Background
Nasopharyngeal carcinoma (NPC) is a malignant tumor of epithelial origin in the nasopharynx, with an incidence rate of up to 2 in 1,000 in Southern China and Southeast Asia [
1‐
4]. Because NPC occurs in the nasopharynx, its early symptoms are not obvious, but NPC have a strong tendency to invade and metastasize. Clinical studies show that approximately 70% of patients with NPC experience metastasis to the lymph nodes in the neck, and almost all patients in advanced stages have invasive growth to the skull base [
5‐
7]. However, to date, the molecular mechanisms involved in the development of NPC, especially in the metastatic process, are still unclear. Therefore, further understanding of the pathogenesis of NPC is needed to develop more effective therapeutic strategies to combat this disease.
Circular RNAs (circRNAs) are a newly discovered class of noncoding RNAs in recent years, and the study of their biological function and relationship with diseases has become a frontier of biomedical research [
8‐
11]. CircRNAs are a group of RNAs with a loop-like structure that lacks a polyA tail. They are not easily degraded by RNA enzymes and have very stable properties due to their special loop structure [
12‐
15]. With the development of the next-generation sequencing technology in recent years, an increasing number of circRNAs have been discovered, and more than 30,000 circRNAs have currently been identified in the transcriptome of various human cell types [
16,
17]. An increasing number of studies have found that circRNAs have important functions in numerous life processes and are involved in the development of many human diseases, including malignant tumors [
18,
19]. Although many circRNAs have been identified, the functions of the vast majority of circRNAs are still unclear, especially in NPC.
Intracellular protein ubiquitination is an important post-translational modification that is widely found in eukaryotic cells. The vast majority of intracellular proteins are degraded via the ubiquitin-dependent proteasome pathway. Small ubiquitin-like modifier 2 (SUMO2) is a member of the SUMO family, and SUMOylation is a type of ubiquitin-like protein modification. SUMOylation and ubiquitination often occur at the same lysine residues of a substrate protein, and SUMOylation can sometimes antagonize ubiquitination in the regulation of transcription factors [
20]. SUMOylation can also promote the ubiquitination and degradation of modified cytoplasmic, mitochondrial, and membrane proteins, serving as a signal for the recruitment of ubiquitin E3 ligases in various biological processes [
21].
Glucose transporter proteins (GLUTs) are one of the most important transmembrane proteins in the body that are responsible for glucose transport and reabsorption in different tissues and organs of the body. High expression of GLUT1 in various tumors, such as liver, gastric, and breast cancers, is involved in the regulation of malignant phenotypes, such as tumor metastasis and proliferation [
22,
23]. It provides favorable conditions for glycolysis through massive glucose uptake, providing more energy and synthetic raw materials for tumor cells. In addition, the large amount of lactic acid produced by glycolysis alters the microenvironment of tumor cells, which is more favorable for tumor cell invasion and metastasis.
In this study, we identified a novel circRNA molecule, circRNF13 (circBase ID: has_circ_0001346), which is expressed at low levels in NPC. The proliferation and metastasis of circRNF13 were assessed in NPC cells through in vitro and in vivo experiments. Further studies revealed that circRNF13 directly binds and stabilizes SUMO2 mRNA, which upregulates the protein levels of SUMO2 and promotes GLUT1 degradation by means of SUMOylation and ubiquitination, thus inhibiting the glycolytic process and proliferation, migration, and invasion of NPC.
Methods
NPC clinical samples
In total, 12 non-tumor nasopharyngeal epithelial (NPE) tissues and 36 NPC samples were collected at the Affiliated Cancer Hospital of Central South University (Changsha, China). This study was approved by the Joint Ethics Committee of the Central South University Health Authority, and informed consent was obtained from all participants. Diagnoses of all specimens were confirmed via histopathological examination.
Cell culture, plasmids, and transfection
NPC cell lines 5-8F, HNE2, CNE2, HONE1, and 6-10B and immortalized normal nasopharyngeal epithelial NP69 cells were obtained from the Cell Center of Central South University. All cells were cultured in RPMI-1640 medium (Gibco) supplemented with 10% fetal bovine serum (Gibco). All cells were maintained at 37 °C in a humidified incubator with 5% CO2.
The circRNA expression plasmid pCirc was generously provided by Prof. Li Yong at Baylor College of Medicine for RNA circulation. For
circRNF13 overexpression, cDNAs reverse-transcribed from RNAs of CNE2 cells were used as PCR templates and subjected to amplification of the full-length
circRNF13. The reporter plasmid pMIR-Luciferase-Reporter for SUMO2 and inserts of the SUMO2 3’UTR sequence within the pMIR-Luciferase-Reporter plasmid were constructed with the One Step Cloning Kit (Vazyme). Plasmids were transfected into cells using Lipofectamine 3000 Reagent (Invitrogen, USA). Primers used are listed in Table S
1.
RNA isolation and RT-PCR
Total RNA was extracted using TRIzol reagent (Life Technologies) and subjected to reverse transcription with random primers using the 5 × All-In-One kit (Abm). Then, the expression levels of target RNAs were measured with SYBR Green Master Mix using a StepOnePlus Real-Time PCR System (Applied Biosystems).
β-actin was used as an endogenous control, and the fold change was calculated via the 2
−∆∆CT method. Primers used are listed in Table S
1.
Cytosolic/nuclear fraction assay
Cells were resuspended in hypotonic buffer (25 mM Tris–HCl, pH 7.4, 1 mM MgCl
2, 5 mM KCl) and incubated on ice for 5 min before adding an equal volume of hypotonic buffer containing 1% NP-40 for an additional 5 min. After centrifuging the cells at 5,000 g for 5 min, the supernatant was collected as the cytosolic fraction. Pellets were washed twice with hypotonic buffer and then resuspended in nuclear resuspension buffer (20 mM HEPES, pH 7.9, 400 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 1 mM PMSF). After incubation on ice for 30 min, samples were centrifuged at 12,000 g for 10 min, and supernatants were collected as the nuclear fraction. Reverse transcription of RNA was used for RT-PCR detection. Primers used are listed in Table S
1.
RNase R treatment
RNase R treatment experiments were performed to verify the stability of circRNA. Five micrograms of total RNA was extracted from NPC cells and divided into two groups. One group was incubated with 1 μl RNase R (20 U/μl; Epicenter, WI, USA) at 37 °C for 30 min, and the other group was treated without RNase R. RNA was incubated at 70 °C for 10 min to inactivate RNase R and then reverse-transcribed for RT-PCR detection. Primers used are listed in Table S
1.
Fluorescence in situ hybridization (FISH)
A digoxigenin-labeled specific probe for
circRNF13 was designed and synthesized by Sangon. The FISH experiment was performed according to the manufacturer's instructions (GenePharma). First, cells were fixed and permeabilized using 0.25% Triton X-100. Then, cells were hybridized with specific probes overnight at 37 °C with
circRNF13 in hybridization buffer. Nuclei were counterstained with DAPI (Invitrogen, D1306, USA). Cells were imaged using a confocal microscope (Ultra-View Vox, Perkin-Elmer, Waltham, MA, USA). The
circRNF13 probe used is listed in Table S
1.
MTT and cell cycle assays
For MTT assays, 1,000 cells per well were seeded into 96-well plates for the MTT assay. Cells were incubated with 0.5 mg/mL filtered sterile MTT (Beyotime) at 37 °C for 4 h at the indicated time point. Then, the media were removed and replaced with 200 μL DMSO, and absorbance was measured at 490 nm.
Cell cycle assays were performed using flow cytometry according to the manufacturer's guidelines. To examine the cell cycle, measurements were performed. Briefly, cells were trypsinized, collected and stained in solution with RNase or PI for 15 min at 25 °C. Flow cytometry analysis was immediately performed using a FACSCalibur.
For colony formation assay. In brief, 2,000 cells were plated into 6-well plates after transfected circRNF13 overexpression vector or circRNF13 siRNA for 48 h. The cells were allowed to grow for the next 7 days to allow colony formation and the colonies were visualized using crystal violet staining.
Wound healing assay
Cells were cultured in 6-well plates for 24 h and then wounded using a sterilized pipet tip to make a straight scratch. After gentle washing with D-Hanks, cells were cultured in RPMI-1640 medium with 1% FBS. Pictures were taken using an Olympus digital camera at 0 and 24 h after wounding (at least three randomly selected fields were imaged).
Transwell assays
Cell invasion assays were performed using a Transwell chamber (Millipore). RPMI-1640 medium supplemented with 20% FBS was added to the bottom chambers, and then cells were suspended in RPMI-1640 medium and seeded into the top chamber, which was coated with Matrigel. After incubating at 37 °C for 48 h, cells that did not migrate through the pores were removed using a cotton swab. The Transwell chambers were fixed in 4% paraformaldehyde for 30 min, followed by staining with 1% crystal violet for 10 min. Cells on the bottom of the chamber were counted using an inverted phase-contrast microscope (at least three randomly selected fields were quantified).
Animal experiments
We purchased 4-week-old female BAL B/c nude mice from the Experimental Animal Center of Central South University (Changsha, China) and raised them in an SPF-free barrier environment. For lung metastasis experiments, nude mice were randomly divided into four groups (n = 7 per group). Each nude mouse was injected via the tail vein with 2 × 106 NPC CNE2 cells transfected with the circRNF13 overexpression vector, circRNF13 siRNA, the blank plasmid, or scrambled siRNA. After eight weeks, nude mice were sacrificed by cervical dislocation. Lung tissue was removed, weighed, and imaged, and the number of nodules on the surface of the lung was recorded to assess tumor metastasis. Lung tissues were then subjected to gradient dehydration, sectioned, embedded in paraffin, and stained with H&E for histological examination.
For subcutaneous tumorigenesis experiments, nude mice were randomly divided into four groups (n = 7). Each nude mouse was subcutaneously injected with 3 × 106 NPC CNE2 cells transfected with the circRNF13 overexpression vector, circRNF13 siRNA, or the blank plasmid, scrambled siRNA. Tumor growth was monitored every 3 days. Tumor size was assessed by measuring the largest perpendicular diameters, and tumor volume was calculated as follows: V = 1/2 × (length) × (width) × (width). Twenty-one days after subcutaneous inoculation, mice were euthanized by cervical dislocation, and the tumor tissue was excised. The formed tumor masses were removed and weighed. All animal protocols were approved by the Institutional Laboratory of Animal Care and Use Committee at Central South University.
Hematoxylin–Eosin staining (H&E)
Paraffin mouse tissue sections were first heated at 65 °C for 2 h. After paraffin sections were dewaxed and hydrated, nuclei were stained with hematoxylin solution (Biosharp, Anhui, China), and then cytoplasmic staining was performed with eosin staining solution (Biosharp, Anhui, China). After the slices were dried, the sheets were preserved with neutral resin (SCR, Shanghai, China).
Seahorse assays
Assays were performed using the Seahorse XFp analyzer (Seahorse Bioscience, Agilent) according to the manufacturer’s instructions. Briefly, 8000 cells/well were seeded into an 8-well XF cell culture microplate in growth medium 24 h before the assay. The extracellular acidification rate (ECAR) was measured using an XFp analyzer in XF base medium (pH = 7.4) containing 1 mM glutamine following sequential additions of glucose (10 mM), oligomycin (1.5 μM) and 2-DG (50 mM). Data were analyzed by the Seahorse XF Glycolysis Stress Test Report Generator package.
Liquid chromatography-mass spectrometry (LC–MS/MS)
Mass spectrometry assays according to the manufacturer’s protocol with minor modifications. Briefly, Scrambled siRNA or circRNF13 siRNA was transfected into CNE2 cells for 48 h. Total protein was extracted and digested with protease overnight. The digested peptide mixture was dried and treated with 0.1% trifluoroacetic acid (TFA). After diluting the 5 μL samples, we used an LTQ Orbitrap Velos Pro mass spectrometer (Thermo Scientific, Bremen, Germany) coupled with an Ultimate 3000 RSLC Nano System (Dionex, CA, USA) to recover proteins and perform proteomic analysis of total proteins, which were identified using Proteome Discoverer 1.4 software (Thermo Fisher Scientific, MA, USA), and the resulting original file was imported into the UniProt KB/Swiss-Prot database for searching. For the database search, the mass tolerances of the precursor and fragmentation were set to 10 ppm and 0.8 Da, respectively. Peptides with a false discovery rate < 1% (q value < 0.01) and proteins with an area value lower than 1 × 106 were discarded. Proteins that met the following criteria were considered to be differentially expressed proteins: ≥ 2 peptides and ≥ 95% confidence; and an average fold change in protein levels ≥ 2.00 or ≤ 0.50. (Student's t-test, p < 0.05).
Enrichment analysis
The differentially expressed proteins screened by LC–MS/MS were imported to Ingenuity Pathway Analysis (IPA) software, and canonical pathways of differentially expressed genes were analyzed to obtain enrichment pathways. Fisher's exact test was used to calculate a p-value to determine the probability of each biological function.
RNA pull-down
The biotin-labeled
circRNF13 probe was synthesized by Sangon Biotech, and the RNA pull-down assay was performed as previously described with minor modifications.
CircRNF13-overexpressing HNE2 and CNE2 cells were fixed in 1% formaldehyde for 10 min, lysed, and sonicated. After centrifugation, 50 μL of the supernatant was retained as input, and the remaining part was incubated with a
circRNF13-specific probe streptavidin Dynabeads (M-280; Invitrogen) mixture overnight at 4 °C. The next day, an M-280 Dynabeads probe-circRNA mixture was washed and incubated with 200 uL of lysis buffer. Finally, TRIzol was added to the mixture for RNA extraction and detection. The
circRNF13 probe and primers used are listed in Table S
1.
Dual-luciferase reporter assay
The Dual-Luciferase® Reporter Assay System was used according to the manufacturer’s instructions (Promega). To evaluate the interaction between circRNF13 and SUMO2, circRNF13 and pMIR-Luciferase-Reporter-SUMO23’UTR plasmids were transfected in HNE2 and CNE2 cells. Forty-eight hours later, firefly and Renilla luciferase activity was examined by the Dual-Luciferase Reporter Assay System, and Renilla activity was used to normalize firefly activity.
In situ hybridization (ISH)
The ISH kit was purchased from Boster Biological (CA, USA). Paraffin mouse tissue sections were deparaffinized and rehydrated with gradient alcohol-water solution, and endogenous peroxidase was inactivated with 3% H
2O
2. An appropriate amount of pepsin (1 mL 3% citric acid solution and two drops of concentrated pepsin) was added to tissue specimens and digested at 37 °C for 15 min. The digestion was then quickly terminated with 0.1 mol/L glycine solution. After refixation with 4% paraformaldehyde, the prehybridization solution was used at 37 °C for 30 min, and hybridization was conducted using a digoxin-labeled
circRNF13 probe (Sangon, Biotech, Shanghai, China) overnight at 37 °C. The next day, slides were washed in 2 × SSC, 0.5 × SSC, and 0.2 × SSC. Biotinylated rat anti-digoxigenin, streptavidin–biotin complex (sABC), and biotinylated peroxidase were then added dropwise. After incubation for 30 min, slides were washed with PBS. DAB color developing solution was subsequently added for 5–10 min and then placed in running water to stop the color reaction. Sections were subjected to dehydration using gradient alcohol, and a neutral resin mount was added dropwise. The
circRNF13 probe used is listed in Table S
1.
Protein half-life assay
Initially, cells were transfected with the circRNF13 overexpression vector or circRNF13 siRNA for 48 h. Then, cycloheximide (CHX, 50 μg/mL) was added into cell culture medium at indicated time points.
Immunoprecipitation
For immunoprecipitation, the antibodies were incubated with 35 μL of protein A/G magnetic beads (Bimake, Houston, Texas, USA) with constant rotation at room temperature for 2 h. HNE2 or CNE2 cell lysates were extracted using GLB
+ lysis buffer (150 mM NaCl, 10 mM Tris–HCl pH 7.5, 0.5% Triton X-100, 10 mM EDTA pH 8.0) with a protease inhibitor cocktail (Roche, Basel, Switzerland, USA) on ice for 2 h. Lysates were centrifuged and then incubated with antibody-conjugated beads for 4 °C overnight. Next the antibody-bead complexes were washed 5–6 times with cold GLB
+ lysis buffer. Then the precipitated proteins were resuspended and boiled 10 min using 6 × SDS-PAGE loading buffer. The boiled immune complex was put on ice for 2 min and subjected for SDS-PAGE electrophoresis. The primary antibodies used are listed in Table S
2.
Immunofluorescence
The cultured HNE2 and CNE2 cells were incubated with 4% paraformaldehyde and then blocked with 5% BSA. The cells were treated with specific antibodies at 4 °C overnight and the secondary antibodies at 37 °C for 1 h. And the cells were counterstained with DAPI for 10 min and imaged under a confocal microscope (Ultra-View Vox, Perkin-Elmer, Waltham, MA, USA). The primary antibodies used are listed in Table S
2.
Western blotting
Total protein lysate (40 μg per lane) was loaded on 10–15% Bis–tris polyacrylamide mini gels (Invitrogen). SDS-PAGE was run at 120 V for 1.5 h to 2 h. Proteins were transferred to nitrocellulose or PVDF membranes by wet transfer for 60–90 min at 100 V. Membranes were blocked in 5% nonfat dry milk in Tris-buffered saline supplemented with Tween 20 (0.1%) (TBST) or phosphate-buffered saline (PBS) for 60 min at room temperature. After blocking, membranes were cut horizontally to examine multiple proteins of different sizes on each gel. Membranes were incubated on a plate shaker overnight at 4 °C with primary antibodies diluted in TBS-T. Membranes were extensively washed with TBS-T (minimum 3 × for 10 min), followed by incubation with appropriate horseradish peroxidase-conjugated secondary antibodies diluted in TBS-T with 5% nonfat dry milk for 60 min at RT on a plate shaker. Membranes were extensively washed with TBS-T (minimum 4× for 15 min). Signals were detected using a Luminata Crescendo detection system following the manufacturer’s recommendations. Multiple film (HyBlot CL, Denville) exposures ranging from 2 s to 2 min were performed for optimal image analysis. The antibodies are listed in Table S
2.
Immunohistochemistry
Immunohistochemistry was performed on formalin-fixed paraffin-embedded sections of mouse xenograft tissues. Briefly, tissues were deparaffinized and rehydrated, and samples were subjected to EDTA-mediated high-temperature antigen retrieval; the samples were then incubated overnight at 4 °C with the primary antibodies. The staining was scored according to the staining intensity and the distribution of stained cells. The distribution was evaluated as none (0), ≤ 10% (1), 10%—50% (2), 50%—80% (3), and > 80% (4). Intensity was evaluated as none (0), faint (1), moderate (2), strong (3), or very strong (4). The sections were reviewed by two pathologists. The antibodies are listed in Table S
2.
Statistical analysis
Statistical analysis was performed using GraphPad Prism 7 software. Student’s t-tests were used to evaluate significant differences between any two groups of data, and one-way ANOVA was used to evaluate significant differences for multiple comparisons. All data are represented as mean ± Standard Deviation (SD). Differences were considered significant at p < 0.05. *, p < 0.05; **, p < 0.01; ***, p < 0.001.
Discussion
In this study, we identified a new circRNA,
circRNF13, which is expressed at low levels in NPC and regulates glycolysis in NPC cells by binding to the 3′-UTR of SUMO2 mRNA, to enhance GLUT1 degradation, resulting in the proliferation and metastasis of NPC. We screened the circRNAs from our RNA-seq data on the NPC cell line 5-8F. From the RNA-seq data, some new circRNAs were identified, which have not been reported yet. Some of the other identified circRNAs have already been reported to play important functions in the development of NPC. For example,
circMAN1A2 is highly expressed in the serum of patients, including NPC patients, and can be used as a potential molecular marker for tumors [
24].
In our study, we only selected
circRNF13, because of its high abundance in NPC cells. However, it was more highly expressed in NPE tissues than in NPC tissues. In NPC, some circRNAs have been reported to be oncogenes that promote the development of NPC. Our previous data showed that
circSETD3, acting as a competing endogenous RNA, promotes NPC migration and invasion by completely binding to
miR-615-5p and
miR-1538 with MAPRE1 [
25].
CircCRIM1 promotes NPC metastasis and chemoresistance by upregulating FOXQ1 expression through adsorption of
miR-422a [
26]. Wang et al
. found that
circTGFBR2 acted as a sponge for
miR-107, to upregulate TGFBR2, resulting in the inhibition of NPC progression [
27]. Chen et al
. found that
circHIPK3 promoted NPC proliferation and invasion by eliminating
miR-4288-induced inhibition of ELF3 [
28]. Fan et al
. found that
circARHGAP12 promoted NPC cell invasion and metastasis by binding directly to the 3′-UTR of EZR mRNA and promoting its stability [
29]. These results suggested that circRNAs play an important role in promoting the development of NPC. In this study, we found that
circRNF13, acting as a tumor suppressor, directly binds to and stabilizes SUMO2 mRNA, to promote SUMO2 protein expression, thereby inhibiting the proliferation and metastasis of NPC. Thus, it can serve as a new model for research on the mechanism of circRNA regulation in tumor development and for the development of drug targets.
Metabolic reprogramming and invasive metastasis are important features of tumors that have long been the focus of attention. Recent studies have revealed that these two malignant phenotypes are not independent, but are actually closely linked. For example, the transmembrane glycoprotein CD147 promotes MCT1 expression, leading to increased lactate secretion, which activates the PI3K/Akt/MDM2 pathway, increasing p53 degradation and promoting the proliferation and metastasis of hepatocellular carcinoma cells [
30,
31]. PGC1-α, a core regulator of metabolism, is significantly overexpressed in cancer and leads to a significant increase in tumor invasion and metastasis [
32]. Morrison et al
. found that melanomas with high expression of the monocarboxylate transporter MCT1 could resist oxidative stress by taking up lactic acid in the circulatory system, resulting in a greater metastatic capacity [
33]. In conclusion, abnormal metabolic function of tumor cells can enhance tumor invasion and metastasis through multiple pathways. In this study, we found that
circRNF13 inhibited glycolysis in NPC cells by suppressing the expression of GLUT1, which inhibited proliferation and metastasis. GLUT1 is an important protein in the process of glycolysis, and malignant tumor cells characteristically overexpress GLUT1, which provides favorable conditions for glycolysis through massive glucose uptake, thus providing more energy and synthetic raw materials for tumor cells. In addition, the large amount of lactic acid produced by glycolysis alters the microenvironment of tumor cells, which is more favorable for tumor cell invasion and metastasis. Designing
circRNF13-based inhibitors against GLUT1, to block tumor cell metabolism, is also a promising strategy for future malignancy treatment.
GLUT1 is the most widely distributed glucose transporter protein [
34]. High expression of GLUT1 has been reported in various tumors, such as liver, gastric, and breast cancers [
35‐
38]. While the transcriptional regulation of GLUT1 has been well studied, the post-translational modifications of GLUT1 have relatively few reports, especially ubiquitination modifications of GLUT1. Xu et al
. found that SALL4 recruits the E3 ubiquitin ligase CUL4B to GLUT1, which reduces the expression levels of GLUT1, and subsequently, inhibits glycolysis in cancer cells [
39]. Lin et al
. found that the E3 ubiquitin ligase family SCF complex Skp2 reduces GLUT1 expression and inhibits glycolysis in tumor cells by ubiquitinating Akt [
40]. In this study, we found that SUMO2 directly promotes the degradation of GLUT1 by means of SUMOylation and ubiquitination. This further increases the understanding of the regulatory mechanism of GLUT1. More importantly, blocking the nutritional source of cells by targeting GLUT1 is an important strategy for the development of anti-tumor drugs. Study of GLUT1 ubiquitination and SUMOylation will likely provide new ideas for drug development.
RNF13 is an E3 ubiquitin-protein ligase with a structural domain of a typical RING zinc-finger protein. In this study, we determined that
circRNF13 is a circular splicing of exons 2–8 of the RNF13 gene. Our study demonstrated that
circRNF13 is expressed at low levels in NPC and inhibits its proliferation and metastasis. The expression and function of
circRNF13 are inconsistent with those of its parental gene RNF13, as the latter exerts an oncogenic function. Overexpression of the RNF13 enzyme is apparent in various human cancers, including basal cell carcinoma, melanoma, and ovarian carcinoma [
41,
42]. RNF13, which acts as a ubiquitin ligase, participates in cancer invasion and metastasis. In this study, we analyzed the expression of RNF13 in the online GEO NPC database and our clinical NPC samples. There was no difference in RNF13 expression between the NPE and NPC clinical samples. Knockdown of RNF13 also had no effect on the ubiquitination of GLUT1. RNF13 mRNA could not bind to SUMO2 mRNA in the RNA pull-down experiment (data not shown). These results suggested that the function of
circRNF13 is independent of RNF13 mRNA, in promoting ubiquitination and SUMOylation of GLUT1, by binding to SUMO2 mRNA.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.