*4.1. Design, Optimization and Chemico-Physical Charactrization of the Double-Tail, Dual Targeting Self-Assembling Amphiphilic Dendron AD*/*E16G6RGDK*

In the general design strategy of nanocarriers to achieve effective siRNA (and in general drug) delivery, two well-known criteria must be satisfied: (i) The delivered therapeutic is be able to reach the desired disease (e.g., tumor) site(s) after administration with minimal loss in their quantity and activity during their journey in the blood stream; and (ii) they only affect diseased cells without exerting harmful effects to healthy organs and tissues. These requirements may be enabled using two strategies: Passive and active drug targeting of drugs. The former approach relies on the (still controversial) intrinsic enhanced permeability and retention effect (EPR) [19,20], according to which molecules with size in the nanometer range tend to accumulate in diseased (especially tumor) tissues much more than they do in normal tissues, thereby leaving little space for specific molecular design and optimization. Active targeting employs directed intermolecular interactions (e.g., ligand binding to disease-overexpressed receptors or other molecular recognition processes) to confer more specificity to the delivery system. In the end, besides enhanced therapeutic action, active targeting can result in reduced non-specific interactions and localization of the drug in peripheral tissues, thus minimizing unwanted side-effects. One of the biggest challenges in active targeting consists in designing new nanovectors (or modifying existing performing ones) with targeting moieties that do not interfere with its cargo loading capacity and its final mission purposes. Obviously, these chemical entities must be fully available to interact with their cellular counterparts and be endowed with strong affinity towards their targets (e.g., surface receptors) in order to trigger endocytosis.

With the goal of empowering the high-performance double-tail self-assembling amphiphilic dendron **AD** with active targeting ability for siRNA specific delivery to cancer cells, the authors considered the RGDK peptide as the starting point for molecular design and optimization. The RGDK peptide is an ideal moiety for cancer targeted nanomedicine in that it features a dual targeting capacity within a single, small molecular scaffold. On the one side, the RGD sequence is known to be able to target tumor endothelium via interaction with ανβ3-integrin (a receptor overexpressed in tumor vasculature [21]); while on the other side, the RGDK sequence has been shown to bind the neutropilin-1 (Nrp-1, a transmembrane hub receptor with multiple ligands that is abundant on the surface of cancer cells [22]), and in so doing, promoted cell uptake and penetration.

Several sequence rounds of computer optimization led to the final design of the E16G6RGDK peptide shown in Figure 13a [8]. This new chemical entity features three distinct segments serving different functions, as follows: (1) A highly negatively charged tail composed of 16 glutamic acid residues (E16), for anchoring the peptide to the positively charged surface of the self-assembled **AD** micelles via strong, favorable electrostatic interactions; (2) the positively charged RGDK warhead for fostering cancer cell targeting and homing by virtue of the dual interaction with ανβ3-integrin and Nrp-1 receptors; and (3) a neutral spacer of optimal length composed of 6 glycine residues (G6) separating the oppositely charged segments, (1) and (2).

**Figure 13.** (**a**) Chemical structure of the E16G6RGDK peptide designed as a double targeting moiety for enhancing the siRNA delivery to cancer cells mediated by the double-tail self-assembling **AD** dendron. The light red box encases the E16 fragment, the light green box marks the G6 spacer while the light blue box highlights the RGDK warhead. (**b**) TEM imaging of the siRNA/**AD**/ E16G6RGDK nanosized micelles formed upon simple addition of the targeting peptide to pre-formed siRNA/**AD** complexes. (**c**) Integrated ITC profile for the titration of siRNA/**AD** complexes with the E16G6RGDK peptide. The solid red line represents data fitting with a sigmoidal function. The insert shows the corresponding ITC raw data. Adapted from [8] with permission of the American Chemical Society, 2018.

With the optimized new peptide structure at hand, and the previous knowledge about the self-assembling process of **AD** and its related performance as siRNA nanocarrier (Section 3), the authors initially verified the interaction of the E16G6RGDK molecules with pre-formed siRNA/**AD** complexes in solution by means of different experimental techniques. Both DLS and TEM confirmed the formation of nanosized (30–45 nm in diameter) spherical micelles (shown in Figure 13b), characterized by a ζ-potential value of +15 mV, approximately equal to half of the value measured for the siRNA/**AD** assemblies (+32 mV). This substantial decrease in positive surface charges further supported the interaction of the negatively charged (−15) peptide with the siRNA/**AD** complexes.

However, the ultimate confirmation of the effective binding of the E16G6RGDK targeting peptides to the siRNA/AD nanomicelles was obtained by isothermal titration calorimetry (ITC). ITC is a straightforward and noninvasive titration-based method for binding interaction analysis performed in solution at constant temperature [23]. The analysis of ITC curves directly yields the binding enthalpy (ΔHbind) and, upon data fitting with a suitable thermodynamic model, the binding constant Kb can be obtained. Once Kb is known, the free energy of binding ΔGbind is calculated using the relationship ΔGbind = −RT lnKbind, where R is the universal gas constant and T is the absolute temperature. Finally, the variation in entropy upon binding is derived from the fundamental Gibbs equation ΔGbind = ΔHbind − TΔSbind. The binding thermodynamics between the targeting peptides and the siRNA/AD micelles was found to be characterized by a favorable (i.e., negative) enthalpic variation (ΔHbind = −5.0 ± 0.2 kcal/mol), as evidenced by the exothermic peaks in the corresponding thermogram (Figure 13c). Also, a small, favorable (i.e., positive) entropic change (TΔS = 1.9 kcal/mol) was estimated, leading to an overall favorable (i.e., negative) ΔGbind value of −6.8 kcal/mol. The substantial enthalpic nature of the binding between the RGKD peptide and the siRNA/**AD** micelles is the result of electrostatic forces as the main supramolecular interaction drivers. However, the favorable entropic contribution was ascribed to the stabilizing peptide/micelle hydrophobic interactions and the concomitant release of ions and water molecules into the bulk solvent [24–26].

The structure of the siRNA/**AD**/E16G6RGDK was next investigated at the molecular level via computer simulations. Extensive (1 μs) atomistic molecular dynamics (MD) simulations revealed the formation of stable, Janus-like nanoparticles featuring a siRNA molecule bound on one face of each **AD** micelle and 4 peptides adsorbed on the opposite face (Figure 14a). Further, MD simulations showed that the negatively charged peptide tails (E16G6) were adsorbed onto the positively charged surface micelle while, most importantly, the 4 RGDK warhead was free from sterical hindrance and protruded deeply into the solvent (Figure 14b).

**Figure 14.** (**a**) Image extracted from the equilibrated portion of the MD simulation of a siRNA/**AD** micelle in complex with 4 E16G6RGDK peptides performed in 150 nM NaCl water solution. The 4 peptides are numbered and colored as follows: E16G6RGDK1, tail (E16G6) sandy brown, head (RDGK) firebrick; E16G6RGDK2, tail light green, head olive drab; E16G6RGDK3, tail khaki, head golden rod; E16G6RGDK4, tail plum, head magenta. The **AD** micelle is in a dark slate blue sticks-and-balls representation, the siRNA molecule is highlighted by its gray van der Waals surface, the peptides are shown as colored spheres, Na<sup>+</sup> and Cl<sup>−</sup> ions are depicted as purple and green hollow spheres, while a light cyan transparent surface is used to represent the water solvent. (**b**) Zoomed view of the image in panel (**a**), showing the E16G6RGDK4 peptide with its tail (plum) attached to the micelle surface and its RGDK warhead (magenta) protruding into the solvent. In this panel, water and ions are omitted for clarity. (**c**) Radial distribution function (RDF) from the **AD** center of mass for the positively charged **AD** terminal groups (dark slate blue), and for each of the four RGDK warhead groups (colors as in panel (**a**)). (**d**) Free energy of binding (ΔGbind) of each E16G6RGDK peptide on the siRNA/**AD** micelle. The patterned bars represent the ΔGbind values for the entire peptide molecule, while the solid bars show the contribution to ΔGbind afforded by the negatively charged (E16G6) tails only. Adapted from [8] with permission of the American Chemical Society, 2018.

These pictorial views were quantified first by calculating the radial distribution function (RDF) from the **AD** micelle center of mass for the positively charged **AD** terminal groups and the 4 targeting peptide RGDK warheads, as shown in Figure 14c. As seen from this Figure, all 4 RGDK moieties are positioned far away from the nanomicelle periphery. Accordingly, they are potentially available for interaction with their cellular target receptors. Next, the free energy of binding ΔGbind of each single E16G6RGDK peptide with the siRNA/**AD** nanomicelle was estimated by processing the MD data according to the MM/PBSA methodology [9–15]. Figure 14d shows not only that all peptides are characterized by favorable ΔGbind values (from −7.0 to −6.2 kcal/mol), but also that the major contribution to binding is afforded by the strong electrostatic interaction between the negatively charged peptide tails (E16G6) and the positively charged micellar surface, as expected by the original molecular design.
