Recognizing synapses in the zfish dataset
When tracing in zfish, it is sometimes useful to know when you have an axon or a dendrite. For example, you may be wondering if a branch you've added belongs. If you could go back and see that the trace before the connection is an axon, and the trace after the connection is a dendrite, then you know you have a merger, because an axon never turns into a dendrite. Likewise, a dendrite never turns into an axon. (Sometimes an axon will start on what looks like a dendritic branch, but whenever this happens, it occurs not far from the cell body, and I (twister2) think the branch before the axon splits off is not yet dendritic nor axonal.)
Another situation where it's useful to know if you have an axon or dendrite, is when you suspect you have two axons in a cell. All cells should have no more than one axon. If you can recognize that you have two different axons, you know at least one of them must be a merger.
Sometimes you can tell from overview when you have an axon or dendrite, but for more certainty, you'll want to go inside some cubes and look in the 2D panel for synapses along the borders of your neuron. When you've identified a synapse, you can then tell which side is a dendrite and which side is an axon.
To identify synapses in zfish cells, look for 2 things:
- A short stretch of border between 2 cells, that is darker and thicker than the rest of the border.
- An area of increased neurotransmitter vesicle density on one side of the dark border. (Neurotransmitter vesicles are little dark spots about 4-8 pixels across. You can see what they look like in the examples below.)
When you've identified a synapse, it should be easy to tell which side of the border is the axon and which is the dendrite. The area of higher vesicle density marks the presynaptic neurite, or axon. The other side is the postsynaptic neurite, or dendrite.
Note: there is such a thing as a postsynaptic density, but I'm not sure if that is visible in zfish slides. Ironically, the presynaptic side is visibly denser than the postsynaptic side due to the neurotransmitter vesicles.
Below is an example of a synapse, from cube #2134873, cell 77327. (Click to enlarge)
The axon is circled red and the dendrite circled green. The yellow square roughly indicates the borders of the cropped images in the bottom row. Each of those images is the same image with different annotations drawn on them (except for the first image, which allows you to see the original greyscale micrograph with no obstructing highlights/markings). The bottom row of images, from left to right, show:
- Original greyscale image
- Dark/thick border (circled red), which along with NT vesicles in #3, indicates a possible synapse
- Area of higher neurotransmitter vesicle density (circled red), which indicates the axon
- Axon (outlined in red) vs dendrite (outlined in green)
Several more examples of synapses in zfish will be added here over time. The images below are thumbnails. Please click them to view at full size.
All examples will show the original greyscale image on the left, and an outlined version of it on the right. The outline colors are red for axons, green for dendrites, and yellow for areas of high vesicle density. The synapses themselves aren't outlined, but they should be obvious: they're the dark borders located where the red, green and yellow outlines come together.
Cube #2676281: (multiple synapses in one shot!)
Cube #2752336: (a pair of synapses!)
Ok, I guess multiple synapses in a single image are going to be pretty common. :p Three synapses coming right up! Cube #2756772:
Just one synapse. Cube #2729001:
I like the way this synapse curves. Cube #2676710: