Tumor-Associated Macrophages and Mast Cells Positive to Tryptase Are Correlated with Angiogenesis in Surgically-Treated Gastric Cancer Patients

Mast cells and macrophages can play a role in tumor angiogenesis by stimulating microvascular density (MVD). The density of mast cells positive to tryptase (MCDPT), tumor-associated macrophages (TAMs), and MVD were evaluated in a series of 86 gastric cancer (GC) tissue samples from patients who had undergone potential curative surgery. MCDPT, TAMs, and MVD were assessed in tumor tissue (TT) and in adjacent normal tissue (ANT) by immunohistochemistry and image analysis. Each of the above parameters was correlated with the others and, in particular for TT, with important clinico-pathological features. In TT, a significant correlation between MCDPT, TAMs, and MVD was found by Pearson t-test analysis (p ranged from 0.01 to 0.02). No correlation to the clinico-pathological features was found. A significant difference in terms of mean MCDPT, TAMs, and MVD between TT and ANT was found (p ranged from 0.001 to 0.002). Obtained data suggest MCDPT, TAMs, and MVD increased from ANT to TT. Interestingly, MCDPT and TAMs are linked in the tumor microenvironment and they play a role in GC angiogenesis in a synergistic manner. The assessment of the combination of MCDPT and TAMs could represent a surrogate marker of angiogenesis and could be evaluated as a target of novel anti-angiogenic therapies in GC patients.


Introduction
Mast cells (MCs) can play a role in tumor angiogenesis and their involvement has been found in spontaneous animal tumor models and human malignancies [1][2][3]. MCs are able to secrete classical pro-angiogenic factors, such as vascular endothelial growth factor (VEGF), fibroblast growth factor-2 (FGF-2), and thymidine phosphorylase (TP), and they are also capable to secrete non-classical pro-angiogenetic factors. [4,5]. Among these, tryptase is the most abundant factor stored in MC secretory granules and it can be degranulated in several ways, for example, C-Kit receptor activation. Published data has indicated that tryptase stimulates endothelial cell (EC) proliferation in a matrigel assay and induces microvascular proliferation in the chick embryo chorioallantoic membrane. In the last experimental model, microvascular proliferation was suppressed by gabexate, a tryptase inhibitor. It is well demonstrated that tryptase binds protease-activated receptor-2 (PAR-2) expressed on ECs, stimulating microvascular formation . Among the stromal cells homing to tumor microenvironments, macrophages (Ms) play several important roles [37]. In tumor microenvironments there are two types of Ms, Ms1 and Ms2 [37]. In particular, Ms1 are involved in inflammation and immune activity against tumor [38]. Ms2 are identified as tumor-associated macrophages (TAMs) and they play a role as pro-angiogenic cells [39,40]. It has been observed that Ms synthesize well known pro-angiogenic factors, particularly VEGF, TP, FGF-2, tumor necrosis factor-α (TNF-α), and interleukins (IL-1, -6, and -8). After exocytosis from Ms2, the above factors stimulate ECs towards proliferation, differentiation, survival, migration, and vascular permeability, thus leading to new microvessel formation [41,42]. Ms also induce angiogenesis by producing and releasing metalloproteinase-9 (MMP-9) that is able to degrade the extracellular matrix and then cause the release of VEGF [43,44].
In this pilot study, we analyzed MCDPT, TAMs and MVD in tumor tissue (TT) and in adjacent normal tissue (ANT) by immunohistochemistry and image analysis from 86 GC patients who had undergone radical surgery. The correlation between the studied parameters and the main clinico-pathological features has also been determined.

Results
Immunostained MCs appear as ovoidal-elongated cells with a red marked cytoplasm and a blue stained nucleus. MCs were scattered in the stromal microenvironment and it was evident that MCDPT were increased in TT vs. ANT ( Figure 1A vs. Figure 1B). For each serial section of TT and ANT, the five most immunostained areas (hot spots) were selected and MCDPT was counted in each hot spot and then the mean for each section and in the global series was calculated (Table 1).  Immunostained TAMs were identified as red marked cells with a blue nucleus. They were more pleomorphic and more abundant compared to MCs. With the image analysis system, two main shapes of Ms were seen: the first with an irregularly defined cytoplasm and a waving plasmalemma (like ameboid cell) with central nucleus (Figure 2A, single small arrow in the upper part); the second with an elongate cytoplasm and a well evident central nucleus (Figure 2A, single small arrow in the lower part). A higher count of TAMs was detected in TT vs. ANT (Figure 2A vs. Figure 2B). For each serial section of TT and ANT, the five most immunostained areas (hot spots) were selected and MCDPT were counted in each hot spot and then the mean for each section and in the global series was calculated (Table 1). Immunostained TAMs were identified as red marked cells with a blue nucleus. They were more pleomorphic and more abundant compared to MCs. With the image analysis system, two main shapes of Ms were seen: the first with an irregularly defined cytoplasm and a waving plasmalemma (like ameboid cell) with central nucleus (Figure 2A, single small arrow in the upper part); the second with an elongate cytoplasm and a well evident central nucleus (Figure 2A, single small arrow in the lower part). A higher count of TAMs was detected in TT vs. ANT (Figure 2A vs. Figure 2B). For each serial section of TT and ANT, the five most immunostained areas (hot spots) were selected and MCDPT were counted in each hot spot and then the mean for each section and in the global series was calculated (Table 1). For MVD, each red immunostained cell was considered a single microvessel and the diameter of the microvessel and the number lumen and red blood cells present in the lumen were considered when identifying a microvessel. Microvessels were scattered within stromal tissue. The blue nuclei of endothelial cells were often evident and higher MVD was observed in TT vs. ANT ( Figure 3A vs. Figure 3B). For MVD, each red immunostained cell was considered a single microvessel and the diameter of the microvessel and the number lumen and red blood cells present in the lumen were considered when identifying a microvessel. Microvessels were scattered within stromal tissue. The blue nuclei of endothelial cells were often evident and higher MVD was observed in TT vs. ANT ( Figure 3A vs. Figure 3B). The mean value ± standard deviation (SD) regarding MCDPT, TAMs, and MVD in TT and ANT was 11.38 ± 4.32, 49.17 ± 17.56, 28.12 ± 8.98 and 2.98 ± 1.45, 15.34 ± 6.21, 10.39 ± 5.62, respectively, and these differences were significant (p ranged from 0.001 to 0.002; Table 1). The actual values of MCDPT, TAMs, and MVD in TT and ANT are shown in the histograms (Figures 4-6) respectively. Interestingly, in TT, a higher MVD was associated with higher MCDPT and TAMs, thus supporting the role of these stromal cells in neovascularization.  The mean value ± standard deviation (SD) regarding MCDPT, TAMs, and MVD in TT and ANT was 11.38 ± 4.32, 49.17 ± 17.56, 28.12 ± 8.98 and 2.98 ± 1.45, 15.34 ± 6.21, 10.39 ± 5.62, respectively, and these differences were significant (p ranged from 0.001 to 0.002; Table 1). The actual values of MCDPT, TAMs, and MVD in TT and ANT are shown in the histograms (Figures 4-6) respectively. Interestingly, in TT, a higher MVD was associated with higher MCDPT and TAMs, thus supporting the role of these stromal cells in neovascularization. From a statistical point of view, significant correlations between MCDPT and TAMs (r = 0.77, p = 0.02), MCDPT and MVD (r = 0.74, p = 0.02), and TAMS and MVD (r = 0.79, p = 0.01) (Figure 7) were shown. With special regard to TT, no correlation between MCDPT, TAMs, MVD, and the main clinic-pathological features was found ( Table 2).

Discussion
Substantial data has indicated the key role of angiogenesis in GC development and progression, but little data related to the role of both MCDPT and TAMs in GC angiogenesis have been published [40,51,52]. Tumor angiogenesis is the process of new blood vessel formation from the pre-existing vascular network. This process takes place in the tumor microenvironment where stromal cells, particularly macrophages and mast cells, induce microvessel formation by means of pro-angiogenic factors [53][54][55][56]. In the stromal microenvironment, MCs and TAMs can be attracted by the pro-inflammatory cytokines secreted from tumor cells and other inflammatory cells. A lot of research has indicated that either MCDPT and TAMs are correlated with MVD degree in animal and human malignancies [39][40][41][42], but little data have been reported concerning the concomitant evaluation of MCDPT and TAMs and their correlation with tumor angiogenesis in terms of MVD.
With special regard to MCs, Mukherjee et al. [57] assessed MC density in tissue from patients affected by gastric ulcers and by GC. Data from this research group were obtained utilizing the histochemical stain of toluidine blue to identify and count MC density. Results indicated that MC density increased in benign gastric ulcers and in cancers compared to control tissue, and furthermore that MC density correlated with angiogenesis.
Ribatti et al. [47] studied TT from GC patients using immunohistochemistry with anti-tryptase and anti-chymase antibodies to stain MCs. This study showed a correlation between MVD and tryptase and chymase-positive MCs, demonstrating the role of MCs in neovascularization.
Recently, we published data indicating that MCDPT may induce angiogenesis in bone metastases from gastric cancer patients, suggesting that MCDPT is able to stimulate angiogenesis in metastatic tumor sites [31].
In the tumor microenvironment, MCs can be activated in different ways, including c-kit receptor stimulation by its ligand, stem cell factors, the IgE-dependent mechanism mediated by T lymphocyte-MC interaction, and TLR activation by other microenvironmental stimuli [58,59]. After activation, intensive or piecemeal degranulation of secretory granules occurs and MC-derived tryptase is released into the tumor microenvironment.
In bench experimental matrigel assays and in chick embryo chorioallantoic membrane assays, tryptase induces EC proliferation and neovascularization which was suppressed by tryptase inhibitors. Tryptase is an agonist of PAR-2 on vascular ECs that induces their proliferation. Following PAR-2 stimulation, the MAPK pathway is stimulated, leading to proliferation and angiogenesis .
MCDPT and TAMs are linked via Toll-like receptors (TLRs). Recent published pilot data from our group demonstrated that MCDPT and TAMs paralleled each other in colon cancer tissue and they increase with increased angiogenesis. TLRs are a family of membrane-spanning, non-catalytic receptors expressed in immune stromal cells including TAMs and MCs [40,41,59].
Increased expression of TLRs has been demonstrated in tumor cells, tissues, and tumor cell lines. TLR4 has been observed as over-expressed in human and mouse colorectal neoplasia, and on the other hand, TLR4-deficient mice are refractory to colon carcinogenesis, indicating that the higher TLR presence on tumor cells induces tumor development directly or indirectly by mean of angiogenesis [61].
According to these biological backgrounds, our results suggested that the increment of both MCDPT and TAMs were correlated with increased MVD. Our results also demonstrated that MCDPT, TAMs, and MVD are much higher in TT compared to ANT, supporting the theory that the above cell types play a role in tumor development and angiogenesis.
To the best of our knowledge, no other study has been published regarding the concomitant assessment of both MCDPT and TAMs and neovascularization in TT and ANT from GC patients. We propose that the coupled MCDPT and TAMs could be considered as a surrogate biomarker of the degree of GC angiogenesis [50,[63][64][65]. From a therapeutic point of view, the data obtained support a novel possibility to inhibit GC angiogenesis by targeting the coupled TAMs and MCDPTs by mean of several agents (e.g., trabectedin, peptide M2, PLX3397, STI571, AB1010). Finally, available tryptase inhibitors such as Gabexate or Nafamostat mesilate may be tested in future clinical trials as an innovative anti-angiogenic strategy [66][67][68][69][70].

Study Population
Eighty-six GC patients diagnosed by preoperative gastric endoscopy were selected to undergo potential curative surgery resection. The surgical techniques were open total and sub-total gastrectomy with D2 lymph node dissection. Selected cases were staged as T 2-3 N 2-3 M 0 (by AJCC for Gastric Cancer 7th Edition) according to the American Joint Committee on Cancer 7th edition (AJCC-TNM) classification [71][72][73][74]. Clinical staging was performed by computed tomography (CT) of the thorax, abdomen, and pelvis. All enrolled patients had adenocarcinomas. The main clinico-pathological characteristics of the patients are reported in Table 2. The research was developed according to the Declaration of Helsinki, and the study was approved by the Ethics Committee of the "Mater Domini" Hospital, "Magna Graecia" University, Catanzaro (No. 242; 22 December 2016). Signed consent from each patient was obtained.

Immunohistochemistry
MCDPT, TAMs, and MVD were detected by immunohistochemistry using a three-layer biotin-avidin-peroxidase system [75]. Briefly, 6-µm-thick serial sections of formalin-fixed and paraffin-embedded TT and ANT were cut. Obtained slides were processed with a microwave oven at 500 W for 10 min, and then the endogenous peroxidase enzyme was inhibited with 3% hydrogen peroxide solution. Subsequently, slides were posted with the following primary antibodies: anti-tryptase (clone AA1; Dako, Glostrup, Denmark) diluted 1:100 for 1 h at room temperature (for MCs identification), anti-CD68 (clone KP1; Dako, Glostrup, Denmark) diluted 1:100 for 1 h at room temperature as the TAMs marker, anti-CD31 antibody (QB-END 10; Bio-Optica Milan, Milan, Italy) diluted 1:50 for 1 h at room temperature as a pan-endothelial marker. The immunoreactivity was detected by employing a biotinylated secondary antibody, and the avidin-biotin peroxidase complex yielded a red chromogen (LPS, K0640, Dako, Glostrup, Denmark). Cell nuclei were stained with Gill's haematoxylin No. 2 (Polysciences, Warrington, PA, USA). No primary antibody was posted in negative controls.

Morphometrical Assay
Light microscopy integrated with an image analysis system (AXIO, Scope A1, ZEISS, Germany) was utilized [75]. For each serial section of TT and ANT, the five most immunostained areas (hot spots) were selected at low magnification. Next MCDPT, TAMs, and MVD were assessed at ×400 magnification (0.19 mm 2 area) in the five identified hot spots areas for each serial section, respectively ( Figure 1A,B, Figure 2A,B and Figure 3A,B). With special reference to MCDPT and TAMS, each immunostained cell was considered in their count. Furthermore, MVD was detected by counting single red-brown stained endothelial cells, endothelial cell clusters, and microvessels, clearly separated from adjacent microvessels.

Statistical Analysis
Mean values for each section and in the global series was obtained for all studied parameters in both TT and ANT groups. The difference between groups was measured by Student's t-test. Mean values ± 1 Standard Deviation (SD) of all the evaluated tissue parameters are reported in Table 1.
Correlations between MCDPT, TAMs, and MVD were calculated using Pearson's (r) analysis ( Figure 4). Correlations among all the analyzed parameters and the main clinico-pathological features listed in Table 2 were performed by the χ-square test (χ 2 ). All analyses were considered statistically significant with a p < 0.05. Statistical analyses were performed with the SPSS statistical software package (SPSS, Inc., Chicago, IL, USA).