The landscape of long noncoding RNA during cutaneous squamous cell carcinoma progression

Abstract

BACKGROUND: Cutaneous squamous cell carcinoma (cSCC) is a common cancer with a high morbidity rate and poor prognosis for metastatic disease. Disease may progress from premalignant actinic keratosis to invasive and metastatic cSCC, but it is perhaps best characterized as a disease continuum progressing from a differentiated to a progenitor-like state. The critical molecular mediators of this process remain poorly defined. Long noncoding (lnc)RNAs, a relatively unexplored class of RNA molecules > 200 nucleotides long, are likely to have important functional roles in cSCC. OBJECTIVES: To provide a comprehensive landscape of lncRNA expression during the cSCC continuum and to identify potentially functional lncRNA drivers of disease progression. METHODS: We interrogated bulk RNA sequencing (RNAseq) data from 110 patient samples, encompassing healthy sun-exposed skin (n = 26), actinic keratosis (n = 14), primary cSCC (n = 66) and metastases (n = 4), to identify changes in lncRNA expression during disease progression. We developed a bioinformatics pipeline to infer lncRNA function based on co-expression patterns and generated a lncRNA signature score, which we validated in head-and-neck squamous cell carcinoma (HNSC) and pancreatic adenocarcinoma (PAAD). We performed bulk RNAseq on 15 patient-derived cell lines and integrated these data to identify tumour cell-specific lncRNAs and validated our findings in multiple other cSCC gene expression cohorts. Using in vitro knockdown approaches we investigated the functional role of LINC00941. RESULTS: We found that lncRNA expression alone is sufficient to identify disease states and progression along the cSCC disease continuum. Correlation analysis revealed potentially functionally relevant lncRNAs and the processes they may regulate. We developed a 267 lncRNA signature that correlates with a progenitor-like state and predicts poor prognosis in HNSC and PAAD. Bulk RNAseq of patient-derived cell lines revealed tumour cell-specific lncRNAs, and knockdown of LINC00941 indicated that it is required for cell proliferation and colony formation in vitro. CONCLUSIONS: Our findings provide a comprehensive description of lncRNA transcriptomic changes in cSCC and demonstrate their functional relevance as biomarkers and drivers of disease progression in this and, potentially, other cancers.; Cutaneous squamous cell carcinoma (or ‘cSCC’ for short) is a type of skin cancer. It is one of the most common cancers worldwide. cSCC is usually caused by prolonged skin exposure to ultraviolet (UV) light (sunlight). This damages the genetic material of skin cells and leads to mutations that cause cancer. Mutations to a specific type of skin cell known as ‘keratinocytes’ lead to the formation of pre-cancerous lesions known as ‘actinic keratoses’. cSCC commonly develops from these lesions. cSCC is usually treated by removing it surgically. However, if it is left untreated, it can become invasive and grow as tumours in other parts of the body. When this happens, it is known as ‘metastatic cSCC’ and the prognosis for the patient is much worse. This will happen in about 1% to 5% of people with cSCC. Treatment in these cases is limited. In our study, we looked at a specific family of genes called ‘long non-coding RNAs’ (lncRNAs). We did this to try and understand how they might contribute to the progression of cSCC. To do this, we studied lncRNA expression in healthy skin and compared the information with skin affected by cSCC. We found that lncRNA expression is enough to tell us about the stage of cSCC development. We identified 267 lncRNAs that may be important in the disease progression of cSCC. We developed a scoring system for these lncRNAs and showed how this can help with the prognosis of head and neck cancer and in pancreatic cancer. We looked in detail at the function of one of these lncRNA molecules (called ‘LINC00941’) and found that it is required for cancer cells to grow. Our findings include a description of lncRNA changes in cSCC. They also show how lncRNAs could be used to spot the disease and understand how it progresses.