Implication of microRNA-regulated PTEN expression in the clinico-pathology & survival outcomes in advanced ovarian cancer

Study participant selection and sample collection

Newly diagnosed individuals with OC (n=104) who were treated from January, 2018 to December, 2020 were enrolled and clinicopathological characteristics were recorded. The inclusion criteria for selecting the participants were ≥ 20 yr who were diagnosed with epithelial OC and did not receive treatment; those with I-IV staging as per International Federation of Gynecology and Obstetrics (FIGO); predicted survival of patients must not be lower than one month; and participants willing to take part in the treatment process with follow up. The exclusion criteria was (i) history of other tumours or those who underwent complete oophorectomy and received any treatments like radiotherapy, chemotherapy previously, prenatal or nursing women; (ii) individuals with severe illness associated with hepatitis, active infections, or any other serious conditions of lungs, kidney, and heart and (iii) those with unrestrained substantial autoimmune ailment. A tissue sample (500 mg-1g) was obtained from the study participants during primary or interval debulking surgery, and normal ovary tissues were taken via resection as a control. Samples were collected in RNA Later (HiMedia, India) and 10 per cent neutral buffered formalin and preserved at -80°C for the study of mRNA, miRNAs, and protein expression (Supplementary Fig. 1).

Follow up and clinical response

The expression level of PTEN was considered as the primary predictor variable, whereas other characteristics (age, tumour stage, grade, size, treatment module, response, and miR expression) were taken as secondary predictor variables. The major outcome was overall survival (OS) and follow up was carried out telephonically until death or last follow up upto June 2023. As per our previous study,²² the clinical efficacy of OC patients was classified as Complete Responders (CRs), Partial-Responders (PRs), and Non-Responders (NRs).

Immunohistochemical expression of PTEN

An immunohistochemical (IHC) assay was performed for the expression of PTEN protein in the formalin-fixed and paraffin-embedded tissue specimens of the OC patients. This was done using the corresponding rabbit polyclonal antibody (AbCam, UK) with a dilution of 1:500 and secondary goat against rabbit IgG antibody (AbCam, UK) following the pre-treatment in citrate buffer in microwave (pH-6.0). The staining was visualised with 3,3’-diaminobenzidine tetrahydrochloride (DAB) followed by counterstaining with haematoxylin. The degree of immunostaining of PTEN cells was examined for the scoring which varied from 0-100 per cent, and a three-grade scoring system was used for evaluation of PTEN expression as (1) negative (0-10% stained cells); (2) + (11-50%); and (3) ++ (>50%)²³.

Western blot

Tissues were homogenised for 10 min in radio-immunoprecipitation assay (RIPA) lysis buffer. Cellular protein was obtained from cell pellets using the whole cell extract method in a protein lysis buffer that contains protease and phosphatase inhibitors. Protein concentration was checked and 50-100 μg proteins were separated on 10 per cent SDS-PAGE, and the gel was subsequently shifted onto a polyvinylidene fluoride (PVDF) membrane (Merck Life Science, USA). Then, the PDVF filter was immersed in the five per cent non-fatty milk in TBS-Tween (TBST) (pH 7.4) and incubated for 1-3 h at room temperature with antibodies against PTEN. Specific proteins were detected with primary antibody of Rabbit anti-PTEN (AbCam, UK),with 1:1000 dilution, incubated in a shaker for 1-3 h at room temperature or overnight at 4˚C, and then washed with TBST thrice. The secondary goat anti-rabbit antibody was used at a 1:5000 dilution (AbCam, UK), followed by shaking at room temperature for 1-3 h. It was then washed with TBST thrice. Finally, visualisation of the protein band was done through chemiluminescence (ECL) reagent (sc-2048; SANTA CRUZ Biotechnology, USA) and imaged with the ChemiDoc gel documentation system (Biorad, USA).

mRNA and miRNA expression detection

Tissues were homogenised for 10 min, and TRIzol (Invitrogen, Thermo Fisher Scientific, Inc., USA) was added for the extraction of total RNA. Nanodrop (Genetex, Korea) was used to determine the purity and concentration of the isolated RNA in its entirety. Using EasyScriptTM (applied biological materials, Canada), reverse transcription of cDNA was done using the isolated RNA based on the manufacturer’s instructions. Applied Biosystems 7500 Real-Time PCR equipment (Applied Biosystems; Thermo Fisher Scientific, Inc., USA) was used to assess the degree of mRNA (PTEN) and miRNAs (miRNA-214, miRNA-433, miRNA-100, and miRNA-152) using Low ROX-Evagreen 2xqPCR (Applied Biological Materials, Canada) from the synthesised cDNA. The sample comprised 10 μl of the SYBR green mix, 0.6 μl of ROX, ≤500 ng/reaction cDNA sample, and 0.6 μl of primer (forward and reverse) (Supplementary Table). Then, nuclease-free water was added to bring the volume to 20 μl. The reaction was conducted at 95˚C for 10 min, then underwent 40 cycles for 10 sec at 95˚C, 30 sec at 60˚C, and 1 min at 72˚C. All tests were performed in triplicate, normalising PTEN mRNA with 18S, and miRNAs with GAPDH, and the results were analysed by calculating the relative fold change (2^-ΔΔCt).

Statistical analysis

SPSS software (Version 26.0; IBM Corp., TX, USA) was used for the statistical analysis. Paired t-test analysis was done to check the differential expression of selected biomarkers in normal and OC tissue. Chi-square (Crosstabs) test was used to find the relationship between PTEN expression and clinicopathological parameters, and Pearson correlation coefficient was used to predict the association of miRs and PTEN. The survival was estimated using the Kaplan-Meier survival curve after monitoring the patients for three years, and the log-rank test was used for the comparison. It was also used to analyse survival associated with PTEN and miRNA expression. Cox regression was done to check the hazard ratio of death associated with various parameters. Receiver Operating Characteristic (ROC) analysis was performed to check the diagnostic value of PTEN and miRNAs. The criteria for statistical significance were P<0.05.

Clinicopathological features study

A total of 104 newly diagnosed OC patients were recruited in this study (Table I). The mean age was 49.29±9.68 years, ranging between 20-80 yr, and the majority were within the age group of 41-60 (67.8%). The larger portion of the participants were presented with advanced stages (74.5% and 21.1% FIGO stage III and stage IV, respectively), and a preponderance of them had serous carcinoma (70.5%) with high grade (48.5%) followed by mucinous (15.0%), clear cell (12.6%), and endometroid (1.9%) EOC at the time of diagnosis. 49.6 per cent had primary debulking followed by adjuvant chemotherapy (ACT) and 50.4 per cent of participants underwent interval debulking surgery following neo-adjuvant chemotherapy (NACT). Response Evaluation Criteria in Solid Tumors (RESIST) and Gynecological Cancer Intergroup (GCIG) criteria-based evaluation categorised 34.6, 37.5, and 27.9 per cent of participants as CR, PR, and NRs, respectively. In addition, within the follow up period of 36 months, 39.9 per cent of the participants succumbed to the disease.

Association of PTEN expression with tumour progression and clinicopathologic parameters

The PTEN protein was found to be positive for all normal control tissues after IHC analysis in comparison to the malignant tissues (Fig. 1), in which only 15 per cent of cells showed nuclear expression. Whereas, in the case of OC, 11.15 per cent of cells had nuclear staining. Further stratification for OC based on staining intensity depicted 91.56 per cent, 87.66 per cent, and 86.25 per cent of cells had cytoplasmic staining with notably lower nuclear stain (Supplementary Fig. 2). Similarly, significant dysregulation in PTEN mRNA expression was found in the cancerous versus normal controls (P=<0.001) (Fig. 2A). Deficient PTEN expression was observed in 53 of 104 tumours (50.96%). The Western blots showed dysregulated expression of PTEN in cancer tissue with differential treatment (ACT/NACT) and normal tissue (Fig. 2B and Supplementary Fig. 3).

Close

Fig. 1.

Representative IHC staining of ovarian samples. (A) PTEN was positively expressed in normal tissue; (B) Negative expression for PTEN; (C-F) low-moderate PTEN expression was observed in some OC tissue samples.

Export to PPT

Close

Fig. 2.

Quantitative Realtime PCR and Western Blot analysis of PTEN messenger RNA (mRNA) and protein in individuals with ovarian cancer (OC). (A) The level of PTEN mRNA is significantly lower in cancer tissue than in normal tissue. (B) The level of PTEN protein (Normal: Lane 1; ACT: Lane 2, 3; NACT: Lane 4, 5) is dysregulated in the tissue of OC patients with different treatment responses compared with normal tissue. The P^**< 0.001 was considered significant. ACT, adjuvant chemotherapy; NACT, neoadjuvant chemotherapy; NRs, non-responders; Rs, responders.

Export to PPT

The chi-square test using crosstab analysis showed a weaker expression among individuals with stage IV (72.7%). Furthermore, in high-grade serous ovarian carcinoma (HGSOC) there were 10 per cent and 20.9 per cent to flow-grade serous ovarian carcinoma (LGSOC) samples had high PTEN expression, and moderate expression (52.9%) of PTEN was detected in well-differentiated cancers, whereas minimal or negative expression (75% and 61.5%) was observed in moderate to poorly differentiated cancers, which suggested that PTEN expression may be regulated by the tumour microenvironment. Furthermore, the expression of PTEN protein did not have any significant association with age and tumour histology (P≥0.05) but was significantly correlated with clinical parameters such as stage (P=0.002), grade (P=0.041), mode of surgery (P≤0.001) of the tumour, and treatment outcome (P≤0.001) (Table I).

Correlation of PTEN and miRNAs dysregulation in OC

Paired sample t-test revealed that the levels of miR-214 and miR-433 were significantly elevated but miR-100 and miR-152 were suppressed in tumour tissues in comparison to normal tissues (P≤0.001) (Fig. 3). Nevertheless, the expression of miR-214, miR-433, and miR-152 suggestively varied between ACT and NACT cases (P≤0.05), though it was not relevant statistically for miR-100 (P≥0.05). Furthermore, the correlation coefficient test showed a strong positive association between PTEN mRNA and protein in OC (r=0.823, P≤0.05) (Fig. 4A). Additionally, both PTEN mRNA and protein expressions are negatively correlated with miR-214 (r= -0.453, P=<0.001; r=-0.542, P≤0.001) and miR-433 (r= -0.413, P≤0.001; r=-0.459, P≤0.001) level; however, with the miR-100 (r= 0.314, P≤0.005; r= 0.431, P≤0.001) and miR-152 (r= 0.344, P≤0.001; r= 0.354, P≤0.001) exhibit positive association with PTEN mRNA and protein (Fig. 4B-C).

Close

Fig. 3.

Quantitative real time PCR analysis of miR-214, miR-433, miR-100, and miR-152 in individuals with OC. The level of miR-214 & miR-433 is significantly dysregulated in the tissue of OC individuals compared with normal tissue and with different treatments (ACT/NACT). The P^*< 0.05; ^**< 0.001 was considered significant.

Export to PPT

Close

Fig. 4.

Scatter plots for correlation analysis between PTEN protein, mRNA, and microRNA levels in ovarian carcinoma tissue samples. (A) between PTEN protein and mRNA; (B) between PTEN mRNA and miRNAs; (C) between PTEN protein and miRNAs. PTEN protein expression was divided based on immunohistochemistry score (1-negative; 2-+; 3-++).

Export to PPT

Overall survival and its association with PTEN and miRNAs dysregulation

The survival probabilities over time were estimated using the Kaplan-Meier estimator, and the log-rank test was employed to compare survival distributions among various patient groups. This test indicated a significant decrease in survival associated with age (15.71 months), tumour stage (15.6 months), grade (10.5 months), treatment response (18.54 months), PTEN expression (20.71 months), miR-214 (18.86 months), miR-433 (18.92 months), miR-100 (18.47 months),and miR-152 (19.60 months) expression (P≤0.05). Median survival times and corresponding 95 per cent confidence intervals (CIs) are reported in table II. Additionally, Cox proportional hazards regression was used to identify potential prognostic factors linked to OS. The Cox model was used for estimating hazard ratios (HRs) and 95% confidence interval (CI)s, adjusting for relevant covariates such as age, treatment type, pathological parameters, and biomarker expression levels. It showed age, tumour stage, grade, treatment, clinical response, PTEN level, and expression of only miR-214, and miR-152 are risk factors (HR>1) for predicting survival among OC patients; however, it was not found to be statistical significance with age (P=0.001) and treatment outcome (P=0.02) (Table II). Furthermore, a concurrent comparative analysis of PTEN and miRNAs presented that cumulative dysregulation in their expression levels had a considerable impact on the OS of the participants (Table III). Low PTEN levels with high expression of miR-214 (18) and miR-433 (12) were found to have poor survival (P<0.05), while lower levels of PTEN alliance with declined miR-100 (16) and miR-152 (16), exhibiting worse survival (P<0.05) (Table III). This suggests that the OS improved with the increased PTEN level.

The ROC analysis revealed a sensitivity of PTEN 91.11 per cent, whereas the specificity was 62.5 per cent. In addition, miRNAs showed sensitivity ranging between 91-97 per cent and specificity extending from 77-95 per cent. The area under the curve (AUC) of PTEN protein, miR-214, miR-433, miR-100, and miR-152 was 0.774, 0.983, 0.989, 0.912, and 0.995, respectively, with significant prognostic values (P≤0.001) to find efficacy between response groups (Table IV).