Timed-pregnant Fisher F344 rats were obtained from Charles River Laboratories (Raleigh, NC). Eight-week-old male C57BL/6J mice, weighing 25–28 g, were maintained on a 12:12 h light:dark cycle and fed ad libitum. PHOX-deficient (gp91 PHOX^-/-) and wild-type C57BL/6J (gp91 PHOX^+/+) mice were obtained from the Jackson Laboratory (Bar Harbor, ME). Animals were adapted for 2 weeks to the conditions described above before experimentation. To examine the effect of FLZ on MPTP-induced neurotoxicity, mice received daily MPTP injections [15 mg/kg of MPTP.HCl, s.c.] for 6 consecutive days. From the third day on, FLZ (75 mg/kg, p.o.) was administered 30 min before every MPTP injection for the last 4 days. Housing, breeding, and experimental use of the animals were performed in strict accordance with the National Institutes of Health guidelines.

FLZ was kindly provided by Professor Xiaotian Liang in the Department of Pharmaceutical Chemistry, Institute of Materia Medica, Chinese Academy of Medical Science. It is a white powder with 99% purity. The polyclonal anti-tyrosine hydroxylase (TH) antibody was a generous gift from Dr. John Reinhard (GlaxoSmithKline, Research Triangle Park, NC). LPS (strain O111:B4) and 2', 7'-dichlorofluorescin diacetate were purchased from Calbiochem (San Diego, CA). [³H]-DA (28 Ci/mmol) was purchased from PerkinElmer Life Sciences Inc. (Boston, MA). The mouse monoclonal antibody raised against OX42 antigen was purchased from Chemicon International (Temecula, CA). Rabbit anti-p47^phox was obtained from Millipore Corporation (Bedford MA). FITC-conjugated goat anti-rabbit IgG antibody was obtained from Jackson ImmunoResearch Laboratories Inc. (West Grove, PA). Rabbit anti-GAPDH antibody was obtained from Abcam (Cambridge, MA). Mouse anti-gp91^phox antibody was purchased from BD Transduction Laboratories (San Jose, CA). All other reagents came from Sigma (St Louis, MO).

Primary rat mesencephalic neuronal-glial cultures were prepared as described previously 18. Briefly, after the ventral mesencephalic tissues were removed and dissociated by a mechanical triturating, cells were seeded at 5 × 10⁵/well to 24-well culture plates or 1.5 × 10⁵/well to 96-well culture plates in maintenance medium. Seven-day-old cultures were used for treatment. Immunocytochemical analysis indicated that at the time of treatment the cultures were made up of ~10% microglia, ~50% astroglia, and ~40% neurons, of which 1 ~2% were TH-immunoreactive neurons.

Primary microglia-enriched cultures were prepared from the whole brains of 1 or 2-day-old mice, as described previously 19. Briefly, brain tissues, devoid of meninges and blood vessels, were dissociated by mechanical trituration. The isolated cells (5 × 10⁷) were seeded in 150 cm² culture flasks in Dulbecco's modified Eagle medium. Upon reaching confluence (12–14 days), microglia were separated from astroglia by shaking the flasks at 180 rpm for 1 h. Purity of the microglia-enriched cultures was >98%, as determined by immunocytochemical staining.

The rat microglia HAPI cells were generous gifts from Dr. James R. Connor (Pennsylvania State University, Hershey, PA). They were maintained at 37°C in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 50 U/ml penicillin, and 50 μg/ml streptomycin in a humidified incubator with 5% CO₂/95% air.

Degeneration of DA neurons was assessed by measuring the ability of cultures to take up DA. DA uptake assays were performed as previously described 18. Briefly, after two washes with warm Krebs-Ringer buffer the cultures were incubated for 20 min at 37°C with 1 μM [³H]-DA for DA uptake. Then the cultures were washed 3 times with ice-cold KRB, and the cells were then dissolved in 1 N NaOH. Radioactivity was determined by liquid scintillation counting. Nonspecific DA uptake, determined in the presence of mazindol (10 μM), was subtracted.

DA neurons were recognized with an anti-TH antibody, and microglia were detected with the OX-42 antibody, as described previously. For morphological analysis, images were acquired using an inverted microscope (Nikon, Tokyo, Japan) connected to a camera (DAGE-MTI, Michigan City, IN) operated with the MetaMorph software (Universal Imaging Corporation, Downingtown, PA). To quantify cell numbers, total TH positive cells in entire wells were counted by three individuals. The average of these scores was reported.

For the TNF-α assay, culture supernatant was collected after 3 h of stimulation with LPS. The concentration of TNF-α was measured with a mouse TNF-α enzyme-linked immunosorbent assay kit from Genzyme (Cambridge, MA).

The production of NO was assessed as the accumulation of nitrite in the culture supernatant, using a colorimetric reaction with the Griess reagent. The culture supernatants were collected after 24 h of stimulation with LPS and mixed with equal volumes of the Griess reagent. The absorbance at wavelength 540 nm was measured with a UV MAX kinetic microplate reader (Molecular Devices).

The production of PGE₂ in the enriched microglial cultures was evaluated 24 h after LPS treatment using a PGE₂ EIA kit from Cayman (Ann Arbor, MI), according to the manufacturer's instructions.

Production of superoxide was determined by measuring the superoxide dismutase (SOD)-inhibitable reduction of the water-soluble tetrazolium salt (WST-1) 30 min after LPS treatment 20. Microglia-enriched cultures in 96-well culture plates were washed twice with Hanks' Balanced Salt Solution (HBSS) without phenol red. Cultures were then incubated at 37°C for 1 h with vehicle control or FLZ in HBSS (50 μl/well). Then, 50 μl of HBSS with and without SOD (50 U/ml) was added to each well along with 50 μl of WST-1 (1 mM) in HBSS and 50 μl of vehicle or LPS (2 ng/ml). Thirty minutes later, the absorbance at 450 nm was read with a SpectraMax Plus microplate spectrophotometer (Molecular Devices, Sunnyvale, CA).

The production of intracellular ROS was measured by 2', 7'-dichlorofluorescin diacetate oxidation 2 h after LPS treatment. Microglia-enriched cultures were seeded (5 × 10⁴) in 96-well plates and then exposed to 20 μM 2', 7'-dichlorofluorescin diacetate for 1 h, followed by pretreatment with FLZ for 1 h and treatment with HBSS containing LPS. After incubation, the fluorescence was read at the 485 nm excitation and 530 nm emission on a fluorescence plate reader. Cell-free experiments with and without FLZ were conducted to determine that the reagents themselves did not alter fluorescence.

The primers synthesized from Sigma Genosys were used as follows: TNF-α, forward (TCGTAGCAAACCACCAAGCA) and reverse (CCCTTGAAGAGAACCTGGGAGTA); iNOS, forward (GTGCTAATGCGGAAGGTCATG) and reverse (CGCTTCCGACTTTCCTGTCT); COX₂, forward (CCAGCAGGCTCATACTGATAGGA) and reverse (GCAGGTCTGGGTCGAACTTG); GAPDH, forward (CCTGGAGAAACCTGCCAAGTAT) and reverse (AGCCCAGGATGCCCTTTAGT); and gp91, forward (CCTGCAGCCTGCCTGAATT) and reverse (AAGGAGAGGAGATTCCGACACA).

HAPI cells were treated with 10 ng/ml LPS for 24 h in the presence or absence of 10 μM FLZ. The cells were removed from culture and washed twice with cold FACS buffer (saline with 5% normal goat serum). Cells were incubated with FcR block for 30 min at 4°C and then incubated for 45 min at 4°C with PE-conjugated OX6 antibody or the appropriate isotype control antibody (BD Pharmingen). After antibody binding, cells were washed and fixed. Fluorescence was analyzed on a FACSCalibur (BD Biosciences). The mean fluorescent intensity (MFI) of experimental group was determined by subtracting the MFI of the isotype control from the MFI of the unstimulated or LPS-stimulated microglia.

Subcellular fractionation was performed as described previously 21. HAPI cells, seeded in dishes at 5 × 10⁴ cells/well, were treated with 10 ng/ml LPS for 10 min in the absence or presence of 10 μM FLZ for 1 h. The cells were then lysed in hypotonic lysis buffer, incubated on ice for 30 min, and subjected to Dounce homogenization (~20–25 strokes, tight pestle A). The lysates were loaded onto sucrose in lysis buffer and centrifuged at 1,600 g for 15 min; the supernatant above the sucrose gradient was used as the cytosolic fraction after centrifugation at 150,000 g for 30 min. The pellets, solubilized in 1% Nonidet P-40 hypotonic lysis buffer, were used as the membranous fraction. For western blot analysis, the primary antibodies included anti-p47^phox (1: 2000), anti-GAPDH (1:2000) and anti-gp91^phox (1:2000).

Quantitative measurements of motor coordination in mice were performed using an accelerating rota-rod (model 7650, UGO Basile, Comerio, VA, Italy). The rota-rod consisted of a plastic rod (diameter = 3 cm; length = 30 cm) partitioned off with round plates to prevent the mice from escaping from the sides of the rod. The rod was covered with smooth plastic tubing and suspended 16 cm above five plastic levers attached to timers that stop when mice land on the lever surface. The mice were placed on the rod that rotated at 3 rpm at 0 min and accelerated from 4 to 20 rpm over 5 min. Mice were oriented perpendicular to the long axis of the rod, such that the mice had to make forward walking movements to avoid falling. The latency time to fall was measured using mice 28 days after their first dose of MPTP, with each mouse experiencing three trials per day with an intertrial interval of 10 min. The average latency of each mouse was calculated.

Mouse brains were cut on a horizontal sliding microtome into transverse free-floating sections, 35 μm in thickness. Six to eight brain sections of SNpc were collected at intervals of 140 μm. For immunohistochemistry, the following concentrations of antibodies were used: anti-TH (1:5000), anti-Iba-1 (1: 1000), biotinylated anti-mouse IgG (1:1000). Digital images of TH neurons in SNpc were acquired on an Olympus microscope (Olympus^®, Tokyo, Japan) using an attached Polaroid digital microscope camera (Polaroid^®, Cambridge, MA, USA). Twenty four consecutive brain slices (35-μm thickness), which encompassed the entire substantia nigra compacta, were collected. A normal distribution of the number of TH-IR neurons in the SNpc was constructed based on the counts of 24 slices from saline-treated mice, MPTP-treated mice and FLZ-treated mice. The distribution curves from these groups superimpose and show no difference in number and shape of the curves. Six to eight evenly spaced brain slices from different groups of animals were immunoreacted using an antibody against TH, and immunoreactive cells were counted. The distribution of cell numbers from each animal was matched with the normal distribution curve to correct for errors resulting from the cutting. Three individuals performed counting in a double-blind manner. Conclusions were drawn only when the interobserver differences were within 5%. A mean value for the number of SNpc TH neurons was then deduced by averaging the counts of 6–8 sections for each animal.

The data are presented as mean ± S.E.M. For multiple comparisons of groups, two-way ANOVA was used. Statistical significance of differences between groups was assessed using paired Student's t test, followed by Bonferroni correction using the JMP program (SAS Institute, Cary, NC, USA). A value of P < 0.05 was considered statistically significant.

Mesencephalic neuronal-glial cultures were pretreated with FLZ or vehicle for 1 h, or post-treated with 10 μM FLZ at 0, 0.5, 1, 2 and 3 h after LPS treatment. Seven days later, the degeneration of DA neurons was assessed. [³H]DA uptake assays indicated that LPS treatment reduced the uptake capacity to about 40% of that of vehicle-treated control cultures (Fig. 2A). FLZ attenuated LPS-induced DA uptake decreases in a dose-dependent manner; whereas 10 μM FLZ alone did not affect DA uptake in cultures. Morphologically, in addition to reductions in the abundance of TH neurons, the dendrites of the remaining TH neurons in LPS-treated cultures were significantly less elaborate than those in the control group (Fig. 2B). FLZ significantly attenuated these morphological changes in TH neurons induced by LPS. These protective effects of FLZ were confirmed by the cell count numbers of TH neurons (Fig. 2C). For the post-treatment groups, 10 μM FLZ showed a protective effect even when given 2 h after LPS treatment (Fig. 2D). However, when 10 μM FLZ was added at 3 h post-LPS treatment, no significant protective effect was observed.

In addition to neurons, the mesencephalic neuron-glia cultures also contain ~10% microglia and ~50% astrocytes. To evaluate the potential contribution of glia in mediating the neuroprotective effect of FLZ, we performed cell reconstitution experiments. As shown in Fig 3, 1-methyl-4-phenylpyridine (MPP⁺) reduced DA uptake by about 52% in neuronal-glial cultures, but 10 μM FLZ significantly protected neurons against this damage. However, 10 μM FLZ failed to show any protective effect against the toxicity of MPP⁺ in microglia-depleted cultures, which contain only neurons and astrocytes. In contrast, when 10% microglia were added back to the neuron-enriched cultures, the MPP⁺-induced reduction in DA uptake could be reversed by 10 μM FLZ. These results demonstrate that microglia, but not astroglia, mediate FLZ's neuroprotective effect against LPS-induced neurodegeneration.

Immunohistochemical studies using the OX-42 antibody, a marker for rat microglia, showed LPS-induced morphological evidence of microglial activation in mesencephalic neuronal-glial cultures. LPS treatment transformed microglia from a predominantly resting, round, small cell morphology to an activated, rod- and/or amoeboid-shaped cell morphology with intensified immunoreaction (Fig. 4A). LPS-stimulated activation of microglia was partially suppressed by pretreating the cultures with 10 μM FLZ, whereas the same concentration of FLZ alone did not show significant effects on microglial morphology. The effect of FLZ on activated microglia was further confirmed by flow cytometric analysis using an anti-MHC-II antibody, a marker for the activated microglia. Quantitative analysis revealed that the degree of 10 ng/ml LPS-induced activation of microglia was decreased by 60% by pretreatment with 10 μM FLZ (Fig. 4B). Rat HAPI-immortalized microglial cells were used for this flow cytometric analysis because most of the primary microglial cells died during the experimental procedure.

Accumulating evidence indicate that proinflammatory factors released by over-activated microglia in dysregulated inflammatory conditions play a key role in DA neurodegeneration 181922. Therefore, we hypothesized that FLZ would dampen LPS-stimulated production of proinflammatory factors in microglia-enriched cultures. We found that pretreatment with FLZ at concentrations ranging from 2.5 to 10 μM significantly blocked LPS-stimulated production of proinflammatory mediators and their mRNA expression in a dose-dependent manner. These gene products include TNF-α (Fig. 5A and 5B), inducible nitric oxide synthase (iNOS) (Fig. 5C and 5D), and COX-2 (Fig. 5E and 5F). In addition, FLZ significantly attenuated LPS-induced ROS production including extracellular superoxide (Fig. 5G) and intracellular ROS (Fig. 5H).

Among the proinflammatory factors released from over-activated microglia, production of superoxide from PHOX has been shown to exert direct neurotoxicity and to regulate gene expression for other proinflammatory factors 19. To further investigate the role of ROS in FLZ-elicited neuroprotective effects, we used neuronal-glial cultures from mice deficient in the catalytic subunit (gp91^PHOX) of PHOX, which is the key enzyme required for the production of ROS. As shown in Fig. 6A, LPS treatment of neuronal-glial cultures prepared from PHOX^+/+ mice substantially reduced [³H]-DA uptake and, similarly, FLZ attenuated this decrease. In contrast, LPS treatment also showed a significant but smaller reduction in [³H]-DA uptake capacity in cultures from PHOX^-/- mice, while FLZ's protective effect was reduced significantly. Further, results from real-time PCR showed that FLZ inhibited the gp91 mRNA expression induced by LPS treatment (Fig. 6B). Therefore, our data support the idea that PHOX plays an important role in FLZ-mediated protection against LPS-induced neuronal degeneration.

It has been reported that activation of PHOX holoenzyme is required to initiate the translocation of its cytosolic components p47^phox into the plasma membrane 23. Because of the potent inhibition of ROS production by FLZ, we sought to determine whether FLZ inhibits PHOX activation by preventing this translocation. Data from subcellular fractionation and western blot analysis clearly showed an increase in immunoreactivity of p47^phox in the membrane of microglia 10 min after LPS treatment (Fig. 6C), and this increase in immunoreactivity was significantly blocked in the presence of FLZ. Thus, this result further proves that FLZ-mediated inhibition of superoxide production by LPS is mediated through suppression of p47^phox membrane translocation; that is, activation of PHOX.

To further substantiate the neuroprotective effect of FLZ shown in these cell culture studies, we performed an animal study with daily injections of MPTP for 6 days.

Immunohistochemical analysis (Fig. 7A) revealed that MPTP injection caused significant DA neuronal loss in SNpc. Both number of DA neurons and extent of neuronal processes were significantly lost in MPTP-treated group mice. For a second group, FLZ was given to the mice starting on the third day of MPTP treatment and injections of FLZ were continued daily for the next four days. This post-treatment regimen with FLZ significantly reduced MPTP-induced DA neuronal loss. Cell count analysis showed that MPTP treatment decreased the number of DA neurons in the SNpc to about 45% of the vehicle treated group, while post-treatment with FLZ increased the number of remaining neurons to about 75% of the vehicle-treated group. For a third group of mice that received only FLZ, the number of DA neurons did not differ from the vehicle-treated group (Fig. 7B). Morphological changes of microglia revealed by immunoreaction with Iba-1 antibody were used as an index for inflammatory conditions. MPTP treatment transformed most of the microglia from resting cells to activated, large cells, and this activation of microglia was significantly suppressed by FLZ post-treatment (Fig. 7C). To obtain quantitative data, the midbrains of the mice were dissected out and changes in Iba-1 immunoreactivity were determined using western blot assays. The results confirmed that FLZ suppressed MPTP-induced microglial activation (Fig. 7D). To access the locomotor activities of the mice, the rota-rod test was used. At 28 days after the first MPTP injection, the length of time that MPTP-group mice remained on the rod was much less than that of vehicle-group mice, while post-treatment with FLZ significantly prolonged the time that the mice were able to remain on the rod (Fig. 7E). These results from rota-rod testing thus corroborated the immunohistochemical results showing loss of DA neurons.