Restriction Digest Woes - (Aug/16/2005 )

I have been attempting to knockout a gene from the MHV genome for quite some time and I am stuck on my final digestion step. I am attempting to cut my vector (and insert) with SexAI and either ClaI or SpeI, depending on the knockout construct. The vector is a pSMART vector containing a fragment of the MHV genome that contains the gene I want to delete. My problem is this: I am able to cut the vector with each of the aforementioned enzymes by themselves just fine. When I attempt a double digest (SexAI and ClaI have 100% activity in NEB4, and SpeI has 75% activity), I get very inefficient cutting of the vector. When I attempt to cut first with SexAI and then either ClaI or SpeI (or vice versa), using a PCR purification column to clean the reactions before the second enzyme, I only get cutting with the initial enzyme. I carry out my reactions between 2-3 hours and am using 3-5U of enzyme per ug of DNA. It has been suggested to me to try chloroform extraction of my DNA between cuts in the event that the PCR purification column is leaving something behind that is inhibiting the second digest. Does anyone know of this happening? And are there any other suggestions as to where to go from here? Thanks for any help in advance!!

I should add that the restriction sites in the vector area bout 2kb apart, so the enzymes should not be inhibiting eachother's recognition sequence.