Edward Y. Estrada

Matrix Metalloproteinases and TIMPs Are Associated With Blood-Brain Barrier Opening After Reperfusion in Rat Brain

BACKGROUND AND PURPOSE Reperfusion disrupts cerebral capillaries, causing cerebral edema and hemorrhage. Middle cerebral artery occlusion (MCAO) induces the matrix-degrading metalloproteinases, but their role in capillary injury after reperfusion is unknown. Matrix metalloproteinases (MMPs) and tissue inhibitors to metalloproteinases (TIMPs) modulate capillary permeability. Therefore, we measured blood-brain barrier (BBB) permeability, brain water and electrolytes, MMPs, and TIMPs at multiple times after reperfusion. METHODS Adult rats underwent MCAO for 2 hours by the suture method. Brain uptake of 14C-sucrose was measured from 3 hours to 14 days after reperfusion. Levels of MMPs and TIMPs were measured by zymography and reverse zymography, respectively, in contiguous tissues. Other rats had water and electrolytes measured at 3, 24, or 48 hours after reperfusion. Treatment with a synthetic MMP inhibitor, BB-1101, on BBB permeability and cerebral edema was studied. RESULTS Brain sucrose uptake increased after 3 and 48 hours of reperfusion, with maximal opening at 48 hours and return to normal by 14 days. There was a correlation between the levels of gelatinase A at 3 hours and the sucrose uptake (P<0.05). Gelatinase A (MMP-2) was maximally increased at 5 days, and TIMP-2 was highest at 5 days. Gelatinase B and TIMP-1 were maximally elevated at 48 hours. The inhibitor of gelatinase B, TIMP-1, was also increased at 48 hours. Treatment with BB-1101 reduced BBB opening at 3 hours and brain edema at 24 hours, but neither was affected at 48 hours. CONCLUSIONS The initial opening at 3 hours correlated with gelatinase A levels and was blocked by a synthetic MMP inhibitor. The delayed opening, which was associated with elevated levels of gelatinase B, failed to respond to the MMP inhibitor, suggesting different mechanisms of injury for the biphasic BBB injury.

Immunohistochemistry of matrix metalloproteinases in reperfusion injury to rat brain: activation of MMP-9 linked to stromelysin-1 and microglia in cell cultures

Reperfusion damages the blood-brain barrier (BBB). Matrix metalloproteinases (MMPs) are associated with the opening of the BBB, but their cellular localization and activation mechanisms are uncertain. We used immunohistochemistry to determine the cellular localization of the MMPs in reperfused rat brain, and cell cultures to study their activation. Spontaneously hypertensive rats (SHR) had a 90 min middle cerebral artery occlusion (MCAO) followed by reperfusion for times from 3 h to 21 days. Frozen sections were immunostained with antibodies to gelatinase A (MMP-2), stromelysin-1 (MMP-3), and gelatinase B (MMP-9). Sham-operated control rats showed MMP-2 immunostaining in astrocytic processes next to blood vessels. After 3 h of the onset of reperfusion MMP-2 immunostaining increased in astrocytes. At 24 h immunoreactivity for MMP-3 and MMP-9 appeared. MMP-3 co-localized with activated microglia (Ox-42+) and ischemic neurons (NeuN+). MMP-9 immunostaining was seen at 48 h in endothelial cells, neutrophils, and neurons. At 5 and 21 days intense MMP-2 staining was seen in reactive astrocytes around the ischemic core. Studies of activation of the MMP were done in lipopolysaccharide (LPS)-stimulated astrocyte and microglia cultures. Stimulated astrocytes produced an activated form of MMP-2. When microglia were stimulated, they activated MMP-9. Immunostaining showed MMP-3 in cultures of enriched microglial cells. The hydroxymate-type, MMP inhibitor, BB-1101, blocked the activation of MMP-2 and MMP-9 by LPS in mixed glial cultures. We propose that MMP-2 is normally present in astrocytic end feet, and that during ischemia MMP-9 and MMP-3 are produced. MMP-3 in microglia/macrophages may be activating proMMP-9. Our results show that a differential expression of MMPs by astrocytes, microglia, and endothelial cells at the blood vessels is involved in the proteolytic disruption of the BBB.

TIMP-2 reduces proteolytic opening of blood-brain barrier by type IV collagenase

Intracerebral hemorrhage occurs in tumors, stroke and head trauma. Proteolysis of the extracellular matrix around cerebral capillaries by naturally occurring mammalian 72-kDa type IV collagenase may initiate this pathologic event. To investigate this hypothesis adult rats underwent intracerebral injection of type IV collagenase purified from human melanoma cells. Histologically, at 4 h there was perivascular cellular infiltration with hemorrhage, and by 24 h there was infarction with necrosis, edema and hemorrhage. Ultrastructurally, the basal lamina of endothelial cells was disrupted at 2 h. Brain uptake of [14C]dextran and [3H]sucrose increased after intracerebral injection of type IV collagenase compared to controls (P less than 0.0001). Tissue inhibitor of metalloproteinase-2 (TIMP-2) reduced the tracer uptake (P less than 0.02). Metalloproteinase inhibitors reduce extracellular matrix proteolysis and protect the blood-brain barrier.

Tumor necrosis factor-α-induced gelatinase B causes delayed opening of the blood-brain barrier: an expanded therapeutic window

Proteolytic damage is a late event in the molecular cascade initiated by brain injury. Earlier, we proposed that matrix metalloproteinases (MMPs) and urokinase-type plasminogen activator (uPA) are important in secondary brain injury. We have shown that intracerebral injection of activated 72-kDa type IV collagenase (gelatinase A) opens the blood-brain barrier, and that during hemorrhagic brain injury there is endogenous production of 92-kDa type IV collagenase (gelatinase B) and uPA. Therefore, to study the functional link between proteolytic enzymes and blood-brain barrier damage, we induced MMP expression by infusing tumor necrosis factor-alpha (TNF) intracerebrally in rats. Initially, the effect on capillary permeability of increasing doses of TNF, using [14C]sucrose uptake, was measured. Then, the time-course of the capillary permeability change was studied at 4, 16, 24 and 72 h. Expression of MMP and uPA was measured by zymography at 24 h after TNF injection and compared to saline-injected controls. A dose-dependent increase in capillary permeability was seen 24 h after TNF injection. Maximal uptake of [14C]sucrose occurred at 24 h compared to saline-injected controls (P < 0.05). Zymography showed production of gelatinase B, which was significantly greater than in saline-injected controls at 24 h (P < 0.05). Batimastat, a synthetic inhibitor to metalloproteinases, reduced sucrose uptake at 24 h (P < 0.0001), and was effective even when given 6 h after TNF (P < 0.01). Thus, gelatinase B is the intermediate substance linking TNF to modulation of capillary permeability. Agents that interfere with transcription of proteolytic enzymes or block their action may reduce delayed capillary injury, extending the therapeutic window.

Bacterial collagenase disrupts extracellular matrix and opens blood-brain barrier in rat ☆

Bacterial collagenase causes hemorrhagic necrosis of brain. We studied the enzymes effect on blood-brain barrier (BBB) permeability and extracellular matrix (ECM) structure by radiolabeled tracers and electron microscopy. Adult rats had intracerebral injection of bacterial collagenase. Brain uptake from blood of [14C]sucrose was measured in 24 rats 0.5 h to 14 days after injection. 12 rats had ultrastructural studies 1 h after collagenase injection. Brain uptake of [14C]sucrose is maximally increased at 0.5 h, remaining significantly increased for 7 days. Ultrastructurally, some vessels had widening of basal lamina while others had severe disruption of basal lamina with stretching of endothelial cells. We conclude that bacterial collagenase disrupts ECM and opens BBB.

Neutral proteases and disruption of the blood-brain barrier in rat.

Blood-brain barrier disruption is common in many neurological diseases. Matrix metalloproteinases are induced in brain injury and increase capillary permeability by attacking the extracellular matrix around cerebral capillaries. Other neutral proteases are also increased in sites of secondary injury, and may contribute to the proteolysis of the blood-brain barrier. Therefore, we studied capillary permeability and histological tissue damage after intracerebral injection of neutrophil elastase, cathepsin G, heparatinase and plasmin. Adult rats were injected intracerebrally with an enzyme. After 1, 4 or 24 h, measurements were made of brain uptake of a radiolabeled tracer, [14C]sucrose. Enzymes that significantly increased capillary permeability were injected into other rats for histological assessment of tissue damage. Elastase increased capillary permeability significantly when compared with controls; maximal damage was seen at 4 h. Plasmin produced smaller increases in permeability at 4 h, exerting its maximal effect on sucrose uptake at 24 h. Cathepsin G had a small effect at 4 h. Heparitinase had no effect. Histologic examination of elastase-injected brains at 24 h revealed multifocal perivascular and intraparenchymal acute hemorrhages accompanied by a polymorphonuclear cell infiltrate. Elastase-injected brains were microscopically similar to saline-injected brains at 1 and 4 h. Plasmin produced fibrinoid changes in the blood vessels at 24 h, coinciding with the maximal increase in capillary permeability. We conclude that neutrophil elastase attacks the capillary extracellular matrix, causing extensive hemorrhage, while plasmin leads to increased vascular permeability and fibrinoid necrosis of blood vessel walls. Differential effects of neutral proteases released secondary to injury could be important in both the acute changes in blood vessel permeability and long-term alterations in vessel structure.

Arginine vasopressin V1-antagonist and atrial natriuretic peptide reduce hemorrhagic brain edema in rats.

Background and Purpose: Injection of arginine vasopressin into the cerebral ventricles in animals with brain injury increased brain water, whereas injection of atrial natriuretic peptide reduced water content. Therefore, to determine the role of endogenous arginine vasopressin in brain edema, we attempted to inhibit edema from a hemorrhagic lesion with an arginine vasopressin V1 receptor antagonist or atrial natriuretic peptide. Methods: Adult Sprague-Dawley rats with hemorrhages induced by 0.4 IU bacterial collagenase were treated with 75 ng (n=9) or 8 μg (n=9) of the vasopressin V1 receptor antagonist d(CH2)5Tyr(Me)Arg, 3.2 μg (n=4) atrial natriuretic peptide injected intracerebrally, or 5 μg/kg per hour (n=7) atrial natriuretic peptide intraperitoneally. They were compared with control groups injected with 0.4 IU collagenase only. Brain water and electrolytes were measured 24 hours later. Brain uptake of [14C] sucrose was measured 30 minutes after lesions were induced by 0.4 IU collagenase alone (n=5) or after collagenase injection and 50 μg/kg per hour n=5) atrial natriuretic peptide injected intravenously. Results: The arginine vasopressin V1 receptor antagonist and atrial natriuretic peptide significantly (p<0.05) reduced water and sodium contents in the posterior edematous regions. Brain uptake of [14C] sucrose was significantly reduced by intravenous atrial natriuretic peptide. Conclusions: Antagonists to arginine vasopressin V1 receptors and atrial natriuretic peptide both significantly reduce hemorrhagic brain edema, and atrial natriuretic peptide appears to protect the blood-brain barrier.

Atrial Natriuretic Peptide Blocks Hemorrhagic Brain Edema After 4-Hour Delay in Rats

BACKGROUND AND PURPOSE Atrial natriuretic peptide (ANP) and arginine vasopressin regulate brain water and electrolytes. Treatment with ANP at the onset of a hemorrhagic injury reduces edema. Clinically, however, hemorrhagic masses form too rapidly for preventive treatment. Therefore, we measured the effect of ANP on brain edema after the hemorrhagic mass was formed. METHODS Adult rats had hemorrhagic lesions produced by the intracerebral injection of 0.4 U bacterial collagenase. Four hours later, an infusion of ANP (120 or 700 ng/kg per 20 hours) was begun into the peritoneum using an implanted miniosmotic pump. Twenty-four hours after the injury, brain water and electrolyte values were measured. The mechanism of ANP action was explored in other groups of rats that either had osmolality increased with mannitol or were injected with the cyclic GMP analogue, 8-bromo-cGMP. RESULTS Atrial natriuretic peptide given after a 4-hour delay significantly reduced brain water and sodium 24 hours after the injury (P < .05). However, neither mannitol nor 8-bromo-cGMP affected brain edema. CONCLUSIONS Delayed administration of ANP reduces brain edema secondary to a hemorrhagic mass. Because it is effective after the mass has formed, ANP may be useful in treatment of edema secondary to intracranial bleeding.

Autoradiographic Patterns of Brain Interstitial Fluid Flow After Collagenase-induced Haemorrhage in Rat

Cerebral oedema accompanies intracerebral haemorrhage. We induced intracranial bleeding by the intracerebral injection of bacterial collagenase. There was oedema observed both at the haematoma site in the caudate/putamen and bilaterally in the hippocampal regions. To determine the role of vasogenic oedema spread from the site of injury, we studied by autoradiography the distribution of extracellular markers injected along with the collagenase. Both 14C-dextran (m.w. 70,000) and 14C-sucrose (m.w. 341) spread away from the injection site into both hippocampal regions in a similar pattern, suggesting bulk flow. Vasogenic oedema secondary to a haemorrhagic lesion in the caudate/putamen is an important cause of the oedema observed in both hippocampal regions in our model.

The effect of arginine vasopressin and V1 receptor antagonist on brain water in cat

Arginine vasopressin (AVP) is important in brain water regulation. To better understand the effect of AVP released by extrahypothalamic fibers in brain, we microinfused AVP into intact brain and studied its effect on brain water and electrolytes. Adult cats had 5 ng of AVP infused into the caudate nuclei. Four h after infusion the brains were removed for measurement of water and electrolyte contents. Animals infused with AVP were compared to controls infused with saline. AVP increased water content significantly in gray and white matter sites, while electrolyte content was unchanged. Another group of animals had intracerebral infusions with 5 ng of AVP and 50 ng of a V1 receptor antagonist, (d(CH2)5Tyr-(Me)AVP). The antagonist blocked the increase in water, suggesting a V1 receptor mediated the action.