Combining imaging mass spectrometry and immunohistochemistry to analyse the lipidome of spinal cord inflammation

Inflammation is a complex process that accompanies many pathologies. Actually, dysregulation of the inflammatory process is behind many autoimmune diseases. Thus, treatment of such pathologies may benefit from in-depth knowledge of the metabolic changes associated with inflammation. Here, we developed a strategy to characterize the lipid fingerprint of inflammation in a mouse model of spinal cord injury. Using lipid imaging mass spectrometry (LIMS), we scanned spinal cord sections from nine animals injected with lysophosphatidylcholine, a chemical model of demyelination. The lesions were demonstrated to be highly heterogeneous, and therefore, comparison with immunofluorescence experiments carried out in the same section scanned by LIMS was required to accurately identify the morphology of the lesion. Following this protocol, three main areas were defined: the lesion core, the peri-lesion, which is the front of the lesion and is rich in infiltrating cells, and the uninvolved tissue. Segmentation of the LIMS experiments allowed us to isolate the lipid fingerprint of each area in a precise way, as demonstrated by the analysis using classification models. A clear difference in lipid signature was observed between the lesion front and the epicentre, where the damage was maximized. This study is a first step to unravel the changes in the lipidome associated with inflammation in the context of diverse pathologies, such as multiple sclerosis. Graphical abstract Supplementary Information The online version contains supplementary material available at 10.1007/s00216-024-05190-3.

).In those cases in which the initial segmentation was not able to describe adequately the lesion, the segment(s) containing it was segregated and re- analysed (C, H, M, R, and W).IHC in S has the colours reversed: Iba1 in green and MBP in red.Segment colours were assigned from the colour bar in the image, according to the correlation between them.All LIMS experiments carried out in negative-ion mode at 10 m/pixel.Scale bar: 200 m Table S1.Performance of the four models tested to classify the lipid signatures extracted from the sections.All models exhibit outstanding performance but in the case of those based on Random Forest and Logistic regression algorithms the classification is perfect, with absence of false positive and false negative cases.

Figure S1 .
Figure S1.Comparison between the IHC images (A, D, F I, K, N, P, S and U) and the segmentation images of the corresponding LIMS experiment, obtained setting the number of segments to between 5 and 10 (B, E, G, J, L, O, Q, T and V).In those cases in which the initial

Figure S2 .
Figure S2.Relative abundance of the lipid classes studied in this work in the three areas identified in the samples.The asterisks correspond to the significance of the changes observed: *: p < 0.05; **: p < 0.01; ***: p<0.0001 in a T-test.

Figure S3 .
Figure S3.Confusion matrices for the four classification models tested.A) support vector machine (SVM); B) Random Forest; C) Naïve Bayes; D) Logistic Regression.Perfect classification of the samples was achieved with Random Forest and Logistic Regression.

Figure
Figure S4.A) Fragmentation spectrum of m/z = 716.524 in negative polarity using a laser energy of 40 µJ.The fragments point to B) PE (16:0_18:1) or C) PE (18:1_16:0).Position of the double bonds was not determined.

Figure
Figure S5.A) Fragmentation spectrum of m/z = 722.513 in negative polarity using a laser energy of 40 µJ.The fragments point to B) PE P-16:0/20:4.Position of the double bonds was not determined.

Figure
Figure S7.A) Fragmentation spectrum of m/z = 730.538 in negative polarity using a laser energy of 40 µJ.The fragments point to PE 15:0_20:1 and C) PE 15:1_20:0 (blue labels) and D) PE 17:0/18:1 (red labels).Position of the double bonds was not determined.
Figure S8.A) Fragmentation spectrum of m/z = 738.508 in negative polarity using a laser energy of 40 µJ.The fragments point to B) PE 16:0/20:4.Position of the double bonds was not determined.

Figure
Figure S9.A) Fragmentation spectrum of m/z = 742.542 in negative polarity using a laser energy of 40 µJ.The fragments point to B) PE 18:1/18:1.Position of the double bonds was not determined.
Figure S12.A) Fragmentation spectrum of m/z = 762.507 in negative polarity using a laser energy of 40 µJ.The fragments point to B) PE 16:0/22:6 and PS 34:0 (blue).Position of the double bonds was not determined, neither the chains for the PS.

Figure
Figure S13.A) Fragmentation spectrum of m/z = 764.526 in negative polarity using a laser energy of 40 µJ.The fragments point to B) PE 20:4_18:1.Position of the double bonds was not determined.

Figure
Figure S15.A) Fragmentation spectrum of m/z = 770.572 in negative polarity using laser energy of 40 µJ.The fragments point to B) PE 18:1_20:1 and C) PE 20:2_18:0 (blue).Position of the double bonds was not determined.

Figure
Figure S17.A) Fragmentation spectrum of m/z 774.547 in negative polarity using laser energy of 40 µJ. A. The fragments point to B) PE P-18:0/22:6 and PS 35:1 (blue).Position of the double bonds was not determined.

Figure
Figure S18.A) Fragmentation spectrum of m/z = 788.520 in negative polarity using laser energy of 40 µJ.The fragments point to B) PE 18:1/22:6 and C) PS 18:0_18:1 (blue).Position of the double bonds was not determined.

Figure
Figure S21.A) Fragmentation spectrum of m/z = 810.530 in negative polarity using laser energy of 40 µJ.The fragments point to B) PS 18:0/20:4.Position of the double bonds was not determined.

Figure
Figure S22.A) Fragmentation spectrum of m/z = 823.546 in negative polarity using laser energy of 40 µJ.The fragments point to B) PI 18:0_15:0.Position of the double bonds was not determined.