The interdisciplinary field of spatial biology continues to connect omics research with the goal of understanding the spatial distribution of biomolecules that influence biological processes and functions. Advanced imaging techniques continue to emerge on the market, but matrix-assisted laser desorption ionization (MALDI) Imaging, a mass spectrometry-based technique, is a widely accepted methodology for determining spatial localization of analytes on tissue and has been around for more than 20 years.
Visualizing metabolites and lipids allows for the connection of the immediate cellular metabolic state with the enzymes and proteins that are doing the work of the cell. MALDI Imaging offers the only unlabeled spatial analysis technique for metabolites and lipids. Additional workflows make released glycans and intact proteins accessible for multiomic connections.
With this technology applied to fresh frozen or FFPE samples, correlating metabolite, lipid, glycan, and protein information to histology becomes increasingly easy at high spatial resolution for faster and more effective analysis of tissue morphological features. This tutorial will highlight the targeted protein workflow portion of MALDI Imaging capabilities, which has recently garnered attention as a breakthrough method for integration across the spatial omics space.
Methods
FFPE human kidney tissues were prepared using the standard MALDI HiPLEX-IHC workflow,1,2 which is described in Figure 1. (Human kidney tissue was kindly provided by the Medical University of South Carolina.) Briefly, the slides were heated at 60°C and transferred through xylene to a Tris-buffered saline rehydration gradient to remove the wax. The tissue then underwent antigen retrieval in a basic buffer, followed by a tissue blocking step. Next, antibodies of choice (with photocleavable peptide tags) were placed on the tissue and allowed to incubate at 4°C overnight.
The peptide tags were then released using ultraviolet light. MALDI matrix (α-cyano-4-hydroxycinnamic acid) was applied using established protocols on a pneumatic M3+ sprayer (HTX Technologies, Chapel Hill, NC). Finally, recrystallization of the matrix was performed, and the tissue was run on a Bruker timsTOF fleX MALDI-2 instrument at 5 µm lateral spatial resolution using microGRID technology.
After MALDI imaging was performed, matrix was washed off and hematoxylin and eosin (H&E) staining was done in accordance with standard procedures. Data was analyzed in SCiLS™ Lab software with corresponding H&E staining integrated with pathological annotations corresponding to protein expression for defining key histological features.
Results
Two series of experiments were run for proof of concept of instrument and workflow capabilities. Initial experiments were run with three antibodies on human FFPE tissue at both 20 µm and 5 µm spatial resolution to demonstrate the resolution enhancement and identification of molecular markers for key histological features. Additional experiments were done on serial tissue sections with higher complexity of antibodies to give a more comprehensive picture of protein evaluation. H&E staining was done post analysis and incorporated pathologists’ annotations showing correlation between protein expression and histological features.
The data shown in Figure 2 represent the initial experiments that were done with three different antibodies—specifically, antibodies against vimentin, histone H2A, and ATPase-1A1—to preliminarily evaluate the MALDI HiPLEX-IHC workflow on FFPE human kidney tissues. An image at 20 µm spatial resolution was captured in one area of the tissue, and a subsequent image at 5 µm spatial resolution was obtained in a different area of the tissue. The three peptides associated with the antibodies were at m/z 1222.79 (for ATPase-1A1), m/z 1230.84 (for vimentin), and m/z 1226.82 (for histone H2A).
For maximum clarity in visualization, mass channels for adducts (protonated peptide and sodium adduct) were combined. Closer examination of the 5 µm spatial resolution data is shown in Figure 3. Overlay of the three corresponding masses demonstrated significant localization of the peptides to areas predicted to be rich in the protein of interest (vimentin marker: glomeruli; histone H2A marker: nuclei; and ATPase-1A1 marker: proximal convoluted tubules).
A secondary, higher multiplexed experiment was conducted, including additional markers for CD68 (m/z 1216.75) and collagen 1A1 (m/z 1234.87). The experiment was repeated on serial sections of FFPE kidney tissue and run under the same conditions with two adjacent sections at 20 µm and 5 µm spatial resolution. Both imaging runs successfully localized and identified all antibodies used while containing minimal to no artifacts as a result of the 5 µm spatial resolution. These corresponding protein images were then directly compared to pathologists’ annotations of the histological features of the human kidney tissue (Figure 4).
Conclusion
This work describes the MALDI-HiPLEX-IHC workflow, which can incorporate microGRID and timsTOF flex technology to deliver 5 µm spatial resolution. Other capabilities include the correlation of complex information about intact proteins and the identification of morphological tissue features. In addition, pathologists’ annotations of histologically stained tissues can contribute to analyses performed by the Bruker SCiLS™ Lab software solution.
Kate Stumpo, PhD, is senior market manager, Imaging Business Unit, Bruker Scientific.
References
1. Yagnik G, Liu Z, Rothschild KJ, Lim MJ. Highly Multiplexed Immunohistochemical MALDI-MS Imaging of Biomarkers in Tissue. J. Am. Soc. Mass Spectrom. 2021; 32(4): 977-988. DOI: 10.1021/jasms.0c00473.
2. Lim MJ, Yagnik G, Henkel C, et al. MALDI HiPLEX-IHC: multiomic and multimodal imaging of targeted intact proteins in tissue. Front. Chem. 2023; 11: 1182404. DOI: 10.3389/fchem.2023.1182404.