Innovative microfluidic technology has enabled massively parallelized and extremely efficient biological and clinical assays. Many biological applications developed and executed with traditional bulk processing techniques have been translated and streamlined through microfluidic processing with the notable exception of sample volume reduction or centrifugation, one of the most widely utilized processes in the biological sciences. We utilize the high-speed phenomenon known as inertial focusing combined with hydraulic resistance controlled multiplexed micro-siphoning allowing for the continuous concentration of suspended cells into pre-determined volumes up to more than 400 times smaller than the input with a yield routinely above 95% at a throughput of 240 ml/hour. Highlighted applications are presented for how the technology can be successfully used for live animal imaging studies, in a system to increase the efficient use of small clinical samples, and finally, as a means of macro-to-micro interfacing allowing large samples to be directly coupled to a variety of powerful microfluidic technologies.

Microfluidic blood processing is used in a range of applications from cancer therapeutics to infectious disease diagnostics. As these applications are being translated to clinical use, processing larger volumes of blood in shorter timescales with high-reliability and robustness is becoming a pressing need. In this work, we report a scaled, label-free cell separation mechanism called non-equilibrium inertial separation array (NISA). The NISA mechanism consists of an array of islands that exert a passive inertial lift force on proximate cells, thus enabling gentler manipulation of the cells without the need of physical contact. As the cells follow their size-based, deterministic path to their equilibrium positions, a preset fraction of the flow is siphoned to separate the smaller cells from the main flow. The NISA device was used to fractionate 400 mL of whole blood in less than 3 hours, and produce an ultrapure buffy coat (96.6% white blood cell yield, 0.0059% red blood cell carryover) by processing whole blood at 3 mL/min, or ∼300 million cells/second. This device presents a feasible alternative for fractionating blood for transfusion, cellular therapy and blood-based diagnostics, and could significantly improve the sensitivity of rare cell isolation devices by increasing the processed whole blood volume.

Haptotaxis is critical to cell guidance and development and has been studied in vitro using either gradients or stripe assays that present a binary choice between full and zero coverage of a protein cue. However, stripes offer only a choice between extremes, while for gradients, cell receptor saturation, migration history, and directional persistence confound the interpretation of cellular responses. Here, we introduce nanodot stripe assays (NSAs) formed by adjacent stripes of nanodot arrays with different surface coverage. Twenty-one pairwise combinations were designed using 0, 1, 3, 10, 30, 44 and 100% stripes and were patterned with 200 × 200, 400 × 400 or 800 × 800 nm2 nanodots. We studied the migration choices of C2C12 myoblasts that express neogenin on NSAs (and three-step gradients) of netrin-1. The reference surface between the nanodots was backfilled with a mixture of polyethylene glycol and poly-d-lysine to minimize nonspecific cell response. Unexpectedly, cell response was independent of nanodot size. Relative to a 0% stripe, cells increasingly chose the high-density stripe with up to ~90% of cells on stripes with 10% coverage and higher. Cell preference for higher vs. lower netrin-1 coverage was observed only for coverage ratios >2.3, with cell preference plateauing at ~80% for ratios ≥4. The combinatorial NSA enables quantitative studies of cell haptotaxis over the full range of surface coverages and ratios and provides a means to elucidate haptotactic mechanisms.