Bacteria infections are a major threat to human health. Antimicrobial resistance is increasing, and vaccines are still lacking for several major pathogens. To identify novel avenues for control strategies, we need to know more about how pathogens respond to the host immune system and to antibiotics in vivo in the infected tissues.
Surviving bacteria can cause relapses if antimicrobials fail to eradicate an infection. Using serial tomography, we show that typhoid fever-causing salmonella can persist during therapy in a specific part of the spleen. Inflammation is weak in this region, which disrupts the critical synergy between antibiotics and inflammation for effective clearance.
Bacteria are micrometer-sized objects. It is difficult to localize such small objects in their tissue context. Serial two-photon tomography can detect one single bacterium in centimeter-sized organs and determine its replication rate.
We can also detect immune cells, such as neutrophils, and determine their spatial relationship with the bacteria. We study the activities of bacteriostatic antibiotics. These antibiotics can cure infections, although they can only inhibit bacterial growth without actually killing the pathogen.
We also determine the spreading of bacterial mutants that lack certain virulence factors and the impact of immune defects on infection and treatment. To begin, pick up the thawed samples of salmonella-infected murine tissues with a pair of tweezers. Insert it quickly into the bottom of a plastic mold containing liquid agarose.
Once polymerization is complete, use a scalpel to open the plastic mold from the corners. After placing a blade on a microtome, transfer the sample with gloved fingers into a sample box. Position the sample under the microtome such that the upper surface of the agarose is at the same level as the blade of the microtome.
Move the stage to place the agarose cube at the center of the objective. Click on the slice button in the tomograph software until an entire intact slice of the cube is obtained. Next, center the objective lens on the tissue sample in X and Y directions.
Launch the laser software, adjust the wavelength to 800 nanometers, and then turn on the laser. Close the cabinet doors of the microscope. Then, switch off all the light sources in the room.
Turn on the PMTs and set the voltage to 750 volts. Now set the microscope shutter to automatic, then set the voltages of V1 and V2 to 20 and 1.71 respectively. Adjust the Z axis with the Z-piezo device until it reaches a tissue surface.
Cut multiple slices of 50 micrometer thickness. Next, define the sample edges. Confine the imaging area to the tissue, with minimal imaging of the surrounding agarose.
After turning off the PMTs, set the laser wavelength to 800 nanometers. Use the laser spot to find the coordinates of the tissue edges while shifting the stage. After confirming the coordinates, position the laser spot at the right front center.
Now, change the wavelength to 940 nanometers. Set the microscope shutter to automatic mode and set the shutter voltage to 20 for V1 and 1.71 for V2.Press the 3D mosaic setting option to begin scanning. After imaging, collect the tissue sections from the water tank.
To stitch the tile images, first, transfer data from the tomograph server to the Linux computer. Open the MATLAB script, StepOneStitchingAndArchive. m, and locate the source folder.
Switch to the editor tab and click run. The progress information will be displayed in the command window. Find the stitched images in a subfolder called stitchedimages_100 in the source folder.
Compress the raw data using the tar command and save it as a single file with the file name extension tar.bz2. To train the support vector machine, preview the stitched images. Open one image with the same file name from each subfolder with Fiji.
Merge the three channels into one color image and adjust the brightness of each channel until clear signals are seen from bacteria and tissue autofluorescence. Note the adjusted maximum intensity level of each channel. Next, open the MATLAB script StepTwoSegmentationAndAnalysis.m.
Define the source folder and the image names for training, then go to the editor tab and click on run. When a dialog box asking for color thresholds appears, fill it with the values based on the earlier manual check with Fiji. Select regions for background and regions of interest according to the dialogs.
Graphs will appear showing how well the model has been trained and asking if more regions need to be added. Add more regions of interest and background until the bacteria can be clearly segmented. The segmentation process will automatically run, and the progress can be seen in the command window.
Next, open the blue channel images, which contains second harmonic signals of collagen. Bin down the first channel images tenfold in both X and Y axes and save the downsized images in a new folder. Make three replicates of downsized images within the same folder.
For the 3D visualization, launch the Imaris visualization software package in the arena view. Now, open the IMS or TIF file. Adjust the color representations of the different channels in the display adjustment window.
Click on image properties under the edit menu to manually set the minimum or maximum values and a value for gamma correction. After adjusting the appearance of the image, export the current view using the snapshot tool. Use the navigation pointer to find a view, then use plus add to add keyframes.
Use the animation icon to represent the 3D data as a movie. Press the red record button to build the movie, then save it at the desired destination folder and file type. Individual salmonella cells were detected in entire mouse organs, such as the spleen, liver, mesenteric lymph nodes, and Peyer's patches.
Segmentation of fluorescent bacteria was confirmed by immunohistochemistry. Host cells were stained in vivo by injecting an antibody to surface markers prior to perfusion.
