Human neuroblastoma cells (SH-SY5Y) were purchased from Type Culture Collection of Chinese Academy of Sciences. SH-SY5Y cells were cultured in DMEM-F12 medium (HyClone, USA) supplemented with 10% heat-inactivated new-born calf serum (Gibco, USA) in a humidified incubator with 5% CO₂ and 95% fresh air.

Total proteins were extracted from the cell lysates using RIPA lysis buffer (Cell Signaling Technology, #9806) supplemented with protease and phosphatase inhibitors (Cell Signaling Technology, #5872). Protein concentrations of the extracts were measured with BCA assay kit (Cell Signaling Technology, #7780). Afterwards, supernatant was boiled with 5× SDS loading buffer (Boster, China) for 10 min and provisionally stored at −80°C. Approximately 60 μg proteins were loaded onto SDS-PAGE, then electrophoretically transferred to PVDF (PerkinElmer, USA) membrane and blocked with 5% milk dissolved in Tris-buffered saline containing 0.1% Tween 20 (TBST) for 1 h at room temperature. The membranes were then blocked with 5% defatted milk at room temperature for 1 h and probed overnight at 4°C with the following primary antibodies: anti-HMGB1 antibody (ab79823; Abcam, Cambridge, MA, USA; 1:1000 dilution); anti-α-synuclein antibody (ab51253; Abcam, Cambridge, MA, USA; 1:500 dilution); anti-Beclin1 antibody (ab51031; Abcam, Cambridge, MA, USA; 1:1000 dilution) and anti-p62 antibody (Cell Signaling technology, NO.8025S; 1:1000 dilution); anti-LC3A/B antibody (Cell Signaling technology, NO.12741S; 1:1000 dilution); and anti-β-actin (ACTB) antibody (CW0281A; Comwin Biotech, USA; 1:3000 dilution). After incubation with horseradish peroxidase-conjugated secondary antibodies for 1 h at room temperature, the signals were detected by Kodak Professional Film Developer. Relative band intensities were quantified using Quantity One software.

SH-SY5Y cells were lysed at 4°C in ice-cold immunoprecipitation (IP) assay lysis buffer (Beyotime, China; P0013J) for 10 min, then the cells lysates were centrifuged for 3 min at 12,000 g. Concentrations of proteins in the supernatant were determined by bicinchoninic acid assay (Abcam, UK; ab207003). Then the extracted proteins were divided into two parts, one contained whole cell lysis (WCL) which would be assessed by Western blot (WB) directly and another prepared for co-immunoprecipitation (Co-IP). Besides, a negative control group was established, which contained normal cell lysates and would be incubated by IgG in Co-IP. Before Co-IP, the samples containing equal amounts of proteins were pre-cleared with protein A or protein G agarose/sepharose (Cell Signaling Technology, #9863) at 4°C for 3 h and subsequently incubated with irrelevant IgG or antibodies (2–5 μg/ml) in the presence of protein A or G agarose/sepharose beads (Cell Signaling Technology, #9863) for 2 h at room temperature or overnight at 4°C with gentle shaking. After incubation, agarose/sepharose beads were extensively washed with PBS, and proteins were eluted by boiling in 2× SDS sample buffer before SDS-PAGE electrophoresis.

Plasmid encoding HMGB1 fused with GFP was provided by Dr. George Hoppe at University of Pittsburgh and Dr. Marco Bianchi at San Raffaele University. Control DNA poly β siRNA and siRNA were purchased from Guangzhou RiboBio Co. Ltd, China, and were operated according to standard transfection protocols for cell cultures. GFP-HMGB1 or siRNA-HMGB1 was transfected into SH-SY5Y cells using Lipofectamine 2000 reagent (Invitrogen, Waltham, MA, USA) according to the manufacturer’s instructions. In accordance with the experimental protocols, the cells were prepared for subsequent treatments 48 h after the transfection procedures.

SH-SY5Y cells were seeded on 12-well plates and fixed in 4% paraformaldehyde for 10 min before detergent extraction with 0.1% Triton X-100 for 15 min at room temperature. The wells were saturated with 1% bovine serum albumin in phosphate-buffered saline (PBS) for 1 h at room temperature and processed for immunofluorescence with anti-HMGB1 antibodies followed by Fluorescein Isothiocyanate (FITC)-conjugated anti-rabbit IgG. Nuclear morphology was analyzed after staining with the fluorescent dye Hoechst 33342 (Sigma-Aldrich, USA) for 10 min. Cells were washed with PBS three times for 5 min each between incubation steps. Images were taken under a confocal microscope.

Prepared cell samples were quickly rinsed in PBS 3× and fixed with a 0.1 mol/L phosphate buffer solution (pH 7.4) containing 3% glutaraldehyde plus 2% paraformaldehyde. Next, the cell samples were fixed by 1% OsO₄ for 1 h to produce osmium black before dehydration. Sections of the cells were prepared with a diamond knife on a Reichert-Jung Ultracut-E ultra-microtome. Then the cell samples were stained with UrAc (20 min) followed by 0.2% lead citrate (2.5 min) for observation under a JEM 1011CX electron microscope (JEOL USA, Inc.). Digital images were acquired by the Advanced Microscopy Techniques imaging system.

SH-SY5Y cells were cultured in 6-well plates and treated with the indicated drug combination for 24 h after transfecting with/without HMGB1 plasmid or siRNA. The cell morphology was observed under an inverted microscope (IX70, Olympus, Japan) and recorded using a CMOS camera (Infinity Lite, USA). For Hoechst staining, cells cultured on coverslips in 12-well plates were fixed in 4% paraformaldehyde for 30 min after predesigned treatments. Hoechst 33258 (Sigma, USA) was used to stain the cell nuclei after PBS washing for 3× and incubated for 10 min before observing. The positively stained cells with nuclear deformities were counted in 10 random fields under the confocal microscope. For flow cytometry analysis, cells were washed three times with PBS after harvest and stained with Annexin V-FITC and propidium iodide (PI) according to the manufacturer’s instructions. The cells were then washed and analyzed with a FACScan flow cytometry (Becton Dickinson, Franklin Lakes, NJ, USA).

All statistical analyses were performed using SPSS 20.0 (SPSS, Chicago, IL, USA). Data are expressed as mean ± standard error (SEM). Differences among groups were analyzed by one-way analysis of variance (ANOVA) followed by Least square difference’s post hoc test. P < 0.05 was considered significantly different. All the data were obtained from three independent experiments.

Real-time PCR and WB were used to investigate the expression level of HMGB1 and α-synuclein upon rotenone exposure. Results showed that the HMGB1 expression (both at mRNA and protein level) was elevated in a concentration dependent manner. Moreover, HMGB1 protein expression pattern detected by WB was consistent with that of HMGB1 mRNA (Figures 1A,C). While in the time course of HMGB1 mRNA expression, with persistent 1 uM rotenone incubation, it revealed a constant increase which peaked at 24 h and then slightly decreased at 48 h (Figure 1B). In contrast, α-synuclein expression demonstrated a time and concentration dependent manner under rotenone treatment (Figures 1C,D).

The interaction between HMGB1 and α-synuclein was primarily investigated by laser confocal microscopy and then confirmed by Co-IP. The laser confocal microscopy indicated that HMGB1 was mainly located in the nucleus and scarcely existed in the cytoplasm in control SH-SY5Y cells (Figure 2A). Nevertheless, after 1 uM rotenone incubation for 24 h, HMGB1 expression levels were markedly elevated, with some shuttling from the nucleus to the cytoplasm (Figure 2A). Besides, the HMGB1 distribution pattern has changed as well, transforming from a granularly concentrated pattern to a muddily dispersive one (Figure 2A). Moreover, when compared with control group, HMGB1 and α-synuclein had been confirmed to form co-precipitation complex after rotenone exposure as proven by Co-IP (Figure 2B). In conclusion, the outcomes above indicated that HMGB1 could present overexpression, cytosolic translocation and interaction with α-synuclein upon rotenone exposure.

To further explore the relationship between HMGB1 and α-synuclein, wild-type HMGB1-GFP fusion protein and HMGB1-siRNA were transiently expressed in SH-SY5Y cells with or without rotenone exposure (Figure 3A). What is interesting is that HMGB1 expression level, compared with respective control group, were significantly elevated upon rotenone exposure, no matter the upward or downward manipulation of HMGB1. Moreover, the HMGB1 distribution patterns demonstrated similar changes to that identified in Figure 2A under rotenone treatment, regardless of the HMGB1 manipulation approaches (Figure 3A). The expression level of HMGB1 had also been proven to coincide with that of α-synuclein both in the context of HMGB1 manipulation and rotenone exposure. In particular, HMGB1 overexpression could increase α-synuclein expression and the positive effect could be further intensified by rotenone incubation (Figure 3B). While in the context of HMGB1 suppression, it was found that α-synuclein expression decreased as well, but partly restored by rotenone exposure (Figure 3C). Therefore, the study results above indicated that HMGB1 might regulate the expression of α-synuclein, while rotenone exposure could exert a positively intensifying effect.

Given that HMGB1 could modulate α-synuclein expression and abnormal α-synuclein aggregates were usually degraded by autophagy-lysosome pathway, whether HMGB1 could exert a modulatory effect on autophagy was thereby investigated. In comparison with control group (Figure 4A), after incubating SH-SY5Y cells with 1 μM rotenone for 24 h, it revealed autophagosome formation and mitochondrial deformation under transmission electron microscope (Figure 4B), which could be further aggravated by HMGB1 overexpression (Figure 4C). Nevertheless, the mitochondrial deformation in SH-SY5Y cells could be partly alleviated when transfected with siRNA-HMGB1, but still accompanying sporadic autophagosomes (Figure 4D).

To confirm the effect of HMGB1 on autophagy, we explored the expression changes of autophagy associated proteins including microtubule-associated protein light chain 3-II (LC3-II) and p62 upon HMGB1 manipulation (Figures 5A,B). As a result, we found that HMGB1 overexpression could remarkably augment autophagy marker protein (LC3-II and p62) expression no matter with or without rotenone pretreatment (Figure 5A). In contrast, inhibition of HMGB1 expression by siRNA could obviously reduce LC3-II and p62 expression, but partly restored by rotenone incubation (1 μM, 24 h; Figure 5B). In general, the expression level of LC-I, LC-II and p62 signified the feasibility of autophagy flux. When revealing more conversion of LC-I to LC-II and less p62 expression, it indicated feasibility of the autophagy flux; while the more conversion of LC-I to LC-II and elevated p62 expression implied normal initiation but not sustaining of autophagy flux. In conclusion, the study outcomes above indicated that HMGB1 could initiate but not sustain the autophagy flux.

To further clarify how HMGB1 modulated the autophagy flux and whether Beclin1 or Vps34 was involved in this process, we transiently transfected SH-SY5Y cells with GFP-HMGB1 or siRNA-HMGB1 and performed Co-IP to explore potential interaction. As a result, it revealed that endogenous Beclin1 normally bind to Vps34 but not HMGB1, as shown by Co-IP in control group (Figure 5C). However, after rotenone treatment, Beclin1 turned to bind to HMGB1 to form HMGB1-Beclin1 complex. Considering that HMGB1 expression could be upregulated upon rotenone treatment (Figures 1–3), it could be presumed that HMGB1 could disintegrate the Beclin1-Vps34 complex (Figure 5C). Moreover, HMGB1 overexpression maintained the HMGB1-Beclin1 interaction no matter rotenone was added or not. In contrast, when cells were transfected with siRNA to reduce HMGB1 expression, the interaction between Vps34 and Beclin1 in cytoplasm would restore (Figure 5C). Hence, these results suggested that HMGB1 had a higher affinity for Beclin1 and thereby suppress autophagy process by blocking the Beclin1-Vps34 complex formation, all of which could be positively intensified by rotenone exposure.

Morphological changes in SH-SY5Y cells were assessed by phase-contrast microscope, which indicated that HMGB1 overexpression or rotenone exposure could contribute to cellular shrinkage and nuclear structure destruction (Figures 6A,B). Hoechst 33258 staining demonstrated nuclear fragmentation, karyopyknosis, karyolysis and chromatin margination (Figure 6B, marked by arrows), which could be used to indirectly indicate cell viability (Figure 6C). Moreover, GFP-HMGB1 plasmids transfection further aggravated the cellular morphology and nuclear structure damages. Nevertheless, HMGB1 suppression could partly alleviate these phenomena (Figures 6A–C). The cell apoptosis rate was determined by flow cytometry. Data showed that rotenone treatment could induce an apoptosis rate of 12.00 ± 0.23% and HMGB1 overexpression resulted in a higher apoptosis rate of 29.24 ± 0.21%, while HMGB1 suppression pre-treatment (siRNA-HMGB1 transfection) only result in an apoptosis rate of 7.95 ± 0.14% (Figure 6D).