Targeting peptides provide a non-invasive and receptor-mediated strategy to shuttle therapeutics across the blood-brain barrier (BBB) through exogenous transcytosis mechanisms. The sub-nanometre dimensions and solid-phase tractability of peptides enable rational design of motifs that target transferrin, LDL, or glutamate receptors which are highly expressed on the cerebro-microvasculature and could allow siRNA, mRNA, or CRISPR RNPs to selectively enter neurons and glioma cells without disruptive osmotic mannitol or intracranial catheters. However, the physicochemical properties that promote BBB permeation—low mass and rapid renal clearance—also limit residence time, while receptor saturation and competition with endogenous ligands could also reduce transcytosis efficiency, requiring a trade-off between affinity and circulation half-life.
The BBB is still the major bottleneck of CNS therapeutics as its tight junctions and efflux transporters exclude ～98 % of small molecules and nearly all biologics. To date, the only solution has been to administer very high systemic doses, which can lead to unwanted peripheral toxicity without any certainty of brain exposure. Active targeting through peptides or ligand-conjugates is therefore an attractive physiological alternative that could allow a preferential accumulation of drugs in the brain parenchyma and sparing of off-organs, but the key to its success is to select receptors that cycle fast and transcytose efficiently without risk of saturation under therapeutic load.
Fig. 1 Structure of the neurovascular unit. 1,5
Claudin and occludin strands weld the endothelial cells together to create a blockade of paracellular flux. Peptides are therefore required to be transported by transcytosis; simply making a drug small and below 1 nm does not guarantee that it will pass. Peptide receptors therefore become the selection criteria, with only those ligands which can initiate vesicular shuttling capable of jumping the junctional firewall without promoting junctional opening and ensuing neuroinflammation. Nutrient entry is regulated by carrier-mediated and receptor-mediated systems. Peptides can bind to these transporters only when they resemble their endogenous ligands, but high-affinity binding saturates these pathways and leads to down-regulation. In addition, active efflux pumps such as P-gp eject many peptide conjugates, and this means that linker design should release cargo after transcytosis and not during transit.
In general, most peptides transcytose inefficiently across cerebral capillaries; the peptide is able to bind to the luminal membrane receptor but is unable to dissociate from the receptor abluminally to cross the endothelial cell, so the peptide returns to the blood without ever entering the brain parenchyma. Cleavable linkers or pH-sensitive bonds that cause release once inside the endothelial cell have been used to overcome this "sticking" phenomenon. However, in addition to increasing synthetic complexity, these must be designed to avoid premature cleavage in plasma. Agents that increase permeability, such as hyperosmotic mannitol or focused ultrasound, may also allow the entry of toxins and immune cells that can cause seizures or edema. Peptides that transiently affect the tight junctions (such as cadherin peptides) must be shown to be reversible and not release pro-inflammatory cytokines. Targeting peptides are a more specific strategy. However, this approach has a narrow safety margin, as over-stimulation of the transporters through over-attachment to the receptor can down-regulate these transporters that are important for brain homeostasis and cause a secondary pathology even if the attached drug gets into the brain.
Table 1 BBB Challenge vs Peptide Design Response
| Barrier Mechanism | Peptide-Level Solution | Translational Risk |
| Tight junctions | Receptor-mediated transcytosis | Receptor saturation |
| Efflux pumps | Cleavable linker post-transit | Premature cleavage |
| Enzymatic degradation | D-amino acid backbone | Immunogenicity |
| Heterogeneous BBB | Multi-receptor targeting | Complexity |
BBB-targeting peptides are designed as molecular 'ferrymen' to first open the BBB gate and then deliver the therapeutic to the target cell. As a result, their amino acid sequence encodes for two sets of instructions: an endothelial internalizing receptor-binding epitope, and a pH-responsive release module to release the RNA or protein payload prior to lysosomal fusion. The physicochemical structure of the peptide needs to thus be simultaneously soluble in plasma and able to insert in a lipid bilayer in the acidic vesicles. Peptides in these applications have an architecture that balances hydrophilic receptor recognition with a brief hydrophobic domain - an amphipathic design that replaces both antibody and lipidoid in a single, biodegradable chain.
Fig. 2 Overview of trans-BBB delivery of cargos using BBB-penetrating and cell-penetrating peptides.2,5
Peptides hijack native transcytosis pathways by mimicking transferrin, insulin or LDL motifs. They are anchored to clathrin-rich plaques in the luminal endothelial surface. Cyclisation or D-amino-acid insertion protect the ligand from luminal proteases, to ensure that the epitope is not degraded during the 8-12 second transit through the cerebral micro-vessels. Following abluminal release the same peptide can present a second, neuron- or microglia-specific loop that was sterically occluded by the carrier during transcytosis. This "zip-code" switch ensures that passage across the BBB is immediately followed by re-targeting to parenchymal cells, rather than perivascular astrocytes, avoiding sequestration in the neurovascular unit.
The most exploited portals include the low-density lipoprotein receptor-related protein-1 (LRP1), the transferrin receptor (TfR), and the insulin receptor (IR). Each has different trafficking kinetics (e.g. TfR recycles quickly enabling multiple rounds of transcytosis whereas LRP1 accommodates larger payloads such as liposomes) and peptide ligands are chosen for nanomolar affinity and fast on-rates to compete with endogenous ligands without saturating the transporter (to avoid compromising barrier integrity). Transcytosis is saturable and dependent on receptor density, so excessive doses of the peptide can lead to down-regulation of the transporter, paradoxically resulting in reduced brain uptake over time. In addition, peptide conjugates may be actively pumped out again by efflux pumps such as P-gp. Cleavable linkers that release the cargo after transcytosis can prevent the peptide itself from accumulating in the brain endothelium and depleting the transporter, so as to maintain transport capacity for repeated dosing.
Table 2 BBB Peptide Design Objectives vs Translational Hurdles
| Design Objective | Peptide-Level Solution | Key Translational Risk |
| Cross endothelium | High-affinity RMT ligand | Receptor saturation |
| Detach post-transit | pH-cleavable linker | Premature plasma cleavage |
| Target neurons | Second parenchymal motif | Dual-specificity synthesis |
| Avoid efflux | Cargo release post-exit | P-gp substrate liability |
Targeting peptides are required to simultaneously meet a number of criteria: the peptide needs to have the appropriate affinity, length and stability, as well as correct release properties. They need to bind to endothelial cell receptors with adequate avidity to initiate transcytosis, yet release the associated cargo intracellularly, prior to degradation or pumping back into the blood. The peptide design needs to address a number of tradeoffs including the need for a high-affinity ligand against potential losses in the off-rate and the need to consider both protease stability of the targeting ligand backbone, as well as pore size limitations of the BBB's tight junctions. Each of these parameters must be fine-tuned: the peptide needs to be of appropriate affinity to avoid receptor saturation, it needs to be small enough to diffuse across the endothelial glycocalyx, and a cleavable linker needs to be placed in an appropriate location to promote release of the cargo inside the brain parenchyma rather than inside the endothelial cells.
Nanomolar affinity is important to ensure the engagement of receptors during their limited exposure in the capillaries. However, a binding that is too strong may be difficult to reverse inside the endothelial cells, preventing the abluminal release of the peptide-cargo complex. This is why some designs include an intentional off-rate or a pH-labile linker that allows a weakening of the receptor interaction once the vesicle is acidified. This causes the cargo to be deposited into the brain tissue, while the peptide is recycled back to the blood. Peptide conjugates can be a target for efflux pumps such as P-glycoprotein which could transport them back into the lumen. To avoid this, the drug is linked by a disulfide or hydrazone bond that is cleaved in the reducing or acidic environment of early endosomes, thus releasing the drug before the transporter recognizes it. This way, the peptide is cleared through the kidney while the therapeutic molecule is delivered into the brain and the endothelial cells do not accumulate large amounts of the therapeutic, thus preserving transporter availability for subsequent doses.
Hydrodynamic radius must be <～5 nm to penetrate endothelial glycocalyx and tight-junction pores under physiological flow conditions without experiencing shear stress. Reduced flexibility upon cyclization or introduction of D-amino acids also allows the peptide to resist entanglement with basement-membrane proteoglycans while remaining resistant to proteases. Rigid scaffolds additionally help to pre-orientate the receptor-binding epitope in a conformation complementary to that of the transporter pocket, a geometry less readily achieved by flexible linear peptides. Backbone modifications (e.g. N-methylation, thioether staples) can provide additional aminopeptidase resistance to the naturally abundant enzymes in cerebral capillaries. These are introduced external to the binding interface to retain affinity but significantly increase half-life from minutes to hours to remain intact during vesicular trafficking from luminal to abluminal membrane.
Serum proteases and brain-capillary aminopeptidases rapidly cleave linear peptides within minutes of exposure. Use of non-natural amino acids, β-peptides or cyclization will produce a protease-resistant scaffold which is stable in systemic circulation but still cleavable in neurons, in the presence of intracellular reducing conditions, which confers an additional layer of safety post-delivery. The peptide must remain intact long enough to interact with receptors throughout the cerebral vascular bed yet short enough to avoid accumulation in healthy brain regions. This can be accomplished by fusing a high-affinity binding domain to a cleavable, hydrophilic C-terminal tag that will rapidly clear the peptide via the kidneys once the cargo is released, which provides an automatic off-switch that limits off-target exposure.
Transport of peptides across the BBB is limited by at least three factors: low transcytosis efficiency, peripheral sink effects, and sequestration after transcytosis. The main causes for these limitations are the limited number, saturability and low affinity of the receptors involved in transcytosis as well as their ubiquitous presence in peripheral organs. Moreover, the endosomal sorting pathways involved in the uptake of the peptide into the cell are also involved in intracellular trafficking to the lysosome. For these reasons, even a construct with high affinity for the receptors of interest may only lead to modest brain penetration. Thus, there is a delicate balance between affinity and receptor conservation, circulation time and degradation/cleavage rate, escape from the endosomes and cell specificity, which makes this strategy only partially translated into clinics, even though it may be shown to be effective by pre-clinical imaging data.
Endocytosis by cerebral capillary endothelium is inefficient under basal conditions; only a small proportion of surface-bound peptides are internalized and an even smaller fraction completes the transcytosis cycle. Moreover, due to the scarcity of these transcytotic portals, when high-affinity ligands are used, they can easily saturate the few transport sites and, as a result, paradoxically cause decreased brain uptake at high doses. As a result, affinities must be adjusted to an intermediate level that is high enough to bind on the luminal surface but also low enough to dissociate from the receptor once at the abluminal surface. This is often achieved with pH-sensitive linkers that cause receptor dissociation once the vesicle is acidified. Systemic doses needed to bind a sufficient number of receptors are often higher than the renal clearance limit, causing loss of the peptide in the kidney and, therefore, accumulation in other off-target organs. Split dosing or receptor "priming" (for example, LRP1 transient up-regulation) can increase transcytosis without increasing overall exposure, but they complicate the clinical regimen and can lead to receptor down-regulation upon repeated dosing.
In addition to being highly expressed on BBB, many BBB receptors (e.g. transferrin receptor) are also abundant on liver, lung and bone marrow. As a result, peptides are sequestered in these "sink" organs, which depletes the available circulating pool for brain uptake. Shielding strategies - including zwitterionic flanking sequences or transient albumin binding - can minimize peripheral adhesion. But such masking must not obscure the receptor-binding epitope or otherwise impair transcytosis kinetics. Even when BBB transcytosis is achieved, uptake in competing organs (spleen, kidney) can result in brain-to-plasma ratios falling below the therapeutic threshold. This is addressed by using cleavable linkers that lose peripheral-binding domains after systemic distribution, leaving only intact, BBB-targeted constructs at the cerebral capillaries, and allowing sink organs to clear the cleaved fragments.
Peptides that transcytose can still become trapped in the perivascular space if their molecular size, charge, or binding to basement-membrane heparan sulfate prevents further diffusion into parenchyma. This results in a false-positive imaging readout (brain-specific high radioactivity but little pharmacology) since the encoded RNA or delivered drug cannot reach target neurons or glia. Addition of protease-sensitive spacers or removal of surface charge facilitates tissue diffusion, but these changes can also cause premature systemic cleavage, so these must be balanced. Cargo may also be sequestered in astrocytic end-feet or trafficked to lysosomes after transcytosis into neurons, both of which quench activity. Thus, pH-responsive or reductile linkers that release their cargo in the neuronal cytosol are also required. These conditional chemistries must remain intact during the acidic, oxidative transit across the endothelium, so linker optimization that tolerates both BBB transit and intracellular activation is often iterative.
BBB-targeting peptides have a delicate balance: every improvement in affinity, stability or multifunctionality is usually accompanied by a commensurate penalty to transport efficiency, safety or manufacturability. The same receptor that enables brain access is limited, saturable and shared with systemic organs; the same chemistry that stabilizes the peptide may trap it after transcytosis; and the same dual-function architecture that bestows neuronal targeting capacity can overwhelm synthetic feasibility. Success thus relies on a tradeoff of accepting sub-optimal performance on individual steps in exchange for a global window where crossing, release and brain-cell engagement all occur at useful efficiencies—a systems-level negotiation often lost when developers optimize single metrics in isolation.
Ultra-high-affinity binders (<1 nM) saturate the finite number of luminal receptors in minutes. This fast rate of binding traps the intact complex into a recycling endocytic loop that returns the complex to blood, preventing completion of transcytosis. The "affinity barrier" therefore limits brain exposure at high doses and leads to a non-monotonic dose-response relationship where increasing the concentration of drug in systemic circulation paradoxically reduces cerebral uptake. For this reason, designers tune dissociation constants to an intermediate range that is still sufficient for saturating luminal receptors but permissive for dissociation on the abluminal side. Many attempts include pH-labile linkers that are expected to reduce the receptor binding strength after vesicle acidification. This triggers peptide dissociation and cargo release into brain parenchyma while the peptide continues to recycle back to blood. Moderate affinity (10-100 nM) increases the chance that both binding arms of a bivalent construct do not engage the same receptor at the same time, which should reduce the chances of capillary trapping and enable forward trafficking. A moderate off-rate also enables repeated rounds of transcytosis per peptide molecule, multiplying the total amount of brain exposure without the need to increase systemic dose. The necessary kinetic balance is very difficult to achieve with rigid small molecule ligands but can be optimized by directed evolution of peptide libraries or AI-assisted affinity maturation that specifically penalizes ultra-tight binding.
Systemic stability is often obtained by D-amino acids, cyclization or backbone N-methylation, but these modifications may also "lock" the peptide into a conformation with insufficient conformational flexibility to undergo receptor-induced endocytosis. Further, excessive peptide stability in plasma can result in extensive peripheral receptor (liver, lung) binding and sequestration, leading to a "sink" effect and a reduced circulating pool of peptide available for transport into the brain. As a result, peptide designers often make use of "conditional" chemistries, e.g., serum-stable thioethers that can be reduced to disulfides in the endosome to preserve the integrity of the peptide in circulation while allowing the necessary conformational flexibility for transcytosis to occur. In addition, hydrophobic patches on the peptide surface that stabilize α-helices also result in non-specific adhesion to basement-membrane proteoglycans and a consequent reduction in the free fraction of peptide that reaches cerebral capillaries. An optimal balance of hydrophobic residues with zwitterionic or sulfated residues can mask this off-target binding and retain the helical confirmation necessary for receptor binding. The peptide should also be soluble at the millimolar concentrations at which it is reacted in the conjugation chemistry without forming amyloid-like fibrils, which are known to activate immune surveillance. This is a formulation consideration that is often not addressed during the initial peptide design stage.
Peptides targeting solely endothelial receptors, e.g. the transferrin receptor, LRP1 (Monovalent), represent the easiest synthetic route and the simplest regulatory path; however, the "task" of parenchymal cell (brain-cell) targeting is left solely to the payload, if it's any. A minimalist strategy can be adequate for small, diffusible payloads, but will not work well for larger cargos such as RNAs or proteins. Large cargos will typically require a secondary "targeting" motif to facilitate binding to the desired neuron or astrocyte. Single-target constructs are not able to make up for receptor down-regulation in diseased states, and will have variable brain exposure in a patient cohort. Dual-function peptides, on the other hand, include both an endothelial "shuttle" motif as well as a parenchymal "homing" sequence, on the same scaffold (Dual-Targeted). This forms a cascade of target binding; first cross the barrier, then bind to parenchymal neurons or glia. This architecture can deliver more cargo into specific cell types within the brain, but doubles the synthetic complexity and may double the molecular weight (larger constructs are more likely to be filtered out, or to trigger immune responses). Further, the two motifs must be arranged such that they do not sterically hinder each other; and they must either cleave off, or some other way reconfigure after transcytosis so as not to re-bind to endothelial receptors once on the abluminal side, i.e. must have spatial and temporal choreography that is challenging to design a priori without an iterative in-vivo testing.
BBB-targeting peptides are best exploited when three biological conditions coincide: (i) the small/packagable size of the payload can withstand vesicular trafficking; (ii) the receptor is available at useful densities on luminal endothelium; and (iii) a favorable therapeutic index emerges after cargo release abluminally. The small size (<5 nm) and modular chemistry of peptides has enabled developers to fuse a BBB-shuttle motif (e.g., angiopep-2 for LRP1) with a secondary neuronal ligand in the same chain, imparting a sequential "cross-then-home" logic unattainable for bulkier antibodies. Efficacy therefore depends on tuning peptide kinetics (high on-rate, intermediate off-rate) to the receptor's recycling rate: slow enough to escape lysosomal sorting but fast enough to avoid capillary saturation, all the while ensuring the linker chemistry is resistant to serum proteases yet cleavable intraneuronally to release functional RNA or drug.
Ideal cargos are therefore siRNA duplexes, antisense oligos, or small proteins (<30 kDa) that are active following transient exposure to the acidic environment of endosomes. Payloads are condensed or covalently attached by peptides, which unlike liposomes, are below the ～5 nm pore limit of the BBB and do not experience steric hindrance. Chemical modifications (2'-O-methyl RNA, phosphorothioate backbones, or disulfide-locked peptides) stabilize the construct against capillary aminopeptidases but are still cleavable by cytosolic nucleases or reductases so that potency is maintained from plasma to cytosol. Therapeutics that act in the cytosol (siRNA, CRISPR guides) or at the membrane (receptor antagonists) are preferred to those requiring nuclear entry since nuclear pores are rare in mature neurons and endosomal escape is already the rate-limiting step. Peptides can be modified with histidine-rich tails that protonate in late endosomes, destabilizing the vesicle and releasing RNA directly into the cytosol—an escape mechanism that cannot be easily added to larger antibody conjugates.
In cases where the therapeutic index requires fast, whole-brain delivery (as with emergency enzyme replacement therapy), focused ultrasound or osmotic mannitol may have advantages over peptide shuttles by opening transient paracellular gaps that allow large proteins to pass. The same can be said for non-specifically membrane-adsorbing cell-penetrating peptides, if a more rapid delivery of bulky mRNA vaccines is required than is provided by receptor-limited transcytosis, and the attendant transient barrier disruption is acceptable. In fibrotic pathologies or diseases with a dense microvasculature, BBB receptors may be occluded by thickened basement membranes or down-regulated by hypoxia. In such cases, receptor-independent methods of delivery (lipidized small molecules that diffuse through the endothelium, or macrophage-mediated carriage that physically shuttles nanoparticles across the vessel wall) often result in deeper parenchymal penetration than peptide conjugates, whose effectiveness is dependent on luminal epitope availability.
Drug delivery across the blood-brain barrier (BBB) is constrained not only by permeability, but by the efficiency and fidelity of transport mechanisms. Many BBB-targeting programs fail because peptide designs prioritize binding strength over transport behavior, leading to endothelial trapping or peripheral sequestration rather than effective brain delivery. Our targeting peptide services support BBB delivery programs through transport-aware design and feasibility-driven evaluation, focusing on whether peptides can enable functional delivery beyond the BBB, not just barrier interaction.
Transport-Aware Sequence Engineering: BBB transport via receptor-mediated transcytosis requires a delicate balance between affinity and release. Our transport-aware sequence engineering approach designs peptides specifically to support transcytosis rather than prolonged endothelial binding. Peptide sequences are optimized to engage BBB transport receptors with sufficient affinity to initiate uptake, while avoiding excessive binding that can result in receptor trapping and reduced transport efficiency. Structural features such as size, flexibility, and charge distribution are tuned to maintain compatibility with transcytotic pathways and minimize degradation during endothelial transit. This approach ensures that BBB interaction promotes forward transport into the brain rather than accumulation at the barrier interface.
Functional Delivery Optimization: Crossing the BBB is only the first step; effective CNS delivery requires that payloads reach relevant brain compartments in a functional form. Our design strategy therefore extends beyond BBB penetration to optimize downstream delivery behavior. Peptide architectures are evaluated for their ability to release or distribute payloads after transcytosis, avoiding designs that successfully cross the BBB but remain endosomally trapped or fail to reach target brain cells. Optimization focuses on aligning peptide behavior with the intended mechanism of action, whether neuronal, glial, or region-specific delivery is required.
Target and Pathway Evaluation: Not all BBB receptors or transport pathways are equally suitable for peptide-mediated delivery. Our feasibility assessment begins with a target and pathway evaluation that examines receptor expression, transport capacity, recycling behavior, and compatibility with peptide architectures. This analysis helps determine whether a given pathway can realistically support meaningful brain exposure at achievable doses, or whether biological constraints are likely to limit delivery efficiency regardless of peptide optimization.
Early Risk Identification: BBB delivery programs often encounter late-stage failures due to risks that are identifiable early, such as peripheral sink effects, low transcytosis rates, or decoupling between BBB crossing and brain tissue distribution. We focus on early risk identification to flag these limitations before extensive development resources are committed. By understanding feasibility boundaries upfront, teams can adjust design strategies, refine expectations, or explore alternative delivery approaches while development remains flexible.
If your BBB delivery strategy achieves receptor binding but limited brain exposure—or if CNS activity does not scale with dose—an early technical discussion can help clarify whether peptide design or transport pathway selection is the limiting factor. Discuss your BBB targeting challenge with our scientists to evaluate feasibility, identify transport bottlenecks, and define a peptide design strategy aligned with functional CNS delivery goals.
References
