Smolke:Research

=Research Overview= Engineered biological systems that process information, materials, and energy hold great promise for developing solutions to many global challenges, including energy and food production, materials synthesis, and medical advancement. Our ability to engineer biological systems is limited by the foundational tools available for programming cellular behavior (i.e., getting information into and acting on information in living systems) and our understanding of how such systems should be constructed (i.e., probing and accessing information in living systems). By developing genetically encoded technologies for reporting on, responding to, and controlling intracellular components in living systems, we are addressing prominent challenges faced in basic and applied biological research. Our research efforts focus on the design of new molecular tools for performing information processing, computation, and control functions in living systems and the application of these tools to programming biological systems. The resulting advances in our ability to transmit information to and from living systems and implement control within cells themselves, will broadly transform how we interact with and program biology, providing access to otherwise inaccessible information on cellular state and allowing sophisticated exogenous and embedded control over cellular functions.

=Component and Device Engineering=

Engineering RNA-based cellular information processing, communication, and control devices
Researchers: Ryan Bloom, Drew Kennedy, Jay Vowles, Josh Wolf

We have developed generalizable design strategies for a broad class of RNA molecules, called RNA devices, that process and transmit user-specified molecular input signals to targeted protein outputs, thereby linking molecular computation to gene expression. We proposed a first-generation framework for the construction of single input-single output RNA devices based on the modular assembly of three components exhibiting basic functions (Win et al., Chem Biol 2009): a sensor component, made of an RNA aptamer or binding element; an actuator component, made of an RNA gene regulatory element; and a transmitter component, made of a sequence that couples the sensor and actuator components. The transmitter component regulates the activity of the actuator component based on the binding state of the sensor component, providing genetic insulation and a standardized communication interface. We initially demonstrated RNA devices that function as single-input Buffer and Inverter gates that convert a molecular signal to increased and decreased gene expression output, respectively (Win et al., PNAS 2007).

One of the key contributions of our work to the field of molecular design is the development of composition frameworks supporting forward engineering and modular assembly approaches. Our design approach thus enables a modular device platform that supports the tailoring of device function (sensing, actuation, computation) through the direct swapping of component modules without time-intensive redesign of the device. Using our design strategy, we have demonstrated novel cis-acting RNA devices that function through a variety of regulatory mechanisms, including ribozyme cleavage (Win et al., PNAS 2007; Win et al., Science 2008), alternative splicing (Culler et al., Science 2010), and RNase III cleavage (Babiskin et al., Mol Sys Biol 2011; Babiskin et al., Nuc Acids Res 2011). We have also demonstrated novel trans-acting RNA devices that function through the RNA interference pathway (Beisel et al., Mol Sys Biol 2008; Beisel et al., Nuc Acids Res 2010) and antisense silencing (Bayer et al., Nat Biotech 2006). By accessing a variety of gene regulatory mechanisms, we have built tailored RNA devices that function in diverse cell types and organisms and that respond to endogenous and exogenous small molecule inputs (Win et al., PNAS 2007; Beisel et al., Mol Sys Biol 2008) and endogenous and heterologous protein inputs (Culler et al., Science 2010). Ongoing research is extending the capabilities of these design platforms by optimizing device architectures and integrating new actuation mechanisms.

Back to Top

Engineering higher-order cellular information processing devices
Researchers: Leo d’Espaux, Katie Galloway, Yen-Hsiang Wang

We have demonstrated the extension of the single-input RNA device frameworks to the construction of higher-order devices that perform multi-input signal processing in living systems (Win et al., Science 2008; Culler et al., Science 2010), further supporting the power of the developed modular assembly strategy. As an example, my laboratory described extended architectures for rationally assembling RNA components (sensors, actuators, transmitters) into multi-input signal integration devices and built genetic devices that function as logic gates, signal and bandpass filters, and exhibited cooperativity. Our research has demonstrated that the developed design frameworks provide a general approach for the forward engineering of multi-input devices, supporting the combinatorial assembly of many information processing, transduction, and control devices from a smaller number of components. More generally, this work has demonstrated the application of synthetic biology design strategies to scalable platforms for genetic device design and that advances in engineering design can transform the scale, efficiency, and speed with which we can engineer cellular behaviors.

Ongoing research is extending the complexity of information processing schemes that can be reliably implemented in cellular systems by building regulatory networks composed of distinct genetic devices. For example, we are examining device design schemes that will allow for ratiometric or differential sensing of multiple molecular signals, signal amplification, error detection, and signal restoration. We are also examining the implementation of single and multi-layered architectures in genetic pathways that will allow scaling of computational complexity (i.e., multi-input processing and multi-output control), while maintaining reliable information transmission between different layers (i.e., genetic constructs or transcripts).

Back to Top

=Enabling Technology Development=

Developing computer-aided design (CAD) tools for RNA device design and optimization
Researchers: Ryan Bloom

The integration of engineered information processing and control devices with cellular networks to effectively control systems-level functions often requires precise tuning of the response properties of the engineered molecules to match the quantitative signaling range of the system of interest. To address this challenge, we have been developing computational design tools that enable the rational tuning and optimization of RNA device function to streamline the process of integrating these genetic elements into biological systems. Specifically, we developed kinetic models of RNA device function based on detailed molecular descriptions of functional mechanisms (Beisel et al., PLoS Comp Biol 2009). We used these models to examine the effects of rate constant variation on the resulting response curve to elucidate strategies for rationally tuning the performance of an RNA device through modulation of rates associated with different mechanistic steps. Our modular assembly platform results in device functions (sensing, actuation, computation) being encoded in distinct and independent components, thereby resulting in individual model parameters mapping to individual components. We used this unique property of our molecular platform to map parameter tuning strategies predicted through the computational models to direct sequence changes of the appropriate components in the device to generate an early sequence-to-function framework for optimizing RNA switch function (Beisel et al., Mol Sys Biol 2008).

Ongoing work is focused on refining and improving the CAD tool for RNA device design. We are examining different RNA folding algorithms and determining the most appropriate way to process the structures and energies that are predicted from the programs to obtain an accurate predictive measure of conformational partitioning in vivo (Beisel et al., Mol Sys Biol 2008), which can then be fed into our computational models to predict the quantitative device response from a primary RNA sequence. Future work will integrate structural studies with our computational models and cellular systems to achieve an integrated understanding of device folding, dynamics, and activity. The longer-term goal of these efforts is to develop a robust sequence-to-function framework that supports the in silico design and optimization of the quantitative response properties of RNA devices from well-characterized component libraries.

Back to Top

Developing high-throughput strategies for the generation of new RNA component functions
Researchers: Andy Chang, Joe Liang, Brent Townsend, Jay Vowles

The implementation of genetic devices within biological systems is broadly limited by our ability to tailor the input-output activities of these devices to applications of interest. Libraries of well-characterized RNA components (encoding sensing, actuation, and computation functions) that are compatible with the device platforms are necessary to realize the full advantages of our design approach. As one example, the generation of RNA aptamers exhibiting desired in vivo affinities and specificities is a current challenge in device design due to inefficiencies and limitations in scaling with existing selection strategies. We are developing a high-throughout in vivo selection strategy in which RNA sensors are screened for desired binding activities within the device platform to enable a solution-based selection for aptamer sequences that are compatible with the device framework and exhibit desired binding properties in the cellular environment. We have established an efficient high-throughput screening strategy for device activity based on fluorescence activated cell sorting by linking the output of an RNA device to the expression of a fluorescent protein, where sequences within the RNA device can be randomized to generate device libraries to search for novel component functions (Liang et al., in preparation). We are applying this in vivo screen to the generation of RNA components exhibiting enhanced or novel activities.

The de novo generation of RNA aptamers will likely require searching a larger sequence space than can be directly achieved in vivo. As such, we are currently developing a high-throughput in vitro selection strategy for the RNA device platform where partitioning of active members will be conducted through PCR recovery, bead-based partitioning, or capillary electrophoresis. We will develop a streamlined in vitro-in vivo selection strategy that will be applied with high throughput characterization strategies, such as surface plasmon resonance (Win et al., Nuc Acids Res 2006), to develop libraries of novel sensing functions to diverse ligands of interest, including natural product families, central metabolites, clinically-relevant drug molecules, disease markers, and exogenous chemicals with properties appropriate for use as trigger molecules.

Back to Top

=Systems Engineering of Cellular Behaviors=

Engineering yeast as a natural product biosynthesis platform
Researchers: Stephanie Galanie, Josh Michener, Michael Siddiqui, Kate Thodey, Isis Trenchard

Natural products and compounds derived from or inspired by natural products make up a large fraction of drug molecules. Traditional synthesis strategies based on recovery from natural sources and chemical synthesis approaches present many challenges associated with the purity, scale, and complexity of the compounds. The engineering of biosynthetic pathways in microbial hosts represents a newer approach to chemical synthesis with exciting potential. We are integrating recent advances in synthetic biology to transform the complexity of genetic networks that can be engineered in biological systems to engineer scalable cellular biosynthesis schemes for important classes of natural products. Specifically, we have focused our efforts on purine alkaloid (Win et al., in preparation) and benzylisoquinoline alkaloid (BIA) biosynthesis pathways (Hawkins et al., Nat Chem Biol 2008). The BIAs are a large class of plant secondary metabolites that exhibit diverse pharmacological activities, including anti-HIV, antimicrobial, anticancer, antineoplastic, vasorelaxation, and cholesterol-lowering activities, and activities for treating cardiovascular and autoimmune diseases. Although the BIAs populate a chemical space with many compelling activities, there currently exists no general source for the BIAs as many of the molecules are too complex for synthetic chemical methods and only a select few accumulate to substantial levels in the native plant hosts. The complexity associated with the BIA biosynthesis pathway, in terms of number of enzymes and complexity of chemistries and regulatory strategies, requires the integration of new approaches to cellular biosynthesis for effective implementation. We demonstrated one of the first examples of biosynthesis of an array of BIA molecules in a microbial host (Hawkins et al., Nat Chem Biol 2008), through the integration of enzymes from plants, bacteria, and humans. Ongoing research efforts are directed to the extension of the synthetic BIA pathway into key branches - the early BIA branch (to enable total synthesis from common precursors) and specialty chemical branches (to enable synthesis of morphinan, benzophenanthridine, and bis-BIAs).

In conjunction with the above pathway reconstruction work, we are developing synthetic biology platforms that will advance the application of cellular biosynthesis strategies to natural product drug discovery, development, and production. As one example, we are pioneering approaches for noninvasive and real-time sensing of metabolite levels based on implementing RNA devices that sense key metabolite or cofactors and regulate fluorescent proteins in response to changing input concentrations (Win et al., PNAS 2007). This tool is currently being used to screen enzyme libraries for enhanced activities. As a second example, we will apply RNA devices that regulate the levels of target pathway enzymes in response to changes in the concentrations of key metabolites and cofactors to implement closed loop embedded control of biosynthesis system behavior. As a third example, we are developing approaches for biosynthesis compartmentalization and specialization. These new approaches will provide a general platform for scalable production of diverse natural product families.

Back to Top

Engineering next-generation molecular and cellular therapies
Researchers: Ryan Bloom, Leo d’Espaux, Katie Galloway, Megan Palmer, Jay Vowles, Kathy Wei, Josh Wolf

Cellular behavior is encoded and controlled by complex genetic networks. Synthetic genetic devices that interface with native pathways can be used to change natural networks to implement new forms of control and behavior. We have integrated engineered RNA devices into biological systems to reprogram cellular behavior. Our efforts to date have focused on the design of next-generation molecular and cellular therapeutic strategies. As one example, we demonstrated the application of engineered RNA devices as autonomous controllers over cellular behavior, and specifically as molecular therapies targeted to diseased cells (Culler et al., Science 2010). We engineered RNA devices that detect increased signaling through disease-associated pathways in human cells and rewire these pathways to produce new behaviors, thereby linking disease markers to noninvasive sensing and reprogrammed cellular fates. As another example, we demonstrated the application of engineered RNA devices as key controllers of cell-fate in cellular therapeutic strategies (Chen et al., PNAS 2010). In particular, we implemented drug-responsive RNA devices in mammalian and human T cells that target specific components of the T-cell signaling pathway to regulate T-cell proliferation in vivo through an exogenously applied drug. This work has highlighted advantages afforded by these RNA-based devices in translation to therapeutic applications and addressing key challenges in the design of safer and more effective therapeutic strategies.

Ongoing research efforts are focused on extending the integration of RNA devices with different cellular pathways to achieve reprogramming of diverse cellular behaviors. As one example, we are continuing efforts on the T-cell engineering project to develop tailored RNA devices that respond to clinically-approved drug molecules and to develop integrated systems designs that provide a more robust response. The longer-term goal of this research will be to conduct systemic in vivo studies and ultimately human clinical trials (in collaboration with Dr. Michael Jensen at the Seattle Children’s Research Institute). Future work will explore the implementation of these genetic devices in other cellular therapy applications, such as stem cells. We are also examining the implementation of RNA devices in pathways associated with MAPK signaling and cell cycle.

Back to Top