Peptide phosphorylation in the design of a vector for intracellular drug delivery based on the cell-penetrating peptide penetratin

The fruitful research in molecular biology over the course of the last 50 years has revolutionized our comprehension of the processes that happen inside a cell. Understanding of intracellular regulatory pathways and the protein synthesis by translation of gene transcripts allow the development of powerful methods to fight a broad spectrum of human health issues that are traditionally very hard to treat. An important requisite, however, is access of the agent to the cell interior. In this respect, the plasma membrane of eukaryotic cells poses an efficient barrier for many potential agents and methods that allow an agent’s transit across it are therefore in high demand.
Promising vectors in this regard are cell-penetrating peptides (CPPs): short polycationic peptides that were shown to be capable of transporting compounds of interest inside eukaryotic cells. To date, the mechanism of their translocation is still under much debate. Also, their application as drug vector is potentially delicate because some CPPs showed a concentration dependent toxicity for cells.
Penetratin (pen-Antp) is among the best studied CPPs. Interestingly, it does not show translocation across model membranes such as unilamellar vesicles (LUVs). A more hydrophobic pen-Antp mutant called pen-2AL, however, does show permeation of model membranes. We were mainly interested in the potential modification of pen-2AL with a phosphorylated tyrosine (named pen-A(pY)L) in order to create a CPP which is only active after dephosphorylation e.g. by a protein phosphatase. In doing so, we aimed for an inducible CPP that would only be activated by cellular phosphatases.
In this thesis we discuss various aspects of the design of the pen-A(pY)L peptide and the investigation of its effect on both model and biological membranes. The thesis is divided into 5 chapters which deal with distinct parts of the peptide design, the investigation of its effect on different model membranes. In chapter 1 we present the methodical basis for the investigation of the interaction of CPPs with model membrane systems. Both the creation of lipid model membranes and their thermodynamical characterization in presence of CPPs are described and supplemented with minute protocols for every method. Exemplary data show that pen-2AL destabilizes model membranes in a detergent-like manner whereas pen-Antp does not. Furthermore, we show that the use of multivalent fluorescent dyes can introduce a critical measurement bias upon interaction with CPPs.
In chapter 2 we discuss the design of the pen-A(pY)L peptide and show its effect on LUVs and CHO cells by means of a permeation (leakage) assay and confocal microscopy, respectively. Our leakage data suggest that the phosphorylated pen-A(pY)L does not permeate LUVs at low micromolar concentrations whereas the unphosphorylated pen-AYL shows strong permeation at these conditions. We could also successfully activate pen-A(pY)L by dephosphorylation as demonstrated by inducible dye leakage from LUVs after addition of a phosphatase. Lastly, CHO cells show uptake of TAMRA-labeled pen-A(pY)L after incubation with 20 µM of the CPP. The peptide seems also to successfully reach the cytosol without damaging the cells.
We then investigated the effect of the used medium during incubation of CHO cells with either peptide as discussed in chapter 3. Using plain phosphate buffered saline (PBS) as incubation medium led to strong detachment of the cells during incubation with low peptide concentration. In contrast, this could not be observed after incubation with either peptide dissolved in DMEM/F12. Furthermore, the images made with differential interference contrast suggest that incubation of CHO cells with 20 µM TAMRA-pen-AYL is toxic for the cells.
In chapter 4 we describe the observed propensity of pen-A(pY)L for gel formation at low millimolar concentration. The peptide gets compared with its unphosphorylated counterpart in order to assess their tendency to aggregate under various conditions in order to identify the potential molecular interactions that promote gel formation. We thereby find that it is probably hydrophobic attraction between the peptide that leads to the clustering of pen-A(pY)L which may have been introduced by the tyrosine’s phenol group. However, the bulk of peptide seems to remain dissolved at low micromolar concentrations i.e. at the relevant concentration for a potential application of the CPP.
Lastly, as a consequence of the earlier observed bias due to the interaction of multivalent fluorescent dyes with certain CPPs we propose in chapter 5 an alternative molecular probe for the detection of membrane permeation by CPPs. Instead of the dequenching of fluorescent dye we exploit the concentration dependence of the electron paramagnetic resonance (EPR) signal of spin labels to detect their leakage out of LUVs. We can show that the EPR signal is, in contrary to fluorescent dyes, independent of the presence of CPPs. However, we also experienced difficulties in enclosing sufficient spin label concentrations into LUVs. Nevertheless, we value electron-spin labels as promising option for such an assay.