| Summary: | <p>With the current increase in life expectancy comes a higher prevalence of age-associated neurological conditions, thus increasing social and economic pressure to improve maintenance of neurological health. Limited sample accessibility and availability place a technological demand on methods compatible with profiling central nervous system tissue which maximise sample utility. Here we present a label-free multi-modal molecular profiling approach which yields spatially resolved transcriptomic, proteomic and biochemical profiles of brain and spinal cord tissue sections, thus addressing this technological demand. This approach was used to profile samples from individuals without neurological condition (n=19) and individuals with secondary progressive multiple sclerosis (SPMS) (n=18), which is characterised by significant architectural alterations to central nervous system identifiable by histological characterisation. Using machine learning and cluster correlation, we have identified two tissue profiles associated with controls which corresponded to white and grey matter regions based on macromolecular, compositional, transcriptomic and proteomic features. By combining the unsupervised clustering across modalities, we identified and profiled a total of 6 degrees of pathology severity. </p>
<p>Apart from overall tissue classification, this approach has also generated valuable mechanistic biological insight: we have been able to not only recapitulate previously reported SPMS pathological biochemical features, but have also linked these processes to specific cell types and tissue-specific processes. This includes evidence of oxidation and alterations to protein processing/folding in oligodendrocytes and neurons. The protein alterations we report include increased β-sheet content and decreased α-helix content in SPMS, a feature normally associated with neurodegenerative conditions with protein folding alterations, and which had never been, to our knowledge, reported in SPMS and adds to its neurodegenerative-like pathological features. We have also outlined processes in SPMS analogous to injury-associated central nervous system scarring, and pathology-associated deviations from typical responses. In particular, we observed niches with high expression of S100β, a protein which has variable cellular function highly dependent on its concentration within tissue: the highly localised profile of spatial gene expression we observe provides insight into a previously unappreciated mechanism by which gene expression can have a deleterious impact attributable to its spatial expression concentration. We believe these findings support the utility of this technological approach not only for disease characterisation and hypothesis generation, but also generating mechanistic and clinical insight for diseases with limited tissue availability, thus enabling the development of treatments to promote healthy ageing of the CNS. </p>
|
|---|