br Fig Fluorescence emission spectra of A D C B
Fig. 5. Fluorescence emission spectra of (A, D) C3, (B, E) C5 and (C, F) C8 with increasing concentration of AT11 G4 (top) or AT11-B0 G4 (bottom). Insets: fraction of ligand bound plots fitted to the Hill saturation binding model. Spectra acquired in 20 mM potassium phosphate buﬀer containing 65 mM KCl, pH 7.
Apparent dissociation constants (KD) obtained from fluorescence titration ex-periments.
Ligand KD (M) n KD (M) n
Rahman barostat. Finally, explicit solvent MD runs of 50 ns were per-formed for each G4-nucleolin complex at constant temperature (300 K) and pressure (1 atm). The Particle mesh Ewald (PME) method was used for calculating long range electrostatic interactions. A 1 nm cut-oﬀ was applied to short range Lennard-Jones interactions. Coordinates were collected in trajectory files every 10 ps. All molecular images were rendered using UCSF Chimera 1.11.2.
3. Results and discussion r> Nucleic APTSTAT3-9R aptamers have been employed to deliver an extensive array of ligands with therapeutic interest such as doxorubicin or 5-fluorouracil to cancer cells (Zhu et al., 2015). Our idea is based on the fact that sequences with the potential to form G4s are attractive nu-cleolin binders, since nucleolin has high aﬃnity to G4s and is over-expressed in HeLa cervical cancer cells (Bates et al., 2009). AT11 and AT11-B0 G4s can gain intracellular access by endocytosis after binding to nucleolin present at the cell surface, and benefit of nucleolin shut-tling to the nucleus. Based on the presumed mechanism, we have evaluated if AT11 and AT11-B0 G4s can still recognize nucleolin while carrying ligands for an intracellular targeted delivery with anticancer eﬀects.
The ligands selected were acridine orange derivatives (C3, C5 and C8) that were already evaluated in HeLa cancer cells and non-malignant
cells, showing in both cases remarkable cytotoxicity (Fig. 1). Through a supramolecular strategy, the AT11 and AT11-B0 G4s-ligand conjugates were tested for their ability to decrease the oﬀ-target toxicity associated with the acridine orange ligands towards non-malignant cells while still maintaining their eﬀects in HeLa cells.
Firstly, we studied the formation and topology of the AT11 and AT11-B0 G4s by TDS absorbance spectroscopy. The diﬀerence between the folded and unfolded spectra of a oligonucleotide may provide va-luable information about its conformation (Mergny et al., 2005). The TDS spectra of AT11 and AT11-B0 are depicted in Fig. 2. A common feature to both spectra is the negative signal at 295 nm which is in-dicative of a G4 structure (Karsisiotis et al., 2011; Mergny et al., 2005). The TDS factor A240nm/ A295nm for AT11 and AT11-B0 was 3.5 and 2.3, respectively. The values above 2 suggest parallel-stranded G4 topologies for both AT11 and AT11-B0 sequences (Karsisiotis et al., 2011).
Afterwards, the formation of complexes AT11- and AT11-B0 G4s-ligand was studied using CD and steady state fluorescence spectro-scopies. The CD spectra of AT11 and AT11-B0 G4s are shown in Fig. 3. CD spectroscopy is a useful technique to infer on the G4 topology of a given G-rich oligonucleotide and to evaluate the eﬀect of ligand binding on the pre-formed topology (Carvalho et al., 2017). Both AT11 and AT11-B0 present the characteristic spectral features of parallel-stranded G4s as seen by a positive peak around 260 nm and a negative peak around 240 nm (Do et al., 2017), in agreement with the conclusions reached by TDS analysis. Upon titrating the ligands with the AT11 G4 solution, an overall increase in ellipticity was observed together with the retention of the characteristic bands (Fig. 3A). This indicates that the ligands are able to bind the G4 structure without disrupting/ changing its pre-folded topology. In the case of AT11-B0 G4, ligands C3 and C5 had little or no eﬀect on the ellipticity while C8 promoted an increase (Fig. 3B).
Then, the eﬀect of ligand binding on the thermal stability of the AT11 and AT11-B0 G4 structures was assessed by CD melting experi-ments. The potential stabilization of AT11 and AT11-B0 G4 structures
may be relevant in two diﬀerent ways: first, it may be used as a measure of how stable the complex is in solution; and second, it might indicate how the interaction with the ligand may enhance the targeting prop-erties of the aptamers as G4 conformations (Bates et al., 2009). The CD melting data of each ligand is shown in Fig. 4. The melting tempera-tures (Tm) of AT11 and AT11-B0 G4s were 42.0 and 53.7 °C, respec-tively. Despite the adjustment of the salt conditions in the melting ex-periments for each aptamer (65 mM KCl vs 10 mM KCl for AT11 and AT11-B0, respectively), the Tm value of AT11-B0 G4 is significantly higher than that of AT11 G4 due to the modification introduced in the sequence. AT11-B0 derives from AT11 by the elimination of the bulge residues which leads to an increased thermal stability (Do et al., 2017). Upon addition of 4 M equivalents of acridine orange ligands, a clear increase in the Tm values was observed which indicates ligand-induced stabilization of the G4 structures (Table 1).