Different TLR4 expression and microglia/macrophage activation induced by hemorrhage in the rat spinal cord after compressive injury

Abstract

Background: Hemorrhage is a direct consequence of traumatic injury to the central nervous system and may cause innate immune reactions including cerebral Toll-like receptor (TLR) 4 upregulation which usually leads to poor outcome in the traumatic brain injury. In spinal cord injury (SCI), however, how hemorrhage induces innate immune reaction in spinal parenchyma remains unknown. The present study aimed to see whether blood component and/or other factor(s) induce TLR4 and microglia/macrophages involved innate immune reactions in the rat spinal cord after traumatic injury.

Methods: Using the compressive SCI model of the rat, hemorrhage in the spinal cord was identified by hematoxylin-eosin staining. Microglia/macrophage activation, TLR4 expression, and cell apoptosis were investigated by immunohistochemistry. Nuclear factor (NF)-κB p50 level of the two segments of the cord was detected by western blotting assay. With carbon powder injection, blood origination of the hematoma was explored. The blood-spinal cord barrier (BSCB) states of the lesion site and the hematoma were compared with immunohistochemistry and tannic acid-ferric chloride staining.

Results: Histological observation found blood accumulated in the center of compression lesion site (epicenter) and in the hematoma approximately 1.5 cm away from the epicenter. TLR4 expression, microglia//macrophage activation, and subsequent apoptosis in the area of far-away hematoma were late and weak in comparison to that in epicenter. In addition, TLR4 positive microglia/macrophages appeared to be phagocytotic in the far-away hematoma more obviously than that in the epicenter. Injected carbon powder indicated that accumulated blood of the far-away hematoma originated from the bleeding of the lesion epicenter, and the BSCB around the hematoma was not compromised in the early phase. Accordingly, at 3 days post injury, NF-κB p50 was upregulated based on the similar levels of blood component hemoglobin, and cell apoptosis was obvious in the epicenter but not in the far-away hematoma.

Conclusion: These data suggest that besides blood component, BSCB compromise and the extent of tissue injury contribute more to TLR4 and microglia/macrophage responses to the spinal cord hemorrhage. Therefore, the innate immune environment is a necessary consideration for the SCI therapy targeting TLR4 and microglia/macrophages.

Figure 1

The device for the compressive spinal cord injury model of the rat. (A) The metal tube and the rod with a plastic plate attached onto the tip. (B) The rod is held by the tube which could be connected with the stereotaxic apparatus. (C) Schematic diagram of compressive injury to the spinal cord by the plastic plate attached to the rod tip, (a) front view and (b) side view.

Figure 2

Compressive spinal cord injury caused distal hematoma far away from the lesion site. (A) A segment of rat spinal cord taken out immediately after compression. Aster shows the compressive point, and the arrow shows the distal hematoma which could be seen grossly. (B-D) H-E staining of the spinal cord sections indicated blood in the compressive lesion center and far-away hematoma which forms spindle-shaped foci at 6 h, 3 days, and 14 d post injury, respectively. (E-G) Higher magnification of the boxes in the pictures above accordingly. Note that cavity formed from the hematoma at 14 days post injury. Bar = 1 mm (A-D), 200 μm (E-G).

Figure 3

Blood origination of the far-away hematoma. (A) H-E staining pictures show that there is no hematoma far away from lesion site in the spinal cord with semi-transection (arrow) aside the lesion. (B) Distal hematoma contained carbon powder injected to the compression center immediately after injury, indicating that blood originated from the lesion site. (C) and (D) are higher amplified pictures of the boxes located in the lesion site (indicated by letter L) and distal hematoma (indicated by letter H) in (B), showing carbon powders (arrows) inside the hemorrhagic foci. Bar = 2 mm in (A) and (B). Bar = 50 μm in (C) and (D).

Figure 4

Microglia/macrophage activation in the spinal area adjacent to the lesion site but not around the distal hematoma at 3 days post injury. (A) Montage picture of longitudinal spinal cord section for IBa-1 immunolabeling, showing lesion site (indicated by letter L) and hematoma (indicated by letter H) at 3 days post injury. (B) Amplified picture of the box adjacent to the lesion site, showing many cells are labeled with both IBa-1 (red), the marker of microglia, and ED-1(green), the marker of activated microglia. (C) Amplified picture of the box adjacent to the far-away hematoma in which only IBa-1 positive cells with thin processes could be seen. Bar = 2 mm in (A). Bar = 50 μm in (B) and (C).

Figure 5

Immunofluorescent labeling for TLR4 and IBa-1 in the lesion site and the hematoma at 3 days post injury. In the lesion site (A-C), there were abundant IBa-1 positive cells, quite part of which express TLR4, showed by confocal microscopy. Most of the microglia/macrophages were round with short or blunt processes. To the contrast, in the distal hematoma and the area adjacent to it (D-F), there were few TLR4 positive cells, and IBa-1 labeled cells were small with thin processes. TLR4 immunoreactivity: red; IBa-1 immunoreactivity: green; Hoechest 33342: blue. Bar = 200 μm.

Figure 6

Representative pictures of immunofluorescent labeling for TLR4 and IBa-1 in the far-away hematoma at 6 h, 3 days, and 14 days post injury. Note that IBa-1 positive cells are small and scattered in the area adjacent to the hematoma at 6 h (A, E, I) and appear in the hematoma at 3 days post injury (B, F, J). At 14 days post injury (C, G, K), TLR4 immunoreactive product is seen in the area of the hematoma, and abundant IBa-1 positive cells are large and phagocytotic and with TLR4 immunoreactivity inside the cytoplasm. All of these observation is similar to the lesion site at 3 days post injury (D, H, L), and the stack of scanning pictures of confocal microscopy confirmed that TLR4 immunoreactive product was located in the cytoplasma of IBa-1 positive cells (L). TLR4 immunoreactivity: red; IBa-1 immunoreactivity: green; Hoechest 33342, blue. L, lesion site; H, hematoma. Bar = 50 μm.

Figure 7

Immunofluorescent labeling for caspase-3 at 3 days and 14 days post injury. Representative microscopic photos show numbers of caspase-3 immunoreactive cells inside the lesion area at 3 days post injury (A), while only few cells are caspase-3 positive (arrow) in the area of or adjacent to hematoma (B). At 14 days post injury, there are casepase-3 positive cells in the lesion site but decreased in number (C); while in the area adjacent to the hematoma, caspase-3 immunoreactive cells are markedly increased (D), compared to that at 3 days. TLR4 immunoreactivity, red; IBa-1 immunoreactivity, green; Hoechest 33342, blue. L, lesion site; H, hematoma. Bar = 50 μm.

Figure 8

Western blotting assay for NF-κB p50, phosphorylated-IκB in the lesion segment and far-away hematoma segment of the spinal cord at 3 days post injury. (A) Representative picture of immunoblots for NF-κB p50, β-actin, phosphorylated-IκB (p-IκB), and hematoglobulin (Hb)-α, respectively. (B) Bar graph of ratio of the blot pixels for NF-κB p50 and p-IκB calibrated by Hb-α in lesion and hematoma segments, respectively (n = 4 for each group, *P <0.05). H, hematoma; L, lesion site.

Figure 9

Microvascular and capillary integrity of the spinal cord in the lesion site and the far-away hematoma. (A) Representative picture of the longitudinal spinal cord section labeled for RECA without ferric-chloride perfusion at 6 h post injury. (B) and (C) are amplified pictures of the boxes in lesion site and hematoma, respectively, showing that small blood vessels were there in the lesion site and in hematoma as well. (D) Representative picture of the longitudinal spinal cord section labeled for RECA after tannic acid and ferric-chloride perfusion, showing nearly no RECA immunoreactivity. (E) and (F) are amplified pictures of the boxes in lesion site and hematoma, respectively, showing that RECA immunolabeling only in the lesion site. Bar = 2 mm in (A) and (D), 50 μm in (B), (C), (E) and (F).

Figure 10

Representative picture of immunohistochemistry for rat immunoglobulin G (IgG) in the spinal cord at 3 days post injury. (A) IgG immunoreactive product is obvious in the lesion area (left box), while in the far-away hematoma (right box), there is little immunolabeling for IgG. (B) and (C) are higher magnification of the boxes in the areas adjacent to the lesion site and to the hematoma, respectively. Bar = 2 mm in (A), 200 μm in (B) and (C).