GTPaseActivatingActivity?
WeDon'tNeedNoStinkingGTPaseActivatingActivityVille
As we discussed in earlier posts, I spent some time many years ago working on an oncogene called 'RAS' under experimental conditions in the lab.
At the time, I was lucky enough to make a small scientific contribution to a massive experimental literature that accumulated late in the last century which demonstrated that the oncoproteins coded by RAS oncogenes cause cells to divide ceaselessly in certain tissue types under certain conditions. Around the same time, clinical researchers determined that RAS oncogenes drive the formation of a bunch of different tumour types in real, actual people, one of which is the aggressive, and ultimately deadly, pancreatic ductal adenocarcinoma, also known by the acronym 'PDAC'.
So, given all this, the great hope at the turn of the century was that we would be able to quickly develop a drug that blocks the ability of RAS oncoproteins to cause cells to divide.
Easy-peasy, right?
Wrong.
****
Why has it been so hard to develop a clinically useful RAS blocker?
First, a little bit of a cell biological digressive explainer (feel free to skip this section if you like)...
It turns out that there are normal RAS genes in every single cell of our bodies. These genes code for normal RAS proteins that respond to transient external growth signals and magically turn them into signals inside cells that turn on cell division. This is given a fancy name called 'signal transduction'.
Specifically, when a growth signal, say a hormone, binds to the outer surface of a normal cell RAS proteins just inside the cell grab onto a molecule called GTP, which is a funky little molecule called guanosine with three phosphates a hanging off it (this is the small grey bit shown in the image above).
Binding GTP switches RAS proteins into the 'ON' position (the purple blob in the image above). When they are in the 'ON' position, normal RAS proteins then bind to a bunch of 'effector' proteins in the cell's cytoplasm. This binding of effector proteins next flips a set of cytoplasmic relay switches, a subset of which moves into the nucleus where they turn DNA synthesis to double up all the cell's chromosomes. This drives cell division in a process called 'mitosis' that just about everybody reading this likely learned about in grade 5 or 6.
Therefore, the growth signals that arrive at the outer cell surface are sometimes called mitogens and normal RAS proteins are often called transducers of the mitogenic signal that drives cell division. Already, I figure you are probably getting a good idea of what might happen if this mitogenic signal transducing activity were to be enhanced by, say, oncogenic mutations in the RAS gene - but first, let's finish talking about what happens with normal RAS proteins.
When the growth signal/homone is no longer present at the cell surface, normal RAS proteins activate their intrinsic enzymatic activity, called a GTPase, that rips one of the phosphates off the bound GTP. This gives rise to a guanosine with just two phosphates, which is also known as a 'd'iphosphate. Thus, the bound molecule is known as GDP. The key here is that normal RAS proteins bound to GDP no longer bind the effector proteins that drive DNA synthesis and cell division stops. Put another way, normal RAS proteins bound to GDP move into the OFF position. As a result, when things are working properly normal cells only divide when they receive appropriate external mitogenic signals in healthy tissues.
In contrast, oncogenic mutations in RAS genes cause amino acid changes in the resulting RAS oncoproteins that destroy the protein's GTPase activity. This means that RAS oncoproteins can't rip off the third phosphate stuck to guanosine to form GDP. Therefore, oncogenic RAS proteins are always bound to GTP and stuck in the ON position - that's what is really depicted in the diagram above.
The take home message here is that cells with oncogenic RAS mutations don't give a hoot-in-heckfire about whether they see appropriate mitogenic signals. Instead, because they are always have GTP stuck to them (RAS ON) they are constantly binding to effector proteins the DNA synthesis/mitosis switches that drive cell division. This is especially true in cells that make up the ducts that drain little tiny glands in the pancreas that secrete the digestive juices into the small intestine. Thus, oncogenic RAS mutations drive the growth of these pancreatic ductal cells which results in the formation of the aggressive, hard to treat lesions known as the pancreatic ductal adenocarcinoma (PDAC) tumours mentioned above ('adeno' stands for gland and 'carcinoma' stands for the lining tissues that give rise to ducts and glands in the pancreas - thus, the tumour name).
OK.
Cell biology digression over...
****
Most of the new anti-cancer drugs that aren't antibody-based are small molecules that can slide into cells where they bind to an oncoprotein and block its action by messing up the protein's intrinsic enzymatic activity. Thus, the name of these drugs often end in 'ib', short for 'inhibitor'.
One of the first of these drugs was imatinib, or Gleevac, which was developed thirty years ago to treat a chronic form of leukemia. Imatinib binds to and blocks the enzymatic activity of an oncoprotein called Bcr-Abl. In this case the oncoprotein is an enzyme called a kinase that likes to stick phosphates on Bcr-Abl-specific effector proteins that also help drive cell division.
Unfortunately, taking this approach to develop a drug to inhibit RAS oncoproteins jammed in the ON position has turned out to be very, very hard for at least two reasons.
First, RAS oncoproteins have a really smooth outer surface that makes it hard to find small molecules that can bind and stick to them on their own.
Second, in a weird twist of fate, you actually have to activate a RAS oncoprotein's enzymatic activity (i.e. the GTPase activity described in the section above) to switch it to the OFF position. Finding a small molecule that can do this kind of thing is really tough because it is way easier to find a compound that binds to and jams up/inhibits an enzyme's activity (i.e. what imatinib does to Bcr-Abl) than it is to find a compound that binds to/activates an already messed up enzymatic activity (i.e. the intrinsic GTPase activity that is abolished up by mutations in oncogenic RAS).
By analogy, if you think of an oncoprotein as a spark plug that drives the cell division engine, imatinib binds to and sits in the gab between the electrodes at the top of the Bcr-Abl to gum up and block it from firing. Conversely, in the case of RAS, you would have to develop a drug that sits in the gap and causes the busted plug to spark . Obviously, fixing busted enzymes/spark plugs with a small molecule is way more difficult just gumming up/messing up the firing of the plug as happens with imatinib.
As a result, it turns out that finding a small molecule that can first bind to the slippery surface of the RAS oncoprotein and then fix its messed up spark plug firing turned out to be almost impossible.
Thus, starting as far back as the early 2000's, a lot of very smart people began to think that RAS oncoproteins were undruggable.
Luckily, a few even smarter people kept trying, using different approaches to the one described above. One of these was developed by an academic research scientist named Gregory Verdine at Harvard who had been working on small molecule molecular glues that cause proteins to stick together. One of these glues, which was originally isolated from bacteria, acts as a molecular chewing gum that sticks a big, honking protein called 'Cyclophilin A' to a bunch of other proteins inside mammalian cells. That kind of work, which at the beginning seems like it might not be immediately applicable to any specific clinical problem, is the type pure, fundamental discovery-based research that is almost always publicly funded.
Anyway, after the original discovery work had been done, Verdine came up with the idea that his molecular glue could be chemically modified such that it could stick cyclophilin A specifically to RAS oncoproteins. To start working on this 'translational' research problem, Verdine and his colleagues raised venture capital money and formed a biotech company called Warp Drive Bio.
Why move away from public funding to do this kind of translational work in attempt to develop a compound that could turn out to be clinically useful?
Well.
You could argue that it is done to try and hit the financial biotech jackpot. And, there is a some of that going on. In fact, most public institutions want their researchers to form such spin-offs because they usually own the intellectual property rights.
However, there are also a couple of other major factors that work against researchers who try to translate their fundamental research findings into something clinically useful using public funding alone. First, this translational stuff often costs more money than most publicly funded research grants can provide. Second, it is grinding, step-by-step work that is often viewed as incremental by public research grant adjudicators who are trained to instead reward novelty. Thus, even getting public funding for this type of work can sometimes become very difficult
Regardless, Warp Drive Bio never quite got there translationally and a number of years after its founding the company was acquired by a second company called Revolution Medicines that soon set their team of excellent molecular modellers and synthetic chemists to work modifying Verdine's molecular glue. In doing so, the RevMed folks developed a slew of new lead compounds, a few of which did the trick and stuck the Cyclophilin A protein to RAS oncoproteins.
Miraculously, these new compounds worked. One of them, which RevMed calls 'daraxonrasib', is represented as the little bit of orange chewing gum that is labelled RAS(ON) inhibitor in the image above
The fact that the inhibitor worked was somewhat dogma busting because it has no effect on the RAS oncoprotein's messed-up GTPase activity (i.e. it does not fix the busted enzymatic sparkplug). Instead, it turns out that the Cyclophilin A molecule (i.e. the turquise blob in the image above) that is stuck to the purple RAS(ON) protein by the chewing gum inhibitor is so big and bulky that it prevents the effector proteins from binding to the complex.
And that's the key here - when the effector proteins can't bind there is no switch flipping resulting in DNA synthesis. As a result, the RAS(ON) oncoproteins can no longer drive tumor cell division.
All of which is great translational science but the real proof is in the clinical trial pudding.
And that pudding was cooked in a recent phase three clinical trial, the results of which were published in the New England Journal of Medicine a few weeks ago. The following is straight from the paper's abstract:
... A total of 500 patients, including 91.8% with (oncogenic) RAS G12 mutations, were randomly assigned to receive daraxonrasib (248 patients) or (standard) chemotherapy (252 patients). The median overall survival in the RASG12 population was 13.2 months with daraxonrasib and 6.6 months with chemotherapy...
Now.
That doubling of survival in RAS oncogene-driven PDAC patients, which amounts to an improvement of over just six months, may not seem like much, but...
Don't forget that this improvement is for the worst-of-the-worst of all the RAS oncogene-driven tumours where no significant improvement in chemotherapeutic treatment has been seen for years.
Imagine that!
______
The image at the top of the post is a schematic/cartoon from the Revolution Medicines website...A paper describing the translational generation and characterization of a number of their RAS(ON) inhibitor compounds, including daraxonrasib is here.
.