19 Feb 2016
Preclinical studies suggest that blocking histone deacetylases, which modulate gene transcription, could potentially treat neurodegeneration, but in practice, the theory has not lived up to its promise. Most HDAC inhibitors do not cross the blood-brain barrier, and those that do are toxic because they alter the expression of many genes. To date, the Food and Drug Administration has approved only a handful of HDAC inhibitors, to treat peripheral T-cell lymphomas.
In the February 17 Science Translational Medicine, researchers led by Kasturi Haldar at the University of Notre Dame, Indiana, describe a new strategy for blocking HDACs in the central nervous system that appears to clear some of these hurdles. The authors encased the general HDAC inhibitor vorinostat in the lipid chelator 2-hydroxypropyl-β-cyclodextrin (HPBCD), then coated the combination with polyethylene glycol (PEG). The trifecta persists longer in plasma than the free HDACi, allowing vorinostat to reach higher concentrations than it would otherwise. In mice that model the rare neurodegenerative disorder Niemann-Pick type C disease, the chelated HDAC inhibitor preserved cerebellar neurons, delayed onset of symptoms, and doubled lifespan, the authors reported. Cyclodextrin alone also extends life in NPC model mice, but the new formulation improved survival by an additional 74 days. Moreover, the high vorinostat concentrations achieved in the brain allowed the researchers to dose only once per week, avoiding side effects associated with constant HDAC inhibition. “A key point in this work is that there is no metabolic toxicity long-term,” Haldar told Alzforum. She hopes to take this formulation into clinical trials for the human disorder.
Commenters praised the strategy, while noting that questions remain about how big a role cyclodextrin played in the improvements, and how well the therapy would translate to people. “This is an interesting approach to solving the problem of getting HDAC inhibitors into the brain. If [researchers] can demonstrate that this formulation allows vorinostat to cross the primate blood-brain barrier too, that would be very powerful,” said Yiannis Ioannou at Mount Sinai School of Medicine, New York. Stephen Sturley at Columbia University, New York, agreed. “I find the results very impressive. [Vorinostat and cyclodextrin] clearly have some kind of synergy.”
Neuronal Rescue. Purkinje cells (green) in the cerebellum (left) degenerate in mice that model Niemann-Pick type C disease (middle), but treatment with a combination of an HDAC inhibitor and cyclodextrin (right) preserves some cells. [Courtesy of Science Translational Medicine/AAAS.]
Treatments for Niemann-Pick type C disease are desperately needed. This fatal childhood illness is caused by recessive mutations in the NPC1 or NPC2 gene. Their protein products help transport cholesterol from lysosomes to the endoplasmic reticulum for recycling. The mutations are believed to destabilize the proteins and result in a severe shortage. The ensuing accumulation of cholesterol leads to swollen axons, demyelination, and neurodegeneration of cerebellar Purkinje cells, among others. Children with the disorder lose cognitive function and develop problems walking, moving, and swallowing, and most die as children or teenagers.
No treatments exist that slow disease progression, but both cyclodextrin and vorinostat are under investigation. Cyclodextrin extends life in mouse models and is currently in a Phase 1 trial for children with NPC. Because cyclodextrin does not cross the blood-brain barrier, researchers deliver it directly into the cerebrospinal fluid by lumbar puncture. Meanwhile, vorinostat, which bumps up NPC1 transcription and mops up cellular cholesterol in cell culture, has entered a Phase 1 trial for adults with NPC (see Pipalia et al., 2011; Munkacsi et al., 2011). However, vorinostat dissolves poorly in blood and barely penetrates the brain.
Haldar and colleagues wondered if the two drugs together might work better than either alone. First author Suhail Alam complexed vorinostat with an eightfold excess of cyclodextrin and smothered both in PEG. Because cyclodextrin and PEG bind hydrophobic compounds such as vorinostat, they may help solubilize the inhibitor and then release it slowly over time, Haldar said. Supporting this idea, mice that received the triple formulation had two to three times the plasma concentrations of vorinostat as animals treated with the HDAC inhibitor alone. This higher plasma concentration may allow more vorinostat to cross the blood-brain barrier, while cyclodextrin remains behind in plasma, Haldar suggested. Vorinostat alone barely budged histone acetylation in mouse brain, but this measure rose between two- to ninefold in mice that received the combination. Transcription of NPC1 doubled, while protein levels jumped by eightfold, bringing NPC1 concentrations up to about one-quarter of the level in healthy control mice.
To test the therapeutic effects of this combination, the authors injected it intraperitoneally into Npc1nmf164 transgenic mice, which carry a destabilizing NPC1 mutation. These animals develop severe motor problems, weight loss, and die at around four to five months of age. The researchers administered the drug once weekly from 21 days of life until the animals died. They found that the triple formulation preserved 25-30 percent of Purkinje cells, compared to their near-total annihilation in Npc1nmf164 mice treated with vehicle control (see image above) or vorinostat alone. Mice treated with the triple formulation maintained their motor skills for about a month longer than controls, and survived about twice as long, some making it to nearly 300 days. Vorinostat alone had no benefit. Meanwhile, mice that received cyclodextrin alone reaped some benefits in better Purkinje cell health, delayed symptoms, and longer lifespan, but none as great as with the triple formulation. Each treatment group consisted of about 10 mice.
The formulation also appeared safe, the authors report, showing no signs of the liver or kidney toxicity known to accompany chronic HDAC administration. This may be because the weekly administration gives tissues a “rest period” from HDAC inhibition, Haldar suggested. Cyclodextrin also can have side effects. In animal models, high doses and CNS delivery of this chelator cause irreversible hearing loss (see Ward et al., 2010). Anecdotally, some reports indicate the same problem is cropping up in the Phase 1 trial, Sturley said. Haldar speculated that the triple formulation may avoid this issue, because it uses lower doses of cyclodextrin that do not cross into the brain. However, this remains to be tested.
Commentators wondered, however, how much of the triple formulation’s benefit is due to cyclodextrin. Sturley noted that the authors injected all treated mice with cyclodextrin alone at days 7 and 15 after birth, before adding vorinostat or the triple formulation from day 21. It is unclear if the triple formulation alone, without pre-treatment, would have provided the same benefits, he said. Ioannou added that cyclodextrin treatment for NPC works better in mice than people, hinting that there may be differences between the species in how the compound is metabolized or transported. “It’s difficult to interpret if this [triple formulation] will be a viable strategy in people,” Ioannou said. Haldar pointed out, however, that the triple formulation preserved neurons much better than cyclodextrin alone, indicating a substantial brain benefit from the addition of vorinostat.
Because the triple formulation helps HDAC inhibitors enter the brain, it could theoretically benefit other neurodegenerative diseases where HDAC inhibition is being explored as a treatment, the authors suggest. The cyclodextrin component itself would not necessarily benefit diseases where cholesterol does not accumulate. That said, Haldar noted, its peripheral delivery is regarded as safe by the FDA and tends to lower peripheral inflammation, which might indirectly help patients with an age-related neurodegenerative disease. She plans to explore the potential of the triple formulation in animal models of other diseases.
Meanwhile, commentators suggested that another option would be to use brain-penetrant HDAC inhibitors. One such drug, panobinostat, was approved to treat multiple myeloma. Haldar noted that it is more toxic than vorinostat, killing NPC model mice. Even so, complexing it with cyclodextrin might allow researchers to use a low dose that could be safer, she suggested. Another HDAC inhibitor is currently in Phase 2 for frontotemporal dementia (see Nov 2014 conference news). Several companies are trying to develop HDAC inhibitors for neurological indications, for example Acetylon and the startup Rodin Therapeutics.—Madolyn Bowman Rogers