Abstract
Background/Aim: To determine the differential protein profiles of cervical cancer cell lines in order to find potential targets that can be used as biomarkers in low-grade squamous intraepithelial lesions (LSIL) diagnosis. Materials and Methods: Proteomic analysis was performed on cervical cancer cell lines by 2D electrophoresis and liquid chromatography–mass spectrometry. Biomarker validation was performed in histological samples by immunofluorescence. Results: Aldo-keto reductase C1 (AKR1C1) and transketolase-like 1 (TKTL1) proteins were selected as biomarkers and their expression was increased in samples with LSIL diagnosis. TKTL1 in combination with AKR1C1 increased sensitivity and specificity to 75% and 66%, respectively, with an area under curve of 0.76 in receiver operating characteristics curve analysis. Conclusion: AKR1C1 and TKTL1 showed potential as biomarkers for diagnosis of LSIL in Mexican women, with similar sensitivity and specificity to the biomarkers used in clinical trials for diagnosis of LSIL.
Cervical cancer (CC) takes second place in incidence and mortality by cancer in women between 15 and 60 years old (1). CC has a high impact in developing countries; in Africa and Central America it is the leading cause of cancer-related death among women with 85% of cases (2), and takes second place for women in Mexico (3). For all CC stages, the 3- to 5-year survival rate in undeveloped countries is less than 50%, with best results being obtained with radical hysterectomy (4). Screening methods for CC and squamous intraepithelial lesions or cervical intraepithelial neoplasia (CIN) include cytology (conventional and liquid-based) and human papillomavirus (HPV) test, alone or in combination. It is common that colposcopy and histopathology accompany these methods according to a woman's age, screening history, risk factors, and the choice of screening test available (5).
Low-grade squamous intraepithelial lesion (LSIL) has been defined as a cytological diagnosis for patients with smears showing cytological criteria of permissive HPV infection. In histopathological diagnosis LSIL and CIN1 are used synonymously, and the gold standard for the definition of uterine cervical disease is histopathological evaluation (6). However, there is a discrepancy among pathologists about LSIL management due to variable progression: In 70% of cases, lesion regression occurs and only 10% progress to a high-grade squamous intraepithelial lesion (HSIL) (7). The use of cancer-specific biomarkers has been evaluated for diagnosis and treatment with a minimally invasive effect and the potential to lower the cost of diagnosis (8). Ki-67 and p16INK4A are biomarkers proteins that play an important role in clinical diagnosis and differential diagnosis between dysplastic and non-dysplastic lesions (9). The grade of dysplasia is correlated with Ki-67 expression in the progression of LSIL, however, Ki-67 staining cannot distinguish between dysplasia and immature squamous metaplasia. p16INK4A is a useful diagnostic adjunct for SIL, also discriminating it from non-neoplastic lesions but is present at low expression in low-grade lesions, which can generate false positives in diagnosis (10). Biomarkers used for diagnosis of SIL, such as the combination of p16/Ki-67 for detection of HSIL or CIN2/3 have a sensitivity and specificity of 95% and 61% respectively, but in the diagnosis of LSIL or CIN1, no significant difference is reported between these lesions and other inflammatory alterations that are classified as atypical glandular cells of undermined significance (11).
Therefore, biomarkers for LSIL diagnosis with higher sensitivity and specificity are required to enable better diagnosis and management of cervical lesions. Advances in genomic and proteomic technologies such as DNA and tissue microarray, two-dimensional gel electrophoresis, liquid chromatography–mass spectrometry (LC-MS/MS), and protein assays coupled with advanced bioinformatics tools, make it possible to develop biomarkers that can reliably and accurately predict outcomes during cancer management and treatment (12, 13). In addition, biomarkers can be used in cancer diagnosis and therapy, based on their molecular functions in a more precise manner (8, 14). This work aimed to search for unique proteins that are differentially expressed between cervical carcinoma cell lines, which have potential as biomarker candidates for the diagnosis of LSIL in histological tissue samples from Mexican women.
Materials and Methods
Cell lines and culture. CC cell line HeLa (HPV18-positive) was purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA) and was cultured in Dulbecco's modified Eagle's medium (Gibco, Thermo Fisher Scientific, Carsbad, CA, USA) supplemented with 10% fetal bovine serum (Biowest, Nuaillé, France) and 1% penicillin/streptomycin (Corning, Manassas, VA, USA). CC cell lines Rova (HPV18-positive) and Vipa (HPV-negative) cells lines were derived from Mexican women (15) and were a kind gift from Dr. Alberto Monroy-García (México); these were cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin. An epithelial ectocervix HPV-E6E7 transformed (Ect1/E6E7) cell line was purchased from the ATCC and cultured in keratinocyte-serum free medium (Gibco) with 0.1 ng/ml human recombinant epidermal growth factor (Gibco), 0.05 mg/ml bovine pituitary extract and additional calcium chloride (44.1 mg/l). Breast adenocarcinoma MDA-MB-231 cell line used as external control was purchased from the ATCC and cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin. All cell lines were incubated at 37°C in 5% CO2.
Cervical histological samples. A total of 29 formalin-fixed paraffin-embedded cervical tissue samples from Mexican women aged between 25 and 40 years with LSIL diagnosis by histology and HPV diagnosis by polymerase chain reaction test were obtained from the Environmental Cytopathology Laboratory's Sample Bank (Morphology Department, Instituto Politécnico Nacional, Mexico City, Mexico) (Table I).
Two-dimensional (2D) sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Membrane protein extracts from cell lines were obtained using MEM-PER™ Plus Membrane Protein Extraction Kit (Thermo Fisher Scientific, Rockford, IL, USA) and quantified by Bradford assay using a bovine serum albumin (BSA) curve. Using 300 μg, each sample protein was diluted in rehydration buffer DeStreak (GE Healthcare, Piscataway, NJ, USA) and applied to 11 cm immobilized pH gradients (IPG) strips, pH 3-11, using a non linear gradient. IPG strips were run using IPGPhor system (GE Healthcare, Fisher Scientific, Pittsburgh, PA, USA) with the following conditions: 1 h at 500 V, 1.25 h at 800 V (gradient), 0.70 h at 11300 V (gradient) and 2.75 h at 2900 V. The strips were equilibrated with 50 nM Tris, 6M urea, 2% SDS, and 30% glycerol containing 1% dithiothreitol (DTT) for 10 min and the same buffer containing 2.5% iodoacetamide twice. To perform the second dimension analysis, strips were run into 10% SDS-polyacrylamide gels in a Ruby electrophoresis system (GE Healthcare, Piscataway, NJ, USA) at 60 mA with 200 V for 30 min, and 74 mA with 200 V for 5 h. 2D gels were stained with Coomassie blue and images were obtained using an ImageQuant LAS 3000 analyzer (GE Healthcare, Piscataway, NJ, USA).
Diagnosis of the samples used in this study (n=29).
Image analysis. 2D electrophoresis gel images were analyzed using ImageMaster 2D Platinum (v7.0; GE Healthcare, Piscataway, NJ, USA). Protein spots were marked and selected with changes in abundance ratio of >1.5-fold. Unique differentially expressed protein spots were defined as those statistically significantly altered based on >2-fold change and a significant change in expression intensity between cell lines (p<0.05).
Protein identification by nano-LC-MS/MS. Spots of interest were cut out, destained with 50% methanol and 5% acetic acid for 12 h and then were washed with distilled water and incubated with 10 mM trypsin. Spots were reduced in 50 mM dithiothreitol for 45 min followed by alkylation in 30 mM iodoacetamide. Spots were washed three times with 100 mM ammonium bicarbonate and dehydrated with 100% acetonitrile (ACN); after that, samples were dried in a speed vacuum. Spots were digested with 20 ng/μl trypsin for 18 h at 37°C. Digested spots were extracted twice with 50% ACN and 100% ACN. Mass spectroscopy was performed using a Synapt G2S MS instrument (Waters Corporation, Milford, MA, USA) with a time of-flight detector. Samples were run in a nanoACQUITY LC system (Waters Corporation) using a 1.8 μm nanoACQUITY UPLC column (HSST3 C18, 75 μm by 150 mm). The column temperature was 35°C and mobile phase A was 0.1% formic acid in water (LC-MS grade, J.T. Baker, Radnor, PA, USA) and phase B was 0.1% formic acid in ACN (LC-MS grade, J.T. Baker), chromatographic gradient: 0.0 min 7% B, 54.67 min 40% B, 56.33 min 85% B, 59.64 min 85% B and 61.30 min 7% B at a speed of 0.4 nl/min. The ionization chamber temperature was 70°C, with a capillary voltage of 3.00 kV, and a cone voltage of 30 V. Mass interval was between 50 and 2,000 Da. For data processing, Protein Lynx Global SERVER (PLGS) 2.5.1 (Waters Corporation) was used, and for protein identification, UNIPROT-Human database was used (https://www.uniprot.org/). Proteins identified with a confidence interval (CI) higher than 95% were considered as positive identifications.
Bioinformatics analysis. Differentially expressed proteins were identified by MASCOT Peptide Mass Fingerprint and were compared with the Swissprot human proteome database (https://www.uniprot.org). Protein database STRING was used for gaining insight into the molecular mechanism of the proteins through interaction networks, cellular pathways, and functions.
Protein expression analysis by western blot. Total protein (20 μg) of Ect1/E6E7, HeLa, Vipa, Rova, and MDA-MB-231 (non-cervical cells control) were run by SDS-PAGE electrophoresis (90 V for 2 h) and transferred to nitrocellulose membrane (Biorad, Hercules, CA, USA). Membranes were blocked with 5% BSA (Sigma-Aldrich) for 1 h and incubated with primary monoclonal antibodies against annexin A1 (ANXA1), cofilin 1 (CFL1), aldo-keto reductase C1 (AKR1C1), transketolase like 1 (TKTL1) and β-actin (housekeeping) (GeneTex, Irvine, CA, USA) at 1:1,000 dilutions at 4°C overnight. After incubation, blots were washed three times with Tris-buffered saline with 0.05% Tween 80 and incubated with goat anti-mouse horseradish peroxidase-conjugated antibody (Merck, Darmstadt, Germany) at 1:5,000 dilution for 1 h. Blots were revealed with the ECL Prime system (GE Healthcare, Piscataway, NJ, USA) using an ImageQuant LAS 3000 analyzer (GE Healthcare). Quantification and intensity measurement of protein bands were analyzed by ImageJ software (v.1.48; NIH Bethesda, MD, USA) and relative expression was determined based on expression of the housekeeping protein.
Immunofluorescence staining. AKR1C1 and TKTL1 protein expression was evaluated using immunofluorescence staining in cervical biopsies samples obtained from Mexican women with HPV-positive/-negative LSIL, and non-lesion tissues. Tissue samples 3 μm-thick were de-paraffinized and rehydrated with xylol, absolute ethanol, and 90%, 80%, and 70% ethanol, and double-distilled water. Tissues samples were washed with phosphate-buffered saline (PBS) and blocked with 5% BSA in PBS 1X with 0.1% Tween and incubated with primary monoclonal antibodies against AKR1C1 and TKTL1 (GeneTex, Inc. Irvine, CA, USA) diluted 1:200 overnight. After three washes with PBS, tissue samples were incubated for 1 h with fluorescein isothiocyanate-labeled secondary antibody diluted 1:1,000 (Jackson Inmunoresearch, West Grove, PA, USA). Finally, tissues were counterstained with 0.1 μg/ml Hoechst and observed under fluorescence microscopy (Carl-Zeiss, Oberkoche, Germany). Protein expression was evaluated using a semi-quantitative score system to determine fluorescence intensity in the tissue, assigning values from 0 to 4 from the lowest intensity to the highest (16). Negative results (low expression) were considered those with a score between 0 to 2 and positive (high expression) between 3 to 4. The total score of combined biomarkers (TKTL1 with AKR1C1) was also obtained taking as low expression scores between 0 to 3 and as high expression scores from 4 to 6. To determine the sensitivity and specificity of the biomarker, the histopathology test was used as the reference method and the immunofluorescence test with the selected biomarkers was used as the method evaluated.
Statistical analysis. Western blot data were analyzed using GraphPad Prism v.5 (GraphPad Software, Inc., San Diego, CA; USA). Data are presented as the mean±standard deviation. To determine the difference between groups, one-way analysis of variance with a post-hoc Tukey's test was performed. Analysis of biomarkers for immunofluorescence between two clinical groups was analyzed with Fisher's exact test and by plotting the receiver operating characteristic curves (ROC), calculating the area under the ROC curve (AUC) using SigmaPlot v.12.0 (Systat Software Inc., San José, CA, USA). Differences with values of p<0.05 were considered statistically significant.
Results
Differential pattern expression of proteins in the cell lines. The protein profile of each cell line (HeLa, Vipa, Rova, and Ect1/E6E7) was obtained. Protein profile images were acquired in triplicate for each cell line, Figure 1 shows the most representative ones. The total number of spots found for each cell line was: 103 for Ect1/E6E7, 96 for HeLa, 97 for Rova, and 58 for Vipa. Seventeen spots were differentially expressed between cell lines and five proteins that were uniquely expressed in the cell lines: ANXA1 in Ect1/E6E6; AKR1C1 and TKTL1 in Rova; and CFL1 and acidic ribosomal phosphoprotein P1 (RPLP1) in HeLa (Table II).
Interactions of differential proteins. Protein–protein interaction analyses were performed with the STRING database (https://string-db.org) to analyze the five identified proteins. Results showed interactions only between CFL1 and ANXA1. The network of interactions for each differential protein was also obtained (Figure 2). The database showed protein interactions with biological processes associated with inflammatory response (two proteins), chemotaxis (two proteins) and apoptosis (one protein) with ANXA1; actin cytoskeleton organization (three proteins) and regulation of component size (two proteins) with CFL1; cellular hormone metabolic process (four proteins) and oxidation-reduction process with AKR1C1; monosaccharide metabolic process (three proteins), nicotinamide nucleotide metabolic process (one protein) and coenzyme metabolic process (three proteins) with TKTL1; and ribosome biogenesis (three proteins) cytoplasmatic translation (one protein) and translation initiation (one protein) with 60S acidic ribosomal protein P1 (RPLP1). This protein was not considered in the subsequent experiments.
Differential expression of proteins in the cell lines. Western blot analysis of proteins identified by MS as being differentially expressed was performed to determine their expression level in cervical cell lines and MDA-MB-231 cells. The results showed an increase in ANXA1 in the HeLa cell line up to 10-fold compared with Ect1/E6E7 cells, however, this was not consistent with the results obtained by 2D electrophoresis, where it was mostly expressed in the Ect1/E6E7 cell line. CFL1 expression was equivalent in all cell lines, while AKR1C1 showed a significant increase up to two-fold in the Rova cell line and a decrease in Vipa cells by 5-fold compared with Ect1/E6E7 cells. TKTL1 had a significant increase in Rova cell line respect to Ect1/E6E7 cell line, which agrees with the expression of this protein in results obtained by 2D electrophoresis (Figure 3).
Representative images from protein profiles of HeLa, Rova, Vipa and Ect1/E6E7 cell lines by two-dimensional electrophoresis. The image shows the isoelectric point with pH 3-10 and molecular weight expressed. Red circles represent the differential spots and yellow circles the selected differential spots characterized by liquid chromatography–mass spectrometry.
Expression of AK1RC1 and TKTL1 in cervical histological samples. According with the AKR1C1 and TKTL1 expression found by western blot assay in the cell lines, these proteins were sought in LSIL and HPV samples by immunofluorescence. Protein expression of AKR1C1 and TKTL1 was observed in the superficial epithelium in HPV-positive samples and in all epithelial layers in samples with LSIL (Figures 4 and 5). AKR1C1 showed a sensitivity of 66% and specificity of 60% for LSIL, and TKTL1 showed sensitivity and specificity of 65%. However, no significant p-value (p=0.1364 and 0.0959, respectively) was found for these biomarkers in diagnosis. With the combination of TKTL1 and AKR1C1, the sensitivity and specificity for LSIL were 75% and 66%, respectively (p=0.0376) (Table III). In the ROC curves, AKR1C1 gave an AUC higher that for TKTL1, of 0.74 and 0.71 respectively, while the combination of AKR1C1 and TKTL1 for detection of LSIL had an AUC of 0.76 (Figure 6).
Discussion
The use of biomarkers has improved prognosis, progression, recurrence monitoring and allows specific treatments for patients with different pathologies, including cancer (12). The diagnosis of LSIL and HSIL is crucial for appropriate treatment before they progress to invasive carcinoma (17). Therefore, biomarkers have been used to differentiate between LSIL and HSIL, which is of vital importance for better diagnosis and timely treatment (18). This work was focused on differential proteins obtained from HeLa, Rova, and Vipa CC cell lines, and the immortalized non-CC cell line Ect1/E6E7. From the 2D electrophoresis, out of 146 analyzed spots in these cell lines, five differential proteins were selected.
Functional association network of selected differential proteins. Protein–proteins interactions analysis by STRING database (https://string-db.org/) of annexin A1 (ANXA1), cofilin-1 (CFL1), aldo-keto reductase family 1 member C1 (AKR1C1), transketolase like 1 (TKTL1) and 60S acidic ribosomal protein P1 (RPLP1). Each protein is represented as a node with edged interactions. Known and predicted interactions are represented with black edges and line thickness indicates the strength of data support.
Identification of proteins using ProteinLynx Global SERVER™ ver 2.5.1 and Uniprot-Human analyzed by liquid chromatography mass spectrometry for cervical cell lines.
In work carried out by Pappa et al., 2D electrophoresis of total proteins from C33A, SiHa and HeLa CC cells and the non-cervical tumor cell line KCK1T showed 18 proteins that were differentially expressed and associated with signaling pathways involved in the development of cancer including HIPPO, phosphatidylinositol-4,5-bisphosphate 3-kinase/AKT serine/ threonine kinase 1, G2/M cell-cycle checkpoint and EIF2 (19). Proteins differentially expressed in CC cell lines have also been reported to be related to signaling pathways that participate in the regulation of cell proliferation, growth metabolism, cell motility, and apoptosis (20). In our study, the first protein analyzed was ANXA1, expressed in the cervical cell line Ect1/E6E7. ANXA1 plays a role in the innate immunity as a glucocorticoid-mediated response and a regulator of the inflammatory process. It also modulates the acute and chronic immune response, fever, intracellular vesicular trafficking, arachidonic acid release, leukocyte migration, tissue growth, and apoptosis (21). A decrease of the ANXA1 protein level has been observed in lymphoma, and prostate, cervical, esophageal, and oral squamous cell cancer (22). A study in histological samples from patients with SIL and CC showed a decrease in ANXA1 where the lesion degree was more severe (23). Nevertheless, in a more recent study, an increase in ANXA1 expression was observed in HPV16 and HPV18-positive cell lines that was associated with carcinogenesis regulation by HPV (24). This might explain the increased ANXA1 expression observed in the HeLa cell line, which is positive for HPV18. In our results, CFL1 was differentially expressed in the HeLa cell line. CFL1 belongs to the actin-depolymerizing factor (ADF)/cofilin family, involved in normal cell functions such as cytokinesis, endocytosis, apoptosis, and cell migration; their activity is mainly based on the participation in the machinery that promotes cell motility (25). In cancer, this protein has been observed to be important in invasion and metastasis of solid tumors (26). Proteomic analysis revealed the deregulation of proteins associated with the cytoskeleton in HeLa, SiHa, and C33A cell lines; among them, CFL1 had a high expression level in HeLa and C33A (20). CFL1 is also involved in the regulation of stromal cells in women with endometriosis by reducing the invasive capacity of the stromal cells and CFL1 overexpression is associated with invasiveness in stromal cells in patients without endometriosis (27). AKR1C1 and TKTL1 were differentially expressed in the Rova cell line. The AKRs participate in the metabolism of carbohydrates, prostaglandins, aldehydes and ketones through NADPH-dependent pathways; therefore, the expression of these proteins occurs mainly in liver cells. However, in cancer AKRs have been associated with the proliferation and angiogenesis of cells (28). In our analysis, an increase in the AKR1C1 expression was observed in histological samples diagnosed with LSIL and in HPV-positive samples. In work by Torres-Mena et al., AKRs were used as biomarkers for the detection of hepatocellular carcinoma in an animal model and human histological samples, showing an increase in the early stages, which suggests that AKRs might be potential biomarkers for use in diagnosis of early stages of cancer (29). In the endometrium, myometrium and cervix, an increase in mRNA level of AKR1C1 has been observed in the secretory phase and in CC the increment of AKR1C1 protein was observed in 75% of HPV-positive cases (30). The increase of AKR1C1 expression observed in the Rova cell line, which is positive for HPV, is in agreement with these data, while in the HPV-negative Vipa cell line, a significant decrease was observed. It has been shown that HPV-positive cell lines, SiHa and HeLa, exhibit an increase in the transcript and protein levels of AKR1C1 and AKRC3; moreover, site-directed mutagenesis of 16E6 in SiHa cells resulted in a significant loss of AKR1C1 expression (31).
Western blot analysis for differential proteins in cell lines. A: Representative images of western blot of proteins from cervical cell lines (Ect1/E6E7, HeLa, Rova and Vipa) and non-cervical cell line control (MDA-MB-231). B: Densitometry of western blot analysis of the proteins differentially expressed in cell lines employing β-Actin as the loading control. *Significantly different from Ect1/E6E7 at p<0.05 (two-way analysis of variance and post hoc Tukey test).
Immunofluorescence analysis of aldo-keto reductase family 1 member C1 (AKR1C1) in cervical samples. AKR1C1 was expressed in tissue samples with low-grade squamous intraepithelial lesions (LSIL) and human papilloma virus (HPV) positive diagnosis. Original magnification ×20.
Immunofluorescence analysis of transketolase like 1 (TKTL1) in cervical samples. TKTL1 was expressed in tissue samples with low-grade squamous intraepithelial lesions (LSIL) and human papilloma virus (HPV)-positive diagnosis. Original magnification ×20.
Protein expression of aldo-keto reductase family 1 member C1 (AKR1C1) and transketolase-like protein 1 (TKTL1) in cervical histological samples with low-grade squamous intraepithelial lesion (LSIL).
Receiver operating characteristics curves of aldo-keto reductase family 1 member C1 (AKR1C1) and transketolase like 1 (TKTL1) alone (A) and in combination (B) for low-grade squamous intraepithelial lesions (LSIL) diagnosis. AUC: Area under the curve.
Overexpression of AKR1C1 has been associated with proliferation and migration in lung cancer, and drug resistance in bladder cancer cells (32, 33). In cancer diagnosis, AKR1C1 has been used as a biomarker of prognosis associated with lower survival in patients with liver cancer (34). In addition to these results, in our study, we observed an increase in the expression level of AKR1C1 in the breast cancer cell line MDA-MB-231. AKR1C1 participates in progesterone metabolism and since this hormone is associated with breast cancer development, AKR1C1 has also been used as a prognostic marker in early stages of breast cancer (35, 36).
On the other hand, TKTL1 was differentially expressed in the Rova cell line, this protein belongs to the enzymes group that participates in the non-oxidative pentose phosphate pathway. In cancer, the reprogramming of metabolic pathways is essential for increasing cell growth and proliferation (37). In cancerous processes, TKTL1 contributes to an increase of glycolysis and the synthesis of fatty acids that promote cell proliferation. TKTL1 also increases the flux of the pentose phosphate pathway that generates ribose-5-phosphate necessary for fatty acids synthesis, and maintenance of homeostasis that protects cells against death by apoptosis (38). These changes in cell metabolism show that TKTL1 overexpression in patients with cancer is associated with malignancy, invasion, resistance, and poor prognosis (39). In LSIL, TKTL1 is correlated with the degree of progression to invasive carcinoma (40). In the same way as AKR1C1, an increase in the expression of TKTL1 was observed both in LSIL and HPV-positive samples. Recent studies have shown a high expression of this protein in patients diagnosed with atypical squamous cells of undetermined significance and HPV (41), which confirms our results in LSIL histological samples.
Finally, the sensitivity and specificity for diagnosis of LSIL and HPV in histological samples were evaluated. We observed that the combination of AKR1C1 and TKTL1 increased the sensitivity and specificity in the diagnosis of LSIL to 77.78% and 65%, respectively, in a similar way that previously reported for the Ki67 and p16 combination (11). Some biomarkers used for diagnosis of LSIL have shown similar sensitivity, however, the specificity remains low due to the similar response in healthy tissue or non-specific inflammatory changes. Our results with the combination of AKR1C1 and TKTL1 gave an AUC of 0.76 for diagnosis of LSIL, which is similar to that using Ki-67 as a biomarker (AUC of 0.72) (42). In the diagnosis of LSIL, AUC values from 0.70 to 0.80 can be obtained, while in HSIL and CC, AUC values were between 0.80 and 0.95 using biomarkers based on microRNAs and biomarkers obtained from serum; however, these biomarkers have a low rate of detection in early LSIL and HSIL (43, 44). Therefore, the biomarkers proposed in this work can be used for the diagnosis of LSIL, where the specificity of other markers is low. In addition, the sensitivity and specificity of AKR1C1 with TKTL1 in HSIL and invasive CC should be evaluated, since the combination of TKTL1 and cancerous inhibitor of protein phosphatase 2 (CIP-2A) as biomarkers predictive of risk had a sensitivity of 0.91-0.93 and specificity of 0.77-0.83, with high expression in double-positive patients with high-risk HPV-DNA genotype infection and diagnosis of atypical glandular cells of undermined significance (41).
Conclusion
The differential protein profile from cervical cell lines, two of which were obtained from Mexican women, showed that AKR1C1 and TKTL1 have high expression in tissues diagnosed as LSIL, with specificity and sensitivity comparable to biomarkers currently used in the clinic for the diagnosis of HSIL and invasive carcinoma. The clinical application of AKR1C1 and TKTL1 as biomarkers can help to better the diagnosis of LSIL improving follow-up and early treatment. However, the study of these proteins in a greater number of samples with low and high grade lesions, and invasive carcinoma, must be considered for strength the validation of these proposed biomarkers of diagnosis.
Acknowledgements
This project was financed by Consejo Nacional de Ciencia y Tecnología (CONACyT) through project A1-S-21548 and Secretaria de Investigación y Posgrado of Instituto Politécnico Nacional through projects 20160665, 20170863 and 20202051. Our gratitude to Dr. Alberto Monroy-García (UIMEO, México) for providing us with the Mexican Rova and Vipa cell lines, María del Carmen Silva Lucero Ph.D.(Department of Physiology, School of Medicine, UNAM, México) for unconditional support on western blot performance and standardization, to Research and Industry Support Services Unit (USAII, Faculty of Chemistry, UNAM) for it support on proteins characterization, Mr. Fermín Florentino Luna Coronel for support with the processing of histological samples. The Authors were grateful for the laboratory and the equipment support from Insignia-IPN project 2016. A.A.S.J. was awarded CONACyT and Programa Institucional de Formación de Investigadores del IPN (PIFI) scholarships. E.R.G. is Comisión de Operación y Fomento de Actividades Académicas (COFAA) of IPN, Estímulo al Desempeño de los Investigadores (EDI) of IPN, and Sistema Nacional de Investigadores (SNI) grant fellow.
Footnotes
Authors' Contributions
AASJ: Conceptualization, methodology, investigation, formal analysis, writing - original draft preparation. AJ, AEM and MCCA: Methodology, formal analysis, investigation, writing-reviewing and editing. ERG: Conceptualization, methodology, supervision, formal analysis, resources, project administration, funding acquisition and writing- reviewing and editing.
Conflicts of Interest
The Authors have declared no conflicts of interest in this study.
- Received September 17, 2020.
- Revision received October 2, 2020.
- Accepted October 8, 2020.
- Copyright© 2020, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.