Heterogeneity of tyrosine-based melanin anabolism regulates pulmonary and cerebral organotropic colonization microenvironment of melanoma cells

Abstract

Background: Dietary tyrosine regulating melanoma progression has been well-recognized. However, whether tyrosine-based melanin anabolism contributes to pulmonary and cerebral organotropic colonization of melanoma remains elusive. Furthermore, approaches based on targeting tyrosinase activity to inhibiting multi-organ metastasis of melanoma cells need to be designed and validated. Methods: Patients derived melanoma cells and mouse B16 melanoma cells with different pigmentation were employed in this investigation. Tyrosine content dynamics in tumors and multiple organs during the melanoma progression was monitored, and tyrosine-based melanin synthesis of melanoma cells derived from multi-organ was determined. Additionally, we also adopted RNA-seq, flow cytometry, real-time PCR and composite metastasis mouse model to analyze organotropic colonization and to validate designed therapeutic strategies. Results: B16 melanoma cells with high activity of tyrosinase and sensitivity of tyrosine utilization for melanin synthesis (Tyr-H cells) easily colonized in the lung, while B16 melanoma cells lacking above characteristics (Tyr-L cells) exhibited potent proliferation in the brain. Mechanistically, Tyr-H cells recruited and trained neutrophils and macrophages to establish pulmonary metastatic niche dependent on highly secreted CXCL1 and CXCL2 and an excessive melanosome accumulation-induced cell death. Tyr-L cells enhanced PD-L1 expression in tumor-infiltrated macrophages when they are progressing in the brain. Accordingly, intervention of tyrosinase activity (2-Ethoxybenzamide or hydroquinone) in combination with inhibitors of phagocytosis (GSK343) or chemotaxis (SB225002) suppressed organotropic colonization and significantly improved the survival of melanoma- bearing mice treated with immune checkpoint blockade (PD1 antibody). Conclusions: The heterogeneity of melanoma cells in utilization of tyrosine is associated with organotropic colonization, providing the basis for developing new strategies to combat melanoma.

Keywords: Melanoma; immune checkpoint; melanin anabolism; organotropic colonization; tumor heterogeneity.

Figure 1

The heterogeneity of tyrosine-based melanin synthesis affects organotropic colonization of melanoma cells. A. Microscope photos of melanoma cells derived from patients. Left photo (255776) showed amelanotic melanoma cells and right photo (238818) showed pigment melanoma cells. Scale bar is 100 μm. B. Human melanoma cell line A375 cells were mixed with 255776 cells or 238818 cells and co-transplanted into NOD-SCID mice by i.v. injection. The pulmonary macrometastases were counted on day 30. Injection of 255776 cells, 238818 cells or A375 alone as the control. C. Tyr gene expression level measured by RNA sequencing technology. D-E. Relative cell viability was evaluated by MTS assay. Tyr-H cells and Tyr-L cells were respectively cultured in the standard medium for 24 and 48 h (D), or in the conditional medium with different concentration of tyrosine (0, 1.25, 2.5, 5, 10 and 20 times higher than the standard concentration, 0 T, 1.25 T, 2.5 T, 10 T, 20 T) for 24 h (E). F. Representative microscope photos of Tyr-L cell clone (left) and Tyr-H cell clone (right) under the 10 T tyrosine treatment for 24 h. G. Absolute intracellular accumulation of melanin in the Tyr-H cells. Tyr-H cells were treated with 5 T and 10 T tyrosine for 30 h. H-K. Schematic showing a distant metastasis model of melanoma (H). Representative results of the lung with metastasis lesions (I), percent survival (J), and mass of tumor in muscle (K). L-M. Tyr-L cells and Tyr-H cells were injected intravenously. Representative photos of the lung with metastasis lesions (L), and percent survival (M). N-O. Photos and mass of the melanoma focus localized in brain from experimental cerebral metastasis model. All data given as mean ± SEM., n ≥ 3. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns, not significant. All results shown are representative of three independent experiments.

Figure 2

Malignant behaviors of tyrosine-resistant melanoma cells in vitro and in vivo. A-B. Relative cell viability was measured by MTS assay. The same number of tyrosine trained cells were cultured in the fine-tuned medium with different tyrosine addition for 48 h (A), or in the standard medium for 6 days (B). C-D. Representative photos and counts of the lung with metastasis lesions from experimental pulmonary metastasis model on day 15. Normal (0 T-), 2 T- , 5 T- and 10 T- B16 cells were injected intravenously. E-F. Photos and mass of the melanoma focus in brain from experimental cerebral metastasis model. G-H. Schematic showing Tyr-L cells enrichment procedure in vivo (G). Photos of B16 subcutaneous tumor. Tumor bearing mice were treated with intratumor injection of tyrosine (H). I-J. Photos and counts of the lung with metastasis lesions from experimental pulmonary metastasis model. Normal B16 cells (PBS-B16) and tyrosine-enriched B16 (Tyr-B16) cells were injected intravenously. K-L. Photos and mass of the brain with metastasis lesions from experimental cerebral metastasis model. All data given as mean ± SEM., n ≥ 3. *P < 0.05, **P < 0.01, ***P < 0.001, ns, not significant. All results shown are representative of three independent experiments.

Figure 3

Increased adhesiveness of Tyr-H cells. A-C. Photos and counts of cell colony. Same number of Tyr-L cells and Tyr-H cells were seeded in the 6-well plate for 60 s, 120 s, 180 s or 300 s, and then discard the medium containing unattached cells. Cell colonies were counted after 10 days culture (A-B), or immediately counted and calculated the remained cells (C). D. Based on RNA sequencing, differentially expressed genes associated with cell adhesion between Tyr-L cells and Tyr-H cells were enriched by KEGG analysis and presented by heatmap. E-J. Based on RNA sequencing, expression of key genes associated with cell adhesion were showed. K. The ability of Tyr-L cells and Tyr-H cells to reside in the pulmonary tissues. Tyr-L cells and Tyr-H cells were labeled by tdTomato. All data given as mean ± SEM., n ≥ 3. **P < 0.01, ***P < 0.001, ns, not significant. All results shown are representative of three independent experiments.

Figure 4

Tyr-H cells recruit neutrophils and macrophages to construct pulmonary metastatic niche. A-C. Gene expression profiles of Tyr-L cells and Tyr-H cells were analyzed by RNA sequencing. Differentially expressed genes participated signaling pathways were enriched by KEGG analysis (A). Volcano plot showing the differentially expressed genes (B). Cytokine-cytokine receptor interaction involved genes were represented by heatmap (C). D-E. Real-time quantitative PCR was used to analyze expression of Cxcl1 gene and Cxcl2 gene. Brain-B16 cells and Lung-B16 cells were isolated from experimental metastasis lesions respectively. F-G. Percentage of pulmonary tissues infiltrated CD11b⁺Ly6G (1A8)⁺ neutrophils, CD11b^-F4/80⁺ alveolar macrophages (AMs) and CD11b⁺F4/80⁺ interstitial macrophages (IMs) gated in CD45⁺ cells (F). Absolute number of above cells in the lung was calculated (G). Lung tissues were collected on day 5 after the experimental pulmonary metastasis model was built. H-J, Neutrophils were depleted by Ly6G antibody in vivo before 48 h of B16 cell transplantation. Depletion efficiency and infiltration were detected by flow cytometry (H). Photos and counts of macrometastases in the lung from above mice (I-J). K. Representative photos of the metastasis lesions in the lung from experimental pulmonary metastasis mice. Normal B16 cells were intravenously injected into CD11b-DTR mice with or without diphtheria toxin treatment. L-M. Representative photos of the lung with metastasis lesions from experimental pulmonary metastasis model (L). Normal B16 cells were injected intravenously, with the treatment of PBS, control liposomes, and clodronate liposomes respectively. Percent survival was calculated (M). All data given as mean ± SEM., n ≥ 3. *P < 0.05, ***P < 0.001, ns, not significant. All results shown are representative of three independent experiments.

Figure 5

Tyrosine-induced melanosome -dependent cell death supported the acquisition of tumor-promoting phenotype by macrophages. A. Relative cell viability was evaluated by MTS assay. The same number of B16 cells were cultured in the conditional medium with different tyrosine addition for 24 h. B-C. Cell apoptosis was examined through Annexin V / PI staining. B16 cells were treated with different concentration of tyrosine or ch282-5 (Bcl-2 inhibitor, positive control). D-E. Cell cycle was analyzed by PI staining. F-G. Mitochondria was quantized by Mito-tracker staining. H-J. Mitochondrial membrane potential (MMP) dependent TMRE signaling was detected by flow cytometry (H-I). MMP was reflected by the MFI ratio of TMRE to Mito-tracker (J). K-L. Lysosome were quantized by Lyso-tracker staining. M. Representative photos of Tyr-L cells and Tyr-H cells with 10 T tyrosine for 24 h. N-O. Representative photos of B16 cells. B16 cells with 10 T tyrosine combining with or without deoxyarbutin or hydroquinone for 48 h. The percentage of pigment cells were calculated. P. Peritoneal macrophages were co-cultured with the EMCDRs (labeled with ZsGreen) in different ratios. Consuming efficiency were identified by flow cytometry. Q-R. Photos of the lung with metastasis lesions from experimental pulmonary metastasis model (Q). Normal B16 cells were injected intravenously, and mice were treated with PBS, macrophages, or EMCDRs-educated macrophages respectively. Percent survival was calculated (R). Mφ, macrophages; EMCDRs-Mφ, EMCDRs educated macrophages. All data given as mean ± SEM., n ≥ 3. *P < 0.05, **P < 0.01, ***P < 0.001, ns, not significant. All results shown are representative of three independent experiments.

Figure 6

Tyr-L cells enhance PD-L1 expression in the cerebral melanoma infiltrated macrophages. A. Comparing to Tyr-H cells, interaction of high expression of (left) and low expression of (right) genes (red nodes) and their associated metabolites (green nodes) in Tyr-L cells were analyzed by the web of OmicsNet. B-C. mRNA level of core genes enriched in the above analysis. D-E. B16 melanoma cells were transplanted into the mice by stereotactic injection. PD-L1 expression of macrophages in the naïve mouse brain, para-tumor tissue and tumor were analyzed by flow cytometry (D), and the MFI of PD-L1 was calculated (E). F. Bone marrow derived macrophages (BMDMs) were co-cultured with Tyr-L cells in the mixture co-culture system or transwell co-culture system for 48 h. PD-L1 expression were detected by flow cytometry. G-H. Splenocytes were co-cultured with Tyr-L cells educated BMDMs. Relative splenocyte viability were measured by MTS assay. All data given as mean ± SEM., n ≥ 3. ***P < 0.001, ****P < 0.0001, ns, not significant. All results shown are representative of three independent experiments.

Figure 7

Targeting tyrosine utilization augments the therapeutic effects of PD1 monoclonal antibody. A. The schedule showing composite melanoma model and treatment. B-F. Composite melanoma mice were treated with E.T.G. (2-Ethoxybenzamide, tyrosine, plus GSK343), PD1 monoclonal antibody (mAb), and E.T.G. plus PD1 mAb, PBS treatment as a control. Percent survival (B), in situ tumor mass (C), pulmonary macrometastases (D), cerebral colonization (E) and cerebral tumor weight (F) were observed. G-K. Composite melanoma mice were treated with hydroquinone plus CXCR2 inhibitor SB225002 (Hy + CXCR2i), PD1 mAb, and Hy + CXCR2i plus PD1 mAb, PBS treatment as a control. The indicators as panel B-F were observed. All data given as mean ± SEM., n ≥ 3. *P < 0.05, ***P < 0.001, ns, not significant. All results shown are representative of three independent experiments.