Cell culture
SH-Sy5y cells were purchased from the European Collection of Cell Cultures (Public Health England, UK). Cells used in this study were between passage 8 and 15. Cells were incubated in DMEM/F12 (1:1) supplemented with 10% human serum (Sigma, UK) due to high level of endogenous glucosylsphingosine present in standard 10% FBS supplemented culture medium. Cell viability was assessed using AlamarBlue™ HS cell viability reagent (Invitrogen, Thermo Fisher) as per manufacturer’s instructions. Cells were plated in 6 well plates at the density of 3 ×105 cells/well and treated with lyso-Gb3/GlcSph (Avanti lipids) dissolved in DMSO or vehicle for 24 h at the 20 ng/mL dose and 72 h at the 200 ng/mL dose. Cells were harvested using trypsin and centrifuged at 500 × g for 5 min, and further washed with PBS.
Label Free proteomics and deep proteomic phenotyping
Pelleted cells were lysed in PBS by three freeze-thaw cycles using 37 °C bath and dry ice. Cell lysate was ice-cold acetone precipitated at a volume of 1:3 and left at –20 °C for 4 h, then centrifuged to pellet precipitated proteins. The supernatant was removed and the protein pellet air dried. Precipitated protein was digested as described previously44. Briefly, protein was solubilised by the addition of 20 µL of 100 mM Tris buffer (pH 7.8) containing 6 M urea, 2 M thiourea and 2% (w/v) amidosulfobetain-14. Disulphide bridges in proteins were broken by the addition of 45 µg of dithioerythritol prepared fresh in 100 mM Tris buffer, pH 7.8 and left to shake for 60 min. Carbamidomethylation of cysteines was performed by the addition of 108 µg of iodoacetamide prepared fresh in buffer as before, and samples incubated in the dark for 45 min. To reduce urea concentration in the samples to below 1 M 160 µL of Milli-Q water was added to samples. 1 µg of trypsin Gold (Promega, Germany) was added to each sample and incubated at 37 °C for 16 h. Peptides were subject to high pH, low pH fractionation using Isolute 100 mg C18 cartridges (Biotage, Sweden) and eluted into 5 fractions at 6–50% ACN. Peptide eluents were dried by centrifugal evaporation and resuspended at a concentration of 300 ng/µL in 3% ACN, 0.1% FA.
Ultra-definition mass spectrometry (UDMSE) analysis was performed as previously described44. Briefly, 0.3 µg of peptides were loaded onto a Waters NanoAquity Liquid Chromatography (LC) system and separated over 60 min on a 75 µm × 150 mm, 1.7 µm Peptide BEH C18 analytical column (Waters, UK), then injected into a Synapt-G2-Si mass spectrometer with Ion Mobility Separation (IMS) (Waters, UK).
Raw data was imported to Progenesis QI version 4.1 (Waters, UK) and each fraction was processed separately before all five fractions were combined into one experiment. MS/MS spectra in the MS raw data files were searched against the UniProt most recent published human reference proteome database with manual addition of porcine trypsin (P00761). Enzyme digestion was set to trypsin, maximum missed cleavage number was set to 3. Cysteine carbamidomethylation was set as a fixed modification, methionine oxidation, protein N-terminal acetylation, glutamine deamidation and protein N-terminal pyrrolidone carboxylation were set as variable modifications. False discovery Rate (FDR) cut off was set at 1% at both peptide and protein level. Protein quantification was performed using unique peptides only. The resulting protein identifications and quantitative data were exported to Excel (Microsoft) for further analysis.
The immobilisation and the identification of lyso-lipid interacting proteins
Glucosylsphingosine and lyso-Gb3 standards in MeOH were evaporated by centrifugation, then glycosphingolipids were reconstituted in 150 mM phosphate buffered saline (PBS, pH 7.4) to a final concentration of 1 mg/mL. Five milligrams of Dynabeads® M-270 epoxy (Invitrogen, UK) were resuspended in 2 mL of dimethylformamide organic buffer, vortexed and aliquoted into 7 vials (blank, sample and a duplicate) at 3.3 ×108 beads per vial. Buffer was removed from the beads using a magnet. Beads were washed with 1 mL of 150 mM PBS (pH 7.4), resuspended in 0.1 mL of 100 mM sodium phosphate buffer (pH 7.4) and vortexed. One hundred microlitres of 1 mg/mL lysoGb3, glucosylsphingosine, or PBS (blank) was added to each aliquot of beads. To enhance glycosphingolipid-bead binding 0.1 mL of ammonium sulphate buffer (1 M final concentration) was added. The mixture was incubated at room temperature while mixing on a rotor for 24 h. The beads with bound lyso-glycosphingolipids were placed on a magnet and supernatants removed. The beads were washed with 0.5 mL of 150 mM PBS (pH 7.4) containing 2.5 mg/mL of blocking agent horse myoglobin and further washed in PBS. Lyso-glycosphingolipid bound beads were resuspended in 0.1 mL of 100 mM sodium phosphate buffer (pH 7.4), 0.035 mg/mL protein from SH-SY5Y cell lysate diluted in 150 mM PBS (pH 7.4) added and incubated at room temperature for 1 hour while mixing on the rotor. Supernatants were removed and beads washed with 0.5 mL of 150 mM PBS containing 0.1% Triton-100 (pH 7.4) for 15 min. Triton wash was removed and beads were washed with 150 mM PBS (pH 7.4) twice. Protein-glycosphingolipid bound partners were eluted from the beads by the addition of 40 µL of 100 mM Tris (pH 7.8), containing 6 M urea, 2 M thiourea, 0.2% ABS-14 (w/v), subjected to in-solution digestion without fractionation and LC-MS/MS proteomics analysis44. Data were imported to ProteinLynx Global SERVER v3 (Waters, UK), protein identification was performed using UniProt’s most recent published human reference proteome FASTA database (https://www.uniprot.org/proteomes/UP000005640) with manual addition of porcine trypsin (P00761) and horse myoglobin (P68082). Protein quantitative data were exported to Excel (Microsoft) for further analysis. Proteins identified were considered to interact with a glycosphingolipid if they were not detected in the blank – those found to interact with the beads without glycosphingolipid bound to the beads.
Ubiquitinated protein analysis
Mouse IgG1 antibody that recognises mono and polyubiquitin chains (Clone FK2, Sigma, UK) on proteins was diluted in 200 µL of PBS containing 0.02% of Tween 20 and coupled to protein G magnetic beads (Dynabeads®, Invitrogen, Sigma) at 4 µg antibody/0.5 mg beads overnight while on a rotor at 4 °C. The bead-antibody complex was placed on a magnet, the supernatant removed, the bead-antibody complex washed twice with 200 µL of PBS-Tween 20, then the bead-antibody complex was incubated with 250 µL horse myoglobin (2.5 mg/mL) for 15 min at room temperature while shaking to reduce nonspecific binding. Horse myoglobin was washed off twice with 200 µL of PBS-Tween 20. The bead-antibody complex was incubated for 1 hour at room temperature on the rotor with SH-SY5Y cell lysates prepared as described above. Bead-antibody-antigen complex was washed with 200 µL of PBS-Tween 20 twice, following which antibody-antigen complex was eluted from the beads with 40 µL of digest buffer containing 100 mM Tris (pH 7.8), 6 M urea, 2 M thiourea, 2% (w/v) ASB-14 while rotating for 2 min. Supernatants containing the antibody-antigen complex were subject to proteomic analysis44. Identification of proteins present in the FK2 antibody immunocapture was performed as described in the previous section with the only difference that mouse immunoglobulin gamma-1 chain C (P01868) from the commercial antibody was also added to the database. Relative protein quantitation was achieved based on the amount of mouse immunoglobulin gamma-1 chain C peptide added to beads. Data were exported to Microsoft Excel for further analysis.
Tubulin polymerisation assay
The assay was performed according to the manufacturer’s protocol using Cytoskeleton BK011P kit. GlcSph and lyso-Gb3 stocks in DMSO and DMSO buffer control were diluted with Milli-Q water at 10 x desired concentrations. Taxol in DMSO and colchicine were diluted in Milli-Q water at 30 µM. Infinite 200 plate reader (Tecan, Switzerland) was set into fluorescence (excitation 340 nm, emission 435 nm) top reading kinetic mode at 37 °C with a 30 s interval, 30 nm gain and 3 flashes per reading. A black, flat bottom, 96-well pate (Corning, costar, #3686) was placed into the reader for 10 min to warm up. DMSO, taxol, colchicine, and GlcSph and lysoGb3 samples were pipetted in duplicates into the plate and an assay buffer containing final concentration of GTP (1 mM), tubulin (2 mg/mL), glycerol (15%, v/v) was added to each well. The plate was immediately placed into the reader, shaken for 10 s in orbital mode, then fluorescence was recorded for 60 min. The data were exported into Excel (Microsoft Inc) for further processing and visualised using Graghpad, v 10.4 (Prism).
Analysis of cellular energy charge
Five hundred millilitres of 1 M ice-cold perchloric acid was added to the cells, they were scraped and collected into chilled tubes. Two hundred microliters of 0.5 M Potassium Hydrogen Carbonate in 1 M Potassium Hydroxide were combined with 250 µL of cell extract, vortexed, centrifuged at 14,000 × g for 5 min at 4 °C and the supernatant collected into a clean tube. This was stored at −80 °C until derivatisation. One hundred µL of 1 M sodium acetate, pH 4.5 and 10 µL of 4 M chloracetaldehyde were added to 100 µL of cell extract. The mixture was incubated at 60 °C for 40 min and the reaction quenched by putting the sample on ice. Once cool the samples were analysed at 4 °C using High-Performance Liquid Chromatography (HPLC) coupled to fluorescence detector. HPLC equipment consisted of PU-1580 intelligent pumps, AS-950 autosampler and DG-1580-53 in-line mobile phase degasser (Jasco Inc.). Etheno-adenine nucleotides were chromatographically separated over a 150 mm × 4.6 mm, 3 µm ODS C18 analytical column (Hypersil) using a 100 – 0% 0.2 M potassium phosphate, pH 5 and 0.2 M potassium phosphate with 10% acetonitrile, pH 5 over a 31-min linear gradient at a flow rate of 0.8 mL/min. The FP-920 fluorescence detector (Jasco) was maintained at 290 nm excitation and 415 nm emission wavelength respectively and the data collected with EZChrom Elite v3.17 software (Jasco Inc.). Chromatographic peaks acquired for ATP, ADP and AMP were integrated using elution times of corresponding standards and were used to calculate adenylate energy charge (AEC) according to equation equation45:
$${AEC}=frac{left[{{{rm{ATP}}}}right]+frac{1}{2left[{{{rm{ADP}}}}right]}}{left[{ATP}right]+left[{ADP}right]+left[{AMP}right]}$$
Analysis of glutathione reductase activity
Cells in PBS were scraped into microcentrifuge tubes, PBS removed by centrifugation for 5 min at 4 °C at 1000 × g, 200 µL of ice-cold assay buffer added, the pellets centrifuged for 15 min at 10,000 × g at 4 °C and supernatants collected and stored at −80 °C until ready for analysis. The assay was performed using ab83461 GR assay kit (Abcam, UK) according to the manufacturer’s instructions. GR reduces GSSG, which in turn reacts with 5,5’-dithiobis-2-nitrobenzoic acid (DTNB) to generate a coloured product (TNB) with strong absorbance at 405 nm. Endogenous GSH was destroyed with 3% hydrogen peroxide (5 µL) added to the sample (100 µL) for 5 min at room temperature, followed by the treatment with catalase (5 µL) for another 5 min. A 0–50 nmol/well TNB standard curve was prepared using a TNB standard and ran alongside the samples, which were added to a 96-well plate (30 µL). A reaction mixture containing GR assay buffer, DTNB, GSSG and NADPH solutions was added to each well (50 µL) and the absorbance read immediately at 405 nm in an Infinite 200 plate reader at room temperature generating an initial reading (A1). The samples were incubated at first for 10 min and the absorbance read again in a kinetic mode with an interval of 1 min for the next hour. TNB standard curve was plotted and a change in absorbance between the initial reading and another time point determined to fall within the linear range of the reaction was applied to the standard curve to interpolate the amount of TNB in each sample. The activity of GR was calculated as follows: GR activity = (∆B*Sample dilution factor)/((T2–T1)*0.9*V) (nmol/min/mL), where ∆B (nmol) is the amount of TNB interpolated from the standard curve, T1 (min) is the time of the initial reading, T2 (min) is the time of the second reading, V (mL) is the sample volume and 0.9 is the sample volume change factor. The resulting value was normalised to total protein determined by the Bradford method to yield GR activity (nmol/min/mg).
Analysis of malate dehydrogenase activity
Cells were collected by scaping. PBS was removed by centrifugation for 5 min at 4 °C at 1000 × g and pellets homogenised in 100 µL of ice-cold assay extraction buffer for 10 min on ice. The pellets were centrifuged at 10,000 × g for 5 min at 4 °C, the supernatants collected and stored at −80 °C until ready for analysis. The assay was performed using an MDH assay kit (Abcam, ab183305, UK) according to the manufacturer’s instructions. One hundred microlitres of ice-cold assay buffer was added to each pellet, placed on ice for 10 min and spun at 10,000 × g for 5 min at 4 °C, the supernatants collected and diluted 1 in 10 in the assay buffer. Ten microlitres of each sample dilution was aliquoted out for total protein measurement by the Bradford method. Samples were analysed in duplicates. A 6-point calibration curve containing 0–12.5 nmol of NADH standard was prepared in duplicates and sample blanks for each sample dilution was run alongside. The reaction mixture for the samples and standards contained samples/standards, assay buffer, enzyme mix, developer and substrate. Sample blanks had all the above except for the substrate. After an initial shake the plate was incubated at 37 °C in Infinite 200 plate reader, the initial absorbance measured at 450 nm, then the absorbance was measured in a kinetic cycle for 30 min every 5 min until the value of the most active sample was greater than the value of the highest standard. The activity of MDH was calculated as follows: ({{{rm{MDH; activity}}}}=frac{{{{rm{Sa}}}}}{left({{{rm{Reaction; Time}}}}right){{{rm{x; Sv}}}}}) where Sa is the amount of NADH (nmol) generated in the sample between the final and initial time of reading (interpolated from the standard curve) ({{{rm{Reaction; Time}}}}) is the difference between the final and initial time of reading (min) ({{{rm{Sv}}}}) is the sample volume used in the rection (mL). The resulting value was normalised to total protein to yield MDH activity (nmol/min/mg).
Analysis of mitochondrial malate dehydrogenase activity
Samples were prepared in the same manner as for MDH assay with the exception that the pellets were incubated in 200 µL of assay extraction buffer for 20 min on ice. The sample protein was determined by the Bradford assay and each sample was diluted to contain 150 µg of protein. The assay was performed using MDH2 activity kit (ab119693, Abcam, UK) according to the manufacturer’s protocol. Samples were diluted with incubation buffer and together with a dilution series of a control sample (100 µL) were added to the wells to capture mitochondrial MDH2 and incubated at room temperature while shaking at 300 rpm for 3 h. The buffer was removed, samples washed with blocking buffer and the wash repeated. An activity solution (100 µL) containing sodium malate, NAD+, coupler, reagent dye and assay buffer giving the final concentration of malate and NAD+ 5 mM and 4 mM respectively was added to the samples, absorbance read at 450 nm in a kinetic cycle for 25 min with a 20 s interval with shaking before and between readings. MDH2 activity was calculated using the Beer-Lambert law and the dye extinction coefficient was 37 mM−1 cm−1.
Analysis of citrate synthase activity
CS activity was determined by a modified method based on the premise that upon the action of CS Coenzyme A and citrate are liberated and Coenzyme A reacts with DTNB, which causes a colorimetric change and can be monitored at 412 nm38. Cells were collected on ice and centrifuged at 4 °C, the PBS removed and the pellet re-suspended in 500 µL of 10 mM Tris buffer, pH 7.4 containing 320 mM sucrose and 1 mM EDTA. Each pellet was split into two 250 µL aliquots, one for the CS the other for complex I activity assays, and quickly snap frozen on dry ice before storage at −80 °C prior to analyses. Cell lysates were thawed and snap frozen twice using ethanol and dry ice and each sample cuvette was prepared to contain 950 µL of 100 mM Tris/ 0.1% Triton, (v/v), pH 8.0, 20 µL of cell homogenate, 10 µL of 10 mM acetyl-CoA and 10 µL of 20 mM DTNB. The corresponding reference cuvettes were prepared in the same way but had 960 µL of 100 mM Tris/0.1% Triton, (v/v). The cuvettes were covered with parafilm, gently inverted twice before they were put into a Kontron Uvikon 941 spectrophotometer (Nothstar Scientific) maintained at 30 °C. Ten microlitres of 20 mM oxaloacetate was added to the sample cuvettes only, gently mixed, and the reaction monitored at 412 nm for 30 min and measured every 30 s. CS activity was calculated using the Beer-Lambert Law, where path length was 1 cm, total volume was 1 mL and the extinction coefficient of DTNB was 13.6 mM−1 cm−1. CS activity was expressed in nmol/min/mg.
Analysis of mitochondrial complex I activity
A modified method of Ragan et al. was used to determine complex I activity46. Each sample was prepared to contain 800 µL of 25 mM phosphate buffer containing 10 mM magnesium chloride, pH 7.2, 10 µL of 100 mM potassium cyanide, 50 µL of 50 mM BSA, 30 µL of 5 mM NADH, 20 µL of cell homogenate and 80 µL of water. The corresponding reference cuvettes were prepared in the same way but contained 90 µL of water. The cuvettes covered with parafilm were gently inverted twice before being put into a Kontron Uvikon 941 spectrophotometer maintained at 30 °C. Ten microlitres of Coenzyme Q1 was added to the sample cuvettes only, gently mixed and left to reach 30 °C, and the reaction measured at 340 nm for 5 min at 30 s intervals. After 5 min elapsed 20 µL of 1 mM rotenone was added to the sample cuvettes only and after a 2–3 min wait the measurement continued for a further 5 min. Using the Beer-Lambert Law the concentration of oxidised NADH was calculated where the change in absorbance at 340 nm after rotenone addition was subtracted from the change in absorbance before rotenone addition, the extinction coefficient of NADH was 6.22 mM−1 cm−1, path length was 1 cm and volume was 1 mL. The resulting value expressed as nmol/min/mg was divided by the value of CS activity to obtain a ratio.
Analysis of mitochondrial complex II/III activity
The activity of complex II/III was assayed based on the King method47. The oxidation of succinate to fumarate by complex II shuttling electrons from Coenzyme QH2 to reduce cytochrome c by complex III and is measured spectrophotometrically at 550 nm as a succinate dependent antimycin A (AA) sensitive reduction of cytochrome c. Each sample cuvette was prepared to contain a final concentration of 1 mM potassium cyanide, 300 µM EDTA, 100 µM cytochrome c, in a 166 mM potassium phosphate buffer pH 7.4 with addition of 185 µL of water. The corresponding reference cuvette was prepared in the same manner but 225 µL of water was added. 20 µL of sample was then added to each sample cuvette only, gently mixed and left to warm to 30 °C in the spectrophotometer. The reaction was initiated by the addition of 40 µL of 0.5 mM succinate into the sample cuvettes and the absorbance followed for 5 min. The absorbance for each sample was determined by subtracting the difference in absorbance prior to and after the addition of AA and Beer-Lambert law was applied using extinction coefficient of cytochrome c as 19.2 ×103 M−1 cm−1. The resulting values were normalised to the activity of CS and expressed as ratios.
Analysis of alpha enolase activity
The activity was examined using an Abcam ab117994 kit according to manufacturer’s protocol. Cells were washed with ice-cold PBS and collected on ice by scraping. PBS was removed by centrifugation at 500 × g for 5 min at 4 °C and aspiration, pellets washed with 100 µL ice-cold PBS, centrifugation repeated and PBS removed. Pellets were dissolved in 200 µL of the kit extraction buffer while incubated on ice for 20 min, spun at 16,000 × g at 4 °C for 20 min, supernatants collected, total protein determined by the Bradford method and the supernatants stored at −80 °C till ready for analysis. Samples were diluted with incubation buffer to contain 150 µg of protein and a control sample dilution series from 0 to 250 µg of total protein was prepared. The native ENO1 was immunocaptured from the samples during a 2-h incubation at room temperature while shaking at 300 rpm. The buffer was removed, samples washed with blocking buffer and the wash repeated. An activity solution (100 µL), containing 0.5 mM NADH, 0.25 mM 2DG, PK, LDH and assay buffer was added to the samples and absorbance was immediately recorded at 340 nm over a period of 40 min with an interval of 20 s at 37 °C in an Infinite f200 plate reader. ENO1 activity was determined as the change in absorbance per min per µg of protein being proportionate to the consumption of NADH in the reaction mixture by the Beer-Lambert Law using the extinction coefficient of NADH as 6.22 mM−1 cm−1.
Analysis of ATP production and glycolytic rates using Agilent XFp Seahorse analyser
Sample preparation
XFp cartridges were left hydrating overnight prior to the day of the assay in MilliQ water. Alamar blue cell viability reagent was added to each well (1/10th volume) and left to permeate the cells overnight. On the day of the analysis cell viability was assessed by measuring colour change and fluorescence of the dye (excitation 530 nm, emission 590 nm) using an Infinite f200 plate reader (Tecan, Switzerland). Fresh XF DMEM assay medium pH 7.4 was prepared with 1 mM sodium pyruvate, 2 mM glutamine and 10 mM glucose on the day of the assay. Following the 72-h GlcSph treatment cell culture medium was carefully removed and cells checked under the microscope for any perturbations. They were then washed with XF medium, 180 µL of assay medium was added to each well, the plate checked under the microscope again and incubated for 1 h at 37 °C with no CO2 for 45–60 min to remove CO2. After degassing cells were washed with the assay medium, checked under the microscope and the plate placed in the Seahorse analyser for the analysis.
ATP production rate
Port A of the XFp cartridge was designated for acute treatment of both control and lyso-GSL treated cells with 3 µM UK5099 (mitochondrial pyruvate shuttle inhibitor). Manufacturer’s protocol was followed using ATP rate assay kit where port B contained 1.5 µM oligomycin, port C contained 0.5 µM rotenone/antimycin A. Data were exported into Seahorse Wave software (v 2.6.1, Agilent), normalised to cell viability and an ATP rate report was generated.
Glycolytic rate
Port A contained 0.5 µM rotenone/antimycin A and port B contained 500 µM 2-deoxy-D-glucose – an allosteric hexokinase inhibitor for glycolysis inhibition. The data were exported into Seahorse Wave software (v 2.6.1, Agilent), normalised to cell viability and a glycolytic rate report was generated.
Statistics and reproducibility
Data from the ATP and glycolytic rate reports and functional assays were exported into Graphpad Prism software (V9, USA) and analysed by a 2-way ANOVA with a Sidak’s post-hoc multiple comparison test for the ATP production and glycolytic rates, and a Kruskal–Wallis test with a Dunn’s post-hoc test for the activity of mitochondrial complexes I-III, citrate synthase, malate dehydrogenases, glutathione reductase, alpha enolase and cellular energy charge assays. Results were considered to be statistically significant with p value < 0.05.
Proteomics data were analysed for pathway enrichment using IPA (QIAGEN Inc, https://digitalinsights.qiagen.com/products-overview/discovery-insights-portfolio/analysis-and-visualization/qiagen-ipa/). Input variables were set to proteins that demonstrated a significantly altered expression between the control and lyso-lipid treated cells, with fold-change as expression observation. The output pathways were set to those with p < 0.05. Gene ontology annotations were extracted using PANTHER bioinformatics tool (https://www.pantherdb.org/).
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.