Background

Cancer is a common cause of human mortality and a severe obstacle to the goal of improving life expectancy worldwide. Cancer morbidity and mortality are growing globally, with over 19 million newly diagnosed individuals and more than 10 million deaths occurring in 2020 [1]. Traditional cancer treatments, such as radiation, chemotherapy, and targeted therapies, benefit cancer patients; however, some limitations of these treatments have been identified in recent years. For example, they may cause irreparable DNA damage, produce more off-target effects, and disrupt the tumor microenvironment [2,3,4]. For this reason, combining immunotherapy and conventional therapies has emerged as a powerful clinical strategy for cancer treatment. Thus, performing multi-center studies on various cancers is important for exploring potential biomarkers suitable for multiple cancers, as this may provide a basis for investigating cancer immunotherapy targets [5].

One promising target is the protein encoded by centrosomal protein 55 (CEP55) (also known as CT111 and C10orf3). CEP55 is an essential component of the CEP family and an important factor regulating mitotic termination and cytoplasmic division, with localization at different locations important in the cell cycle, such as the centrosome or spindle [6, 7]. Previous reports have demonstrated elevated CEP55 expression in breast cancer [8], liver cancer [9], colon cancer [10], and lung adenocarcinoma (LUAD) [11]. This upregulated expression has also shown significant correlations with stage, metastasis, and unfavorable clinical outcome in various tumors [7]. CEP55 can affect the PI3K/AKT pathway and cell cycle, and the knockdown of CEP55 can inhibit tumor cell viability and proliferation and can even lead to tumor cell death [12, 13]. Some studies have also reported that a high expression of CEP55 promotes tumor development, metastasis, and aggression by activating the PI3K/AKT signal pathway [14]. Increasing evidence now suggests a clear association between CEP55 upregulation and the development and progression of multiple malignancies. However, research on the involvement of CEP55 in pan-cancer is lacking, and the clinical significance and potential mechanisms of this gene in multiple cancers require further elucidation.

This study is the first to focus on CEP55 in pan-cancer. Public and in-house data were used to investigate the expression of CEP55 at the messenger RNA (mRNA) and protein levels, with the aim of revealing the conspicuous clinical relevance of CEP55 in multiple cancers. An extensive investigation of lung squamous cell carcinoma (LUSC), based on internal tissue microarrays and multi-center samples, was also performed to validate the clinical significance of CEP55 in cancers. In conclusion, the findings of this study, which focused on the importance of CEP55 expression in various tumors, identified CEP55 as a potential novel immune-related predictive and prognostic marker for multiple cancers.

Materials and methods

Collection of pan-cancer expression data

The Depmap Portal is an extensive database of missing cell function screens that can be used to study the underlying dependence of multiple genes in human cancer cell lines, thereby allowing assessment of the essential roles of those genes [15, 16]. For this study, a dataset (n of samples = 1,068) including CRISPR (clustered regularly interspaced short palindromic repeats) Chronos scores of CEP55 were downloaded from the Depmap Portal.

The Cancer Genome Atlas (TCGA) sample data, including 33 cancer types and their 21 control tissue types, were downloaded from the Xena database. Three sample categories – primary tumors, normal solid tissue, and primary hematogenous cancer peripheral blood – from the TCGA cohort were also included in this study, including 9,358 cancer samples and 722 control samples. Details of the 33 cancers in the TCGA cohort are listed in Table S1, which also provides the abbreviations for these cancers.

The pan-cancer protein level data (n of samples = 1,719) [17,18,19,20,21,22,23] used in this study originated from the Clinical Proteomic Tumor Analysis Consortium, available from Proteomic Data Commons (https://pdc.cancer.gov).

Clinical relevance of CEP55 expression in pan-cancer

CRISPR screens allow phenotypic screening and high-throughput sequencing analysis of target cells through the construction of small guide RNA libraries, leading to the exploitation of critical genes [24]. Chronos is a model of cell group dynamics in CRISPR knockout screens [25], and this study identified the critical role of CEP55 in multiple cancers through Chronos. Data from the cell depletion assay can be indicated by the Chronos score, and a Chronos score of < 0 indicates that the gene is necessary for selected cell lines.

The capacity of CEP55 expression to distinguish between tumor and control tissues can be ascertained using receiver operating characteristic (ROC) curves and summary ROC (sROC) curves for sensitivity, specificity, and area under the curve (AUC) values. Larger sensitivity, specificity, and AUC values indicate a more conspicuous ability of CEP55 to identify cancers and their controls.

Prognosis is an essential clinical parameter for cancer patients. This study considered four clinical outcomes reflecting patient prognosis: overall survival (OS), disease-specific survival (DSS), progression-free interval (PFI), and disease-free interval (DFI). For single cancer, the correlation between CEP55 expression and survival indicators for patients with that cancer was investigated using univariate Cox analysis and Kaplan-Meier curves. Cancer patients with various clinical features may have different prognoses; therefore, the relationship between CEP55 expression and the clinical characteristics of the patients was also evaluated. Prognostic data and clinical characteristics (age, sex, and AJCC [American Joint Committee on Cancer] stage) for the above analysis were available from the Xena database.

Data on tumor mutation burden, microsatellite instability, neoantigen number, and immune microenvironment

SangerBox (v3.0) [26] provides data related to tumor mutation burden (TMB), microsatellite instability (MSI) [27], and neoantigen counts [28]. The TIMER algorithm [29] can be applied to detect the level of infiltration by six immune cell types: B cells, CD4 + T cells, CD8 + T cells, neutrophils, macrophages, and dendritic cells. ESTIMATE [30, 31] reflects the immune environment of cancer patients by assigning three other scores: immune score, stromal score, and ESTIMATE score [32]. The ESTIMATE algorithm was used to calculate three scores to evaluate the correlation between CEP55 and the immune microenvironment. The infiltration levels of the six immune cell types calculated by TIMER were available from the official TIMER website, while the three scores identified by ESTIMATE were downloaded from SangerBox (v3.0).

Potential mechanisms of CEP55 in multiple tumors

In this study, the “clusterProfiler” package [33] was applied to perform gene set enrichment analysis (GSEA) [34], with the aim of allowing further exploration of the potential involvement of CEP55 in signaling pathways associated with various cancers. The signaling pathways examined in this analysis were acquired from the Kyoto Encyclopedia of Genes and Genomes (KEGG) database [35,36,37,38].

Research on CEP55 in LUSC

The mRNA expression of CEP55 in LUSC was explored using multi-center microarrays and RNA-Seq datasets [39, 40] from the ArrayExpress, Gene Expression Omnibus, TCGA, and GTEx [41] databases. Datasets were included using the following selection criteria: [1] microarrays or RNA-Seq; [2] tissue samples acquired from Homo sapiens; and [3] containing CEP55 mRNA expression data. The standards for data exclusion were the following: [1] datasets with duplicate data and [2] fewer than three LUSC samples or control samples in merged datasets. This study covered 35 datasets containing 1,390 LUSC and 1,418 non-LUSC samples (Table S2). In addition, 288 LUSC cases from five datasets (GSE29013, GSE30219, GSE37745, GSE73403, and GSE81089) were obtained to analyze the possible association between CEP55 expression with OS in LUSC patients.

The datasets available from public databases were standardized using log2 (x + 1) and the “limma” package [42]. Since each dataset was derived from a different batch of experiments, the original datasets were merged according to the same platform number using the “SVA” package [43] to eliminate batch effects [44]. For instance, the gene expression data from the GSE31552 and GSE44077 datasets were detected using the GPL6244 platform, leading to their consolidation into a merged cohort named GPL6244. In this study, the 35 datasets were divided into 11 new cohorts with the same platform numbers.

To clarify the differences in protein expression between LUSC and non-LUSC patients, immunohistochemistry (IHC) experiments were conducted on four in-house tissue microarrays (LUC481, LUC482, LUC483, and LUC1021). This study used a rabbit anti-human CEP55 monoclonal antibody (EPR11944[B], ab170414, dilution ratio 1:100) and a second antibody-labeled horseradish peroxidase (D-3004-15, Changdao Biotechnology Co., Ltd., Shanghai, China). The steps of the IHC experiments and the protein score criteria were published previously [45].

Statistical analysis

This study evaluated the variance of CEP55 expression levels among tumor and control groups (e.g., LUSC vs. control) using the Wilcoxon rank-sum test and standardized mean difference (SMD). For the SMD, a random effects model was used when significant heterogeneity existed between the data sets, as determined by I2 values > 50% and a chi-square test p-value < 0.1. Otherwise, a fixed-effects model was applied. The outcomes of the SMD were statistically significant in specific cases (the corresponding 95% confidence interval [CI] excludes 0). Begg’s test was applied to evaluate the publication bias of the SMD, and a p value of less than 0.1 indicated a significant publication bias for SMD outcomes.

The Wilcoxon rank-sum test was used to evaluate the correlation of CEP55 expression with age, gender, and AJCC stage. Correlations among the CEP55 expression and TMB, MSI, neoantigen count, TIMER score, and ESTIMATE score were analyzed using Spearman’s correlation coefficient, with p < 0.05 considered to indicate a significant statistical difference. For the hazard ratio (HR), 95% CI excluding 1 or p < 0.05 indicated statistical significance. Stata (v15.0) was used to plot the sROC, and R (v4.1.0) was used to complete all the remaining computation and visualization steps. The research overflow of this study is shown in Fig. 1A.

Results

Expression of CEP55 and its essential role in various cancers

The Chronos score reflected the importance of CEP55 in various cancer cell lines, and a value of <0 indicated that decreased expression of CEP55 increased cancer cell death. According to Fig. 1B, the Chronos scores of CEP55 in the majority of 28 cancer cell lines were less than 0, indicating that CEP55 was essential for tumor cells in some organs, such as bladder cancer, bone cancer, and gastric cancer.

Among the 21 cancers explored, CEP55 mRNA was overexpressed in 20 cancer tissues (except kidney chromophobe [KICH]) when compared with the respective normal tissues (p < 0.05; Fig. 1C). CEP55 protein expression data for nine of the 20 cancers available from the Clinical Proteomic Tumor Analysis Consortium were collected for this research. The nine cancers were BRCA, GBM, HNSCC, KIRC, LIHC, LUAD, LUSC, PRAD, and UCEC. CEP55 protein levels were higher in cancer tissues than in normal tissues in eight of the nine cancers (except for LUSC) (Fig. 1D; p < 0.05), and the findings were consistent with the conclusions corroborating the results at the mRNA level.

Clinical significance of CEP55 expression in pan-cancer

Given the essential roles and distinct expression levels of CEP55 in a wide range of cancers, the clinical relevance of CEP55 in multiple cancers was further investigated. Sixteen of 21 cancer types had an AUC value of > 0.9 for CEP55 expression, indicating a significant ability of CEP55 expression to distinguish these 16 cancer tissues from their control tissues (Fig. 2A). For the 21 investigated cancers, the results of the sROC analysis showed that CEP55 expression was highly accurate in identifying 21 types of cancers (sensitivity = 0.91, specificity = 0.93, AUC = 0.97; Fig. 2B).

The study also explored the prognostic value of CEP55 in distinct cancers. Upregulation of CEP55 was associated with poor OS and DSS in patients with ACC, KICH, KIRC, KIRP, LGG, LIHC, LUAD, MESO, PAAD, PRAD, and UVM (HR > 1, p < 0.05; Fig. 3A–D). Elevated CEP55 expression also indicated unfavorable DSS results in GBM patients (HR > 1, p < 0.05; Fig. 3 C–D). By contrast, in patients with BRCA or THYM, upregulated CEP55 expression implied that they had longer OS (HR < 1, p < 0.05; Fig. 3A–B). Furthermore, high CEP55 expression suggested shorter DFI for patients with certain cancers (KIRP, LIHC, LUAD, PAAD, SARC, and THCA) and shorter PFI for those suffering from ACC, ESCA, KICH, KIRC, KIRP, LGG, LIHC, LUAD, MESO, PAAD, PCPG, PRAD, SARC, and UVM (HR > 1, p < 0.05; Fig. 4A–D). Notably, upregulated CEP55 expression represented poor prognostic indicators (i.e., OS, DSS, DFI, and PFI) for patients with one of four cancers: KIRP, LIHC, LUAD, and PAAD.

Fig. 1
figure 1

The research overflow, essential roles, and expression of CEP55 in cancers. Panel A: The research overflow of this study. Panel B: Identification of essential roles of CEP55 for multiple cancers. Panel C: The differential expression of CEP55 mRNA between cancers and controls; p-value was based on the Wilcoxon rank-sum test with a false discovery rate. Panel D: The differential levels of CEP55 protein between cancers and controls. ns/NSp > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001

Full size image
Fig. 2
figure 2

The ability of CEP55 to differentiate the tumor tissue from control tissue. Panel A: CEP55 can accurately distinguish cancer tissues from control tissues in some cancers. Panel B: CEP55 distinguishes cancers well from control tissues in 21 cancer types

Full size image
Fig. 3
figure 3

Relation of CEP55 expression with overall survival and disease-specific survival of cancer patients. Panels A and C: CEP55 expression was related to the prognosis of patients in most cancers. Panels B and D: CEP55 expression was related to the prognosis of patients in some cancers; the red curve represents the high-CEP55 expression group, while the blue curve represents the low-CEP55 expression group. *p < 0.05

Full size image
Fig. 4
figure 4

Relation of CEP55 expression with disease-free interval and progression-free interval of cancer patients. Panels A and B: CEP55 expression was related to the prognosis of patients in most cancers. Panels B and D: CEP55 expression was related to the prognosis of patients in some cancers; the red curve represents the high-CEP55 expression group, while the blue curve represents the low-CEP55 expression group. *p < 0.05

Full size image

Correlation of CEP55 expression with clinical features

As shown above, CEP55 may be a prognostic marker for numerous cancers. Therefore, the association between CEP55 expression levels and clinical characteristics (e.g., distant metastasis) of cancer patients was examined. Fig. 5 A reveals higher CEP55 expression levels in patients with distant metastasis than in those without for seven cancers: ACC, LUSC, PAAD, PCPG, PRAD, SARC, and UVM (p < 0.05). Notably, CEP55 emerged as a prognostic risk factor in almost all seven of these cancers, as depicted in Figs. 3 and 4. Thus, both results verified each other. In contrast to the aforementioned cancers, a statistical difference was not noted in CEP55 expression levels for patients with distinct metastasis status in the remaining 15 cancers (Fig. 5A). Similar results were observed for different AJCC stages, ages, and genders (Fig. S1–S3). These findings suggest that, for most cancers, CEP55 expression is not affected by the distant metastasis status, AJCC stages, ages, or genders of cancer patients and can be considered an independent marker.

Relevance of CEP55 expression to TMB, MSI, neoantigen count, and immune microenvironment

TMB is recognized as a potential prognostic biomarker for cancers [46, 47]. CEP55 expression was positively related to TMB in ACC, PAAD, LGG, SARC, and BRCA. On the contrary, the expression levels of CEP55 in THYM, KIRP, THCA, and UVM were negatively correlated with TMB (Fig. 5B).

A microsatellite is a short type of tandem repeat DNA train consisting of one to ten nucleotides [48]. CEP55 expression demonstrated a positive correlation with MSI in six cancers—TGCT, SARC, LUSC, OV, KIRC, and BRCA—and a negative relevance with PRAD, HNSCC, and THCA (Fig. 5C).

A few somatic mutations in tumor DNA can generate immunogenic neoantigens, and these immunogenic peptides can recognize the immune system and target activated T cells [47]. The finding of a correlation of CEP55 expression with TMB and MSI prompted an investigation of the relationship between CEP55 expression and neoantigens. The results revealed a weak correlation between CEP55 expression and neoantigen counts in four tumors (ACC, LUAD, COAD, and PRAD) (Fig. 5D).

The immune response plays a vital role in antitumor function; therefore, the immune correlation of CEP55 in pan-cancer was also explored. The TIMER algorithm demonstrated that CEP55 expression levels tended to positively correlate (except for neutrophils in THYM) with increased infiltration levels of six types of immune cells (especially for B cells, macrophages, and dendritic cells) in THYM, THCA, and LIHC (ρ > 0.3, p < 0.05; Fig. 6A). Furthermore, based on the ESTIMATE algorithm, a moderate positive relationship between CEP55 expression and scores of stromal, immune, and ESTIMATE was detected in THCA and KIRC, while a weak negative correlation between CEP55 expression and the three scores in STAD and SKCM was observed (p < 0.05; Fig. 6B).

Mechanistic prediction of CEP55 in multiple tumors

GSEA was performed in this research to explore the underlying mechanisms of CEP55 in multiple tumors. The peak appeared in the high-expression group, indicating a positive correlation between these pathways and CEP55 expression, and that these signaling pathways were more active when CEP55 expression was elevated. CEP55 expression was correlated with at least five KEGG signaling pathways in six cancers (Fig. 7). Notably, CEP55 was most likely to affect the “olfactory transduction” pathway in cancers (up to 18 cancer types) and it may could play an important role in “metabolism of xenobiotics by cytochrome P450,” “drug metabolism cytochrome P450,” and “cytokine-cytokine receptor interaction” (Table S3).

Fig. 5
figure 5

Relation of CEP55 expression with tumor distant metastasis (panel A), tumor mutational burden (panel B), microsatellite instability (panel C), and neoantigen (panel D) of cancer patients. nsp > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001

Full size image
Fig. 6
figure 6

Relation of CEP55 expression and infiltration levels of immune cells. Panel A: TIMER algorithm; B: ESTIMATE algorithm

Full size image
Fig. 7
figure 7

Signaling pathways potentially affected by CEP55 in multiple cancers

Full size image

Expression of CEP55 in LUSC

The differential expression and important clinical value of CEP55 have been observed in several cancers, but little is known about the gene in LUSC. Therefore, using multi-center data, a comprehensive investigation of CEP55 expression in LUSC was performed in this research. The random effects model showed elevated CEP55 mRNA expression in the LUSC group but not in the non-LUSC group (SMD = 4.07, 95% CI: 3.22–4.91) (Fig. 8A), and no significant publication bias was detected (p > 0.1; Fig. 8B). Regarding each dataset of the 11 merged datasets, the LUSC group showed increased levels of CEP55 mRNA expression compared with the non-LUSC group (p < 0.05; Fig. 8C).

An in-house IHC experiment conducted to verify the CEP55 expression in LUSC at protein levels revealed substantially higher CEP55 protein levels in LUSC tissues than in non-LUSC tissues (p < 0.05; Fig. 8D), consistent with the CEP55 expression at the mRNA levels. Positive CEP55 protein staining was clearly observed by microscopy in LUSC tissues but not in control tissues (Fig. 8E and F).

Clinical relevance of CEP55 expression in LUSC

As shown in Fig. 9A, ROC curves confirmed the ability of CEP55 mRNA expression to distinguish LUSC samples from control samples with high accuracy (AUC > 0.8), similar to the findings from the analysis of various cancers. Moreover, the sROC analysis revealed that SCLC samples could be distinguished from non-SCLC samples based on CEP55 expression (AUC > 0.95, Fig. 9B). These results demonstrate the conspicuous potential for using CEP55 expression to distinguish LUSC patients from individuals without LUSC.

As shown in Fig. 9C, high CEP55 mRNA expression was correlated with unfavorable OS. In detail, LUSC patients with overexpression of CEP55 had more pessimistic OS. The statistically significant result was evident in the GSE29013 (p < 0.05; Fig. 9C), while the remaining four cohorts also suggested a trend to an adverse risk associated with CEP55 expression in the prognosis of LUSC patients (Fig. 9D).

Fig. 8
figure 8

CEP55 mRNA and protein levels in LUSC. Panel A: CEP55 mRNA expression forest plot in LUSC and control tissues. Panel B: Begg’s test for SMD results. Panel C: Violin plots of CEP55 mRNA expression in each dataset. Panel D: The violin plot of CEP55 protein levels. Panels E–F: Microscopic images of CEP55 protein levels in non-LUSC (E) and LUSC (F) tissues; the value (e.g., 100X) at the bottom of each microscope image represents the magnification scale of the microscope. The p-value was calculated based on the Wilcoxon rank-sum test. *p < 0.05; ***p < 0.001

Full size image
Fig. 9
figure 9

Clinical value of CEP55 expression in LUSC. Panels A–B: CEP55 distinguishes LUSC well from control tissues. Panels C–D: The correlation of CEP55 expression with the overall survival of LUSC patients; the red curve represents the high-CEP55 expression group, while the blue curve represents the low-CEP55 expression group

Full size image

Discussion

To our knowledge, previous comprehensive research is lacking regarding the expression, clinical value, and potential mechanisms of CEP55 in pan-cancer, making the present report the first relevant study. The essential roles of CEP55 in multiple cancers were identified using a CRISPR dataset of 1,068 specimens. Overall, the findings from 11,799 samples of various cancers revealed elevated CEP55 expression levels in 20 cancers. Overexpression of CEP55 mRNA was a significant predictive and prognostic indicator in various cancers. The expression of CEP55 was relevant to TMB, MSI, neoantigen counts, and immune cell infiltration in a variety of cancers, indicating that the gene may have an essential role in the immune response. Taking LUSC as an example, the results of this study partly verified the expression level and clinical relevance of CEP55 in cancers based on the analysis of internal tissue microarray and multi-center LUSC data (n of samples = 2,956). In conclusion, CEP55 may be an immune-related predictive and prognostic marker for multiple cancers, including LUSC.

The Chronos scores indicated that, CEP55 is essential for various tumors originating from some organs, based on the comprehensive pan-cancer analysis of CEP55 performed in this study. Differential expression of CEP55 is common in cancer, and its elevated expression has been reported several times. For example, Hauptman et al. [49] demonstrated the upregulation of CEP55 in colorectal cancer using in-house clinical samples and public data. Yang et al. [9] determined overexpression of CEP55 in LIHC and attributed it to DNA hypomethylation. Jiang et al. [50] identified elevated CEP55 expression in non-small-cell lung cancer using 203 specimens. However, no previous investigation has explored CEP55 expression in pan-cancer. Therefore, our study attempted to fill this void, and its findings revealed upregulated CEP55 expression in 20 cancers (e.g., BLCA). High CEP55 expression has also been verified previously in BRAC, CESC, ESAC, and KIRC [51,52,53,54]. Our study results also identified overexpression of CEP55 in CHOL, KICH, KIRP, PRAD, and UCEC. The distinct CEP55 expression between cancers and controls implies that CEP55 may have underlying clinical significance in an extensive range of cancers.

Our findings identify CEP55 expression as a predictor for determining cancer status and indicate a relationship with the prognosis of several cancers. CEP55 expression provided the ability to distinguish specific cancer tissues (KICH, etc.) from control tissues with high accuracy. To the best of our knowledge, this result has not been previously reported, demonstrating the novelty of our study. To date, previous studies have investigated the prognosis value of CEP55 expression in some cancers, including ACC, BRCA, GBM, KIRC, LIHC, LUAD, and PAAD [9, 11, 14, 55,56,57,58]. The inclusion of more than just these cancers in our study allowed for an analysis of the prognostic value of CEP55 expression in 33 tumors. Based on our study findings, CEP55 expression is an OS risk factor for certain cancers (ACC, KICH, KIRC, KIRP, LGG, LIHC, LUAD, MESO, PAAD, PRAD, and UVM) and a protective factor for BRCA and THYM. For DSS, CEP55 can also be a risk factor for 12 cancers (ACC, etc.). Furthermore, high expression of CEP55 indicated an unfavorable DFI for patients with KIRP, LIHC, LUAD, PAAD, SARC, and THCA, as well as poor PFI for those diagnosed with one of 14 cancers: ACC, ESCA, KICH, KIRC, KIRP, LGG, LIHC, LUAD, MESO, PAAD, PCPG, PRAD, SARC, and UVM. Overall, CEP55 may represent a predictive and prognostic marker for multiple cancers.

CEP55 may have different effects on the immune microenvironment in different cancers. On the one hand, CEP55 may participate in the negative immune repose in certain cancers (e.g., STAD, and SKCM), since it shows an adverse relationship with the immune microenvironment in these cancers. On the other hand, CEP55 may present an immune response activation factor for the following reasons. First, high levels of TMB and MSI frequently promote the occurrence of neoantigens, and neoantigens tend to stimulate an immune response [46]. Notably, based on our study results, CEP55 was not only positively related to TMB and MSI in several cancers (SARC, etc.) but it was also positively related to neoantigens in certain cancers (ACC, etc.). Second, dendritic cells are often considered the most functional professional antigen-presenting cells in the adaptive immune response of humans, responsible for transporting tumor antigens and activating antitumor T cells [59]. Interestingly, CEP55 expression was positively correlated with six immune cell types (not only dendritic cells, but also B cells, CD4 + T cells, CD8 + T cells, neutrophils, and macrophages) in specific cancers, namely — THYM, THCA, and LIHC (except for neutrophils in THYM). Third, a weak positive correlation was detected between CEP55 expression and three immune scores in certain cancers (e.g., THCA and KIRC). Thus, our study results indicated a close relationship between CEP55 expression and the immune environment. However, the current findings require further experimental validation.

The underlying mechanisms of CEP55 in various cancers remain complex; for example, CEP55 expression was correlated with at least five KEGG signaling pathways in six cancers (ACC, etc.). Some signaling pathways may be critical for CEP55 involvement in multiple cancers, including “olfactory transduction,” “metabolism of xenobiotics by cytochrome P450,” “drug metabolism cytochrome P450,” and “cytokine-cytokine receptor interaction” signaling pathways. Among these signaling pathways, the CEP55-affected activation of “cytokine-cytokine receptor interaction” signaling pathway directly highlights the association between CEP55 expression and immune cell infiltration levels to some extent. Cytokines are secreted proteins primarily produced by various cell types, particularly immune cells, and some of them play a vital role in the antitumor process [60, 61]. During the antitumor process, immune cells are activated and produce cytokines. When these cytokines bind to their respective receptors, they transmit signals between cells and regulate various biological processes, including cell growth, differentiation, proliferation, and cell death. Notably, certain cytokines such as IL-2 and IL-15 can influence the proliferation of immune cells, thereby contributing to the regulation of immune cell infiltration levels [60, 62]. Thus, a correlation exists between the level of immune cell infiltration and some signaling pathways (at least the “cytokine-cytokine receptor interaction” signaling pathway). Considering the association of CEP55 expression with immune cell infiltration and the potential activation status of the “cytokine-cytokine receptor interaction” signaling pathway, CEP55 may play an important role in tumors through its effects on these two factors. Taken together, our study results provide a clue for further experimental explorations into the possible molecular mechanisms of CEP55 in pan-cancer.

With the view that CEP55 is likely to play a vital role in various cancers, an attempt was made to validate the expression and clinical significance of CEP55 in one specific tumor type (i.e., LUSC). Previously, Fu et al. [63] and Shi et al. [64] investigated CEP55 in LUSC; however, they included small-sized samples (n < 200), and they did not verify their results with multi-center and internal samples, raising the possibility of some limitations. For instance, Fu et al. [63] did not find an association of CEP55 with the prognosis of patients with LUSC, which was identified in our study in one of the five cohorts; this also indicates an advantage of studies based on multiple cohorts. Notably, the use of internal tissue microarrays and multi-center LUSC data in our study revealed elevated CEP55 mRNA and protein levels in LUSC. Furthermore, CEP55 could identify LUSC, and high CEP55 expression was associated with poor prognosis in LUSC patients, which partly supports the results in the pan-cancer analysis.

Our study had some limitations. The number of body-fluid samples used in this study could be expanded, as this would facilitate the validation of the mRNA and protein levels of CEP55 in pan-cancer analysis and the correlation between CEP55 and prognosis. A need remains to include in vivo and in vitro experiments to support the study of the molecular mechanisms of CEP55 in multiple cancer cells. Future studies should include internal samples from multiple cancers, not just from LUSC.

In conclusion, our study provides findings that offer a comprehensive assessment of CEP55 and identify the significant clinical value of CEP55 in pan-cancer. CEP55 may be an immune-related predictive and prognostic marker for certain cancers, including LUSC.