HMGB1 combining with tumor-associated macrophages enhanced lymphangiogenesis in human epithelial ovarian cancer Wenqi Zhang, Jing Tian & Quan Hao
Tumor Biology Tumor Markers, Tumor Targeting and Translational Cancer Research ISSN 1010-4283 Volume 35 Number 3 Tumor Biol. (2014) 35:2175-2186 DOI 10.1007/s13277-013-1288-8
1 23
Your article is protected by copyright and all rights are held exclusively by International Society of Oncology and BioMarkers (ISOBM). This e-offprint is for personal use only and shall not be self-archived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”.
1 23
Author's personal copy Tumor Biol. (2014) 35:2175–2186 DOI 10.1007/s13277-013-1288-8
RESEARCH ARTICLE
HMGB1 combining with tumor-associated macrophages enhanced lymphangiogenesis in human epithelial ovarian cancer Wenqi Zhang & Jing Tian & Quan Hao
Received: 22 August 2013 / Accepted: 2 October 2013 / Published online: 22 October 2013 # International Society of Oncology and BioMarkers (ISOBM) 2013
Abstract Within tumor microenvironment, high-mobility group box protein 1 (HMGB1) and tumor-associated macrophages (TAMs) are able to influence ovarian cancer development and progression via facilitating tumor lymphatic metastasis. However, little is known about the association between HMGB1 and TAMs on lymphangiogenesis in epithelial ovarian cancer (EOC). To investigate the effect of HMGB1 and TAMs on lymphangiogenesis in EOC, immunohistochemistry was performed to determine the expressions of HMGB1, TAMs, and lymphatic vessel density (LVD) in a total of 108 ovarian tissue specimens. Then, the relationships between HMGB1 or TAMs and LVD were assessed by correlation test. In our in vitro study, TAMs were isolated from ascites of EOC patients. Effects of HMGB1, TAMs, and HMGB1 combining with TAMs on lymphatic endothelial cell (LEC) proliferation, migration, and the capillary-like tube formation were measured. Results showed that the expression of HMGB1 and the number of TAMs infiltration were overexpressed in malignant ovarian tumors compared with that in normal ovarian and were closely associated with lymph node metastasis. Positive correlations existed between HMGB1 expression or TAMs count and LVD determination. In an in vitro study, data demonstrated that either HMGB1 or TAMs could facilitate lymphangiogenesis by inducing LEC proliferation, migration, and capillary-like tube formation. Meanwhile, HMGB1 combining with TAMs may augment the pro-lymphangiogenic W. Zhang : J. Tian : Q. Hao (*) Department of Gynecologic Oncology, Tianjin Medical University Cancer Institute and Hospital, National Clinical Research Centre of Cancer, Huanhuxi Road, Hexi District, Tianjin 300060, China e-mail:
[email protected] W. Zhang e-mail:
[email protected] W. Zhang : J. Tian : Q. Hao Key Laboratory of Cancer Prevention and Therapy, Tianjin, China
property. Our data suggest that either HMGB1 or TAMs could facilitate lymphangiogenesis, while HMGB1 coculture with TAMs may strengthen the pro-lymphangiogenic potential, which may serve as a therapeutic target for ovarian cancer. Keywords High-mobility group box 1 protein . Tumor-associated macrophages . Ovarian cancer . Lymphangiogenesis
Introduction Epithelial ovarian cancer (EOC) was revealed to be the fifth leading cause of cancer death in females and the most lethal gynecologic malignancy in the world [1]. Despite advances in detection and therapy, the 5-year survival rate for ovarian cancer patients is about 44 % [1]. Tumor metastasis is responsible for the poor outcome and cancer death. Under the clinical situation of EOC, lymph node is a common channel by which ovarian cancer cells spread and invade. Therefore, targeted disruption of established lymphatic vessels or inhibition of local lymphatic neoangiogenesis is a promising and effective measure to reduce metastasis [2]. However, in the past decades, the unique contributions of tumor microenvironment, especially the role of cytokines and tumor stromal cells in lymphangiogenesis, have not been widely explored. High-mobility group box protein 1 (HMGB1) is a highly versatile protein with both intra- and extracellular functions. Extracellular HMGB1 displays signaling activities on several cell types, such as endothelial cells, enterocytes, macrophages, and monocytes. The effect of HMGB1 may be mediated by an interaction with multiple surface receptors, such as the receptor for advanced glycation end products (RAGEs) and the Toll-like receptors (TLRs) [3–5]. It could work as a potent activator of monocyte cytokine synthesis that mediates a
Author's personal copy 2176
delayed and prolonged phase of monocyte activation [6]. HMGB1 has been reported to be closely related with tumorigenesis, angiogenesis, and metastasis in a number of malignancies [7–10]. In our previous observations, overexpressed HMGB1 is involved in the invasion and lymph node metastasis of cervical squamous cell carcinoma (CSCC), and HMGB1/RAGE pathway plays an important role in the metastasis of CSCC [11]. Tumor-associated macrophages (TAMs), which are generally considered to be polarized to M2 macrophages by cancerderived factors such as interleukin-6 (IL-6), leukemia inhibitory factor, and macrophage colony-stimulating factor (MCSF) in the ascites of advanced EOC patients, could in turn facilitate cancer metastasis and progression by modifying the tumor microenvironment [12, 13]. Increased TAM influx has been found in breast, prostate, endometrial, bladder, and ovarian cancers and correlated with inferior outcome [14–18]. Recent studies on cervical cancer and invasive breast cancer have shown that the VEGF-C- and VEGF-D-producing TAMs promote tumor lymphangiogenesis and lymph vascular invasion [19]. Increasing evidences compellingly demonstrate that both HMGB1 and TAMs are involved in lymphatic metastasis. Furthermore, HMGB1 could act as a cytokine that specifically stimulates cytokine synthesis in human monocytes [6]. Although we have identified that complicated interactions between inflammatory mediators and resident cells play indispensible role in the development and progression of tumor [20, 21], there has been no study investigating whether HMGB1 co-treated with TAMs could potentiate lymphangiogenesis in ovarian cancer. Hence, in our finding, we explored the influence of HMGB1 combining with TAMs on EOC lymphangiogenesis as well as for prognostic and treatment purposes.
Materials and methods
Tumor Biol. (2014) 35:2175–2186
Counting Kit-8 (CCK-8) was purchased from Beyotime (Shanghai, China). Transwell chamber was from Corning Company (Corning, NY, USA), and Matrigel was the product of BD Biosciences (Bedford, MA, USA). Cell lines and cell culture Human lymphatic endothelial cells (LECs) and complete endothelial cell medium (ECM) containing endothelial cell basal medium (EBM), 5 % fetal bovine serum (FBS), 1 % endothelial cell growth supplement (ECGS), and 1 % penicillin/ streptomycin solution (P/S) were purchased from ScienCell (Carlsbad, CA, USA). LECs were seeded in fibronectin (Millipore, Billerica, MA, USA)-coated culture bottles (Corning, Corning, NY, USA), and then, cells were incubated with 5 % CO2 at 37 °C. To digest LECs, 0.125 % trypsin– EDTA (Hyclone, UT, USA) was used. CD14+ cells (TAMs) were isolated from ascites of EOC patients (n =5) by standard density gradient centrifugation with Ficoll-Paque (Tianjin, China). Patient samples were obtained with written informed consent in accordance with the requirements of the Tianjin Medical University Cancer Institute and Hospital Ethics Committee. Monocytes were obtained from the mononuclear cell layer according to the method of Denholm et al. [22]. CD14+ cells (TAMs) were purified by positive selection using magnetic-activated cell sorting (MACS) technology (Miltenyi Biotec, Bergisch Gladbach, Germany) in accordance with the method of Duluc et al. [13]. After purification, cells were incubated in ECM at 37 °C with 5 % CO2 for 1 h, and then, nonadherent cells were vigorously washed off. To measure the purity of CD14+ cells, 1×105 cells were resuspended in PBS containing 2 % FBS and 0.01 % NaN3. Incubation was performed with 20 μl PE-CD14 mAb (BioLegend, San Diego, CA, USA) or isotype-matched control (BioLegend) for 30 min at 4 °C, and labeled cells were analyzed using the flow cytometry method. Results showed that CD14+ cells purity was always >90 %.
Tissue specimens and reagents Preparation of conditioned media A total of 108 human ovarian tissue specimens, including 88 EOC tissue specimens and 20 normal ovarian tissue specimens (diagnosed between 2007 and 2009), were obtained from the Department of Pathology, Tianjin Medical University Cancer Institute and Hospital. All of the EOC patients were clinically staged according to the International Federation of Gynecology and Obstetrics (FIGO) staging system, and none of them received adjuvant therapy prior to the primary surgery. The study was approved by the Institutional Medical Ethics Committee of Tianjin Medical University Cancer Institute and Hospital. Human recombinant HMGB1 and recombinant VEGF-C were obtained from R&D (Minneapolis, MN, USA). The Cell
CD14+ cells (TAMs), digested with 0.125 % trypsin–EDTA were seeded in 12-well plates at 1×106 cells/ml in ECM and then incubated at 37 °C with 5 % CO2 for 2 h. At the end of the incubation time, the supernatants were aspirated and replaced by ECM in the presence or absence of rHMGB1 (2 μg/ml). After 12 h, the conditioned media were collected and centrifuged to remove cellular debris, and the supernatants, including TAMs pretreated with rHMGB1 (2 μg/ml) and TAMs alone, were stored at 4 °C. Conditioned media were used without dilution. Then, ECM containing 2 μg/ml rHMGB1 or 5 ng/ml rVEGF-C or ECM alone was prepared for further use.
Author's personal copy Tumor Biol. (2014) 35:2175–2186
Immunohistochemistry In brief, paraffin-embedded specimens were cut into 4-μm sections and baked at 65 °C for 60 min. After sections were deparaffinized and rehydrated, they were submerged into antigenic retrieval buffer (pH 6.0 citric acid) for heatmediated retrieval. Then, the sections were treated with 3 % hydrogen peroxide to quench the endogenous peroxidase activity. Rabbit mAb HMGB1 (1:500 dilution; Abcam, Cambridge, MA, USA) and mouse mAbs including CD68 (1:300 dilution; Abcam) and D2-40 (1:40 dilution; Abcam) were incubated with the sections overnight at 4 °C. The detection was performed using a standard streptavidin peroxidase technique and chromogen diaminobenzidine (Shanghai, China). Normal ovary tissues from 20 patients were obtained as negative staining controls. A semiquantitative scoring system based on intensity of staining and distribution of positive cells was used to evaluate HMGB1 expression. The intensity of HMGB1-positive staining ranged from 0 to 3 (negative=0, weak=1, moderate=2, or strong=3), and the percentage of positively stained cells was scored as 0 (0 %), 1 (1 to 25 %), 2 (26 to 50 %), 3 (51 to 75 %), and 4 (76 to 100 %). The sum of the intensity and percentage score was used as the final staining scores (0 to 7). The sum indexes (−), (+), (++), and (+++) indicated final staining scores of 0, 1–3, 4–5, and 6–7, respectively. For statistical analysis, sum indexes (−) and (+) were defined as low HMGB1 expression, while sum indexes (++) and (+++) were defined as high HMGB1 expression. We determined the number of CD68+ cells which had infiltrated into cancer nests or stroma (intratumor TAMs). To count TAMs, each section was scanned at low (×40 and ×100) magnifications, and three representative areas were identified. Necrotic areas were excluded when calculating the number of intratumor TAMs. The TAMs were counted at ×200 magnification, and the average value of the three representative areas was used for statistical analysis. Quantification of lymphatic vessels was conducted according to the methods previously described [23]. Each specimen was scanned at low (×40 and ×100) magnifications to identify vascular hot spots. Areas of the greatest lymphatic vessel density were then examined and counted under a higher (×200) magnification. Lymphatic vessel density (LVD) was determined as the mean value of vessel counts. Cell proliferation assay in vitro LECs proliferation was determined by CCK-8 assay. Ninetysix-well plates were coated with fibronectin (2 μg/cm2). LECs at 90 % confluence were harvested and seeded in 96-well plates at a density of 4×103 cells/well containing 200 μl ECM. After cultivation for 6 h, the medium was aspirated, and the LECs were added with 200-μl aliquots of supernatants from cultures
2177
of TAMs alone or TAMs treated with rHMGB1 or ECM supplemented with rHMGB1 (2 μg/ml), while ECM supplemented with rVEGF-C (5 ng/ml) served as a positive control, and ECM alone served as a negative control. After 24, 48, or 72 h of incubation, all the media were removed, and cells in each well were incubated with 20 μl CCK-8 at 37 °C for 30 min. The absorbance of the dissolved formazan in each well was measured with a microplate spectrophotometer at 450 nm. Cell proliferation rate was calculated by the following formula: (treated group Abs450/untreated group Abs450)×100 %. Cell migration assay in vitro Migration assay was performed as previously described [24]. Twenty-four-well polycarbonate Transwell migration inserts (8 μm pore sizes) were coated with fibronectin. Cells were serum starved for 6 h, and cell suspensions containing 1×104 LECs were seeded on the upper compartments. The lower chambers were added with 1 ml ECM containing rHMGB1 at 2 μg/ml or 5 ×105 TAMs or 5 ×105 TAMs treated with rHMGB1 (2 μg/ml). ECM supplemented with rVEGF-C (5 ng/ml) served as a positive control, and ECM alone served as a negative control. Then, cells were allowed to migrate for 6 h toward the lower chamber at 37 °C with 5 % CO2. After incubation, cells on the upper surface of the filter were removed using a cotton swab; cells on the lower surface were fixed in precooled methanol (−20 °C) and stained with Thermo Scientific Three-Step Kit. Migrating cells were counted under the Olympus microscope in 10 randomized fields at ×200 magnification. Capillary-like tube formation assay in vitro Each well of 96-well plates was coated with 50 μl Matrigel stock solution, and the culture plates were kept on ice for 20 min and 37 °C for 30 min, respectively. Then, 100-μl aliquots of conditioned media, including supernatants from cultures of TAMs alone or TAMs pretreated with rHMGB1 or ECM supplemented with rHMGB1 (2 μg/ml), were mixed with LECs (1×104/well), while ECM supplemented with rVEGF-C (5 ng/ml) served as a positive control, and ECM alone served as a negative control. After incubation at 37 °C for 3, 6, 12, or 24 h, the capillary-like tube formation images were captured at a magnification of ×100 with a digital microscope camera system (Olympus, Tokyo, Japan). The level of the tube formation was quantified by counting the number of capillary-like tubes in five randomly fields from each well. Statistical analysis Statistical significance of differences was analyzed by twosided Student's t test, Χ 2 test, and one-way analysis of variance (ANOVA). Correlation test was conducted to investigate
Author's personal copy 2178
the relationships between the presence HMGB1 and LVD or TAMs and LVD. Survival curves were performed using the Kaplan–Meier method, and the log-rank test was used for the comparison of survival differences between groups. The Cox proportional hazard model was used for multivariate analysis. Statistical analyses were carried out using SPSS software 17.0. P value of 7.61 and ≤7.61 per ×200 field were distinguished into the high LVD and low LVD groups, respectively. Of 49 EOC specimens with high LVD, 39 were found to be HMGB1 overexpressed (79.6 %), while in the low LVD group, there were only 12 out of 39 specimens with elevated HMGB1 expression (30.8 %) (P 54 FIGO stage I to II III to IV Differentiation Well Moderate Poorly Histological type Serous cystadenocarcinoma Mucinous cystadenocarcinoma Endometrioid carcinoma Clear cell carcinoma Ascites
P 54
55.3±10.0 27–83 46 42
FIGO stage I to II III to IV Differentiation Well Moderate Poorly Histological type Serous cystadenocarcinoma Mucinous cystadenocarcinoma Endometrioid carcinoma Clear cell carcinoma Ascites Yes No Lymph node metastasis Yes No
P 35 Lymphatic vessel density Low High
number of soluble factors and cell–cell interactions. With reference to previous studies, HMGB1 and TAMs are two important participants of the regulatory machinery governing lymphangiogenesis within tumor microenvironment. However, the association between HMGB1 and TAMs on lymphangiogenesis in the human EOC had not been reported before. In this research, with immunohistochemistry, we observed that HMGB1 and TAMs were overexpressed in malignant EOC specimens and were associated with lymph node metastasis. Data here also revealed the independent prognostic value of HMGB1 for ovarian cancer, which was consistent with the previous study that Jie Chen et al. [30] had reported. In further analysis with correlation test, we identified that the HMGB1 expression and TAM infiltration were positively associated with LVD. We hypothesized that both may facilitate lymphangiogenesis and conducted an in vitro model to verify our assumption. Data showed that HMGB1 as well as
TAM number (mean ± SD)
P value 0.542
50.07±16.21 47.76±19.09 0.001
33 55
41.06±13.97 53.71±17.92
8 27 53
40.38±17.70 47.11±16.74 51.21.±17.80
46 6 28 8
49.70±18.65 39.83±18.48 52.07±15.10 40.75±17.08
69 19
49.78±18.21 46.00±15.16
18 70
65.06±8.01 44.83±17.00
11 77
40.45±15.30 50.18±17.64
39 49
41.26±13.59 55.10±18.10
0.217
0.231
0.409
0.000
0.086
0.000
TAMs isolated from ascites of EOC patients could promote lymphangiogenesis by inducing LEC proliferation, migration, and tube formation. Although the similar results that HMGB1 could induce lymphangiogenesis had been proved by Qiu et al. [31], interestingly, we performed a model of TAMs and HMGB1 co-treatment system, and the results here showed that TAMs co-treated with HMGB1 may exert strengthened effects on LEC biological behaviors. Multiple lines of evidence have demonstrated that the inflammatory tumor microenvironment predisposes tumorigenesis and orchestrates tumor progression [32]. The tumor microenvironment is quite complicated which consists of cancer cells as well as noncancer cells, including endothelial cells (ECs), cancer-associated fibroblasts (CAFs), TAMs, and noncellular components (pro-tumoral mediators) [33]. In our current study, we focus on two important participants: one is HMGB1, and the other is TAMs. HMGB1 works as a cytokine or a growth factor in septic inflammation and neoplasm.
Author's personal copy 2182
Tumor Biol. (2014) 35:2175–2186
Fig. 2 Correlations of HMGB1 expression and TAMs count with clinical prognoses. a Correlation between TAMs count and LVD (Pearson correlation coefficient, R 2 =0.242; P =0.000). b–c Kaplan-Meier estimates
Table 3 Multivariate Cox proportional hazards regression analysis of prognostic factors for survival of EOC Variables
Assigned score
FIGO stage I to II 0 III to IV 1 HMGB1 level Low 0 high 1 Lymph node metastasis No 0 Yes 1
HR (95 % CI)
P value
1 4.763 (1.564–14.503)
0.006
1 2.331 (1.030–5.277)
0.042
1 2.491 (1.174–5.286)
0.015
P
