Investigation of the CTLA-4-CD28 Axis in Oral Squamous Cell Carcinoma

Abstract

Background: Oral squamous cell carcinoma (OSCC) is a common head and neck cancer with low survival rates, especially in advanced stages, despite improved therapies. New developments show that immune checkpoint inhibitors (ICIs) are promising treatment options. A better understanding of immune suppression in OSCC could enable new therapeutic approaches and effective ICI combinations. Methods: The aim of this cross-sectional study was to investigate the significance of the differential expression of cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), CD28 and their ligands CD80 and CD86 for the diagnosis and treatment of OSCC. To this end, mRNA expression was analysed by RT-PCR and compared in 65 healthy oral mucosa samples (NOM) and 104 OSCC samples. Results: The expression of CTLA-4 (a soluble and membrane-bound isoform) was increased in OSCC by 1.72-fold (p = 0.004) and 6.88-fold (p < 0.001), respectively. There was no significant difference for CD28 (p = 0.283), nor for the soluble isoform of CD86 (p = 0.845). The membrane isoform of CD86 was increased in OSCC by a factor of 1.39 (p = 0.009) and CD80 by 6.11-fold (p < 0.001). Conclusions: The results show a significant association between CTLA-4, CD80 and membrane-bound CD86 expression and diagnosis. They could improve diagnostics in multi-marker approaches and serve as therapeutic targets for ICI strategies. In particular, the data indicate a stronger immunosuppressive role of CD80 compared to CD86 in a tumor tissue context, suggesting the exploration of anti-CTLA-4 and anti-CD80 antibody combinations in animal models.

Keywords: CD 28; CD80; CD86; CTLA-4; OSCC; immune checkpoints; oral squamous cell carcinoma.

Figure 1

Comparison of the interaction between CTLA-4 and CD80/CD86 with that between CD28 and CD80/CD86. The outcome of the interaction between the T cell receptor (TCR) and the presented antigen is determined by the expression and interaction of additional receptors on the T cell. Without a second/costimulatory signal, the T cell becomes anergic or undergoes apoptosis. However, if CD28 recognises B7, the T cell becomes activated, which leads to proliferation and differentiation. CTLA-4 is then expressed on the T cell. Recognition of B7 by CTLA-4 results in cell cycle arrest, thereby preventing further T cell activation. Therefore, the effects of CD28 and CTLA-4 on the T cell are opposing. Although both receptors bind the same ligands, CTLA-4 has a higher affinity for B7 molecules. This figure is based on the presentation by Babamohamadi et al. [11].

Figure 2

Comparative analysis of the differential expression of CTLA-4_var1 and var2, CD28, CD80, CD86_Iso1 and Iso3 by RT-qPCR in OSCC vs. NOM. The data are presented in the form of box plots. Expression levels are presented as median ΔCT values. Higher average ΔCT values indicate lower expression. p-values were calculated using Mann–Whitney U test. FC = Fold change = Relative change in expression level between the groups. (A) The expression of CTLA-4_var1 is increased in OSCC tissue. (B) The expression of CTLA-4_var2 is significantly increased in OSCC tissue. (C) No significant change in CD28 expression was observed between the groups. (D) The expression of CD80 is significantly increased in OSCC tissue. (E) The expression of CD86_Iso1 is significantly increased in OSCC tissue. (F) No notable alteration in CD86_Iso3 expression was observed between the groups.

Figure 3

Assessment of statistical relevance of differential expression in OSCC vs. NOM. Receiver operating characteristic (ROC) curves were generated. The AUC (area under the curve) values demonstrate the significant association between the overexpression of the corresponding gene and malignant tissues. The circles indicate the highest Youden indices, which were used to determine the cut-off point (COP). (A) ROC of CTLA_var1. (B) ROC of CTLA_var2. (C) ROC of CD80. (D) ROC of CD86_Iso1.

Figure 4

Analysis of association between diagnosis and positivity of increased expression. The OSCC and NOM groups were divided into positive and negative subgroups for overexpression based on the determined COPs. The association between malignancy and over expression = value under the COP was tested by the Chi-square test. The results demonstrated a significant association between malignancy and over expression of (A) CTLA-4_var1, (B) CTLA-4_var2, (C) CD80, and (D) CD86_Iso1.

Figure 5

Relationship between differential mRNA expression rates of CD80 and clinical stages of disease. Statistical evaluation of differential mRNA expression level of CD80 in tissue specimens of OSCC between early and late clinical status. (A) Box plot showing the comparison of expression level between the groups. Higher average ΔCT values indicate lower expression. p-values were calculated using Mann–Whitney U test. FC = Fold change = Relative change in expression levels between the groups. The expression of CD80 is increased in OSCC tissue with late clinical stages. (B) A receiver operating characteristic (ROC) curves was generated. This was achieved by plotting the sensitivity against the specificity (1-specificity). The AUC (area under the curve) value substantiates the significant association between the overexpression of CD80 and late clinical stages. The circle indicates the highest Youden index, which was used to determine the ideal cut-off point (COP) for distinguishing between early and late stages. (C) The OSCC group was subdivided into positive and negative subgroups for overexpression based on the identified COP. The results generated by Chi-square test revealed a statistically significant association between late clinical stages and overexpression of CD80.

Figure 6

Spearman correlation analysis of the expression levels of the investigated immunomodulators. (a) Spearman correlation analysis of CTLA-4 and CD28 and their ligands CD80 and CD86 in total sample with ΔCT of CTLA-4_var1 as dependent variable at the Y-axis. The data is presented in the form of scatter plots. The expression levels are presented as ΔCT values. p-value was calculated by Spearman’s rho correlation test. ρ = Correlation coefficient. (A) A strong positive correlation was observed between the expression of CTLA-4_var1 and CD28. (B) A moderate positive correlation was observed between the expression of CTLA-4_var1 and CD80. (C) A strong positive correlation was observed between the expression of CTLA-4_var1 and CD86_Iso1. (D) A strong positive correlation was observed between the expression of CTLA-4_var1 and CD86_Iso3. (b) Spearman correlation analysis of CTLA-4 and CD28 and their ligands CD80 and CD86 in total sample with ΔCT of CD28 as dependent variable at the Y-axis. The data is presented in the form of scatter plots. The expression levels are presented as ΔCT values. p-value was calculated by Spearman’s rho correlation test. ρ = Correlation coefficient. A strong positive correlation was observed between the expression of CD28 and (A) CTLA-4_var1 and (B) CTLA-4_var2. (C) A moderate negative correlation was observed between the expression of CD28 and CD80. A strong positive correlation was observed between the expression of CD28 and (D) CD86_Iso1 and (E) CD86_Iso3.

Figure 6

Spearman correlation analysis of the expression levels of the investigated immunomodulators. (a) Spearman correlation analysis of CTLA-4 and CD28 and their ligands CD80 and CD86 in total sample with ΔCT of CTLA-4_var1 as dependent variable at the Y-axis. The data is presented in the form of scatter plots. The expression levels are presented as ΔCT values. p-value was calculated by Spearman’s rho correlation test. ρ = Correlation coefficient. (A) A strong positive correlation was observed between the expression of CTLA-4_var1 and CD28. (B) A moderate positive correlation was observed between the expression of CTLA-4_var1 and CD80. (C) A strong positive correlation was observed between the expression of CTLA-4_var1 and CD86_Iso1. (D) A strong positive correlation was observed between the expression of CTLA-4_var1 and CD86_Iso3. (b) Spearman correlation analysis of CTLA-4 and CD28 and their ligands CD80 and CD86 in total sample with ΔCT of CD28 as dependent variable at the Y-axis. The data is presented in the form of scatter plots. The expression levels are presented as ΔCT values. p-value was calculated by Spearman’s rho correlation test. ρ = Correlation coefficient. A strong positive correlation was observed between the expression of CD28 and (A) CTLA-4_var1 and (B) CTLA-4_var2. (C) A moderate negative correlation was observed between the expression of CD28 and CD80. A strong positive correlation was observed between the expression of CD28 and (D) CD86_Iso1 and (E) CD86_Iso3.