CARP-1 Functional Mimetics: A Novel Class of Small Molecule Inhibitors of Medulloblastoma Cell Growth

Medulloblastomas (MBs) constitute an aggressive class of intracranial pediatric tumors. Current multimodality treatments for MBs include surgery, ionizing radiation, and chemotherapy. Toxic side effects of therapies coupled with high incidence of recurrence and the metastatic spread warrant development of more effective, less toxic therapies for this disease. CARP-1/CCAR1 is a peri-nuclear phospho-protein that is a co-activator of the cell cycle regulatory anaphase promoting complex/cyclosome (APC/C) E3 ligase. CARP-1 functional mimetics (CFMs) are a novel class of small molecule compounds that interfere with CARP-1 binding with APC/C subunit APC-2, and suppress growth of a variety of cancer cells in part by promoting apoptosis. Here we investigated MB growth inhibitory potential of the CFMs and found that CFM-4 inhibits growth of MB cells in part by inducing CARP-1 expression, promoting PARP cleavage, activating pro-apoptotic stress-activated protein kinases (SAPK) p38 and JNK, and apoptosis. Gene-array-based analysis of the CFM-4-treated Daoy MB cells indicated down-regulation of a number of key cell growth and metastasis-promoting genes including cell motility regulating small GTP binding protein p21Rac1, and extracellular matrix metallopeptidase (MMP)-10. Moreover, CFM-4 treatment stimulated expression of a number of molecules such as neurotrophin (NTF)3, and NF-κB signaling inhibitors ABIN1 and 2 proteins. Overexpression of NTF3 resulted in reduced MB cell viability while knock-down of NTF3 interfered with CFM-4-dependent loss of viability. CFMs also attenuated biological properties of the MB cells by blocking their abilities to migrate, form colonies in suspension, and invade through the matrix-coated membranes. Together our data support anti-MB properties of CFM-4, and provide a proof-of-concept basis for further development of CFMs as potential anti-cancer agents for MBs.


Introduction
Medulloblastoma is a common childhood brain cancer. It is a highly malignant tumor type with poor overall prognosis [1]. Current treatment options include a combination of surgery, radiation and chemotherapy. The disease however remains incurable in about a third of the patients, and the therapyassociated severe neurological toxic side effects often result in significant morbidity [2].
Although it is well known that most MBs originate from the distinct germinal zones of the cerebellar cortex [3,4], the transforming events that initiate and/or sustain these cancers are yet to be elucidated. Differential expression of some antigens and receptors such as neurotrophin receptor p75 NTR /TrkC is often noted in common variants of MB and serves as a marker of favorable outcome [5,6]. Moreover, recent gene expression profiling studies have defined four MB subgroups that include sonic hedgehog subgroup, the WNT subgroup, and subgroups 3 and 4 [7]. Whole genome sequencing of a number of primary medulloblastoma tumors representing all the four subgroups further revealed that mutations in different epigenetic modifiers may distinguish MB subgroups 3 and 4, thus presenting potential for targeting of subgroup-specific alterations for therapeutic benefit [8].
Emerging evidence indicates that although majority of MBs harbor wild-type p53 [9], the tumor suppressor function of p53 is often impacted by the associated oncogenes such as MDM2 and/ or WIP1/PPMID [10,11]. In this context, recent, proof-ofprinciple studies have revealed that tumor suppressor functions of p53 can be restored in MBs. The micro-RNA miR-34a was found to sensitize MB cells to chemotherapy in part through its modulation of p53 [12], while a pharmacological inhibitor of MDM2 (nutlin-3) re-activated p53 function and suppressed growth of MB cells in vitro and in vivo [11]. Another recent study explored utility of modified measles virus for treatment of MBs in vitro and in intracerebral murine xenograft model [13]. Thus although significant advances have been made in understanding the biology of MBs, the current treatment modalities remain insufficient to effectively treat and manage this disease, and therefore, warrant development of new anti-MB treatment strategies.
We previously identified and characterized a peri-nuclear phospho-protein, termed CARP-1/CCAR1 [14,15]. CARP-1 is a co-activator of the cell cycle regulatory anaphase promoting complex/Cyclosome (APC/C) E3 ligase [16] and the p53 [17], and is a key transducer of cell growth as well as chemotherapy (adriamycin, etoposide, or iressa)-dependent inhibitory signaling [14,15,17]. The APC/C co-activator function of CARP-1 was recently exploited to identify a number of small molecule inhibitors (SMIs) of CARP-1 binding with APC-2 subunit of APC/C. We have previously shown that these compounds, termed CARP-1 Functional Mimetics (CFMs), suppress growth of a variety of cancer cells in part by stimulating apoptosis [16]. Here we investigated MB growth inhibition by CFMs and the underlying molecular mechanisms. CFMs stimulated pleiotropic anti-MB cell growth signaling that included activation of apoptosis and attenuation of cell growth and survival. Last but not least, CFMs also interfered with biological properties of colony formation, invasion and migration of the MB cells. Thus, our proof-of-concept studies underscore MB inhibitory properties of CFMs that could facilitate development of CFMs or their derivatives/analogs as potential anti-MB modalities.

Cells and Reagents
The human medulloblastoma Daoy and UW-228-1 cells were obtained from Dr. Ronaldo DelMaestro (Brain Tumor Research Center, Montreal Research Institute and Hospital, Montreal, Canada) and were maintained essentially as described before [18]. All the cells were routinely cultured in DMEM supplemented with 10% fetal bovine serum (FBS), 100 units/ml of penicillin, and 100 mg/ml of streptomycin. Cells were maintained at 37uC and 5% CO 2 and were passaged weekly. CFM-1, 4, and 5 were obtained from ChemDiv and/or ChemBridge, San Diego, CA, and dissolved in dimethyl sulfoxide (DMSO; Sigma; St. Louis, MO) at a stock concentration of 50 mM, aliquoted and stored at -20uC, and appropriate working concentrations were prepared with the cell culture medium immediately before the experiments. Clinical grade Adriamycin was from the Harper Hospital Pharmacy, Wayne State University, Detroit, MI. Purified anti-NT3 (neurotrophin 3) mouse monoclonal antibody, clone 5A2, and human NTF3 expression plasmid pCMV6-XL5-NTF3 variant 1 (catalog # sc316918) were obtained from Origene Technologies, Inc., Rockville, MD. Anti-Trk (C-15) rabbit polyclonal, anti-ABIN2 mouse monoclonal, and anti-APC-2 antibodies were purchased from Santa Cruz Biotechnology, Inc., CA. In addition, anti-TrkC mouse monoclonal (62CT51.7.7) and rabbit monoclonal (C44H5) antibodies were purchased from LifeSpan Biosciences, Inc., Seattle, WA and Cell signaling Technology, Beverly, MA, respectively. Anti-cyclin B1 and anti phospho-JNK (Threonine183/Tyrosine 185) G9 mouse monoclonal antibodies, anti-JNK (56G8) rabbit monoclonal antibody, and rabbit polyclonal antibodies for PARP, phospho and total p38 SAPK, ABIN1, IkBa, and IkBb proteins were obtained from Cell Signaling Technology (Beverly, MA). Anti-p21 Rac1 mouse monoclonal antibody was purchased from BD Biosciences, San Jose, CA. Generation and characterization of the anti-CARP-1/ CCAR1 rabbit polyclonal antibodies have been described before [14]. Enhanced Chemiluminescence Reagent was purchased from Amersham Biosciences (Piscataway, NJ) and the Protein Assay Kit was purchased from Bio-Rad Laboratories (Hercules, CA), while Anti b-Actin mouse monoclonal antibody, and chemicals including 3-4, 5-dimethyltiazol-2-yl-2.5-diphenyl-tetrazolium bromide (MTT), cremophor were obtained from Sigma-Aldrich (St. Louis, MO). Purified, lyophilized recombinant human pro-neurotrophin 3 and mature neurotrophin 3 proteins were obtained from BioVision Inc., Milpitas, CA and Cell Signaling, Beverly, MA, respectively, and resuspended in sterile PBS to obtain a 100 mg/ml stock of the respective proteins. The reconstituted protein stocks were aliquoted and stored at 220uC, and appropriate working concentrations were prepared immediately before the experiments. The ON-Target plus SiRNAs for knock-down of NTF3 and DharmaFECT transfection reagent for si-RNA transfections were purchased from Dharmacon Inc., Thermo Fisher Scientific (Lafayette, CO). The lipid-based transfection reagent Lipofectin 2000 for transfection of plasmid DNAs was purchased from Life Technologies, Grand Island, NY.

MTT Assay, Caspase Activity Measurement and Western Blot (WB)
MB cells (5610 3 ) were seeded in a 96-well culture plate and subsequently treated with respective CFM at different concentrations for noted times. Control cells were treated with 0.1% DMSO in culture medium. After indicated periods of treatment, the cells were incubated with 1 mg/ml of MTT reagent at 37uC for 4 hours and then MTT was removed and the purple formazan crystals formed at the bottom of the wells were dissolved using 100 ml of DMSO, followed by colorimetric analysis using a multilabel plate reader at 570 nm (Victor 3 ; PerkinElmer, Wellesley, MA).
Activation of caspases was measured by utilizing the ApoAlert Caspase profiling plate (Clontech) following the manufacturer suggested methods. Cell lysates derived from vehicle (DMSO) or CFM-treated cells were added to the chambers containing immobilized fluorogenic caspase-2, -8, -9, and -3 substrates. The fluroscence signal released from the activated caspase-mediated cleavage of the respective substrate was detected by a plate reader at the excitation and emission wavelengths of 380 and 460 nm, respectively.
For WB analyses, cells were harvested and lysed in RIPA buffer (50 mM Tris-HCI, pH 8.0, 150 mM sodium chloride, 1.0% NP-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS), and 0.1% of protease inhibitor cocktail) for 20 min at 4uC. The lysates were centrifuged at 14,000 rpm at 4uC for 15 min to remove debris. Protein concentrations of whole cell lysates were determined using the Protein Assay Kit. Supernatant proteins, 50 mg from each sample, were separated by SDS-10% polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene difluoride (PVDF) membrane (Bio-rad, Hercules, CA) by standard procedures. The membranes were hybridized with primary antibodies followed by incubation with appropriate secondary antibodies. The antibody-bound proteins were visualized by treatment with the chemiluminescence detection reagent (Amersham Biosciences) according to manufacturer's instructions, followed by exposure to X-ray film (Kodak X-Omat). The same membranes were re-probed with the anti-b actin antibody, which was used as an internal control for protein loading.

Isolation of RNA and Microarray Analysis
Total RNA was extracted from untreated or CFM-4-treated Daoy MB cells. The MB cells were treated with 2 mM, 6 mM and 8 mM dose of CFM-4 for 12 h in serum-free medium. At the end of treatments, the untreated and treated cells were harvested and total RNA were isolated, and purified using the RNeasy Mini kit and RNase-free DNase Set (Qiagen, Valencia, CA) according to the manufacturer's protocols. CFM-4-dependent changes in gene expression in MB cells were performed at the Genomic Core Facility, Karmanos Cancer Institute utilizing Illumina BeadChipH Arrays essentially according to manufacturer's instruction (Illumina Inc., San Diego, CA). Briefly, 0.5 mg total RNA was biotinlabeled and hybridized with BeadChips. The signal was detected with streptovadin-Cy3 according to manufacturer's instruction (Illumina). The imaging of the BeadChips was conducted using a Bead Array Reader in conjunction with Bead Studio software (Illumina). Normalization of the data was carried out using a quantile based approach which transforms the raw data so that the resulting normalized expression values of each sample have the same distribution [19,20]. Differential gene expression between the treated and untreated groups was tested using a moderated ttest. Genes were selected as significant using a cutoff of 0.1 on the nominal p-values and a cutoff of 1.5-fold change in expression between groups.

NTF3 Knock-down Experiments
Daoy MB cells were transfected with scrambled or NTF3 SiRNAs utilizing Dharmafect reagent essentially following the manufacturer suggested protocols. After 96 h incubation with the SiRNAs, the cells were either lysed, and protein extracts analyzed by WB for expression of pro-NTF3, or cells were treated with various doses of CFM-4 and their viabilities were determined by the MTT assay as detailed above. In addition, a plasmid encoding variant 1 of the human pro-NTF3 (cat. # SC316918) was obtained from Origene Technologies, Rockville, MD. UW-228-1 MB cells were transfected with the vector or the pro-NTF3 expression plasmid. The levels of pro-NTF3 in the transfected cells and the viabilities of the transfected cells that were either untreated or treated with CFM-4 or CFM-5 were determined by WB and MTT assays, respectively. For stable NTF3 knockdown, we first PCR amplified pro-NTF3 cDNA from positions 191-698 (NM_001102654.1) and subcloned this PCR amplified cDNA in an antisense orientation to obtain a recombinant plasmid (pcDNA3/NTF3-AS Clone1). Daoy MB cells were then transfected with pcDNA3 vector or pcDNA3/NTF3-AS clone 1 plasmids, and the transfected cells were grown in the continuous presence of 200 mg/ml neomycin (G418) as described before [21]. Multiple, stable, neomycin-resistant sublines were isolated followed by determination of pro-NTF3 levels by WB essentially following the previously published methods [14,16]. The stable sublines expressing vector or the reduced levels of pro-NTF3 were then treated with CFM-4 (4 mM) and CFM-5 (10 mM) for various time periods, and their viabilities were determined by the MTT assays as above.

Cell Migration, Invasion, and Clonogenic Assays
The effects of CFMs on migration of MB cells were measured by the ''scratch'' assay. Cells were grown in a 6-well plate (,10,000 cells/well), and a scratch was created in the cell monolayer using sterile pipette tip. The cells were then allowed to grow in the absence (Control) or presence of 10 mM dose of each of the CFMs for a period of 72-96 hours. Images were captured at the beginning and at regular intervals during cell migration to close the scratch, and the images were compared to quantify the migration rate of the cells essentially as described before [22]. The cells were photographed under different magnifications utilizing Zeiss microscope with attached 35 mm camera for recording the photomicrographs.
Clonogenic assay. Cells were sandwiched between 0.6% and 0.3% agarose in DMEM medium containing 5% FBS in a six-well chamber (500 cells/chamber), and treated with buffer (Control), or respective CFM (10 mM) for 9 days at 37uC humidified CO 2 incubator. The colonies from multiple random fields were counted, compared to control and photographed essentially as above.
Invasion assay. Basement membrane is a thin extracellular matrix (ECM) that underlies epithelia and endothelia and separates epithelial cancer cells from the stroma. Tumor cells produce proteases that degrade ECM to cross the basement membrane to invade stroma and establish distant metastases. The in vitro assay using Matrigel is the most reliable, reproducible, and representative of in vivo invasion. In this Boyden Chamber assay, cancer cells are placed in the upper chamber that is separated from the lower chamber by a porous membrane coated with Matrigel [23]. A cell invasion Boyden chamber assay kit (Chemicon International, CA) was utilized to measure invasion properties of the cancer cells in the absence or presence of CFMs. Briefly, pre-warmed serum free medium (300 ml) was used to hydrate the ECM layer of each chamber for 15-30 minutes at room temperature. Approximately 2-2.5610 5 MB cells were seeded in the upper chamber in a serum-free medium without or with 10 mM dose of the respective CFM. The lower chamber was supplied with full medium containing 10% FBS that served as chemo-attractant to stimulate migration. After an interval, tumor cells present on the lower side of the membrane in the lower chamber were stained, and photographed as above. In addition, the stained cells from the lower side of membrane of some wells were dissociated, lysed in a buffer, followed by quantitation using a fluorescence plate reader with 480/520 nm filter set. The measurements were then plotted as bars in histogram.

Immunocytochemical Analysis
Apoptosis in MB cells was determined by TUNEL assay using in situ cell Death Detection kit from Roche Applied Science (Indianapolis, IN) according to the manufacturer's instruction. Approximately 5610 3 cells were seeded in each well of an 8chamber slide (LabTek, CA), and allowed to grow overnight at 37uC incubator. The cells were then either untreated or treated with respective CFM, and processed for staining essentially following our previously described procedures [24]. The cells on the slides were then stained for presence of CARP-1, phospho-p38, TrkC, and NTF3 by utilizing respective antibodies, and were then photographed under different magnifications using Zeiss microscope with a 35 mm camera attached for recording the photomicrographs. H&E counter-staining of the cells was performed following our previously described methods [24].

Detection of MMP Expression in Human MB Cells
Daoy cells were separately treated with CFM-4 or CFM-5, followed by their homogenization in RIPA buffer (500 ml of lysis buffer per 1610 6 cells). The cell lysates were centrifuged at 10,0006g for 5 min, and the protein concentration in the supernatant of the respective lysate was determined by using Bicinchoninic acid assay. The lysates were stored at 280uC until further use. MMP activation in each lysate was measured using the Quantibody reverse phase human MMP array kit according to manufacturer's instructions (RayBiotech, Norcross, GA). Fluorescence images were detected using a GenePix 4100A Scanner, and data was analyzed using the QAH-MMP-1 GAL software based on the instruction provided by the array manufacturer.

Statistical Analysis
Where appropriate, statistical analysis was performed using unpaired Student's t-test. A p value less than 0.05 between treatment groups is considered significantly different.

CFMs Inhibit MB Cell Proliferation in Part by Stimulating Apoptosis
We first tested the growth-inhibitory effects of CFMs on two human MB cell lines, Daoy and UW-228-1. CFM-1, 4, 5 inhibited the viability of both cell lines in a dose and time-dependent manner ( Figure 1). In general, Daoy cells were more sensitive towards all the CFMs, showing ,50% reduced viability in the presence of 20 mM CFM-1 ( Figure 1A). Doses of 3 mM and 2 mM of CFM-4 and CFM-5, respectively, over a 24 h treatment elicited ,50% reduced viability of Daoy cells ( Figure 1B). The UW-228-1 cells were relatively less sensitive to all the CFMs since a 20 mM dose of CFM-1 caused ,33% reduction in their viability ( Figure 1A). Approximately 50% attenuation of the viability of UW-228-1 cells was noted when treated with a dose of 4 mM and 2 mM of CFM-4 and CFM-5, respectively, over a 24 h period ( Figure 1B). In addition, since the Daoy and UW-228-1 MB cells were relatively more sensitive to CFM-4 or CFM-5, these cells were separately treated with various doses of CFM-4 and CFM-5 for a period of 24 h to determine the lowest effective dose for these agents. Although 500 nM dose of either compound failed to affect MB cells viability, a 1 mM dose of the either agent elicited a 20-40% loss in the viability of both the cell lines ( Figure 1C, D). These data suggest that the viabilities of the MB cells are affected by the 1 mM or higher doses of each of the compound. We next investigated whether CFMs promoted apoptosis to inhibit MB cell growth. Given that a 12 h treatment of both the MB cells with 10 mM dose of CFM-1 resulted in a modest, 20-30-% loss of their viabilities, while a 12 h exposure to 10 mM dose of CFM-4 or CFM-5 suppressed viabilities of the MB cells by a 60-70%, we chose to utilize a 10 mM dose of each compound to determine induction of apoptosis and the underlying molecular mechanism(s). For immunocytochemical analyses, the MB cells were directly cultured in 8-well chamber slides and treated with CFMs as detailed in Methods. Treatment of the MB cells with CFM-1, 4,  Figure 2C). Since the DNA-damage promoting chemotherapeutic Adriamycin (ADR) has been shown to induce PARP-1 cleavage and apoptosis in a number of cancer cells [25], our WB analysis also revealed cleavage of PARP in Daoy cells following treatment with ADR in a dose and time-dependent manner ( Figure 2C). In light of our earlier studies indicating requirement of various caspase activities in transduction of CFM-4-dependent growth inhibitory signaling in human breast cancer cells [16], and the fact that CFMs promoted PARP cleavage in MB cells ( Figure 2C), we next determined whether caspases were activated following exposure of the MB cells to CFMs. For this purpose, we profiled activation of caspases-2, -9, -8, and -3 in UW-228-1 cells following their treatments with CFM-1, -4, or -5 utilizing a fluorescence-based quantitative assay as noted in methods. As expected, treatments of MB cells with CFM-1, -4, or -5 resulted in activation of caspases-2, -9, -8, and -3 ( Figure 2D). The data in figures 1 and 2 therefore suggest that CFMs are novel inhibitors of MB cell growth, and although the MB cells display varying levels of sensitivity towards these compounds, all the three CFMs significantly attenuate MB cell growth in part by stimulating apoptosis.

Apoptosis Signaling by CFMs Involves Activation of Proapoptotic p38 SAPK/MAPK, and Elevated Expression of a Novel Transducer of Apoptosis CARP-1/CCAR
We and others have previously reported identification and characterization of CARP-1/CCAR1 that is a novel and biphasic transducer of cell growth and apoptosis signaling. While CARP-1 was found to be a co-activator of nuclear, steroid/thyroid transcription factors, it was also demonstrated to be a key regulator of p53 function and a transducer of apoptosis signaling by DNA damaging agents such as ADR [14,17]. Since both the MB cells displayed increased sensitivity to inhibition by CFM-4 and CFM-5, we next determined whether exposure of MB cells to these compounds caused increase in CARP-1 expression. In the first instance, the Daoy and UW-228-1 cells were either untreated or treated with CFM-4, and CARP-1 levels were analyzed by immuno-cytochemical staining utilizing anti-CARP-1 a2 antibodies as noted in Methods. Exposure to CFM-4 resulted in increased staining for CARP-1 in both the MB cells ( Figure 3A, B). Consistent with elevated CARP-1 in CFM-4-treated Daoy cells, WB analysis of the lysates derived from ADR or CFM-4-treated Daoy cells revealed a robust increase in CARP-1 expression while CARP-1 presence was undetectable in the lysates from the untreated control cells ( Figure 3C). Moreover, since CFM-5 inhibited growth of the Daoy cells (see figure 1) it also provoked an increase in CARP-1 levels ( Figure 3C). The data in figures 3A-C demonstrate that ADR or CFM-4 enhance CARP-1 expression in the MB cells.
Our earlier studies have also indicated that in addition to regulating apoptosis signaling, CARP-1 regulated expression of a number of key cell cycle and cell growth regulatory molecules. In particular, knockdown of CARP-1 resulted in elevated levels of mitotic cyclin B1, cell growth and motility regulatory small GTPbinding protein p21Rac1, and oncogene c-myc [14]. Moreover, treatment of human breast cancer cells with CFM-4 resulted in reduced expression of cyclin B1 [16]. In light of these observations, we further investigated whether exposure of the MB cells to CFMs also results in loss of cyclin B1, p21Rac1, and c-myc proteins. As shown in figure 3 C, D, and E, CFM-4 or CFM-5 treatments suppressed expression of cyclin B1, p21Rac1, and c-myc proteins in Daoy MB cells.
We have previously observed that increased expression of CARP-1 stimulated apoptosis that involved activation of p38, while blockage of p38 activation abrogated CARP-1-mediated apoptosis [15]. In light of earlier findings demonstrating p38 activation during MB cell apoptosis [21], and together with our recent report describing identification and characterization of novel small molecule CFMs, in particular the CFM-4 that binds with CARP-1, increases CARP-1 levels, p38 activation and apoptosis [16], we speculated that CARP-1 expression and p38 activation are key events in MB cell apoptosis signal transduction in the presence of ADR or CFMs. To test this possibility, we determined whether and to what extent CFMs can stimulate p38 activation in the MB cells. Immuno-cytochemical and WB analyses were performed to determine CFM-mediated changes in p38 activation. Treatments with CFM-1, -4, or -5 resulted in elevated staining for phosphorylated (activated) p38 in Daoy ( Figure 4A) and UW-228-1 ( Figure 4B) MB cells when compared with their untreated counterparts. WB analysis of the cell lysates further corroborated a robust activation of p38 in ADR, CFM-4, or CFM-5-treated Daoy ( Figure 4C) and UW-228-1 ( Figure 4D) MB cells when compared with their untreated controls.

CFMs Suppress MB Cell Growth in Part by Inducing Expression of Cell Growth and Survival Inhibitory Molecules
In an attempt to further elucidate molecular mechanisms of MB cell growth inhibition in the presence of CFMs, we first conducted a gene-array-based analysis. Daoy MB cells were either untreated or separately treated with three doses of CFM-4 as described in Methods. The RNAs from each group were hybridized with genearray chips, and the data were analyzed to identify genes that had a significant 1.5-fold or higher altered expression following CFM-4 treatments. A select subset of these genes is indicated in table 1 (see Table S1 for complete list of CFM-4-regulated genes in Daoy MB cells). Interestingly, the array data revealed down-regulation of a vast majority of genes while a small number of gene were upregulated. Among the genes that were down-regulated were cell growth and metastasis-promoting genes such as matrix metalloproteinase (MMP)7, MMP10, cyclin E1/E2, BUB1, transcription factor E2F2, E3 ubiquitin ligase SKP2, and small GTP binding proteins CDC42 and p21Rac1 (Table 1). Of note is the fact that CFM-4-dependent down-regulation of p21Rac1 in our array experiment is supportive of our WB data in figure 4E where both the CFM-4 and CFM-5 treatments attenuated levels of p21Rac1 in Daoy cells. The array data also revealed up-regulation of a number of cell growth inhibitory genes in the CFM-4-treated Daoy cells. In particular, expression of the tumor necrosis factor alpha inhibitory protein (TNFAIP)3-interacting proteins (TNIP) 1 and 2 [also known as A20-binding inhibitor of NF-kB activation (ABIN) 1 and 2], as well as the nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor beta (NFkBIB, also known as IkBb) were up-regulated suggesting that the CFM-4 treatment likely affected the NF-kB-dependent survival signaling in MB cells. In addition, CFM-4-treated MB cells had elevated expression of the receptor-interacting protein kinase (RIPK)4 and neurotrophin (NTF)3 genes (Table 1), that are known to transduce cell growth inhibitory functions in a context and signal-dependent manner.
Nerve growth factor (NGF) and other members of the NTF family that function by activating receptor tyrosine kinases p75NTR and tropomyosin related kinases (trk) A, B, and C play important roles in survival, differentiation, and apoptosis of neurons [26][27][28]. In regard to the MBs, activation of TrkA and C receptors has been shown to stimulate apoptosis, while overexpression of TrkC receptor in childhood MBs is associated with favorable clinical outcome [21,29,30]. In light of the fact that NTF3 is a preferred ligand for TrkC receptor, and although low levels of 5.5-kb size alternate splice variant of TrkC mRNA were present in Daoy cells, no detectable levels of the 14-kb trkC, p75NTR, trkA, or trkB mRNAs as well as full-length p145 trkC receptor protein were found in these cells ( [21] and refs within). Consequently, Daoy cells were not inhibited by NTF3 treatment, while stable expression of wild-type TrkC receptor in Daoy MB cells sensitized them to the inhibitory effects of NTF3 [21].
Since gene-array experiment revealed upregulation of NTF3 in CFM-4-treated MB cells (see table 1), we next examined whether treatments of MB cells with CFM-4 or CFM-5 increased NTF3 levels. UW-228-1 cells were either untreated or treated with CFM-1 or CFM-5, and NTF3 levels were analyzed by immunocytochemical staining utilizing anti-NT3 mouse monoclonal antibody (clone 5A2) as noted in Methods. Exposure to CFM-5 resulted in increased staining for NTF3 in the UW-228-1 MB cells ( Figure 5A). WB analysis further revealed a moderate increase in pro-NTF3 in the Daoy cells following their treatment with CFM-4 or 5 for 3 h period while a robust increase in mature NTF3 was noticed in these cells exposed to CFM-4 or 5 over a 6 h period ( Figure 5B). A similar, albeit less robust, increase in the pro and mature NTF3 proteins was also noted in CFM-4 or CFM-5treated UW-228-1 cells ( Figure 5C). To determine the extent to which NTF3 is required in CFM-4-mediated growth suppression of MB cells, we conducted a siRNA-based knockdown of NTF3 in Daoy cells. In the first instance, the cells were transfected either with the scrambled siRNAs or NTF3-siRNAs as detailed in Methods, and the cell lysates analyzed by WB. Presence of NTF3-SiRNAs resulted in ,50% loss of pro-NTF3 peptide when compared with its levels in scrambled siRNA-transfected counterparts ( Figure 5D). We next transfected Daoy cells with scrambled or NTF3-siRNAs, followed by their treatments with vehicle DMSO (Control) or various doses of CFM-4 as in Methods. The viabilities of the siRNA-transfected, and control or CFM-4-treated cells were measured by the MTT assays essentially as in figure 1. MB cells transfected with scrambled siRNAs lost their viabilities in the presence of CFM-4 in a dose-dependent manner with a greater than 80% reduction in their viability by the 4 mM dose ( Figure 5E). Knock-down of NTF3 however resulted in just ,50% reduction in the number of viable cells in the presence of 4 mM CFM-4, thereby eliciting a statistically significant level of protection of MB cells from inhibitory effects of CFM-4. To further elucidate role of NTF3 in transducing MB-inhibitory effects of CFMs, we first generated and characterized stable, neomycin-resistant Daoy sublines that express NTF3 antisense as detailed in Methods. WB analysis revealed that over-expression of NTF3 antisense resulted in reduced levels of pro-NTF3 in a number of sublines when compared with their vector-transfected counterparts ( Figure  S1A). Two sublines, each expressing vector or NTF-3 antisense, were then independently treated with CFM-4 and CFM-5 for various time periods followed by determination of their viabilities as in Methods. Consistent with our data in figure 5E, stable knockdown of NTF3 interfered with the growth inhibitory effects of CFM-4 and CFM-5 ( Figure S1B, C). These data therefore collectively suggest that MB cell growth inhibition by CFM-4 is regulated in part by NTF3-dependent signaling mechanisms.
In order to determine the mechanism by which increased expression of NTF3 by CFM-4 negatively regulated growth of the MB cells, we first clarified whether the Daoy and the UW-228-1 MB cells expressed any variants of TrkC, and to the extent to which CFMs can alter levels of these receptors to render cells sensitive to inhibition by NTF3. CFM-5-treated Daoy cells as well as CFM-1-treated UW-228-1 cells elicited elevated immunostaining for TrkC protein ( Figure S2A). In addition, cellular proteins derived from untreated (Control) and CFM-treated Daoy and UW-228-1 MB cells were subjected to WB analysis utilizing multiple anti-TrkC antibodies as described in Methods. The anti-TrkC mouse monoclonal antibody reacted with ,100, 65, and 55 kDa size proteins in the Daoy cells while reacting only with ,100 kDa size protein in the UW-228-1 cells (Figure S2B, C). Of note is the fact that although a modest increase in the 100 kDa TrkC protein was noted in CFM-4 as well as CFM-5-treated UW-228-1 cells, expression of the 100 kDa species was reduced while the levels of 65 and 55 kDa peptides was significantly increased in the CFM-4-treated Daoy cells ( Figure S2B). Although the fulllength TrkC receptor protein has been reported to be ,145 kDa size, the available GenBank database information indicates presence of shorter sized, splice variants of the TrkC receptor  figure S2B and S2C corresponds to any of the above splice variant of TrkC, and the extent increased expression of the 65 and 55 kDa species in CFM-4-treated Daoy cells is due in part to apoptotic signaling induced by CFM-4 remain to be clarified. Additional WB analysis of the lysates derived from UW-228-1 MB cells utilizing anti-TrkC rabbit monoclonal antibody (Cell Signaling Technology) also revealed presence of ,100 kDa size peptide ( Figure S2D). Taken together, our WB analyses in conjunction with multiple anti-TrkC antibodies suggest for the presence of a variant or a TrkC-like peptide(s) in the MB cells. We next addressed whether the TrkClike peptide(s) contributed to NTF3-mediated MB cell growth suppression in the presence of CFM-4. MB cells were separately treated with CFM-4, affinity-purified pro-NTF3, affinity-purified mature NTF3, or a combination of CFM-4 and pro/mature NTF3, and cell viability was measured by MTT assay as in figure 1. Although treatment with a 5 and 50 ng/ml dose of pro-NTF3 elicited a modest (,20%) growth inhibition of the UW-228-1 MB cells, neither mature NTF3 nor a combination of CFM-4 and NTF3 resulted in a greater growth inhibition when compared with that noted with the CFM-4-treated cells ( Figure S3A, S3B). These data collectively suggest that although these MB cells express TrkC-like peptides, the elevated levels of NTF3 following treatments with CFM-4 or CFM-5 are not likely to be involved in extrinsic, receptor-dependent growth inhibitory signaling.
We next tested the possibility that elevated levels of intracellular NTF3 in the presence of CFM-4 or CFM-5 contribute to MB cell growth inhibition in the absence of functional TrkC. For this purpose, a plasmid encoding variant 1 of the human pro-NTF3 protein was transfected in the MB cells. Transfection of the NTF3 expression plasmid however resulted in elevated levels of NTF3 in the UW-228-1 cells ( Figure S3C). The vector or NTF3 plasmid transfected UW-228-1 cells were then either untreated or treated with CFM-4 or CFM-5, followed by determination of live, viable cells essentially as in figure 1. As shown in figure 5F, expression of NTF3 significantly attenuated viabilities of the UW-228-1 MB cells when compared with their vector-transfected controls. Moreover, the extent of the loss of viability of UW-228-1 cells following NTF3 expression was similar to that noted for the CFM-4 or CFM-5-treated, vector-transfected cells ( Figure 5F). Importantly, treatments of NTF3-expressing UW-228-1 cells with CFM-4 or CFM-5 resulted in greater reductions in their viability when compared to their CFM-4 or CFM-5-treated, or NTF3-transfected  Figure 6A, C). These data suggest that p38 activation likely precedes JNK activation, and taken together indicate that activation of both the pro-apoptotic MAPKs contribute to the MB cell growth inhibitory effects of CFM-4. In light of our gene-array data that revealed elevated levels of a number of key molecules that are well known to inhibit canonical NF-kB signaling (Table 1), we further clarified the extent to which CFM-4 treatments can upregulate levels of ABIN1, ABIN2, IkBa and IkBb proteins in MB cells. MB cells were treated with CFM-4 essentially as is figure 6A and C, and expression/levels of NF-kBinhibitory proteins were determined by WB. A moderate increase in levels of ABIN1 and 2 proteins was noticeable within 1 h of CFM-4 treatment of Daoy cells ( Figure 6B), while only ABIN1 levels increased after 3 h and 6 h of CFM-4-treated UW-228-1 cells ( Figure 6C). Interestingly, although a moderate increase in IkBb was noted within 1 h of CFM-4 treatment, no significant increase in IkBa levels was evident in the Daoy cells. In the case of UW-228-1 cells however CFM-4 exposure resulted in initial decline in IkBa levels followed by re-expression of IkBa at 3 h and 6 h treatment periods ( Figure 6C). These data would collectively suggest for a biphasic regulation of NF-kB signaling by CFM-4 in the MB cells. Since ABIN1 and 2 proteins negatively regulate NF-kB function in part by interfering with activation of its upstream kinase IKK [31,32], it is likely that a modest increase of ABIN1 and 2 proteins in the presence of CFM-4 likely inhibits canonical NF-kB signaling as an early response while prolonged exposure results in degradation of/reduced expression of IkBa and/or IkBb proteins in both the MB cells. It remains to be clarified whether loss of IkBa and/or IkBb following prolonged exposure to CFM-4 results in NF-kB activation, and to what extent activated NF-kB functions to stimulate apoptosis or promotes cell survival to overcome CFM-4-mediated stress.

CFM-4 Blocks Migration, Colony Formation, and Invasion of MB Cells
To investigate whether CFMs inhibit biological properties of migration, invasion and colony formation by the MB cells, we performed wound-healing, soft-agar, and matrigel-invasion assays, respectively, as detailed in Methods. Presence of CFM-1, -4, or -5 prevented the UW228-1 cells from growing in the areas of wound created by a scratch, and also caused a greatly reduced number and size of their colonies in soft agar when compared with their respective, untreated controls ( Figure 7A). The fact that CFM-4 treatment of Daoy cells resulted in a 2-fold down-regulation of MMP10 (Table 1), and that the activities of MMPs were found to be associated with glioblastoma and MB cell invasiveness [33][34][35], we determined whether and to what extent exposure to CFMs resulted in reduced activities of various MMPs. For this purpose we conducted an antibody-based array analysis to determine activation status of various MMPs in control (untreated) versus treated Daoy MB cells as indicated in Methods. These data revealed that exposure of Daoy cells to CFM-4 or CFM-5 resulted in attenuation of MMP-1, -2, -9, and -10 activities ( Figure 7B). SiRNA-mediated specific knock-down of either membrane type 1-MMP (MT1-MMP; MMP-14; a direct activator of MMP-2) or MMP-9 was recently found to reduce capacity of the Daoy cells to generate neurospheres while concomitantly abrogating their invasive capacity [36]. Whether CFM-dependent attenuation of MMP activities interfered with invasive properties of the MB cells was determined next by testing the extent to which CFMs blocked the ability of MB cells to invade through the matrigel-coated membranes. As expected, treatment of MB cells with CFM-1, -4, or -5 resulted in a significantly reduced number of MB cells that were able to migrate across the matrigel-coated membranes ( Figure 7C). Taken together, these data show that CFMs, in particular CFM-4 and -5, have the ability to interfere with MB cell invasion and metastasis-inducing pathways.

Discussion
The proof-of-concept studies described here highlight MB cell growth inhibitory properties of CFMs, a novel class of compounds. Consistent with the ability of CFMs to inhibit growth of the breast and other cancer cells [16], we found that CFMs stimulate apoptosis signaling while suppressing molecules/pathways regulating MB cell cycle and cell growth. Anti-cancer molecules/ compounds that simultaneously block cancer cell growth and survival pathways while activating apoptosis are likely to be superior in effectively suppressing cancer. Consistent with this expectation, our current studies demonstrate that CFM-4 suppressed MB cells growth by activating caspases and inducing apoptosis signaling as well as by diminishing the levels of cell cycle regulatory protein cyclin B1. In addition to stimulating CARP-1 expression, activation of pro-apoptotic SAPKs (p38 and JNK), and various caspases, our current studies demonstrate for the first time that MB cell growth inhibitory signaling activated by CFMs involved up-regulation of NTF3 and biphasic regulation of NF-kB signaling, while key transducers of invasion and metastasis pathways were down-regulated.
A number of published studies have described pleiotropic effects of NGF and various other neurotrophins in neuronal cell types that range from regulation of survival, proliferation and cell death [37,38]. Moreover presence of neurotrophin and their receptors has also been demonstrated in primary MB specimens while presence of TrkC receptors has been shown to correlate with a better response to therapy [6,39]. Growth inhibitory effects of neurotrophins, particularly NGF and NTF3, have been further highlighted in the Trk-negative cells that were engineered to overexpress TrkA or TrkC receptors [21,28,30]. In addition, Proneurotrophins have also been shown to promote neuronal cell death in the presence or absence of Trk receptors. In the cells that lack Trk receptors, the pro-apoptotic signaling by pro-neurotrophins often involve p75NTR and/or sortilin receptors [40]. Our current data in table 1 and figure 5 indicate that levels of both the pro and mature NTF3 were elevated in MB cells following their treatments with CFMs. Since MB cell lines generally do not express functional NTR and Trk receptors [30], and although our WB analysis in conjunction with multiple anti-TrkC antibodies showed presence of TrkC-like peptides in the MB cells, their treatments with purified pro-NTF3 or mature NTF3 not only failed to elicit growth inhibition but also did not enhance inhibitory effects of CFMs (see figures S2 and S3). Interestingly however knock-down of NTF3 interfered with CFM-4-dependent inhibition of Daoy cells ( The neuronal apoptosis signaling by the NTFs in the NTRexpressing cells has been shown to involve activation of the cJun N-terminal kinase (JNK) while NTR proteolysis and subsequent apoptosis was abrogated in the sympathetic neurons of jnk32/2 mice, indicating a requirement for JNK signaling in neuronal apoptosis [41][42][43]. Although CFM-4 has been demonstrated to activate pro-apoptotic p38SAPK in breast cancer cells [16], our current studies indicate that in addition to p38SAPK, CFM-4 also activates JNK in MB cells (see figures 4 and 6). It however appears that a robust activation of p38 occurs earlier than robust JNK activation in both the MB cells ( Figure 6). These data underscore pleiotropic nature of apoptosis signaling by CFMs, and suggest that combined activation of p38 and JNK likely amplify the apoptosis signals in the MB cells independent of the involvement of the cell surface NTR and Trk receptors. Moreover, JNK activation has also been implicated in processing of NTR in the neuronal cells [40]. Since CFM-4 and CFM-5 also enhanced production of mature NTF3 in both the MB cells ( Figure 5), to the extent JNK activation also contributed to CFM-dependent processing of pro-NTF3 independent of NTR or Trk receptors in these MB cells remains to be clarified.
The NF-kB family of proteins and their signaling is well known to play a crucial role in organismal physiology and pathologies such as chronic inflammation and cancer. NF-kB signaling however also antagonizes apoptosis induced by TNF-a due in part to suppression of JNK activation [44]. Consistent with these findings, activation of pro-apoptotic MAPKs (p38 or JNK) serves to attenuate NF-kB activation in different stress-induced apoptotic contexts [45,46]. Our data in figure 6 show that CFM-4 exposure over a periods of upto 6 hours results in a slight to modest decline in levels of NF-kB inhibitory IkBa and/or IkBb proteins. Together with our data indicating a modest increase in NF-kB signaling inhibitory proteins ABIN1 and 2 [31,32] following treatments of MB cells with CFM-4 (figure 6), it is likely that growth inhibitory signaling by CFM-4 impacts the canonical, IKK-dependent NF-kB signaling pathway [47]. Interestingly prolonged exposure to CFM-4 resulted in diminished expression of the NF-kB inhibitory IkBa and/or IkBb proteins ( figure 6). Whether this CFM-4-dependent loss of IkBs is due to degradation of cellular proteins as a consequence of apoptosis induction or a signal for NF-kB activation to antagonize or promote apoptosis is unclear. The activation of NF-kB following prolonged exposure to CFM-4 would nevertheless be consistent with the pro-survival functions of NF-kB signaling to antagonize/overcome the stress of CFM-4 exposure. On the other hand, given that a number of recent studies have revealed a pro-apoptotic role for NF-kB signaling [48][49][50], it is also likely that NF-kB activation following prolonged exposure to CFM-4 serves to potentiate/support apoptosis in the MB cells. Further studies will be necessary to clarify the precise role of NF-kB signaling in pleiotropic cell growth inhibitory effects of CFM-4.
Last but not least, the gene array data together with our observations in figure 7 suggest that CFMs impact the biological properties of migration and invasion by MB cells. Since small GTP-binding protein p21Rac1 is well known to regulate processes of cellular motility, loss of p21Rac1 expression following treatment of MB cells with CFMs would be consistent with their inhibition of MB cell motility and migration. In addition, expression/activation of various MMPs was also impacted by CFMs as indicated by our data in figure 7 and table 1. Given that MMPs are downstream targets of NF-kB signaling and their expression and activation contribute to the NF-kB-dependent survival and metastatic spread of the cancer cells, our findings would further highlight antiinvasion and metastasis-inhibitory properties of CFMs. The precise mechanisms by which CFMs impact p21Rac1 and MMP expression are the subjects of our on-going studies.
In summary, the data presented here convincingly demonstrate that CFMs activate multiple cell growth inhibitory and apoptosis pathways to suppress MB cell growth, survival and metastasis processes, and underscore their potential as novel class of anti-MB agents. Figure S1 Stable knock-down of NTF3 interferes with inhibitory effects of CFMs. (A) MB cells were transfected with vector or NTF3-AS clone 1 plasmids, and stable, neomycinresistant sublines were isolated as in Methods. Cell lysates were analyzed by western blotting for levels of pro-NTF3 and actin proteins as in figure 6. (B, C). The indicated sublines were either untreated (Control) or treated with noted doses of respective CFM for various times. Determination of viable/live cells was carried out by MTT assay as in figure 1