siRNA "knock-out"/transfection "knock-in" - (Dec/30/2007 )

HA-HA.. my boss suggested I try to do the exact same thing with my gene of interest! I can do the siRNA with excellent protein knockdown but now we want to see if we can rescue specific functions (but not others) with mutants. I bought the siRNA from Dharmacon. They offer really good "SMARTpool" siRNAs which they guarantee. Now I need to design the mutants with additional silent mutations that make them siRNA resistant. Because the pool contains four different siRNAs I have to address each and every one. At least they give you the sequences. However, I'm thinking of redesigning the siRNA for the 5' and 3' UTR of the endogenous gene but I'll have to start over on an efficient knockdown protocol. But at least this way I know there won't be any problems with the expressed mutants. Any ideas or suggestions on which route would be best?