Cervical cancer is the fourth most common malignant tumor in women worldwide in terms of both incidence and fatality (Sung et al., 2021). Globally, there were about 604,127 new cervical cancer diagnoses in 2020, and 341,831 fatalities (Sung et al., 2021), which seriously threatened women’s health. Currently, cervical cancer can be treated through diverse approachcng surgery (Greggi et al., 2020), chemotherapy (Crafton and Salani, 2016), radiotherapy (Tan et al., 2019), and immunotherapy (Ferrall et al., 2021; Feng et al., 2020). However, the 5-year survival rate was only 59.8% (Zeng et al., 2018), signifying a substantial challenge in boosting the survival rates of individuals diagnosed with cervical cancer.

Cervical cancer development is closely associated with ongoing infection with high-risk human papillomaviruses (HPV). Studies have shown that the most predominant oncogenic HPV types are 16 (57%) and 18 (16%) (Crosbie et al., 2013). The E6 and E7 proteins of these high-risk HPV types possess oncogenic characteristics. However, not all patients infected with HPV will eventually develop cervical cancer, as the progression of the disease is regulated by numerous factors. The specific mechanisms of these factors are still not fully understood (Okunade, 2020). Therefore, further research on the pathogenesis of cervical cancer, together with the discovery of novel treatment targets and sensitive and specific diagnostic markers, has significant clinical value for improving the efficiency of cervical cancer diagnosis and treatment.

Long noncoding RNAs (lncRNAs) are transcripts that are more than 200 nucleotides long (Chen et al., 2022). They have the ability to control how genes that code for proteins are expressed by exerting an influence on the transcription, post-transcriptional and translation modifications of mRNAs (Quinn and Chang, 2016; Choi et al., 2019), thereby leading to tumorigenesis and progression (Bai et al., 2023; Wu et al., 2022). Depending on their position in relation to protein-coding genes, lncRNA can be classified as sense, antisense, intron, intergenic, and bidirectional (St Laurent et al., 2015). Antisense lncRNAs constitute a relatively well-studied category. More than 30% of human annotated transcripts have corresponding antisense lncRNA, which regulates the corresponding justice mRNA through various mechanisms (Faghihi and Wahlestedt, 2009).

In order to examine the impact of differential lncRNA on the tissues of cervical cancer, this study selected cancer tissues and adjacent non-cancerous tissues from three pairs of HPV16-positive cervical squamous carcinoma patients, all clinically diagnosed at stage IB2. Arraystar LncRNA and mRNA expression profiling microarrays were utilized via high-throughput sequencing technology. Bioinformatics analyses were carried out to identify lncRNAs and mRNAs, and to predict relevant downstream regulatory signaling pathways and key genes. For mRNAs, GO functional analysis, KEGG pathway enrichment analysis, and PPI assessments were used. At the cellular level, the characteristics of these lncRNAs, the signaling pathways they participate in, and their relationship to HPV16 infection were all examined using RT-qPCR validation. This study provides theoretical support for bioinformatics in outcome prediction and contributes to the diagnosis of cervical cancer.

With the emergence of high-throughput technologies, a substantial number of dysregulated lncRNAs have been discovered in cervical cancer (Bai et al., 2023; Liu et al., 2023). These lncRNAs have been increasingly regarded as signature molecules of cervical cancer, and have drawn extensive attention. Notable examples include H19 (Dai et al., 2023; Lee et al., 2003), HOTAIR (Sun et al., 2021), and MALAT1 (Hao et al., 2020). These lncRNAs show high expression levels in cervical cancer tissues and are associated with the proliferation and migration of cervical cancer cells. Chronic infection with high-risk HPV subtypes is closely related to the development of cervical cancer. E6 and E7 are identified as key “HPV carcinogens” that drive the progression of cervical cancer (Fan and Shen, 2018). The dysregulated lncRNAs and their associated regulatory pathways in cervical cancer may serve as potential molecular targets for its diagnosis and treatment.

In this study, we analyzed the differential expression profiles of lncRNA and mRNA in cervical cancer patients via high-throughput RNA sequencing The results showed that there were 7,991 differentially expressed lncRNAs in cervical cancer tissues, with 3,608 being upregulated and 4,383 downregulated (Fold change >2 and P < 0.05). Additionally, 5,886 mRNAs were also significantly differentially expressed, including 3,666 that were upregulated and 2,220 that were downregulated. Interaction analysis of the proteins encoded by these significantly different mRNAs demonstrated a strong interaction among the proteins CTNNB1, ATXN2, PRKCD, FCGR3B, UQCRH, RPS28, and KIF15.

The BP, CC, and MF were analyzed by GO enrichment analysis as follows: Upregulated mRNAs are primarily involved in cellular molecular metabolic processes, the cell cycle, organelle organization, growth and development, and other biological processes. These mRNAs are mostly associated with cytoplasmic and organelle components and display molecular functions such as nucleic acid binding, protein binding, and catalytic activity. Conversely, the downregulated mRNAs participate in multicellular biological processes, may have bioadhesion functions, and are connected to myofibrils, the extracellular matrix, and plasma membrane components. Their molecular functions include protein binding and metal ion binding, among others. KEGG pathway enrichment analysis showed that the upregulated mRNAs may be involved in the regulation of the cell cycle, splicing body, and cellular senescence, and are associated with Alzheimer’s disease, the Fanconi anemia pathway, and Parkinson’s disease. The downregulated mRNAs are primarily involved in the regulation of the cell cycle, spliceosome activity, and cellular aging. Additionally, these downregulated mRNAs are associated with insulin and bile secretion and are connected to signaling pathways such as cAMP and MAPK, which may regulate neuroactive ligand-receptor interactions and other pathways. Moreover, analysis of the 20 antisense lncRNAs with the most significant changes revealed that these lncRNAs might exert their biological functions by influencing the corresponding mRNAs, indicating their potential as marker molecules for studying the progression of cervical cancer.

To validate the results of the microarray analysis, we utilized a combination of predictions from the GEPIA database and RT-qPCR assays. We identified the upregulated lncRNAs CDKN2B-AS1, MIR205HG, and HAGLROS, which showed significant differences from the microarray results, as well as the downregulated lncRNAs GATA6-AS1 and DICER1-AS1. The results of GEPIA database showed that CDKN2B-AS1, MIR205HG, and HAGLROS were highly expressed in cervical cancer, and patients with high expression levels had poorer prognoses. Conversely, GATA6-AS1 and DICER1-AS1 were found to be lowly expressed in cervical cancer, and patients with low expression levels had poorer prognoses. We measured the relative expression of these five lncRNAs in HcerEpic and cervical cancer cell lines using RT-qPCR. The results were consistent with the predictions from the GEPIA database, and the differences were statistically significant (P < 0.05). These findings show that the observed trends are in line with the microarray results, thereby confirming the reliability of the lncRNA microarray data.

HPV infection is a major risk factor for cervical cancer. In a study, E6 and E7 oncoproteins were found to be strongly associated with cell adhesion in HPV-induced immunosuppression (Dai et al., 2022). In addition to cervical cancer, HPV16 E6/E7 has been associated with head and neck squamous cell carcinoma (HNSCC), e.g., HPV-positive and HPV-negative HNSCC require different B-cell-centered targeting strategies to induce antigen-specific intratumor maturation of TLS and driver B cells (Ruffin et al., 2023). HPV is also capable of preventing a strong immune response in infected cells by reducing chemokine expression, among other mechanisms (Attademo et al., 2020). In this study, to investigate the relationship between differentially expressed lncRNAs and high-risk HPV infection, we constructed a cell line stably overexpressing HPV16 E6/E7 by transfecting HcerEpic cells with the recombinant lentiviral vector CV224-HPV16 E6/E7. The results of the RT-qPCR assay showed that in HcerEpic cells overexpressing HPV16 E6/E7, the expression levels of CDKN2B-AS1 and HAGLROS were significantly increased, while the expression of GATA6-AS1 was decreased. These findings imply that CDKN2B-AS1, HAGLROS, and GATA6-AS1 may be regulated by HPV16 E6/E7 in human normal cervical epithelial cells.

Although the biological functions of microarray-screened lncRNAs have been studied in cervical cancer, such as the knockdown of MIR205HG, which significantly reduced the proliferation, migration, and invasive ability of cervical cancer cells (Yin et al., 2022), and the oncogenic role of LINC000511 in regulating the expression of PGK1 (Xin et al., 2023), the vast majority of differentially expressed lncRNAs are still poorly understood in terms of their roles in cervical cancer. In this study, we screened a large number of dysregulated lncRNAs in cervical cancer using high-throughput sequencing and confirmed our findings through a combination of GEPIA and RT-qPCR. However, this study did not explore how HPV16 E6/E7 regulates these lncRNAs and their regulatory mechanisms on cervical cancer cells. We plan to continue to supplement the regulatory experiments of HPV16 E6/E7 on lncRNAs and the functional experiments of lncRNAs to provide evidence for their impact on the molecular mechanisms of cervical cancer.

The rapid development of machine learning (ML) methods in recent years has revolutionized the research on the association of long chain non-coding RNAs (lncRNAs) with human diseases. Compared with the time-consuming and labor-intensive characteristics of traditional experimental methods, machine learning-based methods achieve high-throughput prediction and prioritization of disease-associated lncRNAs by integrating multi-omics data. For example, Guo et al. (2022) proposed a machine learning-based model called LDACE to predict potential lncRNA-disease associations by combining extreme learning machine (ELM) and convolutional neural network (CNN). Similarly, Chen et al. (2018b) proposed a novel framework called IMCMDA for inferring potential miRNA-disease associations. The powerful advantage of machine learning techniques in revealing lncRNA and miRNA disease regulatory mechanisms. Yi et al. (2020) developed a stacked integrated computational framework, RPI-SE, for efficient prediction of ncRNA-protein interactions. Meanwhile, k-mer sparse matrices were used to extract effective features of ncRNA sequences. In addition, Bao et al. (2019) who provided LncRNADisease 2.0 to organize the study of lncRNA-disease association, which provided new ideas for potential clinical applications related to lncRNA. Whereas this study relies on traditional methods of bioinformatics analysis and lacks machine learning methods to advance the prediction of lncRNA-disease associations, we will incorporate computational methods of machine learning to enhance the exploration of cutting-edge research in the next study.

In conclusion, this study identified the lncRNAs related to cervical cancer, along with their associated biological processes, molecular functions, and molecular biological pathways. This work lays a molecular foundation for predicting biomarkers for cervical cancer and provides bioinformatic data to support its treatment.