Western Blotting problem - p21 detection in MCF7 lysates (Sep/06/2007 )

Hi,
I spent a lot of time trying to detect p21 by western in human lymphoma cells. As I recall, I had the best luck with a rabbit monoclonal from cell signaling:
http://www.cellsignal.com/products/2947.html
It's likely that you're getting non-specific bands. I would run a western with a "lysate curve". Run 10ug, 20ug, 30ug, 40ug, 50ug, & 60ug. Use less of the primary antibody than you're presnetly using (cut it by 25-50%).
As for secondaries, I'm VERY particular to use only those from Jackson laboratories:
http://www.jacksonimmuno.com/home.asp
I'm always suspicious when one must use a 1:1000-2000 dilution with their secondaries. With the Jackson secondaries, I use a 1:20,000 dilution and get excellent signal/no background. If you keep what you have, try a 1:4000 dilution.

Also, are you blocking prior to primary incubation? If not, try either 5% non-fat dry milk/TBST or 3% BSA/TBST. These may reduce any non-specific antibody interactions.

Good luck!