Early release of high-mobility group box 1 (HMGB1) from neurons in experimental subarachnoid hemorrhage in vivo and in vitro

Abstract

Background: Translocation of high-mobility group box 1 (HMGB1) from nucleus could trigger inflammation. Extracellular HMGB1 up-regulates inflammatory response in sepsis as a late mediator. However, little was known about its role in subarachnoid hemorrhage-inducible inflammation, especially in the early stage. This study aims to identify whether HMGB1 translocation occurred early after SAH and also to clarify the potential role of HMGB1 in brain injury following SAH.

Methods: Sprague-Dawley (SD) rats were randomly divided into sham group and SAH groups at 2 h, 12 h and on day 1, day 2. SAH groups suffered experimental subarachnoid hemorrhage by injection of 0.3 ml autoblood into the pre-chiasmatic cistern. Rats injected by recombinant HMGB1(rHMGB1) solution were divided into four groups according to different time points. Cultured neurons were assigned into control group and four hemoglobin (Hb) incubated groups. Mixed glial cells were cultured and stimulated in medium from neurons incubated by Hb. HMGB1 expression is measured by western blot analysis, real-time polymerase chain reaction (PCR), immunohistochemistry and immunofluorescence. Downstream nuclear factor kappa B (NF-κB) subunit P65 and inflammatory factor Interleukin 1β (IL-1β) were measured by western blot and real-time PCR, respectively. Brain injury was evaluated by cleaved caspase-3 staining.

Results: Our results demonstrated HMGB1 translocation occurred as early as 2 h after experimental SAH with mRNA and protein level increased. Immunohistochemistry and immunofluorescence results indicated cytosolic HMGB1 was mainly located in neurons while translocated HMGB1 could also be found in some microglia. After subarachnoid injection of rHMGB1, NF-κB, downstream inflammatory response and cleaved caspase-3 were up-regulated in the cortex compared to the saline control group. In-vitro, after Hb incubation, HMGB1 was also rapidly released from neurons to medium. Incubation with medium from neurons up-regulated IL-1β in mixed glial cells. This effect could be inhibited by HMGB1 specific inhibitor glycyrrhizic acid (GA) treatment.

Conclusion: HMGB1 was released from neurons early after SAH onset and might trigger inflammation as an upstream inflammatory mediator. Extracellular HMGB1 contributed to the brain injury after SAH. These results might have important implications during the administration of specific HMGB1 antagonists early in order to prevent or reduce inflammatory response following SAH.

Figure 1

Schematic observation of the brain after surgery. Top: rat brains harvested at different times after subarachnoid hemorrhage (SAH) onset. Rat brains in the sham group were harvested one day after the surgical procedure (A). We can conclude that blood clot in the subarachnoid space gradually disappeared with time (B to E). Bottom: rat brains harvested at different times after subarachnoid injection of recombinant High-mobility group box 1 (rHMGB1) saline (G to J). Rats in the control group were injected with 0.9% saline, and sacrificed one day after the surgical procedure (F).

Figure 2

Representative immunohistochemistry analysis of High-mobility group box 1 (HMGB1) in the brain cortex. (A) HMGB1 expression in the sham group: HMGB1 could be observed in the nucleus in the sham group, cells were seldomly stained positive for cytosolic HMGB1. (B, C) HMGB1 staining in the 2-h and day-1 groups post subarachnoid hemorrhage (SAH), respectively. The quantity of cells positive for cytosolic HMGB1 obviously increased in the SAH groups at 2 h and day 1 after SAH. (D) Black boxes show the view of observation. (E, F and G) Enlarged images of A, B and C, respectively. Positive cells were defined as presenting buffy grains in cytoplasm as shown by arrows. Scalar bars present 20 μm. (H) Quantification of density of cells positive for cytosolic HMGB1. Values were obtained from averaging three sections per animal. SAH induced translocation of HMGB1 in the cortex near the subarachnoid space as early as 2 h post SAH. Bars represent the mean ± standard error (n = 6); ^#P < 0.01 compared with the sham group.

Figure 3

Cytosolic expression of high-mobility group box 1 (HMGB1) in Neuron-specific nuclear protein(NeuN)-positive cells of cortex from animals with subarachnoid hemorrhage (SAH). (A, B, C) HMGB1 immunostaining images obtained from the cortex of a sham (A), a 2-h SAH (B), and a day-1 (C) animal respectively. (D, E, F) Merged images of HMGB1 immunostaining (red) and 4',6-diamidino-2-phenylindole(DAPI) nuclear staining (blue). (G, H, I) Merged images of HMGB1 immunostaining (red) and NeuN immunostaining (green). Enlarged images on the right side of each panel highlight the increased number of NeuN-positive cells with cytosolic staining of HMGB1 in cortex from SAH animals. Arrows indicate the cytosolic HMGB1. The mark (>) indicates co-localization of cytosolic HMGB1 and NeuN, The results indicate that HMGB1 translocation occurred as early as 2 h and advanced in the process after SAH. Scale bars represent 20 μm.

Figure 4

Cytosolic expression of High-mobility group box 1 (HMGB1) in Glial fibrillary acidic protein (GFAP)-positive cells of cortex from animals with subarachnoid hemorrhage (SAH). (A, B) HMGB1 immunostaining images obtained from cortex of a sham (A), and a day-1 (B) animal respectively. (C, D) Merged images of HMGB1 immunostaining (red) and 4',6-diamidino-2-phenylindole (DAPI) nuclear staining (blue). (E, F) Merged images of HMGB1 immunostaining (red) and GFAP immunostaining (green). Enlarged images (a, b, c, d, e, f) on the right side of each panel also highlight the increased number of cytosolic HMGB1-positive cells in cortex from SAH animals while few cytosolic HMGB1-positive cells were also positive for GFAP. Arrows indicated the cytosolic HMGB1. The results indicated that astrocytes were not the main source of released HMGB1, at least at this early time point. Scale bars represent 20 μm.

Figure 5

Cytosolic expression of High-mobility group box 1 (HMGB1) in ionized calcium binding adaptor molecule 1 (Iba1)-positive cells of cortex from animals with subarachnoid hemorrhage (SAH). (A, B) HMGB1 immunostaining images obtained from cortex of a sham (A) and a day-1 SAH (B) animal. (C, D) Merged images of HMGB1 immunostaining (red) and 4',6-diamidino-2-phenylindole (DAPI) nuclear staining (blue). (E, F) Merged images of HMGB1 immunostaining (red) and Iba1 immunostaining (green). Enlarged images (a, b, c, d, e, f) on the right side of each panel highlight the increased number of cytosolic HMGB1-positive cells (arrows) in cortex from SAH animals and some cytosolic HMGB1-positive cells were also positive for Iba1(shown as >). Scale bars represent 20 μm. Arrows indicat the cytosolic HMGB1. The mark (>) indicates co-localization of cytosolic HMGB1 and Iba-1. The results indicate that microglia were also a source of extracellular HMGB1. Scale bars represent 20 μm.

Figure 6

Expression of High-mobility group box 1 (HMGB1) protein and mRNA level in brain cortex after subarachnoid hemorrhage (SAH). (A) Western blot analysis of HMGB1 expression in the total protein extraction of the cortex after SAH showed that the HMGB1 protein level was significantly increased as early as 2 h and peaked approximately on day 1. (B) Western blot analysis of HMGB1 in the cytosolic protein fraction of the cortex. The quantity of HMGB1 protein in cytosolic protein fraction was also detected to increase in the 2-h, 12-h, day-1 and day-2 groups, which indicated that the HMGB1 translocation occurred as early as 2 h after SAH onset. (C) Real-time PCR analysis of HMGB1 mRNA level demonstrated the HMGB1 mRNA level was also up-regulated as early as 2 h after SAH. Bars represent the mean ± standard error (n = 6, each group). ^#P < 0.01 compared with the sham group; *P < 0.05 compared with sham group.

Figure 7

Representative photomicrographs show brain cortex double-staining of immunofluorescent (High-mobility group box 1 (HMGB1), green) and PI staining (red) in the sham (A - E), 2-h (F - J), day -1 (K - O), and another day-1 (P - T) group. The nucleus was counterstained with 4',6-diamidino-2-phenylindole (DAPI) (blue) in the same view in each section. (D, I, N, S) Merged images of HMGB1(green) and DAPI (blue); (E, J, O, T) merged images of propium iodide (PI) (red) and HMGB1 (green). Compared with the sham group (D), the subarachnoid hemorrhage (SAH) groups (I, N, S) show translocation of HMGB1 from nuclear to cytoplasm (the mark (>) indicates the cytosolic HMGB1). Few PI-positive cells could be observed in the sham group (B, E) while patches of cells positive for PI could be detected in the SAH groups (G, L). Overlapping images (J, O) showed that most cells were positive for both cytosolic HMGB1 and PI staining (arrows indicate the co-localization of cytosolic HMGB1 and PI). However, positive staining for cytosolic HMGB1 but not positive for PI staining could also easily found in the day-1 group post SAH (T). Scale bar: 20 μm. These results support the theory that both passive and active release of HMGB1 are involved in the translocation process.

Figure 8

Addition of recombinant High-mobility group box 1 (rHMGB1) in the subarachnoid space triggered inflammatory response in vivo. NF-κB was measured by the western blot of its P65 subunit in the nucleus. Toll-like receptor (TLR)4 protein level was also detected by western blot analysis. IL-1β was measured by real-time PCR. (A) rHMGB1 up-regulated P65 subunit protein level in the nuclear protein in cortex cells; P < 0.05 between the 12-h, day-1 group and control group, P < 0.01 between the day-2 group and control group. (B) rHMGB1 increased TLR4 protein level. P < 0.05 between the 12-h and control group, P < 0.01 between the day-1, day-2 group and control group. (C) rHMGB1 upregulated IL-1β mRNA expression in cortex cells: P < 0.05 between the 12-h, day-2 group and control group, P < 0.01 between the day-1 group and control group. (D) Western blot analysis of histone 3 content in rHMGB1(left lane) and the nuclear protein extraction(right lane). The result could exclude histone 3 contamination in rHMGB1 products. Histone 3 was predominant composition of histone protein, thus, our result indicated the rHMGB1 used in the study had good purity. Bars represent the mean ± standard error (n = 6): *P < 0.05 compared with the control group; ^#P < 0.01 compared with the control group.

Figure 9

Representative photomicrographs showed brain neurons double immunofluorescent staining for cleaved caspase 3 (red) and Neuron-specific nuclear protein (NeuN), a neuron cell marker (green) in vivo in the control (A to D) and recombinant High-mobility group box 1 (rHMGB1) treatment group (E to H). The nucleus was counterstained with 4',6-diamidino-2-phenylindole (DAPI) (blue) in the same view in each section. (D, H) Merged images of cleaved caspase 3 (red) and NeuN(green). Compared with the control group (A), more cleaved caspase 3-positive cells were detected in the cortex after rHMGB1 treatment (E). Especially, overlapping images (H) showed that the number of cells positive both for cleaved caspase 3 and NeuN increased compared with control group (D). The marks (>): profiles positive for cleaved caspase 3 and DAPI but negative for NeuN showed activation of caspase 3 in non-neuronal cells. Arrows: profiles positive for cleaved caspase 3 and NeuN showed activation of caspase 3 in neurons. Scale bar: 20 μm. These results indicate that rHMGB1 addition increased the cleaved-caspase 3 positive cells, especially the neurons. RHMGB1 might promote the cell apoptosis.

Figure 10

Representative photomicrographs showed brain neurons double immunofluorescent staining for High-mobility group box 1 (HMGB1) (red) and Neuron-specific nuclear protein (NeuN), a neuronal cell marker (green) in vitro in the control (C, E) and Hb incubation group (D, F). (A) Immunofluorescence staining shows more than 98% cells were positive for both Neuron and 4',6-diamidino-2-phenylindole (DAPI), which suggests the high percentage of Neuron cells in the primary cultured cells. (B) Result of western blot analysis of concentrated conditioned culture media, which showed that HMGB1 could be detected in the media. (C, D) Images of cultured neurons in light microscopy to watch its sharp; light micrograph of neurons shows cellular morphology in sham group (C) and Hb incubation group (D). (E, F) Merged images of HMGB1(red) and DAPI (blue) in cultured neurons. Compared with the sham group (E), the Hb incubation groups (F) showed translocation of HMGB1 in cytoplasm (as shown by white arrows). Scale bar: 20 μm.

Figure 11

Conditioned medium from Hb-treated neurons induced IL-1β in cultured mixed glial cells, which could be inhibited by High-mobility group box 1 (HMGB1)-specific inhibitor. Cultured mixed glial cells were arranged into three groups. In the control group mixed glial cells were treated with control medium; in the medium group mixed glial cells were treated with neuron medium; in the medium + glycyrrhizic acid (GA) group, after mixed glial cells were treated with neuron medium, a special inhibitor of HMGB1 (GA) was added in the medium to silence the activity of HMGB1. IL-1β was measured by real-time PCR. Bars represent the mean ± standard error (n = 6), *P < 0.05 compared with the medium group; ^#P < 0.01 compared with the control group.