HLA-DR regulates macrophage phenotypic transformation and affects malignant behavior in esophageal squamous cell carcinoma

Background Tumor-associated macrophages (TAMs) are an important immune cell component of the tumor microenvironment. This study aimed to explore the molecular mechanism of TAMs phenotype transformation and the role in the development of esophageal squamous cell carcinoma (ESCC). Methods Co-culture conditions were employed to determine the phenotypic effects of TAMs on ESCC cell biological behavior. Tumor metastasis related molecules VEGF-C and MMP-9 produced by TAMs was evaluated by qRT-PCR and western blot. Expression of HLA-DR was knocked down in TAMs in vitro to determine the effects on macrophage polarization and the biological behavior of ESCC. We determined whether co-injection with M2 TAMs and macrophages depletion affected tumor growth in vivo tumor challenge model. Associations between HLA-DR, TAM density, and clinical outcomes were evaluated in patients with ESCC. Results TAMs in ESCC samples were found to closely reect the M2 phenotype of TAMs, and exhibited low expression of HLA-DR. Which was involved in ESCC tumor invasion and metastasis. Low expression of HLA-DR positively correlated with high-density of M2 TAMs, indicating high invasiveness and poor prognosis in patients with ESCC. Downregulation of HLA-DR in TAMs led to additional M2-type TAM polarization and more VEGF-C and MMP-9 secretion, promoted the malignant transformation of ESCC. Conclusions These results demonstrate that downregulation of HLA-DR promote the transformation of M2 TAMs, and participate in the invasion and metastasis of ESCC.


Background
The aggressive invasion and high metastatic capacity of esophageal squamous cell carcinoma (ESCC) are important factors contributing to the poor prognosis of this disease [1]. The occurrence and development of tumor from the participation of a series of complex factors, and the role of the surrounding tumor microenvironment is crucial [2].The tumor microenvironment of ESCC contains various in ammatory and immune cells, and tumor associated macrophages (TAMs) comprise the main component of the immune cell compartment [3]. TAMs can induce epithelial-mesenchymal transition (EMT) in ESCC cells to promote tumor invasion and metastasis [4][5][6]. Increased TAMs density in tumor tissue is a poor prognostic factor in patients with ESCC [7]. Macrophages can differentiate into two polarized states according to their function and phenotype: tumor-supportive (M2 type) macrophages and tumor suppressive (M1 type) macrophages [8]. Macrophages have functional plasticity and can alter their functional pro les in response to environmental stimuli. For example, when macrophages are exposed to lipopolysaccharide (LPS) or gamma interferon (IFN-γ), they are polarized to the M1 phenotype and exhibit antitumor activities. However, exposure of macrophages to Th2 cytokines, such as interleukin-4 (IL-4) and IL-13, results in polarization to an M2 phenotype [9]. Anti-CD47 treatment alone can shift the phenotype of macrophages toward an M1 subtype in vivo [10]. The regulation of macrophage phenotypes is the key to the role of macrophages in promoting or suppressing tumor progression. However, the molecular signals that regulate these phenotypes and the corresponding regulatory mechanisms remain to be elucidated.
The expression of MHC class II proteins (human leukocyte antigen class II, HLA-II) on antigen-presenting cells (APCs) promotes CD4 + T-helper-mediated activation of B cells and cytotoxic CD8 + T cells [11].
Macrophages are a classic type of APC cells. Studies have shown that there is a signi cant negative correlation between the expression of HLA class II molecules (HLA-DR, DP, DQ) and CD68 + CD163 + TAMs, and low expression of HLA-II and high-density TAMs are associated with poor prognosis of cancer patients [12,13]. HLA-DR as an important antigen of HLA-II gene, is associated with the immune response to tumor cells, which can lead to poor outcome of tumor patients [14]. Suggested that HLA-DR is involved in the regulation of tumor progression, which may be mediated by TAMs, but the regulatory effect and mechanism have not been reported. In the present study, we evaluated the effects of HLA-DR in TAMs polarization and the impact of it on malignant behavior of ESCC.

TAMs promoted migration, and invasion of ESCC
To understand the role of TAMs in the progression of ESCC, we co-cultured macrophages (PMA-treated THP-1) with ESCC cells (EC109 and EC9706). Compared with ESCC cells cultured alone, co-culture of ESCC with macrophages for 36 h altered cell morphology from the original round, oval-shaped morphology with epithelial characteristics to an elongated circle-shaped or spindle-shaped mesenchymal morphology, with enlarged gaps between cells, diffuse growth, and decreased adhesion (Fig. 1a). CCK8 and plate cloning experiments showed that TAMs did not signi cantly increase the proliferative capacity of ESCC cells (Fig. 1b-c). However, transwell assay with migration and invasion experiments demonstrated that TAMs signi cantly promote the migration and invasion of ESCC cells (Fig. 1d-e).
To interrogate the role of TAMs in ltration in ESCCs in vivo, we applied nude mice tumor formation experiment and human tissue for analysis. Tumor formation experiments in nude mice showed that the volume and weight of tumors from ESCC cells with chlorophosphate-depleted macrophages were signi cantly smaller than untreated group ( Fig. 1f-h). We used human ESCC tissues for further veri cation, and found that increased number of tumor-in ltrating CD68 + TAMs was closely related to lymphatic invasion, lymph node metastasis, and clinical progression of ESCC (P < 0.05, Fig. 1i). It suggested that there is a signi cant correlation between TAMs and tumor invasion and metastasis. In our previous research, we found that the VEGF-C in the VEGF family and MMP-9 in the tumor microenvironment have been proved possess a certain effect on ESCC metastasis [15].To further explore macrophage speci c regulatory role on ESCC invasion, we evaluated expression of VEGF-C and MMP9 (considered to be key genes that promote tumor invasion and metastasis) in tumor tissue. We found that the expression of VEGF-C and MMP9 in ESCCs was signi cantly higher than that in normal tissues (Fig. 2a--b), and there was a signi cant higher expression of VEGF-C and MMP-9 in TAMs compare with no-co-culture group (Fig. 2c). Consistently, it was con rmed by qRT-PCR (Fig. 2d), suggesting that TAMs may promote the invasion and metastasis of ESCC, which is closely related to VEGF-C and MMP9 secreted by TAMs.
M2 macrophages were the primary TAM phenotype that promoted malignant behavior of ESCC cells Macrophages are generally classi ed as M1 or M2 macrophages [16,17]. Immuno uorescence (IF) imaging was used to evaluate changes in macrophage phenotypes in a co-culture system, and found that the number of CD163-positive M2-type TAMs was signi cantly increased following co-culture with ESCC cells, while the number of M1-type macrophages (indicated by co-staining of HLA-DR and CD68) was signi cantly reduced ( Fig. 3a-b). We further evaluated macrophage phenotypes in vitro by qRT-PCR, and found that macrophages co-cultured with ESCC cells exhibited molecular characteristics of M2 phenotype macrophages, including signi cantly increased IL-10 expression and decreased TNF-α and HLA-DR expression (Fig. 3c). These data suggest that there is phenotypic transformation of macrophages in the ESCC microenvironment, and M2 TAMs may be the main component of macrophages in ESCC microenvironment.
Nude mice in Tumor formation experiments, we induced macrophages to an M2 phenotype in vitro, then coinjected these M2 macrophages along with ESCC cells as subcutaneous tumors. Found that tumors were larger and heavier in the M2 TAMs co-culture ESCC cell group compared with the ESCC cell-alone injection group ( Fig. 3d-f). Further veri cation in human ESCC tissues, we found that the increase in the number of in ltrating CD163-positive M2 TAMs was closely related to lymphatic invasion, lymph node metastasis, and clinical progression in patients with ESCCs (Fig. 3g). It is suggests that M2 macrophages are the primary TAMs phenotype that promotes malignant behavior of ESCC cells

Hla-dr Mediated Macrophage Phenotypic Transformation
To screen for molecular mechanisms that may be related to the M2 polarization of macrophages, we analyzed microarray data (dataset GSE95405 from the GEO database) to compare gene expression between THP-1 cells that were induced into M1 macrophages by treatment with phorbol ester (PMA), lipopolysaccharide (LPS) and IFN-γ, and THP-1 cells that were induced into M2 macrophages by treatment with IL-4 and IL-13 (Fig. 4a). Signi cant differences in the expression of multiple members of the HLA-DR family between M1 and M2 type macrophages were identi ed. Compared with M1-type macrophages, M2type macrophages expressed less HLA-DR (Fig. 4b). To address whether HLA-DR downregulation is a key molecular event involved in the transformation of macrophages into a cancer-promoting M2 phenotype, we transfected macrophages with HLA-DR lentiviral shRNA. HLA-DR expression in macrophages was signi cantly lower after shRNA transfection (Fig. 4c-e). Silencing HLA-DR signi cantly increased expression of the macrophage molecular marker IL-10, and signi cantly reduced expression of TNF-α (Fig. 4f). These data demonstrate that downregulation of HLA-DR promoting the macrophages polarization to M2 phenotype.

Effects of HLA-DR knockdown on TAM-mediated malignant behavior of ESCC
To investigate the effects of HLA-DR knockdown in macrophages on the biological behavior of ESCC tumor cells, we co-cultured HLA-DR knockdown macrophages and ESCC cells in vitro, and found that HLA-DR knockdown macrophages signi cantly increased the proliferation of ESCC cells compared with control macrophages (P < 0.05, Fig. 5a). Transwell migration and invasion assays showed that HLA-DR knockdown macrophages signi cantly promoted the migration and invasion of ESCC cells (P < 0.05, Fig. 5b-c). In addition, we found that knockdown of HLA-DR signi cantly increased the expression of VEGF-C and MMP-9 in TAMs (P < 0.05, Fig. 5d-e), suggesting that downregulation of HLA-DR in macrophages induces TAM M2 phenotypic transformation, promotes secretion of VEGF-C and MMP-9, and participates in the malignant transformation of ESCC.
Low expression of HLA-DR positively correlated with high-density of M2 TAMs, indicating high invasiveness and poor prognosis in ESCC Our experiments in vitro demonstrated that downregulation of HLA-DR contributes to macrophage M2 polarization and promotes malignant behavior of ESCC cells. To further verify the pro-cancer effect of HLA-DR downregulation, the expression of HLA-DR in clinical ESCC samples was evaluated, and found that HLA-DR is mainly expressed in stromal cells in ESCC, and that macrophages constitute an important population of HLA-DR expressing cells (Fig. 6a). Moreover, we observed that some CD68-positive TAMs did not express HLA-DR, suggesting that some TAMs may have lost HLA-DR, and represent a group of M2 phenotype TAMs ( Fig. 6a). Identi cation of M2 TAMs with CD163 staining revealed that ESCC areas with low levels of HLA-DR expression contained high densities of M2 TAMs. Furthermore, in areas with high expression of HLA-DR, there were fewer M2 TAMs (Fig. 6b). Comparing the distribution of M2 TAMs with clinicopathological information, we found that the lower expression of HLA-DR was positively correlated with an increase in the number of M2 TAMs, and positively correlated with invasion depth, lymph node metastasis, and clinical stage of ESCC (Table 1). Analysis of gene expression from the TCGA database and the prognostic data from patients in this study, revealed that lower expression of HLA-DR was closely related to poor prognosis ( Fig. 6c-d). These data suggest that lower expression of HLA-DR may be related to macrophage transformation to an M2 phenotype and associated with increased invasiveness and poor prognosis of ESCC. chemokines, and angiogenic factors [18]. TAMs have been reported to be related to the malignant behavior of tumors and to be important factors leading to poor prognosis in a variety of tumors, such as pancreatic ductal carcinoma and breast cancer [19][20][21]. It is crucial to further de ne the mechanisms of TAM phenotypic transformation and their cancer-promoting mechanisms.
In our study, co-culture of macrophages and tumor cells did not signi cantly change the proliferation of ESCC cells, but did signi cantly enhance the invasion and migration of ESCC cells. In vivo tumorigenesis experiments in nude mice further con rmed that depletion of TAMs by chlorophosphite treatment signi cantly inhibited tumor growth. These results are similar to a study by Fan et al, in which it was reported that TAMs can promotes tumorigenicity, tumor invasion and metastasis [22]. Our results further support the pro-tumor role for TAMs, as our analysis of human ESCC tissue specimens, revealed that the density of CD68 + TAMs in ESCC was positively related to vascular in ltration and lymph node metastasis.
This phenomenon has also been reported in variety of other tumors [23]. But speci c mechanisms by which TAMs promote the development of ESCC still need to be further explored. MMP-9 and VEGF have been investigated as important factors related to invasion and metastasis in tumors [24]. In this study, VEGF-C and MMP-9 were found to be highly expressed in ESCC, and TAM was an important source of them compared to normal macrophages. These indicated that TAM is an important factor in promoting the invasion and metastasis of ESCC, mechanistically due to its production of VEGF-C and MMP-9. It is similar to Li et al report that, the induction of VEGF and MMP9 expression by macrophages from the tumor microenvironment is also one of the important sources [25].
Previous studies have found that the TAMs in ESCCs are mainly M2 type. M2 TAMs are considered to have low antigen delivery capacity, which may facilitate immunosuppression and promote tumor vascular growth and lymph node metastasis [26]. In the present study, immuno uorescence and qRT-PCR were used to evaluate changes in TAM phenotypes in a co-culture system with ESCC cells. This study demonstrated that the number of CD163-positive M2-type TAMs was signi cantly increased following co-culture with ESCC cells, while the number of M1-type  [29,30]. Suggesting the loss of macrophage HLA-DR mediates immune escape and is an upstream molecular event that leads to M2 macrophage polarization.
In the co-culture system, we found that HLA-DR knockdown macrophages signi cantly increased the proliferation, migration and invasion of ESCC cells. In addition, we found that knockdown of HLA-DR signi cantly increased the expression of VEGF-C and MMP-9 in TAMs, suggesting that downregulation of HLA-DR in macrophages induces TAM M2 phenotypic transformation, promotes secretion of VEGF-C and MMP-9, and participates in the malignant transformation of ESCC.
Human ESCC tissue samples showed that HLA-DR was negatively correlated with the number of M2 type TAMs. This supports the ndings from our in vitro experiments. We also demonstrated that low expression of HLA-DR in ESCC tumor tissues was associated with low survival rates of ESCC patients. This is consistent with previous studies, which report that low HLA-DR expression is a risk factor for poor prognosis in various tumors [31,32]. Therefore, we believe that the low expression of HLA-DR has predictive value for the recurrence and metastasis of ESCC.

Western blotting
Western blotting was performed as previously described [33].

Microarray analysis from bioinformatics database
The Oncomine database (https://www.oncomine.org/resource/login.html) was used to analyze the expression of VEGF and MMP-9 in ESCC and normal tissues [34]. Microarray data were obtained from Gene Expression Omnibus (Geo, http://www.ncbi.nlm.nih.gov/geo) from the GSE95405 study. GSE95405 microarray data included gene expression data from M0 and M1 macrophages from six cases. The expression of HLA-DR-related molecules was analyzed from this dataset. Data Were log 2 transformed and the resulting heatmap was constructed using hemi software [35].
To further explore the expression and prognostic value of genes of interest in ESCC, we used a newly developed interactive website, ualcan, (http://ualcan.path.uab.edu/analysis. HTML) to analyze ESCC gene expression data from the TCGA database [36].

Statistical analysis
Statistical analysis was performed using SPSS 20.0 software (Chicago, IL, USA).. Measurement data between two groups were compared using paired tests and/or independent sample t-tests. Measurement data between three groups were compared using single factor analysis of variance. A difference was considered to be statistically signi cant at P < 0.05.

Consent for publication
Not applicable.

Availability of data and materials
All data generated or analyzed during this study are included in this published article .

Competing interests
The authors declare that they have no competing interests.