Modeling Sialidosis with Neural Precursor Cells Derived from Patient-Derived Induced Pluripotent Stem Cells

Sialidosis, caused by a genetic deficiency of the lysosomal sialidase gene (NEU1), is a systemic disease involving various tissues and organs, including the nervous system. Understanding the neurological dysfunction and pathology associated with sialidosis remains a challenge, partially due to the lack of a human model system. In this study, we have generated two types of induced pluripotent stem cells (iPSCs) with sialidosis-specific NEU1G227R and NEU1V275A/R347Q mutations (sialidosis-iPSCs), and further differentiated them into neural precursor cells (iNPCs). Characterization of NEU1G227R- and NEU1V275A/R347Q- mutated iNPCs derived from sialidosis-iPSCs (sialidosis-iNPCs) validated that sialidosis-iNPCs faithfully recapitulate key disease-specific phenotypes, including reduced NEU1 activity and impaired lysosomal and autophagic function. In particular, these cells showed defective differentiation into oligodendrocytes and astrocytes, while their neuronal differentiation was not notably affected. Importantly, we found that the phenotypic defects of sialidosis-iNPCs, such as impaired differentiation capacity, could be effectively rescued by the induction of autophagy with rapamycin. Our results demonstrate the first use of a sialidosis-iNPC model with NEU1G227R- and NEU1V275A/R347Q- mutation(s) to study the neurological defects of sialidosis, particularly those related to a defective autophagy–lysosome pathway, and may help accelerate the development of new drugs and therapeutics to combat sialidosis and other LSDs.


Introduction
Sialidosis is a lysosomal storage disease (LSD) characterized by an abnormal accumulation of sialylated oligosaccharides and glycolipids resulting from NEU1 loss of function [1,2]. Sialidosis-causing mutations in the NEU1 gene, located on chromosome 6p21.3, eliminate the activity or stability of lysosomal sialidase (also known as neuraminidase 1; EC 3.2.1.18), catalyzing the hydrolysis of terminal sialic acid molecules (N-acetylneuraminic acid or NANA) linked to the saccharide chains of glycolipids, glycoproteins, and oligoand polysaccharides [1][2][3][4]. Lysosomal sialidase forms a multienzyme complex with other lysosomal hydrolases, such as protective protein/cathepsin A (PPCA) and glycosidase β-galactosidase (β-GAL), to promote their canonical degradative activities [5,6]. NEU1 mutations can cause disruption of the multienzyme complex, ultimately resulting in loss of enzymatic activity [5,7]. None of the other three sialidases, which are distributed in the cytosol (NEU2 neuramidase 2), plasma membrane (NEU3 neuramidase 3), or mitochondrial/lysosomal/intracellular membranes (NEU4 neuramidase 4), appears to compensate for the deficiency of lysosomal sialidase in sialidosis patients [8]. Genetic testing of sialidosis patients has revealed various pathological variants of the NEU1 gene that are related to a broad range of neurophysiological abnormalities, including visual defects, myoclonus syndrome, cherry-red macular spots, ataxia, hyperreflexia, and/or seizures [9,10]. The clinical symptoms, severity, and penetrance of sialidosis appear to correlate with the types of NEU1 mutations, which closely reflect the overall biochemical properties of sialidase, including the levels of residual sialidase activity.
Animal models harboring disease-causing mutations are a crucial tool for identifying and determining the common molecular mechanisms and biochemical impact of NEU1 deficiency on sialidosis-related phenotypes and pathogenesis. Preclinical research conducted in Neu1 mutant mouse models revealed defects in virtually all systemic organs, including the nervous system and connective tissues that are similar or consistent with the clinical features of patients with sialidosis [11][12][13][14][15]. A wide range of pathological phenotypes observed in sialidosis mouse models appear to be shared across multiple LSDs [12,13,15]. Homozygous Neu1−/− knockout mouse models display unique pathological phenotypes of sialidosis, but also share some clinical and histopathological features with galactosialidosis caused by a primary defect of the PPCA gene (chromosomal locus, 20q13.1) [12]. The brains of Neu1−/− mouse models were shown to exhibit features of Alzheimer's disease (AD)-related pathology and phenotypes, including increased β-amyloid plaque formation and secretion [13]. Although animal models are very valuable, major disadvantages of these models, including the inability to explicitly define and delineate sialidosis-specific pathological traits from those of other LSDs and the inability to fully recapitulate equivalent human phenotypes, remain. To correctly interpret the possible pathological and clinical contribution of NEU1 variants observed in sialidosis patients and to develop effective diagnostic and therapeutic strategies, large numbers of highly representative human mutant models are required.
Induced pluripotent stem cell (iPSC) technology, providing a variety of previously inaccessible disease-relevant target cells, opens up a new realm of possibilities for the development of in vitro cell-based disease models. Development of neural lineage cells from disease-specific iPSCs, particularly those derived from neuropathic LSD patients, has contributed to a greater understanding of the neuropathophysiology of LSDs, yielding an effective cellular platform for identifying safer and more effective therapeutic targets [16][17][18][19][20]. Likewise, the development of sialidosis-and tissue-specific cells from iPSCs with sialidosisassociated mutations also offer emerging opportunities to investigate sialidosis at the cellular and molecular level, but they are less studied than other LSDs [21][22][23][24]. However, there have been a limited number of modeling studies for sialidosis. A recent study using sialidosis neurons from two sialidosis-iPSCs with NEU1 mutations either of p.D176G or of p.P316S revealed functional and molecular abnormalities such as defective presynaptic exocytosis and excessive enhancement of α-amino-3-hydroxyl-5-methyl-4-isoxazolepropionate receptor (AMPAR)-evoked Ca 2+ influx in sialidosis patients [24]. Nevertheless, the neuropathological features characteristic of sialidosis, involving primary defects in the lysosomal and autophagic machinery, have not been fully elucidated across different NEU1 mutation types. Large numbers of neurological cell models, covering various types of heterogeneous pathological mutations in NEU1 gene, are highly desirable to clearly define the neuropathological defects of sialidosis caused by NEU1 mutation and to develop novel therapeutic strategies.
Previously, we found the novel NEU1 mutation (p.R347Q) in sialidosis patient fibroblasts showing sialidosis-associated phenotypes [25]. In this study, we aim to establish new sialidosis-specific neural models by producing sialidosis-specific iPSCs and NPCs with NEU1 mutations (p.V275A/R347Q and p.G227R) to investigate the effects of NEU1 deficiency on autophagy-lysosome pathways and their relevant contribution to neurological defects. Here, we show that sialidosis-iNPCs carried NEU1 G227R or NEU1 V275A/R347Q mutations that closely recapitulate disease-specific phenotypes, such as reduced sialidase activity and autophagy and lysosome dysfunction. Of note, we show that rapamycin treatment effectively rescues sialidosis-related phenotypes such as autophagy and lysosome dysfunction and defective NPC and oligodendrocyte and astrocyte differentiation. We believe that our iPSC models for sialidosis would aid in the elucidation of the molecular pathogenesis as well as in drug discoveries for sialidosis.
Immunofluorescence analysis of the sialidosis-iNPCs with an anti-NEU1 antibody showed a similar punctate endo-/lysosomal staining pattern as seen in the normal-iNPCs, while the expression of NEU1 in the sialidosis-iNPCs (G227R-iNPCs and V275A/R347Q-iNPCs) had a more perinuclear distribution compared with that of the normal-iNPCs ( Figure 3A). Notably, lower protein levels of NEU1 were detected in the sialidosis-iNPCs (G227R-iNPCs and V275A/R347Q-iNPCs) compared to normal-iNPCs ( Figure 3B). Accordingly, we observed decreased sialidase activity in the sialidosis-iNPCs [approximately 10% (G227R-NPCs) and 16% (V275A/R347Q-NPCs) of normal control] ( Figure 3C). The activity and stability of Neu1 is known to be exclusively influenced by association with other lysosomal enzymes such as protective protein/cathepsin A (PPCA) and β-galactosidase [5,6]. Western blot analysis also showed that the protein expression levels of PPCA were not notably different between the normal-and sialidosis-iNPCs ( Figure 3B), suggesting that NEU1 G227R and NEU1 V275A/R347Q mutation-mediated loss of NEU1 activity is not correlated with PPCA expression.

NEU1 Deficiency-Associated Autophagy-Lysosomal Dysfunction in Sialidosis-iNPCs
Sialidase NEU1, a known negative regulator of lysosomal exocytosis, is responsible for the processing of sialic acids on lysosomal-associated membrane protein-1 (LAMP-1), a lysosomal exocytosis effector [26]. Similar to other NEU1-deficient models [26,27], the expression of LAMP1 was markedly increased in the sialidosis-iNPCs (G227R-iNPCs and V275A/R347Q-iNPCs), compared with the normal-iNPCs ( Figure 3B). These results suggest that NEU1 deficiency due to NEU1 G227R and V275A/R347Q mutations may result in impaired lysosomal exocytosis in sialidosis-iNPCs and also demonstrate that the features of neuronal and biochemical defects associated with NEU1 G227R and NEU1 V275A/R347Q mutations can be effectively recapitulated in our sialidosis-iNPC models.

Rapamycin Partly Restores Impaired Function of Sialidosis-iNPCs
As the sialidosis-iPSCs, relative to the normal-iPSCs, were found to be less efficient in the production of NPCs ( Figure 5A,B), we next evaluated whether the differentiation capability of the sialidosis-iPSCs into NPCs could be restored by activation of autophagy with rapamycin. Notably, rapamycin treatment effectively increased the production yields of the sialidosis-iNPCs, as determined by increases in spherical NPC size and numbers ( Figure 5A,B). While the normal-and sialidosis-iNPCs did not differ in their capabilities to differentiate into neurons, the sialidosis-iNPCs showed reduced differentiation towards oligodendrocytes and astrocytes ( Figure 5C,D). Significantly, the defective differentiation potential of the sialidosis-iNPCs, especially into oligodendrocytes and astrocytes, was notably rescued by rapamycin treatment (Figure 5C,D). These observations demonstrate that dysfunction of sialidosis-iPSCs and -iNPCs can be rescued, at least in part, by autophagy activation. In addition, the reversal of the phenotypic defects of sialidosis-iNPCs with rapamycin can provide future opportunities for therapeutic strategies.  Figure 4A images to measure the relative expression of fluorescence intensity. The data are the means ± SDs of at least 3 independent experiments. * p < 0.05, ** p < 0.01, and *** p < 0.001, compared to control normal-iNPCs (t-test).  Figure 5C images to measure the relative expression of fluorescence intensity. The data are the means ± SDs of at least 3 independent experiments. * p < 0.05, ** p < 0.01, and *** p < 0.001, compared to control normal cell line (t-test).

Discussion
Here, we present new sialidosis-iNPC models carrying NEU1 G227R or NEU1 V275A/R347Q mutations generated from the fibroblasts of two independent patients with p.G227R (c.677G > A) and p.V275A (c.824T > C)/p.R347Q (c.1040G > A) mutations [25,32]. All NEU1 G227R -or NEU1 V275A/R347Q -mutated iNPC models showed NEU1 deficiency-mediated autophagy and lysosomal dysfunction and loss of differentiation capacity, which were closely related to neurological dysfunction that occurred because of sialidosis ( Figure 6). Figure 6. Schematic summary of sialidosis-specific neurological impairments characterized in sialidosis-iNPCs. The sialidosis-iNPC model reveals molecular defects in sialidosis, such as NEU1 mutation, deficient sialidase activity, and increased LAMP1 expression. Deficient NEU1 enzyme activity caused by genetic mutations in the NEU1 gene results in impaired autophagy and lysosomal pathways in sialidosis-iNPCs. Induction of autophagy by rapamycin effectively rescues sialidosis-related defects in sialidosis-iNPCs, suggesting an important role of the autophagy-lysosome pathway in neurological abnormalities associated with sialidosis.
Similarly to sialidosis-NPCs carrying NEU1 D176G or NEU1 P316S [21], our sialidosis-iNPC models carrying NEU1 G227R or NEU1 V275A/R347Q mutations revealed correlations of NEU1 mutations with sialidosis-specific molecular defects and disease phenotypes, which are specifically associated with abnormal lysosomal dynamics. NEU1 G227R -and NEU1 V275A/R347Q -iNPCs showed 10% and 29% residual activity respectively, compared to the normal control cells. These results correspond to those determined in the parental patient fibroblasts [25] and other pathological mutations identified in the NEU1 gene [4,24,32]. The expression levels of PPCA, a major constituent of the multienzyme lysosomal complex, were not notably changed in NEU1 G227R -and NEU1 V275A/R347Q -iNPCs compared to normal control cells. However, there is the possibility that NEU1 G227R and NEU1 V275A/R347Q mutations may affect the NEU1-PPCA interaction essential for the correct trafficking of NEU1 to lysosomes, where the NEU1 enzyme is processed into its mature, active form [7,43,45]. As NEU1 acts as a negative regulator of lysosomal exocytosis mediating cellular clearance by processing its substrate LAMP1 sialic acids, NEU1 deficiency-associated accumulation of over-sialylated LAMP1 at the lysosomal membrane appears to be responsible for excessive lysosomal exocytosis [13,26]. Consistent with previous observations in sialidosis-NPCs carrying NEU1 D176G or NEU1 P316S [21], loss of NEU1 enzyme activity in NEU1 G227R -and NEU1 V275A/R347Q -iNPCs is coupled with increased expression of LAMP1, which is potentially suggestive of impaired lysosomal exocytosis in NEU1 G227Rand NEU1 V275A/R347Q -iNPCs, but the detailed molecular machinery and regulatory mechanisms require further investigation.
The role of autophagy malfunction has been extended to LSDs [28,29] but has not been clearly investigated in sialidosis in terms of neurological dysfunction. Our observations provide the first evidence that NEU1 deficiency alters not only lysosomal dynamics but also autophagic activity in NEU1 G227R -and NEU1 V275A/R347Q -iNPCs, whereas we could not exclude the possibility of retrograde trafficking pathway involvement ( Figure 6) [46]. Although NEU1 G227R -and NEU1 V275A/R347Q -iNPCs were found to be less sensitive to rapamycin than normal-iNPCs, rapamycin treatment effectively recovered autophagic flux in NEU1 G227R -and NEU1 V275A/R347Q -iNPCs. Significantly, autophagy activation with rapamycin effectively ameliorated the accumulation of enlarged lysosomes in NEU1 G227Rand NEU1 V275A/R347Q -iNPCs, suggesting a connection between impaired autophagic flux and lysosomal dysfunction. Importantly, the defective differentiation potential of both NEU1 G227R -and NEU1 V275A/R347Q -iPSCs into iNPCs and NEU1 G227R -and NEU1 V275A/R347Q -iNSCs into oligodendrocytes and astrocytes was effectively rescued by rapamycin treatment, demonstrating the potential of targeting these glial cells for therapeutic intervention. Consistently, many studies indicate the contribution of impaired astrocytes, triggered by autophagic/lysosomal dysfunction, to driving neuronal dysfunction in other LSDs, such Alzheimer's disease, Parkinson's disease, and Huntington disease, and highlight the potential of targeting these cells as a beneficial therapeutic strategy for LSDs [47][48][49][50]. However, it still remains unexplored in the context of sialidosis neuropathogenesis. Moving forward, it would be important to address whether and how glial cells play a role on sialidosis progression and pathobiology and to determine their impact on neuronal circuits and networks.
In pursuit of more accurate and reproducible modeling of sialidosis, multiple iPSC models, covering various types of heterogeneous pathological mutations in NEU1 gene, a variety of donors, and differentiation variation, are highly desirable for comparative studies, but still are not available. Our findings show that sialidosis-iNPCs with G227R or V275A mutations share some common pathological and biochemical features with other mutations types of sialidosis, and the utility of our NEU1 G227R -and NEU1 V275A/R347Q -iNPCs can be expanded to comparative and interdisciplinary research with other sialidosis models to better understand phenotype-genotype correlations and common and distinct pathological mechanisms. Recently, gene editing technology, such as the easily accessible CRISPR/Cas9 system, has been applied for therapeutic purposes and disease modeling of LSDs by correcting or inducing disease-causing mutations [51]. Together with patientderived sialidosis-iPSC models, with the use of CRISPR gene-edited sialidosis-iPSC models, will allow us to gain a deeper understanding of the molecular mechanisms underlying the impact of disease-causing mutations, and new therapeutic approaches.
In conclusion, our new NEU1 G227R -and NEU1 V275A/R347Q -iNPC models described here closely mimic the altered autophagy and lysosomal dysfunction associated with neurological dysfunction in sialidosis. The findings of this study also highlight the importance of understanding the mechanisms that control the autophagy-lysosome pathways for the development of effective therapies for sialidosis.

cDNA Sequencing
Total RNA was extracted from fibroblasts, iPSCs, and iNPCs using RNiso Plus reagent (Takara, Shiga, Japan) and was reverse transcribed using the Superscript III first-strand synthesis system (Invitrogen) according to the manufacturers' instructions. Full-length NEU1 cDNAs were amplified by PCR and sequenced as previously described [25].

Alkaline Phosphatase (ALP) Analysis
Normal-and sialidosis-iPSC colonies were fixed in a citrate-acetone-formaldehyde solution and stained for ALP activity using an ALP detection kit (Sigma). Cell images were captured using an Olympus microscope (IX51, Olympus, Shinjuku-ku, Tokyo, Japan).

In Vitro EB Formation and Trilineage Differentiation
The iPSC colonies cultured in feeder-free conditions were dissociated with 1 mg/mL collagenase IV. Embryoid body (EB) formation was carried out in suspension culture by culturing of iPSC aggregates with EB medium consisting of knockout Dulbecco's modified Eagle's medium (DMEM, Thermo Fisher Scientific), 10% KnockOut serum replacement (KSR), 1× NEAA, 0.1 mM β-mercaptoethanol, and 1 mM L-glutamine in ultralow attachment plates (Corning, Corning, NY, USA) for 5 days. For spontaneous trilineage differentiation, the EB aggregates were transferred to gelatin-coated plates, cultured for an additional 10 days, and analyzed for trilineage marker expression by immunocytochemistry. The medium was changed every other day. The antibodies utilized are summarized in Supplementary Table S2.

G-Banding Karyotyping and Short Tandem Repeat (STR) Profiling
To monitor genomic stability, after more than 20 passages in feeder-free medium, iPSCs were processed for standard chromosomal G-band analysis by GenDix, Inc. (Seoul, Korea). A representative image was obtained by ChIPS-Karyo (Chromosome Image Processing System, GenDix, Inc.). STR profiling of the original sialidosis patient cells (6 passages) and iPSCs (15 passages) was performed by HumanPass, Inc. (Seoul, Korea) to confirm that all the iPSCs had the same origin as patient fibroblasts. The absence of mycoplasma contamination was regularly checked by a polymerase chain reaction (PCR)-based method using an EZ-PCR TM Mycoplasma detection kit (Biological Industries, Beit-Haemek, Israel).

Immunocytochemistry
Cells were fixed with 4% formaldehyde (Sigma) for 10 min, permeabilized with 0.1% Triton X-100 for 30 min, and blocked with 3% bovine serum albumin (BSA) in phosphate-buffered saline (PBS) for 2 h. The cells were then incubated overnight at 4 • C with primary antibodies diluted in blocking solution. After washing three times with 0.1% Tween-20-containing PBS at room temperature (RT), the cells were further incubated with fluorescence-conjugated secondary antibodies for 2 h and then counterstained with 300 nM 4 , 6-diamidino-2-phenylindole (DAPI, Invitrogen) at RT for 15-30 min. The images were obtained by using an Axio VertA.1 microscope (Carl Zeiss, Oberkochen, Germany) or an Olympus microscope (Olympus, Japan). The antibodies utilized are summarized in Supplementary Table S2.

Western Blot
Cell lysates (20 µg protein) prepared in RIPA buffer containing 50 mM Tris-HCl, pH 8.0, 150 mM sodium chloride, 1% Triton X-100, 0.5% deoxycholic acid, and 0.1% SDS were separated by 4-20% SDS-PAGE and blotted onto PVDF membranes. After blocking in blocking buffer containing 5% skim milk and 0.1% Tween-20 in Tri-buffered saline (TBST) for 2 h at RT, the blots were washed three times for 5 min with TBST and incubated with primary antibodies diluted in 3% skim milk in TBST. After washing the membrane three times with TBST for 5 min, the horseradish peroxidase (HRP)-conjugated secondary antibodies against each primary antibody species was diluted in 3% skim milk in TBST and incubated for 2 h at RT. The bands were detected using an ECL substrate kit (Thermo Fisher Scientific) and normalized to the density of the loading control, β-actin. The antibodies utilized are summarized in Supplementary Table S2.

Autophagy and Lysosome Staining
The iNPCs were seeded on 18-mm cover-glass bottom dishes and untreated or treated with 500 ng/mL rapamycin for overnight. After the treatments, the cells were stained with a CYTO-ID ® autophagy detection kit (Enzo, Farmingdale, NY, USA) and LYSO-ID ® Red detection kit (ENZ-51005-0100, Enzo) according to the manufacturer's instructions. Images were acquired by an Axio VertA.1 microscope (Carl Zeiss).

Statistical Analysis
The results are presented as the mean ± SE. Student's unpaired t-test was used for statistical evaluation, and * p < 0.05, ** p < 0.01, and *** p < 0.001 were considered significant. Image results were quantified by ImageJ software (NIH, Bethesda, MD, USA).