Protective role of mouse mast cell tryptase Mcpt6 in melanoma

Abstract Tryptase‐positive mast cells populate melanomas, but it is not known whether tryptase impacts on melanoma progression. Here we addressed this and show that melanoma growth is significantly higher in tryptase‐deficient (Mcpt6−/−) versus wild‐type mice. Histochemical analysis showed that mast cells were frequent in the tumor stroma of both wild‐type and Mcpt6−/− mice, and also revealed their presence within the tumor parenchyma. Confocal microscopy analysis revealed that tryptase was taken up by the tumor cells. Further, tryptase‐positive granules were released from mast cells and were widely distributed within the tumor tissue, suggesting that tryptase could impact on the tumor microenvironment. Indeed, gene expression analysis showed that the absence of Mcpt6 caused decreased expression of numerous genes, including Cxcl9, Tgtp2, and Gbp10, while the expression of 5p‐miR3098 was enhanced. The levels of CXCL9 were lower in serum from Mcpt6−/− versus wild‐type mice. In further support of a functional impact of tryptase on melanoma, recombinant tryptase (Mcpt6) was taken up by cultured melanoma cells and caused reduced proliferation. Altogether, our results indicate a protective role of mast cell tryptase in melanoma growth.

response to various ligands to the MRGPRX2 receptor (McNeil et al., 2015). When MCs are activated, they can respond by degranulation, whereby the preformed contents of their granules are released.
MC activation can also drive the synthesis of a number of additional compounds, such as prostaglandins and leukotrienes, as well as numerous cytokines, chemokines, and growth factors (Galli, Nakae, & Tsai, 2005).
MCs are undoubtedly mostly known for their detrimental impact in allergic conditions. However, they are also known to contribute in a variety of additional disorders. In particular, there is a vast documentation supporting an involvement of MCs in various malignancies (Marichal, Tsai, & Galli, 2013;Oldford & Marshall, 2015;Ribatti & Crivellato, 2009;Varricchi et al., 2017). In this respect, a large number of clinical studies have revealed that MCs show a strong tendency to accumulate at malignant lesions. Typically, MCs are most frequently found in the tumor stroma but they can also be found within the tumor parenchyma. Clinical studies have also examined the correlation between MC presence and clinical outcome, and these studies have often led to the conclusion that MC density is associated with a poor outcome (Marichal et al., 2013;Oldford & Marshall, 2015;Ribatti & Crivellato, 2009;Varricchi et al., 2017).
However, there are also numerous occasions in which MC presence appears to associate with good outcome (Marichal et al., 2013;Oldford & Marshall, 2015;Ribatti & Crivellato, 2009;Varricchi et al., 2017). Altogether, there is thus some controversy with regard to how MCs impact on cancer.
In melanoma, a number of clinical studies have revealed a strong accumulation of MCs (Ribatti, Ennas, et al., 2003;Siiskonen et al., 2015;Toth-Jakatics, Jimi, Takebayashi, & Kawamoto, 2000). To gain more mechanistic insight into the role of MCs in melanoma, we previously examined melanoma progression in wild-type (WT) mice versus mice that lacked connective tissue-type MCs altogether (Dudeck et al., 2011), using a model of melanoma colonization to the lung. In that study, we found that MCs had an overall detrimental impact (Öhrvik et al., 2016) and to dissect the underlying mechanism we assessed mice lacking all of the MC proteases in the same model. Strikingly, we found that mice with global MC protease deficiency, contrary to the mice lacking MCs altogether, developed larger tumors than did WT mice (Grujic et al., 2017). This suggests that the combined action of the MC proteases serves a protective function in this model of melanoma, that is, despite the overall negative impact of MCs. Further, the data indicated that the protective function of the MC proteases could be related to CXCL16/CD1ddependent effects on iNKT cell populations (Grujic et al., 2017).
Collectively, these findings indicate that MCs harbor both detrimental and protective activities with respect to tumor progression, and the overall impact of MCs may thus reflect the balance between such activities.
The aim of this study was to further dissect the impact of the MC proteases on melanoma progression. We thereby focused on tryptase, based on our previous observation that tryptase in certain settings can have an anti-proliferative action (Melo et al., 2017).
For this purpose, we evaluated mice lacking tryptase Mcpt6 in a subcutaneous model of melanoma. Our findings reveal a protective function of tryptase in melanoma progression, hence shedding further light into the role of MCs in tumor development.

| Mice
WT and Mcpt6-deficient mice were all on C57BL/6J genetic background. Eight-to 16-wk-old mice were used in all experiments. All experiments were approved by the Local Ethics Committee (Uppsala djurförsöksetiska nämnd).

| Cells, tumor inoculation, and tissue collection
The cell line B16.F10 (ATCC; CRL-6475) was a gift from A.R. Thomsen (Copenhagen University, Denmark). Tumor cells were cultured in DMEM supplemented with 10% FBS, 1% L-glutamine, and 1% penicillin and streptomycin solution. Prior to subcutaneous injections, cells reaching approximately 90%-100% confluency were trypsinized, resuspended in Hanks' balanced salt solution, and counted using trypan blue in order to adjust the cell concentration to 500,000 cells/ml. A total of 50,000 B16F10 cells (100 µl of cell suspension) were injected subcutaneously in the hip region (both sides). From day 7 post-injection and every two days, the mice were examined for tumor growth. The size of tumors (a-length, b-width, tumor volume = (axb 2 )/2) was determined with a caliper, and the mice were sacrificed when the tumor volume reached 1,100 mm 3 (all mice within an experimental group were sacrificed when one of the animals reached a tumor volume of 1,100 mm 3 ). Blood was collected from B16F10 cell-injected and naïve mice into 1.5-ml microtubes and left to coagulate at room temperature for at least 1h. Blood samples were centrifuged at 2000g (4°C, 20 min), and serum was aliquoted and stored at −80°C for ELISA analysis. Tumors and inguinal lymph nodes were frozen on dry ice and stored at −80°C for gene array and qPCR analysis. Tumors were alternatively placed in 4% formalin (PBS-buffered) solution for histological analysis and immunohistochemistry.

Significance
Mast cells are known to be present in melanoma, but their role in tumor progression has not been clarified previously. Here we show that one of the enzymes expressed in large amounts by mast cells, tryptase, has a protective role in melanoma progression in a mouse model. Thereby, it is conceivable that modulation of tryptase function can represent a novel principle for interfering with melanoma growth.

| Histochemistry and immunohistochemistry
Sections from paraffin-embedded tissue were deparaffinized and rehydrated, and epitope retrieval was performed by covering sections with 20 μg/ml of proteinase K in 50 mM Tris, pH 8.0, 1 mM EDTA, and 0.5% Triton X-100 and incubated for 15 min at 37°C in a humidified chamber. Samples were brought to room temperature for 10 min, followed by rinsing with PBS-T (10 mM phosphate buffer, pH 7.4, 2.7 mM KCl, 140 mM NaCl, and 0.1% Tween-20) and blocking for 10 min with Background Sniper solution (Biocare Medical, Pacheco, CA) at room temperature. Incubation with rabbit anti-CPA3 immune serum (1:500) (Rönnberg & Pejler, 2012) was performed overnight at 4°C, followed by addition of goat anti-rabbit Alexa 633 (1:1,000) at room temperature for 1h. Rabbit anti-Mcpt6 immune serum (1:500) (Rönnberg & Pejler, 2012) was added overnight at 4°C, followed by incubation with goat antirabbit Alexa 488 (1:1,000) at room temperature for 1h. Controls were prepared in parallel by using immune serum at the same concentration as the primary immune serum. Samples were extensively rinsed between all steps with PBS-T. ActinRed TM 555 and NucBlue probes were
EdU staining and flow cytometry were performed as described

| RNA extraction, quantitative RT-PCR (qPCR), and gene array analysis
Mice were euthanized, and tumors were collected, frozen on dry ice, and stored at −80°C until use. Tissues were homogenized in TRIzol reagent (Thermo Scientific, Wilmington, DE) using a polytron PT 1200 (Kinematica AG, Luzern, Switzerland). The homogenate was centrifuged at 12,000 g for 1 min, and 500 µl of the supernatant (corresponding to ~50 mg tissue) was used for total RNA isolation using the Direct-zol Subsequently, qPCR was performed using up to 100 ng cDNA, 400 nM primers (indicated in Supporting Table S1) and iTaq Universal SYBR Green Supermix (Bio-Rad, Hercules, CA), following the PCR cycling conditions recommended by the manufacturer, on the C1000 Touch Thermal Cycler instrument (Bio-Rad, Hercules, CA). Each sample was run in duplicates/triplicates, and qPCR data analysis was performed using the Bio-Rad CFX Maestro program. Gene expression levels were presented relative to the house keeping gene (glyceraldehyde 3-phosphate dehydrogenase; GAPDH) and relative either to WT inoculated mice or to respective non-inoculated naïve mice.
For analysis of miR3098 and miR669b, first-strand cDNA was synthesized using Qiagen miRCURY LNA RT kit (cat.# 339340) followed by qPCR with the Qiagen miRCURY LNA SYBER Green PCR kit (cat.# 339345) and miRCURY LNA miRNA PCR assays (primers) specified in Supporting Table S1. qPCR samples were run in duplicates, and gene expression levels were presented relative to non-coding 5S-rRNA and relative to WT inoculated mice.

| Statistical analysis
All analyses were performed in GraphPad Prism using two-tailed as mean values ± SEM. A p value ≤ .05 was considered statistically significant.

| Tumors develop more rapidly in Mcpt6 −/− than in WT mice
To study the impact of tryptase on tumor progression, we injected

| MCs are present in the stroma of subcutaneous melanoma tumors
To approach how MCs impact on the subcutaneous melanoma tumors, we first assessed the presence and location of MCs in the tumors. To identify MCs, we used toluidine blue staining. As seen in

| Mcpt6 is released from tumor-infiltrating MCs and enters melanoma cells
As judged by toluidine blue staining, the tumor-associated MCs were predominantly located in the tumor stroma and were poorly visible in the tumor parenchyma ( Figure 2). However, when using confocal microscopy, MCs were more readily detected within the tumor parenchyma (Figure 3a). Similar to the MCs in the stroma, these MCs were double-positive for Mcpt6 and CPA3, indicating that they were of CTMC subtype. An intriguing observation was that Mcpt6 was released from the parenchymal MCs, and was widely spread within the tumor parenchyma. An assessment of the distance between released Mcpt6 + granules and tumor-associated MCs revealed that granules can be found up to ~40 µm from the closest located MCs within the tumor parenchyma and up to ~60 µm from the closest located MCs in the tumor stroma ( Figure 4). Moreover, the confocal analysis suggested that the released Mcpt6 was in fact found inside the tumor cells, partially localized to the tumor cell nuclei (Figure 3b; Supporting Video S1).
CPA3 was also released from the MCs populating the tumor parenchyma, although the extent of release and uptake into tumor cells was less pronounced (Figure 3b).

F I G U R E 4
Mcpt6 + granules can be found at a large distance from MCs in the tumor stroma and parenchyma. 3D images from Z-stack confocal sections of tumor stroma (a) and parenchyma (b) were used to measure the distance between Mcpt6 + granules and the closest located MC. (c) Distances between MCs and individual granules were measured in µm. Scale bars = 10 μm

| Recombinant Mcpt6 is taken up by melanoma cells and causes reduced proliferation
To provide further mechanistic insight into how tryptase affects melanoma cells, we generated recombinant Mcpt6 and assessed how it affects cultured B16.F10 melanoma cells. As seen in Figure 5a Tables S2 and S3). To verify and quantify these data, we performed qPCR analysis, by focusing on a number of genes of interest. These included the IFNγ-dependent genes Cxcl9, Gbp2, and Gbp10, as well as Tgtp2. qPCR analysis showed that all of these genes were significantly higher expressed in WT versus Mcpt6 −/− tumors. There was also a trend of higher Gbp2 expression in WT versus Mcpt6 −/− tumors. In contrast, the IFNγ gene was not differently expressed ( Figure 7). The gene array screen also indicated that a number of microRNAs were differently regulated as a consequence of Mcpt6 deficiency. Indeed, qPCR analysis confirmed that one of these, 5p-miR3098, was significantly lower expressed in tumors from WT versus Mcpt6 −/− mice (Figure 7). To verify the effect on Cxcl9 at the protein level, ELISA analysis was conducted. In agreement with the gene array and qPCR data, we found that CXCL9 levels were significantly higher in serum from WT versus Mcpt6 −/− mice (Figure 8). We also measured IFNγ levels in serum, but found no significant difference between WT and Mcpt6 −/− animals ( Figure 8).
Next, we used qPCR analysis to evaluate whether the absence of

| D ISCUSS I ON
The bona fide role of MCs in malignancies is still enigmatic. Intriguingly, their role may vary considerably among different types of cancers and perhaps also in the various stages of a malignant condition (Varricchi F I G U R E 6 Recombinant Mcpt6 is taken up by melanoma cells. B16F10 melanoma cells were cultured with or without 50 nM recombinant Mcpt6 as indicated. After 24h, the cells were stained with anti-Mcpt6 antibody and were analyzed by confocal microscopy. Cells were co-stained using a nuclear marker (Hoechst 33342; blue) and an actin probe (ActinRed; red). The figure depicts representative images. White arrows indicate positive Mcpt6 staining. Note that Mcpt6 staining is seen at the cell surface and also in the cytoplasm (lower panel), the latter indicating uptake of Mcpt6. Note also that no positive staining is seen in the absence of added Mcpt6 (control; left panels) et al., 2017). Clearly, since a profound MC accumulation is evident in essentially all investigated malignant settings (Marichal et al., 2013;Oldford & Marshall, 2015;Ribatti & Crivellato, 2009;Varricchi et al., 2017), it is reasonable to assume that they indeed play a significant role. It is notable that MCs are often among the first types of immune cells that arrive at a malignant lesion (Dalton & Noelle, 2012;Oldford & Marshall, 2015), and we may thus envision that effects of MCs are of particular importance in the early phases of cancer development.
However, MCs are consistently found also in fully developed cancers and are thereby likely to affect also late stages of malignant processes.
However, it remains to be demonstrated by using direct experimental approaches (e.g., MC-conditional deletion of the respective genes) that MCs affect malignant development by affecting angiogenesis.
In this study, we show that MC tryptase, Mcpt6, affects the outcome in a subcutaneous model of melanoma. Our data indicate that Mcpt6, under the experimental conditions employed, has a protective role, which is in contrast with the more common view of MCs as detrimental players in melanoma and other cancers. However, it is notable that our findings are in line with a previous study where we showed that the collective absence of all of the proteases expressed by connective tissue-type MCs led to a higher colonization of lungs with melanoma cells following their i.v. administration (Grujic et al., 2017). Moreover, our findings are consistent with clinical evidence in support of a correlation between MC protease expression and protection against melanoma (Crincoli et al., 2019;Siiskonen et al., 2015;Stieglitz et al., 2019). A protective role of MCs is also supported by a study where secreted MC mediators showed anti-proliferative effects on cultured melanoma cells, although the nature of the active MC mediator(s) was not revealed (Stieglitz et al., 2019).
An important issue is to define how MC tryptase exerts its protective function in melanoma. To address this, we first assessed whether MCs were present in the melanomas, as well their location within the tumor and their phenotype. In agreement with many previous studies, our findings reveal that MCs are predominantly F I G U R E 7 Mcpt6 deficiency affects gene expression in melanoma tumors. Total RNA was isolated from tumors of WT and Mcpt6-deficient mice. Subsequently, total RNA was subjected to qPCR analysis of expression of Cxcl9, Tgtp2, Gbp10, Gbp2, and miR3098. Expression of genes was evaluated relative to either GAPDH (proteinencoding genes) or 5S-rRNA (miRNAs), and normalized to WT mice. Note the increased expression of Cxcl9, Tgtp2, and Gbp10, and lower expression of miR3098 in tumors from WT versus Mcpt6 −/− mice. Results are presented as mean values ± SEM (n = 7-13); Mann-Whitney test. *p ≤ .05 detected in the tumor stroma, but were more rare in the tumor parenchyma by conventional staining techniques. However, when instead using confocal microscopy assessment with an anti-Mcpt6 antibody, we found that MCs were in fact relatively abundant also within the tumor parenchyma. An important implication of these data was also that the tumor-associated MCs, both in the stroma and within the parenchyma, express tryptase. Since they also expressed CPA3, this indicates that the tumor-associated MCs are of CTMC rather than MMC subtype. This is thus in agreement with a previous study where it was shown that the tumor-infiltrating MCs in glioblastoma were of CTMC type (Polajeva et al., 2011).
Another important observation, from both conventional and confocal microscopy assessment, was that the tumor-associated MCs frequently showed signs of degranulation, hence suggesting that they release their preformed granule compounds into the tumor microenvironment. Further, an intriguing observation was that tryptase-positive granules were widely distributed within the tumor, not only in the close vicinity to MCs. This suggests that MCs have the capacity to influence tumor cells that are located both in their close proximity but also at a larger distance. Clearly, this can explain why MCs, despite being relatively rare within tumors, can have an impact on the overall tumor progression.
Intriguingly, tryptase released from MCs in the tumor parenchyma was frequently found within the tumor cells, that is, indicating cellular uptake. This was also supported by our in vitro approach, F I G U R E 8 Higher levels of CXCL9 but not of IFNγ in serum of WT compared to Mcpt6 −/− mice. 50,000 B16F10 cells were injected subcutaneously in the hip region of WT or Mcpt6 −/− mice. After termination of the experiment, blood was collected. Blood was also collected from naïve WT and Mcpt6 −/− mice. (a-b) Levels of CXCL9 in serum from (a) B16F10-injected (n = 12) and (b) naïve (n = 11) mice, measured by ELISA. Note the higher levels of CXCL9 in serum from B16F10-injected WT versus Mcpt6 −/− mice, whereas no differences in CXCL9 levels were seen between WT and Mcpt6 −/− naïve mice. (c) IFNγ levels in serum from B16F10-injected WT and Mcpt6 −/− mice, measured by ELISA (n = 12). Results are presented as mean values ± SEM; Mann-Whitney test. *p ≤ .05 F I G U R E 9 qPCR analysis of tumors for the expression of the cell-specific markers CD4, CD8, F4/80, CD11c, CD206, CD19, and CPA3. Total RNA was isolated from tumors of WT and Mcpt6 −/− mice. Total RNA was subjected to qPCR analysis for expression of the CD4, CD8, F4/80, CD11c, CD206, CD19, and CPA3 genes. Gene expression was evaluated relative to GAPDH (housekeeping gene) and presented as 1/∆Ct. Note that the qPCR analysis indicated that all of these populations were equally represented in WT and Mcpt6 −/− mice. Results are presented as mean values ± SEM (n = 9-11). **p ≤ .01; ****p ≤ .0001 where it was revealed that recombinant Mcpt6 is taken up by cultured melanoma cells. In an earlier study, we showed that Mcpt6 can have an impact on gene expression in MCs (Melo et al., 2014) and we also showed recently that human tryptase impacts on gene expression in human melanoma cells (Rabelo Melo et al., 2019).
Prompted by these findings, we evaluated whether Mcpt6 can affect gene expression patterns within melanoma tumors. Indeed, we found that a number of genes were differently expressed in tumors lacking Mcpt6. Among these were genes associated with IFNγ signaling, including Cxcl9 and the intracellular GTPase Gbp10. Increased Cxcl9 expression was also supported by protein analysis, revealing higher levels of CXCL9 protein in serum from WT versus Mcpt6 −/− tumor-bearing animals. CXCL9 is implicated as a tumor suppressor, due to its ability to recruit CD8 + cells (Harlin et al., 2009). However, we did not see any apparent impact of Mcpt6 deficiency on CD8 expression in the tumors, suggesting that any impact of CXCL9 on tumor progression may be explained by effects beyond its influence on CD8 + T-cell recruitment. To date, the role of mouse Gbp10, one of eleven members of the Gbp family, in tumor biology is unclear.
However, in human cutaneous melanoma, several Gbp mRNAs are associated with favorable prognosis (Wang et al., 2018). Hence, the higher expression of Gbp10 in WT versus Mcpt6 −/− mice may represent a protective function in melanoma development. We also found upregulated expression of an additional GTPase, Tgtp2, in WT versus Mcpt6-deficient animals. However, the role of Tgtp2 in tumor settings remains to be explored. We also found that the absence of Mcpt6 was associated with elevated expression of the micro RNA miR3098. Interestingly, miR3098 has previously been implicated in lung cancer (Wang, Xu, & Wang, 2017) and diabetes (Rubin, Salzberg, Imamura, Grivitishvilli, & Tombran-Tink, 2016), although the exact role miR3098 under these conditions has not been clarified.
To provide further mechanistic insight into how tryptase affects the melanoma cells, we performed experiments in which the effect of recombinant Mcpt6 on cultured melanoma cells was assessed. Interestingly, these analyses revealed that Mcpt6 caused a substantial decrease in the proliferation of the melanoma cells.  (Marichal et al., 2013;Oldford & Marshall, 2015;Ribatti & Crivellato, 2009;Varricchi et al., 2017). We may thus envision that the exact expression profile of MCs present in a given tumor setting may determine their net impact on tumor growth. On a speculative angle, it is plausible that tumor-associated MCs may exhibit polarization into pro-tumorigenic (e.g., expressing high levels of pro-tumorigenic growth factors such as VEGF) and anti-tumorigenic subtypes (e.g., expressing high levels of tryptase), that is, in analogy with the polarization seen in the macrophage niche. However, further investigations are needed to evaluate this hypothesis.