We have shown recently that cell accumulation in culture of antisense DNA is strongly influenced by the presence of a 99mTc-MAG3 group for radiolabeling. We have now compared the in vitro and mouse in vivo behavior of 99mTc when radiolabeled to one antisense phosphorothioate DNA by three different methods. The 18-mer antisense DNA against the RIalpha subunit of PKA was conjugated via a primary amine on the 5'-end with the NHS esters of HYNIC and MAG3 and by the cyclic anhydride of DTPA. Surface plasmon resonance measurements revealed that the association rate constant for hybridization was unchanged for all three chelators as compared with that of the native DNA. Size exclusion HPLC showed rapid and quantitative protein binding for all three chelators upon incubation of labeled DNAs in 37 degrees C serum and cell culture medium. However, in each case, radiolabeled and intact oligonucleotide was still detectable after 24 h. Cellular uptake was tested in an RIalpha mRNA-positive cancer cell line. The order of cellular accumulation of 99mTc was DTPA>HYNIC(tricine) >MAG3, with the differences increasing with time between 4 and 24 h. The rate of 99mTc egress from cells was found to be MAG3>HYNIC>DTPA, which may explain the order of cellular accumulation. The biodistribution in normal mice was heavily influenced by the labeling method and followed a pattern similar to that seen previously by us for peptides labeled with the same chelators. In conclusion, although these studies concerned only one antisense DNA in one cell line, the results suggest that the success of antisense imaging may depend, in part, on the method of radiolabeling.