I've been trying to do immunohistochemical detection of GAD67 in mouse brain structures. I've been tested some anti-GAD67 antibodies, but they didn't work well.
Moreover, are there any "hints" someone can give to get it working? Thanks.
you might find this link extremely useful: http://onlinelibrary.wiley.com/journal/10.1002/%28ISSN%291096-9861/homepage/jcn_antibody_database.htm
It provides you with a huge non-commercially motivated database on ABs and their respective references. As Martin mentioned, Anti-GAD 67 by Chemicon (Millipore) works great in the hands of several independent research teams.
As an aside, there's a monoclonal antibody to GAD65 that also works beautifully: clone GAD-6, which is available from the Developmental Studies Hybridoma Bank, University of Iowa, subsidized by NIH. I don't know about its availability outside North America, but we can get it in Canada and we are not NIH-funded. see: dshb.biology.uiowa.edu/
We bought some anti-GA67 clone 1G10.2 (Millipore #MAB5406) and compared it side-by-side with some older anti-GAD67 clone K-87 (Abcam #ab26116) that we have.
We ran an adult mouse brain homogenate across an entire minigel, immunoblotted (using Amersham Hybond Low Fluorescence 2 um PVDF) and cut the blot into strips and incubated the strips with dilutions of 1/500, 1:1,000, 1/2000, 1/4,000, 1/8,000 or 1/16,000 of each antibody. The secondary was horse anti-mouse IgG HRP-linked (Cell Signaling #7076) at 1/2000. For detection we used the Pierce ECL 2 kit. Both antibodies worked very well. Both stained only a single band at about 67K with no other bands visible even at 1/500. I’d say the optimal concentration is about 1/2,000 for both antibodies, but they both still have good signal at 1/16,000.
For immunohistochemistry we followed each supplier’s recommendations (e.g., antigen retrieval prior to K-87) and used a 1/4,000 dilution of each antibody on adult mouse brain sections. Both antibodies worked very well for labeling neuronal cell bodies. The K-87 staining extended more into arborizations as well as nerve terminals (e.g., in cerebellar cortex). GAD67 primarily forms homodimers with itself and stays in the cell body, but some GAD67 forms heterodimers with GAD65 and gets transported to synapses (e.g., PMID: 20805323) so such labeling is expected. If you want to visualize those details the K-87 would be preferable. Otherwise, either 1G10.2 or K-87 would be good choices. Both are often used in the literature. For GAD65 labeling, the GAD6 mAb is excellent for immunoblotting and immunohistochemistry and is widely used in the literature.