Enhancing the performance of Magnets photosensors through directed evolution

Abstract

Photosensory protein domains are the basis of optogenetic protein engineering. These domains originate from natural sources where they fulfill specific functions ranging from the protection against photooxidative damage to circadian rhythms. When used in synthetic biology, the features of these photosensory domains can be specifically tailored towards the application of interest, enabling their full exploitation for optogenetic regulation in basic research and applied bioengineering. In this work, we develop and apply a simple, yet powerful, directed evolution and high-throughput screening strategy that allows us to alter the most fundamental property of the widely used nMag/pMag photodimerization system: its light sensitivity. We identify a set of mutations located within the photosensory domains, which either increase or decrease the light sensitivity at sub-saturating light intensities, while also improving the dark-to-light fold change in certain variants. For some of these variants, photosensitivity and expression levels could be changed independently, showing that the shape of the light-activity dose-response curve can be tuned and adjusted. We functionally characterize the variants in vivo in bacteria on the single-cell and the population levels. We further show that a subset of these variants can be transferred into the mOptoT7 for gene expression regulation in mammalian cells. We demonstrate increased gene expression levels for low light intensities, resulting in reduced potential phototoxicity in long-term experiments. Our findings expand the applicability of the widely used Magnets photosensors by enabling a tuning towards the needs of specific optogenetic regulation strategies. More generally, our approach will aid optogenetic approaches by making the adaptation of photosensor properties possible to better suit specific experimental or bioprocess needs.

Light-directed evolution of dynamic, multi-state, and computational protein functionalities

“…Molecular properties that are not under selection pressure can consequently deteriorate during directed evolution experiments. Not only does the existing work-around for this problem, to run directed evolution campaigns for each functionality sequentially, require repeated interventions and take substantial time 27,29,34 , it is also inherently challenging: The basal activity of wild-type El222 prevented us from evolving the on (1) output state using the -Ura/+FOA method. Then, in attempting to reduce the leakiness of El222, we did not succeed in finding any mutations in El222 in the two hundred 5-FOA-resistant colonies we tested.…”

“…LOV domains make up the core of most non-ion-channel optogenetic systems [39][40][41][42][43][44][45][46][47][48][49][50][51][52][53] , presumably because they are small, reliable, and do not need exogenous chromophores. Interest in LOV domains is reflected in multiple diversification campaigns 34,[54][55][56] . The LOV transcription factor El222, in particular, is widely used in many biological models from cell-free systems 57 , to bacteria 58 , yeast 59,60 , mammalian cell lines 61,62 , and zebrafish 63 .…”

“…24 For example, long-term directed evolution of GPCRs under constant ligand concentrations reduces ligand responsiveness. 25 To cope with this limitation, researchers resort to alternating directed evolution campaigns for each of the different states of the POI at a time [26][27][28][29][30][31][32][33][34][35][36][37] . For example, a bioengineer will assay for on (1) activity under inducing conditions, select candidate mutants, then assay for off (0) activity under non-inducing conditions, select mutants, and repeat this process.…”

“…We apply optovolution to three systems, the optogenetic transcriptional system El222 and the optogenetic dimerization switch PhyB-Pif3, which relay light input by switching on (1) or off (0), and an AND gate, in which a degron-tagged rtTA transcription factor is controlled by transcription and doxycycline: LOV domains make up the core of most non-ion-channel optogenetic systems 39–53 , presumably because they are small, reliable, and do not need exogenous chromophores. Interest in LOV domains is reflected in multiple diversification campaigns 34,54–56 . The LOV transcription factor El222, in particular, is widely used in many biological models from cell-free systems 57 , to bacteria 58 , yeast 59,60 , mammalian cell lines 61,62 , and zebrafish 63 .…”

“…The dynamic, multi-state properties of these proteins make them unsuitable for existing continuous directed evolution paradigms: LOV domains make up the core of most non-ion-channel optogenetic systems 72–80,80–86 , presumably because they are small, reliable, and do not need exogenous chromophores. Interest in LOV domains has spurred past diversification work 54,87–89 . The LOV transcription factor El222, in particular, is widely used in many biological models from cell-free systems 90 , to bacteria 91 , yeast 92,93 , mammalian cell lines 94,95 , and zebrafish 96 .…”