https://blog.superhyydl.org/tags/%E5%85%8D%E7%96%AB%E6%8A%91%E5%88%B6/Recent content in 免疫抑制 on Superhyydl's BlogHugozh-cnTue, 01 Sep 2026 00:00:00 +0000https://blog.superhyydl.org/reading/lif-induced-tumor-plasticity-establishes/Tue, 01 Sep 2026 00:00:00 +0000https://blog.superhyydl.org/reading/lif-induced-tumor-plasticity-establishes/<h2 id="一句话亮点">一句话亮点</h2> <p>LKB1缺失让肺癌细胞自己分泌LIF，通过自分泌环路把癌细胞&quot;往回推&quot;成一个Sox17阳性的未分化炎症状态，这个状态负责招募一群&quot;拉偏架&quot;的髓系细胞，把T细胞按在地上摩擦。用抗体把LIF中和掉，这个状态就没了，T细胞重新上岗。</p> <h2 id="背景痛点">背景/痛点</h2> <p>LKB1（也叫STK11）是肺癌里一个相当不招人喜欢的突变基因——它跟KRAS共突变时，患者对免疫检查点抑制剂响应极差，总生存也短。大家都知道LKB1突变会让肿瘤免疫微环境变得很&quot;冷&quot;，但具体怎么搞冷的，一直不太清楚。</p> <p>另外，肿瘤细胞本来就异质性很强，不同细胞状态对免疫系统的影响不一样。但大部分研究用的是移植瘤模型，细胞系在体外传了几十代，早就&quot;同质化&quot;了，很难看到体内那种复杂的细胞状态分化。这篇工作的厉害之处在于，他们用了原位基因编辑的GEMM模型，让肿瘤从小鼠肺里自己长出来，最大程度保留了进化的自然过程。</p> <h2 id="推理链分步拆解">推理链分步拆解</h2> <h3 id="-lkb1突变重塑髓系微环境多了拉偏架的少了站岗的">① LKB1突变重塑髓系微环境：多了&quot;拉偏架&quot;的，少了&quot;站岗&quot;的</h3> <p>他们首先用KPC（Kras/p53/Cas9）GEMM模型，通过气管滴注带sgRNA的慢病毒，特异性在肺上皮里敲除Lkb1。然后对肿瘤里的免疫细胞做单细胞测序。</p> <p>结果很有意思：Lkb1突变肿瘤里，中性粒细胞显著增多，而且这群中性粒细胞高表达SiglecF——之前有文献报道过，SiglecF高的中性粒细胞是促肿瘤的。巨噬细胞那边也一样，Lkb1突变肿瘤里多了高表达Arg1的间质巨噬细胞，肺泡巨噬细胞反而减少了。</p> <p>那这个现象意味着啥？他们进一步确认，这些变化是肿瘤基因型驱动的，而不是肿瘤负荷大了&quot;顺带&quot;的——在第6周（两组肿瘤大小还没差异时），Lkb1突变组的髓系免疫就已经开始往 immunosuppressive 方向偏了。</p> <p>他们还用病人样本做了验证：KRAS/LKB1共突变的LUAD里，Arg1+间质巨噬细胞 signature 富集，而且这个 signature 越高，病人总生存越差。用氯膦酸盐把间质巨噬细胞清掉，T细胞功能恢复，肿瘤也缩小了。</p> <p>@方法论点评：这里最关键的是因果链的第一步——他们先用GEMM忠实地观察到了表型，再用病人数据验证相关性，最后用氯膦酸盐做功能验证（而不是只停留在相关性），确认这群巨噬细胞确实在&quot;干坏事&quot;。</p> <p>![Fig. 1：Lkb1-mutant tumors create an immunosuppressive myeloid niche. A, Schematic of the GEMM to generate autochthonous lung tumors with loss-of-function mutations in Lkb1. B, Uniform manifold approximation and projection (UMAP) showing neutrophils from Lkb1 WT (sgNeo) and Lkb1-mutant (sgLkb1) lung cancer model (n = 2 per condition) (A). Quantification of neutrophil subclusters is shown to the right. C, UMAP with average gene expression of SiglecF in neutrophils. D, UMAP showing macrophages from the tumor model (n = 2 per condition) in A. Quantification of macrophage subclusters is shown to the right. E, UMAP with average gene expression of Arg1 in macrophages is shown. F, Immunofluorescence staining for SPP1 and ARG1 in murine lung tumors. Scale bar is displayed. G, snRNA-seq analysis of myeloid cells from human LUAD. Quantification of myeloid populations in KRAS (n = 14) and KRAS/LKB1 (n = 4) mutant tumors is shown to the right. H, Enrichment of murine Arg1+ interstitial mac­ rophage signature in KRAS and KRAS/LKB1 mutant human lung tumors using bulk RNA-seq data from the TCGA. I, Survival analysis based on strati­ fication of high vs. low murine Arg1+ interstitial macrophage signature in human LKB1 WT LUAD using data from TCGA. J, Mice bearing Lkb1-mutant tumors were treated with clodronate, and tumor volume was measured by MRI after 3 weeks of treatment. Each data point represents an individual mouse. Mean and standard error of the mean (SEM) are displayed. Mann–Whitney U and Log-rank test was used for statistical analysis. , P &lt; 0.05; , P &lt; 0.0001. <a href="https://blog.superhyydl.org/images/reading/lif-induced-tumor-plasticity-establishes/figure-01.png">A, Created in BioRender. Pillai, R. (2026) https://BioRender.com/moxmx7v.）</a></p>