Short-chain fatty acids (SCFAs), metabolites found in fermented foods and produced by colonic bacterial fermentation of dietary fibers, may mediate gut-brain axis crosstalk influencing the course of multiple sclerosis (MS). In this study, SCFAs are measured in plasma and paired stool samples by targeted mass spectrometry in a large cohort of deeply phenotyped individuals shortly after MS onset. Propionate is consistently associated with a protective effect on three clinical and imaging activity-related outcomes. Findings are reproduced in an equally large external cohort for plasma and stool measurements. Despite the correlation between plasma SCFAs and SCFA-producing gut microbial taxa, propionate levels are not correlated with dietary fiber, suggesting that intake may not be sufficient to modulate its levels and ensuing inflammation. In conclusion, propionate is consistently associated with lower disease activity, and exogenous administration may be more likely than fiber supplementation to confer protective effects on MS activity in proof-of-concept clinical trials.
T cell senescence is an important dysfunctional state associated with tumor progression, but whether senescent T (Tsen) cells directly reprogram tumor cell behavior remains unknown. Here, single-cell transcriptomic analyses identify abundant Tsen cells across breast, colorectal, and lung cancers. Tsen cells actively promote tumor growth, metastasis, stemness, and chemotherapy resistance by reprogramming tumor cells in vitro and in vivo. Mechanistically, lipid-enriched senescence-associated secretory phenotypes (SASPs) from Tsen cells drive metabolic remodeling and activate STAT1/STAT3 signaling in tumor cells, establishing a pro-tumorigenic state. Therapeutically, depletion of Tsen cells or disruption of tumor lipid metabolic reprogramming suppresses Tsen-mediated tumor progression and metastasis in vivo. Our findings establish Tsen cells as active orchestrators of tumor cell biological functions through lipid-rich SASP-dependent intercellular communication and reveal a previously unrecognized mechanism linking immune senescence to cancer progression. Targeting Tsen cells or their metabolic crosstalk with tumor cells can enhance anti-tumor immunity and improve therapeutic responses.
Artificial intelligence in medical imaging is shifting from task-specific algorithms toward adaptable foundation models (FMs) and vision-language models (VLMs). However, applying these models to MRI remains challenging because clinical interpretation requires integrating multi-parametric, volumetric imaging across multiple sequences with clinical context. This systematic review evaluates 35 studies on unimodal FMs and multimodal VLMs in MRI, summarizing emerging architectures, downstream applications, and challenges to clinical translation. Current research is concentrated in neuroimaging, with diagnostic classification representing the most mature application. Segmentation and outcome prediction also show promise, whereas automated report generation remains limited by poor clinical fidelity despite strong linguistic fluency. Broader adoption is hindered by heterogeneous MRI acquisition, limited large-scale paired image-text datasets, and inadequate safety-focused evaluation metrics. Advancing clinical translation will require standardized multicenter datasets, stronger integration with electronic health records and pathology, robust evaluation frameworks, and improved model explainability.
Reducing dietary carbohydrates can improve metabolic health, yet the precise influence of carbohydrate restriction on the gut microbiome remains uncertain. Here, we report an ancillary study to the Framingham State Food Study, or FS(2), a controlled, randomized feeding trial involving adults with overweight and obesity. After a calorie-restriction phase to induce weight loss, we randomized participants to diets varying in energy from carbohydrate and fat (low-carb: 20% carb, 60% fat; moderate-carb: 40% carb, 40% fat; high-carb: 60% carb, 20% fat) for weight-loss maintenance for 20 weeks. Weight loss results in significant microbiome shifts, mirroring known associations. We do not detect compositional changes in the microbiome when diets vary carbohydrate-to-fat ratio, while controlling for other potentially confounding dietary factors. Changes in Shuttleworthia abundance correlate positively with maintenance of total energy expenditure following weight loss, suggesting a link between the gut microbiome and energy metabolism.
We characterize the spatiotemporal dynamics of the tumor microenvironment (TME) in bladder cancer (BLCA) samples before and after neoadjuvant immunotherapy in a multicenter real-world cohort (n = 606). We identify two distinct differentiation outcomes in BLCA epithelial cells: sensitivity and resistance to immunotherapy. Additionally, we identify SERPINH1 as a common target of acquired immunotherapy resistance in BLCA. Furthermore, we characterize an immunotherapy-induced tumor-myeloid barrier, which leads to an excluded immune phenotype and immunotherapy resistance. Finally, we identify T cell terminal exhaustion and a senescent inflammatory cancer-associated fibroblast phenotype as key factors contributing to immunotherapy resistance. Collectively, we reveal the dynamic phenotypic evolution of the TME in BLCA during neoadjuvant immunotherapy and identify the characteristics of residual lesions after immunotherapy. These findings enrich our understanding of immunotherapy resistance in BLCA and contribute to developing combination therapies to improve patient prognosis.
