SUMOylation of nonstructural 5A protein regulates hepatitis C virus replication.

Viruses exploit cellular SUMOylation machinery to favour their own propagation. We show that NS5A is a target protein of small ubiquitin-like modifier (SUMO) and is SUMOylated at lysine residue 348. We demonstrated that SUMOylation increased protein stability of NS5A by inhibiting ubiquitylation, and SUMOylation was also required for protein interaction with NS5B. These data imply that SUMO modification may contribute to HCV replication. Indeed, silencing of UBC9 impaired HCV replication in Jc1-infected cells, and HCV replication level was also significantly reduced in SUMO-defective subgenomic replicon cells. Taken together, these data indicate that HCV replication is regulated by SUMO modification of NS5A protein. We provide evidence for the first time that HCV exploits host cellular SUMO modification system to favour its own replication.

Nonstructural 5A protein of hepatitis C virus regulates heat shock protein 72 for its own propagation.

We identified heat shock protein 72 (Hsp72) as a host factor that was differentially expressed in cells expressing nonstructural 5A (NS5A) protein. To investigate how NS5A modulates Hsp72 in hepatitis C virus (HCV) life cycle, we examined the role of Hsp72 in HCV replication and virus production. NS5A specifically interacted with Hsp72. Both Hsp72 and nuclear factor of activated T cells 5 (NFAT5) levels were increased in cells expressing NS5A protein. Treatments of N-acetylcysteine and glutathione markedly reduced protein levels of both NFAT5 and Hsp72. Knockdown of NFAT5 resulted in decrease in Hsp72 level in cells expressing NS5A. Importantly, silencing of Hsp72 expression resulted in decrease in both RNA replication and virus production in HCV-infected cells. These data indicate that NS5A modulates Hsp72 via NFAT5 and reactive oxygen species activation for HCV propagation.

Comparison of gadoxetic acid-enhanced MRI and superparamagnetic iron oxide-enhanced MRI for the detection of hepatocellular carcinoma.

AIM: To compare the diagnostic accuracy and sensitivity of gadoxetic acid-enhanced MRI and superparamagnetic iron oxide (SPIO)-enhanced magnetic resonance imaging (MRI) for the detection of hepatocellular carcinoma (HCC). MATERIALS AND METHODS: Eighty-nine patients (118 HCCs) who underwent three-dimensional gadoxetic acid-enhanced MRI and SPIO-enhanced MRI with a mean interval of 4.7 days (range 3-7 days), were included in this study. Two observers reviewed the gadoxetic acid set (unenhanced, early dynamic, 10 and 20 min hepatocyte-phase images) and SPIO set [unenhanced and ferucarbotran-enhanced T1-, T2-turbo spin-echo (TSE), and T2* weighted imaging (WI)] in consensus. Diagnostic accuracy and sensitivity were evaluated using the alternative-free response receiver operator characteristic (ROC) method. RESULTS: The area under ROC curve (Az value) and sensitivity of the gadoxetic acid set (Az 0.964; sensitivity 90.7%) were significantly higher than those of the SPIO set (Az 0.830; sensitivity 84.7%; p<0.05). There were 14 and seven lesions that were verified only on the gadoxetic acid set and only on the SPIO set, respectively. Four HCCs were clearly revealed as hypointense only on gadoxetic acid-enhanced hepatocyte phase imaging, but were occult on other sequences, including the SPIO set. CONCLUSION: Gadoxetic acid-enhanced MRI is better than SPIO-enhanced MRI for the detection of HCCs.

Monoclonal antibody recognizing N-terminal epitope of hepatitis C virus nonstructural 5B inhibits viral RNA replication.

The nonstructural 5B (NS5B) protein of hepatitis C virus (HCV) is an RNA-dependent RNA polymerase (RdRp) with a key role in HCV replication. To characterize the functional roles of NS5B in HCV replication, we produced a panel of 10 monoclonal antibodies (mAbs) directed against NS5B protein from mice immunized with functionally active RdRp. The epitopes of eight mAbs are localized in the middle region (amino acid 240-263) of NS5B protein. On the other hand, the epitopes of two mAbs are mapped to amino acids 67-88 at the N-terminus of NS5B protein. To examine the effects of mAbs on HCV-RNA replication, we performed in vitro RdRp assay using either the 3'-untranslated region (UTR) or the full-length of HCV-RNA as a template in the presence of each mAb. mAbs specific for the middle region of NS5B had no effect on RdRp activity. Surprisingly, mAb recognizing the N-terminal region of NS5B inhibited RdRp activity in a dose-dependent manner. We have confirmed the same result using the other subclass of mAb, whose epitope is also localized to the same N-terminal region of NS5B. These data show that NS5B contains a B-cell epitope located between amino acid residues 67 and 88. Binding of this epitope with an antibody interferes with the enzymatic function of NS5B.

Acute carbon monoxide poisoning: MR imaging findings with clinical correlation.

PURPOSE: To assess the magnetic resonance imaging (MRI) findings, including diffusion-weighted imaging (DWI) in patients with acute carbon monoxide (CO) poisoning and correlate MRI findings with carboxyhemoglobin levels. MATERIALS AND METHODS: The MRI examinations and medical records of seven men with a mean age of 43±16.0years (SD) (range: 25-63 years) with acute CO poisoning were reviewed. MRI examinations were analyzed with respect to lesion location, imaging presentation on T1- and T2-weighted images, and diffusion characteristics on DWI and apparent diffusion coefficient (ADC) maps. We also evaluated clinical features and laboratory findings including the presenting symptoms and signs, carboxyhemoglobin level, and treatment. RESULTS: All seven patients presented with mental status change. The level of carboxyhemoglobin ranged between 8.3% and 34.8% (normal<1.5%). All seven patients (7/7, 100%) showed restricted diffusion of the lesions on ADC maps and bilateral involvement of globus pallidus. The mean ratios of ADC values was 0.63±0.15 (SD) (range: 0.46-0.92) on bilateral globi pallidi. Cerebral cortex, cerebral white matter, cerebellum, hippocampus, amygdala, splenium of corpus callosum, midbrain and insula were also involved. CONCLUSION: Bilateral globi pallidi with restricted diffusion may be a characteristic MRI feature in patients with acute CO poisoning. However, the relationship was not certain between the carboxyhemoglobin levels and the variety or severity of MRI findings.

Associations between Cerebral Embolism and Carotid Intraplaque Hemorrhage during Protected Carotid Artery Stenting.

BACKGROUND AND PURPOSE: Carotid artery stent placement in patients with intraplaque hemorrhage remains controversial because of the incidence of cerebral embolism after the procedure. The purpose of this study is to determine if intraplaque hemorrhage is a significant risk factor for cerebral embolism during carotid artery stent placement. MATERIALS AND METHODS: This prospective study assessed 94 consecutive patients with severe carotid stenosis. These patients underwent preprocedural carotid MR imaging and postprocedural DWI after carotid artery stent placement. Intraplaque hemorrhage was defined as the presence of high signal intensity within the carotid plaque that was >200% of the signal from the adjacent muscle on MPRAGE. We then analyzed the incidence of postprocedural ipsilateral ischemic events on DWI and primary outcomes within 30 days of carotid artery stent placement. RESULTS: Forty-three patients (45.7%) had intraplaque hemorrhage on an MPRAGE image. There was no significant difference in the incidence of postprocedural ipsilateral ischemic events and primary outcomes between the intraplaque hemorrhage and non-intraplaque hemorrhage group. However, postprocedural ipsilateral ischemic events were more frequently observed in the symptomatic group (17/41 [41.5%]) than in the asymptomatic group (8/53 [15.1%]; P = .005). CONCLUSIONS: Intraplaque hemorrhage was not a significant risk factor for cerebral embolism during carotid artery stent placement in patients with severe carotid stenosis. Symptomatic patients should receive more careful treatment during carotid artery stent placement because of the higher risk of postprocedural ipsilateral ischemic events.

Specific binding of tritium-labeled inositol 1,4,5-trisphosphate to human platelet membranes: ionic and GTP regulation.

Specific, saturable and reversible binding of tritium-labeled inositol 1,4,5-trisphosphate [( 3H]Ins(1,4,5)P3) to human platelet membranes is demonstrated. The Ins(1,4,5)P3-binding sites are abundant and display high selectivity for Ins(1,4,5)P3. Other inositol phosphates exhibit much lower affinity for this site. The specific [3H]Ins(1,4,5)P3 binding was found to be modulated by pH, monovalent and divalent cations, and GTP. A sharp increase in binding occurs at slightly alkaline pH. The monovalent cations, Na+, K+ and Li+ almost double the binding at 30 mM. Mg2+ inhibits the specific [3H]Ins(1,4,5)P3 binding. At low concentrations of Ca2+, the binding is inhibited, but at concentrations higher than 5 mM the binding is potentiated and increases by almost 5-fold at 100 mM. Similar pattern of the effects is also observed for Mn2+ and Sr2+. The specific [3H]Ins(1,4,5)P3 binding is specifically inhibited by GTP. Other nucleotides also inhibit the binding but at higher concentrations. From saturation binding studies, Ca2+ potentiation seems to be due to the conversion of the receptor from the low-affinity state to the high-affinity one. In the absence of Ca2+, the Scatchard plot is nonlinear and concave, and statistically can be fitted best with two equilibrium dissociation constants (Kd values), 0.19 +/- 0.11 and 13.2 +/- 18.1 nM, respectively, for high- and low-affinity binding sites. However, in the presence of 100 mM CaCl2, the Scatchard plot reveals only the high-affinity binding sites with a Kd value of 0.32 +/- 0.15 nM. The specific Ins(1,4,5)P3 receptor in human platelets could therefore exist in multiple conformational states to regulate the intracellular Ca2+ concentration.

Transient photovoltages in purple membrane multilayers. Charge displacement in bacteriorhodopsin and its photointermediates.

The photovoltaic properties of bacteriorhodopsin molecules and their photochemical intermediates have been investigated in an experimental cell consisting of multilayered films of highly oriented, dry fragments of purple membrane and lipid sandwiched between two metal (Pd) electrodes. The electrical time constant of these sandwich cells containing between 5 and 30 layers is less than 10(-5) S. Bright illumination of these cells with actinic flashes of approximately 1 ms duration generates transient photovoltages. These photovoltages, which make the extracellular surface of purple membrane positive with respect to the intracellular surface, follow the time course of the flash with no detectable latency. The amplitude of the photovoltages increases linearly with light intensity and their action spectrum matches the absorption spectrum of the light-adapted state of bacteriorhodopsin, BR570. In these dry multilayer cells, the slow photointermediates of bacteriorhodopsin, M412, N520 and O640 are long lived. Illumination of the sandwich cells with long duration (200 ms) pulses of light results, therefore, in the formation of photomixtures containing all these slow photointermediates. Flash illumination of the sandwich cells immediately following the conditioning pulse produces photovoltages whose action spectra match the absorption spectra of the M412 and N520 photointermediates. The M412 photovoltages, like the BR570 photovoltages, follow the time course of the actinic flash with no detectable latency and increase in amplitude linearly with light intensity. But, unlike the BR570 photovoltage, the M412, N520 and O640 photovoltages make the extracellular surface of purple membrane negative with respect to the intracellular surface. Through the of their specific photovoltaic signals, M412 and N520 are shown to be kinetically distinct photointermediates of bacteriorhodopsin. Detection of fast photovoltages with these characteristics in the absence of any ionic solution, and in parallel with spectrophotometric changes, suggest that they arise from charge displacements in the bacteriorhodopsin molecules and their photointermediates as they undergo photochemical conversion in response to the absorption of photons.

Proton transport by bacteriorhodopsin in planar membranes assembled from air-water interface films.

Bacteriorhodopsin, in known amounts and controlled orientation, is incorporated into planar membrane films. These films are formed by the sequential transfer of two air-water interface films onto a thin, hydrophilic, electrically conductive support cast from nitrocellulose. The films are easily accessible to electrical measurements and to control of the ionic milieu on either side of the membrane. The area of the assembled membrane films can be varied between 2.3 x 10(-2) cm2 and 0.7 cm2. Illumination of these films produces photocurrents, photovoltages, and changes in the pH of the surrounding medium. The peak amplitude of the photocurrent increases linearly with light intensity for dim lights, and it approaches a saturating value for brighter lights. In the linear range, the stoichiometry of transport is 0.65 +/- 0.06 protons/absorbed photon. The rate of transport is linearly proportional to light at all intensities tested. The amplitude and kinetics of the photovoltage measured are accurately predicted by the photocurrent generated and the passive electrical features of the film. Parallel measurements of pH and photocurrent reveal that the light-induced changes in pH are fully accounted for by the rate and amount of charge transport across the membrane. Preceding the transport of protons, a transient photovoltage is detected that exhibits no detectable latency, reaches peak in about 80 microseconds, and probably arises from light-induced intramolecular charge displacements.

Cell cycle arrest mediated by hepatitis delta antigen.

Hepatitis delta antigen (HDAg) is the only viral-encoded protein of the hepatitis delta virus (HDV). This protein has been extensively characterized with respect to its biochemical and functional properties. However, the molecular mechanism responsible for persistent HDV infection is not yet clear. Previously, we reported that overexpression of HDAg protects insect cells from baculovirus-induced cytolysis [Hwang, S.B. Park, K.-J. and Kim, Y.S. (1998) Biochem. Biophys. Res. Commun. 244, 652-658]. Here we report that HDAg mediates cell cycle arrest when overexpressed in recombinant baculovirus-infected insect cells. Flow cytometry analysis has shown that HDAg expression in Spodoptera frugiperda cells causes an accumulation of substantial amounts of polyploid DNA in the absence of cell division. This phenomenon may be partly responsible for the persistent infection of chronic HDV patients.

Membrane effects of antiinflammatory agents. 2. Interaction of nonsteroidal antiinflammatory drugs with liposome and purple membranes.

The interaction of the nonsteroidal antiinflammatory drugs (NSAIDS) indomethacin, diflunisal, and flurbiprofen and the active sulfide metabolite of sulindac with phosphatidylcholine (PC) liposomes was investigated using differential scanning calorimetry (DSC). These biologically active structures decrease the phase transition temperature and broaden the transition peak with increasing concentration, but without affecting the enthalpy change for the transition on the thermal scan. A comparison with the effects of the prodrug sulindac and its inactive sulfone metabolite suggests that the main action of NSAIDS on membranes is a reduction of the cooperative interaction between phospholipid molecules. The probable positions of these compounds in the bilayer are inferred from similar DSC effects of several reference compounds whose mode of binding to the PC bilayer have previously been described. The active antiinflammatory structures appear to insert deeply into the hydrocarbon region of the bilayer, whereas the inactive compounds probably bind mainly to the carbonyl region near the surface. Using purple membrane as a model to study the drug effect on protein--protein interaction in this membrane system, low concentrations of active NSAIDS effectively dissociate the bacteriorhodopsin lattice. These results suggest that the active NSAIDS studied here are able to partition deeply into the hydrocarbon region of the bilayer and interact with a membrane protein imbedded inside the bilayer. The prodrug sulindac per se is devoid of any significant membrane effects.

Conformation and activity of tetrahydrofuran lignans and analogues as specific platelet activating factor antagonists.

The six (racemic or meso) isomers of 3,4-dimethyl-2,5-bis(3,4-dimethoxyphenyl)tetrahydrofuran and four corresponding desmethyl analogues were prepared and assayed as inhibitors of platelet activating factor (PAF) receptor binding to rabbit platelet plasma membranes. The inhibition by these isomers is stereodependent and varies with the gross shape of the molecules as determined by the molecular mechanics program MM2. The most potent PAF antagonist in this group of compounds is trans-2,5-bis(3,4,5-trimethoxyphenyl)tetrahydrofuran (L-652,731, 14) with an IC50 of 0.02 microM.

Structure-activity relationships of kadsurenone analogues.

Kadsurenone, a specific receptor antagonist of platelet-activating factor (PAF), and its analogues were prepared from derivatives of cinnamyl alcohol and (allyloxy)phenol. Racemic kadsurenone, resolvable by a Chiralpak column at low temperatures, has an IC50 value of 2 X 10(-7) M, which is about 50% of the activity of the natural product (IC50 = 1 X 10(-7) M). The structural specificity of kadsurenone was further demonstrated by the low PAF-receptor-blocking activities of denudatin B, mirandin A, desallylkadsurenone, and the 2-epimer of kadsurenone.

(+/-)-trans-2-[3-methoxy-4-(4-chlorophenylthioethoxy)-5-(N-methyl-N- hydroxyureidyl)methylphenyl]-5-(3,4, 5-trimethoxyphenyl)tetrahydrofuran (CMI-392), a potent dual 5-lipoxygenase inhibitor and platelet-activating factor receptor antagonist.

By incorporating an N-hydroxyurea functionality onto diaryltetrahydrofurans, a novel series of compounds was investigated as dual 5-lipoxygenese (5-LO) inhibitor and platelet-activating factor (PAF) receptor antagonist. These dual functional compounds were evaluated in vitro for 5-LO inhibition in RBL cell extracts and human whole blood, and PAF receptor antagonism in a receptor binding assay. PAF-induced hemoconcentration and arachidonic acid- and TPA-induced ear edema in mice were used to determine in vivo activities. The structure-activity relationship analysis to define a preclinical lead is presented. (+/-)-trans-2-[3-methoxy-4-(4-chlorophenylthioethoxy)-5-(N-methyl- N-h ydroxyureidyl)methylphenyl]-5-(3,4, 5-trimethoxyphenyl)tetrahydrofuran (40, CMI-392) was selected for further study. In the arachidonic acid-induced mouse ear edema model, 40 was more potent than either zileuton (a 5-LO inhibitor) or BN 50739 (a PAF receptor antagonist), and it demonstrated the same inhibitory effect as a physical combination of the latter two agents. These results suggest that a single compound which both inhibits leukotriene synthesis and blocks PAF receptor binding may provide therapeutic advantages over single-acting agents. The clinical development of compound 40 is in progress.

Hepatitis C virus core protein potentiates c-Jun N-terminal kinase activation through a signaling complex involving TRADD and TRAF2.

The hepatitis C virus (HCV) core protein is a multifunctional viral nucleocapsid protein. Previously, it has been demonstrated that the HCV core protein interacts with the cytoplasmic domain of tumor necrosis factor receptor 1 (TNFR1). Since the TNFR1 is engaged in stimulation of transcriptional factor NF-kappaB and AP-1 through activation of IkappaB kinase and c-Jun N-terminal kinase (JNK, or stress-activated protein kinase), respectively, we have examined whether the interaction between core protein and TNFR1 can modulate JNK. In this study, we demonstrate that the HCV core protein synergistically activates TNFalpha-induced JNK at a core concentration dependent manner in human embryonic kidney (HEK) 293 cells. HCV core-mediated synergism of JNK activation was also detected in stable cells expressing HCV core protein. Furthermore, we demonstrate that HCV core protein does not compete with TNF receptor-associated death domain (TRADD) for its interaction with the death domain of TNFR1. Our in vivo data show that HCV core and TRADD form a ternary complex with TNFR1. These findings suggest that the HCV core protein modulates TNFR1 signaling and may, thus, play a role in chronic infection of HCV patients.

Species difference in the specific receptors of platelet activating factor.

Relative potencies of platelet activating factor (PAF) and PAF analogs and several PAF receptor antagonists when inhibiting the [3H]PAF specific binding to human and rabbit platelet membranes and membrane fragments of human lung tissues were compared. In rabbit platelets, L-652,731 was found to be most potent in the list of PAF receptor antagonists with an equilibrium inhibition constant (Ki) of 9.83 (+/- 2.92) X 10(-9) M followed by L-653,150 greater than kadsurenone congruent to Ono-6240 greater than ginkgolide B greater than CV-3988 greater than L-651,142, whereas in human platelets the relative potencies of these PAF receptor antagonists were as follows: Ono-6240 greater than L-653,150 congruent to L-652,731 congruent to kadsurenone greater than ginkgolide B greater than CV-3988 greater than L-651,142. Ono-6240 was the most potent one with a Ki of 4.86 (+/- 1.44) X 10(-8) M which was roughly two times more potent than that in rabbit platelets, whereas the affinity of L-652,731 was about ten times less in human platelets (Ki = 1.03 (+/- 0.15) X 10(-7) M) compared to that in rabbit platelets (Ki = 9.83 (+/- 2.92) X 10(-9) M). These variations between species among PAF antagonists strongly suggest that there exists a species difference at or near the binding site of the receptor of platelet activating factor. The relative potency of these PAF receptor antagonists in human lung membranes differed very little from that in human platelets and was found to be Ono-6240 greater than L-653,150 congruent to kadsurenone congruent to L-652,731 greater than ginkgolide B greater than CV-3988 greater than L-651,142. Even though C16-PAF showed slightly higher potency in human lung, and CV-3988 and Ono-6240 showed slightly lower, the difference was too small to suggest that there is a difference in the PAF receptors between human platelets and human lung tissues.

Detection of Cellular Proteins and Viral Core Protein Interacting with the 5' Untranslated Region of Hepatitis C Virus RNA.

Hepatitis C virus (HCV) is a pesti- and flavi-like virus, which contains a highly conserved 5'-untranslated region (UTR). This region is implicated in the regulation of both translation and RNA replication. To examine the possible cellular factors involved in HCV replication, we performed UV cross-linking experiments to detect cellular protein binding to 5'-UTR of HCV RNA. No cytoplasmic proteins were found to cross-link to 5'-UTR. Surprisingly, when nuclear extracts were used for UV cross-linking, a major protein of 110 kD and several other minor proteins were detected. Competition assays confirmed that the binding of the 110-kD protein was specific to the 5'-UTR. The protein-binding site was mapped within the 78-nt region between nucleotides 199 and 277 from the 5' end of the viral RNA. This protein was present in several different cell lines tested. No cellular proteins specifically bound to the complementary strands of the 5'-UTR. We have also shown by an RNA-protein blotting assay that 5'-UTR bound to the HCV core protein, which can be translocated to the nuclei. These findings suggest that HCV RNA may enter nuclei by complexing with the viral core protein and interact with nuclear proteins that are involved in the regulation of RNA replication or translation. It is thus possible that HCV employs a replication strategy distinct from its related pestiviruses or flaviviruses. Copyright 1995 S. Karger AG, Basel

Rac GTPase activity is essential for EGF-induced mitogenesis.

Rac, a member of the Rho family GTPases, has been implicated in the regulation of a wide range of biological processes including actin remodeling, cell transformation, G1 cell cycle progression, and gene expression. To determine whether Rac GTPase activity is required for epidermal growth factor-induced mitogenesis, Rat-2 stable cells expressing a dominant-negative Rac1 mutant, RacN17, were prepared. Exposure to EGF exhibited a significantly restricted growth response in Rat-2-RacN17 cells compared to Rat-2 parental cells, suggesting an essential role of Rac in EGF-induced mitogenesis. In contrast, addition of lysophosphatidic acid exerted the same level of growth in Rat-2 and Rat-2-RacN17 cells. To gain further evidence for the essential role of Rac in EGF-induced mitogenesis, we performed the microinjection experiment. EGF-induced DNA synthesis was significantly blocked by microinjection of recombinant RacN17 protein, and not control IgG. Our further study to analyze the downstream mediator of Rac in EGF-signaling to mitogenesis demonstrated that Rac-activated phospholipase A2 plays a critical role. Taken together, our results suggest that the "Rac and Rac-activated PLA2" cascade is one of the major mitogenic pathways induced by EGF.

High affinity receptor binding of platelet-activating factor in rat peritoneal polymorphonuclear leukocytes.

Rat peritoneal polymorphonuclear leukocytes (PMNs) showed a single class of high affinity binding sites for platelet-activating factor (PAF) with an equilibrium dissociation constant (KD) of 4.74 (+/- 2.59) nM. Each cell contained 2.79 (+/- 1.40) x 10(4) receptors. In the isolated membranes at pH 7.0 in 10 mM MgCl2, 10 mM Tris and 0.25% bovine serum albumin, the KD value was 0.61 (+/- 0.1) nM, which is roughly identical to the KD values reported previously for various membrane systems under identical ionic conditions. The receptors were highly specific for PAF. Several receptor antagonists that are reported to inhibit the binding of [3H]PAF and the PAF-induced function in platelets could fully displace the binding. The biologically inactive enantiomer (enantio-C16-PAF), a PAF analog, azido-PAF, and an indene derivative of the PAF receptor antagonist, L-651,142, had different potencies to inhibit [3H]PAF binding to rat and human PMN membranes. L-652,731, a tetrahydrofuran analog of the PAF receptor antagonist was about 10 times more potent to inhibit the binding in rat liver tissues than in rat PMNs. These results suggest that PAF receptors on human and rat PMNs are not identical and that PAF receptor subtypes may exist in rat liver and PMNs.

Release of platelet activating factor and its involvement in the first phase of carrageenin-induced rat foot edema.

Platelet activating factor (PAF), a potent lipid-like vasoactive agent, induced rat foot edema when it was injected subplantarly. The edema reached its maximum 1 h after PAF challenge. Indomethacin did not inhibit the peak edematous response whereas both PAF antagonists, kadsurenone and L-652,731, inhibit the PAF-induced rat foot edema (PFE). Both PAF antagonists also partially block the first phase of the carrageenin-induced rat foot edema (CFE). Using the inhibition of [3H]PAF receptor binding to prepared rabbit platelet membranes, release of PAF or PAF-like materials in carrageenin-injected rat hindpaw was observed. These results suggest that the released PAF or PAF-like materials together with the released histamine and kinin evoke the first phase hindpaw edema in the rats. Indomethacin or PAF antagonist, administered alone, does not block the first phase or the second phase of CFE, respectively. However, PAF antagonist potentiated the inhibitory effects of indomethacin suggesting that the released PAF may also be involved in the biosynthesis of prostaglandins to initiate the second phase of rat CFE.