(Received for publication, October 18, 1996, and in revised form, December 10, 1996)
From the Cardiovascular Research Institute,
Departments of § Pediatrics and
Medicine, University
of California, San Francisco, California 94143
The Schizosaccharomyces pombe cdc5 gene product is a cell cycle regulator that exerts its effects at the G2/M transition in fission yeast. We describe the cloning of a putative human transcription factor, pombe Cdc5-related protein (PCDC5RP), which bears significant homology to S. pombe Cdc5 and to expressed sequences in mouse, nematode, and budding yeast. PCDC5RP is expressed widely in normal adult human tissues and thus may have an important general function that has been preserved evolutionarily. PCDC5RP contains two tandem repeats of a helix-turn-helix DNA binding motif, four consensus nuclear localization signals, and a hydrophilic, proline-rich central region similar to the transcriptional activating domain in Myb family members. Remarkably, PCDC5RP moved rapidly from cytoplasm to nucleus upon serum stimulation of cultured cells. This movement correlated temporally with an increase in PCDC5RP phosphorylation. Thus, PCDC5RP is a presumed transcription factor that appears to transduce cytoplasmic signals to the nucleus upon serum stimulation.
The Schizosaccharomyces pombe cdc5 gene product is required for entry into mitosis (1, 2). Haploid yeast bearing a temperature-sensitive mutation in the cdc5 gene arrest with a diploid complement of DNA, single nucleus, and decondensed chromosomes without evidence of mitotic arrest or defective DNA replication (1, 2). In addition, S. pombe Cdc5 shares sequence similarity with the proto-oncogenic transcription factor, c-Myb (2). Thus, S. pombe Cdc5 appears to regulate entry into mitosis at the level of gene transcription.
We report the cloning and characterization of a cDNA encoding a novel human phosphoprotein with significant homology to S. pombe Cdc5 (2). This protein, designated PCDC5RP,1 contains two tandem repeats of a helix-turn-helix DNA binding domain similar to that seen in c-Myb and Myb-related proteins and a central, proline-rich, hydrophilic region that may confer transactivating ability. Widespread expression of PCDC5RP in normal human tissues suggests a general function. Homology with expressed sequence tags in Mus musculus, Caenorhabditis elegans, and Saccharomyces cerevisiae demonstrates that PCDC5RP is a member of a protein family that has been conserved throughout evolution. Remarkably, PCDC5RP translocates from the cytoplasm to the nucleus of mammalian cells in response to stimulation with serum. These findings suggest that PCDC5RP may provide a novel pathway from the cytoplasm to the nucleus in mitogen-activated signal transduction.
A novel 1.45-kb cDNA, clone 67, was isolated during a yeast two-hybrid screen of a HeLa cDNA library in pGADGH (3) (provided by R. Derynck, University of California San Francisco) with cytoplasmic domains (amino acids 775-799, 1094-1115, and 1274-1512) of the human thrombin receptor (4) in the GAL4 binding domain vector, pAS1-CYH (5) (provided by S. Elledge, Baylor College of Medicine). This was sequenced (6) and used to screen a HeLa cDNA library in Uni-ZAP XR (Stratagene) for full-length clones. The 1.45-kb cDNA insert from clone 67 was labeled using the ECL enhanced chemiluminescence system (Amersham Corp.) according to the protocol provided, and plaques were screened using standard techniques (7) as modified in the ECL protocol. Eleven Uni-ZAP XR clones were excised in vivo into SK phagemid according to the protocol from Stratagene. The 2.8-kb insert from phagemid clone 67.1 was sequenced (6), and identity with pGADGH clone 67 was established. Clone 67.1 was modified after ultimate codon 802 to add an epitope (DYKDDDK) recognized by monoclonal antibody M2 (Kodak/IBI) using site-directed mutagenesis (8). The epitope-tagged insert was then excised from SK at NotI/ApaI restriction sites and subcloned into these same sites in pcDNA3 (Invitrogen) (7). This clone, pcD67F, was used in all further studies, as described below.
Related sequences were identified using the basic local alignment search tool (9) and sequence data bases available from the National Center for Biotechnology Information. Alignments and Pustell dot matrix homology analyses were performed using MacVector sequence analysis software (Oxford).
Northern AnalysisThe human multiple tissue Northern blot
(Clontech) was prehybridized in 750 mM sodium chloride, 50 mM sodium phosphate, pH 7.4, 5 mM EDTA,
0.2% Ficoll, 0.2% polyvinylpyrrolidone, 0.2% bovine serum albumin,
50% deionized formamide, 2% sodium dodecyl sulfate, 100 µg/ml
sheared salmon sperm DNA (Sigma) at 42 °C for 4 h, then hybridized in the same solution containing 1.2 × 106
cpm/ml labeled probe at 42 °C for 24 h. Following
hybridization, the blot was rinsed three times in 300 mM
sodium chloride, 30 mM sodium citrate, pH 7, 0.05% sodium
dodecyl sulfate at room temperature, then three times in the same
solution at room temperature for 10 min, then two times in 15 mM sodium chloride, 1.5 mM sodium citrate, pH
7, 0.1% sodium dodecyl sulfate at 50 °C for 20 min. The washed
membrane was exposed to x-ray film for 24 h with one intensifying
screen at 70 °C. The NotI/ApaI fragment of
pcD67F or a human
-actin cDNA control (Clontech) was labeled to
a specific activity of 3 × 108 cpm/µg using the
Prime-It II random primer labeling kit (Stratagene) according to the
manufacturer's instructions. The blot was stripped between
hybridizations by washing two times in 0.5% sodium dodecyl sulfate at 100 °C for 10 min.
COS-7 and CV-1 cell lines were maintained in DMEM H-16 medium with 3 g/liter glucose (Life Technologies, Inc.) and 10% bovine calf serum (Life Technologies, Inc.). Transfections were performed using LipofectAMINE (Life Technologies, Inc.) according to the protocol provided by the manufacturer. Transiently transfected cells were allowed to recover in DMEM with serum for 12-18 h, then incubated an additional 18-24 h in serum-free DMEM containing 0.1% bovine serum albumin. Serum-deprived cells then were stimulated with DMEM with or without 10% bovine calf serum or 0.1% bovine serum albumin.
Western AnalysisCell monolayers were placed on ice and rinsed with cold phosphate-buffered saline with protease/phosphatase inhibitors (100 µg/ml phenylmethylsulfonyl fluoride, 2 µg/ml aprotinin, 1 µg/ml leupeptin, 1 µg/ml soybean trypsin inhibitor, 1 µg/ml pepstatin A, 10 µg/ml benzamidine, 20 µM okadaic acid, 200 µM sodium orthovanadate, 10 mM sodium pyrophosphate). Cells were then scraped into the rinsing buffer and pelleted at 1,500 × g at 4 °C. Pellets were resuspended in an equal volume of 2 × gel loading buffer (100 mM Tris, pH 6.8, 4% sodium dodecyl sulfate, 20% glycerol, 200 mM dithiothreitol, 0.2% bromphenol blue) and sheared by passage through 21- and 27-gauge syringe needles. Samples were analyzed by electrophoresis in 8% polyacrylamide and transferred to nitrocellulose according to standard protocols (7).
Membranes were blocked in TNT (50 mM Tris, pH 7.4, 0.5% sodium chloride, 0.05% Tween 20) with 5% non-fat dried milk at room temperature for 30 min, incubated in the same solution with 10 µg/ml M2 antibody at 4 °C for 18 h, washed in TNT, incubated with horseradish peroxidase-conjugated goat anti-mouse antibody (Bio-Rad) at 1:20,000 dilution in TNT, 5% non-fat dried milk at room temperature for 30 min, and washed in TNT. Blots were rinsed in 50 mM Tris, pH 7.4, 0.5% sodium chloride, then developed using ECL according to the manufacturer's instructions.
Immunofluorescence MicroscopyCOS-7 or CV-1 cells were plated on sterile 4.8-cm2 glass coverslips and transfected as described above. Following transfection and treatment with serum or albumin, cells were fixed at room temperature in 4% paraformaldehyde in phosphate-buffered saline. Coverslips were then rinsed with blocking solution (150 mM sodium acetate, pH 7, 0.1% non-fat dried milk in phosphate-buffered saline), permeabilized with 0.5% Triton X-100 in blocking solution, rinsed with washing solution (15 mM sodium acetate, pH 7, 0.1% non-fat dried milk in phosphate-buffered saline), incubated with 10 µg/ml M2 antibody in washing solution, rinsed with washing solution, incubated with 2 µg/ml fluorescein isothiocyanate-conjugated goat anti-mouse antibody (Life Technologies, Inc.) in washing solution, and rinsed with phosphate-buffered saline, all at room temperature. Coverslips were mounted on slides using Slow Fade Light Anti-fade (Molecular Probes) and examined using a Nikon Microphot-FXA fluorescence microscope.
Labeling of Cells and Immunoprecipitation of PCDC5RPCOS-7
cells were transiently transfected with pcD67F, serum-deprived, and
labeled with [32P]orthophosphoric acid (DuPont NEN)
according to standard methods (10). Following serum stimulation with
phosphate-free DMEM (Life Technologies, Inc.) and 10% dialyzed fetal
calf serum (Life Technologies, Inc.), cells were placed on ice, scraped
into cold phosphate-buffered saline with protease/phosphatase
inhibitors, and pelleted at 1,500 × g at 4 °C. Cell
pellets were resuspended in RIPA buffer (10 mM Tris, pH
7.4, 500 mM NaCl, 1 mM EDTA, 1% Triton X-100,
1% sodium deoxycholate, 0.5% sodium dodecyl sulfate) with
protease/phosphatase inhibitors and sheared by passage through 21- and
27-gauge syringe needles. Cell lysates were extracted for 60 min at
4 °C and then centrifuged at 15,000 × g for 15 min
at 4 °C. Supernatants were precleared with protein A-Sepharose
(Pharmacia Biotech Inc.) for 60 min at 4 °C and then incubated with
10 µg/ml M2 antibody for 3 h at 4 °C. Protein A-Sepharose was
added, and incubations were continued for 60 min. Immunoprecipitates
were washed with RIPA buffer and then suspended in gel loading buffer
and boiled prior to electrophoresis in 8% polyacrylamide. Dried gels
were exposed to x-ray film with two intensifying screens for 4-72 h at
70 °C.
Densitometric analysis was accomplished by scanning representative autoradiographs into Photoshop (Adobe), then measuring signal densities using NIH Image (NIH). Signal densities were normalized against chemiluminescence signals obtained from Western blots of similarly treated, unlabeled samples.
PCDC5RP was identified as an apparently false positive in a yeast two-hybrid screen intended to look for novel proteins that associate with the cytoplasmic domains of the cloned human thrombin receptor (4). Although subsequent studies failed to demonstrate either association or functional coupling of PCDC5RP and the thrombin receptor in mammalian cells, PCDC5RP's ubiquitous expression and relatedness to a cell cycle control element in fission yeast (as described below) prompted further study.
The 1.45-kb partial cDNA identified during the yeast two-hybrid
screen contained an open reading frame encoding a potential DNA binding
domain. This cDNA was used to isolate a 2.85-kb cDNA from a
HeLa cell cDNA library which contained a complete open reading
frame encoding an 802-amino acid protein (Fig.
1A). The protein sequence contained two
tandem repeats of a helix-turn-helix DNA binding motif (11), four
consensus nuclear localization signals (12), a proline-rich,
hydrophilic region similar to known activating domains (13, 14), and
potential sites for phosphorylation by intracellular kinases (Fig. 1,
A and B).
A search for related cDNA sequences revealed significant homology
with the cdc5 gene product in S. pombe (2). The
two proteins were 75% identical over the 223 amino acids comprising
the DNA binding and nuclear localization domains and 17% identical
over the subsequent 535 amino acids (Fig.
2A). Amino acids 4-56 and 57-107 of PCDC5RP
represent two tandem repeats of a helix-turn-helix motif similar to the
DNA binding domain of Myb-related proteins (Fig. 2B). This
region was 83% identical to the analogous region (amino acids 2-103)
of S. pombe Cdc5 (2) and 36-38% identical to the
corresponding domains in the human Myb subfamilies (15, 16). Carboxyl
to this region was a hydrophilic stretch of 223 residues (amino acids
378-500), with some homology to the activating domains of a- and b-Myb
(15, 16) and S. pombe Cdc5 (2) (Fig. 2C). This
analysis suggests that PCDC5RP and S. pombe Cdc5 are highly
related proteins, with significant similarity to the Myb family.
A search of expressed sequences revealed that homologous open reading frames are present in M. musculus (GenBank Accession no. W82296[GenBank]), C. elegans (17), and S. cerevisiae (GenBank Accession no. Z49809[GenBank]) (Fig. 2A). The murine expressed sequence tag aligns with the carboxyl terminus of PCDC5RP and is 97% identical over the 127 amino acids available, suggesting that this is almost certainly the murine homolog of PCDC5RP. The 528-amino acid polypeptide encoded by a C. elegans gene is 80% identical over the first 205 residues, including the DNA binding domain and nuclear localization signals of PCDC5RP, and 30% identical over the remaining 323 amino acids. An open reading frame in S. cerevisiae reveals 54% identity over the first 230 residues and 14% identity over the remaining 332 amino acids. The finding of open reading frames encoding proteins with similar domain arrangements and amino acid sequences in human, mouse, nematode, and budding and fission yeast indicates that PCDC5RP is a member of a protein family that has been highly conserved throughout evolution.
Northern analysis at high stringency revealed a dominant band at 3.4 kb
in all human tissues examined (Fig. 3). In skeletal muscle, heart, pancreas, and placenta, a less prominent 8-9-kb band
was observed. How this larger mRNA species relates to the dominant
3.4-kb species remains to be determined. Of note, the wide expression
of PCDC5RP in normal adult human tissues contrasts with c-Myb, which is
expressed almost exclusively in hematopoietic cells (18, 19). These
findings suggest that PCDC5RP may serve a more general function in
transcriptional regulation.
To determine its location in mammalian cells, an epitope-tagged
version of PCDC5RP was transiently expressed in CV-1 cells. Cells grown
in standard culture medium with 10% bovine calf serum demonstrated
nuclear localization of PCDC5RP (data not shown), consistent with a
role in transcriptional regulation. Remarkably, however, PCDC5RP was
found exclusively in the cytoplasm in transfected CV-1 cells deprived
of serum (Fig. 4A). When these serum-deprived cells were then stimulated with 10% bovine calf serum for 5, 15, or 60 min, PCDC5RP was found solely in the nucleus (Fig. 4B). This
same phenomenon was observed in transiently transfected COS-7 cells
(data not shown). To the extent that the results obtained with
epitope-tagged PCDC5RP expressed in CV-1 or COS-7 cells accurately reflect the behavior of endogenous PCDC5RP, the rapid nuclear translocation of PCDC5RP in response to serum stimulation suggests a
role in relaying signals from the cell surface to the nucleus.
Many transcription factors contain phosphorylation sites that regulate
nuclear localization as well as DNA binding and transactivation (20).
PCDC5RP contains 28 potential phosphorylation sites: 14 for recognition
by casein kinase II (S/T-X-X-D/E), 9 for protein kinase C
(S/T-X-R/K), 2 for protein kinase A
(R/K-X-X-S/T), and 3 for MAP kinase (P-X-S/T-P or
P-X-X-S/T-P) (see Fig. 1). Western blot analysis of whole
cell lysates from COS-7 cells expressing epitope-tagged PCDC5RP
demonstrated a single, transfection-dependent band at
~105 kDa (Fig. 5A). This is greater than
the predicted size of 92.2 kDa, suggesting the possibility that PCDC5RP
either maintains a structure that slows its mobility on gel
electrophoresis or carries post-translational modifications.
To test the hypothesis that PCDC5RP is a phosphoprotein, immunoprecipitates of 32P-labeled COS-7 cells expressing epitope-tagged PCDC5RP were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Autoradiography revealed a transfection-dependent, 32P-labeled band of the appropriate molecular weight (Fig. 5B). The intensity of this band doubled within 10 min of serum stimulation, as quantitated by densitometry. Thus, although recombinant PCDC5RP was phosphorylated in non-stimulated COS-7 cells, the degree of phosphorylation or the amount of phosphoprotein available for immunoprecipitation increased with serum stimulation.
Precedent for phosphorylation-dependent translocation of transcription factors is well established (20). Several potential phosphorylation sites flank the four consensus nuclear localization signals in PCDC5RP (Fig. 1). In addition, these nuclear localization signals and phosphorylation sites are conserved in S. pombe Cdc5. Whether phosphorylation regulates translocation, DNA binding, or activation for these putative transcription factors remains to be determined.
The relatedness of PCDC5RP to S. pombe Cdc5 prompts speculation about its possible function. The S. pombe molecule has been implicated in the regulation of cell division, specifically at the G2/M transition (1, 2). PCDC5RP is almost certainly a serum-regulated transcription factor. Taken together with its similarity to S. pombe Cdc5 and its ubiquitous expression, this suggests a possible role for PCDC5RP in cell cycle control.
Similarities and differences between PCDC5RP and c-Myb also are worthy of comment. The DNA binding ability of the helix-turn-helix motif seen in PCDC5RP has been well characterized (11). In contrast with the DNA binding domain seen in Myb, both PCDC5RP and S. pombe Cdc5 contain only two tandem repeats of the helix-turn-helix motif, whereas Myb family members possess three. Moreover, within this domain both PCDC5RP and S. pombe Cdc5 bear a valine to leucine substitution at a position critical for DNA binding specificity (21, 22). PCDC5RP and S. pombe Cdc5, therefore, may differ from Myb in their DNA binding properties. As noted above, the expression pattern of PCDC5RP differs from that of the Myb family. These observations suggest that although PCDC5RP and S. pombe Cdc5 share similarity with the Myb family, they are likely to have distinct biological roles.
In summary, we describe a novel putative human transcription factor, PCDC5RP, with significant homology to a cell cycle regulator in fission yeast, S. pombe Cdc5, as well as to related molecules in mouse, nematode, and budding yeast. PCDC5RP contains an amino-terminal DNA binding domain related to that seen in Myb family members but with several distinguishing features. PCDC5RP is widely expressed in adult human tissues, suggesting a general function. Remarkably, PCDC5RP undergoes rapid nuclear translocation in response to stimulation with serum.
Extracellular signals are transduced from the cell surface by a variety of schemes. Many growth factor-activated signaling pathways converge on members of the MAP kinase family (23, 24). For these pathways, MAP kinase represents the final element in a cytoplasmic signaling cascade, as stimulation of fibroblasts with serum results in MAP kinase translocation to the nucleus (25-29). In the nucleus, MAP kinase phosphorylates and thereby regulates the activity of transcription factors (30). Other strategies for relaying cytoplasmic information to the nucleus include the direct activation and nuclear translocation of cytoplasmic transcription factors by phosphorylation, as is the case for signal transducer and activator of transcription (STAT) proteins (31), and the release of transcription factors from cytoplasmic retention proteins, as exemplified by the Rel family of DNA-binding proteins (32). Although the precise nuclear function of PCDC5RP remains to be elucidated, its behavior in serum-stimulated mammalian cells suggests that this presumed transcription factor may provide a novel pathway from the cytoplasm to the nucleus in mitogen-activated signal transduction.
HeLa cDNA library in pGADGH was a generous gift from Dr. Rik Derynck (UCSF). We thank Dr. David O. Morgan (UCSF) and Dr. Henry R. Bourne (UCSF) for helpful discussion and comments on the manuscript.