UC Berkeley

Dynamic regulation of cell signaling can drive cellular decisions, whereby a cell can choose divergent fates based on the strength, timing, or location of even an individual protein signal. However, the chemical and genetic tools widely used to probe cell signaling suffer from poor spatiotemporal resolution and cannot easily recapitulate the dynamics with which protein signals affect cell fate choices. Optogenetic techniques have recently enabled rapid and reversible cellular protein activation through light inducible protein heterodimerization, homodimerization, and gene transcription within living cells. Protein oligomerization, however, represents a distinct and important mode of activation for numerous cell signaling events, yet its study and control remain challenging due to the lack of tools to inducibly modulate a protein's oligomeric state. In this dissertation, I will describe the discovery and characterization of a novel method to optically regulate protein oligomerization within cells. This discovery has allowed us precise control over intracellular signaling pathways and cell fate decisions that are a consequence of these signals.

Cryptochrome2 (Cry2) is a light sensitive protein that forms light-dependent oligomeric "photobodies" in Arabidopsis thaliana. We demonstrated that these photobodies can be formed inducibly, reversibly, and tunably in mammalian cells, and we co-opted Cry2 oligomerization to optically regulate mammalian signaling cascades. Using end-to-end protein fusions of Cry2 to specific signaling proteins, we first enabled optical control over the b-catenin pathway, achieving a higher transcriptional response than obtained with the natural pathway ligand Wnt 3a. We then used Cry2 clustering to robustly and dynamically activate RhoA signaling with light. To our knowledge this is the first demonstration of clustering as a mode of activation for this well-studied protein. We have also shown conservation of activation via clustering among Rho family members Rac1 and Cdc42, enabling photoactivation of all three GTPases through identical Cry2 fusions.

In subjecting multiple distinct signaling pathways to optical regulation via Cry2, we showed that Cry2 may be abstracted as a clustering module capable of regulating numerous signaling networks at nodes sensitive to clustering. Supporting and expanding on this hypothesis, we developed an extension of this method called Clustering Indirectly with Cryptochrome 2, or CLICR. CLICR enables Cry2-mediated clustering of binding partners, which can be applied to regulate both cytoplasmic and membrane-bound signaling proteins and, excitingly, can be engineered to modulate endogenous protein activity through fusion of an appropriate adapter molecule to the Cry2 module. We used this generalizable approach to optically cluster and regulate endogenous transmembrane receptor tyrosine kinase (RTK) and b-integrin activity with high spatiotemporal precision, enabling inducible and reversible control of transmembrane receptors without overexpression of exogenous signaling molecules. This methodology broadly extends optogenetic capabilities to a large class of endogenous intracellular and transmembrane signaling proteins important in cell fate decisions and disease progression.

In developing optogenetic Cry2 clustering, we demonstrated its utility in uncovering oligomerization as a mode of regulation for certain signaling proteins. Since the Rho GTPases were not known to be modulated in this manner, we further examined the mechanism by which clustering enhances Rho GTPase membrane translocation. We found that, although cluster-induced membrane translocation is dependent on interaction with its activating GEF enzyme, GEF catalysis is not necessary for this localization, while the presence of the C-terminal lipid tail is critical. We propose a model of activation based on these findings. Further, we developed sensitive methods for detection of cluster-induced GTPase activation, and we identify and test the role of putative elements postulated to enhance GTPase activity and inactivation upon oligomerization.

In addition to uncovering novel mechanistic protein activity, the Cry2 optogenetic tool has tremendous potential for studying the dependence of signaling dynamics on cell fate decisions. Adult neurogenesis is of great interest for ameliorating numerous neurodegenerative diseases, and b-catenin signaling plays a central role in this process. We demonstrated that, upon further protein engineering and illumination-hardware optimization, Cry2 clustering could be used to strongly induce b-catenin signal within neural stem cells and enable optogenetic neurogenesis. We further showed that neurogenesis could be tuned by modulating the activating light source, setting the stage for studies interrogating the neural stem cell response to defined timing and intensity of the b-catenin pathway, yielding insight into how these cells may integrate neurogenic cues in their native stem cell niche.

In summary, we have developed a modular, single-construct, and genetically encoded method to rapidly and reversibly induce oligomerization within mammalian cells, and we have shown its utility and portability in regulating several diverse proteins pathways on fast timescales in response to blue light illumination. We have used Cry2 to both uncover novel protein signaling properties and to regulate and interrogate important cell fate decisions, and we anticipate its modularity will allow broad expansion of the optogenetic toolbox to a large class of signaling proteins important in cell fate decisions and disease progression.