1. Introduction
According to the World Health Organization (WHO), almost all cervical cancer cases (99%) are linked to infection with high-risk human papillomavirus (HPV), an extremely common virus transmitted through sexual contact [
1]. Cervical cancer is the fourth most common cancer in women globally with approximately 660,000 new cases and 350,000 deaths in 2022. The highest rates of cervical cancer incidence and mortality are in low- and middle-income countries, reflecting major inequities driven by the lack of access to national HPV vaccination, cervical screening and treatment services, and social and economic determinants [
2]. HPV DNA can be detected using methods such as hybrid capture and polymerase chain reaction (PCR) [
3]. The WHO endorses HPV-DNA testing as the primary screening technique for cervical cancer prevention [
4]. Emerging research, however, suggests that the oncogenic activity of HPV is a more critical clinical marker of developing cervical lesions and cervical cancer than the mere presence of HPV DNA [
5,
6]. For example, Cuschieri et al. found that detection of HPV E6/E7 transcripts is more specific for disease diagnosis at follow-up than detection of HPV DNA, indicating that positivity for HPV mRNA transcripts at baseline significantly increases the likelihood of persistent infection over DNA detection alone [
7]. Similarly, Molden and associates argue that the prognostic value of DNA testing for identifying severe dysplasia is limited, whereas HPV E6/E7 mRNA detection in combination with cytology is a more reliable prognostic indicator [
8,
9,
10]. Additional studies highlight the occurrence of HPV DNA in cervical tissues without signs of active infection, pointing to a possible latent infection under immunologic control [
11]. These insights support a nuanced screening approach that integrates DNA and mRNA testing to enhance our understanding of cervical cancer progression. Incorporating protein-level detection could significantly augment the efficacy of such a comprehensive screening strategy.
The noninvasive screening using urine instead of invasive tests, such as Pap testing, for infectious HPV will likely increase the number of people screened for HPV. Women report being more comfortable with providing urine specimens [
12,
13], and more attention is being focused on screening urine from cancer patients for HPV before and after treatment [
14]. Therefore, it may be useful to establish a protein-level test to accurately measure the oncogenic activity of cervical cancer based on the quantification of HPV16 E7 oncoproteins in urine collected from women with cervical intraepithelial neoplasia (CIN; pathologic diagnosis of cervical biopsy specimens); for example, using an ultrasensitive enzyme-linked immunosorbent assay (ELISA) with thionicotinamide-adenine dinucleotide (thio-NAD) cycling (TN-cyclon
TM) [
15]. In the present study, we evaluated the validity of using urine in our proposed protein-level test by comparing the ELISA index (threshold of the presence of E7 oncoproteins), HPV typing with uniplex PCR (presence of E7 DNA), and the CIN grade.
We focused on HPV16 E7 oncoprotein because the pathogenesis of cervical cancer and oropharyngeal cancer is causatively associated with HPV infections, especially with HPV16 (50% of cases), HPV16-related types (types 31, 33, 35, 52, 58, and 67), and HPV18 and its related types (types 18, 39, 45, 59, 68, 70, 85, and 97) [
1,
16]. HPV16 and HPV18-related types are considered to be the highest-risk types, and they are thought to be the most malignant types as the causative agents of the development of cervical cancer. Therefore, we have also focused on the related types, not only HPV16 and HPV18. The E7 oncoprotein in these viruses prevents phosphorylation of the retinoblastoma protein (pRB), leading to uncontrolled cell proliferation [
17], and is involved in cell transformation, mitosis, and cervical cancer cell immortalization [
18,
19,
20]. The presence of E7 is sufficient to immortalize epithelial cells [
21], and E7 has a major role in cervical cancer development in transgenic mouse models [
22]. The importance of the E7 oncoprotein in the development of HPV-related malignancy demands its careful quantification in human specimens [
23].
4. Discussion
The ability to precisely quantify the amount of E7 oncoproteins in specimens is critical for gaining insight into the development and progression of cervical cancer. The measurement results using our ultrasensitive ELISA system provide accurate information regarding oncogenic activity in cervical cancer, indicating the potential suitability of this system as the primary screening method, as well as mRNA testing, for a noninvasive diagnosis using urine samples. The present study demonstrated how E7 oncoproteins in high-risk HPV16 and its related types can be quantified using an ultrasensitive ELISA with thio-NAD cycling. Accurate measurements were obtained to determine the concentration of even trace amounts of E7 oncoproteins present in urine specimens collected from patients. The ultrasensitive ELISA detected attomolar levels of E7 oncoprotein and approximately 500 cells/mL of HPV16-positive SiHa cells in solution. The ELISA index, which was determined by dividing the absorbance of the sample by the absorbance of the blank, was set to determine the threshold between E7-containing and E7-non-containing specimens, allowing us to examine the relationships among the ELISA index, uniplex E6/E7 PCR DNA types, and CIN grade.
The most important finding in the present study was the discrepancy between the results of the ultrasensitive ELISA (i.e., the presence of protein) and the results of uniplex E6/E7 PCR (i.e., the presence of DNA). We consider that such a discrepancy could be explained by a variation in the HPV life cycle or in the oncogenic activity. We previously developed an ultrasensitive ELISA antigen test for use in dengue fever [
26], and these studies have produced excellent results comparable to those of PCR-based tests, demonstrating that our ultrasensitive ELISA can be used to detect precise amounts of trace proteins. Therefore, it is reasonable that some specimens in the present study containing no or very low levels of E7 oncoproteins, contrary to the uniplex E6/E7 PCR-based HPV typing, represent inactive infections.
The present results demonstrated a higher prevalence of E7-positive cases with the lower-grade CIN lesions. Many studies to date have only measured mRNA, not protein, and thus the relationship between the transcription of the HPV 16 E7 oncogenes and the translation level of E7 protein levels in HPV infections is not fully established. Therefore, the present data are, to our knowledge, the first accurate research results at the protein level. As shown in
Figure 3, although it is unclear due to the small number of specimens, there was a tendency for CIN1 and 2 to have a higher ELISA index than CIN3. Because many HPV researchers think that the onset of cervical carcinogenesis is intricately linked to HPV infection, specifically through the transcription of high-risk HPV E6 and E7 oncogenes, leading to elevated levels of their mRNA, the present results may be considered controversial [
30,
31,
32]. Although the expression of E7 protein may have increased according to an increase in the CIN grade before starting the experiments, we could not confirm this because of the limited number of cases analyzed. We suggest the following two possibilities. (1) Protein expression levels may vary according to different stages of the HPV life cycle. Therefore, differences between the expression levels of E7 mRNA and E7 protein must be cautiously interpreted. A recent study demonstrated that E7 protein could not be detected while E7mRNA was highly expressed in oropharyngeal cancer, suggesting that the levels of protein expression may not reflect the levels of mRNA expression [
33]. (2) Another review suggests that E6 is translated from full-length E6-E7 mRNA, whereas E7 is translated from spliced forms of E6/E7 mRNA (E6*I, E6*II, etc.) [
34]. Different spliced mRNA products are likely to interfere with each other. In the viral replication stage of HPV, E7-mediated cell cycle re-entry in the middle layers of the squamous epithelium, caused by inhibition of retinoblastoma and its associated protein function, is most important for viral replication [
35]. Thus, E7 is likely to have a greater influence in this step than E6.
Furthermore, the E7 oncoprotein binds to the pRB via the
N-terminal LXCXE motif, inactivating pRB. The E7 oncoprotein of high-risk HPV promotes pRB degradation [
35]. When the expression of E6 and E7 mRNAs in the basal cells increases, the cells take over many steps necessary for cancer, such as suppression of immune response, immortality, transformation, apoptosis, and suppression of differentiation. This E6/E7 mRNA increase is expected to be required for CIN2. As the progression from CIN2 to CIN3, cells derived from basal cells that highly express E6 and E7 mRNAs are also located in differentiated layers (parabasal cell layers and spiny cell layers). We believe that in CIN3 the E6 oncoproteins, rather than E7, may have an important role. Highly expressed E6 oncoproteins inactivate p53 and simultaneously suppress the expression of the
Notch1 tumor suppressor gene, and cells are thought to acquire resistance to differentiation and abnormal proliferation ability. In this regard, we believe that it is necessary to distinguish carefully between CIN2 and CIN3, and this issue will be settled by examining the quantitative change in E6 oncoprotein in the future.
Many researchers believe that the deregulation of both E6 and E7 mRNA expression plays a critical role in cancer progression after HPV DNA is integrated into the host cell genome. The expression levels of E6 and E7 may differ at different stages of the HPV life cycle. Some studies suggest that high-risk HPV E6 may more actively induce malignant progression than low-risk HPV E6, as high-risk HPV E6 inhibits p53 (inhibition of apoptosis and cell cycle deregulation of cells with DNA damage), inhibition of PDZ-binding domain proteins (disruption of cell polarity and dedifferentiation), and activation of telomerase (immortalization) [
35]. Thus, E7 expression is also required, but perhaps not as much in the premalignant stage. In fact, cell proliferation is limited within the epithelium of high-grade intraepithelial lesions (CIN2, 3), whereas cell proliferation is more evident in benign HPV diseases such as condyloma. The cytotoxic T lymphocyte response to clear HPV16 infection is only attributed to its response against E6 and not E7 peptides in women with persistent HPV16 infection, as suggested by Nakagawa et al. [
36]. Our team previously published data showing that anti-E6 antibodies are more prevalent than anti-E7 antibodies in CIN3 patients, even though healthy individuals do not have these antibodies, and both antibodies are detected equally in cervical cancer patients [
37]. Thus, HPV16 E7 may be important in the stage of HPV replication (CIN1) but not as important in the stage of establishing high-grade lesions (CIN2, 3). Another simple explanation is that the present reaction to the E7 protein observed in CIN1 may be a reaction not only to the HPV16 E7 protein but also to E7 of other HPV16-related types, because multiple HPV-type infection was more frequently observed in CIN1 cases than in CIN3 or cancer cases. We cannot deny the possibility that, regarding the low positive rate for CIN3, it may be possible that cells are less likely to be mixed in the urine of CIN3 patients. In any event, further studies are needed to clarify this point. In the near future, we will develop our assay method to detect E6 oncoprotein expression to understand cancer progression better.
Holzinger et al., found DNA-positive and mRNA-negative samples, suggesting that latent infection samples may be present in oropharyngeal squamous cell carcinoma [
38]. On the other hand, Maglennon et al., attempted to detect both DNA and proteins of rabbit oral papillomavirus in epithelial basal layers, and although they detected low levels of RNA transcripts in latently infected tissues, they did not detect late viral proteins by immunofluorescence [
39]. Thus, they concluded that in a latent virus infection, the life cycle does not proceed, and the production of new viral proteins does not occur. Agreement may not necessarily exist between DNA, mRNA, and protein results. Furthermore, they demonstrated that in tissues of rabbit oral papillomavirus under latent infection, E7 and E6 mRNA levels were lower than the E2 level, which was low even during latent infection [
40]. These results are not inconsistent with the fact that E5, E6, and E7 proteins are required to produce a cellular environment that supports viral DNA replication [
41].
Current HPV testing is described in detail in the recent paper by Poljak et al. [
42], which lists the main groups and subgroups of commercial HPV tests available on the global market in 2023: (1) high-risk HPV DNA screening test without genotyping; (2) high-risk HPV DNA screening test with concurrent partial (HPV16/18/45), concurrent extended, or reflex partial genotyping for main high-risk HPV genotypes, and we expect that the number of different tests in this category will increase remarkably, reflecting updated guidelines and changes in screening and triage practices; (3) HPV DNA full genotyping test; (4) HPV DNA genotype- or group-specific genotyping tests; (5) high-risk HPV E6/E7 mRNA tests. There may be more tests using mRNA in the future; (6) in situ hybridization HPV tests; and (7) HPV DNA tests targeting multiple non-Alpha HPV genotypes.
The possibility for self-collection of urine specimens for HPV screening is strongly desired [
43,
44]. A recent survey reported in a Minnesota study [
45] showed an unexpectedly low rate of routine HPV screening in 21- to 29-year-old women. Reasons why some women avoid screens for cervical cancer include the following: embarrassment, fear of detrimental test results, time and cost, discomfort with male doctors, and an assumption of sexual surveillance, i.e., that cervical screenings are being used as a proxy to monitor their sexual activity [
46]. As a potential solution to increase the screening rate, self-collection of cervicovaginal specimens at home with a return by mail for HPV testing is an approach that may alleviate barriers to clinic-based cervical cancer screening [
47]. Although no statistically significant preference was demonstrated with urine self-sampling versus clinician sampling [
48], an easy-to-use self-collection of urine may be more accommodating for examinees. Furthermore, it is necessary to consider why E7 oncoprotein is mixed in the urine. It is difficult to imagine that it enters the bloodstream from the body, passes through the kidneys, and then comes out in the urine. It is reasonable to assume that E7 oncoprotein is contained in cells that are shed from the vaginal opening along with the vaginal discharge, and that it is mixed with the urine that comes out of the urethra.
It was not possible to provide a control population of women not infected with HPV because female patients with some symptoms of cervical cancer visited the obstetrics and gynecology department of the hospital where cervical biopsy specimens were obtained and classified according to the pathologic diagnosis of CIN. The biopsy results were CIN1, 2, and 3, and then the urine specimens were collected. This means that the hospital cannot collect biopsy specimens and urine specimens from a control population of women not infected with HPV. This raises the question of whether collecting a urine specimen from a normal, healthy woman meets the control conditions, and the answer is no because there is no guarantee that the urine specimen is from a woman who is not infected with HPV as she may be asymptomatic.
As a limitation of the present study, we acknowledge that the reliability of the marker of E7 oncoprotein in urine in the diagnosis of CIN is limited due to the lack of comparison between cervical and urine specimens. Furthermore, a greater number of patient samples is needed to evaluate the presence of E7 oncoproteins from high-risk HPV strains such as HPV18. Accurate detection, however, will depend on the availability of high-quality antibodies for use in our ultrasensitive ELISA method. Due to this limitation, we cannot test for E7 oncoprotein from HPV18 in urine at this time, although we may be able to produce quality antibodies against HPV18 E7 in the future.