You could try the following, in case you haven’t already:

1.     Grow cells to mid-log phase

2.     Pellet and wash cells with 0.5 M sucrose 3-4 times

3.     Finally suspend cells in 0.5 M sucrose and leave on ice for 15-30 min before aliquoting and storing at -80°C

4.     Thaw and mix ~40 μL cells with your DNA and transfer the mix to a 0.1 cm cuvette (BioRad)

5.     Apply a single pulse using the settings (BioRad GenePulser): Voltage (1.8 kV); Resistance (100 Ω); Capacitance (25 μF)

6.     Typically a time constant (τ) of 2.4 ms is observed

7.     Plate on the selective medium after recovery

However, since you already tried several protocols and none worked suggests either the strain isn’t accepting the DNA or the product (if being over-expressed from a multicopy plasmid/hot promoter) is toxic to the cells. As you are aware that staphylococci may not accept foreign DNA (extracted from routinely used E. coli) directly owing to an active restriction/modification system, the usual way is to first transform a recipient strain (e.g. RN4220) and then move the DNA to your strain of choice using either phage-mediated transduction, or transformation.

 Hope things work out for you.

Thank you very much Krishan I will try your protocol.

As you mentioned RN4220 is a good choice as an intermediate strain in the transformation of clinical isolates. you said move the DNA to other strains do you by any chance have a protocol of plasmid extraction from S. aureus that had worked for you. 

Moving DNA between E. coli and S. aureus is not as straightforward as it sounds. I agree with Krishan's excellent answer - electroporating plasmids isolated from standard E. coli strains is rarely successful due to staph's impressive restriction barrier. Rather than going through the RN4220 intermediary, I suggest using the engineered E. coli strains DC10B or IM08B. They modify plasmids consistent with S. aureus restriction patterns and DNA isolated from these E. coli strains can be directly electroporated into S. aureus. I've been able to do all my standard cloning in these strains and make all my S. aureus mutants easily with this method.

If you intend to use RN4220, I usually add 500 µl glass beads to overnight culture (1-5 ml) in a microfuge tube and vortex for 2 minutes to physically break down the cell walls, then proceed with standard plasmid prep protocols (commercial kits work). Cheaper and often more effective than lysostaphin / lysozyme treatment.

Additionally, I typically substitute 10% glycerol for the 500mM sucrose in Krishan's protocol, but either works as a non-ionic solute and cryoprotectant. Good luck!

NB - the IM08B strain is analogous to DC10B but optimized to have the methylation patterns recognized by clonal complex 8 S. aureus.

http://www.ncbi.nlm.nih.gov/pubmed/22434850

1.     Pellet ~10^9 cells (one mL of overnight grown culture or something like that), wash once with chilled water, and resuspend in 250 μL of P1 Buffer (Qiagen) containing RNAse.

 2.     Add 5 μL of 5 mg/mL lysostaphin and incubate at 37°C till the cells are completely lysed. Alternatively, like Andrew said, instead of adding lysostaphin, you could use silica beads/bead beater to lyse the cells (cheap and effective).

 3.     Add 250 μL of P2 Buffer and mix gently.

 4.     Add 350 μL of P3 Buffer and mix thoroughly.

 5.     Centrifuge to clear the debris and load the supernatant on Qiagen mini-prep column. Wash, and elute the plasmid DNA.