TPP1 Regulates hTERT Expression and Predicts Early Malignant Event and Prognosis of Cervical Cancer


 Background: Cervical cancer is one of the most common deadly cancer in women worldwide. However, identifying specific biomarkers is still needed. Telomere-binding protein 1 (TPP1) is vital to telomerase activity. However, the role of TPP1 in cervical cancer and its association with human telomerase reverse transcriptase (hTERT) is unclear.This study aimed at exploring the role of telomere-binding protein 1 (TPP1) in cervical cancer development and progression, and potential mechanisms.Methods: Tissue samples from a total of 274 participants were enrolled for the evaluation of protein expression,156 of whom diagnosed withcervical cancers, 102 with cervical intraepithelial neoplasia (CIN) and 16 with normal cervix. In addition, in vitro cellular models with cervical cancer cell lines Hela, Siha, and C33a were transfected by TPP1-siRNAand protein expression of TPP1 and hTERT were assessed. Results: Compared with normal cervix, TPP1 expression was significantly higher in CIN-III and cervical cancers (P<0.001 for both). High expression of TPP1alone (Plog-rank=0.047)andhigh co-expression of TPP1/hTERT (Plog-rank=0.005)weresignificantly associated with worse survival of cervical cancer patients.After adjusting for well-known prognosis factors, hazard ratio was 2.03(95% confidence interval [CI] 0.99-4.16)for high expression of TPP1 and 2.01(95% CI 1.10-3.67) for high co-expression of TPP1/hTERT. TPP1 and hTERT expressions were positively correlated atall levels of cervical lesions (r=0.524, P<0.001). Knockdown of TPP1 decreased hTERT mRNA and protein expression.Conclusions: High expression of TPP1 might be an early event during cervical cancer development and could be served as apotential prognosis biomarker, especially when used together with hTERT. TPP1 might regulate hTERT expression with detailed underlying mechanisms warrant further investigation.


Introduction
In 2018, the Global Cancer Observatory (GLOBOCAN) estimated 569,847 new cervical cancer cases and 311,365 cancer-related deaths, making it the fourth most common cancer and the fourth most deadly cancer in women worldwide [1]. Cervical cancer severely affects women's health with more young women aged 15 to 49 years diagnosed during the last decade [2]. To detect precancerous lesions, to screen early cancers, and to improve patient survival, novel biomarkers and therapeutic targets are urgently needed.
Telomere, a DNA-protein complex locating at the end of chromosomes, together with telomere-binding proteins preserves chromosomal integrity and regulating cell cycles. Maintenance of telomere length and function is crucial during carcinogenesis, which is usually regulated by both telomerase and telomerebinding proteins [3]. Telomere-binding protein 1 (TPP1, also known as TINT1 or PTOP or PIP1) is a vital component of the telomere-binding proteins shelterin complex [4]. TPP1 binds to telomerase and protects chromosome ends from cellular DNA damages. TPP1 also initiates the functions of DNA repair machineries by recruiting telomerase to telomeres. These functions stimulate telomerase activity and promote telomere elongation in telomerase-positive cells [5][6][7]. Additionally, TPP1 combines with another shelterin complex component named protection of telomeres 1 (POT1) to form a stable heterodimer, which maintains genome stability and regulates telomerase-mediated telomere extension [6,[8][9][10].
Human telomerase reverse transcriptase (hTERT) is a fundamental element of telomerase and is known to be associated with cell stemness, cell proliferation and resistance to chemotherapy and radiotherapy in various cancers [11][12][13][14]. Similar to the nding that inhibition of hTERT activity shortens the telomere length and hinders its functions, TPP1 silent mutation using the CRISPR-Cas9 technology is also su cient to shorten telomeres in human cells [15]. It is suspected that TPP1 may regulate telomere length partly through hTERT. hTERT was shown to be a good biomarker for identifying and treating several cancers, but it has limited effect on cervical cancer [16][17][18]. Identifying speci c biomarkers for cervical cancer is therefore needed. This study aims to clarify the expression and clinical value of TPP1 and its association with hTERT in cervical cancer development and prognosis. Overall survival (OS) time was de ned as the date of surgery to the date of death from any cause, or the last date of follow-up, whichever came rst.

Immunohistochemical staining
Hematoxylin and eosin stain were used to con rm the primary pathological diagnosis.
Immunohistochemistry (IHC) test was performed to assess TPP1 and hTERT expression ( Fig. 1A and Fig.  S1A) following manufacturer's instructions. Primary antibodies against TPP1 (Abcam, ab57595, dilution of 1:100) and hTERT (Abcam, ab183105, dilution of 1:30) were incubated on 4 µm-section samples for 1.5 hours at room temperature. Two pathologists blinded to the outcome independently assessed the protein expression using a prede ned rating system. We used IHC rating scales for hTERT according to the previously published method [19], the total score of hTERT was rated by a combination of intensity and percentage of positive cells (Table S1). TPP1 was nuclear expressed and almost all the positive cells were strongly expressed. Therefore, TPP1 score was de ned according to the percentage of positive cells: < 25% of positive cells as negative (-), 10-49% as weakly positive (+), 50-74% as moderately positive (++), and ≥ 75% as strongly positive (+++). For each cervical tissue sample, two blocks per case were evaluated and an averaged score was applied. A discrepancy of scores from two pathologists was solved by re-evaluating and rescoring the samples together. For better interpretations, we de ned both negative (-) and weakly positive (+) rates as low expression, and moderately positive (++) and strongly positive (+++) rates as high expression.

Cell transfection
Three human cervical cancer cell lines, Hela, C33a, and Siha were purchased from the Institute of Life Sciences, Chinese Academy of Science (Shanghai, China). Cells were cultured in DMEM (Gibco, Life Technologies Corporation, UK) supplemented with 10% fetal bovine serum (FBS) and 100 UI/mL penicillin/streptomycin and dissociated by Accumax (Gibco). The plasmids (purchased from GenePharma, Shanghai) was used to knock down TPP1 with TPP1-siRNA sequence of 5'-GTGGTACCAGCATAGCCT-3' by means of Lipofectamine (LipofectamineTM2000, Invitrogen, US). Cells were plated in two six-well plate at 2 × 10 5 cells per well. After 24 hours of incubation, 5 µl of transfection reagent Turbfect (Thermo Fisher Scienti c, USA) was mixed with 8 µl of siRNA in 500µl DMEM and then the mixture was added to the cell culture medium. Cells were collected after 48 hours from the transfection to perform transfection validation, RNA and protein extraction. Transfection e ciency was presented as the intensity and the percentage of uorescent cells. TPP1-siRNA cell lines were compared with FAM-marked negative control (siNC group) and blank control (mock group).
RNA extraction and quantitative real-time polymerase chain reaction (PCR) Total cell RNA was extracted using Trizol reagent (Invitrogen, US). The concentration and quality of RNA were evaluated by Nanodrop 2000 Spectrophotometer (Thermo Fisher Scienti c). Reverse transcribed cDNA was synthesized from 0.1 µg-5 µg of total RNA by the RecertAidTM rst strand cDNA Synthesis Kit (Fermentas, Canada) at 42°C for 60 min, followed by 70°C for 5 min and 4°C afterward. Real-time PCR was performed with SYBR Premix Ex TaqTM (Takara, Japan) in a 12.5 µl reaction volume using the CFX ConnectTM Real-Time PCR detection system (Bio-Rad, US). Primers designed for GAPDH, hTERT, and TPP1 were shown in Table S2. All values were normalized to GAPDH and the 2-ΔΔCt method was used to estimate the fold change of gene expression over control samples.

Protein extraction and Western Blot
Cells were digested using 1 ml Trypsin-EDTA Solution (Beyotime, China) and then decomposed by 500 µl RIPA Lysis Buffer per 100 packed cell volume (PCV). Western blot was carried out to evaluate TPP1 and hTERT protein expression. Primary antibodies targeting TPP1 (ab57595, Abcam, USA, dilution of 1:750), hTERT (ab32020, Abcam, USA, dilution of 1:1000), and β-actin (Santa Cruz, US, dilution of 1:2000) were incubated at 4°C overnight. ECL (Advansta, USA) was used to visualize the speci c bands. Auto radiographs were recorded onto X-ray lms (Eastman Kodak Co, USA) and the ImageJ software was applied to analyze the signal intensity of bands.

Gene expression of TPP1 in publicly available databases
Eligible microarray dataset (GEO accession number: GSE7803) including gene expressions of normal cervical tissues, CINs and cervical cancer tissues was used to assess the expression of TPP1 in different cervical lesions. Test of ANOVA was applied. Oncomine gene expression array dataset was queried to analyze TPP1 mRNA expression in cervical cancer and normal tissues. The de ned parameters used to lter datasets were p-value < 1E − 4 (Student's t-test), fold-change > 2, and genes ranking in the top 10%. One dataset of Zhai Cervix met the criteria and was selected to assess expression of TPP1 in cervical cancer tissues and normal tissues.

Statistical analysis
The association of TPP1 expression and clinical-pathological characteristics was assessed using the Chisquare test. Kaplan-Meier curve with a log-rank test was performed to compare overall survival of cancer patients by different expressions of TPP1 or hTERT. To test the prognosis value of TPP1 high expression or TPP1/hTERT high co-expression, multivariate analysis using the Cox proportional hazards model was performed, adjusted for other well-known prognosis factors, i.e., age, tumor diameter, FIGO stage, differentiation, lymphatic invasion, vagina invasion, parauterus invasion, pathological subtype. Hazard ratio (HR) with a 95% con dence interval (CI) was estimated. The mutual relationship of TPP1 and hTERT expression in different cervical lesions was assessed by bivariate correlation analysis (Spearman coe cients). To assess the association between TPP1 high expression and different cervical lesions, odds ratio with 95% CI was estimated using binary logistic regression. Results of in vitro analysis were given as mean ± standard error of the mean (SEM) with a minimum of three repetitions and a student's ttest was conducted. Statistical analyses were performed in SPSS 26 (SPSS Inc., Cary, NC, USA). All statistical analyses were two-sided and a P value of < 0.05 was considered statistically signi cant.

Results
Expression of TPP1 in normal cervix, CINs and cervical cancers Figure 1A  cervical cancer (SCC or AC/ASC patients, P < 0.001) showed a signi cantly higher odds ratio of TPP1 high expression (Fig. 1B).
In concert with high TPP1 protein expression detected in cervical cancer samples, higher expression of the TPP1 gene was validated in external and publicly available databases (GSE7803 and Oncomine). In the dataset of GSE7803, 10 normal cervical tissues, 7 CINs, and 21 cervical cancers were identi ed. When compared with normal tissues, TPP1 was higher expressed in CINs (P < 0.05) and highest expressed in cervical carcinoma (P < 0.001, Fig. 1C). Similarly, in the dataset of Zhai Cervix from Oncomine including 10 normal cervical tissues and 21 cervical cancer cases, TPP1 gene was signi cantly overexpressed in cervical cancers compared with normal cervical tissues in (P < 0.001, Fig. 1D). These results suggested that high TPP1 expression was associated with cervical cancer development and disease progression.
Association of TPP1 expression with clinical-pathological factors and overall survival of cervical cancer patients Table 1     The prognostic value of TPP1 expression in overall survival was further evaluated using Kaplan-Meier analysis and multivariable COX regression. Among 156 cervical cancer patients, 65 (41.7%) of them died during the follow-up. The median overall survival was 108 months and the ve-year survival rate was 68.8%. Compared with TPP1 lowly expressed cervical cancer patients, a signi cantly worse survivorship was found in their counterparts with high expression of TPP1 (P log−rank =0.047, Fig. 2A). Multivariable analysis showed that high expression of TPP1 was marginally associated with a poor prognosis of cervical cancer (HR = 2.03 [95% CI 0.99-4.16], Table 2). Kaplan-Meier analysis showed that higher expression of hTERT trended to be associated with worse but non-signi cant survival (P log−rank =0.060, Fig. S1B). 64.7% (101/156) of cancer patients had high coexpression of both TPP1 and hTERT ( Table 2).
Patients with high coexpression of TPP1/hTERT showed signi cantly worse overall survival (P log−rank =0.005, Fig. 2B), carrying 2.01 (95% CI 1.10-3.67) times higher mortality risk than those without high coexpression (Table 2). These data indicated that TPP1 and hTERT was associated with patient survival in cervical cancers.

Correlation between TPP1 and hTERT expressions in cervical lesions
Association between TPP1 and hTERT expression was further tested in different cervical lesions, both of which were measured in the same batch of samples (n = 274). Similar to TPP1, the percentage of patients with high expression of hTERT increased with CINs (57.7% [30/52] in CIN-I, 59.1% [13/22] in CIN-II and 60.7% [17/28] in CIN-III) and reached 76.3% (106/139) in SCC and 88.2% (15/17) in AC/ASC (Table 3). Spearman correlation analysis revealed positive correlations between TPP1 and hTERT expression both in the overall levels (r = 0.524, P < 0.001) and in participants with different cervical lesions (Table 3).
These ndings demonstrated an inter-regulation between TPP1 and hTERT that might determine cervical cancer development.

TPP1 regulated the expression of hTERT in cervical cancer cells
To further explore the regulatory mechanism of hTERT by TPP1, we transfected TPP1-siRNA into three human cervical cancer cell lines, i.e., Hela, Siha, and C33a cells. Fluorescence microscope, real-time PCR, and western blot were used to con rm transfection e ciency and depletion of TPP1 (Fig. 3A, B and D).
The transfecting e ciencies in Hela, Siha, and C33a cells were 95%, 90%, and 85%, respectively (Fig. 3A). Compared with control cells, TPP1 expression in three cell lines was signi cantly reduced in the TPP1-siRNA transfected cells, both at mRNA (Fig. 3B) and protein levels (Fig. 3D), suggesting a satis ed knockdown effect of TPP1. We found that hTERT protein expression was downregulated in TPP1 depleted cells at both the mRNA (Fig. 3C) and protein levels (Fig. 3E) in all cell lines. Compared with their mock controls, hTERT mRNA expressions in Hela, Siha, and C33a cells reduced by 38.5%, 81.7%, and 51.5%, respectively (Fig. 3C). And the deducted expressions of the hTERT protein were 70.7%, 57.0%, and 85.2%, respectively (Fig. 3E). These results suggested that TPP1 regulated hTERT expression in cervical cancer cells.

Discussion
A substantial increase of cervical cancer was seen in developing countries, especially among young females [20], making it urgent to understand the driving reason for cancer development and to nd new biomarkers for early detection of this disease and potential therapeutic targets. This study indicated that high TPP1 expression was signi cantly related to CIN-III and cervical cancers. High TPP1 expression was a strong risk factor of poor survival for cervical cancer patients, especially when co-highly expressed with hTERT. TPP1 expression positively correlated with the expression of hTERT in cervical cancer and depletion of TPP1 substantially decreased hTERT expression. These ndings indicated that TPP1 might be both a clinical biomarker for early malignant tumors and a prognostic factor in cervical cancer, and it might also be served as a therapeutic target by regulating hTERT expression.
Previous studies found that higher expression of TPP1 was involved in telomere elongation and contributed to the increase of malignant potential in the pathogenesis of B-cell leukemia and colorectal cancer [21][22][23]. Huang et al. also found that TPP1 was overexpressed in gastric cancer compared to adjacent normal tissue, and further con rmed that TPP1 promoted cancer cell proliferation [24]. We found that TPP1 was related to the progress of cervical cancer malignancy, as shown by continuously increasing positive expression rates from CIN-I, CIN-II, CIN-III, to SCC and AC/ASC. Ki-67 encoded by the MKI67 gene is a marker of cellular proliferation closely related to tumor malignancy. We also assessed the correlation between TPP1 and MKI67 gene expressions using TCGA data from GEPIA database in cervical cancer. Interestingly, we found mRNA expression of TPP1 was positively correlated with the expression of MKI67 (Fig. S2), suggesting that TPP1 could be linked to proliferation and malignancy of cervical cancer. In addition, our results suggested that increased TPP1 expression was an early event during epithelial malignant transformation and cervical cancer initiation, and a substantial-high expression rate started in CIN-III. The result was veri ed by a small external analysis when we analyzed the expression of TPP1 in the dataset of GSE7803. Our study suggested that TPP1 might be a potential predictive biomarker for early cervical cancers. TPP1 is responsible for recruiting telomerase to the telomere and maintaining telomerase processivity with the inseparable participation of hTERT [25]. But the interaction of TPP1 and hTERT in cervical cancer was still unclear. Correlative analysis of clinical samples in the current study revealed that TPP1 and hTERT expressions were positively correlated in cervical cancer. TPP1 might regulate hTERT expression as it was crucial for hTERT recruitment to telomeres and telomere elongation [26]. In vitro results veri ed that the expression of hTERT was decreased at both mRNA and protein levels when inhibiting TPP1 expression. Although there lacks direct proof of hTERT expression by TPP1 for the moment, we speculated that the hTERT might be regulated by TPP1 through several pathways. One possible way could be that the TEL patch on the surface of TPP1 directly interacted with the TEN domains of hTERT to recruit telomerase to telomeres [27][28][29]. Knockdown of TPP1 could decrease the telomerase activity by affecting TPP1 and hTERT combination, which further reduced hTERT expression. Additionally, some studies reported that TPP1 bound to POT1 and interacted with POT1 to affect telomerase regulation at chromosome ends [30]. The C-terminal portion of human POT1 (POT1C) was vital for the POT1C-TPP1 interaction and disruption of the interaction was essential to prevent initiation of genome instability permissive for tumorigenesis [12]. HTERT activity could be partially activated by overexpression of TPP1-POT1 via the insertion of ngers domain (IFD) at hTERT in a TPP1-dependent manner [31]. Both of these two pathways were critical for the functions of TPP1-Htert [25]. Moreover, c-Myc was involved in regulating the TPP1-hTERT complex via phosphatidylinositol 3-kinase/Akt, which regulated vascular smooth muscle cell cycle progression and cell proliferation [32]. Besides, other research demonstrated that TPP1 might be also involved in regulation of telomere shortening by interacting with hTERT as an upstream regulator in human cells [33]. It is possible that TPP1 expression is elevated earlier than the activation of hTERT and hTERT activation by the TPP1 signaling pathway might lead to the occurrence or progression of cervical cancer.
It was interesting to note that high TPP1 expression was a strong predictor of poor survival for cervical cancer and patients with high co-expression of TPP1 and hTERT exhibited a much worse survival than those without such high co-expression. TPP1 expression was related to tumor differentiation, local invasion and distant metastasis and its expression in AC/ASC was higher than that in SCC, indicating a positive correlation between TPP1 expression and the aggressiveness of cancer. High expression of TPP1 in tumor cells can protect telomere end structure from fusion and DNA damage, and thus enhance genomic stability and improve cellular immortalization potential [34]. Telomerase of tumor cells might be highly activated when both TPP1 and hTERT were overexpressed, which could further result in higher invasiveness and poorer prognosis. We also did an exploration analysis to the expression difference of TPP1 gene in other cancer types and their corresponding controls using Oncomine database. The results revealed that TPP1 was also overexpressed in several other cancers such as esophageal carcinoma, stomach adenocarcinoma and liver hepatocellular carcinoma, Increased expression of TPP1 was associated with worse survival in these cancers (Fig. S3). Overexpressed TPP1 might promote tumor malignancy in speci c cancers, implying the possibility of TPP1 being a prognosis predictor of these TPP1-highly-expressed cancers.
The strengths of the study include rigorous clinical data collection, a long-term systematic and prospective follow-up, and exploration of potential molecular mechanisms concerning TPP1 and hTERT and extended bioinformatics analysis. In vitro experiments using three different well-known cervical cancer cell lines were well designed and each nding was validated through several methods, ensuring robustness of the results. However, several limitations warrant attention. Despite that we have enrolled all cervical cancer patients during the study period, it is a single hospital-based study with limited study sample size, which could hinder its generalizability. Therefore, further large prospective studies in different populations are warranted to evaluate the clinical application of TPP1. How TPP1 interacts with and regulates hTERT, and by which means TPP1 promotes the malignant phenotype of cervical cancer are under active investigation. Conclusions TPP1 expression increased signi cantly in late precursor lesions (CIN-III) and cervical cancers compared with normal cervical tissue. High expression of TPP1 indicated a worse survival in cervical cancer and cohigh expression with hTERT can better predict a worse survival. Our results also revealed that TPP1 and hTERT expression were positively correlated by both in vivo and in vitro analysis. High expression of TPP1 may be an early event during cervical cancer development and a prognosis factor, making it a potential predictor for early cancer detection and potential target for cervical cancer treatment.

Consent for publication
Not applicable.

Availability of data and materials
Data available from the authors upon reasonable request and with permission of Zhongnan Hospital of Wuhan University in China.

Competing interests
All authors con rm and declare that no con icts of interest exist.