Netrin-1 Induces Apoptosis in Human Cervical Tumor Cells via the TAp73α Tumor Suppressor

Alterations in the molecular integrity and expression of deleted in colorectal Cancer (DCC) and netrin-1 have been reported in several human cancers. In a large fraction of colorectal cancers, loss of heterozygosity (LOH) of chromosome 18q, including the DCC locus, has been described (1). Similarly, homozygous deletions and point mutations have been detected in colorectal, testicular, and pancreatic cancers. Netrin receptors DCC and UNC5 family are down-regulated in over half of all colorectal cancers. In addition, epigenetic inactivation of DCC and UNC5C genes has been described after promoter hypermethylation and gene silencing, suggesting that they function as putative tumor suppressor genes in many tumor types, including human solid tumors and acute lymphoblastic leukemia (2–6). These genetic, epigenetic, and molecular alterations are also found in brain, breast, urogenital, and digestive cancers (7). The tumor suppressive functions of DCC and UNC5H are linked to their ability to induce apoptosis when they are not engaged with their ligands netrins (7, 8). The three secreted netrins-1, netrin-3, and netrin-4 in mammals belong to a conserved family of guidance clues related to the matricellular protein laminin. The netrin signaling pathways are also initiated through the transmembrane receptors neogenin and integrins, and multiple downstream cascades using focal adhesion kinase (FAK), src and Fyn tyrosine kinases, and the Rho-GTPases. Thus, netrins signal via the simultaneous activation of several receptor subtypes and signal transduction pathways involved in critical cellular and molecular mechanisms linked to tumor progression, including cancer cell survival, tumor angiogenesis, invasion, and metastasis (9–14).

Despite intensive efforts to delineate the intracellular mechanisms of netrin signaling in normal and transformed cells, little is known on the pathways underlying the antagonism between DCC and netrin in cancer cell survival. Apoptosis induced by DCC is dependent on caspase activation but independent of the intrinsic mitochondrial and the extrinsic death receptor–mediated pathways (7, 14). However, the possible implication of the p53 family members in the netrin-1 and DCC pathways on these mechanisms has not been explored, until now. The p53 family members include the p63 and 73 tumor suppressor proteins involved in cell cycle regulation, DNA repair, apoptosis, and transcription of cell proliferation and survival genes (15). In addition, structural homologies, as well as molecular interactions and permissive or antagonistic regulations between these wild-type (wt) and mutated proteins are detected in several models (16–18). Similarly to p53, p73 can promote growth arrest or apoptosis when overexpressed in certain p53-null tumor cells. All p53 family members contain the NH₂-terminal transactivation (TA) domain for recruitment of core transcriptional factors, a DNA-binding domain for recognition of promoter sequences, and an oligomerization domain for tetramerization. Similarly to the TP63 gene, the TP73 encodes several NH₂- and COOH-terminal isoforms with opposing functions and is rarely mutated in human tumors (15). These additional motifs found in both p63 and 73, but not in p53, are created by a combination of two alternative promoters, leading to the synthesis of NH₂-terminal isoforms (TAp73 and ΔNp73), and alternative splicing variants at the COOH-terminal exons 10 to 14. TAp73α is transactivation active and is proapoptotic, whereas ΔNp73α is not. The COOH-terminal isoforms TAp73α are different according to the sterile α motif (SAM) domain and the transcription inhibitory domain (17, 19). The NH₂ terminally truncated ΔNp73α proteins interfere with the transactivation function of TAp73α and p53 in a dominant-negative manner, inhibiting apoptosis (20, 21). Moreover, ΔNp73α is up-regulated in several cancers and is therefore associated with advanced pathologic states and poor prognosis features in human colon and breast cancers and in patients suffering from neuroblastoma (19, 22). Because the TP53 tumor suppressor gene is frequently lost and invalidated by LOH and mutational alterations in human solid tumors, we decided to focus on the two p53 homologues p73 and p63, rarely mutated in cancer, investigating their potential role in the netrin-1 and DCC antagonism on cell viability and apoptosis. For this purpose, we examined the relationships between netrin-1, cancer cell survival, and the p73 and p63 status in the p53-deficient HeLa and HEK-293 cell lines. Indeed, viral oncoproteins such adenovirus 5 E1B in HEK-293 cells, human papillomavirus E6 in HeLa cells, as well as SV40 large T, inactivate p53 by protein-protein interaction but do not seem to interact with p73 (23).

Cell culture conditions, expression vectors, and transfections. Human cervical and kidney cell lines HeLa and HEK-293 cells were cultured at 37°C in DMEM without antibiotics (Invitrogen) supplemented with 10% FCS (PAA Laboratories). Ectopic expression of netrin-1 and DCC was achieved by transient transfection of HeLa and HEK-293 cells (2.5 × 10⁵) using either the pcDNA3.1-Netrin-1 vector encoding myc-tagged chicken netrin, or the pCMV-DCC vector encoding the full-length human DCC, and their combination, as described (14). The pcDNA3 vector encoding E2F1 (24) was a generous gift from Dr. C. Prives, Columbia University, New York, NY. Empty pcDNA3.1 vector was used to keep the amount of DNA constant. Cells were transiently transfected for 6 h at 70% to 90% confluency, using the plasmid DNA (up to 2 μg) mixed with the Lipofectamine 2000 reagent (Invitrogen) at the DNA (μg)/lipid (μL) ratio of 1:2.5. Transfected cells were then cultured in fresh medium for up to 24 to 48 h and harvested for transgene expression and other assays. Silencing of p73 was performed with SureSilencing p73-shRNA plasmids (Super Array-Bioscience Corporation) containing 4 pooled shRNA sequences to ensure an effective depletion of the six p73 isoforms in HeLa and HEK-293 cells. Each vector expresses a short hairpin RNA (shRNA), under the control of the U1 promoter. The shRNA-p73 sequences are as follows: (+1474) 5′-TGTCCAAACTGCATCGAGTAT-3′; (+655) 5′-GGCAATAATCTCTCGCAGTAT-3′; (+750) 5′-CACCATCCTGTACAACTTCAT-3′; (+178) 5′-TCTGTCATGGCCCAGTTCAAT-3′, and GGAATCTCATTCGATGCATAC for the scrambled control sequence (sh-CON). For stable transfection, HeLa cells were selected for 2 wk in 1 mg/mL G418-containing medium. Two control clones (sh-CON, clones 1 and 3) and two clones of p73-silenced HeLa cells (shRNA-p73, clones 1 and 2) were then further selected and characterized.

Semiquantitative and real-time quantitative reverse transcription PCR analysis. Total RNA was extracted from HeLa and HEK293 cells using the RNAeasy Mini kit (Qiagen). For semiquantitative and qRT-PCR, total RNA (2–2.5 μg) was reverse transcribed using the Superscript III First-Strand Synthesis kit (Invitrogen). The relative levels of the β-actin and p73α isoforms transcripts were evaluated using a denaturation step at 95°C for 5 min, followed by 30 cycles at 95°C for 30 s. Primers were as follows: forward 5′-ACCAGACAGCACCTACTTCG-3′, reverse 5′-TCGAAGGTGGAGCTGGGTTG-3′ (TAp73α, amplicon: 252 bp); forward 5′-AAGCGAAAATGCCAACAAAC-3′, reverse 5′-CACCGACGTACAGCATGGTA-3′ (ΔN-p73α, 213 bp); forward 5′-AACGTCACGCTCACACTGTC-3′, reverse 5′-GAGTCGTCCTCGTTCTCGTC-3′ (Netrin-1, 404 bp); forward 5′-AGAAAATCTGGCACCACACC-3′, reverse 5′-CCATCTCTTGCTCGAAGTCC-3′ (β-actin, 434 bp). The cDNAs used in quantitative PCR were synthesized from 2 μg total RNA extracted from 2 independent experiments using the netrin-1, DCC, and E2F1 expression vectors. The real-time PCR reactions typically contained 1 μL of each gene-specific primer (final concentration, 0.5 μmol/L), 10 mL SYBR Green PCR Master Mix (2×; from Roche Applied Biosystems), and 2 mL of the diluted cDNAs (1:10) in a total volume of 20 mL. Primers for qRT-PCR analysis were as follows: TAp73α, 106 bp: forward 5′ CCTCTGGAGCTCTCTGGAAC 3′, reverse 5′ GAAGACGTCCATGCTGGAAT 3′; ΔNp73α, bp107: forward 5′ CAGCCAGTTGACAGAACTAAGG 3′, reverse 5′ AGAGGCTCCGCAGCTAGTGA 3′; and Actin, 90 bp: forward 5′ GGATGCAGAAGGAGATCACTG 3′, reverse 5′ CGATCCACACGGAGTACTTG 3′. The quantitative PCR reactions were performed in duplicate using a LightCycler 480 Real-Time PCR Systems (Roche Applied Science). The thermal cycling conditions included an initial denaturation step at 95°C for 10 min, 40 cycles at 95°C for 15 s, and 63°C for 1 min.

Western blot analysis. Immunoblot analysis was performed essentially as described (14). The blots were probed for 1 h at room temperature with one of the after primary monoclonal (mAb) or polyclonal (pAb) antibodies: the pAb directed against the SAM domain SAM of human p73α (1:2,000 dilution) and p63α (1:1,000) as described (25); the pAb directed against phosphorylated p73α at Y-99 (1:1,000; Genescript Corporation); the pAbs against β-actin (1:1,000), human p63α (1:500), DCC (1:500), Yap-1 (1:250), and Bax (1:250) are from Santa Cruz Biotechnology; the mAbs against cleaved caspase-3 (1:500) and chicken netrin-1 (1:250) are from Cell Signaling and R&D Systems, respectively; the Itch pAb (1:500) and the mAbs against Bcl-2 (1:500) were from BD Transduction Laboratories. Membranes were reprobed for 1 h with the appropriate peroxidase-conjugated secondary antibodies in blocking buffer (1:10,000; Santa Cruz Biotechnology). Signals were visualized with the enhanced chemiluminescence Western blotting detection system (GE Healthcare), and quantified by densitometric analysis, using ImageQuant software and digital scanning (Agfa).

3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, apoptosis, and fluorescence-activated cell sorting analyses. HeLa and HEK-293 cells (2 × 10⁴ cells per well) transiently transfected for 24 h by the control vector and the netrin-1 and DCC expression vectors were treated for 48 h with either 10 μmol/L cisplatin or 0.6 μmol/L doxorubicin. Then, cultured cells were subjected to the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay (Sigma). Cell cycle phase distribution and apoptosis in sub-G₁ fractions were analyzed by flow cytometry of propidium iodide–labeled DNA content profiles, using a Becton Dickinson FACSCalibur. Labeling DNA breaks by the terminal deoxynucleotidyltransferase-mediated dUTP nick-end labeling (TUNEL) assay was performed using the Apo-BrdUrd kit. All assays were performed in triplicates for each condition. Results are presented as means ± SE. Statistical analysis was performed using the Student's t test.

TAp73α ubiquitination assays and stability. HEK-293 cells were transiently cotransfected with expression vectors encoding TA-p73α (26) and His₆-c-Myc–tagged human ubiquitin (27), in the presence of the control, netrin-1, and DCC expression vectors. Transfected cells were cultured for 32 h in standard culture conditions, and then treated for 8 h with the proteasome inhibitor MG132 before harvesting (10 μmol/L; Stressgen). Cell extracts were prepared in NP40 lysis buffer and Myc-ubiquitin–conjugated proteins were separated by overnight incubation at 4°C, using Ni-NTA-agarose beads (Qiagen). The beads were washed thrice with the NP40 lysis buffer, and the nickel-binding proteins were resuspended in 1× Laemmli sample buffer, denatured for 5 min at 95°C, and loaded onto 10% polyacrylamide-SDS gels before being subjected to Western blotting with the p73α anti-SAM domain pAb (25). For the determination of TA-73α protein stability, HEK-293 cells were cotransfected for 24 h with a 1:5 ratio for the TAp73α/pcDNA3 and TAp73α/Netrin-1 plasmids (3 μg total DNA). Cells were then treated with 50 μg/mL cycloheximide (Sigma), and whole cell lysates were processed at the indicated time points for immunoblot analysis. The relative amount of TA-p73α protein was evaluated by densitometry and normalized to β-actin signals.

Netrin-1 affects cell death and survival. To examine the role of netrin-1 on cell viability and cell cycle progression, we transfected p53-deficient HeLa cancer cells with expression vectors encoding either netrin-1, wt-DCC, and their combination. We noted that netrin-1 or DCC expression reduced HeLa cell viability by 30% (Fig. 1A). This cellular response was attenuated by ectopic expression of both DCC and netrin-1, as expected. Our results suggested that netrin-1 reduces cancer cell survival while counteracting the established proapoptotic functions of the DCC tumor suppressor. To further document this observation, we next examined the role of netrin-1 on cell cycle progression and apoptosis by fluorescence-activated cell sorting (FACS) analysis. Both netrin-1 and DCC induced, respectively, a 2.6- and 2.2-fold increase of the apoptotic cell fraction in cultured HeLa cells (sub-G₁ fraction) without any change in the distribution of HeLa cells at the successive transitions of the cell cycle (Fig. 1B). As observed in the MTT assay, simultaneous expression of both netrin-1 and DCC neutralized their respective deleterious functions on HeLa cell apoptosis. We subsequently confirmed these data using the TUNEL assay showing that netrin-1 increased the percentage of apoptotic cells 4-fold, DCC 6-fold, and their combination, only 2-fold (Fig. 1C).

Induction of TAp73α protein by netrin-1 expression. We next evaluated whether the p53-related proteins p73α and p63 were involved in the apoptosis induced by netrin-1 in the p53-deficient cell lines HeLa and HEK-293. First, we investigated the effect of netrin-1 ectopic expression on the accumulation of their transcripts and protein isoforms. As shown in Fig. 2A, netrin-1 selectively induced a robust accumulation of the TAp73α protein in HeLa cells (73 KD; 2.6 ± 0.06-fold increase; n = 3), without any significant change in the transcript levels of TAp73α and ΔNp73α (Fig. 2C) by semiquantitative (left) and real-time reverse transcription-PCR (RT-PCR; right). Our data indicate that netrin-1 is acting on TAp73α expression through posttranscriptional mechanisms. Consistently, ectopic netrin-1 failed to up-regulate ΔNp73α and p63α proteins in HeLa cells (Fig. 2A). In keeping with our data on cell survival and apoptosis in HeLa cells (Fig. 1), netrin-1 down-regulated the anti-apoptotic protein Bcl-2 (5.3-fold) and up-regulated the proapoptotic protein Bax (1.4-fold) in HEK-293 cells (Fig. 2A), resulting in a decreased Bcl-2/Bax ratio, an index of the apoptosis signature. This conclusion was further validated by the detection of the cleaved caspase-3 protein in netrin-1–transfected HEK-293 cells. As shown in Fig. 2B, the induction of the TAp73α protein by netrin-1 was confirmed and extended in the p53-deficient HEK-293 cell line (3 ± 0.7-fold increase; n = 4 experiments).

Implication of TAp73α in the apoptosis induced by netrin-1. We sought to determine whether TAp73α expression was essential in the netrin-1 apoptotic function. Expression of the TAp73α isoform was inhibited by transient transfection of HeLa cells by shRNA, and its reduced protein expression was tested by immunoblot analysis. As shown in Fig. 3A, the p73 shRNA reduced by 29% and 70% endogenous and netrin-1–induced TAp73α levels in HeLa cells, respectively (n = 3 experiments), whereas the control scrambled RNA vector had no interference on the induction of TAp73α by netrin-1 in sh-CON cells (5.5 ± 0.6-fold increase). Results obtained in HeLa cells were confirmed in HEK-293 cells transfected by the netrin-1 vector in the presence and absence of the p73 shRNAs versus the scrambled sequences (data not shown). Most importantly, the p73 shRNAs reversed the induction of the p73α protein by netrin-1 to control levels (scrambled) and was able to reduce by 65% ± 3.3% the netrin-1–dependent apoptosis in HeLa cells (Fig. 3A). Our data imply that netrin-1 mediates apoptosis through induction of the TAp73α proapoptotic form.

Role of the netrin-1 receptor DCC in the induction of TAp73α by netrin-1. We next examined the molecular status of the DCC tumor suppressor gene in HeLa cells. Using combined PCR and sequencing methods to amplify and characterize the molecular integrity of the DCC gene in this model, we detected an internal deletion of six bp within exon 26, leading to the RTV sequence in HeLa cells instead of RSQV in the wt-DCC protein sequence (Supplementary Fig. S1). Such a molecular alteration in the DCC gene has been previously reported in the IMR-32 neuroblastoma cell line (28). Taking into account the DCC molecular status in our model, we next explored the effect of wt-DCC on p73α protein levels after transient transfection of HeLa cells for 24 and 48 hours with the netrin-1 and DCC expression vectors. As shown in Fig. 3B, reconstitution with wt-DCC in HeLa cells failed to elevate endogenous p73α levels but instead decreased by 15% and 32% the p73α levels induced by netrin-1 at the 24- and 48-hour time points. Of note, loss of the DCC signal observed at 48 hours in the presence of netrin-1 is associated with amplification of the TAp73α signal. As expected, simultaneous expression of netrin-1 with wt-DCC decreased and abolished DCC levels in HeLa cells at 24 and 48 hours, respectively, in agreement with previous reports showing that DCC is subjected to ubiquitin-dependent degradation after interaction with its ligand, netrin-1 (29). Consistently, the stability of the ectopic wt-DCC protein was greatly enhanced by MG132 in HeLa cells (Fig. 3C). In contrast, endogenous DCC was not detected in the presence and absence of this proteasome inhibitor, suggesting that the resident DCC protein is either (a) expressed at very low levels, (b) is expressed under a mutated/truncated form, as shown in HeLa cells, or (c) is not expressed at all, due to other mutations or truncations in coding/noncoding regions of the DCC gene. Because introduction of wt-DCC in HeLa and HEK-293 cells induced apoptosis, one can postulate that endogenous DCC is not functioning as a dominant-negative form of the transfected wt-DCC form. It is therefore likely that the endogenous DCC pathway is inactivated in these two models.

Implication of the ubiquitin-proteasome pathway in the induction of p73α by netrin-1. Because p73α stability is regulated by the ubiquitin-proteasome pathway (30), we next examined whether the proteasome inhibitor MG132 affected TAp73α induced by netrin-1 in the presence and absence of wt-DCC. As shown in Fig. 3C, MG132 strongly potentiated TAp73α levels induced by netrin-1 in the presence and absence of the wt-DCC vector. In contrast, wt-DCC was ineffective in the induction of the TAp73α protein in HeLa cells incubated in the presence or absence of the proteasome inhibitor. DCC decreased netrin-1–induced TAp73α levels under both conditions, in spite of the depletion of wt-DCC protein induced by netrin-1. We found that netrin-1 had no effect on the expression levels of the p73 ubiquitin ligase Itch and p73 modular adapter protein Yap-1 in HeLa cells (data not shown). Both Yap-1 and Itch bind the same PPPY motif of p73, resulting in stabilization of p73 by Yap-1 and prevention of Itch-mediated ubiquitination of p73 (31). Our data favor the hypothesis that netrin-1 targets the activation status of p73 ubiquitin/deubiquitin ligases.

We therefore sought to determine the effect of netrin-1 and DCC on the ubiquitination levels of p73α after cotransfection of HEK-293 cells with expression vectors encoding His₆-c-Myc–tagged human ubiquitin, β-galactosidase, and TAp73α (Fig. 4A). Quantification of TAp73α ubiquitination (bottom) was performed by densitometry analysis of the wt-TAp73α protein (arrow) and its cleaved forms, both normalized to the relative amounts of total TAp73α detected by direct Western blot (top). In two separate experiments, we observed that netrin-1 reduced by 62% and 63% the ubiquitination levels of the full-length and cleaved forms of the TAp73α protein in HEK-293 cells. Conversely, wt-DCC increased by 15% and 61% the ubiquitination levels of the cleaved and wt forms of TAp73α, respectively. Upon cotransfection with netrin-1 + DCC, these two ubiquinated bands were respectively reduced by 61% and 22%. We next examined the turnover of TAp73α in the presence and absence of ectopically expressed Netrin-1 (Fig. 4B). Twenty four hours posttranfection of HEK-293 cells with TAp73α and Netrin-1 or control vectors, these cells were treated with the translation inhibitor cycloheximide for 0, 1, 2, 4, and 6 hours. TAp73α protein levels were monitored by Western blot and normalized to β-actin. As shown in Fig. 4B, netrin-1 clearly stabilizes TAp73α protein levels leading to a significant increase in protein half-life.

Netrin-1 and DCC selectively potentiate alterations of cell viability induced by cisplatin. Given that TAp73α is recognized as an important mediator of the apoptosis induced by cisplatin, we next tested whether netrin-1 and DCC interfere with alterations of HeLa cell viability induced by cisplatin and doxorubicin, two chemotherapeutic drugs with different modes of action. As a topoisomerase type II inhibitor and intercalating agent, the anthracycline antibiotic doxorubicin induces breaks in the genomic DNA and blocks DNA synthesis and transcription. The cytotoxicity of the DNA-damaging agent cisplatin is achieved in part through inhibition of DNA replication and transcription, and induction of the DNA damage response signaling pathways involved in the transduction of genotoxic stress signals to the p53 and p73 regulators of cell cycle arrest and apoptosis. First, we established the inhibitory potency of cisplatin and doxorubicin in control HeLa cells (Fig. 5, top). We observed that 10 μmol/L cisplatin and 0.6 μmol/L doxorubicin induced a reduction of ∼50% in HeLa cell viability. Both netrin-1 and DCC ectopic expression increased the cytotoxicity of cisplatin (Fig. 5, bottom), compared with control vector-transfected cells (P < 0.001). It is noteworthy that when HeLa cells were transfected with the same netrin-1 and DCC-encoding vectors, we observed no change in cell viability over the doxorubicin-only treatment.

Effect of external addition of the netrin-1 peptide on TAp73α levels and HeLa cell viability. Further proof of the role of netrin-1 on TAp73α expression and cell viability was provided from HeLa cells stably transfected by expression vectors encoding either p73 silencing shRNAs (sh-RNA-p73) or control sh-scrambled sequences (sh-CON). The effect of external addition of the netrin-1 peptide was investigated on TAp73α protein and the MTT test. As shown in Fig. 6A, the robust induction of p73α protein levels by cisplatin, observed in the sh-CON HeLa cells, was alleviated by the p73 shRNAs, as expected from the validation of the silenced versus the scrambled control clones. In agreement, p73α silencing exacerbates the growth of HeLa tumor xenografts in nude mice, with no effect on the invasive potential of HeLa cells in collagen type I gels (Fig. 6B) and in subcutaneous tumors (Supplementary Fig. S2B). In both groups, ∼75% of tumor xenografts were found invasive. The percentage of Ki-67–positive tumors in p73-silenced cells (26.8%) was significantly higher than the percentage in the corresponding control xenografts (7.6%; P < 0.01), whereas the number of TUNEL-positive cells was higher in control versus p73-silenced cells (respectively, 22.4% and 3.1%; P < 0.01). Consistently, stable silencing of p73α resulted in remarkable induction of the antiapoptotic protein Bcl-2 in sh-RNA-p73 HeLa cells (data not shown). Taken together, our data support the functional validation of the p73α-silenced HeLa cells in the context of cancer cell proliferation and apoptosis in vitro and in vivo. As shown in Fig. 6C, the netrin-1 peptide and cisplatin induced a robust accumulation of TAp73α and compromised HeLa cell viability to similar extent in sh-CON HeLa cells but not in silenced p73-shRNA cells.

The implication of the p53, p63, and p73 family proteins on gene transcription and regulation of several effectors involved in cell cycle control, tumor progression, cancer cell survival, and apoptosis has been largely documented in the literature. In the present study, we bring original data on the conditional regulation of apoptosis induced by netrin-1 and DCC in p53-deficient HeLa and HEK-293 cells. The pertinence of these two models for our study and objectives is validated by the observation that the TP53 tumor suppressor gene is commonly lost or mutated in human solid tumors. Deficiency of p53 disrupts anticancer drug–induced apoptosis and promotes resistance to chemotherapy and a more malignant phenotype. In contrast, the p73 and p63 proteins are rarely mutated in human tumors and are mainly regulated via complex transcriptional and posttranscriptional mechanisms. Because extensive crosstalk among the p53 family members and their isoforms has been reported, our approach therefore facilitates the interpretation of our data under a relevant context of carcinogenesis.

We showed that netrin-1 exerts two divergent functions on cell survival by inducing apoptosis in the absence of DCC, and blocking the apoptosis induced by the DCC tumor suppressor protein, according to the dependence receptor hypothesis. Our data indicate that the cell death pathways induced by netrin-1 and DCC are mutually exclusive. Netrin-1 induced convergent deleterious signals against cancer cell viability and survival as evidenced by the MTT test and determination of apoptosis in the sub-G₁ fraction by FACS analysis and TUNEL assay. Netrin-1 induced caspase-3 cleavage and decreased the Bcl-2/Bax ratio, underlying its capacity to induce apoptosis through the mitochondrial pathway. Phenotypic effects of netrin-1 on viability and apoptosis were associated with the selective induction of the long TAp73α protein isoform in HeLa and HEK-293 cells, suggesting that netrin-1 is a potential regulator of apoptosis and gene expression via TAp73α. Indeed, the TAp73α isoform exhibits proapoptotic functions and transactivates several p53 target genes involved in cell cycle arrest and apoptosis. In our study, however, the induction of TAp73α protein by netrin-1 did not correlate with any change in the distribution of HeLa cells within the different phases of the cell cycle, suggesting a selective effect of netrin-1 on survival in cancer cells with compromised p53. In addition, no induction of the ΔNp73α short isoform was seen by qRT-PCR and immunoblotting in netrin-1–expressing cells. The ΔNp73α short isoform is associated with cell survival, and functions as a dominant negative competitor of TAp73α, due to inhibitory hetero-oligomer formation (32). Consistently, the ΔN isoforms of p73 and p63 are frequently overexpressed in cancer. However, no effect on p63 protein levels was detected after netrin-1 expression, thus showing selective netrin-1 regulation at the p73 signaling network. This observation is compatible with the notion that p63 regulates progenitor cell populations and morphogenesis at the proliferation-differentiation interface in the epithelial and mesenchymal cell lineages, as shown in p63 loss-of-function mouse models (33). Knockdown of p63 expression by shRNA in mammary epithelial cells caused both cell detachment and anoïkis associated with down-regulation of cell adhesion-associated genes encoding β1, β4, and α6 integrins, as well as laminin-γ2 and fibronectin (34). Interestingly, both α6β4 and α3β1 integrins are now considered as functional netrin-1 receptors (35).

In this report, we have shown that netrin-1 selectively up-regulates TAp73α levels through posttranscriptional mechanisms, including inhibition of p73α ubiquitination, stabilization of the p73α protein, with no change in the expression levels of the p73α ubiquitination effectors Itch and Yap-1 (31, 36). In agreement, the induction of the TAp73α protein levels by netrin-1 synergizes with the proteasome inhibitor MG132. Of note, netrin-1 was ineffective on the Tyr-99 phosphorylation levels of TAp73α targeted by the abl tyrosine kinase (data not shown), and involved in the stability of p73. Implication of cyclin-Cdk complexes, acetylation-dependent mechanisms, and ubiquitin-independent pathways using the 20S proteasome were also reported for regulating p73 stability and degradation (37–39). Another possibility is that netrin-1 regulates the translation of the p73α mRNA as recently illustrated by the elegant study of Tsai showing that netrin-1 alleviated the Grb7-mediated translational repression of the κ opioid receptor mRNA via a netrin-1/integrin/FAK and Grb-7 pathway (40). Other key signaling molecules involved in cancer cell survival and progression, including DCC and src, are potentially involved in this FAK molecular scaffold (41). Consistent with this hypothesis, our data indicate that the cell death factor DCC is not competent to induce TAp73α protein levels but conversely counteracts netrin-induced TAp73α protein accumulation. Another divergent mechanism underlying the apoptosis induced by netrin-1 and DCC is illustrated by the observation that DCC is highly efficient in TAp73α ubiquitination. Additional mechanisms may include the participation of microRNAs in regulating apoptosis sensitivity to netrin-1, DCC, and TAp73α (42).

The implication of the TAp73α signal on the apoptosis induced by netrin-1 was shown by transient and stable silencing the endogenous p73α protein levels. Our data therefore indicate that TAp73α is a relevant mediator of the apoptosis induced by netrin-1. Identification of the netrin signaling effectors and target genes induced via TAp73α should provide more information on the signaling networks involved in the apoptosis induced by netrin-1. This question is sustained by the positive cooperativity we have observed between netrin-1/DCC on the cisplatin cytotoxicity in HeLa cells. Recent advances in the field indicate that TAp73α increases the sensitivity to chemotherapeutic drugs and oxidative stress and is involved in the limitation of anchorage-independent growth (43).

In the present study, we found that netrin-1 induced apoptosis in DCC-deficient HeLa cells. Emerging evidence underline the molecular and functional diversity of the netrin ligands, receptors and coreceptors (DCC-adenosine A2B, UNC5A to 5D, neogenin, integrins/FAK, as well as unidentified netrin-1 receptors), and their interplay or antagonism regarding angiogenesis, apoptosis, and survival in cancer cells (35, 44–46). Another level of complexity is illustrated by the formation of homodimers and heterodimers between the cell death family netrin receptors, as shown during axon attraction and repulsion (47). We have shown that DCC-deficient human colon cancer cells and tumors retain UNC5 netrin receptors A to B and C (14). In contrast, expression of the netrin-1 gene NTN1 is reduced or absent in brain and prostate cancers (48, 49), suggesting that cell death signals induced by netrin-1 can be abrogated during oncogenesis. Divergent biological roles of netrin-1 in attraction, retraction or repulsion of neuronal, endothelial, and epithelial cells are also reported according to the cellular and signaling context (12, 46, 50). Therefore, different known or unknown netrin receptors can be sequentially and reciprocally activated, desensitized/internalized, or degraded. Recent data indicate that the dependence receptors UNC5A regulate naturally occurring apoptosis independently of netrin-1 and, therefore, do not provide support for the dependence ligand hypothesis (45). Netrins are secreted and diffusible molecules that are stored together with several growth factors and matricellular proteins at the vicinity of cancer cells and tumor stromal cells. These paracrine and autocrine factors have potential long-lasting effect on the survival and spreading of cancer cells in primary human solid tumors and their metastases, including the modulation of cell death and DNA damage signals induced during cancer therapeutics.

No potential conflicts of interest were disclosed.

Grant support: Institut National de la Sante et de la Recherche Medicale, ARC n°3765, IPSEN, the Fund for Scientific Research Flanders (Brussels, Belgium), and the Scientific Exchange Program between the Flemish community and France (Grant I. 2007.03).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.