Niemann-Pick type C (NPC) is a tragic childhood neurodegenerative disease. The majority of NPC cases (95%) are caused by mutations in the npc1 gene. Npc1 is a late endosomal/lysosomal (LEL) protein that is necessary for proper lipid trafficking. When Npc1 is defective, autophagy is abnormal. In addition, cholesterol and glycosphingolipids accumulate in the LEL. Extensive research by many labs indicates that liberation of these lipids from the LEL of neurons should slow disease progression of NPC.
2-hydroxypropyl-B-cyclodextrin (HP-B-CD) prevents neurodegeneration in NPC mice. HP-B-CD forms a donut with a hydrophobic core, and is used in several applications to solubilize hydrophobic molecules. For instance, it is used in pharmaceutical formulations as an excipient to dissolve and deliver hydrophobic drugs, and is the active ingredient in Febreeze — it binds hydrophobic odor molecules and shields them from our odorant receptors. Remarkably, HP-B-CD liberates cholesterol from the LEL of npc1 mutant cells. These data imply that liberation of cholesterol from the LEL of neurons will prevent their degeneration. Recent reports from Vtesse, Inc and the NIH indicate that HP-B-CD performed well in the Phase 1/2A clinical trial and it slowed disease progression. In addition, the dose-finding portion of the ongoing Phase 2B/3 trial is now complete. Two downsides to HP-B-CD are its ototoxicity (hearing loss) and its delivery through lumbar intrathecal injections so that it can gain access to the brain. However, if successful to FDA approval, many NPC families will likely consider HP-B-CD a miracle, as they currently have no approved treatment.
Suberoylanilide hydroxamic acid (SAHA) was found to rescue slow growth of an NPC yeast model and LEL cholesterol accumulation in patient derived cells. SAHA’s mode of action in NPC is likely through boosting the levels of partially defective Npc1 proteins. How SAHA does this is still a mystery, but might be through regulating the acetylation of chaperon proteins like HSP90. SAHA stabilizes hypomorphic alleles of GBE in Gaucher’s patient derived cells by regulating the association of HSP90. Perhaps in NPC cells, SAHA leads to an improved chaperon activity of HSP90 upon mutant Npc1 proteins, leading to their stability. That theory has not been tested, or is at least not published. SAHA in NPC continues to gain interest because when it is delivered to the brain of NPC mice, it can prevent neurodegeneration and improve lifespan. SAHA (trade name Vorinostat) is FDA approved for cutaneous T cell lymphoma, so continued encouraging results would make it a candidate for repurposing and yield a faster route to provide NPC patients a much-needed treatment.
Drosophila with defective npc1 exhibit the same cellular defects as NPC patient cells, where accumulated cholesterol in the LEL can be observed with the filipin stain (Figure 1). We model rare diseases in simple organisms like Drosophila, and use them to discover small molecule therapies. We reasoned that since HP-B-CD and SAHA can remediate human NPC cellular defects, if they work in flies, we would have further justification that this invertebrate NPC model is valid to the human disease. We would also have more peace of mind that we can use NPC flies to discover new small molecules with therapeutic potential for NPC. Here I’ll describe data generated by me, Kiran Singh, and Tamy Portillo Rodriguez indicating that both HP-B-CD and SAHA rescue our Drosophila NPC model.
Figure 1. Drosophila ring gland where npc1 has been knocked down with RNAi and stained with filipin to reveal puncta of trapped cholesterol in late endosomes and lysosomes (courtesy Luis Milla, Universidad de Santiago de Chile).
We collaborated with the University of Utah’s Mutation Generation and Detection facility (Kelly Beumer and Timothy Dahlem) and Rainbow Transgenic Flies, Inc. to CRISPR generate mutations in the npc1 gene. Unlike our strategy with ngly1, where we used CRISPR to insert a selectable marker bearing transgene into the coding sequence, with npc1 mutagenesis we introduced small INDELS. From a practical standpoint, introduction of a transgene with a selectable marker simplifies pinpointing correctly generated mutant animals. When introducing INDELS, PCR based strategies or complementation tests must be employed. Squishing flies and doing DNA preps or setting up a few hundred fly crosses can make nice people grumpy. However, grumps can still do good things and we discovered cool new alleles. One of which is a 4bp deletion leading to a frameshift and an Npc1 protein possibly truncated between the 8th and 9th transmembrane domain. Accordingly, we named this allele npc18TMs, though we might rename it if we eventually find that the transcript is degraded through nonsense mediated decay. We also created an npc12TMs.
Like previously documented npc1 mutants, npc18TMs larvae are developmentally arrested at the 1st instar stage and eventually die (Figure 2). Though the npc18TMs larvae appear to get bigger, they are essentially chubby 1st instars and never make it to the 2nd instar. The reason npc1 mutants don’t develop is because they are deficient for the steroid hormone ecdysone (20E) that drives developmental progression. 20E is synthesized in mitochondria, and likely is not synthesized as well in npc1 mutants because its precursor, cholesterol, is trapped in the LEL.
Figure 2. Developmental progression of wild-type and npc18TMs larvae. Wild-type (npc1/+) larvae are GFP+. Npc18TMs larvae arrest at the 1st instar.
Supplementing the diet of npc18TMs larvae with sterols (cholesterol or 7-dehydrocholesterol (7-d)) rescues their 1st instar developmental arrest and lethality. We found that the EC50 of 7-d is ~5uM, which enables about 50% of npc18TMs larvae to develop beyond the 1st instar. Those larvae also develop into pupae. We fed npc1 larvae 5uM 7-d along with other small molecules of interest, and measured the % that developed into pupae. HP-B-CD (2% and 0.2%) and SAHA (10uM) both provide strong rescue of the npc1 larvae to pupae (Figure 3).
Figure 3. Rescue of npc18TMs larvae to pupae when fed a diet supplemented with 5uM 7-d along with other small molecules of interest. SAHA (10uM), HP-B-CD (2% and 0.2%), and 50uM “7-d” provide strong rescue.
As expected, HP-B-CD and SAHA also rescue larval developmental delay and viability (Figure 4A-D). When npc18TMs larvae are fed 2.5uM 7-d, which is a suboptimal dose, HP-B-CD and SAHA supplementation rescues complete lethality (Figure 4E).
Figure 4. HP-B-CD and SAHA rescue npc18TMs larval developmental delay and lethality.
HP-B-CD improves solubility of cholesterol, so initially we thought rescue by HP-B-CD might indicate that we have simply increased the effective concentration of sterols in the fly food or fly tissues. To disprove these uninteresting explanations, we also tested sulfobutylether-B-CD (SBE-B-CD), a variant that does not solubilize cholesterol, but binds it and is capable of transferring it from membrane to membrane (eg, lysosome to plasma membrane or mitochondria). SBE-B-CD rescued npc18TMs larval developmental delay almost as well as HP-B-CD (Figure 4A). Presumably, the HP-B-CD and SBE-B-CD are ingested by the larvae, they travel through the hemolymph (blood), penetrate the tiny cluster of prothoracic gland (PG) cells where 20E is synthesized, remove trapped cholesterol from the LEL, and allow it to reach the mitochondria to enable 20E synthesis. Amazing!
Why SAHA rescues npc18TMs mutants is mysterious. SAHA has been shown to increase the levels of human and mouse Npc1 mutant proteins that carry missense mutations. SAHA also only rescues the LEL cholesterol accumulation phenotype of cells carrying such missense mutations, not npc1 null cells. So, it’s been inferred that SAHA works by increasing the level of certain Npc1 mutant proteins that are otherwise misfolded and susceptible to degradation. It is hard to imagine how this chaperoning mechanism could provide much benefit to the Drosophila Npc18TMs mutant protein, which should be a pretty drastic loss-of-function allele. However, npc18TMs might not be that disrupted. When we tested various concentrations of 7-d to rescue npc1A1 mutants, which have a small deletion that takes out the start codon, we found that the EC50 is ~300uM. Since the EC50 of npc18TMs is 5uM 7-d, we are considering it a weak/moderate loss-of-function allele. Therefore, perhaps the hypomorphic Npc18TMs protein can be restored by SAHA through a chaperoning mechanism. Given the power of Drosophila genetics, the npc18TMs fly strain is a valuable tool to determine the mechanism of action of SAHA in NPC.
HP-B-CD and SAHA rescue of npc18TMs indicates that it is a clinically relevant model of NPC. Both of these small molecules have mounting support for their ability to rescue human and mouse NPC cellular and organismal defects. The data presented here indicate that if they had been in an unbiased small molecule screen with our npc18TMs flies, we would have discovered them. Therefore, we can discover new small molecules that will also have clinical relevance. Furthermore, HP-B-CD and SAHA rescue of npc18TMs indicates that new molecules with similar mechanisms of action can be discovered. This NPC disease model is sure to yield new and interesting biology and might enable the discovery of new NPC therapies.