Peptide-guided and ligand-conjugated carriers both address active targeting, but they differ in molecular design and translational feasibility. Peptides offer single-chain specificity, faster synthesis, and renal clearance, while larger ligand conjugates (antibodies, sugars, aptamers) offer multivalent avidity but more complexity and longer half-life. The decision between them will be dictated by receptor biology, manufacturing scale-up, and desired therapeutic window, rather than any intrinsic "affinity superiority". Below, we describe when each modality is beneficial, and why passive accumulation alone will not cut it for next-generation RNA therapeutics.
Active targeting is also required as passive distribution based on vascular leakiness and organ tropism is insufficient for the increasingly high degree of specificity required by the next generation of RNA drugs. In the lack of ligand-directed uptake, mRNA or siRNA often has significant accumulation in liver and spleen, which may result in under-treatment of extra-hepatic malignancies, fibrotic tissues or neuronal populations. Active targeting can enhance the cellular-level payload localization, enabling lower systemic doses, less immune stimulation, and measurable dosimetry, which can correlate the imaging signal to therapeutic response. This is especially important for RNA therapeutics, whose effectiveness is often outweighed by their small therapeutic index and by the need for repeated dosing.
Fig. 1 mRNA delivery: Challenges, goals, widely explored polymer-based systems, and intended applications of this technology.1,5
Active targeting is necessary since passive targeting, i.e. utilizing vascular leakiness and organ tropism, is insufficient for the precise requirements of today's RNA drugs. Without ligand-mediated uptake, extra-hepatic levels of mRNA or siRNA are low, resulting in insufficient treatment of extra-hepatic malignancies, fibrotic tissue or neuronal populations. The active engagement of the payload to the cellular target also allows for lower systemically administered doses with less immune activation, as well as providing a measurable dosimetry with an imaging signal that is directly proportional to the therapeutic effect, something that is important for RNA therapeutics whose efficacy is countered by low therapeutic windows and the need for multiple doses. Passive systems cannot distinguish between cancer and normal cells in the same organ, leading to the non-selective delivery of mRNA or siRNA to all cells that have increased permeation. The subsequent off-target protein translation or gene silencing can lead to toxicity, unwanted immune activation or off-target changes in cell phenotype. Ligand-mediated cell-level specificity is therefore required to recognize surface molecules that are uniquely or significantly over-expressed on diseased cells.
Conjugation of peptides, antibodies, or sugars to RNA delivery vehicles initiates clathrin- or caveolae-mediated endocytosis and sequesters the payload in the cytosol of target receptor-positive cells. Peptide ligands can be optimized for nanomolar binding affinity and fast on-rates to ensure cellular uptake within the narrow time window between injection and renal excretion. In contrast to passive enrichment, this process is saturable and measurable, allowing the generation of dose–response curves to directly link receptor occupancy to therapeutic effect. Active targeting also shifts the whole-body pharmacokinetics of the nanocarrier: antibodies against the transferrin receptor can mediate transcytosis across the blood–brain barrier, while integrin-targeting peptides preferentially deposit RNA within angiogenic tumor vasculature. The biodistribution of these carriers is largely conserved across payload types (mRNA, siRNA, CRISPR RNP) because the same ligand can be used for conjugation, streamlining regulatory toxicology. Finally, the target receptor density can be imaged in the clinic prior to treatment to stratify patients and personalize dosing, a feat impossible with passive systems.
Table 1 Conceptual comparison of peptide- vs ligand-conjugate targeting for RNA therapeutics
| Feature | Peptide-directed systems | Ligand-conjugate systems |
| Molecular weight | Low–intermediate | Variable (small molecule to antibody) |
| Chemistry flexibility | High (solid-phase, cyclisation) | Medium (linker optimisation critical) |
| Immunogenic potential | Generally low | Antibody or protein ligands may activate immunity |
| Receptor scope | Any receptor with known peptide binder | Any receptor with validated ligand |
| Manufacturing step | Peptide-carrier coupling before particle assembly | Often post-insertion onto pre-formed particles |
| Typical stability concern | Proteolysis | Ligand desorption or masking |
Targeting peptides and ligand-conjugates are two of the modular approaches used to impart cell-type specificity to RNA drugs. Peptides provide single-chain precision and specificity, fast solid-phase synthesis, and renal clearance properties suitable for acute dosing and theranostic use. Ligand-conjugates, in which a targeting small molecule, sugar, or aptamer is attached, on the other hand, can offer different pharmacokinetic properties, simpler chemistry, but potential for bulk or immunogenicity. Although these strategies both seek to supplant passive accumulation with active, receptor-mediated uptake, they have key differences in ease of design, regulatory path, and scalability. The decision to use either strategy is based on the receptor biology, size of the therapeutic payload, and desired therapeutic window rather than a general advantage of affinity for either class.
Targeting peptides are small (typically 5–30 residues), linear or cyclic, designed amino-acid sequences that bind cell-surface receptors over-expressed on diseased tissue. Discovered via phage display, in-silico screening, or rational design, targeting peptides are often further engineered to optimize nanomolar affinity, protease resistance and low immunogenicity. Their small size allows for fast tumor penetration, easy solid-phase synthesis, and orthogonal installation of chelators, fluorophores, or RNA-loading handles without disrupting receptor binding. Since peptides are biodegradable and renally cleared, they can be repeatedly administered without risk of accumulation, making them particularly well-suited to chronic RNA regimens which often require multiple dosing cycles.
Ligand-conjugates covalently link natural or synthetic receptor ligands (e.g. galactose, folate, or aptamers) directly to the RNA sequence or to carrier polymers. They are known to engage well-characterized uptake pathways (asialoglycoprotein receptor, folate receptor, etc.), they are often smaller than peptides and thus may cause less steric hindrance during RNA hybridization. They also have established safety profiles (many of these are metabolites or vitamins), and their biodistribution is well known to regulatory agencies. But they lack capacity for affinity maturation or engineering for multiple functions. Also, their fixed chemistry is not always compatible with the multiple conjugation chemistries that are needed to covalently link functional groups to mRNA, self-amplifying RNA, or CRISPR guide RNAs. Selection of a targeted delivery technology therefore largely depends on whether the target receptor justifies peptide level customization or can be addressed sufficiently with a simple metabolite based ligand.
Peptides possess an unrivaled degree of precision, tunability, and payload versatility that remains difficult to achieve in larger ligand platforms. Their compact size enables solid-phase synthesis with single-amino-acid resolution, and thus iterative optimization of affinity, charge, hydrophobicity and protease resistance without the need to redesign an entire scaffold. At this molecular level of granularity, the same backbone can be evolved from a micromolar binder into a picomolar ligand, while also introducing pH-responsive or endosomolytic motifs at specific residues. This enables developers to tune the tradeoff between targeting efficiency, intracellular release and safety within a single molecular entity, accelerating pre-clinical development and minimizing manufacturing risk.
Fig. 2 Peptide-functionalized nanoparticles (NPs) for targeted therapeutic delivery across the blood–brain barrier (BBB).2,5
Each residue can be individually changed, cyclized or N-methylated to adjust receptor fit, serum stability, or linker geometry. Phage display or AI-based generators search spaces of millions of sequences in weeks. The resulting peptides can have switched specificity between receptor subtypes, or newly acquired dual-binding affinity without added molecular weight. Such precision is not possible with small-molecule ligands, whose scaffolds are fixed after synthesis. Affinity can be tuned from micromolar to picomolar range by adjusting charge distribution, adding non-natural amino acids or dimerizing the peptide. Specificity can be adjusted by negative-selection screening to undesirable off-target homologues, resulting in ligands that reject point-mutant variants of the same receptor. Such fine-tuning allows RNA to be delivered only where intended, limiting dose-limiting exposure to liver or kidney.
The same peptide scaffold can be used to conjugate 19-mer siRNA, 2'-O-methyl antisense oligos or 2-kb mRNA by simply changing the linkage chemistry (disulfide for siRNA, amide for ASO, or click-chemistry for mRNA) without redesigning the recognition domain. In this way, one targeting ligand can support many therapeutic programs reducing regulatory qualification and cost of goods. Orthogonal handles (azide-alkyne, thiol-maleimide, or hydrazone bonds) allow site-specific attachment at 5' phosphate, 3' terminus, or internal nucleobase. Spacers of variable lengths (PEG, β-alanine, rigid diketopiperazine) prevent steric interference, allowing RNA hybridization or cap-dependent translation to remain unaffected after conjugation.
Head-to-tail cyclization or lysine branching grafts two separate recognition motifs onto a single peptide. The resulting OR-logic gate broadens tumor coverage and minimizes escape through antigen loss. AI-assisted design that maintains affinity of both arms despite close proximity produces constructs that silence two oncogenes simultaneously without increased dose. Dendrimeric peptides presenting 4-8 copies of the same ligand exploit avidity to reach picomolar effective affinity on cells with modest receptor density, while remaining monovalent in circulation where density is low. This self-limiting behavior concentrates RNA in malignant tissue without raising systemic exposure, which is not possible with monovalent small-molecule ligands.
Conversely, ligand-conjugates simplify the bench to bedside transition, since they capitalize on clinically validated receptors, whose expression maps, internalization and intracellular trafficking are known. The safety dossier of the conjugated moiety is also largely pre-existent since the ligand is typically a metabolite, sugar or vitamin analogue. In addition, conjugation chemistry is limited to a short heterobifunctional spacer, and does not require multi-step solid phase cycles needed for peptide optimization. The result is a single molecular entity with defined stoichiometry that can be more easily characterized, has simpler regulatory CMC packages and is easier to scale up under current good manufacturing practice.
Natural monosaccharides or vitamins have evolved sub-nanomolar residence times for their cognate lectins or nutrient transporters. In the context of multivalent ligands, the avidity when presented in a trivalent or hexavalent array collapses the apparent dissociation constant into the picomolar range without further medicinal chemistry. The binding event is pH-sensitive and therefore the ligand dissociates from its receptor inside acidified early endosomes, leaving receptor available for recycling and preventing competitive blockade from subsequent doses. Such quantified thermodynamic profiles enable relatively straightforward modelling of dose–occupancy relationships and therefore spare extensive in-vivo titration campaigns. Receptors such as the asialoglycoprotein or mannose lectins traffic through clathrin-coated pits whose adaptor proteins, endosomal acidification schedules and lysosomal escape propensities are textbook knowledge. Ligand-conjugates therefore inherit a predefined intracellular itinerary: rapid internalization within minutes, residence in mildly acidic early endosomes and either recycling to the surface or onward movement to late compartments. This choreography can be exploited to synchronize RNA release with endosomal maturation, maximizing cytosolic exposure and minimizing lysosomal degradation.
A trivalent GalNAc cluster linked by a short PEGylated linker is chemically monodisperse, with a single molecular weight, a single stereochemistry and a single net charge. The absence of batch-to-batch heterogeneity found in peptide-directed platforms (e.g. batch-dependent oxidation, incomplete cyclisation or diastereomeric contamination that can all impact PK) means that the conjugate can be fully characterized by routine HPLC-MS, NMR and elemental analysis, thus offering unambiguous identity tests that meet both European and US pharmacopoeial monographs, without the need for peptide mapping or chiral purity assays. The non-immunogenicity of the ligand itself and the linker design, which ensures the linker fragments into innocuous byproducts, mean that standard ICH Q3A/B impurity thresholds apply directly, with no requirement for additional immunogenicity panels or ADAb assays. Additionally, the absence of a secondary structure means that no circular dichroism or FT-IR confirmation is needed, and release testing is therefore accelerated with the possibility of real-time stability assessment under accelerated temperature and humidity conditions.
When choosing between peptides or ligand-conjugates, formulators must balance the competing requirements of sequence freedom versus frozen chemistry, extensive versus specific targeting, and prolonged circulation versus sustained receptor blockade. Peptides offer a mutable framework that can be systematically mutated, cyclized, and pseudouridinated, but each design tweak has implications for metabolism, immunogenicity, and price. Ligand-conjugates sacrifice design flexibility for a predefined scaffold whose receptor binding, internalization pathway, and degradation are already optimized in the clinically approved transport protein, thus shortening development but risking loss of target if receptor levels change.
The peptide backbone is a software layer in solid-phase synthesis. Residues can be exchanged, D-enantiomers can be inserted or hydrocarbon staples can be installed within the same production run. This can shift receptor affinity by several orders of magnitude, while the chemistry of the carrier remains unaltered. The resulting latitude allows for on-demand pivoting from a high-affinity tumor address to a low-affinity vascular zip-code by editing a three-residue epitope, a manoeuvre that is impossible once a small-molecule ligand is welded to the RNA vehicle. Conversely, the sugar, vitamin or metabolite ligand arrives with a thermodynamic signature forged by evolutionary selection; medicinal chemists can lengthen spacers or swap anomeric linkers, but the core binding motif is immutable. This constancy translates into predictable dose–occupancy curves and simplified regulatory specification, but it also means that receptor down-regulation or polymorphic variation cannot be countered by iterative lig editing, forcing developers to accept reduced margins of safety or to reformulate the entire carrier.
A linear peptide, conversely, can be fused to two cell-penetrating tags specific to a given receptor and separated by a protease-sensitive linker. This contraption, an implementation of a Boolean "OR" gate, then has the ability to accumulate in any tissue expressing either one of the two markers. Alternatively, branched dendrons presenting two different loops can promote avidity-dependent clustering of hetero-receptors, extending tropism while avoiding synthetic multiplication. While such a multiplexing feature can be helpful in non-homogeneous tumors with patchy antigen expression, it can also conjure the ghost of on-target/off-tissue toxicity if both targeted receptors fortuitously also densely populate healthy organs. Ligand-conjugates, in comparison, are molecularly monogamous: one cluster of galactose is bound to the asialoglycoprotein receptor, one folate group is taken up by the folate transporter, and no formulation tweaking will make the payload go elsewhere. Such constancy is beneficial in terms of therapeutic window and ease of pharmacokinetic modelling, but also dooms the therapy to failure should the target receptor be downregulated or lost through clonal evolution or transcriptional drift, a problem not shared by peptide cocktails that opportunistically pick the next best target.
Free peptides are susceptible to cleavage by aminopeptidases, chymotryptic-like serine proteases and oxidative modification near the surface of the NP. All of these processes can lead to rapid desorption of the targeting peptide within minutes of intravenous administration. Chemical modifications such as cyclisation, N-methylation or retro-inversion can extend the half-life of the targeting peptide. However, every additional chemical modification increases the hydrophobicity of the peptide which can lead to increased complement activation and/or sequestration by macrophages, perpetuating a constant trade-off between metabolic stability and immune evasion. Ligands such as sugars or vitamins are also more resistant to enzymatic cleavage and remain on the NP surface for days, maintaining receptor-mediated uptake, but may lead to prolonged receptor saturation, dampening repeat dosing. Additionally, the gradual hydrolytic cleavage of linker moieties can lead to the formation of membrane-impermeable adducts that accumulate in renal tubules. This can subtly change the osmotic load, with potential impact on long-term tolerability that is often not considered for rapidly eliminated peptide fragments.
Table 2 Conceptual trade-offs between targeting peptides and ligand-conjugates for RNA therapeutics
| Attribute | Targeting peptide | Ligand-conjugate |
| Structural latitude | High (sequence editing) | Low (core motif fixed) |
| Receptor repertoire | Multi-receptor capable | Single-receptor focused |
| Circulation lifetime | Tunable, but prone to proteolysis | Intrinsically long, risk of saturation |
| Immunogenic risk | Edits may raise new epitopes | Low (ligand is nutrient) |
| Regulatory predictability | Evolving | High (precedent exists) |
This decision-making logic is contingent upon which axis of uncertainty the program can afford to leave unaddressed. If the receptor repertoire is very poorly annotated, or if there is any cause to suspect phenotypic drift, the editability of peptides becomes an insurance policy. On the other hand, if the target is a well-characterized membrane protein whose density and trafficking have been quantified across patient cohorts, then the plug-and-play predictability of a metabolite or sugar ligand shortens both pre-clinical and regulatory timelines. Payload chemistry, linker lability and endosomal escape kinetics must be overlaid on this map to ensure that the ligand class of choice doesn't introduce a new source of uncertainty just as the target biology becomes deterministic.
Peptides also are more forgiving of low-density receptors as multivalent display or cyclisation can artificially enhance avidity, whereas monovalent ligand-conjugates require higher surface abundance to reach significant occupancy. Internalization kinetics are also distinct: peptide carriers can be designed to release at pH 6.0, in step with early-endosomal maturation, whereas sugar ligands remain bound to their receptor until it is degraded in the lysosome, fixing a long intracellular residency which may or may not be required by the payload activation timing. If the receptor's physiological ligand is a circulating micromolar level compound (folate, galactose or transferrin), then the exogenous conjugate must either be able to out-compete it by being multivalent, or resign to being partially blocked. As peptides are not physiological ligands, there are rarely any soluble competitors, but serum albumin binding can still displace a peptide conjugate via hydrophobic adsorption, a competition which can be adjusted by modulating loop polarity or staple position.
For bulky mRNA or self-amplifying RNA payloads, only flexible, long spacers will do. These can be added at a large number of sites along a peptide backbone without perturbing folding. Small-molecule ligands, in contrast, have only one or two anchor points for conjugation. Steric clash with the RNA can diminish binding as well as translation. For smaller payloads, such as 2'-O-methyl ASOs, the small size of a ligand-conjugate mitigates the risk of hybridization interference and eases analytical characterization. Where the therapeutic index mandates dual-receptor AND logic or conditional activation in the acidic tumor micro-environment, only peptides can be rationally designed to incorporate pH-sensitive histidine clusters or cleavable second arms. Ligand-conjugates remain the default for "one-receptor, one-payload" situations where the main goal is to redirect liver-avoiding biodistribution without additional biological bells and whistles.
Choosing between targeting peptides and ligand-conjugates is not a binary decision driven by preference, but a strategic choice shaped by target biology, payload requirements, and delivery objectives. Our peptide design services support RNA targeting strategies by providing rational peptide engineering and comparative feasibility analysis, enabling teams to make informed, data-driven decisions early in development. Rather than positioning peptides as universal replacements for ligand-conjugates, we focus on identifying when peptide-based targeting offers a true functional advantage—and when alternative approaches may be more appropriate.
Target-Driven Sequence Engineering: Effective RNA targeting begins with target biology. Our target-driven sequence engineering approach designs peptides around receptor expression patterns, internalization behavior, and tissue specificity rather than relying on generic targeting motifs. Peptide sequences are engineered to support receptor-mediated uptake compatible with RNA payload requirements, ensuring that targeting interactions facilitate productive intracellular delivery rather than surface binding alone. This biology-first approach improves the likelihood that peptide-mediated targeting translates into functional RNA activity.
Selectivity-Focused Optimization: In RNA therapeutics, improved uptake is not inherently beneficial if it comes at the cost of selectivity. Our optimization strategy prioritizes selectivity-focused design, tuning peptide affinity and physicochemical properties to minimize non-specific uptake driven by charge or hydrophobicity. By aligning peptide binding strength with receptor density and internalization capacity, we help ensure that delivery is driven by active targeting rather than passive accumulation—an important distinction when comparing peptides to ligand-conjugates with fixed binding profiles.
Design Trade-Off Analysis: Targeting peptides and ligand-conjugates each offer distinct advantages and limitations. Our comparative feasibility assessment includes a structured design trade-off analysis that evaluates flexibility, predictability, stability, and scalability in the context of the intended RNA modality. This analysis helps clarify whether the tunability of peptides outweighs the established behavior of ligand-conjugates for a given target, or whether ligand-based approaches may provide a more reliable delivery route.
Data-Driven Go/No-Go Guidance: Rather than advancing designs based on theoretical advantages, we provide data-driven go/no-go guidance grounded in feasibility, risk, and alignment with program goals. Early evaluation of targeting performance, selectivity boundaries, and payload compatibility informs clear recommendations on whether peptide-based targeting is likely to deliver incremental value. This approach reduces late-stage pivoting between targeting strategies and helps teams commit resources to the most appropriate delivery platform.
If you are evaluating whether to pursue targeting peptides or ligand-conjugates for your RNA therapeutics program, an early technical discussion can clarify which approach best aligns with your target biology and delivery objectives. Discuss the right targeting strategy for your RNA therapeutics program with our team to assess feasibility, compare design trade-offs, and define a targeting path that balances selectivity, robustness, and functional delivery.
References
