Interesting How about a "disarmed retrovirus" that encodes an antisense mRNA copy of the receptor protein for a particular retrovirus. Could this protect cells against infection?