Induction and characterization of oral submucous fibrosis model with different pathological stages in rats and mice

Abstract

Oral submucous fibrosis (OSF) is a chronic and progressive fibrosis disease. Although its pathological stages are well characterized in clinical settings, corresponding animal models remain lacking, which has significantly hindered in-depth mechanistic studies and the development of targeted interventions. Herein, we developed new methods in creating OSF models in rats and mice by different injection frequency of bleomycin (BLM) to simulate early and middle stages of fibrosis. The results showed both low-frequency (once a week) and high-frequency (twice a week) of injection can cause significant fibrosis characteristics, including mucosal pallor, limitation of mouth opening, collagen deposition and inflammatory response. Histological and molecular analyses confirm the stage-dependent pathological changes. Furthermore, to validate the pathological relevance of the model, human oral mucosal tissues from OSF patients and healthy individuals were also analyzed. The result exhibits that, compared with low-frequency injections, high-frequency injections of BLM can lead to more serious fibrosis and inflammatory responses in rats and mice, which are corresponding with the early and middle characteristics of human OSF. This work developed stable and repeatable OSF models of rats and mice in different pathological stages, which offer valuable tools for mechanistic studies of OSF and further precise stage-specific therapies.

Introduction

Oral submucous fibrosis (OSF) is a long-lasting and progressive oral mucosa disease, which is mainly prevalent in the areas with areca nut (AN) chewing habits popularity. Globally, there are about 600 million people have the habit of chewing AN, and about 5% of them suffer from OSF^1,2. Moreover, the World Health Organization (WHO) has listed OSF as a kind of oral potentially malignant disorder for the unreversible pathological states. The pain and discomfort caused by this disease can lead to serious oral dysfunction, including dysphagia, limited mouth opening, even facial deformities, significantly damaging both the physical and psychological well-being of patients. In addition to AN, there are also other factors implicated in OSF, such as chemical and physical irritation, genetic susceptibility and metabolic disturbance³. Although it is well studied that the occurrence and development of OSF can be influenced by many factors, the exact pathogenesis of OSF is unclear. What’s more, current treatments can only alleviate the patients’ subjective symptoms to some extent, but the underlying pathological changes cannot be halted or reversed^{4, 5–6}. The pathological features of OSF are mainly collagen deposition, inflammatory cell infiltration, epithelial atrophy, vascular changes³. Even though no standardized animal model can summarize all the characteristics of human OSF, the pathological manifestations of OSF can partially reproduced, which greatly hinder the development of potentially effective treatment drugs. As the method of constructing an animal model is the groundwork for consecutive research, evaluating these research results effectively is challenging due to the lack of standardized OSF animal methods and the variabilities in model quality. Hence, developing an animal model that accurately reflects the typical characteristics of human OSF remains a key challenge.

Currently, various methods are available for constructing OSF models in vivo, in which the AN induction is the most common^{7, 8–9}. While this method is characterized by a prolonged induction period, it also suffers from a lack of standardized dosage and timing. Furthermore, the tolerance of animals to AN component varies significantly from that of humans, resulting in poor reproducibility of human OSF characteristics. Mechanical stimulation serves as another common approach, but the complex process and low success rate with only local trauma led to disappointing results. Bleomycin (BLM) is a chemical agent widely used in inducing fibrosis disease models, including pulmonary fibrosis, skin fibrosis and liver fibrosis^{10, 11–12}. The advantages of BLM are mainly reflected in the following aspects: (1) efficient induction of fibrosis; (2) excellent controllability and high experimental reproducibility; (3) the induced fibrosis closely resembles the pathological characteristics observed in human fibrosis. (4) BLM is extensively utilized across various animal models. The advantages mentioned above rendered it highly suitable for constructing OSF models to explore its underlying mechanisms and potential therapeutic strategies. Although previous study by Zhang et al. have confirmed the effectiveness of using BLM to construct OSF characteristics in Sprague-Dawley (SD) rats¹³. However, the model is relatively simple, requiring a long induction period, only exploring phenotypic changes in collagen deposition, and failing in reflecting the staged progression of the disease. What’s more, lacking mouse OSF models makes it difficult to investigate the molecular pathogenesis of OSF. Therefore, it is necessary to establish a more comprehensive and refined BLM-induced OSF animal model.

According to the current clinical guideline, the pathological and clinical features of OSF vary across different stages of disease progression, necessitating stage-specific treatment strategies¹⁴. Notably, when OSF progresses to an advanced stage, treatment outcomes are almost disappointing. It suggests that developing OSF animal models at different pathological stages holds significant value for guiding clinical prevention and treatment. Overall, this study aims to establish rat and mouse models of OSF at distinct pathological stages through administering of BLM via local injections at varying frequencies. The degree of fibrosis and local immune changes induced will be assessed by using H&E, Masson’s trichrome stain, Immunohistochemistry, quantitative real-time Polymerase Chain Reaction (RT-qPCR), and Scanning electron microscopy (SEM). This comprehensive evaluation will help determine assess the reliability of the model and provide valuable insights into the pathogenesis of OSF, thereby contributing to the development of improved strategies for clinical prevention and treatment.

Materials and methods

Animals

Male SD rats (7 weeks old) and male C57BL/6j mice (7 weeks old) were obtained from Hunan SLK Jingda Laboratory Company (Changsha, China). All animals were segregated and housed in a specific pathogen-free environment at a temperature of 25℃ and a humidity 40%-70%, with 12 h light/dark cycle. All animals were allowed to acclimate to this environment with free access to standard fodder and distilled drinking water for at least 1 week before the experiment. Twenty rats and twenty mice were randomly divided into four groups (n = 5 per group for each species): (1) Control-W group, submucosal injection of 100µL 0.9% saline once weekly; (2) OSF-W group, submucosal injection of 100µL BLM (Japan Chemical Corporation, Tokyo, Japan) once weekly; (3) Control-D group, submucosal injection of 100µL 0.9% saline twice weekly; (4) OSF-D group, submucosal injection of 100µL BLM twice weekly. All injections were administered for a total period of four weeks. All animals were anesthetized with 1% pentobarbital (40 mg/kg) by intraperitoneal injection in each procedure, and the deep anesthesia was confirmed throughout the operation. All methods are reported in accordance with Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines. The whole experimental procedures were performed in accordance with relevant guidelines and regulations, and all the experimental procedures were approved by the Institutional Animal Care and Use Committees of Hunan Cancer Hospital (KNZY-202216;KNZY-202305) and every effort has been made to reduce the suffering.

Oral submucous fibrosis model induced

We induced OSF animal models in SD rats and C57BL/6j mice by oral submucosal injection of BLM according to the schematic view to simulate different stages of OSF, and the timeline is shown (Fig. 1A). BLM was dissolved in 0.9% saline at a concentration of 1 mg/mL. For rats, we used a 29G needle aspirate 100µL BLM for local injection, positioning the needle approximately 2 mm from the unilateral buccal mucosa (Fig. 1B), and the dosage is based on a previously established OSF rat model¹³. For mice, given the small size of the mouse oral cavity and the difficulty in exposing the mucosa, we used a custom-made mouth gag to improve visibility and facilitate access to the oral mucosa, using a 34G needle, 100µL of BLM at the same concentration was locally injected into the position, which is between the mucosal ligaments. (Figure 1C and E). After four weeks of experiment, all animals were euthanized by CO₂ and their oral submucosal tissues were collected.

Fig. 1 [Images not available. See PDF.]

OSF models induced by BLM in rats and mice. (A) Schematic overview of OSF model induced by BLM in SD rats and C57BL/6j mice. (B) Schematic diagram of anatomical injection site of oral mucosa in rats. (C) Schematic diagram of anatomical injection site of oral mucosa in mice. (D) Schematic diagram of rat mouth opening measurement. (E) A custom-made mouth gag was used in this study to expose the visual field of oral mucosa in mice.

Experimental observation and measurement of weight, mouth opening and mucosal lesion area

The vitality of all the rats and mice was observed daily during the whole experiment. During the modeling period, body weight was measured weekly. After intraperitoneal anesthesia with 1% pentobarbital, the mouth opening of rats was measured by fixing the upper incisors and applying a force of 1.9 N to the lower incisors using a force gauge, and the distance between the upper and lower incisors represents the mouth opening degree (Fig. 1D). The changes of the oral mucosa were photographed after 4 weeks of treatment, and the lesion area was measured and quantified by the ImageJ software.

Patients and tissue samples

Oral mucosal tissues from healthy and OSF patients are collected to compare the modeling characteristics of OSF in rats and mice. A total of 6 tissue samples were obtained from patients with OSF, who underwent surgical resection at Xiangya Stomatological Hospital between 2022 and 2023, including 3 early OSF samples and 3 middle OSF samples. All the OSF patients were diagnosed by professional pathologists. Patients who had received prior treatment for oral mucosal lesions or system diseases were excluded. All of 3 normal human oral mucosa samples were obtained when the lower third molars were surgically removed. The inclusion criteria for healthy samples included: no history of areca nut chewing, no systemic diseases, no oral mucosal lesions, and no periodontal disease. Buccal mucosal biopsies were taken from the OSF lesion area in patients and the corresponding buccal area in healthy individuals. All OSF patients had a clear history of areca nut chewing. The collected tissues of animals and human were fixed in 10% neutral buffered formalin or at -80℃ for subsequent experiments.

The study was approved by the Ethics Committee of Xiangya Stomatological Hospital of Central South University (201703104) and informed consent form was signed by all participants. All experiments were performed in accordance with relevant guidelines and regulations.

OSF tissues collection

For animals, after euthanasia, they were placed in the supine position. Using tissue scissors cut along the midline between the mandibular central incisors to separate mandibular bones, thus completely exposing the buccal mucosa of both sides. Oral mucosal tissue samples were collected with an 11# surgical scalpel in area of local injection. For rats, samples with a size of about 0.4 cm × 0.6 cm was cut off, while for mice, samples with a size of about 0.2 cm × 0.4 cm were collected. The depth of biopsy included the whole thickness of the mucosa, extending down to but not including the underlying muscle layer. The collected tissues of animals and human were fixed in 10% neutral buffered formalin or at -80 °C for subsequent experiments.

Histopathologic examination

The oral mucosa tissue from rats and mice and human OSF mucosa were fixed in 10% neutral buffered formalin for 24–48 h, and dehydrated with gradient ethanol, clearing with xylene, then embedded in paraffin. The paraffin-embedded mucosa samples were sliced with a thickness of 4 μm. The tissue sections were placed on glass slides, dewaxed in xylene and a series of ethanol baths, and then HE staining was performed to observe pathological changes; Masson’s trichrome stain was performed to determine the area and degree of collagen deposition. Image analysis of the Masson’s trichrome stained slides was conducted using ImageJ software.

Immunohistochemical staining

Immunohistochemistry staining was employed to assess the expression of fibrosis and inflammation related markers. All the collected paraffin-embedded mucosa samples were deparaffinized, and rehydrated, followed by Immunohistochemistry staining protocol. Subsequently, slides were incubated within 3% hydrogen peroxide for 15 min to block endogenous peroxidase activity and antigen retrieval was performed by microwave oven. Next, the slides were blocked in 5% goat serum for 15 min at 37℃, Slides were incubated overnight at 4℃ with the following primary antibodies: α-Smooth muscle actin (α-SMA) polyclonal antibody (1:1000; Proteintech, Wuhan, China), CD68 polyclonal antibody (1:1000; Proteintech, Wuhan, China), F4/80 polyclonal antibody (1:1000; Proteintech, Wuhan, China). Myeloperoxidase (MPO) (1:500; Abcam, UK). α-SMA was a well-established marker of myofibroblasts (MFBs), which are key effector cells involved in the development and progression of tissue fibrosis. CD68 was used to evaluate the infiltration of macrophages in rats and human, and F4/80 was used as the marker for macrophages in mice. MPO is a classic marker of neutrophils, which serves as a sensitive index to detect neutrophil infiltration. Finally, slides were observed under Leica light microscope. The average positive areas of the α-SMA, CD68, F4/80, MPO were quantified using the ImageJ software.

RT-qPCR

All collected mucosa tissue was taken and ground under liquid nitrogen freezing, and then transferred to a homogenization tube, then the total RNAs was extracted by TRIzol reagent (Invitrogen, USA). NanoDrop (Thermo Fisher, USA) was used to detect RNA concentrations. According to the manufacturer’s instructions for the cDNA synthesis kit (Vazyme, China). PCR amplification was performed to detect the transcript levels of genes by using the real-time PCR system (Applied Biosystems, United States). GAPDH or Gapdh were used as the reference gene and calculated by 2^−ΔΔCT method. The primer sequences of related genes are shown (Table S1).

SEM characterization

All the collected paraffin-embedded mucosa samples were deparaffinized and rehydrated, and these unstained oral mucosa histological thin sections were coated with 10 nm of carbon and imaged with a FIB scanning electron microscope (TESCAN-AMBER) with an accelerating voltage of 20 kV. Using SEM to observe histological sections is useful to visualize the characterization of collagen fibers.

Statistical analysis

In this study, all results were expressed as mean ± standard deviation. GraphPad Prism 9.0 was employed for data analysis. Normality of distribution was assessed using the Shapiro-Wilk test, and homogeneity of variance was tested with Levene’s test. The one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparison post hoc test was used for the comparison of multiple groups. Two-way ANOVA followed by Tukey’s test was used for the analysis of body weight changes and mouth opening among different treatment groups. P < 0.05 was regarded as statistically significant and the asterisk represent the degree of difference (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001). All the staining micrographs were obtained by Leica microscopes, and image analysis of the stained slides and lesion area were conducted using ImageJ software (v1.8.0; https://imagej.net/ij/download.html).

Results

General observations of rats and mice induced by BLM

We find that after 4 weeks of post-BLM injection, rats of OSF-W group (once a week) exhibited slight whitening lesion of the oral mucosa without any fibrous bands (P = 0.011). In contrast, comparing to the OSF-W group, rats of OSF-D (twice a week) group showed obvious pale and rigid changes on the oral mucosa, with fibrous band discovered (P = 0.0009). What’s more, there is no significant changes was found in the both of two control groups (P = 0.99; Fig. 2A and B). During the entire experiment period, no evident body weight loss was observed in any of the rat groups. At the time of 4 weeks, the width of mouth opening of both two experimental groups was significantly decreased compared to the control group (P = 0.0022, P = 0.0256 and P = 0.0013; Fig. 2E). The average body weight of OSF-W group was 456.7 ± 34.5 g, and that of OSF-D group was 447.3 ± 18.3 g, In comparison, the mean weight of Control-W and Control-D group were 486.1 ± 22.0 g and 478.0 ± 9.7 g. respectively (all P > 0.05; Fig. 2F).

Fig. 2 [Images not available. See PDF.]

OSF models are induced by BLM in rats and mice. (A) Representative images of the oral mucosa of rat treatment with BLM once a week and twice a week. Lesion areas are outlined in red. (B) Quantitative analysis of the lesion area in rats. (C) Representative images of the oral mucosa of mice treatment with BLM once a week and twice a week. Lesion areas are outlined in red. (D) Quantitative analysis of the lesion area in mice. (E) Mouth opening measurements in rat treatment with BLM for 4 weeks. (F) The changes of body weight in rat treatment with BLM for 4 weeks. (G) The changes of body weight in mice treatment with BLM for 4 weeks (n = 5 per group). Data are presented as means ± SD (n = 5 per group). Data are presented as means ± SD. *P < 0.05, **P < 0.01, ****P < 0.0001, ns, not significant.

By the end of week 4, both OSF-D and OSF-W groups of mice exhibited noticeable pallor in oral mucosa (P < 0.0001 and P = 0.0002, respectively). However, the whitening lesion area and fibrous band were more obvious in the OSF-D group (P = 0.0005). Besides, the texture of oral mucosa in OSF-D group became significant stiff (Fig. 2C and D). Meantime, both experimental groups showed signs of restricted mouth opening. In the latter 2 weeks of the model establishment, it became increasingly difficult to insert the injector and observe the status of oral mucosa, due to the progressive narrowing of the oral cavity of the test mice. Throughout the entire experiment period, both of two OSF groups showed a tendency toward body weight loss in the last two weeks, but the differences were not statistically significant. The average body weight of OSF-W group were 23.3 ± 0.4 g, and 22.8 ± 1.2 g for the OSF-D group, while the mean weight of Control-W and Control-D group were 24.2 ± 0.9 g and 23.9 ± 0.5 g, respectively (all P > 0.05; Fig. 2G).

H&E staining of rats and mice induced by BLM

We find that rats in both OSF-W and OSF-D groups showed abnormality of different degree in epithelial layer, submucosal inflammatory cell infiltration, blood vessels and collagen deposition. After 4 weeks of weekly BLM treatment, most test rats exhibited thinner and more atrophic mucosal epithelium, with the epithelial rete ridges appearing shorter and flatter. There were a few infiltrated inflammatory cells can be observed in lamina propria, depositing collagen fibers beneath the lamina propria. In contrast, the OSF-D group, showed broader collagen deposition within the lamina propria, much more atrophic epithelial layer, almost disappearance of epithelial rete ridges, increased inflammatory cells infiltration and partial vascular occlusion (Fig. 3A).

Fig. 3 [Images not available. See PDF.]

Oral mucosa H&E staining of rats, mice and human. (A) Representative oral mucosa H&E stained images of SD rats in Control-W group, OSF-W group, Control-D group and OSF-D group (n = 5 per group). (B) Representative oral mucosa H&E-stained images of C57BL/6j mice in Control-W group, OSF-W group, Control-D group and OSF-D group (n = 5 per group). (C) Representative H&E-stained images of oral mucosal tissues from healthy and OSF patients. Scale bar, 100 μm.

H&E staining results showed mice in the OSF-D group exhibited pathological features consistent with middle-stage of OSF lesions. After 4 weeks of induction, the oral mucosa epithelium appeared atrophic and thinner, with flattened and reduced epithelial rete ridges. There were also a large number of pink filament deposition can be observed in the lamina propria. In addition, vascular density was decreased, with a little inflammatory cells infiltration and thickening of the lamina propria observed. These characters are similar to that of OSF patients, suggesting that BLM can promote the deposition of collagen in oral submucosa of mice. In contrast, the symptoms of atrophy of oral mucosal epithelium in the OSF-W group is relatively slighter than in the Control-W group. Specifically, in the OSF-W group, epithelial rete pegs still partially preserved, and a little pink filament deposition, immune cells infiltration and the increased number of vessels can be seen in the lamina propria, resembling early-stage OSF changes in patients (Fig. 3B and C). Furthermore, histopathological evaluations showed that there are no significant alterations in organs such as lung, liver, heart, kidney and tongue (Figs. S1 and S2), confirming that submucosal injection of BLM would only lead to local fibrosis, rarely led to other organ damage in both rats and mice.

Masson’s trichrome stain of rats and mice induced by BLM

Collagen deposition is a hallmark feature of OSF. We further analyzed collagen deposition in OSF models and found that the collagen change of rats in the OSF-W group were similar to those of human mucosa, especially with increased collagen deposition in the lamina propria, disarranged and significantly thickened compared with that of the Control-W group (P = 0.0001). Meanwhile, in the OSF-D group of rats, the area of the lamina propria of oral mucosa expanded, and the collagen fibers deposited were thick and densely arranged into collagen fiber bands of varying thicknesses (P < 0.0001; Fig. 4A and D).

Fig. 4 [Images not available. See PDF.]

Oral mucosa Masson’s trichrome stain of rats, mice and human. (A) Representative oral mucosa Masson’s trichrome stain images of SD rats in Control-W group, OSF-W group, Control-D group and OSF-D group (n = 5 per group). Collagen fibers are stained blue. (B) Representative oral mucosa Masson’s trichrome stain images of C57BL/6j mice in Control-W group, OSF-W group, Control-D group and OSF-D group (n = 5 per group). Collagen fibers are stained blue. (C) Representative Masson’s trichrome stain images of oral mucosal tissues from healthy and OSF patients. Collagen fibers are stained blue. (D) Quantitative analysis of Masson’s trichrome stain in oral mucosa tissue of rats. (E) Quantitative analysis of Masson’s trichrome stain in oral mucosa tissue of mice. (F) Quantitative analysis of Masson’s trichrome stain in oral mucosa tissue of human. Scale bar, 100 μm and 50 μm.

The oral mucosa of mice in the OSF-W group showed a modest presence of blue-violet stained collagen fibers in the lamina propria (P = 0.0005). These collagen fibers appeared thin, straight, and loosely organized. The terminal branches of these collagen bundles intertwined beneath the epithelial basal cells, forming a network of collagen bundles. Notably, the dense collagen fibers in the OSF-D group exhibited markedly stronger and wider collagen staining area than the OSF-W group (P = 0.0055; Fig. 4B and E).

In order to verify the reliability of animal models, we further compared it with human OSF tissue samples. The results demonstrated that OSF models have exhibited a high degree of similarity to clinical OSF pathological characteristics, supporting the validity of models. In normal oral mucosa, the content of collagen fibers in connective tissue is low, arranged loosely and orderly, and light blue in Masson’s trichrome stain. In the early stage of OSF, collagen fibers gradually increased, became more densely and irregularly arranged, and show deeper staining (P < 0.0001), accompanied by mild epithelial atrophy or proliferation. In the middle stage, collagen fibers thickened further and deeply stained (P = 0.0004), tightly packed, and disrupt normal tissue architecture. The above characteristics have been well reproduced in OSF models of this study, which proves that the model can effectively simulate the histological changes in different pathological stages of OSF (Fig. 4C and F).

Fibrosis levels of rats and mice induced by BLM

In order to further elucidate the pro-fibrotic effect of BLM in vivo, the evaluation of collagen accumulation via immunohistochemistry and RT-qPCR were conducted. Immunohistochemical analysis of α-SMA was performed on the lamina propria of oral mucosa in rats and mice to evaluate the numbers of MFB, which is the main collagen source¹⁵. In rats, the expression of α-SMA in the mucosa lamina propria was slightly increased in the OSF-W group compared to the Control group (P = 0.024), whereas OSF-D group exhibited significantly stronger α-SMA expression (P < 0.0001 and P = 0.0002; Fig. 5A and D). Similarly, in the mice models, we also found that increased expression of α-SMA in the OSF-W group compared to the Control-W group (P = 0.0007). Notably, the OSF-D group displayed a significantly higher percentage of α-SMA–positive cells than the OSF-W group (P < 0.0001 and P = 0.0023; Fig. 5B and E). Additionally, oral mucosal tissues from OSF patients were analyzed. Immunohistochemistry results showed that the expression level of α-SMA was markedly upregulated in the middle-stage OSF samples of human compared to early-stage OSF samples of human. (P = 0.017; Fig. 5C and F).

Fig. 5 [Images not available. See PDF.]

Fibrosis level in OSF rats, mice and human samples. (A) Representative immunohistochemical images of α-SMA in oral mucosa tissue of SD rats in Control-W group, OSF-W group, Control-D group and OSF-D group (n = 5 per group). The red arrow represents α-SMA positive cells. (B) Immunohistochemical quantitative analysis of α-SMA expression in oral mucosa tissue of SD rats. (C) Representative immunohistochemical images of α-SMA in oral mucosa tissue of C57BL/6j mice in Control-W group, OSF-W group, Control-D group and OSF-D group (n = 5 per group). The red arrow represents SMA positive cells. (D) Immunohistochemical quantitative analysis of α-SMA expression in oral mucosa tissue of C57BL/6j mice. (E) Representative immunohistochemical images of α-SMA in oral mucosa tissue of human. The red arrow represents α-SMA positive cells. (F) Immunohistochemical quantitative analysis of α-SMA expression in oral mucosa tissue from healthy and OSF patients. (G) mRNA levels of fibrosis-related genes, such as Acta2, Col1a1 and Fn1 were determined by RT-qPCR in oral mucosa tissue of SD rats. (H) mRNA levels of Acta2, Col1a1 and Fn1 were determined by RT-qPCR in oral mucosa tissue of C57BL/6j mice. (I) mRNA level of ACTA2 were determined by RT-qPCR in oral mucosa tissue from healthy and OSF patients (n = 5 per group). Data are mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. Scale bar, 20 μm.

Furthermore, the mRNA expression level of fibrosis-related genes in oral mucosa of both rats and mice were evaluated by RT-qPCR. The results in rats showed that compared to the control group, the expression level in OSF-W of fibrosis-related genes, especially Acta2 mRNA and Col1a1 mRNA, increased (P = 0.03 and P = 0.049), but lower than that in OSF- D group (P = 0.0002 and P = 0.0004; Fig. 5G). As well, the expression level of Fn1 mRNA was increased slightly in OSF-D (P = 0.047). A similar pattern was observed in mice, Acta2 mRNA and Col1a1 mRNA were significantly upregulated in the OSF-W group compared to control group (P = 0.046 and P = 0.018, respectively), and further increased in the OSF-D group compared to OSF-W (P = 0.0001 and P = 0.019, respectively). Fn1 mRNA expression also showed a mild increase in the OSF-D group (P = 0.013; Fig. 5H).

These results indicated that BLM upregulates the expression of α-SMA in correlation with lesion severity in OSF-D group, thus contributing to the progression of OSF. This finding is consistent with the mRNA expression levels in human OSF samples (P = 0.0047 and P = 0.0024; Fig. 5I), suggesting that BLM can effectively promote the formation of MFB in the lamina propria of oral mucosa. Moreover, OSF models findings closely align with those observed in human tissue samples.

Immune changes of OSF rats and mice

Previous studies have shown that inflammation is a key feature in the onset of OSF, marked by the significant upregulation of inflammatory factors and infiltration of inflammatory cells. As studied before, neutrophils and macrophages are the main inflammatory cells involved in OSF development^16,17. Therefore, to explore neutrophils and macrophage involvement during OSF progression, immunohistochemical staining for MPO, CD68 or F4/80 in the oral mucosa of rats and mice was performed. The results reveal evident macrophage infiltration in the OSF tissues of both rats and mice, with more pronounced infiltration in the OSF-D group compared to the OSF-W group (P = 0.012 and P = 0.0006; Fig. 6A-B and G-H). We also found that macrophages infiltration aggravated with the increase frequency of BLM injection. In human samples, the number of CD68⁺ macrophages rose with the progression of OSF (P = 0.001 and P = 0.012; Fig. 6C and I). In addition, the results showed that neutrophils also infiltrated in OSF tissues of rats and mice, however, compared with OSF-D group, the infiltration in OSF-W group of rats and mice was more obvious (P = 0.006 and P = 0.024 Fig. 6D-E and J-K). This result is consistent with the pathological manifestations in human samples, that neutrophil infiltration significantly increased in the early stage of the OSF (P = 0.0005, Fig. 6F and L). The findings indicated that influence of BLM injection frequency in the animal models effectively mimic inflammatory features observed at different pathological stages of human OSF.

Fig. 6 [Images not available. See PDF.]

Alterations in immune cell populations in OSF rats, mice and human samples. (A) Representative immunohistochemical images of CD68⁺ macrophage in oral mucosa tissue of SD rats in Control-W group, OSF-W group, Control-D group and OSF-D group. The red arrow represents CD68 positive cells. (B) Representative immunohistochemical images of F4/80⁺ macrophage in oral mucosa tissue of C57BL/6j mice in Control-W group, OSF-W group, Control-D group and OSF-D group (n = 5 per group). The red arrow represents F4/80 positive cells. (C) Representative immunohistochemical images of CD68⁺ macrophage in oral mucosa tissue of human. The red arrow represents CD68 positive cells. (D) Representative immunohistochemical images of MPO⁺ neutrophils in oral mucosa tissue of SD rats in Control-W group, OSF-W group, Control-D group and OSF-D group. The red arrow represents MPO positive cells. (E) Representative immunohistochemical images of MPO⁺ neutrophils in oral mucosa tissue of C57BL/6j mice in Control-W group, OSF-W group, Control-D group and OSF-D group. The red arrow represents MPO positive cells. (F) Representative immunohistochemical images of MPO⁺ neutrophils in oral mucosa tissue of human in Control-W group, OSF-W group, Control-D group and OSF-D group. The red arrow represents MPO positive cells. (G) Immunohistochemical quantification of CD68 expression in rat model (n = 5 per group). (H) Immunohistochemical quantification of F4/80 expression in mice model (n = 5 per group). (I) Immunohistochemical quantification of CD68 expression in human samples (n = 3 per group). (J) Immunohistochemical quantification of MPO expression in rat model (n = 5 per group). (K) Immunohistochemical quantification of MPO expression in mice model (n = 5 per group). (L) Immunohistochemical quantification of MPO expression in mice model (n = 3 per group). Data are mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Scale bar, 20 μm.

In rats, we observed that the expression levels of TNF-α and IL-6 were significantly elevated OSF-W group compared with the Control-W group (P = 0.0432 and P = 0.0081, Fig. 7A). Furthermore, the expression of IL-1β was markedly increased in the OSF-D group, and the levels of TNF-α, IL-6, and IL-1β were all significantly higher than those in the OSF-W group (P = 0.0066, P = 0.0006 and P = 0.0052, Fig. 7A). In mice, TNF-α and IL-1β were significantly upregulated in the OSF-W group compared to the control group (P = 0.049 and P = 0.026), while IL-6 showed significant upregulation in the OSF-D group (P = 0.0002, P = 0.0007, and P = 0.0039; Fig. 7B). Moreover, the expression levels of TNF-α and IL-1β were significantly higher in the OSF-D group than in the OSF-W group (P = 0.0025 and P = 0.0449, Fig. 7B). These results indicate that the upregulation of inflammatory markers induced by BLM may contribute to the development of OSF and serve as a key driver of fibrosis, which is consistent with the results observed in human OSF samples (Fig. 7C).

Fig. 7 [Images not available. See PDF.]

Changes in inflammatory cytokine expression levels in OSF rats, mice and human samples. (A) Oral mucosa tissue of Tnfa, Il1b and Il6 mRNA were determined by RT-qPCR in oral mucosa tissue of SD rats (n = 5 per group). (B) Oral mucosa tissue of Tnfa, Il1b and Il6 mRNA were determined by RT-qPCR in oral mucosa tissue of C57BL/6j mice (n = 5 per group) (C) Oral mucosa tissue of Tnfa, Il1b and Il6 mRNA were determined by RT-qPCR in oral mucosa tissue of human (n = 3 per group). Data are mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Characterization of collagen fibers of OSF rats and mice

SEM offers an in-depth view of the collagen architecture in the oral mucosa, identifying key structural differences between normal and pathological tissues at various magnifications. Notably, the differences in collagen architecture between normal oral mucosa and OSF tissues are significant¹⁸. In rats, the oral mucosa in the Control-W and Control-D showed a relatively smooth, well-organized network of collagen fibers. These fibers were thin, aligned in parallel bundles, and uniformly spaced, with minimal evidence of entanglement or irregularity, contributing to the overall structural integrity of the tissue. While in the OSF-D group, obvious alternations of collagen morphology are exhibited, thicker and denser fibers with evident bundle fusion. The intermittent space between the fibers was significantly reduced, leading to a more compact and rigid structure. However, the OSF-W group exhibited intermediate changes. Although the fibers became thicker and denser than control groups, they were not completely fused into bundles (Fig. 8A). A similar pattern was observed in mice. SEM images also showed a marked alteration in collagen morphology. In the OSF-D group, the collagen fibers become more disorganized, and more densely packed with bundle fusion, comparing to the Control-D group. In the OSF-W group, collagen fibers also appeared twisted, entangled, and in haphazard orientations, but the level of bundle fusion and cross-linking of collagen fibers was slighter than OSF-D group (Fig. 8B).

Fig. 8 [Images not available. See PDF.]

Characterization of collagen fibers organization in OSF rats, mice and human samples. SEM micrographs of unstained oral mucosa histological sections of SD rats, C57BL/6j mice and OSF patients at high magnification, respectively. (A) SEM micrograph of collagen fiber structure in oral mucosa of SD rats (micrometric). Scale bar, 1 μm and 5 μm. (B) SEM micrograph of collagen fiber structure in oral mucosa of C57BL/6j mice (micrometric). (B) SEM micrograph of collagen fiber structure in oral mucosa of OSF patients (micrometric). Scale bar, 1 μm and 5 μm.

To confirm that the collagen changes in our animal models align with the clinical features of OSF, we examined collagen morphology in human oral mucosa samples from normal tissue, early-stage OSF, and middle-stage OSF (Fig. 8C). In normal mucosa, collagen fibers were sparse, loosely and orderly arranged, and distributed in a reticular or wavy pattern, with smooth surfaces and well-defined structures. In the early stage of OSF, local collagen fibers began to thicken and showed a tendency toward bundle formation, with increasingly disorganized alignment. By the middle stage, the fibers have become markedly thickened and densely packed, forming irregularly fused bundles. The arrangement of fiber was disordered, with surfaces appearing rough, and the overall tissue structure became compact.

Discussion

OSF is a chronic progressive fibrosis disease related to many factors, among which AN chewing is recognized as the most critical risk factor. Thus, using BLM to induce OSF in animal models remains controversial. Although AN plays an important role in the occurrence and development of OSF, there are several limitations in the induction of OSF model in AN extract and arecoline so far. These include a long induction period, lack of standardized protocols, significant interspecies differences in tolerance between animals and humans to AN component, and poor model stability. Previous studies have confirmed that BLM is a commonly used fibrosis inducer, which has been proved effective in inducing OSF in rats¹³. Consistent with the previous findings, our study proved that BLM can effectively induce collagen deposition and simulate the key pathological features of OSF. To emphasize, there have been no reported studies utilizing BLM to induce an OSF model in mice. We conducted a preliminary attempt to apply the same dosage for local injections in mice, based on the mature and feasible scheme in rats (100µL, 1 mg/mL). The results demonstrated that this approach successfully induced fibrosis without causing significant systemic toxicity, which verified the effectiveness and safety of this attempt in mice.

A series studies have demonstrated the pathogenesis of OSF may associated with the abnormal changes in oral local immune microenvironment, such as increased fibrotic cytokine (IL-1β, IL-6, TNF-α, TGF-β), chemokines (CXCL9, CCL2) and infiltration of inflammatory cells (such as T cells, neutrophils, mast cells and macrophages)^{19, 20, 21, 22, 23, 24–25}. A current study by Wang et al. also demonstrated T cells may communicate with epithelial cells to contribute OSF development, suggesting that OSF may be an immune-related disease²⁶. Moreover, we find that the bleomycin-induced organ fibrosis is also achieved by mediating the release of fibrotic cytokines and persistent inflammatory response. For example, in idiopathic pulmonary fibrosis, bleomycin can cause the infiltration of macrophages and T cells, and the secretion of fibrotic cytokines, thus leading to the activation of fibroblasts^27,28. Similarly in the study of systemic sclerosis, bleomycin can induce not only the infiltration of macrophages and T cells, but also the emergence of autoantibodies such as antinuclear antibody (ANA), contributing to further tissue damage and fibrosis^29,30. Besides the reasonable mechanisms, the advantages of BLM are reflected in the following aspects: efficient induction of fibrosis, good controllability and reproducibility, precise control of dosage and administration timing, and close simulation of the pathological characteristics of human fibrosis³¹. From the perspective of immunological pathogenesis and the given attributes, using BLM as an inducer for an in vivo OSF model is reasonable.

An important question remaining consideration is whether the pathological changes in this model accurately reflect the pathological stages observed in OSF patients. In the early stage, key features include collagen edema, vasodilation, fibroblast increase, and inflammatory cell infiltration with a generally normal epithelium. In the middle stage, features such as subepithelial vitreous degeneration, collagen thickening, vasodilation, and chronic inflammatory infiltration appear, with epithelial ridge shortening. As the disease advances to later stages, progressive epithelial dysplasia, thickened collagen bundles, then extensive fibrosis, vascular regression, fibroblast depletion in vitreous areas, and epithelial flattening are observed³². As studied before, OSF models have primarily focused on extending the injection period neglecting the impact of injection frequency^7,13. In this study, we demonstrated that altering the BLM injection frequency effectively affect the characteristics of fibrosis in both rats and mice, closely resembled the progression of early and middle stages of fibrosis in OSF patients. Specifically, H&E and Masson’s trichrome stain revealed distinct mucosal changes in OSF rats and mice induced by different BLM injection frequencies. Higher injection frequencies led to more pronounced epithelial abnormality, increased submucosal collagen deposition, and enhanced inflammatory responses. Furthermore, immunohistochemical staining showed that increased BLM injection frequency resulted in higher levels of α-SMA expression and greater macrophage infiltration, indicating exacerbated fibrotic activity and immune involvement. Consistent with these histopathological findings, RT-qPCR analysis demonstrated that genes related to inflammation and fibrosis were significantly upregulated in both rat and mouse OSF model tissues compared with the control group. Notably, gene expression levels were positively correlated with BLM injection frequency extent, demonstrating more severe disease progression. Collectively, these findings suggest that modifying the BLM injection frequency not only modulates the severity of fibrosis but also reflects the distinct pathological features of OSF at different stages. Therefore, optimizing the BLM injection frequency is essential for developing a more accurate OSF model and exploring the disease’s pathogenic mechanisms across different pathological stages.

In this study, both SD rats and C57BL/6j mice were employed to comprehensively evaluate the effectiveness of BLM in inducing OSF models. Rats, due to their larger size and wider mouth opening, have been the preferred species currently for OSF modeling owing to the ease of handling and injection^33,34. However, OSF rat model can only be used for the research of OSF therapeutic or preventive interventions, limited in studying the pathogenesis of OSF. In contrast, mouse models offer considerable advantages disease mechanistic study because of the high genomic homology with humans³⁵. Nonetheless, constructing an OSF model in the oral cavity of mice is particularly challenging due to the small size of the oral cavity, difficulty in maintaining consistent injection sites, and high mortality rates, which hindered the induction of OSF model in mice. To address the limitations, we used a custom-made mouth gag to facilitate consistent and accurate oral injections in mice. The use of mouth gag reduced variability in surgical operations, ensured more uniform distribution of fibrotic lesions, simplified the procedure, and lowered animal mortality. Hence, it enabled the advanced establishment of a stable and efficient OSF mouse model. Furthermore, in both animal models, this study proved that different frequencies of BLM administration resulted in varying degrees of fibrosis that closely simulated the pathological features of early and middle stages of clinical OSF. Specifically, the low-frequency BLM-induced model (once a week) induced mild fibrosis could serve as a valuable tool for investigating the early pathogenesis and prevention strategies of OSF, while higher-frequency injections (twice a week) for middle stage. These findings underscore the potential of our model and provide a flexible platform for OSF-related research across multiple disease stages.

Conclusion

In this study, OSF models were successfully constructed in rats and mice, and different pathological stages were reproduced by modulating the frequency of BLM administration. These models mirror the pathological progression of clinical OSF, particularly the early and middle characteristics of the disease. Notably, the present study provided a reliable tool for investigating stage-specific mechanisms of OSF and exploring stage-adapted therapeutic approaches.

Acknowledgements

This study was supported by the National Key Research and Development Program of China (2022YFC2402900); National Natural Science Foundation of China (No.82470989; 52103327); The Joint Funds of the Hunan Provincial Natural Science Foundation (2023JJ60509); The Science and Technology Talent Support Project of the Hunan Provincial Science Popularization Special Project (2023TJ-Z08); The Science and Technology Innovation Program of Hunan Province (2024RC3068); The Natural Science Foundation of Hunan Province (2023JJ30813).

Author contributions

Conceptualization, OL and XD; Investigation, JT and ZZ; Visualization, JT, ZZ.; Supervision, GW, JJ; Validation, ZZ; Funding acquisition, XD and OL; Project administration, HT and XC; Writing-Original Draft Preparation, JT and ZZ; Writing—Review & Editing, JT, ZZ. All authors reviewed the manuscript.

Data availability

The datasets generated during the current study are available from the corresponding author upon reasonable request.

Competing interests

The authors declare no competing interests.

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1. Warnakulasuriya, S; Chen, T. Areca nut and oral cancer: evidence from studies conducted in humans. J. Dent. Res.; 2022; 101, pp. 1139-1146. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35459408]

2. Warnakulasuriya, S et al. Oral potentially malignant disorders: a consensus report from an international seminar on nomenclature and classification, convened by the WHO collaborating centre for oral cancer. Oral Dis.; 2021; 27, pp. 1862-1880. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33128420]

3. Tang, J et al. Oral submucous fibrosis: pathogenesis and therapeutic approaches. Int. J. Oral Sci.; 2025; 17, 8. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39890798][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11785813]

4. Gupta, D; Sharma, S. Oral submucous fibrosis—a new treatment regimen. J. Oral Maxillofac. Surg.; 1988; 46, pp. 830-833. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/3171741]

5. Chen, X; Xie, H; Guo, J. Drug treatment for oral submucous fibrosis: an update. BMC Oral Health; 2023; 23, 748. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37828490][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10568776]

6. Chole, RH et al. Review of drug treatment of oral submucous fibrosis. Oral Oncol.; 2012; 48, pp. 393-398. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/22206808]

7. Sumeth Perera, M et al. Development of an in vivo mouse model to study oral submucous fibrosis. J. Oral Pathol. Med.; 2007; 36, pp. 273-280. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17448137]

8. Wen QiTao, W. Q. et al. Development of a mouse model of arecoline-induced oral mucosal fibrosis (2017).

9. Tang, S et al. Comparative analysis of two arecoline-induced oral submucous fibrosis models. Oral Dis.; 2024; 30, pp. 3897-3911. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38098250]

10. Kandhare, AD; Bodhankar, SL; Mohan, V; Thakurdesai, PA. Glycosides based standardized Fenugreek seed extract ameliorates bleomycin-induced liver fibrosis in rats via modulation of endogenous enzymes. J. Pharm. Bioallied Sci.; 2017; 9, pp. 185-194. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28979073][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5621181]

11. Yamamoto, T et al. Animal model of sclerotic skin. I: local injections of bleomycin induce sclerotic skin mimicking scleroderma. J. Invest. Dermatology; 1999; 112, pp. 456-462.

12. Della Latta, V; Cecchettini, A; Del Ry, S; Morales, MA. Bleomycin in the setting of lung fibrosis induction: from biological mechanisms to counteractions. Pharmacol. Res.; 2015; 97, pp. 122-130. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25959210]

13. Zhang, SS et al. A rat model of oral submucous fibrosis induced by bleomycin. Oral Surg. Oral Med. Oral Pathol. Oral Radiol.; 2016; 122, pp. 216-223. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27422420]

14. Association, CS. Guideline for the diagnosis and clinical management of oral submucous fibrosis. Chin. J. Dent. Res.; 2023; 26, pp. 271-285.

15. Chang, YC et al. Arecoline-induced myofibroblast transdifferentiation from human buccal mucosal fibroblasts is mediated by ZEB 1. J. Cell. Mol. Med.; 2014; 18, pp. 698-708. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24400868][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4000120]

16. Pereira, T; Naik, S; Tamgadge, A. Quantitative evaluation of macrophage expression using CD68 in oral submucous fibrosis: an immunohistochemical study. Annals Med. Health Sci. Res.; 2015; 5, pp. 435-441.

17. Zhou, R; Yin, L; Zhang, X; Zhu, K. SFRP1 reduces neutrophil infiltration and inhibits the Wnt/β-catenin pathway to alleviate oral submucous fibrosis. Vitro Cell. Dev. Biology-Animal; 2024; 60, pp. 1034-1045.

18. Nag, R; Pal, M; Paul, RR; Chatterjee, J; Das, RK. Segmentation and analysis of surface characteristics of oral tissues obtained by scanning electron microscopy to differentiate normal and oral precancerous condition. Tissue Cell.; 2019; 59, pp. 82-87. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31383292]

19. Sabarinath, B; Sriram, G; Saraswathi, T; Sivapathasundharam, B. Immunohistochemical evaluation of mast cells and vascular endothelial proliferation in oral submucous fibrosis. Indian J. Dent. Res.; 2011; 22, pp. 116-121. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/21525689]

20. Chiang, C; Wu, H; Liu, B; Wang, J; Kuo, M. Quantitative analysis of immunocompetent cells in oral submucous fibrosis in Taiwan. Oral Oncol.; 2002; 38, pp. 56-63. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/11755822]

21. Sarode, G; Sarode, SC; Deshmukh, R; Raktade, P; Patil, S. Myofibroblasts could be recruited in a chemokine (C-C motif) ligand 2‐dependent manner in pathogenesis of oral submucous fibrosis. J. Oral Pathol. Med.; 2017; 46, pp. 443-447. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28029690]

22. Khan, I; Kumar, N; Pant, I; Narra, S; Kondaiah, P. Activation of TGF-β pathway by Areca nut constituents: a possible cause of oral submucous fibrosis. PLoS One; 2012; 7, e51806. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/23284772][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3526649]

23. Haque, M; Meghji, S; Khitab, U; Harris, M. Oral submucous fibrosis patients have altered levels of cytokine production. J. Oral Pathol. Med.; 2000; 29, pp. 123-128. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/10738939]

24. Kaur, J; Jacobs, R. Proinflammatory cytokine levels in oral lichen planus, oral leukoplakia, and oral submucous fibrosis. J. Korean Association Oral Maxillofacial Surg.; 2015; 41, 171.

25. Sodhi, S et al. Expression of tumor necrosis factor α and its correlation with severity of oral submucous fibrosis: a case-control study. Oral Surg. Oral Med. Oral Pathol. Oral Radiol.; 2014; 117, pp. 704-708. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24787450]

26. Wang, S et al. DPSCs regulate epithelial-T cell interactions in oral submucous fibrosis. Stem Cell Res. Ther.; 2024; 15, 113. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38650025][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11036714]

27. Misharin, AV et al. Monocyte-derived alveolar macrophages drive lung fibrosis and persist in the lung over the life span. J. Exp. Med.; 2017; 214, pp. 2387-2404. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28694385][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5551573]

28. Deng, KM et al. Deleterious role of Th9 cells in pulmonary fibrosis. Cells; 2021; 10, 3209. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34831433][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8621886]

29. Ishikawa, H; Takeda, K; Okamoto, A; Matsuo, S; Isobe, K. -i. Induction of autoimmunity in a bleomycin-induced murine model of experimental systemic sclerosis: an important role for CD4 + T cells. J. Invest. Dermatol.; 2009; 129, pp. 1688-1695. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/19158840]

30. Liang, Y et al. Dynamic pathological analysis reveals a protective role against skin fibrosis for TREM2-dependent macrophages. Theranostics; 2024; 14, 2232. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38505612][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10945340]

31. Liang, M et al. A modified murine model of systemic sclerosis: bleomycin given by pump infusion induced skin and pulmonary inflammation and fibrosis. Lab. Invest.; 2015; 95, pp. 342-350. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25502178]

32. Shih, YH; Wang, TH; Shieh, TM; Tseng, YH. Oral submucous fibrosis: a review on etiopathogenesis, diagnosis, and therapy. Int. J. Mol. Sci.; 2019; 20, 2940. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31208114][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6627879]

33. Yang, B; Fu, MF; Tang, ZG. Rat model with oral submucous fibrosis induced by Arecoline and mechanical stimulation. West. China J. Stomatology; 2019; 37, pp. 260-264.

34. Guo, ZX et al. A biomaterial-based therapy using a sodium hyaluronate/bioglass composite hydrogel for the treatment of oral submucous fibrosis. Acta Biomater.; 2023; 157, pp. 639-654. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36509401]

35. Rydell-Törmänen, K; Johnson, JR. The applicability of mouse models to the study of human disease. Methods Mol. Biol.; 2019; 1940, pp. 3-22. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30788814]