Advanced multi-modal mass spectrometry imaging reveals functional differences of placental villous compartments at microscale resolution

Abstract

The placenta is a complex and heterogeneous organ that links the mother and fetus, playing a crucial role in nourishing and protecting the fetus throughout pregnancy. Integrative spatial multi-omics approaches can provide a systems-level understanding of molecular changes underlying the mechanisms leading to the histological variations of the placenta during healthy pregnancy and pregnancy complications. Herein, we advance our metabolome-informed proteome imaging (MIPI) workflow to include lipidomic imaging, while also expanding the molecular coverage of metabolomic imaging by incorporating on-tissue chemical derivatization (OTCD). The improved MIPI workflow advances biomedical investigations by leveraging state-of-the-art molecular imaging technologies. Lipidome imaging identifies molecular differences between two morphologically distinct compartments of a placental villous functional unit, syncytiotrophoblast (STB) and villous core. Next, our advanced metabolome imaging maps villous functional units with enriched metabolomic activities related to steroid and lipid metabolism, outlining distinct molecular distributions across morphologically different villous compartments. Complementary proteome imaging on these villous functional units reveals a plethora of fatty acid- and steroid-related enzymes uniquely distributed in STB and villous core compartments. Integration across our advanced MIPI imaging modalities enables the reconstruction of active biological pathways of molecular synthesis and maternal-fetal signaling across morphologically distinct placental villous compartments with micrometer-scale resolution.

Fig. 1. Placenta histology.

Syncytiotrophoblast (STB) - outer villous compartment that along with discontinuous cytotrophoblast cover the entire surface of the villous tree. STB provides a barrier between maternal blood and fetal blood. Core – the inner villous compartment that consists of fetal blood vessels, mesenchymal stromal cells, macrophages and connective tissue. pFEC placental – fetal endothelial cell.

Fig. 2. Advanced MIPI approach.

Schematic workflow of the advanced MIPI approach that combines multi-modal MALDI-MSI for comprehensive lipidome and metabolome imaging with complementary microscale proteome profiling by microPOTS. 4-APEBA 4-(2-((4-bromophenethyl)dimethylammonium)ethoxy) benzenaminium dibromide, DHA 2,5-dihydroxyacetophenone, DHB 2,5-dihydroxybenzoic acid.

Fig. 3. Multi-modal MALDI MSI.

a MALDI-MSI categorized two main subregions, STB and core. b Lipidomic imaging by MALDI-UHMR-HF Orbitrap MS, performed on two adjacent sections. H&E-stained optical image and example ion images of section 1, from left to right in the following order: lipids localized in core and stem villi; lipid localized in STB; lipids present in both, STB and core, compartments; lipids detected in stem villi. Lipidomic imaging of placental section 2 is in Supplementary fig. 1. c Metabolomic imaging by MALDI-12T-FTICR MS, performed on two adjacent sections. H&E-stained optical image of placental section 3 and example ion images of acetoacetate and palmitoylcarnitine detected in core and STB, respectively. Metabolomic imaging of placental section 4 is in Supplementary fig. 1.

Fig. 4. Unveiling subregion-specific enzymes using microPOTS processing.

a LCM collection of placenta villous subregions. b Volcano plot for the abundance-based model comparing the tissue subregion means for each protein. A mixed effects linear model, as fully described in the Methods section, with a conditional Gaussian distribution was fit to the normalized log relative abundance data for each protein. A two-sided likelihood ratio test for lack-of-fit was conducted and a Benjamini-Hochberg multiple comparison adjustment was calculated on all p-values. Source data are provided as a Source Data file. c Volcano plot for the probability of detection-based model comparing the tissue subregion mean detection probabilities for each protein. A generalized mixed effects linear model, as fully described in the Methods section, with a conditional binomial distribution was fit to the binary detect/not detected outcome data for each protein. A two-sided likelihood ratio test for lack-of-fit was conducted and a Benjamini-Hochberg multiple comparison adjustment was calculated on all p-values. Source data are provided as a Source Data file. Red hexagons and blue hexagons represent proteins identified exclusively in the STB and Core, respectively. The color of each hexagon corresponds to the scale bars below the graph, which indicate the number of proteins (N). In the present volcano plot, extreme values were not shown. The plot with results for all proteins can be found in Supplementary Statistical Methods.

Fig. 5. Spatial multi-omics integration unravels unique biological pathways in morphologically distinct placenta villi compartments.

a Advanced metabolomic imaging captured chemically derivatized progesterone in STB. Metabolomic imaging was performed on two adjacent sections. b Reconstructed estrogen synthesis and signaling pathways. Microscale proteomics was performed on three adjacent sections on 14 STB and 13 core subregions. c Lipidomic imaging revealed lipid signatures in STB. Lipidomic imaging was performed on two adjacent sections. d Reconstructed de novo synthesis pathway of ceramide in STB. Microscale proteomics was performed on three adjacent sections on 14 STB and 13 core subregions. e Linoleyl carnitine imaged in STB and corresponding linoleic fatty acid in both STB and core, indicating fatty acid transport across villi functional units. Metabolomic imaging was performed on two adjacent sections. f Reconstructed model of fatty acid transport and energy production pathway. Microscale proteomics was performed on three adjacent sections on 14 STB and 13 core subregions. g Ketone bodies prevalently detected in the villous core. Metabolomic imaging was performed on two adjacent sections. h Proposed ketone body oxidation pathway in the villous core. Microscale proteomics was performed on three adjacent sections on 14 STB and 13 core subregions. i Overlayed ion images of acylcarnitine and ketone body, demonstrating compartmentalization of metabolic pathways in placental villi. Metabolomic imaging was performed on two adjacent sections. Enzymes in red color – detected in the STB; Enzymes in blue color – detected in the core; Proteins highlighted in grey were detected in both compartments. Molecules highlighted in orange, blue, and green were detected by MALDI-MSI with localization in STB, core, and both compartments, respectively.