We are ectopically expressing splice variants of ING1 (a tumor suppressor) in some head and neck cancer cell lines using pcDNA3.1/His B plasmid. For some reason, the cells receiving empty vector are dying at the same rate and about the same time as the cells receiving the vector with our gene of interest. We have sequenced our vectors in-house and everything looks as it should. Could endotoxin contamination cause this? Any other ideas?