W e are investigating the use of antisense oligodeoxynucleotides to selectively suppress expression of the mutant type I collagen allele in osteogenesis imperfecta (OI). In this report, we target a human collagen mutation in its natural cellular context. We used cultured fibroblasts from a case of type IV OI, in which the mutant alpha 2(I) allele produces mRNA with exon 16 deleted due to a point mutation in the splice donor site. Lipid-mediated transfection was used to deliver antisense, sense and missense phosphorothioates targeted to both the abnormal mRNA exon 15/17 junction and the nuclear level point mutation. Significant suppression of the mutant protein chain and mRNA was achieved with antisense oligonucleotide to both mRNA and nuclear levels. Mutant protein was suppressed to 44-47% and mutant alpha 2(I) mRNA to 37-43% of their levels in control cells, indicating decreased mRNA as the basis for suppression. Selectivity of mutant allele suppression was better with an mRNA target: suppression was sequence specific and normal mRNA was expressed at 79% of its level in untreated cells. With a nuclear target, significant suppression of mutant mRNA occurred not only with antisense and sense, but also with missense oligonucleotide, which suppressed mutant mRNA to 60% of its level in untreated cells. We also investigated the time course of suppression of protein and mRNA in response to a 4 h transfection of antisense oligonucleotide. From 24-72 h after transfection, mutant protein was suppressed to approximately 50% of its untreated level and suppression of mutant message was significantly greater than that of normal message. The suppression achieved in these studies is insufficient for clinical intervention, but our results provide support for further development of antisense therapy as an approach to the treatment of dominant negative disorders.