Central nervous system (CNS) diseases are a leading cause of death and disability worldwide, and their incidence is expected to increase in tandem with global life expectancies. Many available CNS therapies only ameliorate disease symptoms rather than halting or reversing the actual disease burden, and potential CNS drugs also have lower success rates at all stages of the clinical development pipeline compared to non-CNS candidates. To combat these issues, we currently focus our efforts in the following general areas:
pioneering new routes for stem cell differentiation
building representative neurovascular models using stem cells
understanding neurovascular function and disease mechanisms using genetic engineering and high throughput screening
developing new methods to visualize, target, and control cell phenotypes and behaviors
Building CNS tissue models using human induced pluripotent stem cells (iPSCs)
Animal models of CNS disease can provide excellent insights into causative mechanisms, but therapies that are successful in rodents too often fail in human trials. Moreover, animal models are inherently low throughput for identifying effective therapeutic candidates. Human induced pluripotent stem cells (iPSCs), which grow indefinitely in culture and can be coaxed to form any cell type found in the human body, have been hailed as a potential alternative to animal models. For our purposes, iPSCs can produce an unlimited supply of human cells found in the vascularized CNS, including brain microvascular endothelial cells (BMECs), pericytes, neurons, astrocytes, oligodendrocytes, and microglia. We are interested in using a number of engineering techniques and classical differentiation methods to effectively build representative CNS models from iPSCs. Ultimately, these models will be used for applications in drug screening and disease modeling.
Interrogating neurovascular function with targeted and high throughput approaches
The neurovascular unit (NVU) and blood-brain barrier (BBB) are disrupted during the majority of CNS diseases, but little is known about the underlying mechanisms, in part because BBB function itself is poorly understood. Owing to our ability to make high-fidelity BMECs from iPSCs, we have gathered a wealth of information on possible contributors to the BBB phenotype using standard molecular profiling techniques (Approach 1). We mine this data and perform targeted perturbations to assess the relevance of each gene, protein, and signaling pathway towards maintenance of the BBB phenotype. We are also applying high throughput CRISPR and siRNA approaches (Approach 2) to interrogate specific BBB functions that can be assayed based on fluorescent readouts.
Improving the specificity of molecular targeting
Signaling pathways are incredibly complex, yet have surprising redundancy. For example, the >30 members of the transforming growth factor beta (TGFβ) superfamily signal through only 7 type I and 5 type II receptors that form homomeric or heteromeric complexes which dictate how the signal is processed. So, for example, if we want to use a TGFβ protein for a therapeutic purpose, how do we ensure that it only signals through the appropriate receptor complex that generates the desired response? To address this problem, we are developing new in vitro selection workflows to isolate biomolecules with high affinity and specificity against soluble proteins and cell surface receptors. While our immediate interests apply to the CNS and brain drug delivery, we anticipate these technologies could be applied to numerous diseases or used to intimately probe biological function.
Conditionally-fluorescent molecular biosensors
Assessment of cell phenotypes in real-time remains challenging. Performing genetic screens with non-fluorescent readouts (e.g. uptake and trafficking of synthetic small molecules) is also quite difficult. To address these issues, we are using in vitro selection and directed evolution techniques to develop molecular scaffolds that fluoresce only in the presence of specific analytes.