Computationally Assisted Noncanonical Amino Acid Incorporation

Abstract

Genetic encoding of noncanonical amino acids (ncAAs) with desired functionalities is an invaluable tool for the study of biological processes and the development of therapeutic drugs. However, existing ncAA incorporation strategies are rather time-consuming and have relatively low success rates. Here, we develop a virtual ncAA screener based on the analysis and modeling of the chemical properties of all reported ncAA substrates to virtually determine the recognition potential of candidate ncAAs. Using this virtual screener, we designed and incorporated several novel Lys and Phe derivatives into proteins for various downstream applications. Among them, the genetic encoding of an electron-rich Phe analog, 3-dimethylamino-phenylalanine, was successfully applied to enhance the cation-π interaction between histone methylation and its reader proteins. Thus, our virtual screener provides a fast and powerful strategy to efficiently incorporate ncAAs with diverse functionalities.

Figure 1

Schematic of computationally assisted noncanonical amino acid incorporation.

Figure 2

Development of the ncAA virtual screener. (A) Schematic of the construction of virtual ncAA screener. (B)–(E) The virtual ncAA screener based on Log (P) and ΔG values for Pyl system (B), chPhe system (C), EcTyr system (D), EcLeu system (E), respectively. Overlapping of four virtual ncAA screeners based on Log (P) and ΔG values (F), and Log (S) and ΔG values (G).

Figure 3

The computer-assisted incorporation of the malonylation lysine and glutamylation lysine. (A) Predicated recognizability of Lys derivatives including MalK, GluK, meMalK, meGluK and PrS-GluK with our established virtual ncAA screener of PylRS. (B) Molecular structure of the Lys derivatives. (C) The amber suppression efficiency was tested by a GFP reporter assay with and without the addition of ncAAs for two variants of PylRS. The fluorescent intensity in each group was normalized by the GFP-WT set as 1. The variants were named as meMalKRS and meGluKRS respectively. The error bars represent ± standard error of the mean from three biologically independent experiments. (D) Mass spectrometry characterization of the fidelity of meMalK and meGluK incorporation into GFP. The expected molecular wight (MW) value of GFP incorporated with meMalK and meGluK was 27836 and 27864 Da, and the observed MW value was 27836 and 27864 Da, respectively. The peaks with N-Met cleavage were also detected.

Figure 4

Computer-assisted incorporation of ncAAs with high cation-π interaction potential. (A) Statistics of the presence of Tyr or Phe residues in aromatic boxes of seven types of histone methylation reader domains, including Chromo (chromatin organization modifier), PHD (plant homeodomain), PWWP (domain with a conserved Pro-Trp-Trp-Pro motif), Tudor, MBT (malignant brain tumor), SPIN (Spindlin), and BAH (bromo adjacent homology). (B) Molecular structure of the Tyr and Phe derivatives designed in the study. The electrostatic potential (ESP) maps of the designed derivatives side chain and the cation−π binding energy (CπBE) of the designed derivatives side chain with Na⁺ were calculated and shown. (C) Predicted recognizability of our designed ncAAs with the established virtual ncAA screener of chPheRS. (D) The amber suppression efficiency of different ncAAs was tested by a GFP reporter assay using chPheRS-3 variant. The fluorescent intensity in each group was normalized by the GFP-WT set as 1. Error bars represent ± standard error of the mean of three biologically independent experiments. (E) Mass spectrometry characterization of the fidelity of 3DMAF incorporation into GFP. (F) Flow cytometry analysis of the amber suppression efficiency of 3DMAF incorporation in HEK293T cells. The mean fluorescence intensity ratio of EGFP to mCherry in the absence of 3DMAF was set as 1. Error bars represent the ± standard error of the mean from three biologically independent experiments. Statistical significance was quantified with t test (****p < 0.0001). (G) Mass spectrometry characterization of the fidelity of 3DMAF incorporation into EGFP in HEK293T.

Figure 5

Design of a reader domain with improved histone methylation recognition. (A) Structure of the BPTF-PHD finger (cyan) in complex with the H3K4me3 peptide (yellow) (PDB: 2F6J). The key Tyr residues were showed as sticks. (B) Mass spectrometry characterization of the fidelity of 3DMAF incorporation into BPTF-PHD. The expected molecular mass (MW) value of BPTF-PHD-WT, BPTF-PHD-Y10-3DMAF, and BPTF-PHD-Y17-3DMAF with N-Met cleavage were 19045, 19072, and 19072 Da, and the observed MW value were 19043, 19070, and 19071 Da, respectively. (C) Microscale thermophoretic analysis of BPTF-PHD-WT, BPTF-PHD-Y10-3DMAF and BPTF-PHD-Y17-3DMAF with the H3K4me3 peptide. H3K4me3 was labeled with a fluorescent dye molecule, fluorescein isothiocyanate (FITC). Error bars represent ± standard error of the mean from three biologically independent experiments.