Measuring spine density

The purpose of this tutorial will be to describe to a observer, blind to the conditions of the experiment, how to determine dendritic spine densities in 2-photon micropscopy images.

1) The ImageJ program is required and is free to download (http://rsbweb.nih.gov/ij/download.html).

2) You will ideally be provided with a folder of randomly labeled TIF images such as:

Control conditions (http://www.mediafire.com/?mtdvxy4pkhu7e74)

Frames that were out of focus or otherwise indistinct were discarded but should you have any trouble distinguishing the dendrites or spines please bring this to the attention of the experimenter. Likewise, while the dendrite in question should be obvious, if it is not, please also notify the experimenter.

3) As the first part will be to identify the number of spines on a particular dendrite, there may several different images of the same dendrite, from various magnifications, depths, and views in order to aid you in differentiating what is and isn't a spine.

The long, uninterrupted, branching lines are "dendrites."

The small, bulbous protrusions from the dendrites are "spines" or "dendritic spines."

"Dendritic spine density" will be determined by:

Dendritic spine density = (# of Dendritic spines) / ((Measured Length of dendrite / Measured Length of scale line) *

Scale)

E.g.: (13 Dendritic spines) / ((3.34 / 1.094) * 10 micrometers) = .4258 spines per micrometer

Open the first image in ImageJ. It may look something like this:

The dendrite in question here is the long line running horizontally through the image. You may ignore the text at the top. The bottom line and number is the scale of the image: thus in this particular image, the length of the line is 10 micrometers. This will be used to measure the length of the dendrite.

4) A spreadsheet program is not necessarily required but may be useful to keep track of certain numbers.

To determine the number of dendritic spines, you may need to compare various images of the same spine using ImageJ. For an object to be considered a spine, it needs to meet several criteria:

4.1) First, it has to clearly protrude out of the dendrite shaft by at least 3 pixels (0.5 m) or has to retract completely into the shaft of the dendrite.

4.2) Second, it has to be clearly identifiable (i.e., show high signal-to-noise ratio) without any large changes in image quality between sessions. For example, any spines from images that do not appear sharp or crisp, or are perhaps attributable to sudden movement, must be excluded from the analysis. Thus, it cannot be a shadow or blur of another spine.

4.3) Third, it must be distinct from a crossing dendrite.

* Several features can facilitate making these distinctions:

Edge Detection

Connection of the spine to the dendrite can be more clearly seen in this mode.

Unsharp Mask

This mode can improve the brightness and contrast, as some of the images may be too dark.

* As a note: Because Knott et al., 2006 showed that it is difficult to distinguish dendritic filopodium from thin structures with optical microscopy, all dendritic protrusions are considered spines.

As you work your way along the length of the dendrite, you may encounter an ambiguous object that may or may not be a spine - if you do, go ahead and stop there. Ideally, you will want to only count the number of distinguishable spines in a row. Make sure to notate how many spines there were on that given length of dendrite.

5) Go to the Line Tool (5th from the left), right click and select Freehand Line.

Trace the length of the dendrite that you examined.

Remember to save these numbers.

6) Do this for each image. After collecting the data, calculate dendritic spine density

"Dendritic spine density" will be determined by:

Dendritic spine density = (# of Dendritic spines) / ((Measured Length of dendrite / Measured Length of scale line) * Scale)

E.g.: (13 Dendritic spines) / ((3.34 / 1.094) * 10 micrometers) = .4258 spines per micrometer