Mass spectrometry imaging (MSI) and spatially resolved transcriptomics (SRT) are invaluable -omics techniques that measure the molecular and gene expression levels in a sample and the regions of the tissue where they are present. Scientists normally apply only one of these techniques to an individual sample because they believe that one protocol would destroy the sample before running the second one. Sometimes, however, scientists have extremely rare and valuable biological samples, such as those from human donors. Having a technique that combines metabolite and transcriptome analyses would allow scientists to extract more information from these samples, potentially leading to valuable insights into disease mechanisms.
With this goal in mind, researchers at Uppsala University developed an approach that combines MSI and SRT in the same biological sample. In a study published in Nature Biotechnology, they reported that they can measure low molecular weight metabolites and mRNA transcripts in one sample and validated the results in a mouse model of Parkinson’s disease and in postmortem human brain samples (1).
It's like drinking a cup of cappuccino: We just look at the bubbles, but then we can drink the coffee. With the laser, you ionize only the top of the tissue, and then you can check the mRNA in the bottom or middle of the sample.
- Joakim Lundeberg, Uppsala University
While most researchers think that MSI and SRT are incompatible in the same sample, Joakim Lundeberg, a coauthor and researcher at Uppsala University, and his team decided to test that idea. “We challenged the conception that RNA is degraded” after MSI, Lundeberg said.
Lundeberg and his team first mounted their sample on commercially available Visium Gene Expression array glass slides. They then added the matrix needed for mass spectrometry, performed MSI, stained the sample, imaged the sample with bright field microscopy, and lastly, ran the SRT. The only modification they made to the MSI or SRT protocols was to add three washes of cold methanol at the end of the MSI to wash away the matrix. To his team’s surprise, the mRNA in the tissue was still present after MSI.
“It's like drinking a cup of cappuccino: We just look at the bubbles, but then we can drink the coffee. With the laser, you ionize only the top of the tissue, and then you can check the mRNA in the bottom or middle of the sample,” Lundeberg explained.
To validate their new approach, the team analyzed brain samples from a mouse model of Parkinson’s disease. They first performed MSI and detected dopamine in the intact striatum and substantia nigra, but not in the lesioned contralateral nuclei, as expected in this condition. The researchers then performed SRT on the same samples and found that the expression level of key dopaminergic pathway genes correlated with dopamine presence in the substantia nigra.
The scientists also probed frozen human Parkinson’s disease postmortem striatal brain samples and found similar metabolite and transcripts results as when using each technique separately. They found higher dopamine levels in the medial division of the ventral caudate nucleus compared with other areas of the same nucleus as seen in a previous study, and transcript levels correlated with publicly available human data (2).
“In this study, we look at dopamine in Parkinson’s, which is a slowly developing disease, where the gene expression differences and the mass spectrometry images are pretty subtle, making it hard to become significant,” said Lundeberg. By measuring both parameters in the same sample, “you increase the measurement power, and suddenly things pop up,” he said.
The combination of both techniques increases the value of each individual measurement by decreasing batch effects and technical differences obtained when going to different samples. This is especially important for drug discovery where small changes can determine the efficacy of a drug.
Lundeberg plans to develop a protocol to apply the combined technologies on formalin-fixed paraffin-embedded samples. This will allow researchers to analyze even more samples such as those found in human biobanks.
Erwan Bezard, a Parkinson’s disease researcher at Bordeaux Neurocampus said, “It's absolutely fantastic! The only thing I regret is that there aren’t more people doing that.” Although the combination of MSI and SRT won’t require more equipment than needed for the separate experiments, Bezard said that only about four or five groups of scientists around the world have these technologies available in their labs.
“It’s a fantastic tool for back validation,” Bezard said. “You can generate new hypotheses and new ideas based on the analysis of the human tissues, which would be of course, more thorough [than animal models], so it's simply great.”
The application of this multimodal spatial technology won’t be limited to neurobiology. “The biggest promise is in tumors,” said Lundeberg. “Tumor [molecular composition] changes quite dramatically over space. So, that's one of the reasons for thinking that this is pretty good to do everything on the same tissue section.”
References
- Vicari, M., Mirzazadeh, R., Nilsson, A. et al. Spatial multimodal analysis of transcriptomes and metabolomes in tissues. Nat Biotechnol (2023).
- Shariatgorji, M., Nilsson, A., Fridjonsdottir, E. et al. Comprehensive mapping of neurotransmitter networks by MALDI-MS imaging. Nat Methods 16(10):1021-1028 (2019).
