2017 Volume 40 Issue 4 Pages 402-412
Heat shock protein 70 (Hsp70) is not only a molecular chaperone in cytosol, but also presents in synaptic plasma membranes. To detect plasmalemmal Hsp70 (pl-Hsp70), neurons were immunostained with anti-Hsp70 antibody without permeabilization and fixation. Dotted immunofluorescent signals at neuronal cell bodies and neurites indicated the localization of Hsp70 on the neuronal cell surface. To target only pl-Hsp70, but not cytosolic Hsp70, the anti-Hsp70 antibody was applied without permeabilization in the primary culture of rat cortical neurons. The antibody induced neuronal cell death in a concentration-dependent manner. The anti-Hsp70 antibody activated ubiquitin-proteasome pathway, but inactivated caspase-3. A lag time was required for the neurotoxicity of anti-Hsp70 antibody. Hydrogen peroxide was increased in the anti-Hsp70 antibody-treated neurons during the lag time. Catalase suppressed the anti-Hsp70 antibody-reduced cell viability via the plausible inhibition of hydrogen peroxide generation. One of down-streams of hydrogen peroxide exposure is activation of the mitogen-activated protein kinase (MAPK) signaling cascade. The neurotoxicity of anti-Hsp70 antibody was partially ascribed to c-Jun N-terminal kinase among MAPKs. In conclusion, the anti-Hsp70 antibody targeted pl-Hsp70 on the neuronal cell surface and induced neuronal cell death without complement. Furthermore, hydrogen peroxide appeared to mediate the neuronal cell death, which was accompanied with the enhancement of the ubiquitin–proteasome pathway and the suppression of caspase in a different fashion from the known cell death.
Heat shock proteins (Hsps) are stress-induced proteins that are involved in cellular protection and repair mechanisms.1) Hsps are distributed across different subcellular compartments including the cytosol, nucleus, endoplasmic reticulum and mitochondria. Hsps also exist as extracellular proteins and are released via exosomes in both the resting and stress-induced state.2) In tumor cells, the dotted staining pattern of anti-Hsp70 antibody reveals the localization of membrane-bound Hsp70 in lipid rafts.3) In peripheral tissues, Hsp70 is presented on the cell surface4) and released from endothelial cells.5) Although Hsp70 is also released from neurons,6) it has not yet been reported to be detected on the neuronal cell surface. The Hsp70 family comprises stress-inducible Hsp70 and constitutively expressed heat shock cognate protein 70 (Hsc70). However, neuronal cells do not induce or induce lower levels of Hsp70 in response to stressful stimuli.2,7) Compared with non-neural tissues, brain tissue shows constitutive expression of Hsp70.8,9) Hsp7010,11) and Hsc7012,13) are present in synaptic plasma membranes. In the present study, we ascertained whether Hsp70 family proteins are presented on the neuronal cell surface.
Hsp70 family is the target for 15-deoxy Δ12,14-prostaglandin J2 (15d-PGJ2),14) and plasmalemmal Hsp70 (HSPA8) has been suggested to be one of membrane targets for 15d-PGJ2.15) 15d-PGJ2 has been reported to be increased in cerebrospinal fluid after the transient focal cerebral ischemia.16) Among arachidonate metabolites, 15d-PGJ2 was found to be the most potent inducer of neuronal apoptosis.17,18) Hsp70 protects neurons from lethal damage induced by environmental stress including ischemia.19) The elevated level of anti-Hsp70 antibody is an independent risk factor for ischemic stroke.20) However, it has not yet been ascertained whether anti-Hsp70 antibody is neuroprotective or neurotoxic.
15d-PGJ2 acts as a neuroprotectant at the lower concentration and a neurotoxicant at the higher concentration.21) Its nuclear receptor, peroxisome proliferators-activated receptor γ (PPARγ) is involved in the neuroprotective effect of 15d-PGJ2.22) Neither PPARγ nor its membrane receptor, chemoattractant receptor homologous molecule expressed on T-helper 2 (Th2) cells (CRTH2), was involved in 15d-PGJ2-induced neurotoxicity.23) 15d-PGJ2 activated caspase and induced neuronal apoptosis.23,24) This apoptosis is accompanied with activation of caspase24) and inhibition of ubiquitin–proteasome pathway (UPP).25) UPP is responsible for the degradation of mutant or misfolded proteins in cells and plays a critical role in maintaining cell homeostasis.26) 15d-PGJ2 interrupts UPP function, resulting in accumulation and aggregation of ubiquitinated (Ub) proteins. In the present study, Hsp70 was found on the neuronal cell surface and localized at neuronal cell bodies and neurites in primary cultures of rat cerebral cortices. Generally, molecular biological approaches are used to analyze protein function. However, it is difficult to specifically target plasmalemmal proteins using such approaches without affecting the cytosolic proteins. Thus, we used anti-Hsp70 antibody to specifically analyze the function of plasmalemmal Hsp70 (pl-Hsp70).
Leibovitz’s L-15 medium, trypsin, fetal bovine serum and horse serum were obtained from Invitrogen (Carlsbad, CA, U.S.A.). Horseradish peroxidase-linked antibody against biotine was obtained from Cell Signaling Technology (Boston, MA, U.S.A.). The protein concentration was measured using the bicinchoninic acid (BCA) protein assay reagent obtained from Pierce (Rockford, IL, U.S.A.). 1-Ethyl-3-(dimethylaminopropyl)carbodiimide (EDC) was obtained from Pierce. An anti-Hsp70 antibody (goat polyclonal, [sc-1060]) and goat normal immunoglobulin G (IgG) [sc-2028] were purchased from Santa Cruz (CA, U.S.A.). Anti-pyruvate kinase M (PKM) antibody (goat polyclonal, [ab-6191]) and anti-aldolase C (AldC) antibody (goat polyclonal, [ab-6190]) were purchased from Abcam (Cambridge, U.K.). Anti-ubiquitin antibody (Z0458) was purchased from Dako Japan (Tokyo, Japan). Anti-c-Jun N-terminal kinase (JNK) antibody (#9252) and phospho-specific antibody (#4671) were purchased from Cell Signaling Technology (Danvers, MA, U.S.A.). Propidium iodide was purchased from DOJINDO (Kumamoto, Japan). 15d-PGJ2 was obtained from Cayman Chemicals (Ann Arbor, MI, U.S.A.; Cabru, Milan, Italy). All other chemicals were of reagent grade.
AnimalsAll procedures were conducted in accordance with NIH guidelines concerning the Care and Use of Laboratory Animals and with the approval of the Animal Care Committee of the Himeji Dokkyo University.15) Pregnant female Wistar rats (E17) were used at the beginning of the experiment. Rats were individually housed in macrolon cages with free access to food and water and maintained on a 12 h light/dark cycle, at 25°C room temperature.
Tissue Cultures and Measurement of Cell ViabilitiesCortical neurons were prepared from rat fetal cerebral cortices and cultured as previously reported.17) Cortical neurons were primarily cultured at a density of 2.5×105 cells/cm2. As shown in Supplementary Fig. 1, the 2 d-old cultures contained neurons (>95%) and few glial cells. Neurons were treated without or with normal goat IgG (sc-2028) or anti-Hsp70 antibody (sc-1016) in the absence or presence of various drugs. Two different methods were employed for assessment of neurotoxicity of anti-Hsp70 antibody. 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide dye (MTT) reduction assay reflecting mitochondrial succinate dehydrogenase activity was employed for assessment of neuronal cell death as previously reported.18) Cell death was also measured by manually counting the percentage of neurons that stained with propidium iodide (PI, 0.1 mg/mL). Nuclei stained with PI were counted from 12 fields with data expressed as percentage PI-stained cells normalized to the vehicle-treated group.
Immunofluorescence MicroscopyMicroscopic immunofluorescence studies were performed with or without permeabilization and fixation. Neurons were primary cultured in glass bottom dishes (Nunc, Rochester, NY, U.S.A.) at a cell density of 20000 cells/well. To confirm that Hsp70 is a cytosolic protein, neurons were fixed with 3.7% formaldehyde in phosphate-buffered saline (PBS) for 15 min and permeabilized with 0.1% Triton-X100 at room temperature. After two washing steps in PBS, neurons were immunostained with normal goat IgG and anti-Hsp70 antibody (dilution 1 : 100) for 30 min at room temperature. To target plasmalemmal Hsp70 but not cytosolic ones, without permeabilization and fixation, neurons were immunostained with anti-Hsp70 antibody (dilution 1 : 100) for 30 min at 4°C. Following immunostaining, neuronal cultures were fixed with 3.7% formaldehyde in PBS for 15 min at room temperature. After two washing steps in PBS, Alexa Fluor 488-conjugated anti-goat IgG antibody was used as a secondary antibody. Alexa Fluor 488-conjugated antibodies were excited at 490 nm and detected at 525 nm. Neurons were treated with 1 µM Hoechst33342 in PBS for 30 min at room temperature. Hoechst33342 was excited at 490 nm and detected at 525 nm. Fluorescent signals were analyzed by confocal laser scanning microscopy (Carl Zeiss, LSM510). Although the anti-Hsp70 antibody used in the present study was commercially available, information on its specificity have not yet been disclosed. Another anti-Hsp70 antibody (ab2787, Abcam) has been reported to cross-react with Hsc70 and Grp78.27) To our knowledge, neurotoxicities of antibodies against Hsc70 and Grp78 have not yet been reported.
Detection of Chromatin Condensation (Fluorescence Microscopy)For nuclei staining, neurons (2 d-old) were treated with normal goat IgG (sc-2028) or anti-Hsp70 antibody (sc-1016) antibody at 200 ng/mL for 48 h. The nuclear chromatin of trypsinized cells was stained with 10 µM Hoechst 33342 (Nacalai Tesque, Kyoto, Japan) in the dark at room temperature for 15 min.28) They were then observed with a bright field fluorescent microscope (VANOX; Olympus, Tokyo, Japan) under UV excitation. Chromatin-condensed cells were photographed at 40-fold magnification. In addition, at 20-fold magnification, more than 100 cells with condensed chromatin were counted in each experiment and their percentage was calculated.
Fluorimetric Assay of Caspase-3 ActivityCaspases-3 activity was assessed using a Caspase-3 Assay Kit, Fluorimetric (SIGMA, St. Louis, MO, U.S.A.) as described previously.15) After exposure to antibodies or 15d-PGJ2 for 24 h on day 2, the supernatants were aspirated and cells were harvested with lysis buffer (50 mM N-(2-hydroxyethyl)piperazine-N′-2-ethanesulfonic acid (HEPES), pH 7.4, 5 mM CHAPS and 5 mM dithiothreitol (DTT)). The reaction buffer, including acetyl-Asp-Glu-Val-Asp-7-amido-4-methylcoumarin (Ac-DEVD-AMC), caspase-3 specific substrates, was added to wells, and the production of AMC was sequentially detected in a CytoFluor® Plate Reader at an excitation wavelength of 360 nm/emission 460 nm. Enzyme activities were determined as initial velocities expressed as nmol AMC/min/mL. They were then corrected with the quantity of protein in each well detected by BCA protein assays (Thermo Fisher Scientific, Waltham, MA, U.S.A.).
Western BlottingUbiquitin immunoblot analyses were performed after exposure to normal goat IgG (sc-2028) or anti-Hsp70 antibody (sc-1016) for 24 h on day 2 neuronal cells. Phosphorylated and total JNK immunoblot analyses were performed after exposure to anti-Hsp70 antibody for 15 min on day 3 neuronal cells. After exposure to each antibody, the supernatants were aspirated and cells were harvested with lysis buffer (50 mM HEPES, pH 7.4, 5 mM CHAPS and 5 mM DTT).29) Neuronal cell lysates were subjected to Western blotting. Fifteen micrograms of total protein per lane was separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose (Amersham Biosciences, U.S.A.). Blots were blocked for 1 h in 5% nonfat dry milk-Tris-buffered saline (TBS)-0.1% Tween 20 and incubated overnight with anti-ubiquitin antibody (1/1000) in 5% milk-TBS-0.1% Tween 20 at 4°C (in case of phospho-Abs, 3% bovine serum albumin (BSA) was used instead of milk). This was followed by 1 h incubation with appropriate secondary peroxidase-conjugated Ab (1/2000; Sigma-Aldrich). Immunoreactivity was detected using the ECL detection method, according to the manufacturer’s instructions (Amersham Biosciences).
Measurement of Hydrogen PeroxideHydrogen peroxide activity was assessed using a Red Hydrogen Peroxide Assay Kit® (Enzo Life Sciences, Farmingdale, NY, U.S.A.) according to the manufacturer’s instructions.30) Briefly, cells were seeded into 24-well plates at a density of 25000 cells/cm2. Neurons were treated without or with normal goat IgG (sc-2028) or anti-Hsp70 antibody (sc-1016) in the absence or presence of catalase. After exposure to drugs for 12 h, the supernatants were aspirated and cells were harvested with 0.1% deoxycholic acid. Cell lysates were added to the reaction buffer, including Red Peroxidase Substrate and horseradish peroxidase. The production of hydrogen peroxide was sequentially detected in a Micro plate reader at an excitation wavelength of 570 nm/emission 590 nm.
Statistical AnalysisData are given as the mean±standard error (S.E.) (n=12) or mean±standard deviation (S.D.) (n=3). Representative data from three independent experiments with similar results are shown. Data were analyzed statistically by use of Student’s nonpaired t-test for comparison with the control group, and data on various groups were analyzed statistically by use of two-way ANOVA followed by Dunnett’s test.
To confirm that Hsp70 is a cytosolic protein, we immunostained neurons with permeabilization and fixation at room temperature. A phase contrast photograph was shown in Fig. 1A (left upper panel). Immunofluorescence with anti-Hsp70 antibody indicated that Hsp70 was localized at neuronal cell bodies and neurites (Fig. 1A: right upper panel). Neuronal nucleus were stained with Hoechst33342 (Fig. 1A: left lower panel). The merged photograph indicated that Hsp70 was localized in the cytosol around nuclear (Fig. 1A: left right lower panel). We confirmed that Hsp70 was constitutively localized in the cytosol of primary neurons.7) On the other hand, little immunofluorescent signals was detected with control IgG at room temperature in permeable and fixed neurons (Fig. 2A).
(A) Neurons were permeabilized, fixed and stained at room temperature with anti-Hsp70 antibody (green fluorescence) and Hoechst33342 (blue fluorescence). (B) Without permeabilization and fixation, neurons were stained with anti-Hsp70 antibody (green fluorescence) and Hoechst33342 (blue fluorescence) at 4°C. A phase-contrast photograph (left upper panel), anti-Hsp70 antibody-stained Hsp70 (right upper panel), Hoechst33342-stained neuronal nucleus (left lower panel), and the merged photograph (right lower panel). Scale bar=10 µm.
(A) Neurons were permeabilized, fixed and stained at room temperature with normal goat IgG (green fluorescence) and Hoechst33342 (blue fluorescence). (B) Without permeabilization and fixation, neurons were stained with normal goat IgG (green fluorescence) and Hoechst33342 (blue fluorescence) at 4°C. A phase-contrast photograph (left upper panel), normal goat IgG (right upper panel), Hoechst33342-stained neuronal nucleus (left lower panel), and the merged photograph (right lower panel). Scale bar=10 µm.
Hsp70 is associated with synaptosomal membranes,11) suggesting its localization to the neuronal cell surface. To target plasmalemmal Hsp70, cytosolic Hsp70 staining was excluded, as the entire staining procedure was performed at 4°C without permeabilization and fixation. After the treatment with the first antibody, neurons were fixed and treated with Alexa Fluor 488-conjugated anti-goat IgG antibody as a secondary antibody. Figure 1B illustrated staining patterns of the anti-Hsp70 antibody to neurons in a different fashion from Fig. 1A. The immunostaining at 4°C without permeabilization and fixation revealed a typical ring-shaped surface staining pattern (Fig. 1B: right lower panel). The dotted staining pattern around neuronal cell bodies and neurites reflected the localization of membrane-bound Hsp70 in lipid rafts.5) On the other hand, little immunofluorescent signals were detected by control IgG without permeabilization and fixation (Fig. 2B).
Neurotoxicity of Anti-Hsp70 Antibody in the Primary Culture of Cortical NeuronsSignificances of pl-Hsp70 have not yet been sufficiently clarified. How did anti-Hsp70 antibody influence neurons? Since 15d-PGJ2 induces neuronal cell death, its membrane target, Hsp70, might be involved in the neuronal cell survival. Therefore, we examined effects of anti-Hsp70 antibody on neuronal cell survival (Fig. 3). To target only pl-Hsp70, but not cytosolic Hsp70, anti-Hsp70 antibody was applied without permeabilization for 48 h in primary cultures of rat cortical neurons. Among goat polyclonal antibodies, anti-Hsp70 antibody exhibited neurotoxicity, but neither anti-PKM antibody nor anti-AldC antibody did (Fig. 3A). Commercially available antibodies contain some additive agents, such as sodium azide and albumin. To rule out the possibility that some additive agents may affect the cell viability, we tested a normal goat IgG (goat polyclonal, Santa Cruz) which was solved in the same vehicle of the anti-Hsp70 antibody (goat polyclonal, Santa Cruz). As shown in Fig. 3A, normal goat IgG did not affect neuronal cell survival.
(A) Neurons were treated with normal goat IgG and antibodies against Ald (Aldolase C), PK (pyruvate kinase M) and Hsp70 at 200 ng/mL for 48 h. Cell viabilities were determined by MTT reducing activity. Data are expressed as the mean±S.D. (n=3). ** p<0.01, compared with control. (B) Neurons were treated with 50 ng/mL goat IgG and anti-Hsp70 antibody in the absence (open columns) or presence (closed columns) of 3.5 µM 15d-PGJ2 for 48 h. Cell viabilities were determined by MTT reducing activity. Data are expressed as the mean±S.D. (n=3). ** p<0.01, compared with control. ## p<0.01, compared with 15d-PGJ2 alone. (C) Anti-Hsp70 antibody increased PI-stained nuclei. Neurons were untreated (control) or treated with normal goat IgG (200 ng/mL), anti-Hsp70 antibody (200 ng/mL) or 15d-PGJ2 (5 µM) for 24 h (2 d-old cultures). PI-stained neurons were photographed. Scale bar=50 µm. (D) PI-stained neurons were counted. Cell death is expressed as % PI normalized to control goat IgG. Data are expressed as the mean±S.E. (n=12). * p<0.05, ** p<0.01, compared with control.
15d-PGJ2 can be a neuroprotectant or a neurotoxicant, depending on its concentration.21) We examined effects of 15d-PGJ2 on the anti-Hsp70 antibody-induced neuronal cell death. Cortical neurons were treated with 50 ng/mL goat IgG or anti-Hsp70 antibody in the absence or presence of 3.5 µM 15d-PGJ2 for 48 h. The value of EC50 of 15d-PGJ2 against neuronal cells was about 3.5 µM in the presence of serum. 15d-PGJ2 did not suppress the neurotoxicity of the anti-Hsp70 antibody at the low concentration less than 1 µM. On the other hand, the neurotoxicity of 15d-PGJ2 at the high concentration, 3.5 µM, and that of the anti-Hsp70 antibody were additive (Fig. 3B).
Alternatively, we evaluated the neurotoxicity of anti-Hsp70 antibody with PI staining (Figs. 3C, D). Neurons treated with normal goat IgG (200 ng/mL), anti-Hsp70 antibody (200 ng/mL) or 15d-PGJ2 (5 µM) for 24 h. Little neuron was stained with PI under control condition. 15d-PGJ2 stained neurons markedly as a positive control, whereas goat IgG did not as a negative control (Fig. 3C). Consistent with the MTT assay results, there was a significant increase in the percentage of PI-stained neurons in anti-Hsp70 antibody-treated cultures as compared to control goat IgG-treated ones (Fig. 3D).
Anti-Hsp70 Antibody Induced Neuronal Cell Death in a Concentration- and Time-Dependent MannerAs shown in Fig. 4A, anti-Hsp70 antibody induced neuronal cell death in a concentration-dependent manner (LC50=105 ng/mL). To rule out non-specific neurotoxicities of antibodies, we tested effects of control goat IgG on the neuronal survival. Since goat IgG did not induce neuronal cell death at 200 ng/mL, neurotoxicities of anti-Hsp70 antibody were ascribed to pl-Hsp70. Four hundred nanograms per milliliter anti-Hsp70 antibody had no effect within 24 h, but induced neuronal cell death in a time-dependent manner (Fig. 4B).
(A) Neurons (2 d-old cultures) were treated with normal goat IgG (open circles) or anti-Hsp70 antibody (closed circles) at indicated concentrations for 48 h. (B) Neurons (2 d-old cultures) were treated with normal goat IgG (open circles, 400 ng/mL) and anti-Hsp70 antibody (closed circles, 400 ng/mL) at indicated times. (C) Neurons were treated with normal goat IgG (open circles) and anti-Hsp70 antibody (closed circles) at the indicated concentrations for 48 h (7 d-old cultures). (D) Neurons were treated with normal goat IgG (open circles, 200 ng/mL) and anti-Hsp70 antibody (closed circles, 200 ng/mL) at the indicated times (7 d-old cultures). Cell viabilities were determined by MTT reducing activity. Data are expressed as the mean±S.D. (n=3). ** p<0.01, compared with control.
When neurons were cultured for 7 d, neurons matured on the monolayer of astrocytes (Supplementary Fig. 1). In the mature culture, we could not rule out the possibility that neurotoxic effects of the anti-Hsp antibody might be mediated indirectly by non-neuronal cells. On the other hand, the culture contained little non-neuronal cells such as astrocytes and microglia on day 2 after seeding (Supplementary Fig. 1).17) The concentration- (Fig. 4C) and the time-dependent neurotoxicities (Fig. 4D) of anti-Hsp70 antibody were also detected in the mature neurons as well as in the immature ones. Anti-Hsp70 antibody required a lag time to cause neuronal cell death, suggesting generation of neurotoxic agents within 24 h. To avoid indirect effects of anti-Hsp70 antibody via non-neuronal cells, immature neurons were used for detecting neurotoxic agents.
Effect of Anti-Hsp70 on Biochemical Apoptotic FeaturesWe ascertained whether anti-Hsp70 antibody condensed chromatin, a characteristic feature of apoptosis. Neurons were stained with Hoechst33342 fluorescent dye (Supplementary Fig. 2). In control and the anti-Hsp70 antibody-treated cultures, cells showed little fluorescence in the nucleus. The antibody did not condense chromatin. Next, we ascertained whether anti-Hsp70 antibody caused activation of caspase-3, another apoptotic feature (Fig. 5A). We confirmed that 15d-PGJ2 activated caspase-3 significantly. Caspase-3 activity in goat IgG-treated neurons was similar to that in control. In contrast, the anti-Hsp70 antibody did not activate caspase-3 at 6 h (Fig. 5A: upper panel) and inhibited it at 48 h (Fig. 5A: lower panel).
(A) Neuronal caspase-3: Neuronal caspase-3 activity was assayed after the treatment with normal goat IgG (200 ng/mL), anti-Hsp70 antibody (200 ng/mL) or 15d-PGJ2 (3.5 µM) for 6 h (upper panel) or 48 h (lower panel). Data are expressed as the mean±S.D. (n=3). ** p<0.01, compared with control. (B) Neuronal accumulation of ubiquitinated proteins. Neurons were treated without or with normal goat IgG (200 ng/mL), anti-Hsp70 antibody (200 ng/mL) or 15d-PGJ2 (3.5 µM) for 36 h. Ubiquitinated proteins of neuronal cell lysates were separated by SDS-PAGE, and detected by Western blotting with anti-ubiquitin antibody (upper panel). Ubiquitinated proteins were quantified by the luminoimage analyzer LAS3000 (lower panel), according to the manufacturer’s instructions (GE Healthcare Science).
15d-PGJ2 activates caspases and inhibits UPP.25) Ub proteins are distributed evenly throughout the cytoplasm in neurons, while large perinuclear aggregates containing Ub-proteins are frequently observed in neurons treated with 15d-PGJ2.31) To investigate if the anti-Hsp70 antibody affected the accumulation of Ub proteins in cortical neurons, Western blots of total lysates from rat primary cultures were probed with a specific antibody to detect ubiquitin protein conjugates (Fig. 5B). Ub-proteins in goat IgG-treated neurons were detected similarly to those in control neurons. We confirmed that 15d-PGJ2 increased the accumulation of Ub proteins in neurons. However, the anti-Hsp70 antibody stimulated the degradation of Ub proteins, indicating that it activated UPP.
Catalase Rescued Neurons from Anti-Hsp70 Antibody-Induced Neuronal Cell DeathReactive oxygen species (ROS) including hydrogen peroxide can contribute to the neurotoxicity. To ascertain whether ROS were involved in anti-Hsp70 antibody-induced neuronal cell death, we examined effects of antioxidants on the neurotoxicity of anti-Hsp70 antibody. N-Acetyl cystein and glutathione contain thiol groups, reduce hydroxyl radicals and rescue neurons from 15d-PGJ2-induced apoptosis.32) Vitamin E reduces hydroxyl radicals, singlet oxygen, and neurotoxicity of amyloid β.33) However, we could not detect the neuroprotective effects of N-acetyl cystein, glutathione and vitamin E on anti-Hsp70 antibody-induced neuronal cell death. Among tested antioxidants, catalase suppressed the neurotoxicity of anti-Hsp70 antibody significantly. Catalase protected neurons from the antibody-induced cell death in a concentration-dependent manner (Fig. 6A).
(A) Concentration: Cortical neurons were treated with catalase at indicated concentrations in the presence of normal goat IgG (open circles: 200 ng/mL) or anti-Hsp70 antibody (closed circles: 200 ng/mL). Data are expressed as the mean±S.D. (n=3). ** p<0.01, compared with normal goat IgG. # p<0.05, ## p<0.01, compared with anti-Hsp70 antibody alone. (B) Time course: Cortical neurons were treated with normal goat IgG (open column; 200 ng/mL) or anti-Hsp70 antibody (closed column; 200 ng/mL). Then, neurons were applied with 400 U/mL catalase at the indicated times. Data are expressed as the mean±S.D. (n=3). ** p<0.01, compared with normal goat IgG. ## p<0.01, compared with anti-Hsp70 antibody alone. (C) Catalase prevented the anti-Hsp 70 antibody-induced hydrogen peroxide generation. Cortical neurons were treated with catalase at the indicated concentrations in the absence (circles) or presence of 400 ng/mL normal goat IgG (triangles) or 400 ng/mL anti-Hsp70 antibody (squares). Catalase was added at the indicated concentrations for 12 h. Hydrogen peroxide were quantified at 12 h by Red Hydrogen Peroxide Assay Kit. Data are expressed as the mean±S.D. (n=3). * p<0.05, ** p<0.01, compared with catalase (—).
The anti-Hsp70 antibody required a lag time to exhibit its neurotoxicity. Did catalase suppress the neurotoxicity by post-treatment as well as co-treatment? We evaluated neuroprotective effects of catalase at indicated times after treatment of neurons with anti-Hsp70 antibody (Fig. 6B). The neuroprotective effect of catalase was significantly detected by post-treatment within 24 h, although it was decreased in a time-dependent manner. Catalase catalyzes decomposition of hydrogen peroxide to water and oxygen. Hydrogen peroxide causes neurotoxicity in primary cultures of rat cortical neurons.34) Since anti-Hsp70 antibody had been suggested to generate neurotoxic agents during a lag time, we evaluated intracellular hydrogen peroxide in the antibody-treated neurons within 24 h. Anti-Hsp70 antibody elevated the level of hydrogen peroxide at 12 h (Fig. 6C). Catalase also suppressed the generation of hydrogen peroxide from anti-Hsp70 antibody-treated neurons in a concentration-dependent manner. Contrary to our expectation, normal goat IgG also increased hydrogen peroxide in spite of little neurotoxicity. Thus, we could not detect the significant difference between treatments with normal goat IgG and anti-Hsp70 antibody.
A JNK Inhibitor Prevented Neurons from Undergoing the Anti-Hsp70 Antibody-Induced Neuronal Cell Death PartiallyThe cascade of mitogen-activated protein kinases (MAPKs) is the downstream of hydrogen peroxide.35) There are three MAPKs, JNK, extracellular signal-regulated kinase (ERK) and p38-mitogen activated protein kinase (p38-MAPK). To target pl-Hsp70, but not cytosolic Hsp70, an anti-Hsp70 antibody was applied without permeabilization for 48 h in the primary culture of rat cortical neurons. Each inhibitor alone did not affect neuronal cell survival and neuronal morphology. U0126, an ERK inhibitor, and SB202190, a p38-MAPK inhibitor enhanced neuronal cell death (Fig. 7A) and enhanced shortening of neurites and atrophy of neuronal cell bodies. On the other hand, a JNK inhibitor, SP600125, significantly prevented neurons from undergoing anti-Hsp70 antibody induced neuronal cell death (Fig. 7A). A JNK inhibitor suppressed the neurotoxicity of anti-Hsp70 antibody in a concentration-dependent manner (Fig. 7B). Next, we ascertained whether the anti-Hsp70 antibody phosphorylated JNK by Western blotting with anti-JNK and anti-phosphorylated JNK antibodies (Fig. 7C). In the brain, JNKs are alternatively spliced to yield proteins, which nearly all fall into two molecular weight categories of 54 and 46 kDa.36) The two masses of proteins having 54 and 46 kDa were detected as the total JNK by the anti-JNK antibody, which recognizes phosphorylated and unphosphorylated JNKs (Fig. 7C). The total JNK was not significantly altered among control, normal goat IgG- and anti-Hsp70 antibody-treated neurons. Phosphorylated JNK (54, 46 kDa) was specifically detected by the anti-phosphorylated JNK antibody. Phosphorylation of the anti-Hsp70 antibody-treated JNK appeared to be higher than that of control. The JNK inhibitor appeared to restore shortening of neurites and atrophy of neuronal cell bodies partially. However, we have not succeeded in detecting the significant difference between normal IgG and anti-Hsp70 antibodies-phosphorylated JNK. The contribution of JNK pathway to the neurotoxicity of anti-Hsp70 antibody might be too small (about 20%) to detect its activation.
(A) Neurons were treated with 5 µM U0126, 5 µM SB202190 or 5 µM SP600125 in the absence (open columns) or presence of anti-Hsp70 antibody (closed columns; 200 ng/mL). (B) Neurons were treated with SP600125 at the indicated concentrations in the absence (open circles; control) or presence of normal goat IgG (open triangles; 200 ng/mL) or the anti-HSP antibody (open squares; 200 ng/mL). Cell viabilities were determined by MTT reducing activity. Data are expressed as the mean±S.D. (n=3). ** p<0.01, compared with the normal goat IgG alone and # p<0.05, ## p<0.01, compared with the anti-Hsp70 antibody alone. (C) Neurons were treated with normal goat IgG (200 ng/mL) or the anti-HSP antibody (200 ng/mL) for 15 min. Phosphorylated JNK was specifically detected by the anti-phosphorylated JNK antibody. Both phosphorylated and unphosphorylated JNK are detected as the total JNK by the anti-JNK antibody.
Morphologic alterations were evaluated according to morphologic criteria; neurons with intact neurites and a smooth, round soma were considered viable, whereas those with degenerated neurites and an irregular soma were considered nonviable. To target pl-Hsp70, but not cytosolic Hsp70, anti-Hsp70 antibody was applied without permeabilization for 48 h in primary cultures of rat cortical neurons. In control (Fig. 8A) and goat IgG-treated cultures (Fig. 8B), neurons had extended neurites and smooth, round cell bodies. On the other hand, anti-Hsp70 antibody shortened neurites, but did not fragment. After 24 h, the antibody shrank neuronal cell bodies, lost bright phase-contrast appearance and reduced neurons in a time-dependent manner (Fig. 8C). Catalase did not affect morphologies of control (Fig. 8D) and goat IgG-treated neurons (Fig. 8E), whereas it inhibited almost completely the anti-Hsp70 antibody-shortened neurites and -shrunk neuronal cell bodies (Fig. 8F). In comparison with the anti-Hsp70 antibody, 15d-PGJ2 shortened and fragmented neurites. As well as the anti-Hsp70 antibody, it shrank and disrupted neuronal cell bodies at 3 µM (Figs. 8G, H). 15d-PGJ2 exacerbated the anti-Hsp70 antibody-induced morphological neurodegeneration at 3 µM (Fig. 8I).
Neurons were treated without (A, D, G) or with 100 ng/mL goat IgG (B, E, H) or 100 ng/mL anti-Hsp70 antibody (C, F, I). Neurons were treated without (A, B, C) or with 100 u/mL catalse (D, E, F) and 3 µM 15d-PGJ2 (G, H, I). Cortical neurons were photographed by phase-contrast microscopy 48 h later. Scale bar=100 µm.
In the present study, we provided the first evidence that Hsp70 was also detected on the neuronal cell surface. By the immunostaining without permeabilization and fixation, the dotted staining pattern at neuronal cell bodies and neurites might also reflect the localization of membrane-bound Hsp70 in lipid rafts.5) Our observation was supported by the report that the immunoreactivity of Hsp70 family was detected on the synaptic plasma membrane by immunoelectron microscopic analysis.13) The basal level of Hsp70 in neurons is higher than those in non-neural tissues, whereas the inducibility of Hsp70 in neurons is lower than those in non-neural tissues.8,9) Although stimuli such as stress was absent, pl-Hsp70 was present on the neuronal cell surface.
The anti-Hsp70 antibody induced neuronal cell death, suggesting the impairment of neuroprotective function of Hsp70.37) Since inactivated serums were used in the present culture, the anti-Hsp70 antibody did not require complement to induce neuronal cell death. Commercially available antibodies contain some additive agents, such as sodium azide and albumin, suggesting that they may affect the cell viability. To rule out the possibility, we tested normal goat IgG, which were solved in the solution same as that of the anti-Hsp70 antibody. No neurotoxicity of normal goat IgG indicated that the neurotoxicity of anti-Hsp70 antibody was not ascribed to additive agents. The anti-Hsp70 antibody elevated the intracellular level of hydrogen peroxide during the lag time required to elicit neuronal cell death despite of insignificant increase compared with normal goat IgG. Post-treatment of neurons with catalase after the application of anti-Hsp70 antibody conferred neuroprotection as well as the co-treatment protocol. Combined with the neurotoxic effect of hydrogen peroxide,38) the neuroprotective effect of catalase suggested that the anti-Hsp70 antibody caused neuronal cell death via hydrogen peroxide.
The neurotoxicity of anti-Hsp70 antibody was detected in both immature and mature neurons. When serum was included in the culture medium, neurons matured on a monolayer of glial cells in 7 d-old cultures.17) On the other hand, the 2 d-old cultures contained >95% neurons and few glial cells (Supplementary Fig. 4). The LC50 of anti-Hsp70 antibody in mature neurons (131 ng/mL) was greater than in immature neurons (105 ng/mL). The rate of anti-Hsp70 antibody induced-neuronal cell death in mature neurons was slightly lower than that in immature neurons. The neurotoxicity of anti-Hsp70 antibody was not related to non-neuronal cell growth, suggesting that the neurotoxicity is not ascribed to non-neuronal cells. Thus, anti-Hsp70 antibody acted on neurons directly rather than indirectly via non-neuronal cells.
Neither N-acetyl cysteine, glutathione nor vitamin E prevented neurons from anti-Hsp70 antibody-induced cell death (data not shown). On the other hand, catalase suppressed the neurotoxicity of anti-Hsp70 antibody. The enzyme reduces hydrogen peroxide-induced protein oxidation, DNA damage, mitochondrial membrane transition pore opening and loss of cell membrane integrity. Catalase protected cultured neurons from hydrogen peroxide-induced oxidative stress. How could an enzyme suppress the production of intracellular hydrogen peroxide, when it was too large to diffuse into the cellular cytoplasm? The plasma membrane permeability coefficient for hydrogen peroxide 2×10−4 cm s−1 and the delay establishing a stable gradient was calculated to be approximately 0.9 s, indicating that hydrogen peroxide diffuses rapidly across the plasma membrane.39) Thus, intracellular hydrogen peroxide appeared to diffuse to the extracellular space, where it was reduced in the presence of catalase.
The anti-Hsp70 antibody did not induce the production of hydrogen peroxide immediately, suggesting neuroprotective effects of catalase on the neurotoxicity of anti-Hsp70 antibody by the post-treatment. Actually, post-treatment of neurons with catalase after the application of anti-Hsp70 antibody conferred neuroprotection as well as the co-treatment protocol. Reduced nicotinamide adenine dinucleotide phosphate (NADPH) oxidase has been reported to be involved in the generation of hydrogen peroxide in primary cortical neurons.40) However, NADPH oxidase inhibitors (apocynin, diphenylene iodonium and VAS2870) could not prevent neurons from undergoing the anti-Hsp70 antibody-induced hydrogen peroxide generation and cell death (data not shown). The neuroprotective effect of other inhibitors of hydrogen peroxide-generating enzymes including monoamine oxidase B and xanthine oxidase could not be detected (data not shown). Anti-Hsp70 antibody inhibits the NADH+-dichlorophenol-indophenol oxidoreductase activity of trans-plasma-membrane oxidoreductases, suggesting the down-regulation of ROS reducing activity.12) Since ROS can be generated as byproducts of intracellular enzymatic reactions, anti-Hsp70 antibody might suppress the reductase activity of trans-plasma-membrane oxidoreductases for a period equivalent to the lag time, thereby allowing hydrogen peroxide levels to become elevated as a byproduct. Further studies are required to clarify the mechanism underlining the ability of anti-Hsp70 antibody to cause hydrogen peroxide accumulation after a lag time.
The elevated level of anti-Hsp70 IgG antibody is an independent risk factor for stroke.20) In the stroke, arachidonate cascade including secretory phospholipase A2 (sPLA2) was stimulated.41) Among arachidonate metabolites generated by sPLA2, 15d-PGJ2 can be contributed to the neurotoxicity of sPLA2s.17,18) 15d-PGJ2 is increased progressively after ischemia-reperfusion,42) and exhibits opposite actions as a neuroprotectant and a neurotoxicant in the ischemic brain. Although PPARγ is involved in the neuroprotective effects of 15d-PGJ2, no receptor has been related to its neurotoxicity. 15d-PGJ2 induces neuronal cell death via apoptosis.23,24) Apoptosis is accompanied with plasma membrane blebs, caspase activation and chromatin condensation. The anti-Hsp70 antibody did not condense chromatin (Supplementary Fig. 2). Neurotoxicities of anti-Hsp70 antibody and 15d-PGJ2 were additive, suggesting that the anti-Hsp70 antibody caused neuronal cell death in a different fashion from apoptosis. 15d-PGJ2 activates caspase, whereas it inhibits UPP and accumulates Ub proteins.25) 15d-PGJ2 regulates reciprocally the two proteolytic pathway of caspase and proteasome. Inhibition of caspase results in suppressing the neurotoxicity of 15d-PGJ2.24) In contrast, the anti-Hsp70 antibody activated proteasome and inhibited caspase, reciprocally. To our knowledge, neither necrosis, apoptosis nor autophagy is accompanied with the stimulation of proteasome and the suppression of caspase. Thus, the anti-Hsp70 antibody caused neuronal cell death in a novel fashion.
Increases in oxidative stress are regarded as a sign of stroke pathophysiology.41) sPLA2 enhances ROS production43) and stimulates oxidative signaling pathways in the stroke brain.44) 15d-PGJ2 increases ROS at neurotoxic concentrations.21) Similarly, anti-Hsp70 antibody generated ROS in a concentration-dependent manner. The plasmalemmal Hsp70 is targeted for autoantibodies against Hsp70, which are upregulated in neurologic diseases such as stroke. The targeted Hsp70 might contribute to the generation of neurotoxic ROS and neuronal cell death. To prove this hypothesis, it is required to ascertain whether anti-Hsp70 autoantibodies recognize the cell surface Hsp70 and cause cell death via ROS in human neurons or not.
Hydrogen peroxide can activate the MAPK cascade.35) The ERK pathway is one of several neuroprotective mechanisms that are activated by stress to counteract death signals in central nervous system.45) In addition, p38-MAPK also mediates neuroprotective effects.46) Inhibition of ERK and p38-MAPK resulted in exacerbating the neurotoxicity of anti-Hsp70 antibody. On the other hand, a JNK inhibitor significantly attenuated toxicity of anti-Hsp70 antibody. JNK is activated and involved in neurotoxicities of hydrogen peroxide.38) In conclusion, we provided the first evidence that pl-Hsp70 was localized on the neuronal cell surface. When pl-Hsp70 was targeted with the antibody, neuronal cell death could be induced by hydrogen peroxide. Among the downstream of hydrogen peroxide, JNK appeared to be partially involved in the neurotoxicity of anti-Hsp70 antibody. The anti-Hsp70 antibody stimulated the proteasome activity and suppressed the caspase activity. To our knowledge, this neuronal cell death targeted for pl-Hsp70 was caused in a different fashion from the known death such as necrosis, apoptosis and autophagy.
The authors thank Mr. Eisuke Ohta (Himeji Dokkyo University) and Dr. Minoru Takahashi (Himeji Dokkyo University) for their experimental supports. We are grateful for salutary advice of Prof. Tsutomu Nakamura (Osaka University of Pharmaceutical Sciences) and the dedicated efforts of Prof. Kenkichi Takase (Jichi Medical University). The work presented in the submitted manuscript was funded by Himeji Dokkyo University.
The authors declare no conflict of interest.
The online version of this article contains supplementary materials.
Supplementary Fig. 1. Neurons and glial cells in primary cultures of rat cortical neurons. Cells were cultured for 2 d (A, C, E) and 7 d (B, D, F). Morphologies were examined by phase-contrast microscopy (A, B). Anti-microtuble-associated protein 2 (MAP2) antibody and anti-glial fibrillary acidic protein (GFAP) antibody were obtained from Sigma (St. Louis, MO, U.S.A.). Cortical cultures were immunostained with anti-MAP2 antibody (titer: 1 : 500) specific for neurons (C, D) and anti-GFAP antibody (titer: 1 : 100) for glial cells (E, F). Immunostained neurons and glial cells were detected with ABC-PO kit. On day 2, the present cultures contained neurons at least 95%. Scale bar=100 µm.
Supplementary Fig. 2. Anti-Hsp70 antibody condensed chromatin. Neurons were treated with normal goat IgG and anti-Hsp70 antibody at 200 ng/mL for 48 h (2 d-old cultures). Hoechst33342-stained neurons were photographed (a). Scale bar=100 µm.
