Background & Aims
Pain is a complex, multidimensional, and critical percept that initiates appropriate protective behaviors by integrating incoming sensory information with ongoing brain states and imposes an immense societal burden when it persists. At the molecular level, proteins are key targets of analgesics, and their altered expression can initiate and maintain chronic pain. Although transcriptomes are incredibly informative, growing evidence indicates that gene and protein expression can be discordant. There are also many molecular changes—such as protein-protein interactions and post-translational modifications—that are missed by transcriptomics. To enable insights into molecular changes after injury and in chronic pain conditions, and to identify novel therapeutic targets, we utilize advances in proteomic instrumentation, techniques, and analysis to study the pain generating neuroaxis.
Methods
We assembled murine whole tissue proteomes from dorsal root ganglia (DRG), dorsal (DH) and ventral horns of the spinal cord, and multiple brain regions (frontal cortex and amygdala) in inflammatory (Complete Freund’s Adjuvant; CFA) or nerve (spared nerve injury; SNI) injury pain models. All injured mice used for downstream analysis displayed mechanical hypersensitivity. To improve the size and quality of the dataset, we optimized tissue processing protocols, analyzed samples with high-sensitivity mass spectrometry (Bruker TimsTOF Pro), and used a well-validated analysis pipeline that identified 4 to over 6 thousand proteins per sample. To reduce batch effects and bias, we pseudo-randomized sample processing order and processed tissues blinded.
Results
Clustering analysis revealed high intra-tissue proteome overlap and distinct peripheral and central nervous system proteomes. Across all tissues, inflammatory and nerve injury induced hundreds of significantly up- or down-regulated proteins. Comparing our proteomes to DH and DRG transcriptomes revealed a correlation between protein abundance and gene expression. However—and consistent with many prior studies in cell cultures, animal models, and humans—only a small number of genes/proteins were significantly modulated in both transcript and protein. To broaden the candidate of pain-modulated proteins, we next took advantage of the wealth of protein network data and recent algorithms. To identify additional hits, which we term pain predicted proteins (PPP), we used network propagation approaches[1] that “diffuse” information from our hits across data-driven protein interaction networks. Validating this approach, we identified several PPPs that have implications in migraine and pain.
Conclusions
These proteomes provide a rich molecular resource across multiple tissues, time points, and pain models. The breadth of our datasets and analysis can help identify the evolution of the proteome, from acute to chronic pain, as well as molecular pathways engaged and modulated across pain conditions. Importantly, the network approaches that we applied suggest that existing and new datasets can be further mined to identify new targets. This ability to collect and analyze high-quality proteomes will allow the field to gain new insights into the molecular mechanisms of pain and potentially identify novel targets for pain therapy.
References
[1] Cowen, Lenore, et al. “Network propagation: a universal amplifier of genetic associations.” Nature Reviews Genetics 18.9 (2017): 551-562.
Presenting Author
Biafra Ahanonu
Poster Authors
Topics
- Informatics, Coding, and Pain Registries