Due to the heterogeneity of tumor perfusion, I would say that immersion is the better method here. If you plan to use much immunohistochemistry, I would use paraformaldeyhde, or restrict fixation time to 24-48 hours in neutral buffered formalin.

They are not mutually exclusive. If you're concerned about loss of antigen or induction of signals via tissue manipulation, pump-controlled slow perfusion of 10 mls 1x PBS, then 40-50 mls 4% paraformaldehyde followed by immersion is best. Paraformaldehyde overnight (for paraffin) or for 3 hours (for fixed/frozen) is my fixation of choice for maintaining tissue architecture.

Dr. Crawford is correct, they are not mutually exclusive. In my opinion, the added benefits to morphology of perfusion over immersion may not be significant enough to warrant the extra time to perfuse the animals. I find that the type of fixative is often more important than the means of fixation.

As I had mentioned previously, the benefits of perfusion, especially with something of minimal structure like a xenograft, is the in situ fixation aspect, minimizing the induction of signals that result from the tissue harvest. For instance, phosphorylation of ERK is quite prominently induced with tissue manipulation. So, for an accurate representation of the signaling environment in vivo, it's best to fix in situ.

Immersion in 10% neutral buffered formalin, for 48-72 hr is what I usually use for histology and also for immunohistochemistry with paraffin-tested antibodies.

I completely agree with Dr. Crawford. I only wanted to add one small suggestion if you intend to use frozen sections: in that case, I would also add a sucrose step, which acts as cryoprotectant, preventing ice crystal formation during freezing: transfer the tissue to 30% sucrose in PBS and leave at 4°C. Typically when the tissue sinks, it is fully impregnated and can then be rapidly frozen.

Dr. Soucek is correct with the 30% sucrose step, which I left out until I knew what you wanted. I would add that we generally equilibrate the tissue after the (generally overnight) 30% sucrose step, by transfering for 30 minutes or so into 50:50 30% sucrose:OCT, then into 100% OCT before freezing.

This discussion is very interesting and provides very good tips. We are mostly performing tissue collections of larger cohorts; in this case immersion in our hands is easier to handle. Since we are also interested in molecular characterization of the samples and are not always able to divide the sample, we are now using frequently PAXgene Tissue System from Qiagen. The morphology is (almost) as good as FFPE, but for RNA and DNA extraction it is far better.

The best way is to fix overnight in 4% PFA at 4 C. Then wash with PBS overnight at 4 C, dehydrate in ethanol, and embed in Paraffin blocks.

I know its an old discussion topic, but I am going to harvest tumor samples from mice and need to divide part of the tumor for protein lysate preparation, and the rest for fixation, and immunohistochemistry eventually. I am trying to decide what is the best fixative - I have a protocol that calls for 4% PFA overnight to 48 hours. But, since my animal will not be perfused, I am wondering whether a lower percent (like 2%) of PFA for longer time (48 hours) will be better. After the PFA, I would add a sucrose gradient of 10,20, and 30% until the tissue sinks, and then proceed with flash freezing, OCT mounting, and cryosectioning. Any suggestions whether I am going with correct approach? Thanks.