I found a lot of paper using whole cell dot immunoblots to detect surface protein. Antibody may be specific IgG or serum, and working with wildtype of strain. But protein A will bind to any IgG, should it affect the results?
If you run western blots instead of dot blots, you will be able to disinguish between protein A and your protein, provided that the have a different size. With dot blots, a protein A-negative strain would be the best option. Theoretically, you could try to saturate protein A by flooding it with an excess of non-specific IgGs, but then you would have to label your specific antibody (monoclonal or affinity-purified) directly by chemically coupling it to some enzyme or fluorophore, since using a second anti-IgG antibody would be ruled out.