(586a) In silico investigation of the Loading Mechanisms of Rosette Nanotubes As a Drug Vehicle for paclitaxel

Paclitaxel (PTX) is used as a treatment for breast, ovarian, bladder, lung, prostatic, melanoma, and esophageal cancers, as well as Kaposi sarcoma, and other solid tumors. It is cytotoxic not only to cancer cells, but to other fast-growing and rapidly multiplying cells as well. This can cause severe side effects such as gut irritation, bleeding, hair loss, and inflammation. Targeted delivery by nano-based drug delivery systems (NDDS) can reduce severe side effects while maintaining therapeutic efficacy. NDDS, such as the carbon-based variety, reportedly affect the human biological functions due to their interactions with biomolecules. Liposomal formulations, on the other hand, were reported to have limitedly improved the overall survival of treated cancer patients despite improvements. Thus, there is a need for further development of nano-based drug delivery systems.

Rosette nanotubes (RNT) are DNA-inspired self-assembling materials made from guanine-cytosine motifs. They employ the Watson-Crick hydrogen bonding network (donor-donor-acceptor : acceptor-acceptor-donor) to form ring stacks or helical structures with a cavity for drug loading. They are a promising drug delivery vehicle of anti-cancer drugs due to their biocompatibility, amphiphilicity, low toxicity, and tunable surface for a wide range of functional groups for targeted delivery. The amphiphilic property of the RNT can help in the encapsulation and delivery of hydrophobic drugs.

In this study, the potential of Lysine-functionalized RNTs as a drug vehicle for PTX is explored. Molecular dynamics (MD) simulations and Molecular Mechanics Poisson-Boltzmann Surface Area (MMPBSA) analysis were used to investigate the structural stability of RNT with PTX and provide insights on the drug loading mechanisms of PTX onto the RNT. MD simulations were done using GROMACS 2021.4 and MMPBSA analysis were performed using gmx_mmpbsa tool in AMBERTOOLS22. Three types of RNT were adapted from a study by Yamazaki and Fenniri (2016)⁵; K1, xK1, and iEt-xK1. K1 is a bicyclic RNT, while xK1 and iEt-xK1 are both tricyclic but iEt-xK1 has an ethyl group attached in the center ring.

The results show that PTX can remain in the inner channel of each type of RNT in a vertical position along the length of RNT. But MD trajectories revealed a split in the middle of K1 after the PTX was initially placed horizontally in the inner channel (shown in Figure A). That is due to the limited confined space when PTX was initially placed parallel to the x-axis of K1 where its atoms overlap with the atoms from K1 resulting to steric repulsion. Interestingly, stability was observed after the repulsion while the PTX remained in between which suggests the possibility of sandwiching drug molecules by two RNTs but further investigation is necessary. As to other two types of RNT, when PTX was initially placed horizontally in their inner channels, it moved towards the upper end of xK1 and iEt-xK1 and remained stable there as shown in figure B. Thus, PTX will most probably remain in the inner channel when it is in vertical position as shown in Figures C & D. Generally, a stability throughout the simulation was observed in the RMSD analysis when the PTX has remained on either end and in the inner channel of each type of RNT, and instability when the PTX was initially placed away from the RNT as it randomly moves around the whole system (refer to Figures E, F and G). This is further elucidated by the radial distribution analysis, where PTX are shown on either end of each type of the RNT. The key interactions that held the PTX on either end are π–π stacking and van der Waals interactions while hydrogen bonding held the PTX inside the channel. MMPBSA analysis (Table A) revealed that the binding free energies (-69.29 to -117.74 kJ/mol) when PTX has settled on either end are more stable than the binding free energies (-40.88 to -53.51 kJ/mol) when PTX has remained in the inner channel. Therefore, the most probable way of loading the PTX in each type of RNT is by attachment on either end after each RNT has self-assembled and formed into a tubular structure. Loading the PTX inside the channel of each RNT is less likely to occur since a steric repulsion occurred for K1 and since the PTX eventually moved out from the inner channels of xk1 and iEt-xK1. However, Encapsulation is still possible due to the favorable binding energies when PTX has remained inside the channel despite being weaker than others, although more studies are needed to prove this. From the results, xK1 and iEt-xK1 are good drug delivery candidate for PTX while K1 is not ideal as a drug vehicle for the PTX due to the steric repulsion that has occurred that resulted to significantly large binding free energy of 3384.14 kJ/mol. The insights gained here will be fundamental in achieving an efficient and safe formulation of RNTs as a drug delivery system.