In general, there is not a problem using IRES plasmids with Nucleofection®, with one important caveat. The levels of expressed protein for the first and second genes will not be identical, and this can create problems with analysis and interpretation (see Wong, et al.; Gene Ther. 2002 Mar; 9(5):337-44). As a result, the true transfection efficiency of the plasmid can be underrepresented due to the lower expression level of GFP. As long as this is understood and accounted for during analysis, there is no problem using IRES plasmids with Nucleofection®. 

With IRES plasmids, the promoter drives the expression of two genes, the cloned gene of interest (the upstream gene), and a reporter gene, usually encoding a GFP protein (the downstream gene).
The mRNA expressed from an IRES plasmid is a bicistronic message, meaning that both genes are present on the same mRNA molecule. Thus, equal amounts of the message encoding each gene are present in the mRNA population. However, the efficiency of initiation of translation of the two genes differs significantly. Ribosome binding to the initiation region of the upstream gene is very efficient, while the IRES allows ribosome binding and translation initation for the downstream gene only at a sigificantly lower level. Since the downstream gene is usually GFP, the expression level of the GFP reporter will be lower than would be normally seen when compared to plasmids without an IRES-sequence. Thus the percentage of cells expressing detectable GFP fluorescence is underrepresenting the transfection efficiency, due to the lower expression level of GFP in these cells.

This reduction of GFP expression is primarily independent on the presence or absence of an ORF upstream of the IRES-sequence. An ORF containing RNA-destabilizing elements will further reduce GFP-expression. 

In this case, even if the transfection efficiencies (defined as the percentage of cells that received plasmid DNA) were identical between an IRES-devoid vector control and an IRES containing plasmid (with or without an additional ORF upstream of the IRES), one might incorrectly conclude that the transfection did not work, since there were fewer green cells. This discrepancy is especially evident in cells generally showing low expression levels or with plamids containing weak promoters connected to an IRES and GFP. Equal amounts of the protein of interest and GFP are only achieved with a fusion of both proteins, but potentially leading to impairment of function.