ARTICLES https://doi.org/10.1038/s41589-019-0357-8 1 Department of Biomedical Engineering, Duke University, Durham, NC, USA. 2 Shenzhen Institute of Synthetic Biology, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China. 3 School of Life Sciences, Peking University, Beijing, China. 4 College of Life Science and Technology, Beijing University of Chemical Technology, Beijing, China. 5 Center for Genomic and Computational Biology, Duke University, Durham, NC, USA. 6 Department of Molecular Genetics and Microbiology, Duke University School of Medicine, Durham, NC, USA. 7 Present address: Shenzhen Institute of Synthetic Biology, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, Guangdong, China. *e-mail: you@duke.edu B acteria are a common host to produce diverse biologics, accounting for ~30% of biopharmaceuticals 1,2 . Synthesis of recombinant proteins using bacterial hosts entails multiple steps including culturing, disruption of bacteria by physical or chem- ical means and subsequent isolation and purification of the desired product. For industrial operations, these steps are usually carried out on a large scale; consequently, each step requires a sophisticated and delicate infrastructure to ensure efficiency and product qual- ity 3,4 . While critical for producing molecules in large amounts, this format is not flexible or economically suited for producing or char- acterizing diverse biologics when only a small amount is needed for each 5–7 . Moreover, standard biomanufacturing is not accessible in remote or underdeveloped areas that lack the basic infrastructure (for example, transportation, equipment or electricity) and person- nel with necessary technical training. Conventional manufacturing of certain products in advance could also result in wasted labor and resources owing to the high cost of production, the need for a cold chain for transportation and storage, and the short shelf-life of bio- logics 8 . Therefore, there is a critical need to develop technologies for versatile and scalable on-demand production of diverse biologics, as well as subsequent analysis and purification. This need has driven the development and adoption of single- use technologies (SUTs) for biomanufacturing 3,9 . To date, however, SUTs have focused on replacing traditional stainless-steel-based fixed reactors with flexible and disposable parts 10 or the direct miniaturization of reactors 11 , without changing the fundamental structure of the manufacturing process 12,13 . In particular, distribut- ing production of biologics into multiple steps is a standard practice and important for implementing the production and purification as unit operations, where each unit operation can be individu- ally optimized for large-scale biomanufacturing 3 . However, each step can lead to increased complexity in operation, particularly for small-scale operations. For certain proteins, engineered secre- tion can facilitate the extraction process 14 . Yet, the diverse secre- tion mechanism of bacteria makes the selection of the appropriate secretion pathway for each recombinant protein complicated and time consuming 15 . In addition, the efficiency of secretion is often limited in bacterial hosts, and some large cytoplasmic proteins may be physically impossible to translocate 16 . Induced lysis can allow effective release of diverse proteins 17 , but it does not simplify the process of cell debris separation and downstream purification 18 . As another alternative for biomanufacturing, cell-free systems have the advantages of not being constrained by a cell wall or limited by the toxicity of products 19,20 . In comparison to living cells, however, cell-free systems tend to have a lower efficiency, such as low protein production rates and a short reaction duration owing to the rapid depletion of energy resources and the accumulation of inhibitory by-products 21,22 . To overcome these limitations, we take advantage of recent devel- opments in synthetic biology and stimulus-responsive biomaterials to integrate the multiple steps of production, disruption and separa- tion into a concise format. Our central design concept is to couple programmed lysis of engineered bacteria and controlled periodic volume phase transition in capsule. When the local cell density inside the capsule is sufficiently high, autonomous partial lysis will occur and allows the cells to release their contents, including the protein product of interest. The bacterial growth changes the local environmental conditions (pH and ionic strength), driving the vol- ume phase transition of the encapsulating material. The transition leads to the squeezing effect, which greatly facilitates the transport of the released protein from the interior to the exterior, while cells and large debris are trapped inside the capsules 23–25 . We use nutrient replenishment as a cue to swell the capsule again, while it also resets the capsule environment and allows the cell density to recover after Versatile biomanufacturing through stimulus-responsive cell–material feedback Zhuojun Dai 1,7 , Anna J. Lee 1 , Stefan Roberts 1 , Tatyana A. Sysoeva 1 , Shuqiang Huang 2 , Michael Dzuricky 1 , Xiaoyu Yang 3 , Xi Zhang 2 , Zihe Liu 4 , Ashutosh Chilkoti 1 and Lingchong You 1,5,6 * Small-scale production of biologics has great potential for enhancing the accessibility of biomanufacturing. By exploiting cell– material feedback, we have designed a concise platform to achieve versatile production, analysis and purification of diverse proteins and protein complexes. The core of our technology is a microbial swarmbot, which consists of a stimulus-sensitive polymeric microcapsule encapsulating engineered bacteria. By sensing the confinement, the bacteria undergo programmed partial lysis at a high local density. Conversely, the encapsulating material shrinks responding to the changing chemical envi- ronment caused by cell growth, squeezing out the protein products released by bacterial lysis. This platform is then integrated with downstream modules to enable quantification of enzymatic kinetics, purification of diverse proteins, quantitative control of protein interactions and assembly of functional protein complexes and multienzyme metabolic pathways. Our work demon- strates the use of the cell–material feedback to engineer a modular and flexible platform with sophisticated yet well-defined programmed functions. NATURE CHEMICAL BIOLOGY | www.nature.com/naturechemicalbiology