Broken tumor-suppressor proteins make notoriously poor targets for cancer therapies designed to activate them. Responsible for keeping cell growth division in check, tumor-suppressors are often mutant or missing in cancer cells, giving the cells free rein to proliferate. Hobbling such derelict proteins with drugs – the standard tactic in targeted medicine – might only make matters worse.
But, as Dana-Farber Cancer Institute scientists show in a new study in the journal Cell, some tumor-suppressor proteins that are intact but inactive can be roused with small-molecule agents and reassert their control over cell growth. In T-cell acute lymphoblastic leukemia (T-ALL) cells and tissue samples, they found that a class of small molecules that switch on a specific tumor-suppressor complex can bring cell division to a screeching halt, resulting in the death of the cells.
One of the most extraordinary aspects of the discovery was the economy with which the small molecules work. “By causing the tumor-suppressor’s component parts to assemble into active complexes, they had the desired anti-cancer effects and no side effects on normal cells,” says the study’s senior author, A. Thomas Look, MD, of Dana-Farber. “When the molecule was tested in mice implanted with human T-ALL tissue, it caused tumor cells to die but left normal cells unchanged – something we rarely see with targeted anti-cancer drugs.”
The trail that led to the new study began with research that focused on a drug for a condition as far afield from cancer as can be imagined. In 2008, Alejandro Gutierrez, MD, of Look’s lab conducted a screen of nearly 5,000 drugs in animal models of T-ALL. The most effective was perphenazine, a 50-year-old drug used to treat schizophrenia.
In 2014, Alex Kentsis, MD, PhD, of Look’s lab and Jon Aster, MD, PhD, of Dana-Farber and Brigham and Women’s Hospital found that perphenazine achieved its anti-leukemia effect by activating an enzyme known as PP2A. But because perphenazine can produce severe movement disorders, it has largely been replaced by other drugs for schizophrenia and wasn’t a practical option for patients with T-ALL.
The fact that perphenazine was effective against T-ALL led Look and his Dana-Farber colleagues Eric Fischer, PhD, and Nathanael Gray, PhD, to search for related drugs that also activate PP2A but have fewer side effects.
PP2A is not a single enzyme but a family formed by at least 15 different combinations of multiple subunits that form active enzymes. Individual PP2A enzymes act as tumor suppressors, interfering with specific signals required by cells to maintain control of their growth.
In T-ALL and other human cancers, PP2A is often shut down by certain inhibitory proteins. Perphenazine works, and awakens the cell’s growth-control mechanism, by overriding one of those inhibitory proteins and activating a specific form of PP2A, called PP2A-B56ε.
PP2A enzymes are known as phosphatases; they work by removing bundles of atoms known as phosphate groups from proteins. Phosphate groups can be thought of as spurs to action: a protein laden with phosphate groups is provoked to function; an enzyme such as PP2A that strips phosphate groups from proteins essentially removes those goads, causing the proteins to become idle and be destroyed. The process of taking away an activating phosphate group is known as dephosphorylation.
Look, first author Ken Morita, MD, PhD – a postdoctoral fellow in Look’s lab – and their colleagues worked to uncover which specific form of PP2A perphenazine targets.
PP2A enzymes are a highly diverse bunch; each is made up of three classes of subunits, which can combine to form at least 15 different phosphatases. In cancers with non-functioning PP2A, it’s rare for any of these subunits to be defective. Rather, PP2A’s inactivity is caused by a profusion of inhibitory proteins.
To identify agents that might switch on PP2A without causing movement problems, Morita screened 90 different compounds, most of which were structurally related to perphenazine. One, which investigators dubbed iHAP1, was 10 times more effective than perphenazine in stirring PP2A into action but doesn’t block receptors in the central nervous system associated with the toxicity of perphenazine.
By creating cells lines capable of forming different versions of PP2A – each one missing just one of PP2A’s subcomponents – and treating them with iHAP1, Morita zeroed in on the precise subcomponent targeted by the agent – a subunit known as PP2A-B56ε.
Morita and his colleagues then set out to find the target of PP2A-B56ε – that is, which protein does PP2A dephosphorylate when stimulated by iHAP1? To do so, they used mass spectrometry to examine proteins in cells treated with iHAP1. The most extensively dephosphorylated protein – and the one whose dephosphorylation caused cancer cells to die – was MYBL2.
MYBL2 and other members of the MYB family have important roles at various phases of cell growth and differentiation. The genetic circuit activated by MYBL2 comes into play at the mitosis, or M, phase, when a cell prepares for and undergoes division. By this point, the cell has made duplicate copies of its chromosomes, which should align in pairs at the center of the cell. MYBL2’s job is to round up the many different types of proteins that will be needed when division begins, like a movie producer assembling the cast and crew of a feature film. Among these proteins are some that serve as tiny motors to pull the paired chromosomes apart, towing each member of the pair into one of the newly forming daughter cells.
“MYBL2 is a master regulator of proteins required during the M phase of the cell cycle for every cell in the body,” Look comments. “Cancer cells can be particularly dependent on it.”
Now the mechanism behind iHAP1’s ability to kill T-ALL cells without harming normal cells was clear: iHAP1 switches on PP2A-B56ε, which tears a phosphate group off of MYBL2, knocking MYBL2 out of commission. Without MYBL2 to marshal the proteins needed for cell division, mitosis comes to an abrupt halt. Instead of giving rise to two daughter cells, the process is aborted. Microtubules within the cell – tiny tubes that normally drag the chromosomes apart into two daughter cells – form a disorganized, star-like shape. “The cancer cell cannot recover from this condition, and commits suicide,” Look explains.
The identification of iHAP1 and the discovery of its mode of action point to an intriguing new therapeutic approach to cancers like T-ALL that have intact, but idled, PP2A-B56ε tumor-suppressor genes and proteins, the investigators say.
“Our results illustrate that small molecules can be used to activate PP2A complexes with extreme selectivity,” Morita remarks. “This specificity makes them useful as probes to determine whether activating a particular PP2A complex is sufficient to inhibit cancer cell growth, an important step in developing targeted drugs that selectively activate these complexes.”
The other co-authors of the study are Shuning He, PhD; Radosław P. Nowak, PhD; Jinhua Wang, PhD; Mark W. Zimmerman, PhD; Cong Fu, PhD; Adam D. Durbin, MD, PhD; Megan W. Martel, and Nicole Prutsch, PhD; of Dana-Farber.
The study was funded by the National Institutes of Health (grants R35 CA210064; R01 CA214608; and R01 CA218278); the Lauri Strauss Leukemia Foundation; and the Thrasher Research Fund.