Promiscuity of pre-mRNA spliceosome-mediated trans splicing: a problem for gene therapy?

Trans splicing of messenger RNA has been used in experimental settings to replace mutant RNA sequences. We investigated the feasibility of utilizing trans splicing to replace a mutant RET protooncogene sequence known to inappropriately activate this tyrosine kinase receptor. We constructed a pre-trans-splicing molecule (PTM) consisting of a binding domain complementary to the target intron, the 3' splicing signal sequence (3'ss), derived from adenovirus major late transcript intron 1 and a molecular tag sequence. Accurately targeted trans splicing between the human RET exons and the PTM was demonstrated in NIH 3T3 cells cotransfected with the human RET minigene and the PTM. The efficiency of specific trans splicing was estimated to be no more than 15% in the cotransfection experiment. However, in addition to the targeted trans splicing, nontargeted trans splicing to RET exons was observed. Furthermore, the rapid amplification of 5' cDNA ends (5' RACE) analysis demonstrated that nontargeted trans splicing occurred with endogenously expressed pre-mRNAs in TT cells and that specific trans splicing to RET was a rare event. Attempts to reduce nonspecificity by the addition of a stem-loop to the trans-splicing construct designed to suppress nonspecific splicing failed to have the desired effect. These observations suggest that overexpression of a trans-splicing construct containing a 3'ss results in promiscuous trans splicing and raise significant questions about the specificity and usefulness of currently used trans-splicing approaches. In addition, these findings raise the possibility that nonspecific spliced products may be produced by a variety of gene therapy constructs.