1 June 2018 in Cell Reports
Age-associated decreases in primary CD8+ T cell responses occur, in part, due to direct effects on naive CD8+ T cells to reduce intrinsic functionality, but the precise nature of this defect remains undefined. Aging also causes accumulation of antigen-naive but semi-differentiated "virtual memory" (TVM) cells, but their contribution to age-related functional decline is unclear. Here, we show that TVM cells are poorly proliferative in aged mice and humans, despite being highly proliferative in young individuals, while conventional naive T cells (TN cells) retain proliferative capacity in both aged mice and humans. Adoptive transfer experiments in mice illustrated that naive CD8 T cells can acquire a proliferative defect imposed by the aged environment but age-related proliferative dysfunction could not be rescued by a young environment. Molecular analyses demonstrate that aged TVM cells exhibit a profile consistent with senescence, marking an observation of senescence in an antigenically naive T cell population.
28 October 2019 in Journal of Proteome Research
1 May 2018 in Cell Metabolism
11 November 2021 in Nature Communications
Multipotent mesenchymal stromal cells (MSCs) ameliorate a wide range of diseases in preclinical models, but the lack of clarity around their mechanisms of action has impeded their clinical utility. The therapeutic effects of MSCs are often attributed to bioactive molecules secreted by viable MSCs. However, we found that MSCs underwent apoptosis in the lung after intravenous administration, even in the absence of host cytotoxic or alloreactive cells. Deletion of the apoptotic effectors BAK and BAX prevented MSC death and attenuated their immunosuppressive effects in disease models used to define MSC potency. Mechanistically, apoptosis of MSCs and their efferocytosis induced changes in metabolic and inflammatory pathways in alveolar macrophages to effect immunosuppression and reduce disease severity. Our data reveal a mode of action whereby the host response to dying MSCs is key to their therapeutic effects; findings that have broad implications for the effective translation of cell-based therapies.
1 October 2020 in Cell Stem Cell
6 June 2025 in Nature Medicine
Diagnosing and treating skin diseases require advanced visual skills across domains and the ability to synthesize information from multiple imaging modalities. While current deep learning models excel at specific tasks such as skin cancer diagnosis from dermoscopic images, they struggle to meet the complex, multimodal requirements of clinical practice. Here we introduce PanDerm, a multimodal dermatology foundation model pretrained through self-supervised learning on over 2 million real-world skin disease images from 11 clinical institutions across 4 imaging modalities. We evaluated PanDerm on 28 diverse benchmarks, including skin cancer screening, risk stratification, differential diagnosis of common and rare skin conditions, lesion segmentation, longitudinal monitoring, and metastasis prediction and prognosis. PanDerm achieved state-of-the-art performance across all evaluated tasks, often outperforming existing models when using only 10% of labeled data. We conducted three reader studies to assess PanDerm’s potential clinical utility. PanDerm outperformed clinicians by 10.2% in early-stage melanoma detection through longitudinal analysis, improved clinicians’ skin cancer diagnostic accuracy by 11% on dermoscopy images and enhanced nondermatologist healthcare providers’ differential diagnosis by 16.5% across 128 skin conditions on clinical photographs. These results show PanDerm’s potential to improve patient care across diverse clinical scenarios and serve as a model for developing multimodal foundation models in other medical specialties, potentially accelerating the integration of artificial intelligence support in healthcare.