Singlet oxygen-mediated photochemical cross-linking of an engineered fluorescent flavoprotein iLOV

Abstract

Genetically encoded photoactive proteins are integral tools in modern biochemical and molecular biological research. Within this tool box, truncated variants of the phototropin two light-oxygen-voltage flavoprotein have been developed to photochemically generate singlet oxygen (¹O₂) in vitro and in vivo, yet the effect of ¹O₂ on these genetically encoded photosensitizers remains underexplored. In this study, we demonstrate that the "improved" light-oxygen-voltage flavoprotein is capable of photochemical ¹O₂ generation. Once generated, ¹O₂ induces protein oligomerization via covalent cross-linking. The molecular targets of protein oligomerization by cross-linking are not endogenous tryptophans or tyrosines, but rather primarily histidines. Substitution of surface-exposed histidines for serine or glycine residues effectively eliminates protein cross-linking. When used in biochemical applications, such protein-protein cross-links may interfere with native biological responses to ¹O₂, which can be ameliorated by substitution of the surface exposed histidines of improved" light-oxygen-voltage or other ¹O₂-generating flavoproteins.

Keywords: flavoprotein; histidine; oligomerization; protein chemical modification; protein cross-linking; reactive oxygen species (ROS).

Figure 1

Recombinant expression, purification, and MALDI-ToF analysis of cross-linked iLOV.A, SDS-PAGE gel depicting the purification of iLOV. Lanes: one molecular weight marker, 2 Escherichia coli cell lysate, three soluble lysate, four cell debris, five immediate Ni-NTA flow-through, 6 to 9 Ni-NTA flow-through during the imidazole wash, 10 Ni-NTA eluted (H)₆-SUMO-iLOV, 11 desalted (H)₆-SUMO-iLOV, 12 ULP1-digestion product, 13 ULP1-digested Ni-NTA flow-through, 14 ULP1-digestion Ni-NTA elution, 15 desalted ULP1-digested Ni-NTA flow-through. B, MALDI-ToF mass spectrum of purified iLOV suggests oligomerization of WT iLOV. iLOV, improved light-oxygen-voltage; Ni-NTA, nickel nitrilotriacetic acid; ULP1, ubiquitin-like protease 1.

Figure 2

Photochemical dimerization of iLOV.A, SDS-PAGE analysis of illumination time-dependent dimerization of iLOV. B, UPLC-MS analysis of total ion counts of trypsin digested SDS-PAGE gel-excised monomer (navy) and dimer (orange) bands. Peptides lost upon illumination in the dimer band are indicated by asterisks. iLOV, improved light-oxygen-voltage.

Figure 3

Cross-linking of iLOV in reducing or anaerobic conditions.A, illumination of anaerobic iLOV. B, illumination of an aerobic samples of iLOV in the presence of 10 mM DTT. iLOV, improved light-oxygen-voltage.

Figure 4

iLOV-mediated photochemical¹O₂generation, observed through SOSG fluorescence in the presence (+) and absence (−) of O₂and illumination (hν). The fluorescence response in all conditions is normalized to the SOSG signal of iLOV illumination aerobically (navy). Anaerobic SOSG responses shown in stripped bars. Photochemical response of SOSG alone is shown in orange and yellow. iLOV, improved light-oxygen-voltage; SOSG, singlet oxygen sensor green.

Figure 5

The role of histidines in iLOV cross-linking.A, an X-ray crystal structure of iLOV (PDB: 4EES) depicting the locations of each histidine residue. H − 2 is not present in the crystal structure as it was introduced only in this work. B, the same crystal structure depicting the van der Waals protein surface. The native histidine residues and the N-terminal I1 (as a proxy for the N-terminal H − 2) are depicted in blue spheres. C, SOSG response following 30 min illumination of iLOV (white), iLOV^ΔH (orange, p-value of 0.21), iLOV^ΔYW (navy, p value of 0.00015), and a –iLOV control (gray). D, effect of histidine residues on photochemical cross-linking measured by illumination time-dependent SDS-PAGE. E, the proposed mechanism of ¹O₂/histidine-mediated protein cross-linking. iLOV, improved light-oxygen-voltage; SOSG, singlet oxygen sensor green.

Figure 6

Predicted O₂diffusion into iLOV and role of A40 in access to FMN cofactor.A, molecular dynamics simulations predict three sites occupied by O₂ within iLOV labeled “X,” “Y,” and “Z” near A40. B,¹O₂ generation by iLOV and various A40 mutants determined by SOSG fluorescence after illumination. p values, calculated using a t test against iLOV^WT were 0.0043, 0.000095, and 0.0048 for A40G, A40V, and A40I mutants, respectively. FMN, flavin mononucleotide; iLOV, improved light-oxygen-voltage; SOSG, singlet oxygen sensor green.