Machine Learning-Assisted Engineering of Light, Oxygen, Voltage Photoreceptor Adduct Lifetime.

Naturally occurring and engineered flavin-binding, blue-light-sensing, light, oxygen, voltage (LOV) photoreceptor domains have been used widely to design fluorescent reporters, optogenetic tools, and photosensitizers for the visualization and control of biological processes. In addition, natural LOV photoreceptors with engineered properties were recently employed for optimizing plant biomass production in the framework of a plant-based bioeconomy. Here, the understanding and fine-tuning of LOV photoreceptor (kinetic) properties is instrumental for application. In response to blue-light illumination, LOV domains undergo a cascade of photophysical and photochemical events that yield a transient covalent FMN-cysteine adduct, allowing for signaling. The rate-limiting step of the LOV photocycle is the dark-recovery process, which involves adduct scission and can take between seconds and days. Rational engineering of LOV domains with fine-tuned dark recovery has been challenging due to the lack of a mechanistic model, the long time scale of the process, which hampers atomistic simulations, and a gigantic protein sequence space covering known mutations (combinatorial challenge). To address these issues, we used machine learning (ML) trained on scarce literature data and iteratively generated and implemented experimental data to design LOV variants with faster and slower dark recovery. Over the three prediction-validation cycles, LOV domain variants were successfully predicted, whose adduct-state lifetimes spanned 7 orders of magnitude, yielding optimized tools for synthetic (opto)biology. In summary, our results demonstrate ML as a viable method to guide the design of proteins even with limited experimental data and when no mechanistic model of the underlying physical principles is available.

Residue alterations within a conserved hydrophobic pocket influence light, oxygen, voltage photoreceptor dark recovery.

Light, oxygen, voltage (LOV) photoreceptors are widely distributed throughout all kingdoms of life, and have in recent years, due to their modular nature, been broadly used as sensor domains for the construction of optogenetic tools. For understanding photoreceptor function as well as for optogenetic tool design and fine-tuning, a detailed knowledge of the photophysics, photochemistry, and structural changes underlying the LOV signaling paradigm is instrumental. Mutations that alter the lifetime of the photo-adduct signaling state represent a convenient handle to tune LOV sensor on/off kinetics and, thus, steady-state on/off equilibria of the photoreceptor (or optogenetic switch). Such mutations, however, should ideally only influence sensor kinetics, while being benign with regard to the nature of the structural changes that are induced by illumination, i.e., they should not result in a disruption of signal transduction. In the present study, we identify a conserved hydrophobic pocket for which mutations have a strong impact on the adduct-state lifetime across different LOV photoreceptor families. Using the slow cycling bacterial short LOV photoreceptor PpSB1-LOV, we show that the I48T mutation within this pocket, which accelerates adduct rupture, is otherwise structurally and mechanistically benign, i.e., light-induced structural changes, as probed by NMR spectroscopy and X-ray crystallography, are not altered in the variant. Additional mutations within the pocket of PpSB1-LOV and the introduction of homologous mutations in the LOV photoreceptor YtvA of Bacillus subtilis and the Avena sativa LOV2 domain result in similarly altered kinetics. Given the conserved nature of the corresponding structural region, the here identified mutations should find application in dark-recovery tuning of optogenetic tools and LOV photoreceptors, alike.

The molecular basis of spectral tuning in blue- and red-shifted flavin-binding fluorescent proteins.

Photoactive biological systems modify the optical properties of their chromophores, known as spectral tuning. Determining the molecular origin of spectral tuning is instrumental for understanding the function and developing applications of these biomolecules. Spectral tuning in flavin-binding fluorescent proteins (FbFPs), an emerging class of fluorescent reporters, is limited by their dependency on protein-bound flavins, whose structure and hence electronic properties cannot be altered by mutation. A blue-shifted variant of the plant-derived improved light, oxygen, voltage FbFP has been created by introducing a lysine within the flavin-binding pocket, but the molecular basis of this shift remains unconfirmed. We here structurally characterize the blue-shifted improved light, oxygen, voltage variant and construct a new blue-shifted CagFbFP protein by introducing an analogous mutation. X-ray structures of both proteins reveal displacement of the lysine away from the chromophore and opening up of the structure as instrumental for the blue shift. Site saturation mutagenesis and high-throughput screening yielded a red-shifted variant, and structural analysis revealed that the lysine side chain of the blue-shifted variant is stabilized close to the flavin by a secondary mutation, accounting for the red shift. Thus, a single additional mutation in a blue-shifted variant is sufficient to generate a red-shifted FbFP. Using spectroscopy, X-ray crystallography, and quantum mechanics molecular mechanics calculations, we provide a firm structural and functional understanding of spectral tuning in FbFPs. We also show that the identified blue- and red-shifted variants allow for two-color microscopy based on spectral separation. In summary, the generated blue- and red-shifted variants represent promising new tools for application in life sciences.