Instability of the peptide ligand is the single most frequent reason why peptide–drug conjugates (PDCs) never reach the clinic, whereas antibody–drug conjugates (ADCs) more often fail because of linker or payload issues. Linear peptides are degraded within minutes by serum proteases, lose their three-dimensional binding conformation, and are cleared renally before reaching the tumour, translating into sub-therapeutic payload levels and off-target toxicity.
Fig. 1 Schematic structure of receptor targeting drug conjugates. 1,5
Unstable peptides lead to undesirable consequences in all subsequent drug actions: rapid degradation reduces half-life and tumor penetration and produces inactive metabolites that may compete for the target; unfolding reduces affinity and leads to a mixed biodistribution; and lastly, rapid renal clearance will increase the systemic dose and decrease the therapeutic index.
Chemical stability is the preservation of peptide backbone and linker integrity; biological stability requires preservation of the binding conformation and target affinity during circulation. A chemically intact but biologically inactive peptide conveys neither payload nor efficacy. Circulatory half-life is the time the conjugate persists in plasma, and functional targeting integrity is the time the peptide is capable of receptor recognition. As peptides are small, they are rapidly filtered by the kidney, so functional half-life is often shorter than chemical half-life.
Premature release of the payload results in systemic toxicity without tumor kill. Premature degradation results in pharmacologically inactive fragments that dilute the dose and saturate receptors. Loss of targeting specificity redistributes the cytotoxic moiety to healthy tissues. The resulting clearance from fragmented peptides that is faster than the intact drug, which shifts exposure from tumor to kidney and liver, increasing off-target toxicity. Altered PK invalidates pre-clinical dosing models, leading to failed efficacy endpoints.
ADCs have the added bulk of the cytotoxin; the peptide needs to remain unfolded to avoid steric clash with the antibody Fc. PDCs are smaller, allowing the peptide to be in a tighter turn, but increasing its protease susceptibility. ADCs will be exposed to endosomal proteases after FcRn recycling, and PDCs will be exposed to serum proteases and renal peptidases immediately after injection. ADCs are administered intravenously at low dose, while PDCs may be subcutaneously or orally administered, and may be exposed to gastric proteases as well as first-pass hepatic clearance.
Table 1 Stability Failure Modes in PDC vs ADC.
| Failure Mode | PDC Typical Cause | ADC Typical Cause | Mitigation Shared |
| Premature cleavage | Serum proteases | Endosomal cathepsins | Cyclisation, D-amino acids |
| Loss of affinity | Renal oxidation | Endosomal acidification | Methionine → norleucine |
| PK shift | Rapid renal filtration | FcRn saturation | PEGylation, albumin fusion |
Both PDCs and ADCs are exposed to both enzymatic and chemical stresses that are powerful enough to destroy the conjugate before it even has the opportunity to reach its target. Peptides are much more exposed than whole antibodies, which are protected by an Fc domain. Peptide stability, both in terms of efficacy and safety, is therefore important: a lack of stability results in premature payload release and systemic toxicity. The challenges can be grouped into four main areas: proteolytic digestion in vivo, chemical degradation during formulation and storage, conformational distortion upon payload coupling and linker miscleavage.
Fig. 2 Schematic representation of the modular components of an Antibody-Drug Conjugate (ADC). 2,5
Physiological substrates of serum proteases and peptidases are characterized by a linear peptide backbone. Following administration, the targeting function is quickly lost, often within minutes, as these enzymes hydrolyze exposed amide bonds. Internal sequences are cleaved by endoproteases such as cathepsins and matrix metalloproteinases, while exopeptidases sequentially remove N- and C-terminal residues, progressively degrading the binding epitope. Linear motifs are especially prone to proteolysis because their flexible conformation exposes many cleavage sites; although cyclic or stapled analogs are more resistant, they are still vulnerable to degradation by proteases with broad specificity. In addition to removing the homing moiety, this rapid degradation results in the formation of pharmacologically inactive fragments that compete with the intact conjugate for receptor binding and thus "dilute" therapeutic efficacy and promote off-target payload distribution. The short half-life in circulation also requires frequent dosing, which can increase toxicity and limit patient compliance.
The most common chemical degradations that occur for peptide-drug conjugates include oxidation of sulfur containing residues, deamidation of side chains of asparagine and glutamine, and hydrolysis of labile ester or imine bonds. Oxidative conditions or high temperatures that are often involved in conjugation chemistry promote these modifications at sites such as methionine and cysteine which may be prone to sulfoxide formation. Deamidation can occur spontaneously under neutral-to-alkaline conditions through a cyclic imide intermediate and results in negative charge that may affect peptide folding. Backbone cleavage and aggregation caused by temperature variations and the presence of residual moisture during long term storage may lead to batch to batch heterogeneity in both binding affinity and payload content. The chemical instabilities of PDCs can have implications for their preclinical pharmacokinetic reproducibility and may result in unpredictable immunogenicity or loss of targeting functionality prior to patient administration.
Coupling of cytotoxic warheads contributes significant steric hindrance and electrostatic alterations to the targeting peptide. In many cases, this obliterates secondary structural features, including alpha helices and beta sheets that may be required for receptor binding. Bulky moieties sterically interfere with binding surface, and hydrophobic payloads enhance self-aggregation and solvent-expose buried residues. Increased solvent exposure can make the peptide more vulnerable to proteolytic degradation. In general, even the addition of small-molecule linkers can change the peptide's dynamics by constraining its conformational freedom and therefore the population in the bioactive conformation, leading to orders of magnitude reduction in binding affinity. The entropy reduction in the system has also been shown to be deleterious not just to tumor targeting, but also to internalization kinetics since many receptors have strict peptide orientation requirements for efficient endocytosis. Thus, payload conjugation often has to strike a balance between therapeutic efficacy and the maintenance of the peptide's three-dimensional pharmacophore.
The chemistry of the linker used in a DC vaccine is important as it modulates the circulating stability and intratumoral activation. The use of cleavable and non-cleavable linkers are both associated with unique risks. Cleavable linkers like protease-sensitive dipeptides (Val-Cit) or pH-labile hydrazone groups allow selective release of the payload in the lysosomal compartment due to its acidity and abundance of cathepsins. However, these linkers are prone to premature breakdown in circulation or in non-target tissues with high expression of these proteases. Disulfide linkers may be susceptible to reductive cleavage in circulation by thiol-rich proteins such as serum albumin resulting in premature drug release despite also being activated by high glutathione levels in the tumor cytosol. Non-cleavable linkers like thioethers result in a stable conjugate that only releases its payload upon complete degradation of the antibody or peptide, which prevents leakage of the drug but also may limit intratumoral availability of the payload and bystander killing. In this regard, the ideal linker would take into account the heterogeneity of the tumor microenvironment, the distribution of proteases or other cleaving enzymes in the body, and the toxicity of the payload being used.
Table 2 Comparative Stability Profiles of Targeting Peptides in PDC versus ADC Platforms
| Stability Domain | PDC-Specific Vulnerabilities | ADC-Specific Vulnerabilities | Common Failure Mechanisms |
| Proteolytic Resistance | Entire peptide backbone exposed; rapid renal filtration reduces exposure time | Antibody scaffold provides partial steric shielding; FcRn recycling extends half-life | Serum protease recognition; lysosomal cathepsin degradation |
| Chemical Integrity | Direct modification of peptide residues; limited conjugation sites | Linker-antibody interface instability; payload loss during circulation | Oxidation of sulfur residues; deamidation of Asn/Gln; hydrolysis at labile bonds |
| Structural Preservation | Payload disrupts peptide fold; loss of binding affinity; increased immunogenicity | Minimal impact on antibody structure; peptide linker region may lose flexibility | Steric hindrance; electrostatic repulsion; conformational entropy loss |
| Linker Reliability | Requires minimal immunogenicity; limited linker chemistries available | Complex linker-payload chemistry; drug-to-antibody ratio heterogeneity | Premature payload release; off-target activation; incomplete intratumoral cleavage |
Linker chemistry allows tuning of the "therapeutic window" of circulating stability versus intratumoral activation, and cleavable and non-cleavable linkers offer different side-effect profiles. Dipeptide linkers sensitive to lysosomal proteases like Val-Cit and pH-sensitive hydrazone bonds provide efficient payload release in lysosomes of targeted tumor cells, but may also be cleaved off prematurely in the circulation, or in other organs with similar enzymatic profiles. Reduction-sensitive disulfide linkers are cleaved upon entering the higher glutathione environment of the tumor cytosol, but also can be prematurely cleaved by reducing agents such as thiol-containing plasma proteins. Non-cleavable linkers like thioethers offer greater stability from premature leakage, but are only released from ADC upon degradation of the antibody or peptide moiety itself, which may lead to less intratumoral payload delivery and bystander killing. The ideal linker choice must balance heterogeneity in tumor microenvironment, systemic exposure to cleavage enzymes and the therapeutic index of the drug payload.
The rapid clearance of linear peptides is due in part to the high susceptibility of exposed N- and C-termini and backbone amide bonds to serum aminopeptidases, carboxypeptidases, and endoproteases (trypsin-like serine proteases). The cyclization of peptides through disulfide bonds, lactam bonds, or hydrocarbon staples removes free termini and locks the backbone conformation in a protease-resistant orientation. This change alone can improve serum half-lives from minutes to hours. However, this stabilization is usually accompanied by loss of receptor availability: cyclic backbones can immobilize the binding epitope and preclude induced-fit interactions needed for high affinity binding. Furthermore, too-tight macrocycles may have reduced internalization rates as they no longer have the conformational flexibility to undergo the transition needed for receptor-mediated uptake. Designers must therefore consider both minimizing proteolysis and maintaining a flexible but well-defined pharmacophore.
Replacement of protease-cleavable amino acid residues with noncanonical or D-enantiomeric amino acid surrogates significantly enhance peptide metabolic stability. Specifically, D-amino acid incorporation at scissile bonds frustrates protease recognition. N-methylated residues prevent backbone hydrolysis but preserve key receptor contacts. Removal of cleavage signals such as the dibasic Arg-Lys recognition motif for trypsin or Asn-Gly motifs subject to deamidation decrease off-pathway degradation during storage and distribution. Peptidase resistance can also be increased by introduction of β-amino acids or peptoid residues at non-binding interfaces, also preserving the three-dimensional presentation of key side chains. All such changes at the sequence level should be rationally guided by structure-activity maps so that backbone hardening does not have unintended consequences on target selectivity or immunogenicity, to finely balance the tradeoff between stability and activity.
Helical conformations protected by intramolecular hydrogen bonds and further stabilized by hydrocarbon staples show increased resistance to proteolysis when compared to unstructured peptides. This is because the periodic secondary structure sequesters backbone amides away in a protected core. In contrast, flexible or extended conformations leave multiple cleavage sites accessible and are subject to entropic unfolding in a physiological buffer. An additional factor in the structure–function interplay is that stabilizing groups like lactam bridges, disulfide tethers, or side-chain cross-links need to be offset from the receptor-binding surface in order not to sterically interfere with binding. Excessive stabilization will trap the peptide in a non-binding conformation and decrease on-target potency. Insufficient stabilization causes the peptide to be rapidly cleared from serum. A successful design will therefore use only a minimal number of well-placed constraints to nucleate the bioactive fold while allowing enough dynamics to enable both receptor binding and endosomal escape.
Table 3 Design Strategies and Stability Outcomes in Targeting Peptides
| Design Strategy | Mechanism of Stability Enhancement | Potential Trade-Offs |
| Backbone Cyclization | Eliminates free termini; reduces endoprotease access | May rigidify binding epitope; slows internalization |
| D-Amino Acid Substitution | Impairs protease stereospecific recognition | Risk of immunogenicity; altered receptor compatibility |
| Hydrocarbon Stapling | Locks α-helical conformation; buries amide bonds | Synthetic complexity; possible off-target hydrophobic interactions |
| N-Methylation | Blocks protease hydrogen bonding requirements | Can disrupt intramolecular hydrogen bonds; reduces solubility |
| Protease-Motif Removal | Prevents cleavage at dibasic or polar residues | Limits sequence space; may reduce binding affinity |
| Disulfide Bridging | Reversible constraint; enhances serum stability | Susceptible to reduction in cytosol; may cause aggregation |
The method of conjugation impacts the molecular architecture, batch consistency, and in vivo stability of a PDC. Non-specific conjugation results in a heterogeneous library of compounds with varying drug loads and non-specific and unpredictable stabilities, while site-specific conjugation produces a homogenous library with a known stoichiometry and stable binding motifs. The different conjugation strategies impact the stability of the PDC to enzymatic degradation, chemical degradation, and conformational strain, all of which affect the therapeutic window and development success.
Site-specific conjugation can also provide better batch-to-batch consistency compared to random conjugation methods. This is because when conjugating at a single defined residue, there is a uniform drug-to-peptide ratio, in contrast to random conjugation methods, which give a distribution in drug-to-peptide ratios. In addition, by installing the payload at a single site, it can be ensured that the conjugation does not disrupt the native conformation of the peptide or its binding surface, e.g. by cysteine or enzymatic tagging at a defined site followed by conjugation to the preinstalled linker. For random conjugation strategies, however, multiple positional isomers are often formed, and attachment close to the epitope in contact with the receptor could sterically interfere with binding and/or increase clearance in serum. In addition, random conjugates can often be challenging to characterize and qualify due to their heterogeneity, while for site-specific conjugates it is easier to understand and predict their structure–function relationship and thus qualify them for regulatory submission. This trade-off may be counterbalanced by the greater development work and protein engineering needed for site-specific conjugation upfront, which may complicate the clinical manufacture of these drugs.
Chemistries like enzyme-cleavable linkers (i.e. dipeptide motifs cleaved by lysosomal cathepsins), can allow for intracellular release of the payload, but are susceptible to in vivo cleavage by circulating proteases (and others expressed on non-target organs/tissues) which would lead to degradation of the peptide before reaching the tumor site. pH-sensitive linkers like hydrazone bonds, can be cleaved in an acidic tumor microenvironment, but are at risk of premature hydrolysis in endosomal compartments during recycling of the receptor back to the cell surface which would lead to loss of the drug cargo and fragmentation of the peptide. Linkers sensitive to redox conditions (disulfide linkers) could be cleaved by higher intracellular glutathione, but are also reduced by thiol-rich proteins in plasma which would result in destabilization of the conjugate and potential systemic toxicity. All of these chemistries have different points of failure (cleavage by other proteases for enzyme-cleavable, hydrolysis during storage for pH-labile, and thiol exchange reactions for redox-active linkers). The design should allow stability in circulation and efficient activation in the tumor, while also not damaging the peptide backbone.
The addition of a small-molecule payload group itself will introduce local steric bulk which can deform the secondary structure of the peptide linker. Bigger cytotoxic warheads will have the same effect but will further introduce conformational restriction and hydrophobicity. A small payload group may only partially shield the binding interface and have only minor effects on affinity; however, a large bulky toxin can lead to global unfolding and result in aggregation as well as exposure of residues that are normally hidden from proteolytic degradation. Drug-to-peptide ratio will also impact this effect: a low ratio will decrease steric and conformational crowding but may not provide an optimal drug load, and therefore intratumoral concentrations, in the target tissue; a higher ratio will increase steric crowding which will also increase serum clearance rate and immune system recognition due to greater hydrophobic exposure. In addition, too high of a peptide: payload ratio can lead to decreased peptide flexibility, which is important for both induced-fit binding and endosomal escape, thus causing inefficient internalization. Thus, a balance must be reached in which there is enough payload present to be a therapeutic dose while not having such a large payload group that it impairs the structural integrity or receptor-mediated kinetics of the peptide.
Drug developers need access to correlated data on in vitro degradation rates, in vivo PK behavior, metabolite identification and conjugation-specific stability assessments to confidently predict in vivo performance. Serum half-lives presented without supporting evidence of the degradation mechanisms, cleavage site preferences, effects of drug: peptide ratios on stability, and effects of formulation buffers on chemical and enzymatic sensitivity is of limited value when comparing potential candidates, with only a more complete picture of evidence allowing informed decisions on whether a candidate will persist in circulation or break down before it can reach a tumor site.
As early stage methods for stability assessment, peptide conjugates are incubated in fresh, pooled plasma at 37 °C and samples are analyzed over time for the loss of parent compound by LC-MS or HPLC. These serum stability experiments highlight the peptide's sensitivity to a wide array of serine proteases, metalloproteinases, and peptidases recapitulating catabolism in the circulatory system. Drawbacks to the serum assay include the variability of enzyme composition and levels between serum pools, incomplete protein precipitation leaving low-abundance protein fragments, and the non-specific nature of the serum incubation that does not allow for control of competing chemical breakdown pathways like oxidation of Met or deamidation of Asn side chains. For these reasons, a peptide that shows long serum half-life in one preparation might exhibit very rapid clearance in another. Therefore, serum incubations are often standardized between different batches of serum and replicate samples are tested. Enzymatic assays offer higher resolution for characterizing the stability of peptide sequences by incubating the peptide with a single enzyme like cathepsin B, trypsin/chymotrypsin, or dipeptidyl peptidase to identify hotspots along the peptide sequence. These assays are useful for validating that sequence changes like D-amino acids or N-methylations sterically hinder recognition of the peptide backbone by proteases. However, these assays are also limited by their reductive design: the peptide is exposed to only one protease enzyme, which does not accurately model the competition and synergism of multiple proteases acting on the peptide simultaneously in vivo.
The most conclusive data for stability assessment comes from pharmacokinetics (PK) and biodistribution studies in preclinical species. These studies provide an endpoint of stability in terms of plasma half-life, clearance and volume of distribution, after administration into the systemic circulation. In order to conduct meaningful PK experiments, a bioanalytical method needs to be developed that can resolve intact conjugate and free payload, as well as peptide and peptide metabolites. If the latter cannot be separated, then PK data will overestimate the stability of the conjugate and result in an inaccurate dose for preclinical studies. Sample preparation can involve immunoprecipitation or size-exclusion chromatography, which when coupled with high resolution mass spectrometry allows the parent compound to be quantitated and monitored. Payload release should also be followed as a secondary readout of linker stability. In addition to PK, the biodistribution of the peptide can show tissue and organ specific accumulation and degradation to indicate whether the peptide is actually reaching the tumor target or being taken up by the liver, spleen or kidney where degradation enzymes and clearance processes are more prevalent. This can be combined with metabolite identification by LC-MS/MS mapping of tissue extracts and excreta to locate specific cleavage points and to identify off-target modifications such as glutathione adduction or sulfation, which can be useful for a root-cause analysis of unexpected toxicity or clearance. It is important to note however that in vivo studies are expensive and the number of sampling time points possible is limited.
The most common cause of failures in stability is the use of serum half-life as the sole measure of stability. Promising candidates may look robust in buffer but degrade quickly in vivo as a result of processes such as metabolism in the liver, receptor-mediated internalization and lysosomal degradation, or filtration in the kidneys. This is more likely to happen if assays are conducted against the free peptide rather than the entire conjugate, as conjugation with linkers and payloads will increase steric strain and may alter the overall charge and can also unmask cryptic sites of cleavage in the unconjugated ligand. The second major oversight is not accounting for formulation effects: the addition of surfactants, antioxidants, or pH buffers can significantly impact chemical stability, but stability profiles are often generated in simple phosphate-buffered saline that does not resemble the intended drug product. Third, developers often neglect to consider the effect of drug to peptide ratio on aggregation tendency, as higher payload loadings are more likely to cluster together and promote proteolysis and immune recognition, yet stability screens often only test a single conjugation stoichiometry. Lastly, there is a tendency to overlook long-term stability under stressed temperature and humidity conditions, leading to last-minute discovery of deamidation, oxidation, or linker hydrolysis that affect shelf life and batch consistency. These causes of attrition contribute to high failure rates, as peptides that appear stable early on may fail in preclinical toxicology or early clinical trials. These pitfalls can be mitigated with more comprehensive testing, including screening of the conjugate in its final formulation, at a range of payload loadings, and in both acute and chronic conditions to simulate clinical storage and administration.
Stability issues in peptide–drug conjugates rarely originate from a single factor. In most PDC and ADC programs, instability emerges from the interaction between peptide sequence, three-dimensional structure, conjugation chemistry, and biological exposure. Our targeting peptide development services are specifically designed to address these interconnected challenges through a stability-first, application-aware approach. Rather than treating peptide stability as a downstream optimization problem, we integrate stability considerations from the earliest design stages, ensuring that targeting performance and structural robustness evolve together.
Structure-Guided Sequence Optimization: Peptide instability is often driven by local structural features that expose cleavage-prone regions or disrupt receptor binding after conjugation. Our design strategy begins with structure-guided sequence optimization, where predicted secondary structure, flexibility, and solvent exposure are evaluated alongside target binding requirements. By identifying structurally labile regions and non-essential residues, we rationally modify sequences to improve conformational stability while preserving targeting function. This includes optimizing amino acid composition to reduce excessive flexibility, stabilizing key binding motifs, and minimizing sequence elements that are prone to chemical or enzymatic degradation. The result is a targeting peptide scaffold that maintains functional integrity under conjugation stress and systemic exposure.
Protease-Resistant Design Strategies: Proteolytic degradation remains one of the most common causes of peptide instability in vivo. To address this, we incorporate protease-resistant design strategies tailored to the intended biological context of PDC and ADC applications. These strategies may include selective amino acid substitutions at known cleavage hotspots, backbone modifications that reduce protease recognition, and architectural adjustments that shield vulnerable regions without compromising target accessibility. Importantly, protease resistance is balanced against receptor binding and internalization requirements, avoiding designs that improve stability at the expense of targeting efficacy.
Payload-Compatible Peptide Scaffolds: Many targeting peptides perform well in isolation but lose stability or affinity once conjugated to cytotoxic payloads or linkers. Our conjugation-aware peptide engineering approach addresses this issue by designing peptide scaffolds specifically for compatibility with downstream payloads. We evaluate how payload size, hydrophobicity, and chemical functionality influence peptide folding, exposure of cleavage sites, and overall stability. Based on these insights, peptide architectures are optimized to tolerate conjugation-induced steric and chemical stress, reducing the risk of destabilization after drug attachment. This ensures that peptide stability is preserved not only in its free form, but also in its final conjugated configuration.
Site-Specific Conjugation Support: Random or poorly controlled conjugation can introduce heterogeneity and destabilize targeting peptides. To mitigate this risk, we support site-specific conjugation strategies that preserve both peptide structure and targeting function. By identifying conjugation sites that are spatially separated from critical binding motifs and structurally stable regions, we help maintain consistent peptide behavior across batches. Site-specific approaches also reduce unintended modifications that can accelerate degradation or alter biodistribution, contributing to improved stability and reproducibility in PDC and ADC systems.
in vitro and in vivo Stability Testing: Stability assessment is most informative when performed across multiple biological contexts. We integrate both in vitro and in vivo stability testing into our development workflow to capture degradation pathways that may not be apparent in simplified systems. in vitro assays are used to evaluate serum stability, enzymatic susceptibility, and chemical robustness under controlled conditions. These results are complemented by in vivo studies that assess peptide integrity, metabolite profiles, and circulation behavior in relevant biological environments. Together, these datasets provide a realistic picture of peptide stability throughout the delivery process.
Data-Driven Go/No-Go Recommendations: Rather than generating data in isolation, we apply a data-driven decision framework to stability evaluation. Stability results are interpreted in the context of targeting performance, conjugation behavior, and intended application requirements. This enables clear go/no-go recommendations early in development, helping teams avoid advancing peptide candidates with intrinsic stability limitations that are unlikely to be resolved downstream. When optimization is feasible, data are used to guide targeted redesign rather than broad, trial-and-error iteration.
If stability issues are limiting the performance or progression of your PDC or ADC program, an early technical discussion can clarify whether the root cause lies in peptide design, conjugation strategy, or biological exposure. Discuss your targeting peptide stability challenge with our scientists to evaluate feasibility, identify key risks, and define a stability-focused optimization path tailored to your application.
References
