Single-cell and spatial analysis reveal interaction of FAP+ fibroblasts and SPP1+ macrophages in colorectal cancer
Nat Commun.(IF16.6), 13, 1742 (2022). DOI: https://doi.org/10.1038/s41467-022-29366-6.
Key point:
FAP+Fibro和SPP1+Macro互作形成免疫屏障促进肿瘤生长
NitcheNetR用于细胞类型间关键配受体及靶基因的筛选

结直肠癌(CRC)是继肺癌和乳腺癌后的第三大恶性癌症，按照肺癌乳腺癌的方法治疗CRC，仅对部分癌症类型有效(high microsatellite instability (MSI-H))，因此探究结直肠癌的病理机制以及潜在治疗靶点是非常有必要的。

However, the PD-1-targeting antibody,pembrolizumab, is only effective for mismatch repair-deficient tumors with high microsatellite instability (MSI-H), which account for <5% of metastatic CRC cases. Therefore, it is necessary to understand the mechanism of cellular and molecular remodeling in the tumor microenvironment (TME) of CRC and find potential intervention targets to enhance the therapeutic efficacy of immunotherapy.

近期研究表明基质细胞和髓系细胞能够形成与众不同的位点促进肿瘤生长和迁移，这可能会作为潜在的细胞靶点。

Recent research has revealed that stromal cells and myeloid cells may form a distinctive niche for tumor growth and metastasis, making them potential therapeutic targets.

本项工作证明，高表达的FAP或SPP1抑制PDL1相关的免疫治疗，总的来说，发现一项潜在的结直肠癌治疗的潜在靶点。

Furthermore, high expression of either FAP or SPP1 contributes to resistance to PDL1 blockade immunotherapy. Together, this work unravels the complex interplay between stromal and macrophages subsets, which could serve as potential targets for CRC immunotherapy.

通过2例结肠和3例直肠癌样本进行癌和癌旁的10X单细胞测序并获得54103高质量细胞，以及Seurat分析鉴定9大类细胞(Epithelial, T/ILCs, B, PlasmaB, Myeloid, Mast, Endothelial, Mesenchymal stromal cells(MSCs), Glial(肠神经细胞))

The cells were classified into nine major cell types (Fig. 1b–d), including epithelial cells (n=8940) identified by the expression of EPCAM and CDH1, T/ILCs cells (n=17,420) which expressed the T-cell receptor (TCR) signaling mediators CD3E and CD3G15, B cells (n=2998) marked by MS4A1 and CD79A, plasma cells (n=7252) identified by SDC1 and MZB1 expression, myeloid cells (n=4617) which were positive for CD14 and FCGR3A expression, mast cells (n=2781) defined by their classical markers KIT, IL1RL1, and MS4A2, endothelial cells (ECs; n=2205) marked by PECAM1 and CDH5, mesenchymal stromal cells (MSCs; n=7451) marked by COL1A1 and COL3A1, and glial cells (n=439) marked by S100B and CDH2.

HallMark signature分析对比推断不同细胞类型在癌和癌旁的功能差异，如免疫细胞多涉及IL2/STAT5 signaling, and IL6/JAK/STAT3 signaling

Hallmark gene set for hypoxia was more enriched in MSCs, ECs, and myeloid cells from tumors than those of normal samples tumorinfiltrating myeloid cells and T/ILC exhibited greater enrichment of metabolism-related genes, including fatty acid metabolism, xenobiotic metabolism, bile acid metabolism, and cholesterol homeostasis, than those cells from normal mucosa. Taken together, these findings indicated that the regulatory pathways of major cell types were shaped in the CRC TME.

鉴于样本量少问题，文章通过CIBERSORTx方法模拟TCGA和GEO数据中细胞组分来验证细胞类型丰度，同时通过14个TCGA数据和98+55个GEO数据计算不同细胞类型间的相关性（pairwise Spearman correlations），发现MSCs and myeloid cells显著相关。通过生存分析评估，MSCs和Myeloid的存活周期都显著较短(OS/PFS)

We analyzed pairwise Spearman correlations within the infiltration patterns of nine major cell types across 14 independent CRC cohorts. We observed a significantly positive correlation between MSCs and myeloid cells in all interrogated cohorts (Spearman correlation coefficient [|Rs|] > 0.3 and false discovery rate [FDR] < 0.05 were considered significant correlation), with the Rs ranging from 0.47 in the GSE18105 dataset with 98 CRC samples to 0.76 in the GSE17537 dataset with 55 CRC samples. These analyses revealed that CRC patients in the TCGA CRC cohort (Fig. 2d, e) with higher MSC infiltration were associated with worse OS (log-rank test, p=0.02) and PFS (logrank test, p=0.00071). Greater infiltration of myeloid cells was also associated with worse OS (logrank test, p=0.026) and PFS (log-rank test, p=0.033; Fig. 2d, e).

文章根据细胞表达特征将MSCs分为10个亚类：ICAM1+ telocytes (n = 1016) and ICAM1− telocytes (n = 550), DES+ myofibroblasts, MFAP5+ myofibroblasts, CD24+fibroblasts (n = 228), NT5E+fibroblasts(n = 930), FGFR2+ fibroblasts (n = 2023), FAP+fibroblasts49 (n = 1091), Pericytes, Proliferating fibroblasts. 通过比例统计发现FAP+fibroblast、pericyte、proliferating fibroblasts是在肿瘤组织显著富集的。当前推测Diff应该是N/T占比平均值之差即FAP+/Total Fibro in T - FAP+/Total Fibro in N; p值是两组间Percent的双尾配对t检验。并且在不同数据中TCGA/GSE通过CIBERSORTx计算的比例在N/T组间显著差异。

The FAP+ fibroblasts (Diff=42.4%, p=0.0052), proliferating fibroblasts (Diff=2.29%, p=0.0099), and pericytes (Diff= 10.4%, p=0.031) were markedly enriched in tumor tissue as compared to that from adjacent normal tissue, while NT5E+ fibroblasts (Diff = 14.7%, p=0.0053), FGFR2+ fibroblasts (Diff = 19.3%, p=0.015), ICAM1− telocytes (Diff=−6.53%, p=0.0071), and MFAP5+ myofibroblasts (Diff = 9.84%, p=0.0066) were enriched in tumor-adjacent normal tissue.

FAP+fibroblasts来源是通过Velocyte.R得到的箭头直接判定，并通过pySCENIC计算N/T各Fibroblasts细胞的top转录因子及top细胞marker的表达，并发现TWIST1可能是FAP+fibroblasts产生的关键转录因子。

we performed RNA “velocity” analysis, a computational approach that utilizes nascent transcription in scRNA-seq datasets to infer differentiation trajectories of FAP+ fibroblasts. This analysis predicted that tumor-specific FAP+ fibroblasts likely originated from FGFR2+ fibroblasts or ICAM1+ telocytes. We found that TWIST1 had the highest expression and activity levels in the regulatory network of FAP+ fibroblasts and may represent the master TF driving this differentiation pathway.

髓系细胞主要分为3种DC，4种巨噬，嗜中性粒细胞和增殖髓系细胞. 不论细胞占比还是生存分析，或是CIBERSORTx分析结果都显示SPP1+ macrophage在肿瘤像本显著富集。通过Velocyte.R和pySCENIC进行起源和转录因子分析获得起源细胞类型和关键转录因子。

Three dendritic cell (DC) subtypes were identified (Fig. 4a). Activated DCs (n = 73) were marked by high expression of CCR7, required for innate lymphoid cell trafficking to draining lymph nodes55, and FSCN1, essential for DC maturation56 (Fig. 4a and Supplementary Fig. 2c). XCR1 and CLEC9A were highly expressed in cDC1 (n = 91), while FCER1A and CD1C57 were upregulated in cDC2 (n = 487), THBS1+macrophages (n = 1028) were characterized by THBS1 expression, which was shown to activateM1-like TAMs and promote the malignant migration of cancer58. In addition, VCAN+ macrophages (n = 600) were identified by high expression of VCAN, and another macrophage subtype, C1QC+ MRC1− macrophages (n = 1480), were positive for expression of the complement component gene C1QC but lacked MRC1 expression macrophages showed high expression of SPP1 and the scavenger receptor MARCO59 (n = 443). Furthermore, we observed a neutrophil population characterized by CSF3R expression57 and proliferating myeloid cells known for MKI67 expression.

为了进一步探究具体亚类的细胞功能和关系，通过计算细胞亚类间的spearman相关性，结果显示SPP1+macrophage和FAP+fibroblast相关性极高。并且比较不同组合的细胞类型组合的生存分析显示高表达的FAP+/SPP+最短的存活周期。通过TCGA的组间差异(SPP+/FAP+ VS SPP-/FAP-)和富集结果显示这些差异主要涉及ETM/Hypoxia/TNFα等肿瘤生长相关通路和功能。

We found that the FAP+ fibroblasts and SPP1+ macrophages were the most highly correlated populations across all examined cohorts (Fig. 5a, Supplementary Fig. 7). To further uncover the clinical implication of such close correlation between these two cell types, we compared the PFS of patients with different levels of FAP+ fibroblasts and SPP1+ macrophages. Patients with both high FAP+ fibroblasts and SPP1+ macrophages exhibited the shortest PFS compared with the signature combination groups, suggesting that these two cell types can synergistically promote tumor progression. To further understand the potential triggers or downstream signals that induce FAP+ fibroblasts and SPP1+ macrophages, we calculated the differentially expressed genes between FAP+ fibroblastsHigh SPP1+ macrophagesHigh and FAP+ fibroblastsLow SPP1+ macrophagesLow groups in a TCGA-COAD cohort and performed GSEA analysis subsequently. Genes upregulated in samples with FAP+ fibroblastsHigh SPP1+ macrophagesHigh showed enrichment of epithelial–mesenchymal transition signatures (Fig. 5c). These samples also displayed highly enriched hypoxia gene set (Fig. 5c), which is consistent with the supposed hypoxic environment of tumors53. Furthermore, TNFα signaling and IL2/STAT5 signaling pathways were also enriched in FAP+ fibroblastsHigh SPP1+ macrophagesHigh tumor samples (Fig. 5c). These findings suggest that FAP+ fibroblasts and SPP1+ macrophages respond to different stimuli and signals within the TME of CRC.

空间转录组结果证明SPP1+macrophage和FAP+fibroblast有共定位的现象，且主要集中在肿瘤区域。

Most FAP+ fibroblasts and SPP1+ macrophages were colocalized at the same spots, and a spot in the 10x Genomics Visium platform could accommodate up to 10 cells, suggesting that there was a physical interaction between the two cell types. Furthermore, we found spots with FAP+ fibroblasts and SPP1+ macrophages highly active in pathways that contribute to the formation of desmoplastic structures in four ST datasets, including extracellular matrix organization, collagen fibril organization, and response to TGF-β (Fig. 6l). T cells or B cells were excluded from tumor core. These results suggest the desmoplastic microenvironment may be regulated by the interaction of FAP+ fibroblasts and SPP1+ macrophages to limit the infiltration of immune cells into the tumor core.

NicheNet分析结果显示FAP+fibroblast和SPP1+macrophage 通过配受体 COL1A1/LAMA1-ITGB1直接互作来重构组织环境便于肿瘤生长，主要通过促进SPP1+macrophage的炎症反应，以及通过WNT5A-FZD2 and CCL3-CCR5 pairs 招募SPP1+macrophage，而且这些基因的表达都相对于其他细胞亚类高。