中国P站

TSGs: Tumor Suppressor Genes; GEO: Gene Expression Omnibus; PCA: Principal Component Analysis; PPI: Protein-Protein Interaction; DMEM: Dulbecco's Modified Eagle's Medium; PG: Prodigiosin; FBS: Fetal Bovine Serum; PBS: Phosphate-Buffered Saline; PI: Propidium Iodide; ANOVA: One-Way Analysis of Variance; GO: Gene Ontology; KEGG: Kyoto Encyclopedia of Genes and Genomes; MCC: Maximal Clique Centrality; DEGs: Differentially Expressed Genes; Cis: Cisplatin; IHC: Immunohistochemistry

Cervical cancer, representing the second most common and prevalent malignancy among women globally, inflicts significant economic and medical burdens on families [1]. The etiology of cervical cancer is multifaceted, encompassing genetic and environmental factors [2]. Cervical Cancer (CC) is the fourth recorded widespread cancer among women globally [3]. Currently, various chemical drugs such as bevacizumab, topotecan, and cisplatin constitute the first-line treatment for cervical cancer. However, their use is often marred by severe side effects and the emergence of drug resistance [4]. Consequently, the development of new therapeutic methods is of utmost importance. We have identified a natural compound, Prodigiosin (PG) [5], which exhibits a targeted apoptosis-inducing effect on cancer cells. Thus, the exploration of PG's anticancer mechanism and potential clinical applications holds promise.

Building on the above, PG is a dark red bioactive secondary metabolite synthesized by Actinomycetes, Serratia marcescens, and other bacteria [6]. It exhibits a range of biological properties, including antibacterial, antiprotozoal, anti-malarial, immunosuppressive, and anticancer activities. Extensive research affirms that PG triggers apoptosis in various human cancer cells, while demonstrating comparatively lower toxicity towards normal cells [7]. Recent findings in the context of breast cancer reveal that PG can inhibit the phosphorylation of LRP6, DVL2, and GSK3β, thereby blocking Wnt/β-catenin signal transduction and diminishing the expression of cyclin D1, consequently slowing tumor progression [8]. Furthermore, PG exhibits cytotoxic effects on multidrug-resistant human cancer cells. Studies have reported that PG can induce autophagic death in Dox-S and DoxR lung cancer cells by inhibiting the Akt/PI3K-p85/mTOR signaling pathway, suggesting its potential as a therapeutic option for lung cancer [9]. Nonetheless, comprehensive research into the therapeutic role of PG in cervical cancer is currently lacking.

Recognizing the importance of molecular-level understanding in cancer therapeutics, high-throughput platform-based microarrays have emerged as effective tools for investigating the roles of various genes in disease mechanisms. These gene chips have found extensive application in diverse biological and medical research fields, particularly for identifying Differentially Expressed Genes (DEGs). With the aid of bioinformatics analysis, it is possible to pinpoint oncogenes and Tumor Suppressor Genes (TSGs) exhibiting abnormal methylation patterns and differential expression in cervical cancer tissues. Furthermore, this analysis facilitates the elucidation of associated pathways and functions. Such knowledge contributes to the development of biological markers and therapeutic targets, thereby enabling more precise diagnosis and treatment strategies for cervical cancer.

In our present study, we leverage these technological advancements to further understand the role of PG in cervical cancer. Our initial step involves the analysis of cervical cancer target genes using the Gene Expression Omnibus (GEO) database. Subsequently, we conduct experimental verification of these genes. This dual-approach not only aids in the exploration of the mechanism behind Prodigiosin (PG)- induced apoptosis in cervical cancer cells but also provides clarity on its potential therapeutic targets.

Patient tissue samples and data information

This research was approved by the Ethics Committee of Xiang yang Central Hospital, Hubei, China (Ethics Number: 2023-022). 10 CC tissues and 10 normal cervical tissues were collected and informed consent was obtained from all subjects and/or their legal guardians. We confirm that all methods were performed in accordance with the relevant guidelines and regulations.

All selected cases were histologically verified by pathologists, with none of the patients having undergone chemotherapy or radiation therapy prior to surgery. The normal tissue samples were sourced from women who had hysterectomies due to non-malignant conditions. Subsequently, all the collected tissues were preserved in paraffin blocks.

Data information

We sourced cervical carcinoma-related datasets GSE127265, GSE9750, and GSE173097 from the NCBI GEO. These datasets were based on three different platforms. The first, GSE23126, operated on the GPL23126 platform (Clariom_D_Human) Affymetrix Human Clariom D Assay. This dataset facilitated the identification of differentially expressed genes in cervical carcinoma patients through comparative transcriptome analysis, utilizing seven samples of cervical carcinoma and three normal cervix samples. The second, GSE9750, employed the GPL570 platform (HG-U133 Plus 2) Affymetrix Human Genome U133 Plus 2.0 Array, with data collected from 33 samples of cervical carcinoma and 24 normal cervix samples. The third, GSE23126, was based on the GPL173097 platform Agilent-045997 Arraystar human lncRNA microarray V3. This platform enabled the identification of a metabolic-related risk signature that predicts the prognosis in cervical carcinoma and correlates with immune infiltration, using data from five samples of cervical carcinoma and six normal cervix samples. We used the R package GEO-query to download, process the data, construct the expression matrix, and match each probe to its corresponding gene symbol.

Identification of DEGs

We utilized the 'limma' package to explore DEGs within the GSE127265, GSE9750, and GSE173097 datasets. DEGs were identified using the cutoff criteria |logFC|>1 and FDR<0.05. Once these DEGs were determined, we proceeded to perform a series of analyses. This included PCA plots, Volcano plots, and hierarchical cluster analysis, conducted with the assistance of R packages ggplot2 and heatmap, respectively. Furthermore, to identify core genes, we constructed Venn diagrams utilizing the online tool available.

Function enrichment analysis

We utilized the online DAVID database for functional exploration of the DEGs. This tool enabled us to delve into the gene functions extensively. Furthermore, to investigate the principal functions of the selected genes and their involvement in various signaling pathways, we carried out GO functional and KEGG analyses.

PPI network construction and hub genes selection

To scrutinize the interactions among the DEGs identified earlier, we constructed a PPI network. This was achieved by using the Search Tool for Interacting Genes (STRING) available on the STRING database. Following this, we visualized the obtained PPI network using the CytoHubba plugin in Cytoscape software. Subsequently, we employed the MCC algorithm to identify the top 10 most crucial genes in the network, designating them as the hub genes.

Cell cultivation

We obtained the HeLa cell line, a human cervical carcinoma strain, from the Cancer Center of the Chinese Academy of Medical Sciences. The cells were cultured in flasks maintained at 37°C under a humid atmosphere with 5% CO2 . For this process, we used DMEM (Gibco, USA) enhanced with 10% FBS (Invitrogen, USA) and 1% penicillinstreptomycin solution (containing 100 U/mL penicillin and 100 μg/mL streptomycin). PG (CAS No.: 82-89-3) for the experiments was sourced from MCE (Shanghai, China).

Cell viability assay

HeLa cells were seeded in 96-well plates at a density of 1 × 10⁵ cells/mL, using 100 μL of DMEM medium supplemented with 10% FBS in each well. After 24 h, the cells were exposed to varying concentrations of luciferin: 0, 1, 10, 50, and 100 μM for another 24 h period. Cell viability was then assessed using the Cell Counting Kit-8 (CCK8, MEC, Shanghai, China).

DAPI staining

To observe nuclear morphological changes in apoptotic cells, we employed DAPI staining. HeLa cells, at a density of 1 × 10⁵ cells/ well, were cultured overnight in 24-well plates using DMEM medium supplemented with 10% FBS. Following this, they were exposed to varying concentrations of luciferin (0.1, 1, 2, 10, 50, and 100 μM) for 24 h. Subsequently, the medium was discarded and cells underwent two washes with cold PBS. They were then fixed in 100% ethanol at room temperature for 20 minutes and washed with PBS twice more. Finally, we visualized the cells under a fluorescent microscope (IX70- SIF2 Olympus; Olympus).

Apoptosis flow-cytometry assay

To evaluate PG-induced apoptosis, we employed double staining using Annexin V-FITC and PI. HeLa cells were cultured for 24 h in DMEM medium supplemented with 10% FBS and 1 μM luciferin in 4 cm² dishes. The cells were subsequently harvested, washed twice with PBS, and a count of 5 × 105 cells was resuspended in binding buffer. These were then stained with Annexin V-FITC and PI for 20 minutes using the Annexin V-FITC Apoptosis Detection Kit (Biyuntian, Nanjing, China). Post-staining, flow cytometry analysis was promptly performed (FACScan; BD Biosciences, Milano, Italy). On the flow cytometry readouts, cells in the Q2 (upper right), Q3 (lower left), and Q4 (lower right) quadrants signify early apoptotic, surviving, and late apoptotic cells, respectively.

Real-time PCR analysis

We utilized real-time PCR to assess gene expression in sample tissues and in PG-treated HeLa and A549 cells. The cells were washed with cold PBS and treated with trypsin. Subsequent to this, the cell suspension underwent centrifugation at 1000 × g for 10 minutes. The resulting cell pellet was resuspended in 1000 μL of ice-cold PBS and then centrifuged again under similar conditions. The harvested cells were stored at -80°C for further use. Total RNA extraction from tissues and cells was carried out within 24 h of treatment, using TRIzol reagent per the manufacturer's instructions. RT-PCR was conducted with Biosharp One-Step RT-PCR (BL698A; Invitrogen) and Biosharp Taq DNA polymerase kit (BL699A; Invitrogen). The expression of target genes in control and treated samples, normalized to GAPDH mRNA, was determined using the 2-ΔCt method. The primers used for CDK1 were (forward: 5′- GGAGAAGGTACCTATGGAGTTGTG-3′, reverse: 5′- AGCACATCCTGAAGACTGACTAT-3′), for TOP2A (forward: 5′- ACGGAATGACAAGCGAGAAGTAA-3′, reverse: 5′- GCCAAAGCTGAGCATTGTAAA-3′), for AURKB (forward: 5′- TGCATCACACAACGAGACCTATC-3′, reverse: 5′- GAGTGAATGACAGGGACCATCAG-3′) and for GAPDH (forward: 5′-GGAGTCCACTGGCGTCTTCA-3′, reverse: 5′- GTCATGAGTCCTTCCACGATACC-3′).

Statistical analysis

Each experiment was conducted at least three times, and consistently yielded similar results. Data are represented as mean ± SD. To compare two groups, we used a t-test. When comparing more than two groups, ANOVA was implemented. If ANOVA identified a significant difference, then multiple-comparison tests were subsequently employed. We considered a p-value of less than 0.05 as statistically significant.

Identification and analysis of differentially expressed genes in cervical cancer

We employed the R package limma to identify Differentially Expressed Genes (DEGs) between cervical cancer samples and normal control samples. An analysis of the expression matrix of GSE127265 revealed 1313 DEGs, which included 690 up-regulated and 623 downregulated genes. These genes were represented in a volcano plot and visualized in a clustering heat map (Figure 1a, b). In a similar vein, we identified an equivalent count of DEGs in the expression matrix of GSE9750 and GSE173097, illustrated in corresponding volcano plots and clustered heat maps (Figure 1c-f). To highlight the co-expressed genes across all three datasets (GSE127265, GSE9750, and GSE173097), we created Venn diagrams (Figure 1). This revealed 106 overlapping genes, which we subsequently deemed as core genes for additional analysis.

 

Figure 1: Differential gene expression in cervical cancer across multiple datasets. (a) DEG analysis from GSE127265. (b) Hierarchical clustering of DEGs in GSE127265. (c) DEG analysis from GSE9750. (d) Hierarchical clustering of DEGs in GSE9750. (e) DEG analysis from GSE173097. (f) Hierarchical clustering of DEGs in GSE173097. (g) Co-expression patterns across GSE127265, GSE9750, and GSE173097 datasets.

Analysis of DEGs' functional enrichment

We employed Gene Ontology (GO) function and Kyoto Encyclopedia of Genes and Genomes (KEGG) signaling pathway analyses to understand the roles and pathways associated with the DEGs. Our GO enrichment analysis indicated that the core genes are mainly involved in biological processes such as cell division. These genes are also prevalent in cellular components like chromosomal regions and DNA replication, and they play a role in molecular functions such as integrin binding, chemokine receptor binding, and DNA catalytic activity (Figure 2a). Additionally, our KEGG pathway enrichment analysis suggested that the core genes are primarily active in signaling pathways such as the cell cycle, DNA replication, and the p53 signaling pathway (Figure 2). This indicates that DEGs play a crucial role in modulating these vital biological processes.

 

Figure 2: Functional characterization and pathway analysis of differentially co-expressed genes. (a) GO functional annotation for 106 differential genes. The vertical axis represents the signaling pathway, the horizontal axis represents the gene count within each pathway, and color denotes enrichment significance. (b) KEGG pathway analysis for the 106 differential genes. Circle size corresponds to the number of enriched genes, while color indicates the significance of enrichment.

PPI network development and hub gene identification

A radial graph was employed to illustrate the biological processes and pathways associated with the Differentially Expressed Genes (DEGs) (Figure 3a). Utilizing the 106 ascertained DEGs, we established a Protein-Protein Interaction (PPI) network using the STRING database (Figure 3b). We identified the top 10 genes, characterized by high Maximal Clique Centrality (MCC), as hub genes within the PPI network through the CytoHubba plugin's MCC algorithm in Cytoscape software. The three genes, namely CDK1, TOP2A, and AURKB, with the most pronounced scores, were pinpointed as the pivotal genes within the PPI network (Figure 3).

 

Figure 3: Identification of core genes within the PPI network of coexpressed genes. (a) PPI network formation for co-expressed genes. (b) Identification of the top three core genes. The colors red, orange, and yellow correspond to genes with decreasing MCC values.

Association of CDK1, TOP2A, and AURKB expression with clinicopathological features in cervical cancer

To investigate the potential implication of CDK1, TOP2A, and AURKB in the initiation and progression of cervical cancer, we conducted an evaluation of their expression in 10 cervical cancer samples through Immunohistochemistry (IHC). We observed upregulated levels of CDK1, TOP2A, and AURKB, at both protein and gene levels, in the cervical cancer tissues (Figure 4). Significantly, the expression patterns of these genes echoed the data derived from our database. Such findings suggest that CDK1, TOP2A, and AURKB could be utilized as prognostic indicators for cervical cancer.

 

Figure 4: CDK1, TOP2A, and AURKB gene and protein expressions in cervical cancer specimens. (a) H and E staining. (b-d) Immunohistochemically staining images for CDK1, TOP2A, and AURKB in normal cervical and cancerous tissues, respectively. (e-g) Gene expression levels of CDK1, TOP2A, and AURKB in both normal and cancerous cervical tissues. Data represent mean values from three independent experiments: **p<0.01; ***p<0.001 in comparison to corresponding control values. Scale bar: 40 μm.

Selective cytotoxic effects of PG on HeLa, H8, and A549 cell lines

To delineate the cytotoxicity of PG on distinct cell types, we examined its impact on the HeLa cervical cancer cell line, the H8 normal human cervical epithelial cell line, and the A549 lung cancer cell line. Upon exposure to PG concentrations ranging from 1-200 μM, HeLa cells exhibited marked concentration and time-dependent growth inhibition relative to untreated counterparts (Figure 5a, b). In contrast, H8 cells, emblematic of normal cervical epithelial cells, remained largely unaffected at lower PG concentrations (Figure 5c), although elevated PG levels induced a decrement in their viability over prolonged durations (Figure 5d). This pattern underscores the selective cytotoxic potency of PG on HeLa cells. In a parallel observation, the A549 cell line, representative of a cancer phenotype, manifested significant growth suppression post PG treatment spanning 1-200 μM concentrations (Figure 5).

 

Figure 5: PG's impact on the viability of HeLa, H8, and A549 cells. (a-b) CCK8 assay depicting HeLa cell viability post-luciferin treatment at 24 and 48 h. (c-d) CCK8 assay illustrating H8 cell viability after luciferin exposure at 24 and 48 h. (e-f) CCK8 assay showing A549 cell viability following luciferin treatment at 24 and 48 h. Data represent mean ± SEM from three independent experiments: *p<0.05; **p<0.01; ***p<0.001 relative to respective control values.

PG induces apoptosis in HeLa cells

To explore the potential apoptotic effects of PG on HeLa cells, we employed DAPI staining and Annexin V FITC/PE double staining techniques (Figure 6). Following a 24 h exposure to PG, an approximate 30% of HeLa cells demonstrated apoptosis relative to the control group. Such findings underscore PG's possible efficacy in inducing apoptosis in HeLa cells.

 

Figure 6: Apoptosis assessment in PG-treated HeLa cells using flow cytometry. (a-f) Depict the flow cytometry results. Data represent the mean from three independent experiments: ***p<0.001 in comparison to corresponding control values.

Impact of PG treatment on CDK1/TOP2A/AURKB expression in HeLa cells

To further elucidate the influence of PG treatment on the expression levels of CDK1/TOP2A/AURKB at the protein and gene levels in HeLa cells, we undertook Western blot and RT-PCR assays. The results revealed a notable reduction in both protein and mRNA levels of CDK1/TOP2A/AURKB in the PG-treated cells compared to the controls (Figure 7).

 

Figure 7: CDK1, TOP2A, and AURKB expression in HeLa cells post-PG treatment. After PG treatment, CDK1, TOP2A, and AURKB protein and gene levels in HeLa cells were reduced in comparison to control (a-e). Data represent the mean from three independent experiments: *p<0.05; **p<0.01; ***p<0.001 relative to corresponding control values.

Cis is one of the most effective anticancer drugs for neoadjuvant therapy and advanced cervical carcinoma. Its prolonged use, however, can diminish drug sensitivity, leading to drug resistance. Separately, PG has been identified to have targeted cytotoxicity towards cancer cells, allowing the selective elimination of these malignant cells while sparing normal cells. Notably, PG has been reported to induce apoptosis in various tumor cells, including UCCs, U87MG, HT-29 and MCF-7. Our research echoes these findings, revealing that PG can trigger apoptosis in Hela cells a finding that aligns with the results presented by Lin et al. This apoptotic action in Hela cells involves the activation of molecules like Bcl-2, Bax, and caspase-3, although the precise mechanism remains to be elucidated. In addition, while low concentrations of PG showed no toxicity towards H8 cells, it did cause morphological alterations in A549 cells, further underscoring its targeted impact on tumor cells.

In our research, we extracted 106 differentially co-expressed genes from the GEO database. These DEGs predominantly feature in processes like cell division, proliferation, chromosomal regions, integrin binding, cell cycle, DNA replication, and the p53 signaling pathway. Utilizing the CytoHubba plugin of the Cytoscape software, we identified 10 pivotal genes: CDK1, TOP2A, AURKB, RRM2, MAD2L1, BUB1B, CDKN3, BUB1, KIF11, and CCNB2. A deeper dive into their expression profiles indicated that, relative to normal cervical tissues, all these genes showed pronounced upregulation in cervical carcinoma samples. Notably, CDK1, TOP2A, and AURKB stood out due to their vivid representation, and there was a marked correlation between their expression and survival outcomes in cervical carcinoma patients.

CDK1, TOP2A, and AURKB are pivotal genes for DNA replication. Post-PG treatment, a reduction in their expression levels was noted in both HeLa and A549 cells. Specifically, CDK1 facilitates the G2/M transition by orchestrating the centrosome cycle and initiating mitosis, underscoring its importance in the eukaryotic cell cycle. Interestingly, elevated levels of CDK1 have been detected in breast cancer tissues. Diminishing its expression has been linked to increased apoptosis in these cancer cells. Topoisomerase IIA, on the other hand, modulates the DNA topological state during transcription and is intrinsically linked to processes like DNA replication, chromosome segregation, and condensation. Elevated expression of TOP2A, more profound in lung adenocarcinoma tissues than in their normal counterparts, correlates with enhanced proliferation, migration, and invasion of lung cancer cells in vitro. Aurora B, also recognized as AURKB, is fundamental for chromosome segregation and cytokinesis. Intriguingly, elevated Aurora B levels were observed in ovarian cancer A2780 cells. Suppressing its expression led to an augmented proportion of G2/M phase cells and polyploidy in these cells, a scenario followed by increased cell apoptosis and hindered proliferation. This study reveals PG's concentration-dependent inhibitory effect on cancer cell proliferation. At non-cytotoxic levels, PG compromises HeLa cell viability and quantity. At cytotoxic concentrations, PG disrupts key cancer cell growth pathways, including protein kinase signaling and mitochondrial cell death pathways, triggering apoptosis. PG's interaction with cancer cell DNA facilitates oxidative DNA cleavage, leading to oxidation, breakage, and apoptosis.

In conclusion, our bioinformatics analysis unveiled the DEGs associated with cervical carcinoma, along with their related biological processes and pathways. Through validation with the GEO database, genes CDK1, TOP2A, and AURKB emerged as potential biomarkers for the precise diagnosis and treatment of cervical carcinoma. We also ascertained that PG can both trigger apoptosis in HeLa cells and inhibit the overexpression of pivotal genes such as CDK1, TOP2A, and AURKB. Thus, our results lay the groundwork for considering PG as a therapeutic option for cervical carcinoma. However, our study isn't without limitations. While we pinpointed the targets of PG in cervical carcinoma cells, an in-depth exploration of the underlying molecular mechanisms remains a future endeavor. Furthermore, our research was confined to human cervical carcinoma cell lines. To fully comprehend PG's in vivo effects on cervical carcinoma, subsequent investigations should integrate animal models, enriching the depth and breadth of this line of research.

The authors have no conflicts of interest to declare.

This study protocol was reviewed and approved by the Ethics Committee of Xiangyang Central Hospital (approval number: 2023-022).

All subjects or their immediate family members provided written informed consent before participation.

This research was supported by National Natural Science Foundation of China (81972449), Medical Research Project of HuBei Pediatric Alliance (HPAMRP202402), Foundation of Hubei University of Arts and Science (XK2019046) and Xiangyang Central Hospital research project (2023YZ04, 2024YJ11A).

Conceptualization, experiment and data analysis, ZZ and QC; resources, HX and MT; writing the main manuscript text, ZZ and MT. All authors reviewed the manuscript.

All data generated or analyzed during this study are included in this published article and its supplementary information files.

1. Dhanaworavibul K (2023) . Oncol Res Treat 46: 192-200.

   [] [] []

2. Cohen PA (2019) . Lancet 393: 169-182.

   [] [] []

3. Kumar Kore R (2024) . Cancer Invest 5: 1-16.

   [] [] []

4. Rosen VM (2017) Int J Gynecol Cancer 27: 1237-1246.

   [] [] []

5. Cheng SY (2018) . Apoptosis 23: 314-328.

   [] [] []

6. Darshan N, Manonmani HK (2015) . J Food Sci Technol 52: 5393-5407.

   [] [] []

7. Sumathi C (2014) . Sci World J 290327.

   [] [] []

8. Prabhu VV (2016) Cancer Res 76: 1989-1999.

   [] [] []

9. Espona-Fiedler M (2012) . Biochem Pharmacol 83: 489-496.

   [] [] []