In recent years, synthetic immunology has allowed us to probe the complex interactions between immune cells and their environment, and to develop novel strategies for treating diseases. Cell-based immunotherapies, particularly chimeric antigen receptor (CAR) T cells, have made advances in the clinic for treatment of hematological malignancies. However, there is a broader potential for cell-based immunotherapies to impact the treatment of many challenging diseases such as solid tumors or autoimmunity. Immune cells can be engineered with synthetic circuits to carry out more precise molecular recognition and therapeutic action. Synthetic reconstitution of immune signaling by engineering systems from the bottom-up allow us to identify molecular features sufficient to achieve particular sets of behaviors. Here, we engineered T cells using synthetic Notch (synNotch) receptors that induce the local production of therapeutic payloads. First presented is an exploration of T cell circuits that induce the production of pro-inflammatory cytokine IL-2 specifically at the site of a tumor, bypassing tumor suppression to clear challenging solid tumors without inducing systemic toxicity. Second presented is the study of T cell circuits that drive local immune suppression to block off-target CAR T cell toxicity or protect transplants from cytotoxic T cell killing without systemic immune suppression. Together, these studies demonstrate how synthetic T cell circuits can be used to perturb immune microenvironments for therapeutic applications.

Eukaryotic cells must process large amounts of information and respond correctly to stimuli in order to maintain the health of an organism. Much of this information processing is done by networks of signal transduction proteins. Many signaling proteins behave as allosteric systems that have both active and inactive states. These proteins must recognize specific molecular inputs and, in response, modify their catalytic or binding activities to create the appropriate output. One such allosteric switch is the GTPase-binding domain (GBD) of the actin-regulatory protein WASP (Wiskott-Aldrich syndrome protein). Basally, the GBD autoinhibits WASP's actin polymerization activity through an intramolecular interaction with the C helix, but intermolecular binding of the GTPase Cdc42 to the GBD disrupts this autoinhibitory interaction. The inspiration behind this work was to gain an understanding of some of the mechanisms by which the regulation of signaling proteins such as WASP can be rewired. Specifically we were interested in strategies for bypassing the endogenous regulation of these proteins and making them responsive to different inputs that we can control. We took two approaches to accomplish this: (1) engineering novel protein switching components and (2) studying the mechanism by which the pathogenic bacterium EHEC is able to hijack WASP regulation.

In the first case, we found that by overlapping the primary sequences of pairs of protein interaction modules, we could render their two interactions mutually exclusive. We characterized the behaviors of these engineered interaction switches and also incorporated them into naturally-occurring signaling proteins and pathways. These modified components were able to rewire signaling in vitro and in vivo.

In the second, we characterized the EHEC protein EspFU, which potently activates host cell WASP, leading to actin polymerization. We found that EspFU contains a repeated mimic of the autoinhibitory C helix from WASP, and that this motif competitively disrupts WASP autoinhibition. It is also functionally significant that this motif is repeated six times in EspFU, because potent activity seems to be dependent on EspFU's ability to coordinate the simultaneous activation of at least two WASP proteins.

Many cellular processes require telling time. Cells may use kinetic filtering, the ability to respond to sustained but not transient stimulation, to decode information stored in dynamical profiles and exert control over relative timing of events. Previous work has found only one circuit architecture, the AND-gated coherent feed forward loop, capable of such behavior. We enumerate the space of 1-, 2-, and 3-node networks to ask what additional architectures can drive kinetic filtering, which is quantified by measuring temporal dose response steepness and trigger time. We find two types of positive feedback architectures, two types of buffering architectures, and the coherent feed forward loop with AND logic are the only kinetic filters of 3 or fewer components. The buffering double inhibition circuit architecture is modular, reversible, capable of the widest range of trigger times, and most suitable as a template for future engineering efforts.

In eukaryotic cells, networks of signaling proteins are responsible for converting environmental inputs into the appropriate regulatory outputs. The proteins that comprise signaling networks are highly modular, and are typically made up of multiple, functionally distinct domains. These domains are of two varieties: 1) catalytic domains (eg. kinases, GTPases), which leave a chemical imprint on a downstream target, and 2) interaction domains, which facilitate recognition between network members. As such, information flow in signaling networks is potentiated by catalytic domain function, while protein-protein interaction domains direct this catalytic function to specific targets, thereby defining network connectivity.

In signaling networks, groups of proteins that mediate a particular response (pathways) are often found co-localized with one another into complexes. These complexes are coordinated by scaffolds proteins, which typically are composed of multiple, modular interaction domains. By defining which catalytic components are co-localized within a pathway complex, the interaction domains in a scaffold encode the connectivity for that pathway. It has been hypothesized that scaffolding may have facilitated the evolutionary elaboration of signaling by acting as a modular hub for the addition of new linkages in a pathway: since scaffolds create specificity by co-localization, addition of new components to a pathway may be easily accomplished by addition of a new interaction domain to the scaffold.

We reasoned that the same modularity that may have made scaffolded signaling pathways highly evolvable might also make them engineerable. In the present work we provide evidence that supports this premise: scaffolding can be exploited to generate novel, synthetic pathway linkages, and thus can be used to reshape the I/O of a signaling pathway. We reprogrammed the I/O behavior of the mating response pathway in S. cerevisiea, a classically studied MAP kinase signaling pathway that utilizes a scaffold protein, Ste5, to coordinate members of the pathway. We demonstrate that Ste5 can be used as an assembly platform for the synthetic recruitment of factors that positively and negatively regulate mating pathway activity. By incorporating these recruited modulators into simple transcriptional feedback loops, we demonstrate that the scaffold can be used as a junction to incorporate artificial feedback control into the pathway's response, dramatically altering its I/O. Using a limited set of molecular components, we constructed a variety of feedback circuit architectures and were able to generate diverse classes of behavior.

The results of the work described herein demonstrate the essential plasticity of a signaling network that utilizes scaffolding to define its connectivity. By exploiting these features, we were able to easily introduce new network linkages into a signaling pathway and dramatically reshape its signaling behavior. The variability of circuit behaviors achieved from a small collection of constituent parts supports the idea that the facile rewiring of regulatory linkages using scaffolding may have played an important role in facilitating the evolution of protein signaling networks. Furthermore, the approaches we used for rewiring the mating pathway may be generalizable to a variety of other signaling pathways both in yeast and other organisms.

To survive, cells must be able to properly sense and respond to information in a fluctuating environment. Inside a cell, information flows through discrete signaling pathways built around specific networks of protein interactions. However, it is difficult for the cell to ensure signaling fidelity in the face of a broad range of stimuli, a high internal concentration of proteins - many of which are related, and signaling pathways that use shared elements. For example in the yeast mating mitogen-activated protein kinase (MAPK) pathway, extracellular input activates a three-tiered kinase cascade (MAPKKK Ste11 phosphorylates MAPKK Ste7 which in turn phosphorylates MAPK Fus3), and leads to the mating response. However a similar, but distinct pathway, the filamentation pathway also uses the same MAPKKK Ste11 and MAPKK Ste7 to activate a related MAPK, Kss1. How are these pathways assembled via protein interactions and what prevents signals from crossing over to the wrong pathway?

Cells utilize a general strategy of subcellular compartmentalization to direct the flow of information. Spatial regulation is achieved by targeting proteins to organelles and membranes, or at a molecular level by assembling proteins into distinct signaling complexes. Modular docking interactions are the basis for wiring most three-tiered MAPK kinase cascades. However, docking interactions alone are not sufficient for the MAPKK Ste7 to discriminate between Fus3 and Kss1 MAPKs; they bind competitively to the same motif on Ste7. Instead a scaffold protein, Ste5, is required to selectively activate Fus3 during the mating response. Unexpectedly, the scaffold functions not only as a passive tethering platform but also actively controls signaling by co-catalyzing phosphorylation between the MAPKK Ste7 and the MAPK Fus3 (enhancing kcat nearly 5000-fold). This unique scaffold mechanism answers a long-standing question in the field about why there is limited signal crosstalk between two pathways that share the same upstream components, Ste11 and Ste7.