Bcl-2 Protein Is Required for the Adenine/Uridine-rich Element (ARE)-dependent Degradation of Its Own Messenger*

We have shown previously that the decay of human bcl-2 mRNA is mediated by an adenine/uridine-rich element (ARE) located in the 3′-untranslated region. Here, we have utilized a non-radioactive cell-free mRNA decay system to investigate the biochemical and functional mechanisms regulating the ARE-dependent degradation of bcl-2 mRNA. Using RNA substrates, mutants, and competitors, we found that decay is specific and ARE-dependent, although maximized by the ARE-flanking regions. In unfractionated extracts from different cell types and in whole cells, the relative enzymatic activity was related to the amount of Bcl-2 protein expressed by the cells at steady state. The degradation activity was lost upon Bcl-2 depletion and was reconstituted by adding recombinant Bcl-2. Ineffective extracts from cells that constitutively do not express Bcl-2 acquire full degradation activity by adding recombinant Bcl-2 protein. We conclude that Bcl-2 is necessary to activate the degradation complex on the relevant RNA target.

The bcl-2 gene plays an important role in cell survival, differentiation, and oncogenesis (1,2). A large number of human cancers, including solid tumors, express abnormally high amounts of Bcl-2 protein (3). Bcl-2 expression has prognostic relevance (4) as well as importance for therapy, since it influences cell sensitivity to chemo-and radiotherapy (5). Despite extensive efforts, the molecular mechanisms leading to increased expression of bcl-2 in human cancer have not been elucidated (6,7). Both transcriptional and post-transcriptional events might contribute to determining the rationale for overexpression of Bcl-2 in cancer cells (8,9). Many attempts have been made to develop biological or pharmacological means of controlling bcl-2 expression (10 -12), including the use of antisense compounds (13,14) currently under clinical investigation for chemosensitizing activity (15).
In the regulation of steady-state levels of transcripts in cells, mRNA decay has received much emphasis recently (16). The role of cis elements flanking the coding region has been clearly recognized (17). The adenine/uridine-rich elements (AREs), 1 identified in a large number of genes, including immediateearly genes (18), cell-cycle genes (19), and cytokines and their receptors (20), are discrete nucleotide structures located in the 3Ј-UTR of mRNAs. The importance attributed to ARE results from the discovery of its specific fine-tuning of steady-state levels of many transcripts effectively regulating the level of the encoded protein products in cells (21,22). Actually, ARE functions as the binding site for a number of proteins, the adenine/ uridine binding elements or AUBPs (23)(24)(25) that recruit a complex of exonucleases, termed exosome (26).
In previous studies we described an ARE located in the 3Ј-UTR of bcl-2 mRNA (b-RNA) whose nucleotide structure is similar to the ARE in GM-CSF mRNA (27) and is required for bcl-2 mRNA decay in vivo (28). We have also shown that the activation of an apoptotic program specifically induces AREdependent bcl-2 mRNA degradation (29).
Until now, specific ribonuclease activity in extracts of cells from different lineages has not been evaluated using a transcript carrying the ARE motif. It is possible that the nuclease specificity or efficiency in different lineages regulates the amount of the Bcl-2 protein. It might be of practical importance to understand the determinants of mRNA decay of genes such as bcl-2, expressed at different levels in tissue-specific programs (30) or under different metabolic conditions (31). Cellfree degradation assays have been very useful in analyzing mRNA turnover (32,33), and important findings have been made in recent years by taking advantage of such systems (34,35,26).
Here, we investigated the elements that regulate either the decay rate of the ARE-containing b-RNA in cis or in trans, taking advantage of an in vitro degradation assay. In addition to defining the essential cis elements required for the exonuclease activity, an essential role was found for Bcl-2 protein in activating the degradation machinery on the relevant target.
ics, and 10% fetal calf serum added under standard conditions. Viability was monitored by the trypan blue exclusion assay.
Plasmids-The 5Ј-primer AGATCTAGTCAACATGCCTGC and the 3Ј-primer GGATCCGGTGATCCGGCCAACAAC (flanked at the 5Ј-end by BglII and BamHI restriction sites, respectively) were used in PCR to amplify the 396-bp U1 segment containing the 3Ј-UTR ARE sequence from the human bcl-2 cDNA fragment (48). This segment was cloned in the TA cloning site of the pCRII plasmid according to the TA Cloning Kit specification (Invitrogen Corp., San Diego, CA), to produce the pCR-U1 plasmid. A fragment from this plasmid amplified with the 5Ј-primer GACCATGGGTCGAATCAGCTATTTAC and the 3Ј-primer GACCCGGGGATTTCCAAAGACAGGAG was subcloned in pCRII to produce the pCR-U3 plasmid used for in vitro transcription of b-ARE. pCR-U1 was also amplified with the following pairs of primers: 5Јprimer AGATCTAGTCAACATGCCTGC/3Ј-primer TTCGACGTTTTG-CCTGAAGACT; and 5Ј-primer CAAAACGTCGAACGACCACTAATTG-CCAAGC/3Ј primer GGATCCGGTGATCCGGCCAACAAC. The PCR products are partially complementary, so they were annealed and subsequently amplified with the external primers to generate the ⌬ b-RNA fragment. It was cloned in plasmid pCRII to produce plasmid pCR-U2.
A 394-bp segment of the human c-myc 3Ј-UTR, was obtained by PCR amplification using 5Ј-GCGGGGCCCACTTTTTTATGCTTACCATC-3Ј and 5Ј-GCGCCCGGGCAATAGAAAAAAATCAACTT-3Ј as forward and reverse primers, respectively. The amplification product of the myc 3Ј-UTR segment was inserted into the cloning vector pGEMT, provided with the easy vector system (Promega Italia, Milan, Italy), to obtain pGEMT-MYC-RNA.
A DNA fragment containing the 3Ј-UTR of IL-2 (590 -780 bp) was subcloned between the HincII and KpnI sites of pCY2 as described previously (49).
The control plasmids pTRI-IGFR and pTRI-GAPDH (Ambion, Austin, TX) were linearized with the restriction enzyme BamHI (Amersham Biosciences) and used for in vitro transcription reactions with SP6 RNA polymerase (Promega Italia, Milan, Italy). The pTRI-GAPDH-Rat antisense control template contains a 316-bp fragment of the rat glyceraldehyde-3-phosphate dehydrogenase gene that is derived from exons 5-8. The pTRI-IGFR-human transcription template contains a 276-bp cDNA fragment of the human insulin-like growth factor-1 receptor gene, which spans exons 7-8 in the antisense orientation. The pSS transcription plasmid, obtained by cloning rat somatostatin cDNA into plasmid pSP65 (Promega Italia, Milan, Italy), was kindly provided by Dr. A. Torsello.
Preparation of Cell Lysates and Immunodepletion-Cells (20 ϫ 10 6 ) were centrifuged at 1200 rpm for 10 min at 4°C; the pellet was washed three times with phosphate-buffered saline at 4°C and resuspended in 0.25 ml of Buffer A (10 mM Tris acetate, 40 mM potassium acetate, 3 mM magnesium acetate, pH 7.6, 5% glycerol; after autoclaving, we added 2 mM dithiothreitol, 0.57 mM phenylmethylsulfonyl fluoride, 0.1 mM sodium vanadate, 50 mM sodium fluoride, 2 g/ml leupeptin, and 1 g/ml aprotinin). The cells were vortexed, incubated at 4°C for 30 min, and lysed by sonication. The lysate was centrifuged at 12,000 ϫ g for 20 min at 4°C, and the supernatant was used immediately or frozen. Protein concentrations were measured using the Micro BCA Protein Assay (Pierce). S100 preparation from Jurkat cells and immunodepletion were performed as described previously (26).
In Vitro Transcription-Transcripts were synthesized in vitro using SP6 or T7 (for pGEMT-MYC-RNA) RNA polymerase according to the manufacturer's instructions (Promega Italia) incorporating DIG-labeled NTPs (Roche Diagnostics, Milan, Italy). The 32 P transcripts were obtained as described previously (49).
In Vitro Degradation System-The degradation reaction mixture contained 1 g of lysate, 10 g of carrier tRNA, and 5 ng of transcript for each time point. Samples of 3 l of degradation reaction mixture were taken and mixed into 10 l of urea buffer (7 M urea, 2% sodium dodecyl sulfate, 0.35 M NaCl, 10 mM EDTA, 10 mM Tris-HCl, pH 7.6, and 1 mg/ml carrier tRNA). 10 l of phenol/chloroform (1:1), 30 l of deionized formamide, and 5.75 l of gel loading buffer (50% glycerol, 1 mM EDTA, pH 8.0, 0.25% bromphenol blue, 0.25% xylene cyanol) were added, the mixture was vortexed, heated at 70°C for 5 min, electro-phoresed on a 5% polyacrylamide gel containing 8 M urea, and transferred to a nylon filter (Amersham Biosciences) by a Semi-Dry transfer cell (Bio-Rad). The DIG-labeled transcripts were detected with the digoxigenin system (Roche Diagnostics). For competition experiments, 0.3 g of cell extract were incubated on ice for 10 min with unlabeled competitor RNAs in 100-fold molar excess over labeled transcript before the incubation with DIG-labeled RNA.
In the experiments using rBcl-2, corresponding to amino acids 1-205 of Bcl-2 of human origin (Santa Cruz Biotechnology, Santa Cruz, CA), the cellular extracts (4 g) were preincubated with rBcl-2 at 25°C for 10 min and then incubated with the transcripts. Recombinant Myc protein (rMyc), corresponding to amino acids 1-262 of Myc of human origin (Santa Cruz Biotechnology), was used as control. In vitro analysis of 32 P-labeled RNA decay was described previously (49).
Determination of bcl-2 mRNA Decay in Cells-Cells were treated with the transcription blocker 5,6-dichloro-1-␤-D-ribofuransoylbenzimidazole (DRB) (Sigma-Aldrich) at a final concentration of 20 g/ml for 6, 7, 8, 10, 12, and 14 h. Total cellular RNA was extracted at the indicated times using the NucleoSpin RNA II columns (Macherey-Nagel, Dü ren, Germany) according to the manufacturer's instructions. The RNA was treated with RNase-free DNase (Invitrogen) in order to avoid possible DNA contamination. The purity and integrity of the RNA were checked spectroscopically and by gel electrophoresis, respectively, before carrying out the analytical procedures. Levels of bcl-2 and GAPDH mRNAs were determined by the semi-quantitative RT-PCR method. For synthesis of cDNA, 1 g of total RNA was reverse-transcribed with Random Hexamers (Promega Italia) using standard manufacturer's conditions. The RT reaction was split into 2 PCR reactions with 2 sets of specific primers. The PCR primer sequences are as follows: bcl-2 (139-bp PCR product) forward: 5Ј-GTCATGTGTGTGGAGAGCGT-3Ј and reverse: 5Ј-ACAGTTCCACAAAGGCATCC-3Ј; GAPDH (195-bp PCR product) forward: 5Ј-ATGACAACTTTGGTATCGTG-3Ј and reverse: 5Ј-CAGTGAG-CTTCCCGTTCA-3Ј. The PCR reactions were carried out in a total volume of 100 l in a Hybaid Omn-E thermal cycler under the following conditions: denaturation at 94°C for 1 min (94°C for 2 min in the initial cycle), annealing at 60°C for 1 min, and elongation at 72°C for 1 min, followed by a final step of 7 min at 72°C. For every oligonucleotide pair, a preliminary analysis was conducted to define the appropriate range of cycles consistent with an exponential increase in the amount of the DNA product. For each sample 20 l of PCR amplification product were analyzed on a 3% agarose gel and stained with ethidium bromide. A standard DNA molecular weight ladder (GeneRuler 100-bp DNA Lad-der, M-Medical, Genenco, Firenze, Italy) was run to provide appropriately sized markers. The amount of each amplification product was determined by densitometer analysis. The mRNA half-life was determined from a semilogarithmic transformation of the densitometric data plotted against the time of incubation with DRB.
Western Blot Analysis-Samples of 4 ϫ 10 6 cells were collected by low-speed centrifugation and washed twice in ice-cold phosphate-buffered saline. The cell pellet was resuspended in 60 l of lysis buffer (ice-cold radioimmune precipitation assay buffer with freshly added protease inhibitors), vortexed for 3 s, and incubated on ice for 30 min. The lysates were centrifuged at high speed for 20 min at 4°C. 5 l of the supernatant was removed for protein determination, transferred to a microcentrifuge tube, mixed with reducing buffer, and heated to 99°C for 2 min. Equal amounts of proteins were analyzed by 12% SDS-PAGE, blotted onto polyvinylidene difluoride membranes (Immobilon P, Millipore, Bedford, MA) in a Bio-Rad Trans-blot apparatus at 100 V for 90 min. Blots were processed by an enhanced chemoluminescence (ECL Plus) detection kit as instructed by the supplier (Amersham Biosciences). The blots were probed with a mouse anti-Bcl-2 (Santa Cruz Biotechnology) and a rabbit anti-actin (Sigma-Aldrich) antibody, followed by a horseradish peroxidase-conjugated secondary antibody.
RNA Electrophoretic Mobility Shift Assay (REMSA)-REMSAs were performed by incubating the radiolabeled b-RNA and ⌬ b-RNA riboprobes (5 ϫ 10 5 cpm) with 100 ng of rBcl-2, which were then digested for 20 min at room temperature with 5 units of RNase T1 (La Roche Ltd.), which cuts RNA downstream to guanosine residues into a number of fragments. Samples were separated on a native polyacrylamide gel (6% polyacrylamide/bisacrylamide, 60:1).
UV Cross-linking Assay (UVCA)-rBcl-2 protein was incubated with the 32 P-labeled b-RNA and ⌬ b-RNA riboprobes (10 9 cpm/g, 3 ϫ 10 5 cpm) in the presence of 0.5 mg/ml heparan sulfate and 2 g of tRNA in a microplate (total volume, 10 l) at room temperature for 10 min. The RNA-protein complexes were cross-linked on ice by exposure to UVC for 5 min with 3000 milliwatts/cm 2 in a Stratalinker 1800 (Stratagene, La Jolla, CA). Samples were incubated with RNase A (1 l, 1 mg/ml) for 30 min at 37°C to digest unbound RNA. Proteins were separated by 12% polyacrylamide-SDS gel electrophoresis (SDS-PAGE) under reducing conditions. The images were generated with a Cyclone imager (Packard BioScience).

ARE-dependent b-RNA Degradation in Vitro-
We previously reported that the bcl-2 ARE located in the 3Ј-UTR is a cisacting, regulatory element that can target a reporter mRNA for degradation in vivo (29). To examine ARE-directed RNA decay in a cell-free system, 3Ј-UTR fragments of bcl-2 obtained by PCR were cloned in a transcription vector and used for in vitro synthesis of transcripts (Fig. 1) that were labeled by incorporation of the precursor DlG-UTP. We chose to utilize transcripts that were neither capped at the 5Ј-end with 7 mGpppG nor polyadenylated at the 3Ј-end. Our rationale was to examine ARE-directed RNA decay in its simplest form in the absence of these structures, since ARE-directed degradation of an mRNA   Fig. 3. Bcl-2 levels were evaluated by Western blotting (Fig.  6B) and shown by others. The degree of RNA decay was determined by autoradiography followed by scanning laser densitometry. Data are representative of three independent experiments. CELL   can occur in a cap-independent fashion (50 -52) and in the absence of a poly(A) tract (26,53). Thus, at least some of the steps in the ARE-directed mRNA decay pathway can be uncoupled from processes that involve the cap structure and poly(A) tract. For decay assays, unfractionated cell lysates, derived from exponentially growing cells, were mixed with the relevant transcripts and incubated at 37°C for various times. RNA was purified and fractionated in gels for detection and analysis. The in vitro assay shown here, although reproducing findings obtained with the S100 preparations, does not require biochemical fractionations, making the comparisons of extracts from different cell sources easier. Fig. 2A shows that b-RNA was rapidly degraded by a soluble extract prepared from the DOHH 2 follicular B-cell lymphoma cell line. The half-life of 30 min was calculated by plotting the intensity of the bands analyzed by densitometry (lanes 1-5 and right panel). In the same assay, the GAPDH control transcript was not significantly degraded, the calculated half-life was longer than 2 h (lanes 6 -10 and right panel). The specificity of the reaction was further verified by adding to the same reaction the b-RNA and the ARE-negative IGFR control transcript for the decay assay (Fig. 2B). The two transcripts were degraded with different kinetics. The calculated half-life for the b-RNA was 30 min while that for the IGFR transcript was more than 120 min (Fig. 2B, right panel). Multiple control RNAs were used in Fig. 2, A and B in order to provide substantial evidence that ARE-negative, synthetic RNAs are not significantly degraded under these assay conditions. We conclude from Fig. 2, A and B that an ARE-containing transcript is specifically degraded in this in vitro system.
To directly investigate the contribution of the bcl-2 ARE (b-ARE) to b-RNA decay in vitro, the central core of the 107-nt ARE was deleted ( ⌬ b-RNA). ⌬ b-RNA was incubated with the DOHH 2 extract for the decay assay. As shown in Fig. 2C, the ⌬ b-RNA lacking the consensus sequence degraded more slowly (half-life, 75 min) than the wild-type b-RNA (half-life, 30 min). This result suggests that, while the b-ARE is required for b-RNA decay, ARE-flanking regions in the 3Ј-UTR contribute to the full destabilizing potency of the bcl-2 3Ј-UTR.
The relevance of b-ARE shown in these studies prompted us to evaluate its decay kinetics. The b-ARE transcript and an unrelated transcript were prepared and incubated with the DOHH 2 extract. Fig. 2D shows that the b-ARE was degraded much faster than the control RNA. The calculated 40 -45 min half-life of b-ARE shows the crucial relevance of b-ARE for the nuclease activity (Fig. 2D, right panel). There was no measurable degradation of both transcripts when incubated under the same conditions in buffer alone.
The effects of competitor RNAs on b-RNA degradation in vitro were next evaluated. Extracts were preincubated with up to a 100-fold molar excess of unlabeled b-RNA or with equivalent amounts of a 500-nt rat somatostatin (rSS) synthetic RNA   Table I summarizes these results together with the measured half-lives of all the transcripts used. Taken together, these results suggest that the b-ARE may function through interactions with one or more factors that specifically recognize the ARE.
Decay Activity of Extracts from Different Cell Lines-Using this cell-free mRNA decay system, b-RNA decay was investigated using extracts from a variety of human cell lines representing a spectrum of cell types. Our two goals were, first, to validate the cell-free system by comparing the relative decay rates obtained in vitro with those obtained in whole cells and, second, to determine if cell type-specific degradation of b-RNA occurs. The decay kinetics of RNA in extracts derived from different cell types, evaluated during a 60-min time course, are shown in Fig. 3. Although extracts from all the cell lines degraded b-RNA faster than the control IGFR RNA (half-life Ͼ120 min), the b-RNA half-lives varied widely as shown at the left in Fig. 3.
Extracts from the three cell lines containing a t (14, 18) translocation (DOHH 2 , DHL-4, and Karpas 422) and the two neuroblastoid cell lines (LAN-5 and SHSY5Y) displayed the fastest decay rates (b-RNA half-life: 25-30 min). Extracts from the B-and T-lymphoid cell lines (Namalwa, Raji, and MOLT-4) displayed intermediate decay rates (b-RNA half-life: 50 -60 min), and extracts from myeloid (HL60) and epithelial (MCF-7) cell lines displayed the slowest decay rates (b-RNA half-life: 70 -80 min). These differences were not due to differences in the quality of the extracts, since at least three preparations for each cell line were assayed, and the decay of the stable IGFR RNA was included as a standard control (Table II).
For comparison, the stability of bcl-2 mRNA was examined in vivo by treating cells with DRB, a selective inhibitor of RNA polymerase II, which can prevent transcriptional initiation with little effect on mRNA stability (54). An RT-PCR assay was used to measure levels of bcl-2 and GAPDH mRNAs versus time. The logarithmic phase of bcl-2 and GAPDH cDNA amplification was determined by PCR using progressive numbers of cycles (Fig. 4B).
Based upon these results, 20 cycles for GAPDH and 30 cycles for bcl-2 were employed for quantification. Fig. 4A shows the decay kinetics of bcl-2 mRNA as obtained from the semiquantitative RT-PCR assay along with the data quantified by densitometry (right panel). Similar to the in vitro results, the DOHH 2 B-cell line and the neuroblastoid SHSY5Y cell line displayed the shortest bcl-2 half-lives of the group (ϳ3 h), while HL60 and MCF-7 cells displayed the longest in vivo half-lives (6 and 8 h, respectively). We conclude that the cell-free system reproduces the rank order of bcl-2 mRNA decay rates exhibited among the different cell types.
Bcl-2 Is Required for the Rapid Degradation of b-RNA-First we investigated the role of the Bcl-2 protein in regulating b-RNA degradation using the cell-free system. While extracts prepared from Daudi cells that do not express the Bcl-2 protein (Ref. 55 and Fig. 6B) were ineffective to degrade the b-RNA ( Fig. 5A and B, lanes 1-4), they were able to degrade myc-RNA (Fig. 5A). Most relevant, the addition of rBcl-2 protein to the degradation mixture makes Daudi cell extracts able to degrade b-RNA (Fig. 5B, lanes 5-8). In contrast, the addition of rMyc protein was unable to activate the degradation complex (Fig.  5B, lanes 9 -12). Therefore, findings in Fig. 5, A and B show that the Daudi cell lysates are endowed with a functional degradation machinery although they require Bcl-2 protein to degrade b-RNA.
Three independent factors further substantiated the specificity of the role of Bcl-2 in b-RNA degradation. First, extracts from Jurkat cells over-expressing Bcl-2 degrade b-RNA faster than extracts from mock-transfected cells (Fig. 6A). Second, Jurkat S100 extracts from which Bcl-2 was removed by immunodepletion (Fig. 7C), were not capable of degrading the radiolabeled b-RNA (Fig. 7A). In contrast, another ARE-containing RNA (IL-2 3Ј-UTR) was efficiently degraded by Bcl-2-depleted S100. Furthermore, while 32 ng of rBcl2 were able to reconstitute the b-RNA degradation activity of Bcl-2-depleted extracts, the same amount of rMyc was ineffective (Fig. 7B). Third, the b-RNA transcript but not the core-deleted ⌬ b-RNA can directly bind rBcl-2 in the REMSA assay, as shown in Fig. 8A. The UVCA confirmed the dose-dependent binding of the Bcl2-protein to its own RNA. The rBcl-2 binding to the mutated RNA deleted of the core ARE element is significantly reduced (Fig.  8B). The physical association of Bcl-2 with b-RNA might be the basic elements for the activation of the degradation complex.
These findings also indicate the equivalence between the cell-free system based on the S100 extract-radiolabeled b-RNA and the modified method as proposed in this work. Taken together the data provide strong evidence for a specific role of the Bcl-2 protein to activate the degradation complex on the relevant RNA. DISCUSSION The turnover of mRNA is an important means of regulating both the level and the timing of gene expression (16) along with the metabolic state of the cells and during differentiation (1,2). Along this line, the amount of cellular Bcl-2 protein is finely tuned, and its deregulation can be associated with the pathogenesis of many human diseases and with the responsiveness to cancer therapy (3,5).
In a previous work we showed that the rapid decay of b-RNA in the cells is dependent on a conserved ARE motif (27). Although the ARE-dependent decay of mRNAs has been extensively studied, the efficiency of the elements able to regulate the decay of a given transcript both in cis and in trans in different cell lines has not been evaluated in detail. Here we have investigated some mechanistic aspects by which the b-RNA turnover is regulated.
We used a cell-free system in which the decay kinetics of the synthetic b-RNA are regulated by the ARE motif to investigate the b-RNA decay activity in extracts from different cell types. The transcripts utilized for our degradation studies do not contain a poly(A) tract, since we decided to focus on the degradation of the mRNA body which follows deadenylation (50). Indeed, ARE-mediated mRNA deadenylation in the eukaryotic cells can be uncoupled from degradation of the mRNA body (53,56) as recently discussed in detail (26).
We found that unfractionated cellular extracts degraded ARE-containing b-RNA faster than other transcripts of similar length not containing the ARE motif. Since ⌬ b-RNA was still degraded faster than control RNAs, it could be argued that the regions flanking the ARE motif also contribute to the decay rate. Specificity of the decay system was further demonstrated by mixing b-RNA with unlabelled RNA in the same reaction. The rank order of RNA degradation is similar in vitro and in whole cells. These observations suggest that the soluble-based biochemical system performs with the same fidelity as a polysome-based system and can thus reconstitute physiologically relevant mRNA decay processes.
In vitro decay assays might prove invaluable for investigating the mechanistic aspects of RNA turnover and for assisting studies of alterations associated with metabolic or pathological conditions (32,33).
By using different cell lines, we thought to get some insight into the mechanism by which the AUBPs and the exosome complex might modulate the rate of degradation of b-RNA. Moreover, understanding of how each cell line can degrade b-RNA at a different rate may shed light on how the degradation complex distinguishes its substrates and target mRNAs specifically.
It was found that the degradation rate depends on the cell lineage. The half-life of b-RNA was between 25-30 min and 70 -80 min and was highly reproducible in different preparations from each cell type. The highest decay rate was exhibited by extracts from the three t (14,18) cell lines and the two neuroblastoid cell lines. These cells typically contain the highest levels of Bcl-2 protein at the steady state, as reported by other authors (57)(58)(59)(60)(61) and confirmed by Western blotting (Fig. 6B). A second group of extracts obtained from the t (14, 18)-negative B-cell lines or from one T-cell line, showed slower decay rates, the b-RNA half-life being 50 -60 min. In these cells the level of Bcl-2 is constitutively lower than the Bcl-2 levels in the cell types described above. The slowest decay was shown by extracts from cells that express hardly detectable amounts of Bcl-2. The halflives of the control transcripts were regularly longer than 120 min, consistent with the long half-life in cells (62).
The differences in activities between fast and slow degrada- tion observed with the in vitro RNA decay assays reflect the in vivo situation. These findings indicate the possibility that metabolic conditions of the cells might regulate the rate of the mRNA decay.
The cell-free assay was used to test the concept of whether the level of Bcl-2 controls the decay rate of b-RNA. Daudi cells do not express Bcl-2 and their extracts are not able to degrade b-RNA quickly in the cell-free assay. The addition of recombinant Bcl-2 to the Daudi cell extracts fully activates the degradation process. Therefore, Daudi cell extracts contain the entire enzymatic machinery, capable of degrading myc-RNA, although requiring exogenous Bcl-2 for the b-RNA degradation.
Six distinct pieces of evidence indicate that Bcl-2 specifically activates b-RNA degradation: (a) extracts from Daudi, a cell line not expressing Bcl-2, degrade myc-RNA, while having no effect on b-RNA stability, (b) Daudi extracts supplemented with rMyc protein do not degrade b-RNA, but by Bcl-2 protein, they acquire full-degrading activity on b-RNA, (c) cellular extracts immunodepleted of Bcl-2 lose b-RNA degradation activity, (d) rBcl-2 reconstitutes the degrading activity in Bcl-2-depleted extracts, (e) Jurkat cell extracts, genetically modified to overexpress Bcl-2, degrade b-RNA faster than extracts from mocktransfected cells, and (f) rBcl-2 protein binds the b-RNA in vitro.
Together, these data show that Bcl-2 is crucial in permitting the ARE-dependent degradation of bcl-2 RNA and is essential in regulating the decay rate of its messenger. The most recent models of the ARE-dependent rapid degradation of unstable mRNAs depict the 3Ј-to-5Ј degradation by the exosome complex recruited by the AUBPs (24,26). This stimulating model does not involve the different rate of degradation of a given transcript by different cell types. Most importantly, previous models have not been focused on how the exosome, recruited by the AUBPs, might degrade the relevant RNA among a large pool of messengers endowed with the ARE motif. Here, evidence is provided on the role of the Bcl-2 protein to address the enzymatic activity to the correct target. Moreover, a negative feedback mechanism based upon the concentration of the Bcl-2 protein might not be excluded.
The molecular mechanism by which Bcl-2 regulates the fast degradation of its own RNA remains unknown. The specificity of the reaction might suggest the displacement of a stabilizing factor or, alternatively, the recruitment of a destabilizing factor caused by the presence of Bcl-2 within the b-RNA. Although the findings discussed here may have implications for other AREcontaining RNAs, the molecular mechanism by which a protein can regulate its own RNA degradation remains to be determined.