Synthetic Gene Networks

That is an innovative approach to control gene expression, which may help to turn on and off of oncogene and tumor suppressors in oncology AI-designed DNA sequences regulate cell-type-specific gene expression Researchers have used artificial-intelligence models to create regulatory DNA sequences that drive gene expression in specific cell types. Such synthetic sequences could be used to target gene therapies to particular cell populations. Machine-guided design of cell-type-targeting cis-regulatory elements Cis-regulatory elements (CREs) control gene expression, orchestrating tissue identity, developmental timing and stimulus responses, which collectively define the thousands of unique cell types in the body1,2,3. While there is great potential for strategically incorporating CREs in therapeutic or biotechnology applications that require tissue specificity, there is no guarantee that an optimal CRE for these intended purposes has arisen naturally. Here we present a platform to engineer and validate synthetic CREs capable of driving gene expression with programmed cell-type specificity. We take advantage of innovations in deep neural network modelling of CRE activity across three cell types, efficient in silico optimization and massively parallel reporter assays to design and empirically test thousands of CREs4,5,6,7,8. Through large-scale in vitro validation, we show that synthetic sequences are more effective at driving cell-type-specific expression in three cell lines compared with natural sequences from the human genome and achieve specificity in analogous tissues when tested in vivo. Synthetic sequences exhibit distinct motif vocabulary associated with activity in the on-target cell type and a simultaneous reduction in the activity of off-target cells. Together, we provide a generalizable framework to prospectively engineer CREs from massively parallel reporter assay models and demonstrate the required literacy to write fit-for-purpose regulatory code. https://lnkd.in/eaMDb3YF

Researchers at POSTECH, led by Professor Jongmin Kim, have developed a Synthetic Translational Coupling Element (SynTCE) that enhances the precision and integration density of genetic circuits in synthetic biology. Synthetic biology aims to assign new functions to organisms by utilizing natural and synthetic genetic regulatory tools. A key aspect is the polycistronic operon system, where multiple genes are expressed together to perform specific functions. However, designing sophisticated genetic circuits requires minimizing interference between biological parts and increasing encoding density for efficient integration. Traditional synthetic RNA-based translation regulatory parts have faced challenges in regulating multiple genes with high precision due to interferences in the protein translation process. To address this, Professor Kim's team focused on translational coupling—a natural gene regulation mechanism in operons where the translation of upstream genes affects the translation efficiency of downstream genes. By designing SynTCE to mimic this mechanism, they successfully integrated it with synthetic biological RNA devices, creating more efficient genetic circuits. Incorporating SynTCE into an RNA computing system significantly enhanced the integration density of genetic circuits, allowing precise transmission of input signals to downstream genes. This advancement enables systems capable of simultaneous control of multiple inputs and outputs within a single RNA molecule. Notably, SynTCE allows precise control of protein N-terminals and eliminates interference in protein translation. This capability can be applied in biological containment technology to selectively eliminate targeted cells and direct proteins to specific cellular locations. Such technology is expected to advance precise functional control and facilitate desired biological operations in cells. Professor Kim stated, "This research marks significant progress in enabling sophisticated and accurate genetic circuit design. This new design will be applied in various fields such as customized cell therapeutics, microorganisms for bioremediation, and biofuel production." The study detailing this work was published in the journal Nucleic Acids Research. For more detailed information, you can refer to the original publication: Goh, H., Choi, S., & Kim, J. (2024). Synthetic translational coupling element for multiplexed signal processing and cellular control. Nucleic Acids Research, 52(21), 13469. 

Richard Murray, a professor at Caltech, made this beautiful chart showing how the complexity of gene circuits (as measured by their number of "parts," or components) has scaled over time. (I'm sharing it below with permission.) We can learn many things from this chart. First, academic laboratories have been able to make some *really* complicated gene circuits. My friend, Jai Padmakumar, made the largest gene-circuit ever reported; it was described in a 2024 paper. Jai assembled 1.1 million bases of synthetic DNA into 110 distinct logic gates, and then partitioned that DNA across 66 strains of E. coli. Together, these engineered cells could compute the MD5 hashing algorithm. The problem is that the larger you make your gene circuit, the less "robust" or reliable the engineered cell becomes. Living organisms did not evolve to carry human-made gene circuits! Therefore, many synthetic biology efforts fail to scale into the real-world. The more complex a gene circuit, or the more genes it has, the less likely that it will be robust over time. More genes have more opportunities to break. (Note that this is not always the case in natural organisms. Many cells have evolved overlapping ways to regulate genes, such that if one breaks, others can fill in the gap. We're not good at emulating this synthetically, though.) The chart below shows this trend via the red, dotted line. Engineered cells that have been *commercialized* tend to have only a small number of engineered components; usually less than 10 synthetic genes in total. There is a drop-off in number of components as we move from the laboratory to the real-world. How can synthetic biologists solve this, and begin to build large gene circuits that are robust over time? Perhaps we should make it standard to grow engineered cells in a small bioreactor, perturb them with various stressors, and see how well the engineered cells hold up over time. We could record the number of generations that pass before a cell's functions break, and then report that value in the paper. (This is sometimes done, but not often.) Another option is to "merge" human-made designs, or AI-generated DNA, with continuous evolution. If we wanted to engineer a cell to break down plastic and recycle the atoms into a medicine, for example, then we could first build dozens of different gene circuit architectures (using high-throughput DNA assembly methods), put each gene circuit into a cell, and then do continuous evolution on each of them to see which one holds up best over time, with various stressors. We could sequence the populations over time, see which sequences hold up well, and use the data to train predictive models of "cellular robustness."

Design principles in synthetic biology Synthetic biology holds so much promise, but in reality things just don't work. We design a circuit, the parts work in isolation, we assemble them, and the thing doesn't behave. The underlying problem is that lack design principles for most of the functions we care about; electrical engineering had Kirchhoff’s laws and lots more. In biology the same parts behave differently in different contexts. The reasons are manyfold but it does not help that the machinery we need to assemble is made up of molecules, the signals we want to process are molecules, and the energy we need to invest are yet again molecules. There is no insulation. It’s all “wetware”. In my group’s work we have been interested in design principles (I hate using the word design in a biological context with a passion but cannot think of a better one in this context) related to Turing patterns: the stationary spatial patterns that emerge from reaction-diffusion systems and have been proposed as the mechanism behind everything from zebrafish stripes to digit spacing. A beautiful theory but fiendishly hard to implement using the toolbox of synthetic biology. And when you get there you really have to make sure that you convince others that it is a bona-fide Turing pattern, not just some other structure.  Together with Natalie Scholes, Dr. David Schnoerr, and Mark Isalan we undertook a comprehensive search: we analysed 7,625 candidate two- and three-node Turing networks across roughly 3 × 10¹¹ parameter combinations.  In a nutshell we found that over 61% of topologies can produce Turing patterns, far more than the literature had suggested. But the parameter regions where they actually do are tiny, often under 0.1% of the sampled space. Through painstaking analysis we found two compositional motifs that are almost necessary and almost sufficient (nothing in biology seems to be 100% necessary or sufficient): a positive feedback on at least one diffusing node, and a diffusion-mediated negative feedback loop on both diffusers. 96% of working networks contain both motifs, and 92% of networks containing both motifs work. A follow-up paper took the most robust three-node topology from the atlas and built it in E. coli. This is the first reasonably convincing engineered Turing pattern in a synthetic genetic system. A few things generalise to design principles in synthetic biology more broadly. 1. Network topology alone does not determine function. We had argued that before with Piers Ingram.  2. Robustness should be a design objective from the start. The most-studied Turing topology in the literature is nowhere near the most robust. 3. Theory and experiment have to be tightly coupled; without that cannot tell whether a failure is a design problem or a parameter problem. 4. Exhaustive enumeration beats clever case studies. You do not know which topology is best until you have checked them all. Abstraction or exhaustiveness are our only options.

🟥 Dual and Multiplex CRISPR Systems for Simultaneous Regulation and Editing of Genes CRISPR technology has moved beyond single gene targeting, paving the way for dual and multiplex CRISPR systems capable of simultaneous regulation and editing of multiple genes. These advances are essential for studying complex genetic networks, polygenic diseases, and synthetic biology applications, making gene editing more efficient, scalable, and precise. By allowing coordinated activation, repression, or modification of multiple genetic elements, these systems open up new possibilities for precision medicine, functional genomics, and cellular engineering. A key innovation in this field is the development of dual-function CRISPR systems, where catalytically inactive Cas9 (dCas9) is fused to different effector domains to activate one gene while silencing another in the same system. For example, dCas9-VP64 promotes gene activation, while dCas9-KRAB represses gene expression. Similarly, dCas9-p300 (a histone acetyltransferase) enhances transcriptional accessibility of chromatin, while dCas9-DNMT3A (a methyltransferase) promotes gene silencing through DNA methylation. These dual-function approaches are particularly beneficial for cancer research, as oncogenes can be silenced while tumor suppressor genes can be reactivated, creating more effective therapeutic strategies. In addition to dual-function applications, multiplexed CRISPR systems allow for the simultaneous targeting of multiple genes in a single experiment. One of the most promising strategies involves Cas12a (Cpf1), which can process multiple guide RNAs (gRNAs) independently, thus streamlining the editing of multiple disease-associated genes. In addition, polycistronic gRNA arrays enable coordinated control of gene networks involved in polygenic diseases such as diabetes, neurodegenerative diseases, and autoimmune diseases. These multiplexed approaches enhance our ability to correct multiple mutations simultaneously, making them extremely valuable for future gene therapy applications. Dual and multiplexed CRISPR systems are becoming more precise, efficient, and scalable with continued advances in AI-optimized gRNA design, improved Cas enzyme variants, and advanced delivery methods. These innovations are expected to revolutionize synthetic biology, regenerative medicine, and personalized gene therapy, enabling complex genetic modifications with greater accuracy and reduced off-target effects. As these technologies mature, they will unlock the full potential of CRISPR for multi-gene regulation, whole genome editing, and complex disease treatment. References [1] Nicholas McCarty et al., Nature Communications 2020 (https://lnkd.in/e8XzQAzG) [2] Amalie Brokso et al., Molecular Therapy 2025 (https://lnkd.in/eqcbi24g) #GeneEditing #MultiplexCRISPR #GenomeEngineering #GeneticTherapy #AIinBiotech #BiomedicalInnovation #BiotechBreakthroughs #CSTEAMBiotech

🌱 Advancing CRISPRi-Based Synthetic Gene Circuits in Plants Genetic (or gene) circuits are engineered networks of regulatory elements, for example promoters, transcription factors, and guide RNAs, that work together to control when, where, and how strongly genes are expressed. Much like electronic circuits, they can be designed to turn genes on or off, integrate multiple inputs, or respond dynamically to environmental or developmental signals. While gene circuits are well established in microbes and mammalian systems, their development in plants has lagged - largely due to limited modular tools and slow, low-throughput testing approaches. A new protocol published in Nature Protocols helps close this gap by providing a practical, high-throughput framework for designing and testing CRISPR interference (CRISPRi)–based gene circuits in plants. The protocol describes: - Design principles for CRISPRi-based gene circuits - Protoplast isolation from Arabidopsis leaves and roots, wheat and canola leaves, and moss protonemal tissue - A 96-well, high-throughput protoplast transfection pipeline with a dual-luciferase readout for rapid circuit testing Based on earlier CRISPRi circuit work, this workflow extends circuit testing to root protoplasts and enables functional validation within ~4 weeks across diverse plant systems, without stable transformation. Khan A, Herring G, Zhu JY, Petterson M, Lister R. Designing and testing CRISPRi-based synthetic gene circuits in plants. Nat Protoc. 2026 Feb 4. doi: 10.1038/s41596-025-01312-y. Epub ahead of print. PMID: 41639251. Publication: https://lnkd.in/eT_XhYJg #PlantSyntheticBiology  #CRISPRi  #GeneCircuits  #AgBiotech

Ever wondered how our cells achieve precise control over gene expression during differentiation, even when master regulatory proteins are expressed in overlapping patterns? 7 notes and all this music! Using 64,400 fully synthetic DNA sequences, Froemel et al. set out to uncover the hidden design principles in blood stem cell differentiation. Three surprising mechanisms allow enhancers to convert broad transcription factor (TF) gradients into highly specific gene expression: Occupancy-Dependent Duality: A single TF motif can act as both an activator and a repressor, simply depending on how much of the TF is predicted to occupy the enhancer. This creates a "filter" for specific TF activity, not just maximal or minimal. Cell-State-Dependent Duality: The same TF motif can be interpreted differently across various cell states, influenced by the cellular environment, co-factors, or post-translational modifications. Combinatorial Duality (The Big Surprise!): Combinations of activating TF binding sites can actually neutralize each other or even become repressive. This "negative synergy" is crucial for converting quantitative imbalances in TF expression into binary (on/off) activity patterns, ensuring mutual exclusivity of stem and progenitor cell programs. These principles allow to design enhancers from scratch with specificity to user-defined hematopoietic progenitor cell states. This work highlights the critical role of pairwise TF interactions in achieving regulatory specificity and offers transparent insights into how cells precisely control their fate. This research challenges previous assumptions, especially given observations in some cancer cell lines, and provides a foundational understanding of gene regulation in primary blood progenitors. #GeneRegulation #CellDifferentiation #Enhancers #Hematopoiesis #SyntheticBiology #Genomics #TranscriptionFactors #Biotechnology #Immunology #ImmuneCells https://lnkd.in/em-DkjtY

🚀Exciting news in the world of #SyntheticBiology and #GeneTherapy! 🎉 A team at Massachusetts Institute of Technology has introduced a powerful new framework called #DIAL - a system of programmable promoter editing that enables fine-scale, heritable, tunable control of transgene expression. Here’s what makes this work stand out: • With DIAL, the researchers insert “spacers” between the binding sites of a synthetic zinc-finger transcription factor and a core promoter, then remove them via recombinase to generate a range of expression “setpoints” from the same promoter. • The system supports temporal control via small-molecule modulation of transcription factors and recombinases meaning you can guide when and how much expression happens. • It works in primary cells and induced-pluripotent stem cells (iPSCs), and the edited promoter states are heritable - enabling mapping of transgene level → phenotype → cell fate. • This opens doors to more predictable and high-performance gene circuits, improved cell engineering, better translational potential for cell-based therapies. Imagine the implications: Instead of relying on a strong “on/off” promoter, we could dial in the right level of gene expression for a given therapeutic, differentiation protocol, or synthetic circuit. More control = fewer surprises, better outcomes. Feel free to share your thoughts — how might you use a system like DIAL in your own work? If you’re working in #CellEngineering, #RegenerativeMedicine, #TherapeuticGenes, or #Biotechnology, this is a must-watch development. #PromoterEngineering #TransgeneExpression #GeneCircuits #CellFate #iPSC #PrimaryCells #ZincFinger #Recombinase #TemporalControl #HeritableExpression #PrecisionBiology #NextGenTherapy #Bioengineering #MIT #NatureBiotechnology

Just recorded with Ron Weiss at MIT - one of the pioneers of synthetic biology who's been at this since 1996, back when he was helping set up a wet lab in MIT's CS department. The conversation really clarified something for me about why synthetic biology is finally working. Early on, Ron and others tried to build digital logic circuits in cells - AND gates, NOT gates, the whole thing. Computer scientists thought it made perfect sense. Biologists thought they'd come from outer space. ("Why would you want to make bacteria blink?") But digital logic doesn't scale in biology. You hit a wall around 4-input gates. Cells are already running near full capacity. Digital circuits need massive protein overproduction to maintain clear on/off states. The cells resist. They mutate your carefully designed DNA. They push back. The breakthrough wasn't abandoning the computational approach. It was realizing biology already computes - just not digitally. It uses analog signals, graded responses, intermediate values. More like neural networks than logic gates. We can now start to build actual neural networks inside cells. Perceptrons that do weighted calculations using RNA and proteins. These networks can be designed by biocompilers in silico and compiled into DNA sequences. Ron's work on self-amplifying RNA with Jacob Becraft and Tasuku Kitada led to Strand Therapeutics. One RNA molecule becomes 200,000 copies, each carrying both therapeutic payload and computational logic to detect cancer cells. They started first-in-human trials just over a year ago. Some melanoma patients - who'd exhausted all other options - have had complete remissions. The REACT project with Ken Shepard pushes even further: implantable devices that use optogenetics to control engineered cells, with your iPhone as the interface. One day, you might be able to literally dial up or down therapeutic protein production. We spent decades trying to make therapeutics that compute digitally, in binary, using yes and no, presence or absence. Turns out biology was already computing in a far more sophisticated way. We just had to learn its language. Ron started programming mainframes with punch cards at age 7. Now he's teaching living cells to think. This is the century of biology. And it's just the beginning. Full episode out now on frameshifts.bio.

Mapping 3.4 billion gene circuit designs with AI Designing synthetic gene circuits is like tuning a complex instrument in the dark. You have dozens of genetic parts—promoters, transcription factors, binding motifs—that must work together precisely, yet each combination behaves unpredictably due to context-dependent molecular interactions. Traditional approaches test circuits one at a time, making optimization painfully slow. Kshitij Rai and coauthors just changed the rules of the game. Their platform, CLASSIC (Combining Long- And Short-range Sequencing to Investigate genetic Complexity), combines Nanopore and Illumina sequencing to profile over 100,000 multi-kilobase gene circuit designs in a single experiment—then uses machine learning to predict the behavior of billions more. The workflow is elegant: pooled DNA assembly with barcodes, long-read sequencing to index composition-to-barcode mappings, phenotypic sorting in human cells, and short-read sequencing to link barcodes to function. The result? Quantitative expression data for 121,000 single-input circuits and 128,000 dual-input circuits, used to train neural networks that predict circuit behavior with r² values of 0.86–0.90. The insights are remarkable. High-fold-change circuits don't emerge from a single "optimal" design but from multiple balanced combinations of medium-activity components. AND-gate logic requires clustered transcription factor binding sites; OR-gates need interspersed patterns. These rules were invisible before—now they're learnable from data. The message: by scaling the design-build-test-learn cycle by orders of magnitude and combining it with ML, we can finally navigate genetic design spaces too vast for human intuition, accelerating everything from metabolic engineering to cell therapies. Paper: https://lnkd.in/eD9GPbuJ #SyntheticBiology #GeneCircuits #MachineLearning #Bioengineering #NGS #Nanopore #DeepLearning #Biotechnology #GeneticEngineering #AIforScience #CellTherapy #AIinBiology #SystemsBiology #Genomics #ProteinEngineering #BiologicalDesign