Spatial omics accelerate the study of cancer-associated fibroblasts in non-small lung cancer

Tumour microenvironment (TME) is one of the important factors associated with cancer progression. TME is a multicellular system composed of fibroblasts, endothelial and immune cells, distributed in the extracellular matrix (ECM), and closely interacts with tumour cells to promote the occurrence and development of cancers. In TME, secreted products of various immune and non-immune cell types, such as cytokines and chemokines, as well as metabolites, hypoxia, angiogenesis and ECM remodelling drive chronic inflammation.¹ Recent studies have shown that cancer-associated fibroblasts (CAFs) are important regulators of anti-tumour immune response. CAFs can reshape TME by secreting a variety of cytokines, thereby promoting immune escape. Therefore, targeting CAFs may improve the effectiveness of immunotherapy.^{2, 3} For example, NOX4 inhibitors can reverse the formation of CAFs, thereby restoring anti-tumor immunotherapy efficacy.³ However, due to the high heterogeneity of the CAF population, its origin and function remain unclear, and the lack of understanding of these issues greatly limits the clinical transformation of CAFs.⁴

The development of single-cell transcriptome sequencing (scRNA-seq) has brought fundamental advances in cancer research. The subpopulations and functions of CAFs in non-small-cell lung cancer (NSCLC) have been described in several studies using scRNA-seq.^5-7 However, due to the loss of spatial information, most studies limited their research focus to immune cells in tumours. The emergence of spatial omics technology has made up for the shortcomings of single-cell sequencing, using spatial transcriptome technologies, it is expected to characterize the molecular characteristics and immune regulatory functions of CAFs in cancer.

In a recent study by Xu et al., a subpopulation of CAFs, POSTN⁺ CAFs were found to have a close localization with SPP1⁺ macrophages, and correlated with exhausted phenotypes and lower infiltration of T cells in NSCLC.⁸ Initially, the study identified diverse fibroblast subpopulations in NSCLC through the integration of fibroblasts in the Peking cohort (N = 1986) with fibroblasts in Samsung cohort⁶ (N = 3499) and Tongji cohort⁷ (N = 4497). Several iCAF subpopulations (clusters C01_CCL11, C05_IGF1 and C06_CCL2), adventitial fibroblasts (C03_PI16), alveolar fibroblasts (C04_COL13A1), as well as myCAF subgroups (C02_POSTN, C07_MKI67 and C09_MYH11) were identified. Specifically, they found that POSTN⁺ CAFs (C02_POSTN) were enriched in tumour/metastatic samples compared to normal samples. Gene Set Variation Analysis was performed and POSTN⁺ CAFs showed activities in several pro-invasive pathways such as “angiogenesis” and “epithelial-mesenchymal transition”, suggesting the pro-invasive function of these cells.

Furthermore, a new spatial RNA-seq method called spatially enhanced resolution omics-sequencing (stereo-seq) was performed to illustrate the location of POSTN⁺ CAFs in five tumour samples. Stereo-seq utilized mRNA capture in situ by DNA nanoballs (DNBs) with approximately 220 nm diameter and a centre-to-centre distance of 500 nm.⁹ Seurat was used to cluster the spatial gene expression of each specimen separately at bin100-defined unit (100 × 100 DNBs, i.e. ∼5 x 50 µm area). At a resolution of bin100, spatial clustering could not achieve single-cell resolution but still reflected the spatial proximity of different cell types. Interestingly, four out of five samples showed that POSTN⁺ CAFs and SPP1⁺ macrophages were in proximity. Several ligand-receptor pairs between POSTN⁺ CAFs and SPP1⁺ macrophages were identified, such as COL4A1-ITGB1, Tenascin-C-integrins, which might enhance the immune-modulatory activities of SPP1⁺ macrophages and the phenotypes of POSTN⁺ CAFs through regulating various target genes. Moreover, scRNA-seq data suggested that POSTN⁺ CAFs were associated with lower infiltration and exhausted phenotype of T cells. To validate their findings, Xu et al. performed multiplexed immunohistochemical staining and CIBERSORTx analysis using FFPE tumour samples and The Cancer Genome Atlas (TCGA)-NSCLC samples, respectively. In addition, clinical analysis using TCGA-NSCLC samples suggested that POSTN⁺ CAFs were associated with cancer progression and poor clinical outcomes.

The recent discovery by Xu et al. revealed the pro-tumor and immunosuppressive roles of POSTN⁺ CAFs, which may be targeted to improve immune checkpoint inhibitor response in NSCLC. This is a timely study given the interest in single-cell and spatial technologies and will likely be of great interest and impact in the field. The study's objective is significant because immunotherapies are still not effective for NSCLC and characterizing the TME is a promising approach to designing successful therapies. However, there are still several questions that remain to be answered: How do POSTN⁺ CAFs interact with other cell types? What is the relationship between POSTN⁺ CAFs and T-cell exhaustion, and what is the molecular mechanism behind it? What are the function and spatial distribution of other CAF subpopulations? Recently, a study by Li et al. revealed that POSTN⁺ CAFs serve as pivotal components within the onco-fetal ecosystem of hepatocellular carcinoma (HCC), promoting the progression of HCC through a process termed onco-fetal reprogramming.¹⁰ As POSTN⁺ CAFs also exist in NSCLC, it remains to be determined whether this reprogramming process also operates in NSCLC. In summary, this study is a catalyst for further research into the role of CAFs. Further studies are expected to reveal the distribution trends and molecular characteristics of different subpopulations of CAFs in tumours and clarify their broader significance in the TME.

Haozhen Liu, Chao Chen: Conceptualization; original draft writing and editing. Jixian Liu, Chao Chen: Critical revision, editing and approval of the final version of the article.

The authors declare no conflict of interest.

Not applicable.