It means thinking a little bit more creatively about the screen assay design. I think that can be validated by identifying inhibitors that work in novel ways and with much greater selectivity than ATP site-directed inhibitors.
Fox Chase Team Uses Murine Cell Assay to Find Novel Kinase Inhibitor
Investigators at the Fox Chase Cancer Center in Philadelphia this week announced that they have identified a molecule that can shut down the p21-activated kinase 1, or Pak1, enzyme before it becomes active.
The researchers evaluated the ability of the molecule, called IPA-3, to inhibit platelet-derived growth factor-stimulated Pak activation in mouse embryonic fibroblasts. They found that both basal and PDGF-stimulated Pak activity was inhibited by 30 µM of IPA-3, as assessed by an in-gel kinase assay.
The researchers developed a biochemical screen for allosteric inhibitors of targeting Pak1 activation that identified IPA-3. They found that preactivated Pak1 was resistant to IPA-3.
Writing in the April 21 issue of Chemistry & Biology, the investigators also found that IPA-3 inhibits the activation of related Pak isoforms regulated by autoinhibition. However, it did not inhibit more distantly related Paks, nor did it inhibit the more than 200 other kinases tested.
Jeffrey Peterson, an associate member of the Fox Chase Cancer Center and corresponding author on the paper, spoke to CBA News this week about his team’s work and urged scientists to think ‘outside the active site box’ in cancer drug discovery.
Peterson said that he and his team plan to follow up this work with cell-based assays using breast cancer and neurofibromatosis cell lines.
Can you give me some background on this work?
My lab is interested in the Rho family of GTP-binding proteins and the signaling pathways regulated by the Rho proteins. One of the important effectors of the Rho proteins is the p21-activated kinase 1, or Pak1 kinase.
We are particularly interested in Pak because of its association with human cancer, particularly breast cancer. For example, Pak kinase is frequently hyperactivated in breast cancer, and molecular mechanisms have been worked out that link hyperactive Pak to increased proliferation, survival, and migration of breast cancer cells.
Our primary objective is to develop a small-molecule inhibitor targeting Pak — first, as a tool to understand how Pak functions in a normal setting, and to identify what pathways are relevant to cancer, and second, as a proof of principle to validate inhibitors of Pak as a therapeutic strategy against cancer.
Another objective we had was to try to inhibit this enzyme in a somewhat unusual way. Protein kinases are perhaps the most important therapeutic target in oncology. However, many of the inhibitors that currently target kinases all do so by the same mechanism. They bind in the ATP-binding pocket and displace ATP, which is the substrate of the phosphorylation reaction.
This is a very well-validated approach to targeting kinases, but there is one major drawback. The ATP-binding pocket in Pak kinase is very conserved to the ATP-binding pocket in many protein kinases.
Empirically what has been found is that most kinase inhibitors that target this ATP-binding pocket also inhibit additional kinases. This can result in off-target toxic effects of the drug.
Pak is regulated by autoinhibition — it inhibits itself in the normal state. We wondered if we could exploit this autoinhibitory regulation to identify an inhibitor of Pak that would work not by targeting the ATP binding pocket, but would perhaps stabilize or “clamp down” this autoinhibited state of the enzyme.
The autoregulatory mechanism is unique to the Pak kinases. So we thought again, by targeting this unusual regulatory mechanism, we could identify an inhibitor with much greater kinase selectivity compared to those targeting the ATP-binding pocket.
How exactly did the assay that you developed work?
Let me set it up by telling you how most pharmaceutical companies develop kinase inhibitors. They just take the kinase domain itself and remove all of the extraneous bits, so that the only inhibitors you can identify are those that target the active site.
What is unusual about our screen is that we began with full-length Pak proteins, just as they are found in human cells. These full-length proteins are normally, as I said, autoinhibited. An additional component of the reaction is the protein called Cdc42 that normally activates Pak.
What we are trying to do is recreate all of the biologically relevant activation steps in Pak kinase, instead of the traditional paradigm, which is strip away all of these regulatory things and just look for things that bind in the active site.
So we are trying to reproduce the normal biological regulation, so that we can exploit it in the mechanism of action of our drug candidate.
This screen did not use cell lines. It was a purified protein screen. The only components in the reactions were pure Pak, pure Cdc42, and a protein substrate that we use that becomes phosphorylated by Pak when it is active.
Did you do cell-based studies?
Once we identified the inhibitor that is described in the manuscript, IPA-3, and characterized how it inhibited Pak kinase activity, we wanted to know if it worked in cells.
The final figure in our paper documented that not only in a test tube can we block the activation of Pak, but we can also block it in the context of live mammalian cells. As a test case for our cell-based studies, we used a system that is well-known, whereby you activate Pak kinase with platelet-derived growth factor.
We treated the cells with IPA-3 or not, and then stimulated all the cells with PDGF, and asked if Pak kinase became active or not. Then, as we would predict, the IPA-3 was able to block Pak activation stimulated by PDGF.
What do you see as the next step in this work?
There are, I think, two major implications, and each of those have different future directions.
The first is directly related to the inhibition of Pak. Pak hyperactivity, as I said, has been linked to cancer and a condition called neurofibromatosis. In both of these cases, you have hyperactive Pak enzymes. We are now using this compound to determine whether inhibition of Pak is a viable therapeutic strategy for treating these diseases.
How are you doing that?
By taking, for example, cell lines derived from a variety of cancers that have hyperactive Pak, treating them with the inhibitor, and determining if it slows their proliferation, for example, and other aspects of their behavior.
Are you planning to publish this work?
Absolutely. We do not yet know when, however, because it is still preliminary.
Do you see drugs that work via the mechanism of action described here as an adjuvant therapy to currently used kinase inhibitors, or will they replace currently used kinase inhibitors?
At this point, they would play more of an adjuvant role. We still need to validate inhibition of Pak as a therapeutic strategy.
The Pak kinase is not targeted by currently used therapies. What we first need to do is establish that inhibition of Pak does indeed offer a therapeutic benefit.
What I did want to mention is that there are two main implications: this work provides a mechanism for testing whether Pak is a valid therapeutic target; this work also demonstrates that there are many different ways of inhibiting kinases beyond the typical ATP kinase inhibitor traditionally found in pharmaceutical company kinase screens.
I would like to encourage drug developers to broaden their screens in order to potentially capture these novel opportunities for inhibiting kinases. In other words, I would encourage investigators to think outside of the ‘active site box.’
Can you suggest ways that they might do that?
Typically now, people start out with just the isolated kinase domain in a constitutively active form, and try to inhibit it. That is throwing away a lot of the normal biological regulation that controls the enzyme.
Making the screen slightly more complex to incorporate the native regulation of enzyme activity offers additional opportunities for drugs to inhibit the activity of the enzyme, beyond the active site.