Molecular Histology & Metabolite Profiling driven by MSI: Fine Biological Understanding of Gastrointestinal Diseases
Inflammatory Bowel diseases are characterized by relapsing-remitting inflammatory responses that have a tendency to develop where the bacterial load is greatest. Among industrialized countries, the increased incidence of Inflammatory Bowel diseases is presumably due to changes of the gut microbiota (herein referred as dysbiosis). It is indeed worth noting that Crohn’s disease recurrence is linked to a lower prevalence of major member of Firmicutes, including the anti-inflammatory commensal Faecalibacte-rium prausnitzii. More importantly, the intestinal microbiota has been highlighted to regulate intestinal homeostasis through the secretion of a large set of metabolites . Herein, genetically predisposed animals have been used to reveal the impact of two specific single gene mutations on the modulation of the metabolome within the colonic mucosa. Due to the complexity of colon tissue at the histological level and the importance of correlating meta-bolite distribution with specific location within tissue, the use of Mass Spectrometry Imaging (MSI) appears relevant to provide a better understanding of intestinal luminal meta-bolome.
Our aim is to enhance the knowledge on intestinal metabolome thanks to in-situ analysis of metabolite profile and distribution within colon tissue models. The use of MSI enables a spatially resolved and unlabeled imaging of di+erent metabolites directly in their micro-envi-ronment and provides a molecular profile to specific histological substructures.
Thanks to a collaboration with the INSERM research team headed by Dr. M. Chamaillard, we provide the two following examples of MSI applications in gastroenterology which illustrate the benefits given by MSI:
1) High spatial resolution imaging of colon tissue for molecular histology: Identification of colon tissue substructures using MSI
2) Comparison of metabolite profile and localization in knockdown models of mouse colon: In situ profiling of disease biomarker by MSI
– Animal: Terminal colons from Wild-Type and mutant mice were removed, snap frozen, embedded in CMC and stored at 80°C.
– Sectioning: Colons embedded in CMC were sectioned following transversal plan (12 μm of thickness) using Microm HM560 cryostat (Thermo Scientific, Germany) at -20°C and mounted on ITO conductive glass slides (Delta Technology, USA). HE Staining was performed on adjacent tissue sections for better visualization of histological regions.
– Matrix: 2.5 DHB powder (150 mg) was used and vaporized on tissue sample using home built sublimation apparatus (150°C, 8 min, 2.10-3mbar).
– Mass spectrometry imaging: Autoflex Speed LRF MALDI-TOF (Bruker Daltonik, Germany) with SmartBeam II laser. Positive mode (100-1000 Da) at 20 μm spatial resolution.
– Software: All presented MS images are from MutltimagingTM software 1.1 (ImaBiotech, France).
High spatial resolution imaging of colon tissue for molecular histology
Results & discussions:
High spatial resolution imaging of colon enables the visualization of histological layers and regions within tissue. An example of transversal colon sections mass spectrometric image at high spatial resolution (20 μm) is presented in figure
1. It shows the benefit given by matrix sublimation method to
yield high-quality molecular images with no analyte delocalization. Some contrast ionic species are used to accurately differentiate small histological structures of the colon. These ions are probably related to lipid species but taking into account limited ability of TOF mass spectrometer in terms of accurate mass measurement (10 ppm) and the lack of MS/MS
data, the precise identification remains di)icult. Nevertheless, we are able to propose some potential lipids identification for
ion at m/z 734.57 (red filter) which can be related to Phosphatidylcholine or Phosphatidylethanolamine molecules; PE
(35:0) or PC (32:0). The two others ions at m/z 609.91 (green
filter) m/z 703.59 (blue filter) cannot be accurately identified.
Thus, histological substructures can be easily discriminated
on an overlay molecular image from coronal section.
The colonic mucosa is primarily constituted by several
immune cells population (including lymphocytes and
myeloid cells) and the epithelium. The conjonctive tissues are
well localized on molecular image thanks to blue m/z filter
whereas green m/z filter is associated to external layer of the
mucosa, villi & crypts of Lieberkühn.
Crypts are intestinal glands which contain a large
variety of secretory cells (such as Paneth cells, goblet cells and enteroendocrine cells). Paneth cells secrete substantial quantities of antimicrobial molecules which are key mediators of
host-microbe interactions, and their dysfunction may contribute
to the pathogenesis of chronic inflammatory bowel disease .
Interestingly, the red m/z filter on overlay image
differentiates a specific region of the colon which can be
identified as Peyer’s patches (PPs). PPs are constituted by
isolated or aggregated lymphoid follicles from Gut-Associated
Lymphoid Tissue (GALT) . The induction of immune
tolerance or defense against certain pathogens is a function of
PPs resulting from complex process involving immune cells
located in the lymphoid follicles and the follicle-associated
It plays an important role in some intestinal illnesses,
including Crohn’s Disease (CD) and Gra: versus Host Disease
(GVHD). For all these reasons, the ability to map molecular
changes in these aforementioned specific regions of the colon
(crypts of Lieberkühn or Peyer’s patches) might be useful to a
better understanding of the pathophysiology of gastrointestinal
Comparison of metabolite profiles and localization in genetically predisposed mice
– Animal: Same as previously described
– Sectioning: Same as previously described
– Matrix: 9AA (10mg/ml, Methanol) was chosen and deposited using TLC sprayer (Sigma Aldrich, Germany).
– Mass spectrometry imaging: Solarix 7.0T FTICR (Bruker Daltonik, Germany) with SmartBeam II laser. Positive fullscan mode (100-800 Da), 300 shots at 40 μm spatial resolution.
– Software: All presented MS images are from MultimagingTM software 1.1 (ImaBiotech, France).
Results & discussion
The application of Mass Spectrometry Imaging in the
study of endogenous metabolites from biological tissues is a
recent technique which o »ers the simultaneously monitoring
of several compounds (lipids, small metabolites or drugs) with
spatiotemporal information about molecular behavior
[4,5,6,7]. The combination of these factors allows us to
perform the identification and validation of new biomarkers
of intestinal microbiota closely related to inflammatory bowel
diseases. The figure 2 displays some molecular images
obtain from each colon tissue models in negative detection
mode (anionic species) at 40 μm of spatial resolution. Haematoxylin & Eosin staining of corresponding tissue sections are also presented. Twelve ribonucleotides related metabolites
were selected to provide an overview of metabolome changes
between conditions. Others metabolites classes such as
nucleosides, amino acids, phospholipids or other nucleotides
have been detected but not shown here. Table 1 summarizes
the ion species detected with some information about their
relative concentration between conditions and fold di »erence
values. Figure 3 shows the profile of each ribonucleotides on
colonic resection specimens from wild-Type (WT) and mutant
Ribonucleotides are involved in many biological
process such as intracellular signal transduction, energy
metabolism or others cellular functions. They may have
several forms, including mono-, di- or triphosphate moieties
and undergo some chemical modification such as a reduction
(Desoxy-) or a cyclization (Cyclic-). Moreover, the reduced form
of ribonucleotide is required for both synthesis and repair of
Likewise, cyclic ribonucleotides such as cAMP or cGMP
act as second messenger in organisms. Most ribonucleotides
are abundant in WT model compared to transgenic one except
for ADP, ATP and IMP which have a very low intensity on mass
spectra. Molecular histological specificities are observed on MS
images, especially in one of the mutant colon which included a
Peyer’s patche (PPs). Notably, we noticed an accumulation of
several metabolites in this area compared to mucosa or
submucosa layers. It is the case for cTMP, dTMP (high intensity),
CMP, cAMP and dAMP (low intensity) whereas no specific localization was highlighted for GMP, cGMP, UMP and AMP. This
information might be useful for the understanding of PPs
function and its role in the development of CD and/or GVHD.
Matsumoto et al.  have compared the intestinal luminal metabolome of germ free (GF) with Ex-GF mice i.e. the intestinal microbiota composition. They have observed that some metabolites were up regulated (GF>Ex-GF) or down regulated (GF
We provide a link between a di »erential metabolome and risk to develop inflammatory bowel diseases. Biological
processes involving specific metabolite were directly assessed on histological substructures such as mucosa and submucosa layers
or highly disease relevant tissue (as for example Peyer’s patches or crypts of Lieberkühn). High spatial resolution molecular imaging allows following molecules/metabolites at the cellular level (such as Paneth or goblet cells, which are related to the development of several gastrointestinal diseases).