MIT researchers have built what may be the closest biological equivalent yet to a printed circuit board: a network of living bacterial colonies that can receive molecular inputs, process information, and send chemical signals to specific destinations. Instead of moving electrons through wires, the system moves small molecules between engineered cells. Its basic components are bacterial “transistors” that act as biological switches, along with relay strains that pass signals from one colony to the next. Printed onto a nutrient-rich agar surface inside a Petri dish, these colonies form living circuits capable of performing logic operations, combining signals, and directing information through predetermined pathways. The work suggests that computation could eventually become an integrated function of living systems, allowing plants, microbes, or other biological environments to sense conditions and respond without conventional electronic hardware.
The system was developed by a team at the Massachusetts Institute of Technology led by Christopher Voigt, head of MIT’s Department of Biological Engineering. Hamid Doosthosseini PhD ’25, an MIT postdoctoral researcher, is the lead author of the study, which was published in Nature Chemical Biology. The researchers designed two bacterial transistor types and three additional strains that function as molecular relays. Together, these five strains provide a modular toolkit for constructing a broad range of biological circuits. In their experiments, the team assembled networks that could add two or three inputs, route a signal to a selected destination, and perform several kinds of logic operations. The largest circuit demonstrated in the study contained 24 separate bacterial colonies linked through chemical communication.
In an electronic circuit, a transistor controls the movement of electrical current. It can be switched on or off and can help determine how signals are processed. The MIT researchers created an analogous system in which bacterial cells control the production and movement of signaling molecules. Their transistors are based on Pantoea agglomerans, a bacterium that commonly lives on surfaces, including plants. Each engineered strain responds to a molecule called OC-6, which serves as a control input. One transistor is activated when OC-6 is present, while the second is inhibited by the same molecule. Both types also detect another molecule, OC-12. Depending on whether OC-12 is present and whether the transistor is active, the cells produce a third compound, OHC-14, which serves as the output signal.
This arrangement allows a single bacterial colony to behave differently depending on the combination of inputs it receives. OC-6 effectively determines whether the transistor is permitted to operate, while OC-12 supplies the information that the circuit must interpret. The resulting OHC-14 signal can then be detected by another engineered strain. Rather than placing every component inside one cell, the researchers distributed the functions across different bacterial populations. This separation is important because conventional synthetic biology circuits often rely on many transcription factors operating in the same cell. If too many components are packed together, they can interfere with one another, create unwanted molecular crosstalk, or consume so much of the cell’s protein-production capacity that the system becomes unreliable.
The relay strains solve part of that problem by translating one chemical signal into another. When a relay colony detects OHC-14, it produces a molecular output that can be recognized by a downstream transistor. This creates a biological equivalent of wiring. The researchers printed colonies approximately five millimeters apart on agar, carefully arranging them so that each colony primarily communicated with its nearest neighbor. The spacing limits the range of each chemical signal and helps establish a directional flow of information across the plate. In this way, the geometry of the printed pattern becomes part of the circuit design. A colony’s position determines which signal it receives and which neighboring colony receives its output, much as the paths etched into an electronic circuit board determine how current travels.
Using this spatially organized system, the researchers demonstrated several logic functions. A transistor could perform different operations depending on where it was positioned and what signals reached it. The circuits included multi-input operations, “or” gates, and implication gates, in which one condition determines whether another condition is permitted to produce an output. The team also built networks that combined multiple signals and used a biological demultiplexer. In electronic systems, a demultiplexer takes one incoming signal and directs it to one of several possible outputs according to a control signal. In the bacterial version, chemical inputs determined which relay pathway became active, allowing the signal to be routed to a selected group of downstream colonies.
The most complex demonstration was a circuit capable of adding two inputs together. In digital electronics, addition is built from logic operations that calculate both a sum and a carry signal. The bacterial network used linked colonies to process those inputs through successive molecular steps. Although the biological implementation operates far more slowly than an electronic processor, it illustrates the central principle of modular computation: relatively simple components can be connected to produce a more complicated function. The researchers also showed that their architecture could process several signals at the same time. Because each colony performs a defined task, new circuits can potentially be created by changing the number, identity, and arrangement of the printed strains rather than redesigning one enormous genetic program inside a single cell.
The speed of the living circuits is one of their clearest differences from computer hardware. Each calculation takes approximately eight hours, as the bacterial populations must grow, produce proteins, release molecular signals, and allow those signals to reach and activate neighboring colonies. For most electronic applications, such a delay would make the system impractical. The researchers, however, are targeting biological settings in which the relevant timescales are measured in hours, days, or seasons rather than microseconds. A circuit operating overnight could be fast enough to monitor a plant’s condition during a growing season. “We’re not trying to replace computers,” Voigt said. “We want to put computational control into biology.”
That goal could eventually lead to living sensors embedded on plants or around their roots. A printed bacterial circuit might detect combinations of chemical signatures associated with drought, fungal infection, insect attack, or nutrient imbalance. Instead of merely reporting the presence of one molecule, a network could evaluate several conditions at once and trigger a tailored response only when the correct pattern appears. For example, a circuit might activate the production of a fungicide when it detects both a pathogen-associated signal and evidence of plant stress. Similar systems could potentially be designed to regulate microbial communities, monitor environmental pollutants, or control the release of therapeutic compounds. The use of Pantoea agglomerans, a surface-associated bacterium already linked to plants, makes the platform particularly relevant to agricultural applications, although substantial work would be required before such circuits could operate reliably outside a laboratory.
The study represents a shift in how synthetic biological computation can be designed. Instead of asking one cell to contain an entire circuit, the researchers treat populations of specialized cells as interchangeable computational units. Their colonies can be printed into patterns, connected through molecular relays, and reconfigured to produce new functions. The approach remains slow, dependent on carefully controlled growth conditions, and limited by the stability and communication range of living cells. Yet its potential lies precisely in the fact that it is biological. These circuits can exist where electronic devices are difficult to deploy, respond directly to chemical conditions, and in principle connect sensing with a biological action. The MIT team’s living circuit boards therefore offer a provocative vision of future computation: not machines placed into biology, but biology engineered to compute for itself.
Subject of Research: Living bacterial transistors and printed biological circuit boards for molecular computation
Article Title: Living circuit boards built by printing bacterial transistors
News Publication Date: 17-Aug-2026
Web References: https://www.nature.com/articles/s41589-026-02300-3
References: Nature Chemical Biology, DOI: 10.1038/s41589-026-02300-3
Keywords: Bacteria, Pantoea agglomerans, synthetic biology, biological circuits, bacterial transistors, molecular computation, bioengineering, biotechnology, agriculture, plant sensing, logic gates, living circuit boards
