Associate professor, pediatric/molecular biology and genetics, Johns Hopkins University School of Medicine
• Postdoc, Johns Hopkins University School of Medicine — 2003-2004
• MD, Johns Hopkins University School of Medicine, — 2003
• PhD, genetics, Johns Hopkins University School of Medicine — 2001
• BA, biology/genetics/development, Cornell University — 1996
This month, Johns Hopkins researcher Joshua Mendell and colleagues reported that the activation of Ras oncogenes leads to the repression of two specific microRNAs, miR-143 and miR-145, in human, mouse, and zebrafish cells, and that the restoration of the miRNAs can block pancreatic tumorigenesis.
Gene Silencing News spoke with Mendell this week about the findings.
You were focusing on Kras in the paper. Can you give a little bit of background on it?
My lab has devoted significant effort in the past to try to understand how microRNAs participate in cancer by investigating how they participate in important pathways that are known to be critical for cancer. For example, we have previously identified microRNAs that function in the MYC oncogenic and p53 tumor suppressor pathways.
In the past, this strategy has proven to be a very good way to identify microRNAs whose gain or loss of function directly contributes to tumorigenesis. Our interest in the regulation of microRNAs by Ras is a natural extension of this strategy.
The Ras family of oncogenes is among the most commonly activated … in human cancer. [Members] are typically activated by specific mutations in cancer, as opposed to other oncogenes like MYC, which is often … involved in translocations or amplifications that cause it to be over-expressed. In the case of Ras, the mutations that occur in cancer cells cause it to be locked in an active configuration.
Ras functions as a signaling intermediate, a protein that helps receive signals from outside the cell, such as growth-stimulatory signals, and propagates those signals to downstream pathways. Normally, Ras receives a signal from outside the cell and is only transiently activated. As a result of the mutations that occur in cancer, it's locked in the active conformation such that Ras constantly stimulates downstream signaling pathways even without a signal coming from outside the cell.
When Ras is activated in cancer, it engages a number of downstream pathways that are important, including the MAP kinase pathway, which is a series of successive phosphorylation reactions that help promote cell survival and proliferation.
We were interested in knowing if, in addition to altering canonical signaling pathways, the activity of mutant Ras also changes the microRNA expression in cancer cells to promote tumorigenesis.
Mutations in Ras oncogenes are common in many types of tumors. In particular, mutations in Kras are almost ubiquitous in pancreatic adenocarcinoma, a common and devastating form of pancreatic cancer with few therapeutic options. For this reason, we chose to study the role of microRNAs in the Ras pathway in the setting of pancreatic cancer.
How did you go about exploring that?
We used a number of different model systems to ask how Ras signaling affects microRNA expression. The first thing we did was a simple experiment in which we took a non-cancerous pancreatic duct cell line — pancreatic adenocarcinoma arises from the pancreatic ducts — and added mutant Kras to it. We then used a custom microarray we have in our lab that can monitor the expression of several hundred microRNAs simultaneously to see if any microRNA levels changed.
We found that a very small number were increased and a very small number were decreased. We focused our efforts in this paper on two microRNAs that were decreased: miR-143 and miR-145. We were interested in those for several reasons. First of all, they are in a cluster, which means they are co-transcribed as part of a single unit … [and] we saw that they were coordinately down-regulated by Kras.
Second of all, there is actually a fairly extensive body of literature about these microRNAs specifically showing that they have anti-tumorigenic properties in different kinds of cancer. No one has yet described their role in pancreatic cancer, but they've been implicated in colon cancer and lymphoma, for example.
That led us to believe there might be a functional significance of Kras down-regulating these anti-tumorigenic microRNAs.
We next looked in a panel of cell lines derived from pancreatic cancer that nearly all have activating Kras mutations, and we found that both miR-143 and miR-145 are expressed at very low levels. That, in and of itself, does not show that downregulation of these miRNAs in these lines is due to activated Kras per se, but it's an important correlation.
So then you looked in vivo?
We did that in a couple of ways.
We utilized multiple models in which we could control Kras activity to confirm our earlier findings with the cell lines. The first was a mouse model that was described by David Tuveson in which a mutant allele of Kras is expressed in the pancreas.
We took pancreatic tissue from these mice and compared [them] to wild-type litter mates. We found that in pancreas expressing mutant Kras, we could detect the repression of the microRNAs, a very encouraging result.
We also took advantage of a [model] … that our collaborator Steve Leach [at Johns Hopkins] developed … [in which] mutant Kras is expressed in the pancreas of zebrafish. In the tumors that arise in these zebrafish, the microRNAs are expressed at very low levels compared with normal pancreas.
All of those studies were done in the setting of the pancreas, so lastly we studied their expression in mouse fibroblasts, a commonly used model system for Ras-mediated transformation. When we expressed mutant Kras in these mouse fibroblasts, we again saw that miR-143 and miR-145 were repressed. So in multiple species, in a variety of cell lineages, we consistently saw that when we activate Ras, the microRNAs are repressed.
After documenting that Ras signaling regulates miR-143/145 in this way, we next asked whether this results in any phenotypic change in cancer cells. Do tumor cells that have activated Kras even care if miR-143/145 levels are reduced? To answer this, we restored miR-143/145 expression to physiologic levels in the presence of activated Kras in pancreatic cancer cells, put those cells into mice, and found that they can no longer form tumors.
This results told us that the repression of these microRNAs is an important component of the oncogenic program that is activated by mutant Kras in cancer cells.
What's the current state of understanding as to the specific function of these microRNAs?
That's an ongoing question that we're still working on, but we're beginning to get some initial answers.
Another part of the story that will be relevant to [that question] is that we showed when Ras is activated, it activates another protein called Ras-responsive element-binding protein 1, or RREB1, which is a transcription factor that binds to the promoter of miR-143 and -145 and represses their transcription.
Interestingly, when we started looking at the targets of the microRNAs to try to understand their functions, we saw that one of the targets that had been previously reported is Kras itself. Another target is RREB1. So we proposed that when Kras is activated, repression of miR-143/145 through RREB1 further enhances Kras activity.
Since microRNAs regulate many targets we certainly don't believe that regulation of Kras and RREB1 is the whole story. We think that there are likely other important targets that help the microRNAs perform their anti-tumorigenic functions. Identification of such critical targets is a major priority for future work.