This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Hu, M. G. Articles by Wong, D. T. W. Search for Related Content PubMed PubMed Citation Articles by Hu, M. G. Articles by Wong, D. T. W.

Role of p12^CDK2-AP1 in Transforming Growth Factor-ß1-Mediated Growth Suppression

¹ Department of Pathology, and
² Center for Biochemical and Biophysical Sciences and Medicine, Harvard Medical School, Boston Massachusetts;
³ Laboratory of Head and Neck Cancer Research, Dental Research Institute,
⁴ Jonsson Comprehensive Cancer Center, and
⁵ Molecular Biology Institute, University of California at Los Angeles, Los Angeles, California

	ABSTRACT

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
p12^CDK2-AP1 (p12) is a growth suppressor isolated from normal keratinocytes. Ectopic expression of p12 in squamous carcinoma cells reversed the malignant phenotype of these cells, in part due an ability of p12 to bind to both DNA polymerase /primase and to cyclin-dependent kinase 2 (CDK2), thereby inhibiting their activities. We report in this article that in normal epithelial cells, transforming growth factor ß1 (TGF-ß1) induces p12 expression transcriptionally, which, in turn, mediates the growth inhibitory activity of TGF-ß1. We created inducible p12 antisense HaCaT cell lines [ip12 (-) HaCaT] and showed that selective reduction of cellular p12 resulted in an increase in: (a) CDK2-associated kinase activity; (b) protein retinoblastoma (pRB) phosphorylation; and (c) [³H]thymidine incorporation, and partially reversed TGF-ß1-mediated inhibition of CDK2 kinase activity, pRB phosphorylation, and cell proliferation. Furthermore, we generated p12-deficient mouse oral keratinocytes (MOK^p12-/-) and compared their growth characteristics and response to TGF-ß1 with that of wild-type mouse oral keratinocytes (MOK^WT). Under normal culture conditions, the number of MOK^p12-/- in S phase is 2-fold greater than that of MOK^WT. Concomitantly, fewer cells are in G₂ phase in MOK^p12-/- than that in MOK^WT. Moreover, response to TGF-ß1-mediated growth suppression is compromised in MOK^p12-/- cells. Mechanistic studies showed that MOK^p12-/- have increased CDK2 activity and reduced sensitivity to inhibition by TGF-ß1. Collectively our data suggest that p12 plays a role in TGF-ß1-mediated growth suppression by modulating CDK2 activities and pRB phosphorylation.

	INTRODUCTION

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
p12^CDK2-AP1 (p12; official Human Genome Organization designation is p12^CDK2-AP1) is a highly conserved growth suppressor identified and isolated from normal keratinocytes (1) . p12 cDNA has been cloned from human, mouse, and hamster (1, 2, 3, 4) . Human p12 consists of 115 amino acid residues with a molecular weight of M_r 12,400. Human and rodent p12 share 97% identity, and the mouse and hamster p12 sequences are identical. Ectopic expression of p12 in squamous carcinoma cells increased doubling time and reversed in vitro transformation phenotypes including anchorage independence, suggesting that p12 is a growth suppressor (5) . The clinical behavior of p12 was evaluated recently by immunostaining in 100 cases of oral squamous cell carcinoma (6) . p12 nuclear staining was reduced or absent in 69 cases (69%) of oral squamous cell carcinoma. Decrease or loss of p12 staining correlated positively with the degree of tumor invasion, proliferative index, and outcome of patient 10-year survival. These data suggest that reduction or loss of p12 expression is a negative prognostic indicator in patients with surgically resected oral squamous cell carcinoma. Mechanistic studies have shown that p12 binds to DNA polymerase /primase, thereby inhibiting DNA replication (5) . More recently, p12 has been shown to associate with and negatively regulate cyclin-dependent kinase (CDK) 2 activities by sequestering the monomeric (inactive) pool of CDK2 and by targeting CDK2 for proteolysis on overexpression of p12 in the cells (7) . However, the physiological relevance of the interaction between p12 and CDK2 remain unclear.

Transforming growth factor (TGF)-ß1 is a cytokine that has profound growth suppressive effects on epithelial, endothelial, and hematopoietic cells (8, 9, 10) . The mechanism of TGF-ß1 inhibition in epithelial cells is mediated in part by two classes of rapid gene responses: (a) induction of cell-cycle-inhibitory proteins, such as cyclin-dependent kinase inhibitory (CKI) proteins p15^INK4B and p21^CIP1; and (b) down-regulation of proliferative proteins such as c-myc (11) , which prevents activation of genes that encode CKIs, such as p15^{INK4B and}p21^CIP1, by TGF-ß1 (12 , 13) . p15^INK4B and p21^CIP1 directly inhibit the activity of G₁ phase CDKs (CDK4, CDK6, and CDK2) by associating with their target cyclins, CDKs, or cyclin-CDK complexes to inhibit their activities and subsequent protein retinoblastoma (pRB) phosphorylation (14 , 15) . It has been suggested that cyclin D- and E-dependent kinases, complexed with CDK4/6 and CDK2, respectively, contribute sequentially and coordinately to phosphorylate pRB at the restriction point (16 , 17) . CDK2-mediated pRB phosphorylation is believed to be responsible for the final inactivating phosphorylation of pRB before G₁-S transition (17) . A model was proposed to explain the cell cycle events leading to TGF-ß-mediated growth inhibition in proliferating Mv1Lu cells (15) . The increased p15^INK4B protein induced by TGF-ß1 binds to cyclin D-kinase complexes and facilitates the subsequent binding of p27^KIP1 to CDK2-cyclin E complexes, resulting in inhibition of CDK2, pRB phosphorylation, and G₁ arrest (12) . However, the increased expression of known CKIs in response to TGF-ß1 is not sufficient to confer cell cycle arrest by TGF-ß1. Genetic evidence strongly suggests that TGF-ß1 can still induce arrest in cells derived from either p27^KIP1 (18) or p15^INK4B knockout animals (19) . The absence of p15^INK4B is not compensated by increased levels of expression or increased association with CDK4 or other members of Ink4 or Cip/Kip families (19) . Moreover, in many cell types, including intestinal epithelial cells, p15^INK4B is not increased by TGF-ß1treatment, although TGF-ß1 is a potent growth inhibitor of exponentially proliferating cultures of intestinal epithelial 4–1 cells (20 , 21) , suggesting that the loss of p27^KIP1 or p15^INK4B is likely to be compensated for by some other yet undiscovered physiological CDK inhibitors. We present data to show that p12 is one of such inhibitors.

The signaling pathways downstream of TGF-ß1 receptor complexes that lead to increased expression of target genes are still incompletely understood. It has been proposed that signaling mediators of the SMAD protein family, Smad2, Smad3, and Smad4, participate in transcriptional activation of target genes (22, 23, 24, 25, 26, 27) through direct interaction with a specific DNA sequence (i.e. CAGA), the Smad binding elements (SBEs; Refs. 28 , 29 ). It has also shown that Sp1 and Sp3 are involved in the regulation of the target genes (30) . We show here that Smad proteins may participate in TGF-ß1-mediated p12 expression, further supporting a direct role for p12 as an additional CDK inhibitor that mediates the growth suppression activity of TGF-ß1.

	MATERIALS AND METHODS

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Cell Culture.
Human primary foreskin keratinocytes were a gift of Dr. Karl Münger (Harvard Medical School). The human primary foreskin keratinocytes were maintained in keratinocyte serum-free medium, containing 0.1% gentamicin and 1% antibiotic-antimycotic (Invitrogen, Grand Island, NY). The mink lung epithelial cell line Mv1Lu, HaCaT, and HeLa cells were obtained from the American Type Culture Collection (Rockville, MD). These cells, and the inducible p12 antisense cell lines [ip12 (-) HaCaT] and vector control cell lines (pMTCB6⁺-HaCaT) were maintained in DMEM supplemented with 5% fetal bovine serum, 0.1% gentamicin, and 1% antibiotic-antimycotic.

Generation of Mouse Oral Keratinocytes (MOKs) Cell Lines.
MOKs were isolated from BALB/c wild-type and p12 knockout mice.⁴ Tongue and palatal tissues were dissected from mice that were asphyxiated with CO₂. The tissues were incubated in 1x HBSS without calcium and magnesium chloride (Invitrogen) supplemented with streptomycin 100 µg/ml, gentamicin 5 µg/ml (Invitrogen), and nystatin 100 units/ml (Sigma). Tongue samples were cut longitudinally into three to four pieces. Palates were left as whole pieces. All of the tissues were digested with 1 unit/ml dispase (Roche Molecular Biochemicals) in HBSS at 4°C for 18 h to separate the epithelium from the connective tissue. The epithelium was cut into smaller pieces using sterile disposable scalpels. The tissue was incubated in trypsin-EDTA (Invitrogen) for 5 min. DMEM (Invitrogen) supplemented with 10% fetal bovine serum (JRH) was added to inactivate trypsin. The cell pellets were centrifuged and washed with 1x PBS. The pellets were resuspended in keratinocyte serum-free medium without calcium chloride (Invitrogen) supplemented with 45 µg/ml bovine pituitary extract (Invitrogen), 1 ng/ml recombinant epidermal growth factor (Invitrogen), 14% DMEM (Invitrogen), 0.1% fetal bovine serum (JRH), 0.018 mM calcium chloride (Fisher), 100 units/ml penicillin (Invitrogen), and 5 µg/ml gentamicin (Invitrogen). Cultures were grown at 37°C with 5% CO₂. Cells were subcultured when they reached 50–70% confluence.

Preparation of Inducible p12 Antisense Cell Lines.
A DNA fragment (74 nucleotides) from exon 1 of p12 cDNA starting at the first ATG was amplified by reverse transcription-PCR with human placental total RNA (Clontech Laboratories, Palo Alto, CA) as the template. Primers for the PCR were 5'-AAT TCG GTA CCT TGA GGG CGG CGG CGG GCA TGT G-3' containing a KpnI site sequence (forward primer), and 5'-ACC CGG GGA TCC ATG TCT TAC AAA CCG AAC TTG G-3' containing a BamHI site (reverse primer). PCR reaction was performed as follows, 94°C for 5 min, then 30 cycles of 94°C for 1 min, 55°C for 1 min, and 72°C for 1 min. The PCR amplicon was isolated from 2% agarose gel, and digested with KpnI and BamHI endonucleases to create the cloning sites. The cDNA fragment was cloned into KpnI/BamHI cloning sites of the zinc-inducible expression vector pMT/CB6⁺ (a gift of Dr. Shuki Mizutani, The National Children’s Medical Research Center, Japan; Ref. 31 ) in antisense orientation, and transformed into XL-1 Blue cells. The plasmid was transfected into HaCaT cells using Lipofectamine (Invitrogen) according to the manufacturer’s protocol. Stable transfectants were selected with 1 mg/ml G418 for 2 weeks, and individual clones were isolated. The presence of antisense transcripts was confirmed by reverse transcription-PCR with primers designed in pMT/CB6⁺ vector sequence and human p12 sequence. Three clones (#1, #4, and #9) were selected that demonstrated Zn-inducible down-regulation of endogenous p12. Each clone was independently expanded. Clone#1 [ip12(-)HaCaT #1] was used in this article [hereafter referred as ip12(-)HaCaT]. Five individual clones stably transfected with the pMT/CB6⁺ parental vector were selected in which it was demonstrated that endogenous p12 is not affected by Zn treatment. Control clone#2 (pMTCB6⁺-HaCaT #2) was used in this article (hereafter referred as pMTCB6⁺-HaCaT).

Immunoassays for p12, p15^INK4B, p21^CIP1, p27^KIP1, Total pRB, Phosphorylated pRB, CDK2, and CDK6.
The immunoassays were performed as described (32 , 33) . Briefly, cells were washed twice with ice-cold PBS and lysed in 1x lysis buffer containing 20 mM Tris (pH 7.4), 150 mM NaCl, 1 mM EGTA, 1 mM EDTA, 1% Triton X-100, 2.5 mM sodium PP_I, 1 mM ß-glycerophosphate, 1 mM sodium orthovanadate, 1 mM phenylmethylsulfonyl fluoride, 1 µg/ml leupeptin, 1 µM lactacystin ß-lactone, and 10 µg/ml aprotinin. The cells were frozen at -80°C for 15 min and then thawed at 4°C with gentle shaking. The lysed cells were sonicated for 15 s and then centrifuged at 10,000 x g for 15 min. The protein concentration of the resultant supernatant was determined using the Bio-Rad DC protein assay system with BSA as standard. Equal amounts of protein (30–50 µg) were subjected to SDS-PAGE (7.5–15%) and transferring onto polyvinylidene difluoride membranes in 3-(cyclohexylamino)propanesulfonic acid buffer [10 mM 3-(cyclohexylamino)propanesulfonic acid and10% methanol (pH 11.0)]. The membranes were blocked with 5% nonfat milk in Tris-buffered saline solution containing 0.05% Tween 20 and then incubated with antibodies specific for phospho-RB (NEB) against serine 795; total pRB (BD PharMingen); CDK2 (BD Transduction Laboratories); CDK6 (Santa Cruz Biotechnology); actin (Sigma); p12 (pAb86 rabbit polyclonal antibody made by Strategic Biosolutions (Newark, DE); p27^KIP1 (Transduction Laboratories); p15^INK4B (Santa Cruz Biotechnology) for human and mouse; and p21^CIP1 (BD PharMingen), respectively, according to suppliers’ protocols. The membranes were treated with appropriate secondary antibodies conjugated with horseradish peroxidase and visualized with ECL (Amersham Pharmacia).

Coimmunoprecipitation and Western Blot Analysis.
Coimmunoprecipitations were performed as described (32) . For analysis of direct protein-protein interaction between p12 and CDK2, cell lysates were obtained as described above, and equal amounts of protein from each treatment were immunoprecipitated with 1 µg of antibodies against CDK2, CDK6, or normal mouse IgG by incubating for 1 h at 4°C followed by incubating with protein A-Sepharose for 16 h at 4°C. The beads were centrifuged and repetitively washed. After the final wash, the beads were resuspended in 35 µl of 2x SDS-PAGE sample buffer and boiled for 5 min, and the supernatants were separated by 15% SDS-PAGE, transferred to polyvinylidene difluoride membranes, and blotted with antibodies against CDK2, p27^KIP1, and p12.

RNA Isolation and Northern Blot Analysis.
These procedures were essentially the same as described (32 , 33) . A cDNA probe for 18S rRNA (18S rRNA) was used to normalize loading. The cDNA probes for human and mink p12 were from the corresponding p12 coding region obtained by reverse transcription-PCR with forward and reverse primers of 5'-TTC CAC CAG CAT GGC AAC GTC TTC-3' and 5'-GGC ATT CCG TTC CGT TTC TG-3', respectively. The cDNA probe for mouse was obtained by cutting pCDNA3-mouse-p12 with EcoR1. The DNA fragment was isolated by gel purification kit (Qiagen).

Kinase Assays.
Anti-CDK2, anti-CDK6, and normal mouse IgG immunoprecipitates were washed three times with 2x lysis buffer followed by 2x kinase buffer containing 25 mM Tris (pH 7.5), 5 mM ß-glycerolphosphate, 2 mM DTT, 0.1 mM Na₃VO₄, and 10 mM MgCl₂. The pellet was suspended in 30 µl kinase buffer plus 5 µCi of [-³²P]ATP (NEN; 3000 Ci/mmol), 50 µM ATP, and 1 µg of pRB-c fusion protein (NEB; RB protein COOH-terminus residues 702–928 fused maltose binding protein). The reaction mixture was incubated for 30 min at 30°C, stopped by addition of 10 µl of 4x SDS-PAGE sample buffer, and then subjected to SDS/PAGE. The gels were cut into two pieces M_r 50,000. The top portion was dried and the M_r 68,000 pRB-C was visualized by autoradiography. The bottom portion was used for detecting IgG light chain, CDK2, and CDK6 proteins by Western blotting analyses.

[³H]Thymidine Incorporation and Cell Number Determination.
[³H]thymidine incorporation was performed as described (33) . Cells were seeded in 12-well plates at 6 x 10⁴ cells/well and incubated with [³H]thymidine (0.5–1 µCi/ml) under various conditions. For analysis of cell number, ip12(-)HaCaT cells were seeded in 12-well plates at 3 x 10⁴ cells/well. The cell numbers were determined by a Coulter Counter (Beckman). All of the experiments were done in triplicate and repeated at least three times with different cell preparations.

Analysis of Cell Cycle Distribution.
For fluorescence-activated cell sorting experiments, keratinocytes from MOK^p12-/- and MOK^WT were harvested in PBS after treatment with various concentrations of TGF-ß1 for 24 h. Cells were then fixed with 50% ethanol and stained with propidium iodide for DNA content. Cell cycle distribution was analyzed by flow cytometry using a Coulter cytometer and the Multicycle DNA analysis program.

Cloning of Human p12 Promoter.
Human p12 promoter region was cloned from human bacterial artificial chromosome genomic DNA library by screening with a 5'-untranslated region probe through the library screening service of Genome Systems Inc. The isolated 23kb EcoRI fragment was additionally digested with HindIII and shotgun subcloned. After screening again with a 5'-untranslated region probe, the clone of 6-kb HindIII fragment containing human p12 promoter region was identified, and the 6-kb fragment was subcloned into pGL3 basic vector (Promega).

Reporter Constructs.
A 800-bp PCR fragment containing 170-bp minimal human p12 promoter was cloned into pCR2.1 vector, and its HindIII-EcoR I fragment was subcloned into pBlueScript SKII(-) vector. The 1.8kb XhoI-HindIII fragment was retrieved from the 6.0-kb human p12 promoter clone and ligated with 0.8-kb fragment to generate 2.6-kb clone. After digesting with KpnI and SacI, the 2.6-kb was subcloned into pGL3 basic vector to generate the -2031 reporter construct. The -170-bp reporter construct was made by subcloning blunt-ended 800-bp BamH I fragment into SmaI site of pGL3 basic vector. The -1697, -1173, -770 constructs were made by digesting the -2031 construct with KpnI and BsmB I, KpnI and BstE II, or KpnI and PvuII, respectively. The -170-bp reporter construct was made by subcloning the blunt-ended 800-bp BamH I fragment into the SmaI site of the pGL3 basic vector containing the luciferase reporter gene to create p12/p170-luc. The plasminogen activator inhibitor type-1 promoter construct, PAI-1/p800-luc, was a gift from Dr. Magdalena Koziczak (Friedrich Miescher Institute, Basel, Switzerland; Ref. 34 ). HeLa and HaCaT cells were transiently transfected with these luciferase reporter constructs and pRSV-Gal (Promega) using Lipofectamine PLUS Reagent (Invitrogen) as specified by the manufacturer. Luciferase activities were measured using the Promega luciferase assay kit and a Berthold luminometer (Perkin-Elmer). Galactosidase activities were measured with Labsystems Multiscan MCC/340 plate reader at 405 nm. Data shown are normalized luciferase activity against the galactosidase activities in the same cell lysate.

Electrophoretic Mobility Shift Assay.
Electrophoretic mobility shift analysis was carried out as described (32 , 33) . Nuclear extracts were prepared by the method of Schreiber et al. (35) from monolayer cultures in 100-mm Petri dishes, and 5 µg protein was used for electrophoretic mobility shift analysis. The double-stranded oligonucleotides with the consensus binding motif for Smads (CAGAC) were end-labeled with [-³²P]ATP and T4 polynucleotide kinase, and purified by P-25 Biogel columns. Nuclear extracts were added to ³²P-labeled Smad oligonucleotide (3–4 x 10⁴ cpm/reaction) in a binding buffer containing 3 µg of poly(deoxyinosinic-deoxycytidylic acid) (Pharmacia), 20 mM HEPES (pH 7.8), 10% glycerol, 1 mM EDTA, 5 mM MgCl₂, 1 mM phenylmethylsulfonyl fluoride, 1 mM DTT, and 100 mM NaCl. The reaction mixtures were incubated for 30 min at room temperature, and the DNA-protein complexes were resolved on a 4% nondenaturing polyacrylamide gel. In competition experiments, unlabeled oligonucleotide was added to the nuclear extracts and incubated for 15 min before addition of the radiolabeled probe. Antibodies (1 µg of Smad 2, 3, and 4, and p53/reaction) were added after the radiolabeled oligonucleotide had reacted with the nuclear extracts for 30 min and then incubated for an additional 60 min at room temperature. The sequence of double-stranded oligonucleotides used as a probe in Fig. 3B was 5'-TCG AGA GCC AGA CAA AAA GCC AGA CAT TTA GCC AGA CAC-3' and its complementary strand (Santa Cruz Biotechnology). The mutant oligonucleotide was identical to the sequence described above with the exception of a "C" "T" and "G" "C" substitutions in the CAGAC binding motif. The sequence of double-stranded oligonucleotide p1 was 5'-GTT AAA AAC CAC TTC ATA CAG ACA GAG ACA TGA AAT TTA-3', and its complementary strand, which includes the -165/-125 of the p12 promoter bearing the -147/-143 CAGAC box. The sequence of p1 mutant was identical to the sequence described above with the exception of a "C" "T" and "G" "C" substitutions in the CAGAC binding motif. The sequence of double-stranded oligonucleotide P2 used as a probe in Fig. 3B was 5'-GGG GCT TGG CCA GAC CTG GGC GTC TGG AAT-3', and its complementary strand, which includes the -96/-60 of the p12 promoter bearing the -79/-75 and -68/-64 CAGAC boxes. The sequence of p2 mutant was identical to the sequence described above with the exception of a "C" "T" and "G" "C" substitutions in the CAGAC binding motif. Nuclear factor B gel shift oligonucleotide was from Santa Cruz Biotechnology. The sequence was 5'-AGT TGA GGG GAC TTT CCC AGG C-3'. Anti-Smad antibodies used for supershift (Fig. 3D) were from Santa Cruz Biotechnology. The antibody recognizing p53 was from Calbiochem-Oncogene Research Products.

View larger version (45K): [in this window] [in a new window]   Fig. 3. Characterization of the activation of Smads by transforming growth factor (TGF)-ß1. A, schematic diagram of the p12-p170 promoter-luciferase report constructs showing the three CAGAC Smad protein binding motifs. B and C, electrophoretic mobility shift analysis demonstrating interaction of double-stranded oligonucleotide probes with nuclear protein extracts from untreated (control) or TGF-ß1-treated (TGF-ß1) HaCaT cells (5 ng/ml TGF-ß1 for 30 min). The probes used in B are commercially available wild-type or mutant oligos containing CAGAC or TACAT motifs as described in "Materials and Methods." The probes used in C are probes 1 and 2 spanning the p12 promoter as shown in A and described in "Materials and Methods." Arrowheads denote TGF-ß1-enhanced complexes. D, an EMSA was performed using 32P-labeled probe p1 containing CAGAC sequence and nuclear extract from HaCaT cells uninduced or induced by TGF-ß1 for 30 min. Bands corresponding to specific TGF-ß1-induced complexes are indicated. Ten or 100 molar excesses of various nonradiolabelled oligonucleotides were added as competitors, including the wild-type and mutant "P1" sequence, as well as an unrelated nuclear factor B sequence. E, specific anti-Smad antibodies and unrelated p53 antibody were added to the mixtures 30 min after incubating TGF-ß1-induced HaCaT nuclear extracts with radiolabeled "P1" probe. The supershifted complexes are indicated.

Data Presentation.
All of the experiments were performed at least three times with different preparations of cells, and the data shown are from a representative experiment. Densitometric analyses of the blots were performed using a PDI scanner (model 420 oe), and the results are reported as fold increase over respective controls or actual fold increase.

	RESULTS

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
p12 Expression Is Induced in Epithelial Cells by TGF-ß1.
The effect of TGF-ß1 on p12 expression was determined by examining the steady-state mRNA and protein levels in TGF-ß1-treated rodent and human diploid epithelial cells. Fig. 1, A and B , show the representative data for p12 mRNA and protein in Mv1Lu, human primary foreskin keratinocyte, and HaCaT cells treated with and without TGF-ß1 (5 ng/ml) over a 24-h period. In all three of the epithelial cell types, TGF-ß1 treatment increased p12 mRNA and protein levels. Mv1Lu and human primary foreskin keratinocyte cells showed strong responses (4–5 fold), whereas the response in HaCaT cells is moderate (3–4-fold). Although the extent of increase varies with cell types, in general, TGF-ß1 induces a 3–5-fold increase in p12 mRNA and protein levels. Thus, TGF-ß1 treatment increased the cellular levels of p12 in diploid epithelial cells.

View larger version (42K): [in this window] [in a new window]   Fig. 1. Induction of p12 mRNA and protein by transforming growth factor (TGF)-ß1. Exponentially growing cells (Mv1Lu, human primary foreskin keratinocyte, and HaCaT) were maintained in growth medium in the presence or absence of TGF-ß1 for 24 h. p12 mRNA (A) and protein levels (B) were examined. Fold change relative to control (Fold ) were determined from the densitometric data after being normalized to 18S for p12 mRNA and to actin for p12 protein. The control value (no TGF-ß1) is arbitrarily set at 1.0. C and D, exponentially growing HaCaT cells were treated with or without TGF-ß1 for the times indicated. C, total RNA was isolated, and Northern blot analysis was performed with the use of cDNA probes for either p12 or 18S rRNA. D, the protein level was analyzed by Western blot analysis with the anti-p12 polyclonal antibody pAb86. Actin is used as loading control. E, HaCaT cell lysates were harvested and immunoblotted with the antibodies to p12, p15, p27, p21, and actin. Fold change relative to the control (Fold ) were determined from the densitometric data after being normalized to actin for p12 protein. The control value (no TGF-ß1) is arbitrarily set at 1.0. F, TGF-ß1 increases p12 and p27 associated with cyclin-dependent kinase (CDK) 2. Cell lysates from the exponentially growing HaCaT cells in the presence and absence of TGF-ß1 were prepared, and aliquots (600 µg each) were immunoprecipitated with 1 µg of anti-CDK2 antibody. Western blot analysis was performed on the resuspended immunoprecipitates by using CDK2 (top panel), p27 (bottom right panel), and p12 antibodies (bottom left panel). Light chain stands for light chain of IgG.

TGF-ß1-mediated growth suppression is thought to be dependent on the ability of TGF-ß1 to increase the synthesis of p15 or p21 in different cell types (14) . We confirmed that these CKIs were increased in our HaCaT working model. Fig. 1E shows that TGF-ß1 induced a striking up-regulation of p15^INK4B (8-fold), and moderate increase of p21^CIP1 (3-fold) and p12 (5-fold), whereas it had no effect on the level of p27 ^KIP1. These results are consistent with previous reports (15 , 36 , 37) . Thus, in HaCaT cells, in addition to p15^INK4B and p21^CIP1, p12 is also induced by TGF-ß1. These results suggested that p12 may also be involved in TGF-ß1-mediated growth suppression.

We have shown previously that p12 can interact with CDK2 in p12 overexpressing cells and in vitro binding assays (7) . To determine whether the increased p12 expression induced by TGF-ß1 led to an increase in its association with CDK2, the levels of cellular CDK2-associated p12 were determined by sequential immunoprecipitation/immunoblot analysis. Fig. 1F showed that the association of CDK2 with p12 and with p27^KIP1 increased after a 24-h treatment with TGF-ß1. It should be noted that in HaCaT cells, although TGF-ß1 did not increase the total p27^KIP1 levels (Fig. 1E) , it increased the level of CDK2-bound p27^KIP1 by approximately 5–7-fold (Fig. 1F , right). This is consistent with previous reports that TGF-ß1 causes a redistribution of p27 from CDK4/CDK6 complexes to CDK2 complexes via induction of p15 (15 , 36 , 37) . Taken together, the data indicate that TGF-ß1 increases cellular levels of p12 and its interaction with CDK2, concomitant with known alterations of other CKI proteins.

TGF-ß1 Enhances the p12 Promoter Activity.
To determine whether the up-regulation of p12 by TGF-ß1 is mediated at the level of gene transcription, we examined p12 mRNA levels in response to TGF-ß1 in the presence or absence of actinomycin D (5 µM), an inhibitor of transcription. TGF-ß1-mediated up-regulation of p12 mRNA was completely blocked by pretreatment of cells with actinomycin D (data not shown), indicating that TGF-ß1 activates the transcription of p12 gene. We also determined that the half-life of p12 is not changed by TGF-ß1 (data not shown). Those results suggested that TGF-ß1 treatment did not alter the stability of pre-existing mRNA of p12, but rather increased p12 mRNA level by de novo mRNA synthesis.

To examine the mechanisms by which TGF-ß1 increases cellular levels of p12, a 23-kb human p12 promoter fragment was cloned from a human bacterial artificial chromosome genomic DNA library by screening with a p12 5'-untranslated region probe. The isolated 23-kb EcoRI fragment was additionally digested with HindIII and shotgun subcloned. After screening again with the p12 5'-untranslated region probe, a 6-kb HindIII fragment containing human p12 promoter sequences was identified and subcloned into pGL3 basic vector. To define the basal and TGF-ß1-inducible elements in the p12 promoter, a series of 5' deletion constructs were created spanning the 6-kb p12 promoter fragment using convenient restriction sites (Fig. 2A) . Transfection of these constructs into the highly transfectable HeLa cells revealed that all of the constructs mediated both basal and TGF-ß1-inducible promoter activities (Fig. 2B) . Of note is that deletion of most but the most proximal 170 bp of the p12 promoter (-170p12) resulted in maximal TGF-ß1 induction (approximately four times) of the p12 promoter. The -2031, -1697, and -770 deletion constructs were less responsive to TGF-ß1, ranging from 1- to 2-fold. Thus, -170p12 promoter fragment contained the inducible elements necessary for transcriptional activation by TGF-ß1.

View larger version (27K): [in this window] [in a new window]   Fig. 2. Transforming growth factor (TGF)-ß1 enhances p12 promoter activity. A, schematic diagram of deletion mutants of p12 promoter-luciferase report constructs used in transient transfection assays. Sequential deletions from the 5' end of p12 promoter are shown. B, HeLa cells were transiently transfected with wild-type or deletion constructs as indicated. In each experiment equal amounts of total DNA were transfected. Cells were treated with or without TGF-ß1 for 48 h before lysis. Luciferase activity was normalized to ß-galactosidase activity and expressed as mean of triplicate from a representative experiment; bars, ± SD. These experiments were performed three times in triplicate, and similar results were obtained in each repeat. C and D, pGL3 basic vector (Basic V), p12-p170-luc, or PAI-1–800-Luc were transfected into HaCaT cells. TGF-ß1 treatment and luciferase assays were performed as described above (B). Data shown are normalized luciferase activity against the galactosidase activities in each sample.

To examine the involvement of Smad proteins in TGF-ß1-mediated transcriptional activation of the p12 promoter constructs, we screened for the presence of SBEs (CAGAC) and found three such motifs in the -170p12 minimal promoter region (28) , at positions -68/-64, -79/-75, and -147/-143 relative to the mRNA initiation sites, respectively (Fig. 3A) . To determine whether Smad proteins are activated by TGF-ß1 treatment and are involved in the transcriptional activation of the p12 promoter, electrophoretic mobility shift analyses were performed by using the nuclear extracts from HaCaT cells and ³²P-labeled probes containing the SBE motifs from the minimal -170p12 promoter regions designated as P1 (containing the -147/-143 SBE) and P2 (containing the -68/-64 and -79/-75 SBEs; Fig. 3A ). P2 contains the 2 Smad sites because they are too close to be separated, and the oligo needs to be at least 25–30 nucleotides to be effective. Commercially available Smad-binding oligonucleotides (sc-2597), which contain three copies of CAGAC, and mutant oligos (sc-2598) identical to sc-2597 with the exception of a "C" "T" and "G" "C" substitutions in the CAGAC binding motif, were used as positive and negative controls. TGF-ß1 caused a 3- to 5-fold increase in the band designated as the TGF-ß1-enhanced complexes when the probes of P1, P2, or commercial available positive control Smad-binding oligonucleotides were used (Fig. 3, B and C) . No slower migrating complexes were formed when the mutant probes (sc-2598; Fig. 3B , left panel) was used. Maximum binding requires TGF-ß1 treatment for 30 min, but the complex can be clearly observed after 15-min treatment with TGF-ß1 (data not shown). The DNA-binding complexes are specific. Addition of 100-fold excess of unlabeled P1 oligonucleotide completely eliminated the slower migrating complex (Fig. 3D , Lane 4), whereas an unrelated oligonucleotide with sequence corresponding to that of nuclear factor B (Fig. 3D , Lane 5) binding sites or the mutant version of P1 (Fig. 3D , Lane 6, with "C" "T" and "G"; "C" substitutions in the CAGAC binding motif) had little effect. These results indicated that TGF-ß1 induced binding of Smads to the p12 promoter.

To additionally confirm that Smad proteins were present in the CAGAC binding complex, nuclear extracts were incubated with specific antibodies to Smad 2, 3, and 4. Supershift of the TGF-ß1-dependent binding complex was observed with anti-Smad 3 and anti-Smad 4 antibodies (Fig. 3E , Lanes 5 and 6). The complex was not totally supershifted with either antibody probably because the complex contains not only one SBE but also other cofactors. In contrast, addition of an antibody specific for Smad 2 to the mixture actually reduced the formation of the complex activated by TGF-ß1 (Fig. 3E , Lane 3), whereas addition of antibodies to another transcription factor p53 (Fig. 3E , Lane 4) had no effect. In separate experiments, antibodies to other transcription factors were also ineffective in altering the complex formation, although we were unable to effectively demonstrate a supershift with the use of several commercially available Smad 2 antibodies. Nevertheless, these data suggested that TGF-ß1 induced Smad proteins to bind at the proximal promoter region of the p12 gene.

CDK2 Kinase Activity and Cell Proliferation Are Up-Regulated in p12 Antisense Cell Lines.
To examine the functional and physiological relevance of the TGF-ß1-mediated p12 induction, inducible antisense p12 HaCaT cell lines [ip12(-)HaCaT] were created to determine the role of p12 in TGF-ß1-mediated phenotypes. Exon 1 of the human p12 in antisense orientation was placed under the control of the Zn-inducible metallothionein promoter (pMT/CB⁶⁺). Stable ip12(-)HaCaT cell lines with robust antisense p12 inducibility and control HaCaT cell lines with the parental vector (pMT/CB⁶⁺) were generated. We have similarly created tetracycline-inducible p12 transgene constructs but have found them to be not effective in the transgene induction. Fig. 4A shows that ZnSO₄ treatment of the ip12(-)HaCaT cells for 48 h reduced cellular levels of endogenous p12 mRNA and protein by 3.3- and 4-fold, respectively. No alteration was observed for p15, p21, and p27 (Fig. 4A) . Similar results were obtained from the two other independent p12 antisense cell lines. ZnSO₄ treatment of the control vector cell lines showed no appreciable effect of p12 expression (Fig. 4A) .

View larger version (36K): [in this window] [in a new window]   Fig. 4. Cyclin-dependent kinase (CDK) 2 kinase activity and thymidine incorporation in ip12(-)HaCaT cell lines. A, selective and specific down-regulation of p12 mRNA and protein in the ip12(-)HaCaT cell lines. The ip12(-)HaCaT and pMTCB6+-HaCaT cell lines were treated with or without 100 µM ZnSO4 for 48 h. Northern blot analysis for p12, and Western blot analyses for p12, p15, p21, and p27 were performed as described in "Materials and Methods." B, CDK2 immunoprecipitation of cell lysates (500 µg each) prepared from induced or uninduced antisense cell lines were immunoblotted with anti-CDK2 (top panel) and anti-p12 antibodies (bottom panel). C, in vitro kinase assay. The CDK2 and CDK6 kinase activities in the inducible p12 antisense HaCaT cell lines were determined by using pRBc as substrates. The top panel showed the autoradiographic results from pRBc kinase activities of immunoprecipitated normal mouse IgG, CDK2, and CDK6 complexes. The bottom panel showed immunoblots of light chain of normal mouse IgG, CDK2, and CDK6 levels from same solubilized normal mouse IgG, CDK2, and CDK6 immunoprecipitates used for kinase assays. D, thymidine incorporation in ip12(-)HaCaT cell lines. ip12(-)HaCaT cell lines were treated with or without 100 µM ZnSO4 for 48 h. The results are expressed as mean; bars, ±SD (n = 3).

We next examined CDK2 kinase activity using the pRB COOH-terminal 702–928 (pRBc) peptide as the substrate. In all three of the ip12(-)HaCaT cell lines, the CDK2-associated pRBc phosphorylation levels were elevated 7-fold after ZnSO₄ treatment (Fig. 4C , top middle panel). The increase in pRB phosphorylation is not due to the increase in CDK2 protein level, confirmed by anti-CDK2 immunoblot (Fig. 4C , bottom middle panel). The effect of ZnSO₄ treatment is specific for CDK2 because immunoprecipitation for CDK6 and subsequent kinase assay for pRBc demonstrated no difference in pRBc phosphorylation with or without ZnSO₄ treatment (Fig. 4C , right panels). When normal mouse IgG was used for immunoprecipitation, pRBc phosphorylation was hardly detectable (Fig. 4C , left panels). The pRBc phosphorylation levels were not altered when similar experiments were done on vector control cell lines (data not shown). These data demonstrate that the ZnSO₄ treatment of the ip12(-)HaCaT cells specifically modified CDK2-associated pRB kinase activity.

The effect of selective reduction of intracellular p12 by ZnSO₄ on cellular proliferation in the ip12(-)HaCaT cells was examined by [³H]thymidine incorporation assay. ZnSO₄ treatment resulted in a significant increase in [³H]thymidine incorporation in ip12(-)HaCaT cells (38% increase; P < 0.01; Fig. 4D ). Similar results were seen in the other two cell lines (data not shown). ZnSO₄ treatment of the five control vectors HaCaT cell lines did not result in significant difference in [³H]thymidine incorporation (data not shown). Together, these data allow us to conclude that p12 is an endogenous negative regulator of CDK2-associated kinase activities and is a physiological growth suppressor.

p12 Regulates CDK2-Mediated Phosphorylation of Endogenous pRB.
The above experiments demonstrated that endogenous p12 is a negative regulator of CDK2. Because it is known that CDK2/cyclin E contributes to the phosphorylation of pRB at the G₁-S transition (16 , 17) , we examined the effect of endogenous p12 levels on the phosphorylation status of endogenous pRB. Cellular p12 levels were measured in the ip12(-)HaCaT cells in the presence or absence of TGF-ß1 Fig. 5, A and B , show that, similar to the parental HaCaT cells, p12 is TGF-ß1 responsive in the ip12(-)HaCaT cell lines both in the absence and presence of ZnSO₄. ZnSO₄ treatment reduced the steady-state levels of p12 mRNA (Fig. 5A) and protein (Fig. 5B) by 40% in the absence (Fig. 5 B, Lane 3 versus Lane 1) and presence (Fig. 5 B, Lane 4 versus Lane 2) of TGF-ß1. These results indicate that both the basal level and TGF-ß1-induced expression of p12 are inhibited to the same degree by the p12 antisense transgene.

View larger version (47K): [in this window] [in a new window]   Fig. 5. Reduction of p12 partially reverses transforming growth factor (TGF)-ß1-mediated inhibition of cyclin-dependent kinase (CDK) 2 kinase activity, pRB phosphorylation, and cell proliferation. A, Northern blot analysis of p12 mRNA and 18S rRNA levels in inducible p12 antisense cell lines in the absence or presence of TGF-ß1. The bar graph in the lower portion summarizes results from three separate experiments using the densitometric analysis of the ratio from the p12 and the 18S rRNA. The values are expressed as mean; bars, ±SD (*,P < 0.05, Lane 2 versus Lane 1; #, P < 0.01, Lane 3 versus Lane 1; and , P < 0.05, Lane 4 versus Lane 3). B, Western blot analysis of p12 protein and actin in ip12(-)HaCaT cell lines in the absence or presence of TGF-ß1. C, Western blot analysis. Top panel, the phosphorylation level of pRB (and p130). Middle panel, the total pRB level. Bottom panel, actin level. D, in vitro kinase assay. The top panel showed the autoradiographic results from pRBc kinase activities of immunoprecipitated CDK2 complexes. The bottom panel showed immunoblots of the CDK2 levels from same solubilized CDK2 immunoprecipitates used for kinase assays. E, cell growth inhibition by TGF-ß1 in the presence or absence of ZnSO4. ip12(-) HaCaT cells (3 x 104) were seeded in 12-well plates and treated as follows, (1) TGF-ß1 (T); (2) ZnSO4 (Zn); (3) ZnSO4 + TGF-ß1 (Zn +T); and (4) control. Cell numbers were determined by a Coulter counter (Beckman) after 24, 48, 72, and 96 h of respective treatments. Data are expressed as mean of triplicate determinations and are plotted as % inhibition of cell numbers relative to corresponding controls; bars,±SD. *, P < 0.05 (T versus Zn+T at 48 and 72 h); *, P < 0.01, (T versus Zn+T at 96 h).

To determine whether the observed elevation of pRB phosphorylation was exerted by cellular CDK2 kinase, the CDK2 complexes were immunoprecipitated and used to phosphorylate the pRBc substrate. This confirmation is particularly important because Serine 795 on pRB can be phosphorylated by all three of the CDKs (CDK2, CDK4, and CDK6; Refs. 38, 39, 40 ). Fig. 5D shows that CDK2-mediated phosphorylation of the pRBc is reduced 3-fold by TGF-ß1 treatment (Fig. 5 D, Lanes 1 and 2). In the presence of ZnSO₄, CDK2-mediated phosphorylation of pRBc is increased 8-fold (Fig. 5 D, Lane 4). However, in the presence of both TGF-ß1 and ZnSO₄, phosphorylation level of pRBc is intermediate between that of ZnSO₄ (Fig. 5 D, Lane 4) and TGF-ß1 (Fig. 5 D, Lane 2) treatment alone. Thus, the kinase activity of CDK2 paralleled the phosphorylation status of endogenous pRB both in the presence and in the absence of TGF-ß1.

To additionally correlate pRB with the effect of the suppression of p12 expression on TGF-ß1-mediated growth inhibition in this system, we examined the effect of TGF-ß1 on the proliferation of the ip12(-) HaCaT cell line with different combinations of treatment for the time periods (Fig. 5E) . In the presence of TGF-ß1, inhibition of proliferation occurred 48 h after TGF-ß1 treatment as determined by measuring accumulated cell number. The percentage inhibition of cell growth were 21% ± 4%, 39% ± 3%, and 44% ± 6%, respectively, after 48, 72, and 96-h incubation with TGF-ß1relative to control. However, in the presence of TGF-ß1 and ZnSO₄ for the same time intervals, the inhibition of cell numbers were 15% ± 6%, 27% ± 6% and 25% ± 3%, respectively, relative to control. The inhibition exerted by TGF-ß1 treatment was significantly attenuated when the cells were treated with ZnSO₄, suggesting that the suppression of p12 expression can partially counteract TGF-ß1-mediated inhibition of cellular proliferation in this inducible p12 antisense HaCaT cells. These data suggest that the reduction of p12 expression in the ip12(-) HaCaT cells increased CDK2-mediated phosphorylation of pRB, and counteracted TGF-ß1-mediated suppression of pRB phosphorylation and cell proliferation. Taken together, these data indicate that endogenous p12 modulates TGF-ß1-mediated inhibition of pRB phosphorylation and cellular proliferation by negatively regulating CDK2 kinase activity.

TGF-ß1-Mediated Growth Suppression Is Attenuated in p12-Deficient Keratinocytes.
To additionally investigate whether TGF-ß1 induction of endogenous p12 is involved in TGF-ß1-induced growth suppression, mouse oral keratinocytes from p12 gene knockout and wild-mice were generated. p12^-/- mice were created by targeted gene disruption of exon 3 in BALB/c mice.⁴ Mouse oral keratinocytes (MOK) from wild (MOK^WT) and p12^-/- (MOK^p12-/-) mice were prepared and subjected to treatment of increasing concentrations of TGF-ß1 for 24 h. Cell cycle distributions were determined by fluorescence-activated cell sorting analysis (Fig. 6A) . The percentage of cells in S phase was increased 2-fold in MOK^p12-/- cells (17.4%) as compared with MOK^WT cells (8.8%). In the presence of 0.01 ng/ml of TGF-ß1, the percentages of cells in S phase were 15.8 and 5%, respectively, in MOK^p12-/- and MOK^WT cells, representing a 9% and 43% decrease from that in the absence of TGF-ß1 When the concentration of TGF-ß1 was increased to 0.1 ng/ml, the percentages of cells in S phase were 6.4% and 1.9% in MOK^p12-/- and MOK^WT cells, respectively, a 63% and 78% decrease from the corresponding controls. Fig. 6B summarizes results for three separate experiments, and the data are expressed as percentage of inhibition of cells in S phase relative to corresponding control that did not receive TGF-ß1 These data indicated that TGF-ß1-mediated growth suppression was clearly attenuated in the MOK^p12-/-, especially at low concentration of TGF-ß1. In addition to the changes in S phase, fluorescence-activated cell sorting analysis showed that TGF-ß1 at 0.01 ng/ml did not change G₁ phase distribution (from 63.2 to 63.7%) in MOK^p12-/- but increased slightly (4.8%) in MOK^WT (from 55.9% to 58.6%). However, in the presence of 0.1 ng/ml of TGF-ß1, there was a significant increase (20%) in the percentage of cells in G₁ phase in MOK^p12-/- (from 63.2% to 75.8%). Importantly, the increase of cells in G₁ phase induced by 0.1 ng/ml of TGF-ß1 in MOK^WT was not as great (from 55.9% to 60.8%, representing a 8.7% increase). Furthermore there was also a 2-fold decrease in G₂ phase in MOK^p12-/- compared with the MOK^WT, suggesting that p12 may play a specific role in G₂ phase in response to TGF-ß1.

View larger version (27K): [in this window] [in a new window]   Fig. 6. Cell cycle profiles of the keratinocytes from wild-type and p12-/- mice in the absence or presence of transforming growth factor (TGF)-ß1. A, cell cycle distribution of keratinocytes; exponentially growing keratinocytes from wild-type and p12-/- mice were incubated with increasing concentration of TGF-ß1 in keratinocyte serum-free medium for 24 h. The cell cycle distribution was analyzed by flow cytometry. The data shown are from a representative experiment. B, the plot showed the percentage of cells inhibited in S phase. Values shown are mean from three independent experiments and plotted as percentage of inhibition of cells in S phase relative to different concentration of TGF-ß1; bars, ±SD. Significance of difference from the corresponding controls indicated by asterisks (*, P < 0.05; **, P < 0.01).

View larger version (37K): [in this window] [in a new window]   Fig. 7. Transforming growth factor (TGF)-ß1-mediated changes in cyclin-dependent kinase (CDK) inhibitors and CDK2 kinase activity in MOKWT and MOKp12-/-. A, the MOKWT and MOKp12-/- were treated with increasing concentration of TGF-ß1 for 24 h, followed by analysis of p12 mRNA by Northern blot and respective protein levels by Western blot with the indicated antibodies. B, In vitro kinase assay. The MOKWT and MOKp12-/- were treated with increasing concentration of TGF-ß1 for 24 h, followed by kinase assays using pRB-C as substrate.

	DISCUSSION

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
In the present study we explored the role that p12 plays in TGF-ß1-mediated growth suppression. We show that TGF-ß1 transcriptionally induces the expression of p12 that, in turn, inhibits CDK2-catalyzed phosphorylation of pRB. Reduction of p12 expression results in a partial reversal of the ability of TGF-ß1 to decrease CDK2-mediated pRB phosphorylation and cell proliferation in ip12(-)HaCaT cell lines. The deficiency in p12 expression confers partial resistance to TGF-ß1-mediated inhibition of cell proliferation in mouse keratinocytes, because the other pathways, such as up-regulation of p15^INK4B and p21^CIP1 or down-regulation of c-myc (11, 12, 13) by TGF-ß1, are not compromised. Our results suggest that p12 is an effector of TGF-ß1-mediated growth suppression. Indeed, it has been reported that deletion of known effectors of TGF-ß1, p15^INK4B and p27^KIP1, does not totally abrogate growth arrest (18) . p12 may, thus, act independently of these known inhibitors to effect TGF-ß1-induced growth arrest.

Other mechanisms involved in TGF-ß1-mediated growth arrest have been reported, but it is not clear that these are primary responses to TGF-ß1 signaling through Smads or if they are secondary to initial growth arrest. For instance, it is known that TGF-ß1-induced cell arrest is accompanied by repression of Cdc25A, a tyrosine phosphatase known to activate CDKs (in tumor cells that have lost the INK4b locus; Ref. 41 ). However, this is likely secondary in that E2F/p130 mediates this repression presumably after the initial drop in CDK activity (42) . More recently, it has been reported that TGF-ß1 mediates cell growth arrest of human HepG₂ hepatocellular carcinoma cells by inhibiting a CDK2-activating kinase (43) . However, it is not known how this inhibition is achieved nor if it is direct. In this case, accumulation of non-cyclin-dependent kinase-activating kinase phosphorylated CDK2 is observed. Thus, this could be influenced by p12, as we have found this protein to bind the inactive CDK2 (7) . This, in turn, could start the initial drop in CDK activity in p15^INK4B-defective human tumor cell lines (42) . In contrast, TGF-ß1 regulation of p12 is likely a direct effect of SMADs activation. We show that SMAD proteins are rapidly activated and specifically bind to a CAGAC motif-containing p12 promoter region as early as 15 min after TGF-ß1 treatment. p12 RNA accumulation was detected after 2–4 h of treatment with TGF-ß1, suggesting that SMADs activation occurs before increasing p12 expression.

Interestingly, p12^-/- cells are less sensitive to low amounts of TGF-ß1, but still respond to high amounts, suggesting some hierarchy of response. In addition, p12^-/- cells specifically fail to accumulate in G₂, suggesting that p27^KIP1 and p12-mediated CDK2 inhibition may target different cell cycle stages.

Fig. 8 depicts our current understanding of the role of p12 in the TGF-ß1 response based on the findings presented in this article. At G₁-S transition, the phosphorylation of pRB is accomplished consecutively by CDK4/6-cyclin D followed by CDK2-cyclin E (16 , 17) . Whereas the CDK4/6-cyclin D-mediated pRB phosphorylation is mitogen dependent, phosphorylation by CDK2-cyclin E is not (cyclin E dependent). TGF-ß1 exerts its biological effect by inducing a spectrum of CDK inhibitors, p15^INK4B, p21^CIP1, p12, and p27^KIP1. Therefore, TGF-ß1 can be viewed to execute its antiproliferative effects at G₁-S transition by suppressing pRB phosphorylation at both the CDK4/6-cyclin D and CDK2-cyclin E steps by inducing universal CKI (p21^CIP1), specific CKI for CDK4/6 (p15^INK4B) and specific CKI for CDK2 (p12). Thus, p12 may serve as a specific CDK2 inhibitor that provides cells an additional level of control before full commitment to the cell division cycle.

View larger version (24K): [in this window] [in a new window]   Fig. 8. Working model of p12 in the transforming growth factor ß1-mediated responses serving as a specific negative regulator of cyclin-dependent kinase 2 activities in pRB phosphorylation at G1-S transition.

	FOOTNOTES

 
Grant support: NIH Grants RO1 DE14857 (D. T. W. W.) and F32 DE05763 (M. G. H.).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Requests for reprints: David T. W. Wong, Laboratory of Head and Neck Cancer Research, Dental Research Institute, University of California Los Angeles, School of Dentistry, 10833 Le Conte Avenue, 73–017 CHS, Los Angeles, CA 90095. Phone: (310) 206-3048; Fax: (310) 825-0921; E-mail: dtww{at}ucla.edu

⁴ J. McBride et al., manuscript in preparation.

Received 7/25/03. Revised 10/ 8/03. Accepted 11/ 3/03.

	REFERENCES

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Todd R., McBride J., Tsuji T., Donoff R. B., Nagai M., Chou M. Y., Chiang T., Wong D. T. Deleted in oral cancer-1 (doc-1), a novel oral tumor suppressor gene. FASEB J., 9: 1362-1370, 1995.[Abstract]
2. Tsuji T., Duh F. M., Latif F., Popescu N. C., Zimonjic D. B., McBride J., Matsuo K., Ohyama H., Todd R., Nagata E., Terakado N., Sasaki A., Matsumura T., Lerman M. I., Wong D. T. W. Cloning, mapping, expression, function, and mutation analyses of the human ortholog of the hamster putative tumor suppressor gene doc-1. J. Biol. Chem., 273: 6704-6709, 1998.[Abstract/Free Full Text]
3. Gordon H. M., Kucera G., Salvo R., Boss J. M. Tumor necrosis factor induces genes involved in inflammation, cellular and tissue repair, and metabolism in murine fibroblasts. J. Immunol., 148: 4021-4027, 1992.[Abstract]
4. Daigo Y., Suzuki K., Maruyama O., Miyoshi Y., Yasuda T., Kabuto T., Imaoka S., Fujiwara T., Takahashi E., Fujino M. A., Nakamura Y. Isolation, mapping and mutation analysis of a human cDNA homologous to the doc-1 gene of the Chinese hamster, a candidate tumor suppressor for oral cancer. Genes Chromosomes Cancer, 20: 204-207, 1997.[CrossRef][Medline]
5. Matsuo K., Shintani S., Tsuji T., Nagata E., Lerman M., McBride J., Nakahara Y., Ohyama H., Todd R., Wong D. T. p12(DOC-1), a growth suppressor, associates with DNA polymerase /primase. FASEB J., 14: 1318-1324, 2000.[Abstract/Free Full Text]
6. Shintani S., Mihara M., Terakado N., Nakahara Y., Matsumura T., Kohno Y., Ohyama H., McBride J., Kent R., Todd R., Tsuji T., Wong D. T. Reduction of p12DOC-1 expression is a negative prognostic indicator in patients with surgically resected oral squamous cell carcinoma. Clin. Cancer Res., 7: 2776-2782, 2001.[Abstract/Free Full Text]
7. Shintani S., Ohyama H., Zhang X., McBride J., Matsuo K., Tsuji T., Hu M. G., Hu G., Kohno Y., Lerman M., Todd R., Wong D. T. p12(DOC-1) is a novel cyclin-dependent kinase 2-associated protein. Mol. Cell. Biol., 20: 6300-6307, 2000.[Abstract/Free Full Text]
8. Alexandrow M. G., Moses H. L. Transforming growth factor ß 1 inhibits mouse keratinocytes late in G₁ independent of effects on gene transcription. Cancer Res., 55: 3928-3932, 1995.[Abstract/Free Full Text]
9. Massague J. The transforming growth factor-ß family. Annu. Rev. Cell. Biol., 6: 597-641, 1990.[CrossRef][Medline]
10. Polyak K. Negative regulation of cell growth by TGF ß. Biochim. Biophys. Acta, 1242: 185-199, 1996.[Medline]
11. Alexandrow M. G., Kawabata M., Aakre M., Moses H. L. Overexpression of the c-Myc oncoprotein blocks the growth-inhibitory response but is required for the mitogenic effects of transforming growth factor ß 1. Proc. Natl. Acad. Sci. USA, 92: 3239-3243, 1995.[Abstract/Free Full Text]
12. Claassen G. F., Hann S. R. A role for transcriptional repression of p21CIP1 by c-Myc in overcoming transforming growth factor ß-induced cell-cycle arrest. Proc. Natl. Acad. Sci. USA, 97: 9498-9503, 2000.[Abstract/Free Full Text]
13. Warner B. J., Blain S. W., Seoane J., Massague J. Myc downregulation by transforming growth factor ß required for activation of the p15(Ink4b) G(1) arrest pathway. Mol. Cell. Biol., 19: 5913-5922, 1999.[Abstract/Free Full Text]
14. Hannon G. J., Beach D. p15INK4B is a potential effector of TGF-ß-induced cell cycle arrest. Nature (Lond.), 371: 257-261, 1994.[CrossRef][Medline]
15. Reynisdottir I., Polyak K., Iavarone A., Massague J. Kip/Cip and Ink4 Cdk inhibitors cooperate to induce cell cycle arrest in response to TGF-ß. Genes Dev., 9: 1831-1845, 1995.[Abstract/Free Full Text]
16. Sherr C. J., Roberts J. M. CDK inhibitors: positive and negative regulators of G₁-phase progression. Genes Dev., 13: 1501-1512, 1999.[Free Full Text]
17. Lundberg A. S., Weinberg R. A. Functional inactivation of the retinoblastoma protein requires sequential modification by at least two distinct cyclin-cdk complexes. Mol. Cell. Biol., 18: 753-761, 1998.[Abstract/Free Full Text]
18. Nakayama K., Ishida N., Shirane M., Inomata A., Inoue T., Shishido N., Horii I., Loh D. Y. Mice lacking p27(Kip1) display increased body size, multiple organ hyperplasia, retinal dysplasia, and pituitary tumors. Cell, 85: 707-720, 1996.[CrossRef][Medline]
19. Latres E., Malumbres M., Sotillo R., Martin J., Ortega S., Martin-Caballero J., Flores J. M., Cordon-Cardo C., Barbacid M. Limited overlapping roles of P15(INK4b) and P18(INK4c) cell cycle inhibitors in proliferation and tumorigenesis. EMBO J., 19: 3496-3506, 2000.[CrossRef][Medline]
20. Mulder K. M., Morris S. L. Activation of p21ras by transforming growth factor ß in epithelial cells. J. Biol. Chem., 267: 5029-5031, 1992.[Abstract/Free Full Text]
21. Hartsough M. T., Mulder K. M. Transforming growth factor-ß signaling in epithelial cells. Pharmacol. Ther., 75: 21-41, 1997.[CrossRef][Medline]
22. Macias-Silva M., Abdollah S., Hoodless P. A., Pirone R., Attisano L., Wrana J. L. MADR2 is a subst