Targeting the Endothelin-1 Receptors Curtails Tumor Growth and Angiogenesis in Multiple Myeloma

The endothelin-1 (ET-1) receptors were recently found to mediate pro-survival functions in multiple myeloma (MM) cells in response to autocrine ET-1. This study investigated the effectiveness of macitentan, a dual ET-1 receptor antagonist, in MM treatment, and the mechanisms underlying its activities. Macitentan affected significantly MM cell (RPMI-8226, U266, KMS-12-PE) survival and pro-angiogenic cytokine release by down-modulating ET-1-activated MAPK/ERK and HIF-1α pathways, respectively. HIF-1α silencing abrogated the ET-1 mediated induction of genes encoding for pro-angiogenic cytokines such as VEGF-A, IL-8, Adrenomedullin, and ET-1 itself. Upon exposure to macitentan, MM cells cultured in the presence of the hypoxia-mimetic agent CoCl2, exogenous ET-1, or CoCl2 plus ET-1, down-regulated HIF-1α and the transcription and release of downstream pro-angiogenic cytokines. Consistently, macitentan limited significantly the basal pro-angiogenic activity of RPMI-8226 cells in chorioallantoic membrane assay. In xenograft mouse models, established by injecting NOG mice either via intra-caudal vein with U266 or subcutaneously with RPMI-8226 cells, macitentan reduced effectively the number of MM cells infiltrating bone marrow, and the size and microvascular density of subcutaneous MM tumors. ET-1 receptors targeting by macitentan represents an effective anti-proliferative and anti-angiogenic therapeutic approach in preclinical settings of MM.


INTRODUCTION
Multiple myeloma (MM) is a B cell malignancy arising from post-germinative B cells or plasma cell (PC) precursors and characterized by the accumulation of malignant PCs in the bone marrow (BM), the onset of monoclonal gammopathy, and eventually a significant morbidity due to organ dysfunction (1,2). In spite of recent developments in novel therapies, MM remains an incurable disease, accounting for about 10% of all hematologic malignancies (3). Although various genomic aberrations have been shown to provide PCs with the ability to proliferate in an uncontrolled manner, increasing evidence suggests critical roles for surface receptors with restricted expression in malignant PCs and for BM microenvironment in mediating MM survival, proliferation and resistance to therapy (4).
Endothelin-1 (ET-1), originally isolated from endothelial cells (ECs), as well as its receptors A (ET A R) and B (ET B R), referred to as the "ET-1 axis", exert key physiological functions in the human cardiovascular system and in other normal tissues (5,6). In the last two decades the ET-1 axis has been also implicated in the development of an increasing number of tumors via autocrine and/or paracrine activation of pathways involved in cell proliferation, migration, invasion, epithelial-mesenchymal transition, osteogenesis and angiogenesis (7).
We have previously demonstrated that the ET-1 axis supports MM cell viability through autocrine activation, also anticipating a possible paracrine triggering (8). Accordingly, we found that both primary malignant PCs and MM cell lines were constitutively ET A R positive, expressing ET B R in roughly half of cases on an epigenetic dysregulation basis. MM cells and BM microenvironment, in particular ECs, were also found to express and release ET-1 (8). Importantly, in vitro experiments consisting in ET A R or ET B R blockade with selective antagonists or with the dual receptor antagonist bosentan (5) resulted in a significantly decreased viability of MM cell lines (8). More recently, preliminary in vitro results have demonstrated that the dual receptor antagonist macitentan also decreases MM cell viability (9). Interestingly, both bosentan and macitentan are orally active drugs already licensed in clinics for the treatment of pulmonary arterial hypertension (PAH) (10). Macitentan, designed to improve the efficacy and tolerability of bosentan, presents as main pharmacokinetic advantage a longer and sustained ET A R and ET B R occupancy that allows both an effective drug activity in the presence of elevated ET-1 levels and a one-daily dosing (10). At present, the mechanisms underlying the inhibitory activity of ET-1 receptor (ET-1R) antagonists remain to be established in MM. In this tumor context, angiogenesis is a constant hallmark of BM microenvironment during MM progression (11). Several studies have demonstrated that the hypoxia inducible factor-1alpha (HIF-1a) stabilization, favored by the hypoxic MM BM niche, induces the transcription of a few pro-angiogenic factors, including Vascular Endothelial Growth Factor A (VEGF-A), both in tumor and stromal cells (11). In solid tumors HIF-1a may be controlled also in a hypoxia-independent manner by a number of molecules, including ET-1 (12)(13)(14)(15). Importantly, ET-1 stimulates hypoxic pathways promoting further ET-1 expression, thus maintaining angiogenic and pro-tumoral responses in a positive feedback system (16). Despite the available evidence demonstrating that HIF-1a may be regulated in a hypoxia-independent fashion even in MM (17), the regulatory effects of ET-1 on HIF-1a expression and related pro-angiogenic activities in MM is currently unknown. However, a recent network-based analysis of BM samples from patients affected by monoclonal gammopathy of undetermined significance (MGUS) and MM has highlighted the significant enrichment of ET-1 in the HIF-1a signaling pathway of primary MGUS and MM PCs (18).
Based on the aforementioned premises, aims of our work were: i) to evaluate in vitro the mechanisms underlying the antiproliferative and pro-apoptotic effects exerted by macitentan in MM cells, ii) to establish in vitro the eventual anti-angiogenic role of macitentan and the underlying mechanisms, and ii) to test the effectiveness of macitentan as therapeutic agent against MM in xenograft mouse models and in a chick chorioallantoic membrane (CAM) assay.

In Vivo Studies
NOD/Shi-scid/IL-2Rgnull (NOG) mice were purchased from Taconic (Germantown, NY, USA). Four groups of five four-to six-week-old female mice were subcutaneously injected with 5 × 10 6 RPMI 8226 cells. At day +7 post-injection, mice were daily treated with macitentan (30 mg/kg) (21) or vehicle (Methocell 0.05% and TWEEN 80 0,05%) per os for four weeks. At the end of treatment mice were sacrificed and tumor masses were measured with a caliper (tumor mass in mm 3 was calculated using the formula p/6 larger diameter x smaller diameter 2 ) and analyzed by immunohistochemistry (IHC). Simultaneously, four groups of five 4-6-week-old mice animals totally irradiated (1.2 Gy, 137 Cesium source) were injected into the tail vein with 2 × 10 6 U266 cells. At day 7 post-injection, mice were daily treated with macitentan (30 mg/kg) or vehicle (Methocell 0,05% and TWEEN 80 0,05%) per os. Following a 7day treatment, mice were sacrificed, and MM BM burden was evaluated as percentage of viable human CD45 + CD138 + cells over total viable human CD45+ TOPRO3 negative cells. The day of beginning and the length of macitentan treatment were established based on preliminary experiments. RPMI-8226 and U266 cell lines were chosen on the basis of the literature data for subcutaneous and intra-tail injection of MM cell lines, respectively (22).

Chorioallantoic Membrane Assay
CAM assay was established as previously described (23). Briefly, alginate beads were prepared by dissolving 6% (w/v) alginic acid sodium salt (Sigma Aldrich) in sterile milliQ H 2 O overnight at 4°C. Then, 2 µL of the solution were added with 2 µL of PBS containing 1.5 x10 4 RPMI-8226 cells, in the presence of either 0.1% DMSO (vehicle) or macitentan to a final concentration equal to 20 or 40 pmol/pellet. The gelation reaction was obtained by exposing the pellets to 0.1 M CaCl 2 . Alginate beads were then placed on the top of the embryo CAM of fertilized White Leghorn chicken eggs at day 11 of incubation. After 72 h, newly formed microvessels converging towards the implant were counted under a stereomicroscope (STEMI-SR, Zeiss, Germany) at × 5 magnification.

Statistical Analysis
Mice sample size was chosen as previously described (24). ANOVA test was used to analyze the differences among multiple means. Expression of treatment with respect to controls was analyzed using Student's t test. Statistical analyses were performed by GraphPad PRISM ® version 5-0c.

Macitentan Exerts Anti-Proliferative Effects in Multiple Myeloma Cells by Inhibiting the MAPK/ERK and Pro-Survival Pathways
Preliminary results had shown that macitentan exerts significant anti-survival and anti-proliferative activities towards RPMI-8226, U266, and KMS-12-PE cells (9) (Supplementary Figure  1), the first two cell lines expressing both ET A R and ET B R on the membrane, the third one (KMS-12-PE cells) ET A R only. To evaluate the mechanisms underlying the anti-proliferative action exerted by macitentan, we examined the MAPK/ERK signaling pathway, which is known to be involved in ET-1-mediated autocrine and paracrine pro-tumorigenic functions (7). As shown by the Western blots in Figure 1A, basal levels of p-erk1/2 were found in RPMI-8226, U266, and KMS-12-PE cells under resting conditions. Upon 48 h culture with ET-1, MM cell lines further increased p-erk1/2 expression. By densitometric analysis, macitentan significantly reduced p-erk1/2 levels both in resting and ET-1-stimulated cell lines ( Figure 1A), thus showing its capacity to antagonize either the autocrine (8) and paracrine activation of ET-1 axis in MM cells. We then observed that the expression of cleaved caspase-3 increased in both resting and ET-1-stimulated cells treated with macitentan, thus better characterizing the pathways downstream of ET-1Rs that may be involved in the resistance to apoptosis ( Figure 1B). Taken together, our findings confirm the pro-survival activity of the ET-1 axis and its role as therapeutic target in MM. The capacity of macitentan to interfere with major signaling pathways elicited by both autocrine and exogenous ET-1 is also highlighted.

Macitentan Inhibits Tumor Growth in Multiple Myeloma Xenograft Mouse Models
We next investigated whether macitentan may affect MM cell viability and growth in vivo. In order to address this question, we analyzed the effects of treatment with macitentan in NOG mice injected with MM cells either via intra-caudal vein or subcutaneously.
A first group of 20 mice was injected intra-venously with 2x10 6 U266 cells. Starting from the engraftment (day +7 from injection) and for the subsequent 7 days, mice were treated orally either with macitentan at a dosage of 30 mg/kg (n=10) or with vehicle only (n=10), and then sacrificed. Notably, following injection of U266 cells, control mice displayed hair alterations, weight loss and limbs paralysis as for disease progression (not shown). BM cells, collected by tibial flushing from macitentantreated and untreated mice, were then analyzed for viability by flow cytometry after anti-CD45 and anti-CD138 staining (Supplementary Figure 2). The viable (i.e., TOPRO-3 negative) CD138 + cell fraction was significantly reduced in BM samples from macitentan-treated mice as compared to controls (11 vs 44%, p=0.001) (Figure 2A), thus showing that macitentan can reach effective therapeutic concentrations in the tissue primarily affected by MM cell expansion.
A second group of 20 mice was injected subcutaneously with 5x10 6 RPMI-8226 cells and monitored for subcutaneous tumor mass development. Starting from day +7 after injection, mice were treated with a macitentan daily oral dose of 30 mg/kg, 5 days a week, for 4 weeks (n=10) or with vehicle only (n=10), and then sacrificed. At the end of treatment all animals developed well-delimited tumors in the site of injection that were removed and measured. As shown in Figure 2B, tumor masses from macitentan-treated mice had a median size significantly reduced as compared to controls (125 mm 3 vs 370 mm 3 , p=0.003). Consistently with previous in vitro findings, higher expression of cleaved caspase-3 was observed in histological specimens from macitentan-treated mice as compared to controls ( Figure 2C). Interestingly, PECAM1 staining revealed a significantly diminished microvascular density (MVD) in RPMI-8226 tumor masses from macitentan-treated mice, as compared to controls ( Figure 2D), thus suggesting that in MM macitentan could also exert an anti-angiogenic activity.
Overall, our in vivo findings extend previous in vitro results regarding the anti-proliferative effects of ET-1Rs blockade in MM cells and support the role for macitentan as an effective therapeutic agent against MM. In addition, the significantly reduced MVD observed in tumors from treated mice as compared to controls anticipates a possible anti-angiogenic activity of macitentan in MM.

Macitentan Downregulates Hypoxia Inducible Factor-1alpha, a Crucial Mediator of Multiple Myeloma Angiogenesis
HIF-1a plays a crucial role in the adaptive responses to hypoxia of tumor cells through transcriptional activation of downstream genes required for tumor survival and progression, including those for angiogenesis (25). It is known that MM cells upregulate HIF-1a in the hypoxic BM niche, which in turn promotes MM progression (26). Previous studies demonstrated the capacity of ET-1 to induce HIF-1a in solid tumors (12)(13)(14). In addition recent network-based analysis highlighted the significant ET-1 gene enrichment in the HIF-1a signaling pathway of PCs from BM samples of patients affected by MM and MGUS (18). Consequently, we tried to assess whether ET-1 was capable of up-regulating HIF-1a in MM, similarly to hypoxia. We therefore evaluated HIF-1a protein levels in RPMI-8226, U266, and KMS-12-PE cells cultured for 48 h in the presence of the hypoxia-mimetic agent CoCl 2 , ET-1, or their combination. As shown in Figure 3 both CoCl 2 and ET-1 as well as their combination induced HIF-1a in all cell lines. Interestingly, ET-1 induced HIF-1a similarly to CoCl 2 in RPMI-8226 and, at higher levels, in U266 and KMS-12-PE cells (p<0,001). The combination of ET-1 plus CoCl 2 promoted a further induction of HIF-1a in RPMI-8226 and U266 cells. Importantly, macitentan significantly inhibited both CoCl 2 -and ET-1-induced HIF-1a, also reducing the combined effect of CoCl 2 plus ET-1 (Figure 3). Taken together these results demonstrate that, in MM, the ET-1 axis activates and potentiates HIF-1a expression, which is significantly down-

Macitentan Disrupts the Mutual Dependence Between ET-1 Axis and Hypoxia Inducible Factor-1alpha
Our previous work had shown that both MM and BM microenvironment cells, including ECs, express and release ET-1, Overall, these findings indicate that ET-1 axis and HIF-1a pathway in MM are mutually dependent in an autocrine and paracrine fashion, and that macitentan may target crucial HIF-1a-dependent pathways.

Macitentan Exerts a Significant Antiangiogenic Activity by Downmodulating Cytokine Expression and Release by Multiple Myeloma Cells
Angiogenic switch represents a crucial step towards MM progression (27). Several experimental evidences have already shown that, in MM cells, hypoxic condition targets the transcription of a number of pro-angiogenic cytokines genes via HIF-1a, including VEGF-A, IL-8, and ADM (28,29). As ET-1 up-regulates HIF-1a, we evaluated whether the transcription of HIF-1a-dependent pro-angiogenic genes in MM may be mediated by the ET-1 axis. We thus evaluated by RT-qPCR the transcription of VEGF-A, IL-8 and ADM in RPMI-8226, U266 and KMS-12-PE cells exposed for 3 h to CoCl 2 , ET-1, or their combination. ET-1, either alone or in combination with CoCl 2 , up-regulated in all cell lines the expression of VEGF-A, IL-8 and ADM genes, which in turn were effectively down-modulated by macitentan ( Figure 5). The capacity of ET-1 to increase VEGF-A, IL-8 and ADM gene transcription was significantly reduced in HIF-1a-silenced MM cells (Supplementary Figure 5), further reinforcing the existence of an interplay among ET-1 axis, hypoxia and angiogenesis in MM. The pro-angiogenic activity of the ET-1 axis in MM was then evaluated by measuring VEGF-A, IL-8 and ADM protein levels in RPMI-8226, U266, and KMS-12-PE cells cultured in the presence of CoCl 2 , ET-1, or their combination. As shown by Western blots in Figure 5B, VEGF-A, constitutively expressed at low levels by all cell lines, was significantly up-regulated by either CoCl 2 or ET-1, alone or in combination. In all experimental settings, macitentan downregulated the expression of VEGF-A. As demonstrated by ELISA analysis, the supernatants from RPMI-8226, U266 and KMS-12-PE cells cultured with CoCl 2 , ET-1, or their combination, contained increased concentrations of IL-8 and ADM, which in turn were significantly reduced by the addition of macitentan ( Figure 5C). Overall, these findings indicate a crucial role of the ET-1 axis as a mediator of angiogenesis in MM. More importantly, our data highlight the capacity of macitentan to interfere with the expression and release of pro-angiogenic cytokines, such as ET-1 itself, VEGF-A, IL-8 and ADM through the down-modulation of HIF-1a in MM cells.
RPMI-8226 cells were therefore grafted in an alginate pellet and the number of newly formed vessels converging towards them was counted. As shown in Figure 6, when grafted on the top of the chick embryo CAM, RPMI-8226 cells induced a potent angiogenic response, which was significantly inhibited by the presence of increasing concentrations of macitentan, clearly unveiling its anti-angiogenic activity in MM.

DISCUSSION
In this study we show that the ET-1 axis increases MM cell growth and angiogenesis both in an autocrine and paracrine manner. Of translational relevance, we demonstrate that in xenograft mouse models macitentan impairs MM cell growth and decreases MM-related vascularity. In our opinion, these data reinforce and extend the role of ET-1 axis as therapeutic target in MM (8,9,30). Our in vitro experiments have shown that MM cells express and release increased levels of ET-1 when cultured with CoCl 2 , the latter mimicking the hypoxic conditions of the BM niche.
Similar to CoCl 2 , exogenous ET-1 increased HIF-1a levels in MM cells by up-regulating the transcription of HIF-1adependent pro-angiogenic genes such as VEGF-A, IL-8, ADM, and ET-1 itself. Under the same experimental settings, macitentan effectively down-modulated HIF-1a, the expression  of HIF-1a-dependent pro-angiogenic genes, and the production/ release of the corresponding cytokines. These findings appear of particular interest, as HIF-1a may increase angiogenesis in MM BM under the combined influence of hypoxia and ET-1. In agreement with such hypothesis, HIF1-a-silenced MM cells failed to express VEGF-A, IL-8 and ADM, other than ET-1 itself, upon exposure to exogenous ET-1. Significant antiproliferative effects in terms of MM cell viability in BM and subcutaneous tumor growth were observed in vivo in MM mouse models treated with macitentan. In the same models, residual viable MM cells in BM showed a reduced expression of HIF-1a, while subcutaneous masses displayed a significantly lower MVD than controls. In line with these findings, macitentan effectively impaired the induction of angiogenesis by MM cells in CAM assay. Overall, the set of experimental evidence here provided underscores a particular sensitivity of MM cells towards ET-1Rs blockade, which can hamper ET-1-triggered crucial MM prosurvival and pro-angiogenic signaling pathways, such as those involving MAPK/ERK (31) and HIF-1a (28,32), respectively. MM is known to interplay through mutual pro-survival/ proliferative signals with the BM microenvironment (4,33), the latter being characterized by angiogenesis as a hallmark of disease progression. Novel drugs targeting not only tumor cells, but also the BM microenvironment have been proved to be highly effective in MM (31). Under this perspective, oral immunomodulatory imide drugs, including thalidomide, lenalidomide, pomalidomide as well as proteasome inhibitors, such as bortezomib, are currently used in clinical protocols due to their anti-proliferative and antiangiogenic proprieties (27). Our experimental evidence together with literature data indicate that macitentan may also target both tumor and BM microenvironment, as ET-1 axis drives autocrine and paracrine pro-survival and pro-angiogenic signals in both malignant PCs and ECs. Indeed, not only MM cells (8), but also ECs (10) release increased levels of ET-1 through a HIF-1adependent activation (34), further reinforcing the hypothesis of an ET-1 axis-mediated mutual support between MM cells and ECs in the hypoxic BM microenvironment. Furthermore, based on ET A R and ET B R expression by MM cells (8) and ET B R by ECs (10), ET-1Rs blockade could target not only MM cells, but also the formation of new vessels, therefore exerting a potent therapeutic action against both MM and angiogenesis. Worthy of note, the expression of HIF-1a by ECs in BM has been also pointed as a therapeutic target in MM (35). At the same time, ET-1Rs blockade could inhibit the autocrine activation of ET-1-axis in both MM cells and ECs (10,35,36), thus interrupting an autocrine other than a paracrine pathogenic loop. Of interest, due to its pharmacokinetic properties (10), macitentan can be effective towards MM cells expressing ET A R only or ET A R and ET B R, as well as towards ET B R positive ECs (Figure 7).
ET-1 axis has been considered for a long time a possible therapeutic target in a number of solid tumors, though with fluctuating results (16). As previously stated, macitentan is a second-generation dual receptor antagonist that has been approved since 2013 for the therapy of PAH in the United States and Europe (37,38). Macitentan does not appear to require dosage adjustment in patients with hepatic and renal impairment (37), thus presenting a toxicity profile potentially suitable for the treatment of MM, a disease often associated with renal failure. Additionally, macitentan was shown to sensitize tumor cells to different cytotoxic and targeted agents in various preclinical tumor models, including colorectal cancer, glioblastoma, breast, and lung brain metastasis (39). In this regard, our preliminary in vitro findings indicated that macitentan can be administered in combination with other drugs already used for MM treatment. Accordingly, we found that macitentan exerts a synergistic action towards MM cells with the proteasome inhibitor bortezomib (9) and with the oral immunomodulatory imide drug pomalidomide (R.A. and C.T. unpublished results).
Overall, our data point out a potential therapeutic role for macitentant in MM patients that deserves to be exploited in further preclinical and phase 1-2 clinical studies, either alone or in combination with other anti-MM drugs.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material; further inquiries can be directed to the corresponding authors.

ETHICS STATEMENT
The animal study was reviewed and approved by Technical and Scientific Committee of University of Verona and Italian Ministry of Health.