Cell type-specific gene network reconstruction framework

Today, we'll have a look at a new study identifying genes associated with breast cancer metastasis. Breast cancer is a common and deadly disease worldwide, with up to 30% of breast cancer deaths caused by metastasis. Functional studies at the genomic level have been conducted to understand the high variance in time for relapse. However, differences in tumor microenvironments and stemness qualities of specific cancer clones that may provide important mechanistic insight are expected to be masked at the bulk tissue level. Single-cell transcriptome data has become popular in studying the heterogeneity of normal and disease-associated tissues, and it provides new opportunities to explore particular cell types. However, direct measurement of gene co-expression is considered inaccurate due to the high dropout rate and probable transcriptional bursts present in single-cell data. To overcome this, data imputation has been proposed to retrieve functional gene associations via co-expression analysis of cells.

In a new study, researchers proposed a cell type-specific gene network reconstruction framework with data imputation to infer functional connections from single-cell RNA sequencing (scRNA-seq) data to identify novel genes associated with breast cancer metastasis. Two publicly available scRNA-seq datasets for breast cancer were used to construct breast cancer cell gene networks (CGNs) through a log likelihood score (LLS) scheme based on the Bayesian statistical framework. The study identified several novel genes associated with breast cancer metastasis, including differentially expressed genes (DEGs) that were highly ranked in the CGNs of metastatic cancer cell lines.

The researchers reconstructed the networks for primary and metastatic cancer cells and compared their topologies to identify the genes involved in the metastatic properties of cancer cells, particularly breast cancer. To assess the validity of their network-based predictions, they collected 664 genes associated with cancer metastasis from previous databases and published articles. They also evaluated the enrichment of the known breast cancer metastatic genes for the top candidates using DEGs or network hubness.

The researchers used a gene knockdown (KD) assay in the MDA-MB-231 cell line, which is a highly invasive and aggressive breast cancer cell line. They selected four genes - FAM111A, CDK5RAP2, GGCT, and CCT2 - which were predicted to be among the top 20 candidates for metastatic genes. They found that knockdown of the CCT2 gene significantly reduced relative wound density compared to the negative control and was comparable to the positive control, suggesting that CCT2 may play a role in breast cancer metastasis. CCT2 is a molecular chaperone gene that has been suggested to play a role in breast tumorigenesis, and higher CCT2 expression in patients with colorectal cancer indicates poor prognosis.

Overall, this study provides a novel approach for identifying genes associated with breast cancer metastasis using single-cell transcriptome data and data imputation. The results of the study could potentially lead to the development of new therapeutic targets for breast cancer metastasis.