Inhibition of PI4KIIIα as a Novel Potential Approach for Gaucher Disease Treatment

Dear Editor,

Gaucher disease (GD) is the most common lysosomal storage disease (LSD) caused by an insufficiency of the lysosomal enzyme glucocerebrosidase (GCase) [1]. GCase insufficiency produces the excessive lysosomal accumulation of unmetabolized glycolipid substrates including glucosylceramide (GlcCer), leading to the disruption of the structure and function of tissues and organs, including the blood system, viscera, brain, bones, and cartilage. GD was initially classified into 1–3 types based on the variation of severity and progression in neuropathic manifestation, which is early onset, most severe and acute in type 2, later onset and chronic in type 3, and not obvious in type 1. However, accumulating evidence has shown that type 1 GD is clearly associated with Parkinson’s disease and related synucleinopathies [1]. Currently, there are multiple available ways of treating GD patients, including enzyme replacement therapy, substrate reduction therapy, and pharmacological chaperone therapy [2]. However, the therapeutic effect of these approaches used either alone or in combination in GD patients is unsatisfactory. New drug candidates have been studied for a long time in order to develop promising therapeutic approaches to benefit GD patients, especially those with pathological features in the central nervous system.

The pathogenesis of GD involves autophagy-lysosome pathway (ALP) defect [3]; besides, hyperactivity of the mammalian target of rapamycin complex 1 (mTORC1) has been found in GD neurons derived from induced pluripotent stem cells (iPSC), and pharmacological inhibition of glucosylceramide synthase enzyme to reduce the accumulation of unmetabolized substrates of GCase reverses mTORC1 hyperactivity [4]. Moreover, lysosomal accumulation of unmetabolized glycolipid substrates produces overload and dysfunction of lysosomes, reduces the efficiency of lysosomal digestion and recycling as well as the efficiency of lysosomal fusion with autophagosomes, affects the upstream of ALP, and even may prevent lysosome reformation, resulting in further impairment of the clearance or recycling of metabolic wastes and forming a vicious cycle. Indeed, more and more defects in the ALP have been uncovered in the cells of GD patients and animal models as well as GD cell models. Besides digesting or recycling cellular wastes, lysosomes in virtually all cell types can undergo exocytosis or secretion as a secretory compartment to release intraluminal content and exchange membrane components with the plasma membrane (PM). And facilitation of lysosomal exocytosis can ameliorate the excessive lysosomal accumulation of unmetabolized cellular materials in many LSDs with neuropathological changes and neural symptoms [5].

Phosphatidylinositols and their metabolizing enzymes are involved in a variety of membrane trafficking between different cellular compartments in all cell types, particularly, phosphatidylinositol-4 phosphate (PI4P) is involved in autophagolysosome formation. Four types of phosphatidylinositol-4 kinase (PI4K) generate PI4P at different compartments: PI4KIIα, PI4KIIβ, and PI4KIIIβ produce PI4P in the Golgi apparatus, whereas PI4KIIIα generates PI4P and tightly controls its level at the PM [6, 7]. Of particular relevance here, we previously reported that phenylarsine oxide (PAO), an inhibitor of PI4KIIIα [8], facilitates the exocytosis of Lyso-Tracker-labeled granules and the release of ATP from cultured microglia [9], and facilitates the cellular secretion of Aβ₄₂ from Aβ₄₂- or APP-expressing cells or tissue in association with a reduction of intraneuronal accumulation of Aβ₄₂, indicating that the inhibition of PI4KIIIα is a potential solution for the cellular accumulation of lysosomes with unmetabolized substrates in cells.

In this study, we investigated the effects of chemical and genetic inhibition of PI4KIIIα on the impairment of ALP and cellular accumulation of glycolipids in SH-SY5Y cells, which were treated with Conduritol B epoxide (CBE), an irreversible inhibitor of GCase [10]. We found that inhibition of PI4KIIIα had therapeutic effects on CBE-treated cells, including activation of the ALP, and the amelioration of cell death, cellular accumulation of glycolipids, and overloading of lysosomes, indicating that PI4KIIIα may be a potential therapeutic target and its inhibitors be therapeutic chemicals for treating GD, and presumably for other lyososomal storage disorders and other neurodegenerative diseases.

CBE is frequently applied to cultured cells and animals for generating in vitro and in vivo models of GD. To reveal the proper concentration of CBE for studying the effect of GCase insufficiency-induced impairment in neuronal cells, we examined the viability of SH-SY5Y cells treated at different concentrations of CBE for 48 h. As shown in Fig. 1A, CBE reduced the cell viability in a dosage-dependent manner, and the reduction became significant when the concentration of CBE was at and above 100 μmol/L. Then 100 μmol/L was chosen for inhibiting GCase in the following experiments.

Fig. 1.

PAO protects SH-SY5Y cells from CBE-induced cell death and reduces the pathological accumulation of glycolipids and lysosomes in CBE-treated SH-SY5Y cells. A The viability of SH-SY5Y cells treated with CBE at various concentrations. Data are normalized to control, n = 5 per group, **P <0.001, ***P <0.0001 versus control. One-way ANOVA followed by Dunnett’s test, if not specifically indicated. B The effect of PAO at different concentrations on the CBE-induced reduction of viability in SH-SY5Y cells. Data are normalized to control. n = 5 per group, ^###P <0.0001 versus control, **P <0.001 and ***P <0.0001 versus 100 μmol/L CBE-treated alone in B. C Representative DAPI and PI staining for the nucleus and cell death respectively. Scale bar, 100 μm; n = 3 per group. D Normalized PI staining. n = 3 per group. ^###P <0.0001 versus control, *P <0.05, **P <0.001, ***P <0.0001 versus 100 μmol/L CBE-treated alone. E, F Representative images and normalization of Lyso-Tracker labeling in control and CBE-treated SH-SY5Y cells with or without 10 min PAO treatment. n = 5, ^##P <0.001 versus control, **P <0.001 and ***P <0.0001 versus 100 μmol/L CBE-treated alone. Scale bars, 50 μm. G, H The levels of GlcCer species from cell lysis (G) and culture medium (H) measured by LC-MS/MS. n = 3. Data were analyzed with two-way ANOVA with Turkey’s multiple comparison. Data are presented as the mean ± SEM.

Next, we examined the effect of PAO at different concentrations on the viability of CBE-treated SH-SY5Y cells. After CBE treatment for 24 h and followed by starvation treatment [removal of fetal bovine serum (FBS) from the culture medium] for 24 h, PAO was applied to the culture medium at different concentrations and co-incubated with CBE for 24 h. Then, cells were fixed for the examination of viability with MTT assays. PAO at 25 nmol/L and above increased the viability (Fig. 1B). To further determine the protective effect of PAO, we examined the permeability to propidium iodide (PI) in CBE-treated cells with and without PAO treatment. Consistently, the ratios of PI-positive cells in 25 nmol/L to 75 nmol/L PAO-treated groups significantly decreased when compared to the non-treated group (Fig. 1C, D). These data demonstrated that PAO protects neuronal cells from GCase insufficiency-induced death.

The cells with GCase deficiency typically accumulate unmetabolized glycolipids and lysosomes [1]. To further investigate the therapeutic effects of PAO on CBE-treated SH-SY5Y cells, we incubated control and CBE-treated SH-SY5Y cells with Lyso-Tracker Red DND-99 for 30 min, then the culture medium was replaced with fresh medium containing PAO at the concentrations of 0, 50, or 75 nmol/L for 10 min. In comparison with control, 100 μmol/L CBE treatment dramatically increased the Lyso-Tracker signal in cells, indicating accumulation or enlargement of lysosomes, and PAO at both concentrations greatly restored this alteration (Fig. 1E, F). We next tested whether this effect of PAO was associated with a corresponding change in the cellular accumulation of glycolipids. We measured the levels of a series of GlcCer species in cells and the culture medium by LC-MS/MS. CBE treatment for 48 h remarkably increased the level of each GlcCer in cells, with a concomitant decrease in the level of each GlcCer in the culture medium (Fig. 1G, H). Application of PAO at 50 nmol/L for 24 h reversed the CBE-induced change of glycolipids in both cells and extracellular spaces (Fig. 1G, H). These results indicated that PAO removed the cellular accumulation of glycolipids and lysosomes, and restored the morphological change in lysosomes in GD cell models by facilitating the lysosome-mediated secretion of glycolipids, confirming the previous report that PAO facilitates lysosomal exocytosis [9] and the facilitation of lysosomal exocytosis can promote clearance of unmetabolized substrates and restore the pathological changes in LSD cells [5].

To test whether the effect of PAO on lysosomal alteration in CBE-treated cells is associated with a restoration of autophagic flux, we incubated control and CBE-treated SH-SY5Y cells with PAO at different concentrations or rapamycin for 24 h, and then analyzed the most widely used autophagy biomarkers LC3B and p62 by either western blot or immunostaining. The representative immunoblots and normalized quantifications of LC3B and p62 are shown in Fig. 2A–C. Indeed, application of PAO remarkably increased the level of LC3B with a concomitant dramatic decrease in p62 in a dosage-dependent manner, which was comparable to the effect of rapamycin (the well-known activator of ALP) on both proteins. Consistently, similar effects of PAO on L3CB and p62 in CBE-treated cells were visualized by immunostaining (Fig. 2D). These results indicate that PAO did restore the autophagic flux in CBE-treated cells.

Fig. 2.

PAO activates ALP, the protective effect of PAO is blocked by Baf-A1, and knockdown of PI4Ka reduces the accumulation of lysosomes in CBE-treated SH-SY5Y cells. A–C Representative immunoblots (A) and quantification (B and C) of the expression of LC3B and p62 in cells under different conditions. n = 5, ^#P <0.05 versus control, *P <0.05, **P <0.001 versus 100 μmol/L CBE-treated alone. D Representative images of LC3B and p62 immunostaining in cells under different conditions. Scale bar, 50 μm. E SH-SY5Y cells incubated with 100 μmol/L CBE for 48 h and then co-incubated with or without PAO or 50 nmol/L Baf-A1 for another 24 h, normalized to control. n = 5, data are presented as the mean ± SEM. ^###P <0.0001 versus control, **P <0.001 versus 100 μmol/L CBE-treated alone. F, G Representative and quantification of the expression level of PI4KIIIα protein in cells with control (sh-control) or PI4Ka-targeting vectors (sh1-PI4Ka, sh2-PI4Ka and sh3-PI4Ka) expressing double-stranded RNAs against the mRNA of PI4Ka. H, I Representative and normalization of LC3B protein levels in cells with control or PI4Ka-targeting vectors, and without [H(a) and I(a)] or with [H(b) and I(b)] CBE treatment. n = 5, *P <0.05 versus sh-control group. J, K Representative images and normalization of Lyso-Tracker labeling in cells with or without CBE treatment, which were infected by either sh-control or sh1-PI4Ka (GFP). n = 5, mean ± SEM, ***P <0.0001. Scale bar, 50 μm.

To further test that the therapeutic effects of PAO on GCase deficiency-caused pathological changes could be ascribed to the activation of ALP, we examined whether the protective effect of PAO on CBE-treated cells could be abolished by Bafilomycin A1 (Baf-A1), a specific inhibitor of vacuolar H⁺-ATPase (V-ATPase), which blocks autophagosome-lysosome fusion and inhibits acidification and substrate degradation in lysosomes. MTT assay results manifested that the cell protective effect of PAO was reversed by Baf-A1 (Fig. 2E).

To test whether PAO produced the foregoing protective effect against GCase deficiency via inhibiting PI4KIIIα, we generated three recombinant lentiviral vectors that expressed three different double-strand RNAs to specifically knockdown the mRNA of PI4Ka (encoding PI4KIIIα). As shown in Fig. 2F and 2G, all of the three vectors reduced the amount of PI4KIIIα protein in SH-SY5Y cells. Then, we infected cells with control or targeting vectors, followed by vehicle- or CBE-treatment for 48 h. We found that knockdown of PI4KIIIα increased the level of LC3B (Fig. 2H, I), but reduced the cellular accumulation of lysosomes in both vehicle- and CBE-treated cells (Fig. 2J, K). Thus, reduction of PI4KIIIα expression mimicked the effect of PAO in terms of activating ALP in cells with GCase insufficiency.

Previously we reported that inhibition of PI4KIIIα by PAO facilitates lysosome exocytosis in glial cells [9] and now we found that down-regulation of PI4KIIIα restored the cellular pathological changes in neuronal cells with GCase deficiency via facilitating lysosome exocytosis and activating autophagic flux. Both together indicate that PI4KIIIα and its inhibitors might be a potential new therapeutic target and compounds respectively for treating GD. The exact mechanism of how inhibition of PI4KIIIα facilitates lysosome exocytosis is unknown. PAO and PI4Ka knockdown could have biased the conversion of phosphatidylinositol into PI3P and PI(3,5)P₂, in turn upregulated the activation of TRPML1 (the principal Ca²⁺ channel mediating lysosomal exocytosis), and therefore ameliorated lysosomal accumulation of the unmetabolized substrates and alteration of lysosomes. On the other hand, PAO and PI4Ka knockdown could have simultaneously reduced the levels of PI(4,5)P₂ and PI(3,4,5)P₃ at the PM, then the activity of the Akt-mTOR pathway, leading to the disinhibition of autophagy and the restoral of autophagy flux and the ALP in CBE-treated cells.

The ALP and lysosomes are involved in a variety of cellular activities. Accumulating evidence demonstrates that in addition to regulating the final steps of catabolic processes, lysosomes are essential up-stream modulators of autophagy and other pathways implicating lysosome [11, 12]. Therefore, lysosomal dysfunction or alteration has a profound impact on cell homeostasis, resulting in manifold pathological situations, including neurodegeneration, cancer, infectious diseases, inflammation, and aging, in addition to LSD. Aged cells and cells with either LSD or neurodegeneration display blockade of ALP flux [13]. By facilitating lysosomal exocytosis and restoring ALP to relieve the block of autophagic flux, inhibition of PI4KIIIα may provide a common therapeutic approach for treating LSD and neurodegenerative diseases via removal of excessive metabolic wastes and damaged organelles. It is noteworthy that partial inhibition of PI4KIIIα could be therapeutically useful for the treatment of GD and other disorders while full inhibition of PI4KIIIα might trigger cell senescence and/or cell death [14].

In conclusion, we demonstrated that PAO, an inhibitor of PI4KIIIα produced protective effects on CBE-treated SH-SY5Y cell models with activation of ALP and lysosomal exocytosis, and reduction of the accumulation of substrates and lysosomes. In addition, our data indicate that knocking down PI4Ka mimicked the effects of PAO. Therefore, the evidence above suggests that down-regulation or inhibition of PI4KIIIα can be regarded as a potential therapeutic strategy for GD and other LSDs.

Supplementary Information

Below is the link to the electronic supplementary material.

Conflict of interest

The authors disclose no relevant conflict of interests with this research.