Molecular basis for the increased affinity of an RNA recognition motif with re-engineered specificity: A molecular dynamics and enhanced sampling simulations study

Abstract

The RNA recognition motif (RRM) is the most common RNA binding domain across eukaryotic proteins. It is therefore of great value to engineer its specificity to target RNAs of arbitrary sequence. This was recently achieved for the RRM in Rbfox protein, where four mutations R118D, E147R, N151S, and E152T were designed to target the precursor to the oncogenic miRNA 21. Here, we used a variety of molecular dynamics-based approaches to predict specific interactions at the binding interface. Overall, we have run approximately 50 microseconds of enhanced sampling and plain molecular dynamics simulations on the engineered complex as well as on the wild-type Rbfox·pre-miRNA 20b from which the mutated systems were designed. Comparison with the available NMR data on the wild type molecules (protein, RNA, and their complex) served to establish the accuracy of the calculations. Free energy calculations suggest that further improvements in affinity and selectivity are achieved by the S151T replacement.

Fig 1

(A) NMR structure of pre-miR20b RNA (pdb 2n7x). (B) NMR structure of the Rbfox•pre-miR20b complex (pdb 2n82). The residues at the binding interface are highlighted. The amino acids labelled with a red square correspond to the mutated residues in Rbfox* and in Rbfox* S151T. (C) Left: nucleotide sequence of the pre-miR20b and of the mutant pre-miR20b*. Right: amino acids sequence of the Rbfox, Rbfox* and Rbfox* S151T mutants. Highlighted in green and in orange are the amino acids corresponding to β strands and α helices, respectively. The mutated nucleotides and amino acids are underlined.

Fig 2. Fluctuations of the PAD angle in the simulations of free (top; see Table 1, sim. 1) and bound Rbfox protein (bottom; see Table 1, sim.

9–13). F indicates fluctuations; t short transitions and T long transitions. The structured regions of the protein are highlighted: the grey and yellow regions correspond to the β-strands and helices, respectively.

Fig 3

(A) 2D representation of the free pre-miR20b stem-loop structure. (B) Time development of the εRMSD of the pre-miR20b loop (r-U₂₇GGCAUG₃₃) in the six χ_OL3 MD simulations (Table 1, sim. 2–7). C. RNA backbone dihedral angle histograms calculated over the aggregated simulations. The green dots indicate the values of the angles in the lowest energy structure of the NMR ensemble 2n7x from which the simulations were started.

Fig 4

(A) Bird’s-eye view of the Rbfox•pre-miR20b complex. (B)-(H) Close-up views of the interactions observed at the binding interface during MD simulations. The H-bonds are indicated by dotted red lines. The depicted snapshots belong to the representative structures of the 20 clusters (“MD-adapted structure ensemble”), which have the highest agreement with NMR NOE data. (I) Scheme of the interactions. Circle and arrowheads depict interaction with RNA bases or phosphate groups, respectively.

Fig 5

PAD and tag analyses for Rbfox* free (top) and bound to pre-miR20b* RNA (bottom). “F” indicates fluctuations; “t” short transitions and “T” long transitions. The secondary structure regions of the proteins are highlighted as in Fig 1.

Fig 6

(A) Bird’s-eye view of the Rbfox*•pre-miR20b* complex. (B) and (C) Close-up views of the interactions established by the mutated residues with A₃₀ and C₃₃, respectively. (D) Scheme of the interactions between pre-miR20b* and Rbfox* observed in the MD simulations. Circle and arrowheads depict interaction with RNA bases or phosphate groups, respectively.

Fig 7

(A) Close up view on the interactions established by A₃₀ with T151 and S155 and by U₂₈ with S155 and F126. (B) Comparison between β₂β₃ loop and adjacent residues PAD values of the Rbfox* and S151T mutant in complex with pre-miR20b* RNA. The PAD values relate to the flexibility of the complex.