Ectopic Expression of p73α, but Not p73β, Suppresses Myogenic Differentiation*

The TRP73 gene, a member of the p53 family, encodes several variants through differential splicing and use of alternative promoters. At the N terminus, two different promoters generate the full-length and the ΔN isoforms, with or without the transactivating domain. At the C terminus, seven isoforms generated through alternative splicing have been cloned. Previous studies have demonstrated that ΔN-p73 interferes with p73-induced apoptosis. However, there has been no evidence for functional diversity of the C-terminal p73 variants. In this study, we found that p73α and p73β exerted differential effect on the differentiation of C2C12 myoblasts. Although p73β lacked any detectable effect on differentiation, p73α caused a substantial delay in the expression of muscle-specific genes. In co-transfection experiments p73α, but not p73β, attenuated the transcriptional activity of MyoD. Microarray-based gene profiling confirmed the protraction of MyoD-dependent gene expression in C2C12 cells stably expressing p73α. Notwithstanding the differential effect on differentiation, p73α and p73β showed similar activity in sensitizing C2C12 myoblasts to cisplatin-induced cell death. These results demonstrated a functional diversity between the two C-terminal variants of p73 and suggested that p73α can regulate cellular differentiation in addition to its role in stimulating cell death.

The TRP73 gene, a member of the p53 family, encodes several variants through differential splicing and use of alternative promoters. At the N terminus, two different promoters generate the full-length and the ⌬N isoforms, with or without the transactivating domain. At the C terminus, seven isoforms generated through alternative splicing have been cloned. Previous studies have demonstrated that ⌬N-p73 interferes with p73-induced apoptosis. However, there has been no evidence for functional diversity of the C-terminal p73 variants. In this study, we found that p73␣ and p73␤ exerted differential effect on the differentiation of C2C12 myoblasts. Although p73␤ lacked any detectable effect on differentiation, p73␣ caused a substantial delay in the expression of muscle-specific genes. In co-transfection experiments p73␣, but not p73␤, attenuated the transcriptional activity of MyoD. Microarray-based gene profiling confirmed the protraction of MyoD-dependent gene expression in C2C12 cells stably expressing p73␣. Notwithstanding the differential effect on differentiation, p73␣ and p73␤ showed similar activity in sensitizing C2C12 myoblasts to cisplatin-induced cell death. These results demonstrated a functional diversity between the two C-terminal variants of p73 and suggested that p73␣ can regulate cellular differentiation in addition to its role in stimulating cell death.
The TRP73 gene, a member of the p53 gene family, encodes several variants of the p73 protein. At the 5Ј-end of the gene, two alternative promoters drive the expression of N-terminal variants of p73, one of which encodes a transactivation (TA) 1 domain, whereas the other (⌬N) lacks the TA domain. At the 3Ј-end, alternative splicing gives rise to seven different isoforms (denominated ␣, ␤, ␥, ␦, ⑀, , and ) with disparate C-terminal sequences (1,2). Among these splicing variants, p73␣ contains the longest C-terminal sequence, and it is the predominant p73 isoform expressed in cultured human cell lines (3)(4)(5)(6).
The p73 gene products regulate apoptosis (7)(8)(9). Disruption of the mouse p73 gene caused resistance to cell death initiated by E2F-1 or induced by TNF␣ (10 -12). Current data suggest that p73 can bind to a number of p53-regulated promoters, because all p73 variants contain DNA binding and oligomerization domains that share homology with p53 (2-4, 13, 14). The TA-p73 variants can up-regulate the expression of PUMA, a BH3-only member of the Bcl-2 family, to activate mitochondriadependent apoptosis (15). TA-p73, similar to p53, can also up-regulate BAX, which collaborates with PUMA to trigger cytochrome c release from mitochondria (16,17). Other proapoptotic genes up-regulated by TA-p73 include p53AIP1, NOXA, IGF-BP3, and PERP, which are also activated by p53 (16,18). In addition, a set of genes regulated by p73, but not p53, has also been identified (16,19). The ⌬N-p73 variants that lack a TA domain antagonize TA-p73 by blocking the expression of pro-apoptotic genes. Survival of sympathetic neurons depends on the expression of ⌬N-p73, which is maintained by nerve growth factor (20). The p73 Ϫ/Ϫ mice lacking all p73 splice variants have half as many cortical neurons as their wild type littermates (21). Ectopic expression of ⌬N-p73 protects cortical neuron from diverse apoptotic stimuli (21). The ⌬N-p73 promoter can be up-regulated by either TA-p73 or p53 in a negative feedback loop to mitigate the apoptotic function of TA-p73 and p53 (22). Taken together, the N-terminal variants of p73 have opposing activities in the regulation of cell death.
The TRP73 gene function is not limited to apoptosis. Evidence has accumulated to suggest an involvement of p73 in differentiation. Mice deficient for p73 display pleiotropic phenotypes including hippocampal dysgenesis, hydrocephalus, chronic nasal inflammation, and defective pheromone sensing (23). Ectopic expression of TA-p73 proteins promotes differentiation of neuroblastoma cells in the absence of retinoic acid (24). With explanted neuronal precursor cells, ectopic expression of TA-p73␣ facilitated differentiation toward neuronal myelin-producing oligodendrocytes in response to thyroid hormone, retinoic acid, or the withdrawal of platelet-derived growth factor (25). In the same experimental system, expression of ⌬N-p73␣ severely restrained differentiation (25). Furthermore, the expression of p73 gene products appears to be regulated during differentiation. Cultured neuronal or hematopoietic progenitor cells do not express p73 until the onset of differentiation (24,26). During zebra fish development, TA-p73 is the predominant isoform and its expression is restricted to subsets of cells in olfactory system, telencephalon, dorsal diencephalon, and pronephric ducts (27). The TA-p73␣ expression is also detected in differentiating slow muscle cells of somites and pharyngeal endoderm (27). These observations indicate that p73 may participate in the regulation of development.
A previous study has found that TRP73 gene expression is suppressed in C2C12 myoblasts (28). To determine whether this suppression of p73 is relevant to the differentiating potential of C2C12 cells, we examined the consequence of ectopically expressing p73 on myogenic differentiation of C2C12 myoblasts. The two major splice variants of TA-p73, p73␣ and p73␤, were expressed transiently or stably in C2C12 myoblasts, and their effects on differentiation were examined. We found that p73␣, but not p73␤, suppressed myogenic differentiation, but both isoforms promoted apoptotic response of C2C12 cells to genotoxic stress.

MATERIALS AND METHODS
Cell Lines, Antibodies, and Plasmid Constructs-Murine C2C12 myoblasts were purchased from ATCC and routinely cultured in Dulbecco's modified Eagle's medium plus 20% fetal bovine serum. For myogenic differentiation, the cells were changed into Dulbecco's modified Eagle's medium containing 5% horse serum and 50 ng/ml IGF-1 (Sigma). A monoclonal anti-HA antibody was purchased from BABCO. Anti-MHC antibody was procured from Sigma. Anti-MyoD, anti-myogenin, anti-p21Cip1, and anti-caspase 3 antibodies all were obtained from Santa Cruz Biotechnology. Human p73 cDNAs were from Dr. Gerry Melino of University of Rome, Italy. Human p73␣ and p73␤ were HA-tagged at the N terminus in pcDNA3 mammalian expression vector (Invitrogen). Murine MyoD cDNA was cloned in pCMV vector. Gal4-MyoD, p4RE-Luc, and pGal-Luc plasmids (29,30) were from Dr. Lorenzo Puri at the Burnham Institute, La Jolla, CA and the Telethon Dulbecco Institute, Italy.
Luciferase Reporter Assay-Luciferase activity was measured with Promega's dual luciferase assay system according to the manufacturer's instructions.
Immunofluorescence Staining-Transfected cells were fixed with 3% formaldehyde in PBS for 15 min and permeabilized with 0.1% Triton X/PBS for 5 min at room temperature. The coverslips were blocked with 10% goat serum in 0.1% Triton X/PBS for 30 min. Primary and secondary antibodies were sequentially incubated with the cells in 0.1% Triton X/PBS for 1 h. Coverslips were finally washed with 0.1% Triton X/PBS, mounted on slides, and examined under an Axioplan microscope (Carl Zeiss). Images were captured and analyzed with ImagePro-PLUS (Media Cybernetics). Statistical analysis was performed with GraphPad Prism 4 software.
Gene Expression Profiling with Affymetrix DNA Microarrays-Affymetrix GeneChip mouse oligonucleotide arrays (MG_U74AV2) were employed to profile gene expression associated with myogenic differentiation of C2C12 myoblasts. Total RNAs were isolated with Trizol (Invitrogen) from C2C12 cells stably expressing p73␣ or p73␤ cultured in differentiation media for 24 h, as well as from vector-mocked C2C12 cells cultured in regular growth media. Total RNA preparations were further purified with RNeasy columns (Qiagen). Double-stranded cDNAs were made with a T7 promoter-oligo(dT) primer priming the synthesis of the first strand. Biotinylated cRNA probes were synthesized off double-stranded cDNA as template with an in vitro transcription kit involving T7 RNA polymerase and biotin-dUTP. The labeled cRNA was fragmented to 35-200 base fragments by metal-induced hydrolysis in a fragmentation buffer containing Tris acetate, pH 8.1, potassium acetate, and magnesium acetate. About 50 g of fragmented cRNA probe was used to hybridize an Affymetrix GeneChip at 45°C in a GeneChip hybridization oven. Staining was done in a three-step procedure starting with a streptavidin-phycoerythrin staining solution, followed by incubation with biotinylated anti-streptavidin, and finally a second staining with streptavidin-phycoerythrin. Stained arrays were scanned with a GeneArray scanner from Affymetrix. Analyses of data were performed using the Affymetrix software microarray suite Gene-Spring from Silicon Genetics and various freeware analysis packages such as D-Chip, Cluster, and Treeview.
Cell Death Assay-C2C12 myoblasts stably expressing p73␣ or p73␤ were treated with 25 M cisplatin (CCDP) for 24 h. Thereafter, death of these cells was measured by trypan blue exclusion. Lysates of these cells were examined with an antibody recognizing either pro-caspase 3 or cleaved caspase 3 on Western blots. A, immunofluorescence staining for MHC of C2C12 myoblasts transiently transfected with p73␣, p73␤, or ⌬N-p73␣. C2C12 myoblasts were transiently transfected with GFP or GFP plus HA-tagged p73␣, p73␤, or ⌬N-p73␣. The transfected cells were then cultured in DM for 48 h to allow myogenic differentiation. Thereafter, the cells were fixed and stained with an anti-MHC monoclonal antibody and subsequently with a Cy3-conjugated secondary antibody to visualize MHC staining (red). The images of co-transfected GFP (green) were overlaid with those of MHC staining (red). B, percentages of MHC-positive cells among p73␣-, p73␤-, or ⌬N-p73␣-transfected C2C12 myoblasts 48 h into myogenic differentiation. Increasing amounts of p73␣-, p73␤-, or ⌬N-p73␣-encoding plasmids (0.1-3 g) were transfected into C2C12 myoblasts. GFP was co-transfected to earmark transfected cells. The cells were then cultured in DM for up to 48 h and processed thereafter as described in panel A. The percentages of MHC-positive cells among the GFP-positive cell population (a rough equivalent of cells transiently expressing p73␣ or p73␤) were enumerated. The averages of percentages from four independent experiments were plotted as a function of plasmid transfected. S.E. is shown on top of bars. C, luciferase assay for MyoD transcription activity of 4xRE-Luc in the presence of p73␣, p73␤, ⌬N-p73␣, or p53. A luciferase reporter plasmid (p4RE-Luc) containing four copies of E boxes was transfected into C2C12 myoblasts along with p53, p73␣, p73␤, or ⌬N-p73␣. A plasmid encoding Renilla luciferase was included in all transfections for normalizing the activities of firefly luciferase encoded on p4RE-Luc. At 48 h of transfection, the cells were lysed and assayed for firefly luciferase activity as well as that of Renilla luciferase, which was then used to normalize the former to arrive at an arbitrary unit reflective of MyoD transcriptional activity. The transcriptional activity of MyoD in the absence of p53 or p73 was arbitrarily set at 100, relative to which MyoD transcriptional activities in other conditions were expressed in percentages. The averages of percentages from three independent experiments were plotted as a function of plasmids transfected. S.E. is shown on top of bars. D, luciferase assay for the transcription activity of Gal4-MyoD fusion of GAL-Luc in the presence of p73␣, p73␤, ⌬N-p73␣, or p53. Same as in panel A except for GAL4-MyoD fusion and pGAL-Luc, respectively, in place of MyoD and p4RE-Luc. myogenic differentiation, p73␣, ⌬N-p73␣, or p73␤ was transiently expressed along with green fluorescence protein (GFP) in C2C12 myoblasts. GFP served as an indicator for transfected cells. After 24 h of transfection, the cultures were shifted into differentiation media (DM) to allow the activation of MyoD and the onset of myogenic differentiation (29,30). After 48 h in DM, the cultures were examined by immunofluorescence staining for MHC (myosin heavy chain), a muscle differentiation marker (Fig. 1A) (29,30). The percentages of MHC-positive cells among the GFP-positive population were scored (Fig. 1, A  and B). Ectopic expression of p73␣ reduced the incidence of MHC-positive cells by ϳ30-fold as compared with cells expressing GFP alone. Similar levels of suppressive effect could be achieved with the ectopic expression of ⌬N-p73␣ (Fig. 1, A and  B). By contrast, ectopic expression of p73␤ had no detectable effect on the extent of myogenic differentiation (Fig. 1B).

Inhibition of Myogenic Differentiation by Transiently
Suppression of MyoD Transactivation by p73␣, but Not p73␤-The transcription factor MyoD, a master regulator of myogenesis, is constitutively expressed in C2C12 myoblasts and drives the expression of muscle-specific genes during myogenic differentiation (31-33). The significant inhibition by p73␣ of myogenic differentiation prompted us to examine whether p73␣ affected MyoD activity. A luciferase reporter construct containing four MyoD binding sites (4RE) (29,30) was transfected either alone or with p53, p73␣, ⌬N-p73␣, or p73␤ into C2C12 myoblasts. Luciferase activity was measured at 48 h after shifting the transfected cultures into DM. Consistent with the inhibition of differentiation (Fig. 1A), p73␣ suppressed MyoD-dependent transcription by as much as 90% (Fig. 1C). ⌬N-p73␣ was equally effective at suppressing MyoD transactivation (Fig. 1C). By comparison, p73␤ or p53 was far less potent at attenuating MyoD activity (Fig. 1C).
To determine whether p73␣ directly interferes with the transactivation function of MyoD, we examined its effect on GAL4-MyoD. GAL4-MyoD is a fusion of MyoD and the Gal4 DNA binding domain (29,30). By using a luciferase reporter downstream of five GAL4-binding sites in tandem, it was possible to measure the transactivating function of the GAL4-MyoD fusion protein (29,30). Unlike the 4RE-Luc, co-expression of p73␣ or ⌬N-p73 with GAL4-MyoD had negligible effect on the expression of luciferase. Again, p73␤ or p53 did not affect the expression of GAL-Luc when coexpressed with GAL4-

FIG. 2. Delay in C2C12 myogenic differentiation by p73␣ in stable expression.
A, stable expression of HA-tagged p73␣ or p73␤ in C2C12 myoblasts. HA-tagged p73␣ or p73␤ was cloned in pMSCVhyg retroviral vector. Retroviral particles were produced by transfecting the constructs into BOSC packaging cells. Media harvested from the packaging cells 3 days after transfection were incubated with C2C12 myoblasts to attain viral infection. The infected myoblasts were selected with hygromycin to enrich the p73-expressing cell population. After drug selection, a portion of cells was lysed and examined with anti-HA antibody (␣-HA) on Western blot (WB) for the expression of p73␣ or p73␤. B, nuclear localization of p73␣ or p73␤ in C2C12 myoblasts. C2C12 myoblasts stably expressing HA-tagged p73␣ or p73␤ as established in panel A were fixed and stained with anti-HA antibody. The subcellular distribution of p73 proteins was visualized with a Cy3-conjugated secondary antibody (red). The cells were also counterstained phalloidin (green) and 4Ј,6,-diamidino-2-phenylindole (DAPI) (blue) to visualize actin and nuclei, respectively. C, immunofluorescence staining for MHC of C2C12 myoblasts stably expressing p73␣ or p73␤ during myogenic differentiation. C2C12 myoblasts stably expressing p73␣ or p73␤ were changed into differentiation media for myogenic differentiation. At days 0, 3, and 5 into myogenic differentiation, the cells were fixed and stained for MHC (red). The cells were also counterstained with DAPI to visualize nuclei. D, percentages of MHC-positive multinucleate cells among total cell population stably expressing p73␣ or p73␤ during myogenic differentiation. The cells were subject to myogenic differentiation and stained for MHC as described in panel C. The numbers of multinucleate MHC-positive cells were enumerated, and percentages relative to total cell population were plotted as a function of days into myogenic differentiation. The averages of percentages from four independent observations were used for plotting along with S.E. on top of the bar graphs. Two-way analysis of variance statistical analyses were performed to drive the p values of difference between p73␣or p73␤-expressing cells and vector-mocked cells on days 3 and 5 into myogenic differentiation. MyoD (Fig. 1D). These observations suggested that p73␣ might interfere with the interaction of MyoD with the 4RE promoter, but it did not directly inhibit the transactivating function of MyoD.
Substantial Delay in the Differentiation of C2C12 Myoblasts Stably Expressing p73␣-To corroborate the conclusions drawn from transient transfection experiments, HA-tagged p73␣ or p73␤ was stably expressed in C2C12 myoblasts through retroviral-mediated gene transduction. We were unable to maintain the stable expression of ⌬N-p73␣ in C2C12 myoblasts; hence, its effect was not further examined. Immunoblotting with anti-HA antibody confirmed stable expression and subcellular localization of both proteins ( Fig. 2A). Both p73␣ and p73␤ localized to the nuclei of C2C12 myoblasts (Fig. 2B). The cells were then placed in DM and myogenic differentiation assessed by immunofluorescence staining for MHC at days 3 and 5 (Fig.  2C). p73␣-expressing C2C12 cells differentiated substantially less than either vector-mocked or p73␤-expressing cells at both time points (Fig. 2C). The percentages of nuclei in MHC-positive multinucleate cells relative to those of total cell population were enumerated (Fig. 2D). C2C12 myoblasts stably expressing p73␣ were reduced almost by half in their potential for myo-genic differentiation as compared with vector-mocked C2C12 myoblasts (Fig. 2D). The stable expression of p73␤, on the other hand, had no apparent effect on the differentiation of C2C12 myoblasts (Fig. 2D). Results from stable cell line systems further supported that p73␣ is distinct from p73␤ in its ability to suppress myogenic differentiation.
Delay by p73␣ in the Expression of Muscle Differentiation Markers-Transcription cascades activated by MyoD initiate and maintain the expression of an array of muscle-specific genes (31)(32)(33). Because p73␣ interfered with MyoD function in transient transfection experiments (Fig. 1), we wanted to ascertain whether the stable expression of p73␣ would compromise MyoD function in C2C12 myoblasts. C2C12 myoblasts stably expressing p73␣ or p73␤ were cultured in DM for 6 days to allow myogenesis. At 1-day intervals the cells were harvested and processed for immunoblotting with antibodies that recognize MHC, myogenin, MyoD, and p21Cip1 (Fig. 3A). The up-regulation of MHC, myogenin, and p21Cip1 has been shown to be dependent upon the activity of MyoD (34 -37). In p73␣expressing cells, the MHC level was much lower than vectortransfected myoblasts on days 3 and 4 into myogenic differentiation (Fig. 3A). MHC expression in p73␣-expressing cells FIG. 3. Delay in the up-regulation of muscle differentiation markers by p73␣ in stable expression. A, protein expressions of MHC, myogenin, MyoD, and p21Cip1 in C2C12 myoblasts stably expressing p73␣ or p73␤ during myogenic differentiation. C2C12 myoblasts stably expressing p73␣ or p73␤ along with the vector-mocked control cell line were cultured in differentiation media to allow myogenic differentiation. At days 0 -6 into the differentiation, cells were harvested for Western blotting with antibodies against MHC, myogenin, MyoD, or p21Cip1 after the cell lysates were normalized by tubulin. B, a survey with DNA microarray of muscle-specific gene expressions suppressed by p73␣ in C2C12 myoblasts. C2C12 myoblasts stably expressing p73␣ or p73␤ were cultured in differentiation media for 24 h and thereafter were processed for RNA extraction. As a control, RNA was also prepared from undifferentiated vector-mocked C2C12 myoblast cultured in normal growth media. cRNA probes made from the RNAs were hybridized to Affymetrix mouse cDNA array, and signals were read and captured with an Affymetrix scanner. Data normalization and gene clustering were performed with D-Chip software. The table includes the genes related to muscle differentiation with their expression levels expressed in units relative to those from undifferentiated cells cultured in normal growth media (GM), which are arbitrarily set at 1. These genes consist of those (excluding the ones investigated by immunoblotting in panel A) that vary in mRNA abundance by more than 3-fold across the samples. became detectable on days 5 and 6, indicating a considerable lag in muscle gene expression. The suppression of myogenin expression in p73␣-expressing cells was evident on day 1 into myogenic differentiation. This suppression persisted through day 4 and languished thereafter. The induction of p21Cip1 was detected in vector or p73␤-expressing myoblasts on day 3 into differentiation, but not in p73␣-expressing cells. These results suggested that p73␣ delayed the expression of MyoD-driven gene expression in C2C12 cells. On the other hand, p73␤ had little effect on the expression of these muscle differentiation markers throughout the experimental time course (Fig. 3A).
Investigation of gene expression through immunoblotting is limited in scope by the availability of specific antibodies. We therefore used mouse oligonucleotide microarrays (MG_U74AV2; Affymetrix) to examine the effect of p73␣ or p73␤ on gene expression during C2C12 differentiation. p73␣-or p73␤-expressing C2C12 myoblasts were cultured in DM for 1 day prior to being subject to gene profiling with Affymetrix mouse DNA microarrays. The expression profile of C2C12 myoblasts cultured in regular growth media (GM) was also examined (Fig.  3B). Comparison of the expression profiles of a cohort of muscle-specific genes revealed their up-regulation in p73␤-expressing cells, but not in p73␣-expressing cells cultured in DM (Fig.  3B). With the hybridization values in GM-cultured cells set at 1, the levels of expression of this cohort of muscle-specific genes in DM-cultured cells were evaluated as -fold increase over the baseline (Fig. 3B). These genes can be categorized into three groups according to the development of cardiac/skeletal muscles, namely actin, myosin, and troponin. E boxes have been identified in promoters of cardiac/skeletal actin genes to mediate direct or indirect activation by MyoD (38 -40). The same holds true of cardiac/skeletal genes of myosin light/heavy chains and troponins (41)(42)(43). The gene profiling results were therefore consistent with the conclusion that MyoD activity was compromised by the ectopic expression of p73␣ in C2C12 cells.
Comparable Stimulation of Apoptosis by p73␣ and p73␤ in Cisplatin-treated C2C12 Cells-In cultured human cell lines, p73 is activated by DNA damage to promote apoptosis (7,8,44). Recently, p73 was shown to also promote tumor necrosis factor ␣-induced cell death (12). When cultured in differentiation media, C2C12 cells undergo myogenic differentiation and cell death because of the withdrawal of fetal bovine serum. Stable expression of p73␣ or p73␤ exacerbated cell death of C2C12 from serum deprivation (data not shown). Addition of IGF-1 (insulin-like growth factor 1) to the differentiation media suppresses death without compromising differentiation. The suppression of myogenic differentiation by p73␣ was observed in the presence of IGF-1 (Figs. 2 and 3). To determine whether the inhibitory effect of p73␣ on myogenic differentiation is related to its pro-apoptotic activity, we compared the efficiencies of p73␣ and p73␤ at sensitizing C2C12 cells to death induced by genotoxic stress. C2C12 myoblasts stably expressing p73␣ or p73␤ were treated with 25 M cisplatin (CDDP, cis-diamminedichloroplatinum) for induction of apoptosis by DNA damage. By comparison with vector-expressing myoblasts, both p73␣ and p73␤ caused a higher level of cell death at 24 and 48 h of cisplatin treatment (40 -50% in cell death versus Ͻ10% of vector-controlled myoblasts at 24 h) (Fig. 4A). The attendant proteolytic cleavage of pro-caspase 3 provided for a biochemical measurement of apoptosis, and this event was stimulated by the stable expression of either p73␣ or p73␤ in C2C12 myoblasts (Fig. 4B). p73␣ and p73␤ were equally effective at mediating DNA damage-induced death of myoblasts (Fig. 4, A and B). Therefore, the ability of p73␣ to suppress myogenic differentiation reflects functions other than its pro-apoptotic activity. DISCUSSION In this study, a biological distinction was made between p73␣ and its splice variant p73␤. p73␣ can suppress myogenic differentiation, whereas p73␤ does not have the same effect in C2C12 myoblasts. The distinction arose from the unique ability of p73␣ to attenuate MyoD activity, on which the myogenic differentiation program depends to regulate the expression of genes required for muscle formation. Both p73␣ and p73␤, however, sensitized myoblasts to cisplatin-induced cell death.
A previous study found that p73 gene expression is proactively suppressed in C2C12 myoblasts by a zinc finger homeodomain transcriptional repressor, ZEB (28). This suppression ends shortly after the onset of myogenic differentiation to allow p73 expression (28). It was unclear whether p73 expression contributed to the process of myogenic differentiation. Results from this study suggest that expression of p73 in myoblasts would be detrimental to myogenesis as p73␣ can substantially delay the onset of myogenic differentiation. Myoblasts may therefore selectively repress p73 expression to avoid the inhibitory effect of p73␣. A suppression of p73 expression until after the onset of differentiation was also observed in promyeloblasts, teratocarcinoma, and neuroblastoma (24,26,27). Regulating p73 expression could be a means of controlling differ- FIG. 4. Equivalent apoptosis induced by p73␣ and p73␤ upon DNA damages in myoblasts. A, cisplatin (CCDP)-induced apoptosis of myoblasts stably expressing p73␣ or p73␤. C2C12 myoblasts stably expressing p73␣ or p73␤ and mock of an empty vector were treated with 25 M CCDP for 24 or 48 h, at which time they were assayed for apoptosis. The percentages of dead cells among total cell population were quantified, and the averages were plotted with S.E. on top of the bar graphs. B, cisplatin (CCDP)-induced cleavage of caspase 3 in C2C12 myoblasts stably expressing p73␣ or p73␤. C2C12 myoblasts stably expressing p73␣ or p73␤ or an empty vector were treated with 25 M CCDP for 24 or 48 h, at which time they were assayed for cleavage of caspase 3. The cell lysates were probed with antibodies recognizing specifically either the full-length or cleaved form of caspase 3 by Western blotting. The amount of protein in each sample was normalized by tubulin expression. entiation. For instance, tumor necrosis factor ␣ (TNF␣) can up-regulate p73 through the activation of ABL (12) or ASK1 kinases (45) to induce apoptosis. Interestingly, TNF␣ has also been shown to interfere with myogenic differentiation (46 -48). The up-regulation of p73␣ could, in part, mediate this inhibitory effect of TNF␣.
Defects in terminal differentiation are a hallmark of cancer cells. Despite the initial hypothesis that p73 is a tumor suppressor, the human TRP73 gene is rarely mutated in sporadic cancers (2,4,49). In fact, p73 is found frequently overexpressed in human breast, bladder, liver, colon, and thyroid cancers, with p73␣ being its predominant form (50 -54). We also detected elevated p73␣ expression in rhabdomyosarcoma (data not shown). These cancers could exploit p73␣ to suppress differentiation in aid of their malignant transformation. Other mechanisms, such as nuclear exclusion of ABL tyrosine kinase (55), may be simultaneously enacted to undercut pro-apoptotic activity associated with p73 up-regulation in human cancers. Beyond its involvement in apoptosis, future characterization of p73 physiological functions will rely on the identification of factors that interact with different p73 variants in a cell contextdependent manner to regulate differentiation and possibly other biological processes.