I am not sure about CD3 but I have had a colleague who looked at the down regulation of TCRb following activation in the presence of Salmonella. The trick to it is taking the geometic mean of your sample. It will be a subtle shift if you are looking at a histogram plot. Typically overlays are not able to conclusively demonstrate it. First label the amount of surface receptor with one fluorophore, then permeablize the cells and relabel with a second antibody/fluorophore pair which will label both inside and outside. That way you can use the ratio to determine down regulation.
We do it for TCR on T cells which are primed with APCs pulsed with cognate ligand and we see very beautiful down-regulation with titrating doses of ligand. My guess is that CD3 will also be down-regulated since it is associated with TCR.
As we all expect, the histogram shifts would be very subtle but if you plot your data on linear scale then you can see dramatic differences.
What I would also suggest is to look at both surface and total( by adding a permeabilizing agent such as tritonX-100) CD3 levels.