通过微流控技术解析急性髓系白血病中功能失调的免疫突触

Minimal residual disease (MRD) in acute myeloid leukemia (AML) arises from residual leukemic cells that resist chemotherapy and evade immune surveillance, notably through the PD-1/PD-L1 immune checkpoint. These persistent cells are the main source of relapse. Our working hypothesis is that alterations in calcium signaling, a key regulator of T-cell activation, contribute to PD-1/PD-L1-mediated immune escape in AML. The project aims to characterize the calcium signaling signature associated with PD-1/PD-L1 axis activation during the formation of the immunological synapse (IS) between T cells and leukemic cells from AML patients, to better understand immune dysfunction and therapy failure. To facilitate interactions between rare T cells and leukemic cells, we utilize microfluidic devices that enable controlled cell-cell contact and real-time monitoring of calcium dynamics. An AI-based algorithm performs real-time analysis of IS formation, processing about 100 events per image every five seconds. Calcium fluctuation data are extracted and mathematically modeled to quantify responses and classify patient-derived T cells as “responders” or “non-responders” using machine learning. Immunofluorescence is used to identify key molecular components at the IS, while RT-qPCR quantifies the expression of calcium-regulating genes such as ORAI, STIM, and NFAT isoforms in both T cells and leukemic cells. Our first results reveal altered calcium mobilization in AML-derived CD8⁺ T cells compared with healthy donors, along with PD-1-dependent inhibition of store-operated calcium entry linked to reduced ORAI1 activity. Ongoing single-cell RNA sequencing (scRNA-seq) analyses aim to identify transcriptional profiles of T-cell subpopulations, both at rest and during IS formation. By integrating functional calcium signatures and transcriptomic data, we seek to uncover signaling pathways responsible for impaired IS formation in AML. Ultimately, this work will identify molecular targets within the calcium signaling network that could be modulated to restore the cytotoxic activity of exhausted CD8⁺ T cells against therapy-resistant leukemic cells.