A Perspective on the Prospective Use of AI in Protein Structure Prediction.

AlphaFold2 (AF2) and RoseTTaFold (RF) have revolutionized structural biology, serving as highly reliable and effective methods for predicting protein structures. This article explores their impact and limitations, focusing on their integration into experimental pipelines and their application in diverse protein classes, including membrane proteins, intrinsically disordered proteins (IDPs), and oligomers. In experimental pipelines, AF2 models help X-ray crystallography in resolving the phase problem, while complementarity with mass spectrometry and NMR data enhances structure determination and protein flexibility prediction. Predicting the structure of membrane proteins remains challenging for both AF2 and RF due to difficulties in capturing conformational ensembles and interactions with the membrane. Improvements in incorporating membrane-specific features and predicting the structural effect of mutations are crucial. For intrinsically disordered proteins, AF2's confidence score (pLDDT) serves as a competitive disorder predictor, but integrative approaches including molecular dynamics (MD) simulations or hydrophobic cluster analyses are advocated for accurate dynamics representation. AF2 and RF show promising results for oligomeric models, outperforming traditional docking methods, with AlphaFold-Multimer showing improved performance. However, some caveats remain in particular for membrane proteins. Real-life examples demonstrate AF2's predictive capabilities in unknown protein structures, but models should be evaluated for their agreement with experimental data. Furthermore, AF2 models can be used complementarily with MD simulations. In this Perspective, we propose a "wish list" for improving deep-learning-based protein folding prediction models, including using experimental data as constraints and modifying models with binding partners or post-translational modifications. Additionally, a meta-tool for ranking and suggesting composite models is suggested, driving future advancements in this rapidly evolving field.

Evaluation of borinic acids as new, fast hydrogen peroxide-responsive triggers.

Hydrogen peroxide (H2O2) is responsible for numerous damages when overproduced, and its detection is crucial for a better understanding of H2O2-mediated signaling in physiological and pathological processes. For this purpose, various "off-on" small fluorescent probes relying on a boronate trigger have been prepared, and this design has also been involved in the development of H2O2-activated prodrugs or theranostic tools. However, this design suffers from slow kinetics, preventing activation by H2O2 with a short response time. Therefore, faster H2O2-reactive groups are awaited. To address this issue, we have successfully developed and characterized a prototypic borinic-based fluorescent probe containing a coumarin scaffold. We determined its in vitro kinetic constants toward H2O2-promoted oxidation. We measured 1.9 × 104 m-1⋅s-1 as a second-order rate constant, which is 10,000-fold faster than its well-established boronic counterpart (1.8 m-1⋅s-1). This improved reactivity was also effective in a cellular context, rendering borinic acids an advantageous trigger for H2O2-mediated release of effectors such as fluorescent moieties.

Quantitative live-cell imaging and 3D modeling reveal critical functional features in the cytosolic complex of phagocyte NADPH oxidase.

Phagocyte NADPH oxidase produces superoxide anions, a precursor of reactive oxygen species (ROS) critical for host responses to microbial infections. However, uncontrolled ROS production contributes to inflammation, making NADPH oxidase a major drug target. It consists of two membranous (Nox2 and p22phox) and three cytosolic subunits (p40phox, p47phox, and p67phox) that undergo structural changes during enzyme activation. Unraveling the interactions between these subunits and the resulting conformation of the complex could shed light on NADPH oxidase regulation and help identify inhibition sites. However, the structures and the interactions of flexible proteins comprising several well-structured domains connected by intrinsically disordered protein segments are difficult to investigate by conventional techniques such as X-ray crystallography, NMR, or cryo-EM. Here, we developed an analytical strategy based on FRET-fluorescence lifetime imaging (FLIM) and fluorescence cross-correlation spectroscopy (FCCS) to structurally and quantitatively characterize NADPH oxidase in live cells. We characterized the inter- and intramolecular interactions of its cytosolic subunits by elucidating their conformation, stoichiometry, interacting fraction, and affinities in live cells. Our results revealed that the three subunits have a 1:1:1 stoichiometry and that nearly 100% of them are present in complexes in living cells. Furthermore, combining FRET data with small-angle X-ray scattering (SAXS) models and published crystal structures of isolated domains and subunits, we built a 3D model of the entire cytosolic complex. The model disclosed an elongated complex containing a flexible hinge separating two domains ideally positioned at one end of the complex and critical for oxidase activation and interactions with membrane components.

The NADPH Oxidase and the Phagosome.

The key purpose of phagocytosis is the destruction of pathogenic microorganisms. The phagocytes exert a wide array of killing mechanisms that allow mastering the vast majority of pathogens. One of these mechanisms consists in the production of reactive oxygen species inside the phagosome by a specific enzyme, the phagocyte NADPH oxidase. This enzyme is composed of 6 proteins that need to assemble to form a complex on the phagosomal membrane. Multiple signaling pathways tightly regulate the assembly. We briefly summarize key features of the enzyme and its regulation. We then focus on several related topics that address the activity of the NADPH oxidase during phagocytosis. Novel fluorescence microscopy techniques combined with fluorescent protein labeling of NADPH oxidase subunits opened the view on the structure and dynamics of these proteins in living cells. This combination revealed details of the role of anionic phospholipids in the control of phagosomal ROS production. It also added critical information to propose a 3D model of the complex between the cytosolic subunits prior to activation, in complement to other structural data on the oxidase.

Bimodal Detection of Proteins by 129 Xe NMR and Fluorescence Spectroscopy.

A full understanding of biological phenomena involves sensitive and noninvasive detection. Herein, we report the optimization of a probe for intracellular proteins that combines the advantages of fluorescence and hyperpolarized 129 Xe NMR spectroscopy detection. The fluorescence detection part is composed of six residues containing a tetracysteine tag (-CCXXCC-) genetically incorporated into the protein of interest and of a small organic molecule, CrAsH. CrAsH becomes fluorescent if it binds to the tetracysteine tag. The part of the biosensor that enables detection by means of 129 Xe NMR spectroscopy, which is linked to the CrAsH moiety by a spacer, is based on a cryptophane core that is fully suited to reversibly host xenon. Three different peptides, containing the tetracysteine tag and four organic biosensors of different stereochemistry, are benchmarked to propose the best couple that is fully suited for the in vitro detection of proteins.

Biosensors for spatiotemporal detection of reactive oxygen species in cells and tissues.

Redox biology has become a major issue in numerous areas of physiology. Reactive oxygen species (ROS) have a broad range of roles from signal transduction to growth control and cell death. To understand the nature of these roles, accurate measurement of the reactive compounds is required. An increasing number of tools for ROS detection is available; however, the specificity and sensitivity of these tools are often insufficient. Furthermore, their specificity has been rarely evaluated in complex physiological conditions. Many ROS probes are sensitive to environmental conditions in particular pH, which may interfere with ROS detection and cause misleading results. Accurate detection of ROS in physiology and pathophysiology faces additional challenges concerning the precise localization of the ROS and the timing of their production and disappearance. Certain ROS are membrane permeable, and certain ROS probes move across cells and organelles. Targetable ROS probes such as fluorescent protein-based biosensors are required for accurate localization. Here we analyze these challenges in more detail, provide indications on the strength and weakness of current tools for ROS detection, and point out developments that will provide improved ROS detection methods in the future. There is no universal method that fits all situations in physiology and cell biology. A detailed knowledge of the ROS probes is required to choose the appropriate method for a given biological problem. The knowledge of the shortcomings of these probes should also guide the development of new sensors.

pH sensitivity of FRET reporters based on cyan and yellow fluorescent proteins.

It is generally acknowledged that the popular cyan and yellow fluorescent proteins carried by genetically encoded reporters suffer from strong pH sensitivities close to the physiological pH range. We studied the consequences of these pH responses on the intracellular signals of model Förster resonant energy transfer (FRET) tandems and FRET-based reporters of cAMP-dependent protein kinase activity (AKAR) expressed in the cytosol of living BHK cells, while changing the intracellular pH by means of the nigericin ionophore. Although the simultaneous pH sensitivities of the donor and the acceptor may mask each other in some cases, the magnitude of the perturbations can be very significant, as compared to the functional response of the AKAR biosensor. Replacing the CFP donor by the spectrally identical, but pH-insensitive Aquamarine variant (pK1/2 = 3.3) drastically modifies the biosensor pH response and gives access to the acid transition of the yellow acceptor. We developed a simple model of pH-dependent FRET and used it to describe the expected pH-induced changes in fluorescence lifetime and ratiometric signals. This model qualitatively accounts for most of the observations, but reveals a complex behavior of the cytosolic AKAR biosensor at acid pHs, associated to additional FRET contributions. This study underlines the major and complex impact of pH changes on the signal of FRET reporters in the living cell.

Cadmium disorganises the scaffolding of gap and tight junction proteins in the hepatic cell line WIF B9.

BACKGROUND INFORMATION: Hepatocytes, which perform the main functions of the liver, are particularly vulnerable to toxic agents such as cadmium, an environmental pollutant. To identify the molecular targets for cadmium in hepatocytes, we have studied the effects of CdCl2 on the hybrid cell line WIF-B9 that exhibits stable structural and functional hepatocytic polarity. RESULTS: We showed that the toxicity of CdCl2 (1 µM, 24 h) resulted in a reduction in direct intercellular communication (via gap junctions) and in an increase in paracellular permeability (decrease in the sealing of tight junctions). These effects were not related to changes in the expression of the key proteins involved, Cx32 and claudin 2, the first being constitutive of gap junctions and the second of tight junctions in this cell line. Using immunofluorescence experiments, we observed a change in the location of Cx32 and claudin 2: these two proteins were less often found in the tight junction network that closes the bile canaliculi (BC). In control cells, 'Proximity Ligation Assay' (PLA Duolink®) has confirmed in situ that molecules of claudin 2 and Cx32 are very close to each other at the BC (probably less than 16 nm). This was no longer the case after treatment with CdCl2 . Localisation of occludin and Cx32 relative to each other was not modified by CdCl2 , but CdCl2 increased the PLA signal between molecules of JAM-A and Cx32. Finally, examination of freeze-fracture replicas obtained from cultures treated with CdCl2 showed the disruption of the network of tight junctions and the depletion or the disintegration of the junctional plaques associated with tight junctions. CONCLUSIONS: This study demonstrates in situ the changes induced by cadmium on the organisation of cell-cell junctions and points out the importance of the association Cx32/claudin 2 for the maintenance of normal hepatocyte functions.

The enhanced cyan fluorescent protein: a sensitive pH sensor for fluorescence lifetime imaging.

pH is an important parameter that affects many functions of live cells, from protein structure or function to several crucial steps of their metabolism. Genetically encoded pH sensors based on pH-sensitive fluorescent proteins have been developed and used to monitor the pH of intracellular compartments. The quantitative analysis of pH variations can be performed either by ratiometric or fluorescence lifetime detection. However, most available genetically encoded pH sensors are based on green and yellow fluorescent proteins and are not compatible with multicolor approaches. Taking advantage of the strong pH sensitivity of enhanced cyan fluorescent protein (ECFP), we demonstrate here its suitability as a sensitive pH sensor using fluorescence lifetime imaging. The intracellular ECFP lifetime undergoes large changes (32 %) in the pH 5 to pH 7 range, which allows accurate pH measurements to better than 0.2 pH units. By fusion of ECFP with the granular chromogranin A, we successfully measured the pH in secretory granules of PC12 cells, and we performed a kinetic analysis of intragranular pH variations in living cells exposed to ammonium chloride.

An analytical workflow for the molecular dissection of irreversibly modified fluorescent proteins.

Owing to their ability to be genetically expressed in live cells, fluorescent proteins have become indispensable markers in cellular and biochemical studies. These proteins can undergo a number of covalent chemical modifications that may affect their photophysical properties. Among other mechanisms, such covalent modifications may be induced by reactive oxygen species (ROS), as generated along a variety of biological pathways or through the action of ionizing radiations. In a previous report [1], we showed that the exposure of cyan fluorescent protein (ECFP) to amounts of (•)OH that mimic the conditions of intracellular oxidative bursts (associated with intense ROS production) leads to observable changes in its photophysical properties in the absence of any direct oxidation of the ECFP chromophore. In the present work, we analyzed the associated structural modifications of the protein in depth. Following the quantified production of (•)OH, we devised a complete analytical workflow based on chromatography and mass spectrometry that allowed us to fully characterize the oxidation events. While methionine, tyrosine, and phenylalanine were the only amino acids that were found to be oxidized, semi-quantitative assessment of their oxidation levels showed that the protein is preferentially oxidized at eight residue positions. To account for the preferred oxidation of a few, poorly accessible methionine residues, we propose a multi-step reaction pathway supported by data from pulsed radiolysis experiments. The described experimental workflow is widely generalizable to other fluorescent proteins, and opens the door to the identification of crucial covalent modifications that affect their photophysics.

Förster resonance energy transfer efficiency measurements on vinculin tension sensors at focal adhesions using a simple and cost-effective setup.

Significance: Forces inside cells play a fundamental role in tissue growth, affecting important processes such as cancer cell migration or tissue repair after injury. Förster resonance energy transfer (FRET)-based tension sensors are a remarkable tool for studying these forces and should be made easier to use. Aim: We prove that absolute FRET efficiency can be measured on a simple setup, an order of magnitude more cost-effective than a standard FRET microscopy setup, by applying it to vinculin tension sensors (VinTS) at the focal adhesions of live CHO-K1 cells. Approach: Our setup located at Université Paris-Saclay acquires donor and acceptor fluorescence in parallel on two low-cost CMOS cameras and uses two LEDs for rapid switching of the excitation wavelength at a reduced cost. The calibration required to extract FRET efficiency was achieved using a single construct (TSMod). FRET efficiencies were measured for VinTS and the tail-less control VinTL, lacking the actin-binding domain of vinculin. Measurements were confirmed on the same cell type using a more standard intensity-based setup located at Rutgers University. Results: The average FRET efficiency of VinTS (22.0%±4%) over more than 10,000 focal adhesions is significantly lower (p<10-6) than that of VinTL (30.4%±5%), our control that is insensitive to force, in agreement with the force exerted on vinculin at focal adhesions. Attachment of the CHO-K1 cells on fibronectin decreases FRET efficiency, thus increasing the force, compared with poly-lysine. FRET efficiency for the VinTL control is consistent with all measurements currently available in the literature, confirming the validity of our measurements and hence of our simpler setup. Conclusions: Force measurements, resolved spatially inside a cell, can be achieved using FRET-based tension sensors with a cost effective intensity-based setup. This will facilitate combining FRET with techniques for applying controlled forces such as optical tweezers.

Consequences of the constitutive NOX2 activity in living cells: Cytosol acidification, apoptosis, and localized lipid peroxidation.

The phagocyte NADPH oxidase (NOX2) is a key enzyme of the innate immune system generating superoxide anions (O2•-), precursors of reactive oxygen species. The NOX2 protein complex is composed of six subunits: two membrane proteins (gp91phox and p22phox) forming the catalytic core, three cytosolic proteins (p67phox, p47phox and p40phox) and a small GTPase Rac. The sophisticated activation mechanism of the NADPH oxidase relies on the assembly of cytosolic subunits with the membrane-bound components. A chimeric protein, called 'Trimera', composed of the essential domains of the cytosolic proteins p47phox (aa 1-286), p67phox (aa 1-212) and full-length Rac1Q61L, enables a constitutive and robust NOX2 activity in cells without the need of any stimulus. We employed Trimera as a single activating protein of the phagocyte NADPH oxidase in living cells and examined the consequences on the cell physiology of this continuous and long-term NOX activity. We showed that the sustained high level of NOX activity causes acidification of the intracellular pH, triggers apoptosis and leads to local peroxidation of lipids in the membrane. These local damages to the membrane correlate with the strong tendency of the Trimera to clusterize in the plasma membrane observed by FRET-FLIM microscopy.

Cyan fluorescent protein carries a constitutive mutation that prevents its dimerization.

The tendency of GFP-like fluorescent proteins to dimerize in vitro is a permanent concern as it may lead to artifacts in FRET imaging applications. However, we have found recently that CFP and YFP (the couple of GFP variants mostly used in FRET studies) show no trace of association in the cytosol of living cells up to millimolar concentrations. In this study, we investigated the oligomerization properties of purified CFP, by fluorescence anisotropy and sedimentation velocity. Surprisingly, we found that CFP has a much weaker homoaffinity than other fluorescent proteins (K(d) ≥ 3 × 10(-3) M), and that this is due to the constitutive N146I mutation, originally introduced into CFP to improve its brightness.

Quenched or alive quantum dots: The leading roles of ligand adsorption and photoinduced protonation.

HYPOTHESIS: The fluorescence emission of water-soluble CdTe quantum dots (QDs) capped with mercaptocarboxylic acids (MCAs) is known to be pH-dependent. However, this behaviour is quite different from a study to another, so that literature suffers from a lack of coherence. Here we assume that the QD fluorescence efficiency is actually driven by the acid-base equilibrium of MCA thiol groups, and that light-excited QDs open a non-radiative relaxation path through photoinduced protonation. EXPERIMENTS: We address this issue by examining colloidal CdTe QDs with (time-resolved) fluorescence spectroscopy under various conditions of acidity and light excitation. FINDINGS: It appears that the emission of QDs is quenched below a critical pH value of 6.87, and that light excitation power strengthens this quenching. We thus demonstrate the existence of an additional photochemical process and developed a mathematical modeling accounting for all our experimental results. With only three parameters, it is possible to accurately predict the fluorescence decay of QDs over time, at any pH. Further, we also related the critical pH value of 6.87 to QD surface properties, explaining why observations may differ from a study to another and making the literature much more coherent.

tdLanYFP, a Yellow, Bright, Photostable, and pH-Insensitive Fluorescent Protein for Live-Cell Imaging and Förster Resonance Energy Transfer-Based Sensing Strategies.

Yellow fluorescent proteins (YFPs) are widely used as optical reporters in Förster resonance energy transfer (FRET)-based biosensors. Although great improvements have been done, the sensitivity of the biosensors is still limited by the low photostability and the poor fluorescence performances of YFPs at acidic pH values. Here, we characterize the yellow fluorescent protein tdLanYFP, derived from the tetrameric protein from the cephalochordate Branchiostoma lanceolatum, LanYFP. With a quantum yield of 0.92 and an extinction coefficient of 133,000 mol-1·L·cm-1, it is, to our knowledge, the brightest dimeric fluorescent protein available. Contrasting with EYFP and its derivatives, tdLanYFP has a very high photostability in vitro and in live cells. As a consequence, tdLanYFP allows imaging of cellular structures with subdiffraction resolution using STED nanoscopy and is compatible with the use of spectromicroscopies in single-molecule regimes. Its very low pK1/2 of 3.9 makes tdLanYFP an excellent tag even at acidic pH values. Finally, we show that tdLanYFP is a valuable FRET partner either as a donor or acceptor in different biosensing modalities. Altogether, these assets make tdLanYFP a very attractive yellow fluorescent protein for long-term or single-molecule live-cell imaging including FRET experiments at acidic pH.

Mass spectrometry-based structural dissection of fluorescent proteins.

Fluorescent proteins (FPs) are essential for live cell studies using fluorescence microscopy. To date, the molecular basis for FPs' irreversible photobleaching and the nature of the associated photoproducts are a matter of debate. Mass spectrometry, which should be an ideal technique for the structural dissection of FPs, cannot be harnessed efficiently due to their extreme resistance to trypsinolysis, due to the compactness of the barrel structure containing the chromopeptide. We devised a mild endoproteolysis procedure that affords a peptide mass fingerprint almost totally covering the sequence, thus allowing high-resolution mass spectrometric investigations of the protein structure.

Kinetic Analysis of Phagosomal ROS Generation.

Phagosomal ROS generation is critical for our immune defense against microbial infections. Quantitative assessment of phagosomal ROS production is required to understand the complex relationship between the phagocyte and the microbe, in particular for pathogens that resist phagosomal destruction. ROS detection is difficult due to the transient nature of the reactive species and their multiple interactions with the environment. Direct labeling of phagocytic prey with a ROS-sensitive dye allows to target the dye into the phagosome and to follow the kinetics of phagosomal ROS production on a single phagosome base. Here we describe the basic labeling procedure, the quality assessment, and the imaging technique to achieve this kinetic analysis.

Optimized FRET Pairs and Quantification Approaches To Detect the Activation of Aurora Kinase A at Mitosis.

Genetically encoded Förster's Resonance Energy Transfer (FRET) biosensors are indispensable tools to sense the spatiotemporal dynamics of signal transduction pathways. Investigating the crosstalk between different signaling pathways is becoming increasingly important to follow cell development and fate programs. To this end, FRET biosensors must be optimized to monitor multiple biochemical activities simultaneously and in single cells. In addition, their sensitivity must be increased to follow their activation even when the abundance of the biosensor is low. We describe here the development of a second generation of Aurora kinase A/AURKA biosensors. First, we adapt the original AURKA biosensor-GFP-AURKA-mCherry-to multiplex FRET by using dark acceptors as ShadowG or ShadowY. Then, we use the novel superYFP acceptor protein to measure FRET by 2-color Fluorescence Cross-Correlation Spectroscopy, in cytosolic regions where the abundance of AURKA is extremely low and undetectable with the original AURKA biosensor. These results pave the way to the use of FRET biosensors to follow AURKA activation in conjunction with substrate-based activity biosensors. In addition, they open up the possibility of tracking the activation of small pools of AURKA and its interaction with novel substrates, which would otherwise remain undetectable with classical biochemical approaches.

Complex fluorescence of the cyan fluorescent protein: comparisons with the H148D variant and consequences for quantitative cell imaging.

We have studied the fluorescence decays of the purified enhanced cyan fluorescent protein (ECFP, with chromophore sequence Thr-Trp-Gly) and of its variant carrying the single H148D mutation characteristic of the brighter form Cerulean. Both proteins exhibit highly complex fluorescence decays showing strong temperature and pH dependences. At neutral pH, the H148D mutation leads (i) to a general increase in all fluorescence lifetimes and (ii) to the disappearance of a subpopulation, estimated to be more than 25% of the total ECFP molecules, characterized by a quenched and red-shifted fluorescence. The fluorescence lifetime distributions of ECFP and its H148D mutant remain otherwise very similar, indicating a high degree of structural and dynamic similarity of the two proteins in their major form. From thermodynamic analysis, we conclude that the multiexponential decay of ECFP cannot be simply ascribed, as is generally admitted, to the slow conformational exchange characterized by NMR and X-ray crystallographic studies [Seifert, M. H., et al. (2002) J. Am. Chem. Soc. 124, 7932-7942; Bae, J. H., et al. (2003) J. Mol. Biol. 328, 1071-1081]. Parallel measurements in living cells show that these fluorescence properties in neutral solution are very similar to those of cytosolic ECFP.

Triangulation mapping of oxidative bursts released by single fibroblasts by amperometry at microelectrodes.

It has been previously established that a lesion created by a microcapillary in the membrane of a single aerobic cell (from skin or immune origin) was sufficient to induce a local membrane depolarization and the ensuing release of oxidative bursts. Their kinetic and quantitative features reveal the activity of cell constitutive enzymes, namely, NADPH oxidases and NO synthases, prone to produce rapidly reactive oxygen and reactive nitrogen species. Until now, the spatial resolution provided by microelectrodes has been exploited in this context to characterize the chemical composition of oxidative bursts at several cell types with high collection efficiency. In the present work, spatial features of the oxidative bursts from single human fibroblasts were investigated using a step-by-step geometrical mapping approach. The spatial locations of cell active zones and of the extent of the activated area, when a cell membrane was stressed by a microcapillary's tip of 1-microm radius, have been addressed. On cells of large dimensions such as fibroblasts, ROS and RNS emission originated from a disk surface of the membrane limited to approximately 15-microm radius around the approximately 1-microm hole created by the microcapillary. This experimental result was rationalized through a simple physicochemical model designed to portray the extent of the membrane activated area due to ion concentration variations resulting from the pinhole channel created across the cell membrane. This is consistent with the fact that the activation of constitutive enzymatic complexes (NOX and NOS) is hypothesized to be a consequence of local variations of ion concentrations such as K(+), Na(+) or possibly Ca(2+). Our results showed that the calculated area near the cell membrane where the ion concentration gradients are significant was equivalent to the area of species release measured experimentally.