Optical antisense imaging of tumor with fluorescent DNA duplexes

Antisense targeting of tumor with fluorescent conjugated DNA oligomers has the potential of improving tumor/normal tissue ratios over that achievable by nuclear antisense imaging. When administered as a linear duplex of two fluorophore-conjugated oligomers arranged in a manner that inhibits fluorescence as the duplex and designed to dissociate only in the presence of the target mRNA, the fluorescence signal should in principle be inhibited everywhere except in the target cell. Optical imaging by fluorescence quenching using linear fluorophore-conjugated oligomers has not been extensively investigated and may not have been previously considered for antisense targeting. We evaluated in cell culture and in KB-G2 tumor bearing nude mice a 25-mer phosphorothioate (PS) anti- mdr1 antisense DNA conjugated with the Cy5.5 emitter on its 3' equivalent end and hybridized as a linear duplex with a shorter 18-mer phosphodiester (PO) complementary DNA (cDNA) with the Black Hole inhibitor BHQ3 on its 5' end. In serum environments, 90% of the DNA25-Cy5.5 fluorescence was inhibited immediately following addition of the cDNA18-BHQ3 and showed only slight loss of inhibition over 24 h at 37 degrees C. As evidence of antisense specific binding, when incubated with the DNA25-Cy5.5/cDNA18-BHQ3 duplex, the fluorescence was lower in KB-31 (Pgp +/-) cells compared to KB-G2 (Pgp++) cells, but when incubated with the control cDNA18-Cy5.5/DNA25-BHQ3 duplex in which the fluorophores were reversed, the fluorescence of both cell types was low. As further evidence of specific binding, the fluorescent intensity of total RNA from KB-G2 cells incubated with the study duplex showed evidence of dissociation and hybridization with the target mRNA. Furthermore, the fluorescence microscopy images of KB-G2 cells incubated with DNA25-Cy5.5 as the singlet or study duplex show that migration in both cases is to the nucleus. The animal studies were performed in mice bearing KB-G2 tumor in one thigh and receiving iv the study or control duplexes. The tumor/normal thigh fluorescence ratio was clearly positive as early as 30 min postinjection in the study mice and reached a maximum at 5 h. By contrast, much lower fluorescence was observed in mice receiving the control duplex at the same dosage. Fluorescence microscope imaging showed that the Cy5.5 fluorescence was much higher in tumor sections from the animal that had received the study rather than control duplex. Thus combining a fluorophore-conjugated antisense DNA with an inhibitor-conjugated shorter complementary cDNA inhibited fluorescence both in cell culture and in tumored animals except in the presence of the target mRNA. This proof of concept investigation of optical antisense targeting therefore suggests that further studies including optimization of this approach are appropriate.