This study adhered to the principles of the Declaration of Helsinki and was approved by the Institutional Review Board/Ethics Committee of The University of Hong Kong (UW 21-114; approval date 2 August 2021). Written informed consent was obtained from all participants prior to enrolment.

Fresh surgically resected colorectal tumor tissues and matched adjacent normal mucosa were collected from 64 patients undergoing surgery at the Department of Surgery, Queen Mary Hospital, The University of Hong Kong. The series comprised normal colon (N = 64), colorectal adenoma (CoAd) (N = 13), non-metastatic CRC M0 (N = 47), metastatic CRC M1 (N = 17), and liver metastasis (N = 9). Immediately after resection, tissue specimens were snap-frozen in liquid nitrogen and stored at −80°C until analysis. In addition, serum samples were obtained from 26 patients with colorectal adenoma (prior to colonoscopy), 38 patients with CRC (pre-operative), and 36 healthy control individuals without a history of CRC or adenomatous polyps. Whole blood was centrifuged at 400× g for 10 min, and the clarified serum was aliquoted into sterile 2 mL tubes and stored at −80°C until analysis. Clinicopathological variables were retrieved from the patients’ medical records.

HCT116 (CCL-247) and HT29 (HTB-38) CRC cell lines were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). Cell line authentication was performed by Short Tandem Repeat (STR) profile analysis at Centre for PanorOmic Sciences (Hong Kong, China). No mycoplasma contamination was detected throughout the experiments. All cells used in these experiments were cultured in Dulbecco’s Modified Eagle Medium (DMEM) (Gibco, Cat. No. 11965-092, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS) (Gibco, Cat. No. 10099-141) and 1% penicillin–streptomycin (Gibco, Cat. No. 15140-122). Cells were kept at 37°C in a humidified incubator containing 5% CO₂. For transfection, 2 × 10⁵ cells were seeded into each well of a 12-well plate and transfected with piR-37524 inhibitors or negative control (NC) oligonucleotides using Lipofectamine 3000 (Invitrogen, Cat. No. L3000008, Waltham, MA, USA) according to the manufacturer’s protocol. The piR-37524 inhibitors and NC oligonucleotides were designed and synthesized by RiboBio (Guangzhou, China).

Total RNA was extracted from tissue specimens and serum samples with the mirVana™ miRNA Isolation Kit (Thermo Fisher Scientific, Cat. No. AM1560, Lenexa, KS, USA). All procedures were carried out in accordance with the manufacturer’s protocol. cDNA synthesis and quantitative real-time PCR were performed as previously described, with minor modifications [8]. Primer sequences used are listed in Table S1. For tissue specimens, piRNA expression levels were normalised to RNU6B, whereas serum piRNA levels were normalised to miR-92a-3p. Relative expression was expressed as ΔCt values.

Cell proliferation was quantified using an MTT assay (Invitrogen, Cat. No. M6494). After transfection, 1 × 10³ cells were seeded per well in 96-well plates and cultured for 24, 48, 72, and 96 h. 10 μL MTT solution (5 mg/mL) was added to each well, and the plates were incubated for a further 3 h. The resulting formazan crystals were dissolved in DMSO (Synth, Cat. No. D2650, Diadema, SP, Brazil), and absorbance at 570 nm (with 630 nm as reference wavelength for background subtraction) was measured using a Multiskan™ FC microplate reader (Thermo Fisher Scientific, Model 51119000).

Transfected cells were seeded in 6-well plates at a density of 1 × 10³ cells per well and cultured for 10–14 days until visible colonies formed. Colonies were then washed twice with PBS, fixed with 100% ice-cold methanol for 30 min, and stained with 0.2% crystal violet (Sigma-Aldrich, Cat. No. C0775, St. Louis, MO, USA) for 15 min at room temperature. Colony numbers were counted under a light microscope (Olympus, Model CKX53, Tokyo, Japan).

Cell migration was evaluated using a wound-healing assay. Transfected cells (2 × 10⁵/well) were seeded in 12-well plates and grown to a confluent monolayer. A linear scratch was introduced into the monolayer using a 10 µL sterile pipette tip, and detached cells were removed by washing with PBS. Fresh serum-free culture medium (DMEM, Gibco, Cat. No. 11965-092) was added, and images were captured at 0, 48, and 72 h using an Olympus CKX53 microscope (Evident Corporation, Model CKX53, Tokyo, Japan). Wound closure was quantified using ImageJ software (Version 1.54f, National Institutes of Health, Bethesda, MD, USA).

Total RNA was isolated from transfected cells using TRIzol reagent (Invitrogen, Cat. No. 15596026) following the manufacturer’s protocol. RNA integrity was assessed with an Agilent 2100 Bioanalyzer (Agilent Technologies, Model G2939BA, Santa Clara, CA, USA), and only samples with an RNA integrity number (RIN) > 7 were used for library preparation. Libraries were generated using the TruSeq RNA Sample Preparation Kit (Illumina, Cat. No. RS-122-2001, San Diego, CA, USA) and sequenced on an Illumina HiSeq 2500 platform. Sequencing reads were quality-filtered, and high-quality reads were aligned to the human reference genome (GRCh38) using HISAT2. Differential expression analysis was carried out with the DESeq2 R package (Version 1.38.3), applying a threshold of p < 0.05 and absolute log₂ fold change (FC) > 1. Differentially expressed genes (DEGs) were visualised in a volcano plot generated with ggplot2 in R. Overlaps among DEG sets were summarised using Venn diagrams generated with the online tool at https://bioinformatics.psb.ugent.be/webtools/Venn/.

Potential target genes of piR-37524 were predicted using miRanda software (http://www.bioinformatics.com.cn). Candidate interactions between piR-37524 and mRNA 3^′-UTRs were selected based on sequence complementarity and binding energy criteria.

20 µg total protein lysates were separated by 10% SDS–polyacrylamide gel electrophoresis at 120 V for 90 min and electrotransferred onto PVDF membranes (Bio-Rad, Cat. No. 1620177, CA, USA). Membranes were blocked in TBST containing 5% BSA (Sigma-Aldrich, Cat. No. A215) for 1 h at room temperature. Primary antibodies were used to incubate overnight at 4°C against E-cadherin (1:1000, Cell Signaling Technology, Cat. No. 3195, Danvers, MA, USA), β-catenin (1:1000, Cell Signaling Technology, Cat. No. 8480), Snail (1:1000, Cell Signaling Technology, Cat. No. 3879), TNFAIP3 (1:1000, ABclonal, Cat. No. A1741, Woburn, USA), and GAPDH (1:5000, Santa Cruz Biotechnology, Cat. No. sc-47724, Dallas, TX, USA). After three washes in TBST, membranes were incubated with horseradish peroxidase-conjugated secondary antibodies (1:10000, Invitrogen, Cat. No. 31460) for 1 h at room temperature. Protein bands were detected using enhanced chemiluminescence (ECL) reagents (Thermo Fisher Scientific, Cat. No. 32106).

For tissue and serum piRNA measurements, relative expression was expressed as −ΔCt values. The distribution of the data was assessed using the Kolmogorov–Smirnov or Shapiro–Wilk test. Differences between two groups were analysed using Student’s t test or the Mann–Whitney U test, as appropriate, and comparisons among more than two groups were evaluated using one-way ANOVA or the Kruskal–Wallis test. Receiver operating characteristic (ROC) curves were constructed to evaluate the diagnostic performance of serum piR-37524. A p-value < 0.05 was considered statistically significant. All statistical analyses were performed using GraphPad Prism version 8.0.1 (GraphPad Inc., San Diego, CA, USA).

To discover unexplored transcriptional addiction in CRC, we surveyed our previously published small RNA sequencing dataset generated from 18 CRC patients (18 CRC specimens and 17 tumor-adjacent non-cancerous tissues) [8] and identified 6 piRNAs (piR- 37524, piR-31143, piR-39614, piR-49145, piR-42111, and piR-35406) showing differential expression (log₂-fold change ±2) between CRC tissues and adjacent normal tissues (Fig. 1A–F). These piRNAs were selected based on their differential expression and relevance in cancer, as supported by existing literature (Table S2).

To validate our results from the next-generation sequencing dataset, we performed a preliminary profiling of the expression of these piRNAs in a small patient group comprising 6 primary CRC tissues and the corresponding adjacent non-malignant tissues. The expression of piRNAs within the 6 paired tumor and non-tumor resection specimens was normalized against the gene expression of RNU6B, and the data were represented (as−ΔCt (negative delta Ct). Only piR-37524 and piR-31143 demonstrated a statistically significant change in expression within the CRC tissues relative to the adjacent normal (piR-37524: p = 0.0081; piR-31143: p = 0.0186) (Fig. 2A). While piR-31143 had been previously explored in CRC (16), we focused on the novel piRNA, piR-37524.

To further expand our pilot results from the preliminary profiling, we performed a validation of the expression of piR-37524 using real-time PCR in a larger patient group comprising 64 primary CRC tissues and the corresponding adjacent normal resection specimens. The patient cohort had a median age of 67 years, with primary tumors located in the colon (71.8%), and rectum (including rectosigmoid region) (28.1%) (Table 1). Most patients were male (57.8%), and tumors were classified according to AJCC guidelines as stage I (4.6%), II (28.1%), III (40.6%), or IV (26.6%) (Table 1).

Our data demonstrated that piR-37524 was significantly overexpressed, by over 30-fold in primary CRC tissues relative to the adjacent normal tissues (median: −16.41 vs. −19.18; p < 0.001) (Fig. 2B), with 71.8% of patients showing a T/N >2-fold change in the expression of the piRNA. We subsequently performed a clinicopathological analysis to identify the correlation of the expression of piR-37524 with the overall development and progression of CRC within the patient cohort. Statistical analysis using unpaired Student’s t-test demonstrated that the expression of piR-37524 positively correlated with the primary tumor size (Fig. 2C). Specifically, 68% of patients with tumors ≥ 5 cm (39% of total) showed median-high piR-37524 expression, compared to 38.5% of patients with smaller tumors (<5 cm) (p = 0.0031) (Table 1). Poorly differentiated tumors also showed higher piR-37524 expression than well to moderately differentiated tumors (median: −15.19 vs. −16.55, p = 0.0264) (Fig. 2D). Expression of piR-37524 was also found to reflect the metastatic potential of the primary CRC tumors. About 70% of the patients exhibiting distant metastases showed a median-high expression of piR-37524, while only 42.5% of the non-metastatic patients expressed a median-high level of the piRNA (p = 0.0476) (Fig. 2E). Other clinicopathological factors did not correlate with piR-37524 expression.

Since the expression of piR-37524 correlated clinically with tumor size, differentiation, and metastasis, it was imperative to investigate this piRNA across multiple developmental stages of CRC to get a better understanding of the distinct roles of piR-37524 in the disease. Consequently, we profiled the expression of piR-37524 within CoAd (N = 13), non-metastatic CRC M0 (N = 47), metastatic CRC M1 (N = 17), as well as distant liver metastases (N = 9). Within the study population, while CoAd and liver metastases represented the preliminary stage or risk factor for CRC or new cancer (at a distant site) development, metastatic tumors formed the late stage of CRC progression. Our results demonstrated that overall, the expression of piR-37524 was significantly higher in CoAd, CRC, and liver metastases than in adjacent non-tumor tissue (N = 64) (p < 0.001) (Fig. 3). Notably, a significantly higher expression of the piRNA was observed within the individuals with adenomatous polyps relative to the non-metastatic CRC (median: −11.12 vs.−16.61, p < 0.001) and metastatic CRC (median: −11.12 vs. −15.13, p = 0.0006) (Fig. 3). Liver metastases also showed higher piR-37524 expression than the M0 (median: −11.43 vs. −16.61, p = 0.0003) and the M1 (median: −11.38 vs. −15.13, p = 0.0024) CRC groups (Fig. 3). However, there was no significant difference in piR-37524 expression between CoAd and liver metastases (median: −11.12 vs. −11.38, p = 0.9783).

Given the high piR-37524 expression in early CRC development and distant metastasis, we evaluated its potential as a screening biomarker. We measured serum piR-37524 expression in 64 patients with CoAd (N = 26) or CRC (N = 38) and in 36 healthy controls. Expression was normalized to miR-92a-3p (−ΔCt).

Serum piR-37524 expression was significantly higher in patients with CoAd (median: −11.44 vs. −13.44, p < 0.001) and CRC (median: −12.22 vs. −13.44, p = 0.0078) compared to healthy controls (Fig. 4A). There was no significant difference between CoAd and CRC serum levels. High serum piR-37524 expression correlated with poor overall survival in CRC patients (p = 0.0325) (Fig. 4B).

We subsequently plotted receiver operating characteristic (ROC) curves to have a better understanding of the diagnostic accuracy of piR-37524 within serum samples. As shown in Fig. 4C, piR-37524 significantly differentiates between the CoAd patients and normals with an AUC of 0.8365 (95% CI: 0.7215–0.9515, Sensitivity: 86.9%, Specificity: 76%, p < 0.0001). However, for the CRC vs. normal group, we obtained a lower AUC of 0.6700 (95% CI: 0.5304–0.8096, Sensitivity: 62.5%, Specificity: 56%, p = 0.0287) (Fig. 4D). Although by combining the CoAd and the CRC patients in a single group, we managed to enhance the AUC slightly to 0.7396 (95% CI: 0.6270–0.8522, Sensitivity: 72.73%, Specificity: 72%, p = 0.0006), thus improving the detection accuracy of CRC patients and individuals at risk of CRC (CoAd) from normal (Fig. 4E).

Given the upregulation of piR-37524 in CRC tissues and its correlation with tumor progression, highlighting its potential as a diagnostic biomarker, we investigated whether piR-37524 functions as an oncogene in colorectal cancer progression. We first evaluated piR-37524 expression across a panel of colorectal cancer cell lines and found that HCT116 and HT29 cells displayed markedly higher piR-37524 levels than the other lines (Fig. 5A). We then used an MTT assay to determine the proliferation rate of colorectal cancer cells transfected with piR-37524 inhibitors or NC. Our results showed that knockdown of piR-37524 significantly reduced the proliferation of HCT116 and HT29 cells (Fig. 5B,C). To further validate the effect of piR-37524 on cell proliferation, we performed a colony formation assay. In this assay, HCT116 and HT29 cells transfected with piR-37524 inhibitors were allowed to grow for a prolonged period, and the number of colonies formed was counted. The results revealed that knockdown of piR-37524 significantly reduced the colony-forming ability of both HCT116 and HT29 cells (Fig. 5D,E), further confirming its inhibitory effect on cell proliferation. Additionally, we performed wound healing assays to assess the effect of piR-37524 knockdown on cell migration. The results indicated that knockdown of piR-37524 markedly inhibited the migration of HCT116 and HT29 cells compared to the control treatment (Fig. 5F,G). WB analysis revealed a notable reduction in Snail protein expression within the HCT116 cell line following piR-37524 inhibition, and elevated protein levels of E-cadherin and β-catenin were observed in both HCT116 and HT29 cell lines. These alterations are indicative of changes in the epithelial-mesenchymal transition (EMT) process (Fig. 5H). Collectively, these findings demonstrate that inhibition of piR-37524 significantly impairs migration-related phenotypes of colorectal cancer cells in vitro. Effects on invasive capacity were not assessed in the present study and will be addressed in future work.

Accumulating evidence indicates that piRNAs can interact with mRNAs via partial sequence complementarity, thereby repressing or enhancing gene expression. To explore putative targets of piR-37524, we conducted RNA sequencing (GEO ID. GSE310036) on HCT116 and HT29 cells transfected with piR-37524 inhibitors or the corresponding negative control. In HCT116 cells, 83 differentially expressed genes (DEGs) were identified (p < 0.05, |log₂ fold change| > 1), with 27 upregulated and 56 downregulated upon piR-37524 inhibition (Fig. 6A). In HT29 cells, 71 DEGs were identified, with 63 upregulated and 8 downregulated (Fig. 6B). KEGG enrichment analysis indicated that DEGs in both cell lines were associated with TNF and TGF-beta signaling pathways (Fig. 6C,D).

To further explore these potential target genes, we performed an intersection analysis of the upregulated DEGs in both HCT116 and HT29 cells and identified six genes consistently upregulated in both cell lines (Fig. 6E), of which ALOXE3, HMOX1, and TNFAIP3 were validated as significantly differentially expressed by RT-qPCR (Fig. 6F,G). Using miRanda software, we further analyzed and identified TNFAIP3 as a potential target gene of piR-37524. Our analysis suggested that piR-37524 might interact with the 3^′UTRs of TNFAIP3 mRNA, which was upregulated by piR-37524 inhibitors (Fig. 6H). Thus, TNFAIP3 was identified as a candidate gene with a predicted 3^′UTR binding site for piR-37524, supported by consistent upregulation at transcriptional and translational levels upon piR-37524 inhibition.

To further investigate the mechanism underlying migration inhibition, we evaluated the NF-κB signaling pathway in HCT116 and HT29 cells. Given that our previous RNA sequencing data suggested TNFAIP3 as a potential target of piR-37524, and TNFAIP3 is a known regulator of NF-κB signaling. First, TNFAIP3 expression was markedly increased in both HCT116 and HT29 cells following piR-37524 inhibition; then, we also examined the phosphorylation status of IKK α/β, a key component of the NF-κB pathway. Upon piR-37524 knockdown, we observed a decrease in the expression of phosphorylated IKK α/β (pIKK α/β), while the expression of total IKK α remained relatively unchanged (Fig. 6I). These findings suggest that piR-37524 may promote CRC cell migration and EMT by modulating the NF-κB signaling pathway, likely through a regulatory association with TNFAIP3.