Sage Journals: Discover world-class research

Abstract

OBJECTIVE:

MED subunits have been reported to be associated with various types of tumors, however, the potential role of MED7 in hepatocellular carcinoma (HCC) was still unclear. The aim of the study was to explore the role of MED7 in HCC.

METHODS:

In this study, MED7 mRNA expression levels between HCC and adjacent normal tissues were first analyzed by several public datasets. Then we utilized a tissue microarray (TMA) to investigate the clinical role of MED7 in HCC by immunohistochemistry (IHC). Meanwhile, the potential mechanisms of MED7 based on gene-gene correlation analyses were also explored.

RESULTS:

High mRNA level of MED7 correlated with advanced stage and worse grade of differentiation. IHC results showed that MED7 protein level was upregulated in HCC and associated with Edmondson grade and Microvascular invasion in 330 cases of HCC. GO (Gene Ontology) and KEGG (Kyoto Encyclopedia of Genes and Genomes) analysis revealed that MED7 co-expressed genes participate primarily in ribonucleoprotein complex biogenesis, protein targeting, mRNA processing and nucleoside triphosphate metabolic process et cetera. Further analysis also revealed that MED7 mRNA level has significant correlation with immune cells infiltration levels.

CONCLUSION:

MED7 was upregulated in HCC and correlated with progression of HCC. Meanwhile, MED7 may promote HCC through participating in multiple gene networks to influence tumorigenesis as well as immune response in HCC microenvironment.

Keywords

MED7 hepatocellular carcinoma tumorigenesis progression

1. Introduction

Hepatocellular carcinoma (HCC) is the leading cause of cancer-related deaths with 841,080 new cases and 781,631 deaths in 2018 [1]. To date, several preventive and therapeutic strategies have been significantly improved, but the five-year survival rate for advanced HCC remains low. Intrahepatic and extrahepatic metastases of HCC are the common reason for poor clinical prognosis of HCC patients [2]. Therefore, identification of the possible molecular mechanisms of malignant biological behaviors of HCC will provide therapeutic benefit for HCC.

The Mediator complex is an evolutionarily conserved RNA polymerase II (Pol II) coregulator required for the transcription of nearly all protein-coding genes [3]. Mediators control transcription in part by acting as a link between Pol II and transcription factors, which regulate their activity during initiation and elongation [4]. The destruction of mediator subunits may affect many different cell functions and fates, some of which may be related to carcinogenesis [5].

Since Med complex plays an important role in the transcription of all eukaryotic genes, alteration in its function and/or composition may have important consequences, and may lead to a variety of disease, including cancer [6]. To date, several MED subunits have been reported to be associated with various types of tumors, including HCC. Zou et al revealed that down-regulation of MED19 could significantly inhibit cell proliferation, induce delay in the G1-phase transition, resulted in tumor formation suppression in HCC in vitro and in vivo, which indicating an oncogenic role of MED19 in the HCC progression [7]. MED7, one of the Mediator subunits, was reported to play an important role in gonadal development and embryogenesis [8]. Recent studies also showed that MED7 may involve in tumorigenesis and gene expression profiling. MED7 downregulation was shown to be associated with an increased tumor risk and could therefore potentially be a marker of favorable prognosis in gastrointestinal stromal tumours [9]. Joseph et al also revealed that high MED7 mRNA and protein expression was associated with good prognostic factors in breast cancer, which may be used as a biomarker for good prognosis in invasive breast cancer, especially ER+ luminal subtypes [10]. Altogether, these findings indicate that MED7 may be used to identify the possibility of clinical outcomes, and contributed to further optimize treatment of patients. However, the potential role of MED7 in HCC was still unclear.

In this study, MED7 mRNA levels between HCC and normal tissues were first analyzed by several public datasets. Then we utilized a tissue microarray (TMA) to investigate the clinical role of MED7 protein level in HCC by immunohistochemistry (IHC). Meanwhile, the potential mechanisms of MED7 based on gene-gene correlation analyses were also explored.

2. Materials and methods

2.1 MED7 mRNA level in HCC based on public datasets

UALCAN (http://ualcan.path.uab.edu) was used to detect differential mRNA level of MED7 in HCC. By UALCAN, relative expression level of genes across tumor and normal samples, as well as in various tumor sub-groups such as cancer stage, tumor grade or other clinicopathological features were also analyzed. Meanwhile, survival analysis was also performed by UALCAN.

The Gene Expression Profiling Interactive Analysis (GEPIA) database (http://gepia.cancer-pku.cn/) containing RNA sequence expression data of 9736 tumors and 8587 normal tissue samples from TCGA and the GTEx projects was also used for analysis. An overall survival (OS) curve was generated by GEPIA based on MED7 mRNA level with the log-rank test in HCC.

2.2 Analysis of MED7and immune infiltration by TIMER database

TIMER (https://cistrome.shinyapps.io/timer/) includes 10,897 samples across 32 cancer types from TCGA, which can be used for systematic analysis of immune infiltrates across diverse cancer types. In the present study, overall survival analysis of differential MED7 mRNA level in HCC was performed, and then the correlation of MED7 expression with the abundance of immune infiltrates, including B cells, CD4+ T cells, CD8+ T cells, neutrophils, macrophages, and dendritic cells, as well as the tumor purity in HCC was analyzed.

2.3 MED7 co-expression analysis by LinkedOmics Database

The LinkedOmics database (http://www.linkedomics. org/login.php) contains 32 TCGA cancer-associated multi-dimensional datasets, which can be used for gene co-expression analysis. Co-expression genes of MED7 were analyzed and presented in volcano plots and heat maps. Function module of Gene Ontology biological process (GO_BP), KEGG pathways were performed by the gene set enrichment analysis (GSEA), and the rank criterion was FDR $<$ 0.05 and 500 simulations were performed.

2.4 Patient samples and construction of tissue microarray (TMA)

We collected 330 cases of HCC paraffinized specimens from the Department of Surgery (Zhejiang Provincial People’s Hospital, Hangzhou, China) between 2008 and 2015. All HCC participants had not received any chemotherapy or radiotherapy before operation, and the project was approved by the ethics committee of Zhejiang Provincial People’s Hospital. The HCC patients were followed up by telephone, which were acquired from patient records. The follow-up deadline was December 2016 or date of death. Patient survival time was calculated from the date of surgery to the follow-up cutoff point. Then a TMA containing these 330 HCC samples was constructed by Shanghai Outdo Biotech Company (Shanghai, China). Furthermore, key information of the patient will not be disclosed, and there is no risk to the patients, so the Ethics Committee exempted this study.

2.5 Patient and public involvement

The patients samples used in the experiment are obtained after previous operations, so the patients or the public were not involved in this study.

2.6 Immunohistochemical staining

First, the paraffin sections of TMA were baked at 60 ${}^{\circ}$ C for 2 h, then deparaffinized with dimethylbenzene and hydrated in gradient ethanol. Then high pressure induced antigen retrieval by citrate sodium buffer (pH 6.0) for 3 min was performed on TMA tissue. After the TMA tissue was cooled at room temperature, 3% H ${}_{2}$ O ${}_{2}$ was used to block endogenous peroxidase for 10 min at room temperature, then the tissues were treated with 5% normal goat serum to inhibit non-specific binding. Primary Rabbit anti-MED7 polyclonal antibody (PA5-40770, ThermoFisher, US) and was then used to incubate TMA tissues with a 1:600 dilution in phosphate-buffered saline at 4 ${}^{\circ}$ C overnight. After that, biotin-labeled secondary antibody was used followed by horseradish peroxidase-linked antibody (Zymed, San Francisco, CA), and then stained with 3,3-diaminobenzidine and counterstained by hematoxylin.

2.7 Evaluation of Immunohistochemical staining

TMA tissues were evaluated independently by two different investigators with no prior knowledge of clinical information of patients. MED7 expression score was based on intensity score and the proportion score of positive staining tumor cells. Intensity score was recorded according to the following criterion: No staining (0); low intensity of staining (1); moderate staining (2); high intensity of staining (3). The proportion score was graded according to the following criterion: 0 ( $<$ 5%), 1 ( $\geqslant$ 5% and $<$ 25%), 2 ( $\geqslant$ 25% and $<$ 50%), 3 ( $\geqslant$ 50% and $<$ 75%), and 4 ( $\geqslant$ 75%). Then the total score was obtained by the intensity score multiplying proportion score, and the range was 0–12. A total score $<$ 6 was considered as low expression of MED7, and a score of $\geqslant$ 6 was regarded as high level of MED7 expression.

2.8 Statistical analysis

Statistical Package for the Social Sciences (version 13.0; SPSS Inc, Chicago, IL) was used to perform all statistical analyses in the present study. $\chi$ 2 or Fisher’s exact test was performed to analyze the categorical data. The overall survival of HCC patients according to levels of MED7 was analyzed by Kaplan-Meier survival accompanying the log-rank test. For all tests, the $P$ value was obtained from two tailed statistical test and $P<$ 0.05 was considered statistically significant.

3. Results

3.1 MED7 mRNA level was elevated in HCC

We initially evaluated MED7 mRNA levels by UACAN. The results revealed that mRNA level of MED7 was significantly higher in HCC tissues than in adjacent normal tissues (Fig. 1). Further sub-group analysis of multiple clinic-pathological features in UALCAN database also showed elevated mRNA level of MED7 was significantly associated with stage, grade of HCC. High mRNA level of MED7 correlated with advanced stage and worse grade of differentiation.

Table 1
Expression of MED7 in 330 cases of hepatocellular carcinoma tissues

Clinical parameters	Number	MED7 expression		$\chi^{2}$	$P$ value
		Low (%)	High (%)
Age (years)				0.210	0.646
$<$ 55	128	55 (43.0)	73 (57.0)
$\geqslant$ 55	202	92 (45.5)	110 (54.5)
Gender				0.154	0.695
Male	268	118 (44.0)	150 (56.0)
Female	62	29 (46.8)	33 (53.2)
Tumor size				0.006	0.940
$\leqslant$ 5 cm	192	85 (44.3)	107 (55.7)
$>$ 5 cm	130	57(43.8)	73 (56.2)
Edmondson grade				4.172	0.041
I $+$ II	205	98 (47.8)	107(52.2)
III	119	43 (36.1)	76 (63.9)
Metastasis				2.732	0.098
M0	297	127 (42.8)	170 (57.2)
M1	27	16 (59.3)	11 (40.7)
Microvascular invasion				5.708	0.017
Absence	124	66 (53.2)	58 (46.8)
Presence	121	46 (38.0)	75 (56.4)

Figure 1.

MED7 mRNA level in HCC, as well as in subgroups of patients with HCC stratified based on grade, metastasis and stage by UALCAN. MED7 mRNA level was upregulated in HCC, and high MED7 mRNA level was associated with grade, metastasis and stage. Grade 1: Well differentiated, Grade 2: Moderately differentiated, Grade 3: Poorly differentiated, Grade 4: Undifferentiated. Stage: TNM (Tumor node metastasis) stage. N0: No regional lymph node metastasis, N1: Metastases in 1 to 3 axillary lymph nodes. ${}^{*}P<$ 0.05, ${}^{**}P<$ 0.01.

Figure 2.

MED7 protein level was detected by Immunohistochemical staining in 330 cases of HCC. (A) Weak staining of MED7 in HCC. (B) Strong staining of MED7 in HCC. Magnification: the original magnification $\times$ 40 and $\times$ 200.

3.2 MED7 protein level was associated with clinic pathological features of HCC

Immunohistochemistry results of 330 cases of HCC showed that MED7 was mainly located in nucleus and cytoplasm (Fig. 2). Then the correlation between MED7 expression and clinicopathological features was analyzed and showed that MED7 upregulation was closely associated with Edmondson grade and Microvascular invasion (Table 1). High expression ratio of MED7 in patients with Edmondson grade I $+$ II was 52.2% (107/205), which was lower than that in patients with Edmondson grade III (63.9%, 76/119, $\chi$ 2 $=$ 4.172, $P=$ 0.041). Meanwhile, patients with microvascular invasion showed higher MED7 protein level (56.4%, 75/121) than that in patients without microvascular invasion (46.8%, 58/124, $\chi$ 2 $=$ 5.708, $P=$ 0.017).

3.3 MED7 was correlated with prognosis of HCC

Next, we also performed Kaplan-Meier survival analysis to explore the association between MED7 expression and the survival outcomes of HCC with published database and our present study (Fig. 3). Generally, the patients with high MED7 mRNA level had significantly shorter overall survival (log-rank test, $P<$ 0.05) compared to those with low MED7 mRNA level by UALCAN, GEPIA and TIMER database. In our present study, the mean survival time in patients with high level of MED7 was 43.46 $\pm$ 3.66 months, and was 50.45 $\pm$ 3.31 months in patients with low level of MED7. However, there was no significant difference between MED7 protein level and prognosis of HCC (Fig. 3, log-rank test, $\chi$ 2 $=$ 0.660, $P=$ 0 .417).

Figure 3.

The association between MED7 and survival outcome of HCC was detected by public database and present IHC result. Overall survival (OS) analysis detected by UALCAN, GEPIA, TIMER and in our cohort of 330 cases of HCC detected by IHC.

3.4 MED7 co-expression networks in HCC

In order to gain the insight of MED7 molecular mechanism in HCC, we used function module of LinkedOmics to examine MED7 co-expression networks in HCC. Figure 4A showed that 5,102 genes (dark red dots) significant positive correlated with MED7, whereas 5,044 genes (dark green dots) negatively correlated with MED7, and the top 50 significant genes positively and negatively correlated with MED7 were shown in the heat map (Fig. 4B).

Figure 4.

Co-expression genes of MED7 in HCC were analyzed by LinkedOmics. (A) The highly correlated genes of MED7 identified by Spearman test in HCC. (B) Top 50 genes positively and negatively correlated with MED7 mRNA level in HCC were showed by heat maps. Positively correlated genes were shown in red and negatively correlated genes were shown in blue. (C) Analysis of GO annotations and KEGG pathways of MED7 in HCC.

Furthermore, we also performed Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis. GO by gene set enrichment analysis (GSEA) showed that co-expressed genes of MED7 participate primarily in ribonucleoprotein complex biogenesis, protein targeting, mRNA processing and nucleoside triphosphate metabolic process et cetera, while the process like fatty acid metabolic process, cellular carbohydrate metabolic process, and cell-substrat adhesion were inhibited (Fig. 4C). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis showed enrichment in the ribosome, oxidative phosphprylation, spliceosome, et cetera (Fig. 4C).

3.5 MED7 was correlated with immune infiltration in HCC

Next, we investigated the correlation of MED7 level with immune infiltration levels in HCC by TIMER database. As shown in Fig. 5, MED7 level has significant correlated with immune cells infiltration levels, such as B cell ( $r=$ 0.316, $P=$ 1.97E-09), CD8+ T cell ( $r=$ 0.244, $P=$ 4.99E-06), CD4+ T cell ( $r=$ 0.236, $P=$ 9.58E-06), macrophage ( $r=$ 0.352, $P=$ 2.19E-11), neutrophil ( $r=$ 0.321, $P=$ 1.05E-09), dendritic cell ( $r=$ 0.285, $P=$ 8.61E-08).

Figure 5.

MED7 expression level was associated with immune cells infiltration in HCC. MED7 mRNA level correlated with infiltrating levels of CD8+ T cells, CD4+ T cells, macrophages, neutrophils, and dendritic cells significantly in HCC.

Furthermore, we also performed IHC staining of CD8 (provided as supplementary materials) and analyzed the correlation of MED7 and CD8. The results showed that the protein level of CD8 was associated with metastasis (Table 2), which suggested that CD8 may associate with the disorder of immune status in tumor microenvironment of HCC. However, further analysis showed that correlation between MED7 and CD8 was not significant (Spearmen’s rho test, $r=$ 0.057, $p=$ 0.304).

Table 2

Expression of CD8 in hepatocellular carcinoma tissues

Clinical parameters	Number	CD8 expression		$\chi^{2}$	$P$ value
		Low (%)	High(%)
Age (years)				0.901	0.342
$<$ 55	126	57 (45.2)	69 (54.8)
$\geqslant$ 55	198	79 (39.9)	119 (60.1)
Gender				2.413	0.120
Male	263	105 (39.9)	158 (60.1)
Female	61	31 (50.8)	30 (49.2)
Tumor size				3.558	0.059
$\leqslant$ 5 cm	188	71 (37.8)	117 (62.2)
$>$ 5 cm	128	62 (48.4)	66 (51.6)
Edmondson grade				0.593	0.441
I $+$ II	200	81 (40.5)	119 (59.5)
III	118	53 (44.9)	65 (55.1)
Metastasis				7.081	0.008
M0	291	117 (40.2)	174 (59.8)
M1	27	18 (66.7)	9 (33.3)
Microvascular invasion				2.783	0.095
Absence	122	47 (38.5)	75 (61.5)
Presence	120	59 (49.2)	61 (50.8)

4. Discussion

MED7, as mediator subunits, was reported to play an important role in gonadal development and embryogenesis [8]. Recently, study found that disruption of mediator subunits could affect many different cellular functions, including carcinogenesis [5]. However, the potential role of MED7 in HCC was still unclear.

In our present study, the functional relationships of MED7 with various clinicopathological variables in primary HCC were assessed. We found that high mRNA level of MED7 correlated with advanced stage, worse grade of differentiation and metastasis, and upregulation of MED7 protein was closely associated with Edmondson grade and microvascular invasion. These suggested that high expression levels of MED7 were significantly associated with poorer behaving tumour characteristics. However, Koschubs et al revealed that MED7 downregulation was significantly associated with increased risk of gastrointestinal stromal tumours [11]. Joseph et al also found that high MED7 mRNA and protein expression was associated with good prognostic factors [10]. Although MED7 is a highly conserved mediator subunit, it is not essential for viability of all species. The results revealed that the role of MED7 in cancer tissue may be diverse due to tissue-specificity. However many limitations existed in the present study including small sample size and limitation during collection, the conclusion should be made with caution.

Some studies revealed that high MED7 level was associated with good prognostic factors in breast cancer and gastrointestinal stromal tumours [9, 10]. However, in our present study, we found that patients with high MED7 mRNA level had poorer prognosis in HCC by public database, and there was no significant difference between MED7 protein level and prognosis of HCC. These findings indicate that MED7 could have a function in tumor pathogenesis with tissue specificity, which should be validated by further analysis in vitro and in vivo.

Furthermore, we also explore the molecular mechanism of MED7 in HCC by GO and KEGG analysis, and the results showed that co-expressed genes of MED7 participate primarily in ribonucleoprotein complex biogenesis, protein targeting, mRNA processing and nucleoside triphosphate metabolic process, oxidative phosphprylation, et cetera. Studies also found that Mediator plays an integral role in transcription at least in part by serving as a link between Pol II and transcription factors [4]. This indicated that MED7 level may affect multiple gene networks which are able to influence many cellular processes related to tumourigenesis in HCC. Further analysis by TIMER database tools showed that MED7 mRNA level has significant correlation with immune cells infiltration levels. However, there was no correlation between MED7 and CD8 protein level in our present study detected by IHC method. We considered that IHC staining may not reflect the whole level of immune cells comprehensively and accurately due to factors such as local sampling method and treatment process. Further study in vivo experiment will be performed in the future.

In summary, we found that MED7 was upregulated in HCC and correlated with progression of HCC. Meanwhile, MED7 may promote HCC through participating in multiple gene networks to influence tumorigenesis as well as immune response in HCC microenvironment.

Contributorship statement

Conception: Jun-Gang Zhang.

Interpretation or analysis of data: Zheng-Lin Chen, Ying-Yu Ma.

Preparation of the manuscript: Zheng-Lin Chen.

Revision for important intellectual content: Jun-Gang Zhang, Xiao-Zhou Mou.

Supervision: Xiao-Zhou Mou.

Funding

Not applicable

Data sharing statement

The data that support the findings of this study can be available from the corresponding author on reasonable request.

Ethics statement

This work was approved by ethic committee of Zhejiang provincial people’s hospital.

Supplementary data

The supplementary files are available to download from http://dx.doi.org/10.3233/CBM-220439.

sj-pdf-1-cbm-10.3233_CBM-220439.pdf - Supplemental material

Supplemental material, sj-pdf-1-cbm-10.3233_CBM-220439.pdf

Footnotes

Conflict of interest

The authors declare no conflict of interest.

References

Bray

Ferlay

Soerjomataram

Siegel

R.L.

Torre

L.A.

and Jemal

, Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries, CA: A Cancer Journal for Clinicians 68 (2018), 394–424.

Aldrighetti

Pulitano

Catena

Arru

Guzzetti

Halliday

and Ferla

, Liver resection with portal vein thrombectomy for hepatocellular carcinoma with vascular invasion, Annals of Surgical Oncology 16 (2009), 1254.

Gustafsson

C.M.

and Samuelsson

, Mediator–a universal complex in transcriptional regulation, Molecular Microbiology 41 (2001), 1–8.

Allen

B.L.

and Taatjes

D.J.

, The Mediator complex: a central integrator of transcription, Nature reviews, Molecular Cell Biology 16 (2015), 155–166.

Schiano

Casamassimi

Rienzo

de Nigris

Sommese

and Napoli

, Involvement of Mediator complex in malignancy, Biochimica et Biophysica Acta 1845 (2014), 66–83.

Hasegawa

Sumitomo

Fujita

Aritome

Mizuta

Matsui

Ishino

Inoue

Urahama

Nose

Mukohara

Kamoshida

Roeder

R.G.

and Ito

, Mediator subunits MED1 and MED24 cooperatively contribute to pubertal mammary gland development and growth of breast carcinoma cells, Mol Cell Biol 32 (2012), 1483–1495.

Zou

S.W.

K.X.

Wang

Z.G.

Yuan

Yan

and Zheng

, The role of Med19 in the proliferation and tumorigenesis of human hepatocellular carcinoma cells, Acta pharmacologica Sinica 32 (2011), 354–360.

Kwon

J.Y.

Park

J.M.

Gim

B.S.

Han

S.J.

Lee

and Kim

Y.J.

, Caenorhabditis elegans mediator complexes are required for developmental-specific transcriptional activation, Proceedings of the National Academy of Sciences of the United States of America 96 (1999), 14990–14995.

Hur

Lee

H.J.

Woo

J.H.

Kim

J.H.

and Yang

H.K.

, Gene expression profiling of human gastrointestinal stromal tumors according to its malignant potential, Digestive Diseases and Sciences 55 (2010), 2561–2567.

10.

Joseph

Macnamara

Craze

Russell

Provenzano

Nolan

C.C.

Diez-Rodriguez

Sonbul

S.N.

Aleskandarany

M.A.

Green

A.R.

Rakha

E.A.

Ellis

I.O.

and Mukherjee

, Mediator complex (MED) 7: a biomarker associated with good prognosis in invasive breast cancer, especially ER+ luminal subtypes, British Journal of Cancer 118 (2018), 1142–1151.

11.

Koschubs

Seizl

Lariviere

Kurth

Baumli

Martin

D.E.

and Cramer

, Identification, structure, and functional requirement of the Mediator submodule Med7N/31, The EMBO Journal 28 (2009), 69–80.

Upregulation of MED7 was associated with progression in hepatocellular carcinoma