MiR-548d may play a super-regulatory role in pancreatic cancer, affecting multiple components of the signaling pathways associated with the disease, a recent study by Hannover Medical School researchers has reported.
Using a bioinformatics-based approach, the group analyzed potential miRNA-binding sites in pancreatic cancer-associated pathways, calculating the key miRNAs in these regulatory networks and narrowing the set down to those with the most central role.
The researchers picked miR-548d-3p as their lead candidate and validated its possibly superordinated function in a cell culture model, which showed that over-expression of the miRNA had a tumor-suppressive effect, and may also sensitize resistant pancreatic cancer cells to treatment with the drug gemcitabine. Their study appeared last month ahead of print online in the Journal Pancreas.
Holger Heyn, one of the study authors, told Gene Silencing News this week that the research, though early, suggests miR-548d-3p as a potential prognostic and therapeutic target, and also supports the utility of the group's bioinformatic approach to target miRNAs with potentially superordinated function in cancer regulatory pathways, which he said was a novel strategy.
Most microRNA studies "begin with expression analysis of cancer or disease tissue [to] identify differentially methylated or expressed miRNAs and then look for the target and then the phenotype in the end," he said.
The team's hypothesis "is that miRNAs… are not targeting single molecules, but function more as superordinated molecules to regulate whole pathways and whole tissue types," he said. "We wanted to try something new that would reflect that, more coming from the bioinformatic side, not from the experimental."
This approach could be particularly beneficial for the development of miRNA-based treatments for cancer, according to Heyn. "There are some miRNA treatments already [being developed] in lung and liver cancer. Some work very well. But you might improve the efficiency of these treatments by targeting a miRNA that has more global function related to the disease."
He said that his group hoped that publishing its early pancreatic cancer results would help spread the bioinformatics-led approach among other researchers.
In their study, the team "used two different databases to compare [cancer-associated] pathways with microRNA targets," Heyn said.
First the researchers used the Kyoto Encyclopedia of Genes and Genomes database to identify the key pathway components and their interactions. They then used the DIANA-mirPath software package developed by the Alexander Fleming Biomedical Sciences Research Center in Greece to compare the KEGG data with miRNA targets predicted using the Whitehead Institute's TargetScan.
The DIANA system provides a score that represents the "probability of one microRNA being involved in certain pathways," the group wrote in its report. Applying this analysis suggested miR-548d-3p as the molecule with this superordinated potential. "We compared miRNA targets and pathways in cancer, and ended up, within pancreatic cancer, with only one high-scoring miRNA," Heyn said.
The analysis predicted miR-548d-3p to bind 14 individual targets in the analyzed pathways, the authors wrote. Specifically, miR-548d-3b was linked to antiapoptotic cascades, molecules in the proliferation-promoting KRAS pathway, several pro-angiogenic factors, and a DNA repair factor (RAD51), according to their analysis.
The group also found that a second molecule, miRNA-548d-5p, was encoded on the passenger strand of the miR-548d precursor and targeted additional pancreatic cancer-associated pathways, which they wrote suggested an "additive effect" of the two miRNAs encoded in one "transcriptional unit."
To validate these findings, the group set up a pancreatic cell culture model that transiently over-expressed both miR-548d-3p and miR-548d-5p. In the model, the over-expression reduced cell viability by 17 percent as compared to controls. The group also found that adding the drug gemcitabine further reduced viability by a further 30 percent when combined with miR-548d-3p and 25 percent when added to miR-548d-5p.
"From the targets predicted by the database, the miRNA was supposed to be more of a tumor suppressor," Heyn said. "And when we over-expressed the miRNA and found that it up-regulated apoptosis and the cell lines were more sensitive to [the] drug, gemcitabine."
"In the end it all fit together," he said. "So this gives us confidence that the bioinformatic approach is truly valuable."
Heyn said that because the miRNA his team identified targets anti-apoptotic and cell survival pathways as well as proliferation-promoting genes, the group believes that there is the possibility that down-regulation of miR-548d could be associated with poorer survival in pancreatic cancer, and thus could serve as a prognostic marker. "It has to be proven, but the idea is that you will find miRNA that will give you an idea of the prognosis of the patient," Heyn said.
The potential for developing new treatment is also suggested, though only hypothetically, in the group's results. Heyn said one possibility would involve treating directly, by mimicking the tumor suppressive action of miR-548d. Additionally, he cited a February study in PNAS by researchers at Bellvitge Institute for Biomedical Research in Barcelona suggesting that a small antibiotic molecule called enoxacin can "boost general miRNA processing," and inhibit the growth of a broad spectrum of cancer cells.
Enoxacin was originally identified by Emory University researchers as having the ability to enhance siRNA-mediated mRNA degradation and endogenous microRNA biogenesis. (GSN 7/24/2008)
"In our cases, this might be of high impact as well," Heyn said, "because in the paper they showed that they could treat almost all kinds of cancer, and that [the approach] increases expression of tumor-suppressing miRNAs, overruling the effect of the oncogenic ones."
Heyn has since left the Hannover team to join the Bellvitge Institute, but he said that his former group had begun expanding on the study before he left, and is likely continuing that research. The first follow-on project has been to examine whether miR-548d is down-regulated in pancreatic cancer cells from primary tissue, to "support the idea that our microRNA is a tumor suppressor," he said.
The group also analyzed miRNAs that could be superordinated in other cancer types, finding several, according to Heyn, though they decided to confine their initial publication to pancreatic cancer.
Before he left the team, Heyn said that they had also connected with the developers of the DIANA software, who expressed some interest in making some small changes in the software so other groups might more easily follow the Hannover researcher's approach.
Have topics you'd like to see covered in Gene Silencing News? Contact the editor at dmacron [at] genomeweb [.] com.