Penetratin is a positively charged cell-penetrating peptide (CPP) that has the ability to bind negatively charged membrane components, such as glycosaminoglycans and anionic lipids. Whether this primary interaction of penetratin with these cell surface components implies that the peptide will be further internalized is not clear.
Using mass spectrometry, the amount of internalized and membrane bound penetratin remaining after washings, were quantified in three different cell lines: wild type (WT), glycosaminoglycans- (GAGneg) and sialic acid-deficient (SAneg) cells. Additionally, the affinity and kinetics of the interaction of penetratin to membrane models composed of pure lipids and membrane fragments from the referred cell lines was investigated, as well as the thermodynamics of such interactions using plasmon resonance and calorimetry.
Penetratin internalized with the same efficacy in the three cell lines at 1 µM, but was better internalized at 10 µM in SAneg>WT>GAGneg. The heat released by the interaction of penetratin with these cells followed the ranking order of internalization efficiency. Penetratin had an affinity of 10 nM for WT cells and µM for SAneg and GAGneg cells and model membrane of phospholipids. The remaining membrane-bound penetratin after cells washings was similar in WT and GAGneg cells, which suggested that these binding sites relied on membrane phospholipids. The interaction of penetratin with carbohydrates was more superficial and reversible while it was stronger with phospholipids, likely because the peptide can intercalate between the fatty acid chains.
These results show that accumulation and high-affinity binding of penetratin at the cell-surface do not reflect the internalization efficacy of the peptide. Altogether, these data further support translocation (membrane phospholipids interaction) as being the internalization pathway used by penetratin at low micromolecular concentration, while endocytosis is activated at higher concentration and requires accumulation of the peptide on GAG and GAG clustering.