The OGT–c-Myc–PDK2 axis rewires the TCA cycle and promotes colorectal tumor growth

Deregulated glucose metabolism termed the “Warburg effect” is a fundamental feature of cancers, including the colorectal cancer. This is typically characterized with an increased rate of glycolysis, and a concomitant reduced rate of the tricarboxylic acid (TCA) cycle metabolism as compared to the normal cells. How the TCA cycle is manipulated in cancer cells remains unknown. Here, we show that O-linked N-acetylglucosamine (O-GlcNAc) regulates the TCA cycle in colorectal cancer cells. Depletion of OGT, the sole transferase of O-GlcNAc, significantly increases the TCA cycle metabolism in colorectal cancer cells. Mechanistically, OGT-catalyzed O-GlcNAc modification of c-Myc at serine 415 (S415) increases c-Myc stability, which transcriptionally upregulates the expression of pyruvate dehydrogenase kinase 2 (PDK2). PDK2 phosphorylates pyruvate dehydrogenase (PDH) to inhibit the activity of mitochondrial pyruvate dehydrogenase complex, which reduces mitochondrial pyruvate metabolism, suppresses reactive oxygen species production, and promotes xenograft tumor growth. Furthermore, c-Myc S415 glycosylation levels positively correlate with PDK2 expression levels in clinical colorectal tumor tissues. This study highlights the OGT–c-Myc–PDK2 axis as a key mechanism linking oncoprotein activation with deregulated glucose metabolism in colorectal cancer.

. The lysates were mixed with protein A/G agarose (MedChemExpress, HY-K0202-1) following incubation with the designated antibodies.

Quantitative RT-PCR
Trizol reagent (Takara, 9109) was used to extract total RNA.The quantity and quality of RNA were measured using a Thermo Fisher Scientific NanoDrop 2000 spectrophotometer.Following the manufacturer's instructions, Superscript reverse transcriptase (TAKARA, DRR047S) was used to reverse transcribe RNA.cDNA, genespecific primers, and SYRB Green Master Mix were used in quantitative real-time PCR (yesean, 11201ES08), which was detected using the Bio-Rad CFX96TM Real-Time System.The primers used in this work are presented in Supplementary Table 2. Fold changes were quantified using the 2-rCt method and normalized to β-actin.

Analysis of c-Myc glycosylation
Chemoenzymatic labeling was used to identify the O-GlcNAcylation of the protein.
Cell lysate (500 µg) was conjugated with an alkyne-biotin compound utilizing the Click-iT Protein Analysis Detection Kit (Invitrogen) after being labeled with an azide group using β-1,4-galactosyltransferase GalT (Y289L).Control experiments were conducted simultaneously in the absence of UDP-GalNAz or GalT.After neutralizing biotinylated lysates in a buffer (150 mM NaCl, 100 mM Na2HPO4, and 6% NP-40), the samples were treated with Pierce streptavidin resin for an overnight period at 4 °C.The resin was washed three times with low salt buffer (0.1 M Na2HPO4,0.15M NaCl,0.1% SDS,1% Triton-X100,0.5% sodium deoxycholate, pH 7.5) and high salt buffer (0.1 M Na2HPO4, 0.5 M NaCl, 0.2% Triton-X100, pH 7.5), respectively.The loading buffer was used to elute the bound proteins for western blotting analyses.

Site mapping of c-Myc glycosylation
c-Myc was isolated from HT-29 cells that were coexpressing Flag-tagged c-Myc and HA-tagged OGT.The protein was separated using 10% SDS/PAGE and stained with Coomassie blue R250 (Bio-Rad).The c-Myc band was excised, rinsed with Milli-Q water, and destaining buffer was applied twice for 30 minutes.After being dehydrated in acetonitrile and rehydrated with 10 mM DTT in 50 mM IAM for 45 minutes at 56 °C, proteins in-gel were digested with Trypsin/GluC for 16 hours at 37 °C.As directed by the manufacturer, the resulting peptides were desalted using Millipore's C18 Zip-Tips.
Following vaporation in a vacuum centrifuge, the samples were reconstituted in water and subjected to LC-MS/MS analysis.The materials were identified using the ETD fragmentation mode of the Thermo Fisher Scientific Orbitrap Elite mass spectrometer.

Analysis of metabolite by HPLC-MS/MS
Cells were collected and resuspended with prechilled 80% methanol by well vortex.

RNA-seq analysis
We used the following publicly available gene expression datasets: GSE232258.
Novogene carried out RNA-seq.Briefly stated, Trizol Reagent (Takara, 9109) was used to extract the total RNA from HT-29 cells following Scramble or OGT depletion, respectively.Agilent Technologies, CA, USA's Bioanalyzer 2100 system's RNA Nano 6000 Assay Kit was utilized to assess the overall amounts and quality of RNA.The RNA-Seq libraries were made with the Agilent SureSelect Strand-Specific RNA Library Preparation Kit.Next, AMPure XP Beads (Beckman Coulter, Brea, CA, USA) were used to select the library sizes.The libraries were then pooled based on the target amount of data off the machine and the effective concentration, and the Illumina NovaSeq 6000 was used to sequence the data.

ECAR and OCR analysis
The Glycolysis Stress Test Kit (Angilent, 103020-100) was used to calculate ECAR based on the manufacturer's instructions.In summary, 1.0 × 10 4 cells were seeded into an XF96 plate overnight.After switching to XF medium, the media were incubated for one hour.To detect the ECAR, more glucose (10 mM), oligomycin (1 μM), and 2-deoxy glucose (2-DG) (100 mM) were added.The Cell Mito Stress Test Kit (Angilent, 103015-100) was used to determine OCR.To detect the OCR, 1 μM rotenone, 0.5 μM FCCP, and 1 μM oligomycin were automatically added into the cartridge.The Seahorse Bioscience XF96 Extracellular Flux Analyzer was used to determine and normalize OCR and ECAR values to cell counts.

Measurement of Intracellular ROS Levels
Dichlorofluorescein diacetate (DCFH-DA), a fluorescent dye (Beyotime, S0033S), was used to detect the levels of ROS.DCFH-DA was diluted 1:1000 to a final concentration of 10 mmol/L in serum-free medium.The cells were collected and suspended in diluted DCFH-DA at a concentration of 1 to 20 million/ml and cultivated for 20 minutes at 37ºC in a cell incubator.Every three to five minutes, flip and stir to ensure the probe is in complete contact with the cells.To ensure that no DCFH-DA entered the cells, the cells were carefully washed three times using serum-free cell culture media.A Multi-Mode Plate Reader (BioTek) with an excitation wavelength of 488 nm and an emission wavelength of 525 nm was used to measure the fluorescence intensity.

Measurement of intracellular or tissue NADPH and NADP + levels
To measure NADPH and NADP + levels, the NADP + /NADPPH Quantitation Kit (Biovision, K347-100) was utilized.Briefly, Cells were lysed in extraction buffer at 4°C for 20 minutes and centrifuged at 12,000 g for 10 minutes.To analyze total NADP + /NADPH, 50 μL supernatant was added to 100 μL NADP + reaction mixture and incubated at room temperature for 30 minutes.After being recorded, the signal at 450 nm was adjusted for protein concentration.Following the elimination of NADP + , NADPH levels were measured as previously mentioned, by heating for 30 minutes at 60°C.

Measurement of intracellular or tissue GSH and GSSG levels
Using a GSH/GSSG Quantitation Kit (Biovision, K264) and following the manufacturer's instructions, the levels of GSH and GSSG were determined.After adding the cell to the GSH Assay Mixture, it remained at room temperature for half an hour.A Multi-Mode Plate Reader (BioTek) was used to detect the absorbance at 412 nm and then normalized to the concentration of protein.

Generation of c-Myc reconstituted stable cell lines
To generate c-Myc reconstituted cell lines, pLenti-FlagN-shRNA vector that affords small hairpin RNA (shRNA)-mediated knockdown of endogenous c-Myc while coexpressing exogenous Flag-tagged c-Myc (WT, S415A) was employed.To deplete endogenous c-Myc, the 3′ UTR sequence of c-Myc gene was inserted into pLenti-FlagN-shRNA vector.The rescued c-Myc sequences (flag-tagged WT, S415A) were inserted into the vector.The sequences were present in Supplementary Table 2. Cells were infected with the lentiviruses and stable cell lines were selected with 4 g/ml puromycin for 2 weeks.

Measurement of glucose uptake, ATP levels, lactate production,
Glucose uptake was assessed using the glucose uptake-GloTM test kit (Promega, J1341) following the manufacturer's instructions.Briefly, 2,000 cells were seeded per well in a 96-well plate.Cells were then washed with PBS.50 μl of prepared 1mM 2DG was added to each well and incubated at room temperature for 10 minutes.25 μl Stop Buffer was added to terminate the reaction.Followed by the addition of 25 μl Neutralization Buffer, 100 μl 2DG6P Detection Reagent was added and incubated at room temperature for 0.5-5 hours.The signal integration time was selected as 0.3-1 SEC on the luminescence detector (BioTek), and the luminescence signal was recorded.
ATP levels were determined by ATP assay kit (Beyotime, S0026).Briefly, cells were lysed and centrifuged at 12,000 g for 5 minutes at 4°C.The supernatant was with incubated with100 microliters of the ATP test solution for five minutes at room temperature.the RLU values were measured using a luminometer (BioTek).
To determine the amount of lactate released by cells, the culture media was collected.
A lactate test kit (Solarbio, BC2235) was used to measure lactate levels in accordance with the manufacturer's instructions.The extraction solution to volume (ml) ratio is 500~1000:1 (it is advised that 5 million cells be added to 1ml of extraction solution); After centrifuging for 10 minutes at 4°C and 12000g, supernatant was removed of 0.8 ml , then 0.15 mL of extract solution 2 was gradually added and mixed.The absorbance at 570 nm was recorded using a microplate reader (BioTek).

Cell proliferation assay
Cell proliferation rate was annlyzed by Cell Counting Kit-8 (Beyotime, C0039).In a 96-well plate, 2000 cells were planted and cultivated for the specified amount of time.
After adding CCK-8 solution to each well and incubating for one hour at 37 °C, the Multi-Mode Plate Reader (BioTek) was used to measure the absorbance at 450 nm.

Colony formation assay
Lentivarus was used to infect HT-29, HCT116, and NCM460 cells.Then 1000 cells were seeded into each well of six-well plates, and the cells were grown for two weeks at 37 °C in new media.After 30 minutes of fixing in 4% paraformaldehyde (Beyotime, P0099), the cells were stained for 20 minutes with 1% crystal violet (Beyotime, C0121).

Determination of c-Myc half-life
To repress the synthesis of new proteins, 50 μM cycloheximide (CHX) (MedChemExpress, HY-13259) was added to c-Myc WT or S415A rescue HT-29 cells.
The stated time points (0, 0.5, 1 or 2 hours) were used to harvest the cells after treatment.
After that, c-Myc levels were analyzed by western blotting, and the relative half-life was calculated .

Dual luciferase reporter assays
Using the primers listed in Supplementary Table1, the upstream (-1498/+1) region of the PDK2 gene promoter was cloned into the pGL3-basic expression vector.Using Polyjet (Sinagen, SL100688), cells were transfected with pGL3 and pGL3-PDK2 expression vectors.Renilla luciferase expression vector (pRL) was also transfected into cells as a transfection control.After 48 h of transfection, luciferase assays were performed using a Dual Luciferase Reporter Assay System (Beyotime, RG088S) according to the manufacturer's instructions.The promoter activities were normalized to the corresponding values of Renilla luciferase.

Chromatin immunoprecipitation assay (ChIP)
Plvx-HA control or plvx-HA c-Myc WT/S415A vectors were transfected into HT-29 cells for a duration of 48 hours.The cells were collected and washed three times with PBS.1% formaldehyde was added and incubated at room temperature for 10 minutes.
The addition of 125 mM glycine then halted the process.The cells were collected by centrifugation at 300 × g at 4°C and then resuspended it in 1ml ChIP sonication buffer (10 mM Tris-HCl pH 7.4, 25 mM KCl, 5 mM MgCl2) and DNA was sheared by a sonicator to fragment size.Debris from pellet cells were removed by centrifuging at 14,000 × g for 15 minutes at 4 °C.10% of the supernatant was set aside as input, and the remaining supernatant was incubated with an anti-c-Myc antibody overnight at 4 °C.
The corresponding IgG was used as a negative control.The supernatant was mixed with Protein A/G Magnetic Beads and then incubated for 4 hours at 4 °C on a rotary homogenizer.Following that, the beads were washed using the following buffers: low salt buffer (0.1% SDS, 1.0% TritonX-100, 2 mM EDTA, 20 mM Tris-HCl pH 8.0, 150 mM Nacl), high salt buffer (0.1% SDS, 1.0% TritonX-100, 2 mM EDTA, 20 mM Tris-HCl pH 8.0, 500 mM Nacl), LiCl wash buffer.Using 300 μL of elution buffer, the sheared DNA was eluted.The products were treated with RNaseA and Proteinase K after the crosslinking was reversed by adding 5 M NaCl (final concentration, 0.2 M) and cubating at 65 °C for the entire night.Using a PCR purification kit (Qiagen, 28024), the immunoprecipitated DNA fragments were recovered and subjected to RT-PCR.The crosslinking was undone by adding 5 M NaCl (final concentration: 0.2 M) and letting it sit at 65 °C for an entire night.The products were then treated with RNaseA and Proteinase K. Quantitative RT-PCR was used to evaluate the immunoprecipitated DNA fragments after isolated using a PCR purification kit (Qiagen, 28024).

Gene set enrichment analysis
Based on log2-fold changes, we used the signal-to-noise measurement derived by GSEA to rank the genes according to their connection with the Scramble (n = 3) and shOGT (n = 3) groups.Gene sets related to the glucose metabolic process were gathered from the following database: https://www.gsea-msigdb.org/gsea/index.jsp.

Transcription factors prediction
The putative transcription factors of PDK2 were scanned using the online bioinformatics tools TRANSFAC (http://gene-regulation.com/pub/databases.html)and Animal TFDBI (http://bioinfo.life.hust.edu.cn/AnimalTFDB/).For every transcription factor, the sites that had the highest binding score were set aside.Two transcription factors that were chosen were evaluated for their binding scores using JASPAR (http://jaspar.genereg.net/).

Xenograft model in nude mice and PET/CT imaging
6-week-old male nude BALB/c mice (n = 5 per group) were randomly and blindly assigned to groups.4 × 10 6 HT-29 cells were injected into mice flanks.Once a week, the tumors' volume was measured.When tumor diameter reached 400 mm 3 , 0.15 mCi 18 F-FDG was injected intravenously after starvation for one night.After 1 hour, 18 F-FDG micro-PET-CT scanning (Siemens, Berlin, Germany) was conducted.A pseudocolor map was used to display the PET acquisition images, with red representing a significant absorption of 18 F-FDG.The 18 F-FDG-PET activity was measured using SUVmax and MTV.Tumor samples were gathered and weighed.

Immunohistochemistry staining
Tissue blocks impregnated in paraffin were placed on glass slides that were charged with polylysine.30 minutes of exposure to 3.0% H2O2 inhibited endogenous peroxidase

Figure S3 (
Figure S3(A) The heatmap showing the genes that were differentially expressed in OGTknockdown HT-29 cells and control cells.(B)Quantitative PCR analysis was performed to detect the expression of the downregulated genes in glucose metabolism from the RNA-seq analysis in HT-29.n = 3; Data are presented as means ± SD.P-values were determined by unpaired two-tailed Student's t-tests.

Figure S7 (
Figure S7 (A-D) Immunoblotting analysis of c-Myc protein expression and c-Myc ubiquitination in HT-29 cells knocked down MAGI3 (A and B) and over-expressed MAGI3 (C and D).Data are presented as means ± SD.P-values were determined by unpaired twotailed Student's t-tests.(E) Analysis of c-Myc-MAGI3 interaction in 293T cells overexpressing WT or S415A Flag-tagged c-Myc and HA-tagged MAGI3 with or without OGT overexpression.Immunoprecipitation was performed using HA-tagged antibody.Immunoblotting analyses were performed with the indicated antibodies.
(B) Quantitative PCR analysis of c-Myc and PDK2 expression in HT-29 cells with c-Myc knockdown and reconstituted expression of shRNA-resistant WT, S415A c-Myc in the presence or absence of OGT overexpression.(C) Quantitative PCR analysis of PDK2 expression in HT-29 cells expressing Scramble or shc-Myc with or without OGT overexpression.(D) ChIP-qPCR analysis of c-Myc binding to PDK2 promoter in HT-29 cells with c-Myc knockdown and reconstituted expression of shRNA-resistant WT, S415A c-Myc upon TMG treatment.Ectopic c-Myc was pulled down by the anti-c-Myc antibody.n = 3; Data are presented as means ± SD.P-values were determined by unpaired twotailed Student's t-tests.