Depicting novel tumour cell states through single-cell transcriptome profiles

November 28, 2022

Colorectal cancer (CRC) is among the most common and lethal cancers in the world, being a considerable public health challenge. One possible explanation for this is that CRC presents a high inter- and intra-tumoral heterogeneity, which hampers its treatment. Therefore, depicting novel tumour cell states is becoming urgent to deepen our understanding of CRC and improve the management of the disease. In particular, studies at a single-cell resolution are fundamental to capturing tumour variability. Here, we used publicly available single-cell RNA-sequencing (scRNA-seq) data from normal and tumour colorectal tissues to study the existence of a confused cell state, in which epithelial cells simultaneously express gene signatures of terminally differentiated epithelial cell types that belong to distinct lineages. In fact, we hypothesised that this confused cell state is a feature of CRC and, possible, of cancer in general. To test this hypothesis, we trained a single-cell classifier based on the epithelial cell types and corresponding gene signatures that we previously determined using normal colorectal samples. Then, we used this classifier to annotate tumour samples and identify cells with multiple identities. Finally, we also compared the classification results obtained for normal and tumour tissues. Overall, we were able to identify cells in a confused cell state, not only in tumour samples but also in normal tissues. Furthermore, the percentage of these cells in both conditions was not very different. Thereby, our results indicate that the confused cell state is a normal feature of the intestinal epithelium rather than a hallmark of CRC.

Authors in the community:

• 

  M

  Maria Francisca Telles de Freitas Xara Brasil

  ist189825

Supervisors of this institution:

• 

  C

  Cláudia Lobato da Silva

  ist31519

industrial-biotechnology - Industrial Biotechnology

eng - English

Instituto Superior Técnico