Baseline gene signatures of reactogenicity to Ebola vaccination: a machine learning approach across multiple cohorts.

Introduction: The rVSVDG-ZEBOV-GP (Ervebo®) vaccine is both immunogenic and protective against Ebola. However, the vaccine can cause a broad range of transient adverse reactions, from headache to arthritis. Identifying baseline reactogenicity signatures can advance personalized vaccinology and increase our understanding of the molecular factors associated with such adverse events. Methods: In this study, we developed a machine learning approach to integrate prevaccination gene expression data with adverse events that occurred within 14 days post-vaccination. Results and Discussion: We analyzed the expression of 144 genes across 343 blood samples collected from participants of 4 phase I clinical trial cohorts: Switzerland, USA, Gabon, and Kenya. Our machine learning approach revealed 22 key genes associated with adverse events such as local reactions, fatigue, headache, myalgia, fever, chills, arthralgia, nausea, and arthritis, providing insights into potential biological mechanisms linked to vaccine reactogenicity.

Transcriptomics of natural and synthetic vitamin D in human hepatocyte lipotoxicity.

Vitamin D (VD) has been used to prevent nonalcoholic fatty liver disease (NAFLD), a condition of lipotoxicity associated with a defective metabolism and function of this vitamin. Different forms of VD are available and can be used for this scope, but their effects on liver cell lipotoxicity remain unexplored. In this study we compared a natural formulation rich in VD2 (Shiitake Mushroom extract or SM-VD2) with a synthetic formulation containing pure VD3 (SV-VD3) and the bioactive metabolite 1,25(OH)2-D3. These were investigated in chemoprevention mode in human HepaRG liver cells supplemented with oleic and palmitic acid to induce lipotoxicity. All the different forms of VD showed similar efficacy in reducing the levels of lipotoxicity and the changes that lipotoxicity induced on the cellular transcriptome. However, the three forms of VD generated different gene fingerprints suggesting diverse, even if functionally convergent, cytoprotective mechanisms. Main differences were (1) the number of differentially expressed genes (SV-VD3 > 1,25[OH]2-D3 > SM-VD2), (2) their identity that demonstrated significant gene homology between SM-VD2 and 1,25(OH)2-D3, and (3) the number and type of biological functions identified by ingenuity pathway analysis as relevant to liver metabolism and cytoprotection annotations. Immunoblot confirmed a different response of VDR and other VDR-related proteins to natural and synthetic VD formulations, including FXR, PXR, PPARγ/PGC-1α, and CYP3A4 and CYP24A1. In conclusion, different responses of the cellular transcriptome drive the cytoprotective effect of natural and synthetic formulations of VD in the free fatty acid-induced lipotoxicity of human hepatocytes.

Immune Memory After Respiratory Infection With Streptococcus pneumoniae Is Revealed by in vitro Stimulation of Murine Splenocytes With Inactivated Pneumococcal Whole Cells: Evidence of Early Recall Responses by Transcriptomic Analysis.

The in vitro stimulation of immune system cells with live or killed bacteria is essential for understanding the host response to pathogens. In the present study, we propose a model combining transcriptomic and cytokine assays on murine splenocytes to describe the immune recall in the days following pneumococcal lung infection. Mice were sacrificed at days 1, 2, 4, and 7 after Streptococcus pneumoniae (TIGR4 serotype 4) intranasal infection and splenocytes were cultured in the presence or absence of the same inactivated bacterial strain to access the transcriptomic and cytokine profiles. The stimulation of splenocytes from infected mice led to a higher number of differentially expressed genes than the infection or stimulation alone, resulting in the enrichment of 40 unique blood transcription modules, including many pathways related to adaptive immunity and cytokines. Together with transcriptomic data, cytokines levels suggested the presence of a recall immune response promoting both innate and adaptive immunity, stronger from the fourth day after infection. Dimensionality reduction and feature selection identified key variables of this recall response and the genes associated with the increase in cytokine concentrations. This model could study the immune responses involved in pneumococcal infection and possibly monitor vaccine immune response and experimental therapies efficacy in future studies.