Targeting Mutated p53: Naivete and Enthusiasm to Attempt the Impossible

Abstract

Tumor suppressor TP53 is an important gene in human cancer because it is mutated in the majority of tumors, leading to loss-of-function or gain-of-function phenotypes. Mutated TP53 acts like an oncogene, driving cancer progression and causing poor patient outcomes. The role of mutated p53 in cancer has been known for over three decades, yet there is no FDA-approved drug to address the problem. This brief historical perspective highlights some of the insightful advances as well as challenges in therapeutic targeting of p53, especially the mutated forms. The article focuses on a functional p53 pathway restoration approach to drug discovery that years ago was not mainstream, encouraged by anyone, taught in textbooks, or embraced by medicinal chemists. With some knowledge, a clinician scientist's interest, and motivation, the author pursued a unique line of investigation leading to insights for functional bypass of TP53 mutations in human cancer. Like mutated Ras proteins, mutant p53 is fundamentally important as a therapeutic target in cancer and probably deserves a "p53 initiative" like the NCI's "Ras initiative.” There is a link between naivete and enthusiasm for pursuing difficult problems, but important solutions are discovered through hard work and persistence. Hopefully, some benefit comes to patients with cancer from such drug discovery and development efforts.

Years ago, as a fully trained medical oncologist and physician-scientist, and as my lab was getting established so I could do some higher risk drug discovery research, I decided to pursue what I thought was a novel approach to targeting tumors with mutated p53.

Wild-Type p53 and MDM2

There was plenty of effort 20 years ago to activate wild-type p53 in cancer with chemotherapeutic agents or with what at the time was a new class of MDM2-specific inhibitors developed at Roche in Nutley, New Jersey by Lubo Vassilev and colleagues. MDM2 is an ubiquitin E3 ligase that targets p53 for derogation. The MDM2 therapeutic targeting field has grown much so that several companies in 2022 have agents in the clinic targeting MDM2 or MDM2 and MDM4 (MDMX). Our recent contributions with regard to this area have been through an interest in the phenomenon of hyperprogression experienced by a minority of patients after immune checkpoint blockade, where MDM2 was implicated. Razelle Kurzrock and colleagues recognized the association between MDM2 amplification and hyperprogression. In preclinical studies, we demonstrated that MDM2 blockade stimulates T-cell killing of MDM2-amplified tumor cells and that there is a further boost to tumor killing with combining with anti–PD-1 therapy. We developed a humanized mouse model to investigate hyperprogression although the model is not driven by MDM2 and we are still refining its various applications and unraveling the mechanisms. We recognize MDM4 as a bona fide therapeutic target for which more effort needs to be dedicated to address unmet needs in the clinic. Much more remains to be done in treating cancer through these often-amplified targets including to improve the efficacy of immunotherapy.

Mutant p53 as a Very Important Therapeutic Target in Cancer

But, back to mutant p53 because it has been important to remain focused on mutant p53 as a very important therapeutic target that has been thought of as “undruggable.” Mutant p53, discovered as a tumor antigen in 1979, has always been attractive as a target and recent progress in immune targeting of p53 by Shibin Zhou and Bert Vogelstein represents exciting progress (1). Arnold Levine recently wrote about strategies for restoration to restore p53 suppression of tumors and some of the approaches described are already in the clinic (2). By contrast, as described below, the approach we have pursued is a functional restoration of downstream effector genes that mediate tumor suppression through p53 bypass pathways. While most in the field are well aware of the connections between p53 and viral oncoproteins, it is important to remember that mutant p53 was discovered as a target for anti-sera raised against chemically induced sarcomas (3). There are other important approaches, and many of them including the immune targeting approach, are focused on specific hotspot p53 mutations. In the era of precision oncology this is very reasonable and if successful these strategies will likely help patients.

Inspiring Early Work That Suggested Feasibility and a Path Forward

Twenty years ago, I was inspired by two papers in the literature that suggested there was a scientific rationale and perhaps hope for the therapeutic targeting of tumors with mutated p53. One paper in Nature Medicine by Thanos Halazonetis showed proof of concept by introducing second site mutations in mutated p53 that there was a restoration of transcriptional activity by such mutants (4). This suggested that perhaps modifying p53 conformation could restore p53 function and that there was a physical structural basis for the feasibility of doing so. A second paper by Frazan Rastinejad, then at Pfizer in Groton, CT, published in Science demonstrated that a small molecule, CP31398, could in fact alter mutated p53 to a conformation recognized by wild-type p53–specific antibodies and that such p53 was functionally active. Pfizer shut down their p53 program, and NCI developed interest in the approach with some early efforts in preclinical studies to further develop the approach to treat skin cancer. This was an easier approach for a topical therapeutic with possibly an easier path in the clinic. However, years later this has not come to fruition. For the past two decades, Klas Wiman and colleagues have pursued targeting the restoration of p53 function in tumors with mutated p53 and such efforts led to APR-246 that continues to be tested in clinical trials (5).

Naivete and Possible Insanity Are Sometimes Good

Coming out of a great training environment at Johns Hopkins in the mid-1990′s, and with no better oncology fellowship basic science mentor than Bert Vogelstein, I had learned that taking chances at the bench often with some naivete on my part about likelihood of success could pay off. This was the path to discovery of the consensus binding site of human p53 described in Nature Genetics in 1992 and the discovery of p21(WAF1) in Cell in 1993. But, when I started my own lab and more actively started doing drug discovery research around 2002, we set up a functional approach of using p53 reporters to screen for therapeutic molecules in tumors with mutation or loss of p53. Mentors were skeptical because “you can't really fix something that's lost or deleted” unless you pursue gene therapy as mentioned by Levine (2) but this has its own challenges.

Whether it was more naivete on my part, ignoring mentors’ views, persistence without knowing when to quit (I was told early in my career that one of my gifts is persistence in science and in the pursuit of goals), or possible insanity, we forged ahead with what 20 years ago was technologically innovative. We used tumor cell lines with mutant TP53 with a luciferase reporter and used live cell imaging for high throughput screening (6). It became clear that we could see reporter upregulation at early times and lower doses and that at later times or higher doses cells were killed. This coupling of activation of a molecular response to tumor cell death we believed was a unique and powerful screening method, and while it was disclosed to University of Pennsylvania as a novel screening technology with a published patent, there was never much interest as anyone can do it and I was educated that it is impossible to enforce its patent rights. Such a platform was novel at the time and I remember giving talks at Pfizer and Schering Plough and throughout the country about the approach.

The idea of restoring p53 function or perhaps bypassing it, in a manner that wasn't mutation-specific, was appealing as it means that if successful, in this case, one wouldn't need to develop different therapeutic agents for each TP53 mutation. There are many mutations in TP53 well beyond the various DNA-binding domain hotspots. This doesn't mean that a specific small molecule treatment would necessarily work equally well for all TP53 mutations in different tumors as it is clear different mutations have heterogeneous effects and different gain-of-function effects in addition to the loss of tumor suppressive function. Of course, in the end, in oncology we combine drugs. As long as there is some efficacy and evidence of pathway activation or effective bypass, there would be opportunity to improve upon it in various ways including more potent molecules and drug combinations.

Another recent example of an approach that perhaps had little chance to yield what it was actually designed to do led to the identification of two very interesting microRNA's. There was some serendipity there. One microRNA mIR-6883-5p (or miR-149*) inhibit expression of CDK4/6 and may open up a field to block CDK4/6 in a new way at the protein expression level as opposed to kinase activity inhibition. That work also suggested poorly understood regulation and possible interrelationships between cell-cycle regulation and circadian rhythm regulation. More recently, another hit from the microRNA library was miR-3132, which is the first miR that appears to upregulate TRAIL and may offer an alternative or complementary way to stimulate the innate immune pathway in cancer therapy.

Actual Molecular Mechanisms for Bypass and Functional Restoration of a p53 Signaling Pathway in Tumor Cells with Mutated p53

What became clear as we carried out the screening on a smaller scale was that there were hits that activate a p53 reporter and upregulate endogenous p53 targets (6). Early on it was clear that some of the hits required p73 for the observed reporter activity (6). The early hits were not particularly well suited for development and we realized we needed to do more screening for better compounds. These efforts benefited from funds through the NCI cancer prevention program with collaboration by Dr. Levi Kopelovich that at the time in the mid-2000′s to about 2014 or so supported our screening efforts of about 200,000 compounds from a large NCI DTP library and a ChemBridge library. These efforts led to the identification of prodigiosin analogs, NSC59984 and CB002 or other xanthine analogs (7–9).

NSC59984 Requires p73 and Triggers Mutant p53 Degradation through an ROS–ERK2–MDM2 Axis

While it is evident that NSC59984 requires p73 to activate a tumor suppressive p53 pathway response, it is also capable of triggering a degradation of mutant p53. The degradation mechanism appears to involve an ROS–ERK2–MDM2 axis in cancer cells (9). We have wanted to screen for molecules that can more efficiently do both things (which is very feasible) but have not had appropriate resources to do so in a rapid efficient manner. It has also been interesting for us to see that all analogs of NSC59984 through the NCI libraries became “permanently unavailable.”

NSC59984, Maleimide Derivatives, 9-ING-41, and GSK3

We are continuing to study NSC59984 and analogs and we actually recognized recently that NSC59984 is a maleimide derivative. We had been working on a commercially available GSK3 inhibitor called 9-ING-41 (being developed by Actuate Therapeutics) that has promising activity in clinical trials and uncovered an immune stimulatory mechanism not previously recognized (presented at 2022 AACR meeting by Kelsey Huntington). In the process of looking for available analogs, amazingly, NSC59984 came up and we have been looking to see if it has activity towards GSK3. We had tested NSC59984 for immune stimulatory activity and it wasn't particularly robust although we think this is another property that could be and should be optimized.

Integrated Stress Response Activation Is a Valid Therapeutic Target in Cancer Therapy

Our functional cell-based molecular phenotypic screening approach in drug discovery to target tumors with mutated TP53 became more interesting in recent years with a recognition that we were tapping into the integrated stress response. We had success with the discovery of TIC10/ONC201 using a TRAIL gene promoter luciferase-reporter construct lacking p53 DNA-binding sites, and so as proof of concept with such a drug shrinking brain tumors in patients who received the original compound from the NCI diversity set, that was all I needed to persist with the validity of the mutant TP53 therapeutic targeting approach. ONC201/TIC10 also highlighted another important controversy in the field regarding whether we should block or activate the integrated stress response in cancer therapy. Leaders in the field including Glimcher or Koumenis suggest ER stress should be blocked in cancer therapy or suggest the ISR is tumorigenic. However, it is unmistakable that ONC201 activates the integrated stress response with upregulation of ATF4, CHOP, and TRAIL receptor DR5, and interestingly DR5 is a p53 target gene.

Focusing on the Integrated Stress Response to Bypass Mutant p53 in Cancer Therapy

What became evident and really interesting was that prodigiosin analogs and CB002 analogs appeared to be tapping into the integrated stress response (7, 8). This is interesting because while the integrated stress response can under certain circumstances promote cell survival, the fact that small molecule therapeutics activate the ISR, leading to tumor cell death, has therapeutic relevance.

In recent years, we have begun to more fully map out the transcriptomic and proteomic drug responses that accompany p53 pathway activation (7, 8). In addition, we have recognized and have begun to systematically unravel the p53-dependent (and p53-independent) transcriptomes and cytokinomes following treatment of tumor cells with classical chemotherapy that is known to signal activation of p53. It is clear while there is much basic science focused on p53 and therapeutics, our knowledge in the field of oncology with regard to classical drug mechanisms remains primitive.

As a clinical oncologist who sees patients with colorectal cancer every week, I know that agents such as 5-FU, irinotecan or oxaliplatin can activate p53. But, is that activation relevant or does it predict anything, and what about the TP53-independent effects of the compounds, or what happens when they are combined? How much do these effects impact on the drug toxicities and what more can be learned from the cytokine activation profiles?

Different Drugs That Partially Restore a p53-Transcriptome, Require Different p53 Targets Such as Puma, Noxa, or DR5 to Kill Tumor Cells

For reasons that still are mysterious, PG3-Oc upregulates Puma, which is required for the compounds antitumor effect (7). It does upregulate DR5 but this is only required when the compound is combined with TRAIL (7). On the other hand, CB002 and its analogs upregulate Noxa, which is required for tumor cell death, and both PG3-Oc and CB002 or its other xanthine analogs do not require TP73 (7, 8). Our ongoing work is suggesting that the mechanism by which PG3-Oc engages the integrated stress response is distinct from how ONC201/TIC10 does it. It does appear to involve eIF2-alpha upstream kinase HRI, which it has in common with ONC201. Nonetheless, it is clear that the integrated stress response may be a final common pathway for drug activation in cancer therapy and that activation of the ISR may effectively bypass tumor progression and therapy resistance in tumors with mutated p53.

There are many open questions that we and others are pursuing in the field. One of the big questions is what determines cell fate in the integrated stress response in tumor cells. Is it persistent stress or is there more regulation through protein interactions, posttranslational modifications or other mechanisms such as microRNA or cross-talk with other pathways? Why and how are specific p53 targets that are required for cell death activated and why do different p53-pathway restoring (mutant p53 bypassing compounds) drugs involve different TP53 targets in cell death.

Concluding Remarks

The message of this article is that as a community we need to remain open minded about important clinical problems such as targeting mutant TP53 in cancer therapy as there are many complementary approaches. The undertaking is exceedingly complex and requires basic science, clinical insights, and translation.

I have wondered for years why there isn't a national effort such as the Ras initiative for p53. With the resources put into that program at NCI Frederick and nationally we are starting to see some fruits with 2 drug approvals in 2021 targeting KRAS G12C. There is so much more to do there but the future does look brighter now with drugs targeting other KRAS mutations.

Similar large-scale efforts are needed to address the problem of mutant TP53 in cancer, whether it is through Moonshot funds or other mechanisms that support worthwhile initiatives. A national p53 initiative is overdue and arguably is critical to address to impact on the majority of human cancers with mutated p53. A national p53 initiative certainly deserves its place in discussions planning big science programs, difficult problems whose solutions are expected to impact on patient outcomes, and maybe even cancer prevention. A national p53 initiative certainly deserves the investments being allocated to worthy efforts such as Moonshot and strategies that can impact on a variety of tumor types including brain tumors.

It is clear in oncology both that combination therapies ultimately have a chance to bend the Kaplan–Meier curves and that such efforts are desperately needed for targeting or pathway active agents that can synergize with immunotherapy and other treatment modalities. These ideas shouldn't be too controversial for those who want to do some good, and I think a bit of naivete and persistence is not such a bad thing.

Authors' Disclosures

W.S. El-Deiry is the scientific founder and shareholder of Oncoceutics, Inc. (acquired by Chimerix), p53-Therapeutics, Inc., and SMURF-Therapeutics, Inc.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.