Same-Dendrite Same-Axon (SDSA) Analysis is used to compare the properties of dendritic spines which share similar pre- and post-synaptic activation histories.

Getting Started

The first thing to do when beginning an SDSA analysis is to populate a spreadsheet with appropriate information. This includes the starting and ending sections of protrusion stamps, contact flat area measurements, and spine names. To do this, filter the Reconstruct object list for your desired traces and export the list to a CSV. [Objects > List > Save]

Be sure to sort the spreadsheet by protrusion starting section as some protrusions could be added after a first pass. Protrusions may follow the numbered naming count but not be placed sequentially. For example, a missed protrusion could be noticed later and named “d01p33,” but actually occur between protrusions 13 and 14. 

Selecting the “Middle 15” Synapses

Using this ordered spreadsheet select the middle viable protrusion of the dendrite. If there are an even number of protrusions there will be 2. From the middle protrusion(s) count UP and DOWN the dendrite to include 7 more viable protrusions in each direction for a total of 15-16 initial sample synapses.

NOTE: For the purposes of this analysis a ‘viable’ protrusion is defined as a protrusion leading to a fully formed spine that includes an excitatory synapse. Therefore the protrusion stamps for shaft synapses, inhibitory spines, non-synaptic spines, and incomplete spines should be skipped. Multisynaptic and branched spines share protrusions, though all synapses will be included in the analysis so each synapse should be individually counted.

Protrusion Overlap

So as to not exclude relevant data and to respect protrusion frequency between dendrites, also include any synapses stemming from viable protrusions which overlap with the start and end of the selected “middle 15.” 

This inclusion is contiguous. Once a new protrusion is added to the selection, any other protrusion which overlaps with it must also be included. This includes any non-viable protrusion stamps as they may themselves overlap with viable ones. Continue this chain until a clean break is reached where no other protrusions overlap with the selection (i.e. there are no other protrusion stamps on the same section at the top and bottom of the selected region). While the initial selection was for the “middle 15” synapses, the total number of synapses analyzed for a dendrite can be more. 

It may be helpful, but not necessary, to create a Protrusion Map for each dendrite (see below).

In this partial SDSA spreadsheet for a dendrite, column A is used to highlight the selected middle 15+ synapses and their associated protrusions. 

The original 15 (yellow) are flanked by additional protrusions included due to overlap (green) and clean breaks between protrusions (red). Column B is a running count of viable protrusions along the whole dendrite. In this example there will be 17 axons traced and analyzed. 

Protrusion Maps (Optional)

Using the list of protrusions for each separate dendrite, color and name the corresponding cells in a spreadsheet (with rows being a proxy for section #) where the protrusion stamps occur. If stamps overlap with one another, place them in separate columns. 

In this example map you can see blocks representing protrusions for a dendrite in a series. The row numbers on the left correspond to section numbers in the series. Each block is labeled for a protrusion stamp and colored based on its categorization: yellow for typical single-synaptic excitatory spines, blue for branched spines, pink for multisynaptic spines, and orange for shaft synapses.

Determining ROI Analyzed Length

Once a sampling selection has been made, determine the z-length of the region of interest per dendrite. This will be one of the following two options, whichever is longer:

1. From the lowest # section of the first INCLUDED protrusion to the lowest # section of the next EXCLUDED protrusion following the ROI (red)
2. From the highest # section of the last INCLUDED protrusion to the highest # section of the first EXCLUDED protrusion preceding the ROI (blue)

Using the example spreadsheet from above, the two options for z-lengths are:

1. From section 64 (the lowest # section of the first INCLUDED protrusion) to section 130 (the lowest # section of the first EXCLUDED protrusion following the ROI), giving a total of 66 sections. 
2. From section 58 (the highest # section of the first EXCLUDED protrusion preceding the ROI) to section 128 (the highest # section of the last INCLUDED protrusion), giving a total of 70 sections. 

In this example the analyzed length would be the z-length between sections 58 and 128 (option 2) because it is longer. 

Tracing Axons

Once the spreadsheet is complete and synapses for a dendrite have been selected for analysis it is time to start tracing axons. For each of the selected sample synapses, trace its associated axon out until it reaches a distance of around 2 microns away from its partner’s dendritic shaft or meets the edge of a series. 

Axons can be tricky and mercurial. Sometimes they can be running directly away from a dendrite then loop back around to form another synapse. Other times they stick close to and run alongside a dendrite for most of a series.

This 2 micron distance is a soft limit. Use your own judgment to determine if an axon is likely to get close to another spine from the dendrite of interest by tracking the axon’s directionality and looking at the axon and spine traces in 3D space. 

Because of the typical series sectioning orientation, dendritic shafts are usually in cross section making them easy to follow and trace. However, axons usually run perpendicular-ish to the dendritic arbor. This means that most axons will be horizontally or obliquely sectioned. Be sure to follow the standard Grey Wall tracing method for these objects to ensure accurate tracing and subsequent measurement, such as for axonal bouton surface area.

To be added:

• Types of Analysis. Originally Spine Head Volume was used as a proxy for synaptic potency (See Bromer et al. 2018) along with a signal detection theory approach when calculating synaptic size categories. Mohammad Samavat is currently working on a re-analysis of this work using Contact Flat Area with an information theory approach. We are unsure of a set standard of analysis going forward, but this process of data collection should remain unchanged.