Live Imaging and Characterization of Microglia Dynamics and Interactions with Synapses in Diseased Murine Retina

The scope of our lab's research is to study the underlying mechanisms of glaucoma and microglia in synapse remodeling during development. The current experimental challenges revolve around whether or not microglia are actively disassembling retinal ganglion cell synapses or whether they're just passively clearing debris. Begin with mouse retina positioned on a filter paper and dry it with a laboratory wipe.

Invert the retina onto a MatTek Petri dish. Weigh down the retina using either metal beads or a ring. Then, add three to five drops of artificial cerebrospinal fluid.

Place the sample under a spinning disc confocal microscope and adjust the settings to preview the retina. Collect images along the Z axis at 10 to 20 micrometers total depth using a step size of 0.3 micrometers. Maintain the time interval between frames at 20 to 30 seconds to allow for reliable cell tracking.

To perform surface rendering of the microglia, convert the time lapse file into IMS format using the file converter software. Input the voxel sizes based on the previously acquired image properties. In the analysis software arena, click on the observe folder to open the converted file and ensure the image is in 3D view.

To detect individual microglia, click the add new surface icon in the object toolbar. Select segment only a region of interest, track surfaces over time, classify surfaces, and object-object statistics under algorithm settings. Then, click the blue play button to proceed.

Choose the channel for surface rendering. Optionally, enable smooth to smooth the surface creation and enter the surface detail. Under thresholding, choose machine learning segmentation for detecting cells.

Under training data, use the background and foreground paintbrush tools. Draw paintbrush strokes by selecting the desired paintbrush and using the shift plus left click. Ensure interpolate display is enabled, then select train and predict.

Use the yellow pointer in the middle to navigate across the Z axis. Draw multiple strokes across the X, Y, and Z planes, and across a few timeframes. Use three to five short strokes per entry, iteratively adjusting until an accurate surface is made.

Use the yellow rectangle to toggle between the training display and the actual surface created as a final check. Also, check across the time lapse that the surface matches the cell morphology. Once done, deselect split objects.

Adjust the voxel histogram threshold to remove small unconnected objects not part of the primary surface. Then, edit the surface manually to remove extraneous objects or cut edges using the scissor tool. If desired, set a threshold of gaps allowed for object and select the tracking algorithm for the moving cell.

Filter track objects not associated with the main surface. Once finished, set the moving cells display to a transparent profile to visualize the puncta for the next steps. To detect individual PSD95 puncta, click on the blue surface icon in the object toolbar.

Then, select track spots over time, classify spots, and object-object statistics and press play. Select the source channel for the synaptic marker. Turn off slicer and other channels to visualize only PSD95.

Use the control button and the pointer selection tool to estimate the XY diameter of the puncta. Pick a few bright puncta that persist across frames and turn off background subtraction. Next, add a filter based on spot quality.

Adjust the histogram to include accurate PSD95 puncta across all time frames. Remove the false positive spots via edit. Toggle between the cube or circle cursor to delete spots lasting only one or two frames.

If desired, set a threshold of gaps allowed and choose the tracking algorithm for synaptic spots. Filter spot tracks not associated with real puncta using a histogram threshold to exclude small object tracks. Define classifications using the shortest distance to surface classification.

Set engulfed as any puncta less than zero micrometers, contacted as approximately between zero to 0.5 micrometers from the surface, and uncontacted as 0.5 plus micrometers. Assign labels to events using the classifications to find above. Once the surfaces and spots appropriately represent microglia and puncta, export all statistics under the statistics tab.

Microglia exhibited increased process displacement length after laser-induced ocular hypertension compared to the control. Microglial speed was significantly higher in laser-treated eyes than in controls. The number of contact events between microglia and PSD95 puncta increased following laser treatment.

Time-lapse recordings showed that laser-treated retinas exhibited greater microglial motility and increased interactions with PSD95 puncta compared to control retinas. Our significant findings have shown that there's increased microglia number, complexity, and motility, and synaptic colocalization following transient intraocular pressure elevation in mice. Our protocol offers ex vivo imaging, which helps avoid cataract and corneal opacity issues, enabling easier setup for faster experiments and overall higher throughput.

Our findings advance research by enabling visualization of microglia-neuron interactions, cellular communication and dynamics using transgenic fluorescent lines, as well as high-resolution time-lapse imaging.