Background

In the world, colorectal cancer (CRC) ranks third in terms of prevalence and second in terms of mortality [1, 2]. The incidence of CRC is increasing annually, owing to recent changes in lifestyle and dietary structures [3]. Early detection, timely diagnosis, and radical surgery are crucial for the successful treatment of CRC. However, early-stage clinical symptoms are not often apparent. Most CRC patients are diagnosed at an advanced stage where surgery is not an option. Presently, the gold standard screening methods for CRC are colonoscopy and tissue biopsy [4]; however, these methods are not practical for large-scale screening for CRC owing to their high cost and poor patient compliance [5]. Moreover, CRC is commonly detected using blood-based biomarkers, such as carcinoembryonic antigen (CEA), carbohydrate antigen 125 (CA125), and carbohydrate antigen 19-9 (CA19-9) [6, 7]. However, the diagnostic performance of these biomarkers is unsatisfactory owing to low diagnostic sensitivity or specificity [8]. As a result, a blood biomarker for the clinical diagnosis of CRC is urgently needed.

N6-methyladenosine (m6A) is the most prevalent and conserved transcriptional modification [9, 10]. A methyltransferase complex composed of METTL3, METTL14, and WTAP catalyzes this modification. Additionally, FTO and ALKBH5 are m6A demethylases that regulate the reversibility of the m6A modification. Dysregulated m6A-related genes have been reported in CRC cells [11]. Low FTO expression in patient-derived CRC cell lines increases m6A levels in mRNAs, resulting in enhanced tumorigenicity and chemoresistance in vivo [12]. ALKBH5 plays an antitumor role in CRC by increasing the stability of FOXO3 by attenuating the level of its m6A modification, and FOXO3 targets miR-21 and promotes the expression of SPRY2, providing a new direction for CRC therapy [13]. Moreover, the levels of m6A and METTL3 were increased in CRC tissues, and high m6A or METTL3 levels predict poor prognosis [14]. Additionally, CRC tumorigenesis can be facilitated by the activation of the glycolysis pathway by m6A methylation [15]. Thus, the modification of m6A is crucial in the development and progression of CRC.

Patients with cancer, including gastric cancer [16], breast cancer [17], and non-small-cell lung carcinoma [18], have elevated levels of m6A in their peripheral blood (PB). The levels of m6A in PB correlate with tumor stage and can be used as diagnostic biomarkers. Moreover, these patients exhibit lower expression of m6A demethylase genes FTO and ALKBH5 in PB. However, the diagnostic potential of the m6A modification for patients with CRC has not been investigated.

In the current study, we aimed to investigate the diagnostic potential of levels of PB RNA m6A levels for CRC by comparing the same in HCs and patients with CRC. In addition, we demonstrated in vitro that CRC cells could regulate the levels of m6A in PB mononuclear cells (PBMCs) and detected the levels of PB RNA m6A in a mouse model of CRC.

Materials and methods

Sample collection

PB samples were collected from 78 patients with CRC and 44 HCs without a history of primary or chronic diseases at the Fujian Union Hospital using EDTA tubes. All patients with CRC were diagnosed based on histopathology via endoscopic examination or biopsy, and PB samples were obtained at diagnosis before surgery or radio/chemotherapy. The demographic characteristics of patients with CRC and HCs are listed in Table 1 and Additional file 1: Table S1. Ethical approval (2021QH036) was granted by the Ethics Committee of Fujian Union Hospital.

Table 1 Correlation between the levels of m6A and clinicopathological characteristics in CRC
Full size table

After blood collection, 1 mL of EDTA + blood was treated with 3 mL erythrocyte lysate (Cat. No. R1010; Solarbio) twice to obtain leukocytes, then 1 mL of Trizol reagent (Cat. No. 15596-026; Invitrogen) was added to stabilize the RNA, and the samples were preserved at − 80 ℃ until RNA extraction.

RNA isolation

Total RNA was extracted using the TRIzol reagent. The integrity of RNA was evaluated using agarose gel electrophoresis. RNA yield and purity were measured using the NanoDrop 1000 (Gene Company Limited, Hong Kong, China).

Quantitative real-time polymerase chain reaction (qRT-PCR)

Quantitative RT-PCR was performed using the PerfectStart Green qPCR SuperMix (Cat. No. AQ601-04; TransGen Biotech, Beijing, China) and a 7500 Real-Time PCR System (Thermo Fisher Scientific). The cycling parameters were as follows: 94 ℃ for 30 s, followed by 40 cycles of 94 ℃ for 5 s, 60 ℃ for 15 s, and 72 ℃ for 10 s. To calculate the ΔCq values, the Cq value for the human actin beta [ACTB] reference gene and the mice actin beta [Actb] gene was subtracted from the original Cq value. Normalization of targeted gene expression in each independent sample was performed using the reference (ACTB) gene. The primers for the target genes are listed in Additional file 1: Table S2. The 2−ΔΔCt method was used to calculate the absolute expression.

Quantification of m6A in PB RNA

The EpiQuik m6A RNA Methylation Quantification Kit (Colorimetric; EpiGentek, Farmingdale, NY, USA) was used to measure m6A levels in total RNA by the manufacturer’s instructions. First, 80 µL binding solution was added to the assay well, and then 200 ng RNA. Following a 90-min incubation at 37 ℃, the plates were washed thrice, and the assay wells were sequentially treated with diluted solutions of the primary and detection antibodies and the enhancer. Subsequently, the color reaction was initiated by adding the developer and stop solutions to each well, and the absorbance was measured at 450 nm wavelength. Finally, the standard curve was generated to determine the m6A levels.

Cell lines and cell experiments

The human CRC cell line (HCT116) was obtained from the Chinese Academy of Sciences (Beijing, China). SW480 and MC38 murine colon adenocarcinoma cells were kindly provided by Dr. Mi Zhang of the Basic Medicine of Fujian Medical University (Fuzhou, China). PBMCs were isolated from healthy human PB using the human PB lymphocyte separation medium. SW480 and HCT116 cells were cultured in RPMI 1640 medium (Gibco, Thermo Fisher Scientific) supplemented with 10% fetal bovine serum and 1%Pen/Strep (Thermo Fisher Scientific), and PBMCs were cultured in human PB lymphocyte particular medium in an incubator at 37 ℃ and 5% CO2.

Six-well Transwell chambers (0.4-µm pores, Corning Transwell; Corning Inc, Corning, NY, USA) were used for the co-culture assay. A total of 5 × 105 HCT116 or SW480 cells were seeded in the lower Transwell chamber, and 1 × 106 PBMCs were seeded in the upper chamber of the co-culture system for 48 h. PBMCs were collected separately for qRT-PCR and ELISA analyses.

Animal study

C57BL/6 male mice (6–8-week-old) were purchased from the Sibeifu Animal Center (Beijing, China). All animal experiments were conducted following the protocols approved by the Fujian Medical University of Medicine Policy on the Care and Use of Laboratory Animals. MC38 colon carcinoma cells were injected in 100 µL phosphate-buffered saline (PBS) subcutaneously into the right flank of C57BL/6 mice (n = 14, male) to construct an MC38 murine colon carcinoma cancer model (MC38 cancer model). C57BL/6 mice were also injected with 100 µL of PBS into their right flank as the control group (n = 8). Mice were euthanized 15 day after cells/PBS injection or if the longest dimension of the tumor was as large as 1.5 cm within 20 day. Retroorbital blood was obtained from the mice immediately following euthanasia using EDTA tubes.

Statistical analysis

Statistical analysis was conducted using IBM SPSS Statistics 20.0 (IBM Corp., Armonk, NY, USA) and GraphPad Prism version 7.0 (GraphPad Software, San Diego, CA, USA). Differences between groups were analyzed using an unpaired 2-tailed parametric Student’s t-test for normally distributed data. Otherwise, the nonparametric Mann–Whitney U tests were used to analyze the data. Data were analyzed using a one-way analysis of variance followed by Bonferroni-corrected posthoc tests to allow for comparison of more than two groups. A receiver operating characteristic (ROC) curve was plotted to determine the diagnostic value of the biomarker. The standard deviations are represented by error bars. Statistical significance was set at P < 0.05.

Results

Levels of m6A in the PB of patients with CRC and HCs

The EpiQuik m6A RNA Methylation Quantification Kit was used to determine the levels of PB RNA m6A of patients with CRC (n = 78) and HCs (n = 44). As depicted, patients with CRC had significantly higher PB RNA m6A levels than HCs (P < 0.0001) (Fig. 1A). Moreover, we investigated whether PB RNA m6A levels could be used to distinguish between different pathological stages of CRC (stage I, n = 13; stage II, n = 20; stage III, n = 31; stage IV, n = 12). As depicted, PB RNA m6A levels of the stage II group were significantly higher than those of the stage I group (P < 0.05) (Fig. 1B). However, PB RNA m6A levels of patients with CRC with or without metastasis were of no significant difference (P > 0.05; Fig. 1C).

Fig. 1
figure 1

The levels of N6-methyladenosine (m6A) in RNA of the peripheral blood of healthy controls (HCs) and patients with colorectal cancer (CRC). A The m6A levels of peripheral blood RNA in HC (n = 44) and patients with CRC (n = 78); B m6A levels in peripheral blood RNA of patients with different clinical stages of CRC (stage I, n = 13; stage II, n = 20; stage III, n = 31; stage IV, n = 12); C the m6A levels of peripheral blood RNA in patients with CRC (n = 8) before surgery and after seven days of surgery; D comparison of m6A levels in peripheral blood RNA between patients with CRC with (n = 16) and without (n = 56) metastasis. The bars represent the mean ± standard deviation (SD) of the results from three replicate measurements; *P < 0.05, ****P < 0.0001. HC: healthy control, CRC: colorectal cancer

Full size image

Moreover, we analyzed PB RNA m6A levels in patients with CRC (n = 8) before and after surgery (after 7 day) to determine whether m6A levels could be used as a potential follow-up biomarker. The results suggested that PB RNA m6A levels in patients with CRC significantly decreased after surgery (P < 0.05; Fig. 1D). Collectively, these results indicated that PB RNA m6A levels in patients with CRC are upregulated with the progression of CRC and may serve as a potential follow-up biomarker after surgery for patients with CRC.

Levels of m6A in PB RNA exhibit potential as a biomarker for patients with CRC

To determine the diagnostic value of PB RNA m6A levels for patients with CRC, the ROC curve was plotted. As depicted, the area under the curve (AUC) for PB RNA m6A levels was 0.886 (95% confidence interval CI 0.799–0.974; P < 0.0001) (Fig. 2A), which could be used to distinguish patients with CRC from HCs. PB RNA m6A had an optimal cutoff value of 0.236 (diagnostic sensitivity 84.7%; diagnostic specificity 80%) (Fig. 2B). In comparison to CEA (AUC 0.834) and CA199 (AUC 0.666), the diagnostic value of PB RNA m6A was significantly better (Fig. 2C). Furthermore, combining PB RNA m6A, CEA, and CA199 levels improved the AUC of the ROC curve to 0.935 (95% CI 0.882–0.989, P < 0.0001) (Fig. 2D), indicating the excellent diagnostic potential for patients with CRC.

Fig. 2
figure 2

Clinical utility of m6A levels in peripheral blood RNA, carcinoembryonic antigen (CEA), and carbohydrate antigen 199 (CA199) in diagnosing CRC. A Receiver operating characteristic curve and B cutoff value for m6A for patients with CRC and HCs; C receiver operating characteristic curves for m6A, CEA, and CA199 alone or together, and D cutoff values for m.6A, CEA, and CA199 in combination for patients with CRC and HC. HC: healthy control, CRC: colorectal cancer, CEA: carcinoembryonic antigen, CA199: carbohydrate antigen 199

Full size image

Expression of FTO and ALKBH5 decreased in the PB RNA of patients with CRC

Next, we measured the mRNA expression of FTO and ALKBH5 in the PB RNA of patients with CRC and HCs. The results suggested that both FTO and ALKBH5 expression were significantly downregulated in the PB RNA of patients with CRC compared with those in HCs (P < 0.01) (Fig. 3A and B). These results implied that the downregulation of FTO and ALKBH5 expression may be associated with elevated PB RNA m6A levels in patients with CRC.

Fig. 3
figure 3

Expression levels of FTO and ALKBH5 in patients with CRC. A Quantitative real-time polymerase chain reaction for the relative mRNA expression of FTO and ALKBH5 (B) in the peripheral blood of patients with CRC (n = 26; n = 34) and HC (n = 13; n = 23). **P < 0.01. CRC: colorectal cancer

Full size image

Co-culture with CRC cells increases m6A levels in PBMCs in vitro

To investigate the underlying mechanisms of the upregulation of PB RNA m6A levels in patients with CRC, CRC cells were co-cultured with PBMCs in a Transwell chamber. The total RNA of PBMCs showed an increase in m6A levels upon co-culture with CRC cells, namely HCT116 and SW480 (Fig. 4A). Similarly, co-culture with CRC cells decreased mRNA expression of FTO and ALKBH5 in PBMCs (Fig. 4B and C). These results indicate that the upregulation of m6A levels and downregulation of FTO and ALKBH5 expression in PBMCs could be influenced by CRC cells.

Fig. 4
figure 4

Levels of m6A and demethylases in blood cells cocultured with CRC cells. A m6A levels in PBMCs cocultured with or without CRC cells for 48 h; B mRNA levels of FTO in PBMCs cocultured with or without CRC cells for 48 h; C mRNA levels of ALKBH5 in PBMC cells cocultured with or without CRC cells for 48 h. **P < 0.01, ****P < 0.0001. CRC: colorectal cancer, PBMC: peripheral blood monocular cells

Full size image

High m6A levels in the PB-RNA of an MC38 cancer model

To assess whether CRC cells could increase m6A levels in PB-RNAs in vivo, we constructed an MC38 CRC model and measured PB RNA m6A levels. A significant increase in PB-RNAs m6A levels was observed in the MC38 cancer group when compared to the control group (P < 0.0001) (Fig. 5A). Moreover, the mRNA expression of FTO and ALKBH5 in the PB of MC38 cancer model mice was also decreased compared with those in the PB of the control mice (P < 0.05) (Fig. 5B and C). These results indicate that the MC38 cancer model exhibited increased PB RNA m6A levels, which was also accompanied by the downregulation of FTO and ALKBH5 expression in vivo.

Fig. 5
figure 5

Levels of m6A and demethylases in the peripheral blood RNA of C57 mice. A Levels of m6A in the total peripheral blood RNA of C57 mice, control (n = 8), and MC38 cancer model (n = 14); B, C relative mRNA levels of FTO (B) and ALKBH5 C in the blood of C57 mice. *P < 0.05, ****P < 0.0001. CRC: colorectal cancer

Full size image

Discussion

Early diagnosis is crucial for improving CRC prognosis. Current invasive or noninvasive screening methods fail to achieve satisfactory early screening results owing to their limitations. Although multiple randomized controlled trials have shown that guaiac-based fecal occult blood testing and sigmoidoscopy reduce CRC mortality, their effectiveness in reducing CRC mortality may be diminished if patients do not repeat screening [19,20,21,22]. The serological markers CEA and CA199, currently used to diagnose CRC, have low specificity and sensitivity, both individually and in combination [23, 24]. Therefore, identifying simple and novel early diagnostic biomarkers for CRC is urgently required. In the current study, we observed that PB RNA m6A levels could effectively differentiate patients with CRC from HCs. The diagnostic value of PB RNA m6A levels was significantly improved when combined with the biomarkers CEA or CA199.

To date, few studies have investigated the diagnostic value and underlying mechanism of PB RNA m6A levels in patients with CRC. Our study observed that PB RNA m6A levels in patients with CRC were significantly increased compared with those in HCs, which is consistent with the findings of Xie et al. [25]. However, our results showed no significant differences in PB RNA m6A levels in patients with CRC with or without distant metastasis. The growth and progression of cancer are regulated by crosstalk between writers, readers, and erasers of m6A [26]. Although m6A levels may vary in different tumors, previous reports have suggested that PB RNA m6A levels in patients with gastric cancer [16], non-small cell lung carcinoma [17], breast cancer [18], and rheumatoid arthritis [27] were much higher than those in the corresponding HCs. Therefore, further investigating the diagnostic value of PB RNA m6A levels in other tumors and diseases is crucial.

Our results indicated that PB RNA m6A levels had a promising diagnostic potential for CRC. The AUC of m6A (0.886) for discriminating patients with CRC from HCs was the highest, followed by those of CEA (0.825) and CA199 (0.671). Moreover, CEA and CA199 in combination with PB RNA m6A improved the AUC to 0.935 in patients with CRC. Previous studies and meta-analyses have demonstrated that some miRNAs and circRNAs could serve as potential biomarkers for CRC; however, the AUC of the RNAs was not considered high. For example, a meta-analysis including six studies revealed that the AUC of miR-92a in the diagnosis of CRC was 0.772 [28]. Another meta-analysis, including 18 studies involving 2021 individuals, reported that the AUC of circRNA in the diagnosis of CRC is 0.81 [29]. In addition, our results showed that PB RNA m6A levels in postoperative CRC patients were decreased following surgery, indicating its potential as a biomarker for postoperative monitoring. However, further studies using more clinical samples are required to confirm whether it can be used as a potential follow-up biomarker. Overall, our results indicate that PB RNA m6A levels can act as a better diagnostic biomarker for CRC than the currently used biomarkers.

In the current study, we observed that the expression of both FTO and ALKBH5 was decreased in patients with CRC compared with that in HCs. There may be a relationship between the increased PB RNA m6A levels in patients with CRC and the downregulation of FTO and ALKBH5. Moreover, co-culture with CRC cells results in an increase in m6A levels and a decrease in FTO and ALKBH5 expression in PBMCs. A previous study has reported that imbalanced regulation of m6A strongly confers immune destruction and tumor evasion [30]. Consistently, in our study, PB RNA m6A levels in the MC38 cancer model were significantly increased and were accompanied by a decrease in the expression of FTO and ALKBH5. These results indicate that increased PB RNA m6A levels induced by CRC may be owed to the decreased expression of FTO and ALKBH5.

Our study has some limitations as no significant correlation was observed between PB RNA m6A levels and expressions of FTO or ALKBH5 in the total RNA of PB cells. The possible reasons for the lack of correlation could be: (1) other unidentified methylases and demethylases that require further exploration [31] are also involved in the regulation of m6A; (2) the regulation of m6A in the total RNA of PB cells may require the participation of methyltransferases; (3) interactions between methyltransferases and demethylases and their regulatory factors may also contribute to the changes in m6A level. Therefore, changes in PB RNA m6A levels may not solely depend on the expression of the demethylases investigated in this study. Further investigation is needed to understand the mechanism behind the upregulation of m6A levels and the downregulation of demethylases induced by tumors.

Conclusions

In conclusion, our results revealed that PB RNA m6A levels in patients and the mouse model of CRC are accompanied by downregulation of the demethylases FTO and ALKBH5 at the transcriptional level. Our research suggests that PB RNA m6A levels are a promising biomarker for CRC diagnosis.