LDL receptor blockade reduces mortality in a mouse model of ischaemic stroke without improving tissue-type plasminogen activator-induced brain haemorrhage: towards pre-clinical simulation of…

Be’eri Niego, Brad R. S. Broughton, Heidi Ho, Christopher G. Sobey, Robert L. Medcalf

Symptomatic intracerebral haemorrhage (sICH) following tissue-type plasminogen activator (rt-PA) administration is the most feared and lethal complication of thrombolytic therapy for ischaemic stroke, creating a significant obstacle for a broader uptake of this beneficial treatment. rt-PA also undermines cerebral vasculature stability in a multimodal process which involves engagement with LDL receptor-related protein 1 (LRP-1), potentially underlying the development of sICH. We aimed to simulate rt-PA-induced haemorrhagic transformation (HT) in a mouse model of stroke and to assess if it drives symptomatic neurological deterioration and whether it is attenuated by LDL receptor blockade. rt-PA (10 mg/kg) or its vehicle, with or without the LDL receptor antagonist, receptor-associated protein (RAP; 2 mg/kg), were intravenously injected at reperfusion after 0.5 or 4 h of middle cerebral artery occlusion (MCAo). Albumin and haemoglobin content were measured in the perfused mouse brains 24 h post MCAo as indications of blood-brain barrier (BBB) compromise and HT, respectively. rt-PA did not elevate brain albumin and haemoglobin levels in sham mice or in mice subjected to 0.5 h MCAo. In contrast, administration of rt-PA after prolonged MCAo (4 h) caused a marked increase in HT (but similar changes in brain albumin) compared to vehicle, mimicking the clinical shift from a safe to detrimental intervention. Interestingly, this HT did not correlate with functional deficit severity at 24 h, suggesting that it does not play a symptomatic role in our mouse stroke model. Co-administration of RAP with or without rt-PA reduced mortality and neurological scores but did not effectively decrease brain albumin and haemoglobin levels. Despite the proven causative relationship between severe HT and neurological deterioration in human stroke, rt-PA-triggered HT in mouse MCAo does not contribute to neurological deficit or simulate sICH. Model limitations, such as the long duration of occlusion required, the type of HT achieved and the timing of deficit assessment may account for this mismatch. Our results further suggest that blockade of LDL receptors improves stroke outcome irrespective of rt-PA, blood-brain barrier breakdown and HT.