Binding, uptake, and intracellular trafficking of phosphorothioate-modified oligodeoxynucleotides.

An enhanced appreciation of uptake mechanisms and intracellular trafficking of phosphorothioate modified oligodeoxynucleotides (P-ODN) might facilitate the use of these compounds for experimental and therapeutic purposes. We addressed these issues by identifying cell surface proteins with which P-ODN specifically interact, studying P-ODN internalization mechanisms, and by tracking internalized P-ODN through the cell using immunochemical and ultrastructural techniques. Chemical cross-linking studies with a biotin-labeled P-ODN (bP-ODN), revealed the existence of five major cell surface P-ODN binding protein groups ranging in size from approximately 20-143 kD. Binding to these proteins was competitively inhibited with unlabeled P-ODN, but not free biotin, suggesting specificity of the interactions. Additional experiments suggested that binding proteins likely exist as single chain structures, and that carbohydrate moieties may play a role in P-ODN binding. Uptake studies with 35S-labeled P-ODN revealed that endocytosis, mediated by a receptor-like mechanism, predominated at P-ODN concentrations < 1 microM, whereas fluid-phase endocytosis prevailed at higher concentrations. Cell fractionation and ultrastructural analysis demonstrated the presence of ODN in clathrin coated pits, and in vesicular structures consistent with endosomes and lysosomes. Labeled ODN were also found in significant amounts in the nucleus, while none was associated with ribosomes, or ribosomes associated with rough endoplasmic reticulum (ER). Since nuclear uptake was not blocked by wheat germ agglutinin or concanavalin A, a nucleoporin independent, perhaps diffusion driven, import process is suggested. These data imply that antisense DNA may exert their effect in the nucleus. They also suggest rational ways to design ODN which might increase their efficiency.