Shikonin inhibits IgE-mediated histamine release by human basophils and Syk kinase activity.

OBJECTIVE: Shikonin, a component of the herbal medicine "Shikon", is known to suppress inflammatory reactions, but its molecular targets are not identified. This study examines the effect of shikonin on human basophil degranulation response and aims to identify its targets. MATERIALS: Human basophils in isolated leukocytes from healthy volunteers' peripheral blood; recombinant human Syk and Lyn tyrosine kinases. METHODS: Histamine release from basophils stimulated with anti-IgE antibody was analyzed fluorimetrically. Syk and Lyn kinase activities were tested in Vitro with recombinant proteins and analyzed by off-chip mobility shift assay. RESULTS: Shikonin dose-dependently inhibited the histamine release from basophils induced by anti-IgE antibody (IC50 = 2.6 +/- 1.0 microM; mean +/- SEM). A search for the target(s) of shikonin in the signal cascade of IgE-mediated activation showed that it strongly inhibits Syk (IC50 = 7.8 microM, in the recombinant kinase assay), which plays a pivotal role in the degranulation response. A less significant inhibition was found for Lyn, which phosphorylates FcepsilonRI-betagamma subunits and also Syk. CONCLUSIONS: These results indicate that the inhibition of Syk-dependent phosphorylation events might underlie the blocked histamine release from human basophils, thus contributing to the anti-inflammatory effects of shikonin.

Fungal gliotoxin targets the onset of superoxide-generating NADPH oxidase of human neutrophils.

Gliotoxin from Aspergillus, bearing a S&bond;S bond in its structure, prevented the onset of O(-)(2) generation by the human neutrophil NADPH oxidase in response to phorbol myristate acetate (PMA). Gliotoxin affected the activation process harder than the activated oxidase, as shown by its stronger inhibition when added to neutrophils prior to, than post-PMA at maximum enzyme turnover. Decreased O(-)(2) generation persisted even if cells treated with gliotoxin were subsequently washed, with half-inhibition concentrations (IC(50)) of 5.3, and 3.5 microM for treatments of 15 and 30 min, respectively. In addition, gliotoxin made neutrophils reduce cytochrome c regardless of absence of PMA, through its reaction with intracellular reductants in an oxygen-dependent process, named redox cycling. Thus, we next tested whether preincubation of neutrophils with gliotoxin under hypoxic conditions would relieve the inhibition of NADPH oxidase. Instead, this prevention of redox cycling significantly favored damage to the NADPH oxidase with an IC(50) of 0.009 microM. Moreover, conversion of gliotoxin to its dithiol derivative by addition of reduced dithiothreitol during incubation protected cells from losing oxidase activity. These findings support that the disulfide form of gliotoxin targets NADPH oxidase activation.

Mutation at histidine 338 of gp91(phox) depletes FAD and affects expression of cytochrome b558 of the human NADPH oxidase.

Defective NADPH oxidase components prevent superoxide (O-2) generation, causing chronic granulomatous disease (CGD). X-linked CGD patients have mutations in the gene encoding the gp91(phox) subunit of cytochrome b558 and usually lack gp91(phox) protein completely (X91(0)). gp91(phox) is considered to be a flavocytochrome that contains binding sites for NADPH, FAD, as well as heme. We here report a rare X-linked CGD patient whose neutrophils entirely failed to produce O-2, but presented a diminished expression of gp91(phox) containing about one-third of the heme present in normal individuals by Soret absorption. Translocation of cytosolic factors p67(phox) and p47(phox) was normal. However, the FAD content in his neutrophil membranes was as low as that of X91(0) patients, suggesting complete depletion of FAD in his gp91(phox). This was in agreement with the finding that a single base substitution (C1024 to T) changed His-338 to Tyr in gp91(phox) in a predicted FAD-binding domain of the flavocytochrome model. The loss of FAD could not be corrected even after addition of reagent FAD or a FAD-rich dehydrogenase fraction isolated from normal neutrophils to the patient's membranes, in a reconstitution in vitro with normal cytosol. These results indicate that His-338 is a very critical residue for FAD incorporation into the NADPH oxidase system. This is the first such mutation found in CGD.

Involvement of p40phox in activation of phagocyte NADPH oxidase through association of its carboxyl-terminal, but not its amino-terminal, with p67phox.

Phagocyte NADPH oxidase, dormant in resting cells, is activated upon cell stimulation to produce superoxide anion, a precursor of microbicidal oxidants. Active NADPH oxidase is found on the membrane as an enzyme complex, composed of membrane-integrated cytochrome b558 (gp91phox and p22phox subunits) and two cytosolic factors (p47phox and p67phox), each of the latter containing two src homology 3 (SH3) domains. Recently, we radioactively identified a third cytosolic factor, p40phox, as a molecule that associates with p67phox in human neutrophils. Although it has been found that this p40phox protein is defective in patients with chronic granulomatous disease (CGD) who lack p67phox, evidence to functionally relate it to the NADPH oxidase system has hitherto been lacking. In this study, we raised separate antibodies against both the COOH- and NH2-terminal polypeptides of p40phox as well as against the COOH-terminal polypeptide of p67phox to examine the mode of interaction between p40phox and p67phox in a complex. The antibody against the COOH terminus of p67phox was able to communoprecipitate p40phox in conjunction with p67phox itself as was expected. Very interestingly, however, the antibody against the COOH terminus of p40phox completely dissociated the p67phox molecule from the p40phox-p67phox complex unit without any detectable coimmunoprecipitation of p67phox, despite their tight association, whereas that against the NH2 terminus of p40phox had absolutely no dissociation effect. Similar results were found regarding their effects on the O2-generating ability of cytosol in a cell-free activation system, i.e., inhibition was noted with the COOH terminus antibody but not with that for the NH2 terminus of p40phox. However, this dissociation did not affect the translocation of the cytosolic components including p47phox to the membrane. Once the NADPH oxidase was activated, the antibody for the COOH terminus did not show any inhibitory effect on catalysis by the activated enzyme. The stimulators of NADPH oxidase, MA and SDS, did not dissociate the p40phox-p67phox complex. These results provide the first demonstration that p40phox is practically involved in the activation of NADPH oxidase through the association of its COOH-terminal, but not its NH2-terminal, with p67phox.

Electron spin resonance studies on neutrophil cytochrome b558. Evidence that low-spin heme iron is essential for O2-. generating activity.

Cytochrome b558 purified from pig neutrophils was studied to characterize the spin state of the heme iron in relation to its O2-. generating activity. ESR spectra of cytochrome b558 either from resting or stimulated neutrophils showed a low-spin hemoprotein with g1,2,3 of 3.2, 2.1, and 1.3 (estimated). At physiological pH, the oxidized cytochrome b558 is in a purely low-spin state. On lowering or raising pH from 7, the spin state changes to high-spin. The ESR spectrum of high-spin cytochrome b558 was identical to that of methemoglobin, suggesting that the axial-ligand type in both hemoproteins may be the same, i.e. histidine is the fifth ligand. The ratio of the low-spin to high-spin heme in cytochrome b558 was evaluated by magnetic circular dichroism spectroscopy. The pH of cytochrome b558 was varied to form different ratios of the low-spin to high-spin states of the heme, and its O2-. generating activity was examined in cell-free systems. O2-. forming activity decreased concomitant with loss of the low-spin heme, which provides direct evidence that the low-spin state of cytochrome b558 is essential to generate O2-. and the heme retains the low-spin state through the redox cycle.

Phosphatidylcholine peroxidation and liver cancer in mice fed a choline-deficient diet with ethionine.

A high incidence of hepatocellular carcinoma (HCC) was observed in mice fed a choline-deficient diet containing 0.1% ethionine (CDE) for 19 months. HCC was present in 85% of CDE mice and in 22% of choline-deficient (CD) mice not receiving ethionine. This strong hepatocarcinogenicity of the CDE diet was concomitant with a severe decrease in plasma and liver alpha-tocopherol (Toc) to 60 and 35%, respectively, of those contained in choline-supplemented (CS) control mice. We previously found that this dietary-induced HCC was preceded at 4-week feeding by a depletion of Toc and a remarkable increase of phosphatidylcholine hydroperoxide (PCOOH) in the livers of CDE mice. When HCC was prominent in CDE mice, PCOOH was still elevated. Mouse glutathione S-transferase (GST) M II isozyme, which is related to rat GST-P form, a positive marker for rat hepatic preneoplastic and neoplastic lesions, revealed an inverse histochemical pattern as that seen in rats (i.e., the HCC lesions tended to decreased staining). The aforementioned results taken together indicate that decreases in Toc and enhanced PC peroxidation are important events in CDE-induced mice liver tumors.

Reconstitution of superoxide-forming NADPH oxidase activity with cytochrome b558 purified from porcine neutrophils. Requirement of a membrane-bound flavin enzyme for reconstitution of activity.

Cytochrome b558 of pig blood neutrophils was purified from the membranes of resting cells to examine its ability to reconstitute superoxide (O2-)-forming NADPH oxidase activity in a cell-free assay system containing cytosol and fatty acid. The membrane-associated cytochrome b558 was solubilized with a detergent, n-heptyl beta-thioglucoside, and purified by DEAE-Sepharose, heparin-Sepharose, and Mono Q column chromatography. The final preparation of cytochrome containing 11.5 nmol of protoheme/mg of protein gave bands of the large and small subunits on immunoblotted gel. The cell-free system with the purified cytochrome alone as a membrane component showed little O2(-)-generating activity in the absence of exogenous FAD. However, the system showed high O2(-)-generating activity of 31.8 mol/s/mol of cytochrome b558 (52.5% of the original O2(-)-generating activity of the solubilized membranes) in the presence of a nitro blue tetrazolium (NBT) reductase fraction that was separated from the cytochrome b fraction by heparin-Sepharose chromatography. Heat treatment of the NBT reductase fraction resulted in loss of the O2(-)-generating activity in the reconstituted system. The O2(-)-forming activity of the reconstituted system was markedly decreased by removal of FAD from the NBT reductase fraction and was restored by readdition of FAD to the FAD-depleted reductase. The reconstituted system containing purified cytochrome b558 plus the NBT reductase showed approximately 100 times higher O2(-)-generating activity than a system containing rabbit liver NADPH-cytochrome P-450 reductase instead. These results suggest that both the FAD-dependent NBT reductase and cytochrome b558 are required as membrane redox components for O2(-)-forming NADPH oxidase activity. The present data are discussed in comparison with previously reported results on reconstituted systems containing added free FAD.

Determination of flavin contents in neutrophils by a sensitive chemiluminescence assay: evidence for no translocation of flavoproteins from the cytosol to the membrane upon cell stimulation.

A sensitive and specific chemiluminescence (CL) method with bacterial luciferase was adapted for accurate measurement of the flavins FAD and FMN in the membrane and cytosolic fractions of neutrophils prepared from pig and human blood. The FAD and FMN contents (FAD/FMN = 100:2) in the membranes were essentially the same in resting (R) and myristate-stimulated (S) cells, although O2(-)-generation was markedly enhanced exclusively in S membranes. The O2(-)-forming activity of S samples remained unchanged or even increased after washing the membranes with buffer, although one-third of the FAD was lost during washing (a decrease from 140 to 95 pmol/10(8) cell-equivalent (CE) during washing). The cytosol is known to contain at least three components that are essential for O2- production (p47-phox, p67-phox, and a G-protein), and that are translocated to membranes upon activation, but its flavin content was one tenth of that of the membranes. The cytosol was treated with fatty acids in the absence of membranes to induce substantial precipitation of p47-phox, p67-phox and a protein of 32 kDa. No difference relative to a solvent-control was noted in the low flavin content of the precipitate indicating that these cytosolic components are not flavoproteins. These results do not support the possibility of translocation of a cytosolic flavoprotein to the membrane upon activation of the respiratory burst.

Liver phosphatidylcholine hydroperoxidation provoked by ethionine-containing choline-deficient diet in mice.

It is shown that peroxidation of phosphatidylcholine (PC) is enhanced in liver of mice fed a hepatocarcinogenic choline-deficient diet containing 0.1% w/w ethionine. Mice were divided into 4 groups and fed for 4 weeks one of the following diets: choline-supplemented; choline-supplemented containing ethionine; choline-deficient; and choline-deficient containing ethionine. Phosphatidylcholine hydroperoxide (PCOOH) of liver lipids was measured by high performance liquid chromatography using a chemiluminescence detector. Mice fed a choline-deficient diet containing ethionine showed 6-fold higher PCOOH levels than the choline-supplemented control mice: the PCOOH/PC molar ratios of liver lipids were 32.3 X 10(-5) and 5.6 X 10(-5), respectively. In addition to this remarkable degree of lipid peroxidation in liver of mice fed the choline-deficient diet containing ethionine, we also observed a significant liver fatty infiltration, a decrease in plasma and liver alpha-tocopherol, and an increase in liver injury-indicative enzyme activities. Also, marker enzymes for hepatocarcinogenesis, glucose-6-phosphatase and gamma-glutamyl transpeptidase were affected. These data suggest that enhanced hydroperoxidation of phosphatidylcholine may participate in hepatocarcinogenesis provoked by choline deficiency in the presence of ethionine.

Spontaneous and luminol-dependent chemiluminescences from tissue preparations of benzo[a]pyrene-injected mice.

Chemiluminescence (CL) from tissue preparations of mice i.p. injected with and without the chemical carcinogen benzo[a]pyrene (BP) was detected by a single photon counting apparatus. The spontaneously emitted CL (spontaneous CL) and the CL after luminol addition (luminol-CL) were measured for mice with and without previous induction of their liver mixed-function oxidases (MFO) by phenobarbital. In MFO-non-induced mice, although the spontaneous CL was not notably modified by BP injection, the kidneys presented three times greater luminol-CL after BP injection. On the other hand, MFO-induced mice had higher spontaneous CL of plasma, liver, kidneys, and lungs, as well as higher luminol-CL of liver after BP injection, when compared with the respective MFO-induced but BP-non-injected mice. The luminol-CL of liver was suppressed by scavengers of active oxygen and free-radicals such as TIRON, butylhydroxytoluene and in a less extent by superoxide dismutase. The CL detected from tissue preparations of mice after BP treatment is thought to reflect the formation of oxygen radicals and electronically excited-species during BP metabolism.