While I've never used one myself, I believe most of the clinical staining on large batches of samples are done with autostainers. Dako and Leica both make a line of them. One of the Dako models (I forget which one) can process something like 48 slides at a time, and shortens the staining procedures to as little as 3 hours - so you could run multiple 48 slide batches in a day. The systems are fairly expensive - if you have a clinical pathology department near you, perhaps they'd let you use one after hours?

By hand - I'm like you, I max out around 24 slides. Any more than that and you risk overdeveloping the AEC or DAB color development steps due to too many slides at once (IF obviously doesn't have that issue).

Joseph is right, if you have many slides, automatic slide stainer is the way. You can get it done from core facilities which are for pathology department. But mainly for research purpose, we do not need hundreds, we deal with 10-20 slides at a time. Keep time same for every set of slides,

An automatic stainer would definitely be the preferred method of choice but sadly is not always available. I assume you are only interested in the difference between group A and B and not in absolute levels. If this is the case I have an idea:

Make subgroups of your samples, for example 10 slides of group A and 10 slides of group B. Always stain the same number of slides for each group. Try to keep the conditions for every staining as constant as possible. Make an average of your values for each group A staining subset. Now normalize the values from each group A and B set with the corresponding group A average. Since you normalized each data set the values should now be comparable to each other.

I think this could work. What do you people think?

All suggestions are very good so far. If you can use a service with automated staining, use it. Otherwise you should first go through the plan of your samples and make sure you need all the stainings and tissues, basically reduce the number of samples. Then make sub-groups with manageble sample numbers and do several rounds of staining. When you do the normalization as Todor suggested, your samples will still be comparable.

Last tip: don't do it alone, get a student or colleague to help you ;)

Best of luck!

For Hand staining, I used to use a Thermo Shandon Sequenza (x2) which would hold 100 slides. You had to be set up, organised, smart and not DISTRACTED to do so many by hand. Dewax, DTW ,H202 block & HIER was done in staining racks/troughs of 25. Then covertiles, on to the Sequenza for the washes, serum block, primary overnight at 4C, washes, secondary... Back into staining racks/trough for the DAB and Heamatoxylin counterstain. DCM.

However nowadays I use a Leica BOND RX which takes 36 in a run, 3 runs a day.

Tips: do as Tilo suggests above. Use a well documented primary (ABCAM?) avoid using mouse primaries on mouse tissue (MOM interactions). Use a known positive control and do negative controls. Use a polymer based secondary, they are far better than ABC secodnaries and avoid any endogenous biotin background problems. Do a test run(s) before you commit a load of slides to the actual staining... Good luck!

All of the suggestions I have read so far are great. If you are going to do 100 slides in batches of 20 then you really need suitable controls to look for batch to batch variability.

As long as you are not worried about tissue biomarker heterogeneity, you could also construct a tissue microarray containing core from all the tissues on one block.

If you do use a Sequenza to do the IHC-F then protect the late stage fluorophore reagents from light with a sheet of aluminium cooking foil over the slides.