Cancer stem cell subpopulations in metastatic melanoma to the brain express components of the renin-angiotensin system

Aim: There is increasing appreciation of the role of the renin-angiotensin system (RAS) in carcinogenesis with recent evidence showing expression of the RAS by cancer stem cells (CSCs) in different types of cancer. We have recently demonstrated the presence of three CSC subpopulations within metastatic melanoma (MM) to the brain: a Melan-A subpopulation and a Melan-A subpopulation within the tumor that express OCT4, SALL4, SOX2 and NANOG; and a pSTAT3 subpopulation localized to the CD34 endothelium of microvessels within the tumor. In this study we investigated the expression and localization of components of the RAS in relation to these CSCs in MM


INTRODUCTION
Metastatic melanoma (MM) to the brain occurs in 30% of melanoma patients and accounts for 5%-8% of all brain tumors [1] . For many years, the median survival of MM to the brain is 6-9 months following mainstay treatments including surgical excision, stereotactic radiosurgery and external beam radiotherapy [2][3][4][5] . Investigations into MM to the brain have identified real time preferential hematogenous dissemination of tumor cells [6] , with survival of these cells within brain capillaries, migration to breach the blood brain barrier, angiogenesis and intraparenchymal proliferation [7] .
Although immunotherapies that enhance T-lymphocyte activity such as Ipilimumab, anti-cytotoxic T-lymphocyte associated antigen-4 and anti-programmed cell death-1 inhibitors can improve the median survival, some cause significant toxicities [8,9] . Up to 50%-60% of MM patients harbor the BRAF V600 somatic mutation [10] . Vemurafenib, a BRAF inhibitor, increases the median survival to 16 months [11] but despite an initially good response, progression-free survival remains poor, at seven months [5] . Used together, kinase inhibitors, BRAF-targeted and MEK-targeted small molecule drugs relieve intracranial symptoms and induces remission but cessation due to toxicity causes rapid intracranial and/or extracranial disease progression [12] . The short period of remission after neuro-mimicry and individualized tumor targets [13] has been attributed to the initiation, propagation and differentiation of cancer stem cells (CSCs) [14] .
The CSC concept proposes that cancer is caused by CSCs that possess the ability for uncontrolled growth and propagation [15] . CSCs have been demonstrated in many types of cancer including breast carcinoma [16] , glioblastoma [17] and oral cavity squamous cell carcinoma (OCSCC) [18][19][20] . We have recently identified and characterized three CSC subpopulations within MM to the brain: a Melan-A + subpopulation and a Melan-Asubpopulations that express embryonic stem cell (ESC) markers OCT4, SALL4, SOX2 and NANOG within the tumor, and a pSTAT3 + subpopulation localized to the CD34 + endothelium of microvessels within the tumor [21] .
The RAS has been linked to carcinogenesis for some time [22] . In the classical RAS, angiotensinogen (ANG) is converted to angiotensin I (ATI) by renin. ATI is then converted to angiotensin II (ATII) by angiotensin converting enzyme (ACE). The RAS is implicated in central nervous system disorders such as Parkinson's disease and dementia with neuroprotective and neurotoxic features although its role in carcinogenesis is not discussed in a recent review [23] . GB cells in culture express ANG, pro-renin, ACE, ATII, angiotensin II receptor 1 (ATIIR1) and angiotensin II receptor 2 (ATIIR2); and renin inhibitors decrease DNA synthesis, induce apoptosis and reduce viable cell numbers [24] . Renin and its precursor pro-renin bind to pro-renin receptor (PRR) to activate the MAPK signaling cascades. PRR is associated with increased cell proliferation, decreased apoptosis and highly expressed in pancreatic ductal adenocarcinoma [25] . CSC subpopulations in GB [26] and OCSCC of different subsites [27][28][29] express components of RAS.
We here hypothesized expression of the RAS by CSCs in MM to the brain and investigated the expression and localization of components of the RAS: PRR, ACE, ATIIR1 and ATIIR2, in relation to the putative CSC subpopulations we have recently identified [21] , using immunohistochemical (IHC) staining, Western blotting (WB) and NanoString mRNA analysis.

Tissue samples
Samples of MM to the brain from ten patients were sourced from the Gillies McIndoe Research Institute Tissue Bank and used in this study that was approved by the Central Health and Disabilities Ethics Committee (Ref. 15CEN28) with written informed consent from all participants.

Histology and IHC staining
Hematoxylin and eosin (H&E) staining was performed on 4 μm thick formalin-fixed paraffin-embedded sections of MM to the brain from ten patients included in our previous study [21] , to confirm the presence of the tumor by an anatomical pathologist (HDB). Marque, Rocklin, CA, USA), diluted with Bond TM primary antibody diluent (cat# AR9352, Leica), was performed as previously described [30] .
To localize the components of the RAS in relation to the putative Melan-A + and Melan-A -OCT4 + / SALL4 + /SOX2 + /NANOG + CSC subpopulations within the tumor, and the pSTAT3 + subpopulation on the endothelium of microvessels within MM to the brain [21] , immunofluorescence (IF) IHC staining was performed on two representative samples of MM to the brain from the original cohort of ten patients included for DAB IHC staining. Components of the RAS were co-stained with ESC markers OCT4 or NANOG as the surrogate markers for the Melan-A + and the Melan-A -OCT4 + /SALL4 + /SOX2 + /NANOG + CSC subpopulations, or endothelial markers ERG or CD34 for the pSTAT3 + subpopulation on the endothelium of microvessels within the tumor. Antibodies used for PRR and ATIIR2 detection were VectaFluor Excel anti-rabbit 594 (ready-to-use; cat# VEDK-1594, Vector Laboratories) and Alexa Fluor anti-mouse 488 (1:500; cat#A21202, Life Technologies). Antibodies for ACE and ATIIR1 detection were VectaFluor Excel anti-mouse (ready-to-use; cat# VEDK2488, Vector Laboratories) and Alexa Fluor antirabbit 594 (1:500; cat# A21207, Life Technologies).
Human tissues used for positive controls were placenta for PRR, liver for ATIIR1, and kidney for ATIIR2 and ACE. Negative controls were used in secondary and tertiary antibody staining by omitting the primary antibodies on a randomly selected sample of MM to the brain.

Image analysis
DAB IHC stained-slides were viewed and images were captured with an Olympus BX53 light microscope fitted with an Olympus SC100 digital camera (Olympus, Tokyo, Japan), and processed with the CellSens 2.0 software (Olympus). IF IHC-stained images were captured using an Olympus FV1200 biological confocal laser-scanning microscope and processed with CellSens Dimension 1.11 software using 2D deconvolution algorithms (Olympus).

NanoString mRNA analysis
RNA extraction, as previously described [27] , was performed on four snap-frozen samples of MM to the brain from the same cohort of ten patients included in DAB IHC staining underwent. RNA was subjected to NanoString nCounter TM Gene Expression Assay (NanoString Technologies, Seattle, WA, USA) as completed by New Zealand Genomics Ltd (Dunedin, NZ). Probes for the genes encoding PRR (NM_005765.    Figure 1A]. The tumor tissue widely expressed PRR [ Figure 1B], ATIIR1 [ Figure 1C] and ATIIR2 [ Figure 1D], mostly in the cytoplasm.

Histology and DAB IHC staining
Human tissues used for positive controls: placenta for PRR, liver for ATIIR1, and kidney for ATIIR2 and ACE demonstrated the expected staining patterns [Supplementary Figure 1]. The omission of the primary antibody provided a control for the secondary antibody (data not shown).

IF IHC staining
IF IHC staining showed cytoplasmic expression of PRR [ Figure 2A, red] on the OCT4 + [ Figure 2A, green] CSCs within MM to the brain. ACE [ Figure 2B, green] was expressed only on the ERG + [ Figure 2B, red] endothelium of the microvessels. ATIIR1 [ Figure 2C, green] was expressed by the NANOG + [ Figure 2C, red] CSCs, while ATIIR2 [ Figure 2D, red] was expressed on the OCT4 + [ Figure 2D, green] CSCs in MM to the brain.
Supplementary Figure 2A-H displays individual IF IHC stains shown in Figure 2. Supplemental Figure 2I shows the appropriate negative control on a section of MM to the brain by the omission of the primary antibody.
WB PRR, at a molecular weight of 40 kDa, was present in all five samples of MM to the brain [ Figure 3A]. The negative control of mouse kidney also showed a comparatively faint band. ACE, at 190 kDa, was detected in all five samples, with multiple smaller molecular weight bands detected in all the samples [ Figure 3B], possibly due to degraded protein. ATIIR1 was seen in two of the five samples at 45 kDa, with a 28-kDa  Figure 3C], possibly due to degradation. ATIIR2 was seen in all five samples at 50 kDa [ Figure 3D].

NanoString mRNA analysis
NanoString mRNA analysis revealed very high levels of mRNA expression for PRR, moderate levels of ACE and low levels of ATIIR1 mRNA expression, while ATIIR2 was below detectable levels [ Figure 4].

DISCUSSION
We have recently identified three CSC subpopulations within MM to the brain: a Melan-A + subpopulation and a Melan-Asubpopulations that express the ESC markers OCT4, SALL4, SOX2 and NANOG, and a pSTAT3 + subpopulation localized to the CD34 + endothelium of the microvessels within the tumor [21] . This study demonstrated the expression of PRR, ATIIR1 and ATIIR2 by these CSCs subpopulations we have identified. The demonstration of widespread expression of these cell surface receptors by DAB IHC staining, may be due to their relatively abundant expression within the samples we have examined. Using Positive controls were mouse brain extract for PRR (A), mouse lung protein extract for ACE (B), mouse brain for ATIIR1 (C), PC3 cell lysate for ATIIR2 (D). Negative controls were mouse kidney for PRR (A), human tonsil for ACE (B), mouse kidney for ATIIR1 (C), and NTERA2 for ATIIR2 (D). PRR: pro-renin receptor; ATIIR1: angiotensin II receptor 1; ATIIR2: angiotensin II receptor 2; ACE: angiotensin converting enzyme OCT4 and NANOG as surrogate markers, IF IHC staining revealed localization of PRR, ATIIR1 and ATIIR2 to the OCT4 + /OCT4 + /SALL4 + /SOX2 + /NANOG + CSC subpopulations while ACE was localized the CSC subpopulation on ERG + endothelium of the microvessels which we have shown to express pSTAT3 previously. The expression of PRR, ACE, ATIIR1 and ATIIR2 was confirmed by WB. NanoString mRNA analysis showed high levels of expression of PRR, lower expression levels of ACE and ATIIR1. ATIIR2 was below detectable levels which may be due to mRNA degradation or its transient presence during transcription within the samples examined.
We have previously reported expression of PRR, ATIIR1 and ATIIR2 by the SOX2 + CSC subpopulation, and exclusive expression of ACE by the subpopulation on the endothelium of the microvessels, in human isocitrate dehydrogenase-wildtype GB [26] and OCSCC affecting different subsites [27][28][29] . mRNA expression of renin, ACE, ATIIR1 and ATIIR2 has been demonstrated on primary cultured human keratinocytes, melanocytes, dermal fibroblasts and dermal capillary endothelial cells, but ATIIR2 is not detected in melanocytes [31] . Cellular proliferation associated with ATI-treated and ATII-treated infantile hemangioma cell culture has shown to be diminished by ACE inhibitor ramipril and the ATIIR2 antagonist PD123319, and is enhanced by ATIIR2 agonist CGP42112 [32] , implying a role for the RAS peptides in stem cell proliferation.
Blockade of the RAS leads to inhibited growth of colorectal cancer liver metastases in the regenerating liver [33] and recent reports of increased overall survival of GB patients treated with angiotensin receptor blockers [34] , support the role of the RAS in cancer. It is exciting to speculate that CSCs in MM to the brain maybe a novel therapeutic target by modulating the RAS, although more work including a larger sample size, control tissue samples and in vitro and in vivo functional work is needed to determine the precise role of the RAS in this aggressive cancer.
In conclusion, in this study we have demonstrated the expression and localization of components of the components of the RAS: PRR, ACE, ATIIR1 and ATIIR2 in MM to the brain. PRR, ATIIR1 and ATIIR2 are localized to the CSC subpopulations within the tumor while ACE is expressed by the CSC subpopulation on the endothelium of the microvessels. Targeting the CSCs using modulators of the RAS may be a novel therapeutic approach for MM to the brain.