Adrenomedullin Secreted by Melanoma Cells Promotes Melanoma Tumor Growth through Angiogenesis and Lymphangiogenesis
, Isabelle Nanni 4, Nathalie Beaufils 4, Kamel Mabrouk 5, Jean-Jacques Grob 2 and L’Houcine Ouafik 1,4,*
Round 1
Reviewer 1 Report
The manuscript by Benyahia et al. describes the distribution of AM and its receptor components in melanoma tissues and cell lines, the overexpression of AM by a hypoxia mimetic, and the consequences of blocking the AM autocrine loop in vitro and in vivo. The study has been correctly performed and the results are properly presented. I only have a few comments:
1) In the introduction, I miss a description of the distribution of AM and its binding sites in human skin, the normal tissue from which melanomas arise (cite, for instance, doi: 10.1210/endo.138.12.5622).
2) Also, previous descriptions of AM in melanoma must be referenced in the introduction to put the study in perspective. Otherwise readers may believe this is the first study on melanoma and AM (cite, for instance, doi: 10.1158/1078-0432.CCR-11-1354; doi: 10.1093/cvr/cvw166; doi: 10.1210/endo.138.12.5622).
3) Authors claim to have studied the impact of hypoxia on the expression of AM, but they have used desferrioxamine (please, check the spelling) mesylate. This is a “hypoxia mimetic” and must be described as such, since the physiological impact is similar but not equal to reducing O2 tension. To demonstrate the effects of hypoxia, a real hypoxia chamber must be used. Otherwise, the authors must say, in the text and figure legends, they demonstrated the effects of a hypoxia mimetic (not of hypoxia).
4) Authors say they performed preabsorption of the antibodies with synthetic peptides as specificity controls (lines 108-110). This is the most convincing approach to demonstrate antibody specificity and it should be shown, perhaps as supplementary material.
5) When describing the methods for immunostaining (lines 129-132), authors say they “examined by immunofluorescence…. using a Vectastain… kit”. I guess this is a carry out mistake. Please, correct.
6) In the methods for the cell proliferation assay (lines 134-139), the number of repeats should be indicated.
7) When describing the methods for migration and invasion (lines 141-143), authors just reference some of their papers. It should be good to briefly expand this description so that readers can understand the methodology without being forced to go to other papers.
8) In the section of the matrigel plugs (lines 145-157), a description of the methodology used to quantify the positive structures (how fields were selected, how many fields per animal, etc.) must be added.
9) In the statistical analysis section (lines 185-189), authors say “differences were considered statistically significant when *, p < 0.05; **, p < 0.01; ***, p < 0.001”. Actually, the criterion for establishing statistical significance must be unique. So, I guess that sentence must be “differences were considered statistically significant when p < 0.05”.
10) In Results, the sentence “Previous research demonstrated similar patterns of AM localization in serial sections among specimens of kidney cancer, colorectal cancer, prostate cancer, and epithelial mesothelioma [12, 20, 27, 28]” should be moved to the discussion.
11) The negative results of the hypoxia mimetic on CLR, RAMP2, and RAMP3 (lines 232-233) could be presented as supplementary materials.
12) Melanoma is misspelled several times in the legend of Figure 2.
13) In the discussion, the sentence “this study reveal that AM and the AM receptors AM1 and AM2 are present in melanoma tissue, which supports the view that AM may play a role as an autocrine/paracrine growth factor in melanoma” (lines 442-444) could be misleading, almost indicating this is the first time this was shown. Authors must say something like “this study confirms previous reports….”.
Author Response
Reviewer 1.
We thank the reviewer for the effort and propositions to improve the overall quality of the study.
1) It has been reported that AM and its receptors are present in all the epithelial cells in human skin, the normal tissue from which melanomas arise (Martinez et al, ref 12). The expression has been reported in keratinocytes of the epidermis and their follicles, as well as cells of the glands and secretory ducts (Martinez et al, ref 12). In addition, AM and its receptors were found in skin tumors of different histologies (Martinez et al, 1997). Previous work has reported that a major source of AM in melanoma is tumor-associated macrophages (TAM) (Chen et al, ref 13). This paragraph is now added to the introduction (please see lines 80-84).
2) Now the previous description of AM in melanoma is reported into the introduction (lines 84-85).
3) To the paragraph 2.2 we added “The induction of hypoxia was achieved by using the hypoxia mimetic 260 mM desferrioxamine mesylate (DFX) (Sigma, Paris, France). After cells had grown to confluence of 70%, the medium was changed, and cells were incubated with new medium containing 260 mM DFX for 24 and 48 hours” for more precision (lines 127-130).
4) Immunohistochemistry of melanoma tissues with antiserum preincubated with human synthetic AM, CLR, RAMP2 and RAMP3 (50mM) is reported as “Supplementary Figure 1” in the text (lines 249-250). The supplementary Figure 1 with the legend “Immunohistochemistry of melanoma tissues with antiserum preincubated with human synthetic AM, CLR, RAMP2, and RAMP3 at 50 mM each is shown. No staining for AM, CLR, RAMP2, and RAMP3 can be observed suggesting the specificity of the staining reported in Figure 1.” are reported in the supplementary data.
5) For more precision the previous paragraph “2.4. Immunostaining of the melanoma cells” was replaced by “The fluorescence microscopy analyses of AM, CLR, RAMP2 and RAMP3 were performed on the A375, SK-MEL-28, and MeWo cells. Briefly, after cells were fixed in 4% paraformaldehyde and permeabilized with 0.1% Triton X-100, cells were incubated with polyclonal antibodies against AM (1:2000), CLR (1:2000), RAMP2 (1:1500), and RAMP3 (1:1000) overnight at 4°C, then washed and incubated with Alexa Fluor-conjugated antibodies (1:300; Vector Laboratories) for 45 min at room temperature (RT). After washing, the samples were mounted in VETASHIELD (Vector Laboratories) and analyzed by fluorescence microscopy” (lines 138-145).
6) More precision is reported to the cell proliferation assay “After six days treatment of six wells treated with AM, aAMR, or purified IgG, the effects were examined by 3-(4,5-dimethylthiazol-2, 5-diphenyltetrazolium bromide) (MTT) assay. The Bio-Tek microplate was used to determine the change in the number of viable cells from dye reduction measured by absorbance at 570 nm. The values represent the mean ± SD of five independent experiments with six wells each” (lines 152-157).
7) The whole protocol for the migration and chemoinvasion of the cells is now fully described in the Materials and Methods section 2.6. “Briefly, for chemoinvasion, the filter was coated with a layer of Matrigel (0.5 mg/ml, Becton Dickinson, Paris, France). A375, SK-MEL-28, and MeWo cells were harvested by trypsinization, collected by centrifugation and resuspended in DMEM (A375 and SK-MEL-28 cells) and MEM (MeWo cells) containing 0.5% bovine serum albumin (BSA) and soybean trypsin inhibitor, washed 3 times with the medium containing BSA as above and suspended in the same medium at a concentration of 2x105 cells/ml for A375 and SK-MEL-28 cells, and 1x106 cells/ml for MeWo cells. The bone marrow derived cells (BMDCs) were prepared in DMEM medium at 5x106 cells/ml. 100 ml of this suspension was added to the upper compartment (24-multiwell chemotaxis Boyden microchamber). The lower compartment of the chamber was filled with chemoattractant AM diluted in DMEM (A375, SK-MEL-28 cells and BMDCs) or in MEM (MeWo cells) (n=4 in triplicate). Where indicated, cells were preincubated for 30 min at 37°C with aAMRs, or with preimmune serum (affinity purified IgG). After incubation at 37°C for 4 hr, cells that had not migrated from the top of the filter were scraped away with a cotton applicator. The filter was fixed for 30 min at RT with 3.7% paraformaldehyde, washed two times with PBS, and stained with 4’,6’-diamidino-2-phenylindole (DAPI), and the number of cells that migrated to the lower surface of each membrane was counted at 50X magnification using a microscope. The control well was filled with DMEM or MEM containing 2% FBS. Data are expressed as the number of migrated cells in 10 high-power fields, and the values represent the mean ± SD of four independent experiments, each performed in triplicate” (lines 162-182).
8) CD31- and LYVE-1-positive cells are shown; they were analyzed based on 5 magnification fields (400 x) per section. Immunohistochemical staining of the endothelial cell surface marker CD31 was used to determine microvessel density. The blood vessels were counted randomly from non-necrotic areas in each Matrigel section in an x200 microscope field, on CD31-stained Matrigel sections. Quantitative assessment of the density of cells that stained positive for CD31, or LYVE-1 was conducted for the entire surface of the corresponding slides using CALOPIX software (n=6 per animal, total number of animals (n=5)). MBF_Image J1.43U software was used for the analysis. This paragraph was added to the section 2.7 in the Materials and Methods for more clarification of the analysis process that concern in vivo Matrigel plugs assays (lines 197-204).
9) For all tests, differences were considered statistically significant when p <0.05 is fixed in the text (line 235).
10) The sentence “Previous research demonstrated similar patterns of AM localization in serial sections among specimens of kidney cancer, colorectal cancer, prostate cancer, and epithelial mesothelioma“ in results has moved to the discussion (lines 472-475).
11) Legend of supplementary Figure 2: Expression of AMR (CLR, RAMP2, and RAMP3) in melanoma cell lines. Total RNA (1mg, DNA free) prepared from MeWo, SK-MEL-28 and A375 cells were reverse transcribed into cDNA under normoxia and hypoxia conditions. Relative human CLR (A), RAMP2 (B), RAMP3 (C) and GAPDH mRNAs levels were amplified, detected, and quantified in real time by using an LC480 polymerase chain reaction (PCR) system (Roche Diagnostics, Meylan, France) as described previously (Berenguer et al, ref 27). No significant differences between cells treated with hypoxia mimetic DFX and untreated control cells in CLR, RAMP2, and RAMP3 expression. Each experiment is representative of five independent experiments. This paragraph is now added to the supplementary data section with the Figure. The supplementary Figure 2 is mentioned in the text (line 177).
12) The misspelling of “Melanoma” in the legend of Figure 2 is now fixed.
13) We agree with the reviewer that using the word “reveal” is misleading in sense that can be understood as is the first time that expression of AM in melanomas is demonstrated. This was never our goal neither our intention and now we use the word “confirms” instead of “reveal” as proposed by the reviewer (line 490).
Reviewer 2 Report
The authors vigoursly investigated the effect of adrenomedullin and its receptor on melanoma cell growth both in vitro and in vivo experiments. The data presented are sound and support the conclusions. Authors discuss the different findings in adrenomedullin effect on lympangiogenesis partially and their claims are somewhat acceptable. However I have concern if blockade of adrenomedullin system can prevent distant metastasis of melanoma cells. Melanoma cells prone to metastatize via lymph node and lymph vessels. Tanaka M and her colleagues reported that knocking out RAMP2 system can increase melanoma cell metastasis to the lung (PMID: 27307317). The authors should cite this article and discuss about the discrepant findings from your research work.
Author Response
Reviewer 2.
We thank the reviewer for the effort and propositions to improve the discussion of the manuscript.
The finding reported by Tanaka et al (2016, ref 52 in the text) indicates that the deletion of RAMP2 from endothelial cells suppresses growth of locally transplanted B16F10 melanoma cells. Spontaneous lung metastasis was analyzed using B16BL6 established from B16F10 melanoma cells showed that the incidence of metastasis and the number of metastatic lesions were higher in drug-inducible endothelial cell-specific RAMP2 knockout mouse (DI-E-RAMP2-/-), which could be due to the endothelial cells that were deformed and facilitating the infiltration of the inflammatory cells of the vessel walls. The inflammatory cells could express the chemotactic factors S100A8/9 and SSA3, which attract tumor cells and mediate the formation of a pre-metastatic niche (Tanaka et al, ref 52), contrary to the present study where no metastasis could be found in all the animals treated with aAMR, meanwhile three animals in the control group do show metastasis in the liver. Different point of view could explain this discrepancy. First, endothelial cells participating in neovessel assembly are in a dynamic state during tumor angiogenesis and are thus not firmly attached to the extracellular matrix or to peri-endothelial cells such as pericytes. AM blockade using aAMR exerts an anti-vascular and anti-angiogenic effect presumably by taking advantage of the relative instability of tumor vasculature and its supporting structures, thereby inducing endothelial cells death (Khalfaoui-Bendriss et al., ref 53) and a collapse and regression of tumor vascular and lymphatic neovessels, the routes used by tumor cells to metastasis. Second, contrary to the DI-E-RAMP2-/- that could affect most of the endothelial cells in the animal, as in the lung for example, the AM system blockade using aAMR did not disrupt the normal vasculature of different organs in animal bearing tumors such as in kidney (Khalfaoui-Bendriss et al., ref 53), prostate (Berenguer et al., ref 22), lung and liver (unpublished data) probably because of the AM system which, is expressed at the very low level in non-angiogenic endothelium that couldn’t be detected by the aAMR in vivo. Intravenous injection of fluorescent aAM into animal bearing U87 xenograft in the brain localized specifically at the tumor site 24 hr later without any fixation in the rest of the body (unpublished data). Third, it is also possible that aAMR treatment could prevent any growth of metastatic niches by inhibiting vascular neoangiogeneis and lymphangiogenesis to impair any growth of secondary tumor site. This paragraph is now added to the discussion (lines 575-603).
Round 2
Reviewer 1 Report
I commend the authors for their effort in answering all the reviewers´ comments.
Nevertheless, when answering my comment number 5, about the methods for tissue staining, they explain now very well their technique for immunofluorescence but have completely deleted all explanation for immunohistochemistry. Since this technique is required to understand several important Figures, I suggest the authors add a paragraph explaining this methodology.
Author Response
Author Response File: 
Author Response.docx
Reviewer 2 Report
The authors revised properly.
Author Response
We thank the reviewer for the thoughtful review of our manuscript entitled “Adrenomedullin secreted by melanoma cells promotes melanoma tumor growth through angiogenesis and lymphangiogenesis”.
