Quantification of Dendritic Spines Using Confocal Microscopy Imaging

Take pyramidal neurons that are transduced and immunolabeled to fluoresce green.

Acquire images of the neurons using a confocal laser-scanning microscope with settings adjusted to ensure optimal image quality.

Open an image using specialized software and use the Gaussian filter to reduce the background noise.

Perform semi-automatic dendrite tracing by first estimating the dendrite diameter. Then, select the dendrite's starting point, path, and endpoint.

For automated spine segmentation, select rebuild dendrite diameter and set dendrite volume threshold to represent the actual dendrite volume.

Specify values for the smallest spine head diameter and the maximal spine length.

Adjust the threshold for points representing spines, ensuring their localization on actual spine heads.

Set parameters to classify the spines based on length, head diameter, and shape.

Generate a 3D model of the dendrite with spines and export the data for further analysis.

Under a confocal microscope, select at least 10 healthy neurons per condition with a pyramidal morphology and a full dendritic arborization. To quantify 60 to 100 micrometers per dendrite, acquire images with a 40x, NA 1.3 oil objective and a 488-nanometer laser with a typical peak power at sample level around 20 microwatts.

Set the pixel size around 80 nanometers to properly sample dendritic spines. To sample the whole neuron volume, acquire a Z stack with a z spacing from 150 to 300 nanometers, yielding 20 to 30 z slices. To carry out 3D quantification of dendritic spines, use the following settings for the analysis software.

For pre-processing, use Gaussian filtering by choosing Image Processing, Smoothing, Gaussian filter. Set the filter width equal to the pixel size in the XY plane. To perform a semi-automatic tracing of dendrites, from the filament tracer module in the Slice tab, use the Distance tool to estimate the dendrite diameter.

Next, in the Surpass tab, click on the Filaments tool and choose Skip Automatic Creation for better robustness. In the Draw tab now presented, select AutoPath as a method, Dendrite as a type, and input the estimated dendrite diameter. Using the select mode of the pointer, turn the cursor into a box, and choose Shift and right-click on the dendrite starting point. The software will perform initial calculations.

Now, move the pointer along the dendrite. From the starting point, a yellow line representing the most likely dendrite path is shown. Press Shift and left-click on the dendrite endpoint. To carry out automated spine segmentation, in the module interface, click on the Creation tab. In the rebuild drop list, choose Rebuild Dendrite Diameter and check the Keep Data box. Then click Rebuild.

Set the threshold so that the segmented volume corresponds to the actual dendrite volume. As an algorithm, select the shortest distance from distance map. Then click the Next button. Under the Slice tab, determine the smallest spine head diameter and the maximal length.

Then under the Surpass tab, enter the parameters. Values around 200 to 300 nanometers for minimal diameter and 4 micrometers for maximal length are good starting points. Without checking the Allow Spines box, click the Next button.

Next, adjust the seed points threshold so that the blue points representing spines localize to the actual spine heads. Then click Next. The core calculation will now be performed and can take time. To classify spines, under the Tools tab of the module interface, click on Classify Spines.

After verifying that there are four classes as outlined in the text protocol, the module interface will generate four new filament objects with the results of the classification. Finally, export the statistical data, under the Statistics tab by clicking on Export All Statistics to File.