NEW YORK (GenomeWeb News) – Gauging the expression of genes coding for Dicer and Drosha — two proteins involved in processing small RNAs — can help predict ovarian cancer outcomes, new research suggests.
In a paper appearing online today in the New England Journal of Medicine, a group of researchers from across the US used quantitative RT-PCR to assess the levels of Dicer and Drosha mRNAs in more than 100 ovarian tumor samples. The team found that diminished expression of either Dicer or Drosha was associated with poor ovarian cancer outcomes. But the shortest survival times were linked to a combination of both low Dicer and low Drosha expression.
“[T]hese findings support the hypothesis that miRNAs have an underlying role in cancer progression,” senior author Anil Sood, a gynecologic oncology researcher at the University of Texas MD Anderson Cancer Center, and his co-workers wrote.
Although the precise mechanism behind the RNAi processing-survival link remains unknown, the work provides a peek into ovarian cancer biology — and underscores the potential limitations of small RNA therapeutics that rely on cancer cells’ abilities to process short hairpin RNAs, or shRNAs.
Human cells contain thousands of non-coding RNAs. And with previous research suggesting miRNA levels are sometimes altered in human cancers, including ovarian cancer, some have speculated about potential roles for miRNA regulation and RNAi-processing machinery in cancer.
For the latest study, Sood and his team focused on Dicer and Drosha, two proteins that are critical to small RNA processing. Drosha, an RNAse III enzyme, chops double-stranded RNA into short, double-stranded hairpin RNAs that are 60 to 70 nucleotides long. The RNAs are shuttled out of the nucleus to the cytoplasm where another RNAse III enzyme, Dicer, carves them into mature miRNAs that are 19 to 21 nucleotides long.
These newly formed miRNAs can curb mRNA translation by targeting specific sequences and binding the RNA-induced silencing complex or RISC. Small interfering RNA or siRNA production is similar, but doesn’t rely on Drosha processing.
Sood told GenomeWeb Daily News that he and his team have been working on developing RNA-based cancer therapeutics with nanoparticle delivery systems for several years. But prior to this study, he explained, it was unclear whether it would be better to employ short siRNAs or longer shRNAs in such therapeutics.
The researchers attempted to answer that question by examining the endogenous RNA processing machinery in ovarian cancer samples. They measured Dicer and Drosha mRNA levels in 111 invasive epithelial ovarian cancer specimens and 11 benign ovarian epithelial samples using Applied Biosystems’ TaqMan gene expression assay kit.
Of the ovarian cancer samples tested, 60 percent had lower than normal Dicer mRNA levels, just over half had reduced Drosha levels, and 39 percent had low Dicer and Drosha levels.
Decreased levels of either Dicer or Drosha were associated with significant reduced survival times, Sood said. But those with low levels of both Dicer and Drosha mRNA tended to have the worst ovarian cancer outcomes.
Those with high Dicer mRNA levels had a median survival time of 9.25 years, compared to just 2.33 years in those with low Dicer expression. Individuals with high Drosha mRNA levels had a median survival time of 7.92 years compared to 2.74 years for those with low Drosha levels. Meanwhile, those with both high Dicer and high Drosha levels had increased median survival time (more than 11 years), whereas the median survival time was only 2.66 years for those with low Dicer and low Drosha expression.
The curbed Dicer and Drosha expression in ovarian tumors did not seem to result from gene mutations, though. When the researchers sequenced Dicer (DICER1) and Drosha (RNASEN) exons and flanking sequences from 37 ovarian tumors, they detected only a few mutations, which did not correlate with altered gene expression levels.
The team confirmed their results in follow-up studies of 132 ovarian cancer patients, 129 breast cancer patients, and 91 lung cancer patients. Again, high Dicer and high Drosha mRNA levels predicted longer disease-free and overall ovarian cancer survival times. In breast and lung cancer, high Dicer, but not Drosha, mRNA levels were associated with better outcomes. The team found similar results when analyzing microarray data from two additional studies of 159 and 249 breast cancer patients.
By transfecting high or low Dicer expressing ovarian cancer cell lines with either shRNAs or siRNAs targeting a gene called galectin-3, the researchers confirmed that low Dicer expression translated into poor miRNA processing. The silencing of galectin-3 was poor in cells with low Dicer expression when shRNA was added. On the other hand, the siRNA constructs silenced the gene regardless of whether Dicer expression was high or low.
In an accompanying editorial in the same issue of NEJM, Yale University researchers Frank Slack and Joanne Weidhaas said the paper “provides evidence for a simple mechanism, based on the biologic characteristics of microRNAs, for formulating a prognosis and potentially guiding therapy in ovarian cancer.”
Lots of studies have shown that miRNA levels are mis-regulated in cancers, Slack told GenomeWeb Daily News, but just a few have started looking at Dicer and Drosha levels.
Slack noted that Dicer and Drosha levels likely won’t be used to guide treatment yet, given the limited treatments available for ovarian cancer at this point. But they may help direct treatments down the road, he said. And, Slack added, the work will likely prompt others to look for the prognostic value of these markers in other cancers.
The new paper may also have implications for those developing small RNA-based treatments, Slack said. Because small RNA processing is hampered in many of the ovarian tumor samples, treatments that require such processing likely won’t be effective. “If one is going to be treating some of these patients with therapeutic miRNA, one wouldn’t want to rely on the ability of the cell to process these RNAs,” he explained.
For his part, Slack, who holds two pending patents on using miRNAs as potential therapeutic and diagnostic cancer markers, said he’s still trying to get delivery mechanisms worked out so that RNAs go where they’re intended. Once that is sorted out, he said, the next step will be getting more sophisticated about the types of miRNAs that are used therapeutically.
At this time, Slack said, no one has quite nailed down the best way to deliver processed miRNAs. Generally, he explained, people deliver miRNAs through a virus or vector from which long RNAs are processed or as pre-RNAs. “We need to get more sophisticated about how we provide miRNA mimics,” Slack said.
Based on their work, Sood said he and his team are focusing on developing therapeutics based on siRNAs rather than shRNA fragments that require more processing. They are working on creating an siRNA that targets cancer-specific genes and is delivered via fatty nanoparticles.