Multitargeting peptides can be envisioned as a next step from precision oncology. Tumor heterogeneity and the limitations of single targets are tackled by therapeutics capable of binding to more than one tumor-associated receptor or pathway. Peptides are well suited to multitargeting strategies because of their modular chemical nature that allows for the rational incorporation of individual binding motifs.
Structural classifications of bispecific peptides in immunotherapy 1,5
High affinity monovalent peptides remain vulnerable to stochastic loss or modification of their single target antigen and once that binding site is lost, masked or saturated pharmacologically the entire drug delivery strategy is rendered ineffective. This is further complicated by renal ultrafiltration, leading to a reduced timeframe in which even sensitive tumors are exposed to the cytotoxic payload before clearance. Objective responses to mono-epitope peptide–drug conjugates are therefore often short-lived and quickly followed by rapid regrowth leading to a shift towards multi-epitope recognition.
Each patient and, in some cases, each nodule within a patient may have its own unique repertoire of expressed receptors as a result of differing driver mutations, stromal influences and immune contexture. A peptide specific for one lineage therefore only engages a subset of malignant cells, and collateral clones are missed. These unreached populations then continue to expand, generating mixed imaging responses and incomplete pathological remissions that can be mistaken for peptide failure rather than biological inadequacy. Antigen expression can also be transient, changing in response to therapeutic stress, hypoxia or epigenetic reprogramming. A receptor that is highly expressed at baseline may be down-regulated within days of treatment initiation, leaving the original peptide ineffective while no other epitopes are targeted. This temporal drift diminishes the durability of single-target approaches and makes a strong case for multi-epitope constructs that can recognize a number of different, dynamically regulated targets to maintain therapeutic pressure across disease evolution.
Tumor escape is easy with monovalent peptides. The cell can downregulate, mutate the extracellular domain, or sequester the target. Escape mechanisms need not be as dramatic as a genomic change; transcriptional or post-translational alterations are enough to prevent ligand binding. The offending cells can then recolonize the lesion as the triggering peptide is no longer bound, so its toxin or effector is not delivered. Because the load is dependent on that one interaction, it is no longer delivered, and the lesion rapidly regrows, despite being initially sensitive to the peptide. Reliability requires reproducible binding in the disease state, in different patient populations, and across treatment cycles, which monovalent peptides are unlikely to achieve in heterogeneous malignancies. Clinically, this means that the apparent non-response rate to these therapies is unacceptably high, so they may never make it to clinical development, even if the underlying biology has simply outsmarted the probe. Multitargeting peptides have less dependence on a single target, because the overall binding event is dependent on simultaneous binding to several non-overlapping epitopes.
Table 1 Single- vs Multi-Targeting Peptides in Heterogeneous Tumors
| Parameter | Single Target | Multi-Target | Clinical Implication |
| Antigen loss escape | High | Reduced | Longer disease control |
| Imaging false negatives | Common | Less likely | More accurate staging |
| Manufacturing complexity | Low | Moderate | Higher cost, scalable by solid-phase |
| PK variability | Predictable | Multi-exponential | Requires careful dosing algorithms |
| Regulatory path | Established | Evolving | Needs new biomarker strategies |
Multitargeting peptides are molecular chimeras that display two or more distinct binding motifs within the same synthetic peptide backbone. Instead of a single receptor, they target multiple receptors either in a multivalent fashion with repeated units that bind to the same receptor or by multiple, distinct motifs that bind to different proteins. This strategy of distributed recognition is analogous to the polypharmacy approach with small molecule drugs, while preserving the rapid synthesis, modular chemistry, and fast clearance of peptides. The result is a single molecule that can promote cooperative binding events, overcome heterogeneity, and limit the development of adaptive resistance.
Multivalency with same target: multiple binding units provide avidity-derived avidity: multiple low-affinity arms work together to form a high-avidity complex that does not dissociate even if the density of individual epitopes is low. It is a strategy that can be exploited where receptor expression is heterogeneous or where a fast dissociation rate is desired to allow deeper penetration; the spacing between the units is adjusted to the expected average distance between epitopes such that simultaneous binding is geometrically likely on tumor cells but unlikely on normal cells. Introducing different targets within the same construct provides an orthogonal mode of selectivity: the peptide must encounter both antigens on the same membrane before stable binding can occur, effectively increasing the barrier for off-tumor binding. Length and flexibility of the linker is adjusted to suit the expected distance between the two receptors; a too-short tether results in steric strain while a too-long linker introduces entropy costs that weaken the interaction.
In contrast to combinatorial regimens in which two independent agents are administered, multitargeting peptides are a single chemically defined entity. This simplifies PK, reduces regulatory burden, and eliminates the drug–drug interaction uncertainties that can arise with co-formulations. The entire construct is cleared as a unit so relative exposure ratios are set, not patient dependent. Spatial coupling of the two binding events is also enforced by tethering of the motifs. Receptor clustering, cross-talk or simultaneous internalization can be incorporated into the scaffold to create biological effects that cannot be achieved by a mixture of independent ligands. Coordination can also be used to time-lock downstream signalling to synergise therapeutic or imaging contrast beyond that possible with additive monotherapy.
Multitargeting peptides condense multiple recognition codes into one scaffold. They represent a single injection translating into a synchronized multi-pronged attack, that leverages each target to amplify the effects of others. Engaging two or more tumor-restricted epitopes concurrently, multitargeting peptides spread the therapeutic load, allow for compensation for antigen dropout, and capitalize on receptor co-operativity in a way that a monovalent ligand could never achieve. The following sections will explain how this redundancy translates into broader lesion coverage, more robust retention and a higher barrier to resistance while remaining a chemically defined entity that simplifies manufacturing and regulatory oversight.
Solid tumors are highly phenotypic in which receptor expression may vary by orders of magnitude between adjacent tumor cells. Therefore, a single peptide ligand may fill the high-expressing areas in a lesion, yet spare low-expressing regions within the same tumor. Low-expressing pockets are missed by the radiotracer, but often act as reservoirs for eventual tumor relapse. Multitargeting constructs, on the other hand, recognize two different antigens, one highly and one poorly expressed. In this way, every sub-clone of tumor cells is exposed to at least one of the two ligands (biological OR-logic). This should improve targeting without changing injected dose, and therefore detect and treat micrometastases that may have been missed otherwise. If a lesion lacks the antigen necessary for binding of a given probe, the result is a false negative. The probability that two independent binding motifs are both absent can be reduced by using multitargeting peptides, which display two unrelated binding moieties. Consequently, there should be fewer "invisible" lesions in the clinical setting, and a larger proportion of total tumor volume should receive an effective radiation dose from radioligand therapy, resulting in better staging and progression free survival without increased scan time and radiation exposure.
In this situation, the dissociation of either docking arm of a bivalent peptide from its adjoined receptor is kinetically hindered since their coalesced release is a rare event. As a consequence of the avidity effect, the persistence of the radiolabel or cytotoxic warhead in cancer cells increases even though its monovalent equivalent would have already been effluxed. The prolonged dwell time at late time points in combination with low background allows for late imaging with improved contrast. Furthermore, if one of the two epitopes is temporarily obscured by the stroma, glycocalyx or antigen shedding, the other ligand will be able to occupy an accessible receptor. Since the peptide is still bound to the membrane, the resulting increase in overall residence time permits a larger number of copies of the construct to be internalized through endocytosis, thus boosting the local concentration of the delivered drug or isotope in a self-perpetuating manner that cannot be matched by monospecific ligands once the only tether is gone.
In addition, tumors may escape single-target agents by transcriptional repression or mutation of the antigen of interest. Since a multitargeting peptide offers two distinct molecular handles, concurrent inactivation of both recognition loci is a highly improbable event, and thus there is a good chance that at least one recognition pathway remains. This intrinsic redundancy acts as a biological fail-safe mechanism, helping to ensure that continuous therapeutic pressure is maintained and the development of escape variants is forestalled, without the need for dose escalation and schedule intensification. Since the global signal is the sum of two independent binding events, changes in the number of one receptor population are compensated for by the other, leading to an overall more stable pharmacokinetic behavior both between subjects and between cycles. This can simplify quantitative imaging and dose planning, decrease the inter-subject variability in clinical trials, and give clinicians more confidence that the uptake seen in the tumor is a true reflection of the total tumor burden rather than a transient antigen modulation.
Fuse two or more high-affinity codes to a single backbone and you have a pharmacological Swiss-army knife, but each additional domain also comes with added synthetic burden, steric bulk, and unmodelled contact points. The same avidity that can boost a tumor signal can equally boost weak interactions with serum proteins, basement membranes or off-target receptors, transforming a sniper rifle into a scattergun. Developers must therefore carefully negotiate a multidimensional maze in which coverage gains are exchanged for increases in molecular weight, immunogenic epitope count, batch-to-batch heterogeneity, and so on, while still maintaining a single, regulator-acceptable pharmacokinetic signature.
The emergent biophysics of the resulting construct is also rarely predictable a priori from balancing two dissociation constants, linker lengths, and epitope orientations in silico. In practice, small changes in the sequence such as shuffling up or down or adding in a protease site can flip the dominant clearance mechanism from renal to hepatic or create cryptic protease recognition sites. Iterative cycles of solid-phase synthesis, purification and in vivo imaging are also longer and more involved than for monovalent peptides. Structure–activity relationships also tend to reach plateaus more quickly than their monovalent counterparts, as improving the affinity of one domain could have the unforeseen consequence of causing a conformational distortion in the second. Computational tools trained on monovalent, single-target datasets also perform worse than they should, thus reducing the confidence of scientific teams in these methods and pushing them back towards empirical library screening and animal-heavy validation pipelines. Pharmacology of multitargeting peptides can also be non-linear: a weak arm's effect can be masked by avidity, or one arm can have a disproportionately high effect on uptake skewing the biodistribution profile in a way that neither monomer recapitulates. Emergent properties like unexpected lung retention or biliary metabolite formation arise from the new contact surfaces made at the junction between the domains, and thus make early toxicology screens harder to interpret and require greater metabolite identification effort than conventional peptides.
Additional motifs cast a wider biological net, increasing the likelihood that a receptor with a low but wide distribution will be captured. In this case, ten-fold lower affinity for these sites can manifest as a quantifiable background, as the cumulative surface area of normal tissue outweighs the tumor mass. Designers accept this tradeoff of partial coverage (dropping the ubiquitous antigens) or they introduce conditional chemistries (pH-cleavable spacers) that mute one arm outside the tumor microenvironment, with a cost increase in synthesis complexity. Multivalent constructs will also adhere to anionic basement membranes and serum albumin more readily, extending blood-pool signal that can obscure nearby lesions. Grafting or cyclization to a negative-charge will reduce this adhesion but may also decrease tumor penetration, forcing a middle-ground where coverage is sufficient and background is below PET detection limits.
Chain lengths >40 residues inherently force solid-phase synthesis to lower crude purity. The number of orthogonal protecting groups required increases if D-amino acids, PEG spacers or click handles are added. Longer sequence lengths are also more likely to aggregate during oxidative folding or cyclization steps, and thus require solvent screens and time-consuming controlled dilution protocols that incur additional cost and prolong timeline. Minor process drifts in resin loading or coupling time may also produce positional isomers where one arm is missing or mis-cyclized. Since these impurities generally have partial activity, their separation requires orthogonal purification trains (ion-exchange then size exclusion) and multi-attribute analytical methods (LC-MS, peptide mapping, NMR) that all take up a larger QC footprint compared to that of simple linear peptides. For these reasons, demonstrating reproducibility in the kilogram-scale lots requires tighter in-process controls and broader release specifications, which directly impacts timeline and facility investment.
Ideally, a multitargeting peptide should bind to two or more targets without mutual interference (steric hindrance), be eliminated from the body at the same rate (same pharmacokinetic), and display a similar payload ratio regardless of the targeted receptor. As a result, a multitargeting peptide should be a compromise between each of these attributes, rather than an ideal for each target. A multitargeting peptide with high affinity for each target may be undesirable, so design factors such as selecting targets with similar biology, spacing the motif regions so they can both bind at the same time, and adjusting the affinity of each to similar values is critical to their success.
Diverse modifications of peptides to improve pharmacokinetic properties and enhance biological activity 2,5
The expression of each part of an ideal pair is minimal or mutually exclusive in normal tissue, but they are often co-expressed on cancer cells, thus creating an OR-gate, maximizing the fraction of tumors targeted and minimizing the risk of doubling non-specific binding. Receptors are then sought in transcriptome atlases and proteomes at single-cell resolution, whose expression in healthy tissue is not detectable by PET, for example because of low expression level (generally below the milligram range); a combination of a hypoxia-regulated protein with an onco-growth-associated epitope often meets this criterion as healthy tissues do not normally express these two signals concurrently. Bioinformatic filters also screen-out housekeeping targets expressed in all cells to ensure that the coincidence of dual positivity is significantly enriched for cancer cells. Combining two widely-known but functionally disparate receptors for which reagents already exist is a very common but often a failed approach as downstream signalling cross-talk will often repress one epitope in response to binding of the other. Instead, targets are chosen that are in convergent pathways (e.g., the angiogenesis and immune evasion pathways) such that binding at both control points produces additive or synergistic anti-tumor effects, and thus converting the peptide from a molecular address label to a pathway-blocker.
The inter-arm distance is optimized to be on the order of the average inter-receptor distance on the cell membrane: too short a tether and both arms would bind to the same micro-domain (competing against each other); too long a linker and the hydrodynamic radius of the bivalent would become too large for efficient renal clearance. The standard rigid α-helical or Gly-Ser repeats are thus pre-screened in silico for their end-to-end lengths (iterating through commonly 2–6 nanometer steps) until FRET or SPR assay demonstrate co-binding of both motifs to the target receptors, in an orientation that generally has to be pre-configured such that the two arms can simultaneously dock without steric clashes. It is crucial that the structural integrity of each binding motif fold independently into its active conformation, unassisted by the other domain. A short PEG or β-alanine bridge can be inserted to provide some rotational freedom for the two halves, and cyclization of each individual arm would further pre-lock it into a receptor-ready shape that no longer relies on the conformation of the neighboring sequence to be maintained. Protease-cleavable inserts can be incorporated as a back-up release mechanism: if one arm of the linker binds to an off-target site, local enzymatic cutting releases it and lets the other half of the molecule continue on to the tumor.
Preventing a single binding site from monopolizing interactions starts with balanced affinities: a picomolar and a micromolar arm, for example, would result in the higher affinity arm saturating the target, with the result being functionally monovalent. Arms are optimized in parallel until both fall in the same nanomolar range, where probability of binding is set by antigen density instead of an inherent KD. The resultant matched binding kinetics are adjusted such that on-rates are fast, but off-rates are similar: this allows a peptide to disassociate from the first high density epitope it encounters, enabling a second and third engagement during later rounds of circulation. The final construct is more evenly dispersed throughout the tumor mass, producing a signal that is representative of overall disease burden, instead of the distribution of the most prevalent antigen alone.
Multitargeting peptides are not a panacea. The importance of these peptides will depend on the tumor biology. Multitargeting is advantageous where heterogeneity, low density of individual markers or adaptive mechanisms may compromise single target approaches. On the other hand, where a single marker has been extensively validated and is expressed in high density providing predictable imaging or therapeutic results, multitargeting approaches only complicate the process with no added clinical value. The choice between multitargeting and single targeting approaches is a balance between biological coverage and molecular complexity, i.e., the decision to use multitargeting only when the clinical benefit clearly outweighs the extra effort required for design, manufacture and regulatory approval.
Solid tumors are frequently mosaics in which receptor density differs by several orders of magnitude between neighboring nodules. A monovalent probe, while saturating hot spots, misses cold areas and seeds later recurrence. Bivalent peptides can engage both hot and cold lesions by pairing a highly abundant marker with a rare but cancer-restricted epitope in an OR-logic, thereby minimizing false-negative imaging and maximizing the fraction of lesion mass that can be targeted to therapeutic payload. Cooperative internalization further enriches payload accumulation inside double-positive cells, thereby turning biological heterogeneity into a pharmacological strength without the need to increase the injected dose. In case neither antigen alone is present in sufficient density to reach the detection threshold of PET, avidity-driven simultaneous binding of both may boost total uptake above background. This is frequent in early-stage or post-therapy lesions in which transcriptional activity is globally repressed. Linking two moderate-affinity arms can convert picomolar surface density into a measurable signal and thus enable earlier detection or use of lower radionuclide doses, while limiting radiation dose to normal organs.
When a receptor is available only to tumor tissue in quantitatively significant amounts (think oncofetal antigen with no or limited normal adult expression), then a single ligand is generally your best option. The second arm of a bivalent peptide could go after a closely related isoform expressed in normal tissue and obfuscate the high selectivity that is the hallmark of such targets. Monovalent peptides require less synthesis, QC, and regulatory effort to produce, decrease time to first-in-human, and minimize off-target exposure to radiation or drug without loss of potency. In some diagnostic indications (e.g. where a positive scan will lead to irreversible decisions of surgery, radiation) false positives are extremely significant. Monovalent peptides with picomolar affinity and rapid renal elimination provide the highest tumor-to-background ratio, while dual binders, even if carefully designed, can saturate and accumulate in normal single-positive cells and muddy up quantitative cutoffs. Therefore, when false positives are at a premium (e.g. early cancer screening, or minimal residual disease assessment) then streamlined monovalent ligands are still the format of choice.
Table 2 Decision Matrix for Mono- vs Multitargeting Peptides
| Clinical Context | Target Biology | Recommended Format | Rationale |
| Early detection | One ultra-specific antigen | Single-target | Highest contrast, lowest noise |
| Relapsed lesion | Antigen loss variants | Multi-target | OR-logic reduces false negatives |
| Adjuvant therapy | Moderate expression | Multi-target | Avidity lifts signal above threshold |
| Pediatric use | Narrow therapeutic index | Single-target | Minimizes non-tumor exposure |
| Neoadjuvant imaging | Heterogeneous staining | Multi-target | Covers all sub-clones |
Multitargeting peptides introduce opportunities to address tumor heterogeneity, but they also introduce layers of complexity that are often underestimated. Simply combining multiple binding motifs does not guarantee improved targeting performance and can, in many cases, reduce specificity or predictability. Our peptide design services are built to enable multitargeting strategies through architecture-driven engineering and feasibility-first decision making. Rather than assuming that multitargeting is inherently superior, we focus on designing peptide systems where added complexity translates into measurable functional benefit.
Architecture-Driven Design: The performance of a multitargeting peptide is determined not only by which targets are chosen, but by how binding elements are spatially arranged and functionally integrated. Our architecture-driven design approach considers peptide topology, spacing between binding motifs, and overall molecular flexibility as primary design variables. By controlling the relative orientation and distance of targeting elements, we aim to preserve independent binding functionality while minimizing steric interference and unintended cross-interactions. This structural awareness helps ensure that multitargeting peptides engage intended receptors cooperatively rather than competitively or nonspecifically.
Functional Validation Workflows: Multitargeting peptides require validation beyond single-target binding assays. We implement functional validation workflows that assess whether multiple targeting elements contribute meaningfully to overall targeting performance. This includes evaluating cooperative binding effects, selectivity across heterogeneous cell populations, and the impact of multitarget engagement on biodistribution behavior. Functional validation allows us to distinguish true multitargeting benefit from designs where one interaction dominates and negates the intended advantage of complexity.
Early Complexity Assessment: In multitargeting strategies, added complexity can quickly outweigh functional gain. Our feasibility-first approach begins with an early assessment of complexity drivers, including target co-expression patterns, binding compatibility, and the risk of increased off-target interactions. By analyzing these factors upfront, we help determine whether multitargeting is likely to improve robustness in a given oncology context or simply introduce unnecessary design risk.
Risk-Managed Design Pathways: When multitargeting is justified, we apply risk-managed design pathways that prioritize incremental complexity. Rather than fully committing to highly complex architectures early, designs are developed and evaluated stepwise, allowing performance data to guide subsequent elaboration. This approach reduces the likelihood of late-stage failure caused by overengineered peptide systems and helps ensure that each added targeting element contributes clear functional value.
Multitargeting peptides can offer meaningful advantages in certain oncology contexts, but they are not universally beneficial. An early technical discussion can help clarify whether tumor biology, target expression patterns, and program goals support a multitargeting strategy—or whether a simpler approach may deliver better performance. Discuss whether multitargeting is right for your oncology program with our scientists to evaluate feasibility, assess complexity-related risks, and define a design strategy aligned with your precision oncology objectives.
References
