ADP-dependent glucokinase regulates energy metabolism via ER-localized glucose sensing

Modulation of energy metabolism to a highly glycolytic phenotype, i.e. Warburg effect, is a common phenotype of cancer and activated immune cells allowing increased biomass-production for proliferation and cell division. Endoplasmic reticulum (ER)-localized ADP-dependent glucokinase (ADPGK) has been shown to play a critical role in T cell receptor activation-induced remodeling of energy metabolism, however the underlying mechanisms remain unclear. Therefore, we established and characterized in vitro and in vivo models for ADPGK-deficiency using Jurkat T cells and zebrafish. Upon activation, ADPGK knockout Jurkat T cells displayed increased cell death and ER stress. The increase in cell death resulted from a metabolic catastrophe and knockout cells displayed severely disturbed energy metabolism hindering induction of Warburg phenotype. ADPGK knockdown in zebrafish embryos led to short, dorsalized body axis induced by elevated apoptosis. ADPGK hypomorphic zebrafish further displayed dysfunctional glucose metabolism. In both model systems loss of ADPGK function led to defective N- and O-glycosylation. Overall, our data illustrate that ADPGK is part of a glucose sensing system in the ER modulating metabolism via regulation of N- and O-glycosylation.


Cell death analysis
Cell death was determined using "FITC Annexin V Apoptosis Detection Kit" (BD Pharmingen) according to the manufacturer's protocol using a cytometer

SDS-PAGE and western blotting:
Immunoblotting was performed as described previously [1]. In brief, protein quantification was carried out using Lowry protein assay. Protein was loaded in SDSgels of percentage adjusted to protein size of interest. Blotting was performed with a semi-dry blotting system and remaining gels were stained with Commassie to guarantee complete / even transfer. Blots were typically blocked with milk in TBS-Tween or PBS-Tween for 1 hour, washed three times and inoculated over night with first antibodies before washing and treatment with horseradish-peroxidase-coupled secondary antibodies and another washing step the following day before detection.
Western blots were analyzed using a Fusion-SL Advance 4. 2

Lowry protein assay
Protein quantification was determined in a modified form of the calorimetric assay, originally described by Lowry et al, using the Protein Assay Reagents A and B (Biorad). For quantification, a standard curve was generated using bovine serum albumin (BSA) at concentrations ranging from 0.0 to 2.0 mg/ml in 1% SDS. 25µl of reagent A (alkaline copper tartrate solution, allowing for complex binding of copper ions and peptide bonds of proteins) was added to samples and standards. After brief shaking, 200µl of reagent B was added, followed by 15 minutes of incubation on a plate shaker at room temperature. Measurement was performed in quadruples at 750nm wavelength on a standard spectrophotometer (Spectramax Plus Microplate reader, Molecular Devices, CA, USA).

Photometric lactate and pyruvate measurement
Lactate and pyruvate content was determined as previously described [5] using an Olympus AU400 system (Olympus, Tokyo, Japan). For lactate measurement cell or zebrafish homogenate was mixed with NADH and LDH. For pyruvate measurement cell or zebrafish homogenate was mixed with reaction buffer 1 containing NADH, 1.5M Tris Base and 0.2 % HClO 4 and then with reaction buffer 2 containing LDH.
One day after 1:2 dilution with fresh medium, cells were counted and seeded into 6 well plates in glucose-free RPMI (Thermo Fisher Scientific). After 3 hours of glucosedeprivation, 100µg/ml of 2-NBDG was added and cells were stimulated with PMA (10 ng/ml) or PMA + Iono (10 µM) for 1 hour. For measurement cells were washed and resuspended in Cell-Based Assay. Cells were gated towards viable cells in forward-/site-scatter distribution. Unstained cells and cells treated with 0.1 mM Apigenin in DMSO to inhibit Glucose uptake were included as negative controls.

Quantification of adenosine, nucleotides and nucleotide sugars
Metabolite analysis in cell and zebrafish lysates was performed according to previously published protocols by reversed-phase HPLC electrospray ionization tandem mass spectrometry [2] [3]. Briefly, metabolites were extracted from 30 zebrafish embryos snap-frozen or about 40 × 10 6 cells per condition with 0.9% NaCl.
Cells were split into three parts (adenosine analysis, nucleotide analysis and protein quantification), pelleted and snap-frozen and stored at -80°C until analysis at the the Metabolomics Core Technology Platform at the University of Heidelberg.
AMP, ADP and ATP were derivatized with chloroacetaldehyde at 80 o C, followed by injecting into an Acquity H-class UPLC system for separation. The UPLC was carried out by an Acquity BEH C18 150 mm × 2.1 mm, 1.7 µm with mobile phase A (5.7 mM TBAS, 30.5 mM KH 2 PO 4 pH 5.8) and B (2/3 acetonitrile in 1/3 buffer A) at a flow rate of 0.45 ml/min. The separated derivates were detected by fluorescence (Acquity FLR detector, Waters, excitation: 280 nm, emission: 410 nm, gain: 100) and quantified using ultrapure standards. The Adenosine nucleotides were also analysed by HPLC: Acquity BEH C18 column (150 mm × 2.1 mm, 1.7 µm) connected to an Acquity Hclass UPLC system was used for seperation. Following separation, adenosines were detected by fluorescence using an Acquity FLR detector (Waters, excitation: λ = 280 nm, emission: λ = 410 nm, gain: 100) and quantified using ultrapure standards. The Empower3 software suite (Waters) was used to acquire and process all data.

Subcellular fractionation and coupled enzyme assays
Enzyme activities in lysates and subcellular fractions of Jurkat T lymphocytes were monitored according to previously published protocols and normalized to protein content, determined via Lowry assay [1].
-Lactate dehydrogenase (LDH) activity was recorded as NADH oxidation in ETC buffer containing 1mM pyruvate and 0.5mM NADH.

Mitochondrial membrane potential / JC-1-assay
Mitochondrial membrane potential was measured cytometrically with "JC-1 Mitochondrial Membrane Potential Assay Kit"' (Cayman Europe) according to the manufacturer's protocol. In brief, the fluorescent dye JC-1, which exhibits an emission shift from green (~529 nm) to red (590 nm) upon depolarization of mitochondria was used to stain cells and red/green-fluorescence-intensity-ratio was determined. One day after 1:2 dilution with fresh medium, cells were counted and plated into 6 well plates. After one hour of acclimatization time, cells were stimulated with PMA (10 ng/ml) or PMA + Iono (10 µM) for 24 hour. After stimulation, 100µl of 1:10 diluted JC-1 staining solution were added per 1 ml of cell suspension and cells were incubated for staining in a cell culture incubator for another 30 minutes. Stained cell lysates were washed with assay buffer, transferred into tubes and directly subjected to cytometric analysis on a FACSVerse cytometer (Becton Dickinson).

Acridine orange staining of cells
Determination of acidic vesicles was performed by staining cells with Acridine orange as described previously [4]. Cells were harvested and washed in colorless RPMI 1640, incubated with 1µg/ml Acridine orange at 37°C for 20 minutes and washed with ice-cold colorless RPMI for several times to rinse away any acridine orange residues.

Real time-quantitive PCR
Jurkat T Cell RNA and zebrafish RNA was prepared using "RNeasy kit" with on column-DNA-digestions (Qiagen) and Trizol® reagent with chloroform/isopropanol, respectively. The purified RNA was reverse-transcribed into cDNA using "Maxima First Strand cDNA Synthesis Kit" (Thermo Fisher Scientific). Quantitative real time PCR was performed using "SensiFast SYBR Hi-ROX Kit" (Bioline) and a CFX Connect Real Time system (Bio-Rad). Human genes were normalized to 18S rRNA except for metabolic genes, which were normalized to GAPDH. Zebrafish genes were normalized to ef1alpha. A list of RT-qPCR primer is in suppl. table 5.

Whole mount in-situ hybridization (WISH)
DIG-labelled in situ hybridization antisense mRNA probes were synthesized and transcribed from a linearized PCRII dual promotor (ThermoFisher) vector. Sense probes were used as control. For respective primers see suppl. table 5. WISH was performed as shown by [5]. Briefly, Zebrafish embryos were collected at the desired

TUNEL assay
Apoptosis in zebrafish embryos was detected via TUNEL assay (Roche, Basel, Switzerland). Embryos were fixed with 4 % PFA in PBS overnight. After fixation, embryos were washed twice with PBS for 5 minutes, dechorionated, and dehydrated in 50 % methanol in PBST for 5 minutes and 100 % methanol for 5 minutes.
Afterwards, embryos were washed with acetone for 20 minutes and rehydrated by serial incubation in 75, 50, 25 % methanol and 100% PBST for 5 minutes. They were permeabilized with proteinase K for 5 minutes and fixed with 4% PFA for 20 minutes.
The last step was repeated once. Embryos were washed 3 times with PBS-T for 20 minutes and stained with the TUNEL assay reagent for 1 hour at 37 o C in the dark.
Stained embryos were washed with PBS twice for 5 minutes and observed under a fluorescence microscope.

Acridine orange staining of zebrafish
Live imaging of Acridine orange staining was used to determine cell death. Embryos were stained with 5 µg/µL Acridine orange in E3 medium for 30 minutes followed by three times PBS wash. Stained embryos were observed under a fluorescence microscope.

Cell cycle analysis
Cell cycle analysis was performed with "Propidium Iodide Flow Cytometry Kit" (Abcam) according to the manufacturer's protocol using a FACSVerse cytometer (Becton Dickinson) with gating to viable cells in forward/site scatter. One day after 1:2 dilution with fresh medium, cells were counted and plated into 6 well plates.
Following one hour of acclimazitation, cells were stimulated with/without PMA/Iono for various times before stimulation was terminated via washing with ice-cold PBS.

Whole mount zebrafish glucose measurement
Zebrafish glucose content was measured by a "glucose colorimetric assay kit"

N-Glycan Labelling, Clean-up, and Enzymatic Processing
Released N-glycan were reconstituted in 50 µL of 1% formic acid, thus converted to reducing aldoses, lyophilized and derivatized for two hours at 65°C with 10 µL 2aminobenzamide (2-AB) via reductive amination with sodium cyanoborohydride in 30% v/v acetic acid in DMSO. Removal of excess labelling agent was conducted by frontal HILIC purification using a Thermo UltiMate3000 RS UHPLC system (Thermo

Weak anion exchange separation of labeled N-glycans
Fractionation of N-glycan based upon their degree of sialylation was conducted by anion exchange chromatography using a Waters BioSuite DEAE 10 µm AXC, 7.5 x 75 mm column. Glycans were eluted by a 35 minute linear gradient of 100 mM acetate, pH 7.0 in 20% v/v acetonitrile at 0.75 mL/min.

Hydrophilic interaction UPLC-FLR-MS analysis was conducted on a Waters
Acquity™ UPLC system with online fluorescence detection hyphenated to a Waters Xevo G2 QToF mass spectrometer through an electrospray ionization interface. A linear gradient of 74-55 % acetonitrile in 38 minutes was applied at 0.15 mL/min and at 60 °C using a glycan BEH Amide column, 1.7 µm, 1.0 x 150 mm. 8 µL sample was injected in 75 % v/v acetonitrile. Negative ionisation mode with a capillary voltage of 1.80 kV was applied. Ion source and nitrogen desolvation gas (600 L/h flow rate) temperatures were set to 120 °C and 400 °C, respectively and the cone voltage was kept at 50 V. Full-scan MS data scan range was set to 450-2500 m/z. Data collection and processing was carried out using MassLynx 4.1. Glycan structures are presented using CFG symbol nomenclature [6]. Label free comparison of N-glycan structures was performed with ProgenesisQI software (Waters).    Table 2 Metabolic characterization of control cells and ADPGK KO clones without treatment.  Table 3 2-AB derivatized N-glycan analyses of control and KO1 cells with and without 24 PMA/Iono activation analyzed via HILIC-FLR-MS.

Mean of N=4 independent experiments of CTRs (TF-CTR and WT-CTR) and
Quantitative comparison of the samples was performed through label-free data analysis. All structures presented showed changes with Anova p-value < 0.05. Full images of western blots of ER-fractions stained for SRPRbeta (Fig. 1a) Size appropriate, depiction of size marker missing 1 2 3 4 5 6 7 8 9 10 11 12 13 kDa
Full images of Western blots for Bip (Fig. 2e).