Drugs known as PARP-1 inhibitors have emerged as an important but limited treatment option for certain cancers. Now scientists at Oregon Health & Science University have uncovered a new class of PARP-1 inhibitors with unique and powerful anticancer properties that could make them more widely effective.
“We think it’s going to open up therapeutic possibilities beyond what current PARP-1 inhibitors are used for,” said Michael Cohen, Ph.D., associate professor of chemical physiology and biochemistry in the OHSU School of Medicine and senior author of a paper describing the discovery in Cell Chemical Biology. OHSU has exclusively optioned the potential cancer treatment to a startup company co-founded by Cohen.
Photo: Michael Cohen in his lab with Moriah Arnold, co-author of the study unveiling a new class of PARP-1 inhibitors. (OHSU/Christine Torres Hicks)Through intensive basic science, opening a new window on the workings of PARP-1 inhibitors and investigating even the smallest molecular differences between potential treatments, Cohen’s lab has discovered a candidate for an anti-cancer drug — a molecule whose sticky properties make it toxic to cancer cells at very low doses.
“This is something totally novel from what the existing compounds are doing,” Cohen said. “For all intents and purposes it just locks on to its target and almost never lets go. And we think that’s why it is incredibly cytotoxic in cancer cells.”
The discovery started off with a puzzling observation by Moriah Arnold, a research assistant in Cohen’s lab and now an M.D./Ph.D. student at OHSU. In cancer-initiating cells, Arnold noted that an experimental PARP inhibitor called AZ0108 triggers DNA replication stress, which is the slowing or stalling of the process that cells use to copy their genes during cell division.
This was an eye-opening difference from the PARP-1 inhibitor drug olaparib. Like similar drugs approved by the Food and Drug Administration, olaparib targets tumors that have defects in DNA repair, and works by stopping the PARP-1 protein from helping cancer cells repair damaged DNA. These FDA-approved drugs have yet to find broad use as single agents in treating cancers that do not have DNA repair defects. They do not act by triggering replication stress in multiplying cancer cells, as AZ0108 does.
Curiosity aroused, Cohen decided his lab should dig deeper.
Control of PARP-1 activity involves an effect called “allostery”: When one signaling molecule binds to the protein, it changes how that protein binds to another molecule at a distant site. When PARP-1 binds to damaged or single-stranded DNA, it causes a long-range conformational change, meaning change in shape, that allows PARP-1 to bind to its substrate molecule, NAD+, which leads to PARP-1 activation.
Allostery also works in reverse with PARP-1: Molecules that mimic the substrate molecule NAD+ increase PARP-1’s affinity for binding to DNA. Researchers have speculated that certain PARP-1 inhibitors used in cancer treatment work by means of reverse allostery, but more recent studies have shown that olaparib and other FDA-approved PARP-1 inhibitors do not.
Although the experimental PARP-1 inhibitor, AZ0108, and the FDA-approved olaparib are structurally similar molecules, they differ in how they engage the binding site on PARP-1, Cohen said. Instead, AZ0108 appeared work by reverse allostery.
In further experiments, Cohen’s team delved into the mechanism by which AZ0108 acts as a reverse allosteric inhibitor. First, they tested whether mutations in the PARP-1 protein could disable the reverse allosteric signal. They discovered that changing a single amino acid in a specific part of the protein could silence the signal. While AZ0108 still binds to the mutant PARP-1, the binding no longer triggers the allosteric change at the distant DNA binding site.
“What this says is at that particular position on PARP-1, there’s some interaction that’s key for driving this reverse allosteric change and enhancing DNA binding,” Cohen said.
His team then generated a series of molecules similar to AZ0108, but with subtle changes at one site they predicted would be physically close to the position on PARP-1 that they found was critical to allosteric signaling.
“These changes didn’t much impact the ability of these compounds to inhibit the catalytic activity of PARP-1,” Cohen said. “What they really impacted was this reverse allosteric enhancement of PARP-1 binding with DNA.”
They also found that the extent of inhibitor-induced replication stress correlated with the magnitude of the reverse allosteric effect on DNA binding.
“The synthesis of these molecules showed that the replication stress that we see, induced by these compounds, is really due to the ability of these compounds to lock PARP-1 onto DNA by this reverse allosteric mechanism,” Cohen said.
Quite unexpectedly, several of the synthesized molecules were substantially more potent than AZ0108 in inhibiting cancer cell growth. One of the synthesized molecules, dubbed Pip6, proved to be an incredible 90 times more toxic to cancer cells than AZ0108.
“This was surprising because AZ0108 and Pip6 are almost identical compounds,” Cohen said.
It was hard to explain the massive difference in toxicity to cancer cells, Cohen said. Replication stress is fundamental to the toxicity of these compounds, but in measures of replication stress, Pip6 scored slightly lower than AZ0108.
“This is where things got a little bit confusing for us,” Cohen said. “We thought it was a nice clean story and I was like: okay, now what?”
The explanation came as another surprise. It comes down to the stickiness of the Pip6 molecule in binding with PARP-1. Experiments showed that Pip6 stays persistently bound to the protein, which, in turn, persistently maintains the reverse allosteric signal that keeps PARP-1 bound to DNA.
Tilikum Therapeutics Inc., the startup Cohen co-founded, is aiming to develop the next generation of inhibitors of PARP-1, which is a critical target in ovarian, breast and prostate cancers.
The attributes of Pip6 make it a good candidate for an anti-cancer drug. It is toxic to cancer cells at very low doses. Its persistent binding to the target protein suggests that it could be given less frequently than existing PARP inhibitors. And its unique mechanism of action compared with clinical PARP-1 inhibitors means that it could have much broader clinical use in cancers.
The next step, Cohen said, is testing Pip6 and related molecules in animal models to assess toxic side effects, minimum dosing levels, and how often doses need to be given to maintain effectiveness.
“We’re hoping that we can dose at very low levels,” he said. “And because it has this long residence time, that maybe it’s not even a daily dose that is required. Maybe it’s just every few days or once a week.”
Allosteric regulation of DNA binding and target residence time drive the cytotoxicity of phthalazinone-based PARP-1 inhibitors by Moriah R. Arnold, Michael S. Cohen, and others. Cell Chemical Biology (Sep. 15, 2022)