Development and Evaluation of p16 based Double Antibody Sandwich ELISA for Detection of Cervical Precancer and Cancer

Objective: Cervical cancer is the third most common cancer in women, worldwide. This study was designed to develop an affordable, accurate and simpler screening test like Enzyme-linked immunosorbent assay (ELISA) which is low cost and will help in bringing down the disease burden in resource poor countries. Methods: In this study, we have raised and evaluated monoclonal antibodies against recombinant p16 using immunohistochemistry (IHC), western blot, immunoprecipitation and ELISA. Double antibody sandwich ELISA (DAS-ELISA) and cytokeratin ELISA was designed for screening women with cervical dysplasia and cancer. Results: Cloned, expressed and purified recombinant p16 were used for generation of monoclonal antibodies. After initial screening, six clones were selected, and affinity purified. Except 155D11G10, which was isotype Immunoglobulin (Ig) G1 all the others were found to be IgG2b. 133A6G5 and 151A7B9 were found to be best for p16 IHC, both showed 70 – 80% and 80 – 90% of nuclear staining respectively. All the antibodies positively detected p16 from the HeLa lysates in western blot except 133A6G5. Studies using immunoprecipitation showed 133A6G5, specifically detected p16. DAS-ELISA developed using a combination of our p16 monoclonal antibodies showed sensitivity of up to 2pg. A pilot study using DAS-ELISA and cytokeratin ELISA in cervical samples revealed the assay sensitivity and specificity as 100% and 80%, respectively. Conclusion: Using combination of DAS-ELISA and cytokeratin ELISA we have developed an accurate and reliable method for the early detection of cervical cancer in a subject, with minimal false results. In the future after large scale validation, p16 ELISA could be used as a reliable tool for diagnostic purposes.


Introduction
Cervical cancer is the third most common cancer in women, worldwide (Ferlay et al., 2021). About 88% of cervical cancer cases occur in developing countries where, in many regions, it is the most common cancer. The higher incidence and mortality in cervical cancer has been mainly associated in countries with lower human development index (Singh et al., 2023). In India, lack of awareness about the symptoms, associated risk factors, availability of screening programmes and knowledge about the preventive measures has contributed to the higher rate of mortality (Kadian et al., 2021). Conventional techniques used for the detection of cervical cancer include pelvic examination, visual inspection with acetic acid / Lugol's iodine and Pap smears. While these techniques can be easily performed even in populations with limited access to healthcare, these cytology based screening has low sensitivity (Malhone and Longatto-Filho, 2019). HPV-DNA is the recommended and widely used marker for cervical cancer detection and it has better sensitivity than PAP smear. However, mere detection of HPV does not mean the subject has cancer or will develop cancer. Expression of p16 is linked to the oncogenic risk of HPV infection resulting in cervical cancer, so p16 over expression is used as a surrogate marker for HPV in cervical cancer detection (Volkova et al., 2021). p16, a tumour suppressor protein inhibits cyclin-dependent kinases (CDK)-4 and -6 and is essential for the regulation of cell cycle. Inactivation of p16 has been detected in various human malignant tumors (Netto and Epstein, 2011;Liu et al., 2015;Venkatesh et al., 2021). In cervical carcinomas, human papillomavirus (HPV) viral DNA integration into the host genome may result in disruption of the E2 open reading frame, resulting in unregulated overexpression of HPV oncogenes E6 and E7. E7-mediates catabolism of Rb resulting in overexpression of p16 (Munger et al., 2013;Pal and Kundu, 2019). Thus, p16 overexpression is a downstream effect of high-risk HPV infection. The present study was taken up based on our preliminary results from cervical cancer microarray data, where p16 has emerged as a potential biomarker of diagnostic importance (Rajkumar et al., 2011). The current study was designed to develop an ELISA-based assay for p16 which would help screen women for cervical dysplasia and cancer in large-scale screening programmes.

Patient sample
This study was performed with approval from institutional review boards at Cancer Institute (WIA), Chennai. All the samples used in the study were collected after obtaining appropriate informed consent from the participants in the study. Samples were collected from patients and volunteers reporting to outpatient department of radiation oncology unit of Cancer Institute (WIA) and screening programme organized by Nargis Dutt Memorial cancer hospital, Barshi, India. Briefly, epithelial cells from ectocervical, endocervical and transformation-zone were collected using Cervex-Brush® and rinsed immediately into the PreservCyt Solution vial by vigorous swirling and stored at 4 O C, until further use.

Cloning, expression and purification
Total RNA was isolated from HeLa lysates using RNeasy mini kit (Qiagen, Hilden, Germany) and cDNA synthesis was done using High Capacity cDNA Reverse Transcription Kits, according to the manufacturer's protocol. p16 primers were designed with restriction enzymes flanking the complete coding sequence. (p16 ORF FP5' GGATCCATGGAGCCGGCGGCG 3' and RP5' GCGGCCGCTCAATCGGGGATGTCTG 3'). p16 gene was amplified from the HeLa cell line cDNA by using AmpliTaq Gold DNA polymerase. 25 μl of PCR reaction mix consisted of 2.5 μl of 10x buffer, 0.2 μl of dNTPs, 1.5μl of MgCl 2 , 10 picomoles of forward and reverse primers, 0.2μl of AmpliTaq Gold DNA polymerase, and 100 ng of cDNA. The cycling conditions included the initial denaturation step of 95°C for 10 mins followed by a cycle of denaturation at 94°C for 1 min, annealing at 80°C for 1 min and extension at 72°C for 2 min for 35 cycles followed by a final extension step at 72°C for 7 mins and finally cooled to 4°C. The PCR product was resolved on ethidium bromide stained 1% agarose gel and documented.
Further, full length p16 was cloned into pCR™2.1-TOPO VECTOR as per TOPO® TA Cloning® Kit protocol. Sub cloning of full length p16 into pET 28a+ expression vector was done using standard techniques. The recombinant pET 28a+ -p16 ligated plasmid was transformed into chemically competent BL21 (DE3) cells (Stratagene, La Jolla, CA). Transformed cells were grown on an LB plate containing kanamycin (50mg/ml) at 37 o C overnight. A single colony was then inoculated into 10 ml of LB broth containing 50mg/ml of kanamycin and grown overnight at 37°C with shaking. The overnight culture was diluted 1:50 and grown for 2 more hours at 37°C to get the optical density of 0.6 (log phase) and then the culture was equilibrated at 25°C. Protein production was induced by adding 0.4 mM isopropyl β-D-1-thiogalactopyranoside (IPTG), and bacteria were cultured for an additional 3 hrs at 25°C, and cells were pelleted. Cell pellet corresponding to 100 ml of induced bacterial culture was resuspended in 10 ml of Lysis buffer (50 mM Tris -HCl, 150 mM NaCl, 0.1% NP40 (pH 8) containing lysozyme and protease inhibitors. To facilitate lysis, the sample was sonicated for 10 min and debris removed by centrifugation at 13000 g for 15 min at 4°C. The supernatant was applied to Ni-NTA affinity column (Invitrogen) and incubated at room temperature with constant rocking for 1 hr, after the incubation period column was settled under gravity. The column was washed sequentially with 5 ml of wash buffer containing 25, 50, and 100 mM imidazole. Protein was then eluted from the column using elution buffer containing 250mM imidazole (50 mM Tris -HCl, 150 mM NaCl, 0.1% NP40 (pH 8). His tagged -p16 fusion protein was cleaved at the thrombin site to remove the His-tag from the full length recombinant p16 using THROMBIN CleanCleave TM KITaccording to the manufacturer's instruction. Cleaved full length p16 was further concentrated using Amicon Ultra centrifugal filters (3 KDa) and confirmed by SDS -PAGE and western blotting.

Generation Of Monoclonal Antibodies
The generation of mouse monoclonal antibodies was outsourced to BioKlone Biotech Pvt Ltd, after providing the recombinant p16 protein. Entire selection and evaluation of individual clones were done in our lab. Culture supernatants of the individual parent clones from the hybridoma were screened for their specificity and sensitivity in the detection of p16 by IHC and ELISA. Clones showing high specificity and sensitivity by IHC and ELISA were further selected and subjected to limiting dilution to identify a single specific clone secreting highly specific monoclonal antibodies. Finally, after screening a total of 231 clones, six clones were selected for further affinity purification and isotyping of the monoclonal antibody.

We s t e r n b l o t , I m m u n o p r e c i p i t a t i o n a n d Immunohistochemistry
Affinity purified monoclonal antibodies were further evaluated for their specificity and sensitivity using western blotting, immunoprecipitation, and IHC for p16. Western blotting using our mouse monoclonal antibody was done as described previously (Pauck et al., 2014). Immunoprecipitation of p16 was done as described previously, with slight modifications (Sawicka et al., 2013). Briefly, 500 µl (1mg/ml) of recombinant p16 was incubated along with 100µg of 133A6G5 at 4oC in a blood rotator. 50µl of protein G-Sepharose were washed twice with RIPA buffer and mixed with an overnight incubated antigen-antibody complex and incubated for 3 hours at 4 o C in a blood rotator. Beads containing immune complexes were washed 5 times and 50µl of sample loading buffer was added to the bead pellet. Finally, the bead pellet was

Cloning, Expression, Purification of Full Length Recombinant P16 Protein
The sequences encoding the p16 were amplified using p16 ORF primers and cloned into TOPO TA vector and further subcloned into pET 28a+ vector using standard techniques. Expression in BL21 (DE3) Cells and purification studies revealed a specific band consistent with His -tagged p16 was in the soluble fraction. His tagged -p16 fusion protein was cleaved at the thrombin site to remove the His-tag from full length recombinant p16 using THROMBIN CleanCleaveTMKIT. A single specific band consistent with full length p16 (16 kDa ) was observed in the lane containing cut p16 when run along with uncut p16 in SDS -PAGE and purified full length p16 was further confirmed by western blotting (Results not shown).

Screening of Clones After Limiting Dilution
Purified full length recombinant p16 was used for immunization and generation of mouse monoclonal antibodies. Culture supernatants of the individual parent clones from the hybridoma were screened for their specificity and sensitivity in detection of p16 by IHC and ELISA. Clones showing high specificity and sensitivity by IHC and ELISA were further selected and subjected to limiting dilution to identify a single specific clone secreting highly specific monoclonal antibodies.

Evaluation of Mouse Monoclonals Raised Against P16
Finally, after screening a total of 231 clones, six clones were selected for further affinity purification. Isotyping of selected antibodies revealed that except 155D11G10, all of them were of IgG2b. 155D11G10 was found to be isotype heat-denatured at 95 o C and the supernatant was spun down and loaded in SDS-PAGE followed by western blotting.
Immunohistochemistry for p16 was done as described previously (Vijayalakshmi et al., 2007), briefly, by using wet autoclaving with a hold time of 5 minutes. Mouse monoclonal E6H4 antibody (MTM, Laboratories) was used at 1 in 200 dilutions as a positive control, and affinity purified monoclonal antibodies were evaluated.

Double Antibody Sandwich -ELISA For P16
Sandwich ELISA for p16 was set up by coating microtiter plates in 0.1 M carbonate buffer with 50µl combination of p16 monoclonal antibody 133A6G5 and 151A7B9 (250ng) overnight at 4 o C. Antibody coated plates were then blocked with 2% milk for 1 hour at 37 o c followed by three washes with PBS-Tween20 (0.05%). Recombinant p16 protein standards (250pg -0.12pg) in 2 % BSA were then added and incubated for 2 hours at 37 o c. Following wash thrice with PBS-Tween20, 50 µl detecting antibody (biotinylated p16 monoclonal antibody 155E11G3 and 155D11G10 (1:5,000) was added to each well and incubated for 30 minutes at 37 o c. Finally, Streptavidin-HRP conjugate was added in 1:500 dilutions to each well and incubated for 30 minutes. The plates were then developed after final wash, using OPD substrate buffer and reaction stopped with 2N H 2 So 4 . The absorbance of each well at 492 nm was recorded using ELISA plate reader.

ELISA for Cytokeratin
Direct ELISA for Cytokeratin was set up briefly by coating 50µl of HeLa lysates from different cell numbers (standards) overnight at 4 o C in 0.1 M carbonate buffer. Antigen coated plates were then blocked with 2% milk for 1 hour at 37 o c followed by three washes with PBS-Tween20 (0.05%). 50µl (1:150) of mouse Monoclonal (AE1/AE3) antibody (M3515, Dako) in 2 % BSA were added and incubated for 2 hours at 37 o C. Following wash thrice with PBS-Tween20, 50 µl (1:5,000) anti-mouse IgG, Biotinylated (RPN1001, Amersham™) was added and incubated for 30 minutes at 37 o c. Finally, Streptavidin -HRP conjugate (RPN1231, Amersham™) was added in 1:500 dilutions to each well and incubated for 30 minutes. The plates were then developed after the final wash, using OPD substrate buffer and reaction stopped with 2N H 2 SO 4 . The absorbance of each well at 492 nm was recorded using ELISA plate reader.

Pilot study
A pilot study to establish a proof of principle was designed using DAS-ELISA and Cytokeratin ELISA. Fifty-five cervical scrape samples with known histological / cytological diagnosis were lysed using sonication. The lysates were cleared by centrifugation at 14000rpm for 15 min at 4 o C. 50µl of the supernatant from each sample was used for the ELISA. Samples were run along with p16 standards and HeLa lysates from different cell numbers (standards). The ratio of p16 OD to Cytokeratin OD were used to set the cut-off value based on which the samples were scored positive or negative.   Figure 2. (a), Double antibody -Sandwich ELISA (DAS -ELISA) developed using combination of our p16 antibodies showed greater detecting limits. Sensitivity of the assay was recorded up to 2pg. (b), Direct ELISA for cytokeratin and DAS -ELISA for p16 using HeLa lysates showed detecting limit from lysates up to 5X10 2 cells.
IgG1 (Table 1). In immunohistochemistry, the staining pattern differed among the selected p16 antibodies, E6H4 antibody was used as a positive control.
Except 133A6G5, all the antibodies positively detected p16 from the HeLa lysates in western blotting (Figure 1.a) and immunoprecipitation using 100 µg of 133A6G5 confirmed a specific p16 band being detected at 16 kDa (Figure 1 b).
Double antibody Sandwich ELISA using combination of 133A6G5 and 151A7B9 as capture antibodies and biotinylated 155E11G3 and 155D11G10 showed better sensitivity in picking up the p16 even at lower concentrations. The detecting limits of this ELISA format was recorded up to 0.1 O.D which was equivalent to 1.95pg of p16 when compared with the standards (Figure 2a).

Evaluation Of Cytokeratin ELISA
Parallel to p16 ELISA, we have also evaluated cytokeratin ELISA using AE1/AE3 antibody and HeLa lysates as standards. Lysates were prepared from serially diluted HeLa cells ranging from 5 X10 5 to 0. 5 X10 1 . Direct   ELISA for cytokeratin using AE1/AE3 antibody showed that signals from lysates as low as 500 cells (HeLacells) were detected using this ELISA format (Figure 2b).

Proof of Principle
Towards establishing a proof of principle in our pilot study we have used fifty-five cervical smear samples with known histological / cytological diagnosis. The results were analyzed by comparing values of p16 OD and cytokeratin OD and the samples scored positive, if the p16 O.D was 0.2 or more or if the ratio of p16 OD to cytokeratin OD is greater than 1.35. If the ratio of p16 OD to cytokeratin OD was less than 1.35 it was scored as negative for precancerous change or cervical cancer and if values of both p16 OD and cytokeratin OD were less than 0.1 it indicates the inadequacy of the sample and demands another fresh sample (Table2). An application for patent for the assay had been submitted (Patent Application number -475/CHE/2014, Status -Application Awaiting Examination)

Discussion
Cervical cancer is the second most common cancer among Indian women and the mortality due to cervical cancer is the highest as compared to any other countries (Arbyn et al., 2020). Our earlier studies using microarray have revealed several novel genes to be differentially expressed in cervical cancer (Rajkumar et al., 2011). p16 was reported to be one of the potential biomarkers of diagnostic importance. Although several groups have reported p16 as a potential diagnostic marker and as a surrogate marker for detection of high risk HPV in pre -neoplastic lesions, the interpretation of p16 immunohistochemical results remains a matter of debate, due to the discrepancies in scoring p16 positive cells among different authors (Volgareva et al., 2004;Yoshida et al., 2004;Bose et al., 2005;Holladay et al., 2006). To overcome such discrepancies combined scoring method has been proposed by Thai group, which showed good reproducibility and correlation when compared with histology (Vinyuvat et al., 2008). Several researchers have reported that HPV testing with p16-INK4A triage produces a significant increase in sensitivity compared with conventional cytology, with no substantial increase in referral to colposcopy (Carozzi et al., 2008;Tsoumpou et al., 2011;Roelens et al., 2012).
Although p16 as potential biomarker has been well documented for over a decade, its use in routine diagnostics and in cervical cancer screening program is very limited. Some of the reasons include unsolved issues in implementing LAST (lower anogenital squamous terminology) guidelines and the need for judicious interpretation of staining patterns (Reuschenbach et al., 2014;Yang et al., 2018). To this end, the current study was taken up with the objective to develop a sandwich ELISA for detection of p16. We have done the cloning, expression and purification of recombinant p16. Our sequencing analysis from the positive colonies after cloning showed 100 % sequence similarity to Homo sapiens cyclin-dependent kinase inhibitor 2A, transcript variant 1. Analysis of soluble and insoluble fractions after expression and purification from induced culture revealed highly expressed His tagged -p16 fusion protein. Ni-NTA column purification of the soluble fraction showed single specific band of His tagged -p16 fusion protein which was then confirmed with western blot using anti-p16 antibody. Further, in our quest to generate specific monoclonal antibodies against p16, we tried to purify the full length p16 without the His-tag. To this end, we have exploited the presence of thrombin sites in the His-tagged p16 fusion protein. Thrombin site present after the 6 x histidine tag and before the full length p16 was cleaved and the full length p16 was used for generation of mouse monoclonal antibodies. Initial screening and selection of individual parent clones from the hybridoma were done by IHC and ELISA. Finally, six clones were selected based on their specificity and sensitivity in IHC and ELISA. Isotyping of selected antibodies revealed that except 155D11G10, all of them were of IgG2b. 155D11G10 was found to be isotype IgG1.
We then validated all the mouse monoclonal antibodies generated using our recombinant p16 by Western blotting, IHC and ELISA. Individual affinity purified antibodies were evaluated by its ability to detect p16 from HeLa lysates. Except 133A6G5, all the antibodies detected p16 at 1:200 dilutions by Western blotting. Additionally immunoprecipitation was done using 133A6G5, which confirmed the specificity of this antibody for p16 in HeLa lysates. Interestingly, 133A6G5 which could not detect p16 in western blot at 1:200 dilutions and 151A7B9 which showed a faint band were found to best in p16 IHC. Both 133A6G5 and 151A7B9 showed 70 -80 % and 80 -90 % of nuclear staining, respectively, with similar intensities (+++). This result suggests that different epitope is being recognized by these antibodies. We have also evaluated individual monoclonal antibody as either capture or detecting antibody in sandwich ELISA format. 133A6G5, 155E11G3, 151A7B9 and 155D11G10 showed better sensitivity when used as capture antibody. Average detecting limit of these antibodies were up to 15.62pg.133A6G5 and 155D11G10 monoclonal antibodies were found to better as detecting antibody with their average detecting limit recorded up to 31.25 pg.
Previous studies have shown that a Double-Antibody Sandwich ELISA (DAS -ELISA), using a combination of a monoclonal and a polyclonal antibody as the capture and detecting antibodies respectively, resulted in improved specificity compared to normal ELISA or to a sandwich ELISA using polyclonal as the capture and monoclonal antibody as the detecting antibody (Kolbe and Kubicek, 1990;Riske et al., 1990). Based on the results from validation of our antibodies, we have designed a Double antibody Sandwich ELISA (DAS -ELISA) with 133A6G5, 151A7B9 as capture antibodies and biotinylated 155E11G3 and 155D11G10 as detecting antibodies. Our ELISA using this combination of p16 mouse monoclonal antibodies showed greater detecting limits with sensitivity of up to 2pg.
As of now, only one group has come out with an ELISA Kit for the measurement of solubilized p16 from the lysed cervical samples. They have reported a positive correlation between histologically confirmed high-grade cervical intraepithelial neoplasia (CIN) and level of solubilized p16 protein detected by ELISA (Mao et al., 2007). Evaluation of their prototypic ELISA reported an independent relationship between p16 ELISA level and lesion size and the possible lower sensitivity of the p16 ELISA test in detecting small lesions or lesions that shed few abnormal cells (Balasubramanian et al., 2009). One of the major concerns in the ELISA assay is determining sample adequacy, which can be easily detected in a PAP smear. However, once cells are lysed for the assay and estimating p16 levels alone runs the risk of false negatives due to inadequate number of cells. To overcome this, we have introduced a second Cytokeratin ELISA. If the values of both p16 OD and Cytokeratin OD is less than 0.1 it indicates the inadequacy of the sample and demands another fresh sample.
Although Pap smears are used as the most cost-effective cancer screening test developed to date (Greenberg et al., 1995;Denny et al., 2006), management of ambiguous or low-grade cytological results (ASCUS and LSIL) is very controversial, mainly because of the nature of this morphology-based test, which inevitably leads to interobserver variability and some Pap test discordance with histological follow-up. Meta-analysis of Pap test accuracy has revealed that the mean sensitivity of primary Pap tests is ~58%, and the accuracy of a repeat test is only ~66% (Fahey et al., 1995).
HPV-DNA testing detects high-risk strains of HPV (hr-HPV) which cause almost all cervical cancers. Unlike tests that rely on visual inspection, HPV-DNA testing is an objective diagnostic, leaving no space for misinterpretation of results. However, its positive predictive value of pre-cancerous lesions is less than 10 percent especially in CIN2+ stages (Guzel et al., 2021).
Although multiple molecular markers has been studied including, viral markers (E6, E7), cellular markers (Ki67/ MIB-1, MYC, Survivin), nucleic acid changes (DNA methylation, miRNAs) their usage in population based screening are limited.Many studies have evaluated p16 as a surrogate marker for HR -HPV infection using IHC (Duncan et al., 2013;Gronhoj Larsen et al., 2014). p16 / Ki-67 dual immunostaining has been reported as better approach for cervical cancer screening compared to conventional methods (Eyituoyo Okoturo, 2023). Some of the drawbacks in using p16 IHC are (i) need for trained cytologists / pathologists. (ii) Scoring is subjective and discordant interpretation among pathologists. To overcome these drawbacks, we have developed an ELISA based assay for detection of p16 and cytokeratin in cervical smear lysates. Such a simple calorimetric based ELISA assays can be highly sensitive and specific, does not require trained pathologist / cytologists and can be cost effective with a possibility of automation.
In summary, we have cloned, expressed and purified full length p16 for generation of monoclonal antibodies. We have also validated individual antibodies for their sensitivity and specificity. We have developed a more sensitive Double-Antibody Sandwich ELISA using our p16 mouse monoclonal antibodies and have shown that the sample inadequacy issue can be addressed by introducing an additional Cytokeratin ELISA. We have also established a proof of principle in our pilot study which showed the sensitivity and specificity of our ELISA to be 100% and 80%, respectively. In future studies, if we could demonstrate similar sensitivity with accuracy and reliability in large number patient samples, the p16 DAS-ELISA could be effectively used as a reliable tool for diagnostic purpose in large scale screening programmes.

Author Contribution Statement
MB carried out all the experiments and was involved in data analysis and drafting of the manuscript. S.S.S did the histopathological and Immunohistochemical assessment. GS was involved in the clinical management of cervical cancer patients. A.C and S.J.H screened and recruited patients for the study. TR designed the study; analyzed & interpreted the data, drafted and revised the manuscript..