Cell Lysis and Protein Extraction in a Microfluidic Device with Detection by a Fluorogenic Enzyme Assay

Eric A. Schilling, Andrew Evan Kamholz and, Paul Yager. Analytical Chemistry 2002 74 (8), 1798-180.

Summary:

A novel Mylar microfluidic device for lysis of bacterial cells and detection of b-galactosidase was developed and described. The chemical-based lysis device consisted of an H-filter with three inlets and two outlets (Figure 1). A suspension of E. coli (~10⁸ cells/mL) and the lytic agent were pumped in separately and met in the main lysis channel, which depends on diffusion of the detergent to lyse cells. Bacterial Protein Extraction Reagent (BPER) was chosen as the lytic agent in order to permeabilize the bacterial cell membranes and allow release of b-galactosidase without affecting the catalytic activity of the enzyme. The lysis channel split into two channels downstream, a controlled outlet and a detection channel. The geometry of the device allowed cells to be discarded out of the outlet, while smaller molecules had sufficient time to diffuse into the side of the channel that split off into the detection channel. A fluorogenic assay downstream of the lysis allowed for detection of b-galactosidase. b-D-galactopyranoside (RBG) was pumped into the third outlet, which flowed into the detection channel to allow for fluorescent detection of b-galactosidase. b-gal catalyzed the reaction of RBG into resorufin and D-galactose. Resorufin fluoresces when excited at 570 nm and can be detected at 594 nm. A numerical model of the system allowed the researchers to determine the concentration of b-gal in the lysate by integrating Michaelis-Menten kinetics and diffusion coefficients of b-gal, RBG, and resorufin.

Before experimenting with the device, the researchers conducted 3D modeling using CoventorWare to create a velocity profile for b-gal (Figure 2). The researchers assumed that b-gal flowed into the cell inlet, rather than modeling the time taken for lysis to occur. They stated that this error is negligible because previous experiments demonstrate a lysis time of several seconds, which is two orders of magnitude less than the residence time b-gal is present in the channel, 190 seconds.

The researchers also ensured that no cells entered the detection channel by taking dark-field images of the functional device within seconds of fluorescent images. The fluorescent images displayed resorufin fluorescing while the dark-field images showed no cells. In order to back up these findings, the researchers also introduced a fluorescent DNA stain (SYTO 9) to the cells before flowing them into the device and used fluorescent imaging to detect resorufin (red) without SYTO 9 (green) in the detection channel, and both present in the cell outlet.

Significance:

The paper states that the ability to integrate cell lysis into microfluidic chips already developed for analysis would increase the portability and ease of cellular analysis. Analytical techniques such as immunoassays, PCR amplification, and DNA analysis are some of the possible downstream uses of an integrated lysis-analysis microfluidic chip. Several other on-chip devices for lysis have been developed, but these require an external power source and the researchers expressed concerns about the cost of such devices. The researchers feel that their microfluidic device will be both easier to use and fabricate and less expensive to manufacture than these other devices. Although the researchers have thus far only worked with E. coli cells, they state that the dimensions of the device as well as the reagents used could be easily changed in order to lyse and analyze a wide range of cell types. Also, although b-gal is the only molecule detected in the experiments, the researchers have stated that changes in flow rates will allow for smaller molecules to be fractioned out of the cell stream. They chose b-gal because it was a relatively large molecule and could be detected with a simple fluorogenic assay.

Critique:

The researchers do a good job of explaining their choices of reagents, flow rates, and geometry and sufficiently describe controls and experimental checks, but they do not address the feasibility of using the device for further analysis. According to the papers, b-gal has the ability to diffuse almost 100um from the place it was originally released from the cell during the 190 seconds that it resides in the lysis channel. With a channel of 1000m wide and 100um deep, it seems as though a great deal of b-gal would remain in the left half of the channel. This b-gal would be discarded with the cells through the controlled outlet. Therefore, the amount of b-gal harvested by the microfluidic device is much less than the amount initially released by the cells. In fact, according to the paper, only 10% of the b-gal present in the cell stream actually flowed into the detection channel. Although the researchers have included information about the b-gal concentration in the detection channel, there is no discussion about the amount needed for downstream analytical techniques. This paper is a proof-of-concept that E. coli cells can be lysed on a microfluidic chip using chemical lysis, but it does not address the feasibility of using the lysate for later analysis. Figures such as amount of enzyme harvested from the final outlet versus number of cells pumped through the device should be included in future work.