I do not have any first-hand experience with exosomes, but maybe some techniques I know for synaptosomes are applicable.
I suggest to coat clean glass cover slips in a 24-well plate with 5% BSA in PBS over night. Then discard the coating medium, apply some of your exosome isolate (worth a few µg of exosome particles; a few µl can be enough), and centrifuge the plate hard on 4 degrees Celsius for as long as it takes (might be a few hours, exosomes are quite small and not very dense, after all; I lack exosome experience to make a more specific suggestion). After that, fix with 4% PFA for 10' on ice, followed by 30' on RT. Then wash, quench, and process like any ICC sample.
Should it prove impossible to centrifuge the exosomes onto the cover slip, I have another, more adventurous idea: incubate the isolated exosomes with an antibody (which you know works well) that has previously been coupled to magnetic beads. Then apply the exosomes to a cover slip prepared as above, place onto a magnetic plate, and fix then. The exosomes should be pushed onto the BSA coat and then crosslink to it in sufficient numbers.
The trick is just to find a way to immobilise the exosomes on a coated glass surface for long enough to achieve fixation to that surface.
First, you must know that it's quite unlikely that you see anything using IF on exosomes. They are so small that it is pretty complicated to get enough ab binding to be detected (though there are some groups that somehow have managed to do so). This said, I don't know if you have any experience with exosomes but to pellet them you need, at least, 100000g. These speeds are only reachable by ultracentrifugation and they are not supported by any type of plates/coverslips. We usually do transmission electron microscopy dying our exosomes with gold with an anti-CD63, but since we send the samples to the imaging platform I don't have a clear protocol to give you. Sorry and good luck!
we have performed a lot of electron microscopy on immunolabeled exosomes. drop me an email on [email protected] and I can send you the protocol that we use. We have also done transmission electron microscopy on exosomes isolated with magnetic beads.
I think the Lotvalls editorial also suggests that the use of technology such as NTA for size and conc estimations should always be combined with electron microscopy as this technology does not distinguish between exos and other vesicular structures or protein aggregates.