Targeting structural features of viral genomes with a nano-sized supramolecular drug

Figures

• 

  Fig. 1. (A) P and M enantiomers of the iron cylinder [Fe2L3]4+ optimized by DFT. (B) Distribution of partial charges for P enantiomer as calculated by ssb-d-D3/Def2-SVP level of DFT theory, visualized by VMD, and also showing approximate cylinder size. (C) Surface of the RNA 3-way junction cavity stabilized by the M enantiomer of the cylinder from the crystal structure pdb 4JIY.43 (D) Stacking of RNA bases to the cylinder in the centre of the 3-way junction in pdb 4JIY.43 Analogous stacking is also seen with cylinders located at the terminal base pairs of the strands (see ESI part B‡). Hydrogen are omitted for clarity in A, C and D.

• 

  Fig. 2. (A) Summary of the 1ANR 20 NMR solutions presented in Leontis Westhof (LW) nomenclature. (B) LW nomenclature of 10 μs simulation and PCA and TICA free energy surfaces, demonstrating: how the simulation reproduces 1ANR NMR structure but also reveals transient pairings (LW yellow) not well defined by (but nevertheless noted in) NMR; the greater richness of information in TICA analysis over PCA; the many conformations (TICA minima) that are accessible in the simulation at this temperature (310 K). (C) LW nomenclature of 6 μs simulation and PCA and TICA free energy surfaces. (D) Combined results of seven 2 μs simulations (see ESI Methods‡) starting with different NMR solutions.

• 

  Fig. 3. Left: end-capping of the cylinder observed in an MD simulation. Right: the end-stacking experimentally observed in crystal structure 4JIY43 (red), overlain with that observed in an MD simulation followed by DFT optimisation (blue).

• 

  Fig. 4. Exemplar bulge-binding interactions observed in the simulations, en route to bulge insertion, including bridging from bulge to loop. The right hand figure has the cylinder in the position where a cyclic peptide has been observed to bind to TAR.

• 

  Fig. 5. The bulge insertion mode: the surface of the RNA cavity shows the extremely high contact surface for (A) the M enantiomer and (B) the P enantiomer, and the similarity to each other and to the 3WJ-binding (compare Fig. 1C).

• 

  Fig. 6. (A–E) Entry of M enantiomer: (A) starting position of M cylinder on 1ANR1. (B) Cylinder rotates to split the U25 G26 and (C) aligns in parallel to the G26:C39 base pair (order of microseconds). (D) After relaxation of the backbone (order of microseconds), (E) the cylinder is inserted into the cavity (order of nanoseconds). In contrast to the P cylinder the M cylinder splits the C39 U40 and makes contact transiently stacking the 3 nucleotides of the bulge. (G–L) Entry of P enantiomer: (G) starting position of P cylinder on 1anr1. (H) Cylinder splits the CU nucleotides at the non-bulged strand and (J) pushes the AU base pair (order of microseconds). (K) The bp opens and the cylinder aligns parallel to the GC base pair (order of nanoseconds) and (L) after the AU closes the P cylinder is in the centre of the bulge. (M) Transition timescales for the M cylinder between states.