Abstract

Gene expression based consensus molecular subtypes (CMS) and non-negative matrix factorization (NMF) sub-clusters are robust colon cancer classification systems. Although, the molecular features are clear, colon cancer subgroups based interventions are limited. To address this problem, we analyze the CMS and NMF subgroup guided drug sensitivity in colon cancer cell lines. CMS3 subtype cells are sensitive to 5-Fluorouracil, while, CMS4 subtype cells are sensitive to cisplatin treatment. In NMF classification, a sub-cluster is specifically sensitive to chemotherapy, BRAF inhibitors, PI3K-mTOR inhibitors and NOTCH inhibitor treatment. This sub-cluster has low frequency of TP53, POLE, PIK3CA and BRAF mutation. Transcriptional analysis demonstrates low NOTCH signaling activity, low CDX2 and VDR expression in this sub-cluster. CDX2 and VDR are significantly associated with the sensitivity of chemotherapy, BRAF inhibitors and PI3K-mTOR inhibitors. Moreover, a positive correlation between VDR and CDX2 is identified. VDR and CDX2 mediated regulatory networks are constructed. At last, three or four sub-clusters classification is validated in colon cancer patients. Overall, our results suggest a molecular sub-cluster of colon cancer cells with low CDX2 and VDR expression is sensitive to chemotherapy, BRAF inhibitors and PI3K-mTOR inhibitors treatment and provide an example of translation of cancer classification to subgroup guided therapies.

Highlights

  • Colon cancer is a heterogeneous disease with distinctive genetic and epigenetic alterations [1, 2]

  • Our results suggest a molecular sub-cluster of colon cancer cells with low CDX2 and VDR expression is sensitive to chemotherapy, BRAF inhibitors and PI3K-mTOR inhibitors treatment and provide an example of translation of cancer classification to subgroup guided therapies

  • We find that a molecular sub-cluster of colon cancer cells with low CDX2 and VDR expression is sensitive to chemotherapy, BRAF inhibitors and PI3K-mTOR inhibitors treatment

Read more

Summary

Introduction

Colon cancer is a heterogeneous disease with distinctive genetic and epigenetic alterations [1, 2]. In 2015, Justin Guinney and colleagues integrated the expression data of 4,151 patients from 18 published colon cancer datasets and proposed the CMS classification of colon cancer, including CMS1 microsatellite instability (MSI) immune, CMS2 canonical, CMS3 metabolic and CMS4 mesenchymal four classes [12]. In 2013, Anjuraj Sadanandam and colleagues analyzed the expression data of 1,290 colon cancer patients from published datasets and divided those colon cancer patients into goblet-like, enterocyte, stem-like, inflammatory and transit-amplifying five subtypes based NMF classification [15]. The stem-like colon cancer was associated with the clinical benefit of FOLEIRI treatment. The transit-amplifying colon cancer was associated with the clinical benefit of EGFR inhibitor cetuximab or c-MET inhibitor treatment. Other subgroup based targeted interventions were not further analyzed

Methods
Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call