Lower levels of FTO and ALKBH5 work together to boost FOXO signaling by modifying HK2 mRNA, increasing glycolysis in colorectal cancer

Thirty-six pairs of frozen tissues were obtained from patients diagnosed with CRC who also underwent surgical excision at the Jiangsu Province Hospital. All protocols were reviewed and approved by the Ethics Committee of Nanjing Medical University. SW620 and HCT116 cell lines were purchased from the Chinese Academy of Science (Shanghai, China) and respectively cultured in a Leibovitz’s L-15 medium (Fuheng, Shanghai, China) and RPMI-1640 medium (Biological Industries, Israel). DLD1, RKO and NCM460 cells were gifts from Changhong Miao of Fudan University Shanghai Cancer Center. Cell culture mediums were supplemented with 10% fetal bovine serum (FBS; Yeasen, Shanghai, China) and 1% penicillin–streptomycin (Yeasen) in a humidified incubator with 5% CO₂ at 37 °C.

ALKBH5, FTO, METTL3, METTL14, WTAP, HK2 and IGF2BPs knockdowns in addition to over-expression plasmids were synthesized by Shanghai Genomeditech Biotech Co. Ltd. 293 T cells were used for lentivirus packaging with PEIMAX treatment (Polysciences, USA). After infection with a concentrated virus and 5 μg/ml polybrene (Genomeditech) for 48 h, stably transfected cells were screened through treatment with 5 μg/mL puromycin (Yeasen). Western blotting was performed to verify the efficiency of transfection efficiency. All shRNA targets were listed in Additional file: Table S1. 6

For the cell counting kit 8 (CCK-8) assay, 5 × 10³ CRC cells were seeded in 96-well plates and counted 72 h in the respective media with 10% FBS. Cell proliferation was measured at 450 nm absorbance by adding 10 μL in CCK-8 (Yeasen) after 2 h. For the colony formation assay, 1 × 10⁴ cells were seeded in 6-well plates and cultured in respective media with 10% FBS for 7 days. Images were then captured after the cells were fixed with 4% formaldehyde for 30 min and stained with 2% crystal violet (Yeasen) for another 30 min. For the EdU assay, cells seeded in 96-well plates were cultured with 50 μM EDU for 2 h and then fixed with 4% paraformaldehyde. The cells were then incubated with 1 × Apollo reaction cocktail (RiboBo, Guangzhou, China) for 30 min after permeabilization with 0.5% Triton-X. After DNA staining with Hoechst33342, the images were finally visualized using a fluorescence microscope (Zeiss, Germany).

Cells were suspended in a medium without FBS and seeded in 8 μm pore inserts (Corning, USA) at a density of 2 × 10⁴ cells per insert. The inserts were placed into 24-well plates containing 800 μL of 30% FBS medium. After culturing for 48 h, the cells on the upper side of the filter were fixed with 4% formaldehyde for 30 min and stained with 2% crystal violet for another 30 min. The cells on the upper side were then wiped with swabs, while the cells on the lower side were imaged randomly under a microscope.

Total RNA was extracted from cells using TRIzol reagent (Invitrogen, USA), and cDNA was synthesized using 4xHifair^® III SuperMix plus (Yeasen) according to the manufacturer’s protocols. qRT-PCR was finally performed with a Roche machine using SYBR Green PCR master mix (Yeasen). GAPDH was used as the internal reference. The qPCR primers used in this study are listed in Additional file 6: Table S2.

Cells were lysed in NP40 buffer containing protease and phenylmethanesulfonyl fluoride (PMSF, 2 mM) for 30 min and then boiled with 1 × loading buffer at 105 °C for 10 min. The extracted proteins were separated using sodium dodecyl sulfate–polyacrylamide gel (SDS-PAGE) electrophoresis and transferred onto a nitrocellulose filter membrane. The membranes were blocked using 8% nonfat milk for 2 h and then washed three times with Tris-Buffered Saline Tween-20 buffer. The blocked membranes were incubated with primary antibodies at 4 ℃ overnight, followed by respective secondary antibodies (1:5000) at room temperature for 2 h. Antibody information is listed in Additional file: Table S3. The signals were developed using the image lab software with an enhanced chemiluminescence reagent kit (New Cell & Molecular Biotech, Suzhou, China). 6

N6-methyladenosine RNA methylation was quantified using an m6A RNA methylation quantification kit. Briefly, 300 ng of RNA were bound to wells and then incubated with capture antibody for 1 h. The mixture was then incubated with detection antibodies for 30 min and reacted with an enhancer solution for another 30 min at room temperature, the reaction signal was read using a microplate spectrophotometer at 450 nm.

Cells were lysed with RIP lysis buffer and then incubated with specific antibodies at 4 ℃ for 2 h. The mixture was then incubated with protein A/G magnetic beads at 4 °C overnight. The bound RNA was separated from the beads by washing with RIP buffer solution. The eluted RNA was finally extracted and reverse-transcribed into cDNA for further analysis using qRT-PCR.

Human CRC and paired adjacent normal tissues were soaked in paraffin and fixed with 4% formaldehyde. After cutting and dewaxing, the slices were incubated in citric acid antigen retrieval buffer and blocked with 3% BSA blocking buffer, followed by incubation with primary antibodies at 4 ℃ overnight. Next, slices were incubated with secondary antibodies for 1 h at room temperature after being washed three times. The DAB color-developing solution was added and hematoxylin staining was performed. Finally, the slices were dehydrated and images were taken randomly by scanning.

For glucose assay, the cell supernatant was collected, 20ul of the sample to be tested was added to each well, and 100ul of the working solution of the substrate was added and incubated at 37 ℃ for 5 min to read the OD value A1. Then each well was added with enzymic antibody working solution of 20ul, and the reaction plate was placed at 37 ℃ for 15 min after shaking and mixing. The absorption value A2 was measured at 550 nm with enzymic marker within 30 min. All OD values are calculated by subtracting A1 from A2. For ATP level assay, the culture solution was removed, and the lysate was added in the ratio of 200 µl to each well of the 6-well plate. The lysate was blown repeatedly to fully lysate the cells. After lysate, the supernatant was taken for subsequent determination after centrifugation at 12000 g at 4 ℃ for 5 min. Then, 100 µl of ATP detection solution was added to each well and placed it at room temperature for 3–5 min, so that all the background ATP is consumed, thus reducing the background. A sample of 20 µl was then added to the test hole and the value was measured with a chemiluminescence instrument.

For the mouse xenograft model, 2 × 10⁶ cells were injected subcutaneously into the flank regions of female BALB/c nude mice (4–5 weeks). The animal procedures were approved by the Animal Care and Use Committee of Nanjing Medical University. Five weeks after injection, the tumors were excised from the mice and the weight, width (W), and length (L) were measured. The volume (V) of each tumor was calculated using the formula: V = (W² × L/2).

All experiments were independently repeated at least three times. The means between groups were compared using unpaired or paired student’s t-tests. All data were presented in the form of mean ± SD and p < 0.05 was considered statistically significant.

To evaluate the expression of FTO and ALKBH5 in human CRC samples, immunohistochemistry (IHC) staining was used to visualize the results of the tissue microarrays (TMAs) that contained 36 samples (Fig. 1A). The positive percentages of FTO and ALKBH5 were significantly lower in CRC tissues than in normal tissues (Fig. 1B, C). The protein expression was also analyzed showing low expression levels of FTO and ALKBH5 in several CRC cell lines than that in normal epithelial cell lines NCM460 (Fig. 1D). Similarly, four representative CRC tissues showed lower FTO and ALKBH5 expression levels than the corresponding para-cancer tissues (Fig. 1E). As FTO is known as an obesity-associated protein, we aim to explore whether obese CRC patients are more susceptible to suffering from CRC through dynamic m6A changes. Interestingly, the protein expression levels of FTO and ALKBH5 were lower in the obese CRC patients (Fig. 1F). A high-fat environment was also achieved using a palmitic acid-treated medium instead of a normal medium. The results revealed that FTO and ALKBH5 expression levels significantly decreased with an increase in the concentration of palmitic acid (Fig. 1G).

To examine the effect of FTO expression on CRC growth, stably transfected cell lines with FTO knockdown and over-expression were constructed using lentiviral transfection technology. The efficiency of FTO knockdown and over-expression was verified by western blotting (Fig. 2A, Additional file 1: Figure S1A). Subsequently, CCK-8 assays were performed, which showed that FTO knockdown significantly promoted proliferation in CRC cells while FTO over-expression inhibited their proliferation (Fig. 2B, C, Additional file 1: Figure S1B, C). The results of EdU assays also demonstrated that FTO deficiency remarkably expedited the growth of CRC cells, whereas FTO up-regulation restricted the proliferation of CRC cells (Fig. 2D, E, Additional file 1: Figure S1D, E). Similar findings were obtained using the colony formation assay (Fig. 2F, G, Additional file 1: Figure S1F, G). Additionally, FTO knockdown promoted the migration and invasion of CRC cells, as indicated by the transwell assays (Fig. 2H–J, Additional file 1: Figure S1H–J). Therefore, the above results demonstrate that the increased expression of FTO in CRC cells suppresses cell proliferation, migration, and invasion in vitro.

To explore the role of ALKBH5 expression in CRC, we established a stable ALKBH5 knockdown and over-expression model in both HCT116 and SW620 cell lines using specific shRNAs and found that the protein levels of ALKBH5 were significantly inhibited and over-expressed respectively (Fig. 3A, Additional file 2: Figure S2A). The cell growth curve indicated that ALKBH5 knockdown accelerated cell proliferation in two CRC cell lines, however, ALKBH5 over-expression showed the opposite results (Fig. 3B, C, Additional file 2: Figure S2B, C). Similarly, the EdU assays showed an anti-proliferative role of ALKBH5 in CRC cell lines (Fig. 3D, E, Additional file 2: Figure S2D, E). Colony formation experiments revealed that ALKBH5 inhibition increased the number of colonies, whereas ALKBH5 over-expression suppressed colony formation (Fig. 3F, G, Additional file 2: Figure S2F, G). Additionally, the results from the transwell assay showed that ALKBH5 knockdown significantly facilitated cell migration and invasion of CRC cells, whereas, up-regulation had the opposite effect (Fig. 3H–J, Additional file 2: Figure S2H–J). In conclusion, these results show that ALKBH5 can hinder cell proliferation, migration, and invasion of CRC cells.

To investigate the underlying molecular mechanisms of FTO and ALKBH5 in CRC progression, we conducted RNA sequencing (RNA-seq) and m6A-modified RNA immunoprecipitation sequencing (MeRIP-seq) in CRC cells with stable FTO over-expression, ALKBH5 over-expression, and control cells. The differential mRNA and differential m6A peaks modified genes are shown in volcano plots (Fig. 4A, B). Through KEGG pathway enrichment analysis, FOXO signaling was identified as a common pathway (Fig. 4C, D). Furthermore, the common target gene HK2, regulated by FTO and ALKBH5, was identified through the intersection between RNA-seq and MeRIP-seq (Fig. 4E). Therefore, we verified whether HK2 is regulated by FTO and ALKBH5 in an m6A-dependent manner. m6A IP was conducted, which demonstrated the successful interaction between m6A and HK2 mRNA (Fig. 4F). Moreover, knockdown of FTO/ALKBH5 increased HK2 m6A modification, while over-expression them decreased this modifications (Additional file 5: Figure S5A).

The stable cell lines of FTO and ALKBH5 knockdown and over-expression were verified by RT-PCR (Fig. 5A, C). Moreover, we checked the m6A level after the up/down-regulation of FTO/ALKBH5. As expected, down-regulating FTO or ALKBH5 increased, while up-regulation decreased the total m6A level (Fig. 5B, D). The mRNA levels of HK2 also indicated a negative relationship between FTO and ALKBH5 (Fig. 5E, F). Moreover, the protein levels of HK2 were significantly increased when FTO and ALKBH5 were depleted, whereas over-expression of FTO and ALKBH5 inhibited HK2 protein levels (Fig. 5G). To verify the activation of the FOXO signaling pathway medicated by down-regulation of FTO and ALKBH5 in CRC cells, we analyzed the protein expression of FOXO1 through western blotting. These results showed that the FOXO signaling pathway was activated in FTO and ALKBH5 knockdown cell lines, whereas up-regulation of FTO and ALKBH5 inhibited the FOXO signaling pathway (Fig. 5G).

Additionally, rescue experiments further validated the functional relationship between HK2 and FTO/ALKBH5. We performed EdU and colony formation assays and found that FTO/ALKBH5 knockdown accelerated the proliferation of CRC cells, whereas HK2 knockdown in FTO/ALKBH5 stable knockdown CRC cells decreased cell proliferation (Fig. 5H–K). Additionally, we validated the effects of migration and invasion for HK2 in FTO and ALKBH5 stable knockdown CRC cells. The results showed that silencing HK2 led to the attenuation of cell migration and invasion induced by FTO/ALKBH5 knockdown (Fig. 5L–N). In addition to knocking down HK2, we also utilized hexokinase inhibitor, 2-Deoxy-D-glucose (2DG), to treat HCT116 cells. Similarly, 2DG restored the FOXO pathway activation (Additional file 3: Figure S3A) and increased cell proliferation (Additional file 3: Figure S3B), colony formation (Additional file 3: Figure S3C, D), migration and invasion (Additional file 3: Figure S3E–G) induced by FTO and ALKBH5 silence. Furthermore, treating HCT116 cells with 2DG alone also inhibited cell proliferation (Additional file 4: Figure S4B), colony formation (Additional file 4: Figure S4C–D), migration and invasion (Additional file 4: Figure S4E–F) by inactivating FOXO pathway (Additional file 4: Figure S4A). Overall, these results indicates that a deficiency in FTO and ALKBH5 functions as oncogenes via the promotion of FOXO1 mediated by HK2 in CRC.

To confirm the bio-function of FTO and ALKBH5 in vivo, we subcutaneously injected CRC cells over-expressing FTO and ALKBH5 as well as an empty vector into nude mice. The groups with over-expressed FTO and ALKBH5 showed a significant decrease in tumor growth, both in terms of tumor weight and volume, compared with those of the control group (Fig. 6A–C). Further IHC and western blot assay indicated that up-regulation of FTO and ALKBH5 decreased the level of the proliferation marker Ki67 level, as well as HK2 and FOXO1 in vivo(Fig. 6D–H). As HK2 is a rate-limiting enzymes of the glycolytic pathway, we also explore the effect of FTO and ALKBH5 on glucose metabolism. Results showed down-regulation of FTO and ALKBH5 consumed more glucose and produced more ATP (Additional file 5: Figure S5B, C). To study whether FTO/ALKBH5 has an effect on apoptosis, we performed flow cytometry, tunel staining and western blot for Bcl2 and Caspase3, however, it showed no significant difference with FTO/ALKBH5 over-expression (Additional file 5: Figure S5D–F). Together, these results demonstrate FTO and ALKBH5 inhibits tumor growth in vivo and may participates glucose metabolism.

Since HK2 was positively regulated by m6A methylation, we further explore which m6A reader participates in m6A methylation of HK2 mRNA. So, we searched in RM2Target database(http://rm2target.canceromics.org↗) and found IGF2BP family proteins may affect the m6A modification of HK2. To validate our hypothesis, we silenced IGF2BP1, IGF2BP2, and IGF2BP3 respectively in HCT116 cells, and QPCR indicated that only IGF2BP2 down-regulation led to HK2 decrease (Fig. 7A). Then, RIP-QPCR was performed using anti-IGF2BP2 to confirm the interaction between HK2 mRNA and IGF2BP2 in CRC cells. Notably, IGF2BP2 also bound to the mRNA of HK2 in CRC cells (Fig. 7B). Next, we performed western blot and the results showed both HK2 and FOXO1 were inhibited by IGF2BP2 knockdown (Fig. 7C). Moreover, cell functional assay demonstrated IGF2BP2 silence significantly decreased cell proliferation (Fig. 7D), colony formation (Fig. 7E–F) as well as cell migration and invasion of CRC cells (Fig. 7G–I). In addition, we also knocked down IGF2BP2 in FTO and ALKBH5 down-regulation stabled-transfected cells. The results indicated that IGF2BP2 silencing decreased the malignant behaviors of CRC cells in proliferation (Fig. 7J), colony formation (Fig. 7K–L), migration, and invasion (Fig. 7M–O) induced by FTO and ALKBH5 silence. In summary, FTO and ALKBH5 regulates HK2 in m6A-IGF2BP2 dependent manner.

We assumed whether FTO and ALKBH5 could influence other m6A regulators such as METTL3, METTL14, and WTAP expression. Therefore, the levels of m6A writers were detected in FTO/ ALKBH5 knockdown cells and it revealed that METTL3 was negatively correlated with FTO and ALKBH5, while METTL14 was positively correlated with FTO and ALKBH5 (Fig. 8A, B). Subsequently, we investigated the bio-function of METTL3 and METTL14 in CRC cells. Stable CRC cell lines with METTL3 and METTL14 knockdown were constructed and examined by western blot (Fig. 8C, K). CCK-8 assays demonstrated that METTL3 knockdown significantly decreased CRC cell proliferation (Fig. 8D). Meanwhile, both colony formation and EdU assays showed that METTL3 silencing inhibited the proliferation of CRC cells (Fig. 8E–H). Moreover, transwell assays indicated that METTL13 weakened the migration and invasion of CRC cells (Fig. 8I, J). We also knocked down METTL14 in CRC cells using shRNA. The results demonstrated that METTL14 silencing increased the malignant behaviors of CRC cells in proliferation (Fig. 8L–N), colony formation (Fig. 8O, P), migration, and invasion (Fig. 8Q, R). Furthermore, western blot indicated that down-regulation of METTL3 suppressed HK2 thus inhibiting FOXO pathway, while silence METTL14 promoted FOXO pathway through facilitation HK2 (Additional file 4: Figure S4G).

Animal study was approved by Institutional Animal Care and Use Committee (IACUC)of Nanjing Medical University. The discarded tumor tissue was obtained with patients inform and was approved by The First Affiliated Hospital of Nanjing Medical University(Jiangsu Province Hospital).

Not applicable.

The authors have declared that no competing interest exists.

All data generated or analyzed during this study are included in this published article.