Fabrication of an Open Microfluidic Device for Immunoblotting

Abstract

Given the wide adoption of polydimethylsiloxane (PDMS) for the rapid fabrication of microfluidic networks and the utility of polyacrylamide gel electrophoresis (PAGE), we develop a technique for fabrication of PAGE molecular sieving gels in PDMS microchannel networks. In developing the fabrication protocol, we trade-off constraints on materials properties of these two polymer materials: PDMS is permeable to O₂ and the presence of O₂ inhibits the polymerization of polyacrylamide. We present a fabrication method compatible with performing PAGE protein separations in a composite PDMS-glass microdevice, that toggles from an "enclosed" microchannel for PAGE and blotting to an "open" PA gel lane for immunoprobing and readout. To overcome the inhibitory effects of O₂, we coat the PDMS channel with a 10% benzophenone solution, which quenches the inhibiting effect of O₂ when exposed to UV, resulting in a PAGE-in-PDMS device. We then characterize the PAGE separation performance. Using a ladder of small-to-mid mass proteins (Trypsin Inhibitor (TI); Ovalbumin (OVA); Bovine Serum Albumin (BSA)), we observe resolution of the markers in <60 s, with separation resolution exceeding 1.0 and CVs of 8.4% for BSA-OVA and 2.4% for OVA-TI, with comparable reproducibility to glass microdevice PAGE. We show that benzophenone groups incorporated into the gel through methacrylamide can be UV-activated multiple times to photocapture protein. PDMS microchannel network is reversibly bonded to a glass slide allowing direct access to separated proteins and subsequent in situ diffusion-driven immunoprobing and total protein Sypro red staining. We see this PAGE-in-PDMS fabrication technique as expanding the application and use of microfluidic PAGE without the need for a glass microfabrication infrastructure.

Figure 1.

A composite PDMS-glass microdevice for protein immunoblotting. (A) Photograph of a composite PDMS-glass chip (gel is visualized with Allura red stain). Neither the PDMS substrate nor the glass slide are plasma treated, creating a reversible bond to facilitate delamination of the PDMS from the glass slide. After delamination, proteins immobilized in the microgel are accessible for subsequent manipulation (e.g., immunoprobing, staining, collection, tandem analyses). (B) Microchannels are formed by mating a precast PDMS substrate and a glass slide. Channels are imbibed with benzophenone, which diffuses into the PDMS to form an O₂ shield and promotes PA polymerization upon UV exposure. (C) The fabrication workflow consists of wicking a polymer precursor solution having a high %T acrylamide concentration into the microchannels. A gel pore-size discontinuity is created by masking areas to avoid photopolymerization in the upper channels. Unpolymerized precursor is suctioned out of the channel; a low %T PA precursor solution is introduced, and the entire chip is exposed to UV. (D) Protein separation and blotting: Proteins are analyzed by PAGE and immobilized to the gel via photocapture (blotting) prior to chip delamination and immunoprobing.

Figure 2.

Native protein PAGE in the composite PDMS and glass device. (A) Inverted fluorescence micrographs of TI, OVA, and BSA exhibit sample stacking at the large-to-small pore-size discontinuous gel interface. Note the reduced peak width at the gel interface (E = 250 V/cm, discontinuous 5–15%T PA gel, scale bar = 50 μm). (B) Velocities of TI, OVA, and BSA peaks are affected by local gel pore size. (C) Electrophoretic mobility of TI, OVA, and BSA decreases after migrating from the largepore gel to the small-pore gel. (D) Stacking factors (SF) for TI, OVA, and BSA calculated as the ratios between the bandwidth before and after gel discontinuity as well as the mobility before and after gel discontinuity: SF is equal in both cases. Error bars report standard deviations.

Figure 3.

Protein sizing in a composite PDMS and glass device. (A) Inverted fluorescence micrographs show time evolution of a PAGE analysis of TI, OVA, and BSA in a discontinuous 5–15%T gel (electric field = 250 V/cm, scale bar = 50 μm), with the three protein bands after 60 s of migration. (B) Time evolution of the average SR between OVA-BSA and OVA-TI peak pairs (n = 3 chips, errors bars = SD).

Figure 4.

Diffusion-driven immunoprobing of a PA gel housed in a delaminated, “open” PDMS trench. From the left panel to the right panel: (i) Micrographs of PAGE analysis (E = 250 V/cm) of OVA and BSA in a discontinuous 5–8%T PA gel; (ii) after photoblotting, PDMS substrate release, and immunoprobing; (iii) immunoprobed OVA; (iv) fluorescence trace of immunoblot.

Figure 5.

Protein staining of a protein ladder in the exposed protein-decorated PA gel. In (A–C), PAGE and blotting of Alexa 488 conjugated BSA, OVA, and TI are completed; then, the protein-decorated gel is subjected to Sypro red staining. (A) Protein staining under native protein conditions, with each ladder component initially at ~200 nM. Comparison of two Sypro red incubation periods shows no appreciable difference in detectable BSA signal. (B, C) Protein staining under 0.05% SDS conditions, with each ladder component initially at ~200 nM (B) and ~400 nM (C). Micrograph and intensity traces are after 30 min of SDS incubation and 10 min of Sypro red incubation. (D) Same conditions as in (C), with the addition of 400 nM turboGFP to the protein ladder. 45s PAGE duration; discontinuous 5–15%T PA gel; E = 250 V/cm.