From single DNA molecule to nanoparticle formation: mechanistic basis for monocationic aromatic drug-induced DNA frameworks

Abstract

We investigated the interaction between the monocationic aromatic drug propranolol (PPL) and double-stranded DNA (dsDNA) to elucidate how small molecules can drive higher-order DNA frameworks and nanoparticles (NPs) formation. Single-molecule force spectroscopy with optical tweezers revealed that, at concentrations below 4 mM, PPL interacts with dsDNA through an intercalation-like mode, altering contour length, persistence length, and stretch modulus. At higher concentrations, PPL induced dsDNA compaction, corroborated by atomic force microscopy imaging of condensed structures. Multimolecular assays supported these findings: electrophoretic mobility shift assays revealed progressive mobility loss with increasing PPL concentrations, consistent with aggregate formation, while UV-vis spectroscopy confirmed intercalation-like behavior and strong binding affinity (K_b=1.67 × 10⁶ M⁻¹). At millimolar PPL/DNA ratios (10–14), NPs formulations were obtained with hydrodynamic diameters of 120–244 nm, low polydispersity (0.19–0.30), negative zeta potential (-25 to -35 mV), and particle concentrations up to 5.26 × 10¹¹ NPs/mL. These NPs exhibited very high drug loading (59–72%) and stability under both biological and storage conditions. Collectively, our results demonstrate that PPL engages dsDNA through intercalation-like behavior, compaction, aggregation, and stabilization processes, uncovering a previously unreported mechanism for a monocationic aromatic drug and allowing the efficient formation of NPs. This work expands the current understanding of small molecule-DNA interactions and may be extended to other hydrophilic aromatic drugs, positioning DNA as a versatile building block and ultimately for the development of nucleic acid-based nanomedicines.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04427-w.

Introduction

The development of nanomedicines has gained significant attention due to their potential to improve therapeutic efficacy and selectivity, particularly for chronic diseases such as cancer, cardiovascular diseases, diabetes, and neurodegenerative disorders, which remain leading causes of death and morbidity worldwide [1–3]. In this context, considerable efforts have been devoted to the design of nanoparticulate systems capable of efficiently encapsulating and delivering bioactive molecules.

Among the strategies for advancing nanomedicines, exploiting the specific molecular interactions between components (e.g., polymers, biomolecules, therapeutic agents, and others) has emerged as a promising approach to enhance drug entrapment, stability, selectivity, and controlled release. In particular, ionic and aromatic interactions between hydrophilic, low molecular-weight drugs (HALMD) and polymeric water-soluble excipients represent a promising strategy to overcome formulation challenges associated with conventional nanocarriers [4, 5]. Many clinically relevant drugs possess low molecular-weight (less than 1000 Da), weakly ionizable functional groups and at least one aromatic ring (e.g., propranolol, salbutamol, amitriptyline, chlorpheniramine, doxorubicin, among others) [6–8]. Despite their efficacy, these features often limit their retention within conventional delivery platforms such as hydrogels [9–11], w/o/w emulsions [12–14], liposomes [15–17] and microspheres [18], leading to low encapsulation efficiency, premature leakage, and burst release.

To address these limitations, nanomedicines based on aromatic interactions between HALMDs (e.g., amitriptyline, imipramine, cyclobenzaprine, promethazine, chloroquine) and synthetic non-biodegradable aromatic polymers have demonstrated promising results yielding stable nanoparticles with narrow size distributions, high drug loading, and extended stability under biological and storage conditions [19–21]. However, the non-biodegradability and lack of biological activity of charged aromatic polymers is recognized as a drawback for such nanomedicines. This has motivated growing interest in DNA as an alternative charged aromatic framework, owing to its polyanionic backbone, stacked aromatic bases, biocompatibility, and biological functionality [22].

Aromatic interactions play a central role in DNA stabilization and molecular recognition, governing base stacking, protein-DNA binding, and interaction with small-molecule ligands as intercalators [23–27]. In addition, nucleic acid-based molecules, including DNA, have emerged as powerful therapeutic agents due to their ability to regulate gene expression, modulate immune responses, and act as diagnostic tools. However, their inherent instability and susceptibility to enzymatic degradation pose significant challenges for formulation, effective delivery and clinical application.

The interactions between DNA and HALMD are of particular interest in therapeutic design [28, 29]. HALMDs typically interact with DNA by one of three mechanisms (intercalation, groove binding, or condensation) each of which induces distinct structural effects. While various nanomedicines based on HALMD-DNA combinations have been reported (including DNA origami [30–33], hydrogels [34, 35], and DNA-coated inorganic materials such as gold nanoparticles [36, 37]), their fabrication often relies on complex, multistep protocols involving thermal cycling for hybridization, auxiliary excipients or use of specific solvents. Such approaches frequently require extensive purification limiting practical implementation [38–40]. Despite these advances, a mechanistic understanding of how HALMD-DNA interactions propagate from the unimolecular scale to higher-order framework and nanoparticle formation remains incomplete. Addressing this gap is essential for the rational design of DNA-based nanomedicines that rely on minimal components and simple assembly routes.

In this study, we applied a multiscale experimental approach (from unimolecular to multimolecular) to investigate the interaction between double-stranded DNA and the hydrophilic, low molecular-weight drug propranolol (PPL). To achieve this, we combined: (1) single-molecule force spectroscopy to quantify the mechanical responses of dsDNA upon PPL binding; (2) atomic force microscopy (AFM) to visualize conformational changes and aggregation; (3) electrophoretic mobility shift assays (EMSA) and UV-vis spectroscopy to evidence chemical interactions; and (4) dynamic light scattering (DLS), scanning transmission electron microscopy (STEM) and nanoparticle tracking analysis (NTA) to characterize the resulting nanoparticles (NPs). Our results reveal that a monocationic aromatic molecules can drive a hierarchical transition from molecular-scale binding to higher-order DNA frameworks, ultimately enabling the spontaneous formation of stable DNA-based NPs through straightforward aqueous mixing. This work provides mechanistic insight into how electrostatic and aromatic interactions can be exploited to design DNA-based nanomaterials.

Materials and methods

Materials

Propranolol hydrochloride (295.8 g/mol), metoprolol tartrate salt (684.82 g/mol) and doxorubicin hydrochloride (579.98 g/mol) were purchased from AK Scientific (California, USA) and was used as received. Double-strand λ-phage DNA (48502 bp) was purchased from the New England Biolabs company (Ipswich, USA) and did not go through purification before use. Primers for PCR (5’-TTAAGTCGCTTGAAATTGCTATAAGCAGAG-3’, 5’-TGATCAACTGGCTTTCCAAACTCGTATTCG-3’, 5’-AGTGCTGGCTGAATACCACAAACAGATTGA-3’) were synthetized by Integrated DNA Technologies, Inc (Iowa, USA). Dreamtaq™ DNA polymerase, GelRed Nucleic Acid Gel Stain 10.000X, 6X TriTrack DNA loading dye, GeneRuler 1 kb ladder and GeneArt™ Linear pUC19L vector were purchased from Thermo Fisher Scientific (Waltham, USA). Calf Thymus dsDNA from Merck Millipore (cat. No.: 2618, Germany) was dissolved in water (1 mg/mL) and dialyzed using a Pur-A-Lyzer™ Maxi Dialysis Kit (MWCO, of 3500 Da, Sigma Aldrich, USA) for 24 h, room temperature, gentle agitation and using Milli-Q water as a receptor medium. The dsDNA concentration (ng/µL) was estimated by measuring the absorbance at 260 nm in a NanoDrop 1000 UV-vis spectrophotometer (Thermo Scientific, USA). Subsequently, the molar concentration was calculated considering the nucleotides as the monomeric units, thus an average MW of 330 g/mol. The anionic aliphatic polymer poly(sodium vinylsulfonate) (PVS) (130.1 g/mol of monomeric units) was purchased from Sigma Aldrich (USA) and was used as received. The pH was adjusted with an Edge^® HI2002 pH meter (Hanna Instruments, USA) with minimum amounts of NaOH and HCl (Merck, Germany). All the solutions were prepared with Milli-Q water obtained from a Simplicity SIMS 00001 equipment (Millipore, France).

Methods

DNA stretching experiments

Standard molecular biology protocols were used to generate, label, and purify the dsDNA fragment/molecule [41–44]. A single 10 kbp dsDNA fragment/molecule was generated by polymerase chain reaction (PCR) using the λ-phage as a template (48502 bp, New England Bioloabs, USA) and primers labeled with biotin (5´ end) and digoxigenin (3´ ends). These modifications allowed tethering of the DNA molecule between streptavidin-coated and anti-digoxigenin-coated beads in the optical tweezers experiments. Details of PCR procedure are provided in supplementary information (S1), and the obtained PCR product was purified using the Wizard^® SV Gel and PCR Clean-Up System (Promega, USA).

Stretching experiments were performed using a dual-trap C-Trap instrument (Lumicks, Netherlands) with a five-chamber microfluidic flow cell. Measurements were conducted in PBS buffer to ensure molecular stability, controlled pH, and appropriate ionic strength. The use of this buffer also enables reliable comparison of dsDNA mechanical properties, including the characteristic overstretching transition (~ 65 pN), thereby confirming proper single-molecule tether formation. A single dsDNA molecule was tethered between two polystyrene beads (3.1 μm streptavidin-coated and 2.1 μm anti-digoxigenin-coated). To that, 10 µL of 10 kbp dsDNA (1:50 diluted PCR product), 5 µL streptavidin-coated polystyrene beads and 15 µL of buffer phosphate PBS (Corning^® PBS, USA) was incubated during 15 min and then adjusted to 300 µL. In addition, 1 µL of anti-digoxigenin-coated polystyrene beads were diluted with buffer phosphate PBS to 1 µL. Measurements were conducted at increasing PPL concentrations (0.1–50 mM), and stretching rate of 150 nm/s. At least three molecules and 10–30 cycles per concentration were analyzed. Only tethers with a validated contour length of ~ 3.4 μm were used. Force-distance curves were fitted to the extensible worm-like chain (eWLC) model using Python scripts (available at Lumicks). Key mechanical parameters (contour length Lc, persistence length Lp, and stretch modulus St) were extracted for each condition.

The fractional elongation (q), an indicator of the extent of drug binding to dsDNA, was determined from the fractional lengthening at defined force and ligand concentration, following Eq. (1):

1

where corresponds to experimental dsDNA lengthening at a given force (5–40 pN) and PPL concentration (1–4 mM), corresponds to the dsDNA lengthening without PPL at the same force range, and corresponds to the maximum dsDNA lengthening observed at that force range.

To quantitatively analyze the cooperativity of PPL binding to dsDNA, the fractional elongation data were fitted to the Hill equation:

2

were is the fractional elongation, is the ligand (PPL) concentration and is the apparent dissociation constant. represents the Hill coefficient and quantifies the degree of cooperativity, a Hill coefficient indicates positive cooperative binding among ligand molecules [45, 46].

Atomic force microscopy

Images were obtained using Veeco IIIa Digital Instruments AFM, in air tapping mode via a Hi-RES AFM tip (match Company, 160 ~ KHz resonance frequency). Images (512 × 512 pixels, 2 × 2 µm²) were captured from samples deposited on freshly cleaved muscovite V-1 mica. For dsDNA imaging, 20 µL of pUC19L vector (78 pM; expressed as nucleotide concentration, assuming an average molecular weight of 330 g/mol per nucleotide) was mixed with 2 mM MgCl₂, applied to mica for 2 min, rinsed with Milli-Q water (1 mL) and dried with under nitrogen flux. For PPL-dsDNA complex, increasing concentrations of PPL (6.25, 25 and 100 mM) were incubated with dsDNA for two min before deposition. Occupancy area was calculated using MatLab Software and following the procedure previously described [47].

Electrophoretic mobility shift assay (EMSA)

The EMSA assay was performed using dsDNA (> 10 kbp) at a constant concentration of 6 × 10^− 5 M (expressed as nucleotide concentration, assuming an average molecular weight of 330 g/mol per nucleotide), while PPL was tested in a concentration ranged from 0 to 1 × 10^− 2 M. For each reaction, 5 µL of dsDNA were mixed with 5 µL of PPL solution and 2 µL of 6X TriTrack DNA loading dye. The mixtures were incubated at room temperature for 5 min to allow complex formation. Subsequently, 10 µL of each sample were loaded onto 0.5% (w/v) agarose gels pre-stained with GelRed for DNA visualization. Electrophoresis was conducted in 1X TAE buffer (40 mM Tris-acetate, 1 mM EDTA), commonly used to preserve DNA integrity and ensure appropriate electrophoretic resolution during analysis, at 100 V for 60 min. DNA migration shifts caused by complexation with PPL were visualized under a UV transilluminator, and gel images were analyzed using ImageJ software.

UV-vis absorption spectra

The UV-vis spectra were acquired with a Duetta spectrophotometer (Horiba Scientific, Japan) using a quartz cuvette (Hellman^®, Germany) of 1 cm of path length. Equal volumes (500 µL each) of PPL and dsDNA were mixed and incubated for 5 min at room temperature. PPL concentration was fixed at 7.5 × 10^-5 M, while varying the dsDNA concentration from 0 to 5 × 10^-4 M (expressed as nucleotide concentration, assuming an average molecular weight of 330 g/mol per nucleotide). The absorption spectra (λ = 200–800 nm) were recorded and plotted to illustrate spectroscopic features such as hypochromic and bathochromic shifts. These effects became more pronounced with increasing dsDNA concentrations and provide qualitative insight into PPL-dsDNA interactions. To determine intrinsic binding constant, K_b, the concentration of PPL was fixed at 7.5 × 10^-5 M, while the concentration of dsDNA was varied between 1 × 10^-6 and 5 × 10^-5 M (expressed as nucleotide concentration, assuming an average molecular weight of 330 g/mol per nucleotide). The selected PPL and dsDNA concentrations are in the linear range of absorbance, in accordance with the Lambert-Beer law. Calibration curves, linear equations and linear adjustments (R²) for both components are provided in the Supplementary Information (Figure S1). Absorbance values at 295 nm (the characteristic wavelength of PPL) were recorded and fitted to Eq. (3), derived from the classical binding model described by Wolfe-Shimer for drug-DNA interactions [48]:

3

where ε_a, ε_f, and ε_b are the apparent, free and bound complex extinction coefficients, respectively. The value of ε_f was determined from the calibration curve of free PPL, and ε_a was determined as the ratio between the measured absorbance and the PPL concentration, A_obs/[PPL]. A plot of [DNA]/(ε_a-ε_f) versus [DNA] gave a slope of 1/(ε_b-ε_f) where the y-intercept equals to 1/K_b(ε_b-ε_f).

Preparation and physicochemical characterization of PPL/dsDNA nanoparticles

Preparation of PPL/dsDNA nanoparticles

The PPL/dsDNA formulations were synthetized according to the method previously reported by our group [19–21]. Briefly, 500 µL of cationic PPL was added dropwise to 500 µL of anionic dsDNA under continuous stirring (room temperature, 5 min) in aqueous solution, and absence of any buffer, to minimize ionic shielding able to affect ionic and aromatic interactions between components [49–51], thereby promoting interactions between the oppositely charged functional groups and aromatic cyclic groups belonging to PPL and dsDNA. Final concentrations were adjusted to achieve the desired molar ratios ([PPL]/[dsDNA]). The dsDNA concentration was kept constant (1.44 × 10^{− 3} M; expressed as nucleotide concentration, assuming an average molecular weight of 330 g/mol per nucleotide), while the PPL concentration was varied to obtain final molar ratio between 0.1 and 22 (1.44 × 10^{− 4} − 3.2 × 10^{− 2} M). The anionic polyelectrolyte PVS was used as a non-aromatic control.

Physicochemical characterization of PPL/dsDNA nanoparticles

The presence of dispersed particles in aqueous medium was initially analyzed by turbidimetry in a Duetta spectrophotometer (Horiba, Japan). For that, the absorbance of 1 mL of each formulation was measured at a wavelength where none of the compounds (PPL and dsDNA) absorb (λ = 650 nm). Each sample was analyzed in triplicate at 25 °C. The results are shown as the mean ± standard error of the mean for n = 3.

The hydrodynamic diameter and zeta potential of the formulations were determined by DLS and laser Doppler anemometry (LDA), respectively, using a Zetasizer NanoZS (Malvern Instruments, UK) equipped with a standard λ = 633 nm laser as the incident beam. The formulations were diluted in Milli-Q water and loaded into a disposable folded capillary cuvette (DTS1070). The results were analyzed using the ZetaSizer v7.12 software. Each analysis was performed in triplicate at 25 °C. The results are shown as the mean ± standard error of the mean for n = 3.

The determination of the nanoparticle concentration was conducted in a NanoSight NS300 (Malvern Instruments, UK). The samples were diluted from 5 to 10 times with Milli-Q water to achieve an optimum concentration range of 10⁷-10⁹ particles/mL. A minimum of five videos (one min each one) of the particles moving under Brownian motion were captured. The videos were analyzed for size distribution and particle concentration using the built-in NTA v3.0 software (Malvern Instruments, UK) [52].

The morphological characterization was carried out in scanning transmission electron microscope (STEM), model Inspect F-50 (FEI, Holland). STEM images were obtained by sticking a droplet (20 µL) of the formulation to a copper grid (200 mesh, covered with Formvar) for 2 min, then removing the droplet with filter paper avoiding the paper touching the grid, then washing the grid twice with a droplet of a Milli-Q water for 1 min and removing the droplet with a filter paper. Subsequently, the sample was stained with a solution of 1% (w/v) phosphotungstic acid by adding a droplet to the grid for 2 min and then removing with filter paper. The grid was dried at room temperature for at least 1 h prior to being analyzed [53].

Drug association and loading of PPL/dsDNA nanoparticles

Drug association and loading were obtained as previously described [20]. The association efficiency of PPL in the nanoparticles was determined by analyzing the ratio between the amount of drug associated in the formulation (experimental value) and the total initial drug (known value). The drug loading (% w/w) was calculated as the ratio between the associated drug mass and the theoretical total nanoparticle mass (defined as associated drug mass + initially added dsDNA mass). The amount of drug associated with the nanoparticles was determined indirectly by quantifying the free (non-associated) drug in the medium. Free PPL was separated from nanoparticles using Vivaspin^® 6 centrifugal tubes (MWCO 3.5 kDa, 5000 G x 40 min), which operate as a centrifugal filtration through a semipermeable membrane [19–22, 54–56]. The quantification of PPL was done by measuring the absorbance at 295 nm using a Duetta spectrophotometer (Horiba Scientific, Japan). The standard curve of PPL was linear (R² > 0.999) in the range of concentrations between 1.5 × 10^{− 4} M and 3 × 10^{− 5} M (molar extinction coefficient was 5412.8 M^{− 1}cm^{− 1}). The added dsDNA was previously purified (Pur-A-Lizer™ Maxi dialysis kit, MWCO 3.5 kDa), quantified (Nanodrop 1000 UV-vis, Thermo Scientific, USA) and freshly used. The results are shown as the mean ± standard error of the mean for n = 3.

Stability assays of PPL/dsDNA nanoparticles

The stability of the formulations was evaluated in terms of hydrodynamic diameter and zeta potential at different temperatures (20–55 °C), pH variations (2–10), and storage time (12 weeks, 4 °C) using a ZetaSizer NanoZS. For pH variations, HCl (0.25 − 0.01 M) and NaOH (0.25 − 0.01 M) solutions were selected and controlled with the automatic titrator. The temperature of the samples was modified directly in the NanoZS instrument (20–50 °C), with a thermal equilibrium time of 15 min for each measurement of hydrodynamic diameter and zeta potential. The results are shown as the mean ± standard error of the mean for n = 3.

Results and discussion

Unimolecular dsDNA Analysis Upon Interaction with PPL

The structural properties of dsDNA have been extensively studied to gain insight into the mechanism underlying its compaction, through both theoretical and experimental approaches. Most studies indicate that dsDNA compaction occurs when approximately 90% of the negative charge is neutralized by positively charged low-molecular weight species with a valence of + 3 or higher [57–59]. The selected drug for this work (PPL) shows two condensed aromatic rings and a flexible aliphatic tail ending in a weakly basic amine group (Fig. 1). Importantly, due to the pKa of PPL (9.45), it only shows a single positive charge at physiological pH. Interestingly, PPL has been shown to interact with DNA and influences various cellular processes, such as proliferation [60–63], migration [62, 63], and apoptosis [60, 61, 64].

Fig. 1.

Structure of propranolol, showing two condensed aromatic rings and a basic amine tail

In this study, we initially aimed at investigating the mechanical alterations induced by PPL on a single dsDNA molecule, to further elucidate how PPL-dsDNA interactions affect the structural and physicochemical properties at a molecular level. This single dsDNA molecule approach enabled the identification of specific changes in DNA elasticity and conformation, providing the arguments of the interactions, in addition to mechanistic insights, guiding subsequent analyses. A single 10 kbp dsDNA fragment/molecule was tethered between two optically trapped beads and subjected to repeated stretching-relaxation cycles, both in the absence and in the presence of increasing concentrations of PPL (until 50 mM). Fragments/molecules of 10 kbp were selected as the most suitable length for these experiments, as shorter fragments/molecules were difficult to trap and manipulate. For each tested PPL concentration, between 3 and 5 independent dsDNA assays were analyzed (10 to 30 cycles were recorded per dsDNA). The experimental setup is illustrated in Fig. 2A and an experimental result of force-distance profile corresponding to the selected dsDNA in their native form (without PPL) is shown in Fig. 2B. As evidenced, this curve is characterized by a rapid increase in force over a short extension interval (phase 2), which is attributed to the elastic deformation of B (helicoidal)-form dsDNA structure. The force-distance curve was fitted to the extensible worm-like chain (eWLC) model to extract mechanical parameters such as the contour length (Lc), persistence length (Lp), and stretch modulus (St), as summarized in Table 1. The obtained Lc (3.266 ± 0.001 μm) and Lp (59.999 ± 0.003 nm) values correspond to the maximum extension and flexibility of the dsDNA in their B-form and are consistent with those expected for a 10 kbp dsDNA fragment/molecule (theoretical Lc of 3.4 μm corresponding to 0.34 nm per base pair).

Fig. 2.

(A) Optical tweezers experimental setup model. (B) Experimental force-distance profile (extension is represented by a continuous line and relaxation by dashed line) for the selected 10 kbp dsDNA fragment/molecule displaying four distinct phases: Phase 1: At low forces (< 10 pN) the stretching profile is dominated by an opposing, primarily entropic force generated by the reduction in accessible conformations as the dsDNA is stretched. Phase 2: At higher forces (up to 35 pN) elastic deformation of B-form dsDNA occurs increasing the force needed to stretch the dsDNA. Phase 3: Around 65 pN, the molecule undergoes an overstretching transition, often associated with a force-induced melting or structural transition of the double helix. A hysteresis effect is observed, as evidenced by the non-overlapping extension and relaxation curves. Phase 4: after the overstretching transition, most of the dsDNA is converted into ssDNA but a few GC rich regions may hold the two strands together

Table 1.

Fitted parameters of the eWLC model applied to force-distance curves of the 10 kbp dsDNA fragment/molecule in the absence and in the presence of increasing concentrations of PPL

[PPL] (mM)	Lc (µm)	Lp (nm)	St (pN)
0	3.266 ± 0.001	59.999 ± 0.003	1225.3 ± 11.9
1	3.259 ± 0.002	59.999 ± 0.006	1411.1 ± 34.8
1.5	3.242 ± 0.007	59.999 ± 0.024	1218.3 ± 33.5
2	3.676 ± 0.001	56.351 ± 0.003	220.3 ± 2.1
2.5	3.839 ± 0.004	43.479 ± 0.012	264.7 ± 2.3
3	3.902 ± 0.004	21.588 ± 0.016	282.3 ± 2.5
4	4.528 ± 0.020	15.108 ± 0.088	522.8 ± 31.4

The force-distance curves of a 10 kbp DNA fragment/molecule in the presence of increasing concentrations of PPL (0–4 mM) are shown in Fig. 3A. Table 1 summarizes the corresponding parameters obtained from fitting the curves to the eWLC model. As evidenced in Fig. 3A, in the presence of PPL we observed a progressive increase in Lc (from 3.266 ± 0.001 to 4.528 ± 0.020 μm) which is commonly associated with intercalation-like interactions with dsDNA. Previous studies have shown that distinct DNA binding modes generate characteristic force-extension responses in optical tweezers experiments. Minor groove binders generally do not produce a significant increase in Lc and tend to stabilize the overstretching transition by shifting it to higher forces [65]. In contrast, major groove association [66] or mixed binding modes [67] may generate mechanical responses that are not readily distinguishable from intercalation-like behavior, thus groove binding or mixed binding modes cannot be excluded. This process occurs under equilibrium conditions, as demonstrated by the near-complete overlap of the extension and release curves (supplementary information, Figure S2), indicating that binding PPL to dsDNA is fast with respect to the time scale of DNA stretching. Similar results in terms of Lc increase and overlap of the extension and relaxation curves has been reported in the presence of other intercalating drugs such as mitoxantrone (MTX) [68], doxorubicin (DOX) [69], hydroxychloroquine (HCQ) [70], actinomycin D (ACD) [71]. Intercalating drugs bind by inserting between adjacent base pairs, physically separating them and increasing the Lc value [72–74]. Among the physicochemical characteristics that the above drugs possess, we underline their cationic character (2 charges for MTX and HCQ, 1 for DOX and ACD), aromaticity (3 conjugated aromatic rings for ACD, MTX and DOX and 2 for HCQ) and different Log P (DOX:1.3, MTX:1.4, ACD:2.5, and HCQ:3.8). For the case of PPL, this drug shows a single cationic functional group, two condensed aromatic rings, and a Log P of 2.5. Despite lacking the typical structural features of classical intercalators (i.e., a rigid and planar extended aromatic system), our results provide experimental evidence of intercalation-like behavior for PPL. To our knowledge, this is the first study to provide experimental evidence of intercalant-like behavior for any molecule with similar characteristics to PPL (in terms of charges, aromaticity, and Log P) into dsDNA.

Fig. 3.

(A) Force-distance curves of a 10 kbp dsDNA fragment/molecule in absence and in the presence of increasing concentrations of PPL (0–4 mM). (B) Overstretching forces in the absence and in the presence of PPL at 1.5 and 2.5 mM (solid lines: stretching; dashed lines: relaxation). (C) Fractional elongation per base of the 10 kbp dsDNA versus PPL concentration (0–4 mM) at different forces (10–40 pN) (symbols: experimental data; solid lines: Hill equation fits)

In addition, a significant decrease in both Lp (from 59.999 ± 0.003 nm to 15.108 ± 0.088 nm) and St (from 1225.3 ± 11.9 pN to 220.3 ± 2.1 pN) is evidenced after exposing PPL to dsDNA (Table 1). The impact on Lp appears to be more variable and depends on the specific intercalating agent and their concentration. For instance, studies with the classic intercalator ethidium bromide have shown that the binding can either decrease, increase, or have minimal effect on the apparent value of Lp [75]. Kreft et al. reported a decrease in the DNA Lp in the presence of MTX [68]. In contrast, Lima et al. observed an initial increase in Lp at low MTX concentrations (≤ 0.6 µM), followed by a marked decrease at higher concentrations (up to 3 µM) [76]. PPL binding may induce changes in the dsDNA hydration shell, contributing to significant structural modifications and a consequent decrease in the Lp value [74, 77, 78]. Notably, St tends to decrease upon intercalation-like binding, indicating a reduction in the stiffness of the dsDNA molecule under tensile stress, likely due to the disruption of self-stacked structure of adjacent DNA bases, and consequent hydrogen bonding weakening between the dsDNA pair bases [27, 79, 80], providing a less organized structure.

Together with the above observations, another important aspect indicative of the intercalation-like effect of PPL on dsDNA is the strength-increasing during the overstretching stage (phase 3, Fig. 3B) [74, 81]. At high forces (> 60 pN), the native B-form dsDNA undergoes an overstretching transition, during which the double helix partially unwinds [82–84]. This transition appears as a plateau in the force-distance curve, where the dsDNA molecule gains ∼70% in Lc over a narrow force range [74, 81]. As also shown in Fig. 3B, the overstretching force rises from 60 pN in the absence of PPL, to ~ 80 pN (PPL at 1.5 mM), and ~ 120 pN (PPL 2.5 mM) in the presence of the drug. Simple models suggest that molecules bound between base pairs or interacting in an intercalation-like manner to B-form DNA confer additional stability to the dsDNA. Consequently, higher forces are required to induce the transition to the overstretched state (phase 3) [85]. Therefore, the observed changes in dsDNA mechanical properties (Lc, Lp y St), along with the rise in overstretching forces, align well with the characteristic behavior of high-affinity intercalating consistent with intercalation-like DNA-binding molecules [68–70, 81].

In addition to providing mechanical parameters, force-distance curves of dsDNA fragment/molecule also contain relevant information regarding the PPL-dsDNA intercalation-like mechanism. In Fig. 3C, the fractional elongation (q), calculated as the ratio between the dsDNA extension at a given PPL concentration and the extension at saturating PPL levels, is shown. For classical intercalators, the fractional elongation can often be described by the non-cooperative McGhee-von Hippel binding isotherm, which typically displays a hyperbolic profile [86–88]. The sigmoidal behavior observed for PPL (Fig. 3C) deviates from this pattern. To quantitatively assess this behavior, the data were fitted using the Hill equation (see Eq. 2 in Material and Methods), yielding a Hill coefficient between 12 and 15 across all force conditions (supplementary information, Table S1). The corresponding fitted curve is shown in Fig. 3C. This value is consistent with a cooperative binding behavior indicating that binding events are not independent [45, 46]. This cooperative/complex/hierarchical mechanism could be indicative of a dynamical aggregation behavior, valuable to obtain larger and tight interacting self-aggregated structures [21].

In additional force-distance experiments, periodic jumps were detected at testing PPL at 10 mM and 50 mM (Fig. 4A). These patterns indicate a higher degree of molecular organization induced by increasing PPL concentration and are interpreted as signatures of dsDNA compaction in the presence of the drug. The force-distance curves exhibited significant force jumps and different apparent Lc (supplementary information, Figure S3). Each jump corresponds to abrupt specific changes in the dsDNA conformation due to the interacting PPL at increased concentrations (compared with patterns observed with PPL ≤ 4 mM), and has been widely described as a compaction phenomenon [84, 89–91]. Interestingly, as observed in Figure S3, the extension and relaxation curves in the stretching-relaxation cycles do not overlap indicating hysteretic behavior, suggesting that the force-induced conformational transitions in this regime are not fully equilibrated within the experimental timescale. At higher PPL concentrations (> 10 mM), the force-extension curves display discrete rips, indicating cooperative structural transitions associated with PPL-induced DNA compaction. Hysteresis likely reflects kinetic barriers associated with the rupture of compacted structures during stretching and their delayed reformation during relaxation. In addition, such interactions may preferentially form when the DNA molecule is in a relaxed configuration, where the reduced bead-to-bead distance facilitates intramolecular contacts mediated by PPL. Importantly, dsDNA compaction has only been reported upon exposure the dsDNA to molecules with 3 positive charges (spermidine), at elevated concentrations of agents with 2 positive charges (magnesium), as well as to the presence of precipitation agents (ethanol) [58, 92, 93]. Together, optical tweezers experiments demonstrate that PPL predominantly interacts with a dsDNA molecule via intercalation-like binding, inducing significant mechanical changes and leading to single dsDNA compaction.

Fig. 4.

(A) Force-distances curves in the presence of PPL at concentrations of 10 and 50 mM, displaying the stick-release pattern indicative of dsDNA compaction. AFM micrographs (uppercase) and their calculated occupation area (lowercase) of dsDNA (linear plasmid pUC19L) (B-b) in the absence and in the presence of (C-c) 6.5 mM, (D-d) 25 mM and (E-e) 100 mM of PPL

In order to analyze low, intermediate, and high interaction regimes, while promoting PPL-dsDNA interactions at unimolecular dsDNA level, atomic force microscopy (AFM, in air tapping mode) was used to visualize the spatial conformation of single dsDNA fragments/molecules in the presence of PPL. For these studies, a shorter commercially available linearized pUC19 plasmid (2870 bp) was used since longer dsDNA fragments/molecules, such as those used in optical tweezers, tend to overlap on the mica surface and hinder visualization of individual molecules. The selected concentration of dsDNA was fixed to 78 pM and incubated with increasing concentrations of PPL at 0, 6.25, 25, 100 mM to promote the visualization of the fragments/molecules at different regimes. Consistent with previous results, increased drug concentrations promoted the single dsDNA chain condensation and compaction, resulting in the formation of molecular frameworks (Fig. 4B-E; and supplementary information Figure S4 to Figure S7). AFM in air tapping mode is a well-established and widely validated technique for the morphological characterization and comparative analysis of dsDNA and supramolecular assemblies [94–96]. Numerous studies have demonstrated that appropriate immobilization protocols on mica (i.e., using divalent cations, such as in this work) produce minimal alterations in the mechanical properties of dsDNA, such as Lc, Lp and in the promotion of aggregation [97–100]. As a control, dsDNA was imaged under the same air tapping-mode AFM conditions in the absence of PPL, exhibiting no aggregation or significant changes in Lc or Lp (Supplementary Information, Figure S8 and Table S2). Therefore, the aggregation observed in the AFM images can be attributed to the presence of PPL. Quantitative analysis of the AFM micrographs revealed a significant decrease of 67% in DNA occupancy area, confirming the formation of PPL/dsDNA frameworks (Fig. 4b-e). These AFM observations are consistent with the stick-release events observed during the extension of single dsDNA fragment/molecule (Fig. 4A) and point to a dual and concentration-dependent mechanism by which PPL alters the dsDNA structure. At low concentrations, PPL interacts with the dsDNA by an intercalation-like behavior, binding cooperatively along the double helix. At higher concentrations, it induced dsDNA compaction, visible as aggregates. This suggests that the interaction may actively facilitate the onset of dsDNA condensation, revealing an unrecognized functional versatility of PPL by modulating nucleic acid structure. As PPL has only 1 positive charge; we hypothesize that the presence of the two condensed aromatic rings within PPL could promote the observed behavior and support the compaction. The results obtained in this study point to a mechanistic scenario in which PPL shows an intercalation-like behavior within the dsDNA, subsequently promoting their compaction. Although previous studies suggested that PPL may preferentially interact through a groove binding mechanism [101], our analyses do not rule out this interaction pathway. It is plausible that an initial recognition within the major groove could constitute a step in the overall process and facilitate the transition toward more compact DNA states. These observations from AFM studies provide direct evidence of PPL-induced dsDNA compaction at a unimolecular level, consistent with the single-molecule results and confirming the formation of condensed dsDNA structures at increasing PPL concentrations. To the best of our knowledge, this is the first work describing intercalation-like behavior and significant compaction of dsDNA by using biophysical (optical tweezers) and microscopical (AFM) methodologies using a single dsDNA molecule. There exists only one work describing this dual mechanism using a bicationic drug (MTX) [76] and analyzing compaction by optical tweezers at very low stretching-relaxing forces (up to 2.5 pN). In addition, our demonstration includes a monocationic drug and high forces (until 140 pN), thus providing significant reliability. These results strongly promote further analyses in order to identify, at a multimolecular dsDNA level, the formation of larger structures using the monocationic PPL with potential uses in biotechnology, pharmacology, and drug delivery.

Multimolecular dsDNA Analysis Upon Interaction with Propranolol

Following the dsDNA single-molecule analysis, we proceeded to investigate PPL-dsDNA massive interactions at a multimolecular dsDNA scale. This approach allows us to assess whether the effects observed with single dsDNA molecule allow the formation of larger structures in a more concentrated regime. To this end, gel electrophoretic mobility assays (EMSA), a well-established and reliable technique for evaluating ligand-dsDNA complex formation, were conducted [102–104]. In this method, reductions in electrophoretic mobility and/or increase in the intensity of the band corresponding to non-migrating molecules are indicative of strong and massive intermolecular attractive interactions, typically reflecting the formation of larger structures. Figure 5A-B shows the electrophoretic migration of dsDNA (> 10 kbp, 0.06 mM) incubated with increasing concentrations of PPL (between 0.006 and 100 mM). As the PPL concentration increased, a small decrease in dsDNA migration was observed up to 4 mM PPL. Above this concentration, a pronounced aggregation leading to the complete arrest of migration is observed. These results suggest that PPL promotes the formation of higher-order drug-dsDNA structures. There exists evidence that a bicationic and polyaromatic molecule (e.g.: polypyridyl) is able to provide similar results in EMSA; the authors explained this phenomenon due to the presence of condensed dsDNA molecules in the band corresponding to non-migrating molecules [103]. This behavior supports the hypothesis that PPL-dsDNA interactions can drive condensation into larger structures, consistent with the compaction observed using single dsDNA molecule, analyzed by optical tweezers and AFM.

Fig. 5.

(A) Agarose gel of dsDNA migration in the presence of increasing concentrations of PPL (between 0.006-100 mM). (B) Quantification of relative dsDNA migration. (C) UV-vis spectra (between 250 and 350 nm) of PPL (7.5 × 10⁻⁵ M), in the absence (a) and in the presence of dsDNA [1 × 10^− 6 M (b), 5 × 10^− 6 M (c), 1 × 10^− 5 M (d), 2.5 × 10^− 5 M (e), 5 × 10^− 5 M (f), 2 × 10^− 4 M (g), 3 × 10^− 4 M (h)]. The inset shows the fitting of the data to the Wolfe-Shimer equation (mean ± S.D., n = 3) (C)

UV-vis spectroscopy was used to follow the spectroscopic changes in PPL upon interaction with dsDNA and to determine the apparent association constant between the components. In all experiments, the concentration of PPL was kept constant (7.5 × 10⁻⁵ M) while the dsDNA concentration varied from 0 to 5 × 10⁻⁴ M. As shown in Fig. 5C, increasing dsDNA concentration altered the chemical environment of PPL, as evidenced by hypochromic and bathochromic shifts in the absorption band of PPL at 295 nm. These changes are characteristic of intercalation events; in addition, the reduction in the absorption (up to 40% in our case) reflects the electronic interaction between the drug and DNA bases [105–107]. Simultaneously, the shift (up to 4 nm) is indicative of the presence of aromatic interactions between the aromatic rings of PPL and dsDNA base pairs. By fitting the Wolfe-Shimer equation (see Eq. 3 in Material and Methods) to the absorbance values at 295 nm as a function of dsDNA concentrations, we calculated a K_b of 1.67 ± 0.41 × 10⁶ M^-1 for the PPL-dsDNA complexes (see inset in Fig. 5C). This value is consistent with association constants reported for other intercalating molecules with values ranging between 10⁵ and 10⁶ M^-1 [105–107]. These UV-vis results provide quantitative confirmation of strong binding affinity between PPL and dsDNA, supporting the intercalation-like interaction mechanism proposed from the biophysical and microscopic analyses.

In order to study the development of NPs in this study, the concentration regime was increased compared with the multimolecular dsDNA experiments (EMSA and UV-vis). We explored PPL/dsDNA molar ratios, keeping the concentration of dsDNA constant at 1.4 × 10⁻³ M while systematically increasing the concentration of PPL until a significant increase in the absorbance (650 nm) was observed. The use of millimolar dsDNA concentration was intentional, based on previous studies involving synthetic polymers with physicochemical properties similar to DNA (i.e., aromatic and negative charged), that evidenced NPs formation at concentrations around 1 × 10⁻³ M when mixing with cationic hydrophilic drugs [17–19]. At definite concentrations of the polymer, smooth transition from transparent complexes to NPs, microparticles, and macroprecipitates is typically observed when titrating charged aromatic polymer solutions in the presence of aromatic counterions at increasing concentrations [19–22]. The obtained turbidimetric results are indicative of stable colloidal structures at PPL/dsDNA molar ratios between 10 and 14, as evidenced by the increased and stable absorbance values at 650 nm (coming from scattered light, Fig. 6A). At lower PPL/dsDNA molar ratios (0.1-8), the medium remained transparent, indicating the presence of soluble complexes. In contrast, ratios above 14 led to the formation of aggregates and massive precipitation. In order to validate the critical role of the aromatic rings in the polymeric component (dsDNA), we developed a control experiment using the non-aromatic and strongly polyanionic polymer polyvinylsulfonate (PVS). As evidenced in Fig. 6A, no turbidity appeared at the tested concentrations of PPL and PVS. Nanoparticle tracking analysis (NTA) revealed the presence of NPs and whose concentration (NPs/mL) strongly depended on the PPL/dsDNA charge ratio, increasing by nearly two orders of magnitude (from 7.1 × 10⁹ to 5.2 × 10¹¹ NPs/mL, Fig. 6A) at increasing the PPL concentration. These values are commonly observed for other NPs developed after mixing two aqueous solutions [19–21].

Fig. 6.

(A) Turbidimetry of PPL/dsDNA formulations at molar ratios between 0.1 to 22 (rhombuses: dsDNA, squares: PVS) and concentration of PPL/DNA NPs (NPs/mL, evidenced by nanoparticle tracking analysis, NTA). (B) Hydrodynamic diameter (bars) and zeta potential (squares) of the PPL/dsDNA formulations at molar ratios between 2 to 20. (C) STEM micrograph (150.000X) of the PPL/dsDNA NPs in molar ratio 10

Dynamic light scattering (DLS) showing NPs with hydrodynamic diameters ranging from 120 to 244 nm, low polydispersity indexes (PDI = 0.19–0.30) and zeta potential between − 25 to -35 mV (Fig. 6B). Interestingly, despite the use of high PPL/dsDNA molar ratios (10–14), the resulting NPs consistently exhibited a negative zeta potential (-25 to -35 mV). Notably, the PPL concentrations at which NPs formation occurs (1.44 × 10^-2-2.02 × 10^{− 2} M) are below both the solubility limit (2.5 × 10^-1M [108]), and the critical aggregation concentration (CAC) of PPL (≈ 1 × 10^{− 1} M [108–110]). This observation indicates that the NPs result from specific PPL and dsDNA interactions, rather than from PPL self-aggregation. In addition, DLS measurements showed that nanoparticle size increased from 120 to 240 nm as the PPL/dsDNA ratio increased. When considered together, the increased NPs concentration (Fig. 6A) and the particle size (Fig. 6B), this indicate that an increasing in the PPL concentration promotes nucleation and growth processes. Scanning transmission electron microscopy (STEM) provided insights into particle morphology, revealing the presence of structures in the size range presented by DLS (Fig. 6C). These results show that further increasing the PPL/dsDNA molar ratios leads to the formation of reproducible and stable NPs, reflecting a concentration-dependent nucleation and growth process. This confirms the interaction between PPL and dsDNA can generate nanoparticulate structures under the experimental conditions studied.

The resulting NPs exhibit favorable PPL encapsulation parameters, with drug association efficiencies ranging from 64 to 77% (PPL/dsDNA 10: 77.2%, PPL/dsDNA 12: 75.5%, PPL/dsDNA 14: 64.3%) and drug loading in the range of 59–72% (PPL/dsDNA 10: 71.8%, PPL/dsDNA 12: 70.3%, PPL/dsDNA 14: 58.5%). The AE and DL values were analyzed to relate the presence of PPL and dsDNA in each nanoparticle and their influence in the total number of NPs. As the PPL/dsDNA ratio increased, the incorporation of PPL (from 9.4 × 10⁸ to 1.5 × 10⁷ molecules) and dsDNA (from 1.2 × 10⁸ to 1.6 × 10⁶ nucleotides) per nanoparticle decreased while increasing the number of formed NPs (from 7.1 × 10⁹ to 5.2 × 10¹¹ NPs/mL, Fig. 6A), reflecting a dynamic and productive performance. The high encapsulation parameters obtained in our investigation are consistent with those reported for other systems developed using HALMD and synthetic polymers that share key physicochemical features with dsDNA: negatively charged groups and aromatic moieties capable of establishing complementary electrostatic and aromatic interactions. In those studies, AE above 82% and DL values up to 67% were obtained [19–22, 54–56]. This is explained by the fact that the HALMD is not only a drug to be carried, but a fundamental structural constituent of the NPs, requiring only two components for its formation and increasing the DL values compared with other formulations where surfactants, cross-linkers and/or additional excipients provide lower DL. In this kind of system, centrifugal filtration is preferred over dialysis to isolate the non-encapsulated drug and calculate AE and DL. Due to the dynamic structuration of dsDNA and PPL to produce the NPs, elution during the dialysis process should promote the concomitant release of free and encapsulated PPL, as observed for other systems [19, 21]. On the contrary, centrifugal ultrafiltration is a widely validated method for determining AE and DL for this kind of systems, enabling comparison with other studies [19–21, 54–56]. This technique provides fast separation kinetics and avoids NP dilution. It is not ruled out that AE can be overestimated using centrifugal filtration, due to NPs concentration during filtration. However, investigations focused on aggregation mechanisms that involve aromatic interactions between complementary charged species have demonstrated that molecular deprotonation can accompany aromatic stacking, promoting the formation of more hydrophobic and stabilized aggregates [111–112]. This phenomenon provides a coherent explanation for the obtained results in this work: PPL, initially intercalated within the dsDNA structure, may undergo partial deprotonation that enhances aromatic interactions, promoting its accumulation within a hydrophobic microenvironment. This process would effectively generate a nucleation core inside the NPs, thereby supporting the high encapsulation parameters measured for the PPL/DNA NPs. This evidence suggests a hierarchical and unified mechanism that integrates intercalation-like behavior, controlled molecular reorganization, and subsequent DNA compaction.

To gain a deeper understanding into the role of the aromatic interactions in the ability of monocationic HALMD to interact with dsDNA and promote nanoparticle formation, we selected two additional HALMD (metoprolol and doxorubicin) with different aromaticity degree. Their combinations with dsDNA and the resulting systems were characterized in terms of hydrodynamic diameter, PDI, zeta potential, number of particles, AE and DL. Considering that the main difference among these molecules is the number of cyclic aromatic rings (PPL = 2 rings, metoprolol = 1 ring, and doxorubicin = 3 rings), the results are expected to depend on the capacity of these aromatic rings to interact with the aromatic bases of dsDNA. The results evidence that, under identical dsDNA concentrations, the polycyclic doxorubicin formed nanostructures at significantly lower HALMD/dsDNA molar ratios (between 0.5 and 0.6) compared with the bicyclic PPL (molar ratios between 10 and 14). As observed, in the case of the monocyclic metoprolol, it was not possible to obtain NPs although wide HALMD/dsDNA molar ratios were tested (0.5 to 15). Notably, once nanoparticle formation is achieved, although interacting with significantly different HALMD (PPL and doxorubicin), they exhibited similar colloidal characteristics (i.e., hydrodynamic diameter, PDI and zeta potential, concentrations in terms of NPs/mL) and similar encapsulation parameters (AE and DL) (see supplementary information, Table S3). This suggests that the characteristics of the obtained NPs depend on the elaboration strategy and in the dsDNA characteristics. Interestingly, the absence of NPs formation when using the highly anionic and aliphatic poly(vinyl sulfonate) (PVS) instead of aromatic dsDNA confirms the critical role of the aromatic components in the formulations to obtain NPs. This is the first work describing that a non-toxic, monocationic, and bicyclic HALMD (PPL) is able to form NPs after interacting in aqueous medium with dsDNA. While the ability of DOX to interact with DNA and affect both cancerous and non-cancerous cells is widely described, this work opens new possibilities for testing existing HALMD (such as PPL) and new synthetized monocationic bicyclic HALMD to interact with polynucleotides and/or for obtaining NPs to create new medicines.

To further evaluate the performance of the obtained formulations, stability studies of the PPL/dsDNA NPs (molar ratio 10) were carried out by exploring changes in the hydrodynamic diameter and zeta potential under different conditions, including temperature (20–50 °C), pH (2–10), and storage time (until 12 weeks). Figure 7 shows that NPs exhibited good thermal stability across the tested temperature range (Fig. 7A). In terms of pH, a clear trend to destabilization was observed under extreme acidic (pH 2) and alkaline (pH 10) conditions, likely associated with the high ionic strength at those pH values (Fig. 7B) [21]. To evaluate the reversibility of aggregation and precipitation under extreme pH conditions, we developed PPL/dsDNA formulations (molar ratio 10, pH ~ 6, ~235 nm, PDI ≤ 0.1), submitted to pH 10 (resulting in aggregation and precipitation), and restoring the pH to 6. In every case, larger structures appeared (≥ 1000 nm and precipitates, see supplementary information, Figure S9), indicating that the process to recover the original NPs after extreme pH changes is irreversible. Additionally, an increase in hydrodynamic diameter was detected until the second week, from ≈ 120 nm to ≈ 180 nm, then remained stable up to 12 weeks (Fig. 7C). Importantly, PDI values stayed below 0.3 and the zeta potential below − 28 mV, indicating that the system maintained a stable colloidal state. This size increase does not appear to be consistent with particle aggregation or coalescence, which would typically result in increased polydispersity and reduction in the absolute value of the zeta potential. In consequence, it could be attributed to a structural reorganization and water inclusion into the PPL-dsDNA nanostructure [113].

Fig. 7.

Stability study of PPL/dsDNA NPs (molar ratio 10) observed through the hydrodynamic diameter (bars, numbers above the bars correspond to the polydispersity index, PDI) and zeta potential (lines) as a function of temperature (A), pH (B) and time (C). (Mean ± SD; n = 3)

To support with a schematic representation linking the mechanistic insights in this work, from unimolecular to multimolecular scale, Fig. 8 is presented.

Fig. 8.

Proposed mechanism for the progressive interactions between PPL and dsDNA. dsDNA and PPL cooperatively bind (A) through electrostatic and aromatic interactions, inducing an intercalation-like association that increases dsDNA flexibility (B). As PPL concentration increases, dsDNA progressively condenses through intermolecular interactions (C). Higher PPL concentration ultimately leads to the formation of compact nanoparticles (~ 120–240 nm) composed of condensed dsDNA and PPL without the need for additional excipients (D)

Conclusions

This study uncovers a hierarchical and multimodal interaction between the monocationic aromatic drug propranolol (PPL) and dsDNA, driven by combined electrostatic and aromatic forces. PPL binds DNA in a concentration-dependent manner, exhibiting intercalation-like and/or mixed binding behavior at low concentration and acting as a condensing agent at higher concentrations, promoting the formation of DNA frameworks and nanoparticle assembly. Interestingly, a clear stepwise progression is evidenced from unimolecular (optical tweezers, AFM) to multimolecular dsDNA levels (EMSA, UV) and until obtaining reproducible and stable NPs. These findings broaden the current paradigm of small molecule-DNA interactions, which has been largely restricted to classical intercalators and multivalent condensing agents. Importantly, we demonstrate that this binding mechanism can be exploited to provide reproducible and high yielding drug-DNA NPs without auxiliary excipients. Beyond PPL, the provided methodological strategy may be extended to other aromatic drugs, positioning DNA as a versatile building block and ultimately for the development of nucleic acid-based nanomedicines.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

María Gabriela Villamizar-Sarmiento: Conceptualization, Formal analysis, Methodology, Validation, Writing - original draft. Romina Muñoz Buzeta: Methodology, Formal analysis. Rodrigo Rivera: Conceptualization, Methodology, Formal analysis. Francisco Melo: Methodology, Formal analysis. Juan Ruso: Methodology, Formal analysis. Ignacio Moreno-Villoslada: Conceptualization, Methodology, Formal analysis, Writing-review & editing. Felipe Oyarzun-Ampuero: Conceptualization, Methodology, Formal analysis, Writing-review & editing. Mauricio Báez: Conceptualization, Methodology, Formal analysis, Writing-review & editing.

Funding

This work was supported by FONDECYT 3210549, FONDECYT 11251306 (M.G.V-S), FONDECYT 1250392 (I.M-V.), FONDECYT 1241624 (F.A.O-A.), FONDECYT 1231276 (M.B.), FONDEQUIP EQM180114, FONDEQUIP EQM160157 (F A. O-A), FONDEQUIP EQM230061 (F.M.), ANID/ACT240058 (F A. O-A), FONDAP 15130011 and 1523A0008 (F A. O-A), Proyecto de Continuidad Transitoria Centros de Excelencia Universidad de Chile (2026-2027), ANID/ACT250073 (F A.O-A) and Dicyt Usach for partial support through project 042431MH_Ayudante (F.M.).

Data availability

The authors declare that the data supporting the findings of this study are available within the paper and its Supplementary Information files. Should any raw data files be needed in another format they are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.