Cell-Penetrating Peptides: Why Uptake Does Not Prove Delivery

• Cationic (positively charged): TAT (from HIV), and polyarginines like R8 and R9. The positive charge helps them stick to the negatively charged outer membrane.
• Amphipathic: penetratin (from a fruit-fly protein), transportan, and MPG. These have both water-loving and fat-loving faces, which helps them insert into membranes.
• Anionic and specialized: p28 is an unusual, negatively charged example.
• Chemistry can change behavior: cyclizing the ring, adding a fat tail (lipidation), or attaching a polymer coat (PEGylation) all shift how a CPP acts. And the cargo itself can mask or change the peptide's features, so a CPP that works bare may behave differently once it's loaded.

CPPs have been used to deliver peptides, proteins, siRNA, and plasmid DNA in the lab.

Energy-dependent uptake (endocytosis). The cell wraps the peptide in a bubble of membrane and pulls it inside. There are several flavors — clathrin-mediated, caveolae-mediated, and macropinocytosis — and the peptide can switch between them as the dose changes. The problem: whatever comes in this way starts out trapped inside that bubble (an endosome).

Direct translocation. Here the peptide crosses the membrane without the cell swallowing it, through models with names like the "carpet," "pore," and "inverted micelle," plus an "arginine-phosphate hopping" idea for the positively charged ones.

A cautionary tale from the field's history: some early reports of dramatic direct entry turned out to be artifacts of how cells were fixed before imaging. When researchers looked at living cells instead, a lot of that apparent entry disappeared. It's a good reminder to be skeptical of a clean story.

• Arginine over lysine. Arginine's guanidinium group forms stronger, more useful bonds with the membrane's phosphate groups than lysine's amine does.
• Charge density and amphipathicity. More positive charge, and a structure that folds into a helix against the membrane, both tend to help.
• Cargo. Its size, charge, and how it's linked to the peptide all change uptake.
• The membrane itself. Lipid composition and membrane potential matter.
• Lab conditions. Temperature and concentration shift results, which is why one number from one experiment can mislead.
• A useful concept — CP50: the concentration at which half-maximal uptake happens, a way to compare peptides on an even footing (though exact values depend heavily on the assay).

A tip from me: test several concentrations, and test them with your real cargo attached — not just the bare peptide. The bare peptide's behavior often won't survive contact with the payload.

• Live-cell imaging rather than fixed cells, to avoid fixation artifacts.
• Trypsin or acid washes to strip off surface-bound peptide so it isn't counted as "inside."
• Orthogonal assays such as CAPA (a chloroalkane penetration assay) and mass spectrometry, which measure cytosolic arrival in different ways.
• Uptake inhibitors as clues only. They're blunt tools and rarely prove a single mechanism on their own.

The review In Vitro Assays: Friends or Foes of Cell-Penetrating Peptides (PMID 32630650) is a good, honest tour of how these assays can both help and mislead.

A 2025 Nanoscale review (doi:10.1039/D5NR00617A) surveys how CPPs are combined with nanocarriers, including work on getting past mucus and the blood–brain barrier.

Two clinical touchpoints, described neutrally:

• p28, an unusual anionic peptide, was tested for safety in a small Phase I trial (PMID 23449360). I'm noting only that a first-in-human safety study was done.
• Certepetide (also called CEND-1 or LSTA1), a tumor-penetrating peptide, has been studied in clinical trials, including pharmacokinetic work in people (PMID 39789733). I mention it only to show the field is moving into human studies; I'm making no claims about how well it works.

Endosomal escape strategies include pH-sensitive elements that activate in the acidic endosome, membrane-destabilizing sequences, stimuli-responsive carriers, and co-delivered "disrupting" partners. None is a solved problem.

The tradeoffs:

• Make a peptide better at breaking membranes, and it often gets more toxic.
• Add features, and you can trigger an immune response.
• Chase raw potency, and you can lose specificity for the right cells.
• And the peptide still has to survive in blood long enough to arrive.

A tip from me: a functional readout — did the cargo actually do its job inside the cell? — beats a bright image every time.

One study tested five of these peptides (Epitalon, Vilon, Thymogen, Thymalin, and Chonluten) in the human THP-1 monocyte/macrophage cell line (PMID 35408963). In that cell culture only work, the peptides modulated proliferative signaling, and some lowered TNF-α and IL-6 release after an inflammatory (LPS) challenge and reduced monocyte adhesion.

Two things to keep straight:

• This is cell culture, not animals or people. Results in a dish don't tell you what happens in a body.
• The gene-regulation mechanism is a hypothesis. It's an interesting idea with early support, not a settled fact.

For plain-language summaries, see our library entries on the THP-1 study, Vilon, and Vesugen.

Bottom line: CPPs are couriers you attach cargo to. Khavinson peptides are studied as messages in their own right. Don't blur the two.

• Rational tweaks: adjusting charge, adding arginines, or stabilizing a helix.
• Chemistry: cyclization, lipid tails, or PEG coats to improve stability or uptake.
• Responsive elements: parts that switch on only in the acidic environment of an endosome, to help with escape.
• Simulations and machine learning: molecular dynamics and predictive models can shortlist promising sequences fast.

The honest caveat: computation shortens the candidate list; it doesn't replace the experiments that tell you what actually happens in cells.

• Membrane damage: assays like LDH leakage (did the peptide poke holes and let cell contents out?) and viability tests.
• Hemolysis: does it rupture red blood cells?
• Immune response: could the sequence provoke an unwanted reaction?
• Biodistribution: where does it actually go in a living system?

When a CPP is coupled to a therapeutic cargo, the resulting conjugate is reviewed as a drug — with attention to how it's characterized, how the body handles it (pharmacokinetics), its toxicity, and its immunogenicity. Anyone heading in that direction should consult the relevant regulator early.