Gallic Acid–Triethylene Glycol Aptadendrimers Synthesis, Biophysical Characterization and Cellular Evaluation

Herein, we describe the synthesis of an aptadendrimer by covalent bioconjugation of a gallic acid–triethylene glycol (GATG) dendrimer with the G-quadruplex (G4) AT11 aptamer (a modified version of AS1411) at the surface. We evaluated the loading and interaction of an acridine orange ligand, termed C8, that acts as an anticancer drug and binder/stabilizer of the G4 structure of AT11. Dynamic light scattering experiments demonstrated that the aptadendrimer was approximately 3.1 nm in diameter. Both steady-state and time-resolved fluorescence anisotropy evidenced the interaction between the aptadendrimer and C8. Additionally, we demonstrated that the iodine atom of the C8 ligand acts as an effective intramolecular quencher in solution, while upon complexation with the aptadendrimer, it adopts a more extended conformation. Docking studies support this conclusion. Release experiments show a delivery of C8 after 4 h. The aptadendrimers tend to localize in the cytoplasm of various cell lines studied as demonstrated by confocal microscopy. The internalization of the aptadendrimers is not nucleolin-mediated or by passive diffusion, but via endocytosis. MTT studies with prostate cancer cells and non-malignant cells evidenced high cytotoxicity mainly due to the C8 ligand. The rapid internalization of the aptadendrimers and the fluorescence properties make them attractive for the development of potential nanocarriers.


Introduction
Cancer is a major public health problem that, unfortunately, is growing worldwide. Prostate cancer, like any other cancer, is a cellular disease caused by the alteration of specific genes or changes in cellular signals that lead to abnormal cell growth and proliferation

Synthesis of 2[G4]-N 3 Dendrimer
The repeating unit of GATG dendrimers ( Figure 1A) was synthesized according to [20] using an efficient, safe (green chemistry principles), and cost-effective synthetic route of 4 steps. This repeating unit was employed for the divergent synthesis of dendrimers as previously reported [21]. Briefly, the GATG repeating unit was firstly treated with hydroxybenzotriazole (HOBt) and 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC) to promote the coupling of two repeating units to a 2,2 -(ethylenedioxy)bis(ethylamine) linker and produce 2[G1]-N 3 , a dendrimer of first generation with 6 terminal azides. After the reduction of the azides with triphenylphosphine (Ph 3 P), via the Staudinger reaction, the resulting terminal amines were again reacted with a second layer of repeating units, and these steps were repeated until the dendrimer achieved the fourth-generation 2[G4]-N 3 with 162 terminal azides ( Figure 1B). The bioconjugation process of the aptamer to the 2[G4]-N 3 dendrimer was carried out in two phases. Both steps are described in detail in the Supporting Information. First, the AT11 aptamer, modified with an NH 2 group, was functionalized with a DBCO-PEG 13 -NHS ester ( Figure 2A). The long PEG 13 linker was selected to increase the distance between the dendrimer and the aptamer and, thus, to reduce possible electrostatic repulsions between aptamers. The dibenzocyclooctyne (DBCO) was chosen as a chemical handle for the subsequent bioconjugation step between the dendrimer and AT11-PEG 13 -DBCO by means of a strain-promoted azide-alkyne cycloaddition (SPAAC) [22][23][24] ( Figure 2B). The reaction progress of the SPAAC was monitored by polyacrylamide gel electrophoresis (PAGE). Finally, the aptadendrimer was labelled with a Cy5-DBCO fluorophore to allow its visualization in confocal microscopy experiments. ( Figure 2B). to promote the coupling of two repeating units to a 2,2′-(ethylenedioxy)bis(ethylamine) linker and produce 2[G1]-N3, a dendrimer of first generation with 6 terminal azides. After the reduction of the azides with triphenylphosphine (Ph3P), via the Staudinger reaction, the resulting terminal amines were again reacted with a second layer of repeating units, and these steps were repeated until the dendrimer achieved the fourth-generation 2[G4]-N3 with 162 terminal azides ( Figure 1B).

Aptadendrimer Synthesis: Surface Functionalization of 2[G4]-N3 with AT11 Aptamer and Fluorophore Labelling
The bioconjugation process of the aptamer to the 2[G4]-N3 dendrimer was carried out in two phases. Both steps are described in detail in the Supporting Information. First, the AT11 aptamer, modified with an NH2 group, was functionalized with a DBCO-PEG13-NHS ester (Figure 2A). The long PEG13 linker was selected to increase the distance between the dendrimer and the aptamer and, thus, to reduce possible electrostatic repulsions between aptamers. The dibenzocyclooctyne (DBCO) was chosen as a chemical handle for the subsequent bioconjugation step between the dendrimer and AT11-PEG13-DBCO by means of a strain-promoted azide-alkyne cycloaddition (SPAAC) [22][23][24] (Figure 2B). The reaction progress of the SPAAC was monitored by polyacrylamide gel electrophoresis (PAGE). Finally, the aptadendrimer was labelled with a Cy5-DBCO fluorophore to allow its visualization in confocal microscopy experiments. ( Figure 2B).

Aptadendrimer Characterization
Dynamic light scattering (DLS) measurements were performed on a Malvern Nano ZS (Malvern Instruments, Malvern, UK), operating at 633 nm with a 173° scattering angle, at 25 °C. DLS mean diameters were obtained from the volume particle size distribution provided by the Malvern Zetasizer Software. DLS histograms were obtained from the volume particle size distributions. Z-potential values were obtained by laser Doppler anemometry (LDA), measuring the mean electrophoretic mobility (Malvern Zetasizer Nano ZS, Malvern Instruments).

Aptadendrimer Characterization
Dynamic light scattering (DLS) measurements were performed on a Malvern Nano ZS (Malvern Instruments, Malvern, UK), operating at 633 nm with a 173 • scattering angle, at 25 • C. DLS mean diameters were obtained from the volume particle size distribution provided by the Malvern Zetasizer Software. DLS histograms were obtained from the volume particle size distributions. Z-potential values were obtained by laser Doppler anemometry (LDA), measuring the mean electrophoretic mobility (Malvern Zetasizer Nano ZS, Malvern Instruments).

Molecular Docking
The three-dimensional structure of the AT11 aptamer was downloaded from the PDB database (https://www.rcsb.org/, accessed on 5 February 2022; PDB entry 2N3M). The dock preparation (DockPrep tool), the assignment of Gasteiger charges and polar hydrogens for G4 nucleic acid were performed using Chimera 1. 16. Docking experiments were carried out by the AutoDock 4.2 program using the Lamarckian genetic algorithm (25 runs in an initial population of 150 random individuals, a maximum number of evaluations of 2.5 × 10 6 , rate of mutation and crossover of 0.02 and 0.8 and, finally, elitism value of 1). Using a box (125 × 125 × 125 Å along the x, y, and z axes) with a grid spacing of 0.6 Å, the AT11 structure was centered and left rigid while the ligand was allowed full flexibility. Additionally, the same protocol was applied to understand the interaction between the C 8 ligand and the GATG repeating unit.

Loading Efficiency Determination
The loading efficiency of the C 8 ligand by the aptadendrimers was determined, based on the amount of C 8 entrapped, using Equation (1).
Briefly, the aptadendrimer (10 µM of aptamer) was incubated with C 8 (1 µM) under constant stirring using a Hula Mixer Sample Mixer (Thermo Fisher Scientific, Waltham, MA, USA). The measurements of the amount of C 8 were based on the fluorescence intensity of the ligand in the supernatant and following centrifugation at 3000× g for 15 min using a Vivaspin 2 kDa (Sartorius, Gottingen, Germany) in the flowthrough. For fluorescence measurements, a high-precision 3 mm × 3 mm cell (Hellma Analytics, Jena, Germany) was used in a Horiba FluoroMax4 fluorometer (Kyoto, Japan) using suitable excitation and emission wavelengths (492 and 526 nm, respectively). The wavelengths were determined after excitation and emission spectra in a fluorometer.

Drug Release
The C 8 release profile of aptadendrimer was studied applying the method reported by Carvalho et al. [25]. Using a Slide-A-Lyzer™ MINI Dialysis Device with a 3.5 kDa of molecular cut-off (Thermo Fisher Scientific, USA), 100 µL of aptadendrimer/C 8 complex was added and dialysis was performed in 1 mL of the final volume of buffer solution under constant stirring using an HulaMixer Sample Mixer (Thermo Fisher Scientific, USA). Samples were collected from the bottom tube (100 µL) along the time window, and buffer was added to the dialysis medium in order to maintain constant volume during the experiment. The released C 8 content was quantified by measuring the fluorescence of the samples using a high-precision 3 mm × 3 mm cell (Hellma Analytics, Germany) in a Horiba FluoroMax4 fluorometer (Kyoto, Japan) using suitable excitation and emission wavelengths (492 and 526 nm, respectively).

Steady-State Fluorescence Intensity and Anisotropy Measurements
Steady-state fluorescence experiments were performed in a Horiba Jobin Yvon Fluorolog 3-22 (Tokyo, Japan) spectrofluorometer using 0.5 cm × 0.5 cm quartz cuvettes (Hellma Analytics, Germany) at room temperature. The emission spectra of 0.5 µM C 8 in buffer and upon incubation with increasing aptadendrimer concentrations were recorded with excitation at 488 nm with 2 and 4 nm slits in the excitation and emission, respectively. For fluorescence anisotropy experiments, the samples were excited at 488 nm and the polarized emission was recorded at 520 nm using 5 and 10 nm slits in the excitation and emission, respectively. The steady-state fluorescence anisotropy, r , was calculated using Equation (2) as previously described [26]: where I VV and I V H are the intensities of the vertically and horizontally polarized fluorescence emission upon excitation with vertically polarized light, respectively. The G factor (G = I HV /I HH , with horizontal excitation components) accounts for the monochromator transmission efficiency to the polarization of the light. Data are shown as mean ± standard deviation of 10 measurements.

Time-Resolved Fluorescence Intensity and Anisotropy Measurements
Time-resolved fluorescence experiments were carried out by the single-photon-timing technique as previously described [27,28]. The samples were excited at 488 nm using a BDS-SM-488FBE pulsed picosecond diode laser from Becker & Hickl (Berlin, Germany), and the fluorescence emission was recorded at 520 nm. The fluorescence intensity decays, I(t), were collected with the emission polarizer set at the magical angle regarding the vertically polarized excitation beam. For anisotropy decays, the parallel and perpendicular polarized components of the fluorescence-I VV (t) and I V H (t), respectively-to the plane of the excitation beam were alternatively recorded. The instrument response function (IRF) was obtained from the excitation light scattered by a Ludox solution (silica, colloidal water solution, Sigma-Aldrich).
The decays were analyzed using the TRFA Data Processer Advanced (version 1.4) from the Scientific Software Technologies Centre (Belarusian State University, Minsk, Belarus) as previously detailed [27,28]. The reduced χ2 value (<1.3) and the random distribution of weighted residuals/autocorrelation plots were used to evaluate the goodness of the analysis.
Fluorescence intensity decays were analyzed using a sum of discrete exponential terms [26]: here α i and τ i are the pre-exponential (amplitude) and the lifetime of the ith decay component of fluorescence, respectively. The amplitude-weighted mean fluorescence lifetime, τ , was then calculated according to: The fluorescence anisotropy decays, r(t), were analyzed by a sum of discrete exponential terms as previously described [26]: where β i and φ i stand for the normalized amplitude and the rotational correlation time of the ith anisotropy decay component, respectively.

Cytotoxic Studies of Aptadendrimers
The cytotoxicity of the aptadendrimer, C 8 -loaded aptadendrimer, AT11 aptamer, and AT11-C 8 was evaluated in PNT1A, PC-3, and DU-145 cell lines by the MTT assay. The cells were seeded in 96-well plates (2 × 10 4 cells/mL) and after 24 h for cell adhesion, they were treated with different concentrations of aptadendrimers and complexes for 72 h. In IC 50 experiments, the C 8 ligand was successively diluted (from 3.12 µM to 0.024 µM). The wells containing untreated cells were used as control. Then, fresh media containing 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide salt (MTT; Sigma-Aldrich, St. Louis, MO, USA) was added, followed by further incubation at 37 • C in a humidified atmosphere containing 5% of CO 2 for 4 h. Formazan crystals were dissolved in DMSO, and their absorbance was recorded at 570 nm using a BioRad ×Mark™ microplate reader (BioRad, Hercules, CA, USA). The cell viability percentages of different treatments were calculated by considering the absorbance of the control as 100% viability. Cell viability was expressed as mean ± SD from at least three different plates with each condition tested in quadruplicate wells. GraphPad Prism 8 (San Diego, CA, USA) was used for data treatment.

Results and Discussion
For the preparation of the aptadendrimer, we selected a fourth-generation dendrimer of the GATG family, herein referred to as 2[G4]-N 3 ( Figure 1B), previously described by some of us [21]. GATG dendrimers are composed of a gallic acid core, responsible for the multivalency, and long triethylene glycol spacer arms that give flexibility to the dendritic structure. They have been exploited for the preparation of drug [29,30] and gene delivery [31] systems, and the construction of monodisperse nanotools to modulate the complex mechanisms governing multivalent interactions [32][33][34][35]. 2[G4]-N 3 , which is prepared divergently from a triethylene glycol diamine core and the GATG repeating unit ( Figure 1A), carries 162 terminal azides that we have used for the bioconjugation of the AT11 aptamer by means of the strain-promoted azide-alkyne cycloaddition (SPAAC) [22][23][24].
With this aim, AT11-NH 2 (a version of AT11 with a terminal amino group) was first functionalized with a dibenzocyclooctyne (DBCO) using an active ester equipped with a long PEG 13 linker (DBCO-PEG 13 -NHS, DMSO, 25 • C). The long PEG-based linker was chosen over a long carbon spacer to increase the distance between the dendrimer and the aptamer and consequently, the flexibility of the aptadendrimer (Figure 2A). Additionally, the PEG polymer coating has the advantage improving the biophysical and chemical properties of nanoparticles (such as high hydrophilicity, spatial repulsion, and electrical neutrality) that resulted in higher biocompatibility and blood circulation half-life [36]. Additionally, some therapeutical approaches using PEGylation have already been approved by the Food and Drug Administration of the USA [36]. Despite these positive points, PEG conjugation can present some drawbacks and limitations as adverse side effects caused by the polymer itself or by side products formed during synthesis, unexpected changes in the pharmacokinetic behavior and non-biodegradability of PEG [37]. After purification by dialysis, AT11-PEG 13 -DBCO was obtained in 89% yield. The complete functionalization of AT11 with DBCO was confirmed by UV-Vis ( Figure S1) based on their characteristic absorptions at 280 and 309 nm, as described in the Supporting Information.
The subsequent SPAAC bioconjugation of AT11-PEG 13 -DBCO to 2[G4]-N 3 was performed in DMSO at 37 • C ( Figure 2B). Monitoring of the reaction progress was carried out by PAGE until no free aptamer was detected ( Figure S2). Several loadings of AT11 aptamer per dendrimer were tested, namely 8, 17, and 34. Since, for the lower loadings, some aptadendrimer aggregation was observed, a loading of 34 was defined for this initial exploration of the aptadendrimer properties and activity. After completion of the AT11 bioconjugation, the aptadendrimer was labelled with a cyanine-5 (Cy5) fluorophore also using a SPAAC. After purification by ultrafiltration (Amicon) to remove any potential unreacted Cy5, 2[G4]-(N 3 ) 127 /(AT11) 34 /(Cy5) 1 was obtained in 97% yield. In this conjugate, a Cy5 loading of 1 was determined by UV-Vis ( Figure S3) and checked by fluorescence spectroscopy (Figure S4), as described in the Supporting Information.
Once the synthesis of the aptadendrimer had been performed, we checked the G4 formation of AT11 conjugated on the dendrimer surface by CD spectroscopy (Figure S5). The results showed a characteristic profile of parallel G4 topology (positive and negative bands at 260 and 240 nm, respectively) as previously reported for this aptamer [11]. In addition, it was possible to observe an extra band around 280 nm, a chirality signal that can be explained by the conjugation of AT11 with the PEG 13 spacer and the dendrimer. This band is attributed to the π-π* transition of the carbonyl groups of the GATG repeating unit that was embedded in a chiral entity becoming CD active [38,39] or can be explained by conjugation with the dendrimer that alters the secondary structure of AT11.
To evaluate if the aptadendrimer can benefit from the enhanced permeability and retention (EPR) effect, we performed a dynamic light scattering (DLS) assay to evaluate its size. The aptadendrimer exhibited a mean particle size of 3.1 nm ( Figure S6), which is in accordance with the reference values [40][41][42]. In addition, this nanosize is suitable for cellular internalization (≤100 nm). Additionally, the surface charge of aptadendrimer was measured. A zeta potential of −22.2 mV was revealed for the aptadendrimer, an anionic character with a value in line with others previously described for G4 aptamer-based nanoparticles [43] and the charge adopted by G4 vs. duplex DNA [44]. The negative zeta potential of the aptadendrimer confirms the incorporation of AT11 (highly negative because of the phosphate groups) on the surface of the dendrimer, which is a neutral macromolecule (no charge on the surface of 2[G4]-N3). Not unexpectedly, the data obtained are in agreement with that for Au nanoparticles functionalized with aptamers [43].
Next, we evaluated the binding mode and interactions between the acridine orange ligand C 8 and AT11 aptamer by molecular docking. C 8 was chosen not only because of its promising anticancer effects [18,25,45,46], but also because it binds NCL-targeted aptamers and can potentially stabilize G4 structures. A previous report showed that interactions among the C 8 ligand and G4 structures are mainly end-stacking interactions between the planar acridine moiety and the top and bottom G-quartets [47], resulting in the stabilization of G4 as evidenced by the thermal stability increase [45,46,[48][49][50].
However, the free C 8 ligand has a high toxicity in both malignant and normal cell lines [25,43]. For the molecular docking experiments, we used the tridimensional structure of AT11 previously determined by Do et al. [11] and deposited in the Protein Data Bank (PDB entry 2N3M). As expected, the results depicted in Figure 3 showed that C 8 interacts via π-π stacking with the tetrad (G18-G21-G24; π-π interactions). This can be explained by the planar nature of the acridine moiety that favors these interactions and the protonated nitrogen atoms that are placed near the central carbonyl groups in the G-tetrads.
Additionally, it was observed the interaction of the ligand alkyl linker with the groove (G17-G27-T16-T28). Molecular docking simulations carried out between C 8 and the GATG repeating unit also confirmed contacts between the acridine scaffold and the gallic acid core of the dendritic structure ( Figure S7).
Following this, the aptadendrimer encapsulating C 8 was prepared and characterized. There were no significant changes in size and zeta potential between the unloaded and C 8 -loaded aptadendrimers, suggesting that the nanoparticle maintained the structure.
To analyze the loading efficiency of the aptadendrimer with C 8 , we incubated C 8 with the aptadendrimer and the fluorescence of free C 8 ( Figure S8) in the mixture was measured before and after centrifugation ( Figure S9) [25]. A percentage of C 8 -loaded aptadendrimer around 98% was revealed through Equation (1).  Following this, the aptadendrimer encapsulating C8 was prepared and characterized. There were no significant changes in size and zeta potential between the unloaded and C8-loaded aptadendrimers, suggesting that the nanoparticle maintained the structure.
To analyze the loading efficiency of the aptadendrimer with C8, we incubated C8 with the aptadendrimer and the fluorescence of free C8 ( Figure S8) in the mixture was measured before and after centrifugation ( Figure S9) [25]. A percentage of C8-loaded aptadendrimer around 98% was revealed through Equation 1.
The C8 association to aptadendrimer was also monitored by both steady-state and time-resolved fluorescence and anisotropy measurements. The variation in the C8 fluorescence intensity and the amplitude-weighted mean fluorescence lifetime, 〈τ〉, upon complexation with aptadendrimers are shown in Figure 4A,B, respectively. Notably, the fluorescence intensity increased more than 20-fold and the 〈τ〉 enhanced from 0.86 ns in the buffer to 4.05 ns upon C8 binding to the aptadendrimers, supporting that C8 fluorescence in the buffer is severely quenched. Considering the C8 structure ( Figure 1D), the iodine atom acts as an effective intramolecular quencher of the conjugated dye in an aqueous solution due to the heavy atom effect; upon its binding to the aptadendrimer, C8 must adopt a more extended conformation that prevents the iodine quencher effect. The C 8 association to aptadendrimer was also monitored by both steady-state and timeresolved fluorescence and anisotropy measurements. The variation in the C 8 fluorescence intensity and the amplitude-weighted mean fluorescence lifetime, τ , upon complexation with aptadendrimers are shown in Figure 4A,B, respectively. Notably, the fluorescence intensity increased more than 20-fold and the τ enhanced from 0.86 ns in the buffer to 4.05 ns upon C 8 binding to the aptadendrimers, supporting that C 8 fluorescence in the buffer is severely quenched. Considering the C 8 structure (Figure 1D), the iodine atom acts as an effective intramolecular quencher of the conjugated dye in an aqueous solution due to the heavy atom effect; upon its binding to the aptadendrimer, C 8 must adopt a more extended conformation that prevents the iodine quencher effect.  To further confirm the C8-aptadendrimer complex formation, fluorescence anisotropy experiments were also carried out. A relatively high steady-state fluorescence anisotropy in buffer was obtained for a small fluorophore (〈r〉 = 0.062 + 0.006), which is explained in part by the short fluorescence lifetime of C8 in an aqueous solution. Remarkably, upon increasing the aptadendrimer concentration, the anisotropy of C8 steadily aug- To further confirm the C 8 -aptadendrimer complex formation, fluorescence anisotropy experiments were also carried out. A relatively high steady-state fluorescence anisotropy in buffer was obtained for a small fluorophore ( r = 0.062 + 0.006), which is explained in part by the short fluorescence lifetime of C 8 in an aqueous solution. Remarkably, upon increasing the aptadendrimer concentration, the anisotropy of C 8 steadily augmented until attaining a plateau at r ~0.20 + 0.001 ( Figure 5A). These data reveal that the hydrodynamic volume of C 8 increased in the presence of the aptadendrimer, confirming its complexation and the slowing down of the overall rotational tumbling of the complex during its fluorescence lifetime. Moreover, time-resolved fluorescence anisotropy experiments further confirmed the C 8 complexation. Indeed, the fluorescence anisotropy decay of C 8 was greatly affected by the addition of the aptadendrimer ( Figure 5B). First, the fast rotational correlation time (φ 1 ) of C 8 was enhanced upon complexation towards 1.4 ns, which reflects the segmental motion of C 8 bound to the PEG 13 -AT11 in the dendrimer. Moreover, the longer rotational correlation time (φ 2 ), assigned to the overall rotational motion, increased significantly from 0.14 ns for C 8 in solution to 10-12 ns upon binding to the aptadendrimer. Altogether, our results clearly show that C 8 binds to the aptadendrimer. The release profile of C8 from the aptadendrimer complex ( Figure 6) was evaluated in 20 mM KPi (pH = 6.9) supplemented with 65 mM KCl for 48 h, with samples collected at different time points. The results evidenced that the release is fast and most noticeable after 4 h and it has a maximum cumulative release profile at 12 h. The release profile of C 8 from the aptadendrimer complex ( Figure 6) was evaluated in 20 mM KPi (pH = 6.9) supplemented with 65 mM KCl for 48 h, with samples collected at different time points. The results evidenced that the release is fast and most noticeable after 4 h and it has a maximum cumulative release profile at 12 h.
Next, we performed in vitro studies to access the cellular internalization and viability of the aptadendrimer in two prostate cancer cell lines (PC-3 and DU-145) and one nonmalignant cell line (PNT1A). The cell viability was determined by the MTT assay after incubation of the AT11, C 8 -AT11 complex, aptadendrimer, and C 8 -aptadendrimer complex for 3 days. The results are presented in Figure 7. Next, we performed in vitro studies to access the cellular internalization and viability of the aptadendrimer in two prostate cancer cell lines (PC-3 and DU-145) and one nonmalignant cell line (PNT1A). The cell viability was determined by the MTT assay after incubation of the AT11, C8-AT11 complex, aptadendrimer, and C8-aptadendrimer complex for 3 days. The results are presented in Figure 7.  Table S1.
Free AT11 elicited only a minor effect on the viability of both PC-3 and PNT1A cells (80% of mean viability) and 58% on the viability of DU-145 cells, in which DU-145 and PC-3 were both androgen-independent with moderate and high metastatic potential, respectively [51]. On the other hand, free C8 reduced the viability of all cell lines ( Figure  S10), showing a non-selective cytotoxic effect, as previously seen [43,49].
The aptadendrimer presented higher cytotoxicity than free AT11 in all cell lines. Indeed, incubation of the aptadendrimer (at the same concentration of AT11 as above) for 3 days dramatically increased its toxicity by 3.56 times for PNT1A, and 2.23 and 2.25 times for PC-3 and DU-145 cells (ratios among the viabilities of free AT11 and aptadendrimer), respectively (Figure 7). The cell viability decreased to ca. 23% for PNT1A, 34% for PC-3, and 26% for DU-145 cells. Regarding C8-aptadendrimer, after 3 days, it caused a ca. 15%, 16%, and 22% reduction in viability of PNT1A, PC-3, and DU-145 cells, respectively. The similar toxicity for both cancer and non-malignant cells points to a lack of selectivity for the aptadendrimer to deliver C8 to the target cancer cells.
In order to evaluate the uptake, we treated the cancerous and the non-malignant cell  Table S1.
Free AT11 elicited only a minor effect on the viability of both PC-3 and PNT1A cells (80% of mean viability) and 58% on the viability of DU-145 cells, in which DU-145 and PC-3 were both androgen-independent with moderate and high metastatic potential, respectively [51]. On the other hand, free C 8 reduced the viability of all cell lines (Figure S10), showing a non-selective cytotoxic effect, as previously seen [43,49].
The aptadendrimer presented higher cytotoxicity than free AT11 in all cell lines. Indeed, incubation of the aptadendrimer (at the same concentration of AT11 as above) for 3 days dramatically increased its toxicity by 3.56 times for PNT1A, and 2.23 and 2.25 times for PC-3 and DU-145 cells (ratios among the viabilities of free AT11 and aptadendrimer), respectively (Figure 7). The cell viability decreased to ca. 23% for PNT1A, 34% for PC-3, and 26% for DU-145 cells. Regarding C 8 -aptadendrimer, after 3 days, it caused a ca. 15%, 16%, and 22% reduction in viability of PNT1A, PC-3, and DU-145 cells, respectively. The similar toxicity for both cancer and non-malignant cells points to a lack of selectivity for the aptadendrimer to deliver C 8 to the target cancer cells.
In order to evaluate the uptake, we treated the cancerous and the non-malignant cell lines with the aptadendrimer and C 8 -aptadendrimer complex labelled with Cy5 and recorded images using a confocal laser scanning microscope (CLSM). Firstly, we performed a live imaging of the aptadendrimer over 1 h in PC-3 and PNT1A cells (Figures S11 and  S12). The results evidenced a fast internalization of the aptadendrimer. It was not possible to observe colocalization of the aptadendrimer with an NCL antibody in both cell lines, suggesting that internalization is not actively mediated by cell surface NCL. In a previous report, we showed that free Cy5-AT11 and C 8 -AT11 were able to colocalize NCL in HeLa cancer cells and that the complex remained stable during cell trafficking, while free C 8 localized with nucleoli [18]. To clarify the cellular uptake mechanism of the aptadendrimer, we incubated the cells with aptadendrimer at a low temperature (4 • C). At lower temperatures, cells reduced their metabolism, and the membrane increased rigidity and blocked energy-dependent uptake and passive diffusion [52]. The passive diffusion mechanism is predominantly limited to small, uncharged molecules that travel down concentration gradients. Our results ( Figure S13) evidenced the absence of an aptadendrimer fluorescence signal for both cell lines at low temperatures. Of note, incubation at 4 • C also influenced Hoechst migration and, consequently, the labeling of nuclei. This result demonstrates that the uptake mechanism is not based on passive transport but rather on another pathway. Usually, the cellular uptake of nanocarriers requires endocytosis [53]. In the specific case of dendrimer nanocarriers, it is extensively described that internalization can occur through different endocytic routes [54][55][56][57][58]. We propose that in this case, internalization proceeds via endocytosis through an NCL-independent way. We can infer that the uptake mechanism can be intrinsically related to the high toxicity of aptadendrimers in both cell lines. Previous studies suggested that endocytosis pathways can be identified by the combination of different inhibitors, endocytic markers and genetic approaches [59]. Nevertheless, there is a possibility of side effects, upregulation of compensatory mechanisms and their effect can vary between different cell lines [59].
Subsequently, whether the conjugation of the C 8 ligand with the aptadendrimer influences the internalization process was analyzed. For this, cell lines were incubated for 1 h with the C 8 -aptadendrimer complex and after, imaged by CLSM. According to Figure 8, it is possible to claim that the aptadendrimer internalizes differently than free AT11 and that it is localized in the cytoplasm of the cells, demonstrating that the association of C 8 to the aptadendrimer did not affect the internalization and uptake into the cells.
Considering that the lysosomal escape of nanoparticles is a crucial parameter for an efficient intracellular delivery [60], we decided to analyze the capacity of the developed aptadendrimer to escape from the lysosomal compartmentalization in PC-3 and PNT1A cells. From CLSM and using the LysoView 540 probe to stain lysosomes, we observed that PC-3 cells seemed to easily internalize the aptadendrimer and presented a higher fluorescence intensity than PNT1A ( Figure S14). Regarding colocalization coefficients (Table S2), they were similar and relatively high, which could indicate that the aptadendrimer was unable to escape lysosome degradation in both cell lines. Although, the free aptadendrimer that did not colocalize with LysoView 540 (in both cell lines) could influence the similar results in the cell viability after aptadendrimer treatment. The colocalization of the aptadendrimer with the lysosomal marker corroborates the previous experiments where we confirmed that the passive diffusion mechanism is not the uptake route for aptadendrimer internalization. A recent study conducted by Moreno-Echeverr et al. summarizes a detailed experimental protocol for sample preparation, staining and imaging to improve the reproducibility of nanoparticles' colocalization with lysosomes [61].
Subsequently, whether the conjugation of the C8 ligand with the aptadendrimer in-fluences the internalization process was analyzed. For this, cell lines were incubated for 1 h with the C8-aptadendrimer complex and after, imaged by CLSM. According to Figure  8, it is possible to claim that the aptadendrimer internalizes differently than free AT11 and that it is localized in the cytoplasm of the cells, demonstrating that the association of C8 to the aptadendrimer did not affect the internalization and uptake into the cells.

Conclusions
This study reports the synthesis of an aptadendrimer, a fourth-generation GATG dendrimer functionalized on the surface with 34 copies of the AT11 aptamer loading the acridine orange ligand C 8 . The aptamer AT11 adopts a G4 structure confirmed by CD spectroscopy before and after conjugation with the dendrimer. The aptadendrimer was prepared by conjugation via SPAAC, leading to nanostructures with a particle size of 3.1 nm suitable for in vitro studies. The association of the aptadendrimer with C 8 , an acridine orange ligand with anticancer properties that binds/stabilizes the G4 structure of AT11, was confirmed by steady-state and time-resolved fluorescence and anisotropy measurements. Fluorimetry experiments also showed a high loading capacity of the aptadendrimer with C 8 . The release of C 8 from the aptadendrimer was significant after 4 h and showed a maximum cumulative release profile after 12 h. Confocal microscopy indicated that the C 8 -aptadendrimer complex was efficiently taken up into the cells and it was maintained during cell internalization and trafficking. The internalization of the aptadendrimer or complex with C 8 was not mediated by NCL or passive diffusion but by other internalization pathways such as endocytosis. The compartmentalization of aptadendrimers in the lysosomes was also verified. Cell viability studies showed that the aptadendrimer and C 8 complex presented a non-selective cytotoxic effect.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.