A leaky blood-brain barrier, fibrinogen infiltration and microglial reactivity in inflamed Alzheimer's disease brain

Abstract

This study has used immunohistochemical examination of tissue obtained from Alzheimer's disease (AD) brains and rat hippocampus injected with Abeta(1-42) peptide to determine effects of induced inflammatory reactivity on integrity of blood-brain barrier (BBB) and viability of neurons. Tissue from AD, but not non-demented, brains exhibited a diffuse pattern of staining for fibrinogen and immunoglobulin (IgG) indicative of BBB leakiness with considerable fibrinogen immunoreactivity (ir) appearing in association with Abeta deposits. Immunostaining for the endothelial cell specific glycoprotein, von Willebrand factor, showed morphological evidence for altered blood vessels in AD tissue. AD brains also demonstrated extensive areas of fibrinogen ir in association with microglial reactivity. In vivo, intra-hippocampal injection of Abeta(1-42) caused time-dependent (1-7 days after injection) increases in double staining of fibrinogen with areas of microgliosis. Two independent pharmacological strategies were employed to examine how Abeta(1-42) stimulation (7 days injection) may be linked to neurodegeneration. The defibrinogenating compound, ancrod, reduced inflammatory reactivity, levels of parenchymal fibrinogen and IgG, and was neuroprotective. These results prompted use of Abeta(1-42) plus fibrinogen as a novel in vivo inflammatory stimulus and this combination significantly enhanced inflammatory reactivity, vascular perturbations and neuronal damage compared to Abeta(1-42) alone. A second approach, using anti-Mac-1 (antibody for antigen CD11b) to block activation of microglia, was highly effective in attenuating effects of Abeta(1-42) plus fibrinogen amplification of inflammatory and vascular responses and conferred significant neuroprotection. The overall findings from study of AD tissue and in vivo in Abeta(1-42) and Abeta(1-42) plus fibrinogen stimulated rat hippocampus suggest microglial responses to promote increased extravasation of blood protein as a critical component in amplifying inflammatory reactivity and causing neuronal damage in inflamed AD brain.

Figure 1

Fibrinogen immunostaining of tissues from the entorhinal cortex of non-demented (ND) and Alzheimer’s disease (AD) brain. (A) Low (left panels, scale bar = 500 μm) and high (right panels, scale bar = 100 μm) magnifications of fibrinogen immunoreactivity (ir). The lower right panel shows a detailed view of an AD section with arrow indicating fibrinogen ir in proximity to a blood vessel. The upper right panel is a representative image of an AD section stained with fibrinogen pre-absorbed with fibrinogen antibody. (B) Quantification of fibrinogen ir in ND (n= 7 cases) and AD (n= 8 cases) brain tissue; *indicates P < 0.05. (C) Representative double staining of fibrinogen with Aβ peptide deposits. Scale bar represents 200 μm. The magnified inset in the lower right panel (scale bar: 30 μm) shows a detailed view of an AD section indicating fibrinogen ir in proximity to peptide.

Figure 2

Representative fluorescent double staining for fibrinogen, von Willebrand factor (vWF) and immunoglobulin (IgG). (A) Double immunofluorescent staining for vWF (red, left panels) and fibrinogen (green, middle panels) in non-demented (ND) and Alzheimer’s disease (AD) sections; the merged image is presented in right panels. Scale bar shown is 60 μm. (B) IgG immunoreactivity (ir) for ND/AD tissue with low magnification. Scale bar = 100 μm. (C) Quantification of IgG ir in ND (n= 7 cases) and AD (n= 8 cases) sections. *Indicates P < 0.05.

Figure 3

Representative double immunofluorescence staining of fibrinogen with microglia (HLA-DR marker) and astrocytes (GFAP marker) in the entorhinal cortex of non-demented (ND) and Alzheimer’s disease (AD) brain. (A) ND tissue (upper panels) showed low levels of fibrinogen immunoreactivity with microglia whereas AD tissue (lower panels) was characterized by extensive areas of diffuse fibrinogen deposition in association with microglia. Scale bar = 100 μm. (B) Representative staining for astrocytes in ND (upper panels) and AD (lower panels) sections. Double staining shows merged areas of markers. Scale bar = 100 μm.

Figure 4

Representative double immunohistochemical staining for microglia (HLA-DR marker) and Aβ (6F/3D marker) in the entorhinal cortex of non-demented (ND) and Alzheimer’s disease (AD) tissue. Sections were incubated with HLA-DR and Aβ followed by visualization using DAB/nickel ammonium sulphate (anti-HLA-DR, dark purple colour) or DAB (anti-Aβ, brown colour). Staining shows blood vessels (v) with AD tissue demonstrating close association between Aβ, microglia and vessel. Scale bar represents 50 μm. Note the predominant ramified and ameboid morphologies of microglia in ND and AD tissue, respectively.

Figure 5

Fibrinogen and microglia (OX-42) double immunofluorescent staining in Aβ_1-42-injected rat hippocampus. (A) In controls (7 days injection of PBS, upper panel or reverse peptide Aβ_42-1, second panel) low levels of fibrinogen (left column) and numbers of microglia (middle column) are evident. The right column shows merged staining. Subsequent panels show progressive time-dependent changes in fibrinogen/OX-42/merged immunoreactivity for 1, 3 and 7-day durations of Aβ_1-42 injection. Scale bar represents 200 μm. (B) Double staining for Aβ_1-42 and fibrinogen (7 days after injection). Scale bar = 80 μm.

Figure 6

Fibrinogen and astrocyte (GFAP) double immunofluorescent staining in Aβ_1-42-injected (7 days) hippocampus. GFAP immunoreactivity (ir; middle panels) was increased with Aβ_1-42 compared to controls (PBS and Aβ_42-1). Typical merged staining is shown in the right panels. Scale bar = 200 μm. The inset in the lower right panel presents a higher magnification showing negligible association of fibrinogen with GFAP ir astrocytes.

Figure 7

Effects of the defibrinogenating compound, ancrod, on parenchymal fibrinogen and IgG in Aβ_1-42-injected hippocampus. (A) Representative immunoreactivity (ir) for fibrinogen (upper left panel) and IgG (lower left panel) in 7 days Aβ_1-42-injected hippocampus. Effects of ancrod treatment with Aβ_1-42 are presented in the right panels. The IgG staining pattern shows well-defined blood vessels. Scale bars are 100 μm (for fibrinogen) and 50 μm (for IgG). (B) Quantification of ancrod effects on fibrinogen ir (upper bar graph) and IgG ir (lower bar graph). Data are means ± S.E.M. from six animals. *Denotes significant difference for P < 0.05.

Figure 8

Effects of ancrod on gliosis and neuronal viability in Aβ_1-42-injected hippocampus. (A) Representative microglial (Iba-1 marker, upper panels), astrocyte (GFAP marker, middle panels) and neuronal (NeuN marker, lower panels) staining at 7 days with PBS, Aβ_1-42 and Aβ_1-42 plus ancrod treatment. Scale bar represents 100 μm. (B) Quantification of ancrod effects on microgliosis (upper bar graph), astrogliosis (middle bar graph) and neuronal viability (lower bar graph). Results (Iba-1 and NeuN) are expressed as the number of cells per high-power (×400) field. Data are means ± S.E.M. from six animals. *Denotes significant difference for P < 0.05. ^#Denotes P < 0.05 between Aβ_1-42 and Aβ_1-42 plus ancrod.

Figure 9

Microglial (Iba-1 marker) and neuronal (NeuN marker) staining following 7 days intra-hippocampal injection of Aβ_1-42 and Aβ_1-42 plus fibrinogen in the absence/presence of anti-Mac-1 treatment. (A) Controls (left panels) show that PBS and Aβ_42-1 injections are associated with low numbers of microglia. Injection of Aβ_1-42 (upper middle panel) or Aβ_1-42 plus fibrinogen (upper right panel) progressively increases microgliosis. Effects of anti-Mac-1 on microgliosis with Aβ_1-42 (lower middle panel) and Aβ_1-42 plus fibrinogen (lower right panel) injections. Scale bar represents 100 μm. (B) Quantification (n= 6 animals) of Iba-1 immunoreactivity (ir) for the different treatments. Results are expressed as the number of Iba-1 ir cells per high-power (×400) field. Data are means ± S.E.M. from six animals. *P < 0.05 versus PBS; ^#P < 0.05 versus Aβ_1-42 or Aβ_1-42 plus fibrinogen; **P < 0.05 versus Aβ_1-42. (C) Representative NeuN ir for controls (left panels), Aβ_1-42 and Aβ_1-42 plus fibrinogen (middle and right upper panels, respectively) and Aβ_1-42 and Aβ_1-42 plus fibrinogen with anti-Mac-1 treatment (middle and right lower panels, respectively). Scale bar represents 100 μm. (D) Quantification (n= 6 animals) of NeuN ir for the different treatments. Results are expressed as the number of NeuN ir cells per high-power field. Data are means ± S.E.M. from six animals. *P < 0.05 versus PBS; #P < 0.05 versus Aβ_{1- 42} or Aβ_1-42 plus fibrinogen; **P < 0.05 versus Aβ_1-42.

Figure 10

Effects of anti-Mac-1 on IgG staining in Aβ_1-42 and Aβ_1-42 plus fibrinogen-injected hippocampus. (A) Typical patterns of IgG immunoreactivity (ir) in controls (left panels) and with Aβ_1-42 (upper middle panel) and Aβ_1-42 plus fibrinogen (upper right panel). Effects of anti-Mac-1 are shown in lower middle and right panels. Scale bar represents 100 μm. (B) Quantification (n= 6 animals) of IgG ir for the different treatments. *P < 0.05 versus PBS; #P < 0.05 versus Aβ_1-42 or Aβ_1-42 plus fibrinogen; **P < 0.05 versus Aβ_1-42.