Expression of quantitative trait loci (eQTL) co-localization analyses have identified NTN4 as being a new candidate breast cancer (BC) risk gene, Australian researchers reported in a study published in the August 31, 2020, online edition of the American Journal of Human Genetics.

"Although NTN4 has previously been associated with BC progression, here we provide the first functional evidence that it may also be involved in disease susceptibility," said study co-leader Stacey Edwards.

Functional assays suggest that the NTN4 gene plays a role in BC etiology, but further work is required to confirm the role of NTN4 and other candidate BC risk genes or associated signaling pathways in BC development, which may lead to new prevention or treatment approaches.

Such new approaches are necessary, since "other than tamoxifen and anastrozole, there are very few BC risk reducing medications available," said Edwards, an associate professor in the Cancer Division of the QIMR Berghofer Medical Research Institute in Brisbane.

"Moreover, treatment uptake is low among women at increased risk, due to potentially serious side effects, hence the urgent need to develop safer and more effective risk reduction medications, as well as better treatments," she told BioWorld Science.

In eQTL analyses, genetic variants are tested for association with gene expression levels to identify genes regulated by trait-associated variants.

For BC, several studies have used eQTL data from tumor and normal tissue data-sets to identify candidate genes, while recent studies have also shown that BC risk variants could regulate genes in tumor microenvironment cells, including immune cells and fibroblasts.

Because eQTLs are widespread, overlap between genome-wide association studies (GWASs) and eQTL may occur due to chance, hence the importance of showing that the same genetic signal underlies gene expression and disease susceptibility, to reduce false positive findings.

"It is tempting to use eQTL catalogs to look at whether an SNP [single nucleotide polymorphism] found through a trait association study is linked to the expression of any genes," said study co-leader Jonathan Beesley, a senior research officer at QIMR Berghofer.

However, "this can lead to spurious findings, because eQTLs are so widespread in the genome there is probably an association between every gene and at least one nearby SNP," he cautioned.

As a consequence, "more sophisticated statistical analysis must be performed to show that the same SNPs underlie both associations with the trait and gene expression."

Several statistical co-localization approaches have already been developed to determine whether molecular traits such as gene expression and a disease trait do indeed share common causal variants.

The simplest model used in tools such as the COLOC test for co-localization for two traits determines whether they are driven by distinct variants or share a single causal signal.

Breast cancer GWASs have previously identified 150 genomic risk regions containing more than 13,000 causal variants (CCVs), which are predominantly noncoding and enriched in regulatory elements, but the genes underlying BC risk associations are largely unknown.

In their new AJHG study, researchers co-led by Edwards, Beesley and their QIMR Berghofer colleague, associate professor Juliet French, used genetic eQTL co-localization analysis to identify loci at which gene expression could potentially explain BC risk phenotypes.

Using data from the Breast Cancer Association Consortium (BCAC) and QTL from the Genotype-Tissue Expression (GTEx) project and The Cancer Genome Project (TCGA), they identified shared genetic relationships and revealed new associations between cancer phenotypes and effector genes.

This analysis identified 17 genes, including NTN4, as being potential mediators of BC risk.

For NTN4, they demonstrated that the rs61938093 CCV at this region was located within an enhancer element that physically interacts with the NTN4 promoter, and that the risk allele reduced NTN4 promoter activity.

"This study provides an example of combining genetic epidemiology, molecular biology and animal models to demonstrate how an associated genomic variant might influence risk of BC," said Edwards.

"Knowledge of such functional mechanisms might help in the interpretation of other association study findings," she told BioWorld Science.

Moreover, knockdown of the NTN4 gene in breast cells was demonstrated to increase cell proliferation in vitro and tumor growth in vivo.

"For the in vitro anchorage-dependent and -independent cell growth assays, knockdown was achieved using short interfering RNAs in human BC cells, in which NTN4 knockdown increased breast cell growth by approximately 50%," explained Edwards.

"In the in vivo assays in female mice, knockdown was achieved using CRISPR (clustered regularly interspaced short palindromic repeats)-interference based NTN4 repression in human BC cells, in which NTN4 knockdown increased breast tumor growth by more than 150%."

Together, these data provide evidence linking risk-associated variation to genes that may contribute to BC predisposition, which has prevention/treatment implications.

Treatment implications

"This unbiased approach to GWAS can lead to unanticipated discoveries, such as the involvement of genes and pathways not previously linked to BC development," said Edwards.

"While it is not directly targeted by current treatments, it is possible that knowledge of NTN4's role in BC may open up new avenues to drug development."

"Our genetic and experimental evidence suggests that reduced NTN4 expression is associated with increased breast cell growth, suggesting it would not be a suitable drug target in BC.

"However, NTN4 participates in a variety of signaling pathways and it may be possible to target upstream regulators to maintain or increase its expression."

Looking forward, she said, "we are currently performing large-scale pooled CRISPR knockout and activation screens of all GWAS-identified target genes to evaluate their potential roles in BC development." (Beesley, J. et al. Am J Hum Genet 2020, 107: 1).