Abstract: Fluorescent proteins (FPs) with emission wavelengths in the far-red and infrared regions of the spectrum provide powerful tools for deep-tissue and super-resolution imaging. The development of red-shifted FPs has evoked widespread interest and continuous engineering efforts. In this article, based on a computational design and genetic code expansion, we report a rational approach to significantly expand and red-shift the chromophore of green fluorescent protein (GFP). We applied computational calculations to predict the excitation and emission wavelengths of a FP chromophore harboring unnatural amino acids (UAA) and identify in silico an appropriate UAA, 2-amino-3-(6-hydroxynaphthalen-2-yl)propanoic acid (naphthol-Ala). Our methodology allowed us to formulate a GFP variant (cpsfGFP-66-Naphthol-Ala) with red-shifted absorbance and emission spectral maxima exceeding 60 and 130 nm, respectively, compared to those of GFP. The GFP chromophore is formed through autocatalytic post-translational modification to generate a planar 4-(p-hydroxybenzylidene)-5-imidazolinone chromophore. We solved the crystal structure of cpsfGFP-66-naphthol-Ala at 1.3 Å resolution and demonstrated the formation of a much larger conjugated π-system when the phenol group is replaced by naphthol. These results explain the significant red-shifting of the excitation and emission spectra of cpsfGFP-66-naphthol-Ala.
Keywords: Computational design; Green fluorescent protein; Red-shift; Unnatural amino acids.