Exogenous GM1 can partially restore GM1 content reduced by PDMP treatment
To reduce the content of GM1 in the plasma membrane, SH-SY5Y cells were treated with 20 µM PDMP, an inhibitor of glucosylceramide synthesis, which can partially reduce the level of GM1. To examine the effect of PDMP on cell viability, cells were treated with various concentrations of PDMP in 96-well plates (0µM, 10µM, 20µM, 30µM, 40µM, 50µM, 60µM, 70µM, 80µM, 90µM, 100µM, 120µM, 160µM). We found that the cell proliferation decreased with increasing PDMP doses, and that PDMP had no significant effect on cell viability at 20 µM (Fig. 1A).
To confirm the inhibitory effect of PDMP on GM1 content, CtxB-Alexa 647 was used to label GM1. Treatment with 20 µM PDMP significantly reduced the quantity of GM1. Notably, treatment with 100µM exogenous GM1 restored the GM1 content to almost to the basal level (Fig. 1B, C). This result is consistent with the report by Martino Calamai et al[26].
GM1 deficiency inhibits GDNF-induced cell proliferation
To investigate the effect of GM1 deficiency on the neurotrophic effects of GDNF, CCK-8 assays and EdU staining assays were performed to measure the proliferation of SH-SY5Y cells induced by GDNF after PDMP treatment. Cells gradually proliferated upon GDNF treatment, and the proliferation was significantly enhanced after 36 hours of treatment. In contrast, starting at 24 hours of treatment, cells treated with PDMP+GDNF proliferated significantly less than those treated with GDNF alone, and the change was reversed after the addition of exogenous GM1 (Fig. 2A). EdU staining results showed that after PDMP treatment, the proportion of EdU-positive cells induced by GDNF decreased significantly. The addition of 100 µM GM1 significantly restored the effect of GDNF on the proliferation of SH-SY5Y cells, even to a degree comparable to those treated with GDNF alone (Fig. 2B, C).
In conclusion, reducing the content of GM1 can inhibit the effect of GDNF on proliferation, and exogenous GM1 can reverse this change.
GM1 deficiency inhibits GDNF-induced cell differentiation
To examine the effects of GM1 on the pro-differentiation effect of GDNF, SH-SY5Y cells were treated under four conditions (Ctrl, GDNF, PDMP+GDNF and PDMP+GM1+GDNF), and microtubule-associated protein 2 (MAP2) immunofluorescent staining was performed. The axon length and the number of SH-SY5Y cells with axons were determined using ImageJ software. As shown in Fig. 3A, GDNF treatment increased the average axon length and the number of cells with axons, indicating the pro-differentiation effect of GDNF. Pretreatment with PDMP reduced the average axon length and the number of cells with axons induced by GDNF. Quantitative analysis revealed that exogenous GM1 increased the axon length at least by 0.1 inch compared to the PDMP treatment group (Fig. 3B). The number of cells with axons was also significantly increased by exogenous GM1 (Fig. 3C). In addition to the longer protrusions and increased number of cells with protrusions, the cells cultured with exogenous GM1 also maintained high tyrosine hydroxylase (TH) expression (Fig. 3D), which is specifically expressed by dopaminergic cells. In conclusion, reducing the content of GM1 can inhibit the effect of GDNF on differentiation induction, and exogenous GM1 can reverse this change.
GM1 deficiency impairs GDNF-RET signaling
Stimulation with 50 ng/mL GDNF increased the phosphorylation of the primary GDNF signaling molecules RET, Erk and Akt. Pretreatment with PDMP reduced GDNF-RET signaling. Fig. 4 demonstrates that p-RET, p-Erk and p-Akt induced by GDNF were reduced to basal levels when GM1 was depleted. The crucial role of GM1 in these changes was further illustrated by the increase in the levels of p-RET, p-Erk and p-Akt when GM1 was exogenously supplied to the cells pretreated with PDMP.
GM1 deficiency disturbs the assembly of lipid rafts
To investigate the alteration of lipid raft assembly, we performed Triton X-100 solubilization and OptiPrepTM density gradient centrifugation to isolate lipid rafts and examined the alterations in lipid raft markers in each sample by Western blotting. Immunoblotting of cell extracts from Triton X-100 solubilization samples revealed that treatment with 20 µM PDMP largely dispersed caveolin-1 and flotillin-1, the markers of lipid rafts, from the DRMs, also known as lipid rafts, to the detergent-soluble, non-raft membrane domains, and that this could be partially reversed by the addition of exogenous GM1. However, the total protein levels of CD71, a non-lipid raft marker, were not different between the GM1-deficient cells and the GM1-supplemented cells (Fig. 5A, B).
Fractions from OptiPrepTM density gradient centrifugation were also analyzed by Western immunoblotting. Not surprisingly, the results were consistent with those described above (Fig. 5C). Caveolin-1 and flotillin-1 in SH-SY5Y cells were mainly localized in lipid raft fractions, while CD71 was distributed in non-lipid rafts. After pretreatment with PDMP, caveolin-1 and flotillin-1 tended to disperse from lipid raft to non-raft fractions, which could be partially reversed by the addition of GM1. At the same time, the localization of CD71 in non-lipid rafts was not changed by PDMP treatment. These results suggested that GM1 deficiency could result in changes in lipid raft compositions.
RET translocation into lipid rafts is blocked in GM1-reduced SH-SY5Y cells
Many studies have shown that lipid rafts play a vital role in the transmission of GDNF-RET signals. Upon GDNF treatment, RET translocates into lipid rafts and activates RAS/Erk, PI3K/Akt and various other signal transduction pathways. To investigate the effect of lipid raft alterations induced by GM1 deficiency on GDNF-RET signals, we isolated lipid rafts and determined the translocation of RET into lipid rafts. First, we performed Triton X-100 solubilization to isolate lipid rafts. The results showed that, without GDNF treatment, RET was primarily located in detergent-soluble, non-raft membrane domains, and very little RET was detected in lipid raft fractions. After 20 minutes of treatment with 50 ng/mL GDNF, RET in lipid raft membrane domains increased. Pretreatment with PDMP inhibited the translocation of RET into lipid rafts induced by GDNF. In addition, GDNF-induced RET distribution in lipid rafts was partially restored following the restoration of GM1 content in PDMP pretreated cells (Fig. 6A).
To confirm these results, we used another biochemical method, OptiPrepTM density gradient centrifugation, to isolate lipid rafts. Consistent with the observations from Triton X-100 solubilization experiments, the majority of RET was located outside of DRMs without GDNF treatment, and after GDNF treatment, RET was translocated into lipid rafts. When GM1 was reduced, GNDF-induced RET translocation into lipid rafts was inhibited, which could be restored by GM1 addition (Fig. 6B).
GM1 is decreased in MPP+- induced PD cell model
MPP+ treatment is frequently used to establish a cell model of PD[27, 28]. In this study, SH-SY5Y cells were treated with 2.5 mM, 5 mM, or 7.5 mM MPP+ for 24 h, and cell viability was determined by CCK-8 assay. The results showed that at 5 mM, MPP+ significantly reduced cell viability to approximately 70% of that of the control group (Fig. 7A). Therefore, we used 5 mM MPP+ to treat SH-SY5Y cells. To investigate the effect of 5 mM MPP+ on GM1 content, we labeled GM1 with CtxB-Alexa 647 and quantified it by ImageJ software. We found that GM1 was significantly decreased after MPP+ treatment (Fig. 7B, C). This change could be partially reversed after adding exogenous GM1.
Synergistic protective effect of GM1 and GDNF in an MPP+-induced PD cell model
To determine the synergistic protective effect of GM1 and GDNF on the MPP+-induced PD cell model, we first selected the appropriate concentration of GDNF. MPP+-injured SH-SY5Y cells were treated with various concentrations of GDNF (10 ng/mL, 20 ng/mL, 30 ng/mL, 40 ng/mL, 50 ng/mL, 60 ng/mL) for 36 h. Subsequently, cell viability was determined by CCK-8 assay. The results showed that 10 ng/mL GDNF had no significant protective effect on MPP+-injured SH-SY5Y cells (Fig. 8A). We selected this concentration of GDNF to investigate the synergistic protective effect of GM1 and GDNF.
SH-SY5Y cells were treated with 20 µM PDMP with or without GM1 for 24 h followed by incubation in serum-free medium for 4 h. Subsequently, cells were stimulated with GDNF for 36 h. Cell viability was determined by CCK-8 assay and the expression of TH protein was evaluated by Western blot assay. The results showed that treatment with GM1 alone at 40 µM or 100 µM had no protective effect on injured cells, while 10 ng/mL GDNF together with 40 µM or 100 µM GM1 increased cell viability and the expression of TH protein in MPP+-injured SH-SY5Y cells (Fig. 8B, C).