NEW YORK (GenomeWeb News) – In a paper appearing in the early, online version of the Proceedings of the National Academy of Sciences last night, a London-based research team described how it used an RNA interference screen in conjunction with massively parallel sequencing to track down nearly two dozen genes that seem to influence tamoxifen response.
Their results suggest resistance to the breast cancer drug was tied to small RNA-induced silencing of 11 of the genes detected in the cell-based screen. Conversely, sensitivity to the drug was bumped up in breast cancer cells when another 11 genes were silenced. Meanwhile, gene expression patterns from a handful of patients who received tamoxifen as part of their breast cancer treatment seem to support the patterns detected in the screen, researchers found, indicating that levels of some of these genes coincided with clinical outcomes.
"This screen and the subsequent validation experiments define a constellation of genes that modulate the cellular response to this widely used drug," co-corresponding author Alan Ashworth, a molecular biology researcher with the Institute of Cancer Research in London, and co-authors wrote.
Because of its ability to inhibit estrogen receptor signaling, tamoxifen has become a commonly used drug for treating estrogen receptor positive breast cancers, researcher explained. Even so, they noted, tamoxifen resistance and relapse remain important obstacles to effective breast cancer treatment.
In an attempt to track down genetic factors mitigating tamoxifen response, the team came up with a high-throughput RNAi screen that used a set of 56,670 short hairpin RNAs coded by lentiviral sequences to systematically knock down 16,487 genes across the genome.
After using pools of these shRNAs to infect a breast cancer cell line called MCF7 — which is known to be estrogen receptor positive and sensitive to the active tamoxifen metabolite 4-hydroxytamoxifen — researchers exposed the cells to a 21-day 4OHT treatment.
They then used massively parallel sequencing to compare the shRNA repertoires in the treated cells with those found in mock-treated cells, using this information to extrapolate insights into which genes contribute to tamoxifen resistance and sensitivity.
Once they had narrowed in on 121 candidate genes that seemed to contribute to drug sensitivity and 131 potential resistance genes, the team used a second RNAi screen to verify authentic interactions for 23 of the genes.
Subsequent dose-response experiments showed that 11 of these genes were linked to drug resistance and 11 to sensitivity. Among them: genes coding for proteins involved in sister chromatid cohesion, chromatin remodeling, RAS signaling, and more.
Clinical data seems to support a role for at least some of the genes in tamoxifen response, researchers found. When they looked at gene expression data for five individuals who had been treated with tamoxifen, they found that relapse risk was higher in individuals with lower than usual levels of the neurofibromin gene NF1, one of the RAS signaling-related genes identified in the screen.
By bringing together information garnered from their RNAi screen with previously known tamoxifen-response genes, the team also identified so-called response "metagenes" — sets of eight genes predicting tamoxifen resistance and six tied to sensitivity to the drug.
Based on their tamoxifen-related findings, those involved in the study argue that a similar screening approach also holds promise for doing other kinds of disease or drug response studies.
"[T]he experimental setup applied here, with the deconvolution of pooled shRNAs by massively parallel sequencing, allows genome-wide screens to be performed at a fraction of the cost and time compared with genome-wide plate arrayed screens," they concluded, "and will likely enable the dissection of additional disease-based phenotypes."