1. Introduction

In 2003, Torlakovic et al. [1] described unusual proliferation of colorectal serrated polyps and introduced the terms “sessile serrated polyp” and “sessile serrated adenoma (AD)” to classify these lesions. Currently, the World Health Organization [2] recommends referring to these lesions as sessile serrated lesions (SSLs). SSLs are regarded as early precursors in the serrated neoplasia pathway, leading to colorectal carcinomas characterized by microsatellite instability-high [3]. The serrated pathway is characterized by mutations in BRAF, disruptions to the Wnt/β-catenin signaling pathway, and widespread methylation of CpG islands [4–7]. This pathway is believed to differ from the traditional AD–carcinoma sequence [8], in which ADs progress to invasive colorectal carcinomas through a series of genetic changes, including mutations in APC and KRAS [6, 7]. Different apoptotic activities and p21WAF1/CIP1 expressions have been found in SSLs compared with ADs [9]. Moreover, there were differences in the expressions of HIF1α, EphB2, and DNA repair proteins and in the frequency of BRAF/KRAS mutations in those two lesions [10]. Wnt/β-catenin signal activation mediated by the methylation of SFRP4, MCC, and AXIN2 was dissimilar in the candidates between the serrated pathway and conventional AD routes [11].

MicroRNAs (miRNAs) are small noncoding RNAs consisting of 18–25 nucleotides that play an important role in numerous biological processes, including cell proliferation, differentiation, and apoptosis [12], as well as in the development and progression of colorectal carcinomas [13–22]. The expressions of miR-20a and miR-21 have been evaluated in the tissues of candidates for traditional colorectal carcinogenesis [13–17]. Furthermore, several studies have indicated that miR-181b is implicated in the clinical prognosis of colorectal cancer [18–20]. Alternatively, miR-93, a negative regulator of colorectal carcinogenesis, represses the Wnt/β-catenin signaling pathway [21, 22]. In addition, a study on benign colorectal serrated lesions showed a differential expression of miR-21 and miR-181b in SSLs and hyperplastic polyps [23].

To date, very few studies have analyzed the association of specific miRNAs with the serrated neoplasia pathway and traditional AD–carcinoma sequence. This study was aimed at quantitatively assessing the expression of miRNAs, including miR-20a, miR-21, miR-93, and miR-181b, and elucidating the potential roles of the two distinct pathways in colorectal carcinogenesis.

2. Methods

2.2. Histological Criteria

Histological diagnosis followed the WHO diagnostic criteria for SSLs [2], focusing on features such as crypt serrations, basal crypt dilation, distorted crypts, and growth along the muscularis mucosae. The histologic features of HD were assessed according to a previous description [11] as follows: a tubular, tubulovillous, or fused glandular pattern, mimicking conventional adenomatous HD or a serrated glandular pattern, preserving the serrated or saw-toothed structure with infolding of the crypt epithelium, which consisted of cuboidal and eosinophilic dysplastic cells with substantially larger nuclei and irregular thickening of the nuclear membrane (so-called serrated-type HD). For SSLs with HD or SC, either a normal SSL component at the lesion edge or an abrupt transition to dysplasia or carcinoma within a single tissue fragment was required. Two experienced gastrointestinal pathologists (H.M. and T.Y.) independently reviewed the samples, which were included only if they confirmed the diagnosis. Typical morphologies of the representative cases of the SSL series are shown in Figure 1.

3. Results

3.2. Expressions of miRNAs

Detailed data on the expression levels of miR-20a, miR-21, miR-93, and miR-181b in the lesions are presented in Table 3. Figure 2 illustrates the expression levels of these miRNAs in the SSL series, AD lineage, and normal mucosa.

Table 3. Expression levels of miRNAs in colorectal lesions and normal mucosa.

	SSLs (n = 25)	SSL-HD (n = 19)	SSL-SC (n = 13)	ADs (n = 19)	AD-HD (n = 26)	AD-SC (n = 25)	Normal mucosa (n = 15)
miR-20a	0.68 ± 3.09	0.85 ± 0.76	0.53 ± 0.29	0.65 ± 1.19	0.77 ± 0.31	1.48 ± 1.46	0.90 ± 0.43
	(0.24–16.07)	(0.21–3.25)	(0.28–1.35)	(0.26–5.73)	(0.27–1.52)	(0.42–7.42)	(0.16–1.75)
miR-21	2.05 ± 2.04	2.07 ± 2.68	3.22 ± 1.67	0.59 ± 1.24	1.23 ± 0.72	3.04 ± 2.04	0.98 ± 2.72
	(0.53–7.82)	(0.60–9.75)	(0.53–6.64)	(0.21–5.83)	(0.37–3.51)	(0.55–8.17)	(0.54–11.18)
miR-93	1.23 ± 1.22	1.27 ± 1.06	0.92 ± 0.43	1.88 ± 2.74	2.59 ± 1.19	3.18 ± 1.91	1.00 ± 0.73
	(0.57–5.26)	(0.36–3.17)	(0.29–1.71)	(0.48–10.30)	(1.18–6.53)	(0.98–9.81)	(0.31–2.90)
miR-181b	1.96 ± 1.60	1.17 ± 0.89	0.66 ± 0.39	1.07 ± 2.06	1.03 ± 0.66	1.45 ± 1.08	0.99 ± 0.91
	(0.68–8.61)	(0.26–3.84)	(0.35–1.80)	(0.48–8.81)	(0.44–3.12)	(0.54–5.90)	(0.19–3.97)

• Note: Data are represented as median ± SD (range).

4. Discussion

This study sheds light on the differential expressions of miR-20a, miR-21, miR-93, and miR-181b in candidates along the serrated and traditional AD–carcinoma routes. In colorectal carcinogenesis, miR-20a regulates BID, a gene associated with apoptosis, and affects sensitivity to TRAIL, a TNF superfamily member that induces exogenous apoptosis [14]. Moreover, miR-20a promotes epithelial–mesenchymal transition by suppressing the expression of genes such as SMAD4 and GABBR1, thereby facilitating invasion in colorectal cancer [15]. In this study, high miR-20a expression was observed in AD-SC, suggesting that miR-20a may be involved in the early step of invasion through its association with antiapoptotic mechanisms and epithelial–mesenchymal transition in the AD–carcinoma sequence.

We additionally explored the associations between the expression of miRNAs and methylation of Wnt/β-catenin signaling–related genes, AXIN2 and MCC, in the SSL series, including 25 SSLs, 9 SSL-HDs, and 3 SSL-SCs (Table S1). A significant positive correlation was observed between miR-20a expression and MCC methylation, suggesting that miR-20a may correlate with Wnt/β-catenin signaling activation exclusively in the MCC-methylated SSL series. Considerable evidence has shown that miRNAs are implicated in the activation of Wnt/β-catenin signaling [13], including overexpression of miR-942 inducing this signaling in AD–carcinoma progression [14]. Further analysis is needed to clarify the relationship between the Wnt/β-catenin signaling pathway and miRNA expression in SSL carcinogenesis.

The expression of miR-21 is frequently upregulated in colorectal cancer, and the target genes regulated by miR-21 include PDCD4, RhoB, and TGFβR2 [16]. In the present study, miR-21 levels increased from ADs through AD-HD to AD-SC, whereas no significant changes were observed in SSLs, SSL-HD, and SSL-SC. These results suggest that miR-21 is a positive regulator of the AD–carcinoma sequence but not of the serrated pathway. Previous studies have shown that KRAS mutations have significant links to the AD series [6, 7]. KRAS mutation was infrequent in colorectal cancers with high miR-21 expression compared with those with low-level expression [17]. These findings were discordant with miR-21 expression in the AD series in our study.

miR-93 suppresses CCNB1 protein expression, leading to cell cycle arrest and repression of ERBB2, p21, and VEGF, all of which play important roles in cell proliferation [21]. In colon cancer cells, miR-93 downregulated the Wnt/β-catenin pathway, which was confirmed by measuring the expression of β-catenin, AXIN, c-Myc, and cyclin D1 in this pathway [22]. Murakami et al. [11] described that upregulation of the Wnt/β-catenin signaling pathway, such as nuclear accumulation of β-catenin (Figure 1f), contributes to the serrated neoplasia pathway. In the current study, the expression of miR-93 was significantly lower in SSL-SC than in SSLs. Thus, suppression of miR-93 may enhance Wnt/β-catenin signaling in carcinogenesis of the serrated pathway.

miR-181a has strong tumor-promoting effects by inhibiting the expression of WIF-1 and has a potential role in inducing epithelial–mesenchymal transition [18]. miR-181b is associated with colorectal carcinogenesis by targeting PDCD4 [19], similar to miR-21 [16]. Epigenetic silencing of miR-181b contributes to tumorigenicity in colorectal cancer by targeting RASSF1A [20]. In a methylation-specific PCR study using 20 primers, including APC and RASSF1A, SSLs with higher histological grades showed more extensive methylation than those with lower grades [3]. In our study, miR-181b expression was significantly higher in SSLs but lower in SSL-SC than in normal mucosa, suggesting that the epigenetic silencing mechanism of miR-181b is associated with the early phase of invasion in the serrated pathway. Similar to our results, in a previous study, miR-181b was expressed at higher levels in SSLs than in normal mucosa [23]. Additionally, miR-181b expression levels were significantly higher in AD-SC than in AD-HD and normal mucosa. Accordingly, miR-181b may play diverse roles in the serrated pathway and AD–carcinoma sequence.

We found distinct differences in the expression of miR-20a, miR-21, and miR-181b between the SSL series and AD lineage. A schematic depiction of the differences in the expression of the four miRNAs compared with normal mucosa (baseline expression) between the two routes is shown in Figure 3. The expression of miR-93 showed a stepwise increase during tumor progression in the AD lineage, whereas the values did not change during this step in SSL carcinogenesis. The expression levels of miR-20a increased in the early phase of carcinoma invasion in the AD lineage but decreased in this phase in the SSL series. Recently, miR-21 and miR181a-2 were reported to be associated with carcinogenesis via the serrated pathway [24]. Colorectal cancers with high-level miR21 expression frequently harbor BRAF mutations and a CpG island methylator phenotype-high [17], which showed a significant link to the SSL series [5, 7, 10]. This is consistent with the results of our study representing that miR-21 expression is significantly higher in the SSL series than in normal mucosa. Sugai et al. [25] demonstrated no substantial overlap in the expression of miRNAs/messenger RNAs in colorectal cancers with microsatellite instability and precursor SSLs, and they speculated that the network patterns of miRNAs/messenger RNAs change during carcinogenesis in the serrated pathway.

The differential expression levels of miR-20a, miR-21, miR-93, and miR-181b that were observed in this study may have significant clinical implications. These miRNAs may potentially serve as noninvasive biomarkers for the early detection and risk stratification of colorectal neoplasms, particularly those arising from the serrated pathway. Given that serrated lesions—particularly SSLs—are often endoscopically subtle and thus easily missed, the use of circulating or stool-based miRNA assays may further enhance early detection strategies for these malignancies. Furthermore, the expression patterns of these miRNAs may help clinicians differentiate between serrated lesions and conventional ADs, thus facilitating personalized patient surveillance strategies. For example, higher expression levels of miR-21 and miR-181b, which are associated with microsatellite instability-high or high-risk lesions, might flag patients who would benefit from more intensive follow-ups.

Our study focused on a selected panel of four miRNAs (miR-20a, miR-21, miR-93, and miR-181b). Although this targeted approach may not have fully captured the full complexity of the serrated neoplasia pathway, these miRNAs were chosen based on previous evidence suggesting their involvement in colorectal cancer progression—particularly in the contexts of serrated and microsatellite instability [13–23]. Future studies should consider a broader range of miRNAs to further elucidate the molecular basis of colorectal tumorigenesis. For example, miR-31 and miR-135b have both been implicated in the serrated neoplasia pathway—particularly in lesions with high microsatellite instability—high and BRAF mutations—whereas miR-143 and miR-145 are commonly associated with AD–carcinoma sequences and KRAS or APC mutations [26]. Including these miRNAs may enhance our understanding of distinct carcinogenic pathways and support the development of miRNA-based diagnostic or prognostic tools.

This study had several limitations. First, the sample size of each subgroup was relatively small. Additionally, this study focused on the expression of only four miRNAs. Larger cohorts and analyses of more miRNAs are needed to validate the findings and ensure the reproducibility of the differences in miRNA expression. Second, the study relied on FFPE tissue samples, in which RNA was potentially degraded. However, miRNA expression profiles of FFPE tissues closely resemble those of fresh tissues, underscoring the suitability of FFPE tissues as appropriate resources for miRNA analysis [27]. miRNAs are less affected by the fixation/embedding process and are stable (intact) in FFPE samples, and this stability is largely independent of formalin fixation time and duration of tissue block storage [12].

In conclusion, the present study emphasized that the differential expression of miR-20a, miR-21, miR-93, and miR-181b is presumably related to genetic and/or epigenetic variations in candidates between the serrated neoplasia pathway and AD–carcinoma sequence, highlighting their potential roles in carcinogenesis. Our results pave the way for the development of miRNA-based biomarkers for colorectal lesions.

Ethics Statement

This study was approved by the Institutional Review Board and ethical committee of our hospital (reference number 2017166). The study was performed in accordance with the principles of the Declaration of Helsinki.

Consent

The need for informed consent was waived by the Institutional Review Board and ethical committee of Juntendo University School of Medicine.

Conflicts of Interest

The authors declare no conflicts of interest.

Author Contributions

Takashi Murakami is the guarantor of the integrity of the entire study. Takashi Murakami and Hiroyuki Mitomi contributed to the study concepts, study design, and literature research. Takashi Murakami, Naoki Tsugawa, Yudai Otsuki, Eiji Kamba, and Tomoyoshi Shibuya contributed to acquiring the clinical data. Hiroyuki Mitomi, Yuichiro Kadomatsu, Takuo Hayashi, Tsuyoshi Saito, and Takashi Yao pathologically reviewed the cases. Takashi Murakami and Hiroyuki Mitomi contributed to the clinical studies, experimental studies, and statistical analysis. Takashi Murakami contributed to manuscript preparation. Hiroyuki Mitomi contributed to manuscript editing. Takashi Yao and Akihito Nagahara contributed to the critical revision of the manuscript.

Funding

The study was supported the Japan Society for the Promotion of Science (10.13039/501100001691) (#24590429).