https://orcid.org/0000-0003-1597-4919
Abstract
Background: This study focussed on exploring the prognostic prediction performance of the growth arrest and DNA damage-inducible 45 beta (GADD45B) and its associations with T-cell activity and immune soakage in different malignancies, especially lung squamous cell carcinoma (LUSC). Methods: We applied TIMER database for comparing the expressions of GADD45B among different cancers. OncoLnc, Gene Expression Profiling Interactive Analysis 2 (GEPIA2), and Kaplan-Meier Plotter were utilised to evaluate the prognostic prediction performance of GADD45B. Besides, the associations of GADD45B with clinical stage, associated gene markers, and immune infiltration were examined through TISIDB, GEPIA2, and Tumour Immune Estimation Resource (TIMER). Biological processes (BPs) and KEGG enrichment analyses were performed to illustrate the possible role of GADD45B in LUSC. The miRWalk database was adopted to analyse the gene miRNA interaction network of GADD45B in LUSC. Results: GADD45B expression was decreased in most of the malignancies, with relation to the poor prognosis in LUSC. GADD45B also significantly affected the survival of LUSC subgroups divided by clinic data. GADD45B significantly correlates with and may stimulate T cell exhaustion in LUSC. Conclusions: GADD45B is a prognostic indicator in multiple tumours, especially in LUSC. Moreover, modulating GADD45B expression may improve immunotherapy efficacy in LUSC.
1. Introduction
Lung cancer (LC) contributes to most of the cancer-associated mortality globally, with 2,093,876 newly diagnosed and 17,607 deaths in 2018 alone [Citation1]. Non-small cell lung cancer (NSCLC) is categorised into two main subtypes [Citation2]. Among the histological subtypes of NSCLC, LUSC is a prominent subtype and is one of the deadliest cancers in the world [Citation3–5]. LUSC constitutes approximately 30% of the lung malignancy clinical cases globally and brings about nearly 400,000 deaths every year [Citation6,Citation7]. Although much progress has been achieved in the prevention, illness diagnoses, and therapy of lung cancer, its clinical results are still not ideal. There is substantial evidence indicating that the tumour microenvironment influences the malignant phenotype of the tumour [Citation8–10]. Lung cancer is an immune-sensitive malignant tumour, which involves the infiltration of several immune cells, such as monocytes, T cells, mast cells, B cells, dendritic cells (DCs), macrophages, eosinophils, and natural killer (NK) cells. Identifying effective biomarkers contributes to the diagnosis and prediction of the prognosis of cancer, which can help in the reduction of LC-induced mortality. However, only a few studies have systematically probed into the relationship between the poor prognosis of lung cancer and the microenvironmental immunophenotype.
GADD45B gene belongs to the growth arrest and DNA damage-inducible 45 (GADD45) gene family, which also includes GADD45A and GADD45G. GADD45B is related to stress signals in physiological and environmental stress responses and is implicated in oncogenic stress that can lead to DNA repair, arrest of the cell cycle, cell apoptosis, survival, and ageing [Citation11]. GADD45B can activate the p38/JNK pathway mediated by combining and activating the specific MTK1/MEKK4 kinase in response to environmental stress [Citation12]. GADD45B can change the epigenetic program of α2-adrenergic receptors and can regulate the generation of local pro-inflammatory factors within the rodent amygdala; therefore, GADD45B regulates the social behaviours of adolescents epigenetically [Citation13]. In stage-II colorectal cancer patients with reduced OS (overall survival) and PFS (progression-free survival) [Citation14] and those suffering from papillary thyroid cancer receiving iodine radiotherapy and total thyroidectomy [Citation15], the upregulation of GADD45B serves as a factor to independently predict patient prognosis. GADD45B is involved in the resistance to chemotherapy and aggressive growth of human embryonic carcinoma side population cells [Citation16]. Silencing of GADD45B can reduce the migration and invasion of cholangiocarcinoma cells [Citation17]. As demonstrated by these results, GADD45B promotes malignant phenotypes in multiple cancers. However, the related mechanism remains largely unclear.
In this study, through an integrated analysis applying a variety of databases, we illustrated the differential expression and prognostic signature of GADD45B across diverse cancer types. We also elucidated the relevance of GADD45B in tumour immune infiltration, and investigated its potential biological functions. Our study demonstrated the significance of GADD45B as a prognostic biomarker for LUSC and presented novel evidence for understanding the interactions among GADD45B expression, tumour invasion and T cell failure.
2. Materials and methods
2.1. Analysis based on TIMER database
The TIMER database provides application for calculating the abundance of tumour-infiltration immune cells (TIICs) [Citation18], which are divided into 6 subsets that are neutrophils, macrophages, DCs, B cells, and CD4+ and CD8+ T cells based on The Cancer Genome Atlas (TCGA) with the constrained least-squares fitting adopted [Citation19]. The TIMER database mainly covers 10,897 samples, which are from the TCGA database. We applied this database to analyse the pan-cancer expressions of GADD45B and its association with the abundance of the six TIIC subsets. The Wilcoxon test determined whether the difference was significant. We also performed purity-adjusted partial Spearman’s correlation analyses. Furthermore, tumour infiltration degrees among tumours with diverse somatic copy number alterations (SCNAs) which include high amplification (2), arm-level gain (1), diploid/normal (0), arm-level deletion (-1), and deep deletion (-2) were detected. In consistent with previous literatures, the "Correlation module” examined whether the expressions of GADD45B and the gene markers of multiple immune cells, including the functional cells like effector Treg and exhausted T cells, were correlated [Citation20–23]. Log2 RSEM (RNA-Seq by Expectation-Maximization) was calculated to represent the expression.
2.2. Analysis based on GEPIA2 database
The GEPIA2 database provides an online method for interactively analysing the diverse mRNA expression of an indicated gene in tumour vs. normal tissues, with tumour data from TCGA and the normal tissue expression datum also obtained from the Genotype-Tissue Expression (GTEx), it included 9736 tumours and 8587 normal samples. This database offers a series of tailored functions [Citation24]. It was used to determine the association of GADD45B expression with pan-cancer prognosis, and the connection between the expression of GADD45B and tumour-infiltration immune cell (TIIC) markers. Only datasets of tumour tissues were utilised in our study.
2.3. Analysis based on Kaplan-Meier (KM) plotter database
The KM Plotter database is a web-based resource that explores the prognostic signature in 21 types of malignancies. It includes four large datasets comprised of 6234 breast cancer (BC), 2190 ovarian cancer (OC), 3452 LC and 1440 gastric cancer (GC) cases [Citation25]. We asked whether GADD45B expression correlated with survival in these four malignancies using this database, and presented the results with survival curves and log-rank P and hazard ratio (HR) with 95% confidence interval (CI) values.
2.4. Analysis based on TISIDB database
With data retrieving from 7 public datasets and 988 reports, the TISIDB database integrated high-throughput screening techniques to provide multi-omics data for immunological signature exploration [Citation26]. This database was applied to determine the logical relationship of GADD45B expression with immunomodulators and lymphocytes.
2.5. Analysis based on HPA database
The HPA database incorporates pathologic and genetic expression data gather in plenty studies rooted in dissimilar cell lines and histologic types [Citation27]. It was adopted in this article to detect the levels of GADD45B within diverse histologic types, as well as the localisation of GADD45B mRNA in cells.
2.6. Analysis based on MEXPRESS database
MEXPRESS represents an approach to visualise data regarding the status of DNA methylation status, TCGA expression, clinical information, as well as the underlying associations [Citation28]. In this study, we used MEXPRESS to measure the methylation status of GADD45B and the correlation of its expression with diverse clinical parameters of LUSC.
2.7. Analysis based on LinkedOmics database
The GADD45B co-expression was analysed via Pearson test and visualised by volcano plot, heatmap, or scatter plot, using the "LinkFinder" module of LinkedOmics. This conducted the GO and KEGG enrichment analyses through GSEA using the "LinkInterpreter" module. The enriched pathways showing the false discovery rate (FDR) < 0.05 and simulations of 500 were considered significant [Citation29].
2.8. Analysis based on miRWalk database
GADD45B was imported into the miRWalk2.0 website for screening the miRNAs modulating GADD45B. The identified miRNAs were considered as the possible regulatory miRNAs of GADD45B [Citation30].
2.9. Analysis based on OncoLnc database
OncoLnc provides an interactive approach to study the survival associations of lncRNAs, mRNAs, and miRNAs. OncoLnc covers the survival data of 21 types of TCGA-derived cancers, as well as the corresponding data of MiTranscriptome, mRNAs and miRNAs. It can view KM chart results of at least one cancer at the same time, provide Cox regression data, and facilitate the extraction of sufficient data for analysis. In addition, users can also simultaneously detect the prognostic significance of the detected genes in 21 types of cancer, which helps to study the important functions of certain genes in cancer survival.
2.10. Statistical analysis
Box plot was generated to display the GADD45B expression suggested by TIMER, and Wilcoxon test was conducted to examine the clinical significance. The GEPIA2 and KM Plotter were employed to obtain survival curves with log-rank P values, HRs and Cox P values calculated. The immunocytes infiltration degree and its association with SCNA cancer and non-cancer tissues were detected by the Wilcoxon rank-sum test (two-sided). Spearman’s correlation analyses were employed for assessing the pertinence between the expressions of GADD45B and immune-related genes within LUSC. The expression levels of GADD45B in each subtype of lung squamous cell carcinoma were analysed, and the statistical significance of this correlation coefficient was tested by Kruskal-Wallis (-log10pv) test. Highly correlated gene markers co-expressed with GADD45B were analysed by Pearson ‘s test. The thresholds were set as the medians, unless otherwise noted. The difference was considered significant with a P-value ≤ 0.05. shows specific information on all databases used in the study.
Table 1. Detailed information of databases applied in the present study.
3. Results
3.1. Expression of GADD45B mRNA in different human cancers
To assess the levels of GADD45B in human malignant tumours, this study used the RNA-seq data from TCGA-derived tumours. The GADD45B expression in different TCGA-derived cancers was compared in tumour vs. normal tissues (). GADD45B levels were markedly decreased in CHOL (cholangiocarcinoma), BLCA (bladder urothelial carcinoma), ESCA (esophageal carcinoma), HNSC (head and neck cancer), COAD (colon adenocarcinoma), KICH (kidney chromophobe), KIRC (kidney renal clear cell carcinoma), KIRP (Kidney renal papillary cell carcinoma), PRAD (prostate adenocarcinoma), LUSC (Lung squamous cell carcinoma), LUAD (lung adenocarcinoma), LIHC (liver hepatocellular carcinoma), STAD (stomach adenocarcinoma), SKCM (skin Cutaneous Melanoma), UCEC (uterine corpus endometrial carcinoma), and THCA (thyroid carcinoma), by comparison with those in adjacent normal tissues.
Figure 1. Expression of GADD45B within different TCGA-derived cancers illustrated through TIMER. Significance levels: *p < 0.05, **p < 0.01, ***p < 0.001.
3.2. Prognostic prediction potential of GADD45B in cancers
To assess the prognostic prediction performance of GADD45B for cancer, this study conducted a comprehensive analysis on the associations of the levels of GADD45B with patient survival from three large online public cancer databases comprising diverse specimens. Then we explored the prognostic role of GADD45B in different types of cancer, details shown in Table S1. The GADD45B level was significantly associated with survival in eight cancers, including LUSC, ESCA, COAD, KIRC, READ, LIHC, STAD, and SARC ().
Table 2. The effect of GADD45B level on cancer prognosis based on OncoLnc analysis.
We assessed the prognostic prediction performance of GADD45B based upon RNA-seq data from 33 TCGA-derived tumours using GEPIA2. shows the significance of GADD45B for predicting the survival in various cancers. The association of increased GADD45B expression with poorer disease-free survival (DFS) and OS was observed in LUSC (p = 0.0015, HR = 1.6; p = 0.0024, HR = 1.7, respectively) (). Poor OS of THCA was related to higher GADD45B levels (p = 0.034, HR = 3.2), whereas dismal DFS of THCA was related to decreased GADD45B levels (p = 0.025, HR = 0.51) (). Dismal OS of HNSC were related to higher GADD45B levels (p = 0.035, HR = 1.3) (). Dismal DFS of LUAD were related to higher GADD45B levels (p = 0.047, HR = 1.4) (). Dismal OS of SARC was related to decreased GADD45B levels (p = 0.048, HR = 0.67) (). Dismal OS of SKCM was related to decreased GADD45B levels (p = 0.033, HR = 0.75) (). Moreover, GADD45B levels significantly affected the OS of STAD (stomach adenocarcinoma), UVM (uveal Melanoma) (Figure S1A and B).
Figure 2. Prognostic prediction performance of GADD45B in different cancers assessed through the GEPIA2 Plotter (a-g) and KM (h-k). (A) Survival heatmap regarding GADD45B levels within 33 TCGA-derived malignant tumours. The heatmap displays the log10 HRs with different colours. Blue and red squares indicate the lower and higher risks, respectively. The boxed rectangles represent the significance of the prognostic analysis. DFS and OS curved lines of LUSC (n = 482) (B), THCA (n = 510) (C), OS curves of HNSC (n = 518) (D), DFS curves of LUAD (n = 478) (E), OS curves of SARC (n = 262) (F), and OS curves of SKCM (n = 488) (G). DFS and OS curves of BC (n = 1879, n = 4929) (H), DFS and OS curves of LC (n = 1925, n = 982) (I), GC (n = 875, n = 640) (J), and OC (n = 1656, n = 1435) (K). OS-overall survival; RFS-relapse free survival; DFS-disease-free survival; PFS-progression-free survival.
Thereafter, this study researched the relevance of the levels of GADD45B with the patient prognostic outcome from four large-scale cancer datasets (including LC, BC, GC, and OC) based on KM Plotter. Decreased levels of GADD45B were observed to be related to the dismal survival rate in LC (OS: P = 3.4e−6 HR = 0.7 [0.6–0.81]; PFS: P = 3.8e−7, HR = 0.53 [0.41–0.68]), Higher GADD45B levels were related to dismal prognostic outcome in ovarian cancer (OS: p = 0.0082, HR = 1.19 [1.05–1.35]; PFS: p = 0.01, HR = 1.18 [1.04–1.35]) (). The levels of GADD45B suggested the poor prognosis of LUSC patients. Accordingly, the present research subsequently examined the related mechanisms based on the KM plotter to evaluate the relevance of the GADD45B level with sufferer clinicopathological features. GADD45B level was markedly relevant to OS, first progression, sex, histology, stage, grade, and smoking history, except race ().
Table 3. Prognostic prediction performance of GADD45B within different LUSC subtypes based on Kaplan-Meier Plotter.
3.3. Effect of GADD45B on the regulation of immune molecules
The TISIDB database was applied to assess the relevance of GADD45B level and immunomodulators (). The relationship between GADD45B expression level and tumour-infiltrating lymphocytes (TILs) was displayed in . The three most significant TILs associated with GADD45B included the mast cells (Spearman: ρ = 0.57), NK cells (Spearman: ρ = 0.528), and eosinophils (Spearman: ρ = 0.525), with each P value < 2.2e−16 (). The relationship between GADD45B expression level and immuno-inhibitors was displayed in . The three most significant immuno-inhibitors included ADORA2A (Spearman: ρ = 0.435), CSF1R (Spearman: ρ = 0.432), and HAVCR2 (Spearman: ρ = 0.404), with each P value < 2.2e−16 (). As shown in , the three immuno-stimulators representing the strongest associations with GADD45B expression were TNFSF13 (Spearman: ρ = 0.442), CXCR4 (Spearman: ρ = 0.442), and CD40LG (Spearman: ρ = 0.441), with each P < 2.2e−16 (). Besides, as displayed in , the MHC molecules showing the three strongest correlations with GADD45B were HLA-DOA (Spearman: ρ = 0.48), HLA-DPB1 (Spearman: ρ = 0.472), and HLA-DPA1 (Spearman: ρ = 0.443), with each P < 2.2e−16 (). Therefore, we suggested that GADD45B could likely modulate these immune molecules.
Figure 3. Spearman’s relevance between GADD45B, immunomodulators, and lymphocytes (TISIDB). (A) Association of the abundance of TILs with the level of GADD45B (B) The three most significant TILs associated with GADD45B expression level. (C) Associations of immuno-inhibitor abundances with GADD45B expression level. (D) The three most significant immuno-inhibitors associated with GADD45B expression level. (E) Associations of immuno-stimulator abundances with GADD45B expression level. (F) The three most significant immuno-stimulators associated with GADD45B expression level. (G) Association of MHC molecules with GADD45B expression level. (H) The three most significant MHC molecules associated with GADD45B expression level. Blue or red boxes indicate negative or positive correlations, respectively. The correlation strength is in direct proportion to the colour intensity. MHC, major histocompatibility complex; TILs, tumour-infiltrating lymphocytes.
3.4. GADD45B is related to the degrees of immune infiltration in LUSC
The frequency of lymphocyte infiltration in tumours independently predicts lymph node metastasis (LNM) and cancer survival [Citation31–33]. We further examined if GADD45B was involved in the immune cell infiltration in the 39 malignancies through TIMER analysis (Figure S2). We observed that GADD45B expression impacted the tumour purity in 23 of the malignancies and B cell infiltration in 12 of the malignancies. Additionally, the GADD45B expression also indicated altered infiltration level of neutrophils, macrophages, DCs and CD4+ or CD8+ T cells in 26, 24, 26, 21, and 17 of the malignancies, respectively. Whereas these significant associations were not observed in Lymphoid Neoplasm Diffuse Large B-cell Lymphoma (DLBC) (Figure S2-2J).
We also suggested that GADD45B expression level presented a strong association with macrophage (R = 0.443, P = 2.44e-24), CD4 + T cell (R = 0.439, P = 8.23e-24) and DC (R = 0.419, P = 1.53e-21) infiltrations, presented a modest association with neutrophil (R = 0.359, P = 6.79e-16) infiltrations and the purity level (R = −0.326, P = 2.65e-13), and presented a weak but significant association with CD8+ T cell (R = 0.224, P = 8.51e-07) and B cell (R = 0.213, P = 3.00e-06) infiltrations in LUSC (). The KM plots revealed that the expression of GADD45B significantly correlated with LUSC survival (p = 0.003) (). Notably, the deletions or normal copy numbers of the GADD45B gene locus were related to higher levels of immune cells infiltration (). The present findings revealed the significant effect of GADD45B on immune infiltrations of LUSC.
Figure 4. Associations of GADD45B with degrees of immune infiltration within LUSC. (A) Associations of GADD45B expression level with immune infiltration degrees of CD8+ or CD4+ T cells, B cells, neutrophils, macrophages, and DCs and tumour purity. (B) TIMER plots showing the correlation of immune infiltrations or GADD45B expression with survival in LUSC, its sample contained 482 patients of which 205 died. (C) Tumour-infiltration degrees were associated with GADD45B SCNAs in LUSC. Significance levels: * p < 0.05; ** p < 0.01; *** p < 0.001.
3.5. Association of GADD45B with TIIC gene markers
To clarify the impact of GADD45B on tumour immune infiltrations, the associations of GADD45B with TIIC markers in LUSC were assessed employing the GEPIA2 and TIMER databases. The gene markers for functional T cells and other immune cells were identified as recorded in the previous literatures [Citation21, Citation34,Citation35]. The correlation analytical results after tumour purity adjustment in LUSC are shown in . The expression of GADD45B significantly correlated with that of the markers of multiple immune cells and a majority of functional T cells, as exhibited in . Notably, these findings indicated the robust association between GADD45B and functional T cell infiltration, and revealed the correlation between GADD45B and exhausted T cell infiltration, which is supported by limited research till now [Citation36,Citation37].
Table 4. Relationships between GADD45B and TIIC markers based on the TIMER in LUSC.
To validate these results, we also evaluated the correlations between the expressions of GADD45B and the TIIC markers in both the LUSC and non-cancer tissues using GEPIA2 (). GADD45B was positively correlated with mast cells, M2 macrophages, DCs, TAMs, monocytes, neutrophils, T cells, B cells, as well as multiple functional T cells, especially central memory T cells, effector T cells, and exhausted T cells. In our study, GADD45B significantly correlated with the expressions of several key genes referring to exhausted T cells, including TIM-3, PD-1 and TIGIT, with each P-value < 0.0001, and sequentially decreased correlation strength (0.36, 0.2 and 0.19), which are also vitally involved in cancer immunotherapy. Elevated GADD45B level was also correlated with a higher degree of DC infiltration in LUSC. Consistently, the DC markers CD209, HSD11B1, HLA-DRA, HLA-DPA1, HLA-DPB1, HLA-DQB1, BCDA-1, CD11c, and BDCA-4 were correlated with GADD45B expression, suggesting that GADD45B is significantly correlated with the DC infiltration in tumours. DCs can facilitate the degree of tumour metastasis by promoting the responses to Tregs while restraining the cytotoxicity of CD8+ T cells [Citation38]. Further inquiry is needed to probe the role of GADD45B in adjusting the penetration of DCs and tumour metastasis. The above findings confirmed the important role of GADD45B in the degree of infiltration of different critical immune cells. GADD45B was remarkably connected with several crucial genes related to exhausted T cells, indicating that GADD45B played an important role in anti-LUSC immunotherapy. Besides, GADD45B was observed to be related to mast cells in LUSC, indicating its significance and the necessity of additional studies to investigate its related mechanisms.
Table 5. The correlation of GADD45B with TIIC gene markers in LUSC and normal samples using GEPIA2.
3.6. GADD45B expression in LUSC
GADD45B expression was detected in five subtypes based on the difference in immunological characteristics, wound healing type (C1), interferon γ (IFN-γ) dominant type (C2), inflammatory type (C3), lymphocyte depleted type (C4) and the TGF-b dominant type (C6). GADD45B was expressed as the highest and the lowest in the C3 and C4 types, respectively (). This study also detected GADD45B in different LUSC molecular subtypes in TISIDB, including classical, basal, secretory, and primitive types. Based to our consequences, the highest and lowest GADD45B expression levels were measured in secretory and classical subtypes in TISIDB, respectively (). Based on these findings, GADD45B was tightly associated with the tumour immune microenvironment (TIME). In line with GEPIA2 database-based data analysis, the differences between LUSC stages from certain datasets were not significant (). As disclosed by HPA database analysis, high GADD45B staining was observed in LC specimens compared to non-carcinoma lung tissue (). Moreover, MEXPRESS-based analysis revealed the association between the levels of GADD45B mRNA and the histological type and sample type ().
Figure 5. GADD45B expression in LUSC. (A) GADD45B expression in different TISIDB-derived LUSC immune subtypes. (B) GADD45B expression in different TISIDB-derived LUSC molecular subtypes. (C) GADD45B expression in different GEPIA2-derived LUSC stages. GADD45B gene expression data were calculated relative to log counts per million mapped reads (log2 CPM) in (A, B) and log2 (TPM + 1) in (C). (D) HPA-based immunohistochemistry of GADD45B. GADD45B in normal tissue: https://www. proteinatlas.org/ENSG00000099860-GADD45B/tissue/lung; GADD45B in tumour tissue: https://www.proteinatlas.org/ENSG00000099860-GADD45B/pathology/lung+ cancer#img. T: GADD45B pcarotein expression in cancer samples (high staining, strong intensity, quantity >75%); N: GADD45B protein expression in non-carcinoma samples (low staining, weak intensity, quantity 75%-25%). (E) GADD45B mRNA expression is correlated with the histological and sample types (n = 758).
3.7. GADD45B co-expression networks in LUSC
We identified the co-expressing genes of GADD45B applying the LinkFinder module in the LinkedOmics for determining the potential biological functions of GADD45B in LUSC. As shown in , a total of 11264 genes displayed positive correlations and 8838 genes displayed negative correlations with GADD45B, respectively (p < 0.05). exhibited the heat maps displaying the 50 most significant GADD45B-related genes (either positive or negative).
Figure 6. Co-expression and potential function analyses of GADD45B in LUSC (LinkedOmics). (A) GADD45B-related genes in LUSC identified through the Pearson test. Green and red dots separately represent genes with notably negative and positive correlations with GADD45B. (B and C) Heatmaps presenting the 50 most significantly correlated (both positive and negative) GADD45B-related genes in LUSC. (D and E) Markedly associated GO: BP annotations along with KEGG pathway analysis for GADD45B in LUSC. (F and G) Survival heatmaps revealing the 50 most significantly correlated (both positive and negative) GADD45B-related genes in LUSC. Survival heatmaps revealing the log10 HRs of different genes. Blue and red squares denote decreased and increased risks, respectively. Blocks with frames indicate significant positive and negative consequences of the prognostic factors analysis (p < 0.05). (H) GADD45B and its predicted miRNAs (GADD45B is displayed as blue rounds and targeted miRNAs as yellow rounds. The mutual effect between GADD45B and its interrelated miRNAs is presented as lines). GO: Gene Ontology; KEGG: Kyoto Encyclopaedia of Genes and Genomes; FDR: false discovery rate.
On the grounds of the GSEA-noted GO terms, the genes showing co-expression of GADD45B were mostly connected with extracellular structure organisation, leukocyte migration, regulation of inflammatory response, response to molecules of bacterial origin, and adaptive immune response (). As displayed by the KEGG analysis, these genes were mostly connected with cell adhesion molecules, coagulation and complement cascades, osteoclast differentiation, phagosomes, and interaction between cytokines and cytokine receptors. In contrast, the RNA degradation, cell cycle, ribosomes, RNA transport, and spliceosome process were inhibited (). In , the hsa04010: MAPK, hsa04012: ErbB, hsa04115: P53, and hsa04110: cell cycle signal transduction pathways were related to tumorigenesis and pathogenic mechanisms in NSCLC.
Figure 7. NSCLC pathway that is regulated by the alteration of GADD45B.
Notably, of the 50 most significant genes that were positively associated with GADD45B, 22 were found with HR > 1 (each p < 0.05). Conversely, of the 50 most significant genes that were negatively associated with GADD45B, only one gene was found with HR <1 (each p < 0.05) ().
3.8. miRNA screening of regulatory GADD45B
The miRWalk was applied to screen the miRNAs potentially targeting GADD45B and to construct the miRNA-gene network, which consists of GADD45B and 402 miRNAs (). The number of lines indicates the contribution level of each miRNA. Additionally, the 20 most significant miRNAs potentially targeting the GADD45B are shown in .
4. Discussion
GADD45B belongs to the growth-arrest DNA damage-inducing gene family [Citation14], which performs the common function of the gene family and is related to cell growth, DNA damage repair, anti-cancer immune response, and apoptosis [Citation39,Citation40]. However, none of the existing studies have comprehensively analysed the associations of the level of GADD45B with T cell activity, survival, and immune infiltration in different malignancies. This study investigated, for the first time, the strong relationship between GADD45B and LUSC using several large databases. This study used cancer specimens from several large-scale databases for analytical investigation. The results indicated that the expression of GADD45B was concerned in the survival of patients suffering from different cancers, including LUSC. The co-expressed genes of GADD45B are also suggested to have great prognostic value for LC. In addition, the level of GADD45B had a positive relation with the degrees of immune infiltration in LUSC. GADD45B was previously identified as a tumour suppressor gene [Citation41], which appeared to contradict this study, wherein it was observed that the upregulation of GADD45B was an unfavourable prognostic factor for LUSC. Similarly, the promotion of GADD45B expression in tumorigenicity or rapid progression of the disease has been reported in colorectal cancer [Citation14, Citation42], GC [Citation43], as well as OC [Citation44]. The present work suggested the feasibility of GADD45B as a prognostic biomarker for LUSC, offering novel avenues to understand the associations of GADD45B with the function of T cells and immune infiltration. The research detected the expression of GADD45B with a systematic prognostic landscape for different forms of cancers based on large datasets in TIMER. GADD45B was differently expressed in tumour and non-carcinoma samples of various cancers. Analysis according to TIMER database, the expression of GADD45B decreased in BLCA, CHOL, ESCA, COAD, KICH, HNSC, KIRP, KIRC, LUAD, LIHC, PRAD, LUSC, STAD, SKCM, UCEC, and THCA (), compared to non-carcinoma samples.
Human Protein Atlas data further confirmed the expression of GADD45B in LC using Immunohistochemistry (). We found no statistical significance between the expression of GADD45B and LUSC stage (). As suggested by OncoLnc database-based analysis, the level of GADD45B predicted the dismal survival in several cancers (LUAD, BRCA, UCEC, HNSC, BLCA, GBM, OV, LUSC, KIRC, LGG, COAD, STAD, ESCA, and READ) (). Besides, as suggested by GEPIA2-based data analysis, the GADD45B level predicted the dismal prognostic outcome of LUSC. However, Kaplan-Meier Plotter-based data analysis suggested that lung cancer with the upregulation of GADD45B had a better prognosis. Such an unusual trend potentially suggested that there might be other unclear mechanisms requiring further investigation. The expression of GADD45B also correlated significantly with univariate analysis, multivariate analysis, as well as patient gender, histology, stage, and grade of smoking history, except race (). Overall, GADD45B may act as a prognostic biomarker for LUSC.
This study evaluated the relativity of the level of GADD45B with the immune system using the TISIDB database. Our results demonstrated that the expression of GADD45B was the most significantly related to lymphocytes (including mast cells, NK cells, and eosinophils), immunostimulators (including TNFSF13, CXCR4, and CD40LG), immuno-inhibitors (including ADORA2A, CSF1R, and HAVCR2), and MHC molecules (including HLA-DMA, HLA-DPB1, and HLA-DPA1). Therefore, GADD45B might provide new directions for studying the immune escape in LC cells and a therapeutic target for anti-LC immunotherapy.
However, lung cancer is by no means a single illness and is further divided into multiple molecular subtypes. According to a TISIDB database-based analysis, the GADD45B gene showed the highest expression in the secretory subtypes, followed by the basal subtype, and was the least expressed in the classical and primitive subtypes. Differential GADD45B expression within LUSC of varying immune subtypes was tested. The outcomes presented that the C3 subtype displayed the greatest expression of GADD45B compared to that in the residual five subtypes. The final generalised analysis of GADD45B levels in different LUSC subtypes from various databases suggests the important role of GADD45B in the TIME characteristics.
Since GADD45B had a vital impact on the immune system and forecasting the prognosis of LUSC, The present research explored the associations of GADD45B with degrees of immune infiltration in LUSC (). The upregulation of GADD45B was compactly concerned in high degrees of immune infiltration of several subpopulations of immune cells, such as B cells, CD8+ T cells, neutrophils, CD4+ T cells, macrophages, monocytes, and DCs. The different SCNA for GADD45B markedly affected the degrees of immune infiltration within LUSC (), and great concern was paid to the close link between GADD45B and immune cells. As indicated by subsequent analyses of the relationships of GADD45B with TIIC gene markers, GADD45B interacted with several immune cells as well as functional T cells, including central memory T cells, effector Treg T cells, effector T cells, and exhausted T cells ( and ). Since the exhaustion of T cells is a principal cause for noneffective anti-cancer immunity [Citation45–47], the step for its prevention is vital for anti-cancer immunotherapy. Our study illustrated that the increased GADD45B expression indicated a higher expression of various critical genes referring to the exhausted T cells, including TIM-3, PD-1, and TIGIT, which are presently therapeutic targets or participate in immunotherapy [Citation48,Citation49].
We analysed the correlation between the expression level of GADD45B and immune cell infiltration based on the TIMER database in an attempt to find evidence that this biomarker regulates certain immune cell infiltration in lung squamous cell carcinoma. As shown in , the expression of GADD45B was highly correlated with DC cell infiltration in lung squamous cell carcinoma, and the reliability of the data was again demonstrated in the analysis of somatic copy number changes [Citation50]. The difference of expression level of GADD45B in different immune cell infiltration suggests that GADD45B is specific in immune infiltration of lung squamous cell carcinoma. Lung cancer usually has a high burden of tumour mutations, and it also has a strong immunosuppressive microenvironment, especially destroying and interfering with DC cells and T cells [Citation51]. DC-based immunotherapy is safe, well tolerated, and can elicit antitumor immune responses in many lung cancer patients. Combining DC cell-based immunotherapy with other anticancer therapies, such as chemotherapy, radiotherapy, and/or checkpoint inhibition, may improve their efficacy. Alternatively, the selection of antigens based on novel epitopes expressed in lung cancer cells has been shown in previous studies not only to induce immune responses but also to contribute to clinical responses [Citation52]. A study showed that a mixed vaccine combining dendritic cells and tumour cells (Lewis lung carcinoma cells) used as immunotherapy significantly delayed tumour growth in decimal lung cancer models, and this treatment strategy may explore new directions for the treatment of lung cancer in the future [Citation53].
This study also identified that GADD45B is associated with mast cells in LUSC. Mast cells exert the effector activity in patients with TH2-skewed autoimmune and allergic inflammation, enhancing the inflammatory responses and activating T cells in co-operation with DCs [Citation54]. Several near-term studies have suggested that mast cells have significant effect on the TIME conformation or promoting cancer development [Citation55,Citation56]. Since GADD45B upregulation manifested the higher expression of some pivotal gene markers of mast cells, we speculated that GADD45B exerts important functions in enhancing the inflammatory responses and activating T cells.
Our results suggest that GADD45B is an important biomarker for poor prognosis. By analysing the data, we can conclude that the upregulation of GADD45B predicts the dismal prognostic outcomes of diverse types of cancers, including LUSC, while synchronously inducing T cell exhaustion, resulting in ineffective antitumor immunity. The simultaneous occurrence of these two conditions further confirms that GADD45B is a rare independent prognostic marker. Therefore, GADD45B may play a crucial part in normal immunity and regulating TIME.
One previous study had demonstrated that the upregulation of GADD45B was closely related to three KEGG pathways, namely, the activated ‘MAPK signal transduction pathway’, the inactivated ‘cell cycle’, and the ‘P53 signal transduction pathway’ [Citation57]. These results were in line with those of our pathway analysis (). Therefore, GADD45B is identified to play a vital role in the pathogenesis of LC.
Collectively, the results of this work collectively demonstrate the potential of GADD45B as a prognostic biomarker for numerous types malignancies, notably LUSC. The upregulation of GADD45B is related to higher degrees of immune infiltration of neutrophils, DCs, mast cells, T cells, and several functional T cells. GADD45B exerts a vital effect on immunity and is highly correlated with exhausted T cells, which might serve as a principal element to facilitate the exhaustion of T cells in LUSC. Detection of the level of GADD45B possibly contributes to prognostic prediction and modulation of the GADD45B levels within exhausted T cells, offering a novel management strategy for the optimisation of the efficacy of anti-LUSC immunotherapy.
5. Conclusions
In summary, our study provides complete proof for the value of GADD45B in the disease progression of patients with lung squamous cell carcinoma and its potential as a biological target and predictive prognostic indicator for lung squamous cell carcinoma. The expression of GADD45B is not only closely related to the prognosis of patients with lung squamous cell carcinoma, but also suggests that it induces T cell failure and is highly related to immune infiltration. Therefore, we speculate that GADD45B plays an important role in enhancing inflammation, normal immunity and regulating TIME, and may become a novel target for researching the immune escape of lung squamous cell carcinoma cells and a therapeutic target for anti-lung cancer immunotherapy. Furthermore, the connection between GADD45B and DC cell markers is also worthy of attention, which may be a new direction for future LUSC research.
Data accessibility
All data generated or analysed during this study are included in this published article and its supplementary information files.
Supplemental Material
Download MS Word (5.1 MB)Acknowledgements
TCGA and GEO belong to public databases. The patients involved in the database have obtained ethical approval. Users can download relevant data for free for research and publish relevant ar-ticles. Our study is based on open source data, so there are no ethical issues and other conflicts of interest.
Disclosure statement
No potential conflict of interest was reported by the authors.
Data availability statement
Data supporting our findings are already included in the manuscript.
Additional information
Funding
References
- Bray F, Ferlay J, Soerjomataram I, et al. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68(6):1–17.
- Gridelli C, Rossi A, Carbone DP, et al. Non-small-cell lung cancer. Nat Rev Dis Primers. 2015;1:15009.
- Gandara DR, Hammerman PS, Sos ML, et al. Squamous cell lung cancer: from tumor genomics to cancer therapeutics. Clin Cancer Res. 2015;21(10):2236–2243.
- Chen Z, Fillmore CM, Hammerman PS, et al. Non-small-cell lung cancers: a heterogeneous set of diseases. Nat Rev Cancer. 2014;14(8):535–546.
- Perez-Moreno P, Brambilla E, Thomas R, et al. Squamous cell carcinoma of the lung: molecular subtypes and therapeutic opportunities. Clin Cancer Res. 2012;18(9):2443–2451.
- Huang J-Z, Chen M, Chen D, et al. A peptide encoded by a putative lncRNA HOXB-AS3 suppresses Colon cancer growth. Mol Cell. 2017;68(1):171–184.e6.
- Sorber L, Zwaenepoel K, Deschoolmeester V, et al. Circulating cell-free nucleic acids and platelets as a liquid biopsy in the provision of personalized therapy for lung cancer patients. Lung Cancer. 2017;107:100–107.
- Mony JT, Schuchert MJ. Prognostic implications of heterogeneity in intra-tumoral immune composition for recurrence in early stage lung cancer. Front Immunol. 2018;9:2298.
- Liu X, Wu S, Yang Y, et al. The prognostic landscape of tumor-infiltrating immune cell and immunomodulators in lung cancer. Biomed Pharmacother. 2017;95:55–61.
- Xiong Y, Wang K, Zhou H, et al. Profiles of immune infiltration in colorectal cancer and their clinical significant: a gene expression-based study. Cancer Med. 2018;7(9):4496–4508.
- Liebermann DA, Tront JS, Sha X, et al. Gadd45 stress sensors in malignancy and leukemia. Crit Rev Oncog. 2011;16(1–2):129–140.
- Takekawa M, Saito H. A family of stress-inducible GADD45-like proteins mediate activation of the stress-responsive MTK1/MEKK4 MAPKKK. Cell. 1998;95(4):521–530.
- Kigar SL, Chang L, Auger AP. Gadd45b is an epigenetic regulator of juvenile social behavior and alters local pro-inflammatory cytokine production in the rodent amygdala. Brain Behav Immun. 2015;46:60–69.
- Zhao Z, Gao Y, Guan X, et al. GADD45B as a prognostic and predictive biomarker in stage II colorectal cancer. Genes. 2018;9(7):361.
- Barros-Filho MC, de Mello JBH, Marchi FA, et al. GADD45B transcript is a prognostic marker in papillary thyroid carcinoma patients treated with total thyroidectomy and radioiodine therapy. Front Endocrinol (Lausanne). 2020;11:269.
- Inowa T, Hishikawa K, Matsuzaki Y, et al. GADD45β determines chemoresistance and invasive growth of side population cells of human embryonic carcinoma. Stem Cells Int. 2010;2010:782967.
- Myint KZ, Kongpracha P, Rattanasinganchan P, et al. Gadd45β silencing impaired viability and metastatic phenotypes in cholangiocarcinoma cells by modulating the EMT pathway. Oncology Letters. 2018;15(3):3031–3041.
- Li T, Fan J, Wang B, et al. TIMER: a web server for comprehensive analysis of tumor-infiltrating immune cells. Cancer Res. 2017;77(21):e108–e110.
- Li B, Li JZ. A general framework for analyzing tumor subclonality using SNP array and DNA sequencing data. Genome Biol. 2014;15(9):473.
- Bhattacharya S, Dunn P, Thomas CG, et al. ImmPort, toward repurposing of open access immunological assay data for translational and clinical research. Sci Data. 2018;5:180015.
- Danaher P, Warren S, Dennis L, et al. Gene expression markers of tumor infiltrating leukocytes. J Immunother Cancer. 2017;5:18.
- Mermel CH, Schumacher SE, Hill B, et al. GISTIC2.0 facilitates sensitive and confident localization of the targets of focal somatic copy-number alteration in human cancers. Genome Biol. 2011;12(4):R41.
- Nirmal AJ, Regan T, Shih BB, et al. Immune cell gene signatures for profiling the microenvironment of solid tumors. Cancer Immunol Res. 2018;6(11):1388–1400.
- Tang Z, Kang B, Li C, et al. GEPIA2: an enhanced web server for large-scale expression profiling and interactive analysis. Nucleic Acids Res. 2019;47(W1):W556–W560.
- Nagy Á, Lánczky A, Menyhárt O, et al. Validation of miRNA prognostic power in hepatocellular carcinoma using expression data of independent datasets. Sci Rep. 2018;8(1):9227.
- Ru B, Wong CN, Tong Y, et al. TISIDB: an integrated repository portal for tumor-immune system interactions. Bioinformatics. 2019;35(20):4200–4202.
- Uhlén M, Fagerberg L, Hallström BM, et al. Proteomics. Tissue-based map of the human proteome. Science. 2015;347(6220):1260419.
- Koch A, de Meyer T, Jeschke J, et al. MEXPRESS: visualizing expression, DNA methylation and clinical TCGA data. BMC Genomics. 2015;16(1):636.
- Vasaikar SV, Straub P, Wang J, et al. LinkedOmics: analyzing multi-omics data within and across 32 cancer types. Nucleic Acids Res. 2018;46(D1):D956–D963.
- Sticht C, de La Torre C, Parveen A, et al. miRWalk: an online resource for prediction of microRNA binding sites. PLoS One. 2018;13(10):e0206239.
- Ravelli A, Roviello G, Cretella D, et al. Tumor-infiltrating lymphocytes and breast cancer: beyond the prognostic and predictive utility. Tumour Biol. 2017;39(4):1010428317695023.
- Azimi F, Scolyer RA, Rumcheva P, et al. Tumor-infiltrating lymphocyte grade is an independent predictor of sentinel lymph node status and survival in patients with cutaneous melanoma. J Clin Oncol. 2012;30(21):2678–2683.
- Ohtani H. Focus on TILs: prognostic significance of tumor infiltrating lymphocytes in human colorectal cancer. Cancer Immun. 2007;7:4.
- Sousa S, Määttä J. The role of tumour-associated macrophages in bone metastasis. J Bone Oncol. 2016;5(3):135–138.
- Siemers NO, Holloway JL, Chang H, et al. Genome-wide association analysis identifies genetic correlates of immune infiltrates in solid tumors. PLoS One. 2017;12(7):e0179726.
- Liu L, Tran E, Zhao Y, et al. Gadd45 beta and Gadd45 gamma are critical for regulating autoimmunity. J Exp Med. 2005;202(10):1341–1347.
- Lu B, Ferrandino AF, Flavell RA. Gadd45beta is important for perpetuating cognate and inflammatory signals in T cells. Nat Immunol. 2004;5(1):38–44.
- Sawant A, Hensel JA, Chanda D, et al. Depletion of plasmacytoid dendritic cells inhibits tumor growth and prevents bone metastasis of breast cancer cells. J Immunol. 2012;189(9):4258–4265.
- Liebermann DA, Hoffman B. Myeloid differentiation (MyD) primary response genes in hematopoiesis. Blood Cells Mol Dis. 2003;31(2):213–228.
- Ju S, Zhu Y, Liu L, et al. Gadd45b and Gadd45g are important for anti-tumor immune responses. Eur J Immunol. 2009;39(11):3010–3018.
- Zumbrun SD, Hoffman B, Liebermann DA. Distinct mechanisms are utilized to induce stress sensor gadd45b by different stress stimuli. J Cell Biochem. 2009;108(5):1220–1231.
- Wang L, Xiao X, Li D, et al. Abnormal expression of GADD45B in human colorectal carcinoma. J Transl Med. 2012;10:215.
- Lee K-W, Lee SS, Hwang J-E, et al. Development and validation of a six-gene recurrence risk score assay for gastric cancer. Clin Cancer Res. 2016;22(24):6228–6235.
- Li L, Cai S, Liu S, et al. Bioinformatics analysis to screen the key prognostic genes in ovarian cancer. J Ovarian Res. 2017;10(1):27.
- Wherry EJ. T cell exhaustion. Nat Immunol. 2011;12(6):492–499.
- Wherry EJ, Kurachi M. Molecular and cellular insights into T cell exhaustion. Nat Rev Immunol. 2015;15(8):486–499.
- Zarour HM. Reversing T-cell dysfunction and exhaustion in cancer. Clin Cancer Res. 2016;22(8):1856–1864.
- Ribas A, Wolchok JD. Cancer immunotherapy using checkpoint blockade. Science. 2018;359(6382):1350–1355.
- Anderson AC, Joller N, Kuchroo VK. Lag-3, tim-3, and TIGIT: co-inhibitory receptors with specialized functions in immune regulation. Immunity. 2016;44(5):989–1004.
- Kastenmüller W, Kastenmüller K, Kurts C, et al. Dendritic cell-targeted vaccines–hope or hype? Nat Rev Immunol. 2014;14(10):705–711.
- Pyfferoen L, Brabants E, Everaert C, et al. The transcriptome of lung tumor-infiltrating dendritic cells reveals a tumor-supporting phenotype and a microRNA signature with negative impact on clinical outcome. Oncoimmunology. 2017;6(1):e1253655.
- Stevens D, Ingels J, van Lint S, et al. Dendritic cell-based immunotherapy in lung cancer. Front Immunol. 2020;11:620374.
- Chen X, Liu Z, Huang Y, et al. Superior anti-tumor protection and therapeutic efficacy of vaccination with dendritic cell/tumor cell fusion hybrids for murine Lewis lung carcinoma. Autoimmunity. 2014;47(1):46–56.
- Dudeck A, Köberle M, Goldmann O, et al. Mast cells as protectors of health. J Allergy Clin Immunol. 2019;144(4S):S4–S18.
- Eissmann MF, Dijkstra C, Jarnicki A, et al. IL-33-mediated mast cell activation promotes gastric cancer through macrophage mobilization. Nat Commun. 2019;10(1):2735.
- Komi DEA, Redegeld FA. Role of mast cells in shaping the tumor microenvironment. Clin Rev Allergy Immunol. 2020;58(3):313–325.
- Jin X, Liu X, Zhang Z, et al. Identification of key pathways and genes in lung carcinogenesis. Oncol Lett. 2018;16(4):4185–4192.