Fluorescent dyes alter intracellular targeting and function of cell-penetrating tetrapeptides

Fluorescent labels are commonly used to investigate the mechanisms of cellular uptake and intracellular distribution of cell-penetrating peptides. However, labels such as fluorescein and rhodamine are relatively large and very lipophilic and may significantly alter physicochemical properties of small peptides. To minimize the impact of the fluorescent probe on a tetrapeptide, we substituted one of the amino acids (Lys4) in a tetrapeptide ([Dmt1]DALDA, Dmt-D-Arg-Phe-Lys-NH2 where Dmt = 2',6'-dimethyltyrosine) with two different fluorescent amino acids (beta-dansyl-L-alpha,beta-diaminopropionic acid (dnsDap4) or beta-anthraniloyl-L-alpha,beta-diaminopropionic acid (atnDap4)). Initial studies with confocal laser scanning microscopy (CLSM) showed very different localization patterns for the two fluorescent analogs, with [Dmt1,atnDap4]DALDA showing mitochondrial localization and [Dmt1,dnsDap4]DALDA showing diffuse cytoplasmic localization. Studies with isolated mouse liver mitochondria suggested that [Dmt1,dnsDap4]DALDA targeted the mitochondrial matrix resulting in mitochondrial depolarization, opening of the permeability transition pore, mitochondrial swelling, and rapid release of the peptide into the cytoplasm. In contrast, [Dmt1,atnDap4]DALDA was retained in the inner mitochondrial membrane and did not induce mitochondrial swelling. Furthermore, [Dmt1,atnDap4]DALDA protected mitochondria against Ca2+-induced swelling. Importantly, the unlabeled parent peptide [Dmt1]DALDA behaved like [Dmt1,atnDap4]DALDA and was mitoprotective. These findings suggest that experimental results obtained with fluorescent labels must be interpreted with caution, and the use of multiple fluorophores, together with confirmation using the original or radiolabeled molecule, is recommended.