pGEX derivatives from GE is used extensively for purification of GST-tagged proteins for immunization. U can also use the pET expression systems as described in the detailed protocol article below: http://www.ncbi.nlm.nih.gov/pubmed/18428660
Whether u need a large tag like GST or can use a small His tag depends on the solubility and nature of ur protein (GST makes proteins more soluble, but since the fusion tag is big, u can have GST-specific antibodies if u do not chop off the tag).
It's important to know whether this is a secreted or intracellular protein. S. aureus secretes lots of proteins, and purification from the medium may be the best way to isolate some of these. Also, many of these extracellular proteins have disulfide bonds, which won't form correctly during standard intracellular expression in E. coli.
If it is an intracellular protein, then recombinant expression in E. coli might be your best shot, as Manoj and Arvind mention. I'd advise you to look at Current Protocols (the big red binders of protocols that many labs have) for the details.
Be aware that S. aureus has a very AT rich genome, and E. coli is comparatively GC rich. That means that the codon usage of an S. aureus gene may not be well-optimized for expression in E. coli; see http://www.embl.de/pepcore/pepcore_services/protein_expression/ecoli/optimisation_expression_levels/ for details.
There are strains of E. coli with extra copies of rare tRNAs which can help alleviate the codon-bias problem: Rosetta from Novagen and CodonPlus from Stratagene.
Thank you so much Wes. We are looking at intracellular protein and i think i will have to go with cloning. I will look at the suggestion you gave me. I really appreciate it.