Hidden Chemistry Gap Leaves Peptide-Targeted Liposome Drug Delivery Flying Blind

One of the most promising strategies in nanomedicine may be quietly undermined by a chemistry problem that almost nobody measures. Peptide-decorated liposomes, designed to ferry anticancer drugs directly to liver tumors, rely on a well-known coupling reaction between thiol groups on peptides and maleimide groups anchored to the nanoparticle surface. Yet according to a new Perspective published in Frontiers in Drug Delivery by Kashif Maroof and Rükan Genç of Sabancı University and Mersin University, the number of maleimide groups that are actually available and reactive at the moment of coupling is a critical, underreported variable that researchers routinely conflate with the amount of lipid they initially put into the formulation.

The authors examined seven published hepatic drug-delivery systems, including liposomes decorated with cyclic RGD, the EGFR-targeting GE11 peptide, the cell-penetrating peptide BR2, a plasmin-binding tripeptide, an anti-GPC3 antibody paired with the CPP44 peptide, and the HBV-derived lipopeptide Myrcludex B. Across this entire set, only one study, Calandro and colleagues’ plasmin-binding tripeptide system, actually quantified conjugation efficiency as a fraction of peptide input. In that case, a nominal four-to-one maleimide-to-peptide ratio yielded just 33 percent conjugation, corresponding to roughly 300 peptide molecules per liposome. In the other six systems, nominal stoichiometry and identity-confirming assays such as MALDI-TOF or SDS-PAGE were used in place of true quantitative yield measurements.

Why does this matter so much? Because the biological claims built on these formulations hinge on ligand density. Du and colleagues reported more than tenfold higher accumulation of cRGD-labeled liposomes in activated hepatic stellate cells, while Lin and colleagues achieved 1.89-fold higher HepG2 uptake and 63.4 percent tumor inhibition with their dual-ligand arsenic trioxide liposomes. Zhang and colleagues found that 0.5 mol percent BR2 gave uptake indistinguishable from unmodified liposomes, whereas 2 mol percent increased uptake significantly. But if the actual number of peptides on each particle is unknown, these density-response relationships cannot be reliably interpreted, and apparent synergy in dual-ligand systems may partly reflect uncharacterized differences in effective density between comparators.

The chemistry itself is individually well understood. Maleimide rings can hydrolyze into non-reactive maleamic acid during aqueous processing, with the extent depending on the formulation and the process conditions. Work by Oswald and colleagues demonstrated the scale of this loss: when maleimide-functionalized lipid was incorporated during liposome formation, only 63 percent of active maleimide survived preparation, falling to 32 percent after dialysis, whereas post-insertion of pre-formed micelles retained 76 percent. In their solvent-injected DMPC and cholesterol system, roughly 37 percent of surface-accessible maleimide functionality was lost during pre-insertion processing, due to lipid orienting inward and rapid initial hydrolysis.

These retention figures, however, cannot simply be transplanted into other studies. The authors stress that Oswald’s measurements come from a specific lipid composition and preparation route, and comparable hydrolysis kinetics have not been established for the other six systems. A peptide-in-excess formulation reported by Du and colleagues, for example, corresponds to roughly one maleimide for every sixteen peptide molecules if calculated on total lipid, yet conjugation efficiency was never measured. In Lin’s dual-ligand system, neither the maleimide-to-antibody ratio nor the coupling efficiency for either ligand was reported, and the published CPP44 sequence lacks the cysteine stated in the Methods section, leaving the source of its reactive thiol unresolved.

The peptide side of the reaction carries its own unmeasured losses. Reduction with TCEP or DTT, or chemical thiolation with reagents like SATA, generates the reactive thiols that peptides need for coupling, but this pool degrades through re-oxidation, incomplete deprotection, disulfide formation, and solubility limits in the reaction buffer. None of the seven systems tracks either active maleimide or reduced peptide thiol over the course of the reaction, so coupling yields cannot be attributed to any specific cause. Reaction pH varied across studies, from Cheng’s pH 7.4 coupling of Myrcludex B to Zhang’s pH 6.5 coupling of BR2, both within the general 6.5 to 7.5 window of maleimide-thiol selectivity, but pH alone cannot explain yields that were never quantified.

To close this gap, Maroof and Genç propose a reactive-site accounting framework that separates six quantities often blurred together in practice: nominal maleimide input as stated in the lipid composition, reactive maleimide as the outward-facing unhydrolyzed fraction actually present at coupling, available peptide thiol as the free-thiol fraction surviving synthesis and handling, conjugation efficiency as the proportion of input peptide that attaches, final conjugated ligand density in molecules per particle, and biologically accessible ligand density as the fraction capable of engaging its receptor in the biological milieu. The framework requires two parallel accounting terms rather than one, because conjugation efficiency is set jointly by whichever reactive pool is limiting under the reaction conditions used.

The proposed workflow is concrete. Before conjugation, researchers should quantify active maleimide using a calibrated thiol-probe assay, incubating a liposome aliquot with a known excess of cysteine or glutathione under the same buffer and temperature conditions used for coupling, then quenching with excess thiol scavenger and measuring residual probe by HPLC or a validated colorimetric or fluorometric assay. In parallel, the accessible peptide-thiol pool should be measured immediately before coupling with Ellman’s reagent or a validated fluorescent probe. Both measurements should be normalized to recovered lipid, determined independently from phosphorus analysis or a validated lipid-specific assay, rather than to nominal input. After purification, residual free peptide, total conjugated peptide, and total recovered lipid should be used to calculate conjugation efficiency and molecules per particle. During biological characterization, biologically accessible ligand density and functional avidity should be determined, along with storage stability and batch-to-batch variation of the conjugation layer.

The authors emphasize that route selection must be evaluated experimentally rather than assumed. Post-insertion of pre-formed DSPE-PEG-maleimide micelles into pre-extruded liposomes better preserves maleimide reactivity than incorporating the maleimide lipid during initial liposome formation, but insertion efficiency depends on liposome size, lipid phase behavior, PEG-lipid structure, temperature, incubation time, and removal of residual micelles. Evidence from outside the hepatic set shows what complete reporting can look like: Gray and colleagues documented conjugation efficiencies consistently above 90 percent across three maleimide densities using a validated fluorescence assay, while Martínez-Jothar and colleagues found that for one peptide ligand a two-to-one maleimide-to-thiol ratio gave 84 percent conjugation, whereas a five-to-one ratio with longer incubation gave only 58 percent, demonstrating that no single ratio can be transferred across ligands and formulations.

The implications reach beyond liver targeting. Hydrolysis, steric crowding, and thiol re-oxidation are inherent to thiol-maleimide chemistry regardless of target tissue, and the seven hepatic systems serve as an illustrative case set for a broader reporting problem in nanocarrier functionalization. The authors caution that their evidence base demonstrates deficiencies in reporting rather than proving maleimide availability to be the primary cause of performance differences. They argue that reactive-site stoichiometry should be treated as a critical formulation variable, reported alongside size, polydispersity, encapsulation efficiency, and zeta potential, consistent with the FDA’s risk-based framework for liposomal drug products. They also call for single-particle methods capable of quantifying particle-to-particle ligand-density distributions and propose a factorial experiment varying the reactive-site ratio, concentration, pH, time, temperature, PEG architecture, and liposome size as a direct test of the framework. The takeaway is stark: compelling biology built on unmeasured surface density is biology that no one can reproduce, and the measurements needed to fix this are available today.

Subject of Research: Maleimide availability and reactive-site accounting in thiol-maleimide peptide conjugation on PEGylated liposomes for hepatic drug delivery

Article Title: Maleimide availability in thiol-maleimide peptide conjugation on PEGylated liposomes: a reactive-site accounting framework illustrated using hepatic drug-delivery systems

Article References: Maroof, K., & Genç, R. (2026). Maleimide availability in thiol-maleimide peptide conjugation on PEGylated liposomes: a reactive-site accounting framework illustrated using hepatic drug-delivery systems. Frontiers in Drug Delivery, 6, Article 1964437. https://doi.org/10.3389/fddev.2026.1964437

Image Credits: AI Generated

DOI: 10.3389/fddev.2026.1964437

Keywords: PEGylated liposomes, thiol-maleimide conjugation, maleimide hydrolysis, hepatocellular carcinoma, peptide targeting, conjugation efficiency, ligand density, DSPE-PEG-maleimide, nanocarrier functionalization, hepatic drug delivery, reactive-site accounting, surface functionalization