Skip Navigation

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (14)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Göpfert, U.
Right arrow Articles by Hengst, L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Göpfert, U.
Right arrow Articles by Hengst, L.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Human Molecular Genetics, 2003, Vol. 12, No. 14 1767-1779
DOI: 10.1093/hmg/ddg177
© 2003 Oxford University Press

Cell cycle-dependent translation of p27 involves a responsive element in its 5'-UTR that overlaps with a uORF

Ulrich Göpfert, Michael Kullmann and Ludger Hengst*

Max-Planck-Institut für Biochemie, Am Klopferspitz 18a, D-82152 Martinsried, Germany

Received January 21, 2003; Revised April 13, 2003; Accepted April 24, 2003


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
p27Kip1 regulates cell proliferation by binding to and modulating the activity of cyclin-dependent kinases. The CDK inhibitor is haploinsufficient for tumor suppression and reduced p27 activity is fundamental for the development of many human malignancies. Consistently, reduced p27 protein provides independent prognostic information in various tumors including breast, prostate, colon and gastric carcinomas. In normal cells, p27 protein increases in growth arrest but also oscillates during cell cycle progression. Expression of p27 is regulated through mechanisms including transcription, translation and ubiquitin-mediated degradation. Each of these pathways may contribute to deregulation of p27 in hyperproliferative diseases. p27 translation increases in proliferating cells during G1 phase and declines as cells enter S phase. To investigate the mechanisms of p27 translational control, we analyzed fragments of the p27 transcript for their contribution to cell cycle regulated translation. We found that an element in the p27 5'-UTR can render reporter translation cell cycle sensitive with maximal translation in G1-arrested cells. This novel element of 114 nt contains a G/C-rich hairpin domain that is predicted to form multiple stable stemloops and also overlaps with a small upstream ORF (uORF). Both structures contribute to cell cycle-regulated translation. The uORF can be translated in vitro and its sequence and position are highly conserved in mice and chickens. Interestingly, the precise sequence or the length of the uORF-encoded peptide are not important for p27 translation, consistent with the idea that ribosomal recruitment to its initiation codon rather than the translation product itself contributes to the regulation.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Oscillating activity of cyclin-dependent kinases (CDKs) is a prerequisite for all major cell cycle phase transitions (1,2). CDKs are heterodimeric protein kinases that are composed of a catalytic subunit called CDK and a positive regulatory subunit, the cyclin (2). The first CDK complex that becomes activated in G1 is cyclin D associated with CDK4 or CDK6, followed by the activation of E/CDK2 at the G1/S boundary (1). The activation of the kinase complexes is subject to diverse regulatory mechanisms. These can include a cell cycle phase-dependent expression of most cyclins and regulation of their complex formation with the cognate CDK. A second level of control relies on the requirement for activating phosphorylation of the CDK subunit at its T-loop (2,3). The third level involves inhibitory phosphorylation at the N-terminus of the CDK subunit by tyrosine or dual specific kinases (4). The inhibitory phosphate can be removed by members of the Cdc25 family of phosphatases (5).

This regulatory network is superimposed by binding of negative regulatory subunits to monomeric CDKs or to the CDK complex. In mammalian cells, two families of CDK inhibitory proteins (CKIs) bind to and regulate CDK/cyclin activities. The Ink4 family has four members, p15INK4b, p16INK4a, p18INK4c and p19INK4d (6). These proteins bind CDK4 or CDK6 subunits opposite to the cyclin binding site. Binding of Ink4 proteins distorts the structure of the catalytic cleft and the cyclin binding site. This results in impaired ATP binding and a reduced CDK/cyclin interface (7). The second family of CDK inhibitors, the Cip/Kip family, has a wider spectrum of specificity and binds diverse cyclin/CDK complexes. The three members of this family, p21Cip1/Waf1/Sdi1, p27Kip1 and p57Kip2 share an N-terminal CDK-inhibitory domain (8–10). The structure of the inhibitory domain of p27 bound to cyclin A/CDK2 confirmed that the inhibitor binds to both subunits of the kinase complex. It occupies the ATP binding pocket of the kinase (11). Binding of a single molecule of p21 is therefore sufficient for inactivation of a cyclin A/CDK2 complex (12). Paradoxically p21 and p27 do not only inhibit CDK kinase activity but are also involved in the activation of cyclin D/CDK4 complexes (13–16).

The p27Kip1 inhibitor associates with and regulates CDK activity during G1 progression and in quiescence (17–21). p27 was first identified in TGF-ß arrested mink lung epithelial cells (17,18,22) and in lovastatin arrested HeLa cells (19); a p27 cDNA was also isolated in an interaction screen using cyclin D1 as a bait (20). Levels of the inhibitor oscillate during cell cycle progression with a peak in G1 phase, while p27 is barely detectable during the remaining cell cycle phases. Ectopic overexpression blocks the cell cycle in G1 (17,20). Accumulation of the protein appears to be essential for proper restriction point control, as reduction of p27 levels in fibroblasts using antisense approaches suppresses quiescence by supporting serum independent proliferation (23,24). An important function of p27 in growth control was demonstrated after deletion of the p27 gene in mice (25–27). p27-/- animals are characterized by hyperproliferation of multiple tissues, leading to multiorgan hyperplasia. Both p27 heterozygous and p27 nullizygous animals are predisposed to tumorigenesis in multiple tissues after challenge with carcinogens or {gamma}-irradiation (28). However, primary fibroblasts from p27-/- animals grow strictly serum-dependent, suggesting possible redundancy in this pathway (25–27).

p27 expression is regulated at different levels including transcriptional (29–32), translational (33–35) and post-translational mechanisms, including protein stability (36–38), complex association (13) and its localization (39–43). The complex control indicates that the inhibitor may serve as a platform to integrate diverse mitogenic and anti-proliferative signals.

In many scenarios, including cell cycle progression of HeLa cells, p27 mRNA level do not change, even when an induction or reduction of the protein is observed (8). Two post-transcriptional mechanisms, translational control and regulation of protein stability, are mainly responsible for the oscillation of p27 observed during the cell cycle or the induction of p27 in growth-arrested cells (35,36). The stability of p27 is regulated by proteasome-dependent degradation (36,44), which requires a CDK2-mediated phosphorylation of threonine 187 at the G1/S transition (40,45–48), and is controlled by alternative mitogen-dependent pathways during G1 and the G0/G1 transition (37,38).

Growth factor-mediated reduction of p27 level during the G0/G1 transition is initially a result of translational mechanisms (33–35). Similarly, the decrease of p27 in the S-phase involves a reduction in its rate of translation (34,35). The alterations in p27 translation should contribute to precisely adjusting inhibitor level to allow or prevent cyclin E/CDK2 activation. Since activated CDK2 triggers a feedback loop by phosphorylating p27 on T187 and initiating its ubiquitin-dependent degradation, activation of CDK2 kinase should prevent reaccumulation of the inhibitor, even if translation rates are increased. Therefore precise tuning of p27 translation during the G1/S transition may be important for proper cell cycle control.

Several independent mechanisms were recently uncovered that regulate p27 translation. Increased translation of the p27 mRNA correlates with its enhanced polyribosomal association (34), suggesting that the initiation step may be rate limiting. Sustained translation of p27 under poor growth conditions also relies on internal initiation of translation that depends on an IRES element in the 5'-UTR (49,50). A Rho-dependent regulation of p27 synthesis at the G0/G1 transition involves an element of 300 nucleotides in the 3'-UTR of the p27 mRNA that represses translation (51).

To investigate mechanisms that regulate p27 translation during the cell cycle, we studied the contribution of different elements of the p27 mRNA. In this study we describe the identification of a cell cycle regulatory element (CCRE) in the non-translated leader of the p27 mRNA. This element overlaps with a small upstream ORF (uORF). The uORF can be translated in vitro and its initiation codon is required for maximal cell cycle regulation of p27 translation.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
p27 translation is regulated during the cell cycle and peaks in G1 phase (34,35). Since p27 transcripts of various length have been described that differ in length of their 5'-UTR between 152 and 575 nucleotides (50,52–55), we wished to determine if alternative transcription start sites might contribute to the regulated translation of p27. First, we compared RNA isolated from a transformed human cell line (HeLa) and human normal diploid fibroblasts (HS68) that were either growing asynchronously or contact inhibited. Northern blots were hybridized with a probe containing either the very 5' end of the p27 mRNA or a fragment including the coding region (Fig. 1). Identical hybridization patterns were obtained for both fragments, suggesting no major change in the transcriptional start site under these conditions. To investigate a potential tissue specific expression of truncated forms of the p27 mRNA, the same fragments were used to hybridize polyA+ RNA isolated from various human tissues. As for the cell lines, both probes recognized one single RNA species of identical mobility in most tissues (Fig. 1B). Therefore we conclude that the 5' end of the p27 mRNA is broadly expressed in diverse human cells and tissues and that variations in growth conditions do not lead to major alterations in the p27 transcriptional start site.



View larger version (51K):
[in this window]
[in a new window]
  		Figure 1. Analysis of p27 mRNA expression. (A) Schematic representation of the p27 transcript. The radioactive probes used in (B) are indicated. (B) The longest form of the p27 5'-UTR is expressed in a broad range of human cell lines and tissues. Northern blot analysis of total RNA from tissue culture cells (left) and polyA+ RNA from various human tissues (right). Blots were sequentially hybridized with a probe corresponding to the very 5' end of the p27 cDNA (nucleotides -575 to -462), a larger fragment including parts of the p27 ORF and 3'-UTR (nucleotides 120–1511) and a probe of GAPDH cDNA (loading control). The size of standard RNAs is indicated in kb.

 
The 5'-UTR of the p27 mRNA can mediate cell cycle-regulated translation
In order to investigate the mechanism of cell cycle dependent translation of p27, we next analyzed the contribution of the 5' and 3' non-translated regions. The abilities of the longest 5' and 3' non-coding sequences to confer cell cycle regulated translation to a transcript were investigated using luciferase reporter constructs. To measure cell cycle phase-dependent regulation, reporter activities in lovastatin-arrested cells were compared with activities in thymidine-arrested or asynchronous cells. The p27 5'-UTR was inserted upstream and the 3'-UTR downstream of the firefly luciferase reporter gene (Fig. 2A). After transfection, HeLa cells were either grown asynchronously or synchronized in G1 by lovastatin or in S phase by thymidine treatment. The 5'-UTR caused a 2.4-fold increase in relative luciferase activity of lovastatin-treated cells compared with thymidine arrested cells (Fig. 2B). No cell cycle-dependent regulation was observed for the 3'-UTR, however, general translation efficiency was reduced. Combination of 5'- and 3'-flanking regions did not enhance the regulation conferred by the 5'-UTR (Fig. 2B). In these experiments, transfection efficiencies were normalized by cotransfection with a ß-galactosidase reporter. The same degree of cell cycle regulation was obtained in experiments where luciferase activity was normalized for luciferase mRNA level (not shown), excluding that changes in transcription rates or mRNA stabilities contribute to the cell cycle phase-dependent regulation of luciferase activity.



View larger version (18K):
[in this window]
[in a new window]
  		Figure 2. The p27 5'-UTR mediates cell cycle regulated translation. (A) Schematic drawing of the reporter plasmids pGL3 Control, pGL5'-UTR, pGL3'-UTR and pGL5'+3'-UTR. The p27 uORF is represented as gray box. (B) Plasmids were co-transfected into HeLa cells with pSV-ß-Gal as an efficiency control. After transfection, cells were either grown asynchronously or arrested in G1 phase by lovastatin or in S phase by thymidine. The luciferase activities were determined and normalized to ß-galactosidase activities and expressed relative to the values obtained from pGL3 control. The ratio between luciferase activities in lovastatin arrested cells and thymidine-arrested cells is indicated below the diagram. The figure represents the results of four independent experiments.

 
To confirm that the p27 5'-UTR leads to cell cycle phase-dependent regulation of p27 translation and to exclude potential drug-specific effects, we followed reporter translation during the cell cycle. HeLa cells were synchronized by a thymidine/nocodazole block release protocol and luciferase translation was determined as cells progressed through the cell cycle. Since luciferase protein is relatively stable, we determined reporter synthesis rates rather than activities by measuring 35S incorporation in pulse experiments. Overall protein synthesis rates were low in G2/M arrested cells and increased after release from the mitotic block. In comparison to the vector control, addition of the complete p27 5'-UTR led to a cell cycle-regulated translation of the reporter (Fig. 3A). A truncated version of the 5'-UTR containing a recently identified IRES element failed to exhibit this regulation (Fig. 3A). This indicated that elements required for cell cycle-dependent regulation may reside at the very 5' end of the p27 5'-UTR. As an efficiency control for the release of the mitotic block, we determined the cell cycle phase distribution at each time point by flow cytometry (Fig. 3B). This analysis includes the majority of non-transfected cells. Since we noticed a clear delay in cell cycle progression of the transfected cell population, the flow cytometry data represent the cell cycle phase distributions of the majority of cells, while the transfected cells are predicted to delay progress through the cell cycle.



View larger version (19K):
[in this window]
[in a new window]
  		Figure 3. Cell cycle regulated translation in synchronized HeLa cells depends on the 5'-UTR. (A) HeLa cells were transfected with luciferase expression plasmids with or without the indicated fragments of the p27 5'-UTR as leaders. Twenty-four hours after transfection, cells were synchronized by a thymidine/nocodazole block/release protocol. At times indicated after release from the nocodazole block, cells were labeled with 35S methionine for 1 h. Subsequently, luciferase was immunoprecipitated and incorporated radioactivity was quantified using a phosphorImager. RNA integrity and levels were determined in northern blots and the synthesis rate is shown after normalization for mRNA levels. (B) The cell cycle phase distribution was determined by flow cytometry analysis of the DNA content, without discrimination between transfected and untransfected cells in the samples. The DNA content of asynchronous proliferating cells prior to the synchronization is shown as a control (asyn.).

 
Cell cycle-dependent translation requires a 114 nt element in the p27 5'-UTR
To determine elements in the 5'-UTR that regulate translation during the cell cycle, we constructed a series of deletion mutants (Fig. 4A). Cell cycle-dependent translation rates were measured in lovastatin-arrested, thymidine-arrested or untreated asynchronous cells (Fig. 4B and C). Northern blots confirmed unaltered reporter mRNA levels under the different growth conditions (not shown). A 2.4-fold increase in translation was observed for the complete 5'-UTR of p27 in lovastatin arrested cells (Fig. 4B). Extended deletions from the 3' end of the 5'-UTR had no or little effect on the cell cycle regulated translation (Fig. 4B and C). Even the smallest 5' fragment of only 114 nt ({Delta}1) showed regulated reporter gene activity upon lovastatin arrest. Therefore elements sufficient for the increased translation of p27 in lovastatin-arrested cells are present in this minimal region that we designated the cell cycle responsive element (CCRE). The CCRE contains only a 5' portion of an uORF and in addition a G/C rich sequence. Presence of the CCRE element strongly inhibits reporter gene translation (over 10-fold in asynchronous cells, Fig. 4B). This inhibition was reduced in larger 5' fragments and neutralized in the complete 5'-UTR that contains the modular IRES (Fig. 4B). Small deletions of the 5' end of the 5'-UTR (between 35 and 158 nt) reduced and finally completely abolished the cell cycle response (Fig. 4). Even a fragment containing the entire uORF sequence ({Delta}a) showed a strongly reduced cell cycle response, suggesting that the presence of the uORF is not sufficient for cell cycle regulation.



View larger version (26K):
[in this window]
[in a new window]
  		Figure 4. Deletion analysis of the p27 5'-UTR. (A) Schematic drawing of complete and truncated p27 5'-UTR fragments. 3' truncations are indicated by numbers, 5' truncations by letters. The uORF is represented by a black box. Owing to a point mutation in fragment {Delta}c, these data should be considered confirmatory. (B) Relative luciferase activities under different growth conditions. The fragments in (A) were ligated upstream of the firefly luciferase coding sequence of pGL3 control and tested as described in Figure 2. Each fragment was tested in at least three independent experiments. (C) Induction of luciferase expression in lovastatin arrested cells compared to thymidine arrested cells. The values were calculated from the data shown in (B) and expressed as relative regulation compared to the complete 5'UTR (100%) over the non-regulated vector (0%).

 
A G/C-rich hairpin element is required for maximal cell cycle regulated translation
The fragment ({Delta}a) lacking 35 nt of the very 5' end of the 5'-UTR showed a severe reduction in cell cycle regulated translation (Fig. 4C). Reduced regulation may result from altered initiation at the uORF due to shortening of the distance between the uORF and the m7G cap structure. Alternatively, this domain might regulate translation through its structure or by binding of regulatory factors. A G/C-rich domain in this element is predicted to form a number of stable stemloop structures with similar free energy (between -23 and -24 kcal/mol) and is located just upstream of the initiation codon of the uORF (Fig. 5A).



View larger version (19K):
[in this window]
[in a new window]
  		Figure 5. A G/C rich element is required for maximal cell cycle regulation by the p27 5'-UTR. (A) Predicted energy minimized secondary structures of 5' end of the p27 5'-UTR are shown up to the initiation codon (box) of uORF (86,87). This region contains the G/C rich element that is predicted to undergo several hairpin structures with similar free energy. Structures with a {Delta}G>-23 kcal/mol are shown (1–4). Predicted secondary structure of the mutated sequence {Delta}G/C (5), where the distance between the m7G cap and the AUG codon of the uORF is maintained. The arrow indicates the end of the exchanged sequence of 35 nt. (B) Relative luciferase activities under the different growth conditions. The empty vector pGLControl and the vector pGL5'-UTR containing the p27 5'-UTR were as described in Figure 2. Thirty-five nucleotides of the G/C-rich domain upstream of the uORF in the vector pGL5'-UTR were exchanged to an unstructured A/T-rich sequence to obtain the plasmid pGL{Delta}G/C. Plasmids were analysed in synchronized cells as described in Figure 3. Each fragment was tested in six independent experiments.

 
To distinguish between a sequence dependent or a sequence independent mechanisms, we replaced 35 nucleotides of the G/C rich element by an A/T rich sequence of equal length, keeping the distance to the cap constant. The G/C content of the region upstream of the uORF is thereby reduced from 79% G/C to 45%. The novel sequence ({Delta}G/C) is predicted to lack strong secondary structure and the calculated {Delta}G drops from -23.9 kcal/mol to -9 kcal/mol (Fig. 5A, structure prediction 5). This exchange nearly completely abrogated cell cycle regulated translation (Fig. 5), suggesting that the G/C rich hairpin domain may play a key role in inducing translation during G1 progression. A prominent role for the G/C element is supported by a high degree of sequence conservation in vertebrates (Fig. 6A and B). The nucleotide sequences preceding the uORF are as highly conserved in human and chicken as the nucleotide sequences of the conserved p27 coding region.



View larger version (41K):
[in this window]
[in a new window]
  		Figure 6. The human uORF is highly conserved in mice and chicken. (A) A p27 mRNA of gallus gallus was assembled out of 2 EST clones (GenBank accession no.: AJ447507 and BU298808) and compared with a human cDNA. The positions of one predicted uORF and the p27 ORF are indicated. Sequence identities between the human and the avian sequences are indicated for different regions. The highest degree of sequence identity is found for the uORF sequence. (B) Comparison of nucleotide sequences of the human, mouse and chicken G/C-rich region upstream of the uORF and the uORFs. Identical nucleotides are shown in black boxes. (C) Comparison of the peptides encoded by the human, mouse and chicken uORFs. Identical amino acids are shown in black boxes. The small mouse ORF has not been shown to be part of the mouse p27 mRNA.

 
The uORF in the p27 5'-UTR is highly conserved and translated in vitro
The G/C-rich domain directly precedes a short open reading frame, coding for a peptide of 29 amino acids. This uORF is also highly conserved in vertebrates in its position, length and sequence. Homologous elements are present in the genomic sequence in mice (50) and in an EST cDNA clone of a chicken cDNA library (Fig. 6A). In all organisms, the uORF is positioned in a similar distance from the p27 ORF (458 nt in mice, 427 nt in humans and 405 nt in chicken). While the sequence identity in this spacer is very low between human and chicken cDNAs (24.3%), the nucleotide sequences of the avian and human uORF are 87% identical (Fig. 6A and B). This degree of identity even exceeds the amino acid sequence conservation of 69% for the encoded peptide (Fig. 6C) or the conservation of the p27 coding regions in chicken and men (77% identical nucleotides; Fig. 6A). This extensive conservation suggests an important role for the uORF that might include functions in translational regulation.

uORFs are present in several mRNAs encoding regulatory proteins connected to cell growth regulation including c-mos, Bcl-2, PDGF, MDM2, SCL or C/EBP proteins (56–59). Since ribosomal recognition of the initiation codon of many uORFs is crucial for their role in regulating downstream translation (56), we wished to determine if the human uORF can be expressed. First, we translated the p27 5'-UTR in rabbit reticulocyte lysates, resulting in synthesis of a small peptide of the expected size (pCR-{Delta}a, Fig. 7A and B). To confirm this observation, we introduced a frameshift mutation in the uORF sequence leading to an extension of the coding sequence (Fig. 7A). Translation of the otherwise identical construct led to synthesis of an extended translation product around the expected size of 16.8 kDa, replacing the 3.2 kDa peptide of the wild-type 5'-UTR (pCR-{Delta}a'; Fig. 7B).



View larger version (27K):
[in this window]
[in a new window]
  		Figure 7. The uORF in the p27 5'UTR is translated in vitro. (A) Schematic drawing of constructs analyzed in coupled in vitro transcription/translation reactions by [14C]leucine incorporation. pCR{Delta}a contains the wt p27 uORF, coding for a peptide of 3.2 kDa (seven leucine residues). pCR{Delta}a-mut contains a frameshift mutation leading to an extended uORF, coding for a 16.8 kDa protein (23 leucine residues). Open reading frames are shown in black. The transcription start sites of SP6 RNA polymerase are marked by arrows. (B) Plasmids indicated in (B) were translated in rabbit reticulocyte lysates. In vitro translation products were separated by SDS–PAGE and analyzed by fluorography. Lane ‘pCR’: a control reaction with the empty vector. The molecular masses of standard proteins are indicated in kDa.

 
uORF motifs are frequently inhibitory for translation of a downstream ORF because they inhibit ribosomal scanning. However, there are examples where they stimulate translation by enabling ribosomes to bypass cis-acting inhibitory elements or do not alter translation at all (59). For example, deletion of the initiation codon of an uORF in the PDGF mRNA failed to enhance translation of PDGF (60). To analyze the effect of the uORF on p27 translation, we changed its ATG start codon to ATT. In vitro transcribed wild-type and ATT mutant p27 mRNAs were translated in HeLa cell extracts or in rabbit reticulocyte lysates. A significant increase in p27 synthesis was observed after elimination of the initiation codon, independent of whether capped or uncapped transcripts were used (Fig. 8). Therefore the uORF can inhibit p27 translation in vitro. The presence of the m7G cap increased p27 translation 3- to 4-fold (Fig. 8), indicating that under these conditions translation initiation mainly occurred through a cap-dependent mechanism.



View larger version (42K):
[in this window]
[in a new window]
  		Figure 8. Efficient translation of p27 mRNA in vitro is inhibited by the uORF. (A) Schematic drawing of template DNA used in vitro to generate capped and uncapped p27 transcripts. A single point mutation in pCRkipATT eliminates the start codon of the uORF. (B) Capped (+) and uncapped (-) transcripts were translated in HeLa extracts or in rabbit reticulocyte lysates (RRL) in the presence of 35S-methionine. Translation products of rabbit reticulocyte lysates were analyzed directly by SDS–PAGE and fluorography, whereas p27 expressed in HeLa extracts was immunoprecipitated prior to analysis. Lane C: a control reaction in the absence of added transcript.

 
The initiation codon of the uORF but not its sequence contribute to cell cycle regulated translation in vivo
To investigate the role of the uORF in mediating regulated translation of p27 during the cell cycle, we compared the 5'-UTR with a number of point mutants that alter uORF structure. In the first mutant we deleted the uORF by converting its start codon to ATT. This mutation reduced cell cycle regulation of reporter translation to about 50% (Fig. 9), indicating that the uORF may enhance cell cycle-dependent regulation of p27.



View larger version (17K):
[in this window]
[in a new window]
  		Figure 9. The initiation codon of the uORF enhances cell cycle regulation. (A) Schematic drawing of the wild type (5'-UTR) and mutant fragments of the p27 5'-UTR. uORFs are shown as black boxes. Fragments were ligated upstream of the firefly luciferase coding sequence and analyzed as described in Figure 2. (B) Relative luciferase activities. The luciferase activities were normalized to ß-galactosidase activities and expressed relative to the values obtained from the parental vector. The figure represents the results of four independent experiments where each fragment was tested at least three times. (C) Induction of luciferase expression in lovastatin arrested cells compared to thymidine arrested cells. Values calculated from the data shown in (B) are expressed relative to the induction obtained for the complete 5'-UTR (100%). The nonregulated parental vector was set to 0%.

 
Additional mutants were constructed to investigate whether position or length of the uORF or the sequence of the encoded peptide are prerequisite for the enhancement of cell cycle regulation. In one mutant, we moved the start codon 33 nucleotides downstream of the original start site by reintroducing a start codon in the ATT mutant. With respect to cell cycle regulation, the mutant with the truncated uORF behaved identical to that lacking the entire uORF (‘ATT/ATG’; Fig. 9C). Finally we constructed a mutant where the initiation codon of the uORF was left unchanged. In this mutant, a frameshift in the third codon leads to a truncated uORF of eight codons. Only two initial amino acids of the uORF were conserved. Interestingly, this mutant maintained cell cycle regulation at levels comparable to the intact 5'-UTR (‘prestop’; Fig. 9).

These experiments clearly demonstrate that the ATG sequence of the uORF has an impact on cell cycle regulated translation. This could be due to its function as a start codon of the uORF or because the initiation codon could be part of a binding site for a yet unknown factor. If the ATG plays a role as initiation codon of the uORF, the strength of the initiation codon might alter its function in regulating translation. Therefore we performed experiments where the start site strength was altered by mutating the sequence surrounding the initiation codon.

The translation start site of the uORF is predicted to be weak. While the G in position +4 corresponds to the optimal consensus sequence, the C residue in position -3 should be a G or A base for optimal initiation (61,62). To investigate whether the rate of initiation of the uORF plays a role in regulating translation, we modified the strength of the initiation codon of the uORF. Minimal start site strength was generated by mutating the G in positon +4 to C (Fig. 10A). As for the ATT mutant, translation control was reduced to about 50%, consistent with the expected severely reduced translation of the uORF. An optimized initiation context was obtained by introducing the sequence ‘GCCACC’ immediately upstream of the ATG codon of the uORF, generating the optimized initiation context (61,62). This required an exchange of five nucleotides. The 5'-UTR containing this optimized consensus sequence also reduced cell cycle regulation to 50%. This indicates that the intermediate strength of the initiation codon is critical for cell cycle regulation, possibly by making it sensitive to alterations in cap-dependent translation. Taking together, our data demonstrate that the precise sequence surrounding the initiation codon of the uORF is important for regulating p27 translation during the cell cycle.



View larger version (16K):
[in this window]
[in a new window]
  		Figure 10. A weak translational start site of the uORF is required for maximal cell cycle regulation of translation. (A) Comparison of sequence surrounding the p27 uORF initiation codon with an optimized or minimized initiation context. (B) Relative luciferase activities under different growth conditions. 5'-UTRs with the different initiation contexts represented in (A) were inserted upstream of the firefly luciferase coding sequence of pGL3 Control plasmid and tested as described in Figure 2. Each fragment was tested in six independent experiments. (C) Induction of luciferase expression in lovastatin arrested cells compared with thymidine arrested cells. The values were calculated from the data shown in (B) and expressed relative to the induction obtained for the complete 5'-UTR (100%) over the non-regulated vector (0%).

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 REFERENCES
 
In order to investigate pathways that contribute to p27 translation we have identified novel elements on the p27 mRNA that are involved in its translational control. We found that the 5'-leader of the p27 mRNA controls translation in a cell cycle-dependent manner, stimulating translation during G1 progression or in lovastatin-arrested cells. The domain responsible for this regulation (cell cycle regulatory element, CCRE) is located at the very 5' end of the 5'-UTR. The CCRE includes a G/C-rich element predicted to form stable hairpin structures and the 5' end of an uORF. Both structural elements contribute to regulation and are highly conserved among vertebrates.

Recently, Vidal and coworkers published a study on p27 translation using a 5'-UTR that missed the CCRE identified here (51). They found that a RhoA-dependent regulation of reporter translation in non-proliferating cells involved a synergistic regulation of translation that required both, a short 5'-UTR of 151 nt and the 3' end of the p27 3'-UTR. Under our conditions, we did not observe a significant influence of the 3'-UTR in combination with the longer 5'-UTR. This might be due to different folding of the longer 5'-UTR or due to the different cell lines used.

The uORF is translated and inhibitis of p27 synthesis in vitro
The p27 uORF is translated and strongly inhibits cap-dependent translation of the downstream p27 coding region in vitro. While the uORF is required for efficient regulation of translation during the cell cycle, it does not inhibit downstream translation level in vivo. In vivo, p27 translation can occur through internal initiation through an IRES element that is positioned between the uORF and the p27 ORF (50). This IRES element appears to be inactive or assembled improperly under in vitro conditions, similar to the lack of in vitro internal initiation by the c-myc IRES (63). Factors required for internal initiation may be limiting in vitro. An example for limiting factors is the rhinovirus IRES, where additional polypyrimidine-tract binding protein (PTB) and unr protein are required for efficient internal initiation of translation in reticulocyte extracts (64,65). In fact, we have not yet been able to reconstitute IRES activity in an in vitro system. The interpretation of insufficient IRES activity is consistent with the observations that a cap structure is required in vitro for efficient p27 translation and that, in vivo, uORF-containing fragments strongly inhibit reporter translation only in the absence of the IRES.

The uORF enhances cell cycle-regulated translation
uORFs can regulate translation in lower and higher eukaryotes (59,66,67). A number of uORFs have been described that are involved in the translational regulation, especially of growth regulatory proteins. These include also cell cycle regulatory proteins like the yeast cyclin CLN3. Its synthesis can be regulated through a short uORF in its 5' leader (68). Here, we show that the p27 uORF also contributes to the cell cycle phase-dependent regulation of translation.

While the uORF nucleotide and amino acid sequences are highly conserved from mammals to birds, sequence or integrity of the uORF are surprisingly not involved in regulating translation during the cell cycle. Instead, integrity of its initiation codon and the surrounding sequence is necessary to maintain a maximum level of regulation. The intermediate strength and the position of the uORF appear to be critical for cell cycle regulation of p27 translation. Increased or decreased strength of the initiation sequence of the uORF both reduce regulation to level in the absence of the uORF, suggesting that any modulation of the start site strength interferes with efficient regulation. While we cannot formally exclude that mutagenesis of the initiation site or surrounding sequences always interferes with the binding site of an unknown factor, we favor the hypothesis that a weak initiation codon in combination with a G/C rich upstream sequence plays a central role in regulating p27 translation during the cell cycle.

A G/C-rich hairpin element contributes to cell cycle regulation
Deletion of the uORF diminishes, but does not abolish cell cycle regulation. We therefore investigated whether alternative structures present in the CCRE might contribute to cell cycle regulation. Upstream of the uORF resides a G/C-rich element that is predicted to form a number of strong secondary structures. Similar to uORFs, stable hairpins can also reduce cap-dependent translation. Deletion of 35 nucleotides and thereby destroying the predicted hairpin structure diminished the cell cycle-dependent regulation to about 50%, similar to disruption of the uORF. Surprisingly, a more dramatic loss of cell cycle regulation was observed when the G/C rich hairpin element was not deleted but replaced by an unstructured A/T rich sequence of identical length.

Since deletion of the G/C rich element has less impact on regulated translation than replacement by unstructured sequence, this domain seems to act in cooperation with the flanking m7G cap structure or downstream sequences. The function of the G/C-rich element probably involves its ability to undergo different hairpin structures with similar free energy; binding of proteins may favor one of the predicted structures and could be regulated. Changes in secondary structure in this domain may influence folding of the entire 5'-UTR or alter binding of factors to the 5'-UTR, or it may regulate the initiation strength of the uORF, for example by controlling ribosomal scanning or by bringing the cap structure in proximity to its AUG codon.

The predicted stem-loop structures start in direct proximity to the cap, either including the first nucleotide or starting 9–10 nt downstream of the cap. M. Kozak observed that a stem loop with a {Delta}G of -30 kcal/mol prevented mRNAs from engaging 40S ribosomal subunits only when the hairpin occurred 12 nt from the cap and had no deleterious effect when it was repositioned to 52 nt from the cap (69). Naturally occurring stem-loops that precede and regulate uORFs or IRES-dependent translation are primarily found in viruses, for example the hepatitis C virus (70). For efficient translation of the single ORF of the bovine viral diarrhea virus, proper folding of a stem-loop at the very 5' end of its RNA is essential (71). The trans-activation response element (TAR) at the 5' end of the human immunodeficiency virus type 1 (HIV-1) mRNA forms a stable hairpin that is a target for binding of viral and cellular proteins. The viral Tat protein binds to the hairpin and activates viral gene expression, the cellular La autoantigen, that also binds to the TAR hairpin, relieving its sequence-specific cis-inhibitory repression of translation (72). In a similar way, different hairpin structures in combination with different proteins may regulate p27 translation.

Together with our observation that the strength of the initiation at the uORF seems to be crucial for its cell cycle regulatory function, one possible model for CCRE function may be that the G/C-rich hairpin is regulated by cell cycle-specific expressed proteins and cooperates with the uORF as sensor of cap-dependent translation levels. A detailed analysis of proteins or protein complexes that are able to bind to the CCRE preferentially in a cell cycle-controlled manner and a structure-probing analysis of the hairpin domain should help to determine the exact molecular mechanism of the cell cycle regulation.

p27 translation and hyperproliferative diseases
Reduction of p27 expression is emerging as a fundamental step for the development of human malignancies (73,74). Various signals converge in the regulation of p27, altering level of the protein by using diverse mechanisms including transcription (29–32), translation (33–35), p27 stability (36–38) or localization (36–38). This may allow integration of multiple growth regulatory pathways at the level of p27 expression. Consistent with the prominent role of p27 in oncogenesis, it was discovered that reduced levels of the inhibitor provide independent prognostic information in many human tumors (75–77) and that p27 cooperates with mutations in several oncogenes and tumor suppressor genes to facilitate tumor growth (77,78). Increasing evidence also suggests that the deregulation of p27 in cancer is a result of deregulation of distinct pathways including p27 mislocalization to the cytoplasm (41–43,79) or interference with its ubiquitin-dependent degradation (80,81). In addition, deregulation of translation is associated with a broad range of human cancers (82).

We observed roughly 2-fold changes in translation rates during the cell cycle. Biological significance of two fold changes was clearly demonstrated in mice lacking only one allele of p27 (25–27). The heterozygous animals display an intermediate phenotype of multiorgan hyperplasia and develop tumors when challenged with irradiation or carcinogens (28), consistent with the notion that p27 is haplo-insufficient for tumor suppression (28,77). Consequently, loss of p27 translation during the cell cycle may well contribute to tumorigenesis. Therefore detailed understanding of the mechanisms of p27 translation should not only help to better understand its deregulation in hyperproliferative diseases but may also support the development of novel approaches to the treatment of cancer and other hyperproliferative diseases.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
REFERENCES
 
DNA manipulations
The isolation of the p27 5'-UTR was described previously (50). The complete 3' end of the p27 cDNA was obtained as EST clone from the Resource Center of the German Genome Project (cDNA clone IMAGp998F02581) (83). A 234 bp XbaI–NotI fragment from this clone was inserted into pUG32 creating a complete 3'-UTR in the plasmid pUG216.

Deletion mutants of the p27 5'-UTR were generated by PCR using oligonucleotide primers containing HindIII or NcoI sites (primer sequences: see below). Owing to a mutation in the primer, the C residue in position-457 of the original cDNA was lost in the fragment {Delta}c spanning nucleotides-461 to -3 of the 5'-UTR. The data obtained with the fragment should therefore be considered as confirmatory. Products were inserted into the HindIII and NcoI sites of pGL3 control (Promega) upstream of the firefly luciferase reporter gene. Point mutations (‘ATT’, G-517->T; ‘prestop’, insertion C-514; and ‘ATT/ATG’, G-517->T, G-486->A, T-483->G) were introduced using primers as indicated below. The p27 3'-UTR was isolated by PCR and inserted into the XbaI site of pGL3 control and pGL5'UTR creating the plasmids pGL3'-UTR and pGL5'+3'-UTR respectively. The constructs pGLoptimal and pGLminimal were obtained by PCR-based overlap extension procedure (84) with the appropriate primer pairs indicated below. As a template for PCR, the complete p27-5'-UTR in a bicistronic reporter construct (50) was used. In the case of cloning pGLoptimal, the PCR products obtained after overlap extension were digested with HindIII and NcoI and ligated into similarly cut pGL3 control reporter. For pGL3-minimal, PCR-products were digested with HindIII and NarI and ligated into similarly cut pGL3 control vector. The construct pGL{Delta}G/C contained a 5'-UTR-sequence equal in length to pGL5'-UTR; it only differed in the first 35 bp which were exchanged from G/C- to A/T-rich, thereby destroying the highly structured 5' end of p27-5'-UTR. This was achieved by annealing {Delta}G/C primers indicated below. The duplexes were phosphorylated and ligated into HindIII-digested, calf intestinal phosphatase-treated pGL{Delta}a.

For in vitro translation experiments, an interfering initiation codon of the vector pCRII (between the SP6 transcription start site and the HindIII cloning site) was removed by site-directed mutagenesis to create the vector pCR'. pCRkip1 was obtained after insertion of the p27 5'-UTR and the p27 ORF and 3'-UTR in pCR'. The initiation codon of the uORF was removed in pCRkip1ATT. pCR{Delta}a and pCR{Delta}a-mut were generated by ligating the p27 5'-UTR fragments {Delta}a and {Delta}a-mut into the HindIII and NcoI sites of pCR'. {Delta}a-mut differs from {Delta}a by a single base deletion in position-499 resulting from PCR mutagenesis.

The following oligonucleotides were used for plasmid construction: {Delta}1—5'-AAGCTTCCACCTTAAGGCCGCGCT-3' and 5'-CCATGGTCTCTGACGAAGAAGAAAATGATTG-3'; {Delta}2—5'-AAGCTTCCACCTTAAGGCCGCGCT-3' and 5'-CCATGGTTCTCGGGGAGAAAAACACCCCGAAA-3'; {Delta}3—5'-AAGCTTCCACCTTAAGGCCGCGCT-3' and 5'-CCATGGTTCTCCTTGCCGGCGTCGGAGTCGCAG-3'; {Delta}4—5'-AAGCTTCCACCTTAAGGCCGCGCT-3' and 5'-CCATGGTTCTCCAAGCGGAGAGGGTGGCAAAG-3'; {Delta}a—5'-AAGCTTCGGCTCCCGCCGCCGCAACCA-3' and 5'-CCATGGTTCTCCCGGGTCTGCACG-3'; {Delta}b—5'-AAGCTTCTTCTTCGTCAGCCTCCCT-3' and 5'-CCATGGTTCTCCCGGGTCTGCACG-3'; {Delta}c—5'-AAGCTTGGATCCGCGGCCTCCTTCCACC-3' and 5'-CCATGGTTCTCCCGGGTCTGCACG-3'; {Delta}d—5'-AAGCTTCGTCTTTTCGGGGTGTTTTTC-3' and 5'-CCATGGTTCTCCCGGGTCTGCACG-3'; {Delta}e—5'-AAGCTTCCGACGCCGGCAAGGTTTGGA-3' and 5'-CCATGGTTCTCCCGGGTCTGCACG-3'; 3'UTR—5'-AAAAAAACTAGTACAGCTCGAATTAAGAATATGTTTCCT-3' and 5'-AAAAAAACTAGTGAATAGCTATGGAAGTTTTCTTTATTGA-3'; {Delta}G/C—5'-AGCTTTAAGTAAATAAAAACCAGTTTCGTTAATAA-3' and 5'-AGCTTTATTAACGAAACTGGTTTTTATTTACTTAA-3'; ATT—5'-CCGCCGCAACCAATTGATCTCCTCCTCTG-3' and 5'-CAGAGGAGGAGATCAATTGGTTGCGGCGG-3'; prestop—5'-GCCGCAACCAATGGATCCTCCTCCTCTGTTTA-3' and 5'-TAAACAGAGGAGGAGGATCCATTGGTTGCGGC-3'; ATT/ATG—5'-CTGTTTAAATAGACTCGCCATGGCAATCATTTTCTTCTTCGTC-3' and 5'-GACGAAGAAGAAAATGATTGCCATGGCGAGTCTATTTAAACAG-3'; optimal—5'-GCTCCCGCCGCCGGCCACCATGGATCTCCTCCTCTG-3' and 5'-CAGAGGAGGAGATCCATGGTGGCCGGCGGCGGGAGC-3'; minimal—5'-GCCGCAACCAATGCATCTCCTCCTCTG-3' and 5'-CAGAGGAGGAGATGCATTGGTTGCGGC-3'; U40—5'-GCGTATCTCTTCATAGCCTT-3'; U39—5'-GCAAGAAGATGCACCTGATG-3'; pCRII Vector mutagenesis—5'-GACACTATAGAATACTCAAGCTCATCAAGCTTGGTACCGAGC-3' and 5'-GCTCGGTACCAAGCTTGATGAGCTTGAGTATTCTATAGTGTC-3'.

RNA analysis
Multiple tissue northern blots were obtained from Clontech. Total RNA was isolated with TriReagent (Sigma) or the RNeasy Mini Kit (Qiagen). RNA was denatured, separated on a 1% formaldehyde-agarose gels, transferred to a nylon membrane and UV-crosslinked. Radioactive-labeled probes were generated by random prime labeling or PCR amplification. Hybridizations were performed as described previously (85). Blots were analyzed by autoradiography and by using a phosphorImager.

RNA structure prediction was performed using the program mfold version 3.1. (www.bioinfo.rpi.edu/applications/mfold/) (86–88). Structures are predicted for 37°C under ionic conditions of 1 M NaCl.

Cell culture and cell cycle synchronization
HeLa cells were arrested in G1 phase by 24 h treatment with 66 µM lovastatin (MSD) and in S phase by treatment with 2 mM thymidine as described (35,89). Flow cytometry analysis of the DNA content to determine the cell cycle phase distribution was performed using standard procedures and FITC coupled antibodies (BD) according to the manufacturers instructions. Cells were transfected with SuperFect reagent (Qiagen). A 0.2 µg sample of reporter plasmid was co-transfected with 1.8 µg pSV-B-Gal vector (Promega). After 3–6 h the transfection medium was removed and the cells were incubated for additional 24 h in medium containing either 66 µM lovastatin, 2 mM thymidine, or solvent. To synchronize HeLa cells, 24 h after transfection cells were incubated in 4 mM thymidine for 16 h, released for 6 h and incubated in 50 ng/ml nocodazole for 10 h. Synchronisation of cells for pulse experiments after nocodazole release were performed as described previously (35).

Protein analysis and reporter gene activity assays
To determine protein synthesis rates and stabilities, pulse-chase experiments of HeLa cells were performed as described earlier (35). Cells transfected with monocistronic reporter plasmids were lysed in Reporter Lysis Buffer (Promega) and firefly luciferase expression was determined using the Luciferase Assay System (Promega). ß-Galactosidase activity was quantified in the same lysate using the ß-Galactosidase Enzyme Assay System (Promega).

In vitro transcription
A standard reaction mixture for in vitro transcription contained 0.5–1 µg linearized DNA template, 20 U SP6 RNA polymerase (Roche), 2 µl 10-fold transcription buffer (Roche) and 20 U RNasin (Promega) in a final volume of 20 µl. For synthesis of capped RNA 1 mM ATP, 1 mM CTP, 1 mM UTP, 0.1 mM GTP and 1 mM m7GpppG were included. For synthesis of uncapped RNAs m7GpppG was omitted and ATP, CTP, UTP and GTP were used at 0.5 mM concentration. After incubation for 1–2 h at 37°C, 1 U RNase-free DNaseI (Promega) was added. After additional 20 min at 37°C transcripts were purified by gelfiltration on Sephadex G-25, phenol/chloroform extraction and ethanol precipitation and quantified by measuring absorbance at 260 nm.

In vitro translation
Coupled in vitro transcription/translation reactions were performed using the TNT Coupled Reticulocyte Lysate System (Promega). A 1 µg sample of supercoiled plasmid DNA was translated in a 50 µl reaction in the presence of 250 nCi [14C] leucine (314 mCi/mmol). After 90 min at 30°C, proteins were separated by SDS–PAGE and analyzed by fluorography. In vitro translation with the rabbit reticulocyte lysate system (Promega) was performed for 90 min at 30°C in a reaction volume of 20 µl containing 100 ng RNA, 13.2 µl lysate, 2 µl 1 mM aminoacid mixture without methionine, 20 U RNasin (Promega), 8 µCi [35S] methionine (1000 Ci/mmol), 80 mM K+ and 1.4 mM Mg2+. Translation products were resolved by SDS–PAGE and visualized by fluorography. RNA transcripts were recovered from an aliquot of the reaction mixture and analyzed for integrity by northern blot analysis.

HeLa lysates were prepared as described (90). For in vitro translation 20 µl HeLa lysate, 10 µCi [35S] methionine (1000 Ci/mmol), 125 pmol methionine and 125 ng RNA template were combined in a total volume of 25 µl and incubated for 1 h at 30°C. p27 was immunoprecipitated and [35S] methionine incorporation was analyzed by SDS–PAGE and fluorography.


   ACKNOWLEDGEMENTS
 
We thank Sandera Kölle from MSD Sharp & Dohme for lovastatin. We also thank members of the laboratory for stimulating discussions and Frauke Melchior for critical reading of the manuscript. This work was funded by the Max-Planck Society, the government of Oberbayern and Roche, Penzberg.


   FOOTNOTES
 
* To whom correspondence should be addressed. Tel: +49 8985783969; Fax: +49 8985782361; Email: hengst{at}biochem.mpg.de Back


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 

  1. Ekholm, S.V. and Reed, S.I. (2000) Regulation of G(1) cyclin-dependent kinases in the mammalian cell cycle. Curr. Opin. Cell Biol., 12, 676–684.[CrossRef][ISI][Medline]

  2. Morgan, D.O. (1997) Cyclin-dependent kinases: engines, clocks, and microprocessors. A. Rev. Cell Dev. Biol., 13, 261–291.[CrossRef][ISI][Medline]

  3. Kaldis, P. (1999) The cdk-activating kinase (CAK): from yeast to mammals. Cell Mol. Life Sci., 55, 284–296.[CrossRef][ISI][Medline]

  4. Lew, D.J. and Kornbluth, S. (1996) Regulatory roles of cyclin dependent kinase phosphorylation in cell cycle control. Curr. Opin. Cell Biol., 8, 795–804.[CrossRef][ISI][Medline]

  5. Nilsson, I. and Hoffmann, I. (2000) Cell cycle regulation by the Cdc25 phosphatase family. Prog. Cell Cycle Res., 4, 107–114.[Medline]

  6. Carnero, A. and Hannon, G.J. (1998) The INK4 family of CDK inhibitors. Curr. Top. Microbiol. Immunol., 227, 43–55.[ISI][Medline]

  7. Jeffrey, P.D., Tong, L. and Pavletich, N.P. (2000) Structural basis of inhibition of CDK-cyclin complexes by INK4 inhibitors. Genes Dev., 14, 3115–3125.[Abstract/Free Full Text]

  8. Hengst, L. and Reed, S.I. (1998) Inhibitors of the Cip/Kip family. Curr. Top. Microbiol. Immunol., 227, 25–41.[ISI][Medline]

  9. Sherr, C.J. and Roberts, J.M. (1995) Inhibitors of mammalian G1 cyclin-dependent kinases. Genes Dev., 9, 1149–1163.[Free Full Text]

  10. Sherr, C.J. and Roberts, J.M. (1999) CDK inhibitors: positive and negative regulators of G1-phase progression. Genes Dev., 13, 1501–1512.[Free Full Text]

  11. Russo, A.A., Jeffrey, P.D., Patten, A.K., Massague, J. and Pavletich, N.P. (1996) Crystal structure of the p27Kip1 cyclin-dependent-kinase inhibitor bound to the cyclin A-Cdk2 complex. Nature, 382, 325–331.[CrossRef][Medline]

  12. Hengst, L., Göpfert, U., Lashuel, H.A. and Reed, S.I. (1998) Complete inhibition of Cdk/cyclin by one molecule of p21(Cip1). Genes Dev., 12, 3882–3888.[Abstract/Free Full Text]

  13. Soos, T.J., Kiyokawa, H., Yan, J.S., Rubin, M.S., Giordano, A., DeBlasio, A., Bottega, S., Wong, B., Mendelsohn, J. and Koff, A. (1996) Formation of p27-CDK complexes during the human mitotic cell cycle. Cell Growth Differ., 7, 135–136.[Abstract]

  14. Blain, S.W., Montalvo, E. and Massague, J. (1997) Differential interaction of the cyclin-dependent kinase (Cdk) inhibitor p27Kip1 with cyclin A-Cdk2 and cyclin D2-Cdk4. J. Biol. Chem., 272, 25863–25872.[Abstract/Free Full Text]

  15. LaBaer, J., Garrett, M.D., Stevenson, L.F., Slingerland, J.M., Sandhu, C., Chou, H.S., Fattaey, A. and Harlow, E. (1997) New functional activities for the p21 family of CDK inhibitors. Genes Dev., 11, 847–862.[Abstract/Free Full Text]

  16. Cheng, M., Olivier, P., Diehl, J.A., Fero, M., Roussel, M.F., Roberts, J.M. and Sherr, C.J. (1999) The p21(Cip1) and p27(Kip1) CDK ‘inhibitors’ are essential activators of cyclin D-dependent kinases in murine fibroblasts. EMBO J., 18, 1571–1583.[CrossRef][ISI][Medline]

  17. Polyak, K., Lee, M.H., Erdjument-Bromage, H., Koff, A., Roberts, J.M., Tempst, P. and Massague, J. (1994) Cloning of p27Kip1, a cyclin-dependent kinase inhibitor and a potential mediator of extracellular antimitogenic signals. Cell, 78, 59–66.[CrossRef][ISI][Medline]

  18. Polyak, K., Kato, J.Y., Solomon, M.J., Sherr, C.J., Massague, J., Roberts, J.M. and Koff, A. (1994) p27Kip1, a cyclin-Cdk inhibitor, links transforming growth factor-beta and contact inhibition to cell cycle arrest. Genes Dev., 8, 9–22.[Abstract/Free Full Text]

  19. Hengst, L., Dulic, V., Slingerland, J.M., Lees, E. and Reed, S.I. (1994) A cell cycle-regulated inhibitor of cyclin-dependent kinases. Proc. Natl Acad. Sci. USA, 91, 5291–5295.[Abstract/Free Full Text]

  20. Toyoshima, H. and Hunter, T. (1994) p27, a novel inhibitor of G1 cyclin-Cdk protein kinase activity, is related to p21. Cell, 78, 67–74.[CrossRef][ISI][Medline]

  21. Harper, J.W., Elledge, S.J., Keyomarsi, K., Dynlacht, B., Tsai, L.H., Zhang, P., Dobrowolski, S., Bai, C., Connell-Crowley, L., Swindell, E. et al. (1995) Inhibition of cyclin-dependent kinases by p21. Mol. Biol. Cell, 6, 387–400.[Abstract]

  22. Slingerland, J.M., Hengst, L., Pan, C.H., Alexander, D., Stampfer, M.R. and Reed, S.I. (1994) A novel inhibitor of cyclin-Cdk activity detected in transforming growth factor beta-arrested epithelial cells. Mol. Cell. Biol., 14, 3683–3694.[Abstract/Free Full Text]

  23. Coats, S., Flanagan, W.M., Nourse, J. and Roberts, J.M. (1996) Requirement of p27Kip1 for restriction point control of the fibroblast cell cycle. Science, 272, 877–880.[Abstract]

  24. Rivard, N., L'Allemain, G., Bartek, J. and Pouyssegur, J. (1996) Abrogation of p27Kip1 by cDNA antisense suppresses quiescence (G0 state) in fibroblasts. J. Biol. Chem., 271, 18337–18341.[Abstract/Free Full Text]

  25. Fero, M.L., Rivkin, M., Tasch, M., Porter, P., Carow, C.E., Firpo, E., Polyak, K., Tsai, L.H., Broudy, V., Perlmutter, R.M. et al. (1996) A syndrome of multiorgan hyperplasia with features of gigantism, tumorigenesis, and female sterility in p27(Kip1)-deficient mice. Cell, 85, 733–744.[CrossRef][ISI][Medline]

  26. Kiyokawa, H., Kineman, R.D., Manova-Todorova, K.O., Soares, V.C., Hoffman, E.S., Ono, M., Khanam, D., Hayday, A.C., Frohman, L.A. and Koff, A. (1996) Enhanced growth of mice lacking the cyclin-dependent kinase inhibitor function of p27(Kip1). Cell, 85, 721–732.[CrossRef][ISI][Medline]

  27. Nakayama, K., Ishida, N., Shirane, M., Inomata, A., Inoue, T., Shishido, N., Horii, I. and Loh, D.Y. (1996) Mice lacking p27(Kip1) display increased body size, multiple organ hyperplasia, retinal dysplasia, and pituitary tumors. Cell, 85, 707–720.[CrossRef][ISI][Medline]

  28. Fero, M.L., Randel, E., Gurley, K.E., Roberts, J.M. and Kemp, C.J. (1998) The murine gene p27Kip1 is haplo-insufficient for tumour suppression. Nature, 396, 177–180.[CrossRef][Medline]

  29. Inoue, T., Kamiyama, J. and Sakai, T. (1999) Sp1 and NF-Y synergistically mediate the effect of vitamin D(3) in the p27(Kip1) gene promoter that lacks vitamin D response elements. J. Biol. Chem., 274, 32309–32317.[Abstract/Free Full Text]

  30. Shaffer, A.L., Yu, X., He, Y., Boldrick, J., Chan, E.P. and Staudt, L.M. (2000) BCL-6 represses genes that function in lymphocyte differentiation, inflammation, and cell cycle control. Immunity, 13, 199–212.[CrossRef][ISI][Medline]

  31. Kolluri, S.K., Weiss, C., Koff, A. and Gottlicher, M. (1999) p27(Kip1) induction and inhibition of proliferation by the intracellular Ah receptor in developing thymus and hepatoma cells. Genes Dev., 13, 1742–1753.[Abstract/Free Full Text]

  32. Medema, R.H., Kops, G.J., Bos, J.L. and Burgering, B.M. (2000) AFX-like Forkhead transcription factors mediate cell-cycle regulation by Ras and PKB through p27kip1. Nature, 404, 782–787.[CrossRef][Medline]

  33. Agrawal, D., Hauser, P., McPherson, F., Dong, F., Garcia, A. and Pledger, W.J. (1996) Repression of p27kip1 synthesis by platelet-derived growth factor in BALB/c 3T3 cells. Mol. Cell. Biol., 16, 4327–4336.[Abstract]

  34. Millard, S.S., Yan, J.S., Nguyen, H., Pagano, M., Kiyokawa, H. and Koff, A. (1997) Enhanced ribosomal association of p27(Kip1) mRNA is a mechanism contributing to accumulation during growth arrest. J. Biol. Chem., 272, 7093–7098.[Abstract/Free Full Text]

  35. Hengst, L. and Reed, S.I. (1996) Translational control of p27Kip1 accumulation during the cell cycle. Science, 271, 1861–1864.[Abstract]

  36. Pagano, M., Tam, S.W., Theodoras, A.M., Beer-Romero, P., Del Sal, G., Chau, V., Yew, P.R., Draetta, G.F. and Rolfe, M. (1995) Role of the ubiquitin-proteasome pathway in regulating abundance of the cyclin-dependent kinase inhibitor p27. Science, 269, 682–685.[Abstract/Free Full Text]

  37. Hara, T., Kamura, T., Nakayama, K., Oshikawa, K. and Hatakeyama, S. (2001) Degradation of p27(Kip1) at the G(0)–G(1) transition mediated by a Skp2-independent ubiquitination pathway. J. Biol. Chem., 276, 48937–48943.[Abstract/Free Full Text]

  38. Malek, N.P., Sundberg, H., McGrew, S., Nakayama, K., Kyriakides, T.R., Roberts, J.M. and Kyriakidis, T.R. (2001) A mouse knock-in model exposes sequential proteolytic pathways that regulate p27Kip1 in G1 and S phase. Nature, 413, 323–327.[CrossRef][Medline]

  39. Rodier, G., Montagnoli, A., Di Marcotullio, L., Coulombe, P., Draetta, G.F., Pagano, M. and Meloche, S. (2001) p27 cytoplasmic localization is regulated by phosphorylation on Ser10 and is not a prerequisite for its proteolysis. EMBO J., 20, 6672–6682.[CrossRef][ISI][Medline]

  40. Connor, M.K., Kotchetkov, R., Cariou, S., Resch, A., Lupetti, R., Beniston, R.G., Melchior, F., Hengst, L. and Slingerland, J.M. (2003) CRM1/RAN-mediated nuclear export of p27Kip1 involves a nuclear export signal and links p27 export and proteolysis. Mol. Biol. Cell, 14, 201–213.[Abstract/Free Full Text]

  41. Viglietto, G., Motti, M.L., Bruni, P., Melillo, R.M., D'Alessio, A., Califano, D., Vinci, F., Chiappetta, G., Tsichlis, P., Bellacosa, A. et al. (2002) Cytoplasmic relocalization and inhibition of the cyclin-dependent kinase inhibitor p27(Kip1) by PKB/Akt-mediated phosphorylation in breast cancer. Nat. Med., 8, 1136–1144.[CrossRef][ISI][Medline]

  42. Liang, J., Zubovitz, J., Petrocelli, T., Kotchetkov, R., Connor, M.K., Han, K., Lee, J.H., Ciarallo, S., Catzavelos, C., Beniston, R. et al. (2002) PKB/Akt phosphorylates p27, impairs nuclear import of p27 and opposes p27-mediated G1 arrest. Nat. Med., 8, 1153–1160.[CrossRef][ISI][Medline]

  43. Shin, I., Yakes, F.M., Rojo, F., Shin, N.Y., Bakin, A.V., Baselga, J. and Arteaga, C.L. (2002) PKB/Akt mediates cell-cycle progression by phosphorylation of p27(Kip1) at threonine 157 and modulation of its cellular localization. Nat. Med., 8, 1145–1152.[CrossRef][ISI][Medline]

  44. Harper, J.W. (2001) Protein destruction: adapting roles for Cks proteins. Curr. Biol., 11, R431–435.[CrossRef][ISI]