Open Access

Show Abstract ▼

Abstract: Macromolecular crystallography provides mechanistic understanding of biological processes and can be applied in drug design. Nowadays, the use of robotic systems for crystal growth and diffraction analysis is widespread and high-throughput protein-to-structure pipelines for ligand and fragment screening are revolutionizing the field. However, the identification of crystals is still largely carried out through manual inspection, sometimes involving tens of thousands of images, which represents a bottleneck in an otherwise highly automated process. Here we describe AXIS, an AI-based Crystal Identification System combining the DINOv2 computer vision model, state-of-the-art transfer learning and MARCO, the largest crystallization dataset available to date, for automated crystal detection. AXIS can operate with both visible and UV light images and integrates a Lab-in-the-Loop approach combining ML and expert inputs for iterative learning and specialization. AXIS enables automated annotation of large crystallization image datasets with performance and accuracy comparable to that of human experts, and the Lab-in-the-Loop approach introduced here enables efficient adaptation to local conditions, facilitating widespread application, which has been a major limitation to date. AXIS can help to correct human errors in image annotation and removes critical bottlenecks, particularly in the context of extensive crystallization screens or high-throughput applications like fragment and ligand screening, unlocking the potential for higher levels of automation that are key in both fundamental and translational research.

Open Access

Show Abstract ▼

Abstract: The red fluorescent protein mCherry is one of the most widely used fluorescent proteins in biology. Here, we have changed the chromophore chemistry by converting the thioether group of M66 to a thiol group through mutation to cysteine. The new variant, termed mCoral (due to its orange fluorescence hue), has similar brightness to mCherry but improved resistance to hydrogen peroxide. The variant is also responsive to pH with low and high pKa forms that have distinct spectral properties, which DFT analysis suggests is due to protonation state changes in the newly introduced thiol group, as well as the phenol group. The structure of mCoral reveals that the M66C mutation creates a space within the β-barrel structure that is filled by a water molecule, which makes new polar interactions, including the backbone carbonyl group of F65. Molecular dynamics simulations suggest that this additional water molecule, together with local solvation around the chromophore, could play a role in promoting planarity of the full conjugated system comprising the chromophore. The mCoral chromophore makes slightly more H-bonds with water than mCherry. The main water exit point for mCherry is also narrower in mCoral, providing a potential explanation for increased resistance to hydrogen peroxide. Overall, a small structural change to mCherry has resulted in a new fluorescent protein with potentially useful characteristics and an insight into the role of dynamics and water in defining the structure–function relationship in red fluorescent proteins.

Open Access

Show Abstract ▼

Abstract: Cytochromes P460 oxidise hydroxylamine within the nitrogen cycle and contain as their active site an unusual catalytic c-type haem where the porphyrin is crosslinked to the protein via a lysine residue in addition to the canonical cross links from cysteine residues. Understanding how enzymes containing P460 haem oxidise hydroxylamine into either nitrous oxide or nitric oxide has implications for climate change. Interestingly the P460-containing hydroxylamine oxidoreductase utilises a tyrosine cross link to haem and performs similar chemistry. Previous crystal structures of cytochrome P460 from Nitrosomonas europaea (NeP460) clearly show the existence of a single crosslink between the NZ atom of lysine and the haem porphyrin, with mutagenesis studies indicating roles for the crosslink in positioning a proton transfer residue and/or influencing the distortion of the haem. Here we describe the evidence for a novel double crosslink between lysine and haem in the cytochrome P460 from Methylococcus capsulatus (Bath). In order to understand the complexities of this enzyme system we applied high resolution structural biology approaches at synchrotron and XFEL sources paired with crystal spectroscopies. Linked to this, we carried out QM/MM simulations that enabled the prediction of electronic absorption spectra providing a crucial validation to linking simulations and experimental structures. Our work demonstrates the feasibility of a double crosslink in McP460 and provides an opportunity to investigate how simulations can interact with experimental structures.

Show Abstract ▼

Abstract: Modern biology and biomedical research have been transformed by advances in optical techniques that enable the sensitive detection and spatially resolved characterization of biomolecules. Among these, fluorescence microscopy has emerged as a pivotal tool for visualizing cellular structures and dynamics, allowing researchers to observe biological systems with high specificity and temporal resolution. Having the ability to selectively label cellular components with fluorophores has revolutionized the fields of cell biology, neurobiology and medical diagnostics. Central to the success of fluorescent-based approaches has been the engineering of fluorescent proteins. From the discovery of the first fluorescent protein, the green fluorescent protein (GFP), to its subsequent optimization that led to a diverse family of genetically encoded fluorescent reporters and later the discovery and development of a repertoire of di>erent coloured proteins, there is a constant need for fluorescent proteins (FPs) with better features to fit di>erent analysis requirements. This thesis explores the engineering of fluorescent proteins (FPs) to enhance their utility in bioimaging and Raman-based spectroscopy. Focusing on mCherry, mNeptune, and superfolder GFP (sfGFP), the work employs site-directed mutagenesis and non-natural amino acid (nnAA) incorporation to modulate fluorescence properties and introduce vibrational signatures detectable by Raman spectroscopy. In Chapter 3, nnAAs—p-cyano-phenylalanine (pCNPhe) and p-ethynyl- phenylalanine (pCCPhe)—were incorporated into specific residues of mCherry and mNeptune to introduce C≡N and C≡C vibrational bonds. Structural modelling guided residue selection near the chromophore to maximize Raman coupling. Although protein expression was successful, many mutants lacked colour, suggesting disrupted chromophore maturation. Nonetheless, this work laid foundational strategies for genetically encoded Raman-active probes. Chapter 4 details the creation of a novel mCherry variant (mCherryM66C), where cysteine was introduced at position 66 of the chromophore. Spectral analysis showed a 21 nm blue shift in excitation and a 25 nm shift in emission, along with a modest increase in quantum yield. mCherryM66C displayed dual pH-dependent transitions (pKa = 5.7 and 8.8), consistent with a three-state chromophore model involving sequential protonation. While Raman signal was reduced compared to wild-type mCherry, the variant’s pH sensitivity and environmental responsiveness suggest potential use as a biosensor. Chapter 5 investigates the sfGFP H148S mutant (YuzuFP), developed through molecular dynamics simulations to enhance chromophore hydrogen bonding. Replacing histidine with serine resulted in increased brightness (~1.5-fold), improved absorbance, and significantly enhanced photobleaching resistance (~3-fold). Despite minimal change in pKa, YuzuFP maintained fluorescence more e>ectively across a wide pH range, a>irming its stability and suitability for imaging applications. Collectively, this work demonstrates the e>ectiveness of combining computational design with biochemical engineering to create next-generation FPs with enhanced spectral properties and environmental sensitivity. These engineered variants hold promise for advanced imaging techniques, including multiplexed fluorescence and Raman-based biosensing.

Open Access

Show Abstract ▼

Abstract: Nitroalkene fatty acids (NO2-FAs) are formed endogenously. They regulate cell signaling pathways and are being developed clinically to treat inflammatory diseases. NO2-FAs are electrophilic and form thioether adducts with glutathione (GSH), which are exported from cells. Glutathione transferases (GSTs), a superfamily of enzymes, contribute to the cellular detoxification of hydrophobic electrophiles by catalyzing their conjugation to GSH. Herein, we evaluated the capacity of five human GSTs (M1-1, M2-2, M4-4, A4-4, and P1-1) to catalyze the reaction between nitrooleic acid (NO2-OA) and GSH. The reaction was monitored by HPLC-ESI-MS/MS and catalytic activity was detected with hGSTs M1-1 and A4-4. Using stopped-flow spectrophotometry, a 1400 and 7500-fold increase in the apparent second-order rate constant was observed for hGST M1-1 and hGST A4-4, respectively, compared to the uncatalyzed reaction (pH 7.4, 25 °C), in part due to a higher availability of the thiolate. The crystal structure of hGST M1-1 in complex with the adduct was solved at 2.55 Å resolution, revealing that the ligand was bound within the reaction center, and establishing a foundation to build a model of hGST A4-4 in complex with the adduct. A larger number of interactions between the enzyme and the fatty acid were observed for hGST A4-4 compared to hGST M1-1, probably contributing to the increased catalysis. Altogether, these results show, for the first time, that hGSTs can catalyze the reaction between GSH and NO2-FAs, likely affecting the signaling actions of these metabolites and expanding the repertoire of GST reactions.