NEW YORK (GenomeWeb) – Structural rearrangements affecting the androgen receptor may contribute to as many as one third of castration-resistant prostate cancer tumors, new research suggests.
Researchers from the University of Minnesota and elsewhere used targeted sequencing to interrogate androgen receptor gene (AR) locus patterns in multiple samples from 15 individuals with castration-resistant prostate cancer. Based on patterns in dozens of tumor samples from these and other prostate cancer patients, they characterized a wide range of genomic structural rearrangements involving the androgen receptor, often leading to amped up activity in the androgen receptor signaling pathway.
Senior author Scott Dehm, a cancer researcher from the University of Minnesota, and his co-authors noted that these and other findings — published today in Nature Communications — suggest androgen receptor genomic structural rearrangement (AR-GSRs) are "important drivers of persistent AR signaling in [castration-resistant prostate cancer."
Past studies have provided a look at the mutations and copy number changes associated with resistance to androgen deprivation treatments that target the androgen receptor, the team explained, and there is evidence that altered AR transcript splicing can affect response to such treatments as well. Consequently, the group decided to delve into the rearrangements that might impact the locus in individuals with prostate cancers that have advanced despite treatment.
"The presence of AS-GSRs in clinical prostate cancer has not been described by previous large-scale tumor genomic profiling studies, probably because these studies have employed copy number arrays, whole-exome sequencing, or moderate-coverage whole-genome sequencing, which are not powered for robust structural variant discovery in a high repeat-content gene such as AR," the authors wrote.
To directly assess the nature and prevalence of AR gene rearrangements, the researchers used a liquid phase bait panel, coupled with Illumina HiSeq 2500 sequencing, to do targeted sequencing on a region spanning almost 200,000 bases at the AR locus in 30 metastatic, castration-resistant prostate cancer tumor samples obtained by rapid autopsy from 15 individuals who died of the disease between 2000 and 2013.
They uncovered AR gene amplifications in a dozen tumor samples from six of the patients, while six more metastatic tumor samples harbored missense mutations involving the AR gene. Meanwhile, genomic structural rearrangements affecting the AR locus turned up in 10 metastatic tumor samples from six individuals, including three copy-number neutral metastases and seven tumor samples from individuals who also had AR amplifications.
When the team sequenced the AR gene locus in localized tumor samples from six more individuals with castration-resistant prostate cancer and 21 individuals with prostate cancers that hadn't been treated with hormone therapy, it found AR-GSRs in two samples from castration-resistant prostate cancer patients — rearrangements that were not found in any of the hormone-naïve tumor samples.
Using PCR amplification and Sanger sequencing, the researchers went on to refine their view of the AR-GSR breakpoints in the tumor sample, identifying multiple AR rearrangements in seven of the castration-resistant prostate cancer tumors. They also pulled in quantitative RT-PCR and/or RNA sequence data for the castration-resistant prostate cancers to explore relationships between AR splicing, expression, and rearrangements at the AR locus.
"Our integrative analysis indicated that tumors harboring AR-GSRs at high variant allele fractions expressed tumor-specific [AR messenger RNA splicing variants (AR-Vs)]," the authors noted. "Conversely, reconstruction and validation of certain AR-GSR events present at low variant allele fractions revealed that some of the resultant AR gene architecture may not be compatible with AR-V, or even full-length AR, expression."