Effect of incomplete double strandedness of test DNA on DNA binding results.

Four commercial radiolabeled "double-stranded" DNA preparations were chromatographed on BNDC to determine proportion of incompletely ds-DNA in each. Sera from 10 patients with rheumatic diseases other than SLE, chosen for strong binding to one of the commercial DNAs, were retested with each of the other DNAs. Binding to each commercial DNA by all 10 sera was proportional to percent of that DNA containing ss-regions and in nine cases declined to zero for completely ds-DNA.