MIT study reveals gene syntax shapes synthetic circuit protein output

Gene order is becoming a design rule

The most important lesson from MIT’s new synthetic biology study is simple: DNA sequence alone does not determine what a circuit will make. The physical arrangement of genes along the chromosome, what the team calls gene syntax, changes transcriptional feedback, chromatin behavior, and ultimately protein output.

Published in *Science* on April 30, 2026, the paper, *Gene syntax defines supercoiling-mediated transcriptional feedback*, shows that tandem, divergent, and convergent layouts can push neighboring genes in very different directions. For engineers building protein production systems, gene therapies, or cell-based medicines, that matters because predictable expression is the whole game.

What the MIT team set out to test

Christopher P. Johnstone, the lead author and a postdoc at MIT, and senior author Kate E. Galloway built the study around a question that has practical consequences for synthetic circuit design: if two genes sit next to each other, does their order and orientation change how much protein each one makes?

To answer that, the researchers used synthetic two-gene reporter circuits integrated into mammalian cell lines at a fixed safe-harbor locus. That placement choice was important because it controlled for genomic location effects, letting the team isolate the influence of gene arrangement itself. They then induced one gene with a small molecule, creating a controlled burst of transcription that let them watch how supercoiling spread through the local DNA neighborhood.

The team measured the output at multiple levels. They tracked protein and mRNA in single cells by flow cytometry, then paired that with CUT&Tag, GapRUN, and Region Capture Micro-C to examine chromatin state, supercoiling density, and chromatin folding. The result is a layered picture of how a synthetic circuit behaves not just as a sequence, but as a physical structure.

Why syntax changes expression

The core mechanism reported by MIT News is a structural one. When transcription runs through DNA, it does not leave the helix unchanged. Upstream DNA becomes looser, while downstream DNA becomes more tightly wound. That changes access for RNA polymerase and creates a feedback loop between one gene’s activity and its neighbor’s response.

In other words, the orientation of a gene pair can help determine whether transcription helps or hinders the next promoter in line. Gene syntax is no longer just a naming convention for biologists reading a plasmid map. It is a physical variable that shapes how the circuit behaves in real cells.

The study frames this as supercoiling-mediated transcriptional feedback, which is a useful way to understand why adjacent genes do not always act independently. The DNA itself carries part of the signal.

The syntax patterns that mattered most

The data showed clear differences between the major layouts:

• Divergent syntax produced amplified, correlated expression.
• Tandem syntax reduced expression from the downstream gene.
• The effects generalized across integration method and cell type.

That is the practical takeaway for circuit builders. Divergent arrangements can make two genes rise and fall together, which may be useful when a system needs coordinated output. Tandem arrangements, by contrast, can dampen the gene that sits downstream, which could be used deliberately when one component needs to be restrained.

The fact that these effects held across delivery methods and cell types makes the findings much more than a one-off quirk of one experimental system. The bioRxiv preprint behind the work also emphasizes that syntax can tune mean, variance, and stoichiometries of expression across different delivery routes and cell types, which is exactly the kind of control synthetic biology has been chasing.

Why this changes how engineers should think about circuits

For years, circuit design has leaned heavily on biochemical logic: activators, repressors, promoters, enhancers, and regulatory proteins. Those tools still matter, but this study argues that physical placement on DNA is an additional control layer. That is a big deal for anyone trying to build compact systems where every base pair has to earn its keep.

The strongest engineering implication is that syntax can be used to shape not only average expression, but also noise and relative abundance between two genes. That opens a route to tighter control over stoichiometry, which is essential in systems where protein balance determines whether a circuit works cleanly or fails noisily.

This is especially relevant for:

• Protein production platforms that need stable yield
• Gene therapy constructs where overexpression can be harmful
• Cell-based medicines that depend on precise ratios of engineered proteins
• Multi-gene circuits that must coordinate outputs without bulky regulatory logic

A more compact path to predictable synthetic biology

MIT’s work builds on earlier modeling from 2022 that explored tandem, divergent, and convergent arrangements. What changed here is the experimental mechanism. The 2026 study adds direct evidence that supercoiling-mediated feedback explains how neighboring genes influence one another in mammalian cells.

That shift matters because it turns a once-theoretical design idea into something engineers can use. Instead of relying only on extra regulatory parts to tune expression, designers may be able to get more predictable behavior by placing genes in the right orientation from the start. In compact circuits, where space is limited and complexity raises the risk of unwanted interactions, that could simplify construction without sacrificing control.

Kate E. Galloway’s group and collaborators from institutions including the Massachusetts Institute of Technology, the University of British Columbia, Leiden University Medical Center, Caltech, the Ragon Institute of MGH, MIT and Harvard, and others connected across the Boston/Cambridge area, Pasadena, California, Leiden, the Netherlands, and Vancouver, British Columbia, have effectively expanded the synthetic biology toolkit. Gene syntax now joins the list of design variables that can be tuned with intent rather than discovered by trial and error.

The bigger message for biotech

The headline result is not just that one layout outperformed another. It is that DNA folding and transcription are linked tightly enough to make local architecture a meaningful engineering parameter. For synthetic biology, that means more than better reporter constructs. It points toward a future where the order and orientation of genes are chosen as deliberately as the promoters that drive them.

For anyone building the next generation of engineered cells, that is the shift to watch: predictable output may depend as much on how genes are arranged in three dimensions as on what those genes are.