Surface Charge Governs Pullulan Nanoparticle–Human Serum Albumin Complexation and Controlled Drug Release

Background

Targeted nano‑drug delivery has emerged as a cornerstone of modern oncology, offering sustained release and reduced systemic toxicity (1–3). For a nanoparticle (NP) to reach tumor tissue, it must navigate blood circulation, traverse the vascular endothelium, penetrate the extracellular matrix, and enter cancer cells (4–6). Protein adsorption—particularly of high‑abundance plasma proteins such as human serum albumin (HSA)—forms a “protein corona” that can alter NP biodistribution, cellular uptake, and drug release (10–14). The corona’s composition is driven by protein concentration, NP affinity, and surface properties, including charge, hydrophobicity, and size (15–19).

Pullulan, a neutral polysaccharide, can be chemically modified to introduce hydrophobic cholesterol groups, yielding amphiphilic polymers that self‑assemble into core–shell NPs. Previous work has shown that cholesterol‑modified pullulan (CHP) can encapsulate anticancer drugs and that HSA binds to CHP via hydrophobic interactions (28–31). Building on this, we engineered two additional variants: CH‑modified animated pullulan (CHAP) bearing amino groups and CH‑modified carboxylated pullulan (CHSP) bearing carboxyl groups. These modifications enable systematic exploration of how surface charge influences protein adsorption and drug release.

Methods

Materials

HSA (Sigma‑Aldrich), N,N‑imidazole, ethylenediamine, succinic anhydride, and other reagents were analytical grade.

Synthesis of CHP, CHSP, and CHAP

CHP was prepared by esterifying pullulan with cholesterol succinate (CHS) in DMSO, followed by purification (Fig. 1). CHAP was obtained by reacting CHP with N,N‑diimidazole and ethylenediamine; CHSP was synthesized via succinic anhydride coupling. Structural confirmation was performed by FTIR and ¹H‑NMR, yielding cholesterol substitution degrees of 4.5 % (CHP), 7.8 % (CHSP), and 14.1 % (CHAP) (Fig. 2–3).

Preparation and Characterization of NPs

Nanoparticles were formed by dialysis of the polymers in DMSO against water, followed by filtration (0.45 µm). Size and zeta potential were measured by dynamic light scattering (DLS). CHP, CHAP, and CHSP NPs exhibited mean diameters of 73.1 nm, 116.9 nm, and 156.9 nm, with zeta potentials of −0.698 mV, +12.9 mV, and −15.4 mV, respectively (Fig. 4).

ITC

Isothermal titration calorimetry quantified the thermodynamics of HSA binding to each NP type at 25 °C (Fig. 5). Exothermic binding was observed for CHP and CHSP, while CHAP exhibited an initial exothermic phase followed by endothermic binding, reflecting the interplay of hydrophobic and electrostatic forces.

Fluorescence Spectroscopy

HSA fluorescence (excitation 280 nm, emission 290–450 nm) was monitored after mixing with each NP type. Quenching analysis using a modified Stern–Volmer equation yielded binding constants of 2.02, 2.99, and 4.72 × 10⁵ M⁻¹ for CHSP, CHP, and CHAP, respectively, indicating the strongest interaction for the positively charged CHAP.

Circular Dichroism Analysis

CD spectra (200–250 nm) assessed changes in HSA secondary structure upon NP binding. α‑helical content decreased from 55 % (free HSA) to 48–52 % for the complexes, with the most pronounced reduction for CHP–HSA (Fig. 8).

Drug Release In Vitro

Mitoxantrone (MTO) was encapsulated in each NP via dialysis. Release profiles were measured in PBS at 37 °C, with and without HSA (0.1 mg/mL). The free drug released 99.8 % within 4 h. For NPs alone, 48‑h release rates were 53.7 % (CHP), 58.5 % (CHAP), and 63.2 % (CHSP). HSA presence reduced these rates to 33.9 % (CHP–HSA), 32.5 % (CHAP–HSA), and 35.8 % (CHSP–HSA) (Fig. 9).

Results

Characterization of CHP, CHSP, and CHAP Polymers

FTIR spectra confirmed successful functionalization: ester carbonyl peaks at 1731 cm⁻¹ for CHP, amide peaks at 1648 cm⁻¹ for CHAP, and carboxyl peaks at 1710 cm⁻¹ for CHSP (Fig. 2). NMR analysis quantified substitution degrees (Table 1).

Properties of CHP, CHSP, and CHAP NPs

Size and surface charge directly influenced NP self‑assembly: negatively charged CHSP and positively charged CHAP formed larger particles (156.9 nm and 116.9 nm) compared with neutral CHP (73.1 nm). Zeta potentials reflected the incorporated functional groups (−15.4 mV, +12.9 mV, −0.698 mV). The presence of charged groups disrupted core packing, yielding looser, larger structures (Fig. 4).

Thermodynamic Analysis

ITC data (Table 2) revealed exothermic binding for CHP (ΔH = 42.8 kJ mol⁻¹) and CHSP (ΔH = 80.4 kJ mol⁻¹), and a mixed exo‑endo profile for CHAP (ΔH = 22.4 kJ mol⁻¹). Entropy changes suggested easier HSA binding to CHP and CHAP (ΔS = 0.251 kJ mol⁻¹ K⁻¹ and 2.775 kJ mol⁻¹ K⁻¹) than CHSP (ΔS = 0.201 kJ mol⁻¹ K⁻¹).

Fluorescence Spectroscopy

Fluorescence quenching confirmed a rapid adsorption phase followed by slower complex formation. The binding constants derived from Stern–Volmer analysis matched ITC results, underscoring the dominant role of electrostatics for CHAP and hydrophobicity for CHP (Fig. 7).

CD Spectrum Analysis

CD measurements showed progressive loss of α‑helix upon complexation: CHP–HSA dropped to 48.0 %, CHAP–HSA to 48.6 %, and CHSP–HSA to 52.0 % (Fig. 8). The fastest helicity reduction occurred for CHP, consistent with its smaller, densely packed core.

Drug Release

HSA adsorption consistently slowed drug release across all NP types. The hierarchy of release rates (48 h) matched the sequence of NP size and surface charge: CHSP > CHAP > CHP for free NPs, and CHSP–HSA > CHP–HSA > CHAP–HSA for protein‑bound complexes (Fig. 9). These findings highlight how the protein corona modulates diffusion and degradation pathways.

Discussion

Our data confirm that surface charge is a decisive factor in HSA corona formation and subsequent drug release. The positively charged CHAP formed the strongest NP–HSA complex, attributable to combined electrostatic attraction and hydrophobic anchoring. In contrast, the negatively charged CHSP exhibited the weakest binding, limited by charge repulsion and a more porous shell that slowed protein migration to the core. Neutral CHP occupied an intermediate position, displaying the highest HSA coverage due to its compact, hydrophobic core.

These interactions influence the in‑vivo fate of drug‑loaded NPs. A dense protein corona can impede endothelial traversal, alter organ distribution, and reduce cellular uptake (15–18). By tuning surface charge, we can modulate corona composition and thus fine‑tune pharmacokinetics and therapeutic windows.

Conclusions

We successfully engineered three pullulan‑based nanoparticles with distinct surface charges. Size, charge, and hydrophobicity collectively governed HSA adsorption, binding strength, and drug‑release kinetics. Positively charged CHAP exhibited the strongest protein binding and the slowest drug release, whereas neutral CHP showed the highest protein coverage but the fastest release among free NPs. These insights provide a blueprint for designing nano‑carriers with tailored protein corona profiles to optimize targeted delivery and controlled release.