If your goal is to cause a specific protein to be expressed transiently via an episomally maintained gene, could you just exclude all introns and the exons not associated with that transcript variant when choosing what DNA to package into your plasmid for that gene? So far as I know, the role of introns is usually in allowing alternative splicing and in modifying gene expression. But if you're only interested in one transcript variant, you wouldn't need alternative splicing to occur on the episome, and gene expression is handled in plasmid design by the selected promoters and enhancers.
For those reasons, could the introns (and exons not used for that transcript variant) be excluded? And would that affect the protein produced?