Investigate heterogeneous neutrophils in NSCLC

为了靶向非小细胞肺癌GEM模型中的肿瘤浸润中性粒细胞亚群（SiglecF+中性粒细胞），我们决定创建中性粒细胞中缺乏SiglecF表达的小鼠模型。不幸的是，尽管多次尝试，本应只在中性粒细胞中失去SiglecF表达的小鼠模式有一个不可预见的缺陷：它不能在任何细胞类型中表达SiglecF，可能是由于未知的表观遗传效应。因此，我们决定寻找调节肿瘤内中性粒细胞发育的上游调节因子，并基于我们的单细胞RNA测序和单细胞ATAC （Assay for Transposase-Accessible Chromatin）测序数据确定RUNX1是一个潜在的转录因子。我们建立了只在成熟中性粒细胞（Ly6G-Cre; RUNX1-flox）中缺乏RUNX1的小鼠模型，发现肿瘤浸润中性粒细胞中SiglecF+中性粒细胞显著增加。这是第一个能够调控肿瘤浸润中性粒细胞发育的转录因子，为理解肿瘤浸润中性粒细胞的异质性奠定了基础。为了进一步研究其在GEM模型中对肿瘤进展的影响，我们建立了成熟中性粒细胞中缺乏RUNX1免疫系统的骨髓嵌合体GEM模型。为了对NSCLC GEM模型中的肿瘤负荷进行更精确和公正的测量，我们与CCR人工智能资源合作。我们开发了一种用于肿瘤自动量化的人工智能算法。我们的论文《基于人工智能的肺腺癌小鼠模型肿瘤分割》发表在《病理信息学杂志》上。

To target a subgroup of tumor-infiltrating neutrophils (SiglecF+ neutrophils) in the GEM model of NSCLC, we decided to create a mouse model lacking SiglecF expression in neutrophils. Unfortunately, despite multiple attempts, the mouse mode supposed only to lose SiglecF expression in neutrophils has an unforeseen defect: it can't express SiglecF in any cell type, likely due to unknown epigenetic effects. Therefore, we decided to find upstream regulators modulating intratumor neutrophil development and identified RUNX1 as a potential transcription factor based on our single-cell RNA sequencing with single-cell ATAC (Assay for Transposase-Accessible Chromatin) sequencing data. We generated a mouse model lacking RUNX1 only in matured neutrophils (Ly6G-Cre; RUNX1-flox) and discovered a significant increase of SiglecF+ neutrophils in tumor-infiltrating neutrophils. This is the first transcription factor that can regulate tumor-infiltrating neutrophil development, laying the groundwork for understanding the heterogeneous tumor-infiltrating neutrophils. To further study its impact on tumor progression in a GEM model, we have created a bone marrow chimera GEM model with an immune system lacking RUNX1 in matured neutrophils. To have a more precise and unbiased measurement of tumor burden in the GEM model of NSCLC, we collaborated with CCR Artificial Intelligence Resource. We developed an artificial intelligence algorithm for automatic tumor quantification. Our manuscript, "Artificial Intelligence-based Tumor Segmentation in Mouse Models of Lung Adenocarcinoma," was published in the Journal of Pathology Informatics.