UV-dependent Alternative Splicing Uncouples p53 Activity and PIG3 Gene Function through Rapid Proteolytic Degradation*

The p53-inducible gene 3 ( PIG3 ) is a transcriptional target of the tumor suppressor protein p53 and is thought to play a role in apoptosis. In this report, we identify a novel alternatively spliced product from the PIG3 gene that we call PIG3AS (PIG3 alternative splice). PIG3AS results from alternative pre-mRNA splicing that skips exon 4 of the five exons included in the PIG3 transcript. The resulting protein product shares its first 206 amino acids with PIG3 but has a unique 42-amino acid C terminus. In unstressed cells and after most DNA damage conditions that induce transcription from the PIG3 gene, production of the PIG3 transcript dominates. However, in response to UV light, pre-mRNA splicing shifts dramatically in favor of PIG3AS. Unlike the PIG3 protein, the PIG3AS protein is rapidly degraded with a short half-life and is stabilized by proteasome inhibition. Our results illustrate the first example of an endogenous, UV-inducible, alternative splicing event and that control of the splicing machinery is involved in the cellular DNA damage response. They also suggest that rapid proteolytic degradation represents a cellular mechanism for uncoupling p53 activity from PIG3 gene activation that is independent of promoter selectivity. The transcription of most eukaryotic genes results in pre-mRNAs consisting of coding regions, called exons, separated then both made from pBabe with overlap extension PCR using the following primers:

The transcription of most eukaryotic genes results in pre-mRNAs consisting of coding regions, called exons, separated by non-coding regions, called introns. During pre-mRNA processing, intron sequences are removed and exons are spliced together to produce mature mRNA transcripts. Often, numerous mRNAs are produced from the same pre-mRNA because of unique exon ligations generated in the process of alternative splicing. This enables the production of numerous unique protein coding transcripts from a single gene and can greatly modify gene function. Recent data suggest that the number of genes in the human genome subject to alternative splicing may be as high as 60% (1). Therefore, alternative splicing could explain much of the complexity built into the human genome considering the unexpectedly low number of annotated genes (30,000 -40,000) compared with simpler organisms such as Drosophila melanogaster (14,000;Ref. 2) and Caenorhabditis elegans (19,000;Ref. 3).
Alternative splicing is involved in many cellular processes and is under extensive intracellular control (4). The inclusion of alternative exons can occur in a developmental or tissue-specific manner (5,6). A growing number of extracellular signals have also been shown to induce changes in splicing (7). Although the mechanisms mediating many of these inducible alternative splicing events remain to be determined, a few have been shown to depend on conserved signal transduction pathways that ultimately influence proteins of the splicing machinery (7). p53 function as a tumor suppressor depends on its ability to respond to a wide array of oncogenic stimuli and activate the transcription of numerous genes involved in cell cycle arrest, apoptosis, and DNA repair (8 -10). One such gene is the p53inducible gene 3 (PIG3). 1 PIG3 was originally discovered in a serial analysis of gene expression study designed to identify genes induced by p53 before the onset of apoptosis (11). It is unique among p53-responsive genes in that its expression is mediated through the microsatellite sequence (TG(C/T)CC) n (12) rather than the classic consensus element (two or more copies of the 10-bp half-site 5Ј-PuPuPuC(A/T)(A/T)GPyPyPy-3Ј separated by up to 13 bp) (13). There is some evidence to suggest that the PIG3 protein is involved in the generation of reactive oxygen species (ROS) (11), which are important downstream mediators of the p53-dependent apoptotic response (14,15). First, PIG3 expression precedes the appearance of ROS in p53-induced apoptosis (11). Second, PIG3 shares sequence similarity with numerous NAD(P)H quinone oxidoreductases shown to be potent inducers of ROS (11). Third, certain p53 mutants capable of inducing cell cycle arrest but not apoptosis retain the ability to activate target genes such as the cyclindependent kinase inhibitor p21, but not PIG3 (16,17). However, because PIG3 expression alone is insufficient to cause apoptosis, it is suspected that many factors cooperate to cause ROS-dependent cell death (11). At this time, the exact role of PIG3 in p53-dependent apoptosis remains to be determined.
In this report, we characterize a novel splice variant from the PIG3 gene called PIG3 alternative splice (PIG3AS). In undamaged cells and after most forms of DNA damage that induce p53-dependent transcription from the PIG3 gene, PIG3AS remains the minor mRNA species produced. However, in response to UV radiation, an inducible shift in pre-mRNA splicing takes place that almost exclusively favors the production of the PIG3AS transcript. Unlike PIG3, we find that the PIG3AS protein product is rapidly degraded by the proteasome. Our results demonstrate the first example of an endogenous gene subject to UV-dependent alternative splicing. They also illustrate that alternative splicing coupled to proteolytic degradation represents a means of uncoupling p53 activity from target gene function.

EXPERIMENTAL PROCEDURES
Cell Culture and Treatment-MCF7 breast adenocarcinoma cells were cultured in DMEM supplemented with 10% fetal calf serum, 50 units/ml penicillin G, and 50 g/ml streptomycin sulfate (Invitrogen). HCT-116 p53ϩ/ϩ and p53Ϫ/Ϫ colorectal carcinoma cells (a kind gift of Dr. Bert Vogelstein (18)) were grown in McCoy's 5A (Invitrogen) supplemented with 10% fetal bovine serum, 50 units/ml penicillin G, and 50 g/ml streptomycin sulfate. Doxorubicin, etoposide, actinomycin D, and MG-132 (Sigma) were added to cell medium at a final concentration of 0.2 g/ml, 50 M, 10 g/ml, and 10 M, respectively. For UV treatment, cells were rinsed in warm phosphate-buffered saline and exposed to a 252 nm germicidal UVC lamp (GE) in a minimal volume of phosphate-buffered saline at a dose of 2.8 J/m 2 /s. The UV lamp was routinely calibrated with a Blak-Ray UV Intensity Meter (UVP; Model J-225). Ionizing radiation was performed with a 137 Cs irradiator at a rate of ϳ2.5 grays/min.
Cloning and Production of Stable Cell Lines-PIG3 and PIG3AS were first amplified from cDNA made from MCF7 cells using the following primers: PIG3 5Ј clone (KpnI), ATGGTACCGACAATATGTTA-GCCGTGCAC; PIG3 3Ј clone, ATGAATTCTTCACTGGGGCAGTTCC-AGGA; and PIG3AS 3Ј clone, ATGAATTCCTATGGGCCTCCTG-GATTTCG. The resulting PCR products were cloned into the KpnI and EcoRI sites of pcDNA3.1(ϩ)(Invitrogen). PIG3 and PIG3AS cDNA was then excised from pcDNA 3.1 with KpnI and EcoRI and subcloned into BglII and BamHI sites of pBabe Puro using the following linker: top strand, P-GATCCAGCTAGATCTAGCTGTAC; bottom strand, P-AGC-TAGATCTAGCTG. This linker added a BglII site to the multiple cloning site of pBabe Puro when ligated with PIG3 excised from pcDNA 3.1 with KpnI and EcoRI. After preliminary experiments were performed, an N-terminal FLAG epitope (DYKDDDDK) was added to both pBabe PIG3 and pBabe PIG3AS using PCR with the following primers: PIG3 5Ј-FLAG, 5Ј-GATAGATCTGGTACCGACAATATGGATTACAAGGAC GATGACGATAAGGGTGGTTTAGCCGTCCACTTTGACAAGCC-3Ј; PIG3 3Ј clone, sequence above; and PIG3AS 3Ј clone, sequence above. PCR products were purified and cloned into the BglII and BamHI sites of pBabe PIG3. FLAG-PEST 1 (F-PEST 1) was made from pBabe F-PIG3AS with overlap extension PCR using the following primers: PEST 1 and 2 5Ј-FLAG, GATAGATCTGGTACCGACAATATGGA; PEST 1 3Ј, GGCAGCAGCTACTCCTCCCCCTCCGTGGAGAAGTGAG-G; PEST 1 5Ј, GGAGGAGTAGCTGCTGCCGGTTCTGGACAGAATC; and PIG3AS 3Ј clone, sequence above. First, two separate PCR products were made from PEST 1 and 2 5Ј FLAG/PEST 1 3Ј and PEST 1 5Ј/PIG3AS 3Ј clone primer sets and gel purified. The two products were then mixed and re-amplified with PIG3 5Ј clone/PIG3AS 3Ј clone. F-PEST 2 was made from pBabe F-PIG3AS using PCR with the following primers: PEST 1 and 2 5Ј FLAG, sequence above; PEST 2 3Ј, ATGAA-TTCCTATGGGCCTCCTGGATTTCGCAATGCTCCTA GATTCTGTCC-AGAACCGGCA. F-PEST 1/2 was made by PCR from pBabe F-PEST 1 using the PEST 1 and 2 5Ј FLAG and PEST 2 3Ј primers. F-PIG3AS⌬42 was made with the following primers: PEST 1 and 2 5Ј FLAG, sequence above; and F-PIG3AS⌬42, ATGAATTCCTATTTGGTGAATTTCAGCG-TTGCTTC. The resulting PCR products from F-PEST 1, -2, -1/2, and F-PIG3AS⌬42 were cloned into the BglII and EcoRI sites of pBabe PIG3. PCR amplifications used for cloning was done using Proof Start Polymerase (Qiagen) according to the manufacturer's instructions. Each clone was sequenced to confirm that no mutations were acquired.
To construct polyclonal stable cell lines overexpressing F-PIG3, F-PIG3AS, F-PEST 1, F-PEST 2, F-PEST 1/2, and F-PIG3AS⌬42, the pBabe retroviral system was used (19). In brief, pBabe vectors were transfected into Bosc23 packaging cells using SuperFect (Qiagen) according to manufacture's instructions. 24 h after transfection, medium containing recombinant retrovirus was collected and used to infect the second packaging cell line PT67. All infections were performed in 6-well plates with the cells at ϳ40% confluence. Infection mixtures contained 0.5 ml of medium containing virus, 0.5 ml of fresh medium, and 8 g/ml polybrene (Sigma). 48 h after infection, the cells were split to 5% confluence and selected successively in 2-4 g/ml puromycin (Sigma). After selection, recombinant retrovirus was harvested in medium lacking puromycin and used to infect HCT-116 p53Ϫ/Ϫ cells. Cells were once again successively selected with puromycin.
A typical PCR reaction of 50 l contained 1ϫ PCR buffer (Invitrogen), 1. Densitometry was done using Quantity One software (Bio-Rad) on unaltered images after normalization to ␤-2-M levels. To determine the percentage PIG3AS of total new transcripts in Fig. 2, PIG3AS was first multiplied by 2.15 to compensate for differences in fragment sizes. We then used the calculation below.
Production of PIG3AS Specific Antibody-The C-terminal 42 amino acids of PIG3AS was amplified from pcDNA 3.1 PIG3AS using the following primers: GST exon 5 5Ј, CGGGATCCGTACAAGCAAATGCT-GGTGAATGC; and PIG3AS 3Ј clone, sequence above. The resulting PCR product was then cloned into the BamHI and EcoRI sites of the GST-fusion vector pGEX-4T-2 (Amersham Biosciences). GST-fusion protein was then produced in Escherichia coli and used by the Southern Alberta Cancer Research Center Hybridoma Facility (University of Calgary) to generate rabbit polyclonal antibodies reactive against PIG3AS. To remove nonspecific antibodies in the collected serum reactive against GST, the serum was diluted 1:10 in Tris-buffered saline and incubated in two successive batches of GST bound to glutathione agarose (Amersham Biosciences) overnight at 4°C. 0.1% sodium azide was added to the remaining supernatant and the antibody was used directly for Western blotting.
Pulse-Chase Analysis-HCT-116 p53Ϫ/Ϫ cells stably overexpressing F-PIG3 or F-PIG3AS were seeded into 6-well plates. Once they reached 60% confluence, they were rinsed once in phosphate-buffered saline, rinsed once in DMEM lacking methionine and cysteine (Met-Cys-DMEM; Invitrogen), and incubated in Met-Cys-DMEM supplemented with 10% dialyzed fetal calf serum (Invitrogen), 50 units/ml penicillin G, and 50 g/ml streptomycin sulfate. The cells were then pulsed for 2 h in Met-Cys-DMEM containing 5 l/ml 35 Met (530 MBq/ml; Amersham Biosciences). After pulsing, 35 Met-containing medium was removed and the cells were incubated in a large volume of McCoy's 5A medium (containing methionine and cysteine) before being harvested at the appropriate time points. Cells were harvested by trypsin and resuspended in 1 ml of FLAG immunoprecipitation buffer (50 mM Tris, pH 7.4, 150 mM NaCl, 1 mM EDTA, and 1% Triton X-100) for 30 min on ice. After centrifugation, immunoprecipitations were performed on soluble cell extracts using Anti-FLAG M2 Affinity Gel (Sigma) according to the manufacturer's instructions. After extensive washing, 50 l of protein sample buffer was added to the beads and they were incubated at 95°C for 5 min. Approximately 25 l of each sample was loaded on a 12% polyacrylamide gel. After running, each gel was dried and exposed to Kodak BioMax MR film at Ϫ70°C.

PIG3AS Is an Alternatively Spliced Transcript from the PIG3
Gene-While performing RT-PCR to examine the relative levels of PIG3 transcripts in MCF7 cells, we noticed the presence of a second PCR product (Fig. 1A); we reasoned that this could represent an alternatively spliced transcript from the PIG3 gene. To determine its identity, and to eliminate the possibility of nonspecific amplification, we purified and sequenced this band. Results confirmed that the unidentified band was part of an alternatively spliced PIG3 transcript in which exon 4 was absent (Fig. 1B). The presence of this alternatively spliced form of PIG3, which we refer to as PIG3AS, was also verified in a number of primary and transformed cell lines and with a multiple tissue Northern blot (data not shown). In each case, PIG3AS remained the minor mRNA isoform produced from the PIG3 gene.
The predicted protein product from PIG3AS contains 248 amino acids compared with the 332 of PIG3 (Fig. 1C). PIG3AS shares its first 206 amino acids with PIG3 but has a unique C terminus consisting of 42 amino acids as a result of a frameshift caused by the ligation of exon 3 to exon 5. This also causes PIG3 and PIG3AS to use unique stop codons in exon 5. The unique 42-amino acid C terminus of PIG3AS shares no significant similarity with proteins in the public databases. In addition, PIG3AS lacks a substantial C-terminal domain (126 amino acids compared with PIG3) that shares similarity with many quinone oxidoreductases that would probably render it catalytically inactive (11).
UV Light Induces the Signal-dependent Alternative Splicing of PIG3 Pre-mRNA Favoring PIG3AS Transcripts-Because PIG3 has been shown to be induced in a p53-dependent manner after application of numerous DNA damaging agents, we sought to determine whether PIG3AS behaved in a similar manner. To this end, MCF7 cells were exposed to doxorubicin, etoposide, UV radiation, or ionizing radiation and harvested at various times for both total RNA and total protein. To examine transcripts semiquantitatively, we used the primer drop RT-PCR method with ␤-2-microglobulin (␤-2-M) as an internal standard (see "Experimental Procedures"). Each treatment induced both PIG3 and PIG3AS but to varying degrees ( Fig. 2A,  top). Doxorubicin, etoposide, and ionizing radiation increased levels of both PIG3 and PIG3AS transcripts; PIG3 remained the dominant isoform. It was surprising to find that there was a dramatic shift in the pre-mRNA splicing of PIG3 transcripts in response to UV light; PIG3 levels remained relatively constant, whereas PIG3AS increased significantly. Densitometric analysis indicated that at each time point after UV irradiation, more than 75% of new PIG3 transcripts consisted of PIG3AS ( Fig. 2A, lanes 8 -10). This can be compared with untreated cells, where PIG3AS represents only 17.1% of total PIG3 transcripts ( Fig. 2A, lane 1). Although increases in PIG3AS transcript percentage are found in response to other forms of DNA damage, the effect was less intense and more transient in nature compared with UV treatment. Additional stimuli known to activate p53, including hypoxia, hydrogen peroxide, and methyl methanesulfonate, were also tested but induced neither PIG3 nor PIG3AS in MCF7 cells (data not shown).
Although the data presented thus far indicate that the production of PIG3AS transcripts dominate after UV exposure, the stability of both PIG3 and PIG3AS transcripts needed to be analyzed to differentiate between an alternative splicing event being specifically activated in response to UV light versus a differential effect on mRNA stability. To this end, MCF7 cells were incubated in the presence of actinomycin D for various times with or without a prior UV exposure, and RT-PCR analysis was performed (Fig. 2B). As can be seen, both PIG3 and PIG3AS transcripts are relatively stable, similar to that of ␤-2-M, with little quantifiable difference in half-life that could account for the UV-induced accumulation of PIG3AS (Fig. 2B,  top). As a control for this method, RT-PCR was also done using primers designed to amplify two alternatively spliced Cdc2-like kinase transcripts (21); both species have short half-lives (Ͻ4 h; Fig. 2B, bottom). Together, these results show that PIG3AS is induced in response to stimuli capable of activating p53-dependent transcription from the PIG3 gene. Although PIG3 usually remains the dominant isoform produced, UV almost exclusively favors the alternative splicing of PIG3 pre-mRNA and accumulation of PIG3AS. To our knowledge, this represents the first example of an endogenous UV-inducible alternative splicing event.
To further characterize the induction of PIG3AS, MCF7 cells were exposed to increasing amounts of UV radiation ranging from 10 -50 J/m 2 and harvested 24 h later (Fig. 2C). At a low dose (10 J/m 2 ), nearly equal amounts of PIG3 and PIG3AS were induced (Fig. 2C, lane 2). However, as cells were exposed to increasing amounts of UV, the ratio of the two transcripts shifted in favor of PIG3AS. This effect peaked at a dose of 40 J/m 2 , above which the ratio of new PIG3AS produced no longer increased. These results indicate that PIG3 pre-mRNA splicing is sensitive to the amount of UV radiation; increased UV exposure maximizes the percentage of PIG3AS produced.
Because alternative splicing has also been shown to be promoter-dependent (22,23), and considering that different transcription factor binding sites may exist in the PIG3 promoter (24), we wanted to determine whether the induction of PIG3AS after UV was p53-dependent. For this, we exposed the isogenic cell lines HCT-116 p53ϩ/ϩ and HCT-115 p53Ϫ/Ϫ (18) to UV light and analyzed total PIG3 transcripts using RT-PCR (Fig.  2D). The results show that PIG3AS induction after UV exposure is entirely p53-dependent. It is noteworthy that baseline transcription of PIG3 or PIG3AS does not require p53 (Fig. 2C,  compare lanes 1 and 3). However, we could not determine whether an endogenous shift in pre-mRNA splicing occurs in HCT-116 p53Ϫ/Ϫ cells after UV exposure because of the long half-life of PIG3 and PIG3AS mRNA (Fig. 2B) coupled with the low baseline rate of transcription in these cells.
Unlike PIG3, PIG3AS Is Rapidly Degraded by the Proteasome-Next, it was important to confirm that PIG3AS encoded the predicted protein (Fig. 1C) and to determine whether its cellular levels increase after UV exposure. For this, some of the cells used in Fig. 2A were harvested for total protein and analyzed with Western blotting. Using transient transfection experiments, we previously determined that a commercially available PIG3 antibody (Exalpha Biologicals) was capable of recognizing PIG3AS as well (data not shown). Although each type of DNA damage was found to increase PIG3 protein levels (Fig. 3, middle), no PIG3AS could be detected, even after UV exposure.
Because PIG3 but not PIG3AS was readily detectable by Western blotting, it was possible that PIG3AS was being rapidly degraded by the proteasome. To test this hypothesis, we first created polyclonal cell lines stably overexpressing PIG3 (F-PIG3) and PIG3AS (F-PIG3AS), each containing an N-terminal FLAG epitope using the pBabe retroviral system (see "Experimental Procedures"). The identity of each cell line was verified using RT-PCR (Fig. 4A). A control cell line (pBabe) and those overexpressing either F-PIG3 or F-PIG3AS were then treated with the proteasome inhibitor MG-132, harvested for total protein, and subjected to Western blotting with an anti-FLAG antibody. Although the expression of both F-PIG3 and F-PIG3AS transcripts was similar, only the F-PIG3 protein could be detected in untreated cells (Fig. 4B, compare lanes 3  and 5). However, after incubation with MG-132, F-PIG3AS was stabilized to the extent that it could be detected (Fig. 3B, lane  6). To further confirm these results, we performed pulse-chase experiments to examine the stability of both proteins. Although the half-life of F-PIG3 was more than 24 h, the half-life of F-PIG3AS was only ϳ1 h ( Fig. 4C and data not shown).
PIG3AS Degradation Does Not Depend on a Putative PESTlike Motif or its Unique C-terminal Domain-Next, we wanted to determine why PIG3AS, but not PIG3, is rapidly degraded by the proteasome. Upon closer examination of the unique C-terminal region, we noticed a potential PEST motif between amino acids 225 and 248 (Fig. 5A). PEST motifs are stretches of amino acids rich in proline, glutamic acid, serine, and threonine that target proteins for proteolytic degradation (25). To study whether this region is responsible for mediating the rapid degradation of PIG3AS, we created polyclonal cell lines overexpressing PIG3AS proteins containing a FLAG epitope and harboring mutations at amino acids critical to the makeup of the PEST domain (Fig. 5A). Once again, the cell lines were constructed using the pBabe system and verified using RT-PCR (Fig. 5B, left). We expected that mutation of the PEST domain would result in the stabilization of the F-PIG3AS mutants without MG-132. We were surprised to find, however, that each mutant behaved similarly to F-PIG3AS and required proteasome inhibition to become stabilized (Fig. 5C, left).
It was next necessary to determine whether other sequence information contained within the C-terminal domain of PIG3AS protein contributed to its rapid proteolytic degradation. To do this, we generated a FLAG-tagged version of PIG3AS devoid of its unique C-terminal 42 amino acids (F-PIG3AS⌬42). After confirmation of this cell line (Fig. 5B, right), we found that similar to F-PIG3AS, F-PIG3AS⌬42 protein was only stabilized in the presence of MG-132 (Fig. 5C, right). These results indicate that PIG3AS is degraded independently of a putative PEST-like motif or any region located in its 42 C-terminal amino acids.
Endogenous PIG3AS Is Also Rapidly Degraded by the Proteasome-Although our results indicate that overexpressed F-PIG3AS is rapidly degraded by the proteasome, we wanted to determine whether the same was true for endogenous protein.
To accomplish this, MCF7 cells were exposed to 20 J/m 2 UV for 12 h and then incubated with MG-132 for an additional 12 h. Although this dose of UV induces both PIG3 and PIG3AS transcripts (Figs. 2C and 6), it was chosen because it minimized cell death when combined with MG-132. Consistent with this, Western blot analysis revealed an increase in PIG3 protein after UV exposure (Fig. 6, top). In contrast, endogenous PIG3AS induced by UV is only detectable in the presence of MG-132 (Fig. 6, middle). It is unknown why PIG3 protein levels FIG. 3. PIG3 protein, but not PIG3AS, is readily detectable after various forms of DNA damage. Western blots were performed on total cell extracts from MCF7 cells harvested in Fig. 2A using anti-p53 (top), anti-PIG3 (middle), and anti-QM (bottom; loading control) antibodies. The asterisks represent nonspecific bands reactive against the PIG3 antibody. decrease after combined UV and MG-132 treatments compared with UV alone in this experiment (Fig. 6, top). A similar, but less dramatic effect was also observed when using MG-132 in situations in which PIG3 is overexpressed (see Fig. 4B and data not shown). These results illustrate that endogenous PIG3AS is targeted for rapid proteolytic degradation by the proteasome. They also suggest that although the PIG3 gene is activated in response to UV light in a p53-dependent manner, little corresponding protein accumulates (especially after high doses of UV radiation) as a result of the proteolytic degradation of PIG3AS protein.

DISCUSSION
In this report, we describe a novel splice variant produced from the PIG3 gene that we call PIG3AS. To our knowledge, PIG3AS represents the first example of an endogenous UVinducible alternative splicing event and illustrates the involvement of the splicing machinery in the cellular DNA damage response. We also propose that the proteolytic degradation of PIG3AS represents a novel mechanism to regulate PIG3 gene function that is independent of p53 promoter selectivity.
UV-inducible Alternative Splicing of PIG3AS-One way living organisms respond to their environment is through changes in gene expression. This can occur by increasing or decreasing the levels of a particular transcript as well as altering pre-mRNA splicing patterns (alternative splicing). Whereas bioinformatics approaches and microarray analysis have enabled the identification of numerous alternatively spliced transcripts, only a limited number of inducible alternative splicing events have been described previously (7). In this report, we used RT-PCR to characterize the inducible alternative splicing of PIG3AS in response to UV light. This effect is limited to UV light in that a similar effect is not seen in response to other forms of DNA damage that up-regulate PIG3 gene expression (Fig. 2).
From our data, it is clear that UV light alters the splicing of PIG3 pre-mRNA to skip exon 4. A number of proteins that affect splice site selection and thus alternative splicing have been described and can generally be divided into two main groups, serine-arginine (SR) proteins and heterogeneous ribonuclear proteins (hnRNPs) (23). These proteins bind to specific RNA sequences and either enhance or repress the activity of the spliceosome toward a particular exon. The activity of certain SR and heterogeneous ribonuclear proteins can be modified by phosphorylation status (26 -28) and changes in subcel-  A) and C-terminal deletion (⌬42) mutants of PIG3AS. Total RNA was isolated from these cells, those infected with an empty retrovirus (pBabe), or those overexpressing F-PIG3AS (as a control for expression) and subject to RT-PCR using PIG3 primers (lanes 1-5; 26 cycles), PIG3 exon 3 5Јand F-PIG3AS⌬42 primers (lanes 6 -7; 26 cycles), and ␤-2-M primers (lanes 1-7; 22 cycles). C, the stability of F-PIG3AS is not enhanced by mutations in a PEST-like motif or deletion of its unique C-terminal domain. Stable cell lines overexpressing F-PIG3AS, F-PIG3AS mutated in its putative PEST domain, or F-PIG3AS⌬42 were either left untreated or incubated with 10 M MG-132 for 12 h as indicated and harvested for total protein. Western blots were performed using anti-FLAG (top) and anti-QM (bottom; loading control) antibodies. The asterisk represents a nonspecific band reactive against the FLAG antibody. lular localization (29,30) that depend on conserved cellular signaling networks (7,30). Of particular interest with respect to the data presented here is the inducible alternative splicing of CD44 variable exon 5. In response to T-cell activation, this particular alternative splicing event has been shown to be dependent on the extracellular signal-regulated kinase-mitogen-activated protein kinase signaling pathway (ERK-MAPK) (31,32), a cascade also known to be activated in response to UV light (33). It is noteworthy that UV can also induce the cytoplasmic translocation of heterogeneous ribonuclear proteins A1 through MKK 3/6 -p38 signaling and alters the splicing of an adenovirus reporter construct (30). It is possible that the splicing of PIG3AS could also be regulated by the MAPK pathway and/or hetergeneous ribonuclear proteins A1. At present, the signaling pathway and identity of splicing factors required for the alternative splicing of PIG3AS are unknown.
Our results also implicate the splicing machinery as a component of the cellular response to DNA damage. It is well documented that proteins involved in the DNA damage response are frequently mutated in human cancers. For example, p53, xeroderma pigmentosa, and Cockayne syndrome proteins, required for the identification and repair of UV-induced DNA damage, are also frequently associated with skin cancer (34 -36). The splicing process has also been identified as a major cause of disease (37). It is estimated that ϳ15% of single basepair mutations causing human genetic disease are the result of splicing defects (38). Because of the link between DNA damage and alternative splicing described here, one might predict that mutations in genes that influence splice site selection would be found in human tumors and may facilitate the development or progression of cancer. Consistent with this, numerous misspliced transcripts (37,39,40), including those encoding tumor suppressors and oncogenes (41)(42)(43), have been reported in human cancers without underlying causes (i.e. mutation or change in function of a splicing factor).
Is PIG3AS a Non-functional Protein?-PIG3 was first identified in a serial analysis of gene expression screen designed to identify p53-regulated genes before the induction of apoptosis (11). Because of its similarity to NAD(P)H quinine oxidoreductases, PIG3 is suggested to promote the production of ROS and apoptosis (11). We have performed a number of experiments in our lab to determine the function of PIG3. Like others (11,24), we have also found that the overexpression of PIG3 does not induce apoptosis or have any detrimental effects on cell growth (data not shown).
The proteolytic degradation of PIG3AS is strikingly similar to that of the polymorphic P187S variant of NAD(P)H quinine oxidoreductase 1 (44), a protein it shares significant similarity with. Cell lines homozygous for this version of NAD(P)H quinine oxidoreductase 1 lack enzymatic activity (45). It was later shown that this protein, unlike its wild-type counterpart, is targeted for destruction by the proteasome, perhaps being recognized as a misfolded protein (44,46). Whatever the function of PIG3, we favor the notion the PIG3AS represents a nonfunctional gene product similar to P187S because both proteins are unstable and rapidly degraded.
Uncoupling p53 Activity and PIG3 Gene Activation-p53 is an intimate player in the cellular decision between life and death through the selective induction of genes involved in either cell cycle arrest or apoptosis (10,47). Regulation of p53 activity in this manner seems to depend on specific post-translational modifications of p53 itself that influence its affinity for a specific consensus sequence or regulate its association with other co-activators of transcription. Here we describe a novel mechanism by which a cell can regulate p53-dependent gene activation that is independent of promoter selectivity and tran-scriptional activation. Although high doses of UV light (Ն40 J/m 2 ) induce transcription from the PIG3 gene in the form of PIG3AS, we have shown that little corresponding protein accumulates. One could therefore reason that the activated form of p53 in these situations cannot differentiate the PIG3 binding site as it can for those in the promoters of other genes. This may be a result of its unique and newly evolved microsatellite structure compared with the classic consensus element found in other p53-responsive genes (12,48). In this context, UV-regulated PIG3 gene activation may represent an undesirable consequence of p53 activation that is corrected by alternative splicing and proteolytic degradation.