Cell-Penetrating Peptides: Drug Delivery Research

Reviewed by PeptideJournal.org Editorial Team | Updated February 2026

Most therapeutic molecules — proteins, nucleic acids, even many small drugs — cannot cross the cell membrane on their own. The membrane is a gatekeeper, and it is ruthlessly selective. For decades, this biological reality limited what researchers could deliver into living cells. Then, in 1988, two independent labs noticed something strange about the HIV-1 TAT protein: it walked right through the membrane as if it wasn't there.

That observation launched an entire field. Today, more than 4,200 unique cell-penetrating peptides (CPPs) have been cataloged in the CPPsite3 database, and researchers are testing them as delivery vehicles for everything from cancer drugs to gene therapies. No CPP-based formulation has yet won FDA approval — but several are in clinical trials, and the science is moving fast.

Here's what the research shows so far.

Table of Contents

What Are Cell-Penetrating Peptides?
A Brief History: From TAT to Thousands
How CPPs Cross the Cell Membrane
Three Classes of CPPs
What Can CPPs Deliver?
CPPs in Cancer Research
Crossing the Blood-Brain Barrier
Activatable CPPs: Adding a Targeting Switch
Clinical Trials: Where Things Stand
Challenges and Limitations
The Bottom Line
References

What Are Cell-Penetrating Peptides?

Cell-penetrating peptides are short sequences of amino acids — typically 5 to 30 residues — that can cross biological membranes and carry molecular cargo along with them. Unlike most delivery systems, CPPs do not rely on specific cell-surface receptors. They interact directly with the lipid bilayer through electrostatic and hydrophobic forces, which means they can enter virtually any cell type.

This versatility is both their greatest strength and their most significant limitation. CPPs are remarkably efficient at getting molecules inside cells. They are far less efficient at getting molecules into the right cells.

Most CPPs carry a net positive charge at physiological pH, thanks to clusters of arginine and lysine residues. That positive charge is what pulls them toward the negatively charged phospholipid headgroups in the cell membrane — the first step in a complex entry process.

A Brief History: From TAT to Thousands

The field began with two accidental discoveries:

1988: Frankel and Pabo reported that the full-length TAT protein from HIV-1 could cross cell membranes and reach the nucleus — no receptor binding required.
1991: Alain Prochiantz's group at the Ecole Normale Superieure in Paris found that the Antennapedia homeodomain protein from Drosophila could enter neuronal cells.
1994: Prochiantz's team identified the specific 16-amino-acid sequence responsible — a peptide they named penetratin (RQIKIWFQNRRMKWKK). This was the first peptide formally recognized as a CPP.
1998: Lebleu's group pinpointed the minimal TAT sequence needed for membrane crossing: just 11 amino acids (YGRKKRRQRRR, residues 47-57).

From those two parent peptides, the field exploded. Researchers designed synthetic variants — polyarginine chains, chimeric sequences, cyclic forms — each tuned for different cargo types or uptake profiles. The CPPsite database grew from 741 unique sequences in 2012 to 1,700 by 2015 and to over 4,285 by 2024, a nearly six-fold increase in just over a decade.

How CPPs Cross the Cell Membrane

There is no single answer to this question. After more than 30 years of study, researchers have identified two principal routes — and the real answer is usually "both, depending on conditions."

Direct Translocation

Some CPPs physically punch through the membrane. Positively charged arginine side-chains interact with phosphate groups on both sides of the lipid bilayer, nucleating the formation of a transient pore. The peptide diffuses along the pore surface and arrives in the cytoplasm. This process is fast, energy-independent, and does not require the cell's endocytic machinery.

Direct translocation tends to dominate at higher peptide concentrations. It has been observed with TAT, penetratin, and various polyarginine peptides, particularly in the presence of amphiphilic counteranions like pyrenebutyrate that accelerate the process.

Endocytosis

At lower concentrations, most CPPs are internalized through endocytic pathways — macropinocytosis, clathrin-mediated endocytosis, and caveolae-mediated endocytosis among them. The peptides get trapped inside endosomes, and a major research challenge is engineering their escape before lysosomal degradation destroys the cargo.

A 2024 study in Molecular Pharmaceutics by Pirhaghi et al. noted that newer research has identified a Rab14-dependent endocytic pathway — distinct from the classical Rab5/Rab7 route — that some CPPs appear to use preferentially.

What Determines the Route?

Multiple factors tip the balance: peptide concentration, sequence and structure, cargo size and hydrophobicity, cell type, membrane potential, and temperature. For large cargo molecules like proteins or nanoparticles, endocytosis predominates. For free CPPs at high concentrations, direct translocation can be the primary route.

This complexity matters because endosomal entrapment is one of the biggest practical obstacles in CPP-based drug delivery. If your cargo gets stuck in an endosome, it never reaches its intracellular target.

Three Classes of CPPs

Based on their physicochemical properties, CPPs fall into three broad categories:

Class	Charge Profile	% of Known CPPs	Key Examples
Cationic	Net positive; rich in arginine and lysine	~22%	TAT (47-57), polyarginine (R8, R9), VP22
Amphipathic	Both hydrophobic and hydrophilic domains	~44%	Penetratin, Pep-1, MPG, transportan, CADY
Hydrophobic	Predominantly apolar residues	~15%	SG3, Pep-7, pepducins, stapled peptides

Amphipathic CPPs are the largest group, accounting for over 40% of all known sequences. They adopt alpha-helical or beta-sheet conformations that let them insert into the lipid bilayer directly. Penetratin, one of the original CPPs, is classified as a secondary amphipathic peptide — it appears unstructured in solution but folds into an alpha-helix when it contacts the membrane.

Cationic CPPs like TAT and polyarginine rely almost entirely on charge-based interactions. They are generally simpler to synthesize and have been the most widely studied in clinical contexts. However, amphipathic CPPs like penetratin and MAP tend to show higher internalization efficiency in head-to-head comparisons.

A newer, still-small class of anionic CPPs challenges the old assumption that positive charge is mandatory. Peptides like SAP(E), which replaces basic residues with glutamic acid, can penetrate membranes effectively — suggesting that amphipathicity may matter more than net charge alone.

What Can CPPs Deliver?

The cargo list is remarkably broad:

Small-molecule drugs (doxorubicin, temozolomide, cyclosporine)
Proteins and peptides (p53-derived peptides, frataxin, antigens)
Nucleic acids (siRNA, plasmid DNA, antisense oligonucleotides, peptide nucleic acids)
Nanoparticles (liposomes, polymeric nanoparticles, quantum dots)
Imaging agents (fluorophores for surgical visualization)

CPPs attach to cargo through two general strategies. Covalent conjugation links the CPP directly to its payload via disulfide bonds, amide bonds, or cleavable linkers — a tight connection that keeps the cargo nearby through membrane transit. Non-covalent complexation relies on electrostatic or hydrophobic interactions to form stable CPP-cargo assemblies, an approach particularly useful for nucleic acid delivery.

For researchers working on peptide bioavailability, CPPs represent a fundamentally different approach to the delivery problem — rather than protecting a peptide drug from degradation in the gut or bloodstream, CPPs escort it directly through the cell membrane.

CPPs in Cancer Research

Cancer is the most active area of CPP drug delivery research, and for good reason. Tumors are notoriously difficult to penetrate with conventional therapeutics, and many promising anticancer molecules — tumor suppressor proteins, siRNAs, immune modulators — cannot cross the cell membrane without help.

p28: A First-in-Class CPP Cancer Therapy

One of the most advanced CPP-based cancer candidates is p28, a 28-amino-acid peptide derived from azurin, a protein produced by Pseudomonas aeruginosa. p28 preferentially enters cancer cells, where it binds to both wild-type and mutant p53, blocking COP1-mediated ubiquitination and preventing p53 degradation. The result: p53 levels rise, and the cell cycle arrests at G2/M.

In a phase I trial involving 15 adults with advanced solid tumors, p28 produced no dose-limiting toxicities, no significant adverse events, and no immune response. Trends toward reduced tumor biomarkers appeared across all dosing cohorts. A separate phase I trial in pediatric patients with recurrent CNS tumors confirmed the safety profile. The FDA granted p28 both Orphan Drug and Rare Pediatric Disease designations.

p28 also shows antiangiogenic activity through a separate, non-p53 mechanism: it inhibits VEGFR2 and FGFR1 kinases in endothelial cells, cutting off the blood supply tumors need to grow. Researchers at Rush University Medical Center are studying p28 in combination with temozolomide for glioblastoma, where it appears to amplify the chemotherapy drug's cytotoxic activity.

CPPs in Cancer Immunotherapy

A growing body of work connects CPPs to cancer immunotherapy research. Because activated immune cells often struggle to infiltrate solid tumors, CPPs can help by shuttling immune-modulatory payloads past the membrane.

In 2024, Shi's group developed cell-penetrating peptide-induced chimeric conjugates (cp-PCCs) that degrade the enzyme DHHC3, which palmitoylates PD-L1 and anchors it to the membrane. By stripping PD-L1 from the tumor cell surface, cp-PCCs effectively re-expose tumors to immune attack — a novel angle on checkpoint blockade.

CPP-antigen conjugates are also gaining traction in peptide vaccine research. Backlund et al. conjugated CPPs with 10 different tumor-associated antigens and neoantigens. Nine out of ten showed stronger T-cell priming compared to unconjugated controls, suggesting CPPs can boost vaccine immunogenicity by improving antigen uptake into dendritic cells.

Crossing the Blood-Brain Barrier

The blood-brain barrier (BBB) blocks over 98% of small-molecule drugs and essentially all large biologics from reaching brain tissue. CPPs are among the few delivery technologies that can breach it.

TAT-conjugated cargo has crossed the BBB in preclinical models of Alzheimer's disease, binding to heparan sulfate glycosaminoglycans in cerebral amyloid deposits. Penetratin and polyarginine R8 have been tested for ocular drug delivery, another application where biological barriers block most therapeutics.

A 2024 review in Molecular Pharmaceutics by Pirhaghi et al. identified CPPs as "promising therapeutics and drug-delivery systems for neurodegenerative diseases," noting that the TAT sequence remains the most widely studied BBB-penetrating CPP, though newer amphipathic designs are closing the gap.

For the broader picture on non-invasive peptide administration, see our review of oral peptide delivery technology.

Activatable CPPs: Adding a Targeting Switch

The biggest knock against classical CPPs is that they enter every cell indiscriminately. Activatable cell-penetrating peptides (ACPPs) address this by adding a built-in off switch.

An ACPP has three parts:

A polycationic CPP domain (typically nine D-arginine residues)
A polyanionic masking domain (nine D-glutamic acid residues) that neutralizes the CPP's positive charge
A cleavable linker between the two domains, designed to be cut by a specific protease

In healthy tissue, the masking domain keeps the CPP inactive. In the tumor microenvironment, where proteases like MMP-2 and MMP-9 are overexpressed, the linker gets cleaved. The masking peptide falls away, the CPP activates, and it penetrates nearby cells along with its cargo.

Roger Tsien's group at UC San Diego pioneered this concept, and it has since been validated in multiple xenograft models and a transgenic breast cancer mouse model (MMTV-PyMT). Accumulation concentrates at the tumor-stromal interface, with spatial resolution under 50 micrometers.

ACPPs have been conjugated with doxorubicin for tumor-targeted chemotherapy and with fluorescent dyes for surgical imaging. AVB-620, a protease-cleavable ACPP conjugated with Cy5 and Cy7 fluorophores, was developed for real-time tumor visualization during surgery and has been tested in clinical settings.

pH-responsive ACPPs offer an alternative trigger. Histidine-glutamic acid dipeptide repeats fused to CPPs activate at mildly acidic pH (below 6.5) — characteristic of the tumor microenvironment — while remaining inactive at normal physiological pH. Dual-stimulus systems combining MMP cleavage with pH activation are now under investigation.

This kind of targeted design connects to broader work on peptide-drug conjugates and cyclic peptide engineering, where researchers use structural modifications to improve selectivity.

Clinical Trials: Where Things Stand

Despite thousands of publications and dozens of preclinical successes, no CPP-based drug has received FDA approval. But several programs have reached the clinic:

Candidate	CPP Used	Indication	Stage	Status
p28 (NSC745104)	Azurin-derived p28	Advanced solid tumors, pediatric CNS tumors	Phase I (completed)	FDA Orphan Drug designation
Nomlabofusp	TAT	Friedreich's ataxia	Phase I (ongoing)	Showing increased frataxin levels (NCT06681766)
PGN-EDO51	EDO platform CPP	Duchenne muscular dystrophy	Phase I	Discontinued for DMD; company pivoted to DM1
Delcasertib (KAI-9803)	TAT (47-57)	Acute myocardial infarction	Phase II/III (completed)	Failed primary endpoint in PROTECTION AMI trial
AVB-620	ACPP (protease-cleavable)	Surgical tumor imaging	Clinical evaluation	Active
Polyarginine-cyclosporine	Polyarginine	Psoriasis (topical)	Phase II	Completed by CellGate Inc.

The results so far tell a mixed story. p28 showed a clean safety profile and preliminary anticancer activity. Nomlabofusp, a TAT-frataxin fusion, is producing measurable increases in frataxin levels in Friedreich's ataxia patients — a rare genetic disease with no approved treatments targeting the underlying deficiency. But PGN-EDO51 failed to hit dystrophin targets in DMD, and delcasertib showed no benefit over placebo in a well-powered 1,010-patient trial for heart attack.

The pattern suggests that CPP technology works well for some applications (intracellular protein replacement, cancer cell targeting) and less well for others (acute cardiac emergencies where intravenous distribution may dilute the drug before it reaches ischemic tissue).

Challenges and Limitations

Several problems remain unsolved:

Lack of specificity. Classical CPPs enter all cell types, raising the risk of off-target effects. Activatable and tumor-homing CPPs are partial solutions, but none have reached late-stage trials.
Serum instability. Most CPPs are rapidly degraded by proteases in the bloodstream. D-amino acid substitutions, cyclization, and PEGylation improve stability, but each modification can alter uptake behavior.
Endosomal entrapment. CPPs internalized by endocytosis often get stuck in endosomes and are degraded in lysosomes before reaching their intracellular targets. Endosomal escape remains perhaps the single largest engineering challenge.
Immunogenicity concerns. While p28 showed no immune response in clinical trials, longer CPP-cargo conjugates could potentially trigger antibody production with repeated dosing.
Scale and manufacturing. Synthesizing CPP-drug conjugates at pharmaceutical scale adds complexity and cost compared to small-molecule drugs.

The Bottom Line

Cell-penetrating peptides represent one of the most versatile drug delivery platforms in peptide research. They can shuttle nearly any molecular cargo — from small drugs to large proteins to nanoparticles — across the cell membrane, a feat that most delivery technologies cannot match. The CPPsite3 database now catalogs over 4,200 unique sequences, and clinical trials are testing CPP-based therapies for conditions ranging from cancer to Friedreich's ataxia to muscular dystrophy.

The technology is not yet mature. No FDA-approved CPP drug exists, and fundamental challenges around targeting specificity and endosomal escape remain. But the pace of research — particularly in activatable CPPs, nanoparticle conjugation, and cancer immunotherapy — suggests this field is closer to a clinical breakthrough than it has ever been.

If you're following peptide drug delivery research, CPPs are worth watching closely.

References

Frankel AD, Pabo CO. Cellular uptake of the Tat protein from human immunodeficiency virus. Cell. 1988;55(6):1189-1193. PubMed
Derossi D, Joliot AH, Chassaing G, Prochiantz A. The third helix of the Antennapedia homeodomain translocates through biological membranes. J Biol Chem. 1994;269(14):10444-10450. PubMed
Vivès E, Brodin P, Lebleu B. A truncated HIV-1 Tat protein basic domain rapidly translocates through the plasma membrane and accumulates in the cell nucleus. J Biol Chem. 1997;272(25):16010-16017. PubMed
Gori A, et al. Cell penetrating peptides: classification, mechanisms, methods of study, and applications. ChemMedChem. 2023;18(17):e202300236. Wiley
Agrawal P, et al. CPPsite 2.0: a repository of experimentally validated cell-penetrating peptides. Nucleic Acids Res. 2016;44(D1):D1098-D1103. PMC
Kumar V, et al. CPPsite3: an updated large repository of experimentally validated cell-penetrating peptides. Drug Discov Today. 2025. PubMed
Pirhaghi M, et al. Cell-penetrating peptides: promising therapeutics and drug-delivery systems for neurodegenerative diseases. Mol Pharm. 2024;21(5):2097-2117. PubMed
Du R, et al. Applications of cell penetrating peptide-based drug delivery system in immunotherapy. Front Immunol. 2025;16:1540192. PMC
Yang H, et al. Advances and prospects of cell-penetrating peptides in tumor immunotherapy. Sci Rep. 2025;15:3674. Nature
Warso MA, et al. A first-in-class, first-in-human, phase I trial of p28, a non-HDM2-mediated peptide inhibitor of p53 ubiquitination in patients with advanced solid tumours. Br J Cancer. 2013;108(5):1061-1070. Nature
Langer HT, et al. Phase I trial of p28 (NSC745104) in pediatric patients with recurrent or progressive central nervous system tumors. Pediatr Blood Cancer. 2016;63(8):1464-1470. PMC
Bates SE, et al. Tumor-targeting cell-penetrating peptide, p28, for glioblastoma imaging and therapy. Front Oncol. 2022;12:940001. Frontiers
Lincoff AM, et al. Inhibition of delta-protein kinase C by delcasertib as an adjunct to primary percutaneous coronary intervention for acute anterior ST-segment elevation myocardial infarction. Eur Heart J. 2014;35(37):2516-2523. PubMed
Olson ES, et al. Activatable cell penetrating peptides linked to nanoparticles as dual probes for in vivo fluorescence and MR imaging of proteases. Proc Natl Acad Sci USA. 2010;107(9):4311-4316. PNAS
Wu J, et al. Cell-penetrating peptides for transmucosal delivery of proteins. J Control Release. 2024;366:864-878. PubMed
Gostaviceanu A, et al. Cancer-targeting applications of cell-penetrating peptides. Int J Mol Sci. 2025;26(1):2. MDPI
Dissanayake S, et al. Recent developments in anticancer drug delivery using cell penetrating and tumor targeting peptides. J Control Release. 2017;250:62-76. PubMed