A murine model of inflammation-induced cerebral microbleeds

Background: Cerebral microhemorrhages (CMH) are tiny deposits of blood degradation products in the brain and are pathological substrates of cerebral microbleeds. The existing CMH animal models are beta-amyloid-, hypoxic brain injury-, or hypertension-induced. Recent evidence shows that CMH develop independently of hypoxic brain injury, hypertension, or amyloid deposition and CMH are associated with normal aging, sepsis, and neurodegenerative conditions. One common factor among the above pathologies is inflammation, and recent clinical studies show a link between systemic inflammation and CMH. Hence, we hypothesize that inflammation induces CMH development and thus, lipopolysaccharide (LPS)-induced CMH may be an appropriate model to study cerebral microbleeds. Methods: Adult C57BL/6 mice were injected with LPS (3 or 1 mg/kg, i.p.) or saline at 0, 6, and 24 h. At 2 or 7 days after the first injection, brains were harvested. Hematoxylin and eosin (H&E) and Prussian blue (PB) were used to stain fresh (acute) hemorrhages and hemosiderin (sub-acute) hemorrhages, respectively. Brain tissue ICAM-1, IgG, Iba1, and GFAP immunohistochemistry were used to examine endothelium activation, blood-brain barrier (BBB) disruption, and neuroinflammation. MRI and fluorescence microscopy were used to further confirm CMH development in this model. Results: LPS-treated mice developed H&E-positive (at 2 days) and PB-positive (at 7 days) CMH. No surface and negligible H&E-positive CMH were observed in saline-treated mice (n = 12). LPS (3 mg/kg; n = 10) produced significantly higher number, size, and area of H&E-positive CMH at 2 days. LPS (1 mg/kg; n = 9) produced robust development of PB-positive CMH at 7 days, with significantly higher number and area compared with saline (n = 9)-treated mice. CMH showed the highest distribution in the cerebellum followed by the sub-cortex and cortex. LPS-induced CMH were predominantly adjacent to cerebral capillaries, and CMH load was associated with indices of brain endothelium activation, BBB disruption, and neuroinflammation. Fluorescence microscopy confirmed the extravasation of red blood cells into the brain parenchyma, and MRI demonstrated the presence of cerebral microbleeds. Conclusions: LPS produced rapid and robust development of H&E-positive (at 2 days) and PB-positive (at 7 days) CMH. The ease of development of both H&E-and PB-positive CMH makes the LPS-induced mouse model suitable to study inflammation-induced CMH.