Circular RNA Hsa_circ_0062403 Acts as a Potential Prognostic Biomarker for Early Recurrence of Hepatocellular Carcinoma

Introduction

Primary liver cancer is the sixth most frequently diagnosed cancer and the third leading cause of cancer death worldwide, according to global cancer statistics 2022.¹ Primary liver cancer mainly includes 75%-85% hepatocellular carcinoma (HCC) and 10%-15% intrahepatic cholangiocarcinoma (ICC). And 21%-55% of liver cancer cases are mainly related to chronic hepatitis B and C virus infection.^2,3 In contrast to the declining proportion of virus-related HCC, the incidence of non-viral HCC, particularly that driven by metabolic dysfunction-associated steatotic liver disease (MASLD), is increasing rapidly. MASLD affects 38% of the global population, and it is estimated that the global prevalence of MASLD will reach 56% within the next decade.⁴ However, HCC is frequently diagnosed at an advanced stage, thus precluding surgical resection and conferring a poor prognosis.⁵ Recurrence and metastasis remain the leading causes of poor prognosis and long-term mortality in patients with resectable HCC. Despite remarkable advances in the treatment of hepatocellular carcinoma (HCC) in recent years—including interventional therapy, local ablation, chemotherapy, targeted therapy, and immunotherapy—the prognosis of patients with HCC remains dismal, with a 5-year overall survival rate of less than 18%.⁶

Circular RNAs (circRNAs) are a unique class of non-coding RNAs (ncRNAs) defined by their covalently closed loop structure, a hallmark feature that distinguishes them from linear ncRNA species. CircRNAs exert pivotal regulatory roles in transcriptional and post-transcriptional gene expression via diverse mechanisms, including acting as miRNA sponges, encoding functional polypeptides, serving as protein scaffolds, and forming stable RNA-protein complexes to modulate downstream biological processes.⁷ It is well documented that dysregulated circRNAs are critically involved in the tumorigenesis and malignant progression of diverse human cancers, including HCC,^8–10 lung cancer,^11,12 gastric cancer,¹³ breast cancer,^14,15 colorectal cancer,¹⁶ hematological malignancies,¹⁷ and neuroblastoma,¹⁸ and are indispensable modulators of the establishment and maintenance of tumor drug-resistant phenotype.¹⁹ Given that dysregulated circRNAs are stably expressed and readily detectable in tissues, peripheral blood, and body fluids, circRNAs hold great promise to serve as novel biomarkers and therapeutic intervention targets for the diagnosis and treatment of malignant tumors.²⁰ Tang et al²¹ indicated that circRNA_0009792 enhances the stemness of HCC cells via the PSIP1/c-Myc signaling pathway, and ultimately drives HCC progression and the development of lenvatinib resistance. In another study, researchers found that higher circMRCKα levels presented significantly reduced overall survival and greater cumulative recurrence. circMRCKα promotes tumor glycolysis and progression via USP22 binding and HIF-1α upregulation.²² These findings highlight that circRNAs, acting as new biomarkers, exert critical regulatory roles in the recurrence and metastasis of HCC; however, further research is still needed.

In this study, we identified a novel circRNA (hsa_circ_0062403) in HCC, which was named circLZTR1, which is encoded by the Leucine zipper–like post-translational regulator 1 (LZTR1) gene. In our present study, we revealed that the knockdown of LZTR1 can promote HCC cell proliferation and metastasis, and might be used as an effective predictive marker for HCC recurrence. Mechanistically, LZTR1 downregulation increased RAS protein expression, leading to MAPK pathway activation and enhanced epithelial-mesenchymal transition. Although a substantial number of circRNAs have been documented to participate in the tumorigenesis and malignant progression of HCC, few studies have specifically focused on early postoperative recurrence. Furthermore, the biological function and clinical value of circLZTR1 in HCC early recurrence remain largely unreported. Therefore, in this study, we aims to investigate the predictive value of circLZTR1 for early recurrence of HCC, and to clarify its intrinsic correlation with the malignant biological phenotype of HCC.

Materials and Methods

Patients and Tissues

91 HCC patients who underwent surgical resection were included in this study. The inclusion criteria are as follows: (1) All patients were pathologically diagnosed with HCC; (2) Not receiving radiotherapy, immunotherapy, or chemotherapy before surgery. The exclusion criteria were as follows: (1) history of other cancers; and (2) incomplete clinical or follow-up data. Early recurrence of HCC was uniformly defined as tumor recurrence and metastasis occurring within 1 year after radical hepatectomy for HCC. This study process was approved by the Ethics Committee of the Ningbo Medical Center Lihuili Hospital (Ethical approval number KY2022PJ020). All tissue samples were collected during liver cancer surgical resection, stored at low temperature (−80°C) with RNA Preservation Kit (RNAhold, TransGen Biotech, Beijing, China).

RNA Extraction, Reverse Transcription, and Quantitative Real-Time Polymerase Chain Reaction (qRT-PCR)

Total RNA was extracted from HCC tissues using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). The concentration and purity of total RNA were verified using the NANO Quant infinite M200PRO spectrophotometer (TECAN, Männedorf, Switzerland). Complementary DNA (cDNA) was synthesized via reverse transcription using the AMV Reverse Transcription System (Invitrogen, Carlsbad, CA, USA) with 2 μg of qualified total RNA as input template.

qRT-PCR was performed with 2×PCR Mix (QIAGEN, Hilden, Germany) on the ABI ViiA 7 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) following the manufacturer’s instructions. For circular RNA expression analysis, back-spliced junction-specific primers for circLZTR1 were designed. The primer sequences for hsa_circ_0062403 were 5’-TACATGCCAAGGGCCCTAAA-3’ (forward) and 5’-CGGCTGGAGAAACCACATTT-3’ (reverse). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as the endogenous normalizer, with primer sequences of 5’-AGCCACATCGCTCAGACAC-3’ (forward) and 5’-GCCCAATACGACCAAATCC-3’ (reverse). All primers were synthesized by Shanghai Jierui Bioengineering Co., Ltd. (Shanghai, China). For the low abundance of circRNA, a 12-cycle pre-amplification was performed prior to formal qRT-PCR amplification. qRT-PCR was carried out in triplicate for each sample, and the specificity of amplification was confirmed via melt curve analysis. The relative expression level of circLZTR1 was calculated using the 2^−ΔΔCt method.²³

Statistical Analysis

Statistical analyses were performed using SPSS 25.0 (IBM Corporation, 2020, USA). Numerical data were expressed as means and standard deviations (SDs) for normally distributed data, while categorical data were expressed as numbers (%). We used the chi-square test and t test to compare the clinicopathological characteristics and circLZTR1 levels between the early recurrence group and the non-early recurrence group (Control). Patients were divided into different risk groups (Low-; Moderate-; High-) with the cut-off points of circLZTR1 automatically calculated by X-tile software using an exhaustive method and Log rank test (Yale University, CT, USA). Subsequently, Survival curves were calculated to compare the recurrence-free survival (RFS) rates in three different risk groups using the Kaplan-Meier method and the Log rank test. A multivariate Cox proportional hazards regression model was performed to identify predictors of early recurrence. Variables showing significant differences in univariate t or χ²-tests (P-value < 0.05) were included in the multivariate Cox regression model for identification of independent prognostic factors. Receiver operating characteristic curve (ROC) analysis was performed to assess the predictive performance of circLZTR1 for early recurrence. Statistical significance was defined as P-value < 0.05.

Results

Baseline Characteristics of Enrolled HCC Patients

The baseline characteristics are shown in Table 1. In this cohort, 75/91 (82%) patients were male, and 16/91 (18%) patients were female. 27/91 (29.7%) patients were diagnosed with early recurrence within 1 year after radical resection. The average age of the early recurrence group was 59.5 years (59.5±13.8), 61.4 years (61.4±12.3) in the control. Regarding tumor profiles, 35/91 (38.5%) were MVI positive, 7/91 (7.7%) were PVTT positive, 18 (19.8%) had tumors in the AJCC T3 and T4 stages, and 35/91 (38.5%) had poor pathological differentiation. The chi-square and t test found that hepatitis B virus (HBV) infection, alpha-fetoprotein (AFP) grade, direct bilirubin (DB), international normalized ratio (INR), microvascular invasion (MVI), tumor diameter, number of tumors, and portal vein tumor thrombosis (PVTT) were significant difference between early recurrence and control group (P < 0.05). In addition, the lower circLZTR1 expression level was also observed in MVI and PVTT positive patients, as shown in Figure 1 (P < 0.05). Post-hoc power analysis showed a power of 0.82 for independent t-test and 0.998 for Log rank test by circLZTR1 expression level.

Table 1 Baseline Characteristics of the HCC Patients

Figure 1 The association between circLZTR1 and HCC malignant phenotypes. (A) Association with MVI. *P < 0.05 (B) Association with PVTT. *P < 0.05 (C) Association with tumor number. (D) Association with AFP levels.

Risk Stratification and Prognostic Assessment for circLZTR1

The optimal cut-off points of circLZTR1 expression level were auto-calculated by the X-tile software and divided HCC patients into the high risk group (< 0.10), moderate risk group (0.10–0.25), and low risk group (> 0.25). The chi-square test indicated that there was a significant difference between the early recurrence and control group (χ² = 7.60; P = 0.022). Survival curves were calculated to compare the RFS rates in three different risk groups using the Kaplan-Meier method, and the results showed a significant discriminatory ability for recurrence risks (P < 0.05, Figure 2).

Figure 2 Recurrence-free survival curve of circLZTR1 in three different risk groups.

Prognostic Factors for Early Recurrence in Multivariate Cox Regression Analysis

Based on the results of univariable analysis in Table 1, variables with statistical significance (P < 0.05) were entered into the multivariable Cox proportional hazards regression model, including HBV infection, AFP grade, DB, INR, MVI, tumor diameter, number of tumors, PVTT and circLZTR1. Multivariable Cox regression analysis showed that AFP grade (>400 vs <20: HR = 7.880, 95% CI: 2.499–24.844, P < 0.001), INR (>1 vs ≤1: HR = 4.203, 95% CI: 1.133–15.588, P = 0.032), number of tumors (HR = 3.795, 95% CI: 1.541–9.346, P = 0.004), PVTT (HR = 3.438, 95% CI: 1.050–11.258, P = 0.041), tumor diameter (HR = 1.306, 95% CI: 1.101–1.550, P = 0.002), and circLZTR1 (>0.25 vs <0.10: HR = 0.142, 95% CI: 0.042–0.482, P = 0.002) were independent influence factors for HCC early recurrence (Table 2). The results indicated that a lower circLZTR level was an independent risk factor for HCC early recurrence.

Table 2 Multivariate COX Analyse for Early Recurrence of HCC Patients

Predictive Performance of circLZTR1 for Early Recurrence by ROC Analysis

We established three prognostic models for predicting HCC early recurrence based on multivariable Cox regression analysis results: ①Clinical Model, a clinicopathological feature-based model incorporating AFP grade, INR, tumor number, PVTT, and tumor diameter; (2) CircLZTR1 Model, a univariable model with circLZTR1 expression as the only predictor; (3) Cox Model, the full model integrating all the above clinicopathological variables and circLZTR1. ROC curve analysis was conducted to assess the predictive performance of each model and AJCC T stage, with the area under the curve (AUC) calculated to quantify the predictive accuracy. The results revealed that the circLZTR1 alone model exhibited a favorable predictive value for early HCC recurrence (AUC = 0.661, 95% CI: 0.534–0.787, P = 0.016), with an AUC comparable to that of the Clinical Model (AUC = 0.655, 95% CI: 0.534–0.776, P = 0.020). More importantly, the combined Cox Model that incorporated circLZTR1 into the clinicopathological features (AUC = 0.767, 95% CI: 0.666–0.868, P < 0.001) yielded a significantly higher AUC than the Clinical Model and AJCC T stage (AUC = 0.707, 95% CI: 0.579–0.835, P = 0.002), indicating a remarkably enhanced predictive efficacy for early recurrence in HCC patients. The results were shown in Table 3 and Figure 3.

Table 3 Comparison of ROC Curve with and Without circLZTR1

Figure 3 Receiver operating characteristic curve of circLZTR1 in prediction for HCC early recurrence.

Discussion

CircRNAs are characterized by a single-stranded, covalently closed circular structure, which lacks 5’ caps and 3’ poly(A) tails as a class of non-coding RNAs. This unique feature renders them resistant to ribonuclease degradation, with a significantly longer half-life than linear mRNAs. CircRNAs can regulate the transcriptional and post-transcriptional expression of their host genes, modulate pre-mRNA alternative splicing, serve as competing endogenous RNAs (ceRNAs) to sponge microRNAs (miRNAs), bind and sequester RNA-binding proteins (RBPs), and act as scaffolds to orchestrate protein-protein interactions.²⁴ Beyond these well-established roles, a growing body of evidence has revealed that a subset of circRNAs can be translated into functional peptides or proteins via cap-independent translation pathways.²⁵ In this study, we found that lower circLZTR1 expression level was closely associated with the MVI and PVTT, and may serve as a novel, promising prognostic biomarker for predicting early recurrence in patients with HCC. In our present study, the knockdown of LZTR1 can promote HCC cell proliferation and metastasis, and was associated with early recurrence of HCC. Mechanistically, LZTR1 downregulation increased RAS protein expression, leading to MAPK pathway activation and enhanced epithelial-mesenchymal transition.²⁶ Consequently, we put forward the hypothesis that circLZTR1 may exert its recurrence-promoting and pro-metastatic effects in HCC predominantly through regulating the expression of its cognate host gene LZTR1.

Kurahashi et al²⁷ first reported that LZTR1 was encoded by a gene mapping to chromosome 22q11.2, and identified via sequence homology with leucine zipper-containing proteins. Subsequent studies identified LZTR1 as a member of the BTB–Kelch superfamily. Functionally, LZTR1 acts as the substrate-specific adaptor component of the Cullin 3 (CUL3) ubiquitin ligase complex, which drives the ubiquitination and subsequent proteasomal degradation of canonical RAS proteins as well as the non-canonical RAS-family member RIT1, and consequently restrains the activation of the RAS/MAPK signaling pathway.^28–30 However, the RAS/MAPK signaling pathway plays a fundamental role in the malignant progression and short survival of human HCC.³¹ Yıldız et al³² identified that the loss of the LZTR1 gene can reduce the vimentin expression and motility, and then exhibited reduced transwell migration and invasion in LZTR1 knockout (KO) Hep3B HCC Cells.

In this study, we first found that circLZTR1 (hsa_circ_0062403), a circular RNA derived from the back-splicing of its host gene LZTR1, acts as a promising prognostic biomarker for predicting early recurrence of postoperative HCC patients. Our previous study²⁶ demonstrated that LZTR1 downregulation markedly activates the oncogenic RAS/MAPK cascade and induces epithelial-mesenchymal transition (EMT), both of which are core events driving HCC malignant progression, metastasis, and early recurrence, as Zhao et al indicated.³³ These findings support the clinical value of circLZTR1 and imply that it may participate in HCC progression via regulating LZTR1 and its downstream signaling. Notably, this potential mechanistic linkage is merely a scientific hypothesis deduced from previously published evidence, rather than a conclusion verified by functional experiments in the present study. In addition, we also observed a significant downregulation of circLZTR1, which was associated with MVI or PVTT, and may serve as a critical mediator of HCC malignant phenotypes. As Yıldız et al³² reported, the loss of LZTR1 protein could remodel EMT and mesenchymal plasticity (EMP). Notably, profound intratumoral heterogeneity and dynamic phenotypic plasticity are core drivers of tumor progression and metastatic dissemination in HCC.³⁴ Mounting evidence confirmed that EMT enhanced the invasiveness of circulating tumor cells (CTCs) and played a pivotal role in promoting HCC metastasis and MVI progression.^35–37 Therefore, we hypothesize that circLZTR1 may promote the malignant progression and metastatic dissemination of HCC by regulating the EMT and EMP process. However, the specific functional mechanism of circLZTR1 in HCC, including whether it exerts biological effects by directly regulating its host gene, acting as a miRNA sponge, or interacting with functional proteins, has not yet been elucidated. Further in-depth studies are still needed to clarify its underlying regulatory network.

Accumulating evidence indicates that circRNAs can serve as important biomarkers for predicting poor prognosis in HCC patients. Jang et al³⁸ discovered that high hsa_circ_0003570 expression was an independent prognostic factor for progression-free survival (PFS, HR=0.633). In another study, high hsa_circ_0004018 and hsa_circ_0003570 were inversely independent risk factors for PFS in patients with HBV-HCC.³⁹ A study on cholangiocarcinoma (CCA) complicated with biliary obstruction reported that a CCA-specific circRNA panel consisting of hsa_circ_0000367, hsa_circ_0021647, and hsa_circ_0000288, which can be detected in both bile and serum exosomes, has been identified as a promising biomarker for postoperative early recurrence monitoring of CCA.⁴⁰ Our study first found that circLZTR1 can act as an important biomarker for early recurrence (1-year) of HCC (ACU = 0.767), this result providing a novel approach for preoperative risk stratification and postoperative surveillance of HCC.

Certainly, there are some limitations of this study as follows: First, this study was a single-center retrospective study with a relatively small sample size of 91 HCC patients, which may lead to inherent selection bias and limited statistical power, and needs to be further validated in larger multi-center cohorts. Second, we only proposed the potential regulatory mechanism of circLZTR1 in HCC based on our previous functional research on its host gene LZTR1, while systematic in vitro and in vivo functional experiments and mechanistic validation at cellular and animal levels have not been performed. Third, circLZTR1 expression was not validated in non-invasive body fluid samples, including serum, plasma, and urine, which limits the exploration of its clinical translational potential as a liquid biopsy biomarker.

In conclusion, the downregulation of circLZTR1 is closely associated with the acquisition of invasive phenotypes, vascular invasion, and subsequent early recurrence of HCC patients after surgery, suggesting that circLZTR1 may serve as a novel, potential prognostic biomarker for predicting early recurrence in patients with HCC. Given the limitations of the present study, large-scale cohorts, external validation and mechanistic research are warranted to confirm our observations.

Conclusion

In conclusion, the downregulation of circLZTR1 is closely associated with the acquisition of invasive phenotypes, vascular invasion, and subsequent early recurrence of HCC patients after surgery, suggesting that circLZTR1 may serve as a novel, promising prognostic biomarker for predicting early recurrence in patients with HCC.