Roles of Beclin1 protein expression in cervical cancer: a meta-analysis and bioinformatics analysis

Abstract

Beclin1 is a key regulator of a family of autophagy-related proteins. The aim of our study was to elucidate the clinicopathological and prognostic significance of Beclin1 expression which is a positive regulator of autophagy in cervical cancer. The results showed that a total of 2682 patients were enrolled in 21 case-control studies. The results showed that, as for Beclin1 expression, significant differences were found in cervical cancer vs. normal cervical tissues (p<.00001) and cancer tissues with vs. no lymph node metastasis (p<.00001); tumour diameter no less than vs. less than 4 cm (p=.001), myometrial invasion depth no less than vs. less than 1/2 and FIGO I vs. II (p=.02); relationship between Beclin1 expression and prognosis of cervical cancer (p=.03). Kaplan–Meier’s plotter showed that Beclin1 expression was negative. It was associated with overall, post-progressive and distant metastatic survival. According to the Oncomine database, Beclin1 mRNA expression in cervical cancer tissues was higher than that in normal tissues. Cox multivariate showed that lymph node metastasis and TNM stage were important factors affecting the survival time of patients. Beclin1 expression can be used as an indicator of prognosis in patients, and provide methods and ideas for prevention and treatment.

Keywords:

• Beclin1
• cervical cancer
• meta analysis
• clinicopathological features
• prognosis

Introduction

Cervical cancer is the second leading cause of cancer death in women aged 20–39 years worldwide (Haran et al. 2021). Most cervical cancers are squamous cell carcinomas, which originate in the squamous epithelial cells of the cervix (Zhong et al. 2020). Studies have shown that persistent infection with high-risk human papilloma virus (hrHPV) has been identified as a major factor that leads to cervical cancer. As the extensive screening programs, we adopt periodic surveillance through hrHPV and Pap smear-based testing can play a key role in reducing the associated morbidity (Tsikouras et al. 2016; Naz et al. 2018), so its prolonged pre-cancerous phase can last 10–20 years (Bray et al. 2018). Whereas, cervical cancer is aggressive and often detected in advanced stages. In the present situation, we shed light on our current understanding and knowledge of racially disparate outcomes in cervical cancer (Olusola et al. 2019) with concepts and knowledge about its prevention and treatment evolving rapidly, emerging new concepts and technologies for cancer interventions, and more urgently (Zheng and Ding 2018).

Beclin1, also known as BecN1, is considered to play an important role in cellular postures and inflammatory responses (De et al. 2019). It is the first mammalian autophagy gene identified, is an important positive regulator of autophagy and also a marker of autophagy initiation. Beclin1 contains 450 amino acid sequences and with a protein molecular weight of 60 kDa (Gao J et al. 2018; Chen M et al. 2019; Leonard et al. 2019). It can interact with members of the Bcl family of anti-apoptotic proteins and interact with a variety of proteins to regulate autophagosome formation and maturation (Sun Y et al. 2011). Overexpression of Beclin1 may enhance apoptosis signalling in cervical cancer (Sun Y et al. 2009).

Previous studies have shown that the autophagy gene Beclin1 is directly related to the occurrence and development of tumours by regulating autophagy activity. In common gynaecological malignancies such as cervical cancer, ovarian cancer and endometrial cancer, there is a phenomenon of decreased autophagy activity (Huang et al. 2013). Beclin1 expression was decreased in cervical cancer tissues, and it was also closely related to pelvic lymphatic metastasis and tumour histological grade. Beclin1, as a tumour suppressor gene, is characterised by low expression or deletion of a single allele in various tumours of different races. Studies have confirmed that Beclin1 can form different chelates with related molecules, including Beclin1-UVrag-PI3K, Beclin1-UVrag-Rubicon-PI3K and Beclin1-Atg14-PI3K, suggesting that Beclin1-mediated autophagy of tumour cells is closely related to the EGFR/PI3K/Akt signalling pathway. The relationship between Beclin1 expression and cervical cancer with its clinicopathological features is not identical (Zhu et al. 2012).

Studies have suggested that Beclin1 protein expression in cervical cancer is closely related to the differentiation degree of cervical cancer and size of lesion (Zhang Q et al. 2019). Some studies have shown that Beclin1 expression is related to cervical cancer accompanied by lymph node metastasis, and other studies have manifested that Beclin1 expression is related to the invasion depth of cervical cancer (Gao JX et al. 2017). Therefore, in this study, we sought to conduct a meta-analysis to estimate the prognostic importance of elevated Beclin1 expression for survival among patients with cervical cancer.

Materials and methods

Published literature search and data extraction

In this research, we discovered the papers collected in PubMed, Medline, ScienceDirect, and Chinese Journal of Science on January 31 2021. The key words were Beclin1 and the cervical (cervix OR cervix uteri OR neck of uterus or uterine neck) AND (cancer OR carcinoma OR tumour OR squamous cell carcinoma). Study selection criteria included: (1) published literature in both Chinese and English was restricted to cervical cancer patients; (2) Beclin1 expression and prognosis in patients with cervical cancer were detected by immunohistochemistry; (3) none of the cervical cancer patients received preoperative radiotherapy or chemotherapy. Exclusion criteria included: (1) reviews, case reports, abstracts and meeting minutes; (2) repeat publications; (3) unclear diagnosis. The quantity of the literature is poor and cannot be used, and for duplicate publication or data. The similarity studies included only good methodological quality.

Data extraction and methodological quality evaluation of included studies

As shown in Table 1, two evaluators (Ma GY and Zhang ZG) were selected independently according to inclusion and exclusion criteria: literature, extraction of data, and evaluation of methodological quality of included studies. In case of disagreement, the dispute shall be resolved through discussion or adjudicated by a third party. Using self-made data extraction table to extract data, the extraction content mainly includes: title, author, publication year, original literature source, sample size and uterus; neck cancer and lymph node metastasis, clinical stage, histological grade, etc. The methodological quality of the included studies was assessed using the NOS scale, which included three dimensions (nine points) of study subject selection (four items, four points), intergroup comparability (one item, two points) and outcome measurement (three items, three points). If there is a disagreement, we discussed and solved it. If necessary, we discussed with the third party and finally reached a consensus.

Table 1. Main characteristics of eligible studies.

Author	Country	Antibody supplier	Cases	Ctr	Risk to cancer	Follow-up outcome	Quality
Zhang XN et al. (2018)	China		72	20	Reduced	Positive	7
Zhou M et al. (2020)	China	R&D	80	20	Reduced		8
Zhang J et al. (2019)	China	Nakasugi Jinqiao	100	100	Reduced		9
Zhang J et al. (2017)	China	Abcam	50	20	Reduced		8
Zhang XN et al. (2016)	China		90	20	Reduced		8
Wu RH and Zheng (2010)	China	Santa Cruz	85	30	Reduced		8
Wu L and Guo (2017)	China		50	20	Reduced		8
Li XL et al. (2020)	China	ABGENT	50	50	Reduced		7
Meng et al. (2019)	China	Abcam	54	13	Reduced	Positive	7
Li Y et al. (2013)	China		50	12	Reduced		7
Wang ZH et al. (2010)	China	Santa Cruz	81	20	Reduced		7
Zhou GQ et al. (2019)	China	Abcam	120	50	Reduced		8
Wu GR et al. (2020)	China	Abcam	47	15	Reduced		7
Zhao et al. (2017)	China	Santa Cruz	80	20	Reduced		8
Wang XW et al. (2018)	China	Santa Cruz	30	30	Reduced		7
Li Y et al. (2018)	China		45	45	Reduced		7
Hang et al. (2020)	China	Abcam	80	30	Reduced		7
He and Ren (2019)	China	Abcam	43	20	Reduced		8
Zhang XN et al. (2017)	China		43	40	Reduced		8
Zhu et al. (2012)	China		50	26	Reduced	Positive	7
Cheng HY et al. (2012)	China		122	31	Reduced		8

Bioinformatics analysis

The expression level of Beclin1 gene was analysed using Oncomine (www.oncomine.org), the world's largest oncogene chip database and integrated data mining platform. Kaplan–Meier’s plot was used to analyse the prognostic significance of Beclin1 mRNA expression. The expression level of Beclin1 was divided into high expression group and low expression group according to the optimal threshold value given by Kaplan–Meier’s graph. Beclin1 expression was associated with overall survival (OS), relapse-free survival (RFS), distant metastasis-free survival (DMFS) and post-progressive survival (PPS) of all patients. There are two ways to do this. Multiple analysis (multiple change) was used to calculate the expression rate of each gene under the two conditions, which was generally in the range of 0.5–2.0, and there was no significant difference in the expression of this gene. T test (p value), the gene whose T statistic exceeds a certain value is detected as abnormal. The confidence of the difference was calculated to analyse whether it was statistically significant. Beclin1 mRNA levels in 13 cervical carcinoma tissues and 60 normal cervical epithelial tissues were compared. We performed a logarithmic transformation of all data, centred the median of each array, and normalised the standard deviation of each array to a single value.

Statistical analysis

RevMan 5.2 software (version 5.2; www.cochrane.Usinges/Download/Files/revman.HTM); meta analysis, Beclin1 expression was estimated according to the clinicopathological parameter OR and its 95%CI effect size in cervical cancer patients. First, χ² test was used to test the heterogeneity of the original studies included in the study. When the heterogeneity was not significant, i.e. p ≤ .1, there was heterogeneity among the results, and the fixed effect model (Mantel-Haenszel method) was used. If p > .1, the random effects model (Der Simonian and Laird method) was used. I² was used to quantitatively analyse the heterogeneity effect at the cut-off values of 25, 50 and 75%. When p < .1 and I² > 50%, significant heterogeneity was considered among the results. A meta-analysis was performed based on the model of machine effect. If the heterogeneity was too large, only descriptive analysis was performed. If p > .1 and I² < 50%, no significant heterogeneity was considered to exist among the study results, and a fixed-effect model was used for meta-analysis. Funnel plots were used to evaluate publication bias, and Begg's test and Egger's test were used to evaluate whether the funnel plots were consistent. Funnel plots were used to evaluate publication bias, and Begg's test and Egger's test were used to evaluate whether the funnel plots were consistent. Meta-analysis was performed using RevMan 5.2 software and SPSS software (version 10.0; SPSS, Inc., Chicago, IL) and t-test. Bilateral p < .05 was considered to be statistically significant.

Results

Literature retrieval results

As shown in Figure 1 and Table 1, duplicate studies, including review literature and animal experiments were excluded by reading the abstract. After preliminary search, a total of 128 literatures were selected. After further screening, a total of 33 literatures (Zhang XN et al. 2018; Liu Y et al. 2019; Zhou M et al. 2020) were included in qualitative synthesis. Finally, only 21 literatures discussing the relationship between Beclin1 expression and clinicopathological or prognostic expression indexes of cervical cancer were included (Zhang J et al. 2019; Li XL et al. 2020; Sun XD et al. 2021). Exclusion criteria are as follows: (1) repeated publication or repetition of research. (2) only proceedings, reviews and abstracts of research; (3) to study the expression of maspin by Western blot, RT-qPCR, cDNA microarray or transcriptome sequencing; (4) insufficient data (Figure 1).

Figure 1. Flow diagram of article selection.

We included a total of 2682 cases in 21 controlled studies, including 1460 cases in the cervical cancer group, 590 cases in the cervical intraepithelial neoplasia (CIN) group and 632 cases in the normal cervical tissue group. The correlation between Beclin1 expression and cervical cancer clinicopathological parameters was plotted by dominance ratio (OR) forest plot. There was moderate heterogeneity in the results of each study (p = .04, I² = 37%), which may be due to potential selective bias of subjects in each study and no history of disease in the control group. Meta-analysis random-effects models showed that Beclin1 expression was significantly different in cervical cancer tissue vs. normal cervical tissue, with Beclin1 expression being lower in the cervical cancer group than in the normal cervical group ((OR = 0.10, 95%CI 0.08–0.13, p < .00001) (Figure 2(A))).

Figure 2. Forest plot of the relationship between Beclin1 expression and clinicopathological features of cervical cancer. (A) Cancer and normal tissue. (B) Cancer and CIN. (C) CIN and normal tissue. (D) Lymph node metastasis. (E) Degree of differentiation. (F) Depth of invasion. (G) Tumour size. (H) Patient survival rate. CIN: cervical intraepithelial neoplasia.

A total of 13 studies reported Beclin1 expression in the cervical cancer group and the CIN group, including 895 cases in the cervical cancer group and 575 cases in the CIN group; meta-analysis results of the fixed-effect model showed that Beclin1 expression in the cervical cancer group was lower than that in the CIN group, and the difference was statistically significant (OR = 0.26, 95%CI 0.18–0.38, p < .0001)) (Figure 2(B)). A total of 13 studies reported CIN group with normal palace neck tissue Beclin1 expression, among which 590 cases of CIN, normal cervix group of 331 as example, the results had moderate heterogeneity (p = .41, I² = 4%), the heterogeneity may have originated from different potential system characteristic differences between the research object. Results of meta-analysis of random response model showed that Beclin1 expression in the CIN group was lower than that in the normal cervical tissue group, and the difference was statistically significant (OR = 0.36, 95%CI 0.26–0.50, p = .41) (Figure 2(C)). A total of 17 studies reported Beclin1 expression in the group with lymph node metastasis and the group without lymph node metastasis, including 430 patients in the group with lymph node metastasis and 794 patients without lymph node metastasis. There was no statistical heterogeneity among the results (p = .0002, I² = 64%). Meta-analysis results of the fixed-effect model showed that Beclin1 expression in the group with lymph node metastasis was lower than that in the group without lymph node metastasis, with a statistically significant difference (OR = 0.43, 95%CI 0.26–0.72, p < .00001) (Figure 2(D)). A total of 15 studies were conducted in the middle and low differentiation group vs. the high differentiation group. Beclin1 expression was reported in the medium and poorly differentiated group and the highly differentiated group, including 509 cases in the medium and poorly differentiated group and 636 cases in the highly differentiated group. There was a high degree of heterogeneity among the results (p < .0001, I² = 71%), which may be due to the influence of exposure factors and potential confounding factors. Meta-analysis results of random effects model showed that Beclin1 expression was significantly different between low and medium differentiation groups and high differentiation groups (OR = 2.02, 95%CI 1.14–3.55, p = .02) (Figure 2(E)). Beclin1 expression was reported in seven studies, including 213 cases in the group with muscle invasion depth ≥1/2 and 266 cases in the group with muscle invasion depth <1/2. There was no heterogeneity (p = .13, I² = 39%). Meta-analysis results of the fixed-effect model showed that there was a statistically significant difference in Beclin1 expression between the two groups (OR = 1.60, 95%CI 1.07, 2.39, p = .02) (Figure 2(F)). Beclin1 expression was reported in the tumour ≥4 cm group (329 cases) and the tumour <4 cm group (410 cases) in 4 cm group. There was a statistical difference between the results (p = .41, I² = 3%). Meta-analysis of the fixed-effect model showed that there was a statistically significant difference in Beclin1 expression between the two groups (OR = 1.71, 95%CI 1.23–2.37, p = .03) (Figure 2(G)).

According to the above results, Beclin1 expression was up-regulated in normal mucosal tissues compared with that in cervical cancer and precancerous lesions (Figure 2(A,B)). Forest maps associated with OR and Beclin1 expression were divided according to gender (Figure 2(C)), depth of invasion (Figure 2(D)), lymph node metastasis (Figure 2(E)), TNM stage (Figure 2F) and Lauren’s classification (Figure 2(G)).

The pathological data presented in each of the literatures are different. If a specific clinicopathological feature is missing, the article is excluded. As a consequence, the survival data are shown in Figure 2(H), accordance with the four data sets mentioned above. We investigated the association between Beclin1 expression and decreased survival in patients with cervical cancer and found that it was significant (HR, 0.39; 95% confidence interval, 0.17, 0.90; p < .0001).

Publication bias

The mining of publication bias can be prevented quantitatively using funnel plots, as shown in Figure 3. Individuals removed these studies from the pooled analysis, and sensitivity analysis was used to assess the individual's impact on the pooled results of the study. According to Eger's tests, this meta-analysis had no significant publication bias.

Figure 3. Funnel plot for testing publication bias between Beclin1 expression cervical cancer. Publication bias was analysed according to the degree of risk of Beclin1 expression in (A) cervical cancer. Publication bias was also tested between Beclin1 expression and clinicopathological features of cervical cancer, including (A) cancer and normal tissue. (B) Cancer and CIN. (C) CIN and normal tissue. (D) Lymph node metastasis. (E) Degree of differentiation. (F) Depth of invasion. (G) Tumour size. (H) Patient survival rate. CIN: cervical intraepithelial neoplasia; SE: standard error; RD: risk difference.

Relationship between Beclin1 expression and bioinformatics characteristics of cervical cancer

Bie's and D 'Pyeon's data sets showed that, based on bioinformatics characteristics, Beclin1 mRNA expression in cervical cancer tissues was higher than that in normal tissues (Figure 4(B,C), p < .05). According to Kaplan–Meier’s diagram (Figure 4(A)), higher Beclin1 mRNA expression was positively correlated with OS of all cervical cancer patients. The OS time of cervical cancer patients with high Beclin1 expression was higher (p < .05), and only significantly correlated with the OS rate (p < .05).

Figure 4. Beclin1 mRNA expression was positively correlated with (A) overall survival in patients with cervical cancer. HR: risk ratio. Beclin1 mRNA expression in the occurrence of cervical cancer. Bioinformatics analysis was performed using Bie's and D 'Pyeon's data sets to analyse the expression of Beclin1 mRNA during the development of cervical cancer. Beclin1 expression was lower in (B) cervical cancer tissues than in (C) normal cervical epithelial tissues.

Discussion

Beclin1 is a tumour suppressor gene, so the higher the expression level of Beclin1, the better the prognosis of tumours, and plays a significant part in the development of tumour diseases (Stang 2010; Wang J et al. 2017; Wang YY et al. 2017). Beclin1 is a gene that positively regulates autophagy and can regulate the process of autophagy. Beclin1 suppresses tumours primarily by blocking mutations in cancer-related genes. It can induce autophagy apoptosis of tumour cells. It can inhibit the angiogenesis of tumour tissue (Duan and Peng 2011). Cheng et al. conducted an in vitro culture of cervical cancer cells with rectal metastases and found that Beclin1 was able to inhibit the metastasis and invasion of the metastatic cell line HeLa cells, suggesting that the inhibitory effect of Beclin1 on cell proliferation was relevant (Cheng YX et al. 2016). Although there are few studies on Beclin1 expression in cervical cancer at home and abroad, some studies have found that Beclin1 protein expression is decreased in brain tumours, lung cancer, ovarian cancer, breast cancer and endometrioid adenocarcinoma, especially the progression and prognosis of cervical cancer is associated with decreased Beclin1 expression level (Cerit et al. 2018). Studies have shown that the positive rate of Beclin1 in cervical cancer tissues, cervical squamous intraepithelial lesions and normal mucosa increases successively, and Beclin1 gene inhibits the formation and development of cervical cancer, which has a certain correlation with its prognosis (Liu C et al. 2013). This is consistent with the trend of the results of our experimental analysis.

Previous studies have shown that Beclin1 expression levels were significantly lower in the cervical cancer group than in normal cervical tissue, while reduced Beclin1 expression was associated with poor prognosis in human cervical cancer, and Beclin1 mRNA expression was also negatively correlated with the OS of cervical cancer patients (Ou et al. 2014; Su and Liu 2014). This is consistent with the results of our meta-analysis, indicating that Beclin1 expression levels are closely associated with different clinicopathological features of cervical cancer, and that when Beclin1 expression is reduced it increases the risk of cervical cancer development. Meanwhile, our study found that Beclin1 expression was reduced in patients with lymph node metastasis, suggesting that Beclin1 protein expression is a promoter of lymph node metastasis in cervical cancer, which is consistent with the findings of Ma XH et al. In the comparison of the degree of cervical cancer differentiation, FIGO clinical stage, tumour diameter and depth of myxoid infiltration, Beclin1 expression level was closely correlated with the above clinical case information, and the difference was statistically significant (p < .05), which was consistent with the findings of Li et al. (Li Y, et al. 2021; Ma XH et al. 2021). Studies have shown that the abnormal phenomenon of autophagy is common in tumours, and the duality of autophagy in the occurrence, development, invasion and metastasis of tumours has made it a research hotspot. In various cancers such as ovarian cancer, breast cancer, liver cancer and prostate cancer, the activity of autophagy-related pathways or proteins has been down-regulated or deleted, which has an important impact on the survival and prognosis of patients. Studies have found that when Beclin1 expression is significantly up-regulated, Beclin1 mRNA expression is also significantly increased. This suggests that autophagy plays an important role in cervical carcinogenesis. Beclin1 plays an important role in the initiation of autophagy, mainly by forming trimer with PI3K and Atg14 and continuously recruiting autophagy-related proteins to mediate the initiation of autophagy (Altman and Nelson 2016; Gadducci et al. 2017; Chen H et al. 2021). Beclin1 is located on human chromosome 17q21 and plays a dominant role in autophagy as an indispensable regulator of autophagy by regulating type III PI3K-dependent phosphatidylinositol 3-phosphate (PI3P) production and subsequent supplementation of ATG proteins to promote autophagosome formation (Wong et al. 2018). Beclin1 is a homologous gene of yeast Atg6/VPS30. It binds to PI3K through a conserved region of its structural domain to form a complex involved in autophagic lysosome formation and induces cellular autolysis. Beclin1 was found to be a single allele deletion or mutation in approximately 55% of human colorectal and breast cancers, and its expression was downregulated in uterine smooth muscle sarcoma (Zhu et al. 2012; Hu et al. 2015; Mousavi et al. 2016).

In conclusion, the current evidence suggests that Beclin1 protein expression is significantly associated with cervical cancer and its different clinicopathological features, while Beclin1 mRNA expression is negatively correlated, suggesting that abnormal Beclin1 expression may predict the pathological behaviour of cervical cancer. These results suggest that the expression levels of Beclin1 mRNA gene can be used to predict the corresponding protein levels, suggesting that Beclin1 may play an important role in the development of cervical cancer. Therefore, our systematic evaluation may enrich the study of the pathogenesis of cervical cancer more accurately, provide new ways for cervical cancer screening and provide new targets for cervical cancer gene therapy. However, due to the limitation of the number and quality of included studies, it cannot be used as an independent factor to judge the prognosis of cervical cancer. Therefore, the above findings need to be validated by more high-quality, large-sample and rigorously designed studies. There are several limitations to our meta-analysis. First, the potential publication bias comes from the fact that the published results were predominantly positive. Second, the patients included in the study were only from Asia and the United States. Due to the different levels of medical development in different regions, the experimental methods used to detect Beclin1 expression may also differ, which may affect the results. Third, survival data were extracted from the survival curves, which may affect the results. Fourth, the small sample size may affect the relevance of some articles. In conclusion, Beclin1 protein is down-regulated during the development of cervical cancer. Beclin1 was negatively correlated with cervical cancer infiltration depth, TNM stage, lymph node metastasis and survival prognosis. Beclin1 plays an inhibitory role in the development, progression and prognosis of cervical squamous cell carcinoma. Therefore, Beclin1 protein expression can be used as a potential marker to predict the prognosis of cervical cancer patients, which is expected to provide new strategies for the treatment of cervical cancer.

Conclusions

Beclin1 plays a complex role in the development of tumours. Meanwhile, Beclin1 expression was up-regulated in patients with cervical cancer. Beclin1 was negatively correlated with TNM stage, histological grade and lymph node metastasis in patients with cervical cancer. Beclin1 expression can be used as a good marker for prognosis of cancer patients, providing new methods and ideas for prevention and treatment.

Acknowledgements

All procedures carried out in studies involving human participants are in accordance with the ethical standards of institutions and/or national research councils, as well as with the 1964 Helsinki Declaration and its subsequent revisions or similar ethical standards. Formal consent is not required for this type of research.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

All data used in this paper are from published articles and public data platforms, including Kaplan–Meier plotter and Oncomine. The prognostic significance of Beclin1 mRNA was analysed using Kaplan–Meier plotter (http://kmplot.com). Additionally, the expression level of Beclin1 was analysed using Oncomine (www.oncomine.org), a cancer microarray database and web-based data mining platform for a new discovery from genome-wide expression analyses.

Additional information

Funding

This work is supported by Science and Technology Bureau of Hebei Province.