N-FORMYL-MET-LEU-PHE-INDUCED OXIDATIVE BURST IN DMSO-DIFFERENTIATED HL-60 CELLS REQUIRES ACTIVE HSP 90 , BUT NOT INTACT MICROTUBULES

In this study we examined whether microtubules and heat shock protein 90 (Hsp90) are involved in phorbol myristate acetate (PMA) and N-formyl-Met-Leu-Phe (fMLP)-induced oxidative burst in DMSO-differentiated HL-60 cells. Our results showed that microtubule interfering agents, paclitaxel (1–5 μM), colchicine (1–100 μM), nocodazole (1–20 μM), and vincristine (1–50 μM), did not affect either PMA or fMLP-induced oxidative burst. In contrast, radicicol, an inhibitor of Hsp90, inhibited fMLP-induced oxidative burst in time and concentration-dependent manner where IC50 value for 30 min pre-incubation was 16.5 ± 3.5 μM radicicol. We conclude that both PMA and fMLP-induced oxidative burst in DMSO-differentiated HL-60 cells is microtubule-independent while the latter requires Hsp90 activity.


INTRODUCTION
Superoxide radical generated during oxidative burst by phagocyte NADPH oxidase (EC 1.6.99.6) is a precursor of other reactive oxygen spe cies which are used for host defence against bacterial and fungal pathogens 1 .During the oxidase activation, cytosolic components of the oxidase (p67 PHOX , p47 PHOX , p40 PHOX , and small G-proteins Rac and Rap1A) are induced to migrate to the cell membrane and form an active enzyme complex with membrane-bound flavocytochrome b 558 consisting of gp91 PHOX and p22 PHOX (ref. 2 ).NADPH oxidase may be activated by various stimuli triggering the oxidative burst either via binding to receptor (e.g.Fc receptor, complement receptor or chemokine receptor) or by direct interaction with a component of a signalling cascade 3 .Although cytoskeleton plays a role in regulation of the NADPH oxidase assembly, the influence of the different states of the cytoskeletal proteins on the access of cytosolic components to flavocytochrome is unknown.In the resting cells, p47 PHOX and p67 PHOX are associated with the cytoskeleton, particularly with moesin, coronin and actin 4 , which was shown to interact also with recombinant Rac1 and Rac2.Furthermore, actin enhances activation, while actin depolymerizing agents facilitate deactivation of NADPH oxidase in cell-free system consisting of cytosol and plasma membrane 5 .Less is known about the possible role of microtubules in the oxidative burst.Reibman et al. 6 showed that microtubule disruption in neutrophils treated with colchicine did not affect superoxide radical production stimulated via Fc receptors or receptors for N-formyl-Met-Leu-Phe (fMLP), thus the assembly of NADPH oxidase did not require intact microtubules.Treatment of neutrophils with vincristine, an inhibitor of microtubule assembly, did not impair myristate-induced superoxide release as well 7 .However, Wiles et al. 8 demonstrated that colchicine attenuated and microtubule stabilization by paclitaxel increased fMLPinduced oxygen radical production by human neutrophils while experiments of Piazzolla et al. 9 showed that paclitaxel inhibited fMLP-induced superoxide release by human neutrophils and colchicine had no effect.
The discrepancies in observed action of paclitaxel and colchicine inspired us to examine the effect of microtubule interfering agents, paclitaxel, colchicine, nocodazole, and vincristine, on fMLP and phorbol myristate acetate (PMA)-induced oxidative burst in DMSO-differentiated HL-60 cells.We included in our study also radicicol, an inhibitor of heat shock protein 90 (Hsp90) 10 , to investigate possible involvement of Hsp90 in signalling pathways leading to the oxidative burst, though radicicol was found to have no effect on fMLP-induced superoxide generation in primed or unprimed human neutrophils 11 .

HL-60 cell line maintenance and differentiation into neutrophils
Human promyelocytic cell line HL-60 (ECACC No. 98070106) was maintained in RPMI-1640 medium containing 5 mM L-glutamine, 100 U/ml penicillin, 0.17 mM streptomycin, and 10 % (v/v) fetal bovine serum in a humidified incubator at 37 °C and 5 % CO 2 .Cell density was maintained between 1 and 10 × 10 5 cells/ml.Prior to differentiation cells were collected by centrifugation for 3 min at 240 g, resuspended in fresh medium and diluted to density 5 × 10 5 cells/ml.DMSO was added to reach 1.25 % (v/v) and the cells were cultured for 7 days in a humidified incubator at 37 °C and 5 % CO 2 .On day three fresh medium with DMSO was added to prevent cell starvation.Differentiated cells were collected by centrifugation as described above, washed once in PBS buffer (137 mM NaCl, 2.7 mM KCl, 9 mM Na 2 HPO 4 , 1.5 mM KH 2 PO 4 , pH 7.4), and finally resuspended in PBS supplemented with 0.5 mM CaCl 2 , 1 mM MgCl 2 , and 30 mM D-glucose (PBS+).Cells were then kept on ice and used for further experiments only when capable of superoxide generation, which was sensitive to inhibition by DPI and SOD.

Chemiluminescent detection of superoxide formation
Differentiated HL-60 cells (final density 10 6 cells/ml) were transferred into fluorescence quartz cuvette containing PBS+ buffer pre-warmed to 37 °C.DMSO (control) or a solution of tested substance in DMSO or DPI (10 µM)  or SOD (100 U/ml) were added 5, 30 or 60 min prior to cell stimulation.MCLA (4µM) was added 30 s prior to start of measurement and the background chemiluminescence was obtained.Superoxide generation was initiated 60 s after the start of measurement by fMLP (5 µM) or PMA (0.1 µM).Maximum DMSO concentration in assay was 0.95 % (v/v).
The time course of MCLA-dependent chemiluminescence was monitored on Perkin Elmer LS50B luminometer (Perkin Elmer, Norwalk, CT, USA) equipped with total emission mirror and thermostatted cuvette holder allowing continuous stirring.The areas under chemiluminescence peak were integrated and used for calculation of half-maximal concentration (IC 50 ) necessary for inhibition of oxidative burst.Activity detected in samples containing 100 U/ml SOD was used as 100 % inhibition.

RESULTS AND DISCUSSION
The DMSO-differentiated HL-60 cells respond to chemotactic peptide fMLP or phorbol ester PMA by triggering the superoxide radical generation.To explore the possible involvement of microtubules and Hsp90 in the oxida tive burst, cells were pre-incubated with paclitaxel (1-5 µM), colchicine (1-100 µM), nocodazole (1-20 µM), vincristine (1-50 µM), and radicicol (1-50 µM) for 5 or 30 min at 37 °C and subsequently stimulated by fMLP or PMA.Tested microtubule interfering agents act by different mechanisms resulting in microtubule stabilization (paclitaxel) or microtubule disruption (colchicine, nocodazole, vincristine) 12 , but none of these substances markedly af-N-formyl-Met-Leu-Phe-induced oxidative burst in DMSO-differentiated HL-60 cells requires active Hsp90, but not intact microtubules PMA interacts directly with protein kinase C (PKC) activating NADPH oxidase 15 .We hypothesize that Hsp90 acts upstream of PKC and in agreement with experiments of Inanobe et al. 16 , which demonstrated that Hsp90 from HL-60 cells interacted with free βγ subunits of heterotrimeric G-proteins, we suggest Hsp90 as a possible target for modulation of the oxidative burst.fected either PMA or fMLP-induced oxidative burst in differentiated HL-60 cells (Table 1).While nocodazole appears to partially inhibit fMLP-induced oxidative burst, this inhibition may be attributed to non-specific effects due to substance precipitation which was noticeable even with the 20 µM concentration of nocodazole.In contrast, radicicol, an inhibitor of Hsp90 (ref. 10 ), inhibited fMLPinduced oxidative burst in time (Table 2) and concentration-dependent manner (Fig. 1).Cells of later passage, i.e. cells cultured for 6 or more weeks before differentiation, were more responsive to radicicol than cells differentiated 2 weeks after resuscitation (Table 2).In case of PMA stimulation we observed only partial inhibitory effect that required treatment of cells with 50 µM radicicol for 30 min (Table 1).Moreover, radicicol at concentrations 1-50 µM did not inhibit sodium dodecyl sulfate-induced activation of NADPH oxidase in cell-free system containing cytosol and membranes obtained from unstimulated DMSO-differentiated HL-60 cells (data not shown).
Cell resistance towards microtubule interfering agents may result from β-tubulin gene mutations or multidrug resistance allowed by P-glycoprotein 13 .Because tested compounds bind to tubulin via different binding sites 12 and neutrophils lack P-glycoprotein, which may eliminate colchicine 14 , paclitaxel, and vincristine 13 , we conclude that both PMA and fMLP-induced oxidative burst in DMSOdifferentiated HL-60 cells is microtubule-independent.The requirement for Hsp90 function in fMLP-induced oxidative burst results probably from the triggering mechanisms where binding of fMLP to its G-protein-coupled receptors activates multiple signalling pathways while

Fig. 1 .
Fig. 1.Dose-response curve for radicicol inhibition of fMLP-induced oxidative burst in DMSO-differentiated HL-60 cells of early passage.10 6 cells/ml were pre-incubated with different concentrations of radicicol for 30 min prior to stimulation by fMLP.Inhibition was calculated from chemiluminescent trace areas where 0 % inhibition represents superoxide generation in the presence of fMLP alone and 100 % inhibition in the presence of 100 U/ml SOD and fMLP.The IC 50 value derived from three experiments was 16.5 ± 3.5 µM radicicol.
col prior to stimulation by fMLP.Dose-response curves were constructed and respective IC 50 s were calculated.Data are means from at least two experiments.Values in column IC 50 were determined with differentiated HL-60 cells of early passage, in column IC 50 * with cells of later passage.