Tissue mechanics and expression of TROP2 in oral squamous cell carcinoma with varying differentiation

The protocols in this study were approved by the research ethics committee of Lanzhou University. Tumor samples were collected from patients after obtaining written informed consent. A total of 108 patients with oral squamous cell carcinoma (OSCC) were registered at the second hospital of Lanzhou University between January 2013 and March 2019. Among these samples, 36 samples each showed high, moderate, and low levels of differentiation. The experimental group comprised 60 males and 48 females aged 41–68 years (average: 51 years). All patients were diagnosed with OSCC based on surgery and pathology; patients did not undergo radiotherapy, chemotherapy, or immunotherapy before surgery. Pathological analysis after tumor biopsy was performed by two experienced pathologists, after which the diagnosis of other diseases (including inflammation at other sites and secondary tumors) were excluded. Cancer and cervical lymph node tissues were collected after maxillofacial surgery. All specimens were sampled from typical areas of the lesion and fixed with 10% neutral-buffered formalin followed by conventional paraffin embedding. Among them, 42 and 66 patients exhibited no and cervical lymph node metastases, respectively. Clinical TNM staging was performed according to the 7th edition TNM staging classification standard jointly developed by the International Union for Cancer Control and American Joint Committee on Cancer [24] and World Health Organization guidelines [25].

OSCC tissues were fixed overnight using 10% neutral formalin (Solarbio, Beijing, China), paraffin embedded, sliced into 4-μm thick sections, dewaxed using xylene, and rehydrated using different concentrations of ethanol. The sections were stained with hematoxylin for 5 min and hydrochloric acid-ethanol and eosin for 3 min followed by gradient dehydration, transparentization, sealing, and neutral resin sealing (Solarbio, Beijing, China). Sections were visualized and imaged using the Olympus BX53 at magnifications of × 10, 20, and 40.

HE sections were subjected to the S-P method to detect TROP2 expression. The sections were incubated overnight with the primary antibody against TROP2 (1:1000, Abcam, USA) at 4 °C followed by incubation with biotin-labeled goat anti-rabbit IgG (1:5000, Abcam, USA) at 37 °C for 1 h. The sections were then developed using DAB (Beijing Zhongshan Golden Bridge Biotechnology, China), dehydrated, transparentization, and film and neutral resin sealed. The prepared sections were visualized using microscopy (Olympus BX53, Japan).

Fixed tissues were placed in Petri dishes containing phosphate-buffered saline. All analyses for mechanical properties were performed using the biological atomic force microscope (BioAFM; NanoWizard III, Bruker, USA). Silicon AFM probes from the Pointprobe® series with a force constant of 0.2 N/m (CONTR-reflex coating, NanoWorld, USA) were used. The spring constant of the probe was calibrated using built-in thermal vibration before measuring the resonance frequency of 13 kHz and thickness of 2 μm. AFM was performed using the contact model and a scan rate of 0.5 Hz/s in air. Force-distance curves are generated when the probe contacts the tissue following which, the structure, morphology, and mechanical properties of samples are measured at 5 μm/s [26]. Six random sites were selected for each sample and each site was measured 15 times. We used the modified Hertz contact model to analyze force-distance curves [27] and calculate Young’s modulus and roughness of OSCC tissues with varying differentiation.

Tumor cells from poorly differentiated OSCC samples exhibited characteristic atypia, poor differentiation, and irregular morphology (Fig. 1). However, the number, volume, atypia, nuclear pyknosis, and mitotic structures decreased in tumor cells from highly differentiated OSCC as compared to those in poorly differentiated cells. TROP2 primarily localized in the cytoplasm of tumor cells, but not in adjacent normal epithelial cells. We observed that low differentiation and high malignancy of OSCC was associated with higher TROP2 expression (Fig. 2). The average optical density of TROP2 among the low, medium, and highly differentiated OSCC tissues were 0.32 ± 0.042, 0.25 ± 0.018, and 0.11 ± 0.013, respectively (Fig. 3).

We analyzed the clinicopathological characteristics of patients with OSCC with varying degree of differentiation. Differential expression of TROP2 was associated with patient age, tumor differentiation, tumor size, TNM stage, percutaneous nerve filtration, and vascular invasion (Table 1, P < 0.05). Patients with poorly differentiated tumors were more likely than patients with well and moderate differentiated tumors to have high TROP2 expression (P < 0.001). However, there was no association between the expression of TROP2 and patient gender, tumor location, lymph node metastasis, or distant metastases (P > 0.05).

Using Kaplan-Meier survival curves, we observed that an increase in TROP2 expression negatively correlated with the overall survival of patients (Fig. 4). And low/no of TROP2 expression group’s 3-years survival rate was 55.6%, a 20.4% for high expression group and 5-years rate were 45.4 and 10.2% respectively. TROP2 expression was associated with patient age (P = 0.039, OR = 0.437, 95% CI [0.198–0.966]), tumor differentiation (Well vs. Moderate, P > 0.05, OR = 5.645, 95% CI [0.625–50.987]; Moderate vs. Poor, P < 0.001, OR = 105.400, 95% CI [19.053–583.063]; Well vs. Poor, P < 0.001, OR = 595.000, 95% CI [51.529–6870.366]), tumor size (P < 0.05, OR = 13.148, 95% CI [5.060–34.168]), TNM stage (P < 0.05, OR = 0.141, 95% CI [0.082–0.244]), vascular invasion (P < 0.05, OR = 14.587, 95% CI [4.653–45.729]), and peripheral nerve invasion (P < 0.05, OR = 11.062, Table 2). High TROP2 expression was detected in older patients with low degree of differentiation, larger tumor volume, higher TNM staging, and vascular and peripheral nerve invasion, thereby resulting in lower overall survival. Thus, TROP2 may be a prognostic indicator for survival in OSCC patients.

The surface morphologies of OSCC tissues with varying degrees of differentiation were analyzed (direct topographical imaging) using BioAFM. Figure 5 shows the representative image from each tissue acquired during the cantilever-based AFM nano- indentation test. The tissue interface varied with tumor differentiation, indicating that highly differentiated OSCC tissues had a regular and flat morphology. OSCC tissues with low differentiation exhibited an overall irregular morphology with distinct modulation and loose tissue organization. Figure 6 summarizes the roughness of OSCC tissues with varying differentiation. The average surface roughness of low, medium, and highly differentiated OSCC tissues were 448.9 ± 54.8, 792.7 ± 83.6, and 993.0 ± 104.3 nm, respectively. Roughness of the tissue surface was enhanced with increasing differentiation of OSCC tissues.

We used BioAFM to determine Young’s modulus based on the mechanical properties of 108 OSCC tissues with varying degrees of differentiation. We randomly selected six contact points from each slice and each contact point was measured 15 times. Force-distance curves were generated for each slice and the JPK Data Processing software (5.1.8 version) was used to calculate Young’s modulus. Figure 7 shows the average variation in stiffness within individual tissues in the range of 1–8 kPa. In the low differentiation samples, we observed low stiffness as compared to that in high or medium differentiation samples (P < 0.05). Thus, tissue differentiation was positively associated with its stiffness (Fig. 7).

The Pearson coefficient showed a negative association between the stiffness of OSCC tissues and TROP2 expression (Fig. 8, r = − 0.84, P < 0.01). Thus, we detected an increase in stiffness with varying differentiation in the tumor samples.