Niemann Pick Disease Type C (NPC) is a rare autosomal recessive disorder characterized by lipid accumulation and neurodegeneration (Vanier, 2010). NPC primarily affects infants, children, and young adults, and has an estimated prevalence of 1 in every 100,000 births (Berry-Kravis, 2021). Average life expectancy for NPC patients is 13 years of age, although there is high heterogeneity in age of onset and disease progression (Bianconi et al., 2019). In 2024 the U.S. Food & Drug Administration approved its first two treatments for NPC, representing an exciting breakthrough for researchers and patients: Miplyffa (arimoclomol) was approved to treat NPC in adults and children aged 2 years and older, and Aqneursa (N-acetyl-L-leucine) was authorized to address neurological symptoms associated with NPC in both adult and pediatric patients weighing at least 15 kg (Jan et al., 2025). However, both drugs are aimed at managing symptoms, and the search for curative treatments continues.

NPC is neurovisceral disorder that causes symptoms in both the central nervous system (CNS) and peripheral organs, and clinical manifestations of NPC are typically age dependent. In infancy and early childhood, NPC primarily presents as a liver and spleen disease accompanied by subtle neurological symptoms, such as delayed developmental milestones (Geberhiwot et al., 2018). Adolescent and adult patients are more likely to present with psychiatric symptoms that can mimic schizophrenia or mood disorders (Rego et al., 2019). Neurodegenerative symptoms including motor control issues (clumsiness, seizures, gaze palsy), behavioral changes, and memory loss are common in patients of all ages (Berry-Kravis, 2021), and nearly 70% of patients experience cognitive decline at a high enough severity to meet criteria for a dementia diagnosis (Rego et al., 2019). Indeed, NPC is often called childhood Alzheimer’s due to its similarities with the pathology and clinical symptoms of age-related Alzheimer’s disease (AD)‑ (Malnar et al., 2014). However, NPC is an extremely heterogeneous disease resulting from an array of genetic mutations, and significant variation in age of disease onset, types of symptoms, and rate of progression has been observed between patients.

Across patients of all ages, the cellular hallmark of NPC is abnormal cholesterol accumulation within late endosomes and lysosomes and consequent disruption of the endolysosomal pathway (Vanier, 2010). This pathway is also responsible for producing extracellular vesicles (EVs). EVs are membranous nanoparticles that are produced by all cells and can be taken up by distinct recipient cells, allowing them to play important roles in cell-to-cell communication. EV molecular cargo can include proteins, lipids, nucleic acids, and metabolites from the parent cell. EV cargo is influenced by disease state and accordingly, EVs can play significant roles in pathology and serve as biomarkers for both lipid storage disorders (Eitan et al., 2016; Gleason et al., 2021) and neurodegenerative diseases (Rastogi et al., 2021; Soliman et al., 2021). Interestingly, many elements of EV biology are likely impacted by the various cellular disruptions that occur in NPC (Wheeler and Sillence, 2020). However, little is known about how mutations in the NPC proteins affect EVs.

In this review, we will examine both NPC and EV biology and discuss our current understanding of the role of EVs in NPC. Throughout, our focus is to highlight current understanding and opportunities for further research. We note that of the 18 peer-reviewed articles that directly examine EVs in the context of NPC, 13 were published in the last 5 years indicating growing scientific interest and the opportunity to uncover both novel disease biology and therapeutic targets.

NPC is caused by mutations in either the NPC1 gene (95% of cases) or the NPC2 gene (5% of cases), which, respectively, code for the NPC1 and NPC2 proteins (Vanier, 2010). Over 400 pathogenic mutations in NPC1 and NPC2 have been identified (Mavridou et al., 2017). The most common NPC1 mutation (I1061T) is found in 20% of patients, while the remaining 80% have mutations that are often unique or shared among only a handful of other patients (Millat et al., 1999). To add further complexity, NPC patients are often compound heterozygotes, meaning that they inherit a different disease-causing allele from each parent and thus have two different NPC mutations. This genetic variety likely contributes to the observed heterogeneity in patient disease onset and progression, as each mutation impedes the function of the NPC1 or NPC2 proteins to a different extent. The high genetic and clinical heterogeneity is also a key challenge to understanding NPC biology, as studies that focus on a single mutation or use genetic knockouts generate results that may only be relevant to a small proportion of NPC patients. Building in vitro or in vivo models that represent the full spectrum of NPC patients is challenging, and given the rarity of NPC, human studies struggle to obtain truly representative sample sizes.

Both NPC1 and NPC2 are evolutionarily conserved and highly expressed in all human cells (Carstea et al., 1997). NPC1 is a large protein (1,280 amino acids) that resides in the endolysosomal membrane, while the smaller NPC2 protein (130 amino acids) is localized to the endolysosomal lumen (Vanier and Millat, 2004). The role of NPC1 and NPC2 is to transport cholesterol out of endolysosomes (Figure 1). Following endocytosis, cholesterol resides in early endosomes; these organelles then mature into late endosomes and finally merge with lysosomes to form endolysosomes. In the endolysosomal lumen NPC2 binds cholesterol at its sterol-sensing domain. The NPC2/cholesterol complex then binds to NPC1, which has both cholesterol and NPC2 binding domains on its luminal side (Trinh et al., 2018). NPC1 then transports cholesterol across the endolysosomal membrane and into the cytosol, likely by passing the lipid along the sterol-sensing domain embedded in the NPC1 transmembrane domains (Pfeffer, 2019). Mutations in any of the key domains of NPC1 or NPC2 interrupt this pathway, causing endolysosomes to accumulate significant amounts of cholesterol while the rest of the cell enters a state of cholesterol starvation (Bajwa and Azhar, 2021).

Importantly, the connection between disrupted cellular cholesterol trafficking and NPC disease pathology is not fully elucidated, especially regarding alterations in the CNS. The neurobiology of NPC is complex and incompletely understood, but evidence suggests that NPC1 loss in neurons is a key driver of disease pathology (Lopez et al., 2011). Purkinje neurons of the cerebellum are known to die in an anterior-to-posterior pattern as NPC progresses, and neuronal cells of the frontotemporal cortex and other brain regions can also be affected (Rego et al., 2019). An early study provided evidence in an NPC knockout mouse model that Purkinje cell loss did not proceed by an apoptotic pathway that can be inhibited by Bcl-2 or minocycline, a drug that inhibits apparent apoptotic death in other mouse models of neurological disease (Erickson and Bernard, 2002). Why neuronal cells are particularly sensitive to the loss of the NPC proteins is currently unknown; neurons do not express NPC1 or NPC2 at particularly high levels, and patterns of cholesterol accumulation do not correlate with patterns of neuronal death (Martin et al., 2020; Cariati et al., 2021). Some researchers have hypothesized that neurodegeneration in NPC is related to impaired crosstalk between neurons and other cells. Indeed, loss of NPC1 results in inflammatory responses across brain regions, and corresponding glial defects such as hypomyelination, microgliosis, and reactive astrocytosis (Rego et al., 2019). For instance, early studies identify glial cells as a major target for pathology in a mouse model of NPC, and that a lack of myelin within the corpus callosum may be related to the loss of nerve fibers in this structure (German et al., 2002). Further, the finding of an increase in cathepsin-D-laden microglial cells, in brain regions that undergo neurodegeneration, support that microglia contribute to the neurodegenerative process by phagocytosing dead neurons. And the finding that activation of astrocytes in regions that undergo neurodegeneration was consistent with a role for these glial cells in the neurodegenerative process (German et al., 2002). While each of these could impair neuronal health, the exact mechanism linking loss of NPC1 and NPC2 to clinical neurodegenerative symptoms is not well understood.

NPC also causes pathologies with the inner ear, including auditory and balance issues that are common in older NPC patients (Senirli et al., 2016). Early studies to examine patients with NPC1 and hearing loss suggested that multiple steps along the auditory pathway are affected, suggesting that hearing loss is an early and sensitive marker of disease progression (King et al., 2014). One compound that has been evaluated as a treatment for NPC is 2-hydroxypropyl-beta-cyclodextrin (HPβCD). While HPβCD can reduce some debilitating symptoms and extend life span, the therapeutic doses used to treat the disease cause hearing loss by selectively damaging only cochlear outer hair cells (Ding et al., 2021). Earlier studies on the mechanism(s) underlying HPβCD-induced hearing loss found that a prestin-dependent mechanism contributes to HPβCD ototoxicity (Takahashi et al., 2016), that HPβCD-induced outer hair cell death is attributed to the structural role of prestin in maintaining the outer hair cells lateral membrane, rather than its motor function (Zhou et al., 2018), and that cochlear damage by HPβCD is unrelated to reactive oxygen species formation (Lee et al., 2019).

The build-up of endolysosomal cholesterol in NPC also creates a number of other cellular problems that may contribute to disease biology (Wheeler and Sillence, 2020). Cholesterol accumulation severely disrupts the lipid profile of NPC endolysosomes, including a buildup of sphingomyelin, glycosphingolipids, and ceramides that together slow endolysosomal processing and disrupt the organelle’s ability to process cellular waste or transport materials (Newton et al., 2018; Tobias et al., 2019). Mitochondria also become disrupted, showing increased contacts with endolysosomes and resulting in increased reactive oxidative species and cell stress signaling (Vázquez et al., 2012; Torres et al., 2017). Finally the autophagy pathway becomes dysfunctional, causing autophagosomes to accumulate which can eventually result in cell death (Pacheco and Lieberman, 2008). Recently, there has been growing interest in how endolysosomal disruptions in NPC may also affect EVs, which can be produced through the endolysosomal pathway. Though not yet fully understood, emerging evidence suggests that EV concentration and cargo is altered by NPC mutations, making EVs important players in disease pathology.

EVs are small particles that are released from nearly every known cell. Typically 50 to 500 nm in size, EVs are enclosed by a lipid bilayer and contain molecular cargo such as proteins, DNA, coding and noncoding RNA, and metabolites (Thery et al., 2018). While technically accurate, these definitions bely the incredible complexity and opportunity contained in these tiny packages. The cargo carried by EVs reflects their cell of origin but does not perfectly mirror it, and because of this cargo EVs can have dramatic effects when taken up by recipient cells (Sardar Sinha et al., 2018; Upadhya et al., 2020; Van Hoecke et al., 2021a). Although EV science is still an emerging field, EVs have been implicated in nearly every type of human disease, including cancer (Richards et al., 2017; Zhou et al., 2021), infection (Costafreda et al., 2020; Picca et al., 2020), neurodegeneration (Rastogi et al., 2021; Soliman et al., 2021), and, as discussed here, rare diseases such as NPC (Tatiana et al., 2020; Gleason et al., 2021).

Secreted EVs can be found in many biological sources, including cell culture media, biofluids, and tissues (Welsh et al., 2024). EVs collected from patient biofluids can be secreted from any organ and thus have organ-specific effects on diseases. However, it is not always possible to determine the origin of EVs isolated from human biofluids, which complicates interpretation of their cargo. For example, there is great interest in the use of blood EVs as surrogates for cerebrospinal fluid EVs and as biomarkers for neurodegenerative diseases. Yet the results from such studies have not been consistent in their conclusions thus far. For example, L1CAM protein has used for the isolation of brain EVs in blood and the potential diagnosis of AD. While L1CAM protein is highly expressed in brain, it also highly expressed in peripheral organs such as kidney, urinary bladder, and skin.¹ Thus, some studies using multiple methods have not supported L1CAM as a marker to enrich EVs of neuronal origin (Norman et al., 2021; Kadam et al., 2025). In contrast, a recent study provided evidence from single-vesicle analyses that L1CAM immunoprecipitation supported the existence of L1CAM positive EVs in human blood that also carried multiple neuronal markers GAP43, β-III-tubulin and VAMP2, demonstrating that L1CAM can be used as a target for the isolation of neuronal-derived EVs from blood (Nogueras-Ortiz et al., 2024). Thus, the rapid development of combined sensitive methods to assess and validate brain-derived EVs and their cargo in blood provides confidence in the development of non-invasive methods for elucidating the roles of EVs in the mechanisms underlying of brain diseases, and ultimately for the development of non-invasive strategies for the diagnosis and classification of neurodegenerative brain diseases from blood.

The most common methods to separate EVs from other extracellular nanoparticles include ultracentrifugation, size exclusion chromatography, and immunocapture, often in combination, and each method can enrich for slightly different EV populations (Livshits et al., 2015; Konoshenko et al., 2018; Doyle and Wang, 2019; Stam et al., 2021). Methods to analyze EVs may also be biased towards certain EV subpopulations; for example, size-based methods may not be able to detect the smallest EV populations, and methods that include labelling the EV membrane may preferentially dye EVs with certain lipid compositions. Thus, when comparing EV studies, it is important to consider the method used to prepare the EV samples and downstream analytics.

The most commonly studied EVs include exosomes, which are formed through the endolysosomal pathway, and microvesicles (also termed ectosomes) that bud from the plasma membrane. Exosome biogenesis begins when the membrane of late endosomes curve inward to form small vesicles in the endosome lumen, termed intraluminal vesicles (ILVs) (Figure 2A) (Gurung et al., 2021). Due to their appearance in electron microscopy, late endosomes containing ILVs are termed multivesicular bodies (MVBs). After ILV formation, an MVB fuses with the plasma membrane wherein the ILVs are released into the extracellular space as exosomes. Exosome biogenesis can be driven by a family of proteins known as the endosomal sorting complex required for transport (ESCRT). ESCRT-independent pathways have also been observed, such as the conversion of sphingomyelin to ceramide by the neutral sphingomyelinase 2 (nSMase2) protein (Trajkovic et al., 2008).

Microvesicles are formed by outward budding of the plasma membrane (Figure 2B). Microvesicle biogenesis pathways involve complex rearrangement of both cytoskeletal and membrane components, facilitated in part by a local increase in calcium ions and proteolytic enzymes that disassemble the cytoskeleton (Kalra et al., 2016). Membrane curvature is facilitated by lipid-anchor modifications on proteins (e.g., myristylation or palmitoylation), and the GTPase ADP-ribosylation factor (ARF) is required for secretion (Yu et al., 2024). Some proteins involved in exosome biogenesis are also implicated in microvesicle biogenesis, such as TSG101 and ALIX. Although they overlap in size, it is generally thought that exosomes are smaller (50–120 nm) than microvesicles (70–200 m).

One of the most interesting aspects of EVs is their molecular cargo, such as proteins, miRNAs, and lipids, which can be enriched disproportionate to what is found in their parent cell (Jeppesen et al., 2019; Couch et al., 2021; Gurunathan et al., 2021). This suggests that EVs are mechanistically packaged with specific molecules, either to remove them from the parent cell or transfer them to a recipient cell. Whether EV cargo is randomly integrated, intentionally curated, or some mix of both, is unknown. Yet in either case, EV cargo offers useful insight into the state of the parent cell. Cellular perturbations, either through disease or experimental manipulation, are known to alter EV cargo. Further, the functional effects that EVs have on recipient cells can be altered based on changes to the parent cell. This suggests that, random or not, EV cargo has important implications for parent cell biology and cell-to-cell communication, especially in the context of disease.

At the time of this writing, we are aware of only 18 peer-reviewed publications that have directly examined EVs in the context of NPC (Table 1). Sixteen of these studies analyzed EVs in response to NPC1 mutations, while only two examined NPC2 mutations. Four studies were performed in vivo, using either the NPC1^m1N/J or NPC1^nmf174/J mouse. The remaining experiments were performed using in vitro models, including immortalized cell lines, primary murine cells, or patient-derived fibroblasts. Many studies used genetic models, while others treated cells with U18666A (U18), a small molecule inhibitor that specifically inactivates the NPC1 sterol-sensing domain and reproduces NPC endolysosomal cholesterol accumulation (Lu et al., 2015). Three studies used cells that mimic neurological phenotypes, including oligodendrocyte precursors (Oli-neu), rat primary astrocytes, and rat primary aged cortical neurons. Only four examined EVs in NPC human samples, each of which relied on fibroblasts derived from a single patient. Given the high degree of heterogeneity among NPC patients, studies on larger cohorts of NPC human samples are needed.

The existing studies on EVs in NPC have primarily focused on the mechanistic role of EVs in NPC pathways, or on how NPC disease affects EV production. It is important to consider that protocols for EV enrichment and analysis vary widely between publications, and results must be interpreted within their methodological context. Despite these limitations, there is growing evidence that the NPC proteins play direct roles in EV biology; that EV concentration and cargo is altered by NPC pathology; and that EV packaging and release may be one method by which NPC attempts to remove cholesterol from endolysosomes in cells. To further understand how EVs contribute to or are affected by NPC cellular disruptions, these studies must be carefully considered and built upon.

Current studies suggest that EVs are altered in NPC, and that these changes are relevant to NPC pathology. Though limited, the existing literature raises several interesting questions about the relationship between NPC and EVs. The conflicting evidence on EV concentration suggests that the ways in which the NPC proteins affect EV formation and release are complex, and may be dependent on cell type, model system, and the subspecies of EVs analyzed (exosomes, microvesicles, etc.). Of particular relevance is whether there are CNS-specific changes to EVs in NPC which may help explain the neurodegenerative aspects of the disease. It is also interesting to note that loss of NPC1 protein function generates EVs that carry protein and lipid content related to NPC disease pathology. This could suggest that EVs are a useful method for NPC mutant cells to export harmful material; however, as EVs are often taken up by recipient cells, this may also implicate EVs in the spread pathologic molecules throughout the body.

Critical gaps in the current literature also limit our understanding of EVs in NPC. For example, there is not yet any information on how EV miRNAs are altered in NPC, and full proteomic analyses have not yet been conducted. As with other aspects of NPC the majority of studies have focused on the impact of NPC1 loss, and although the NPC2 protein plays known roles in EV biogenesis, no studies to date have examined how NPC2 mutations affect EV concentration or cargo. Additionally, the lack of studies in human samples makes it difficult to determine how EV alterations impact NPC patients. As no studies to-date have examined EVs collected from patient biofluids, it is currently unknown how EVs from differently impacted organs in NPC—such as the liver, brain, or inner ear—may contribute to disease pathology. Patient heterogeneity must also be considered. Many of the current EV NPC studies utilized in vitro systems with full knockdown of the NPC1 protein, either by genetic alterations or drug treatment. However, the degree of protein dysfunction in NPC patients varies widely, and full loss of NPC1 or NPC2 protein expression is less common in pediatric or adult patients. How EVs are affected by incomplete loss NPC1 or NPC2 is unknown, and it is possible that this too varies by individual.

Finally, the molecular connection between loss of NPC1/NPC2 protein function and EV alterations is also currently unknown. Some research suggests that NPC1 and NPC2 play direct roles in EV release and uptake yet is also possible that EVs are impacted by other cellular changes caused by NPC1/NPC2 mutations. For example, endolysosomal dysfunction, oxidative stress, and altered autophagic flux are known to occur in NPC (Wheeler and Sillence, 2020), and evidence from other disease states suggest that these cellular perturbations impact EV biology (Eitan et al., 2016; Chiaradia et al., 2021; Ai et al., 2024). Conversely, discoveries about the impact of NPC cellular pathology on EVs may shed light on other diseases that share similar mechanisms. Indeed, outside of NPC, there is a growing appreciation of how endolysosomal injury affects EV production. Thus, further research into NPC and EVs will contribute to our growing understanding of EV cellular biology and reveal novel information about a childhood disease that currently has no cure.