Properly orienting the specimen is a crucial step in this protocol. Because we are using anatomical features of the dorsal surface of the brain to locate LC (boundary between cerebellum and inferior colliculus), it is important that the sections be aligned properly. This requires care in properly setting the brain into the mouse brain slicer matrix. We recommend cutting ~500 μm more tissue anterior and posterior to LC to avoid missing the nucleus. The most common mistake is to cut too few sections that results in missing the LC entirely. Thus, for one’s first time following this protocol, we recommend cutting more sections than necessary. Careful study of the brain atlas images prior to staining is very helpful. The appearance of the brainstem changes appreciably every few hundred microns and, with some experience, it is possible to know what sections are worth staining simply by the macroscopic appearance.
During the process of localizing the LC, there might be variations in the signal depending on how well the brain was oriented during sectioning. When cutting through the center of the LC, the signal is bright and covers a larger area as compared to the edge of the LC, which will show up as a signal over a much smaller area. In the case that coronal slices are slightly tilted, the LC of one side of the 4th ventricle might be apparent and the one on the other side might only be visible in an adjacent slide. Thus, one cannot always expect the appearance of both LC regions at maximum intensity within one brain slice. This artifact can be avoided by cutting the brain exactly coronal in the mouse brain slicer matrix and carefully embedding the brain into the cubical embedding mold with OCT.
Immunostaining, at least with the anti-tyrosine hydroxylase antibody, is extremely forgiving and, in our experience, works on sections up to 100 μm in thickness. We have found that blocking solution is not necessary for high signal-to-noise staining of LC, reducing cost and reducing the amount of time needed to locate LC. In our experience, the staining protocol can be sped up – albeit with reduced quality and penetrance of staining – by reducing permeabilization to 2 h, primary antibody for 8 h, and secondary antibody for 2 h. Additionally, if one is simply interested in locating LC (e.g., for validating injection of a virus/tracer), sections from a fixed brain can be cut on a vibratome at 100 μm thickness.
One limitation of this protocol is it, by design, requires euthanizing the animal and removing the brain. Therefore, it is not useful for in vivo localization (e.g., electrophysiologic recordings). Another limitation is that this protocol requires PFA fixation which might alter the native state of the tissue. These alterations include the elemental content such as copper, calcium, iron and zinc36. The actual alteration of metal distribution caused by PFA fixation may be tested in one sample which can be run in parallel to a non-fixed sample. A comparison of the metal distribution between these two samples will provide evidence on the effect of PFA fixation on the distribution of the metal which is of interest in a certain study. If PFA fixation must be avoided, the general principle of this protocol (locating LC by immunostaining and using adjacent sections for follow-up experiments) can be extended to frozen sections without fixation.
We note that this protocol is largely a refinement of existing methods to solve this problem. The novelty exists in tailoring previous approaches to locate a very small, easily missed nucleus. We expect that this protocol can be easily modified and extended based on need (e.g., using transgenic animals expressing fluorophores in LC to avoid immunostaining).