The research in my laboratory is focused on (1) deciphering the mechanisms that control normal development of the brain and how aberrations of such mechanisms produce diseases, and (2) investigating how such knowledge can be translated into improved patient care. We begin by studying molecular mechanisms and then build on the lessons learned from those studies using a multi-disciplinary approach that encompasses genomics, bioinformatics, biochemistry, cell biology and mouse genetics. Our work involves close collaboration between basic scientists and clinicians.
Working model of HRHMMB
Maintenance of stemness is one of the critical mechanisms in medulloblastoma
One of the first projects in my laboratory focused on advancing
understanding of the childhood brain tumor medulloblastoma (MB) as a
foundation for patient-specific therapeutic approaches. In the course
of our early work, we discovered that the transcriptional repressor
REST has a new function in that it maintains self-renewal and blocks
differentiation of normal neural stem cells. REST is aberrantly
overexpressed in a subclass of MB tumors, and a unique role of REST in
these MB tumors is to block differentiation and maintain stemness of
the cerebellar stem/progenitor cells. This work was exciting because
we and others suggested that REST inhibitors could be utilized to
block MB. Additional work suggested that REST cooperates with cMYC to
form this unique subgroup of MB (High REST, High cMYC MB; HRHMMB). We
are currently generating a genetically engineered mouse model of this
Stem and progenitor cells are more flexible than previously thought
We were one of the first to show muscle progenitor cells (mesodermal lineage) can be reprogrammed by a genome-wide transcriptomic shift into a physiologically active neuronal phenotype (ectodermal lineage) by simply activating REST target genes with a single recombinant transcription factor. Although not universally accepted in 2003-2004, our studies provided some of the earliest evidence that cells are more flexible than was previously thought: cell fates can be switched by simple manipulation of a few transcription factors. Later elegant studies from several laboratories have shown that mouse and human fibroblasts can be reprogrammed to an induced-pluripotent state (iPS) by the transfer of a few transcription factors and that these iPS cells can then be differentiated into various cell-types, supporting our original findings. These studies have far-reaching implications in stem cell biology and human health.
Embryonic stem cell pluripotency is context-dependent
We discovered that REST maintains pluripotency of embryonic stem cells by maintaining the expression of the known pluripotency regulators, including Oct4, Nanog, and Sox2, through a novel microRNA-mediated mechanism. Our further work results resolved some of the contradictions in the literature and showed that the REST-mediated regulation of ES cell pluripotency depends on the cell-type (not all ES lines are the same) as well as the culture conditions, indicating how various factors form part of an interconnected genome-wide network influencing each other.
REST in glioblastoma stem cells and in Precision Medicine
More recently, we were also the first to show a REST-mediated regulation of oncogenic properties of glioblastoma stem cells via maintenance of stemness. Using bioinformatics and biochemical validations, we recently found a new mechanism with implications in Personalized Medicine. We found that the stemness regulator Sox2 is a new, clinically important target of microRNA-21 (miR-21) in patient glioblastoma tumors, with implications for prognosis. Using the miR-21-Sox2 regulatory axis, glioblastoma tumors can be classified into “stem-like” versus “neuronal progenitor-like” subtypes. The miR-21-Sox2 axis was also found in mouse neural stem cells and in the mouse brain at different developmental stages, suggesting a role in normal development. Importantly, this classification is a better predictor of patient survival than currently used parameters. Thus, this mechanism-based classification identifies a distinct population of glioblastoma patients with distinguishable phenotypic characteristics and clinical outcomes. In addition, we recently discovered two new mechanisms in glioblastoma. First, using genome-wide expression analysis followed by biochemical validations, we found that a new REST target in patient glioblastoma tumor-derived stem cells (GSCs) is miR-203, and that the REST-miR-203 axis specifically regulates invasion, and not proliferation or apoptosis. Second, REST represses a new target, the Dopamine Receptor 2 (DRD2) gene transcription to control tumorigenic properties of a subclass of GSCs. The latter project identified how REST can regulate GBM biology through neurotransmitter signaling. We are currently examining the potential use of FDA-approved drugs in this class of GBM tumors.
Chromatin-mediated regulation in vivo: zygotic gene transcription, locomotion and chronic pain
I began my independent laboratory with two projects, the MB project described above, and a second project focused on determining how chromatin states regulate transcription at the beginning of mammalian development. By modifying chromatin through histone manipulation in mouse 1-cell and 2-cell embryos, we discovered that the chromatin regulation of transcription begins at the 2-cell stage of development, coinciding with the onset of zygotic transcription. Further, this function requires a unique coactivator activity. This was an exciting finding, but at the same time, we also discovered the important role of REST in MB. Because the latter discovery had potential relevance to MB patient care, we redeployed all our lab resources to that project. Recently, our work on REST in normal mice came full circle. Although REST overexpression (OE) has been implicated in many diseases and behavioral disorders, there has been a critical lack of a conditional REST OE mouse model creating a roadblock to mechanistically study the role of REST OE in these disorders under physiological conditions. To this end, we became involved in addressing this problem. We have now created the first conditional REST OE knockin mouse line. Our published results confirm that tissue-specific REST OE in these mice is physiologically relevant. Genome-wide analyses followed by biochemical and behavioral assays indicated that REST regulates spontaneous locomotion by repressing a new target, Dopamine Receptor 2. In another line of work and in collaboration with Hui-Lin Pan, we discovered that REST regulates chronic pain after nerve injury. We are currently working to decipher these mechanisms.