By Community Editor, Aakriti Jain and Guest Contributor, Grant Vousden – Dishington
Last weekend, approximately 300 scientists, industry professionals, and students gathered at MIT Building 26 for The Mammalian Synthetic Biology Workshop 2.0. On Twitter, the hashtag #msbw2 was used to cover the event live. The first mSBW workshop took place two years ago, and there were many new developments to be discussed in that short time span. Dozens of labs since 2013 have made use of CRISPR/Cas9-based genomic editing to pioneer everything from logic gates implemented with designer proteins to customized immune cells and in vitro disease modeling with organoids. Descriptions of these advancements were spread over more than 25 talks during the day-and-a-half meeting, and recurring trends arose quickly in the domain of immunological engineering, biocomputation, and ex vivo disease modeling.
After a brief introduction by MIT Professor Ron Weiss, the opening keynote of workshop was delivered by Professor Elaine Fuchs from The Rockefeller University. This plenary didn’t address synthetic biology per se but instead set the stage for later presentations by addressing mechanisms of skin stem cell differentiation and tumorigenesis. Stem cells are often envisioned solely as progenitors of terminal and functional cell types, but Dr. Fuchs reminded us that many of these cells are quiescent. They themselves do not differentiate but nonetheless exchange transcriptional signals, including Noggin and Wnt, with their active counterparts. The build-up of these intercellular signals is crucial for certain cell fates and in transitioning quiescent stem cells to active status. The plenary concluded with a preview of how synthetic biology might play a role in cancer therapy; specifically, knock-out of genes Cdkn1a or Nfe2/2 in tumor cells makes them more responsive to drug treatment.
Wyss Institute and Harvard Medical School Professor George Church continued to build momentum after the keynote by discussing foundational technologies that make the new wave of synthetic biology possible, as well as their associated concerns and risks. He also advocated for a more standardized design process for CRISPR/Cas9 systems. Chief among these issues was the potential of an off-target mutation, which can activate or inactivate genes never intended to be modified. Professor Church mentioned the recent publication of Genome-wide Unbiased Identification of DSBs Enabled by Sequencing (GUIDE-Seq) for identifying unintended double strand breaks (DSB) and cleavages in CRISPR/Cas9 RNA-based editing, which doesn’t fix the problem but makes it easier to identify when off-target editing has occurred. This was only one of many other tools mentioned in the talk that hold the power to expedite the synthetic biology design process.
Professor Christina Smolke of Stanford University was slated to give the third talk, but due to unforeseen circumstances, her talk titled RNA Devices to Control T Cells was instead delivered by her graduate student, Remus Wong. This presentation was the first to mention what would become a recurring concept at the workshop: chimeric antigen receptors (CAR). When genetically modified to express CARs, T cells are able to identify tumor cells and participate in antitumor activity. However, what makes the approach innovative is the combined use of a ribozyme switch to modulate the CAR transgene and a microRNA switch to modulate the endogenous gene, IL-2, which regulates T cell growth and survival. Wong demonstrated that these two can be controlled by the clinically relevant drug Leucovorin (Folinic acid), introducing the possibility of physicians enhancing or attenuating the therapeutic effects of designer T cells on cancer patients.
Directing Cell Phenotype
Not even an hour after George Church raised questions about CRISPR/Cas9 specificity, Duke University Professor Charles Gersbach discussed epigenomic engineering to control cell phenotypes that. Intentionally or not, this talk answered several concerns raised by Church. In the Gersbach lab, students Pablo Perez-Pinera and Isaac Hilton have been pursuing methods for, as Professor Gersbach puts it, programming gene expression based on TALE transcription factors and acetyltransferase, respectively. These programmed modifications to CRISPR/Cas9 complexes delivered with lentiviral vectors allowed the researchers to modulate activation of endogenous genes, like MyoD, via control of histone mechanisms. The work demonstrates one avenue for enhancing specificity of genomic engineering techniques: combining multiple regulators, each tuned by a different signal.
UCSF Professor Wendell Lim took the use of orthogonal control of specific cells a step further in his ambitiously named presentation, Redesigning The T Cell. Reiterating the importance of CAR in engineering T cells to recognize specific antigens, Professor Lim mentioned one drawback to their clinical application: they are successful in treating B cell cancers but not solid tumors. To augment the clinical use of designer T cells, he proposed methods to not only target disease and cancer-specific pathogens, but also to give patients and physicians temporal control of synthetic T cells and where they operate within the body. Genetic modifications could allow for titratable control over immune responses, not unlike the Smolke lab’s use of Leucovorin-modulated switches. Before concluding his talk, Professor Lim spoke of using multiple CAR species to parallelize and specify T cell activity, essentially an immunological AND-gate, introducing logic gates implemented with proteins as a second theme that became common in later talks.
Professor Wilson Wong of Boston University put the emphasis on biocomputation into full swing, introducing his lab’s new, as yet unpublished Boolean Logic and Arithmetic through DNA-Excision (BLADE) platform for robust and scalable design of boolean control systems in situ. The short talk was followed by a somewhat rapid change in subject by Michael Todhunter, a graduate student inZev Gartner’s lab at UCSF, who focused on using synthetic biology to rapidly create 3D tissue structures. In particular, Todhunter emphasized the tissue of mammary glands for their diverse cellular content, which includes endothelial, epithelial, fibroblast and other species of interest. By printing spots of designer DNA on the surface of particular cells, they can adhere to one another via DNA hybridization. After many iterations of this simple method, dense and complex tissues of many chosen cell-types can be created.
Design and Optimization of Genetic Circuits
Professor Michael Elowitz of Caltech resumed the discussions after lunch, kicking off a session dedicated to the optimization of genetic circuits. The talk emphasized four chromatin regulators implicated in nearly all biochemical mechanisms – or at least the ones the Elowitz lab is interested in: EED, KRAB, Dnmt3b, and HDAc4. The two former regulators modulate histone methylation while the latter two are involved in DNA methylation. One of the many benefits of epigenetic regulation, according to Professor Elowitz, is the variety of types and timescales of “memory,” in this case referring to the ability to tune the chromatin regulators based on previous states of activity or silence. By combining the intrinsic memories of the four chromatin regulators above, the Elowitz lab is able to customize cell responses to certain stimuli to be either stochastic or all-or-nothing, as dictated by the needs of the physician or researcher. Despite the fact that they are epigenetic mechanisms, Professor Elowitz interest in these processes lies in understanding synthetic biology at the level of the individual cell. This is in contrast to his colleagues at the workshop proposing control of entire immune systems, seeing each cell as its own agent and decision maker. Being able to design entire physiological systems will depend crucially on an understanding of cell-to-cell variability in their mechanisms.
Following the talks about epigenetic regulation, Professor Leonidas Bleris from UT Dallas took aim at one of the fundamental problems in mammalian synthetic biology: controlled gene delivery in vector-based systems. Even in organisms where it is known how to deliver the transgene of interest, controlling the duration and number of copies of the gene in a given cell is a separate problem. Self-destruction, or in the case of genetic engineering with CRISPR systems, self-cleaving, is one conceptually simple mechanism for adjusting these variables. A recent paper from the Bleris lab suggested a protocol for creating such a system, where the Cas9 protein is guided via RNA to cleave the delivery vector at a specified location, inactivating further copies of the gene. The experiment was performed in human embryonic kidney tissue, further validating its usefulness in mammalian systems and, potentially, in the clinic.
The two short-talks wrapping up this session were given by two MIT researchers, Silvana Konermann, a graduate student in Feng Zhang’s lab, and Dr. Tasuku Kitada, a post-doctoral fellow in Ron Weiss’s lab. Both talks concerned issues relevant to the application of custom RNA sequences to disease. Acknowledging that the set of biological problems that can be solved and addressed by targeting single genes is relatively small, Konermann spoke of work stemming from her recent Nature paper on using single-guide RNA (sgRNA) to transcriptionally activate ten genetic targets simultaneously. Dr. Kitada’s talk also focused on RNA-based delivery methods, particularly the problem of off-target toxicity from “spilling” RNA to unintended tissues. Drawing on content from his recent review paper, he suggested ways to target RNA to specific types of tissue, e.g. HeLa cells. His idea forms the basis for what he calls “smart vaccines,” injected agents that remain dormant until they come into contact with one of the tissues they’re designed to destroy.
Functions of Multicellular Systems
Professor Ed Boyden, Principal Investigator of MIT’s Synthetic Neurobiology Group and co-inventor of optogenetics, provided a survey of the many different methods now available for monitoring and controlling biological systems in vivo. Neuroscientists are no strangers to Professor Boyden’s work and its applications to controlled stimulation of neural circuits and genetically identifiable neuron sub-species, but he was quick to affirm that the technique is more widely applicable. The light-sensitive ion channels and proton pumps that make optogenetics possible can also be installed in genetically modified cells from any other mammalian system. Furthermore, Professor Boyden expanded on the newest technique to come from his lab, Expansion Microscopy – stylishly abbreviated as “ExM.” In this technique, cells are fixed, labeled with immunofluorescent markers for relevant cytological features, treated with polyacrylate to create a swellable gel, and digested with enzymes. Hydrating the gel then expands the sample, still containing the fluorescent markers, augmenting the microscopic resolution by a factor of roughly 4.5, sufficient to image nanoscale details in many samples.
Emulate President and CSO Dr. Geraldine Hamilton gave a talk that switched gears a little with a presentation on her companys development of a variety of organ-on-chips methods. Notably, their lung-on-chips showed great promise with on-chip simulations of endothelial-epithelial tissue-tissue interactions, lung inflammation immune response, and even beating cilia. The frequencies of these beats were almost indistinguishable from in vivo measurements. Dr. Hamilton also introduced various other technologies that Emulate is currently working on, including human gut-, liver-, and kidney-on-chip, for a wide range of applications in drug discovery & development. These include safety and therapeutic target identification, testing of everyday chemicals, such as in cosmetics and household chemicals, clinical trials on chips for larger and more diverse cohorts, and other uses for induced pluripotent stem cells (iPSC). While not substitutes for full organ systems, these in vitrotissues maintain critical features at the multicellular scale to understanding their mechanisms. Indeed, the model of the human gut even included a partial microbiome. With a growing pipeline of organs and disease models and extensive characterization of morphology and function, this technology is sure to be of great use to synbio researchers in the future.
Jonathan Brunger, a student researcher in orthopedics at Duke University, and Dr. Daniel Woodsworth, a researcher at the Michael Smith Genome Sciences Centre in British Columbia, rounded off this session with more example applications of synthetic biology to therapeutics. Brunger’s current aim is to use the tools discussed at the meeting to regulate and reduce inflammation responses in tissues treated with regenerative medicine techniques, improving the prospects and outcomes of these still-new approaches. In particular, he studies the pro-inflammatory physiology involved in osteoarthritis, a condition especially affecting aging population and common cause of chronic pain. Dr. Woodsworth broached the question originally poised by Professor Bleris of how to deliver medicine to biological systems of interest. He is especially interested in lymphocytes and the granzyme-perforin pathway as a hackable delivery system for anti-tumor treatments. His experiments achieved success in apoptosis-resistant tumors by fusing an “orthogonal” toxin to granzyme that selectively affected cancerous tissue, effectively turning the lymphocyte into a cytotoxic vector.
Gene Regulation and Epigenetic Control
Professor Elowitz’s description of four chromatin regulators being used for epigenetic control was strongly supported and augmented by Professor Karmella Haynes from Arizona State University. A2011 paper by Professor Haynes and her collaborator uses a human chromatin protein, Polycomb, and its xenogenic homologues to reactivate silenced loci, again demonstrating that genetic modulation is possible in vivo via epigenetic mechanisms. With proper design, synthetic chromatin regulators could enhance the effectiveness of CRISPR editing, perhaps even the shortcomings mentioned in prior talks. The professor ended the presentation by challenging intentions to use genomic editing techniques in human disease treatment with a central question: can histones and chromatin regulators interfere with CRISPR/Cas9 tools?
The final talk of the first day, delivered by Professor Yaakov “Kobi” Benenson of ETH Zurich, returned from pre-clinical applications to basic research by discussing the design of synthetic biomolecular systems not only to perform a specific function but also to engage those functions only when certain conditions are true. First-order logical operations – AND, OR, NOT, XOR, NOR, NAND, etc. – form the basis of computation in silicon systems, and the same conditional logic can be applied to biology. Whereas computers rely on transistors, Professor Benenson and his colleagues propose using RNA interference (RNAi) as the basis for biomolecular logic. A simple but novel application of such a system would be to switch on or off a feedback-amplified biosensor. Normally, such systems are dubious, because when there is no input to the sensor, there is usually “leakage” in the system that leads to self-amplification. With an AND-gated version of such a sensor, the feedback mechanism can be disabled when there is no input present in the environment.
Aside from discussions of how and when to implement logic design into synthetic biological circuits, Professor Benenson’s lecture was an apt end to the first day of plenaries for another reason: he reminded the audience that the concept and promises of synthetic biology today are not actually new. In particular, he cited work from Motoyosi Sugita and colleagues who described theoretical applications of logic circuits to analyze in-vivo physiological activity. The work was published in the 1960s.