Human papillomavirus, Epstein–Barr virus, and Candida albicans co‐infection in oral leukoplakia with different degrees of dysplasia

Abstract Objectives To identify human papillomavirus (HPV), Epstein–Barr virus (EBV), and Candida albicans in oral leukoplakia with different degrees of dysplasia. Materials and methods An observational, cross‐sectional, descriptive study was performed using 30 formalin‐fixed and paraffin‐embedded tissues from patients with clinical suspicion of leukoplakia and confirmed diagnosis of oral dysplasia. Histological analyses were performed by two pathologists (interobserver) and dysplasias were classified as mild, moderate, or severe. Conventional PCR was used to detect HPV and EBV viruses and C. albicans. To determine the association between each microorganism with different degrees of dysplasia a Chi‐square test was employed. Results The tongue was the most common site for leukoplakias (71.4%) in females with a mean age of 50 years (ranging between 30 to 50 years old; 57.1%). EBV was the most frequently detected (73.3%), followed by HPV (43.3%), mainly of type 16 (40%), and C. albicans (23.3%). Significant differences were observed between degrees of dysplasia and HPV presence (p = 0.005). In lesions positive for HPV, EBV, and C. albicans the most frequent histological changes were hyperkeratosis, irregular interpapillary ridges, and loss of basal stratum cell polarity. Conclusion Co‐infection with human papillomavirus, Epstein Barr virus, and Candida albicans in oral leukoplakia could be associated with dysplastic changes.

According to the Uppsala Symposium, Sweden 1994, clinical types of leukoplakia are classified as homogeneous and non-homogeneous. Homogeneous leukoplakias are generally asymptomatic and are recognized by a uniform white lesion slightly elevated compared to surrounding mucosa with a thin appearance. The surface can often present fissures, wrinkles, or a corrugated surface in appearance.

Epstein-Barr virus (EBV) has been implicated in multiple malignant neoplasms. Still, a causal association between leukoplakia and
EBV is yet unclear (Braz-Silva et al., 2008;Milagres et al., 2007). EBV replication increases in oral leukoplakias through mechanisms that favor its latency and persistence within the stratified squamous epithelium (Guidry et al., 2018).
Candida fungus is a common opportunistic pathogen in mucous membranes within the oral cavity and its activity increases in individuals with weakened immune systems (Sankari et al., 2015). Candida albicans is the most studied, representing a risk factor for premalignant transformation to malignant lesions in the oral cavity (Gainza-Cirauqui et al., 2013;Verma et al., 2015). Furthermore, several mechanisms, such as nitrosamine and acetaldehyde formation have been associated with histopathological changes, such as epithelial hyperplasia, hyperkeratosis, micro-abscesses, and chronic inflammation (Salvatori et al., 2016).
Because cancer is a progressive disease evolving from premalignant lesions to malignant stages; it is important to develop prognosis strategies during the early stages of the disease resulting from viral and fungal infections. These infections can result in tissue dysplasia associated with cancer advancement (Escribano-Bermejo & Bascones-Martínez, 2009;Pitiyage et al., 2009;Reibel, 2003).

| MATERIALS AND METHODS
An observational cross-sectional descriptive study was carried out with signed informed consent from all participants. The study was approved by the institutional ethics committee No. 014-2015 in compliance with the Helsinki declaration protocols. A total of 45 samples were collected and analyzed; of these, only 30 met the selection criteria and were included for further analyses. Exclusion criteria included patients with oral cancer diagnosis, under antibiotic medication, and consuming alcohol. An oral biopsy was obtained from patients with presumptive oral leukoplakia diagnosis for histopathological dysplasia analysis.

| Clinical and histological analysis of leukoplakias
After clinical examination, lesions were classified as homogeneous or non-homogeneous. A biopsy was collected and divided into two parts; one half was used for diagnostic purposes. To this end, tissue specimens were formalin-fixed, and paraffin embedded and stained with hematoxylin and eosin for histological analysis by two pathologists performing an inter-observer and unbiased analysis. Leukoplakias were classified using the Smith and Pindborg's classification based on their degree of dysplasia as mild, moderate, or severe, (Krogh et al., 1987;Pindborg et al., 1997). Histological analyzes included: loss of polarity of the basal stratum cells, presence of more than one basal cell layer with basaloid appearance, increased nuclear/cytoplasmic ratio, drop-shaped rete ridge, irregular epithelial stratification, increased number of mitotic figures with atypical cell shape, abnormally superficial mitoses, cellular and nuclear pleomorphisms, nuclear hyperchromatism, increased number and size of nucleoli, loss of cell cohesion, keratinization of isolated cells or in groups. If findings presented large discrepancies between both pathologists, then a third pathologist analyzed the slides and emitted his concept. This last evaluation was compared to the first two assessments, those with the highest concordance rate were selected (Kappa index >0.8).

| DNA extraction, quantification, and quality
The other half of the biopsy was used for DNA extraction using the Qiagen FFPE kit for paraffin-embedded tissue sections, following the manufacturer's instructions. DNA was quantified and quality assessed using a Nanodrop T2000 and 3% agarose electrophoresis. DNA quality was verified by amplifying the beta-globin gene (Table 1) by conventional PCR.

| HPV identification by conventional PCR
Specific primers were used to detect HPV in oral leukoplakias (Table 1), under the following PCR conditions: 1X GoTaq Buffer, 1 mM MgCl 2 , 1 μM GP5 +/GP6 + primers, 1.25 U GoTaq polymerase, 0.2 mM dNTPs, and 2 μl DNA sample (0.5 ng) in a total volume of 20 μl. Thermal cycling conditions were: 95 for 5 min, 95 for 1 min (1 cycle), 55 for 1 min, 72 for 1 min and 5 s (35 cycles), and one last cycle at 72 for 5 min. HeLa cells (positive for HPV) were used as a positive control and water as a negative control. Human papilloma virus PCR products were visualized in 3% w/v agarose gels in a Bio-Rad Gel Doc XR+ documentation system. Each experiment was carried out in triplicate and internal controls were used.

| Identification of HPV16 by conventional PCR
HPV-positive samples were confirmed with specific primers for HPV16 (Table 1)  Thermal cycling conditions were: 95 for 5 min, 95 for 1 min (1 cycle), 60 for 1 min, 72 for 1 min and 5 s (35 cycles), and the last cycle at 72 for 5 min. An HPV16 plasmid was used as a positive control (donated by Doctor Zur Hausen of the Deutsches Krebsforschungszentrum-DKFZ of Germany), and water was used as a negative control. PCR products for HPV 16 were visualized in 3% wt/vol agarose gel in a Bio-Rad Gel Doc XR+ documentation system.
Each experiment was carried out in triplicate and internal controls were used.

| Identification of EBV by conventional PCR
EBV identification was carried out using specific primers for EBV (Table 1)  water was used as a negative control. PCR products for EBV 16 were visualized in 3% wt/vol agarose gels in a Bio-Rad Gel Doc XR+ documentation system. Each experiment was carried out in triplicate and internal controls were used.
2.6 | Identification of C. albicans by conventional PCR C. albicans was identified by specific primers (Table 1) water was used as a negative control. PCR products for C. albicans were visualized in 3% wt/vol agarose gels in a Bio-Rad Gel Doc XR+ documentation system. Each experiment was carried out in triplicate and internal controls were used.

| Statistical analysis
Data were analyzed using means, standard deviation, and table of frequencies to determine whether there was a relationship between HPV, EBV, and C. albicans and different grades of dysplasia. A Chisquare test was performed using the SPSS statistical software, version 2016, with a significance of p < 0.05.

| Characteristics of the population
The study cohort consisted mainly of females (78.5%); with age ranging between 30 to 50 years old (57.1%). A higher incidence was observed in urban areas (64.2%), where the tongue was the most common site for leukoplakias (21.4%; Table 2).
Out of the samples analyzed 53.3% of the leukoplakias were of the homogenous clinical type. In this group 37.5% were positive for HPV, 62.5% for EBV, and 18.7% for C. albicans. In contrast, 46.6% of leukoplakias of the non-homogenous clinical type, 50% were positive for HPV, 85.7% for EBV, and 28.5% for C. albicans. (Table 3). However, no significant differences were observed between the clinical types and the presence of HPV, EBV, and C. albicans (Table 3).
Respecting histological characteristics, 43.3% had mild dysplasia, of which 23.0% were positive for HPV, 61.5% for EBV, and 7.6% for C. albicans. For the 36.6% with a moderate degree of dysplasia, 81.8% were positive for HPV and EBV, whereas 18.1% were positive for C. albicans. Last, for the 20% with a severe degree of dysplasia, 16.6% were positive for HPV, 83.3% for EBV and 66.6% for C. albicans. Interestingly, significant differences were observed between the degree of dysplasia and HPV and C. albicans with a p < 0.05 (Table 3) (Reibel et al., 2017).
Although several classifications of dysplasia have been proposed through the years, they all show a certain degree of limitation regarding subjectivity in determining changes and variability in interobserver results. Collectively, what becomes relevant is the identification of epithelial cell changes in dysplastic lesions to recognize their potential malignancy (Koch et al., 2011).
Therefore, in the present study the Pindborg classification was used (Krogh et al., 1987;Pindborg et al., 1997) taking into account the 13 criteria for classifying mild, moderate and severe dysplasia, determining the severity of the dysplasia, and the association with viral and fungal infections.
In the study of pre-malignant and malignant lesions presence of infectious agents in the oral cavity has gained importance because some lesions develop in the absence of the main risk factors, such as alcohol and tobacco consumption. Moreover, certain pathogenic microorganisms such as HPV (Nielsen et al., 1996), EBV (Guidry et al., 2018) and Candida albicans (Chiu et al., 2011;Verma et al., 2015) can alter cellular function and induce histological alterations in oral cavity tissues.
In regard to gender, 78.5% of the women presented oral leukoplakia, similar to what was reported by Kristoffersen et al., 2012, who studied 50 samples of leukoplakia and found females were the most frequently affected (44%; Kristoffersen et al., 2012). in patients with oral leukoplakia (Yang et al., 2009). In 2012, Kristoffersen reported an average age of 56 years (Kristoffersen et al., 2012) and, Szarka et al. (2009) reported an average age of 56 years (Szarka et al., 2009).
The aforementioned data shows that although oral disorders with malignant potentials, such as leukoplakia, are diagnosed more frequently in advanced age; timely diagnosis is challenging, due to lack of clinical examination follow-up and absence of symptoms (Warnakulasuriya, 2018). Regardless, based on evidence it has been demonstrated the natural history of the disease is progressive and preventable (Goodson et al., 2015).
We determined that leukoplakias were most frequently infected by EBV (73.3%), followed by HPV (43.3%), in addition to C. albicans infection (23.3%). The main HPV serotype was 16, which was found in 40% of the samples included. Presence of these infections could lead to dysplasia in oral cavity tissues and induce malignant processes (Stojanov & Woo, 2015).
Human papillomavirus infection is an important risk factor for oral leukoplakia and has been associated with carcinogenic processes (Angiero et al., 2010;Chen & Zhao, 2017). Since viruses are obligate intracellular microorganisms, they can induce cellular alterations, which can result in dysplasias (Lerman & Woo, 2014;Stojanov & Woo, 2015). This was evidenced in this study, where a significant difference (p = 0.005) was observed between presence of HPV and cellular atypia at the level of dysplasia (mild, moderate, and severe; Table 2). According to our results, the most frequent histological changes were hyperkeratosis, irregular ridges, teardrop-shaped papillae, more than one basal cell layer with basaloid appearance, cellular and nuclear pleomorphism, and cells with koilocytic appearance in HPV-infected moderate-grade leukoplakia.
Some HPV mechanisms that could cause malignant transformation include E6 and E7 expression, two oncoproteins that interfere with p53 and pRB tumor suppressor protein activity within the cell cycle, thus promoting uncontrolled proliferation and, subsequent cancer (Erira et al., 2015;Rampias et al., 2009). These mechanisms could also be related to dysplastic changes at early stages (Chen & Zhao, 2017;Khanal et al., 2017).
Furthermore, EBV infection has also been frequently reported in healthy individuals in addition to leukoplakia (Kikuchi et al., 2016;Sand et al., 2002). Even though it is an important risk factor for malignant transformation (Acharya et al., 2015), the association between infection, premalignant lesion onset, and cancer is still controversial (Kikuchi et al., 2017;Søland et al., 2020). Several mechanisms responsible for these changes could be based on LMP1 protein expression, which allows for viral replication within cells and could alter cell function and induce dysplastic changes (Ali et al., 2015;Kikuchi et al., 2016). The findings of the present study include several histological changes, such as cellular and nuclear pleomorphism, more than one layer of basal cells with basaloid appearance, and hyperkeratosis in leukoplakia positive for EBV.
Co-infection with HPV and EBV has been reported in precancerous lesions and OSCC. (Jalouli et al., 2010(Jalouli et al., , 2012Jiang et al., 2015;Sharma et al., 2019). This viral interaction has allowed us to evidence that HPV infection is important, however, not sufficient, for cancer development, since not all individuals infected with HPV progress into cancer (Guidry & Scott, 2017). However, presence of other microorganisms, such as EBV and C. albicans, may trigger a synergistic mechanism or enhance HPV's malignancy.
It is important to highlight that this fungus can invade superficial epithelial layers, and most strains perform what is known as "commutation," which allows the fungus to change phenotypes with different morphological and functional properties (Sitheeque & Samaranayake, 2003). This, in turn, could induce histological changes, such as those reported in the present investigation, which has also been reported in other studies (Gujral et al., 2010).
This fungus increases its activity in humans with certain immune deficiencies (Salvatori et al., 2016). Moreover, within the oral cavity it is a risk factor for premalignant lesion transformation into malignant lesions (Gainza-Cirauqui et al., 2013;Verma et al., 2015), mainly in the presence of tissues injured by trauma, maceration, atrophy or friction.
Some mechanisms, such nitrosamine and acetaldehyde production have been associated with histopathological changes, including epithelial hyperplasia, hyperkeratosis, microabscesses, and chronic inflammation (Salvatori et al., 2016). However, this study identified the most frequent histological changes were loss of basal stratum cell polarity, atypical mitoses, nuclear hyperchromatism, and acanthosis in severe grade leukoplakias positive for C. albicans.

| CONCLUSION
Leukoplakia on the tongue was most frequently infected by EBV, and the presence of HPV and C. albicans could be associated with dysplastic disorders.

CONFLICT OF INTEREST
The authors declare there is no conflict of interest.

ETHICS STATEMENT
The study was approved by the Ethics committee of the Universidad Cooperativa de Colombia-Bogot a Colombia (No. 014-2015). All samples were collected from participants after signed informed consent under the ethical norms of the declaration of Helsinki.

DATA AVAILABILITY STATEMENT
The data is not available because it compromises ethical standards.