Pleiotropic Anti-Angiogenic and Anti-Oncogenic Activities of the Novel Mithralog Demycarosyl-3D-ß-D-Digitoxosyl-Mithramycin SK (EC-8042)

Demycarosyl-3D-ß-D-digitoxosyl-mithramycin SK (DIG-MSK) is a recently isolated analogue of mithramycin A (MTA) that showed differences with MTA in the DNA binding strength and selectivity. These differences correlated with a better therapeutic index and less toxicity in animal studies. Herein, we show that DIG-MSK displays a potent anti-tumor activity against different types of cancer cell lines, ovarian tumor cells being particularly sensitive to this drug. Of relevance, DIG-MSK exerts low toxicity on fibroblasts and peripheral blood mononuclear cells, this toxicity being significantly lower than that of MTA. In correlation with its antitumor activity, DIG-MSK strongly inhibited Sp1-mediated transcription and endogenous Sp1 mRNA expression, which correlated with the inhibition of the expression of key Sp1-regulated genes involved in tumorigenesis, including VEGFA, BCL2L1 (Bcl-XL), hTERT, BRCA2, MYC and SRC in several ovarian cells. Significantly, DIG-MSK was a stronger inhibitor of VEGFA expression than MTA. Accordingly, DIG-MSK also exhibited potent anti-angiogenic activity on microvascular endothelial cells. Likewise, it significantly inhibited the gene expression of VEGFR1, VEGFR2, FGFR, PDGFB and PDGFRA and, additionally, it induced the expression of the anti-angiogenic factors angiostatin and tunstatin. These effects correlated with a pro-apoptotic effect on proliferating microvascular endothelial cells and the inhibition of the formation of endothelial capillary structures. Overall, the pleiotropic activity of DIG-MSK in inhibiting key oncogenic and angiogenic pathways, together with its low toxicity profile, highlight the therapeutic potential of this new drug.


Introduction
Induction of angiogenesis is a hallmark of the tumorigenic process that plays a key role in tumor progression and metastasis [1]. Cancer cells secrete pro-angiogenic factors that stimulate capillary growth into the tumor bed. Vascular endothelial growth factor (VEGF) is a crucial angiogenic factor that comprises two structurally and functionally related proteins termed VEGFA and VEGFB, being VEGFA the dominant angiogenic factor in most settings [2,3]. VEGF functions as a ligand for two tyrosine kinase receptors: vascular endothelial growth factor receptor 1 (VEGFR1) and 2 (VEGFR2), both expressed on the surface of endothelial cells (ECs). Additionally, a redundant number of angiogenic factors, including angiopoietins, platelet-derived growth factor (PDGF) and fibroblast growth factor (FGF), collaborate with VEGF in the angiogenic process [4].
Anti-angiogenic drugs, mainly targeting VEGF and PDGF, have been successfully introduced in the anti-cancer therapy [5,6]. Bevacizumab, a humanized anti-VEFGA antibody, significantly improved the prognosis of colorectal cancer patients [7]. Multiple inhibitors targeting VEGFR2 and other angiogenic factors, such as sunitinib and sorafenib, displayed promising effects on renal carcinoma patients [5,6]. However, many tumors sooner or later become refractory or resistant to VEGF blockade. This resistance is frequently caused by the activation of alternative angiogenic proteins, such as FGFs, due to preexisting multiple redundant pro-angiogenic molecules and pathways [5,6,8].
Mithramycin A (MTA) is an aureolic acid-type polyketide antibiotic produced by various species of Streptomyces [9]. MTA binds to GC-rich regions in DNA and interferes with the action of several transcription factors; including Sp1. Toxic side effects of MTA have limited its clinical use to the treatment of Paget's disease and hypercalcemia [10]. Nevertheless, the interest in MTA as an anti-cancer drug has been renewed in 2012, when two clinical trials (NCT01610570 and NCT01624090) were launched at the National Cancer Institute to treat Ewing sarcoma and non-small cell lung cancer. The expansion of mithramycin chemical diversity through combinatorial biosynthesis and biocatalysis has yielded dozens of mithralogs [11]. A promising mithralog with improved therapeutic window, named DIG-MSK (demycarosyl-3D-ß-D-digitoxosyl-mithramycin SK; EC-8042) has been reported (Fig 1). We showed that DIG-MSK is as efficient as MTA in inhibiting Sp1 function, but it differs from MTA in the strength and selectivity of DNA binding [12]. Likewise, DIG-MSK is a stronger inhibitor of endogenous SP3 gene expression and Sp3-driven transcription than MTA [12]. In vivo, the mithralog EC-8042 exerts a potent antitumor activity against ovarian, colon, breast, melanoma, non-small-cell lung and central nervous system cancers, but it is one order of magnitude less toxic than MTA [13].
Sp1 regulates key angiogenic factors, including VEGF, FGF and FGF receptor (FGFR); and the therapeutic inhibition of Sp1 by MTA directly correlated with potent anti-angiogenic effects in vitro and in vivo [14][15][16][17][18]. Herein, we report that DIG-MSK has a pleiotropic antitumor activity, targeting key oncogenic and angiogenic pathways, including those mediated by VEGF, FGF or PDGF. The potent and pleiotropic biological activities exerted by DIG-MSK and its better toxic profile highlight the potential relevance of this agent to treat cancer.

Mithralogs
MTA and DIG-MSK (EC-8042) were isolated from Streptomyces argillaceus, purification was carried out by preparative HPLC and they were identified by HPLC-MS [13]. Their purity was checked by HPLC (~97%). Stock solutions of the different analogues were prepared in DMSO DIG-MSK were provided by the company EntreChem (http://entrechem.com/es/about_us/index.html). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. EntreChem provided support in the form of salaries for authors LEN and FM, but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the 'author contributions' section.
Competing Interests: The drugs MTA and DIG-MSK were provided by company Entrechem (http:// entrechem.com/es/about_us/index.html), the employer of Luz-Elena Núñez and Francisco Morís. Francisco Morís has stocks in Entrechem. There are no further patents, products in development or marketed products to declare. This does not alter the authors' adherence to all the PLOS ONE policies on sharing data and materials, as detailed online in the guide for authors. and kept at -20°C until use. This study was approved by the Ethics Committee of the Hospital Universitario Central de Asturias (Spain) and written informed consent was obtained from all donors.

SRB assay
The effect of MTA or DIG-MSK on cell viability was determined by using the SRB-method, as described elsewhere [19]. Cell viability was measured after 48 hours of treatment and IC50 values were calculated using the SRB assay carried out in triplicate.

VEGFR1 and VEGFR2 expression analysis
Cell surface protein expression of VEGFR1 and VEGFR2 was evaluated by flow cytometry by using the following primary antibodies: anti-VEGFR1 (sc-271789, Santa Cruz Biotechnologies) and anti-VEGFR2 (sc-504, Santa Cruz Biotechnologies). After washing, cells were stained with a fluorescein isothiocyanate (FITC)-conjugated secondary antibody (AbD Serotec) and were analyzed by flow cytometry (BD FACScanto, Becton Dickinson).

Apoptosis and cell cycle analyses
Cells were treated with 200 nM MTA or DIG-MSK and DMSO for 48 hours. Adherent and detached (floating) cell populations were pooled together, and the percentage of cell death by apoptosis and necrosis was determined by flow cytometry (BD FACScanto, Becton Dickinson) after double staining with anti-annexin V antibody (Immunostep) and 7-AminoActinomycin D (7-AAD) (Immunostep). Primary apoptosis was determined as the Annexin-V positive/ 7-AAD negative cell population. Necrotic cells were characterized as Annexin-V negative/ 7-AAD positive (primary necrotic cells) and Annexin-V positive/ 7-AAD positive (necrosis arising from apoptotic cells or primary necrosis). Specific cell death was calculated using the following formula: % specific cell death = (% death cell of sample-% death cells of negative control)/(100 -% death cells of negative control) x 100.
The activity of caspases 3 and 9 was measured by using the Caspase 3 Colorimetric Kit (Life Technologies) and the Caspase 9 Fluorimetric Kit (Invitrogen), following the manufacturer's recommendations.
For cell cycle analysis, cells were treated with MTA (200 nM) or DIG-MSK (200 nM) for various periods of time; cells were harvested, fixed with 70% ethanol, stained with propidium iodide (PI, Sigma) and analyzed by flow cytometry.

Migration assays
HUVEC and HMEC-1 cells were cultured until convergence and, then, cells were wounded by scratching, rinsed with PBS, and incubated with DMSO, or several doses of MTA or DIG-MSK. Wound closure was monitored over time by using an inverted microscopy (Motic AE2000) and a compatible camera (Motican 1000, Motic). Images were captured at 0, 16, 24, 48 and 72 hours and analyzed with ImageJ software (NCBI). Experiments were made by triplicate and cell migration was expressed as the percentage of distance migrated divided by the length of the initial wound.

Capillary-like Structure Formation Assay
HMEC-1 cells were incubated in 96 well plates treated with Matrigel, and treated with several doses of MTA or DIG-MSK. After 48 hours, capillary-like structures formation was captured and processed with Angiodraw Software. The angiogenic index (AI) was calculated as the number of branch points in a field. The experiments were made by triplicate and four different fields were counted for each experiment.

Quantitative Real Time PCR (qRT-PCR)
Total RNA was isolated from cells treated with DMSO, MTA (200 nM), or DIG-MSK (200 nM) for 8 hours using the High Pure RNA Isolating kit (Roche). cDNA was obtained using High Capacity cDNA Reverse Transcription Kit (Invitrogen). Quantitative Real Time PCR (qRT-PCR) was performed in triplicate by using HT7900 RT-PCR and SYBR Green PCR Master Mix (Applied Biosystems). The sequences of the oligos used for real-time PCR are shown in Table 1. The thermal profile was 95°C for 3 minutes, followed by 40 cycles of 95°C for 20 seconds; annealing temperature (specific for each gene, see Table 1) for 30 seconds and 70°C for 30 seconds. Relative expression values of the different genes were calculated from the threshold cycle (Ct) following the 2 −ΔΔCT method and using GAPDH as internal housekeeping control gene for normalization. Table 1. qRT-PCR primers used and thermal profiles.

Statistical analysis
Continuous variables were compared with Mann-Whitney U test. The p values <0.05 were considered statistically significant.

Results
Effect of DIG-MSK on the toxicity and the cell cycle of cancer cell lines and mononuclear blood cells The effect of MTA and DIG-MSK on the viability of a panel of cell lines was analyzed by determining the drug concentrations required for inhibiting cell viability by 50% (IC50) [19]. MTA and DIG-MSK strongly reduced the viability of most of the cell lines analyzed at low nanomolar doses (  Table 2). Of note, both drugs were particularly efficient on ovarian cells, in which low nanomolar doses successfully inhibited cell viability. Conversely, the toxicity of these analogues in embryo fibroblasts 3T3 was significantly lower than in cancer cell lines. The toxic effect of these drugs on the ovarian cells was further analyzed by measuring apoptosis (Fig 2A and 2B). Ovarian cells were treated with 200 nM MTA or DIG-MSK and cell death was analyzed by annexin V staining. A strong cytotoxic effect was observed with both analogues in the three cell lines analyzed, causing cell death by both apoptosis and necrosis (ranging from 40% to 60%). Of note, the induction of apoptosis in PBMCs obtained from 6 healthy donors was very low (Fig 2B), causing DIG-MSK significantly lower toxicity than MTA (2% vs. 12%; p<0.001). To extend these observations, caspase 3 and caspase 9 activities were analyzed. Both drugs induced a significant increase in caspase 3 and 9 activities in ovarian cells ( Fig 2C and 2D). In agreement with the cell death, both MTA and DIG-MSK had a significantly lower cytotoxic effect on healthy PBMCs than in ovarian cells, although the caspase 3 activity induced by DIG-MSK was significantly lower than MTA (p<0.01).  Next, we analyzed the effect of MTA and DIG-MSK on the proliferation of ovarian tumor cells by analyzing the cell cycle distribution. As shown in Fig 3, both analogues halted the cell cycle and induced a G1 arrest in IGROV-1 and OVCAR-3 cells, whereas no such effect was observed in A2780 cells.

DIG-MSK inhibits the expression of Sp1-target genes in ovarian cancer cells
To get further insight into the signaling pathways underlying the activity of MTA and DIG-MSK, the effect of these compounds on the expression of a transfected luciferase reporter vector containing several Sp1-response elements was determined. Both drugs were potent inhibitors of Sp1-mediated transcriptional activity in ovarian cells (Fig 4A). In agreement, they were also potent inhibitors of endogenous Sp1 expression and the transcription of eight Sp1-target genes involved in key aspects of the ovarian tumorigenesis as determined by qRT-PCR (Fig 4B and 4C). MTA and DIG-MSK significantly down-regulated the transcript levels of VEGFA, BCL2L1 (Bcl-XL), hTERT, BRCA2, MYC and SRC genes. DIG-MSK showed, in general, a similar or stronger inhibitory effect than MTA on the transcription of such Sp1-regulated genes. Particularly, DIG-MSK was more efficient than MTA in inhibiting VEGFA gene expression in A2780 cells (1.3-fold), OVCAR-3 cells (2.8-fold, p<0.05) and IGROV-1 (3.2-fold, p<0.001) cells ( Fig 4C). However, a stronger induction of P21 and MICA expression was observed upon treatment with MTA, the latter being a gene involved in the activation of the anti-tumor immune response [20].

DIG-MSK exerts a pro-apoptotic effect and suppresses tube formation by microvascular endothelial cells
To further analyze the anti-angiogenic properties of DIG-MSK, its effect on microvascular endothelial cells (ECs) was analyzed. No significant effect was observed on cell cycle progression in HUVEC or HMEC-1 cells (Fig 5A). Likewise, no EC migration was not significantly affected by DIG-MSK, as determined by the wound/scratch-healing assay (not shown). However, DIG-MSK treatment induced significantly lower levels of cell death in HUVEC cells than MTA, with no marked differences observed in HMEC-1 cells (Fig 5B).
Next, the formation of new blood vessel was analyzed by conducting the endothelial tube formation assay in which Matrigel was used as a substrate. The integrity of the capillary-like structure formed by HMEC-1 cells was significantly inhibited by DIG-MSK in a dose-dependent manner, and this effect was observed at non-cytotoxic doses (Fig 6 and S1 Fig). However, MTA was only able to inhibit capillary-like formation at high doses, which was associated with a high toxicity in HMEC-1 cells.

MTA and DIG-MSK modulate the expression of angiogenic genes in human microvascular endothelial cells
To gain further insight into the mechanism underpinning the suppression of capillary formation by DIG-MSK, the impact of this drug on the transcription of key angiogenic genes (VEGFA, VEGFB, VEGFC, VEGFD, VEGFR1, VEGFR2, VEGFR3, FGF, FGFR, PDGFA, PDGFB, PDGFRA, PDGFRB, EGF and HIF1A) in ECs was analyzed by qRT-PCR. As shown in Fig 7A, a significant reduction in VEGFR1 and VEGFR2 gene expression in HUVEC cells (ranging from 40 to 50% of inhibition). A down-regulation of VEGFR2, FGFR and PDGFRA gene expression was observed with both analogues in HMEC-1 cells. DIG-MSK also inhibited VEGFR1 and PDGFB expression in these cells (Fig 7A and 7B). Contrarily, MTA significantly up-regulated the angiogenic factor PDGFA in HUVEC cells. Similarly, their effect on the transcript levels of key anti-angiogenic genes encoding angiostatin, tunstatin, endostatin and thrombospondin-1 was analyzed (Fig 8). Angiostatin gene expression was up-regulated by both analogues and tunstatin was induced by DIG-MSK in HUVEC cells. Conversely, a reduction of endostatin expression was observed in HUVEC cells treated with both drugs.

Effect of MTA and DIG-MSK on the secretion of VEGF and the surface expression of VEGFR1 and VEGFR2
Given the strong inhibition of VEGF gene expression in ovarian cancer cells, and the reduction of the mRNA expression of VEGF receptors observed in ECs upon exposure to MTA and DIG-MSK, we next studied the modulation of these key regulators of the angiogenic process at  (Fig 9A). Further, both drugs completely inhibited the production of VEGFA in IGROV-1 cells (15.4-fold MTA, 19.63-fold DIG-MSK, p<0.05). Of note, in OVCAR-3 cells, DIG-MSK was significantly more efficient than MTA in inhibiting the secretion of VEGFA (1.45-fold, p<0.05). Likewise, flow cytometry analyses also showed that surface expression of VEGFR1 and VEGFR2 was significantly down-regulated by MTA and DIG-MSK in HUVEC and HMEC-1 cells (p<0.01) (Fig 9B). Pleiotropic Anti-Angiogenic and Anti-Oncogenic Activities of DIG-MSK Overall, our results show that DIG-MSK has a pleiotropic anti-angiogenic effect by up-regulating the expression of key angiogenic genes as well as repressing the transcription of key antiangiogenic genes, indicating that this new drug may be useful for anti-cancer therapies that target angiogenic pathways.

Discussion
MTA is a DNA-binding drug that shows preference for G/C-rich DNA sequences, and, therefore, a significant part of the anti-tumor activity of MTA is thought to be mediated by the blockade of the binding of transcription factors such as Sp1 [21]. Sp1 expression and activity are up-regulated in many cancers, correlating with advanced stage and poor prognosis [22][23][24][25][26]. Accordingly, Sp1 inhibition may decrease tumor formation, growth and metastasis  [14,27,28], representing an attractive target, despite being considered undruggable by conventional drug discovery approaches [29].
DIG-MSK is a novel analogue of MTA that shares the general features of other mithralogs, such as the preference for binding to G/C-rich DNA sequences and inhibition of Sp1-driven transcription [21]. Nevertheless, DIG-MSK has significant differences in the strength and specificity of DNA binding [12]. Thus, DIG-MSK is a stronger inhibitor of the expression of Sp3 and Sp3-driven transcription than MTA. In spite of the fact that Sp1 and Sp3 recognize the same DNA element with similar affinity, these two transcription factors possess different functions, and the relative level of both proteins differentially regulates the expression of several genes [22,30,31,32]. Overall, these differences between MTA and DIG-MSK point that different therapeutic activity and toxicity between both analogues may exist. In fact, initial studies reported that DIG-MSK has an improved therapeutic index compared to the parental MTA, since it is one order of magnitude less toxic than MTA in vivo [13]. In agreement, in our study we report that DIG-MSK displays low toxicity in primary cells. The toxicity of MTA and DIG-MSK was low in mouse embryo fibroblasts and mononuclear blood cells obtained from healthy donors, and, in these cells, DIG-MSK showed significantly lower toxicity than MTA. These results are in agreement with in vivo studies were DIG-MSK has been well tolerated and no evidence of toxicity or cell death was observed in ovarian cells and other tissues [13].
Noteworthy, DIG-MSK retains a potent anti-tumor activity against cells obtained from different types of cancers, but ovarian cells were particularly sensitive to this drug. Similar to the parental MTA, DIG-MSK potently inhibited Sp1-mediated transcription in these ovarian tumor cells, and, given that the Sp1 gene is autoregulated in a positive feedback, DIG-MSK was also a potent inhibitor of endogenous Sp1 expression. In correlation, DIG-MSK inhibited Sp1-regulated genes involved in the expression of key oncogenic processes, such as DNA repair (BRCA2), cell proliferation (MYC, SRC), telomere maintenance (hTERT) and cell survival (BCL-XL). Notably, p21 WAF1 /CDKN1A, which is involved in halting cells at G1 and G2/M phases, was induced in OVCAR-3 and IGROV-1, which correlated with a halt of the cell cycle and G1 arrest.
Of note, one of the genes strongly inhibited by DIG-MSK was VEGFA, which is a wellknown Sp1-regulated gene in cancer [33,34]. In fact, it has been reported that MTA is a strong inhibitor of VEGF and it shows potent anti-angiogenic effects [14][15][16][17][18]. Nevertheless, we observed that DIG-MSK was more efficient than MTA in inhibiting key angiogenic genes. Thus, DIG-MSK was a strong inhibitor of VEGF pathway; but, contrasting with other antiangiogenic drugs, DIG-MSK showed a dual effect inhibiting both VEGFA expression in cancer cells and VEGFR1 and VEGFR2 expression in ECs. Further, it also reduces the expression of other key angiogenic receptors, such as FGFR, PDGFRA and PDGFRB. Additionally, it up-regulates the expression of the anti-angiogenic genes encoding for angiostatin and tunstatin. The up-regulation of anti-angiogenic factors by MTA treatment has been previously described and it has been shown to exert anti-myeloma effect in vivo [17]. Conversely, MTA treatment increased PDGFRA mRNA expression and both drugs inhibited the expression of endostatin in HUVEC cells. However, these effects are cell-specific and could be linked to the specific characteristics of the cell line or with the transformation mechanisms of these cells. Further studies are needed to elucidate the role of DIG-MSK in the regulation of these genes in ECs. Nevertheless, the final effect of DIG-MSK treatment was a pro-apoptotic effect on proliferating microvascular ECs and the inhibition of the formation of capillary structures at non-toxic doses, which may be due to the effect of these angiogenic factors on the formation and stabilization of new blood vessels. Of note, MTA has a more limited effect on the inhibition of angiogenic factors and it was unable to inhibit the formation of capillary-like structures at non-toxic concentrations, suggesting that DIG-MSK has improved anti-angiogenic properties compared with the parental MTA.
Several anti-angiogenic drugs, mainly targeting VEGF and PDGF, have been successfully introduced in the anti-cancer therapy [5,6]. The majority of these drugs are highly specific and they are associated with low toxicity. However, many tumors become sooner or later refractory or resistant to VEGF blockade. Compensation by other angiogenic factors, such as FGF and angiopoietins, is a major factor that contributes to tumor resistance to VEGF blockade [35][36][37]. Contrasting the majority of drugs against angiogenesis, DIG-MSK shows pleiotropic effects on DNA targets that affect a wide range of Sp1-regulated genes. This array of anti-angiogenic and anti-oncogenic activities has important advantages and disadvantages. On one hand, they may limit the development of drug resistance. On the other hand, this multi-target activity may be associated with elevated toxicity, which may limit its therapeutic use. Notably, among many derivatives of MTA, DIG-MSK has been selected because it has showed a lower toxic profile than MTA, which highlights the potential of this drug as an efficient and safer anti-cancer agent.