FLUID-CELL: Flow-enabled Light and Ultrastructural Imaging Device for Correlative Electron and Light Localization

Abstract

We present a novel microfluidic flow cell, or Flow-enabled Light and Ultrastructural Imaging Device for Correlative Electron and Light Localization (FLUID-CELL), that connects fluorescence light microscopy (FLM) and cryo-electron microscopy (cryo-EM) for advanced biological imaging and ultrastructural and structural analyses. The design of the FLUID-CELL features a precisely engineered microchannel that maintains native cell culturing conditions, supporting correlation and enabling real-time observation by FLM, as well as subsequent cryo-EM analysis. In this study, this device enabled practical FLM imaging over extended experimental periods, consistent sample handling, and the ability to perform correlative imaging. This capability connects dynamic cellular events imaged by fluorescence light microscopy with high-resolution ultrastructural data collected with cryo-EM. Our dual-modality approach streamlines the workflow and opens new possibilities for investigating the relationship between cellular function and molecular architecture at the nanoscale.

Keywords: bacteria; biofilm; cryo-electron microscopy (cryo-EM); cryo-electron tomography (cryo-ET); microfluidic; three-dimensional (3D) printing.

Figure 1.. Design of the primary FLUID-CELL components.

A) An overhead view of the system, with the gaskets (in black) and the Luer-Lok adaptor inlet and outlet ports (in transparent gray) visible. B) A view underneath the flow cell, the coverslip, and a centered grid is visible. C) A view of the flow cell from within the chamber. (D) Technical drawing describing the bottom section of the flow cell. Details of all relevant dimensions. (D-1) The base of the flow cell well shows the thickness of the chamber wall that compresses the gasket. (D-2) Receiving well for an EM grid. The central circle was chosen to allow for a small tolerance in grid placement. The smaller spaces to the left and right of the central circle are provided to allow for the development of flow progression before encountering the grid. (D-3) Receiving well for external clamping. (D-4) Cross section of the flow cell, along the path of the flow chamber. (D-5) Close up to the inflow port. Shows the height of walls surrounding the flow chamber. (D-6) Details the grid well depth as well as the depth of the coverslip well. (E) Technical drawing describing the top section of the flow cell. Details of all relevant dimensions. (E-1) Extruded ‘Bananas’ which keep the grid in the focal plane of the microscope, on a glass coverslip. Geometry was determined from the intersection of two arcs, one from the center of the grid well with a radius of 1.45 mm. The second was chosen as the center point of the previous edge, with a chosen radius of 2.53 mm. (E-2) Note the connections from the bottom of the inlet ports to the interior wall of the ports and the main body. (E-3) Cross section of the flow cell, showing the dimensions of the ports and the extrusion of the ‘bananas.’ (E-4) Shows the size of the gasket reception well. (E-5) Details the receiving well for clamping. (E-6) Shows the width of the threaded adaptor. (E-7) Shows rise and turn of the thread.

Figure 2.. Design of peripheral FLUID-CELL components.

(A) Technical drawing of the external securing bars and their dimensions. The bars are made of Durable V2 resin and are used to provide external pressure on the flow cell, compressing the gasket and sealing the flow cell chamber to make it leakproof. (B) Technical drawing of the sealing gasket and its dimensions. The sealing gasket is made of Silicone 40A resin and serves as a compressible seal between the two flow cell chamber halves, making it leakproof.

Figure 3.. Diagram depicting flow cell setup.

Diagrams are of the experimental configurations for bacterial cell culture (A) and nanobead flow experiments (B). (A) Four replicate FLUID-CELL devices were used in each experiment. However, other numbers of FLUID-CELL devices can be connected and used in similar configurations, depending on existing instrumentation. Once EM grids were inoculated, inlet flasks were filled with a rich PYE media, allowing the EM grids to be flushed with a fresh, sterile, and steady flow of media while microcolonies and biofilms developed. (B) The syringe with the nanobead solution was connected to the Gilson MINIPULS 3 peristaltic pump line far enough away from the flow cell and grid to allow for complete mixing between the two streams before reaching the observation point in the flow cell chamber.

Figure 4.. Velocity profile of the FLUID-CELL chamber with an EM grid in place.

Velocity was calculated by passing fluorescent beads into the flow stream set at 100 μl/min and capturing a time-lapse image. Analysis was performed with FIJI software using the TrackMate plugin. (A) Field of view (FOV) data were collected using brightfield light microscopy. (B) View of this same FOV under Texas Red (TXR) fluorescence, with fluorescent beads visible moving across the top of the grid. (C) Single frame from a collected movie showing the flow paths that were processed from the movies, paths are colored based on a velocity gradient seen in D. (D) Collection of all flow paths from an entire data movie, depicts velocity gradients across the EM Grid as well as the observable streamlines that the fluid moves along. Scale bars of 500 μm (A-C).

Figure 5.. Fluid flow experiment with Caulobacter crescentus CB15 cells.

Caulobacter crescentus CB15 cells were inoculated into the FLUID-CELL and incubated under static conditions for 2 hours to allow for grid attachment. Next, the media was flushed through at a rate of 200 μL/min, and brightfield (BF) images were collected every 24 hours over a 48-hour period. Time 0 represents the point at which the flow was initiated. (A) View of an EM grid within the receiving well. (B) 0 hours of flow, 40x BF image of a center square. (C) 24 hours of flow, the same grid square. (D) 48 hours of flow, the same grid square. (E-H) An equivalent experiment conducted with an Ibidi μ-Slide VI^0.5 flow cell, cells were inoculated and allowed to incubate in static conditions for 2 hours, then media flow was started at a rate of 200 μL/min, and BF images were collected at regular time points of a central section of the flow chamber. (E) A BF image of an area of several starting biofilms, 4 hours after flow began. (F) A BF image of the exact location, with a notable biofilm colony developing, 10 hours after the flow started. (G) A BF image of the same area, with the same large biofilm colony, and other various colonies growing, 16 hours after flow started. (H) A BF image of the same area shows that the extensive biofilm had grown to a height at which the fluid flow swept it out of the frame; however, other, smaller colonies continued to develop 24 hours after the flow began. Scale bars of 500 μm (A), 25 μm (B-H).

Figure 6.. Light microscopy and electron microscopy images of Caulobacter crescentus cells at select time points during a flow experiment.

(A) A fluorescence image of C. crescentus cells attached to an EM grid surface (whole square view) after 16 hours of fluid flow. (B) A brightfield image of the same square (A) after 16 hours of fluid flow. (C) An enlarged view of B, where a distinct ‘W’ shape of an initial C. crescentus cell community is visible. (D) A correlated negative stain EM image of this same C. crescentus cell community ‘W’. Scale bars of 20 μm (A-B), 10 μm (C), and 2 μm (D).