1. ¹Department of Medicine, 840 S Wood St, University of Illinois at Chicago, Chicago, IL 60612, USA
2. ²Department of Biochemistry and Molecular Genetics, 900 S Ashland Ave., University of Illinois at Chicago, Chicago, IL 60607, USA

Correspondence: AL Gartel, E-mail: agartel@uic.edu; AL Tyner, Department of Medicine, 840 S Wood St, University of Illinois at Chicago, Chicago, IL 60612, USA. E-mail: atyner@uic.edu

³Current address: Howard Hughes Medical Institute, University of Wisconsin-Madison, Madison, WI 53706, USA

⁴Current address: Kookmin University, School of Techno-Science, Seoul, 136-702, Korea

Received 26 February 2004; Revised 8 April 2004; Accepted 16 April 2004; Published online 13 September 2004.

Abstract

p21^WAF1/CIP1 is an important transcriptional target of p53 and it plays a critical role in growth arrest after DNA damage. Here, we report the identification of a novel alternate mouse p21 transcript that is conserved in evolution. It differs from the classical p21^WAF1/CIP1 transcript in the first exon, which is located at approximately 2.8 kb upstream of transcriptional start site of p21^WAF1/CIP1 and is sandwiched between two p53 binding sites. This novel p21 transcript is present in most mouse tissues with highest levels of expression in the spleen. In contrast to the classical p21^WAF1/CIP1 transcript, this new transcript is highly dependent on p53 for its basal expression, as evidenced by its absence in nearly all of p53^-/- mouse tissues. This transcript is also absent at nonpermissive temperature in a 10-1 mouse cell line lacking endogenous p53 and harboring temperature-sensitive p53 mutant. However, this novel transcript is induced to appreciable levels in the presence of high p53 activity at the permissive temperature. Our data suggest that p53-dependent induction of p21 may be an additive effect conferred by individual increases in the alternate and classical p21 transcripts.

p21^WAF1/CIP1 is the founding member of the Cip/Kip family of cyclin-dependent kinase inhibitors (CKIs) (reviewed in Gartel et al., 1996). It inhibits the activity of cyclin/Cdk2 complexes and plays an important role in cell cycle regulation (Brugarolas et al., 1999). p21 is one of the main downstream effectors of p53 function upon DNA damage (el-Deiry et al., 1993). Expression of p21 is controlled mostly at the transcriptional level by both p53-dependent and -independent mechanisms (reviewed in Gartel and Tyner, 1999). Overexpression of p21 arrests cells primarily in the G1 phase of the cell cycle (Harper et al., 1995).

Alternative splicing or alternate promoter usage can lead to the expression of different transcripts from the same locus (reviewed in Landry et al., 2003). These mechanisms have been shown to regulate a number of CKIs, including p15^INK4B (Tsubari et al., 1997), p16^INK4a-ARF (Duro et al., 1995; Mao et al., 1995; Quelle et al., 1995), p18^INK4c (Phelps et al., 1998), p57^KIP2 (Tokino et al., 1996) and human p21^WAF1/CIP1 (Nozell and Chen, 2002). Transcripts from a single locus may encode the same protein, as in p18^INK4c (Phelps et al., 1998), or they may encode different proteins as seen in p16^INK4a-ARF locus, where one transcript encodes INK4, which functions as a CKI, and the other encodes ARF, which is an inhibitor of MDM2 (Quelle et al., 1995). The human p21^WAF1/CIP1 locus produces three different transcripts utilizing two different promoters (Nozell and Chen, 2002). p21B and p21C are expressed from a promoter located within the first intron of human p21A (the classical human p21^WAF1/CIP1 transcript; el-Deiry et al., 1993). While p21C encodes the normal p21 CKI, the p21B transcript produces a short protein of 123 amino acids that causes apoptosis when overexpressed. Upon DNA damage, p53 induces p21B and p21C transcripts through a newly identified p53 response element in their proximal promoter (Nozell and Chen, 2002).

In this study, we have analysed the mouse p21 genomic locus and found that a novel transcript with a different first exon is expressed, and that its expression is highly dependent on p53. We discovered this alternatively spliced p21 transcript using the BLAST search tool against the mouse EST database. The first exon of this spliced form (hereafter called transcript-1) is approximately 2.8 kb upstream from the transcription start site of the classical p21^WAF1/CIP1 transcript (el-Deiry et al., 1993) (hereafter called transcript-2) and it is conserved in rat and human (Figure 1b). These two transcripts are shown in the genomic context in Figure 1a, along with one other transcript (hereafter called transcript-3) that has been described previously (Huppi et al., 1994).

Figure 1.

Identification of a novel alternate mouse p21 transcript. (a) The p21 locus in the mouse genome. The sequences for transcript-2 and -3 were obtained from Genbank (Accession numbers U24173 and U09507, respectively). The sequence for transcript-1 was derived using EST sequences with Accession numbers BY747770, BY206886, BE135541, AA793510 and AI877450. The putative protein coding region for the three transcripts is shown as shaded area. Also indicated are the different primers used in RT–PCR to amplify the transcripts. (b) First exon of mouse p21 transcript-1 is conserved in rat and human. Sequence alignment of the first exon of mouse p21 transcript-1 with rat and human sequences from http://pipeline.lbl.gov/rat/, June 2003 release (Couronne et al., 2003) is shown

Full figure and legend (114K)

In order to determine if these p21 transcripts are expressed in a tissue-specific manner, we examined their levels in different tissues of wild-type mice using RT–PCR (Figure 2a). Expression of transcript-2 was found in all tissues examined. Within each tissue, the level of transcript-2 was higher than the other two transcripts. Highest levels of expression of transcript-1 were observed in the spleen where they were comparable to transcript-2 levels (Figure 2a). Appreciable amounts of transcript-1 were also detectable in kidney, thymus, muscle, colon and the reproductive tissues like ovary, uterus and cervix. Transcript-3 was detectable in all tissues except liver.

Figure 2.

Novel p21 transcript-1 is highly dependent on p53 for its basal expression in mouse tissues. (a) C57BL/6J female mice were fed a commercial diet and water ad libitum. They were killed at 8 weeks of age and total RNA from different tissues were isolated using Trizol Reagent (Invitrogen). RNA was treated with RQ1 DNase (Promega) to get rid of any genomic DNA contamination and RT–PCR was performed with RNA samples as follows. cDNA was synthesized using 0.5 g of DNase-treated RNA with the SuperScript First Strand Synthesis Kit according to the manufacturer's recommendations (Invitrogen). The three transcripts were identified by PCR and the primer locations are shown in Figure 1. The transcript-1 primers 5'-GCA TGA ATG GAG ACA GAG ACC-3' (primer 1.1) and 5'-ACA CCA GAG TGC AAG ACA GC-3' (primer 1.2) amplified a 432 bp product, the transcript-2 primers 5'-GCA GAT CCA CAG CGA TAT CC-3' (primer 2.1) and primer 2.2 (which is the same as primer 1.2) amplified a 398 bp product and the transcript-3 primers 5'-AGC CTG AAG ACT GTG ATG GG-3' (primer 3.1) and 5'-AAA GTT CCA CCG TTC TCG G-3' (primer 3.2) amplified a 220 bp product. PCR products were separated on a 2% agarose gel and visualized with ethidium bromide staining. (b) p53^-/- female mice (C57BL/6J-Trp53^tm1Tyj) were obtained from the Jackson laboratories. The mice were killed at 8 weeks of age and both RNA preparation and PCR product analysis was carried out as described above

Full figure and legend (103K)

To investigate how p21 transcripts are regulated by p53, we assessed their levels in different tissues from p53^-/- mice using RT–PCR (Figure 2b). The classical p21^WAF1/CIP1 (transcript-2) was expressed in all tissues examined, at levels comparable to those detected in tissues of wild-type mice. Transcript-3 was detected in kidney, thymus, uterus, cervix, duodenum, ileum, colon and stomach. In contrast, transcript-1 was absent in all tissues examined, except muscle, where a weak basal-level expression was seen. When the expression profiles of p53-null mice were compared with those of wild-type animals, the expression of transcript-2 was independent of p53 in many tissues as expected (Macleod et al., 1995), whereas transcript-1 was mostly absent when compared with the wild-type control. These data suggest that transcript-1 is highly dependent on p53 for its basal expression in murine tissues. In contrast to a previous report that basal p21 transcription in the liver is not dependent on p53 (Nunez et al., 2000), we found that transcript-2 levels are obviously reduced, although not completely absent, in p53^-/- liver tissue (Figure 2b) when compared with wild-type animals (Figure 2a), while the ₂-microglobulin mRNA levels assessed as internal control were found to be similar (data not shown).

Our studies indicate that the novel p21 transcript-1 is highly regulated by p53 when compared to the classical p21^WAF1/CIP1 (transcript-2) that is additionally regulated by p53-independent mechanisms (Gartel and Tyner, 1999). It has been reported previously that the basal transcription of p21^WAF1/CIP1 in the spleen is p53 dependent (Macleod et al., 1995). Interestingly, we were able to see appreciable level of transcript-2 in the spleen tissue derived from p53-null mice, but there were no detectable levels of transcript-1 (Figure 2b).

In order to further investigate how these different p21 transcripts are regulated by p53, we used a p53-null 10-1 mouse fibroblast cell line that ectopically expresses a temperature-sensitive p53 allele (described in Zambetti et al., 1992). Transcript-2 was expressed even in the absence of functional p53 and in the presence of p53 its expression was strongly induced as expected (Figure 3). Transcript-1 was not detectable in the absence of functional p53, even with additional cycles of PCR amplification. However, expression of transcript-1 was increased to appreciable levels in the presence of p53 (Figure 3). The dependence of transcript-1 on p53 is not surprising, because its start site is sandwiched between two p53 binding sites, which were previously reported to control the expression of the classical p21^WAF1/CIP1 (transcript-2) (el-Deiry et al., 1995). Our results suggest that p53-dependent induction of p21 may be an additive effect conferred by individual increase in alternate transcript-1 and classical transcript-2.

Figure 3.

The novel p21 transcript-1 is highly inducible by p53 in a mouse fibroblast cell line. The 10-1 cells expressing a murine temperature-sensitive p53 (a generous gift from Dr Nissim Hay, University of Illinois at Chicago) were grown in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% FCS (Atlanta Biologicals), penicillin and streptomycin in 5% CO₂. The temperature-sensitive p53 gene contains an alanine-to-valine substitution at amino-acid residue 135. At 39°C (nonpermissive temperature), more than 80% of the p53 is in a mutant conformation, while at 33°C (permissive temperature), 80% of the p53 is in wild-type conformation. The cells were grown to 50% confluency and then incubated at either 39 or 33°C for 24 h. RNA was extracted from the cells using RNeasy kit (Qiagen). cDNA preparation and PCR were carried out as described in legend to Figure 2. In addition, for each combination of primers, the kinetics of PCR amplification and primer annealing efficiency were studied and PCR was performed within the exponential range. While the number of cycles for PCR with ₂-microglobulin was set at 25, the amplification of all p21 transcripts was carried out for 32 and 30 cycles, respectively, for the samples derived from nonpermissive and permissive temperatures. Transcript-3 was not seen under these conditions, but it was detected at the permissive temperature with an increased number of PCR cycles or if the amount of input cDNA used was increased (data not shown)

Full figure and legend (33K)

p21 expression is often controlled at the level of transcription. We show here the existence of three different p21 transcripts with different 5' sequences due to alternative transcription start sites. Apart from a difference in dependence on p53 for their expression, it is also possible that the different 5' sequences may contribute to post-transcriptional regulation. This clearly needs further study.

The p21 genomic locus is organized differently in rodents and humans. The human p21 locus encodes the CDK inhibitor p21 and an alternate p21B transcript encoding a novel protein capable of causing apoptosis (Nozell and Chen, 2002). The p21B protein lacks homology with known mouse or rat sequences; however, the first 38 amino acids of its CDS correspond to an apparent MIR repeat (GenBank Z85996), suggesting a recent origin by retrotransposition. Despite these differences, the first exon of p21 transcript-1 is conserved in mouse, rat and human (Figure 1b) both in terms of sequence composition and the positioning between the two p53 binding sites, implying that this new transcript may play an important role in the downstream effects mediated by p53. Characterization of a human p21 transcript-1 in response to p53 activation and various stimuli is currently underway (Radhakrishnan and Gartel, manuscript in preparation).

It is generally believed that changes in gene regulation account for much, if not most evolutionary change (Levine and Tjian, 2003). Acquisition of a novel enhancer for one gene can result in a different structure and allow survival in a new environment (Wang and Chamberlin, 2004). However, the mechanisms by which new promoters or enhancers are added to a gene, or whether regulatory elements from an upstream promoter can be incorporated into the regulation of a downstream transcription site for more efficient mRNA production, remain unclear. Intriguingly, we have found that the p53-responsive enhancer of transcript-2 is positioned as the core promoter of transcript-1. As transcription through regulatory chromatin can derepress downstream promoters (Gribnau et al., 2000; Hogga & Karch, 2002), the production of transcript-1 might enhance transcript-2. It is possible that the transcript-1 represents an intermediate stage between a stand-alone promoter and an intergenic transcript and/or enhancer element. If so, further study of p21 transcriptional regulation may provide general insights into the modes of evolution of gene structure.