FTO downregulation mediated by hypoxia facilitates colorectal cancer metastasis

Fat mass and obesity-associated protein (FTO), an N6-methyladenosine (m6A) demethylase, participates in tumor progression and metastasis in many malignancies, but its role in colorectal cancer (CRC) is still unclear. Here, we found that FTO protein levels, but not RNA levels, were downregulated in CRC tissues. Reduced FTO protein expression was correlated with a high recurrence rate and poor prognosis in resectable CRC patients. Moreover, we demonstrated that hypoxia restrained FTO protein expression, mainly due to an increase in ubiquitin-mediated protein degradation. The serine/threonine kinase receptor associated protein (STRAP) might served as the E3 ligase and K216 was the major ubiquitination site responsible for hypoxia-induced FTO degradation. FTO inhibited CRC metastasis both in vitro and in vivo. Mechanistically, FTO exerted a tumor suppressive role by inhibiting metastasis-associated protein 1 (MTA1) expression in an m6A-dependent manner. Methylated MTA1 transcripts were recognized by an m6A “reader”, insulin-like growth factor 2 mRNA binding protein 2 (IGF2BP2), which then stabilized its mRNA. Together, our findings highlight the critical role of FTO in CRC metastasis and reveal a novel epigenetic mechanism by which the hypoxic tumor microenvironment promotes CRC metastasis.

horizontal line. Washed the plate with PBS, added serum-free medium and the wound 23 healings were observed at 0, 12, 24 h by a microscope. Image J software was used to 24 calculate the healing area of scratches.

25
Transwell migration and invasion assays 26 Transwell migration assays were performed using a 24-well Transwell chamber system 27 (Corning, USA), and invasion assays were performed using a 24-well invasion chamber were mapped to the human genome version hg19 by using HISAT2 software [2]. The 44 expressed genes (DEGs) between the FTO knockdown and control groups were 46 analyzed by DESeq2. DEGs with log2 > 0 or < 0 and P value < 0.05 were defined as 47 upregulated genes or downregulated genes.

48
Vector and m 6 A mutation assays 49 The online tool SRAMP (http://www.cuilab.cn/sramp/) was used to predict potential 50 m 6 A modification sites on MTA1 RNA sequences. Full-length MTA1 transcripts, the 51 MTA1 CDS region, the MTA1 three prime untranslated region (3′-UTR), and the m 6 A 52 motif mutant CDS and 3′-UTR regions were cloned into pcDNA3.1 constructs by 53 OBiO Technology (Shanghai) Corp., Ltd. and used for the RNA pulldown assay.

56
RiboMinusTM Eukaryote Kit v2 (A15020, Invitrogen) was used for RNA isolation and 57 purification. Total RNA (50 µg) was sheared to ~100 nt fragments by RNA 58 Fragmentation Reagents (AM8740, Invitrogen), and 1/10 of the fragment pool was 59 conserved as an input control for further sequencing. The fragmented RNA and anti-60 m 6 A antibody (202003, Synaptic Systems) were incubated together for 1 hour at 4°C.

61
PierceTM Protein A/G Magnetic Beads (88803, Thermo Scientific) were washed and 62 resuspended in immunoprecipitation (IP) buffer. Then, the fragmented RNA and m 6 A 63 antibody mixture were added to the beads for conjugation by gentle rotation at 4°C 64 overnight. The m 6 A antibody was digested using proteinase K digestion buffer. MeRIP-65 seq library preparation and sequencing on an Illumina HiSeq 2500 system were carried 66 algorithm from the input (RNA-seq) and m 6 A IP (MeRIP-seq) sequencing libraries with 69 a q-value (P value with false discovery rate (FDR) correction) of 0.05. The m 6 A 70 modification peaks that appeared in at least two samples were retained for further 71 analysis. Homer was used for m 6 A peak annotation, BEDTools was used for m 6 A read 72 extraction[6], and our custom Perl script was used for conversion to RPKM. The data 73 were normalized using the "limma" package in R software [7].