NEW YORK (GenomeWeb) – A team led by researchers at the University of Michigan has catalogued and started characterizing circular RNAs (circRNAs) found across cancer types, including proposed circRNA markers in the urine of prostate cancer patients. Independently, another group led by researchers at the University of Toronto has identified circRNAs with functional roles in prostate cancer.
As they reported online today in Cell, the UMich researchers and their colleagues used exome capture and RNA sequencing to profile circRNAs in more than 800 tumor, cell line, or pooled normal samples assessed for the Michigan Oncology Sequencing Center (MI-ONCOSEQ) program. The analyses led to nearly 129,000 circRNAs in 40 cancer types, which they compared to more than 60,000 circRNAs previously compiled in the CircBase database.
Along with circRNA expression data for more than two dozen matched tumor-normal pairs, the team's cancer-associated circRNA collection — known as MiOncoCirc — helped the researchers focus in on circRNAs associated with metastatic castration-resistant prostate cancer or aggressive neuroendocrine prostate cancer cases, as well as potential circRNA biomarkers found in urine samples from prostate cancer patients in general.
"We were able to show that these circRNAs exist in urine and that those coming from prostate cancer can be detected," co-senior and co-corresponding author Arul Chinnaiyan, director of UMich's Michigan Center for Translational Pathology, said in a statement. "Our future studies will explore these circRNAs as urine- or blood-based cancer biomarkers."
The researchers used Agilent SureSelect exon capture probes to target nearly 21,000 genes and more than 334,000 exons in cDNA from 868 clinical samples, cell lines, and pooled normal tissue samples, following by sequencing with Illumina technology and identifying candidate circRNAs with an analytical pipeline called CIRCexplorer.
When the team compared and validated this method against established Ribo-Zero sequencing and exoribonuclease RNase R-based enrichment approaches in a prostate cancer cell line, it found that the exome capture approach appeared to pick up some circRNAs missed by Ribo-Zero, while maintaining "circular-to-linear ratios" that cannot be gleaned with the RNase R method.
Along with their search for possible prostate cancer circRNA biomarkers, the researchers used MiOncoCirc and another in-house computational pipeline called CODA to look at the prevalence of so-called "read-through circRNA" (rt-circRNA) transcripts comprised of exons stemming from adjacent genes. Those analyses suggested that rt-circRNAs made up around 2.5 percent of circRNAs in each sample, on average, leading to a handful of rt-circRNAs with apparent tissue-specific expression.
Based on these and other findings, the authors argued that the open-access MiOncoCirc resource holds promise for identifying possible cancer biomarkers and teasing out more general circRNA features and functions.
"[M]iOncoCirc will serve as an important resource for scientists who wish to explore the lineage-specific and expression patterns of circRNAs in cancer as well as the intriguing mechanism of read-through splicing," they concluded. "Such studies may shed light into the function of circRNAs and help develop the use of circRNAs in diagnostic medicine."
In a related study published in Cell today, an independent international team led by investigators at the University of Toronto, Princess Margaret Cancer Center, and Ontario Institute for Cancer Research described 76,311 circRNAs identified with ultra-deep RNA sequencing on 144 sporadic intermediate-risk localized prostate cancers from the Canadian Prostate Cancer Genome Network.
There, investigators found more than 7,200 circRNAs per prostate tumor, on average, and narrowed in on 171 circRNAs with functional roles in prostate cancer cell proliferation in their follow-up loss-of-function screening analyses.
"The global burden of circRNAs was correlated with tumor aggressivity," the authors wrote, adding that the cancer transcriptome in general "harbors numerous functional entities not observable with traditional molecular strategies."