Investigating fibroblast heterogeneity in head and neck squamous cell carcinoma

Abstract

Head and neck cancer is the 8th most common cancer in the UK, with head and neck squamous cell carcinoma (HNSCC) accounting for 90% of cases. Cancer-associated fibroblasts (CAF) form a prominent part of the tumour microenvironment (TME) in a number of solid cancers including HNSCC, where they are associated with protumourigenic functions and poor prognosis. CAF are often marked by alpha-smooth muscle actin (aSMA) and have a myofibroblastic phenotype (myCAF). However, CAF are heterogenous and comprise of various populations, including inflammatory CAF (iCAF). The aim of this work was to provide a thorough characterisation of the fibroblast landscape of HNSCC.
A range of transcriptomic methods were used to identify CAF and corresponding TME associations in HNSCC, including bulk RNA-Seq (TCGA) and deconvolution techniques. In vitro gene expression experiments evaluated plasticity and key pathways connected to fibroblast heterogeneity. Single cell RNA-Sequencing (scRNA-Seq) was used to identify and subsequently characterise fibroblast phenotypes in terms of gene expression profiles, differential abundance, active pathways and transcription factors. Spatial transcriptomics (10x Visium) analysis was used to identify the spatial positioning of fibroblast subsets within HNSCC using deconvolution algorithms, colocalisation with other cell types and ligand-receptor interactions governing fibroblast phenotypes. Multiplexed Immunofluorescence imaging (MxIF) was used to validate spatial transcriptomic cell type associations.
Initial analysis of HNSCC extracellular matrix (ECM) genes showed that CAFhigh tumours were associated with poor survival and the mesenchymal molecular subtype. Two ECM signatures were identified, reflecting two different CAF subsets. This was confirmed by analysis of scRNA- Seq data, and included myCAF and a phenotype resembling universal fibroblasts, which were also present in normal mucosa. Transforming Growth Factor–b (TGFb) signalling linked the two CAF subsets.
Novel HNSCC scRNA-Seq and spatial transcriptomics datasets were then analysed, exploring human papillomavirus (HPV)+ (immune hot) and HPV- HNSCC (immune cold) TMEs. This revealed six clusters of fibroblasts, including myCAF, PI16+ universal fibroblasts and two inflammatory fibroblasts [(IL11+) iCAF and fibroblastic reticular (FRC)-like]. (IL11+) iCAF were spatially associated with inflammatory monocytes, regulated through IL1b and TNFa signalling, and associated with poor survival. FRC-like were enriched in immune hot tumours, associated with CD4+ T cells and B cells in tertiary lymphoid structures, regulated through LTbR signalling and were associated with improved survival and response to checkpoint immunotherapy.
Pan-cancer analysis identified various ‘iCAF’ clusters including IL11+ (i)CAF and IGF1+ (i)CAF. IL11+ CAF reflected a particularly inflammatory subset associated with inflammatory monocytes and were also found in other gastrointestinal and skin cancers, implicated with barrier dysfunction. IGF1+ CAF, transcriptomically similar to iCAFs previously described in pancreatic cancer, were found pan-cancer and reflected a low fibroblast activation state. Inflammatory fibroblasts identified in cancers were also present in inflammatory diseases.
This work clarifies and develops upon our understanding of immunomodulatory fibroblasts in HNSCC and across cancer types, highlighting vital spatial and immunological niches.

More information

Identifiers

ORCID for Benjamin Henry Jenkins: orcid.org/0000-0002-7588-0044

ORCID for Chris Hanley: orcid.org/0000-0003-3816-7220

Catalogue record

Export record