A study on the correlation of MyD88 expression with gastric cancer

Abstract

Objective: The myeloid differentiation factor-88 (MyD88) plays a key role in mediating the innate immune signal transduction of toll-like receptor (TLR) and interleukin-1 (IL-1) family members, and it also participates in the regulation of tumorigenesis in various cancer models Our study sought to determine whether there is any correlation with MyD88 and the development of gastric cancer and, if such a correlation exists, to find out whether it can be used to improve the prognosis of gastric cancer patients. Patients and methods: The expression of MyD88 in 108 cases of gastric cancer specimens, 15 cases of adenoma, and 15 cases of normal mucosa was detected by immunohistochemistry, and the correlations of the MyD88 expression with clinicopathologic changes (including disease-free survival [DFS] and overall survival [OS] were analyzed. The level of MyD88 was detected in well-differentiated MGC-803 and poorly-differentiated BGC-823 cell lines by qPCR and western blot. The expression of MyD88 was then measured by western blot after the treatment of an MyD88 overexpression vector or MyD88 inhibitor. Cell proliferation was determined by overexpression or suppression of MyD88. Results: In clinical cases, MyD88 was highly expressed in 23% of patients with gastric cancer as compared to those in normal mucosa and adenoma. There was a significant correlation of MyD88 overexpression with gastric metastasis (P<0.01). The overexpression of MyD88 significantly promoted the proliferation of MGC-803 and BGC-823 cell lines in gastric cancer. According to the single factor analysis, a high expression of MyD88 was strongly associated with poor DFS and OS (P<0.01), and MyD88 was an independent prognostic factor of OS. Conclusion: This study demonstrates that a high expression of MyD88 is associated with the gastric cancer patients with liver metastasis, and facilitates the proliferation of gastric cancer cells. MyD88 is an independent predictive factor for the poor prognosis of gastric cancer patients, which provides a potential tool for future clinical diagnosis.

Introduction

Gastric cancer is the second-leading cause of cancer death worldwide, and its major risk factor is persistent gastritis due to the colonization of Helicobacter pylori. Though all the individuals with the colonization of Helicobacter pylori suffer from a certain degree of gastritis, gastritis in only 2.9% of patients will develop into gastric cancer, suggesting that the additional factors contribute to the development of gastritis into gastric cancer [1-3]. In this context, the role of pathogen recognition receptors, especially those in the TRL family, has attracted much attention and has been investigated as a potential promoter of gastrointestinal cancer in recent years [4].

Myeloid differentiation factor-88 (MyD88) is regarded as a kind of key adaptor molecule which links IL-1R1 to IL-1 receptor associated kinase (IRAK) complex [5]. The module component of MyD88 consists of the death domain (DD) at the end of NH2 and the Toll/IL-1R (TIR) domain at the end of COOH. The adaptor molecule modulates the signal transduction of innate immunity (such as TLR members and IL-1 family) and thus affects cancer promotion and antitumor response [6]. The MyD88 protein directly triggers the protein signal transduction of TLR/IL-1R and the signal transduction of the interferon (IFN)-γ receptor in the innate immune responses [7,8].

The inflammatory responses promote the occurrence of cancer through multiple mechanisms [9]. The nuclear factor-κB (NF-κB) is the key factor in inflammation, and it plays an essential role in regulating tumors caused by chronic inflammation or exogenous mutagen [10]. NF-κB is activated by a variety of stimuli [11]. However, it is not yet clear which pathway is potentially involved in the cancer-related inflammation and tissue repair responses. The roles of inflammation and tissue repair responses other than chronic inflammation or the use of exogenous carcinogen in the spontaneous cancer are not yet fully understood. Recent study indicated that MyD88 (the signal transducer of TLRs) employs the toll-like receptor (TLR) in the innate immune system during the control of the tissue regeneration responses [12].

The classic TLR2 signal model involves the adaptor protein MyD88 and Mal (also known as TIR domain-containing adaptor protein [TIRAP]), thus activating the downstream pathways of NF-kB, p38, extracellular regulated kinase, c-Jun N-terminal kinase and mitogen-activated protein kinase (MAPK) [13]. However, a recent in vitro study on macrophages indicated that the important role of MyD88 in TLR2-induced activation is different from activation in the signal pathways of NF-κB and MAPK, and it highlighted the necessity that TLR2 relies on Mal to activate the signal cascades in response to the low concentration of ligand.

In this study, the expression of MyD88 in normal gastric mucosa, adenoma and gastric cancer was systematically analyzed. We aim to determine the possible correlation of MyD88 with DFS and OS in clinical diagnosis.

Materials and methods

Information resource and cell line

Human gastric cancer well-differentiated cell line MGC-803 and poorly-differentiated cell line BGC-823 were purchased from Genomeditech (Shanghai) Co., Ltd., and they were cultured in Dulbecco’s modified Eagle medium (DMEM) containing 10% fetal bovine serum and 2 mM glutamine and cultured under 10% CO₂ at 37°C. A total of 108 patients with gastric cancer (including 62 males and 46 females) subjected to surgery in the Cancer Hospital of Wuhan University from 2010 to 2015 were investigated in this study. Demographic information (gender and age) and tumor features (anatomic site, histology, vascular invasion, lymphatic vessel invasion, lymph node metastasis, liver metastasis, peritoneal metastasis and TNM stage) of the patients was obtained from clinical and pathological records (Table 1). The tumor locations were classified into proximal or distal positions. The disease stage was classified on the basis of the standard proposed by the American Joint Committee on Cancer (AJCC) [14]. The disease-free survival (DFS) was defined as the time interval between the day of surgery and the day of recurrence detected for the first time. If the recurrence was not detected, the date of death or of the last follow-up was used instead. The overall survival (OS) was defined as the time interval between the surgery and the patient’s death. The follow-up period of DFS and OS after the first surgery was 5 years. Patients with cancer in relation to gastritis ulcers, Crohn’s disease, and familial adenomatous polyposis were excluded. The study was approved by the Ethics Committee of Beijing First Hospital of Integrated Chinese And Western Medicine. All patients enrolled in this study signed written informed consent after a detailed explanation.

Table 1.

Clinicopathological statistical results of 108 patients

Clinicopathological factor	p	High expression of MyD88 (n=25)	Low expression of MyD88 (n=83)
Age
65	0.915	11	37
<65		14	46
Gender
Male	0.5054	13	49
Female		12	34
Tumor site
Distant	0.3113	23	68
Proximal		2	15
Histology
Good	0.0076	4	36
Medium		18	43
Poor		3	4
Vascular invasion
Yes	0.0987	15	32
No		10	51
Lymphatic invasion
Yes	0.3038	20	59
No		5	24
Liver metastasis
Yes	0.0001	10	12
No		15	71
Peritoneal metastasis
Yes	0.0611	2	7
No		23	76
TNM staging
I, II	0.0197	7	31
III		7	35
IV		11	17

Main reagents

MyD88 antibody (HFL-296, Santa Cruz Biotech, Santa Cruz, CA, USA), streptavidin (Dako-Cytomation, Glostrup, Denmark), Lipofectamine 2000 (Thermo Fisher Scientific, Waltham, MA, USA), MyD88 inhibitor (NBP2-29328) (Novus, Littleton, CO, USA), MyD88 over expression vector (Genomics-online, Beijing, China), CCK8 kit (Kumamoto, Japan), RNeasy mini kit (Qiagen, Chatsworth, CA, USA), quantitative reverse transcription-polymerase chain reaction (qRT-PCR) micro ribonucleic acid (miRNA) detection kit (Ambion, USA), RT-buffer (Toyobo, Osaka, Japan), SYBR Green real-time PCR mix (Toyobo, Osaka, Japan), primer synthesis (Invitrogen), nitrocellulose membrane (Millipore, Billerica, MA, USA), anti-MyD88 primary antibody (Santa Cruz Biotechnology, CA, USA), β-actin (Sigma), StatView software (Abacus Concepts, Berkeley, CA, USA), and ABI PRISM 7700 sequence detection system (Applied Biosystems, Hamilton, New Zealand).

Immunohistochemical detection

Formalin-fixed and paraffin-embedded tissue specimens of 15 cases of normal mucosa, 15 cases of adenoma, 108 cases of gastric cancer, and 14 cases of liver metastatic lesion were used in the immunohistochemical detection.

First, 4 μm-thick tissue sections were labeled with streptavidin-biotin, dewaxed in xylene and rehydrated in ethanol, followed by treatment with 0.3% hydrogen peroxide to block the endogenous peroxidase activity for 30 min. Then the tissue sections were incubated with the serum-free protein blocker and then incubated with the mouse anti-rabbit MyD88 polyclonal antibody (1:100) at 4°C. Biotinylated secondary antibody and peroxidase-labeled streptavidin were applied at room temperature for 50 min, respectively. The antigen-antibody complex was developed using 3,3’-diaminobenzidine. After slight counterstaining with hematoxylin, the glass slides were examined under a microscope. Human tonsil tissues were used as the positive control for MyD88 immunoreactivity, while phosphate buffered saline without primary antibody was used as the negative control.

Evaluation of immunohistochemical results

Each section was evaluated independently by two pathologists who were blinded to the clinical data. MyD88 in most specimens displayed a similar staining intensity (moderate-strong), so the staining intensity was not evaluated in this study. Membrane and/or cytoplasm staining in any intensity indicated the positive staining of MyD88. Several high-power fields (×200) were selected from regions with different staining densities (high, medium, low, and negative) using a digital camera. Photos were printed on plain paper, on which a grid was drawn. An average of 2000 (1500-2500) tumor cells were selected in each tumor, and the percentage of tumor cells with positive staining was calculated. After that, the percentage of MyD88-positive tumor cells was scored based on the 0-4-point scale (0 point: no staining, 1 point: ≥10%, 2 points: 11-30%, 3 points: 31-50%, and 4 points: >50%). In addition, the expression level of MyD88 was divided into the following two groups according to the score: the low expression group (0, 1 and 2 points) and the high expression group (3 and 4 points). Also, the sum of the MyD88 staining scores were divided into the following two groups: the <5 group and the =5 group.

Over expression or inhibition of MyD88 in cells

Cells were passaged once every 2 to 3 days and cells at logarithmic growth were used in the experiment. Cell transfection was performed 24-72 h after passage. Cells in 6-well plates were transfected with the overexpression vector (1 μg) Lipofectamine 2000 or treated with the MyD88 inhibitor (100 μM). The growth and proliferation of cells were then observed.

Cell proliferation assay

Cells were seeded in 96-well plates and cultured at 37°C and 5% CO₂. A cell proliferation assay was performed at 48 h after transfection or treatment. Cells were washed with PBS 3 times. 100 μL CCK8 mixture (CCK8 Reagents: medium =1:10) was added to each well and incubated at 37°C for 2 h in the dark. Absorbance at the 450 nm wavelength was measured by a microplate reader. Each group has 5 wells.

RNA extraction and qRT-PCR

RNA extraction and qRT-PCR were performed as described previously. Total RNA was isolated from the fixed paraffin specimens or cells [16]. RNA was extracted from cells using the RNeasy mini kit, and reversely transcribed using RT-buffer according to manufacturer’s instructions, followed by qPCR using SYBR Green real-time PCR mix. Primers were designed as follows: MyD88 (sense: CTCCTCCACATCCCTTCC, antisense: CCGCACGTTCAAGAACAGAGA), glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (sense: TGAAGCAGGCATCTGAGGG, antisense: CGAAGGTGGAAGAGTGGGAG). PCR was performed using the ABI PRISM 7700 sequence detection system, and reaction conditions are as follows: at 95°C for 10 min, at 95°C for 30 s, at 60°C for 1 min, a total of 40 cycles. The relative quantitative value of MyD88 expression was normalized based on the expression level of the reference gene GAPDH in each case.

Western blot

Tissues were washed with Hanks balanced salt solution twice and directly lysed in a radioimmunoprecipitation assay buffer. Cell or tissue lysates (60 mg) were isolated via sodium dodecyl sulfate polyacrylamide gel electrophoresis, transferred onto a nitrocellulose membrane, incubated with anti-MyD88 primary antibody or mouse β-actin monoclonal antibody at room temperature for 2 h, fully washed with 0.1% phosphate buffered saline with Tween-20 and incubated with the horseradish peroxidase-labeled secondary antibody (diluted at 1:3000). Finally, the signal was enhanced via chemiluminescence, thus enabling visualizing.

Statistical analysis

GraphPad Prism 5.0 software was used for the analysis. DFS and OS measured on the day of each operation were the main statistical results. DFS and OS were estimated using the Kaplan-Meier curve, and the curve was compared using a log-rank test. Recurrence and date of death were analyzed using the univariate and multivariate Cox proportional hazard model. In addition, the hazard ratio (HR) and 95% confidence interval were calculated. The probability value (p) <0.05 suggested that the difference was statistically significant.

Results

MyD88 expressions in normal gastric mucosa, para-carcinoma tissues, adenoma and gastric cancer

The immunolocalization of the MyD88 protein was observed in the cytoplasms. Basically, there was no or very weak MyD88 expression in the normal mucosa, para-carcinoma tissues and adenoma. MyD88 expression was displayed in 93 out of 108 (86%) gastric cancer specimens, and the high-level expression was detected in 25 kinds of cancer (23%). The MyD88 expression was positive in all 14 cases of liver metastatic lesions, 12 cases of which (86%) showed high expression (Figure 1).

Figure 1.

Differences in the MyD88 expression among normal gastric mucosa, para-carcinoma tissues, adenoma and gastric cancer. (A) The expression of MyD88 in different groups. The representative images of MyD88 expression in normal gastric mucosa (B), para-carcinoma tissues (C), adenoma (D) and gastric cancer (E) by immunohistochemical detection (×200). The level of MyD88 in gastric cancer was remarkably higher than that in normal gastric mucosa, para-carcinoma tissues, and adenoma. **The difference is extremely significant, P<0.01.

MyD88 expressions in MGC-803 and BGC-823 cell lines

We then measured the MyD88 expressions in human gastric cancer well-differentiated MGC-803 cell lines and poorly-differentiated BGC-823 cell lines via qPCR and western blotting. Results revealed that the expression of MyD88 in MGC-803 cells at both the mRNA and protein levels was significantly higher than it was in the BGC-823 cells (P<0.05), suggesting that the MyD88 expression is correlated with the differentiation degree of gastric cancer cells (Figure 2).

Figure 2.

MyD88 expressions in MGC-803 and BGC-823 cell lines. It was shown that the level of MyD88 in MGC-803 was significantly higher than that in BGC-823 by the detection of qPCR and western blot. *, P<0.05.

Effects of MyD88 on the proliferation of different gastric cancer cell lines

Based on the results above that MyD88 was related to the development of gastric cancer, we then evaluated the effect of MyD88 on the proliferation of gastric cancer cells. The expression of MyD88 was initially upregulated or downregulated by the overexpression vector and inhibitor. Of note, the level of MyD88 was obviously mediated in both cell lines, respectively (Figure 3A, 3B). Afterwards, our data showed that the over expression of MyD88 significantly increased the proliferation of both MGC-803 and BGC-823 cells (P<0.05). By contrast, the inhibition of MyD88 statistically suppressed the cell proliferation (Figure 3C, 3D).

Figure 3.

Effects of overexpression and inhibition of MyD88 on proliferation of gastric cancer cells. The expression of MyD88 was regulated by the overexpression vector or inhibitor in MGC-803 (A) and BGC-823 (B) cell lines by western blotting. (C) Over expression of MyD88 significantly elevated the proliferation of both cell lines (D) while the inhibition of MyD88 impeded the cell growth. *, P<0.05. **, P<0.01.

Clinicopathological significance of MyD88 expression

The high expression of MyD88 was related to the histology (P=0.0076), liver metastasis (P=0.001), and TNM staging (P=0.0197) of the cancer (Table 1). In addition, the MyD88 expression level was higher in gastric cancer accompanied with liver metastasis than that in gastric cancer without liver metastasis (P<0.05) (Figure 4).

Figure 4.

Difference in the immunostaining score of MyD88 between gastric cancer with liver metastasis (+) (n=22) and gastric cancer with liver metastasis (-) (n=86). The expression level of MyD88 in gastric cancer with liver metastasis (+) was significantly higher than it was in gastric cancer with liver metastasis (-) (P<0.05).

Clinicopathological parameters and survival rate of patients with gastric cancer

During the 5-year follow-up, there were 53 cases of tumor recurrence (DFS=49%) and 46 deaths (OS=43%). According to the univariate analysis, histology (DFS: P=0.0322, OS: P=0.0084), TNM staging (DFS: P<0.0001, OS: P=0.0001), vascular invasion (DFS: P<0.0001, OS: P=0.0001), lymphatic invasion (DFS: P=0.001, OS: P=0.0121), lymph node metastasis (DFS: P=0.0001, OS: P=0.0017), liver metastasis (DFS: P<0.0001, OS: P<0.0001) and peritoneal metastasis (DFS: P=0.0386, OS: P=0.007) were important factors related to DFS and OS (Table 2). The patient’s age, gender and tumor site were not associated with DFS or OS (Table 2). Independent prognostic factors were determined via multivariate analysis. All factors with significant prognostic value in the univariate analysis, including histopathological variables and tumor markers, are shown in Table 2.

Table 2.

Statistical results of clinicopathological parameters of patients with gastric cancer

Variable	5-year DFS HR	p	5-year OS HR	p
Age (65 years old vs. <65 years old)	0.84 (0.48-1.45)	0.5355	1.17 (0.65-2.08)	0.5934
Gender (male vs. female)	0.95 (0.55-1.63)	0.8568	0.95 (0.53-1.71)	0.8781
Tumor site (distal end vs. proximal end)	2.89 (1.16-8.97)	0.0656	3.17 (0.98-10.23)	0.0534
Histology (poor vs. medium vs. good)	1.95 (1.05-3.6)	0.0322	2.57 (1.27-5.19)	0.0084
TNM staging (III/IV vs. I/II)	8.07 (3.2-20.33)	<0.0001	6.26 (2.46-15.88)	0.0001
Vascular invasion (yes vs. no)	3.28 (1.86-5.78)	<0.0001	3.35 (1.82-6.16)	0.0001
Lymphatic invasion (yes vs. no)	4.68 (1.86-11.79)	0.001	3 (1.27-7.09)	0.0121
Lymph node metastasis (yes vs. no)	3.64 (1.87-7.09)	0.0001	3.07 (1.52-6.2)	0.0017
Liver metastasis (yes vs. no)	5.04 (2.85-8.9)	<0.0001	5.35 (2.94-9.74)	<0.0001
Peritoneal metastasis (yes vs. no)	2.32 (1.05-5.18)	0.0386	3.06 (1.35-6.91)	0.007
MyD88 (high vs. low)	2.33 (1.31-4.13)	0.0038	3.03 (1.67-5.48)	0.0002

MyD88 expression and survival rate of patients with gastric cancer

The effect of high expression of MyD88 on the survival of patients with gastric cancer was evaluated. The high expression of MyD88 was significantly associated with low DFS (P=0.0018) and low OS (P=0.015) (Figure 5).

Figure 5.

Kaplan-Meier survival curves of DFS and OS in gastric cancer patients with different expressions of MyD88. A high level of MyD88 was associated with low DFS and OS.

Discussion

According to a recent study, the MyD88 signal drives the tumor growth in a variety of organs [8]. Moreover, it has been revealed that MyD88 is crucial for promoting the diethylnitrosamine-induced hepatocellular neoplasm [7]. In the spontaneous (Apcmin/+) and carcinogen (azoxymethane)-induced intestinal tumor model, it has also been proved that MyD88 is important in the development of tumors [8]. In addition, MyD88 is a chemical-induced positive regulator of skin and connective tissues [10]. It was observed in this study that the MyD88 signal was often upregulated in gastric cancer compared with those in normal mucosa and adenoma, and the correlation between MyD88 expression and gastric cancer was further validated. The above findings suggest that the MyD88 signal transduction, as a carcinogenic factor, favors the development of gastric cancer.

Whether MyD88 is involved in the occurrence of tumors remains unclear [15]. Chronic infection and inflammation in humans are regarded as the two most important factors resulting in the occurrence and progression of tumors [16]. TLR signal transduction possibly plays an important role in various cancers, including gastric cancer, ovarian cancer, lung cancer, pancreatic cancer and liver cancer. It has been proved to be associated with local chronic inflammation [17-20]. Upregulation of TLR may lead to genomic DNA damage, mutation, and abnormal chromosomal translocation [8]. The activation of TLR4 signal seemingly promotes the development of gastritis-related cancers through such mechanisms as enhanced cyclooxygenase-2 (Cox-2) expression and increased epidermal growth factor receptor (EGFR) signal transduction [18]. Besides gastritis-related cancer [21], symbiotic bacteria are also involved in the development of sporadic gastric cancer and may promote gastric cancer by inducing chromosomal instability [22]. Results of this study manifested that MyD88 signal transduction contributed to tumorigenesis in both gastritis-related cancer and sporadic gastric cancer.

Results of this study clearly demonstrated that overexpression of MyD88 was an independent and significant prognostic factor for the 5-year DFS and OS. Previous evidence demonstrated that the increase in downstream signals of the TLR4/MyD88 pathway, such as Cox-2 and nuclear factor-κB (NF-κB), was associated with the survival of gastric cancer patients. Cox-2 is also considered to play an important role in the occurrence of rectal cancer. According to a recent study by Ogino et al., Cox-2 overexpression is associated with the poor survival rate of colon cancer patients [23]. NF-κB is the terminal point of the MyD88 signal transduction pathway. Evidence suggests a link between NF-κB activation and cancer development, and the constitutive NF-κB activation has also been detected in most cancer cell lines and various tumor tissues [24]. Scartozzi et al. reported that the NF-κB nuclear expression may indicate the response and survival in irinotecan-refractory metastatic gastric cancer treated with cetuximab-irinotecan therapy [25]. These findings illustrate that the MyD88 signal pathway in tumor cells plays a crucial role in the prognosis of gastric cancer patients, and blocking this pathway can greatly benefit patients with gastric cancer.

In conclusion, this study demonstrates that MyD88 is associated with an increased risk of liver metastasis and a decreases survival rate, which suggests a possible method by targeting the MyD88 signal in the treatment of gastric cancer.

Disclosure of conflict of interest

None.