Synchrotron-based correlative imaging of metals and proteins in neuronal cells: state of the art and future challenges in neurometallomics

Abstract

Metal homeostasis in the nervous system is subtly regulated and changes in metal distribution or content, either increases or decreases, are associated with neurodegeneration or cognitive impairment. Determining the localization and quantification of metals in different types of neurons is important information for understanding their role in neurobiology. Synchrotron X-ray fluorescence imaging is a powerful technique that provides very high sensitivity and high spatial resolution for imaging metals in cells. However, additional biological information is often required to correlate the subcellular localization of metals with specific proteins or organelles. The purpose of this article is to review the studies in neuroscience that correlate metal imaging by synchrotron X-ray fluorescence with protein localization by other techniques. This article highlights the diversity of correlative modalities that have been used, from fluorescence to super-resolution and infrared microscopy, and the wealth of information that has been extracted, but also discusses some current limitations. Future developments are needed, particularly for direct imaging of metals and proteins with a single instrument.

Graphical Abstract

Open in new tabDownload slide

Introduction

Metals such as iron (Fe), copper (Cu), zinc (Zn), and manganese (Mn) play an essential physiological role in all human cells, including cells from the nervous system. These metals are involved in broad physiological functions such as metabolism, energy production, protein synthesis, DNA synthesis and repair, detoxification, protection against oxidative stress, and more. They are also involved in more specific functions in neurons, for instance, Zn and Cu can modulate synaptic activity and plasticity, influencing learning and memory [1, 2]. The homeostasis of physiological metals is tightly regulated in the nervous system, and any alteration in metal distribution and availability, either increases or decreases, can lead to cellular dysfunctions, contributing to neurodegenerative diseases and cognitive decline [3, 4]. Metal dyshomeostasis is observed in most neurodegenerative disorders such as Alzheimer's, Parkinson's, and Huntington's diseases, amyotrophic lateral sclerosis, and multiple sclerosis [5–10]. Understanding and targeting metal homeostasis in the brain is thus critical for developing therapeutic interventions for neurodegeneration. On the other hand, neurotoxic metals such as lead (Pb), mercury (Hg), arsenic (As), and cadmium (Cd) are known to induce deleterious effects on the nervous system through mechanisms that have been increasingly investigated. Although significant progress has been made, including studies using synchrotron-based techniques [11], the precise cellular and molecular pathways of neurotoxicity remain incompletely understood [12].

Knowledge of the subcellular distribution of physiological or toxic metals in nervous cells can contribute to revealing their neurobiological functions or their mechanisms of neurotoxicity. However, imaging metals in neurons at high spatial resolution presents significant challenges due to their low concentrations, at trace and ultra-trace levels. Synchrotron X-ray fluorescence (SXRF) microscopy has emerged as one of the most appropriate analytical techniques for precise subcellular metal detection [13–15]. In the latest generation synchrotron radiation facilities, SXRF microscopy can be performed at spatial resolutions down to few tens of nanometers, with extremely brilliant X-ray sources allowing ultimate sensitivity in metal detection. While SXRF is very useful for determining the subcellular distribution and quantification of metals, it requires additional cytological information to understand metal functions. Several research teams have developed correlative imaging approaches to explain metal distributions obtained by SXRF in their biological context [14, 16]. To pinpoint the precise distribution of elements, it is helpful to investigate their co-localization with proteins or with cellular organelles; that can be achieved by labeling the organelle-specific proteins.

In this article, we will briefly describe the principle of SXRF microscopy, and review the development of subcellular metal and protein correlative imaging combining SXRF with various other protein imaging modalities such as epifluorescence, confocal, and super-resolution microscopy or infrared microscopy. We will discuss the critical question of sample preparation that should address both SXRF and protein identification constraints, most frequently requiring cryogenic protocols to preserve the native state of the samples. Imaging metals and proteins in neurons remains a challenging area of research, requiring continued development of innovative tools. We will discuss new perspectives of improvement by fully cryogenic correlative modalities, or by designing metal-binding molecules to perform metal and protein localization at the same time by SXRF.

SXRF for metal imaging

When electrons are accelerated in a synchrotron-type facility, the light emitted by these relativistic electrons undergoing centripetal acceleration is called synchrotron radiation. The characteristics of synchrotron radiation make it far more suitable for spectroscopic analysis than conventional light sources [17]. Spectral continuity, from the far-infrared to X-rays, is not achieved by any other source over such a wide range of wavelengths. The low emission divergence in the vertical plane, of the order of 0.1 mrad, leads to a brightness several orders of magnitude greater than that of X-ray tubes, and enables excellent beam focusing. Because of all these features, several synchrotron radiation facilities have been built around the world. The high flux and low divergence of synchrotron radiation can be exploited by focusing systems (Fresnel lenses, Kirkpatrick–Baez-type mirrors) to reduce beam size to micrometric or even nanometric dimensions [18, 19]. Thanks to these focusing optics, operating in the hard X-ray energy spectrum, SXRF technique can be implemented at micro- and nanoscopic scales.

SXRF is based on the photoelectric principle. When a photon beam interacts with the atoms of a sample, the photon can be absorbed and an electron can be ejected from the inner shells of the atom, causing an electron from a higher shell to fall into this lower shell to fill the vacancy left behind, simultaneously emitting an X-ray with an energy equal to the energy difference between the two electron shells involved (Fig. 1a), this energy being characteristic of the element that emitted the X-ray (Fig. 1c). SXRF is an analytical method that can achieve high spatial resolution and low detection limits, making it ideal for imaging low-concentration chemical elements in biological systems [20–24]. The detection limit of SXRF is typically around 0.1 µg g⁻¹ [22], which is compatible with the imaging of most metals of interest in the brain, either exogen ous essential metals such as Fe (20–50 µg g⁻¹ range), Zn (10–20 µg g⁻¹ range), or Cu (2–5 µg g⁻¹ range), or exogenous contaminants such as As, Ni, Pb, and Cr present in the 0.1–1 µg g⁻¹ range [25]. The X-ray beam energy can be finely selected with high energy resolution, ΔE/E typically in the range of 10⁻³–10⁻⁴, just above the absorption edge of the element of interest, which significantly increases the signal-to-noise ratio and reduces the detection limits compared to conventional X-ray sources. SXRF allows multielement analysis (Fig. 1c and d). SXRF is also a quantitative analytical technique. There is no standardized procedure for quantitative data analysis, as it varies according to the detection system of each synchrotron beamline. Element quantification is usually carried out in close collaboration with beamline staff. Detailed procedures describing data processing and element quantification in biological samples have been published previously [21, 24, 26–29]. Common general principles can be highlighted. First, SXRF spectra acquired on samples are fitted with appropriate software and then the quantitative distribution of elements is calculated in proportion to the signal obtained on certified reference materials. The elemental content is expressed in terms of mass per unit area, usually in ng mm⁻². These values can be converted to ng g⁻¹ or µM concentrations when the thickness and density of the samples are known.

Multielement and quantitative analytical capabilities are of great interest to compare metal concentrations at the subcellular level in different neuropathological processes to detect accumulation, deficiency, or dyshomeostasis of any element in the samples. For example, neurotoxic metals may substitute for essential metals as they enter the cell, thereby disrupting their function and localization [12]. The capabilities of SXRF are therefore of particular interest as elemental co-localization and quantification can be demonstrated at the single-cell level.

Sample preparation

Sample preparation is a critical issue, as chemical elemental composition and distribution must be preserved, since SXRF has very low detection limits and very high spatial resolution. Protein labeling must be compatible with SXRF analysis and avoid elemental modification. Sample preparation protocols must meet three criteria: (i) preserve cell structure; (ii) not alter the distribution of chemical elements, especially diffusible ions; and (iii) avoid the introduction of exogenous chemical elements or contaminants.

The invasive sample preparation protocols typically used in biology for immunolabeling include chemical fixation with detergents and aldehydes, and dehydration steps with organic solvents. These steps often lead to significant changes in the elemental composition and distribution in biological samples [30–38]. For example, many studies have systematically shown that aldehyde-based fixation disrupts the integrity of the plasma membrane, creating pores through which diffusible ions can escape, resulting in a drastic loss of diffusible elements such as K, Ca, Cl, and Br [30–35, 38], but also of more tightly bound elements such as P, Mn, Cu, and Zn [31, 34, 36, 37]. It is important to note that membrane permeabilization not only affects the cell composition in terms of inorganic elements but also induces the loss of up to 40% of the intracellular organic composition [31]. In addition, the chemicals used for fixation and dehydration introduce exogenous elements that dramatically alter the content and distribution of intracellular elements such as P, Na, S, Ca, Fe, Cu, and Zn [30, 32–35]. Cryogenic fixation methods have been shown to be more effective in preserving the native state of cells, avoiding the loss of diffusible elements, preventing redistribution, and exogenous element contamination [30, 31, 33–35]. These methods are particularly suitable for SXRF analysis and correlative imaging as they maintain the elemental and structural integrity of the sample. Although protocols of immunolabeling are not optimal for the study of intracellular elements due to these limitations, they can still be valuable in certain contexts, such as the study of tightly bound elements. This highlights the importance of understanding the limitations of each preparation method in order to select the most appropriate protocol depending on the scientific question to be addressed.

Cryogenic methods have been developed for the rapid immobilization of the sample using plunge-freezing (Fig. 2), thus maintaining cellular structures and elemental distributions at a given instant close to their native state. Typically, to perform cryogenic preparation of cultured cells, such as dissociated neurons, cells are cultured on thin silicon nitride membranes, which are today the most commonly used supports for synchrotron imaging [30]. These membranes are first sterilized, and then coated with appropriate cell adhesion factors such as polylysine. Cultures are maintained at 37°C with a 5% CO₂ atmosphere and neurons can be exposed to extracellular compounds, such as metals, or fluorescently labeled molecules for live-cell imaging. Extracellular culture medium is rinsed with ammonium acetate solution of the same pH and osmolarity and the excess rinsing solution is blotted with a filter paper. Cells are then cryofixed by plunge freezing in liquid ethane. Samples can be analyzed by SXRF at room temperature after freeze-drying, or in their frozen hydrated state [39].

Correlation of SXRF imaging with light microscopy for protein localization

To investigate the role of physiological metals in neurobiology and their dyshomeostasis in neurodegenerative diseases, it is needed to colocalize the proteins with which these metals interact at the subcellular level. In general, metals will be associated with metal transport proteins, metal storage proteins, and with metalloproteins where they exert their function as enzymatic or structural cofactors. Proteins of interest can be labeled by various means and observed in different ways. In this review article, we have structured the following sections according to the method of observation.

Correlative imaging of SXRF with epifluorescence and confocal microscopy

Fluorescence microscopy is an imaging technique widely used in biomedical research to visualize and study biological molecules at the subcellular level. The principle of the method is based on the excitation of samples labeled with fluorochromes, molecules capable of emitting light when exposed to a light source of a specific wavelength. Coupling fluorochromes to proteins of interest is a fairly common and routine procedure for imaging proteins inside cells using epifluorescence or confocal microscopy. A widely used approach is immunofluorescence, in which a fluorescent antibody is introduced into the cell to bind to the protein of interest. However, this protocol requires chemical fixation and permeabilization of cell membranes. This type of procedure is generally avoided for the study of metal quantification and imaging in biological samples due to the labile nature of metal binding as explained in the previous section.

An alternative strategy is to use fluorescent proteins that are compatible with live-cell observations. One of the most widely used approaches is cellular transfection with plasmids encoding fusion proteins, in which a protein of interest is fused to a fluorophore such as green fluorescent protein (GFP) (Fig. 3). This method is often used to visualize protein localization and dynamics in living cells. Fluorochromes can also be directly coupled to proteins of interest by specific chemical reactions, e.g. with organic fluorescent dyes (Figs 3 and 5).

In an epifluorescence microscope, the emitted fluorescence is separated from the light of the source by the use of a spectral emission filter. The source of illumination is typically provided by a mercury lamp, or more recently by Light-Emitting Diodes (LEDs). Depending on the excitation filter, it is possible to select the wavelength of light that passes through the objective and reaches the sample. By selecting an incident wavelength that excites the fluorochrome in the cell, the fluorescent molecules are excited and emit light of a specific wavelength in response. These photons are emitted in all directions and only a fraction is collected by the objective. This light emitted back by the sample passes is filtered according to the absorption filters and sent to the camera port above. By completely filtering the excitation light without blocking the emitted fluorescence, it is possible to see only those objects that are fluorescent. Epifluorescence microscopy has its limitations, in particular the contribution of out-of-plane fluorescence, which can degrade image quality and reduce contrast. This out-of-plane fluorescence results from the excitation of fluorophores located at different depths in the specimen, making it difficult to distinguish between the different layers of a thick specimen. Confocal microscopy overcomes this limitation.

Confocal microscopy is an advanced optical microscopy technique for obtaining high-resolution, three-dimensional images of thick samples. Unlike epifluorescence microscopy, which illuminates the entire sample, confocal microscopy uses a focused point of light to illuminate the sample and a pinhole placed at the conjugation of the focal plane to eliminate out-of-focus light. The pinhole blocks light from the out-of-focus planes, allowing only light from the focal plane to be collected. By scanning this focal plane through the sample, layer by layer, it is possible to reconstruct a three-dimensional image of the sample, a process automated by a laser scanning system coupled to a detection device such as a photomultiplier or CCD camera.

For neurological studies, SXRF has been combined with epifluorescence and confocal microscopy to correlate subcellular metal distribution with organelle localization using organelle-specific fluorescent dyes compatible with live-cell imaging. For example, Grubman et al. studied the subcellular distribution of Ca and Zn in a cellular model of a childhood neurodegenerative disorder, neuronal ceroid lipofuscinosis (Batten disease) [40]. Ca and Zn distribution was compared with mitochondrial and endoplasmic reticulum localization, stained with MitoTracker^® Deep Red and ER-Tracker™ Green, respectively, and observed by confocal microscopy. In this study, SXRF revealed significant subcellular mislocalization of both Ca and Zn in a diseased cell model compared to control cells. Other organelle fluorescent labeling platforms are available, such as the CellLight^TM reagents which are fluorescent protein-signal peptide fusions for live-cell imaging. Cellular labeling utilizes BacMam technology, which uses an insect cell virus (baculovirus) coupled with a mammalian promoter. For instance, CellLight^TM Golgi-RFP (red fluorescent protein) was used to reveal the accumulation of Mn in cells expressing a mutation in the Slc30a10 gene that has been identified in new forms of familial parkinsonism [41]. CellLight^TM Golgi-RFP targets the sequence of the human Golgi-resident enzyme N-acetylgalactosaminyltransferase 2. In this same study, the localization of the Slc30a10-GFP protein enabled identification of the cells that were correctly transfected and showing GFP fluorescence (Fig. 3). In another study, Gräfenstein et al. correlated SXRF imaging of P, S, K, and Zn and fluorescence distribution of HeLa cells overexpressing the fluorescence construction of huntingtin protein exon 1 [42]. They compared the native protein that remains soluble and the pathogenic case where it forms aggregates, and they found a homogenous partitioning of elements for both cytoplasmic and aggregate forms.

Correlative imaging of SXRF with super-resolution microscopy

The recent development of various light fluorescence microscopy techniques that enables super-resolution imaging of biological samples with spatial resolution not limited by the diffraction limit of light has considerably improved the investigation of protein arrangements in subcellular compartments, especially in neurons [43]. One notable combination is the integration of stimulated emission depletion (STED) live-cell microscopy [44, 45], with subsequent nano-SXRF imaging to achieve detailed visualization of the distribution of metals and proteins within neuronal compartments [46]. For STED microscopy, in addition to the excitation laser as used for confocal microscopy, a torus-shaped laser depletes the peripheral fluorescence of the focal point to keep only the core fluorescence. There is no isotropic fluorescence emission around the excitation point and the diffraction spot then becomes smaller, which allows for a considerable increase in spatial resolution.

In the study by Domart et al. [46], correlative imaging was employed to investigate the cosegregation of Cu and Zn with cytoskeletal proteins (tubulin and F-actin) fluorescently labeled with silicon-rhodamine dyes, in the dendrites and synapses of cultured rat hippocampal neurons. This correlative approach, using STED to localize proteins and SXRF to map metals, both performed at similar spatial resolution (35 nm), revealed critical interactions between Cu and Zn with tubulin and F-actin, which are involved in synaptic plasticity and memory formation.

In another study, the same group provided critical insights into the structural organization of chemical elements in growth cones during neuronal development. By employing a combination of STED super-resolution microscopy and SXRF, they demonstrated that elements such as Ca, Zn, and K are asymmetrically distributed within active growth cones, particularly in regions rich in F-actin, which are essential for neurite pathfinding. These findings highlighted the critical role of biologically active metals in the dynamic organization of the neuronal cytoskeleton, reinforcing the importance of correlative imaging techniques in elucidating metal–protein interactions at the nanometric scale [47]. The experimental validation of correlative nanoimaging of metals and proteins in primary neurons using SXRF and STED microscopy has been published by the same group. This methodology enabled the nanometric comparison of metal distribution with that of synaptic proteins such as postsynaptic density 95 (PSD95) and cytoskeletal components. The ultrasensitive nature of SXRF allowed for precise mapping of diluted essential metals, including Zn and Cu, providing valuable insights into their roles in synaptogenesis and neuronal plasticity [48].

The examples above demonstrate that super-resolution imaging techniques such as STED microscopy offer significant advantages for correlative imaging with SXRF. By combining the spatial resolution of SXRF, for metal mapping, with the molecular specificity and high spatial resolution of STED, for protein localization, a more comprehensive understanding of the underlying biological functions of metals can be achieved.

Correlative imaging of SXRF with Fourier transform infrared

Fourier transform infrared (FTIR) imaging is a powerful technique that combines infrared spectroscopy with imaging technology to provide detailed chemical information at high spatial resolution about chemical bonds and molecular structure. FTIR imaging can be used to visualize proteins by detecting specific vibrational modes in the infrared spectrum that correspond to the chemical bonds and secondary structures of proteins. For instance, amide I band (around 1600–1700 cm^−¹) is primarily associated with the C=O stretching vibration of the peptide bond and provides information on the protein's secondary structure (α-helices, β-sheets, turns, and random coils). Amide II band (around 1500–1600 cm^−¹) arises from the N–H bending and C–N stretching vibrations of the peptide bond. FTIR imaging is useful for detecting and mapping areas of protein aggregation, which can occur in diseases such as amyloidosis or neurodegenerative disorders. FTIR imaging informs not only about chemical bounds associated to proteins but also about some other important biomolecules, such as lipids. When combined together, SXRF and FTIR are very complementary, informing both on the inorganic and organic chemical distributions.

The combination of FTIR and SXRF in the context of neurodegenerative disorders has provided valuable information on the composition of amyloid-β plaques in Alzheimer's disease (AD). In brain tissues of AD, Miller et al. identified by FTIR elevated β-sheet content in amyloid plaques as well as hot spots of Zn and Cu by SXRF [49]. In another study, FTIR and SXRF were used to image aggregated proteins and metals (Fe, Cu, and Zn) in the brain hippocampus of a murine model of AD [50]. FTIR and SXRF were performed on serial tissue sections of mouse hippocampus, showing intense Fe, Cu, and Zn enrichment within amyloid-β plaques (Fig. 4). Levels of Cu, lipid ester, and lipid methylene groups were particularly elevated at the plaque periphery, and levels of Zn, Fe, and aggregated protein were elevated in the plaque core. Similar results were later observed in human samples showing the presence of Cu, Fe, and Zn in amyloid plaques [51]. Two types of amyloid deposits were identified in tissue sections of the cortex of patients in advanced stages of AD, fibrillary and nonfibrillary plaques. The analysis of the FTIR spectra has allowed correlation of lipid oxidation with the presence of nonfibrillary plaques and XRF showed higher accumulation of Fe in fibrillary plaques than in nonfibrillary plaques. These results revealed a different metal composition in the two identified types of aggregated amyloid plaques.

FTIR spectro-microscopy and SXRF were combined by Ducic et al. to analyze aggregates in an Amyotrophic Lateral Sclerosis (ALS) model of astrocytes. They demonstrated changes in the Cu concentration, in the amount of antiparallel β-sheet structures and in lipid localization and composition, within cells of ALS models, indicating a link between increased Cu concentration and protein aggregation [52].

FTIR imaging is usually performed at spatial resolutions >3 µm with commercial instruments, but recent developments allow IR spectroscopy to be performed at higher resolution, down to 300 nm. For example, optical photothermal infrared (O-PTIR) is a label-free super-resolution IR microspectroscopy technique that can be applied for subcellular imaging. The combination of O-PTIR and SXRF nanoimaging allowed capture of subcellular Fe clusters colocalizing with elevated amyloid β-sheet structures and oxidized lipids in cultured primary neurons at 300 nm spatial resolution [53].

Perspectives and challenges

Full cryo-correlative imaging

In the context of correlative imaging of metals and proteins, full cryo-correlative imaging, which combines cryo-fluorescence light microscopy (FLM) and cryo-SXRF, can be particularly useful. Cryo-FLM is a technique derived from traditional fluorescence microscopy, designed to observe biological samples under cryogenic conditions. The fundamental principle of cryo-FLM is based on preserving the sample in a vitrified state, i.e. rapidly frozen without the formation of ice crystals, thus preserving its hydration and native structure and maintaining the distribution and organization of intracellular elements close to the in vivo state. After labeling the molecules of interest with specific fluorophores, the sample is vitrified using a cryogenic fluid (i.e. liquid ethane), and then stored in liquid nitrogen. In addition, cryogenic conditions offer greater fluorophore stability, considerably reducing the risk of photobleaching during sample illumination, a crucial advantage for a prolonged or repeated observation of samples. Cryo-FLM is therefore particularly useful for studying delicate cellular structures, or for preserving samples in a state as close as possible to that observed under native conditions.

Cryo-FLM can be combined with cryo-SXRF when synchrotron beamlines are equipped with cryogenic sample stages as developed on several nanoimaging beamlines [54–57]. The implementation of such cryostats in synchrotron systems exemplifies the technological advancements enabling nanometric analyses under cryogenic conditions, preserving sample integrity for prolonged studies while minimizing radiation damage risks. Correlative cryo-SXRF and cryo-FLM have been applied to a variety of studies. For example, Conesa et al. combined in cryo-conditions SXRF to image an iridium-based anticancer agent within MCF7 cancer cells with the cryo-FLM of mitochondrial and acidic organelles markers [58]. Similarly, the cellular distribution and quantification of iridium was correlated to actin protein, mitochondria, and nucleus distributions using correlative cryo-SXRF and cryo-FLM [59]. In another study of anticancer agents, the intracellular behavior of functionalized Se nanoparticles, which showed antiproliferative effects on prostate cancer cells, correlative cryo-SXRF and cryo-FLM demonstrated that Se reaches lysosomes in prostate cancer cells labeled with lysosomes-RFP CellLight fluorescent probe [60].

Cryo-FLM combined with cryo-SXRF has also been applied in the field of neuroscience, enabling the imaging of developing hippocampal neurons [34]. Cytoskeletal proteins (tubulin and F-actin) were labeled with silicon rhodamine dyes designed to bind specifically to tubulin or F-actin, vitrified and observed by cryo-FLM. Subsequently, the elemental composition of neurons was imaged by cryo-SXRF. This study revealed the interactions between cytoskeletal structure and physiological metals (Cu and Zn), offering new perspectives for studying the functions of these metals in neurodevelopment (Fig. 5) [39].

One of the main limitations to the dissemination of this fully cryo-correlative imaging approach is the limited availability of cryo-fluorescence light microscopes. These microscopes are still expensive and rare, but are available at an increasing number of facilities, mainly for cryo-light electron microscopy (CLEM) applications, and can be adapted for cryo-light X-ray microscopy. Cryo-FLM is already available directly at some synchrotron facilities, and cryo-FLM observations can also be performed independently at CLEM facilities, where samples can be kept frozen in liquid nitrogen until synchrotron analysis. Cryo-correlative imaging already stands out as a powerful tool for exploring the structural and chemical composition of cellular and biological tissue components under near-native conditions, enabling precise analysis of both biological and therapeutic elements at nanometric resolution. This approach will open vast research opportunities, particularly in biomedicine, by providing insights into cellular mechanisms and advancing the development of targeted treatments with minimized sample alteration.

Direct correlative imaging of metals and proteins with metal tags

The design of metallic tags to label proteins is of great interest since SXRF imaging would enable imaging of the proteins of interest and the cellular metal content at once. The proof of principle was provided in 2006 by McRae et al. [61]. In this study, they coupled FluoroNanogold, an organic fluorophore with a 1.4 nm gold cluster, as a secondary antibody to a monoclonal antibody associated with mitochondria or Golgi apparatus proteins. Combining the micro-XRF technique and optical immunofluorescence microscopy, it was possible to compare the distribution of organelles labeled with gold to intracellular elements such as Zn in cultured cells. However, since cell permeabilization and chemical fixation were needed to introduce the antibodies into the cells, this protocol may induce element loss and redistribution. In 2020, the use of lanthanide-binding tags for 3D X-ray imaging of proteins in cells with nanoscale resolution emerged as a novel tool to enable precise visualization of the localization and structure of proteins within cells, facilitating detailed studies of protein interactions and cellular functions [62]. While lanthanide-binding tags were initially used in simple organisms such as bacteria, their application can be extended to mammalian systems. Recently, lanthanide-binding tags have been developed for the immunolabeling of numerous proteins with different lanthanides, allowing the simultaneous imaging of multiple proteins using the multielement capability of SXRF [29]. This technology cannot be directly adapted for correlative imaging of lanthanide-labeled proteins together with endogenous metals, as this would require native sample preparation. However, in the future, lanthanide-binding tags (LBTs) can be seamlessly integrated into target protein sequences using advanced genome editing tools such as CRISPR/Cas9, which are becoming increasingly reliable in mammalian cells. Recently, cobalt oxide nanoparticles have also been synthesized with the final aim of serving as protein-labeling vectors for SXRF imaging [63]. As recently reviewed, the development of metal-based probes for X-ray fluorescence microscopy will enable the provision of invaluable information about metal distribution and their roles in various biological processes [15, 64, 65].

The mentioned emerging X-ray-sensitive probes for synchrotron radiation are revolutionizing in situ bioimaging at the nanoscale. These advancements are poised to significantly impact various fields, including cell biology, pathology, and pharmacology, for drug tracking and efficacy for therapeutics, by providing detailed and dynamic views of biological systems in their native environments. They allow multiple challenges such as monitoring biological processes (e.g. enzyme activity, ion transport, and metabolic changes in real time) or investigating disease mechanisms, (e.g. cancer progression and neurodegenerative disorders, at the molecular level). New avenues for metal-tool development are directed toward enhancing their biocompatibility, specificity, and X-ray-sensitivity to minimize toxicity and maximize imaging contrast. Moreover, advanced algorithms and computational methods are required to handle the large volumes of data generated by synchrotron imaging techniques, so a higher effort would be recommended on that side.

In parallel, combining synchrotron X-ray imaging with other imaging modalities, such as magnetic resonance imaging (MRI) and fluorescence microscopy, could definitely provide comprehensive insights into biological systems and continued development in this area holds promise for new discoveries and improved diagnostic and therapeutic strategies. For example, Antharam et al. used high-field magnetic resonance microscopy (14 T) and SXRF mapping to study hippocampal tissue from AD patients and healthy controls [66]. They reported a positive correlation between tissue iron and MRI relaxation rates, illustrating the ability of these techniques to reveal molecular changes preceding significant atrophy. Similarly, Finnegan et al. demonstrated a linear relationship between local iron concentrations determined by SXRF imaging and MRI parameters, such as R2 and R2*, in postmortem basal ganglia tissue at 9.4 T [66]. These studies highlight the potential of SXRF to enhance our understanding of the role of iron in neurodegenerative diseases by enabling precise quantification of iron distributions. Taken together, these findings highlight the potential of combining synchrotron-based imaging modalities with MRI to provide complementary information on iron homeostasis and its perturbations in neurodegenerative diseases.

Conclusion

In summary, understanding the cellular and molecular mechanisms involved in the neurobiological or neurotoxic functions of metals requires the development of new analytical techniques and new imaging protocols using multiple techniques for the same object of study. This field of research has benefited greatly from recent advances in synchrotron radiation imaging, which has reached levels of sensitivity and spatial resolution that allow metals to be localized very precisely in subcellular structures, including synaptic compartments. These metal localization data need to be correlated with other biological information, such as interactions with the biomolecules present, particularly proteins. Recent technological advances have also led to significant improvements in protein imaging, thanks to new super-resolution microscopy techniques and improved fluorescent molecules for labeling. These methodological advances will continue to bring new data on metals in neurosciences.

Special attention must be paid to sample preparation methods, as those typically used for protein observation lead to significant changes in the distribution of chemical elements, such as metal loss, and redistribution of exogenous metal contamination. Milder preparation methods based on cryogenics should be used, which limits the potential applications due to the scarcity and high cost of cryogenic equipment. Some applications do not require cryogenic processing, typically when the interaction of the metals with the biological structures is very stable, such as for covalent binding or solid phases (precipitates, calcifications, nanoparticles). Sample preparation is an integral part of the experimental design and should be chosen very carefully according to the objectives of the study.

The use of direct correlative microscopy for metals and proteins offers an interesting perspective compared to multimodal imaging. However, direct correlative microscopy requires the development of biomolecular tools, such as proteins labeled with exogenous elements such as lanthanides, for direct observation of metal-tagged proteins and endogenous metals by synchrotron imaging spectroscopy. Another interesting alternative is the use of optical fluorescence microscopy for both protein and metal imaging with fluorescent metal sensors at high spatial resolution using super-resolution microscopy [67], but in this case, only the labile fraction of metals is observable. Label-free imaging in the postmortem brain using scanning transmission X-ray microscopy is also a very interesting approach, combining the mapping of trace metals such as Fe, with the imaging of organic structures, in particular the protein density, allowing the identification of subcellular entities such as the nucleoli or neuromelanin distribution [68].

Our knowledge of metal functions in neuroscience and neurotoxicology is advancing, but much remains to be done to develop new tools given the multiple indications of metal imbalance observed in many, if not all, neurological disorders.

Conflict of interest

The authors declare no conflict of interest.

Funding

None declared.

Data availability

No new data were generated or analyzed in support of this research.

References

Author notes