Unraveling the sterol-trafficking defect in Niemann-Pick C disease

The interorganellar transfer of lipids, particularly cholesterol, is imperfectly understood but is a key component of membrane homeostasis as shown by the lethal disorders that are associated with its derangement. For unknown reasons, the neuron is particularly susceptible to excessive lipid accumulation. In the case of Niemann-Pick type C (NPC) disease, a lysosomal lipid storage disorder, the accumulation of cholesterol and sphingolipids manifests as a fatal neurovisceral degenerative disorder. Liu et al. (1) report in this issue of PNAS that defective trafficking of cholesterol and disease severity can be circumvented in the npc1^−/− mouse by a single and timely injection of the cholesterol-binding agent, 2-hydroxypropyl-ß-cyclodextrin (CYCLO).

Genes of the NP-C Pathway

Two clinically and pathologically indistinguishable diseases (NPC1 and NPC2) are associated with mutations in either of two distinct genes (NPC1 or NPC2). In vitro studies have documented binding of cholesterol to both NPC1 (2) and NPC2 (3) proteins, albeit with marked differences in affinity. The most recent data suggest that cholesterol is exchanged in a tandem fashion between these 2 proteins although the specter of a third protein has recently loomed (4) consistent with the observation that mutations cannot be identified in NPC1 or NPC2 in some patients with NPC disease. In vivo studies, which largely use the npc1-null mouse model, have demonstrated that all organ systems are affected by NPC mutations, but only agents that appear to cross the blood–brain barrier alleviate neurodegeneration and extend lifespan (1). The authors suggest that the cholesterol-binding properties of CYCLO may allow this agent to substitute for dysfunctional NPC1 protein in the npc1-null mouse. In solubilizing the acid lipase-generated free cholesterol that becomes trapped in lysosomes, CYCLO is also taken up into organelles by bulk phase endocytosis and may then be able to transfer the cholesterol to the NPC2 protein pool, thus reestablishing transfer through the NPC pathway. This novel and enticing hypothesis can now be tested in the NPC2 mouse model (5). CYCLO may provide a useful tool to probe the interaction of NPC1 and NPC2 proteins, which modulate the relocation of lysosomal cholesterol to regulatory cytosolic pools.

Role of Cholesterol in NP-C

In a series of earlier publications, the Dietschy and Repa laboratories elegantly demonstrated the central role that defective cholesterol trafficking out of lysosomes plays in the cellular pathogenesis of this disorder in the NPC mouse. (6). The current observations suggest that this offending metabolite has been sufficiently mobilized in a key tissue, such as the brain, to impact disease. However, the possible pathogenic role of many other lipids as additional “offending metabolites” remains unsettled (6, 7). Sterol and sphingolipids frequently colocalize with important consequences, and may even be cotransported. The authors have not yet monitored other lipids that are known to accumulate in NPC disease.

CYCLO may provide a useful tool to probe the interaction of NPC1 and NPC2 proteins.

The effect of CYCLO in NPC mice has exciting heuristic implications for a broad spectrum of important mechanistic and perhaps, ultimately, therapeutic questions. Lessening the cholesterol metabolic burden in the NPC mouse has been shown to clearly moderate the severity of the disorder in this model (1), thereby indicating that reversing the sterol-trafficking defect itself might improve the course of NPC. Early support for this notion was found in a study reporting that blocking chylomicron delivery to the liver lowers cholesterol content and improves liver function. However, neither CNS lipid storage nor disease symptoms were ameliorated by this intervention (8). A second study showed that LXR agonists increase cholesterol loss from the brain without altering synthesis (9). Neither of these physiological manipulations appeared to alter cholesterol permeability across the limiting membrane of the lysosomes in which the cholesterol was trapped; this may explain their limited effects. The unique aspect of the current study is that administration of CYCLO to the npc1^−/− mice appeared to reestablish sterol movement out of lysosomes.

Mechanism of CYCLO Action

The next critical question is how CYCLO might induce such responses in these animals. An important and somewhat surprising observation in the current study was that CYCLO administration to normal mice showed no effect on whole body cholesterol turnover. In the npc1^−/− mouse treated with CYCLO the 3 cellular signals of rapid lysosomal cholesterol depletion into the cytosolic compartment were clearly documented: (i) an immediate increase in cholesteryl esters, (ii) a suppression of the SREBP target genes, and (iii) an up-regulation of LXR target genes. It is clear from this and other studies that the timing of CYCLO administration is critical. Seven-day-old mice (perhaps especially those with NPC disease) may have a more permeable blood–brain barrier that becomes progressively less permeable during development. Liu et al. (1) show remarkable lessening of the cholesterol burden from a single administration at this juncture that continued to be effective despite the virtual disappearance of CYCLO from the animal within 24 h (only 9% of the administered dose remained in the animals, at this time point). Whether this reflects an epigenetic effect of CYCLO treatment or a singular salvage event that transiently spares neurons from the ravages of loss of these functional NPC1/NPC2 proteins, remains to be addressed. The bottom line is that a single application of CYCLO and the ensuing homeostatic responses led to net loss of sterol from the whole animal and a marked improvement in the clinical state of the mutant mouse.

Many critical questions must be answered before CYCLO or a related compound reaches the clinic. What are the permeability coefficients for CYCLO across the capillary beds of every organ, including the central nervous system, as a function of age? Would repeated injections of CYCLO cure the liver and lung, but not the CNS? Would direct administration of cyclodextrin into the CNS improve half-life and response? Finally, it is important to bear in mind that contrary to the NPC mouse model where cholesterol has been shown to act as an offending metabolite, a similar role for this sterol in human NPC disease remains to be fully established. Such a role has been suggested from documentation of delayed appearance of LDL-derived cholesterol in the plasma in NPC patients (10). At present, CYCLO represents an exciting tool for exploring the cell biology of NPC disease that needs further experimental validation in animal models before it is considered as a therapeutic agent.