Parallel Multi-Time Point Cell Stimulus and Lysis In a Microfluidic Device Using Chaotic Mixing and Pressure Resistance
Alison Hirsch1, Catherine Rivet2, Boyang Zhang1, Melissa Kemp3, and Hang Lu1. (1) School of Chemical and Biomolecular Engineering, Georgia Institute of Technology, 311 Ferst Dr. NW, Atlanta, GA 30332, (2) Interdisciplinary Bioengineering Program, Georgia Institute of Technology, 311 Ferst Dr. NW, Atlanta, GA 30332, (3) The Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology, 311 Ferst Dr. NW, Atlanta, GA 30332
To build a dynamics model of a complex cellular pathway, it is necessary to acquire a large data set, such as gene expression and protein activity at different time points upon stimulation by external signals. Many important protein activation events occur within minutes after stimulation. However, with conventional multi-well plate assays by manual handling, it is difficult to achieve adequate resolution in the appropriate time scales. Microfluidics is a capable alternative that can provide uniformity in sample handling to reduce error between experiments. We present a device for multi time-point lymphocyte stimulation and lysis for downstream analysis of protein activation. Mixing and even splitting of reagents into each time-point channel while using minimal amounts of reagents are key device features. Our device is capable of eight time points with controlled rapid mixing, precise timed stimulation, and rapid lysis. During operation, a syringe pump drives the flow to only three inlets (cells, stimulus, and lysis buffer), cells and stimulus are mixed and split into eight equal streams in the stimulation chip. The majority of the incubation time occurs in the tubing leading to the lysis chip, where the reaction is quenched and cells are ruptured with lysis buffer to extract intracellular proteins. Rapid mixing is essential to our design for precise definition of stimulation time. A staggered herringbone array was used to achieve full mixing of reagents with minimal shear in less than 0.2 sec on the stimulation chip (and <0.9 sec on the lysis chip), an extremely small fraction of the stimulation time. Because our reagents have different fluidic properties, a COMSOL model was created to probe the mixing effects of viscosity and density. Our model suggests the effect is solely due to viscosity, not the change in density. Further confocal microscopy experiments were used to visualize the mixing of solutions with mismatched viscosity. Equal flow rates in each stream were achieved by balancing the channel resistance: all channels have the same small dimensions with pressure drops orders of magnitude greater than the tubing. Although to set up multiple stimulation times we use different lengths of tubings (with varying inner diameters), the flow rates remain uniform. This configuration gives large flexibility of the specific time points we want to achieve. With commercially available tubings, we can easily achieve stimulation times ranging from 30 seconds to 1 hour. We show that multiple protein activation events in lymphocyte stimulation by antibodies can be detected in a much smaller number of cells using our system than in conventional assays.