p53 inhibits mRNA 3&prime - Nature

Oncogene (2011) 30, 3073–3083

& 2011 Macmillan Publishers Limited All rights reserved 0950-9232/11 www.nature.com/onc

ORIGINAL ARTICLE

p53 inhibits mRNA 30 processing through its interaction with the CstF/BARD1 complex FI Nazeer, E Devany, S Mohammed, D Fonseca, B Akukwe, C Taveras and FE Kleiman Chemistry Department, Hunter College, City University of New York, New York, NY, USA

The mechanisms involved in the p53-dependent control of gene expression following DNA damage have not been completely elucidated. Here, we show that the p53 C terminus associates with factors that are required for the ultraviolet (UV)-induced inhibition of the mRNA 30 cleavage step of the polyadenylation reaction, such as the tumor suppressor BARD1 and the 30 processing factor cleavage-stimulation factor 1 (CstF1). We found that p53 can coexist in complexes with CstF and BARD1 in extracts of UV-treated cells, suggesting a role for p53 in mRNA 30 cleavage following DNA damage. Consistent with this, we found that p53 inhibits 30 cleavage in vitro and that there is a reverse correlation between the levels of p53 expression and the levels of mRNA 30 cleavage under different cellular conditions. Supporting these results, a tumor-associated mutation in p53 not only decreases the interaction with BARD1 and CstF, but also decreases the UV-induced inhibition of 30 processing, all of which is restored by wild-type-p53 expression. We also found that p53 expression levels affect the polyadenylation levels of housekeeping genes, but not of p21 and c-fos genes, which are involved in the DNA damage response (DDR). Here, we identify a novel 30 RNA processing inhibitory function of p53, adding a new level of complexity to the DDR by linking RNA processing to the p53 network. Oncogene (2011) 30, 3073–3083; doi:10.1038/onc.2011.29; published online 7 March 2011 Keywords: p53; cleavage-stimulation factor; BARD1; DNA damage; RNA processing

Introduction The p53 gene is the most commonly mutated target in human tumors. Activation of p53 affects the expression level of a large set of genes and mediates several cellular responses such as DNA repair, cell cycle arrest and/or apoptosis (Vogelstein et al., 2000; Levine et al., 2006). Although it is well established that p53 is a transcriptional Correspondence: Dr FE Kleiman, Chemistry Department, Hunter College, City University of New York, 695 Park Avenue, New York, NY 10065, USA. E-mail: [email protected] Received 3 May 2010; revised 7 January 2011; accepted 18 January 2011; published online 7 March 2011

regulator, transactivation-independent functions of p53 have been described (reviewed by He et al., 2007; Takwi and Li, 2009). For example, certain microRNAs are transactivated by p53, and these microRNAs cause dramatic changes in gene expression, offering an indirect p53-mediated control of gene expression at the posttranscriptional level (Chang et al., 2007). It has been described that the induction of p53 expression upon ultraviolet (UV) treatment is associated with changes in the levels of total poly(A) þ mRNA (Ljungman et al., 1999; McKay and Ljungman, 1999). Interestingly, this p53associated changes in poly(A) þ RNA levels might also be functionally related to the UV-induced inhibition of mRNA polyadenylation (Kleiman and Manley, 2001). As mRNA poly(A) tails are important for the regulation of mRNA stability, it is possible that these changes of poly(A) þ mRNA levels might represent another mechanism of p53-mediated control of gene expression. The 30 end of the mRNA is processed by the cleavage of the mRNA followed by the addition of a nontemplated polyadenylated tail, which in mammalian cells is of approximately 200–300 adenosines. The assembly of the cleavage/polyadenylation machinery requires specific signal sequences in the mRNA precursor as well as interactions of a large number of protein factors (reviewed by Mandel et al., 2008). It has been shown that the regulation of mRNA 30 end formation can have significant roles in cancer (Kleiman and Manley, 2001; Topalian et al., 2001; Scorilas, 2002; Rozenblatt-Rosen et al., 2009). Most importantly, alternative mRNA cleavage and polyadenylation changes the length of the 30 -untranslated region and regulates gene expression of different mRNAs in cancer cells (Mayr and Bartel 2009; Singh et al., 2009) and during cell differentiation (Sandberg et al., 2008; Zlotorynski and Agami, 2008; Ji et al., 2009). Cleavagestimulation factor (CstF) is one of the essential 30 processing factors and is most likely active as a dimer of an heterotrimer, consisting of three protein factors called CstF3, CstF2 and CstF1. CstF2 interacts directly with the mRNA, and cells deficient in CstF2 undergo cell cycle arrest and apoptotic death (Takagaki and Manley, 1998). Both the CstF1 and CstF3 subunits interact specifically with the C-terminal domain of RNA polymerase II, likely facilitating the RNA polymerase II-mediated activation of 30 end processing (McCracken et al., 1997; Hirose and Manley, 1998). After DNA damage, mRNA 30 processing is inhibited as a result of

Role of p53 in mRNA 30 processing FI Nazeer et al

3074

SN

)

PD

5%

GST-BARD1

t(

t( pu

SN PD

in

in

GST-CstF1

pu

5% t(

BRCT

BARD1 (1-777)

pu

ankyrin repeats

RING

5%

)

It has been shown that p53 interacts with BARD1 independently of BRCA1 to induce an apoptotic response to genotoxic stress (Feki et al., 2005). A

in

Results

germline mutation in BARD1 (Gln564His) reduces the binding to p53 and induction of apoptosis (IrmingerFinger et al., 2001). BARD1 can also interact with the 30 processing factor CstF1, inhibiting mRNA 30 processing and linking it to the DNA damage response (DDR) (Kleiman and Manley, 1999). Interestingly, the Gln564His mutation of BARD1 also reduces the binding of CstF1 to BARD1, interfering with the role of BARD1 in mRNA 30 processing (Kleiman and Manley, 2001). Taken together, these studies suggest a functional interaction between BARD1, CstF1 and p53 under DNA-damaging conditions. To further investigate this possibility, we first examined the physical association of p53 with BARD1 and CstF1 using the recombinant proteins shown in Figure 1a. The results showed that both full-length His-p53 and the C-terminal fragment of p53 interacted directly in vitro with not only GST-BARD1 (Figure 1b, lane 6), but also GST-CstF1 (lane 3). Neither CstF1 nor BARD1 bound to the two

)

CstF/BARD1/BRCA1 complex formation (Kleiman and Manley, 1999) and of the proteasome-mediated degradation of RNA polymerase II (Kleiman et al., 2005), suggesting the existence of possibly redundant mechanisms to explain the inhibitory effect of UV irradiation. Herein, we examine the functional interaction of p53 with BARD1 and CstF1 and its effect on mRNA 30 processing as well as on the polyadenylation of cellular RNAs. Our results provide new insights into p53 function and into the mechanisms behind the regulation of mRNA 30 processing in different cellular conditions.

GST

SN

PD

7xWD-40

-p53

CstF1 (1-397) DNA transactivation binding regulation

-p53 (1-293)

p53 (1-397) p53 (1-293)

-p53 (94-293)

p53 (94-393) p53 (94-293)

-p53 (94-393) 1

2

3

4

input +

α-CstF2 CoIP IP SN -

pull-down GST-BARD1 His-p53 SN

PD

UV:

no His-p53 SN

PD

SN

PD

SN

-

5

6

7

8

IP +

SN +

PD

- Topo II - CstF1

- BARD1

- p53 1

2

3

4

5

6

7

- CstF2

8

- p53 1

pull-down GST-CstF1 His-p53 SN

1

PD

2

3

PD

4

SN

5

PD

6

SN

7

2

3

4

5

6

IP +

SN +

α-p53 CoIP

no His-p53 SN

9

PD

UV:

input +

IP -

SN -

- BARD1

- Topo II

- p53

- BARD1 - CstF2

8

- p53 1

2

3

4

5

6

Figure 1 CstF1 interacts with both BARD1 and p53 to form a protein complex. (a) Diagram of BARD1, CstF1 and p53 derivatives. (b) Requirement of p53 C-terminal domain for both CstF1 and BARD1 interaction. Recombinant His-p53 or the indicated His-p53 derivatives were incubated with purified GST, GST-CstF1 or GST-BARD1. Protein samples were treated with RNase A. Bound proteins were eluted and analyzed by western blot with anti-p53 antibodies. Five percent of His-p53 or His-p53 derivatives used in the reaction are shown as input. (c) CstF1, BARD1 and p53 can coexist in complexes. Either GST-BARD1 or GST-CstF1 was immobilized in glutathione-Sepharose beads to test its ability to bind CstF1/BARD1 and p53 from NE of HeLa cells, while increasing concentrations of recombinant His-p53 (5, 10 and 20 ng) were added. Bound proteins were detected by immunoblotting with CstF1 and p53 antibodies. The pellets (PD) and the supernatant (SN) were analyzed. (d) p53, CstF and BARD1 co-immunoprecipitate from NEs of MCF7 cells treated with UV irradiation. CstF and p53 co-immunoprecipitation is irrespective of UV irradiation. Co-immunoprecipitation assays were performed using NEs of cells exposed or not to UV irradiation and allowed to recover for 2 h as described before. NEs were immunoprecipitated with either anti-CstF2, to ensure that any detected interactions were between p53 and intact CstF, or anti-p53 antibodies. Protein samples were treated with RNase A. Equivalent amounts of the pellets (IP) and the supernatants (SN) were resolved by sodium dodecyl sulfate–polyacrylamide gel electrophoresis and proteins were detected by immunoblotting with antibodies against BARD1, CstF2 and p53. Antibodies against Topo II were used as a control of specificity. Positions of Topo II, CstF2 and p53 are indicated. Twenty percent of the NE used in the immunoprecipitation reaction is shown as input. Oncogene


3075

derivatives that lacked the C-terminal domain of p53 or to GST alone (lane 9). Taken together, these results indicate that the C-terminal domain of p53, which has been described to have regulatory functions in the DDR (Sauer et al., 2008), constitutes the CstF1 and BARD1 interaction domain. To test whether p53 might also be part of the previously described CstF1/BARD1 complex, competition assays were performed by incubating GST-BARD1 or GSTCstF1 immobilized on glutathione beads with nuclear extracts (NE) of non-UV-treated HeLa cells and increasing amounts of His-p53 (Figure 1c). The effect of exogenously added recombinant p53 on the complex formation was easily assessed in this assay because HeLa cells, which are HPV-E6 transformed, have low levels of p53 expression (Wei, 2005). As shown in Figure 1c, increasing amounts of His-p53 increased both the amount of CstF1 ‘pulled down’ by GST-BARD1 (upper panel, lanes 1–8) and the amount of BARD1 ‘pulled down’ by GST-CstF1 (lower panel, lanes 1–8), indicating that p53 stabilized the CstF1/BARD1 complex. To examine the status of these p53/BARD1/CstFassociated complexes in NEs, we analyzed extracts from different cell lines by co-immunoprecipitation assays. As earlier studies had shown a link between these proteins and the UV-induced DDR, we extended these studies to DNA-damaging conditions. We used cells expressing different levels of p53, such as MCF7 cells (normal levels) and HeLa cells (low levels). HeLa cells were included in this study because most of the functional properties of the 30 processing machinery have been described in this biological system (Takagaki et al., 1989; Hirose and Manley, 1998; Kleiman and Manley, 1999, 2001). As described before, p53 accumulation was detected in samples from MCF7 cells exposed to UV treatment (Figure 1d, compare lanes 1 and 2). Interestingly, antibodies against CstF2 co-precipitated a significant amount of p53 independently of UV treatment (Figure 1d, upper panel, lanes 3 and 5). Similar results were obtained in the reciprocal co-immunoprecipitation analysis with p53 antibodies (lower panel). Interestingly, a significant amount of BARD1 co-precipitated with both CstF and p53 only in UV-treated samples (upper and lower panel, lane 5). The increased amount of p53 in the extracts of UV-treated cells solely cannot explain the increased association of these three proteins. Although similar amounts of p53 and CstF are immunoprecipitated by their own antibodies in samples exposed or not to UV, a complex formation with BARD1 is only detected in UV-treated cells. Similar results were obtained in the co-immunoprecipitation with BARD1 antibodies and with HeLa cells (not shown). Although these results do not show how many complexes p53 can form with CstF and BARD1, they clearly show that UV treatment induced the interaction between those three factors, supporting the idea that p53 and CstF might be simultaneous binding partners of BARD1. These studies also showed that the co-precipitation of p53 with CstF was irrespective of DNA damage, indicating that the formation of the CstF/p53 complex might be independent of BARD1.

To test whether p53 has a role in mRNA 30 processing through its interaction with CstF and BARD1, we performed in vitro RNA cleavage assays with limiting amount of NE of HeLa cells and addition of increasing amounts of different His-tagged p53 derivatives. As mentioned before, HeLa cells provided a good system to study exogenously added p53 as these cells have very low levels of p53 and most of the functional studies on 30 processing have been carried out in these cells. Interestingly, only increasing concentrations of either full-length His-p53 (Figure 2a, lanes 2–4) or the C-terminal fragment of p53 (lanes 14–16) into the reaction mix significantly reduced the 30 cleavage of the radiolabeled adenoviral L3 pre-mRNA. However, neither the two derivatives that lacked the C-terminal region of p53 (lanes 8–13) nor GST alone (lanes 5–7) had an effect on the cleavage reaction. These results indicate that the same region of p53 required for binding CstF1 and BARD1 (Figure 1b) is necessary for inhibiting mRNA 30 cleavage. Besides, our results indicate that the addition of recombinant His-p53 wild-type (WT) to NE of nontreated HeLa cells in a cell-free assay is sufficient to induce inhibition of mRNA 30 end processing, suggesting that the p53-mediated inhibition of 30 processing is independent of transactivational functions of p53. To further characterize the role of p53 in mRNA 30 processing, we performed siRNA-mediated knockdown of p53 in cells expressing different levels of p53 (HeLa and MCF7 cells) and then analyzed the UV-induced inhibition of 30 cleavage. Figure 2d shows that a 48 h siRNA treatment resulted in a substantial depletion of p53 (E90%) in NEs of both cell lines (lanes 5 and 6 and 11 and 12), independently of the UV treatment. Importantly, the expression levels of p53 in all the MCF7 samples were much higher than that in the HeLa samples analyzed. Consistent with previous results (Noda et al., 2000), UV treatment significantly increased accumulation of p53 and this was not affected by control siRNAs (Figure 2d). The depletion of p53 in HeLa cells abolished the UV-induced inhibition of mRNA 30 end cleavage (Figure 2b, compare lanes 4 and 6), indicating that p53 has an inhibitory effect on mRNA 30 cleavage under DNA-damaging conditions. Consistent with previous results from our lab, NEs from control siRNA-treated cells showed the UV-induced inhibition of 30 processing. As similar results were observed with the siRNA-mediated knockdown of BARD1 in HeLa cells (Kleiman et al., 2005), we propose that p53 together with the CstF/BARD1 complex might play a role in the UV-induced inhibition of 30 processing. Strikingly, NEs from cells expressing normal levels of p53 showed no significant levels of 30 cleavage using this assay (Figure 2c, lanes 1–4). The lack of detectable levels of 30 processing in NEs of MCF7 cells was irrespective of control siRNA (lanes 3 and 4) and UV treatments (lanes 2 and 4). These results are consistent with the possibility that high levels of p53 inhibit 30 mRNA processing. Supporting this idea, siRNA-mediated knockdown of p53 in MCF7 cells resulted in extracts exhibiting significant levels of 30 cleavage (lane 5) and UV-induced inhibition of 30 Oncogene


3076 NE

GST

His-p53

1-293

94-393

94-293

- AdL3 - 5’

1

2

3

4

5

RNAi: untreated cont + UV: +

6

7

8

9

11

12

RNAi: untreated

p53 -

10

+

UV: -

+

13

14

15

cont -

16

p53 +

-

+

- AdL3

- AdL3

- 5’ - 5’

1

RNAi: UV:

2

3

untreated +

4

5

HeLa cont +

6

1

p53 -

+

untreated +

2

3

MCF7 cont +

4

5

6

p53 -

+ - Topo II - p53

1

2

3

4

5

6

7

8

9

10

11

12

Figure 2 The C-terminal domain of p53 inhibits 30 cleavage in vitro and the levels of 30 cleavage correlate with expression levels of p53. (a) The C-terminal domain of p53 is necessary for inhibition of 30 cleavage. NEs from HeLa cells were preincubated with no addition or increasing amounts of the indicated His-p53 derivatives or GST proteins (40, 80 and 120 ng). After 15 min, L3 pre-mRNA was added and incubation continued for 90 min. RNAs were purified and analyzed by denaturing PAGE. Positions of pre-mRNA and the 50 cleavage product are indicated. (b) siRNA-mediated knockdown of p53 affects the UV-induced inhibition of 30 cleavage in HeLa cells. NEs from cells treated with p53/control siRNA and UV irradiation, and allowed to recover for 2 h, were analyzed for L3 pre-mRNA 30 cleavage. Positions of pre-mRNA and the 50 cleavage product are indicated. (c) siRNA-mediated knockdown of p53 induces 30 cleavage in MCF7 cells. NEs from cells were analyzed as described in (b). (d) Comparison of p53 expression levels in NE from HeLa and MCF7 cells. Protein concentrations in NE from siRNA treated/untreated cells were normalized by immunostaining with antibodies against Topo II.

cleavage (lane 6). It is important to highlight that p53 siRNA treatment did not completely deplete p53 in UVtreated MCF7 cells (Figure 2d, lane 12). Interestingly, the samples from p53 siRNA-treated MCF7 cells showed similar levels of both p53 expression and 30 cleavage to samples from control siRNA-treated HeLa cells (for protein expression: compare lanes 3 and 4 to lanes 11 and 12 in Figure 2d; and for 30 cleavage: compare lanes 3 and 4 in Figure 2b to lanes 5 and 6 in Figure 2c). These results indicate that p53 has an inhibitory effect on mRNA 30 cleavage and that this effect is dependent on the cellular levels of p53. Oncogene

We extended these studies to the colon carcinoma RKO cells that express normal levels of p53 (Figure 3c, lanes 1–6) and the isogenic cell line RKO-E6 that expresses low levels because it contains a stably integrated papilloma virus (Kessis et al., 1993; Figure 3c, lanes 7 and 8). The analysis of the UVinduced inhibition of 30 cleavage in samples from p53 siRNA-treated RKO cells showed similar results to that observed in samples from MCF7 cells treated in similar conditions (compare Figure 3a, lanes 1–6, to Figure 2c). Interestingly, samples of RKO-E6 cells behaved similarly to those of HeLa cells (compare Figure 3a, lanes 7


3077 RNAi: untreated UV: +

cont +

p53 -

+

-

H-1299

untreated + +

RKO-E6

a

RKO-E6

H-1299

H eL

RKO

NE

-

His-p53

NE

His-p53

- AdL3

- AdL3 - 5’

- 5’

3

4

5

6

7

RNAi: untreated UV: +

8

9

RKO cont +

10

11

p53 +

1

2

RKO-E6 H-1299 untreated + +

3

4

5

6

7

8

a

2

H eL

1

- Topo II - p53

1

2

3

4

5

6

7

8

9

10

11

Figure 3 The levels of 30 cleavage inversely correlate with expression levels of p53. (a) Samples from cells expressing different levels of p53 show different levels of mRNA 30 processing. NEs from RKO, RKO-E6 and H-1299 cells were treated with UV irradiation, allowed to recover for 2 h and then were analyzed for L3 pre-mRNA 30 cleavage. NEs from RKO cells treated with p53/control siRNA and UV irradiation were also analyzed. HeLa cell samples were also included as a control. Positions of pre-mRNA and the 50 cleavage product are indicated. (b) p53 can inhibit 30 cleavage in samples from RKO-E6 and H-1299 cells. NEs from RKO-E6 and H-1299 cells were preincubated with no addition or increasing amounts of recombinant full-length His-p53 derivative (40, 80 and 120 ng). After 15 min, L3 pre-mRNA was added and incubation continued for 90 min. RNAs were purified and analyzed by denaturing PAGE. Positions of pre-mRNA and the 50 cleavage product are indicated. (c) Comparison of p53 expression levels in NE from RKO, RKOE6, H-1299 and HeLa cells. Protein concentrations in NE from UV-treated/untreated and siRNA-treated/untreated cells were normalized by immunostaining with antibodies against Topo II.

and 8, to Figure 2b (lanes 1 and 2), siRNA-treated samples are not shown). Like with MCF7 and HeLa cells, NEs from both RKO and RKO-E6 cells exposed to UV treatment showed the formation of the CstF/ BARD1/p53 complex (not shown). Samples from lung cancer H-1299 cells, which carry an homozygous partial deletion of p53 and lack expression of the protein (Figure 3c, lanes 9 and 10), showed high levels of 30 processing independently of UV treatment (Figure 3a, lanes 10 and 11). As with HeLa cells (Figure 2a), the addition of increasing amounts of full-length His-p53 to limiting amounts of NE from RKO-E6 (Figure 3b, lanes 2–4) and H-1299 cells (lanes 6–8) significantly reduced the 30 cleavage of the radiolabeled adenoviral L3 premRNA. Taken together, these results indicate that this p53 function in RNA 30 processing is not a cell typespecific effect and that p53 can inhibit the 30 cleavage step of the polyadenylation reaction in an expression level-dependent manner. To further determine the role of p53 in mRNA 30 processing, we took advantage of the isogenic colon cancer cell lines, DLD-1 and D-A2. DLD-1 cells carry the tumor-associated S241F mutation in p53 and lack the expression of WT-p53, which results in a defective p53 pathway (Yu et al., 1999). A two-step procedure was used to establish a modified tetracycline-off system for controlled WT-p53 expression in D-A2 cells, leaving the expression of WT-p53 under the regulation of doxycycline (Dox; Yu et al., 1999). To investigate the possible effects of this mutation on 30 processing, first we examined the effect of the S241F mutation on the p53, BARD1 and CstF interaction by analyzing NEs from DLD-1 and D-A2 cells by co-immunoprecipitation. As shown in Figure 4a, UV treatment induced the

accumulation of the mutant p53 in samples of DLD-1 cells (compare lanes 1 and 2). Interestingly, antibodies against p53 did not co-immunoprecipitate either CstF or BARD1 from NEs of DLD-1 cells irrespective of UV treatment (Figure 4a, top panel, lanes 3–6). However, the reciprocal co-immunoprecipitation analysis with CstF2 antibodies showed binding to BARD1, but not to p53 in samples from UV-treated DLD-1 cells (Figure 4a, bottom panel, lanes 3–6), indicating that BARD1 can form a complex with CstF independently of p53. Similar results were observed using NEs from D-A2 cells with no induction of WT-p53 expression (not shown) and with samples from H-1299 cells (not shown). Importantly, the co-immunoprecipitation of p53, CstF and BARD1 with both p53 and CstF2 antibodies was rescued in NEs from D-A2 cells following the induction of WT-p53 (Figure 4b, lanes 3–6). The association of the three proteins in extracts from non-treated and UV-treated cells likely reflects the high levels of p53 expression in D-A2 cells growing in induction medium without Dox (Figure 4b, compare lanes 1 and 2). Altogether, these results indicate that the Ser241 and/or some structural feature around this residue are important for the interaction of p53 with BARD1 and CstF. Consistent with this, the S241F mutation in p53 prevented the formation of the p53/ BARD1/CstF complex in extracts from DLD-1 cells. However, the S241F mutation in p53 did not affect the CstF/BARD1 interaction, indicating that BARD1 can bind to CstF independently of p53 and that p53 might have a role stabilizing the complex. Next, we determined whether the S241F mutation influences the inhibitory effect of p53 on 30 end processing using NEs of DLD-1 and D-A2 cells upon Oncogene


3078 input +

UV:

α-p53 CoIP IP SN -

IP +

SN +

UV:

input + -

α-p53 CoIP IP SN IP +

SN +

UV:

- Topo II

- Topo II

- BARD1

- BARD1

0.5

0

2

hr UV:

+Dox +

-

-Dox +

1

3

4

-AdL3-5’-

- CstF2

- CstF2

- p53

- p53

input UV: +

α-CstF2 CoIP IP SN IP +

SN +

UV:

input +

α-CstF2 CoIP IP SN IP +

SN +

- Topo II

- Topo II

- BARD1

- BARD1

- CstF2

- CstF2

- p53

- p53

1

2

3

2

3

4

5

6

1

2

3

4

5

His-p53 S241F

His-p53

NE

1

2

- AdL3 - 5’

6

1

2

3

4

5

6

7

Figure 4 The S241F mutation in p53 disrupts the BARD1, CstF and p53 interaction and decreases the inhibitory effect of p53 on 30 cleavage, and all this is restored by WT-p53 expression. (a) CstF, p53 and BARD1 do not co-immunoprecipitate from NE of DLD-1 cells that express S241F mutant p53. NEs were immunoprecipitated with either anti-CstF2 or anti-p53 antibodies. Protein samples were treated with RNase A. Equivalent amounts of the pellets (IP) and the supernatants (SN) were resolved by sodium dodecyl sulfate– polyacrylamide gel electrophoresis and proteins were detected by immunoblotting with antibodies against BARD1, CstF2 and p53. Antibodies against Topo II were used as a control of specificity. Positions of Topo II, BARD1, CstF2 and p53 are indicated. Twenty percent of the NE used in the immunoprecipitation reaction is shown as input. (b) CstF, BARD1 and p53 co-immunoprecipitate from NEs of D-A2 cells following induction of WT-p53 expression. D-A2 cells were grown in the absence of Dox to induce the expression of WT-p53. Samples were analyzed as in (a). (c) The induced expression of WT-p53 in D-A2 cells inhibits pre-mRNA 30 cleavage. NEs from DLD-1 (left panel) and D-A2 cells (right panel) non-treated or treated with Dox and/or UV irradiation, and allowed to recover for the times indicated in the figure, were analyzed for L3 pre-mRNA 30 cleavage. Positions of pre-mRNA and the 50 cleavage product are indicated. (d) The S241F mutation in p53 abolishes the inhibition of 30 cleavage. NEs from HeLa cells were preincubated with no addition or increasing amounts of recombinant His-p53 or His-p53 (S241F) derivative (40, 80 and 120 ng). After 15 min, L3 pre-mRNA was added and incubation continued for 90 min. RNAs were purified and analyzed by denaturing PAGE. Positions of pre-mRNA and the 50 cleavage product are indicated.

DNA-damaging conditions. Samples from DLD-1 cells showed high levels of 30 processing that are similar to those observed in HeLa, RKO-E6 and H-1299 cells, and those levels are slightly decreased by UV treatment (Figure 4c, left panel), suggesting that the S241F mutation reduces p53 inhibitory functions on 30 cleavage. As the UV-induced inhibition of mRNA 30 processing is not completely abolished by the S241F mutation in p53, it is possible that other mechanism(s) might be involved in regulating 30 processing in this response. Supporting this idea, the formation of the CstF1/BARD1 complex after UV treatment is not abolished in samples from DLD-1 cells (Figure 4a), indicating that the mRNA 30 processing inhibitor BARD1 (Kleiman and Manley, 1999) can bind CstF independently of p53 and that p53 might have a role stabilizing the complex. These results indicate that the decrease in the CstF/BARD1/p53 complex formation in DLD-1 cells (Figure 4a) is associated with a decrease in p53-induced inhibition of 30 processing (Figure 4c, left panel), suggesting that serine 241 and/or some structural feature around this residue is important for the CstF/ BARD1/p53 complex formation and, consequently, Oncogene

for the function of p53 as an inhibitor of mRNA 30 cleavage. To further confirm the involvement of p53 in this response, we used extracts from D-A2 cells either grown with or without Dox in mRNA 30 cleavage assays. Like the samples from DLD-1 cells, D-A2 cells depleted of WT-p53 (plus Dox) showed high levels of 30 processing independently of UV treatment (Figure 4c, right panel, lanes 1 and 2). Strikingly, the induced expression of WTp53 by the removal of Dox was sufficient to inhibit 30 cleavage in NEs from D-A2 cells (lanes 3 and 4). As p53 expression in D-A2 is induced by the removal of Dox and not by UV treatment, the 30 cleavage inhibition observed in those cells was irrespective of UV treatment. These results again show a reverse correlation between the levels of pre-mRNA 30 processing and the levels of p53 expression. To characterize the effect of the S241F mutation of p53 on 30 cleavage, we constructed a His-p53 recombinant protein containing the mutation (His-p53 S241F). The S241F mutation significantly reduced binding to GST-BARD1 and GST-CstF1 (not shown), indicating that this p53 residue is important for optimal interaction


3079

-

+

-

+

4 2

+ D-A2

-

-

+

-

+

control

6

-

0

p53

8

0

2

+ DLD-1

4 2 + D-A2

2 0

-

-

+

-

control

6

-

4

p53

8

0

6

+ DLD-1

+

4 2 + D-A2

6 4 2 0

-

-

+

-

control

6

-

8

p53

8

0

c-fos poly(A)/total RNA ratio

0

4

8

+ DLD-1

c-fos poly(A)/total RNA ratio

2

p21 poly(A)/total RNA ratio

4

6

p21 poly(A)/total RNA ratio

GAPDH poly(A)/total RNA ratio

6

treated MCF7 cells (Figure 5a) and D-A2 cells minus Dox (Figure 5b). Samples from those cells showed high levels of WT-p53 expression (Figures 2d and 4b) and very low levels of mRNA 30 cleavage (Figures 2c and 4c, right panel). Interestingly, our results showed that GAPDH and b-actin mRNAs were enriched B5-fold in the poly(A) þ preparation over the total RNA fraction in non-UV-treated samples from p53-depleted MCF7 cells and DLD-1 cells. Importantly, samples from these cells not treated with UV showed detectable levels of mRNA 30 cleavage (Figures 2c and 4c, left panel). After UV treatment, samples from p53-depleted MCF7 cells showed not only a strong decrease in the poly(A) þ /total RNA ratio to the levels observed for control siRNA (Figure 5a), but also very low levels of 30 cleavage (Figure 2c, lane 6). Samples from DLD-1 cells treated with UV showed a slight decrease not only in the poly(A) þ /total RNA ratio (Figure 5b), but also in the levels of 30 cleavage (Figure 4c, left panel). As the DLD1 cell line does not express WT-p53, our results suggest that other mechanism(s) might also be involved in regulating the poly(A) þ /total RNA ratio of the housekeeping genes and the levels of 30 processing in this response. Taken together, these results are consistent with our earlier studies (Mirkin et al., 2008) and others previous observations (Akeo et al., 2007; Maccoux et al., 2007) that showed that GAPDH RNA expression can change significantly in different biological systems and under different conditions. Consistent with previous observations (Kleiman et al., 2005; Mirkin et al., 2008; Fox et al., 2009), our results did not show a change in the protein expression of those genes (not shown). Given the complexity of the protein expression pathway and the limitations of the techniques used in the

8

GAPDH poly(A)/total RNA ratio

UV: cell line:

8

control

actin poly(A)/total RNA ratio

UV: siRNA:

actin poly(A)/total RNA ratio

with BARD1 and CstF1. Then, we monitored 30 cleavage of the adenoviral L3 pre-mRNA in NEs from untreated HeLa cells either with no addition or with increasing amounts of His-p53 S241F. As observed in Figure 2a, addition of increasing amounts of purified WT-His-p53 to reaction mixtures effectively inhibited 30 cleavage, almost completely at the highest concentrations (Figure 4d, lanes 2–4). Importantly, the His-p53 S241F derivative was without detectable effect on 30 processing at all concentrations tested (lanes 5–7). Thus, our results indicate that p53 serine 241 is important for the binding of BARD1 and CstF and, consequently, for the function of p53 as an inhibitor of mRNA 30 cleavage. The data presented above provide evidence that p53 inhibits mRNA 30 processing through its interaction with the CstF/BARD1 complex. To get an insight into the molecular mechanisms of this p53 transactivationindependent function, we tested for enrichment of different mRNAs in the poly(A) þ RNA population before and after UV treatment in MCF7 cells treated with p53 siRNA (Figure 5a) and in the isogenic cell lines DLD-1 and D-A2 (Figure 5b). This approach allowed us to analyze the effect of p53 on mRNA 30 processing (poly(A) þ RNA population) over the effect of p53 on transcription and stability (total RNA fraction). We analyzed housekeeping genes, such as b-actin and GAPDH, as well as genes involved in the DDR, such as the p53-regulated gene p21 and the oncogene c-fos. Quantification of quantitative reverse transcription (qRT) –PCR results shows that the studied mRNAs did not show a significant change in the enrichment in the poly(A) þ preparation over the total RNA fraction after UV treatment of samples from control siRNA-

+ p53

8 6 4 2 0

D-A2

+

-

+ DLD-1

Figure 5 p53 expression induces the enrichment in the poly(A) þ preparation over the total RNA fraction for b-actin and GAPDH mRNAs, but not for p21 and c-fos mRNAs. Real-time PCR analysis of b-actin, GAPDH, p21 and c-fos mRNAs polyadenylation. Total and poly(A) þ RNA were prepared after UV treatment from (a) MCF7 cells treated with control/p53 siRNA or (b) DLD-1 (p53 S241F/)/D-A2 (p53 S241F/ þ ) cells following the removal of Dox. RNA purification and RT–PCR reactions were performed as described in Materials and methods. Equal volumes of both total and poly(A) þ RNA samples were used as a template in the RT reactions. Equal amounts of cDNAs were used in qRT–PCR reactions with primers specific for GAPDH, b-actin, c-fos and p21 mRNAs. Relative quantification was achieved using standard curves of known amounts of total cDNA. The results shown are the average of four PCRs from two different RNA extractions. Oncogene


3080

measurements of mRNA and protein levels (Fu et al., 2007; Gry et al., 2009), the observed lack of correlation between mRNA and corresponding protein expression levels is not unusual. These results show a correlation between the expression of WT-p53 and a decrease in the levels of not only mRNA 30 cleavage (Figures 2 and 4), but also of the poly(A) þ /total RNA ratio of the housekeeping genes analyzed in this study. A change in the poly(A) þ /total RNA ratio could reflect a change either in the levels of polyadenylated mRNAs or in the levels of deadenylated mRNAs, whose poly(A) tails have been removed during mRNA turnover, in the total mRNA fraction. Supporting this idea, we have recently shown that the DDR involves not only the inhibition of 30 cleavage, but also the activation of deadenylation, which results in a decrease of mRNA stability (Cevher et al., 2010). Although further studies are necessary to determine the functional relevance of p53 in the regulation of deadenylation, our data indicate that p53 is an inhibitor of the 30 cleavage reaction of polyadenylation. Interestingly, neither the depletion of p53 expression nor the expression of mutant p53 had an effect on either p21 or c-fos poly(A) þ mRNA enrichment independently of the UV treatment (Figures 5a and b), suggesting that the steady-state levels of those mRNAs are not regulated by 30 processing in a p53-dependent manner. Consistent with previous observations (Fong et al., 2010), our results showed the UV-induced increase in the protein expression of c-fos and p21 (not shown). These results indicate that the expression levels of WTp53 has important roles in the regulation of the levels of different endogenous mRNAs not only by its transactivation functions, but also by transactivation-independent functions decreasing the levels of polyadenylated mRNAs by inhibition of mRNA 30 cleavage. As we proposed before, it is possible that the association of p53, BARD1 and CstF in different cellular conditions could play a role in regulating mRNA 30 processing and, therefore, mRNA levels of different genes. Discussion To our knowledge, this is the first study showing a 30 RNA processing inhibitory function of p53, adding a new level of complexity to the DDR by linking RNA processing to the p53 network. Several lines of evidence support the hypothesis that p53 is an inhibitor of mRNA 30 processing. First, the direct interaction of the C-terminal domain of p53 with CstF1 and BARD1, both of which are involved in the UV-induced inhibition of mRNA 30 processing, and the existence of protein complexes of these factors in extracts of different cell lines (Figure 1). Second, p53 can inhibit the 30 cleavage reaction in vitro and p53 expression levels inversely correlate with levels of mRNA 30 cleavage (Figures 2 and 3). Third, the tumor-associated S241F mutation in p53 reduces binding to BARD1 and CstF and disrupts the CstF/BARD1/p53 complex association, and this complex formation is restored by the induced expression Oncogene

of WT-p53 (Figure 4). Furthermore, cells expressing this mutant p53 show reduced inhibition of mRNA 30 cleavage, which is restored by the induced expression of WT-p53. Fourth, we determined that the expression of WT-p53 has different effects on the polyadenylation state of the housekeeping genes analyzed in this study (Figure 5). Taken together, our results suggest a novel function of p53 as an inhibitor of the mRNA 30 processing machinery. We have proposed in previous studies that there is a general effect of DNA-damaging conditions on mRNA levels, and that the UV-induced inhibition of 30 end processing plays an important role in decreasing the total cellular mRNA levels as part of this response (Cevher and Kleiman, 2010). Consistent with this, the levels of poly(A) þ mRNA of genes not involved in DDR decrease after DNA damage (Ljungman et al., 1999; Dheda et al., 2004; Akeo et al., 2007; Maccoux et al., 2007; Mirkin et al., 2008). It has been shown that full recovery of total mRNA levels within 6 h after the DNA-damaging exposure correlates with cellular protection against apoptosis in a p53-dependent manner (McKay et al., 2001). On the other hand, there is also a gene-specific effect of DNA-damaging conditions on the levels of poly(A) þ mRNAs of genes involved in the DDR, those genes are either down- or upregulated at different time points after DNA damage (reviewed by Cevher and Kleiman, 2010). In this work, we have discovered that under DNA-damaging conditions, p53 in association with the 30 processing factor CstF and the tumor suppressor BARD1 can control mRNA 30 processing of housekeeping genes, but not of p21 and cfos genes, which are involved in DDR. On the basis of our studies, we propose that p53, a protein with compromised expression in most cancers, plays a role in the general response to DNA damage by stabilizing the CstF/BARD1 complex and inhibiting 30 processing of aborted nascent RNA products, allowing the elimination of prematurely terminated transcripts to avoid the expression of deleterious proteins and facilitating DNA repair by clearing the area around the lesion. Defective polyadenylation of prematurely terminated transcripts is known to activate a nuclear surveillance pathway, allowing the elimination of those mRNAs by exosomemediated degradation (reviewed by Cevher and Kleiman, 2010). As DNA repair proceeds, the levels of p53 expression decrease allowing the recovery of total mRNA levels. Our results indicate that different cell lines exhibit different 30 processing profiles depending on p53 expression levels, consistent with the idea proposed by Singh et al. (2009) that the interaction of the 30 processing machinery and factors involved in the DDR/ tumor suppression might result in cell-specific 30 processing profiles. Supporting the idea that the 30 processing machinery is interconnected with the p53 pathway, it has been shown that Rbbp6, a p53-binding protein, and Ku-70 subunit of DNA-PK, which is involved in the BARD1mediated phosphorylation of p53 Ser15 upon DNA damage (Fabbro et al., 2004), are part of the pre-mRNA 30 processing complex (Shi et al., 2009). Other tumor


3081

suppressors, such as CSR1 and Cdc73, have also been shown to functionally associate with 30 processing factors, such as CPSF3 (Zhu et al., 2009) and CPSF/ CstF (Rozenblatt-Rosen et al., 2009). Recently, it has been shown that the use of alternative mRNA 30 cleavage and polyadenylation sites can control the expression of certain genes by eliminating or including several cis-acting elements, such as microRNA target sites and AU-rich elements, in cancer cells and during development (Sandberg et al., 2008; Zlotorynski and Agami, 2008; Ji et al., 2009; Mayr and Bartel, 2009; Singh et al., 2009). Although more work is necessary to determine the functional relevance of the CstF/BARD1/ p53 interaction in the regulation of expression of specific genes involved in DDR, it is possible that the p53mediated inhibition of mRNA 30 processing might also play a role in the selection of different alternative mRNA 30 cleavage sites and, consequently, in the regulation of the mRNA levels of genes involved in DDR. Considering that the p53 pathway is tightly controlled in cells following DNA damage (reviewed by Vousden, 2006), the p53-associated control of mRNA 30 processing could be an effective mechanism employed to control gene expression in cells upon DNA-damaging conditions. Taken together, our studies identify a novel 30 RNA processing inhibitory function of p53 and suggest that the CstF/BARD1/p53 interaction contributes to UV-induced inhibition of pre-mRNA 30 processing, providing evidence of another link between mRNA 30 processing and tumor suppression.

Materials and methods Tissue culture methods and DNA-damaging agents HeLa, MCF7 RKO and RKO-E6 cells were cultured in Dulbecco’s modified Eagle’s medium–10% fetal bovine serum. NCI-H1299 cells were cultured in RPMI 1640 media supplemented with 10% fetal bovine serum. The isogenic colon cancer cell lines DLD-1 and D-A2 were generously provided by Dr Vogelstein (Johns Hopkins School of Medicine, Baltimore, MD, USA; Yu et al., 1999). DLD-1 cells were grown in McCoy’s 5A media containing 10% fetal bovine serum (v/v) and 2500 U of pen–strep. D-A2 cells required an additional 0.4 mg/ml G418, 20 ng/ml Dox and 0.25 mg/ml hygromycin B. D-A2 cells were grown in the presence or absence of Dox as indicated. Ninety percent confluent cultures were exposed to UV and harvested after 2 h. UV doses (20 J/m2) were delivered in two pulses using a Stratalinker (Stratagene, La Jolla, CA, USA). Before pulsing, medium was removed and replaced immediately after treatment. Nuclear extracts preparation and immunoblotting analysis After UV treatment, NEs were prepared from harvested cells as described (Kleiman and Manley, 2001). Sixty micrograms of each NE was analyzed by immunoblotting with antibodies against BARD1 (H-300, Santa Cruz Biotechnology, Santa Cruz, CA, USA), p53 (SC-126, Santa Cruz), CstF2 (generously provided by Dr Manley, Columbia University, New York, NY, USA), CstF1 (10064-2-AP, Protein Tech Group, Chicago, IL, USA) and Topoisomerase II (Topo II, SC-25330, Santa Cruz).

siRNA knockdown of p53 expression in HeLa and MCF7 cells The siRNA specific for human p53 was synthesized by Qiagen (Valencia, CA, USA) and the control RNA duplex used as non-silencing siRNA was obtained from Dharmacon RNA Technologies (Lafayette, CO, USA). HeLa, RKO and MCF7 cells were grown in a 10-cm plate in complete Dulbecco’s modified Eagle’s medium. At 60–70% confluence, the cells were transfected with p53 siRNA or control siRNA and Lipofectamine 2000 (Invitrogen) according to the manufacturer’s protocol. After culturing the cells for an additional 24 h, the transfection procedure was repeated, and the cells were harvested for analysis 48 h after the initial transfection. A fraction of the cells was exposed to UV and harvested after 2 h. NEs were prepared and analyzed by western blot and used in 30 cleavage reactions. Immunoprecipitation analysis One hundred microgram of total protein from each NE was immunoprecipitated with the anti-CstF2 (generously provided by Dr Manley, Columbia University), CstF1 (10064-2-AP, Protein Tech Group) or p53 (SC-126, Santa Cruz) monoclonal antibodies bound to protein A-Sepharose beads. The beads were recovered by centrifugation and treated at 4 1C with 50 mg of RNase A/ml for 10 min. Immunoprecipitations were carried out as described (Kleiman and Manley, 2001), except that washing was with buffer A (1 phosphate-buffered saline: 137 mM NaCl, 3 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, 0.01% Nonidet P-40, 0.5 mM phenylmethylsulfonyl fluoride and 0.04% bovine serum albumin). Aliquots of pellets and supernatants were analyzed by sodium dodecyl sulfate– polyacrylamide gel electrophoresis and immunoblotting as described above. Recombinant proteins GST-BARD1 and GST-CstF1 plasmids were constructed as described (Kleiman and Manley, 2001). The plasmid encoding His-p53 was generously provided by Dr Prives (Columbia University). His-p53 derivatives were obtained by PCR amplification with primers containing BamHI and NdeI restriction sites followed by inserting the PCR fragment into pET14b vector (Novagen, Gibbstown, NJ, USA). The Ser241Phe mutation on p53 was introduced by site-directed mutagenesis using Quickchange kit (Stratagene) according to the manufacturer’s protocol. Plasmids encoding BARD1, CstF1, p53 and p53 derivatives were expressed in Escherichia coli, and purified by binding to and elution from either glutathione-agarose or Ni-agarose columns as described (Kleiman and Manley, 2001). Protein–protein interaction assays The GST fusion protein interaction assays with His-p53 and His-p53 derivatives were performed as described (Kleiman and Manley, 2001). Alternatively, 30 ml of HeLa, RKO-E6 and NCI-H1299 cell NEs were incubated with 1 mg of the indicated GST fusion proteins and increasing amounts of His-p53; the binding and washing conditions were as before. Protein samples were treated at 4 1C with 50 mg of RNase A/ml for 10 min. Equivalent amounts of pellets and supernatants were analyzed by immunoblotting. 30 Cleavage assays P-labeled L3 pre-mRNA substrates were prepared as described (Kleiman and Manley, 1999). Protein concentrations of the extracts were equalized by Bradford assays (BioRad, Hercules, CA, USA) before use in processing reactions. 32

Oncogene


3082 Cleavage assays with equivalent amounts of total protein were carried out as described (Kleiman and Manley, 2001). Analysis of endogenous mRNAs by qRT–PCR Total nuclear RNA (200 mg) was purified from MCF7 cells treated with control/p53 siRNA, DLD-1 and D-A2 cells using the RNeasy (Qiagen) and the Oligotex resin (Qiagen) following the manufacturer’s instructions, resulting in 6–8 mg of polyA þ RNA. The RNA concentration of the total RNA samples obtained under different conditions was equalized. Poly(A) þ RNA was prepared by treating total nuclear RNA samples with Oligotex resins. Equivalent amounts of purified RNA were used as a template to synthesize cDNA using random hexamer primers and Moloney murine leukemia virus reverse transcriptase (Promega, Madison, WI, USA) according to the manufacturer’s protocol. Commercially available primers were used in the qRT–PCR reactions (GAPDH: 4333764F; ACTB: 4333762F; c-Fos: Hs00170630-m1; p21: Hs00355782-m1; Applied Biosystems, Foster City, CA, USA). RT reactions were performed using random primers and equal volume of both total RNA and polyA þ RNA samples. Equal amounts of total or poly(A) þ cDNAs were used in the qRT–PCR

reactions to observe the poly(A) þ enrichment as described (Gomes et al., 2006; Cevher et al., 2010). Relative quantification of both cDNA samples was carried out by using standard curves of known amounts of total cDNA.

Conflict of interest The authors declare no conflict of interest.

Acknowledgements We thank Dr JL Manley for anti-CstF2 antibodies, Dr B Vogelstein and Dr J Bargonetti for cell lines DLD-1 and D-A2, Dr EK Boamah for technical advice, Dr C Prives for p53 encoding plasmids, Dr R Baer for BARD1 encoding plasmids and Dr S Pinõl-Roma and Dr MA Cevher for advice and discussion. This work is supported by National Institute of General Medical Sciences Grant SC1GM083806 to FEK and by Minority Access to Research Careers Program (MARC) to BA and CT.

References Akeo K, Funayama T, Hamada N, Akeo Y, Hiramitsu T, Kobayashi Y. (2007). The effects of ultraviolet irradiation and hypoxia on expression of glutathione peroxidase and glyceraldehyde 3-phosphate dehydrogenase in the cultured retinal pigment epithelium. Tissue Culture Res Commun 26: 149–157. Cevher MA, Kleiman FE. (2010). Connections between 30 end processing and DNA damage response focus article in Wiley interdisciplinary reviews WIREs RNA. http://wires.wiley.com/ WileyCDA/WiresArticle/wisId-WRNA20.html. Cevher MA, Zhang X, Fernandez S, Kim S, Baquero J, Nilsson P; et al. (2010). Nuclear deadenylation/polyadenylation factors regulate 30 processing in response to DNA damage. EMBO J 29: 1674–1687. Chang TC, Wentzel EA, Kent OA, Ramachandran K, Mullendore M, Lee KH et al. (2007). Transactivation of miR-34a by p53 broadly influences gene expression and promotes apoptosis. Mol Cell 26: 745–752. Dheda K, Huggett JF, Bustin SA, Johnson MA, Rook G, Zumla A. (2004). Validation of housekeeping genes for normalizing RNA expression in real-time PCR. Biotechniques 37: 112–119. Fabbro M, Savage K, Hobson K, Deans AJ, Powell SN, McArthur GA et al. (2004). BRCA1-BARD1 complexes are required for p53Ser-15 phosphorylation and a G1/S arrest following ionizing radiation-induced DNA damage. J Biol Chem 279: 31251–31258. Feki A, Jefford CE, Berardi P, Wu JY, Cartier L, Krause KH et al. (2005). BARD1 induces apoptosis by catalysing phosphorylation of p53 by DNA-damage response kinase. Oncogene 24: 3726–3736. Fong S, King F, Shtivelman E. (2010). CC3/TIP30 affects DNA damage repair. BMC Cell Biol 11: 23. Fox JT, Shin WK, Caudill MA, Stover PJ. (2009). A UV-responsive internal ribosome entry site enhances serine hydroxymethyltransferase 1 expression for DNA damage repair. J Biol Chem 284: 31097–31108. Fu N, Drinnenberg I, Kelso J, Wu JR, Paä¨bo S, Zeng R et al. (2007). Comparison of protein and mRNA expression evolution in humans and chimpanzees. PLoS One 2: e216. Gomes NP, Bjerke G, Llorente B, Szostek SA, Emerson BM, Espinosa JM. (2006). Gene-specific requirement for P-TEFb activity and RNA polymerase II phosphorylation within the p53 transcriptional program. Genes Dev 20: 601–612. Oncogene

Gry M, Rimini R, Stro¨mberg S, Asplund A, Ponteń F, Uhleń M et al. (2009). Correlations between RNA and protein expression profiles in 23 human cell lines. BMC Genom 10: 365. He X, He L, Hannon GJ. (2007). The guardian’s little helper: microRNAs in the p53 tumor suppressor network. Cancer Res 67: 11099–11101. Hirose Y, Manley JL. (1998). RNA polymerase II is an essential mRNA polyadenylation factor. Nature 395: 93–96. Irminger-Finger I, Leung WC, Li J, Dubois-Dauphin M, Harb J, Feki A et al. (2001). Identification of BARD1 as mediator between proapoptotic stress and p53-dependent apoptosis. Mol Cell 8: 1255–1266. Ji Z, Lee JY, Pan Z, Jiang B, Tian B. (2009). Progressive lengthening of 30 untranslated regions of mRNAs by alternative polyadenylation during mouse embryonic development. Proc Natl Acad Sci USA 106: 7028–7033. Kessis TD, Slebos RJ, Nelson WG, Kastan MB, Plunkett BS, Han SM et al. (1993). Human papillomavirus 16 E6 expression disrupts the p53-mediated cellular response to DNA damage. Proc Natl Acad Sci USA 90: 3988–3992. Kleiman FE, Manley JL. (1999). Functional interaction of BRCA1associated BARD1 with polyadenylation factor CstF-50. Science 285: 1576–1579. Kleiman FE, Manley JL. (2001). The BARD1-CstF-50 interaction links mRNA 30 end formation to DNA damage and tumor suppression. Cell 104: 743–753. Kleiman FE, Wu-Baer F, Fonseca D, Kaneko S, Baer R, Manley JL. (2005). BRCA1/BARD1 inhibition of mRNA 30 processing involves targeted degradation of RNA polymerase II. Genes Dev 19: 1227–1237. Levine AJ, Hu W, Feng Z. (2006). The P53 pathway: what questions remain to be explored? Cell Death Differ 13: 1027–1036. Ljungman M, Zhang F, Chen F, Rainbow AJ, McKay BC. (1999). Inhibition of RNA polymerase II as a trigger for the p53 response. Oncogene 18: 583–592. Maccoux LJ, Clements DN, Salway F, Day PJ. (2007). Identification of new reference genes for the normalisation of canine osteoarthritic joint tissue transcripts from microarray data. BMC Mol Biol 8: 62–72. McCracken S, Fong N, Yankulov K et al. (1997). The C-terminal domain of RNA polymerase II couples mRNA processing to transcription. Nature 385: 357–361.


3083 McKay BC, Becerril C, Ljungman M. (2001). P53 plays a protective role against UV- and cisplatin-induced apoptosis in transcriptioncoupled repair proficient fibroblasts. Oncogene 20: 6805–6808. McKay BC, Ljungman M. (1999). Role for p53 in the recovery of transcription and protection against apoptosis induced by ultraviolet light. Neoplasia 1: 276–284. Mandel CR, Bai Y, Tong L. (2008). Protein factors in pre-mRNA 30 end processing. Cell Mol Life Sci 65: 1099–1122. Mayr C, Bartel DP. (2009). Widespread shortening of 30 UTRs by alternative cleavage and polyadenylation activates oncogenes in cancer cells. Cell 138: 673–684. Mirkin N, Fonseca D, Mohammed S, Cevher MA, Manley JL, Kleiman FE. 2008. The 30 processing factor CstF functions in the DNA repair response. Nucleic Acids Res 36: 1792–1804. Noda A, Toma-Aiba Y, Fujiwara Y. (2000). A unique, short sequence determines p53 gene basal and UV-inducible expression in normal human cells. Oncogene 19: 21–31. Rozenblatt-Rosen O, Nagaike T, Francis JM, Kaneko S, Glatt KA, Hughes CM et al. (2009). The tumor suppressor Cdc73 functionally associates with CPSF and CstF 30 mRNA processing factors. Proc Natl Acad Sci USA 106: 755–760. Sandberg R, Neilson JR, Sarma A, Sharp PA, Burge CB. (2008). Proliferating cells express mRNAs with shortened 30 untranslated regions and fewer microRNA target sites. Science 320: 1643–1647. Sauer M, Bretz AC, Beinoraviciute-Kellner R, Beitzinger M, Burek C, Rosenwald A et al. (2008). C-terminal diversity within the p53 family accounts for differences in DNA binding and transcriptional activity. Nucleic Acids Res 36: 1900–1912. Scorilas A. (2002). Polyadenylate polymerase (PAP) and 30 end premRNA processing: function, assays, and association with disease. Crit Rev Clin Lab Sci 39: 193–224.

Shi Y, Di Giammartino DC, Taylor D, Sarkeshik A, Rice III WJ, Yates JR et al. (2009). Molecular architecture of the human premRNA 30 processing complex. Mol Cell 33: 365–376. Singh P, Alley TL, Wright SM, Kamdar S, Schott W, Wilpan RY et al. (2009). Global changes in processing of mRNA 30 untranslated regions characterize clinically distinct cancer subtypes. Cancer Res 69: 9422–9430. Takagaki Y, Manley JL. (1998). Levels of polyadenylation factor CstF-64 control IgM heavy chain mRNA accumulation and other events associated with B cell differentiation. Mol Cell 2: 761–771. Takagaki Y, Ryner LC, Manley JL. (1989). Four factors are required for 30 -end cleavage of pre-mRNAs. Genes Dev 3: 1711–1724. Takwi A, Li Y. (2009). The p53 pathway encounters the microRNA world. Curr Genom 10: 194–197. Topalian SL, Kaneko S, Gonzales MI, Bond GL, Ward Y, Manley JL. (2001). Identification and functional characterization of neopoly(A) polymerase, an RNA processing enzyme overexpressed in human tumors. Mol Cell Biol 21: 5614–5623. Vogelstein B, Lane D, Levine AJ. (2000). Surfing the p53 network. Nature 408: 307–310. Vousden KH. (2006). Outcomes of p53 activation—spoilt for choice. J Cell Sci 119: 5015–5020. Wei Q. (2005). Pitx2a binds to human papillomavirus type 18 E6 protein and inhibits E6-mediated P53 degradation in HeLa cells. J Biol Chem 280: 37790–37797. Yu J, Zhang L, Hwang PM, Rago C, Kinzler KW, Vogelstein B. (1999). Identification and classification of p53-regulated genes. Proc Natl Acad Sci USA 96: 14517–14522. Zhu ZH, Yu YP, Shi YK, Nelson JB, Luo JH. (2009). CSR1 induces cell death through inactivation of CPSF3. Oncogene 28: 41–51. Zlotorynski E, Agami R. (2008). A PASport to cellular proliferation. Cell 134: 208–210.

Oncogene