Blood Brain Barrier Modelling

About

Innovation Description A recent analysis of Phase II and III compound failures between 2013-2016 revealed that attrition rates were second highest in central nervous system (CNS) therapeutics after oncology, and were attributed to poor efficacy (Harrison, 2016). Efficacy and safety profiles of therapeutics are determined by the extent of target tissue exposure, whereby poor blood brain barrier (BBB) permeability and low drug residence time in brain tissue is often linked to poor drug activity.   The core premise of the innovative solution offered here is that sophisticated physiologically relevant models are needed to provide the critical information required to make successful early decisions in the lead optimisation/ selection phase of drug development. This value proposition was substantiated by Elebring (2012) who stated ‘focus on science quality should lead to improved drug discovery quality, cost and speed.’ As such, since 2007, the pharmaceutical industry has shifted distribution of failures to earlier in development rather than late clinical trials to reduce clinical costs by 5.5 %, but further improvements in the field of CNS therapeutics are required (DiMasi et al., 2016).   The existing in vitro techniques used routinely in the Pharmaceutical Industry (Caco-2, MDCK-MDR1 cells, PAMPA assays) often lead to poor in vivo predictions due to use of over simplified cellular models from the wrong tissue, wrong species, grown in an irrelevant architecture with under expression of relevant drug metabolising enzymes and transporters. The end result is over prediction of in vitro CNS permeability, leading to poor correlation of in vivo drug disposition, hence often observed as a lack of efficacy in clinical trials due to under exposure of the drug to CNS tissue.   The solution to this problem was to produce a metabolically competent, three-dimensional, all human model of the BBB, derived from primary human cells with dynamic flow. The term barrier suggests a static physical structure, traditionally thought of as a monolayer of endothelial cells, but it is now well established that the BBB metabolic and transport barrier is a dynamic interface, whose phenotype is critically dependent on an intact functional neurovascular unit (NVU), yet currently available models of BBB (endothelial monolayers of non-human origin) have not evolved to reflect this. The NVU is comprised of dynamically integrated endothelial cells, pericytes, astrocytes, neurons and microglia. Tight junctions between endothelial cells prevent paracellular transport and pericytes play a key role in the development of cerebral microvaculature and regulation of blood flow. Astrocytes and the basal lamina support function of neurons, regulate pericyte differentiation and communicate with segments of the microvasculature.   CNS barriers undergo progressive decline in function during aging, as the pro-/anti-inflammatory cytokine balance shifts towards pro-inflammatory cytokines leading to a progressive low-grade inflammatory state thereby disturbing brain homeostasis and leaving the brain vulnerable to neurodegenerative diseases. Interestingly, more focus has recently been placed on the better understanding of the mechanisms of damage at the BBB as an early event in many neurological conditions and therefore it is no surprise there is growing interest in the BBB as a therapeutic target in its own right.   Over the past 30-years there have been many iterations of in vitro BBB models, however recent advances in nanotechnology, microfluidics and three-dimensional bioscaffolds have progressed the field. The widespread acceptance of the importance of 3D hierarchy, extracellular matrix and the presence of multiple cell types required in culture for effective cell-cell communication and NVU phenotype has led to exciting developments in the research area of in vitro modelling. Furthermore, there has been increased emphasis on the importance of species specific models for producing relevant scaling factors for use in in vitro-in vivo correlations.   Dr Jane Alder’s research group have produced an all human dynamic flow 3D in vitro model of the BBB containing cells of the NVU (under patent). The model has been optimised to express maximal trans endothelial resistance, tight-junction (claudin-5 and occludin), transporter (MDR1 and BCRP) and enzyme (CYP2D6 and CYP3A4) expression and activity and validated against commercial instruments such as the CellZScope® (Nanoanalytics, Munster, Germany) and ECIS® (Applied Biophysics, NY, USA).   In summary, the BBB model may be used to predict drug disposition to prove brain targeting or conversely the model may be used to assess the risk of CNS exposure and unwanted off-target toxicity. The model may be adapted to induce a disease state, aiding elucidation of mechanisms of CNS pathophysiology.   Harrison RK, (2016) Nature Reviews: Drug Discovery, 15: 817-8 Elebring T, (2012) Drug Discovery Today, 17:1166-9 DiMasi JA, et al., (2016) Journal of Health Economics, 47: 20-33