By Rich Soll, Senior Advisor, Strategic Initiatives, WuXi AppTec (@richsollwx)
p53, known as the guardian of the genome, occupies a particularly important role in cancer biology as a tumor suppressor by regulating or affecting progression through cell cycle, controlling apoptosis and ensuring genomic stability. It is frequently mutated (~50% of cancers), but even in non-mutated forms of p53+ cancer, the actions of p53 become limiting to due expression of proteins such as MDM-2 (hDM-2 in humans), which cause p53 ubiquitination, leading to p53 degradation.
For years, a major focus of researchers studying many hematological malignancies has been to suppress activity with antagonists of MDM-2. At least 14 companies have been pursuing this approach according to Clarivate Analytics. This important class of drugs bind to and inhibit MDM-2 activity. Alternatively, ablation of the kinase CK1α has also been associated with tumor inhibition but insights into how to capitalize on these seemingly independent effects only recently have been realized.
Professor Yinon Ben-Neriah of the Lautenberg Center of Immunology and Cancer Research at Hebrew University in Jerusalem, Israel is a pioneer in hematological disorders. He cloned cAbl, characterized the leukemic oncoprotein Bcr-Abl, shared the identification of the multidrug resistance protein Mdr1, and identified the first effective Bcr-Abl small molecule inhibitor with Alex Levitzki – which served as the platform for the discovery of Gleevec.
He has been studying CK1α for more than 20 years adding, “Years ago we found that CK1α was a necessary component of the Wnt signaling pathway.” His early work showed that CK1α was found to phosphorylate β-catenin; working together with GSK3, the coordinated events leads to a bis-phosphorylation cascade, which in turn essentially creates a ubiquitin docking site for β-catenin. β-Catenin then becomes degraded by the proteosome.
“Both GSK3 and CK1α are absolutely necessary and this event is the major signaling in Wnt pathway,” notes Ben-Neriah. “The function of this event is to keep Wnt actually low and moderate so as to not have too much β-catenin to be stabilized as this would be turning on the Wnt pathway.” Through ablation of CK1α, β-catenin is stabilized and the Wnt pathway becomes hyperactive and that may lead to tumorigenesis since Wnt pathway plays a key role in tumorigenesis.
Ten years later, another role for CK1α in p53 activation was found. The mechanism is still being elucidated because it is not a typical pathway for p53 activation. Ben-Neriah’s team found that ablation of CK1α leads to activation of DNA damage response leading to p53 activation. While activation of DNA damage response occurs with many chemotherapies, such a doxorubicin, the difference with CK1α inhibition is that there is no detectable DNA damage, thus circumventing the adverse effects of chemotherapies such as tissue damage and genotoxicity. Thus, there is the potential to realize benefits without paying the toll.
CK1α ablation methods were used to activate p53 many years ago, but Ben-Neriah asked ‘what would happen if CK1α inhibitors are used instead?’ He came upon a few hits that were not great so he then formed a partnership with medicinal chemistry team at WuXi, who went through iterations of design and testing, always seeking improvements. “We went through 20-30 cycles during this optimization from hit to lead. Now we have compounds that are working on the order of a 1000 fold better than the initial hits,” stated Ben-Neriah. “We had a particularly flexible chemistry team and designers which made this partnership with our research institute special. Moreover, essentially through this enabling partnership, we produced insightful, breakthrough results.”
It was during the evaluation of these compounds that Ben-Neriah noticed something very interesting. Specifically, the compounds were working better than gene KO of CK1α. Ben-Neriah continued: “We saw the extent of p53 activation/stabilization was well beyond what we could achieve with the CK1α KO. That raised the question regarding what else might be going on.”
One of the first observations was that these compounds were obliterating the transcription of MDM-2. What is the significance of this? Ben-Neriah explains: “We know that knocking out CK1α leads to p53 stabilization. However, that inhibition can only achieve a certain level of performance because MDM-2 still exerts its inhibitory action on p53. Thus, there is a limit to how much one can stabilize p53 because of this negative feedback loop by MDM-2 on p53 activity.”
Moreover, these compounds inhibited MDM-2 independently of their effects of p53 stabilization through CK1α inhibition. It is this combined effects of MDM-2 inhibition and CK1α inhibition that led to profound p53 activation.
Ben-Neriah investigated further. It turned out that these compounds also inhibited the expression of Myc, a target of Wnt pathway activation. A paradox was created because Myc expression should have increased upon Wnt activation through CK1α inhibition. That the compounds inhibited both MDM-2 levels and Myc levels suggested that the compounds impacted transcriptional activation. “In fact, the compounds inhibited other important proteins involved in apoptosis such as MCL-1, the most important anti-apoptotic factor in hematopoietic cells,” observed Ben-Neriah.
Ben-Neriah sought out a common denominator amongst the CK1α-independent observations. His attention was drawn to the work of Rick Young, Professor of Biology at the Whitehead Institute at the Massachusetts Institute of Technology, concerning super enhancers such as Myc and MCL-1, which are under control of the kinases CDK7 and CDK9. It turns out that these compounds are highly selective towards CDK7 and CDK9 within the CDK isoforms. This was striking because CK1α and CDK7/9 do not belong to the same family.
These compounds are potent and specific towards these kinases, with Kd’s on the order of 5 nM. To gain additional insights into the molecular actions of these compounds, attention was turned to structural biology. The Crelux division of WuXi AppTec was engaged for this purpose. Ben-Neriah noted, “The crystallography was particularly difficult in the case of CK1α yet the team persevered and solved this problem.” The binding modes of this series of molecules to these structurally distinct family of kinases revealed that the compounds bind at the ATP binding site and overall, share similar H-bonding interactions with several residues of these proteins.
“We consider ourselves to be lucky because we now have molecules that not only enhanced p53-mediated effects, but also broadly co-targeted super enhancers responsible for apoptosis protection and cancer progression in leukemia,” acknowledged Ben-Neriah. The effects of the compounds on the bone marrow in an AML leukemic mouse model (MLL-AF9-induced AML), representative of a poor prognosis of one of the most aggressive human leukemias, was studied and compared to normal hematopoietic cells.
In this leukemia there are 120 super enhancers that are produced which are not present in normal bone marrow. Upon drug treatment, more than 90 super enhancers – such as the oncogenes Myc, Myb, Runx2 and Meis1 – are disrupted and severely depressed, indeed abolished, by the test compounds. Thus, there is exquisite sensitivity of these super enhancers towards the co-targeting compounds.
What then is the advantage of co-targeting? In more than 90% of human leukemia p53 is not mutated, but the actions of p53 is limiting because inhibitors of p53, such as Mdm-2, are overexpressed, thus inhibiting the effects of p53; there would be a partial effect using only a CK1α inhibitor. To see maximal effects of p53, one needs to maximally activate p53 and reduce/abolish super enhancers.
Ben-Neriah concludes: “These leukemic cells, like many cancers, are ‘transcriptionally addicted’ which means they are very dependent on the super enhancers, unlike normal cells. So in our case, we have a perfect match where we crush the super enhancer addiction while promoting the protective effect p53 action. That is the secret of these amazing compounds.”
The details of this exploration were revealed in a Cell publication entitled ‘Small Molecules Co-targeting CK1α and the Transcriptional Kinases CDK7/9 Control AML in Preclinical Models’, Cell, 2018, 175, 1-15.
Since the publication of the Cell Paper, Ben-Neriah has received multiple letters and emails from patients who have since run out of options. Perhaps now, we may be approaching the possibility that a cure is in sight.
These compounds have been licensed to the US biotech BioTheryX. Phase 1 clinical studies on the lead compound should begin shortly as the IND on the lead candidate has been accepted by the FDA. If patients can tolerate the drug, Ben-Neriah feels that this could be a great drug for AML and potentially other p53-positive cancers.