Robust operations of ultra-high throughput cell lineage tracking for rapid bioengineering and pathogen detection

Bifrost Biosystems – a recently formed biotech spin-out from Harvard University – here presents a platform for ultra-high throughput and ultra-high information screening of cell lineages. The basic platform involves individual lineages growing and dividing in narrow dead-end trenches that open into a feeding channel with flowing media. The media feeds the cells diffusively, and also washes away excess progeny to maintain a linear mini-colony of constant size. We have built over 1,000 different types of such devices, have tracked individual cell lineages for over 700 consecutive generations, have achieved imaging throughput of over 1,500,000 parallel lineages with sufficient resolution to track individual cells, and can image from 10-150X magnification, even with sub-diffractive microscopy. We have successfully used 150 different organisms on this platform, from some of the smallest bacteria all the way to human stem cells and can also grow different kinds of cells together to study interactions. We further developed many additional enabling features within this platform, including: non-intrusive and pristine live cell retrieval post-imaging, selective optical killing of cells, in situ genotyping of every lineage on chip, single-cell longitudinal time series with scRNA-Seq, 96-well drug screening formats where each well has 10,000 cell trenches, concentration gradients for dose-response curves, the possibility of interfacing with droplets, and the ability to run large screens for structured arrayed libraries in addition to pooled ones. The platform has also accurately identified and characterized bacteria, based on hyperspectral imaging, microscopic growth profiles, and quantitative phase microscopy. These features all worked well on a single type of platform in an academic lab setting, where platforms were run and maintained by the teams of world-experts that developed them. Now we propose to bring these capabilities to a broader audience of government labs, academic institutions, and possibly the private biotech sector. Many core aspects of the platform must be standardized and streamlined, so that instruments can run every day and be operated by end-users who have little training. This will require an integrated fluid handling system, ideally stand-alone reusable cartridges, a wide range of optimized microfluidic growth chips for the full range of organisms used, and an integrated platform of algorithms to extract and analyze the TBs of data that are produced every day. Here we propose the necessary steps for such an integrated platform, which in 24 months will be ready for in depth evaluation in the laboratories of transition partners.