Popescu, V. | Search bioRxiv

bioRxiv2022

Rapamycin rescues loss-of-function in blood-brain barrier-interacting regulatory T cells

In many autoimmune diseases, FOXP3+ regulatory T cells (Tregs) skew towards a pro-inflammatory and non-suppressive phenotype and are therefore unable to control the exaggerated autoimmune response. This may largely impact the success of autologous Treg therapy which is currently under investigation for treatment of autoimmune diseases, including multiple sclerosis (MS). Thus, there is a need to ensure in vivo stability of Tregs before successful Treg therapy can be applied. Using a murine genetic fate-mapping model, we demonstrate that inflammatory exFOXP3 T cells accumulate in the central nervous system (CNS) during experimental autoimmune encephalomyelitis (EAE). In a human in vitro BBB model, we discovered that interaction with inflamed blood-brain barrier (BBB)-endothelial cells induces loss of suppressive function in Tregs. Transcriptome analysis further revealed that Tregs which migrated across inflamed BBB-endothelial cells in vitro have a pro-inflammatory Th1/17 signature and upregulate the mTORC1 signaling pathway compared to non-migrated Tregs. These findings suggest that interaction with BBB-endothelial cells is sufficient to affect Treg function, and that transmigration triggers an additive pro-inflammatory phenotype switch, which was also seen in CNS-derived exFOXP3 T cells of EAE mice. In vitro treatment of migrated human Tregs with the clinically-approved mTORC1 inhibitor rapamycin completely restored the loss of suppressive function. Finally, flow cytometric analysis indicated an enrichment of inflammatory, less suppressive CD49d+ Tregs in the cerebrospinal fluid of MS patients, thereby underscoring the relevance of our findings for human disease. In sum, our findings provide firm evidence that the inflamed BBB affects human Treg stability, which can be restored using a mTORC1 inhibitor. These insights can help in significantly improving the efficacy of autologous Treg therapy of MS.

immunology↗

bioRxiv2019

Machine Learning Analysis of Motor Evoked Potential Time Series to Predict Disability Progression in Multiple Sclerosis

BackgroundEvoked potentials (EPs) are a measure of the conductivity of the central nervous system. They are used to monitor disease progression of multiple sclerosis patients. Previous studies only extracted a few variables from the EPs, which are often further condensed into a single variable: the EP score. We perform a machine learning analysis of motor EP that uses the whole time series, instead of a few variables, to predict disability progression after two years. Obtaining realistic performance estimates of this task has been difficult because of small data set sizes. We recently extracted a dataset of EPs from the Rehabiliation & MS Center in Overpelt, Belgium. Our data set is large enough to obtain, for the first time, a performance estimate on an independent test set containing different patients.\n\nMethodsWe extracted a large number of time series features from the motor EPs with the highly comparative time series analysis software package. Mutual information with the target and the Boruta method are used to find features which contain information not included in the features studied in the literature. We use random forests (RF) and logistic regression (LR) classifiers to predict disability progression after two years. Statistical significance of the performance increase when adding extra features is checked with the DeLong hypothesis test.\n\nResultsIncluding extra time series features in motor EPs leads to a statistically significant improvement compared to using only the known features, although the effect is limited in magnitude ({triangleup}AUC = 0.02 for RF and {triangleup}AUC = 0.05 for LR). RF with extra time series features obtains the best performance (AUC = 0.75 {+/-} 0.07), which is good considering the limited number of biomarkers in the model. RF (a nonlinear classifier) outperforms LR (a linear classifier).\n\nConclusionsUsing machine learning methods on EPs shows promising predictive performance. Using additional EP time series features beyond those already in use leads to a modest increase in performance. Larger datasets, preferably multi-center, are needed for further research. Given a large enough dataset, these models may be used to support clinicians in their decision making process regarding future treatment.

immunology↗