That depends on which proteins you want to go for. If you are interested in cytosolic proteins, then use a buffer w/o detergent, and remove organelles by ultracentrifugation. If you are interested in a particular organelle, isolate that first and then solubilise with detergent.

Your buffer obviously needs protease inhibitors. One of those is EDTA, it inactivates metalloproteases. If you can get away without depends on your sample.

Another thing to keep in mind: Many people use buffers with high Na and low K, like PBS. However, this is the composition of the extracellular environment, the cytosol has high K. I'd use a HEPES/KOH buffer pH 6.8, say 20 mM. Then add some salt to reduce unspecific protein-protein interactions (say, 25 mM KCl) and a reducing agent (0.5 mM DTT). Some proteins have special requirements, like Mg. Detergents and (thio-)urea can increase protein yield, but may interfere with plate binding later, only experiments can tell. If you want to use IEF, avoid ionic detergents. 1 mM Na-fluoride can prevent bacterial growth, if that is an issue.

Please clarify which analyte (low or high molecular weight) from sciatic nerve homogenate you want to estimate by ELISA. You have to keep in mind that the buffer and its ingredients you are going to use during lysis does not modify your analyte of interest (thereby no alteration in antigen-antibody interaction) and interfere during analysis (inhibiting the activity of enzyme used in ELISA) other wise you will get absurd/erratic results.