF2α was postulated (Fukunaga et al., 1993).
In explaining the formation of these compounds in vivo, two different mechanisms are discussed. On the one hand, a free radical-catalyzed peroxidation of arachidonic acid with endocyclization, leading to a PGG2-like compound, was suggested. Further reduction or rearrangement in the PGG2-like compound results in F2- or E2/D2-isoprostane formation (Morrow et al., 1990, 1994). Regarding this mechanism, it has been proposed that detection of isoprostanes might offer a quantitative index for the generation of free radicals and lipid oxidation in vivo (Nourooz Zadeh et al., 1995;Roberts and Morrow, 1994). In accordance with this assumption, elevated levels of isoprostanes have been detected in situations associated with increased free radical generation: during coronary reperfusion and in chronic cigarette smokers (Bachi et al., 1996; Morrowet al., 1995; Takahashi et al., 1992).
On the other hand, a COX-dependent isoprostane formation is discussed (Pratico et al., 1995; Pratico and FitzGerald 1996). The COXs are rate-limiting enzymes in the arachidonic acid cascade, providing the endoperoxide PGH2, which in turn is further metabolized by specific enzymes to PGs, thromboxanes and prostacyclin. Several recent reports revealed the existence of two isoforms, COX-1 and COX-2 (for a review, see Otto and Smith, 1995). The former is thought to be constitutively expressed and involved in so-calling house-keeping functions, such as homeostasis, organ function or hormone modulation, whereas the latter is highly inducible and implicated in the signal cross-talk during pathological situations, such as inflammation.
For further investigations into the mechanism of the formation of isoprostanes, especially 8-epi-PGF2α, we established a sensitive and specific assay based on GC/MS/MS for the detection of these compounds in diverse enzymatic and cellular systems as well as in humans and determined its formation under different conditions.
In all systems used, in vitro and in vivo, we observed a COX-dependent formation of 8-epi-PGF2α. The fact that isoprostane formation in human samples could not be fully prevented by COX inhibitors suggests that further mechanisms are involved in isoprostane formation in vivo.
Methods
Materials.
All chemicals other than those listed below were purchased from Sigma Chemical (Deisenhofen, Germany). Arachidonic acid, COX-1 and COX-2 isoenzymes, purified from sheep tissues, as well as murine COX-2 antibodies were purchased from Cayman Chemicals (Paris, France). According to the supplier, the protein purity of COX-1 is >95%, and that of COX-2 is >70%. IL-1β and CGP 28238 were a kind gift of Dr. C. Vosbeck and Dr. I. Wiesenberg (Ciba, Basel, CH). [3,3,4,4-2H4]PGE2, [3,3,4,4-2H4]6-keto-PGF1α, [3,3,4,4-2H4]TXB2and [3,3,4,4-2H4]PGF2αand their nondeuterated analogs were a kind gift from Dr. Udo Axen (Upjohn, Kalamazoo, MI). [18,18,19,19-2H4]PGD2was a kind gift of Dr. C. Meese (Dr. Margarete Fischer-Bosch Institut, Stuttgart, Germany). 8-Epi-PGF2α was purchased from Cayman Chemicals. [18O2]-8-Epi-PGF2αwas prepared from 8-epi-PGF2α and H2 18O as previously described (Lehmann et al., 1992). Ethyl acetate, H2 18O (96.5%) and chloroform were obtained from Promochem (Wesel, Germany). Buturylcholinesterase and O-methylhydroxylamine hydrochloride were from Sigma. Pentafluorobenzyl bromide was from Lancaster (Mühlheim/Main, Germany). Water, methanol and formic acid were from Merck (Darmstadt, Germany). Hexane and sodium acetate were from Riedel-de Haen (Seelze, Germany). Bis-(trimethylsilyl)trifluoroacetamide was purchased from Macherey & Nagel (Düren, Germany). N,N-Diisopropylethylamine was from Pierce (Oud Beijerland, The Netherlands). Silica gel thin-layer chromatography plates (LK6D, 5 × 20 cm) were from Whatman (Maidstone, UK). Helium (99.996%), methane (99.995%) and argon (99.998%) were from Messer Griesheim (Herborn, Germany).
Subjects and study protocol.
Twelve healthy female volunteers (age, 17–38 years) who were selected on the basis of medical history, physical examination and clinical laboratory screening and eight patients (age, 8–14 years; three girls and five boys) with clinical and laboratory characterization of HPS (Seyberth et al., 1985, 1997) were included in the study.
The study protocol was performed with a titrated range of 0.9 to 7.0 mg/kg indomethacin/day on a longterm treatment for HPS-patients and 1.6 to 2.2 mg/kg/day for 2 days for healthy volunteers, respectively.
Urine was collected on day 2 of indomethacin treatment and 2 days after indomethacin withdrawal for prostanoid determination. At this time, no indomethacin was detectable in the plasma. No other anti-inflammatory or analgetic drugs were taken 2 weeks before the study. The study protocol was approved by the Ethics Committee of the Philipps University of Marburg.
Measurement of COX activity by product analysis.
Purified COX isoenzymes (40 units of COX-2) and (80 units of COX-1) were preincubated in buffer (0.05 M Tris·HCl, pH 7.6, 2 mM phenol, 1 μM hematine and 0.1% Tween 80) for 5 min at 37°C with diclofenac (1 μM) or vehicle (ethanol). The reaction was initiated by the addition of arachidonic acid (25 μM) and stopped by acidification with formic acid to pH 2.5 after an incubation time of 150 sec at 37°C. The formed PGs and isoprostanes were extracted by three volumes of ethyl acetate and analyzed by GC/MS/MS.
Cell experiments.
Human washed platelets were isolated from fresh blood as recently reported (Klein et al., 1994). After preincubation (5 min, 37°C) with inhibitors or vehicle, platelets were stimulated by the addition of arachidonic acid (3 μM), calcium ionophore A 23187 (0.5 μM) or thrombin (1 unit/ml) to study prostanoid synthesis from endogenous pools. Reaction was stopped as described above, and ethyl acetate extracts were examined by GC/MS/MS.
Rat mesangial cells were cultured as described (Klein et al., 1994) and used between passages 12–15. For induction of COX-2, cells were starved for 2 days in culture medium containing 0.5% fetal calf serum after treatment with 1 nM IL-1β and different COX inhibitors for 24 hr. For evaluation of endogenous prostanoid production, supernatants were collected after 24 hr and stored at −80°C until sample preparation. To analyze the metabolism of exogenous arachidonic acid, IL-1-stimulated cells were washed twice with phosphate-buffered saline and stimulated at room temperature with 5 μM arachidonic acid in phosphate-buffered saline for the periods indicated. Formed prostanoids were extracted and further processed for GC/MS/MS analysis as described below.
Western blot analysis and RT-PCR.
Western blot analysis of COX-2 was performed as we recently reported (Klein et al., 1994). Briefly, 1 × 106 rat mesangial cells were pelleted, lysed in phosphate-buffered saline/1% Triton X-100 and separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The proteins were transferred onto a nitrocellulose membrane (Amershan, Braunschweig, Germany) and visualized with a 5% Ponceau S solution. After destaining, the membrane was blocked with 5% milk powder, and a COX-2-selective antibody (Cayman Chemicals) was applied in a dilution of 1:100. For visualization, a peroxidase-labeled goat anti rabbit antibody (1:7500; Dianova, Hamburg, FRG) was added that was detected by the enhanced chemiluminescence method after several times of intensive washing. RT-PCR analysis was performed as described (Nüsing et al., 1996).
Sample preparation of in vitro assays for prostanoids by GC/MS/MS-analysis.
Samples were prepared as recently described (Schweer et al., 1994) with minor modifications. Briefly, sample extracts were spiked with ∼10 ng of deuterated internal standards, and solvent was removed. The methoxime was obtained through reaction with an O-methylhydroxylamine hydrochloride-acetate buffer. After acidification to pH 3.5, prostanoid derivatives were extracted, and the pentafluorobenzylesters were formed. Samples were purified by thin-layer chromatography, and two broad zones withR v 0.03 to 0.39 and 0.4 to 0.8 were eluted. After withdrawal of the organic layers, trimethylsilyl ethers were prepared by reaction with bis(trimethylsilyl)-trifluoroacetamide and thereafter subjected to GC/MS/MS analysis.
Sample preparation of urine for GC/MS/MS analysis.
To determine the concentration of 8-epi-PGF2αin urine samples, further steps were necessary. 8-Epi-PGF2α was gained by solid-phase extraction, derivatized to the corresponding pentafluorobenzyl esters and purified by thin-layer chromatography. After elution form the silica, samples were further purified by high-performance liquid chromatography (Schweer et al., 1997) before formation of the trimethylsilylether derivatives and GC/MS/MS analysis.
GC/MS/MS analysis.
A Finnigan MAT TSQ700 GC/MS/MS equipped with a Varian 3400 gas chromatograph and a CTC A200S autosampler was used. Gas chromatography of prostanoid derivatives was carried out on a (J & W) DB-1 (20 m, 0.25 mm i.d., 0.25-μm film thickness) capillary column (Analyt, Mühlheim, Germany) in the splitless mode. GC/MS/MS parameters were exactly as described by Schweer et al. (1994). Products gained from zone 1 were analyzed for PGE2, TXB2, 6-keto-PGF1α, PGF2α, 8-epi-PGF2α and isoprostanes; products from zone 2 were analyzed for PGD2. For quantification of PGF2α, 8-epi-PGF2α and isoprostanes, [P-3(CH3)3SiOH] daughter ions (m/z 299) were used.
Statistical analysis.
Comparisons between groups were made by Student’s t test for paired or unpaired data as appropriate. Values of P < .05 were taken as significant.
Results
PG and 8-epi-PGF2α formation by purified COX isoenzymes.
Opinions differ on whether isoprostane generation occurs via COX-dependent or -independent pathways. To study the possible formation via the enzyme COX, we first used purified COX-1 and COX-2 proteins. As a function of time, PGE2 and PGD2, the decay products of PGH2, reached a plateau after ∼1000 sec in COX-2 enzyme samples incubated with arachidonic acid (fig. 1). Surprisingly, PGF2α and 8-epi-PGF2α formation did not follow with similar kinetics. These two products reached a maximum at ∼150 sec, followed by a decrease to initial quantities after ∼3500 sec. An almost identical kinetic behavior was observed in COX-1 assays (data not shown), albeit the product level was lower, with maximal values of 14 ng of PGF2α and 700 pg of 8-epi-PGF2α compared with 85 and 4 ng, respectively, for COX-2. For both enzymes, COX-1 and COX-2, a similar ratio of PGF2α to 8-epi-PGF2α (18:1) was found. No other isoprostanes except 8-epi-PGF2α were formed. Preincubation with diclofenac, a potent inhibitor for both isoenzymes, totally blocked the formation of PGs, including 8-epi-PGF2α (table1).
Prostanoid and isoprostane formation in human washed platelets and rat mesangial cells.
As a next approach to analyze isoprostane formation by the COX enzymes, we used cellular models with human washed platelets as COX-1 system and IL-1-stimulated rat mesangial cells as COX-2 system. Human washed platelets revealed time-dependent product formation after stimulation with arachidonic acid or calcium ionophore A23187. The synthesis of main metabolite TXB2 as well as isoprostanes and PGE2 showed kinetics of saturation, reaching maximum values after 5 min, followed by a moderate decrease in the following 5 min (data not shown). Values at maximal point are shown in table 2. Furthermore, the addition of a physiological stimulus, such as thrombin, caused a similar formation of 8-epi-PGF2α. However, in contrast to purified COX enzymes, platelets produced an abundant fraction of isoprostanes (eluting profile shown in fig.2). Interestingly, thrombin forced a much larger synthesis of isoprostanes than A23187 or arachidonic acid (table2).
Next, we investigated the effects of the COX inhibitors indomethacin and diclofenac as well as the dual lipoxygenase/COX-inhibitor NDGA on prostanoid and isoprostane formation in platelets (fig.3). Both, indomethacin and diclofenac almost completely suppressed the formation of TXB2, PGE2 and 8-epi-PGF2α, whereas less inhibition was observed within the isoprostane fraction. In comparison to that, NDGA was a more effective inhibitor of the isoprostanes than of the prostanoids and 8-epi-PGF2α. Because NDGA is an antioxidant, we determined the IC50 values for isopostane inhibition in comparison with indomethacin, thereby discriminating the enzyme inhibitory potency from antioxidant potency. For the inhibition of TXB2 and isoprostane formation by indomethacin, we obtained a half-maximal inhibition of 2.8 and 500 nM, respectively, and the IC50 values of NDGA were 5 μM for TXB2 and 10 μM for isoprostanes. The similarity between TXB2 and isoprostane suppression by NDGA points toward COX activity as the primary target of inhibitor action.
Similar results in product formation and inhibition were obtained by using the calcium ionophor A 23187 or thrombin to stimulate the metabolism of arachidonic acid endogenously (data not shown).
Incubation of rat mesangial cells with IL-1 resulted in a time-dependent increase in prostanoid formation (fig.4) with 6-keto-PGF1α as main metabolite. Also, 8-epi-PGF2α release increased and reached 15% of PGF2α formation.
To assess isoprostane formation dependent on exogenous substrate, cells were treated with IL-1 for 20 hr after stimulation with 3 μM arachidonic acid (fig. 5). A similar kinetic of time-dependent product formation was observed for all prostanoids, including 8-epi-PGF2α. Compared with PGF2α 12% 8-epi-PGF2α was formed. As illustrated in figure 6, the COX inhibitors indomethacin and diclofenac completely blocked endogenous PG formation, but 8-epi-PGF2α formation was blocked slightly less efficiently. 8-Epi-PGF2α appeared to be the only isoprostane formed by mesangial cells (data not shown).
To evaluate whether IL-1-induced production of these compounds was dependent on de novo synthesis of COX-2, we investigated its expression. Western blot (fig. 7A) and RT-PCR analysis (fig. 7B) of cell lysates confirmed the induction of transcription and translation of COX-2 after IL-1 treatment. An important finding was that dexamethasone, which is known to suppressde novo synthesis of COX-2, inhibited the formation of the PGs as well as 8-epi-PGF2α at similar extent (fig. 6). Confirmation that product formation was dependent on COX-2 activity was given by the complete suppression of PG and 8-epi-PGF2α formation by CGP 28238, a selective COX-2 inhibitor (Klein et al., 1994, 1996) (table3). In all cell experiments, 8-epi-PGF2α was the only isoprostane formed (data not shown).
Influence of the COX-inhibitor indomethacin on in vivoprostanoid and isoprostane formation.
We investigated the effect of the COX inhibition on human prostanoid and isoprostane formation in two different collectives. In group 1, healthy volunteers were treated with indomethacin for 2 days, and urine samples were collected and analyzed for COX products. After indomethacin withdrawal for 72 hr, urinary prostanoid levels were analyzed again and taken as control values. Comparison between the urine samples taken before and after indomethacin withdrawal showed that indomethacin significantly suppressed the synthesis of TXB2 and of its metabolization products 2,3-dinor-TXB2 and 11-dehydro-TXB2 (fig.8A). Although lowered, prostanoid and 8-epi-PGF2α formation were significantly different, but isoprostane formation was not (P = .09).
Consequently, a second group with HPS (Seyberth et al., 1985) was selected. HPS is a complex congenital tubular disorder associated with excessive renal and systemic PGE2biosynthesis (Seyberth et al., 1997). In this collective, long-term treatment with indomethacin improves the prognosis of HPS by suppressing PGE-M and PGE2 levels. To investigate the inhibitory effect of indomethacin on prostanoid and isoprostane formation, long-term medication with indomethacin was interrupted for 72 hr, and urinary excretion rates of prostanoids and isoprostanes were determined before and at the end of inhibitor withdrawal, according to the study protocol. During this time interval, PGE2 and indomethacin concentrations were monitored to control PG biosynthesis. In contrast to the results gained in the control group, we observed a significant isoprostane inhibition (P = .0015 for 8-epi-PGF2α and P = .029 for isoprostanes) by indomethacin in HPS patients (fig. 8B). Conspicuously, the inhibitory profile of indomethacin was not consistent for all metabolites investigated. The mean of product suppression ranged from 93% to 83% for TXB2 and its metabolites, from 87% to 84% for PGE2 and PGE-M and from 65% to 54% for PGF-M, PGF2α and 8-epi-PGF2α. The lowest inhibitory effect of 40% was obtained within the isoprostane fraction.
Discussion
In this study, we investigated 8-epi-PGF2α and isoprostane formation in diverse in vitro models as well as in humans by means of a selective and sensitive GC/MS/MS technique. The formation of these PGs is discussed controversely in the literature (Morrow et al., 1990, 1993, 1994; Pratico et al., 1995). Based on studies concerning the formation of 8-epi-PGF2α in rats fed with CCL4 (Morrow et al., 1992b), in plasma and in low-density lipoprotein exposed to oxidative stress in vitro (Lynch et al., 1994) or in urine samples of smokers (Bachi et al., 1996; Morrow et al., 1995), free radical-induced generation was suggested. Therefore, isoprostanes were assumed to be an ideal marker for the pathophysiology of oxidative injury. However, because COX-dependent formation of 8-epi-PGF2α in thrombin or collagen activated platelets has been reported (Pratico et al., 1995), the function as a parameter for lipid peroxidation has to be reconsidered. According to our data, isolated COX proteins as well as COX-specific cellular systems are able to synthesize 8-epi-PGF2α. Although the synthetic capacity occurs on a lower level compared with the other prostanoids formed, this generation is definitely enzyme dependent. An interesting fact is that based on specific activity, COX-2 synthesizes ∼10-fold more 8-epi-PGF2α than COX-1. In accordance, similar observations were made recently using LPS-stimulated monocytes (Pratico and FitzGerald, 1996). F2-Isoprostanes are thought to be chemically stable end-products and therefore superior to rapidly decomposing lipid hydroperoxides as analytical marker for lipid peroxidation. Monitoring of 8-epi-PGF2α formation by purified COX protein revealed that most likely 8-epi-PGF2α is further metabolized, thereby escaping continuous monitoring, at least in the cell-free system. However, this must be reconsidered if further metabolism of 8-epi-PGF2α does exist in vivo.
8-Epi-PGF2α is reported to be a potent renal vasoconstrictor, but the cellular origin is unknown. According to our data, the mesangial cells are good candidates for the formation of 8-epi-PGF2α; moreover, the increase in COX-2 expression, known to occur under pathological conditions, was correlated with an increase in 8-epi-PGF2α. In pathological situations, increased arachidonic acid availability was reported, and therefore the increase in 8-epi-PGF2α could be correlated with nephrotoxicity. Intrarenal artery infusion of 8-epi-PGF2α into euvolemic rats significantly reduced single-nephron glomerular filtration and intraglomerular arteriolar resistance (Takahashi et al., 1992). Thus, in patients with inflammatory renal injury, COX-dependent formation of 8-epi-PGF2α could further impair renal blood flow and glomerular filtration.
The comparably small amounts of 8-epi-PGF2α in the assays of the purified COX proteins suggest the synthesis as a by-product during catalysis, as earlier demonstrated for several products of [14C]arachidonic acid-stimulated microsomes from sheep seminal vesicels (Hecker et al., 1987). A plausible explanation would be that radicals formed during the metabolism of PGG2 to PGH2(Kuehl et al., 1977) are responsible for the formation of 8-epi-PGF2α. Accordingly, NDGA, which is known to act as an antioxidant, exerts similar inhibitory potency on isoprostane formation compared with inhibition on TXB2 formation, indicating the involvement of COX activity. However, assuming this mechanism, other isoprostanes should be expected, because a free radical-catalyzed mechanism is unlikely to result in only one specific product. Therefore, the observation of 8-epi-PGF2α as the only isoprostane formed by isolated proteins and rat mesangial cells points against the radical theory and to a more complex mechanism, including enzyme-specific activity. In this context, the decreasing levels of PGF2α and 8-epi-PGF2α after several minutes to basal levels in the protein assays could be a sign of radical cooxidation of these compounds, as has been reported for different xenobiotics as well as for PGs (Eling et al., 1990). In the cell system, this cooxidation may be blocked by reducing agents, such as proteins or glutathione.
In human urinary samples, the generation of 8-epi-PGF2α is not as obvious as in the protein or cell experiments. First, in contrast to the enzymatic formation described, isoprostanes appear as a large fraction in the GC/MS/MS-chromatogram (data not shown), indicating a great variety of compounds. Second, 8-epi-PGF2α levels are in a similar range as PGE2 or PGF2α. Moreover, isoprostanes are the most abundant metabolites detected. This is in clear contrast to the sythesizing ability of the COX proteins and of the used cell systems.
In the control group, indomethacin suppressed mainly the thromboxane metabolites but also significantly 8-epi-PGF2α formation. This is consistent with the higher susceptibility of the systemic prostanoid synthesis toward COX inhibition. In agreement, similar decrease in urinary excretion of 11-dehydro-TXB2 by aspirin in healthy volunteers was reported (Wang et al., 1995). However, although a suppression of serum levels of 8-epi-PGF2α and TXB2was detectable, no decrease in 8-epi-PGF2αlevels was observed. The authors concluded that there was COX-independent 8-epi-PGF2α formation in the kidney. However, in this study, other PGs were not determined; therefore, it is unknown whether COX activity was totally abolished. We observed in the HPS group with long-term indomethacin treatment that all urinary prostanoid products were significantly suppressed, albeit by quite different extents: the inhibitory potency of indomethacin on isoprostanes and 8-epi-PGF2α is reduced in respect to regularly formed prostanoids. These facts point toward the evidence of further pathways in addition to an enzymatic generation.
In conclusion, we assume that COXs are able to synthesize 8-epi-PGF2α but that in vivo, further, most likely radical-based, mechanisms do exist. However, the combined action of these mechanisms may exclude isoprostanes and especially 8-epi-PGF2α as an exclusive parameter for oxidative injuries. Therefore, it might be necessary to determine under such conditions in addition to 8-epi-PGF2α the classical PGs to exclude any contribution by the COX pathway.
Acknowledgments
We are grateful to Dr. J. Pfeilschifter for providing rat mesangial cells and to B. Watzer and D. Cegledi for technical assistance.
Footnotes
-
Send reprint requests to: Dr. Rolf M. Nüsing, Department of Pediatrics, Philipps University of Marburg, 35033 Marburg, Germany. E-mail:nuesing{at}mailer.uni-marburg.de
-
↵1 This work was supported by grants from the Deutsche Forschungsgemeinschaft (Nu 73/2–1, Se 263/11–1).
- Abbreviations:
- PG
- prostaglandin
- COX
- cyclooxygenase
- NSAID
- nonsteroidal anti-inflammatory drug
- GC/MS/MS
- gas chromatography-tandem mass spectrometry
- RMC
- rat mesangial cells
- IL-1
- interleukin-1β
- TXB
- thromboxane
- HPS
- hyperprostaglandin syndrome
- RT
- reverse transcription
- PCR
- polymerase chain reaction
- Received October 28, 1996.
- Accepted May 5, 1997.
- The American Society for Pharmacology and Experimental Therapeutics