Multi-omic Analysis of the Intracortical Microelectrode-Tissue Interface Inflammatory Response

Introduction:: Intracortical microelectrodes (IMEs) are devices implanted in the motor cortex of the brain that can be used to record the electrical activity of local neurons. These devices could be used to power an external device such as a prosthetic or wheelchair to aid those who suffer from paralysis, spinal cord injury, or general loss of motor function (1). However, these devices tend to fail over chronic time points, limiting their clinical use and efficacy (2). Many coatings and drug therapies have been tested to mitigate this inflammatory response (2), yet there is currently no consensus on a therapeutic treatment for IME implantation. A better understanding of the molecular components of this inflammatory response could provide possible targets for therapeutic improvement of IMEs.

This study quantifies the expression of over 21,000 RNA transcripts and 80 proteins within 270 µm of the implant site with 1 µm resolution. RNA was quantified in separate regions of interest based on distance from the implant, and protein expression was further divided into segments based on cell type. Cell-specific capture of expressed proteins allowed for the quantification of protein expression within our selected cell types (neurons and astrocytes). Our study represents the first proteomic or transcriptomic study of IME implant sites to implement cell specificity and is overall the most comprehensive analysis of the molecular components of the IME neuroinflammatory response that has been performed.

Materials and Methods:: Wild-type mice were implanted with silicon, planar Michigan-style electrodes, and the brains were extracted after 4, 8, and 16 weeks. Implanted groups were compared to nonsurgical control mice. The brains of these animals were then sliced perpendicular to the implanted probe, and the slides were used to measure protein and RNA expression using the Nanostring Digital Spatial Profiler system. Slides were stained with fluorescent morphology markers to identify the location of different cell types. Tissue was also stained with fluorescent tags (or “barcodes”) that will bind to our proteins or RNA of interest and be counted by the system. The electrode site was separated into 3 areas of interest: inner, middle, and outer rings. Each ring was 90 µm in thickness, therefore the total collected area was within a 270 µm radius from the electrode site. For our protein measurements, fluorescent images from the morphology markers were used to create additional areas of interest for different cell types, specifically neurons and astrocytes. This created 9 regions of interest, the combinations of the 3 cell types and the 3 rings. The “barcodes” from the protein and RNA were collected and counted separately for each designated area of interest.

Results, Conclusions, and Discussions:: Our proteomic analysis, which is complete, validates that protein expression changes based on the time following implantation, distance from the implant, and cell specificity. When analyzing the protein expression of the entire implant site (no cell or distance segmentation), we find a general trend of downregulation of autophagy proteins (ATG5, ATG12, Beclin-1) and proteins required for healthy neuronal function (Neurofilament light, Myelin basic protein). Very few inflammatory proteins are found to be significantly differentially expressed (Figure 1). The neuronal cell segment protein expression is very similar to our results found in the unsegmented regions (Figures 2,1). We once again see the downregulation of autophagy proteins and proteins required for healthy neuronal function. Autophagy degrades harmful substances within the cell and therefore prevents cell death. The downregulation of these proteins while neuronal dieback is still occurring suggests that by the 8-week time point, the autophagy attempts to save the cell have failed, and cells are resorting to apoptosis or necrosis.

In astrocytes, we see the upregulation of proteins associated with inflammatory pathways (SPP1, CTSD, MHCII) (Figure 3). This upregulation of inflammatory proteins is only found when we segment by astrocytes, showing that our cell segmentation method allows us to investigate new proteins and pathways that would have been lost in the proteins expressed by other cell types. In previous analyses without cell specificity, genes and proteins were not able to be definitively matched to the cell type where they were expressed, and differentially expressed genes and proteins were lost (Figures 1-3).

Our full transcriptomic analysis, which is nearly complete, will be used to supplement the proteomics data and provide further insight into the exact pathways being activated within each group. The 88 proteins quantified in this proteomic analysis cannot cover all of the pathways activated by these damaged cells, but quantifying the entire known mouse transcriptome will allow us to do a full analysis of the gene expression profile. This will allow us to determine which genes and proteins provide the best options for therapeutic IME improvement. Our lab plans to target these molecules ourselves in future experiments.





Acknowledgements (Optional): :

References (Optional): : 1. Ajiboye, B. et al., The Lancet. (2017;389:1821-1830). 2. Jorfi, M. et al., J Neural Eng. (2014;12:011001)