The normal compound fly eye (left) is marred by cell death in a strain (right) with a mutation causing amyotrophic lateral sclerosis.

KIRSTIN MAULDING

To halt brain diseases, drugs take aim at protein traffic jams that kill neurons

The compound eyes of the common fruit fly are normally brick red. But in neurologist Tom Lloyd's lab at Johns Hopkins University School of Medicine in Baltimore, Maryland, many of the fly eyes are pocked with white and black specks, a sign that neurons in each of their 800-odd eye units are shriveling away and dying.

Those flies have the genetic equivalent of amyotrophic lateral sclerosis (ALS), the debilitating neurodegenerative disorder also known as Lou Gehrig's disease, and their eyes offer a window into the soul of the disease process. By measuring the extent of damage to each insect's eyes, researchers can quickly gauge whether a drug, genetic modification, or some other therapeutic intervention helps stop neuronal loss.

Those eyes have also offered an answer to the central mystery of ALS: just what kills neurons—and, ultimately, the patient.

The flies carry a mutation found in about 40% of ALS patients who have a family history of the disease, and in about 10% of sporadic cases. The mutation, in a gene called C9orf72, consists of hundreds or thousands of extra copies of a short DNA sequence, just six bases long. They lead to unusually large strands of RNA that glom onto hundreds of proteins in the cell nucleus, putting them out of action. Some of those RNA-ensnared proteins, Lloyd and his Hopkins colleague Jeffrey Rothstein hypothesized, might hold the key to ALS.

Over many months, the researchers systematically studied the role of each protein by developing fly strains carrying both the ALS mutation and an incapacitated or hyperactive version of each protein's gene. One set of flies, bred to have elevated levels of a protein called RanGAP, stood out. Fifteen days after the flies emerged from their pupal casings, their eyes remained a pure burnt sienna. RanGAP "was by far the most potent suppressor of neurodegeneration," Lloyd says. What was known about its function was tantalizing: It serves as a courier, helping shuttle other proteins across the membrane that divides the cell nucleus from the cytoplasm.

RanGAP Importin Exportin Ran C9orf72 RNA KPT-350 TDP-43 Nuclear pore Stress granule Cytoplasm Fixing traffic jamsThe flow of molecules through pores in the nucleus is key to the health of cells, particularly neurons. Traffic problems may contribute to amyotrophic later- al sclerosis (ALS) and other brain diseases. NormalImportin proteins bring TDP-43and other cargo from the cytoplasmthrough the nuclear pore, whereasexportins take molecules out. ALSIn some ALS cases, abnormal RNAmade by a mutant C9orf72 geneprevents RanGAP from mediatingnuclear import, leading to proteinbuildup in stress granules. ALS + drugA possible drug, KPT-350, impairsthe specific exportin XPO1 and maynormalize nuclear traffic of proteinsor bust up stress granules.
V. ALTOUNIAN/SCIENCE

The team's result would upend neuroscientists' understanding of ALS and brain disease in general, and others were on the same trail. In 2015, two more research teams reported that defects in the cell's nuclear transport system were hallmark features not only of ALS, but also of frontotemporal dementia (FTD), another progressive brain disease caused by C9orf72 mutations. Scientists would soon link dysfunctional trafficking across the nuclear divide to other neurodegenerative diseases—Alzheimer's, Huntington, spinocerebellar ataxia—and even to normal aging. In all those ailments, the resulting abnormal pileups of proteins somehow become rogue neuronal killers.

"I often get queasy when someone makes a discovery and tries to explain the rest of the world with it," says Rothstein, a neurologist who directs the Johns Hopkins Brain Science Institute. But here, he says, it seems to be true.

The findings are not merely academic. They are inspiring therapeutic efforts to address the cause of general age-related neurodegeneration—a goal that has largely eluded drug developers. If the gradual loss of nucleocytoplasmic transport is a conserved feature of the aging brain, says Sami Barmada, a neurologist at the University of Michigan in Ann Arbor, preventing it "might be a really broad and effective therapy."

Several biotech companies have jumped on that idea, exploring it in animal models and planning the first human trials this year. Chief among them: Biogen in Cambridge, Massachusetts, which in 2018 bought the rights to develop a drug compound called KPT-350 that directly targets the nuclear transport pathway. The research underpinning that drug's action is brand new. But, "The biology is there," says Chris Henderson, head of neuromuscular and movement disorders research at Biogen. "Here's a drug with a body of rationale," he adds, "and we're optimistic about getting this into trials."

Crossing the nuclear border

The lipid membrane that divides the DNA-packed nucleus from the rest of the cell is like an international border busy with two-way industrial traffic. DNA-binding proteins and other molecules are constantly flowing into the nucleus to help turn genes on and off, for example. The messenger RNAs produced by those genes stream the other way, into the cytoplasm to protein-assembly platforms. The cell must regulate that traffic through entry points known as nuclear pores. Choke off those portals and it stands to reason cells will suffer.

The first hints that disrupted nuclear transport might underpin ALS came in 2010, when researchers at King's College London, working with human nerve cancer cells, experimentally blocked the expression of proteins involved in the import business. The result was something also seen in cells from ALS patients: clumps of a protein called TDP-43 building up in the cytoplasm.

Few ALS researchers paid much attention to that early report. What might be gumming up the gears of the transport machinery in ALS patients wasn't clear, and the researchers couldn't say whether the buildup of TDP-43—a protein that normally binds both DNA and RNA inside the nucleus to regulate multiple steps in gene expression—was actually killing neurons or was just a consequence of a different toxic process. It would take another 5 years—and Lloyd's and Rothstein's study of the flies with telltale eyes—for ALS scientists to take nuclear transport more seriously.

Here’s a drug with a body of rationale, and we’re optimistic about getting this into trials.

Chris Henderson, Biogen

The Hopkins team's result electrified colleagues in part because it had identified a transport protein, RanGAP, as key to neurodegeneration. The team showed in both the fly model of ALS and in cells from human patients that the lengthy RNA readouts produced by the mutant C9orf72 gene seemed to stick to RanGAP near the nuclear pore and put the protein out of commission. The loss of functioning RanGAP spurred a backup of the nuclear import system, resulting in the cytoplasmic buildup of proteins such as TDP-43—cluttering a cell like bags of rotting trash during a garbage strike.

Just as galvanizing was the team's finding that a potential drug could preserve neuronal health, at least in the flies. "All of a sudden it threw a potential treatment approach into the ring," says Dorothee Dormann, a biochemist from Ludwig Maximilian University in Munich, Germany.

The team had no drug that could boost levels of RanGAP in the cytoplasm and restore enough inflow to rescue the eye neurons. But Lloyd reasoned that blocking outflow of TDP-43 and other nuclear proteins may have the same beneficial effect. An experimental compound called KPT-276 was known to selectively inhibit a key nuclear export protein called exportin 1 (XPO1). The approach was a hack of sorts, marrying two wrongs—defective inflow and outflow—to make a right, but it worked. When Lloyd gave KPT-276 to his ALS flies, their eyes remained pristine.

From cancer fighter to brain protector

KPT is the experimental compound code used by Karyopharm Therapeutics, a small drug company in Newton, Massachusetts. Karyopharm formed in 2008 to develop XPO1 inhibitors for treating cancer, the idea being to trigger a buildup of tumor suppressor proteins in the nucleus, where they carry out their anticancer watchdog function. A decade on, the company's first clinical candidate, a drug for multiple myeloma, is widely expected to win marketing approval in the coming months.

Chemists at Karyopharm developed a suite of XPO1 inhibitors, including KPT-276 and a relative called KPT-350, that had an important attribute: They crossed the blood-brain barrier more readily than other candidates. KPT-350 proved more potent and less toxic in preclinical testing, so the firm looked for ways to use it to treat brain disease and injury.

Lloyd's and Rothstein's results piqued the company's interest. When Sharon Tamir, its head of neurodegenerative and infectious diseases at the time, learned that the Hopkins researchers were working with KPT-276 and not KPT-350, she called them up to propose a collaboration using the "better" compound. Meanwhile, she began to distribute KPT-350 to other groups in Japan, Belgium, and across the United States. Collectively, those scientists showed the drug's neuroprotective effects across a range of human cell, fly, and rodent models of ALS, Huntington, and other brain diseases.

A false-color atomic force micrograph shows the complex pores (green rings) that closely regulate traffic in the cell between the nucleus and cytoplasm.

H.OBERLEITHNER, UNIVERSITY HOSPITAL OF MUENSTER/Science Source

For example, treatment with KPT-350 preserved the health of axons, the long, signal-transmitting extensions of nerve cells, and improved the motor functions of mice with a multiple sclerosis–like condition, a team led by neuroscientist Jeffery Haines at the Icahn School of Medicine at Mount Sinai in New York City showed. And in the Hopkins group's hands, the drug kept alive mouse neurons harboring the mutation associated with Huntington.

"There's still a lot that needs to be explored about why the nuclear pore complex is so susceptible to problems in different types of neurons in different brain regions causing multiple different diseases," says Gavin Daigle, a former postdoc in Rothstein's lab who worked on the Huntington project and helped link disrupted pore function to Alzheimer's disease before joining AbbVie in Cambridge. But he stresses that all the research is showing one thing: "This is a pathway that can be targeted."

The results proved enough to convince Biogen, which bought the rights to test the drug in humans. "The package of preclinical data that Karyopharm was able to amass really justifies the excitement," says Laura Fanning, R&D project leader for KPT-350 at Biogen (which has renamed the molecule BIIB100). "It's not just a blip of efficacy in one strain of mice. It's a broad base of evidence," she says. A first-in-human dose-escalation study of KPT-350 could begin in ALS patients later this year. If the drug shows promise against that disease, Biogen may expand its clinical testing to other conditions, Henderson says.

Moving into the clinic

Although the drug seems to work in the laboratory, why or how is not at all clear. "The story started to get murkier as more data has come in," notes Haines, now at Regeneron Pharmaceuticals in Tarrytown, New York. Initially, most scientists assumed that because it blocks XPO1, the drug prevents proteins such as TDP-43 from piling up in the cytoplasm by trapping them in the nucleus. But last year, Dormann's team and another led by Philip Thomas, a biochemist at the University of Texas Southwestern Medical Center in Dallas, independently reported that TDP-43 and another protein called FUS seem to exit the nucleus by passive diffusion, not through XPO1-mediated transport. (FUS also clumps in the cytoplasm of motor neurons in some patients with ALS or FTD.)

So if KPT-350 is not acting directly on the transport system, what is it doing? "It looks like the drug is targeting some more general neurotoxic pathway," Dormann says, "but it remains to be clarified what the mechanism really is and which nuclear transport defects we're correcting with this drug."

One possibility, recent research suggests, is that the drug actually targets tiny, dense packets of protein and RNA that form during times of cellular stress. In healthy cells, those membraneless "stress granules" generally break down and their components disperse after a viral infection, thermal shock, or some other environmental insult has passed. Not so in the diseased neurons of people with ALS or FTD. In those cells, the stress granules persist and turn sticky, recruiting proteins such as TDP-43 and FUS and eventually transforming into solid, toxic aggregates.

Over the past year, several research teams have shown that components of the nuclear transport machinery—including importers, exporters, and parts of the nuclear pore itself—also can get tangled up in those aggregates. The transportation system falters, and as more TDP-43 and other proteins are added to the stress granules, a feedback loop takes hold that grinds the molecular traffic to a halt. "TDP-43 is not just a victim of nucleo-cytoplasmic transport defects," says Wilfried Rossoll, a neuroscientist at the Mayo Clinic in Jacksonville, Florida. "It's also a perpetrator."

In August 2018, findings from a study led by neurobiologist Ludo Van Den Bosch of VIB–Catholic University of Leuven in Belgium suggested that the transport protein XPO1 itself may play a role in stress granules. That means a drug such as KPT-350 may serve primarily as a stress granule buster, and any impact on transport may be secondary. "Things are more complicated than initially presented," says Van Den Bosch, who has collaborated with Karyopharm.

The open questions about KPT-350 have not discouraged other groups from pursuing additional strategies to sort out nuclear traffic problems. In 2017, for example, Guillaume Hautbergue and his colleagues at the University of Sheffield in the United Kingdom implicated another export factor in the neuronal loss experienced by ALS flies with the C9orf72 mutation. Hautbergue is working on ways to target that protein to prevent the export of mutant RNAs produced by the gene.

Other researchers are focusing on breaking up stress granules. That approach should free up transport factors and pore proteins held hostage in those granules, allowing them to return to their usual posts in the cell, explains James Shorter, a protein biochemist at the University of Pennsylvania. He is developing a way to equip cells with a gene for making a "disaggregase" protein and has begun to test the therapeutic strategy in a mouse model of ALS.

A few drug companies, including Denali Therapeutics of South San Francisco, California, and Aquinnah Pharmaceuticals of Cambridge, are looking for small molecules that can do basically the same thing. Those therapies may not directly target the nuclear transport pathway, but they would get the job done, says Aquinnah co-founder and Chief Scientific Officer Ben Wolozin, a neuropharmacologist at Boston University's School of Medicine, because dismantling stress granules helps restore healthy nuclear transport. "This is all part of an integrated biological response," Wolozin says.

Aquinnah hopes to begin to evaluate its lead compound in ALS patients this year, about the same time that Biogen is aiming to get KPT-350 into the clinic. For now, Biogen scientists are still trying to pin down what the drug is doing in various genetic models of the disease, including the flies with marred eyes. But to some extent, Henderson says, knowing the exact mechanism of action doesn't really matter. "The relevant experiment," he concludes, "is in the human patient."