Introduction

Gastrointestinal (GI) tumors including esophageal carcinoma (ESCA), stomach adenocarcinoma (STAD), pancreatic adenocarcinoma (PAAD), colon adenocarcinoma (COAD), rectum adenocarcinoma (READ) and liver hepatocellular carcinoma (LIHC) are the major malignancies globally with significant mortality and morbidity [1,2,3,4,5,6]. Because the treatment for patients with, especially, advanced GI tumors are limited and the patient prognosis is poor [6,7,8,9], it is necessary to explore novel diagnostic and therapeutic markers for GI tumors.

Centromere protein L (CENPL) is a member of the centromere protein family. It is assembled with other centromere proteins to form the constitutive centromere associated network (CCAN), also known as inner kinetochore [10,11,12]. CENP-L-N complex plays an important role in centromere specificity and kinetochore stability [12]. The depletion phenotype of CENP-L is crucial to the stability of kinetochore microtubules, for a homogenous poleward microtubule flux rate and for the kinetochore pushing force [13]. These studies suggest that CENPL is indispensable during normal cell division. However, the understanding of CENPL in cancer is still limited. Yin et al. found that CENPL might be a diagnostic and prognostic marker for breast cancer [14]. In our previous study it was found that CENPL was overexpressed in hepatocellular carcinoma and predicted the prognosis of patients, and the overexpression of CENPL was positively correlated with the abundance of various tumor infiltrating lymphocytes [15]. Whether CENPL plays a role in the development and progression of other GI tumors remains unknown, and more studies are needed.

In this study, we investigated the expression and prognostic value of CENPL in GI tumors, including ESCA, STAD, PAAD, COAD and READ, by using bioinformatic methods and immunohistochemical staining.

Methods

Analysis of CENPL gene expression

The level of CENPL mRNA in esophageal, gastric, colon and pancreatic tumors and adjacent normal tissues was analyzed with Oncomine (https://www.oncomine.org/) [16] and GEPIA: Gene Expression Profiling Interactive Analysis (http: //gepia.cancer-pku.cn) [17] databases. The default settings were used in all analyses, and Fold Change (FC) of ≥ 2 and a p value of < 0.05 were considered statistically significant.

Gene expression and clinical characteristics

OncoLnc (http://www.oncolnc.org/) [18] and GEPIA database were used to analyze the prognosis of patients with ESCA, STAD, PAAD, COAD and READ. For each cancer, the patients were divided into high- and low-level groups according to the gene expression level, and the cutoff was 50%. UALCAN (http://ualcan.path.uab.edu/index.html) [19] database was used to analyze the relationship between the expression of CENPL mRNA and clinical characteristics including TP53 mutation, age and sex. A p value of < 0.05 was considered statistically different.

CENPL gene expression and clinicopathological data regarding TCGA PAAD was from the UCSC Xena database (https://xenabrowser.net/) [20]. Univariate and multivariate Cox regression analyses were performed using survival package of R software (Version 3.6.3) to assess Hazard Ratio (HR) and p-values of prognostic risks for PAAD. The survminer package was used for visualization. The downloaded PAAD data included 181 patients' CENPL expression (FPKM) and clinicopathological data including sex, age, tumor size and survival. Of those, 82 were female and 99 were male, and 167 had tumor size data, with a median size of 3.5 cm, (Supplementary table 1).

Mutation and levels of immune cell infiltration

The cBioPortal (https://www.cbioportal.org/) [21] database was used to analyze the mutations of CENPL in tumors. The relationship between CENPL mRNA and levels of immune cell infiltration was analyzed with the TIMER: Tumor Immune Estimation Resource (https://cistrome.shinyapps.io/timer/) [22] database. A p value of < 0.05 and a correlation coefficient value of ≥ 0.5 were considered statistically significant.

Analysis of potential regulatory mechanisms

The miRWalk (http://mirwalk.umm.uni-heidelberg.de/) [23] database was used to predict potential microRNAs (miRNAs) that regulate CENPL mRNA. Cytoscape software (Version: 3.7.2) was used to screen and visualize top 9 potential miRNAs associated with CENPL expression. The Encyclopedia of RNA Interactomes (ENCORI, http://rna.sysu.edu.cn/encori/mirTarPathways.php) database was used to analyze the expression of miRNA candidates in cancers [24]. Pan-cancer Analysis Platform model was selected. A p value of < 0.05 were considered statistically different.

Immunohistochemical Staining

The study was conducted in accordance with the principles of the Declaration of Helsinki [25]. The specimen collection procedure was approved by the Hospital Ethics Committee. The need for informed consent was waived by the Ethics Committee of Dongfang Hospital. The investigation was approved by the Ethics Committee of Dongfang Hospital.

Five-μm sections were obtained from paraffin-embedded PAAD tumor and non-tumor tissues (From Department of Pathology, Dongfang Hospital). All these sections were dewaxed in xylene and re-hydrated in alcohol followed by wet autoclave pretreatment in citrate buffer for antigen retrieval (10 min at 120℃, pH = 6.0) and then rinsed with phosphate buffer saline. Immunohistochemical staining for antibody to CENPL (Rabbit: bs13836R, Beijing Biosynthesis Biotechnology, China) was performed using the avidin–biotin- peroxidase complex method. The primary antibody (1:200 dilution for CENPL) was applied to the sections and allowed to react for 1 h at room temperature. The sections were then incubated with biotinylated anti-mouse/rabbit antibody for 30 min and avidin–biotin-peroxidase reagent for 20 min. After color development with diaminobenzidine, the sections were counterstained with hematoxylin.

Results

Levels of CENPL mRNA in ESCA, STAD, PAAD, COAD and READ

Levels of CENPL mRNA in colorectal and gastric cancers and PAAD were higher than those in para-tumor tissues (p < 0.01 for all, Fig. 1A-C and Table 1). The range of fold change (FC) in COAD were 2.168–3.007 (Fig. 1A and Table 1). In STAD it was 2.194–2.502 (Fig. 1B and Table 1). FC was 2.039 in PAAD (Fig. 1C and Table 1). Using the defined threshold, no results were retrieved for esophageal carcinoma. Analysis of GEPIA database showed that the level of CENPL mRNA in 182 cases of ESCA was significantly higher than that in 286 cases of normal tissue (p < 0.01, Fig. 1D). CENPL mRNA was significantly overexpressed in 408 STAD cases than in 211 cases of normal tissues (p < 0.01, Fig. 1E). In 275 cases of COAD and 349 cases of normal tissues, CENPL mRNA was significantly higher in tumor (p < 0.01, Fig. 1F). Results of 92 cases of READ and 318 cases of normal tissues showed significantly increased levels of CENPL mRNA in tumors (p < 0.01, Fig. 1G). The expression of CENPL mRNA in 179 cases of PAAD was significantly higher than that in 171 cases of normal pancreatic tissues (p < 0.01, Fig. 1H).

Fig. 1
figure 1

The level of CENPL mRNA in ESCA, STAD, COAD, READ and PAAD tumor and normal tissues. Higher levels of CENPL mRNA in gastric and colorectal cancers and PAAD tumor than normal tissues, p < 0.01, (A-C), (Oncomine) The level of CENPL mRNA in ESCA, STAD, COAD, READ and PAAD was higher than that in normal tissues, p < 0.01, (D-H), ( GEPIA).

Full size image
Table 1 The FC of CENPL mRNA in gastrointestinal tumors (Oncomine)
Full size table

Relationship between the level of CENPL mRNA and the prognosis of patients with ESCA, STAD, COAD, READ and PAAD

In 174 patients with PAAD, those with higher levels of CENPL mRNA had shorter survival (p < 0.05, Fig. 2A). In 144 patients with ESCA, there was no significant difference in the survival between the high- and low-expression groups (p > 0.05, Fig. 2B). No significant difference was found in the survival of STAD patients between the high- and low-expression groups (p > 0.05, Fig. 2C). In 440 patients with COAD, no significant difference in the survival was observed between high- and low- expression groups (p > 0.05, Fig. 2D). There was no significant difference in the survival of READ patients between high- and low- expression groups (p > 0.05, Fig. 2E).

Fig. 2
figure 2

The prognostic value of CENPL mRNA in gastrointestinal tumors. PAAD patients with high levels of CENPL mRNA had shorter survivalp < 0.05, (A), (OncoLnc). There was no significant correlationship between CENPL mRNA and the survival of patients with ESCA, STAD, COAD and READ, p > 0.05, (B-E), (OncoLnc). PAAD patients with higher levels of CENPL mRNA have shorter OS and DFS, p < 0.05, (F-G), (GEPIA)

Full size image

In addition, GEPIA was analyzed regarding the prognostic value of CENPL mRNA in patients with PAAD. The results showed that patients with higher levels of CENPL mRNA had shorter overall survival (OS) (p < 0.01, Fig. 2F). Similarly, patients with higher expression of CENPL mRNA had shorter disease-free survival (DFS) (p < 0.01, Fig. 2G).

Univariate Cox regression analysis showed that CENPL mRNA was a risk factor for patient prognosis (HR = 1.9, CI = 1.1–3.1, p < 0.05) (Fig. 3A). Multivariate Cox regression analysis included CENPL mRNA, age, sex and tumor size (dichotomized with a cut-off value of 3.5 cm) and CENPL mRNA (HR = 1.74, CI = 1.05–2.9, p < 0.05) and age (HR = 1.03, CI = 1.01–1.1, p < 0.05) were independent risk factors for the prognosis of PAAD patients (Fig. 3B). Sex and tumor size were not independent risk factors forthe prognosis of PAAD patients (both p > 0.05) (Fig. 3B).

Fig. 3
figure 3

Univariate and multivariate Cox regression analyses of prognostic risk factors in PAAD patient. Univariate Cox regression analysis showed that CENPL mRNA was a risk factor for the prognosis of PAAD patients (HR = 1.9, p < 0.05) (A). Multivariate Cox regression analysis showed that CENPL and age were risk factors for patient prognosis (B), whereas sex and tumor size had no significant effects on the prognosis (B)

Full size image

CENPL mutation and the correlation between its expression and tumor infiltrating lymphocytes

Gene mutation analysis showed that only 2% (17/850) of 850 patient from four independent datasets had mutations, and the main type of CENPL mutation was amplification (Fig. 4A). No significant correlations were observed between the expression of CENPL mRNA and PAAD tumor purity and levels of immune cell infiltration including B Cells, CD8 + T Cells, CD4 + T Cells, macrophages, neutrophils and dendritic cells (|Cor|< 0.5, Fig. 4B). However, it was found that the expression of CENPL mRNA was significantly associated with TP53 mutations in PAAD. In 82 PAAD patients with TP53 mutation, the level of CENPL mRNA was significantly higher than that in 93 patients without TP53 mutation (p < 0.01, Fig. 4C).

Fig. 4
figure 4

The mutations of CENPL and the relationship between the expression of CENPL mRNA and the abundance of tumor immune cell infiltration in PAAD. The mutation rate of CENPL in PAAD was 2% (17/850), (A), (cBioPortal). There was no significant correlation between the expression of CENPL mRNA and the abundance of immune cell infiltration in PAAD | Cor |< 0.5, (B), (TIMER). CENPL mRNA was significantly overexpressed in PAAD with TP53 mutation, p < 0.01, (C), (UALCAN)

Full size image

miRNAs potentially regulating CENPL in PAAD

1086 miRNAs (supplement table 2) that potentially regulate CENPL mRNA were identified using the miRWalk database. Cytoscape (Version: 3.7.2) screened the top 9 potential miRNAs with the strongest association with CENPL mRNA: hsa-miR-371a-3p, hsa-miR-382-5p, hsa-miR-340-3p, hsa- miR-331-5p, hsa-miR-324-5p, hsa-miR-433-5p, hsa-miR-483-3p, hsa-miR-484 and hsa-miR-511-5p (Fig. 5A). Further analysis showed significant low expression of hsa-miR-340-3p in PAAD (FC = 0.44, p < 0.01 and FDR < 0.05, Fig. 5B). hsa-miR-484 was also significantly under-expressed (FC = 0.41, p < 0.01 and FDR < 0.05, Fig. 5C). Moreover, it was found that PAAD patients with lower expression of hsa-miR-340-3p had shorter survival (p < 0.05, Fig. 5D). The expression of hsa-miR-484 was not significantly correlated with the survival of PAAD patients (p = 0.177, Fig. 5E).

Fig. 5
figure 5

The miRNAs that potentially regulate the expression of CENPL mRNA in PAAD. The top 9 miRNAs with the potential to regulate CENPL mRNA expression in PAAD, (A), (miRWalk). miR-340-3p and miR-484 were significantly under-expressed in PAAD patients, (B-C), (ENCORI). PAAD patients with lower miR-340-3p expression have shorter survival (p < 0.05) (D), (OncoLnc). There was no significantly correlation between miR-484 expression and the survival of PAAD patients (p > 0.05, OncoLnc) (E)

Full size image

The expression of CENPL protein in PAAD and normal tissues

Immunohistochemical staining of clinical specimens showed that the expression of CENPL protein in PAAD was significantly higher than that in adjacent normal tissues. The expression of CENPL in tumor tissues (42%, 11/26) was higher than that in non-cancer tissues (Fig. 6). CENPL protein was mainly located in cytoplasm and membrane (Fig. 6).

Fig. 6
figure 6

The expression of CENPL protein in PAAD tumor and normal tissues. Immunohistochemical staining showed that the expression of CENPL protein in tumor was higher than in normal tissues

Full size image

Discussion

In this study, the expression of CENPL mRNA in gastrointestinal tumors was analyzed using multiple public databases. CENPL mRNA was overexpressed in ESCA, STAD, PAAD, COAD and READ tumors compared with adjacent normal tissues. Then, the prognostic value of CENPL in patients with gastrointestinal tumors was assessed. The level of CENPL mRNA significantly predicted the prognosis of PAAD patients. Univariate and multivariate Cox regression analyses showed that CENPL mRNA was an independent risk factor for prognosis, suggesting that CENPL may play a critical role in PAAD and is worthy of further study. The mutation rate of CENPL was low and the relationship was not significant between CENPL expression and the abundance of immune cell infiltration in PAAD. However, it was found that the level of CENPL mRNA was significantly correlated with TP53 mutation in PAAD. Immunotherapy plays an important role in the treatment of cancer [26]. Therefore, we analyzed the expression level of CENPL mRNA in skin cutaneous melanoma (SKCM), lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), kidney renal clear cell carcinoma (KIRC), breast invasive carcinoma (BRCA), and bladder urothelial carcinoma (BLCA) by GEPIA [17], for which immunotherapy have been approved. The results showed that the expression of CENPL mRNA was significantly increased in SKCM, LUAD, LUSC and BRCA, and no significant overexpression in KIRC and BLCA (Supplement Fig. 1). This also suggests that CENPL may exert effects in a variety of tumors. miR-340-3p was not only significantly under-expressed in PAAD, but also associated with worse prognosis in these patients, suggestive of a potential role in regulation of CENPL expression in PAAD. Finally, it was verified that CENPL protein was overexpressed in many clinical specimens from PAAD patients.

PAAD is the seventh leading cause of cancer-related deaths worldwide. Notably, the incidence of PAAD is growing, and the patient survival has barely improved [6, 27]. The pathogenesis, early diagnosis and precise treatment remain challenging. Studies have revealed that TP53 mutation occurred in approximately 70% of patients with PAAD and was one of the major factors involved in development and progression of PAAD [28,29,30,31]. In this study, CENPL was rarely mutated in PAAD. However, the level of CENPL mRNA was higher in patients with TP53 mutation. These results suggest that CENPL may have a robust association with TP53 mutations in PAAD patients. CENPL is one of the proteins that constitute kinetosome, and overexpressed CENPL may affect the function of kinetochore. Previous studies have shown that kinetosome dysfunction might be a critical cause of chromosome instability and tumorigenesis [32,33,34]. Therefore, we hypothesize that overexpression of CENPL impairs the function of kinetosome and cause chromosomal instability, thereby causing or accelerating TP53 mutation in pancreatic cancer. which is worthy of further exploration.

miRNA is a single-stranded non-coding RNA molecule with a length ranging from 19 to 25 nucleotides. It is encoded by the endogenous gene. miRNA is involved in biological functions of cells by regulating its target genes [35]. miRNAs widely participate in signal activation, and cell proliferation, differentiation and death [35]. miR-484 is a key factor in cancer and non-cancerous diseases, and its targets in cancer include VEGFB, VEGFR2, MAP2, MMP14, HNF1A, TUSC5, and KLF12 [36]. miR-484 was found to precipitate tumorigenesis and metastasis in liver, prostate and lung cancers [36]. Nan et al. reported that in pancreatic ductal adenocarcinoma, the inhibitory effect of miR-484 on YAP was attenuated and predicted adverse patient outcomes [37]. In this study, we found that CENPL overexpression in PAAD might be the consequence of miR-484 suppression.

Previous studies have shown that miR-340 acted as an oncogenic or tumor suppressor by targeting genes related to proliferation, apoptosis, and metastasis. It showed significance in diagnosis, treatment, chemotherapy resistance, and prognosis [38]. Two studies have found that miR-340-3p might be closely related to the development of laryngeal squamous cell carcinoma and breast cancer [39, 40]. A recent study showed that miR-340-3p regulated by lncRNA HOXA10-AS could down-regulate HTR1D and inhibit the malignant biological behavior of pancreatic cancer through PI3K-Akt signaling pathway [41]. In this study, it was found to have a potential role in PAAD by targeting CENPL. In other words, the overexpression of CENPL mRNA in PAAD may be caused by insufficient miR-340-3p. Therefore, it may be a critical molecule in the pathway that regulates the expression of CENPL. Furthermore, by using the GEPIA [17] and DAVID [42, 43] databases, we found the top 200 genes that were significantly co-expressed with CENPL in PAAD, mainly enriched in endocytosis (Supplement table 3). Endocytosis was revealed to be a pivotal pathway for regulation of metastasis [44]. Therefore, the overexpression of CENPL in PAAD may be involved in this process leading to disease deterioration.

There are some limitations in this study. Laboratory studies of the function of CENPL in PAAD cells is lacking, and the regulatory effects of miR-340-3p on CENPL mRNA expression should be verified through experiments. Furthermore, the number of PAAD patients involved in immunohistochemistry was small.

Conclusion

In summary, the current study suggests that CENPL may play an important pathologic role in PAAD, and CENPL mRNA is a prognostic predictor for PAAD patients.