Three-dimensional Imaging of Bacterial Cells for Accurate Cellular Representations and Precise Protein Localization

Because bacteria are so small, people often forget they are three-dimensional objects. Our protocol allows the accurate reconstruction of the shapes of both flat and curved cells. This protocol only requires slight modifications to a standard microscope and they're readily available in MATLAB scripts.

Before beginning the experiment place two stacks of each of three 20 by 20 millimeter cover slips on the opposite ends of a single glass microscope slide. Pipette 200 microliters of agarose pad solution onto the slide and immediately and firmly place a second slide onto the stack of cover slips to flatten the agarose. After allowing the agarose to solidify for about one minute at room temperature carefully slide off the top slide and use a large end of a 200 microliter pipette tip to cut out individual five millimeter diameter pads from the gel.

When all of the pads have been obtained pipette the E.coli cell culture multiple times to disrupt any cell clumps and to ensure that the cells are homogeneously distributed before adding one microliter of subculture onto each pad. After letting the samples air dry for five to 10 minutes confirm that the droplets have been completely absorbed into the pads before placing a cover slip onto the pads. Then use a cotton swab to gently brush an appropriate sealant around the edge of the cover slip, taking care to keep the sealant away from the top of the cover slip.

Immediately after sealing the cover slip place the slide onto a microscope stage and after five minutes of thermal equilibration use the microscope focus wheels to bring the middle of a cell into focus. In the associated microscope software under ND Acquisition, check the Z box to acquire a Z stack and click the Home button to set the the middle of the cell as the starting point. Set the Step size to 0.1 micrometers and the Range to four micrometers and make sure that the Z Device is set to the Piezo stage.

It is critical that there is a sufficient blurring on both the top and bottom of the cell so that the cell is almost indistinguishable from the background. Set the fluorescent channels under the lambda window to the settings for the fluorescent molecules being imaged and confirm that the order of experiment is set to lambda so that a complete Z stack will be obtained in each color channel before switching. Then click Run now to begin the image acquisition.

When all of the images have been acquired open the files in an appropriate image analysis software program. Draw a box around an individual cell and duplicate that cell two times, once for each channel, making sure the Duplicate hyperstack box is checked, and change the channel to either one or two. Once both stacks are available, select Images, Stacks, Tools, and Concatenate to combine the images.

With the protein channel first and the shape channel second. After saving the new image as a TIFF file make a folder inside a folder on the desktop that contains the cropped images and the cell_shape_settings_tri. txt file from shae-cellshape-public exampleData_tri.

Edit the cell_shape_settings_tri. txt file to have the correct settings for the experiment of interest. And run the Cell_shape_detector3dConvTriFolder function with the string to the folder location followed by the number of the cell on which to start and the number of cells that should be run.

To confirm that the cells have been properly reconstructed run ScreenFits. When the graphical user interface opens click the Select Folder button and select the folder with the reconstruction data files to visually screen the individual cell reconstructions. For any cell that appears misshapen or that lacks full coverage select the box next to the reconstruction to appendage FLAG to the name so that it can be excluded from any downstream analysis.

Run enrichmentSmothingSpline to create an enrichment profile of the relative concentration of the protein of interest as function of the Gaussian curvature at the surface. Then select each folder with the just-created TRI. mat files and select the newly-created curve mat file.

This method is especially useful when dealing curved or abnormally shaped cells as a 2D representation cannot reflect the curvature of the cells accurately. In this experiment a green fluorescent fusion protein to the bacterial actin protein MreB was generated to study the precise localization of the actin protein in both wild type and mutant E.coli cells. By making these measurements in 3D captured images the shapes of both the wild type and rodZ mutant cells were able to be reconstructed.

The localization of MreB was then determined to be enriched at small Gaussian curvatures, a geometric feature that can only be measured in 3D in a rodZ dependent manner. Remember to make sure the Z stack is large enough that the cells are blurred on the top and bottom and that the cells are well spaced from each other. We are showing how to measure the geometric localization of proteins in 3D but other information can be calculated from this data such as the size of tag protein assemblies.