One objective of this investigation was to determine whether chemical modifications of oligonucleotides to permit radiolabeling with gamma- or positron emitters interferes with hybridization and target cell accumulation. A second objective was to establish to a reasonable extent whether cellular accumulation of radiolabeled oligonucleotides can be explained by an antisense mechanism. METHODS: An 18mer uniform phosphorothioate DNA antisense to the messenger RNA (mRNA) of the type I regulatory subunit alpha of cyclic adenosine monophosphate-dependent protein kinase A (RI alpha) was conjugated with the N-hydroxysuccinimidyl derivative of S-acetylmercaptoacetyltriglycine (MAG3) through a primary amine/linker and investigated in vitro in cell culture. RESULTS: By surface plasmon resonance, the association kinetics between native (i.e., without amine/linker) DNA and MAG3-amide/linker-DNA were identical. Melting temperatures were also identical for native DNA, amine/linker-DNA, and MAG3-amide/linker-DNA, indicating that these chemical modifications had no detectable influence on hybridization. However, cellular accumulation of (99m)Tc-MAG3-DNA was lower than that of (35)S-MAG3-DNA, suggesting that chemical modifications can have an important influence on cellular accumulation. In tissue culture studies of ACHN tumor cells (a human renal adenocarcinoma cell type), an antisense effect was suggested by 3 findings: an increased accumulation of (35)S- or (99m)Tc-labeled antisense versus sense DNA, an increased accumulation of (99m)Tc-antisense DNA in another RI alpha-positive tumor cell line (LS174T) but not in a murine transfected control cell line (HC-2), and the disappearance of the increased cellular accumulation of (99m)Tc-antisense DNA with increasing dosage of antisense DNA. Higher than expected cellular accumulations of about 10(5) antisense DNAs per cell over 24 h suggest stabilization of the target mRNA or increased mRNA production by the presence of the antisense DNA. In support of this suggestion, we observed, first, an increased incorporation of uridine-5'-triphosphate into RNA in cells exposed to the antisense DNA but not to the control DNA and, second, an increase in target mRNA expression in cells exposed to the antisense DNA but not to the control DNA. CONCLUSION: This evidence suggests tumor cell accumulation by an antisense mechanism. Moreover, the high level of DNA accumulation suggests that a rapid target mRNA turnover or transcription rate may be an important determinant of tumor counting rates.