In Vitro Model Characterizing Carcinogenic Progression of HPV-Positive Oropharyngeal Cancer

Simple Summary

High-risk human papillomavirus (hrHPV), particularly HPV16, is a major driver of oropharyngeal squamous cell carcinoma, now the most common HPV-related cancer in the U.S., surpassing cervical cancer. Despite available vaccines, low uptake and shifting sexual behaviors have contributed to rising incidence, with projections exceeding 30,000 cases annually by 2029. Early detection remains difficult due to the absence of precursor lesions and long latency between infection and symptom onset. To address these limitations, we developed an in vitro HPV16 oral cancer model using the 3D organotypic raft culture system. This model mimics the progression of HPV16-transfected oral epithelium from precancerous to malignant states and allows for detailed monitoring of structural and biochemical changes. Validated using established markers of high-grade lesions, the model allows us to study the early progression of HPV16-driven disease and supports the search for reliable biomarkers to improve diagnosis and guide treatment.

Abstract

Background/Objective: Human papillomavirus (HPV) represents the most widespread sexually transmitted infection globally, with high-risk strains such as HPV16 driving a rising incidence of oropharyngeal squamous cell carcinoma (OPSCC), particularly in developed countries like the United States and United Kingdom. In the U.S., HPV16-associated OPSCC has surpassed cervical cancer as the most common HPV-related malignancy. Despite the availability of preventive vaccines, uptake remains suboptimal among adolescents and shifting sexual behaviors have contributed to increased disease burden. Early detection remains a major clinical challenge due to the absence of defined precursor lesions and the extended latency between viral exposure and disease onset. Most patients present with advanced-stage disease and no prior clinical history of pre-malignancy, limiting access to early-stage samples and hindering biomarker discovery. Methods: To address these limitations, we developed an in vitro HPV16 oral cancer model, using the three-dimensional organotypic raft culture system that simulates the progression of HPV16-transfected oral epithelium from precancerous states to malignant phenotypes. Results: Using HPV16-transfected human tonsil keratinocytes, we generated stratified and differentiated epithelia that mimic the biochemical and structural changes observed in vivo. This system enables detailed monitoring of epithelial differentiation, biochemical shifts, viral genome status, and key oncogenic and metabolic markers associated with HPV16-driven OPSCC. By aligning expression profiles with clinical datasets, we validated the model through the measurement of virologic markers linked to infection and progression, as well as tissue markers indicative of carcinogenic transformation. Conclusions: This model offers a promising tool for refining early detection strategies and evaluating potential clinical biomarkers, ultimately aiming to improve diagnostic precision and therapeutic outcomes in HPV-associated OPSCC.

1. Introduction

Human Papillomavirus (HPV) is the most widespread sexually transmitted infection, worldwide. High-risk HPV is associated with a notable disease burden and cancer, causing approximately 5% of all cancers across the world [1]. HPV causes over 95% of cervical cancers and roughly 85% of anal cancers. High-risk HPV types also contribute to a substantial and increasing proportion of oropharyngeal head and neck cancers, now representing about half of all oropharyngeal squamous cell carcinoma (OPSCC) cases globally [1]. Approximately 70% of all newly diagnosed OPSCC cases in the United States having detectable HPV DNA, highlighting the viruses’ central role in disease development [1]. Of these, 85–96% are caused by high-risk HPV16 infection [2,3]. The rise in HPV-associated OPSCC incidence has now made it the most common HPV-related malignancy in the United States, surpassing cervical cancer [4,5]. Currently, approximately 20,000 new cases of HPV-associated OPSCC are diagnosed, with projections indicating an upward trend of more than 30,000 cases per year by 2029 [3]. Current HPV vaccination efforts remain stagnant, with statistics estimating 78.2% of adolescents between the ages of 13 and 17 having received the first dose of the vaccine [6]. The increased incidence of HPV⁺ OPSCC has been associated with a shift in risk factors such as increased oral sexual behavior amongst younger age groups, noting a significant correlation between lifetime oral sex partners and HPV⁺ OPSCC incidence [7,8,9]. Other risk factors such as history of tobacco and alcohol use have been associated with OPSCC development. However, incidence of these associated cancers has declined over the past 30 years with HPV16-associated OPSCC taking over as the prominent form of this cancer. Therefore, as rates of HPV-associated OPSCC in the US and other countries increase, it is of great importance to study and comprehend this malignancy.

Early and intermediate detection of HPV-associated OPSCC remains a significant clinical challenge. The absence of a defined precursor lesion and the extended latency between viral exposure and disease onset complicate efforts to establish effective screening protocols [10,11]. Typically, patients present with a lateral neck mass, often the first noticeable symptom, by which regional metastasis has occurred and no prior clinical signs of malignancy are evident. At this stage, patients often experience substantial short- and long-term treatment-related morbidities, including chronic dysphagia, lower cranial neuropathies, difficulty chewing, and voice or speech impairments [12,13]. Although p16 immunohistochemistry, HPV E6/E7 mRNA detection, and HPV DNA identification via polymerase chain reaction (PCR) and in situ hybridization are considered diagnostic benchmarks, alternative approaches such as HPV serology, transcervical ultrasound imaging of the neck and oropharynx, and mucosal imaging have yet to demonstrate consistent reliability [7,14,15,16]. Moreover, HPV-associated OPSCC is the most expensive HPV-related malignancy to treat, with costs approaching $160,000 within the first two years [17]. Multimodality treatment regimens can cost nearly twice as much as single-modality approaches [18]. Therefore, early detection could promote the use of less intensive treatment plans, potentially reducing treatment-related morbidity, mortality, and overall healthcare costs, while improving patient quality of life. These limitations highlight the urgent need for reliable biomarkers and improved screening platforms.

In this study, we established an in vitro oral cancer model using the organotypic raft culture system to address limitations in early detection of HPV-associated OPSCC. This system enables the study of HPV16-transfected oral epithelium and its progression from a precancerous state to a phenotype resembling malignant tissue, thereby offering a promising tool for evaluating early disease markers and refining detection strategies. We monitored phenotypic alterations during successive passaging by assessing established markers of high-grade HPV-associated OPSCC. Due to the limited availability of data on pre-malignant HPV-associated OPSCC lesions, our analysis prioritized gene expression profiles commonly linked to oropharyngeal cancers in clinical studies. Organotypic raft cultures generated from HPV16-transfected tonsil keratinocytes were used to examine shifts in epithelial differentiation, biochemical differentiation expression patterns, viral genome status, and key carcinogenic and metabolic indicators associated with HPV-driven OPSCC progression.

2. Materials and Methods

2.1. Isolation of Primary Tonsil Keratinocytes

Tonsillar specimens were sourced from routine tonsillectomy procedures. All tissues were de-identified prior to processing with removal of information such as name, age, and race. Authorization to use these samples as “discarded samples” was granted by The Pennsylvania State University College of Medicine Institutional Review Board (IRB# 25284). Mixed epithelial cell populations were generated following previously published methods [19]. In summary, the epithelial layer was separated from the underlying connective tissue and dermis, which were discarded. The isolated epithelium was rinsed three times in phosphate-buffered saline (PBS; Corning, Manassas, VA, USA) supplemented with 50 µg/mL Gentamycin sulfate (Gibco BRL, Bethesda, MD, USA) and 2× Nystatin (Sigma Chemical Co., St. Louis, MO, USA). The tissue was then chopped with scalpels and subjected to trypsin digestion in spinner flasks to obtain single-cell suspensions. After trypsinization, the cell mixture was collected and combined with 20 mL of E media containing 5% fetal calf serum (FCS; Cytiva, Logan, UT, USA). Next, centrifugation pelleted the cells, supernatant was aspirated, and pellets were resuspended in 1 mL of 154 Media (Life Technologies, Grand Island, NY, USA). The suspension was transferred to culture dishes containing 7 mL of 154 Media. Two additional rounds of dissociation were then repeated, to increase total cell yield. At 70% confluency, the cells were split and electroporated.

2.2. Electroporation of the HPV16 Genome into Primary Tonsil Keratinocytes

HPV16 epithelial lines were obtained via electroporation of the full length HPV16 genome into low passage primary tonsil keratinocytes and maintained according to established protocols in the laboratory [20]. Cell line populations that stably maintained the HPV16 genome were used for raft culture production. HPV16 virus stocks were produced from raft production. The cloned HPV16 genome: HPV16:114B prototype European variant [21,22], was digested with BamHI to linearize the viral DNA and separate it from the vector sequence. A total of 10 μg viral DNA was electroporated into primary cells using a Gene Pulser (Bio-Rad Laboratories, Hercules, CA, USA). Immortalized keratinocytes stably maintaining HPV16 genomes following electroporation were cultured with mitomycin C-treated J2 3T3 feeder cells and maintained in E-Media.

2.3. Organotypic Raft Culture Production

Raft tissues were generated following established protocols [23,24,25]. 3T3 J2 fibroblasts (feeder cells) were trypsinized and resuspended in 10% reconstitution buffer, 10% 10× Dulbecco’s Modified Eagle Medium (DMEM) (Life Technologies, Gaithersburg, MD, USA), 80% collagen (Dickinson, Franklin Lakes, NJ, USA), and 2.4 μL/mL of 10 M NaOH. Feeder cells were incorporated at a density of 6.25 × 10⁵ cells/mL. 2.5 mL aliquots of the collagen mixture was added to each well of 6-well plates. The aliquots were then incubated for 1 h to solidify, and then the matrices were equilibrated by adding 2 mL of E Media to each well after incubation. The HPV16 tonsil cell lines were trypsinized and resuspended in E Media at 2 × 10⁶ cells/mL, and 1 mL of this suspension was seeded onto each collagen matrix-containing well. The seeded matrices were then incubated for 2 h to permit cell adherence to the dermal equivalent. After removal of the media, the collagen matrices were lifted onto stainless steel grids at the air-liquid interface and fed with E-Media supplemented with 10 μM 1,2-dioctanoyl-sn-glyserol (C_8:0; Cayman Chemical Company, Ann Arbor, MI, USA). Rafts were harvested after 20 days (day 20) for analysis. Feeding and C8 treatment was done every other day. After raft tissues were harvested, preparation of viral stocks and determination of viral titers were performed as described in subsequent sections.

2.4. Preparation of HPV16 Viral Stocks

HPV-transfected raft tissues were harvested as described [24]. To generate CV stocks, two rafts were disrupted using a Dounce homogenizer in 500 μL of phosphate buffer consisting of 0.05 M sodium phosphate (pH 8.0), and 2 mM MgCl₂. An additional 250 μL of the same buffer was then used to rinse the homogenizer. The 750 μL virus prep was then digested with 1.5 μL of benzonase (Millipore Sigma, Darmstadt, Germany) to eliminate non-encapsidated viral genomes for 1 h at 37 °C. Following incubation, 188 μL of ice-cold 5 M NaCl was added to each virus prep to adjust the salt concentration to 1 M. The samples were vortexed thoroughly and centrifuged at 10,500 rpm for 10 min. The resulting supernatants, representing CV stocks, were stored at −80 °C until further use.

2.5. HPV16 Viral Stock Titering

HPV16 titers were quantified using a qPCR-based DNA encapsidation assay as outlined previously [24]. To detect genomes within CV preparations that were resistant to endonuclease digestion, the following protocol was used. 10 μL benzonase-treated CV stock were mixed with 200 μL HIRT DNA extraction buffer (400 mM NaCl/10 mM Tris-HCl (pH 7.4)/10 mM EDTA (pH 8.0)), along with 10 μL 10% SDS and 2 μL of 20 mg/mL proteinase K (Sigma Aldrich, St. Louis, MO, USA), and incubated for 2 h at 37 °C to release viral genomes. After digestion, the DNA was purified twice with phenol-chloroform-isoamyl alcohol (25:24:1), followed by an additional extraction with an equal volume of chloroform. DNA was precipitated overnight at −20 °C using ethanol. The resulting pellet was collected by centrifugation, washed with 70% ethanol, and dissolved in 20 μL of Tris-EDTA, where it was allowed to resuspend overnight. Quantification of viral genomes was performed using the Thermo Scientific Maxima SYBR Green qPCR kit (Thermo Fisher Scientific Baltics UAB, Vilnus, Lithuania). The HPV16 E2 open reading frame (ORF) was amplified using 0.3 μM each of the forward (HPV16E2 5′) and reverse (HPV16E2 3′) primers (Table S1). A standard curve was generated using serial dilutions of pBSHPV16:114B DNA ranging from 10⁸–10⁴ copies/μL. A Bio-Rad iQ5 Multicolor Real-Time qPCR machine and software (Bio-Rad Laboratories, Hercules, CA, USA) were utilized for PCR amplifications and subsequent data analysis.

2.6. Immunohistochemistry and Immunofluorescence Staining

Raft cultures were fixed in 10% buffered formalin and embedded in paraffin, and 4-μm cross sections were prepared. A section from each sample was stained with hematoxylin and eosin as previously described [26].

For immunofluorescence staining, the slides were submerged in xylene for deparaffinization and then were rehydrated. Antigen retrieval was achieved by submerging the slides in Tris-EDTA buffer (pH 9) in a 90 °C water bath for 10 min. The slides were then rinsed with Tris-buffered saline (TBS)–Tween and blocked with the Background Sniper blocking reagent (Biocare Medical, Pacheco, CA, USA). The slides were then stained with the primary antibody overnight at 4 °C. Each sample was stained with antibodies against Keratin 5 (1:1000 dilution; catalog number MA5-12596; Invitrogen, Rockford, IL, USA), Keratin 10 (1:1000 dilution; catalog number MA1-06319; Invitrogen), Keratin 6 (1:1000 dilution; catalog number MS-766-P; NeoMarkers, Fremont, CA, USA), p53 (DO-1) (1:1000 dilution; catalog number sc-126; Santa Cruz Biotechnology, Paso Robles, CA, USA), Rb (Rb1) (1:1000 dilution; catalog number sc-73598; Santa Cruz Biotechnology), Keratin 8 (1:1000 dilution; catalog number MA5-14428; Invitrogen), p16^INK4A (1:1000 dilution; catalog number sc-56330; Santa Cruz Biotechnology), COX5b (C-5) (1:1000 dilution; catalog number sc-374416; Santa Cruz Biotechnology), and p120 (6H11) (1:1000 dilution; catalog number sc-23873; Santa Cruz Biotechnology). The slides were then rinsed with TBS-Tween 3 times and stained with secondary antibody (Alexa Fluor 488; Life Technologies, Eugene, OR, USA) diluted 1:200 for 1 h at room temperature. Next, the slides were stained with Hoechst nuclear stain (1:5000 dilution) for 15 min and rinsed with TBS-Tween twice. All antibodies were diluted in Da Vinci Green diluent (Biocare Medical). A Nikon Eclipse 80i microscope and NIS Elements (v4.4) software (Nikon Instruments Inc., Melville, NY, USA) were used to acquire images.

3. Results

3.1. Development of a Multistep HPV16 Oral Cancer Progression Model

There is an increasing need to develop effective screening techniques to enable early and intermediate detection of disease. However, challenges such as lack of a precursor lesion and a long latency period between virus exposure and disease manifestation complicate this effort. The most common HPV-associated OPSCC symptom is a lateral neck mass. As a result, patients often arrive to the clinic with an advanced stage of disease due to regional metastasis and are without a clinical history of pre-malignancy [27,28,29,30,31]. Immunohistochemical staining for p16, detection of HPV E6/E7 mRNA, and identification of HPV DNA through polymerase chain reaction (PCR) and in situ hybridization are considered the gold standard methods for HPV-associated OPSCC screening [7]. In addition to these, other screening techniques include HPV serology, transcervical ultrasound imaging of the neck and oropharynx, and mucosal imaging. However, none of these techniques are reliable sources for cancer evaluation [14,15,16]. Therefore, the growing demand for reliable clinical biomarkers has driven the development of an in vitro oral cancer model, utilizing the organotypic raft culture system, designed to address current limitations in early detection and disease evaluation (Figure 1). With continuous passaging, our system can mimic in vitro the progression of precancerous HPV16 oral epithelium to resemble cancer-like tissue. To ensure our in vitro system accurately reflects the in vivo environment, we evaluated phenotypic changes during continuous passaging using known markers of high-grade HPV-associated OPSCC. Given the scarcity in pre-malignant HPV-associated OPSCC lesion data, we focused on the expression of genes that are associated with oropharyngeal cancers in vivo as reported in literature. Using organotypic raft tissues derived from HPV16-transfected tonsil keratinocytes, we have evaluated changes to differentiation and keratin expression, the HPV genome state, and carcinogenic and metabolic markers of HPV-associated OPSCC progression.

3.2. Validation of HPV16 Virus Production and Infectivity in Tonsillar Epithelial Cells

We then tested and validated the established oral cancer model on a productive HPV16 life cycle by quantifying HPV16 titers though quantitative RT-PCR analysis. Organotypic raft cultures of HPV16-transfected HTLK cells were grown for 20 days and treated with 1,2-dioctanoyl-sn-glyserol (C8), as described in the Materials and Methods section. Raft tissues were harvested, and viral titers were quantified by amplifying the E2 open reading frame, as described previously [23]. HPV16 virion production in both male and female HTLK lines showed an initial passage dependent increase in the amount of encapsidated genomes. Peak encapsidated HPV16 virus production was observed by passage 20, with a significant decline noted beyond passage 20. Nonetheless, both male and female HPV16-transfected tonsil tissue initially exhibit a productive viral life cycle (Figure 2, Panels B and C). As stated, continuous passaging of both male and female HTLK lines result in decreased viral titers, suggesting integration of the HPV16 DNA into the host genome. When infected with HPV16, the viral genome can exist in the cell as an episome or can integrate into the human genome. HPV episomes are detected in non- and pre-malignant tissues, while integrated genomes are detected in malignancies [32,33]. The disruption of E2 gene expression is correlated with integration of the viral DNA into the human genome, a milestone in HPV-associated cancer development as it results in the dysregulation of the E6 and E7 oncoprotein expression. Using the same HPV16 tonsil lines, we previously published findings showing a decrease in the E2/E6 ratio in both male and female HTLK lines with continuous passaging, suggesting integration of HPV16 into the host genome, a key milestone in HPV-associated cancer development, with integration also shown previously by sequencing [34].

3.3. Morphological Differentiation and Stratification of Tonsil Keratinocytes in Raft Cultures

Phenotypic changes in HPV infected tissue occurs during progression to cancer which can be utilized as carcinogenic markers. The organotypic raft culture system is an important method for studying both normal epithelia and epithelial carcinogenesis in vitro. Not only can organotypic cultures mimic the in vivo progression of non-transformed epithelia but progression to cancer can be observed within this in vitro system [35,36,37,38,39,40,41,42,43,44]. Primary tonsillar epithelial cells were isolated from human tonsil samples and established in E Media as described in the Materials and Methods. Multiple mixed pools of tonsillar tissue were obtained from patients undergoing tonsillectomies. The isolated primary tonsil epithelial cell lines are then transfected with the HPV16 genome via electroporation and cultured as monolayers. HPV16 tonsil cell lines were also produced through natural HPV16 infection; however, genome delivery via electroporation proved to be a more efficient method. The HPV16-transfected tonsil keratinocytes are continuously passaged and at every 5th passage, the cells are grown in organotypic culture for analyses (Figure 1). Epithelial cells transfected with HPV16 DNA, induces cellular immortalization, and when grown in organotypic culture, the tissue presents an altered differentiation phenotype, similar to what is seen in vivo. The HPV16-transfected tonsil lines were grown in the organotypic culture system at increasing passages to measure the progression of morphological differentiation and stratification. Male and female tonsil rafts were harvested for paraffin embedding after a 20-day incubation period as described in the Materials and Methods. Thin sections of the raft cultures were histochemically analyzed and compared to normal primary tonsil keratinocyte raft cultures. Hematoxylin and eosin (H&E) staining was done to examine tonsillar epithelial morphology and stratification (Figure 3). Human epidermal cells in vivo, undergo extensive stratification, marked by a sequential progression of morphological changes characteristic of terminal differentiation [45,46,47,48,49]. Prior studies have confirmed that the entire HPV16 genome, in particular, the requirement of E6 and E7 oncogenes expression, can alter morphological differentiation in vitro similarly to biopsies of HPV16 anogenital lesions in vivo [38,43]. H&E staining’s of HPV16-transfected tonsil rafts were basal-like and progressively declined in stratification and terminal differentiation. Raft tissues of lower passage 7 (P7) HPV16 tonsil cells exhibited organized and defined epithelial layers such as a smooth basement membrane (indicated by arrows in Figure 3 Panels B and C P7). However, continuous passaging resulted in: (i) the increased disorganization and cell overcrowding in the parabasal layer (indicated by an arrow in Figure 3 Panel C P43), (ii) vacuolation (indicated by arrows in Figure 3 Panel B and C P27), (iii) abnormal nuclear appearance by shape and enlarged size, (iv) koilocytosis in the male and female tonsil raft tissues (indicated by arrows in Figure 3 Panel B and C P52 and P22, respectively), (v) and keratin pearl formation (indicated by arrows in Figure 3 Panel B P17 and P42, respectively) (Figure 3). These features are often observed in oral intraepithelial neoplasias in vivo. Additionally, primary human tonsil keratinocytes grown in organotypic culture system exhibited morphological characteristics resembling those observed in vivo. An overall impairment of the tissue architecture and stratification was observed with continuous passaging in the developed in vitro model, successfully mimicking its in vivo counterparts during progression to carcinoma.

3.4. Measuring of Biochemical Expression Patterns of Differentiation and Wound Healing Markers of Tonsil Keratinocytes in Raft Cultures

As keratinocytes commit to terminal differentiation, they initiate a series of biochemical transitions, with keratin synthesis and expression serving as a key hallmark [47,50,51]. To assess the expression of biochemical markers of differentiation, primary and HPV16-transfected tonsil rafts were harvested for paraffin embedding as described in the Materials and Methods. Thin sections of the raft cultures were subjected to immunofluorescence staining and compared to primary tonsil rafts. The results are summarized in

Table 1. Results of biochemical differentiation and cancer marker expression in male and female HPV16-transfected human tonsil keratinocyte rafts grown at increasing passages. “+” indicates expression of the marker at the corresponding passage.

Passage #	M19					M28					F25					F26
	K5	K6	K10	K8	P16	K5	K6	K10	K8	P16	K5	K6	K10	K8	P16	K5	K6	K10	K8	P16
P7	+	+	+	+	+	+	+				+	+				+	+
P12	+	+	+			+	+				+	+				+	+
P17	+	+	+			+	+				+	+				+	+
P22	+	+				+	+				+	+				+	+
P27	+	+				+	+				+	+				+	+
P32	+	+				+	+				+	+				+	+		+
P37/39	+	+	+	+		+	+				+	+				+	+	+	+
P42/43	+	+	+	+		+	+				+	+				+	+	+	+
P47/48	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
P52	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
P62	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
P76	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+

Each color represents a different marker. Blue: K5, Purple: K6, Orange: K10, Green: K8, Red: P16.

. High levels of keratin 5 (K5) are present in the epithelial basal layer at all sites in the human oral cavity. This keratin plays a significant role in maintaining cellular proliferation and preserving tissue integrity [52,53]. Continuous passaging of the HPV16 HTLK cell lines reveal a downregulated expression pattern of K5 within the raft tissues of males and females. Instead of remaining in the basal layer (example indicated by arrows in Figure 4 Panels A and B P7), K5 expression is present across the epithelia at all passages (Figure 4, Panels A and B). Keratin 10 (K10) protects epithelia tissue from trauma or damage by forming dense bundles that confer mechanical strength to the epidermis. It also inhibits keratinocyte proliferation. K10 is typically absent in normal non-keratinizing oral epithelia, such as tonsils. In contrast, K10 is typically observed at homogenous levels in the suprabasal layers of keratinizing oral epithelia, such as gingiva. Its expression marks terminal differentiation and keratinization [53,54]. Continuous passaging of the HPV16 HTLK cell lines resulted in expression of K10, in later passages (Figure 5, Panels A and B). However, as passaging increased, K10 expression increased and localized across the tissue in males and females, consistent with its in vivo counterparts. Keratin 6 (K6) is a wound healing marker expressed in the suprabasal layer of the oral epithelia. In the occurrence of injury to stratified epithelia, activated keratinocytes at the wound edge express K6 [55]. Compared to primary tonsil rafts, an increased K6 expression pattern was observed across the HPV16-transfected tonsil raft tissue with continuous passaging (Figure 6, Panels A and B). The altered expression of K6 suggests a compromised wound healing response in the developed model. Overall, the progressive and aberrant expression patterns of differentiation and wound healing markers in our in vitro model effectively recapitulates the in vivo transition to carcinoma, further validating its relevance as a tool for studying HPV16-associated oral cancer progression.

3.5. Measuring Expression Patterns of Carcinogenic and Metabolic Markers of Tonsil Keratinocytes in Raft Cultures

Expression of additional markers of HPV-associated carcinogenesis and metabolism were also investigated in the HPV16-transfected rafts. Keratin 8 (K8) is a tumor-associated antigen expressed during HPV-associated malignant transformation and has progressive potential in lesion development [56]. Immunofluorescence staining for the K8 marker revealed a low expression pattern in earlier passages of the HPV16-transfected HTLK cell lines. Earlier passages (Figure 7, Panels A and B P7) showed low to no K8 expression. This pattern remained consistent until later passages (Figure 7, Panels A and B P37-52) where stronger K8 staining is observed and localization is present across the tissue layers. p16 overexpression is regarded as an adequate surrogate marker of oncogenic HPV infection [57,58]. Normally silenced by epigenetic repressors, p16 is upregulated in HPV+ cells through E7 oncoprotein-mediated retinoblastoma (Rb) expression loss and activation of the KDM6B demethylase, which lifts repression of the p16 locus. In the Rb deficient cells, p16 becomes essential for survival, highlighting that p16 can adopt oncogenic roles, ensuring cancer cell survival [59]. Primary and HPV16-transfected rafts were harvested for paraffin embedding as described in the Materials and Methods. Sections of the raft cultures were subjected to p16 immunofluorescence staining and compared. Staining revealed a stronger p16 presence in later passages. Lower passages revealed small amounts of p16 expression across the male and female raft tissues (Figure 8, Panels A and B P12 and P24; and Table 1); however, continuous passaging revealed increased p16 expression across the basal and differentiating layers as the passage numbers increased (Figure 8, Panels A and B P48 and P76; and Table 1).

Integration of the HPV16 genome results in increased expression of the E6 and E7 oncoproteins, which drive oncogenic transformation of host epithelial cells. E6 and E7 reprogram the host by targeting key tumor-suppressive pathways, including p53 and retinoblastoma (pRb), to support continuous proliferation. Specifically, HPV16 E6 promotes degradation of p53, while HPV16 E7 functionally inactivates pRb, thereby reducing cell-cycle control enabling sustained use of host replication machinery [60]. Because direct detection of E6 and E7 proteins is often unreliable in stratified epithelial systems, and consistent with prior findings demonstrating that E6-mediated p53 loss and E7-dependent disruption of pRb are hallmark functional readouts of high-risk HPV activity, we used p53 reduction and pRb loss as surrogate markers of E6 and E7 function in our established model [61,62,63]. Immunofluorescence staining of p53 and pRb revealed low levels of both proteins in later-passage tonsil raft tissues, comparable to levels observed in primary tonsil raft tissue not transfected with the HPV16 genome (Figure 9, Panels A and B; Figure 10, Panels A and B).

Cytosolic host protein p120 catenin (p120) expression is indicative of HPV infection as it targets the virus $\gamma$-secretase for host-cell membrane insertion [64]. Additionally, p120 catenin functions as a critical regulator of the cadherin-catenin adhesion complex, which mediates cell-cell adhesion and governs signal transduction pathways that influence cellular processes such as growth, differentiation, polarity, and migration. Mutagenic and epigenetic downregulation of p120 is implicated in cancer development and is associated with invasiveness and progression of human epithelial tumors, such as OPSCC [65,66,67,68]. After examining the tonsil raft tissues for p120 expression by immunofluorescence staining, high p120 levels were noted in the basement membrane of the tissue at lower passages (indicated by arrows at P27 in Figure 11, Panels B and C). As passaging continued, p120 levels progressively diminished until little to no detectable signal remained (Figure 11, Panels B and C).

Tumors are known to favor glycolysis over cellular respiration for energy production. Cytochrome c subunit 5B (COX5B), a metabolic marker localized in the inner mitochondrial membrane, contributes to proton gradient formation during ATP synthesis and is studied for its role in characterizing tumor metabolism. In normal oral mucosa, COX5B localizes in the basal layer and is expressed at little to no levels [69]. After staining for COX5B in the raft tissues, there was low to absent expression of COX5B in the tissue. However, continuous passaging revealed an increased COX5B expression pattern in the tissue, indicating that our HPV16 tonsil raft tissues are metabolically highly active (Figure 11, Panel D). Therefore, progressive passaging of HPV16-transfected tonsil raft cultures revealed dynamic expression changes in relevant markers of HPV-associated carcinogenesis and metabolism, including late-stage induction of p16, attenuation of p120, and upregulation of COX5B, further supporting the in vitro model’s relevance to HPV16-associated oral transformation.

4. Discussion

HPV replication is closely tied to the differentiation program of its host tissue, the squamous epithelium. Our laboratory was the first to report the in vitro propagation of HPV using a cervical neoplasia-derived cell line [26], and we have since expanded our repertoire to include multiple keratinocyte types. We have extensive experience generating unique primary cell lines from individual tissue samples, including oral epithelial tissues [70,71,72,73,74,75]. In the present study, we report for the first time the development of a three-dimensional in vitro model of HPV16-associated oropharyngeal cancer progression, derived from primary tonsil epithelial cells using the organotypic raft culture system. The raft culture system mimics essential morphological and physiological elements of epithelial differentiation [76]. To date, the organotypic raft culture system remains as the sole in vitro method proven to reliably mimic epithelial differentiation to the degree that enables the complete study of the HPV16 life cycle and production of infectious virions [26,77,78,79]. Previous reports have demonstrated that cell lines derived from human tonsil samples, when grown in the organotypic raft system, generate tissues that closely resemble their in vivo counterparts in both morphology and HPV16 biosynthesis [5,71,77,80,81,82]. Furthermore, both our group and others have demonstrated that progression to cancer can be observed using the organotypic raft culture system [35,36,38,39,40,41,42,43,44]. The ability to produce human epithelial tissue in vitro that supports the full HPV life cycle creates an experimental system for examining the mechanism by which HPV16 drives the development of oropharyngeal squamous cell carcinoma (OPSCC) in tonsillar epithelium. Among all cancers, the incidence rates of HPV-associated oropharyngeal cancers continue to rise, with HPV⁺ OPSCC now taking over as the prominent form of this cancer and overtaking cervical cancer as the most common site of HPV cancers in the US [4,5]. As rates of HPV-associated OPSCC continue to rise with no validated methods to screen for early and intermediate stages of the cancer, it is of great importance to study and comprehend these malignancies.

We report the establishment of primary tonsil keratinocyte cell lines that were transfected with HPV16 DNA via electroporation. Consistent with prior reports, we found these cell lines grown in the organotypic raft system to exhibit morphological features similar to in vivo epithelial tissues [70,71,80]. When allowed to stratify and differentiate in the raft culture system, the established HPV16 cell lines successfully underwent virion morphogenesis, as indicated by the titer results (Figure 2, Panels B and C). These results are consistent with previous tonsil, cervical and foreskin HPV16 and 18 cell lines already established in our lab [71,81,82]. We observe a consistent decrease in HPV16 titers as passaging progressed in both male and female cell lines. However, an exception was noted in Male 19 (M19) raft cultures, which showed an inconsistent decline in titers. This variation is likely attributable to the integration status of the viral genome in this cell line, which will be discussed further ahead. The productive HPV life cycle is linked to the terminal differentiation of keratinocytes in stratified epithelial tissue [83]. However, persistent high-risk HPV viral infection is associated with dysregulation of the HPV oncogenes, prolonging their pro-proliferative effects on the host cells, leading to increased genomic instability and eventual integration of the viral episomes into the host genome [80,81]. This integration results in cellular dysplasia and cancer development. Importantly, integration and carcinogenesis are not part of the natural viral life cycle but instead represents a “dead end” for the virus, signifying a non-productive state due to loss of cellular differentiation and episomal viral DNA, both of which are required for productive HPV replication [84,85,86,87]. Therefore, the observed reduction in viral titers suggests an impairment of the HPV16 life cycle in our developed model, likely due to viral DNA integration into the host genome, as supported by the E2 (episomal)/E6 (total copies) ratio data from our previous publication [34]. The measurement of E2 and E6 protein in HPV cell lines is used as an indicator of the HPV genomic state. Our tonsil cell lines had E2/E6 ratios <1, and continued to decline with successive passaging, implying that the viral genomes were transitioning to a fully integrated state in the established tonsil cell lines. By passage 40 (P40), we observed near-complete depletion of viral titers and E2/E6 ratios approaching zero, with the exception of the M19 cell line. Revisiting the M19 titer data, we noted a consistent E2/E6 ratio close to 1.0, suggesting that most viral genomes remained episomal, consistent with the line’s sustained titer data. Further genomic analysis of the same HPV16-transfected tonsil keratinocyte lines confirmed viral integration in five of the six lines, while the sixth line (M19) retaining an episomal state. Genomic instability was noted in early passages across the lines, but subsided following integration, except in M19 where instability persisted into later passages [34]. Although all the cell lines accumulated structural variants at comparable rates prior to integration, four out of the five lines that underwent integration showed a marked reduction in the rate of structural variant accumulation afterward [34]. The decline in structural variant formation occurs despite continued E6/E7 expression, as integration eliminates the episomal replication stress that drives structural variant formation. Once integrated, the cells transition to a more stable state characterized by constitutive, but much more consistent E6/E7 expression [88,89,90,91,92]. Interestingly, the increased transition to an integrated state indicated by our genomic analysis data, along with concurrent abnormalities contribute to the reduced differentiation capacity observed with continuous passaging, provides strong evidence supporting a role of HPV viral integration in the progression of HPV-associated tumorigenesis [93,94]. These findings underscore the utility of our model in capturing key molecular events associated with malignancy advancement. Collectively, these observations suggest that the developed in vitro model faithfully recapitulates key features of a persistent high-risk HPV16 infection, including progressive viral genome integration and life cycle disruption, as observed in vivo during epithelial carcinogenesis.

Our data show epithelial stratification and differentiation were impaired with successive passaging in both male and female HPV16-transfected tonsil raft tissues (Figure 3). HPV16 is capable of immortalizing primary human epithelial cells, and prior studies have confirmed that transfection of HPV16 into keratinocyte cell lines alters differentiation morphology when stratified in the organotypic raft system, compared to primary keratinocytes [38,43]. Our HPV16-transfected tonsil keratinocytes grown in organotypic raft cultures at low passages exhibit epithelial morphology consistent with low-grade precancerous lesions, whereas higher passages display cancer-like lesions. This progression supports the involvement of secondary cellular events in driving the acquisition of this morphology [38,40,43]. The development of parakeratotic bodies and koilocytes are consistent with our previous studies investigating the morphology of HPV16 cell lines grown in raft cultures [81]. Additionally, the observed histological abnormalities in our tonsil raft tissues are comparable to in vivo abnormal anogenital and oropharyngeal intraepithelial neoplasia development [43,95,96]. Normal epithelia architecture consists of defined epithelial layers with proliferative cells restricted to the basal layer; however, our HPV16 tonsil raft tissue structure is attenuated with continuous passaging. Consistent with in vivo neoplasia tissue, our tonsil raft cultures show koilocytosis, parakeratosis, undergo abnormal differentiation, and show a presence of nucleated basal-like cells across the tissue—features consistent with in vivo HPV16-associated transformation and studies we have previously done [35,72,73,74,97,98].

Upon stratification in the organotypic raft system, our HPV16 tonsil keratinocyte cell lines exhibit loss of differentiation both morphologically (Figure 3) and biochemically (Table 1 and Figure 4, Figure 5 and Figure 6) when stratified in the organotypic raft system. The expression patterns of specific keratins depend on tissue type, the state of differentiation or development, and pathologic conditions [99,100]. When analyzing the differentiation marker profile, we observe the dysregulation of a variety of keratin markers in the tonsil rafts with successive passaging. Prior studies have supported that keratins associated with differentiation are variably expressed during tissue injury, the re-epithelialization process, and transformation [54,55,99,100,101,102,103]. It is important to recognize that keratin expression in oral epithelium differs from that observed in cervical and foreskin epithelia [87,88]. K5 is usually expressed in the mitotically active basal layer of oropharyngeal epithelia, and its expression is attenuated as differentiation progresses [52,53,104]. In our present study, the expression pattern of K5 is altered and downregulated in the raft tissues of our continuously passaged tonsil cell lines. Similar K5 expression patterns were observed by Vasca et al. in their head and neck squamocellular carcinoma studies [105]. K10 is a terminal differentiation marker (typically absent in oral epithelia) expressed in the suprabasal layer of keratinizing epithelia. As epithelial cells undergo terminal differentiation and keratinization, the expression pattern of K10 increases [53,54]. Our results show an elevated expression pattern of K10 in the tonsil rafts with continuous passaging. Consistent with previous reports, the K10 expression pattern observed in our in vitro model closely resemble that of native epithelial tissue undergoing transformation [103,106]. K6 serves as a wound-responsive marker, typically expressed in the suprabasal layer of oral epithelia, and is upregulated following epithelial injury [55]. Our in vitro model demonstrates increased K6 expression in raft tissues with continuous passaging, mirroring the expression patterns observed in in vivo OPSCC samples [107,108]. Notably, in primary untransfected raft tissues, we observe K6 expression in the suprabasal layer. Its presence in the untransfected raft cultures reflects the mechanically stimulated, regenerating, and proliferative state of the tissue in this model, consistent with the way organotypic raft cultures mimic a repairing, hyperproliferative stratified epithelium [83]. Organotypic tissues from our immortalized HPV16-transfected tonsil keratinocytes became more altered with successive passaging. These phenotypic and biochemical changes are consistent with the concept that differentiation and malignancy are inversely correlated [109]. Overall, our findings reveal that the biochemical expression patterns in our in vitro oral cancer model closely mimics those observed in vivo, reinforcing its relevance for studying HPV-associated epithelial transformation.

We have demonstrated that our in vitro model expresses carcinogenic and metabolic markers associated with OPSCC development, consistent with observations made in vivo. Upon continuous passaging, we observed increased levels of the carcinogenic markers K8 and p16 in later passages (Figure 7 and Figure 8), whereas the oncoprotein surrogate markers p53 and pRb decreased following integration (Figure 9 and Figure 10), with expression levels comparable to those seen in primary tonsil raft tissue. These findings align with our previously published E2/E6 ratio data and previous integration studies [34], which collectively show a decrease in episomal HPV16 genomes with successive passaging. This shift reflects viral integration, dysregulation of E6 and E7 oncoprotein expression, subsequent upregulation of K8 and p16, and the reduction or loss of p53 and pRb. Notably, our K8 expression results mirror those seen in patient samples [110,111]. While p16 expression has historically been controversial in the context of OPSCC due to variability among patient samples, recent studies support its role as a biomarker for HPV-associated OPSCC in vivo [112,113,114]. In our cell line collection, p16 presence increased with continuous passaging, particularly in passages identified as having undergone integration based on our E2/E6 ratio analysis. Interestingly, both p16 and K8 showed higher levels in the later passage raft cultures derived from the male HPV16-transfected tonsil cell lines compared to the female lines. This observation is noteworthy given that HPV-associated OPSCC incidence is higher in males than in females. Regarding the oncoprotein surrogate markers, we observed an accumulation of p53 and pRb in passages leading up to integration, followed by a marked loss of both markers in passages confirmed to have undergone integration. This pattern is expected, as integration disrupts E2-mediated regulation of E6 and E7, resulting in enhanced targeting of p53 and pRb for degradation and functional inactivation of these tumor-suppressive pathways. This outcome mirrors what has been observed previously in patient samples [115]. Additionally, the atypical expression pattern of p120 in our raft tissues marks loss of cell adhesion, coinciding with previously published patient sample results [53]. Finally, our COX5B data reveal a metabolically active phenotype in HPV16 raft tissues, consistent with its in vivo counterparts. This suggests an increase in respiratory metabolism with continuous passaging, reflective of changes typically observed during OPSCC carcinogenic progression [69]. These observations indicate a progressive shift toward OPSCC-associated phenotypes in our in vitro model, mirroring those observed in vivo.

We recognize that there are limitations to the in vitro model. The organotypic raft culture system successfully mimics epithelial stratification; however, it does not fully capture the complexity of native tissue. Key components such as immune cells, vasculature, and broader stromal interactions are absent, all of which play important roles in shaping epithelial behavior in vivo. The organotypic cultures have a maximum culture period of approximately 30 days, making it difficult to study chronic or long-term processes. However, the development and validation of this model represent a significant step forward for oropharyngeal cancer research. Currently, there are no validated systems in the context of the complete viral life-cycle that allow investigators to study the early stages and carcinogenic progression of HPV-associated OPSCC in a manipulable format that also yields sufficient material for downstream analyses. This model not only enables examination of the complete HPV16 life cycle but also provides a platform to investigate the molecular mechanisms of HPV integration and the earliest events in OPSCC development, with the goal of identifying biomarkers of early disease. This is particularly important given that most patients present with advanced-stage tumors and no clinical history of pre-malignancy.

5. Conclusions

In summary, we present a novel in vitro model employing the organotypic raft culture system to investigate HPV16-associated OPSCC development using human tonsil tissue. Our findings demonstrate a productive HPV16 life cycle and progressive viral genome integration into the host cells. We also confirm expression patterns of differentiation, carcinogenic, and metabolic markers consistent with those observed in vivo. Collectively, these results validate the utility of this model for studying HPV16-associated OPSCC progression in an in vitro setting. This model provides a valuable opportunity to examine the HPV16 life cycle in the oropharynx and investigate carcinogenic progression, with the goal of identifying potential early and intermediate OPSCC biomarkers and therapeutic targets.