Mechanism Matters: A Taxonomy of Cell Penetrating Peptides

Abstract

The permeability barrier imposed by cellular membranes limits the access of exogenous compounds to the interior of cells. Researchers and patients alike would benefit from efficient methods for intracellular delivery of a wide range of membrane-impermeant molecules, including biochemically active small molecules, imaging agents, peptides, peptide nucleic acids, proteins, RNA, DNA, and nanoparticles. There has been a sustained effort to exploit cell penetrating peptides (CPPs) for the delivery of such useful cargoes in vitro and in vivo because of their biocompatibility, ease of synthesis, and controllable physical chemistry. Here, we discuss the many mechanisms by which CPPs can function, and describe a taxonomy of mechanisms that could be help organize future efforts in the field.

Keywords: cell penetrating peptide (CPP); cytosolic delivery; membrane translocation; transient permeabilization.

Figure 1

(Key Figure). Mechanisms of known cell penetrating peptides (CPPs). A: A putative landscape for CPP mechanisms based on how they interact with cells and with cellular membranes. The X- and Y-axes represent the mode of interaction with cell membranes, and the response of the cell to the peptide, respectively. Toxicity (Z-axis) results from plasma membrane permeabilization, but not endosomal permeabilization. For this reason,, low elevation regions of the landscape (blue) are useful for CPPs. B: The taxonomy of CPP mechanism, which is a portion of the taxonomy for all membrane active peptides.

Figure 2

Frequency distributions of various physical properties 747 non-redundant cell penetrating peptides. A: Total peptide length. B: Net charge assuming pH 7.4, C-terminal amidation, and no contribution from histidine. C: Wimley-White interfacial hydrophobicity score[26] for the CPP sequences, assuming an C-terminal amide and an N-terminal amino group. Positive is unfavorable for partitioning. D: Fraction of residues that are cationic, excluding histidine. E: Helical hydrophobic moment[97]. F: Absolute value of the β-sheet hydrophobic moment calculated assuming an unbroken diad repeat motif. On each histogram, we show the values for the five representative CPPs described in the text.

Figure 3

Schematic illustration of some of the overlapping mechanisms by which a peptide may cross a membrane. The effect of peptide on the membrane is also indicated. A: Spontaneous translocation of monomeric peptide that occurs without membrane disruption. B: Translocation that requires some peptide self-assembly, but not membrane disruption. C: Translocation that requires significant peptide aggregation and minor membrane disruption. D: Translocation that requires peptide self-assembly and moderate (i.e. selective or short-lived) membrane disruption. E: Translocation associated with peptide self-assembly into a structure that drives non-selective or long-lived membrane permeabilization. In this scenario, even large polar molecules pass through the membrane. This mechanism will be acutely toxic if it occurs in plasma membranes, but not if it is confined to endosomal membranes. These illustrations represent snapshots of states in a continuum of mechanisms. The arrows show some of the possible transitions between mechanisms that could arise from many experimental factors, including peptide sequence and physical; chemistry, local peptide concentration, and anionic lipid content.

Figure 4

Schematic illustration of some of the various mechanisms by which a CPP and attached cargo may be internalized into a cell. The fate of an unattached small molecule and macromolecule are also shown. A: Spontaneous membrane translocation across the plasma membrane, which occurs without peptide self-assembly or membrane disruption (See Fig. 5B & E). B: Transient plasma membrane permeabilization. C: Endocytosis of membrane bound peptide-cargo complex, along with unattached small and large molecule cargoes. D: Endosomal membrane lysis, or large scale disruption, releases the CPP-cargo conjugate and all co-encapsulated cargoes. E: Translocation across the endosomal membrane delivers CPP and attached cargo, but not co-encapsulated cargo. F: Degradation or recycling of CPP and all cargoes will occur rapidly if the other mechanisms do not enable delivery to the cytosol (See Fig. 5A&C). The mechanisms depicted are not mutually exclusive; they can happen concurrently.

Figure 5

Example confocal microscopy images of fates of dye-labeled CPPs in Chinese hamster ovary cells under various conditions. Here we show behaviors of a classical, highly cationic CPP, Arg₉, and a spontaneous membrane translocating peptide (SMTP) TP2. Both. Both peptides, which were described earlier, are labeled with the red dye tetramethylrhodamine (TAMRA). In some images, the cell media contains Alexafluor488-labelled dextran (green) an aqueous phase probe that is passively entrapped in endosomes. A: Endosome entrapment. Cells incubated at 37°C for 2 hours with 2 µM Arg₉-TAMRA show punctate red intensity that is always associated with endosomes. At concentrations less than 5 µM Arg₉, delivery to the cytosol is insignificant, and the entrapment of the TAMRA cargo does not change even after 24 hours. B: Spontaneous membrane translocation. Cells incubated with 2 µM TP2-TAMRA at room temperature for as little as 10 minutes show diffuse cytosolic fluorescence, indicating plasma membrane translocation. C: Active uptake. Cells simultaneously incubated with 2µM Arg₉-TAMRA and labeled dextran at 37°C show that dextran fluorescence often overlaps with TAMRA fluorescence in intracellular organelles. This observation demonstrates that Arg₉, at low concentration, enters cells only through endocytosis. D: No uptake. Cells simultaneously incubated with 2µM Arg₉-TAMRA and labeled dextran at room temperature, which inhibits endocytosis, show that neither Arg₉ nor dextran enter cells appreciably under these conditions. However, it is clear that Arg₉ binds strongly to the plasma membrane. Above 10 µM Arg₉, on the other hand, delivery of TAMRA to the cytosol and nucleus by Arg₉ is significant[34]. E: Spontaneous membrane translocation. Incubation of cells with TP2-TAMRA and labeled dextran at room temperature show that TP2 is internalized into cells without simultaneous entry of dextran. This observation, and others[33], indicate that endocytosis is not required for TP2 internalization.