Microfluidic preparation of anchored cell membrane sheets for in vitro analyses and manipulation of the cytoplasmic face

Molecular networks on the cytoplasmic faces of cellular plasma membranes are critical research topics in biological sciences and medicinal chemistry. However, the selective permeability of the cell membrane restricts the researchers from accessing to the intact intracellular factors on the membrane from the outside. Here, a microfluidic method to prepare cell membrane sheets was developed as a promising tool for direct examination of the cytoplasmic faces of cell membranes. Mammalian cells immobilized on a poly(ethylene glycol)-lipid coated substrate were rapidly and efficiently fractured, with the sheer stress of laminar flow in microchannels, resulting in isolation of the bottom cell membrane sheets with exposed intact cytoplasmic faces. On these faces of the cell membrane sheets, both ligand-induced phosphorylation of receptor tyrosine kinases and selective enzymatic modification of a G-protein coupling receptor were directly observed. Thus, the present cell membrane sheet should serve as a unique platform for studies providing new insights into juxta-membrane molecular networks and drug discovery.


Fluorescently labeled PEG-lipids
Fluorescently labeled PEG-lipids ( Fig. S1) were synthesized essentially as previously reported [S1] . N-boc-amine-PEG-NHS was lipidated with DOPE using the amine coupling reaction and, after deprotection, FITC and Alexa647-NHS were reacted with the amine moiety of the PEG-lipid. Figure S1. Chemical structures of reagents used for membrane staining.

Detailed protocols of microfluidic preparation of cell membrane sheets
To prepare cell membrane sheets by microfluidic laminar flow, a microfluidic device immobilizing cells in the flow path was prepared by combining the PEG-lipid coated glass substrate with a sticky-slide (catalog number: 81128, ibidi GmbH, Munich, Germany) which had a microchannel. The microchannel had a width, length and height of 5, 48.5 and 0.1 mm, respectively. When the sticky-slide was fitted to the glass, the microchannel was closed by the PEG-lipid coated surface of the glass, establishing a flow path with a PEG-lipid coated bottom surface. The glue on the surface of this sticky-slide was removed and the slide incubated in 1% BSA (Sigma Aldrich) for 10 min. After washing the sticky-slide with MilliQ water, both the sticky-slide and the glass were fitted together in PBS to prevent trapping of air in the flow path. The combined sticky-slide and glass were fixed in place with a set of handmade stainless steel clips (Fig. S2). After formation of the flow path, 100 μL of a cell suspension was poured into each micro flow path, followed by incubation for more than 10 min at room temperature to enable immobilization. Then, the flow path was connected to a syringe pump (Legato 100, KD Scientific, Holliston, MA, USA) with tubes (LMT-55, Saint-Gobain, Courbevoie, France) and connectors (VRF106 and VFT106, ISIS Co., Ltd., Osaka, Japan). PBS was poured into the flow path at a variety of velocities, from 0.1 to 50 mL/min, for 1 min, to induce cell fracture.

Confocal fluorescence imaging of the cell membrane sheets of He/La cells
He/La cells were harvested and immobilized on the PEG-lipid surface, and after incubation for 10 min, similar to Ba/F3 cells, the cell membrane sheets were obtained by the present microfluidic cell fracture (Fig. S3a,c). In the case of adherent cells, it is previously reported that the immobilized cells adhere and expand on the PEG-lipidmodified collagen surface after long incubation under physiological conditions [S2] . In this study, He/La cells were also confirmed to expand on the surface after incubation for more than 1 hour at 37 °C. After adhesion and expansion, the efficiency of the cell membrane sheet formation from the well-spread cells was observed to decrease (Fig.   S3b,c). The reason of this decrease can be explained in accordance with decrease in cell height and formation to the streamline form of the spread cells. These changes in cell shape are assumed to lead to reduction of shear stress on the cells.

Immunostaining of whole cells expressing the antigen on cell surfaces
In addition to the samples shown in Fig. 3, whole cells expressing HA-G2A, not exposed to microfluidic laminar flow, were also observed after immunostaining (Fig.   S4). In these samples, on the bottom plasma membrane of the immobilized cells, the HA-tag on the extracellular domain of G2A did not stain with anti-HA antibody FITC conjugate, whereas it was clearly stained on the top and side plasma membrane. These results strongly suggested that the anti-HA antibody FITC conjugate could not access the space between the membrane sheet and the glass substrate.

Scanning electron microscopy (SEM) imaging
In addition to those shown in Figs. 4, SEM images of a whole cell and membrane sheets are shown in Fig. S5. Traces were observed behind the fractured membrane sheets in Fig. 4a and Fig. S5b. Therefore, the enlarged images of the traces in Fig. 4a were also obtained (Fig. S6).

Fluorescence recovery after photo-bleaching (FRAP) analysis
The time-course plots of the normalized fluorescence intensities obtained from the FRAP experiment are shown in Fig. S7. and with the sonication method (closed diamonds). Raw data were normalized to intensities obtained just before bleaching. Values are means ± standard deviation (n = 10).