This page describes the formation of the "anchoring segment" which is subsequently ligated to the insert oligo to create the base unzipping construct.
The goal is to obtain a double-stranded DNA (dsDNA) fragment of about 1 kilobase pair (kb) length, with a digoxigenin (dig) label on one blunt end, and an unlabeled sticky end on the other. The method we used was to do a PCR reaction with a dig-labeled primer, followed by a restriction digest of a single recognition site close to the non-labeled end. The short fragment was discarded while the large fragment was purified by gel extraction (and / or PEG precipitation).
Our specific protocol was evolved from existing PCR recipes involving the plasmid pRL574 from the Bob Landick lab. It's kind of funny how many single-molecule tethering experiments have used this plasmid, without any need for the underlying gene it encodes. We, too, do not require this plasmid exactly, but use it for convenience. If you're going to redesign, starting with a commercial plasmid that you can buy large amounts of purified plasmid prep would make sense.
There are many possible alternatives for preparing this piece. One would be to restriction digest a plasmid and ligate a dig-labeled hairpin or duplex onto one end. This would probably work very well, and allow for multiple-dig labels and / or a hairpin to increase the anchoring segment stability. We have not yet tried this as of 2007.
pRL574 purified plasmid DNA, concentration about 1.5 nanogram / microliter. This is about a 1:100 dilution of miniprep DNA. We obtained this plasmid from Bob Landick lab. The sequence of the plasmid is here.