Phenol has been superficially explored as a protein solvent for proteomics. Factors influencing extraction of various protein types aren't yet well understood, nor are factors influencing precipitation. The composition of the phenol phase will affect extraction and solvent composition will affect precipitation.
An important variable
Large protein pellets are difficult and time consuming to wash. For this reason precipitate only enough protein (from phenol) to run a few gels.
Improving phenol extraction
Hydrogen bonding of phenol to the peptide backbone as well as hydrophobic interactions promote protein solubility. Solubility in phenol requires (for most proteins) denaturation. Phenol itself is a good denaturant, but heating and the presence of SDS (and in some cases salts) facilitates dissolution. Effects of pH (adding various amounts of base, e.g. N-ethylmorpholine) have not been explored. Adding stronger phenolic acids (e.g. p-nitrophenol) or multivalent phenols (e.g. resorcinol) can improve solubility. Effects of salts haven't been explored. Glycoproteins with sizeable carbohydrate compositions will not be soluble in phenol.
Improving protein precipitation
Methanol is a good precipitant, but some proteins are alcohol soluble (zein is a classic). Acetone is a better precipitant, but also precipitates phospholipids. Ethers (e.g. THF) are also potential precipitants.
Ideally one will use a model system containing a (limited) variety of proteins varying in isoelectric point and hydrophobicity. By analyzing total proteins alongside phenol extracted proteins one can in principle identify factors influencing extraction.
With commercial zein as a model phenol extracts all components (visible on a 1d SDS gel) and acetone, but not methanol, precipitates the components.
Some useful tips
Most biological dyes partition into the phenol phase. Ponceau S is a useful aqueous phase marker although at higher salt (LiCl) concentrations it distributes detectably into phenol. Extracting proteins from 40% sucrose solutions keeps the phenol phase on top; however sucrose partitions significantly into phenol. Acetone, but not methanol precipitates this sucrose. Either avoid sucrose when precipitating with acetone or wash the phenol phase to remove residual sucrose (or other low MW contaminants). Salt (LiCl) facilitates extraction of basic proteins from PBCV1, a large DNA containing virus.
Phenol is denser than water and naturally forms the lower phase. Phase positions can be changed by adjusting densities. Adding sucrose to the aqueous phase is one strategy, adding lighter hydrocarbons to the phenol phase is another. Alkanes (e.g. hexane) are the lowest density hydrocarbons but are immiscible with phenol. 1/3 volume (relative to phenol) of toluene makes the phenol layer less dense than the aqueous phase. (see also) With PBCV virion proteins, protein recovery is slightly less, but there's no obvious selective loss.