Recombinant protein production in bacteria is limited by costly cell lysis and multi-step purification. To bypass these limitations, leveraging native bacterial secretion systems to secrete proteins offers a promising alternative. The type 3 secretion system (T3SS) of Salmonella enterica serovar Typhimurium can secrete proteins at hundreds of milligrams per liter. While promising, higher titers from this and other natural systems are needed to be relevant commercially. A major engineering target to enhance protein secretion is the T3SS secretion apparatus. However, efforts to quantify its assembly are hindered by bottlenecks in scalability and complexity. Assembly involves ~20 structural proteins forming with precise stoichiometry under dynamic regulation. Membrane- and periplasmic-embedded components are difficult to probe without costly, time-intensive methods. To fully realize the T3SS as a tool for scalable protein production, new tools are needed for rapid and accurate characterization of the assembled apparatus. Here, we establish a high throughput flow cytometry method in S. Typhimurium by quantifying the abundance of the needle-bound tip protein SipD as a proxy for T3SS apparatus assembly. We then adapted a super resolution microscopy method, known as Structured Illumination Microscopy (SR-SIM), to visualize the presence of SipD and validate the flow cytometry results. Applying this approach, we revealed how overexpression of key T3SS regulators hilA and hilD impact assembly, expanded the assay with a secretion-compatible fluorescent reporter to link assembly with secretion, and uncovered how a PrgI variant impacts apparatus architecture. Together, these tools enable rapid insights into T3SS assembly and advance heterologous secretion platform development.
The Hidden Power of Normalization: Exponential Capacity Control in Deep Neural Networks
arXiv:2511.00958v1 Announce Type: cross Abstract: Normalization methods are fundamental components of modern deep neural networks (DNNs). Empirically, they are known to stabilize optimization dynamics and


