Extra Cellular Matrix Derived Metabolite Regulates Angiogenesis by FasL Mediated Apoptosis

Object Antiangiogenic treatments are beginning to give promising outcomes in many vascular diseases including tumor angiogenesis. In this current study the antiangiogenic and pro-apoptotic actions of α1(IV)NC1 and its N- and C- peptides α1S1(IV)NC1, α1S2(IV)NC1 were investigated in-vitro and in-vivo. Study Method Endothelial cells (ECs) were treated with α1(IV)NC1, α1S1(IV)NC1, α1S2(IV)NC1 and in-vitro proliferation, migration, tube formation and apoptotic assays were executed. FasL, Fas, Caspase-8, -3 and PARP activations were studied using immunoblotting analysis using specific antibodies. Also the in-vivo antiangiogenic and pro-apoptotic effects were tested using α1(IV)NC1 in a mice model. Results Like α1(IV)NC1, its N- and C- terminal α1S2(IV)NC1 and α1S1(IV)NC1 domains posses anti-proliferative, pro-apoptotic activity and inhibit ECs migration and tube formation in-vitro. Both α1S1(IV)NC1 and α1S2(IV)NC1 domains promote apoptosis by activating FasL and down stream apoptotic events including activation of caspase-8, -3 and PARP cleavage in a dose dependent manner in-vitro in ECs. Tumors in mice showed apoptotic TUNEL positive microvasculature upon α1(IV)NC1 treatment, indicating inhibition of tumor angiogenesis and tumor growth. Further, the antitumor activity of α1(IV)NC1 was abrogated when caspase-3 inhibitor was used. These results conform additional properties of α1(IV)NC1 as an endogenous angioinhibitor that induces apoptosis in-vitro and in-vivo by activating FasL mediated caspase-3. Significance α1(IV)NC1 and its N- and C- terminal α1S1(IV)NC1 and α1S2(IV)NC1 domains also posses pro-apoptotic and angioinhibitory activity in-vitro and in-vivo. α1(IV)NC1 regulates tumor angiogenesis by activating FasL mediated apoptosis in-vitro and in-vivo. These results demonstrate that α1(IV)NC1 and its peptides inhibit neo-vascular diseases.


Introduction
Angiogenesis, the formation of new blood vessels from preexisting blood vessels, is a very stringently controlled program and normally does not occur, except during development and wound repair processes [1,2]. This stringent regulation of angiogenesis is manifested by a balance between pro-and antiangiogenic factors, which keep angiogenesis in check [2]. However, the dynamic equilibrium between pro-angiogenic and anti-angiogenic factors are controlled under many pathological settings, including tumor angiogenesis in cancer progression and other incidents like as age-related macular degeneration, retinopathy of prematurity and diabetic retinopathy resulting in the growth of abnormal new blood vessels [3][4][5].
Vascular basement membranes (VBM) constitute an important component of blood vessels [6]. Makeover of VBM can provide vital pro-and anti-angiogenic molecules to control formation of new blood vessels [7][8][9]. Type IV collagen is a major component of VBM and plays a critical role in new blood vessel development [6]. Proteolytic degradation of type IV collagen in the VBM generates numerous antiangiogenic molecules [7,[10][11][12]. One such antiangiogenic molecule derived from type IV collagen noncollagenous (NC1) domain a1 chain, a1(IV)NC1, has been tested in variety of tumor angiogenesis studies in mice [13][14][15]. However, the molecular and cellular mechanism(s) responsible for inhibition of angiogenesis is not yet clearly understood. The in-vitro and in-vivo studies have demonstrated that a1(IV)NC1 can directly affect endothelial cell migration and impact their proliferation and sprouting [14]. Earlier we have demonstrated that a1(IV)NC1 promotes apoptosis via activation of caspase-3 and PARP cleavage by inhibiting FAK/p38-MAPK/Bcl-2 and Bcl-x L signaling cascade [15]. These results provide a clear understanding about the apoptotic signaling and therapeutic potential of a1(IV)NC1 molecule in neovascular diseases. However, the effects of a1(IV)NC1 and its N-and C-terminal domains a1S1(IV)NC1 and a1S2(IV)NC1 on endothelial cell apoptosis and neo-vascularization have not been previously studied.
In the present study, we demonstrate that a1(IV)NC1 and its Nand C-terminal domains a1S1(IV)NC1 and a1S2(IV)NC1 are potent inhibitors of endothelial cell proliferation, migration and tube formation in-vitro and tumor angiogenesis in-vivo. a1(IV)NC1 promotes apoptosis via activation of caspase-3 and PARP cleavage, presumably by inhibiting FAK/p38-MAPK/Bcl-2 and Bcl-x L signaling cascade [15]. Here in this study, we show that Nand C-terminal domains of a1(IV)NC1 cross talk with FasL and activate FasL and its downstream apoptotic missionary including caspase-8, caspase-3 and PARP cleavage in-vitro. Furthermore, we identified that a1(IV)NC1 promotes apoptosis in tumor vasculature and inhibits angiogenesis and this effect was reversed by a caspase-3 specific inhibitor DEVD in-vivo. These findings contribute significantly towards understanding the apoptotic activation in proliferating ECs and therapeutic potential of endogenous angioinhibitor a1(IV)NC1 and its N-and C-terminal a1S1(IV)NC1 and a1S2(IV)NC1 domains in tumor growth and tumor angiogenesis.

Materials and Methods
Fetal calf serum (FCS), Endothelial basal medium (EBM-2) and Endothelial cell growth medium (EGM-2) were obtained from Fischer Scientific Inc. Penicillin and streptomycin and low melting agarose were purchased from Sigma-Aldrich and cell stains hematoxylin and eosin (H&E) were purchased from Fischer Scientific Inc. Sephadex TM -G 100, -G 25 and -G 200 were purchased from GE Healthcare Bio-Sciences AB. BD Matrigel TM Matrix (14.6 mg/ml) was purchased from BD Biosciences Discovery Laboratory. T 4 -DNA ligase (bacteriophage ligase), different restriction enzymes and polymerases were purchased from New England Biolabs. SCC-PSA1 tumor cells were purchased from ATCC. Caspase inhibitors z-DEVD-fmk was from Enzyme System Products. H&E staining kit form Fisher Scientific Inc. CD31 antibody was purchased from Upstate. Caspase-3, -8, PARP, FasL and Fas antibodies were purchased from Cell Signaling. MTT assay kit was purchased from Chemicon.
Ethics Statement. Boys Town National Research Hospital IACUC committee approved all mice studies.
Cloning, Expression and Purification of a1(IV)NC1 and its N-and C-terminal domains a1S1(IV)NC1 and a1S2(IV)NC1 The coding sequence corresponding to the N-and C-terminal non-collagenous domains (NC1) from human Collagen type IV a1 chain was isolated from the placental cDNA using one-step reverse transcriptase-PCR (RT-PCR) (Invitrogen, CA). The N-terminal subunit 330-bp from full length a1(IV)NC1 was amplified using the forward primer: ATACATATGGGCTTCCTTGTGAC-CAGGCATA and the reverse primer: CACAAGCT-TAGGCGCCTCACACACAGCAC and cloned between 'NdeI' and 'XhoI' sites of pET22b. The C-terminal subunit 330-bp from full length a1(IV)NC1 was amplified using the forward primers: GTGCATATGGTGCACAGCCAGACCAT and the reverse primer: GTGGCAGCAGCCAACTCA and cloned between 'NdeI' and 'XhoI' sites of pET22b. Amplification and cloning was carryout similarly as reported in our earlier publication [18]. The positive clone was used to transform E. coli strain BL21 for protein expression and purification was performed similarly as reported earlier [18].

Proliferation assay
A suspension of 7000-cells/well mouse choroidal endothelial cells (ECs) in a 96 well plate was used in proliferation assay. Cells were grown in 96 well plate under 0.5% FBS supplemented with heparin, endothelial mitogen, glutamine and penicillin/streptomycin. After 24-hrs, medium was replaced with ECs medium containing 10% FCS and different concentrations of a1(IV)NC1 or its N-and C-terminal a1S1(IV)NC1 and a1S2(IV)NC1 domains (0.25 and 2.0 mM) and after 48-hrs relative levels of methylene blue incorporation was measured as reported [14,17].
Migration assay About 1.0610 4 cells/well of ECs were seeded in serum free medium with and without recombinant a1S1(IV)NC1 and a1S2(IV)NC1 (1.0 mM). Medium containing 10 gg/ml of VEGF was placed into the bottom wells of the Boyden chamber and incubated for about 48-hrs at 37uC with 5% CO 2. The numbers of ECs that were migrated and attached to the bottom side of the Boyden chamber membrane were counted as reported earlier [14,19].

Tube formation assay
Briefly, Matrigel matrix about 250 ml was thawed overnight on ice-cold room and added to each well of a 24-well plate and allowed to solidify for 30-min at 37uC culture incubator. A suspension of about 50610 3 ECs in medium without antibiotic was plated on top of the Matrigel matrix. The ECs were than incubated with and without a1(IV)NC1 or its N-and C-terminal a1S1(IV)NC1 and a1S2(IV)NC1 domains (1.0 mM) for 48-hrs at 37uC and viewed using a CK2 Olympus microscope [14,20,21].
FasL and caspase activation assay About 1.0610 6 serum starved EC were collected and suspended in serum free medium. These ECs were pretreated with a1(IV)NC1 N-and C-terminal a1S1(IV)NC1 and a1S2(IV)NC1 domains (1.0 mM) and incubated on 10-cm 2 dishes for 6 and 18hrs. After 6 and 18-hrs, the floating ECs were collected, adherent cells washed once with cold PBS. Adhering and floating cells were lysed on ice in RIPA lysis buffer and centrifuged at 4uC for 30-min at 13000-rpm. About 30 mg of cytosolic extract/lane was separated using 10% SDS-PAGE followed by immunoblotting with anti-FasL, Fas, caspase-8, caspase-3 and PARP antibodies. Immunoreactivity of FasL, Fas, cleaved caspase-8, caspase-3 and PARP proteins were visualized using ECL detection kit as reported [15].
In-vivo tumor studies About 1.0610 6 SCC-PSA1 teratocarcinoma tumor cells were implanted subcutaneously into each mouse at dorso-lateral sides. Six male 129/Sv mice were used in each treatment and control groups. Once the tumors reached around 100-mm 3 sizes, 30 mg of a1(IV)NC1 or in combination with 10 mg of caspase-3 specific inhibitor DEVD in 100 ml total volume was intravenously injected to each mouse daily for 15-days. All the experimental mice were scarified and tumors and other organs were collected [21,31].

Immunohistochemical staining
Tumors from control and a1(IV)NC1 or in combination with DEVD treated mice were embedded in OTC compound and snap-frozen in liquid nitrogen. Tumor sections were incubated for 2-hrs with 5% BSA in PBS to block any nonspecific binding and then incubated with anti-CD-31 antibody for 1-hr followed by incubation with peroxidase-labeled goat anti-rabbit secondary antibody at 37uC. The number of CD-31 positive blood vessels per 10-microscopic fields from four tumor sections per condition was counted [14,23].

TUNEL assay for apoptosis
To evaluate the relative levels of CD-31 positive vascular or cellular apoptosis, and to quantify cellular apoptosis in tumors from control and a1(IV)NC1 or in combination with DEVD treated mice with and without a1(IV)NC1 treatment, we used ApopTag apoptosis detection kit. Briefly the tumor sections were incubated with 5% BSA in PBS to block any nonspecific binding and stained with ApopTag apoptosis detection reagents following the manufacturer's instructions. To determine whether CD-31 positive blood vessels were undergoing apoptosis, control and a1(IV)NC1 or in combination with DEVD treated tumor sections were co-incubated with both ApopTag reagents and anti-CD-31 antibody. Tumor sections were washed twice and incubated with rhodamine-conjugated goat anti-rabbit secondary antibody for 1hr at 37uC. The fluorescence staining was analyzed using a fluorescence microscope [14,23].

Results
Distinct anti-angiogenic activities of a1(IV)NC1 and its Nand C-terminal domains a1S1(IV)NC1 and a1S2(IV)NC1 a1(IV)NC1 is an endogenous metabolite generated from the non-collagenous (NC1) domain of a1 chain of type IV collagen by matrix metalloproteinases-9 (MMP-9) [24,25]. It was discovered as an angioinhibitory protein with substantial anti-tumor activities [13,14]. The present study was aimed at understanding the molecular mechanism(s) underlying angioinhibition by a1(IV)NC1 and its N-and C-terminal a1S1(IV)NC1, a1S2(IV)NC1 domains and its implications in prevention of tumor angiogenesis. Our studies tried to identify the angioinhibitory and pro-apoptotic activities of a1(IV)NC1 and its N-and C-terminal a1S1(IV)NC1, a1S2(IV)NC1 domains in mouse choroidal endothelial cells (ECs).
We have performed different angiogenesis experiments to determine the angioinhibitory potential of a1(IV)NC1 and its Nand C-terminal domains a1S1(IV)NC1 and a1S2(IV)NC1 using ECs. Many serum proteins are major pro-angiopromoting factors, elevated levels or imbalance of serum proteins (example VEGF and bFGF) are responsible for the majority of ocular angiogenesis driven by ischemia and also tumor angiogenesis. We first determined serum stimulated angioinhibitory activity of a1(IV)NC1 and its N-and C-terminal domains a1S1(IV)NC1 and a1S2(IV)NC1 by measuring ECs proliferation. The antiproliferative effect of a1(IV)NC1 and its N-and C-terminal domains were tested in ECs using methylene blue incorporation. Fetal calf serum (10% FCS) stimulated proliferation of ECs were significantly inhibited by a1(IV)NC1 its N-and C-terminal a1S1(IV)NC1 and a1S2(IV)NC1 domains in a dose dependent manner ( Figure 1A). Interestingly, both N-and C-terminal a1S1(IV)NC1 and a1S2(IV)NC1 domains were showing similar proliferation inhibitory activity when compared to its parent molecule a1(IV)NC1.
Migration of endothelial cells is basically essential during neovascularization [14,26]. Migration of ECs across a PVD membrane towards VEGF in a Boyden chamber was inhibited by 1.0 mM N-and C-terminal a1S1(IV)NC1 and a1S2(IV)NC1 domains ( Figure 1B). We additionally confirmed the angioinhibitory action of a1(IV)NC1 and its two domains a1S1(IV)NC1 and a1S2(IV)NC1 by a functional assay of ECs tube formation [20]. Tube formation on Matrigel is associated with ECs proliferation, migration and survival [1]. Treatment of ECs with a1(IV)NC1 and its domains a1S1(IV)NC1 and a1S2(IV)NC1 inhibited tube formation equally on Matrigel matrix ( Figure 1C). Our previous study reported that the angioinhibitory activity of a1(IV)NC1 is mediated through a1b1 integrin [14,21]. Surprisingly, in this study we noticed that ECs when treated with N-and C-terminal domains and full length a1(IV)NC1 (1.0 mM) were appeared rounded (apoptosis-like) and some of the cells detach from the Matrigel matrix (data not shown). This might be due to activation caspases and apoptosis in ECs treated with a1(IV)NC1 or its N-and C-terminal domains a1S1(IV)NC1 and a1S2(IV)NC1 similarly as reported earlier [15].
Further we also tested the angioinhibitory activity at different doses of a1(IV)NC1 and its N-and C-terminal domains a1S1(IV)NC1 and a1S2(IV)NC1 (0, 0.25, 0.5, 1.0, 1.5, 2.0 mM) by MTT cell viability assay after 10% FCS stimulation in ECs. The results reveal that ECs proliferation was significantly increased by FCS stimulation which was inhibited by a1(IV)NC1 and its N-and C-terminal domains dose dependently after 24-hrs ( Figure 2).

FasL mediated in-vitro apoptotic activity of a1(IV)NC1 and its N-and C-terminal a1S1(IV)NC1 and a1S2(IV)NC1 domains
Caspase-3 is a pivotal molecule mediating cellular apoptosis [27]. Earlier we demonstrated that a1(IV)NC1 induces apoptosis in endothelial cells by activating Caspase-3 [15]. Here in this study, we tested, whether a1(IV)NC1 and its N-and C-terminal domains posses pro-apoptotic activity or not? Interestingly, ECs incubated with a1(IV)NC1 and its N-and C-terminal domains showed dose and time dependent activation of FasL without affecting Fas expression compared to control untreated cells ( Figure 3A & B). These results suggest both a1S1(IV)NC1 and a1S2(IV)NC1 domains may cross talk with death receptor, activating FasL mediated apoptosis in ECs. These results demonstrate that a1(IV)NC1 and its two domains a1S1(IV)NC1 and a1S2(IV)NC1 promoted apoptosis through FasL in addition to MAPK signaling inhibition [15].
We further assessed whether a1S1(IV)NC1 and a1S2(IV)NC1 induces activation of FasL downstream different caspases such as caspase-8 and caspase-3. We treated ECs with a1S1(IV)NC1 and a1S2(IV)NC1 and observed that activation of caspase-8 ( Figure 4A and B). These results demonstrate that full length a1(IV)NC1, both its N-and C-terminal domains a1S1(IV)NC1 and a1S2(IV)NC1 induced cellular apoptosis by activating FasL and activating downstream caspase-8. Among the known caspases, caspase-3 is an important effector molecule for most cellular apoptosis [27]. To study whether caspase-3 could be activated by a1S1(IV)NC1 and a1S2(IV)NC1 domains, we treated ECs with different doses of a1S1(IV)NC1 and a1S2(IV)NC1 and observed activation of caspase-3 in a time dependent manner ( Figure 4C  and D). Here, we also identified that activation PARP cleavage which is further down stream to FasL, caspase-8 and caspase-3 in ECs treated with a1S1(IV)NC1 and a1S2(IV)NC1 time and dose dependently ( Figure 5A and B). These results demonstrate that activation of ECs apoptosis by a1S1(IV)NC1 and a1S2(IV)NC1 domains by activating FasL and its down stream caspases and PARP cleavage.

Neutralizing anti-tumorigenic activity of a1(IV)NC1 by caspase-3 inhibitor
To further conform the pro-apoptotic activity of a1(IV)NC1 also partly regulated by caspase-3 activation and apoptosis, we carried in-vitro tumor studies with a1(IV)NC1, or a1(IV)NC1 coadministrated with caspase-3 inhibitor DEVD to tumor bearing mice. Consistent to our earlier findings, administration of a1(IV)NC1 significantly inhibited SCC-PSA1 tumor growth and tumor angiogenesis in-vivo [27]. In contrast, treatment with caspase-3 specific inhibitor DEVD alone showed significant effect on inhibition of tumor growth. The antitumor activity of a1(IV)NC1 was significantly attenuated when a1(IV)NC1 and DEVD were co-administrated to tumor bearing mice (data not shown).
To assess the antitumor effects of a1(IV)NC1 and its ability to activate microvasculature apoptosis, control and a1(IV)NC1 treated mice tumor sections were stained with anti-CD31 antibody (Fig. 6, middle panel). In control tumors, less TUNEL positive apoptotic staining was detected whereas elevated levels of apoptosis was observed in tumors that were treated with a1(IV)NC1 (Fig. 6). DEVD treatment alone meagerly affected overall tumor cell apoptosis when compared with a1(IV)NC1. Further, a significant number of ECs in tumor vasculature underwent apoptosis upon a1(IV)NC1 treatment when compared to control tumors, as observed through CD31/TUNEL dual staining (Fig. 6, right panel yellow). Interestingly, co-administration of a1(IV)NC1 with DEVD to tumor bearing mice significantly reduced EC apoptosis in microvasculature that was induced by a1(IV)NC1. Further apoptotic tumor cells were observed in a1(IV)NC1 treated tumors using TUNEL staining. This could be the result of an indirect effect since the microvasculature is affected by a1(IV)NC1 treatment, and the tumor cells possibly suffer from oxygen supply and nourishment resulting in apoptosis. This effect was observed at elevated levels in cells undergoing apoptosis, other than CD-31 positive ECs (Fig. 6, left and right panels). We also quantified the microvascular density and number of apoptotic vessels through TUNEL staining and by counting the number of CD-31 positive blood vessels (data not shown). Collectively, our results demonstrate that a1(IV)NC1 and its -N and -C terminal domains induced activation of FasL mediated caspase-3 activation and apoptosis in proliferating endothelial cells.

Discussion
Many endogenous metabolites generated from type IV collagen were identified as pro-apoptotic, antiangiogenic and anti-tumorigenic in nature [12][13][14][15]21,25,[28][29][30][31][32]. The non-collagenous (NC1) domain released by proteolytic activity from type IV collagen a1 chain (a1(IV)NC1) was identified as an antiangiogenic molecule, where as its N-and C terminal domains apoptotic activity is not yet studied. Antiangiogenic activity of a1(IV)NC1 is mediating through a1b1 integrin [14,21]. However, it is critical to examine a1(IV)NC1 and its both its N-and C-terminal domains a1S1(IV)NC1 and a1S2(IV)NC1 effects in several well defined relevant in-vitro experiments before confirming that these domains are also antiangiogenic. In this study, we tested the antiangiogenic/pro-apoptotic effects of a1(IV)NC1 and its N-and Cterminal domains in different in-vitro and in-vivo experiments.
We demonstrate in this study for the first time, that a1(IV)NC1 and its N-and C-terminal domains inhibits serum induced EC proliferation, migration and tube formation. In addition, we also identified that both a1S1(IV)NC1 and a1S2(IV)NC1 domains activates caspase-8, caspase-3/PARP cleavage through FasL activation in proliferating endothelial cells. This is coherent with the earlier studies revealing that antiangiogenic activity of a1(IV)NC1 is mediated through a1b1 integrin signaling and apoptosis [14,33]. a1(IV)NC1 and its N-and C-terminal a1S1(IV)NC1 and a1S2(IV)NC1 domains activated FasL mediated apoptosis in ECs. Understanding the mechanism(s) of action of a1S1(IV)NC1 and a1S2(IV)NC domains is crucial for their therapeutic development and use. Thus, a1(IV)NC1 and its N-  and C-terminal domains promotes apoptosis in proliferating ECs and inhibits tumor angiogenesis, these endogenous molecules may be an effective therapeutic candidate for treatment of many neovascular diseases. Further evaluation through extensive laboratory studies on these molecules is needed to address the function of these angioinhibitors to be considered for the clinical trials.
Our findings suggest that a1(IV)NC1 and its N-and C-terminal domains a1S1(IV)NC1 and a1S2(IV)NC1 may also be effective in inhibition of tumor angiogenesis by activating caspase-3 in new blood vessels. This is further supported by earlier findings that another angioinhibitor, a3(IV)NC1 regulate angiogenesis in a number of in-vitro and in-vivo models [10]. Regression of tumor growth in mice upon a1(IV)NC1 treatment is associated with reduced tumor vasculature and increased TUNEL positive endothelial and tumor cells when compared to control tumors, indicating in-vitro activation of caspase-3. This suggests that a1(IV)NC1 activates caspase-3 and inhibits tumor angiogenesis and tumor growth. Thus, this work not only supports our efforts in development of a1(IV)NC1, its N-and C-terminal domains a1S1(IV)NC1 and a1S2(IV)NC as a potential candidate for tumor angiogenesis neovascular diseases.