Multimodal imaging combining Raman and mass spectrometry applied to tuberculosis – MultiRaMaS

To meet the strategic objectives, MultiRaMaS is organized into 5 tasks:
- Task 1 «Sample preparation«: This one consists of using mouse and rabbit lungs infected with tuberculosis and treated with a panel of antibiotics. Once the lungs are collected, they will be sectioned. The sections will be placed on a support suitable for both methods. This step will include a validation phase to verify the integrity of the sections, as well as define the anatomical regions to be imaged. Histological staining protocols will also be included in this task.
- Note that in order to make acquisitions within the same anatomical regions and facilitate the alignment of the two modalities, a fiduciary marking will be used.
- Task 2 «Raman Imaging«: This will focus primarily on the optimization of Raman parameters in order to determine the optimal conditions for image acquisitions.
- Task 3 «MALDI mass spectrometry imaging«: In order to carry out a multi-omic study, several approaches will be considered. The idea is to be able to detect and image, towards an adequate sample preparation, all biomolecules related to tuberculosis including antibiotics, lipids, and peptides / proteins. Likewise, the acquisition parameters will also be optimized.
- Task 4 «Quantification by MALDI imaging«: This one is directly related to the developments and results obtained in the previous task. The quantification of antibiotics and lipids will receive special attention.
- Task 5 «Data processing for multimodal imaging«: This one concerns the implementation of a computational strategy dedicated to multimodal imaging. For this purpose, a new software will be designed including basic functionalities specific to each modality, as well as more advanced functionalities allowing the correlation of the two modalities as well as their fusion.

The results obtained during this first period are in line with what was expected in the MultiRaMaS project. The optimizations made as part of the first task made it possible to finalize the outlines. The methodology is now mastered and robust. The cross-shaped fiduciary landmarks were optimized using the laser beam and are now used routinely, making it easier to overlay multimodal images. Concerning the second task, few results were generated in Raman spectroscopy. However, some developments have been made and the preliminary results have made it possible to make interesting advances in the optimization of image acquisition parameters. The third task focuses on developments in MALDI mass spectrometry imaging for a multi-omic study. For this, different matrices and deposition conditions have been produced. The best conditions were applied to first study the spatial distribution at 5 µm of chlofazimine (CFZ) in the granuloma area. CFZ has been shown to be localized in macrophages and foamy macrophages. Some lipids related to tuberculosis, in particular PIM, have also been mapped. Based on previous results, MALDI imaging quantification (qMSI) was applied on the CFZ. This development is based on the use of a tissue homogenate and the development of a calibration curve that will be used to determine the concentration of CFZ. The qMSI will then be compared with the results of laser microdissection followed by quantification by liquid chromatography coupled with a mass spectrometer. Finally, the processing of imaging data is currently carried out with pre-existing tools. This fifth task will be initiated very soon and should lead to the design of new software dedicated to the processing of multimodal imaging.

In the coming deadlines, our ambition is to continue our efforts to carry out this MultiRaMaS project. We will first focus on the finalization of the qMSI method, as well as its publication in a peer-reviewed journal. As described previously, the idea is to be able to perform a multi-omic investigation based on MALDI imaging, so we will continue our efforts on the detection and mapping of anti-tuberculosis drugs, as well as lipids. We will then focus on the development of MALDI imaging for the detection of peptides / proteins. Once the methodological developments related to MALDI imaging are finalized, multimodal imaging will be highlighted. At the same time, the writing of the new software will have started and should allow us to automate the data processing of multimodal images, required for the interpretation of the results. We also aim to extend our panel of modalities to be implemented in order to improve significantly our knowledge of this infectious pathology.

M. Tuck, L. Blanc, R. Touti, N.H. Patterson, S. Van Nuffel, S. Villette, J.-C. Taveau, A. Römpp, A. Brunelle, S. Lecomte, N. Desbenoit. Multimodal Imaging based on Vibrational Spectroscopies and Mass Spectrometry Imaging applied to Biological Tissue: A Multiscale and Multi-omics Review. Analytical Chemistry, 2021, 93, 445-477. DOI: 10.1021/acs.analchem.0c04595. (https://hal.archives-ouvertes.fr/hal-03107524v1). Invitation de N. Desbenoit à rédiger une revue.

L. Blanc, A. Brunelle, C. Courrèges, N. Desbenoit, I. Fournier, J. Franck, L. Labeyrie, S. Mounicou, J.-Y. Salpin, D. Schaumlöffel, M.A. Subirana, D. Touboul, M.D. Tuck, S. Van Nuffel, et M. Wisztorski (par ordre alphabétique). Imagerie par Spectrométrie de Masse : Principes et Applications. ISTE, Encyclopédie Science, 2021, Chimie Analytique, Chapitre XXX, Soumis et accepté. Invitation de N. Desbenoit à rédiger un chapitre.

In public health contexts, where results of conventional imaging methodologies are often disappointing, molecular multimodal imaging is within a realm of possibility. Various imaging techniques are informative for functional groups, molecular weights, or special recognition sites, but no individual technique provides optimal answers to all questions. Thus, combining information from two or more measurement platforms is highly attractive. Such an approach is required to elucidate the complex spatial distribution of biomolecules in tissues, opening the way for a qualitative and quantitative multi-omics overview, both targeted and unbiased, of lipids, proteins, peptides, antibiotics, nucleic acids as well as glycans. Herein, we propose to combine vibrational spectroscopy, namely Raman Scattering (RS) and Mass Spectrometry Imaging (MSI) in a single workflow. Strengths and weaknesses of these technologies make them highly complementary. Key objectives of MultiRaMaS are both methodological and biological. On the one hand, as an emerging and challenging strategy multimodal imaging combining RS and Matrix-Assisted Laser Desorption/Ionization (MALDI) MSI needs further improvements: 1) optimization of a stringent and common sample preparation including cryo-sectioning and accurate identification of anatomical structures, 2) optimization of settings for fast multi-omics screening by RS imaging, 3) optimisation of settings to access the high chemical specificities and structure details by MALDI-MSI, 4) quantification of biomolecules of interest by MALDI-MSI, and 5) data processing including visualization, quantification, co-registration as well as fusion of hyperspectral imaging data. On the other hand, this original and remarkable approach, not fully explored, will be applied throughout a single tissue section and methodological outputs will be applied for the first time to the characterization of tuberculosis (TB), an infectious disease which affects each year 10 million humans and causes 1.7 million deaths. Mycobacterium tuberculosis (Mtb), the etiological agent of TB, establishes a durable lung infection in complex lesions where it is found intracellularly in various immune cell types and extracellularly in the central necrotic core of these lesions. Our current efforts are focused on increasing spatial resolution to visualize and identify the structures of molecules of interest at the cellular and subcellular level, i.e. lifting a scientific barrier. Of particular interest is the mapping and quantification of anti-TB drugs and biomarkers in the phagolysosome of macrophages where M. tuberculosis bacilli reside. Thus, the major biological objectives are: 1) to decipher the architecture and the microenvironment of tuberculous lesions, 2) to map, identify and quantify anti-TB drugs and biomarkers in the phagolysosome of macrophages, and finally 3) to evaluate the effect on Mtb of anti-TB drug concentration and duration of treatment. As an ultimate objective, all imaging data will be used to apply artificial intelligence and machine learning to facilitate the automation of the superimposition of imaging data onto digitized and fully annotated histological images. To meet the strategic objectives, MultiRaMaS is organized into 5 tasks. MALDI-MSI experiments shall rely on a highly performant combination of a source with high spatial resolution and high resolution mass spectrometer combining high accuracy and efficient MS/MS capability. This project gathers the complementary expertise of the project coordinator and his renowned collaborators to ensure the development of a multimodal imaging workflow to investigate tuberculosis. Complementary resources and expertise are assembled to ensure the feasibility of the proposed scientific program. This multidisciplinary and interdisciplinary project is original because of the technologies implemented which have rarely been combined, as well as by its applications to investigate tuberculosis.