Stalling chromophore synthesis of the fluorescent protein Venus reveals the molecular basis of the final oxidation step

Fluorescent proteins (FPs) have revolutionised the life sciences, but the mechanism of chromophore maturation is still not fully understood. Here we show that incorporation of a photo-responsive non-canonical amino acid within the chromophore stalls maturation of Venus, a yellow FP, at an intermediate stage; a crystal structure indicates the presence of O2 located above a dehydrated enolate form of the imidazolone ring, close to the strictly conserved Gly67 that occupies a twisted conformation. His148 adopts an “open” conformation so forming a channel that allows O2 access to the immature chromophore. Absorbance spectroscopy supported by QM/MM simulations suggests that the first oxidation step involves formation of a hydroperoxyl intermediate in conjunction with dehydrogenation of the methylene bridge. A fully conjugated mature chromophore is formed through release of H2O2, both in vitro and in vivo. The possibility of interrupting and photochemically restarting chromophore maturation and the mechanistic insights open up new approaches for engineering optically controlled fluorescent proteins.


Structure determination of Venus 66azF
For dark state Venus azF66 sample, a concentrated purified sample (~20 mg/ml) was used for screening crystal formation using the PACT premier TM HT-96 screen and JCSG-plus TM HT-96 screen (Molecular Dimensions, UK). The sitting drop vapour diffusion method was used in a sealed tray and kept at 20°C. All the steps of crystal preparation were performed in the dark. The growth of crystals was monitored under a light microscope, and grown crystals were harvested by picking individual crystals in a mounted litholoop (Molecular Dimensions) and plunging them into liquid nitrogen. X-ray diffraction collection, analysis and structure determination are outlined in the main text.

Bacterial live cell imaging.
Widefield fluorescence microscopy was used for bacterial live cell imaging. For photoactivation of Venus Y66azF , E. coli Top10 cells were first induced to express Venus 66azF . A Coverwell TM imaging chamber (Sigma-Aldrich) was fixed on a glass slide and 0.2 ml of 1% agarose was applied before the induced bacterial cells (0.2 ml) were added and covered with a cover slide. A sample of uninduced bacterial cells was prepared in the same manner as a control. Slides for induced and uninduced bacterial cells were exposed to UV-light for 1 min at a distance of about 1 cm. Transmitted and widefield fluorescence images visualised using an inverted Olympus IX73 widefield fluorescence microscope. Images were collected with a Hamamatsu Orca flash 4.0 camera with x100 objective lens using HCImaging software and a Prior Lumen200Pro light source. Fluorescence emission was separated by a multiband dichroic emission filter set 69002 (Chroma), a wavelength of 450 nm was used for the excitation.

Mass spectrometry.
Protein liquid chromatography-mass spectrometry (LC-MS) was acquired on a Waters Acquity H-Class UPLC system coupled to a Waters Synapt G2-Si quadrupole time of flight mass spectrometer. The column used was a Waters Acquity UPLC Protein C4 BEH column 300 Å, 1.7 μm (2.1 x 100 mm) held at 60 °C with samples elutes over an acetonitrile/water gradient containing 0.2% formic acid. The flow rate was 0.2 mL/min. Mass spectrometry data was collected in positive electrospray ionisation mode and the data analysed using Waters MassLynx 4.1.
Deconvoluted mass spectra were generate using the MaxEnt1 software. Supporting Figure S1. The mass analysis of the sample prior to irradiation provides additional evidence that Venus 66azF has undergone dehydration and retains the phenyl azide group in the chromophore. If hydration had not taken place, we would expect to see mass peaks of 16-18 Da higher than that observed. Irradiation. On irradiation, the samples of Venus 66azF used here replicated the data shown by samples used to generate data in the main manuscript. After 1 min UV irradiation, Venus 66azF transform to an intermediate that directly converts to the 440 nm absorbing species (isobestic point ~390 nm). We then followed the reaction by SDS-PAGE and found a fragment at ~20kDa appeared after 1 min irradiation. The strength of this band remained relatively constant between 5-30 mins (20-25% as based on densitometry analysis using ImageJ). Given that we still see absorbance changes and fluorescence increase between 1-30 mins, we think that fragmentation is not a product of photolysis but the more commonly observed, denaturation/unfolding induced cleavage at the chromophore (Ref 4 and references therein). It does appear that the product after photolysis is more prone to fragmentation. This could be due to conversion to the phenyl amine, which can generate resonant forms that in turn may promote cleavage between the proposed C65-Ca65 bond that will become solvent accessible once denatured. Under the conditions of the ESI-ToF (water/acetonitrile mix with 0.2% formic acid at 60°C), the fragmentation appears more extensive than under SDS-PAGE, with the major product being 19958-19959 Da. It is known that harsher denaturation conditions can promote fluorescent protein fragmentation (Ref 4 and references therein). The major product equates to conversion of the phenyl azide to the phenyl amine and cleavage between C65-Ca65 (see Figure S2d for proposed structures) to generate a hydroxylated product (Product A). We cannot fully rule out a methylated product (Product B) potentially due to cleavage between N65-Ca65 but Product A is more likely to result from a tentative hydrolysis reaction. There is precedence that such fragmentation products of denatured GFP-derived proteins with modified chromophore compositions can occur as demonstrated by Getzoff and colleagues 5 using protein X-ray crystallography. There is no evidence that additional covalent changes have occurred in the fragment, such as UV induced decarboxylation of Glu222 reported previously 6 . Thus, we believe that fragmentation is due to exposing the phenyl amine version of Venus 66azF to denaturing conditions so promoting cleavage between C65-Ca65 rather than any UV induced cleavage events.

Supporting
Supporting Figure S3. (a) Overlay of the Venus WT (yellow) and im-Venus 66azF (grey). The chromophore is shown as sticks. The RMSD between backbone is 0.39 Å. Chromophore structure of im-Venus 66azF . (b) Electron density map (2Fo-Fc, 1.0 sigma) of the CRO. Each residue component of the original sequence is shown. The C65 carbon where a hydroxyl group would be observed prior to the dehydration step is outlined; no electron density for the expected hydroxyl group is observed. The C66-O66 bond indicated by the O66 annotation, is consistent with a single bond at 1.48Å.
(c) The final coordinates of the chromophore superposed on the omit map in that region. Positive difference density contoured at +3σ, negative difference density contoured at -3σ. (d-e) Omit map density near the chromophore, in the early stages of REFMAC refinement before including nonprotein atoms. Observed density (bluegrey) contoured at 1σ, positive difference at +3σ (green), and negative difference at -3σ (red). Automatic placement of waters by COOT in this map located a single water molecule at the site later described as an O2 molecule. Note the excess density beyond the water site, which persisted throughout refinement until replaced by an O2 molecule ( Figure 2, main manuscript). Also, note G67 in the canonical C67-O67 conformation observed for wild-type Venus results in negative density, which is resolved through formation of the non-canonical twisted arrangement shown in Figure S3b and Figure 2 in the main manuscript. Supporting Figure S5. Comparison of the chromophore structure of im-Venus 66azF (grey) and unpublished GFP mutant Q183E (PDB 2qt2; orange). The GFP mutant has a partially formed chromophore and the same orientation of the Gly67 carbonyl oxygen observed for im-Venus 66azF .

a b
Methylene bridge I ring P ring Gly67 Gly67 Figure S6. Schematic outline of the reaction of O2 with im-Venus 66azF , including the potential intermediates and the role of residue E222.