Introduction: Diabetes, especially Type 1 Diabetes, poses a significant challenge to global public health, with islet cell transplantation offering a new hope for patients. However, the phenomenon of immune rejection limits its long-term success rate. Recent research has focused on the complexity of the post-transplant immune microenvironment.
Methods: We analyzed islet cell subsets in allogeneic and syngeneic islet transplant mouse models using single-cell sequencing technology. Differential expression analysis revealed variations in the expression of mitochondrial genes. Additionally, predictive models of islet transplant immune rejection were constructed using machine learning algorithms.
Results: We identified three islet cell subsets, finding that these subsets play different roles in islet transplantation. Specifically, changes in the expression of mitochondrial genes were closely related to immune rejection. Our machine learning model accurately predicted immune rejection in islet transplantation, achieving high AUC values (AUC>0.8) in the training and validation sets.
Conclusion: This study is the first to explore in-depth the changes in mitochondrial gene expression within islet cells after islet transplantation and their role in immune regulation, revealing the central role of islet cells in the immune microenvironment. Moreover, the developed machine learning predictive model provides new tools for reducing immune rejection and improving the success rate of islet transplantation, offering scientific bases for future diabetes treatment strategies.