Development Of Hammerhead Ribozymes As An Agent For Gene Therapy In Osteogenesis Imperfecta: Allele-Specific Suppression Of Mutant Type I Collagen mRNA.

Joan C. Marini, Paul Dawson and Antonella Forlino
Heritable Disorders Branch, NICHD, NIH, USA

Purpose: For dominant negative disorders, such as Osteogenesis Imperfecta, a gene replacement approach to gene therapy will not be effective. We are developing a suppression approach to OI gene therapy, aiming to selectively inactivate the mutant type I collagen mRNA while leaving expression from the normal allele intact. At full effectiveness, a mutant allele would be converted to a functionally null allele. This approach should mimic the molecular situation in type I OI and modulate disease severity.

Methods: To study the effectiveness and specificity of hammerhead ribozymes in OI fibroblasts, we selected cells carrying a mutation which generates a ribozyme cleavage site, and alters a restriction enzyme site. At the protein level, the G85V substitution creates a pepsin sensitive site. Active and inactive ribozyme templates were cloned into expression vectors. Stably transfected cells were selected. Quantitative competitive RT-PCR analyses, normalized to b-actin expression levels, were used to quantitate absolute levels of normal and mutant collagen mRNA and ribozymes.

To study ribozyme effectiveness in a mouse model, we have generated expression constructs specific for the murine ribozyme cleavage region. In vitro effectiveness and specificity have been demonstrated.

Results: Hammerhead ribozymes with an active catalytic core specifically decreased (50-60%) the level of mutant a1(I) collagen mRNA, while normal a1(I) mRNA levels were comparable to untransfected control. Collagen mRNA levels in cells transfected with inactive ribozyme were comparable to untransfected cells. Ribozyme expression persisted through high cell passage numbers. In clonal cell lines, the level of ribozyme expression correlated with extent of mutant mRNA reduction. In cells expressing ribozyme, we demonstrated a decrease in the proportion of collagen triple helices containing mutant chain. Stable expression of active ribozyme did not affect cell growth rates.

For ribozyme studies in the murine Brtl IV system, in vitro cleavage experiments were done using ribozymes with 13 nt binding arms specific for the region flanking the GUC ribozyme cleavage site introduced into the murine a1 (I) Exon 23. Synthetic normal and mutant target templates were generated from subcloned murine cDNA fragments. Complete digestion of mutant template was obtained, there was no cleavage of normal template.

Conclusions: Ribozymes cause specific suppression of an endogenous mutant type I collagen mRNA in fibroblasts from an OI proband. These results are encouraging as a starting point for in vivo applications. We have developed a knock-in mouse model with non-lethal OI (Brtl IV) in which a ribozyme cleavage site has been introduced by site-directed mutagenesis. For proof-of-principle experiments, we plan to by-pass vector delivery approaches and deliver ribozymes to the Brtl mouse by mating with a transgenic mouse expressing a specifically- targeted ribozyme.

Reference: Proceedings of the 7th International Conference on Osteogenesis Imperfecta. Montreal, Canada, 1999.